Roots, Bugs and Venison: Prehistoric Cuisine at Swartkrans Cave by Jason Heaton | Papers by Jason

View of the current excavations at Swartkrans in the 1.8-million-year-old depositis, which have yielded two types of early hominin, stone and bone tools and the butchered remains of large animals. Image: Jason Heaton Early lime miners played a crucial role in alerting the scientific community to the presence of hominin and other Pleistocene fossils in many South African caves. Image: Jason Heaton ROOTS, BUGS AND VENISON: Prehistoric cuisine at Swartkrans Cave Travis Pickering and Jason Heaton show us how the fossil record has led us towards an understanding of human evolution. Mines and anthropology The study of human evolution in South Africa owes a great debt to an unlikely source. Long maligned in textbooks and secondary academic sources as destroying fossils, it was early 20th century lime miners who first exposed and alerted scientists to the rich treasure troves of human prehistory encased in caves like Taung, Sterkfontein, Kromdraai and Swartkrans. Gold and lime mining were intimately linked in those pioneer days, because lime is a catalyst in the Macarthur-Forrest cyanide process used for extracting gold from low-quality ore. In order to recover lime from the underground caves in which it was contained, miners also had to remove, by hammering, digging and dynamiting, calicified cave sediments (breccia) that were found along with the lime. In the process, the miners also discovered that the sediments contained fossilised bones. The Macarthur-Forrest process This is a process of gold cyanidation and is a metallurgical technique for extracting gold from low-grade ore by converting the gold to water-soluble aurocyanide metallic complex ions. It is the most commonly used process for extracting gold from ore, but is highly controversial because cyanide is so poisonous. Early hominins An accompanying article in this issue, by Bob Brain, tells the story of Raymond Dart’s fortuitous receipt of a block of such fossil-bearing sediment, which contained the skull of the Taung Child. After careful study, Dart determined that the skull represented an extinct species of human he calledAustralopithecus africanus – the southern ape of Africa. Although the Taung Child’s skull was in many ways ape-like, with a small braincase and large jaws and teeth, a key human trait was obvious in the base of its cranium. The hole through which the spinal cord travels to connect with the brain, the foramen magnum, is situated toward the front of the child’s skull. This morphology is typical of an obligate biped (like a human and unlike an ape) whose head rests on top of an upright spine. In comparison to modern humans, it is natural to think of Australopithecus africanus, which we now know dates between 3–2 million years old, as primitive. But, research since Dart’s day has revealed a myriad of species that are ancestral to humans and predate Australopithecus africanus by millions of years. Biomolecular evidence indicates that the taxonomic group to which modern humans belong – the hominins – diverged from the chimpanzee lineage somewhere in Africa about 8–5 million years ago. And, recent palaeontological research across Top: Piles of fossil-bearing cave sediments from Swartkrans. These calcified sediments are called breccias and require hammers and chisels to remove them from the cave and then subsequent processing in acetic acid to remove fossils without inflicting damage on them. Image: Jason Heaton Above: View of the bases of a modern human (left) and modern chimpanzee cranium, showing the relative positions of the foramen magnum in an obligate biped (the human) and a quadrupedal ape. Image: Jason Heaton ▲ ▲ Quest 5(2) 2009 3 Above: Side views of human and chimpanzee jaws, illustrating a fundamental anatomical difference between hominin and non-hominin primates. Living and extinct hominins lack(ed) an upper canine honing complex in their lower jaws, while apes posses this feature. Image: Jason Heaton Left: Hypothesised relationships of modern humans and African apes and their extinct ancestors, based on genetic and palaeontological data. Top: An Ardipithecus fossil jaw from the Gona region of Ethiopia.Image: Sileshi Semaw Above: A cast of the remains of Lucy. Image: Wikimedia commons the continent is augmenting the fossil record of this time span. Currently, there are three possible candidates for the oldest known hominin species. One of these, Ardipithecus kadabba, was discovered by palaeoanthropologist Yohannes HaileSelassie and his colleagues in geological deposits that occur in the Middle Awash River Valley of Ethiopia, and are dated between 5.8–5.2 million years old. Ten years earlier, in 1994, a research group led by Tim White announced the first known species of Ardipithecus, Ardipithecus ramidus, which was found in the same region as Ardipithecus kadabba, but in a geological context about one million years younger. White, Haile-Selassie and their coworkers argue that Ardipithecus kadabba and Ardipithecus ramidus are time-successive (that is, chrono-) species that represent the earliest appearance of the hominin lineage. The genus Ardipithecus is more closely allied to modern apes in much of its cranial and dental anatomy than it is to lateroccurring hominin species. Like modern chimpanzees and unlike hominins, Ardipithecus fossils possess large canine teeth and relatively small molars capped with thin enamel. While little is published about the skeleton of Ardipithecus below the cranium, the few arm bones that have been described display a mix of human-like and ape-like features. Human or ape? With so many morphological similarities to apes, it is reasonable to wonder why Ardipithecus is considered a human ancestor rather than an ape ancestor. Supporters of the hominin status of Ardipithecus point to two important anatomical features that it shares with later hominins. First, the position of the foramen magnum is towards the front of skull in the Ardipithecus, as it is in Australopithecus africanus and in modern humans. The second feature that purportedly links Ardipithecus more closely to living humans than to living apes is subtle, but just as critical as the position of the foramen magnum. Although Ardipithecus canine teeth are large, they are also blunt and less projecting and dagger-like than those of apes. In addition, apes have what are called sectorial lower third premolars, the front faces of which slope backward so that the upper canine locks between it and the lower lateral incisor when the upper and lower teeth are in contact. Thus, the back edge of the upper canine glides across this slope as an ape opens and shuts its mouth, honing that canine edge. In contrast, no hominin, living or extinct, possesses this canine honing system. Although the upper canine and lower third premolar of Ardipithecus kadabba interlocked, the system was not so developed as to sharpen the top fang as thoroughly and consistently as seen in apes. 4 Quest 5(2) 2009 A challenge to the hominin status of Ardipithecus has been offered by a Kenyan and French team of palaeoanthropologists working in 6.0 million-year-old deposits in the Tugen Hills of Kenya. Team leaders Brigitte Senut and Martin Pickford stress the apelike features of Ardipithecus and relegate it to a position as an ape ancestor, while offering their own recent discovery, Orrorin tugenensis, as the ancestor of all species leading to modern humans. Orrorin is represented by various teeth, mandible fragments and pieces of arm and leg bones, and like Ardipithecus, retains many primitive aspects in its anatomy. The same can be said for the remarkable 7.0 million year old cranium recently recovered by palaeontologist Michel Brunet and his team in the Djurab Desert of northern Chad. Critics of the hominin status of this cranium, assigned to the novel taxon Sahelanthropus tchadensis, claim that it is actually just an ape, with no indication of obligate bipedal locomotion, crucial to a species’ inclusion in the hominin lineage. Other researchers, such as White and Haile-Selassie, see enough anatomical continuity between Sahelanthropus, Orrorin and Ardipthecus to suggest that all three should be included in the same genus— which, if correct, would push the earliest known appearance of Ardipithecus one million years deeper into prehistory. Thus, there is still no clear picture of this earliest phase of the human story and as the dust settles around these new and exciting discoveries, many experts are taking a wait-and-see stance before naming the definitive human ancestor. The origins of humankind Regardless, it was not until long after the extinction of Ardipithecus, Orrorin and Sahelanthropus, some time around 2.3 million years ago, that the first species of our own genus, Homo, appeared, as indicated by fossil evidence from the fossil locality of Hadar, in northeastern Ethiopia. In the time span between the demise of Ardipithecus, Orrorin and Sahelanthropus and the rise of early Homo, hominins were represented on earth by the so-called ape-men or australopithecines, members of the genus Australopithecus. Dart’s Taung Child, Australopithecus africanus, was only one of several species included in this widely distributed and long-lived group of hominins. For instance, the most famous australopithecine skeleton, Lucy, belonged to a geologically older species called Australopithecus afarensis, known from 3.6–3.0-millionyear-old sites in Ethiopia, Kenya and Tanzania. And Australopithecus afarensis was preceded in East Africa by Australopithecus anamensis (4.2–3.9 million years old). The adaptation that most specifically sets the genus Australopithecus apart from the provisional hominins of earlier times and the modern apes is called megadontia. Megadontia refers to enlargement of the premolars and molars, the cheekteeth. But, not only did the cheekteeth of Australopithecus increase in absolute size through time, from species to species, but they also increased in size relative to body size. They also changed shape in some species, with the premolars becoming molar-like in their anatomy. In addition, the cheektooth enamel of Australopithecus is much thicker than that possessed by the earliest putative hominins and the apes. Collectively, these features functioned as part of an adaptive complex that involved increasingly more powerful chewing capabilities as the australopithecines evolved. The apex of this unique morphology occurs in a group of ape-men referred Top left: Comparison of the lower jaws and teeth of a modern human and Australopithecus robustus from Swartkrans. Note the extreme robusticity of the latter’s mandible and its enormous premolars and molars, a condition called megadontia and one that, in part, defines the genus Australopithecus. Image: Jason Heaton Top right: Australopithecus afarensis reconstruction. Displayed at Museum of Man, San Diego, California. Image: Wikimedia commons Above: Cranium SK 48 from Swartkrans, a representative of Australopithecus robustus, one of three species of robust australopithecine and the only one known from South Africa. Australopithecus aethiopicus and Australopithecus boisei are the two known species of robust australopithecine from East Africa. Image: Jason Heaton to as robust australopithecines. The particulars of skull morphology that link the robust australopithecines together as a distinct sub-group of Australopithecus are anatomical consequences of their extreme tooth size and their need for hyper-chewing power. For example, a high ridge of bone – the sagittal crest – that runs down the midline of the male robust australopithecine cranium functioned to anchor immense chewing muscles originating on the mandible. The mandible itself is massively deep and the cheekbones are placed far forward on the face of robust ape-men so that this huge jaw musculature could be further accommodated on the sides of the head. Quest 5(2) 2009 5 ▲ ▲ Above left: Swartkrans is remarkable for its large collection of Australopithecus robustus fossils, which derive from over 100 individual ape-men. Image: Jason Heaton Above right: Swartkrans was the first site in the world to yield evidence of two contemporaneous types of early hominin, Australopithecus robustus and Homo ergaster) (represented by this beautifully preserved cranium SK 847). Subsequently, the coexistence of robust ape-men and early Homo has also been confirmed in East African palaeoanthropoligical contexts. Image: Jason Heaton Top: Charles Darwin (1809–1882). Image: Wikimedia commons Above: The late South African palaeoanthropologist John Robinson proposed the influential ‘dietary hypothesis’, which argued that robust ape-men were specialised herbivores. Image: University of the Witwatersrand Herbivory and carnivory The earliest known robust ape-man species, Australopithecus aethiopicus, appears in the fossil record of East Africa around 2.7 million years ago. That emergence is at roughly the same time as the first appearance of stone tools in the archaeological record (2.6 million years ago) and makes Australopithecus aethiopicus a near contemporary of Australopithecus garhi (2.5 million years ago), the species that might have given rise to the genus Homo soon after. The coincidence of these three events—the rise of the robust australopithecines, the rise of early humans (Australopithecus garhi) and the invention of stone tools – is exciting and complicating for palaeoanthropologists. But, that was not always the case. Conventional wisdom has long held that that Africa experienced an extreme change to more open, savanna habitats between 3.0 and 2.5 million years ago and that this change caused the split of the hominin lineage into the robust ape-men and early Homo. The idea that adaptation to savannahs drove human evolution is even older than the discovery of hominin fossils from Africa. As early as the 1870s, Charles Darwin argued that open habitats were the crucible of human evolution – the idea being that forest fruits favoured as food by our ape ancestors would no longer be available to proto-hominins among the grassland flora. As the fossil record filled in throughout the 20th century and scientists began to gain a more sophisticated understanding of the ancient African environment, the ‘savannah hypothesis’ became increasingly developed—and Swartkrans Cave, in particular, played and continues to play a crucial role in that development. By the late 1950s, South African palaeoanthropologist John Robinson proposed a refinement of the ‘savannah hypothesis’ called the ‘dietary hypothesis’. At that time, many researchers assumed that Dart’s Australopithecus africanus was directly ancestral to the genus Homo and that it was also of the same period as a robust apeman from South Africa, now called Australopithecus robustus. Swartkrans has yielded the largest number of Australopithecus robustus fossils in the world, as well as several of an early form of Homo, Homo ergaster, and was the site upon which many of Robinson’s ideas were developed. Robinson argued that Australopithecus robustus adapted to the savannah by maintaining an herbivorous diet, but that it diverged from its forest ancestors because it specialised on the coarse Gracile, in this context, refers to hominin forms with more lightly built skulls, lacking the large teeth and jaws of the robust species. 6 Quest 5(2) 2009 Above: African bovids can be grouped as browsers versus grazers. The type of vegetation they consume – browse or graze – will be differentially expressed in their 12C/13C ratios. Image: Jason Heaton Above right: The 12C/13C ratio of a carnivore will reflect that of its prey – crucial information in a palaeontological context that can assist in reconstructing the diet of extinct animals. Image: Jason Heaton Right: Hyenas are carnivores; ape-men were long presumed to be mostly herbivorous. Thus, it was surprising when Australopithecus africanus fossils yielded 12C/13C ratios similar to those of hyenas. Image: Jason Heaton vegetable matter available in open country. These kinds of resources, such as nuts and plant roots, are notoriously tough to chew and it was in this light that the large jaws and teeth of the robust australopithecines made adaptive sense. In contrast the teeth and jaws of the presumptive Homo ancestor Australopithecus africanus are more gracile, and Robinson contended that that species (and later, Homo) solved the problem of savannah survival by increasingly incorporating meat into its diet. Meat, readily available on the African savannah, is soft and does not require its consumer to have a massive dental battery to break it down. Muscle does, however, adhere to bones and comes in large packages, so a cutting technology would be most useful for a blunt-toothed hominid that began to exploit this resource. Thus, the invention of stone tools becomes completely explicable and assignable to proto-Homo with the resource partitioning envisaged in Robinson’s dietary hypothesis. Further, acquiring meat requires a smarter brain than does picking nuts or grubbing for roots. It was not surprising therefore that our large-brained, small-faced lineage persists and the small-brained, large-faced robust ape-men eventually went extinct. Opportunistic feeders Of course, the model was too simple to ever hold up. Three avenues of recent investigation argue against an exclusive specialised nut or root diet for the robust australopithecines. First, nutritional analyses of those rather low quality savanna plant foods suggest that it would be exceedingly difficult, if not impossible, for a large-bodied primate like Australopithecus robustus to meet its daily caloric and protein requirements by relying on them. Second, studies of damage (scratches and pits) on the chewing surfaces of robust australopithecine teeth, when compared to the damage seen on modern primate teeth of known dietary specialisation, demonstrate that they ate a variety of food types. Third, isotopic analyses of robust australopithecine teeth indicate that the group was comprised of diverse feeders. Isotope research on fossil hominins requires some explanation. All plants take up two types of stable carbon (C) isotopes, 12C and 13C, during photosynthesis. However, tropical grasses and sedges can convert 13C into sugars and other tissues more easily than can tropical trees and shrubs. As a result trees and shrubs tend to ‘select against’ 13C during photosynthesis, and thus have lower levels of this isotope than do tropical grasses and sedges. Thus, trees and shrubs form a distinct group as opposed to grasses and sedges when analysing their proportions of 12C to 13C. Animals that eat these plants contain the same proportion of 12C to 13C as the type of plant group they consume—trees and shrubs versus grasses and sedges. Likewise, meat-eating animals reflect the 12C/13C ratios of the herbivore prey that they consume. In a groundbreaking 1999 study, palaeoanthropologists Matt Sponheimer and Julia Lee-Thorp demonstrated that Australopithecus africanus teeth from the South African site of Makapansgat have 12C/13C ratios similar to those of meat-eating hyena teeth also recovered there. This suggested to the researchers that Australopithecus africanus was sometimes carnivorous, feeding at least occasionally on the carcasses of grass-eating ungulates. (See Carbon isotopes, photosynthesis and archeology in this issue for more details on this line of research) Digging for termites? However, critics argued that the 12C/13C ratio in Australopithecus africanus can be explained by other factors, such as its consumption of grass-eating insects or water-loving sedges. When a similar 12C/13C ratio was identified in Australopithecus robustus, the supposed specialised nut or root feeding vegetarian, understanding the non-ungulate sources of the grass signal became even more crucial. Researchers, working within the frame of Robinson’s dietary hypothesis, originally classified artefacts that were made from antelope bones and that were found prominently at Swartkrans (and later at other Australopithecus robustus sites like Sterkfontein and Drimolen) as root-digging tools. Specifically, Bob Brain’s inspired skills Quest 5(2) 2009 7 ▲ ▲ Left: Early Stone Age technology was simple but effective. Sharp flakes were used to cut meat from bones and large cobbles were used to crack open antelope leg bones for access to nutrient-rich marrow. Image: Jason Heaton Bellow left: CK (Bob) Brain, legendary South African natural historian and palaeontologist. Brain’s long-standing excavations at Swartkrans revealed many remarkable aspects of early hominin adaptation and behaviour including the idea that Australopithecus robustus created and used bone tools to harvest edible roots. other researchers to explore the idea that there may have been an alternative underground food that Australopithecus robustus was instead targeting with its digging activities. In particular, it was argued, based on a new round of experimental work, that the bone tools were rather used by Australopithecus robustus to open the mounds of grass-eating termites. However, more testing and logical scrutiny is required before the rootdigging scenario is invalidated and can be abandoned in favor of the termiteforaging hypothesis. For instance, root extraction necessarily requires digging—and digging with tools if you were a prehistoric hominin lacking claws. A termite mound, in contrast, can be breached or even toppled simply by kicking or hurling a rock at it. It needs to be demonstrated that termite harvesting is somehow more efficient by digging holes than by causing swarms by other means of penetrating it. Additionally, Matt Sponheimer analysed termite isotopes and concluded it was unlikely that termites alone could explain the high C4 signal in Australopithecus robustus. Sponheimer’s data show a 3–40% C4 component in ape-man diets; if that was gained entirely from termite consumption, it would indicate a diet of nearly 100 % termites—a very unlikely situation. of observation and lateral thinking led to the important scientific epiphany that the wear patterns on the fossil bone tools and the digging damage incurred on screwdrivers he used to excavate Swartkrans were nearly identical. Brain further reasoned (very sensibly!) that Australopithecus robustus wasn’t digging at Swartkrans over a million years ago for fossils, but must have instead been interested in some subterranean resource important for its survival. This led him to experiment with modern bones that he used to dig for common edible resources still growing on Swartkrans Hill today, the underground bulbs of lilies and grass stars. That damage, too, matched the fossil bone wear and thus the hypothesis of a bone tool root-digging adaptation for Australopithecus robustus was well accepted. The new isotope results on ape-man fossils that give a grass eating signal, in part, prompted Was Homo the only early carnivore? The termite foraging hypothesis has, though, gained a real foothold in the secondary literature of palaeoanthropology. This probably has more to do with our perceptions of what Australopithecus should have been doing, rather than what it was doing. For many, eating the flesh and marrow of large vertebrates cannot be attributed to the ape-men; hominin meat-eating has to stay the exclusive capacity of the Homo lineage (probably including Australopithecus garhi). Carnivory is, after all, one of the primary factors that made us human—supplying the high energy and quality nutrients to grow big, smart brains, ensuring the survival and eventual primacy of our species. Australopithecus was a small-brained idiot that went extinct. However, another school of opinion has emerged that is willing to consider the isotopic data in a more straighforward way. If those data are saying that the robust ape-men could have consumed large vertebrate flesh, then maybe they did. How, then, could one explain the extraordinary jaws and teeth of those hominins? Wild animals experience fat and lean times throughout a year. Good times of year provide an abundance of the animal’s preferred foods. Harsher periods challenge the animal to subsist on other, non-preferred resources – called fallback foods. Among modern apes, gorillas and chimpanzees both prefer fruit, but each can also subsist on lower quality leaves and piths as fallback foods when seasons change and fruit becomes scarce. The difference between these two species, though, is that gorillas are able to subsist entirely on fallback foods, while chimpanzees cannot: they still require some fruit to survive. Researchers Greg Laden and Richard Wrangham have suggested the extreme morphology of the robust australopithecine skull and teeth might be an indication that these hominins followed a gorilla-like strategy of relying completely on fallback foods when preferred foods like fruit and meat were in short supply. On the ancient African savannah such fallback foods probably included roots and perhaps nuts. Adaptations for exploiting these types of resources are the ones that should be prominently expressed in the anatomy of an animal that takes a gorilla-like approach. This is because efficient exploitation of a single (or small range of) non-preferred resource(s) is essential for an individual’s survival during that time period. Species that adopt a more chimpanzee-like strategy – only using fallback foods to supplement the diet in a season when preferred foods become rare – are released from the pressure to specialise morphologically. Hence, the Homo lineage, the presumed chimpanzee-like hominin group, shows fewer extreme specialisations in its craniodental anatomy than seen in the robust australopithecines. The place of stone tools We must, of course, also consider the role of emergent stone technology in such a scenario and its potential to further release Homo from evolutionary pressures that were outside its body structure. There is a complication, though: robust australopithecines are found in archaeological association with stone tools almost as frequently as is 8 Quest 5(2) 2009 early Homo. Further, the isotopic data discussed above implicates ape-men in meat-eating to some degree, and much experimental and archaeological work, as well as simple logic, indicates that stone tools were probably invented to process animal carcasses. No hominin that ever existed possessed the imposing strength and speed, tearing claws, and fleshslicing and bone-crushing dentitions of the lions, leopards, hyenas and host of extinct predators with whom they competed for animal carcasses in the distant past. Those facts, while disadvantages for hominins millions of years ago, are happy coincidences for the scientists who attempt to reconstruct their foraging behaviors. In place of claws and teeth, early hominins used stone tools to process carcasses and those tools left incidental traces on the butchered bones. It is through analysis of those traces that palaeoanthropologists have begun to build a picture of hominin meat-eating. The two major categories of butchery damage imparted on bones by hominid tools are cutmarks, made by the sharp edge of a cutting tool when it accidentally sliced through overlying meat into underlying bone, and percussion marks, created when hominins used river cobbles to pummel open and harvest the marrow of animal leg bones. Cutmarks and percussion marks are seen alongside the stone tools that produced them, starting in the earliest recognised archaeological record from East Africa at 2.6 million years ago. This is not surprising because those first stone artifacts come in two basic forms – sharp cutting flakes and heavy cobble hammers—both of which are ideal for butchery and less perfect for other activities. As with the earlier East African occurrences, this trace evidence of butchery on animal bones is a common component of the Swartkrans archaeological record between 1.8–1.0 million years ago. It documents a very successful strategy of animal food exploitation, in which hominins appear to have enjoyed primary access to preferred, heavily meatladened parts of carcasses before those portions could be consumed by the host of carnivorous competitors that shared the Swartkrans landscape. Conventional wisdom, following from Robinson’s ‘dietary hypothesis’, touted these as exclusive abilities of Homo erectus, a more ‘human-like’ early hominin than was Australopithecus robustus and the other ape-man species. But, as reviewed here, a truly unbiased reading of the available evidence shows that this conclusion rests on a shaky scientific foundation. Changing perceptions and paradigm shifts are indications of a healthy and dynamic academic discipline, and Swartkrans Cave and its remarkable record of prehistoric life continues to catalyse palaeoanthropology in both ways. The stark dietary—and thus, by extension, lifestyle—separation for the robust ape-men and early Homo modelled by Robinson’s ‘dietary hypothesis’ now seems unlikely. With two contemporaneous species of large-brained, omnivorous hominins sharing the Swartkrans environs 1.8–1.0 million years ago, the challenges of differentiating their behavioral and cultural uniqueness will remain a daunting but exciting challenge for students of human evolution. ■ Travis Rayne Pickering is an Associate Professor and the Director of the Swartkrans Palaeoanthropological Research Project. He works at the Department of Anthropology University of Wisconsin-Madison, Madison, Wisconsin, USA and the Institute for Human Evolution, University of the Witwatersrand, Johannesburg, South Africa. Jason L. Heaton is an Adjunct Professor in the Department of Biology at BirminghamSouthern College, Birmingham, Alabama, USA. He is a project palaeontologist on the Swartkrans Palaeoanthropological Research Project. Above: Numerous bone tools, similar to the modern experimentally produced one illustrated here, have been recovered from Swartkrans and other robust ape-men cave sites in South Africa. It is hypothesised alternatively that they were used by australopithecines to dig up edible roots (left) or termites (right). Image: Jason Heaton Further Reading Brain CK. The Hunters or the Hunted? An Introduction to African Cave Taphonomy. Chicago & London: University of Chicago Press, 1981. Brain CK and Shipman P. The Swartkrans bone tools. In Swartkrans: A Cave’s Chronicle of Early Man, ed. Brain CK, pp. 195–215. Transvaal Museum, Pretoria,1993. Laden G. and Wrangham R. The rise of the hominids as an adaptive shift in fallback foods: plant underground storage organs (USOs) and australopith origins. Journal of Human Evolution 2005; 49, 482–498. Pickering, TR. Subsistence behaviour of South African Pleistocene hominids. South African Journal of Science 2006; 102, 205–210. Pickering TR, Domínguez-Rodrigo M, Egeland CP and Brain CK. New data and ideas on the foraging behaviour of Early Stone Age hominids at Swartkrans Cave, South Africa. South African Journal of Science 2004; 100, 215–219. Robinson JT. The origin and adaptive radiation of the australopithecines. In Evolution und Hominisation, ed. G. Kurth, pp. 120–140. Fischer, Stuttgart, 1962. Sponheimer M. and Lee-Thorp JA. Isotopic evidence for the diet of an early hominid, Australopithecus africanus. Science 1999; 283, 368–370. Sponheimer, M., Passey, B., de Ruiter, D., Guatelli-Sternberg, D. Cerling, T. and Lee-Thorp, J. Isotopic evidence for dietary flexibility in the early hominin Paranthropus robustus. Science 2006; 314, 980–982. Acknowledgements Pickering thanks first and foremost Bob Brain, who inspires his career in palaeoanthropology and, incredibly, exceeds all expectations as a collaborator and friend. It is through his patient and gracious mentoring that I have the honor of directing the Swartkrans Palaeoanthropological Research Project (SPRP) under his coordination. Thanks to the other members of the project for their invaluable contributions and camaraderie and to my family for their unconditional support. The SPRP is supported by grants to Pickering from the National Science Foundation (USA), the LSB Leakey Foundation (USA) and the Palaeontological Scientific Trust (South Africa). Heaton would like to think his wife Alison for her support throughout the SPRP and thanks also Stephany Potze for coordinating photographs of the fossil specimens. We both thank Bridget Farham for the invitation to contribute to this special issue of QUEST and for her saintly patience. Quest 5(2) 2009 9