The Thoracic Cage of KSD-VP-1/1

  • Bruce M. Latimer
  • C. Owen Lovejoy
  • Linda Spurlock
  • Yohannes Haile-Selassie
Part of the Vertebrate Paleobiology and Paleoanthropology book series (VERT)


Ribs are naturally fragile and, as a consequence, are rarely preserved in the fossil record. The costal elements recovered from the KSD-VP-1/1 partial skeleton are important evidence allowing reconstruction of the hominin thoracic cage. The ribs of KSD-VP-1/1 are examined with respect to their implications for the evolution of Australopithecusafarensis thoracic morphology. Angulation and torsion along the rib corpus and rib declination indicate a broad upper thorax and a deeply invaginated thoracic vertebral column. Implications for the early hominin thoracic bauplan are discussed.


Australopithecus Thorax Ribs 


The discovery of the 3.6-million-year old Australopithecus afarensis (KSD-VP-1/1) partial skeleton (Haile-Selassie et al. 2010a, b) from Woranso-Mille, Ethiopia provides rare evidence regarding the evolutionary development of the human and great ape thoracic cages. Differences in thoracic skeletal anatomies between modern African apes and humans have been noted earlier (Keith 1923; Schultz 1961; Latimer and Ward 1993; Jellema et al. 1993; Lovejoy 2005; Gomez-Olivencia et al. 2009; Ward et al. 2012; Bastir et al. 2013). However, owing to their natural fragility, there has been a lack of sufficient fossil rib material that could allow reconstruction of the chronology of these changes. Indeed, debate continues regarding the anatomy of the primitive hominin thoracic shape with some researchers suggesting that the chimpanzee/human last common ancestor (CLCA) had an African ape-like, inverted, “funnel-shaped” thorax (Schmid 1991; Schmid et al. 2013), while others reconstruct the CLCA as having a human-like “barrel-shaped” thorax (Lovejoy 2005; Haile-Selassie et al. 2010b). This simplistic dichotomy between “funnel-shaped” and “barrel-shaped” may too greatly simplify a more complex problem, a possibility that is discussed below.

The six ribs included in the KSD-VP-1/1 partial skeleton (Fig. 7.1), even while fragmentary, shed light on this issue, and allow the reconstruction of the thoracic shape in Au.afarensis. Because the features that characteristically distinguish hominin ribs are functionally interrelated (declination, torsion, and neck/shaft angle), even fragmentary ribs can be highly informative allowing reliable discrimination between hominoid thoracic shapes. They also provide evidence regarding the associated functions of the shoulder, the pelvis, and the lumbar and thoracic spines. As many of the anatomical features that we discuss below are unique to humans (and our bipedal ancestors), we also, where appropriate, include brief mention of certain pathological conditions that are similarly confined to humans. These conditions shed light on the selective pressures encountered by early hominins and are the direct result of our habitually upright posture.
Fig. 7.1

Cranial views of the ribs of KSD-VP-1/1. a KSD-VP-1/1n, left second rib, reversed in image; b KSD-VP-1/1q, R. fifth, sixth or seventh rib fragment; c KSD-VP-1/1s, mid-thoracic rib fragment; d KSD-VP-1/1o, R. seventh or eighth rib fragment; e KSD-VP-1/1p, R. eighth or ninth rib fragment; f KSD-VP-1/1r, L. eleventh rib

Anatomical Descriptions

The following descriptions are intended for use by researchers utilizing cast replicas of the original specimens. Real anatomy is often difficult to discern from casts particularly if there is significant postmortem damage, reconstruction, and/or supplementation of the original fossil prior to casting. Adherent matrix can also cause difficulties when using replicas. Following this, it should be noted that postmortem damage has removed many of the traditional anatomical landmarks and that caution should be practiced when reviewing the measurements presented here.

In addition, confusion can also result when using qualitative and relative comparative terminology (Franciscus and Churchill 2002) such as “highly curved,” “roughened,” or “robust.” In the following descriptions, the reader should be aware that such qualitative terminology refers to ranges of variability normally encountered in human anatomy. When used, comparisons with other taxa will be explicitly noted in an attempt to reduce potential ambiguity. All measurements are presented in Table 7.1.
Table 7.1

Linear measurements of the preserved KSD-VP-1/1 ribs








Maximum preserved chord length (MCL)







Maximum perpendicular from MCL to pleural margin






Transverse corpus dimension, estimated original mid-shaft – transverse






Craniocaudal corpus dimension, estimated original mid-shaft – cranial–caudal






Center of articular head to the center of tubercle


All measurements are in millimeters (mm)

KSD-VP-1/1n(Figs. 7.1a,7.2, and7.3)

KSD-VP-1/1n is a fragmentary left second rib. Approximately two-thirds of the corpus is present from the head to the most ventral point, which is obliquely fractured such that the ventral third of the specimen is not present. Several postmortem transverse fractures occur along the body, most of which oppose well and do not alter the original anatomy. A large triangular fragment is absent in the vicinity of the posterior angle, the size of which required reinforcement of the rib body (see Figs. 7.1a, 7.2, and 7.3). This loss of bone has resulted in damage to the tubercle and required reconstruction that may have slightly altered the rib’s original anatomy in the direction of increasing the curvature of the pleural margin (decreasing the radius of curvature). As a consequence, estimates of the original curvature and copular volume may be somewhat underestimated. Nevertheless, its curvature radius falls well within the human range and substantially below those of the African apes (see below and Haile-Selassie et al. 2010b). Despite the postmortem damage, it is nevertheless apparent that the curvature of the pleural margin indicates a human-like neck-corpus angle and the deep invagination of the vertebral column into the thorax. Several flakes of bone are missing from the cranial and caudal surfaces of the corpus but these do not alter overall anatomy. The crest for M. scalene posterior, although somewhat eroded, is visible along the dorsal margin of the corpus. There is a roughened tuberosity for the attachment of M. serratus anterior. The caudal surface is smooth and the costal groove, which begins immediately ventral to the tubercle, is shallow. The rib corpus is typically flat ventral to the tubercle.
Fig. 7.2

CT-based views of KSD-VP-1/1 ribs (KSD-VP-1/1n, b, q). Postmortem damage is evident but the original curvature of the pleural margin remains largely undistorted. Dashed lines indicate areas of cross-sections. Figure from Ryan and Sukhdeo (2016)

Fig. 7.3

KSD-VP-1/1n, second left rib. Cranial and caudal views, right and left, respectively

KSD-VP-1/1q (Fig. 7.1b)

KSD-VP-1/1q is a fragmentary right fifth, sixth, or seventh rib. Postmortem damage does not allow for a more accurate numerical designation. The head and neck are eroded and no original subchondral surfaces remain. A triangular flake of bone is lost associated with a well-opposed transverse fracture across the costal angle. It is nevertheless apparent from the remaining portions that the costal angle is flexed indicating a deep invagination of the mid-thoracic vertebral column. In addition, despite the significant surface flaking and cortical exfoliation, the original curvature of the rib is apparent especially along the pleural margin. If oriented so that the undamaged pleural surface of the head and neck is in a vertical plane (see Jellema et al. 1993), it is clear that the remaining ventral portion of the rib corpus declines inferiorly and sigmoidally twists (caudal margin inferomedially, cranial edge superolaterally) along its long axis.

KSD-VP-1/1s (Fig. 7.1c)

KSD-VP-1/1s is a badly damaged middle segment of a mid-thoracic rib. Judging from its size and curvature, it is likely from a sixth, seventh, or eighth rib. Much of the original cortical surface is exfoliated, and no original subchondral surfaces remain intact.

KSD-VP-1/1o (Fig. 7.1d)

KSD-VP-1/1o is a fragment from the dorsal half of a right seventh or eighth rib. Much of the original cortical surface is damaged particularly around the head and neck, and no subchondral surfaces remain. Adhering matrix remains in several areas producing a roughened surface that should not be interpreted as anatomical. Despite the significant postmortem damage, the overall curvature and shaft dimensions are interpretable. When the pleural surface of the head and neck are oriented vertically (see Jellema et al. 1993), the corpus of the rib demonstrably declines and twists about its long axis. The costal neck axis is strongly flexed indicating posteriorly directed transverse vertebral processes and an invaginated thoracic vertebral column. The ventral third of the shaft is flattened and has a sharp inferior margin. Although erosion has damaged much of the pleural surface, a shallow costal groove is apparent.

KSD-VP-1/1p (Fig. 7.1e)

KSD-VP-1/1p is the dorsal half of a right eighth or ninth rib. Several transverse cracks occur along the shaft but these oppose well and do not distort the original curvature or torsion. The surface is eroded over much of the specimen although small areas of subchondral bone remain on the head and neck. The posterior angle is rugose and ventral to it the corpus flattens and presents a sharp caudal margin. The costal groove begins at the posterior angle and is deep. The shaft declines markedly such that when the cranial surface of the specimen is placed on a horizontal plane the ventral half of the corpus does not touch the surface. Associated with this declination is the obvious sigmoidal torsion along the rib corpus.

KSD-VP-1/1r (Fig. 7.1f)

KSD-VP-1/1r is the dorsal half of a left eleventh rib. Much of the surface is eroded, and several transverse cracks occur along the corpus but these oppose well and do not distort the original curvature and torsion. The surface around the head and neck is damaged and no original subchondral bone remains. The cranial margin of the shaft is sharp particularly along the ventral third of the fragment. A costal groove is apparent extending from the posterior angle ventrally. When the specimen is placed on a horizontal planar surface with its cranial edge facing down, the ventral half of the fragment does not touch the surface, an exercise that demonstrates rib declination. Associated with this declination is the sigmoidal torsion along the corpus.

Thoracic Shape and Function

In order to provide an appropriate anatomical context within which to view the KSD-VP-1/1 thoracic skeleton, it is necessary to compare and contrast human and African ape thoraces (see Fig. 7.4). Below we describe the anatomy and function of the thoraces of modern human and African apes and the implications for KSD-VP-1/1 and Australopithecus in general.
Fig. 7.4

A comparison of the thoracic cages and the pelvic girdles in a chimpanzee (left), KSD-VP-1/1 (center), and a human (right). The “barrel-shaped” human rib cage is transversely broad in its upper portion and narrowed at its bottom to conform to the sagittalized ilia. In contrast, the ape rib cage is narrowly constricted in its cranial segment and flares in its inferior segments to conform to the transversely broad false pelvis – forming the inverted “funnel-shaped” thorax. The KSD-VP-1/1 thorax is broad in its upper portion like the human but flares somewhat in its inferior portion to conform to the typical Australopithecus platypelloid pelvis – thus forming a “bell-shaped” thorax. Also note the long and flexible lumbar columns in the hominins in direct contrast to the ape’s short and inflexible lumbar column

Modern Human

The human thoracic bauplan includes several linked, diagnostic features in the rib cage, which allow it to be reliably distinguished from that of the African apes. These include a relatively broad cupola or superior thoracic area, a ventrally invaginated vertebral column, and the declination and torsion of the rib bodies (Schultz 1961; Latimer and Ward 1993; Jellema et al. 1993; Franciscus and Churchill 2002; Haile-Selassie 2010b; Ward et al. 2012).

As noted, modern humans are distinguished from the African apes by having a “barrel-shaped” thorax in contrast to what is described as an inverted “funnel-shaped” thorax in chimpanzees and gorillas (Schultz 1961; Latimer and Ward 1993; Jellema et al. 1993). The “barrel-shaped” human condition is largely a product of the relatively broad upper thorax juxtaposed to the constricted and transversely narrow lower rib cage (see Fig. 7.4; Latimer and Ward 1993 and discussion below). The tapering or “waisting” of the caudal half of the human thorax is a reflection of the marked changes in the pelvis and iliocostal space of hominins and is the consequence of anatomical modifications required for habitual bipedality (Lovejoy 1974; Latimer and Ward 1993; Jellema et al. 1993; Lovejoy 2005; Lovejoy et al. 2016). The modifications of the hominin false pelvis are the result of the “sagittalization” of the iliac blades to provide a functioning abductor complex allowing stabilization of the torso during single limb stance.

In addition to the dramatic, easily recognized changes in the human pelvis are features associated with the long, flexible human lumbar spine. These vertebral characteristics are a necessary requirement for habitual bipedality and likely represent one of the earliest adaptations to this peculiar form of locomotion. As noted by Lovejoy and McCollum (2010), the early availability of at least a partial lumbar lordosis negated any requirement for the earliest bipeds to have ever utilized a, bent-hip-bent-knee (BHBK) gait. The unique series of alternating curvatures in the human spine enable the superjacent torso and head to balance over the supporting limbs, thus not requiring a BHBK gait. This contrasts markedly to that seen in the African apes, wherein the abbreviated lumbar spine is “entrapped” between the elongated iliac blades and is virtually immobile. The resulting inability to lordose the lumbar spine leads to the BHBK posture during bouts of facultative bipedality as this is the only mechanism available to allow balancing over the single supporting limb.

Because several of the features characterizing human ribs are related directly to the long, flexible thoracolumbar column, their ontogenetic development warrants some discussion. The development of the human thoracic cage is basically biphasic with the first major phase taking place during infancy and the second occurring during adolescence. During the first two years of life, human infants have a pyramidal-shaped thorax exhibiting a circular cross section with little or no rib declination and no rib torsion (Keith 1923; Bastir et al. 2013; Openshaw et al. 1984). As a consequence of this orientation of the ribs, the breathing of infants is largely diaphragmatic. Importantly, during this early period, infants also have nearly straight thoracolumbar vertebral columns with minimal spinal curvatures. Upon beginning to walk, children start to develop an incipient lumbar lordosis and a structural thoracic kyphosis. These two curves develop simultaneously as a balancing mechanism. During this early period, the developing thoracic kyphosis is relatively deeper than is the comparatively undeveloped lumbar lordosis, a condition that results in the somewhat rounded back profile characteristic of young children. The developing thoracic kyphosis causes the growing ribs to begin to decline such that their anterior ends are positioned caudally relative to their vertebral attachments.

Rib declination in young children is largely the result of the strong anterior wedging in the upper thoracic vertebrae (Latimer and Ward 1993) with further rib growth enhancing declination. It is also during this early period that the spinal column begins to invaginate into the thorax (Bastir et al. 2013) as part of the general hominoid adaptation that places the scapula dorsally for improved humeral circumduction. In humans, the invagination of the thoracic vertebral column is also a way to improve balance and improve leverage for the epaxial muscles. This anterior migration of the thoracic vertebral column also positions the column closer to the upper body’s centroidal axis thereby reducing potentially damaging eccentric loading of the vertebral bodies and adjacent intervertebral discs (Latimer and Ward 1993; Ward et al. 2012). Prior to this invagination process, the most dorsal structures in an infant’s back are the spinous processes of the thoracic vertebrae. In contrast, in older children and adults, the most posterior structures are the costal angles, confirming this anterior migration of the vertebral column into the thoracic cage.

The final adult configuration of the human thoracic cage occurs during adolescence, wherein further development of the lumbar lordosis and associated thoracic kyphosis cause the ribs to increasingly decline and twist along their corpora. Additional growth along the rib’s anterior terminus results in additional rib declination and torsion. Another transformation that takes place during this period is the increased broadening of the upper thoracic cage (ribs 2–5), ultimately resulting in the adult human transversely broad upper torso (Jellema et al. 1993; Bastir et al. 2013).

During adolescence, several additional secondary structural modifications of the thorax take place leading to the eventual achievement of the adult thoracic shape. The lumbar lordosis increases, especially in females (Latimer and Ward 1993; Masharawi et al. 2010). These changes are, among extant hominoids, unique to humans, and as noted above, allow balancing of the upper body in a sagittal plane during bipedal walking and running. The transversely broad rib cage also increases the ability to control rotation and balance around the long, highly flexible lumbar region. This latter feature, the flexible lumbar region, is entirely lacking in extant African apes (Lovejoy 2005). In order to achieve the hominin lordotic curve, the articular facets of the lumbar vertebrae must become progressively more separated moving caudally down the column. This increasing interfacet distance permits hyperlordosis of the lumbar column and prevents impingement by the intervertebral facets of the intervening laminar region known as the pars interarticularis. Inadequate separation between the lumbar intervertebral facets can result in the condition known as spondylolysis and its pathological sequela spondylolisthesis, two unfortunate conditions confined solely to humans (Latimer and Ward 1993).

Ontogenetic changes in the thoracolumbar vertebrae also include differential craniocaudal growth along the anterior vertebral margins resulting in an increased lumbar lordosis (negative wedging) and a slightly decreased thoracic kyphosis (reduced anterior wedging) relative to the earlier condition wherein there is a somewhat greater thoracic kyphosis linked to a minimal lordosis (Scoles et al. 1991; Latimer and Ward 1993). Failure of this vertebral growth pattern can result in a continuation of the adolescent spinal curvatures and vertebral wedging patterns resulting in the condition described as Scheuermann’s or adolescent kyphosis (Scoles et al. 1991). It should be noted that the condition described as Scheuermann’s kyphosis is largely confined to humans and that the excessive anterior wedging associated with the condition usually occurs at the depth of the developing thoracic kyphosis – T7, T8, and T9 – the vertebral elements under the greatest compressive stress as a consequence of their residing within the depth of the anteriorly directed thoracic concavity (Scoles et al. 1991; Christiansen and Bouxsein 2010; Cotter et al. 2011). Another uniquely human consequence of this positioning within the thoracic kyphosis is the observation that the seventh and eighth thoracic vertebrae are also the most commonly fractured vertebral elements (wedge fractures) usually owing to the bone mineral loss that accompanies human senescence (Christiansen and Bouxsein 2010; Cotter et al. 2011). These vertebral fractures can result in the initiation of the “vertebral fracture cascade” (Christiansen and Bouxsein 2010). Such fractures, while relatively common among humans, do not occur in ape vertebrae (Cotter et al. 2011). The inflexible and stiff thoracolumbar spine described in the African apes as well as their higher bone mineral density and accompanying greater compressive strength of the vertebral bodies (Cotter et al. 2011) provides a protective, stress shielding mechanism mitigating against vertebral fractures. It is interesting that several Australopithecus specimens for which there are relevant vertebral elements (A.L. 288-1; Sts 14; StW 431) also display a spinal condition similar in location and morphology to that seen in humans with Scheuermann’s kyphosis (Cook et al. 1983; Scoles et al. 1991; Ward et al. 2012). This particular condition is not seen in the African apes and is related to the unique loading trajectories imposed upon the flexible and sinusoidally curved vertebral column of an habitual biped (Scoles et al. 1991; Ward et al. 2012).

Also included in the normal development of the human spine, are the changes in the secondary ossification centers for the transverse processes of the thoracic vertebrae (Latimer and Ward 1993; Jellema et al. 1993). These modifications alter the angular orientation of the transverse processes relative to the vertebral centra and in so doing, change the associated rib costal axes resulting in further declination and torsion of the ribs. These ontogenetic alterations are unique to Homo and Australopithecus (Latimer and Ward 1993; Jellema et al. 1993; Ward et al. 2012) and can have profound effects upon the overall shape of the thoracic cage (decreased anteroposterior dimension) and the mechanics of breathing (see below).

Associated with these alterations in rib orientation is the so-called “descent” of the manubrium, sternum, and clavicles (Todd 1912; Keith 1923; Ohman 1986; Basir et al. 2013). These ontogenetic alterations further reduce the relative anteroposterior dimensions of the thorax. This “descent of the shoulder” (Todd 1912) is a consequence of the enhanced lumbar lordosis and the associated increasing declination of the underlying rib cage. It is likely that the single vertebral facet on the first rib among hominins (Ohman 1986; Schmid et al. 2013) and the anterior attachment of M. deltoideus on the hominin clavicle (Ohman 1986; Ward et al. 2012) are also related to this process. Indeed, essentially all described Australopithecus clavicles (A.L. 333x-6/9, A.L. 438-1, A.L. 333-94, StW 431, StW 582, and KSD-VP-1/1) also show this hominin modification in muscular attachment suggesting a horizontally oriented clavicle, a descended shoulder and a Homo-like pectoral girdle (Ward et al. 2012).

It should be noted that in human females several associated features are exaggerated relative to males. For instance, among females, the manubrium and sternum are situated somewhat lower along the thoracic column (Keith 1923; Aiello and Dean 1990; Bastir et al. 2013). They also have greater declination and twisting to their ribs (Bellemare et al. 2003, 2006), as well as a greater lumbar lordosis (Masharawi et al. 2010). These differences between men and women occur during puberty and are driven by the increased lumbar lordosis in females and the subsequent changes in the thoracic spine and torso. It is, moreover, likely that these developmental processes are pathologically involved in the development of adolescent idiopathic scoliosis, a condition more common in females and one that is known only in humans and no other primate.

At this point, some discussion of the functional reasons for these described evolutionary modifications in the human vertebral column, shoulder, and thoracic cage is necessary. That is, why do humans alone descend their shoulders and decline and twist their ribs in a manner so unlike the African apes? A potential explanation resides in the early and dramatic modifications seen in the hominin pelvis (see Lovejoy et al. 2016). The “sagittalization” and the transverse narrowing of the hominin false pelvis that occurred in order to allow an abductor complex coupled with the long, flexible, lordosed lumbar column resulted in the tapering of the inferior thoracic cage. This “waisting” of the elongated hominin iliocostal space necessarily resulted in the reduction of the surface area of the hominin diaphragm relative to the condition in the apes, wherein the transversely broad inferior thorax provides an expansive area of attachment for this critical respiratory muscle (the diaphragm). This means that in order to produce similar inspiratory flow rates, hominins must spatially displace their relatively smaller diaphragms more as a consequence of the reduced radius of curvature and abbreviated cross sectional area of the muscle attachment area. To compensate for this somewhat reduced ability to diaphragmatically breathe, hominins declined (and twisted) their ribs allowing the addition of “thoracic” breathing, which is to say, they elevated the caudally angled ribs and in so doing increased the anteroposterior dimension of the thorax (the so-called “pump handle” effect). Thus, the addition of this thoracic/costal breathing mechanism to what would otherwise be primarily the diaphragmatic ventilator mechanism accompanied the structural changes in the pelvis and lumbar column required for habitual bipedality.

This hypothesis is supported by the observation that prior to the ontogenetic modifications of the thoracic cage described above, young children largely breathe utilizing the diaphragm and do not, and cannot, effectively use the thoracic mechanism until the ribs begin to decline. Similarly, the African apes with their horizontally oriented ribs (i.e., non-declined) are much less able to effectively utilize the anterosuperior rib elevating mechanism. Habitual, obligate bipeds have, as a consequence of their erect posture, also forfeited their ability to utilize the “visceral pump” used by quadrupeds (Bramble and Carrier 1983; Daley et al. 2013). This decoupling of the locomotor-respiratory systems in hominins has resulted in the much greater flexibility in breathing patterns in hominins (Daley et al. 2013). In addition, women with a greater decent and declination of the rib cage use a greater percentage of thoracic breathing than do men (Bellemare et al. 2003, 2006), offsetting their smaller diaphragms with enhanced volume displacements produced by thoracic rib elevation. In sum, the highly distinctive changes in the hominin rib cage can be viewed as a way of compensating for the modified pelvis and the long, flexible lumbar spine – both of which are necessary for habitual bipedality.

In light of these described ontogenetic alterations (and associated uniquely hominin pathological conditions), many of what would otherwise be viewed as isolated features in the adult human thorax are in fact associated morphologies that are functionally and structurally interrelated, so that even fragmentary skeletal elements such as the KSD-VP-1/1 ribs allow important inferences to be made about adjoining structures.

African Ape Thoracic Shape

As described above, the “barrel-shaped” human thorax contrasts markedly with the inverted “funnel-shaped” or “frustum-shaped” thorax of the African apes, wherein the cupola is transversely constricted and the inferior thorax is markedly broadened in order to conform to the transversely elongated (broadened) iliac blades (see Fig. 7.4). In dramatic contrast to humans who have narrowed the inferior portions of their rib cage (waisting) to conform to the narrow false pelvis, the African apes have markedly expanded their lower thoracic cage to conform to the “coronalization” of their ilia. The resultant broad and tall pelvis in combination with the reduced number of lumbar vertebral elements (average = 3.5 vertebral elements) results in the virtual immobility of the African ape vertebral column (Lovejoy 2005).

As discussed above, additional changes seen in the African ape thorax include the ligamentous and skeletal “entrapment” of the inferior most lumbar vertebrae between the elongated iliac blades and the rigid fixation of the lower ribs upon the lowermost thoracic vertebrae (Ward et al. 2012). All of these changes act to dramatically stiffen the African ape thoracolumbar spine, essentially reducing it to an immobile “poker spine.” This condition is an adaptation for climbing in a large-bodied ape to reduce potentially damaging shear and bending stresses on the lumbar column during active and vigorous bouts of arboreality (Lovejoy 2005). Moreover, as a consequence of these adaptations and the lack of a long, flexible lumbar lordosis, African ape ribs neither decline inferiorly nor do they demonstrate torsion along the rib corpus (Jellema et al. 1993). Indeed, the ribs of African apes cannot decline inferiorly as a consequence of their markedly abbreviated iliocostal space. The reduced iliocostal region in the African apes also prohibits any lateral mobility of the vertebral column, further stiffening the spine. In addition, the orientation of the vertebral transverse processes and costal curvatures (Latimer and Ward 1993; Jellema et al. 1993; Ward et al. 2012) indicates less spinal invagination and reduced costal neck angles. It is noteworthy that chimpanzees have vertebral columns that are less anteriorly invaginated than are those of gorillas (Kagaya et al. 2008), raising the possibility that the smaller African apes have secondarily modified their thoracic skeletons. Inasmuch as the thoracic vertebral counts differ significantly in chimpanzees and bonobos (McCollum et al. 2009), further examination of their rib elements would be a valuable area for future research.

In view of the obvious shape differences between human and African ape thoraces, it seems prudent to pose the question as to why the African great apes have markedly narrowed their upper thoracic cages. Traditionally, this thoracic shape has been attributed to climbing (see Hunt 1991; Schmid 1991; Schmid et al. 2013) with the implication that Australopithecus (and perhaps other early hominins) maintained this so-called primitive “funnel-shaped” thorax and that, furthermore, this was evidence of adaptively significant amounts of arboreality in this genus. However, the KSD-VP-1/1 ribs and pelvis (see Lovejoy et al. 2016) fail to support this scenario and instead point to a complete lack of evidence suggesting that Australopithecus ever possessed a “funnel-shaped” thorax. Because features like rib declination and torsion are evidence of an elongated and flexible lordotic region in KSD-VP-1/1, and because these differ dramatically from the African ape inflexible poker spine, it seems highly unlikely that Australopithecus ever engaged in ape-like arboreality. In fact, these early hominins would have been physically incapable of African ape-like climbing (contra Stern and Susman 1983; Stern 2000), and quite obviously, show no adaptations to it.

In an earlier analysis of the KNM-WT 15000Homoerectus thoracic cage by Jellema et al. (1993), the lack of relevant fossil material led to the erroneous conclusion that humans had expanded their cupular region in compensation for their narrowed caudal rib cage. With the discovery of KSD-VP-1/1 this now seems improbable and instead it now appears that early hominins never possessed a “funnel-shaped” thorax.

A re-examination of the mechanical reasons for the African ape morphology can now be addressed. Instead of being viewed as a climbing and suspensory adaptation, the narrowed operculum can instead be seen as an adaptation to the unique African ape locomotor form, knuckle-walking. While the transversely broad thorax in the lesser and greater apes realigns scapular orientation to permit enhanced shoulder mobility and suspensory climbing, it also places the glenohumeral joint in an exceptionally vulnerable position during terrestrial locomotion.

Owing to the extremely low coefficient of friction in any healthy synovial joint it is necessary that the major transarticular forces remain perpendicular to the adjoining chondral surfaces (Burstein and Wright 1994). This requirement has profound consequences for large-bodied animals seeking to combine arboreal and terrestrial locomotor modalities. In chimpanzees and gorillas, the scapulae are positioned on the dorsum of the thorax such that the glenohumeral joints are directed superiorly and laterally, an orientation suitable for heightened shoulder mobility and suspensory climbing. However, this same orientation becomes highly deleterious during knuckle-walking because, unless modified, it would subject the glenohumeral joint to considerable shear stresses caused by the superoposteriorly directed ground reaction forces during the knuckle-strike phase. This is problematic in an extremely mobile articulation like the glenohumeral joint wherein the surrounding soft tissue envelope that largely maintains the joint’s integrity provides little resistance to shear forces. The stiff, extended elbow used during knuckle-walking and the abbreviated olecranon process in apes (Drapeau 2004; Simpson et al. nd) both greatly reduce the capacity of the M. triceps brachii to eccentrically and/or isometrically contract (dissipating reaction forces), further exacerbating this deficiency. The tapering of the upper thorax in the African apes is likely a parallelism (as is knuckle-walking) that ameliorates this fundamental problem by allowing the scapulae to rotate about the top of the “frustum-shaped” thorax (reduced radius of curvature). This permits the glenohumeral joint to align and face both ventrally and inferiorly thereby opposing ground collision forces and thus reducing potentially damaging shear stress across the shoulder. This altered position of the scapulae would also enhance the ability of M. serratusanterior to participate in impact attenuation through isometric and eccentric contraction.

The elevated shoulder orientation and oblique clavicular angle of the African apes (shrugged shoulders) are also related to reducing shear in the glenohumeral joint during knuckle-walking. The hands and wrists also demonstrate adaptations for this highly unusual mode of locomotion (Simpson et al. in review). As African apes are relatively large bodied and spend greater than 80% of their total locomotor repertoire (Doran 1997) engaging in bouts of knuckle-walking, it is not surprising that numerous skeletal and soft tissue adaptations to this activity are evident. Importantly, the fact that these highly specialized adaptations are not seen in KSD-VP-1/1 indicates that early hominins never knuckle-walked and, furthermore, that this unusual locomotor mode was independently derived in chimpanzees and gorillas (White et al. 2009).

The KSD-VP1/1 Thorax

The distinct and easily recognized anatomical contrasts between the thoracic skeletons of humans and the African apes provide a valuable context within which to examine the KSD-VP-1/1 rib remains (see Fig. 7.4). Moreover, this process should shed light on the thoracic skeleton of the CLCA (see Haile-Selassie et al. 2016) and also provide evidence regarding the evolutionary development of knuckle-walking and the “funnel-shaped” African ape thoracic cage.

The second rib has previously been demonstrated to be especially useful in discriminating the upper thoracic shapes between humans and African apes (Jellema et al. 1993; Haile-Selassie et al. 2010b). Thus, KSD-VP-1/1n, an essentially complete second rib, is particularly valuable in assessing the conformation of the cranial thorax in this specimen. As we noted earlier, the rib curvature in this specimen is more similar to humans than it is to chimpanzees (Fig. 7.5). The rib curvature index (Haile-Selassie et al. 2010b and see Fig. 7.5) clearly separates humans and KSD-VP-1/1n from the African apes indicating that the upper rib cage of this specimen was transversely expanded like humans and was not constricted as it is in Pan and Gorilla (see Fig. 7.4). Moreover, the neck/corpus angle indicates that the spine was invaginated deeply within the thorax, another distinguishing feature of hominins (Latimer and Ward 1993; Jellema et al. 1993; Lovejoy 2005; Ward et al. 2012). It is clear that the thoracic cupola of KSD-VP-1/1 was transversely broad and differed fundamentally from the condition in the African apes. As human costal declination and twisting of the rib corpus is especially apparent in the seventh and eighth ribs (Jellema et al. 1993), and these particular elements are also preserved in the KSD-VP-1/1 partial skeleton (KSD-VP-1q, 1 s, 1o, and 1p), additional information can also be gleaned about the caudal half of the rib cage of Au. afarensis.
Fig. 7.5

Method of rib curvature index assessment used for the second rib (KSD-VP-1/1n) and comparisons with Pan, Gorilla, and Homo second rib indices. Although one Gorilla specimen (CMNH-B 1781; N = 33) fell within the human range, it is clear that KSD-VP-1/1n lies well within the human range. Modified from Haile-Selassie et al. (2010b)

Although the ribs are fragmentary and are not associated with thoracic vertebrae, it is nevertheless possible to orient the pleural surface of the rib heads into appropriate anatomical position and to estimate rib orientation from such an examination (see Jellema et al. 1993 for additional details). The preserved portions of KSD-VP-1/1o (seventh or eighth rib) and KSD-VP-1/p (eighth or ninth rib) indicate an oblique, anteroinferior orientation, and obvious declination of the rib corpus. Axial torsion along the costal corpus is apparent, and this is consistent with the rib declination and a long, flexible human-like lumbar vertebral column. This also suggests a lumbar lordosis and the progressive diminishment of the anteroposterior thoracic diameter relative to the transverse breadth of the thoracic cage. This latter observation also agrees with the pelvic morphology of the KSD-VP-1/1 partial skeleton (see Lovejoy et al. 2016). Because several of the characteristic features of the human thoracolumbar vertebral column are ontogenetically and structurally linked to the costal cage (see above), it is possible to reconstruct additional associated anatomies in the KSD-VP-1/1 partial skeleton.

It is evident that the KSD-VP-1/1 specimen did not have an ape-like, constricted upper thorax but rather had a transversely expanded operculum like that seen in modern humans. Moreover, the declination and torsion of the KSD-VP-1/1 costal elements indicate that this specimen had a long lower torso, a long and flexible lumbar column, and a high degree of spinal column invagination relative to the condition seen in the African apes. That the KSD-VP-1/1 thorax exhibits an anteroinferior declination of its costal elements is also clear, again paralleling the condition described in humans (Jellema et al. 1993; Bastir et al. 2013). All of the elements representing the thorax in KSD-VP-1/1 suggest a Homo-like rib cage and one that differs dramatically from the African ape condition.

These observations obviously indicate that the thoracic cage of KSD-VP-1/1 was not “funnel-shaped”, but rather was transversely broad in its cranial half and relatively tapered (compared to the African apes) in its lower portion (see Fig. 7.4). It should be noted, however, that the relative mediolateral breadth of the pelvis in Australopithecus was greater than in modern humans, necessitating some amount of lateral flaring in the lower thorax (see Latimer and Ward 1993; Lovejoy et al. 2016). In light of this, the thoracic shape of early hominins (pre-Homo sapiens) should be more aptly described as “bell-shaped.” This designation would eliminate the overly simplistic dichotomy of either barrel or funnel shaped and instead would recognize that Australopithecus would have flared slightly more at its thoracic bottom (less waisting than in modern humans, more waisting than in African apes) as a consequence of their broad bi-iliac breath induced by the platypelloidy of early (and even later) members of the species (Tague and Lovejoy 1986; Simpson et al. 2008; Lovejoy et al. 2016).

Following the same reasoning, Sawyer and Maley (2005) described the flaring of the caudal portion of the Neanderthal thoracic cage (owing to the broad false pelvis) as having created a “bell-shaped” thorax. With this in mind, perhaps the “barrel-shaped” descriptor should be reserved exclusively for modern Homosapiens. It is now abundantly clear that these early hominins did not have “frustum-shaped” or “funnel-shaped” thoraces. Furthermore, it now is also obvious that they possessed Homo-likerib declination, and a long flexible, lordosed lumbar region. This latter feature, an elongated lumbar spine, clearly contravenes any selectively significant amount of African ape-like climbing in Au.afarensis. As importantly, the complete lack of a structural lordotic spine and the complete inability to even facultatively assume a lordosis in the African apes virtually precludes their assuming hominin-like bipedality, casting serious doubt upon the use of extant chimpanzees and/or gorillas as appropriate models for early hominin locomotion. Similarly, earlier suggestions that Australopithecus locomotion was confined to a BHBK posture during walking (Stern and Susman 1983; Stern 2000) are now contravened by all available evidence from a now dramatically improved fossil record for the genus. It is now clear that Australopithecus possessed long, flexible, lumbar columns that were fully capable of achieving structural lordotic curvatures; indeed, early hominins likely never went through an ape-like lumbar entrapment phase. Thus, earlier suggestions of ape-like arboreality in Australopithecus must be re-examined in this light.


The KSD-VP-1/1 thorax was “bell-shaped,” clearly lacking an opercular constriction as in the African apes. Australopithecus did not have a frustum or “funnel-shaped” thorax. This conclusion is further supported by the somewhat less direct evidence provided by the Ardipithecusramidus false pelvis (Lovejoy et al. 2009a, b). Following this, several points become clear. First, the African ape thoracic shape with its constricted cupola is independently derived in Pan and Gorilla. Furthermore, this thoracic form is likely an adaptation to terrestrial knuckle-walking in a large-bodied, suspensory, and vertically climbing hominoid. Second, there is now no reliable evidence that any hominin ever possessed a narrow, constricted upper thorax. Finally, a broad upper thorax is probably primitive and the CLCA did not exhibit a conically shaped thorax.



We thank the Authority for Research and Conservation of Cultural Heritage (ARCCH) of the Ministry of Culture and Tourism of Ethiopia and administrative offices of the Afar Regional State of Ethiopia for field and laboratory research permits. We thank M. Decker, S. Simpson, and two anonymous reviewers for valuable comments. We also thank D. Su for help with figures. The Woranso-Mille project was financially supported by grants from The Leakey Foundation, The Wenner-Gren Foundation, The National Geographic Society, The Cleveland Museum of Natural History, and The National Science Foundation (BCS-0234320, BCS-0321893, BCS-0542037, BCS-1124705, BCS-1124713, BCS-1124716, BCS-1125157, and BCS-1125345).


  1. Aiello, L., & Dean, C. (1990). An introduction to human evolutionary anatamy. London: Academic Press Harcourt Brace & Company.Google Scholar
  2. Bastir, M., Martinez, D. G., Recheis, W., Barash, A., Coquerelle, M., Rios, L., et al. (2013). Differencial growth and development of the upper and lower human thorax. PLoS ONE, 8, e75128.CrossRefGoogle Scholar
  3. Bellemare, F., Jeanneret, A., & Couture, J. (2003). Sex differences in thoracic dimensions and configuration. American Journal of Respiratory and Critical Care Medicine, 168, 305–312.CrossRefGoogle Scholar
  4. Bellemare, F., Fuamba, T., & Bourgeault, A. (2006). Sexual dimorphism of human ribs. Respiratory Physiology & Neurobiology, 150, 233–239.CrossRefGoogle Scholar
  5. Bramble, D. M., & Carrier, D. R. (1983). Running and breathing in mammals. Science, 219, 251–256.CrossRefGoogle Scholar
  6. Burstein, A. H., & Wright, T. M. (1994). Fundamentals of orthopaedic biomechanics. New York: Williams & Wilkins.Google Scholar
  7. Christiansen, B. A., & Bouxsein, M. L. (2010). Biomechanics of vertebral fractures and the vertebral fracture cascade. Current Osteoporosis Reports, 8, 198–204.CrossRefGoogle Scholar
  8. Cook, D. C., Buikstra, J. E., DeRousseau, C. J., & Johanson, D. C. (1983). Vertebral pathology in the afar Australopithecines. American Journal of Physical Anthropology, 60, 83–101.CrossRefGoogle Scholar
  9. Cotter, M. M., Loomis, D. A., Simpson, S. W., Latimer, B., & Hernandez, C. J. (2011). Human evolution and osteoporosis-related spinal fractures. PLoS ONE, 6, e26658.CrossRefGoogle Scholar
  10. Daley, M. A., Bramble, D. M., & Carrier, D. R. (2013). Impact loading and locomotor-respiratory coordination significantly influence breathing dynamics in running humans. PLoS ONE, 8, e70752.CrossRefGoogle Scholar
  11. Doran, D. M. (1997). Ontogeny of locomotion in mountain gorillas and chimpanzees. Journal of Human Evolution, 32, 323–344.CrossRefGoogle Scholar
  12. Drapeau, M. S. (2004). Functional anatomy of the olecranon process in hominoids and plio-pleistocene hominins. American Journal of Physical Anthropology, 124, 297–314.CrossRefGoogle Scholar
  13. Franciscus, R. G., & Churchill, S. E. (2002). The costal skeleton of Shanidar 3 and a reappraisal of Neandertal thoracic morphology. Journal of Human Evolution, 42, 303–356.CrossRefGoogle Scholar
  14. Gómez-Olivencia, A., Eaves-Johnson, K. L., Franciscus, R. G., Carretero, J. M., & Arsuaga, J. L. (2009). Kebara 2: New insights regarding the most complete Neandertal thorax. Journal of Human Evolution, 57, 75–90.CrossRefGoogle Scholar
  15. Haile-Selassie, Y., Saylor, B. Z., Deino, A., Alene, M., & Latimer, B. M. (2010a). New hominid fossils from Woranso-Mille (Central Afar, Ethiopia) and taxonomy of early Australopithecus. American Journal of Physical Anthropology, 141, 406–417.Google Scholar
  16. Haile-Selassie, Y., Latimer, B. L., Alene, M., Deino, A., Gibert, L., Melillo, S. M., et al. (2010b). An early Australopithecus afarensis postcranium from Woranso-Mille, Ethiopia. Proceedings of the National Academy of Sciences USA, 107, 12121–12126.CrossRefGoogle Scholar
  17. Haile-Selassie, Y., Latimer, B. M., Lovejoy, C. O., Melillo, S. M., & Meyer, M. R. (2016). Conclusion: Implications of KSD-VP-1/1 for early hominin paleobiology and insights into the chimpanzee/human last common ancestor. In: Y. Haile-Selassie & D. F. Su (Eds.), The postcranial anatomy of Australopithecus afarensis: New insights from KSD-VP-1/1 (pp. 179–188). Dordrecht: Springer.Google Scholar
  18. Hunt, K. D. (1991). Mechanical implications of chimpanzee positional behavior. American Journal of Physical Anthropology, 86, 521–536.CrossRefGoogle Scholar
  19. Jellema, L. M., Latimer, B., & Walker, A. (1993). The rib cage. In A. Walker & R. E. Leakey (Eds.), The Nariokotome Homo erectus Skeleton (pp. 294–325). Cambridge: Harvard University Press.CrossRefGoogle Scholar
  20. Kagaya, M., Ogihara, N., & Nakatsukasa, M. (2008). Morphological study of the anthropoid thoracic cage: Scaling of thoracic width and an analysis of rib curvature. Primates, 49, 89–99.CrossRefGoogle Scholar
  21. Keith, A. (1923). Man’s posture: Its evolution and disorders. British Medical Journal, 1, 451–454.CrossRefGoogle Scholar
  22. Latimer, B., & Ward, C. V. (1993). The thoracic and lumbar vertebrae. In A. Walker & R. E. Leakey (Eds.), The Nariokotome Homo erectus skeleton (pp. 266–293). Cambridge: Harvard University Press.CrossRefGoogle Scholar
  23. Lovejoy, C. O. (1974). The gait of australopithecines. Yearbook of Physical Anthropology, 16, 18–30.Google Scholar
  24. Lovejoy, C. O. (2005). The natural history of human gait and posture II: hip and thigh. Gait and Posture, 21, 129–151.Google Scholar
  25. Lovejoy, C. O., & McCollum, M. (2010). Spinopelvic pathways to bipedality: why no hominids ever relied on a bent-hip–bent-knee gait. Philosophical Transactions of the Royal Society of London, 365, 3289–3299.CrossRefGoogle Scholar
  26. Lovejoy, C.O., Latimer, B. M., Spurlock, L., & Haile-Selassie, Y. (2016). The pelvic girdle and limb bones of KSD-VP-1/1. In Y. Haile-Selassie & D. F. Su (Eds.), The postcranial anatomy of Australopithecus afarensis: new insights from KSD-VP-1/1 (pp. 155–178). Dordrecht: Springer.Google Scholar
  27. Lovejoy, C. O., Simpson, S. W., White, T. D., Asfaw, B., & Suwa, G. (2009a). Careful climbing in the Miocene: the forelimbs of Ardipithecus ramidus and humans are primitive. Science, 326, 70e71–70e78.Google Scholar
  28. Lovejoy, C. O., Suwa, G., Simpson, S. W., Matternes, J. H., & White, T. D. (2009b). The great divides: Ardipithecus ramidus reveals the postcrania of our last common ancestors with African apes. Science, 326, 100–106.Google Scholar
  29. Masharawi, Y., Dar, G., Peleg, S., Steinberg, N., Medlej, B., Kay, H., et al. (2010). A morphological adaptation of the thoracic and lumbar vertebrae to lumbar hyperlordosis in young and adult females. European Spine Journal, 19, 768–773.CrossRefGoogle Scholar
  30. McCollum, M. A., Rosenman, B. A., Suwa, G., Meindl, R. S., & Lovejoy, C. O. (2009). The vertebral formula of the last common ancestor of African apes and humans. Journal of Experimental Zoology Part B Molecular and Developmental Evolution, 314B, 123–134.Google Scholar
  31. Ohman, J. C. (1986). The first rib of hominoids. American Journal of Physical Anthropology, 70, 209–229.CrossRefGoogle Scholar
  32. Openshaw, P., Edwards, S., & Helms, P. (1984). Changes in rib cage geometry during childhood. Thorax, 39, 624–627.CrossRefGoogle Scholar
  33. Ryan, T. M., & Sukhdeo, S. (2016). KSD-VP 1/1: Analysis of the postcranial skeleton using high-resolution Computed Tomography. In Y. Haile-Selassie & D. F. Su (Eds.), The Postcranial Anatomy of Australopithecus afarensis: New Insights from KSD-VP-1/1 (pp. 39–62). Dordrecht: Springer.Google Scholar
  34. Sawyer, G. J., & Maley, B. (2005). Neanderthal reconstructed. The Anatomical Record (part B: New Anatomy) 238B, 23–31.Google Scholar
  35. Schmid, P., Churchill, S. E., Nalla, S., Weissen, E., Carlson, K. J., et al. (2013). Mosaic morphology in the thorax of Australopithecus sediba. Science, 340, 1234598.CrossRefGoogle Scholar
  36. Schmid, P. (1991). The trunk of the Australopithecus. In Y. Coppens & B. Senut (Eds.), Origine(s) de la bipedie chez les hominides (pp. 225–234). Cahiers de Paleoanthropologie, Editions du CNRS. Paris.Google Scholar
  37. Schultz, A. H. (1961). Vertebral column and thorax. In H. Hofer, A. H. Schultz, & D. Starck (Eds.), Primatologia-Handbook of Primatology (Vol. 4, pp. 1–66). Basel/New York, NY: Karger.Google Scholar
  38. Scoles, P. V. (1991). Vertebral alterations in Scheuermann’s kyphosis. Spine, 16, 509–515.CrossRefGoogle Scholar
  39. Simpson, S. W., Quade, J., Levin, N. E., Butler, R., Dupont-Nivet, G., Everett, M., et al. (2008). A female Homo erectus pelvis from Gona, Ethiopia. Science, 322, 1089–1092.CrossRefGoogle Scholar
  40. Simpson, S. W., Lovejoy, C. O., & Latimer, B. M. (in review). Why do knuckle-walking African apes knuckle-walk? The Anatomical Record.Google Scholar
  41. Stern, J. T. (2000). Climbing to the top: A personal memoir of Australopithecus afarensis. Evolutionary Anthropology, 9, 113–133.CrossRefGoogle Scholar
  42. Stern, J. T., & Susman, R. L. (1983). The locomotor anatomy of Australopithecus afarensis. American Journal of Physical Anthropology, 60, 279–317.CrossRefGoogle Scholar
  43. Tague, R. G., & Lovejoy, C. O. (1986). The obstetric pelvis of AL 288-1 (Lucy). Journal of Human Evolution, 15, 237–255.CrossRefGoogle Scholar
  44. Todd, T. W. (1912). The decent of the shoulder after birth. Anatomischer Anzeiger, 41, 385–397.Google Scholar
  45. Ward, C. V., Kimbel, W. H., Harmon, E. H., & Johanson, D. C. (2012). New postcranial fossils of Australopithecus afarensis from Hadar, Ethiopia (1990-2007). Journal of Human Evolution, 63, 1–51.CrossRefGoogle Scholar
  46. White, T. D., Asfaw, B., Beyene, Y., Haile-Selassie, Y., Lovejoy, C. O., & Suwa, G. (2009). Ardipithecus ramidus and the paleobiology of early hominids. Science, 326, 75–86.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2016

Authors and Affiliations

  • Bruce M. Latimer
    • 1
    • 2
  • C. Owen Lovejoy
    • 2
    • 3
  • Linda Spurlock
    • 3
    • 2
  • Yohannes Haile-Selassie
    • 2
  1. 1.Department of Orthodontics School of Dental MedicineCase Western Reserve UniversityClevelandUSA
  2. 2.Department of Physical AnthropologyCleveland Museum of Natural HistoryClevelandUSA
  3. 3.Department of Anthropology and Division of Biomedical SciencesKent State UniversityKentUSA

Personalised recommendations