The archaeological deposits at Mumba rockshelter, northern Tanzania, have been excavated for more than 70 years, starting with Margit and Ludwig Köhl-Larsen in the 1930s. The assemblages of Middle Stone Age (MSA) and Later Stone Age (LSA) artefacts collected from this site constitute the type sequences for these cultural phases in East Africa. Despite its archaeological importance, however, the chronology of the site is poorly constrained, despite the application since the 1980s of several dating methods (radiocarbon, uranium-series and amino acid racemisation) to a variety of materials recovered from the deposits. Here, we review these previous chronologies for Mumba and report new ages obtained from optically stimulated luminescence (OSL) and infrared stimulated luminescence (IRSL) measurements on single grains of quartz and multi-grain aliquots of potassium (K) feldspar from the MSA and LSA deposits. Measurements of single grains of quartz allowed the rejection of unrepresentative grains and the application of appropriate statistical models to obtain the most reliable age estimates, while measurements of K-feldspars allowed the chronology to be extended to older deposits. The seven quartz ages and four K-feldspar ages provide improved temporal constraints on the archaeological sequence at Mumba. The deposits associated with the latest Kisele Industry (Bed VI-A) and the earliest Mumba Industry (Bed V) are dated to 63.4 ± 5.7 and 56.9 ± 4.8 ka (thousands of years ago), respectively, thus constraining the time of transition between these two archaeological phases to ∼60 ka. An age of 49.1 ± 4.3 ka has been obtained for the latest deposits associated with the Mumba Industry, which show no evidence for post-depositional mixing and contain ostrich eggshell (OES) beads and abundant microlithics. The Nasera Industry deposits (Bed III) contain large quantities of OES beads and date to 36.8 ± 3.4 ka. We compare the luminescence ages with the previous chronologies for Mumba, and briefly discuss how the revised chronology fits in the context of existing archaeological records and palaeoclimatic reconstructions for East Africa.
Read full abstract