Researchers used a clever trio of tricks to figure out what 2 million-year-old human ancestors ate by analyzing the stuff on their teeth, and they found something unexpected: They ate more like chimpanzees than like humans. Their analysis could point to a reason why our species triumphed while some of our long-lost cousins failed.

"Our results suggest that there was more variation among hominins around 2 million years ago, in terms of what they ate and where they lived," Amanda Henry of Germany's Max Planck Institute for Evolutionary Anthropology explained in an email, using the technical term for humans and their close fossil relatives.

The various species spread out into different environments, ranging from forests to grasslands, but each species had a preferred environment, in part due to diet specialization. "Then, with the evolution of Homo erectus around 1.8 million years ago, we see a big shift," Henry said. "Homo erectus thrived in a whole variety of environments, and was even able to migrate out of Africa."

Homo erectus eventually gave rise to Homo sapiens — that is, us. Meanwhile, the hominins that were tied down to Africa's forests — perhaps including Australopithecus sediba, the South African species that Henry and her colleagues studied — faded away in the competition with other apes. The latest research, published online today by the journal Nature, doesn't address that big evolutionary issue directly. But it does provide plenty of food for thought.

Decoding the diet
Australopithecus sediba was discovered in 2008, and has come to be seen as a transitional species between the relatively small, ape-like pre-humans known as australopiths and our closer ancient relatives in the genus Homo. Only a few specimens of the species have been recovered, but they appear to reflect chimp-like as well as human-like characteristics.

So which way did Au. sediba swing when it came to its diet? The researchers focused on the teeth of two specimens, a juvenile male and an adult female.

First, the researchers zapped the tooth enamel with a laser and analyzed the vapor that was given off. The composition of the enamel preserves a record of what the individual ate while it was growing up. In this case, Johns Hopkins University geochemist Benjamin Passey used a mass spectrometer to look at the balance of carbon-12 and carbon-13 isotopes in the vapor.

Lee Berger

The teeth of a juvenile male Australopithecus sediba had a dark layer of calculus, which is clearly visible in this close-up picture.

If the level of carbon-12 is relatively high, that suggests a diet heavy in forest products such as leaves, fruits and the critters that eat those things. Those are known as C3 foods. But if carbon-13 is high, that points to foods from the African savanna, such as seeds, roots and grasses. These are C4 foods.

"It seems like a hallmark of human evolution to be able to use savanna resources," Passey told me. "Today, most of our energy comes from grass in one way or another, either from grain or from animals that eat grain and grass."

Tests on sediments and various animal bones suggested that the area where the specimens were found was a savanna 2 million years ago. However, when Passey analyzed the samples from Au. sediba, he found that the carbon was almost pure C3, suggesting a diet heavy in forest products.

Two other lines of evidence confirmed that preference: Deposits of plaque on the teeth contained bits of silica, known as phytoliths, that were linked to forest vegetation. Also, the pits on the teeth were characteristic of creatures that favored a forest diet. All the evidence pointed to the conclusion that these particular pre-humans went to the forests for virtually all of their food, even though their remains were left behind in the savanna.

"This astonished us," Passey said.

The meaning of diet
Upon reflection, it shouldn't be that surprising that Australopithecus sediba passed up what the African savanna had to offer, said the University of Colorado's Matt Sponheimer, another co-author of the Nature paper. "We know living apes don't seem to like such foods much — or at least they are never a large portion of their diets," he told me in an email. "It is likely that this is at least partly due to their masticatory (and probably digestive) anatomy being suboptimal for such diets."

Modern-day savanna chimpanzees engage in similar behavior, the researchers noted.

The findings are consistent with the view that "early hominins were quite flexible with respect to diet, with different populations preferring different parts of the available plant resource base, depending on when they lived, and where," said Bill Kimbel, who serves as director of Arizona State University's Institute of Human Origins but was not involved in the Nature study. "The plant phytoliths are a nice addition to the repertoire of perspectives on the issue."

In an email exchange, Kimbel cautioned against reading too much into Au. sediba's preference, as reflected by the results from two individuals. He took issue with my suggestion that the results put the species in the "chimp camp" rather than the "human camp," and noted that other hominin species had different preferences for C3 vs. C4 foods.

"I don't think sediba should be seen as 'remarkable' in this context," he told me. "And sediba no more belongs in the 'chimp camp' than it does in the 'giraffe camp' (with which it also shares dominant C3 values). This is not a useful analogy."

Sponheimer declined to say whether he thought Australopithecus sediba died out because it didn't shift its diet to the foods of the savanna, "but one could make such an argument if indeed this was an organism tied to very specific microhabitats by a limited dietary repertoire." A different argument could be made as well: Perhaps the descendants of Au. sediba eventually adapted to a diet that took in C4 as well as C3 foods, and thus contributed to the rise of early Homo.

Passey, for one, would love to study more specimens. "It would be nice to analyze those and see if all members of that species have that same forest behavior," he told me. However those future experiments turn out, the findings reported today show how novel techniques — ranging from precision laser blasting to tooth-crud analysis — can shed light on the origin of our species.   

More about human origins:

• Was there a fork in our family tree?
• Humans had sex with now-extinct relatives
• Fossils shake up humanity's family tree
• Savannas may have nurtured human ancestors

In addition to Henry, Sponheimer and Passey, the authors of "The Diet of Australopithecus Sediba" include Peter S. Ungar, Lloyd Rossouw, Marion Bamford, Paul Sandberg, Darryl J. de Ruiter and Lee Berger. The paper, published online today, will appear in a future print edition of Nature.

Alan Boyle is msnbc.com's science editor. Connect with the Cosmic Log community by "liking" the log's Facebook page, following @b0yle on Twitter and adding the Cosmic Log page to your Google+ presence. You can also check out "The Case for Pluto," my book about the controversial dwarf planet and the search for new worlds.