Bipedalism is a locomotory mode whereby an organism moves within an environment on two legs. Humans are unique amongst the extant primates in being obligate bipeds [Harcourt-Smith & Aiello 2004] but are postulated to have evolved from an arboreal, quadrupedal, knuckle-walking ancestor [Richmond et al. 2001; Schmitt 2003; Sockol et al. 2007; Thorpe et al. 2007]. The transition within the human lineage towards bipedality has been accompanied by the accumulation of a multitude of adaptive musculoskeletal traits. The majority of these traits are post-cranial in that they manifest inferior to the head in a fully erect posture. This suite of adaptations includes direct evolutionary modifications to the weight-bearing structures: the vertebral column, the pelvis, the lower limb and the foot. Other indirect modifications for example in the upper limb, present as a result of our abandoning of arboreal climbing, will also be discussed. An examination of these characters exhibited in modern humans will be compared and contrasted with corresponding character states in extant and fossil apes. Evidence will come from morphological and developmental studies of modern humans, Great Apes, fossil hominins such as Ardipithecus and Australopithecus & the remarkably informative fossil apes Dryopithecus and Oreopithecus.
Given the effects of gravity, the maintenance of bipedal posture requires an effective transmission of the upper body weight from the trunk, through the pelvic girdle to the lower limbs. In humans the centre of mass (COM) lies centrally and directly vertical above the hips, thus the upper body weight is absorbed at the foot of the vertebral column. Therefore we expect to see weight-bearing adaptations in this region of the back. One such adaptation is known as a lumbar lordosis, a posterior concavity of the vertebral column in the lumbar region. Dorsal wedging of the L1-L5 vertebrae creates a curvature in the lower spine that, much like a spring, helps absorb body weight and allows flexion & extension of the trunk. The angle created is typically 30° but can reach 50° in a pregnant female [Whitcome et al. 2007]. Lovejoy  describes this feature as ‘uniquely human’ however analyses of the Upper Miocene hominoid Oreopithecus bambolii and Australopithecus afarensis suggest that they too possessed lordoses indicative of bipedal posture [Köhler & Moyà-Solà 1997; Sanders, 1998]. Extant Great Ape vertebral columns lack lumbar lordoses and are therefore restricted in their lower back mobility (Figure 1). Further, the lumbosacral angle in humans averages 60°, almost double that seen in the Great Apes [Aiello & Dean, 1990]. Again this illustrates how lower back flexibility is essential for cushioning the impacts of walking or running.
Figure 1. Blue highlighted lumbar regions of a modern human (left) and chimpanzee (right). The lordotic human curvature allows lower back flexibility in contrast with the ape’s rigid condition. Arrows indicate centre of mass/gravitational direction. Source: Adapted from Whitcome et al, 2007.
Other weight-loading adaptations in the lower spine include the lengthening and thickening of the lumbar region. Humans usually possess 5 lumbar vertebrae with progressively widening zygapophyses [Lovejoy 2005]. This contrasts with chimpanzees who have between 3 and 4 lumbar segments with narrowing facets. In addition, continuing caudally along the column, human vertebral bodies become larger in overall size [Hauesler et al. 2002].
The human pelvis has undergone significant remodelling in order to maximise the efficiency of bipedal locomotion. Compared to the chimpanzee pelvis, the human version is considerably shorter and broader (Figure 2). This has had the effect of distancing the thorax from the ilium and freeing the lumbar region from within the pelvis [Lovejoy 2005]. Consequently the human lower back is capable of a greater range of movement than a Great Ape’s. This alteration has also brought into close proximity the sacroiliac and hip joints, reducing force transmission stress on the ilium [Aiello & Dean 1990].
Joint surface sizes are skeletal indicators of bipedalism because they are proportional to the quantity of force passing through them [Aiello & Dean, 1990]. It follows that the lower limb joints of a human are larger relative to body size than other apes because of the body weight they perpetually support in bipedal stance [Jungers, 1988]. The auricular surface of a human hip for example is comparable to a gorilla’s of twice the body weight [Schultz, 1961].
Figure 2. Anterior pelvic view of a) chimpanzee, b) Australopithecus afarensis Al-288-1, c) Homo sapiens female and d) Homo sapiens male. Note the tall and narrow chimpanzee ilium in contrast to the short and broad human condition. The Australopithecine pelvis showcases a transitional morphology with extreme lateral iliac flare. The broadened human sacrum frees the lumbar region which is sunken rigid within the chimpanzee hip. Source: Lovejoy, 2005.
Other modifications to the human pelvis have repositioned muscle attachment sites relative to those seen in the apes. Humans have an expanded, anteriorly-projecting iliac blade with respect to the ischium unlike the chimpanzee’s whose is more aligned. The resulting hip structure facilitates a reorganisation of gluteal musculature without compromising the hamstring lever arm [Sockol et al 2007]. Whereas in apes the gluteals function as extensors of the thigh at the hip joint, in humans they primarily assist in abduction thanks to their more anterior origins [Aiello & Dean 1990].
A broader human sacrum and an expanded posterior superior iliac spine greatly improve the surface attachment and lever advantages of the posture-maintaining erector spinae muscle group. Significantly, the prominent human anterior inferior iliac spine (which provides attachment for the knee-extending rectus femoris muscle and the trunk-balancing iliofemoral ligament) is absent or reduced in apes [Aiello & Dean 1990].
A shift towards the human pelvic condition is documented in fossils of A. afarensis such as Al 288-1, better known as ‘Lucy’. Superiorly Lucy’s hip showcases adaptations to a bipedal gait including a broad sacrum, a widened interacetabular distance and pronounced lateral iliac flare [Berge 1994]. Unlike in humans though, Lucy’s hip was subject to large rotatory movements as her gluteals did not function as stabilisers of the joint [Ward, 2002]. The human-like pattern of iliac cancellous bone in O. bambolii indicates that it too was probably a biped [Rook et al, 1999].
The femoral bicondylar angle is a classic feature that distinguishes bipedal humans from quadrupedal apes. In humans this angle averages 8-11° compared to 1-2° in the African apes. A larger bicondylar angle aligns the distal end of the femur, knee and lower leg with the body’s midline thus minimising the distance the body’s COM must be displaced in walking to lie directly over the stance leg [Aiello & Dean, 1990]. Interestingly, the biped A. afarensis exhibits a larger 12-15° angle perhaps as a consequence of its exceptionally broad pelvis (Figure 3).
Figure 3. Femoral bicondylar angles in a) modern humans b) Australopithecus and c) chimpanzees. Source: Shefelbine et al, 2002.
Another noticeable difference between humans and great apes is found in limb proportions. Humans compared to apes have significantly longer lower limbs relative to upper limbs (Figure 4). Ruff  finds that although human infants exhibit a characteristically adult femoral: humeral length ratio, their femoral: humeral shaft strength only increases following the adoption of bipedalism at the age of 1, hinting at a developmental response to stress. Tardieu  illustrates how the tibio-femoral angle reverses in a juvenile from markedly abducted in a newborn to 6° adducted in a 6-7 year old.
Figure 4. Femoral to humeral length in 100 modern humans (open circles) and 100 non-human Catarrhines (filled squares). Humans consistently show greater femoral lengths and overlap only with leaping colobine monkeys. Source: Ruff, 2003.
Other osteological aspects of the femur betray bipedal locomotion. Some help to diagnose bipedalism in fossil hominins, for example the 6 million year-old proximal femur of Orrorin tugenensis [Richmond et al 2008]. Human features include a marked obturator externus groove indicating full hip extension of a habitual biped. An enlarged femoral head, a broader neck, a marked intertrochanteric line, a prominent gluteal tuberosity and a raised linea aspera are all evidence of greater weight-bearing abilities and enhanced lever arm attachments for muscles & ligaments in humans that help maintain upright posture. Such features are lacking or reduced in apes.
The enlarged, elliptical profile of a human’s distal femoral condyles (when viewed inferiorly) compared to the chimpanzee’s maximises tibial contact in extension of the knee joint, effectively distributing the weight load. Due to the differential length of the articular femoral condylar surfaces, human knees are capable of assuming a close-packed position in full extension whereby the femur rotates medially on the tibia. The structure of the cruciate ligaments and menisci within the chimpanzee knee do not permit this [Aiello & Dean, 1990]. Hence a human can stand stationary upright for long periods whereas a chimpanzee or bonobo must exhibit an unspecialised, bent hip-bent knee (BHBK) gait when walking bipedally [D’Août et al 2004]. Horizontally-angled tibial plateaus (when viewed medially) are characteristic of fully-extended modern human knees whereas continually flexed Great Ape knees tend to produce reclined angles in excess of 110° to the vertical.
Ankle & Foot
The plane of the tibiotalar (ankle) joint in humans is near perpendicular to the long axis of the tibia. This helps position the leg vertically to the foot which is advantageous to a biped as it provides stability in walking. An ape’s ankle joint is considerably more flexible and is characterised by a 20-35° tibiotalar angle [Aiello & Dean 1990].
A human foot exhibits multiple osteological features which are beneficial to bipedalism compared to the ape’s which is adapted for grasping (Table 1). The nature of the human calcaneus reflects its function in absorbing the force of heel strike in walking. The human metatarsal (MTS) robusticity pattern provides evidence of lateral to medial weight transfer across the foot. Strikingly the human hallux has been laterally rotated and adducted in humans to realign it with the other digits. This, combined with a greater robusticity, provides a sturdier platform and increases the power of toe push-off achieved by the strong flexor hallucis longus muscle. In apes the hallux is widely abducted and mobile due to a convex, medially facing MTS1 facet on the medial cuneiform. In addition, curved phalanges equipped with broad bases and flexor sheath ridges reflect the function of the ape’s foot in grasping.
Human feet exhibit a unique longitudinal arch structure supported by (plantar and spring) ligaments & a well developed plantar aponeurosis. The thick aponeurosis absorbs 60% of the stress on the foot, cushioning it on ground contact. Enhanced plantar ligament attachments sites in humans provide greater stability to the foot which acts as a lever. The shared presence of an os peroneum facet (sesamoid bone of the peroneus longus muscle) on the cuboid of humans and Ardipithecus ramidus suggests that the p. longus tendon supports the longitudinal arch [Lovejoy et al, 2009]. Its general absence in apes is a pointer to greater foot flexibility. Wedged cuneiforms contribute to the transverse arch in humans and apes however only in the former do the metatarsal heads contact the substrate.
The functional interpretations of the famous Stw573 “Little Foot” Australopithecus africanus and OH8 Homo habilis feet are strongly debated due to their mosaic of human-ape features [Clarke & Tobias 1995; Kidd & Oxnard, 2005]. In contrast to humans, the fossil ape O. bambolii exhibits a more medial lever axis and abducted lateral metatarsals. However a near vertical calcaneal tuberosity and tall medial trochlea, unlike in Dryopithecus and extant apes, suggest a parallel tibial axis and COM as in humans [Köhler & Moyà-Solà 1997].
One cranial trait used to diagnose bipedality in fossil hominids is the position of the foramen magnum at the base of the skull. Humans possess an anterior location of the foramen magnum in contrast to a chimpanzee’s more posterior position relative to the bitympanic line. The human condition when viewed in norma basilaris helps position the vertebral column and body weight directly under the cranium. Brunet et al  used this indicator to assign bipedality to the 6-7 million year old Sahelanthropus tchadensis type specimen, though their conclusions are hotly debated [Wolpoff et al. 2002].
Thorax and upper limb
Humans in contrast to the African apes have a barrel-shaped thorax containing 12 pairs of medio-laterally flattened ribs. This serves to centralise the axis of the body’s COM whilst the lack of a 13th rib pair increases lower trunk mobility. Arboreality in apes is associated with a more superiorly-oriented glenoid fossa. This permits a range of upper limb movements above the head for example suspension in the fossil ape Dryopithecus [Köhler & Moyà-Solà 1996]. In contrast, humans exhibit a more laterally-facing glenoid fossa; a derived trait assuming they had an arboreal last common ancestor (LCA). The human condition reflects how bipedalism has freed our upper limbs for use in an anterior position to manipulate objects.
In summary the evolution of habitual bipedalism in humans has been accompanied by numerous musculoskeletal modifications. Structures such as the pelvis, lower limb and foot have been remodelled to improve their weight-bearing abilities. Such remodelling has facilitated subtle alterations in the attachment sites and functioning of key postural muscles like the erector spinae group and important walking muscles like the gluteals. The hip, knee and ankle joints also document osteological and anatomical changes which betray a human’s obligatory locomotor mode. In contrast, extant apes such as the chimpanzee, with which we share a LCA, exhibit tell-tale adaptations to an arboreal lifestyle in their corresponding character states. Some traits shared between humans and fossil hominids like O. tugenensis & A. afarensis could reflect shared ancestry. Other traits present in humans and fossil apes such as O. bambolii probably reflect homoplasy or convergent evolution.