A wide range of archaeological contexts can be sampled for phytoliths and starch, including sediments from activity areas, house floors, refuse or storage pits, cess deposits, burials and other features, as well as cultural materials such as lithics, pottery, tooth tartar, and coprolites. In all cases, the sampling procedure should be designed to answer specific research questions, rather than being an ad hoc process. Strategies should consider how many samples are needed and which contexts need to be sampled in order to answer the research questions. Sampling strategies will also vary depending on site content and complexity. As a general rule, it is better to collect too many samples in the field rather than too few, as this allows greater flexibility in the laboratory if research goals are revised or if new questions arise.

Sampling strategies

There are several approaches to sampling sediments for microremains, the most common of which are discussed below. One or a combination of these approaches may be adopted depending on the types of data needed to answer your research questions.

Generally, it is advisable to avoid collecting samples from highly disturbed deposits, which may derive from mixed contexts. You will also need to consider what additional samples need to be collected in order to interpret the starches or phytoliths recovered from the context of interest. These controls are necessary for characterising an activity area or feature in relation to surrounding loci, and will also help you determine whether the microfossils are in their primary depositional context or are intrusive. At minimum, sediments adjacent to but outside the locus of interest should be collected for comparative analysis. Always remember that the greater your sampling control, the more robust your final interpretations.

Contextual sampling

This strategy involves collecting samples from specific features such as organic accumulations, ash deposits, pits, cess deposits, fire places, ovens, floor areas or burials. For burials, sediments should be collected from directly beneath the skeleton in the abdominal and pelvic areas, as well as from the matrices surrounding the body. While contextual sampling is useful for gaining insight into specific areas where organic remains, including starch and phytoliths, are very likely to occur, it is recommended that spot sampling of features should not replace a more comprehensive sampling strategy that incorporates all parts of a site (Lennstrom and Hastorf 1995:702).

Horizontal sampling

Horizontal sampling is carried out on a single context with several samples collected to investigate synchronic variability. Samples are usually collected systematically over an excavation area or living surface at regular intervals (e.g. 50 cm or 100 cm grid). This approach is most informative about plant economy and use-activities (e.g. use of space, production areas), as it is most likely to detect spatially discrete starch or phytolith deposits occurring on site.

Vertical/column sampling

This strategy is one of the most commonly used for microremain sampling, and it is has been though to give a diachronic perspective of the phytolith/starch assemblages. Because vertical columns are spatially-confined, however, they usually offer a very restricted view of on-site plant use activities. This strategy may be useful for investigating changes in plant use over time if constant depositional contexts such as middens or refuse pits are analysed (Pearsall 2000:487). Otherwise, differences in starch and/or phytolith assemblages between archaeological horizons should be interpreted very cautiously, as these may reflect differences between the discrete contexts sampled rather than temporal trends in plant use.

Micromorphology and phytolith/starch sampling

This method is probably the most precise approach to sampling microremains. When a micromorphology block is collected, it is good practice to collect a second block side by side. This second block can be 'microsampled' for starch and phytoliths later in the laboratory using the micromorphology thin section as a guide. This approach is very useful for studying ancient floors or stabling areas, for example.

Collecting sediments for starch and phytolith analyses

There are a few simple rules that should be followed to minimise the possibility of contamination during on-site sampling for phytoliths and starch.

Generally, only a small amount of sediment (between 30 and 50 grams) is needed for phytolith or starch analysis, but this can vary depending on the type of site being sampled (e.g. Palaeolithic sites might have lower microfossil concentrations and would therefore require larger amounts of sediment) and if multiple analyses are being performed on the same batch (e.g. microremains, chemical analysis, loss on ignition, etc.). It is also good practice to collect enough material to perform several extractions from the same sediment sample, in case of low recovery rates or laboratory mishaps.

Sampling artefacts and other cultural materials

Artefact sampling strategies may be designed to investigate spatial or temporal patterns in plant processing activities, or differential use of morphological or technological tool types. Artefacts with very weathered or degraded surfaces should be avoided, as they are much less likely to retain use-related residues. Less common remains such as coprolites or teeth with calculus may be sampled opportunistically, although cemetery excavations or mass burials may present the opportunity to investigate differential plant food consumption patterns between social groups (e.g. class, gender) through calculus. If possible, collect several calculus from multiple teeth of an individual to maximise starch and phytolith recovery. Preliminary studies also suggest that calculus from posterior teeth (e.g. molars) may be more productive, as they are involved in food processing more than anterior teeth (e.g. canines, incisors) (Boyadjian et al. 2007).

Artefacts recovered from sieves or held in previously excavated collections (which might have been washed) can also be sampled for phytolith/starch residues. Interpretations will need to account for the possibility that fewer sampling controls were in place, precautions to avoid contamination may not have been taken, and that washing could have removed a proportion of the original residue.

References

• Boyadjian, C.H.C., S. Eggers and K.J. Reinhard 2007. Dental wash: a problematic method for extracting microfossils from teeth. Journal of Archaeological Science 34:1622-1628.
• Fullagar, R.L.K., J. Field, T.P. Denham and C.J. Lentfer 2006. Early and mid Holocene tool-use and processing of taro (Colocasia esulenta), yam (Dioscorea sp.) and other plants at Kuk Swamp in the highlands of Papua New Guinea. Journal of Archaeological Science 33:595-614.
• Kealhofer, L., Torrence, R. and Fullagar, R. 1999. Integrating Phytoliths within Use-Wear/Residue Studies of Stone Tools. Journal of Archaeological Science 26:527-546.
• Lennstrom, H.A. and Hastorf, C.A. 1992. Testing old wives' tales in Palaeoethnobotany: A comparison of bulk and scatter sampling schemes from Pancan, Peru. Journal of Archaeological Science 19:205-229.
• Pearsall, D. M. 2000. Palaeoethnobotany. A Handbook of Procedures. Academic Press, New York.
• Piperno, D. 2006. Phytoliths. A Comprehensive Guide for Archaeologists and Paleoecologists. 238 pp. Lanham, New York, Toronto, Oxford: AltaMira Press (Rowman & Littlefield).