Abstract:
The scarcity of human remains from Pleistocene Northern Asia is a significant limiting factor in understanding the course of human evolution in this part of the world. The high degree of fragmentation of bones in Siberia has limited our ability to identify and better understand the hominin groups present in the region between 300 – 50 ka (thousand years ago). In this dissertation, I address the dearth of human fossils in the region by screening thousands of non-diagnostic bone fragments using a peptide mass fingerprinting method known as Zooarchaeology by Mass Spectrometry (ZooMS). ZooMS allows for efficient and reliable taxonomic identification of bone collagen and here I apply it to six sites from Middle and Late Pleistocene contexts. In total, 8,921 unidentified bone fragments were analysed using ZooMS and resulted in the discovery of nine new hominin fragments including the oldest known Denisovan fossils (Denisova 18, 19, 20, 21), two Middle Palaeolithic Neanderthals (Denisova 15 and Denisova 17), two hominins linked with Upper Palaeolithic assemblages (Denisova 14 and Denisova 16), and one which had been previously misidentified as a bear parietal fragment (Denisova 7).
Parallel to the search for human fossils, large assemblages of faunal remains were also identified. These were compared against the zooarchaeological record of Denisova Cave which was established previously using traditional methods. The comparison helped identify the potential biological, depositional, and taphonomic factors responsible for the accumulation and fragmentation of the bones at the site. Bones identified using ZooMS from Denisova Cave were also used to create isotopic baselines, essential for studying Pleistocene hominin diet in the Altai Mountains. This revealed instances of extreme isotopic variation especially with regards to some of the humans, leading to the identification of the highest C and N isotope values currently reported for Neanderthals.
Finally, a subset of bones (2,501) were used to map glutamine deamidation and test the efficacy of this technique in understanding protein degradation as a function of time and as means of assessing the relative age of the hominin fossils. This technique has been used recently to test the antiquity of unprovenanced fossils, however, our results revealed that too much variability is observed even within a single layer to be able to securely identify intrusive or unprovenanced fossils.
This work is the largest application of ZooMS on an archaeological assemblage and the results confirm our hypothesis that bone assemblages which are routinely excavated from Pleistocene-age sites, in Siberia and elsewhere, present an important untapped source of archaeological information. The data reported in this dissertation stem solely from fragmented bones that could not be identified on the basis of morphology. High throughput analysis of such material has the potential for groundbreaking discoveries and should become a routine component of the archaeological tool kit.
Print-on-Demand