Middle Pleistocene subsistence in the Azraq Oasis, Jordan: Protein residue and other proxies
Introduction
Genetic evidence has revealed the complexity of interbreeding and population assimilation that took place across Eurasia between 100,000–45,000 years ago involving modern humans and archaic populations such as the Neanderthals and the Denisovans (e.g., Kuhlwilm et al., 2016, Reich et al., 2010, Simonti et al., 2016). Key to understanding the nature of these interbreeding and assimilation events is learning more about the lives of hominins already subsisting on the landscape when later humans arrived. Modern humans subsisting in extreme environments today often adapt by broadening their foraging strategies to incorporate a wide variety of plants and animals including insects and other arthropods into their diets. Faunal remains from Late Pleistocene sites (130,000–11,000 years ago) in Eurasia show this to be a cumulative trend. A key question in paleoanthropology concerns what strategies earlier hominins followed as they dispersed across a highly variable Eurasian landscape. Here we describe the oldest known family-specific protein residue on stone tools. Using cross-over immunoelectrophoresis (CIEP), seventeen tools from Shishan Marsh 1 (SM-1), a stratified, in-situ site in northeast Jordan dating to approximately 250,000 years ago tested positive for rhinoceros (3), duck (3), horse (5), camel (3), and bovine (3) protein residue. Based on lithic, faunal, paleoenvironmental and protein residue data, we conclude that Middle Pleistocene hominins were able to live in extreme environments through a reliance on surprisingly human-like adaptations including a broadened subsistence base, modified tool kit and strategies for predator avoidance and carcass protection.
SM-1 site is a securely dated and stratified site in the Azraq Oasis, which is located in the center of the endorheic Azraq Basin of Jordan's Eastern Desert (Fig. 1). Historically, the oasis was fed by springs originating from the upper aquifer of the basin, which is recharged in the Jebel Druze basalt massif to the north (El-Naqa, 2010). With a total precipitation in the central basin of 50 mm/year, the environment is arid and able to sustain only Saharo-Arabian vegetation (Al-Eisawi, 1996). The historic spring areas of the oasis, Azraq Druze and Azraq Shishan, have attracted hominin populations for at least 300,000 years, making the oasis one of the richest archaeological and paleontological landscapes in the Middle East. The drying of the marshes in the late 20th century due to considerable pumping of fresh water to supply urban areas permitted the study of deeply buried archaeological layers, which include Lower, Middle, Upper, and Epi-Paleolithic occupations. Moreover, evidence from terraces surrounding the basin and buried lacustrine deposits suggest high lake stands in the past (Abed et al., 2008, Ames and Cordova, 2015, Cordova et al., 2013). The complex stratigraphy of the oasis provides evidence of fluctuations between different depositional contexts, including marsh, playa, lacustrine, deltaic, and aeolian, suggesting extreme environmental changes associated with climatic and hydrological cycles (Abed et al., 2008, Ames and Cordova, 2015, Ames et al., 2014a, Cordova et al., 2013, Jones and Richter, 2011).
In a broader stratigraphic context, the Middle Pleistocene occupations at SM-1 are associated with a transitional environment from a receding lake to marshy ponds formed at the edge of a fan-delta (Fig. 2). Layers 10 and 9 correspond to lake recession with alluvial influx. Marshy ponds with gentle alluvial sediment influx correspond to layers 8, 7c and 7b. Subsequent desiccation is associated with aeolian (sandy silt) deposits represented by Layer 7a.
Large bifacial and flake-based stone artifacts are associated with layers 8, 7c and 7b. Not all layers produced pollen and phytoliths, but the archaeological layers indicate a predominance of aquatic vegetation surrounded by desert scrub typical of the hot deserts of the modern Middle East (Fig. S3). Modern desert plants are dominated by dry-adapted and salt-tolerant chenopod scrub (Chenopodiaceae) and sand dune vegetation (Calligonum comosum). The local aquatics are dominated by grasses (Poaceae), rushes (Juncaceae), cattails (Typha spp.), and sedges (Cyperaceae), among others (Table S1). Grass pollen diameters and phytolith cells suggest that the majority of the grasses were reeds (Phragmites australis) (Tables S1, S2, Fig. S4, Supplementary material). The bioclimatic context of the lithic assemblages more closely resembles modern conditions than those of the Last Glacial Maximum. The high Chenopodiaceae-Artemisia ratio (C/A) (Table S1) suggests that the hominin occupations may have occurred during an interglacial stage or near the transition from an interglacial to glacial stage, or more specifically under warm and dry conditions (see Supplementary material).
The faunal remains recovered to date are poorly preserved, but layers 8, 7b and 7c indicate the presence of gazelle (Gazella sp.), camel (Camelus sp.), wild cattle (Bos cf. primigenius), equids (Equus spp.), an extinct elephant (cf. Elephas), rhinoceros consistent with steppe rhinoceros (Stephanorhinus hemitoechus), probable lion (cf. Panthera leo), and other large carnivores. Steppe rhinoceros, equids, wild cattle, and camel were identified at the nearby Acheulean site C-Spring (Clutton-Brock, 1970, Clutton-Brock, 1989), and steppe rhinoceros, wild cattle, equids, and elephant were identified during previous excavations in the nearly adjacent site of 'Ain Soda (Dirks, 1998, Lister et al., 2013). These taxa are indicators of a dry, open, steppe environment with some shrubs (Bennett and Hoffman, 1999, Davis, 1980, Fortelius et al., 1993). Such animals would be attracted to water margins of the oasis and the associated plant resources, where they may have been ambushed by hominins and/or other predators. The interpretation of 'Ain Soda as a butchery site supports this scenario as one aspect of oasis usage (Lister et al., 2013, Rollefson et al., 1997).
The oldest OSL age estimate at SM-1 is 266 ± 40 kya (Supplementary material), and is assumed to be the minimum date for layer 8 (Fig. 2). Dates from upper layers 7b and 7a, 125 ± 12 kya and 119 ± 40 kya respectively, indicate the minimum age for the burial of the stone tools. However, based on our paleoenvironmental and geomorphic reconstructions, it is unlikely that these minimum dates represent an age for the creation and use of the cultural material. The surface of layer 7a was likely exposed and experienced continuous aeolian reworking for an extended period of time, which would result in a younger date than initial onset of aeolian depositional conditions. Moreover, the base of layer 7b and layer 8 produced pollen, but none was recovered from layer 7a, which suggests that the lithic material is resting in sediments that were relatively quickly buried and in primary context, as pollen is rarely preserved on exposed soil surfaces in warm and arid environments (Cordova, 2007). Continuous aeolian reworking of layer 7a may also have resulted in the incorporation of aeolian fine sand and silt into the upper parts of layer 7b, which would produce a younger age estimate and account for the lack of pollen preservation. Together this evidence points to a large gap in the depositional sequence between the production and use of the stone tools and the final deposition of layer 7a, meaning that for now the age of 266 ± 40 kya more closely represents the age of creation and use of the cultural material. Geomorphic and paleoenvironmental evidence from the surrounding region correlates with our interpretation.
A maximum high lake stand in the Azraq basin is reported between 346 and 316 kya (Abed et al., 2008)—when SM-1 would have been under water—while a relatively dry environment due to deposition of evaporative carbonate is suggested by a U-Th age of 220 ± 30 ka (Macumber, 2001) at a nearby site almost at the level of SM-1. Therefore, the derived proxies from SM-1 (pollen and phytoliths, see Supplementary material) suggest a regional transition from an exceptionally wet period to conditions similar to the present or even more arid. The sedimentary facies at SM-1, which are fluviolacustrine, and the aquatic nature of the local vegetation evident in the pollen and phytoliths, suggests that this locality represented an oasis where animals and hominins were brought together in sufficient numbers to leave substantial traces. Overall, the current geomorphic and paleoenvironmental data from SM-1 and the surrounding region indicate that this locale was repeatedly utilized by biface and flake tool-using populations between at least 300–150 ka—conservatively speaking—as minor fluctuations in climatic cycles reorganized the extent and distribution of water resources in the oasis. This age estimate represents the window of time between the highest known lake stand and the desertification of the local landscape. As more dates are obtained and more details of the paleolandscape reconstructed we expect to refine this window for the age of the Middle Pleistocene occupations at SM-1, particularly concerning the younger end of the timeframe.
Approximately 10,000 artifacts made from local flint nodules and small fluvial clasts from nearby wadi gravels were excavated during three field seasons from 2013 to 2015. Typologically, the archaeological assemblage at SM-1 corresponds to the Late Acheulean of Azraq facies characterized by small to moderately-sized ovate and discoid bifaces and a predominance of flake tools—all with sharp edges with little to no evidence of rolling or post-depositional edge damage (Copeland, 1988). The assemblage differs from a surface collection recovered from ‘Ain Soda (Rollefson et al., 1997) in having fewer cleavers, a moderate use of the Levallois technique and an abundance of small tools (i.e., utilized and/or retouched flakes less than 5 cm in length including scrapers, burins and borers) that were knapped on-site from local fluvial clasts. The condition of the artifacts in conjunction with preliminary analyses of debitage samples and of artifact orientation data suggest that the artifacts were recovered in primary context (Ames et al., 2014b).
Section snippets
Cross-over immunoelectrophoresis
Cross-over immunoelectrophoresis (CIEP) detects residues based on reactions between antibodies and antigens, wherein antibodies are used to detect unknown antigens. CIEP is more sensitive to proteins than other methods of protein residue analysis including enzyme-linked immunosorbent assay, radioimmune assay, and Western blot analysis, being able to detect 10−8 g of protein in a 5 μl sample (Culliford, 1964). CIEP has been used to detect protein residues from a variety of fish species on
Results
As noted in Section 2.1., a total of 44 tools out of nearly 7000 stone artifacts from SM-1 were chosen for residue analysis based on the presence of pronounced use-wear, which was evidenced by microflaking and rounding. An initial sample of 6 artifacts was tested against antisera from bovine, camel, goat, and horse with one artifact testing positive for horse protein residue (Table 1). On the strength of these results we carefully chose an additional 38 artifacts to be tested against camel,
Discussion
The results presented here are the oldest identifiable protein residues in the world and constitute direct evidence of the exploitation of specific taxa in the Levant by Middle Pleistocene hominins. As protein residue analyses become more widely employed in archaeological contexts, previously held suppositions about the preservation of biological evidence may need to be reconsidered (see discussion in Moore et al., 2016). Our data join a growing body of work that demonstrates the survival of
Conclusion
Positive protein residue results on 17 Middle Pleistocene stone tools from an assemblage dating to approximately 250 ka, in conjunction with associated lithic, faunal and other paleoenvironmental data suggest that these Middle Pleistocene hominins were adaptable, opportunistic and capable of exploiting a wide range of fauna—from waterfowl to rhinoceros—in what was likely one of the last humid refugia in the region. Such insights are all the more important as large swaths of inland Arabia, once
Acknowledgements
We thank the Department of Antiquities of Jordan and the Azraq Wetland Reserve of the Royal Society for the Conservation of Nature (Jordan) for permission to conduct our research and for all of their help, Dr. Eric Klaphake at the Cheyenne Mountain Zoo for providing the black rhinoceros serum and Dr. Leah Minc, Oregon State University Radiation Center and Dr. Loren Davis, College of Liberal Arts, Oregon State University for providing use of the Keyence VHX 600 Digital Microscope. We also thank
References (65)
- et al.
The paleoclimate of the eastern desert of Jordan during marine isotope stage 9
Quat. Res.
(2008) - et al.
Paleoenvironmental change and settlement dynamics in the Druze Marsh: results of recent excavation at an open-air Paleolithic site
Quat. Int.
(2014) - et al.
Interglacial and glacial desert refugia and the Middle paleolithic of the Azraq oasis, Jordan
Quat. Int.
(2013) - et al.
Woodworking activities by early humans: a plant residue analysis on Acheulian stone tools from Peninj (Tanzania)
J. Hum. Evol.
(2001) - et al.
The species identification of very old human protein stains
Forensic Sci. Int.
(1979) Mousterian technology and settlement dynamics in the site of Hummal (Syria)
J. Hum. Evol.
(2011)- et al.
Integration of use-wear with protein residue analysis - a study of tool use and function in the south Scandinavian Early Neolithic
J. Archaeol. Sci.
(2009) - et al.
Paleoclimatic and archeological implications of Pleistocene and Holocene environments in Azraq, Jordan
Quat. Res.
(2011) - et al.
New fossil remains of Elephas from the southern Levant: implications for the evolutionary history of the Asian elephant
Palaeogeogr. Palaeoclimatol. Palaeoecol.
(2013) - et al.
The use of sequential experiments and SEM in documenting stone tool microwear
J. Archaeol. Sci.
(2014)
Mousterian intra-site spatial patterning at Quneitra, Golan Heights
Quat. Int.
Residue and microwear analyses of the stone artifacts from Schöningen
J. Hum. Evol.
New insights on the wooden weapons from the Paleolithic site of Schöningen
J. Hum. Evol.
Evaluating protein residues on Gainey Phase Paleoindian stone tools
J. Archaeol. Sci.
Protein stability in preserved biological remains I: survival of biologically active proteins in an 8-year-old sample of dried blood
Int. J. Biochem.
Protein stability in preserved biological remains II: modification and aggregation of proteins in an 8-year-old sample of dried blood
Int. J. Biochem.
Recovery of protein and DNA trapped in stone tool microcracks
J. Archaeol. Sci.
Protein analysis of bugas-holding tools: new trends in immunological studies
J. Archaeol. Sci.
Journal of
Archaeol. Sci.
Cellular and Molecular Immunology
The Geology of Qasr Amra
Vegetation of Jordan
Middle and Late Pleistocene landscape evolution at the Druze Marsh site in northeast Jordan: implications for population continuity and hominin dispersal
Geoarchaeology
Equus caballus
Mamm. Species
Blood residues on stone tools: indoor and outdoor experiments
World Archaeol.
The fossil fauna from an upper Pleistocene site in Jordan
J. Zoology, Lond.
A re-consideration of the fossil fauna from C-spring, Azraq
Environment, chronology and lower-Middle paleolithic occupations of the Azraq basin, Jordan
Paléorient
Millennial Landscape Change in Jordan: Geoarchaeology and Cultural Ecology
Precipitin reactions in forensic problems
Nature
Late Pleistocene and Holocene equid remains from Israel
Zoological J. Linn. Soc.
Cited by (26)
New insights into early paleoindian (Gainey) associations with proboscideans and canids in the niagara peninsula, southern ontario, canada
2023, Journal of Archaeological Science: ReportsMiddle to Late Quaternary palaeolandscapes of the central Azraq Basin, Jordan: Deciphering discontinuous records of human-environment dynamics at the arid margin of the Levant
2022, Quaternary InternationalCitation Excerpt :Alongside the 2013–2015 archaeological excavations at SM1 (Nowell et al., 2016; Pokines et al., 2019), our team exposed and documented 13 geological test pits distributed around the margins of the Ayn Sawda depression (Fig. 2). The SM1 stratigraphic sequence was first documented in 2013 and consists of 14 macro-stratigraphic units (Nowell et al., 2016, Fig. 3); the SM2 test pit sequence was also documented in 2013. Ten stratigraphic sequences were examined in 2014, SM3 through SM12, and a further two sequences in 2015: SM13 and SM14 (Supplementary Material, Section S1.1).
The palaeoenvironmental potential of the eastern Jordanian desert basins (Qe'an)
2022, Quaternary InternationalThe Middle to Late Pleistocene transition in the Azraq Oasis, Jordan: A phytolith-based reconstruction of wetland palaeoecology
2022, Palaeogeography, Palaeoclimatology, PalaeoecologyCitation Excerpt :This suite of units (8-7c-7b) contains the Late Acheulean archaeological remains, with isolated finds documented in correlated stratigraphic positions across the western Shishan Marsh (Ames et al., 2021). Zone I is capped by a culturally sterile aeolian silt veneer (unit 7a) dating to MIS 5e (Nowell et al., 2016). Overall, the Zone I sequence represents a shrinking, shallow waterbody whose relative position to an alluvial fan changes across late Middle Pleistocene.
Wonderboompoort, South Africa: A natural game funnel for meat harvesting during the later Acheulean
2021, Journal of Archaeological Science: ReportsTorque creation and force variation along the cutting edges of Acheulean handaxes: implications for tip thinning, resharpening and tranchet flake removals
2020, Journal of Archaeological ScienceCitation Excerpt :For example, some researchers have observed microscopic wear traces along the whole working edges of tools (Keeley, 1980; Ollé et al., 2014; Murray, 2017; Hardy et al., 2018), while others describe only their tips exhibiting heavy wear (Keeley, 1980; Soressi and Hays, 2003; Viallet, 2016; Zupancich et al., 2018). Although rarer, similar observations have been made with residue distributions (Dominguez-Rodrigo et al., 2001; Nowell et al., 2016; Hardy et al., 2018). Morphological attributes on artefacts likewise suggest the preferential use of specific edge portions.