Environment, Prehistory & Archaeology of Mount Rainier National Park, Washington
Greg C. Burtchard

Chapter 4:
THE 1995 MOUNT RAINIER ARCHEOLOGICAL RECONNAISSANCE
(continued)

ENVIRONMENTAL CHARACTERISTICS AND PREHISTORIC SITE DENSITY

Use of an environmentally grounded approach toward understanding the archaeological record of the Pacific Northwest is not particularly new. Environmentally based predictions have been employed by federal agencies for some time to attempt to determine the location of prehistoric remains in advance of field inventory. Though causal connections are seldom specified, attributes such as elevation, slope, solar exposure, landform type and (especially) distance to water have been assumed to have conditioned human use of the mountains in the past and, hence, considered to be critically associated with the location of archaeological remains in the present (see Kvamme 1988:331-338). These predictions are then used to develop probability zones that serve to simplify survey obligations by directing inventory coverage differentially toward the higher probability landforms predicted by the model.

Elsewhere, I have been critical of these approaches largely because they tend to lack consideration of underlying causes for the patterns they purport to predict and are self-fulfilling in application, thereby hindering empirical falsification of the models' predictions (see Burtchard and Keeler 1991:107-122). In the Burtchard and Keeler study, Mt. Hood National Forest's predictive model (emphasizing distance to water, landform, and slope) was evaluated in light of the Forest's growing site database. While a strong association was found between the relative density of prehistoric remains to elevation and slope, only weak association was found with distance to water, and none with solar exposure. A tendency for site association with montane basins, ridges and saddles, sideslope benches, and valley floodplain terraces is largely subsumed by elevation and slope (Burtchard and Keeler 1991:115-117). Working with these results, Mt. Hood National Forest restructured site inventory procedures to accommodate critical resource distribution patterns related to a less restrictive range of environmental variables.

The questionable utility of non-causal models notwithstanding, it is not unreasonable to expect prehistoric populations to have selected campsites with an eye toward water availability, flat ground, solar gain and so on–particularly when such characteristics occur in association with economically productive habitats. This section examines the association between the current Mount Rainier site and isolated find database with a set of particular environmental characteristics commonly used in the aforementioned survey systems. Table 4.10 summarizes these basic categories for Mount Rainier National Park. Figures and associated discussion that follow illustrate the results. Patterned association between environment/resource zone is presented first to demonstrate the strong relationship emphasized throughout this report. As will be seen, available data also suggest a strong association with elevation (which, of course, is linked to differential use of subalpine and alpine zones) and with relatively flat ground. As in the Mt. Hood study, patterned association with distance to water, landform type and solar exposure is more equivocal, though the present small sample size may color the result. Please note that in order to maximize sample size, no effort has been made to distinguish between site types as has been done above and as modeled formally in Chapter 5. Prehistoric sites and isolates for which sufficient data are available are considered equally.

Table 4.10 Mount Rainier Site/Environmental Associations

Environmental/Resource Zones and Elevation

Throughout this report, it has been argued that high resource productivity of Mount Rainier's subalpine and alpine habitats should have served to focus human activity on these habitats during the prehistoric past. If so, these zones will display high site density relative to other montane habitats. The prediction is deductive in that it flows directly from environmental considerations rather than the archaeological record itself. Archaeological data provides means to test, rather than to generate expectations. While the quality of the test will improve as additional surveys are completed, available site data from Mount Rainier clearly are consistent with the predicted subalpine/alpine pattern.

Table 4.11 displays site count by the five major resource zones employed to characterize the Park–high energy floodplains, maritime forest, subalpine parkland, alpine tundra, and perpetual snowfields and associated glacial rubble. Area for each of these zones is computed from GIS data used to generate color fold-out Figures 2.10 through 2.13 in Chapter 2. Both site count and site density figures show the clear subalpine pattern. Even if site count for maritime forest were tripled to accommodate poorer surface exposure and lower survey fraction, site density results would remain below alpine and well-below subalpine zones. In my opinion, the pattern is far too strong to be product of either chance or site discovery biases.

Table 4.11 Mount Rainier Site Count and Density, by Environmental Zone

Figure 4.4 shows the environmental zone and site frequency relationship graphically. The graph is conservative in that percentages are computed on the basis of simple site count per zone. Had it been normalized to display site density per hectare as shown in the final column of Table 4.11, the contrast between subalpine and forest zones would have been even more pronounced; the contrast between subalpine and alpine zones, somewhat less so.

Figure 4.4 Site Relative Frequency, by Environmental Zone

Figure 4.5 shows the relationship between prehistoric site frequency and elevation. The similarity between this graph and Figure 4.4, of course, is due to the direct link between elevation shown here and environmental zones shown above. In concert, the figures reinforce the critical role played by upper elevation landscapes in prehistoric land-use practices and, hence, on the archaeological record of that use. Twenty-nine sites and isolates (76% of the sample) are situated in subalpine contexts. Adding sites in alpine settings increases the fraction to 87%. Furthermore, all of the maritime forest sites are located in upper elevation forest contexts between 4,000 and 5,000 ft or higher–locations that arguably provide ready access to subalpine environments and/or were used at times when the closed forest fringe was lower. In short, the upper elevation pattern is clear and robust, and, in light of the manner in which these habitats would have conditioned resource availability throughout the past, should not be surprising.

Figure 4.5 Site Relative Frequency, by Elevation Slope

Figure 4.6 shows the relationship between slope and site frequency in the Mount Rainier sample. Seventy-five percent of the documented sites are found on landforms with a local slope of 5o (8.8%) or less. Assuming that most lithic aggregations represented in the sample are associated with at least short term residence, the tendency toward selection of flat ground is understandable–flat ground simply is best suited to sleep and for conducting a variety of maintenance tasks. Slope is irrelevant, of course, for sites such as quarries (e.g., FS 90-04) and lost implement localities (e.g., IF 10-95), as well as for sites in disturbed, secondary context such as IF 11-95. Cliff and talus rockshelters such as FS 63-01, FS 86-02 and FS 72-02, while often occurring on steep slopes, have interior use spaces which are quite flat. Accordingly, the tendency for sites to be located on low gradient terrain, or to avail themselves of small flat-ground irregularities on otherwise sloping ground, is strong in the Mount Rainier sample (as elsewhere) and may be explained by the tendency to select such ground for a variety of task-specific and residential purposes.

Figure 4.6 Site Relative Frequency, by Slope

Distance to Water

In a similar evaluation of site distribution in the vicinity of Mt. Hood, we found a relatively weak association between lithic scatters and distance to water–47% within 50 m, 59% within 100 m (Burtchard and Keeler 1991:117-119). The water association in the Mount Rainier sample is somewhat stronger– 53% within 50 m, 75% within 100 m (see Figure 4.7). It is not clear, however, if the difference is simply due to small sample size for Mount Rainier or if the distinction is genuine. In dry environments, water can be a significant determinant of human movement and settlement patterns. Water is, after all, heavy, difficult to contain and required daily. In wetter environments such as the western and high Cascades, water imposes less of a constraint on human land-use options simply by virtue of its widespread availability. Because Mount Rainier offers a variety of water sources, the relationship between site location and access to water should be weak. Since water is important, it makes sense to settle near a source when it is handy to do so. In a place with many alternative sources such as Mount Rainier, however, it is equally likely that sites will be located well away from visible sources of supply. In the event the Park implements some form of predictive survey strategy, it is suggested that distance to water not be considered a key variable in structuring the survey design.

Solar Exposure

While solar exposure seldom appears in predictive site location models, it is not unreasonable to expect people to select sites with good solar exposure in cold environmental circumstances. Since Mount Rainier was almost certainly used during summer months, however, we would expect a tendency to maximize solar gain to be offset by countervailing considerations such as visual exposure, wind direction, natural rockshelter orientation and so on. Accordingly, we would expect to see little relationship between site location and solar orientation. Histogram Figure 4.8 suggests that this is so.

Figure 4.7 Site Relative Frequency, by Distance to Water

Figure 4.8 Site Relative Frequency, by Solar Orientation

Landform

A glance at the variety of landscapes represented in site/isolate summary Table 4.10 shows that Mount Rainier's prehistoric sites are found in a variety of upland settings. Other than re-emphasizing the tendency toward subalpine/flat ground site selection, little clear patterning is evident in the available data. Commonly represented in the sample are side slope bench settings, ridge crests and saddle passes between drainages, cirque basins, and, of course, cliff face and talus rubble rockshelters. Figure 4.9 below summarizes the landform variety characteristic of the Mount Rainier sample. As with solar exposure and distance to water, specific landform characteristics do not appear to be significantly associated with the location of prehistoric archeological remains.

Figure 4.9 Site Relative Frequency, by Landform Group

The most salient pattern that has been emphasized in this chapter is the clear site bias toward subalpine to alpine habitats as predicted by theoretically derived expectations and as observed during the early historic period. Basic prehistoric site and isolate data (as well as that for two historic period sites) have been provided in tabular form and plotted onto a general location map. Professional readers may wish to refer to individual site forms in the project's companion 1995 Reconnaissance Data volume (Burtchard and Hamilton 1998) for more complete site-specific detail. In the following chapter, site location and environmental data are combined with theoretical expectations regarding long-term human use of the region to develop 1) a site type and distribution model for Mount Rainier National Park, and 2) a general Holocene land-use model for the Park and the southern Washington Cascades.

Substantial attention also has been given to describing and interpreting lithic assemblages observed in the field and curated at Park headquarters in Longmire. Particular attention was given to deriving artifact classes, isolating and describing raw material patterns, discussing temporal indications of these and other relevant data, and considering several functional site taxonomies implied by patterned variation in artifact and material classes in the Mount Rainier sample. These considerations are formally developed into a site type taxonomy in Chapter 5.