About Phytoliths

by Steve Archer

What is a phytolith?

“Phytolith” comes from Greek roots phyto (plant) and lithos (stone). “Plant stones.” More technically, they are silica casts of plant cells or spaces between cells in plants. Silica, dissolved in groundwater as monosilicic acid, is taken up into the plant during its life. The silica is deposited in the plant as opal silica, chemically identical to the “opal” you know as a semi-precious stone. These tiny opal particles take on the shape of cells, or “negative” casts of the space between cells. Hairs, prickles, and other surface structures of plants also often become phytoliths.

Who discovered phytoliths?

Interestingly, one of the first people to observe and describe phytoliths under the microscope was Charles Darwin during the voyage of the Beagle—where he began to form his theories of natural selection and evolution. He noted the phytoliths while looking under the microscope at dust blown aboard ship from storms. However, his interest stopped with simple description.

Archaeological applications of phytoliths began in 1971 with a paper published by Irwin Rovner, now at North Carolina State University. Major phytolith discoveries came in the late 1970s with Deborah Pearsall using phytoliths to document early maize agriculture in Ecuador.

How do the phytoliths get in the ground?

Phytoliths are released from the rest of the plant’s structure where the plant decays or is burned. This can be either where a plant naturally died and decayed, where humans have disposed of plant residue, or even where animals have eaten plants and deposited their dung, as phytoliths survive the digestive process.

Do all plants have phytoliths?

No. Phytolith production in plants follows evolutionary groupings. Some plants, particularly grasses and other monocots produce many phytoliths, and in fact use them as part of their support structure. Other plants do not produce phytoliths at all. One plant family of particular interest in Virginia, the Solanaceae, does not produce phytoliths. This means that plants in this family, such as tomatoes, potatoes, and, importantly, tobacco, cannot be seen archaeologically using phytoliths.

However, some plant families, such as the grasses (Poaceae) have phytoliths even more diagnostic, or taxonomically specific, than seeds or pollen. This is important because particular grasses have different growth environments, and different geographic origins. This is highly useful in understanding a site’s environment and landscape.

Why study phytoliths?

Phytolith analysis is only one component of a broader sub-field of archaeology called paleoethnobotany or archaeobotany, the study of plant remains in archaeology. Prior to the twentieth century, most material culture was made of plants. Think about a world without construction wood, plant foods, plant medicines, wood furniture, trees, ornamental plants—and you can begin to see the importance of looking at plants to understand the past. Other types of archaeobotanical studies include macrobotanical analysis (the study of charred plant materials such as wood and seeds recovered from flotation), palynology (the study of pollen and spores), and a variety of other microscopic and chemical specialties. At Colonial Williamsburg, we do macrobotanical and phytolith work in-house, and have done other kinds of specialized archaeobotanical studies in conjunction with researchers elsewhere.

Each archaeobotanical “data set” gives slightly different information, and the best understanding of the past is obtained when more than one type of data is used.

Phytoliths are of particular interest to archaeologists because they are highly resistant to decay, more so than most other types of archaeological plant remains. Because they are essentially stone, they preserve indefinitely in all but the most extremely alkaline burial conditions.

Are phytoliths the same as pollen? I’ve read about pollen studies in archaeology.

No. Phytoliths are also microscopic plant remains, but the similarity ends there. Pollen studies have a much longer tradition in archaeological research, and also tell you different things about archaeological sites and past environments. Pollen grains are the male reproductive gametes that fertilize female flowers. They have a tough outer structure that permits them to survive for long periods of time in certain depositional settings. Pollen tends to preserve very poorly on most archaeological sites, and it is often difficult to understand its “formation process” (i.e., how did this pollen get into this deposit, and does it represent a plant used during the actual occupation of the site). Many palynologists tend to focus on stratified lake sediments to understand environmental change over a long period of time. Phytoliths can also be used in this manner, and scientists are beginning to look at phytoliths and pollen together in lake and other natural deposits.

We don’t currently have the capacity to extract or interpret pollen in-house, and that work is generally done in collaboration with other researchers.

Can phytoliths tell exactly what plant or kinds of plants were used in a garden feature?

Short answer, no, with a few exceptions. One thing that is difficult to comprehend—until you’ve spent some time with it—is the sheer complexity of archaeological deposits. Today there are virtually limitless kinds of tests and extractions that generate “data” from “dirt.” We now have the potential to do more “data production” with a cup of soil than an archaeologist in the 1950s could with an entire site. The trick is making good choices in research design, and then figuring out what it all means!

Remember that not all plants produce phytoliths, first of all. That eliminates many possibilities. Second, think about how a garden is actually used. Soil is moved around and brought in from other places. Plants are continuously harvested and planted, with different plants in the same location in different seasons or years. Some, like pumpkin vines, might be plowed into the soil before a next planting. Others, like carrots, might be completely removed from the soil to consume, leaving no trace. Then, think about what goes into garden soil. Fertilizer, which may contain animal manure—and consequently phytolith traces of all the plants the animal ate! Or yard compost containing many other plants from different areas of the site. Now compound these processes over twenty or a hundred years. What plant grew in that soil? Probably many, and we will get traces of some of them, plus traces of all these other kinds of plants. It is important to point out that archaeological sites are dynamic rather than static, and the problem of “what plant grew there” is usually a problem with the question, rather than the phytoliths.

That being said, some plants that leave very distinct and numerous traces, like maize (corn), for example, often give a strong suggestion that a certain area may have been used to grow corn (or squash, etc.), especially if that use remained relatively unchanged for long periods of time.

So, what use are phytoliths if they can’t exactly reconstruct a garden?

We interpret phytoliths as assemblages and in comparison with other samples from the same or different sites. Looking at the overall proportions of phytolith types within a sample (we generally count hundreds of phytoliths from each sample)—we can ask different, and broader kinds of questions. How was this yard space used? How did the site environment change over time? Where did people process their corn? Where was an orchard located? Was this lawn maintained, or was it allowed to go to pasture? What areas of the site were used repeatedly for the same purpose, and what areas changed over time?

It is important to develop our questions in conjunction with both what we know, and what we want to know about a site or group of sites. What is our understanding of the level of disturbance of a site or deposit? If the site is primarily industrial, should we ask foodways questions of this site?

These questions are more about how people interacted with their landscape, and in many ways are more significant and meaningful than “did this garden contain bouncing bet, or johnny jump-up?” As with other artifacts, good phytolith questions further archaeological interpretation—“what was life like in the past?”—rather than being an end in and of itself.

How big are phytoliths?

Phytoliths range between 5 and 200 microns, most being around 10-30 microns. (A micron is 1/1000th of a millimeter.) A human hair, for comparison, is approximately 100 microns in width.

Can I see phytoliths with the naked eye?

No.We use a transmitted light microscope, with magnification between 100X and 1000X to see them. But, you certainly have felt phytoliths at one point or another. If you’ve ever put a blade of grass between your lips and felt the roughness, or cut your bare arms walking through a cornfield, that is due to the silica phytoliths in these plants.

How do you get phytoliths out of soil?

The short answer is, through a combination of sieving with water, gravity sedimentation, acid organic removal, and heavy liquid flotation. The whole process from raw soil to phytolith extract can take up to a month to complete.

In slightly more detail, this is the process:

1. Excavators in the field take small bags of soil from many contexts on a site. The archaeologists work together to decide on a research plan for the site, and how the phytolith analysis may contribute to the overall research goals.

    We use about 80 grams of soil, usually this amounts to about a half cup. The raw soil is first deflocculated. That is, it is stirred in water over several days with detergent (we prefer Cascade) to break up the lumps and aggregated soil. The goal is to have all of the individual soil particles ionically repelling each other to facilitate the next step.

2. The soil is then sieved with water using very fine-meshed screens of 250 and 53 microns. All soil larger than 250 microns (phytolith sized) is discarded. Soil between the sizes of 53 and 250 microns (the “C fraction”) is saved for special studies of larger sized phytoliths.

3. Dissolved soil in water that is smaller than 53 microns is collected and allowed to settle. At this point, the material larger than phytoliths is gone, and we need to remove clay particles smaller than phytoliths (less than 2 microns) that impair phytolith extraction and viewing.

4. To remove clays, the sediment, dissolved in water, is allowed to settle in an 8 cm column of water for one hour. At this point, clay particles are still in suspension, while phytolith-sized particles have settled to the bottom. The water containing clays is poured off, saving the settled soil at the bottom of the beaker. This process is repeated many times until the supernatant water is clear.

5. The soil now has particles both larger and and smaller than phytoliths removed. At this point, organic chemicals (humic colloids) in the soil need to be removed from the sediment. Organic removal is accomplished by heating the soil in a combination of nitric acid and potassium chlorate, sometimes called Schulze solution. This is done in the fume hood on a hot plate.

6. The sediment is then rinsed clean of the acid, using the centrifuge and fresh water.

7. Phytoliths have a slightly lower specific gravity than other components of the remaining soil, such as quartz particles. To separate the phytoliths from the sediment, we prepare a heavy specific gravity liquid using a chemical called sodium polytungstate. The process is like making a “black and tan” beer, or the separation of oil and vinegar in a salad dressing cruet. By adding the sodium polytungstate solution to the test tube and centrifuging, the phytoliths float on top of this liquid, while other minerals remain in the bottom of the test tube. The phytoliths are collected using a pipette (like an eyedropper), and placed into new test tubes.

8. The phytoliths are rinsed clean of any remaining chemicals using water and acetone, and are then ready to be put onto a slide and counted.

How are phytoliths counted?

An analyst systematically counts the slide by moving the field of view across the microscope slide in lines (transects). He or she notes, counts and measures different known types of phytoliths, as well as recording new unknown phytoliths with drawings and photographs. We aim to count a statistically useful number of known phytolith forms, usually 200 per slide.

How are phytoliths identified?

In short, how do you know what you’re looking at? Phytolith research has been a component of archaeology since 1971. Botanists also have been documenting phytoliths in various plant families, so much is known about certain forms that are seen over and over again. In particular, different sub-families of grasses are well studied, as are some major crops like maize and cereals. These have been thoroughly documented in scientific literature. (Search the collection on-line)

Our second component to identifying phytoliths is developing and maintaining a phytolith reference collection of plants likely to be found in our archaeological samples. Using herbarium and other known specimens, these plants are processed to see if they produce phytoliths. We maintain a computer database of over 1000 specimens, including microsocope photographs of phytoliths found in them. This is the only mid-Atlantic historical archaeology phytolith research collection.

How do you interpret phytoliths?

Phytoliths in one sense are like any other kind of artifact. They do not speak for themselves. Each phytolith analysis is a project with a research design. That is, we choose specific contexts on the site to analyze, because we are interested in a particular problem. You don’t just “do” phytolith analysis to see what it says, just as a list of artifacts alone is not informative about a site. Examples of questions we’ve addressed with phytoliths are: How was the Tucker garden managed and maintained? Are there traces of seventeenth-century landscapes still visible in plowzone at the Rich Neck Plantation? How do a series of fills inform us about the landscape of the Wray site during and after its occupation?

Creating a story based on phytolith evidence entails knowledge of both the archaeology of a site and the botanical aspects of the plants producing phytoliths.

As with all other aspects of the archaeological program here, the more analysis we do, the more refined and better our interpretations become, by having the ability to make more comparisons, as well as clarifying unidentified phytolith forms.

What places has the phytolith lab done work on?

In and around Williamsburg: St. George Tucker Garden; Carter’s Grove; Rich Neck Plantation; Yorktown; Jamestown. Elsewhere on the East Coast: Monticello; Charleston, South Carolina; Thomas Jefferson’s home at Poplar Forest.

Who started our phytolith lab?

In 1995, the Department of Archaeological Research funded the creation of the phytolith lab, under the direction of Dr. Lisa Kealhofer. Dr. Kealhofer took a position at Santa Clara University in 1999. Steve Archer, a Ph.D. candidate at the University of California at Berkeley, is currently in charge of the phytolith laboratory.

Steve Archer is a research associate in the Department of Archaeological Research. This paper was written in 2004 to aid students and volunteers working in the Phytolith Lab.