The early hominid Paranthropus robustus might not be as specialized an eater as researchers thought. Using a new laser technique, anthropologists peered into the teeth of these hominids to discover that the primate actually ate a variety of foods.

Between 2.4 and 1.4 million years ago, P. robustus roamed the African savanna. Researchers surmised that, because they had large molars with thick tooth enamel and strong jaw muscles, they ate low-nutrient, fibrous foods whereas their toolmaking relatives, Homo habilis and Homo erectus, ate softer foods such as fruit and meat. "A lot of things made these guys look like chewing machines," says Matt Sponheimer, an anthropologist at the University of Colorado at Boulder. "Human teeth, by comparison, are quite dainty." This apparently rigid diet was blamed for P. robustus's extinction later in the Pleistocene as the climate became drier and more seasonal.

Using 1.8-million-year-old teeth from South Africa, Sponheimer and his colleagues investigated the hominid's diet on a seasonal scale. Tooth enamel, laid down in layers, does not overturn quickly, Sponheimer explains, and yet gives a dynamic picture of the hominid's chemical makeup. Chemicals found in any animal come from what it ingests, such as plants. Plants turn sunlight into energy using photosynthesis, but they rely on two slightly different forms of the process. The two approaches add different amounts of the isotope carbon 13 to the plants' leaves. Bushes, shrubs and trees have less carbon 13 in their fruit or leaves than savanna grasses. The chemicals in the food eaten by P. robustus, including the isotope, were incorporated into the enamel of the hominid's permanent teeth as they formed.

Sponheimer and his colleagues used a laser to remove samples from each of the four P. robustus teeth, leaving only small marks on the precious specimens. Then, they analyzed the carbon isotope ratio in the samples. Instead of only eating tough woody food, the researchers report in this week's Science that Paranthropus ate a variety of foods, perhaps grasses, tubers and even animals--and what they ate varied seasonally and from hominid to hominid.

But if P. robustus relied on a flexible diet, then its food was not the reason it went extinct. Instead, Sponheimer raises other possibilities: If they had a more varied diet, they may have eaten the same foods as H. habilis or H. erectus. Perhaps this direct competition did them in. Or maybe P. robustus lived in smaller groups that were more susceptible to predators. Or possibly human ancestors outbred and overran them. "Really, the sky's the limit," Sponheimer says. "But the simple answer is we do not know."