The archaeological sites in the Owyhee uplands were cataloged by artifact feature types present at each site. These sites were counted and compared on the basis of features present. The distribution pattern of sites within the Owyhee uplands was analyzed, along with possible causes for the distribution.
In the Owyhee uplands of southern Malheur County, 511 archaeological sites are known (Appendix 4). Each of these archaeological sites can have multiple associated features, such as lithic scatter, lithic tools, ground stone, rock shelter and pottery. The counts and proportions of sites with some of these features show the frequent occurrence of lithic scatters (94.1%) and the low frequency of elements like pottery, pit houses and woven items (Table 2). These differences are partly a result of time depth (the extent of time over which people have made stone tools and pottery are different, so stone tools will show up more frequently) and partly a result of the lack of excavations. While lithics have been made into tools since people started inhabiting the Owyhee uplands, pottery was introduced around 500 AD in southeastern Idaho (Butler 1983). The only pit houses and woven items known for the Owyhee uplands are the result of two excavations. Items which only become visible in the archaeological record following excavations are found in low frequency in this summary because only four partial excavations have been conducted (35ML6, 35ML7, 35ML8, 35ML181). At seven other sites test pits have been dug where the sites have been potted (disturbed by collectors). Therefore surface features visible at surface sites are better represented in archaeological records from the Owyhee uplands. Petroglyphs, which are the focus of this study, only occur at 4.9% of archaeological sites in the Owyhee uplands.
The distribution of known archaeological sites in the Owyhee uplands clusters in two areas with greater concentrations (Figure 4). These are in the Trout Creek Mountains (southwestern corner) and along the Owyhee river. The rest of the study area is covered by scattered sites at low densities. The pattern seen in the region may not accurately reflect a thorough site distribution for this region, but it does reflect some of the trends in site distribution.
It would be unrealistic to say that the distribution of sites reflects settlement distribution, because the known sites are not a result of an unbiased sample. Some areas have been extensively surveyed while other areas have been surveyed on a sporadic basis or are relatively unexplored. The Bureau of Land Management (BLM) manages approximately 63% of the land in Malheur County, so their work and records are extremely important for understanding the archaeology of the area. The areas of private land around Jordan Creek and the town of Rome are not well represented in archaeological data because there is no obligation for private landowners to report archaeological sites. Most of the archaeological work which has been done in the county has been under the direction of the BLM on BLM lands. Archaeological surveys have been conducted for long swaths along the Owyhee River from Rome to the reservoir, an area very accessible for rafts. However, the sites on the mesa or less accessible river courses have been sporadically noted in BLM records, only as a rancher or BLM employee tells the archeologist that they have seen a site. It is possible that hundreds of sites wait to be encountered on isolated mesa tops and in smaller river drainages. Throughout BLM land, archaeological surveys are done with other goals in mind: e.g. fire rehabilitation projects, placement of a pipeline, reservoir or fencing or infrequent land transfers. These factors contribute to the distribution of known sites (Figure 4) but do not accurately reflect the density of prehistoric habitation.
Another factor which affects site distribution is site preservation. While lithic scatters may remain in the landscape for extended periods of time, other site types may disappear from the record. Throughout the area, there is very little soil accumulation, so sites remain exposed on the ground surface. The only locations with soil accumulation are the flood plains of the rivers. It remains unclear how the preservation process works in the Owyhee uplands, but it has an effect on the nature and distribution of sites. Sites along river courses are also disturbed by the wide and erratic fluctuation in river flows.
Despite inaccuracy in the distribution pattern, one obvious trend in site distribution emerges. The greatest site concentrations are in areas with contemporary perennial water sources. An arbitrary distance of 1500 meters away from perennial water sources was chosen and plotted in comparison to the distribution of sites (Figure 5). Out of the archaeological sites in the Owyhee uplands 361 of the 511 fell within 1500 meters of perennial water sources. The Trout Creek Mountains (or Oregon Canyon Mountains, southeastern corner of Malheur County) and the Owyhee River system show concentrations. Stream systems starting in the Trout Creek Mountains are fed by higher rainfall than areas of the mesa, but in general they run out of the mountains onto the mesa and dissipate.
Rock art occurs at 4.9% of the archaeological sites in the Owyhee uplands (Figure 6). As the focus of this study, a selection of petroglyph sites are examined based on rock art designs, location in the landscape and associate archaeological features.
Initially the recording of archaeological sites with rock art for this project was based upon the whole of Malheur County. However, current political boundaries do not coincide with features of the landscape, namely drainage basins and plant communities. Of the twenty eight known archaeological sites with rock art in Malheur County, Oregon, twenty five lie within the Owyhee uplands (Appendix 3). The twenty five sites from the Owyhee uplands can be analyzed. While a comparative analysis with surrounding areas would be interesting, the small sample size of sites outside the Owyhee uplands would make a comparison with adjoining areas unrealistic at this time.
Petroglyphs in the Owyhee uplands are made on exposed basalt. All locations that the author has visited have hard, fine to medium grained basalt without air holes. Although the area is largely basalt, very little of it is fine grained or exposed. The basalt flows are largely covered with sagebrush. In addition not all rock in the Owyhee uplands is basalt. Rhyolite, tuffs and sedimentary deposits are also found. Exposed basalt on which petroglyphs are made in the Owyhee uplands includes rim rock, blocky boulders and water worn boulders.
Rock art design types for the Great Basin were designated by Heizer and Baumhoff (1962). In the Owyhee uplands all the petroglyph designs fit these broad classifications: Great Basin Representational, Great Basin Abstract with the subdivisions of Rectilinear and Curvilinear, Pit and Groove and Great Basin Scratched. Out of twenty five sites, the styles were evaluated at the sixteen sites that were visited (Table 3). All of the sixteen sites include elements of the Great Basin Abstract style executed by pecking. Ten sites have representational elements and three sites have pit and groove elements executed by pecking. Scratched elements are present at five sites. Of these sites, there is no systematic duplication of representational elements between sites. Remnants of pigment were found in two petroglyph designs at one location.
Thirteen of the rock art sites were recorded in greater detail, with information on every face and design. These sites range in size, relative to the number of rock faces used and the number of petroglyph designs (Table 4). The smallest sites have five faces and twenty elements or six faces and seventeen elements. The largest site, 35ML1045, has 163 faces and 584 designs. Between the extremes are many sizes of petroglyph sites.
Of 1969 designs recorded at thirteen sites, the majority are abstract in style (81.4%). Small percentages represent the other styles (4.3% are representational, .5% pitted, 1.6% scratching, 11.8% dots). Since the proportion of elements other than abstract are extremely low at all sites, barring 35ML1057, the stylistic classifications based upon the method of production and subject matter do not assist in classifying the sites in the Owyhee uplands. As Francis (2001) suggests for the Great Basin, the different chronological periods Heizer and Baumhoff (1962) designated for these styles are not apparent. There are no regular instances of superimposition or differing coloration which would lead the author to place these stylistic classifications into different chronological periods, even at petroglyph sites in the Owyhee uplands where abstract, representational and scratched elements are all present. However the stylistic classifications can be used as a reference frame to describe the designs found in the Owyhee uplands.
The majority of petroglyph designs in the Owyhee uplands are abstract. This designation corresponds to the Great Basin Abstract style. Abstract designs could have multiple meanings, both in current interpretation and intentions of the original artists. The pecked designs which are not clearly similar to humans, four legged animals or hand or footprints are considered abstract. Designs that could be the sun, plants or animal tracks might have other interpretations. Since no ethnographic records connect the modern interpretations of these designs to the artists intentions, the designs were called abstract. This does not mean that the designs had no meaning when they were made, just that the meaning has been lost with the passage of time.
Anthropomorphic or human like figures have been well documented in the rock art of surrounding areas and, unlike abstract designs, are easy to classify. Despite the low percentage (4.3%) of all representational figures in designs of the Owyhee uplands, the anthropomorphic designs deserve comment. Three different types of anthropomorphic figures are found. These are stick figures (Photo 9), stick figures with lines coming out of their heads (Photo 10), and stick figures where the body of the figure is represented by a circle (Photo 11). The first of these types is recognized throughout the world. The second type is recognized as occurring on the Columbia Plateau and along the lower Snake River in Idaho (Boreson 1998, Nesbitt 1968). The third type is often called the "shield-bearing warrior", this design is found to the north and east of the Owyhee uplands along the Salmon and Snake Rivers (Boreson 1998). The human like figures in the Owyhee uplands reflect styles which are known to occur in areas to the north and east.
The scratched designs in the Owyhee uplands consist of thin lines, probably created using a sharp rock. Common compositions are zigzags, vertical or horizontal lines, crosshatching and fans. However there are other compositions including one four legged animal (Photo 12). The Great Basin Scratched style was made with a sharp rock. "Its elements are straight lines, sun figures, and crosshatching" (Heizer and Baumhoff 1962:208). On the Columbia Plateau, similar compositions of lines, crosshatching and fans have been recorded (Boreson 1998).
Pitted designs were only found at one (35ML1050) of the 16 sites where the elements were detailed (Photo 13). Thus, pitting accounts for an extremely small percentage of the petroglyph designs (0.5%). However two of the riverside petroglyph sites 35ML130 and 35ML168, which were visited by the author, had very deeply pitted petroglyphs on a few boulders (Photo 7). The pits, or circular depressions, which seem to be randomly placed on a rock face, vary in depth from one to multiple centimeters. Many of the pits are very dark in color, similar to the basalt around them. Based on the coloration, or patination, Heizer and Baumhoff (1962) proposed that the Pit and Groove style, which they found at six sites in the northern half of Nevada, was the oldest petroglyph style. Pits are also found on boulders at the rivers edge along the Columbia river system (Boreson 1998).
The antiquity of the Pit and Groove style described by Heizer and Baumhoff (1962) was based on relative dating methods. They utilized the coloration, or patination, of the petroglyphs as a method for determining age. Both patination and superimposition are relative dating techniques (Keyser 2001). These techniques can be employed primarily on individual panels of rock art to determine which designs are older and which are younger. Older designs will be a darker shade, under a newer design, or both.
Petroglyphs in the Owyhee uplands could provide a few clues which could be used in relative dating, if the petroglyphs differed significantly in composition or in method of construction. However the instances of differential patination and superimposition were not frequent and did not show a consistent trend. Often designs which could be classified as abstract were superimposed on or lighter than other designs which could be classified as abstract (Photo 14).
Petroglyphs are located on either basalt boulders or basalt rim rock. The smoothest boulders or rim rock at the petroglyph sites are often the ones chosen for petroglyphs. Additionally, petroglyphs were made on the clearest side of the rock, not the sides where the most lichen and moss accumulates. Petroglyphs in the Owyhee uplands were made in accessible and visible locations.
While it has been noted that the petroglyphs occur on both basalt outcroppings and basalt boulders, the sites and the individual rock faces are in accessible locations. Not only is it possible to walk up and around all of the cliff faces, it is possible to scramble through the boulder piles. More importantly, almost all of the designs could have been made from a comfortable sitting or standing posture because they are on accessible rock faces. Most sites, including 35ML850, 35ML1050 and 35ML1054, had no awkwardly placed petroglyphs. However, occasionally some designs are slightly out of reach for a five foot five inch recorder. At 35ML1044 two petroglyphs out of 293 were above the height, on a vertical face, which the recorder could reach. Elsewhere at the site, one panel required climbing up about a meter between two rocks, however the face itself could have been used as a seat (inclination 50 degrees from vertical). For awkwardly located panels, 35ML605 also comes to mind, one of the seven rocks, used at this location for petroglyphs, is only accessible by leaning over a ledge from above. This rock with five designs, including one that looks like a deer (Photo 15), was about four meters off of the ground. In general, petroglyphs could have been made from a comfortable location to stand or sit since they are at heights easily reached from these positions.
The visibility of rock art can provide clues to the audience that viewed the art. Art along a trail could be visible to everyone, while in the dark recesses of a cave it would be less visible. In the Owyhee uplands visibility of rock art is generally good. Crisp images can be seen from 100 meters away on both basalt outcroppings and basalt boulders. Some lichen filled or patinated images are now difficult to view or photograph, but they would have been bright when created.
Rock art sites in the Owyhee uplands do not contain only art, they can be associated with other artifact feature types. Compared to all archaeological sites, the features found at petroglyph sites occur in different frequencies (Table 5).
The artifact feature types which are found at archaeological sites with rock art are lithic scatters, rock shelters, ground stone, rock features and lithic tools.
Petroglyph sites are more frequently found in association with perennial water than archaeological sites in general. Eighty four percent of petroglyph sites (21 of 25) are within 1500 meters of perennial water sources whereas only 72% of all archaeological sites (369 of 511) are within the same distance.
After the discussions of design and accessibility, it is important to discuss the sites as a whole locality used by native inhabitants of the area. With twenty five sites to discuss, it is helpful to make an initial separation between sites. All archaeological sites with rock art in the Owyhee uplands lie within an area that historically was occupied by one group, the Tag?t?ka, tuber eaters (Steward 1939). While some of the sites with petroglyphs near the Nevada border are at elevations between 1751 and 2000 meters, there is not enough rainfall for mountain big sagebrush, so all of the petroglyph sites are within one vegetative and climatic zone. These being the same, the sites will be discussed in groups by their location in the landscape, specifically whether they are located on the mesa, in the canyons or on the mesa at a water tub.
Before breaking into these groups, it is important to note that none of the twenty five archaeological sites with petroglyphs are known to have pit houses, ceramics or woven material. This is not surprising since each feature is known to exist at fewer than four sites. They have also not been subjected to intensive investigation for habitation sites.
Eleven of the twenty five petroglyph sites in Malheur county are located along the Owyhee River and its tributary, Jordan Creek. The archaeological sites are found on the flood terraces. Some of the sites are on the terrace closest to the river, while others are on terraces over 20 meters elevation above the river. The riparian band along the river is very narrow (no more than five meters), but it contains willows, wild roses, currants and other riparian vegetation. High variability in water flow tends to scour the river banks of trees and large woody vegetation. The Owyhee River has some fish species year-round and prior to reservoir construction supported yearly fish runs. The terraces along the river are flat and open in some areas, despite the general constricted nature of the river. On the terraces away from the river's edge, the vegetation is basin big sagebrush because of the deep soils. All of the rock art located along the river is in the form of petroglyphs on water worn basalt boulders on flood terraces.
Sites along the river are not situated in the narrowest canyons, they are on fairly wide terraces, often where an intermittent drainage cuts down to the river. Much of the Owyhee River has limited access because of uninterrupted basalt cliffs which can tower over fifty meters. Where intermittent drainages join the river, the terrace area is often wide and the stream course has usually broken through the basalt. These locations provide access routes between the river and mesa as well as wider terraces; these spots are currently used for camping spots by river rafters.
Because of their location along the river, these were some of the first archaeological sites to be recorded and pillaged. Most of the large caves along riverside sites in the region have been potted. Artifacts sitting on top of the ground, like projectile points and ground stone have also been picked up. Petroglyphs located on wide terraces where rafters camp have occasionally been vandalized. Because they are near the river and visible from the water, some smaller sites have been recorded. The only recorded sites containing only one or two boulders with petroglyphs are beside the river; most recorded sites have many more boulders. This does not make small sites unique to the river course, but reflects upon how different areas have been documented. The river banks have been canvassed more thoroughly than the mesa and small stream drainages.
Sites 35ML105, 35ML130, 35ML144, 35ML153, 35ML156, 35ML164 35ML168, 35ML196 35ML197, are 35ML692 are located along the Owyhee River. 35ML1057 is beside Jordan Creek. While most of these archaeological sites have more than one feature class, four sites only have petroglyphs: 35ML164, 35ML196, 35ML692 and 35ML1057 ( Table 6). In the case of 35ML1057, any other features which might have been associated with the petroglyphs have been removed by a long history of farming and the recent construction of a levee. Two boulders with petroglyphs are known to have been incorporated into the levee which was built between 1965 and 2001.
The remaining seven archaeological sites with petroglyphs along the Owyhee River are associated with lithic scatters. In addition, three are associated with ground stone (35ML105, 35ML156 and 35ML168) and one is associated with a rock shelter (35ML130). While absence of ground stone is not diagnostic, because collectors remove it, its presence is diagnostic. The four sites with ground stone or a rock shelter are indicative of habitation sites. Therefore four of the archaeological sites with petroglyphs were associated with habitation. On the other hand, four sites have only one artifact class, the petroglyphs. These differing associations with the petroglyph sites make generalizations difficult. In the Columbia Plateau region, petroglyph and pictograph sites are located within a mile of an archaeological site which was used as a fishing place or campsite. Continuing further north into British Columbia, "petroglyph sites appear to be associated with subsistence activities, while those located on the Fraser River are thought to be closely related to fishing" (Boreson 1998:611).
Of the riverside sites, only two have been extensively recorded: 35ML153 and 35ML1057. From this small sample it is currently unrealistic to examine the counts of petroglyph faces and elements as a distinct grouping, different from the other petroglyph sites with this information (Table 4).
From the sample of twenty five archaeological sites with petroglyphs in the Owyhee uplands, eleven are located on the mesa tops. Mesa top sites are located both along basalt rim rock (or out cropping) and in boulder fields (Photos 16 and 17). Some sites are a combination of both locations as the rim rock has broken over time. On top of the mesa, the Wyoming big sagebrush vegetation dominates because of shallow soil, however at the lower edge of most rim rock, basin big sagebrush predominates in the deeper soil, with its associated currants and basin wild rye.
The eleven mesa top petroglyph sites overlook water sources, normally perennial, from higher locations. At two of these eleven archaeological sites only petroglyphs have been found. These are 35ML992 and 35ML1049. 35ML992 is located about 500 meters away from 35ML1044 and has a permanent water source. 35ML1044 lacks a water source but overlooks wetlands. Essentially, the small site 35ML992 could be considered part of the larger 35ML1044. 35ML1049 is perched on the edge of the mesa overlooking the Owyhee River 600 meters (2000 feet) below. The site runs along the near edge of the basalt rim rock, next to where a very small intermittent drainage breaks through the basalt rim rock, and thus is a means of access between the mesa and river below. 35ML1049 has a paucity of petroglyphs, there are ten faces with one or two elements each, but only three faces remain legible.
If these two sites are set aside, the remaining nine archaeological sites with petroglyphs on the mesa top also have rock walls or rock circles associated with them. All of these nine sites overlook water. The rock walls are located on the mesa top, so in profile they are higher than the petroglyphs and both are higher than the water source (Figure 8). Of these eight have associated lithic scatters (only 35ML1053 lacks a lithic scatter). These sites which unite the combination of petroglyphs, rock features and wetlands or perennial water sources are 35ML842, 35ML850, 35ML1044, 35ML1045, 35ML1046, 35ML1050, 35ML1051, 35ML1052, and 35ML1053 (Table 6).
These nine archaeological sites provide a large comparative sample. The rock features, petroglyphs, wetlands and lithic scatters associated with the sites can be compared. Rock features are very unique in the Owyhee uplands. Of 511 known archaeological sites only twenty four have rock features, less than 5%. Rock features have been documented in Owyhee County, Idaho by Mark Plew (1976) but it is unknown in what archaeological contexts they occur. In the Owyhee uplands, the only well documented rock features are found in association with petroglyphs. The context in which the other two thirds of the rock features occur is unknown.
Two types of rock features are found in the Owyhee uplands, rock walls and rock circles. Rock walls are built to follow the contour of the land and often fill in gaps where the rim rock is low or crumbling (35ML850, 35ML1044, 35ML1046, and 35ML1050). The walls would only serve as an effective "fence" from the downhill side, from above they blend into the landscape (Photo 19). Rock circles are built near or at the edge of rim rock, and often stand slightly above the height of the rim rock (Photo 20). Sites where there are rock circles near the edge of the rim rock are 35ML842, 35ML1045, 35ML1050, 35ML1051, 35ML1052, and 35ML1053. The one exception to this is at 35ML850 where the circle is located in the middle of a shallow draw between rock walls (Photo 21).
The lithic scatters associated with eight of these nine archaeological sites have multiple rock types. While obsidian is very common in some lithic scatters, others are very mixed (Photo 22). The flakes in the lithic scatters are composed of obsidian, fine grained basalt, painted jasper, quartzite, opalized chert and siltstone. Many of the components in the lithic scatters are not observed in resident geological deposits at the sites where the principle rock is basalt.
Field notes for upland petroglyph sites record that almost all have currants (Photo 23). Most of these sites had basin wild rye as well, which is expected, because they both grow in deep soils (Photos 24 and 25). Both of these edible plants have ripe fruit and seeds, respectively in mid to late summer. While the sites might have other botanical features in common, a botanical collection was not made at the archaeological sites. An additional challenge is that overgrazing at the end of the 19th and beginning of 20th century resulted in a reduction of some species or reduced ranges of other species.
Water tub petroglyph sites are located on the mesa top, however they are broken off into their own subsection because of the unique nature of the water tub and their dissimilarity to other mesa top petroglyph sites. A water tub acts like a trough and holds runoff water. Water tubs are located in existing intermittent drainages below shallow drop-offs formed by the drainage cutting through basalt layers (Photo 18). Because of the short drop, the water in the intermittent drainage digs a hole. This results in a small depression which holds water after the drainage runs dry.
Three petroglyph sites are located at water tubs, these are 35ML605, 35ML688 and 35ML1054. They are situated at very small rock outcroppings in intermittent drainages. Additionally the three sites are isolated on the mesa because, although they are close to the Owyhee River, a sheer cliff separates them from it. While the three sites have petroglyphs, only one of the sites, 35ML688, has an associated feature, a lithic scatter. 35ML688 is also significantly larger than the other two sites on the basis of rock art. There are sixty nine panels at 35ML688 while 35ML605 and 35ML1054 have seven and six panels, respectively. 35ML688 also covers a greater area, with more exposed basalt faces, however it is unclear why 35ML688 is a larger site. Perhaps the drainage which feeds this site runs more regularly. It appears that the attraction of these sites was the water, available in the water tubs when the rest of this mesa region was dry.
From the previous discussion, it has become apparent that sites are similar based upon their location in the landscape. Within the Owyhee uplands, the twenty five archaeological sites with petroglyphs are found in three locations: along the perennial rivers, at water tubs and on basalt overlooking water holes or wetlands (Figure 12). For the three distinct locations, the landscape as well as the archaeological remains present are slightly different (Table 7).
In general, the sites along the river vary in size and might have been associated with habitations and therefore wintering along the river or harvesting fish. The petroglyph sites at water tubs are relatively small and only one is associated with other archaeological remains, a lithic scatter. The archaeological sites which overlook wetlands are characterized by petroglyphs, lithic scatters, rock features and are perched above a water source.
Since there is no reliable dating method for petroglyphs, rough dates for these sites with rock art must be constructed in a round about manner. The lines of evidence which could be used include chronologically diagnostic projectile points, ethnographic records and relative dating methods.
Of all the lithic scatters visited, only one site produced a datable projectile point. This point was from 35ML1050, an extensive site on the mesa top overlooking a water hole. Interestingly, this point was not found amid the lithic scatter of 35ML1050, which covers approximately 16 acres, but at least 20 meters from its edge and below the rim rock. From a field sketch, it was identified as being a broken Elko Corner Notched point. The identification was made by William Andrefsky (personal communication 2001). This type of point was used over a range of time in the northern Great Basin and an acceptable time range would be from 4,300 to 3,300 years ago. This one point can say that 35ML1050 was visited or inhabited during this time period, but it in no way places a limit on people using the site before or after this date.
In recording the ethnography of Native American groups of the northern Great Basin, Steward (1941) notes no creation of rock art. In fact Steward supposed that rock art could possibly have been made by an earlier group to occupy the region. Since the Shoshone and Paiute were relative newcomers to the areas of southeastern Oregon and Idaho, it is possible that the petroglyph sites were not their creation, rather they were already a part of the landscape when they entered the region.
While Heizer and Baumhoff (1962) propose different dates for abstract, representational and pit and groove petroglyphs, in the Owyhee uplands both abstract elements and one representational element have accumulated varnish, since they were made, that has turned them the color of the rock (Photos 26 and 27). These elements look like the pit elements in the region. As Dorn (2001) mentions, the oldest relative date for a petroglyph is when the varnish is the same color as the rock. If the model of Heizer and Baumhoff (1962) is used, all of the pecked styles for the Owyhee uplands have been used between 5000 and 3000 BC, the dates for the Pit and Groove style. All of these styles then continue through time, because lighter versions are found at the sites, 35ML1050 even has light colored pitted petroglyphs. While Heizer and Baumhoff (1962) placed the earliest date for petroglyphs at 5000 BC, the dates from the Columbia Plateau are later, because the pit and groove forms are excluded. From comparisons between portable art and rock art designs, "the rock art tradition at The Dalles [north central Oregon] emerged as early as 1500 BC out of rudimentary pit and groove forms" (Boreson 1998:617).
Relative dating methods include looking at the color of the petroglyph. On basalt surfaces, petroglyphs get darker with age (Dorn 2001). Petroglyphs in the Owyhee uplands show a range of coloration from almost white to the color of the basalt. The oldest date this method allows is when the petroglyph is the same color as the rock background (Photo 26). Another method of dating is superimposition. There are a few clear examples from the Owyhee uplands (Photo 28), most superimposition is difficult to distinguish because both the early and late elements are of an abstract style (Photo 14).
While this study reports vandalism at some archaeological sites with petroglyphs close to roads or the river, the vandalism generally is not destructive (Photo 31). Rather individuals added additional modern glyphs in the form of pictures or names to rocks around those with petroglyphs and sometimes on the same rock face as where petroglyphs are already present (Photo 32). Some of the smaller rocks which had petroglyphs on them are probably not at the sites, in one instance rocks of a manageable size have been incorporated into a monument (Photo 33)
The only well documented rock features found in the Owyhee uplands were found in association with petroglyph sites. There are two types of rock features, rock walls and rock circles. Generally rock walls follow the contour of the land and create an impenetrable barrier from below. Rock circles are often situated at the edge of the rim rock. These archaeological features are located in positions which are above water holes or wetlands. Both rock walls and rock circles are normally constructed one rock thick and with three or more rock layers vertically (Photo 29).
The rock circles at 35ML1050 are situated near the edge of the mesa. The circles vary in size from one to three meters across and the highest sections currently stand 80 centimeters high (Photo 20). Placement of rock features near the edge of the mesa, where they overlook a water hole or wetland is also found at 35ML992, 35ML1045, 35ML1051, 35ML1052 and 35ML1045. At 35ML1052 there are 14 different rock circles on the edge of the rim rock (Photo 30).
The walls at 35ML850 follow the contour of the hill for 341 meters (Photo 19). Nine segments, ranging between seven and seventy eight meters apiece are interspersed with low rim rock. In places the walls stand over one meter tall. The labor required to build the wall, considering its length and height, suggest repeated or prolonged use of the site. Site 35ML850 is at the top of a draw which overlooks a meandering creek. From the perspective of walking up the draw, the rock wall is on the right hand side. The other side of the draw has tall rim rock. The rock walls could work like a funnel. At the end and center of the funnel there is a rock circle. This circle is 118 meters away from where the rock wall ends.
The occurrence of rock features at petroglyph sites in Nevada was noted by Heizer and Baumhoff (1962:221):
As for fences, we find these at four sites; of these, all have Great Basin Abstract style (both Curvilinear and Rectilinear), three have Great Basin Scratched style, two have Great Basin Painted style, one has Great Basin Representational style, and one has Pit-and-Groove style. Three of the four sites are marked as suitable for antelope [pronghorn] corrals, and it is likely that the fences were parts of the corrals. . . . Blinds are found at five of the sites and associated with a diversity of petroglyph styles similar to that found in combination with fences. Two of the blinds are found together with fences and must have been used in connection with corrals.
While the context in which Heizer and Baumhoff found the rock features may be different, as there are no corral like features in the Owyhee uplands, the recognition that they occur alongside rock art is important.
Rock features cannot be dated in traditional manners, since they do not contain organic material. The rock features in the Owyhee uplands are covered with lichen, but lichen is a difficult dating method for rock art because growth rates are difficult to calibrate in arid regions (Dorn 2001), the same would hold true for rock walls. While most of the rock features have crumbled and look old, this is no scientific method. While Heizer and Baumhoff (1962) noted rock features at seven of fifty eight sites, they did not suggest dates. Agenbroad (1976, cited in Andrefsky and Presler 2000:17) notes that rock hunting alignments are associated with small surface lithic scatters near springs beginning around 8000 BP. The Uplands Camp Phase also has an increase in the occurrence of ground stone which suggests an increase in vegetative food processing (Andrefsky and Presler 2000).
The only other clue available for the dating of rock features comes from ethnography. The Shoshone and Paiute who lived in adjoining areas are only known to have constructed hunting blinds and corrals out of sagebrush, a plentiful desert plant which is easily acquired. Thus the ethnographic records present no leads to the dating of rock features, except that these groups probably were not responsible for the construction of rock features. Therefore it is likely that no rock features were made during the last 1000 years. In sum, a possible range of dates from 8000 to 1000 years ago can be placed on these rock features from association with lithic scatters and ethnographic records.
Considering past research, it is difficult to determine the cultural pattern of the Owyhee uplands. No specific ethnographic study was conducted on the area from which a direct historic method can be employed. Furthermore, the continued debate over a Numic spread questions the cultural continuity of the Northern Paiute groups. Therefore, petroglyph sites within the Owyhee uplands cannot, with certainty, be ascribed a cultural affiliation nor can the rock art be dated by current methods. The following are the models which the examination of the petroglyph sites in southern Malheur County might support.
The model of aggregation presented by Conkey (1980) might be applicable to the hunter-gatherers of this region. Since aggregation and dispersal occurs on a regional scale, it is necessary to include many known sites from the study area. An aggregation site would, by definition, occur in an economic, social or ritual context which would generate the population concentration. By demonstrating that sites with petroglyphs in the Owyhee uplands resulted from population aggregations, it could be postulated that a social or ritual factor contributed to the aggregation and thereby the creation of petroglyphs.
Important to aggregation are areas with the economic capacity for sustaining an aggregated population, especially for any prolonged time period. This requirement is met in neighboring regions by the pinenut harvest in the Great Basin, and in the Snake River basin by salmon runs and camas plains. In the Owyhee uplands it would be necessary to determine resources which could support an aggregation. Possible resources would include camas concentrations, grass seeds, antelope hunting and summer fishing runs on the Owyhee River. These sites would also reflect the congregation of population in remains of camps.
Research has shown that the petroglyphs are in locations which could have been used for hunting or fishing. The possible hunting sites would not be able to support an aggregated population because there is no guarantee that game would be present. Riverside sites could have been used as locations to fish, but not all are known to be campsites as well. If riverside fishing habitation sites are large, the model of aggregation (Conkey 1980) might be applicable to these rock art sites in the Owyhee uplands.
The vegetative food resources which are available in the Owyhee uplands occur in dispersed patches throughout the landscape. The small concentrations are found in areas with slightly higher amounts of moisture, eg. near springs. This factor would be difficult to determine because of vegetation changes over time.
After conducting research, it seems likely that vegetative foods were harvested by yearly movement around the landscape to known locations. The scarcity of resources in any one location might not have allowed for prolonged stops at any of the locations. A frequently used location might be one where vegetative food resources were available and where animal food resources were possibly available.
The animals which are native to the Owyhee uplands have the same constraints as humans to surviving in the habitat. They are constrained in summer range by the location and quantity of watering spots. Pronghorn (Antilocapra americana) commonly go to drink several times a day (Lubinski and Herren 2000). People who want to hunt could optimize the limited access to water resources by waiting at locations where animals will be required to pass. Although the Owyhee River runs year round, many of its tributaries are intermittent. Another limit is the nature of the river and its main tributaries. Many run in deep canyons which have vertical cliff sides. This makes locations with access from the rim down to the water important for both humans and animals. Experience with the landscape has shown that most of these accessible paths occur where a smaller tributary has cut a path to join a larger tributary. This landscape of the Owyhee uplands might create ready made traps in the draws which large animals must pass to reach water when it is scarce.
Very little can be said about the opportunistic procurement of small game. Small game procurement does not rely upon the landscape, but rather happening upon a rattlesnake, sage grouse or rabbit that is nearby and taking advantage of the situation.
If petroglyphs are associated with large game procurement, the locations would be those which provide water. So the petroglyphs would be near water holes, springs or canyons leading to and from water sources.
All of the petroglyph sites recorded by the author were shown to be near water sources which remain after the surrounding area is dry. While none of these sites are associated with constrained canyons leading to or from water, game could be herded from watering holes toward constricted locations at sites 35ML850 and 35ML1046. Despite a general lack of constricted locations, many of the rock circles that sit on rim rock overlooking the water could have been used as blinds for either sighting or hunting game; sites with rock circles include 35ML842, 35ML1045, 35ML1050, 35ML1051, 35ML1052 and 35ML1053. The water's edge might also have been used as a location to kill large game animals. Sites located on the Owyhee River and Jordan Creek might not be situated in positions which would take advantage of large game movements, but would have had fishing resources.
For both humans and animals in the Owyhee uplands, water is a limiting factor. During midsummer and drought seasons, water becomes scarce in this semiarid environment. Therefore it is logical that human populations would prefer their semi-permanent residence for dry months to be near a reliable source of water. To conclude that sites with rock art are related to water supply, the distance of rock art sites to water must be compared to the distance of other sites to water. Also these sites would likely represent extended or yearly habitations.
Research has shown that most of the petroglyph sites in the Owyhee uplands are located close to perennial water resources, but this is also a trend with the majority of archaeological sites from the Owyhee uplands, 72% are within 1500 meters of perennial water sources. It is difficult to show that the water was the reason that individuals were at certain archaeological sites, they may have been interested in the resources which are more prolific around water sources, such as animals and edible plants.
Much of the wide expanse of grazing land has little access to water. Access is limited both by water scarcity and by the tendency of perennial streams to be incised below vertical cliffs throughout the Owyhee uplands. To the north on the Columbia Plateau, "Pictographs (paintings) and petroglyphs (carvings) are found on rock surfaces such as rock shelter walls, escarpments, outcrops and boulders, generally near a permanent source of water" (Boreson 1998:611).