télécharger 352.15 Kb.
|Text S1. Previous evidence of hominin occupation in the Western Rift Valley|
Two relatively well-studied regions of the Western Rift Valley (the Upper Semliki and the Ugandan side of Lake Albert) have yielded long geological sequences with rich fauna and some archaeological remains.
In the Upper Semliki Valley of the Democratic Republic of Congo, the sequence includes a Plio-Pleistocene fluvial lacustrine series (Lusso Formation), a Lower to Middle Pleistocene fluviatile series (Semliki Formation) followed by the Middle to Upper Pleistocene initiation and development of terrace formation along the South flowing Semliki (Kasaka Formation, Katanda Formation, Ishango Gravels Formation, Museya Gravels Formation) . Biface industries with spheroids and some evidence of Levallois technology are documented in several localities within the Semliki Formation  or on the subsequent Upper Terrace (Kasaka Formation) of the river. At Katanda, 6 km north of Ishango, in two locations, excavation of an Oldowan assemblage, associated with Elephas recki in the upper part of the Semliki Formation indicate that the Semliki Formation in large part does not postdate ca. 500 ka, and is likely much older toward the base . These finds and abundant evidence of artefacts in the Ugandan sites of the Albertine Rift Valley  indicate consistent occupation of the Western Rift, presumably by Homo erectus, during the late Lower to early Middle Pleistocene.
At Nyabusosi on the Ugandan side of Lake Albert, towards the top of the Behaga Member of the Nyabusosi Formation, cranial fragments and a worn upper molar, cemented into a ferricrete block and attributed provisionally to “Homo cf. erectus” were recovered in 1986 by Pickford and Senut . The remains were first dated to 0.7 to 1.8 Ma with a preference for the older age , and later revised to "ca. 1.5 Ma"  on the basis of the presence of Kolpochoerus majus and Elephas recki in the fauna collected near the fossil.
The 2.85 - 0.4 Ma time-range of “classic” E. recki  as well as early Middle Pleistocene dates for both Elephas recki and especially Kolpochoerus majus, at other east African sites [8-11] suggest that this hominin specimen could also derive from horizons around or even younger than the Lower/Middle Pleistocene boundary at 780 ka.
However, a tuff in the uppermost Kagusa Member, which overlies the Behaga Member of the Nyabusosi Formation, was potentially correlated chemically with a tuff in the Turkana Basin Natoo Member of the Nachikui Formation (1.65 - 1.3 Ma) implying an older age . Therefore some questions remain concerning the age of this hominin specimen.
While hominin presence during the time span of Homo erectus is uncontroversial, earlier Plio-Pleistocene hominin occupation is more debated. Within the Semliki Valley, lithic artefacts provisionally associated with the Lusso Formation were described from two locations: Kanyatsi 2 on the lake shore ca. 3 km east of Ishango, and Senga 5A, ca. 10 km north of Ishango along the Semliki River. At Kanyasti, a very small number of core artefacts were recovered either in situ or with adhering matrix suggesting attribution to the Lusso Formation. These could also be derived from younger deposits through slumping, although the 1988 excavations did not support this hypothesis .
The Senga 5A site is located on a low ca. 11 m terrace comparable to the Ishango terrace, ca. 10 km north of Ishango [13-15]. The entire site exhibits indications that it was formed or at least affected by Upper Pleistocene and Holocene sedimentary processes. However, a large collection (>5000) of diagnostic Plio-Pleistocene mammalian and fish fossils (e.g. Tragelaphus nakuae, Notochoerus euilus and Metridiochoerus jacksonii) suggesting contemporaneity with the Omo Valley Shungura Formation F-G in the Eastern Rift (2.0 - 2.35 Ma) were recovered from a derived ironstone rubble. These were associated with a much smaller number of artifacts, some of which exhibit typical Lusso Formation iron-staining in the cracks, suggesting possible contemporaneity with the Plio-Pleistocene fossil assemblage . Sediments in the site contained the heavy mineral perovskite, which is a marker of Upper Pleistocene volcanic activity in the nearby Katwe volcanic field. The western portion of the site contained a mixture of Lusso-age fossils and more sub-fossilized specimens of probable Holocene age. On the other hand, faunas from the eastern portion were overwhelmingly of Lusso age and contained little evidence of mixing with extant species. As a whole, the Senga 5A site suggests but does not confirm the possible presence of tool-making hominins in the Semliki by 2 Ma.
Text S2. Ishango site
Text S2.1 History of excavations
The site was first identified by Damas during his 1935-1936 expedition in the Upper Semliki, which focused on recent lake fauna and flora . Damas collected some bone material at Ishango that included a fragment of human mandible at a level equivalent to 180-200 cm below the surface in a partially cemented matrix with large mammal bones .
In 1950, a new expedition was set up by de Heinzelin in order to find more archaeological sites and human fossils in the Upper Semliki region [2, 18-19]. At Ishango, he excavated two trenches, one parallel (N143GE) and the other perpendicular (N43GE) to the river shoreline (Figure S1). The southeastern edge of the N143GE trench bordered the 1935 survey of Damas . The two trenches met up northwestward at the border of the lake cliff. To the west of their junction, limited by gully erosions, an area of 100 m2 preserved only the lower parts of the archaeological levels (Ishango Gravels Formation, TT). De Heinzelin started his excavation on the 23th of April 1950 in this western zone (Corner W), noting later  that most of the human remains, fauna and artefacts came from the area west of the junction between his two trenches (zones A & B on Figure S1). He observed that the archaeological levels gradually disappear all around, though large remnants remain present in both trenches and between the trench N43GE and the cliff.
Further excavations at Ishango include the 1959 extension of the N43GE trench by Splingaer, whose material is unpublished. In 1985-1986, two four-meter-square excavations, one near the junction of the two trenches and one further into the cliff, were carried out by Brooks and members of the Semliki Research Expedition to obtain materials for dating [20-21].
Figure S1. The excavation plan at Ishango 11, modified following de Heinzelin .
A: area preserving only the basal part of the Ishango Gravels Formation (excavated first, between the 23th of April and the 9th of May 1950. This area was initially named "Field E", then "Field W" and finally "Corner W" in de Heinzelin's notes). B: area where the tuffaceous levels where present in addition to the latter ones (excavated between the 17th and 19th of July 1950). No grid system was laid out in areas A and B.
Text S2.2 Stratigraphy and Dating
The lithostratigraphic units identified at the Ishango 11 (Figure S2) are as follows from top to bottom: R/TP/TT/L .
The top part of the sequence are Recent deposits (symbol R) correlated to the uppermost Holocene. At Ishango 11, they include the Post-Emersion Zone (ZPEm) occupation level which was first described by de Heinzelin  and dated between 1,680 ± 80 CalBP to 3,170 ± 90 CalBP by 14C on charcoals .
The informal Semliki Lower Terrace Complex (Symbol T) which precedes the Recent deposits regroups the Ishango Gravels Formation (symbol TT) and the Museya Gravels Formation (symbol TP). These fluvial deposits are located at about 9 to 12 m above the present-day Semliki river . The Museya Gravels Formation (symbol TP) is correlated with the middle Holocene and immediately posterior to the volcanic tuffs [1,19]. The stratotype of the Ishango Gravels Formation (symbol TT) has been defined at IS-11 . This fluvial deposit with linked deltaic and slope deposits, and volcanic admixture, is dated to the uppermost Pleistocene to the lower Holocene . At the time of its deposition, the course of the Semliki River was inverted (south-flowing) compared to the present day [18,22].
Figure S2. Schematized stratigraphic section of the ten first meters of Ishango N43GE trench, modified following de Heinzelin [1,19].
R: Recent Formation; TP: Museya Gravels Formation; TT: Ishango Gravels Formation; L: Lusso Formation. ZPEm: Post Emersion Zone; N.TUF: Tuffaceous levels; SD-SFM: Hardened Sand - Fine Micaceous Sand; NFPr: Principal Fossiliferous level; G.INF: Inferior Gravels. Stars: archaeological assemblages containing human remains.
The detailed lithological and stratigraphic description of the Ishango Gravels Formation is given by de Heinzelin [18-19]. Four major levels are shown on the trench sections (Figure S2):
At IS-11, the lower level of the Ishango Gravels Formation (G.INF) cuts the oldest fossiliferous deposits identified in the Upper Semliki basin, the Lusso Formation (symbol L). The Lusso Formation, first identified as Kaiso beds , is lake deposit mudstones with sands and ironstones [15-16]. They are rich in mammal, fish and mollusk fossils as well as fossil wood. The faunal assemblages at several sites support an age from the later Pliocene to the earliest Pleistocene. The mollusk assemblage of the Lusso Formation has been correlated to the associations GX and GX' from the Kaiso/Hohwa beds, dated to 2.6 - 2 Ma [4, 32-33]. Some divergences between the Lusso mollusk assemblage and the Ugandan references have been underlined by Pickford et al. . They could be related to a slightly earlier age of the Lusso assemblage or to the pene-contemporaneous fragmentation of the Lake Obweruka into two sub-basins (the basins of the proto-lake Edward and of the proto-lake Albert), also suggested by the ichthyofauna . The mammalian fauna indicates a biostratigraphic correlation with the fauna from Members F and G of the Shungura Formation in Ethiopia, between 2.35 and 2 Ma [35-37]. The base of the Lusso Formation has not been observed and the top of the beds is an erosive contact .
Text S2.3 Origin of #Ish25 upper molar
De Heinzelin recorded for each geological sample a clear description of contents, which facilitates partial reconstruction of the spatial organization of the human remains. However, no mention of an isolated tooth was found within any of the human samples described in de Heinzelin's notebooks. By cross-checking the information of several archives from the 1950s excavation (Figures S3-S8) the authors were able to establish the origin of #Ish25. Below are selected translations of hand written letters from de Heinzelin to Twiesselmann dated to the very first days of the excavation at Ishango, and the first page of de Heinzelin's notebook:
- Letter to Mister Twiesselmann dated to Sunday 23rd of April 1950 (Figures S3-S4).
Front page - Black ink:
Here ends the first day of excavation. Or rather of clearing. The stratigraphy starts to show up and we already have two harpoons and some worked quartz. the Director must come tomorrow with the headquarters from the Parcs. It's why I have opened the steam from today. Not yet of humans.
At the bottom of the front page, under de Heinzelin signature, but with a Turquoise ink:
Good news. I have the base of a human tibia and maybe some teeth. Results from 1/4 of m3 from the fossiliferous gravel. Tuesday 25th at 10h the Director has not yet arrived.
Back page - Turquoise ink:
Report Tuesday night 25:
21 fragments of worked quartz unrolled
8 fragments of rolled worked quartz
Hundreds of broken or worked bone fragments
1 complete tibia
1 huge base of tibia
1 tibia head of a immature individual, detached from the cartilage
2 femoral heads very partial human or humanoïd
1 humerus fragment Unbelievable mix of human 2 radius fragments and robust traits
1 cubitus fragment
5 complete or partial foot bones
1 parietal or frontal fragment
5 portions of cranium probably belonging to a large primate
One hundred kg or so of mammal bones, sauria, fishes and maybe birds.
I'm crumbling under the weight of the samples and am eaten by the sun, which is a blessing. I suggest that the first cleaning is done on the premises, despite that mobilize workers, otherwise we will never cope in the lab.
Not yet any sight of the administrative staff.
Figure S3. Letter from de Heinzelin to Twiesselmann dated to Sunday 23rd of April 1950. Recto.
"Dimanche 23/IV/50. Monsieur Twiesselmann. Voici achevées les premières journées de fouilles. Ou plutôt de déblaiement. La stratigraphie commence à apparaître et nous avons déjà deux harpons et des quartz taillés. Mr le Directeur doit venir demain avec tout l'état-major des Parcs. C'est pourquoi j'ai ouvert la vapeur dès aujourd'hui. Pas encore d'humain. Je vous assure que c'est une école pas ordinaire que d'entamer ces fouilles : pas de langage commun avec les travailleurs (20), matériel grossier et incomplet, climat équatorial, avec biana(?) tendu du au climat et à la proximité des éléphants et des hippo qui viennent patrouiller jusque dans le camp. Ne pourriez-vous me faire expédier toute notre provision de petits ?acs. Ces petits ? sont l'un de mes meilleurs soutiens. Journaux et caisses sont presque des denrées rares. Et il faut nourrir des piocheurs. J'espère vivement pouvoir vous donner un jour de bonnes nouvelles. J'arrête d'écrire car suis dans une tempête d'éphémères. Avec mes salutations et meilleur souvenir au laboratoire. J. de Heinz.
Bonne nouvelle. J'ai la base d'un tibia humain et peut-être des dents. Résultat d'1/4 de m³ de gravier fossilifère.
Mardi 25 à 10 h. Mr le Directeur n'est pas encore venu."
«identité méditerranéenne» et un véritable dialogue interculturel, à partir des similarités et des interactions sociales mises en...
«Mettre en œuvre une démarche expérimentale pour mettre en évidence le rôle d’un catalyseur»