ABSTRACT
This paper focuses on two archaeological sites in the Kokytos valley in north-western Greece, discovered by the Thesprotia Expedition in 2004. We stress the discrepancies between surface and subsurface assemblages and try to explain these differences. Initially the sites were found and recorded in a surface survey. After this several multiscale datasets were obtained through different methods, e.g. phosphorus sampling, trial excavations and different geophysical techniques. In combining the different datasets it became quite clear that the surface assemblages are biased mainly through erosion and modern landscaping, but also due to so-called ‘walker effects’. Both sites proved to be more extensive in size and have a longer occupational time span when adding multiscale datasets to the surface collections made in the surface survey.
