Long Bone Morphometrics for Human from Non-human Discrimination

Bree Saulsman (BSc, GDipForSci)

Centre for Forensic Science University of Western Australia

This thesis is presented for the degree of Master of Forensic Science of the University of Western Australia

2010

I declare that the research presented in this 48 point thesis, as part of the 96 point Master degree in Forensic Science, at the University of Western Australia, is my own work. The results of the work have not been submitted for assessment, in full or part, within any other tertiary institute, except where due acknowledgement has been made in the text.

…………………………………………………

Bree Saulsman

Abstract

Forensic anthropologists are frequently required to confirm the human origin of complete and partial skeletal remains. This determination, however, can be difficult for bone fragments with few or no distinctive morphological landmarks. Current methods of distinguishing human from non-human bone fragments include microscopic, immunological and DNA testing, which are each limited to some degree (e.g. time consuming and expensive). The purpose of this study is to investigate an alternative morphometric approach to quantify the external structure of human long bones (humeri, femora, tibiae) compared to quadrupedal (sheep, dog, pig) and bipedal (kangaroo, emu) animals commonly found in Australia.

Eight traditional linear measurements were taken on a sample of 50 human and at least 10 of each of the five non-human species; measurements were then analysed using ANOVA, canonical variates analysis (CVA) and direct discriminant analysis. The results from ANOVA and CVA indicate morphometric variation between the six species (five for the humerus), which were able to be correctly classified into their corresponding groups with a high degree of expected accuracy: humerus – 70-97.8%; femur – 70.9-97.3% and tibia – 72.4-96.2%. Direct discriminant analysis further separated the human from the combined non-human species, with a cross-validated classification accuracy of 95% or higher. More importantly, however, the technique also proved to be accurate if only a fragment of the diaphysis is analysed; classification accuracy 63 – 99%.

As it may not always be possible to accurately determine the known midshaft of a bone (especially if only a relatively small fragment is available) the potential forensic utility of a method requiring a whole bone has to be considered. In order to assess whether the power to discriminate between humans and non-humans is significantly reduced if measurements are taken up to 2cm above and below the known midshaft, a series of measurements (antero-posterior diameter, medio-lateral diameter and midshaft circumference) were taken on 17 human and 50 pooled non-human species (40 for the i

humeri). Overall, it was found that there is only a small degree of size variation within 2cm above or below the known midshaft, with relatively low average standard deviation ranges for the three human and non-human bones: humerus – 0.420-2.2.05 mm; femur – 0.345-0.586 mm; tibia – 1.034-2.676 mm. Not knowing the precise location of the midshaft on classification accuracy was also shown to have a relatively small impact, especially for the tibia, which appears to be the ‘bone of choice’ for attempting to distinguish between the human and non-human species considered in the present study.

This study has provided strong evidence that quantifying the external structure of long bones is a forensically viable technique for species discrimination. This project has also shown that morphometric methods can be successfully applied for human from nonhuman identification when remains are fragmentary and lack diagnostic muscle attachment marks or articular regions. Overall, the results of this study have outlined a forensically useful non-invasive method to distinguish human from non-human bones.

ii

Acknowledgements

This project would not have been completed without the help, support, and guidance of several people. First and foremost I would like to thank Daniel Franklin and Charles Oxnard for their supervision of this project. Their guidance and editorial comments have been instrumental to this thesis. Special thanks to Dan for his continued support and unwavering patience throughout this process, I could not have asked for a better supervisor. Thanks also to Ian Dadour and the Centre for Forensic Science for all their help throughout the last few years.

Thanks are also due to the following for kindly providing access to, and/or assistance with all animal materials used in this project: Chris from A&D Pet Foods, Kip Venn, Ron Porter, and Wally Gibb. Particular thanks to Darryl Kirk, of the School of Anatomy and Human Biology (UWA) for supplying this research with much needed human skeletal material. Many thanks to Shalmira Karkhanis for conquering her fear of maggots and all things wriggly to spend a day helping me acquire the animal limbs. I am eternally grateful for your willingness to participate in this long and arduous preparation process.

Words cannot describe how thankful I am for the unconditional love, support and encouragement my fiancée, Mike, has shown me during this whole process. I would not be where I am today without him. To the Yeats’ and my parents, I am forever grateful for all the love and support you have shown me throughout the many years of study – and there have been many! Finally, I dedicate this thesis to Mike and the much missed Elizabeth Yeats (1945 – 2007).

iii

Table of Contents

Abstract

i

Acknowledgements

iii

Table of Contents

iv

Figures

ix

Tables

xii

Chapter One:

Introduction

1.1

Introduction

1

1.2

Aims

4

1.3

Thesis Outline

5

1.4

Importance of a Comparative Collection

5

1.5

Selected Non-human Species

6

1.6

Limitations of This Study

7

Chapter Two:

2.1

The Skeletal System

Bone Biology

8

2.1.1

General Structure of Bone

8

2.1.2

Cortical Bone Thickness

11

2.1.3

Bone Composition

13

2.1.4

Bone Histology

14

2.2

Bone Biomechanics

16

2.3

The Limbs

17

2.3.1

Fore-limbs

17

2.3.2

Hind-limbs

18 iv

2.4

Bipedal Locomotion

19

2.4.1

Kangaroo

19

2.4.2

Emu

21

2.4.3

Summary

22

Chapter Three:

3.1

Introduction

23

3.1.1

Humeri

23

(i) Proximal-third

24

(ii) Middle-third

26

(iii) Distal-third

26

Femora

27

(i) Proximal-third

27

(ii) Middle-third

29

(iii) Distal-third

29

Tibiae

30

(i) Proximal-third

32

(ii) Middle-third

32

(iii) Distal-third

33

3.1.2

3.1.3

3.2

3.3

Species Morphology, Discrimination and Preparation

Current Methods for Species Discrimination

33

3.2.1

Microscopic Examination

34

3.2.2

Protein Analysis

36

3.2.3

DNA Analysis

37

3.2.4

Summary

37

Defleshing Methods

38

3.3.1

Water Maceration

38

3.3.2

Chemical Maceration

38

3.3.3

Carrion Insects

39

v

3.3.4

Boiling

Chapter Four:

4.1

4.2

4.3

40

Materials and Methods

Materials

41

4.1.1

41

Sources of Skeletal Material

Methods

42

4.2.1

Preparation of Non-human Skeletal Material

42

4.2.2

Whole Bone Data

43

4.2.3

Partial Bone Data

44

4.2.4

Measurement Error Calculations

44

Statistical Analyses

46

4.3.1

ANOVA

46

4.3.2

Discriminant (Canonical) Analysis

47

4.3.3

Direct Discriminant Functions

47

Chapter Five:

Results

5.1

Introduction

49

5.2

One-Way ANOVA

49

5.2.1

Humerus

49

5.2.2

Femur

51

5.2.3

Tibia

51

5.2.4

Summary

54

5.3

Discriminant (Canonical) Analysis

54

5.3.1

Humeri

54

(i) Stepwise Analysis

54

(ii) Shaft Measurements

57

(iii) Midshaft Only

58 vi

5.3.2

5.3.3

5.3.4 5.4

5.5

Femora

60

(i) Stepwise Analysis

60

(ii) Shaft Measurements

62

(iii) Midshaft Only

63

Tibiae

65

(i) Stepwise Analysis

65

(ii) Shaft Measurements

67

(iii) Midshaft Only

69

Summary

70

Direct Discriminant Functions

71

5.4.1. Humeri

71

5.4.2

Femora

72

5.4.3

Tibiae

73

5.4.4

Summary

74

Morphometric Variation around the Known Midshaft

75

5.5.1

Standard Deviation

75

5.5.2

Classification Accuracies

76

5.5.3

Summary

78

Chapter Six:

Discussion and Conclusions

6.1

Introduction

80

6.2

Statistical Analysis

80

6.3

Morphometric Variation: Structure and Function

81

6.4

Forensic Applications

84

6.5

Future Research

86

6.6

Concluding Summary

87

vii

References

88

Appendix I: Long Bone Measurements

98

Appendix II: Measurement Error Calculations

101

Appendix III: Results of Primary Data

102

Appendix IV: Basic Statistics

109

Appendix V: Midshaft Measurements

115

Appendix VI: Group Classification using Midshaft Measurements

121

viii

Figures Page Figure 1.1

2

Intact femoral bones

Figure 1.2

3

Middle-third of femoral shafts

Figure 2.1 Trabecular bone structure

9

Figure 2.2 Femur with all gross anatomical features labelled

10

Figure 2.3 Differences in cortical bone thickness using mean raw measurements

12

Figure 2.4

15

Microscopic structure of a long bone showing the arrangement of lamellae in the Haversian systems

Figure 2.5

18

Comparison of front limb structures of a dog, pig, cow and horse

Figure 2.6

19

Comparison of back limb structures of a dog, pig, cow and horse

Figure 2.7

19

Skeleton of a kangaroo ix

Figure 2.8

22

Emu lower limb

Figure 3.1

25

Anterior and posterior views of human, kangaroo, sheep, pig and dog right humeri

Figure 3.2

28

Anterior and posterior views of human, kanagaroo, sheep, pig, dog and emu right femora

Figure 3.3

31

Anterior and posterior views of human, kangaroo, sheep, pig, dog and emu right tibiae

Figure 3.4

34

Histological section showing osteons, Haversian canals, Volkmann’s canal and interstitial lamellae

Figure 3.5

35

Histological section of a deer humerus showing plexiform bone

Figure 5.1

55

CVA plot of individual scores from the stepwise analysis of seven humeri measurements

Figure 5.2

58

CVA plot of individual scores from the three humeri shaft measurements

Figure 5.3

59

CVA plot of individual scores from the two humeri midshaft measurements

x

Figure 5.4

61

CVA plot of individual scores from the stepwise analysis of seven femora measurements

Figure 5.5

63

CVA plot of individual scores from the four femora shaft measurements

Figure 5.6

64

CVA plot of individual scores from the three femora midshaft measurements

Figure 5.7

66

CVA plot of individual scores from the stepwise analysis of seven tibiae measurements

Figure 5.8

68

CVA plot of individual scores from the four tibiae shaft measurements

Figure 5.9

70

CVA plot of individual scores from the three tibiae midshaft measurements

Figure 5.10

82

CVA plots of the three quadrupedal species for the humerus, femur and tibia samples

xi

Tables Page Table 5.1

50

ANOVA comparison of sample means of the eight humeri measurements for the five species

Table 5.2

52

ANOV comparison of sample means of the eight femora measurements for the six species

Table 5.3

53

ANOVA comparison of sample means of the eight tibiae measurements for the six species

Table 5.4

55

Mean scores, canonical roots and variance of the five species using a stepwise analysis of seven humeri measurements

Table 5.5

56

Standardised coefficients from the stepwise analysis of seven humeri measurements

Table 5.6

57

Mean scores, canonical roots and variance of the five groups computed from three humeri shaft measurements

Table 5.7

59

Mean scores, canonical roots and variance of the five groups computed from two humeri midshaft measurements

xii

Table 5.8

60

Mean scores, canonical roots and variance of the six groups using a stepwise analysis of seven femora measurements

Table 5.9

62

Standardised coefficients from the stepwise analysis of seven femora measurements

Table 5.10

63

Mean scores, canonical roots and variance of the six groups computed from four femora shaft measurements

Table 5.11

64

Mean scores, canonical roots and variance of the six groups computed from three femora midshaft measurements

Table 5.12

65

Mean scores, canonical roots and variance of the six groups using a stepwise analysis of seven tibiae measurements

Table 5.13

67

Standardised coefficients from the stepwise analysis of seven tibiae measurements

Table 5.14

68

Mean scores, canonical roots and variance of the six groups computed from four tibiae shaft measurements

Table 5.15

69

Mean scores, canonical roots and variance of the six groups computed from three tibiae midshaft measurements

xiii

Table 5.16

71

Summary of cross-validated classification accuracies for the humerus, femur and tibia

Table 5.17

72

Discriminant functions computed from eight humeri measurements Table 5.18

73

Discriminant functions computed from eight femora measurements

Table 5.19

74

Discriminant functions computed from eight tibiae measurements

Table 5.20

76

Standard deviations calculated for the mean of the measurements taken at and around (1cm and 2cm above and below) the known midshaft

Table 5.21

77

Overall classification accuracies for the mean and maximum values above and below the known midshaft

xiv

CHAPTER ONE

INTRODUCTION

1.1

Introduction

Forensic anthropologists are often required to examine skeletal remains to answer questions of identification and circumstances surrounding death. This involves the application of physical anthropology to the medicolegal system. Of all the subdisciplines within anthropology, physical anthropology has had the closest relationship with the law, as it occupies the study of ‘humanity’s similarity to and divergence from other animals’ (Ferllini, 2002). As a result, forensic anthropologists require a thorough understanding of both human and non-human osteology.

Analysis of faunal (non-human) remains recovered from archaeological sites have had an important role in archaeological interpretation, as these bones can be used to study past hunting practices, animal husbandry patterns and diet. In modern forensics, recovered bones are frequently referred to an investigating authority, with the role falling to the forensic anthropologist to confirm or exclude their human origin. It is therefore just as important for forensic investigators to be able to differentiate between human and non-human bones, as misidentification can lead to misuse of resources and expenses in police investigations.

Complete, or substantial parts of long bones containing diagnostic features, can usually be readily distinguished as human or non-human (Figure 1.1). However, even a specialist may find this distinction difficult when given an isolated diaphysis (or a portion thereof) with missing articular ends (Figure 1.2), especially in the event of postmortem alterations (Ubelaker, 1989). Carnivores often target the ‘softer’ cancellous portions of the articular regions of long bones (Haglund et al., 1988; Reitz, 2008); rodents can also cause damage by gnawing at similar regions, especially the femoral and 1

tibial condyles (Krogman and Iscan, 1985; Klippel, 2007). It may, therefore, be difficult for an anthropologist to quickly assess whether a bone is human or non-human due to the destruction of these diagnostic morphological features.

Figure 1.1. Intact femoral bones: left to right; human, kangaroo, sheep, pig, dog and emu.

2

Figure 1.2. Middle-third of femoral shafts: left to right; human, kangaroo, sheep, pig, dog and emu.

Microscopic appearances of cortical bone may be diagnostic of non-human species, as the presence of plexiform bone, or a characteristic osteon banding pattern, may rule out human origin (Mulhern and Ubelaker, 2001). However, differences in bone histology throughout the skeleton, as well as among different species, are not fully documented (Hillier and Bell, 2007). Alternative methods also include protein radioimmunoassay (pRIA) and DNA analysis. Ubelaker et al. (2004) recently developed a technique using radioimmunoassay to quantitatively separate human from non-human bones, which has some degree of advantage over DNA analysis, as proteins are able to withstand most environmental influences and are thus less prone to contamination (Cattaneo, 2006). However, even with the advent of immunological, genetic and histological approaches that potentially offer identification from fragmentary evidence, they are not without their own specific limitations, especially in terms of cost and time. Undoubtedly there exists a need for an effective methodology, whereby the separation of ‘human vs. nonhuman’ bones can be achieved rapidly, accurately and cost effectively.

3

Generally forensic anthropologists rely upon both their own intimate knowledge of human anatomy and individual experience, coupled with the use of comparative reference collections of vertebrates specific to their geographic region. Comparative sources, both for domestic (e.g. Sisson and Grossman, 1953; May, 1970) and nondomestic (e.g. Schmid, 1972; Gilbert, 1973; Cornwall, 1974) species are available, but these are generally cursory and cover almost exclusively species not found in Australia. Amongst the few comparative sources on indigenous Australian animals are Merrilees and Porter (1979) and Oxenham and Barwick (2008).

Although morphoscopic or visual assessment is useful when bones are complete, this alone may be inadequate when materials are damaged or lack the diagnostic muscle attachment marks or articular regions. In such cases, alternative techniques are required. The purpose of this research, therefore, is to determine whether alternative quantitative morphometric methods are useful for human from non-human bone identification. The test species are human long bones (humeri, femora, tibiae) in comparison with the same bones of quadrupedal (sheep, dog, pig) and bipedal (kangaroo, emu) animals.

1.2

Aims

This study aims to establish diagnostic criteria for the comparison of human bones with animal remains native to Australia. The specific aims of this project are:

(i)

Determine if human bones can be differentiated from selected non-human species through statistical analyses of linear measurements.

The primary aim of this study is to investigate whether non-invasive standard morphometric methods can be applied when distinctive morphological markers are absent. Though this project will consider measurements across the length of the whole long bone, attention – both qualitative and quantitative – will ultimately be given to those that solely include the diaphysis, or specifically the middle third.

4

(ii)

Develop an Australian reference collection for the rapid and accurate identification of an unknown specimen.

The long-term goal of this study is to create an anthology of known species native to Australia, whereby a forensic scientist can match an unknown bone to one in the reference collection. This will allow an identification key to be developed that provides a quick and easy distinction between several animal species.

1.3

Thesis Outline

Relevant human and non-human skeletal systems are reviewed in Chapter Two, from bone tissue level through to total limb assemblages. This chapter also discusses the fundamental principles of biomechanics and their intrinsic influences on bone form. Chapter Three focuses on the osteology of the individual long bones and the availability of recognisable features found within each third. Also in this chapter is an evaluation of the current methods available for species discrimination, including a review on techniques for defleshing. Descriptions of skeletal preparation and analytical methods are described in Chapter Four. The results of the morphological analyses are outlined in Chapter Five. The results are discussed in Chapter Six, the aims outlined in Chapter One are addressed, and suggestions for future research proposed.

1.4

Importance of a Comparative Collection

Differentiating human from non-human bones depends upon individual training and the quality of the reference material. For that reason, access to a large and diverse reference collection is an important tool for a forensic anthropologist. While there are texts dedicated to human and non-human osteology, there are limited selections that combine the two together, with the exception of the most recent comparative texts by Adams and Crabtree (2008) and France (2009), which explore both domestic and non-domestic species. Earlier works, such as Schmid’s (1972) ‘Atlas of Animal Bones,’ and Gilbert’s (1973) ‘Mammalian Osteo-Archaeology: North America,’ are also available with distribution maps for each species. The paucity of published material is illustrated by 5

most case studies and text books concentrating on limited problems such as the similarities between the metacarpals and metatarsals of the human hand and the bear paw (Stewart, 1979), which is redundant in Australia.

To fully quantify the range of normal biological variation requires large samples. This can be inexpensively obtained from road-kills and abattoirs; the latter requiring care to avoid tool damage and/or confusion on provenance of limbs between different animals. Such materials must also be defleshed; which is labour intensive and time consuming. These limiting factors may explain the lack of comparative material available for the anthropologist.

Whilst high zoological expertise is not expected, a well trained anthropologist needs general familiarity with skeletal characteristics of the major non-human species within their region, as a properly established collection will serve as a guide with which recovered remains may be compared. The first step in compiling a comparative collection is to obtain samples of locally significant species. The most valuable collections will have several specimens for each species, illustrating sex, age, and species variation.

1.5

Selected Non-human Species

The five non-human species studied in the present project were chosen specifically, as not only are they common to Western Australia, but also because of their anatomical similarity to humans (especially in the absence of articular joints) and thus their potential to be mistakenly identified. These include feral pig (Sus scrofa), sheep (Ovis aries), dog (Canis familiaris), kangaroo (Macropus fuliginosus) and emu (Dromaius novaehollandiae).

6

1.6

Limitations of This Study

Time constraints have made it necessary to make a number of restrictions in the present project. First, comparisons are limited to adult bones; subadults are not considered, because there is limited information regarding epiphyseal fusion in many animals both domestic and wild. Second, sexual dimorphism is also not considered, as the primary interest of this study was to elucidate potential human to non-human species level variation as a whole.

7

CHAPTER TWO

THE SKELETAL SYSTEM

2.1

Bone Biology

The term ‘bone’ can refer to an individual element of the skeleton (such as the humerus), or the material from which those elements are built, both of which are described in more detail below. The axial skeleton encases (and consequently protects) the vital organs of the head and chest; the appendicular skeleton is mainly involved in posture and movement.

2.1.1 General Structure of Bone

The bones of the skeleton differ in their form, internal construction and function. They can be cylindrical (long bones of the limbs), flat (such as some skull elements, sternum and the scapulae) or irregular (such as vertebrae and innominates). The elongated cylindrical bones are confined to the limbs; their mechanical function lies in supporting the body and providing a leverage system for the muscles and tendons (Frazer, 1946). The diaphyses of these bones are generally tubular in shape with expanded proximal and distal ends (epiphyses), and an internal marrow cavity. In certain places, the tubular shape may be distorted where the bone is drawn out into flanges and tubercles for mechanical strength and muscle and ligament attachments. Almost all long bones are curved so as to more evenly distribute impressed load bearing stresses (Currey, 2002).

In longitudinal section adult long bones have two basic structural components. The solid, dense bone found in the walls of the diaphysis is the compact or cortical bone. Cortical bone only contains the microscopic spaces of the bone cell system and forms the external layer of all bones. The outer wall of cortical bone varies in thickness, its maximum in a long bone being near the midshaft, and then tapering off towards the 8

metaphyses. It provides protection and support, whilst resisting compressive, tensile and bending stresses produced by weight and movement (Currey, 2002). The second component known as trabecular or cancellous bone has a spongy appearance and is typically found at the distal ends of the long bones (Figure 2.1). The trabeculae in spongy bone tissue helps resist stresses and transfer loads in more complexly shaped regions (Martin et al., 1998; Muscolino, 2006). The internal anatomy of a typical long bone (femur) is shown in Figure 2.2.

Figure 2.1. Image of trabecular bone structure in the distal end of a human femur. (Adapted from Martin et al. 1998:33).

9

Figure 2.2. Adult femur with all gross anatomical features labelled. (Adapted from Muscolino 2006:44).

A thin tissue (periosteum) surrounds the outside of the diaphysis during life and coats all bone surfaces, except for the articular surfaces. The articular regions are not covered by this tough sheath of dense irregular connective tissue, but instead by a thin layer of cartilage, which provides a frictionless surface for articulation (Davis, 1987). The periosteum contains bone forming cells that enable bone to grow in thickness, but not in length. It also protects the bone, assists in fracture repair, helps nourish the bone tissue, and serves as attachments for some ligaments and tendons. The medullary cavity is also lined by a membrane called the endosteum that contains bone forming cells (Frazer, 1946; Muscolino, 2006 and see Figure 2.2). 10

2.1.2 Cortical Bone Thickness

Although the primary focus of this present study is on the macroscopic features of long bone for human from non-human discrimination, further separation between species may also be possible through quantifying the cortical thickness of long bone diaphyses (especially as accompanied evidence to fragmentary remains). Previous research suggests that long bone cortical thickness varies considerably between human and certain non-human species, with adult animals of comparable size to humans having thicker and more compact cortices (Brothwell, 1981; Wolf, 1986; Ubelaker, 1989). It has also been suggested that the cortex in an adult human is roughly one-quarter the total thickness of bone, with larger non-human mammals (such as dog, bear, sheep, etc.) containing thicker cortices at one-third of the total diameter (Wolf, 1986; Thomas, 1995). However, these features are only documented as providing cursory evidence for human from non-human species discrimination, with little or no description on how these figures were obtained.

Preliminary studies have recently been conducted by Croker et al. (2009) investigating the forensic efficacy of this commonly accepted yet sparsely published method for species discrimination. Cross-sections of the femoral midshafts of humans, kangaroos and sheep were obtained to calculate the cortical thickness index by means of the whole shaft and medullary cavity diameters on each bone. Statistically significant differences were found between the three groups with the human femur exhibiting the largest cortical thickness index and the sheep the smallest (Figure 2.3). However, more extensive investigations are needed to determine if these differences are not limited to just the midshaft by investigating multiple sites on other skeletal elements, in addition to other non-human species. Furthermore, the effects of age and sexual dimorphism would also be of significant importance for future studies into the use of cortical bone thickness for species determination.

11

Figure 2.3. Differences in cortical bone thickness using mean raw measurements for (a) human, (b) sheep and (c) kangaroo. (Adapted from Croker et al. 2009).

Long bone diaphyseal cross-sectional properties have also been utilised when addressing issues associated with functional adaptation (i.e. effects of locomotion, body mass, skeletal robusticity and changes during growth and development) on both modern and archaeological samples (e.g. Martin and Armelagos, 1979; Bloom, 1980; Ruff and Hayes, 1983a; Ruff and Hayes, 1983b; Demes et al., 1991; Runestead et al., 1993; McLorg, 2000; Ruff, 2000; Wescott, 2006; Stock and Shaw, 2007). This type of research will thus rely on biomechanical principles, following the approach that crosssectional dimensions of long bones are largely influenced by mechanical loading, possibly resulting in cortical thickness variation between human and selected nonhuman species.

Although Croker et al. (2009) suggest that the differences in modes of locomotion (bipeds versus quadrupeds) between the human and two non-human species may be a contributing factor towards the varying thicknesses observed in cortical bone, their investigation is limited only to kangaroos and sheep. Furthering Croker’s et al. (2009) original work, cross-sectional properties of the same bones used in the present study with species displaying to both bipedal (human, kangaroo and emu) and quadrupedal (sheep, dog and pig) locomotion, could potentially offer more effective comparisons through a wider range of test species, with cortical thickness variation also anticipated between the bipeds due to the specialised locomotory functions unique to each species (walking, hopping and running).

12

Cortical thickness can be quantified in a variety of ways, although ultimate choice of methodology will be dependent on the availability of resources (such as time, funding, availability of equipment and expertise etc.). Croker et al. (2009) have chosen to express cortical width using a derived formula calculating the cortical thickness index as a percentage of total shaft diameter. However, incorporating the use of cross-sectional geometry on long bones may more effectively reflect the effect of biomechanical principles on the amount of cortical thickness present between species. Because of their biomechanical significance, cross sectional geometric properties of long bone diaphysis (cortical area) have been increasingly used in a number of form/function studies, particularly for reconstructing mechanical loading history of long bone shafts (Ruff, 1984; Demes et al., 1991; O’Neill and Ruff, 2004; Wescott, 2006). Cortical area measures the bone’s internal resistance to axial loads, allowing for effective examination of the entire cross-sectional area and not just an average of parts (Martin and Armelagos, 1979; Ruff and Hayes, 1983a). For an overall analysis of crosssectional properties, additional measurements representing both internal and external dimensions could also be documented. These could include documenting ap and ml widths of bone shafts, calculating medullary area, as well as second moments of area (which represent the bending strength of a bone in a particular axis) allowing for a more accurate indication of cortical properties (Ruff and Hayes, 1983a; Demes, 1991; Runestead et al., 1993; Ruff, 2000; Stock and Shaw, 2007).

2.1.3 Bone Composition

Bone can be considered a two- or multi-phase material. The inorganic component is mineral hydroxyapatite, which is a crystalline form of calcium phosphate, and consists approximately 65-70% of the dry weight of bone, has the engineering properties of a ceramic, very strong in compression but weak in tension, and provides the rigidity and compressive strength. The organic content (30-35%) consists of collagen and other organic substances, forming long fibres that are elastic. These structures are very strong in tension but weak in compression. It is the particular combination of the two that gives bone it characteristic properties. The nature of the combination is that the crystals are firmly bonded to the fibres to form a two phase material (Davis, 1987; Martin et al., 1998). Without the hydroxyapatite, bone would be flexible like rubber (a decalcified 13

bone can be tied in a knot), and without collagen it would be brittle like glass or china (and thus easily fracture under tension like a cup handle). It is this particular combination as a two-phase material (a mechanical analogue is fibre-glass) that gives bone its remarkable properties for bearing stresses and strains. Other constituents of bone include water, proteins, polysaccharides, living cells, blood vessels, lymphatics and nerves (Frazer, 1946; Currey, 2002).

2.1.4 Bone Histology

As bone is a living substance, blood vessels are needed throughout to keep the cells (which lie deep within the bones) supplied with oxygen and nutrients; microscopic sections show that bone is organised around a network of capillaries. Both the compact and trabecular components of adult bones are made of lamellar bone tissue, which is uniformly laid down repeatedly on bone surfaces during appositional growth (Hall, 2005).

Compact bone also consists of closely packed secondary osteons or Haversian systems (Figure 2.4), necessary because nutrients cannot be diffused from surface blood vessels due to its dense bone composition (Frazer, 1946). These Haversian systems measure about 200-300 microns in diameter and are about 3-5 mm in length. Each system consists of a central vascular canal called the Haversian canal, through which blood, and sometimes nerve fibers, pass parallel to the long axis of the bone. Volkmann’s canals connect the blood vessel from one Haversian canal to the blood vessel of an adjacent Haversian canal thus creating a network that supplies all nutrients to the cells of long bones (Martin et al., 1998).

14

Figure 2.4. Microscopic structure of a long bone showing the arrangement of lamellae in the Haversian systems. (Adapted from Davis 1987:53).

Each Haversian canal is surrounded by characteristic concentric cylinders of bone (lamellae), and is thus what sets Haversian systems, or secondary osteons, apart from primary osteons. Haversian systems are surrounded by a cement sheath, whereas primary osteons are not (Currey, 2002). Secondary osteons are formed as part of a remodelling process and replace primary bone in which osteoclasts (cells which remove bone) digest long tunnels through the matrix. Each tunnel is then filled by the formation of concentric cylinders of bone secreted by osteoblasts (bone-forming cells responsible for synthesising and depositing bone material) in association with the blood vessels (Davis, 1987; Mulhern and Ubelaker, 2001). Replacement of bone tissue takes place throughout life, with several succeeding generations of Haversian systems being observed in specimens of adult bones under the microscope. It is this continuing internal remodelling of bone through the removal of bone tissue by osteoclasts and the building of new ones by osteoblasts, which allows modification of bone shape and re-adaptation to new mechanical stresses (Davis, 1987).

15

In contrast, spongy or trabecular bone consists of lamellae that are arranged in an irregular lattice of thin columns of bone called trabeculae; receiving nutrition from blood vessels in the surrounding narrow spaces. Though it may seem that the trabeculae are arranged in a very unsystematic manner, they are however organised to provide maximum strength, as they are related to the lines of stress and can realign if the direction of the stress changes (Currey, 2002; Oxnard, 2008) thus adapting to new forces that may come to act on the ends of the bone (e.g. in experimental or clinical situations).

2.2

Bone Biomechanics

The bones from both human and non-human skeletons can provide insight into how they are constructed, and how they work. Although genetic mechanisms determine some aspects of the shapes of whole bones, other aspects of bone shape exist within the limits of their biological functions; also termed biomechanics (Oxnard, 1973). The term biomechanics is defined as the application of mechanical principles to biological systems. The study of biomechanics within anthropology is diverse, ranging from primate locomotion and mastication, to long bone structural analysis (Lovejoy et al., 1976). The arrangement and structure of animal long bones thus relates not only to genetic heritage, but also offers insights into posture and locomotion, attack and/or escape capabilities, as well as the effects of movement in various types of terrain (Oxnard et al., 1990; Searfoss, 1995). Long bones must be structured to withstand compressive loads, in addition to bending movements, without undergoing plastic deformation or fracture (Currey, 2002); it is thus clear that various shapes of whole bones are intimately related to their functions.

As function (therefore loadings) can vary throughout life; a bone must be able to adjust to changing demands. As a result, the amount of compact and trabecular bone are determined in part by function and load, so that the arrangements of the bone provide maximum strength with minimum weight. Due to its hard composition, bone cannot swell or shrink, and because bone may be added or removed from the periosteum or endosteum, all shape changes must occur on surfaces (Currey, 2002). Extra loading 16

results in deposition of new bone at the site of stress, whereas inactivity or conditions of weightlessness result in the resorption of bone and may lead to atrophy. Consequently, bone can adapt in response to changes in mechanical demand (Chaplin, 1971; Frankel and Nordin, 1980).

2.3

The Limbs

The entire skeleton, as well as individual elements, is shaped by function, especially mode of locomotion. Locomotion is the complex set of behaviours that primarily influence the morphology and physiology of animals (Dickinson et al., 2000). Most mammals are quadrupeds, though some, such as humans, are bipeds. In both cases, the fore-limbs attach to the trunk at their most proximal component, the pectoral girdle, and the posterior or hind-limbs connect to the trunk at the pelvic girdle. Although all mammal limbs have similar bone groupings and structural divisions, limb usage can vary greatly, with the front and back limbs of the same animal often being quite different, depending on their function (Davis, 1987).

2.3.1 Fore-limbs

Generally, the three long bones in the fore-limbs (the humerus, radius, and ulna) form the primary weight-bearing supports for the front portion of the body in quadrupeds. The front limbs tend to be used for a wider variety of functions than the hind-limbs, such as handling food, digging, climbing or flying (Figure 2.5). In hoofed animals, such as cows, horses and pigs, the front limbs are used in conjunction with the rear limbs, and perform nearly identical roles whilst walking or running (Searfoss, 1995).

17

Figure 2.5. Comparison of front limb structures of a dog, pig, cow and horse (Adapted from Searfoss 1995:100).

2.3.2 Hind-limbs

The back or hind-limbs are also usually associated with support and movement of the body (Figure 2.6). The acetabulum in the pelvic girdle is much deeper than the glenoid fossa in the scapula, allowing flexibility in the shoulders of fast running animals. The rear limbs often provide most of the thrust in locomotion and therefore generally have much less freedom of movement than the front limbs (Davis, 1987; Searfoss, 1995).

18

Figure 2.6. Comparison of back limb structures of a dog, pig, cow and horse (Adapted from Searfoss 1995:102).

2.4

Bipedal Locomotion

The animals discussed thus far in this chapter have been placental mammals displaying mostly quadrupedal locomotion. Like humans, various marsupial mammals and some large birds are uniquely bipedal in their ranges of movement, from walking to running (cursorial movement) and hopping (saltation). Long bone form and function of saltators and bipedal cursors are generally more specialised than quadrupedal cursors (Hildebrand, 1995).

2.4.1 Kangaroo

Species of Macropodoidea (a superfamily including kangaroos, wallabies and rats kangaroos) range in sizes from approximately 0.5 to 85 kg. Nearly all members of this family employ bipedal hopping as the primary mode of locomotion at high speeds (an awkward shuffling on all four legs and the tail the slowest gait for the kangaroos) and have comparable hind-limb geometry (Alexander, 1992; McGowan et al., 2008). Relative lengthening of the hind-limbs and of their distal segments is more extreme for 19

saltators than for quadrupeds (Hildebrand, 1995); such that their fore-limbs appear quite undeveloped in comparison (Figure 2.7). The much shorter fore-limbs are usually used for slow progression, balance, and manipulation (Reitz and Wing, 2008). Although all segments of the hind leg are elongated, two of them – the tibia (between upper segment and ankle) and tarsus (ankle) segments – are the most elongated. The mechanical demands of bipedal hopping in the larger macropodoid marsupials, such as the kangaroos, can be explained through their capacity to store substantial amounts of elastic energy in their long, relatively thin Achilles tendons (Bennett and Taylor, 1995; McGowan et al., 2008).

Figure 2.7. Skeleton of an adult kangaroo. (Adapted from Oxenham 2008:67).

Simultaneous and repetitive action of the hind legs is employed by kangaroos (ricochetal locomotion). This extreme specialisation of the hind limbs for propulsion has produced the seemingly unbalanced body of all members of the family. Although leaping their bulky bodies along on the natural springs of the hind feet originally 20

provides an adequate burst of speed, it would be disadvantaged over long distances, especially against quadrupedal pursuers (Troughton, 1967; Dagg, 1977).

2.4.2 Emu

Emus are also bipedal, but with a different type of locomotion to that of the kangaroo. They are cursorial birds, escaping by running and unable to fly. Associated with this are the extremely muscular upper legs, thin elongated lower legs, and reduced size and musculature of the wings (Abourachid and Renous, 2000). They have lost the ability to fly and instead have heavy compact bodies and powerful feet. Yet despite their large size, these birds are competent hind limb runners, with almost 1/3 of their body mass, being devoted to leg extensor muscles, the largest recorded for any tetrapod (Hutchinson, 2004). Emus have very short (almost horizontal) femora and very long tibiae; with additional specialisations that include a tibiotarsus (union of proximal tarsal bones with distal end of the tibia) to which the fibula is attached, and a tarsometatarsus (union of the distal tarsal bones with the proximal end of the metatarsus). This complex assemblage, quite unlike those of most mammals described previously, maximises acceleration and mechanical advantage by using both muscle power and elasticity while running (Figure 2.8) (Hutchinson, 2004).

21

Figure 2.8. Emu lower limb. (Adapted from Webster and Webster 1974:105).

2.4.3 Summary

The aforementioned biomechanical parameters, such as limb orientation, relative muscle sizes and muscle fascicle lengths, tendon lengths and elasticities, present locomotor abilities/limits which differ between animals according to their specific needs. Though kangaroos and emus are bipedal and share some similarities to humans, each also has uniquely different adaptations. Kangaroos are different to both humans and emus because they utilise a hopping rather than a striding motion, as a result of their capacity for elastic energy savings in the primary ankle extensor tendons (Hutchinson, 2004; McGowan et al., 2008). Emus are more similar to humans in that they are a habitually striding biped; they differ, however, in having different arrangements of limb segments; with a proportionately smaller size femur compensated by a corresponding increase in relative tarsometatarsus length (Gatesy and Biewener, 1991).

22

CHAPTER THREE

SPECIES MORPHOLOGY, DISCRIMINATION AND PREPARATION

3.1

Introduction

Although the primary purpose of the present study is to formulate a quantitative method for human from non-human species discrimination, a qualitative approach must also be duly considered. The first component of this chapter will thus review the gross morphology of the three long bones (humerus, femur and tibia) for the human and five non-human species (emu excluded for the humerus) to establish distinguishing landmarks and ranges of variation for each one-third: proximal, middle and distal (especially in the event of diaphyseal fragments). It is however, expected that practicing forensic anthropologists will already have the necessary expertise of anatomical features (especially as the basic features of the mammalian skeleton are remarkably constant) required to firstly establish from which skeletal element an unknown fragment belongs, in addition to locating the approximate region on the diaphysis from which the fragment may originate (i.e. proximal, middle or distal-third). Even with the lack of articular regions, the midpoint may frequently be approximated by visual estimation (i.e. the midpoint in the humerus is usually located a few millimetres below the inferior margin of the deltoid tuberosity). The second and third part of this chapter reviews the current methodologies available for species discrimination and skeletal preparation/defleshing and their advantages and disadvantages within a forensic context.

3.1.1 Humeri

A human humerus is relatively long and thin in comparison to a kangaroo, sheep, pig and dog. The shaft of a human humerus is roughly cylindrical proximally, gradually becoming triangular until it is broad and flat distally (Figure 3.1). The kangaroo is the most different macroscopically, exhibiting a distal end similar to a human, but with a 23

more robust and curved diaphysis. A dog humerus has a slight spiral twist, whereas a pig and sheep humeri are rather stout, with the former having an appearance in profile somewhat like an italic f minus the cross-bar due to the marked backward and forward inclination of the proximal and distal ends respectively (Sisson and Grossman, 1953; Oxenham, 2008).

(i)

Proximal-third

The lateral tuberosities (equivalent to the human greater tubercle) in all four animal species is easily distinguished from that of a human, as they are more pronounced and extending further cranially than the articular surface of the head. This difference is greatest in the sheep and pig, with both having very large and curved lateral tuberosities; the latter exhibits a particularly prominent neck and strongly curved humeral head (Sisson and Grossman, 1953; Oxenham, 2008). In sheep, pig and dog humeri, the deltoid tuberosity is confined to the proximal end of the shaft and is less well defined (Bone, 1975; Hillson, 1996).

24

Figure 3.1. Anterior (above) and posterior (below) views of human, kangaroo, sheep, pig and dog right humeri (right to left). 25

(ii)

Middle-third

In the lateral middle third of the human humerus is a characteristically extended, roughened and slightly raised area (the deltoid tuberosity) that serves as the attachment for the tendons of the deltoid muscle (Metress, 1989). The shaft is also roughly cylindrical in this region. The kangaroo has a conspicuously flared and distinctive sharp and raised deltoid crest running along the anterior of the middle third (Oxenham and Barwick, 2008). In sheep the anterior ridge of the radial grove extends distally, giving it a very stout and bulky appearance for the first two-thirds (May, 1970). The shaft of both the dog and pig is laterally compressed, the latter distinctly flattened on the medial aspect (Bone, 1975).

In a human humerus, the major nutrient foramen is located medially and is always directed towards the distal end below the deltoid tuberosity (Bass, 1987). In a kangaroo the major nutrient foramen is located on the antero-medial aspect of the middle onethird. In sheep the major nutrient foramen is on the postero-lateral surface at the junction of the middle and distal thirds. Both pig and dog have the major nutrient foramen on the posterior surface approximately at midshaft (Sisson and Grossman, 1953).

(iii)

Distal-third

The distal end of a human humerus is expanded laterally and medially and flattens antero-posteriorly superior of the supracondylar ridge. The distal end may include a foramen extending through into the olecranon fossa (septal aperture); however this is a trait that may not always be present in humans (Bass, 1987; Adams, 2008). The distal third of a kangaroo humerus is most morphologically similar to a human, both having triangular, broad and flattened distal ends, and prominent medial epicondyles. However, the occurrence of both the hooked process of the supinator (or outer ridge) above the radial groove, and presence of the antero-medial supracondylar foramen, quickly distinguishes it as non-human (Owen, 1866).

26

The distal end of a sheep, pig, and dog humeri is much narrower by comparison (Oxenham, 2008). The olecranon fossa, on the posterior surface of a sheep humerus, is relatively deep and narrow (May, 1970). The olecranon fossa in a pig is also very deep; the plate of bone which separates it from the coronoid fossa is thin and sometimes perforated, creating a supratrochlear foramen analogous to a dog humerus (Sisson and Grossman, 1953; France, 2009). This feature is also sometimes present in humans.

3.1.2 Femora

A human femur has a relatively greater maximum length than in the the other species (Figure 3.2). However, all are superficially similar, with a femoral head set upon a neck for hip articulation and a distal extremity with two condyles for articulation at the knee joint. A human femur is characteristically long and rather slender, the shaft is curved anteriorly, the head is large and globular, and the neck elongated and narrow (Jenkins, 2002).

(i)

Proximal-third

A human femur has a large ball-like head that bulges out from a relatively narrow neck on all sides. The head bears, on its medial side, a prominent pit called the fovea (Hillson, 1996). In contrast, the femoral head of a kangaroo is markedly below the upwardly projecting greater trochanter therefore, the relatively undeveloped head merges gradually with a curved neck (Oxenham, 2008). The head of a sheep femur has a shallow fovea that spreads widely onto the proximal border of the neck; the greater trochanter is only a little more cranial than the head and is relatively small. The head of a pig femur is strongly curved, the greater trochanter more prominent than the lesser trochanter, which is rather small and less developed than the other species considered here (France, 2009). The femoral head of a dog is well developed containing a shallow fovea. The greater trochanter does not extend as high as the head (Sisson and Grossman, 1953). The head of an emu femur is well marked; the greater trochanter less developed than the other species examined here.

27

Figure 3.2. Anterior (above) and posterior (below) views of human, kangaroo, sheep, pig, dog and emu right femora (right to left).

28

(ii)

Middle-third

The shaft of a human femur is bowed, giving it a smoothed flattened appearance anteroposteriorly with rounded medial and lateral borders (edges), and a very prominent linea aspera on its posterior surface (Jenkins, 2002). The major nutrient foramen is approximately centred of the middle one-third, on the posterior surface of the bone, adjacent to, or on, the linea aspera; it is always directed toward the head (Bass, 1987; Metress, 1989). The shaft of a kangaroo femur is particularly robust and more rounded, although slightly convex anteriorly. The nutrient foramen is located at the junction of the proximal and middle one-third on the medial side of the posterior surface, next to a distinct tubercle positioned more centrally on the posterior surface of a kangaroo femur. The shaft of a sheep femur is slightly curved anteriorly, with a very rounded and solid shaft for the greater part of its length, except on the posterior surface, where the smoothness is interrupted by the slight elevation of the linea aspera. The nutrient foramen is located at the junction of the middle and distal-third of the medial side of the posterior surface (May, 1970).

The shaft of a pig femur is only slightly curved and has a prominent ridge running laterally from the greater trochanter to a large lateral supracondyloid crest. The dog femur is relatively more slender and cylindrical than that of a sheep or pig, with the shaft strongly curved in the distal two-third and convex anteriorly (Hillson, 1996). The femur of an emu is cylindrical and slightly curved; the nutrient foramen is on the posterior surface approximately centre of this middle-third (Sisson and Grossman, 1953).

(iii)

Distal-third

The distal end of a human femur distinctively flares to form relatively large articular condyles. The distal end of a kangaroo, sheep, pig and dog femora are by comparison relatively quite narrow (Oxenham, 2008). Trochlea ridges on the sheep femur are parallel, but slightly oblique (May, 1970). In both a pig and dog femur, the ridges of the trochlea are almost sagittal (Sisson and Grossman, 1953). The distal extremity of an emu femur has an anterior trochlea for articulation with the patella, and two posterior condyles for articulation with the tibia and fibula. The lateral condyle is marked by a 29

groove for articulation with the fibula (Sisson and Grossman, 1953). The supracondylar fossa is a rough depression proximal to the lateral condyle, and is especially deep and prominent in the kangaroo, sheep and emu, but not prominent in the human, pig and dog (Hillson, 1996).

3.1.3 Tibiae

The tibia from a large kangaroo can have a greater maximum length than that of a human or emu; it is characterised by a prominent wedge-shaped or triangular proximal half, which gradually becomes thinner and more cylindrical distally. The tibiae of a sheep, pig and dog are all slender and of a similar size and robustness. The emu tibia is a straight cylinder and structurally different both proximally and distally to the other species shown below (Figure 3.3) (Sisson and Grossman, 1953).

30

Figure 3.3. Anterior (above) and posterior (below) views of human, kangaroo, sheep, pig, dog and emu right tibiae (right to left). 31

(i)

Proximal-third

The proximal end of a human tibia is always the broadest part of the bone and is somewhat triangular (often termed the tibial plateau). The condyle outlines are larger and more rounded in a human, and irregular in the non-human species (Hillson, 1996; Jenkins, 2002; Muscolino, 2006). The tuberosity is only moderately developed in a human tibia, with the anterior crest progressing for more than half the entire length of the bone. The posterior surface contains the popliteal line and the major nutrient foramen is on the lateral border, directed toward the distal end (Bass, 1987).

A kangaroo tibia has a characteristically pronounced and very compressed medio-lateral ridge on the upper half of the anterior surface; the major nutrient foramen is on the posterior surface near the junction of the proximal and middle one-third (Sisson and Grossman, 1953; Oxenham, 2008). The tibial tuberosity in a sheep is relatively prominent and angular and joins a crest that is well rounded and developed and extends inferiorly to the midshaft (Hillson, 1996). On the posterior surface the popliteal line is poorly defined and extends distally to the junction of the middle and distal thirds. The nutrient foramen is on the lateral border in the proximal third (May, 1970).

The tibial tuberosity in a pig is massively developed and rounded and joins a prominent outward curving crest that fades gradually into the shaft (Sisson and Grossman, 1953; Hillson, 1996). The crest is short but very prominent in a dog tibia and the major nutrient foramen is on the posterior surface of the lateral border (Sisson and Grossman, 1953). The proximal end of the tibia in an emu has a lumpy flat surface with a crest at the anterior margin. The proximal side of the shaft has a ridge (fibular crest) that is characteristically prominent. The major nutrient foramen is on the posterior surface near the junction of the proximal and middle-third (Sisson and Grossman, 1953).

(ii)

Middle-third

The posterior surface of a human tibia is slightly concave and convex anteriorly, due to a sharp anterior crest (Bass, 1987). The shaft of the tibia is triangular with three borders; anterior (crest), medial (smooth and rounded) and interosseus (lateral border 32

characterised by a crest) with sharp, distinct edges (Metress, 1989; Muscolino; 2006). The shaft of a kangaroo does not contain any sharp or distinct edges, and has a more rounded and straight middle-third. The shaft of a sheep tibia is well rounded in the middle and flattens antero-posteriorly towards the distal third. The shaft of a pig tibia is slightly curved and convex medially (Sisson and grossman, 1953). The tibial shaft in a dog forms a double curve; the proximal part is convex medially and the distal part laterally. The tibia of an emu is smooth and flattened antero-posteriorly, with a slightly elevated ridge running along its medial border.

(iii)

Distal-third

In a human tibia, the anterior crest becomes less defined distally. The distal end of a human tibia is relatively flat and convex with one prominent downward projection called the medial malleolus, which is similar to a kangaroo, but slightly different to that of the sheep, dog and pig (Oxenham, 2008). It also has a prominent lateral groove for the fibula (Hillson, 1996). The distal end of a sheep, pig and dog tibiae is far more sculpted with a medial malleolus that is prolonged inferiorly with a pointed end. Laterally there is a deep narrow groove which separates the two prominences (Sisson and Grossman, 1953). The articular surface of the distal end in the emu is characterised by two prominent rounded condyles that are separated sagittally by a deep groove (sulcus) for articulation with the metatarsus (Hesse and Wapnish, 1985).

3.2

Current Methods for Species Discrimination

Forensic anthropologists can quickly verify intact whole bones (as shown above) as human or non-human through their intimate knowledge of anatomy. However, this assessment can become challenging when presented with fragmented remains (especially the diaphysis) containing few diagnostic features. In such cases forensic investigators may rely on more specialised techniques, some of which are described in more detail below.

33

3.2.1 Microscopic Examination

Although there are several arrangements in the various types of bone comprising the human skeleton, they are all similarly constructed of units called osteons. Each cylindrical osteon has a capillary running along its axis in a canal called a Haversian canal (Martin et al., 1998). Although bones of various animals consist of the same skeletal material, they differ in histological appearance; the presence of dense Haversian bone or secondary osteons potentially distinguishing human (Figure 3.4) from nonhuman bone (Enlow and Brown, 1956; 1957; 1958). These differences can be seen qualitatively (i.e. structure and pattern) and/or quantitatively (also termed histomorphometrics) (Martiniakova et al., 2006).

Figure 3.4. Histological section showing osteons (O), Haversian canals (HC), Volkmann’s canal (VC) and interstitial lamellae (IL) (x80). (Adapted from Ross and Pawlina 2006:229).

The pioneering studies of Enlow and Brown (1956; 1957; 1958) provide thorough descriptions of the types of bone in different species, but without direct comparison or quantitative data. Conversely, Harsànyi (1993) provided quantitative data for Haversian 34

canal diameter, but with limited qualitative description. Cattaneo et al. (1999) formulated a discriminant canonical equation, but was unable to provide precise guidelines for measuring bone microstructure. Owsley et al. (1985) showed that secondary osteons could potentially discriminate human from non-human (deer) humeri. This technique, however, may not be species-specific as some mammals (e.g. other primates, goat, sheep, pig and cow) can share these structural features (Owsley et al., 1985; Hillier and Bell, 2007).

There are, nonetheless, useful pattern differences between human and non-human bone. For example, plexiform bone, a primary tissue characterised by a horizontal and regular rectangular organisation, is commonly found in non-human mammalian bone (Figure 3.5) and only rarely observed in humans. Plexiform bone is characteristic of the long bones of large rapidly growing mammals, such as deer, cow or sheep. It may, however, occasionally be present when a child is rapidly increasing in size during a growth spurt (Mulhern and Ubelaker, 2003).

Figure 3.5. Histological section of a deer humerus showing plexiform bone (x100). (Adapted from Owsley et al. 1985).

To assess whether patterns of osteon organisation can be applied for human from nonhuman discrimination of bone fragments, Mulhern and Ubelaker (2001), compared femoral midshaft sections of 60 human, 9 sheep and 6 miniature swine. They 35

demonstrated that (like most studies) non-human bone may be easily identifiable by the presence of plexiform bone, although, in the absence of such a reliable indicator, other features such as a specific banding pattern may also be useful for non-human identification. However, these observations are mainly limited to subadult specimens, with more research needed on both mature bone as well as additional species, including bones from other parts of the skeleton. Overall, the shape and size of bone cells under magnification are not entirely indicative of human or non-human bone. Histological features differ within an individual skeleton as well as between different species (Pfeiffer, 1998; Hillier and Bell, 2007). As such, histological techniques are not widely applied and more research is needed in determining species-specific patterns, ideally by means of coalescing both qualitative and quantitative techniques (Martiniakova et al., 2006).

3.2.2 Protein Analysis

Ubelaker et al. (2004) modified Lowenstein’s (1980) immunological studies on speciesspecific protein detection in archaeological bone, for the purpose of forensic identification. The technique involves protein extraction followed by a solid-phase double-antibody radioimmunoassay (pRIA) using controls of antisera and radioactive antibody of rabbit gamma globulin. Species determination is based on the evaluation of radioactivity intake. Using this technique, Ubelaker et al. (2004) performed blind tests on six known bone samples (three human and three non-humans) on various skeletal elements, including long bones; correctly distinguished the human from non-human samples. As only small quantities are required, pRIA can be used to test for species prior to DNA analysis (Ubelaker et al., 2004). Importantly proteins are more resistant to many environmental factors compared to DNA (such as in the event of burnt remains), thus detection of species-specific protein in bone can be advantageous (Cattaneo, 2006). However, there is still relatively little research and larger samples and a wider range of species are required.

36

3.2.3 DNA Analysis

Immunological tests and DNA analysis are becoming increasingly successful in confirming the human origin of bone fragments, especially as histological methods are less widely applied (Cattaneo et al., 1999). When gross morphology is compromised, genetic species identification is possible by matching an unknown sample to a known reference, though comparing sequences of genes, usually mitochondrial DNA (mtDNA) loci that are known to vary between species. The sequence analysis of the cytochrome b (cyt b) gene is most frequently used for forensic species identification. This method can be applied to any biological sample, from bone to hair and tissue. The most obvious advantage of DNA analysis is that it utilises the naturally abundant mtDNA genome, therefore rendering it useful for all but the most degraded samples (Parson et al., 2000; Melton et al., 2007; Nakaki et al., 2007). Even though biomolecular methods such as DNA have reduced research into alternative species-specific techniques, confidence in DNA can yet be misleading. For example, with very dry bone, often genetic information is not sufficiently present, which makes DNA extraction impossible. In these instances histological analysis or other techniques are required. In addition, the cytochrome b gene sequence is largely unknown for most of the major vertebrate groups. Contamination can also be an issue, ultimately leading to waste of costly resources (Saferstein, 1993; Martiniakova et al., 2006).

3.2.4 Summary

Protein, DNA, or microscopic analysis may at times be the only applicable methods. They are all, however, costly and time consuming. Thus a simple method for differentiating human from non-human bone is clearly required, which does not require a specific bone or a specific location, has a low risk of contamination, high level of accuracy, and that can be performed rapidly and cost effectively.

37

3.3

Defleshing Methods

In order to expose the underlying osseous material for examination, all non-human samples require extensive skeletal preparation. Processing of bone specimens involves defleshing: removal of tissue from the skeleton, removing the grease from the bones and subsequent bleaching. There are four main methods commonly applied: water maceration; chemical maceration; carrion insects; and boiling. These methods are each summarily considered below.

3.3.1 Water Maceration

Water maceration, both cold and warm (up to 37°C), is the simplest and least aggressive method of skeletal preparation (Fenton et al., 2003). Carcasses are soaked in fresh water-filled containers that accumulate bacteria which aid in flesh decomposition. This is advantageous because it avoids strong chemicals and can be used for smaller specimens. It is, however, slow and produces a putrid odour (Hildebrand and Goslow, 2001); it is time consuming and can take weeks or even months depending on the size of the specimen and prior state of decomposition. It is not often used in routine forensic practice, as subsequent degreasing is required (Mairs et al., 2004).

3.3.2 Chemical Maceration

Two main chemical techniques are commonly used. The first (Snyder et al., 1975) uses an antiformin solution (sodium carbonate and bleaching powder) to macerate both fresh and embalmed cadavers. This takes as little as an hour and is amongst the fastest technique available. The chemicals used, however, are poisonous and must be handled carefully in a well-ventilated area using gloves and goggles. The chemicals can damage the bones, especially the epiphyses (Krogman and Iscan, 1985). The second technique (Stephens, 1979) utilises bleach and sodium hydroxide, and depending on the degree of decomposition may take between one to three days for complete tissue removal. There

38

is, however, a tendency for the cortical surfaces to flake after bleaching and the process requires regular supervision (Krogman and Iscan, 1985; Fenton et al., 2003).

Although chemical maceration is fast, it is relatively expensive, with an increased risk to both the sample and the preparer. Harsh chemicals such as chlorine bleaches, various oxidisers and papain should not be used because of their deleterious effect on bones and their continued reaction following cooking. To prevent bone damage, the chemical concentrations and the duration of maceration must be closely monitored (Fenton et al., 2003).

3.3.3 Carrion Insects

The use of dermestid beetle colonies for removing flesh is the most frequently published technique. It is efficient and commonly used by museums and taxidermists to prepare skeletal material (Mairs et al., 2004). Dermestids are a group of carnivorous beetles whose larvae ingest flesh, leaving the bones unmarked. A large number of adult beetles are required in a working colony, with the growing larvae consuming the greater amount of material (Anderson, 1948). Dermestid cleaning requires initial skinning and stripping of flesh from the remains, after which the remaining flesh is allowed to dry for several hours producing a ‘jerky-like’ texture before being introduced to the colony (Hildebrand and Goslow, 2001). The time required by the insects to clean small skulls usually takes 24 to 48 hours, depending upon the number of bugs and the amount of soft tissue material remaining on the carcass (Hall and Russell, 1933).

The main advantage of this method is that no chemicals are required and the end result is pristine skeletal material (Fenton et al., 2003). A colony, however, must be maintained at 20-30°C and often requires a large space. This method is ideal for smaller specimens, as it is difficult to maintain a colony big enough for large bones. This method does not degrease the bone so they will retain odours and fatty acids (Hildebrand and Goslow, 2001). It is also necessary to regulate the progress, as when deprived of soft tissue, the beetles start ingesting bone (Hefti et al., 1980). As a result, 39

keeping larvae alive when there are no specimens in preparation is problematic. This process is time consuming and obtaining and housing beetles in a bug-proof container at an even temperature can be an expensive undertaking.

3.3.4 Boiling

Boiling fleshed bones in a weak solution of water and commercial enzyme detergent is another method for the rapid removal of flesh (Barker et al., 2008). Cooking produces fully defleshed and degreased bones within a day or two. Mooney (1982) used active biologic enzymes (in detergent) to deflesh a skull over a constant heat. Such detergents are available commercially in the United States (e.g. Biz, Borax and Oxydol) and they have few health and safety issues. They are non-bleaching, do not consume any calcium, and degrease the bone (Krogman and Iscan, 1985; Fenton et al., 2003).

Selecting an appropriate preparation method depends on the required end result and the time and facilities available. The aim of the present project was to apply a relatively quick, easy, inexpensive and safe defleshing method on the specimens, without affecting the underlying bone. The major strength of the boiling method is the safety of the ingredients employed, both to the preparer and to the skeletal specimen, with no volatile compounds that produce noxious fumes. Therefore maceration of specimens using the boiling method was deemed to be the most rapid, safe and effective technique of defleshing.

40

CHAPTER FOUR

MATERIALS AND METHODS

4.1

Materials

For the human sample, a total of 50 of each of the three long bones (humeri, femora and tibiae) were measured. For the five non-human species a minimum of 10 femora (20 for the kangaroo), 10 humeri (emu excluded) and 10 tibiae (15 for emu) were measured. The non-human sample is smaller than the human sample as the primary aim is to elucidate species level variation, as well as the substantial time required for sample preparation.

4.1.1 Sources of Skeletal Material

All of the human long bones were acquired from the School of Anatomy and Human Biology, The University of Western Australia. The 10 sheep were obtained from the School of Animal Biology, The University of Western Australia, and belong to the most common species of domestic sheep (Ovis aries), of which there are over 200 breeds, all weighing between 30-70kg. The 10 feral pigs (Sus scrofa) were donated by the Department of Environment and Conservation (DEC) in the course of their program to cull feral animals. These carcasses weighed between 45-90kg. The 10 dog carcasses (Canis familiaris) were acquired from the City of Armadale pound and were a mixture of various breeds and weights (10-55kg). The 20 kangaroos (Macropus fuliginosus) and 15 emus (Dromaius novaehollandiae) were obtained from Kanga Pet Meats (Midland, WA) and the Toodyay Emu Farm (Toodyay, WA) respectively.

41

4.2

Methods

This section outlines the method for defleshing and preparing the anatomical specimens. The measurements taken on the resulting skeletal material are then outlined; the statistical methods used to analyse that data are then described.

4.2.1 Preparation of Non-human Skeletal Material

The limbs were disarticulated at their respective girdles using a Stanley Retractable Blade Utility Knife. Where possible, attention was confined to the right limbs. As much of the soft tissue as possible was then removed prior to maceration to accelerate skeletal preparation. The materials were macerated by boiling in sixty litre cooking pots on hot plates positioned so that a constant heat source was provided (adapted from Barker et al., 2008). The bones were processed in water using sodium borate (borax) – a non-toxic multi-purpose household product – and potassium hydroxide to act as a catalyst (Simmons and Haglund, 2005). The amount of borax depended on the water volume and the size of the limbs being processed, however, on average approximately 2 cups of powdered detergent and three to four pellets of potassium hydroxide (each weighing approximately 0.14g) were used per 45L.

The animal limbs were immersed in the cooking pots and the water was initially brought to a boil and then reduced to a simmer. The temperature of the water was evenly maintained, preventing vigorous boiling at all times, thus minimising potential damage to bones from hitting the sides of the pot (Barker et al., 2008). A lid was used to prevent water evaporation. When a bone was too long to fit in the pot (e.g. kangaroo and emu) it was processed on one end and then inverted to process the remainder. It was often necessary (depending on decomposition stage) to frequently change the water until maceration was complete.

Following maceration the bones were left to cool slowly (Hildebrand and Goslow, 2001). The remaining flesh was then removed by hand under running water (persistent 42

ligaments and tendons were gently pulled away using plastic tweezers) and the residual waste drained into a sieve. All unwanted biological material was disposed following the appropriate protocols1. Maceration time varied depending on limb size and the amount of remaining flesh. A partially fleshed medium sized animal leg (e.g. sheep) takes approximately 24-48 hours to prepare. For future reference, it is recommended that string is tied to the end of each limb to make checking the process of maceration easier. The limbs were placed in labelled bags, such as stockings or other strong boilable flowthrough bags, to prevent the loss of small bones amongst the debris of boiled flesh (Barker et al., 2008).

The final step in the maceration and preparation process involved bleaching using diluted hydrogen peroxide (15-20%). This whitens the bones and removes any residual fat or grease. Enough bleach was used to completely cover the bones for approximately 24-hours. They were checked frequently to prevent them from becoming chalky and brittle (Searfoss, 1995). Once bleaching was completed, the bones were rinsed with water and air dried.

4.2.2 Whole Bone Data

Eight measurements (adapted from those of Bass (1987), Buck (1988) and Buikstra and Ubelaker (1994)) were taken at several different anatomical locations on the length of each bone; full definitions of all measurements are outlined in Appendix I. Measurements were taken to one decimal place and were acquired using sliding, spreading and vernier calipers and an osteometric board. Though cloth tapes are typically used for measuring circumferences, because they are liable to stretch and wear, new strings were used in their place, and these measured with vernier calipers (Chaplin, 1971).

Measurements were chosen to correspond with those commonly used in traditional anthropometry to quantify overall bone shape in order to increase opportunities for 1

For The University of Western Australia’s guidelines on correct disposal of animal carcasses and waste refer to: http://www.safety.uwa.edu.au/policies

43

comparative studies, and to also define the relatively featureless regions of fragmented bone shaft. All eight measurements require a complete long bone or a portion containing a complete proximal or distal end. Though identification needs also to be possible from partial bones (see below) the main purpose of this part of the project is to conduct a preliminary study to determine whether the use of standard measurements, rather than morphological features alone, can provide the basis for differentiation.

4.2.3 Partial Bone Data

As it may not always be possible to accurately determine the known midshaft of a fragmented bone, the effect of taking a measurement slightly above or below this point is an important consideration. In order to assess the degree of morphometric variation around the known midshaft, a subset of the original bones were re-measured to determine whether the power to discriminate between humans and non-humans is significantly reduced if measurements are taken above or below the known midshaft. The midshaft measurements (antero-posterior and medio-lateral diameters and circumference – see Appendix I) were selected specifically for this purpose, as not only are they used for all three long bones, but also because the definitions of these measurements require the complete length of a bone.

To assess whether the power to discriminate between humans and other animals is significantly reduced if measurements are not taken exactly at the midshaft, the measurements (see above) were taken at five levels (1centimetre and 2cm above and below the known midshaft) on 17 human and 50 pooled non-human species (40 for the humeri).

4.2.4 Measurement Error Calculations

Measurement reliability is a direct indicator of data quality, which is important in morphometrics. The aim is to statistically quantify the errors inherent in bone measurement; systematic, instrumental and personal (Chaplin, 1971; WHO Multicentre 44

Growth Reference Study Group, 2006). The technical error of measurement (TEM) is an accuracy index representing measurement quality and can be defined as the standard deviation of repeated measurements taken independently of one another on the same subject (Norton and Olds, 1996). Technique-based measurement error (intra-observer error) was assessed by measuring four human long bones – humerus, femur and tibia – on four separate occasions, with a minimum of one day between repeats to eliminate the possibility of recalling figures. The formula for TEM is: TEM =

As the size of the TEM is often associated with the mean of the variable, the following formula was used to convert the absolute TEM to a relative TEM (rTEM): rTEM(%) =

The rTEM provides an estimate of the error expressed as a percentage relative to the size of the measurement, and is analogous to the coefficient of variation.

The coefficient of reliability (R) estimates the proportion of total measurement variance that is not due to measurement error. The coefficient indicates the degree to which a given measurement is error free. For example, a measurement reliability R=0.85 indicates that the measurement is 85% error free (Frisancho, 1990).

R= S2 = usual population variance of a character

45

TEM values < 1.0 and R values approaching 1.0 indicate high levels of measurement repeatability. The results of the precision study (which are outlined in Appendix II) demonstrate that half the humeri measurements and the head circumference measurement for the femora produced relatively high TEM values (1.027 – 1.505%). However, the overall error rates for the remaining measurements are low with small TEM values (TEM ≥ 0.25); all have relative TEM scores of < 5%, thus indicating that they are being taken precisely (Weinberg et al., 2005). As the reliability (R) scores range from 0 to 1 (with a value of 1 indicating that no measurement error was present) a value of 0.75 or higher is considered precise (Mueller and Martorell, 1988). Accordingly all measurements have high reliability values (R ≥ 0.71) which indicate high measurement precision and that any variation within the measured samples were primarily due to factors other than measurement error (Weinberg et al., 2005).

4.3

Statistical Analyses

A series of statistical analyses were performed to assess the level of variation present between the species and the level of classification accuracy of the assignment of species at each of two levels: (i) human and all animal species individually (6 groups), and (ii) human versus all animal species grouped (2 groups). All calculations were performed using the Statistical Package for Social Sciences Program (SPSS 17.0).

4.3.1 ANOVA

A one-way analysis of variance (ANOVA) is different to a t-test in that it is used to compare the means of a larger number of populations (i.e. three or more) on a continuous variable. This test was performed to assess the level of morphometric variation between the six groups, with post-hoc comparisons to quantify the differences (Pallant, 2007). If between-group variability is substantially greater than within-group variability, then the means are declared to be significantly (Sig.) different (Voelkl and Gerber, 1999). Hence, if the significance value is less than or equal to 0.05, there is a significant difference somewhere among the mean scores on the dependent variable (measurement) for all 6 species (Pallant, 2007). 46

Under the Levene test, the majority of the measurements produced significance levels greater than 0.05, suggesting that the variances between groups are not equal, therefore violating the assumption of homogeneity of variance. However, post-hoc comparisons are intended to safeguard against the possibility of an increased Type 1 error, resulting in the application of the Games-Howell procedure, as it does not assume equal variance (Field, 2000; Pallant, 2007).

4.3.2 Discriminant (Canonical) Analysis

The main purpose of a discriminant analysis is to maximise differences between a priori defined groups. This procedure is commonly referred to as a multiple discriminant function, or canonical variates analysis (CVA). As a preliminary investigation, all eight measurements were entered into the analysis in a stepwise manner, so that at each step the measurement that contributes most to the separation of the groups is entered into the discriminant function first (Katzenberg and Saunders, 2000). Plots are used to show how each individual (per long bone) is separated into their distinct clusters around their species centroid (Field, 2000). However, as this stepwise analysis does not necessarily reflect real case scenarios (as the forensic anthropologist may not have the whole bone available to them), some predictors (e.g. midshaft only) were independently assessed in the event of fragmented remains.

4.3.3 Direct Discriminant Functions

Linear discriminant analyses are used to provide an indication of the success rate for prediction of group membership (Kinnear and Gray, 2009) for both the primary and secondary sets of data. A series of direct discriminant analyses were performed using a two group sample (human and five non-human species pooled) to calculate standards designed to simply determine if an unknown bone is human or non-human. These standards can then be used to ascertain group membership of unknowns (George and Mallery, 2006). Discriminant scores are obtained by multiplying each variable with its 47

corresponding coefficient, summing them, and then adding the constant. The discriminant score is then compared to the sectioning point, which indicates whether the unknown bone is classified as human or non-human. All discriminant analyses are cross-validated using a ‘leave-one-out’ procedure, which is the iterative removal of each specimen from its sample; that specimen is then treated as though no a priori assumption of identity exists. The identity of the specimen is then estimated using the original samples (including the original n-1 parent sample); cross-validation statistics are then calculated from the known and estimated identities (Polly and Head, 2004).

48

CHAPTER FIVE

RESULTS

5.1

Introduction

This chapter presents the results of the osteological statistical analyses (both primary and secondary sets of data) for the human and five non-human species (four for humeri). The tabulated results of the primary raw data collected from traditional anthropometric standards are presented in Appendix III; the descriptive statistics (means, standard deviations and variances) of that data are outlined in Appendix IV.

5.2

One-Way ANOVA

The raw measurement data for all six groups were compared to see if there were significant differences among the mean measurement values (dependent variable) for each bone. The results show significances at four levels: P ≤ 0.001 (***), P ≤ 0.01 (**), P ≤ 0.05 and NS (not significant).

5.2.1 Humerus

For the humerus the sample means of the human and four non-human groups were compared to assess where, among the eight measurements, significances are greatest. Table 5.1 show that the measurements with the greatest differences between the groups were maximum length, head circumference and head diameter (all P ≤ 0.001). This is expected as the human humerus is relatively longer in length as compared to the other animals, whilst also exhibiting a more prominent ball-like head. The shaft measurements (e.g. midshaft diameters and shaft least circumference) in comparison are not as significantly different between the human and non-human groups.

49

Table 5.1. ANOVA comparison of sample means of the eight humeri measurements for the five species.

Maximum length (A) | Head diameter (B) Human Measurement

A

B

Human Kangaroo

***

***

Dog

***

***

Pig Sheep

*** ***

*** ***

Kangaroo

Dog

Pig

Sheep

A

B

A

B

A

B

A

B

***

***

***

***

***

***

***

***

*

NS

**

NS

***

*

NS

NS

NS

***

NS

*

*

NS

**

NS

NS

*

NS

***

NS

***

NS

*

*** P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, NS: not significant

Head circumference (A) | Surgical neck circumference (B) Measurement Human

Human A B

Kangaroo Dog A B A B NS *** ** *** NS

NS

Kangaroo Dog

*** ***

** NS

NS

NS

Pig

***

**

NS

NS

NS

NS

Sheep

***

NS

NS

NS

**

NS

Pig A B *** **

Sheep A B NS ***

NS

NS

NS

NS

NS

NS

**

NS

*

NS

NS

*

Midshaft m-l diameter (A) | Midshaft a-p diameter (B) Measurement Human Kangaroo Dog Pig Sheep

Human A B

Kangaroo A B NS *

Dog A

B *

NS

*

NS

NS

NS

*

NS

NS

**

NS

**

NS

***

NS

*

NS

NS

Pig

**

**

A

B **

NS

NS

NS

NS

**

*

NS

NS

NS

NS

**

NS

***

NS

**

Sheep A B

NS

***

Shaft least circumference (A) | Epicondylar breadth (B) Measurement Human

Human A B

Kangaroo A B NS

Kangaroo Dog

NS NS

Pig Sheep

NS

NS

Dog A NS NS

B *** ***

***

NS

***

NS

***

NS

***

NS

NS

*

***

NS

***

NS

NS

50

Pig A NS

B ***

Sheep A B * ***

NS

***

NS

***

NS

NS

NS

NS

NS

NS

NS

NS

5.2.2 Femur

For the femur the measurements with the largest statistical differences between the human and non-human groups are maximum length, head circumference and condylar breadth (all P ≤ 0.001 – Table 5.2). Unlike the humerus, however, the three midshaft measurements (a-p diameter, m-l diameter and circumference) are also highly significantly different between the human and non-human groups (P ≤ 0.001). In the kangaroo sample, four out of eight measurements are not significantly different to the human group (Table 5.2). The only other non-significant difference between the human and non-human groups is transverse epicondylar breadth in the emu.

5.2.3

Tibia

Of the three bones examined, the tibia was found to be the most significantly different between the human and non-human groups; with five of the eight measurements (condylo-malleolar length, spino-malleolar length, midshaft a-p diameter, malleolus a-p diameter and maximum epiphyseal breadth) being significantly different (Table 5.3). The shaft measurements of the kangaroo tibia, as for its femur, are also not significantly different to the human group (excluding midshaft a-p diameter).

51

Table 5.2. ANOVA comparison of sample means of the eight femora measurements for the six species. Maximum length (A) | Femoral head circumference (B) Measurement Human Kangaroo Dog Emu

Human A B *** *** ***

*** *** ***

Kangaroo A B *** *** *** *

Dog

Emu

A ***

B ***

A ***

B ***

A ***

B ***

Sheep A B *** ***

***

***

* **

NS

***

**

***

**

***

NS

NS

NS

NS

***

**

***

**

NS

NS

***

Pig

NS

**

*** NS

***

**

NS

***

**

Pig

***

***

***

**

NS

Sheep

***

***

***

**

NS

NS

NS

*** P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, NS: not significant

Subtrochanteric circumference (A) | Midshaft a-p diameter (B) Measurement Human

Human A B

Kangaroo A B NS

NS

NS

NS

Kangaroo Dog Emu

*** ***

*** ***

*** ***

*** **

Pig

*

***

NS

Sheep

***

***

***

Dog

Emu

Pig

A ***

B ***

A ***

B ***

A *

B ***

Sheep A B *** ***

***

***

*** ***

** ***

NS

***

***

NS

NS

NS

***

***

***

*** * *

NS

NS

***

***

***

NS

NS

***

***

***

NS

*

***

*

NS

NS

Midshaft m-l diameter (A) | Midshaft circumference (B) Measurement Human Kangaroo Dog Emu

Human A B

Kangaroo A B NS

NS

*** ***

NS

NS

*** ***

*** ***

*** ***

Dog

Emu

Pig

A ***

B ***

A ***

B ***

A ***

B ***

***

***

*** ***

*** ***

***

***

NS

NS

***

***

***

*** NS

***

***

*

***

***

Pig

***

***

***

***

NS

Sheep

***

***

**

***

*

Sheep A B *** *** ** * ***

*** * ***

**

NS

NS

**

Condylar breadth (A) | Transverse epicondylar breadth (B) Measurement Human Kangaroo Dog Emu

Human A B *** *** ***

Kangaroo A B *** ***

*** ***

***

NS

NS

*** ***

Dog

Emu

A ***

B ***

A ***

***

***

NS

***

Pig B NS

*** ***

***

*** NS

***

***

NS

***

***

Pig

***

***

***

***

NS

Sheep

***

***

***

***

NS

52

A ***

B ***

Sheep A B *** ***

***

***

***

***

NS

NS

NS

NS

***

***

***

***

NS

NS

NS

NS

Table 5.3. ANOVA comparison of sample means of the eight tibiae measurements for the six species. Condylo-malleolar length (A) | Spino-malleolar length (B) Measurement Human Kangaroo Dog Emu

Human A B ** *** *** ***

Pig

***

Sheep

** *** *** *** ***

Kangaroo A B ** **

Dog

Emu

A ***

B ***

A ***

B ***

A ***

B ***

Sheep A B *** ***

***

***

NS

NS

***

***

***

***

***

NS

NS

NS

NS

***

***

***

***

***

***

***

***

NS

NS

***

***

***

NS

NS

***

***

***

NS

NS

***

***

*** ***

***

Pig

***

***

*** P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05, NS: not significant

Midshaft a-p diameter (A) | Midshaft m-l diameter (B) Measurement Human

Human A B

Kangaroo Dog Emu

*** *** ***

Pig Sheep

Kangaroo A B NS ***

NS

*** ***

**

***

NS

**

***

***

**

NS

* *

Dog

Emu

Pig

A ***

B ***

A ***

B ***

A ***

**

*

NS

* ***

**

*

**

*

NS

NS

NS

NS

***

***

***

***

NS

NS

*** ***

***

*

NS

NS

***

***

*

NS

NS

***

***

B

Sheep A B *** ***

NS

NS

NS

Midshaft circumference (A) | Shaft least circumference (B) Measurement Human Kangaroo Dog Emu

Human A B

Kangaroo A B NS

NS

Dog

Emu

B ***

A ***

B ***

A ***

B ***

**

NS

* ***

** ***

NS

NS

*

NS

NS

NS

NS

NS

***

***

***

***

NS

NS

NS

*** ***

*** ***

** *

**

***

***

NS

NS

NS

***

***

NS

NS

NS

***

***

Pig

***

***

Sheep

***

***

*

Sheep A B *** ***

A ***

NS

NS

Pig

NS

NS

NS

Malleolus a-p diameter (A) | Maximum distal epiphyseal breadth (B) Measurement Human Kangaroo Dog Emu

Human A B *** *** ***

*** *** *

Kangaroo A B *** ***

Dog A B *** *** NS

NS

*

Emu A B *** * *** ***

*** ***

***

* ***

***

***

NS

**

NS

***

***

NS

NS

NS

***

***

Pig

***

***

NS

Sheep

***

***

NS

53

Pig

Sheep A B *** ***

A ***

B ***

NS

NS

NS

NS

** ***

NS

NS

NS

***

***

***

NS

*

NS

*

5.2.4 Summary

Analysis of the raw measurement data clearly establish that the tibia is the bone that is most significantly different between the human and non-human groups. It is also evident that for both the femur and tibia, it is the midshaft region of the bone that is most different in overall size. This suggests that even in these almost featureless regions there is enough morphometric variation for discrimination. More significantly, at these middle thirds, both the femur and tibia exclude the kangaroo species as being significantly different to humans, indicating lack of morphometric variation of hind limbs between the two groups at this specific region.

5.3

Discriminant (Canonical) Analysis

The first canonical variates analysis (CVA) includes all eight measurements on all three long bones for the human and five non-human species (excluding emu for the humerus) separately. However, as the practical application of this requires a complete and morphologically diagnostic bone (often unrealistic in a forensic setting), measurements suitable for fragmented material are also analysed.

5.1.1 Humeri

(i)

Stepwise Analysis

The stepwise analysis of all five groups using all eight measurements results in all but one measurement (head circumference) being selected, resulting in 97.8% classification accuracy following cross-validation. The mean scores on the first four canonical variates (CV’s) are listed in Table 5.4, with the first two canonical variates combined accounting for over 96% of the total variation. The first canonical variate accounts for 85% of the variation and clearly separates the human from the non-human species without any overlap in the observed ranges (Figure 5.1). The second canonical variate accounts for 11.3% of the variance (Table 5.4) and primarily separates the kangaroo (but also the sheep) from the remaining groups (including human) (Figure 5.1). The 54

third and fourth canonical variates (not illustrated) are both very small (less than 5% combined) and separate sheep from the pigs, and the pigs from the dogs respectively, with all other groups in-between.

Table 5.4. Mean scores, canonical roots and variance (%) of the five species using a stepwise analysis of seven humeri measurements. Canonical Variate Human Kangaroo Dog Pig Sheep

CV1 5.510 -5.181 -7.040 -9.045 -6.282

CV2 -0.150 5.464 0.316 -1.110 -3.920

CV3 -0.138 1.413 -1.153 -2.058 2.486

CV4 0.015 0.186 -0.894 0.630 0.005

Canonical Roots % of Variance

0.988 85.0%

0.920 11.3%

0.787 3.4%

0.356 0.3%

Figure 5.1. CVA plot of individual scores around corresponding group centroids for the five species from the stepwise analysis of seven humeri measurements.

55

For the first CV, four of the standardised coefficients are positive; the remaining three are negative (Table 5.5). Of these, there is one relatively high positive coefficient (maximum length, +1.316) and two relatively high negative coefficients (midshaft medio-lateral diameter, -0.999 and surgical neck circumference, -0.663). These contrasting measurements indicate that species with greater maximum lengths and larger head diameters are being distinguished from those that are shorter in length and smaller in medio-lateral shaft width.

Table 5.5. Standardised coefficients for CV1-CV4 computed from the stepwise analysis of seven humeri measurements for comparing the five human and non-human groups. Standardised Coefficients CV1 CV2 CV3 CV4 1.316 0.117 -0.435 -0.310 -0.663 -0.264 -0.589 0.538 -0.229 -1.543 0.479 1.033 0.251 0.398 1.202 -0.493 -0.999 1.163 -0.749 -0.606 0.138 1.274 0.393 0.549 0.495 -0.950 -0.101 0.141

Measurement Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Epicondylar breadth Head diameter

For the second CV there are two relatively high positive coefficients (midshaft mediolateral diameter, +1.163 and epicondylar breadth, +1.274) and two comparatively high negative coefficients (shaft least circumference, -1.543 and head diameter, -0.950). Though not large, there is only one further measurement with a coefficient larger than 0.3; midshaft antero-posterior diameter (+0.398). Viewed together, those coefficients contrast species with larger medio-lateral diameters and distal articular ends, to those species that have narrower distal-thirds and smaller head diameters. The positive/negative ratio of the standardised coefficients implies that overall shape differences are an important basis for discriminating between species on CV2.

56

(ii)

Shaft Measurements

As fragmented bones (especially shafts) are often encountered in routine forensic casework, three shaft measurements (least circumference, midshaft a-p and midshaft ml) were analysed to test their group separating/discriminating potential. It was found that 82.2% of specimens could be correctly classified using those three measurements. The mean scores on the first three canonical variates for the five groups are shown in Table 5.6. The first two CV’s combined account for 95.8% of the total variation, with the CV1 accounting for just over 70% of the total variation. This CV effectively isolates the human and sheep groups from both each other and from the rest of the non-human species (Figure 5.2). Although the group centroids are well separated along CV1, it is evident from Figure 5.2 that individual scores overlap considerably. The second CV (24.8% of the total variation, Table 5.6) separates kangaroos from the pigs, with the human, dog and sheep species positioned in-between (Figure 5.2). The third CV (not illustrated) accounts for only a very small amount of the total variance (less than 5%) and separates the sheep from the dogs, with the kangaroos, pigs and humans inbetween, but separated.

Table 5.6. Mean scores, canonical roots and variance (%) of the five groups computed from three humeri shaft measurements.

Canonical Variate Human Kangaroo Dog Pig Sheep

CV1 0.629 -1.939 -2.028 -2.135 2.640

CV2 0.129 1.997 -0.407 -1.780 -0.455

CV3 -0.196 0.496 -0.555 0.466 0.572

Canonical Roots % of Variance

0.848 71%

0.687 24.8%

0.363 4.2%

57

Figure 5.2. CVA plot of individual scores around corresponding group centroids for the five species from the three humeri shaft measurements.

(iii)

Midshaft Only

In the event that only a small fragment of the midshaft is available for examination, the discriminatory power of two measurements (antero-posterior and medio-lateral midshaft diameters) was assessed. Cross-validated classification accuracy was reasonable at 70%, with CV1 accounting for 89% of the total variation (Table 5.7). Although individual scores clearly overlap each other, the group centroids are still well separated along CV1 (Figure 5.3). The second CV accounts for 10.7% of the total variance and separates the kangaroos from all of the other groups.

58

Table 5.7. Mean scores, canonical roots and variance (%) of the five groups computed from two humeri midshaft measurements. Canonical Variate Human Kangaroo Dog Pig Sheep

CV1 0.686 -0.760 -1.877 -2.773 1.982

CV2 -0.180 1.293 -0.536 -0.066 0.207

Canonical Roots % of Variance

0.825 89%

0.451 10.7%

Figure 5.3. CVA plot of individual scores around corresponding group centroids for the five species from the two humeri midshaft measurements.

59

5.3.2 Femora

(i)

Stepwise Analysis

The mean scores for each CV for the human and five non-human species are listed in Table 5.8. The stepwise analysis resulted in seven of the original eight measurements being selected (midshaft a-p diameter excluded) which resulted in 97.3% classification accuracy following cross-validation. The first two CV’s combined account for 96.5% of the total variation, with CV1 accounting for most of the variance (88%). This CV effectively isolates the human group from the rest of the non-human species, with the kangaroo, dog, pig and sheep all clustered together and in-between the human and emu group (Figure 5.4). The second CV accounts for 8.5% of the variance (Table 5.8) and clearly isolates the emus from the rest of the groups, which are relatively closely clustered. It is evident that there is no overlap in the observed ranges between the human and non-human species groups (Figure 5.4). The remaining three CV’s are all very small (each less than 2%) with the human group separated, but still relatively close to the non-human species.

Table 5.8. Mean scores, canonical roots and variance (%) of the six groups using a stepwise analysis of seven femora measurements. Canonical Variate Human Kangaroo Dog Emu Pig Sheep Canonical Roots % of Variance

CV1 7.354 -4.346 -4.725 -10.719 -6.299 -6.336

CV2 0.423 -0.067 -2.719 5.528 -2.893 -1.899

CV3 -0.146 1.926 -1.156 -0.959 -0.811 -0.198

CV4 -0.003 0.246 -0.199 0.113 1.873 -2.262

0.029 -0.135 -0.942 -0.008 0.559 0.516

0.99 88%

0.911 8.5%

0.708 1.8%

0.677 1.5%

0.356 0.30%

60

CV5

Figure 5.4. CVA plot of individual scores around corresponding group centroids for the six species from the stepwise analysis of seven femora measurements.

For CV1, five of the standardised coefficients are positive; the remaining two are negative (Table 5.9). Of these, there is one relatively high positive coefficient (maximum length, +1.599) and two relatively high negative coefficients (midshaft circumference, -1.217 and subtrochanteric circumference, -0.828). These contrasting measurements indicate that species with greater maximum lengths are being distinguished from those that are shorter in length and have smaller shaft circumferences.

For CV2 there are two relatively high positive coefficients (transverse epicondylar breadth, +1.156 and midshaft medio-lateral diameter, +0.986) and two comparatively high negative coefficients (femoral head circumference, -0.734 and midshaft circumference, -0.686). The positive/negative ratio of the standardised coefficients would again suggest that overall shape differences are an important basis for discriminating between species. In this case, species with larger medio-lateral diameters and distal articular ends are being distinguished from those that have narrower head and midshaft circumferences. 61

Table 5.9. Standardised coefficients for CV1-CV5 computed from the stepwise analysis of seven femora measurements for comparing the six human and non-human groups.

Measurement Maximum length Femoral head circumference Subtrochanteric circumference Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epicondylar breadth

(ii)

CV1 1.599 0.344 -0.828 0.178 -1.217 0.062 0.242

Standardised Coefficients CV2 CV3 CV4 -0.108 0.013 0.180 -0.734 -0.482 -0.408 0.186 -0.768 1.649 0.986 -0.107 -0.872 -0.686 1.166 -1.102 0.013 1.738 1.321 1.156 -1.386 -0.630

CV5 -0.592 1.494 0.100 -0.576 1.202 -0.469 -0.534

Shaft Measurements

Of the original eight femoral measurements a total of four were chosen as being more suitable for forensic application (e.g. fragmented bones – midshaft circumference, midshaft a-p diameter, midshaft m-l diameter and subtrochanteric circumference). Using those measurements resulted in 74.5% of specimens being correctly classified following cross-validation. The first two CV’s account for 92.8% of the total variation (Table 5.10). On CV1 (63.1% total variance) the kangaroo group is in the middle with the human and sheep mean centroids on either side; pig and dog groups both have mean scores greater than -3, with the emu clearly separated on the positive end (Figure 5.5). It is also evident that there is a considerable overlap in the observed ranges between these six species. The second CV accounts for nearly 30% of the variance and again separates the emus from the other species. The third and fourth canonical variates (not illustrated) account for a very small amount of the total variance (less than 5%) with the human groups separated but still close to the five non-human groups.

62

Table 5.10. Mean scores, canonical roots and variance (%) of the six groups computed from four femora shaft measurements.

Canonical Variate Human Kangaroo Dog Emu Pig Sheep

1st 0.918 -0.019 -3.118 3.304 -3.204 -1.534

2nd -0.786 -0.197 0.298 2.950 2.134 -1.057

3rd 0.328 -0.405 0.136 -0.367 0.483 -1.083

4th -0.055 0.643 -0.762 -0.357 0.449 -0.339

Canonical Roots % of Variance

0.885 63.1%

0.793 29.7%

0.435 4.1%

0.39 3.1%

Figure 5.5. CVA plot of individual scores around corresponding group centroids for the six species from the four femora shaft measurements.

(iii)

Midshaft Only

The analysis of three midshaft femoral measurements (circumference, a-p diameter, and m-l diameter) resulted in 70.9% classification accuracy following cross-validation. The first two CV’s combined account for 95.7% of the total variation (Table 5.11). On CV1 (86.7% total variance) the emu group is clearly separated from all the other groups; the dogs and pigs are most removed from the emus, with the human, kangaroo and sheep 63

groups situated in-between (Figure 5.6). The second CV only accounts for 9% of the total variation but does separate the human sample from the five non-human species. The third CV accounts for a very small amount of the total variance (less than 5% and not illustrated) and separates the kangaroo from the dog, with the human and three nonhuman groups located in-between, but separated. It is also clearly evident that there is considerable overlap in the observed ranges for each of the six species (Figure 5.6).

Table 5.11. Mean scores, canonical roots and variance (%) for the six groups computed from three femora midshaft measurements.

Canonical Variate Human Kangaroo Dog Emu Pig Sheep

CV1 0.882 -0.032 -3.099 3.438 -3.094 -1.593

CV2 0.559 -0.307 -0.004 -1.249 -0.209 -0.722

CV3 -0.062 0.656 -0.767 -0.356 0.427 -0.304

Canonical Roots % of Variance

0.885 86.7%

0.521 9.0%

0.39 4.3%

Figure 5.6. CVA plot of individual scores around corresponding group centroids for the six species from the three femora midshaft measurements. 64

5.3.3 Tibiae

(i)

Stepwise Analysis

In the stepwise analysis of the six groups using all eight measurements, all but one measurement (midshaft circumference) is selected, resulting in 96.2% classification accuracy following cross-validation. The mean scores for CV’s 1-5 are listed in Table 5.12. The first CV (65.4% of total variance) clearly separates the emu and then the kangaroo groups, from the human, sheep, pig and dog groups, which are clustered closely together (Figure 5.7). There is a considerable overlap in the observed ranges between the sheep, pig and dog species, but no overlap in the individual scores for the human, kangaroo and emu groups (Figure 5.7). The second CV (19.8% of total variance) separates the humans and kangaroos from the negatively scored emu, with the closely clustered sheep, pig and dog centroids all having scores greater than -3. The third CV (not illustrated) separates the emu from the kangaroo, with the human, pig, sheep and dog in-between, but separated.

Table 5.12. Mean scores, canonical roots and variance (%) of the six groups using a stepwise analysis of seven tibiae measurements. Canonical Variate Human Kangaroo Dog Emu Pig Sheep

CV1 -3.063 4.327 -3.629 13.693 -3.149 -2.775

CV2 2.302 4.247 -4.217 -1.325 -6.111 -3.441

CV3 1.652 -7.245 -2.344 2.099 0.081 -1.901

CV4 0.009 0.420 -0.433 -0.115 1.299 -1.161

-0.010 0.060 -0.695 -0.039 0.217 0.529

Canonical Roots % of Variance

0.987 65.4%

0.959 19.8%

0.945 14.2%

0.507 0.6%

0.276 0.1%

65

CV5

Figure 5.7. CVA plot of individual scores around corresponding group centroids for the six species from the stepwise analysis of seven tibiae measurements.

Three of the standardised coefficients for CV1 are positive; the remaining four are negative (Table 5.13). Of these, there is clearly one very high positive coefficient (spino-malleolar length, +10.791) and one very high negative coefficient (condylomalleolar length, -9.748). This positive/negative coefficient ratio would suggest that those species with longer tibiae are being distinguished from those that are shorter in length. The standard coefficients of CV2 include three positive and four negative values (Table 5.13). Of these, there is one relatively high positive coefficient (condylomalleolar length, +3.290) and one relatively high negative coefficient (spino-malleolar length, -2.293). This would again suggest that overall shape differences (i.e. length variation) are important for discriminating between species.

66

Table 5.13. Standardised coefficients for CV1-CV5 computed from the stepwise analysis of seven tibiae measurements for comparing the six human and non-human groups.

Measurement Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

(ii)

Standardised Coefficients CV1 CV2 CV3 CV4 -9.748 3.290 -4.232 2.689 10.791 -2.293 3.384 -2.675 -0.739 0.520 0.795 0.499 0.838 -0.381 -0.668 0.582 -0.527 -0.603 0.222 -0.899 0.357 -0.495 0.563 1.304 -0.800 0.619 0.415 -1.289

CV5 4.136 -3.936 -0.885 -0.351 0.680 1.075 -0.575

Shaft Measurements

Four of the shaft measurements (m-l, midshaft a-p, circumference and least circumference) were analysed, with 78.1% of species being correctly classified following cross-validation. The mean scores for CV’s 1-4 are listed in Table 5.14; the first two CV’s accounting for 97.4% of the total variation. CV1 (75% of total variance) clearly separates the human group from the five non-human groups (Figure 5.8). The second CV accounts for 22.4% of the variance and separates the emus from the dogs, with the kangaroo, human, pig and sheep groups ranged in-between. It is apparent that there is also no overlap in the individual scores between the human and non-human groups on CV1 or CV2 (Figure 5.8). The third and fourth canonical variates are both very small (less than 3% combined and not illustrated) and separates the humans from that of the non-human species along both CV’s.

67

Table 5.14. Mean scores, canonical roots and variance (%) of the six groups computed from four tibiae shaft measurements.

Canonical Variate Human Kangaroo Dog Emu Pig Sheep

CV1 2.830 -2.071 -2.628 -2.923 -2.569 -2.495

CV2 0.046 0.690 -1.973 2.940 -1.569 -1.787

CV3 -0.023 1.073 0.210 -0.351 0.384 -1.023

CV4 -0.002 0.044 -0.231 -0.024 0.161 0.071

Canonical Roots % of Variance

0.941 75.0%

0.836 22.4%

0.455 2.5%

0.093 0.1%

Figure 5.8. CVA plot of individual scores around corresponding group centroids for the six species from the four tibiae shaft measurements.

68

(iii)

Midshaft Only

In the event that only a small fragment of the midshaft is available for assessment, the discriminatory power using only three midshaft measurements (m-l, a-p, and circumference) is still reasonable with an expected cross-validated classification accuracy of 72.4%. The first CV (76% of total variance) separates the human group from the five non-human species (Table 5.15; Figure 5.9). The second CV (24% of variance) places the emu (+2.962) and dog (-1.969) groups at opposite ends, with the kangaroo, human, pig and sheep groups separated, but in-between. Again, there is no overlap in the individual scores between the human and non-human species (Figure 5.9).

Table 5.15. Mean scores, canonical roots and variance (%) of the six groups computed three tibiae midshaft measurements. Canonical Variate Human Kangaroo Dog Emu Pig Sheep

CV1 2.742 -1.747 -2.535 -2.885 -2.417 -2.684

CV2 0.037 0.656 -1.969 2.962 -1.574 -1.738

CV3 0.005 -0.123 -0.262 0.025 0.099 0.225

Canonical Roots % of Variance

0.938 75.9%

0.836 24.0%

0.12 0.2%

69

Figure 5.9. CVA plot of individual scores around corresponding group centroids for the six species from the three tibiae midshaft measurements.

5.3.4 Summary

The first series of canonical variates analyses involved the stepwise analysis of all eight measurements on each individual bone (humeri, femora, tibiae) for the six species separately (five for the humerus). The resulting CVA plots show that for each long bone of each species, all specimens are successfully separated into distinct clusters around their group centroids (Figures 5.1, 5.4 and 5.7). It is also evident in all three stepwise plots that there is a clear separation between the human and all non-human species, with no overlap in the observed ranges. Each species was also correctly classified with a high degree of expected accuracy: humerus (97.8%); femur (97.3%); tibia (96.2%).

However, by selecting measurements (such as those on the shaft) that is more likely to be forensically relevant, results in a reduction in classification accuracy for all three bones (Table 5.16). Nevertheless, there is still a clear separation between the human and all non-human species groups along CV1 for the humerus, femur and tibia. The tibial 70

measurements (Figures 5.8-5.9) yielded the greatest human/non-human separation, which is confirmed by relatively high classification accuracy (range: 72.4-78.1%).

Table 5.16. Summary of cross-validated classification accuracies for the humerus, femur and tibia. Stepwise Humerus 97.8% 97.3% Femur 96.2% Tibia

5.4

Shaft 82.2% 74.5% 78.1%

Midshaft 70.0% 70.9% 72.4%

Direct Discriminant Functions

Two group direct discriminant functions (human and all non-human species pooled) for each long bone (humeri, femora and tibiae) are performed using all eight measurements to determine how well an individual long bone can be assigned to its correct group. The first series of analyses included the stepwise method, thus requiring a complete and otherwise morphologically diagnostic bone. As fragmented remains are often part of routine forensic investigations, a series of standards are subsequently presented that are designed to be applied to incomplete long bones.

5.4.1 Humeri

In the stepwise analysis of all eight variables, maximum length, midshaft anteroposterior diameter, midshaft medio-lateral diameter and head diameter were selected, which resulted in 100% cross-validated classification accuracy (Table 5.17). In the likelihood that a complete shaft is available for examination (Function 2) the expected classification accuracy is 73.3%. The remaining four functions are designed to be applied in cases of fragmentary bones, with classification accuracy ranging from 63.3% (using single measurements taken at both distal and mid-shaft) to 74.4% (fragmented diaphysis).

71

Table 5.17. Discriminant function assignment and classification of human and all nonhuman species pooled computed from eight humeri measurements. †

Group centroids & [sectioning point]

Correctly assigned

H 4.732 [-0.592] A -5.915

H 50/50; A 40/40 [100%]

H -0.488 [0.061] A 0.610

H 41/50; A 25/40 [73.3%]

H -0.488 [0.061] A 0.610

H 42/50; A 25/40 [74.4%]

H 0.127 [-0.016] A -0.159

H 45/50; A 12/40 [63.3%]

H -0.335 [0.042] A 0.419

H 36/50; A 21/40 [63.3%]

H -0.114 [0.042] A 0.143

H 45/50; A 12/40 [63.3%]

*

Equation

Function 1:

stepwise variables

(ml x 0.055) + (ap x -0.141) + (mld x -0.411) + (hd x 0.189) + -9.255 Function 2:

complete diaphysis

(mld x 0.364) + (ap x -0.324) + (slc x 0.019) + -1.866 Function 3:

fragmented diaphysis

(mld x 0.384) + (ap x -0.303) + -1.595 Function 4:

midshaft

(ap x 0.356) + -7.103 Function 5:

midshaft

(mld x 0.340) + -6.752 Function 6:

distal shaft

(slc x 0.160) + -9.371 *

ml: maximum length, ap: midshaft antero-posterior diameter, mld: midshaft medio-lateral diameter, hd: head diameter, slc: shaft least circumference. †

H: Human, A: Non-human

5.4.2 Femora

In the stepwise analysis, six of the eight available measurements were selected (Function 1: maximum length, femoral head circumference, midshaft antero-posterior diameter, midshaft circumference, condylar breadth and transverse epicondylar breadth) with a classification accuracy of 100% (Table 5.18). Function 2 demonstrates that measurements taken on a long bone shaft missing the articular ends results in 81.8% classification. Functions 3 to 5 are intended for fragmentary remains and yielded expected classification accuracies ranging from 78.2% (three midshaft measurements) to 65.5% (single midshaft measurement). 72

Table 5.18. Discriminant function assignment and classification of human and all nonhuman species pooled computed from eight femora measurements.

† *

Equation

Function 1:

H 0.968 [0.081] A -0.806

H 38/50; A 52/60 [81.8%]

H 0.802 [0.067] A -0.668

H 39/50; A 47/60 [78.2%]

H 0.425 [0.036] A -0.354

H 33/50; A 39/60 [65.5%]

H 0.439 [0.037] A -0.366

H 33/50; A 42/60 [68.2%]

midshaft only

(ap x 0.230) + -5.865 Function 5:

H 50/50; A 60/60 [100%]

fragmented diaphysis

(ap x 0.744) + (mld x 0.692) + (mc x -0.425) + -2.781 Function 4:

H 5.405 [0.451] A -4.504

complete diaphysis

(ap x 0.551) + (mld x 0.658) + (mc x -0.257) + (sc x -0.086) + -1.855 Function 3:

Correctly assigned

stepwise variables

(ml x 0.039) + (fhc x 0.036) + (ap x -0.187) + (mc x -0.091) + (cb x -0.120) + (teb x 0.162) + -8.083 Function 2:

Group centroids & [sectioning point]

midshaft only

(mld x 0.219) + -5.293 *

ml: maximum length, fhc: femoral head circumference, ap: midshaft antero-posterior diameter, mld: midshaft medio-lateral diameter, mc: midshaft circumference, sc: subtrochanteric circumference, cb: condylar breadth, teb: transverse epicondylar breadth. †

H: Human, A: Non-human

5.4.3 Tibiae

For the stepwise analysis of the eight measurements a total of six were selected (see Function 1 – Table 5.19) which resulted in 100% classification accuracy. Function 2 demonstrates that if only a shaft (complete diaphysis) is available an expected accuracy of 99% can be achieved employing four shaft measurements (ap, mld, mc and slc). Function 3 (excluding the slc measurement) also provided the same expected classification. Function 4 shows that even a single midshaft measurement (ap) can predict human or non-human group membership with a very high 82.9% accuracy; classification accuracy, however, drops to 66.7% when the single mc measurement is used (Function 5 – Table 5.19). 73

Table 5.19. Discriminant function assignment and classification of human and all nonhuman species pooled computed from eight tibiae measurements.

† *

Equation

Function 1:

H 2.837 [0.129] A -2.579

H 49/50; A 55/55 [99%]

H 2.692 [0.123] A -2.447

H 49/50; A 55/55 [99%]

H 1.237 [0.057] A -1.124

H 46/50; A 41/55 [82.9%]

H 0.226 [0.011] A -0.205

H 34/50; A 36/55 [66.7%]

midshaft only

(ap x 0.269) + -6.369 Function 5:

H 50/50; A 55/55 [100%]

fragmented diaphysis

(ap x 0.535) + (mld x -0.399) + (mc x -0.016) + -3.259 Function 4:

H 3.637 [0.166] A -3.306

complete diaphysis

(ap x 0.476) + (mld x -0.513) + (mc x -0.028) + (slc x 0.087) + -4.242 Function 3:

Correctly assigned

stepwise variables

(ap x 0.453) + (mld x -0.423) + (mde x 0.206) + (mc x -0.019) + (sml x -0.081) + (cml x 0.077) + -7.443 Function 2:

Group centroids & [sectioning point]

midshaft only

(mc x 0.067) + -4.560 *

ap: midshaft antero-posterior, mld: midshaft medio-lateral diameter, mde: maximum distal epiphyseal breadth, mc: midshaft circumference, sml: spino-malleolar length, cml: condylo-malleolar length, slc: shaft least circumference. †

H: Human, A: Non-human

5.4.4 Summary

For all three long bones the discriminant function that provided the highest correct discrimination between the human and non-human group was Function 1; correctly assigning individuals with 100% accuracy when employing the stepwise procedure (Tables 5.17-5.19). For Function 2 (complete diaphysis) the expected classification accuracy ranges from 99% (tibia) to 73.3% (humerus). The accuracy of subsequent functions designed for fragmentary remains is clearly reduced, although still relatively high: humerus – 63.3-74.4%; femur – 65.5-78.2%; tibia – 66.7-99%. It is evident that of the three long bones, the tibia was able to be classified with the highest degree of accuracy (Table 5.19). 74

5.5

Morphometric Variation around the Known Midshaft

An additional secondary set of measurements were taken to examine whether the power to discriminate between humans and non-humans is considerably reduced if measurements are taken slightly above or below the known midshaft. Measurements from Appendix I (ap, mld, and mc) were taken 1cm and 2cm above and below the known midshaft on each bone (humerus, femur and tibia) on the human (n=17) and ten each of the non-human species (n=40 for humeri). The results of the secondary midshaft measurements are outlined in Appendix V, including means and standard deviations for each individual specimen.

5.5.1 Standard Deviation

The average standard deviation for the midshaft measurements (ap, mld and mc) taken at the five levels on the human and pooled non-human samples (Appendix V) are summarised in Table 5.20. For the midshaft anterior-posterior diameter (ap) measurements, the smallest standard deviation in the human sample is for the femur (0.370) and the largest is for the tibia (1.034). In the non-human sample the smallest standard deviation for the ap measurement is also for the femur (0.469), but the bone with the largest variability is the humerus (2.205). The medio-lateral diameter (mld) measurement had the lowest mean standard deviation in the human femur (0.345) compared to that of the human humerus (0.860), with the femur again being least variable (0.586) in the non-human sample, compared to the humerus with over three times the average standard deviation value at 1.897 (Table 5.20). The tibial midshaft circumference measurement (mc) has the largest standard deviation values, for both the human (1.803) and combined non-human (2.676) samples (Table 5.20).

75

Table 5.20. Standard deviations calculated for the mean of the measurements taken at and around (1cm and 2cm above and below) the known midshaft.

Humerus

Human

Non-human Pooled

ap

0.420

2.205

mld

0.860

1.897

Femur

Human

Non-human Pooled

ap

0.370

0.469

mld

0.345

0.586

Tibia

Human

Non-human Pooled

ap

1.034

1.329

mc

1.803

2.676

*

ap: midshaft antero-posterior diameter, mld: midshaft medio-lateral diameter, mc: midshaft circumference.

5.5.2 Classification Accuracies

Of the measurements taken at the five levels on each long bone (Appendix V), the mean and maximum value (2cm above and below the known midshaft) for each individual specimen were used in the appropriate discriminant function from Tables 5.17-5.19; the discriminant scores were compared to the corresponding sectioning point to indicate whether the bone is classified as human or non-human (Appendix VI). Overall classification accuracies for the three analyses (mean, 2cm above and 2cm below) for the two midshaft measurements per long bone are summarised in Table 5.21.

76

Table 5.21. Overall classification accuracies for the mean and maximum values above and below the known midshaft.

HUMERUS

ap

mld

ap FEMUR

mld

ap TIBIA

mc

Mean H 5/17 A 8/40

2cm Above H 10/17 A 1/40

2cm Below H 7/17 A 11/40

13/57 = 22.8%

11/57 = 19.3%

18/57 = 31.6%

H 12/17

H 7/17

H 12/17

A 16/40

A 26/40

A 12/40

28/57 = 49.1%

33/57 = 57.9%

24/57 = 42.1%

H 12/17 A 30/50

H 12/17 A 30/50

H 12/17 A 30/50

42/67 = 62.7%

42/67 = 62.7%

42/67 = 62.7%

H 13/17

H 13/17

H 12/17

A 30/50

A 31/50

A 31/50

43/67 = 64.2%

44/67 = 65.7%

43/67 = 64.2%

H 15/17 A 40/50

H 17/17 A 34/50

H 15/17 A 39/50

55/67 = 82.1%

51/67 = 76.1%

54/67 = 80.6%

H 15/17 A 36/50

H 15/17 A 29/40

H 14/17 A 36/50

51/67 = 76.1%

44/67 = 65.7%

50/67 = 74.6%

*

ap: midshaft antero-posterior diameter, mld: midshaft medio-lateral diameter, mc: midshaft circumference.

When the average and maximum ap measurements, taken above and below the known midshaft on the subset of human and non-human humeri, are inputted into the discriminant function from Table 5.17, the degree of classification accuracies ranges from 19.3 to 31.6% (Table 5.21); a significant reduction from the original 63.3%. For the femur there is a much smaller reduction in classification accuracies using all three ap measurements from 65.5% (Table 5.18) to 62.7%; only a 2.8% decrease. When the average ap measurements are used in the appropriate discriminant function for the tibia (Table 5.19) the degree of classification accuracy is 82.1%; a reduction of only 0.8% from the original standard. If the measurements taken away from the known midshaft are used, the classification accuracy is 76.1% at 2cm above, and 80.6% at 2cm below.

77

The three mld measurements for the humeri produced slightly better accuracies compared to the ap measurements on the same bone (Table 5.21) with classification accuracy ranging from 42.1% (below) to 57.9% (above). However, the fall in accuracy is around 20% from the original 63.3% (Table 5.17). There are a greater number of individuals correctly assigned as belonging to a human or non-human group using the mld measurement in Table 5.21 for the femur (64.2-65.7%) the largest deficit being 4% from the original standard (Table 5.18). For the mc measurement taken 1 or 2cm above the known midshaft (65.7%) there is only a one percent decrease in accuracy from the original 66.7% (Table 5.19), whilst both the average (76.1%) and below known midshaft (74.6%) resulted in accuracies higher than that of the original standards.

5.5.3 Summary

The average standard deviations for the measurements (ap, ml and mc) in Table 5.20 quantifies the variability of values within 2cm above and below the known midshaft. It is evident that the degree of variability around the known midshaft is very small (0.3452.676) for each bone in both the human and non-human samples (Table 5.20). However, of the three long bones, the ap and mc measurements in the tibia appears to be most variable (1.034-2.676) within the 4cm midshaft region for both the human and pooled non-human groups.

Of the three long bones, the classification accuracies for the humeri (Table 5.21) using the mean and maximum values (1cm and 2cm above and below known midshaft) for the two midshaft measurements (ap and mld) resulted in the largest reduction in classification accuracy when compared to the original standards (Table 5.17). Correct human and non-human classification ranged from 57.9% (mld at 2cm above) to as low as 19.3% (ap at 2cm above); a fall in accuracy as large as 44%. When compared to the discriminant functions in Table 5.18, the same two ap and ml measurements resulted in much better classification rates for the femur, with the reduction in classification accuracy being between 2.8% - 4% (Table 5.21). However, once again, it is the tibia that provides the best discrimination between the human and non-human groups when the average ap measurement is used in the appropriate discriminant function from Table 78

5.19. The degree of classification accuracy at this measurement is 82.1%; a reduction of only 0.8% from the original standard. If the two measurements taken 1 to 2cm away from the known midshaft are used, high classification accuracy is still achieved; 76.1% (above) and 80.6% (below). The mc measurements taken at the five levels for the tibia resulted in a classification accuracy range of 65.7-76.1% (Table 5.21), which is not only a negligible reduction in accuracy (1% - above) compared to the original standards (Table 5.19), but has even resulted in an increase in accuracy of up to 9.4% (using the mean of the midshaft measurements – above and below).

79

CHAPTER SIX

DISCUSSION AND CONCLUSIONS

6.1

Introduction

The correct identification of human from non-human fragmentary bones is clearly an important initial step in eliminating non-human remains from further consideration in forensic situations. The primary aim of the present study was to establish whether noninvasive morphometric methods can be applied to fragmentary long bones (humerus, femur and tibia) to differentiate human from selected non-human species (kangaroo, sheep, pig, dog and emu).

6.5

Statistical Analyses

This study has demonstrated that differentiating humans from the five non-human species using quantitative methods is possible through the statistical analyses of linear measurements. Preliminary results from ANOVA (Tables 5.1-5.3) and CVA (Figures 5.1-5.9) showed that there were enough morphometric differences between all six species (five for the humerus) resulting in extremely high classification accuracies for all three long bones (96.2-97.8%). However, this does little more than replicate what can be achieved by visual observation by a competent comparative anatomist or forensic anthropologist.

More significantly, this study has presented a series of standards, designed to be applied to long bone fragments, that are still accurate for human from non-human species discrimination (63 to 99% - see Tables 5.17-5.19). Furthermore, even if the precise midshaft location is unknown (required for most shaft measurements) the degree of size variation within 2cm above and below the known midshaft was shown to be very small (Table 5.20). The subsequent flow-on effect on classification accuracy is relatively small, especially for the tibia, which appears to be the ‘bone of choice’ for attempting to 80

distinguish between the human and non-human species considered in the present study (Tables 5.19; 5.21). Fragments of the femur and humerus can also provide effective discrimination (Tables 5.17-5.18), however if the standards using single measurements are used, then classification accuracy is 65% or less, and thus unlikely to be useful in routine forensic investigation. Unlike the tibia, both the humerus and femur contain diagnostic features that happen to be present at the test location (near midshaft) that allow for morphological identification. These include the deltoid tuberosity (in the human humerus), the linea aspera (in the human femur) and the distinct tubercle on the posterior surface of the kangaroo femur (see Chapter 3 for further discussion on species morphology).

The high degree of classification accuracy achieved using tibial fragments can be correlated to the distinctive features of a bone with prominent diaphyseal characteristics: e.g. triangular proximally and gradually thinner and more cylindrical distally. Consequently, if an investigator is presented with an unknown diaphyseal fragment where external diameters clearly decrease proximally-distally with a simple more or less circular cross-section, this could be of great assistance in identifying the fragment as that of a tibia.

6.3

Morphometric Variation: Structure and Function

The human from non-human distinction that was elucidated is primarily associated with the different functions of bones in these species, as biomechanically, the shapes of whole bones can only exist within the limits of biological function (Oxnard, 1973). In these particular bones, functional differences relate principally to locomotor differences between the species examined. Human, emu and kangaroo each demonstrate their own unique versions of bipedality, from upright walking and running, through body horizontal running, and hopping (saltation), respectively. As a result, long bone form and function of bipedal saltators and bipedal cursors are different from those of quadrupedal cursors (pig, sheep and dog), as it is primarily the hind-limbs that bear the body’s full weight in bipeds (compared to the distributed loading over four limbs in quadrupeds). This distinction is illustrated and confirmed by the CVA stepwise plots 81

(Figure 5.10) where the quadrupedal species are consistently clustered together, but visibly separated from the bipedal groups for all three long bones.

HUMERUS

FEMUR

TIBIA

Figure 5.10. CVA plots of the three quadrupedal species (dog, pig and sheep - circled) for the humerus, femur and tibia samples.

Differences in locomotor gaits between the human and non-human bipedal groups contribute even further to the differences observed: evident in corresponding long bone architecture of hopping bipeds as compared to striding bipeds. The mechanical demands of saltatory hopping in kangaroos result in a relatively more robust femur and tibia; necessary to withstand the large forces during ground contact. Consequently, larger 82

macropodoid species (e.g. kangaroos) require large ankle extensor musculature, with a comparatively greater capacity for elastic strain energy storage and recovery during hopping (Bennett and Taylor, 1995; McGowan et al., 2008).

Although emus are somewhat more similar to humans in that they are a habitually striding biped, they differ in both arrangement and structure of limb segments. Emus have relatively short (almost horizontal) femora compensated by a corresponding increase in relative tibial length. They have additional specialisations that include a tibiotarsus and tarsometatarsus to maximise acceleration and mechanical advantage by using both muscle power and elasticity while running (Gatesy and Biewener, 1991; Hutchinson, 2004). As a result, both the femur and tibiotarsus in emus are relatively straighter, serving to reduce bending strains caused by hip and ankle extensor muscle forces transmitted along the lengths of the bones (Main and Biewener, 2007).

Variation between human and emu hind-limb segments are also related to differences in body and limb posture (upright versus crouched) along with associated differences in the body’s centre of mass (CM). Because humans support their trunk vertically, the CM is located almost directly over the hip, whereas the CM in birds are located well in front of the hip due to the more horizontal orientation of the avian vertebral column. Consequently, compared to the vertically oriented femur in humans, the avian femur projects anteriorly and horizontally to balance the CM over the feet (Gatesy and Biewener, 1991). As a result, the tibiotarsus acts as the major proximal lever (therefore equivalent, mechanically, to the femur in humans) and the tarsometatarsus as the main distal lever (equivalent mechanically to the tibia in humans). These modifications allow emus to achieve greater step lengths and duty factors even with a horizontal femur and more crouched posture (Gatesy and Biewener, 1991; Hutchinson, 2004).

Although mode of locomotion is the feature that most dictates the morphology of skeletal elements, other factors such as age and sexual dimorphism also influence morphological differences between species (Schmid, 1972; Reitz and Wing, 2008). Age at death determines overall body size, size and proportion of skeletal elements, as well as size of muscle attachments; all of which are important in identifying species. 83

Castration and other medical conditions also delay epiphyseal fusion, resulting in relatively longer and slender long bones (Chaplin, 1971; Davis, 1987). Maturation sequences of long bones may differ between wild and domestic animals, even within the same species, with epiphyseal closure found to occur earlier in domestic as compared to feral animals (Morris, 1972; Noddle, 1974). This may be attributed to the care given to domestic animals and to the selective breeding of domestic stock for rapid maturity. Furthermore, animals may also exhibit clear differences between adult males and females, known as sexual dimorphism, with size and robusticity of skeletal elements differing between them. Often, males are typically larger than females among many mammals and birds (Hillson, 1996; Reitz and Wing, 2008).

Environmental setting can additionally influence bone size and shape, and hence contribute to the observed morphometric variation. Under optimum or favourable nutritional conditions (such as those experienced by domestic animals) most animals develop faster and attain larger size at an earlier age than those animals growing in less favourable conditions (such as those experienced by wild animals) (Noddle, 1974; Hillson, 1996). Overall, variations between species are generally the result of both genetic variability in an animal population and the environmental conditions in which an animal grows and develops.

This study, however, has shown that traditional measurements can be employed to assess human versus non-human variation as a result of biomechanics. This is all in accordance with Ruff and Hayes (1983a) who suggest that linear measurements of external contours provide insights into the relationship between biological (mechanical) functions and morphological forms of bones. This relationship is confirmed and extended by the results of the present study.

6.4

Forensic Applications

Although visual assessment for distinguishing human from non-human bones is completely adequate for relatively complete bones, it is widely accepted that this is not 84

possible when the bone(s) are damaged or lack diagnostic features (such as muscle marking and joint surfaces). As a result, for fragmentary long bones, particularly diaphyseal fragments lacking diagnostic details, forensic investigators typically rely on molecular (Cattaneo et al., 1999; Ubelaker et al., 2004; Cattaneo, 2006) or histological (Owsley, 1985; Harsanyi, 1993; Mulhern et al., 2001; Martiniakova et al., 2006) methods for species determination. Even though histological or the introduction of newer, destructive forms of analysis (such as DNA or radioimmunoassay testing) can offer information regarding identification of bone fragments, such methods are inherently more expensive and time consuming and can thus introduce significant delays in a forensic investigation. Furthermore, more research is required to decrease the inaccuracies that exist within the current methodologies (e.g. Hillier and Bell, 2007, Martiniakova et al., 2006; Cattaneo, 2006).

This research, however, has demonstrated that human long bones can be differentiated from selected non-human species using simple linear measurements that can be taken even when distinctive morphological markers are absent. As a result, a series of standards (Tables 5.17-5.19) have been developed that can be applied to distinguish between human and selected non-human (sheep, pig, dog, kangaroo and emu) fragmentary remains (humerus, femur and tibia). More importantly, these standards are cost and time effective and non-invasive, allowing the physical integrity of the specimen to be preserved, an extremely important consideration in any forensic investigation.

The identification criterion developed in this thesis is completely novel and there is no evidence in the literature to suggest otherwise. Though the use of traditional morphometric methods are more commonly applied for formulating, for example, a biological profile (age, sex, stature and ethnicity) within forensic anthropology, they have not, until now, been used for human from non-human species discrimination. Usually, investigators will resort to histological or molecular methods if gross morphology is not viable (see above). This study has, therefore, effectively demonstrated that questions of medicolegal significance can be adequately addressed through quantitative morphometry, making subsequent destructive forms of analysis (i.e. histology and molecular methods) less necessary. Even if only a small fragment of 85

the diaphysis is present, there is still enough morphometric variation (both qualitative and quantitative) between the human and combined non-human species to make effective distinctions.

6.5

Future Research

The findings presented in this research have considerable potential for further development. The non-human sample sizes used here were relatively small (due to the considerable time involved in their procurement and preparation). Larger sample sizes would make further effective distinctions between human and non-human species possible using quantitative morphometry. The data collected on larger samples should not only corroborate on the findings here, but increase their value for broader investigations. Further extension of the present study also requires examination of data from other skeletal elements on both the non-human species already considered, in addition to other non-human species native to Australia. Investigation of the effects of age and sexual dimorphism within and between species would also be of forensic significance.

Technical changes, such as re-defining or simplifying some of the linear measurements used in the present study may also be useful. For example, instead of measuring maximum diameters in the antero-posterior and medio-lateral planes (as described by Bass 1987; see Appendix I), maximum (or minimum) midshaft diameters could alternatively be taken wherever they occur, thus eliminating the necessity of taking measurements in certain planes (Buikstra and Ubelaker, 1994). Coupled with quantifying the external dimensions of long bones, quantifying the internal architecture of the same bones by measuring the cortical thickness (see Chapter 2.1.2), would allow for an even more comprehensive and functionally (mechanically) interpretable data for human from non-human discrimination.

86

6.6

Concluding Summary

It is necessary to differentiate fragmentary human remains from equivalent non-human remains within the forensic context, as misidentification can result in an unnecessary waste of costly resources. Although there are various methodologies available (i.e. microscopic and molecular) for determining the origin of skeletal remains, their applicability is limited by cost and time. This study has provided accurate and easily applied non-invasive standards that allow for human from non-human species discrimination (kangaroo, emu, pig, dog and sheep) using traditional linear measurements of the tibia, femur or humerus animal bones from within the Western Australian milieu.

This study has outlined strong evidence that traditional measurement techniques, based primarily on linear measurements of external contours, is a forensically viable technique for species discrimination. This project has also shown that morphometric methods alone are more than adequate for human from non-human identification when remains are fragmentary and lack diagnostic muscle attachment marks or articular regions. Ultimately, an identification key has been developed that provides a quick and easy distinction between several species, not only indicating if the unknown bone is human, but, if not, also identifying the species to which it belongs (Figures 3.1-3.3).

However, the success of these results is only at a preliminary stage, with much potential for future research. The long-term goal of this study is to create a comparative collection of reliably identified, aged, and sexed skeletons of modern animals native to Australia. The standards formulated in the present study include older adolescent to fully adult specimens and as such are not designed to be applied to immature remains. If anthropological assessment indicates that the remains are clearly skeletally immature, then we suggest applying histological or molecular methods for distinction between human and non-human bones. If remains are also highly fragmentary or burned, it may again be necessary to use alternative methods to determine species.

87

References

Abourachid A, Renous S. Bipedal locomotion in ratites (Paleognatiform) examples of cursorial birds. The Ibis 2000; (142):538-549. Adams BJ, Crabtree PJ. Comparative Skeletal Anatomy: A photographic atlas for medical examiners, coroners, forensic anthropologists, and archaeologists. Totowa, NJ: Humana Press; 2008. Alexander RM. Exploring biomechanics: animals in motion. New York: Scientific American Library; 1992: p.23. Anderson RM. 2nd ed. Methods of collecting and preserving vertebrate animals. Ottawa: National Museum of Canada; 1948: p.140. Barker C, Cox M, Flavel A, Laver J, Loe L. Mortuary procedures II-Skeletal analysis I: Basic procedures and demographic assessment. In: Cox M, Flavel A, Hanson I, Laver J, Wessling R, eds. The Scientific Investigation of Mass Graves: Towards Protocols and Standard Operating Procedures. Cambridge: University Press; 2008: p.299. Bass WM. Human Osteology: A Laboratory and Field Manual. 3rd ed. Missouri: Missouri Archaeological Society; 1987. Bennett MB, Taylor GC. Scaling of elastic strain energy in kangaroos and the benefits of being big. Nature 1995; 378:56-59. Bloom RA. A Comparative Estimation of the Combined Cortical Thickness of Various Bone Sites. Skeletal Radiology 1980; 5:167-170. Bone JF. Animal Anatomy and Physiology. 4th ed. Oregon: O.S.U. Book Stores; 1975. Brothwell DR. Digging up Bones. 3rd ed. London: Oxford University Press; 1981.

88

Buck A. A Morphometric Analysis of Western Australian Aboriginal Post-Cranial Remains. Unpublished MSc Prelim, Department of Anatomy and Human Biology. Perth: University of Western Australia; 1988. Buikstra JE, Ubelaker DH. Standards for data collection from human skeletal remains. Arkansas: Arkansas Archaeological Survey; 1994. Cattaneo C, DiMartino S, Scali S, Craig, OE, Grandi M, Sokol RJ. Determining the human origin of fragments of burnt bone: a comparative study of histological, immunological and DNA techniques. Forensic Science International 1999; 102:181-191. Cattaneo C. Forensic anthropology: developments of a classical discipline in the new millennium. Forensic Science International 2006; 165(2-3):185-193. Chaplin RE. The Study of Animal Bones from Archaeological Sites. London: Seminar Press; 1971. Cornwall IW. Bones for the Archaeologist. Rev. Ed. London: Phoenix House; 1974. Croker SL, Clement JG, Donlon D. A comparison of cortical bone thickness in the femoral midshaft of humans and two non-human mammals. HOMO – Journal of Comparative Human Biology 2009; 60(6):551-565. Currey JD. Bones: Structure and Mechanics. New Jersey: Princeton University Press; 2002: pp.4-211. Dagg AI. Running, Walking and Jumping. London: Wykeham Publications; 1977: pp.99-103. Davis SJM. The Archaeology of Animals. London: B.T. Batsford Ltd; 1987: p.3260. Demes B, Jungers WL, Selpien K. Body Size, Locomotion, and Long Bone CrossSectional Geometry in Indriid Primates. American Journal of Physical Anthropology 1991; 86:537-547. Dickinson MH, Farley CT, Full RJ, Koehl MAR, Kram R, Lehman S. How Animals Move: An Integrative View. Science 2000; 288:100-106. 89

Enlow DH and Brown SO. A Comparative Histological Study of Fossil and Recent Bone Tissue. Part I. Texas Journal of Science 1956; 8:405-443.

Enlow DH and Brown SO. A Comparative Histological Study of Fossil and Recent Bone Tissue. Part II. Texas Journal of Science 1957; 9:186-214.

Enlow DH and Brown SO. A Comparative Histological Study of Fossil and Recent Bone Tissue. Part III. Texas Journal of Science 1958; 10:187-230.

Ferllini R. Silent Witness. London: Quintet Publishing; 2002: p.10. Fairgrieve SI. Of Beasts and Humans: A Case of Recognition. In: Fairgrieve SI, eds. Forensic Osteological Analysis: A Book of Case Studies. Springfield: Charles C Thomas; 1999: p.13.

Fenton TW, Birkby WH, Cornelison J. A Fast and Safe Non-Bleaching Method for Forensic Skeletal Preparation. Journal of Forensic Sciences 2003; 48(2):274-276. Field A. Discovering Statistics: using SPSS for Windows. London: Sage Publications; 2000: pp.276-422. France DL. Human and nonhuman bone identification: a colour atlas. Boca Raton: CRC Press; 2009. Frankel VH, Nordin M. Basic Biomechanics of the Skeletal System. Philadelphia: Lea & Febiger; 1980: p.15. Frazer JE. The Anatomy of the Human Skeleton. 4th ed. London: J & A Churchill Ltd; 1946: pp.2-7. Frisancho AR. Anthropometric standards for the assessment of growth and nutritional status. Anne Arbor: The University of Michigan Press; 1990: p.29. Gatesy SM, Biewener AA. Bipedal locomotion: effects of speed, size and limb posture in birds and humans. Journal of Zoology 1991; 224:127-147. George D, Mallery P. SPSS for Windows Step by Step: A Simple Guide and Reference. 6th ed. Boston: Pearson Education; 2006: pp.278-292. 90

Gilbert BM. Mammalian Osteo-Archaeology: North America. Columbia: Missouri Archaeological Society; 1973. Haglund WD, Reay DT, Swindler DR. Tooth mark artifacts and survival of bones in animal scavenged human skeletons. Journal of Forensic Sciences 1988; 33:985-997. Hall ER, Russell WC. Dermestid beetles as an aid in cleaning bones. Journal of Mammalogy 1933; 14:58-59. Hall BK. Bones and Cartilage: Developmental Skeletal Biology. Burlington: Elsevier; 2005: pp.3-52. Harsanyi L. Differential diagnosis of human and animal bone. In: Grupe G, Garland AN, eds. Histology of ancient bone: methods and diagnosis. Springer-Verlag; 1993: p.79-94.

Hefti E, Trechsel U, Rüfenacht H, Fleisch H. Use of Dermestid Beetles for Cleaning Bones. Calcified Tissue International 1980; 31:45-47.

Hesse B, Wapnish P. Animal bone archaeology: from objectives to analysis. Washington: Taraxacum; 1985.

Hildebrand M. Analysis of Vertebrate Structure. 4th ed. New York: John Wiley & Sons; 1995: p.477-479.

Hildebrand M, Goslow G. Analysis of Vertebrate Structure. 5th ed. New York: John Wiley & Sons; 2001: p.587.

Hillier ML, Bell LS. Differentiating Human Bone from Animal Bone: A Review of Histological Methods. Journal of Forensic Sciences 2007; 52(2):249-263. Hillson S. Mammal Bones and Teeth: An Introductory Guide to Methods of Identification. London: Institute of Archaeology; 1996. Hutchinson JR. Biomechanical modelling and sensitivity analysis of bipedal running ability. I. Extant taxa. Journal of Morphology 2004; 262:421-440. 91

Jenkins DB. Hollinshead’s Functional Anatomy of the Limbs and Back. 8th ed. Philadelphia: W.B. Saunders; 2002: pp.268-329. Katzenberg MA, Saunders SR. Biological Anthropology of the Human Skeleton. New York: Wiley-Liss Inc.; 2000: pp.52-399. Kinnear PR, Gray CD. SPSS 16 Made Simple. New York: Psychology Press; 2009: pp.525-563. Klippel WE, Synstelien JA. Rodents as taphonomic agents: bone gnawing by brown rats and gray squirrels. Journal of Forensic Sciences 2007; 52(4):765-773. Komar DA, Buikstra JE. Forensic Anthropology: Contemporary Theory and Practice. New York: Oxford University Press; 2008: pp.76-85. Krogman WM, Iscan MY. The Human Skeleton in Forensic Medicine. Springfield IL: Charles C Thomas; 1985: pp.37-43. Lovejoy CO, Burstein AH, Heiple KG. The biomechanical analysis of bone strength: a method and its application to platycnemia. American Journal of Physical Anthropology 1976; 44:489-506. Lowenstein JM. Species-specific proteins in fossils. Naturwissenschaften 1980; 67:343-346. Main RP, Biewener AA. Skeletal strain patterns and growth in the emu hindlimb during ontogeny. Journal of Experimental Biology 2007; 210:2676-2690. Mairs S, Swift B, Rutty GN. Detergent: An Alternative Approach to Traditional Bone Cleaning Methods for Forensic Practice. American Journal of Forensic Medicine and Pathology 2004; 25(4):276-284. Martin DL, Armelagos GJ. Morphometrics of Compact Bone: An Example from Sudanese Nubia. American Journal of Physical Anthropology 1979; 51:571-578. Martin RB, Burr DB, Sharkey NA. Skeletal Tissue Mechanics. New York: SpringerVerlag; 1998: pp.32-40. Martiniakova M, Grosskopf B, Omelka R, Vondrakova M, Bauerova M. Differences among species in compact bone tissue microstructure of mammalian 92

skeleton: use of a discriminant function analysis for species identification. Journal of Forensic Sciences 2006; 51(6):1235-1239. May NDS. The Anatomy of Sheep. 3rd ed. Brisbane: University of Queensland Press; 1970. McGowan CP, Skinner J, Biewener AA. Hind limb scaling of kangaroos and wallabies (superfamily Macropodoidea): implications for hopping performance, safety factor and elastic savings. Journal of Anatomy 2008; 212:153-163. McLorg PA. Variability in Humeral Dimensions in a Sample of Modern Human Females: Implications for Measuring Cortical Bone Loss. American Journal of Human Biology 2000; 12:118-127. Melton T, Holland C. Routine Forensic Use of the Mitochondrial 12S Ribosomal RNA Gene for Species Identification. Journal of Forensic Sciences 2007; 52(6):1305-1307. Merrilees D, Porter JK. Guide to the identification of teeth and some bones of native land mammals occurring in the extreme south west of Western Australia. Perth: Western Australian Museum; 1979. Metress SP. Human Osteology for the Archaeologist. Ringe, NH: Fund for Anthropology, Department of Anthropology, Dept. of Anthropology, Franklin Pierce College; 1989. Mooney MP, Kraus EM, Bardach J, Snodgass JI. Skull preparation using the enzyme-active detergent technique. The Anatomical Record 1982; 202:125-129. Morris O. A review of mammalian age determination methods. Mammal Review 1972; 2(3):69-104. Mueller WH, Martorell R. Reliability and Accuracy of Measurement. In: Lohman TG, Roche AF, Martorell R. Anthropometric Standardisation Reference Manual. Illinois: Human Kinetics Books; 1988: p84. Mulhern DM, Ubelaker DH. Differences in Osteon Banding Between Human and Nonhuman Bone. Journal of Forensic Sciences 2001; 46(2):220-222.

93

Mulhern DM, Ubelaker DH. Histologic Examination of Bone Development in Juvenile Chimpanzees. American Journal of Physical Anthropology 2003; 122:127133. Muscolino JE. Kinesiology: the skeletal system and muscle function. Missouri: Elsevier Mosby; 2006. Nakaki S, Hino D, Miyoshi M, Nakayama H, Moriyoshi H, Morikawa T, Itohara K. Study of anial species (human, dog, cat) identification using a multiplex single-base primer extension reaction in the cytochrome b gene. Forensic Science International 2007; 173:97-102. Noddle BA. Ages of epiphyseal closure in feral and domestic goats and ages of dental eruption. Journal of Archaeological Science 1974; 1(2):195-204. Norton K, Olds T. Anthropometrica. Sydney: UNSW Press; 1996. O’Neill MC, Ruff CB. Estimating human long bone cross-sectional geometric properties: a comparison of noninvasive methods. Journal of Human Evolution 2004; 47:221-235. Owen R. On the Anatomy of Vertebrates: Birds and Mammals. Vol. II. London: Longmans, Green, and Co.; 1866: p.353-354. Owsley DW, Mires AM, Keith MS. Case Involving Differentiation of Deer and Human Bone Fragments. Journal of Forensic Sciences 1985; 30:572-578. Oxenham M, Barwick R. Guide to the identifying human skeletal remains in Australia. In: Oxenham M. Forensic Approaches to Death, Disaster and Abuse. Queensland: Australian Academic Press; 2008: pp.1-5. Oxnard CE. Form and Pattern in Human Evolution. Chicago: The University of Chicago Press; 1973: p.19. Oxnard CE, Crompton RH, Lieberman SS. Animal lifestyles and anatomies: the case of the prosimian primates. Seattle: University of Washington Press; 1990. Oxnard CE. Ghostly muscles, wrinkled brains, heresies and hobbits: a leverhulme public lecture series. New Jersey: World Scientific Publishing; 2008: pp.39-80. 94

Pallant J. SPSS: Survival Manual. NSW: Allen & Unwin; 2007: p.103-246. Parson W, Pegoraro K, Niederstätter, Föger, Steinlechner M. Species identification by means of the cytochrome b gene. International Journal of Legal Medicine 2000; 114:23-28. Pfeiffer S. Variability of Osteon Size in Recent Human Populations. American Journal of Physical Anthropology 1998; 106:219-227. Polly PD, Head JJ. Maximum-likelihood identification of fossils: Taxonomic identification of Quaternary marmots (Rodentia, Mammalia) and identification of vertebral position in the pipesnake Cylindrophis (Serpentes, Reptilia). In: Elewa AMT. Morphometrics-Applications in Biology and Paleontology. Heidelberg: Springer-Verlag; 2004: p.197-222. Reitz EJ, Wing ES. Zooarchaeology. 2nd ed. New York: Cambridge University Press; 2008: pp.56-136. Ross MH, Pawlina W. Histology: A text and atlas with correlated cell and molecular biology. 5th ed. Baltimore: Lippincott Williams and Wilkins; 2006: p.229. Ruff CB, Hayes WC. Cross-sectional Geometry of Pecos Pueblo Femora and Tibiae – A Biomechanical Investigation: 1. Method and General Patterns of Variation. American Journal of Physical Anthropology 1983a; 60:359-381. Ruff CB, Hayes WC. Cross-sectional Geometry of Pecos Pueblo Femora and Tibiae – A Biomechanical Investigation: II. Sex, Age, and Side Differences. American Journal of Physical Anthropology 1983b; 60:383-400. Ruff CB. Allometry Between Length and Cross-Sectional Dimensions of the Femur and Tibia in Homo sapiens sapiens. American Journal of Physical Anthropology 1984; 65:347-358. Ruff CB. Body size, body shape, and long bone strength in modern humans. Journal of Humans Evolution 2000; 38:269-290. Runestad JA, Ruff CB, Nieh JC, Thorington Jr RW, Teaford MF. Radiographic Estimation of Long Bone Cross-Sectional Geometric Properties. American Journal of Physical Anthropology 1993; 90:207-213. 95

Saferstein R. Forensic Science Handbook. Vol. III. New Jersey: Prentice Hall; 1993: p. 435. Schmid E. Atlas of Animal Bones. Amsterdam: Elsevier Pub. Co.; 1972. Searfoss G. Skulls and Bones. US: Stackpole Books; 1995: pp.98-225. Simmons T, Haglund WD. Anthropology in a forensic context. In: Hunter J, Cox M. Forensic archaeology: Advances in theory and practice. London: Routledge; 2005: p.174. Sisson S, Grossman JD. The Anatomy of Domestic Animals. 4th ed. Philadelphia: W.B. Saunders Company; 1953. Snyder RG, Burdi A, Gaul G. A rapid technique for preparation of human fetal and adult skeletal material. Journal of Forensic Sciences 1975; 20:576-580. Stephens BG. A simple method for preparing human skeletal material for forensic examination. Journal of Forensic Sciences 1979; 24:660-662. Stevens J. Applied Multivariate Statistics for the Social Science. 3rd ed. New Jersey: Lawrence Erlbaum Associates; 1996: p.239. Stewart TD. Essentials of Forensic Anthropology: Especially as Developed in the Unites States. Springfield: Charles C Thomas; 1979: pp.45-58. Stock JT. A test of two methods of radiographically deriving long bone crosssectional properties compared to direct sectioning of the diaphysis. International Journal of Osteoarchaeology 2002; 12:335-342. Stock JT, Shaw CN. Which measures of diaphyseal robusticity are robust? A comparison of external methods of quantifying the strength of long bone diaphyses to cross-sectional geometric properties. American Journal of Physical Anthropology 2007; 134:412-423. Thomas P. Talking bones: the science of forensic anthropology. New York: Facts on File; 1995. Troughton E. Furred animals of Australia. 9th ed. Sydney: Halstead Press; 1967: pp.151-226. 96

Ubelaker DH. Human Skeletal Remains: Excavation, Analysis, Interpretation. 2nd ed. Washington: Taraxacum; 1989: pp.43-51. Ubelaker DH, Lowenstein JM, Hood DG. Use of Solid-Phase Double-Antibody Radioimmunoassay to Identify Species from Small Skeletal Fragments. Journal of Forensic Sciences 2004; 49(5):924-929. Voelkl KE, Gerber SB. Using SPSS for Windows: Data Analysis and Graphics. New York: Springer-Verlag; 1999: pp.195-206. Webster D, Webster M. Comparative vertebrate morphology. New York: Academic Press; 1974: p.105. Weinberg SM, Scott NM, Neiswanger K, Marazita ML. Intraobserver Error Associated with Measurements of the Hand. American Journal of Human Biology 2005; 17:368-371. Wescott DJ. Effect of Mobility on Femur Midshaft External Shape and Robusticity. American Journal of Physical Anthropology 2006; 130:201-213. White TD. Human Osteology. 2nd ed. San Diego: Academic Press; 2000: p.17-29. WHO Multicentre Growth Reference Study Group. Reliability of anthropometric measurements in the WHO Multicentre Reference Study. Acta Paediatricia 2006; 450:38-46. Wolf DJ. Forensic anthropology scene investigations. In: Reichs KJ. Forensic Osteology: Advances in the Identification of Human Remains. Springfield: Charles C Thomas; 1986: p.18-19. Woo SLY, Kuei SC, Amiel D, Gomez MA, Hayes WC, White FC, Akeson W. The effect of prolonged physical training on the properties of long bone: a study of Wolff’s law. Journal of Bone and Joint Surgery 1981; 63A:780-787.

97

Appendices

Appendix I: Long Bone Measurements Definition of long bone measurements commonly used in traditional anthropometry and the equipment used in this study. A. Humerus 1. Maximum length (ml): [osteometric board] From the most proximal point on the head, to the most distal point on the trochlea (Buck, 1988). 2. Surgical neck circumference (snc): [string] Taken at the point of maximum curvature, distal to the lesser tuberosity (Buck, 1988). 3. Shaft least circumference (slc): [string] This measurement is taken about the second one-third distal to the deltoid tuberosity (Bass, 1987). 4. Midshaft antero-posterior diameter (ap): [sliding caliper] Taken at midshaft, between the most anterior and posterior points (Bass, 1987). 5. Midshaft medio-lateral diameter (mld): [sliding caliper] Located at the same level as measurement 4, between the most medial and lateral points (Bass, 1987). 6. Head circumference (hc): [string] Circumference around the articular margin of the head (Buck, 1988). 7. Epincondylar breadth (eb): [osteometric board] Distance of the most laterally protruding point on the lateral epicondyle from the corresponding projection of the medial condyle (Buikstra & Ubelaker, 1994). 8. Head diameter (hd): [sliding caliper] Direct distance between the most superior and inferior points on the border of the articular surface (Buikstra & Ubelaker, 1994).

B. Femur 1. Maximum length (ml): [osteometric board] From the most proximal point on the femoral head, to the most distal point on the medial condyle (Buck, 1988). 2. Femoral head circumference (fhc): [string] Maximum circumference around the head of the femur (Buck, 1988). 98

3. Subtrochanteric circumference (sc): [string] Taken at the proximal third of the diaphysis, immediately below the lesser trochanter (Buck, 1988). 4. Midshaft antero-posterior diameter (ap): [sliding caliper] Taken at midshaft, between the most anterior and posterior points (Buck, 1988). 5. Midshaft medio-lateral diameter (mld): [sliding caliper] Located at the same level as measurement 4, between the most medial and lateral points (Buck, 1988). 6. Midshaft circumference (mc): [string] Circumference taken at the same level as in measurement 4 and 5 (Buck, 1988). 7. Condylar breadth (cb): [sliding caliper] From the most medial point on the medial condyle, to the most lateral point on the lateral condyle (Buck, 1988). 8. Transverse epicondylar breadth (teb): [osteometric board] Between the most medial point on the medial epicondyle, to the most lateral point on the lateral epicondyle (Buck, 1988).

C. Tibia 1. Condylo-malleolar length (cml): [spreading caliper] From the midpoint of the margin of the articular surface of the medial condyle, to the most distal point on the medial malleolus (Buck, 1988). 2. Spino-malleolar length (sml): [osteometric board] From the most proximal point on the intercondylar eminence to the most distal point on the medial malleolus (Buck, 1988). 3. Midshaft antero-posterior diameter (ap): [sliding caliper] Taken at midshaft, between the most anterior and posterior points on the shaft (Buck, 1988). 4. Midshaft medio-lateral diameter (mld): [sliding caliper] Taken at the same level as in measurement 3, between the most medial and lateral points (Buck, 1988). 5. Midshaft circumference (mc): [string] Circumference taken at the same level as in measurement 3 and 4 (Buck, 1988). 6. Shaft least circumference (slc): [string] Least circumference is usually located on the distal third of the shaft (Buck, 1988). 7. Malleolus antero-posterior diameter (map): [sliding caliper] From the most anterior point on the anterior border, to the most posterior point on the posterior border (Buck, 1988). 99

8. Maximum distal epiphyseal breadth (mde): [osteometric board] Maximum distance between the two most laterally projecting points on the medial malleolus and the lateral surface of the distal articular region (Buikstra & Ubelaker, 1994).

100

Appendix II: Measurement Error Calculations

Humeri: Measurement

Repeatability (S²)

TEM

Individual (S²)

Coefficient of Variation

Reliability

rTEM (%)

Maximum Length Surgical Neck Circumference Shaft Least Circumference Midshaft AP Diameter Midshaft ML Diameter Head Circumference Epicondylar Breadth Head Diameter

0.094 2.266 1.074 1.102 0.637 1.055 0.281 0.293

0.306 1.505 1.036 1.050 0.798 1.027 0.530 0.541

197.406 31.324 34.745 5.645 2.175 67.789 21.844 4.261

0.100 1.674 1.643 4.839 4.047 0.781 0.855 1.226

1.000 0.928 0.969 0.805 0.707 0.984 0.987 0.931

0.099 1.673 1.642 4.841 4.047 0.781 0.855 1.225

Femora: Measurement

Repeatability (S²)

TEM

Individual (S²)

Coefficient of Variation

Reliability

rTEM (%)

Maximum Length Head Circumference Subtrochanteric Circumference Midshaft AP Diameter Midshaft ML Diameter Midshaft Circumference Condylar Breadth Transverse Epicondylar Breadth

0.063 2.086 0.453 0.539 0.371 0.426 0.703 0.250

0.250 1.444 0.673 0.734 0.609 0.653 0.839 0.500

117.422 21.898 29.516 2.488 1.886 7.519 2.668 14.125

0.057 1.073 0.666 2.537 2.248 0.764 1.158 0.610

0.999 0.905 0.985 0.783 0.803 0.943 0.736 0.982

0.057 1.073 0.666 2.536 2.248 0.765 1.158 0.609

Tibiae: Measurement

Repeatability (S²)

TEM

Individual (S²)

Coefficient of Variation

Reliability

rTEM (%)

Condylo-malleolar Length Spino-malleolar Length Midshaft AP Diameter Midshaft ML Diameter Midshaft Circumference Shaft Least Circumference Malleolus AP Diameter Maximum Distal Epiphyseal Breadth

0.469 0.094 0.168 0.305 0.242 0.711 0.414 0.344

0.685 0.306 0.410 0.552 0.492 0.843 0.643 0.586

167.219 176.516 1.229 1.477 17.633 0.852 4.238 5.406

0.186 0.082 1.421 2.598 0.637 1.205 1.663 1.096

0.997 0.999 0.863 0.794 0.986 0.832 0.902 0.936

0.004 0.082 1.421 2.598 0.637 1.204 1.662 1.095

101

Appendix III: Results of Primary Data (mm)

Human: Bone number Side

H U M E R U S

Maximum length Surgical neck circumference Shaft least circumference Midshaft a-p diameter Midshaft m-l diameter Head circumference Epicondylar breadth Head diameter

Bone number Side

H U M E R U S

Maximum length Surgical neck circumference Shaft least circumference Midshaft a-p diameter Midshaft m-l diameter Head circumference Epicondylar breadth Head diameter

Bone number Side

H U M E R U S

Maximum length Surgical neck circumference Shaft least circumference Midshaft a-p diameter Midshaft m-l diameter Head circumference Epicondylar breadth Head diameter

1 R 333 88 57 21.5 18 126 62 42.5

2 L 298 91.5 67.5 24 21 130 61 44

3 4 R L 322 316 93.5 103 63 69.5 23.5 24 19 23.5 139 142.5 65 70 48.5 48

18 L 324 95.5 59 20 21 145.5 64 47

19 R 285 95.5 57.5 20.5 18 123 58 41.5

20 21 22 L R R 327 318 312 97 99.5 100 61 59 64 20 21 22 20 19 20 138 130.5 137.5 67 62 64 45.5 45 46

35 36 37 38 R R L L 318 302 317 303 97 104.5 103 98 53.5 60 59.5 54.5 17.5 20 21 19 17 21 21 18 56.5 129.5 140.5 136 65 62 64 60 43 43 48 46

Bone number

1 R Maximum length 500 Femoral head circumference 148.5 Subtrochanteric circumference 104 Midshaft a-p diameter 34.5 Midshaft m-l diameter 29 Midshaft circumference 94 Condylar breadth 80.5 Transverse epincondylar breadth90

2 L 394 125.5 99 25.5 25.5 78.5 66 75

Bone number

19 20 21 R R R 443 428 468 132 154 135.5 103 101.5 105 28 30.5 30.5 27.5 28 27.5 83 90 89 68 75 69 77 84 78

Side

F E M U R

18 R Maximum length 423 Femoral head circumference 131 Subtrochanteric circumference 96.5 Midshaft a-p diameter 21.5 Midshaft m-l diameter 23.5 Midshaft circumference 78 Condylar breadth 62 Transverse epincondylar breadth75 Side

F E M U R

5 6 R L 317 295 104.5 91 63.5 55 23 19 21 18 136.5 134.5 67 60 46 45

3 L 364 113.5 80.5 23 22 68.5 60 68

7 8 9 10 11 R R R L L 303 304 278 321 322 95.5 91 82 95 94 61 65 52 59.5 62 22.5 24 20.5 21.5 20.5 21 20 18 19 22 129 142 114 138.5 130.5 59 64 55 63 62 44 47.5 40 46 44

23 24 L R 335 283 100.5 82.5 63.5 49 22.5 17 22 16 137.5 116.5 61 52 44.5 40

39 L 271 86.5 52.5 20 16 116 56 36.5

40 L 318 101 65.5 25 22.5 139 64 45.5

4 5 6 L R L 422 405 430 118.5 132.5 113.5 90 95 90 28.5 28 24 26 25 27 83.5 82.5 78 64 70 61 75 82 68

22 23 L R 467 423 137.5 130.5 104 91.5 31 28 28 29 91.5 84 67.5 71 89 79

25 26 27 28 R R L L 320 313 275 314 92 97 75 101.5 56 59 52 61 19.5 22.5 17.5 21 19.5 18 16 21 126 131.5 119.5 141 56 66 53 63 41.5 44 40 48

41 42 43 L R L 310 293 308 96.5 85.5 96.5 59.5 52 58 21 19 18 20 17 19 130 120 127.5 63 57 59 44 38 42.5

13 R 293 90.5 60 19 20 131 67 42

29 R 282 83 52.5 20.5 16 112.5 55 36

30 31 32 33 34 R L R R R 295 272 275 308 287 85 87.5 97.5 98 96.5 47 49 56 57.5 53 16.5 17 21 19.5 20 16 16 17 18 17 114 120 120 135.5 120 55 55 58 66 55 37.5 39 40 45 40

44 45 46 R L R 287 248 289 97.5 85 100 52.5 50 57 20 18 19 15.5 14.5 17 123.5 113.5 58 56 45 59 40 37 40

10 R 431 120.5 90 25.5 26.5 79 62 74

14 L 282 84.5 57.5 21 18 128.5 58 44

15 16 L L 305 314 85 100 58.5 60 20 20 21 20 128 135 59 65 42 46

17 L 331 85 53.5 19 18 137.5 63 45

47 L 346 112 62.5 19.5 21 63.5 66 42.5

48 R 297 104.5 55.5 19.5 18 60 60 42

11 12 13 14 R L L R 407 436 427 428 124 134 138 126 94.5 103.5 103 103 28 29 27.5 30 26 27 27 24 82.5 85.5 80.5 83 69 71 68.5 67 72 82 79 89

15 L 451 134 96.5 31 25 86.5 71.5 80

16 R 444 125 104.5 30 28 90 72.5 80

17 R 449 141 95.5 27.5 27 85 76 83

32 R 422 128 91 29.5 25.5 86 66.5 75

33 L 447 125.5 95 27.5 25 80.5 68 76

34 L 420 125.5 94.5 25.5 24 77 64 72

7 R 424 117.5 91.5 25.5 26 80.5 58.5 70

8 L 434 142 93.5 25.5 27 79 70 82

9 R 433 141 108.5 28 26 82.5 77 89

24 L 425 134.5 101.5 29 29 81.5 63 71

25 L 440 136 95 27 26 81 73.5 83

26 27 28 L R L 456 410 450 146 123.5 120.5 118 80.5 87.5 27.5 23 27 31.5 24 24.5 87.5 71 77 70 65 63 85 71 71

102

12 L 294 87.5 60.5 19 22 129.5 68 41

29 30 31 R R L 430 408 470 142 120 132 89.5 90 102.5 27 25.5 34 25 22.5 28 80.5 74 96 69 63.5 72 80 70 84

49 50 R R 281 320 91 102 54.5 63.5 19 21 16.5 20 57 66 57 67 37 45

Bone number

35 R Maximum length 439 Femoral head circumference 140 Subtrochanteric circumference 96 Midshaft a-p diameter 27 Midshaft m-l diameter 26 Midshaft circumference 85 Condylar breadth 71 Transverse epincondylar breadth83 Side

F E M U R

Bone number Side

T I B I A

Condylo-malleolar length Spino-malleolar length Midshaft a-p diameter Midshaft m-l diameter Midshaft circumference Shaft least circumference Malleolus a-p diameter Max. distal epiphyseal breadth

Bone number Side

T I B I A

Condylo-malleolar length Spino-malleolar length Midshaft a-p diameter Midshaft m-l diameter Midshaft circumference Shaft least circumference Malleolus a-p diameter Max. distal epiphyseal breadth

37 38 R R 484 400 144 116 114.5 93.5 34 27 27 23 93.5 78 75 66 87 71

39 40 41 42 43 R R L L L 372 417 429 428 370 124 130.5 136 129 115 95 98.5 95.5 91.5 90 24 28.5 28 27 22 25 26.5 27 25.5 24 76.5 83.5 85 81 69 68 68 71 71 61 77 79 79 79 69

44 45 L L 455 453 139 130 114.5 102 25.5 27 31 28 88 82 70 74 85 80

46 R 482 141.5 105 29.5 26 87.5 71 77

47 R 405 132 85.5 24 26 72 70 76

48 49 R L 403 477 128 149 93.5 114 22 28 24 30 71.5 91.5 65.5 75.5 73 83

50 L 350 112.5 90.5 23.5 24 74.5 57 65

1 L 343 356 28 22 77 71.5 40.5 53

2 L 324 333 26 19 68 64 34 45

3 R 361 366 27 20 74 74.5 36 50

4 L 321 323 27 27.5 23.5 70.5 38.5 59

5 L 335 343 28 18 62.5 58 34 46

6 L 393 397 35 22 88 78 38 52

7 R 325 333 28 22 79 71 36 51

8 9 L R 352 335 365 348 27.5 27 22.5 20 81.5 76.5 72 70 35 36 50 51

10 R 329 338 27 20 75.5 71 39 50

11 L 307 313 26 16.5 63 58 32 44

12 13 14 15 16 L L L R R 380 375 370 318 347 388 384 379 326 354 31.5 28 29 29 26 23 19.5 21 21 21 83 75.5 76.5 78 75 70 69.5 65.5 71.5 71.5 41.5 38 38 33 37 57 52 50 48 56

17 L 379 390 30 18 71.5 64.5 36 48

18 L 375 384 30 19 77.5 72 37 54

19 R 372 383 30 25 83 73 40.5 58

20 L 375 381 30 21 77 72.5 39 54

21 L 380 388 29.5 19 73 69 37.5 54

22 L 364 373 29.5 21 20 77.5 36 50

23 L 362 357 27 19 70 62 33 43

24 25 26 L R R 362 357 365 372 365 373 31 27.5 31.5 21 21 20.5 73.5 75 79.5 67 71.5 70 34 40.5 35 47 55 48

27 L 334 340 26.5 19 70.5 67 34 47

28 L 331 342 28 20 75.5 68.5 36 50

29 L 362 369 32 20 81 70 34 49

34 R 311 320 22 18 62.5 60 30.5 46

Bone number Side

T I B I A

36 R 403 115.5 88 26 22.5 73.5 56 68

Condylo-malleolar length Spino-malleolar length Midshaft a-p diameter Midshaft m-l diameter Midshaft circumference Shaft least circumference Malleolus a-p diameter Max. distal epiphyseal breadth

35 R 349 358 25 18.5 74 66 34 51

36 R 405 415 32 25 93 85.5 42 53

37 R 388 395 28 20 79 69.5 40 55

38 L 358 365 31.5 21 83.5 74 36 51

39 L 355 363 30 20 74 70 35 50

40 L 315 322 25 17 63.5 59.5 31 46

103

41 L 393 403 35 20 86 75 39 53

42 43 L R 343 359 345 369 25 30 18.5 18 66 77 62 70 33 38 47 56

44 R 334 338 29.5 20.5 79.5 72.5 30 44

45 R 338 344 26 18 70.5 66.5 36 49

30 R 365 372 34.5 26.5 21 75.5 34.5 49

31 32 R R 335 350 341 356 28 23 19 16 72.5 58.5 65.5 54 32.5 33 43 47

46 47 48 L R L 295 330 364 303 338 372 23 22.5 30 18 17 23 61 64 82 62.5 59 77 31.5 32 44 43 45 57

33 L 352 358 25.5 18 65.5 62 33 47

49 50 R R 341 336 349 342 27 27 19 19 78 74.5 72 67.5 38 34 53 46

Kangaroo:

H U M E R U S

Bone number Side Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Head circumference Epicondylar breadth Head diameter

Bone number Side F E M U R

Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth

1 L 232 94.5 88.5 26 20.5 75.5 55 59

2 L 282 96 104 28.5 27 88 58 62

3 4 5 L L R 310 307 226 98 111.5 67.5 115 117.5 78 30 31 21 28 28 19 93.5 95 64 63 65 44 67 68 47

Bone number Side Condylo-malleolar length T Spino-malleolar length I Midshaft antero-posterior diameter B Midshaft medio-lateral diameter I Midshaft circumference A Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

1 L 229 102 62 25 26 97.5 63 26

2 L 186 96.5 58 17 19 86.5 58 26

3 L 203 98 56 19 22 98.5 57 28

4 L 224 106.5 69.5 23 26 100 64 30.5

6 7 8 9 10 R R R R R 234 235 226 230 231 91.5 75 74.5 71 88 88 87.5 80 87.5 88 26 23 21 22 25 21 23 21 22 23 75 72 67.5 69.5 75.5 54 48.5 45.5 47.5 48 58 52 50 51 51

1 R 471 479 23 25.5 79.5 68.5 30.5 45

2 R 526 533 24 25.5 78.5 71.5 30.5 43

104

3 R 548 556 27 30 89.5 77 32 46

5 R 230 101.5 61 23.5 23 94.5 62 28

11 R 216 71.5 76.5 21.5 20 64.5 45 49

4 R 452 462 25.5 25 76.5 69 30.5 42

12 R 238 78 80 22 20.5 66 45 49

5 R 421 428 19 18.5 58.5 53 23 34

6 L 184 100.5 58.5 21.5 24 92.5 60 27.5

13 14 R R 313 270 100 94.5 110 100.5 34 28 28.5 25 98.5 84 63.5 59 68 63

7 R 203 95 56 19 23 97 58 30

15 R 281 94.5 103 29 27 88.5 58.5 62

8 R 211 101.5 64.5 21 27 101.5 62 27.5

16 R 279 89.5 104.5 28 27 86 59.5 63

6 7 R R 409 413 415 420 19.5 20 19 21 60.5 64 53 59 23 23.5 31 34

17 R 294 94.5 105.5 31 25 88.5 61 65

9 10 R R 176 222 99.5 111.5 53.5 70 23.5 25 18 30 81.5 100 57 63 24 31

18 R 274 97.5 99 28 24 80.5 60 64

19 R 293 90.5 99.5 28 27 83.5 55 61

20 R 292 94.5 108.5 30 27.5 90 62.5 67

8 9 10 R R R 396 392 375 404 399 382 25.5 19 18 26 20 19 65.5 60.5 59 55.5 53.5 54.5 23 23 23 34 33 34

Sheep:

H U M E R U S

Bone number Side Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Head circumference Epicondylar breadth Head diameter

1 R 150 85.5 54.5 19 16 88 35 28

F E M U R

Bone number Side Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth

1 2 3 R R R 179 163 177 70 76 74.5 68 75 77.5 18 20 21.5 18 18.5 20 58 61 64.5 35 39 39 41 47 47

Bone number Side Condylo-malleolar length T Spino-malleolar length I Midshaft antero-posterior diameter B Midshaft medio-lateral diameter I Midshaft circumference A Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

1 R 207 212 13 16 49 47 21.5 30

2 3 R R 157 155 96 102.5 61.5 63 21.5 22 16.5 18 93.5 96 39 38 30 30

4 R 168 110 64.5 24 19.5 114 43 34

5 6 R R 161 170 105.5 104 67.5 68 24 25 21.5 21 105.5 101.5 43 42 31 31

4 5 R R 201 178 82 71 84 75 24 21 23 18 75 61.5 42 35 51 43

2 3 4 5 R R R R 198 231 227 213 203 235 232 214 15 17 17 15 16 19 19 18 52 59.5 58 54.5 51 58.5 57 55 21 25.5 24.5 24 33 34 36 34

105

7 R 144 88.5 58 18 16 100 38 28

8 R 166 109 67 23.5 21 111 42 33

9 R 170 107.5 66.5 25.5 20.5 116.5 43 34

10 R 153 103 62.5 22 19 112.5 49 31.5

6 7 8 9 R R R R 203 196 202 185 88 73.5 80.5 73 88.5 76 81 77 22 22.5 23.5 20.5 21 21 23 21 69.5 70 73 68.5 42.5 37.5 41 40 50 47 48 46

10 L 188 70 75.5 21 21 68.5 41.5 48

6 7 8 R R R 231 231 224 236 236 230 16 16 18 18 18 19 56 56 61 59.5 59.5 59.5 27 27 25.5 37 37 39

10 R 220 225 16 17 55 52 25.5 35

9 L 221 225 15.5 16.5 56 52 25 35

Pig: Bone number Side H Maximum length U Surgical neck circumference M Shaft least circumference E Midshaft antero-posterior diameter R Midshaft medio-lateral diameter U S Head circumference Epicondylar breadth Head diameter

1 R 203 127.5 67.5 19 26 93 46 29

6 R 146 102.5 53.5 15.5 19 79.5 38 24

7 R 166 112.5 58 16.5 21.5 95.5 45 29

8 R 148 112.5 59 16.5 20.5 94 44 26

Bone number Side Maximum length F Femoral head circumference E Subtrochanteric circumference M Midshaft antero-posterior diameter U Midshaft medio-lateral diameter R Midshaft circumference Condylar breadth Transverse epincondylar breadth

1 2 3 4 5 6 R R R R R R 217 197 202 208 192 180 79 82.5 86 73 74.5 77.5 97.5 94 89 88.5 87.5 82 24 21 22 23 21.5 21 19 16.5 17.5 20.5 17 16 70.5 62 64.5 70.5 61 59.5 43 45.5 46 42 42 42 48 49 51 46 47 47

7 R 177 78.5 89 20 17 60.5 42 46

8 9 10 R R L 166 145 155 71.5 76.5 73.5 75 76.5 71.5 18 16 17 15 14 17.5 54 49 49 37 39 35 41 42 40

1 2 3 4 5 6 R R R R R R 191 176 175 171 161 178 198 181 180 178 165 186 20.5 18 18 17.5 16 18 21 20.5 19.5 18 18.5 18 65.5 62 60.5 59 57 59.5 58.5 54.5 57 53 53 55 28 25.5 26 27 28 28 33 31 32 34 32 33

7 R 140 145 14 16 47.5 44.5 23 27

8 9 10 R R L 144 127 130 149 131 135 17 13 14.5 19 15 15 57.5 44.5 46 53.5 43 42.5 26 25.5 26 31 29 31

T I B I A

Bone number Side Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

2 R 184 125 73.5 21 27 87.5 44 27

106

3 R 171 120.5 61.5 18 23 101.5 47 32.5

4 R 185 115 62.5 18 24 99.5 47 27.5

5 R 178 115 62.5 18 23 94.5 43 26

9 L 133 92 49.5 13 17 83.5 37 25

10 R 114 96.5 48 14 17 91 43 26

Dog:

H U M E R U S

Bone number Side Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Head circumference Epicondylar breadth Head diameter

1 R 213 118.5 66.5 20.5 25.5 96.5 52 28

F E M U R

Bone number Side Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth

T I B I A

Bone number Side Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

2 R 190 108 64.5 21 23 100 47 29

9 R 117 79.5 41.5 13 14.5 64 43 22

10 R 119 99 40 13 14 86 43 26

1 2 3 R R R 218 221 213 73 78.5 76.5 81.5 88.5 86 19 21 21.5 17.5 21 20 59 66 66.5 40 43 41 45 50 51

4 5 6 7 8 9 R R R R R R 212 180 201 171 157 145 71 69 68.5 62.5 56.5 69 78.5 72 74 66.5 68.5 52 20 18.5 18 15.5 17.5 14 18 18 16 15 16 13 58 58 53.5 49.5 52.5 42.5 40.5 37 36 32 30.5 31.5 44 41 41 48 36 37

10 R 133 69 53.5 14 13.5 42 31 35

1 2 3 R R R 222 222 218 227 226 221 16.5 18 17 18 18 19.5 55.5 58.5 58 50.5 59 59 18 26 26 34 36 36

4 5 6 7 R R R R 184 124 143 157 186 125 145 159 15.5 16 17 13.5 17.5 18 17 16 52 40 37 41 54 41 37.5 47 24.5 17.5 17.5 18 32 24 23 27

10 R 200 202 15.5 18 53 52.5 20 29

107

3 R 201 110 58 17 23 94.5 46 28

4 R 179 97.5 54 18 19 86 39 25.5

5 R 197 104.5 58.5 19 22.5 88 45 27

6 R 164 96 58.5 17.5 20.5 90 42 27

7 R 156 92 48.5 15 19 85 38 24.5

8 R 142 85 50 15.5 18 70.5 37 22

8 R 163 167 13.5 16.5 49 47.5 19.5 29

9 R 219 221 16 17 54 54 22.5 31

Emu: Bone number Side Maximum length F Femoral head circumference E Subtrochanteric circumference M Midshaft antero-posterior diameter U Midshaft medio-lateral diameter R Midshaft circumference Condylar breadth Transverse epincondylar breadth

T I B I A

Bone number Side Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

1 R 411 449 24 30 87.5 78.5 43 54

2 R 424 458 25 27 83.5 76 40 51

1 R 243 86.5 134 31.5 34 104 63.5 80

2 R 239 88.5 131 30 31.5 96 60 74

3 R 246 102 142.5 33 35 106.5 61.5 79

3 4 5 R R R 418 423 419 458 464 456 26 26 25.5 28.5 30.5 31 86 89 89.5 75.5 79.5 83 47 45.5 44 54 57 54

108

4 R 230 78 126 29 30.5 95 55.5 69

6 7 R R 434 403 471 444 25 21.5 31 24 90 74 84 66.5 45 50 57 66

5 R 232 84.5 126.5 27.5 27.5 84.5 50.5 66

8 R 400 437 25 32 90.5 80 42 54

6 7 8 9 R R R R 229 235 230 250 86.5 89 86 92 137 125.5 132.5 140 31.5 31 31 30 34 32 32 33 102 99 99 101.5 59 58 58.5 61 75 74 74 77

9 R 417 456 23.5 29 83.5 77.5 43 56

10 11 12 R L R 448 420 409 488 460 448 27 24 25 30 26 26 91.5 81 82 83.5 71.5 75.5 43 45.5 45 55 50 52

13 R 409 442 23 27 81.5 75.5 45.5 53

10 R 233 92.5 127.5 32 33 100.5 59.5 75

14 R 419 455 23 26 78.5 71.5 47 53

15 L 382 419 23.5 25 78.5 67 45.5 50

Appendix IV: Basic Statistics Basic descriptive statistics on the primary data collection. Human: HUMAN HUMERUS Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Head circumference Epicondylar breadth Head diameter FEMUR Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth TIBIA Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

Mean 303.22 94.1 57.83 20.33 18.88 121.12 60.76 42.92

Stand dev 19.913 7.464 5.013 1.937 2.142 24.298 4.935 3.309

Variance 396.542 55.704 25.129 3.751 4.587 590.404 24.349 10.953

429.52 130.6 97.12 27.32 26.16 81.97 68.09 77.78

30.531 10.237 8.307 2.945 2.103 6.480 5.279 6.322

932.132 104.806 69.006 8.671 4.423 41.994 27.864 39.971

350.98 358.62 28.24 20.17 71.38 68.73 35.92 50.04

24.263 24.678 2.952 2.385 14.628 5.979 3.189 4.175

588.673 609.016 8.717 5.690 213.985 35.747 10.167 17.427

109

Kangaroo: KANGAROO HUMERUS Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Head circumference Epicondylar breadth Head diameter FEMUR Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth TIBIA Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

Mean 206.8 101.25 60.9 21.75 23.8 94.95 60.4 27.85

Stand dev 19.736 4.820 5.656 2.731 3.645 6.461 2.716 2.199

Variance 389.511 23.236 31.989 7.458 13.289 41.747 7.378 4.836

263.15 88.625 96.05 26.65 24.2 80.275 54.875 58.8

33.030 11.715 12.644 3.846 3.172 10.783 7.147 7.281

1090.976 137.234 159.866 14.792 10.063 116.276 51.076 53.011

440.3 447.8 22.05 22.95 69.2 61.45 26.2 37.6

58.386 58.700 3.312 3.940 10.896 9.100 4.050 5.680

3408.900 3445.733 10.969 15.525 118.733 82.803 16.400 32.267

110

Sheep: SHEEP HUMERUS Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Head circumference Epicondylar breadth Head diameter FEMUR Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth TIBIA Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

Mean 159.4 101.15 63.3 22.45 18.9 103.85 41.2 31.05

Stand dev 9.046 8.459 4.405 2.466 2.158 9.615 3.882 2.166

Variance 81.822 71.558 19.400 6.081 4.656 92.447 15.067 4.692

187.2 75.85 77.75 21.4 20.45 66.95 39.25 46.8

13.265 5.912 5.619 1.745 1.833 5.510 2.711 2.974

175.956 34.947 31.569 3.044 3.358 30.358 7.347 8.844

220.3 224.8 15.85 17.65 55.7 55.1 24.65 35

11.206 11.497 1.375 1.203 3.474 4.408 2.028 2.494

125.567 132.178 1.892 1.447 12.067 19.433 4.114 6.222

111

Pig: PIG HUMERUS Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Head circumference Epicondylar breadth Head diameter FEMUR Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth TIBIA Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

Mean 162.8 111.9 59.55 16.95 21.8 91.95 43.4 27.2

Stand dev 27.157 11.692 7.837 2.374 3.466 6.829 3.438 2.452

Variance 737.511 136.711 61.414 5.636 12.011 46.636 11.822 6.011

183.9 77.25 85.05 20.35 17 60.05 41.35 45.7

23.506 4.498 8.506 2.604 1.856 7.632 3.465 3.592

552.544 20.236 72.358 6.781 3.444 58.247 12.003 12.900

159.3 164.8 16.65 18.05 55.9 51.45 26.3 31.3

22.430 23.313 2.274 2.127 7.268 5.881 1.549 2.058

503.122 543.511 5.169 4.525 52.822 34.581 2.400 4.233

112

Dog: DOG HUMERUS Maximum length Surgical neck circumference Shaft least circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Head circumference Epicondylar breadth Head diameter FEMUR Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth TIBIA Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

Mean 167.8 99 54 16.95 19.9 86.05 43.2 25.9

Stand dev 33.990 11.780 8.954 2.823 3.755 11.174 4.566 2.436

Variance 1155.289 138.778 80.167 7.969 14.100 124.858 20.844 5.933

185.1 69.35 72.1 17.9 16.8 54.75 36.25 42.8

32.426 6.360 12.425 2.685 2.606 8.502 4.739 5.770

1051.433 40.447 154.378 7.211 6.789 72.292 22.458 33.289

185.2 187.9 15.85 17.55 49.8 50.2 20.95 30.1

36.456 37.183 1.454 0.985 7.790 7.091 3.500 4.581

1329.067 1382.544 2.114 0.969 60.678 50.289 12.247 20.989

113

Emu: EMU FEMUR Maximum length Femoral head circumference Subtrochanteric circumference Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Condylar breadth Transverse epincondylar breadth TIBIA Condylo-malleolar length Spino-malleolar length Midshaft antero-posterior diameter Midshaft medio-lateral diameter Midshaft circumference Shaft least circumference Malleolus antero-posterior diameter Maximum distal epiphyseal breadth

Mean 236.7 88.55 132.25 30.65 32.25 98.8 58.7 74.3

Stand dev 7.424 6.247 6.079 1.582 2.138 6.088 3.599 4.218

Variance 55.122 39.025 36.958 2.503 4.569 37.067 12.956 17.789

415.73 453.67 24.47 28.2 84.43 76.33 44.73 54.4

15.126 15.674 1.433 2.506 5.233 5.457 2.389 3.888

228.781 245.667 2.052 6.279 27.388 29.774 5.710 15.114

114

Appendix V: Midshaft Measurements Mean and standard deviations calculated for the midshaft measurements (ap, mld and mc) taken at the five levels (1cm and 2cm above and below known midshaft) on each long bone. Humeri: Bone number 2cm above known ap 1cm above known ap known ap midshaft H 1cm below known ap U 2cm below known ap

M MEAN A STANDARD DEV N 2cm above known ml 1cm above known ml known ml midshaft 1cm below known ml 2cm below known ml MEAN STANDARD DEV

1 2 3 4 24 18 20 21 23 18 20 21 22 18 20 21 23 18 20 21 23 18 19 20 23 18 19.8 20.8 0.707 0.000 0.447 0.447 21.5 18 18 18.5 21.5 18 17 18 20.5 18 17 18 20 17 17 17.5 20 17 17 17 20.7 17.6 17.2 17.8 0.758 0.548 0.447 0.570

5 22 22 22 21 21 21.6 0.548 21 21 21 20 20 20.6 0.548

Bone number 1 2 2cm above known ap 27.5 20 1cm above known ap 27.5 20.5 K known ap midshaft 25 21 A 1cm below known ap 24 21 N 2cm below known ap 21 18 G MEAN 25 20.1 A STANDARD DEV 2.716 1.245 R 30 23 O 2cm above known ml 26 20 O 1cm above known ml known ml midshaft 23 18.5 1cm below known ml 21 18 2cm below known ml 20 18 MEAN 24 19.5 STANDARD DEV 3.868 2.121

6 18.5 19.5 20 20 19 19.4 0.652 18.5 18.5 18 18 17 18 0.612

7 20 19 19 19 19.5 19.3 0.447 20.5 20.5 20 19 18.5 19.7 0.908

3 22 22 22 21 18 21 1.732 27.5 22 19 18 18 20.9 4.037

8 19.5 19 18 18.5 19.5 18.9 0.652 20 20 19 18 17 18.8 1.304

4 28 29 29 25 24 27 2.345 31 30 29 29 22.5 28.3 3.347

115

9 24 24 24 24.5 23.5 24 0.354 23 23 23 22 21.5 22.5 0.707

10 20 20 20 19.5 19 19.7 0.447 21 21 20.5 20 19 20.3 0.837

5 28.5 29 28 26 21.5 26.6 3.070 30 27 24 22 20.5 24.7 3.834

11 20 20 19 19 19 19.4 0.548 20 21 19.5 18 18 19.3 1.304

6 23 23 23 22 19 22 1.732 25.5 23 20.5 19 18 21.2 3.054

12 18.5 18.5 18.5 19 18.5 18.6 0.224 18 17.5 16 16 15.5 16.6 1.084

13 18 18 18 18 17.5 17.9 0.224 15.5 15.5 14.5 14 14 14.7 0.758

7 21.5 21 22 21 18 20.7 1.565 21.5 22 19.5 18.5 18 19.9 1.782

14 22 22 21 22 22 21.8 0.447 21.5 21.5 20 21 21.5 21.1 0.652

8 26 26 25 25.5 22 24.9 1.673 30 28 25 22 22 25.4 3.578

15 19.5 19.5 19.5 19.5 19.5 19.5 0.000 20 20 19 19 18 19.2 0.837

16 20 19 19.5 19 19 19.3 0.447 19.5 18 18 16.5 16 17.6 1.387

9 22 23 23 22 18 21.6 2.074 28.5 22.5 19.5 19.5 17 21.4 4.422

17 17.5 18.5 18.5 18.5 19 18.4 0.548 21.5 20.5 20 19 18 19.8 1.351

10 28 28 29.5 25 22 26.5 3.000 33 28 25 23 22 26.2 4.438

Bone number 1 2 2cm above known ap 22 24 1cm above known ap 19.5 22 known ap midshaft 19 20 20 S 1cm below known ap 17.5 2cm below known ap 18 19.5 H E MEAN 19.2 21.1 E STANDARD DEV 1.754 1.884 P 2cm above known ml 17.5 17.5 1cm above known ml 16 17 known ml midshaft 15.5 16.5 1cm below known ml 15.5 16.5 2cm below known ml 15.5 17 MEAN 16 16.9 STANDARD DEV 0.866 0.418

3 25.5 22.5 22 21 20 22.2 2.080 20.5 19 18 18 18 18.7 1.095

4 30.5 27 24 22 21 24.9 3.879 22 21 20 19 19 20.2 1.304

5 26.5 23 23 21 21 22.9 2.247 25 23 21.5 20.5 21 22.2 1.823

6 31 26 25 23 21 25.2 3.768 22 21.5 21 20 20 20.9 0.894

7 23 19 18.5 18 18 19.3 2.110 21 19 17 16.5 17 18.1 1.884

8 23 21.5 20.5 21 21 21.4 0.962 25 23 21 20 20 21.8 2.168

9 29 26 23 22 21 24.2 3.271 24 22 21 20 20 21.4 1.673

10 29 26.5 23 21.5 20 24 3.691 20.5 19 19 18 18 18.9 1.025

Bone number 1 2 2cm above known ap 28.5 28 1cm above known ap 27.5 26 known ap midshaft 26 26 1cm below known ap 25.5 25.5 26 28 P 2cm below known ap I MEAN 26.7 26.7 G STANDARD DEV 1.255 1.204 2cm above known ml 21.5 24.5 1cm above known ml 20 23.5 known ml midshaft 18 21.5 1cm below known ml 17 20.5 2cm below known ml 17 19 MEAN 18.7 21.8 STANDARD DEV 1.987 2.225

3 29 25 24.5 23 23 24.9 2.460 22.5 20.5 17.5 17 16 18.7 2.706

4 25.5 24.5 24.5 23 23 24.1 1.084 25 21 18 18 16 19.6 3.507

5 27 25.5 24 23 25 24.9 1.517 21 19 18 17 16 18.2 1.924

6 24 21 20 19 19 20.6 2.074 19.5 16.5 15 14 15 16 2.151

7 26 24 22.5 21 21 22.9 2.133 23 20 16 16.5 15 18.1 3.324

8 25 22 20 20 19 21.2 2.387 18 16 16 16 17.5 16.7 0.975

9 22 20 18.5 18 19 19.5 1.581 17 15 15 17.5 18.5 16.6 1.557

10 25 21 18 17 18 19.8 3.271 18 14 13 13 14 14.4 2.074

Bone number 1 2 2cm above known ap 28 27 1cm above known ap 28 26.5 known ap midshaft 27 24 1cm below known ap 25 22 2cm below known ap 24 21 D O MEAN 26.4 24.1 G STANDARD DEV 1.817 2.655 2cm above known ml 26 20 1cm above known ml 23 20 known ml midshaft 22.5 20.5 1cm below known ml 20 20 2cm below known ml 20 20 MEAN 22.3 20.1 STANDARD DEV 2.490 0.224

3 27 26 19 22 20 22.8 3.564 17 16.5 16 16.5 16.5 16.5 0.354

4 23.5 21.5 20 20 18 20.6 2.043 17 17 16.5 17 17 16.9 0.224

5 25.5 23.5 23 21.5 20 22.7 2.080 19 18 18 18 18 18.2 0.447

6 25 22.5 21.5 19.5 19 21.5 2.424 18 18 18 18 18 18 0.000

7 21.5 20.5 19.5 17.5 15.5 18.9 2.408 16 15 15 15 15 15.2 0.447

8 22 21 19 17 16.5 19.1 2.408 16 16 16 15 15.5 15.7 0.447

9 20 18 17.5 14 14 16.7 2.636 15 13 13 13 13 13.4 0.894

10 18 17 17.5 18 17.5 17.6 0.418 18 18 17.5 18 17.5 17.8 0.274

116

Femora: Bone number 1 2 2cm above known ap 28.5 28.5 1cm above known ap 28 29 known ap midshaft 28 28.5 1cm below known ap 28 28.5 28 29.5 H 2cm below known ap 28.1 28.8 U MEAN M STANDARD DEV 0.224 0.447 A 2cm above known ml 23.5 26 N 1cm above known ml 23 25 known ml midshaft 23.5 25 1cm below known ml 23.5 25 2cm below known ml 23.5 24.5 MEAN 23.4 25.1 STANDARD DEV 0.224 0.548

3 29 29.5 30 29.5 29.5 29.5 0.354 27 27 28 28 28 27.6 0.548

4 28 28 28.5 28 28 28.1 0.224 25.5 25 25.5 25.5 26 25.5 0.354

5 25 25 25.5 25 25 25.1 0.224 24 23 23 23.5 24 23.5 0.500

6 26 26 26.5 26 26 26.1 0.224 28 28 28 28 28 28 0.000

7 8 9 27.5 27 29 27 27 29 27.5 27.5 29.5 26.5 27 29.5 26 27 30 26.9 27.1 29.4 0.652 0.224 0.418 27 24.5 24 26 24.5 24 26 25 24 25.5 25 24 26 25 24 26.1 24.8 24 0.548 0.274 0.000

4 30 30 31 31.5 31 30.7 0.671 29 29 28 28.5 28 28.5 0.500

10 30 29.5 30.5 31 30.5 30.3 0.570 25.5 25.5 25 24.5 25 25.1 0.418

6 27.5 27.5 28 28 27.5 27.7 0.274 25 24.5 24.5 24.5 24.5 24.6 0.224

12 22 22 22 22 22 22 0.000 26 25 24.5 24 24 24.7 0.837

13 23.5 23.5 23.5 24 24 23.7 0.274 24 24 24 24 24 24 0.000

14 24 24.5 24 24 23.5 24 0.354 28 27.5 27.5 27.5 27.5 27.6 0.224

15 27 27 28 28 28 27.6 0.548 27 26.5 26 26 26 26.3 0.447

16 26.5 27 27 26.5 26 26.6 0.418 25 25 24.5 24.5 24.5 24.7 0.274

17 25 23 23 23.5 23 23.5 0.866 29 29 29 28.5 28 28.7 0.447

Bone number 1 2cm above known ap 32 1cm above known ap 32 known ap midshaft 32 K 1cm below known ap 31.5 A 2cm below known ap 31.5 N MEAN 31.8 G STANDARD DEV 0.274 A 29 R 2cm above known ml 1cm above known ml 29 O 29.5 O known ml midshaft 1cm below known ml 28 2cm below known ml 28 MEAN 28.7 STANDARD DEV 0.671

2 26 26.5 26.5 26 26 26.2 0.274 24 24 24 24 24 24 0.000

3 29 30 30 30 28 29.4 0.894 26.5 26 25.5 25 24.5 25.5 0.791

7 8 9 27 29 31 27 28 31 26.5 27.5 32 27 28 31.5 27 28 31 26.9 28.1 31.3 0.224 0.548 0.447 27 28 29 27 27.5 28.5 26.5 27.5 28 27 27 28 26.5 27 27.5 26.8 27.4 28.2 0.274 0.418 0.570

10 27.5 27.5 28 27.5 27 27.5 0.354 27.5 27.5 27 27 26.5 27.1 0.418

Bone number 1 2cm above known ap 18 1cm above known ap 18 known ap midshaft 18 1cm below known ap 18.5 S 2cm below known ap 18.5 18.2 H MEAN E STANDARD DEV 0.274 E 2cm above known ml 18 P 1cm above known ml 18 known ml midshaft 18 1cm below known ml 18 2cm below known ml 18 MEAN 18 STANDARD DEV 0.000

2 19 19 19 20 20 19.4 0.548 19 19 19 19 20 19.2 0.447

3 4 5 6 7 8 9 22 23 20.5 21.5 22.5 22.5 20.5 21 23 20.5 21.5 23 22 20 21 23 21 22 23 22.5 20.5 21 24 21 22 24.5 23 21 21 24 21 22 25 23 21 21.2 23.4 20.8 21.8 23.6 22.6 20.6 0.447 0.548 0.274 0.274 1.084 0.418 0.418 20 23 17.5 20.5 20 22.5 21 20 23 17 20.5 20 22.5 21 19 23 17 21 20.5 22.5 21 19 23 17.5 21.5 21 23 21.5 19 23 17.5 22.5 23 72 22.5 19.4 23 17.3 21.2 20.9 32.5 21.4 0.548 0.000 0.274 0.837 1.245 22.082 0.652

10 21 21 21 21 21 21 0.000 21 21 21.5 22 22.5 21.6 0.652

117

5 26.5 27 27 26.5 26 26.6 0.418 28 27.5 27 27 26.5 27.2 0.570

11 27.5 27.5 28 28 28 27.8 0.274 30.5 30.5 30.5 30 30.5 30.4 0.224

Bone number 2cm above known ap 1cm above known ap known ap midshaft 1cm below known ap 2cm below known ap P MEAN I STANDARD DEV G 2cm above known ml 1cm above known ml known ml midshaft 1cm below known ml 2cm below known ml MEAN STANDARD DEV Bone number 2cm above known ap 1cm above known ap known ap midshaft 1cm below known ap 2cm below known ap D MEAN O STANDARD DEV G 2cm above known ml 1cm above known ml known ml midshaft 1cm below known ml 2cm below known ml MEAN STANDARD DEV Bone number 2cm above known ap 1cm above known ap known ap midshaft 1cm below known ap 2cm below known ap E MEAN M STANDARD DEV U 2cm above known ml 1cm above known ml known ml midshaft 1cm below known ml 2cm below known ml MEAN STANDARD DEV

1 24 24 24 24 24 24 0.000 20 21 21 20 21 20.6 0.548

2 22 21 21.5 21 22 21.5 0.500 17.5 17 17 17 18 17.3 0.447

3 23 22 22 22 22.5 22.3 0.447 18 17.5 17.5 18 20 18.2 1.037

4 23 23 22 22 22.5 22.5 0.500 20 20.5 20.5 21 21 20.6 0.418

6 21 20.5 20.5 21 21 20.8 0.274 17 16 16 16.5 17 16.5 0.500

7 8 9 20.5 19 17 20 18 16 20 17.5 16 19 17.5 16 20 18 17 19.9 18 16.4 0.548 0.612 0.548 18 15.5 15.5 17 15 14 17 15 14 18 16 15 19 17 15.5 17.8 15.7 14.8 0.837 0.837 0.758

10 18 16.5 17 17 17 17.1 0.548 15 14 14 14 14 14.2 0.447

1 19 19 19 19.5 20 19.3 0.447 18 17 17.5 18 19 17.9 0.742

2 20 20 21.5 21.5 22 21 0.935 20 20 21 21 21 20.6 0.548

3 4 5 6 20 19 18 16.5 20 19 18 17 21 19 19 17 21.5 19 19 17 22 19 19 17 20.9 19 18.6 16.9 0.894 0.000 0.548 0.224 19 17 17 16 19 16 16.5 16 19 17 16.5 16 20 17 17 16.5 21 18 18 17 19.6 17 17 16.3 0.894 0.707 0.612 0.447

7 8 9 15 16 13 15.5 16.5 13 16 17 13 16 17 13.5 16 16.5 14 15.7 16.6 13.3 0.447 0.418 0.447 15 16 13 15 15.5 13 15 15.5 13 16 16 13.5 17 17.5 14 15.6 16.1 13.3 0.894 0.822 0.447

10 18 18 18 18.5 19 18.3 0.447 17 17 18 18.5 19.5 18 1.061

1 32 32.5 32.5 32 32.5 32.3 0.274 33 33.5 33.5 34 35 33.8 0.758

2 32 30 30 30 30 30.4 0.894 31.5 31.5 30.5 31.5 32.5 31.5 0.707

3 32.5 33 34 33 33.5 33.2 0.570 34 35 35 35 35.5 34.9 0.548

7 8 9 29.5 31 30 30 31 30 31 32 30 31 31 30.5 31 30 31 30.5 31 30.3 0.707 0.707 0.447 31 31.5 33 31 31.5 33 31.5 32 33 32 32 34 32.5 33 34 31.6 32 33.4 0.652 0.612 0.548

10 32 32 32.5 31.5 31.5 31.9 0.418 32.5 31.5 31.5 32 34 32.3 1.037

4 31 30 30.5 31 31 30.7 0.447 30.5 30 29.5 29.5 30 29.9 0.418

118

5 22 21.5 21.5 21 21 21.4 0.418 17 17 17 17 18 17.2 0.447

5 26 26.5 27.5 27 27 26.8 0.570 28 27 27.5 27 27.5 27.4 0.418

6 32 32 32 31 31 31.6 0.548 34 34 33 33 35 33.8 0.837

Tibiae: Bone number 1 2 2cm above known ap 29.5 30.5 1cm above known ap 29 30 known ap midshaft 28 29 1cm below known ap 28 28.5 H 2cm below known ap 27.5 28.5 28.4 29.3 U MEAN M STANDARD DEV 0.822 0.908 A 2cm above known mc 78 79 N 1cm above known mc 76.5 78 known mc midshaft 76.5 77 1cm below known mc 74.5 75 2cm below known mc 73 73.5 MEAN 75.7 76.5 STANDARD DEV 1.956 2.236

3 28.5 28.5 28 27.5 26.5 27.8 0.837 78.5 78 77 76 73 76.5 2.179

4 29 28 28 28 26.5 27.9 0.894 76 75.5 73.5 73 71 73.8 2.019

5 29 28 26 25.5 25 26.7 1.718 73 72 71 69 68 70.6 2.074

6 30 30 29 28.5 28 29.1 0.894 81 79 79 78.5 76.5 78.8 1.605

7 8 9 10 11 28 35 30 32.5 30.5 28 34 28.5 31 30.5 28 34 27 32 29.5 27 33 27 31 29 26 33 26 31 27.5 27.4 33.8 27.7 31.5 29.4 0.894 0.837 1.565 0.707 1.245 77 89 79 83 85 74 88 77 82 83.5 73 88 74.5 80.5 81.5 73 88 73.5 79.5 81.5 71.5 87 72.5 79 79.5 73.7 88 75.3 80.8 82.2 2.049 0.707 2.660 1.681 2.110

12 24 23 23 23 22.5 23.1 0.548 65.5 65 64 63 62.5 64 1.275

13 25 24 24 23 22.5 23.7 0.975 68 67.5 66 65 63.5 66 1.837

14 30 29.5 28 28 28 28.7 0.975 79.5 79.5 77.5 77.5 75 77.8 1.857

15 28 28 28 26 26 27.2 1.095 77 75 74.5 73.5 72.5 74.5 1.696

16 28 26.5 25 25 23 25.5 1.871 70.5 69.5 70 68 67 69 1.458

17 31.5 30.5 31 30 29.5 30.5 0.791 79 79 78.5 78 76 78.1 1.245

Bone number 1 2 2cm above known ap 24 25.5 1cm above known ap 23.5 24 known ap midshaft 23.5 23 K 1cm below known ap 23 23 A 2cm below known ap 23 23 N MEAN 23.4 23.7 G STANDARD DEV 0.418 1.095 A 84 81 R 2cm above known mc 1cm above known mc 82 80 O 79.5 78.5 O known mc midshaft 1cm below known mc 78 78 2cm below known mc 76 76.5 MEAN 79.9 78.8 STANDARD DEV 3.170 1.754

3 28 26 27.5 26.5 26 26.8 0.908 95.5 91.5 89.5 87 84 89.5 4.373

4 26.5 26 26 25 25 25.7 0.671 83 79.5 76.5 75 74 77.6 3.664

5 21 21 20 19.5 19 20.1 0.894 64.5 62.5 58.5 56.5 56 59.6 3.748

6 21 20 19.5 20 19 19.9 0.742 65 61.5 60 58 58 60.5 2.915

7 22 21 21 21 20.5 21.1 0.548 67 65.5 64.5 63 62.5 64.5 1.837

8 23 22 22 21 20 21.6 1.140 71.5 69 67 64.5 61.5 66.7 3.883

9 20.5 20 19 19 18.5 19.4 0.822 63.5 62 60.5 60 56.5 60.5 2.622

10 20 20 19.5 19 18.5 19.4 0.652 64 62.5 60.5 57 57 60.2 3.174

Bone number 1 2 2cm above known ap 15 17 1cm above known ap 14 16.5 known ap midshaft 13.5 15.5 1cm below known ap 13 15 2cm below known ap 13 14 S 13.7 15.6 H MEAN E STANDARD DEV 0.837 1.194 E 2cm above known mc 50 53.5 P 1cm above known mc 48 52.5 known mc midshaft 47.5 51 1cm below known mc 48 49.5 2cm below known mc 46.5 49 MEAN 48 51.1 STANDARD DEV 1.275 1.917

3 20.5 19 17 18 16 18.1 1.746 64.5 62 61 59 57.5 60.8 2.706

4 19 18 17 17 16 17.4 1.140 62.5 59 57.5 57.5 57.5 58.8 2.168

5 17 16.5 16 15.5 14.5 15.9 0.962 57 55 53 52.5 52.5 54 1.969

6 16 16 15.5 15 15 15.5 0.500 54 52.5 52 51 50 51.9 1.517

7 18 17 16 16 15.5 16.5 1.000 60 56.5 55 55.5 55 56.4 2.104

8 20 19 18 17 17 18.2 1.415 64.5 61.5 60.5 60 59 61.1 2.104

9 18 17 16 16 15.5 16.5 1.000 58.5 55.5 55 54.5 54.5 55.6 1.673

10 19 18 17 16.5 15.5 17.2 1.351 60 57.5 55 55 52.5 56 2.850

119

Bone number 2cm above known ap 1cm above known ap known ap midshaft 1cm below known ap 2cm below known ap P MEAN I STANDARD DEV G 2cm above known mc 1cm above known mc known mc midshaft 1cm below known mc 2cm below known mc MEAN STANDARD DEV Bone number 2cm above known ap 1cm above known ap known ap midshaft 1cm below known ap 2cm below known ap D MEAN O STANDARD DEV G 2cm above known mc 1cm above known mc known mc midshaft 1cm below known mc 2cm below known mc MEAN STANDARD DEV Bone number 2cm above known ap 1cm above known ap known ap midshaft 1cm below known ap 2cm below known ap E MEAN M STANDARD DEV U 2cm above known mc 1cm above known mc known mc midshaft 1cm below known mc 2cm below known mc MEAN STANDARD DEV

1 25 22.5 19.5 18 16.5 20.3 3.439 74 69 66 63 61 66.6 5.128

2 21 18.5 17 16 14.5 17.4 2.485 71 66 63 60 57 63.4 5.413

3 22 19 18 16 15 18 2.739 68.5 63 61.5 59 57 61.8 4.396

4 24.5 21 18.5 16 15 19 3.857 70.5 64.5 59.5 56.5 54.5 61.1 6.465

5 22.5 18.5 17 15 14.5 17.5 3.221 67.5 62 59.5 57 54.5 60.1 4.992

6 23 20 17.5 16 15 18.3 3.233 69 63 60 57 55 60.8 5.495

9 22 16 13 11.5 11.5 14.8 4.424 62 50.5 46 45 44 49.5 7.416

10 21 17.5 15 13 12 15.7 3.633 63.5 51.5 46.5 44 43.5 49.8 8.289

1 18 17.5 17 16.5 16 17 0.791 59 58.5 56.5 56 56 57.2 1.440

2 20.5 20 19 20 20 19.9 0.548 62 61 60 60 60 60.6 0.894

3 18.5 18 17 17 17 17.5 0.707 61 59.5 59.5 59 59.5 59.7 0.758

4 17.5 16 15.5 15.5 15 15.9 0.962 55.5 53.5 53 53.5 53.5 53.8 0.975

5 17.5 17 16 16 16.5 16.6 0.652 44 41 40 40 39 40.8 1.924

6 7 8 9 13 16 16 18 12 15 15 17 11 13 14 16.5 11 13 14 16.5 11.5 13 13 16 11.7 14 14.4 16.8 0.837 1.414 1.140 0.758 41 51 52 59 39 48 50 55.5 37.5 45 48 53.5 37.5 47 48 52.5 37 47 48 52.5 38.4 47.6 49.2 54.6 1.636 2.191 1.789 2.748

10 17 16 15.5 15.5 15 15.8 0.758 53.5 52 55 56.5 54.5 54.3 1.681

1 25 25 24 25 25 24.8 0.447 84.5 85 85.5 86 87 85.6 0.962

2 24 24 24 24.5 25 24.3 0.447 82.5 83.5 84.5 84.5 84.5 83.9 0.894

3 26 26 26.5 26 26 26.1 0.224 87 87 87 87 86 86.8 0.447

4 26 26 25.5 26 25.5 25.8 0.274 91 91 89.5 91 91 90.7 0.671

5 26 26 25.5 25 25 25.5 0.500 90 89 89.5 90 90 89.7 0.447

6 25 25.5 25 24.5 24 24.8 0.570 89 90 89 89 89 89.2 0.447

120

7 21 17 15 13 11.5 15.5 3.708 60.5 54.5 49 47 45.5 51.3 6.170

8 24 20 17.5 14.5 13 17.8 4.396 70 64 58 55 53.5 60.1 6.841

7 8 9 10 22 25 24 26.5 22 25 24 27 22 25 23.5 26.5 22 25 24 26 21 25 23 26.5 21.8 25 23.7 26.5 0.447 0.000 0.447 0.354 71.5 90 83 91.5 73 90 83 91.5 74 90 84.5 91.5 74 90 84.5 91.5 74 90 84 91.5 73.3 90 83.8 91.5 1.095 0.000 0.758 0.000

Appendix VI: Group Classification using Midshaft Measurements Human and non-human group classification calculated from original discriminant functions using the average and maximum value (1 to 2cm above and below known midshaft) for the humerus, femur and tibia.

Mean Value Above and Below Known Midshaft:

HUMAN KANGAROO SHEEP

Individual 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6

Mean Value 23 18 19.8 20.8 21.6 19.4 19.3 18.9 24 19.7 19.4 18.6 17.9 21.8 19.5 19.3 18.4 25 20.1 21 27 26.6 22 20.7 24.9 21.6 26.5 19.2 21.1 22.2 24.9 22.9 25.2

HUMERUS † *ap (H/A) 1.085 H -0.695 A -0.054 A 0.302 H 0.587 H -0.197 A -0.232 A -0.375 A 1.441 H -0.090 A -0.197 A -0.481 A -0.731 A 0.658 H -0.161 A -0.232 A -0.553 A 1.797 H 0.053 H 0.373 H 2.509 H 2.367 H 0.729 H 0.266 H 1.761 H 0.587 H 2.331 H -0.268 A 0.409 H 0.800 H 1.761 H 1.049 H 1.868 H 121

Mean Value 20.7 17.6 17.2 17.8 20.6 18 19.7 18.8 22.5 20.3 19.3 16.6 14.7 21.1 19.2 17.6 19.8 24 19.5 20.9 28.3 24.7 21.2 19.9 25.4 21.4 26.2 16 16.9 18.7 20.2 22.2 20.9

*mld 0.286 -0.768 -0.904 -0.7 0.252 -0.632 -0.054 -0.36 0.898 0.15 -0.19 -1.108 -1.754 0.422 -0.224 -0.768 -0.02 1.408 -0.122 0.354 2.87 1.646 0.456 0.014 1.884 0.524 2.156 -1.312 -1.006 -0.394 0.116 0.796 0.354

†

(H/A) A H H H A H H H A A H H H A H H H A H A A A A H A A A H H H A A A

PIG DOG

7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9

19.3 21.4 24.2 24 26.7 26.7 24.9 24.1 24.9 20.6 22.9 21.2 19.5 19.8 26.4 24.1 22.8 20.6 22.7 21.5 18.9 19.1 16.7

-0.232 0.515 1.512 1.441 2.402 2.402 1.761 1.477 1.761 0.231 1.049 0.444 -0.161 -0.054 2.295 1.477 1.014 0.231 0.978 0.551 -0.375 -0.303 -1.158

A H H H H H H H H H H H A A H H H H H H A A A

18.1 21.8 21.4 18.9 18.7 21.8 18.7 19.6 18.2 16 18.1 16.7 16.6 14.4 22.3 20.1 16.5 16.9 18.2 18 15.2 15.7 13.4

-0.598 0.66 0.524 -0.326 -0.394 0.66 -0.394 -0.088 -0.564 -1.312 -0.598 -1.074 -1.108 -1.856 0.83 0.082 -1.142 -1.006 -0.564 -0.632 -1.584 -1.414 -2.196

H A A H H A H H H H H H H H A A H H H H H H H

10

17.6

-0.837

A

17.8

-0.7

H

*

ap: midshaft antero-posterior diameter, mld: midshaft medio-lateral diameter, mc: midshaft circumference. †

H: Human, A: Non-human

HUMAN

Individual 1 2 3 4 5 6 7 8 9 10 11 12

Mean Value 28.1 28.8 29.5 28.1 25.1 26.1 26.9 27.1 29.4 30.3 27.8 22

FEMUR † *ap (H/A) 0.598 H 0.759 H 0.92 H 0.598 H -0.092 A 0.138 H 0.322 H 0.368 H 0.897 H 1.104 H 0.529 H -0.805 A 122

Mean Value 23.4 25.1 27.6 25.5 23.5 28 26.1 24.8 24 25.1 30.4 24.7

*mld -0.168 0.204 0.751 0.292 -0.147 0.839 0.423 0.138 -0.037 0.204 1.365 0.116

†

(H/A) A H H H A H H H A H H H

KANGAROO SHEEP PIG DOG

13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

23.7 24 27.6 26.6 23.5 31.8 26.2 29.4 30.7 26.6 27.7 26.9 28.1 31.3 27.5 18.2 19.4 21.2 23.4 20.8 21.8 23.6 22.6 20.6 21 24 21.5 22.3 22.5 21.4 20.8 19.9 18 16.4 17.1 19.3 21 20.9 19 18.6 16.9 15.7 16.6 13.3 18.3

-0.414 -0.345 0.483 0.253 -0.46 1.449 0.161 0.897 1.196 0.253 0.506 0.322 0.598 1.334 0.46 -1.679 -1.403 -0.989 -0.483 -1.081 -0.851 -0.437 -0.667 -1.127 -1.035 -0.345 -0.92 -0.736 -0.69 -0.943 -1.081 -1.288 -1.725 -2.093 -1.932 -1.426 -1.035 -1.058 -1.495 -1.587 -1.978 -2.254 -2.047 -2.806 -1.656

A A H H A H H H H H H H H H H A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A 123

24 27.6 26.3 24.7 28.7 28.7 24 25.5 28.5 27.2 24.6 26.8 27.4 28.2 27.1 18 19.2 19.4 23 17.3 21.2 20.9 22.6 21.4 21.6 20.6 17.3 18.2 20.6 17.2 16.5 17.8 15.7 14.8 14.2 17.9 20.6 19.6 17 17 16.3 15.6 16.1 13.3 18

-0.037 0.751 0.467 0.116 0.992 0.992 -0.037 0.292 0.949 0.664 0.094 0.576 0.708 0.883 0.642 -1.351 -1.088 -1.044 -0.256 -1.504 -0.650 -0.716 -0.344 -0.606 -0.563 -0.782 -1.504 -1.307 -0.782 -1.526 -1.680 -1.395 -1.855 -2.052 -2.183 -1.373 -0.782 -1.001 -1.570 -1.570 -1.723 -1.877 -1.767 -2.380 -1.351

A H H H H H A H H H H H H H H A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A

EMU HUMAN KANGAROO SHEEP

1 2 3 4 5 6 7 8 9

32.3 30.4 33.2 30.7 26.8 31.6 30.5 31 30.3

1.564 1.127 1.771 1.196 0.299 1.403 1.15 1.265 1.104

H H H H H H H H H

33.8 31.5 34.9 29.9 27.4 33.8 31.6 32 33.4

2.109 1.606 2.350 1.255 0.708 2.109 1.627 1.715 2.022

H H H H H H H H H

10

31.9

1.472

H

32.3

1.781

H

Mean Value 75.7 76.5 76.5 73.8 70.6 78.8 73.7 88 75.3 80.8 82.2 64 66 77.8 74.5 69 78.1 79.9 78.8 89.5 77.6 59.6 60.5 64.5 66.7 60.5 60.2 48 51.1 60.8

*mc 0.512 0.566 0.566 0.385 0.170 0.720 0.378 1.336 0.485 0.854 0.947 -0.272 -0.138 0.653 0.432 0.063 0.673 0.793 0.720 1.437 0.639 -0.567 -0.506 -0.238 -0.091 -0.506 -0.527 -1.344 -1.136 -0.486

Individual Mean Value 28.4 1 29.3 2 27.8 3 27.9 4 26.7 5 29.1 6 27.4 7 33.8 8 27.7 9 31.5 10 29.4 11 23.1 12 23.7 13 28.7 14 27.2 15 25.5 16 30.5 17 23.4 1 23.7 2 26.8 3 25.7 4 20.1 5 19.9 6 21.1 7 21.6 8 19.4 9 19.4 10 13.7 1 15.6 2 18.1 3

TIBIA † *ap (H/A) 1.271 H 1.513 H 1.109 H 1.136 H 0.813 H 1.459 H 1.002 H 2.723 H 1.082 H 2.105 H 1.540 H -0.155 A 0.006 A 1.351 H 0.948 H 0.491 H 1.836 H -0.074 A 0.006 A 0.840 H 0.544 H -0.962 A -1.016 A -0.693 A -0.559 A -1.150 A -1.150 A -2.684 A -2.173 A -1.500 A 124

†

(H/A) H H H H H H H H H H H A A H H H H H H H H A A A A A A A A A

PIG DOG EMU

4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9

17.4 15.9 15.5 16.5 18.2 16.5 17.2 20.3 17.4 18 19 17.5 18.3 15.5 17.8 14.8 15.7 17 19.9 17.5 15.9 16.6 11.7 14 14.4 16.8 15.8 24.8 24.3 26.1 25.8 25.5 24.8 21.8 25 23.7

-1.688 -2.092 -2.200 -1.931 -1.473 -1.931 -1.742 -0.908 -1.688 -1.527 -1.258 -1.662 -1.446 -2.200 -1.581 -2.388 -2.146 -1.796 -1.016 -1.662 -2.092 -1.904 -3.222 -2.603 -2.495 -1.850 -2.119 0.302 0.168 0.652 0.571 0.491 0.302 -0.505 0.356 0.006

A A A A A A A A A A A A A A A A A A A A A A A A A A A H H H H H H A H A

58.8 54 51.9 56.4 61.1 55.6 56 66.6 63.4 61.8 61.1 60.1 60.8 51.3 60.1 49.5 49.8 57.2 60.6 59.7 53.8 40.8 38.4 47.6 49.2 54.6 54.3 85.6 83.9 86.8 90.7 89.7 89.2 73.3 90 83.8

-0.620 -0.942 -1.083 -0.781 -0.466 -0.835 -0.808 -0.098 -0.312 -0.419 -0.466 -0.533 -0.486 -1.123 -0.533 -1.244 -1.223 -0.728 -0.500 -0.560 -0.955 -1.826 -1.987 -1.371 -1.264 -0.902 -0.922 1.175 1.061 1.256 1.517 1.450 1.416 0.351 1.470 1.055

A A A A A A A A A A A A A A A A A A A A A A A A A A A H H H H H H H H H

10

26.5

0.760

H

91.5

1.571

H

125

Maximum Value above Known Midshaft:

HUMAN KANGAROO SHEEP PIG

Individual 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4

Max. Value 24 18 20 21 22 19.5 20 19.5 24 20 20 18.5 18 22 19.5 20 18.5 27.5 20.5 22 29 29 23 21.5 26 23 28 22 24 25.5 30.5 26.5 31 23 23 29 29 28.5 28 29 25.5

HUMERUS † *ap (H/A) 1.441 H -0.695 A 0.017 H 0.373 H 0.729 H -0.161 A 0.017 H -0.161 A 1.441 H 0.017 H 0.017 H -0.517 A -0.695 A 0.729 H -0.161 A 0.017 H -0.517 A 2.687 H 0.195 H 0.729 H 3.221 H 3.221 H 1.085 H 0.551 H 2.153 H 1.085 H 2.865 H 0.729 H 1.441 H 1.975 H 3.755 H 2.331 H 3.933 H 1.085 H 1.085 H 3.221 H 3.221 H 3.043 H 2.865 H 3.221 H 1.975 H 126

Max. Value 21.5 18 18 18.5 21 18.5 20.5 20 23 21 21 18 15.5 21.5 20 19.5 21.5 30 23 27.5 31 30 25.5 22 30 28.5 33 17.5 17.5 20.5 22 25 22 21 25 24 20.5 21.5 24.5 22.5 25

*mld 0.558 -0.632 -0.632 -0.462 0.388 -0.462 0.218 0.048 1.068 0.388 0.388 -0.632 -1.482 0.558 0.048 -0.122 0.558 3.448 1.068 2.598 3.788 3.448 1.918 0.728 3.448 2.938 4.468 -0.802 -0.802 0.218 0.728 1.748 0.728 0.388 1.748 1.408 0.218 0.558 1.578 0.898 1.748

†

(H/A) A H H H A H A A A A A H H A A H A A A A A A A A A A A H H A A A A A A A A A A A A

DOG

5 6 7 8 9 10 1 2 3 4 5 6 7 8 9

27 24 26 25 22 25 28 27 27 23.5 25.5 25 21.5 22 20

2.509 1.441 2.153 1.797 0.729 1.797 2.865 2.509 2.509 1.263 1.975 1.797 0.551 0.729 0.017

H H H H H H H H H H H H H H H

21 19.5 23 18 17 18 26 20 17 17 19 18 16 16 15

0.388 -0.122 1.068 -0.632 -0.972 -0.632 2.088 0.048 -0.972 -0.972 -0.292 -0.632 -1.312 -1.312 -1.652

A H A H H H A A H H H H H H H

10

18

-0.695

A

18

-0.632

H

*

ap: midshaft antero-posterior diameter, mld: midshaft medio-lateral diameter, mc: midshaft circumference. †

H: Human, A: Non-human

HUMAN ROO KANGA

Individual 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3

Max. Value 28.5 29 29.5 28 25 26 27.5 27 29 30 27.5 22 23.5 24.5 27 27 25 32 26.5 30

FEMUR † *ap (H/A) 0.69 H 0.805 H 0.92 H 0.575 H -0.115 A 0.115 H 0.46 H 0.345 H 0.805 H 1.035 H 0.46 H -0.805 A -0.46 A -0.23 A 0.345 H 0.345 H -0.115 A 1.495 H 0.23 H 1.035 H 127

Max. Value 23.5 26 27 25.5 24 28.0 27.0 24.5 24.0 25.5 30.5 26.0 24.0 28.0 27.0 25.0 29.0 29 24 26.5

*mld -0.147 0.401 0.620 0.292 -0.037 0.839 0.620 0.072 -0.037 0.292 1.387 0.401 -0.037 0.839 0.620 0.182 1.058 1.058 -0.037 0.511

†

(H/A) A H H H A H H H A H H H A H H H H H A H

SHEEP PIG DOG EMU

4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8

30 27 27.5 27 29 31 27.5 18 19 22 23 20.5 21.5 23 22.5 20.5 21 24 22 23 23 22 21 20.5 19 17 18 19 20 20 19 18 17 15.5 16.5 13 18 32.5 32 33 31 26.5 32 30 31

1.035 0.345 0.46 0.345 0.805 1.265 0.46 -1.725 -1.495 -0.805 -0.575 -1.15 -0.92 -0.575 -0.69 -1.15 -1.035 -0.345 -0.805 -0.575 -0.575 -0.805 -1.035 -1.15 -1.495 -1.955 -1.725 -1.495 -1.265 -1.265 -1.495 -1.725 -1.955 -2.3 -2.07 -2.875 -1.725 1.61 1.495 1.725 1.265 0.23 1.495 1.035 1.265

H H H H H H H A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A H H H H H H H H 128

29 28 25.0 27 28 29 27.5 18 19 20 23 17.5 20.5 20.0 22.5 21.0 21.0 21 17.5 18 20.5 17 17 18 15.5 15.5 15 18 20 19 17 17 16 15 16 13 17 33.5 31.5 35 30.5 28 34 31 31.5

1.058 0.839 0.182 0.620 0.839 1.058 0.730 -1.351 -1.132 -0.913 -0.256 -1.461 -0.804 -0.913 -0.366 -0.694 -0.694 -0.694 -1.461 -1.351 -0.804 -1.570 -1.570 -1.351 -1.899 -1.899 -2.008 -1.351 -0.913 -1.132 -1.570 -1.570 -1.789 -2.008 -1.789 -2.446 -1.570 2.044 1.606 2.372 1.387 0.839 2.153 1.496 1.606

H H H H H H H A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A H H H H H H H H

HUMAN KANGAROO SHEEP PIG

9

30

1.035

H

33

1.934

H

10

32

1.495

H

32.5

1.825

H

Max. Value 78 79 78.5 76 73 81 77 89 79 83 85 65.5 68 79.5 77 70.5 79 84 81 95.5 83 64.5 65 67 71.5 63.5 64 50 53.5 64.5 62.5 57 54 60 64.5 58.5 60 74 71

*mc 0.666 0.733 0.700 0.532 0.331 0.867 0.599 1.403 0.733 1.001 1.135 -0.171 -0.004 0.767 0.599 0.164 0.733 1.068 0.867 1.839 1.001 -0.238 -0.205 -0.071 0.231 -0.305 -0.272 -1.210 -0.975 -0.238 -0.373 -0.741 -0.942 -0.540 -0.238 -0.640 -0.540 0.398 0.197

Individual 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2

Max. Value 29.5 30.5 28.5 29 29 30 28 35 30 32.5 30.5 24 25 30 28 28 31.5 24 25.5 28 26.5 21 21 22 23 20.5 20 15 17 20.5 19 17 16 18 20 18 19 25 21

TIBIA † *ap (H/A) 1.567 H 1.836 H 1.298 H 1.432 H 1.432 H 1.701 H 1.163 H 3.046 H 1.701 H 2.374 H 1.836 H 0.087 H 0.356 H 1.701 H 1.163 H 1.163 H 2.105 H 0.087 H 0.491 H 1.163 H 0.760 H -0.720 A -0.720 A -0.451 A -0.182 A -0.855 A -0.989 A -2.334 A -1.796 A -0.855 A -1.258 A -1.796 A -2.065 A -1.527 A -0.989 A -1.527 A -1.258 A 0.356 H -0.720 A 129

†

(H/A) H H H H H H H H H H H A A H H H H H H H H A A A H A A A A A A A A A A A A H H

DOG EMU

3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9

22 24.5 22.5 23 21 24 22 21 18 20.5 18.5 17.5 17.5 13 16 16 18 17 25 24 26 26 26 25.5 22 25 24

-0.451 0.222 -0.317 -0.182 -0.720 0.087 -0.451 -0.720 -1.527 -0.855 -1.393 -1.662 -1.662 -2.872 -2.065 -2.065 -1.527 -1.796 0.356 0.087 0.625 0.625 0.625 0.491 -0.451 0.356 0.087

A H A A A H A A A A A A A A A A A A H H H H H H A H H

68.5 70.5 67.5 69 60.5 70 62 63.5 59 62 61 55.5 44 41 51 52 59 53.5 85 83.5 87 91 90 90 73 90 83

0.030 0.164 -0.037 0.063 -0.506 0.130 -0.406 -0.305 -0.607 -0.406 -0.473 -0.841 -1.612 -1.813 -1.143 -1.076 -0.607 -0.975 1.135 1.035 1.269 1.537 1.470 1.470 0.331 1.470 1.001

H H A H A H A A A A A A A A A A A A H H H H H H H H H

10

27

0.894

H

91.5

1.571

H

130

Maximum Value below Known Midshaft:

HUMAN KANGAROO SHEEP PIG

Individual 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4

Max. Value 23 18 20 21 21 20 19.5 19.5 24.5 19.5 19 19 18 22 19.5 19 19 24 21 21 25 26 22 21 25.5 22 25 18 20 21 22 21 23 18 21 22 21.5 26 28 23 23

HUMERUS † *ap (H/A) 1.085 H -0.695 A 0.017 H 0.373 H 0.373 H 0.017 H -0.161 A -0.161 A 1.619 H -0.161 A -0.339 A -0.339 A -0.695 A 0.729 H -0.161 A -0.339 A -0.339 A 1.441 H 0.373 H 0.373 H 1.797 H 2.153 H 0.729 H 0.373 H 1.975 H 0.729 H 1.797 H -0.695 A 0.017 H 0.373 H 0.729 H 0.373 H 1.085 H -0.695 A 0.373 H 0.729 H 0.551 H 2.153 H 2.865 H 1.085 H 1.085 H 131

Max. Value 20 17 17 17.5 20 18 19 18 22 20 18 16 14 21.5 19 16.5 19 21 18 18 29 22 19 18.5 22 19.5 23 15.5 16.5 18 19 21 20 17 20 20 18 17 20.5 17 18

*mld 0.048 -0.972 -0.972 -0.802 0.048 -0.632 -0.292 -0.632 0.728 0.048 -0.632 -1.312 -1.992 0.558 -0.292 -1.142 -0.292 0.388 -0.632 -0.632 3.108 0.728 -0.292 -0.462 0.728 -0.122 1.068 -1.482 -1.142 -0.632 -0.292 0.388 0.048 -0.972 0.048 0.048 -0.632 -0.972 0.218 -0.972 -0.632

†

(H/A) A H H H A H H H A A H H H A H H H A H H A A H H A H A H H H H A A H A A H H A H H

DOG

5 6 7 8 9 10 1 2 3 4 5 6 7 8 9

25 19 21 20 19 18 25 22 22 20 21.5 19.5 17.5 17 14

1.797 -0.339 0.373 0.017 -0.339 -0.695 1.797 0.729 0.729 0.017 0.551 -0.161 -0.873 -1.051 -2.119

H A H H A A H H H H H A A A A

17 15 16.5 17.5 18.5 14 20 20 16.5 17 18 18 16 16 15

-0.972 -1.652 -1.142 -0.802 -0.462 -1.992 0.048 0.048 -1.142 -0.972 -0.632 -0.632 -1.312 -1.312 -1.652

H H H H H H A A H H H H H H H

10

18

-0.695

A

18

-0.632

H

*

ap: midshaft antero-posterior diameter, mld: midshaft medio-lateral diameter, mc: midshaft circumference. †

H: Human, A: Non-human

HUMAN ROO KANGA

Individual 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3

Max. Value 28 29.5 29.5 28 25 26 26.5 27 30 31 28 22 24 24 28 26.5 23.5 31.5 26 30

FEMUR † *ap (H/A) 0.575 H 0.92 H 0.92 H 0.575 H -0.115 A 0.115 H 0.23 H 0.345 H 1.035 H 1.265 H 0.575 H -0.805 A -0.345 A -0.345 A 0.575 H 0.23 H -0.46 A 1.38 H 0.115 H 1.035 H 132

Max. Value 23.5 25 28 26 24 28 26 25 24 25 30.5 24 24 27.5 26 24.5 28.5 28 24 25

*mld -0.147 0.182 0.839 0.401 -0.037 0.839 0.401 0.182 -0.037 0.182 1.387 -0.037 -0.037 0.730 0.401 0.072 0.949 0.839 -0.037 0.182

†

(H/A) A H H H A H H H A H H A A H H H H H A H

SHEEP PIG DOG EMU

4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8

31.5 26.5 28 27 28 31.5 27.5 18.5 20 21 24 21 22 25 23 21 21 24 22 22.5 22.5 21 21 20 18 17 17 20 22 22 19 19 17 16 17 14 19 32.5 30 33.5 31 27 31 31 31

1.38 0.23 0.575 0.345 0.575 1.38 0.46 -1.61 -1.265 -1.035 -0.345 -1.035 -0.805 -0.115 -0.575 -1.035 -1.035 -0.345 -0.805 -0.69 -0.69 -1.035 -1.035 -1.265 -1.725 -1.955 -1.955 -1.265 -0.805 -0.805 -1.495 -1.495 -1.955 -2.185 -1.955 -2.645 -1.495 1.61 1.035 1.84 1.265 0.345 1.265 1.265 1.265

H H H H H H H A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A H H H H H H H H 133

28.5 27 24.5 27 27 28 27 18 20 19 23 17.5 22.5 23 23 22.5 22.5 21 18 20 21 18 17 19 17 15.5 14 19 21 21 18 18 17 17 17.5 14 19.5 35 32.5 35.5 30 27.5 35 32.5 33

0.949 0.620 0.072 0.620 0.620 0.839 0.620 -1.351 -0.913 -1.132 -0.256 -1.461 -0.366 -0.256 -0.256 -0.366 -0.366 -0.694 -1.351 -0.913 -0.694 -1.351 -1.570 -1.132 -1.570 -1.899 -2.227 -1.132 -0.694 -0.694 -1.351 -1.351 -1.570 -1.570 -1.461 -2.227 -1.023 2.372 1.825 2.482 1.277 0.730 2.372 1.825 1.934

H H H H H H H A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A H H H H H H H H

HUMAN KANGAROO SHEEP PIG

9

31

1.265

H

34

2.153

H

10

31.5

1.38

H

34

2.153

H

Max. Value 74.5 75 76 73 69 78.5 73 88 73.5 79.5 81.5 63 65 77.5 73.5 68 78 78 78 87 75 56.6 58 63 64.5 60 57 50 53.5 64.5 62.5 57 51 55.5 60 54.5 55 63 60

*mc 0.432 0.465 0.532 0.331 0.063 0.700 0.331 1.336 0.365 0.767 0.901 -0.339 -0.205 0.633 0.365 -0.004 0.666 0.666 0.666 1.269 0.465 -0.768 -0.674 -0.339 -0.238 -0.540 -0.741 -1.210 -0.975 -0.238 -0.373 -0.741 -1.143 -0.841 -0.540 -0.908 -0.875 -0.339 -0.540

Individual 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2

Max. Value 28 28.5 27.5 28 25.5 28.5 27 33 27 31 29 23 23 28 26 25 30 23 23 26.5 25 19.5 20 21 21 19 19 13 15 18 17 15.5 15 16 17 16 16.5 18 16

TIBIA † *ap (H/A) 1.163 H 1.298 H 1.029 H 1.163 H 0.491 H 1.298 H 0.894 H 2.508 H 0.894 H 1.970 H 1.432 H -0.182 A -0.182 A 1.163 H 0.625 H 0.356 H 1.701 H -0.182 A -0.182 A 0.760 H 0.356 H -1.124 A -0.989 A -0.720 A -0.720 A -1.258 A -1.258 A -2.872 A -2.334 A -1.527 A -1.796 A -2.200 A -2.334 A -2.065 A -1.796 A -2.065 A -1.931 A -1.527 A -2.065 A 134

†

(H/A) H H H H H H H H H H H A A H H A H H H H H A A A A A A A A A A A A A A A A A A

DOG EMU

3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9

16 16 15 16 13 14.5 11.5 13 16.5 20 17 15.5 16.5 11.5 13 14 16.5 15.5 25 25 26 26 25 24.5 22 25 24

-2.065 -2.065 -2.334 -2.065 -2.872 -2.469 -3.276 -2.872 -1.931 -0.989 -1.796 -2.200 -1.931 -3.276 -2.872 -2.603 -1.931 -2.200 0.356 0.356 0.625 0.625 0.356 0.222 -0.451 0.356 0.087

A A A A A A A A A A A A A A A A A A H H H H H H A H H

59 56.5 57 57 47 55 45 44 56 60 59.5 53.5 40 37.5 47 48 52.5 56.5 87 84.5 87 91 90 89 74 90 84.5

-0.607 -0.774 -0.741 -0.741 -1.411 -0.875 -1.545 -1.612 -0.808 -0.540 -0.573 -0.975 -1.880 -2.048 -1.411 -1.344 -1.043 -0.774 1.269 1.102 1.269 1.537 1.470 1.403 0.398 1.470 1.102

A A A A A A A A A A A A A A A A A A H H H H H H H H H

10

26.5

0.760

H

91.5

1.571

H

135