By Ann Gibbons
Talk about a high-fiber diet: the newest member of the human family, Australopithecus sediba, ate enough bark, leaves, and fruit that its appetite was more like that of a chimpanzee's than a human's. That is the conclusion of a new study, in which an international team of researchers used state-of-the-art methods to analyze the diet of two australopithecines that fell into a death pit in Malapa, South Africa, almost 2 million years ago.
The creatures' preference for foraging in the forest surprised researchers, who thought that by this time in human evolution, most members of the human family would be eating a broader range of foods from different habitats, such as noshing on savanna grasses and tubers—or the animals that fed on them. Although our lineage had long since split from chimpanzees, Au. sediba's "diet looks as much like a chimp's or a giraffe's than anything else," says co-author Peter Ungar, a paleoanthropologist at the University of Arkansas, Fayetteville.
Ever since the discovery of the remarkably well-preserved partial skeletons of a female and juvenile male in 2008, researchers have wondered where Au. sediba fits in the human family tree—and if members of this species acted more like early members of our genus Homo or Australopithecus, which were both living in Africa at this time. Although classified as an australopithecine, the creatures were partly ape, with a tiny brain, long arms, a chimp-sized body, and a narrow birth canal. But they also were part human, with short fingers, a long thumb used for precision gripping, and a brain that had begun to reorganize more like a human's, prompting some to think they may be a candidate ancestor of our genus. So any new window on their behavior could help researchers understand what they were like "as animals that were once alive," says Ungar. "We thought, let's try to get at their diets to see what they ate and what type of habitat they lived in."
The fossils, which had been encased in calcified rock, were so well preserved that the researchers were able to use three different methods to analyze their diet. Most exciting, they realized that the tooth enamel still had plant phytoliths—microscopic plant fossils that form in distinctive shapes when different plants absorb silica. Paleoanthropologist Amanda Henry of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, extracted 38 phytoliths from two teeth of the young male, in the first extraction of phytoliths from an early hominin (previously, she had extracted phytoliths from Neandertals). She found phytoliths from fruit, leaves, wood and bark, as well as from grasses or sedges that grow in tropical forests or shade, but not open savanna.
Meanwhile, paleoanthropologist Matt Sponheimer of the University of Colorado, Boulder, and his colleagues used a laser to ablate tiny sections of tooth enamel, less than 1 millimeter long, which released two isotopes or forms of the element carbon, C and C. The ratio of those two isotopes reflects the types of vegetation the individuals ate as children when their teeth were forming. Plants with less C are so-called C plants such as fig trees and palms, which grow in the shade or woods. Plants with more C are C plants, including grasses that grow in the sun.
Sponheimer found that the Au. sediba individuals ate more C plants than any other hominin tested. The carbon in its diet was "unusual for hominins," and "more typical of giraffes." Among hominins tested, its diet most closely resembled that of Ardipithecus ramidus, a more primitive hominin that lived 4.4 million years ago in Ethiopia. Ungar also examined the patterns of microscopic wear, or microwear, on the teeth and found that Au. sediba ate harder foods similar to those Au. robustus and H. erectus ate, but not as soft as food typically consumed by Au. africanus, another South African species that some researchers thought was closely related to Au. sediba.
Altogether, "the three lines of evidence suggest these two individuals had an unexpected diet compared to those of African hominins of a similar antiquity," the authors report today in Nature. Au. sediba sought out leaves, fruits, and barks in the forest, rather than the grasses that were abundant in the area or the animals that fed on them. This suggests that Au. sediba preferred to feed in the forest, which fits with the ape-like parts of its anatomy, such as long arms, narrow hips and a primitive heel, that were adapted for climbing trees. Thus its diet most resembles that of a savanna chimp today. This underscores how unusual the broad diets of other hominins were at the time, says Ungar.
This sort of analysis is "the wave of the future in paleoanthropological research," says paleoanthropologist Mark Teaford of High Point University in High Point, North Carolina. The findings "force a reconsideration of the diet and thus of the behavior of our ancestral lineage, as well as our immediate ancestors," adds anthropologist Margaret Schoeninger of the University of California, San Diego. " Australopithecus sediba has a diet more typical of primates and of Ardipithecus ramidus than any of the other australopith diets reported so far." In other words, if Au. sediba was ancestral to humans (which many researchers think unlikely), it would suggest that our ancestors had a wider range of habitats and behavior later than expected in human evolution. "Various species were still exploring a whole range of habitats and environments," she says. "It's not just one type of adaptation."
The works also provides insight to Au. sediba's behavior, says Schoeninger. If it foraged in the woodland, the size of its social group would have been smaller, since it would have faced less fearsome predators than hominins that lived in more open terrain. And the fact that three lines of evidence all show that a member of the human family was still foraging in the woods just 2 million years ago "has to give us pause for thought," says Teaford, if only because scientists now have the tools to find out what was for supper 2 million years ago.
ScienceNOW, the daily online news service of the journal Science