The dawn of Homo floresiensis

New fossil findings demonstrate that the diminutive hominin Homo floresiensis lived on the Indonesian island of Flores at least 700,000 years ago, and may point to its rapid dwarfism from the larger Homo erectus. See Letters p.245 & p.249

Since the first description of Homo floresiensis in 2004 (ref. 1), these little hominins from the Indonesian island of Flores have raised very big questions. Do these skeletal remains represent a new species in the extinct hominin family, or are they modern humans who were pathologically dwarfed, or members of a short-stature population? If they belong to a different species, what was its evolutionary origin? Why was it so different from other hominin species? The most common answer to these questions has been repeated for more than ten years: we need more remains from Flores — especially from different sites and older time periods — to tip the scales. On pages 245 and 249 of this issue, van den Bergh et al.2 and Brumm et al.3 report the finding of those long-awaited remains.

After H. floresiensis was described, many palaeoanthropologists embraced the idea of a new and odd-looking hominin species that had a diminutive brain and body size. Supporters of the pathology hypothesis, however, have been unrelenting in looking for syndromes and conditions that could have been responsible for the unexpectedly small size of these hominins; some suggestions have been published quite recently4. The current findings — consisting of a lower-jaw fragment, an indeterminate cranial fragment and some small teeth from at least three different individuals — confirm beyond any reasonable doubt that H. floresiensis is a distinct hominin species with deep evolutionary roots that trace back more than 700,000 years.

Van den Bergh and colleagues present detailed analyses of the size and shape of the fossils, found at the Mata Menge site on Flores, including comparisons with remains of other hominin species. They show that such tiny teeth are found only in Homo sapiens — whose origin and migration to Asia are substantially later than the age of the new fossils — and in H. floresiensis. Brumm and colleagues report on the open-grassland habitat and stone tools associated with these hominins. They describe these tools as technologically similar to the ones found with the later H. floresiensis individuals from the Liang Bua site, and suggest that this points to the behavioural stability of the hominins from Flores over a long period of time. In addition, Brumm et al. use a combination of dating techniques to provide evidence that the fossils were deposited around 700,000 years ago, thus confirming the early origin of this species.

Although this confirmation finally ends the debate about the validity of H. floresiensis as a species, its evolutionary origins are likely to remain under discussion for much longer. There are two main models (Fig. 1). H. floresiensis may have evolved from the larger Homo erectus through a process of island dwarfing — an extreme reduction in size due to the absence of predators and to resource scarcity that is typical of island ecosystems. Alternatively, it may be descended from the earlier Homo habilis, or even from a small form of hominin from the Australopithecus genus. This second model implies that very primitive hominins would have left Africa by 2 million years ago, but there is no fossil or archaeological evidence for such an early dispersal.

Figure 1: Candidates for the ancestry of Homo floresiensis.

There are two main models for the evolutionary origin of the hominin species H. floresiensis, which inhabited the Indonesian island of Flores and had a particularly small brain and body. One possibility is that Homo habilis, or a similar form that also had a relatively small body and brain, may have left Africa by 2 million years ago and reduced in size even further. But there is no evidence for such early hominins outside Africa. Alternatively, H. floresiensis may be descended from the later and larger-bodied Homo erectus, for which there is evidence on Java around 1 million years ago and earlier. This second model would involve much greater body and brain reduction over a much shorter period of time. (Data on brain and body size are from refs 12–14, and are based on east African specimens for H. habilis, on early Indonesian specimens for H. erectus and on remains from Liang Bua for H. floresiensis.)

Mostly on the basis of the morphology of a lower molar tooth and of general affinities of the jaw fragment, van den Bergh and colleagues claim that the remains from Mata Menge are more closely related to H. erectus than to H. habilis. The reliability of lower-molar morphology to assess species relationships supports their claim5. However, the traits that point to a more primitive ancestor for H. floresiensis mostly come from body parts other than the skull6,7 and cannot be assessed using the Mata Menge sample, which does not include such postcranial remains.

Without further fossil evidence, the discussion between proponents of the two models will continue. Some will think that extreme dwarfing from H. erectus is unlikely, especially to the extent of the dramatic brain-size reduction observed in H. floresiensis8, although empirical data from hippopotamuses suggest that similarly strong brain reduction may occur9. Others will argue that a long-distance migration route for H. habilis, or an earlier form, from Africa to southeast Asia is even more implausible. For now, it seems that all possible explanations remain outside the comfort zone of classic scenarios of human evolution.

Van den Bergh et al. propose that the hominins from Mata Menge might be descended from the hominins that made stone tools at the site of Wolo Sege10, also on Flores, which is dated to approximately 1 million years ago. They further hypothesize that large-bodied H. erectus hominins are the ones that made these tools. This speculation could be proved wrong if remains from other small hominins were found with the tools in the future, but it raises an interesting question: can the extreme reduction of the brain and body of H. floresiensis have evolved over a mere 300,000 years?

Three hundred millennia may not seem a 'short' period of time to many readers. However, no other such dramatic transformation in hominin evolution is known to have occurred over a similarly brief timescale. A quantitative analysis and comparison of evolutionary rates across different hominin species and with H. floresiensis would lend formal support to this informal observation. Alongside such quantification, it might be helpful to look at more-distant species. Some mammals show evidence of even stronger degrees of dwarfing over substantially shorter periods of time, and extremely fast rates of size reduction in island environments11. In addition, we must not rule out the possibility that the direct ancestors of H. floresiensis were not the most typical representatives of their species. Indeed, the strange combination of primitive and derived traits in H. floresiensis anatomy could be the result of a pronounced founder effect, which occurs when a new population is established from a small sample that does not reflect the parental population's diversity and most-common traits.

Some scenarios that look mind-blowing from our anthropocentric point of view become underwhelmingly conventional when we expand our horizons. Rapid island dwarfism is not extraordinary in nature, nor is the founder effect or the long-scale migration of species that lack human-like cognitive abilities. Whatever the actual origin of H. floresiensis, we will be much closer to an answer if we look beyond hominins in our search for explanations.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Brown, P. et al. Nature 431, 1055–1061 (2004).

    ADS  CAS  Article  Google Scholar 

  2. 2

    van den Bergh, G. D. et al. Nature 534, 245–248 (2016).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Brumm, A. et al. Nature 534, 249–253 (2016).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Henneberg, M., Eckhardt, R. B., Chavanaves, S. & Hsü, K. J. Proc. Natl Acad. Sci. USA 111, 11967–11972 (2014).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Gómez-Robles, A., Bermúdez de Castro, J. M., Martinón-Torres, M., Prado-Simón, L. & Arsuaga, J. L. J. Hum. Evol. 82, 34–50 (2015).

    Article  Google Scholar 

  6. 6

    Jungers, W. L. et al. Nature 459, 81–84 (2009).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Tocheri, M. W. et al. Science 317, 1743–1745 (2007).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Martin, R. D., MacLarnon, A. M., Phillips, J. L. & Dobyns, W. B. Anat. Rec. 288A, 1123–1145 (2006).

    Article  Google Scholar 

  9. 9

    Weston, E. M. & Lister, A. M. Nature 459, 85–88 (2009).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Brumm, A. et al. Nature 464, 748–752 (2010).

    ADS  CAS  Article  Google Scholar 

  11. 11

    Evans, A. R. et al. Proc. Natl Acad. Sci. USA 109, 4187–4190 (2012).

    ADS  CAS  Article  Google Scholar 

  12. 12

    Kubo, D., Kono, R. T. & Kaifu, Y. Proc. R. Soc. B 280, 20130338 (2013).

    Article  Google Scholar 

  13. 13

    Grabowski, M., Hatala, K. G., Jungers, W. L. & Richmond, B. G. J. Hum. Evol. 85, 75–93 (2015).

    Article  Google Scholar 

  14. 14

    Baab, K. L. J. Anthropol. Sci. 94; see (2016).

Download references

Author information



Corresponding author

Correspondence to Aida Gómez-Robles.

Related audio

Gerrit van den Bergh explains how new fossils shed light on the mysterious hobbit.

Related links

Related links

Related links in Nature Research

Palaeoanthropology: What teeth tell us

Palaeoanthropology: Homo floresiensis from head to toe

Palaeoanthropology: Human evolution writ small

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gómez-Robles, A. The dawn of Homo floresiensis. Nature 534, 188–189 (2016).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.