Towards producing pure phytolith concentrates from plants that are suitable for carbon isotopic analysis

Phytoliths are micrometric particles of amorphous silica that form inside or between the cells of higher plant tissues throughout the life of a plant. Phytolith morphological assemblages extracted from sediments and buried soils are increasingly used as proxies of grassland diversity and tree cover density. When found in significant amounts in archeological sites they can be used for identifying food habits, cultural and agricultural practices. Phytoliths can contain small amounts of C occluded in their structure (phytC). It is generally assumed that the source of this phytC is atmospheric CO2 that was fixed by the plant via photosynthesis. Isotopic analyses of phytoliths (delta C-13, C-14) were thus expected to inform respectively on the photosynthetic pathway or on the age of the mineralized host plants. However recent C-14 analyses of phytC from phytolith concentrates extracted from soils and harvested grasses yielded unexpected C-14 ages of several hundreds to kyr old. These C-14 phytC results raised the question of a possible source of refractory/old soil organic matter component taken up by roots, which can be attached or occluded in phytoliths. Simultaneously these results highlighted the need for setting standardized protocols leading to concentrates entirely devoid of organic residues, as well as for a robust method for checking phytolith purity. The goal of this work was thus to develop protocols for extracting phytoliths from plants, leading to 100% phytolith purity, as required for phytC analyses. Protocol 1 utilizes a multi-step process of dry ashing and acid digestion, while protocol 2 also uses acid digestion as well as a separate alkali immersion step which removes surface layers. Phytolith concentrate purity was gauged in a semi-quantitative fashion through the use of SEM-EDS analysis. This quality check for phytolith purity can reveal small C particulate contamination of phytolith concentrates that may considerably bias isotopic and quantitative analyses of phytC. Results indicate that the two protocols were able to entirely remove small C particulate contamination. Protocol 1 produced phytolith concentrates with well defined morphologies suitable for both morphological and isotopic analyses. However measurement of C yields showed that protocol 1 probably induced C leakage, leading to lower recovery. Protocol 2 is faster, leads to higher C yield but may lead to a beginning of dissolution. With these protocols on hand, sources of phytC can be properly investigated. (C) 2013 Elsevier B.V. All rights reserved.