Blog científico

Tenemos muchas cosas para contarte

Blog científico

Publicado el 24 de October de 2013 por María Martinón


The cover picture of Science on 18 October 2013 was so spectacular and eloquent that it needed no title or subtitle. Probably, the journal’s editors and graphic designers understood that the ferocious combination of a skull of tiny capacity with a huge face and jaw had sufficient aesthetic impact for anyone, even the most ignorant on palaeontological sciences, to realise that something was boiling inside.


Figure 1. Cover of the journal Science At last, the embargo was lifted – not just on the publication, but on the eight years during which many of us had to enjoy in clandestinity, like a lover, the existence of a man that was different from any other man we had met before. Skull number 5 from Dmanisi (labelled D4500), which fitted a jaw found at the same site over ten years ago (D2600), provided the scientific community (and humankind at large) with one of the oldest and most complete hominin heads known, with some 1.8 million years of age. The way in which the authors of the Science paper analysed and interpreted such a discovery reignited a debate that, while not new, seemed to have been forgotten.


Figure 2. David Lordikipanidze kneeling on the basalt flow that seals the sediment where skull 5 begins to peek out, and María Martinón-Torres, comparing in situ the face of the new arrival to casts of other hominid relatives. Dmanisi has always been the precious pebble in the shoe, a golden pea under the matress where paradigms about human evolution rest. It forced us to review the dates when we thought hominids first left Africa, and the circumstances in which they achieved this feat. Challenging the general belief that we had only left home when we had a considerable physical and intellectual capacity, the Dmanisi men had ventured into Eurasia wielding rudimentary stone tools and a brain that was hardly bigger than that of a chimpanzee. Now Dmanisi shakes us again because, contrary to what usually happens when a new find is published, the picture of human evolution seems to get simpler rather than more complex. Without a doubt, the new skull increases the variability of the sample so far found at the Georgian site. The significant differences between D4500 and the four skulls and/or faces recovered previously are consonant with the differences previously documented between D2600 (the jaw that fits the new skull) and the other two jaws (D211 and D2735).


Figure 3. Drawing by Eduardo Saiz of the front view of the jaw D2600, which fits the new skull from Dmanisi. In 2008, in a widescale comparative study including hominids and other close primates published in the Journal of Human Evolution, Skinner and colleagues argued that the dissimilarity between D2600 on the one hand, and D211 and D2735 on the other, could not be explained by sexual dimorphism  (the differences that would exist between male and female individuals of the same species), unless these humans had a degree of sexual dimorphism comparable to that of gorillas and never recorded for the Homo genus. The latter is of course a possibility, albeit not an insignificant one (nor a parsimonious one). If proved true, it would compel us – as the authors of the Science paper suggest – to reconsider our interpretation of the fossil record for early Homo. (At this point I would like to highlight José María Bermúdez de Castro’s point that, in this hypothetical case of sexual dimorphism, paradoxally the larger individual, “the male”, would have a much lower craneal capacity than the supposed females).


According to Lordkipanidze and his team, the large variability exhibited by the Dmanisi hominids would dwarf the differences so far used to discriminate species such as H. habilis, H. rudolfensis, H. ergaster or H. erectus. All of these would thus belong to the same species, representing regional variants or “races” of a single lineage that would have inhabited virtually the whole of the Eurasian and African continents over hundreds of thousands years (even millions if we cover the time span from H. habilis to H. ergaster). In order to arrive at this conclusion – let’s emphasise this -, these authors depart from the premise that all five individuals recovered at Dmanisi belong to the same population and, hence, to the same species.   Figure 4. Lateral view of jaw D2600 On the dilemma of christening fossils How to identify a species in the fossil record is a problem that will not be solved conclusively in the near future. Species are generally discriminated by their behaviour and morphology. However, when dealing with fossils, in most cases we can simply make taxonomic assessments based on morphology only, given that behaviour as such does not fossilise. Even so, and as discussed below, there are certain aspects of behaviour and biology that we should not disregard in our quest to identify species. A key issue is that, in general, we predict that individuals belonging to the same species will be similar among them but rather different from members of other species – however, nature provides countless exceptions to this seemingly logical assumption.


From a biological standpoint, a species is a group of populations whose members can interbreed and produce fertile offspring, but who cannot do so (or do not do so under normal circumstances) with members of another species. From a genetic and evolutionary perspective, species are groups of organisms who are reproductively homogeneous, in a given time and space. These groups will undergo changes through time and space diversification, which may culminate with the emergence of a new, different species  - a process known as speciation. To cap it all, it is precisely palaeogenetics that has messed it all up and now tells us that well-recognised species in the fossil record, such as H. sapiens and H. neanderthalensis, actually did  interbreed. Furthermore, even though the morphology or external appearance is the best tool available for the study of fossils, this method is not foolproof. There are species whose skeletons are apparently identical (especially if we “take off” hair and soft tissue, as is the case with fossils), but who are however reproductively isolated: they do not interbreed and they would not have fertile offspring (for example, some lemur species). There are also unrelated organisms that adapt to the same circumstances in the same way, and end up looking very similar (convergent evolution) – even if they lack a recent common ancestor. At the other end of the spectrum, there are cases of specimens who may have clearly different appearances while belonging to the same lineage (think of races of dogs). This means that both splitters, who tend to divide the fossil record in various groups, and lumpers, who group them together fewer categories, have sufficient arguments to defend their positions.


The practice of naming new species and questioning those already named pertains all living creatures, including plants and animals alike. It is true, however, that when the human record is concerned, naming a new species invariably brings about some controversy. As Publio Terencio Afer famously said: “I am human: Nothing human can be alien to me”. Hence, when it comes to naming or classifying humans, all of us feel immediately alluded. Sadly, these controversies often cast a shadow on the precise and methodical work behind the analysis and classification of the living world. This may convey the wrong impression that naming or refuting species is a matter of personal preference or even whim. Let us now focus on the conclusions of Lordkipanidze and his colleagues. From a biological perspective, I find it difficult to envisage a human panspecies adapted or pre-adapted to the extremely diverse and changing ecosystems of Africa and Asia over hundreds of thousands, possibly millions of years. In my opinion, the severe climatic fluctuations, geographic barriers (mountains, deserts, oceans) and other environmental constraints, should have conditioned human dispersal and settlement, also constraining the possibility of a substantial genetical interbreed – a necessary prerequisite to maintain species homogeneity for so long. We should analyse the biology of our species in a way no different from that of other animals – be them crows, bears, felines or humans. Some may argue that our own species, Homo sapiens, is a single species and yet it is indeed adapted to practically all coordinates of the globe. I would remind these that our common life as a species is barely 200,000 years old at most (as opposed to the almost 2 million year span covered by all the ancestral Homo). More importantly, I would stress that the adaptation of Homo sapiens is essentially a technological one, not an anatomical one – to the extent that the sick or the disabled may be “the strongest” and outlive those born physically superior. In other words, the selective pressure that the environment exerts on our anatomy is no longer as strong as it used to be. Moreover, one should not forget that a hypersocial species such as ours has practically eradicated a crucial element necessary to allow for speciation: isolation. Owing to sophisticated technologies that have in social media a great example, and thanks to a superpopulation without precedents in the animal kingdom, Homo sapiens is never alone or completely isolated. Furthermore, when we speak of inter-group isolation, we mean an isolation that needs not be exclusively physical or geographic (although I belive this must have played an important role in the early stages of the Pleistocene). Instead, as documented in thousands of current animal species, isolation is fundamentally ecological (preference for a particular habitat) or ethological (for example the difference between diurnal and nocturnal species, which may prevent interbreeding).


In these cases, species are effectively isolated from each other by  barriers that are hard to see in the fossil record. A single species spanning hundreds of thousands or millions of years and square kilometers, avoiding – as if in a bubble – natural selection, genetic drift or founder effect, for example, does not seem to me the most parsimonious option. Especially so when we look at speciation in other animals. In my view, for an evolutionary stasis on this scale, one of two conditions would have had to be met. One possibility would be that part of the ecosystem remained relatively stable and unchanged during this whole period, thus not requiring the need of any change or adaptation. Against this possibility, there is no need to reiterate the environmental, geographic and climatic differences in Africa and Asia (I can even think of many members of our species who I would struggle to envision surviving in other countries, other cultures, other climates – or even other food). Another option allowing such a protracted human homogeneity would be a relatively continuous and regular genetic exchange, which might “compensate” for the variability among individuals in that group. This scenario seems improbable if we consider the low population density of these hominid populations. By this I do not mean that it must have been necessarily either one option or the other. However, we are entering an arena where hypotheses are very difficult to test. Whatever the case, when information about the biology or culture of a given group is available in addition to their morphology, we should not overlook it.   Same place, same time? Having presented some general aspects about biology and speciation, I would like to offer a more personal view of the Dmanisi fossils and the recently published Science paper. This perspective actually is the outcome of joint work by our team during the last ten years, within a collaboration project with Georgia supported by the Duques de Soria Foundation, which has allowed for an excellent research and knowledge exchange relationship with the research team at the National Museum of Georgia. Even if irrelevant (or perhaps not) this collaboration has also led to strong frienships with the inhabitants of this country that, curiously, is also known as Iberia. I would also like to mention, not without pride, that one of the co-authors of the paper in question, Ani Margvelashvili (also the lead author of another paper on D2600 recently published in PNAS and mentioned below), spent some time with our team, and I had the pleasure of supervising her Masters thesis precisely on the teeth from Dmanisi.


To get back to the subject, with this work we are not trying to claim that we are right and those who disagree are wrong. However, we would like to provide first hand data to this discussion and explain why we are not convinced by the conclusions of the article on the new skull 5. Personally, I wish the authors had tested what I deem as the prioritary hypothesis in this case: considering the great variability shown by the Dmanisi collection (accepted by both splitters and lumpers), they should consider the possibility that more than one species is represented in their sample. Lordkipanidze and his team lump all the early forms of Homo in a single lineage because they find that variability among all the analysed skulls of H. habilis, H. ergaster and H. erectus is not larger than that documented in the Georgian skulls, which, they assume, belong to a single population. If all the Dmanisi skulls belong to a single population, they argue, and their variability is equal to, or larger than, that found for all the early Homo specimens, then there is no reason to refute that the latter belong to the same group too. But we should also reverse this question. If the Dmanisi sample shows the same variability as a sample of at least two (possibly three or four) different species of Homo, perhaps we should accept that there is more than one species in Dmanisi. Not only this hypothesis is never falsified in the article – but its opposite is taken as a starting premise (which is never contrasted!), to develop all the subsequent argument. One more thing: even when you cannot refute a null hypothesis, this does not make it true. This simply means that it has not been possible to refute it because the data employed lack the necessary resolving power. Lordkipanidze and his team assume that they are dealing with a single population because all the fossils have been recovered in the same geological horizon, rather “close” in time. Some researchers are critical of this point, however. In particular, the team led by Calvo-Rathert of the University of Burgos, in a paper published in Quaternary Research in 2008, suggested that the accumulation of hominids might have taken place over a period of up to hundreds of thousands of years. Even by more conservative estimates (i.e. those of Lordkipanidze and colleagues), stratigraphic levels with evidence of human presence might span several tens of thousands of years, even if they favour a shorter period. What cannot be denied following previous works on the chrono-stratigraphy, is that mandibleD2600 derives from a horizon with normal polarity  (that is, it belongs to the Olduvai subchron and is thus older than 1.78 Ma), and some of the other fossils have been recovered in layers with reverse polarity (post-Olduvai and hence more recent than 1.78 Ma). This chrono-stratigraphic datum implies that the possible period of human occupation in this Caucasian territory may have been longer. Therefore, the possibility that more than one lineage could have inhabited the same geographic area seems less remote. Dmanisi al dente In 2008 we published in the Journal of Human Evolution an article where we analysed in detail the dentition of the Dmanisi hominids. In this article we highlighted the markedly primitive aspect of some of the Georgian teeth – most notably the canines, which in some cases resembled those of  H. habilis or even the genus Australopithecus. In this sense, it would have been interesting for skull 5 to have been analysed with a larger sample of hominids, including Australopithecus (and, I darsesay, even with Paranthropus, given the massive face and jaw). This is particularly relevant because there is no clear consensus about the thresholds that separate Homo from Australopithecus for these chronologies. For example, as pointed out by its discoverer Lee Berger, it is unfortunate that Ausralopithecus sediba was left out of the comparison, as it represents one of the most complete craneal and postcraneal records of what is often called “the missing link”. The teeth of D2600 – or rather, the roots, since the crowns are very worn – are characterised by very robust and primitive length, number and degree of bifurcation and molarisation – features that only appear together in some specimens of Homo rudolfensis, Australopithecus or even Paranthropus.


In the other jaws, namely D211 and D2735, despite the habilis-like traits, some features are surprisingly modern for a population of this age. These include, among others, teeth with decreasing molar series (the first molar is larger than the second, which is larger than the third). I would not like to bore the reader with exhaustive dental descriptions but I need to emphasise the following: the derived traits that we identified in all the Dmanisi mandibles but D2600 are so atypical that they are not found in the fossil record until the Middle Pleistocene, almost one million years later. This “rarity” was shared by D211 and D2735, which pointed at some close kinship or “family” relation between them, and was absent in D2600, which showed instead some traits that already seemed at the extreme of the known variability for Homo. Based on this evidence, back then we already suggested the possibility that at least two different paleodemes or populations might be represented in the Dmanisy hypodigm (without going into the discussion of whether or not these represented different species). Only a few weeks ago, in a paper published in PNAS, the Dmanisi team suggested that the differences existing among the jaws were due to dentognathic pathologies of this unusual individual. According to this article, led by Ani Margvelashvili, jaw D2600 showed a very severe pattern of dental wear. The loss through wear of part of the crowns would have led to a remodelling of the height of the mandibular bone in order to compensate for this lack of dental “height” and allow for occlusion with the maxillary jaw. According to these authors, this supposed growth would be responsible for the dramatic height differences in the body and symphysis of D2600, and the height of D2735 and D211, as well as for variations in the shape of the arcade. Figure 5. From left to right: Elena Lacasa-Marquina, Pilar Fernández Colón, María Martinón-Torres and Laura Martín-Francés with mandible D2600 at the National Museum of Georgia in February 2012. In an almost vertiginous time coincidence, the journal Comptus Rendus Palevol has just accepted an article by our team, led by Laura Martín-Francés, where we assess the pathologies in the jaw of skull 5 and arrive at two important conclusions relevant to this debate. One of them is that these alleged pathologies are not responsible for the differences in shape and size between the Dmanisi jaws. While it is true that D2600 shows a very severe and peculiar wear (discussed below), the compensation mechanisms would never lead to an increase in mandibular dimensions but to the exact opposite, i.e. a recession of the lower end of the jaw (or alveolar crest), which would become shorter and thicker as a response to the wear. What we do see (and have also identified in the 1.2 Ma jaw from Sima del Elefante in Atapuerca) is a compensatory hyper-eruption  of the teeth, that is, that they have continued to “emerge” or erupt in order to allow for occlusion with the upper jaw. Our study would therefore confirm that the morphological differences among the Dmanisi jaws potentially have a taxonomic origin rather than constituting a “deformation” caused by pathologies. Figure 6. Detail of the severe dental wear of D2600.            


In the face of the recent commotion, many people have wondered: is it so important whether all of these are considered members of the same species? Obviously, nobody’s life depends on the answer. Its significance depends on our interest in finding out what happened. In our case, as anthropologists, it depends on the degree of precision we seek to obtain in our understanding of human evolution. We are ever more demanding, which I think is fabulous as this is the only way to make progress in scientific knowledge – by pushing its limits. In the case of the Dmanisi finds, the possibility that two human species may have coexisted at the same place (if they really did coincide in space and time) would pose an extremely interesting ecological question – rather than a taxonomic one. This is where the second part of the article led by Laura Martín-Francés kicks in. In this article we concluded that the pattern of dental wear recorded in the mandile of the new skull 5 was the consequence of a fibrous and highly abrasive diet, similar to that of chimpanzees or gorillas. The wear pattern was typical of animals that eat a considerable amount of vegetal foodstuffs, such as plants and tubers, which also require pre-masticatory work (such as peeling and leaf removal). These activities also leave their marks on the teeth, macroscopically and microscopically. This indication of a predominantly vegetal diet might suggest an adaptation to a different ecological niche for the owner of skull 5 compared to the other hominids found at the site. This particular use of his teeth explains why, in spite of the great dental wear of D2600, surprisingly this mandile does now show the compensatory mechanisms (such as pronounced mesial drift) that one would normally expect in the teeth of a Homo individual with a similar extent of teeth wear. Even in jaws D211 and D2735, where teeth are a lot less worn, we found these compensatory mechanisms are more pronounced than in D2600. The evidence thus indicates a substantially different use of their teeth for the latter. In our view, this constitutes good empirical evidence of adaptation to different ecological niches – precisely the sort of evidence that, according to Lordkipanidze and his team, was needed for the hypothesis of several palaeospecies in Dmanisi to be considered more parsimonious than that of a single lineage. To sum up, we should not forget that a species is more than its morphology – and much more than its craneal morphology alone. It may be that craneal anatomy may not be a powerful enough tool to allow species discrimination for this period of the fossil record. I am not suggesting that teeth provide the ultimate alternative (though their morphology is subject to fewer changes than the rest of the skeleton because they don’t undergo any remodelling throughout their lives), but I would like to emphasise that, in order to reach conclusions as grand as the claim that all the early Homo are just one, it would be desirable to examine other parts of the skeleton (which are fortunately preserved). Also, where possible, we should complement anatomical studies with other studies that may refine our understanding of the ecology and behaviour of these populations. In the same direction as our palaeopathological study of D2600, some archaeologists such as Javier Baena and his team at the Universidad Autónoma de Madrid acknowledge the possibility that the tools recovered in Dmanisi may have been manufactured by two different groups. All in all, while I am well aware that we need further data before we can claim the existence of two lineages at Dmanisi, I believe that even more data are needed to support the idea that, at the dawn of humankind, there was only a single Homo species.  


María Martinón-Torres Dental Research Group Leader. National Research Center on Human Evolution (Burgos, Spain)   Related references:   David Lordkipanidze, Marcia Ponce de León, Ann Margvelashvili, Yoel Rak, G. P. Rightmire, Abesalom Vekua, Christoph P.E. Zollikofer. 2013. A complete skull from Dmanisi, Georgia, and the evolutionary biology of early Homo. Science 342, 326-331. Ann Margvelashvili, Christoph P.E. Zollikofer, David Lordkipanidze, Timo Peltomäki, Marcia S. Pocne de León. 2013. Tooth wear and dentoalveolar remodeling are key factors of morphological variation in the Dmanisi mandibles. PNAS, doi: 10.1073/pnas.1316052110. María Martinón-Torres, José María Bermúdez de Castro, Aida Gómez-Robles, Ann Margvelashvili, David Lordkipanidze, Abesalom Vekua. 2008.  Dental remains from Dmanisi: Morphological analysis and comparative study. Journal of Human Evolution 55, 249-273. Laura Martín-Francés, María Martinón-Torres, Elena Lacasa-Marquina, Pilar Fernández-Colón, Ana Gracia-Téllez, José María Bermúdez de Castro. 2013. Palaeopathology of the Plio-Pleistocene specimen D2600 from Dmanisi (Republic of Georgia). Comptes Rendus Palevol (in press). Javier Baena, David Lordkipanidze, Felipe Cuartero, Reid Ferring, David Zhavnia, Diego Martín, Teona Sheila, Gioergi Bidzinashvili, Marta Roca, Daniel Rubio. 2010. Technical and technological complexity in the beginning: The study of Dmanisi lithic assemblate. Quaternary International 223-224, 45-53. Calvo-Rathert, M., Goguitachichvili, A.:, Slogashvili D., Villaláin, J.J., Bógalo, M.F., Carrancho, A., Maissuradze, G. 2008. New paleomagnetic data from the hominin bearing Dmanisi paleo-antrhopologic site (southern Georgia, Caucasus).Quaternary Research 69, 91-96.     *Thanks to my brother Marcos Martinón-Torres at UCL, for everything and also for this.