Digging Deeper at Dmanisi

Before the recent excavations all of the hominin fossils and archeological evidence from Dmanisi had come from stratum B. The 73 artifacts and 34 bone fragments recently recovered from stratum A (specifically from subunits A2a–A4a) at M5 have been interpreted as coming from a series of “living surfaces” on ash falls that were exposed long enough for soil formation to begin. Most of the artifacts were flakes, but there are a few cores and choppers; all are consistent with the simplest, mode 1, category of artifacts. The artifacts from stratum A are mostly made on a red tuff that was exploited more intensively than the raw materials used to make the artifacts found in stratum B.

None of the sediments in any of the excavations at Dmanisi can be older than the underlying Masavera Basalt, which has been dated to ca. 1.85 Ma. The sediments that make up stratum B have been dated by magneto- and biochronology and by correlative argon–argon dating to ca. 1.77 Ma. In contrast, the sediments in stratum A all display a normal magnetic signal that is interpreted as the upper Olduvai subchron; this places them between 1.85 and 1.78 Ma (the latter being the age of the Olduvai–Matuyama boundary). The evidence from Dmanisi thus provides a minimum first appearance date (FAD) for hominins in the Caucasus, but it is very

Evidence for H. erectus outside of Africa extends back to at least ca. 1.85 Ma.

likely that hominins were there long before ca. 1.85 Ma, for in the absence of long sections FADs are subject to substantial error (3).

Thus far, no hominin fossils have been recovered from the sediments of stratum A, but the archeological evidence is consistent with the hypothesis that the artifacts in stratum A were made by the hominin taxon that is sampled in stratum B. However, what type of hominin is that? Some have suggested that the hominin fossils recovered from Dmanisi may sample more than one taxon, but most observers see just one morph in the sample. At the time of their discovery the morphology of the first hominin fossils to be recovered did not fit neatly into what was then known about either H. habilis or Homo erectus, but subsequent H. erectus discoveries in East Africa (4, 5) have expanded our knowledge of that taxon. Most experienced observers agree that the Dmanisi hominins are appropriately assigned to H. erectus, but with the caveat that they lie at the primitive end of that taxon's morphological spectrum; some observers have suggested they sample a taxon that is more primitive than H. erectus (6, 7).

A More Primitive African Emigrant?

For a long time now, the conventional wisdom has been that the first hominins to leave Africa were morphologically and adaptively quite distinct from Australopithecus (e.g., they had body proportions much like tropical modern humans and were capable of imagining and manufacturing mode 2 artifacts), but this view has had to be modified because of the hominin fossils assigned to Homo floresiensis from Liang Bua (8). The hominin sample from Liang Bua now includes nearly 100 individually numbered specimens that are estimated to represent fewer than 10 individuals. The taxon H. floresiensis was immediately controversial for at least two reasons. First, the estimated geological age of the specimens of between ca. 17 and ca. 74 ka substantially overlaps the presence of modern humans in the region. Second, although its discoverers and describers acknowledged the small overall size of H. floresiensis (the stature of LB1 is estimated to be 106 cm and its body mass to be ca. 25–30 kg) it had an exceptionally small brain (ca. 417 cm³) for an adult hominin individual. A few researchers have suggested that the hominin fossils from Liang Bua sample a population of Homo sapiens—most likely antecedents of the small-statured Rampasasa people who live on Flores today—afflicted by either an endocrine disorder, or some other syndrome, that includes microcephaly. Both explanations, a unique endemically dwarfed hominin species or pathology, are exotic, but the researchers who espouse a pathological explanation for the individuals represented by LB1–15 need to explain what pathology results in an early Homo-like cranial vault, archaic hominin-like mandibular, dental, carpal, and pedal morphology, and a brain that although very small apparently has none of the morphological features associated with microcephaly. Initially it was suggested that H. floresiensis was an endemically dwarfed H. erectus, but the burden of subsequent analyses suggest that it may be an endemically dwarfed version of a more primitive H. habilis-grade hominin (9–12).

Homo Who?

Although I suggested that many of my colleagues are agreed regarding the “what” with respect to Homo (i.e., that the genus Homo includes the species H. habilis), in the past some of us have argued (13) that if a genus is both a clade and a grade, then there are grounds for excluding H. habilis from Homo because it is very likely not in the same grade as the type species of Homo, H. sapiens. The grade arguments involve a wide range of regional and functional morphology, including the relative size of the postcanine tooth crowns (13) and mandibular corpus (14), the proportions (15) and relative strength (16) of the limbs, carpal morphology (9), and diet as inferred from microwear evidence (17). All these lines of evidence suggest that the adaptive coherence of Homo would be compromised if H. habilis is included in Homo. Thus, if these arguments are accepted the origins of the genus Homo are coincident in time and place with the emergence of H. erectus, not H. habilis.

We have seen from the most recent evidence from Dmanisi (1) that evidence for H. erectus outside of Africa extends back to at least ca. 1.85 Ma, and some claim that fossils of H. erectus from Java may also be in that age range. However, what about the earliest evidence of H. erectus in Africa? There is uncontestable evidence of H. erectus in the form of the KNM-ER 3733 cranium from Koobi Fora, whose geological age is ca. 1.78 Ma, but many point to an occipital fragment, KNM-ER 2598, whose geological age is ca. 1.87 Ma, as the earliest evidence of H. erectus in Africa. Thus, the new evidence from Dmanisi suggests that the FADs for H. erectus both inside and outside of Africa are much the same, but it must be remembered that both FADs are likely to be adjusted upward.

Uncertain Origins

So what does all this mean for the origin of Homo? The following scenarios are compatible with the meager evidence we have; it would be misleading to claim that any of the scenarios are supported by that meager evidence. I am skeptical about the merits of including H. habilis in Homo, but if it is included then Homo most likely evolved in Africa, but “where” in Africa, “when,” and “from what” are all questions presently without reliable answers. If, on the other hand, H. habilis is excluded from Homo, then the meager evidence we have is consistent with H. erectus, and thus Homo, evolving from a H. habilis-grade taxon either within, or outside of, Africa. In the most parsimonious version of the “outside of Africa” scenario for the origin of Homo, H. erectus would have evolved from a H. habilis-grade hominin either in Asia or in Southeast Asia, and then H. erectus would have migrated to Africa some time before 1.87 Ma. This scenario is consistent with the aspects of the morphology of the Dmanisi and the Liang Bua hominins that are more primitive than the condition seen in H. erectus sensu stricto. This primitive morphology is one of stumbling blocks for a “within Africa” scenario for the origin of H. erectus taxon, for it would mean the migration out of Africa of two hominins, first a H. habilis-grade taxon then H. erectus sensu stricto. Another stumbling block for an ancestor-descendant relationship between H. habilis and H. erectus sensu stricto within Africa is that both the ancestor and the descendant overlap in time in East Africa for several hundred thousand years (5).

Even Dmanisi Is Not Enough

We must hope that researchers find more sites as productive as Dmanisi so that today's meager evidence for the origin of Homo will soon be augmented. In the meantime we need to be realistic about what can, and what cannot, be deduced about hominin evolutionary history. It is sobering to realize that even in the case of a taxon such as Homo neanderthalensis that has an order of magnitude better fossil record than for early Homo, we still have much to learn about its origin and evolution.

Acknowledgments

B.W. receives support from the George Washington University Provost and the George Washington University Signature Program. Research by graduate students funded by National Science Foundation Integrative Graduate Education and Research Traineeship Program Grants DGE-0801634 and 9987590 has been used in this review.

Footnotes

References