Application of spatial statistics to the study of site formation processes

Abstract

Space is an ineluctable reference dimension in Palaeolithic research for monitoring site formation and modification processes. Quantitative, hypothesis-driven analyses of spatial patterns are paramount to understand the dynamics of cultural and natural processes and, therefore, to correctly interpret past human behaviours. However, spatial statistics are seldom used in a taphonomic perspective. This cumulative dissertation aims to contribute to the development of a spatial taphonomic approach to the study of site formation processes. Multilevel multiscale spatial statistics are applied to the taphonomic study of I) the Lower Palaeolithic site of Pirro Nord 13, Italy; II) the Middle Pleistocene open-air site of Marathousa 1, Greece and III) the Late Villafranchian palaeontological locality of Tsiotra Vryssi, Greece. The first study evaluates the reliability of the spatial and stratigraphic association of the lithic artefacts with the faunal remains that are used for the biochronological estimate of the site of Pirro Nord 13. The spatial properties of both biostratinomic and diagenetic processes are investigated. Results of our analyses confirm the stratigraphic integrity of the deposit. In the second study, statistical inference is conducted in order to test two contrasting models of deposition - autochthonous vs. allochthonous. A geoarchaeological approach is adopted for the analysis of the orientation and three-dimensional distribution of the finds. Results of our analyses are consistent with minor transportation and substantial spatio-temporal association of the lithic and faunal assemblages, thus supporting the current interpretation of Marathousa 1 as a butchering site. The third study elaborates on a hitherto little investigated aspect of spatial taphonomy: multilevel anisotropy. Recursive anisotropy is detected both at the level of taphonomic elements and at the assem-blage level at Tsiotra Vryssi. Along with analysis of the differential preservation of skeletal elements and sedimentological observations, the taphonomic history of the site is unravelled. Furthermore, by conducting a spatial taphonomic analysis of a non-human related faunal assemblage, this study contributes to the building of an essential referential framework for the interpretation of human-related fossil assemblages deposited in fluvial environments. Spatial taphonomy complements the traditional archaeological, geoarchaeological and taphonomic approaches. It enhances our understanding of the processes forming archaeological and palaeontological assemblages, with implications for palaeoecological reconstruction, biochronological estimates and the interpretation of past human behaviours.