Journal of Archaeological Science: Reports 35 (2021) 102719 2352-409X/© 2020 Elsevier Ltd. All rights reserved. Raw material procurement at Boila Rockshelter, Epirus, as an indicator of hunter-gatherer mobility in Greece during the Late Upper Palaeolithic and Early Mesolithic Paraskevi Elefanti a, * , Gilbert Marshall a , Christos L. Stergiou b , Eleni Kotjabopoulou c a Department of History and Archaeology, National and Kapodistrian University of Athens, Greece b Department of Minerology, Petrology, Economic Geology, School of Geology, Aristotle University of Thessaloniki, Greece c Ephorate of Antiquities in Ioannina, No 6, 25th of March, Ioannina, Greece A R T I C L E INFO Keywords: Palaeolithic Mesolithic Hunter-gatherer mobility Epirus Chipped stone Geochemistry Rare earth elements ABSTRACT Mobility represents one of the most important behavioural strategies of the Palaeolithic and Mesolithic, allowing groups to organise their use of landscapes by positioning themselves close to resources, as well as to regulate group size and social networks. One way of measuring the scale of this mobility is through the transport of lithic raw materials in the form of cores and fnished artefacts. This paper aims to investigate settlement and mobility in Greece during the Late Upper Palaeolithic to Early Mesolithic transition, by using chipped stone technology and raw material transport as an index for wider subsistence and settlement changes associated with the Early Holocene. We take as a case study the Boila Rockshelter, located in the Voidomatis Basin in the rugged uplands of the Tymphi Massif, part of the Pindus Mountains in northwestern Greece. Typological analysis of almost two-thirds of the Boila chipped stone assemblage was carried out, along with geochemical trace element analysis on chert samples from the site, as well as limestone exposures in the local area and further afeld. The geochemical analysis was focused on rare earth and other elements, with the results from Boila and the geological samples then compared. These confrmed that the majority of the chert used at Boila was originally derived from outcrops within the Vikos Gorge and Voidomatis River Valley, with the best pieces from approximately 10 km to the southeast of the site. Abrasion on the surfaces of these indicates that the majority of pieces used at Boila were collected from secondary fuvial deposits, such as those immediately below the site. In contrast, the geochemical results suggested that the reddish-brown varieties which were recovered at low frequency at Boila were probably collected from deposits located further towards the south. Synthesis of the results from the chipped stone and geochemical analyses suggested that during the Early Holocene at Boila, there was a change in the composition of the retouched tool assemblage. Backed bladelets continued to be made, but they were added to by increasing numbers of backed points and geometrics, along with a signifcant increase in the frequency of microburins. In parallel, the use of local black chert increased, albeit from a high level to begin with, while the poorer quality local greyish-white varieties declined, along with reddish-brown. The decline in greyish-white chert probably points to deliberate choice, favouring the better quality local black variety. The decline in reddish-brown chert, likely to have been collected further towards the south, suggests increasingly localised use of the landscape, possibly in response to improving conditions leading to increasing abundance, thus reducing the scale of mobility systems or the need for as frequent site relocation. 1. Introduction Human settlement history during the Final Pleistocene and Early Holocene is one of the most hotly debated issues in Aegean prehistory (Perl` es, 2001; ¨ Ozdo˘ gan, 2011; Horejs et al., 2015; Reingruber, 2017), being closely linked with two current research themes. How did hunter- gatherer societies respond to the changing conditions of the Early Ho- locene and what was the origin of the Neolithic which appeared during the mid-9th millennium? Fundamental to this discussion is the nature of the Mesolithic, the three thousand year period which coincided with the * Corresponding author. E-mail address: paraskevielefanti@gmail.com (P. Elefanti). Contents lists available at ScienceDirect Journal of Archaeological Science: Reports journal homepage: www.elsevier.com/locate/jasrep https://doi.org/10.1016/j.jasrep.2020.102719 Received 1 June 2020; Received in revised form 3 November 2020; Accepted 10 November 2020