Hominin remains from the Late Pliocene of Ledi-Geraru, Ethiopia.

The genera Homo and Paranthropus are common in the fossil record from about 2.0 million years ago. Both are thought to have derived from an earlier Australopithecus ancestor, with the most likely ancestor for the genus Homo generally thought to be Australopithecus afarensis. However, Australopithecus afarensis is not known after 2.95 million years ago, with Hominid fossils being rare over the intervening interval, spanning the latest Pliocene and earliest Pleistocene. Examples of Paranthropus have recently been described from 2.7 million-year-old deposits in the Omo-Turkana Basin of Ethiopia and Nyayanga in Kenya, and the 2.66 million-year-old deposits at Laetoli in Tanzania, while a jawbone attributed to the genus Homo has been found at Ledi-Geraru in Ethiopia which has been dated to 2.78 million years ago, pushing the presence of both these 'Pleistocene' genera back into the latest Pliocene, while a new species of Australopithecus, Australopithecus garhi, has recently been described from 2.5 million-year-old (earliest Pleistocene) deposits in the Afar region of Ethiopia.

In a paper published in the journal Nature on 13 August 2025, a team of scientists led by Brian Villmoare of the Department of Anthropology at the University of Nevada Las Vegas describe a series of recent Hominin finds made by the Ledi-Geraru Research Project in the Afar Basin of Ethiopia.

The Ledi-Geraru sites are located at the northern end of the palaeoanthropological sites of the Afar Basin, and has produced the only known evidence of the genus Australopithecus surviving after 2.95 million years ago, as well as the earliest evidence for the appearance of the genus Homo. Paranthropus has not been found in this area, but it is unclear whether this represents a genuine absence. The two sites of Ledi-Geraru, Lee Adoyta and Asboli are to the west of the Awash River, in an area cut through by the Mille and Geraru rivers and their tributaries. The deposits here are between 2.5 and 3.0 million years old, and have been dated by Argon-Argon radioisotope stratigraphy of volcanic layers, as well as magnitostratigraphy.

Map of the Ledi-Geraru Research Project area. (a) The  Ledi-Geraru Research Project area (yellow star) is located towards the northern  extent of palaeontological sites (red circles) in the Afar depression, Afar Region,  Ethiopia. (b) Within the Ledi-Geraru project area, the Lee Adoyta and Asboli fossil  sites are located approximately 12.5 km apart. Villmoare et al. (2025).

Argon-Argon dating relies on determining the ratio of radioactive Argon⁴⁰ to non-radioactive Argon³⁹ within minerals from igneous or metamorphic rock (in this case volcanic ash) to determine how long ago the mineral cooled sufficiently to crystallise. The ratio of Argon⁴⁰ to Argon³⁹ is constant in the atmosphere, and this ratio will be preserved in a mineral at the time of crystallisation. No further Argon³⁹ will enter the mineral from this point, but Argon⁴⁰ is produced by the decay of radioactive Potassium⁴⁰, and increases in the mineral at a steady rate, providing a clock which can be used to date the mineral.

Magnitostratigraphy uses traces of ancient magnetic fields preserved in iron minerals in rocks to trace ancient pole reversals; the poles only have two possible orientations (north pole in the north/south pole in the south or south pole in the north/north pole in the south) and these occasionally flip, with the poles exchanging positions. Pole reversals happen more-or-less at random, with periods between reversals occurring at intervals ranching from tens of thousands to millions of years, and reflected across the globe. This creates a pattern of magnetic reversals in sedimentary rocks that can be matched in different rocks across the globe.

The first specimen described by Villmoare et al. comes from the Gurumaha Sedimentary Package, which outcrops in narrow fault-bounded exposures in the central Lee Adoyta basin and in drag-faulted blocks adjacent to basalt ridges  bounding the basin to the east. This sedimentary package is cut through by the Gurumaha Tuff, which has been dated to 2.782 million years before the present. This is the unit which previously produced specimen LD 350, a 2.78 million-year-old mandible which is the oldest fossil assigned to the genus Homo. 

The specimen derived from this unit, LD 302-23, is a third right lower premolar found 22 m to the southwest and 7 m bellow specimen LD 350, but still above the Gurumaha Tuff layer. This tooth measures 11.5 mm in length and 10.5 mm in width, and has a fragment of enamel missing from its lingual corner, being otherwise well-preserved. The shape of the premolar is consistent with that seen in some examples from Australopithecus afarensis, but the pattern of cusps is quite different to anything seen in any member of the genus, making it unlikely that this tooth came from an Australopithecus. It also falls within the size range of both species of Paranthropus, but is quite different in shape. Third premolars from early members of the genus Homo are quite variable, but clearly differ from both those of Australopithecus and Paranthropus. Since this tooth falls within the size and shape variation found in these early Homo specimens, Villmoare et al. assign it to the genus Homo. 

New Hominin dentition from the Ledi-Geraru Research project. Right, from top: LD 302-23 P₃,  LD 750 P₄, AS 100 M¹ and AS 100 M².  Left, images show the LD 760 assemblage  (top, from left: maxillary molar, I², I¹, maxillary canine; bottom, mandibular molars). Villmoare et al. (2025).

The second specimen described, LD 750-115670, is an isolated lower fourth premolar, found at the base of an 8 m exposure of fossiliferous mudstones and sandstones at site LD 750. This site is located stratigraphically between the 2.63 million-year-old Lee Adoyta Tuffs and the 2.59 million-year-old Giddi Sands Tuff. 

The tooth crown is unworn, with all cusps preserved, although the root is broken off, giving a maximum root height of about 2 mm. The lack of wear may imply that the tooth was unerupted at the time of death. The tooth is 12.4 mm long and 11.4 mm wide, placing it at the upper end of the size range for Australopithecus afarensis or Australopithecus africanus, and too large for Australopithecus anamensis. No lower jaw or teeth are known for Australopithecus garhi, but the specimen is within a plausible size range for the species. It also falls within the size range of both Paranthropus species, but again is quite different in shape. It does resemble several fourth premolars attributed to early Homo, though Villmoare et al. note that these attributions are provisional, and that it is difficult to distinguish between early Homo and Australopithecus fourth premolars. Since this tooth lacks any distinctively Homo features, Villmoare et al. provisionally assign it to aff. Australopithecus sp..

A set of five associated lower molars were discovered at a site identified as LD 760, a flat sandy area approximately below the 2.63 million-year-old Lee Adoyta Tuffs. These are worn, with dentine exposed on their outer cusps, and wide for their lengths, giving them a squarish profile. Notably, the third molar is larger than the second molar, and the second molar is larger than the first, the first and second molars a relatively square, and the first and second molars lack a seventh cusp, all traits compatible with Australopithecus afarensis, but not early Homo. However, these teeth also differ from those of Australopithecus afarensis in several ways; they do not taper towards the rear, and lack a distinctive bilobate buccal contour. 

LD 760 molars compared to Australopithecus afarensis. Left molars from Ledi-Geraru specimen LD 760 (left) and Hadar specimen A.L. 400-1 (right). Measurements in mm of the LD 760 molars (length × width): LM1: 13.3 × 13.4, LM2: 14.5 × 14.6, LM3: 14.0 (estimated) × 15.7, RM1: 13.2 × 13.1, RM2: 14.8 × 15.2. Specimens are oriented with their buccal surfaces to the left and mesial surfaces up. Villmoare et al. (2025).

A partial upper molar was also recovered from this locality. This preserves the lingual grove of the tooth, which appears to be quite distinct. In the upper molars of Australopithecus garhi this groove is indistinct. Furthermore, the upper molars of Australopithecus garhi has a greatly reduced hypercone cusp, which leads the protocone cusp to take on a triangular shape. The hypercone is absent in the LD 760 specimen, but the protocone is present, and shows no sign of modification due to a reduced hypercone. However, the sample size for Australopithecus gahri is small, so this cannot be ruled out as a natural variation within the species.

Also found at the LD760 site were a right maxillary (upper) canine, a complete left maxillary lateral incisor and a left fragmentary maxillary central incisor. Thecanine (LD 760-115979) is well  preserved, lacking only the tip of the root. This has mesial and distal interproximal contact facets (wear marks caused by contact between teeth), with the interproximal contact facet having a matching distal interproximal contact facet on the second incisor. This means that the second incisor and canine were in contact, something typical of the the genus Homo, and unlike the situation in many Australopithecus specimens (including the Australopithecus gahri maxilla BOU-VP-12/130) where these two teeth are separated by a diastema (this trait is variable in Australopithecus afarensis, which may-or-may not have a diastema, so this could conceivably also be the case in Australopithecus gahri). The canine is also notably large, towards the upper end of the size range seen in Australopithecus (and much larger than anything seen in Paranthropus).

Comparative maxillary canine morphology. (a) Lingual (left) and  labial (right) views of the Ledi-Geraru LD 760-115979 canine (left) with Hadar Australopithecus afarensis specimens A.L. 763-1 (middle) and A.L. 333x-3 (right). Note that the  LD 760 canine is a right canine, whereas the Australopithecus afarensis canines are from the  left and are mirrored in these images. (b)–(d) LD 760-115979 ((b) shown in lingual view) contrasted with Hadar Australopithecus afarensis specimen A.L. 199-1 ((c) right canine  shown; distal to the upper right) and Bouri Australopithecus garhi specimen BOU-VP-12/130  (d) left canine, mirrored; distal is to the right). Note the simple mesial–distal  chisel-like wear pattern on the LD 760 canine (b) in contrast to the complex  multi-faceted wear pattern of Australopithecus afarensis (c) and the broad curved basin on the  distal side of the Australopithecus garhi upper canine ((d) this is seen on both left and right  canines). Images are oriented differently to emphasise the distinctive relevant morphology. Images in (b)–(d) are not to scale. Vallmoare et al. (2025).

The canine of Australopithecus gahri has a unique structure, with a shallow distal basin reminiscent of a talon which is contiguous with a wide wear furrow which runs along the entire post-canine dental row, something not seen in any other Hominin, and absent in the LD 760 canine. Morphologically, this tooth resembles those of Australopithecus afarensis, however, it has different wear patterns. In Australopithecus afarensis wear is mostly seen on the distal crest, whereas in the the LD 760 canine it is predominantly on the apex, suggesting a difference in diet and/or lifestyle.

The LD 760 individual clearly does not belong to Paranthropus, and is not morphologically consistent with any described species of Australopithecus. However, since it resembles Australopithecus afarensis more closely than anything else, Vallmoare et al. refer it to Australopithecus sp. indet.

The final specimens discussed come from the Giddi Sands unit in the Asboli region. These were found immediately below the 2.593 million-year-old Giddi Sands Tuff, and comprise a partial upper left first molar, and two fragments of an upper left second molar, which can be assembled to form a whole crown. They show little wear, and are close in shape to those of Australopithecus afarensis, although they lack the pronounced lingual occlusocervical sloping and general 'puffy' appearance of the molars of that species. They closely resemble the molars of early Homo specimens such as the 2.3 million-year-old specimen from the Busidima Formation at Hadar, or the 2.5-2.4 million-year-old specimen from Mille-Logya. It is quite different in form from the molars of Australopithecus gahri, and is small compared to the molars of either Australopithecus gahri or Paranthropus. 

Vallmoare et al. believe that although the Ledi-Geraru material is very limited, it provides clear evidence that both Australopithecus and Homo were present during the 3.0-2.5 million years ago interval, suggesting that multiple non-robust Hominin lineages were present in East Africa before 2.5 million years ago. 

The molar fragments from Asboli sufficiently resemble the Homo specimens from Hadar and Mille-Logya that Vallmoare et al. are confident that they represent the same species. However, they predate the older of these specimens (Mille-Logya) by at least 150 000 years. This adds to the evidence for an early appearance of Homo as Ledi-Geraru previously established by the LD 350-1 mandible, as does the LD 302-23 premolar which Vallmoare et al. describe from the Gurumaha sedimentary package.

The LD 750 and LD 760 material both come from the Lee Adoyta sedimentary package, although they are separated by 24 m of strata and the 2.63 million-year-old Lee Adoyta Tuffs. Nevertheless, both appear to represent a single species of Australopithecus (an assumption based in part upon the unlikelyhood of two similar species of Austrlopithecus coexisting in the same area). 

Vallmoare et al. consider four potential explanations for this material. Firstly, they might represent a late surviving population of Australopithecus afarensis (approximately 350 000 younger than the current youngest member of the species, from the Kada Hadar 2 Submember at Hadar). Secondly, they may represent an unknown species of Australopithecus ancestral to Paranthropus; the presence of Homo in this region implies that the Homo and Paranthropus lineages had diversified by this time, but Paranthropus itself appears to be absent. However, the oldest currently accepted member of the genus coming from about 2.7 million years ago from Nyayanga in Kenya and 2.66 million years ago from the Upper Ndolanya Beds at Laetoli in Tanzania, which makes this scenario less likely. Thirdly, they may represent earlier examples of Australopithecus gahri, which have not yet developed the distinctive features of the 2.5 million-year-old specimen BOU-VP-12/130, although the lack of similarities makes this unlikely. Finally, these specimens represent a new, as yet undescribed, species of Australopithecus.

Villmoare et al. conclude that were at least three species of Hominins present in the Afar Region between 3.0 and 2.5 million years ago, Australopithecus gahri, an unknown species of Homo, and an unknown species of Australopithecus. At the same time, Australopithecus africanus was still present in South Africa, and Paranthropus had already appeared in Kenya, Tanzania, and southern Ethiopia. The Ledi-Geraru environment was drier and more open than was typical for Australopithecus, very much the sort of environment associated with appearance and proliferation of the genus Homo, suggesting that, at least locally, Australopithecus may have been able to adjust to these more open environments, at least for a while.

See also...

Using a multivariate analysis of the teeth of Sahelanthropus tchadensis to assess its status as a Hominin.

During the twentieth century palaeoanthropologists searching for the origin of the Hominins became convinced that the group has an East African origin. However, in 2001 scientists from the working on Upper Miocene deposits in the Toros Menalla region of Chad uncovered a group of possible Hominin fossils which they assigned to a new species, Sahelanthropus tchadensis, and which included a nearly complete, is distorted, cranium as well as a mandible and some isolated teeth.

Cranium TM 266-01-060-1, nicknamed Toumai, the holotype of Sahelanthropus tchadensis. Smithsonian Human Origins Program.

If correctly interpreted as a Hominin, then Sahelanthropus tchadensis would be the oldest known member of the group. As such the discovery has been subjected to a considerable amount of scrutiny, particularly as the original specimens had been subject to considerable taphonomic (post-mortem) alteration. One of the problems with the original material was that it was had to interpret the position of the foramen magnum (the whole in the skull through which the brain is connected to the central nervous system) because of distortion of the skull. This is important, because in (upright) Hominins it is located roughly in the centre of the bottom of the skull, while in Apes it is typically towards the back of the skull. A virtual reconstruction of the cranium of Sahelanthropus tchadensis has suggested that the foramen magnum would have been on the base. 

Sahelanthropus tchadensis also has a shorter, more vertical face, reduced canines and a related lack of a honing complex (a gap between the lower canine and the first premolar, into which the upper canine fits, allowing the two canines to rub together and sharpen, or 'hone' one-another), and a downward facing lip on the nuchal crest at the rear of the skull, all of which are Hominin traits. However, it also has a small, Ape-like neurocranium, as well as having a size and number of tooth-roots consistent with an Ape, presenting a mosaic of features consistant with an very early Hominin, or possibly an Ape closely related to the earliest Hominins.

Different analyses of a femur and two ulnae found at the same location as the cranium have suggested both that Sahelanthropus tchadensis was either habitually bipedal, or not habitually bipedal, and that it probably spent at least some of its time in the trees. Since a cladistic analysis has recovered Sahelanthropus tchadensis as a Hominin, this raises at least the possibility that Hominins were habitually walking on two legs 7 million years ago.

One aspect of Sahelanthropus tchadensis which has not been studied extensively is its dentition, despite a number of teeth and tooth fragments being available. In a paper published in the South African Journal of Science on 31 July 2024, Walter Neves, Leticia Valota, and Clovis Monteiro of the Institute of Advanced Studies at the University of São Paulo, present the results of a morphometric analysis which compared the upper posterior dentition of Sahelanthropus tchadensis to that of living Apes and fossil Plio-Pleistocene Hominins.

The upper posterior teeth were chosen as these were the teeth for which the mesiodistal and buccolingual diameters could be determined in Sahelanthropus tchadensis. These were compared to a selection of teeth from Pan troglodytes (Chimpanzees), as well as the Hominins Orrorin tugenensis, Ardipithecus ramidus, Australopithecus afarensis, Australopithecus africanus, Paranthropus boisei, Paranthropus robustus, Homo habilis, and Homo erectus.

For each tooth the ratio between the mesiodistal and buccolingual diameters (length and width) was calculated, and these were plotted on a distribution map. The teeth fell into approximately three groups on this map, one comprising modern Chimpanzees, one comprising the 'Robust Autralopithecines' Paranthropus boisei and Paranthropus robustus, and one containing all other Hominins.

Distribution of the species and specimens included in the study along the morphospace defined by principal component 1 (PC1) x principal component 2 (PC2). Neves et al. (2024).

The teeth of Sahelanthropus tchadensis fell within the non-Robust Hominin cluster, plotting closest to Ardipithecus ramidus, a Miocene-Pliocene species with a large number of known specimens, which again has some Ape-like characteristics, but which is generally accepted as a Hominin by palaeoanthropologists today. Neves et al. interpret this as supporting the hypothesis that Sahelanthropus tchadensis is a Hominin rather than an Ape.

See also...

Evidence for shoes in the Middle Stone Age of the Cape Coast, South Africa.

The appearance of clothing is considered to be one of the key stages in the development of Modern Humans, and the development of footwear, a complex form of clothing, is in itself an important technological leap. However, the exact time when footwear first appeared is unknown, with the first foot-coverings presumed to have been made of perishable materials unlikely to have survived. In the absence of preserved footwear, ichnology (the study of tracks and traces) provides the most plausible technique for detecting the first use of footwear among Hominins.

The oldest known shoes in the archaeological record are sandals made of woven Sagebrush bark from Oregon in the United States, the oldest of which have been dated to between 10 500 and 9200 years old. Also from the United States, a variety of sandals, moccasins, and slip-on footwear have been uncovered in Missouri, dating from between 8000 and 1000 years ago. The oldest shoe from Eurasia is a leather wrap-around shoe from southeastern Armenia, dated to between 5600 and 5400 years ago. Also from Eurasia, ‘Ötzi the Iceman' a frozen natural mummy found in the High Alps on the border between Austria and Italy, dated to more than 5000 years old, had a complete set of clothing, including complex footwear made from Bear-skin, Deer hide, and tree bark. A pair of Cow-hide sandals from Israel has also been dated to more than 5000 years ago.

Hominin trackways are known from many sites around the world, and can be identified by features such as the alignment and shape of the hallux, relative digit lengths, and the presence of a prominent medial longitudinal arch. However, these identifying features are based upon the assumption that the track-maker was unshod, and it is unlikely that all would be present were they to have been wearing some form of foot-covering. Thus, unsurprisingly, the majority of known Hominin tracks are considered to have been made by unshod tracemakers, although a few exceptions are known.

The oldest apparently shod Human tracks in the Americas are in Jaguar Cave, Tennessee, which are thought to be about 5000 years old. In Europe such traces include traces associated with a Magdalenian (Upper Palaeolithic) tool assemblage in Fontanet Cave, France, which appear to have been made by an individual wearing a soft shoe or sock. Also in France, Cussac Cave, footprints associated with a Gravettian assemblage and dated to between 38 000 and 31 000 years ago lack any digit impressions, despite these being present on both Human handprints and Bear footprints within the same cave, leading to the possibility that the trace-makers may have been shod.

Footprints without digit impressions are also known from a third site in France, Le Rozel, although these are about 80 000 years old, implying that the trace-makers were Neanderthals rather than Modern Humans. Possible evidence for footwear use by Neanderthals comes from Theopetra Cave in Greece, where a mixture of shod and unshod traces appears to have been left by a group of children around 130 000 years ago.

In South Africa, a number of possible footwear traces have been found on the Cape Coast, an area noted for its excellent Pleistocene tracks, including one trace which appears to show a clear sandal-imprint.

The use of clothing is thought to have been developed by Hominins inhabiting cool environments during Pleistocene cold spells, although the perishable nature of the material from which such cloths are likely to have been made, makes it unlikely that any preserved material will ever be found. Instead, tools used for the processing of skins, leather, and textiles are used as proxies for the materials themselves. This begins with the appearance of scrapers presumed to have been used for the preparation of skins, shortly followed by bone awls, needles and eyed needles, which imply the manufacture of increasingly complex clothing. Based upon this, it has been calculated that the first simple clothing probably appeared around 800 000 years ago, long before the emergence of Modern Humans, with several different Hominin species probably manufacturing and wearing cloths, with the possibility that footwear appeared at the same time. Molecular clock analysis suggests that Body Lice probably diverged from Head Lice between 170 000 and 80 000 years ago, by which time is is assumed that Humans were habitually wearing cloths all the time.

The Contrebandiers Cave site in Morocco, which has been dated to between 120 000 and 90 000 years ago, has yielded a variety of bone tools thought to have been used for the processing of leather and furs, but for the most part Pleistocene sites with evidence for tools likely to have been used in clothing manufacture and tracks made by Hominins are restricted to two areas, Western Europe and the Cape Coast of South Africa.

Such tools in Europe are known from the Middle Palaeolithic of southwestern France, with the oldest dating to about 51 400 years ago, and other examples between 48 000 and 41 000 years ago. These tools were presumably made by Neanderthals, and include lissoirs, tools which can be used to process hides, giving a smooth, tough, and reasonably impermeable finish (although not the same as modern commercially produced leathers).

The oldest sites in South Africa yielding tools which could have been made for making cloths are at Klasies River and Blombos Cave on the Cape Coast, with the Klasies River site dated to about 100 000 years ago and Blombos Cave mostly to about 80 000 years ago, although one awl, made from the bone of a Bird, was dated to at least 125 000 years ago.

Although these tools are taken as the earliest tools which are clearly associated with clothing manufacture, in both Europe and South Africa earlier stone tools could have been used to cut or pierce hides during the manufacture of simple garments, as could sharpened shell tools known from the Cape Coast. However, the fact that the oldest surviving footwear was made from woven Sagebrush rather than leather does suggest that caution should be applied when judging what ancient populations would have seen as suitable materials for making footwear.

It has also been suggested that wearing shoes might affect the way in which the foot develops, something which has been implied for a set of Human remains from Tianyuan in China dated to about 50 000 years ago, although the phalanges of the foot are seldom well preserved, limiting the number of instances in which this method can be applied.

Another Upper Palaeolithic site at Sunghir, in northern Russia, yielded a buried individual with remnant body decoration implying leggings or boots, as well as having extremely gracile lateral phalanges, something thought to be associated with habitual shoe wearing. This was particularly surprising as all Upper Palaeolithic Hominin tracks from Eurasia appear to have been made by barefoot trace-makers.

The relationship between footwear and foot morphology is complicated. Individuals who minimise their use of footwear are thought to have stronger foot muscles, and fewer pathologies of the feet, but this would be unlikely to result in any detectable skeletal difference between someone who never wore shoes and someone who occasionally did so, or even somebody who habitually wore very lightweight shoes.

In a paper published in the journal Ichnos on 28 August 2023, Charles Helm of the African Centre for Coastal Palaeoscience at Nelson Mandela University, Martin Lockley, also of the African Centre for Coastal Palaeoscience at Nelson Mandela University, and of the Dinosaur Trackers Research Group at the University of Colorado Denver, Hayley Cawthra, again of the African Centre for Coastal Palaeoscience at Nelson Mandela University, and of the Minerals and Energy Unit at the South African Council for Geoscience, Jan De Vynck, again of the African Centre for Coastal Palaeoscience at Nelson Mandela University, and of the Evolutionary Studies Institute at the University of the Witwatersrand, Mark Dixon, again of the African Centre for Coastal Palaeoscience at Nelson Mandela University, Renée Rust, again of the African Centre for Coastal Palaeoscience at Nelson Mandela University, and of the School of Geography, Archaeology and Environmental Studies at the University of the Witwatersrand, Willo Stear and Monique Van Tonder, once again of the African Centre for Coastal Palaeoscience at Nelson Mandela University, and Bernhard Zipfel, also of the Evolutionary Studies Institute at the University of the Witwatersrand, describe three Middle Stone Age sites on South Africa’s Cape Coast where tracks appear to have been made by Hominins wearing shoes, and discuss ways in which future studies of this topic could be approached.

Map of the Cape Coast of South Africa, showing Cenozoic deposits and places mentioned in the text. Helm et al. (2023).

Scientists from the African Centre for Coastal Palaeoscience have been studying the trace fossils of the Cape Coast since 2007. Here, a 350 km section of coast has frequent coastal aeolianites (wind-blown sand deposits) which have been buried by subsequent similar deposits than set with a carbonate cement, preserving a record of the people and animals which moved over them during the Middle Stone Age. As well as trackways left by Hominins, these studies revealed imprints left by Crocodiles, Giraffe, breeding Sea Turtles and very large Tortoises, none of which have left skeletal remains in the area.

The Cape Coast region has a remarkably rich archaeological record, tracing the appearance and development of numerous stages considered key steps on the way to Modern Human behaviour, including personal adornments, jewellery, the heat treatment of stone tools, art, and the use of abstract symbols. Palaeoclimatic studies of the same region indicate that the coastal plains upon which these deposits were laid down went through a cycle of exposure and inundation throughout the Pleistocene, which probably helped to maintain an ideal Hominin environment in the region. 

Prior to this project, Hominin tracks had previously been discovered at Langebaan on the west coast of South Africa, and Nahoon on the east coast of the country. Extensive study of the Cape Coast has produced four new track localities, and while another probable set of tracks has recently been identified at Langebaan. Combined with a set of tracks on an ancient lakebed in the Nefud Desert of Saudi Arabia, this represents the entire global inventory of tracks more than 46 000 years old which are attributed to Modern Humans.

All of the Hominin tracks on the Cape Coast are within the aeolianites of the Waenhuiskrans Formation, which together with the cemented foreshore, shoreface and lagoon deposits of the Klein Brak Formation, comprise the Bredasdorp Group. These deposits have been dated to between 400 000 and 36 000 years old, although the majority belong to Marine Isotope Stage 5, making them between 130 000 and 80 000 years old. To the east of Robberg, the correlate of the Waenhuiskrans Formation is the Nahoon Formation, which forms part of the Algoa Group. These Pleistocene aeolianites split readily along their bedding planes, so that tracks are often exposed on fallen blocks on the coast. Thes blocks are ephemeral, quickly being eroded away by the action of the waves.

Due to this ephemeral nature, tracks needed to be recorded quickly when exposed. This was done by recording their locality with a GPS unit and taking numerous photographs, which were then used to build photogrammetric models. Exposed trackways and footprints were also measured for track length, track width, track depth, pace length, stride length, and thickness of foresets, and individual footprints were examined for evidence of strap attachment points. 

In addition, studies were made of newly created tracks on sandy surfaces on the Cape Coast, with varying levels of moisture, slope, and firmness, and a variety of shoe types, including shoes with an open soft sole, an open hard sole, a closed soft sole and a closed hard sole. The best match for the fossil footprints was made by using an open hard soled shoe on soft, moist sand.

Shoe designs for this study were based upon shoes made by examples of sandals made by San peoples of southern Africa from the collections of the Blombos Museum of Archaeology in South Africa and Zambezi Heritage Museum in Namibia. The shoe which gave the closest results to the preserves footprints was based upon two such sandals, and was made from two layers of Cow-hide glued together, and laced by piercing three holes through this sole, one between the big toe and the adjacent digit, and two about half way along the length of the shoe, one at each side. The laces were made from the same Cow-hide as the upper part of the shoe, and were threaded through the holes and knotted underneath, protruding below, although in the museum examples prolonged use had levelled out the knots with the base of the sole. These sandals proved easy to attach to the foot.

Helm et al. present details of three footprint-bearing sites on the Cape Coast, these being, from west to east, Kleinkrantz, Goukamma, and Woody Cape.

At Kleinkrantz a slab measing 55 cm x 55 cm and 20 cm thick with apparent footprints on its upper surface, was found lying on a modern coast dune beneath a vegetated slope, which in turn lies beneath a cliff with an exposure of the Waenhuiskrans Formation. There was no sign of a recent rockfall, and samples taken from the Waenhuiskrans Formation in this area have been dated to between 148 000 and 79 000 years old.

(A) The Kleinkrantz site in the Garden Route National Park; scale bar is 10 cm. (B) Photogrammetry colour mesh of the Kleinkrantz site, using 48 images. Photos were taken average 36.4 cm from the surface. The reprojection error is 0.39 pix. Vertical and horizontal scales are in metres. Arrows indicate possible strap attachment point impressions. Helm et al. (2023).

This slab has two similar depressions, one in front of the other, each with crisp three crisp and well-defined margins, two straight parallel, the third near semicircular. One of these appears to have been a double impression, with a slightly narrower impression overlying an slightly wider one. The wider of these is 9 cm wide, the narrower 7 cm, while the single impression is 8 cm wide. Both features are at least 16 cm long. The narrowest impression also has three depressions within it consistent with the position of the strap knots on the experimental sandal.

Two other possible track prints are also present, although these are more amorphous, and were apparently made by a person travelling in a different direction, though again possible strap-attachment knot impressions are present. 

The second site lies within the Goukamma Nature Reserve, between Sedgefield and Knysna, where the Waenhuiskrans Formation has been dated to between 136 000 and 79 000 years old, although recent dates obtained suggest parts of the formation may be as young as 73 000 years old. Here a fallen slab has three footprints with crisp margins and no signs of digits, one of which particularly resembles the imprint of a shod foot. This print measures 11.5 cm in length and 6 cm in width.

(A) The Goukamma surface; scale bar = 10 cm. (B) Angled view of the Goukamma site; scale bar is 10 cm. (C) Photogrammetry colour mesh of the Goukamma site, using 59 images. Photos were taken average 29.9 cm from the surface. The reprojection error is 0.39 pix. Vertical and horizontal scales are in metres. Helm et al. (2023).

The third site, Woody Cape, is situated in the coastal portion of the Addo Elephant National Park. Here a detached slab from the Nahoon Formation shows a trackway comprising four footprints, one of which is partial. The prints are 10-12 cm long and 5-6 cm wide, with a pace length of 19 cm. Raised areas to the left of each print imply some downslope movement.

Photogrammetry colour mesh of the Woody Cape site, using 37 images. Photos were taken average 29.2 cm from the surface. The reprojection error is 0.57 pix. Vertical and horizontal scales are in metres. Helm et al. (2023).

Helm et al. also attempted to recreate a trackway similar to the Pleistocene examples, using a hard-soled sandal based upon museum specimens, on a modern, moist, soft, level dune surface. This produced footprints with slight out-toeing about 26 cm in length, with a width of 11.5 cm and a pace length of 63 cm. The front and back margins of the prints were semicircular, the sides parallel. Strap attachment marks can be seen.

Photogrammetry colour mesh of the neoichnological trackway (level surface, hard-soled sandal, soft substrate) using 63 images. Photos were taken average 37 cm from the surface. The reprojection error is 2.07 pix. Vertical and horizontal scales are in metres. Helm et al. (2023).

Helm et al. accept that the evidence they provide cannot be seen as irrefutable, however, the presence of similar tracks at three separate sites, combined with the recreation of similar prints with a pair of modern sandals based upon a historic design, does strongly suggest that similar sandals were being worn by Humans in the area more than 70 000 years ago. All three sets of footprints appear to have been made by trackmakers smaller than modern adult Humans, suggesting that they were either adults of smaller stature, or children.

A pair of sandals on exhibit in the Blombos Museum of Archaeology in Still Bay, viewed from above (A), and below (B), showing strap attachment points; scale bars are 10 cm. Helm et al. (2023).

The tracks of bare feet are distinctive, due to the presence of digits and the raised arch of the foot. The impression likely to have been made by a shoe of unknown design is less easy to predict, but it is likely to have followed the general plan of a Hominin footprint, i.e. twice as long as wide, and soft foot-coverings may still preserve the presence of an arch, and rounded front and back margins. It is also quite likely that the front portion of the print will be wider than the rear portion, and that the curve of the front and back parts of the foot will be different. Trackways will reflect the fact that Hominins are bipeds with a narrow stride. In some cases, distinctive traces may be made by the footwear, such as attachment marks from straps.

A pair of sandals on exhibit in the Zambezi Heritage Museum in Katima Mulilo, Namibia, viewed from above (A), and below (B), showing strap attachment points; scale bars are 10 cm. Helm et al. (2023).

Previous attempts to recreate impressions likely to have been made by ancient footwear mostly concentrated on the use of soft leather shoes by persons walking over clay substrates. These studies found that shod tracks were typically longer and narrower than unshod tracks, and that this became more exaggerated as the substrate got moister and the footprints deeper. Shod footprints were also typically simpler than unshod ones, most obviously in the lack of toe-prints. 

These findings may have some implications for studies being carried out on the Cape Coast of South Africa, where some footprints have been considered questionable due to their apparently elongate shape.

Historic sandals in the collections of museums shed light on the appearance of shoes within the region in the recent past, and therefore possibly those that might have been used by people in the remote past armed with similar resources and environmental challenges. These sandals are remarkably symmetrical, with little difference between the width of the front and rear portions of the shoe. The pair from the Zambezi Heritage Museum in Namibia are believed to have been made from Buffalo hide in the 1920s.

The San people who made these sandals are known to have used a variety of Animal skins, with different groups using specific hides, including Eland, Hartebeest, and Wildebeest, while other Animals, such as Gemsbok, were avoided by all groups. Furthermore, different types of sandals were made for different purposes, including special shoes for seniors, shoes for uphill travel, and running sandals with a hook under the toe giving extra purchase when running over soft sand (possibly in imitation of the foot of an Ostrich, the fastest bipedal Animal present in the region).

A pair of ‘running sandals’ in the Fourie Collection in Museum Africa, accession number MM40-69-2416; scale bar is 10 cm. Justine Wintjes in Helm et al. (2023).

Rock art in Southern Africa also provides a record of the use of footwear in the region, albeit one that only stretches back a few thousand years. The artwork at Baviaanskloof in Eastern Cape Province is thought to be about 2000 years old, and includes a 21 cm high figure interpreted as a shaman, naked apart of a pair of laced sandals and a kaross (cloak made from Animal skin). The fact that a largely naked figure is wearing elaborate sandals suggests that these were important items to the people of the region.

(A) A male figure, apparently with footwear and shoelaces (magnified in inset), in a San pictograph from Baviaanskloof.  (B) Rock art of footwear that appears to depict shoelaces, Baviaanskloof. Helm et al. (2023).

In total, four figures with footwear are present within at Baviaanskloof, two of which have identifiable laces. Sandals can also be seen in the rock art of the western US, where Pueblo petroglyphs made between 1000 and 1200 AD also depict footwear.

Helm et al. note that the identification of tracks made by Hominins with footwear can be difficult, and suggest a set of criteria which may make this easier for future researchers. Firstly, any trackway examined should be of sufficient length for the study to be useful. Secondly, Hominins are bipedal, with a narrow gait, and any track made by them should reflect this. Thirdly, Hominin footprints are roughly twice as long as wide, regardless of footwear. Fourthly, Hominin tracks tend to have rounded fronts and backs. Fifthly, any trackway analysed for the presence of footwear should be of sufficient quality, with crisp margins. Sixthly, shod footprints will lack the diagnostic features of unshod footprints, such as digit marks. Finally, the footprints may have diagnostic features associated with the footwear, such as strap attachment points.

See also...

Follow Sciency Thoughts on Facebook.

Follow Sciency Thoughts on Twitter.

Evidence for the construction of wooden structures in the Middle Pleistocene of Zambia.

The first evidence of woodworking is associated with the Early Pleistocene of Oldowan Culture of East Africa, with a number of stone tools showing wear traces and residues left when they were used to work wood. The earliest known wooden tools come from the Middle Pleistocene of Southern Africa, with waterlogged wooden items from several sites, associated with Middle Stone Age and Acheulean stone tools. One such site is at Kalambo Falls in northern Zambia, where a series of wooden artefacts were recovered in the 1950s and 1960s, from horizons which also produced Acheulean tools. These included a wood chip, and three objects with transverse notches interpreted to have been made by intentional shaping of the wood. The Amanzi Springs site in Eastern Cape Province, South Africa, a stick with a possible chop mark was discovered in a waterlogged deposit which also produced Acheulean tools in the 1960s. These deposits were radiometrically dated to between 404 000 and 390 000 years before the present, from unmodified wood within the same strata. A wooden object with cut marks and fine striations from the Florisbad Spring Deposit in Free State, South Africa, was found alongside Middle Stone Age stone tools and remains of the Hominin Homo helmei; this is considered to be the oldest known clearly modified wooden object, but has not been dated beyond Middle Pleistocene.

In a paper published in the journal Nature on 20 September 2023, Larry Barham of the Department of Archaeology, Classics & Egyptology at the University of Liverpool, Geoff Duller of the Department of Geography and Earth Sciences at Aberystwyth University, Ian Candy of the Department of Geography at Royal Holloway, University of London, Christopher Scott, also of the Department of Archaeology, Classics & Egyptology at the University of Liverpool, Caroline Cartwright of the Department of Scientific Research at the British Museum, John Peterson again of the Department of Archaeology, Classics & Egyptology at the University of Liverpool, Ceren Kabukcu, again of the Department of Archaeology, Classics & Egyptology at the University of Liverpool, and of the Interdisciplinary Center for Archaeology and Evolution of Human Behaviour at the University of Algarve, Melisa Chapot, also of the Department of Geography and Earth Sciences at Aberystwyth University, F Melia, again of the Department of Archaeology, Classics & Egyptology at the University of Liverpool, Veerle Rots of the Prehistory TraceoLab  at the University of Liège, Nikki George, again of the Department of Archaeology, Classics & Egyptology at the University of Liverpool, Noora Taipale, also of the Prehistory TraceoLab at the University of Liège, Peter Gethin, once again of the  Department of Archaeology, Classics & Egyptology at the University of Liverpool, and Perrice Nkombwe of the Moto Moto Museum describe recently described evidence for wood being used tob build a structure in the Middle Pleistocene at Kalambo Falls.

New excavations at Kalambo Falls in 2019 recovered five modified wooden artefacts from four different areas, at different layers within the sediments, both above and below the current level of the river. A sixth wooden object showed no signs of intentional modification, but was associated with one of the modified wooden items and some Acheulian stone tools, in a layer below the current river level. One of the other modified wooden objects was also found (separately) associated with Acheulian stone tools, in a layer below the current river level. The remaining modified wooden objects were found without associated stone tools in layers above the current river level, two of them together and one on its own. 

Location of Kalambo Falls archaeological site and excavated areas. (a) Site location in south-central Africa. (b) Course of the Kalambo River (in outline) from around 1956 to 2006 in relation to previous excavations at sites A, B, C, D and C North. Site BLB (2019) and excavation units BLB1, BLB2, BLB3, BLB4 and BLB5 are located along the current main channel (blue). (c) Cross-section of the 2019 excavation units showing the location of 16 luminescence dating samples (KF01–KF17, dark blue circles, uncertainties (±) shown at 1 − σ) by unit. Unit BLB1 is a geological section of the full cliff exposure from ground surface to below water level. The three colour bands indicate clusters of pIR IRSL ages grouped by mean ages and standard error of the mean. The earliest wood objects (BLB5, BLB3) are in the lower green band with a mean age of 476 ± 23 000 years. The blue band has a mean age of 390 ± 25 000 years and incorporates the wood object in BLB2. The overlying yellow band has a mean age of 324 ± 15 000 years and incorporates wood objects in BLB4. Red diamonds indicate modified wood objects. Barham et al. (2023).

At Kalambo Falls a nine-metre-deep exposure of fluvial deposits, interpreted as having been laid down by a laterally migrating river with moderate-to-high energy flows, comprises mostly gravels and course sands, with occasional non-contiguous beds of fine sand, silt, and clay. This deposit has an elevated water table, which has preserved wooden objects and plant matter within the bottom two metres of the sequence. Wooden items are interpreted as either having been deliberately deposited by Hominins or carried to its current location by the flow of the river. 

Modified wood tools from site BLB, Kalambo Falls, 2019. (a) BLB5 structural element (object 1033). (b) BLB3 ‘wedge’ (object 660). (c) BLB2 ‘digging stick’ (object 219). (d) BLB4 cut log. (e) BLB4, tapered piece with single chop-mark. Scale bars are 10 cm. Barham et al. (2023).

Sixteen dates were obtained from this sequence using a combination of optically stimulated luminescence dating of single quartz grains and postinfrared infrared stimulated luminescence dating of potassium-rich feldspars. These dates form a stratigraphic sequence (i.e. the older dates are below the newer dates in the succession), with one exception, and fall into three distinct age clusters. The oldest cluster, centred on 476 000 years ago (and extending approximately 23 000 years in either direction of this date) includes all of the samples from below the current river level, including two modified wooden objects, BLB3, which was found with a piece of unmodified wood and a collection of Acheulian tools, and BLB5, which was found with another group of Acheulian tools. The next cluster, centred on 390 000 years ago and extending 25 000 years on either side of that date, contained a single wooden tool, BLB2, found on its own, above the level of the current river, while the final cluster, centred on 324 000 years ago and extending 15 000 years on either side of that date, includes a pair of wooden artefacts, BLB4, found together.

It was possible to identify the wood from which the artefacts were made, but attempts at carbon dating them produced 'infinite' results, indicating they are more than about 50 000 years old. Infrared spectroscopy of these specimens showed that they had begun to mineralize, with silica beginning to replace the original wood. The tools were interpreted by comparison to later wooden tools from Holocene archaelogical sites in Zambia and the UK, as well as those made by modern Human populations.

The oldest tool, object 1033 from layer BLB5, is a Large-fruited Bushwillow, Combretum zeyheri, log 141.3 cm in length and 25.6 cm in width, which was found overlying another log at an angle of 75°. Importantly, this overlying log has a notch, 13.2 cm long and 11.4 cm wide, intentionally cut into it where it rests on the underlying log, which also shows signs of modification. Bark is still present on both logs, bur is absent around the notch, emphasizing that this was intentionally cut, although this is not in doubt, due to the visible chop and scrape marks on the exposed wood within the notch. 

Structural unit formed by two overlapping logs in BLB5. The underlying log passes through a central notch cut into the upper log (object 1033) and extends into the section. Plan view of the unit (left) and during excavation (right). The numbers refer to the distance in centimetres. Barham et al. (2023).

The notch shows several areas of scrape marks, overlapping in the centre. These are 'v'-shaped in profile, up to 24 mm long and 3.2 mm wide. The end of the log is tapered, and also appears to have been modified.

Annotated images of the BLB5 upper log (object 1033) showing areas of intentional modification. From left to right, the location of the central notch in profile, shaping marks in and on the margins of the notch (a)–(k), the notch in profile from the opposite side. The image on the right shows the upper surface of the log, and the three parts of the log (1)–(3) separated by cracks. White arrows indicate locations of shaping facets on the sides and upper surface of the log. Barham et al. (2023).

The underlying log also shows signs of modification, with a series of small striations running across the grain of the wood at its mid-point, again with 'v'-shaped profiles indicative of scraping, and the end of the log which passes through the notch in the upper log showing signs of having been intentionally thinned by the scrapping of a series of notches along the grain of the wood.

Shaping marks on the upper surfaces of object 1033 and on the underlying treetrunk. Clockwise, from left upper left; chop marks on Part 2; cluster of small convex hewing marks on Part 1, near Part 2; cutmark (upper arrow) and small facets (lower arrows) on Part 1 near Part 3; intercutting chop marks on the upper right edge of the Part 3 taper; underlying log midsection, intersecting cutmarks transverse to the grain (bold arrow, upper left, indicating direction of grain). Marks on underlying treetrunk interpreted as result of scraping, perhaps from debarking. Barham et al. (2023).

Infrared spectroscopy suggests that the notch has also been modified by fire-shaping, something which was reported on another log of similar size (165 cm) previously found with Acheulian artefacts at Kalambo Falls, but not dated, which also had a wide, deep groove cut into it and tapering ends. This wood was presumed to have been part of a structure, and thought to have been modified by the actions of Early Humans. The new discovery of two logs interlocking on a similar notch, strongly supports the idea that the function was constructional in purpose, although the age of the strata from which the wood was recovered considerably pre-dates the oldest estimates for the emergence of Modern Humans.

Shaping marks on the treetrunk underlying object 1033 notch, BLB5. Surface modifications on the treetrunk beneath and just beyond the notch. Images described counter clockwise: white arrows highlight direction of striations, some V-shaped in profile, cutting across the grain from the edge toward the middle of the surface; a set of parallel transverse striations/cutmarks, with V-shaped profiles (white arrows) along ridge descending to a flat tapered surface; inset detail of the striations on the ridge crest; overview of the set of striations along the ridge; profile view of the flattened surface below the ridge; profile view of the ridge and break of slope onto the flat tapering surface which extends beyond the notch; plan view of the flat tapering end with transverse striations and cutmarks with V-shaped profiles (white arrows) interpreted as shaping marks. Barham et al. (2023).

The second collection of material below the modern river level, BLB3, comprises another collection of Acheulian tools, including flake tools, cleavers, handaxes and core axes, as well as two wooden objects, a 'v'-shaped piece of Figwwod, Ficcus sp., showing no signs of intentional modification, and a flattened branch of Sausage Tree, Kigelia africana, 36.2 cm in length and with one end carved to a point.

Shaping and possible use marks on object 660, BLB3. Surface modifications on ‘wedge’ shown clockwise: white arrows indicating location of modifications on base and either side of the tip; basal crack (top arrow) and side split, facet ‘a’ intercepts a striation, possible use damage, and ‘b’ indicates a set of faint transverse striations; ‘c’ and ‘d’ intersecting facets above tip, ‘e’ and ‘f’ are sets of parallel faint transvers striations; tip face with convex step terminations, possibly from use. Barham et al. (2023).

Barham et al. believe that this artefact has been intentionally shaped by repeated high-impact blows, in order to form a wedge. Similar wedges have been recovered from the Mesolithic Starr Car site in the UK, and are known to have been used for a variety of purposes by Aboriginal Australians, but nothing similar has ever been discovered in a Middle Pleistocene deposit.

The BLB2 horizon, dated to about 390 000 years ago produced a single wooden item, broken into two pieces which fitted together; one of these was found still embedded in the sediment, the other in the river directly below, where it had apparently fallen. The combined wooded object is about 62.4 cm in length, made of African Sausagewood, stripped of all bark and intentionally fashioned into a point, being 11.9 cm wide at the base, 6.1 cm wide in the middle, and 1.3 cm wide at the tip. The object has striations cutting across the grain of the wood in several places, these being less than 10 mm wide and convex in profile. The tip of the object is slightly rounded, which might be a sign of wear from use. or derive from its time in the river. The object is interpreted as a digging stick similar to those used historically by foragers in the Kalahari. Similar artefacts are known from the Middle Holocene of Zambia, but again the Middle Pleistocene occurrence is unprecedented. 

Shaping marks on object 219, BLB2. Upper image shows all views of the object with white arrows indicating location of marks on either side of the tip and above the tip. Both sides of the tip (two images on left) have a faint set of transverse striations and two or more small facets away from the tip. (Scale bar is 1 cm.) The vertical image to the right shows four groups of faint parallel transverse striations in an area with two knots, no scale. The object is interpreted as a digging-stick. Barham et al. (2023).

The BLB4 horizon, dated to about 324 000 years ago, produced two wooden artefacts and no stone items. The uppermost of these is a rectangular piece of Large-fruited Bushwillow wood measuring 59.24 cm by 29.34 cm by 7.7 cm, with traces of bark and sapwood still present. This shows signs of having been flattened, but this may have been due to the weight of the sediment above it. It also has a large chop mark running across its obverse surface. Barham et al. interpret this item as a piece of tree trunk which has been intentionally cut to size, indicating a capacity to process large pieces of wood.

Chop marks on ‘cut log’, BLB4. Upper image shows all views of the object and location of modifications (white arrows) on the ends, ‘a’ and ‘b’, two marks on the underlying surface and the preservation of bark on both surfaces. Image below is ‘a’ with its distinctive stepped chop marks (black arrows), and underlying grey image (denatured RTI) shows the width of the step between the first and second chop-mark expanding, left to right, from 13 to 25 mm. The deep, slightly convex cut of the first chop is visible. The underly step expands left to right from 6 mm to 11 mm. A third step starts with a width of 7 mm. Bottom image is end ‘b’ with black arrows indicating location of chop marks. Barham et al. (2023).

The second item is a Sausagewood branch, with a side branch 37.9 cm in length, tapering from a 12.3 cm wide base to a 2.1 cm wide tip. The wood has been split centrally, and has a transverse chop mark above the tip. The purpose of this artefact is unclear.

Images of split branch, BLB4 (‘notched stick’). Above, close-up of the facet/chop-mark with arrow indicating direction of initiation. Below, all views of split branch with chop (white arrow) and close-up of the entry facet. The object was found cracked in situ. Barham et al. (2023).

The Kalambo Falls site has yielded a number of exceptionally well-preserved, intentionally modified wooded objects, with the most recent discoveries confidently dated to the Middle Pleistocene (and previous, undated discoveries likely to be of similar vintage). This has significant implications for our understanding of the technological capabilities of earlier members of the genus Homo, which have previously only been known from stone tools. The implications are clearly that these people were capable of felling large trees, and constructing objects such as platforms, which would have required considerable skills in selecting trees, felling them, and processing the wood obtained. Since this appears to have been done with stone tools, it shows that stone- and woodworking evolved together, and helps to explain some of the larger stone tools produced as part of the Acheulean technology. 

Between about 470 000 and about 274 000 years ago the Kalambo Falls site would have had an extensive forest cover and a high water table, leading to periodic flooding of the river plain. In such an environment the ability to produce raised platforms, wooden walkways, or even structures with solid foundations would have been highly advantageous.

The invention of hafted tools (tools with wooden handles) is thought to have occurred between about 500 000 and about 200 000 years ago, and would have required a good understanding of the properties of wood, and how to shape it by chopping and scraping. The interlocking logs at the lowest dated level at Kalambo Falls show a similar level of technological cognition, as well as a previously unanticipated ability to create the first built environments.

See also...

Follow Sciency Thoughts on Facebook.

Follow Sciency Thoughts on Twitter.