Research Today | Science
A small fragment adds weight to a major human origins question
A fragment of skull from South Africa’s Cradle of Humankind is adding evidence to a long-running question about when Homo erectus first appeared in southern Africa. The specimen is not a complete skull, and researchers are careful about its classification. Yet it matters because it appears to belong to an adult from deposits dating to roughly two million years ago, where the species was previously represented by a young child’s skull.
A peer-reviewed paper published online on 25 September 2026 in Annals of Human Biology describes 18 hominin fossils from Drimolen Main Quarry. Of these, researchers assigned 14 to Paranthropus robustus, one tentatively to Homo erectus in the broad sense, and three only to an unidentified hominin. The findings enlarge the anatomical record for two different branches of the human family tree and allow more detailed comparisons between ancient populations.
Why Drimolen is important
Drimolen Main Quarry lies in the Cradle of Humankind World Heritage landscape, around 40 kilometres northeast of central Johannesburg. Excavations began after the site’s discovery in 1992. Over time, the quarry yielded a substantial fossil collection, including important skulls of Paranthropus robustus and a juvenile specimen, DNH 134, previously attributed to Homo erectus.
The new paper reports that 119 hominin specimens had been recovered from the site. But recovery and scientific description are different stages of research. Some material was carefully excavated decades ago yet remained incompletely described, leaving potentially useful anatomical information out of published comparisons. The researchers returned to this backlog to document fragmentary skull and jaw material, not to announce that all 18 specimens had just been discovered underground.
What the researchers actually examined
Jesse M. Martin, A. B. Leece, Stephanie Baker, Andy I. R. Herries and David S. Strait studied 18 previously undescribed or incompletely described cranial and mandibular specimens excavated between 1992 and 2019. Their approach was comparative anatomy, also called descriptive morphology. They examined preserved bone shapes, sutures, ridges and jaw structures, comparing them with fossils from Drimolen and other South African sites, particularly Swartkrans and Kromdraai. They supplemented visual assessments with anatomical measurements.
The team used original comparative fossils curated at the University of the Witwatersrand and the Ditsong National Museum of Natural History, as well as casts and published descriptions of material from elsewhere in Africa and Eurasia. This is a study of physical fossil anatomy, not a DNA analysis, a clinical experiment or a statistical survey of living people. Its conclusions depend on how much diagnostic anatomy survives and how well the comparative collections represent variation within ancient species.
The 18 fossils, counted carefully
The central numerical finding is straightforward: 14 of the 18 specimens were assigned to Paranthropus robustus, one to Homo erectus sensu lato, and three could not be identified confidently to genus and species. That corresponds to approximately 77.8%, 5.6% and 16.7% of this particular described group, respectively. The percentages describe a selected fossil collection, not the proportions of animals living in the region two million years ago.
Although Paranthropus fossils greatly outnumber Homo fossils in the assemblage, preservation, recovery, the history of excavations and which bones are identifiable all affect the count. The research therefore does not establish a reliable population census or prove that the two species lived at a precise ten-to-one ratio. It does, however, show that the newly described sample is strongly dominated by material attributable to Paranthropus.
DNH 127 and the case for an adult Homo erectus
The most consequential specimen is DNH 127, a partial skull preserving pieces of the temporal, parietal and occipital bones near the back and side of the head. These bones meet around the area known anatomically as the asterion. The fossil is fragmentary: it does not include a complete face, jaw or braincase from which a comprehensive reconstruction could be made.
The authors describe several anatomical clues. The preserved parietal wall extends as much as 44.7 millimetres above one skull suture, and its maximum measured thickness is 7.2 millimetres. The shape of the mastoid region and the relatively limited overlap between skull bones do not fit well with their comparison sample of Paranthropus. A straighter suture pattern resembles features seen in Homo erectus, including a comparative specimen from Dmanisi in Georgia.
These details matter collectively rather than individually. The authors explicitly describe the assignment as tentative and use the qualification sensu lato, meaning Homo erectus in a broad taxonomic sense. The study adds evidence to an existing interpretation; it does not eliminate all uncertainty about the fossil’s identity.
How old is the skull fragment?
The researchers place the relevant Drimolen deposits at approximately 2.04 million to 1.95 million years old. That is a range of about 90,000 years, rather than a single calendar date. The site’s age model draws on geological context and earlier dating work, including a magnetic reversal used to constrain some deposits.
There is an important qualification. DNH 127 came from a loose, partly decalcified breccia block near the DNH 7 pinnacle, rather than from a perfectly preserved in-situ position. The researchers consider it likely that the block broke away from the nearby deposit and that the specimen is broadly contemporary with other fossils from that part of the quarry. This is a reasoned geological attribution, not a direct measurement of the fossil’s individual age.
Some other Drimolen fragments were recovered from disturbed mining rubble, for which the authors say precise dating is less secure. A responsible account must distinguish the general age of the site from the confidence attached to each fossil’s original position.
Why the adult specimen changes the picture
The earlier Drimolen fossil DNH 134 was a young individual, estimated to have died at roughly two to four years of age. Juvenile skulls differ from adult skulls because bones, proportions and muscle attachments change during growth. Having only a young individual therefore limited comparisons with adult Homo erectus remains elsewhere.
DNH 127 supplies a second specimen from the Drimolen Homo record and, crucially, adult anatomy. The team argues that this strengthens the case for the species’ presence in southern Africa around two million years ago. For context, well-known adult specimens from Dmanisi in Georgia are around 1.8 million years old, approximately 200,000 years younger than the broad Drimolen age estimate.
That comparison should not be exaggerated. An older specimen at one site does not, by itself, establish where a species evolved or the route by which it spread. It is evidence about where an early population lived, not a final answer to the geographic origin of Homo erectus.
A second story: change within Paranthropus robustus
The 14 specimens assigned to Paranthropus robustus also address a separate evolutionary question. This extinct human relative had large grinding teeth and powerful chewing anatomy. The researchers compared older Drimolen material with younger samples from Swartkrans and Kromdraai, including deposits around 1.8 million years old.
The anatomical comparisons support a hypothesis that the Drimolen population was distinct from the later samples. The paper identifies an additional jaw characteristic that may help separate the groups. The authors interpret the pattern as compatible with microevolution, meaning relatively small changes in anatomy within a lineage over evolutionary time, rather than the sudden appearance of an entirely unrelated species.
In broader comparisons, later Paranthropus material shows changes associated with larger teeth and more substantial chewing structures. The paper supports the proposed distinction between the Drimolen group, sometimes labelled P. robustus ukusa, and younger P. robustus robustus material. The strength of that classification will depend on future fossils and continued debate about variation within the species.
What the fossils cannot tell us about diet and climate
Large molars and robust jaws can suggest mechanical capacity for processing certain foods, but bones do not directly reveal every meal. Nor does a comparison between sites prove that a specific climatic change caused a particular anatomical trait. Dietary interpretations require additional evidence, such as tooth wear, chemical signatures, ecological reconstruction and ideally independent dating.
The same caution applies to the contrasting trajectories often described for Homo erectus and Paranthropus robustus. They differed in cranial and dental anatomy, but this study does not measure brain size for a complete newly discovered Homo erectus skull, nor does it test whether one diet was more successful than another. The paper’s direct contribution is the documentation and comparison of fossil bones.
Why this matters beyond palaeoanthropology
The research illustrates how scientific advances can come from careful re-examination of existing collections. The newly described material includes fossils recovered long before this paper appeared. Publishing descriptions of small or damaged specimens allows other researchers to test classifications, compare anatomical traits and revisit hypotheses without relying only on spectacular, nearly complete skulls.
It also underscores South Africa’s continuing importance to the global history of human origins. Scientists affiliated with the University of Johannesburg and the University of the Witwatersrand contributed to the work alongside international collaborators. The country’s fossil repositories are not merely tourist attractions; they are research infrastructure that preserves evidence for future generations.
For readers interested in other methods of reconstructing the human past, Research Today has separately examined what traces of ancient human lineages can reveal through DNA. That is a different kind of evidence from the physical skull and jaw anatomy analysed at Drimolen.
The limitations researchers must work around
The sample is only 18 specimens, many of them fragmentary. Species identification is consequently less certain than it would be with complete skeletons. DNH 127’s tentative attribution is especially important to preserve in headlines and interpretation. The fossil’s recovery from a loose breccia block also means its association with the dated deposit is inferred from context rather than recorded as an undisturbed position.
Comparisons between Drimolen, Swartkrans and Kromdraai may reflect chronological change, geographic variation, age, sex, preservation or some combination of these. Fossils are not randomly sampled from ancient populations. The observed dominance of Paranthropus among identified remains should therefore not be read as an exact historical population ratio. Nor do the results prove that southern Africa was the evolutionary birthplace of Homo erectus.
These qualifications do not erase the contribution. They define it: the paper expands a sparse anatomical record, strengthens an early southern African occurrence of Homo erectus and provides further comparative evidence for variation within Paranthropus robustus.
What comes next
More securely dated adult fossils would help clarify whether DNH 127 represents Homo erectus in a narrow sense or belongs within a more complicated early Homo diversity. Additional jaws and skull fragments could test whether the differences between early and later Paranthropus populations are consistent across individuals and anatomical regions. Equally important is the painstaking publication of material already held in collections.
The headline result is therefore not that one fossil has settled human evolution. It is that 18 documented specimens, including a rare adult fragment, have made the evidence from South Africa’s two-million-year-old landscape more informative and more open to scientific scrutiny.
Source Information
Study title: New hominin cranial and mandibular remains from Drimolen Main Quarry, South Africa.
Authors: Jesse M. Martin, A. B. Leece, Stephanie Baker, Andy I. R. Herries and David S. Strait.
Journal: Annals of Human Biology, volume 53, issue 1, article 2725046.
Published online: 25 September 2026.
Research method: Comparative descriptive morphology, qualitative anatomical assessment and quantitative measurements of fossil cranial and mandibular remains.
Research location: Drimolen Main Quarry, Cradle of Humankind, South Africa, with comparative material from other fossil sites.
Study material: 18 previously undescribed or incompletely described hominin fossils, including 14 attributed to Paranthropus robustus, one tentatively to Homo erectus sensu lato and three not identified to genus and species.
DOI: 10.1080/03014460.2026.2725046.
