Scavenging and Human Evolution
Although meat eating helped to shape
the evolution
of human brains, behavior and toolmaking; our early
ancestors seem to have been better scavengers than hunters
by Robert J. Blumenschine and John A. Cavallo
Man the Hunter is a phrase that rings. Who would not rather be numbered with the lion than with the vulture? Hunting seems nobler than scavenging and, at first glance, more profitable, too. What better way to reaffirm our evolutionary success than to portray our earliest hominid ancestors as mighty hunters? Many anthropologists agree that eating the meat of large animals helped to form the physical and social environment that selected for the traits that most distinguish humans from apes. But was that meat acquired by predation or by scavenging? This question matters perhaps as much as any in evolutionary studies because it touches on the definition of human nature. Unfortunately, the answer given by the theory of Man the Hunter is based more on sexual and other prejudices than on the fossil record and the ecology of finding food.
Scavenging has received little notice, we believe, because many anthropologists have been too quick to project current ways of life into the past. They use hunter- gatherers, apes or carnivores as surrogates for aspects of early hominid life that have been obscured by the passage of time a practice that strips the hominids of the very adaptations that made them unique. Advocates of the hunting theory also elevate hominids above other organisms, as if our ancestors were immune to most of the pressures that shape relations between predators and prey. In all these matters, they assume that early hominids found hunting to be bountiful, predictable and safe and scavenging to be marginal, opportunistic and risky.
Our research reaches quite different conclusions. Scavenging may have been more common than hunting two million years ago, at the boundary between the Pliocene and Pleistocene epochs. Flaked stone toolmaking, the practice of butchering large animals and the evolution of big-brained (Homo) all make their first known appearance in the physical record at this time. Because much of the evidence lies at such east African sites as Olduvai Gorge in Tanzania, we attempted to learn how to decipher the residues of ancient subsistence patterns at nearby game reserves: Tanzania's Serengeti National Park and Ngorongoro Conservation Area We also tried to test objectively the prevailing notion that scavenging would have been inferior to hunting.
In independent stints over a period of 20 months, we noted how predators and scavengers got their meat and what they did to the bones they left behind. Our fieldwork thus united ethology with taphonomy the study of how postmortem events alter carcasses in the fossil record. Further, we integrated these results with paleontological and archaeological evidence for the behavior of protohominids. This approach reads into the past only those aspects of present behavior and ecology that leave preservable residues. It thus avoids a wholesale imposition of the way of life of a modern species that happens to suit one's ideals.
That no substitute exists in anthropology for such actualistic studies may be demonstrated by what zoologists have shown about the behavior of the hyena the popular symbol of scavenging and the lion the prototypical predator. Until 30 years ago, no one conceived that each carnivore both hunts and scavenges. If biases can so cloud the truth about living carnivores, how much more careful must scientists be in reconstructing the subsistence of extinct hominids?
The theory of Man the Hunter has never been constrained by fossil evidence. Charles Darwin was the first to present hunting as the behavioral catalyst that selected for an enlarged brain, tool use, reduced canine teeth and bipedalism, thus splitting the lineages of humans and apes. He laid out his hypothesis in The Descent of Man (1871), before any fossils earlier than the Neanderthals had been found. When more ancient specimens turned up in the early decades of this century, workers linked them directly to Darwin's scheme. Raymond A. Dart, discoverer of the Australopithecus genus, spent some 30 years trying to show that this hominid could have hunted the animals whose bones were so often found mingled with its own. To circumvent the problem of the absence of stone tools at these sites, Dart invoked an "osteodontokeratic" tool and weapon kit made from a bones, teeth and horn.
This interpretation gained popular support in the many accounts of humanity's "killer ape" forebears. It fell apart, however, under the critical tests of the pioneering taphonomist C. K. Brain of South Africa's Transvaal Museum. He showed that the australopithecines had played no role in gathering the bones of the animals found in association with their own skeletons. Instead, these studies suggested, both hominids and ungulates had ended together when the leopards that hunted them discarded their carcasses at the base of their favored feeding trees. Yet the hunting hypothesis remained intact; now, however, it was made to apply to the later stage in evolutionary history that began with the appearance of large brained Homo habilis.
The arguments for this theory reached full flower in the papers collected in Richard B. Lee and Irven DeVore's Man the Hunter (1968). The contributors sketched the following scenario. Protohominids encroach on the savanna by eking out their accustomed vegetarian diet with increasing amounts of hunted flesh Hunting puts a premium on foresight and dexterity, selecting for larger brains and nimbler hands. These traits increase the capacity for technology, raising the payoff of intelligence and augmenting the original selective pressure. Hunting becomes the engine of a self-sustaining cycle-of social and intellectual evolution.
This theory prevailed until the late 1970s, when an influential article by the late Glynn Isaac shifted the emphasis from the gathering of meat to the sharing of it [see "The Food-Sharing Behavior of Protohuman Hominids," by Glynn Isaac; SCIENTIFIC AMERICAN, April 1978]. Isaac, an archaeologist at the University of California at Berkeley, showed that early hominids had home bases a behavioral innovation which he argued, implied a sexual division of labor another innovation. To enhance the omnivorous strategy, males ranged far in search of scavengeable meat or hunted quarry, females gathered fruits and tubers nearer home and families shared the take. Eventually this altruistic behavior and social cooperation began to select for intelligence, language and culture.
Lewis R. Binford, now at Southern Methodist University, carried such analysis further in 1981. In a taphonomic reanalysis of Mary Leakey's data from the early Olduvai bone assemblages, Binford argued that neither hunting nor food sharing had evolved by H. habilis times. Hominids had merely processed the meager leftovers of more capable carnivores by breaking open bones to get at the marrow. He said scavenging could not have provided the surpluses of meat needed to sustain food sharing. Instead the social and nutritional aspects of protohominid feeding resembled the mainly vegetarian diets of modern apes.
Binford later argued, on similar grounds, that even the early modem Homo sapiens of southern Africa and the contemporaneous Neanderthals of Europe relied on scavenging to get large animals and hunted only small ones. Thus, Binford, too, retained the hunting hypothesis by moving it closer to the present within the past 100,000 years. His reconstruction accepts that scavenging was a penurious enterprise and that hunting and its attendant pattern of food sharing was a driving evolutionary force, albeit one that took effect very late in our evolution.
We began our critique of this entire approach by appraising the hunting prowess of early hominids. The physiques of Australopithecus and early Homo were unprepossessing. Females stood about four feet tall, males under five; females weighed about 70 pounds and males around 100. Their long arms suggest they still took refuge in trees. No doubt they had frequent occasion to do so, confronted as they were by such proficient predators as lions, sabertoothed cats and hyenas. As for their tools, even Homo wielded a very primitive kit of rough-hewn scrapers and unworked hammerstones. No true weapons are apparent.
Yet the archaeological evidence shows that these puny primates encroached on the large carnivores' niche. At Olduvai and elsewhere, archaeologists have found simple stone artifacts in association with fossilized bone fragments from animals as small as a gazelle and as big as an elephant. Surfaces of some of these bones bear the tooth marks of carnivores. Some of these and other bones also carry cut marks made when the associated tools were used to remove meat and disarticulate bones. Many bones are fractured and marked by a hammerstone, used to get at the marrow. Could protohumans have killed animals as fleet and formidable as these? We think scavenging deserves a closer look.
Proponents of hunting have argued that a diurnal hominid would have had difficulty in locating the kills of wider ranging predators and that any they might have chanced on would have been thoroughly eaten by hyenas, the only animal that can crush bones for marrow with its teeth. But these arguments miss two scavenging opportunities that we identified in Tanzania big- cat kills in riparian woodlands and the carcasses of very large animals that die of disease or by drowning. Hominids foraging in this habitat may have pieced together a niche no other scavenger could exploit so well .
Riparian woodlands would have suited partially arboreal bipeds by providing sanctuary and by hiding carcasses from vultures, the lead spotters of the scavenger tribe. Large ungulate carcasses crop up in these regions mainly in the dry season, when lions abandon their defleshed, zebra size kills. Leopard kills, on the other hand, consist of smaller ungulates and are available year-round. These kills are shielded best of all because they are typically stored in trees. Two million years ago saber-toothed cats may have provided hominids with a third opportunity, also in riparian woodlands. Kills of these extinct predators would have provided very large carcasses and abundant meat.
We hypothesize that scavenging may have been most important in the dry season, when plant foods are scarcest and scavenging opportunities are most diverse. Aside from leopard kills, wet season predation does not have the predictable riparian focus but instead is scattered in broader and more open habitats. Hyenas are quick to find and consume these exposed carcasses. Because scavenging may have made carnivory and herbivory seasonally complementary feeding strategies, we do not assume as proponents of Man the Hunter do that getting meat was the core of hominid adaptation. As dental evidence suggests, hominids have always been omnivores. The mere existence of stone tools and animal hones does not demonstrate that meat eating was common.
Yet scavenging may have made the dry season a time of plenty. It is then that starvation and predation produce many carcasses. Even the most marginal abandoned lion kill, retaining only marrow and brain, could provide much more than an adult's daily caloric requirements at the cost of half an hour's effort with a hammerstone. This rate of food return is higher than can be obtained by harvesting plants. If efficiency guided foraging decisions, hominids would have always preferred scavenging to harvesting, whenever scavenging was possible.
This preference would have been most marked at the height of the dry season, when plant productivity reaches its nadir and big-cat kills become predictable, shortening searches for these resources. A similar economy applies in the comparison with hunting: less energy is spent in picking one's food out of a tree or up off the ground if one does not first have to chase it.
Scavenging also incurs less risk than hunting. Any meat that attracts hominids can also attract lions, which, when they arrive, may ignore the dead in order to pursue the living. Our research showed, however, that large carnivores often leave certain classes of carcasses unattended for long periods. In the interim, the sites would have been safe.
Defleshed lion kills in riparian woodlands are particularly safe. We found that bone-crushing hyenas usually do not discover these carcasses until a day after they are abandoned by lions a good window of opportunity for any hominid capable of wielding a hammerstone. Tree-stored leopard kills provide more food (flesh as well as marrow) at less risk, especially when the cache contains several kills. Leopards tend to be solitary, and even a baboon or chimpanzee can sometimes scare one away. Moreover, leopards often abandon their kills voluntarily for as long as eight to 12 hours during the day, leaving some of them in a complete condition. Hominids in these woodlands would appear to have faced no more danger in scavenging than in foraging for plant foods in the same regions.
But risks may have outweighed benefits in the open plains, despite the plentiful opportunities for displacing the timid cheetah and jackal from their kills, exploiting abandoned lion kills in wet periods and benefiting from natural deaths during drought. The reason for this judgment is the scarcity of trees, which deprived the arboreally adapted hominids of sanctuary. Yet this drawback applies with at least equal force to hunting. Big herbivores have ways of defending themselves, and even if one can be killed, the conspicuousness of the killing quickly attracts scavengers. Many of these would have been more than a match for rock-wielding bipeds.
Hunting enthusiasts might respond that game is more wholesome than carrion. In the Serengeti, however, we found that few carcasses left on the ground retain scavengeable food as long as 48 hours, the time it takes for putrefaction to set in. But even then, most edible tissues remain encased in skin or bone that excludes insects and other postmortem disease vectors. Even carcasses produced by "natural" death generally carry no dangerous parasites, because most such deaths result from malnutrition, not disease.
Scavenging has also been faulted as nutritionally unsound. John D. Speth of the University of Michigan has suggested that animals that died of hunger would provide protein without enough fat for balance, a diet that can lead to a form of starvation. (Backwoodsmen called it "rabbit fever" because it came from living exclusively on rabbits and other lean game.) Yet hominids have always gotten most of their calories from the carbohydrates and oils of plants, and the most regular dry- season scavenging possibilities are predator kills of animals with fat in their marrow.
Which came first, scavenging or hunting? Answers have been proffered on ethological grounds, only to be controverted by new evidence. Hunting was uniquely human until Jane Goodall documented it in chimpanzees. Scavenging was beneath the dignity of a primate until workers discovered chimpanzees and baboons usurping the kills of cheetahs and leopards. It was foreign to human nature until 1988, when an ethnographic study of 20 years' duration documented avid scavenging by the Hadza and San foragers of sub-Saharan Africa. The delayed observation testifies to the prejudice against scavenging.
The earliest hominids probably scavenged and took small prey with their hands, as chimpanzees and baboons do. Only their next step was unique: they began to use tools to butcher large carcasses that nonhuman primates cannot exploit. The difficulty of this leap belies the charge that scavenging offers no challenge that might select for human qualities.
Our fieldwork suggests that scavenging is not at all easy for a slow, small, dull-toothed primate. To locate scavengeable carcasses before others did, we had to learn how to interpret the diverse cues to the presence of a carcass in riparian woodlands. They include the labored, low-level, early- morning, beeline flight of a single vulture toward a kill; vultures perched in mid-canopy rather than at the crown of a tree, where they nest; appendages of a concealed leopard or of its kill dangling from a branch; and tufts of ungulate hair or fresh claw marks at the base of a leopard's feeding tree. At night, the loud "laughing" of hyenas at a fresh kill, the panicked braying of a zebra being attacked, the grunting of a frightened wildebeest all serve notice of where to find an abandoned carcass when morning comes.
Higher primates make "mental maps" of their ranges and use them to predict where the next batch of fruit will ripen. Hominids might have applied this ready-made skill to predict the future availability and location of carcasses. We learned how to do it, with great effort. Every day we monitored the movements, hunting and feeding schedules, and belly sizes of predators, as well as the general activity of their prey. Apart from its possible nutritional payoffs, hominids might have used such information routinely to avoid predators.
Social skills would not have advanced, however, unless scavenging also selected for social cooperation Scavenged carcasses that fed only one individual, leaving no surplus to share, would probably have promoted competition. But if our research results are correct and big-cat kills gave early hominids a food surplus, then Isaac's model of cooperative foraging, processing and food sharing would work. Similarly, if such carcass foods did not usually coincide with plant foods, the emergent social skills might have expanded to include a division of labor, with corporate foraging about a common home base. To add to our ancestors' challenges, one need only hypothesize that they generally found carcasses in one place and stone for butchery tools in another. Uniting the tools with their objects would have thus required deep planning depth, detailed mental mapping and social cooperation.
West African chimpanzees are the only nonhuman primates that have enough planning depth to bring stone tools to food sources, as they do when they transport stone hammers and anvils to break the hard nuts of kola and Panda trees. Still, they do not carry the stone very far. H. habilis transported stone as far as 10 kilometers (six miles) and the nuts are not nearly as ephemeral as the scavengeable carcasses the early hominids butchered.
Technological skills necessary for exploiting most scavenging options are embodied in the earliest, Oldowan, tool kit [see box on opposite page] sharp edged stone flakes to deflesh and disarticulate and natural cobbles to break marrow bones and skulls. No tools clearly designed as weapons are apparent in either this complex or those of the more sophisticated tools of the Acheulean age, which ranged from 1.5 million to 200,000 years ago.
Such considerations lead us to conclude that Oldowan hominids may have created a scavenging niche that can account both for the earliest assemblages of tools and large mammal bones and for many uniquely human traits assumed to have arisen from hunting. A scavenging option may have started as a supplement to plant foraging, appearing in the following stages.
Hominids may have begun eating large animals long before Homo appeared. Australopithecus could have pioneered the strategy when it occupied the savannas and woodlands that became widespread by six million years ago, in the wake of global climatic change. These open environments would have offered far better scavenging opportunities than the closed woodlands and forests of hominid predecessors habitats in which the apes have remained to this day.
The earliest hominids may have come across defleshed kills while foraging for plants in thin ribbons of riparian woodlands. These resources would require only unmodified cobbles to extract the remaining marrow and brain from their bony cases. The butchery of these rich sources of energy and protein might have left a record that has so far eluded archaeologists because it occurred before the invention of flake tools, whose manufacture leaves a conspicuous litter of stone chips.
If so, diurnal hominids may have started to supplant hyenas by getting to the kill first. This hypothesis finds some confirmation in the extinction of several hyena species about two million years ago. The advent of a flaked stone technology by about 2.5 million years ago may then have enabled hominids to encroach on a new component of the large mammal scavenging niche. Now they could get flesh as well as marrow. With flaked stone, hominids held in their hands fabricated versions of the flesh shearing carnassial teeth of carnivores. With them, the hominids could butcher a leopard's tree-stored kills. The flesh from the much larger prey of the saber-toothed cats would also have been accessible, an observation that leads us to suggest that hominids may have had something to do with this species' extinction some 1.5 million years ago. It may be significant that these big cats persisted longer in Europe and the Americas than in Africa, going extinct only after hominids first colonized these continents.
Hunting of very small prey by hominids may have been an ancient strategy, and the late development of projectile weapons made early H. sapiens a predator more capable than any other primate. But scavenging has probably had a much more pervasive effect on human evolution than has hitherto been appreciated.