A few months after the diminutive hominin fossil Homo floresiensis, which because of its relatively large feet was quickly dubbed the ‘Hobbit’, turned out to be considerably older than previously thought it hit has the headlines again because its ancestors may have colonized the Indonesian island of Flores far earlier still. A pair of articles in the 9 June 2016 issue of Nature consider evidence from another site on the island where fluvial sediments offer more easily interpreted stratigraphy than the complex Liang Bua cave assemblage where the original skeletal remains were unearthed. The site in the So’a Basin became an important target for excavation following the discovery there in the 1950’s of stone artefacts, east of Wallace’s Line – a fundamental faunal and floral divide once thought to be due to the difficulty of crossing a deep, current-plagued channel in the Indonesian archipelago. The unexpected presence of artefacts drew palaeoanthropologists from far afield, but it was almost 50 years later before their exploration yielded hominin remains.

One of the papers reports sparse new finds of hominin material from the So’a Basin, a fragment of mandible and 6 isolated teeth thought to be from at least three individuals (van den Bergh, G.D. et al., 2016. Homo floresiensis-like fossils from the early Middle Pleistocene of Flores. Nature, v.  534, p. 245-248). The other covers newly discovered artefacts, the stratigraphic and palaeoecological setting, and radiometric dates of the finds (Brumm, A. and 22 others, 2016. Age and context of the oldest known hominin fossils from Flores. Nature, v.  534, p. 249-253). The jaw fragment shows signs of having once held a wisdom tooth, showing that it belonged to an adult. Yet although it resembles the dentition of the younger Liang Bua specimens, it seems more primitive and is even smaller. The other dental finds are most likely to be deciduous teeth of juveniles. Fission-track, uranium-series and ⁴⁰Ar/³⁹Ar dating indicates that the fossils entered the sediments about 700 ka ago. But tools and remains of prey animals in deeper sedimentary layers here and at other Flores sites indicate the presence of hominins back as far as about 1 Ma, before which there are no such signs.

So, at least a million years ago Flores was colonised by hominins. Either the original immigrants were uniquely small compared with other hominins of that vintage in Asia and Africa, or within 300 ka they had decreased in size through the evolutionary influence of limited resources on Flores and the process of island dwarfism. The second may also have been influenced by an initially small population of migrants or a later population ‘bottleneck’ that added a loss of genetic variability – a founder effect.   These two alternatives may point respectively to either the even earlier migration out of Africa and across most of Asia of perhaps H. habilis, or the dwarfing of a limited population of H. erectus who made their way there from their known occupation of Java. The authors painstaking analysis of the meagre remains suggest a closer dental resemblance to Asian Homo erectus than to earlier African hominins, so the second alternative seems more likely. However, even that scenario poses palaeoanthropology with a major problem; yet another evolutionary process that helps cryptify the links among our earlier relatives. (See also: Gomez-Robles, A., 2016. The dawn of Homo floresiensis. Nature, v.  534, p. 188-189.)