Phytolith Analysis: Applications of a New Paleoethnobotanical Technique in Archeology

Abstract

The purpose of this paper is to acquaint pro-fessionals not directly involved in soil phytolith research with its applications in archeology. The potential and limitations of the technique, nature of the remains, and basic procedures for field sampling and analysis will be discussed. Both current research applying phytolith analy-sis in New World archeology and topics requir-ing basic research will be reviewed. Study of phytoliths (opal silica bodies) in the soil of archeological sites is a new component of paleoethnobotanical research in the Americas. Although use of this technique to identify cul-tivated grasses archeologically has a consider-able history in the Old World (Neotolitzky 1900, 1914; Schellenberg 1908; Edman and Soderberg 1929; Watanabe 1968, 1970), application of phytolith analysis to detection of New World crops dates from the 1960s with investigations at the Kotosh site in Peru (Matsulani 1972). It was not until the 1970s and publication of Rovner's (1971) stimulating article on the potential of phytolith analysis in archeology that interest in the technique began to grow. Subsequently, phytolith analysis of archeo-logical soils has increased in the Americas (e.g., Starna et al. 1980). Growing interest in this application of phytolith analysis has led to refinements in processing procedures, classification and identification techniques, and a broadening of problems ap-proached. Increasing interest among paleoeth-nobotanists and archeologists in phytolith anal-ysis must be tempered with caution, however, because considerable basic research remains to be done. Topics of needed research include movement of phytoliths in soil, development of classfication schemes for regional vegetation formations, investigation of sources of bias in archeological phytolith assemblages, and con-tinued work on classfication and counting pro-cedures. Although archeological gains from phytolith analysis are most striking from contexts where preservation of macrobotanical remains or pol-len is poor, the technique is strongest when ap-plied as one component of a paleoethnobotan-ical study. Because phytoliths are subject to dif-ferent preservation and deposition biases than macrobotanical remains or pollen, phytolith data can complement other data sources. At the same time, some overlap of phytolith and pollen or macrobotanical data can increase confidence in phytolith interpretations. NATURE AND ANALYSIS OF PHYTOLITHS