Download Human evolution: taxonomy and paleobiology and more Lecture notes Paleobiology in PDF only on Docsity!
J. Anat. (2000) 196 , pp. 19–60, with 3 figures Printed in the United Kingdom 19
Review
Human evolution : taxonomy and paleobiology
B E R N A R D W O O D A N D BRIAN G. R I C H M O N D
Department of Anthropology , George Washington University , and Human Origins Program , National Museum for Natural History , Smithsonian Institution , Washington , DC , USA
( Accepted 23 November 1999 )
This review begins by setting out the context and the scope of human evolution. Several classes of evidence, morphological, molecular, and genetic, support a particularly close relationship between modern humans and the species within the genus Pan , the chimpanzee. Thus human evolution is the study of the lineage, or clade, comprising species more closely related to modern humans than to chimpanzees. Its stem species is the so-called ‘ common hominin ancestor ’, and its only extant member is Homo sapiens. This clade contains all the species more closely-related to modern humans than to any other living primate. Until recently, these species were all subsumed into a family, Hominidae, but this group is now more usually recognised as a tribe, the Hominini. The rest of the review sets out the formal nomenclature, history of discovery, and information about the characteristic morphology, and its behavioural implications, of the species presently included in the human clade. The taxa are considered within their assigned genera, beginning with the most primitive and finishing with Homo. Within genera, species are presented in order of geological age. The entries conclude with a list of the more important items of fossil evidence, and a summary of relevant taxonomic issues.
Key words : Hominins ; cladistics ; Homo.
Human evolution : context and scope
Anatomical, molecular and genetic evidence suggests that the animal most closely related to modern humans is the chimpanzee, Pan , with Gorilla being more distantly related. Both of these ape genera are decidedly nonhuman in their appearance and be- haviour, and until recently their anatomical resem- blances had persuaded the majority of commentators to assume that Pan and Gorilla must be more closely- related to each other, and then to Pongo , the orangutan, than to modern humans, but a recent overview of traditional morphology narrowly links Homo and Pan (Shoshani et al. 1996). Prior to this, analyses of proteins (Zuckerkandl et al. 1960 ; Good- man, 1962, 1963 ; Zuckerkandl, 1963) and, more recently, of both nuclear and mitochondrial DNA of the great apes (Ruvolo, 1997), have shown that the similarities between Homo sapiens and Pan are very
Correspondence to Professor Bernard Wood, Henry R. Luce Professor of Human Origins, Anthropology Department, 2110 G St. NW, Washington DC 20052, USA.
close. An increasing number of researchers interpret this evidence as supporting the hypothesis that Homo and Pan share a common ancestry to the exclusion of Gorilla (Ruvolo, 1995). However, other scientists continue to maintain that the relationships between Homo , Pan and Gorilla are so close that their details have not yet been satisfactorily resolved, and suggest that the relationship between the 3 taxa is best treated as an unresolved trichotomy (Green & Djian, 1995 ; Marks, 1995 ; Rogers & Commuzzie, 1995 ; Deinard et al. 1998). Is it possible to determine how long ago a separate human lineage became established? Differences in the amino acid sequences of proteins, and in the base sequences of DNA, can be used to provide an estimate of how long lineages have been independent (Kimura, 1968, 1977). Most naturally-occurring mutations are neutral, conveying no discernible reproductive ad- vantage on the animal. If one makes the reasonable assumption that these neutral mutations have been
H. ergaster A. habilis Q A. rudolfensis Q
A. bahrelghazali
P. aethiopicus
Fig. 1. Hominin phylogram. Species considered to be part of the tribe Hominini, or hominins, as opposed to chimpanzee ancestors, or panins. The horizontal axis spreads the species out according to the relative size of their chewing teeth and brain size. Taxa with large molar and premolar crowns are to the right, and those with smaller postcanine teeth are to the left. Less speciose interpretations of the hominin fossil record do not recognise the taxa that are in bold type. The hypothetical taxa (?) are a reminder that in the relatively unexplored period between 6 and 2 myr ago the number of taxa will probably increase. Although the 2 taxa marked with asterisks are}have conventionally been assigned to Homo , it is likely that they are more closely related to Australopithecus species.
occurring at the same rate in closely-related lineages, then the degree of molecular difference can be used as a ‘ clock ’ to estimate the time elapsed since any 2 lineages separated (Sarich & Wilson, 1967). When this is done for the molecular differences between modern humans and the living African apes, it has been estimated that the human lineage separated from the rest of the hominoids between 5 and 8 myr ago (Ruvolo, 1997). A traditional classification, together with one that incorporates the taxonomic implications of the mol- ecular evidence, is given in Table 1. The new classification means that the vernacular terms we have been using to describe the human clade are no longer applicable. Thus the clade can no longer be described as containing ‘ hominids ’, for the family Hominidae has become more inclusive, and now refers to the common ancestor of the living African apes (i.e. Homo , Pan , and Gorilla ) and all of its descendants.
The appropriate vernacular term for a member of the human clade is now ‘ hominin ’, for this is the way to refer to members of the tribe Hominini, and its 2 component subtribes, the Australopithecina and the Hominina. Thus, ‘ hominid evolution ’ becomes ‘ hominin evolution ’. The vernacular ‘ hominine ’ has also taken on a more inclusive meaning, for the subfamily Homininae now includes both ‘ panins ’, the vernacular term for members of the tribe Panini containing the chimpanzees, and ‘ hominins ’, the vernacular for species in the tribe Hominini. Conse- quently, the term ‘ australopithecine ’, the vernacular for Australopithecinae, the subfamily established by Gregory & Hellman (1939) for the fossils we now allocate to Ardipithecus , Australopithecus and Paran - thropus , no longer applies. We use ‘ australopiths ’ to refer to members of the subtribe Australopithecina (Table 1 a ). Although the molecular data provide powerful
20 B. Wood and B. G. Richmond
average of 12 pairs, humans have fewer ribs than the 13 pairs typically found in African apes, and there are correspondingly fewer thoracic vertebrae (mean 12, range 11–13) in the modern human spine compared to that of African apes (mean 13, range 12–14). The human vertebral column is longer in the lumbar region, with an average of 5 lumbar vertebrae (range 4–6) compared with 3–4 lumbar vertebrae in great apes (range 3–5) (Schultz & Straus, 1945 ; Schultz, 1961). Modern humans are more similar to apes in upper limb than in lower limb morphology. Many human upper limb skeletal characteristics can be related to the loss of habitual weight-bearing function. For example, human upper limb bones are generally straighter and less robust than their great ape counterparts, and muscle insertions are typically designed for less power output (Thorpe et al. 1999), but they permit a greater range of motion, or speed. Relative to body size, the human upper limb is shorter than those of apes, but the difference in length occurs in the forearm and hand, not in the upper arm (Aiello & Dean, 1990 ; Jungers, 1994). Modern humans retain an apelike, mobile, shoulder joint with a few modi- fications, such as relatively small supraspinous and relatively large infraspinous fossae (Roberts, 1974), less cranially-oriented glenoid fossae and lateral clavicular heads (Ashton & Oxnard, 1964 ; Stern & Susman, 1983), features that are related to habitual use of the arm in lowered positions. In African apes and humans, the humeral shaft twists from the humeral head, which faces medially, down to the coronally oriented elbow joint (Evans & Krahl, 1945). Differences in elbow morphology between apes and humans are subtle (Robinson, 1972 ; Aiello et al. 1999). The human distal humerus exhibits an anteriorly oriented (rather than a distally oriented) capitulum, a shallow olecranon fossa, and weak development of the spool shape of the trochlea associated with a relatively modest lateral trochlear ridge. All these characteristics appear to be related to the loss of upper limb weight support in humans (Aiello & Dean, 1990). Great ape radii and ulnae are also more robust and longitudinally curved (Aiello et al. 1999). The most striking adaptations in the human upper limb occur in the wrist and hand, and they relate to improved manual dexterity. The human wrist is capable of more mobility in extension than those of the African apes, and it has been argued that this is an adaptation for wrist movements involved in tool making and tool use, such as hammering and throwing (Marzke, 1971). The long thumbs and relatively short,
straight fingers of the modern human hand are proportioned so that the thumb and fingers can form a precision grip, in which the broad, fleshy fingertips of the thumb and fingers are opposed in order to hold an object between them (Napier, 1961). The human thumb has a saddle-shaped carpometacarpal joint, a relatively broad metacarpal, and refined motor con- trol based on discrete, well-developed flexor pollicis longus and opponens pollicis muscles that enable independent control of the thumb and full oppos- ability (Susman, 1994) ; these 2 muscles are smaller, or absent, in African apes. Compared with apes, human manual digits have unusually broad distal phalangeal tufts and fleshy fingertips that provide a large and highly-sensitive frictional surface (Susman, 1998). Humans have shorter and straighter phalanges, unlike the long, curved proximal and middle phalanges of apes, especially the Asian apes, that improve the latter’s ability to grasp large arboreal supports and reduce the stresses associated with climbing and suspension (Susman, 1979 ; Hunt, 1991 ; Richmond, 2000). Modern human adult locomotion, unlike that of the living apes, is almost exclusively bipedal, and this is reflected in the morphology of the pelvic girdle and the lower back, knee, ankle and foot, and in the disposition of the muscles connecting the lower limb to the pelvis and trunk. The human pelvis is highly derived compared with that of the apes and other primates. Major changes in skeletal design include a craniocaudally-shortened ilium, which brings the sacroiliac joint in closer proximity to the hip joint, and sagittally-oriented iliac blades, which allows the gluteus medius and gluteus minimus muscles to be used as hip stabilisers during the stance phase of bipedal walking (Stern & Susman, 1981). The human ischium is short, with prominent ischial spines for well-developed sacrospinous ligaments that contribute to pelvic stability when standing, walking, or running. The modern human birth mechanism is unique. In nonhuman primates the sagittally-elongated pelvic inlet and outlet allow the newborn to emerge with its face ventrally, related to the pubic symphysis (Stoller, 1995). In modern humans, the pelvic inlet is broadest transversely whereas the outlet is widest sagittally. Thus the large head (Schultz, 1941 ; Jordaan, 1976) of the relatively large-bodied (Sacher & Staffeldt, 1974 ; Mobb & Wood, 1977) modern human neonate has to rotate during its passage through the birth canal (Rosenberg & Trevathan, 1995). The substantial differences between the lower limbs of modern humans and apes are largely attributable to the bipedal locomotion of the former. The most
22 B. Wood and B. G. Richmond
striking difference is the greater absolute and relative length of modern human lower limbs that increases stride length and thus the speed of bipedal walking (Jungers, 1982). Because the lower limbs support the body during bipedal gait, the acetabulum, femoral head and other lower limb joints are relatively larger in humans (Jungers, 1988 c ). Modern human femora are distinctive in that they show the valgus condition (i.e. they converge towards the knee), thus helping to position the feet closer to the midline (Walmsley, 1933 ; Tardieu & Trinkaus, 1994). The greater stresses placed on the lateral side of the knee by the valgus orientation of the distal femoral shaft are resisted by larger lateral condyles in modern human distal femora and proximal tibiae (Heiple & Lovejoy, 1971 ; Ahluwalia, 1997), and by bony buttressing beneath the tibial lateral condyle. Modern human adult femoral condyles are elongated anteroposteriorly (Tardieu, 1986, 1998) with a deep patellar groove, characteristics that increase the moment arm of the quadriceps femoris muscle, and promote the stability of the patella (Heiple & Lovejoy, 1971 ; Wanner, 1977). Lastly, the human foot shows many adaptive changes in skeletal design for bipedalism, including an adducted hallux, a longitudinal arch, long calcaneal tuberosity with a prominent lateral plantar process, and short straight toes (Susman, 1983 ; Lewis, 1989). In addition to the morphological differences be- tween apes and modern humans, there are also contrasts in the rate that their bodies grow and in the order in which structures appear during development (Schultz, 1960). Modern humans reach maturity much more slowly than do apes. They also erupt their teeth in a different order, and the milk, or deciduous, molars wear out before the adult molars have erupted (Smith et al. 1994 ; Macho & Wood, 1995). The time taken to complete tooth crown development differs between apes and humans, but these differences generally reflect differences in crown height. A major contrast between modern humans and apes is that the former have very extended periods of growth for the final stages of crown formation. It is these differences that are largely responsible for the relatively delayed crown formation, eruption, and root completion of modern humans compared with the African apes (Macho & Wood, 1995). There are many important behavioural differences between modern humans and the living apes, such as the former’s elaborate written and spoken language, but most of these behaviours leave little, or no, trace in the hard tissues that make up the hominin fossil record. Thus researchers have turned to other lines of evidence for their reconstruction, and debate is
ongoing about the extent to which these behavioural differences, especially spoken language, can be detected in the paleontological and archaeological records.
Ancestral differences
Although an impressive number of contrasts exists between the morphology of the living apes and modern humans, the differences between the earliest hominins and the late Miocene ancestors of the living great apes are likely to have been more subtle. Some of the features that distinguish modern humans and the living apes, such as those linked to upright posture and bipedalism, can be traced far into human prehistory. Others, such as the relatively diminutive jaws and chewing teeth of modern humans, were acquired more recently and thus cannot be used to discriminate between early hominins and ape ancestors. At least 2 early hominin genera, Australo - pithecus and Paranthropus , had absolutely and rela- tively larger chewing teeth than later Homo (McHenry, 1988 ; Wood & Collard, 2000). This ‘ megadontia ’ may have been an important derived feature of early hominins, but it has been reversed in later hominins. We do not yet have sufficient information about the earliest stages of hominin evolution to determine whether megadontia is con- fined to hominins, but a preliminary analysis of Miocene hominoids suggests that these are also relatively megadont (P. Andrews & B. A. Wood, unpublished data). How, then, are we to tell a late Miocene}early Pliocene early hominin from the ancestors of Pan , or from the lineage that provided the common ancestor of Pan and Homo? The presumption is that the common ancestor and the members of the Pan lineage would have had a locomotor system that is adapted for orthograde arboreality and climbing, and probably knuckle- walking as well (Washburn, 1967 ; Pilbeam, 1996 ; Richmond & Strait, 1999). This would have been combined with projecting faces accommodating elon- gated jaws bearing relatively small chewing teeth, and large, sexually-dimorphic, canine teeth with a honing system. Early hominins, on the other hand, would have been distinguished by at least some skeletal and other adaptations for a locomotor strategy that includes substantial bouts of bipedalism (Rose, 1991), linked with a masticatory apparatus that combines relatively larger chewing teeth, and more modest-sized canines that do not project as far above the occlusal plane. These proposed distinctions between hominins,
Human evolution 23
Table 2. Key to commonly - used fossil hominin site abbreviations
Site abbreviations Explanations for the site-specific prefixes used in the text
AL or A.L. L ower A wash River (Hadar in Afar Depression) ARA Ara mis Formation BC B aringo ( C hemeron Formation) BK B aringo ( K apthurin) BOU-VP Bou ri— V ertebrate P aleontology ER E ast R udolf (now usually called Koobi Fora, or sometimes East Turkana) GVH G ladys v ale H ominin HCRP RC H ominid C orridor R esearch P roject Malema HCRP UR H ominid C orridor R esearch P roject U raha KB K romdraai Site B —Fossils discovered after 1955 KGA K onso Ga rdula (now known as Konso) KNM- K enya N ational M useum (followed by the appropriate site abbreviation e.g. ER, WT etc.) KP K ana p oi KT K oro T oro, Chad LH or L.H. L aetoli H ominin MAK-VP Mak a— V ertebrate P aleontology MLD M akapansgat L imeworks D umps OH or O.H. O lduvai H ominin Omo Designation for fossils recovered by the French-led group, from the Shungura Formation, Ethiopia SE S terkfontein ‘ E xtension Site ’ SH S hungura Formation SK S wart k rans Hominin (SKW— S wart k rans W its ; SKX— S wart k rans E x cavation, refers to specimens recovered by C. K. Brain since 1965) Sts Specimens recovered from S terkfontein Type S ite between 1947 and 1949 Stw, StW, Stw}H, or StW}H S terkfontein W its Hominin—specimens recovered from any part and any member of the Sterkfontein Formation after 1968. TM T ransvaal M useum—the catalogue designation of the following : Sterkfontein—fossils discovered between 1936 and 1938 ; Kromdraai—fossils discovered between 1938 and 1955 UA U adi A alad site WT W est T urkana (including Nariokotome)
are also the most primitive (Fig. 1). Within each genus the order of presentation is such that primitive, and generally geologically older, species precede the more derived ones. Each species’ entry begins with the history of its discovery, then a list of important sites, a summary of the characteristic morphology, and its behavioural implications, available information about the paleohabitat, a summary of the hypodigm, or fossil record, for that species and, lastly, references to any current taxonomic debates involving that species. Explanations of the letter abbreviations used to identify fossils by site and locality are provided in Table 2.
Ardipithecus
Ardipithecus ramidus (White et al. 1994) White et al. 1995
The first creature to show at least some rudimentary human specialisations, and currently the most primi- tive hominin known, is Ardipithecus ramidus (White et al. 1994, 1995). The evidence is in the form of
C 4 ±5 myr-old fossils recovered in late 1992 and thereafter, from a site called Aramis, in Ethiopia. The remains have some features in common with living species of Pan , others that are shared with the African apes in general, and, crucially, several dental and cranial features that are shared with later hominins. Sites. Aramis, Middle Awash, Ethiopia ; perhaps also at Tabarin and Lothagam, Kenya. Characteristic morphology. The case White et al. (1994) put forward to justify their taxonomic judgment centres on the cranial evidence. These researchers claimed that compared with A. afarensis , A. ramidus has relatively larger canines, its first deciduous molars have less complex crowns, the articular eminence is flatter, the enamel thinner, and the upper and lower premolar crowns are more asymmetric, and thus more apelike (White et al. 1994). These workers suggested that A. ramidus should be excluded from the apes because it shares a number of derived anatomical features with later hominins, including relatively small upper central incisors, less projecting canines and a poorly-developed canine honing mechanism, broad mandibular molar crowns,
Human evolution 25
and a foramen magnum that is more anteriorly- situated than in the apes. Behavioural implications. Judging from the size of the shoulder joint, the body mass of A. ramidus was in the vicinity of 40 kg. Its chewing teeth were relatively small, and the position of the foramen magnum suggests that the posture and gait of A. ramidus were, respectively, more upright and bipedal than in the living apes. The relatively large incisors and the thin enamel covering on the teeth suggest that the diet of A. ramidus may have been closer to that of the chimpanzee than is the case for other early hominins. As yet we have no information about the size of the brain, nor any direct evidence from the limbs about the posture and locomotion of A. ramidus. The report on the remains of an associated skeleton that has been found (see below) is awaited with considerable interest. Paleohabitat. It has been reported that the remains of the plants and animals, including a large rep- resentation of extinct colobines, found with A. ramidus suggest that the bones had been buried in a location that was close to, if not actually within, woodland (WoldeGabriel et al. 1994), but the habitat and dietary preferences of fossil Colobus may not match those of extant Colobus. Hypodigm. Holotype : ARA-VP-6}1, an associated partial set of upper and lower teeth. Paratypes : ARA- VP-1}128, another set of associated teeth ; ARA-VP- 1 }4, a right humeral shaft ; ARA-VP-1}500, temporal and occipital remains ; ARA-VP-7}2, a fairly complete left humerus, radius, and ulna, as well as a number of teeth and dental fragments (White et al. 1994). Well- preserved specimens : teeth, ARA-VP-6}1 and 1}128 ; and White et al. (1995) refer to a currently unpublished associated skeleton. With hindsight, the remains from Aramis may not be the first evidence found for this species ; the mandibular fragment from Lothagam in Kenya, that has been dated to around 5 myr (Hill & Ward, 1988), may prove to be more similar to A. ramidus than to A. afarensis. Taxonomy. The new species was initially allocated to Australopithecus (White et al. 1994), but has since been assigned to a new genus, Ardipithecus , which, the authors suggest, is significantly more primitive than Australopithecus (White et al. 1995).
Australopithecus
Australopithecus anamensis Leakey et al. 1995
Fossils dating to between 3.9 and 4.2 myr found by Meave Leakey and her team at Kanapoi and Allia
Bay, in Northern Kenya, have been assigned to a new species of Australopithecus , apparently more primitive than Australopithecus afarensis (see below) (Leakey et al. 1995, 1998). Sites. Kanapoi and Allia Bay, Kenya. Characteristic morphology. Diagnostic features cited by the authors include the small size and elliptical shape of the external auditory meatus, a narrow mandibular arch with parallel mandible corpora, a sloping mandibular symphysis, long and robust canine roots, upper molar crowns that are broader mesially than distally, and a small humeral medullary cavity. A. anamensis displays a number of derived characteristics that distinguish it from A. ramidus , including absolutely and relatively thicker enamel similar to that of A. afarensis , broader molars, and a tympanic tube that extends only as far as the medial edge of the postglenoid process (Leakey et al. 1995). The main differences between A. anamensis and A. afarensis relate to mandibular morphology and details of the dentition. The mandibular symphysis of A. anamensis is steeply-sloping compared with the more vertical symphysis of later hominids, including A. afarensis. In some respects the teeth of A. anamensis are more primitive than those of A. afarensis (e.g. asymmetry of the premolar crowns, less posteriorly- inclined canine root, and the relatively simple crowns of the deciduous first mandibular molars), but in others (e.g. the low cross-sectional profiles, and bulging sides of the molar crowns) they show similarities to more derived, and temporally much later, Paranthropus taxa. Compared with A. afarensis , A. anamensis also exhibits a primitive, horizontal tympanic plate. The few known postcranial fossils preserve portions of the upper and lower limb. Contrary to earlier assessments that it is humanlike, the distal humerus of A. anamensis does not closely resemble extant humans or African apes, and instead resembles other fossil hominins, including A. afarensis , P. robustus , and Homo sp. in overall morphology (Lague & Jungers, 1996). The radius is apelike in several features, including its considerable overall length, the length of a distinct radial neck, and the well-developed brachioradialis insertion, but it lacks the pronounced shaft curvature typical of African apes (Heinrich et al. 1993). The distal end shows a mosaic of Asian ape and African ape features, resembling the former in exhibiting a relatively large articular surface for the lunate, but sharing with African apes a distally- projecting dorsal ridge, relatively coplanar scaphoid and lunate facets, and a large, dorsally-oriented scaphoid notch. The manual proximal phalanx is
26 B. Wood and B. G. Richmond
Australopithecus afarensis Johanson et al. 1978
Some half a million years after the present evidence for A. ramidus , and perhaps contemporaneous with fossils of A. anamensis , there is evidence in East Africa of another relatively primitive hominin, Australo - pithecus afarensis. This was the name given to hominin fossils recovered from Laetoli, in Tanzania, and from the Ethiopian site of Hadar (Johanson et al. 1978). When the classification of the material was first considered it was natural that researchers contem- plated its relationship to Australopithecus africanus Dart 1925, evidence of which had been recovered half a century earlier from a cave site in southern Africa (see below). The results of morphological analyses suggest that there are significant differences between the 2 hypodigms (White et al. 1981 ; Kimbel et al. 1984 ; Johanson, 1985). Support for this assessment comes from the results of cladistic analyses (e.g. Skelton & McHenry, 1992 ; Strait et al. 1997) in which they are rarely related as sister taxa (Fig. 2). Comparisons have also emphasised that in nearly all the cranial characters examined, A. afarensis displays a more primitive character state than does A. africanus (e.g. White et al. 1981 ; Kimbel et al. 1984). The fossil record of A. afarensis is best known from 3 ±4 to 3±0 myr-old sediments at Hadar, older remains are known from Laetoli in Tanzania (3±7 myr) and Fejej in Ethiopia (as old as 4±2 myr ; Kappelman et al. 1996). Thus A. afarensis is presently much better sampled than A. ramidus or A. anamensis , for it includes a skull, (Kimbel et al. 1994), substantial fragments of several skulls, many lower jaws and sufficient limb bones which allow for a reliable estimate of the stature and body mass of A. afarensis. The collection also includes a specimen that preserves just less than half of the skeleton of an adult female, whose field number is A.L.-288, but which is better known as ‘ Lucy ’. Sites. Laetolil Beds at Laetoli (originally ‘ Laetolil ’), Tanzania ; Hadar C Sidi Hakoma, Denen Dora and Kadar Hadar Members ; Middle Awash C Maka and Belohdelie ; Fejej, and Lower Omo Valley C White Sands, all in Ethiopia. Hominin fossils from Koobi Fora, Allia Bay, and South Turkwell, all in Kenya, may also belong to A. afarensis. The taxonomy of the Tabarin mandible needs to be reassessed in the light of the discovery of A. ramidus (see above). Characteristic morphology. All systematic assess- ments of A. afarensis have stressed the primitive nature of the cranium and dentition. Indeed, in their cladistic analysis of 60 cranial and dental characters, Strait et al. (1997) list just 10, the smallest number for
any of the hominins they consider, that distinguish A. afarensis from their Pan } Gorilla outgroup, and they list only 2 A. afarensis autapomorphies (Strait et al. table 4). The features that distinguish the cranium of A. afarensis from that of Pan are mainly related to the smaller canine and larger postcanine teeth of the former, and the influence the smaller canines has on the face of A. afarensis , including the reduced snout and the presence of a canine fossa. Otherwise, apart from the frontals lacking the type of supratoral sulcus seen in Pan (Kimbel et al. 1994), the pattern of ectocranial cresting in A. afarensis is Pan -like, as is the smooth transition between the nasoalveolar clivus and the floor of the nose, the shallow palate, the I#}C diastema (modest though it is), the exaggerated mastoid pneumatisation, and the weakly flexed cranial base (White et al. 1981 ; Kimbel et al. 1984). Most crania show osseous evidence of the type of occipito- marginal sinus venous drainage pattern that also occurs at a high incidence in Paranthropus (Falk & Conroy, 1983). The fossa for the mandibular condyle is apelike ; it is shallow, with little, or no, development of the articular eminence. Apart from their relatively small canines, the mandibles share with the African apes straight postcanine tooth-rows, and tall and narrow corpora with substantial hollowing on the lateral surface. Turning to the dentition, the crowns of the dm"s are intermediate between the simple cusp arrangements seen in Pan , and the more complex cusp patterns of A. africanus and Paranthropus sp. (White et al. 1994). The upper canines show the oblique wear seen in living great apes, the majority of the P$ crowns are unicuspid, and the P% crowns are more asymmetric than in more recent australopith taxa. The incisors are smaller than those of the apes, and the thick-enameled cheek teeth have larger crowns. The subocclusal morphology of the mandibular postcanine teeth is, at least among the hominins studied, distinctive in having narrow root canals and distal root components that project towards the buccal surface of the mandibular corpus, giving a serrated appearance when viewed from the lingual side (Ward & Hill, 1987). Postcranially, A. afarensis provides the first evi- dence that, with the exception of lower limb features related to bipedalism, australopiths retained a gen- erally apelike skeletal design and body shape (McHenry, 1991). Evidence from fossil rib fragments, including the apelike rounded cross-section and absence of flattening in the middle section of the body of the ribs, suggests that the rib cage of A. afarensis was capacious and retained the inverted funnel shape
28 B. Wood and B. G. Richmond
typical of great apes (Schmid, 1983). A derived trait shared with humans is the single articular facet on the first rib in A. afarensis , a feature that appears to be related to habitual orthograde posture (Stern & Jungers, 1990). The vertebrae tend to have long, apelike spinous and transverse processes, and the vertebral bodies are intermediate in size compared with the ape and human conditions. Lumbar vertebrae are wedged such that the anterior length of the body is greater than the posterior length. The upper limb of A. afarensis is shorter than a great ape of comparable mass, but long relative to humans. These differences are driven by variation in radius and ulna length, because the relative humerus length of A. afarensis is comparable to that of African apes and humans (Jungers, 1994). In the shoulder, the scapula retains a primitive cranially-oriented glenoid fossa (Stern & Susman, 1983), and the humeral head is less spherical than in apes, and resembles humans in having a relatively large lesser tubercle (Robinson, 1972). The humeral shaft may exhibit less marked torsion than in Pan or Homo (Larson, 1996), and the distal end exhibits a well-developed, Pan -like, lateral trochlear ridge, but lacks the steep lateral margin of the olecranon fossa typical of African apes. The distal humerus resembles Paranthropus humeri in exhibiting a well-developed, superiorly-positioned, lateral epi- condyle. Like A. anamensis and African apes, the distal radius of A. afarensis has a distally-projecting dorsal ridge, relatively coplanar scaphoid and lunate articular surfaces, and a large, dorsally-situated scaphoid notch (Richmond & Strait, 1999). In the hand, the pisiform is long and the fingers are intermediate in length between the long fingers of extant apes and the short ones in modern humans (Latimer, 1991), but they are longitudinally-curved as in chimpanzees and A. anamensis. The tufts on the distal phalanges are relatively narrow (Bush et al. 1982), suggesting that A. afarensis did not possess broad, fleshy fingertips. Like most apes (except Gorilla ), the pollical metacarpal is not robust (Susman, 1994). The pelvis shows a mixture of primitive and derived features. Apelike morphology includes the coronal orientation of the iliac blades, a somewhat long ischium without a raised tuberosity, a reduced acetabular anterior horn, and evidence of weakly- developed sacroiliac ligaments. However, the pelvis shares with humans a short, wide ilium, a well- developed sciatic notch and anterior inferior iliac spine, and wide sacrum. The femoral head and acetabulum, as well as sacroiliac and lower inter- vertebral joints, are small relative to humans of
comparable size (Jungers, 1988 a ). The femoral neck is long, and the cortical bone is thick inferiorly as in modern humans (Ohman et al. 1997). Although longer than that of apes, the femur is shorter than a human of similar stature (Jungers, 1982). The femur has a bicondylar angle that is even more valgus than in humans, owing to the wide pelvis and short femoral length. The feet also exhibit a mosaic morphology, including a derived adducted hallux, robust calcaneal tuberosity with a lateral plantar process, relatively short toes (compared with apes), and dorsally- oriented metatarsophalangeal joints, combined with primitive features, such as the shape of the talar trochlea, and the curvature and length (greater than humans) of the pedal proximal phalanges (Stern & Susman, 1983 ; Latimer & Lovejoy, 1989, 1990 a , b ). Behavioural implications. To judge from the size of the postcranial remains, the species ranged in body mass from C 25 kg, for a small female, to C 50 kg for a large presumed male (Jungers, 1988 b ; McHenry, 1992). The suggestion that A.L. 288-1, one of the smallest A. afarensis individuals, may be a male (Ha$usler & Schmid, 1995), which would strengthen the case for taxonomic heterogeneity, has been effectively refuted (Wood & Quinney, 1996 ; Tague & Lovejoy, 1998). Stature estimates suggest a range of C 1 ±0–1±5 m. The estimated brain volume of A. afarensis is between 375 and 540 cm$, with a mean of c. 470 cm$. This is larger than the average brain size of a chimpanzee, but, if the estimates of the body size of A. afarensis are anything like correct, then, relative to estimated body mass, the relative brain size of A. afarensis is not much greater than that of Pan. It has incisors that are smaller than those of extant chim- panzees, but the chewing teeth—the premolars and molars—of A. afarensis are relatively larger than those of Pan (McHenry, 1988). The thick enamel of the A. afarensis cheek teeth suggest that nuts, seeds, and hard fruit may have been an important com- ponent of the diet of this species. The shape of the pelvis and the lower limb suggests that A. afarensis was adapted to bipedal walking. This indirect evidence for the locomotion of A. afarensis is complemented by the discovery, at Laetoli, of several trails of fossil footprints (Leakey & Hay, 1979). These provide very graphic, direct, evidence that A. afarensis , or another contemporary hominin, was capable of bipedal locomotion. The size of the footprints, and the length of the stride, are consistent with stature estimates based on information from the limb bones of A. afarensis. These suggest that the standing height of the individuals in this early hominin species was between 1 m and 1±5 m (Jungers, 1988 a ).
Human evolution 29
‘ africanus Weinert 1950 ’ suppressed. The results of the deliberations were published as ‘ Opinion 1941 ’ in the Bulletin of Zoological Nomenclature (ICZN, 1999). In it the ICZN confirmed that ‘ africanus Weinert 1950 ’ be suppressed so that if it is to be removed from Australopithecus , the A. afarensis hypodigm should be referred to as Praeanthropus afarensis.
Australopithecus bahrelghazali Brunet et al. 1996
Hominin fossils collected in Chad, in North-central Africa, and faunally-dated to C 3 ±5 myr (Brunet et al. 1995), have been assigned to A. bahrelghazali. They extend the known geographical range of fossil hominins far beyond East and southern Africa (Wood, 1995). The discovery of these fossils under- scores how little we currently know about the ranges of extinct hominin species and the biogeographical history of hominin evolution (Foley, 1999 ; Strait & Wood, 1999). Site. Bahr el ghazal region, Chad, North-central Africa. Characteristic morphology. The published evidence, a mandible and a maxillary premolar tooth, has been interpreted as being sufficiently distinct from A. ramidus , A. afarensis and A. anamensis to justify its allocation to a new species. Brunet et al. (1996) claim that the thickness of its enamel distinguishes the Chad remains from A. ramidus , and that the more vertical orientation and reduced buttressing of the mandibular symphysis, together with the more symmetric crowns of the P$, separates it from A. anamensis. The complexity of the mandibular premolar roots is the main feature that distinguishes A. bahrelghazali from A. afarensis (but see below), and its more slender corpus, larger incisors and canines and more complex mandibular premolar root system separate it from A. africanus. Behavioural implications. At present little can be said about the behaviour of A. bahrelghazali other than that its similarity to A. afarensis in terms of enamel thickness and dental morphology suggests that the 2 taxa shared a similar diet (e.g. fruit, nuts, and seeds). Paleohabitat. Associated fauna reflect both open and wooded habitats. The remains of some aquatic taxa indicate the presence of a river, or riparian woodland. Thus the paleohabitat of A. bahrelghazali is consistent with that of australopiths from East and southern Africa. Hypodigm. Holotype : KT 12}H1, anterior man- dible. Paratype : KT 12}H2, right P$. Taxonomy. In a recent paper White et al. (2000)
claimed that a complex P$ root system is also seen in a percentage of A. afarensis specimens, and thus it cannot be used to distinguish A. bahrelghazali. Australopithecus africanus Dart, 1925
In 1924, nearly 50 years before the discovery of the East African remains belonging to A. afarensis , an early hominin child’s skull was found among the contents of a small cave exposed during mining at the Buxton Limeworks at Taungs (the name was changed later to Taung) in southern Africa. To judge from the fossil mammals found with it, the Taung hominin was more ancient than any of the hominin remains that had been recovered in Europe, Java or China (see below). The new hominin was described by Raymond Dart, who referred it to a new genus and species, Australopithecus africanus , literally the ‘ southern ape of Africa ’ (Dart, 1925). Dart referred to postcranial remains in his description of the material, but only the skull survives. No other australopiths have been recovered from the Buxton Limeworks. Given the difficulties of assessing a juvenile speci- men, Dart’s analysis of the Taung was remarkably perceptive, for he claimed it was an example of an ‘ extinct race of apes intermediate between living anthropoids and man ’ (ibid, p. 195). This judgment depended heavily on Dart’s interpretation of the relative size of the face, and his conclusion, based on the height of the canine crown and the small size of the gap, or ‘ diastema ’, between the incisors and canine, that the dentition ‘ is humanoid rather than an- thropoid ’ (ibid, p. 196). He also cited the relatively robust mandibular corpus and the vertical and unbuttressed symphysis as further evidence of the Taung child’s human affinities. It is noteworthy that Dart explicitly contrasted the humanoid nature of the Taung symphysis with that of the Piltdown jaw, noting that the symphysis of ‘ Eoanthropus dawsoni scarcely differs from the anthropoids ’ (ibid, p. 197). Dart related the foramen magnum to prosthion, anteriorly, and inion, posteriorly, in a ‘ … head- balancing index … ’. The value for Taung, 60.7, was intermediate between the value for an ‘ adult chim- panzee ’, 41.3, and ‘ Rhodesian man ’, 83±7 (ibid, p. 197). Lastly, Dart interpreted the relatively posterior location of the lunate sulcus as evidence of expansion of the ‘ parietal region ’ of the brain (ibid, p. 198). Since the discovery at Taung, the remains of hominins we now classify as A. africanus have been found at 3 other cave sites in southern Africa. At all these cave sites, as at Taung, early hominin fossils are mixed in with other animal bones in rock and bone- laden, hardened, cave fillings, or breccias. The cave at
Human evolution 31
Fig. 3. Location of cave sites in and around the Blauuwbank Valley, South Africa.
Sterkfontein (Fig. 3) yielded its first hominin fossils in 1936, with further specimens being recovered in 1937 and 1938. When Robert Broom announced the discovery of the cranium TM 1511 in 1936, he expressed the opinion that the new cranium ‘ probably agrees fairly closely with the Taungs ape ’, but he went on to state that ‘ … it advisable to place the new form in a distinct species, … .’ (Broom, 1936 b ). He subse- quently gave it the name Australopithecus transvaalensis (Broom, 1936 a ), but transfered it to a new genus, as Plesianthropus transvaalensis , some 2 years later (Broom, 1938), by which time mandibular (e.g. TM 1515) and postcranial (e.g. TM 1513) evidence had come to light. Excavations at Sterkfontein were held in abeyance until 1947, when Broom and John Robinson restarted them. To date, Sterkfontein has yielded a collection of more than 600 Australopithecus remains, most of them coming from Member 4 (but see below). The first evidence of fossil hominins from Makapansgat, another southern African cave site, was the calvarium MLD 1, found in 1947. Raymond Dart allocated it to a new species, and gave it the name Australopithecus prometheus (Dart, 1948) be- cause he believed that the Makapansgat hominin was capable of making fire. Hominin fossils continued to be recovered from Makapansgat until the early 1960s. In 1951 Sherwood Washburn, a primatologist, and Bryan Patterson, a paleontologist, wrote a joint letter to ‘ Nature ’ suggesting that the taxonomy of the Taung, Sterkfontein and Makapansgat hominins be rationalised, and their proposal received influential support from Sir Wilfrid Le Gros Clark (1955) in his
monograph ‘ The Fossil Evidence for Human Evol- ution ’. Thereafter it became conventional to refer all the ‘ gracile ’ remains from southern Africa to a single genus, Australopithecus , and it was not long before researchers and commentators carried the process of rationalisation a stage further by subsuming A. transvaalensis and A. prometheus into the species of Australopithecus with taxonomic priority, namely A. africanus Dart, 1925. The third site to yield the remains of A. africanus is Gladysvale (Fig. 3). Broom collected fossils there in 1936, but the first hominins, 2 teeth (referred to as GVH 1 and 2 in Berger et al. 1993, but as GVH-7 in Berger & Tobias, 1994) and a phalanx (GVH-8) were recovered nearly 60 years later, in 1991. Until recently (see Partridge et al. 1999), the cave sites in southern Africa could only be dated by comparing the remains of the mammals found in the caves with the mammalian fossils found at the better- dated sites in East Africa. In this, and in other ways, the ages of the A. africanus -bearing breccias have been estimated to be between 2±4 and 3 myr. Claims for a substantially earlier age for Member 2 (Clarke & Tobias, 1995 ; Clarke, 1998 ; Partridge et al. 1999) have been challenged (McKee, 1996). Sites. Taung (D-C), Sterkfontein (Member 4, and probably Member 2, but see below), Makapansgat (Member 3), Gladysvale, all in South Africa. Characteristic morphology. The differences between A. africanus and A. afarensis are set out in detail in White et al. (1981) and Johanson (1985). Cranially the main contrasts are in the A. africanus face, which is broader and less prognathic than in A. afarensis. The mandibles of A. africanus have more robust corpora than those of A. afarensis. The main difference in the teeth is that, relative to A. afarensis , the anterior teeth are reduced in size and the postcanine teeth enlarged in A. africanus. Aside from these differences the crown of the dm" is more complex in A. africanus than in A. afarensis. In most respects, the postcranial skeleton of A. africanus resembles A. afarensis (McHenry, 1986), but there are a few important differences. First, the limb proportions of A. africanus may be less modern humanlike than those of A. afarensis and A. anamensis (McHenry & Berger, 1998). The lower vertebral column known for A. africanus shows that it possessed 6 functionally-defined lumbar vertebrae, more than the 5 typical of modern humans and 3–4 characteristic of great apes. In this way, it resembles Homo ergaster , and suggests that 6 lumbar vertebrae is the primitive condition for hominins. The suggestion that the A. africanus tibia is more chimpanzee-like (Berger &
32 B. Wood and B. G. Richmond
Ardipithecus ramidus , and an important component of the hypodigm of A. afarensis (see above). However, it was 2±5 myr-old hominin fossils (Asfaw et al. 1999) recovered from localities within the Hatayae (ab- breviated to ‘ Hata ’) Member of the Bouri Formation (de Heinzelin et al. 1999), C 30 km to the south of the aforementioned sites, that prompted the recognition of another new australopith taxon. The new species is based on cranial fossils of which the best-preserved is the holotype, BOU-VP-12, from Locality 12. Sites. Bouri, Middle Awash, Ethiopia. Characteristic morphology. The taxon combines a relatively primitive cranium with canines larger than those of A. afarensis , and large-crowned postcanine teeth, especially premolars, that, despite the small size of the Bouri cranium, are as large as those of Paranthropus boisei (see below). However, unlike any Paranthropus species A. garhi possesses a relatively large anterior dentition and its postcanine teeth lack the extreme enamel thickness seen in Paranthropus. The authors of the paper announcing the new species claim that the cranium lacks the derived features of Paranthropus , and suggest that its face, palate and subnasal morphology are more primitive than that of A. africanus and Homo. The essentially primitive nature of A. garhi is suggested by the results of a recent cladistic analysis (Strait & Grine, 1999). Although an associated skeleton, BOU-VP-12}1 A- G, has been recovered from an equivalent horizon, at a nearby locality, the discoverers of both this and the type specimen of A. garhi have resisted making the assumption that the skeleton and the cranium belong to the same species. The skeleton represents the first evidence of femur elongation in the hominin fossil record. However, this individual also exhibits a forearm that is as long or longer, relative to its humerus, as the upper limbs of Pan , A. afarensis and probably A. africanus , and contrasts with that of Homo ergaster (see below). Behavioural implications. Behavioural implications have not yet been discussed in the literature, but the elongated femur suggests anatomical refinements related to bipedalism. However, the retention of long arms and a very high brachial index suggests that arboreality was also a significant component of the locomotor repertoire of whatever taxon is represented by the associated skeleton. Cut-marks on animal bones found at nearby localities suggest that A. garhi , or another contemporary hominin not yet found in the Bouri region (e.g., H. [or A .] rudolfensis or P. aethiopicus ), was exploiting mammalian carcasses as a source of meat. Paleohabitat. The fossil cranium was recovered
from sediments laid down on a floodplain crossed by channels making their way to a lake that fluctuated in size. The antelopes and pigs found from horizons similar to those yielding the hominins suggest a mixed, open woodland, paleohabitat (de Heinzelin et al. 1999). Hypodigm. Holotype : BOU-VP-12}130, a cranium (N.B. the field number given in the formal description [Asfaw et al. 1999] of the holotype, ARA-VP-12}130, is a misprint ; see erratum note in Science , 284 , p. 1623), Bouri, Middle Awash, Ethiopia ; Paratypes : none. Taxonomy. The announcement of A. garhi implied that it is the ancestor of Homo , but its morphology is consistent with other interpretations. For example, it could represent the sister-taxon of a clade comprising A. africanus , Paranthropus , and Homo (Strait & Grine, 1999). At present, the relationships of A. garhi are unresolved, and will remain so until researchers can determine which aspects of its morphology are synapomorphic and which are homoplasic.
Paranthropus
Just as there are East and southern African variants of the so-called ‘ gracile ’ australopiths, there are also regional variants of another type of hominin that many now assign to a separate genus, Paranthropus. They are often referred to as ‘ robust ’ australopiths because of their relatively massive faces and lower jaws.
Paranthropus robustus Broom, 1938 and Paranthropus crassidens Broom, 1949
Remains of Paranthropus robustus come from southern African cave sites, and are dated to between C 1 ±9 and C 1 ±5 myr. The type specimen, an adult, presumably male, cranium, TM 1517, was recovered in June, 1938, at Site B of a cave called Kromdraai, and was announced and described in the same year (Broom, 1938). Kromdraai, like the caves of Swartkrans, Drimolen (see below), and Sterkfontein (see above), is in the Blaaubank Valley (Fig. 3). Subsequent discoveries were made at Kromdraai in 1941 (TM 1536), 1944 (TM 1603) and then again in the middle 1950s. Fossils found in excavations carried out in the 1970s have brought the number of hominin fossils recovered from Kromdraai to close to 20, sampling a minimum of 6 individuals (Vrba, 1981). Recent excavations in the cave have recovered a deciduous molar, KB 5503 (Thackeray, pers. comm.). The first hominin, SK 6, was recovered from
34 B. Wood and B. G. Richmond
Swartkrans in 1948 and was reported a year later (Broom, 1949). Three years of intensive excavation of Member 1 resulted in a rich collection of hominin remains. Hominins attributed to P. robustus have since been recovered not only from Member 1, but also from the Member 1 } 2 interface and from Members 2 and 3 (Brain, 1993, 1994). Nearly all of the research on the interpretation of how the various types of breccia entered the Swartkrans cave has been carried out by C. K. (Bob) Brain. It was also due to his efforts that the role played by predators in the accumulation of the fossil bones in the southern African cave sites was established (Brain, 1993). More recently, P. robustus -like hominins have been recovered from the sites of Drimolen and Gondolin (Fig. 3). The Drimolen site was discovered in 1992 and has already yielded 49 fossil hominins, the vast majority of which are referable to P. robustus. Gondolin was excavated by Vrba in 1979 (Watson, 1993), and the faunal remains now include 2 Paran - thropus teeth, GDA 1 and 2 (Menter et al. 1999). Clarke (1994) reported the discovery of 3 P. robustus - like teeth, including a lower molar (StW 566) and an upper incisor and canine, during recent excavations in Member 5 at Sterkfontein. Sites. Kromdraai B, Swartkrans (Members 1–3), Drimolen, Gondolin, and possibly Sterkfontein (Member 5), all in South Africa. Characteristic morphology. The brain, face and chewing teeth of P. robustus are larger than those of A. africanus , yet the incisor and canine teeth are smaller. The postcanine teeth, like those of P. aethiopicus and P. boisei , have thick enamel. The cranium has ectocranial crests, and the cranial base is more flexed than in A. africanus. The cranial capacity has recently been reassessed to C 475 cm$^ (Falk et al. 2000). It also shares with P. boisei (see below) and A. afarensis a tendency for the intracranial venous blood to drain through a supplementary occipitomarginal system of dural sinuses. Some authors treat this evidence as strong support for a Paranthropus clade (Falk & Conroy, 1983), but others are less inclined to treat it as a phylogenetically-valent trait (Kimbel, 1984). There are quite a few postcranial fossils from Kromdraai and, especially, Swartkrans that probably belong to P. robustus. The uncertainty stems from the fact that craniodental remains of both Paranthropus and Homo cf. erectus have been recovered from the lower members of Swartkrans (Susman, 1988 b ; Trinkaus & Long, 1990). However, because over 95 % of the craniodental fossils are attributable to P. robustus , it is inferred that most of the postcranial
remains probably belong to this taxon (Susman, 1988 b ). With this caveat in mind, the postcranial skeleton of P. robustus retains some primitive features, but in many ways it is remarkably modern humanlike. The distal humerus resembles modern humans in its articular morphology, and the dorsal margin of the distal radius does not project distally as in the knuckle- walking African apes (Susman, 1988 b ; Grine & Susman, 1991). Hand fossils from Swartkrans show a number of derived humanlike features, including a broad pollical metacarpal head, straight-shafted man- ual proximal phalanges with relatively weak flexor sheath markings, and a pollical distal phalanx with a broad apical tuft with spines, and large insertion for a strong flexor pollicis longus muscle. The pelvis and hip joint resembles the morphology of A. afarensis and A. africanus , but the iliac blade is wider and the acetabulum, femoral head and sacral articular surface are smaller (McHenry, 1975). The femur shares with P. boisei and H. habilis femora an anteroposteriorly- flattened neck, and the cortical bone of the proximal femoral shaft of P. robustus is thick, and lacks the mediolateral buttressing seen in H. erectus (Ruff et al. 1999). In the foot, the hallucal metatarsal is strikingly humanlike, with an expanded inferior base, and dorsally-extended distal articular surface (Susman, 1988 b ). Behavioural implications. Average body size esti- mates for P. robustus males (C 40 kg) and females (C 32 kg) suggest substantial sexual dimorphism. Cranial and dental differences between the taxa have led to the suggestion that the diet of P. robustus differed from that of A. africanus. Evidence from studies of dental microwear indicate that P. robustus ate foods that were substantially harder (Grine, 1986), but which considering the small size of their incisors, coupled with the relatively low microwear feature density (Ungar & Grine, 1991), may have required less incisal preparation. Stable isotope analysis of P. robustus tooth enamel suggests that its diet included substantial components of C-4 foods (Lee-Thorp et al. 1994), including grasses, sedges, some tubers, and the animals that eat these plants (Koch et al. 1994). Brain (1994) interprets these data as indicating that P. robustus ‘ were generalized rather than specialized feeders ’ (ibid, p. 222). Wear on bone tools found in the same breccia is consistent with digging, possibly for buried food items such as roots and tubers (Brain, 1988). The similarities in hip and pelvic morphology with A. afarensis and A. africanus suggest that the gait of P. robustus probably resembled that of the ‘ gracile ’ australopiths (Macchiarelli et al. 1999). These simi-
Human evolution 35
Paranthropus boisei (Leakey, 1959) Robinson, 1960
The first evidence of an East African species of hominin resembling P. robustus , 2 deciduous lower teeth, OH 3, a canine and a molar, was found in 1955 at Olduvai Gorge, in Tanzania (Leakey, 1958). The type specimen of the new species, OH 5, a magnificent, undistorted, cranium with a well-preserved dentition, was recovered in July, 1959 (Leakey, 1959). The open sutures, the partially-erupted M$s, and the well- developed sagittal crests point to the cranium being that of an immature male. The new species was initially included in a new genus, Zinjanthropus (Leakey, 1959), but subsequent taxonomic reviews resulted in it being relegated to a subgenus (Leakey et al. 1964), and 3 years later it was proposed that any generic distinction between Zinjanthropus and Australopithecus should be abandoned (Tobias, 1967). It is now usual to refer to the taxon as Australopithecus boisei , or Paranthropus boisei (see also Robinson,
- ; the latter implies a sister group relationship between the ‘ robust ’ australopiths from East and southern Africa (see below). A fragmented cranium (OH 30) and several isolated teeth (OH 3, 26, 32, 38, 46 and 60) from Olduvai have been assigned to the same species. An ulna (OH 36) may also belong to P. boisei (Aiello et al. 1999). Further evidence of P. boisei emerged with the discovery in 1964 of a mandible with a large, robust body, large premolar and molar crowns and small incisors and canines, at the Peninj River, on the shores of Lake Natron, in Tanzania (Leakey & Leakey, 1964 ; Tobias, 1965). Thereafter, some cranial, man- dibular, but mostly dental, remains were recovered from the Shungura Formation in Ethiopia (Howell & Coppens, 1976 ; Coppens, 1980) ; an incomplete cranium (Carney et al. 1971), and cranial and tooth fragments (summarized in Wood, 1999 b ) were found at Chesowanja, in the Chemoigut basin, to the East of Lake Baringo, in Kenya ; and the first skull (KGA 10- 525), and 8 other C 1 ±4 myr-old specimens of P. boisei , were recovered from the site of Konso, in Ethiopia (Suwa et al. 1997). More details of all but the Konso specimens are given in Wood (1991, p. 27 and thereafter). Most recently a poorly-preserved maxil- lary fragment (HCRP RC 911) was recovered from the Chiwondo Beds, at Malema, Malawi (Kullmer et al. 1999). However, the site collection that provides the most comprehensive evidence about P. boisei is that from Koobi Fora, which is on the north-eastern shore of Lake Turkana. The Koobi Fora evidence includes crania, partial crania, many mandibles and
isolated teeth, and is described in detail in Wood (1991). Sites. Olduvai Gorge and Peninj}Natron, in Tanzania ; Shungura Formation, Omo Region and Konso Gardula, in Ethiopia ; Koobi Fora, Baringo Region, and West Turkana, in Kenya ; and Malema, in Malawi. Characteristic morphology. The features that set P. boisei apart are to be found in the cranium, mandible and dentition. Cranially, it is the only hominin that combines a massive, wide, flat, face with a modest- sized neurocranium (C 450 cm$). The face of P. boisei is larger and wider than that of P. robustus , yet its brain volume is the same, or smaller. Some features are apparently unique to P. boisei , such as the complex, overlapping parietotemporal suture, and others, such as the dominance of the occipitomarginal venous sinus system for draining blood from the base of the brain, are shared with other taxa. The flexed cranial base seems to be uniquely organised, with the foramen magnum situated relatively far forward for a hominin with a modest brain size. The articular region of the temporal bone combines a relatively deep, laterally-extensive fossa for the condyle of the man- dible, a pronounced articular eminence and virtually no preglenoid planum (this morphology contrasts with the more primitive mandibular fossa of P. aethiopicus , see below). The mandibles have a larger and wider corpus, than any other hominin. The dentition combines very large-crowned, broad-based and thick-enameled premolar and molar teeth (Wood et al. 1983) with small anterior (i.e. incisor and canine) teeth. The tooth crowns apparently grow at a faster rate than in any other early hominin (Beynon & Wood, 1987). The morphological differences between the southern and the East African forms of Paranthropus are listed in Wood (1991, pp. 258–268, Tables 2±8 and 2±9). Despite the richness of the cranial evidence for P. boisei , there are no postcranial remains that can, with certainty, be ascribed to that taxon. Individual postcranial bones and a partial skeleton from Koobi Fora have been linked with the taxon (Grausz et al. 1988 ; Walker et al. 1989), but the evidence for doing so is far from conclusive (Wood, 1991, p. 182). The partial skeleton is characterised by limb proportions that resemble A. afarensis (Grausz et al. 1988), and are less apelike than those of A. africanus. Behavioural implications. The picture that emerges from the fossil evidence is that P. boisei was a markedly sexually-dimorphic hominin, with the esti- mated average body mass of presumed males (C 50 kg) being much greater than the mass of females
Human evolution 37
(C 34 kg) (McHenry, 1992). The absolutely and relatively small canines mean that if there was intra- male competition for females, then the males used other means to signal threats. The estimated cross- sectional areas of the mandibular corpora are between 2 and 3 times larger than expected for a hominoid of that body size. The large-crowned, thick-enameled, chewing teeth and the large mandibles with wide bodies, have conventionally been interpreted as evidence that the diet of P. boisei was a highly specialised one, devoted to eating seeds or fruits with hard outer coverings. It may be that this is entirely wrong, and P. boisei might have been the higher primate equivalent of a bushpig. In other words although its morphology is specialised, its large teeth and mandibles probably enabled it to cope with a wide range of dietary items, except that its jaws and teeth would have been ill-equipped to slice, or tear, raw meat. If the partial skeleton, KNM-ER 1500, belongs to P. boisei , the limb proportions show that this hominid possessed hindlimb elongation indicative of bipedalism, like that in A. afarensis (Grausz et al. 1988). However, like A. afarensis , the relatively long forelimbs suggest that the locomotor behaviour of P. boisei included an arboreal component. Paleohabitat. Shipman & Harris (1988) suggested that P. boisei specimens were more likely to be found in closed habitats, but a more recent analysis found that P. boisei remains are most commonly associated with relatively open habitats associated with grass- land, including open woodland and scrub woodland, close to a water source (Reed, 1997). Hypodigm. Holotype : OH 5, adolescent cranium found at site FLK, Bed I, Olduvai Gorge, Tanzania. Well-preserved specimens : skulls—KGA 10-525 ; crania—KNM-ER 406, 407, 732, 13750, 23000, KNM-WT 17400 ; mandibles—Peninj 1, KNM-ER 729, 3230, 15930. Taxonomy. Researchers have suggested that the hypodigm of P. boisei may display more variation than can be accommodated within one species (Dean, 1988). However, some of the apparently excessive variation in size is due to taphonomic factors, and the residue does not exceed the variation observed in living higher primate taxa (Silverman et al. unpublished).
Paranthropus aethiopicus (Arambourg & Coppens,
- Chamberlain & Wood, 1985
The earliest East African fossil evidence for ‘ robust ’ australopiths is interpreted by some researchers as
being taxonomically distinct from the main P. boisei hypodigm. One of the oldest of the ‘ robust ’ mandibles recovered from the Shungura Formation was made the holotype of a novel species and genus, Paraustralopithecus aethiopicus (Arambourg & Coppens, 1968), and when a distinctive 2±5 myr-old cranium (Walker et al. 1986) was recovered from sediments at West Turkana it was natural to consider whether it should be assigned to the same taxon. Suwa (1988) has pointed out that the pre-2±3 myr-old dental remains are not as derived as the bulk of the P. boisei sample that is younger than 2±3 myr, and he has suggested that these differences might warrant sep- arate taxonomic recognition. Wood et al. (1994) found that several features of the mandible and the mandibular dentition of the East African Paran - thropus lineage change around 2±3 myr-ago, and they supported the interpretation that the ‘ early ’ and the ‘ late ’ stages of the robust lineage in East Africa should be recognised as different taxa, with the former being referred to as Paranthropus aethiopicus. Sites. Shungura Formation, Omo Region, in Ethiopia ; Nachukui Formation, West Turkana, Kenya. Characteristic morphology. P. aethiopicus has a more primitive cranial vault and base, including a shallow articular fossa, and a low articular eminence continuous with a flat preglenoid planum, along with a more prognathic face, larger incisors and a less- flexed cranial base than P. boisei. No postcranial remains are currently known for this taxon. Behavioural implications. The larger incisors suggest that these teeth played more of a role in feeding and food processing than is the case for P. robustus and P. boisei. Paleohabitat. Associated faunas suggest that the habitat of P. aethiopicus was ‘ more closed ’ than that of P. boisei (Reed, 1997). Hypodigm. Holotype : Omo 18±18 (or 18± 1967 ±18), edentulous adult mandible, locality Omo 18, Section 7, Member C, Shungura Formation, Omo Region, Ethiopia. Well-preserved specimens : cranium— KNM-WT 17000 ; mandible—KNM-WT 16005. Taxonomy. (See above.)
Homo Homo habilis Leakey et al. 1964
In 1960, a year after the discovery of the type specimen of P. boisei , OH 5, from Bed I at Olduvai Gorge, Louis and Mary Leakey recovered substantial parts of both parietal bones, ‘ a large part of a left
38 B. Wood and B. G. Richmond
