Old taphonomy issues, new charcoal data for Mesolithic contexts: Impact of fragment size and sampling context on the assemblages of Escabasses cave (SW France)

Abstract

Does fragment size influence significantly archaeological charcoal spectra? Does the floristic content of charcoal concentrations provide a representative image of the anthracological record? These taphonomy-related questions are as old as the first methodological groundwork in anthracology and therefore, they could be considered as solved. The aim of this paper is to show that they are still topical today, as depositional and post-depositional conditions vary between archaeological contexts, periods, and sites. With this in mind, we revisited the charcoal data of an Early Mesolithic occupation, level 6 of Escabasses cave, which had the particularity of yielding a significant amount of charcoal fragments over 4 mm size and numerous combustion features. This allowed to test the impact of different size classes (mainly ]2-4[ mm vs. > 4 mm) and types of deposits (concentrated vs. scattered charcoal) on taxonomic diversity and proportions, as well as on the abundance of a key-taxon, deciduous oak (Quercus). Our results show that (i) contrarily to what is generally assumed, at Escabasses the proportions between taxa showed more variability according to size class than to types of deposit; (ii) size class doesid not impact floristic diversity; (iii) however, the probability to find higher Quercus frequencies within the ]2-4[ mm size class is significant. These results allude to different stages in the taphonomy history of the charcoals, potentially ranging from the type of occupation and related human practices to a specific postdepositional behavior of Quercus, suggested by previous studies.

A vegetation record based on charcoal analysis from Border Cave, KwaZulu-Natal, South Africa, ∼227 000 to ∼44 000 years ago

“…Consequently, specimen sizes vary between members. This is important because a study of the taphonomy in Escabasses Cave (southwestern France) showed that size class does not affect interpretations of floristic diversity (Henry et al, 2020). However, earlier studies by Chabal (1991Chabal ( ,1992Chabal ( , 1997 4.…”

“…Approximately 80 specimens of charcoal (Stewart et al, 2016) were analysed from two layers from each member, except from Members 5 BS and 4 WA where more layers were examined (Table 1). Standard anthracological methods were applied to identify the charcoal (see for example Figueiral and Mosbrugger, 2000;Allott, 2006;Théry-Parisot, 2010;Cartwright, 2013;Backwell et al, 2014;Bamford, 2015a, b;Stewart et al, 2016;King et al, 2018;Bodin et al, 2020;Henry et al, 2020). Border Cave specimens were selected for identification by means of random sampling within each sample using random number tables (Drennan, 2009).…”

Sampling and quantitative analysis methods in anthracology from archaeological contexts: Achievements and prospects

“…If fragment size class sub-sampling is carried out only on certain size fractions from a given sample (i.e., only from the 2-4 mm fraction) these results are also likely to be biased. Subsequent comparisons of their taxon frequency spectra to those contained in the > 4 mm fraction may also reveal differences in sample composition (and the proportions of individual taxa) (e.g., Henry et al, 2020). This is because in the 2-4 mm fraction the absence of fragments >4 mm means that the most frequent taxa in the assemblage, which are more common in the larger size classes, will be under-represented.…”

First insights into the wood management for the production of lime given by the anthracological study of a 19th century lime kiln from Martinique, Lesser Antilles

“…Another explanation for their dominance could be a differential preservation leading to an over-fragmentation of charcoals. Several studies have shown that post-depositional processes affect charcoal preservation and can induce a bias in the proportions between taxa (Lancelotti et al 2010;Théry-Parisot et al 2010;Chrzazvez et al 2014;Dussol et al 2017;Henry et al 2020). To assess the likelihood of the hypothesis of a differential preservation, we would need this kind of data for each of our taxa, which would require time-consuming experiments.…”