4.1 Introduction

During the Neogene, the earliest, or basal, hominins appear in the record of central and eastern African fossil deposits (Fig. 4.1). Changes in late Miocene environments may have provided the selective force for terrestrial bipedality, a key distinguishing feature of this group. At present, the interplay between morphological (evolutionary) and environmental change remains uncertain. Previous hypotheses suggested that savannas, or a mosaic of savannas and forests, played a significant role in evolution (e.g. Dart, Reference Dart1925; Ardrey, Reference Ardrey1961; Jolly, Reference Jolly1970; Lovejoy, Reference Lovejoy1981; Coppens, Reference Coppens1994; Elton, Reference Elton2008). More recently, there has been debate on the use of the term ‘savanna’ (see Domínguez-Rodrigo, Reference Domínguez-Rodrigo2014) and detailed analysis of the role of this ecosystem in early hominin evolution (WoldeGabriel et al., Reference WoldeGabriel, White, Suwa, Renne, de Heinzelin, Hart and Heiken1994, Reference WoldeGabriel, Ambrose, Barboni, Bonnefille, Bremond, Currie, DeGusta, Hart, Murray, Renne, Jolly-Saad, Stewart and White2009; Cerling et al., Reference Cerling, Levin, Quade, Wynn, Fox, Kingston, Klein and Brown2010, Reference Cerling, Wynn, Andanje, Bird, Korir, Levin, Mace, Macharia, Quade and Remien2011; White et al., Reference White, Ambrose, Suwa and WoldeGabriel2010).

Fig. 4.1. Map of major Mio-Pliocene hominin sites in eastern and southern Africa.

Research at basal hominin sites suggests that bipedalism may have first emerged in more wooded rather than open environments (e.g. White et al., Reference White, Suwa and Asfaw1994, Reference White, Ambrose, Suwa, Su, DeGusta, Bernor, Boisserie, Brunet, Delson, Frost, Garcia, Giaourtsakis, Haile-Selassie, Howell, Lehmann, Likius, Pehlevan, Saegusa, Semprebon, Teaford and Vrba2009a; Pickford and Senut, Reference Pickford and Senut2001). For example, fauna associated with the hominin Orrorin tugenensis suggest a gallery forest fringed with streams, shifting to open woodland away from the water sources (Pickford and Senut, Reference Pickford and Senut2001). Associated with Ardipithecus kadabba fossil remains, the environment varied from closed and open woodland to floodplain grasslands (Su et al., Reference Su, Ambrose, Degusta, Haile-Selassie, Haile-Selassie and WoldeGabriel2009). At Aramis (Ethiopia), slightly younger Ardipithecus fossils suggest less ecological variation. Fossil wood, seeds and associated fauna recovered alongside A. ramidus (WoldeGabriel et al., Reference WoldeGabriel, Ambrose, Barboni, Bonnefille, Bremond, Currie, DeGusta, Hart, Murray, Renne, Jolly-Saad, Stewart and White2009), combined with carbon isotopes from hominin teeth, suggest a woodland-to-forest habitat (White et al., Reference White, Asfaw, Beyene, Haile-Selassie, Lovejoy, Suwa and WoldeGabriel2009b). In their analysis of isotopic data from 1300 paleosols in the Awash and Omo-Turkana Valleys (Kenya), Cerling et al. (Reference Cerling, Wynn, Andanje, Bird, Korir, Levin, Mace, Macharia, Quade and Remien2011) argue that A. ramidus lived in grasslands and wooded grasslands during the late Miocene–early Pliocene (cf. White et al., Reference White, Ambrose, Suwa and WoldeGabriel2010). These reconstructions are vital to understanding the hominin shift to bipedality. The environment may also help explain why these species appear to be stenotopic and geographically restricted.

In the early Pliocene, the appearance of Australopithecus signifies a major commitment to bipedalism. Evidence suggests that Australopithecus was tolerant, possibly more so than its predecessors, to variations in its environment. This may explain how the genus successfully expanded its range into southern Africa near the early to late Pliocene transition. Presently, the earliest South African Australopithecus fossils derive from Sterkfontein, Member 2 (>3.0 Ma; Bruxelles et al., Reference Bruxelles, Clarke, Maire, Ortega and Stratford2014), and Makapansgat, Member 3.

4.2 Origin of the basal hominins

Three genera, Sahelanthropus, Orrorin and Ardipithecus, have been proposed as primitive representatives of our lineage (Fig. 4.2). Recognition of these basal hominins depends upon a suite of characteristics that, taken together, differentiates them from other groups such as Miocene or modern apes. The basal hominins exhibit less prognathism, possess small canines lacking honing wear, and demonstrate varying adaptations to bipedality. Basal hominins have been recovered in Chad (Sahelanthropus) and the East African Rift in Kenya (Orrorin) and Ethiopia (Ardipithecus). Recently, Haile-Selassie et al. (Reference Haile-Selassie, Suwa and White2004) suggested that all three basal hominins may represent a single species.

Fig. 4.2. Spatiotemporal distribution of Neogene hominins

(based upon Wood, Reference Wood2010, his fig. 1).

The oldest of the basal, bipedal group, Sahelanthropus tchadensis, was dated biostratigraphically to 7–6 Ma (Vignaud et al., Reference Vignaud, Duringer, Mackaye, Likius, Blondel, Boisserie, de Bonis, Eisenmann, Etienne, Geraads, Guy, Lehmann, Lihoreau, Lopez-Martinez, Mourer-Chauviré, Otero, Rage, Schuster, Viriot, Zazzo and Brunet2002), but ¹⁰Be cosmogenic nuclide dating on the sediments indicates that the fossils may date to the upper end of that range (7.2–6.8 Ma; Lebatard et al., Reference Lebatard, Bourlès, Duringer, Jolivet, Braucher, Carcaillet, Schuster, Arnaud, Monié, Lihoreau, Likius, Mackaye, Vignaud and Brunet2008). These dates closely approximate estimates from genetic research (Bradley, Reference Bradley2008; Fabre et al., Reference Fabre, Rodrigues and Douzery2009) indicating a divergence between Pan and the hominin lineage near 8–6 Ma. The Sahelanthropus cranium (TM 266-01-060-1) possesses traits such as a small (360–370 ml), low braincase and a relatively vertical face, a short canine with apical wear, implying a lack of a canine-premolar honing complex, and moderately thick enamel, suggestive of a transitional form between the Miocene apes and later hominins (Brunet et al., Reference Brunet, Guy, Pilbeam, Mackaye, Likius, Djimdoumalbaye, Beauvilain, Blondel, Bocherens, Boisserie, De Bonis, Coppens, Dejax, Denys, Duringer, Eisenmann, Fanone, Fronty, Geraads, Lehmann, Lihoreau, Louchart, Mahamat, Merceron, Mouchelin, Otero, Pelaez Campomanes, Ponce De Leon, Rage, Sapanet, Schuster, Sudre, Tassy, Valentin, Vignaud, Viriot, Zazzo and Zollikofer2002, Reference Brunet, Guy, Pilbeam, Lieberman, Likius, Mackaye, Ponce de Leon, Zollikofer and Vignaud2005). Complicating its phylogenetic assessment, Sahelanthropus is only known from craniodental remains. Other researchers have suggested that the cranium belongs to a female ape and not a hominin (Wolpoff et al., Reference Wolpoff, Senut, Pickford and Hawks2002, Reference Wolpoff, Hawks, Senut, Pickford and Ahern2006). Nevertheless, virtual reconstruction of the heavily distorted cranium recently suggests that Sahelanthropus was a biped (Zollikofer et al., Reference Zollikofer, Ponce de Leon, Lieberman, Guy, Pilbeam, Likius, Mackaye, Vignaud and Brunet2005) due to its inferiorly directed nuchal plane and anteriorly placed foramen magnum.

Another basal hominin, Orrorin tugenensis (~6.0 Ma), retains more ape-like dental morphology than the older Sahelanthropus. For example, Orrorin possesses a robust ape-like I¹ and triangular, pointed maxillary canines with mesial grooves, similar to that found in male monkeys and apes (Senut et al., Reference Senut, Pickford, Gommery, Mein, Cheboi and Coppens2001). In contrast, traits such as the long femoral neck, an obturator externus groove, and a shallow trochanteric fossa in the Orrorin femur (BAR 1002’00) strongly imply bipedality (Pickford et al., Reference Pickford, Senut, Gommery and Treil2002) and support its status as an early hominin. Recent morphometric analysis (Richmond and Jungers, Reference Richmond and Jungers2008) argues that Orrorin’s femur was most similar to the morphology, and perhaps gait, observed in Australopithecus and Paranthropus, rather than Homo, as previously claimed (Senut et al., Reference Senut, Pickford, Gommery, Mein, Cheboi and Coppens2001; Pickford et al., Reference Pickford, Senut, Gommery and Treil2002). Almécija et al. (Reference Almécija, Tallman, Alba, Pina, Moyà-Solà and Jungers2013), however, concluded that Orrorin femoral morphology is not intermediate between extant great apes and humans, but rather intermediate between plesiomorphic Miocene apes and Plio-Pleistocene hominins. Additionally, they argue that extant apes are derived relative to the ancestral hominoid morphology and exhibit higher diversity in proximal femoral morphology. Moreover, similarities among the extant great apes appear to be homoplastic (Almécija et al., Reference Almécija, Tallman, Alba, Pina, Moyà-Solà and Jungers2013).

The last of the Mio-Pliocene basal hominin groups is represented by Ardipithecus, which includes two temporally separated species, Ardipithecus kadabba (5.8–5.2 Ma; Haile-Selassie, Reference Haile-Selassie2001; Haile-Selassie et al., Reference Haile-Selassie, Suwa and White2004) and A. ramidus (4.4 Ma; WoldeGabriel et al., Reference WoldeGabriel, Ambrose, Barboni, Bonnefille, Bremond, Currie, DeGusta, Hart, Murray, Renne, Jolly-Saad, Stewart and White2009). The Miocene species Ardipithecus kadabba is known from the Asa Koma (5.54–5.77 Ma) and Kuseralee (~5.2 Ma) sites in the Middle Awash, Ethiopia (Haile-Selassie, Reference Haile-Selassie2001; WoldeGabriel et al., Reference WoldeGabriel, Haile-Selassie, Renne, Hart, Ambrose, Asfaw, Heiken and White2001; Haile-Selassie et al., Reference Haile-Selassie, Suwa and White2004; Renne et al., Reference Renne, Morgan, WoldeGabriel, Hart, Haile-Selassie, Haile-Selassie and WoldeGabriel2009). Initially designated as a sub-species of A. ramidus (Haile-Selassie, Reference Haile-Selassie2001), the potential presence of a primitive C/P₃ honing complex (C¹=ASK-VP-3/400, C₁=STD-VP-2/61 and P₃=ASK-VP-3/403) resulted in elevation to its own species (Haile-Selassie et al., Reference Haile-Selassie, Suwa and White2004). Some post-cranial elements, such as the proximal pedal phalanx (AME-VP-1/71), are suggestive of bipedality (Haile-Selassie et al., Reference Haile-Selassie, Suwa, White, Haile-Selassie and WoldeGabriel2009). It exhibits a dorsally canted proximal articular surface, a feature associated with toeing-off in bipeds (Latimer and Lovejoy, Reference Latimer and Lovejoy1990).

Ardipithecus ramidus, the better known of the two species, is represented by 110 specimens including a partial female skeleton (White et al., Reference White, Asfaw, Beyene, Haile-Selassie, Lovejoy, Suwa and WoldeGabriel2009b). In initial descriptions, White et al. (Reference White, Suwa and Asfaw1994) noted thin enamel, a lower premolar (P₃) with a single cusp, and a small posterior dentition – ape-like traits suggesting closeness to the last common ancestor of chimps and humans. Recent studies (White et al., Reference White, Suwa and Asfaw1994, Reference White, Asfaw, Beyene, Haile-Selassie, Lovejoy, Suwa and WoldeGabriel2009b; Semaw et al., Reference Semaw, Simpson, Quade, Renne, Butler, McIntosh, Levin, Dominguez-Rodrigo and Rogers2005, Lovejoy et al., Reference Lovejoy, Latimer, Suwa, Asfaw and White2009a–Reference Lovejoy, Suwa, Spurlock, Asfaw and Whited; Suwa et al., Reference Suwa, Asfaw, Kono, Kubo, Lovejoy and White2009a, Reference Suwa, Kono, Simpson, Asfaw, Lovejoy and Whiteb) have revealed characteristics linking A. ramidus to later hominins. These include: small canines lacking evidence of a C/P₃ honing complex, an anteriorly positioned foramen magnum, implied lumbar lordosis, an anterior iliac spine, and a greater sciatic notch. Additionally, very little sexual dimorphism was observed among the Aramis A. ramidus individuals. However, A. ramidus also exhibits primitive characters, such as the absence of a longitudinal arch in the foot and relatively equal fore- to hindlimb lengths (Lovejoy et al., Reference Lovejoy, Latimer, Suwa, Asfaw and White2009a, Reference Lovejoy, Suwa, Simpson, Matternes and Whitec). Furthermore, post-cranial evidence reveals adaptations for terrestrial bipedalism and arboreal clambering (Lovejoy et al., Reference Lovejoy, Suwa, Simpson, Matternes and White2009c) with a uniquely competent (and primitive) grasping hallux (Lovejoy et al., Reference Lovejoy, Latimer, Suwa, Asfaw and White2009a). The pelvis of A. ramidus (Lovejoy et al., Reference Lovejoy, Suwa, Spurlock, Asfaw and White2009d) displays a mixture of palmigrade clambering and terrestrial bipedal features – a more primitive form of bipedality than later Australopithecus (White et al., Reference White, Asfaw, Beyene, Haile-Selassie, Lovejoy, Suwa and WoldeGabriel2009b).

4.3 Diversification of the archaic hominins

The term archaic refers to the taxa between the basal hominins and Homo (Wood and Lonergan, Reference Wood and Lonergan2008), including the genera Australopithecus and Kenyanthropus. Much of the australopith anatomy suggests an increasing commitment to bipedality, with current evidence (e.g. Leakey et al., Reference Leakey, Fiebel, McDougall and Walker1995) suggesting that Australopithecus originated in eastern Africa. When compared to A. ramidus (White et al., Reference White, Asfaw, Beyene, Haile-Selassie, Lovejoy, Suwa and WoldeGabriel2009b), the spatial and temporal distribution of Australopithecus localities indicate habitats from closed woodland and forest to open grassland and shrub (Behrensmeyer and Reed, Reference Behrensmeyer, Reed, Reed, Fleagle and Leakey2013; Brown et al., Reference Brown, McDougall, Gathogo, Reed, Fleagle and Leakey2013). However, one should be cautious of assigning individual species to stenotopic or eurytopic categories (Bobe et al., Reference Bobe, Behrensmeyer and Chapman2002). The oldest species (4.2–3.9 Ma), Australopithecus anamensis, is preserved at sites such as Kanapoi and Allia Bay, Kenya (Leakey et al., Reference Leakey, Fiebel, McDougall and Walker1995; Ward et al., Reference Ward, Leakey and Walker2001, Reference Ward, Manthi and Plavcan2013), Asa Issie, Ethiopia (White et al., Reference White, WoldeGabriel, Asfaw, Ambrose, Beyene, Bernor, Boisserie, Currie, Gilbert, Haile-Selassie, Hart, Hlusko, Howell, Kono, Lehmann, Louchart, Lovejoy, Renne, Saegusa, Vrba, Wesselman and Suwa2006) and possibly Fejej, southern Ethiopia (Fleagle et al., Reference Fleagle, Rasmussen, Yirga, Bown and Grine1991). The transition from Ardipithecus to Australopithecus around 4.4–4.2 Ma was rapid. However, the mode of speciation is unresolved (White et al., Reference White, WoldeGabriel, Asfaw, Ambrose, Beyene, Bernor, Boisserie, Currie, Gilbert, Haile-Selassie, Hart, Hlusko, Howell, Kono, Lehmann, Louchart, Lovejoy, Renne, Saegusa, Vrba, Wesselman and Suwa2006) although changes in local to regional environments may have been significant. Existing data suggest a more wooded biome for the Ethiopian material (White et al., Reference White, WoldeGabriel, Asfaw, Ambrose, Beyene, Bernor, Boisserie, Currie, Gilbert, Haile-Selassie, Hart, Hlusko, Howell, Kono, Lehmann, Louchart, Lovejoy, Renne, Saegusa, Vrba, Wesselman and Suwa2006), contrasting with the mosaic wooded/savanna reconstructions for the Kenyan A. anamensis sites (Coffing et al., Reference Coffing, Feibel, Leakey and Walker1994; Leakey et al., Reference Leakey, Fiebel, McDougall and Walker1995; Schoeninger et al., Reference Schoeninger, Reeser and Hallin2003).

Craniodentally, A. anamensis lacked derived traits seen in later Australopithecus. It possessed parallel post-canine toothrows and a distinctive canine-premolar complex with mesiodistally longer honing teeth and large canine roots, especially at Kanapoi (Ward et al., Reference Ward, Manthi and Plavcan2013). While similar, hypodigm maxillae from Kanapoi (KNM-KP 29283) and the Middle Awash (ARA-VP-14/1) vary considerably in molar size and canine shape (White et al., Reference White, WoldeGabriel, Asfaw, Ambrose, Beyene, Bernor, Boisserie, Currie, Gilbert, Haile-Selassie, Hart, Hlusko, Howell, Kono, Lehmann, Louchart, Lovejoy, Renne, Saegusa, Vrba, Wesselman and Suwa2006). Molar crown dimensions of some specimens (e.g. ASI-VP-2/334) from Asa Issie are at the upper end, or above, the previously known ranges for A. anamensis. One canine from Asa Issie (ASI-VP-2/334) is the largest hominin canine yet found (White et al., Reference White, WoldeGabriel, Asfaw, Ambrose, Beyene, Bernor, Boisserie, Currie, Gilbert, Haile-Selassie, Hart, Hlusko, Howell, Kono, Lehmann, Louchart, Lovejoy, Renne, Saegusa, Vrba, Wesselman and Suwa2006). As a group, A. anamensis canines are still smaller than extant apes. All three canines from Asa Issie (ASI-VP-2/2, ASI-VP-2/334 and ASI-VP-2/367) possessed well-developed mesiolingual ridges, intermediate in form between Ardipithecus and Australopithecus afarensis (White et al., Reference White, WoldeGabriel, Asfaw, Ambrose, Beyene, Bernor, Boisserie, Currie, Gilbert, Haile-Selassie, Hart, Hlusko, Howell, Kono, Lehmann, Louchart, Lovejoy, Renne, Saegusa, Vrba, Wesselman and Suwa2006). Additionally, the mandibular symphysis, as seen in the holotype KNM-KP 29281A, reaches as far back as the first molar (M₁). Furthermore, it is also associated with a fragmentary temporal bone (KNM-KP 29281B), a mandibular fossa lacking an articular eminence (Ward et al., Reference Ward, Leakey and Walker2001). Research (Teaford and Ungar, Reference Teaford and Ungar2000) suggests that A. anamensis was the first hominin to exhibit thicker enamel when compared to earlier species, such as Ardipithecus (Suwa et al., Reference Suwa, Asfaw, Kono, Kubo, Lovejoy and White2009a, Reference Suwa, Kono, Simpson, Asfaw, Lovejoy and Whiteb). Australopithecus was better adapted to a more abrasive diet requiring greater mastication (Teaford and Ungar, Reference Teaford and Ungar2000; Ward et al., Reference Ward, Leakey and Walker2001). This dietary shift, however, does not appear to be connected to any global climatic change (Alemseged, Reference Alemseged, Reed, Fleagle and Leakey2013).

Postcranially, A. anamensis demonstrates a continuing adaptation to bipedalism whilst retaining many primitive and ape-like features in the upper skeleton (for a review, see Alemseged, Reference Alemseged, Reed, Fleagle and Leakey2013). For instance, an A. anamensis proximal tibia (KNM-KP 29285) illustrates conformation to a bipedal pattern with an extended knee adapted to habitual bipedality (Ward et al., Reference Ward, Leakey and Walker2001). The upper limb, however, still retained significant primitive characters, such as a long radius and a mildly curved manual phalanx. Ward et al. (Reference Ward, Leakey and Walker2001), however, caution against over-interpretation of these features. Arguments vary on the importance of arboreal traits in Australopithecus, especially with regard to a later hominin (A. afarensis), as reflections of locomotor behaviour (Senut and Tardieu, Reference Senut, Tardieu and Delson1979; Stern and Susman, Reference Stern and Susman1983; Stern, Reference Stern2000; Alemseged et al., Reference Alemseged, Spoor, Kimbel, Bobe, Geraads, Reed and Wynn2006; Green and Alemseged, Reference Green and Alemseged2012) or retentions from a common ancestor (Lovejoy, Reference Lovejoy1981; Latimer, Reference Latimer, Coppens and Senut1991; Ward, Reference Ward, Reed, Fleagle and Leakey2013).

As first proposed by Kimbel et al. (Reference Kimbel, Lockwood, Ward, Leakey, Rak and Johanson2006), new evidence from Woranso-Mille (Ethiopia) supports an ancestor-descendant relationship between A. anamensis and A. afarensis (Haile-Selassie et al., Reference Haile-Selassie, Saylor, Deino, Alene and Latimer2010b). There is also debate as to whether A. afarensis should subsume Kenyanthropus platyops (Leakey et al., Reference Leakey, Spoor, Brown, Gathogo, Kiarie, Leakey and McDougall2001; White, Reference White2003). Biogeographically, A. afarensis ranges as far south as Laetoli, Tanzania (White, Reference White and Hartwig2002), with possibly its earliest occurrence (~3.8 Ma) at Belohdelie, Ethiopia (Asfaw, Reference Asfaw1987). The species’ last occurrence is well-documented at Hadar (Ethiopia) in deposits dated between 3.4 and 2.9 Ma. In addition, the Hadar A. afarensis sample preserves the strongest evidence for australopith eurytopy, where it existed through fluctuations in its environment (Bonnefille et al., Reference Bonnefille, Potts, Chalié, Jolly and Peyron2004; Campisano and Feibel, Reference Campisano and Feibel2007; Reed, Reference Reed2008). Younger deposits in the Omo basin (Brown and de Heinzelin, Reference Brown, de Heinzelin and de Heinzelin1983) portray an important transitional period (2.9–2.7 Ma), but Suwa et al. (Reference Suwa, White and Howell1996) note that the hominin remains are too fragmentary to be diagnostic.

A recent review estimated that cranial capacities for A. afarensis were within and slightly above (380–550 ml) the chimpanzee range (Holloway et al., Reference Holloway, Broadfield and Yuan2004). Relative to earlier hominins, the face was slightly reduced, but still moderately prognathic, possessing an ape-like nasoalveolar clivus (e.g. AL 417-1). In the species, the primitive lower face is paired with a more derived mandible. Although mandibular morphology is highly variable (Kimbel et al., Reference Kimbel, Rak and Johanson2004), a novel feature, an upright mandibular symphysis, is observed for the first time in some specimens (e.g. AL 417-1, AL 444-2) (Kimbel and Delezene, Reference Kimbel and Delezene2009). Lockwood et al. (Reference Lockwood, Kimbel and Johanson2000) recognised that late-occurring A. afarensis sites, such as Hadar, hosted the largest overall mandibles. When compared to the earlier A. anamensis, the maxillary (I¹) and mandibular (I₁) incisors of A. afarensis are smaller (Ward et al., Reference Ward, Leakey and Walker2001). Conversely, canines were not appreciably different between A. anamensis, A. afarensis and A. africanus (Kimbel and Delezene, Reference Kimbel and Delezene2009). Canine attrition was intermediate (e.g. AL 200-1) between ape-like honing wear and the apical form of later hominins. Unlike the basal hominins and A. anamensis, the P₃ of A. afarensis possessed a more symmetrical single cusp with frequent development of a second cusp (Kimbel, Reference Kimbel, Henke and Tattersall2007). However, Alemseged (Reference Alemseged, Reed, Fleagle and Leakey2013) notes that a temporal trend (in P₃ morphology) cannot be discerned in the A. afarensis sample. Even juvenile specimens (Alemseged et al., Reference Alemseged, Spoor, Kimbel, Bobe, Geraads, Reed and Wynn2006) exhibited a moralisation of the premolars (e.g. dp₃). Further advancing megadontia, first noted in A. anamensis, and relative to its body size, A. afarensis displayed large premolars and molars. However, overall molar size between A. anamensis and A. afarensis was similar (Ward et al., Reference Ward, Leakey and Walker2001).

Possibly related to dietary changes, A. afarensis developed weak to moderate articular eminences and more derived mandibular fossae. Kimbel et al. (Reference Kimbel, Rak and Johanson2004) note that some specimens (e.g. AL 444-2 and AL 822-1) fall between the previous ranges (for mandibular fossae) of A. afarensis and the more South African derived A. africanus. In many cases, posteriorly oriented sagittal crests or closely approximated temporal lines can be observed on A. afarensis crania, possibly suggesting changes in chewing stresses (Kimbel and Delezene, Reference Kimbel and Delezene2009).

Post-cranially, A. afarensis exhibited many ape-like features in its wrist, shoulder, primitive limb proportions and long curved phalanges (Susman and Stern, Reference Susman, Stern, Coppens and Senut1991; Stern, Reference Stern2000; Alemseged et al., Reference Alemseged, Spoor, Kimbel, Bobe, Geraads, Reed and Wynn2006; Arias-Martorell et al., Reference Arias-Martorell, Potau, Bello-Hellegouarch and Pérez-Pérez2015). Regarding juvenile scapulae from Dikika, Ethiopia (DIK-1-1), Green and Alemseged (Reference Green and Alemseged2012) argue that A. afarensis exhibited substantial suspensory behaviour. The benefit of cranially oriented glenoid fossae rests in the ability to transfer the weight more directly from the vertebral column to the forelimb when climbing, or hanging from one arm (Hunt, Reference Hunt1991). But a recently recovered adult scapula from Woranso-Mille (KSD-VP-1/1g) appears more Homo-like, questioning the ape-like morphology previously attributed to A. afarensis. Haile-Selassie et al. (Reference Haile-Sellassie, Latimer, Alene, Deino, Gilbert, Meililo, Saylor, Scott and Lovejoy2010a) remark that the scapula does not offer evidence for a suspensory history in A. afarensis. When applied to the fossils, a new methodology (Green et al., Reference Green, Serrins, Seitelman, Martiny and Gunz2015) may further reveal the functional nuances. Kimbel and Delezene (Reference Kimbel and Delezene2009) note that the humerofemoral index in A. afarensis was intermediate (~85) relative to small-bodied Homo sapiens (~74) and P. paniscus (~98). The relatively long upper limb may reflect sleeping and/or feeding activities in the trees, much like their great ape relatives (Sept, Reference Sept1992). However, the intermediate proportions result from a relatively short (ape-like) femur, and not a long (ape-like) humerus (Jungers, Reference Jungers1982; Jungers and Stern, Reference Jungers and Stern1983).

While there is debate on the exact form of locomotion exhibited by A. afarensis (see Stern, Reference Stern2000; Ward, Reference Ward, Reed, Fleagle and Leakey2013, for discussion), data strongly suggest terrestrial bipedality. Hypotheses range from modern, human-like extended knee locomotion (e.g. Latimer and Lovejoy, Reference Latimer and Lovejoy1990) to a bent knee and bent hip gait (e.g. Stern and Susman, Reference Stern and Susman1983). The pelvic evidence reveals changes in the orientation of the iliac blades which are more medially rotated, and dramatically shortened, when compared to extant apes (Fig. 4.3). Because A. afarensis (e.g. AL 288) possessed valgus femora, the feet were placed directly beneath the pelvis. Additionally, footprints preserved in volcanic ash at Laetoli (~3.6 Ma) indicate the presence of longitudinal arches, as well as an adducted hallux, or great toe (Raichlen et al., Reference Raichlen, Gordon, Harcourt-Smith, Foster and Haas2010). Some, however, disagree with the level of hallucial adduction postulated for A. afarensis (Stern and Susman, Reference Stern and Susman1983; Clarke and Tobias, Reference Clarke and Tobias1995). Newly recovered material, a 4th metatarsal (AL 333-160), reveals transverse and longitudinal arches in a foot with no midtarsal break (Ward et al., 2011). The question is thus not whether A. afarensis was bipedal when on the ground (cf. Sarmiento, Reference Sarmiento1998), but whether its arboreal features reflected a significant contribution to its locomotor behaviour (Stern, Reference Stern2000).

Fig. 4.3. Pelvis comparisons. Shown are the complete pelvic girdles of three hominid species: (1) P. troglodytes, (2) A. afarensis (AL 288-1, cast) and (3) H. sapiens. Pan exhibits narrow, elongated ilia whilst in Australopithecus and Homo, they are shortened and flared. Scale = 5 cm

(photo: Jason Heaton).

Reconstructions of sexual dimorphism in A. afarensis vary from Gorilla- and Pongo-like (Gordon et al., Reference Gordon, Green and Richmond2008) to modern human-like (Reno et al., Reference Reno, Meindl, McCollum and Lovejoy2003, Reference Reno, McCollum, Meindl and Lovejoy2010). Ape-like dimorphism would signal substantial male–male competition for reproductive access (Plavcan and van Schaik, Reference Plavcan and van Schaik1997), but a reduction in dimorphism could indicate (Reno et al., Reference Reno, Meindl, McCollum and Lovejoy2003, Reference Reno, McCollum, Meindl and Lovejoy2010) male provisioning of females (Lovejoy, Reference Lovejoy1981). Thus, based upon present evidence, potential dimorphism is unresolved (Gordon, Reference Gordon, Reed, Fleagle and Leakey2013).

4.4 Dispersal of hominins into South Africa

Hominin populations do not appear to have expanded southwards until the Pliocene. At Langebaanweg, a South African Mio-Pliocene (~5 Ma) site, a diverse array of vertebrates is preserved (Hendey, Reference Hendey and Klein1984), but no hominin has been recovered. Current evidence (see Ségalen et al., Reference Ségalen, Lee-Thorp and Cerling2007) suggests that hominin migrations were timed with the expansion of C₄ grasses. Around 4.0–3.0 Ma, African mammal migrations (Strait and Wood, Reference Strait and Wood1999) coincide with the first occurrence of grasses and their associated fauna at Makapansgat, the northernmost South African hominin site (Reed, Reference Reed1997; Sponheimer et al., Reference Sponheimer, Reed and Lee-Thorp1999; Hopley et al., Reference Hopley, Marshall, Weedon, Latham, Herries and Kuykendall2007). Farther south, palaeoenvironmental reconstructions from the Pliocene deposits at Sterkfontein (Vrba, Reference Vrba1974; McKee, Reference McKee1991; Bamford, Reference Bamford1999; Pickering et al., Reference Pickering, Clarke and Heaton2004) and Taung (Berger and Clarke, Reference Berger and Clarke1995) suggest moister, wooded habitats.

In South Africa, three species were initially identified: Australopithecus africanus at Taung (Dart, Reference Dart1925), Plesianthropus transvaalensis at Sterkfontein (Broom, Reference Broom1938) and Australopithecus prometheus at Makapansgat (Dart, Reference Dart1948). Each possessed features that were thought to distinguish them, but Robinson (Reference Robinson1954) grouped all three into A. africanus. Overall, the body of A. africanus was similar to A. afarensis with a few notable exceptions (Ward, Reference Ward, Reed, Fleagle and Leakey2013). An analysis (Clarke and Tobias, Reference Clarke and Tobias1995) of the foot of StW 573, the oldest South African hominin (Bruxelles et al., Reference Bruxelles, Clarke, Maire, Ortega and Stratford2014), suggests a greater range of motion than its predecessor (cf. McHenry and Jones, Reference McHenry and Jones2006). Additionally, the limbs of A. africanus, and their proportions, are suggestive of increased arboreality (Green et al., Reference Green, Gordon and Richmond2007). These may reflect autapomorphic traits resulting from local adaptations to the more wooded environment of the Sterkfontein Valley.

In A. africanus, the cranium (e.g. Sts 5) was more globular and less ape-like than in previous hominins. Compared to modern humans, however, the braincase remained relatively small (~428–515 ml) and the face is bordered by a large brow ridge and sloping forehead. Between the brow and maxilla, the face remains prognathic and chimp-like, but with a flattened nose and nasoalveolar clivus. Anteriorly, the palate is more parabolic, trending toward Homo in morphology. Unlike extant apes and most early hominins, the mandibular fossa is deepened with a distinct articular eminence. Compared to A. afarensis, the incisors and canines are reduced whilst the posterior dentition is slightly enlarged (Kimbel and Delezene, Reference Kimbel and Delezene2009). Adult dental morphology is nearly devoid of primitive features, though the Taung child, a juvenile A. africanus (Fig. 4.4), retains an ape-like deciduous third premolar (dp₃).

Fig. 4.4. Taung Child. An example of a juvenile of the South African species, A. africanus. The lower deciduous third premolar (dp₃) is ape-like in morphology. In later hominins this is replaced by the modern molariform type

(photo: Jason Heaton).

A potential problem for the single species hypothesis, as it relates to South African Pliocene hominins, is the high degree of morphological variation, especially at Sterkfontein. Some researchers (Aguirre, Reference Aguirre1970; Clarke, Reference Clarke and Delson1985, Reference Clarke and Grine1988) have suggested that two distinct species are present. At Makapansgat, Tobias (Reference Tobias1967) noted the presence of ‘robust’ features, suggesting that they were more similar to Australopithecus robustus, a Pleistocene megadont hominin, than the Sterkfontein sample. Several analyses (Kimbel and Rak, Reference Kimbel, Rak, Kimbel and Martin1993; Ahern, Reference Ahern1998; Lockwood and Tobias, Reference Lockwood and Tobias2005) have attempted to address the range of variability, with little agreement (see review by Grine, Reference Grine, Reed, Fleagle and Leakey2013). Kimbel and White (Reference Kimbel, White and Grine1988) suggested that the Sterkfontein crania could also be separated by the degree of prognathism, with Sts 5 and Sts 71 representing the two groups (see Fig. 4.5 for comparison). Recently, Clarke (Reference Clarke, Reed, Fleagle and Leakey2013) assigned two groups at Sterkfontein to separate species, A. africanus and A. prometheus (i.e. the second species). Illustrating his case, a mandible (StW 384) and maxillae (Sts 1 and StW 183) of A. prometheus exhibit characteristics of later robust australopith groups, such as extreme bunodonty and comparatively large molars.

Fig. 4.5. Cranial comparisons. Pan troglodytes (left) exhibits a smaller and less globular braincase when compared to A. africanus (Sts 5-Ms Ples [middle] and Sts 71 [right]). Prognathism is also reduced in Australopithecus. Note that the dentition is missing in Sts 5. Scale = 1 cm

(photo: Jason Heaton).

Preliminary analyses of new postcranial elements, as discussed in Clarke (Reference Clarke, Reed, Fleagle and Leakey2013), provide additional evidence of two morphs (e.g. StW 562 and StW 595), the latter of which is more ape-like. The presence of two species at Sterkfontein would seemingly reject monophyly among the South African (Australopithecus robustus) and East African (A. aethiopicus and A. boisei) ‘robust’ forms. These species possibly originated through cladogenesis in the late Pliocene (3.6–2.6 Ma), but the nature of that event is debated (Grine, Reference Grine1988). Dating of the Sterkfontein deposits is also important to questions of evolutionary change, as Member 4 precedes and possibly bridges the Plio-Pleistocene boundary (Herries et al., Reference Herries, Pickering, Adams, Curnoe, Warr, Latham, Shaw, Reed, Fleagle and Leakey2013).

What is clear is that the Neogene record was dominated by bipedal hominins exhibiting varying degrees of arboreality. Near the close of the Pliocene (~2.6 Ma), a dramatic shift in behaviour occurred, as populations of hominins began to produce stone tools (Semaw et al., Reference Semaw, Rogers, Quade, Renne, Butler, Dominguez-Rodrigo, Stout, Hart, Pickering and Simpson2003). Unfortunately, there are currently no directly associated hominin remains. As a result, a late variant of australopith, A. garhi, is currently the best contender for the earliest toolmaker (Semaw, Reference Semaw2000). This behavioural shift set the stage for subsequent cultural evolution during the Pleistocene and its dramatic impact on the origins and evolution of the genus Homo. Now, evidence suggests that the first occurrence of Homo may have extended into the Pliocene (Villmoare et al., Reference Villmoare, Kimbel, Seyoum, Campisano, DiMaggio, Rowan, Braun, Arrowsmith and Reed2015).

4.5 Summary

During the Neogene, the hominin fossil record suggests a probable evolutionary path from the basal hominins (Sahelanthropus, Orrorin and Ardipithecus) to early forms of Australopithecus. Debate remains on the role that wooded environments had on the origins of bipedalism, a defining hominin trait. Undoubtedly, variations in woody cover played a significant role during early hominin evolution, and adaptations developed as a result of local responses, rather than coinciding with global climatic events. In reaction, the hominin lineage arguably becomes wholly committed to bipedality, which made possible later development of the complex cultural behaviours observed in the Quaternary. Near this Neogene–Quaternary boundary, cladogenesis in archaic hominins produced two lineages: one leading to ‘robust’, megadont forms of Australopithecus, and the other leading to Homo.