JASs Proceeding Paper doi 10.4436/JASS.90018 Journal of Anthropological Sciences Vol. 90 (2012), pp. 99-116 On the trail of the genus Homo between archaic and derived morphologies* Giorgio Manzi SAPIENZA Università di Roma, Dipartimento di Biologia Ambientale, Piazzale Aldo Moro 5, 00185 Roma, Italy e-mail: [email protected] Summary – The topic of this review is the evolution of the genus Homo, focusing on evolutionary transitions that occurred during the Early and Middle Pleistocene. Two crucial issues are addressed in particular: 1) the emergence in the Early Pleistocene of the archaic variant of Homo that might represent the last common ancestor before the emergence of at least two (more probably three) geographically distinct trajectories; and (2) the evolution of these derived lineages, ultimately leading to the allopatric speciations of the most encephalised species of Homo: H. neanderthalensis and H. sapiens. In this framework, the time window between 1.0 million years ago (Ma) and 500 thousand years ago (ka) is of crucial importance, since it is probable that a new kind of humanity emerged in this period and then spread across a wide area encompassing Africa and Eurasia. These humans are represented by a number of specimens that are included within the single, polymorphic, and widespread species H. heidelbergensis. It is suggested that, in the course of the Middle Pleistocene, this species diversified in a number of incipient species – or subspecies – geographically and phenotypically distinct from one another. The case-study furnished by the calvarium found near Ceprano, in Italy, is of great interest in this regard, since it displays the least derived morphology seen among the hypodigm of H. heidelbergensis, and may represent better than other specimens the ancestral morphotype (i.e., the stem subspecies) of this taxon. Keywords - Hominin evolution, Genus Homo, Early Pleistocene, Middle Pleistocene, Lower Paleolithic, Cranial morphology, Homo heidelbergensis, Subspecies. Introduction taken place from the original spread of the genus Homo: a phenomenon known as “out-of-Africa During most of the 20th century, scholars 1”. The model of progressive change then required interpreting the evolution of the genus Homo the persistence of a single, widely-distributed, and shared a paradigm implying the existence of a polymorphic humanity throughout most of the single human species that progressively evolved Pleistocene. It thus assumed the existence of suc- over the entire Pleistocene. This is known as cessive evolutionary stages of an archaic species “the single-species hypothesis” (Weidenreich, – usually referred to as H. erectus – followed by 1947; Mayr, 1950), consistent with the gradual- “grades” of another, more derived species: i.e., H. ist perspective of the “modern synthesis” (Huxley, sapiens. In this perspective, H. erectus was regarded 1942). According to this model, the current vari- as the first human species to leave Africa, equipped ability of H. sapiens was considered to have been with a brain of about 1,000 ml and skilled enough the result of small changes occurring through time to produce the elaborate Acheulean or “Mode in local populations of a single and widespread 2” (Clark, 1968) Paleolithic stone tools. These species. The origin of this lineage would have were the quasi-modern aptitudes that made these * Lecture presented at the International Meeting “Many Times Homo”, Rome 2011. the JASs is published by the Istituto Italiano di Antropologia www.isita-org.com 100 Evolution of the genus Homo humans capable of facing a range of diverse envi- hominin fossil record implied for decades that ronmental conditions. H. erectus was a widespread species that included Despite regional morphs were identified a number of archaic-looking and (by contrast) among either the archaic or the derived spe- culturally quasi-modern fossil samples, which cies, the focus was on stages/grades instead of were considered ancestral in each region to cor- clades (Wood & Lonergan, 2008). Thus, assum- responding populations of H. sapiens. ing that H. erectus was present in Asia, Africa, However, more recently an increasing body and even Europe, each archaic regional variant of data suggests a different scenario (Fig. 1). was viewed as ultimately moving towards a dif- Above all, the evidence from the Georgian site of ferent “race” among the extant human diversity Dmanisi (Gabunia et al., 2002; Rightmire et al., (compare Coon, 1962). At the same time, the 2006; Rightmire & Lordkipanidze, 2009) dem- species H. sapiens included both recent popu- onstrates that hominins engaged in the first out- lations and extinct morphotypes, respectively of-Africa diffusion were not derived, encephalised, ascribed to a small number of subspecies, while and technologically advanced humans, but were the adoption of a trinominal nomenclature gave definitively more archaic than previously believed, rise to nomina such as H. sapiens sapiens and H. with a brain just above the threshold of 500-600 sapiens neanderthalensis (Mayr, 1950; Campbell, ml (maximum encephalic volume of the australo- 1965). Another example might be some Middle pithecines) and a morphology close to the chang- Pleistocene specimens found in Africa (e.g., ing definition and hypodigm of H. habilis (after Kabwe) and Eurasia (e.g., Petralona, Dali) Leakey et al., 1964). Driven by ecological, rather lumped within H. sapiens and often referred to as than by behavioural or ”cultural” motives, these “archaic H. sapiens”, according to such a “gradis- hominins had a tendency to diffuse and to adapt tic” scheme (as in Stringer et al., 1979). to variable non-tropical environments. The theory called “multiregional evolution”, Under this new approach, H. erectus is now introduced by M.H. Wolpoff and colleagues in generally viewed as an Eastern Asian species of the early ‘80s (e.g., Wolpoff et al., 1984), may Homo (contra Asfaw et al., 2002; see e.g. Manzi be considered as a recent version of this view. et al., 2003), distributed in the island of Java Based on the observation that a certain degree of and in Northern China, whereas its African “regional continuity” characterizes the morpholo- counterparts are considered as a distinct species gies of archaic and modern populations within and referred to as H. ergaster (Groves & Mazák, each geographical area (Thorne & Wolpoff, 1981; 1975; Wood, 1991). Alternatively, these two Wolpoff et al., 1984, 2001; Frayer et al., 1981), geographical variants are grouped together as H. this model suggests that the anatomically modern erectus sensu lato, while H. erectus sensu stricto humans of Africa, Asia and Europe emerged from would be the Asian deme of this a multiregional archaic populations that existed in each respective taxon. The same corpus of data suggests also that region. At the same time, during their evolution the earliest out-of-Africa dispersal would have throughout great part of the Pleistocene, all these started well before the appearance of either H. populations were kept as single species by a signifi- erectus sensu stricto, or the Acheulean (which are cant amount and pervasiveness of gene flow. now regarded as geographically distinct and thus independent phenomena), that is between more than 2,0 and 1,6 Ma. A tentative updated scenario At the same time, other species have been named and/or old nomina have been reconsid- Out of Africa 1 ered. These include (ordered according to the Therefore, according to the evolution- respective original denominations): H. heidelber- ary paradigm of the ‘40s (Weidenreich, 1947; gensis (1908), H. rhodesiensis (1921), H. pekinen- Mayr, 1950), even popular interpretations of the sis (1927), H. soloensis (1932), H. helmei (1935), G. Manzi 101 H. mauritanicus (1954), H. leakeyi (1963), H. Mediterranean and continental regions of Europe rudolfensis (1986), H. antecessor(1997), H. geor- (Hovers & Braun, 2009). The earliest appear- gicus (2002), H. cepranensis (2003), H. floresiensis ance of any Paleolithic in Europe is recorded (2004). However, many of these taxa are debated by recent findings in Spain (Toro-Moyano et and/or are not widely acknowledged. In fact, the al., 2003), in Southern France (Crochet et al., most robust scenario is far less speciose than this 2009), and also in Italy, where hominin pres- listing suggests. The identification of a num- ence is suggested by the stone tools found at ber of different species within the genus Homo Pirro Nord, near Apricena in Puglia (Arzarello probably implies an overestimate of interspecific et al., 2007, 2009), in association with a rich diversity, whereas in many cases this diversity was paleontological assemblage biostratigraphically probably intraspecific (and should be referred to referred to a Faunal Unit of the Early Pleistocene the rank of subspecies; see below). Nevertheless (about 1,4 Ma) (Sardella et al., 1998). Also of (according, e.g., to the seminal paper by interest in Italy are sites dated to about 1,0 Ma Tattersall, 1986), this speciose scenario confers such as Monte Poggiolo (Peretto et al., 1998) or a more reasonable significance to human varie- a number of localities in the Ceprano basin and ties that were formerly disguised as variants of H. surroundings that have recently been subject to erectus (sensu lato), or as members of the infor- new excavations and analyses (Segre & Biddittu, mal and confusing entity until recently known as 2009). We may add that, between 950 and 700 “archaic H. sapiens”; and this latter designation is ka, these humans were capable to adapt to higher now thankfully disappearing from the literature. latitudes, as demonstrated by sites in Southern England (Parfitt et al., 2010). Nevertheless, it Across Europe is worth noting that these incursions would Two distinct waves of immigrants In Europe have been strongly influenced by ecological seem to be recognizable at present, respectively in conditions and, thus that hominin presence in the late Early (until 780 ka) and in the Middle Western Europe was presumably discontinuous Pleistocene (after 780 ka). The fossil record until about 600 ka (Agustí et al., 2009), perhaps documenting the former one of these disper- in relationship with distinct waves of diffusion. sals is referred – at least in part – to H. antecessor The major discontinuity was represented (Bermúdez de Castro et al., 1997). Possible rep- by Marine Isotopic Stage (MIS) 16 (see discus- resentatives of this species have been discovered sion in Manzi et al., 2011), which was one of only in two sites of the Sierra de Atapuerca, near the most extreme glaciations of the last million Burgos in Spain (see Fig. 2): in the layer TE9 years, with an ice sheet extension below 50° lati- of the Sima del Elefante, dated to 1,2-1,3 Ma tude in Eastern Europe (Helmke et al., 2003). (Carbonell et al., 2008), and in the layer TD6 MIS 16 represented a possible cause of extinc- of the Gran Dolina, dated to more than 780 ka tion of the earliest Europeans: it is reasonable, in (Carbonell et al., 1995; Bermúdez de Castro et al., fact, that this climatic collapse created a strong 1997). However, a detailed analysis of the mor- environmental barrier along time, and was prob- phology of the partial mandible and teeth from ably associated with a population discontinuity Atapuerca TE9 (Bermúdez de Castro et al., 2011) at the species level. It might not be by chance critically reconsidered the putative attribution to that MIS 16 preceded both the appearance H. antecessor of this “earliest European”, which is of the Acheulean (Mode 2) technology in the not included so far in any named taxon. continent, and the occurrence of a more abun- The fossil record from Atapuerca bracketed dant fossil record with a diverse morphological between 1,3 Ma and 780 ka is in association appearance if compared to H. antecessor and/or with the Lower Paleolithic so-called Oldowan to the sample from Dmanisi (see Fig. 2). or “Mode 1” (Clark, 1968). At the same time, This second dispersal into Europe probably Mode 1 assemblages are widely distributed in occurred after 700-600 ka and was related to www.isita-org.com 102 Evolution of the genus Homo s o y ed lineare alstremel shs ex ce; daspeciee the videner of wher hard fossil emited numbopic stages, nes = en a lie isot d liwerin oletma Homo (bnships bsed on nus atios ba g to the gegenetic relc variation of the fossil record pertainines of diffusion and/or phylod (below) a curve of climati8 and MIS 16 is evidenced. ative chronology and topology ominin occurrence); trajectoriotted lines). It is also reporteriod bracketed between MIS 1 thde Fig. 1 - Ten= inferred indicated (cold/arid p G. Manzi 103 morphologically-derived hominids with clear inappropriate to be also ancestral to the African signs of further encephalization, that are well emergence of H. sapiens. Nevertheless, there are known from a number of sites. The most nota- in Europe other fossil specimens – such as the ble such assemblage of fossil material is again calvarium from Ceprano in Italy (see Fig. 2), but in the Sierra de Atapuerca (e.g., Arsuaga et al., not only – that are penecontemporaneous with 1997), namely at the site with the evocative the material from Atapuerca SH, and might rep- name of Sima de los Huesos (SH): the extraor- resent (far better than the Spanish sample) a pos- dinarily rich and well preserved human sam- sible ancestral morphotype of H. heidelbergensis ple, which has been pushed back to about 600 (see below), supporting the alternative hypothe- ka (Bischoff et al., 2007). These and a number sis that this species – taken as a whole, while join- of other European fossil samples of the Middle ing African and Eurasian fossil samples – might Pleistocene (see below) may be regarded as falling represent the evolutionary stem leading to the within the polymorphic variability of H. heidel- divergence between Neanderthals and modern bergensis (Rightmire, 1998; Mounier et al., 2009, humans (compare, e.g., Endicott et al., 2010). 2011), a species described in Africa and Eurasia that is frequently associated with Mode 2 lower Back to Africa Paleolithic assemblages. African hominins dated to around 1,0 Ma – H. antecessor and H. heidelbergensis are regarded i.e., specimens from sites such as Bouri (Daka), as competing with each other for the same phylo- Buia, and Olorgesailie (Fig. 2) – share morpho- genetic position, as seen in alternative proposed logical affinities with H. ergaster, as pointed out scenarios of human evolution. H. antecessor is by Manzi and colleagues (2003), among others. claimed by the Spanish workers (after Bermúdez In this perspective, these crania of the late Early de Castro et al., 1997) as the stem-species that was Pleistocene are distinct from the African homi- ancestral to the evolutionary divergence between nins of the Middle Pleistocene – like Bodo, and the evolutionary lineage of the Neanderthals in Kabwe – that are referred to H. heidelbergensis Europe and to the origin of our species in sub- (or, alternatively, to H. rhodesiensis). Saharan Africa. Nonetheless, H. heidelbergensis has This observation suggests a taxonomic and also been considered as pertinent in the same cru- phylogenetic discontinuity that ranges across the cial role (Stringer, 1983; Rightmire, 1996), and Matuyama-Brunhes boundary (780 ka). Such a this interpretation appears to me more robust at phenetic distance between H. ergaster (until 900 present (e.g., Manzi & Di Vincenzo, 2012). ka) and H. heidelbergensis (after 600 ka) sup- In a sense, the Sierra de Atapuerca contains ports, here as in Europe, a distinction at the spe- evidence that could resolve this ambivalence, cies level. Might it also signal an allochthonous since this small karstic hill in Northern Spain (non-African) origin of the Mid-Pleistocene includes sites with samples of either H. antecessor taxon? Unfortunately, in sub-Saharan Africa or H. heidelbergensis. It should be remarked that and not only there the period between 900 and the material from Atapuerca SH (H. cf. heidel- 600 ka is very poor in fossil evidence. In addi- bergensis) is clearly Neanderthal-like, being char- tion, representatives of H. ergaster are distinct acterized by a number of features that, later in from Mid-Pleistocene fossils either from Europe the Pleistocene, would typify the Neanderthals – including both the Italian calvarium from (Santa Luca, 1978; Dean et al., 1998). From Ceprano (contra Mallegni et al., 2003) and, even the perspective of published views on Atapuerca more clearly, the assemblage from Atapuerca SH SH (Arsuaga et al., 1997; Rak et al., 2011), – or mainland Asia, as represented by specimens H. heidelbergensis would have the identity of a such as Narmada, Dali, and Jinniushan. These European regional chronospecies antecedent to Eurasian samples are in turn also distinguishable H. neanderthalensis or would be even part of this from H. antecessor in Europe, and from H. erectus taxon. Thus, H. heidelbergensis would emerge as sensu stricto in the Far East. www.isita-org.com 104 Evolution of the genus Homo Fig. 2 - Geographic distribution of sites with fossil hominins mentioned in the text. Therefore, although we still do not know Denisova where and exactly when, it seems that something A possible answer about the last common happened between about 900 ka and 600 ka that ancestor of Neanderthal and modern humans generated a new and more encephalised kind of comes from the complete mitochondrial DNA humanity spreading quite rapidly in Africa and (mtDNA) sequence extracted from a human Eurasia, which may be referred to as the single and phalanx found at the Denisova cave in the Altai widely-diffused species H. heidelbergensis. These mountains, Southern Siberia (Krause et al., “new” humans appear clearly different from the 2010). In the context of episodic occupations of variable assemblage that derived from the earliest this site in the Late Pleistocene, the layer where dispersal out of Africa, including H. ergaster, H. the phalanx was found has been dated to 48-30 antecessor, H. erectus and also, probably, the sin- ka, and contains archaeological assemblages gular diminutive form referred to as H. floresien- including both Middle and Upper Palaeolithic. sis (Brown et al., 2004). At present, however, the By contrast, the mtDNA surprisingly points to chronology, topology, and phylogenetic dynam- humans that were different from both H. nean- ics related to the rather synchronous appearance derthalensis and H. sapiens, but that shared with of Middle Pleistocene humans that we may refer them a common ancestor at less than 1,0 Ma to H. heidelbergensis are still unclear. As a matter (Krause et al., 2010). of fact, we do not know the provenance of the As a working hypothesis, this suggests that the archaic humans that spread geographically and Denisova phalanx may represent a still unknown were ancestral both to Neanderthals and H. sapi- hominin species that originated before the begin- ens (Rightmire, 2008; Hublin, 2009). ning of the Middle Pleistocene, interestingly just G. Manzi 105 before the appearance of H. heidelbergensis in the more complex models could explain the data”. fossil record. As a matter of fact, in my view, the occurrence of Researchers have opted to wait for their gene flow across Eurasia, between the ancestors data to provide a clearer picture before giving of both Neanderthals and the Denisovan homi- this largely hypothetical species a formal name. nins, could be a better explanation of their affini- Nevertheless, it is already possible to speculate ties in nuclear DNA (compare Fig. 5). that the Denisova hominins was in relation with Consistently, the observed genetic diversity a “non-erectus” occupation of mainland Asia dur- between Neanderthals and modern humans ing the Middle Pleistocene. In fact, assuming coalesced at around 500 ka (Green et al., 2008; that H. erectus was one of the evolutionary out- Briggs et al., 2009; Endicott et al., 2010), sub- comes of the earliest diffusion in Eurasia – and stantiating previous conclusions based on mor- therefore excluding this taxon from the scope of phology and palaeogeography. These suggest possibilities – we need to look to other humans isolation and divergence between the European that were in Asia during the Middle Pleistocene. and African lineages during the Middle and the We should thus focus on specimens, such as early Late Pleistocene (after Santa Luca, 1978). Dali and Jinniushan, that in the past have been Indeed, looking at the hypodigm of H. heidelber- ascribed to H. sapiens daliensis (Wu, 1981) and gensis as a whole (compare Fig. 4) it is clear that are currently considered by various authors as a considerable amount of variability characterises representatives of the Easternmost populations this species (Mounier, 2009). In other words, of H. heidelbergensis (after Rightmire, 1996, we should conclude that populations of H. hei- 1998). Which raises the question of whether or delbergensis bore regional features (in Africa, not it would be appropriate to ascribe the pha- Asia, and Europe respectively) that are referable lanx from Denisova, and its precious molecular to a phenomenon known as “isolation by dis- content, to a late variant of the same taxon that tance” (Wright, 1943). At the same time, con- Dali, Jinniushan, and other “non-erectus” speci- siderable phenotypic variation has to be noted mens from the late Middle Pleistocene belong to. even within the European fossil record of the Further analyses on the Denisova material – Middle Pleistocene, recently greatly expanded including exceptionally preserved nuclear DNA by the revised chronology of the calvarium from from the phalanx and the discovery of an upper Ceprano in Italy (Manzi et al., 2010). molar – led the same group of researchers to pub- lish additional data (Reich et al., 2010). These new data, based on sequences of nuclear DNA The case-study of Ceprano and the morphology of the tooth, confirmed that the Denisova individuals exhibit molecular and A cranium for the earliest Europeans? dental features that appear “archaic”. In contrast, At the beginning of the 1990s, a “short chro- the picture that emerges from the analysis of the nology” for the earliest inhabitants of Europe nuclear genome suggests that this human group was introduced (Roebroeks & van Kolfschoten, has affinities with the Neanderthals, closer than 1994), suggesting that no humans were present those expected from the mtDNA. Surprisingly, in Europe before approximately 500 ka. This the scenario that has been suggested according date was claimed consistent with the chronology to these new data places the Denisova hominins of both the oldest human fossil record found in as a sister group of the Neanderthals, “with a Europe since that time (e.g., Roberts et al., 1994), population divergence time of one-half to two- and the earliest documentation of Acheulean thirds of the time to the common ancestor of assemblages discovered in various part of the Neanderthals and modern humans” (Reich et continent (e.g., Piperno et al., 1998). In 1994, al., 2010, p. 1057). However, Reich and col- however, fossils that were considered older than leagues (2010, p. 1057) admit also that “other, 700-800 ka, from both Italy (Ceprano, March www.isita-org.com 106 Evolution of the genus Homo Fig. 3 - Drawing of the Ceprano calvarium compared to a suggestive representation of the human- kind that the Italian fossil might represent (drawings by Maurizio Mei and Carlo Ranzi respectively). 1994; Ascenzi et al., 1996) and Spain (Atapuerca geo-palaeontological data that were available at a TD6, July 1994; Carbonell et al., 1995), falsified micro-regional scale. This stratigraphy describes this hypothesis. Thus, for more than a decade the two main complexes: the layer where the human Ceprano calvarium was considered part of the calvarium was found belongs to the lower por- fossil evidence documenting the human presence tion of the upper stratigraphic complex, indicat- in Europe before 500 ka. ing to Segre a tentative age of about 800-900 ka The Italian specimen (Fig. 3) was discov- (Ascenzi et al., 1996, 2000) and the archaic fea- ered in several fragments in a field known as tures of the calvarium were considered consistent Campogrande, near the town of Ceprano, in with the various Mode 1 techno-complexes from Southern Lazio, less than 100 km South-East of sites scattered in the Ceprano basin (Segre & Rome. Its discovery represented the result of sys- Biddittu, 2009), despite the fact that a number tematic field activities conducted for decades in of Acheulean assemblages are also well known at Southern Lazio by the Italian Institute of Human Campogrande and in its surroundings. Palaeontology, and particularly by I Biddittu. On March 13th 1994, during a survey along The muddle in the middle the trench excavated for a new road, Biddittu In this context, a project of surveys and exca- found a great number of fragments of a single vations started in 2001 under the direction of cranium in the section created by the excavators. I. Biddittu and myself (with the licence of the All the fragments were then carefully extracted Soprintendenza Archeologica del Lazio), with a and sieved from the in situ clayey sediments. The threefold aim: 1) better comprehension of the reconstruction of the cranium from these pieces Pleistocene stratigraphy of the Ceprano basin; required the coordinated efforts of a various 2) definite validation of the available geo-chron- experts and, overall, about five years (Ascenzi et ological model (Ascenzi et al., 1996, 2000); 3) al., 1996, 2000; Clarke, 2000). improvement of the palaeontological and archae- At the same time, for the purposes of chrono- ological records. logical reference, the geologist A.G. Segre sug- Results obtained through a multidiscipli- gested a synthetic stratigraphic column based on nary approach – including stratigraphic and G. Manzi 107 palynological data, combined with sedimen- On the whole, these researches largely sup- tology, geochemistry, soil-micromorphology, ported the following conclusions (see, e.g., taphonomy, and the archaeological evidence Manzi et al., 2001). First of all, though some – showed that the Ceprano calvarium is actu- metric and architectural features of Ceprano ally more recent than was previously believed, approach those shared by fossils referred to H. pointing to a time range close to about 400 ka erectus, the variability of this taxon is unfavour- and, more precisely, to the interval at the begin- ably enlarged when Ceprano is added. Second, ning of MIS 11 bracketed between 430 and 385 there are discrete features that detach Ceprano ka. This chronology is also consistent with the from its claimed “erectus-like” appearance and normal geomagnetic polarity recorded in the may be viewed as derived, suggesting a connec- area of discovery down to a depth of about 50 tion with the Middle Pleistocene fossil record metres (Muttoni et al., 2009). These unexpected from Africa and Europe. Third, Ceprano does results, and the consequent new chronology of not display any Neanderthal traits, while it shows the fossil specimen in the midpart of the Middle some affinities with the African penecontem- Pleistocene, led us to conclude that « the mor- poraneous fossil record that are closer than the phology of the human calvarium from Ceprano - affinities it has with its European counterparts. which lacks Neanderthal traits and does not have A possible conclusion is that Ceprano represents a real counterpart among the continental pene- a mosaic morphological link between the clade contemporaneous fossil record - appears now composed by the group of species referred to as tantalizing», pointing out to «more complex sce- H. erectus sensu lato, on one hand, and samples narios of human evolution in Europe than pre- commonly referred to as H. heidelbergensis on the viously believed, involving either the occurrence other. This in turn suggests that Ceprano might of a considerable intraspecific diversity (with document «the occurrence of an ancestral stock archaeologically distinct settlements) or, alterna- of Homo heidelbergensis/rhodesiensis» (Bruner & tively, the co-existence of different lineages (with Manzi, 2007, p. 365), the cranial morphology of their own respective archaeological traditions) which was partially lost along the subsequent tra- during part of the Middle Pleistocene» (Manzi et jectory of human evolution in Europe, but that al., 2010, p. 584). was preserved elsewhere (including Africa and, This also called for a taxonomic re-evalua- possibly, mainland Asia). tion of the Italian specimen. Originally, Ceprano Mounier and colleagues (2011) recently was attributed to “late H. erectus” (Ascenzi et reconsidered the morphology of the Ceprano al., 1996; Clarke, 2000). Subsequently, two calvarium in a wide comparative framework, papers criticized the H. erectus affinities origi- including the fossil record pertaining to H. heidel- nally claimed (Ascenzi et al., 2000; Manzi et al., bergensis and related species. This new extended 2001), arguing that less than two-thirds of the analysis, which combines geometric morpho- character states were in accord (and not always metrics with the evaluation of discrete features, unequivocally) with those commonly encoun- provides robust and independent corroboration tered in H. erectus sensu stricto, while others to the previous hypothesis and supports the attri- appeared peculiar or progressive. Further studies bution of Ceprano to an archaic variety of H. hei- included a cladistic approach, with the question- delbergensis (see also Manzi, 2011; Manzi & Di able proposal of a new species name (Mallegni Vincenzo, 2012). In conclusion, despite its rela- et al., 2003), whereas the CT scanning of the tively recent age in the Middle Pleistocene, the specimen (Bruner & Manzi, 2005) and other Italian specimen may represent the morphology phenetic data (Manzi et al., 2003; Manzi, 2004; of the as yet undiscovered ancestral stock of the Bruner & Manzi, 2007) produced additional ele- human variability represented in the fossil record ments that were useful to better understand the of the second part of the Middle Pleistocene specimen in a comparative framework. from Africa and Eurasia (Fig. 4). www.isita-org.com 108 Evolution of the genus Homo Fig. 4 - Examples of variability in cranial morphology during the second part of the Middle Pleistocene, i.e. roughly between 500 and 150 ka: a – Ceprano (Italy, Europe); b – Atapuerca SH 5 (Spain, Europe); c – Dali (China, Asia); d – Kabwe (Zambia, Africa). Scale bar = 5 cm. Towards recent Homo emerged from Africa (Asfaw et al., 1992; Krause et al., 2010) or from the Near East (Dennel et al., Neanderthals and modern humans 2011; Bermúdez de Castro & Martinon-Torres, The Ceprano calvarium is thus part of 2012). We also already stressed in this paper that Middle Pleistocene hominins scattered in Europe these Acheulean-bearing humans exhibit a clear from northern latitudes (e.g., Swanscombe in discontinuity in morphology with more archaic England; Mauer, Bilzinsgleben, and Steinheim humans from Europe (i.e., H. antecessor) and in Germany; etc.) to the Mediterranean regions elsewhere (i.e., H. ergaster and H. erectus). (including the large sample from Atapuerca SH Subsequent hominin evolution in Europe in Spain; Arago in Southern France; Petralona in during the Middle Pleistocene agrees with a per- Greece; Venosa and Visogliano in Italy; etc.). As sistent geographic isolation of human popula- we have seen, this a relatively rich fossil record tions to the north of the Mediterranean: a sce- is considered by many authors to be part of nario that is supported both morphologically the hypodigm of the species H. heidelbergensis and genetically (Stringer, 1974; Santa Luca, (Rightmire, 1998; Mounier et al., 2009), named 1978; Green et al., 2008, 2010; Weaver et al., for the European Mauer mandible (Schoetensack, 2008). In sync with this pattern of evolution, the 1908). In addition, data cited in this review sug- so-called “accretion model” (Dean et al., 1998; gest that this Mid-Pleistocene fossil record repre- Hublin, 2000, 2009) proposes that fossil samples sents the dispersal in Western Eurasia of homi- are characterised by an increase of Neanderthal nins bearing the Mode 2 techno-complexes. features, in possible relationship with the dra- Their exact place of origin is still not clear, matic climatic swings bracketed between MIS though it may be assumed that they ultimately 16 and MIS 2. These glaciations might have