Homo Sapiens: Anatomically Modern Humans (AMH)

Uranium-Series Dating
The elemental U and Th values vary greatly along the profiles of the bone sample from the Iwo Eleru skeleton (left-hand diagrams in Figures S1 and S2). The outer surfaces and pores show measureable Th concentrations, which indicate the presence of detrital material. There is a gradient of higher U-concentrations in the centre of the bone (around 80 to 100 ppm) towards the outsides where U-concentrations of 50 to 60 ppm were measured. These lower U-concentrations may either be due to a denser bone structure, with less internal surface area for adsorbing uranium, or to leaching of U from the volumes close to the surface. The apparent U-series age estimates vary between about 10 to 14 ka in the central part of the bone and increase in some laser scans to between about 15 to 17 ka closer to the surface. The right-hand columns of Figures S1 and S2 show the plot of apparent U-series age versus U-concentration.

It is expected from the diffusion-adsorption model for U-uptake [9]–[12] that spatially resolved analyses across a homogeneous bone yield u-shaped or constant U-concentration and apparent U-series age profiles. In ideal circumstances, a plot of apparent U-series age versus U-concentration would either be flat or show increasing U-series age estimates with increasing U-concentrations. This is clearly not the case for the results on Iwo Eleru (Figures S1 and S2, right hand columns). The apparent U-series age estimates are more or less constant in the inner part of the bone fragment and increase towards the outside while the U-concentrations decrease.

One of the problems of U-series dating of bone is that the different domains within a bone may give very different U-series ages. The best age estimate is actually not derived by simply averaging all results, but from identifying the domains that had experienced the fastest U-uptake and have remained a closed system since. This may only apply to very small volumes (see [13]). The U-series analyses of the central part of the bone give a minimum age of the sample. Excluding any results with more than 10 ppb Th (indication of contamination) and U concentrations of less than 75 ppm, all scans give compatible results with mean ages ranging between 10.6±1.7 (Track 2) and 12.7±1.6 ka (Track 4), see right hand columns in Figures S1 and S2. Most apparent U-series age results from the central parts of the bone are within 9 and 14 ka. Virtually all age results overlap within their errors. Most of the scatter is likely due to statistical variation. The best age estimate is thus derived from the mean value of all Th-free age estimates in the central part of the bone (11.7±1.7 ka). This age estimate is derived from more than 80% of the exposed bone area (excluding the pores), which may be taken as an indication that most of the uranium was accumulated within a relatively short time range. The average U-series age increases by less than 1 ka (i.e. less than 10%) by including the Th-free results of the domains with lower U-concentrations. Considering that the outer U-concentrations are lower by up to a factor of four compared to the interior part, U-leaching is apparently not the main cause for the reduced U-concentrations.

The question is whether the specimen could be substantially older. Most older apparent age estimates are associated with relatively high Th and low U concentrations, either close to the outside or in pores (Tracks 1, 3, 4, 11, and 12). Here, the older apparent U-series results may well have been influenced by detrital material as well as U-leaching. Other older results (>15 ka) are seen for example in Track 8 (around cycle 3100), Track 11 (around cycle 430 and 550) and Track 12 (around cycle 1480), see arrows in the respective diagrams in Figure S2. In Track 12, the section between cycle 1481 and 1488 yields an age of 19.3±1.8 ka. The immediately adjacent sections yield 13.9±2.6 ka (cycles 1472 to 1481) and 10.8±2.1 ka (cycles 1488 to 1505). It is a simple statistical fact that some subsamples of a single population will deviate from the other results by more than 2-σ. The most likely apparent age result comes from the average of this section. The same seems to apply to the other marked sections.

Nevertheless, Tracks 9 (between cycles 3076 and 3161) and 11 (between cycles 726 and 791, see circles in the respective diagrams of Figure S2) show wider domains with relatively old apparent U-series ages (15.0±1.3 and 16.3±0.5 ka), which do not depend on the U-variations within these domains. While U-leaching cannot entirely be excluded in these sections, the lower U-concentrations may just as well be due to different amounts of internal surface area. These two sections are close to the lower surface and could have preserved the original U-series isotope signatures. If this is the case, the most likely age of the bone is around 16.3 ka. In summary, the minimum age of the bone is around 11.7±1.7 ka while some domains indicate an age as old as 16.3±0.5 ka.




During a recent examination of our morphological 3-D coordinate dataset, it was ascertained that it included a subadult specimen not listed in the publication. This specimen was not shown in the plots but mistakenly remained in the analysis as part of the early modern human sample (EAM). We thank Dr. Waddell for bringing this issue to our attention. We have re-analyzed our data excluding this specimen. Although some of the numerical results change slightly, the main findings, relationships among samples and our conclusions do not change. The PCA and CVA plots, as well as the minimum spanning tree, are nearly identical to those originally reported (Corrected Figure 3). The new Mahalanobis D2 results are similar in the values obtained and pattern of relationships among samples to the ones originally reported (Corrected SI Table 2). The main difference is the distance between Iwo Eleru and the early modern human Qafzeh-Skhul sample from the Levant (EAM), which dates to between 90-130 ka, in the distances corrected for unequal sample sizes. This group, and not the much later Zhoukoudian Upper Cave sample as originally reported, is the closest neighbor to Iwo Eleru in the corrected Mahalanobis D2. This result, although different from that originally reported, strengthens our conclusion that Iwo Eleru bears greater morphological similarity to early rather than late modern human samples from Africa and Eurasia. The revised dataset and the updated key for the labels used in the dataset are also provided with this Correction. Figure 3:


Discussion
Our analysis indicates that Iwo Eleru possesses neurocranial morphology intermediate in shape between archaic hominins (Neanderthals and Homo erectus) and modern humans. This morphology is outside the range of modern human variability in the PCA and CVA analyses, and is most similar to that shown by LPA individuals from Africa and the early anatomically modern specimens from Skhul and Qafzeh. Iwo Eleru is distinct from the recent African samples used here (although the range of recent modern human variation encompasses relatively low and elongated cranial shapes approaching this condition). Past work has suggested that neurocranial shape reflects population history relatively reliably among modern human populations [14], [15]. Although we did not find unambiguous strong affinities between Iwo Eleru and the samples used here, its overall morphological similarities with early modern humans suggest a link to these early populations and possibly a late Middle-early Late Pleistocene chronology. Nonetheless, the archaeological setting, stratigraphy, previous radiocarbon [see 4] and our new U-series dating indicate a much younger, terminal Pleistocene age for this cranium. Such a late chronology for the Iwo Eleru cranium implies that the transition to anatomical modernity in Africa was more complicated than previously thought, with late survival of “archaic” features and possibly deep population substructure in Africa during this time.

Thus our restudy of the Iwo Eleru cranium confirms previously noted archaic cranial shape aspects, and the U-series age estimates on its skeleton support the previously proposed terminal Pleistocene date for this burial. Our findings also support suggestions of deep population substructure in Africa and a complex evolutionary process for the origin of modern humans [16], [17], [7], [18], [19], [20], [21]. Perhaps most importantly, our analysis highlights the dearth of hominin finds from West Africa, and underscores our real lack of knowledge of human evolution in that region, as well as others. As also indicated by restudy of the Ishango (Congo) fossils [22], Later Stone Age fossils from at least two regions of Africa retain significant archaic aspects in their skeletons. We hope that the next stage of this research will extend studies to the Iwo Eleru mandible and postcrania, and to comparative materials such as those from Ishango.

Harvati K, Stringer C, Grün R, Aubert M, Allsworth-Jones P, Folorunso CA (2011) The Later Stone Age Calvaria from Iwo Eleru, Nigeria: Morphology and Chronology. PLoS ONE 6(9): e24024.

When modern humans started emerging from Africa and spreading throughout Eurasia, they found many places already occupied by older hominids such as Neanderthals and Denisovans. As humans do, we got rather friendly with our new neighbours: evidence of that hanky panky lives on in our DNA today.

After closely analysing the existing literature, Teixeira and his colleague biologist Alan Cooper have identified two such 'ghost' ancestors in modern DNA. The first, identified in Eurasian DNA with the help of artificial intelligence, was widely reported earlier this year.

The second, however, was reported last year, a detail that flew under the radar in a larger paper: a mysterious, and inconclusive, genetic signature exclusively found in the population of Flores, Indonesia. It appears to be as divergent from modern human DNA as Neanderthal or Denisovan DNA is.



As the modern humans moved farther east, across into islands of Southeast Asia, they seem to have run into more groups.

"At least three other archaic human groups appear to have occupied the area, and the ancestors of modern humans mixed with them before the archaic humans became extinct," Teixeira said.nOne of those groups was the Denisovans. The other two remain a mystery.

The first unknown extinct hominid - named EH1 - was roughly genetically equidistant from Denisovans and Neanderthals. The ancestor of all Asian and Australo-Papuan populations bred with EH1, resulting in 2.6 to 3.4 percent shared EH1 ancestry.



It's less strong now, but that genetic signal can still be detected in the DNA of Aboriginal Australians, East Asians and Andaman Islanders. This led the researchers to tentatively conclude that EH1 likely occupied a region in northern India, where a group of modern humans - the migration branch that went on to Asia, Australia and the Papuan islands - encountered them (1 on the map above).

Modern humans also seemed to have interbred with Denisovans in a number of locations, such as East Asia, the Sunda Shelf, and the Philippines (2, 3, and 4 on the map).

Evidence for EH2 - the extinct hominid that interbred with modern humans on Flores - is a little less clear. It only appears in short-statured people that live near Liang Bua Cave - where Homo floriensis was discovered. So it's highly localised, and has somehow remained contained for the roughly 50,000 years since the two groups met (5 on the map).