Recent research by Henrikson and colleagues (Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017) to evaluate the stratigraphic context of Folsom fluted points and mammoth remains recovered from the lower levels in Owl Cave raised concerns regarding the previous claim of a Folsom–mammoth association (see Miller Reference Miller, Ericson, Taylor and Berger1982, Reference Miller, LeMoine and Scott MacEachern1983, Reference Miller, Bonnichsen and Sorg1989; Miller and Dort Reference Miller, Dort and Bryan1978). These concerns stem from the absence of unambiguous evidence of bone tool manufacture or marrow extraction, given other viable explanations for the surface modifications on the mammoth bone. These include noncultural causal agents such as carnivore gnawing and impacts from roof fall.
During the past year, additional documents, maps, profiles, field notes, and photographs created during the last three field seasons at Owl Cave (1975 through 1977) were gifted to the Museum of Idaho (MOI). These records, combined with the donation of the Owl Cave archaeological collections to MOI in 2019, have allowed us to further assess the relationship between the Folsom assemblage and mammoth from Layer 18 at a much higher resolution. In this report, we present a series of new radiocarbon dates and other evidence demonstrating that postdepositional periglacial processes occurring in the cave during the terminal Pleistocene mingled a Folsom assemblage—temporally associated with bison remains—with an older event involving the deposition of mammoth remains.
The Formation of Layer 18
Owl Cave and adjacent Coyote and Dry Cat Caves are collapsed lava tubes located in one of many expansive Pleistocene-aged lava fields on the eastern Snake River Plain (Figure 1). These lava fields originated from periodic eruptions of shield volcanoes, with molten lava spreading across the terrain as pā-hoe-hoe flows or transported via tubes (Keszthelyi and Self Reference Keszthelyi and Self1998; Kuntz et al. Reference Kuntz, Covington, Schorr, Link, Kuntz and Platt1992; Pawar et al. Reference Pawar, Katikar, Vaddadi, Shinde, Rajaguru, Joshi and Eksambekar2015; Wentworth and Macdonald Reference Wentworth and Macdonald1953). Lava tubes are created when the surface of a cooling pā-hoe-hoe flow inflates from trapped gases, forming a hollow conduit within the flow. As a tube-fed flow subsides, sections of the tube may form an unstable dome prone to collapse (Kempe et al. Reference Kempe, Bauer, Bosted, Coons, Elhard and Middleton2010; Lockwood et al. Reference Lockwood, Hazlett and de la Cruz-Reyna2022; Pawar et al. Reference Pawar, Katikar, Vaddadi, Shinde, Rajaguru, Joshi and Eksambekar2015). Owl, Coyote, and Dry Cat Caves represent three closely spaced collapses of the same lava tube dome; these are often referred to collectively as the Wasden site. Although the age of the flow and each cave's collapse is unknown, it appears that sediment only began accruing in Owl Cave's interior very late in the Pleistocene (Miller and Dort Reference Miller, Dort and Bryan1978).
Figure 1. Aerial and topographic view of the Snake River Plain Wasden Site, including Dry Cat Cave, Owl Cave, and Coyote Cave.
Owl Cave's southwest aspect created an ideal setting for the accretion of aeolian sediments, resulting in more than 6 m of stratified loess deposits beneath the cave overhang (see Henrikson et al. Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017). By 1978, roughly 400 m3 had been excavated in the eastern half of the cave, revealing cultural occupations reportedly spanning the last 13,000 years (Figure 2). Folsom points and modified mammoth bone, argued to represent the earliest cultural assemblage in the cave, were recovered from Layer 18 (Figure 3), which was “capped” or “sealed” by a catastrophic roof fall event that occurred after the Folsom occupation (Miller Reference Miller, Ericson, Taylor and Berger1982:84; Miller and Dort Reference Miller, Dort and Bryan1978:131). Layer 18 was characterized as a “discrete” 50–80 cm thick deposit of highly organic “dark-colored silt with a high fraction of basalt grit” between lighter carbonate-rich silts and clays resting above and below (Miller Reference Miller, Ericson, Taylor and Berger1982:84). The cave floor, described as decomposing basalt matrix, rests below a 35 cm deposit of culturally sterile “yellow clay” beneath Layer 18 (see Figure 3).
Figure 2. (a) Plan view of the excavated and unexcavated units in Owl Cave; (b) View of alcove during 1977 excavations of the lowest levels in Owl Cave. (Courtesy of the Museum of Idaho.)
Figure 3. North wall profile of Layer 18, unit E4G, with layers labeled: (a) Layer 16; (b) Layer 18; (c) sterile yellow clay; (d) Intermixed gravels on the cave floor. (Courtesy of the Museum of Idaho.)
Periglacial Conditions
As early as 1967, Dort (Reference Dort1968:33) recognized a variety of deformations in the Owl Cave strata, including numerous vertical intrusions in the profile (Figure 4). Dense concentrations of these deformations appear both above and below the early Holocene bison bone bed (Butler Reference Butler1968). Dort (Reference Dort1968:31) surmised that, prior to the deposition of the bison bone, dramatic seasonal freeze-thaw cycles within the cave allowed for the formation of ice wedges. As these wedges melted during spring thaw, their cavities filled with wind-blown debris and slump, forming what Dort described as ice wedge “casts.” As deeper strata were excavated, evidence of ice wedges became intermixed with distorted sediments (Dort Reference Dort1968:35; see Figure 4), which he interpreted as periglacial “involutions.” (It is unclear if Dort was involved in the 1970s excavations at Owl Cave, or if he had the opportunity to examine the alcove deposits exposed between 1975 and 1977.)
Figure 4. Owl Cave profile showing ice wedges and involutions above and below the early Holocene bone bed. (Courtesy of the Museum of Idaho.)
Although Miller (Reference Miller, LeMoine and Scott MacEachern1983:41) does acknowledge that the deepest strata in Owl Cave were “deformed” after deposition, she asserts that “very little mixing of the distorted sediments at the upper and lower contacts was observed” (Miller Reference Miller, Ericson, Taylor and Berger1982:84). The results of a sediment analysis prompted the conclusion that Layer 18 had retained sufficient “integrity” to demonstrate that the lithics and faunal remains were restricted to this stratum and are therefore “contemporaneous” (Miller Reference Miller, Ericson, Taylor and Berger1982:89). However, Figure 5 shows the variegated floor of the alcove at a depth of roughly 6.4 m below main datum (BMD). This mottled surface suggests that the contacts between Layer 18 and the strata above and below were far from well defined. Field notes also indicate that each 2 × 2 m unit in this area of the cave was excavated in arbitrary 10 cm levels, which would eliminate the stratigraphic controls required to recognize the context of individual artifacts or their relationship to the faunal assemblage. It is also unclear how roof rubble and boulders were removed during the excavations. If the rubble was not excavated cautiously, the provenience of individual artifacts and bones within or below the roof fall would have been difficult to ascertain.
Figure 5. Variegated floor of Layer 18 during excavation of the alcove. (Courtesy of the Museum of Idaho.)
The Roof Fall Episode
The deepest sediment deposits in Owl Cave were clearly altered by a range of postdepositional cryogenic processes, including freeze-thaw cycles (see Anketell et al. Reference Anketell, Cegla and Dzulynski1970; Butrym et al. Reference Butrym, Cegla, Dzulynski and Nakonieczny1964; French Reference French1986). However, the terminal Pleistocene roof fall event that dislodged the massive basalt boulders encountered between 5.0 and 6.0 m BMD (Figure 6) would have likely resulted in further, even catastrophic, impacts to any cultural deposits resting below. The removal of the roof fall was considered imperative, given the discovery of a mammoth bone fragment wedged between boulders during the 1967 field season (Butler Reference Butler1968). Because the deeper portions of the cave contained more manageable rock “rubble” (see Figures 2 and 6), the last phase of excavation focused on this area rather than expanding toward the exterior dripline (Miller Reference Miller, Bonnichsen and Sorg1989). Although Miller (Reference Miller, Ericson, Taylor and Berger1982:84, Reference Miller, Bonnichsen and Sorg1989:384) described the roof fall as “capping” or “sealing” Layer 18, she does acknowledge that this event likely occurred “during or just after deposition” of sediments that were likely “saturated” from periglacial conditions. Field profiles from the alcove reveal boulders and cobbles clearly within Layer 18 (Figure 7), and plan views from the same units show a substantial amount of basalt rubble at the same depth (see Figure 7). Oddly, Miller (Reference Miller, Ericson, Taylor and Berger1982:42) maintained that the fractured mammoth bone was the result of human modification rather than damage from roof fall, arguing that the “bulk of the bone” had likely been “well-preserved by cushioning in the wet sediment” (Miller Reference Miller, Ericson, Taylor and Berger1982:87). This assertion is not supported, given the extensive abrasions and visible impacts on most of the mammoth bone recovered from Owl Cave (see Henrikson et al. Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017).
Figure 6. North wall profile of units in alcove. (Courtesy of the Museum of Idaho.)
Figure 7. Plan view of Layer 18 in Owl Cave, Level 620 BMD, showing rubble. Plan view and profile redrawn from the 1977 field notes.
Deciphering Layer 18 Spatial Relationships
With the entire Owl Cave assemblage now at MOI, we have been able to evaluate some of the scenarios presented by Henrikson et alia (Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017:588), including the argument that the fluted points and mammoth, despite reportedly being from the same stratum, are not associated. Newly acquired images of the alcove excavations (Figure 8) emphasize the postdepositional processes at work in Owl Cave, including what appears to be extensive sediment deformation.
Figure 8. Mammoth bones in situ on the floor of the alcove, with the dark sediments (Layer 18) visible in the profile. (Courtesy of the Museum of Idaho.)
Under periglacial conditions, deformations occur when unfrozen course-grained sediments—such as sand, resting above a frozen, impermeable layer—become saturated (Vandenberghe Reference Vandenberghe2013). During the next cold cycle, as the upper portion of the saturated active layer freezes, unfrozen sediments below are contorted, deformed, or “squeezed” between the frozen mass above and the impermeable layer below (Bertran et al. Reference Bertran, Andrieux, Antoine, Deschodt, Font and Sicilia2017; McArthur and Onesti Reference McArthur and Onesti1970; Sharp Reference Sharp1942). During thaws or cycles involving saturation, fine-grained sediments (including silts and clays), as well as objects, can move vertically through a soil profile (Pollard Reference Pollard, Menzies and van der Meer2018; Wood and Johnson Reference Wood and Johnson1978). Wood and Johnson (Reference Wood and Johnson1978:343) observed particle sorting during an experiment involving saturated sediments. When subjected to repeated freeze-thaw conditions, “coarse particles” migrated upward in the profile, whereas “fines” moved downward. Silts and clays, which are particularly susceptible to frost heaving due to their high porosity, can also form “segregated ice lenses” (Pollard Reference Pollard, Menzies and van der Meer2018:542). As fine sediments refreeze, “frost-jacking” may force objects upward “when an ice lens preferentially forms beneath” (Pollard Reference Pollard, Menzies and van der Meer2018:542; Wood and Johnson Reference Wood and Johnson1978). Figure 8 shows mammoth ribs in situ on the floor of the lava tube well below the boundary of what was identified as Layer 18, clearly indicating that faunal remains were not confined to this stratum. The pronounced involutions also visible in the profile show the magnitude of the periglacial processes occurring in Owl Cave during the terminal Pleistocene. If saturated sediments were subjected to freeze-thaw cycles, there is the possibility that the “stratigraphy” encountered in the bottom of the cave is the product of particle sorting, with coarse-grained sediments (Layer 18) resting above yellow clay.
Although the mammoth took center stage in previous investigations, Miller (Reference Miller, Ericson, Taylor and Berger1982:88, 1983:40) did acknowledge that bison, as well as camel, were also recovered from beneath the roof fall. However, it appears that the bison bone, along with other potentially diagnostic faunal remains, received limited attention at that time. The bison remains were briefly described as “isolated teeth, ribs, long bones and pelves” assigned to Bison antiquus (Miller Reference Miller, Ericson, Taylor and Berger1982:91). However, most of the “Layer 18” faunal assemblage acquired by MOI in 2019 is still in gallon-sized bags labeled as “scrap bone.” There is no documentation to suggest that these remains have been inventoried or analyzed before now. Table 1 presents the new AMS assays generated from diagnostic bison and highly fragmentary mammoth specimens recovered from these bags. The associated provenience information (Figure 9) underscores the extensive vertical and horizontal displacement of artifacts and faunal remains recovered from the lowest levels of Owl Cave, likely as a result of recurring freeze-thaw cycles, frost-heaving, and frost-jacking (see Wood and Johnson Reference Wood and Johnson1978). The mammoth and bison are not contemporaneous; they are mixed throughout the deformed sediments below the roof fall.
Table 1. Recent Owl Cave Radiocarbon Ages on Mammoth and Bison from Layer 18.
Figure 9. Comparison of the vertical distribution of AMS-dated bison and mammoth bone and Folsom projectile points by excavation unit in Owl Cave. Numbers identify AMS-dated bison and mammoth bone (see also Table 1): (1) UGAMS-25078, (2) ACRF-4364, (3) ACRF-4365, (4) ACRF-4366, (5) ACRF-4369, (6) ACRF-4370, (7) ACRF-4373, (8) ACRF-4374, (9) ACRF-4379, (10) Aeon-1182, (11) ISGS-A2303, (12) Beta-322088, (13) Beta-322087, (14) ACRF-4362, (15) ACRF-4367, (16) ACRF-4368, (17) ACRF-4372, (18) ACRF-4381, and (19) ACRF-4382.
The new AMS assays on ultrafiltered collagen from Layer 18 bison remains (see Table 1) correspond to those previously generated by Miller (Reference Miller, Bonnichsen and Sorg1989:383). The possibility that the bison could be associated with the Folsom lithics was only briefly mentioned (Miller Reference Miller, Ericson, Taylor and Berger1982:89, Reference Miller, LeMoine and Scott MacEachern1983:40; Miller and Dort Reference Miller, Dort and Bryan1978:137) but not pursued in favor of the more sensational mammoth bone “quarry” (Miller Reference Miller, Bonnichsen and Sorg1989). Figure 10 shows the summed probability distributions of the calibrated AMS assays on bison and mammoth from the lowest levels of Owl Cave, confirming that the mammoth is pre-Clovis in age, whereas the bison dates fall neatly in line with the age range of Folsom technology recently presented by Buchanan and colleagues (Reference Buchanan, David Kilby, LaBelle, Surovell, Holland-Lulewicz and Hamilton2022).
Figure 10. Summed probability distribution (SPD) of calibrated Layer 18 mammoth and bison AMS dates. SPD was created using rcarbon (Crema and Bevan Reference Crema and Bevan2021) and R package ggplot2 (Wickham Reference Wickham2008).
Discussion
Despite having a temporal association between the Folsom assemblage and bison remains, we find ourselves in a position similar to that of the original Owl Cave researchers. We currently lack the evidence to demonstrate an unequivocal association. However, there is additional circumstantial evidence in the form of other stone tools recovered from the deepest sediments within the alcove (see Figure 7). Bifaces and debitage were briefly mentioned in early reports (Miller Reference Miller, LeMoine and Scott MacEachern1983:41; Miller and Dort Reference Miller, Dort and Bryan1978:131), but their association with the points was never described in detail. As with the faunal remains, these items were assumed to provide further evidence of human predation on “extinct fauna” (Miller and Dort Reference Miller, Dort and Bryan1978). Although Henrikson and colleagues (Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017) proposed that the fragmentary Folsom artifacts recovered from the alcove might be in a secondary context (i.e., deposited via slope wash), their presence in proximity suggests that this part of the cave was not under perpetual periglacial conditions during the Younger Dryas. If seasonal thaws occurred, Owl Cave may have been suitable for short-term human activity.
At this point, there is no compelling evidence to suggest human involvement with the mammoth in the bottom of Owl Cave (Haynes Reference Haynes1991; Henrikson et al. Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017; Karr Reference Karr2015). However, previously reported positive reactions for elephant and horse on two Folsom specimens (see Henrikson et al. Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017) merit attention. Although horse remains were not previously identified in the assemblage, the scrap-bone bags contain a large volume of highly fragmentary bone that has yet to be analyzed. As stated in Henrikson et alia (Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017), periglacial conditions in the deepest parts of the cave could potentially slow the decomposition of a mammoth, allowing for the contamination of stone tools in proximity to the remains (see Popp Reference Popp2024). Because the western half of Owl Cave has yet to be excavated, we can only report on mammoth bones recovered from the alcove. These remains include highly fractured long bones and ribs, a partial scapula, and a molar (Miller Reference Miller, Bonnichsen and Sorg1989:384). Many of southern Idaho's lava tubes were occupied by large carnivores during the terminal Pleistocene, leaving behind fragmentary postcranial remains of extinct ungulates (see Henrikson and Long Reference Henrikson, Long, Graf and Schmitt2007). Although the scapula, long bone, and ribs are suggestive of a den, the isolated molar is harder to explain, unless the rest of the mammoth is still under roof fall in unexcavated portions of the cave. At this point, we only know that mammoth bones were in the cave when the roof fall episode occurred, likely causing the breakage patterns and characteristics initially used as evidence for human modification.
Conclusions
Owl Cave is now typically excluded from lists of North American sites with compelling evidence of archaeological materials in direct association with terminal Pleistocene megafauna based on concerns initially raised by Haynes (Reference Haynes1991:238) and Grayson and Meltzer (Reference Grayson and Meltzer2002:331). These concerns include the disparity between the known age of Folsom points and the radiocarbon ages associated with the mammoth, the plausibility of other causal agents to explain the physical condition of the mammoth bone, and the absence of heavy tools that would have been required to break thick, cortical bone. We would also add that the original reports (see Miller Reference Miller, Ericson, Taylor and Berger1982, Reference Miller, LeMoine and Scott MacEachern1983, Reference Miller, Bonnichsen and Sorg1989) inadequately documented the stratigraphic context of artifacts and fauna, making it extremely difficult to assess the relationship between archaeological materials and the remains of Pleistocene mammals.
Although the remaining intact deposits in Owl Cave likely retain important evidence regarding Folsom activity and its possible association with the faunal assemblage, it is critical to recognize, evaluate, and exhaust the data potential of the existing MOI collections prior to considering further excavations at the site. This is an essential step in the development of well-defined, ethical research objectives and advanced excavation techniques that also recognize and embrace the necessity of preservation, especially for National Register properties such as Owl Cave.
Our ongoing research with the curated assemblage will focus on the lithic debris and tools recovered from within and below the roof fall to decipher possible associations with the Folsom material, with consideration given to potential vertical displacement (see Henrikson and Long Reference Henrikson, Long, Graf and Schmitt2007). The stone tools can also be evaluated for functional evidence that could be associated with butchery activity. The contents of the scrap bone bags will be cataloged and analyzed. Although the bulk of the faunal material in the bags appears to be highly fragmented mammoth bone, there is no indication that these fragments were previously examined for evidence of butchery or modification. Bison and other faunas have already been identified, including extinct taxa. Given the possibility that Owl Cave may also have functioned as a carnivore den during the terminal Pleistocene, it will be important to conduct morphometric analyses as well as obtain additional AMS assays on these specimens.
Owl Cave is the only known site in the Desert West with a Folsom assemblage in a buried context (Henrikson et al. Reference Henrikson, Byers, Yohe, DeCarlo and Titmus2017; Rondeau Reference Rondeau2023). We now know that the age of the bison bone from the lower levels is consistent with Folsom. The locality also contains evidence of a 9,000-year-old mass bison kill that closely resembles Paleoindian kill sites on the Great Plains (Butler Reference Butler1968; Clements Reference Clements2023; Frandson and Henrikson Reference Frandson and Henrikson2023), hinting at the possibility that the cave could have been utilized for the same purpose for several thousand years. If a strong association can be demonstrated between the Folsom assemblage and the bison remains, Owl Cave will represent the only known site of its kind west of the Rocky Mountains.
Acknowledgments
The authors thank the many individuals, including the staff with the Idaho National Laboratory Cultural Resource Management Office, who assisted with the preparation of this manuscript. Randy and Stephen Harris, grandsons of Leonard Wasden, generously donated the Owl Cave collection to the Museum of Idaho (MOI), making this research possible. Kristina Frandson and other dedicated MOI staff have worked diligently to organize the collections and allowed access to newly acquired field notes and documents. MOI provided the loan of Owl Cave materials for these analyses, and Battelle Energy Alliance generously funded the new accelerator dates.
Funding Statement
This work was supported by the Idaho National Laboratory and the Museum of Idaho.
Data Availability Statement
The original field notes and associated documentation used to support the findings of this study are available at the Museum of Idaho, Idaho Falls. Recently generated AMS assays are available at the Cultural Resource Management Office, Idaho National Laboratory, Idaho Falls.
Competing Interests
The authors declare none.
