The recent retrieval of a complete mitochondrial (mt) DNA sequence from a 48–30 ka human bone from Denisova (Siberia) (Krause et al., 2010) is a remarkable achievement fully deserving international acclaim. Without wishing to detract from this feat, however, we wish to challenge their conclusion that the Denisova hominin “derives from a hominin migration out of Africa [ca. 1.0 Ma] distinct from that of the ancestors of Neanderthals and of modern humans” (Krause et al., 2010: 894). In addition, we challenge their assumption that the ancestors of the Neanderthals left Africa between 500–300 ka. In our view, alternative interpretations of the evidence are available and should be considered.Longer excerpts below:
Did the Denisova hominin derive from “a hominin migration out of Africa [ca. 1.0 Ma] distinct from that of the ancestors of Neanderthals and of modern humans” (Krause et al., 2010: 894)?
In support of Krause et al. (2010), there is evidence of hominin dispersals from Africa after the first documented appearance of Homo erectus ca. 1.75 Ma and before 0.5 Ma. Mammalian dispersals from Africa were much easier in the Early Pleistocene than during the Middle Pleistocene (see below) [. . .]
Although hominins could therefore have dispersed from Africa ca. 1.0 Ma, as suggested by Krause et al. (2010), we also need to bear in mind both the small number of relevant African finds from the late Early Pleistocene and the virtual absence of fossil skeletal evidence from the 5,500 miles of continental Eurasia (i.e., excluding Java) between Spain and China (a gap equivalent to the distance from London to Johannesburg). In Africa, the crania from Buia (0.9 Ma) and Daka (ca. 1 Ma) are the most relevant to understanding whatever might have been happening in Asia. Neither are straightforward examples of African H. erectus; [. . .] Neither specimen has obvious counterparts in Asia. Within Eurasia, the fossil hominin evidence from Southwest Asia – the cross-roads between Africa and Europe, but also the “black hole of palaeoanthropology” (Dennell, 2009:192) – between 1.5–0.5 Ma is limited to a few incisors from ‘Ubeidiya, one of which has been identified as H. erectus (Belmaker et al., 2002). The earliest European hominin evidence is the mandible from the Sima del Elefante (ca. 1.3 Ma) (Carbonell et al., 2008) and an assemblage of >150 specimens from the Gran Dolina TD6 level, ca. 0.8–1.0 Ma ([Falguères et al., 1999] and [Berger et al., 2008]) that is attributed to Homo antecessor ([Carbonell et al., 1995], [Carbonell et al., 2008] and [Bermúdez de Castro, 1997]). Putting aside discussions about its taxonomic distinctiveness (e.g., [Lahr and Foley, 1998], [Rightmire, 2008] and [Hublin, 2009]; but also see [Bermúdez de Castro, 1997], Bermúdez de Castro, 2003 J.M. Bermúdez de Castro, M. Martinón-Torres, S. Sarmiento and M. Lozano, Gran dolina-TD6 versus Sima de los Huesos dental samples from Atapuerca: evidence of discontinuity in the European Pleistocene population?, J. Archaeol. Sci. 30 (2003), pp. 1421–1428. Article | PDF (226 K) | View Record in Scopus | Cited By in Scopus (14)[Bermúdez de Castro, 2003], [Delson et al., 2000], [Mounier et al., 2009] and [Stringer, 2009]), morphologically, the European Early Pleistocene populations currently represented by the Sierra de Atapuerca hominins are closer to Early Pleistocene hominins from Asia than those from Africa ([Carbonell et al., 2005], [Carbonell et al., 2008] and [Martinón-Torres et al., 2007]). Archaeologically, the absence of any Acheulean artefacts at Atapuerca would make a Eurasian origin more likely, as hominins dispersing from the Levant after 1.4 Ma would be expected to retain an Acheulean technology. The Chinese evidence ca. 1 Ma is limited to the cranium from Lantian (Gongwangling), for which an age of 1.15 Ma is most commonly cited (An and Ho, 1989), and the two crania from Yunxian, which probably date to 0.8–1.0 Ma. Both have previously been assigned to H. erectus (Wu and Poirier, 1995). Work in progress indicates that the Gongwangling cranium is likely significantly older than 1.15 Ma and nearer in age to that of Dmanisi, and is thus not relevant here. With a cautionary note due to the high distortion of this fossil, the Yunxian crania may show affinities with Homo heidelbergensis ([Stringer, 2002] and [Etler, 2010]; see below), may or may not be derived directly from the population represented by the Gongwangling specimen, and may not be ancestral to the better known H. erectus population at Locality 1, Zhoukoudian.
Although hominins could therefore have dispersed from Africa ca. 1.0 Ma, as suggested by Krause et al. (2010), large-scale dispersals also occurred within Eurasia in the Early Pleistocene. [. . .] Under conditions of repeated fragmentation and recombination, part of the European Early Pleistocene population’s gene pool could have developed in isolation as a separate deme and persisted in time. The mid-Middle Pleistocene dates (0.45 +0.05/−0.10 Ma) for Ceprano (Muttoni et al., 2009), tentatively assigned to H. antecessor (Manzi et al., 2001) because of its primitive morphology and lack of Neanderthal traits (see also Manzi et al., 2010), could support this scenario. Within this frame, 1.0 Ma DNA from Denisova could be another example of lineage persistence in time. Interestingly, the Yunxian evidence has recently been cited as an “ideal candidate” for the ancestor of both the Denisova hominin and H. heidelbergensis in both Europe and Africa, and thus it may also imply a large-scale dispersal westwards after 1.0 Ma (Etler, 2010). Therefore, whilst a dispersal event out of Africa may have occurred ca. 1.0 Ma, dispersal events within Eurasia seem at least as likely. As there is no ancient DNA evidence from Eurasian populations of H. antecessor, H. heidelbergensis, and East Asian H. erectus, it is thus a massive assumption that the source of the Denisova mtDNA was necessarily African.
Did H. heidelbergensis leave Africa ca. 500–300 ka?
A popular scenario is that H. heidelbergensis originated in Africa; some groups left ca. 500 ka, dispersed into Europe and possibly India (Cameron et al., 2004), and thereby introduced an Acheulean bifacial technology into these regions (e.g., Klein, 2009). [. . .] There are several reasons for expressing caution over this scenario. First, archaeologically, there is no need to postulate an immediate African origin for the Acheulean in Europe or India ca. 500 ka, as an Acheulean hand-axe and cleaver tradition with clear African affinities is evidenced at GBY, Israel, ca. 780 ka. The Levant is a far more likely starting point for the Acheulean of Europe and India after 500–600 ka than an East African one. There is also no clear evidence that later Levallois technologies in Eurasia were African in origin (contra Foley and Lahr, 1997), and not independent, indigenous developments within those areas where Acheulean bifacial technologies were used (Tuffreau, 1995). Second, the ambiguities and limitations of the Middle Pleistocene hominin record preclude firm inferences about these events. [. . .] Because the absolute dating of key specimens attributed to H. heidelbergensis or Homo rhodesiensis is still problematic, the relative dating of specimens remains unclear: obvious examples are Kabwe and Sima de los Huesos. At present, it is not even unequivocally clear that the earliest example of H. heidelbergensis (if considered as an Afro-Eurasian lineage) is African rather than European or East Asian. [. . .] As examples of the range of current interpretations on offer, one can argue either that H. heidelbergensis originated in sub-Saharan Africa and dispersed as far as China ([Stringer, 1990], [Stringer, 2002] and [Rightmire, 2001]), or that it originated in China and eventually dispersed as far as sub-Saharan Africa (Etler, 2010)! Alternatively, it has been suggested that African specimens such as Bodo, Kabwe, and Ndutu should be classified as H. rhodesiensis ([McBrearty and Brooks, 2000] and [Hublin, 2001]), and seen as distinct from H. heidelbergensis, which is thus restricted to Eurasia. Consistent with that suggestion, a SW Asian origin of H. heidelbergensis has been proposed, possibly with some traits shared with H. antecessor ([Martinón-Torres et al., 2006], [Martinón-Torres et al., 2007], [Gómez-Robles et al., 2007], [Bermúdez de Castro, 2008], [Carbonell et al., 2008] and [Dennell, 2009]). Here, we are not advocating any one of these scenarios but aim only to point out that there are several alternatives to the one presented by Krause et al. (2010). Based on the differences in the genetic diversity of H. sapiens and H. neanderthalensis (e.g., [Orlando et al., 2006], [Green et al., 2010] and [Krause et al., 2010]), future genetic studies might contribute to this debate by analysing the theoretical genetic models behind “Africa into Eurasia” and “Eurasia into Africa” scenarios.
The third line of evidence requiring interpretive caution includes faunal data, which indicate that Africa and SW Asia were probably largely isolated by the Saharan-Arabian desert barrier between 500–300 ka. [. . .]
We suggest, therefore, that the evidence that the ancestors of Neanderthals dispersed out of Africa between 500–300 ka is, contra Krause et al. (2010), equivocal on chronometric and morphological grounds, and improbable on archaeological and faunal grounds. [. . .]
As outlined above, we urge caution over both the interpretative framework of Krause et al. (2010) and their main conclusion. The evidence that the ancestors of Neanderthals (i.e., H. heidelbergensis) left Africa ca. 500–300 ka is currently inconclusive, and the origin of H. heidelbergensis remains enigmatic. Whilst dispersals out of Africa might have occurred ca. 1.0 Ma, large-scale dispersals within Asia were also probable, and thus an Asian origin of the Denisovans cannot be excluded. These issues cannot be resolved without substantial improvements in the dating of key specimens, without an enlarged Asian fossil hominin record (particularly from SW Asia), and without a much more detailed Middle Pleistocene climatic record from SW Asia and NE Africa. Although the Denisova evidence is undoubtedly a fascinating piece in the jigsaw puzzle of human origins, it would be premature at present to determine the part of the picture to which it belongs.