Archaeological Report
Harney County Oregon, Range Seeding Projects

Prior to undertaking this project the investigators grappled with the problem of defining just what exactly constitutes a "site." They approached the problem from a minimal definition stance; that is, what is the smallest amount of cultural material that can be considered a site. They chose an arbitrary operational definition of ten or more pieces of culturally modified lithics in 100 square meters for lithic scatters, with the explicit assumption that any petroglyphs/pictographs or architectural features encountered would be considered sites. This definition is a start toward systematizing inventory data, but it falls far short of completely defining the implicit assumptions operant in the concept of a "site".

Since lithic scatters are the most frequently encountered cultural remains in our study area, we will limit this discussion, for the moment, to consideration of those attributes that constitute a lithic scatter site. Lithic scatters result from knapping activity, the usual goal of which is to either produce, shape or sharpen tools or tool preforms, although an occasional flake might be produced as a knapper tests flaking instruments or material sources. Tools or flakes may be deposited through loss, discard or some form of post-depositional transportation. Thus lithic tools, broken fragments of tools, preforms and debitage are distributed across the landscape in a non-random fashion that reflects their genesis and post-depositional influences.

Traditional archaeological surveys, including this one, are often oriented towards the delineation of "sites" or the loci of intensive cultural activity as reflected in cultural remains. Cultural material however, is not clearly delineated into "sites" and "isolated" artifacts. It is more accurate to visualize the distribution of cultural material as a continuum, where artifacts are found in varying densities from one area to the next. Sites are usually intuitively and implicitly perceived in areas where there are "concentrations" of material, but how much material is required for a "concentration" and how can the limits of areal extent for those concentrations be defined? In this survey, the minimum amount of material for a site concentration is defined as ten pieces of culturally modified lithics per 100 square meters and site limits were intuitively assessed. This approach does not seem so useful as consideration of the relative density of cultural materials. Consider for example two hypothetical cases. In area "A" there are ten flakes in 100 square meters at the center of a tract of land one kilometer square. Area "B" is also one kilometer square, but this is a quarry area with densities varying from thousands of flakes per 100 square meters to one or two flakes per 100 square meters. Tn the center of area "B" there are ten flakes in an area of 100 square meters surrounded by areas with only two or three flakes per 100 square meters, beyond which the denser concentrations occur. Does the material at the center of area "B" constitute a separate site or is it part of one site that covers all of area "B"? If area "B" is all one site then we can say that the site density is the same for both areas "A" and "B": that is, one site per square kilometer. This is obviously misleading since cultural material is thousands of times denser in area "B".

"Non-site" archaeology is one solution to this problem. In a "non-site" survey, the goal is to record all cultural manifestations, regardless of their age, density or other characteristics. Once the data has been gathered, the density of cultural material can be mapped for the study area using all observations of material of certain ages or types. This is a more flexible arrangement that has yielded some interesting results on the pilot projects that have used it. In one case, for example, the number of projectile points and spent rifle cartridges and shotgun shells were found to correlate, suggesting that good hunting areas were approximately the same prehistorically as they are today. In another study (Dancey 1974), projectile points that would normally have been considered isolated finds were found to concentrate in an area where the topography funneled game animals through a narrow pass. Here we have an example of what could be considered a hunting "site," though the cultural remains don't fit any of the usual preconceived site definitions.

In most of the seeding surveys, only those lithic scatters dense enough for the investigators to consider "sites" were recorded. Isolated artifacts were also included, but apparently only formal tools such as projectile points or groundstone manos were recorded. In a couple cases historic materials were also recorded, but apparently only when the investigators felt the remains to be sufficiently old. The single exception to these generalizations is found in the House Butte survey which was conducted prior to the establishment of the minimal site definition used in the other surveys. Thus, small concentrations of cultural material were recorded as "sites" in the House Butte survey. Also, individual flakes were noted as isolated artifacts, as were historic materials including an insulator and a trap. To be consistent, only those sites that meet the minimal site criteria used on the other surveys are included in this analysis.

A total of 163 prehistoric and three historic sites were recorded during the survey of seventeen seeding projects. A total of 27,715 acres or about 11,222 hectares, were surveyed. The prehistoric site density, then, can be calculated at 1.45 sites per square kilometer. Table 2 shows the sites per square kilometer figure for each seeding. Site density varies widely between seedings; but as the table shows, two areas stand out for their high site density - Venator and Mill Gulch - with 8.75 and 5.19 sites per square kilometer respectively. As was discussed above, however, site density is not always a good indicator of intensity of use for an area because of the wide variance in site dimensions. One way to compensate for size variation is to compute the sum of areas occupied by sites and then compare that figure to the total size of the area surveyed. Table 3 gives the percent of area within each seeding that is covered by sites. Figure 4 illustrates the difference between measuring site density by the number of sites (as in Table 1) and measuring site density by area covered (as in Table 2). Project areas are ranked by each method in Figure 4, and it's clear that the two approaches yield totally different pictures of the intensity of prehistoric activity in the study areas. In all cases except one there was a change in ranking between the two methods, and in one case (Smyth Creek) the seeding project moved from tenth rank up to third place. The cause of this difference is obviously due to the fact that while some areas contain many small sites, other areas include a few large sites or some mix of large and small sites.

The largest site recorded is more than 100,000 times larger than the smallest site recorded during the surveys. Both the number of sites and the area covered by those sites will be considered in this analysis. A third variable, the density of cultural material within each site, would be useful here also, but this information was not consistently recorded during the surveys. In a few cases the total amount of flake debris was estimated, but in most instances estimates concerned only the range of density within the site. making comparisons between sites impossible.

Large sites are usually found adjacent to permanent water sources or lithic resources, while small sites are more variable in location, occurring near permanent, ephemeral or relic water sources or, in some cases, at locations quite distant from the nearest available water.

The distribution of sites in relation to living components of the environment is difficult to assess. No attempt has been made to relate site location to vegetation or wildlife habitat because of the massive changes these systems have undergone since the arrival of Europeans on this continent. The biased nature of our sample also prevented us from drawing any conclusions about site distribution in this regard. These seeding areas were specifically chosen because they were thought to be suitable for crested wheatgrass and because the current forage production was low, in most cases because the areas are predominatly covered with sagebrush now. Thus our sample is drawn from only a small fraction of the ecological diversity of the region.