Zachary Singer, State Terrestrial Archaeologist, Maryland Historical Trust
A primary function of many ceramic vessels is their use as containers for cooking food. Insights into the contents cooked in a pot can be gleaned from analyses of residues preserved on and absorbed within ceramics. Visible residues typically occur as charred encrustations, which are likely the result of a portion of the food being burned during the cooking process. Samples of the charred crust can be removed from the ceramic and then analyzed to learn about the foods that were burned onto the pot during cooking. An example of the types of information that can result from the study of ceramic encrustations is provided by a pilot study of ceramics from the River Farm Site (18AN881).
The River Farm Site is located along the floodplain of the Patuxent River in the Jug Bay Wetlands Sanctuary of Anne Arundel County. The site was first recorded in 1992. Anne Arundel County's Cultural Resources Division and the Lost Towns Project led large-scale investigations of the site between 2015 and 2017. During these excavations, a buried living surface was encountered, which preserved several archaeological features and large ceramic sherds dating to the Middle and Late Woodland time periods.
In 2018, visible encrustations identified on 2 fabric impressed and oyster shell tempered ceramics identified as Late Woodland Townsend ceramics were chosen for study. The PaleoResearch Institute radiocarbon dated the charred encrustations, extracted pollen, starches, diatoms, and phytoliths from the food crusts, and investigated the organic composition of the residues via Fourier Transform Infrared Spectroscopy.
One ceramic sherd with visible encrustations was recovered from a large trash midden, Feature 103 (Figure 1). Carbonized food crust removed from the sherd yielded an AMS radiocarbon date of 865 ± 24 RCYBP (PRI-5875), which has a two-sigma calibrated age
range of AD 1052-1255 or 898-695 cal BP. These dates fit well with other radiometric dates for Late Woodland Townsend wares (ca. A.D. 950 - A.D. 1600).


FTIR analysis of the organic compounds preserved in the charred crust indicated the presence of pectin and cellulose from the cell walls of plants. The encrustation did not yielded pollen nor starch grains. The phytoliths recovered from the food crusts included many palm family phytoliths and a few grass phytoliths. Diatoms and sponge spicules were observed suggesting use of water when cooking.
The analysis of the microbotanical remains and organic residues preserved in the charred
encrustation suggests that this pot was used for boiling food that likely included palmetto, which is a member of the palm family. Presently, the northern distribution of palmetto extends to the Carolinas but not into the Mid - Atlantic. The calibrated radiocarbon dates from the encrustation coincide with the Medieval Warm Period, which dates between A.D. 900-1300 and is a time when global temperatures were warmer than at present. Perhaps the presence of palmetto phytoliths preserved in the food crust on this ceramic reflect a more northerly natural range for palmetto during the Medieval Warm Period and that Townsend ceramic using peoples at River Farm collected and boiled palmetto locally. Alternatively, the palmetto may have been acquired via trade or exchange.
The other ceramic sherd with visible encrustations was recovered from a small trench, Feature 200 (Figure 2). Charred food crust removed from the sherd yielded an AMS radiocarbon date of 772 ± 33 RCYBP (PRI-5877), which has a two-sigma calibrated age range of AD 1219-1284 or 732-666 cal BP. These dates overlap with other radiometric dates for Late Woodland Townsend wares (ca. A.D. 950 - A.D. 1600).


This FTIR signature yielded peaks typical of proteins. No starches were observed. Pollen extracted from this charred food crust indicate the presence of marshelder, ragweed, gum trees, pine, wormwood, grasses, and plants in the goosefoot family, sunflower family, and mustard family. Phytoliths indicative of Zea mays (maize) cobs and glumes were recovered. A single sedge phytolith was found. Diatoms and sponge spicules were also noted.
The combined FTIR, phytolith, and diatom record indicates boiling maize, meat, and possibly sedge roots or tubers in this vessel.
The analysis of charred encrustations from the River Farm Site provided direct evidence of the cooking practices of Late Woodland peoples, revealing both the foods they consumed and the methods they used. Future studies of Woodland period archaeological sites should prioritize the identification of visible encrustations on ceramic sherds during field and laboratory analyses. Sherds with visible encrustations should be minimally handled and not washed to avoid contamination or accidental removal of the crusts. This research highlights the potential of ceramic residue analyses to enhance the study of Woodland period subsistence practices throughout Maryland.
Comments