Dmanisi is a tiny village located at the Georgian Caucasus, the crossroads of Europe and Asia. It is a remarkable site that challenges our understanding of early human expansion into Eurasia 1.8 million years ago, far from the homeland of our earliest ancestors.

A bit of history

1930   Start of the excavation works and study of a medieval fortress outside Dmanisi.

1983   A search for medieval relics beneath the churchyard produced the discovery of some teeth and bones of extinct animals from the early Pleistocene: a rhinoceros tooth and then bones from mammoth and saber-toothed cat.

1984   Stone tools made by humans start to appear. In total more than 2,000 stone tools have been recovered from the site.

1991   The first human fossil appeared: mandible D211 dated to almost 1.8 Ma, a time where no humans were previously thought to have left Africa. 

Key figures

• The site is almost 2,000 km away from any other hominid activity during the Gelasian age (2,588 to 1,806 million years ago).

• There are 4 meters of sedimentary deposits. The primary component is volcanic ashes from regional eruptions. All the specimens were found between two layers conclusively dated between 1.76 – 1.85 million years ago by radiometric methodology.

• 10,000 bones from about 50 extinct animals, including deer, bears and sabre-toothed tigers. Dmanisi sits atop a promontory by the confluence of two rivers, making it an attractive place for mammals but with no ways to escape from carnivores.

• 5 hominin skulls were found from 1991 to 2005. Many postcranial bones between 2005-2007: femur, tibia, kneecap, ankle, upper arm, five vertebrae, etc.  A toe bone was found in 2011. A complete male pelvis in 2014. The fossils are extremely well preserved.

• Along with the bones there is a complete lithic record of more than 2,000 tools. Their industry is surprisingly Oldowan: hominins in Africa at the same time period were making Acheulean tools such as hand axes.

• One fifth of the bones have signs of carnivore predation, while others have tool marks showing that the Dmanisi hominins were predator as well as prey. For example, Skull 5 (described below) was found beside a deer bone and a baby rhinoceros femur that had been chewed. But that deer bone had a stone flake tool embedded in it.

 The hominins

• Jaw D211 found in 1991, is a young adult. It preserves 16 teeth, is narrow and resembles Homo erectus in its mandibular corpus and the tooth crowns. Cranium D2282 found in 1999, matches D211 jaw (Skull 2). It is a gracile individual with light supraorbital tori and vault architecture suggesting to be a young female adult. It shares many morphological features with Homo erectus except for the small capacity c. 650 cc.

• D2280 (Skull 1) is a partial, small cranium, Homo erectus-like.

• Cranium D2700 and the complete jaw D2735 (Skull 3) correspond to a subadult similar to Homo habilis in cranial capacity (600 cc), brow ridge, occipital rounding and midfacial profile. The face is low, prognatic and concave. The dentition is large.

• Skull 4 (cranium D3444 and associated jaw D3900) also shares features with Homo habilis. This small individual lived until c. 40 years old. It had lost almost its entire dentition, which may have left him dependent on others for some years. This could be the earliest example of human compassion.

• D4500 cranium found in 2005 + D2600 jaw found in 2000 (Skull 5) are the most complete adult representative of early Homo ever recovered. The brain is tiny (546 cc) compared to the other Dmanisi specimens. The morphology is robust, the face is massive and projecting. Very tall jaw and very large molars. The incisors and canines show heavy wear. It shows male morphology however the pattern is unexpected given the tendency in higher primates for male to exceed females in brain size by 8 to 15%.  In the vault and cranial base there are some resembles to Homo erectus.

The controversy

The taxonomic identity and paleobiological significance of the Dmanisi assemblage remain controversial. Many of the similarities among the specimens are primitive and shared with early Homo and even with Australopithecines, for example the little cranial capacity could suggest to be descendants of Homo habilis.  Other characters are derived such as the dentition, the form of the brow ridge which is larger and bar-like, a small keeling on the vault, and details of the temporal bone construction. Postcranial remains  also resemble a modern, Homo erectus-like architecture.

The stratigraphic and taphonomic evidence showed that the hominin material was accumulated and then buried by ash falls over a relatively brief interval, probably between only 200-1000 years, in opposition to many other sites in Africa and Asia where the fossils are scattered through a long sequence of deposits covering a long period of time. However, a more recent review of the complex stratigraphy (Bermudez de Castro et al, 2012) suggests that the accumulation could cover an undetermined longer period of time, even several thousands of years.

Is this enough to consider that the sample variation represents more than one species? The controversy can be illustrated with the following different perspectives:

• The variation within Early Pleistocene Homo (erectus/ergaster, habilis, rudolfensis, georgicus) represent a single lineage, based on multivariate comparisons of the cranial shape and size among different samples -> defended by David Lordkipanidze, Christoph Zollikofer, Adam Van Arsdale, Milford Wolpoff…

• The hominins all seem to have a common body plan, all are drawn from just one population. But the Dmanisi assemblage provides a valuable perspective on taxonomic diversity as does not fall neatly into one of the taxonomic packages that have been proposed for Homo in the 2.5-1.5 Ma interval -> Philip Rightmire, John Hawks…

• The Dmanisi assemblage represent at least two (up to four?) early Homo species, based on the wide variation in features such as jaw and brow shape, which not only happens among Dmanisi skulls but also when compared with other Homo erectus fossils -> Jeffrey Schwartz, Ian Tattersall, Zhang Chi…

• In particular Skull 5 is pointed out as a species different from the other four skulls. Indeed, a new species Homo georgicus was defined (Gabunia et al. in 2002) with jaw D 2600 as the type specimen -> José María Bermúdez de Castro, María Martinón-Torres…

• The Dmanisi individuals are not different from Homo habilis. They may come from an early dispersal out of Africa 2.5 million years ago, although no evidence has been found yet -> Michael Chazan, Martha Tappan…

The future

The narrow window of the hominin deposits and the variability (and lack of knowledge) of early Homo tend to suggest a single lineage for the Dmanisi hominins. This lineage could have come from an early dispersal of early Homo out of Africa before 2 MYA which carried the Oldowan technology with them. However, if such window is demonstrated not to be so narrow, then Skull 5 could perfectly belong to a different population or even species that arrived to that site in a very different time. We have still much to discover and clarify in Dmanisi… Most of the site has not been excavated yet. The evolutionary position of the Dmanisi hominins will need to be re-assessed when having more finds and reaching a more integrated view between the cranial and the postcranial remains.

In the meantime, Dmanisi may no longer be the earliest hominin evidence in Eurasia. In 2015, Wei Qi and colleagues claimed to have found stone artifacts at the Heigugou site in the Nihewan basin, China, dated to 1.95-1.77 Ma.