Northern Great Basin Prehistory Project
Archaeological Field School
2014 Summer Session

NGBPP Research at the Paisley Caves

Dennis Jenkins and bison bone
Dennis Jenkins with bison bone from Paisley Caves

by Dennis L. Jenkins
Director, Northern Great Basin Field School

Museum of Natural and Cultural History
University of Oregon

The Paisley Caves are located in the Summer Lake Basin north of Paisley in south-central Oregon. The site is composed of 8 caves and rockshelters in a west facing ridge of scoriacious basalt.

Cut by waves at the height of the Pleistocene, these caves lie at an elevation of 4520 ft. above two prominent beach lines cut into the talus slope. As lake levels fell at the end of the Pleistocene the current Summer Lake Basin was hydrographically separated from the Chewaucan Basin sometime between 17,000 and 18,000 cal. BP. Water levels in both basins continued to fall until about 14,500 cal. BP when a resurgence in the Chewaucan Basin caused lake levels to rise above the 4388 ft. sill separating the two basins. The Chewaucan River then breached the gravel fan separating the two basins and began flowing north into the Summer Lake Basin. Over the next 2000 years high lake stands established prominent shorelines at elevations of roughly 4330, 4360, and 4380 ft. During much of this period water stood within one mile of the Paisley Caves, making occupation of the caves substantially more attractive than it has been since.

Luther Cressman had a crew test the Paisley Caves in 1938, trenching caves 1, 2, and 3 from the mouth to the back wall in each case. His crews returned the following year to complete the excavation of the cave interiors. They removed the deposits in three stratigraphic units: sediments above Mazama ash, in the ash, and below the ash. In the pre-Mazama deposits of Cave 3 they recovered the remains of late Pleistocene camel, bison, and horse in apparent association with artifacts. Cressman returned to the site one last time in 1940 to verify this apparent association of megafauna and artifacts. While Cressman believed he had demonstrated the association of artifacts with extinct Pleistocene fauna, few other researchers have formally accepted his interpretations due to the lack of adequate documentation for these finds.

To test Cressman's theories, the UO field school conducted new excavations at the Paisley Caves during the 2002, 2003, 2007, 2009 and 2010 field seasons. Investigations were conducted in caves 1, 2, 3, and 5. Sediments were removed by arbitrary 5 centimeter levels dug within strata. Students were closely supervised at all times. Our emphasis was on the recovery of in situ bone, coprolites and cultural materials. Photographs were taken of all large mammal bones and artifacts found in situ. Soil samples were recovered from around those that could possibly be late Pleistocene in age.

Suiting up to collect coprolite
Suiting up to collect coprolite

Cave 1, located at the south end of the site, is about 6 m long and 4 to 5 meters wide. The roof at the mouth of the cave has fallen since 1939. A pair of huge boulders now block direct access to the interior of the cave, covering a good portion of the deposits in the cave mouth. A narrow concavity extending under the larger of the two offered the near perfect location in which to excavate a series of 1x2 m test units in predominantly undisturbed deposits. Excavations continued to 230 cm through extremely dry deposits. The upper 80 cm of sediments contained sparse Middle and Late Holocene cultural deposits. Deposits between 80 and 160 cm are, for all practical purposes, culturally sterile though a small hearth and a few artifacts, dated at 8440 cal. BP, were encountered just below a lens of Mazama ash at approximately 120 cm. The oldest and densest cultural remains in the cave were located between 190 and 230 centimeters. Radiocarbon dates of 8440, 7540, and 7445 have been obtained from sage brush charcoal, human feces, and a basketry fragment. Obsidian hydration readings on debitage recovered from 200 cm suggest that the earliest occupation samples date from about 11,000 cal. BP, although a butcher-cut sheep bone has been dated to 11,930 cal. BP and a rabbit bone to 13,770 cal. BP.

Cave 2, located next to Cave 1, is roughly 7 meters long and 6 meters wide. Large boulders extend across most of the entrance at the dripline and access to the lower, central portion of the cave is obtained by way of a narrow path along the northeast wall. Intact deposits were encountered near the bottom of this path. UO excavations reached depths of 240 centimeters along the cave wall. Coprolites and perishable artifacts were relatively common in Late Holocene deposits. A thick lens of Mazama tephra, dated at 7640 cal. BP, marked the upper boundary of Early Holocene deposits. A coprolite located below Mazama tephra was dated to 8620 cal. BP. A small but dense charcoal lens, suggesting the presence of a hearth, was encountered at a depth of 200 centimeters, more than a meter below the tephra lens. What appear to be predominantly artiodactyl bones, split to facilitate the removal of the marrow, were recovered in large quantities around and in this feature along with lithic debitage and tools of various kinds including a short length of sagebrush rope, a wooden peg, a pumice abrader, and scrapers. A short fragment of Western Stemmed or Foliate point base was the only projectile point found in apparent association with the hearth. Charred processed edible tissues recovered from the charcoal lens were radiocarbon dated to a mean age of ca. 12,000 cal. BP. The sagebrush rope also produced a date of ca. 12,000 cal. BP. Horse and camel bones were encountered well above this hearth, but the presence of rodent dens near the location of their recovery suggests they may be in disturbed contexts. A twig, rodent bone, sage grouse bone, and another rodent bonew were dated to 13,450, 13,740, 13,950 and 14,220, respectively. A large mammal bone chip gave a date of 14,320. Camel and horse bones were found in the Qasal sands with these items. Obsidian chips occur in this stratum as well, with large quuantities of bone.

Cave 2 Stratigraphy
Cave 2 Stratigraphy

Cave 3, mistakenly numbered Cave 4 by the field school, is roughly 4 meters deep and 7 meters wide. This is the cave from which Cressman recovered bones of bison, camel, horse, large dog (possibly wolf), fox and probably bear, as well as hawk, sagehen, pintail, and teal. UO excavations conducted during the 2002 field season were situated at the mouth of the cave in the area of a prominent berm composed of disturbed midden deposit. Excavations reached nearly 4 meters and were terminated among the heavily rounded boulders near the floor of the cave but not on bedrock. Cultural materials and bone were bimodally distributed in the undisturbed deposits below the disturbed deposits of the berm. Obsidian hydration measurements indicate the upper meter of deposits contain Middle and Late Holocene cultural materials. Mazama tephra deposits occur at about 130 cm, corresponding well with an obsidian hydration reading of 5.1 microns and a slight increase in cultural materials. After dropping off again to sterile deposits, cultural materials, particularly bone, increased again between circa 280 and 370 cm with the most notable peak occurring between 280 and 290 cm.

Cave 5, located next to Cave 3/4, was not investigated by Cressman but has been vandalized for many years. It is more than 6 m deep and 9 m wide. Excavations skirted the vandalized portions of the cave, focusing on the south-central and north ends of the cave. A large block excavation removed part of the berm of disturbed midden in the south-central portion of cave mouth. Excavations there reached a maximum depth of 300 cm in a fine, moist, gravelly silt. Bone preservation was generally poor below a prominent lens of Mazama tephra due to the saturation of the sediments. Excavations further inside found dry deposits, recovering a willow dart butt, dated to 11,540 cal. BP, and a large bison horn core spatula nearby. A large mammal bone and a coprolite were dated to 14,310 cal. BP, and 14,420 cal. BP, respectively. Obsidian projectile points, bifaces, flakes and scrapers were recovered with horse bones in basal sands dated between 13,000 and 14,600 cal. BP.

Deposits at the north end of the cave were extremely dry, unconsolidated—though generally compact—wood rat midden. This deposit exhibits incredible preservation of perishable materials. These smelly, dusty deposits are largely comprised of rat feces, bone fragments, twigs, straw, sand, and gravel. The bottom of the deposit was generally reached among wave-rounded boulders at depths of about 260 cm though cultural remains continued to depths of 300 cm in pockets among the boulders.

Perishable artifacts comprise the largest portion of the cultural assemblage. Tiny fragments of very fine netting occur in the upper 50 cm of deposits and processed sinew was recovered between 225 and 300 cm in deposits among the boulders. Three strands of processed grass fiber thread were required to get an AMS date of 12,750 cal. BP.

Cassie Albush with Camel Vertebra
Cassie Albush with Camel Vertebra

A camel vertebra was encountered at a depth of 224 cm, a camel astragalus was recovered at 228 cm, and various other camel foot bones were recovered scattered throughout the deposit. The astragalus was AMS dated to 14,290 cal. BP. Bison long bone fragments and a patella were also recovered along with the mandibles of an extinct goat. It is interesting to note that few of these bones can be considered good meat bones and none exhibit obvious evidence of butchery. However, two fragments of long bone do appear to have been broken by impact fractures while they were fresh to facilitate the removal of marrow.

A burned horse phalanx was recovered from a small charred feature that may have been a hearth. This small bowl-shaped ash feature, discovered at a depth of about 200 cm, may be cultural. It appears to have a rock lined base, and is at the depth that large mammal bones, waterfowl, and fish remains first increase dramatically over previous numbers, suggesting that this feature could have had a cultural origin and that it may date from an early time period. However, accepting it as a cultural feature must be approached cautiously since these highly organic deposits are prone to smolder for long periods of time once they begin to burn. Oxygen in rodent passages has allowed smoldering to continue deep below the surface in at least one instance in this portion of the cave. Consequently, this feature has not yet been dated, though ash samples were recovered from it for that purpose. However, a second horse phalanx was recovered from an adjacent 1x1 meter excavation unit at a depth of 210-215 cm. It produced an AMS date of 13,140 cal. BP. Human coprolites from this deposit have produced AMS radiocarbon dates as old as the camel astragalus. Radiocarbon dating of twigs, urine, bone, coprolite, cordate and charcoal have produced more than 40 dates ranging from 1240 to 16,190 cal. BP. Artifacts and coprolites found in well-stratified and soundly-dated contexts have produced C-14 dates between 14,170 and 14,340 cal. BP.

Article by Dr. Dennis Jenkins:

Protocol for Collection of Coprolites for DNA Analysis*

Summary and Conclusions

Human Coprolite
UO field school excavations were conducted in Paisley Caves numbers 1, 2, 3, and 5. Field methods were intensively focused on recovering cultural materials and faunal remains in situ within stratigraphic context. Lithic debris was generally found to be very sparse and the proportion of tools to debitage unusually high. This pattern suggests that occupations were generally limited to very brief stays and that the site was not generally a destination camp. The pattern changes somewhat for the lowest terminal Pleistocene and Early Holocene cultural components, where lithic debitage is much more common though seldom dense.

The majority of the faunal assemblage is comprised of microfauna deposited by raptors and carnivores between human occupations of the caves. Fish and waterfowl are most common in the Late Pleistocene deposits at the bottom of the caves though they also occur in smaller numbers in the Late Holocene deposits. Large mammal bones are most commonly confined to deposits which also include artifacts. However, they may occur practically anywhere throughout the deposits due to various kinds of disturbances.

Consequently, establishing the true associations of materials recovered within the caves is absolutely vital to our analysis. We have responded to this challenge by restricting our radiocarbon dating efforts to the sampling of artifacts and bone that is identifiable to species. AMS dates indicate that the Pleistocene cultural deposits range from ca. 12,000 to 14,340 cal. BP. Camel, bison, horse, and extinct artiodactyl remains have been recovered in deposits with perishable items of human manufacture. One modified large bone has been directly dated at 14,230 cal. BP.

Dry desert caves, like those at Paisley 5 Mile Ridge, are wonderful repositories for perishable items that do not ordinarily survive in open sites. Unfortunately, humans are not the only process by which remains come to be deposited in caves. Rats, squirrels, raptors, carnivores, wind, water, and gravity all contribute to the formation and alteration of cave deposits. This fact makes it extremely dangerous to assume that items found together in caves belong together, or that they even date from the same time period. Cressman failed to adequately account for this fact in his analysis of the oldest deposits in the Paisley Caves. Consequently, his interpretations have not been widely accepted. To verify the association of Pleistocene mammals with humans at the Paisley Caves we have to proceed much more cautiously with our analyses and interpretations. At this point, we have proven that people occupied the caves during the time that camels and horses were present in the region. We are now working to demonstrate that people had a part in the deposition of their bones in the Paisley Caves.

* Requires Adobe Reader to view