Hi, I’m Jennifer, and I’m a phytolith analyst. The term ‘phytolith’ comes from Greek and means ‘plant stone’. It is used to describe microscopic casts of plant cells that are formed from minerals. Usually ‘phytoliths’ refer to silica plant cell casts that are formed when a plant takes up silica in the form of monosilicic acid with the groundwater. This silica enters the groundwater when rocks like feldspar and quartz erode. Other than oxygen, silicon is the most abundant element in the earth’s crust, so the supply of soluble silica in the groundwater is constantly replenished. Once inside the plant, the monosilicic acid is deposited as solid silica, making phytoliths. We are still not entirely sure why plants make phytoliths: it could be defensive, which happens in hairs; it could be for structure, which is why there are phytoliths in fruit rind; or it could be accidental, because the plant has no way to get rid of the silica once it has taken it up with the groundwater. It is probably a combination of these reasons

Figure 1: Millet husk silica skeleton from archaeological sample from Masudpur I

Phytolith identification is based on shape, features and surface texture. Under certain conditions, potentially linked with irrigation, large collections of phytoliths survive in their original anatomical structure. These ‘silica skeletons’ can provide extra information allowing for more specific identification (Figure 1). Unlike macrobotanical remains (see Cemre’s blog post 1st March), phytolith morphology tends to be linked more with genus than species (although it is possible to link phytoliths to species in some cases). We can, however, use phytoliths to work out which parts of the plant are present, because cellular structures vary depending on the organ in the plant. Therefore, a leaf phytolith will look different from a husk phytolith because it has a different function in the plant. A grass leaf phytolith will look different from a palm leaf phytolith (Figure 2) because they are very different plants. It is more difficult to distinguish different grass leaves, as the cell function in grass is often the same, but differences may be seen between arid, tropical or temperate grasses (Figure 3), and in crop plants the husks often show some variation, making them the most useful identification feature.

Figure 2: Date palm leaf phytoliths from archaeological sample from Masudpur I

The kinds of questions we can ask of phytoliths are a bit different from the ones we ask of macrobotanical remains like seeds. Phytoliths are not usually identifiable to species, so instead of asking what crop species were people growing, we might use the different parts of plants that are represented to ask, how people were processing these plants. Luckily we work in a region where the broad cereal groups (wheat/barley, millets and rice) have very distinct phytoliths! We can look at the different parts of the cereal plants, their weeds and think about the stages of processing carried out, and whether this changed as social demands for food and surpluses changed.

Figure 3: Grass leaf short cell phytoliths from various archaeological samples. Top left and bottom right: bilobates; Top right: cross; Bottom left: saddle

At the moment I am very interested in rice, as this is a crop we know was grown in the TwoRains study area from the macrobotanical remains, because we have found both rice grains and spikelet bases. Rice is a big phytolith producer (Figure 4), but we have not yet found any rice phytoliths in our samples, and I am trying to work out why it is not there. In the field season I will be collecting modern and archaeological samples to try to solve this problem!

Figure 4: Left: rice leaf bulliform from archaeological sample from Masudpur I; Right: rice husk phytoliths from modern reference collection in the George Pitt Rivers Lab., McDonald Institute for Archaeological Research, University of Cambridge.