Earliest evidence of elephant butchery at Olduvai Gorge (Tanzania) reveals the evolutionary impact of early human megafaunal exploitation
- Institute of Evolution in Africa (IDEA), Rice University, Archaeological and Paleontological Museum of Madrid and Foundation of the University of Alcalá, Spain
- Department of Anthropology, Rice University, United States
- Archaeological and Paleontological Museum of Madrid, Spain
- PALEVOPRIM laboratory, UMR CNRS 7262, and Université de Poitiers, France
- Department of Geodynamics, Stratigraphy and Paleontology, Complutense University, Spain
- Department of Prehistory and Archaeology, University of Valladolid, Spain
- Osteoarkeologiska forskningslaboratoriet och antikens kultur, Stockholms Universitet, Sweden
- Universidade de Vigo, MAPAS Lab, Spain
- UNIARQ – Centre for Archaeology, School of Arts and Humanities, University of Lisbon, Portugal
- Area of Prehistory, Department of History and Philosophy, University of Alcalá de Henares, Spain
- Ngorongoro Conservation Area Authority, Department of Cultural Heritage, United Republic of Tanzania
- Department of Archaeology, University of Dar es Salaam, United Republic of Tanzania
Figures
Figure 1
Location of EAK in the junction of the main and secondary branches of Olduvai Gorge, main faults and Naibor Soit quartzites (A), within the area where the Bed I sites cluster at the junction (B), and the specific EAK locus with 1 m contour lines (C).
Figure 2
Stratigraphy of EAK.
(A) Location of EAK on the south side of the Main Gorge (view to the east). (B) Stratigraphic section of the gully (Korongo) showing the correlation between the different marker tuffs of Bed I on either side. The landslide displaced the boundary between Bed I and Bed II, including the following sequence: waxy claystone, tuff 1 F, waxy claystone. Vertical displacement is 6.25 m. (C) Detail of the stratigraphy of the site, showing how the archaeological level rests directly on Tuff 1 F and was covered by the clay of the lowermost Bed II. The landslide has created vertical cracks that separate both the Tuff 1 F and the archaeological level itself. The archaeological remains have moved along with the tuff blocks. The original clay of the lowermost layer II protected the fossils from erosion after the landslide. An incipient soil formed on this surface and was subsequently buried by colluvium. At present, erosion is acting on the part closest to the stream, affecting some of the fossils.
Figure 3
Photogrammetry (left) and planimetry (right) of the EAK site in 2022 and 2023.
The site-wide distribution of materials, lithics (blue) and bones (red), topographic profiles, and the 3D terrain model are shown.
Figure 4
Relative risk (i.e. probability of occurrence) of bones and lithic artifacts at EAK, with second-order functions showing regular correlation between both types of items.
Figure 5
Correlation between lithics and bones at EAK (using the elephant bones as a covariate) with the ρ (‘rhohat’) function and a Z 1 and Z 2 Berman–Lawson–Waller.
Figure 6
Left: Intentional shaping of points in a quartzite LCT from FLK West, and on a proboscidean femur shaft.
Arrows indicate conchoidal scars probably caused by use (given their location away from the edge) and their reflection and stepped morphology by the pointed area. Notice the polishing at the point, which contrasts with the unpolished state of the remainder of the artifact. More complete cortical and medullary views of this artifact can be seen in Appendix 1—figures 6–8. Right: Comparison of the configuration strategies identified on the quartz LCT and the LAS proboscidean femur shaft. (A) LCT on a quartz slab from FLK West, Level 6; (B) The proboscidean femoral shaft from LAS in the vicinity of EAK; (C) Diacritical diagram showing point conceptualization sequencing identified in the previous specimens: 1. First thinning work (oblique to the distal end in the quartz LCT and vertical in the bone shaft); 2. Right-sided retouch; 3. Left-sided retouch. Series 2 and 3 determine the final shaping of the distal tip.
Appendix 1—figure 1
Excavation at EAK of the partial articulated rear foot of the juvenile elephant, displaying the calcaneum with unfused epiphysis (upper) and articulated phalanges (lower).
Appendix 1—figure 2
Fragmented and anatomically connected ribs from the proboscidean individual at EAK.
Appendix 1—figure 3
Distribution of elephant and hippopotamus carcasses at the main section of Olduvai Gorge, documented in the survey conducted in June 2024.
The lower (red), intermediate (yellow), and upper (blue) stratigraphic units display a clustered distribution of carcasses. Notice the spatial overlap of those megafaunal clusters with the dense landscape-scale concentrations of stone artifacts (in faded red for the LAS lower unit and faded blue for the upper unit). Elephant carcasses appear numbered and hippopotamus carcasses start with ‘h’ before the number. EGB, elephant green-broken bone. The megafaunal site of TK Sivatherium is outside the map.
Appendix 1—figure 4
Example of elephant green-broken limb bone shaft displaying percussion mark (upper) and overlapping reflected/stepped scars on its medullary surface resulting from hammerstone breaking.
Appendix 1—figure 5
EGB1: proboscidean limb bone shaft with green breaks found at BK (upper unit).
Appendix 1—figure 6
EGB 2: proboscidean femoral shaft with green breaks and a distal portion showing a combination of curved scars and a polished tip, most likely polished anthropogenically.
Appendix 1—figure 7
EGB 2: medullary view of the same proboscidean specimen displaying green breaks and step fractures on its proximal portion.
Appendix 1—figure 8
EGB 2: same specimen showing green fractures on all its sections, as seen from a medullary perspective.
All breaks show sharp edges in contrast with the polished tip.
Appendix 1—figure 9
EGB3: proboscidean limb shaft with smooth green breaks.
It was found in the LAS lower unit.
Appendix 1—figure 10
EGB4: proboscidean limb shaft fragment with green breaks showing multiple overlapping conchoidal scars and hackle marks typical of dynamic loading.
It was found in the LAS lower unit.
Appendix 1—figure 11
EGB4: detail of the hackle marks of the same specimen, typical of dynamic loading.
Appendix 1—figure 12
EGB5: Probable proboscidean flat bone found at SC (upper unit).
It has attributes of green breakage and some pitting on its cortical surface, possibly resulting from percussion.
Appendix 1—figure 13
EGB 5: same specimen seen from the other side.
Appendix 1—figure 14
EGB 6: medullary view of possible proboscidean limb shaft fragment with green break morphology, found in the intermediate unit in the vicinity of FLK west.
Appendix 1—figure 15
EGB 6: same specimen seen from its cortical side.
Notice the pitting on its surface near the notch.
Appendix 1—figure 16
EGB7: proboscidean limb shaft fragment (probably from a femur) with green break attributes, showing multiple overlapping conchoidal scars (center) found at SC (upper unit).
Appendix 1—figure 17
EGB7: other side of the same proboscidean limb shaft fragment with attributes of green-bone breakage, such as smooth fracture surface and conchoidal scars, found at SC (upper unit).
Appendix 1—figure 18
EGB7: same proboscidean limb shaft fragment, cortical view.
Note the marks from abrasive agencies, plus some pitting by both lower edges of breakage planes suggestive of the potential dynamic loading effector.
Appendix 1—figure 19
EGB7: detail of the same limb shaft fragment showing the longest smooth fracture surface typical of a green break with inflection point in the middle.
Appendix 1—figure 20
Example of other megafaunal long bone shafts showing green breaks.
Appendix 1—figure 21
Elephant-sized flat bone fragment with green-bone curvilinear break.
Appendix 1—figure 22
Same specimen with a different view.
Appendix 1—figure 23
Proboscidean limb bone fragment found at EAK, with morphology of a green-bone break.
Red lines show the outline of the overlapping scars.
Appendix 1—figure 24
Mapping of intensity, nearest-neighbor cleaning, and spatial modeling of the SPP of bones at EAK, including graphs of the inhomogeneous second-order functions.
Appendix 1—figure 25
Mapping of intensity, nearest-neighbor cleaning, and spatial modeling of the SPP of lithics at EAK, including graphs of the inhomogeneous second-order functions.
Appendix 1—figure 26
Lithics from EAK.
Bipolar cores: 1 Cubic specimen (77 x 70 × 58 mm, 455 g) from which a medium-sized flake (47 x 63 mm) has been detached; 3. A cubic blank has been exploited in the thickness plane around the perimeter of the piece with a circular tendency (51 x 47 × 32 mm, 194 g). Frosting is observed on the basal area and on some ridges; 5. Triangular morphology and opposed detachments, showing clear frosting on the base (71 x 40 × 35 mm, 118 g). Hammerstone: 2. Oval basalt cobble (78 x 70 × 48 mm, 323 g) showing battering around the perimeter of the piece. Handheld cores: 4. Exhausted multifacial-multipolar core with a cubic shape (41 x 40 × 30 mm, 75 g), showing a minimum number of 10 negative scars crossing in multiple planes.
Appendix 1—figure 27
Flakes from EAK.
Phonolite: 1. Rectangular, Type 6 specimen, unifaceted butt, one dorsal negative scar, with 92 mm of bilateral and distal cutting edge (48 x 38 × 13 mm, 29 g). Quartz: 2. Rectangular, Type 6, unifaceted platform, simple orthogonal dorsal pattern, and 50 mm lateral cutting edge (58 x 39 × 19 mm, 42 g); 3. Triangular, Type 6, unifaceted platform, simple orthogonal dorsal pattern, and 30 mm of latero-distal edge (recent notch on the left side; 43 x 28 × 9 mm, 13 g); 4. Oval, Type 6, lineal platform, opposed dorsal pattern, and 40 mm of transversal cutting edge (35 x 50 × 12 mm, 19 g); 5. Quadrangular, Type 6, unifaceted butt, orthogonal pattern, diffuse bulb on ventral face and 28 mm of distal cutting edge (48 x 37 × 14 mm, 24 g); 6. Oval, Type 6, unifaceted butt, lateral opposed dorsal pattern, and 20 mm of distal cutting edge (36 x 32 × 13 mm, 13 g).
Appendix 1—figure 28
Stereograms, rose diagrams, and results obtained from the application of the Rayleigh, Kuiper, and Watson tests.
Appendix 1—figure 29
Cluster of hyena-ravaged bones from a giraffe at Chobe (Botswana).
Appendix 1—figure 30
Spiral and green-broken planes on both sides of a giraffe humerus (broken by hyenas) from the cluster displayed in Appendix 1—figure 22.
Appendix 1—figure 31
Giraffe scapula ravaged by hyenas displaying overlapping non-invasive notches (blue arrows), continuous green breaks (yellow arrows), and axial reflected scars parallel to the axis of the bone (also documented in hammerstone broken megafaunal limb bones; see Appendix 1—figure 4).
Appendix 1—figure 32
Example of Bos/Bison femoral shaft from the hyena den of Bois Roche (from Villa and Bartram, 1996, Flaked Bone from a Hyena Den, Paleo 8: 143–159).
Notice the amount of scars, their overlapping and incomplete nature, and their continuous trajectory on what seems a pointed bone fragment.
Appendix 1—figure 33
Example of green breakage on a proboscidean femur (Gomphoterium) from the Middle Miocene site of Virgen del Puerto (Madrid, Spain).
Appendix 1—figure 34
Example of an elephant green-broken femoral shaft fragment from BK (Olduvai, Upper Bed II) presented by Leakey as a case of bone tool.
Notice the similar scar overlap and smooth green fracture surfaces resulting from bone breaking. The specimen's length is 46 cm. Currently displayed at the Olduvai Museum.
Appendix 1—figure 35
Tibia fragment from a large medium-sized bovid displaying multiple overlapping scars on both breakage planes inflicted by carnivore damage (red arrows).
Within those scars, there are abundant reflected/incomplete smaller scars (yellow arrows). The specimen was found within the base of the LAS stratigraphic unit, where the most abundant number of megafaunal green-broken bones have been found.
Appendix 1—figure 36
Same specimen as shown in Appendix 1—figure 30, viewed on its medial side.
The profile of the overlapping scars and notches can be clearly seen (red arrows), together with traces of tooth marks (yellow arrows).
Tables
Table 1
African Early Pleistocene archaeological sites with proboscidean carcasses containing potential evidence of hominin involvement.
| Area excavated (m2) | N° bones | N° lithics | Megafaunal taxon | Other non megafauna | Age (Ma) | Green breaks | Cut marks | |
|---|---|---|---|---|---|---|---|---|
| Nyayanga 3 (Kenya) | 49 | 1580*(241 Hippo) | 42† | Hippo | Present | 3–2.6 | Unreported | Ambiguous |
| Nyayanga 5 (Kenya) | 24 | 196* | 14† | Hippo | Present | 3–2.6 | Unreported | Ambiguous |
| FLK North 6 (Tanzania) | 37 | 400 | 123 ‡ | Elephant | Present | 1.8 | Absent | Ambiguous |
| FLK North Deinotherium site (Tanzania) | - | 23 | Elephant | Present | 1.7 | Absent | Absent | |
| Barogali (Djibouti) | 37 | - | 569 | Elephant | Absent | 1.6–1.3 | Unreported | Absent |
| Nadung´a (Kenya) | 53 | 65 | 6797 | Elephant | Present | 1–0.7 | Unreported | Absent |
-
*
Numbers in main text and appendix tables do not coincide.
-
†
Only artifacts reported as associated to the skeletal remains (total in both excavations is 135).
-
‡
Excluding purported manuports.
Appendix 1—table 1
Skeletal representation of the elephant carcass at EAK.
| NISP | MNE | |
|---|---|---|
| Cranium | 12 | 1 |
| Mandible | ||
| Teeth | 44 | 5 |
| Vertebrae | 1 | 1 |
| Ribs | 51 | 8 |
| Pelvis | 11 | 1 |
| Scapula | ||
| Humerus | ||
| Radius | ||
| Ulna | 1 | 1 |
| Carpals | ||
| Metacarpals | ||
| Femur | 5 | 2 |
| Tibia | 2 | 2 |
| Patella | 1 | 1 |
| Calcaneum | 2 | 1 |
| Astragalus | 1 | 1 |
| Tarsals | 5 | 5 |
| Metatarsals | 5 | 5 |
| Phalanges | 12 | 12 |
Appendix 1—table 2
Distribution of the lithic specimens retrieved from EAK sorted by lithic category and raw material.
V, vesicular; B, basalt; P, phonolite; Q, quartz.
| Lithic category | Raw material | Total | ||||
|---|---|---|---|---|---|---|
| V | B | P | Q | n | % | |
| Unmodified | 1 | 1.25 | ||||
| Cobbles | 1 | |||||
| Percussion | 1 | 1.25 | ||||
| Hemmerstones | 1 | |||||
| Cores | 4 | 5 | ||||
| Handheld | 1 | |||||
| Bipolar | 3 | |||||
| Detached | 23 | 28.75 | ||||
| Flakes | 1 | 14 | ||||
| Broken flakes | 4 | |||||
| Bipolar flakes | 4 | |||||
| Waste | 51 | 63.75 | ||||
| Flake fragments | 14 | |||||
| Debris | 24 | |||||
| Shatter | 10 | |||||
| Fragments | 3 | |||||
| Total n | 1 | 1 | 1 | 77 | ||
| Total % | 1.25 | 1.25 | 1.25 | 96.25 | ||
Appendix 1—table 3
Size (L=length, B=breadth; T=thickness) and mass (M) of complete flakes retrieved from EAK.
| Min | Max | Mean | SD | |
|---|---|---|---|---|
| L | 20 | 58 | 32.76 | 10.66 |
| B | 19 | 50 | 29.41 | 7.97 |
| T | 6 | 19 | 11.17 | 3.66 |
| M | 4 | 42 | 13.35 | 10.56 |
Additional files
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Earliest evidence of elephant butchery at Olduvai Gorge (Tanzania) reveals the evolutionary impact of early human megafaunal exploitation
eLife 14:RP108298.
https://doi.org/10.7554/eLife.108298.5
