Geophytes are perennial, herbaceous plants that store their energy in underground organs, such as roots, tubers, corms, and rhizomes (Raunkiaer Reference Raunkiaer1934). They are nutrient-dense vegetables, rich in carbohydrates, fiber, vitamins, and minerals, with caloric return rates as high as 3,500 cal/hr (comparable to those of avian and small mammal resources; Botha et al. Reference Botha, Cowling, De Vynck, Esler and Potts2022; Carney et al. Reference Carney, Tushingham, McLaughlin and Guedes2021; Fowler and Rhode Reference Fowler, Rhode and Ubelaker2006; Madsen and Schmitt Reference Madsen and Schmitt1998; McGuire and Stevens Reference McGuire and Stevens2017; Smith and McNees Reference Smith and McNees2005, Reference Smith and McNees2011; Trammell et al. Reference Trammell, O'Connell, Bush, Parker, Borghi, Hunsaker and Byrd2008). Geophytes are hardy, resilient plants that are tolerant of cold temperatures and drought and are well documented as a reliable food source for hunter-gatherers worldwide (Hawkes and O'Connell Reference Hawkes and O'Connell1992; Hurtado and Hill Reference Hurtado and Hill1987; Pontzer and Wood Reference Pontzer and Wood2021; Pontzer et al. Reference Pontzer, Wood and Raichlen2018). Over hundreds of generations, Indigenous communities in the northern Great Basin developed cultural practices centered around the gathering, preparation, and consumption of geophytes. These practices became deeply embedded in their cultural identity, forming rituals, stories, and traditions around the harvesting and use of these plants (Carney et al. Reference Carney, Tushingham, McLaughlin and Guedes2021; Couture et al. Reference Couture, Ricks and Housley1986; Fowler Reference Fowler1989; Fowler and Lepofsky Reference Fowler and Lepofsky2011; Fulkerson and Tushingham Reference Fulkerson and Tushingham2021; Kelly Reference Kelly1932; Steward Reference Steward1933; Stewart Reference Stewart1939). However, finding evidence of geophyte use in the archaeological record remains a challenge (Copeland and Hardy Reference Copeland and Hardy2018; Fulkerson and Tushingham Reference Fulkerson and Tushingham2021; Mercader et al. Reference Mercader, Akeju, Brown, Bundala, Collins, Copeland and Crowther2018).
Archaeologists in the northern Great Basin—a region encompassing the intersection of the High Lava Plains of the Columbia Plateau and the northernmost Great Basin interior drainage—have long hypothesized that human settlement patterns and foraging behaviors during the Late Pleistocene/Early Holocene were tethered to rich lacustrine environments that included wetlands, marshes, and extensive sand dunes (Bedwell Reference Bedwell1973; Duke and King Reference Duke and King2014; Madsen Reference Madsen, Graf and Schmitt2007; Weide Reference Weide1968). As regional human population densities increased, and aridity desiccated wetland resources, people were forced upslope to exploit a more diverse array of plants, including geophytes (Aikens and Jenkins Reference Aikens, Jenkins, Melvin Aikens and Jenkins1994; Brashear Reference Brashear, Melvin Aikens and Jenkins1994; Helmer and Tushingham Reference Helmer and Tushingham2021; Jenkins Reference Jenkins, Melvin Aikens and Jenkins1994; O'Connell Reference O'Connell1975). Although some research from early Holocene archaeological sites in the region alludes to the use of upland geophytes (e.g., Blong et al. Reference Blong, Adams, Sanchez, Jenkins, Bull and Shillito2019; Bradley et al. Reference Bradley, Smith and Nussear2022; Kennedy Reference Kennedy2018; Kennedy and Smith Reference Kennedy and Smith2016; Kingrey Reference Kingrey2022; McDonough et al. Reference McDonough, Kennedy, Rosencrance, Holcomb, Jenkins and Puseman2022), it has been assumed that geophyte use is a Late Holocene phenomenon based on the establishment of large village sites located in upland environments across Fort Rock Basin (Brashear Reference Brashear, Melvin Aikens and Jenkins1994; Housley Reference Housley, Melvin Aikens and Jenkins1994; Prouty Reference Prouty, Melvin Aikens and Jenkins1994, Reference Prouty Guy1995). For example, at Boulder Village, rock art sites are aggregated near upland playas containing rock ring features, abundant ground-stone tools, and numerous storage pits; the latter two were presumably used to process and store geophytes (Brashear Reference Brashear, Melvin Aikens and Jenkins1994). Botanical surveys conducted around Boulder Village revealed that most sites are associated with occurrences of ethnographically important geophytes, including biscuitroot (Lomatium Raf.), yampa (Perideridia Rchb.), camas (Camassia Lindl.), sego lily (Calochortus Pursh.), and bitterroot (Lewisia Pursh.); these findings led investigators to conclude that root gathering was the primary subsistence focus in this area (Housley Reference Housley, Melvin Aikens and Jenkins1994; Jenkins Reference Jenkins, Melvin Aikens and Jenkins1994; Prouty Reference Prouty, Melvin Aikens and Jenkins1994, Reference Prouty Guy1995).
The Warner Valley uplands in southern Oregon (southeast of Fort Rock Basin) also support large populations of ethnographically important geophytes. Embedded within many of these populations are lava outcrops with thousands of rock art panels. As with Boulder Village, rock ring structures, ground-stone tools, and bedrock grinding surfaces are also documented (Figure 1; Cannon and Ricks Reference Cannon, Ricks and Ames1986, Reference Cannon, Ricks and Quinlan2007; Louderback, Wilks, and Simper Reference Louderback, Wilks and Simper2022; Ricks Reference Ricks1995). It has been proposed that rock art sites in Warner Valley are intentionally located near ethnographically significant plant communities and are associated with the processing of those plant foods (Cannon et al. Reference Cannon, Cliff Creger, Fowler, Hattori, Ricks, Janetski and Madsen1990; Cannon and Ricks Reference Cannon, Ricks and Ames1986; Ricks Reference Ricks1995; Ricks and Cannon Reference Ricks, Cannon and Hyder1993). The current project tests the assumption that geophytes were collected and processed at these upland areas by sampling 58 bedrock metates at three rock art sites east of Warner Valley. Our research tests whether starch granules extracted from open-air bedrock metate surfaces can be identified as originating from Lomatium spp., an ethnographically significant plant that grows abundantly on the landscape near our study sites. The approach to archaeological starch research, used herein, begins with a focused examination of specific plant taxa that are members of the local flora. Systematic studies are then conducted on reference granules from that subset so that observed morphological characteristics could be applied to the identification of starch granules recovered from archaeological contexts (Louderback, Wilks, et al. Reference Louderback, Wilks, Herzog, Howat Brown, Joyce and Pavlik2022). This project focuses on Lomatium because it grows abundantly in the uplands near the sites examined, it has a rich ethnographic record of use in the region, and it is shown to be highly nutritious.
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20250210154442109-0925:S0002731624000428:S0002731624000428_fig1.png?pub-status=live)
Figure 1. Upland habitats east of Warner Valley Oregon, support large populations of ethnographically important plant communities: (a) biscuitroot, (b) yampa, and (c) wild rye (photos by William J. Cannon). Embedded within many of these plant communities are (d) lava outcrops with hundreds of rock art panels and bedrock metates (photo by Stefania L. Wilks).
Ethnographic Evidence of Lomatium
Lomatium is a genus in the carrot family (Apiaceae) that comprises more than 100 native species and subspecies distributed across western North America (Constance and Wetherwax Reference Constance and Wetherwax2023). These species and subspecies are typically associated with foothills and lower montane slopes, but some occur at higher elevations among alpine plant communities. They are perennially herbaceous forbs whose deep taproots prefer shallow, rocky soils (lithosols). Lomatium spp. flower after snow melt in early spring and produce seed before entering dormancy as summer daytime temperatures increase. Traditional ecological knowledge (TEK) describes at least 14 Lomatium species to be among the most important foraged plant foods in the Pacific Northwest: they may represent more than 30% of all ethnographically significant geophyte species harvested (Hunn Reference Hunn and French1981, Reference Hunn1990; Hunn and French Reference Hunn and French1981). All parts of the plant can be eaten fresh, but their fleshy taproots are preferred and prepared by pounding and grinding into dried cakes to be stored as winter reserves (Couture Reference Couture1978; Couture et al. Reference Couture, Ricks and Housley1986; Hunn and French Reference Hunn and French1981; Prouty Reference Prouty Guy1995).
According to ethnographic accounts, Lomatium was traditionally harvested with digging sticks in the spring as part of the systematic exploitation of seasonally available food resources. Harvesting began when flower stalks were still visible and had well-developed umbels, indicating larger taproots (Figure 2).
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20250210154442109-0925:S0002731624000428:S0002731624000428_fig2.png?pub-status=live)
Figure 2. (a) Flowering individual of Lomatium in late spring/early summer (photo by Stefania L. Wilks). Inset showing bulbous taproot, the underground storage organ; (b) Lomatium was traditionally harvested with digging sticks in the spring as part of the systematic exploitation of seasonally available food resources (photo courtesy of Washington State Digital Archives).
Recent experimental studies propose that past Indigenous management practices may have had a positive impact on geophyte growth and production (Botha et al. Reference Botha, Cowling, De Vynck, Esler and Potts2022; Carney et al. Reference Carney, Tushingham, McLaughlin and Guedes2021; Thoms Reference Thoms1989; Trammell et al. Reference Trammell, O'Connell, Bush, Parker, Borghi, Hunsaker and Byrd2008). Digging was restricted to zones of dense plant concentrations, which increased plot productivity by aerating soils and reducing competition for seeds dispersed during the digging process. When found in sufficient quantities and harvested according to traditional practices, Lomatium can provide substantial calories. Preliminary studies of densities and harvest rates suggest these geophyte resources could provide the bulk of the annual calorie requirement (Hunn Reference Hunn and French1981, Reference Hunn1990; Hunn and French Reference Hunn and French1981). Harvest rate estimates indicate that one forager could harvest a year's supply of Lomatium taproots for four people in 400 hours (Hunn and French Reference Hunn and French1981:92). When this TEK became established or how it may have changed over time is unknown. However, ethnobotanical data provided by the Indigenous people whose traditional lands are represented in the northwestern Great Basin clearly indicate that Lomatium was a resilient, primary food source with strong cultural affiliations (Couture Reference Couture1978; Couture et al. Reference Couture, Ricks and Housley1986; Housley Reference Housley, Melvin Aikens and Jenkins1994; Hunn Reference Hunn and Minnis1981; Prouty Reference Prouty Guy1995; Stewart Reference Stewart1939).
Bedrock Metates and Starch Granule Analysis
Bedrock metates are shallow, often highly polished depressions created on the bedrock's surface through grinding or milling activities. They are ubiquitous throughout North America and are largely understood through their contextual relationship with landscape and ethnographic models (Burton et al. Reference Burton, Adams, Willis and Nadel2017; Lynch Reference Lynch2021; Lynch et al. Reference Lynch, Matthews and Noble2017; Wilks et al. Reference Wilks, Paredes and Louderback2024). For example, in central Nevada, oval grinding slicks or depressions on boulders and rockshelter ledges were used to process pine nuts from nearby pinyon woodlands (Tinsley et al. Reference Tinsley, Louderback, Pavlik, Baker, Townsend, Tucker and Wilks2021). In contrast to analyses of ground-stone tools preserved in a buried context, very limited starch research has been conducted on bedrock mortars and metates exposed to natural erosional elements (but see Wilks et al. Reference Wilks, Paredes and Louderback2024). Nevertheless, bedrock milling features have the potential to provide evidence of past human lifeways, including foods collected and processed, settlement patterns and land investment, and the frameworks for social identity (Buonasera Reference Buonasera2012, Reference Buonasera and Tammy2015; Burton et al. Reference Burton, Adams, Willis and Nadel2017; Fulkerson and Tushingham Reference Fulkerson and Tushingham2021; Lynch Reference Lynch2021; Lynch et al. Reference Lynch, Matthews and Noble2017; Shoemaker et al. Reference Shoemaker, Davies and Moore2017; Stevens et al. Reference Stevens, Whitaker and Rosenthal2019; Tinsley et al. Reference Tinsley, Louderback, Pavlik, Baker, Townsend, Tucker and Wilks2021).
The study of starch granules preserved in archaeological contexts can refine our understanding of which plant taxa people were processing. Carbohydrates produced during photosynthesis are stored as starchy reserves in seeds, fruits, and underground organs. Grinding processes related to human cultural activities release the starches that become deeply lodged in the surface cracks and crevices of stone artifacts. Due to their semi-crystalline structure, starch granules can remain preserved in archaeological contexts for millennia, thus providing direct evidence of plant exploitation. Systematic morphometric studies of modern plant starches, including granule size, shape, and 3D-surficial structuring, provide useful taxonomic references through which archaeological identifications can be made with reliable accuracy (Henry Reference Henry and Henry2020; ICSN 2011; Liu et al. Reference Liu, Ma and Cui2014; Louderback et al. Reference Louderback, Herzog and Pavlik2017; Louderback, Wilks, et al. Reference Louderback, Wilks, Herzog, Howat Brown, Joyce and Pavlik2022; Mercader et al. Reference Mercader, Akeju, Brown, Bundala, Collins, Copeland and Crowther2018; Reichert Reference Reichert1913; Wang et al. Reference Wang, Zhao, Wang and Lui2019).
Recent starch analysis on excavated ground stone tools has revealed evidence of geophyte use by Late Pleistocene / Early Holocene hunter-gatherer populations in western North America (Herzog and Lawlor Reference Herzog and Lawlor2016; Joyce et al. Reference Joyce, Louderback and Robinson2022; Louderback and Pavlik Reference Louderback and Pavlik2017; Rankin Reference Rankin2016; Rhode and Rankin Reference Rhode, Rankin and Thomas2020). On the Colorado Plateau in southern Utah, residues found on ground-stone tools buried in North Creek Shelter indicate the use of a native potato, Solanum jamesii, as early as 10,900 years ago (Louderback and Pavlik Reference Louderback and Pavlik2017). Geophyte starch was also identified on early Holocene tools from Hogup Cave in Utah (Herzog and Lawlor Reference Herzog and Lawlor2016). Evidence for the processing of spring parsley (Cymopterus spp.) at Early and Middle Holocene sites (around 9400–4400 cal BP) in the Wyoming Basin was documented on ground-stone tools (Joyce et al. Reference Joyce, Louderback and Robinson2022). And ground stone from high elevations (around 8300–1400 cal BP) as far ranging as the White Mountains of California and Wyoming's Wind River Range (Rankin Reference Rankin2016; Rhode and Rankin Reference Rhode, Rankin and Thomas2020) supports the dietary importance of geophytes throughout the Holocene.
Methods and Materials
Bedrock Metates
Starch residues were extracted from 58 open-air bedrock metates at the three upland rock art sites in southern Oregon: six from Corral Lake (35LK500), 28 from Barry Spring (35LK502), and 24 from Long Lake (35LK514) (Figure 3; supplemental data archived in DRYAD, https://datadryad.org/stash/dataset/doi:10.5061/dryad.tqjq2bw52). The bedrock metate features occur on fine-grained, basalt outcrops containing hundreds of petroglyphs and pictographs and are situated along the edges of upland playa lakes and springs (Cannon and Ricks Reference Cannon, Ricks and Ames1986; Ricks Reference Ricks1995; Ricks and Cannon Reference Ricks, Cannon and Hyder1993).
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20250210154442109-0925:S0002731624000428:S0002731624000428_fig3.png?pub-status=live)
Figure 3. Starch residues were extracted from bedrock metates at Corral Lake, Barry Spring, and Long Lake. Bedrock metate features occur on fine-grained volcanic basalt outcrops containing hundreds of petroglyphs and pictographs situated along the edges of upland playa lakes and springs (photos by Stefania L. Wilks).
Rock art and bedrock metates from all three sites show evidence of patination, a geochemical process thought to develop a varnish on inorganic surfaces over time caused by weathering. Some of the patinated grinding surfaces showed evidence of more recent reuse occurring on top of the patina surface (Figure 4). A subset of bedrock metates in this study were sampled based on the presence of patina (n = 12) or patination with reuse (n = 8) on their surface. The remaining bedrock metates sampled (n = 38) were associated with rock art panels but did not exhibit patina on their surface.
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20250210154442109-0925:S0002731624000428:S0002731624000428_fig4.png?pub-status=live)
Figure 4. (Upper panel) Rock art panels identified as Great Basin Carved Abstract (GBCA) were found buried beneath a Mt. Mazama ash layer in association with Western Stemmed projectile points. (Lower panel) A subset of bedrock metates was sampled based on the presence of patina (left) or reuse over surface patination (right) (photos by Stefania L. Wilks).
Starch Granule Analysis
Reference Materials
Plant lists were compiled and specimens collected in the field to document site floras with particular focus on Lomatium spp. because of its importance to Indigenous communities as well as its abundance and ubiquity across the landscape (Louderback, Wilks, and Simper Reference Louderback, Wilks and Simper2022). Vouchered specimens were mounted, labeled, and reposited in the Garrett Herbarium, Natural History Museum of Utah (NHMU), as well as on the Intermountain Regional Herbarium Network. The collection is titled “Flora of Warner Valley, Oregon-Geophyte Study” (http://intermountainbiota.org/portal/collections/index.php?catid=1) to document site floras (Louderback, Wilks, and Simper Reference Louderback, Wilks, Herzog, Howat Brown, Joyce and Pavlik2022).
Our approach to identifying archaeological starch granules from bedrock metates began with a systematic study of Lomatium granules extracted from extant plant populations within the vicinity of each site (Louderback, Wilks, and Simper Reference Louderback, Wilks, Herzog, Howat Brown, Joyce and Pavlik2022). Three Lomatium species (L. donnellii, L. macrocarpum, and L. triternatum) were collected, and each taproot (three per species) was sampled for starch. Using randomly generated x–y coordinates on the microscope stage, we described and measured morphological characteristics, such as size, shape, and surface features, of 100 starch granules from each taproot (n = 900 granules total; for descriptions of morphological characteristics, see Louderback, Wilks, et al. Reference Louderback, Wilks, Herzog, Howat Brown, Joyce and Pavlik2022). The frequency of characteristics occurring on granules was calculated and expressed as a number between 0.0 and 1.0. This approach allowed us to develop a set of statistically defined characteristics frequently observed in Lomatium granules, which could then be applied to the identification of archaeological granules (Figure 5).
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20250210154442109-0925:S0002731624000428:S0002731624000428_fig5.png?pub-status=live)
Figure 5. (Upper left) Three Lomatium species (L. donnellii, L. macrocarpum, and L. triternatum) were collected near the three study sites and sampled as starch reference material (photo by Lisbeth A. Louderback). (Upper right and lower panel) Diagnostic characteristics of more than 900 starch granules were quantified and used to identify archaeological starch residues (adapted from Louderback, Wilks, and Simper Reference Louderback, Wilks and Simper2022).
Sample Collection in the Field
After collecting starch samples from the surfaces of 58 bedrock metates from all three sites, we assessed environmental contamination by collecting separate metate surface and noncultural control samples. These samples were analyzed in conjunction with the samples extracted from deeper within the bedrock metate's geologic matrix (for the expanded protocol, see Wilks et al. Reference Wilks, Paredes and Louderback2024). All consumable supplies were used only once to prevent cross-contamination during sampling in the field. Loose sediment was brushed from metate surfaces before sampling. Approximately 50 mL of distilled water (DH20) was then added to the stone surface and cleaned with an ultrasonic toothbrush. A sterile syringe was used to transfer the DH20 and sediment to a sterile test tube labeled “surface extract.” This was repeated until the DH20 wash was clear.
Next, 50 mL of a 2.5% solution of sodium hexametaphosphate (Na-Hex) was added to the cleaned bedrock surfaces to deflocculate residue that was more deeply embedded in the stone's interstitial matrix. After about an hour, approximately 50 mL of DH20 was added, and a sterile ultrasonic toothbrush was used to vigorously clean the surface for five minutes. A sterile syringe was then used to transfer the resultant serum to a 50 mL test tube labeled “interstitial extract.”
Control Samples
Thirteen control samples (Barry Spring, n = 6; Corral Lake, n = 1; and Long Lake, n = 6) were also collected from nonworked or noncultural stone surfaces about 5–10 m from the sampled archaeological bedrock metates. These control samples were collected at each site and treated in the same way as the surface and interstitial starch samples. Control samples were labeled as “control” and numbered consecutively as they were collected in the field. The aim was to compare the control sample residue to the residue extracted from the bedrock metate. Although noncultural stone surfaces in the vicinity of bedrock metates may have been contaminated with starch from associated plant materials, this source would yield insignificant levels of starch granules when compared to starch granules ground into the stone matrix of bedrock cultural features and artifacts.
Laboratory Analysis
All samples were processed in the Archaeobotany Laboratory at the Natural History Museum of Utah (NHMU). Each residue serum was sieved through a 125 μm mesh Endecotts sieve into a beaker using deionized water (DiH20). Sample water greater than 125 μm was discarded, while the sampled portion under 125 μm was retained and transferred to a 50 mL test tube and centrifuged for 3 minutes at 3000 RPM. The supernatant was then discarded and the sample pellet transferred to a new 15 mL test tube, resuspended with about 10 mL of DiH20, mixed with a vortex, and centrifuged for 3 minutes at 3000 RPM. Approximately 7 mL of Lithium heteropolytungstate (LST: specific gravity 2.2) was added, resuspended with a vortex mixer, and then centrifuged for 15 minutes at 1000 RPM. The starch samples were extracted from the surface of the heavy liquid using a pipette and transferred to a freshly labeled 15 mL test tube. To remove any residual heavy liquid, each sample was rinsed two to three times with about 10 mL of DiH20, vortexed, and centrifuged for 3 minutes at 3000 RPM. The sample was then decanted and resuspended with about 7 mL of acetone, mixed with a vortex, and centrifuged for 3 minutes at 3000 RPM. The acetone was decanted, and samples were covered and left to dry overnight. Individual samples were reconstituted with 50% DiH20 and 50% glycerol and mounted on a glass slide for observation. 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). Starch granules were photographed under 400× magnification with a digital camera (Zeiss HRc) and measured with Zeiss Zen software.
Results
Bedrock Metates
Twelve bedrock metates displayed visible patina, seven of which yielded a total of 101 starch granules. Of these, five granules (5%) were attributed to Lomatium and six (6%) to Apiaceae. Eight bedrock metates were patinated with evidence of reuse; seven of these surfaces yielded a total of 130 granules, of which nine (7%) were attributed to Lomatium spp. and 18 (14%) to Apiaceae. There is a difference in mean starch yield between patina versus reused surfaces (reused surfaces produced higher yields), although it was not quite statistically significant (t = 1.73, df = 18, p = 0.10).
Starch Granule Analysis
Reference Materials
Granule size distributions for each of the species were pooled (n = 300 per species) and their distributions were plotted. Morphological characteristics occurred more frequently in the top 20% size range of starch granules, and therefore, we report frequencies for those granules (Liu et al. Reference Liu, Ma and Cui2014; Louderback et al. Reference Louderback, Herzog and Pavlik2017; see Figure 5).
Seven diagnostic characteristics were used to identify archaeological granules of Lomatium spp. (Louderback, Wilks, et al. Reference Louderback, Wilks, Herzog, Howat Brown, Joyce and Pavlik2022): size range of 10.7–31.0 μm, circular to oval shape, centric hilum, distinct extinction cross, fissure (transverse/stellate), visible lamellae, and pressure facets. Archaeological starch granules exhibiting six or seven of these characteristics are considered Lomatium spp. (Figure 6). Granules displaying four or five are attributed to their family, Apiaceae. Starches with fewer than three characteristics remain unidentifiable.
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20250210154442109-0925:S0002731624000428:S0002731624000428_fig6.png?pub-status=live)
Figure 6. Micrographs of Lomatium spp. starch granules shown in differential interference contrast (DIC). (Top row) Reference granules from Lomatium spp. harvested from boundaries within each site. (Bottom rows) Archaeological granules extracted from bedrock metates at Long Lake, Barry Spring, and Corral Lake attributed to Lomatium spp (see reference photos for scale).
Laboratory Analysis
Interstitial residues extracted from 43 of the 58 metate surfaces yielded a total of 644 starch granules. Of those 644 starch granules, 81 (13%) were assigned to Lomatium spp., and 107 (17%) were identified as belonging to Apiaceae (Figure 6; Table 1). Control and surface samples associated with each site yielded significantly fewer starch granules than the interstitial samples overall. Control samples yielded four starch granules, whereas surface samples yielded a total of two granules (Wilks et al. Reference Wilks, Paredes and Louderback2024).
Table 1. Interstitial (Archaeological) and Surface Starch Granules Recovered from Bedrock Metates at Corral Lake (n = 6), Barry Spring (n = 28), Long Lake (n = 24), and All Control Samples (n = 13).
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20250210154442109-0925:S0002731624000428:S0002731624000428_tab1.png?pub-status=live)
Although the current study focuses on the presence of Lomatium spp., starch granules from other plant species were also present on the bedrock metates from Corral Lake, Barry Spring, and Long Lake. Granules of Triticeae, the grass tribe that includes wild rye, barley, and wheat, occurred frequently (6%) on the metate surfaces. Two species of wild rye (Leymus cinereus and Elymus elymoides) grow at the sites today, so it is conceivable that people in the past also collected and processed these grasses. Despite our attempts to identify archaeological starch granules, a large number of granules (80%) extracted from the bedrock metates remain unidentified due to enzymatic damage. Nevertheless, every granule observed in this study was photographed and measured to enable future identification (https://doi.org/10.5061/dryad.tqjq2bw52).
Discussion
Despite the historical and contemporary significance of geophytes to Indigenous groups across the globe, archaeological evidence for their use is difficult to reveal. Starch granule analysis is a novel technique used by archaeologists to assess the significance of plants in ancient human diets and expose which plant species were collected and how they were processed. Recent starch granule analyses have shed light on the importance of plants in ancient human diet (Henry Reference Henry and Henry2020; Herzog and Lawlor Reference Herzog and Lawlor2016; Joyce et al. Reference Joyce, Louderback and Robinson2022; Louderback and Pavlik Reference Louderback and Pavlik2017; Rankin Reference Rankin2016; Rhode and Rankin Reference Rhode, Rankin and Thomas2020).
The present study identified starch granules, including biscuitroot (Lomatium spp.), on bedrock metate surfaces at all three rock art sites (Corral Lake, Barry Spring, and Long Lake), thus providing direct evidence for the processing of an ethnographically significant geophyte growing nearby. Some of these bedrock metates had a patina, suggesting older grinding surfaces, whereas others showed evidence of reused surfaces embedded within the patination (Cannon and Ricks Reference Cannon and Ricks1999; Middleton et al. Reference Middleton, Smith, Cannon and Ricks2014). Our results indicate that archaeological starch residues can be extracted from the interstitial matrices of grinding surfaces, regardless of patination and its effects on bedrock metate features.
Rock art sites and bedrock metates in the uplands east of Warner Valley are associated with Late Pleistocene / Early Holocene archaeology (Benson et al. Reference Benson, Larry, Hattori, Southon and Aleck2013; Cannon and Ricks Reference Cannon and Ricks1999; Middleton Reference Middleton2013; Middleton et al. Reference Middleton, Smith, Cannon and Ricks2014; Ricks Reference Ricks1995; Smith and Barker Reference Smith and Barker2017). Rock art panels identified as Great Basin Carved Abstract (GBCA) were found buried beneath a layer of Mt. Mazama ash (eruption around 7,560 ybp) and in association with Western Stemmed projectile points (Middleton et al. Reference Middleton, Smith, Cannon and Ricks2014; Smith and Barker Reference Smith and Barker2017). Western Stemmed projectiles are chronologically associated with Late Pleistocene / Early Holocene archaeological contexts (Brown et al. Reference Brown, Gilmour, Solimano and Ames2019; Jenkins et al. Reference Jenkins, Davis, Stafford, Campos, Hockett, Jones and Scott Cummings2012; Smith et al. Reference Smith, Duke, Jenkins, Goebel, Davis, O'Grady, Stueber, Pratt and Smith2020) and are frequently used as regional measures of relative dating (Rosencrance et al. Reference Rosencrance, McDonough, Holcomb, Endzweig and Jenkins2022). In addition to being associated with Western Stemmed projectile points, the rock art panels and bedrock metates are also embedded within extant populations of ethnographically significant geophytes. These geophyte plant communities would have been a reliable source of carbohydrates to people living in the region (Carney et al. Reference Carney, Tushingham, McLaughlin and Guedes2021; Couture et al. Reference Couture, Ricks and Housley1986; Madsen Reference Madsen, Graf and Schmitt2007; McGuire and Stevens Reference McGuire and Stevens2017; Prouty Reference Prouty Guy1995; Steward Reference Steward1933). Therefore, it is possible that the collecting and processing of upland geophytes were important parts of subsistence lifeways since the transition from the last Ice Age. Whether the bedrock metates in this study were created and used by people living on the landscape during the Late Pleistocene / Early Holocene remains unclear, however. Given the open-air nature of these features, it is possible they were used repeatedly to process geophytes throughout the Holocene.
Although archaeological features such as rock art and bedrock metates are difficult to date, they are contextually and functionally tied to our understanding of past people's relationships with their environment (Buonasera Reference Buonasera2012, Reference Buonasera and Tammy2015; Fulkerson and Tushingham Reference Fulkerson and Tushingham2021; Lynch Reference Lynch2021; Lynch et al. Reference Lynch, Matthews and Noble2017; Shoemaker et al. Reference Shoemaker, Davies and Moore2017; Stevens et al. Reference Stevens, Whitaker and Rosenthal2019; Tinsley et al. Reference Tinsley, Louderback, Pavlik, Baker, Townsend, Tucker and Wilks2021). Bulk resource processing at fixed locations is often done in large family groups or at community-level events (Adams Reference Adams1999; Helmer and Tushingham Reference Helmer and Tushingham2021; Lynch Reference Lynch, Matthews and Noble2017; Lynch et al. Reference Lynch2021; Madsen and Schmitt Reference Madsen and Schmitt1998; McCarthy et al. Reference McCarthy, Hicks and Blount1985; O'Connell et al. Reference O'Connell, Hawkes, Blurton Jones, Kroll and Douglas Price1991). In Warner Valley, all the largest rock art sites recorded—Long Lake, MC Reservoir, Portrait Rim, Point Juniper Reservoir, North of Point Juniper, Barry Spring, Hidden Lake, 44 Lake, and Corral Lake—contain bedrock metates (Cannon and Ricks Reference Cannon, Ricks and Ames1986; Ricks Reference Ricks1995). The presence of geophyte starch granules, specifically Lomatium spp., a plant found growing near the sites today, supports the hypothesis that rock art sites with bedrock metate grinding features are intentionally located near ethnographically significant plant communities and that they are associated with the processing of those plant foods (Cannon and Ricks Reference Cannon, Ricks and Ames1986, Reference Cannon, Ricks and Quinlan2007; Cannon et al. Reference Cannon, Cliff Creger, Fowler, Hattori, Ricks, Janetski and Madsen1990; Ricks Reference Ricks1995).
Indigenous people of the northern Great Basin possess a rich cultural identity with geophytes as essential components of their nutritional subsistence strategies (Couture et al. Reference Couture, Ricks and Housley1986; Fulkerson and Tushingham Reference Fulkerson and Tushingham2021; Helmer and Tushingham Reference Helmer and Tushingham2021; Housley Reference Housley, Melvin Aikens and Jenkins1994; Hunn Reference Hunn and Minnis1981; Kelly Reference Kelly1932; Prouty Reference Prouty Guy1995; Steward Reference Steward1933; Stewart Reference Stewart1939). The gathering of native geophytes, such as biscuitroot, yampa, bitterroot, camas, and sego lily, remains important in traditional cultural practices. Ethnohistoric accounts describe starchy, edible geophytes such as Lomatium as highly valued root crops that were traditionally harvested in the spring, when groups or families migrated to “root-digging” camps in the mountains. Seasonal harvesting campgrounds were social places that brought extended families together, facilitating the exchange of knowledge and trade goods, the arrangement of intertribal marriages, important regional gambling competitions, and strengthening bonds of cooperation required for survival in semiarid landscapes (Couture et al. Reference Couture, Ricks and Housley1986; Fowler Reference Fowler1989; Wewa and Gardner Reference Wewa and Gardner2017). Wilson Wewa, a member of the Warm Springs Paiute and a keeper of traditional knowledge, recalls learning the stories of his people while harvesting geophytes as a young boy with his extended family in the mountains of southeast Oregon (Wewa and Gardner Reference Wewa and Gardner2017).
TEK evolves through generations of holistic practices aimed at increasing the quantity or quality of culturally preferred plants (Anderson Reference Anderson2005; Berkes Reference Berkes2018; Fowler and Leland Reference Fowler and Leland1967; Fowler and Lepofsky Reference Fowler and Lepofsky2011; Fowler and Rhode Reference Fowler, Rhode and Ubelaker2006; Hunn Reference Hunn and Minnis1981; Hunn and French Reference Hunn and French1981; Nabhan Reference Nabhan2000). Hunter-gatherer societies have often been portrayed as passively existing within their environments, but ethnobotanical researchers have demonstrated that these groups actively managed landscapes through sophisticated practices. A growing body of research argues that Indigenous peoples were managing their landscapes long before the arrival of European-based colonists (Anderson Reference Anderson1997, Reference Anderson2005; Carney et al. Reference Carney, Tushingham, McLaughlin and Guedes2021; Deur Reference Deur2009; Deur and Turner Reference Deur and Turner2005; Fowler Reference Fowler, Williams and Hunn1982; Pavlik et al. Reference Pavlik, Louderback, Vernon, Yaworsky, Wilson, Clifford and Codding2021; Peacock and Turner Reference Peacock, Turner, Minnis and Elisens2000; Stewart Reference Stewart, Lewis and Anderson2002; Turner Reference Turner2014, Reference Turner2020a, Reference Turner2020b). Low-intensity land management practices as strategic burning, tilling, and seed dispersal may have contributed to long-term stable economies and social organizations very distinct from traditional agrarian societies (Lightfoot et al. Reference Lightfoot, Cuthrell, Striplen and Hylkema2013). Sustainable Indigenous ecological practices, such as controlled burning to promote the growth of edible plants and enhance animal habitats, the cultivation of “semi-domesticated” plants, such as berry patches, and the management of shellfish beds, ensured long-term productivity and resilience in resource systems (Deur Reference Deur2009; Lepofsky et al. Reference Lepofsky, Hallett, Lertzman, Mathewes, Mchalsie, Washbrook, Deur and Turner2005; Lightfoot et al. Reference Lightfoot, Cuthrell, Striplen and Hylkema2013). In the northwestern Great Basin, Klamath and Modoc communities intentionally modify the landscape to meet their dietary, social, and spiritual needs by managing culturally preferred plant species and their habitats. Likewise, Paiute communities develop cultural practices centered around the gathering, preparation, and consumption of geophytes (Brashear Reference Brashear, Melvin Aikens and Jenkins1994; Couture Reference Couture1978; Couture et al. Reference Couture, Ricks and Housley1986; Deur Reference Deur2009; Deur and Turner Reference Deur and Turner2005; Housley Reference Housley, Melvin Aikens and Jenkins1994; Kelly Reference Kelly1932; Prouty Reference Prouty Guy1995; Steward Reference Steward1933). Connecting past human foodways with the present exemplifies TEK empowers Indigenous food sovereignty and can be used to inform public land management policy.
Conclusion
Plant communities rich in geophyte species are essential to the traditional lifeways of many Indigenous groups across the globe. These species are closely associated with archaeological features in the northern Great Basin, suggesting their cultural and dietary significance throughout time. Starch granule analysis can assess that significance by revealing which plant species were collected and how they were processed. Furthermore, this novel technique often corroborates ethnohistoric and contemporary accounts of traditional plant use.
Geophytes are recognized for their multifaceted roles in Indigenous cultures, serving not only as food resources but also as catalysts for social cohesion, cultural exchange, and cooperative endeavors. Their importance extends beyond mere sustenance to encompass broader aspects of community well-being and identity. Understanding the cultural and ecological significance of geophytes can inform public land management policy, fostering more inclusive and sustainable approaches to natural resource management. Incorporating Indigenous perspectives and traditional ecological knowledge into decision-making processes supports the conservation and stewardship of geophyte habitats.
Acknowledgments
The bedrock grinding surfaces and rock art features described in this research exist on sacred ground that has been respected, cared for, cherished, and protected over the ages and to this day by the Klamath, Modoc, and Paiute Peoples. We commit to follow their lead in the conservation and stewardship of these habitats. We are grateful to the BLM Lakeview Office, especially Carolyn Temple for funding our cooperative agreement, and Grace Haskins, Shelli Timmons, and Cynthia Brown for executing the funding. Kathy Stewardson joined us in the field and helped choose the bedrock metates for sampling: we appreciate her time and advice throughout the project. Thank you to Samantha Paredes for assisting with starch samples in the lab and mounting plant specimens on herbarium sheets. Thank you to Allison Izaksonas for repositing the herbarium collection into the NHMU Garrett Herbarium and imaging them for online access. We also thank Phoebe McNally at the DigitLab (University of Utah) for making the maps, and Fanny Blauer for translating our abstract into Spanish.
Funding Availability Statement
Funding for this project came from the Department of the Interior Bureau of Land Management (Lakeview District Office) Cooperative Agreement (L20AC00050) awarded to Lisbeth A. Louderback.
Data Availability Statement
Starch granule measurements, descriptions, and images are available at https://doi.org/10.5061/dryad.tqjq2bw52.
Competing Interests
The authors declare none.