Phytolith Analysis — Plant Silica Body Analysis

Both recover microscopic plant remains from sediment, but they record different signals. Pollen grains are produced in large quantities and dispersed by wind or insects, so pollen assemblages often reflect regional vegetation including trees and shrubs. Phytoliths are produced within individual plant cells and typically represent local vegetation, particularly grasses and herbaceous taxa. In many soil types — especially well-drained, acidic, or arid soils — pollen does not preserve but phytoliths survive, making the two methods complementary rather than interchangeable.

For Poaceae (grasses and cereals), morphotype resolution can reach genus or sometimes species level. Wheat, barley, rice, and maize produce characteristic husk phytoliths that can be distinguished from wild grass morphotypes with appropriate reference collections. For most non-grass crops (legumes, root crops, many fruits), phytolith evidence is limited or absent because these plants do not accumulate silica at diagnostic levels.

For reliable percentage estimates of morphotype assemblages, a minimum of 200–300 phytoliths per sample is the widely cited threshold. For concentration-based analyses (phytoliths per gram of sediment) the minimum count may be lower, but assemblage composition estimates remain unreliable below 200. Samples yielding very low phytolith concentrations — common in highly leached or coarse sandy sediments — may not meet this threshold regardless of sample volume.

A regional reference collection of modern plant phytoliths is strongly recommended. Without it, identification is limited to broadly recognised morphotypes (e.g., grass short cells) and taxon-level inferences are unreliable. Published morphotype atlases and online databases such as the Global Pollen Database (which also hosts phytolith data) and regional herbarium-based collections can partially substitute when assembling a new reference is not feasible.