Citation
Pollical metacarpal morphology in primitive Homo

Material Information

Title:
Pollical metacarpal morphology in primitive Homo evidence from Homo Naledi
Creator:
Bowland, Lucyna A. ( author )
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file : ;

Subjects

Subjects / Keywords:
Hand ( lcsh )
Fingers ( lcsh )
Homo naledi ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Homo naledi provides important insights into the possible anatomy of primitive members of the genus Homo. The pollical metacarpal displays a suite of morphological characteristics previously unknown within hominoids, including a median longitudinal crest, a narrow proximal base, and broad flaring muscle flanges on the distal end. To better understand the unique morphology of the H. naledi pollical metacarpal, this study employs a 3D geometric morphometric analysis of the pollical metacarpal shaft morphology in a comparative sample of adult Old World monkeys, non-human apes, modern humans, and fossil hominins (n=337). A 20x20 semilandmark grid was placed across the palmar diaphyseal surface of 3D surface renderings and the landmark data were subjected to a principal components analysis of Procrustes shape variables. A multiple regression analysis of log centroid size on Procrustes shape variables was performed to test for the effects of size on shape. The results indicate that H. naledi separates from other hominins based on its narrow proximal base and gracile shaft surmounted by flaring muscle flanges on the distal end. The gracile shaft is most like cercopithecines, Pan, Pongo, and the australopiths, suggesting this is the primitive condition for hominins. In contrast, Neanderthals are characterized by a wide proximal base, a pinched midshaft, and broad distal flanges, while modern human metacarpals have a straight shaft, less robust muscle flanges, and a robust proximal base. The results of this study suggest that pollical metacarpal morphological development within primitive Homo was characterized by an intensified recruitment of intrinsic muscles on an otherwise gracile shaft, adding to the current understanding of the early manipulative behaviors and tool use within the genus Homo.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Lucyna A. Bowland.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
on10611 ( NOTIS )
1061150070 ( OCLC )
on1061150070

Downloads

This item has the following downloads:


Full Text
POLLICAL METACARPAL MORPHOLOGY IN PRIMITIVE HOMO: EVIDENCE
FROM HOMO NALEDI BY
LUCYNA A. BOWLAND B.A., University of West Florida, 2013
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts
Department of Anthropology
2018


This thesis for the Master of Arts degree by
Lucyna A. Bowland has been approved for the Anthropology Program by
Charles Musiba, Chair Caley M. Orr Jill E. Scott Anna Warrener
Date: May 12, 2018


Bowland, Lucyna A. (M. A., Anthropology Program)
Pollical Metacarpal Morphology In Primitive Homo: Evidence From Homo Naledi Thesis directed by Associate Professor Charles Musiba
ABSTRACT
Homo naledi provides important insights into the possible anatomy of primitive members of the genus Homo. The pollical metacarpal displays a suite of morphological characteristics previously unknown within hominoids, including a median longitudinal crest, a narrow proximal base, mid broad flaring muscle flanges on the distal end. To better understand the unique morphology of the H. naledi pollical metacarpal, this study employs a 3D geometric morphometric analysis of the pollical metacarpal shaft morphology in a comparative sample of adult Old World monkeys, non-human apes, modem humans, and fossil hominins (n=337). A 20x20 semilandmark grid was placed across the palmar diaphyseal surface of 3D surface renderings and the landmark data were subjected to a principal components analysis of Procrustes shape variables. A multiple regression analysis of log centroid size on Procmstes shape variables was performed to test for the effects of size on shape. The results indicate that if. naledi separates from other hominins based on its narrow proximal base and gracile shaft surmounted by flaring muscle flanges on the distal end. The gracile shaft is most like cercopithecines, Pan, Pongo, mid the australopiths, suggesting this is the primitive condition for hominins. In contrast, Neanderthals are characterized by a wide proximal base, a pinched midshaft, and broad distal flanges, while modern human metacarpals have a straight shaft, less robust muscle flanges, and a robust proximal base. The results of this study suggest that pollical metacarpal morphological development within
m


primitive Homo was characterized by an intensified recruitment of intrinsic muscles on an otherwise gracile shaft, adding to the current understanding of the early manipulative behaviors mid tool use within the genus Homo.
The form and content of this abstract are approved. I recommend its publication.
Approved: Charles Musiba
IV


TABLE OF CONTENTS
I. POLLICAL METACARPAL MORPHOLOGY IN PRIMATES..............................1
Introduction and Review of the Literature..................................1
Relationship Between Entheseal Morphology, Behavior, and Load-Bearing
Ability.............................................................3
Manual Grips within Hominoids.......................................6
Morphological Variation within the Hands of Lossil and Extant Primates
Morphology of Extant Primates.................................9
Morphology of Fossil Hominins................................15
Theories for the Evolution of the Modem Human Hand.................28
The Rise of Habitual Stone Tool Use..........................29
The Hand is a Pleiotropic Result of Selection for the Foot...30
Conclusion................................................................32
II. MATERIALS AND METHODS.................................................34
Materials.............................................................34
Methods...............................................................38
III. RESULTS..............................................................41
Full Sample...............................................................41
PCI Shape and Groupings............................................41
PC2 Shape and Groupings............................................42
PC3 Shape and Groupings............................................43
Regional Variation within Modem Humans....................................44
Reduced Sample............................................................47
v


Hominoid Sample...................................................49
PCI Shape and Groupings.....................................50
PC2 Shape and Groupings.....................................53
PC3 Shape and Groupings.....................................55
Shape and Size...........................................................57
IV. DISCUSSION AND FUTURE RESEARCH.......................................59
PCl-Breadth of the Metacarpal Shaft......................................59
PC2-Breadth of the Proximal End Relative to the Overall Shaft............60
PC3-Breadth of the Proximal End Relative to the Distal Flanges...........61
Implications for the Morphology of Primitive Homo........................62
Future Research..........................................................63
Variation within Gorilla spp......................................63
Sexual Dimorphism.................................................64
Conclusion...............................................................65
REFERENCES...............................................................66
vi


CHAPTER I
POLLICAL METACARPAL MORPHOLOGY IN PRIMATES Introduction and Review of the Literature
A fully opposable thumb capable of powerful precision grasping is seen as a hallmark of modem humans. Over the course of our evolutionary history, the hand was freed from its locomotor constraints, allowing for its use primarily as a tool for manipulator purposes (Kivell et al., 2011). Humans are unique among the extant great apes in the extensive use of the thumb, which is reflected in their broad trapeziometacarpal joints (TCMJ) and robust metacarpal shafts, both of which are adaptations for enhanced precision grasping within modem humans (Marzke, 1997). Humans are also unique in respect to the musculature of the hands. In contrast to all other extant primates (except for gibbons) the flexor pollicis longus (FPL) in humans is a distinct muscle from the flexor digitomm profundus (FDP), with a tendon that is exclusive to the pollex (Lemelin and Diogo, 2016). This feature is likely related to increased force at the tip of the first metacarpal (Me 1), which in turn aids enhanced precision grasping (Hamrick et al., 1998; Marzke et al., 1998, Lemelin and Diogo, 2016). In contrast, extant non-human primates possess short thumbs relative to finger length, with long phalanges that preclude enhanced precision grasping (Marzke, 1997; Almecija et al., 2015). Extant African apes occasionally use their thumbs for manipulator behaviors, but their hand morphology is highly specialized for their locomotor patterns, in which the pollical metacarpal is rarely engaged (Tuttle, 1969a).
To accurately understand the significance of the manual morphology of extant primates, one must look at the evidence for early hand morphology within fossil hominins. Within the fossil record, there is evidence for hominin hand remains by ca. 6 Ma with
1


Orrorin tugenensis (Almecija et al., 2010). Or. tugenensis remains display several morphologies related to enhanced grasping, including a broad apical tuberosity and an insertion for the FPL on the pollical distal phalanx (PDP) (Marzke 1997; Almecija et al., 2010). Ardipithecus ramidus, at circa 4.5 Ma, had short metacarpals, long fingers, and a mobile wrist. This suite of morphologies contrasts with extant African apes and is more like Old World monkeys and Miocene hominids such as Proconsul (Lovejoy et al., 2009; Kivell, 2015). Australopithecus metacarpals are characterized by an overall gracility, with narrow metacarpal shafts and bases, mid a greatly diminished crest for the opponens pollicis. Homo neanderthalensis metacarpals show a marked hypertrophy of the intrinsic musculature, and their overall morphology appears to be autapomorphic for the taxon. The early anatomically modem human (EAMH) Qafzeh specimens lack the hypertrophy of the intrinsic musculature that is present in Neanderthals, and their pollical metacarpals are overall less robust than Neanderthals, and more like modem humans.
However, noticeably absent from the fossil record is detailed evidence for the manual morphology of primitive members of the genus Homo. The fossil remains for primitive Homo are sparse, including several specimens from Homo habilis (OH7) (Leakey et al., 1964), partial hand remains from Homo /lores! ensis (Brown et al., 2004; Tocheri et al.,
2007), and remains from the newly-discovered species Homo naledi (Berger et al., 2015; Kivell et al., 2015). Neither the remains from Olduvai Gorge (OH7), nor those from H. floresiensis include a complete Mcl, meaning the evidence for the morphology of primitive Homo first metacarpals has been, until recently, poorly understood. However, included within the vast fossil finds from the Dinaledi Chamber in South Africa were seven pollical metacarpals from H. naledi (Berger et al., 2015). All the specimens exhibited the same
2


unusual morphological characteristics: a narrow proximal base surmounted by broad muscle flanges on the distal ends of the shaft (Kivell et al., 2015), which is unlike anything seen previously. It is unclear whether this unique morphology possibly represents the character state for primitive members of the genus Homo, or whether H. naledi represents an autapomorphy in the hominin clade.
This paper attempts to answer the question of whether H. naledi represents an autopamorphic taxon or a transitional form for primitive Homo pollical metacarpal morphology. To answer this question, and to better understand the morphology of primitive Homo pollical metacarpals, this study employed a 3D geometric morphometric analysis of pollical metacarpal shaft morphology in a comparative sample of Old World monkeys, nonhuman apes, modem humans, and fossil hominins (n=337).
Relationship between Entheseal Morphology, Behavior, and Load-Bearing Ability
The intrinsic muscles of the hand have often been a focus for pale o anthropological studies because of their assumed relationship to hand use and function (Ricklan, 1987; Eliot mid Jungers, 2000; Zumwalt, 2006; Marzke et al., 2007). Increased muscle entheseal development is often associated with increased recmitment of the intrinsic musculature, because of the assumption that the increase in muscle recmitment was correlated with the advent of stone tool usage (Susman, 1994).
An enthesis is the site where muscle meets bone, and is the result of the stress concentration where hard (bone) and soft (muscle) tissues meet. Muscle entheses help to transmit forces over osteotendinous surfaces, as well as anchor tendons to bones. The proposed relationship between muscle enthesis development and behavior is based on the idea that repetitive activity stimulates periosteal remodeling and modeling activity by
3


increasing bone cell proliferation through expansion of the capillaries (Schlecht, 2012). While the perceived relationship between the two components appears to be somewhat straightforward, it is important to note that there is no proven relationship between muscle entheseal morphology and the behavior of the organism in question, and previous studies have found no relationship between the two (Rabey et al., 2015; Williams-Hatala et al., 2016). However, it is not currently understood how other factors, such as age, sex, body mass, and genetics influence enthesis morphology, mid these variables are rarely explored in entheseal morphology studies (Schlecht, 2012).
In a recent study, Williams-Hatala et al. (2016), examined the relationship between muscle enthesis development mid function in a sample of cadaveric modem human pollical metacarpals. The study looked at whether any correlations existed between physiological cross-sectional area of the opponens pollicis muscle and enthesis development. The study found no statistically significant (p>0.05) relationship between the muscle architecture and the entheseal surface morphology. The authors questioned the efficacy of using muscle enthesis architecture as a proxy for function in extinct organisms, arguing the musculoskeletal stmcture is too complex for simple causal relationships (Williams-Hatala et al., 2016). However, their sample consisted entirely of cadaveric specimens, whose muscles no doubt underwent some degree of atrophy prior to the time of the study. Additionally, previous research (Zumwalt, 2006) has noted the problematic nature of using fully mature individuals when attempting to assess the efficacy of activity levels and muscle enthesis morphology.
It has long been hypothesized (Marzke et al., 1992; Susman, 1994; Marzke et al., 1998) that a more robust thumb is synonymous with the ability to bear increased loading
4


experienced during tool making. Human pollical metacarpals have abroad shaft and an expanded proximal base, which increases surface area and allows for more even distribution of forces incurred because of powerful precision grasping (Marzke, 1997). To test this hypothesis, Williams et al. (2012) monitored the normal force (N) and pressure (kPa) of six experienced tool makers while they simulated 01 do wan tool making techniques. The participants held the stone tools in a three-jaw chuck grip (Fig. 1) with the hammer stone stabilized by distal phalanges of the first three digits. While they conceded the thumb plcys an important role in stabilizing the object during production, they note it is, "over-built" for tool-making stress levels. The results of their study found forces incurred during 01 do wan tool production concentrated more on the second and third digit, as opposed to the thumb. They concluded the results do not support the hypothesis that the primate thumb evolved as a response to increased loads experienced during 01 do wan tool making.
Figure 1. Extended 3-jaw chuck grip in a modem human. From Marzke et al., 1997.
5


Manual Grips within Hominoids
The ability to engage in pad-to-pad precision grasping is paramount for enhanced manual dexterity within modern humans, and the unique morphology of the human thumb is tantamount to the production of manual grips (Napier, 1962a; Napier, 1962b; Marzke, 1997; Almecija et al., 2010). Precision grip is defined as any grip that utilizes the thumb and at least one other finger, either with or without using the palm as a prop (Napier, 1962b; Marzke, 1997) (Fig. 2). Pad-to-pad precision grip is defined as the proximal pulp of the thumb being in opposition to one or more fingers (Almecija et al., 2010). Human hands are uniquely adapted for pad-to-pad precision grasping, in contrast to other extant apes, whose long fingers, short thumbs, mid reduced intrinsic musculature preclude them from engaging in effective manual grips (Napier and Tuttle, 1993; Almecija et al., 2010). Evidence for enhanced precision grasping capabilities has been found in numerous hominin species, beginning with the Miocene hominin Or. tugenensis (Senut et al., 2001; Almecija et al.,
2010). Morphological features associated with pad-to-pad precision grasping can be found on the PDP of modem humans and select fossil hominin species, and include an insertion point for FPL, the presence of an ungual fossa, and asymmetrical ungual spines (Marzke, 1997). All features of the PDP associated with pad-to-pad precision grasping are found within modem humans, but are absent entirely within extant great apes, and are variably present within the hominin fossil record.
Almecija et al. (2010) compared the morphology of the PDP in extant primates and fossil hominin specimens to explore the morphological affinities between these taxa. Or. tugenensis, the oldest known specimen for which there is evidence of a PDP (Senut et al., 2001), has both an insertion gable for the FPL and a broad apical tuberosity, like modem
6


human PDPs (Almecija et al., 2010). However, the later specimen of OH7, attributed to H. hcibilis, lacks an insertion gable for the FPL and ungual spines, though it does possess a large palmar fossa that extends to the apical tuberosity (Almecija et al., 2010).
Figure 2. Modem human thumb and index finger (right hand) during pad-to-pad precision grasping in ulnar view. From Almecija et al., 2010.
Robust muscle attachment sites for the opponens pollicis and FPL on the thumb have been used to infer enhanced grasping capabilities within early hominins (Marzke et al., 1998; Niewoehner et al., 2003; Maki and Trinkaus, 2011; Marzke, 2013). The opponens pollicis muscle is located on the distodorsoradial portion of the first metacarpal, and the presence of a particularly robust muscle attachment site has been used to infer hypertrophy of the opponens pollicis, as well as provide evidence for an increased moment arm of the muscle (Trinkaus, 1983; Maki and Trinkaus, 2011). Neanderthals are known to have a particularly robust crest for the opponens pollicis muscle attachment (Niewoehner, 2006). Maki and Trinkaus (2011) examined the relationship between muscle enthesis development and function in a study of Neanderthal, modem human, and Middle and Mid Upper Paleolithic pollical metacarpals.
7


Their study found size of the crest, both absolute and relative to overall body size, to be greater in Neanderthals and Middle Paleolithic hominins than in later hominins and modem humans. They hypothesized the decrease in the size of the crest in later hominins is possibly tied to advances in lithic technology as the Paleolithic progressed, requiring less muscle force to be employed than in previous lithic technologies.
Morphological Variation within the Hands of Fossil and Extant Primates
Primate hands are comprised of several sets of bones. The bones of the wrist are known as the carpal bones, the number of which varies among different primate taxa.
Humans have eight carpal bones, while most other primates possess nine. This is due to the fusion of the scaphoid and os centrale bone into one element in humans, while the two are separate bones in many other primate species. The palmar portions of the fingers are known as the metacarpals, while the external portions of the digits are comprised of proximal, intermediate, and distal phalanges (except for the thumb, which contains only the proximal mid distal phalanges) (Kivell, 2016).
There is a wide range of variation within the pollical metacarpals of primates. Figure 3 shows morphological variation for a selection of hominoid taxa. Some taxa, such as Pan mid Hylobates, possess gracile metacarpal shafts with a narrow proximal base. In humans, the metacarpal shaft is broad, with a relatively expanded articular facet on the proximal end. The following section will review some of the key morphological features of extant primates, as well as discuss the functional importance corresponding to each
8


Figure 3. Palmar view of three-dimensional surface renderings of the pollical metacarpal in a selection of hominoid taxa. ^Variously attributed to Australopithecus and Homo.
Morphology of Extant Primates
Homo sapiens. The bony morphology and distinct musculature of the human thumb are indicative of its integral role in human manipulative behaviors (Kivell, 2015). The human hand is distinct from that of other living primates in terms of the long thumb relative to other digit lengths and the broad apical tufts on the distal phalanges, both of which aid precision grasping capabilities (Marzke, 1997; Almecija et al., 2015). Compared to other extant primates, the muscles of the human thumb have a much larger moment arm, which is linked to enhanced mechanical capabilities, such as offering more leverage when attempting to stabilize the hand against an object (Kivell, 2015). H. sapiens hands are also characterized by much lesser degrees of phalangeal curvature than is present in other living primates. Figure 4 describes the morphological adaptations associated with enhanced precision grasping within the human hand, as well as the muscles recruited during precision and power squeeze grips.
9


()
broad apical tufts (with ungual spines) for better manipulative control
(b)
precision grip
(C)
power sqeeze grip
Me proportions facilitate precision and power squeeze grips
styloid process* to stabilize carpal-Mc joints
extended first dorsal interosseous attachment
combination of thumb and fifth digit are critical to controlling and manipulating objects within one hand during precision and power squeeze grips used during tool making and tool use
fifth digit stabilizes hammerstone and core during tool making
robust thumb with FCU, FDM and ADM well-developed active in dominant
muscle attachments and non-dominant
large, mobile and relatively hands during
flat trapezium-Mcl joint tool making
broader boot-shaped trapezoid
proximodistal reorientation of radial carpal and carpal-Mc joints large scaphoid-trapezoid articulation
radiocarpal ability for wrist flick
FPL not strongly active**
OP, first DI and FPB active just before strike in dominant hand
most muscles of thumb have larger moment arms than in Pan OP and AP muslces have large cross-sectional area and potential torque
EPB and independent FPL more frequent in humans than other primates
fifth digit anchors hand during power squeeze grip
Figure 4. Bony and soft tissue considered to be associated with precision and power squeeze grips used during stone tool use and production (a). Human precision grip (b) and a power
squeeze grip (c). From Kivell, 2015.
Humans possess three muscles (an independent FPL, the extensor pollicis brevis, and the first volar interosseous of Henle) which are thought to be integral to their ability to form precision grips (Susman et ah, 1999; Williams et ah, 2012). These three muscles are absent in nearly all other extant primates, except for the Hylobatids, who also possess a FPL that is separate from the flexor digitorum profundus (FDP) (Lemelin and Diogo, 2016). The FPL inserts on the volar side of the distal pollical phalanx and has been widely associated with stone tool usage within the hominin lineage (Susman, 1994; Marzke, 1997; Diogo et al.,
2012; Williams et al., 2012). Apart from the hylobatids (in which the FPL and the FDP are joined via a connective tissue) and Homo sapiens, this muscle is joined with the FDP muscle in all other primates (Lemelin and Diogo, 2016). Humans are therefore unique among extant primates in having an FPL that is free from the other digits, which is responsible for the increased flexion seen in the thumb of modem humans (Hamrick et al., 1998; Williams et al., 2012). However, the discovery of an insertion site for the FPL on the distal pollical phalanx
10


of Orrorin tugenensis (ca. 6 Ma), predating the use of stone tools by at least 2.5 Ma (McPherron et al., 2010), throws doubt on the long-held assumption that the presence of this muscle insertion site is synonymous with tool usage in hominins (Susman, 1994, Almecija et al., 2010).
The extensor pollicis brevis (EPB) muscle inserts on the first phalanx mid is present in roughly 90% of modem humans (Williams et al., 2012). The muscle aids in thumb extension by assisting the adductor pollicis brevis (APB) and the extensor pollicis longus (EPL) and is integral to the ability to stabilize the thumb against an object while extended. The first volar interosseous of Henle is found in roughly 90% of modem humans (Williams et al., 2012), mid less than 50% of other primate taxa, though it is present in both Gorilla and P. troglodytes (Diogo et al., 2012). Its function in unclear, though it is believed to either aid in adducting the thumb at the metacarpophalangeal joint, or possibly to add sensory information about the position of the thumb (Susman et al., 1999; Williams et al., 2012).
The opponens pollicis muscle inserts on the lateral side of the pollical metacarpal (Diogo et al., 2012). In humans, the insertion site presents as a distinct bony flange mnning along the medial portion of the bone. The muscle is responsible for abduction and flexion at the first carpometacarpal joint, allowing for the opposition of the thumb. In modem humans, the opponens pollicis occupies a much larger cross-section and provides more potential torque than what is seen in extant primates (Marzke, 1997; Marzke et al., 1999). Taken together, these two features are thought to limit the fatigue experienced by the thumb during increased loading, which has been shown to occur in experimental studies of Oldowan tool manufacture (Hamrick et al., 1998, Kivell, 2015). As studies of fossil remains must rely solely on the bony morphology to form functional interpretations, the presence of the bony
11


flange on the distodorsoradial portion of the pollical metacarpal has been used to infer hypertrophy of the muscle, as well as a greater moment mm for movement (Maki and Trinkaus, 2011).
Non-human primates. This section will focus on the hand morphology of extant non-human primates. The locomotor patterns of the species in question vary greatly, from committed terrestrial knuckle-walkers to fully arboreal, suspensory primates. Except for Nasalis larvatus, who is a fully arboreal primate, all cercopithecine Old World monkeys included within this study (Erythrocebus patas, Macaca fasicularis, Papio sp.) are terrestrial quadrupeds (Patel, 2010). Gorilla sp. and Pan sp. are terrestrial knuckle-walkers, Pongo sp., divide their time between an arboreal and terrestrial lifestyle, and Hylobates lar is a committed arboreal primate whose morphology is highly specialized for suspensory locomotion (Tuttle, 1969b; Siegel and Pemotto, 1975; Matarazzo, 2008).
Extant non-human apes can be divided into two general categories: African apes, Pan mid Gorilla, and the Asian apes, Pongo and Hylobates. The main morphological differences between the Asian apes and African apes are a result of their different locomotor patterns. The African apes (Gorilla and Pan) possess secondary adaptations for terrestrial locomotion, while Hylobates mid Pongo are both adapted for arboreal locomotion (Tuttle, 1969b). Figure 5 shows differences in the ray morphology and manual proportions among extant great apes.
12


Homo
&
& &
* IflRB
Q> o l>
Pongo
&
B
0
Pan
Figure 5. Articulated metacarpals and phalanges of the left hand, scaled to 1cm. From Patel and Maiolino, 2016.
African apes. Gorilla is divided into two species, the mountain Gorilla (Gorilla beringei), and the lowland Gorilla (Gorilla gorilla). Gorilla beringei are predominantly terrestrial in their feeding, nesting, and locomotor patterns. Gorilla gorilla are generally more arboreal, though this could simply be a consequence of their more forested habitats (Tuttle, 1969a). The hands of gorillas are adapted to bear the intense forces incurred throughout the hand during knuckle-walking (Kivell, 2016). The carpal bones are tightly compressed to help stabilize the wrist and prevent radioulnar deviation or rotation during knuckle-walking (Tuttle, 1969b; Kivell, 2016). In addition to an immobile wrist, Gorilla metacarpals are adapted in response to knuckle-walking locomotion. The breadth of the metacarpal head is large relative to the length of the shaft (Susman, 1994), and there is a transverse ridge at the base of the dorsal articular surface (Tuttle 1969b; Patel and Maiolino, 2016) There is debate
13


as to the functional advantages of this dorsal ridge, with some authors suggesting it acts to help support mid stabilize the metacarpophalangeal joints during locomotion (Tuttle, 1969b), though it is also possible it is unrelated to knuckle-walking and maybe instead correlate to changes in body size throughout ontogeny (Susman and Creel, 1979; Patel and Maiolino, 2016). Almecija et al. (2015) note the morphological state exhibited by gorillas and modem humans appears to be largely plesiomorphic and remains largely unchanged since the time of the last common ancestor (LCA) of Pan and Homo.
The genus Pan consists of two species, Pan paniscus (bonobos), and Pan troglodytes (chimpanzees). Pan are generally terrestrial knuckle-walkers, though they also spend considerable time engaged in climbing and suspensory locomotion in trees (Marzke, 1971). Chimpanzees move quadrupedally with their metacarpophalangeal joint hyperextended and their weight supported mainly on the dorsal aspects of the middle phalanges (Tuttle, 1969b; Marzke, 1971; Matarazzo, 2008). Tuttle (1969a) reported that during quadmpedal locomotion, chimpanzee thumbs never contact the ground, nor are they used for support during locomotion. Relative to Gorilla, the metacarpals of chimpanzees are long and gracile (Susman mid Creel, 1979).
Asian apes. Pongo (orangutans) are arboreal, mid are generally suspensory (Tuttle, 1969b). They possess the most curved phalanges of the extant great apes (Richmond et al., 2016). Compared to the African apes, the metacarpals of Pongo are elongated and have less pronounced muscle and ligament impressions (Susman and Creel, 1979). Their metacarpal heads are relatively much larger than their shafts, in contrast to the morphology seen in gorillas and Homo sapiens (Susman, 1994; Marzke and Marzke, 2000).
14


Hylohates lar (gibbons) are generally categorized as a brachiatoar, though they are capable of palmigrade and pronograde quadrupedal locomotion (Tuttle; 1969b; Van Horn, 1972). Hylobatids most common form of terrestrial locomotion consists of walking quadrupedally with the hand formed into a fist, in which most of their weight is supported on the back of the proximal phalanges (Marzke, 1971). Hylobatid metacarpals are long and thin (Susman mid Creel, 1979). The thumb of Hylobates is short in relation to their other metacarpals, though they possess one of the longest thumbs relative to other primates (Van Horn, 1972; Patel and Maiolino, 2016). Hylobates are the only extant primates, aside from modem humans, that possess a separate tendon belly for the FPL (Lemelin and Diogo, 2016). Morphology of Fossil Hominins
Within the hominin lineage, morphological adaptations associated with enhanced precision grasping first emerge within Or. tugenensis at roughly 6 Ma, mid there is further evidence for early hominin manual morphology through to Au. sediba at roughly 1.2 Ma (Marzke, 1983; Senut et al., 2001; Almecija et al., 2010; Kivell et al., 2011). This section will explore the available evidence early hominin manual morphologies and their possible functional implications.
Ardipithecus. The manual remains of Ardipithecus provide important insights into the locomotor patterns of early hominins. The genus Ardipithecus is comprised of two species, the older Ar. kadabba (5.7-5.2 Ma) (Haile-Selassie, 2001), and the younger Ar. ramidus (ca. 4.4Ma) (White et al., 1994). The remains from Ar. ramidus are much more complete, and so this section will focus on Ar. ramidus. The manual remains consist of nearly two complete hands, from which are missing only the pisiform and some terminal phalanges (Lovejoy et al., 2009).
15


The hands of Ar. ramidus display several important morphological features (Fig.6). The medial metacarpals (Mc2-5) are all short relative to extant great apes. The wrist of Ar. ramidus appears to be highly mobile, which contrasts with extant great apes (Lovejoy et ah, 2009) and more like extant monkeys (Orr, 2017). An immobile wrist helps to dissipate forces dunng loading, such as during suspensory locomotion (Lovejoy et ah, 2009). The thumb of Ar. ramidus is large and robust (Kivell, 2015), with a highly-developed FPL gable on the PDP (Lovejoy et al., 2009). In contrast, the FPL of African apes is variably highly reduced or altogether absent (Lemelin and Diogo, 2016). The articular surfaces of the metacarpophalangeal, carpometacarpal, and interphalangeal joints are larger and more robust compared to Afncan apes, possibly indicative of greater load-bearing ability throughout these joints or of greater thenar mobility (Lovejoy et ah, 2009).
Figure 6. Digitally rendered composite hand of ARA-VP-6/500 (Ardipithecus ramidus) in palmar view. From Lovejoy et ah, 2009.
Lovejoy et al. (2009:74) argue the postcramal morphology of Ar. ramidus is indicative of an above-branch quadruped which they hypothesize is also the likely
16


locomotor form of the last common ancestor of humans and chimps (LCA). At present, it is not clear whether Ar. ramidus represents a hold-out from the Miocene, representing the likely locomotor patterns of the LCA, or whether the species independently evolved for its highly specialized form of locomotion. While both scenarios are likely, Ar. ramidus is roughly 3 million years removed from the hypothesized time of the LCA (Tocheri et ah, 2008), and so any interpretations arguing its morphology closely resembles a basal hominin must bear that fact in mind.
Australopithecus sp. and Paranthropus robustus. By roughly 3 Ma, remains from Australopithecus provide evidence for the advent of morphological adaptations in the hand associated with increased manipulative abilities (Marzke, 1983; Marzke, 1997; Alba et al., 2003). Marzke (1997) noted two important features in the PDP of Stw-294, most commonly attributed to Au. africanus. These features include (1) a broad apical tuft; and (2) evidence for an FPL muscle. The FPL muscle plays an important role in flexing the thumb and providing greater resistance against the radial side of the index finger, allowing for a more forceful and controlled precision grip (Susman, 1998). Ricklan (1987) contends the Am. africanus specimen Stw 64 shows evidence of a styloid process on the third metacarpal, though others have suggested the feature is morphologically distinct from the styloid process that characterizes the genus Homo (Ward et al., 2014). The presence of a styloid process on the third metacarpal is thought to help stabilize the metacarpal while in articulation with the carpal bones, as well as relieve increased stresses incurred during tool-making (Marzke and Marzke, 2000). This feature is present in Neanderthals mid modem humans, but absent in all other extant great apes and early fossil hominins, including Australopithecus (Ward et al.,
17


2014 [but see Ricklan (1987) for a review of its possible appearance 'mAu. africanus (Stw 64))].
Green and Gordon (2008) concluded that the lengths of the of Au. africanus (Stw 418 mid Stw 583) metacarpals were like the more modem human-like lengths seen 'mAu. afarensis. However, while the relative length of the metacarpals overall is more like modem humans, the relative breadth of the first metacarpal is more like extant apes (Susman and Creel, 1979). Susman (1994) linked the breadth of the first metacarpal to a morphology more derived for manipulative tasks such as tool-making. Ricklan (1987) noted curvature in the phalanges of Au. africanus, a trait which is generally associated with arboreal locomotion mid grasping capabilities, though he disputed this was the case for Au. africanus, arguing the hands were more adapted for manipulative tasks than for time spent in the trees.
Further evidence of precision grasping capabilities can be found in Australopithecus afarensis. The skeleton of Lucy (A.L. 288) includes a single non-pollical phalanx, and finds from the Afar locality of A.L. 438 include several medial phalangeal elements (Rohan and Gordon, 2013). The clem' majority of associated hand remains, however, come from the Afar locality of A.L. 333, and as such, this review will focus primarily on remains recovered from this specific site. The hand remains associated with A.L. 333 include complete and fragmentary metacarpals and phalanges (Johanson et al., 1980; Rohan and Gordon, 2013; Ward et al.,2012).
There is some contention regarding the taxonomic affinities of the Au. afarensis phalanges in regards to their relative dimensions. Marzke (1983) and Alba et al. (2003) concluded the manual proportions of Au. afarensis were more modem human-like, with a long thumb relative to shorter digit length. Based on these proportions, both studies
18


suggested that Am. afarensis was most likely capable of a high degree of precision grasping like that seen in modem H. sapiens.
The manual remains of Au. afarensis (A.L. 333) exhibit several derived features thought to be associated with precision grasping (Ward et al., 2014). One of the derived features present is a second metacarpal with an asymmetrical head, which also has a volar-radial projection, as well as an articular surface that is beveled in an ulnar direction dorsally (Marzke, 1997; P anger et al., 2002). Marzke (1997) noted the combination of these features would have caused the index finger to pronate as it flexed, which in turn would bring the entire palmar surface of the finger into direct contact with the surfaces of an object. The joints of the second metacarpal articulate with the capitate mid trapezium in an orientation away from the sagittal plane, which allows the thumb to rotate objects in a secure 3-jaw chuck hold grip (Marzke, 1971; Marzke, 1997). Additionally, the hand exhibits a longer thumb relative to digit length, though likely not long enough to have facilitated powerful precision pad-to-pad grasping (Marzke, 1997; Panger et al. 2002).
Rolian and Gordon (2013) state they believe the dimensions of the Am. afarensis phalanges fall in between those of Gorilla and modern humans, which contrasts with previous results (Alba et al., 2003). Rolian and Gordon (2013) included all known manual remains of Au. afarensis in their sample, as opposed to only selecting certain elements for study, which is likely the cause of their contradictory conclusion. If their results are correct, it would mean Am. afarensis would not have been capable of precision grasping, and by extension, probably could not have manufactured stone tools. However, it is important to note that while there are many known metacarpal remains from Australopithecus, only a few can reliably be attributed to the same individual, making a m eta-analysis of Australopithecus
19


hand remains problematic (Kivell, 2015). Overall, the manual remains of Au. afarensis appear to have been adapted for tool usage, though their ability to manufacture those tools is still doubtful.
Australopithecus sediba (1.97 Ma), is a hominin from Malapa, South Africa (Berger et al., 2010). Malapa hominin 2, (MH2) contains an almost complete right hand, missing only distal pollical phalanges for digits 2-5 and some carpal elements. The pollical distal phalanx displays a medio laterally expanded apical tuft, which more closely resembles later hominins. However, the shaft of the distal pollical phalanx is narrower than what is seen in later hominins. Au. sediba possessed a well-developed FPL muscle, as evidenced by morphology of the distal pollical phalanx, including an ungual fossa. The robust base on the proximal articular facet of the pollical metacarpal indicates the presence of well-developed intrinsic pollical muscles (Kivell, et al., 2011). The insertion site for the opponens pollicis muscle is poorly expressed, indicating weak development of the muscle and a limited ability for opposition of the thumb (Kivell et al., 2011). The morphology of the thumb is overall more gracile than robust, which is the primitive condition, and like earlier hominins such as Au. afarensis. The gracile nature of the MH2 hand bones suggests Au. sediba most likely was not a prodigious tool maker, as studies have shown Oldowan tool making increases the loads experienced by the thumb, mid therefore, enhances the bony morphology of the muscle attachment sites (Kivell et al., 2011).
Remains of Paranthropus robustus, dated to 1.8 million years ago from the site of Swartkrans, in South Africa, demonstrate numerous morphological adaptations associated with enhanced manipulative capabilities (Susman, 1988). Prior to the discovery of the Swartkrans manual remains, many scientists assumed that brain size directly correlated with
20


tool manufacture, but the small endocranial volume of the specimens, coupled with derived manual morphology, put this theory to rest (Susman, 1988). SKX 5016, from Member 1 at Swartkrans and attributed by Susman (1998) to Paranthropus, is a PDP that shows clear markings of derived features. The distal phalanx has a broad apical tuft comparable to the dimensions seen in modem humans, a feature associated with increased precision grasping. The phalanx also shows a well-defined area for the FPL muscle (Susman, 1998), which is functionally important for independent control of the thumb separate from the other digits of the hand (Lemelin and Diogo, 2016).
SKX 5020 is a right pollical metacarpal also attributed to P. robustus. Unlike the long, gracile metacarpals of many Old-World monkeys (Patel and Maiolino, 2016), SKX 5020 is robust, with a broad proximal surface like that seen in modem humans (Marzke et al., 2010). SKX 5020 displays a clem" crest for the opponens pollicis muscle (Susman, 1998), indicating Paranthropus had an increased range of motion in the thumb relative to extant primates and other, earlier, hominins (Marzke, 1997).
Homo. Homo habilis. Evidence for tool usage in the hand of early H. habilis comes from multiple carpal and phalangeal remains widely attributed to OH 7 (Napier, 1962a; Susman mid Creel, 1979; Marzke, 1997). Attribution of the manual remains from OH 7 into the H. habilis holotype met with contention almost immediately upon the initial publication, mid while most scholars agree with the inclusion into H. habilis, there is still dispute. Moya-So la et al. (2008) note that the strongest evidence for including the manual remains with the craniodental remains from the same site has generally rested on the similarity in ages, as both the craniodental and manual remains are clearly from a sub adult individual. The manual remains were not recovered near the cranial remains, casting doubt on the validity of the
21


claim that they warrant inclusion into the OH 7 holotype. Pedal remains recovered from the same site have been attributed to a second individual, OH 8, further complicating the matter of proper placement for the manual remains. For the sake of simplicity, however, this paper will use the current attribution mid refer to the OH 7 hand remains as belonging to H. habilis. For a more in-depth review of the controversies surrounding the taxonomic affinity of the OH 7 hand remains, see Moya-Sola et al. (2008).
The manual remains of OH 7, which include middle and distal phalanges, as well as various carpal elements, have all been assigned as belonging to the right side of a subadult individual (Moya-Sola et al., 2008). Of particular interest for studies concerning precision grip capabilities are the trapezium and the distal phalanges. The distal phalanges have broad tufts, the presence of which is correlated with a higher level of precision grasping capability. The trapezium has a broad metacarpal surface that is flatter than what is seen in modem human trapeziums. Marzke (1997) notes this feature is associated with more even distribution of larger internal forces, which allow for greater stresses to be transmitted through the joint surface during tool manufacture. Unfortunately, there are no known metacarpals associated with the OH 7 hand, and so a discussion on the thenar muscle morphology is not possible (Susman and Creel, 1979).
Homo /lores! ensis. H. floresiensis was a small-bo died hominin from the Indonesian island of Flores, first described in 2004, and dated to between 100-60 ka (Brown et al., 2004; Sutikna et al., 2016). Included within the manual remains for the species are carpals, phalanges, and partial metacarpals from the type specimen LB1 (Tocheri et al., 2007; Larson et al., 2009; Orr et al., 2013), and several carpal bones from other individuals (Larson et al., 2009; Orr et al., 2013). The wrist of H. floresiensis is more like extant African apes than modem humans
22


or Neanderthals (Tocheri et al., 2007; Orr et al., 2013), while the phalanges do show some evidence of derived morphological characteristics (Larson et al., 2009).
Three complete carpals were discovered with the remains of LB1, including a scaphoid, capitate, and trapezoid, as well a partial lunate and hamate (Tocheri et al., 2007; Orr et al., 2013). The three complete carpal bones all display a symplesiomorphic morphology that is shared with extant African apes, including a wedge-shaped trapezoid and a triangular articular surface on the scaphoid, and waisting along the radial aspect of the capitate, which is like what is seen in the australopiths (Tocheri et al., 2007; Larson et al., 2009).
Larson (2009) lists three possible metacarpals known from H. floresiensis. Of these, one is possible a metatarsal, one was lost during transfer, and the last is described here. The solitary metacarpal definitively known from the species lacks both the proximal and distal ends, though the remaining shaft does show moderate signs of muscle markings for the interosseous muscles. The phalanges associated with H. floresiensis are all partial, though still informative. Two distal manual phalanges are known for the species, one a pollical and one a non-pollical phalanx. The PDP has a well-defined pit associated with the insertion for the FPL (Larson et al., 2009), which is associated with enhanced precision grasping abilities (Susman, 1988).
Homo naledi. H. naledi is a recent fossil hominin discovery from the Dinaledi Chamber of the Rising Star Cave System in South Africa, dated to between 236 ka and 335 ka (Berger et al., 2015; Dirks et al., 2017). The find remains the largest single accumulation of hominin bones to date, with nearly 150 elements recovered from the hand alone. The phalanges of H. naledi exhibit markedly primitive morphology, although the wrist of the species is more like
23


later hominins mid modem H. sapiens. Hand 1 is the most complete hand recovered and consists of 26 bones from the right side of one individual. The pollical metacarpal of Hand 1, along with six other pollical metacarpals recovered from different specimens, show that H. naledi possessed pollical metacarpals with a robust distal end, but a narrow proximal base. The pollical metacarpal of Hand 1 exhibits well-developed attachment sites for both the opponens pollicis and the first dorsal interosseous muscle (Kivell et al., 2015). The carpal bones from H. naledi fall within the range of variation seen in Neanderthals and modem humans, and are derived relative to extant great apes. The carpal bones from Hand 1 exhibit a flat TCMJ, a facet on the trapezium that extends onto the scaphoid tubercle, mid a bootshaped trapezoid. These are all features shared with Neanderthals mid modem humans, mid most likely serve to distribute compressive loads incurred during strong precision grips involving the thenar musculature (Marzke, 1997; Kivell et al., 2015).
This morphology contrasts with the more gracile pollical metacarpals of earlier australopiths (Marzke, 1983) and many Old-World monkeys (Patel and Maiolino, 2016), and more closely aligns with the metacarpal morphology seen in modern humans mid Neanderthals (Niewoehner, 2001; Niewoehner, 2006). Additionally, the PDP of H. naledi are robust, with a broad apical tuft, which has previously been discussed as a derived adaptation for precision grasping (Almecija et al., 2010). The combination of these morphological features would have facilitated a forceful pad-to-pad precision grasp like that seen in modem humans (Kivell et al., 2015). In contrast to this suite of derived metacarpal traits, the phalanges of H. naledi retain a primitive morphology, being relatively long and exhibiting a marked curvature. Kivell et al. (2015) note the mean phalangeal curvature seen in H. naledi is nearly the same as that seen in Australopithecus afarensis and does not vary significantly
24


from the curvature exhibited in extant African ape phalanges. The long, curved phalanges seen in H. naledi provide strong evidence that this species spent considerable time in the trees and was adapted to arboreal locomotion. However, the combination of derived wrist structure with primitive phalangeal morphology suggests that H. naledi represents a hominin that practiced arboreal locomotion while possessing the ability to perform enhanced precision grips, which is in contrast with anything previously known from the hominin fossil record (Kivell et al., 2015).
Homo neanderthalensis. Early research into Neanderthal hand morphology, particularly in regards to the thumb, concluded that their thumbs were too short to have facilitated precise manipulative capabilities (Niewoehner et al., 2003). Vlcek (1975) argued that Neanderthals would have been incapable of fine precision grasping based on their thenar muscle morphology. Vlcek (1975) argued Neanderthals would have been incapable the same precision grip seen in modem humans based on the placement of the opponens pollicis muscle, which he stated was shifted too far to the palmar side of the bone to allow for full medial rotation of the thumb. He based his conclusions on specimens from the cave of Kiik-Koba, which is a Mousterian site in Ukraine first excavated in the early 20th century (Vleck, 1973). However, Niewoehner (2006:165) states there is, no anatomical or functional basis for the...claims that Neanderthal thumbs were less mobile than those of modem humans.
Neanderthal thumbs are in fact highly derived for precision grasping, and subsequent digital analyses concluded there is nothing in their morphology which would have precluded them from achieving the same degree of manual dexterity as that seen in modem humans (Niewoehner et al., 2003). In a broad of study Neanderthal manual remains, Niewoehner (2006) concluded Neanderthal fingers showed evidence of hypertrophied muscles, which, in
25


conjunction with other morphologies of the hand, meant they were most likely capable of very powerful grasps. Both the opponens pollicis, which is responsible for much of the opposition of the thumb, and the opponens digiti minimi, which aids in radial rotation of the fifth metacarpal during cupping movements of the hand, showed evidence of hypertrophy. Niewoehner (2006) also noted enlarged tubercles on the trapezium, hamate, and scaphoid, which he notes conferred increased mechanical advantages in the carpometacarpal and carpophalangeal joints. Additionally, the distal pollical phalanges of Neanderthals show evidence of a radioulnarly expanded apical tuft like the dimensions seen in modem humans, which has been previously noted as aiding in precision grasping of the hand (Trinkaus,
1983). This suite of morphological features would have facilitated powerful, precision grasping in Neanderthals.
Morphology of the last common ancestor. There is currently no known specimen representing the LCA of humans and chimpanzees. As such, hypotheses regarding the morphology of the LCA must rely on inferences made based on the known morphological similarities between extant non-human apes, humans, mid fossil hominin specimens. Tocheri et al. (2008) argue that, based on parsimony, the hand of the LCA most likely resembled that of extant African apes, rather than resembling the more arboreal Asian apes. Following this argument, it can be inferred that all hand morphologies shared by extant great apes and non-hominid outgroups are homologous and were most likely present in the hand of the LCA (Tocheri et al., 2008). Table 1 lists a summary of some of the morphological and myo logical features that Tocheri et al. (2008) hypothesize to have been present in the LCA, mid their state within modem humans.
26


Table 1. Inferred morphological and myological hand features of the LCA and their state in modem humans. From Tocheri et al., 2007 and Larson et al., 2009.__________________
Feature Inferred character state of LCA Character state in modem humans
Relative finger length Long fingers relative to thumb Thumb is long relative to other digits
Proximal phalangeal curvature Curved dorso-palmarly Straight dorso-palmarly
Mel robusticity Gracile Robust
Apical tufts (distal phalanges) Narrow Broad
Flexor pollicis longus Absent or reduced with no separate tendon belly Separate tendon belly
Opponens pollicis Most likely occupies a relatively small section of bone Relatively enlarged when compared to Pan
Neck of capitate Capitate has a waisted neck on the radial side* Expanded appearance on the radial side
Styloid process on Mc3 Absent Present
*This feature is also seen in the australopiths and Homo floresiensis, as discussed in an earlier section of this paper.
Theories for the Evolution of the Modem Human Hand
The rise of habitual bipedality, which freed the hands from locomotor restraints, has generally been credited as the catalyst for the evolution of modem human hand morphology (Susman, 1994; Kivell, 2015). Habitual arboreal locomotion is characterized by long, curved phalanges which inhibit precision grasping and manual dexterity (Marzke, 1997). As hominins moved out of the trees and increasingly practiced a terrestrial, bipedal lifestyle, the need for longer, curved phalanges lessened, and the hands instead became increasingly
27


adapted for manipulation, as opposed to suspensory aids (Marzke, 1997). Evidence from hominin remains indicates that the morphological features associated with stone tool usage (broad apical tufts, a long thumb relative to digits, less phalangeal curvature, etc.) arose before the advent of stone tools, likely as a response to increased manipulation related to other activities such as feeding (Kivell, 2015). Recently, it has also been proposed that the modem human hand configuration evolved as a pleiotropic by-product of the need for shorter toes to facilitate bipedal locomotion (Rolian, 2009). As new fossil hominin finds come to light, new evidence suggests the modern layout might in fact be quite primitive, and more like early Miocene apes (Almecija et al., 2010). If this is the case, the hand morphology seen in extant apes and early hominins, characterized by long, curved fingers and a relatively short thumb, would in fact be the derived character state. The following section will further discuss the theories for the evolution of modem human hand morphology, including the rise of habitual stone tool usage as a driver for enhanced precision grasping capabilities.
The Rise of Habitual Stone Tool Use
The shift from an arboreal lifestyle to habitual bipedalism has long been viewed as the catalyst that drove the evolution of modem human hand morphology (Susman, 1994; Panger et al., 2002). As humans increasingly moved about on two feet, the need for long, curved phalanges lessened, leaving the hands free for other, non-locomotor, duties. Likely, the initial driver for increased precision grasping and manipulative capabilities began with foraging behavior, as hominins began to exploit new resources for food (Almecija and Alba, 2014). Extant chimpanzees (Pan troglodytes) have long been known to employ branches for extractive foraging purposes, and it has been proposed that early hominin tool use began in such a fashion, later shifting to stones as a primary tool source (Carvalho et al., 2012).
28


However, these organic materials do not preserve in the fossil record, mid so at present, studies must focus on preservable materials such as lithic and faunal remains to infer the advent of stone tool usage within the hominin lineage.
The discovery at Dikika, Ethiopia of two cut-marked bones push back the emergence of stone tool usage to circa 3.4 Ma, roughly 800 ka earlier than previously thought, though the alleged cut marks on the Dikika bones remains controversial (Semaw et al. 2003; McPherron et al. 2010). The stone tools from the site of Gona, Ethiopia, provide the earliest definitive proof of stone tool usage only within the hominin lineage at roughly 2.5 Ma (Semaw et al., 2003). However, this still begs the question of why stone tool usage only arose within the hominin lineage and not in other primates. Theories abound on this subject, and there is no room in this manuscript for a full review (but see Panger et al., 2002; Sanz and Morgan 2013) for a more thorough review of the matter). For the sake of brevity, the discussion of why tool use arose only within hominins will be limited to one possible theory. The Hand is a Pleiotropic Result of Selection for the Foot
Derived morphological traits within the hand first appear within the hominin fossil record around roughly the same time as adaptations for bipedality (Richmond and Jungers, 2008; Rohan et al., 2011). Or (ca. 6 Ma), whose manual elements represent the earliest hominin hand remains to date, exhibit derived distal phalanges, in conjunction with femoral morphology indicative of a biped (Almecija et al., 2010; Almecija et al., 2013). Or. tugenensis possessed broad apical tufts related to precision grasping, demonstrating these morphological adaptations arose quite early within the hominin lineage. The early emergence of these derived features, coupled with the nearly homologous structures and similar developmental pathways of manual and pedal elements, led to the theory that the adaptations
29


coevolved (Rohan, 2009).
If this hypothesis is correct, the growth of the hands and feet of extant primates would have to follow the same ontogenetic pathways and produce similar phenotypic results. Further, hominin fossils should exhibit similar proportions in the pedal and manual phalanges as is seen in modem humans. The results of the study conducted by Rohan et al. (2009:1563) demonstrated that, in Homo sapiens mid Pan troglodytes, the variation in growth of the manual and pedal phalanges is constrained by the same developmental blueprint, supporting the hypothesis that their ontogeny within primates is linked. The hominin fossil record, however, is currently too sparse to provide either positive or negative proof of their hypothesis. The data do show that by the time of Au. afarensis (ca. 3.5 Ma), manual and pedal proportions within hominins were intermediate between modem humans mid chimpanzees. Additionally, by the time of early Homo (ca. 1.8-1.5 Ma), hominin hand mid foot proportions appear to be quite like those seen in modem humans (Rohan et al., 2009).
This theory contrasts with the longstanding belief that human hands evolved after the shift to habitual bipedalism, independent of the selective pressures acting upon pedal morphology (Marzke, 2013). Rohan et al. (2009) contend their multiple lines of evidence, including genetic, fossil, mid, archaeological, disprove the previous theory and offer a new proposal for the evolution of modern manual morphology. The fossils of Orrorin tugenensis, which demonstrate that derived traits within the hominin hand emerged much earlier than the evidence for stone tool usage (McPherron et al., 2010), make the theory by Rohan et al. (2009) particularly tempting to accept. It is becoming increasingly clear that the morphological traits previously assumed to correlate with precision grasping and stone tool manufacture (Susman, 1994; Marzke, 1997) appeared much earlier than previously thought,
30


mid most likely evolved within the hominin lineage independently of stone tool usage or manufacture.
Conclusion
The hands of living and fossil primates display a wide range of morphological variation, and this is especially true for the pollical metacarpal. The long, robust thumbs of modem humans are adapted for enhanced precision grasping, as reflected in the broad shaft, expanded proximal bases, and broad distal pollical phalanges. In contrast, non-human primates generally possess more gracile metacarpals and longer fingers relative to their thumbs (Napier, 1962a; Susman, 1979; Marzke, 1997). The morphological evidence from the hominin fossil record shows the pollical metacarpal gradually became more robust over the course of human evolution. Australopithecus possessed gracile pollical metacarpals with narrow proximal bases (Susman, 1979; Berger et al., 2010), suggesting this is the primitive character state for hominin pollical metacarpals. By the time of H. neanderthalensis and early anatomically modem humans, the pollical metacarpal had become much more robust, with a distinct crest for the opponens pollicis and a broad TCMJ facet (Niewoehner, 2001; Niewoehner, 2006).
What remains unclear is in what sequence these morphological adaptations arose in the genus Homo. Did adaptations in the thumb of primitive Homo begin with an expansion of the proximal facet, greater recmitment of the intrinsic musculature, resulting in a broader base and more prounced muscle flanges, or did these features evolve in unison? The fossil remains from H. naledi afford the opportunity, for the first time, to study the pollical metacarpals of primitive Homo to better understand this timeline. The following chapters present a geometric morphometric analysis of a broad comparative sample of extant and
31


fossil primate pollical metacarpals aimed at quantifying the shape of the H. naledi first metacarpal to answer the question of whether if. nadedi represents an autapomorphic taxon in terms of pollical metacarpal morphology, or whether it is evidence of a transitional metacarpal morphology in primitive members of Homo.
32


CHAPTER II
MATERIALS AND METHODS Materials
This study includes 3D virtual renderings of Homo sapiens (n=178 [(Table 2]), nonhuman apes (n=86 [Table 3]), Old-World cercopithecine monkeys (n=59 [Table 4]), and fossil hominins (n=14[Table 5]). All left metacarpals were reflected in Geomagic prior to landmarking to create a full right-side sample. While some extant specimens were labelled by sex, time limitations prohibited any analyses based on sex. Additionally, since many specimens could not be identified by sex, analyses based on sex would have been possible only for a very reduced number of specimens. All scans were provided courtesy of Drs.
Caley Orr, Matthew Tocheri, Biren Patel, and Tracey Kivell. Figure 7 shows a 3D representation of the extant hominoid sample. Figure 8 shows a 3D representation of the cercopithecine sample. Figure 9 shows a 3D representation of the fossil hominin sample.
Table 2. Homo sapiens sample.
Population Male Female Unknown sex Total
.African American 6 11 0 17
Aleut/Pre-Aleut 0 0 15 15
Australian 0 0 8 8
Chinese 0 0 24 24
Europe mi 10 4 7 21
European American 12 12 0 24
Indigenous African 11 7 4 22
Indigenous American 0 10 10
Japanese 2 1 0 3
Nub ianEgypti an 6 3 0 9
33


Unknown 0 0 25 25
Total 47 38 93 178
Table 2 >. Non-human ape sample.
Taxon Male Female Unknown sex Total
Gorilla beringei 5 5 1 11
Gorilla gorilla 15 10 0 25
Hylobates lar 7 4 0 11
Pan paniscus 9 7 0 16
Pan troglodytes 11 9 3 23
Pongo sp. 7 4 0 11
Total 54 39 4 97
Table 4. Cercopithecine monkey sample.
Taxon Male Female Unknown Total
Erythrocebus patas 7 5 0 12
Macaca fasicularis 9 1 0 10
Nasalis larvatus 6 3 3 12
Papio sp. 4 5 5 14
Total 26 14 8 48
34


Table 5. Fossil lominin sample.
Specimen Total
Australopithecus afar crisis A.L.333 1
Australopithecus sediba MH2 1
Homo naledi U.W.-101-270 U.W.-101-282 U.W.-101-1321 U.W.-101-1641 4
Homo neanderthalensis Kebara 2 La Chap el le La Ferrassie* Regourdou Shanidar 4 6
Homo sapiens Qafzeh 9 1
SK 84** 1
*Both left and right metacarpals from this specimen were used. The left metacarpal was mirrored in Geomagic to reflect a right. **This specimen has been variously attributed to
Paranthropus (Susman, 1988), Homo, and Australopithecus (Kivell, 2015).
35


Hylobates Pongo Pan
lar pygmaeus troglodytes
Gorilla
beringei
Homo sapiens
Figure 7. Representative 3D surface models of the hominoid sample.
Erythrocebus patas Macaca fasicularis Nasalis larvatus Papio anubis
Figure 8. Representative 3D surface models of the cercopithecine sample.
36


H.
neanderthalensis Homo naledi Au. sediba Au. afarensis
(La Chapelle) (U.W. 1321) SK 84 (MH2) (A.L. 333)
Figure 9. Representative 3D surface models of the fossil hominin sample.
Methods
Using the Patch tool in the Landmark Editor Software (Wiley et al., 2005; Wiley,
2007), nine landmarks were placed across Mel palmar diaphyseal shaft and a 20x20 semilandmark grid was placed across the palmar diaphyseal surface (Fig. 10; Table 6). From the nine landmarks, a 20x20 semilandmark grid was designated across the Mel palmar diaphyseal surface. Fandmarks are point locations that must be biologically homologous between specimens, but for many biological structures, such as the muscle attachments on the pollical metacarpal, landmark positions cannot be made homologous between structures because of the curves of those surfaces. For these analyses, one must use semilandmarks, which allows for the analysis of two- or three-dimensional curves and surfaces (Gunz and Mitteroecker, 2013). Gunz and Mitteroecker (2013) argue it is better to have densely spaced landmarks to capture the greatest detail in terms of morphology, as well as to estimate for
37


missing data. Landmark placement was chosen to best capture the overall morphology of the palmar shaft.
Figure 10. The location of the landmark and semilandmark data points onafi sapiens right Me 1. The numbered yellow dots indicate the original nine landmarks placed on the Mcl surface and the small red dots indicate the semilandmark grid.
38


Table 6. Anatomical description of the landmark placement.
Landmark Number Landmark Placement
1. Most lateral point on the distal end of the shaft
2. Midpoint of the palmar articular margin on the head
3. Most medial point on the distal end of the shaft
4. Most lateral point at midshaft
5. Center of the palmar shaft
6. Most medial point at midshaft
7. Most lateral point of the base
8. Apex of the palmar beak
9. Most medial point of the base
Following Scott (2015), of the nine original landmarks placed across the palmar diaphyseal surface, the four comer points (i.e, landmarks 1, 3, 7, and 9 in Fig. 10) were selected to anchor the sliding semi landmarks, resulting in 396 evenly spaced total semilandmarks. During superimposition, the 396 semilandmarks were slid along the palmar diaphyseal surface by minimizing Procmstes distances (Gunz and Mitteroecker, 2013). Sliding of the semilandmarks ensures optimal spacing between landmarks and semi landmarks, and establishes geometric correspondence of the semilandmarks by removing the arbitrary effects of initial spacing (Gunz and Mitteroecker, 2013). Raw landmark data contains information on size, shape, mid orientation, and so to correct for these factors, the data were transformed into shape variables (Gower, 1975; Rohlf and Slice, 1990; Gunz and Mitteroecker, 2013) using a generalized Procmstes analysis (GPA) performed in the geomorph package in R Console (Adams and Otarola-Castillo, 2013; R Core Team,
2016). The resultant semilandmark data are the Procmstes shape coordinates, which contain
39


information on the original shape of each landmark configuration by specimen (Schroeder et al., 2017).
A principal component analysis (PCA) was then performed on the variance/covariance matrix of the Procrustes shape coordinates to visualize the shape differences between specimens mid to identify the major axes of variation using the geomorph package in R console (Dryden and Mardia, 1998; R Core Team, 2016; Schroeder et al., 2017). The deformation grid of shape change across principal components was done using the geomorph package in R Console. An ANOVA and Tukeys post-hoc pairwise comparison were performed in the Statistica software package to determine between-group differences along principal components axes. Finally, to explore the effects of overall metacarpal size on morphology, a multivariate regression analysis of log centroid size on shape variables was computed by plotting multivariate regression scores against log centroid size. Regression scores predict the location of each individual on the component (DiStefano et al., 2009). Because there was no direct measure for body mass of the individuals, log centroid size was used as a proxy for overall body mass (Parr et al., 2011; Knigge et al., 2014). Reduced major axis lines were computed using the RMA software for reduced major axis regression (Bohonak, 2004) and fitted using the program Statistica.
40


CHAPTER III
RESULTS Full Sample
This section will discuss the results of the principal components analyses of the full sample for Principal Component (PC) 1-3. PC scores beyond 3 represented less than 5% of the total variation within the sample, and so were excluded from discussion.
PCI shape and groupings
PCI captures 38% of the variation within the sample, and represents the variation in shaft breadth between taxa. The first principal component is most useful in separating out the great apes (except for Pan) and the younger fossil hominin sample (Neanderthals, SK 84, and Qafzeh 9) from the Old World monkeys and the australopiths. The plot of PC2 on PCI (Fig. 11) shows that the negative side of PC 1 represents taxa with broader shafts, including H. sapiens, H. neanderthalensis, and Gorilla sp., while the positive end shows taxa with more gracile shafts, such as Pan md Hylobates. The Old World monkeys (Erythrocebus, Macaca, Nasalis, and Papio) as well as the Pongidae, all occupy shape space on the far positive side of PCI, reflecting the gracile first metacarpals of these primates. H. naledi occupies shape space on the positive end of the axis, though they are not as far on the positive end as the Old World monkeys and Pan. Au. sediba mid Au. afarensis both occupy shape space on the positive end of the axis, reflecting their gracile shaft morphology, like what is seen in Pan.
SK 84 and Qafzeh 9 both occupy shape space on the negative side of the axis, clustering closely with if. neanderthalensis and modem humans.
41


Gorilla A H. sapiens # Pan
Full Sample PCI vs.PC2
A Hylobates
Pongo
I Macaca A Erythrocebus
+ Nasalis
Qafzeh 9
I H.neanderthalensis
SK 84
+" H. naledi
a !*
A A
* *
- A v
+ i A A,
Figure 11. Scatterplot of PCI on PC2.
PC2 shape and groupings
PC2 captures 8% of the variation within the sample, and represents the size of the base relative to the breadth of the shaft. Shape space on the negative end of the PC2 axis is occupied by much of the Gorilla sample, as well as roughly half of the Pan sample. Except for some of the Papio specimens, the cercopithecines all occupy shape space on the negative end of the PC2 axis
42


H. neanderthalensis, Qafzeh 9, SK 84, mid H. naledi all occupy shape space solely on the positive end of the PC2 axis. H. naledi occupies shape space almost completely separately from all other taxa with only minimal overlap with H. sapiens and H. neanderthalensis. Pongo, Hylobates, mid modem humans all occupy shape space on both the positive mid negative ends of the axis. Taxa on the negative side of PC2 possess a shaft with a broader distal end and a relatively narrower proximal facet. On the positive side of PC2 are taxa with a broader proximal base mid a narrower distal end. The variation in breadth of the distal end of the shaft is most likely the result of variation in size of the muscle flanges. H. naledi and H. neanderthalensis, both on the positive end of PC2, have very pronounced crests for the opponens pollicis muscle, while taxa on the negative end of PC2, such as the cercopithecine monkeys, have greatly reduced muscle markings for the opponens pollicis muscle.
PC3 shape and groupings
PC3 (Fig. 12) captures 5% of the variation within the sample, mid represents the breadth of the proximal end relative to the medial and lateral flanges on the distal portion of the shaft. Shape space on the negative side of PC3 represents taxa with more of an hourglass or waisted figure (H. neanderthalensis, SK 84, Au. sediba, Pongo, mid most of the Pan and Nasalis sample), while taxa on the positive end are characterized by a more proximodistal uniform shape of the first metacarpal (Papio, Erythrocebus, Macaca, and Hylobates). As in PCI and PC2, modem humans occupy shape space on both the positive mid negative end of the PC3 axis. H. naledi occupies shape space on both the negative mid positive end of PC3.
43


Figure 12. Scatterplot of PCI on PC3.
Regional Variation with Modern Humans
The recent modern human sample (n=178) includes specimens from several geographic regions (Table 1). To assess morphological variation among the populations of recent H. sapiens, a PCA was conducted on the modern human subsample. PCI represents 26% of the variation within the sample, PC2 represents 11% of the variation, and PC3 represents 9% of the variation. PC scores beyond 3 represented less than 5% of the total variation, and so were omitted from analysis.
44


A one-way ANOVA was performed on the results from the PCA to test for
morphological variation between regional groups of modem humans. The results of the oneway ANOVA show significant differences on PCI and PC2. Table 7 shows the results of the one-way ANOVA for the modem human sample. A Tukeys post-hoc pairwise comparison was performed to find the source of variation between groups for PCs 1 and 2 (Tables 8 and 9). Values that are statistically significant (p<0.05) are denoted.
PCI PC2 PC3
/?=0.00 /?=0.00 p> 0.05
Red designates statistically significant at the p<0.05 level.
Table 8. Results of Tukeys post-hoc pairwise comparison results for PCI for the modem human subsample.
Population 1 2 3 4 5 6 7 8 9 10 11
Unknown (1) 1.000 0.992 0.982 0.029 0.986 0.557 0.999 0.986 0.564 0.998
Japanese (2) 1.000 1.000 1.000 0.942 1.000 0.996 0.999 1.000 0.998 0.999
African American (3) 0.992 1.000 1.000 0.648 1.000 0.988 0.875 1.000 0.995 0.709
Indigenous African (4) 0.982 1.000 1.000 0.568 1.000 0.986 0.829 1.000 0.994 0.584
European American (5) 0.029 0.942 0.648 0.568 0.570 1.000 0.035 0.962 0.999 0.001
European (6) 0.986 1.000 1.000 1.000 0.570 0.985 0.844 1.000 0.994 0.619
Indigenous American (7) 0.557 0.996 0.988 0.986 1.000 0.985 0.338 1.000 1.000 0.149
Nubian Egyptian (8) 0.999 0.999 0.875 0.829 0.035 0.844 0.338 0.861 0.353 1.000
Australian (9) 0.986 1.000 1.000 1.000 0.962 1.000 1.000 0.861 1.000 0.760
Aleut/Pre-Aleut (10) 0.564 0.998 0.995 0.994 0.999 0.994 1.000 0.353 1.000 0.124
Chinese (11) 0.998 0.999 0.709 0.584 0.001 0.619 0.149 1.000 0.760 0.124
Red designates statistically significant at the /?<0.05 level.
45


Table 9. Tukeys post-hoc pairwise comparisons results for PC2 for the modem human subsample. _________________________________________________________________
Population 1 2 3 4 5 6 7 8 9 10 11
Unknown (1) 0.999 0.958 1.000 1.000 1.000 0.647 1.000 0.739 0.000 1.000
Japanese (2) 0.999 1.000 1.000 0.999 0.984 0.629 1.000 0.679 0.008 0.996
African American (3) 0.958 1.000 0.996 0.970 0.691 0.102 1.000 0.152 0.000 0.858
Indigenous African (4) 1.000 1.000 0.996 1.000 0.995 0.464 1.000 0.566 0.000 1.000
European American (5) 1.000 0.999 0.970 1.000 1.000 0.609 1.000 0.705 0.000 1.000
European (6) 1.000 0.984 0.691 0.995 1.000 0.940 0.991 0.965 0.001 1.000
Indigenous American (7) 0.647 0.629 0.102 0.464 0.609 0.940 0.542 1.000 0.471 0.808
Nubian Egyptian (8) 1.000 1.000 1.000 1.000 1.000 0.991 0.542 0.617 0.000 0.999
Australian (9) 0.739 0.679 0.152 0.566 0.705 0.965 1.000 0.617 0.481 0.871
Aleut/Pre-Aleut (10) 0.000 0.008 0.000 0.000 0.000 0.001 0.471 0.000 0.481 0.000
Chinese (11) 1.000 0.996 0.858 1.000 1.000 1.000 0.808 0.999 0.871 0.000
Red designates statistically significant at the /?<0.05 level.
Reduced sample
The full sample is quite large, and so the decision was made to attempt to narrow down the sample size for further analyses. Based on the results of the full sample PCA analysis, it appeared Pan, Pongo, Hylobates and the cercopithecines all shared a similar morphology. These taxa all occupy the same shape space on PC 1-3 as the Australopithecus sample, and it is likely they reflect the primitive condition for hominins. As such, the author decided to let Pan represent the primitive condition for hominids and to exclude Pongo, Hylobates, and the cercopithecines from subsequent analyses if no significant variation was
found to exist between the cercopithecines, Pan, Pongo, and Hylobates.
A one-way ANOVA was performed to determine if variation existed between the
46


cercopithecines and the hominid sample. The results show significant variation across PC 1-3 (Table 10). A Tukeys post-hoc pairwise comparison was then performed to determine the source of the variation between the groups (Table 11). Results from the Tukeys post-hoc test showed no variation between Pan, Pongo, Hylobates, and the cercopithecines for PCs 1 and 2. While there is some variation between Pan, Hylobates, Pongo, and the cercopithecines on PC3, the decision was made to narrow down the sample size for further analyses to better focus on more closely related taxa. A one-way ANOVA was then performed to examine variation between the hominid sample. The results showed significant variation between taxa on PC 1-3 (Table 12). A Tukeys post-hoc pairwise comparison was then performed to determine between which taxa there was statistically significant variation (Table 13).
Table 10. Results of a one-way ANOVA for the Cercopithecoidea and hominid sample._________________________________________________________________________
PCI PC2 PC3
/?=0.00 /?=0.00 /?=0.00
Table 11. Tukeys post-hoc pairwise comparison results for PCs 1-3 for the Cercopithecoidea and hominid samples.__________________________________________________________________
Genus Pairing PCI PC2 PC3
Cercopithecidae-Hylobates Not significant Not significant Not significant
C e r c op ith e ci d a e Gorilla p<0.05 p<0.05 p <0.0.5
C e r c o p i th e ci d a e Horn o sapiens p<0.05 p<0.05 Not significant
C e rc op ith e ci d a e Pan Not significant Not significant p <0.0.5
Cercopithecidae-Pongo Not significant Not significant p <0.0.5
47


Table 12. Resuli ts of a one-way ANOVA for t le hominid samples.
PCI PC2 PC3
/?=0.00 /?=0.00 /?=0.00
Red designates statistically significant at the /?<0.05 level.
Table 13. Tukeys post- ioc pairwise comparison results for PCs 1-3 for t re hominid sample.
Genus Pairing PCI PC2 PC3
Homo sapiens-Hylobates /?<0.05 Not significant Not significant
Homo sapiens-Gorilla Not significant /?<0.05 /?<0.05
Homo sapiens-Pan /?<0.05 /?<0.05 /?<0.05
Homo sapiens-Pongo /?<0.05 Not significant /?<0.05
Red designates statistica ly significant at the p<0. 05 level.
Hominoid sample
The hominoid sample includes all specimens of Gorilla sp., Pan sp., H. sapiens, H. neanderthalensis, Au. afarensis, Au. sediba, SK 84, Qafzeh 9, and if. naledi. In the principal component analysis, PCI (Figs. 13 and 14) explained 33% of the total variance, PC2 (Figs.
14 and 15) explained 10% of the total variance, and PC3 (Figs. 16 and 17) explained 6% of the total variance. PC scores beyond 3 represented less than 5% of the total variation within the sample, and so were excluded from this discussion.
PCI shape and groupings
PCI captures 33% of variation within the hominoid sample, and is most informative for distinguishing between extant apes (Figs. 13 and 14). The negative side of the PCI axis is mostly occupied by Homo sapiens, Gorilla, and Neanderthals, while Pan and Hylobates occupy the positive side of the axis. Gorilla beringei and Gorilla gorilla show a clear
48


distinction between the two species in shape space, with Gorilla beringei clustering more on the negative side of the axis, and Gorilla gorilla occupying more space on the positive side of the axis. Qafzeh 9 lots squarely within the Homo sapiens sample. SK 84, also plots within the range of Homo sapiens. Both Au. afarensis and Au. sediba occupy the same shape space as Pan and Hylobates on the first principal component axis, while Homo naledi occupies the shape space on the border between Homo sapiens and Pan. Figure 14 shows a scatterplot of PC2 on PCI for the hominoid sample
Figure 13. The morphological variation in the hominid sample across PCI.
49


Hominoid Sample PC1 vs. PC2
0.05 0.04 0.03 0.02 0.01 0.00 -0.01 -0.02 -0.03 -0.04 -0.05
-0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08
PC 1(33%)
H Gorilla gorilla
Gorilla beringei Homo sapiens
# Pan troglodytes
# Pan paniscus
Qafzeh 9
Neanderthal
SK 84
+ Homo naledi A Au.sediba A Au. afarensis
A
AA
ledi AAA A . A
* I/* * \
A J* A aV J 4a a a
* ** ^
* * \ r*
A
A A
AA"
i vv^a
A .*
tW14
#
A
A H

/
Figure 14. Scatterplot of PC2 on PCI.
The positive side of the first PC axis represents taxa with a narrower metacarpal shaft, both mediolaterally and on both the proximal and distal ends. The positive side of the PC 1 axis is occupied by Pan and the australopiths, reflecting their more gracile musculature compared with the other extant apes and humans. Homo naledi also plots well within the positive side of the PC 1 axis, suggesting the seemingly broader distal end of the metacarpal
50


is a result of greater surface area for muscle entheseal development, rather than a broader overall shaft. H. naledis placement on the positive end of PCI indicates that PCI is not capturing muscle robusticity, as it might appear from qualitative observations, but rather only the overall range of breadth within the shaft itself The negative side of the PCI axis is mostly occupied by Gorilla, H. sapiens, and H. neanderthalensis, all of which have broad metacarpal shafts reflective of relative degrees of muscle development. PCI also appears to capture something related to the size of the proximal mid distal art icular facets, with those taxa on the negative end having larger a surface area for the articulation with the trapezium. PC 2 shape and groupings
PC2 (Figs. 14 mid 15) captures 10% of the variation within the hominoid sample. Both Gorilla gorilla and Gorilla beringei cluster almost entirely on the negative end of the second principal component axis. Both species of Pan cluster mostly on the negative end of the axis. The Neanderthal sample occupies shape space on the positive end of the axis.
Qafzeh 9 and SK 84 both occupy shape space on the slightly positive end of the PC2 axis.
Au. sediba occupies shape space on the positive end of the axis, while Au. afarensis occupies shape space on the negative end of PC2. Homo naledi occupies shape space squarely on the positive end of the second principal component axis.
Shape on the PC2 axis represents the breadth of the proximal end of the shaft relative to the overall size of the shaft, as well as narrowing of the beak medio laterally. The impression of a narrowing and tapering of the shaft is most likely a result of the degree of muscle hypertrophy, which gives the false impression of a broad overall shaft. PC2 captures robusticity of the first metacarpal, which is shown in the position of both Homo naledi mid the Neanderthals. H. naledi is unique in having a small base relative to large, laterally flaring
51


flanges, while Neanderthals have a relatively large base and large flanges.
The negative portion of the PC2 axis represents taxa with a broader overall shaft, as well as more narrowing of the beak in a lateral-to-medial direction. The taxa on the negativeside of the axis all possess robust musculature (opponens pollicis and first dorsal interosseous), which serves to make the shaft broader on the medial and lateral sides of the bone. The positive side of the second principal components axis represents taxa with narrower overall shafts and lesser development of the first dorsal interosseous and opponens pollicis muscles. The shape of the positive side of the second principal component also represents taxa with almost no narrowing of the metacarpal beak, as is seen on the negative
end of the second principal components axis. Figure 15 illustrates morphological variation across PC2 for the hominoid sample.
a!
Gorilla du. afarensis Homo H. Homo
beringei (A.L. 333) sapiens at neanderthalensis naledi
(0,0) (LaFerrassie)
+
Figure 15. The morphological variation in the hominoid sample across PC2.
52


PC3 shape mid groupings
PC3 captures 6% of the variation within the hominoid sample, and is most informative in capturing the breadth of the proximal base relative to the laterally flaring distal flanges associated the opponens pollicis and first dorsal interosseous muscle. PC3 is the only PC in which Homo naledi does not occupy shape space separate from any other taxa, but rather occupies the same shape space as Pan, H. sapiens mid Gorilla gorilla. As in PCI, Homo sapiens occupy shape space across most of both axes, reflecting the varied morphology of modem human hands. PC3 separates out the Neanderthal specimens from other taxa. Pan occupies space mostly on the positive side of the third PC axis, along with Hylobates, which occupies space solely on the positive side of the axis. Au. sediba occupies space on the positive side of the axis, while Au. afarensis occupies space on the negative end.
The positive side of PC3 represents taxa with wide bases above which there is narrowing before flaring out into wide distal flanges, thus creating a waisted shape. Neanderthals occupy shape space exclusively within the positive side of the PC3 axis, reflecting their unique morphology of wide proximal bases as well as laterally flaring distal flanges, mostly related to the opponens pollicis muscle. The negative end of PC3 seems to represent taxa with similarly broad proximal and distal ends, lending a more straight down look to the shaft, in contrast to the impression of waisting on the negative end of the axis. Figure 16 shows morphological variation across PC3 for the hominoid sample. Figure 17 shows a scatterplot for PC3 on PCI for the hominoid sample.
53


Figure 16. The morphological variation in the hominoid sample across PC3.
54


Hominoid Sample PC1 vs. PC3
0.05
0.08
Figure 17. A scatterplot of PC3 on PCI for the hominoid sample.
Shape and size
A multivariate regression analysis of regression scores against log centroid size was perfonned for the entire dataset. Results of the multivariate regression analysis showed no significant relationship between shape and size (p= 0.06). None of the taxa plot along the
55


same allometric trajectory as each other, indicating that a change in size would not induce a change in metacarpal shape. Figure 18 shows the results of the multivariate regression analysis.
0.25
Hominoid Multivariate Regression Analysis
0.20
0.15
0.10
co
o 0.05
0.00
-0.05
-0.10
Gorilla
A Homo sapiens
Pan
Hylotxnes
Pongo
Cercopithecld
Qafzeh 9
Neanderthal
SK 84
+ Homo noted! A Australoptth
-0.15
4.2 4.4 4.6 4.8 5.0 5.2
Log Centroid Size
5.4 5.6
5.8
Figure 18. A plot of the multivariate regression analysis showing the reduced major axis
lines.
56


CHAPTER IV
DISCUSSION AND FUTURE RESEARCH
This study used geometric morphometric techniques to quantify the shape of the first metacarpal in a broad comparative sample of extant and fossil primates. The aim of the study was to determine if H. naledi represents an autapomorphic taxon in terms of first metacarpal morphology. The results presented here indicate that the first metacarpal of H. naledi is not autapomorphic, but rather represents a transitional morphology for primitive members of the genus Homo.
PCl-Breadth of the Metacarpal Shaft
The PCA results demonstrate that taxa can best be differentiated by the overall breadth of the first metacarpal shaft. In both the full sample and the reduced hominoid sample, the most discriminating factor was the overall breadth of the metacarpal shaft. Breadth of the shaft appears to be tied to two major behavioral components: locomotor patterns and manipulator behaviors. On the negative end of the PCI axis are specimens (Homo sapiens, Homo neanderthalensis, Gorilla beringei, SK 84, mid Qafzeh 9) that regularly use their thumbs for manipulator purposes, whether through use of stone tools (Susman, 1988; Niewoehner, 2001), or for breaking down fibrous leaves and pith for consumption (Rogers et al., 2004; Rothman et al, 2007). Except for Gorilla beringei, these taxa locomote bipedally, mid their thumbs are free for use exclusively as a means of manipulation. However, gorillas do not use the pollical metacarpal when engaged in knucklewalking, mid thus their thumbs serve mainly as a manipulator aid during feeding (Tuttle, 1967).
57


Au. sediba and Au. afarensis occupy shape space on the positive end of PCI, reflecting their gracile shafts. Based on the morphology of the australopiths, it appears the primitive character state for hominins is a gracile metacarpal shaft. Additionally, it is likely that the gracile metacarpal shafts seen in Pan and Old World monkeys are plesiomorphic for hominins. H. neanderthalensis, SK 84, Qafzeh 9, and modem humans all occupy shape on the negative side of PCI because of their broader metacarpal shafts, and it is therefore likely that a broader metacarpal shaft is the derived character state for hominins. A broader metacarpal shaft allows for greater development of the intrinsic musculature, and is indicative of taxa that use their thumbs more frequently. It is unclear whether hominins adapted a broader shaft because of increased tool usage (Susman, 1994) or whether this morphology predates the appearance of stone tools.
PC2-Breadth of the Proximal End Relative to Shaft Breadth The second principal component in both the full and reduced analyses captures the variation in the breadth of the base of the pollical metacarpal relative to overall shaft breadth. A broader proximal facet allows more even distribution of compressive loads transferred across the trapeziometacarpal joint surface incurred because of enhanced precision grasping (Marzke, 1997). Gorilla beringei, H. neanderthalensis, and SK 84 all possess the widest proximal bases, suggesting these taxa are especially adapted for incurring large levels of stress at the TCMJ. Marzke et al. (2010) proposed the wider bases observed within Gorilla are a result of the vigorous pulling that occurs during the processing of the fibrous vegetation they regularly consume (Rogers et al., 2004; Rothman et al., 2004). H. sapiens and Qafzeh 9 possess narrower proximal bases than the earlier Neanderthals and SK 84. The reduction in proximal facet breadth could likely be the result of differences in technique employed during
58


stone tool manufacture. Niewoehner (2001) suggests early anatomically modem humans and modem humans adapted to manufacture stone tools in such a way as to alleviate some of the pressure being transferred through the TCMJ. It is possible modem humans manufactured stone tools in a manner that used less powerful grips than were employed by H.
neanderthalensis.
PC3-Breadth of the Proximal End Relative to the Breath of the Distal Flanges
PC3 in both analyses captures the breadth of the metacarpal base relative to the breadth of the distal muscle flanges. The positive and negative ends of PC3 represent variation in levels of adaptations for powerful precision grasping. The positive end of PC3 represents specimens whose metacarpals have a waisted appearance, which is a result of an expanded proximal base combined with broad, mediolaterally flaring distal muscle flanges. The broader muscle flanges are likely a result of increased recmitment of the intrinsic musculature. Increased size of the intrinsic muscles would result in higher levels of compressive loads being transferred through the TCMJ, and the broader proximal facet of the taxa on the positive end of PC3 shows these taxa were adapted for more even distribution of these forces through that joint (Marzke, 1997). Taxa on the negative end of PC3 have a relatively narrow proximal base compared to shaft breadth, mid reduced distal flanges, possibly the result of less well-developed intrinsic musculature.
PC3 is most helpful for pulling out the Neanderthal specimens from the rest of the sample. H. neanderthalensis hands are specifically adapted for powerful precision grasping, as evidenced by the hypertrophy of the thenar muscles mid the broad, expanded bases of the pollical metacarpal (Niewoehner, 2006). Additionally, SK 84 occupies shape space within the Neanderthal sample, suggesting SK 84 was possibly also adapted for powerful precision
59


grasping. While PC3 is most helpful in distinguishing the shape of the Neanderthal specimens from the rest of the taxa, it is important to note the most extreme outlier on PC3 is a Gorilla beringei specimen, (Fig 17). The functional significance of the Gorilla metacarpal morphology is discussed in more detail in a following section.
Implications for the Morphology of Primitive Homo
The only definitively known pollical metacarpals from members of the genus Homo who are thought to represent primitive Homo morphology are those of H. naledi, mid so direct comparison to other fossil hominin specimens is not possible at present. However, the results of this study allow for the formation of hypotheses regarding the morphology of pollical metacarpals of other fossils from primitive Homo. Overall, the hand of H. naledi displays a mosaic of primitive and derived conditions, with primitive, curved phalanges but a derived carpal morphology that is within the range seen in modem humans and Neanderthals (Kivell et al., 2015).
The placement of H. naledi on the first two principal components axes on both the full and reduced sample indicates that the metacarpal shaft is not hyper-robust, as it might appear from initial qualitative observations of the bone. Rather, the shaft itself is gracile, with large flanges superimposed on both the medial and lateral sides. Additionally, the proximal base of H. naledi is narrow, which is a primitive character state shared with Australopithecus. A broad metacarpal base reduces stress across the TCMJ by increasing surface area and minimizing concentrations of force (Marzke et al., 2010). The narrow proximal base indicates H. naledi was not adapted for transferring the increased loads produced for powerful precision grasping. The large flanges are possibly correlated to increased muscle
60


size of the opponens pollicis mid first dorsal interosseous muscles, though this relationship is not definitive (Wallace et al., 2017; Williams-Hatala et ah, 2016).
The morphology of the first metacarpal of H. naledi indicates the character state for primitive members of the genus Homo was a gracile metacarpal shaft, surmounted by broad medial mid lateral flanges. The gracile morphology of the proximal shaft and base is like cercopithecids, Pongo, Pan, andNw. sediba an&Au. afarensis, suggesting that this character is likely plesiomorphic for hominins. In contrast, the broad shaft and metacarpal base, combined with mediolaterally flaring muscle flanges seen in Neanderthals, appear to be an autapomorphy for the species. This hypothesis is supported by the more modem human-like morphology of the Qafzeh specimen, which demonstrates a reduction in both the size of the muscle flanges and the breadth of the proximal base in later Homo specimens. H. naledi represents a transitional state between the overall gracility of the primitive, Australopithecus shaft and the broader, derived shafts of later hominins.
H. naledi demonstrates the piecemeal nature of morphological adaptation within the human hand. The broad flanges indicate greater recruitment of the intrinsic muscles, which, coupled with a derived suite of carpal bones, indicates primitive members of the genus Homo possibly adapted first to produce increased forces, with the expansion of the proximal base following. The broader surface area of the proximal base seems likely to have evolved as a means through which to lessen the forces transferred through the TCM J surface. This suggests the hands of primitive members of the genus Homo were suited for powerful grasping before their joint surfaces could adapt to bear the increased loads being transferred because of the increased muscle mass.
61


Future Research
Variation within Gorilla spp.
Past studies (Rogers et al., 2004; Rothman et al, 2004; Knigge et al., 2014) have documented the morphological differences between eastern (Gorilla beringei) and western (Gorilla gorilla) gorilla populations, which are heavily influenced by the local ecology and habitat of the surrounding area. In a study of diet across six different populations of western gorillas, Rogers et al. (2004:186) found western gorillas have the greatest diversity in their diets of any species of Gorilla, regularly consuming pith, leaves, fruit, though they label pith mid leaves as fallback foods for western gorilla. Western gorillas heavily consume seasonal fruits, with populations from each of the six sites consuming fruits regularly throughout the year. In contrast, the eastern gorilla (Gorilla beringei), inhabits more mountainous regions, where seasonal fruit is either scarce or unavailable. Thus, the diet of eastern gorillas relies heavily on leaves, pith, and stems (Rothman, 2004). Differences in their trophic habits are reflected in the bony morphology of these two groups, with Gorilla beringei having a much broader metacarpal shaft and proximal base than Gorilla gorilla, most likely because of increased use of the thumb due to time spent breaking down pith and fibrous leaves for food consumption. Future research comparing the feeding habits of different species of Gorilla will serve to further illuminate possible reasons behind the metacarpal morphological variation. Additionally, more analyses on different Gorilla skeletal elements will inform as to whether the trend of morphological variation between species of Gorilla continues throughout the skeleton.
Sexual dimorphism
62


Time limitations did not allow for analyses that considered size and shape of the first metacarpal based on sex, though it is very possible differences in metacarpal morphology within the present sample are in some part a result of sexual dimorphism. Gorillas, orangutans, baboons, and proboscis monkeys are the most sexually dimorphic primates, and among the cere op ithe cine monkeys, males tend to be between 30-80% larger than their female counterparts (Plavcan, 2002). Any future analyses on these data should control for sex to explore its possible on shape and size of the first metacarpal.
Conclusion
This work aimed to quantify the shape of the palmar first metacarpal shaft across primate species using a 3D geometric morphometric analysis. The shape of the first metacarpal is influenced by the behavioral repertoire of the organism in question. In primates, the vast differences in forms of locomotor patterns and manipulatory behaviors are reflected in the high levels of morphological variation present across the thumbs of different taxa.
The results of this study demonstrate variation both within and between primate species in regards to the shape of the first metacarpal. The cercopithecines, Pan, and the arboreal apes Hylobates and Pongo are all characterized by gracile metacarpal shafts with reduced muscle entheseal morphology. This morphology is like Australopithecus, suggesting this is the primitive character state for hominin first metacarpals. In contrast, modem humans, Neanderthals, mid Gorilla all possess broader metacarpal shafts with much more defined attachment sites for the intrinsic musculature. There is a pronounced degree of morphological variation between species of Gorilla, which is consistent with previous
63


research (Knigge et al., 2014), and which warrants future research to better understand the behavioral implications behind this variation.
H. naledi is distinguished from all other primates by the combination of a gracile metacarpal shaft, broad flaring muscle flanges, mid a diminutive proximal base. These results suggest that early evolution of the thumb within Homo began with a selection for larger intrinsic musculature, prior to adapting an expanded proximal base to help alleviate the associated increased compressive loads. Based on this study, we expect any subsequent finds of primitive Homo pollical metacarpals to reflect this morphology.
64


REFERENCES
Adams DC, Otarola-Castillo E. 2013. geomorph: an R package for the collection and analysis of geometric morphometric shape data. Methods Ecol Evol. 4:393-399.
Alba DM, Moya-Sola S, Kohler M. 2003. Morphological affinities of thq Australopithecus afarensis hand on the basis of manual proportions mid relative thumb length. J Hum Evol 44(2):225-254.
Almecija S, Moya-Sola S, Alba DM. 2010. Early origin for human-like precision grasping: a comparative study of pollical distal phalanges in fossil hominins. PLoS One 5(7):el 1727.
Almecija S, Tallman M, Alba DM, Pina M, Moya-Sola S, Jungers WL. 2013. The femur of Orrorin tugenensis exhibits morphometric affinities with both Miocene apes and later hominins. Nat Commun 4:2888.
Almecija S, Alba DM. 2014. On manual proportions and pad-to-pad precision grasping in
Australopithecus afarensis. J Hum Evol 73:88-92.
Almecija S, Smaers JB, Jungers WL. 2015. The evolution of human and ape hand proportions. Nat Commun 6:7717.
Berger LR, de Ruiter DJ, Churchill SE, Schmid P, Carlson KJ, Dirks PH, Kibii JM. 2010.
Australopithecus sediba: a new species of Homo-like australopith from South Africa. Science 328(5975): 195-204.
Berger LR, Hawks J, de Ruiter DJ, Churchill SE, Schmid P, Delezene LK, Kivell TL, Garvin HM, Williams SA, DeSilva JM et al. 2015. Homo naledi, a new species of the genus Homo from the Dinaledi Chamber, South Africa, elife 4.
Bohonak AJ. 2004. RMA: Software for Reduced Major Axis Regression. 1.17 ed. San Diego State University.
Brown P, Sutikna T, Morwood MJ, Soejono RP, Jatmiko, Saptomo EW, Due RA. 2004. A new small-bodied hominin from the Late Pleistocene of Flores, Indonesia. Nature 431(7012): 1055-1061.
Carvalho S, McGrew W, and Dominguez-Rodrigo M. 2012. The origins of the Oldowan. Stone Tools and Fossil Bones, p 201-221.
65


Diogo R, Richmond BG, Wood B. 2012. Evolution and homologies of primate and modem human hand and forearm muscles, with notes on thumb movements mid tool use. J Hum Evol 63(l):64-78.
Dirks PH, Roberts EM, Hilbert-Wolf H, Kramers JD, Hawks J, Dosseto A, Duval M, Elliott M, Evans M, Grun R et al. 2017. The age of Homo naledi and associated sediments in the Rising Star Cave, South Africa, elife 6:e24231.
Distefano C, Zhu, M., Mindrila, D. 2009. Understanding and Using Factor Scores:
Considerations for the Applied Researcher. Practical Assessment, Research, and Evaluation 14(20): 1-11.
Dryden IL, Mardia, K.V. 1998. Statistical Shape Analysis. New York: John Wiley & Sons.
Eliot DJ, Jungers WL. 2000. Fifth metatarsal morphology does not predict presence or absence of fibularis tertius muscle in hominids. J Hum Evol 38(2):333-342.
Gower JC. 1975. Generalized Procmstes analysis. Pyschometrika 40(1):33-51.
Green DJ, Gordon AD. 2008. Metacarpal proportions in Australopithecus africanus. J Hum Evol 54(5):705-719.
Gunz P, Mitteroecker, P. 2013. Semilandmarks: a method for quantifying curves mid surfaces. Hystrix 24(1):103-109.
Haile-Selassie Y, Latimer BM, Alene M, Deino AL, Gibert L, Melillo SM, Saylor BZ, Scott GR, mid Lovejoy CO. 2010. An early Australopithecus afarensis postcranium from Woranso-Mille, Ethiopia. Proc Natl Acad Sci USA 107(27):12121-12126.
Hamrick MW, Churchill, S.E., Schmitt, D., Hylander, W.L. 1998. EMG of the human flexor pollicis longus muscle: implications for the evolution of hominid tool use. J Hum Evol 34:123-136.
Johanson DC, Taieb, M., Coppens, Y., Roche, H. 1980. New discoveries of Pliocene
hominids and artifacts in Hadar-international Afar Research Expedition to Ethiopia (4th mid 5th field seasons, 1975-1977). J Hum Evol 9:583-585.
Kivell T, Kibii, JM., Churchill, SE., Schmid, P. Berger, LR. 2011 .Australopithecus sediba Hand Demonstrate Mosaic Evolution of Locomotor mid Manipulative Abilities.
Science 333(6048): 1411-1417.
Kivell TL. 2015. Evidence in hand: recent discoveries and the early evolution of human manual manipulation. Philos Trans R Soc Lond B Biol Sci 370(1682).
Kivell TL, Deane AS, Tocheri MW, Orr CM, Schmid P, Hawks J, Berger LR, Churchill SE. 2015. The hand of Homo naledi. Nat Commun 6:8431.
66


Kivell TL. 2016. The primate wrist. In: Kivell TL, Lemelin, P., Richmond, B.G., Schmitt, D., editor. The Evolution of the Primate Hand. New York: Springer, p 17-54.
Knigge RP, Tocheri MW, Orr CM, and McNulty KP. 2015. Three-dimensional geometric morphometric analysis of talar morphology in extant gorilla taxa from highland and lowland habitats. Anat Rec 298(l):277-290.
Larson SG, Jungers WL, Tocheri MW, Orr CM, Morwood MJ, Sutikna T, Awe RD,
Djubiantono T. 2009. Descriptions of the upper limb skeleton of Homo floresiensis. J Hum Evol 57(5):555-570.
Leakey LS, Tobias PV, Napier JR. 1964. A New Species of the Genus Homo from Olduvai Gorge. Nature 202(4927):7-9.
Lemelin P, and Diogo, R. 2016. Anatomy, Lunction, and Evolution of the Primate Hand Musculature. In: Kivell TL, Lemelin, P., Richmond, B.G., Schmitt, D. editor. The Evolution of the Primate Hand. New York: Springer, p 155-193.
Lovejoy CO, Simpson SW, White TD, Asfaw B, Suwa G. 2009. Careful climbing in the Miocene: the forelimbs of Ardipithecus ramidus and humans are primitive. Science 326(5949):70e71-78.
Maki J, and Trinkaus, E. 2011. Opponens Pollicis Mechanical Effectiveness
in Neandertals and Early Modem Humans. PaleoAnthropology 2011(2011):62-71.
Marzke MW. 1971. Origin of the human hand. Am J Phys Anthropol 34(l):61-84.
Marzke MW. 1983. Joint Functions and Grips of thq Australopithecus afarensis Hand, with Special Reference to the Region of the Capitate. J Hum Evol 12:197-211.
Marzke MW, Wullstein KL, Viegas SF. 1992. Evolution of the power ("squeeze") grip and its morphological correlates in hominids. Am J Phys Anthropol 89(3):283-298.
Marzke MW. 1997. Precision grips, hand morphology, and tools. Am J Phys Anthropol 102(1):91-110.
Marzke MW, Toth N, Schick K, Reece S, Steinberg B, Hunt K, Linscheid RL, An KN. 1998. EMG study of hand muscle recmitment during hard hammer percussion manufacture of Oldowan tools. Am J Phys Anthropol 105(3):315-332.
Marzke MW, Marzke RF, Linscheid RL, Smutz P, Steinberg B, Reece S, An KN. 1999. Chimpanzee thumb muscle cross sections, moment arms and potential torques, mid comparisons with humans. Am J Phys Anthropol 110(2): 163-178.
67


Marzke MW, Shrewsbury MM, Homer KE. 2007. Middle phalanx skeletal morphology in the hand: can it predict flexor tendon size mid attachments? Am J Phys Anthropol 13 4(2): 141 -151.
Marzke MW, Tocheri MW, Steinberg B, Femiani JD, Reece SP, Linscheid RL, Orr CM, Marzke RF. 2010. Comparative 3D quantitative analyses of trapeziometacarpal joint surface curvatures among living catarrhines and fossil hominins. Am J Phys Anthropol 141(1):38-51.
Marzke MW. 2013. Tool making, hand morphology and fossil hominins. Philos Trans R Soc Fond B Biol Sci 368(1630):20120414.
Matarazzo S. 2008. Knuckle walking signal in the manual digits of Pan mid Gorilla. Am J Phys Anthropol 135(l):27-33.
McPherron SP, Alemseged, Z., Marean, C.W., Wynn, J.G., Reed, D., Geraads, D., Bobe,R., mid Bearat, H. 2010. Evidence for Stone-Tool-Assisted Consumption of Animal Tissues before 3.39 Million Years Ago at Dikika, Ethiopia. Nature 466:857-860.
Moya-Sola S, Kohler M, Alba DM, and Almecija S. 2008. Taxonomic attribution of the Olduvai hominid 7 manual remains and the functional interpretation of hand morphology in robust australopithecines. Folia Primatol (Basel) 79(4):215-250.
Napier J. 1962a. The evolution of the hand. Sci Am 207(6): 56-62.
Napier J. 1962b. The prehensile movements of the human hand. Bone Joint J 38(4):902-913.
Napier J, Tuttle, R. H. 1993. Hands: Princeton University Press.
Niewoehner WA. 2001. Behavioral inferences from the Skhul/Qafzeh early
modem human hand remains. PNAS 98(6):2979-2984.
Niewoehner WA, Bergstrom A, Eichele D, Zuroff M, Clark JT. 2003. Digital analysis: Manual dexterity in Neanderthals. Nature 422(6930):395.
Niewoehner WA. 2006. Neanderthal hands in their proper perspective. In: Harvati KaH, T, editor. Neanderthals Revisited: and Perspectives New Approaches: Springer.
Orr CM, Tocheri MW, Burnett SE, Awe RD, Saptomo EW, Sutikna T, Jatmiko, Wasisto S, Morwood MJ, Jungers WL. 2013. New wrist bones of Homo floresiensis from Liang Bua (Flores, Indonesia). J Hum Evol 64(2):109-129.
Orr CM. 2017. Locomotor hand postures, carpal kinematics during wrist extension, mid associated morphology in anthropoid primates. Anat Rec 300(2):382-401.
Panger MA, Brooks AS, Richmond BG, WoodB. 2002. Older than the Oldowan? Rethinking the emergence of hominin tool use. Evol Anthropol ll(6):235-245.
68


Parr WC, Chatterjee HJ, Soligo C. 2011. Inter- and intra-specific scaling of articular surface areas in the hominoid talus. J Anat 218(4):386-401.
Patel BA. 2010. Functional morphology of cercopithecoid primate metacarpals. J Hum Evol 58(4):320-337.
Patel BA. 2016. Morphological Diversity in the Digital Rays of Primate Hands. In: Kivell TL, Lemelin, P., Richmond, B.G., Schmitt, D., editor. The Evolution of the Primate Hand: Springer, p 55-100.
Plavcan JM. 2002. Sexual dimorphism in primate evolution. Am J Phys Anthropol 44:25-53.
R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.
Rabey KN, Green DJ, Taylor AB, Begun DR, Richmond BG, McFarlin SC. 2015.
Locomotor activity influences muscle architecture and bone growth but not muscle attachment site morphology. J Hum Evol 78:91-102.
Richmond BG, Jungers WL. 2008. Orrorin tugenensis femoral morphology and the evolution of hominin bipedalism. Science 319(5870):1662-1665.
Richmond BG, Roach, N.T., Ostrofsky, K.R. 2016. Evolution of the early hominin hand. In: Kivell TL, Lemelin, P., Richmond, B.G., Schmitt, D., editor. The Evolution of the Primate Hand: Springer, p 515-543.
Ricklan DE. 1987. Functional anatomy of the hand of Australopithecus africanus. J Hum Evol 16:643-664.
Rogers ME, Abemethy K, Bermejo M, Cipolletta C, Doran D, McFarland K, Nishihara T, Remis M, Tut in CE. 2004. Western gorilla diet: a synthesis from six sites. Am J Primatol 64(2):173-192.
Rohlf JF, Slice, D. 1990. Extensions of the Procrustes method for the optimal superimposition of landmarks. Syst Biol 39(l):40-59.
Rohan C. 2009. Integration mid Evolvability in Primate Hands and Feet. Evolutionary Biology 36(1):100-117.
Rohan C, Lieberman DE, Zermeno JP. 2011. Hand biomechanics during simulated stone tool use. J Hum Evol 61(1):26-41.
Rohan C., Gordon, A.D. 2013. Metacarpal proportions in Australopithecus africanus. J Hum Evol 54(5):705-719.
69


Rothman JM, Plumptre AJ, Dierenfeld ES, Pell AN. 2007. Nutritional composition of the diet of the gorilla (Gorilla beringei): a comparison between two montane habitats. J Trop Ecol 23(06):673-682.
Sanz CM, Morgan, D.B. 2013. The Social Context of Chimpanzee Tool Use. In: Sanz CM, Call, J., Boesch, C., editor. Tool use in animals: cognition and ecology: Cambridge University Press, p 161-175.
Schlecht SH. 2012. Understanding entheses: bridging the gap between clinical and anthropological perspectives. Anat Rec 295(8):1239-1251.
Schroeder L, Scott JE, Garvin HM, Laird MF, Dembo M, Radovcic D, Berger LR, de Ruiter DJ, Ackermann RR. 2017. Skull diversity in the Homo lineage and the relative position of Homo naledi. J Hum Evol 104:124-135.
Scott JE. 2015. The phylogenetic utility of mentum osseum morphology in Pleistocene Homo. Am J Phys Anthropol 156:283.
Semaw S, Rogers MJ, Quade J, Renne PR, Butler RF, Dominguez-Rodrigo M, Stout D, Hart WS, Pickering T, Simpson SW. 2003. 2.6-Million-year-old stone tools mid associated bones from OGS-6 mid OGS-7, Gona, Afar, Ethiopia. J Hum Evol 45(2): 169-177.
Senut B, Pickford, M., Gommery, D., Mein, P., Cheboi, K., Coppens, Y. 2001. First hominid from the Miocene (Lukeino Formation, Kenya). Earth Planet Sc 332:137-144.
Siegel M, Pemotto, B. 1975. Hand Use and Metacarpal Robusticity in Catarrhini. Primates 16:371-377.
Susman RL, Creel N. 1979. Functional and morphological affinities of the subadult hand (O.H. 7) from Olduvai Gorge. Am J Phys Anthropol 51(3):311-332.
Susman RL. 1988. Hand of Par anthr opus robustus from Member 1, Swartkrans: Fossil Evidence for Tool Behavior. Science 240(4853):781-784.
Susman RL. 1994. Fossil Evidence for Early Hominid Tool Use. Science 265(5178): 1570-1573.
Susman RL. 1998. Hand function and tool behavior in early hominids. J Hum Evol 35(1):23-46.
Susman RL, Nyati L, Jassal MS. 1999. Observations on the pollical palmar interosseous muscle (of Henle). Anat Rec 254(2):159-165.
Sutikna T, Tocheri MW, Morwood MJ, Saptomo EW, Jatmiko, Awe RD, Wasisto S,
Westaway KE, Aubert M, Li B et al. 2016. Revised stratigraphy and chronology for Homo floresiensis at Liang Bua in Indonesia. Nature 532(7599):366-369.
70


Tocheri MW, Orr CM, Larson SG, Sutikna T, Jatmiko, Saptomo EW, Due RA, Djubiantono T, Morwood MJ, Jungers WL. 2007. The primitive wrist of Homo floresiensis mid its implications for hominin evolution. Science 317(5845): 1743-1745.
Tocheri MW, Orr CM, Jacofsky MC, Marzke MW. 2008. The evolutionary history of the hominin hand since the last common ancestor of Pan and Homo. J Anat 212(4):544-562.
Trinkaus E. 1983. The Shanidar Neanderthals. New York: Academic Press.
Tuttle RH. 1967. Knuckle-walking mid the Evolution of Hominoid Elands. Am J Phys Anthropol 26(2): 171-296.
Tuttle RH. 1969a. Knuckle-walking and the Problem of Human Origins. Science 166(3908):953-961.
Tuttle RH. 1969b. Quantitative mid functional studies on the hands of the Anthropoidea. L The Hominoidea. J Morphol 128(3):309-363.
Van Horn R. 1972. Structural adaptations to climbing in the gibbon hand. Am Anthropol 74(3):326-334.
Vleck, E. 1973. Postcranial Skeleton of aNeandertal Child from Kiik-Koba, U.S.S.R. J Hum Evol 2:537-544.
Vleck E. 1975. Morphology of the first metacarpal of neandertal individuals from the
Crimea. B Mem Soc Anhro Par 2(3):257-276.
Wallace IJ, Winchester JM, Su A, Boyer DM, Konow N. 2017. Physical activity alters limb bone structure but not entheseal morphology. J Hum Evol 107:14-18.
Ward CV, Kimbel, W. H., Harmon, E.H., Johanson, D.C. 2012. New postcranial fossils of Australopithecus afarensis from Hadar, Ethiopia (1990e2007). J Hum Evol 63:1-51.
Ward CV, Tocheri MW, Plavcan JM, Brown FH, Manthi FK. 2014. Early Pleistocene third metacarpal from Kenya mid the evolution of modern human-like hand morphology. Proc Natl Acad SciU S A 111(1): 121-124.
White TD, Suwa G, Asfaw B. 1994. Australopithecus ramidus, a new species of early hominid from Aramis, Ethiopia. Nature 371(6495):306-312.
Wiley DF, Amenta, N, Alcantara D.A., Ghosh D., Kil Y. J., Dels on E., Hare ourt- Smith W., Rohlf F.J., St. John K., Hamann B. 2005. Evolutionary morphing. Proc IEEE Visual.
Wiley DF. 2007. Landmark Editor. University of California, Davis: Institute for Data Analysis mid Visualization.
71


Williams EM, Gordon AD, Richmond BG. 2012. Hand pressure distribution during Oldowan stone tool production. J Hum Evol 62(4): 520-532.
Williams-Hatala EM, Hatala KG, Hiles S, Rabey KN. 2016. Morphology of muscle
attachment sites in the modem human hand does not reflect muscle architecture. Sci Rep 6:28353.
Zumwalt A. 2006. The effect of endurance exercise on the morphology of muscle attachment sites. J Exp Biol 209(Pt 3):444-454.
72


Full Text

PAGE 1

POLLICAL METACARPAL MORPHOLOGY IN PRIMITIVE HOMO : EVIDENCE FROM HOMO NALEDI BY LUCYNA A. BOWLAND B.A., University of West Florida, 2013 A thesis submitted to the Faculty of the Graduate S chool of the University of Colorado in partial fulfillment of the requirements for the degree of Master of A rts Department of A nthropology 2018

PAGE 2

ii This thesis for the Master of Arts degree by Lucyna A. Bowland has been approved for the Anthropology Program by Charles Musiba, C hair Caley M. Orr Jill E. Scott Anna Warrener Date: May 12, 2018

PAGE 3

iii Bowland, Lucyna A. (M.A., Anthropology Program) Pollical Metacarpal Morphology I n Primitive Homo : Evidence F rom Homo Naledi Thesis directed by Associate Professor Charles Musiba ABSTRACT Homo naledi provides important insights into the possible anatomy of primitive members of the genus Homo T he pollical metacarpal displays a suite of morphological characteristics previously unknown within hominoids, including a median longitudinal crest, a narrow proximal base, and broad flaring muscle flanges on the distal end. To better understand the unique morphology of the H. naledi pollical metacarpal, this study employs a 3D geometric morphometric analysis of the pollical metacarpal shaft morphology in a comparative sample of adult Old World monkeys, non human apes modern humans, and fossil hominins (n= 337). A 20x20 semilandmark grid was placed across the palmar diaphyseal surface of 3D surface renderings and the landmark data were subjected to a principal components analysis of Procrustes shape variables A multiple regression analysis of log centroid s ize on Procrustes shape variables was performed to test for the effects of size on shape. The results indicate that H. naledi separates from other hominins based on its narrow proximal base and gracile shaft surmounted by flaring muscle flanges on the distal end. The gracile shaft is most like cercopithecines, Pan, Pongo, and the australopiths, suggesting this is the primitive c ondition for hominins In contrast, Neanderthals are characterized by a wide proximal base, a pinched midshaft, and broad distal flanges, while modern human metacarpals have a straight shaft, less robust muscle flanges, and a robust proximal base. The resu lts of this study suggest that pollical metacarpal morphological development within

PAGE 4

iv primitive Homo was characterized by an intensified recruitment of intrinsic muscles on a n otherwise gracile shaft, adding to the current understanding of the early manipula tive behaviors and tool use within the genus Homo. The form and content of this abstract are approved. I recommend its publication. Approved: Charles Musiba

PAGE 5

v TABLE OF CONTENTS I. POLLICAL MET ACARPAL MORPHOLOGY I N PRIMATES .1 Introduction and Review of the Literature ... .1 Relationship Between Entheseal Morphology, Behavior, and Load Bearing Ability ..3 Manual Grips within Hominoid s .....6 Morphological Variation within the Hands of Fossil and Extant Primates Morphology of Extant Primates ...................................................................9 Morphology of Fossil Hominins ................................................................15 Theories for the Evolution of the Modern Human Hand .......................................28 The Rise of Habitual Stone Tool Use ........................................................29 The Hand is a Pleiotropic Result of Selection for the Foot .......................30 Conclusion ................................................................................................................... ......32 I I MATERIALS AND METHOD S .......... ......... .................... .................... .................... ...34 Materials ....................................................................................................................34 Methods .............................................. ........................................................................38 I II RESULTS .... ...... ..........................................................................................................41 Full Sample ........................................................................................................................41 PC1 Shape and G roupings .....................................................................................41 PC2 Shape and G roupings .....................................................................................42 PC3 Shape and G roupings .....................................................................................43 Regional V ariation within Modern H umans .......... ............................................................44 Reduced Sample .................................................................................................................47

PAGE 6

vi Hominoid Sample .............................................. ....................................................49 PC1 Shape and Groupings .........................................................................50 PC2 Shape and Groupings .........................................................................53 P C3 Shape and Groupings .........................................................................55 Shape and Siz e ..................................................................................................................57 I V DISCUSSION AND FUTURE RESEARCH ........ .................... .................... ..............59 PC1 Breadth of the Metacarpal Shaft ................................................................................59 PC2 Breadth of the Proximal End Relative to the Overa ll Shaft .......................................60 PC3 Breadth of the Proximal End Relative to the Distal Flanges .....................................61 Implications for the Morphology of Primitive Homo ...................................................... .62 Future Research .................................................................................................................63 Variation within Gorilla spp .................................................................................63 Sexual Dimorphism ...............................................................................................64 Conclusion .........................................................................................................................65 REFERENCES.... .................... .................... .................... .................... .................... ..........66

PAGE 7

1 CHAPTER I POLLICAL METACARPAL MORPHOLOGY IN PRIMAT ES Introduction and Review of the Literature A fully opposable thumb capable of powerful precision grasping is seen as a hallmark of modern humans. Over the course of our evolutionary history, the hand was freed from its locomotor constraints, allowing for its use primarily as a tool for manipulator purposes (Ki v ell et al., 2011). Humans are unique amon g the extant great apes in the extensive use of the thumb, which is reflected in their broad trapeziometacarpal joints ( TCMJ ) and robu st metacarpal shafts, both of which are adaptations for enhanced pre cision grasping within modern humans (Marzke, 1997). Humans are also unique in respect to the musculature of the hands. In contrast to all other extant primates ( except for gibbon s ) the flexor pollicis longus ( FPL ) in humans is a distinct muscle from the flexor digitorum profundus (FDP) with a tendon that is exclusive to the pollex (Lemelin and Diogo, 2016). This feature is likely related to increased force at the tip of the first metacarpal (Mc1), which in turn a ids enhanced precision grasping (Hamrick et al., 1998; Marzke et al., 1998, Lemelin and Diogo, 2016). In contrast, extant non human primates possess short thumbs relative to finger length, with long phalanges that preclude enhanced precision grasping (Marz ke, 1997; Almecija et al., 2015). Extant African apes occasionally use their thumbs for manipulator behaviors, but their hand morphology is highly specialized for their locomotor patterns in which the pollical metacarpal is rarely engaged (Tuttle, 1969a ). To accurately understand the significance of the manual morphology of extant primates, one must look at the evidence for early hand morphology within fossil hominins Within the fossil record, there is evidence for hominin hand remains by ca. 6 Ma with

PAGE 8

2 Orrorin tugenensis (Almecija et al., 2010). Or. tugenensis remains display several morphologies related to enhanced grasping, including a broad apical tuber osity and an insertion for the FPL on the pollical distal phalanx (PDP) (Marzke 1997; Almecija et al ., 2010). Ardipithecus ramidus at circa 4.5 Ma, had short metacarpals, long fingers, and a mobile wrist. This suite of morphologies contrasts with extant African apes and is more like Old World monkeys and Miocene hominids such as Proconsul ( Lovejoy et al., 2009; Kivell, 2015) Australopithecus metacarpals are characterized by an overall gracility, with narrow metacarpal shafts and bases, and a greatly diminished crest for the opponens pollicis. Homo n eanderthal ensis metacarpals show a marked hypertrophy of the intrinsic musculature, and their overa ll morphology appears to be auta pomorphic for the taxon. The e arly anatomically modern human (EAMH) Qafzeh specimens lack the hypertrophy of the intrinsic musculature that is present in Neanderthals and their pollical metacarpals are overall less robust than Neanderthals and more like modern humans. However, n oticeably absent from the fossil record is detailed evidence for the manual morphology of primitive members of the genus Homo The fossil remains for primitive Homo are sparse, including several specimens from Homo habilis (OH7) ( Leakey et al., 1964 ), partial hand remains from Homo floresiensis (Brown et al., 2004; Tocheri et al., 2007), and remains from the newly discovered species Homo naledi (Berger et al., 2015; Kive l l et al., 2015). Neither the remains from Olduvai Gorge (OH7), nor those from H. floresiensis include a complete Mc1 meaning the evidence for the morphology of primitive Homo fir s t metacarpals has been, unt il recently, poorly understood However, included within the vast fossil finds from the Dinaledi Chamber in South Africa were s even pollical metacarpal s from H. naledi ( Berger et al., 2015). All the specimens exhibited the same

PAGE 9

3 unusual morphological characteristics: a narrow proximal base surmounted by broad muscle flanges on the distal ends of the shaft (Kivel l et al., 2015) which is unl ike anything seen previously. It is unclear whether this unique morphology possibly represents the character state for primitive members of the genus Homo or whether H. naledi represents an auta po morphy in the hominin clade. This paper attempts to answer the question of whether H. naledi represents an autopa morphic taxon or a transitional form for primitive Homo pollical metacarpal morphology. To answer this question, and to better understand the morphology of primitive Homo pollical metacarpals, this study employed a 3D geometric morphometric analysis of pollical metacarpal shaft morphology in a comparative sample of Old World monkeys, non human apes, modern humans, and fossil hominins (n=337). Relationsh ip b etween Entheseal M orphology Behavior, and Load Bearing A bility The intrinsic muscles of the hand have often been a focus for paleoanthropological studies because of their assumed relationship to hand use and function (Ricklan, 1987; Eliot and Jungers, 2000; Zumwalt, 2006; Marzke et al., 2007) I ncreased muscle entheseal development is often associated with increased recruitment of the intrinsic musculature, because of the assumptio n that the increase in muscle recruitment was correlated with the advent of stone tool usage (Susman, 1994) An enthesis is the site where muscle meets bone, and is the result of the stress concentration where hard (bone) and soft (muscle) tissues meet. Muscle entheses help to transmit forces over osteotendinous surfaces, as well as anchor tendons to bones. The proposed relationship between muscle enthesis development and behavior is based on the idea that repetitive activity stimulates periosteal remode ling and modeling activity by

PAGE 10

4 increasing bone cell proliferation through expansion of the capillaries (Schlecht, 2012). While the perceived relationship between the two components appears to be somewhat straightforward it is important to note that there i s no proven relationship between muscle entheseal morphology and the behavior of the organism in question, and previous studies have found no relationship between the two (Rabey et al., 2015; Williams Hatala et al., 2016). However, it is not currently unde rstood how other factors, such as age, sex, body mass, and genetics influence enthesis morphology, and these variables are rarely explored in entheseal morphology studies (Schlecht, 2012) In a recent study, Williams Hatala et al. (2016), examined the re lationship between muscle enthesis development and function in a sample of cadaveric modern human pollical metacarpals. The study looked at whether any correlations existed between physiological cross sectional area of the opponens pollicis muscle and enth esis development The study found no statistically significant ( p >0.05) relationship between the muscle architecture and the entheseal surface morphology The authors questioned the efficacy of using muscle enthesis architecture as a proxy for function in extinct organisms, arguing the musculoskeletal structure is too complex for simple causal relationships (Williams Hatala et al., 2016). However, their s ample consisted entirely of cadaveric specimen s, whose muscles no doubt underwent some degree of atrophy prior to the time of the study Additionally, previous research (Zumwalt, 2006) has noted the problematic nature of using fully mature individuals when attempting to assess the efficacy of activity levels and muscle enthesis morphology. It has long been hypothesized (Marzke et al., 1992; Susman, 1994; Marzke et al., 1998) that a more robust thumb is synonymous with the ability to bear increased loading

PAGE 11

5 experienced during tool making. Human pollical metacarpals have a broad shaft and an expanded proximal b ase, which increases surface area and allows for more even distribution of forces incurred because of powerful precision grasping (Marzke, 199 7). To test this hypothesis, Williams et al. (2012) monitored the normal force (N) and pressure (kPa) of six exper ienced tool makers while they simulated Oldowan tool making techniques. The participants held the stone tools in a three jaw chuck grip (Fig. 1 ) with the hammer stone stabilized by distal phalanges of the first three digits. While they conceded the thumb p lays an important role in stabilizing the object during production, they note it is, over built for tool making stress levels. The results of their study found forces incurred during Oldowan tool production concentrated more on the second and third digit as opposed to the thumb. They concluded the results do not support the hypothesis that the primate thumb evolved as a response to increased loads experienced during Oldowan tool making. Figure 1 Extended 3 ja w chuck grip in a modern human. From Marzke et al., 199 7.

PAGE 12

6 Manual G rips with in H ominoids The ability to engage in pad to pad precision grasping is paramount for enhanced manual dexterity within modern humans and the unique morphology of the human thumb is tantamount to the production of manual grips ( Napier, 1962 a ; Napier, 1962b; Marzke, 1997; Almecija et al., 2010) Precision grip is defined as any grip that utilizes the thumb and at least one other finger, either with or without using the palm as a prop ( Napier, 1 962b ; Marzke, 1997) (Fig. 2 ) Pad to pad precision grip is defined as the proximal pulp of the thumb being in opposition to one or more fingers (Almecija et al., 2010). Human hands are uniquely adapted for pad to pad precision grasping, in c ontrast to other extant apes, whose long fingers, short thumbs, and reduced intrinsic musculature preclude them from engaging in effective manual grips ( Napier and Tuttle 1993; Almecija et al., 2010). Evidence for enhanced precision grasping capabilities has been found in numerous hominin species, beginning with the Miocene hominin Or tugenensis (Senut et al., 2001; Almecija et al., 2010). Morphological features associated with pad to pad precision grasping can be found on the PDP of modern humans and select fossil hominin species, and include an insertion point for FPL the presence of an ungual fossa, and asymmetrical ungual spines (Marzke, 1997). All features of the PDP associated with pad to pad precision grasping are found wit hin modern humans, but are absent entirely within extant great apes and are variably present within the hominin fossil record. Almecija et al. (2010) compared the morphology of the PDP in extant primates and fossil hominin specimens to explore the morpho logical affinities between these taxa. Or tugenensis the oldest known specimen for which there is evidence of a PDP (Senut et al., 2001), has both an insertion gable for the FPL and a broad apical tuberosity, like modern

PAGE 13

7 human PDP 's ( Almecija et al., 201 0). However, the later specim en of OH7, attributed to H. habilis lacks an insertion gable for the FPL and ungual spines, though it does possess a large palmar fossa that extends to the apical tube rosity (Almecija et al., 2010). Figure 2 Modern human thumb and index finger (right hand) during pad to pad p recision grasping in ulnar view. From Almecija et al., 2010. Robust muscle attachment sites for the opponens pollicis and FPL on the thumb have been used to infer enhanced grasping capabilities withi n early hominins (Marzke et al., 1998; Niewoeh ner et al., 2003 ; Maki and Trinkaus, 2011; Marzke, 2013). The opponens pollicis muscle is located on the distodorsoradial portion of the first metacarpal, and the presence of a particularly robust muscle attach ment site has been used to infer hypertrophy of the opponens pollicis, as well as provide evidence for an increased moment a rm of the muscle (Trinkaus, 1983 ; Maki and Trinkaus, 2011). Neanderthals are known to have a particularly robust crest for the opponens pollicis muscle attachment (Niewoehner, 2006). Maki and Trinkaus (2011) examined the relationship between muscle enthesis development and function in a study of Neanderthal, modern human, and Middle and Mid Upper Paleolithic pollical metacarpals.

PAGE 14

8 Their study found size of the crest, both absolute and relative to overall body size, to be greater in Neanderthals and Middle Paleolithic hominins than in later hominins and modern humans. They hypothesized the decrease in the size of the crest in later h ominins is possibly tied to advances in lithic technology as the Paleolithic progressed, requiring less muscle force to be employed than in previous lithic technologies Morphological V ariation within the H ands of Fossil and Extant P rimates Primate hands are comprised of several sets of bones. The bones of the wrist are known as the carpal bones, the number of which varies among different primate taxa. Humans have eight carpal bones, while most other primates possess nine. This is due to the fusion of the scaphoid and os centrale bone into one element in humans, while the two are separate bones in many other primate species. The palmar portions of the fingers are known as the metacarpals, while the external portions of the digits are comprised of proximal, intermediate, and distal phalanges (except for the thumb, which contains only the proximal and distal phalanges) (Kivell, 2016) There is a wide range of variation within the pollical me tacarpals of primates. Figure 3 shows morphological variation for a selection of hominoid taxa Some taxa, such as Pan and Hylobates possess gracile metacarpal shafts with a narrow proximal base. In humans, the metacarpal shaft is broad, with a relatively expanded articular facet on the proximal end. The following secti on will review some of the key morphological features of extant primates, as well as discuss the functional importance corresponding to each

PAGE 15

9 Figure 3 Palmar view of three dimensional surface renderings of the pollical metacarpal in a selection of hominoid taxa *Variously a ttributed to Australopithecus and Homo. Morphology of Extant P rimates Homo sapiens. The bony morphology and distinct musculature of the human thumb are indicative of its integral role in human manipulative behaviors (Kivell, 2015). The human hand is distinct from that of other living primates in terms of the long thumb relative to other d igit lengths and the broad apical tufts on the distal phalanges, both of which aid precision grasping capabilities (Marzke, 1997; Almecija et al., 2015). Compared to other extant primates, the muscles of the human thumb have a much larger moment arm, which is linked to enhanced mechanical capabilities, such as offering more leverage when attempting to stabilize the hand against an object (Kivell, 2015). H. sapiens hands are also characterized by much lesser degrees of phalangeal curvature than is present in other living primates. F igure 4 describes the morphological adaptations associated with enhanced precision grasping within the human hand, as well as the muscle s recrui ted during precision and power squeeze grips. Hylobates lar P an troglodytes Au. afarensis Homo naledi Homo sapiens H. neanderthalensis SK 84

PAGE 16

10 Figure 4 Bony and soft tissue considered to be associated with precision and power squeeze grips used during stone tool use and production (a). Human precision grip (b) and a power squeeze grip (c). From Kivell, 2015. Humans possess three muscles (an independent FPL, the extensor pollicis brevis, and the first volar interosseous of Henle) which are thought to be integral to their ability to form precision grips (Susman et al., 1999; Williams et al., 2012). These three muscles are absent in nearly all other extant primates, except for the Hylobatids, who also possess a FPL that is separate from the flexor digitorum profundus (FDP) (Lemelin and Diogo, 2016). The FPL inserts on the volar side of the distal pollical phalanx and has been widely associat ed with stone tool usage within the hominin lineage (Susman, 1994; Marz ke, 1997; Diogo et al., 2012; Williams et al., 2012 ). Apart from the hylobatids (in which the FPL and the FDP are joined via a connective tissue) and Homo sapiens this muscle is joined with the FDP muscle in all other primates (Lemelin and Diogo, 2016). Humans are therefore unique among extant primates in having an FPL that is free from the other digits, which is responsible for the increased flexion seen in the thumb of modern humans (Hamrick et al ., 1998; Williams et al., 2012). However, the discovery of an insertion site for the FPL on the distal pollical phalanx

PAGE 17

11 of Orrorin tugenensis (ca. 6 Ma), predating the use of stone tools by at least 2.5 Ma (McPherron et al., 2010), throws doubt on the long held assumption that the presence of this muscle insertion site is synonymous with tool usage in hominins (Susman, 1994, Almecija et al., 2010). The extensor pollicis brevis (EPB) muscle inserts on the first phalanx and is present in roughly 90% of moder n humans (Williams et al., 2012). The muscle aids in thumb extension by assisting the adductor pollicis brevis (APB) and the extensor p ollicis longus (EPL) and is integral to the ability to stabilize the thumb against an object while extended. The first vo lar interosseous of Henle is found in roughly 90% of modern humans (Williams et al., 2012), and l ess than 50% of other primate taxa, though it is present in both Gorilla and P. troglodytes (Diogo et al., 2012). Its function in unclear, though it is believe d to either aid in adducting the thumb at the metacarpophalangeal joint, or possibly to add sensory information about the position of the thumb (Susman et al., 1999; Williams et al., 2012). The opponens pollicis muscle inserts on the lateral side of the pollical metacarpal (Diogo et al., 2012). In humans, the insertion site presents as a distinct bony flange running along the medial portion of the bone. The muscle is responsible for abduction and flexion at the first carpometacarpal joint, allowing for th e opposition of the thumb. In modern humans, the opponens pollicis occupi es a much larger cross section and provides more potential torque than what is seen in extant primates (Marzke, 1997; Marzke et al., 1999). Taken together, these two features are thou ght to limit the fatigue experienced by the thumb during increased loading, which has been shown to occur in experimental studies of Oldowan tool manufacture (Hamrick et al., 1998, Kivell, 2015). As studies of fossil remains must rely solely on the bony mo rphology to form functional interpretations, the presence of the bony

PAGE 18

12 flange on the distodorsoradial portion of the pollical metacarpal has been used to infer hypertrophy of the muscle, as well as a greater moment arm for movement (Maki and Trinkaus, 2011) Non human primates This section will focus on the hand morphology of extant non human primates. The locomotor patterns of the species in question vary greatly, from committed terrestrial knuckle walkers to fully arboreal, suspensory primates. Except for Nasalis larvatus who is a fully arboreal primate, all cercopithecine Old World monkeys included within this study ( Erythrocebus patas, Macaca fasicularis Papio sp ) are terrestrial quadrupeds (Patel, 2010). Gorilla sp. and Pan sp. are terrestrial knuckle walkers, Pongo sp., divide their time between an arboreal and terrestrial lifestyle, and Hylobates lar i s a committed arboreal primate whose morphology is highly specialized for suspensory locomotion ( Tuttle, 1969 b ; Siegel and Perno tto, 1975; Matarazzo, 2008). Extant non human apes can be divided into two general categories: African apes, Pan and Gorilla and the Asian apes, Pongo and Hylobates The main morphological differences between the Asian apes and African apes are a result of their different locomotor patterns. The African apes ( Gorilla and Pan ) possess secondary adaptations for terrestrial locomotion, while Hylobates and Pongo are both adapted for arboreal locomotion (Tuttle, 1969 b ). Figure 5 shows differences in the ray mo rphology and manual proportions among extant great apes.

PAGE 19

13 Figure 5 Articulated metacarpals and phalanges of the left hand, scaled to 1cm. From Patel and Maiolino, 2016. African apes Gorilla is divided into two species, the mountain Gorilla ( Gorilla beringei ), and the lowland Gorilla ( Gorilla gorilla ). Gorilla beringei are predominantly terrestrial in their feeding, nesting, and locomotor patterns. Gorilla gorilla are generally more arboreal, though this could simply be a consequence of their more fo rested habitats (Tuttle, 1969 a ). The hands of gorillas are adapted to bear the intense forces incurred throughout the hand during knuckle walking (Kivell, 2016). The carpal bones are tightly compressed to help stabilize the wrist and prevent radioulnar dev iation or rotation during knuckle walking (Tuttle, 1969 b ; Kivell, 2016). In addition to an immobile wrist, G orilla metacarpals are adapted in response to knuckle walking locomotion. The breadth of the metacarpal head is large relative to the length of the shaft (Susman, 1994), and there is a transverse ridge at the base of the dorsal articular surface (Tuttle 1969 b ; Patel and Maiolino, 2016) There is debate

PAGE 20

14 as to the functional advantages of this dorsal ridge, with some authors suggesting it acts to help s upport and stabilize the metacarpophalangeal joints during locomotion (Tuttle, 1969b ), though it is also possible it is unrelated to knuckle walking and maybe instead correlate to changes in body size throughout ontogeny (Susman and Creel 19 79; Patel and Maiolino, 2016) Almecija et al. (2015) note the morphological state exhibited by gorillas and modern humans appear s to be largely plesiomorphic and remains largely unchanged since the time of the last common ancestor (LCA) of Pan and Homo. The genus Pan consists of two species, Pan paniscus (bonobos), and Pan troglodytes (chimpanzees). Pan are generally terrestrial knuckle walkers, though they also spend considerable time engaged in climbing and suspensory locomotion in trees (Marzke, 1971) Chimpanzees move quadrupedally with their metacarpophalangeal joint hyperextended and their weight supported mainly on the dorsal aspects of the middle phalanges (Tuttle, 1969 b ; Marzke, 1971; Matarazzo, 2008). Tuttle (1969 a ) reported that during quadrupe dal locomotion, chimpanzee thumbs never contact the ground, nor are they used for support during locomotion. Relative to Gorilla the metacarpals of chimpanzees are long and gracile (Susman and Creel 1979 ) Asian apes Pongo (orangutans) are arboreal and are generally suspensory (Tuttle, 1969 b ). They possess the most curved phalanges of the extant grea t apes (Richmond et al., 2016) Compared to the African apes, the metacarpals of Pongo are elongated and have less pronounced muscle and ligament impressions (Susman and Creel 1979). Their metacarpal heads are relatively much larger than their shafts, in contrast to the morphology seen in gorillas and Homo sapiens (Susman, 1994; Marzke and Marzke, 2000).

PAGE 21

15 Hylobates lar (gibbons) are generally categ orized as a brachiato a r, though they are capable of palmigrade and pronograde quadrupedal locomotion (Tuttle; 1969 b ; Van Horn, 1972). H ylobatids most common form of terrestrial locomotion consists of walking quadrupedally with the hand formed into a fist, in which most of their weight is supported on the back of the proximal phalanges (Marzke, 1971). Hylobatid metacarpals are long and thin (Susman and Creel 1979). The thumb of Hylobates is short in relation to their other metacarpals, though they possess one of the longest thumbs relative to other primates (Van Horn, 1972; Patel and Maiolino, 2016). Hylobates are the only extant primates aside from modern humans, that possess a separate tendon belly for the FPL (Lemelin and Diogo, 2016). Morphology of Fo ssil Hominins Within the hominin lineage, morphological adaptations associated with enhanced precision grasping first emerge within Or tugenensis at roughly 6 Ma, and there is further evidence for early hominin manual morphology through to Au. sediba at roughly 1.2 Ma (Marzke, 1983; Senut et al., 2001; Almecija et al., 2010; Kivell et al., 2011). This section will explore the available evidence early hominin manual morphologies and their possible functional implications. Ardipithecus T he manual remains of Ardipithecus provide important insights into the locomotor patterns of early hominins. The genus Ardipithecus is comprised of two species, the older Ar. kadabba (5.7 5.2 M a) (Haile Selassie, 2001), and the younger Ar ramidus (ca. 4.4 M a) ( White et al., 1994). The remains from Ar ramidus are much more complete, and so this section will focus on Ar. ramidus. The manual remains consist of nearly two complete hands, from which are missing only the pisiform and some terminal phalanges (Lovejoy et al., 2009).

PAGE 22

16 The hands of Ar. ramidus display several important morphological features (Fig 6 ) The medial metacarpals (Mc2 5) are all short relative to extant great apes. The wrist of Ar. ramidus appears to be highly mobile, which contrasts with extant great apes (Lovejoy et al., 2009) and more like extant monkeys ( Orr, 2017 ). An immobile wrist helps to dissipate forces during loading, such as during suspensory locomotion (Lovejoy et al., 2009). The thumb of Ar. ramidus is large and robust (Kivell, 2015 ) with a highly developed FPL gable on the PDP (Lovejoy et al., 2009). In contrast, the FPL of African apes is variably highly reduced or altogether absent (Lemelin and Diogo, 2016). The articular surfaces of the metacarpophalangeal, carpometacarpal, and interphalangeal joints are larger and more robust compared to African ap es, possibly indicative of greater load bearing ability throughout these joints or of greater thenar mobility (Lovejoy et al., 2009). Figure 6 Digitally rendered composite hand of ARA VP 6/500 ( Ardipithecus ramidus) in palmar view. From Lovejoy et al. 2009. Lovejoy et al. (2009:74) argue the postcranial morphology of Ar. ramidus is indicative of an above branch quadruped which they hypothesize is also the likely

PAGE 23

17 locomotor form of the last common ancestor of humans and chimps (LCA). At present, it is not clear whether Ar ramidus represents a hold out from the Miocene, representing the li kely locomotor patterns of the LCA or whether the species independently evolved for its highly specialized form of locomotion. While both scenarios are likely, Ar. ramidus is roughly 3 million years removed from the hypothesized time of the LCA (Tocheri et al., 2008), and so any interpretations arguing its morphology closely resembles a basal hominin must bear that fact in mind. Australopithecus sp. and Paranthr opus robustus By roughly 3 Ma, remains from Australopithecus provide evidence for the advent of morphological adaptations in the hand associated with increased manipulative abilities ( Marzke, 1983; Marzke, 1997; Alba et al., 2003). Marzke (1997) noted two important features in the PDP of Stw 294, most commonly attributed to Au. africanus These features include (1) a broad apical tuft; and (2) evidence for a n FPL muscle. The FPL muscle plays an important role in flexing the thumb and providing greater resi stance against the radial side of the index finger, allowing for a more forceful and controlled precision grip (Susman, 1998). Ricklan (1987) contends the Au. africanus specimen Stw 64 shows evidence of a styloid process on the third metacarpal, though oth ers have suggested the feature is morphologically distinct from the styloid process that characterizes the genus Homo (Ward et al., 2014). The presence of a styloid process on the third metacarpal is thought to help stabilize the metacarpal while in articu lation with the carpal bones, as well as relieve increased stresses incurred during tool making (Marzke and Marzke, 2000 ). This feature is present in Neanderthals and modern humans, but absent in all other extant great apes and early fossil hominins, inclu ding Australopithecus (Ward et al.,

PAGE 24

18 2014 [but see Ricklan (1987) for a review of its possible appearance in Au. africanus (Stw 64))]. Green and Gordon (2008) concluded that the lengths of the of Au. africanus (Stw 418 and Stw 583) metacarpals were like the more modern human like lengths seen in Au. afarensis However, while the relative length of the metacarpals overall is more like modern humans, the relative breadth of the first metacarpal is more like extant apes (Susman and Creel 1979). Susman (199 4) linked the breadth of the first metacarpal to a morphology more derived for manipulative tasks such as tool making. Ricklan (1987) noted curvature in the phalanges of Au. africanus a trait which is generally associated with arboreal locomotion and gras ping capabilities, though he disputed this was the case for Au. africanus arguing the hands were more adapted for manipulative tasks than for time spent in the trees. Further evidence of precision grasping capabilities can be found in Australopithecus af arensis The skeleton of Lucy (A.L. 288) include s a single non pollical phalanx and finds from the Afar locality of A.L. 438 include several medial phalangeal elements (Rolian and Gordon, 2013). The clear majority of associated hand remains, however, co me from the Afar locality of A.L. 333, and as such, this review will focus primarily on remains recovered from this specific site. The hand remains associated with A.L. 333 include complete and fragmentary metacarpals and phalanges (Johanson et al., 1980; Rolian and Gordon, 2013; Ward et al., 2012). There is some contention regarding the taxonomic affinities of the Au. afarensis phalanges in regards to their relative dimensions. Marzke (1983) and Alba et al. (2003) concluded the manual proportions of Au. a farensis were more modern human like, with a long thumb relative to shorter digit length. Based on these proportions, both studies

PAGE 25

19 suggested that Au. afarensis was most likely capable of a high degree of precision grasping like that seen in modern H. sapie ns The manual remains of Au. afarensis (A.L. 333) exhibit several derived features thought to be associated with precision grasping (Ward et al., 2014). One of the derived features present is a second metacarpal with an asymmetrical head, which also has a volar radial projection, as well as an articular surface that is beve led in an ulnar direction dorsally (Marzke, 1997; Panger et al., 2002). Marzke (1997) noted the combination of these features would have caused the index finger to pronate as it flexed which in turn would bring the entire palmar surface of the finger into direct contact with the surfaces of an object The joints of the second metacarpal articulate with the capitate and trapezium in an orientation away from the sagittal plane, which all ows the thumb to rotate objects in a secure 3 jaw chuck hold grip (Marzke, 1971; Marzke, 1997). Additionally, the hand exhibits a longer thumb relative to digit length, though likely not long enough to have facilitated powerful precision pad to pad graspin g (Marzke, 1997; Panger et al. 2002). Rolian and Gordon (2013) state they believe the dimensions of the Au. afarensis phalanges fall in between those of Gorilla and modern humans, which contrasts with previous results (Alba et al., 2003). Rolian and Gordo n (2013) included all known manual remains of Au. afarensis in their sample, as opposed to only selecting certain elements for study, which is likely the cause of their contradictory conclusion. If their results are correct, it would mean Au. afarensis wou ld not have been capable of precision grasping, and by extension, probably could not have manufactured stone tools. However, it is important to note that while there are many known metacarpal remains from Australopithecus only a few can reliably be attrib uted to the same individual, making a meta analysis of Australopithecus

PAGE 26

20 hand remains problematic (Kivell, 2015). Overall, the manual remains of Au. afarensis appear to have been adapted for tool usage, though their ability to manufacture those tools is sti ll doubtful. Australopithecus sediba (~1.97 Ma ), is a hominin from Malapa, South Africa (Berger et al., 2010). Malapa hominin 2, (MH2) contains an almost complete right hand, missing only distal pollical phalanges for digits 2 5 and some carpal elements. The pollical distal phalanx displays a mediolaterally expanded apical tuft, which more closely resembles later hominins. However, the shaft of the distal pollical phalanx is narrow er than what is seen in later hominins. Au. sediba possessed a well develop ed FPL muscle, as evidenced by morphology of the distal pollical phalanx, including an ungual fossa. The robust base on the proximal articular facet of the pollical metacarpal indicates the presence of well developed intrinsic pollical muscles (Kivell, et al 2011). The insertion site for the opponens pollicis m uscle is poorly expressed, indicating weak development of the muscle and a limited ability for opposition of the thumb (Kivell et al., 2011). The morphology of the thumb is overall more gracile than robust, which is the primitive condition and like earlier hominins such as Au. afarensis The gracile nature of the MH2 hand bones suggests Au. sediba most likely was not a prodigious tool maker, as studies have shown Oldowan tool making increases th e loads experienced by the thumb, and therefore, enhances the bony morphology of the muscle attachment sites (Kivell et al., 2011). Remains of Paranthropus robustus dated to 1.8 million years ago from the site of Swartkrans, in South Africa, demonstrate numerous morphological adaptations associated with enhanced manipulative capabilities (Susman, 1988). Prior to the discovery of the Swartkrans manual remains, many scientists assumed that brain size directly correlated with

PAGE 27

21 tool manufacture, but the small endocranial volume of the specimens, coupled with derived manual morphology, put this theory to rest (Susman, 1988). SKX 5016, from Member 1 at Swartkrans and attributed by Susman (1998) to Paranthropus is a PDP that shows clear markings of derived featu res. The distal phalanx has a broad apical tuft comparable to the dimensions seen in modern humans, a feature associated with increa sed precision grasping. The phalanx also sh ows a well defined area for the FPL muscle (Susman, 1998), which is functionally important for independent control of the thumb separate from the other digits of the hand (Lemelin and Diogo, 2016). SKX 5020 is a right pollical metacarpal also attributed to P. robustus. Unlike the long, gracile metacarpals of many Old World monkeys (Pa tel and Maiolino, 2016), SKX 5020 is robust, with a broad proximal surface like that seen in modern humans (Marzke et al., 2010). SKX 5020 displays a clear crest for the opponens pollicis muscle (Susman, 1998), indicating Paranthropus had an incr ea sed rang e of motion in the thumb relative to extant primates and other, earlier, hominins (Marzke, 1997). Homo Homo habilis. Evidence for tool usage in the hand of early H habilis comes from multiple carpal and phalangeal remains widely attributed to OH 7 ( Napier, 1962a; Susman and Creel, 1979; Marzke, 1997 ). Attribution of the manual remains from OH 7 into the H. habilis holotype met with contention almost immediately upon the initial publication, and while most scholars agree with the inclusion into H. hab ilis there is still dispute Moya Sol a et al. (2008 ) note that the strongest evidence for including the manual remains with the craniodental remains from the same site has generally rested on the similarity in ages, as both the craniodental and manual rem ains are clearly from a sub adult individual. The manual remains were not recovered near the cranial remains, casting doubt on the validity of the

PAGE 28

22 claim that they warrant inclusion into the OH 7 holotype. Pedal remains recovered from the same site have bee n attributed to a second individual, OH 8, further complicating the matter of proper placement for the manual remains. For the sake of simplicity, however, this paper will use the current attribution and refer to the OH 7 hand remains as belonging to H. ha bilis For a more in depth review of the controversies surrounding the taxonomic affinity of the OH 7 hand rem ains, see Moya Sola et al. (2008 ). The manual remains of OH 7, which include middle and distal phalanges, as well as various carpal elements, hav e all been assigned as belonging to the right side of a subadult in dividual (Moya Sola et al., 2008 ). Of particular interest for studies concerning precision grip capabilities are the trapezium and the distal phalanges. The distal phalanges have broad tuft s, the presence of which is correlated with a higher level of precision grasping capability. The trapezium has a broad metacarpal surface that is flatter than what is seen in modern human trapeziums. Marzke (1997) notes this feature is associated with more even distribution of larger internal forces, which allow for greater stresses to be transmitted through the joint surface during tool manufacture. Unfortunately, there are no known metacarpals associated with the OH 7 hand, and so a discussion on the then ar muscle morphology is not possible (Susman and Creel, 1979). Homo floresiensi s. H. floresiensis was a small bodied hominin from the Indonesian island of Flores, first described in 2004 and dated to between 100 60 k a (Brown et al., 2004; Sutikna et al., 2016). Included within the manual remains for the species are carpals, phalanges, and partial metacarpals from the type specimen LB1 (Tocheri et al., 2007; Larson et al., 2009; Orr et al., 2013), and several carpal bo nes from other individuals (Larson et al., 2009; Orr et al., 2013). The wrist of H. floresiensis is more like extant African apes than modern humans

PAGE 29

23 or Neanderthals (Tocheri et al., 2007; Orr et al., 2013), while the phalanges do show some evidence of deri ved morphological characteristics (Larson et al., 2009). Three complete carpals were discovered with the remains of LB1, including a scaphoid, capitate, and trapezoid, as well a partial lunate and hamate (Tocheri et al., 2007; Orr et al., 2013). The thre e complete carpal bones all display a symplesiomorphic morphology that is shared with extant African apes, including a wedge shaped trapezoid and a triangular articular surface on the scaphoid, and waisting along the radial aspect of the capitate, which i s like what is seen in the australopiths (Tocheri et al., 2007; Larson et al., 2009). Larson (2009) lists three possible metacarpals known from H. floresiensis. Of these, one is possible a metatarsal, one was lost during transfer, and the last is describ ed here. The solitary metacarpal definitively known from the species lacks both the proximal and distal ends, though the remaining shaft does show moderate signs of muscle markings for the interosseous muscles. The phalanges associated with H. floresiensis are all partial, though still informative. Two distal manual phalanges are known for the species, one a pollical and one a non pollical phalanx. The PDP has a well defined pit associated with the insertion for the FPL (Larson et al., 2009), which is assoc iated with enhanced precision grasping abilities (Susman, 1988). Homo naledi H. naledi is a recent fossil hominin discovery from the Dinaledi Chamber of the Rising Star Cave System in South Africa, dated to between 236 ka and 335 ka ( Berger et al., 2015 ; Dirks et al., 2017 ). The find remains the largest single accumulation of hominin bones to date, with nearly 150 elements recovered from the hand alone. The phalanges of H. naledi exhibit markedly primitive morphology, although the wrist of the species is more like

PAGE 30

24 later hominins and modern H. sapiens Hand 1 is the most complete hand recovered and consists of 26 bones from the right side of one individual. The pollical metacarpal of Hand 1, along with six other pollical metacarpals recovered from different specimens, show that H. naledi possessed pollical metacarpals with a robust distal end, but a narrow proximal base. The pollical metacarpal of Hand 1 exhibits well developed atta chment sites for both the opponens pollicis and the first dorsal interosseous muscle (Kivell et al., 2015). The carpal bones from H. naledi fall within the range of variation seen in Neanderthals and modern humans, and are derived relative to extant great apes. The carpal bones from Hand 1 exhibit a flat TCMJ a facet on the trapezium that extends onto the scaphoid tubercle, and a boot shaped trapezoid. These are all features shared with Neanderthals and modern humans, and most likely serve to distribute co mpressive loads incurred during strong precision gr ips involving the thenar musculature (Marzke, 1997; Kivell et al., 2015). This morphology contrasts with the more gracile pollical metacarpals of earlier australopiths (Marzke, 1983) and many Old World mo nkeys (Patel and Maiolino, 2016), and more closely aligns with the metacarpal morphology seen in modern humans and Neanderthals (Niewoehner, 2001; Niewoehner, 2006). Additionally, the PDP of H. naledi are robust, with a broad apical tuft, which has previou sly been discussed as a derived adaptation for precision grasping (Almecija et al., 2010 ). The combination of these morphological features would have facilitated a forceful pad to pad precision grasp like that seen in modern humans (Kivell et al., 2015). I n contrast to this suite of derived metacarpal traits the phalanges of H. naledi retain a primitive morphology, being relatively long and exhibit ing a marked curvature. Kivell et al. (2015) note the mean phalangeal curvature seen in H. naledi is nearly the same as that seen in Australopithecus afarensis and does not vary significantly

PAGE 31

25 from the curvature exhibited in extant African ape phalanges. The long, curved phalanges seen in H. naledi provide strong evidence that this species spent consid erable time in the trees and was adapted to arboreal locomotion. However, the combination of derived wrist structure with primitive phalangeal morphology suggest s that H. naledi represents a hominin that practiced arboreal locomotion while possessing the a bility to perform enhanced precision grips, which is in contrast with anything previously known from the hominin fossil record (Kivell et al., 2015). Homo neanderthalensis Early research into Neanderthal hand morphology, particularly in regards to the th umb, concluded that their thumbs were too short to have facilitated precise manipulative capa bilities (Niewoehner et al., 200 3). Vlcek (1975) argued that Neanderthals would have been incapable of fine precision grasping based on their thenar muscle morphol ogy. Vlcek (1975) argued Neanderthals would have been incapable the same precision grip seen in modern humans based on the placement of the opponens pollicis muscle, which he stated was shifted too far to the palmar side of the bone to allow for full media l rotation of the thumb. He based his conclusions on specimens from the cave of Kiik Koba, which is a Mousterian site in Ukraine first excavated in the early 20 th century (Vleck, 1973). However, Niewoehner (2006:165) states there is, no anatomical or func tional basis for the...claims that Neanderthal thumbs were less mobile than those of modern humans ". Neanderthal thumbs are in fact highly derived for precision grasping, and subsequent digital analyses concluded there is nothing in their morphology which would have precluded them from achieving the same degree of manual dexterity as that seen in modern humans (Niewoehner et al., 2003). In a broad of study Neanderthal manual remains, Niewoehner (2006) concluded Neanderthal fingers showed evidence of hypertrophied muscles, which, in

PAGE 32

26 conjunction with other morphologies of the hand, meant they were most likely capable of very powerful grasps. Both the opponens pollicis, which is responsible for much of the opposition of the thumb, and the opponens digiti minimi, which aids in radial rotation of the fifth metacarpal during cupping movements of the hand, showed evidence of hypertrophy. Niewoehner (2006) also noted enlarged tubercles on the trapezium, h amate, and scaphoid, which he notes conferred increased mechanical advantages in the carpometacarpal and carpophalangeal joints. Additionally, the distal pollical phalanges of Neanderthals show evidence of a radioulnarly expanded apical tuft like the dimen sions seen in modern humans, which has been previously noted as aiding in precision grasping of the hand (Trinkaus, 1983). This suite of morphological features would have facilitated powerful, precision grasping in Neanderthals. Morphology of the last com mon a ncestor There is currently no known specimen represe nting the LCA of humans and chimpanzees. As such, hypotheses regarding the morphology of the LCA must rely on inferences made based on the known morphological similarities between extant non human a pes, humans, and fossil hominin specimens. Tocheri et al. (2008) argue that, based on parsimony, the hand of the LCA most likely resembled th at of extant African apes, rather than resembling the more arboreal Asian apes. Following this argument, it can be inferred that all hand morphologies shared by extant great apes and non hominid outgroups are homologous and were most likely present in the hand of the LCA (Tocheri et al., 2008). Table 1 lists a summary of some of the morphological and myological feature s that Tocheri et al. (2008) hypothesize to have been present in the LCA, and their state within modern humans.

PAGE 33

27 Table 1 Inferred m orphological and myological hand features of the LCA and their state in modern humans. From Tocheri et al., 2007 and Larson et al., 2009 Feature Inferred character state of LCA Character state in modern humans Relative finger length Long fingers relative to thumb Thumb is long relative to other digits Proximal phalangeal curvature Curved dorso palmarly Straight dorso palmarly Mc1 robusticity Gracile Robust Apical tufts (distal phalanges) Narrow Broad Flexor pollicis longus Absent or reduced with no separate tendon belly Separate tendon belly Opponens pollicis Most likely occupies a relatively small section of bone Relatively enlarged when compared to Pan Neck of capitate Capitate has a waisted neck on the radial side* Expanded appearance on the radial side Styloid process on Mc3 Absent Present *This feature is also seen in the australopiths and Homo floresiensis as discussed in an earlier section of this paper Theories for the Evolution of the Modern Human Hand The rise of habitual bipedality, which freed the hands from locomotor restraints, has generally been credited as the c atalyst for the evolution of modern human hand morphology (Susman, 1994; Kivell, 2015). Habitual arboreal locomotion is characterized by long, curved phalanges which inhibit precision grasping and manual dexterity (Marz ke, 1997). As hominins moved out of the trees and increasingly practiced a terrestrial, bipedal lifestyle, the need for longer, curved phalanges lessened, and the hands instead became increasingly

PAGE 34

28 adapted for manipulation, as opposed to suspensory aids (Ma rzke, 1997). Evidence from hominin remains indicates that the morphological features associated with stone tool usage (broad apical tufts, a long thumb relative to digits, less phalangeal curvature, etc.) arose before the advent of stone tools, likely as a response to increased manipulation related to other activities such as feeding (Kivell, 2015). Recently, it has also been proposed that the modern human hand configuration evolved as a pleiotropic by product of the need for shorter toes to facilitate b ipe dal locomotion (Rolian, 2009 ). As new fossil hominin finds come to light, new evidence suggests the modern layout might in fact be quite primitive, and more like early Miocene apes (Almecija et al., 2010). If this is the case, the hand morphology seen in e xtant apes and early hominins, characterized by long, curved fingers and a relatively short thumb, would in fact be the derived character state. The following section will further discuss the t heories for the evolution of modern human hand morphology, incl uding the rise of habitual stone tool usage as a driver for enhanced precision grasping capabilities. The Rise of Habitual Stone Tool U se The shift from an arboreal lifestyle to habitual bipedalism has long been viewed as the catalyst that drove the evolution of modern human hand morphology (Susman, 1994; Panger et al., 2002). As humans increasingly moved about on two feet, the need for long, curved phalanges lessened, leaving the hands free for other, non locomotor, duties. Likely, the initial driver for increased precision grasping and manipulative capabilities began with foraging behavior, as hominins began to exploit new resources for food (Almecija and Alba, 2014). Extant chimpanzees ( Pan troglodytes ) have long been known to employ branches for extractive foraging purposes, and it has been proposed that early hominin tool use began in such a fashion, later shifting to stones as a primary tool source (Carvalho et al., 2012 ).

PAGE 35

29 However, these organic materials do not preserve in the foss il record, and so at present, studies must focus on preservable materials such as lithic and faunal remains to infer the advent of stone tool usage within the hominin lineage. The discovery at Dikika, Ethiopia of two cut marked bones push back the emergence of stone tool usage to ci rca 3.4 Ma, roughly 800 ka earlier than previously thought, though the alleged cut marks on the Dikika bones remains controversial (Semaw et al. 2003; McPherron et al. 2010) The stone tools from the si te of Gona, Ethiopi a, provide the earliest definitive proof of stone tool usage only within the hominin lineage at roughly 2.5 Ma (Semaw et al., 2003). However, this still begs the question of why stone tool usage only arose within the hominin lineage and not in other primat es Theories abound on this subject, and there is no room in this manuscript for a full review (but see Panger et al., 2002; Sanz and Morgan 2013 ) for a more thorough review of the matter). For the sake of brevity, the discussion of why tool use arose only within hominins will be limited to one possible theory. The Hand is a Pleiotropic Result of Selection for the F oot Derived morphological trai ts within the hand first appear within the hominin fossil record around roughly the same time as adaptations for bipedality (Richmond and Jungers, 2008; Rolian et al., 2011 ). Or (ca. 6 Ma), whose manual elements represent the earliest hominin hand remains to date, exhibit derived distal phalanges, in conjunction with femoral morphology indicative of a biped (Almecija et al., 2010; Almecija et al., 2013). Or. tugenensis possessed broad apical tufts related to precision grasping, demonstrating these morpholog ical adaptations arose quite early within the hominin lineage. The early emergence of these derived features, coupled with the nearly homologous structures and similar developmental pathways of manual and pedal elements, led to the theory that the adapta ti ons

PAGE 36

30 coevolved (Rolian, 2009 ) If this hypothesis is correct, the growth of the hands and feet of extant primates would have to follow the same ontogenetic pathways and produce similar phenotypic results. Further, hominin fossils should exhibit similar pr oportions in the pedal and manual phalanges as is seen in modern humans. The results of the study conducted by Rolian et al. (2009 :1563) demonstrated that, in Homo sapiens and Pan troglodytes the variation in growth of the manual and pedal phalanges is co nstrained by the same developmental blueprint supporting the hypothesis that their ontogeny within primates is linked. The hominin fossil record, however, is currently too sparse to provide either positive or negative proof of their hypothesis. The data do show that by the time of Au. afarensis (ca. 3.5 Ma), manual and pedal proporti ons within hominins were intermediate between modern humans and chimpanzees. Additionally, by the time of early Homo (ca. 1.8 1.5 Ma), hominin hand and foot proportions appear to be quite like those seen in mo dern humans (Rolian et al., 2009 ). This theory contrasts with the longstanding belief that human hands evolved after the shift to habitual bipedalism, independent of the selective pressures acti ng upon pedal morphology (M arzke, 2013). Rolian et al. (2009 ) contend their multiple lines of evidence, including genetic, fossil, and, archaeological, disprove the previous theory and offer a new proposal for the evolution of modern manual morphology. The fossils of Orrorin tugenensis which demonstrate that derived traits within the hominin hand emerged much earlier than the evidence for stone tool usage (McPherron et al., 2010), make t he theory by Rolian et al. (2009 ) particularly tempting to accept. It is becoming increasingly clear that the morphological traits previously assumed to correlate with precision grasping and stone tool manufacture (Susman, 1994; Marzke, 1997) appeared much earlier than previously thought,

PAGE 37

31 and most likely evolved within the h ominin lineage independently of stone tool usage or manufacture. Conclusion The hands of living and fossil primates display a wide range of morphological variation and this is especially true for the pollical metacarpal T he long, robust thumbs of modern humans are adapted for enhanced precision grasping, as reflected in the broad shaft, expanded proximal bases, and broad distal pollical phalanges In contrast, non human primates generally possess more gracile metacarpals and longer fingers rela tive to their thumbs (Napier, 1962 a ; Susman, 1979; Marzke, 1997) The morphological evidence from the hominin fossil record shows the pollical metacarpal gradually became more robust over the course of human evolution. Australopithecus possessed gracile po llical metacarpals with narrow proximal bases ( Susman, 1979; Berger et al., 2010) suggesting this is the primitive character state for hominin pollical metacarpals. By the time of H. neanderthalensis and early anatomically modern humans, the pollical meta carpal had become much more robust, with a distinct crest for the opponens pollicis and a broad TCMJ facet (Niewoehner, 2001; Niewoehner, 2006) What remains unclear is in what sequence these morphological adaptations arose in the genus Homo Did adaptations in the thumb of primitive Homo begin with an expansion of the proximal facet, greater recruitment of the intrinsic musculature, resulting in a broader base and more prounced muscle flanges, or did these features evolve in unison? The fossi l remains from H. naledi afford the opportunity, for the first time, to study the pollical metacarpals of primitive Homo to better understand this timeline. The following chapter s present a geometric morphometric analysis of a broad comparative sample of e xtant and

PAGE 38

32 fossil primate pollical metacarpals aimed at quantifying the shape of the H. naledi first metacarpal to answer the question of whether H. nadedi repre sents an auta pomorphic taxon in terms of pollical metacarpal morphology or whether it is eviden ce of a transitional metacarpal morphology in primitive members of Homo

PAGE 39

33 CHAPTER II MATERIALS AND METHODS Materials This study includes 3D virtual renderings of Homo sapiens (n=178 [(Table 2 ] ), non human apes (n=86 [Table 3 ] ), Old World cercopithecine monkeys (n =59 [Table 4 ] ), and fossil hominins (n=14 [T able 5 ] ). All left metacarpals were reflected in Geomagic prior to landmarking to create a full right side sample While some extant specimens were labelled by sex, time limitations prohibited any analyses based on sex. Additionally, since many specimens could not be identified by sex, analyses based on sex would have been possible only for a very reduced number of specimens. All scans were provided courtesy of Drs. Caley Orr, Matthew Tocheri, Biren Patel, and Tracey Kivell. Figure 7 shows a 3D representation of the extant homin o id sample. Figure 8 shows a 3D representation of the cercopithecine sample. Figure 9 shows a 3D representation of the fossil hominin sample. Table 2 Homo sapiens sample Population Male Female Unknown sex Total African American 6 11 0 17 Aleut /Pre Aleut 0 0 15 15 Australian 0 0 8 8 Chinese 0 0 24 24 European 10 4 7 21 European American 12 12 0 2 4 Indigenous African 11 7 4 2 2 Indigenous American 0 10 10 Japanese 2 1 0 3 Nubian/Egyptian 6 3 0 9

PAGE 40

34 Unknown 0 0 25 25 Total 47 38 93 178 Table 3 Non human ape sample. Tax on Male Female Unknown sex Total Gorilla beringei 5 5 1 11 Gorilla gorilla 15 10 0 25 Hylobates lar 7 4 0 11 Pan paniscus 9 7 0 16 Pan troglodytes 11 9 3 23 Pongo sp. 7 4 0 11 Total 54 39 4 97 Table 4 Cercopithecine monkey sample. Taxon Male Female Unknown Total Erythrocebus patas 7 5 0 12 Macaca fasicularis 9 1 0 10 Nasalis larvatus 6 3 3 12 Papio sp. 4 5 5 14 Total 26 14 8 48

PAGE 41

35 Table 5 Fossil hominin sample. Specimen Total Australopithecus afarensis A.L.333 1 Australopithecus sediba MH2 1 Homo naledi U W 101 270 U W 101 282 U W 101 1321 U W 101 1641 4 Homo neanderthalensis Kebara 2 La Chapelle La Ferrassie* Regourdou Shanidar 4 6 Homo sapiens Qafzeh 9 1 SK 84 ** 1 *Both left and right metacarpals from this specimen were used The left metacarpal was mirrored in Geomagic to reflect a right. **This specimen has been variously attributed to Paranthropus (Susman, 1988), Homo, and Australopithecus (Kivell, 2015).

PAGE 42

36 Figure 7 Representative 3D surface models of the homin o id sample. Figure 8 Representative 3D surface models of the cercopithecine sample. Hylobates lar P an troglodytes Homo sapiens Pongo pygmaeus Gorilla beringei Erythrocebus patas Macaca fasicularis Nasalis larvatus Papio anubis

PAGE 43

37 Figure 9 Representative 3D surface models of the fossil hominin sample Methods Using the Patch tool in the Landmark Editor Software (Wiley et al., 2005; Wiley, 2007), n ine landmarks were placed across Mc1 palmar diaphyseal shaft and a 20x20 semilandmark grid w as placed across the palmar diaphyseal surface (Fig. 10; Table 6). From the nine landmarks, a 20x20 semilandmark grid was designated across the Mc1 pal mar diaphyseal surface. Landmarks are point locations that must be biologically homologous between specim ens, but for many biological structures, such as the muscle attachments on the pollical metacarpal, landmark positions cannot be made homologous between structures because of the curves of those surfaces. For these analyses, one must use semilandmarks, whi ch allows for the analysis of two or three dimensional curves and surfaces (Gunz and Mitteroecker, 2013). Gunz and Mitteroecker (2013) argue it is better to have densely spaced landmarks to capture the greatest detail in terms of morphology, as well as to estimate for H. neanderthalensis (La Chapelle) SK 84 Homo naledi (U.W. 1321) Au. afarensis (A.L. 333) Au. sediba (MH2)

PAGE 44

38 missing data. Landmark placement was chosen to best capture the overall morphology of the palmar shaft. Figure 10 The lo cation of the landmark and semi landmark data points on a H. sapiens right Mc1. The numbered yellow dots indicate the original nine landmarks placed on the Mc1 surface and the small red dots indicate the semilandmark grid

PAGE 45

39 Table 6 Anatomical description of the landmark placement. Landmark Number Landmark Placement 1. Most lateral point on the distal end of the shaft 2. Midpoint of the palmar articular margin on the head 3. Most medial point on the distal end of the shaft 4. Most lateral point at midshaft 5. Center of the palmar shaft 6. Most medial point at midshaft 7. Most lateral point of the base 8. Apex of the palmar beak 9. Most medial point of the base Following Scott (2015), o f the nine original landmarks placed across the palmar diaphyseal surface, the four corner points (i.e, la ndmarks 1, 3, 7, and 9 in Fig. 10 ) were selected to anchor the sliding semilandmarks, resulting in 396 evenly spaced total semilandmarks. During superimposition, the 396 semilandmarks were slid along the palmar diaphyseal surface by minimizing Procrustes distances ( Gunz and Mitteroecker, 2013 ). Sliding of the semilandmarks ensures optimal spacing between l andmarks and semilandmarks, and establishes geometric correspondence of the semilandmarks by removing the arbitrary effect s of initial spacing ( Gunz and Mitteroecker, 2013 ). Raw landmark data contains information on size, shape, and orientation and so to correct for these factors the data were transformed into shape variables (Gower, 1975; Rohlf and Slice, 1990; Gunz and Mitteroecker, 2013) using a generalized Procrustes analysis (GPA) p erformed in the geomorph package in R Console ( Adams and Otarola Castillo, 2013 ; R Core Team, 2016 ) The resultant semi landmark data are the Procrustes shape coordinates, which contain

PAGE 46

40 information on the original shape of each landmark configuration by spe cimen (Schroeder et al., 2017). A principal component analysis (PCA) was then performed on the variance/covariance matrix of the Procrustes shape coordinates to visualize the shape differences between specimens and to identify the major axes of variation using the geomorph package in R console (Dryden and Mardia, 1998; R Core Team, 2016; Schroeder et al., 2017) The deformation grid of shape change across principal components was done using the geomorph package in R Console. An ANOVA and Tukey s post hoc pairwise comparison were performed in the Statistica software package to determine between group differences along principal components axes Finally, to explore the effects of overall metacarpal size on morphology, a multivariate regression analysis of lo g centroid size on shape variables was computed by plotting multivariate regression scores against log centroid size Regression scores predict the location of each individual on the component (DiStefano et al., 2009). Because there was no direct measure f or body mass of the individuals, log centroid size was used as a proxy for overall body mass (Parr et al., 2011; Knigge et al., 2014 ) Reduced major axis lines were computed using the RMA software for reduced major axis regression ( Bohonak, 2004) and fitted using the program Statistica

PAGE 47

41 CHAPTER III RESULTS Full S ample This section will discuss the results of the principal components analyses of the full sample for Principal Component ( PC ) 1 3. PC scores beyond 3 represented less than 5% of the total variation within the sample, and so were excluded from discussion. PC1 shape and groupings PC1 captures 38% of the variation within the sample, and represents the variation in shaft breadth between taxa. The first principal component is most useful in separating out the great apes ( except for Pan ) and the younger fossil hominin sample (Neanderthals, SK 84, and Qafzeh 9) from the Old World monkeys and the australopiths. The plot of PC2 on PC1 (Fig. 11 ) shows that the negative side of PC1 represents taxa with broader shafts, including H. sapiens H. neanderthalensis and Gorilla sp., while t he positive end shows taxa with more gracile shafts, such as Pan and Hylobates. The Old World monkeys ( Erythrocebus Macaca Nasalis, and Papio ) as well as the Pongidae, all occupy shape space on the far positive side of PC1, reflecting the gracile first metacarpals of these primates. H. naledi occupies shape space on the positive end of the axis, though th ey are not as far on the positive end as the Old W orld monkeys and Pan. Au sediba and Au. afarensis both occupy shape space on the positive end of the axis, reflecting their gracile shaft morphology, like what is seen in Pan. SK 84 and Qafzeh 9 both occupy shape space on the negative side of the axis, clustering closely with H. neanderthalensis and modern humans.

PAGE 48

42 Figure 11 Scatterplot of PC1 on PC2. PC2 shape and groupings PC2 captures 8% of the variation within the sam ple, and represents the size of the base relative to the breadth of the shaft. Shape space on the negative end of the PC2 axis is occupied by much of the Gorilla sample, as well as roughly half of the Pan sample. Except for some of the Papio specimens, the cercopithecines all occupy shape space on the negative end of the PC2 axis

PAGE 49

43 H. neanderthalensis Qafzeh 9, SK 84, and H naledi all occupy shape space solely on the positive end of the PC2 axis. H naledi occupies shape space almost completely separately from all other taxa with only minimal overlap with H. sapiens and H. neanderthalensis Pongo, Hylobates and modern humans all occupy shape space on both the positive and negative ends of the axis. Taxa on the negative side of PC2 possess a shaft with a br oader distal end and a relatively narrower proximal facet. On the positive side of PC2 are taxa with a broader proximal base and a narrower distal end. The variation in breadth of the distal end of the shaft is most likely the result of variation in size o f the muscle flanges. H. naledi and H. neanderthalensis both on the positive end of PC2, have very pronounced crests for the opponens pollicis muscle, while taxa on the negative end of PC2, such as the cercopithecine monkeys, have greatly reduced muscle m arkings for the opponens pollicis muscle. PC3 shape and groupings PC 3 (Fig. 12) captures 5% of the variation within the sample, and represents the breadth of the proximal end relative to the medial and lateral flanges on the distal portion of the shaft. Shape space on the negat ive side of PC3 represents taxa with more of an hourglass or waisted figure ( H. neanderthalensis SK 84, Au. sediba, Pongo and most of the Pan and Nasalis sample ) while taxa on the positive end are characterized by a mo re proximodistal uniform shape of the first metacarpal ( Papio, Erythrocebus, Macaca, and Hylobates ) As in PC1 and PC2 modern humans occupy shape space on both the positive and negative end of the PC3 axis. H. naledi occupies shape space on both the negative and positive end of PC3

PAGE 50

44 Figure 12 Scatterplot of PC1 on PC3. Regional Variation with Modern H umans The recent modern human sample (n=178) includes specimens from several geographic regions (Table 1). To assess morphological variation among the populations of recent H. sapiens a PCA was conducted on the modern human subsample. PC1 represents 26% of the variation within the sample, PC2 represents 11% of the variation, and PC3 represents 9% of the variat ion. PC scores beyond 3 represented less than 5% of the total variation, and so were omitted from analysis.

PAGE 51

45 A one way ANOVA was performed on the results from the PCA to test for morphological variation between regional g roups of modern humans. The res ults of the one way ANOVA show significant differences on PC1 and PC2. Table 7 shows the results of the one way ANOVA for the modern human sample. A Tukey s post hoc pairwise comparison was performed to find the source of variation between groups for PC s 1 and 2 (Tables 8 and 9 ). Values that are statistically significant (p<0.05) are denoted. Table 7 Results of a one way ANOVA for the modern human sample. PC1 PC2 PC3 p =0.00 p =0.00 p > 0.05 Red d esignates statistically significant at the p <0.05 level. Table 8 Results of Tukey s post hoc pairwise comparison results for PC1 for the modern human subsample Population 1 2 3 4 5 6 7 8 9 10 11 Unknown (1) 1.000 0.992 0.982 0.029 0.986 0.557 0.999 0.986 0.564 0.998 Japanese (2) 1.000 1.000 1.000 0.942 1.000 0.996 0.999 1.000 0.998 0.999 African American (3) 0.992 1.000 1.000 0.648 1.000 0.988 0.875 1.000 0.995 0.709 Indigenous African (4) 0.982 1.000 1.000 0.568 1.000 0.986 0.829 1.000 0.994 0.584 European American (5) 0.029 0.942 0.648 0.568 0.570 1.000 0.035 0.962 0.999 0.001 European (6) 0.986 1.000 1.000 1.000 0.570 0.985 0.844 1.000 0.994 0.619 Indigenous American (7) 0.557 0.996 0.988 0.986 1.000 0.985 0.338 1.000 1.000 0.149 Nubian Egyptian (8) 0.999 0.999 0.875 0.829 0.035 0.844 0.338 0.861 0.353 1.000 Australian (9) 0.986 1.000 1.000 1.000 0.962 1.000 1.000 0.861 1.000 0.760 Aleut/Pre Aleut (10) 0.564 0.998 0.995 0.994 0.999 0.994 1.000 0.353 1.000 0.124 Chinese (11) 0.998 0.999 0.709 0.584 0.001 0.619 0.149 1.000 0.760 0.124 Red d esignates statisti cally significant at the p <0.05 lev el.

PAGE 52

46 Table 9 Tukey s post hoc pairwise comparisons results for PC2 for the modern human subsample. Population 1 2 3 4 5 6 7 8 9 10 11 Unknown (1) 0.999 0.958 1.000 1.000 1.000 0.647 1.000 0.739 0.000 1.000 Japanese (2) 0.999 1.000 1.000 0.999 0.984 0.629 1.000 0.679 0.008 0.996 African American (3) 0.958 1.000 0.996 0.970 0.691 0.102 1.000 0.152 0.000 0.858 Indigenous African (4) 1.000 1.000 0.996 1.000 0.995 0.464 1.000 0.566 0.000 1.000 European American (5) 1.000 0.999 0.970 1.000 1.000 0.609 1.000 0.705 0.000 1.000 European (6) 1.000 0.984 0.691 0.995 1.000 0.940 0.991 0.965 0.001 1.000 Indigenous American (7) 0.647 0.629 0.102 0.464 0.609 0.940 0.542 1.000 0.471 0.808 Nubian Egyptian (8) 1.000 1.000 1.000 1.000 1.000 0.991 0.542 0.617 0.000 0.999 Australian (9) 0.739 0.679 0.152 0.566 0.705 0.965 1.000 0.617 0.481 0.871 Aleut/Pre Aleut (10) 0.000 0.008 0.000 0.000 0.000 0.001 0.471 0.000 0.481 0.000 Chinese (11) 1.000 0.996 0.858 1.000 1.000 1.000 0.808 0.999 0.871 0.000 Red d esignates statisti cally significant at the p <0.05 lev el. Reduced s ample The full sample is quite large, and so the decision was made to attempt to narrow down the sample size for further analyses. Based on the results of the full sample PCA analysis, it appeared Pan Pongo Hylobates and the cercopithecines all shared a simila r morphology. T hese taxa all occupy the same shape space on PC1 3 as the Australopithecus sample, and it is likely they reflect the primitive condition for hominins. As such, the author decided to let Pan represent the primitive condition for hominids and to exclude Pongo Hylobates and the cercopithecines from subsequent analyses if no significant variation was found to exist between the cercopithecines, Pan, Pongo, and Hylobates A one way ANOVA was performed to determine if variation existed between the

PAGE 53

47 cercopithecines and the hominid sample. The results show significant variation across PC1 3 (Table 10). A Tukey s post hoc pairwise comparison was then performed to determine the source of the variation between the groups (Table 11). Results from the T ukey s post hoc test showed no variation between Pan, Pongo, Hylobates and the cercopithecines for PCs 1 and 2. While there is some variation between Pan, Hylobates Pongo, and the cercopithecines on PC3, the decision was made to narrow down the sample si ze for further analyses to better focus on more closely related taxa. A one way ANOVA was then performed to examine variation between the hominid sample. The results showed significant variation between taxa on PC1 3 (Table 12). A Tukey s post hoc pairwise comparison was then performed to determine between which taxa there was statistically significant variation ( Table 13). Table 10 Results of a one way ANOVA for the Cercopithecoidea and hominid sample. PC1 PC2 PC3 p =0.00 p =0.00 p =0.00 Table 11 Tukey s post hoc pairwise comparison results for PCs 1 3 for the Cercopithecoidea and hominid sample s Genus Pairing PC1 PC2 PC3 Cercopithecidae Hylobates Not significant Not significant Not significant Cercopithecidae Gorilla p<0.05 p<0.05 p <0.0.5 Cercopithecidae Homo sapiens p<0.05 p<0.05 Not significant Cercopithecidae Pan Not significant Not significant p <0.0.5 Cercopithecidae Pongo Not significant Not significant p <0.0.5

PAGE 54

48 Table 12 Results of a one way ANOVA for the hominid sample s PC1 PC2 PC3 p =0.00 p =0.00 p =0.00 Red d esignates statistically significant at the p <0.05 level. Table 13 Tukey s post hoc pairwise comparison results for PCs 1 3 for the hominid sample Genus Pairing PC1 PC2 PC3 Homo sapiens Hylobates p <0.05 Not significant Not significant Homo sapiens Gorilla Not significant p <0.05 p <0.05 Homo sapiens Pan p <0.05 p <0.05 p <0.05 Homo sapiens Pongo p <0.05 Not significant p <0.05 Red d esignates statistically significant at the p <0.05 level. Hominoid sample The hominoid sample includes all specimens of Gorilla sp. Pan sp., H. sapiens, H. neanderthalensis Au. afarensis, Au. sediba, SK 84, Qafzeh 9, and H naledi. In the principal c omponent analysis, PC1 (Figs. 13 and 14 ) explained 33 % of the total variance, PC2 (Figs. 14 and 15) explained 10 % of the total variance, and PC3 (Figs. 16 and 17 ) explained 6 % of the total variance. PC scores beyond 3 represented less than 5% of the total variation within the sample, and so were excluded from this discussion. PC1 shape and groupings PC1 captures 33% of variation within the hominoid sample, and is most informative for distinguishing between extant apes (Fig s. 13 and 14 ) The negative side of the PC1 axis is mostly occupied by Homo sapiens, Gorilla, and Neanderthals, whil e Pan and Hylobates occupy the positive side of the axis. Gorilla beringei and Gorilla gorilla show a clear

PAGE 55

49 distinction between the two species in shape space, with Gorilla beringei clustering more on the negative side of the axis, and Gorilla gorilla occu pying more spa ce on the positive side of the axis. Qafzeh 9 lots squarely within the Homo sapiens sample. SK 84, also plots within the range of Homo sapiens Both Au. afarensis and Au. sediba occupy the same shape space as Pan and Hylobates on the first principal component axis, while Homo naledi occupies the shape space on the border between Homo sapiens and Pan. Figu re 14 shows a scatterplot of PC2 on PC1 for the hominoid sample Figure 13 T he morphological variation in the hominid sample across PC1 Homo sapiens Au. sediba (MH2) Pan troglodytes H. neanderthalensis (La Chapelle) Homo naledi (U.W. 1321)

PAGE 56

50 Figure 14 Scatterplot of PC2 on PC1. The positive side of the first PC axis represents taxa with a narrower metacarpal shaft, both mediolaterally and on both the proximal and distal ends. The positive side of the PC 1 axis is occupied by Pan and the australopiths, reflecting their more gracile musculature compared with the other extant apes and humans. Homo naledi also plots well within the positive side of the PC1 axis, suggesting the seemingly broader distal end of the metacarpal

PAGE 57

51 is a result of greater surface area for muscle entheseal development, rather than a broader overall shaft. H. na ledi s placement on the positive end of PC1 indicates that PC1 is not capturing muscle robusticity, as it might appear from qualitative observations, but rather only the overall range of breadth within the shaft itself The negative side of the PC1 axis is mostly occupied by Gorilla, H. sapiens and H. neanderthalensis all of which have broad metacarpal shafts reflective of relative degrees of muscle development. PC 1 also appears to capture something related to the size of the proximal and distal articular facets, with those taxa on the negative end having larger a surface area for the articulation with the trapezium. PC 2 shape and groupings PC2 (Figs. 14 and 15 ) captures 10 % of the variation within the hominoid sample. Both Gorilla gorilla and Gorilla beringei cluster almost entirely on the negative end of the second principal component axis. Both species of Pan cluster mostly on the negative end of the axis The Neanderthal sample occupies shape space on the positive end of the axis. Qafzeh 9 and SK 84 both occupy shape space on the slightly positive end of the PC2 axis. Au. sediba occupies shape space on the positive end of the axis, while Au. afarensis occupies shape space on the negative end of PC2 Homo naledi occupies shape space squarely on the po sitive end of the second principal component axis. Shape on the PC2 axis represent s the breadth of the proximal end of the shaft relative to the overall size of the shaft, as well as narrowing of the beak mediolaterally The impression of a narrowing and tapering of the shaft is most likely a result of the degree of muscle hypertrophy which gives the false impression of a broad overall shaft. PC2 captures robusticity of the first metacarpal, which is shown in the position of both Homo naledi and the Neanderthals. H. naledi is unique in having a small base relative to large, laterally flaring

PAGE 58

52 flanges, while Neanderthals have a relatively large base and large flanges. The negative portion of the PC 2 axis represents taxa with a broader overall shaft, as well as more narrowing of the beak in a lateral to medial direction. The taxa on the negative side of the axis all possess robust musculature (opponens pollicis and first dorsal interosseous), which serves to make the shaft broader on the medial and lateral sides of the bone. The positive side of the second principal components axis represents taxa with narrower overall shafts and lesser development of the first dorsal interosseous and opponens pollicis muscles. The shape of the positive side of the second principal component also represents taxa with almost no narrowing of the metacarpal beak, as is seen on the negative end of the second pri ncipal components axis. Figure 15 illustrates morphological variation acro ss PC2 for the hominoid sample. Figure 15 The morphological variation in the hominoid sample across PC2 Gorilla beringei Homo naledi (U.W. 1321) Homo sapiens at (0,0) Au. afarensis (A.L. 333) H. neanderthalensis (La Ferrassie)

PAGE 59

53 PC3 shape and groupings PC3 captures 6% of the variation within the hominoid sample, and is most informative in capturing the breadth of the proximal base relative to the laterally flaring distal flanges associated the opponens pollicis and first dorsal interosseous muscle. PC3 i s the only PC in which Homo naledi does not occupy shape space separate from any other taxa, but rather occupies the same shape space as Pan, H. sapien s and Gorilla gorilla. As in PC 1 Homo sapiens occupy shape space across most of both axes, reflecting the varied morphology of modern human hands. PC3 separat es out the Neanderthal specimens from other taxa. Pan occupies space mostly on the positive side of the third PC axis, along with Hylobates, which occupies space solely on the positive side of the ax is. Au. sediba occupies space on the positive side of the axis, while Au. afarensis occupies space on the negative end. The positive side of PC3 represents taxa with wide bases above which there is narrowing before flaring out into wide distal flanges t hus creating a waisted shape Neanderthals occupy shape space exclusively within the positive side of the PC3 axis, reflecting their unique morphology of wide proximal bases as well as laterally flaring distal flanges, mostly related to the opponens poll icis muscle. The negative end of PC3 seems to represent taxa with similarly broad proximal and distal ends, lending a more straight down look to the shaft, in contrast to the impression of waisting on the negative end of the axis. Figure 16 shows morpho logical variation across PC3 fo r the hominoid sample. Figure 17 shows a scatterplot for PC3 on PC1 for the hominoid sample.

PAGE 60

54 Figure 16 The morphological variation in the hominoid sample across PC3 Homo sapiens Gorilla beringei EAMH (Qafzeh 9) H. neanderthalensis (La Chapelle) Au. afarensis (A.L. 333)

PAGE 61

55 Figure 17 A scatterplot of PC3 on PC1 for the hominoid sample Shape and size A multivariate regression analysis of regression scores against log centroid size was performed for the entire dataset. Results of the multivariate regression analysis showed no significant relationship between shape a n d size ( p =0.06 ) N one of the taxa plot along the

PAGE 62

56 same allometric traj ectory as each other indicating that a change in size would not induce a change in metacarpal shape. Figure 18 shows the results of the multivariate regression analysi s. Figure 18 A plot of the multivar iate regression analysis showing the reduced major axis lines

PAGE 63

57 CHAPTER IV DISCUSSION AND FUTUR E RESEARCH This st udy used geometric morphometric techniques to quantify the shape of the first metacarpal in a broad comparative sample of extant and fossil primates. The aim of the study was to determine if H. naledi represents an autapomorphic taxon in terms of first metacarpal morphology The results presented here indicate that the first metacarpal of H. naledi is not auta pomorphic, but rather represents a transitional morphology for primitive members of the genus Homo. PC1 Breadth of the Metacarpal S haft The PCA results demonstrate that taxa can best be differentiated by the overall breadth of the first metacarpal shaft. In both the full sample and the reduced hominoid sample, the most discriminating factor was the overall breadth of the metacarpal shaft. Breadth of the shaft appears to be tied to two major behavioral components: locomotor patterns and manipulator behaviors. On the negative end of the PC 1 axis are specimens ( Homo sapiens, Homo neanderthalensis, Gorilla beringei SK 84, and Qafzeh 9) that regularly use their thumbs for manipulator purposes, whether through use of stone tools (Susman, 1988; Niewoehner, 2001), or for breaking down fibrous leaves and pi th for consumption (Rogers e t al., 2004; Rothman et al, 2007 ). Except for Gorilla beringei these taxa locomote bipedally, and their thumbs are free for use exclusively as a means of manipulation However, g orillas do not use the pollical metacarpal w hen e ngaged in knuckle walking and thus their thumbs serve mainly as a manipulator aid during feeding (Tuttle, 1967).

PAGE 64

58 Au. sediba and Au. afarensis occupy shape space on the positive end of PC1 reflecting their gracile shaft s Based on the morphology of the a ustralopiths it appears the primitive character state for hominins is a gracile metacarpal shaft. Additionally, it is likely that the gracile metacarpal shafts seen in Pan and Old World monkeys are plesiomorphic for hominins H neanderthalensis SK 84, Qafzeh 9, and modern humans all occupy shape on the negative side of PC1 because of their broader metacarpal shafts and it is therefore likely that a broader metacarpal shaft is the derived character state for hominins. A broader metacarpal shaft allows for greater development of the intrinsic musculature, and is indicative of taxa that use their thumbs more frequently. It is unclear whether hominins adapted a broader shaft because of increased tool usage (Susman, 1994) o r whether this morphology predates the appearance of stone tools. PC2 Breadth of the Proximal End Relative to Shaft B readth The second principal component in both the full and reduced analyses captures the variation in the breadth of the base of the poll ical metacarpal relative to overall shaft breadth A broader proximal facet allows more even distribution of compressive loads transferred across the trapeziometacarpal joint surface incurred because of enhanced precision grasping (Marzke, 1997). Gorilla b eringei, H. neanderthalensis and SK 84 all possess the widest proximal bases, suggesting these taxa are especially adapted for incurring large levels of stress at the TCMJ Marzke et al. (2010) proposed the wider bases observed within Gorilla are a result of the vigorous pulling that occurs during the processing of the fibrous vegetation they regularly consume (Rogers et al., 2004 ; Rothman et al., 2004 ). H. sapiens and Qafzeh 9 possess narrower proximal bases than the e arlier Neanderthals and SK 84 The reduction in proximal facet breadth could likely be the result of differences in technique employed during

PAGE 65

59 stone tool manufacture Niewoehner (2001) suggests early anatomically modern humans and modern humans adapted to manufacture stone tools in suc h a way as to alleviate some of the pressure being transferred through the TCMJ. It is possible modern humans manufactured stone tools in a manner that used less powerful grips than were employed by H. neanderthalensis. PC3 Breadth of the Proximal End R e lative to th e Breath of the Distal F langes PC3 in both analyses captures the breadth of the metacarpal base relative to the bread th of the distal muscle flanges The positive and negative ends of PC3 represent variation in levels of adaptations for powerful precision grasping. T he positive end of PC3 represent s specimens whose metacarpals have a waisted appearance, which is a res ult of an expanded proximal base combined with broad, mediolaterally flaring distal muscle flanges. The broader muscle flanges are likely a result of increased recruitment of the intrinsic musculature. Increased size of the intrinsic muscles would result in higher levels of compressive lo ads being transferred through the TCMJ and the broader proximal facet of the taxa on the positive end of PC3 shows these taxa were adapted for more even distribution of these forces through that joint (Marzke, 1997). Taxa on the negative end of PC3 have a relatively narrow proximal base compared to shaft breadth, and reduced distal flanges, possibly the result of less well developed intrinsic musculature. PC3 is most helpful for pulling out the Neanderthal specimens from the rest of the sample. H. neander thalensis hands are specifically adapted for powerful precision grasping, as evidenced by the hypertrophy of the thenar muscles and the broad, expanded bases of the pollical metacarpal (Niewoehner, 2006). Additionally, SK 84 occupies shape space within the Neanderthal sample, suggesting SK 84 was possibly also adapted for powerful precision

PAGE 66

60 grasping. While PC3 is most helpful in distinguishing the shape of the Neanderthal specimens from the rest of the taxa, it is important to note the most extreme outlier on PC3 is a Gorilla beringei specimen (Fig 17 ). The functional significance of the Gorilla metacarpal morphology is discussed in more detail in a following section. Implications for the M orphology of P rimitive Homo The only definitively known pollical metacarpals from members of the genus Homo who are thought to represent primitive Homo morphology are those of H. naledi and so direct comparison to other fossil hominin s pecimens is not possible at present. However, the results of this study allow for th e formation of hypotheses regarding the morphology of pollical metacarpals of other fossils from primitive Homo Overall, the hand of H. naledi displays a mosaic of primitive and derived conditions, with primitive, curved phalanges but a derived carpal morphology that is within the range seen in modern humans and Neanderthals (Kivell et al., 2015). The placement of H. naledi on the first two pri ncipal components axes on both the full and reduced sample indicates that the metacarpal shaft is not hyper robust, as it might appear from initial qualitative observations of the bone. Rather, the shaft itself is gracile, with large flanges superimposed o n both the medial and lateral sides. Additionally, the proximal base of H naledi is narrow, which is a primitive character state shared with Australopithecus A broad metacarpal base reduces stress across the TCMJ by increasing surface area and minimizing concentrations of force (Marzke et al., 2010). The narrow proximal base indicates H. naledi was not adapted for transferring the increased loads produced for powerful precision grasping. The large flanges are possibly correlated to increased muscle

PAGE 67

61 size o f the opponens pollicis and first dorsal interosseous muscles, though this relationship is not definitive (Wallace et al., 2017 ; Williams Hatala et al., 2016). The morphology of the first metacarpal of H. naledi indicates the character state for primiti ve members of the genus Homo was a gracile metacarpal shaft, surmounted by broad medial and lateral flanges The gracile morphology of the proximal shaft and base is like cercopithecids, Pongo, Pan and Au sediba and Au. afarensis suggesting that this character is likely plesiomorphic for hominins. In contrast, the broad shaft and metacarpal base, combined with mediola terally flaring muscle flanges seen in Neanderthals, appear to be an autapomorphy for the species. This hypothesis is supporte d by the more modern human like morphology of the Qafzeh specimen, which demonstrates a reduction in both the size of the muscle flanges and the breadth of the proximal base in later Homo specimens. H. naledi represents a transitional state between the ove rall gracility of the primitive Australopithecus shaft and the broader, derived shafts of later hominins. H naledi demonstrates the piecemeal nature of morphological adaptation within the human hand. The broad flanges indicate greater recruitment of the intrinsic muscles, which, coupled with a derived suite of carpal bones, indicates primitive members of the genus Homo possibly adapted first to produce increased forces, with the expansion of the proximal base following. The broader surface area of the pr oximal base seems likely to have evolved as a means through which to lessen the forces transferred through the TCMJ surface. This suggests the hands of primitive members of the genus Homo were suited for powerful grasping before their joint surfaces could adapt to bear the increased loads being transferred because of the increased muscle mass.

PAGE 68

62 Future Research Variation within Gorilla spp Past studies (Rogers et al., 2004; Rothman et al, 2004; Knigge et al., 2014) have documented the morphological differences between eastern ( G orilla beringei) and western ( Gorilla gorilla ) gorilla populations, which are heavily influenced by the local ecology and habitat of the surrounding area. In a study of diet across six different populations of western gorillas Rogers et al. (2004:186) found western gorillas have the greatest diversity in their diets of any species of Gorilla regularly consuming pith, leaves, fruit, though they label pith and leaves as fallback foods for western gorilla. Western gorillas hea vily consume seasonal fruits, with populations from each of the six sites consuming fruits regularly throughout the year. In contrast, the eastern gorilla ( Gorilla beringei ), inhabits more mountainous regions, where seasonal fruit is either scarce or unava ilable. Thus, the diet of eastern gorillas relies heavily on leaves, pith, and stems (Rothman, 2004). Differences in their trophic habits are reflected in the bony morphology of these two groups, with Gorilla beringei having a much broader metacarpal shaft and proximal base than Gorilla gorilla most likely because of increased use of the thumb due to time spent breaking down pith and fibrous leaves for food consumption. Future research comparing the feeding habits o f different species of Gorilla will serve to further i lluminate possible reasons behind the metacarpal morphological variation. Additionally, more analyses on different Gorilla skeletal elements will inform as to whether the trend of morphological variation between species of Gorilla continues throughout the skeleton. Sexual dimorphism

PAGE 69

63 Time limitations did not allow for analyses that considered size and shape of the first metacarpal based on sex, though it is very possible differences in metacarpal morphology within the present samp le are in some part a result of sexual dimorphism. Gorillas, orangutans, baboons, and proboscis monkeys are the most sexually dimorphic primates, and among the cercopithecine monkeys, males tend to be between 30 80% larger than their fe male counterparts (P lavcan, 2002). Any future analyses on these data s hould control for sex to explore its possible on shape and size of the first metacarpal. Conclusion This work aimed to quantify the shape of the palmar first metacarpal shaft across primate species using a 3D geometric morphometric analysis. The shape of the first metacarpal is influenced by the behavioral repertoire of the organism in question. In primates, the vast differences in forms of locomotor patterns and manipulatory behaviors are reflected in th e high levels of morphological variation present across the thumbs of different taxa. The results of this study demonstrate variation both within and between primate species in regards to the shape of the first metacarpal. The cercopithecines, Pan and th e arboreal apes Hylobates and Pongo are all characterized by gracile metacarpal shafts with reduced muscle entheseal morphology. This morphology is like Australopithecus suggesting this is the primitive character state for hominin first metacarpals. In co ntrast, modern humans, Neanderthals, and Gorilla all possess broader metacarpal shafts with much more defined attachment sites for the intrinsic musculature. There is a pronounced degree of morphological variation between species of Gorilla which is consi stent with previous

PAGE 70

64 research (Knigge et al., 2014), and which warrants future research to better understand the behavioral implications behind this variation. H. naledi is distinguished from all other primates by the combination of a gracile metacarpal shaft, broad flaring muscle flanges, and a diminutive proximal base. These results suggest that early evolution of the thumb within Homo began with a selection for larger intrinsic musculature, prior to adapting an expanded proximal base to help alleviate the associated increased compressive loads. Based on this study, we expect any subsequent finds of primit i ve Homo pollical metacarpals to reflect this morphology.

PAGE 71

6 5 REFERENCES Adams DC, Otarola Castillo E. 2013. geomorph: an R package for the collection and analysis of geometric morphometric shape data. Methods Ecol Evol. 4:393 399. Alba DM, Moya Sola S, Kohler M. 2003. Morphological affinities of the Australopithecus afarensis hand on the basis of manual proportions and relative thumb length. J Hum Evol 44(2):225 254. Almecija S, Moya Sola S, Alba DM. 2010. Early origin for human like precision grasping : a comparative study of pollical distal phalanges in fossil hominins. PLoS One 5(7):e11727. Almecija S, Tallman M, A lba DM, Pina M, Moya Sola S, Jungers WL. 2013. The femur of Orrorin tugenensis exhibits morphometric affinities with both Miocene apes and later hominins. Nat Commun 4:2888. Almecija S, Alba DM. 2014. On manual proportions and pad to pad precision grasping in Australopithecus afarensis J Hum Evol 73:88 92. Almecija S, Smaers JB, Jungers WL. 2015. The evolution of human and ape hand proportions. Nat Commun 6:7717. Berger LR, de Ruiter DJ, Churchill SE, Sch mid P, Carlson KJ, Dirks PH, Kibii JM. 2010. Australopithecus sediba : a new species of Homo like australopith from South Afric a. Science 328(5975):195 204. Berger LR, Hawks J, de Ruiter DJ, Churchill SE, Schmid P, Delezene LK, Kivell TL, Garvin HM, Williams SA, DeSilva JM et al. 2015. Homo naledi a new species of the genus Homo from the D inaledi Chamber, South Africa. e life 4. Bohonak AJ. 2004. RMA: Software for Reduced Major Axis Regression. 1.17 ed. San Diego State University. Brown P, Sutikna T, Morwood MJ, Soej ono RP, Jatmiko, Saptomo EW, Due RA. 2004. A new small bodied hominin from the Late Pleistocene of Flores, Indones ia. Nature 431(7012):1055 1061. Carvalho S, McGrew W, and Dominguez Rodrigo M. 2012. The origins of the Oldowan. Stone Tools and Fossil Bones. p 201 221.

PAGE 72

66 Diogo R, Richmond BG, Wood B. 2012. Evolution and homologies of primate and modern human hand and fo rearm muscles, with notes on thumb movements and tool use. J Hum Evol 63(1):64 78. Dirks PH, Roberts EM, Hilbert Wolf H, Kramers JD, Hawks J, Dosseto A, Duval M, Elliott M, Evans M, Grun R et a l. 2017. The age of Homo naledi and associated sediments in th e R ising Star Cave, South Africa. e life 6 :e24231 Distefano C, Zhu, M., Mindrila, D. 2009. Understanding and Using Factor Scores: Considerations for the Applied Researcher. Practical Assessment, Research, and Evaluation 14(20):1 11. Dryden IL, Mardia, K.V. 1998. Statistical Shape Analysis. New York: John Wiley & Sons. Eliot DJ, Jungers WL. 2000. Fifth metatarsal morphology does not predict presence or absence of fibularis tertius muscle in hominids. J Hum Evol 38(2):333 342. Gower J C. 1975. Generalized P rocrustes analysis. Pyschometrika 40(1):33 51. Green DJ, Gordon AD. 2008. Metacarpal proportions in Australopithecus africanus J Hum Evol 54(5):705 719. Gunz P, Mitteroecker, P. 2013. Semilandmarks: a method for quantifying curves a nd surfaces. Hystrix 24(1):103 109. Haile Selassie Y, Latimer BM, Alene M, Deino AL, Gibert L, Melillo SM, Saylor BZ, Scott GR, and Lovejoy CO. 2010. An early Australopithecus afarensis postcranium from Woranso Mille, Ethiopia. Proc Natl Acad Sci US A 107( 27):12121 12126. Hamrick MW, C hurchill, S.E., Schmitt, D., Hylander, W.L. 1998. EMG of the human flexor pollicis longus muscle: implications for the evolution of hominid tool use. J Hum Evol 34:123 136. Johanson DC, Taieb, M., Coppens, Y., Roche, H. 1980 New discoveries of Pliocene hominids and artifacts in Hadar international Afar Research Expedition to Ethiopia (4th and 5th field seasons, 1975 1977 ). J Hum Evol 9:583 585. Kivell T, Kibii, JM., Churchill, SE., Schmid, P. Berger, LR. 2011. Australopithecus sediba Hand Demonstrate Mosaic Evolution of Locomotor and Manipulative Abilities. Science 333(6048):1411 1417. Kivell TL. 2015. Evidence in hand: recent discoveries and the early evolution of human manual manipulation. Philos Trans R Soc Lond B Biol Sci 370(1682). Kivell TL, Deane AS, Tocheri MW, Orr CM, S chmid P, Hawks J, Berger LR, Churchill SE. 2015. The hand of Homo naledi Nat Commun 6:8431.

PAGE 73

67 Kivell TL. 2016. The primate wrist. In: Kivell TL, Lemelin, P., Richmond, B.G., Schmitt, D., editor. The Evolution of the Primate Hand. New York: Springer. p 17 54. Knigge RP, Tocheri MW, Orr CM, and McNulty KP. 2015. Three dimensional geometric morphometric analysis of talar morphology in extant gorilla taxa from highland and lowland habitats. Anat Rec 298(1):277 290. Larson SG, Jungers WL, Tocheri MW, Orr CM, Mor wood MJ, Sutikna T, Awe RD, Djubiantono T. 2009. Descriptions of the upper limb skeleton of Homo floresiensis J Hum Evol 57(5):555 570. Leakey LS, Tobias PV, Napier JR. 1964. A New S pecies of the Genus Homo from Olduvai Gorge. Nature 202(4927):7 9. Lemelin P, and Diogo, R. 2016. Anatomy, Function, and Evolution of the Primate Hand Musculature. In: Kivell TL, Lemelin, P. Richmond, B.G., Schmitt, D. editor. The Evolution of the Primat e Hand. New York: Springer. p 155 193. Lovejoy CO, Si mpson SW, White TD, Asfaw B, Suwa G. 2009. Careful climbing in the Miocene: the forelimbs of Ardipithecus ramidus and humans are primitive. Science 326(5949):70e71 78. Maki J, and Trinkaus, E. 2011. Op ponens Pollicis Mechanical Effectiveness in Neandertals and Early Modern Humans. PaleoAnthropology 2011(2011):62 71. Marzke MW. 1971. Origin of the human hand. Am J Phys Anthropol 34(1):61 84. Marzke MW. 1983. Joint Functions and Grips of the Australopith ecus afarensis Hand, with Special Reference to the Region of the Capitate. J Hum Evol 12:197 211. Marzke MW, Wullstein KL, Viegas SF. 1992. Evolution of the power ("squeeze") grip and its morphological correlates in hominids. Am J Phys Anthropol 89(3):283 298. Marzke MW. 1997. Precision grips, hand morphology, and tools. Am J Phys Anthropol 102(1):91 110. Marzke MW, Toth N, Schick K, Reece S, Steinb erg B, Hunt K, Linscheid RL, An KN. 1998. EMG study of hand muscle recruitment during hard hammer percussion manufacture of Oldowan tools. Am J Phys Anthropol 105(3):315 332. Marzke MW, Marzke RF, Linscheid RL, Smu tz P, Steinberg B, Reece S, An KN. 1999. Chimpanzee thumb muscle cros s sections, moment arms and potential torques, and comparisons with humans. Am J Phys Anthropol 110(2):163 178.

PAGE 74

68 Marzke MW, Shrewsbury MM, Horner KE. 2007. Middle phalanx skeletal morphology in the hand: can it predict flexor tendon size and attachments? A m J Phys Anthropol 134(2):141 151. Marzke MW, Tocheri MW, Steinberg B, Femiani JD, Ree ce SP, Linscheid RL, Orr CM, Marzke RF. 2010. Comparative 3D quantitative analyses of trapeziometacarpal joint surface curvatures among living catarrhines and fossil hom inins. Am J Phys Anthropol 141(1):38 51. Marzke MW. 2013. Tool making, hand morphology and fossil hominins. Philos Trans R Soc Lond B Biol Sci 368(1630):20120414. Matarazzo S. 2008. Knuckle walking signal in the manual digits of Pan and Gorilla. Am J Phy s Anthropol 135(1):27 33. McPherron SP, Alemseged, Z., Marean, C.W., Wynn, J.G., Reed, D., Geraads, D., Bobe,R., and Bearat, H. 2010. Evidence for Stone Tool Assisted Consumption of Animal Tissues before 3.39 Million Years Ago at Dikika, Ethiopia. Natur e 466:857 860. Moya Sola S, Kohler M, Alba DM, and Almecija S. 2008. Taxonomic attribution of the Olduvai hominid 7 manual remains and the functional interpretation of hand morphology in robust australopithecines. Folia Primatol (Basel) 79(4):215 250. Na pier J. 1962 a The evolution of the hand. Sci Am 207(6):56 62. Napier J. 1962 b The prehensile movements of the human hand. B one Joint J 38(4):902 913. Napier J T uttle R. H. 1993. Hands: Princeton University Press. Niewoehner WA. 2001. Behav ioral inferences from the Skhul/ Qafzeh early modern human hand remains. PNAS 98(6):2979 2984. Niewoehner WA, Bergst rom A, Eichele D, Zuroff M, Clark JT. 2003. Digital analysis: Manual dexterity in Neanderthals. Nature 422(6930):395. Niewoehner WA. 2006. Neanderthal hands in their proper perspec tive. In: Harvati KaH, T editor. Neanderthals Revisited: and Perspectives New Approaches: Springer. Orr CM, Tocheri MW, Burnett SE, Awe RD, Saptomo EW, Sutikna T, Jat miko, Wasisto S, Morwood MJ, Jungers WL. 2013. New wrist bones of Homo floresiensis from Liang Bua (Flores, Indonesia). J Hum Evol 64(2):109 129. Orr CM. 2017. Locomotor hand postures, carpal kinematics during wrist extension, and associated morphology in anthropoid p rimates. Anat R ec 300(2):382 401. Panger MA, Brooks A S, Richmond BG, Wood B. 2002 Older than the Oldowan? Rethinking the emergence o f hominin tool use. Evol Anthropol 11(6):235 245.

PAGE 75

69 Parr WC, Chatterjee HJ, Soligo C. 2011. Inter and intra specific scaling of articular surface areas in t he hominoid talus. J Anat 218(4):386 401. Patel BA. 2010. Functional morphology of cercopithecoid primate metacarpals. J Hum Evol 58(4):320 337. Patel BA. 2016. Morphological Diversity in the Digital Rays of Primate Hands. In: Kivell TL, Lemelin, P., Richmond, B.G., Schmitt, D., editor. The Evolution of the Primate Hand: Springer. p 55 100. Plavcan JM. 2002. Sexual dimorphism in primate e volution. Am J Phys Anthropol 44:25 53. R Core Team. 2016. R: A language and environment for statistical computing R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R project.org/. Rabey KN, Green DJ, Taylor AB, Begun DR, Richmond BG, McFarlin SC. 2015. Locomotor activity influences muscle architecture and bone growth but not muscle attachmen t site morphology. J Hum Evol 78:91 102. Richmond BG, Jungers WL. 2008. Orrorin tugenensis femoral morphology and the evolution of hominin bipedalism. Science 319(5870):1662 1665. Richmond BG, Roach, N.T., Ostrofsk y, K.R. 2016. Evolution of the early ho minin h and. In: Kivell TL, Lemelin, P., Richmond, B.G., Schmitt, D., editor. The Evolution of the Primate Hand: Springer. p 515 543. Ricklan DE. 1987. Functional anatomy of the hand of Australopithecus africanus J Hum Evol 16:643 664. Rogers ME, Abernethy K, Bermejo M, Cipolletta C, Doran D, McFarl and K, Nishihara T, Remis M, Tutin CE. 2004. Western gorilla diet: a synthesis from six sites. Am J Primatol 64(2):173 192. Rohlf JF, Slice, D. 1990. Extensions of the Procrustes method for the optimal superimp osition of landmarks. Syst Biol 39(1):40 59. Rolian C. 2009. Integration and Evolvability in Primate Hands and Feet. Evolutionary Biology 36(1):100 117. Rolian C, Lieber man DE, Zermeno JP 2011. Hand biomechanics during simulated stone tool use. J Hum Evol 61(1):26 41. Rolian C., G ordon A.D. 2013. Metacarpal proportions in Australopithecus africanus J Hum Evol 54(5):705 719.

PAGE 76

70 Rothman JM, Plumptre AJ Dierenfeld ES, Pell AN. 2007. Nutrit ional composition of the diet of the gorilla ( Gorilla beringei ): a comparison between two montane habit ats. J Trop Ecol 23(06):673 682. Sanz CM, Morgan, D.B. 2013. The Social Context of Chimpanzee Tool Use. In: Sanz CM, Call, J., Boesch, C., editor. Tool use in animals: cognition and ecology: Cambridge University Press. p 161 175. Schlecht SH. 2012. Understanding entheses: bridging the gap between clinical and anthropological perspectives. Anat Rec 295(8):1239 1251. Schroeder L, Scott JE, Garvin HM, Lair d MF, Dembo M, Radovcic D, Berger LR, de Ruiter DJ, Ackermann RR. 2017. Skull diversity in the Homo lineage and the relative position of Homo naledi. J Hum Evol 104:124 135. Scott JE. 2015. The phylogenetic utility of mentum osseum morphology in Pleistoce ne Homo. Am J Phys Anthropol 156:283. Semaw S, Rogers MJ, Quade J, Renne PR, Butler RF, Dominguez Rodrigo M, St out D, Hart WS, Pickering T, Simpson SW. 2003. 2.6 Million year old stone tools and associated bones from OGS 6 and OGS 7, Gona, Afar, Ethiopia. J Hum Evol 45(2):169 177. Senut B, Pickford, M., Gommery, D., Mein, P., Cheboi, K., Coppens, Y. 2001. First hominid from the Miocene (Luke ino Formation, Kenya). Earth Planet Sc 332:137 144. Siegel M, Pernotto, B. 1975. Hand Use and Metacarpal Robusticit y in Catarrhini. Primates 16:371 377. Susman RL, Creel N. 1979. Functional and morphological affinities of the subadult hand (O.H. 7) from Olduvai Gorge. Am J Phys Anthropol 51(3):311 332. Susman RL. 1988. Hand of Paranthropus robustus from Member 1, Swartkrans: Fossil Evidence for Tool Behavior. Science 240(4853):781 784. Susman RL. 1994. Fossil Evidence for Early Hominid Tool Use. Science 265(5178):1570 1573. Susman RL. 1998. Hand function and tool behavior in early hominid s. J Hum Evol 35(1):23 46. Susman RL Nyati L, Jassal MS. 1999. Observations on the pollical palmar interosseous muscle (of Henle). Anat Rec 254(2):159 165. Sutikna T, Tocheri MW, Morwood MJ, Saptomo EW, Jatmiko, Awe RD, Wasisto S, West away KE, Aubert M, Li B et al. 2016. Revised stratigraphy and chronology for Homo floresiensis at Liang Bua in Indonesia. Nature 532(7599):366 369.

PAGE 77

71 Tocheri MW, Orr CM, Larson SG, Sutikna T, Jatmiko, Saptomo EW, Due RA Djubiantono T, Morwood MJ, Jungers WL. 2007. The primi tive wrist of Homo floresiensis and its implications for hominin evolution. Science 317(5845):1743 1745. Toch eri MW, Orr CM, Jacofsky MC, Marzke MW. 2008. The evolutionary history of the hominin hand since the last common ancestor of Pan and Homo J Anat 212(4):544 562. Trinkaus E. 1983. The Shanidar Neanderthals. New York: Academic Press. Tuttle RH. 1967. Knuckle walking and the Evolution of Hominoid Hands. Am J Phys Anthropol 26(2):171 296. Tuttle RH. 1969 a Knuckle walking and the Problem of Human Or igins Science 166(3908):953 961. Tuttle RH. 1969b. Quantitative and functional studies on the hands of the Anthropoidea. I. The Hominoidea. J Morphol 128(3):309 363. Van Horn R. 1972. S tructural adaptations to climb ing in the gibbon hand. Am Anthropol 74(3):326 334. Vleck, E 1973. Postcranial Skeleton of a Neandertal Child from Kiik Koba, U.S.S.R. J Hum Evol 2:537 544. Vl eck E. 1975. Morphology of the first metacarpal of neandertal individuals from the Crimea. B Mem Soc Anhro P ar 2(3):257 276. Wallace IJ, Wi nchester JM, Su A, Boyer DM, Konow N. 2017. Physical activity alters limb bone structure but not entheseal morphology. J Hum Evol 107:14 18. Ward CV, Kimbel, W. H., Harmon, E.H., Johanson, D.C. 2012. New postcranial fossils of Australopithec us afarensis from Hadar, Ethiopia (1990e2007). J Hum Evol 63:1 51. Ward CV, Toche ri MW, Plavcan JM, Brown FH, Manthi FK. 2014. Early Pleistocene third metacarpal from Kenya and the evolution of modern human like hand morphology. Proc Natl Acad Sci U S A 1 11(1):121 124. White TD, Suwa G, Asfaw B. 1994. Australopithecus ramidus a new species of early hominid from Aramis, Ethiopia. Nature 371(6495):306 312. Wiley DF, Amenta, N, Alcantara D.A., Ghosh D., Kil Y.J., Delson E., Harcourt Smith W., Rohlf F.J., St. John K., Hamann B. 2005. Evolutionary morphing. Proc IEEE Visual. Wiley DF. 2007. Landmark Editor. University of California, Davis: Institute for Data Analysis and Visualization.

PAGE 78

72 Williams EM, Gordon AD, Richmond BG 2012. Hand pressure distribution during Oldowan stone tool production. J Hum Evol 62(4):520 532. Williams Ha tala EM, Hatala KG, Hiles S, Rabey KN. 2016. Morphology of muscle attachment sites in the modern human hand does not reflect muscle architecture. Sci Rep 6:28353. Zumwalt A. 2006. The effect of endurance exercise on the morphology of muscle attachment sites. J Exp Biol 209(Pt 3):444 454.