The middle Pliocene gets crowded

New hominin fossils discovered in Ethiopia, dated to between 3.5 million and 3.3 million years ago, suggest that species diversity may have been as high during early human evolution as in later periods. See Article p.483

For many years, human evolution was viewed as a diverse radiation of species emerging a little after 3 million years (Myr) ago following an earlier phase characterized by little or no diversity (Fig. 1). Best known from this earlier period is the middle Pliocene species Australopithecus afarensis, which is documented by fossils between 3.7 and 3.0 Myr old from eastern Africa, including the emblematic skeleton known as Lucy1. On page 483 of this issue, Haile-Selassie et al.2 describe a new species, Australopithecus deyiremeda, which adds to other evidence3,4,5,6,7,8 indicating that A. afarensis was not the only ancient human relative during this time period. Indeed, it seems that the hominins that populated Africa in the middle Pliocene may have been just as diverse taxonomically as later stages of human evolution.

Figure 1: Temporal distribution of known hominin species.

Bars show when evidence of various hominin species occurs in the fossil record between 4 million and 2 million years (Myr) ago. The long-standing interpretation of hominin diversity was that Australopithecus afarensis and its probable ancestor A. anamensis formed a single lineage during the middle Pliocene (blue bars), and that this was followed by a radiation of Australopithecus, Homo and Paranthropus species after 3 Myr ago (green bars). However, more-recent fossil evidence indicates taxonomic diversity during the middle Pliocene (red bars), including Kenyanthropus platyops, multiple species of Australopithecus, and an unassigned foot fossil from Burtele, Ethiopia. Haile-Selassie and colleagues' announcement2 of the species Australopithecus deyiremeda adds to this early diversity. The earliest archaeological evidence for the use of stone tools, from the Lomekwi and Dikika sites, is also indicated.

In the 1970s and 1980s, it was hotly debated whether the middle Pliocene hominin fossils found in Ethiopia and Tanzania represent more than one species, but a broad consensus emerged that they could all be attributed to the single variable species A. afarensis1. Further fossil finds rekindled the debate about taxonomic diversity and led to the naming of two new species: Australopithecus bahrelghazali in 1996, based on a 3.6-Myr-old partial lower jaw from Chad3, and Kenyanthropus platyops in 2001, based on cranial fossils 3.5–3.3 Myr old from Lomekwi, Kenya4. The notion of multiple, contemporary middle Pliocene species was highlighted more recently by the discovery at Burtele, Ethiopia, of 3.4-Myr-old foot bones that are too primitive to belong to A. afarensis7, and by the revised dating of a South African Australopithecus skeleton to 3.7 Myr ago8 (Fig. 1).

It is against this background that Haile-Selassie and colleagues describe A. deyiremeda, which is represented by fossils between 3.5 Myr and 3.3 Myr old discovered at Woranso-Mille in Ethiopia. The location is close to sites of similar age that have produced abundant A. afarensis specimens1, suggesting that the two species lived in close proximity in place as well as in time.

The A. deyiremeda fossils include a partial maxilla (upper jaw bone) and two mandibles (lower jaw bones). The maxilla shows that the cheek bones were situated more forward than in A. afarensis, and the mandibles have a thick body because the ramus (the part connecting with the cranium) ascends from a more-forward position. The cheek teeth are notably small, except for the lower third molars. Such comparisons can be made with confidence because A. afarensis is known from one of the largest samples in the hominin fossil record, providing a good representation of morphological variation1.

But how does A. deyiremeda compare with the other middle Pliocene hominin species? As noted by Haile-Selassie and colleagues, it shares several diagnostic features with K. platyops, including forward cheek bones, small first and second molars, and three-rooted upper premolars4. However, the two species differ in the front part of the maxilla, which is flat and non-projecting in K. platyops but curved and protruding in A. deyiremeda and A. afarensis. Similarly, the A. deyiremeda mandibles share a forward position of the ramus with two possible K. platyops specimens from Lomekwi, but they are more like A. afarensis in the midplane. In broad terms, A. deyiremeda can be seen as sharing newly evolved morphology with K. platyops in the cheek areas of the jaws, while showing a more primitive, A. afarensis-like shape towards the front. Whether the retention of this primitive morphology in A. deyiremeda does indeed warrant species-level distinction from K. platyops must be confirmed by further study and fossil finds. Comparison with A. bahrelghazali is difficult because this species is known only from the front part of a mandible and an isolated premolar. It seems that one or more additional specimens have been recovered1,6 but these remain undescribed.

Finding such taxonomic diversity raises the question of how multiple species could have coexisted over a long period in a stable ecosystem, particularly when they live in close geographic proximity, as seems to be the case with A. deyiremeda and A. afarensis. Niche partitioning, involving diverse dietary preferences, foraging strategies, habitat selection and population movements, will probably be the key factor. However, establishing a concrete link between such characteristics and the morphological differences that distinguish species is often difficult, not least because the morphology may be affected by random genetic drift as much as by selection. A. deyiremeda, A. afarensis and K. platyops differ in their jaws in ways that reflect the position of the chewing muscles and the size of the tooth crowns. These differences provide an opportunity to investigate whether feeding behaviour and diet played a part, by modelling the biomechanics of chewing and assessing the dental wear and stable isotopes present in the fossils, both of which can give an indication of the types of food eaten by the individual.

Any discussion of niche partitioning and foraging behaviour among middle Pliocene hominins should consider the discovery of 3.4-Myr-old cut-marked bones at Dikika9 in Ethiopia, and of 3.3-Myr-old stone tools at Lomekwi10. Differences in the use of stone tools may reflect species-specific behavioural niches, so it would be informative if we could identify the Lomekwi toolmaker. The tools were found in close geographical and temporal proximity to where a key K. platyops specimen (the paratype) was discovered, making this species the default candidate.

However, a further complication now emerges: the K. platyops paratype is a maxillary fragment that was associated with the holotype (main defining specimen) on the basis of three shared features: forward cheek bones, three-rooted premolars, and a small first-molar crown4. But these three features are now also found in A. deyiremeda, and the distinctive front part of the maxilla that sets K. platyops apart is not preserved in the paratype. Hence, it cannot be excluded that it was A. deyiremeda that made an appearance at the tool site. Regardless, associating stone tools with a specific species as the toolmaker is notoriously difficult, and the increasingly rich fossil record of the middle Pliocene provides plenty of opportunity for lively debate.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Kimbel, W. H. & Delezene, L. K. Yearb. Phys. Anthropol. 52, 2–48 (2009).

    Article  Google Scholar 

  2. 2

    Haile-Selassie, Y. et al. Nature 521, 483–488 (2015).

    CAS  ADS  Article  Google Scholar 

  3. 3

    Brunet, M. et al. C. R. Acad. Sci. 322, 907–913 (1996).

    Google Scholar 

  4. 4

    Leakey, M. G. et al. Nature 410, 433–440 (2001).

    CAS  ADS  Article  Google Scholar 

  5. 5

    Spoor, F., Leakey, M. G. & Leakey, L. N. Phil. Trans. R. Soc. B 365, 3377–3388 (2010).

    Article  Google Scholar 

  6. 6

    Guy, F. et al. J. Hum. Evol. 55, 37–47 (2008).

    Article  Google Scholar 

  7. 7

    Haile-Selassie, Y. et al. Nature 483, 565–569 (2012).

    CAS  ADS  Article  Google Scholar 

  8. 8

    Granger, D. E. et al. Nature (2015).

  9. 9

    McPherron, S. P. et al. Nature 466, 857–860 (2010).

    CAS  ADS  Article  Google Scholar 

  10. 10

    Harmand, S. et al. Nature 521, 310–315 (2015).

    CAS  ADS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Fred Spoor.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Spoor, F. The middle Pliocene gets crowded. Nature 521, 432–433 (2015).

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.


Sign up for the Nature Briefing newsletter for a daily update on COVID-19 science.
Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing