News and Views
Variation in the mandibles from Dmanisi, Georgia

https://doi.org/10.1016/j.jhevol.2008.02.003Get rights and content

Introduction

The site of Dmanisi provides abundant evidence for hominin occupation in Georgia, dated at approximately 1.77 Ma (Gabunia and Vekua, 1995, Gabunia et al., 2000, Vekua et al., 2002, Lordkipanidze et al., 2006, Rightmire et al., 2006). Crania, mandibles, and postcranial remains document at least five individuals (Jashashvili, 2005, Lordkipanidze et al., 2007) ranging in age from subadult (adolescent) to older adult, and almost surely sampling both sexes. Our published work suggests that the Dmanisi skulls are likely drawn from a single population, accumulated over a short interval of time, and best referred on current evidence to Homo erectus. However, it has been noted that the Dmanisi mandibles are quite dimorphic (Gabunia et al., 2002), and some workers discern important differences among the crania (Schwartz, 2000, Schwartz and Tattersall, 2003). Such observations raise the question of whether two taxa may be sampled at the site. On the basis of cranial capacity comparisons, Lee (2005) finds no compelling reason to reject the (null) hypothesis that variation in the Dmanisi assemblage conforms to that expected within a single species. Recently, Skinner et al. (2006) have revisited this question, and their treatment of mandibular dimensions supports an alternative interpretation. Whether the fossils should be sorted to one dimorphic group, or several taxa, is clearly a key issue. Resolving it will generate new hypotheses concerning the level and pattern of variation characteristic of ancient Homo.

Skinner et al. (2006) comment on the degree of size difference present in the Dmanisi mandibles, remarking once again on the contrast between D2600 and the smaller D211 jaw. An analysis of corpus measurements suggests to these authors that either the Dmanisi population was so sexually dimorphic as to raise doubts about its status within Homo, or the D2600 mandible should be grouped separately from the other specimens. We support such studies of the fossils. The site offers unprecedented access to an early Pleistocene paleodeme, and there are many novel research opportunities. However, it is important when making use of published measurements to note the cautionary information supplied in primary sources. In their treatment of D2600 and D211, Skinner et al. (2006) have not done this. As a consequence, the basis for their comparisons is suspect, and their conclusions must be questioned.

To assess size and shape of the mandible, Skinner et al. (2006) employ just four measurements. Corpus height and breadth are measured at the symphysis and at the position of M1. For each variable separately, and also for a combined measure of size, the ratio of larger (D2600) to smaller (D211) values is obtained for the Dmanisi specimens. These ratios are then compared to distributions of similar ratios generated by an exact resampling procedure applied to groups of modern humans, chimpanzees, gorillas, and orangutans. Skinner et al. (2006) find that it is particularly corpus height that sets the two Dmanisi mandibles apart. The D2600/D211 height differences are large enough to occur only rarely in a series of resampling experiments carried out for reference populations of extant apes and humans. However, ratios of corpus breadth at least as large as the Dmanisi values are encountered more frequently in all of the living hominoids. Neither for thickness of the symphysis nor for breadth at M1 do any of the comparisons produce significant differences. Another analysis in which the mandibular measurements are standardized by the geometric mean shows that the Euclidean “shape” distance between D2600 and D211 can be matched in recent human samples but is less likely to be equaled or exceeded in the ape populations.

Section snippets

Condition of the Dmanisi mandibles

The comparisons carried out by Skinner et al. (2006) are critically dependent on the four dimensions utilized for the Dmanisi mandibles. Obtaining these data would be straightforward if there were little damage to the specimens, but in many instances, fossils are not perfectly preserved. This is the situation for D211. Here, the symphyseal region is intact, but the corpus is broken inferiorly, and all of the posterior part of the jaw is missing. Breakage is quite evident from the photographs,

Alternative metric approaches

If measuring the Dmanisi mandibles in the way proposed by Skinner et al. (2006) is problematic, then other approaches should be explored. One possibility is to examine only portions of the corpus that have not been damaged. The base of D211 is broken below P4, but for both this specimen and D2600, corpus heights can be estimated more reliably at the position of P3 (Table 1). The height ratio obtained at this location (right side) is 1.66. This ratio can be compared to distributions generated

The importance of age-related changes in morphology

Apart from the evidence supplied by measurements and statistical treatment, it is worth considering in more detail how the morphology of the Dmanisi specimens might be altered by the growth process. Skinner et al. (2006) acknowledge that, given the apparent age separation of D211 from D2600, in vivo changes may represent “a confounding factor” in comparisons. They argue that the effects of age are adequately accounted for in their analysis, as the human and ape reference samples include both

Discussion and conclusions

Our findings parallel those of Skinner et al. (2006) in showing that D211 and D2600 differ in corpus height. This is documented by measurements of the symphyseal region, where both specimens are relatively intact. However, the significance of such observations is limited. Posteriorly, the D211 jaw is broken, and D2600 presents obvious pathology associated with dental wear. Measurements taken in the region of M1 will be inaccurate and should not be used to construct indices of overall size and

Acknowledgements

We appreciate the efforts of all of the dedicated students and professionals who have joined in the excavations and worked in the laboratory at Dmanisi, and we are grateful for support provided by the Georgian National Museum. For assistance with our research, we thank M. Bukhsianidze, R. Ferring, Y. Haile-Selassie, T. Jashashvili, L. Jellema, G. Kiladze, C.O. Lovejoy, R. Meindl, M. Ponce de Leon, M. Tappen, A Vekua, and C. Zollikofer. A. Margvelashvili and S. Kauffman helped to prepare the

First page preview

First page preview
Click to open first page preview

References (24)

  • Y. Kaifu et al.

    Taxonomic affinities and evolutionary history of the early Pleistocene hominids of Java: dentognathic evidence

    Am. J. Phys. Anthropol.

    (2005)
  • Kaifu, Y., Indriati, E., Aziz, F., Kurniawan, I., Jacob, T., Baba, H. Cranial morphology and variation of the earliest...
  • Cited by (0)

    View full text