Measuring up.
Researchers used six common cranial measurements, shown here on a modern human skull, to analyze the hobbit skull.

Lisa Nevell, Adam Gordon, and

Hobbit Skull Suggests a Separate Species

A new study of a "hobbit" skull found in Indonesia scores a point for those who believe that tiny hominids who once lived there should be classified as their own species. The study concludes that the 18,000-year-old skull from the island of Flores is unlikely to be a shrunken or diseased Homo sapiens, as some have argued, and that its ancestry may instead trace back to ancient Homo species in Africa.

The Indonesian hobbits have been a source of controversy since their bones were published in 2004. Some researchers have argued that the hominids were so unusual--with chimpanzee-sized brains, 1-meter-tall stature, and other peculiarities--that they deserve to be classified as their own species, H. floresiensis. Others contend that the hobbits are merely H. sapiens who evolved small stature because they lived on an island (ScienceNOW, 11 March) and that the one tiny skull belonged to a diseased individual (ScienceNOW, 5 March).

To help resolve the debate, paleoanthropologists Adam Gordon, Lisa Nevell, and Bernard Wood of George Washington University in Washington, D.C., compared the size and shape of other hominid skulls with that of H. floresiensis. This skull, named LB1 because it was the first specimen removed from Liang Bua cave on Flores, is astonishingly small, housing a 400-cc brain; in contrast, modern human brains average about 1400 cc. If the hobbits were merely small and/or diseased modern humans, the team reasoned, the skull might resemble that of a shrunken H. sapiens.

The researchers gathered published data on six measurements of skull shape, such as the height of the cranium and the forward jut of the jaw, on 2524 modern humans, 30 ancient hominids of various species, and the hobbit. Statistical analysis showed that the hobbit skull most resembled H. erectus skulls from Africa and Dmanisi, Georgia, dating as far back as 1.7 million years ago. Then, because the skull's tiny size presumably influences its shape in certain ways, the researchers did a second analysis considering the effects of scale--in effect asking what type of hominid, if shrunk to hobbit size, would best match LB1. In this part of the study, LB1 most resembled African H. habilis, the most primitive and small-brained species of our genus, also dated to about 1.7 million years ago.

"This is particularly exciting because ... it suggests that we really do have a hominin lineage that split off from our own as much as 1.7 million years ago, yet persisted up until the time when modern humans started peopling the Americas," says Gordon. "That's pretty cool." The team publishes its findings online today in the Proceedings of the National Academy of Sciences.

The study "makes a significant contribution to the debate," using statistical methods not used before, says Debbie Argue of Australian National University in Canberra, who has reported similarity in the jaw bones of LB1 and H. habilis. "Those of us who think H. floresiensis is a legitimate species are all wondering where it came from, and Gordon et al. nicely add to that discussion," agrees Dean Falk of Florida State University in Tallahassee.

But other researchers say the study doesn't explicitly address the diseased-human hypothesis. The paper has "a basic design flaw" because it includes no data on microcephaly, a genetic disorder causing a small head, says neurobiologist Christoph Zollikofer of the University of Zurich in Switzerland. Gordon responds that published data on microcephalics don't include all six measurements, so such specimens couldn't be included. The paper instead argues that it's unlikely that microcephaly or other pathologies would echo the cranial shape of ancient Homo species, as LB1 does.

Still, Zollikofer, an expert on the evolution of skull size and shape who has remained neutral in the hobbit wars, says this study's six measurements are too simple to capture the complexities of skull shape, a concern shared by others. In his view, this kind of analysis might cluster together skulls that are actually distinct. Depending on the species included, says Zollikofer, the approach could end up finding similarities between LB1 and chimpanzees.

Related site