Heritage Cases

Diepkloof Rock shelter, located in the Western Cape Province is a key Middle Stone Age (MSA) site that has focused interest of researchers. Indeed, it is one of the few MSA sites that display in stratigraphy both Howiesons Poort and Still Bay assemblages (Porraz et al., 2013a,b). Moreover, numerous engraved ostrich eggshells have been found associated with the Howiesons Poort artefacts, contributing to the issue of early origin of symbolism in human behavior. However, “how much early” has remained a major debate in this particular site. Jacobs et al. (2008, 2015, 2017), Tribolo et al., (2009, 2013) and Feathers (2015) have indeed proposed age estimates using luminescence dating that do not agree. Jacobs et al. (ibid) have suggested, after dating sediments of nine MSA sites from all over South Africa, including Diepkloof, that Howiesons Poort and Still Bay were of short duration, separated by a time gap and had the same age in all sites (60-65 ka and 71-72 ka respectively). However Tribolo et al., (ibid), have suggested by dating both sediments and burnt lithics, that in Diepkloof Rock Shelter, the Still Bay started about 110 ka ago and the Howiesons Poort was of long duration, ending 55 ka ago. The dating of both techno-complexes (their first appearances, their durations, their last expressions) has direct implications for the present demographic, cultural and ecological models that aim clarifying the history of Anatomically Modern Humans in Southern Africa (Porraz et al., 2013a; Conard et al., 2014). The chronological discrepancies observed in Diepkloof have tentatively been explained through methodological biases. In particular, Guérin et al. (2013) have suggested that the uncertainties applied by Jacobs et al., were underestimated. The calibration of the beta counter was also criticized and the model for beta dose rate modeling was claimed incorrect by these authors. Galbraith (2015) and Jacobs (2015, 2017) however have dismissed these arguments. On the other hand, Feathers (2015) has published additional luminescence ages that do not agree either with Jacobs et al. estimates. Consequently, we are facing an endless debates with methodological issues for luminescence dating. In order to overcome this impasse, we believe that another dating method must be used, resting on different physical principles than the luminescence dating, and applied on different kind of material. Therefore, we intend to apply a new U (Uranium)-series dating methodology on shells. Previous U-series methodologies implied full disintegration of the sample for analysis of the U and Th (Thorium) content. However, in this case, late (post-depositional) input of U and Th could not be avoided, leading either to minimum age estimates or to a complete failure of the dating. More recent developments use standard laser ablation system in order to evaluate the U and Th incorporation (e.g. Sharp et al., 2016) but the dating still rest on modeling of this incorporation. The new methodology that has been developed by our team (collaboration between the IRAMAT-CRP2A, Bordeaux, IPREM-LCABIE, Pau and LSCE, Gif-sur-Yvette, France) takes advantage of the femto-laser ablation system build at the Pau University (Donard et al., 2015; Pecheyran et al., 2017). This laser system allows virtual beam shaping by combining a high repetition rate with a fast galvanometer movement. This is of great interest in order to adjust the apparent laser beam shape to the structure of the sample, striae for instance (Fernandez et al., 2007), while common lasers only allow circular craters at low repetition rate: this results in a much better signal sensitivity (with limits of detection as low as a few attograms, (Donard et al., 2015, Aramendia et al., 2015) and a better spatial resolution (Fernandez et al., 2007). Therefore, the femto-laser ablation system available in Pau university allows: 1) mapping of the U and Th content at high resolution and detecting parts of the shells that have not been polluted by late U and Th incorporations, and 2) reaching the level of sensitivity required for analyzing the very low amount of U and Th present in these parts of the sample. In consequence, no more modeling of U input is theoretically necessary. This laser facility is unique in the world: these analyses can therefore only be conducted in France. The excavation of Diepkloof has revealed the presence of OES fragments throughout the sequence but the present study will strictly focus on the HP layers. The ostrich eggshell (n = 18) fragments that we are planning to date and for which we are applying for this export permit are small (< 2 cm2) and do not show any engraving. Besides, such remains are abundant in the studied layers (Texier et al., 2013); in other words, their export and partial destruction for analysis will be of no consequence for the general study of the faunal and cultural material, while the results will have main implications for the understanding of the MSA cultural sequence across Southern Africa.