Phytolith transport in sandy sediment: Experiments and modeling

Phytoliths are minerals of amorphous silicon dioxide forming in cell lumina and intercellular spaces of living plants. Because of their morphological specificity and their stability in soils phytoliths have been widely used in paleoenvironmental reconstructions. However, the mechanisms of phytolith displacement in soil are not well understood, which causes uncertainties in the interpretation of phytolith data. The objectives of the present study were (1) to determine phytolith transport in sandy sediment under different rainfall, and (2) to assess the long-term phytolith displacement with moving water by modeling. Phytoliths extracted from Phragmites australis were added to the upper 1-cm layer of columns packed with medium sand. Two rainfall regimes 1) low-frequency irrigation (40 mm × 2 times per month), and 2) high-frequency irrigation (40 mm × 8 times per month) were simulated over a period of 5 months. At the end of this period, the distribution of phytoliths was examined in the columns. The weighted mean travel distance of phytoliths was 2.7 ± 1.6 mm and 3.7 ± 0.2 mm at the low-frequency and high-frequency treatments, respectively. Under the high-frequency irrigation 22% of the applied phytoliths were leached from the application layer. Two modeling approaches were tested against the experimental data of the high-frequency irrigation treatment. The first model was a convection–dispersion model with attachment, detachment and straining terms. In the second model, reversibility of straining was additionally taken into account. Both models resulted in an equally good agreement with observations. However, the long-term predictions calculated with the two approaches were contradictory. The first model showed no significant translocation after a period of thousand years, whereas the second model predicted that phytoliths would on average be displaced by 19 cm. We hence conjecture that our set of observations was not sufficient to distinguish between two models. Long-term experiments on phytolith transport in undisturbed soil are required to ascertain proper modeling of phytolith transport in a sandy soil.