A virtual assessment of the suprainiac depressions on the Eyasi I (Tanzania) and Aduma ADU-VP-1/3 (Ethiopia) Pleistocene hominin crania

Abstract

Despite a steady increase in our understanding of the phenotypic variation of Pleistocene Homo, debate continues over phylogenetically informative features. One such trait is the suprainiac fossa, a depression on the occipital bone above inion that is commonly considered an autapomorphy of the Neanderthal lineage. Challenging this convention, depressions in the suprainiac region have also been described for two Pleistocene hominin crania from sub-Saharan Africa: Eyasi I (Tanzania) and ADU-VP-1/3 (Ethiopia). Here, we use a combined quantitative and qualitative approach, using μCT imaging, to investigate the occipital depressions on these specimens. The results show that neither the external nor the internal morphologies of these depressions bear any resemblance to the Neanderthal condition. A principal component analysis based on multiple thickness measurements along the occipital squama demonstrates that the relative thickness values for the internal structures in Eyasi I and ADU-VP-1/3 are within the range of Homo sapiens. Thus, our results support the autapomorphic status of the Neanderthal suprainiac fossa and highlight the need to use nuanced approaches and multiple lines of evidence.

Introduction

The suprainiac fossa has been widely considered a Neanderthal-derived feature, described as a discrete elliptical depression with an uneven floor, confined to the area directly above inion, between the bilateral arches of the occipital torus (a prominent shelf-like structure between the supreme and superior nuchal lines; Hublin, 1978a, Hublin, 1978b, Hublin, 1983), with an apex on the midline (Hublin, 1978a, Hublin, 1984, Santa Luca, 1978, Stringer et al., 1984, Schwartz and Tattersall, 2005, Harvati et al., 2007, Harvati et al., 2019, Balzeau and Rougier, 2010, Nowaczewska, 2011, Franciscus and Holliday, 2013, Harvati, 2015, Nowaczewska et al., 2019). Initially described by Virchow (1872) as a pathological cranial feature on the Neanderthal holotype specimen, Feldhofer 1, it was later recognized by Klaatsch, 1902, Gorjanović-Kramberger, 1902, and Boule (1911–1913) as a nonpathological trait present on the Krapina (Croatia), Spy (Belgium), and La Chapelle-aux-Saints (France) Neanderthal crania. According to the accretion hypothesis (Dean et al., 1998, Hublin, 1998), which proposes the gradual establishment of specific Neanderthal craniofacial traits across four ‘stages’ of Neanderthal evolution (early pre-Neanderthals, pre-Neanderthals, early Neanderthals, and classic Neanderthals), the suprainiac fossa likely became fixed in the Neanderthal lineage. Some scholars consider the trait to be present in all Late Pleistocene European Neanderthals (e.g., Smith et al., 2005) and, in a less pronounced form, in Middle Pleistocene European specimens that are considered to be early members of the Neanderthal clade (Marston, 1937, Vandermeersch, 1978, Hublin, 1978a, Dean et al., 1998, Balzeau and Rougier, 2010, Arsuaga et al., 2014, Meyer et al., 2016).

Depressions above inion proposed as possibly homologous, or at least related, to the Neanderthal suprainiac fossa have been reported on several adult fossil human crania from the late Middle and Late Pleistocene of Africa (e.g., Haile-Selassie et al., 2004, Trinkaus, 2004, Balzeau and Rougier, 2010), the Late Pleistocene of Europe (e.g., Frayer, 1992, Kramer et al., 2001, Caspari, 2005, Soficaru et al., 2006, Soficaru et al., 2007, Trinkaus, 2007), the Late Pleistocene of the Levant (e.g., Skhūl V, Qafzeh VI, and Manot 1; Kramer et al., 2001, Smith et al., 2005, Hershkovitz et al., 2015, Hershkovitz et al., 2017), the Middle Pleistocene of China (Xuchang 2; Li et al., 2017), and the Late Pleistocene of Australia (e.g., Kow Swamp and Willandra Lakes; Thorne and Macumber, 1972, Wolpoff et al., 2001, Curnoe, 2011). Moreover, similar structures have been described for Holocene specimens from Europe, Africa, and Australia (Mirazón Lahr, 1994, Balzeau and Rougier, 2010, Nowaczewska, 2011, Nowaczewska et al., 2019).

Because of possible morphological similarities between the Neanderthal suprainiac fossae and some of these suprainiac depressions, the etiology and phylogenetic significance of these traits are the subject of continued discussion (e.g., Trinkaus, 2004, Caspari, 2005, Balzeau and Rougier, 2010, Bosman and Harvati, 2019, Nowaczewska et al., 2019). In particular, the presence of similar features on African Middle and Late Pleistocene hominins puts into question the status of the suprainiac fossa as a Neanderthal autapomorphy because the trait could possibly be inherited from the last common ancestor of modern humans and Neanderthals. Currently, however, only a handful of purported early Homo sapiens specimens from the African Middle and Late Pleistocene have been described as exhibiting a suprainiac depression. These include Eliye Springs KNM-ES 11693 (Bräuer and Leakey, 1986), Kébibat (Rabat) 1 (Saban, 1975), Eyasi I (Reck and Kohl-Larsen, 1936, Trinkaus, 2004), and ADU-VP-1/3 (Haile-Selassie et al., 2004). It should be noted that these specimens, with the exception of the Late Pleistocene ADU-VP-1/3 (Haile-Selassie et al., 2004, Yellen et al., 2005), are very poorly dated and can only be broadly placed within the Middle Pleistocene (Stearns and Thurber, 1965, Bräuer and Leakey, 1986, Mehlman, 1987, Domínguez-Rodrigo et al., 2008, Millard, 2008). As such, there is a limited understanding of the morphological variability in this time period.

To this end, we investigate here the proposed suprainiac depressions on the Eyasi I (Tanzania) and ADU-VP-1/3 (Ethiopia) specimens, maximizing the amount of information to be gained from these specimens through a systematic review of both their external and internal morphologies. These fossils are especially relevant for the common ancestry scenario as they date close to currently proposed time of appearance of modern human anatomical traits in eastern Africa (Eyasi I; 375–88 ka; Mehlman, 1987; Domínguez-Rodrigo et al., 2008) and to times of dispersal of anatomically modern humans out of Africa (ADU-VP-1/3; 105–79 ka; Yellen et al., 2005), respectively. Moreover, we focus on the internal morphology of the occipital squama (i.e., external table, diploic layer, and internal table) as recent research has shown that this can be especially informative in the distinction between Neanderthal suprainiac fossae and H. sapiens suprainiac depressions (Balzeau and Rougier, 2010, Balzeau and Rougier, 2013, Bosman and Harvati, 2019, Nowaczewska et al., 2019). Balzeau and Rougier (2010) found that the H. sapiens form of the suprainiac depression is characterized by significant bone remodeling and thinning restricted to the external table. In contrast, the Neanderthal form of the suprainiac depression is characterized by thinning of the diploic layer, whereas the external table exhibits no substantial remodeling and thinning (Balzeau and Rougier, 2010). These results suggest that the suprainiac fossa found in Neanderthals can be considered an autapomorphic (uniquely derived) trait.

If the depressions on Eyasi I and ADU-VP-1/3 are found to be quantitatively and qualitatively similar to the Neanderthal condition, then the status of the suprainiac fossa as a diagnostic Neanderthal autapomorphy must be reconsidered as it might be a homologous ancestral trait shared in the two lineages, thereby complicating the taxonomic assignment of isolated occipital bones in the fossil record. Alternatively, if the depressions above the inion on Eyasi I and ADU-VP-1/3 are found to differ from the Neanderthal condition, then the suprainiac fossa can be retained as an autapomorphic trait, and the common ancestry scenario should be reconsidered. Here, we explore these issues by performing a qualitative assessment of the external morphology of the putative suprainiac depressions and related occipital superstructures in Eyasi I and ADU-VP-1/3 as well as by making both qualitative and quantitative evaluations of their internal structure using microcomputed tomography (μCT). This is followed by a discussion on how our results inform the common ancestry hypothesis and the etiology of the suprainiac depression.