Nos tutelles



Accueil > Equipes > IDentifications, Environments & Anthropometry (IDEA)


par Dumoncel - publié le , mis à jour le

The origin of mankind (genus Homo) and its closest relative, the genus Paranthropus
Prof J. Braga (UMR 5288 CNRS)
As yet, taxonomic and phylogenetic hypotheses about early human evolution in Africa have been hampered by the absence of an undisputable chronology that records (relatively) continuously the faunal and hominin changes that may have occurred in southern Africa over the 3 to 2 (million years ago) Ma period. The 3-to-2 Ma period is critical because a major change in hominin evolution took place in eastern Africa at the 2.8 Ma key faunal change when Australopithecus was replaced by the two parallel branches representing Paranthropus and Homo.
The newly discovered fossil sample from Kromdraai Level 2 represents a high species diversity that includes mammals but also birds, reptiles and amphibians. This hominin-bearing stratigraphic sequence is used to clarify the aforementioned phylogenetic conundrum and to shed new light on early human evolution in southern Africa. The newly-discovered hominin sample from Kromdraai Level 2 contribute to a better understanding of the origin of Paranthropus in southern Africa. It provides a palaeoenvironmental and phylogenetic scenario of the Australopithecus to Paranthropus/Homo replacement in this region. Moreover, in the context of recent discoveries of relatively large Paranthropus specimens at both Gondolin and Swartkrans, the new Paranthropus sample from Kromdraai Level 2 helps determine whether the southern African Paranthropus body size and sexual dimorphism was underestimated due to some taphonomic bias associated with carnivore involvement and selective predation. Further palaeobiological analyses of this Paranthropus sample from Kromdraai Level 2 allow us to determine whether differences observed between southern African Paranthropus samples reflect alpha taxonomic and/or evolutionary factors.

Prof. Francis Thackeray, Dr Berhnard Zipfel & Dr Christine Steininger, Evolutionary Studies Institute at the University of the Witwatersrand, South Africa
Prof. Frederick E. Grine, Department of Anthropology at the State University of New York
Dr Jean-Baptiste Fourvel, UMR 5608 CNRS (TRACES), Toulouse, France
Prof. Jean-Luc Schneider, UMR 5805 CNRS (EPOC), Bordeaux, France
Dr Marco Pavia, Universita degli Studi du Torino, Italie
Dr Laurent Bruxelles, USR 3336 CNRS, IFAS Johannesbourg
Dr Veronika Zimmer, School of Biomedical Engineering & Imaging Science, King’s College London (Postdoctoral Research Fellow)
Dr Jean Dumoncel, UMR 5288 CNRS (AMIS), Toulouse, France
Nonkululeko Ngoloyi, UMR 5288 CNRS (AMIS), Toulouse, France (Erasmus Mundus scholarship) (PhD student)

3D Morphometrics : sexual dimorphism and interspecific variation (taxonomy)
Prof. J Braga (UMR 5288 CNRS)

Assessment of sexual dimorphism in early hominin species is crucial to understanding their paleobiology and has also important implications for reconstructions of social behaviors such as competition, mating strategies and signaling. Discerning sexual dimorphism from interspecific variation is a pivotal challenge and has been prominent in debates concerning the status of several hypodigms (e.g., Australopithecus afarensis, Australopithecus africanus and Homo habilis). A common method is the comparison of the ratios of the extreme dimensions to the ratios of female to male means of extant species, but it is difficult to apply to the paleontological record because the sex of fossil specimens is not known.
In order to distinguish sexual dimorphism from interspecific variation, we use an approach that is already widely applied in the field of “computational anatomy. “Computational anatomy” has the advantage of enabling direct computations of differences (i.e., continuous, optimal, and smooth deformations) between two (or more) 3D shapes that evince distinct morphologies even if the features to be investigated are not homologous. We use nonlinear elastic matching between shapes highlights local correspondences and differences by using full paths of deformations. Our method allows the computation of optimal deformations in a 3D shape space. Optimal deformations represent the best correspondences between points across the two shapes as well as the mapping (vector field) that optimizes the deformation (transformation from one to another). We also visualize the shape means and covariances enabling a full statistical analysis of sexual dimorphism and interspecific variation.

Dr Chafik Samir, Dr Anis Fradi, UMR 6158 CNRS, Université Clermont Auvergne, France
Prof. Jean-Michel Loubes, Dr Laurent Risser, UMR 5219 CNRS, Toulouse, France
Dr Jean Dumoncel, UMR 5288 CNRS (AMIS), Toulouse, France
Dr Veronika Zimmer, School of Biomedical Engineering & Imaging Science, King’s College London (Postdoctoral Research Fellow)
Dr Didier Descouens, Clinique Pasteur, Toulouse, France

• PhyloShapes 3D
Prof. J Braga (UMR 5288 CNRS), Dr C. Samir (UMR 6158 CNRS)
Whereas distinct methods have been proposed to measure and test for phylogenetic signals, all of the approaches previously proposed have used traits in finite-dimensional spaces represented either by a vector of features or parametric functions (e.g., logarithmic, polynomials, b-spline) based on angles and lengths, both leading to strong restrictions when performing statistical studies. In the present “PhyloShapes 3D” project, we investigate 3D anatomical features directly by their real (i.e., nonlinear) shapes (see the “3D Morphometrics” project). This project is motivated by the fact that the early hominin phylogenetic relationships have not yet been resolved using cladistics methods since a variety of cladistic trees representing distinct hypotheses of hominin phylogeny have been produced over the last decades. Our main aim is to predict 3D shapes of taxonomically and phylogenetically meaningful anatomical features at key stages of human evolution.

Dr Chafik Samir, Dr Anis Fradi, UMR 6158 CNRS, Université Clermont Auvergne, France
Prof. Jean-Michel Loubes, Dr Laurent Risser, UMR 5219 CNRS, Toulouse, France
One “80/Prime” CNRS PhD fellow (recruited soon), UMR 6158 & UMR 5288 CNRS, France

• Supercomputing and phenotypic analyses with diffeomorphism
Prof. J Braga (UMR 5288 CNRS), Dr. J. Dumoncel (UMR 5288 CNRS)
Since 2014, we have been collaborating with the regional supercomputing centre CALMIP (p1440 project) in Toulouse for getting access to their supercomputer. We deployed a code in python dedicated to shape analysis : Deformetrica. Defrometrica uses diffeomorphism to capture differences between shapes which can be represented with images, points, curves or surfaces. Deformetrica uses resource-intensive mathematical tool, but the code can use multiprocessing and GPU computation to speed up the computations. We used the supercomputer from CALMIP to compute our analyses on paleoanthropological material. Diffeomorphism can be applied to various kind of anatomical structures, such as molar, endocast or inner ear. We develop tools for the pre- and post-processing and the analyses of meshes used with Deformetrica (

D. Ginibriere, UMR 5288 CNRS (AMIS), Toulouse, France
Dr N. Renon, UMS 3667 (CALMIP), Toulouse, France
Dr G. Subsol, UMR 5506 (LIRMM), Montpellier, France

• Micro-CT scanner
Prof. J Braga (UMR 5288 CNRS), D. Ginibriere (UMR 5288 CNRS)
In collaboration with Carestream, we designed a micro-CT system to investigate small anatomical structures such as teeth or small bones. We used a micro-focus x-ray source which was associated to a system to rotate the objects. By using Carestream Dental reconstruction software, we are able to reconstruct a 3D image of the object with a resolution of 18.5 µm.

Dr J. Dumoncel, UMR 5288 CNRS (AMIS), Toulouse, France
J. M. Inglese, C. Maury, Carestream, France