Scavenging of "Peking Man"

Franz Weidenreich, who in the 1930s studied the fossils of Homo erectus unearthed in China, is caricatured along with Ralph von Koenigswald (wielding the shovel), who found fossils of H. erectus in Java. The fanciful setting is, according to the artist, "any place where the dead are disturbed."

New Evidence shows that a venerable cave was neither hearth nor home.

By Noel T. Boaz and Russell L. Ciochon


China is filled with archaeological wonders, but few can rival the Peking Man Site at Zhoukoudian, which has been inscribed on UNESCO's World Heritage List. Located about thirty miles southwest of Beijing, the town of Zhoukoudian boasts several attractions, including ruins of Buddhist monasteries dating from the Ming Dynasty (1368-1644). But the town's main claim to fame is Longgushan, or Dragon Bone Hill, the site of the cave that yielded the first (and still the largest) cache of fossils of Homo erectus pekinensis, historically known as Peking Man - a human relative who walked upright and whose thick skull bone and beetling brow housed a brain three-quarters the size of H. sapiens's.

A model of an H. erectus skull, based on fossils of several individuals from the Peking Man Site at Zhoukoudian. Most of the missing bones, represented in white, mirror existing parts of the opposite side of the skull.

The remains of about forty-five individuals - more than half of them women and children - along with thousands of stone stools, debris from tool manufacturing, and thousands of animal bones, were contained within the hundred-foot-thick deposits that once completely filled the original cave. The task of excavation, initiated in 1921, was not completed until 1982. Some evidence unearthed at the site suggested that these creatures, who lived from about 600,000 to 300,000 years ago, had mastered the use of fire and practiced cannibalism. But despite years of excavation and analysis, little is certain about what occurred here long ago. In the past two we have visited the cave site, examined the fossils, and carried out new tests in an effort to sort out the facts.

To most of the early excavators, such as anatomist Davidson Black, paleontologist Pierre Teilhard de Chardin, and archaeologist Henri Breuil, the likely scenario was that these particular early humans lived in the cave where their bones and stone tools were found and that the animal bones were the remain of meals, proof of their hunting expertise. Excavation exposes ash in horizontal patches within the deposits or in vertical patches along the cave's wall; these looked very much like the residue of hearths built up over time.

A more sensational view, first advanced by Breuil in 1929, was that the cave contained evidence of cannibalism. If the animal bones at the site were leftovers from the cave dwellers' hunting forays, he argued, why not the human bones as well? And skulls were conspicuous among the remains, suggesting to him that these might be the trophies of headhunter. Perhaps, Breuil even proposed, the dull witted H. erectus has been prey to a contemporary, advanced cousin, some ancestral from of H. sapiens. Most paleoanthropologists rejected this final twist, but the cannibalism hypothesis received considerable support.

In the late 1930s Franz Weidenreich, an eminent German paleoanthropologist working at Peking Union Medical College, described H. erectus remains in scientific detail. A trained anatomist and medical doctor, he concluded that some of the skulls showed signs of trauma, including scars and fresh injuries from attacks with both blunt and sharp instruments, such as clubs and stone tools. Most convincing to him and others was the systematic destruction of the skulls, apparently at the hands of humans who had decapitated the victims and then broken open the skull bases to retrieve the brains. Weidenreich also believed that the large longitudinal splits seen, for example, in some of the thighbones could only have been caused by humans and were probably made in an effort to extract the marrow.


The early invesitgations at Zhoukoudian were coordinated by the Cenozoic Research Laboratory in Beijing. Staff members included (left to right in foreground) Teilhard de Chardin, Franz Weidenreich, Yang Zhongjian, and Bian Meinian.

Others held dissenting views. Chinese paleoanthropologist Pei Wenzhong, who codirected the early Zhoukoudian excavations, disagreed with Breuil and suggested in 1929 that the skulls had been chewed by hyenas. Some Western scientists also had doubts. In 1939 German paleontologist Helmuth Zapfe published his findings on the way hyenas at the Vienna zoo fed on cow bones. Echoing Pei's earlier observations, of which he was aware, Zapfe convincingly argued that many of the bones found at sites like Longgushan closely resembled modern bones broken up by hyenas. In fact, a new term, taphonomy, was coined shortly thereafter for the field Zapfe pioneered: the study of how, after death, animal and plant remains become modified, moved, buried, and fossilized. Franz Weidenreich soon revised his prior interpretation of several H. erectus bones whose condition he had attributed to human cannibalistic activity, but he continued to argue that the long-bone splinter and broken skull bases must have resulted from human action.

Following disruptions in fieldwork during World War II (including the loss of all the H. erectus fossils collected at Longgushan up to that time, leaving only the casts that had been made of them), Chinese paleoanthropologists resumed investigation of the site. While rejecting the idea of cannibalism, they continued to look upon the cave as a shelter used by early humans equipped with stone tools and fire, as reflected in the title of paleoanthropologist Jia Lanpo's book The Cave Home of Peking Man, published in 1975.

About this time, Western scientists began to appreciate and develop the field of taphonomy. A few scholars, notably U.S. archaeologist Lewis R. Binford, then reexamined the Longgushan evidence, but only from a distance, concluding that the burning of accumulated bat or bird guano may have accounted for the ash in the cave. With the founding in 1993 of the Zhoukoudian International Paleoanthropological Research Center at Beijing's Institute of Vertebrate Paleontology and Paleoanthropology, a new era of multidisciplinary and international research at Longgushan began. At the institute, we have been able to collaborate with paleontologists Xu Qinqi and Liu Jinyi and with other scholars in a reassessment of the excavations.

One of taphonomy's maxims is that the most common animals at a fossil site and/or the animals whose remains there are the most complete are most likely the ones to have inhabited the area in life. Standing in the Beijing institute amid row after row of museum cases filled wit mammal fossils from the cave, we were immediately struck by how few belonged to H. erectus - perhaps only 0.5 percent. This suggests that most of the time, this species did not live in the cave. Furthermore, none of the H. erectus skeletons is complete. There is a dearth of limb bones, especially of forearms, hand, lower leg bones, and feet - indicating to us that these individuals dies somewhere else and that their partial remains were subsequently brought to the cave. But how?

Franz Weidenreich at his labartory at the American Museum of Natural Histoy in the 1940s with ape and human skulls.
During the 1930s, excavators dug down through the hundred-foot-thick deposits that contained the remains of "Peking man." The deposits, which also yielded animal bones, stone tools, and layers of ash, had completely filled an ancient cave.

The answer was suggested by the remains of the most common and complete animal skeletons in the cave deposit: those of the giant hyena, Pachycrocuta brevirostris. Had H. erectus, instead of being the mighty hunters of anthropological lore, simply met the same ignominious fate as the deer and other prey species in the cave? This possibility, which had been raised much earlier than PEI and Zapfe, drew backing from subsequent studies by others. In 1970, for example, British paleontologist Anthony J. Sutcliffe reported finding a modern hyena den in Kenya that contained a number of human bones, including skulls, which the animals had apparently obtained from a nearby hospital cemetery. In the same year, South African zoologist C. K. Brain published the findings of his extensive feeding experiments with captive carnivores, akin to those of Zapfe three decades earlier. One of Brain's conclusions was that carnivores tend to chew up and destroy the ends of the extremities, leaving, in the case of primates, very little of the hand and feet.

To test the giant hyena hypothesis, we examined all the fossil casts and the few actual fossils of H. erectus from Longgushan. We looked for both carnivore bite marks and the shallow, V-shaped straight cuts that would be left by stone tools (although we realized that cut marks would probably not be detectable on the casts). We also analyzed each sample's fracture patterns. Breaks at right angles indicate damage long after death, when the bone is fossilized or fossilizing; fractures in fresh bones tend to be irregular, following natural structural lines. Breakage due to crushing by cave rocks is usually massive, and the fracture marks characteristically match rock fragments pushed into the bone.

We were surprised by our findings. Two-thirds of Longgushan's H. erectus fossils displayed what we are convinced are one or more of the following kinds of damage: puncture marks from a carnivore's large, pointed front teeth, most likely the canines of a hyena; long scraping bite marks, typified by U-shaped grooves along the bone; and fracture patterns comparable to those created by modern hyenas when they chew bone. Moreover, we feel that the longitudinal splitting of large bone - a feature that Weidenreich considered evidence of human activity - can also be attributed to a hyena, especially one the size of the extinct Pachycrocuta, the largest hyena known, whose preferred prey was giant elk and woolly rhinoceros. One of the H. erectus bones, part of a femur, even reveals telltale surface etchings from stomach acid, indicating it was swallowed and then disgorged.

A composite image of the skulls of Pachycrocuta and H. erectus, left,shows how the giant hyena may have attacked the face. Beneath is a disgorged piece of an H. erectus thighbone.


The pattern of damage on some of the skulls sheds light on how hyenas may have handled them. Bite marks on the brow ridge above the eyes indicate that this protrusion had been grasped and bitten by an animal in the course of chewing off the face. Most animals' facial bones are quite thin, and modern hyenas frequently attack or bite the face first; similarly, their ancient predecessors would likely have discovered this vulnerable region in H. erectus. Practically no such facial bones, whose structure is known to us from discoveries at other sites, have been found in the Longgushan cave.

The rest of the skull is a pretty tough nut to crack, however, even for Pachycrocuta, since it consists of bones half again as thick as those of a modern human, with massive mounds called tori above the eyes and ears and around the back of the skull. Puncture marks and elongated bite marks around the skulls reveal that the hyenas gnawed at and grappled with them, probably in an effort to crack open the cranium and consume the tasty, lipid-rich brain. We concluded that the hyenas probably succeeded best by chewing through the face, gaining a purchase on the bone surrounding the foramen magnum (the opening in the cranium where the spinal cord enters), and then gnawing away until the skull vault cracked apart or the opening was large enough to expose the brain. This is how we believe the skull bases were destroyed - not by the actions of cannibalistic H. erectus.

We know from geological studies of the cave that the animal bones found there could not have been washed in by rains or carried in by streams: the sediments in which the bones are found are either very fine-grained - indicating gradual deposition by wind or slow-moving water - or they contain angular, sharp-edged shards that would not have survived in a stream or flood. Some of the bones may have belonged to animals that died inside the cave during the course of living in it or frequenting it. Other bones were probably brought in and chewed on by hyenas and other carnivores.

Cut marks we observed on several mammal bones from the cave suggest that early humans did sometimes make use of Longgushan, even if they were not responsible for accumulating most of the bones. Stone tools left near the cave entrance also attest to their presence. Given its long history, the cave may have served a variety of occupants or at times have been configured as several separate, smaller shelters. Another possibility is that, in a form of time-sharing, early humans ventured partway into the cave during the day to scavenge on what the hyenas had not eaten and to find temporary shelter. They might not have realized that the animals, which roamed at twilight and at night, were sleeping in the dark recesses a couple of hundred feet away.

An artist's depiction of the cave shows hyenas consuming the remains of an H. erectus.


What about the ash in the cave, which has been taken as evidence that H. erectus used fire? Recently published work by geochemist Steve Weiner and his team at the Weizmann Institute of Science in Israel suggests that the fires were not from hearths. In detailed studies of the ash levels, they discovered no silica-rich layers, which would be left by the burning of wood. Wood (as well as grass and leaves) contains silica particles known as phytoliths - heat-resistant residues that are ubiquitous in archaeological hearth sites. The results indicate that fire was present in the cave but that its controlled use in hearths was not part of the story.

Still, a human hand may somehow be implicated in these fires. One possibility we are exploring in the next phase of our research is that Longgushan was a place where Pachycrocuta and H. erectus confronted each other as the early humans sought to snatch some of the meat brought back to the cave by the large hyenas. Pachycrocuta would have had the home court advantage, but H. erectus, perhaps using fire to hold the carnivore at bay, could have quickly sliced off slivers of meat. Although today we might turn up our noses at such carrion, it may have been a dependable and highly prized source of food during the Ice Age.