January 23, 2022

Age of the oldest known Homo sapiens from eastern Africa – Nature

Only eight sites in Africa have yielded possible early anatomically modern Homo sapiens fossils from the late Middle Pleistocene (approximately 350–130 thousand years ago (ka))11. Most of these have considerable age uncertainty or debatable H. sapiens apomorphy11. A principal method for constraining the fossil ages is the use of single-crystal 40Ar/39Ar isotope dating applied to stratigraphically associated volcanic ash (tephra) beds12,13,14. However, many distal tephra deposits consist largely of glass and lack suitable crystals for dating. In this case, geochemical fingerprinting can be used to match a tephra layer to more readily dated proximal deposits with larger, more abundant phenocrysts. The most widely accepted fossils that are interpreted as possessing unequivocal modern cranial apomorphies (that is, a tall cranial vault and a chin) and classified as H. sapiens are two Ethiopian finds11,15,16, namely the Omo I1 and Herto specimens4. Accordingly, the evidence that constrains their ages assumes particular importance but is a topic of considerable geochronological controversy3,6,8.

The Omo I remains were discovered in the late 1960s in the lower Omo valley of southern Ethiopia1,14, at the surface of a siltstone near the top of Member I of the Omo-Kibish Formation (Fig. 1a, b). The maximum age of Omo I was derived from the 40Ar/39Ar age of 196 ± 4 kyr (2σ)3,6,17 obtained for alkali feldspar phenocrysts from the three youngest pumice clasts that were sampled from a heterogeneous tuffaceous deposit correlated with the Nakaa’kire Tuff3, which is reported to lie “near, but probably slightly below” the fossils3 (Fig. 1b). Recalculated using a more widely adopted age of 28.201 million years (Myr) for the irradiation monitor (sanidine from the Fish Canyon Tuff of Colorado)18, the Nakaa’kire Tuff age shifts marginally to 197 ± 4 kyr. Owing to the uncertain stratigraphic relationship between this tuff and the hominin fossils19, much attention has been focused on dating the KHS Tuff—a widespread, more-than-2-m-thick deposit of fine ash fallout at the base of Member II of the Omo-Kibish Formation (Fig. 1b). The KHS Tuff overlies Member I, where Omo I was retrieved around 1.4 m lower down section, and is demonstrably younger than the fossils3,9. Although the Nakaa’kire Tuff was identified in several sections below the KHS Tuff, the latter was not found in the same section from which the dated pumice clasts correlated with the Nakaa’kire Tuff (on the basis of major element composition) were sampled3. The fine grain size of the KHS Tuff has precluded direct 40Ar/39Ar dating, and no correlation to a source volcano or proximal pyroclastic unit has to our knowledge been made previously. However, drawing on published major element glass compositions, it has been correlated with both tephra TA-5520,21 from the Konso Formation and the directly 40Ar/39Ar-dated 184 ± 10 kyr unit D22 (recalculated age) of the Gademotta Formation6 (Fig. 1b). Relating the sediment flux in the Omo-Kibish basin with high lake levels that correspond to Mediterranean sapropel deposition9,23, a slightly younger age for the KHS Tuff of around 172 kyr has also been proposed6. Either of these ages (184 or 172 kyr) would be consistent with the proposed age of 197 ± 4 kyr for Omo I.

Fig. 1: Late Middle Pleistocene tephrostratigraphy of the Main Ethiopian Rift.

a, Map of the MER showing silicic volcanoes and the late Middle Pleistocene sedimentary formations and relevant tephra units. White boxes with blue edges depict former correlatives of the KHS Tuff6,8 b, Synthetic stratigraphic logs of the late Middle Pleistocene formations showing former correlations for the Alyio Tuff6 (green), Konso SVT (pink, also identified in the Chew Bahir sediment core33), new correlations for Konso unit TA-56 (yellow), and source eruptions (stars). LHM, lower Herto Member; UHM, upper Herto Member. c, Tephra ETH18-8 above the KHS Tuff at the KS locality in the Omo-Kibish Formation9.

The Herto H. sapiens fossils were recovered in the late 1990s in the Middle Awash4,5 (Afar rift; Fig. 1a). They were preserved in a sandstone within the upper Herto Member of the Bouri Formation (Fig. 1b). This sandstone is capped by the Waidedo Vitric Tuff (WAVT) (Fig. 1b), which is widespread across western Afar and is also present at Gona24, 50 km north of Herto. Direct dating of the WAVT has remained inconclusive owing to crystal contamination, but dating of pumice and obsidian clasts in the fossiliferous sandstone yielded a maximum age of around 160 kyr (ref. 5). The WAVT was identified as a distal correlative of tephra TA-55 (Fig. 1b), on the basis of major element analysis of individual grains and major and trace element analysis of purified bulk separates5,25. In Konso, unit TA-55 lies below the 155 ± 14 kyr Silver Tuff5 (SVT) (recalculated age) (Fig. 1b), suggesting an age for the Herto fossils of around 160–155 kyr (ref. 4). This finding was challenged, however, in a study6 that correlated the Kibish KHS with Konso TA-55, and therefore with the Herto WAVT (Fig. 1b). This argument suggested an age of around 172 kyr for the WAVT, contradicting the established Herto stratigraphy. The Herto research group8 responded by corroborating their original stratigraphy, with the WAVT above the Herto fossils, thus challenging an age of about 172 kyr for the KHS. They concluded that the KHS, Konso unit TA-555, Gademotta unit D (around 184 kyr)22 and WAVT5 could all represent a single tephrostratigraphic marker lying above the Omo-Kibish and Herto H. sapiens fossils, but that multiple eruptive sources would also be plausible8 (Fig. 1b). Given the lingering uncertainties of the stratigraphic relationship of the Nakaa’kire Tuff to Omo I, the age of the KHS Tuff becomes critical to the chronostratigraphy of these sites.

We have re-sampled the KHS Tuff and other pertinent ash deposits at Omo-Kibish, Konso and Gademotta to assess the geochemical correlations from which the ages of the oldest modern human fossils are inferred. While revisiting the sampling locality of the KHS Tuff (KS type section)9 at Omo-Kibish, we sampled another tephra layer in Member II (Fig. 1c) in an outcrop about 100 m from the KS type section. Unit ETH18-8 is an approximately 15-cm-thick, very well-sorted crystal-rich fine sand grey tephra layer situated 40 cm above the KHS Tuff (Fig. 1c). It is ubiquitous between the KHS section (KS) and the Chibele section (CB), and might stratigraphically correspond to unit CRF-23 previously identified above the KHS Tuff at the CB section9, although this cannot be confirmed through geochemical analysis because of the different microprobe conditions used.

In an attempt to identify and date the eruption that generated the KHS tuff, we included samples of ignimbrites from the caldera-forming eruptions of Shala and Corbetti volcanoes. Shala and Corbetti are the only Main Ethiopian Rift (MER) systems known to have produced major eruptions between around 170 ka and 250 ka26. At Shala, the largest caldera in the central MER (Fig. 2a), we sampled at a more-than-20-m-thick exposure of the unwelded Qi2 ignimbrite27 (Fig. 2b, c), southwest of Lake Shala and 350 km northeast of Omo-Kibish (Fig. 2a). We also analysed glass from a welded ignimbrite (COI2E) attributed to the formation of Corbetti caldera, dated at 177 ± 8 kyr (ref. 26). A challenge of geochemical correlations between proximal and distal tephra deposits in the region is similarity in major and trace element compositions between pyroclastic products, not only of the same volcano but of different volcanoes in the MER28. Accordingly, correlations are ideally based on a detailed suite of major, minor and trace element single-grain glass shard or pumice glass analyses.

Fig. 2: Stratigraphy and age of the Shala Qi2 ignimbrite.

a, Location of site ETH17-14 near Lake Shala in the MER. b, Synthetic stratigraphy of the Qi2 ignimbrite of Shala at location ETH17-14. c, Photographs of units 14A, 14B and 14C of the Qi2 ignimbrite at site ETH17-14. Field observations indicate that deposits 14A and 14B are subunits of the same phase of the Qi2 eruption. d, 40Ar/39Ar age pooled data plotted on ideograms for samples 14A and 14C of the Qi2 ignimbrite (bottom) yielding a preferred composite eruption age of 233 ± 22 kyr (top). Data are weighted means. Error bars show data and results at 2σ. 40Ar*, radiogenic 40Ar; MSWD, mean square of weighted deviates; P, probability that residuals are explained by measurement errors exclusively; n, number of accepted grains.

The KHS glass shards are homogeneous pantelleritic rhyolite in composition (77.0 ± 0.3 wt% SiO2, 9.7 ± 0.1 wt% Al2O3, 5.0 ± 0.1 wt% FeO* (FeO* refers to the total Fe as FeO) and 7.1 ± 0.4 wt% Na2O+K2O; Supplementary Table 1). Immobile oxide abundances, including FeO*, CaO, Al2O3 and TiO2 (Fig. 3, Supplementary Table 1), correspond with those of glasses from the proximal products of the Qi2 eruption of Shala volcano (samples ETH17-14A1, B1, B5 and C) (Figs. 2b, c, 3, Supplementary Fig. 4, Supplementary Table 1, Supplementary Information). These correlations are corroborated by comparing immobile trace element ratios for Qi2 and KHS glasses and principal component analysis (Fig. 3, Supplementary Figs. 4, 5, Supplementary Table 2, Supplementary Information).

Fig. 3: Geochemical fingerprints of MER tephra and their sources.

Major element abundances and trace element ratios of glasses from the Shala Qi2 ignimbrite (around 233 kyr), the Corbetti ignimbrite (around 177 kyr), the Gademotta unit D (around 184 kyr), the Kibish KHS and ETH18-8 tuffs, and the Konso TA-56 tuffs (all data from this study). Major element data are normalized to 100% anhydrous. Error bars shown are relative standard deviations derived from repeat measurements of matrix match glass secondary standards STH-S6 (for FeO*, n = 91; Supplementary Table 6) and ATHO-G (for Al2O3, CaO and TiO2, n = 70; Supplementary Table 6). They are plotted in the top right corner of each plot for clarity and rescaled to the value of the centre point. In the case of element ratios, error propagation has been applied using analyses of standard ATHO-G (n = 15; Supplementary Table 7). Additional compositional observations and biplots are presented in Supplementary Fig. 5.

In addition, we find that the COI2E pantelleritic rhyolite glass from the 177 ± 8 kyr (ref. 26) Corbetti ignimbrite (74.3 ± 0.2 wt% SiO2, 9.1 ± 0.1 wt% Al2O3, 5.6 ± 0.2 wt% FeO* and 10.1 ± 0.2 wt% Na2O+K2O) (Fig. 3, Supplementary Fig. 4, Supplementary Table 1, Supplementary Information) has immobile oxides and trace element abundances that match those for Kibish unit ETH18-8 and Konso TA-56 (Fig. 3, Supplementary Figs. 4, 5, Supplementary Table 2, Supplementary Information).

We used the 40Ar/39Ar dating method to analyse 113 individual sanidine crystals extracted from pumice samples ETH17-14A1 (base, 68 crystals) and ETH17-14C (top, 45 crystals) collected from the Shala Qi2 deposits (Fig. 2). The resulting data were filtered to exclude grains with low gas yields, at or below blank level, and xenocrysts with ages significantly older than the mean of the dataset (six grains with ages exceeding 1 Myr). The distributions of ages from each sample were indistinguishable at 2σ uncertainty (Fig. 2d). Combining analyses from both pumice samples yielded a weighted mean of 233 ± 22 kyr at 2σ (Fig. 2d, Supplementary Table 3), thereby dating the Qi2 eruption and the KHS tuff.

An age of 233 ± 22 kyr for KHS is consistent with the 177 ± 8 kyr age that we associate with the overlying ETH18-8 tephra (Fig. 1b). However, it casts doubt on the suggested correlation between high deposition fluxes in the Omo basin with large in-flows of fresh water from the Nile River system into the Mediterranean sea6,7,9, at least during the formation of Member II. Our KHS age is incongruent with the formation of Mediterranean Sapropel S6 at 172 ka6, and instead overlaps the timing of the formation of sapropel S8 (217 ka)9,29. Although the 177 ± 8 kyr age of ETH18-8 is consistent with the formation of sapropel S6 (172 ka)29, only a mudstone unit of around 40 cm thickness separates KHS from ETH18-8, which cannot account for the suggested rapid deposition in the basin concomitant with sapropel S7 (192–199 ka)3.

The revised Omo-Kibish stratigraphy is also incompatible with the 197 ± 4 kyr age reported for the Nakaa’kire Tuff3,7,9, which is found in Member I of the formation3,7,9 and which must therefore be older than 233 ± 22 kyr. The age of 197 ± 4 kyr was inferred from three out of five dated pumice clasts from lenses found in ‘a sandy tuffaceous matrix’7. Although these samples had similar major element compositions to the Nakaa’kire Tuff, they were collected from a lateral outcrop and not in section3,7,9. Given the uncertainty in the age and stratigraphic placement of the Nakaa’kire Tuff, as well as its heterogeneous lithology and geochemistry, the identification of the 233 ± 22 ka Qi2 eruption of Shala as the source of the KHS Tuff provides a more robust minimum age for Omo I H. sapiens.

Furthermore, our glass compositional data, source correlation and age estimate for KHS allow us to re-assess its identification at other archaeological sites in Ethiopia. New lithological examination of the pedogenically altered unit TA-55 at Konso (Supplementary Fig. 1) in grain size fractions of greater than 125 µm, greater than 80 µm and greater than 25 µm, after density separation, failed to identify glass shards in this deposit, which was previously correlated with the WAVT at Herto. This precluded evaluation of the reported correlation with the KHS Tuff6. However, with the underlying unit TA-56 now correlated with Kibish unit ETH18-8 and the 177 ± 8 kyr Corbetti ignimbrite (Fig. 3, Supplementary Figs. 4, 5), it is clear that TA-55 is younger than 177 ± 8 kyr and so cannot correlate with Qi2 or the KHS Tuff.

Although the 184 ± 10 kyr unit D of Gademotta appears close to KHS in major element contents, neither major nor trace element abundances clearly overlap (Fig. 3, Supplementary Figs. 4, 5, Supplementary Information), precluding a match. Immobile trace element ratios and principal component analysis show that unit D also differs from TA-56 (Fig. 3, Supplementary Figs. 4, 5, Supplementary Information).

The correlation of the Herto WAVT with Konso unit TA-555, around 800 km south of Herto, led earlier investigators to accept the 155 ± 14 kyr age of the SVT at Konso as the terminus ante quem of the Herto fossils. This correlation has been debated30 but reinforced by additional geochemical data25. We were unable to find preserved glass in our TA-55 sample but our results undermine the tephrostratigraphic correlations proposed between the Omo-Kibish, Gademotta and Konso formations6 and bracket the age of the Konso TA-55 tuff between 177 ± 8 kyr (TA-56) and 155 ± 14 kyr (SVT). Although its correlation with the WAVT at Herto should be confirmed in the future using grain-discrete single-point glass analyses, this age bracket is consistent with the underlying Herto fossiliferous sandstone (approximately 160 kyr)5, and confirms that the Herto H. sapiens fossils are considerably younger than Omo I at Omo-Kibish.

Our new age constraints are congruent with most models for the evolution of modern humans, which estimate the origin of H. sapiens and its divergence from archaic humans at around 350–200 ka (refs. 16,31,32). The challenge remains to obtain a robust maximum age for Omo I. Our revised tephrostratigraphy demonstrates that the Herto specimens postdate the Omo I remains from Omo-Kibish, and that they do not lie beneath the same tephra horizon as the Kibish fossils, as previously inferred8. Further geochemical data are needed to clarify the relationship between the WAVT and other MER tephra, and may ultimately identify the WAVT source, promising a more reliable minimum age for the Herto fossils. More generally, continued efforts to develop the tephrochronological framework for eastern Africa will help in addressing a range of interrelated volcanological, palaeoenvironmental and palaeoanthropological questions.

Source link