Rimrock Draw Rockshelter (35HA3855): A Summary of Recent Fieldwork at a Stemmed Point Site Near Riley Oregon
by Kelsey Provost, Jessica French, and Pat O'Grady.
Rimrock Draw Rockshelter is located near Riley Oregon in Harney County. During the period from September 12 to September 23 of this year, test excavations occurred at the site to determine its potential for use as a field school research project. This paper is a summary of the recent fieldwork, some artifacts that were found, and plans for future research at the location.
Discovered during surveys by the Burns Bureau of Land Management and volunteers from the Oregon Archaeological Society, the surface finds included 26 stemmed points, a Black Rock Concave Base point, a crescent fragment, Northern Side-Notched points, biface fragments, including a fluted biface, overshot flakes suggestive of Clovis technology, a small number of Elko Series points, and a bedrock mortar. The stemmed points are, by far, the dominant group of projectile points recovered on the surface and they account for all of the diagnostic artifacts collected within the excavation units that are described in greater detail below. Most have rounded bases, but six have squared bases and the majority are representative of Parman Type 1 and Parman Type 2 points, respectively. The fluted biface, in association with the overshot flakes, is of particular interest because, through proximity, they point to the possibility that Clovis-age artifacts might be deposited within the rockshelter itself. Crescents are unique and somewhat enigmatic stone tools that are more frequently found in lakeside settings, so the recovery of a fragment along the stream drainage was surprising. Black Rock Concave Base points are uncommon anywhere in Oregon, but the 10,000 - 12,000 year old points are found more commonly in the area around Riley and Wagontire than elsewhere. All of the surface-collected artifacts were found within a 300 meter radius of the rockshelter.
Encouraged by the finds and noting 6 to 7 feet high sagebrush around the rockshelter apron which is an indicator of deep sediments, Burns BLM Archaeologist Scott Thomas recommended the use of an auger in front of the rockshelter. The auger pulled up an obsidian flake at 180 cm, and reached 200 cm before bottoming out on rock. Thomas decided to promote the excavation at the rockshelter as a means of exploring early Holocene climate change, and funding for the project was approved by the BLM state and national offices.
A crew composed of BLM and U of O archaeologists put in two 2x2 and one 1x1 test units at the site to assess the potential for understanding climate change issues, and with the hope the site could eventually become a University of Oregon field school location. The excavations revealed the presence of two cultural components, including an upper component furnished with a hearth and ground stone implements, and a dark, midden-like lower component that terminated in a compacted occupation surface at 180 cm.
The stratigraphy in Units 1 and 2 indicated water borne silt to coarse sands. The sediments also indicate gradual accumulations over time, consistent with the idea that the small drainage that passes the site today was once perennial.
Unit 1 was located against the rockshelter face. Excavations in Unit 1 ended at 200 cmbd with only a small exposure of sediment at that point. The sediments became extremely dark and rich in charcoal at level seven. A hearth feature was identified in the sidewall, on top of the bedrock that eventually encompassed the majority of the unit. The hearth feature contains two charcoal lenses interbedded with fire affected earth. Obsidian flakes are present in the sidewall in both the charcoal lenses and the fire affected earth. After excavating the feature, sediment samples were taken from both charcoal lenses. Large amounts of charcoal were recovered throughout Unit 1 in level 7. The hearth feature ties in well with an upper cultural component that was equally prominent in Unit 2 just north of Unit 1. There, a mano and a metate were found in levels 5 and 7 along with an abundance of lithic debitage. Pollen analysis is pending on the ground stone.
Auger probes were put in across the site in order to obtain a better sense of the sediments. At about one meter, Auger Probe 2, located northwest of Unit 2, pulled up extremely black sediment, much darker than the previous. The auger work served as a guide to what might be found in the excavation units. The same sediment appeared in Unit 2 at the same depth and was determined to be a distinct cultural component due to the high quantities of artifacts found within. This lower cultural component, distinct from the upper component with the hearth and groundstone, is approximately 50cm thick, made up of loosely consolidated silt, charcoal, and other cultural material. The lower cultural component transitions into a thin, well compacted and discrete occupation surface that displays a clear change from the previous sediments. The occupation surface sloped in a dish shape to the southeast, towards the back of the shelter. Underlying this surface at an abrupt transition is an orange to medium brown sand and gravel layer, in which artifact counts are sharply reduced. The orange layer may prove to be sterile with depth, but our time was too limited to explore this deposit beyond the initial discovery.
The lower component represents a distinct cultural deposit for several reasons. First, the dark color and texture are consistent with other Great Basin examples of concentrated habitation. Second, already consistently high counts of debitage increased within the cultural component, especially at level 12. Third, fifteen tools were collected throughout, including bifaces, scrapers, EMFs, and ground stone. On the deep occupation surface, a cache of debitage was found at 170-180 cmbd in Quad A. The cache was 25 cm x 13 cm in size at 170 cmbd. It was located on a narrow rock ledge that suggested site furniture. The cache consisted of 88 obsidian flakes, an obsidian scraper, and a biface with overshot flakes removed from it. Also found on the occupation surface was a stemmed point preform with weak shoulders, near a ground stone artifact of indeterminate use.
Early in the project, it was decided that a third unit would be opened 10 meters to the west of TU-1 and 2 in a smaller alcove of the rockshelter. Tools were recovered in levels 4-7. They included a scraper, two hammer stones, and a chopper. A stemmed point with an intact base and resharpened blade was recovered from level 2. The unit was so rocky that it was not possible to delineate any cultural components beyond the tool concentration.
Built up against the wall of the rockshelter between Units 1/2 and Unit 3 is a packrat midden that holds the potential for producing perishable cultural material. To illustrate, a pair of feathers bound with sinew were found on the ground surface below the midden after a strong windstorm passed through. Attempts to date the feathers will be made following species identification. Their existence is a positive sign that the packrat midden could hold more artifacts. Future plans includes disassembly of the midden in hopes of finding more perishable material.
The test excavations at Rimrock Draw Rockshelter were extremely successful in demonstrating how important this site could be to Northern Great Basin Archaeology. There is solid evidence for extended habitation of considerable antiquity that carried across two weeks of excavation. Cultural deposits extended to 2 meters and auger probes indicated at least another 50 cm of untested sediments lie below.
At this stage, artifacts are being submitted for analyses of various kinds. The buried lithic tools will be sent for protein residue analysis, and then forwarded for obsidian sourcing and hydration. The ground stone objects will be sent for pollen and macrobotanical analysis. Charcoal from the hearth and from in situ recovery is being identified in preparation for radiocarbon dating, and the feather bundle will be sent away for identification and eventual dating. The University of Oregon plans on running an archaeology field school at the Rimrock Draw Rockshelter beginning next June. A paleobotany field school will also run for three weeks and offer four credits of instruction. The abundance of cultural material makes Rimrock Draw Rockshelter an exciting new site with great potential for multiple lines of research.
Items recovered from the 2012 excavations at Rimrock Draw Rockshelter:
A Glimpse into the 2012 University of Oregon Archaeology Field School at Rimrock Draw Rockshelter
by Patrick O'Grady, Margaret M. Helzer, and Scott P. Thomas
The 2011 fieldwork at Rimrock Draw Rockshelter (35HA3855) was originally brought to the attention of the CAHO readership by Provost and French (2011), who reported on the BLM-sponsored test excavations at the site in September of that year. During that session, three excavation units produced charcoal, debitage, edgemodified flakes, bifaces and other formed tools. Northern Side-notched points (ca. 7,000 to 4,000 BP), and Western Stemmed points (ca. 12,000-7,000 BP) were the only projectile points recovered within the rockshelter at that time, though a wider variety was collected from the landscape surrounding it. The artifacts were encountered through excavations that reached depths of 190 cm. Subsequent augering indicated that the deposits were considerably deeper. Following the 2011 work, we decided to return to the site with the 2012 University of Oregon (U of O) Archaeology Field School.
We convened in June of 2012 for a six-week school with a crew of 23 archaeology students gathered from across the United States (Figure 1), accompanied by four supervisors. Scott Thomas and Chuck Morlan of the Burns BLM were also on hand, providing support and additional supervision. During the first week, Clovis expert Michael F. Rondeau returned to offer his Paleoindian Lithic Workshop, now in its fifth year. Marge Helzer taught a three-week Paleoethnobotany (PEB) Field School alongside the archaeology field school (Figure 2). Six students attended, the maximum number possible for the available PEB facilities. We also had a strong turnout from the Oregon Archaeological Society, our well-trained and highly valued volunteer group based largely in the Portland/Vancouver area. At the peak of operations, we had 29 students, six teacher/supervisors, eight volunteers, and a variety of visiting researchers at the site ... the place was buzzing!
The relationship between the paleoethnobotany and archaeology field schools proved to be very beneficial for both groups. The archaeology field crew had the luxury of being able to call “up the hill” to the PEB crew when an intriguing stain or hearth feature was encountered. They would troop down to the site to collect samples and discuss their context with the archaeology crew. Returning to the lab trailer, the PEB crew would conduct sediment flotation on the spot to extract botanical remains, and the archaeologists would have preliminary information regarding the contents of the samples in one to two days. The interdisciplinary nature of this interaction is in keeping with a tradition first established for the field school by Luther Cressman and carried on through the years by Mel Aikens and Dennis Jenkins. We regularly offer geoarchaeology field instruction as a part of our interdisciplinary focus, but the PEB program was new. The course will be offered again in 2013. The most significant finds from the 2012 botanical work were burned fragments of chenopods, wada, willow, and bulrush (Figure 3). All are edible or utilitarian genera and the latter two offer a view of the riparian environmental conditions at the time of site occupation. Wada is a lakeside plant species that may have been brought to the site for consumption. The burned bulrush seeds are being submitted for radiocarbon dating.
The archaeologists excavated 12 new units during the field school: 11 that were 1x2 m in size, and a 2x2. Unit 2, a 2x2 started in September of 2011 that reached a depth of 190 cm, was reopened with the expectation that we would reach bottom during the six-week field school. That did not prove to be the case. The new units were established at the east, central, and west portions of the rockshelter to enhance our understanding of deposits acros the site. Along with the excavation, extensive pedestrian surveys were made of the ground surface surrounding the rockshelter to learn about the distribution and composition of lithic scatters.
The results were encouraging. The surface surveys yielded more stemmed points. Two distinct artifact concentrations were identified that produced two fluted bifaces, one concave base point, 14 overshot flakes, two unfinished bifaces with overshot scars, a spurred tool, and other artifacts consistent with fluted point technology (Rondeau, personal communication). We noted lithics suggestive of fluted technology in previous surveys, but the concentrations found this summer indicate a more substantial presence than we realized.
The rockshelter excavations revealed a three-part stratigraphic series consisting of eolian sediments underlain by a dense series of silty clay layers, which, in turn, are underlain by an orange sandy clay layer extending to bedrock. The ca. 170 cm-thick eolian deposit has a high degree of mixing, but our work also indicated that there is a consistent 1000 year-old component a meter below the surface across the site with the ca. 1300 year-old Newberry pumice underneath (Foit 2012), and another 4000 year-old component near the bottom. Below that is the second stratum; a series of clays, silts, sands, and occasional lenses of tephra that, thus far, contain nothing more recent than stemmed points in the 7,000 to 12,000 year range. One Haskett point was found in the deepest sediments of this stratum. Two hearth features were also found, which produced the burned willow twigs and bulrush seeds mentioned above. Eight tephra samples from this layer have been submitted to Washington State University for analysis.
Within Unit 2, there are concentrations of roof fall in two separate, deep layers, offering indications of structural changes through time that may have capped and protected deeper deposits. The deepest and most massive of these ancient collapses prevented our progress beyond ca. 250 cm this summer. We decided to return for two additional weeks of work in September, supported by funding provided through Scott Thomas of the Burns BLM and Stan McDonald, the lead archaeologist for the BLM in Oregon. Terry Paddock, a volunteer of long standing at Paisley Caves and former Paisley field school student, arrived in advance of the team to split and remove boulders prior to the September excavations. He exposed the third sediment package, the orange sand/clay layer that was sparsely populated with cultural material. This layer extended to ca. 330 cm before terminating on weathering bedrock. Within the deposit were a few pieces of debitage and one thick cortex flake of chalcedony modified into a tool. The translucent, caramel-colored flake is convex and rectangular, with a single, worn edge with pronounced serrations from rough flaking that were smooth and uniform in height from use, suggesting that cutting was the primary activity for this artifact. The tool was found 285 cm below datum (275 cm deep).
Multiple large tooth enamel fragments were found above the chalcedony artifact, including one concentration at 273 cm and a single piece at 260 cm. Dr. Edward Davis, manager of the U of O Museum of Natural & Cultural History Condon paleontological collection, assisted in the identification of the specimens utilizing the museum's fossil and contemporary comparative collections. The tooth enamel fragments lack clear diagnostic attributes, but share more similarities with camel than horse and are undoubtedly Pleistocene in age. The relationship of the tooth fragments to the flake tool is not clear, complicated by the fact that all were deposited in a rocky stratum where the sequence of deposition is tough to decipher. The discovery of the cultural items in close proximity to Pleistocene fauna, all underlying a dense accumulation of roof fall, is very compelling.
Rimrock Draw has already proven to be a highly significant site, but much more is in store. The field school will be returning again in 2013. By the end of the summer session, the 2012 excavators succeeded in removing most of the eolian sediments in the new units. Next year's excavators will be reopening the units and working into the stemmed point deposits (and beyond) over the course of six weeks. The PEB field school will begin its second year of operation, again offering instruction for up to six students. We also plan to offer two geoarchaeology field schools including a six-week course for undergraduate students and a three-week advanced course for graduate students. Mike Rondeau will return to teach his Paleoindian Lithics workshop.
It is clear that there is still much to learn at this site. The presence of a distinct stemmed point component deep in the deposits is reason enough to return. Artifacts that have associations with fluted point technology are now being found in concentrations on the surface and it is possible that similar material may be recovered in the rockshelter itself. The chalcedony flake tool in the proximity to Pleistocene faunal remains, all of which were buried and separated by heavy roof fall, offers a good reason for optimism about what 2013 might bring.
Acknowledgements: The University of Oregon Archaeology Field School receives annual funding from the Department of Anthropology and the Museum of Natural and Cultural History, and operates under the administration of the Academic Extension Program (formerly Summer Sessions). The field school also receives annual funding from the U. S. Department of the Interior, Bureau of Land Management, under the Clovis Quest and Climate Change programs. Scott Thomas of the Burns BLM was directly responsible for securing this funding, for employing the services of Terry Paddock, and for the tephra analyses. George Wingard graciously provided the trailer that was used for the PEB lab.
2012 Letter Report of two tephra samples submitted from Rimrock Draw Rockshelter (35HA3855). Washington State University School of Earth and Environmental Sciences, Pullman.
Provost, Kelsey, and Jessica French
2011 Rimrock Draw Rockshelter (35HA3855): A Summary of Recent Fieldwork at a Stemmed Point Site Near Riley, Oregon. Current Archaeological Happenings in Oregon 36(4):4-8.
Rondeau, Michael F.
2012 Personal Communication: Email dated October 10, 2012.