Large Footprints, Uncertain Identities: Rethinking Paranthropus at Lake Turkana

Twenty-one hominin footprints cross an ancient surface at GaJi10, on the eastern side of Kenya’s Lake Turkana. Fourteen were unknown when the first excavations exposed tracks in 1978–79. Together, they preserve evidence of several individuals moving through a lakeshore environment approximately 1.43 million years ago. The impressions have become the centre of an awkward anatomical puzzle: their shapes resemble tracks attributed to Paranthropus boisei, while their dimensions suggest bodies larger than skeletal reconstructions had led researchers to expect.

The study, led by Kevin Hatala and published in Proceedings of the National Academy of Sciences on 27 July 2026, retains two explanations. Either P. boisei was larger and more variable than previously recognised, or Homo erectus possessed a much wider range of foot function than existing comparisons suggest. That unresolved attribution is central to the discovery. A large footprint is direct evidence of a large impression; its maker’s species, stature and social identity require further argument.

The Max Planck announcement describes eight probable adults, with estimates for the largest reaching roughly 1.8 metres and 75 kilograms. These are reconstructed dimensions, not measurements of an excavated skeleton, and they should not become a new species average. Nevertheless, the possibility of such individuals changes the comparisons researchers must make when interpreting fragmentary early hominin remains.

For archaeology, the larger interest lies in bringing bodies back into landscapes. Footprints connect anatomy with movement and location. They can also place several individuals within a narrow interval, something that scattered bones rarely accomplish.

A body assembled from incomplete evidence

Paranthropus is often described as robust, but that designation refers especially to its heavily built jaws and enlarged chewing teeth. It does not independently establish a large body. These hominins represent branches of the human family distinct from our own, and the relationship between their conspicuous skull anatomy and their less familiar skeletons has long been difficult to reconstruct.

The difficulty begins with identification. Teeth can carry features that allow a confident species assignment. An isolated arm or leg bone may lack equally diagnostic characteristics, particularly where several hominins occupied the same region. A limb found near a recognisable jaw is therefore valuable, but proximity still has to be tested against stratigraphy, preservation and the possibility that different individuals accumulated together.

The partial skeleton OH 80, reported from Olduvai Gorge in Tanzania in 2013, illustrates that process. Researchers associated teeth and fragmentary limb bones through their restricted distribution, compatible preservation and absence of duplicated skeletal elements. Diagnostic teeth supported attribution to P. boisei; the limb fragments indicated a strongly built individual. The evidence had already complicated a uniformly small-bodied reconstruction well before the latest footprint analysis.

There is a methodological danger here. If large isolated bones are assigned to Homo partly because Homo is expected to be large, those assignments can reinforce the expectation used to make them. The appropriate response is to seek independent anatomical associations and additional kinds of evidence. Tracks offer one such test, although they introduce their own identification problems.

A footprint records a moving foot

The methodological background includes Hatala and colleagues’ 2024 study of approximately 1.5-million-year-old tracks at Koobi Fora. Their analysis identified two patterns of bipedal movement on the same surface and comparable diversity at other sites. They interpreted these as evidence that H. erectus and P. boisei shared lake-margin habitats. The GaJi10 research applies this developing comparative framework to a slightly younger assemblage.

Reading these differences requires more than matching an outline to a modern bare foot. As a foot enters sediment, supports weight and withdraws, it moves material. The resulting depression records that sequence. Its shape depends on motion as well as anatomy, so a preserved contour cannot automatically be read as the contour of the living foot.

Experimental work published in 2023 made this distinction explicit. Using paired X-ray views, three-dimensional animation and particle simulations, Hatala, Stephen Gatesy and Peter Falkingham investigated how tracks form. They found that an arch-shaped footprint is not a straightforward indicator of an anatomical arch: characteristic foot movements can generate that form. Their comparisons distinguished aspects of locomotion recorded at Laetoli from the more modern pattern represented at Ileret.

This changes what the Kenyan evidence can establish. Repeated differences between tracks can support the presence of different walking mechanics. Attaching a species name requires an additional comparison with the fossil record. In the 2026 paper, the authors explicitly preserve an alternative H. erectus attribution, rather than treating the preferred Paranthropus identification as settled.

The distinction also prevents an evolutionary misunderstanding. A walking pattern different from ours does not, by itself, demonstrate defective or inefficient locomotion. Efficiency depends on the animal’s anatomy, activities and environment. The tracks document diversity; they do not rank the success of its participants.

From foot dimensions to kilograms

Estimating body size introduces another layer of comparison. Researchers examine relationships between feet and bodies in living people, then use those relationships to interpret fossil impressions. The approach has an empirical basis, but the relationship is statistical: people with similar foot dimensions do not all have identical heights or masses.

A 2021 study led by Christopher Ruff investigated 193 adults and 50 juveniles from five habitually barefoot or minimally shod populations. Foot size correlated strongly with body mass; predictions using foot area had an average error of approximately 10 per cent in the combined sample. That figure describes performance in the studied modern sample. It is not a guaranteed uncertainty range for an extinct species.

Applying such relationships across species requires assumptions about proportions. A hominin with relatively large feet could produce an overestimate if reconstructed using a differently proportioned reference population. Sediment deformation and variation between successive steps add further uncertainty. The useful conclusion is therefore broader than any single number: the GaJi10 impressions deserve comparison with a wider range of possible bodies.

Earlier Ileret research shows why this matters historically. A 2016 analysis interpreted large tracks and modern-looking locomotor patterns as evidence of H. erectus. Its authors acknowledged the very limited skeletal samples available for comparison with other hominins. Expanding those samples can alter how persuasive body size is as an identification criterion, without invalidating every earlier track assignment.

Eight individuals are not a census

The proposed social interpretation is more conditional still. The researchers suggest that the GaJi10 assemblage included several adult males, potentially travelling together. In the institutional announcement, Neil Roach raises the possibility that males tolerated one another for protection while also competing for mates. This is a behavioural hypothesis drawn from the apparent composition of the assemblage, rather than an activity directly visible in the tracks.

Size-based sex attribution depends on knowing how male and female dimensions overlapped. If the species’ size distribution is itself being revised, that assignment necessarily remains provisional. Even a securely identified group of males would leave several possibilities open: temporary association, joint foraging or membership of a larger social unit.

Nor does an excavated surface necessarily preserve every individual who passed nearby. Earlier footprint research explicitly recognised that sediment strength can determine which body sizes leave discernible tracks. A lighter individual may register poorly where a heavier one leaves a clear impression. Preservation therefore complicates claims that females or juveniles were absent.

The interpretive distinction is consequential. Several individuals using the same patch of ground within a short interval offer a basis for investigating social tolerance. Demonstrating stable group membership, cooperation or a mating system would require a stronger and more repeated pattern. Additional surfaces could show whether the apparent grouping recurs; a single exposure cannot supply an entire social organisation.

Why the lakeshore belongs in the explanation

The earlier Koobi Fora study places this problem within a landscape occupied by multiple hominins. Its footprint evidence supports habitat overlap between the proposed species, raising questions about competition and the division of resources. Occupying the same shoreline does not establish that they ate the same foods or used it in the same way.

Independent dietary evidence helps frame those questions. In 2011, Thure Cerling and colleagues analysed carbon isotopes in P. boisei tooth enamel. Their results indicated substantial reliance on carbon derived from C4 plants, including tropical grasses and sedges. Carbon isotope analysis distinguishes broad dietary sources; it does not identify a particular meal or plant part. The finding makes plant resources relevant to interpretations of wetland use, but cannot demonstrate what the GaJi10 walkers were doing.

Recent skeletal research adds another independent perspective. In 2025, Carrie Mongle and colleagues described hand and foot bones securely associated with P. boisei craniodental remains at Koobi Fora. The hand combined manipulative capabilities with features consistent with powerful grasping and manual food processing. The authors considered some tool manufacture and use anatomically plausible. Such capability does not identify the maker of a particular stone assemblage, but it gives future interpretations of behaviour a firmer anatomical foundation.

These evidence streams should constrain one another. Teeth inform diet, associated bones inform anatomy, and tracks record movement through a particular setting. None can substitute completely for the others.

The immediate task is to test whether the GaJi10 combination of large dimensions and distinctive foot mechanics appears repeatedly. Securely associated skeletons could establish the proportions needed to improve body-size estimates. More track surfaces could test the proposed social pattern. For now, the footprints make a restricted reconstruction of Paranthropus harder to maintain, while leaving its precise stature, the identity of every walker and their relationships unresolved.

Selected Sources and Further Reading

1. Hatala, K. G., et al. (2026). Insights into hominin body size, locomotion, and behavior from Early Pleistocene trackways in northern Kenya. Proceedings of the National Academy of Sciences, 123(31), e2530996123.
2. Hatala, K. G., et al. (2024). Footprint evidence for locomotor diversity and shared habitats among early Pleistocene hominins. Science, 386, 1004–1010.
3. Hatala, K. G., Gatesy, S. M., and Falkingham, P. L. (2023). Arched footprints preserve the motions of fossil hominin feet. Nature Ecology & Evolution, 7, 32–41.
4. Ruff, C. B., et al. (2021). Body mass estimation from footprint size in hominins. Journal of Human Evolution, 156, 102997.
5. Hatala, K. G., et al. (2016). Footprints reveal direct evidence of group behavior and locomotion in Homo erectus. Scientific Reports, 6, 28766.
6. Domínguez-Rodrigo, M., et al. (2013). First Partial Skeleton of a 1.34-Million-Year-Old Paranthropus boisei from Bed II, Olduvai Gorge, Tanzania. PLOS ONE, 8(12), e80347.
7. Cerling, T. E., et al. (2011). Diet of Paranthropus boisei in the early Pleistocene of East Africa. Proceedings of the National Academy of Sciences, 108(23), 9337–9341.
8. Mongle, C. S., et al. (2025). New fossils reveal the hand of Paranthropus boisei. Nature, 647, 944–951.
9. Max Planck Society (2026). Fossil footprints reveal body size and group dynamics of 1.4-million-year-old human relative. Research announcement, 27 July.