Ancient use of the four corners potato

Despite its long history, utilitarian value, and cultural significance to several Indigenous Tribes in the Southwest USA, the extent to which the Four Corners potato (Solanum jamesii Torr.) has been domesticated requires circumscription. Establishing the temporal and spatial dimensions of intentional cultivation would provide an essential component of the domestication argument. This project tests the hypothesis that S. jamesii tubers were processed with ground stone tools from archaeological sites located beyond the natural range of the species, especially where genetic evidence has previously indicated human transport and establishment in gardens. Microbotanical evidence, in the form of starch granules from 401 ground stone tools at 14 archaeological sites, is examined. More than 6,500 starch granules were recovered from the tools; 163 of which were assigned to S. jamesii.  Four sites (North Creek Shelter, Long House/Mesa Verde, Pueblo Bonito/Chaco Canyon, and Point of Pines) show consistent use of S. jamesii (ubiquity >18%), as early as 10,900 cal BP, and well into Puebloan times.  Three of these sites are located far north of the species’ center of distribution in the Mogollon region, across hundreds of kilometers of the Colorado Plateau, and still support an extant population nearby. This suggests an anthropogenic distribution of S. jamesii across the Four Corners region and a unique cultural identity around the use of this native potato.  These findings, combined with ethnographic interviews and nutritional data, provide clear evidence of use in relation to natural and anthropogenic distributions, thereby allowing an assessment of the degree to which these energy-rich, nutritious, and compact tubers were purposely used and transported.

Methods
Sites and Tools

There are many archaeological sites within and beyond the documented range of S. jamesii. Relatively few have been excavated, and fewer still have ground stone tools from buried deposits with published contextual details. Only excavated and recorded sites that had available ground stone tools would be suitable for starch granule analysis.  Fortunately, eight repositories had collections from 14 sites that met these criteria and provided a wide temporal range (early Holocene through Protohistoric) for sampling a sufficient number of properly stored tools.

Inventories of ground stone tools (i.e., manos and metates) from the 14 sites were examined to determine the availability and feasibility for sampling. Objects were not chosen if they 1) were on exhibit, 2) had an unknown location at the repository, 3) were not collected, or 4) were restricted objects. Approximately 98% of the ground stone tools sampled contain site and/or stratigraphic provenience. Furthermore, information regarding curation practices (i.e., handling, cleaning, and storage) was considered so that the probability of cross-contamination was low (i.e., objects bagged individually, stored in sealed boxes vs. open shelves, etc.). A total of 401 ground stone tools from eight repositories (American Museum of Natural History, Arizona State Museum, Brigham Young University Museum of Peoples and Cultures, Edge of the Cedars, Mesa Verde Visitor and Research Center, Museum of Northern Arizona, Natural History Museum of Utah, and the Western Archaeological Conservation Center) were ultimately sampled for starch granule analysis.

Starch Granule Analysis

Starch granules are microscopic plant structures formed from photosynthates by subcellular amyloplasts as energy reserves. Long-term energy reserves, or “storage starches,” are most abundant in seeds, fruits, and underground storage organs (USOs - corms, tubers, taproots, etc.). These starch granules are released from cells during food processing and can subsequently become deposited in the cracks and crevices of archaeological tools. Because of their microcrystalline structure, starch granules are relatively resistant to organic decay and are often preserved for millennia in archaeological contexts.

Consumable supplies used in this study, such as nitrile gloves, weigh boats, toothbrush heads, and test tubes, were only used once to prevent cross-contamination during sampling and processing.  Extraction of starch granules began with placing a portion of each ground stone tool in a weigh boat filled with deionized water (diH20), which was then placed in a sonicator. Stones that were too large to fit in the sonicator were spot sampled with a sonicating toothbrush. Spot sampling uses a smaller surface area than sonication, so starch yield may be affected. Each sample was then filtered with diH20 using a 125 µm  U.S.A. standard test sieve and transferred to a 50 ml sterile test tube. Samples were centrifuged for three minutes at 3000 RPM. The supernatant was discarded, and the re-suspended sample pellet was transferred to a 15 ml sterile test tube.

Heavy liquid was used to isolate starch granules from the samples; 5 ml of lithium heteropolytungstate (LST - specific gravity 2.00 - 2.35) (Central Chemical Consulting, Malaga, Australia) was added to each sample and resuspended with a vortex mixer. The sample was then centrifuged for 15 minutes at 1000 RPM. The suspended fraction was decanted into another sterile set of labeled test tubes. Two more rinses (addition of diH20 and centrifugation, three minutes at 3000 RPM) removed any residual heavy liquid. Acetone was added, then each sample was spun and allowed to dry overnight. This process resulted in the formation of a small pellet of organic material that would include starch granules, if present.

Starch Granule Identification

Following processing, pellets were re-suspended in a few drops of a 50/50 glycerol and diH2O solution and then mounted on a glass slide. Each slide was scanned in its entirety using a transmitted brightfield microscope fitted with polarizing filters and Nomarski optics (Zeiss Axioscope 2, Zeiss International, Göttingen, Germany). A digital camera (Zeiss Axiocam HRc) with imaging and measurement software (Zen Core v2.7) was used to capture images and to make starch granule measurements. Because ground stone tools were made available from repositories over six years, imaging and measurements were performed by multiple collaborators. Consistency was assured by frequent training and assigning each collaborator to all tools from a single site.

Previous studies discuss our approach for identifying S. jamesii starch granules. Those studies developed a set of statistically defined diagnostic characteristics that include: 1) possession of an eccentric hilum, 2) the presence of a longitudinal fissure, 3) the absence of fissure branching, 4) a ratio of fissure width to granule width in the range of 0.21–0.28 (mean ratio ± 99% CI), and 5) mean maximum granule length greater than 34.88 μm (mean minus the 99% CI). Archaeological granules that have an eccentric hilum and possess two or more of the characteristics are taxonomically assigned to S. jamesii. Granules with eccentric hilum and possessing less than two of the characteristics could be S. jamesii, but could also belong to other plant taxa with USOs (e.g., Liliaceae) and in this study, remain unidentified.

Control Samples

Starch occurs naturally on laboratory surfaces and can be airborne, so precautions were taken to minimize contamination during laboratory processing. Passive traps were placed in the NHMU laboratory spaces and collections room and left in place for approximately one week during sample processing. 

In addition to passive traps, control samples were taken from feature fill sediments and/or animal bones to detect background levels of starch at four archaeological sites.  Sediment control samples from North Creek Shelter were processed and analyzed in 2013. However, subsequent research has shown that sediments may not be ideal control samples due to the natural occurrence of starch and enzymatic damage caused by soil bacteria. Therefore, we sampled animal bones from the same cultural deposits where ground stone tools were recovered at Danger Cave, Camel's Back Cave, and Sudden Shelter.

Not all archaeological sites had sediment or faunal bones available for sampling, as many collections were curated decades ago, before the routine practice of collecting associated sediment during excavation. Moreover, submitting additional research requests to sample sediment or faunal material from curated collections was not always feasible. Although feature fill sediments and animal bones located in situ near artifacts may have been exposed to starch from the processing of associated plant materials, the quantity of starch granules from such sources is expected to be minimal compared to starch granules embedded in the cracks and crevices of ground stone tools.

Interviews with elderly people: The elder interviews were recorded, transcribed, and are archived at Utah Diné Bikéya, where the identities of the elders are held incognito. Before the start of our project, we submitted an application to the University of Utah Institutional Review Board (IRB), and that approval letter (IRB_00112805) is attached to the article submission.