AMD002_h.Inventory_strategy.pdf

The Umatilla National Forest covers approximately 1,400,000 acres in north eastern Oregon and southeastern Washington. It lies within a larger geographic area known as the Blue Mountain Physiographic Province of the central Columbia Plateau. The physical characteristics of this area have been conducive to the development of a diversity of cultural remains. Within the forest boundaries the known cultural resources include a wide range of both prehistoric and historic properties. The value of these properties lie in their potential to yield information about the lifeways of earlier people, their cultural adaptations to the environment, and about human behavior in general.

Cultural resources are fragile and non-renewablei when sites are destroyed a part of the human story is lost forever. In response to the need to protect this fragile cultural heritage the American public has passed a number of preservation laws and regulations. The National Historic Preservation Act of 1966, as amended, Executive order 11593, 36 CFR 800 "Protection of Historical and Cultural Properties" and 36 CFR 219 "National Forest System Land and Resources Management Planning" direct the Forest Service to establish a program to inventory, evaluate, and manage prehistoric and historic cultural properties on federal lands.

Inventory is the first step in the legal compliance procedure. When performed properly, it enables land managers to assess the potential effects of projects on cultural properties and to plan accordingly. This document describes the Umatilla National Forest Cultural Resource Inventory Plani its purpose is to explain the procedures followed from survey design to the final report. In an effort to complete an inventory of all National Forest System Lands (FSM 2361.02), in a cost-effective manner, the Umatilla inventory plan is based on a judgmental stratified sampling design with a concurrent random sampling strategy. The judgmental sample is based on general anthropological knowledge and is designed to allow the Forest to comply with Federal law and regulation whereas the random sample is designed to gather data necessary to formulate a statistically valid site location predictive model.

Environmental Background

The present day land surface features influence the environment and are a result of the geology and sequence of geologic events through time. During the Paleozoic era the land masses were under the seai this resulted in the deposition of great layers of sediment. Volcanic activity and related uplifting caused the gradual elevation of the Blue Mountains, as well as the coast and Cascade ranges. Massive lakes were created between the mountain ranges with interconnecting rivers. Extensive lava flows and sediment buildup drained the eastern area except for the high inland lakes.

The Tertiary Period resulted in the Clarno Formation of thick andesite flows, volcanic breccias, mudflows, lacustrine and fluvial tuffaceous sediments and lacustrine silts and clays. These formations contain an extensive fossil record of Tertiary plants and animals. The Miocene epoch again was characterized by volcanism. widespread columbia River Basalt forms the mass of the Blue Mountains between Pendleton and La Grande, Oregon. At Gibbon, Oregon, one exposure shows 2500 feet of basalt (Ehmer n.d.).

The Pliocene Epoch was a transition period between the volcanism of the Miocene and the glaciation of the Pleistocene. During the Pliestocene, continental glaciation pushed southward from Canada into Washington as local glaciers formed in the Blue Mountains. According to Baldwin (1964:107) the eastern unit is characterized by altered volcanic flows and sedimentary rocks and the effects of Pleistocene glaciation as evidenced by numerous cirques, glacial lakes, and moraines.

As the climate warmed and glaciers retreated, volcanic activity continued to uplift the Blue Mountains. Numerous local cones and vents like the Spring, Sugarloaf, Thimbleberry and Green Mountains are the result of this activi ty.

In recent times, aeolian and alluvial sediments have covered the basalt flows.

The most recent major event was the eruption and collapse of Mt. Mazama in the Cascades 6700 years ago. This catastrophic event deposited a layer of volcanic ash over the area which can be used as a time-marker when it appears in cul tural deposits. e. g ., the Marmes Rockshel ter, Stockoff basal t quarry, and other sites.

Soil formation can be divided into four major soil groups. These include: (1) residual soils from basalt or andesite which support predominantly ponderosa pine standsj (2) soils derived from volcanic ash over basalt or andesite residuum which supports predominantly mixed conifer standsj (3) soils of volcanic ash origin over bedrock or buried residuum from basalt, andesite or tuffaceous material which support predominantly lodgepole pine standsj and (4) soils derived from volcanic ash, over tuffaceous basaltic and sedimentary materials which supports stands dominated by white fir (Ehmer n.d.)

The present-day climate is influenced by the Cascade Mountains which act as a barrier to precipitation from the Pacific Ocean. Wide annual and diurnal variations in temperature occur. Summer temperatures occasionally exceed 1000

54 0Farenheit. Winters are generally mild, although temperatures as low as Farenheit have been recorded. Annual precipitation ranges from 20 to 65 inches mostly coming in the form of snow in the winter. Winter snowfall averages 100 inches at higher elevations.

Climatic variations over time have influenced cultural patterns. For example, the second occupation at the Stockoff site appears to coincide with the Altithermal, a period characterized by distinctly higher temperatures and drier conditions than at present. The site saw its most intensive utilization during this time. McPherson et al. (1981), suggest that "During the Altithermal, the resource base on Stockoff, and in the surrounding area, was probably broad enough to allow for the maintenance of a year-round exploitative economy."

Changes in hydrology have also occurred over time. There was much more water in the region during the Pleistocene than at present. Today there are four major drainage basins located on the Forest: the Umatilla, John Day, Grande Ronde, and Lower Columbia. Wi thin these drainage systems are several rivers and creeks of importance as anadromous fish resources. These include the North Fork of the John Day River, the Umatilla River, the North and South Forks of the Umatilla River, the Grande Ronde River, the South Fork of the Walla Walla River, the North and South Forks of Cable Creek, and Desolation, Camas and Hidaway Creeks. Natural lakes are few in number and small in size.

Present day plant species composition and vegetation patterns vary with regard to slope, aspect and elevation. Steep south facing slopes are characterized by grass covered openings with timbered stringers along the drainages. Idaho fescue, big sagebrush, bluebunch wheatgrass, and bitterbrush are examples of grass and shrub species located in these open areas. Grand fir, Douglas fir, and ponderosa pine are primarily found in the stringers. North facing slopes are characterized by a dense covering of mixed conifers interspersed with mountain alder (Hall 1973).

Above 5000 feet the high elevation terrain is a mosaic of forest, rock outcrops, and alpine meadows. The forest species composition is primarily subalpine fir and lodgepole pine. The plateaus of the northern forest at the 4000 to 5000 foot elevations are covered by dense stands of mixed conifers represented by grand fir, larch, lodgepole and ponderosa pines. At the same elevations the southern portion of the forest is characterized by dense stands of lodgepole pine and mixed conifers on north facing slopes. The south facing slopes are characterized by open stands of ponderosa pine and Douglas fir (Hall 1973) .

Modern wildlife includes Rocky Mountain elk, mule deer, black bear, cougar, bobcat, coyote, raccoon, porcupine, rabbits and hares, several species of squirrels and chipmunks, gophers, and others. Bird life is abundant. Upland game birds are the Blue and Ruffed grouse, Band Tailed pigeon, Mourning dove, chuckar, gray partridge and turkey. Raptors include Bald and Golden eagles, goshawks, Cooper's hawks, ferruginous hawks, prairie falcons and ospreys. In addition, there are ten species of woodpeckers, several species of nuthatches, bluebirds and chickadees.

Culture History

A cultural chronology specific to the Blue Mountain province does not exist.

Most of the archaeological investigation on the Central columbia Plateau has occurred along the columbia and Snake Rivers and their tributaries. This work has resulted in a fairly complete understanding of riverine adaptations through time while the upland areas remain poorly known.

The known evidence of prehistoric occupation within the Umatilla National Forest is sparse. Most of the present data are derived from 225 project-specific cultural resource inventories conducted by the Forest Service since 1977 and from a few contracted inventories for proposed land exchanges (Mead and Cinadr 1976i and Mead 1975) .

Combined archaeological data from inventories and text excavations within and adjacent to the Umatilla (Berryman 1987), Malheur, and Wallowa-Whitman National Forests indicate occupation of the area as early as 7000 to 10000 years ago.

Perhaps, the most notable is the stockoff site (35UN52), a large, upland, open site located between La Grande and Baker, Oregon (McPherson et. al. 1981).

Little is known of the earliest occupation which predates the Mt. Mazama ash fall of 6700 B.P. A second, more intensive occupation indicates utilization of the area for hunting, gathering, and quarrying activities. It was during this occupation that the milling stone and a possible net weight are first observed. This occupation appears to coincide with the Altithermal. It is possible that during this period of warmer, drier conditions the site was exploited more intensively and may have been occupied for longer periods of time due to availability of resources. A similar occupational sequence may be seen at the Marshmeadow site (35UN95). Here, however, the majority of artifacts from deposits of the same age relate directly to the exploitation of large game animals (Mcpherson et al. 1981).

The morphology of the nondiagnostic tools and tool fragments at Stockoff were homogeneous over time. There is, however, a break in the well developed proj ectile point technology. About 4000 B. P. there appear three distinct groups of projectile point, which according to McPherson et al. (1981) "represent a hodge-podge of potentially diagnostic variants." Mcpherson also states that at about the same time, the Marshmeadow site was also reoccupied by people with an entirely new tool kit.

Site test excavations (Burney 1985) demonstrate occupation of the northern Blue Mountains of southeastern Washington from as early as 9000 - 7000 B.C. to 1700 A.D. or later. At the Warner Spring site (45GA124), occupation appears to be continuous over that entire time span. A more complete picture of the Umatilla National Forest prehistory will emerge as more information, based on excavation and data recovery, is collected and analyzed.

In the interim, when referring to the general chronology of the area it is convenient to refer to the six-part scheme proposed by Leonhardy and Rice (1970). This scheme proposes four periods further divided into six phases, including the Windust, the Cascade, the Tucannon, the Harder, the Piqunin, and the Numipu to describe the cultural sequences of the Lower Snake River Plain between 10000 B.C. and the early historic period (Figure 2)

McPherson et al. (1981: 65-66) provides a concise profile of the most refined chronology proposed for the Lower Snake River region of the southern Columbia Plateau, based on the cultural chronology proposed by Leonhardy and Rice (1970), and revised by Leonhardy (1975);

The Pioneer period (ca. 10000 - 5000 B.P.) includes the Windust and Cascade phases, which are viewed as a cultural continuum. By the outset of the Windust Phase (ca. 10000 8000 B.P.), lithic technology was well developed; the technique of producing blades struck from polyhedral cores was known. The artifact assemblage diagnostic of this phase includes lanceolate, stemmed projectile points ... cobble choppers, bola stones, edge-ground cobbles, milling stones, and manos (Rice 1972). The subsistence economy at this time was based primarily upon big game hunting; the collection of river mussels and plants was of secondary importance. Artifacts indicative of a fishing economy are not present within the assemblage.

The Cascade phase (ca. 8000 - 5000 B.P.) is characterized by lanceolate "Cascade" projectile points ... large bifacially flaked knives, and edge-ground cobbles. Artifacts present in the earlier Windust phase, excluding stemmed projectile point styles, are also included within the Cascade phase tool assemblage. Coinciding with the eruption of Mt.

Mazama around 6700 years ago, a new artifact, the Cold Springs side-notched projectile point, was added to the existing assemblage ... This type slowly became the predominant form. The subsistence economy at this time was based upon big game hunting, though plant and riverine resources, including fish for the first time, increase in importance.

The Initial Snake River period (ca. 5000 ~ 2500 B.P.), coinciding with the Tucannon marks the beginning of the second cultural tradition in the Lower Snake River region, a tradition which continued into the Ethnographic period ... Proj ectile points manufactured during the previous period were replaced by two types. The first has a short blade, shoulders of varying prominence, and a contracting stem; the second variety is corner~notched with an expanding stem. These types are more crudely manufactured than previous styles, which possibly indicates a shift away from big game hunt The presence of hopper~mortar bases, pestles, and net sinkers implies an increased reliance on seed and root crops, and fishing. Well formed knives are surprisingly absent from the assemblage. River mussel harvesting also increases in importance over previous periods. House pits, indicative of a more sedentary lifestyle, appear on the Plateau for the first time. Their use increased into the Ethnographic period.

During the Snake River period (2500 ~ 250 B.P.), coeval with the Harder phase, extensive house pit viI were constructed along major river courses. Early in this period proj ectile points similar to the large corner notched varieties, manufactured during the previous period, and basal-notched points are the predominant forms ... During the later part of the Snake River period the larger proj ectile point varieties are replaced by smaller, corner~notched, basal~notched, and stemmed forms which continue to be manufactured into the Ethnographic period.

Artifacts indicating a greater reliance upon plant and riverine resources are also included within the assemblage.

The Ethnographic period (250 80 B.P.), coeval with the Numipu phase, is marked by the arrival of the horse on the Plateau and ends with the adoption of the reservation system. Small side-notched, corner notched, and stemmed projectile points coupled with Euro-American trade goods characterize this time period.

A general picture of the aboriginal distribution of Native American groups has been constructed based on ethnohistorical evidence. The Umatilla Indians occupied the northwestern portion of the forest, ranging from Alderdale, Washington, to the Umatilla River. The Walla Walla lived mostly in southeastern Washington. The Cayuse Indians ted the resources of the northern portion, between the Columbia River and the Blue Mountains in Oregon and Washington. Several other tribes inhabited areas immediately adjacent to the present day forest boundaries. The Nez Perce occupied the Imnaha, Grande Ronde, and Snake River drainages to the north and east of the Cayuse. To the south, the Northern Paiute inhabited the John Day, Malheur, and Snake River Drainages.

The historical record of the area demonstrates the potential diversity of kinds of cultural remains which might be encountered on the forest. The 1804 Lewis and Clark Expedition is generally considered the beginning of the historic period. From that date, until the discovery of gold in the 1860s, the forest was subject to brief visits by explorers, traders, trappers, and missionaries.

The discovery of gold led to a large sedentary population which in turn created a demand for convenient goods and services. As a result, merchants, farmers, and stockmen were drawn to the mining areas along the southern boundary of the forest. Gold mining reached its peak in the late 1890s. Several important lode mines were located in the Granite mining district. In 1908 the Fremont Powerhouse was constructed to provide electrical power to the Red Boy Mining Company and others. The Powerhouse is an early example of hydroelectric power generation in the Pacific Northwest and is listed on the National Register of Historic Places.

In the 1840s the Oregon Territory was largely Public Domain. Settlement was controlled and encouraged by federal laws such as the Preemption Act of 1841 and the Homestead Act of 1862> Settlers also wanted clear title to traditional Indian lands. Consequently, by 1855, Indian lands were ceded to the U. S > Government and the Umatilla, Cayuse and Walla Walla Tribes were removed to the Confederated Tribes of the Umatilla Indian Reservation> Later, the Dawes Act of 1887 divided Indian reservation lands into individual holdings for American Indians and declared remaining lands surplus and Public Domain.

The Forest Reserves of 1891 set aside resource areas from Public Domain. In 1906 the Forest Lands Homestead Act provided for the establishment of agricultural lands within the Forest Reserve. In addition to the mining and homesteading activities, the use of Forest rangeland became increasingly important to the cattle and sheep industry between 1860 and 1900. In 1904 the Government permit system was established for public domain land to alleviate conflicts between sheepmen and cattlemen.

Between 1890 and 1930 timber harvest was facilitated by the construction of logging railroads. Timber resources continue to be a primary economic activity in the area. The management of timber and other resources provided the basis for the development of government services to the area. Under the supervision of the U.S. Forest Service, the Civilian Conservation Corps constructed or improved recreational sites and administrative facilities on National Forest lands (Throop 1979)

Previous Inventory

There have been approximately 225 cultural resource inventory, test excavation, and mitigation projects conducted on the Forest in the past 10 years. These projects have accounted for 289 recorded sites, 176 historic and 113 prehistoric. While the vast majority of the prehistoric sites recorded are lithic scatters, other site types include cairns and rockshelters. The preponderance of historic sites are related to mining activities while the remainder are associated with the homestead era and railroad logging with a few associated with military activities.

Table 1 presents the cultural resources inventory effort on the forest to date.

As outlined in the table, approximately one third of the forest has been sample surveyed resulting in a site density of 0.40 sites per section. This may not be an actual representation of site density on the forest as a whole as site density increases by over 400 percent, to 1.76 sites per section, if one looks at the site density of actual surveyed area.

Of the total acres surveyed, 85,284 are in Oregon and 20,036 are in Washington. Similarly, the total represents 357,026 project acres in Oregon and 100,464 project acres in Washington.

Table 1

Inventory Accomplishments

Total Acres Acres Percent Sites Site District Acres Cleared Surveyed Surveyed Recorded Density

(1) (2) (3) (4)

Walla Walla 358531 100072 27690 8 28 37 .25 .86 Pomeroy 365901 96969 19418 5 27 62 .41 2.07 North Fork John Day 465830 181095 37890 8 39 141 .50 2.39 Heppner 212213 79354 20322 10 37 49 .40 1. 53

Forest Total 1402475 457490 105320 8 33 289 .40 1. 76

(1) Acres surveyed

(2) Acres cleared

(3) Sites per section cleared

(4) Sites per section surveyed

Site density on each district is quite variable with the density on the North Fork John Day District being the greatest while Walla Walla District has the smallest. The density on the North Fork District, however, is not representative of the district as a whole as 62 of the sites recorded are concentrated along the North Fork John Day River. If these sites are excluded, the site density would be 0.74 sites per section which is smaller than that on the Walla Walla District.

In addition to the recorded sites, the Umatilla overview (Hudson et. al. 1978) identified over 700 suspected sites on the Forest. The vast majority of these sites have not yet been field verified.

Cultural Resource Inventory Strategy

The Umatilla National Forest began conducting cultural resources investigations in 1977. Since the inception of the cultural resources program the emphasis has been upon compliance with federal regulations concerning undertakings (esp.

36 CFR 60, 36 CFR 296, 36 CFR 800, The National Historic Preservation Act of 1966, and Executive Order 11593). These initial investigations were extremely limited and amounted to little more than record searches. When field reconnaissance became the norm for cultural resource compliance, the surveys consisted primarily of examinations of the locations of selected Forest Service ground disturbing projects in areas considered to have a high potential of containing cultural resources.

During this period of time, few, if any, cultural resources have been located in those areas which were a priori determined to have a low probability of containing sites. There is, however, no way to quantify this finding, as the survey conducted in these areas was not conducted in a sturdy statistical manner.

The utility of such a reactive approach for planning purposes has proven to be limited. The data obtained are not useful in the generation of predictive models, and the net result is that each undertaking requires survey without any means of determining whether such survey will produce redundant or new data.

Louthan (1983) has suggested that predictive modeling is both cost effective and has utility in planning. Whether this is true, or whether such models are of limited value, it is clear that they depend upon a baseline set of data that includes archeological, environmental and geographic information.

Synthesis and interpretation of the data obtained from project-based reactive surveys is also not easily analyzed when a baseline set of data is lacking.

Especially problematical is the relationship of the archeological remains to environmental variables, such as vegetation community, slope, exposure, distance to water, and other critical aspects of the ecosystem. Also of interest is the relationship of the various types of sites to each other (the inter-site patterning), and the relationship of these regional patterns to the various environmental variables.

The Forest Service, and more specifically the Umatilla National Forest has committed to implement a GIS. The basic data needed to create an adequate GIS for the Forest will be available by 1992. Upon implementation of the GIS it will be possible to stratify the Forest based on adequate environmental variables and thereby facilitate cultural resources locational predictive modeling. In order to gain the needed baseline data to create a cultural resources locational predictive model, part of the Umatilla National Forest inventory strategy will include a random sample survey. This survey constitutes a long range effort intended to generate an overall sample of the cultural resources present on the Forest. It will be conducted both in conjunction with project judgmental surveys (see below) and as special projects as funding allows.

Random Sample Research Design

There is a considerable body of literature both within and outside of archeology concerning archeological sampling design (Raab 1977; Donaldson 1977;

Mueller 1974; Kohler and Parker 1986: 403 417; Matson and Lipe 1975). Early papers addressing the utility of sampling include Ruppe (1966), Plog (1968) and Judge (1973). Binford (1964), in what was undoubtedly the seminal contribution to the topic, argued persuasively for probability based sampling within a regional research des

The advantages of sampling primarily stem from its utility for planning.

properly designed and executed sampling can provide cost effective and accurate data needed for planning. It is, precisely how to accomplish this end, to efficiently obtain a representative sample of the archeological population, that various approaches have differed. Probably the most widely used approach has been to stratify the area based on environmental and geographic variables, and then select samples from each of the classes generated (see Judge et. al.

1979, Wood 1978) .

Examination of the data currently available for the Umatilla National Forest led to the decision not to follow this approach. The reasoning behind this was that the knowledge of the environmental variables and the relationships between these variables and the cultural resources needed to adequately stratify the sample was lacking. The boundaries between environmental communities had not been plotted with the degree of accuracy necessary to insure that such environmentally stratified data would be accurate. Furthermore, it is possible that the environmental communities have shifted, and the relationships between archeological sites and environmental communities would thus be unclear.

Another consideration in not stratifying the area had to do with the time frame involved. The complete surveying of a sample was predicted to take several years to accomplish, and the level of environmental data available for the Forest, and the mapping of the environmental variables, will undoubtedly improve during this period. Environmental stratification of the Forest at this point is premature, and would result in the work becoming obsolete before the sample is completed.

In lieu of stratifying the area, a two level hierarchical random sample was designed. The decision was made that the section (one square mile) would be the basic sample unit. Each section containing Forest land was assigned a number from 0 to 9 drawn from a random numbers table. Those sections assigned digits 1, 3, 5, and 7 were then chosen arbitrarily for sampling. A random selection of these determined that those units assigned the digit "3" would constitute the first sample.

In line with Binford (1964), sample units one quarter mile on a side (quarter sections) were then chosen within each section assigned the digit "3". These were chosen by assigning numbers to each quarter (1 through 4) and using a random numbers table to select one quarter from each section. This resulted in an overall 2 1/2 percent sample of the Forest that was relatively evenly distributed, but which retained a random character.

Considerable attention has been devoted in the literature to whether the appropriate type of sample unit is the quadrat or the transect. It has been demonstrated that the transect produces the better sample of the elements of the sampling universe, but the quadrat allows more inferences about the inter-relationships (Mueller 1974, Judge et al. 1979, Plog and Hill 1971, Matson and Lipe 1975:131). As one of the primary thrusts of the sample survey is to assess the inter-relationships of the various sites, and the inter-relationships of these with the environment, the quadrat was selected as the most appropriate unit.

While some authorities have examined a sample as large as 10 percent (Wood 1978), and others have suggested that in some cases as much as 70 percent of an area may need to be examined (Green 1976:6), one widely held view has been that a 1 percent survey is adequate for predictive modeling of cultural resources in an area (see Plog and Hill 1971, Rice et.al. 1980, Plog 1978). The issue here is that statistical it is the sample size which is critical, with the sample fraction of only secondary importance. For proper sample design, the characteristics of the population to be sampled should be considered.

Often, a large sample is more representative of a population than a smaller one. But, the needed size for a representative sample depends on the homogeneity or heterogenity of the popUlation. Furthermore, the size of the sample needed also depends upon the desired precision and confidence level.

Those factors must be carefully balanced with the amount of analysis to be performed on the collected data (the analysis of large data sets can be quite costly), and the administrative costs of the data collection (see Eckhardt and Ermann 1977:161) .

With these considerations in mind, it was felt that the sample needed to be as large as selected (224 sample units out of a universe of 8765 units) in order to achieve the desired precision and confidence level. The method followed allows for a 95 percent confidence interval and between 5 and 10 percent precision even with maximum standard deviation (Yamane 1967:398).

Two qualifications must be made concerning the process followed. If a section contained a quarter or quarters which lacked forest land, and which were privately owned, then these quarters were excluded from the sample and a different quarter was selected. However if a quarter contained any forest land, no matter how little, then it was left in as a possible sample unit.

Also, as the earth is spherical and maps are flat, sections are not always 640 acre squares. Some of the sections in the Forest are distorted. Despite this, those sections were included in the universe, and if selected they were quartered. This resulted in some quadrats more or less than the projected 160 acres.

A staged approach similar to that advocated by Redman (1973) will be followed.

Following the completion of the first 2 1/2 percent sample survey, an assessment will be made. It should be possible to analyze site location information in relation to environmental variables and make a determination as to how to proceed. Options include continuing with an additional 2 1/2 percent sample, stratifying a sample, generating a predictive model from the 2 1/2 percent sample already surveyed, or some combination of the above.

In the event it is determined that enough information has been gathered, a first draft model will be created using those apparent predictor variables.

This model will be tested against known site location from the survey already performed. Depending on the outcome of this exercise, several entire local drainage basins will be surveyed in order to test the derived locational model.

These basins will be selected in both the north and the south end of the Forest. The information from these two tests will be correlated with the information from the random sample and a second order model or models will be derived. Fine tuning of the model will be continual.

A grab sample of diagnostic artifacts and chronometric samples, if available, will be made during the survey. The locations of these specimens will be mapped, and an attempt will be made in the laboratory to assess the relative and absolute dates of the assemblages.

Judgmental Sample Design

The Umatilla National Forest must continue to carry out its responsibilities to inventory in advance of any project which may affect cultural resources. In the absence of a statistically valid sample on which predictive site location modeling may be accomplished, the Forest will employ a probability judgmental sample. This portion of the Umatilla National Forest Cultural Resource Inventory plan is similar to other forest inventory plans in the Pacific Northwest Region (Davis 1983; Reagan 1982). Its purpose is to provide a systematic method of conducting cultural resource surveys which meet the following criteria: 1) Section 106 Compliance, 2) maximum conservation of the cultural resource base and 3) maximum cost-efficiency. As Davis (1983) concludes, the efficiency of every cultural resource survey on the Forest could be improved to cover more acres and, at the same time, be better focused to locate more sites.

A complete inventory is required to meet compliance with cultural resource direction, policy and laws. A complete inventory identifies and accurately describes 100 percent of the locatable cultural resources. A sample inventory investigates only a portion of a project area in order to ascertain the total resources which may be present. From a statistical vievvpoint, a sample survey can be a complete survey when the sample selected includes 100 percent of the cuI tural resources. Selection of the sample thus determines if the "sample inventory" may be termed a "complete inventory" in the legal sense.

There are numerous approaches to sampl ing. The traditional archaeological approach uses known evidence to determine which unknown areas should be investigated. It is this approach, the judgmental sample based on anthropological knowledge, which will be utilized. Most sample surveys assume that the location of cultural properties are correlated with environmental characteristics rather than being randomly scattered over the landscape. This assumption is probably true of prehistoric sites; they appear to be closely tied to environmental variables of slope, geomorphic features, and distance to water.

This generalization often will not apply to the more complex cultural systems represented by historic sites; these site locations are often dependent on cultural factors which may outweigh environmental concerns. According to Sabo et al. (1982), there appear to be definite relationships between site function and specific sets of biophysical associations during the prehistoric. This functional differentiation appears to have exerted minimal influence during the historic period. They conclude that various cultural factors such as economics, social organization, and transportation are the major determinants for the location of historic sites. Locating such sites during the inventory process must then depend on a thorough pre-field search of existing historic records and interviews with local informants.

The Umatilla National Forest will employ a judgmental probability model in order to comply with applicable laws and regulations. Within the framework of the judgmental probability model, areas are not sampled; surveys are conducted in those areas predicted to contain 100 percent of the cultural resource sites. Therefore, 100 percent of the locatable sites should be identified.

The following is a list of the areas which will receive 100 percent survey:

1. The location of previously reported sites and ticklers: such as home steads, mining claims I trails railroad rights of -way I areas sacredI to the practice of Native American religion.

2 . Areas in proximi ty to water; (i. e ., springs and 1akes I and terraces adjacent to permanent streams) .

3. Major ridge systems, including saddles and feeder ridges which may have been used for historic or storic transportation systems.

4. Geologie formations associated with historic mining activities, those which characteristically yield lithic raw material for tool manufacture by prehistoric people and rock overhangs, caves and, boulder and rock faces.

5. Environmental ecotones associated with known subsistence resources, i.e .. camas fields, large animal migration routes, etc.

6. Prominences which offer a panoramic view.

7. Open slopes of less than 15%.

If the survey design for a particular planning unit indicates that less than 20 percent of the area is included in the scheduled survey, additional survey will be conducted in the low probability areas so that at least 20 percent of the area will be surveyed.

Although the above survey areas should contain 100 percent of the locatable cuI tural resources, areas not considered to have a potential for containing cult~ral resources will also be investigated. At least five percent of the low probability areas will be surveyed and documented by Forest Service staff qualified as Cultural Resource Technicians in conjunction with normal duties.

Documentation of the areas surveyed will consist of an appendix to the survey report in which the areas surveyed will be detailed and a statement of findings presented.

There is no reason ,to expect that cultural properties will be located during this inspection of these low probability areas. However, should properties be located, then additional areas similar to those in which the properties were located will be inspected. In addition, the inventory plan will be modified to include those areas as high probability areas. In this way, the inventory plan will remain dynamic, with additional areas added to those acres which are considered to have a high probability to contain cultural properties.

The Forest is in the process of developing agreements with Native American groups who have or are still using the Forest for cultural or religious purposes. These agreements will contain provisions for the identification of areas important to these groups so that expressed concerns for those areas can be addressed. The first of the agreements is to be accomplished with the Confederated Tribes of the Umatilla Indian Reservation. When these agreements are finalized they will be appended to this plan and become a part of the inventory plan.

Basic Inventory and Field Procedures

Cultural resource inventory is both a process and a record resulting from that process. Inventory is the first step in the compliance process, and is required for all activities undertaken directly by, permitted by or licensed by the Forest Service which have the potential to affect significant cultural properties. The general type of inventory design will vary somewhat wi t:h the scope of the proposed project.

The overview is a broad-scale inventory of a large geographic area, based on previously known or recorded information. This document meets inventory requirements for the forest plan and other planning documents. The overview does not meet project specific inventory requirements because it does not locate all cultural resources within the study area. A cultural resource overview of the Malheur, Umatilla, and Wallowa-Whitman National Forests was compiled by Hudson et al. (1978). There is a need for periodic updating of this material both to facilitate the pre-field records check and to comply with a Memorandum of Agreement with the State Historic Preservation Office (USDA Forest Service 1984). In the 10 years that have passed since the overview for the Umatilla was written a fair amount of survey has been conducted on the Forest. Much data has been gathered during these inventories. That information will be incorporated into an updated overview to be completed during FY '90 and then updated on a biannual basis after that.

A cultural resource inventory is necessary as a part of an ongoing forest-wide program of cultural resource management when the project has the potential to affect cultural resources. Examples of activities that may affect cultural resources are disturbance of the ground surface or disturbance of existing structures: transfer of land from federal jurisdiction or ownership: and indirect impacts from increased public access or use.

All of the activities projected to be conducted on the Forest could have a potential to affect cultural resources. Some activities such as hand constructed fences, reconstruction and maintenance of existing improvements, activities within stream channels, resource management which does not involve surface disturbance, installation of sign posts and survey monuments, and personal use fuelwood sales mayor may not affect cultural resources depending on the design of the project.

Activities associated with range management can be divided into two classes:

structural improvements and nonstructural improvements. Structural improvements include, but are not limited to, construction of fences and corrals and numerous water development projects. The installation of pipelines, both on the surface and subsurface, drinkers, trick tanks, and troughs are examples of small water development projects. Larger water development projects include the construction of stock tanks, and installation of storage tanks.

Nonstructural activities associated with range projects can again be divided into two classes: mechanical and nonmechanical. Projects which may be accomplished by mechanical means include pushing of trees and various methods of rangeland planting. Nonmechanical methods of range improvement include such projects as prescribed burns.

Wildlife habitat improvement can be accomplished by a variety of methods.

Nonstructural improvements which will not necessarily be accomplished with mechanical methods include burning, riparian planting, and planting of browse plants and grasses. Fence construction for enclosures, spring head protection, and access I escape mayor may not involve the use of mechanical equipment.

Possible improvements that can be accomplished by mechanical methods are piling of slash and downed timber from clear cutting, road closures, and plowing and drilling for seed introduction. Also included are a variety of water development projects. These projects include the installation of trick tanks, creation of potholes in wetlands, and the rejuvenation of marsh lands through the use of spreader ditches, spring development and construction of earthen dams.

Wilderness value and recreation management will include both mechanical and nonmechanical improvements. These improvements include trailhead construction and reconstruction; trail construction, reconstruction and maintenance;

wilderness clean-up; and construction, reconstruction and maintenance of developed recreation sites. Additional recreation oriented projects include construction of parking facilities and winter sports areas for a variety of activities. Road maintenance reconstruction, and construction will occur toj facilitate recreation, both developed and dispersed.

Soil and watershed management will primarily involve stabilization of water courses and vegetation planting. Water course stabilization will be accomplished by either mechanical or nonmechanical methods and is usually accomplished through the construction of dams with soil, rock, and vegetation.

Planting to abate erosion can be done with or with out mechanical methods.

Timber management will include harvesting and regeneration of tree stands.

Activities associated with timber management include road construction and reconstruction tree felling and skidding, brush piling, fuelwood harvesting,j scarification and planting, and prescribed burns.

The cultural resource inventory process may be divided into three parts: a pre field phase, a field phase, and a post-field phase. In the pre field phase the boundaries of any given survey project are determined. A survey design is developed that includes the proposed impact area, as well as lands adjacent to it: Areas contiguous to specific impact areas are included to eliminate isolated unsurveyed parcels. In some instances, the project survey boundaries may be enlarged with reference to specific geomorphic features or to include a group of small land parcels.

Designing the inventory on the basis of the inventory plan will be done by the Forest Archaeologist. Once the project area is defined, a systematic records check is conducted. This includes, but is not limited to, the current list of the National Register of Historic Places, the Forest Overview, the plat maps of the General Land Office survey, and other historical maps and notes. This procedure is in accordance with the Memorandums of Agreement between the Forest Service and the respective State Historic Preservation Offices (USDA Forest Service 1979, USDA Forest Service, 1988). A literature search provides the background for placing into context cultural sites located during the field inventory. Pertinent data on plant communities, soils, topography, and geology are also reviewed prior to and during field work.

Large projects, those that involve more than 100 acres, will be inventoried according to a sampling design based on the inventory plan. A standard inventory report will be written to document the finding of the survey. Small scale projects that involve less than 100 acres will be 100 percent surveyed and will also be documented with a standard inventory report. The rationale behind the difference is explained as follows large scale projects are generally more complex in the number of proposed activities which may cause disturbance and usually include a variety of strata, while small scale projects usually involve only a single type of activity and a single type of strata.

Since a considerably smaller area is involved in small scale projects, the sampling strata are likely to be limited or invariable.

The field phase of the inventory consists of systematic pedestrian survey by trained personnel. All areas delineated by the inventory design are subject to an intensive field search. Survey transects will be no more than 20 meters apart. In areas of dense ground cover (e.g., vegetation, forest duff) surface inspection will require the periodic removal of ground cover to mineral soil.

Areas such as dense stands of lodgepole may not be inspected due to safety concerns. Meandering transects which maximize coverage of all visible soil surfaces are preferred to straight compass-oriented lines.

A site is defined as a location of purposeful prehistoric or historic human activity. An activity is considered to have been purposeful if it resulted in a deposit of cultural materials beyond the level of one or a few accidentally lost artifacts. A cultural resource qualifying as a site under this definition should exhibit at least one of the following:

a) One or more features.

b) An occurrence of cultural material (such as flakes, ground stone, or historic debris) in a density of at least 10 items.

The above criteria may be modified, where appropriate, by a cultural resource specialist. Locations of cultural material not classifiable as sites by this definition will be considered isolated occurrences.

Si tes will be marked on the ground by attaching an aluminum tag to a tree closest to the center of the site. This tag will have the Forest site number inscribed on it. In addition, the tree will have two, two-inch orange bands painted at about five feet above the ground and a spot painted at the base.

The site boundary will be flagged with orange flagging at such time that the site is in danger of being disturbed.

Documentation

Documentation of the cultural resource inventory results is an essential step in the compliance process. Survey work and report are of equal importance since both are required for compliance. The principal surveyor will complete a written documentation which meets the minimum requirements established by the Pacific Northwest Region and the SHPO(s) .

Each report shall include:

1. Introduction

a. Narrative description of the undertaking, including nature and extent of all land modifying activities

b. Narrative description of the project area, including the legal description(s) , description of the topography, drainage, vegetation, elevation and soils.

c. Description of previous surveys and recorded sites within or immediately adjacent to the project area.

d. Culture history of the Forest.

2. Description of the survey

unsurveyed areas if any tie. thick lodgepole stand).

3. Description of the findings

a. Definition of a site for the purposes of the report (if sites or isolated occurrences were located).

4. Evaluation

a. Evaluation of National Register eligibility, or potential eligibili ty.

5. Protection requirements

a. Discussion of operating conditions and other measures to be taken to protect sites during the undertaking.

b. Discussion of possible mitigation measures, if sites will be affected.

6. Maps

a. A portion of a USGS topographic map at a scale of 7.5 min.

showing the project boundaries.

location and number.

d. Maps will contain the following standard information:

Township and Range North Arrow Scale Project Name Contour Interval The statement "For Official Use Only" The statement "Disclosure of site location information is prohibited (36 CFR 296.18)"

e. All maps will be legible and reproducible on black and white copy machines.

7. References, illustrations, and exhibits.

8. site form(s) (R6 2300 18) and isolated occurance form(s) (each site form will be accompanied by a sketch map of the site, photographs, and, where appropriate, artifact drawings) .

Monitoring

All of the Historic Preservation laws and regulations have been enacted to protect and preserve historic properties. To this end, locating the properties is a necessary first step. Monitoring is the next step in order to insure that those identified properties are protected and preserved. All identified historic properties will be revisited at least once during, or at the conclusion of activities near the site. This monitoring will be reported to the respective SHPO in the form of a letter on or about the first of January of the year following the monitoring.

Inventory Plan Evaluation and Modification

Monitoring and review of the…