Phylogenetic Impact of KNM-WT 17000: A Discussion of the Hypotheses Existing Prior to, Reformulated, and Postulated as a Result of the Unearthing of the 'Black Skull'
by Kimberly Oliveir (2002)

By the 1980s, analyses of excavated sites had been improved through the application of increasingly sophisticated geochronological and palaeoecological frameworks, with the aid of absolute (radiometric) dating. For example, in 1976 it was reported that, "Undatable fossiliferous horizons have in some instances been correlated through careful biostratigraphic studies with sites of known age" (Johanson 1980: 33). The sites of both South and East Africa were considered savanna mosaic and not forest (Reichs 1983: 209). The most common hominid taxonomic construction at this time consisted of a two-pronged fork in a linear progression. The Homo genus progression through time led from the species of H. habilis to H. erectus to H. sapiens, while the Australopithecus genus went from A. africanus to A. robustus to A. boisei.

In the late sixties and seventies, J.T. Robinson, allotted Australopithecus robustus and Australopithecus boisei, often referred to as the 'robust' lineage, to the scientific genus name of Paranthropus, according to the morphological aspects of the mastication trend, to distinguish it from the Homo lineage. His rational was as follows: "Since A. robustus and A. boisei comprise a distinct evolutionary branch (clade), they should be accorded distinct generic separation (Paranthropus)" (Tattersall 1988: 72). A crucial part in his argument for the disregard of the Australopithecus genus name was that the South African "specimens from Taung, Sterkfontein, and Makapansgat could be accorded membership in the genus Homo as the species H. africanus" (Tattersall 1988: 72). But since this genetic distinction creates a polyphyletic (a group of species with more than one common ancestor) and thus invalid group, the use of the name continues to be precarious today. Here it will be used along with the distinction of Australopithecus in recognition of all existing hypotheses.

The morphologically identical Hadar and Laetoli sites in Ethiopia, East Africa, were heavily focused upon by paleoanthropologists during the 1970's. The yeilds included specimens of Homo habilis, Australopithecus/Paranthropus boisei and the then new and equivocal species termed Australopithecus afarensis. Todd Olson was one professional in the field who did not agree with the new species distinction because he saw anatomical evidence for two separate species in A. afarensis. He proposed that all the supposed A. afarensis specimens from Laetoli were morphologically similar to the Homo lineage while equally classified specimens from Hadar belonged in the Paranthropus or Homo line (diagram below). Ian Tattersall points out that even though Olson based his hypothesis on Robinson's, "Robinson recognized the 'Garusi' or Laetoli fossils discovered in the 1930's to be part of the Homo line" (1988: 72) while Olson placed the same specimens in the Paranthropus lineage. This not only reveals the contention about the fossils themselves, but the discrepancies underlying the theory itself. The phylogenetic scheme originally postulated by Robinson and altered by Olson will be referred to as Hypothesis Three.

After a period of disagreement, at the 1978 Royal Swedish Academy of Sciences Nobel Symposium, the paleoanthropological community in general recognized an extremely sexually dimorphic species of Australopithecus, afarensis, in Hadar and Latolil rather than two species, as Olson maintained. This fossil species retains primitive traits (plesiomorphies) which suggest a late divergence of the hominids from the chimpanzee clade, pongidae. A result of anatomical and behavioral evidence and the protein structure and DNA analysis conducted in the mid-seventies, it was "deduced that man chimpanzee and gorilla diverged from a common ancestor as recently as five of six million years ago" (Reichs 1983: 209). (Note: Under the latest proposal (Wood, B. & Richmond, B. 2000. Human evolution: taxonomy and paleobiology. Journal of Anatomy 196: 19-60), pongidae is a Family including the genus Pongo, the genus Gorilla and the genus Pan.) Thus, many new theories were postulated in regard to the construction of hominid phylogeny based on the new means of evidence.

According to Donald Johanson, A. afarensis was ancestral to both the remainders of the Australopithecus lineage and the Homo lineage. I will refer to his hypothesis, also advanced by Tim White (Johanson, D. & White, T. 1979. A systematic assessment of early African hominids. Science 202: 321-330), as Hypothesis One (diagram below). While the Homo lineage experienced encephalization and dental size reduction as time progressed, the remainder of the Australopithecus lineage, consisting of A./P. robustus then A./P. boisei, experienced "increasing specialization in the masticatory apparatus and an apparent lack of brain expansion (which) led ultimately to extinction (Johanson 1980 64)." A. africanus was viewed to be the anatomically intermediate link between A. afarensis and A./P. robustus. At some point between two and three million years ago, A. africanus and H. habilis diverged. As per oldest dated evidence at the time, the Homo lineage evolved in East Africa and the Australopithecus lineage in Southern Africa. Johanson was sure to repeatedly point out that more specimens from East Africa will have to be uncovered to support or oust his theory.

The Single Species Hypothesis, advocated in the 1960's by C. Loring Brace and Milford Wolpoff, viewed a 'characteristically human' behavior: "It was still clearly 'culture', in its tool-using/tool-making manifestation, that translated 'ape' into 'human' through several 'pre-human' phases" (Pilbeam 1980: 267). Defining the means in terms of the ends in this manner (tautology) is generally incorrect and unacceptable, but the fossil evidence did not prove the theory in either direction. The splitting method of classification (treating anatomical differences as inter-species variation) became more common partially in recognition of the lumping tendency to neglect probable species, but either extreme is precarious. The Single Species Hypothesis attributed all anatomical differences to intra-species variation. In this model, only one hominid species could exist at any one time (according to the ecological principle of competitive evolution), resulting linear progression for hominid evolution. The major issue with the idea of gradual evolution is "where do you draw the line and say one species turns into another?" (Leakey & Lewin 1992: 129) Nonetheless, the Single Species Hypothesis was still considered plausible into the 1980's, although the majority of paleoanthropologists began to accept the idea of a multi-branched hominid phylogenic tree. This issue exemplifies how "paleoanthropological data is susceptible to a rather broad range of interpretations" (Pilbeam 1980: 268). Yet the larger and more complete the fossil sample is, the more the specimens can be treated as if extant in reconstructing the phylogeny.

According to the table constructed by David Pilbeam (1980: 280), diet is directly related to climate. The increasing progression of brain and postcanine teeth size through the Hominid lineage, with a reversal in cheek teeth size in later 'Hominine' as a result of the rise in technology. While he definitely attributes dental adaptations to dietary pressures, he tentatively accepts Sacher's 1974 proposal that and "increase in brain size might imply a correlated lengthening of life history processes (1980: 281)." Parameters such as time of maturation, gestation and life have been proven to be relatively associated to body and brain size. Frank Spencer (1997: 150) states that the 1980's provided innovated methods of obtaining information from the fossil record regarding growth and development rate patterns (life history parameters). For example, "studying bone remodeling and tooth development with the use of the scanning electron microscope" (Spencer 1997: 150). Pilbeam states the Hadar and Laetoli specimens suggest that "'hominid' dentitions evolved in non-bipedal and probably non-tool-using creatures" by 3.5 million years ago (with an upper limit of 6 million years ago). The conclusion is that behavior according to diet rather than culture is the driving force in adaptations for early hominid evolution.

Yoel Rak, seemingly confirmed Pilbeam's support of diet being related to climate by exhibiting "how the steady rise in bulkiness of the face through the…Australopithecine line could be explained as graded architectural reinforcement required by increased chewing forces" (Lewin 1986: 721). His analyzation of the face topography supported that A. africanus is the progenitor of A./P. robustus and A./P. boisei, on the basis of mastication specializations increasing through time, while A. afarensis is the primitive-trait-holding (plesiomorphic) common ancestor of both the Homo and ''robust'' lineages. By 1.5 million years ago, Rak postulates that two hominid taxa occupied the same African landscape but specialized in exploiting different ecological niches (1983: 122). While the Australopithecine progressing into the ''robust'' line became specialists in terms of mastication, the Homo lineage became specialists in terms of using their larger brains and increased intelligence in creating ways to exploits the same niche. Although there now existed two branches in hominid evolution, each embodied a linear progression.

Rak stresses that A./P. robustus and A./P. boisei "represent two distinct fundamental structures of the facial architecture" (1983: 121) and thus are deserving of their separate species names but does not advocate the generic distinction of Paranthropus. Based upon the fossil sample that existed at this time, A./P. robustus evidence only came from South Africa and "should probably be considered a relic that continued to exist in the South and that probably represents the prototype from which A. boisei evolved. The appearance of the latter in East Africa may be the result of immigration to a newly opened ecological niche, whose exploitation became possible through A. boisei's further specialization" (Rak 1983: 121). The separate species names are based on morphology, with A./P. boisei being the "final product" in the australopithecines line, rather than following the lumping tendency by placing them in geographical variant subspecies or a separate genus designation.

Prior to Yoel Rak's studies, A. africanus as the last common ancestor between the Homo and australopithecine lines was the accepted theory. In Hypothesis Two postulated by Richard Skelton, Henry McHenry, and G. Drawhorn (diagram), A. africanus takes A. afarensis place in the phylogenetic trees with A. afarensis being ancestral to A. africanus (Tattersall 1988: 72). Although the theory still maintains the two-prong fork appearance, the split occurs after 3.0 million years rather than before. Rak justifies his rational in his hypothesis: "Otherwise, we are forced to view A. africanus as having lost its derived facial features and reassumed a generalized facial appearance in order to evolve into Homo" (1983: 120). He renounces the possibility because it would be contradictory to the theory of gradualism he advocates - towards what he refers to as a "final product", being the brain in the Homo lineage and the boisei-like 'hyper-robust' mastication facial adaptation in the Australopithecus lineage. Yet, the females and the young of A. africanus specimens, particularly TM-1512 and STS-52, sway towards facial morphological similarity to A. afarensis, while the males of A. africanus are similar to the females and young of A. robustus (Tattersall 1988: 72). As the sample size becomes larger, idiosyncratic anatomical variation increases and characterizing traits of species overlap, which does not support gradual linear progression.

Into the 1980's, more successful field research allowed for more accurate knowledge to be expressed in the theoretical framework. In 1984, Richard Leakey and Alan Walker's crew working on the west side of Lake Turkana in Kenya found the 'Turkana Boy', KNM-WT 15000, an almost complete skeleton of African Homo erectus/Homo ergaster (Leakey & Lewin 1992: 122). The following summer, the crew returned in search of the rest of the specimen, with much fruitless effort. Twenty miles/thirty-two kilometers south of Nariokotome and two miles north of the Lomekwi River (the geological context of the localities will be discussed further on in the paper), checking on a hippopotamus skull, Alan Walker came across an upper jaw and skull. With collection of traits expressed like in no other, a small brain combined with large tooth roots, its discovery allowed for the immediate recognition of a need to reformulate their now relatively simple theories of human evolution. Two days later, Pat Shipman found a large dorsal ventral skull piece of the same individual with the largest sagittal crest of all fossil specimens, compounding with the nuchal crest, a feature typical of A. afarensis morphology. In reconstruction, the cranium was not difficult while the face was because it was highly fragmented and so unique: the lateral side of the orbital margin had no edge. Walker speaks of the specimen: "In gorillas, there's no distinct edge, just a smooth curve of bone. This pattern - though less marked - occurs in a few of the 'robust' australopithecines from South Africa, but we had not found it in East Africa" (Leakey & Lewin 1992: 128). The entire face was "vintage A./P. boisei, large and dish shaped" (1992: 128), yet the brain was one of the smallest of all fossil specimens, at 410 cubic centimeters. The specimen came to be known as the 'Black Skull' because the "Maganese salts that penetrated the bone during fossilization had produced a bronze-black color" (1992: 127). Alan Walker tentatively scientifically named it Australopithecus kamoyensis, in the honor of the field team member who brought them to their site that day.

Alan Walker and Richard Leakey wanted to be conservative in allotting a scientific name for the 'Black Skull'. The eastern shore of Lake Turkana, Koobi Fora, has yielded the largest sample of A./P. boisei fossils, the earliest evidence being from the Shangura Formation at Omo River dated between 2.75 and 2.5 million years ago (Wood 1992: 235). This already established fairly contemporaneous origination date and proximity already makes A./P. boisei a probable option to classify the 'Black Skull' in, based on more than anatomical specializations alone. The initial reaction was to now create a three-pronged fork model, with the 'Black Skull' species leading to A./P. boisei in one line, A. africanus leading to A. robustus in another line and Homo in the final line (Lewin 1986: 720). Yet, if placed in A./P. boisei, the anatomical characteristics defining that distinction would be expanded and thus endorse the lumping tendency. Leakey is neither an advocate of gradualism nor the idea of chronospecies, "different segments of an evolving lineage that is arbitrarily divided in time" (Leakey & Lewin 1992: 129), because it obscures the physical response of natural selection to environmental pressures. If not a lumping tendency issue, then the level of homoplasies is extreme. Functional convergence between the A. africanus- A./P. robustus line and the 'Black Skull'-A./P. boisei line is the size of the posterior (cheek) teeth that increases as the anterior teeth decreases. In the Homo lineage, both the posterior and anterior teeth experience a decrease in size.

The convergence does not only lie in masticatory adaptation. Environmental pressures directly affecting diet adaptations indirectly affect other physical aspects. In this case, the reduction in size of the front teeth influences the orthognathism of the face. As Richard Lewin suggests in his report on the find (before Alan Walker and Richard Leakey officially published their research findings in 1986) that all three lines are "undergoing a dramatic shortening of the face…becoming more vertical, as in modern humans. In addition, the base of the cranium in each line is becoming more flexed, probably related to the tucking under of the face" (Lewin 1986: 720). Yet the probability of convergence was very contentious and apparently inevitable in cranial anatomy for a three-pronged taxonomy model. At this preliminary point in research, still essentially the unearthing of the "Black Skull,' the processes of pressure and adaptation cannot quite be deciphered. The specimen might even change what is thought about those processes.

The 'Black Skull' has no direct influence on the reconstructed Homo lineage. The designation of Homo habilis will continue to be considered a recognition of the lumping tendency by categorizing the morphological intermediates between A. africanus and H. erectus. But now, as Richard Lewin mentions, the "ancestor/descendant relationship between A. africanus and H. habilis as a result of a series of advanced features being shared between the two species might now be more tempted to view these as parallelisms" (Lewin 1986: 721). The 'Black Skull' finding best further supports Johanson and White's theory, referred to here as Hypothesis One, only in that A. afarensis is the last common ancestor for the later hominid lineages, be there two, three or several. While the rest of their hypothesis is no longer valid as a result of the discovery, the proprietors were prepared for and accepting of the 'Black Skull'. As Kimbel states, "Well, we've been shown that we didn't get it quite right and that's just fine." Todd Olson, advocate of Hypothesis Three, views the 'Black Skull' as the 'type specimen', if you will, for his theory. He believes the fossil supports his belief in a 'robust' australopithecine in Eastern Africa, the sexually dimorphic male A. afarensis from Laetoli that he considers a separate species.

The geological context of the 'Black Skull' has also given palaeoanthropologists valuable clues and insights into ancient hominid use of their landscape. It is a confirmed prevalent paleoanthropological belief that the world experienced an overall shift to a colder atmosphere at roughly 2.5 million years. Frank Spencer mentions that "various fossil taxa were shuffled and reshuffled in an effort to produce a suitable phylogenetic and taxonomic model that would explain these various forms" (1997: 150), providing a time sequence that encompasses sites between 4.5 to 1.3 million years ago. Throughout Africa, mammalian faunal association has supported "a shift from a more wooded habitat at the time of the earlier gracile forms to a more open country and grass-dependant fauna at the time for which there is the best evidence of the later robust australopithecines" (Wood 1992: 237). Faunal correlation is used mainly to confirm dates obtained through radiometric (absolute) methods because it is an indirect method of dating.

The 'Black Skull' derived from the Pliocene sediments of the Lomekwi drainage on the western shore of Lake Turkana, Ethiopia. Thus, the Kenya Museum number of the specimen is KNM-WT 17000. (Fig. 1 above, Walker et al. 1986: 518) These sediments are "overbank deposits of a large perennial river, probably the ancestral Omo" (Walker et al. 1986: 518). The Lomekwi drainage overall dips to the east while the strata in which the skull was found, Lomekwi I, was "deformed into a syncline by drag along a fault that truncates the section about 50 meters east of the site" (Walker et al. 1986: 517). Being the 'African Horn' region of the rift valley, the strata west of the site, despite smaller and more plentiful vaults, has been correlated via radiometric analysis of the volcanic ash layers: "The volcanic ash layer which caps the section is compositionally indistinguishable from Tuff D of the Shungura Formation" (1986: 517). Conveniently, the Upper Burgi Tuff of the Koobi Fora Formation on the western shore of Lake Turkana has also been radiometrically correlated to Tuft D (Walker et al. 1986: 518). The three stratigraphic layers, Lomekwi I, Tuff D, and the Upper Burgi Tuff, each provide a method of datum confirmation of the other two. In other words, the Plio-Pleistocene sediments of West Turkana, the Lower Omo Valley and East Turkana constitute the same sedimentary group.

Tuff D is absolutely dated at 2.52 +/- 0.05 million years old, which is corroborated by Lomekwi (Walker et al. 1986: 518). The 'Black Skull' comes from a layer at Lomekwi that associates to 3.8 meters below Tuff D, making the date of 2.52 the lower boundary. The next datable volcanic ash layer occurs at 10 meters, which provides an upper boundary to the date. To support the upper boundary, three exposed ash layers at the Lokalalei drainage four kilometers northwest of Lomekwi are utilized (above diagram). The upper two layers relate to the layers at Lomekwi from which the "Black Skull' was derived while the lowest layer correlates with a tuff in submember C9 below Tuff D. In other words, the specimen correlatively lied within submember C9 in the Shungura Formation. "Based on the K/Ar chronology of the Shungura Formation and scaling on the basis of constant sedimentation rates there, the cranium is estimated to be 2.55 million years old" (Walker et al. 1986: 518). Paleomagnetism earth polarity chronology further specified the date at ~2.45 million years old. This date is less certain because there is discrepancy between the paleomagnetic chronology of the Shungura and Lomekwi formations. The accepted absolute age of the Black Skull, allowing for error, is 2.50 +/- 0.07 million years old (Walker et al. 1986: 518 & Leakey & Lewin 1992: 130), a mere thousand years after the beginning of the Late Pliocene.

The use of mammalian faunal correlation as an indirect dating method was successful at West Turkana because of the plentiful (over two hundred) fossil specimens compromising more than forty species. "The skeleton of a ground-dwelling colobine (monkey) and a relatively complete camel mandible" (Walker et al. 1986: 518) support the assumption that the atmosphere during the Pliocene had become drier and possibly hotter while forests did not completely cover the continent any longer. Bovid (hoofed, hollow-horned ruminants) samples, not only being the most prevalent non-hominid sample from the site, are also the species represented in the Shungura formation. This is in compliance with stratigraphic correlation between Tuff D of that formation and Lomekwi, as mentioned above. "Elephas recki shungurensis is the common elephant at the localities considered here while the Notochoerus scotti and Kolpochoerus limnetes are the common suids" (Walker et al. 1986: 519). The three separate tax represented are common throughout Pliocence sites, but these particular specimens are extremely morphologically similar to those from Shungura formation members C and D (diagram above). This evidence, along with the absence of other taxa that are present in older or younger layers of Lomekwi confirm the accepted absolute age for the 'Black Skull.'

The first report on the specimen, written by Alan Walker, Richard Leakey, John Harris, and Frank Brown (all together referred to here as 'the authors'), was printed in Nature in August 1986. The morphological characteristics of the almost complete adult cranium KNM-WT 1700, the 'Black Skull' were thoroughly explained (photographs following cover page, Leakey 1988: 3 & Walker et al. 1986: 519). Luckily the spheroid brain case was not deformed during burial, but it does have some parts missing (Walker et al. 1986: 519). These include the zygomatic arches, facial bone fragments from the maxillary sinus, the temporal plates where they meet with the zygomatics, and most of the frontal processes. The piece of the impressive sagittal crest is preserved on the anterior portion of the parietal while the rest of it is not there, along with the frontals and inferior portion of the nuchal region on the occipital bone of the skull. It does not have bilateral asymmetry but has sharp edges in the orbits and all of the bony contacts.

The specimen has many analogies with A./P. boisei, including a massive palate, face and cranium base. The later two physical qualities comparable to the A./P. boisei type specimen, OH 5, and KNM-ER 406 from across the lake at Koobi Fora (although KNM-WT 17000 palate is slightly larger) (Walker et al. 1986: 519). The 'Black Skull' has the following masticatory specializations that are similar to the 'robusts': the palate is large with a thick roof; the broken root of the molars and fourth premolar are large and therefore the teeth themselves probably had large crowns; anterior zygomatic processes of the maxilla create a flat and flared face and provide enough room for the massater and temporalis to be oriented forwardly and laterally so grinding is most efficient (Kreger 2001: 1 & Wolpoff 1999: 309). Yet it has a primitive small brain vault size of 410 cubic centimeters the smallest size ever published for any fossil hominid (Walker et al. 1986: 519). (The authors believe that A.L 162-28 from Hadar must be smaller.) This measurement is the mean of five water displacements of the reconstructed cranial vault, but had a standard error of 4.32. As a result of the preservation of the parts of the cranial base and frontal orbital plates, the cranial vault was able to be reconstructed. KNM-WT 17000 also exemplifies other primitive characteristics resembling A. afarensis. These include: large anterior tooth sockets; a flattened cranial base with associated flat, non-projecting articular eminence of the mandibular fossa and a shallow palate; petrous pyramid of temporal bone oriented towards the sagittal crest while the foramen magnum is placed posteriorly (relative to modern humans); extreme facial prognathism (Kreger 2001: 2 & Wolpoff 1999: 309). The projection of the massive face and palate combined with the small brain case justifies (in terms of physical balance and support) the largest sagittal crest ever found on a hominid and complete (no intervening bare area) compound temporal-nuchal crest. This exemplifies an even more "extreme development of the nuchal muscles and the more vertical orientation of their attachment on the occipital bone" (Kreger 2001: 2). The right upper third premolar is the only complete tooth crown in the specimen (Walker et al. 1986: 519). Its dimensions are 11.5 mesiodistally (horizontally/laterally) and 16.2 buccolingually (vertically/longitudinally), measurements which are outside the range of A. robustus (9.2 to 10.7 md, 11.6 to 15.2 bl) and at the high end of the range for A./P. boisei (9.5 to 11.8 md, 13.8 to 17.0 bl). It is more square than in the type specimen for A./P. boisei, OH 5, which is bigger mesiodistally at 10.9 and smaller buccolingually at 17.0.

The authors recognize two morphological aspects that are not found in KNM-WT 17000 that are in A./P. boisei. It has, what the authors call, a 'twist' along the length of the supraorbital torus that typical 'hyper-robust's do not have (Walker et al. 1986: 519). Yoel Rak, in his characterizations of the facial robusticy progression through the australopithecine line mentioned earlier, defined A./P. boisei as having anterior palates that are deep, yet KNM-WT 17000 has a palate like A./P. robustus, in which the palate is shallow throughout. The middle and lower facial prognathism of KNM-WT 17000 separates it from all of the 'robust' australopithecines, the latter of which are orthognathic (although a bit of the incisors bulge past the supraorbital torus).

Besides providing detailed descriptions of the specimen the authors rationalized their classification for the 'Black Skull.' Since the majority of anatomical features are similar to A./P. boisei, KNM-WT 17000 was initially categorized as such. This will be referred to here as Hypothesis Four. Those which are not A./P. boisei analogies are either (or both) pleisomorphies (retained primitive traits) or attributed to undocumented intraspecific variation. The latter idea was met with much skepticism in preliminary theories - claiming it advocated lumping, gradualism, or homoplasies - and continues to be today. Hypothesis Four "states that the morphological resemblances between the Southern and Eastern 'robust' forms (A./P. robustus and A./P. boisei respectively) arose through convergent evolution, that is, convergent functional adaptations to a heavily masticated diet" (Tattersall 1988: 73). The authors, in recognition of the issue with convergent evolution, serve a taxonomic condition: "Although future finds may show that KNM-WT 17000 is well within the range of variation of A. boisei, it is also possible that the differences will prove sufficient to warrant specific distinction" (Walker et al. 1986: 521). Just as White and Johanson reserved the right to alter their hypothesis with the future coming of new discoveries, the authors reserved the right to name a new species for the 'Black Skull.' Yet in this case, the authors took up their reservation in advance, allowing the reader to interpret the evidence for himself and chose the distinction he or she prefers. For example, Frank Spencer (1997) and B.A. Wood (1992) both include KNM-WT 17000 in A. boisei for their encyclopedic explanations for Australopithecine evolution.

The scientific name for Omo 1967-18 and accepted name for KNM-WT 17000 here is Australopithecus aethiopicus (Walker et al. 1986: 521). Camille Arambourg and Yves Coppens discovered the toothless mandible in 1967 and gave it the designation of Paraustralotpithecus aethiopicus because "its V-shaped jaw (among other features) distinguished it from the robust australopithecus forms know in the area" (Kreger 2001: 1). The prefix Para- (near) was chosen because "of the uncertainty at the time as to weather it was truly Australopithecus" since "there wasn't much else comparable to it at that age, about 2.6 million years ago" (Leakey & Lewin 1992: 133). The authors sunk the generic classification of Paraustralopithecus for the Omo mandible into Australopithecus, feeling the discovery of KNM-WT 17000 validated that. Omo 1967-18 was also an equivocal specimen, for equal morphological classification issues. In the years to follow, one of its discoverers, Coppens (1979) regarded the Omo mandible as an early A. boisei and others, such as Johanson and White (1979), placed it in A. africanus (Leakey 1988: 22). For the same reason the A./P. boisei classification was suggested, the authors suggested to compare KNM-WT 17000 to the Omo 1967-18 mandible from the Shungura Formation in the Lower Omo Valley in Eithiopia: not only for the morphological similarities, but also because of the fairly contemporaneous dates and stratigraphic proximity of the typical samples across the lake in Koobi Fora to the 'Black Skull. If you recall from the geological and paleoecological discussion above: the Plio-Pleistocene sediments of West Turkana, the Lower Omo Valley and East Turkana constitute the same sedimentary group.

The mandible KNM-WT 16005 (Fig. 3, Walker et al. 1986: 521), the details of which were reported along with the 'Black Skull' in the initial report (which will not be discussed here in concern for focus), was found two miles southeast of Lomekwi at Kangatukuseo and provides further support for the A. aethiopicus distinction. (The location and stratigraphy of this site may be viewed in Fig. 1 above along with Lomekwi.) KNM-WT 16005 "in its size shape and proportions is very similar to" the Omo mandible (Walker et al. 1986: 521). The latter has a larger lower first molar (15.7 md, 14.3 bl mm) and lower fourth premolar (12.0 md, 15.0 bl mm), but the lower second molar is a little smaller (17.0 md, 16.7 bl mm). KNM-WT 16005 is smaller overall and more frontally narrow when compared to the mandible of the 'Black Skull'. Nonetheless, KNM-WT 160005 provides the morphological and geological link between KNM-WT 17000 and Omo 1967-18. In Origins Rediscovered, Leakey and Lewin write: "Through the Byzantine rules of zoological nomenclature, the Black Skull may be the savior of aethiopicus (1986: 133)."

The alternate A. aethiopicus distinction is more accepted today, as well as by the initial postulators and advocates of Hypothesis Four, because of the contemporary rejection of the idea of a gradual evolutionary increase in robusticity with A./P. boisei being the end of the line. What Yoel Rak had supposedly proven turned out to be merely a consequence of an incomplete fossil sample. Walker et al. (1986: 521) clarify the confusion: the 'hyper-robust' skull morphology which characterizes the A./P. boisei classification and is also found in the 'Black Skull' is just that and not the teleological goal of evolution. While Rak said the cresting pattern, orthognathism and deep temporomandibular (jaw joint) became more exaggerated through time in the 'robust' lineage, the authors do not base their classification on this. They merely recognize "the massive size, extremely large palate and teeth, the build of the infraorbital and nasal areas and the anterior position and low take-off of the zygomatic arch" (Walker et al. 1986: 521) as qualities of the 'Black Skull' that are analogous with A./P. boisei and reason for altercation of present hominid phylogeny theories. Therefore, they also view A./P. robustus in the same light, relative to A. aethiopicus: "The new specimen shows that A. robustus is a related, smaller species that was either derived from ancestral forms earlier than 2.5 million years ago and/or has evolved independently in southern Africa" (Walker et al. 1986: 521). KNM-WT 17000's sympleisomorphies are more common with A. anamenisis than A. africanus, further proof that the specimen is not part of the 'robust' lineage. These traits include "very small cranial capacity, low posterior profile of the calvaria, nasals extended above the frontomaxillary suture and well onto an uniflated glabella, low calvaria with frontal squama and extremely convex inferolateral margins of the orbits such as found in some gorillas" (Walker et al. 1986: 521). These aspects of morphology are not found in A./P. robustus or A./P. boisei. Considering each lineage as separate entities, the authors reject the evolutionary progression of A. africanus to A. robustus to A. boisei. Thus, they renounce the 'proof' of gradual masticatory specialization in the 'robust' lineage provided by Yoel Rak in support of Hypothesis One and thus the theory itself.

Although Hypothesis Two became equivocal with the advance of Hypothesis One (the latter exemplifying a need to be reformulated above), the theory still held some justification, especially since the facial morphology of A. africanus specimens were showing morphological similarities to A. afarensis and A./P. robustus. The 'Black Skull' removes A. africanus from the possibility of being the last common ancestor of the Homo and Australopithecus lineages with theories that the specimens which constitute the species have autapomorphies (uniquely derived traits). The authors acknowledge this: "KNM-WT 17000 shows that all known A. africanus share features which are derived relative to it" (Walker et al. 1986 521). Therefore, Hypothesis Two is also rejected. A. africanus remains an acknowledged species today but is equivocal in terms of phylogenetic placement.

During preliminary theoretical formulations, Todd Olson believed KNM-WT 17000 to be the perfect candidate for the 'robust' australopithecine in Eastern Africa, the sexually dimorphic male A. afarensis from Laetoli that he considers a separate species in Hypothesis Three. Although there exists special sympleisomorphies between A. afarensis specimens and KNM-WT 17000 (listed in two paragraphs above as part of the cause to reformulate Hypothesis One), these are only features of the calvaria, the brain case itself (Walker et al. 1986: 522). Further more, "not one individual adult specimen of A. afarensis preserves a facial skeleton attached to the calvaria." Once again, more fossil evidence was necessary to move to the next stage in the formulation of theories. After the 'Black Skull' was found, it seemingly supported Hypothesis Three and initially supported as a possibility by Walker, Leakey, Harris and Brown (1986) in the first report.

Hypothesis Five (diagram below) departed from the separation of Hadar and Laetoli fossils, as advocated by Todd Olson, thus viewing A. afarensis as one sexually dimorphic species (Tattersall 1988: 73). Hypothesis Five may be considered a reformulation of Hypothesis One in that it retains A. afarensis as the last common ancestor to the Homo and Australopithecine lineage. A. africanus is seen to be a predecessor to the Homo lineage after diverging from A. afarensis (Kreger 2001: 3). Walker and Leakey (1986) state that A. africanus may be ancestral to A. robustus, the southern 'robust' species, as Hypothesis One suggests. A. africanus cannot be a predecessor to the northern species because it did not exist there (according to the present fossil sample) and "A. africanus demonstrates many derived features relative to the condition seen in KNM-WT 17000" (Walker and Leakey 1986: 256). The taxonomic placement of A. africanus is equivocal and often is simply removed.

Hypothesis Five is an appreciation of Hypothesis Three in that it recognizes separate Southern (A./P. robustus) and Eastern (A./P. boisei) African 'robust' species, geological variants. A. aethiopicus is morphologically justifiably placed between A. afarensis and A./P. boisei, but not to A./P. robustus (the diagram provided does not properly illustrate this detail). The latter species diverged before A. aethiopicus evolved, a phylogenetic separation based on the distinguishing "A. boisei condition of the teeth, palate, face, and foramen magnum" (Walker and Leakey 1986: 256). Walker and Leakey acknowledge that their belief means that "several features that are common to both 'robust' species must have evolved in parallel" (1986: 256). These parallelisms include (directly listed from the report): increase in cranial capacity; changes in endocranial shape; development of middle branch of meningeal vessels; possible changes in the venous sinuses; changes in orientation of the posterior surface of the petrous pyramid; change in the set of the facial skeleton on the braincase; development of articular eminence; change in orientation of typanic plate to become more vertical; and changes to make the cranial base more flexed (Walker and Leakey 1986: 256). Therefore, the theory indirectly accepts the equivocal genetic distinction of Paranthropus, as postulated some thirty years ago by J.T. Robinson. Ian Tattersall makes a further suggestion: "In a strictly cladistic interpretation of the taxonomy of this phylogenetic hypothesis not only would Paranthropus be the correct genus grouping for P. aethiopicus, P. robustus, and P. boisei but the species 'A.' afarensis should properly be placed in a distinct genus; the name Australopithecus would pertain only to A. africanus" (1988: 73). (I have not found nor heard of any documented evidence supporting or rejecting the latter part of his statement since it was published in 1988.)

Hypothesis Five also suggests that if convergent evolution exists within Paranthropus clade then it "might have characterized all of later hominid evolution." Walker and Leakey list several homoplasies between the Homo lineage and Paranthropus lineage (direct listing here): increase in brain size; increase in flexion of cranium; decrease in orthognathism; reflexion of frontal lobe rostal section; cerebellum goes back under the occipital poles; petrous temporal crest becomes sharper; middle meningeal vessel pattern develops a middle branch; development of the articular eminence; petrous temporal axes become more cronally oriented; longus capitis markings get weaker; eruption patterns become identical (1986: 256). Walker and Leakey state that if not homoplasies, these features evidence that the Homo and Paranthropus lineages shared a common ancestor that was not A. africanus. They point out that "In the unlikely event that the second alternative is true, it follows that KNM-WT 17000, which has not developed most of these features, must be close to that ancestral condition (Walker & Leakey 1986: 256)." The 'Black Skull' is not relatively specialized to later hominoid species of either lineage.

Kimbel, White and Johanson's analysis (1986) created alternative phylogenic constructions of hominid evolution (published in the same book as Walker & Leakey 1986). Following a thorough discussion of the analogies of KNM-WT 17000 (Figure 16.1, Kimbel 1986: 262), they concluded (1986: 266) not only that Australopithecus aethiopicus (not Paranthropus) is a valid taxon but also that it is the link between A. afarensis and A. robustus and/or A. boisei. They eliminated A. africanus as being the last common ancestor of Homo and 'robust' australopithecine lineages or single ancestor of the latter lineage, further making the species' placement in phylogenic constructions difficult. Kimbel, White and Johanson confirm that KNM-WT 17000 should not be classified as A. boisei, but do not advocate a third lineage originating in the mid-Pliocene. The 'Black Skull' does not imply, as Todd Olson had viewed, that "A. afarenisis cranial reconstruction combines two taxa and provides evidence of more than one hominid species at Hadar" (Kimbel 1986: 266). They finalize the rejection of Hypothesis Three with the following statement: "In sum, the discovery of KNM-WT 17000 does not require a re-evaluation of the taxonomic unity of A. afarensis. We further maintain that such a re-evaluation can only be based on the Hadar remains themselves; to employ KNM-WT 17000 as a test of the taxonomic composition of the Hadar sample constitutes an inappropriate use of this otherwise important hominid fossil" (Kimbel 1986: 266). Although contention about the Hadar specimens continues to persist today, Wolpoff supports that KNM-WT 17000, or any unassociated fossils should not be involved in resolving the issue: "The question of variation at Hadar must be resolved in the Hadar sample" (1999: 278). They agree with Walker and Leakey's theory (1986) in that that the 'Black Skull' "demonstrates considerable parallelism in early hominid evolution" (Kimbel 1986: 266).

Other remains were later attributed to A. aethiopicus, supporting the species distinction. Omo 338y-6 provided evidence that the extensive masticatory system formed at an early age for the species. Dated at 2.39 million years ago, this ten-year-old specimen is a cranial vault that lacks the face and frontal bone (Kreger 2001: 2 & Wolpoff 1999: 314). Morphological features that indicate the early formation of the heavy mastication are (directly listed): a large overlap of the temporal onto the parietal bone and a marked development of the superior nuchal line with a strong downward projecting inion at its center. Although the specimen did not have a sagittal crest, it would have developed with age because the temporal lines meet anteriorly in the middle of the skull. The 2.7 million year old mandible fragment, L55s-33, from level C6 at Omo was found before the 'Black Skull.' The single worn fourth premolar that the specimen retained was "too small to be A. boisei (since it was) eighty-three percent the size of the smallest known A. boisei tooth, but did not fit into any other species well" (Kreger 2001: 2 & Wolpoff 1999: 310). Gen Suwa, a specialist in the study of australopithecine molars, "reported that the premolar was unusually think-enameled and had the squared off shape of the hyper-robust species" (Kreger 2001: 2 & Wolpoff 1999: 310). Following the discovery of KNM-WT 17000, his judgement of the premolar form was justified with the L55s-33's placement into A. aethiopicus.

Hypothesis Five's support of the polyphyletic distinction of Paranthropus along with extensive parallelisms, as well as Kimbel, White and Johanson's advocacy of the latter, was met with response to prove its impropriety, causing much cladistic analyzation of the 'Black Skull' in efforts to make a more parsimonious phylogenetic tree. Skelton and McHenry (1992) and Liberman et al (1996), who used different traits in their studies, both came to the same conclusions: "Both see A. aethiopicus as a dead-end sidebranch and Paranthropus as polyphyletic and invalid" (Kreger 2001: 3). Skelton and McHenry view A. aethiopicus as "Rather than giving us insight into the evolution of the other 'robust' australopithecines, it instead gives us unexpected insight into the nature of the last common ancestor of all post-afrarensis hominids and the nature of the heavy chewing adaptation" (Skelton & McHenry 1992: 340). The evidence they present denies more than one lineage leading from A. afarensis and provides for three hypothetical ancestors that will "eventually be discovered in the fossil record" (diagram below, Skelton & McHenry 1992 340). The aethiopicus-like ancestor will have heavy masticatory traits similar to those of A. afarensis or A. africanus, but not of the true A. aethiopicus. "All other post-afarensis hominids are too derived in basicranial flexion, orthognathism and encephalization to have been the ancestor of A. aethiopicus" (Skelton & McHenry 1992: 341). The 'Black Skull's species cannot be an ancestor itself because it would mean a reversal of heavy masticatory traits through to the A. africanus-like ancestor.

Skelton and McHenry's study also does view several homoplasies as necessary in Hominid evolution, citing Darwin's justification of "analogical variation" - similar adaptation of similar creatures in similar environments to similar pressures (Skelton & McHenry 1992: 342). Their study analyzed seventy-seven traits in five functional trait complexes, thirty-four of the traits considered in the masticatory trait complex (Table 2, Skelton & McHenry 1992: 326). The results indicated that the "at least some ways different" masticatory apparatus in A. robustus and A. boisei signify that the similarly heavy chewing adaptation in A. aethiopicus is due to parallel evolution (Skelton & McHenry 1992: 342). The parallelism is most prevalent in traits related to heavy mastication between A. aethiopicus and A. robustus. A. aethiopicus developed "an emphasis on the masseter as the main chewing muscle, with a primitive retention of the emphasis on the posterior temporalis," while A. robustus and A. boisei also emphasized the masseter, but had "the derived condition of the functional de-emphasis of the posterior temporalis" (Skelton & McHenry 1992: 342). Skelton and McHenry state that if the homoplaisies do not lie in the mastication complex traits, then they exist in traits related to the other four trait complexes, anterior dentition, basicranial flexion, prognathism/orthognathism and enchephalization.

While Skelton and McHenry's phylogenetic construction is presently the most predominant, cladistic studies will persist without end. For example, in a more recent study by Strait et al. in 1997, Paranthropus is again accepted as monophyletic, with P. aethiopicus being the recent common ancestor for P. robustus and P. boisei. Yet Wolpoff (1999: 308) supports a phylogeny like Skelton and McHenry's. Evidently to prove which present theories are correct, the fossil sample size will have to increase, as always in paleoanthropology. It is also important to consider that the present is both infinite and unstable.

The 'Black Skull' was found contemporary to a period that geologists and paleontologists believed to be a global cooling period. In hominid evolution, the origin is believed to happen at some point between 7.5 and 4.0 million years ago. At this time a "a single species of bipedal ape established the family and probably began an adaptive radiation" (Leakey & Lewin 1992: 120). Paleoecological research has revealed that the landscape was a woodland environment at this time, rather than a forest or savannah mosaic. The diets of these predecessors was most likely ape-like, but by three million years ago the morphology of the teeth in the fossil record is seen to begin altercations yet still reliant on plant foods. A million years subsequent "some bipedal apes had become small-brained, large-cheek-teeth, specialist plant eaters; others, large-brain, small-cheek-teeth omnivores; some filled niches in between" (Leakey & Lewin 1992: 120). Elisabeth Vrba, via the intense study of the extensive variation of the bovid fossil record also from this time period, believes that "Climate change, through changing vegetation cover and population distribution can drive evolution" (Leakey & Lewin 1992: 130). Her theory of climate-related animal evolution, 'turnover pulse', results from the evidence that "it was 2.8 million years ago that cycles of cooling and drying in East Africa intensified, producing a shift towards grassier environments. These changes, in turn, resulted in the increase in number of bovids adapted to those environments" (Kerr 2001: 237). If environmental conditions are the main pressures for evolution of particular adaptations, then functional convergence between similar organisms is highly likely. Evidence of a paleoclimatic change during the Pliocene supported the concept of a branching out of hominid clades. Therefore, the 'Black Skull,' in its unique combination of 'robust' and 'gracile' features, despite all other implications of equivocal phylogenic theories, eliminated the possibility of linear evolution in the hominid lineage. There is no doubts that this notion is altering theories as the ones that have existed are being read.

To conclude, I leave you with the following quote from Implications of KNM-WT 17000 for the Evolution of 'Robust' Australopithecines, written by W.H. Kimberly, T. D. White, and D.C. Johanson (1986: 265): "From Darwin on, it has been acknowledged that such adaptive morphology in a closely related group of organisms is particularly susceptible to parallelism. The KNM-WT 17000 cranium shows that the Hominidae is no exception to this pattern."