Contents lists available at ScienceDirect Journal of Archaeological Science: Reports journal homepage: www.elsevier.com/locate/jasrep Provisioning on the periphery: Middle Palaeolithic raw material supply strategies on the outer edge of a territory at La Combette (France) Lucy Wilson a, ⁎ , Constance L. Browne a , Pierre-Jean Texier b a Department of Biological Sciences, University of New Brunswick in Saint John, P.O. Box 5050, 100 Tucker Park Road, Saint John, N.B. E2L 4L5, Canada b Aix-Marseille Univ, CNRS, Minist Culture, UMR 7269- LAMPEA, 5 rue du château de l'Horloge – BP 647, F-13094 Aix-en-Provence Cedex2, France ARTICLE INFO Keywords: Lithic provisioning Middle Palaeolithic Flint sources La Combette Bau de l'Aubesier Resource selection ABSTRACT We apply a resource selection model to evaluate the importance of factors influencing raw material provisioning at a Middle Palaeolithic site, La Combette, located on the edge of the territory from which its raw materials came. The model was previously applied to another Middle Palaeolithic site, the Bau de l'Aubesier, which is more centrally located in the same region. La Combette is smaller and in easier terrain, and was used for short-term occupations such as hunting camps, so provisioning would therefore be expected to be strongly influenced by the quality of the raw material, and by the distance to the sources. In addition, the sourcing data for La Combette are less detailed than those available for the Bau, so this tests our model in a less-than-ideal case. The results show that our approach does work under these conditions. At La Combette, variables describing the terrain (distance from the source, extent of the source area, difficulty of the terrain, etc.) are more important than those describing the raw material (quality and size of nodules available). Even in relatively easy terrain, distance is more im- portant than quality when using a site on the periphery. 1. Introduction It is well known that stone tools provide a large part of the in- formation about Palaeolithic lifeways, and that this information comes from both their characteristics as tools (typology, technology, function), and as rocks (petrographic nature, geologic source); see for instance Wilson (2007a, and references therein) for a discussion of the inter- pretation of the choices of which sources were used for raw materials, and the fragmentation of their chaîne opératoire across space. In recent publications (Browne and Wilson, 2011, 2013; Wilson and Browne, 2014), we described the development and application of a new method to investigate the factors that influenced prehistoric choices of which lithic raw materials to use, based on consideration of both the terrain (size of the source, difficulty of access, etc.) and the raw material (quality for knapping, size of nodules or pieces available at the source). That work concerned a Middle Palaeolithic site called the Bau de l'Aubesier (hereinafter called the Bau), located in the Vaucluse, southern France (Fig. 1), and used extensive databases of the lithic types in the site's layer assemblages, as well as the locations and characteristics of over 350 potential lithic sources in the region, all compiled by LW over a period of more than twenty years. We use techniques adapted from wildlife ecology (Boyce and McDonald, 1999; Boyce et al., 2002; Browne, 2010; Browne and Paszkowski, 2014; Johnson et al., 2006; Manly et al., 2002) in order to evaluate the significance and the relative importance of various po- tential factors, and ultimately to find the combination of factors which best accounts for the proportions of raw materials found in the site assemblages. In this way, we discovered that at the Bau, although the same overall set of factors was consistently significant in each layer, in the lower layers the subset of characteristics that describes the sources and the terrain around them was relatively more important, while in the upper layers the characteristics of the raw material itself became more important (Wilson and Browne, 2014). Our work can be described as subscribing to a human behavioural ecology perspective, whereby humans are seen as subject to physical evolutionary pressures (Winterhalder and Smith, 2000; Bird and O'Connell, 2006), and we believe that statistical and mathematical approaches (e.g. Surovell, 2009) can help us reconstruct these pressures and the responses to them. We do not discount the cultural influences on humans, including prehistoric hominins, but we are primarily at- tempting to determine how their lives were shaped by the exigencies of their environment and their needs for food, tools, shelter, etc. We do not assume any particular model of lithic procurement, whether em- bedded (Binford, 1977, 1980, 1989) or otherwise. We also do not in- corporate any data about other resources which might have been available within a lithic source area, which would make that source https://doi.org/10.1016/j.jasrep.2018.07.001 Received 1 March 2018; Received in revised form 15 June 2018; Accepted 2 July 2018 ⁎ Corresponding author. E-mail addresses: lwilson@unbsj.ca, cbrowne@unb.ca (L. Wilson). Journal of Archaeological Science: Reports 21 (2018) 87–98 2352-409X/ © 2018 Elsevier Ltd. All rights reserved. T