Multi-method dating reveals 200 ka of Middle Palaeolithic occupation at Maras rock shelter, Rhône Valley, France

Abstract

The emergence of the Middle Palaeolithic, and its variability over time and space are key questions in the field of prehistoric archaeology. Many sites have been documented in the south-eastern margins of the Massif central and the middle Rhône valley, a migration path that connects Northern Europe with the Mediterranean. Well-dated, long stratigraphic sequences are essential to understand Neanderthals dynamics and demise, and potential interactions with Homo sapiens in the area, such as the one displayed at the Maras rock shelter ("Abri du Maras"). The site is characterised by exceptional preservation of archaeological remains, including bones dated using radiocarbon (¹⁴C) and teeth using electron spin resonance combined with uranium series (ESR/U-series). Optically stimulated luminescence was used to date the sedimentary deposits. By combining the new ages with previous ones using Bayesian modelling, we are able to clarify the occupation time over a period spanning 200,000 years. Between ca. 250 and 40 ka, the site has been used as a long-term residence by Neanderthals, specifically during three interglacial periods: first during marine isotopic stage (MIS) 7, between 247 ± 34 and 223 ± 33 ka, and then recurrently during MIS 5 (between 127 ± 17 and 90 ± 9 ka) and MIS 3 (up to 39,280 cal BP).

Keywords: Bayesian modelling; Chronology; Electron spin resonance; Luminescence; Neanderthal; Radiocarbon; Uranium-series; ZooMS.

Figure 1

Location of Middle and Late Pleistocene sites in the middle Rhône Valley, France (a, modified from), view of the rock shelter (b) and of the stratigraphy (c).

Figure 2

Stratigraphic sequence of Maras rock shelter, with the previous dating results displayed in Table 1 on the left (regular: individual age; bold: weighted mean age; italic: maximum age) and the location of the new samples dated in this study on the right. Note that layer 4 is subdivided into two phases of human occupation (levels 4.1 and 4.2) and that levels 5.1, 5.2 and 5.3 are phases of human occupation of layer 5 upper, present only in the eastern part of the site and thus not represented here.

Figure 3

Location and ages (1σ) of the OSL samples (circles) and of the ESR/U-series sample (hexagon) in layers 6 (orange), and 5 upper (blue), and levels 5.2 (white), 5.1 (yellow) and the transition between 5.1 and 4.2 (green).

Figure 4

Dose response curves obtained on quartz (a: AM-01) and on hydroxyapatite (b: AM-08), of 124 ± 6 Gy and 115 ± 3 Gy respectively.

Figure 5

(A) Events model: the 9 Events each contain between 1 and 9 dates which are assumed to be contemporaneous within each Event; (B) Phases model: the 8 Phases contain 1–2 Events. Stratigraphic relationships are represented by arrows going from the oldest to the most recent phase.

Figure 6

Chronological phasing of the Maras rock shelter: a posteriori date distribution of the beginning and end of each of the 8 phases characterised. The time scale used is cal. BP (before 1950). For phases with only one Event, the start and end distributions are identical. Only phase “5 upper” has two different start and end distributions because it contains two events.

Figure 7

Synthesis of the dating results (1σ) obtained at Abri du Maras. APM (a posteriori mode) for each event are represented, as well as corresponding MIS (marine isotopic stage). Note that due to the scale, the error bars of the radiocarbon dates are not visible, and only the youngest U-series date obtained for layer 5 upper can be represented and thus the APM (388 227 years) as well as the corresponding event (1) are not visible on the figure.