Taphonomy is commonly described as “the study of the environmental conditions affecting the preservation of animal or plant remains”. This broad definition is usually understood to refer to post-depositional processes. One of the main purposes of taphonomic studies is the understanding of possible biases in the fossil record. In this respect, different forms of pre- and post-depositional taphonomy can affect phytolith assemblages that scientists retrieve from sediments and soils. These can be both natural- and human-induced. This paper analyses the forms of possible biases during- and post-depositional, as well as during sampling, recovery and study and proposes some methodological adjustments as to reduce the possibility of errors as well as to check the representativeness of the assemblages retrieved. It is shown that control is possible when human action is involved in the form of attentiveness in sampling and recovery strategies. Pragmatic and statistic methods are proposed so to understand the impact of pre- and post-depositional taphonomic processes on the final phytolithic assemblage. With an increase of stratigraphic control and simple calculation of representativeness it is possible to achieve a confident level of standardisation so to reduce biases at minimum and obtain reliable results with phytolith analysis.

The authors would like to thanks RM Albert, MA Cau and D Cabanes for the invitation to present the current work at an ICREA Conference in 2010. Both authors are part of the AGREST research group (Government of Catalunya) and SimulPast CONSOLIDER-INGENIO2010 project funded by the Ministry of Science and Innovation (MICINN – Government of Spain). CL is part of the AGRIWESTMED Project. This work has been funded by the EU ERC-230561 Project, the SimulPast CD2010-00034 Project and the NoGAP HAR2010-16052 Project

Peer reviewed