OLYMPIC The Evolution and Diversification of Native Land Use Systems on the Olympic Peninsula A Research Design

Definition of Management Zones

The purpose of this chapter is twofold: to divide the region into management/research zones and present archaeological predictions for these zones. The first step will be to list the proposed zones accompanied by their major distinguishing features. In dividing up the Peninsula, it is necessary to make certain assumptions about the relative importance of various environmental features to hunter-gatherers in the past. Both paleoenvironments and human adaptations have changed in substantial ways since the time of initial human occupation of the region. These points suggest that the environmental features which we use in dividing the Peninsula into management zones should be ones that are likely to be of sufficient importance and stability to have been of significance throughout most if not all of the prehistory of this region.

Major physiographic features are strongly correlated with variations in the major classes of fish and marine food resources. These features figure importantly in segmenting the coastline of the Peninsula into useful management zones. Resource classes considered in relationship to physiographic features included salmon, sea mammals, shellfish, and sea birds. Moving inland to the lower river valleys and lowlands of the Peninsula, differences in salmon resources were also a major factor in partitioning of the river valleys. This reflects the importance that has been attached to the distribution of this resource class in determining the presence or absence of a riverine-based settlement strategy. Moving still further inland and up slope, the distribution of terrestrial resources, especially deer and elk, was emphasized in the demarcation of zonal boundaries in the Peninsula's interior. In general, then, the zonal divisions reflect the geography of human land use as this has been depicted in preceding chapters of this report.

It must also be emphasized that these zones should be considered provisional in the light of current information. They will serve no useful purpose if they are considered immutable; as new information becomes available such change might involve further subdivision of the zones presented here to accomodate different or additional environmental variables or it might require actual relocation of zonal boundaries. Such changes may be necessitated by new knowledge about the region's archaeological record or about its past environments or both. During the course of the present study, the zones were redefined more than once; the present set of zones represents a more fine-grained form of an earlier version (see Schalk 1985c).

Although the research design developed in this study is regional in scale and could be used to generate expectations about the character of archaeological variability throughout the Olympic. Peninsula, predictions are focused upon those zones that are included within the Park. A comprehensive development of predictions for the other zones of the Peninsula is beyond the scope of the present effort but may be accomplished in the future in conjunction with other studies.

The Park is divided into four major management zones with subdivisions in two of the four major zones. These zones are as follows:

These zones are illustrated in Figure 8.1 (a large-scale version of this map is available at the Pacific Northwest Region, NPS). In the discussion below, the zonal boundaries are defined and the rationale for their placement is explained. Against this background, attention is also directed to brief summaries of currently available archaeological information for each zone encompassed within the Park. The final section under the discussion of each zone addresses the issue of future directions for archaeological research within that zone. Included here are predictions for the character of the archaeological record represented in each zone, important research domains and their associated data requirements. As in Chapter 2 and the remainder of this chapter as well, the terms residential base, location, field camp, and station follow Binford (1980:10).

Figure 8.1 Proposed Archaeological Management Zones for the Olympic Peninsula Region. (click on image for a PDF version)

I. The Coastal Margin

The coastal margin is operationally defined as 5 km wide band of land paralleling the shoreline around the Peninsula. The definition of the landward margin of this zone seems to be somewhat more arbitrary in an environmental sense than the placement of boundaries on the other zones. The use of 5 km as the inland boundary to this zone amounts to a compromise to at least two considerations. Firstly, in the densely forested landscape characteristic of much of this region, a distance of 5 km probably encompasses most of the places that would be visited on foot and returned from in a single day. Secondly, acknowledging that the sea/land interface has changed throughout the Holocene to varying degrees around the Peninsula (see Chapter 5), a distance of 5 km encompasses most areas of land in the region that are below an elevation of 30 m. Sea levels were apparently not more than 30 m above the modern sea level since the late Pleistocene. The strip of land included between high tides and 5 km inland, therefore, includes most areas that were situated directly on or adjacent to saltwater since the Late Pleistocene. Besides the narrow strip of land along the shoreline of the Peninsula, this zone also includes offshore islets and rocks. The Coastal Margin so defined includes the entire coastal strip of Olympic National Park.

The coastline of the Olympic Peninsula has about as much variability as can be encountered in a single region and this is reflected in the marked variations in the kinds of marine resources that occur in different areas of the Peninsula. The major subdivisions of the Coastal Zone were identified to accomodate this variability and these subdivisions are explained in the discussions below.

The coastal strip of Olympic National Park is by far the best known zone within the Park from an archaeological perspective. Archaeological surveys were carried out here in the 1950s (Stallard and Denman 1955) as were minor excavations at Toleak Point (Newman 1959) and White Rock Village (Guinn 1963). A more recent survey examined the coastal strip of the historic territory of the Makah as well as the river banks of the included streams (E. Friedman 1974; 1976). Relative to much of the Olympic Peninsula's interior, the obstrusiveness of typical sites along the coastal zone (especially shellmiddens), and the geologically dynamic character of this zone, facilitate site discovery.

The most numerically dominant sites recorded in the Coastal Margin zone are those referred to as "shellmiddens". Despite the obtrusiveness of shellfish remains, exploitation of these resources probably represented a relatively minor subsistence activity for the occupants of most shellmiddens. Existing historical and archaeological data make clear that sites as functionally distinct as spring fishing and/or sea mammal hunting camps, permanent winter village settlements, and bluff-top redoubts are lumped together under the generalized term "shellmidden". In fact, petroglyphs represent the only other class of aboriginal sites recorded within this zone and the only site class that typically lacks shellfish. Aside from preliminary site distributional information, most of the substantive knowledge about this zone comes from one large scale excavation project—the Ozette Project. This project is discussed further below under the discussion of Zone Ic.

Zone Ia: Grays Harbor to Cedar Creek

This zone extends from the northern end of Grays Harbor to Cedar Creek, about 5 km south of the Hoh River mouth. The distinguishing features of this zone include productive salmon rivers, a relatively straight coastline with few headlands and rocky islets, rather extensive areas of low gradient sandy beaches south of Point Grenville, and a wide continental shelf (35-50 km). In terms of food resources, this area includes 1) the mouths of two of the Peninsulas most productive salmon streams: the Quinault and the Queets; 2) sandy substrate clams in greater abundance than other outer coastal zones {especially razor clams from Point Grenville south}; 3) a scarcity of sea mammal haul outs and sea-bird rookeries; and 4) poor access to halibut banks and sea mammal migration routes along the continental shelf.

Archaeological Expectations

Only about a 17 km long strip of Olympic National Park falls within this zone. Expectations for this zone would include:

1) Sea level changes in this zone have been of low amplitude during the last 9,000 years with the lowest relative sea levels only 1-2 m below modern levels (ca. 9,250 BP) and highest levels (ca. 4500 BP) less than 10 meters above present levels (see Chapter 5). It follows that the Holocene archaeological record of this zone (and other outer coastal zones—IB and IC) should be exceptionally well preserved compared to other areas of the Northwest. If, as some have argued (Fladmark 1975; 1979), early Holocene adaptations were already heavily oriented toward marine resources, then sites containing evidence of maritime orientation should be found within this zone. On the other hand, it has been argued in this study that systematic use of marine resources in this region developed in the late Holocene. According to this model, most archaeological sites in this zone will be related to procurement of marine resources but they are expected to be less than about 3,000 years old. This model maintains that this zone is not expected to have been important in early and mid-Holocene land use systems. Contrasts between these models and their implications for the Coastal Zone of the Park were developed during the early stages of the present project (see Schalk 1985c). Distinguishing between these models rests heavily on the geological question of whether or not the known archaeological resources of the Coastal Zone are a highly biased sample of those that previously existed there.

2) Sites that are less than 3000 years old in this zone are expected to be associated with a riverine collector land use system. Since the types of systems were present historically in this area, this would imply that there have been no major (systemic) cultural changes during an interval of about three millenia. Intensification of resource exploitation may, however, be reflected in evidence for shorter resting periods between harvests (e.g. younger age structure in shellfish) or addition of previously unexploited low-return resources.

3) In view of the limited marine resources of this zone and the large rivers that empty into the ocean within it, it is expected that there would have been a riverine distribution of winter villages. Sites located on saltwater within this zone, except at large river mouths would have been used seasonally (particularly in spring-summer) by logistical task groups seeking marine resources. This would mean that archaeological sites along the coastal margin of this zone are likely to be either field camps or locations. Both of these site types are expected to lack evidence for permanent housing. Faunal assemblages should be dominated by marine resources, especially shellfish and nearshore marine fish. Artifact assemblages should be dominated by implements used in the harvest and processing of these same resources. The overall character of the faunal and lithic assemblages relative to winter village sites should be one of low diversity. Many of the faunal resources and artifact types found in residential bases should be lacking, reflecting the specialized nature of site occupation.

4) Any winter villages situated in this zone should be located at or very near the mouths of rivers, especially the larger rivers. This reflects the view that fully maritime adaptations were not present in this region of the Olympic Peninsula in prehistoric times and that riverine collectors positioned their permanent winter residences near the places where salmon were obtained in bulk for storage.

5) Site locational patterning should be strongly correlated with the distributional structure of marine resources being taken. The "location" types of sites should have a greater tendency to have faunal remains of resources that occurred in the immediate vicinity of the site. Field camps, on the other hand, are expected to contain resources from a wider range of habitat types. This would imply higher resource species diversity in discrete occupational events. Because village sites are not expected on saltwater within this zone (except at river mouths), most sites less than about 3,000 years old should be either "locations" or "field camps".

6) Reoccupation of the same locations on a cyclical basis should lead to the vertical accretion of both locations and field camps. This is the result of the stability and spatial predictability that is characteristic of many marine resources. In those cases where shellfish were being exploited, these deposits should contain well preserved vertebrate faunal remains if such resources were also being exploited. This expectation is based upon the acid neutralizing nature of mollusc shells. To the extent that most intertidal hardshell clams occur in protected coastal waters, the molluscan resources of this zone are likely to be dominated by the razor clam.

The dominance of this particular species may have interesting implications for both the locational stability of shellmidden deposits and the extent to which vertebrate faunal remains are preserved in shellmiddens. The habitat of the razor clam is best developed along straight stretches of low-relief coastline that lack inlets or embayments. These are high energy beaches and not suitable habitat for hardshell clams. Another characteristic of the razor clam is that it is unusual for the thinness of its shell and, therefore, preservation of its valves is likely to be rather poor compared to most other clams.

There are two points to be made from these observations. The shoreline distribution of these clams tends to be linear rather than patchy like that of the hardshell clam species. These different resource distributional structures imply different patterns of site formation (see Thomas 1983:69-71). The linear nature of razor clam beds along this zone is, therefore, expected to reduce the potential for vertical accretion of midden deposits at discrete locations. Instead, a more linear and horizontally diffuse process of archaeological site formation is expected. Given this process and the perishable quality of razor clam shells, archaeological deposits should be harder to detect and, when detected, are likely to have relatively poor preservation of vertebrate fauna relative to shellmiddens with high proportions of hardshell clam species.

Zone Ib: Cedar Creek to Cape Johnson

Distinguishing features of this zone also include the mouths of two productive salmon rivers, but in contrast to Zone Ia, a complex coastline with numerous headlands and rocky islets and a narrower continental shelf. The southern boundary of this zone was drawn to include Destruction Island. This particular island is unusual for its size, distance offshore, elevation, and archaeological potential. The width of the continental shelf ranges from about 20 km at the southern end of this zone to about 10 km at the northern end. This stretch of coastline has considerably higher potential than Zone Ia for hardshell clams, marine mammals, birds and offshore fishing.

Archaeological Expectations

The archaeology of this zone is expected to share many characteristics with that of Zone Ia. Greater marine resource diversity of marine resources, however, makes it somewhat transitional in character to Zones Ia and Ic.

1) This zone should have intact sites containing evidence for systematic marine resource use spanning most of the Holocene if Fladmark's "early maritime hypothesis" is correct. Another expectation associated with this hypothesis is that sea mammals will be greatly emphasized prior to 5,000 B.P. but that after this time salmon will be of considerable importance. This reflects a postulated surge in salmon production after stabilization of sea levels and stream gradients after about 5,000 B.P. According to the early maritime hypothesis, intensive shellfish use is expected to develop in conjunction with the more sedentary adaptations afforded by salmon exploitation at around 5,000 yrs B.P.

2) According to the arguments developed in Chapter 6, however, usage of this zone prior to 3,000 B.P. is expected to involve no more than casual exploitation of marine resources throughout most of the year. The coastal strip is not a location where people practicing a foraging type land use strategy based upon the hunting of ungulates are expected to be during most of the winter. Hunting efficiency during winter for the Old Cordilleran systems would favor locations where ungulates were more concentrated or where the residential bases were centered within foraging zones that had land on all sides. (Coastal settlement implies a reduction in hunting efficiency for terrestrial resources due to the reduced land area within the foraging radius of a settlement situated "with its back to the water"). This would mean that archaeological remains associated with this zone should be more recent than the estimated emergence of riverine collector systems at around 3,000 yrs B.P. The important exception to these arguments, however, has to do with the protein dynamics problem discussed in Chapter 6. The exploitation of marine resources that are oil-rich or which contain carbohydrates during seasons of low terrestrial productivity and especially late winter/early spring is likely to be the earliest form of marine resource exploitation. This means that archaeological evidence for use of marine resources before 3,000 yrs B.P. should be limited to that season of the year when fats and carbohydrates are in particularly short supply (i.e. late winter/early spring).

3) According to the models set forth in this study, archaeological sites in this zone should be dominated largely by field camps and locations that are less than 3,000 years old; due to the greater abundance of marine resources, however, there is also some potential for coastal residential bases after about 1,000 yrs B.P. The latter will be located in protected settings with good canoe launches and freshwater sources. Remains of permanent houses, winter seasonality indicators (e.g. fur seal bones), and high diversity faunal and artifact assemblages are the expected archaeological signatures of such sites. Sites located at the mouths of the major rivers entering the ocean in this zone are likely to have the longest records of winter residential use.

4) One feature that is quite distinctive about this zone relative to Zone Ia is the presence of numerous rocky islets and headlands. These are important sea mammal and marine bird rookeries. Taken together with the narrowing of the continental shelf in this zone, late prehistoric sites should contain higher frequencies of sea mammal bones, bird bones, egg shells, and offshore fish. These should occur in sites occupied mainly during the spring and summer seasons. Offshore islands with sufficiently level ground for site formation should contain faunal and artifact assemblages characteristic of field camps or location site types. For fully maritime collector systems, the usage of certain islands such as Destruction Island as stations for monitoring sea mammals over a broad area can also be expected.

Some of these same landforms, especially those capable of being easily defended, have potential for the presence of remains of fortresses or redoubts. The age of such remains is not expected to be great and this suggestion is based upon the belief that warfare on the Northwest Coast developed after the emergence of collector land use systems, especially maritime collector systems.

5) The locations for exploiting many of the marine resources available within this zone are extremely stable in their spatial distributions through long periods of time. Also, different marine resources may be available in the same location at different seasons. In addition, areas of level ground that are habitable and suitable for archaeological site formation are somewhat limited. The combined effect of these factors over long time periods will be usage of the same places as locations and field camps in the procurement of a number of different resources. An offshore island with bird rookeries, for example, might be used occasionally as a location during spring forays for birds and bird eggs, as a location (butchering site) by whale hunting parties in the early summer, as a station for monitoring sea mammal movements, as a field camp in late summer for offshore halibut fishing activities, and even as a defensive retreat during intervals of intergroup conflict. Occupations of these different types accumulating in the same place imply complex deposits. Assemblages deriving from such deposits will occur as blends that are less diverse than assemblages from winter residential sites but more diverse than a single function, special purpose field camp or location. Consequently, these deposits and their assemblages may be quite difficult to place into simple site types.

6) The intertidal zone of this management unit has potential for archaeological features such as fish traps and canoe ramps. Both of these types of features are likely to be manifested as alignments of rocks.

7) There is some potential in this zone for the occurrence of bird-netting locations. The netting of birds and waterfowl typically required a natural feature that funneled the flight patterns through defiles where a net could be strung. The most obvious examples of such features would be isthmus settings in which waterfowl fly low to the ground between two bodies of water or embayments.

8) Inland use within this zone is likely to be very thinly represented. Cultural modified cedar trees may occur where canoe-making, plank-splitting, and bark-peeling occurred. But these are likely to be the most obtrusive remains here. The perishability of much of the technology involved, curation of important non-perishable tools (e.g. adzes or axes), and the lack of repetitive use of the same locations are all factors that should contribute to an archaeological record with few material traces. When the very poor conditions for detecting archaeological remains in densely forested settings are also considered, it becomes apparent that archaeological remains situated away from present or former shorelines will be extremely rare.

Zone Ic: Cape Johnson to Cape Flattery

This zone is most salient for the unusual productivity of marine resources and its small, unproductive salmon streams. The continental shelf is quite narrow here (less than 5 km off of Cape Flattery) and this means that sea mammals that follow the shelf in their north-south migrations come much closer to the coastline than they do anywhere else on the Peninsula. Also, halibut can be taken in abundance here for similar reasons. Sea mammal haul outs occur at a number of locations within this zone. This zone is characterized by a high energy shoreline with limited potential for hardshell clam production. On the other hand, shellfish species that are adapted to high-energy, rocky beaches are relatively abundant in this zone. Streams which enter the ocean within this zone are all quite small with limited diversity and productivity of salmon runs. The Ozette River constitutes a somewhat more productive stream than others in the northwestern Peninsula due to its lacustrian rearing habitat for sockeye salmon, but relative to the Peninsula's larger rivers is still a minor salmon-producing stream.

Current Archaeological Information on the Coastal Margin Zone Ic.

Beyond the archaeological survey data mentioned above, current knowledge about the archaeology of this zone comes mainly from the intensive investigations at the Ozette site carried out by Washington State University between 1966 and 1982 (Croes 1977, 1980; E. Friedman 1974, 1976a, 1976b, 1980; J. Friedman 1975; Gleeson 1973a, 1973b, 1973c, 1974a, 1974b, 1980a, 1980b, 1982; Gleeson et al. 1976; Gleeson and Fisken 1977; Gustafson 1968; Huelsbeck 1981a, 1981b, 1983a, 1983b; Kirk and Daugherty 1978; Mauger 1975, 1978, 1980a, 1980b; Wessen 1982; and others). These investigations involved large scale block excavations of water-saturated deposits that span from the late prehistoric period through the early 20th century. The focus of research was in one area of the site (Area B) and especially on deposits estimated to have accumulated during the past four centuries. Minimal information is available on the excavations of earlier deposits at the site but it has been reported that a well developed maritime adaptation has been in existence for 2,000 years B.P. (Gleeson 1980:8; Gustafson 1968; McKenzie 1974 cited in Gleeson 1980:8) and that whales probably have been hunted for at least the past 2,000 years (Kirk and Daugherty 1978:92). A few other sites in Makah traditional territory such as Cannon Ball Island, Warmhouse (45CA204), Tatoosh Island (45CA207), and Sooes (45CA25), were tested in conjunction with the investigations at the Ozette site (E. Friedman 1974).

A collection of approximately 50,000 artifacts as well as more than a million faunal remains was recovered from Ozette (Gleeson 1980b:15; Huelsbeck 1983a:71). The scale of this project has never been equalled in western Washington and the degree of preservation of wood and fiber is quite remarkable. Because the Ozette Project is the centerpiece of all archaeological studies done to date in the Coastal Margin zone, I will attempt to identity some of the more significant contributions of the project as well as the questions for future research that it raises.

Nearly 80% of the artifacts recovered are of materials that are perishable under typical conditions in most archaeological deposits (Gleeson et al. 1976:25). This unusual circumstance was the basis for an equally unusual approach to the investigation in which "...the focus has been on studies that are ethnographic in nature although based upon archaeological materials" (Mauger 1980a:2; Gleeson 1980:1). The Ozette material is characterized (Mauger 1980:2) as "...an essentially ethnographic collection recovered from an archaeological context."

Ethnographic studies are generally synchronic and, with only minor qualifications, the Ozette archaeological investigations are synchronic in nature as well. Accordingly, the study sought to confirm observations made in ethnographic sources (Drucker 1951) and ethnohistoric accounts (Swan 1869). In some areas, such as the question of how social rank might be expressed in the spatial distribution of artifacts and debris on house floors, support was found for the ethnographic sources. For example, higher densities of wood chips (Gleeson 190b:178), fish bones (Huelsbeck 1980:55), rare shellfish (Wessen 1982:177), and ornamented artifacts. (Gleeson 1980b:177, 179) were interpreted as confirmation that the highest ranking individual in House 1 occupied the northeastern corner of the house. In regard to subsistence, minor discrepancies with ethnographic statements about relative importance of fish species were suggested from the analyses of Ozette fish faunal remains. Perhaps the most dramatic lack of correspondence between ethnographic expectations and archaeological observation at Ozette was that fur seal bones were very abundant (Huelsbeck 1981b:24). Ethnographic accounts for the Nootka represent fur seal hunting as an activity that only developed in the late 19th century as a response to the fur trade (Drucker 1951:46). Ethnohistoric accounts of the Makah apparently do not mention fur seal hunting (Huelsbeck 1981b:24).

At least eight houses were identified and three of these were excavated in their entirety (Gleeson 1980b:15). In western Washington and especially for coastal sites, the opportunity for the exposure of entire houses is relatively unique and a number of interesting patterns were identified. It was found that densities of artifacts and debris were highest in exterior midden areas between houses and lowest in central house floors. On the basis of fish species represented in the Ozette fauna (Huelsbeck 1981b:46, 60) and analyses of shellfish growth rings (Wessen 1982:142), it is suggested that some people occupied the Ozette site throughout the year. According to ethnohistoric sources (e.g. Swan 1869), Ozette (Hosett) was one of five Makah winter villages occupied during the late 19th century (Swan 1869) and the occupants of these villages shifted summer residence to two summer fishing camps.

The preservation of house structural elements made it possible to document details about house architecture (Mauger 1978). Information recovered included numerous details about technical solutions to the maintenance of houses, provision of adequate drainage in a setting with heavy precipitation, patterns of refuse disposal, and spatial distribution of materials across the floors of houses.

Potentials for Future Research Raised by the Ozette Investigations

The investigations at Ozette have raised a number of questions likely to be of relevance in future research. Certain questions must be answered before the full potentials of the Ozette data base can be realized. Some of the questions that emerge from a reading of reports and theses available at this time on the Ozette site investigations include:

1. What is the age of the Unit IV and V deposits that were the focus of the excavations at Ozette?

There is uncertainty about the age of the deposits that were the focus of the investigations (Gleeson 1980b:39). Only one radiocarbon date is available, a date of 440 ± 90 yrs BP (WSU-1778), from the Units IV and V deposits. Dendrochronological age estimates (Gleeson 1980a:81-84, 1980b:39), additional radiocarbon dates, and any other means for gaining finer temporal control are critical to placing the site assemblages into a chronological context. As Gleeson (1980b:185) points out, the Ozette data base has great potential for yielding valuable information on the nature of culture change in response to European contact. This potential can not be realized as long as assemblages that straddle the interval before and after European contact are treated as a composite assemblage. More rigorous chronological control would permit diachronic comparisons between the Ozette assemblages or between Ozette assemblages and those from other sites.

2. How does the Ozette data base fit into a regional system of settlement and land use?

This question is similar to the preceding one in that it identifies the desirability of providing better contextual perspective—in this case a spatial context in a settlement geography. To be able to make generalizations about subsistence, faunal frequencies, or artifact assemblage frequencies it is necessary to have an explicit understanding of the role this site played within an annual cycle of seasonal movements. It is, in other words, necessary to identify how this part relates to a larger whole. In efforts to characterize the overall dependence upon certain subsistence resources, there has been a tendency to assume that the patterns revealed at a single site can be extrapolated to the whole cultural system (E. Friedman 1974; Huelsbeck 1983a, Wessen 1982). For example, there is a suggestion of inconsistencies between the ethnohistoric record for the Makah and the archaeological record at Ozette in regard to relative importance of fish species exploited (Huelsbeck 1980a:52). There may be an incongruity between ethnographic observations of a system at multiple points on the landscape throughout a yearly cycle and the archaeologist's view of a system from a single place on the landscape that was probably only occupied for a part of the year. To relate and compare these two rather different kinds of observations in a reliable way will require more systematic attention to models of settlement geography and seasonal round.

Ethnohistoric accounts describe Ozette as a winter village location but there are some faunal indications of year round occupations at this site (Huelsbeck 1981b:60; Wessen 1982). For meaningful comparisons or contrasts to be made between the archaeological and ethnohistoric/ethnographic records, however, further consideration should be given to the complexity of native settlement strategies. For example, Northwest ethnographers frequently describe seasonal rounds with the qualification that certain individuals, especially the elderly and the infirm, might not follow the general pattern of the local group. If similar patterns are considered for the Ozette site, it becomes clear that the archaeological record there may be more complex than has been fully appreciated. Following the implications of this example one step further, can it be assumed that the spatial patterning of artifacts and debris on a house floor accurately reflects the social organization or subsistence pattern of all occupants through a yearly cycle or merely the individuals that happened to be using the structure immediately prior to abandonment? If, for example, the elderly individuals left behind by younger co-residents during the summer disproportionately consumed certain resources (e.g. shellfish), the residues occurring on a house floor in late summer might not be in any sense representative of consumption patterns for the entire household group. Similarly, a house abandoned (or covered by mud) in late winter is likely to have rather different food residues and spatial patterning of those residues on its floor than the same house would have if abandoned in late summer. In other words, archaeological inferences about normative subsistence activities and social class must control for season of site abandonment, age-dependent differences in seasonal subsistence, and site formation processes before patterning at one site can be related to a whole cultural system.

There is an obvious need for better information from sites representative of all components of the seasonal settlement/subsistence cycle. For the northern outer coast of the Olympic Peninsula, information from one of the sites recorded historically as a spring/summer fishing and sea mammal hunting camp would yield invaluable comparative information on functional/seasonal differences in artifact and faunal assemblages. Such data would enhance the capacity for placing the Ozette site data base into a context of regional land use systems. Preliminary efforts along these lines have been undertaken (E. Friedman 1976) but the scale of such efforts has been far too small to produce samples necessary to document frequency differences in faunal assemblages or artifact forms in a representative way.

3. What kinds of archaeological data likely to be present at Ozette should be sought if this site is reopened in the future?

Given the synchronic/ethnographic quality of the work done to date at Ozette, there has been no significant attention to deposits predating the past four centuries although deposits are present which are at least 2,000 years old (Gleeson et al. 1976:1). Despite the large scale of excavations at Ozette, it is interesting to keep in perspective the small fraction of the entire site that has been excavated. The site is described as being almost a kilometer long (Huelsbeck 1983a:1). Considering the complex occupational history that may have extended over several centuries, it is appropriate to ask how the use of one locality might relate to the much larger site. The frequency of massive mudslides at this particular portion of the site raises the possibility that this location was not the best along this stretch of shoreline for house building. What implications might this have for the representativeness of excavations at this locality?

Some of the Ozette project research objectives can only be realized when diachronic data are available. The antiquity of whale hunting, for example, was one of the research topics of the project (Gleeson 1980a:1) but to address this topic will require recovery of information from deposits earlier than those that have been reported from Ozette. The development of fur seal hunting as a component of subsistence is a subject for which similar data are required. According to Philip Drucker (1951:46), the hunting of fur seals by the Nootka was an activity that developed in response to the fur trade. The Makahs also hunted fur seal during the late 1800s (Swan 1869) but the archaeological faunal record from Ozette suggests this activity occurred in precontact times as well (Huelsbeck 1981b:24; Gustafson 1968). When this activity began and the nature of its development are questions that relate directly to the larger issue of how maritime adaptations evolved in this region. Well controlled artifact assemblages minimally spanning the past 2-3,000 years will probably be necessary to address these sorts of diachronic questions effectively.

4. What are the sources for the precontact metal artifacts at Ozette?

The occurrence of iron artifacts (blades and chisel bits) hafted in wooden handles associated with deposits estimated to be about 400 years old (Gleeson 1973a:11-12, 1974c:10, 1980b:11, 50, 52; Mauger 1974:12) is interpreted as "incontrovertible evidence" of the use of iron prior to European contact. Similarly, copper sheet metal was found in the same deposits. Japan has been identified as the most plausible source for the iron and the agent of transport the Japanese Current (Gleeson 1980b:52). Possible sources for the copper include trade from northern Northwest Coast (if native copper) or Japan via the Japanese Current (Gleeson 1982:41).

Obviously, the uncertainties surrounding the age of these deposits are an obstacle to determining the significance of these items. If these metal artifacts eventually prove to date from the late 18th century, the significance that has been attributed to them might be cast in a rather different light. In view of the apparent lack of reported evidence for a precontact iron working industry or reported occurrences of iron in other precontact sites elsewhere in the Northwest, a precontact occurrence of these tools warrants closer scrutiny. This is especially the case inasmuch as the Makah were an important node in native trade networks (Singh 1956; Gleeson and Fisken 1977:36; Croes 1977, 1980) and they apparently had substantial numbers of the metal tools (Gleeson and Fisken 1977:21, 23). Could it be that archaeologists have simply overlooked these items elsewhere? Compositional analysis of the metal artifacts in conjunction with more rigorous chronological control would seem essential to a resolution of this puzzle.

There has been no systematic study and reporting of the historic artifacts from this site. This is essential to accomplishing one of the research objectives identified during the project—the consideration of culture change from the protohistoric into the historic period (Gleeson et al. 1976:24; Gleeson 1980; Mauger 1975:21-22).

5. Social Organization and Intrasite Spatial Analysis

In view of the great anthropological interest there is in when, how, and why socially stratified cultural systems evolved on the Northwest Coast, any means for measuring social rank are of considerable importance. A central theme in the spatial analyses at the Ozette Site and especially House 1 was that social rank was expressed in the spatial distribution of artifacts and debris on the house floor. The various spatial analyses of the floor of House 1 are based upon a model of the living floor preserved in situ by a mudslide:

It is a look at the past never before possible, for no archaeological opportunity anywhere quite matches this one. The mudflow in effect stopped the clock at Ozette, leaving virtually everything preserved in place, somewhat as volcanic ash has preserved a record of everyday life in ancient Pompeii (Kirk and Daugherty 1978:94; emphasis added).

Following upon this view of the Ozette as being a Pompeii-like deposit, the distribution of artifacts and debris were envisioned as being "frozen in time" on a living floor just as they had been moments before the catastrophy occurred. Taking this model of post-depositional events one step further, the spatial analyses of artifacts and debris across the floor of House 1 were viewed as frozen manifestations of intrahouse social organization. Based upon the occurrence of high densities of certain kinds of debris across the north end of House 1, it was argued that the highest ranking individual in the house lived in this corner (e.g. Gleeson 1980a:73; Huelsbeck 1980a:55, 1981b:89-97) as is suggested in Nootkan ethnography (Drucker 1951). This area of House 1 at Ozette was characterized by high frequencies of wood chips, fish bones, and rare species of shellfish. In effect, the argument for the locus of high rank within this house is that it is the messiest area of the house with the greatest quantities of the same kinds of debris that occur in exterior midden deposits.

A particularly problematic aspect of the spatial analyses that were conducted is that they assume but do not demonstrate that the archaeological patterning they identify is the result of socially regulated human activities. Before this approach can be generalized and applied to other sites, there is an alternative model that deserves serious consideration. I am suggesting that the "Pompeii model" may not be an appropriate one for the spatial distribution of debris in portions of the Ozette deposits. The "Pompeii model" is problematic because much of the cultural debris distributed across the areas within the outline of House 1 was quite clearly deposited there by the mudslide itself. The architectural data from the Ozette site (Mauger 1978) support this interpretation and are inconsistent with the assumption of in situ preservation beneath the mudslide. These data may be summarized briefly.

Houses 1, 2, 3, and 4 were in the path of a massive mudslide which both structural and stratigraphic evidence indicate was "rapid and catastrophic" (Mauger 1978:3, 47). The slide moved downhill not merely gently covering everything in its place but instead it smashed into the houses with considerable violence. Portions of houses located uphill may have been distributed downslope where other houses stood (Mauger 1978:161). Walls were collapsed and apparently the roof of House 1 was carried onto the beach beyond the archaeological deposits (Mauger 1978:50, 144). Wooden wall support posts of House 1 were sheared off at the ground along the east wall (Mauger 1978:50, 147, 197). Some midden was "plucked up or ploughed by the slide" (Mauger 1978:47). The exterior midden deposits that had accumulated to a depth of as much as two meters or more against the house walls (Mauger 1978:74) would have been smeared across the floor of House 1. There is a high density of artifacts and debris in exterior midden relative to interior midden (Gleeson and Fisken 1977:12, 39; Wessen 1982:158; see also Mauger 1978:174, 175, 186). It may be no coincidence that the northern portion of House 1 where the effects of plowing were most pronounced is also the part of this structure which proved to have the highest densities of the debris classes that were interpreted as indicators of high rank. Where the effects of the slide were least pronounced in House 1 and where exterior midden deposits were not smeared across the house floor (Mauger 1978: 153), the densities of debris were markedly lower.

While there is no reason to suggest that all spatial patterning on the floors of the Ozette houses was produced by the mudslide itself, there can be little doubt that the "Pompeii model" ignores important evidence for significant spatial displacement of artifacts and debris during the event that produced the excellent organic preservation at Ozette. Numerous figures representing the floor of House 1 show a high density "smear" of materials that extends from east to west across the northern portion of the house floor (e.g. Gleeson 1980:Figure 18; Huelsbeck 1981b:Figures 20, 21, 22, 24, 25, 26, 27, 28, 29, 30; Mauger 1978:Figures 35, 41, 47; Wesson 1983:Figures 21, 22, 23). Before social inferences can be reliably drawn from the Ozette data—the artifact, faunal and debris distribution patterns on the house floors—it is essential that the effects of mudslide displacement be systematically examined and adequately controlled for in any spatial analyses. An explicit discussion of site formation processes would provide the foundation upon which most other analyses and interpretations could be built.

Archaeological Expectations

This zone has little to offer to people who were not fully maritime but, on the other hand, once maritime adaptations developed, the marine resource potentials of this zone were unsurpassed elsewhere in the region. The contrast between marine versus terrestrial and riverine food resource productivity is exceptionally sharp in this zone. For this reason, this zone presents an unequalled opportunity to test models for the emergence of fully maritime systems of land use. Geological conditions for site preservation throughout the Holocene should be as good here as anywhere on the Olympic Peninsula if not the entire Northwest Coast. If the paleoenvironmental data suggesting that the outer coast of the Peninsula has been rising are accurate (see Chapter 5), the absence of archaeological remains associated with early maritime land use systems can not be explained away as a preservation problem. If clear evidence for early and mid-Holocene maritime adaptations can not be found here, it will be difficult to fall back on the sea level change argument.

Some of the many expectations for this zone that flow out of discussions and models presented in previous chapters are as follows:

1) This zone is ideal for testing the "early maritime hypothesis" because, if it is valid there should be a rather well preserved record here of human use spanning much of the Holocene. If marine resources played a major role in early and mid-Holocene subsistence, then the archaeological record of this area should be rich in archaeological remains associated with these time periods. In contrast, the models developed in this study lead to the expectation that this zone will have minimal archaeological evidence predating the development of riverine collector type land use systems on the Peninsula. In fact, much of the prehistoric archaeological record here may postdate the emergence of maritime collector land use systems. Dating of the earliest remains of winter village sites in this zone should be highly informative regarding the question of how long fully maritime adaptations were present on the Olympic Peninsula.

2) This zone will be important for investigating the evolutionary relationship between the riverine and maritime collector land use strategies (see Chapter 7). If the maritime collecting system of land use developed subsequent to the riverine collector systems in this region of the Northwest Coast, there are implications for the archaeological record of this zone. It was postulated that winter-sedentary land use systems were already established along the Peninsula's larger rivers as much as 1,500-2,000 yrs before the appearance of maritime collector systems. This would imply that usage of this section of the coastal zone would have been based out of winter residential sites located outside of this zone (especially in Zone II). Site types expected to occur in this zone might include field camps and locations but not winter residential bases. In other words, there should be an absence of permanent village sites during this interval (ca. 3,000-1,500 B.P).

To the extent that late winter/early spring were always stressful periods for aboriginal subsistence systems of this region, sites associated with this season should be particularly well represented. Such sites are expected to be strongly associated with the exploitation for immediate consumption of marine resources that contain significant amounts of fat or carbohydrate. The importance of protein-sparing foods was discussed in Chapter 6 and there are a location number of marine resources that could fulfill this requirement (e.g. sea mammals, waterfowl, herring eggs, and shellfish).

3) Also assuming that the first collector systems on the Olympic Peninsula would have been centered in villages on the larger rivers in other areas of the Peninsula, there may be some archaeological evidence of usage of this zone by logistic task groups from such villages that visited this area during the summer months for specific resources. Summer season field camps are, therefore, expected to be an important archaeological site type present after the emergence of riverine collector systems in this region (estimated at around 3,000 B.P.). These sites should lack evidence for permanent houses and should be dominated by non winter season seasonality indicators.

4) After the postulated emergence of maritime collector systems in the region around 1,000 yrs B.P., winter villages should appear within this zone. These will contain remains of permanent houses, numerous winter seasonality indicators, and high diversity artifact and faunal assemblages. Evidence of heavy wood-working should be well represented reflecting the importance of house construction as well as the manufacture of large canoes for the increased logistic requirements of the fully maritime land use system.

5) Since the staple storage resources in this region were necessarily non-salmonid fishes (especially halibut), the location of winter village sites should be correlated with environmental factors other than the distribution of salmon. Protection from the weather, access to various marine resources, sources of freshwater and potential for canoe launching should be among the primary variables determining residential site locations.

6) The maritime collector system is a more complex economic system than the riverine collector system for reasons discussed in Chapter 7. This implies larger villages, larger cooperative labor groups, and more complex division of labor with more specialists of various sorts. One of the archaeological correlates to these points is that intrasite variability and intra-household variability in artifact assemblages should be greater than for riverine collectors.

7) The maritime collector system is likely to involve a greater degree of social ranking than the riverine collector system. Despite the great interest that the evolution of ranking has in Northwest Coast anthropology, the archaeological measurement of social rank presents a significant challenge. Mortuary data have most commonly served as the basis for making social inferences in other regions, but can not be expected to do so here. Investigations at the Ozette site have identified archaeological patterns suggestive of differential economic status and access to wealth objects or scarce resources. This is promising but, in general, it seems likely that until a number of more basic questions about prehistoric subsistence and land use are better understood, it will be difficult to address problems related to social organization. Also, it seems likely that questions of social organization will require far more intensive and extensive excavations than have been accomplished to date in the Northwest. Large scale efforts that encompass multiple houses and intrahouse areas may be necessary and, if diachronic or inter-regional comparisons are to be made, such excavations would have to be carried out at multiple sites as well.

8) After the appearance of maritime collector systems in the region, relatively few field camp sites will be produced. This expectation reflects the high degree of logistic mobility characteristic of these systems and the large size of their foraging zones (see Chapter 7).

9) The high degree of logistic mobility associated with the coastal collector system postulated after 1,000 yrs B.P. for this area should manifest itself in residential site faunal assemblages that include marine species from a wide variety of different habitat types that extend well beyond the immediate vicinity of the site itself. In contrast, faunal assemblages from resource procurement sites in this zone that are associated with the earlier riverine collector systems should reflect a much more spatially restricted suite of species and especially those obtained in the immediate vicinity of individual sites.

10) Although the coastline through this zone has been covered in previous archaeological surveys, these investigations have apparently not specifically attempted to identify cultural resources in the intertidal zone (Wessen 1985:30). There is reason to believe from paleoenvironmental evidence (see Chapter 5) that this particular coastline has less potential for submerged archaeological deposits than some others because it is emergent. This is, however, a model the validity of which has yet to be demonstrated. Whether or not there are remains of prehistoric occupations in the intertidal or even subtidal portions of this zone, late prehistoric rock features (e.g. fish traps, canoe ramps) are known to occur and may be found in places where attention in previous surveys has been focused only above the high tide line.

11) Expectations 4 through 8 that were presented under Zone Ib are also relevant to Coastal Zone Ic.

Zone Id: Cape Flattery to Crescent Bay

The distinctive features of this zone are that it has good access to sea mammals and offshore fisheries, a moderate energy shoreline with slightly better shellfish potential than Zone Ic. It is strategically located relative to deep water and especially the highly productive halibut banks to the northwest of Tatoosh Island. Like Zone Ic, this zone is characterized by salmon streams that are small and relatively unproductive. Large scale archaeological excavations have been undertaken over a period of several years at a complex of sites on the Hoko River (Croes and Blinman 1980; Miller 1984; Croes and Hackenberger 1988). Still in progress, this investigation, the Ozette Project, and the Manis Mastodon site represent the only intensive, multi-year site investigations that have occurred on the Olympic Peninsula. No portion of this zone is encompassed within the boundaries of Olympic National Park. Expectations for archaeological patterning in this zone, however, would be rather similar to those for Zone Ic.

Zone Ie: Crescent Bay to Mouth of Hood Canal

This is a lower energy zone with more sandy beaches and spits than Zone Id and its hardshell clam resources are, therefore, relatively productive. Rivers with good (the Elwha) to moderate (e.g. Dungeness) salmon productivity enter the Strait of Juan de Fuca within this zone. Sea-mammals, except for harbor seal, are less abundant than in Zone Id. The diversity of sea bird species is lower than in the Outer Strait and the Outer Coast zones. This zone, along with the river valley and lowlands that lie inland from it (Zone IIc), are exceptional within the Peninsula region for having low precipitation. Within the region, this area was characterized by relatively high productivity of terrestrial plant and animal resources. No portion of this zone is encompassed within Olympic National Park.

Zone If: Hood Canal

The waters of this fiord are the most protected (low energy) of any in the region and this accounts for this area's high shellfish productivity—especially of hard shell clams. In terms of sea mammals, this zone is a cul de sac with both low diversity and abundance. With one important exception (i.e. the Skokomish River), the streams of this zone have limited salmon production and small stretches of spawning habitat in their lower reaches. The archaeological record of this zone is very poorly known. Due to only very restricted areas of low relief along this side of the Peninsula, there has been very substantial historic impact on the archaeological record of this zone (e.g. highway and residential development). No portion of this zone is encompassed within Olympic National Park.

Zone II: River Valleys and Lowlands

This zone includes all areas of the Peninsula that occur between the inland boundary of the Coastal Margin (Zone I; 5 km from the mean high tide line) and the lower elevational boundary of Zone III (the 2000 ft contour interval). This zone is relatively wide along the western and northern Peninsula but quite narrow along the Hood Canal side of the Peninsula. This was an extensive and densely forested zone broken only by a limited number of small prairies and areas where disturbances such as fire temporarily created early seral stages. The entire zone is occupied year round by resident elk and deer populations and, during the winter, by migratory herds of these same animals.

There are basicly three subareas within this broad zone that are expected to be of particular importance to hunter-gatherers: the river terraces, prairies, and the areas between 1,500 and 2,000 ft elevations where winter ranges of migratory game occur. Each of these subareas warrant further comment and may eventually deserve designations as separate zones.

Most of the rivers of the Peninsula have only short reaches of low gradient in their lower portions. Even where the coastal lowlands reach their greatest width on the western Peninsula such as along the Lower Quinault, major portions of these rivers are characterized by braided channels. These channels have changed dynamically even in the past century (Wessen 1978a:13) and, therefore, are not conducive to the preservation of sediments of substantial age. Stream flows are quite variable with order-of-magnitude differences between summer and winter a typical situation. Positioning of winter villages sites, therefore, is not expected to have been within the active flood plain or where flooding was a frequent occurrence.

As was discussed in Chapter 5, the rivers of the Peninsula are characterized by a four terrace sequence. These terraces were formed by climatic events and consequently are of similar ages in different river valleys. In Figure 8.2 a generalized model of the terrace sequence, vegetation types, and ages is illustrated. This geological pattern in terrace formation apparently applies throughout the Peninsula and has profound importance for how archaeological survey is done as well as how the results of survey are interpreted. The river terraces are likely to be major areas in which prehistoric sites of all ages and a wide variety of types occur throughout the Peninsula. In certain zones, they are likely to be the kinds of landforms upon which residential sites will be found (especially Zone IIa, Zone IIc, and Zone IIe). An awareness of the approximate ages of these terraces is invaluable to anticipating the archaeology of this zone because the first two terraces above the modern floodplain are surprisingly recent. In fact, only the third terrace above the alder flats is old enough to have in-place archaeological remains that are greater than 750 years of age.

Figure 8.2 River Terrace Ages (yrs B.P.) and Forest Succession (modified from Wessen 1978a; data from Fonda 1974). (click on image for a PDF version)

Most of the known prairies occur within this zone and the few exceptions occur in the Coastal Zone. Although a few of the prairies are larger (e.g. Quillayute, Forks, and Sequim) most of the recorded prairies are quite small—generally less than 100 acres in extent. Most of the prairies for which detailed locational information was found occur in the western Peninsula, but there are ethnographic and historical accounts of prairies in the northeastern Peninsula as well (see Chapter 4; Gunther 1927; Onat and Larson 1984:43; L. Smith 1951).

A general expectation that applies to all of Zone II is that prairies will generally occur within the foraging radius of winter villages. Where there was a riverine collector land use strategy with villages distributed at some distance upriver as in Zones IIa and IIe, prairies are expected to broadly parallel the river valley distribution of villages. In those areas characterized exclusively by coastal villages (e.g. Zone IIb), it is expected that the distribution of prairies will not parallel river valleys but rather will be situated within about a 10 km radius of saltwater. Examination of the distributions of prairies in Figure 3.3 in this light is intriguing. The general distribution of prairies on the western Peninsula is consistent with this explanation. If the locations of other prairies can be identified, it will be interesting to see if the same relationship is maintained.

It is important to note that many prairies may have been encroached upon by forest vegetation prior to being documented by Euroamericans. This process has been documented for southern Puget Sound and the dramatic differences between prairie and forest soils may disappear quickly as the forest reclaims an area of prairie (Bryan in Tuohy and Bryan 1958:45). On the other hand, at one locality in Clallam County old growth spruce and hemlock were reported growing on what seemed to be prairie soil (Smith 1951:53). Identification of former prairie areas then constitutes an important research challenge for which there are not obvious methodological solutions presently. This is a subject of much needed research on the Olympic Peninsula and western Washington generally.

The areas between roughly 1,500 and 2,000 ft elevation comprise a third kind of setting within Zone II that is of considerable archaeological importance. These areas occur as fingers of land that extend up the river valleys well into the central portion of the Park. The concentrations of animals that occur here in winter are hypothesized to have been the mainstay of winter season subsistence for Old Cordilleran subsistence systems (see Chapter 6). The third terraces above the floodplain and other benches in these "foothill" settings are expected to be areas of high potential for early and mid-Holocene archaeological sites. Locations of particular importance will be those settings with southerly solar exposure. [1] These are settings where winter snow accumulations are reduced and, as a result, where forage availability is maintained for large herbivores. The terraces and benches along the valley walls offer two additional locational advantages for hunters—these places would tend to have commanding views of the surrounding landscape at times in the past when the forest was more open and they tend to be warm places in the winter due to cold air drainage.

At least two archaeological sites that fit this general profile have been identified in the recent past. One of these sites was found along the Elwha River during the archaeological reconnaissance associated with the present study (see Appendix A) and another is known from just outside the Park's southeastern boundary (Schalk 1985b).

Zone IIa: West Slope Rivers and Lowlands

The width of the lowland zone reaches its maximum for the Peninsula in this zone. The rivers here are relatively large and productive and supported riverine settlement patterns during the early 19th century. There are numerous prairies in this zone which were present at the time of initial Euroamerican settlement. Rainforests occur in the river valleys. It is estimated that 85% of the total elk on the Olympic Peninsula occurred in the west slope river valleys in historic times and this zone coincides with the mapped distributions of migratory elk winter ranges (Schwartz and Mitchell 1945:305).

Current Information on the Archaeology of the West Slope Rivers and Lowlands

Archaeological investigations in this zone include the exploratory efforts of Albert Reagan (1917). He provides the first descriptions of archaeological sites at inland locations. One of these sites was located on the Hoh River and was described as follows:

An ancient midden heap is also to be found on the Hoh River some 16 miles inland at a place called the "bench", on a benched area where the Olympic glacier made a stand on its retreat up the mountains from the coast (Reagan 1917:16).

Reagan also mentions a number of other archaeological sites at inland locations—"at Beaver Prairie, Forks Prairie, Quillayute Prairie, and at various camping places along the Quillayute River and its tributaries (Reagan 1917:9)."

Reagan's observations about the kinds of materials recovered from midden sites are not clearly linked to specific places or sites but apparently the "oven-mounds" he described (Reagan 1917:11) were observed on these prairie sites. [2] The study of such sites would obviously offer the rare opportunity to learn something about the prehistoric uses of vegetal food resources in this region.

In 1977, Wessen (1978a) carried out an archaeological reconnaissance of a series of localities along the valleys of the Quinault, Queets, Hoh, and Quillayute rivers. Using the "direct historic approach", this reconnaissance focused on the inspection of places that were reported in ethnographic sources (e.g. Olson 1936; Singh 1956, and others) and by local informants (apparently non-native) as the former locations of "settlements". There are 45 localities listed as places that were surveyed including 25 along the Quinault River, 5 on the Queets, 7 on the Hoh, and 8 on the Quillayute. [3] In an attempt to cope with the problem of site discoverability, soil auguring was emphasized although some other techniques were used in an effort to identify sites (e.g. soil pH and phosphate tests, and vegetation pattern analysis). Although Fonda's (1974) vegetation succession model for river terraces was used in the identification of landforms, it is unclear how the survey was conducted relative to these landforms.

The cultural resources identified were described as mostly "fire hearth areas, old cedar stumps, and historic homesteads (Wessen 1978a:58)". Charcoal-rich stratigraphic units (considered non-cultural) and individual fire-cracked rocks were observed widely throughout the areas examined (Wessen 1978a:60, 65). A variety of lithic artifacts (including a projectile point, "pile driver", cortex spall tools, grooved net weights) was found but "none could be demonstrably associated with any specific site (Wessen 1978a:60)". In general though, it was concluded from the survey that

Either the survey emphasis and procedures were, in fact, inappropriate to the specific requirements of this region, or, the survey work accurately reflects the extremely limited nature of preserved archaeological features. It is the writer's impression that the survey fairly accurately reflects the archaeological potential of these river valleys (Wessen 1978a:69).

Poor site preservation rather than problems of site discovery was considered the primary factor influencing the results of this survey (Wessen 1978a:71). Even though few cultural resources were recognized, it was concluded that the site discovery techniques (mainly soil auguring) "should remain, a mainstay of survey examination."

Despite the explicit intent of this survey to focus upon the locations of relatively recent, ethnographically reported sites, the results of this survey were nonetheless generalized:

While the archaeological potential of the Western Olympic Peninsula can, in no sense, be regarded as great, the possibility of significant fortuitous discovery, particularly with respect to "wet sites," is considered to be real and should not be overlooked (Wessen 1978a:74; emphasis added).

Very large areas of land have been surveyed during numerous surveys on public lands within the River Valleys and Lowlands Zone of the Olympic Peninsula. Virtually without exception, these surveys have failed to identify aboriginal archaeological sites (e.g. Wessen 1977a, 1977b, 1978a, 1978b, WAPORA 1980; Dalan et al. 1981, Dalan et al. 1981).

Before the archaeology of the Western Olympic Peninsula is relegated to waiting for fortuitous discoveries of new "wet sites", there are many questions that will need to be addressed. Some of these questions warrant mention.

What are the material correlates one would expect to find with the various ethnographically recorded "settlements"? Of the 45 localities examined during Wessen's (1978a) survey, only 16 (36%) are described as villages or places where houses were located. All of the other localities are identified as places of fish weirs, fishtraps, or simply named places. Assuming that all of these localities were perfectly preserved as they once existed in the 19th century, which of these sites would be expected to have actual midden deposits associated with them? Here it would be worth noting that many of the coastal midden sites represent accumulations that required a millenium or more to develop. Is is likely that deposits of similar character will occur on a landform that was available for prehistoric occupation for only 200 years? In alluvial settings, how were survey efforts focused relative to the terrace model proposed by Fonda (1974)? The lower terraces are quite young in an archaeological sense. If, for example, the search for archaeological remains was focused on the first terrace which is only 400 years old, what kind of midden development is likely to have occurred in the relatively short period of time such a landform could have been used?

Relative to issues of site discoverability and the use of augurs for site survey, what are the implications of the absence or scarcity of shellfish as a major faunal resource at riverine sites? How effective is a soil probe or augur for identifying subsurface features? What kind of features can archaeologists hope to identify with an augur and how can these be identified? Given the potentially great range of physical evidence that might be encountered at such diverse locations, is it likely that the use of a soil augur can reliably determine the presence of such diverse archaeological resources?

Considering the crude maps provided in most ethnographic sources and the lack of precise locational information for most named places, how much confidence should be placed in these maps as the basis for archaeological survey?

Lastly, is it defensible to extrapolate the results of survey conducted with the "direct historical approach" to the entire archaeological record of the Western Olympic Peninsula? It would seem that the results of these surveys may be more informative about the nature of ethnographic information and the utility of the "direct historical approach" than they are about the archaeological record of the river valleys of the Olympic Peninsula.

In sum, archaeological survey in this zone and in the entire region has been conducted as if the ethnographically described cultural systems were the only ones that have ever existed in this region. While the difficulty of site detection can not be underestimated, it is also clear that the survey methods that have been employed are ones which are appropriate to the detection of large midden deposits. Wide survey transects (up to 100 m or more), auguring, soil tests for pH and phosphate, and vegetation studies are all search techniques that assume that significant archaeological resources will manifest themselves like coastal middens. If this assumption is unfounded, then it might be argued that the survey methods that have been used in the forests of the Olympic Peninsula are for the most part inappropriate to the task. It will suffice to say that accumulating archaeological evidence from the Olympic Peninsula and throughout the Northwest supports the view that archaeological remains do exist in this zone and that the failure to identify archaeological resources in these settings is at least partially a failure of methods. This subject is considered further in Chapter 9.

Archaeological Expectations

1) Old Cordilleran residential sites and locations should occur on upper river terraces, prairies, and in the foothills where the winter ranges of migratory game are located. Assuming that the forest was reasonably open between 10,000 and 6,000 B.P., there should be a widespread but low density distribution of residential sites and locations (especially hunting and butchering sites) throughout much of this zone. These remains are likely to be represented mainly by concentrations of lithics in relatively low density reflecting high mobility. Midden deposits are not expected and preservation of organic material of any kind will generally require unusual conditions (e.g. water saturation). Site features will generally be limited to small hearths composed of rocks but only infrequently containing charcoal.

2) Most Old Cordilleran sites below about 1,000 ft elevation in this zone should be related to spring through fall season occupations. This is postulated to be a season of high mobility and small group size. The determinants of spring through fall season site locations before 3,000 B. P. are not expected to be rigid with the consequence that there is little tendency for reoccupation of the same places or vertical accretion of archaeological debris. This should contribute to the archaeological formation of relatively small, low density lithic scatters. Interception of ungulate resources in low relief areas during this portion of the year would probably require the positioning of hunters near places where encounters have the highest probability of occurring. Water sources such as lakes, springs and streams are likely to be among the most predictable locations for encountering animals in this portion of the regional landscape.

3) Due to the warmer and dryer climate before 6,000 B.P., the ratio of deer to elk is expected to have been higher in the early Holocene than in late Holocene or recent times. In those unusual circumstances in which there is preservation of faunal remains in archaeological deposits that are older than about 6,000 yrs. B.P., the ratio of deer to elk remains should be relatively high.

4) Locations where there are bedrock controlled channels, falls and rapids may be particularly important for their potential to have preserved sediments and archaeological remains that are older than those present on the first two terraces. These sorts of locations are interesting not only for their potential for older sediments but because they would be natural fishing sites for people using devices for taking salmon in a non-intensive way (e.g. dip-nets, leisters, and harpoons). As long as immediate consumption prevailed, there would be no need for mass harvest technology (i.e. seines, weirs, gill nets, traps). Similarly, even after the emergence of collector systems, spring and summer salmon fishing is expected to be mostly for immediate consumption and this would also tend to focus on these same places along the river channels. Because of salmon spawning requirements, the distribution of such sites will have as an upriver limit that zone of the streamway where gravels and cobbles are replaced by boulders due to increasing stream gradient.

5) It is postulated that burning of prairies began after 6,000 B.P. in response to the closure of forests occurring at that time. These prairies should show evidence of continuous usage after 6,000 B.P. Although ungulate resources would almost certainly have been exploited on prairies, carbohydrate-rich root resources are likely to have been of particular importance in the use of these areas. The most obtrusive kinds of archaeological remains in prairie settings are likely to be earth oven features used for processing roots. Such features have been described by Reagan (1917) as being mounds rather than pits. The use of such mounded oven features for root processing may be related to the wet sediments so commonly encountered in a region with such large amounts of rainfall (Alston Thoms, pers. comm. 1988).

6) Winter villages (and the collector land use system) should appear relatively early along the rivers of this zone and prior to the appearance of winter villages on the coast. This is especially the case for the Quinault and Queets rivers which have lengthy periods of annual salmon availability and high productivity. Virtually all types of sites associated with a collector system should occur in this zone once the riverine collector system came into existence after ca. 3,000 yrs B.P. In addition to the remains of winter villages and fishing locations, cultural resources that can be expected to occur along the river channels include cultural modified trees, especially cedars. The extraction of cedar for house planks and canoes as well as the removal of cedar bark for fiber clothing can be expected to have resulted in tangible remains in places along the drainages in this zone. Because of their perishable nature, such remains will necessarily be of late prehistoric age.

7) There should be a general congruence between prairies and the distribution of winter village sites along the rivers. Identification of the latter sites may be more difficult than is the case along saltwater due to an expected scarcity or absence of shellfish remains (except in the lower reaches of rivers) as well as the usual biotic factors that obscure archaeological remains in such settings.

8) Wessen (1978:53) noted two patterns associated with the native "settlements" historically recorded:

Settlements tend to be clustered toward the lower portions of rivers, and they tend to be located near stream confluences and near falls, rapids, or other relatively permanent channel obstructions (Wessen 1978a:53).

For the collector systems of land use that arose in the late Holocene, this pattern can be expected for both winter residential sites and fishing sites because of spawning requirements of salmon. Salmon require sediments in the gravel-to-cobble size range and would not find suitable spawning habitat at points upstream from the distribution of such sediments in the streamway (see Wessen 1978a:10-13; Bauer 1971 cited in Wessen 1978a:10).

Zone IIb: Northwestern Peninsula Rivers and Lowlands

This subarea is distinctive for the low elevations and short rivers with low salmon diversity and productivity. The ungulate resources of this area are limited to the low density, resident populations of deer and elk.

Waterman (n.d.) lists 215 place names for the Makah, most of which are within their traditional territory. All but a very small proportion of these are located near saltwater. It is interesting that the two areas where there are place names shown away from the coastline are the Waatch Creek and Ozette River and Lake. Waatch Creek, a slough-like feature, is an area of archaeological potential on the basis of historical accounts and the fact that overland travel between Mukkaw Bay and Neah Bay must follow this creek. In addition, if this coastline has been uplifting as suggested in Chapter 5, then there is potential for sites that were once on saltwater to be perched at inland locations along this creek. Reagan (1917:17, 20) noted the possibility that Cape Flattery was formerly an island and that there was a "tradition of the Indians that the ocean once flowed through this area between the Strait of Fuca and the ocean." Archaeological deposits located along this slough have potential for having been formed at a time when Cape Flattery was an island.

Archaeological Expectations

1) Early and mid-Holocene archaeological remains should be relatively scarce in this zone reflecting low game densities and the absence of significant winter game ranges.

2) Archaeological sites occurring within this zone will be limited mainly to locations—places that could be reached within the daily foraging radius of residential sites located along the Coastal Margin (Zone I). These are generally expected to be places where some resource procurement activity occurred. Locations where animals were killed and butchered, where cedar trees were exploited for wood or fiber, or where medicinal plants were gathered are three examples of the kinds of locations that might be expected here. With the possible exception of culturally modified cedar trees, archaeological locations in this zone are likely to have extremely low archaeological visibility.

3) Late prehistoric archaeological sites in this zone are mainly expected to be locations within the daily foraging radius of the coastal villages.

4) Due to the coastal distribution of residential sites in this zone, it is expected that anthropogenic prairies here will be situated within about 10 km or less of saltwater residential sites and will generally not occur further inland within this zone.

5) Culturally modified trees associated with such activities as cedar bark stripping and canoe building have a high potential for occurring in this zone. Areas along watercourses have particularly good potential for culturally modified cedars due to the presence of continuously moist sediments and the possibility of transport by water.

Zone IIc: North Slope Rivers and Lowlands

This zone is distinctive for its lack of precipitation. This climatic characteristic is associated with more open forest conditions and more seral vegetation which in turn would mean enhanced terrestrial plant and game resources. However, the width of the coastal lowlands zone here is not great. There are productive salmon rivers in this zone, especially the Elwha and, to a lesser extent, the Dungeness. Only very small areas of Olympic National Park fall within this zone and these are mostly distributed along the upper valley of the Elwha River.

Archaeological Expectations

This zone has the highest potential for archaeological sites of late Pleistocene age. The reason for this is that it was on the leeward side of a system of mountain glaciers. Due to the marked dryness that must have characterized this area during the early Holocene, this area is expected to have had unusually high game densities—especially of deer. Therefore, it is expected that Old Cordilleran residential bases and locations will occur in relatively high density in this zone. The Upper Elwha has high potential for Old Cordilleran sites that were occupied in the winter season as hunting camps and locations. These should occur on the level areas and especially places with southerly solar exposure and at the mouths of tributary canyons. Such sites will be found on the third alluvial terraces or even higher benches. There is some potential for a riverine village settlement pattern along the Elwha river.

Zone IId: Hood Canal Rivers and Lowlands

This zone is exceptionally narrow compared to the other lowland zones of the Peninsula and has only limited winter range for ungulates. The rivers of this zone have their accessible salmon habitat concentrated in their lower reaches. Precipitation is much lower than on the western slopes of the Olympics but much greater than in Zone IIc. In short, this zone is characterized by limited terrestrial and riverine resource productivity. Very small areas of the Park fall within this zone and these areas are along the upper Dosewallips, Duckabush, and Hamma Hamma Rivers.

Archaeological Expectations

1) Archaeological sites associated with the Old Cordilleran land use systems are expected to occur in this zone in quite low densities due to the very limited terrestrial resources. Those that occur are likely to be on the upper river terraces and higher benches of land in the upper river valleys.

2) There should be no late prehistoric winter residential sites ("villages") located in riverine settings here. Sites associated with this interval will be limited mainly to locations; field camps will be very scarce.

Zone IIe: Skokomish River and Lowlands

This is the only large river with extensive salmon habitat on the Hood Canal. The lowland zone is wider here and the potential for terrestrial plant and animal resources is probably only surpassed by Zone IIc within the lowland zones of the Peninsula. The North Fork of the Skokomish River is the only portion of this zone that is encompassed within the Olympic National Park.

1) The potential for Early and Mid-Holocene sites in this zone is good, especially in the upper river valley of the Skokomish. These will occur as winter residential sites and locations associated primarily with hunting. They should be on upper river terraces and higher benches and in locations with southerly exposures.

2) Late prehistoric residential sites should be found on the river terraces above the modern floodplain reflecting a riverine settlement pattern postulated for this area. However, the riverine distribution of such sites would be well outside the boundary of Olympic National Park.

3) Late prehistoric sites occurring on the North Fork of the Skokomish river should include field camps and locations associated with winter hunting activities carried out by logistic parties operating out of winter villages on the lower river.

Zone III: Montane

This zone encompasses the montane and subalpine forests between elevation contours of 2,000 and 4,000 ft. and amounts to a band encircling the high country of the Central Olympics. Both montane and subalpine forest types occur within the zone defined as "Montane" here and the term is not intended to be synonymous with its usage by botanists or foresters (e.g. Fonda and Bliss 1969). During the winter season, this zone receives much of its precipitation in the form of snow. The accumulation of snow in these forest types limits the development of understory vegetation which in turn restricts forage production for deer and elk. In terms of resources for hunter-gatherers, this zone is the most food-scarce of the four major zones that are defined for the Olympic Peninsula. While migratory deer and elk pass through this zone in the spring and fall as they follow the new vegetation upslope into the parklands and meadows of Zone IV, their numbers tend to be dispersed and are probably difficult to intercept except possibly along the main trails that lead through this zone. The only plant food resources of any consequence for humans in this zone are berries and, compared to subalpine settings of the Cascades, the huckleberries here seem diffuse in their distribution and relatively unproductive.

With some exceptions, modern trails tend to follow valley bottoms rather than ridges up to the subalpine parks and meadows. Since the early explorers traversed the Olympics, most travel has followed game trails and these were also the routes originally selected by the Forest Service when it built most of the trail system in what was eventually to become Olympic National Park (Wood 1984:14). Presumably, prehistoric trails followed approximately the same routes though these must inevitably have wondered some due to tree falls, landslides and other dynamic changes over long time spans. Level ground constitutes a small percentage of this zone and restricts the potential for archaeological site formation to a small proportion of this zone.

Throughout the Olympics, the distances between the lowland forests (Zone II) and the Supalpine/Alpine zone are generally less than 10 km if one follows the trails. This is an important point in considering what types of sites might be expected within this zone. It implies that travel through this zone can easily be accomplished in less than a day. Given the scarcity of resources that occur within this zone and the time to move through it enroute to those zones above and below, it is unlikely that overnight (field) camps were often established in this zone.

The archaeological potential of the Montane Zone is the lowest of all the four major management zones identified for the Olympic National Park. Site types that are expected to occur in the Montane Zone are all basicly locations. Some of the activities that might result in production of archaeological locations might include hunting animals moving along trails, resting while in transit through this zone, and reduction of lithic raw materials occurring within this zone. Due to the exposure of bedrock in portions of this zone, there is some potential for lithic quarry sites. However, it is not yet clear that any of the lithics found in archaeological sites on the Peninsula came from bedrock sources. The occurrence of exposed bedrock in this zone also implies the potential for rockshelters. The prehistoric use of such shelters by humans in this zone is again expected to be very limited but this does not eliminate the significance of such locations as possible sources of information about the prehistoric distributions of faunal resources.

Although it is difficult to identity important food resources that are likely to have had dietary importance in this zone, it is more difficult to dismiss the possibility that resources of significance for non-subsistence purposes might occur in the Montane Zone (e.g. medicinal plants). What, if any, material remains might be associated with such activities is difficult to state. For the most part then, the archaeological record of this zone is expected to be relatively sparse and the density of the forest would only add to the unobtrusiveness of such a record. At the lower and upper elevational margins of this zone as it grades into Zones II and IV, potential for archaeological remains is somewhat greater. This is especially the case when long term climatic changes are considered because these changes may have produced at least minor elevational shifting of zonal boundaries. Expectations for the kinds of cultural resources that might occur in the lower and upper margins of this zone are similar to those already discussed for the upper river valleys of Zone II and for the lower portion of Zone IV discussed below.

Zone IV: Subalpine and Arctic

Included within this zone are all those areas of the Park that lie above 4,000 ft elevation—the parklands, meadows, glaciers, and bare rock. Human use of this zone is by necessity limited to summer season after the snow has melted. In most years, the "window of opportunity" here may extend from about late June to late September. As with Zone III, most of this zone is characterized by steep relief with few areas appropriate for site formation to occur.

According to Kuramoto and Bliss (1970:319), the transition zone between the upper subalpine forest and subalpine meadows occurs at about 4,455 ft on the north-central Olympics and, due to less snow, somewhat higher in the northeastern Olympics. In many areas of the Olympics, this transition consists of a parkland with interfingering meadows and forest. Due to the occurrence of forest fires and Holocene climatic variations, the zone of transition is a dynamic one. Therefore, placement of the lower boundary of this management zone is defined by the 4,000 ft elevational contour—an elevation well below the level at which modern subalpine forests begin to give way to subalpine meadows. The intent of using an elevation well below the modern treeline is to insure that most areas that may have been meadows or parkland during the Holocene are not excluded and also to accomodate variations in the elevations of treeline relating to spatial factors such as microclimatic variations and solar exposure.

Within this zone are some substantial areas for which there is no archaeological potential. Glaciers and steep rocky peaks are examples of such areas. For purposes of simplicity, these areas were not distinguished as a separate management zone.

Current Archaeological Knowledge

Until recent archaeological reconnaissance revealed numerous archaeological resources in the subalpine parklands of Olympic National Park (Bergland 1984a), there was a tendency among archaeologists to imagine that "...penetration into the dense forests of the Olympic Peninsula was generally minimal (Mauger 1978:7)." This view was clearly derived from the ethnographic and historical records:

West of this range (the Cascades), the country is a vast forest into which the natives penetrated seldom in pursuit of game, and through which they travelled only along a few trails. They lived along and traveled upon its highways, avoiding its interior forests as much as possible. They were a distinctly maritime people, whose front door opened just above the high tide and looked out upon natural oyster or clam beds, or a salmon fishery. The wilderness behind the house was the gloomy home of Seatco, the demon of the dark forest, with whom was associated a host of minor imps (Wickersham 1900:141).

Similarly, Wood (1976:2) states:

Occasionally the Indians journeyed to the hot springs in the foothills, but as a rule they shunned the interior. According to their legends, a god resembling a gigantic eagle or raven lived among the high peaks and his wrath would descend upon anyone who dared to enter his realm. Moreover, the Indians had little practical reason to visit the mountains. Life was comparatively easy because the sea yielded an abundance of fish, and the supply of clams and oysters was plentiful when the tide was low. The hunters killed game on the lowlands; the woman picked berries and collected roots and bulbs in the forest.

Wood's observations are at least in part derived from the accounts of Lieutenant O'Neils' expeditions into the Olympics in the late 1800s. Members of the O'Neil Expedition mentioned Quinault Indians encountered on the upper Quinault River who admitted no knowledge of the country beyond the foothills (Wood 1976:226-227, 346).

Apparently contradicting these historic sources are explicit statements by ethnographers about the use of the mountains (e.g. Olson 1936; Gunther 1927; Elmendorf 1960). Unfortunately, most of these ethnographic references to the "mountains" lack any locational details and their use of the term is quite vague relative to the management zones defined here. Zones III and IV and portions of Zone II may all be encompassed in the areas referred to as "the mountains" by ethnographers. Although there is a scarcity of references to specific places that were visited relative to these management zones, the seeming contradictions in historic and ethnographic sources regarding usage of the interior of the Olympic Peninsula may be more apparent than real.

Early in the course of the present project, it was postulated that subalpine areas were likely to have been used more intensively during the early and mid-Holocene than after the emergence of semi-sedentism in the region (Schalk 1985). This expectation derives from the fundamental differences between land use systems that emphasize residential mobility versus those which emphasize logistic mobility. The implication of these deductions is that native usage of the Olympic mountains prior to the appearance of the riverine and maritime collecting systems involved systematic exploitation of resources in Zone IV. After the appearance of the collector systems, however, the focus of exploitation shifted downward in elevation onto the ungulate winter ranges that are generally below 2,000 ft in Zone II. The archaeological record of Zone IV, therefore, seemed to hold promise as an area where the implications of this explanation could be evaluated with archaeological data.

Zone IV was selected as the focus of the archaeological reconnaisance carried out during the present project. This zone was selected for several reasons. Firstly, although the interior portion constitutes the largest land area of Olympic National Park, far less is known about the archaeology of this area than the Coastal Zone. In fact, prior to the reconnaissance carried out in 1983 (Bergland 1984a), there had been minimal appreciation for the possibility that there were significant archaeological resources located at any distance from saltwater or major stream corridors.

Secondly, the prehistoric cultural resources of this zone were expected to be subject to some of the greatest impacts presently occurring anywhere in the Park. This expectation was based upon two factors. One was an appreciation of the spatial constraints that are inevitably imposed upon patterns of human movement and land use by mountainous terrain. In other words, modern-day usage of the park was expected to have a substantial degree of spatial congruence with prehistoric usage for the simple reason that many of the same environmental variables would control human usage (e.g. availability of water, slope, solar and wind exposure, etc). Another factor was that Bergland (1984a) had recorded 23 sites several of which have been exposed to disturbance related to Park developments (e.g. parking lots) or visitation.

Thirdly, Zone IV is one of the few areas of the Park where pedestrian archaeological reconnaissance offered the potential for significant archaeological data on site distributions, artifact assemblages, and site structure from surface evidence. The collection of archaeological information through surficial information is limited to a large degree by vegetation conditions and outside the high elevation settings of the Olympic Peninsula, such settings are restricted mainly to areas of recent human disturbance.

A fourth consideration for selecting Zone IV for a reconnaissance has already been mentioned. This zone offered the potential to evaluate a hypothesis regarding early and mid Holocene versus late Holocene use of high elevation settings. The models of land use set forth in the research design postulate that there would be little incentive for travel-into subalpine settings for Late Prehistoric people who practiced land use systems of the collector variety. It was reasoned that in terms of the structure of regional food resource distributions, Zone IV offered little or nothing to collectors that could not be obtained more effectively in other zones. Efficiency and scheduling considerations in collector land use systems would restrict predation on migratory ungulates to their winter ranges. These winter ranges occur mainly in Zone II on the flanks of the mountains. It was, therefore, hypothesized that the Subalpine Zone was best suited to distinguishing between alternative models for the emergence of maritime collecting systems (see Schalk 1985c:121). The fact that archaeological patterning in Zone IV can not reasonably be dismissed as a biased sample that happened to survive the vagaries of changing sea levels promised the opportunity for a more decisive results than would be possible in Zone I, the Coastal Margin. Zone IV seems less vulnerable than others (especially Zones I and II) to recurrent arguments that explain away archaeological patterning based upon negative evidence.

The reconnaissance was focused upon localities in the Subalpine Zone in the eastern Olympics—the upper Elwha, a series of high ridges between Obstruction Peak and the Upper Dosewallips River, and the Hurricane Ridge to Obstruction Peak area. A total of approximately 115 km of the Olympic Trail System was covered and an area of roughly 300 acres examined. Descriptive results of the reconnaissance are provided in Appendix A of this report. A total of 42 cultural resources was recorded and this included 38 lithic scatters, 3 rock/cairn features, and a bark-peeled cedar. In addition, nine isolated lithic items were found. Of the 42 cultural resource sites, 40 were found in the Subalpine/Arctic Zone IV and 2 were found along the Elwha River in Zone IIc.

Site areas ranged from less than 1 m² up to one (45CA288) that was estimated to have an area of 14,130 m². This very large site was located on a bench above the Elwha River at an elevation of about 1600 ft. Excluding this one site, all others recorded appear to be relatively small in area. The mean site area for the Zone IV lithic scatters is about 157 m and the mean site diameter is about 7 m. Clearly, the average site is quite small relative to the transect intervals that have often been maintained during cultural resource surveys in this region. Even though ground surface visibility was relatively good in many areas of Zone IV, it should be emphasized that most of these sites would be missed if survey transects of 30, 50 or even 100 m had been maintained during the reconnaissance. The low density of surficially visible lithic remains compounds this problem of site detection. The average site had only 16 surficially visible lithic items for a density of about 0.1/m².

Cryptocrystalline lithic materials amounted to less than 1 percent of all lithic items observed. All other lithics were fine-grained basalt. The frequent occurrence of cobble cortex and the scarcity of flakes with well developed bulbs of percussion are interpreted as evidence that primary reduction of basalt pebbles occurred and possibly that some of the debitage resulted from a bipolar reduction technique. Substantial differences in the degree of weathering present on the basalt was observed, often between individual items at the same site. Regrettably, field observations on lithic debitage were not sufficiently detailed to offer more than non-quantitative impressions. Further comments on how more systematic observations on debitage might be made during future reconnaissance efforts are offered in the final chapter of this study.

All lines of evidence from the reconnaissance are consistent with the expectation that these Subalpine lithic scatters are functionally related to hunting. Site locational patterns, the forms of the stone tools present (projectile points, cutting and scraping tools) and the types of tools that were not observed (e.g. heavy pounding tools, grinding implements) all support such an interpretation. Surprisingly, the only recognizable hearths encountered were those that appeared to have been used rather recently in the historic period. Although a few scattered fire-altered rocks were noted, contraction-cracked rocks were not observed. The absence of either substantial rock features or quantities of contraction-cracked rocks is consistent with the interpretation that root resources were of little relevance to the use of these localities.

Temporal indicators for these sites as well as those reported by Bergland (1984a) were also largely consistent with expectations. Projectile points were the only temporally diagnostic artifacts observed and all but one of these were lanceolate forms. The only exception was a small stemmed point with a haft element in the atlatl size-range (i.e. > 8 mm). On the basis of projectile points and a single radiocarbon date from Bergland's (1984a) reconnaissance, it would seem that usage of the Subalpine Zone appears to have been greatest during the early to mid-Holocene and greatly reduced during the past 2-3,000 yrs. In sum, the land use models developed in this study predicted the cessation of hunting in Zone IV after the emergence of collector systems in the river valleys. To date, no unambiguous archaeological evidence of Late Prehistoric usage has been identified in Zone IV although more than 60 sites have been recorded in this zone.

Archaeological Expectations

The archaeological potential within this zone varies from very high in the subalpine parkland and meadows to nonexistent in areas presently covered with glaciers or which were covered as recently as the 19th century. Extensive areas of this zone are bare rock and so steep as to be inaccessible for human usage. Most of the archaeological remains in this zone are expected to be within the modern parklands and meadows or at slightly lower elevations in areas that have not been densely forested throughout the Holocene. These are the same areas where Zone IV food resources are concentrated.

Differences in vegetation and fauna (especially deer and elk frequencies) between the windward and leeward sides of the Olympics may well be worthy of further subdivision of this zone but this is not done here because few differences could be deduced for the archaeological records of the two areas. The amount of land area in subalpine parkland and meadows is much greater in the area to the east of the Elwha whereas glaciers are largely confined to the west side of the Elwha (see Figure 8.1).

Probably the most important difference between the two sides, however, is the dominance of elk on the windward side and deer on the drier leeward side. To the extent that there are behavioral differences in these two animals that would have required different hunting strategies, it may eventually be interesting to consider how differences in the relative proportions of these two animals influence the archaeological record. If and when environmental variations between the two areas can be linked to expectations for archaeological patterning, these two areas might be distinguished as different management zones.

Archaeological expectations for this zone will be considered in the discussion that follows in terms of site locational characteristics, assemblage content, and site structure.

1) Exploitation of ungulates in this zone is expected to be mainly carried out in the context of land use systems of the forager type. The high costs of transporting any substantial amount of meat obtained in this zone to a residential base below the winter snow line would limit use of this zone primarily to immediate consumption strategies. Since the ungulate resources that occur here in summer can be taken most efficiently by a collector strategy after the animals move down into the river valleys in the winter, there is little advantage to hunting in this zone. A river valley and winter season focus of hunting for collectors would have the double advantage of providing fresh meat when it is most needed and when it can be taken with the least logistic effort. Use of this zone is expected to have been most intensive during the early Holocene (before 6,000 B.P.) when regional adaptations are postulated to have been highly mobile foragers who depended to a considerable degree on game even during the summer. It is also expected that there will be less intensive use of this zone after 6,000 B.P. and especially after the appearance of collector type land use systems in the late Holocene (after ca. 3,000 B.P.).

2) Site locational patterns are expected to conform to the patterns of ungulate movement and behavior. Since use of this zone is expected to be largely restricted to forager type land use systems, there are three kinds of sites expected: residential bases, stations, and locations. Stations are expected to be located where animals can be monitored and/or intercepted. Locations of high potential include low passes between river valleys, areas surrounding lakes, melt basins, or other water sources, and any terrain that offers panoramic views of expanses of relatively open ground. Sequestered spots in close proximity to water sources should be favored locations for stations because deer and elk do most of their foraging within a 1000 ft of water. Isolated finds or "locations" created by the loss of items while hunting or butchering of kills are expected to occur throughout the meadow/parkland but should occur in highest densities in the vicinity of hunting stations since most animals would be killed and processed near these locations. Due to cool temperatures even during the summer season, field camps or residential camps are likely to be located in areas with good solar exposure.

3) Because modern trails tend to follow topographic "paths of least resistance" through the very high relief of Zone IV, it is expected that there would be a rather high degree of spatial congruence in the ways past and present humans and ungulates move through this landscape. This implies that archaeological traces of trailside resting spots, kill and butchering locations, ambush locations, tool refurbishing locations, and residential sites have a high probability of occurring in or in proximity to the modern trails. Adding to the effects of high relief in this respect is the fact that fallen trees would cause far less long-term wandering of the trails than is expectable through Zones II and III.

4) Due to the focus of activity on hunting and upon immediate consumption in this zone, lithic assemblages should be very low diversity with projectile points the most frequently occurring formal tools. Flake cutting tools are the only other artifact expected in any frequency. Assuming that there are no suitable lithic sources in this zone, most formed tools would be brought here from lower zones where they were made. This should result in a scarcity of primary decortication flakes and the lithic debitage found in the Arctic-Subalpine zone should be limited mainly to that which would result from the maintenance of cutting edges, and sharpening or repair of projectile points.

5) Given negligible rates of sediment accumulation in this environment, there would be continuous and long-term opportunity for recycling of any usable lithic materials on very stable site surfaces. This should be manifest by the scarcity of large pieces of workable debitage or tools that could be reused. It would also be manifest occasionally by the occurrence of different degrees of weathering (patina) on flaked surfaces on the same object.

6) Intersite assemblage variability in lithics both within and between the three classes of sites is expected to be quite low due to the uniformity in the activities conducted. Some tool repair could occur at stations, locations, and residential bases but, on the whole, these site types are expected to be difficult to distinguish on the basis of assemblage variability.

7) The limited extent of sediment deposition in all but unusual situations is of overwhelming importance in conditioning the archaeological remains of this environmental zone. The recurrent execution of the same sorts of activities over restricted areas at locations during long time periods should result in the accumulation of lithic palimpsests. These palimpsests occurring at hunting locations may be difficult to distinguish from stations and residential bases except possibly in having less evidence for tool refurbishing. Residential sites will also tend to become palimpsests with the residues of many distinct occupations mixed together into a single scatter of lithic material more spatially extensive than the area occupied during any single occupation. On such sites, there may be horizontal stratification that could be distinguished by differences in weathering rinds on lithics (see Porter 1975) and/or clustering of lithics into discrete spatial concentrations. Investigators of these sites should anticipate the potential presence of horizontal stratification and use field methods that are appropriate for collecting such data.

8) Recognizable cooking and food processing features are expected to be infrequent, reflecting the lack of energy investment in facilities during brief visits by small hunting groups. Most cooking activities are expected to involve roasting of meat directly over an open fire and evidence of stone boiling is, therefore, not expected to be represented by quantities of fire-broken rocks with thermal contraction fracture patterns. Due to the high frequency of natural fires in this particular environmental zone, detection of thermally altered rocks resultant from human activities will require the capacity to distinguish these against the "background noise" of natural occurrences of such rocks.

9) Due to the substantial age of most archaeological deposits anticipated in this zone, the occurrence of preserved charcoal or other datable organics remains is expected to be rare. These occurrences are of great importance not only for dating purposes but for flotation analysis to identity any vegetal food resources present.

10) Lithic reduction byproducts should reflect two primary "situations"—resharpening and repair of tools that were carried into this zone as finished products and expedient production of tools from locally available raw materials. Because lithic tool usage is expected to be dominated by hunting-related activities in this zone, the production of cutting and scraping tools and the replacement of broken and lost projectile points should be the dominant debitage-generating activities. Lithic analyses that facilitate determinations of the role of individual sites in a settlement system are likely to be particularly informative.

Other Research Domains and Data Requirements

Development of Chronological Controls

It was noted in the discussion of the Old Cordilleran in Chapter 6 that chronological control has been a major problem pertaining to the archaeological record prior to 3,000 B.P. throughout western Washington. Given the usual scarcity of dateable organic materials in Old Cordilleran sites, development of alternatives to radiometric dating are of considerable importance. The Subalpine-Arctic zone may provide such alternatives. Development of a Holocene glacial chronology for the Olympics might provide a way that lower limiting dates could be placed on archaeological assemblages situated down-valley from modern glaciers (see Fonda 1974:928). Tephrachronology and especially the distribution of Mazama ash may also be valuable in this respect. Still another technique worthy of serious consideration involves relative dating based upon degrees of weathering rind development on chipped stone lithics (see Porter 1975). Earlier in this chapter, it was noted that the lower limiting dates on the four major river terraces of this region (Fonda 1974) offer a valuable means for age estimation of archaeological remains when these have been identified and investigated in alluvial settings.

Hypotheses Regarding Faunal Changes in the Past

Some biologists have hypothesized that migratory herds of Roosevelt elk developed as a result of post-contact phenomena—particularly settlement in the lowlands and logging in mountainous areas (Starkey et al. 1982; also Taber 1980). One implication of this hypothesis is that deer rather than elk should be the dominant ungulates represented in faunal collections from prehistoric archaeological sites in the Subalpine Zone throughout the Olympics.

Another hypothesis relating to faunal changes was presented in Chapter 5. It was argued that the warmer and drier climate and more open forest conditions of the early Holocene (prior to 6,000 B.P.) would have favored deer at the expense of Roosevelt elk. The expectations of this hypothesis are that the ratio of deer to elk should be relatively high in archaeological faunal collections from sites older than ca. 6,000 years and relatively low in collections more recent than this—especially in those subalpine areas west of the Elwha. Faunal data from archaeological sites spanning the entire Holocene would be extremely useful for addressing these kinds of problems. Although faunal preservation is likely to be quite poor in open sites within this zone, at least one open site has yielded traces of preserved mammal bones of mid-Holocene age (Seven Lakes Hearth Site, 45CA274; Bergland 1984a:45). Rockshelter sites in this zone that contain dry sediments and/or less acidic soils, also have potential for providing data relevant to this hypothesis.

A related question has been raised—"Were Mountain Goats indigenous to the Olympic Mountains?" The present mountain goat population developed from goats that were introduced to these mountains in the 1920s. In response to their burgeoning population and consequent impact on alpine ecosystems, the National Park Service recently has initiated an effort to remove mountain goats from Olympic National Park. This program is based on a determination that mountain goats were not native to the Olympics and, therefore, are incompatible with the objective of maintaining the Park in the most natural condition possible. Lyman (1988) suggests that these management policies regarding mountain goats may be misguided due to inadequate paleoenvironmental information. He maintains that there is insufficient information to assume that mountain goats were not a part of the indigenous fauna of the region throughout some or even most of the Holocene. Lyman also suggests that mountain goats may have survived throughout the Holocene up to as recently as the late 1700s or early 1800s before they were extirpated (Lyman 1988:21-22).

Noting the absence of mountain goat remains from archaeological sites around the Olympic Peninsula, Lyman questions the relevance of these data to the issue of whether or not mountain goats were indigenous to the Olympics Mountains. According to Lyman, the lack of reported mountain goat remains in archaeological faunal assemblages may simply be the result of the setting of most sites near the coast at some distance from mountain goat habitat, the small number of sites that have been excavated, the small scale of most of the excavations that have been performed, the small size of most recovered faunal samples, and the potential that mountain goat remains simply have not been recognized in faunal collections (Lyman 1988:14). It is Lyman's position that current Park management policies are faulty because they are founded on negative evidence.

To place this conclusion into perspective, some of the available data that Lyman did not make use of deserve consideration. Specifically, there are three sources of information that he does not adequately consider. These include, the accounts of early parties of explorers, the ethnographic records for the various tribes whose territories ring the Peninsula, and archaeological data that Lyman was not aware of.

1. Accounts of Early Expeditions into the Olympics. One source of information about the conjectured existence of mountain goats in the Olympics are the accounts of Euroamerican exploring parties. The first expeditions into the Olympics that resulted in well-documented written accounts did not occur until the late 1880s (Evans 1983:13). The O'Neill and Press expeditions occurred between 1885 and 1890 and left a written record of their experiences in the mountains (Wood 1967, 1976). These accounts make no mention of the observation of mountain goats despite some detailed information about elk, bear and other animals. In sum, first-hand accounts of extensive travels in the interior of the Olympic Mountains did not result in a single reported observation of mountain goats.

2. Ethnographic Evidence. Elmendorf (1960:135) states that both the mountain sheep and the mountain goat were not among the species native to the territory of the Twana. According to Olson (1967:15), Quinault informants also indicated an absence of both mountain goats and mountain sheep (Olson 1967:15) and Olson makes no mention of the use of mountain goat wool in weaving. Mountain goat hides, along with a number of other types of animal hides, were sewn into robes but these, like buffalo hides in the historic period, appear to have been obtained through trade (Olson 1967:57). On the northern margin of the Peninsula, the Klallam are said to have obtained small quantities of mountain goat wool as a minor trade item from the Cowichan on the mainland of British Columbia (Gunther 1927:212). It would appear then that there are independent and corroborative ethnographic accounts from the west, north and east sides of the Peninsula that mountain goats were not a part of the fauna of the Olympics. These ethnographic accounts refer specifically to the early and middle portions of the 19th century but probably also apply to as far back in time as oral traditions can reliably be extended.

Mountain goats were also missing from the native fauna reported historically from Vancouver Island but mountain goat wool robes were worn by high ranking Nootka males (Drucker 1951:9, 114). Mountain goat wool was obtained by the Nootka through their trading contacts with the Kwakiutl. Suttles (1951:95) reported that Straits Salish may have occasionally hunted mountain goats which "lived only in the higher mountains on the mainland, some distance from Straits territory." Mainland groups whose hunting territories extended into the Cascades such as the Snoqualmie, Skagit, and Skykomish hunted mountain goats (Haeberlin and Gunther 1930:25). It appears that articles of clothing (e.g. women's skirts, head bands, mantles, and bonnets) manufactured from mountain goat wool were highly valued markers of high rank throughout Puget Sound (Eells 1985:119-123). Eells (1985:203) suggests that mountain sheep horn dishes came from as far as the Stikine River in northern British Columbia.

To summarize, a substantial amount of ethnographic research conducted among several of the tribes of this region by some of the Northwest Coast's most venerated ethnographers came to the same conclusion regarding mountain goats. The ethnographic statements are explicit that mountain goats were not a part of the regional fauna in the Olympics and that mountain goat horn, hides, and wool were important wealth objects obtained through trade. In short, the ethnographic accounts go well beyond negative evidence or simple omission of this subject. The suggestion that mountain goats may have been present in the late 18th and early 19th centuries is quite difficult to reconcile with the ethnographic record.

3. Archaeological Evidence. With only minor exceptions, the faunal record from archaeological sites on the Olympic Peninsula conforms to the general pattern of western Washington. The acidity of the soil is such that animal bones do not typically survive long except in shell middens where the acidity is neutralized by the shells. In fact, the only reported occurrence of mountain goat remains from archaeological deposits on the Olympic Peninsula comes from shell middens—apparently in the vicinity of LaPush (Reagan 1917:16). In his description of the archaeological middens of the Quillayute Region, Albert Reagan (1917:16) comments as follows:

Bones of animals identified: Elk, big horn, mountain goat, black bear, Putorius, species?, black-tailed deer, wild cat, beaver, raccoon and otter.

In a footnote on the same page that refers to the listing of bighorn sheep and mountain goat, Reagan states that "The latter two are found usually only in the ladle form of the horns." In view of the trade in horn spoons reported ethnographically (Eells 1985:107, 231; Smith 1940:321-322), these occurrences can hardly be interpreted as evidence of the existence of either bighorn sheep or mountain goats in the Olympics. [5] In fact, the absence of habitat suitable to bighorn sheep in the Olympics clearly points to trade as the source of the remains of this species. The most likely explanation for the presence of these particular kinds of mountain goat and bighorn sheep remains in midden deposits anywhere on the Olympic Peninsula is that they resulted from trade with other tribes.

Although not from the Olympic Peninsula, there is another reported archaeological occurrence of mountain goat remains that deserves consideration here. King (1950:91) reports that "a few bones of the mountain goat and mountain sheep occurred in the Maritime strata" at the Cattle Point site on San Juan Island. It is noteworthy that each of these species appears to be represented by a single element (see King 1950:Table 12) but that there was apparently sufficient uncertainty about the accuracy of these identifications that a question mark was placed beside both of these species in a list of the mammals identified in order of descending frequency (see King 1950:91). If these bones were not mistakenly identified and are also not the result of some admixture of bones of domestic sheep and goats from the overlying historic stratum, then these are highly anomalous occurrences. The presence of either mountain goats or bighorn sheep on San Juan Island is inconsistent with the known habitat requirements for both of these species.

An important question emerges here regarding the use of archaeological faunal remains to make inferences about the past distribution of various species. What criteria can be applied in making the distinction between indigenous and non-indigenous specimens in an archaeological faunal assemblage? Clearly, the mere presence of the remains of a species in an archaeological deposit can not be assumed to demonstrate the existence of that species as a former participant in the local ecosystem. Archaeologists rarely make such an assumption when they find dentalia or olivella shells because in such instances they recognize the possible role of trade as a source for those items. In the case of larger animals such as mountain goat, however, it may be necessary to carefully consider the kinds of anatomical parts represented. In particular, it will be necessary to determine if the portions of the anatomy represented are likely to be those that travelled in a trade network or, on the other hand, if they include elements that indicate subsistence usages for immediate consumption. Skeletal elements represented should be scrutinized in the light of the degree to which those particular anatomical parts would be transported for subsistence purposes. Before archaeologists expand the former distribution of a species each time an element of that species is found in an archaeological site (e.g. Lyman 1986, 1988), it is necessary to recognize that archaeological sites are formed in ways quite distinctive to paleontological deposits. Forms of human behavior such as trade and transport play too important a role in cultural systems to be ignored when interpreting archaeological faunal remains.

To summarize, there is to date only one reported archaeological occurrence of mountain goat remains on the Olympic Peninsula. This find or finds is inadequately reported from an excavation(s) carried out over 70 years ago in the vicinity of LaPush. As far back in time as mountain goat horn, hides, and wool have served as trade items, the occurrence of associated anatomical parts of goat in archaeological sites does not demonstrate the existence of that species as a participant in the local ecosystem. If mountain goat remains are indeed present in the Cattle Point site faunal assemblage from San Juan Island, this occurrence strongly demonstrates the point that goat remains can occur in archaeological deposits far from suitable habitat for the species. It is predicted here that mountain goat remains will be found in very low frequencies in shell middens of the Olympic Peninsula whether or not there was a contemporaneous population of this species present in the Olympics. Assessing whether or not goats were present in the Olympics on the basis of evidence from coastal shell middens will minimally require a careful consideration of economic anatomy of goats relative to skeletal elements represented. If only those parts such as horns or skeletal elements that tend to travel as "riders" on a hide are found, it will be difficult to make a case for a local population of mountain goats in the Olympics.

A critical test for determining the presence or absence of mountain goats during the Holocene in the Olympic Mountains would be to recover a sizable faunal collection from a cave or rockshelter high in the Subalpine Zone. A dry deposit with good faunal preservation and potential for dating the recovered fauna would be ideal. Such a deposit need not contain cultural remains to be useful relative to this question. Although no such deposit is known to exist at this time, Appendix C is an attempt at listing some locations within the Park that are worthy of inspection for their potential to yield this kind of faunal evidence. Acknowledging the difficulty of finding deposits with preserved faunal remains, consideration might be given to the analysis of blood residues on stone tools (e.g. Gurfinkel and Franklin 1988).

It should be emphasized that there presently is minimal direct information on the early and middle Holocene mammalian fauna of this region. Although mastodon, bison and caribou are known to have been present on the Peninsula in the late Pleistocene (Gustafson et al. 1979), it is not known when these species went extinct. One could as easily argue for the existence of caribou in the Olympics as the mountain goat and the presence of caribou would probably have much greater significance for prehistoric human subsistence systems.

From the perspective of mammalian biogeography, the Olympic Mountains can be viewed as a "land-bridge island". Inasmuch as all national parks and natural areas are increasingly becoming "land-bridge islands" due to the progressive human modifications of intervening landscapes (Newmark 1987), documentation of the long-term faunal changes that have occurred in the Olympics might have broader scientific implications that go beyond archaeology to issues of wildlife management.

Hypotheses Concerning the Sources of Lithic Raw Materials

It has been suggested that the basalt raw materials which dominate the lithic assemblages described in the mountains of the Park are non-local in origin or that they come from the coastal margin of the Peninsula (Tabor 1984; Bergland 1984a:24). The limited archaeological reconnaissance that was done in conjunction with the present project produced results that seem inconsistent with the non-local origin hypothesis (Appendix A). There was cortex on a significant proportion of the flaking debris on sites in the mountains of the eastern Olympics and it appears that primary reduction of small cobbles and nodules of basalt took place on these sites. It seems unlikely that pedestrian hunter-gatherers would carry unworked basalt cobbles high into the mountains for primary reduction.

Another insight gained from the recent reconnaissance was the finding of several unworked basalt nodules in the interior of the Park that appear to be indistinguishable from the materials found on archaeological sites elsewhere in the Subalpine zone of the Park. These nodules were found on the Long Ridge trail (Elwha Valley, see Appendix A) where it descends a steep slope. Interpretation of the location as a natural occurrence of this type of basalt is complicated by the finding of a single lanceolate point in the trail nearby. No chipping debris was noted at this location that would permit interpretation of the point as the product of on-the-spot tool manufacture at a source lithic raw material source. However, this occurrence again raises the question of whether the basalt might actually occur within or at least near the Park.

There are at least two possibilities to account for the natural occurrence of a type of basalt in the Park that has been interpreted as non-local in origin. The first is that there may be dikes of intrusive basalt of a different variety than that which occurs throughout the northern and eastern sections of the Park. Butler (1961:) suggested that the type of basalt that occurs in Old Cordilleran lithic assemblages from the Puget Sound region came from intrusive basalt dikes. We presently have no knowledge of the occurrence of such dikes on the Olympic Peninsula but the possibility that they may exist must be considered.

The second possibility is that there may be natural occurrences of basalt within or adjacent to the Park that have been glacially rafted from a source location at some distance from the Peninsula. It has been suggested that glacial transport of lithic raw materials in areas formerly glaciated should be considered as an alternative to trade from exotic sources (Meltzer 1984). This mechanism is consistent with information suggesting that the glacial lobe that extended westward through what is now the Strait of Juan de Fuca also extended a considerable distance up the Elwha Valley. It is also consistent with the finding of the basalt nodules along the Long Ridge Trail in the Long Creek valley, a tributary to the Elwha.

There are, then, three hypotheses that have been identified regarding the source of the basalt tool stone that occurs in the Olympic mountains. These hypotheses and deductions from them may be listed:

Hypothesis 1: This basalt is exotic and was traded from outside the region. The implications of this hypothesis are:

a) Early stages of lithic reduction should be poorly represented archaeologically in the Park and on the Peninsula. Lithics will be dominated by the byproducts of the final stages of tool manufacture or tool maintenance and repair.

b) Lithic reduction techniques (e.g. the bipolar technique) that maximize valuable tool stone should be strongly represented and those techniques that do not make efficient use of raw material should be very poorly represented.

c) Natural deposits of similar basalt will not be found within the Park on the Peninsula.

Hypothesis 2: The basalt occurs in dikes of intrusive basalt that are within the Park or at least on the Olympic Peninsula. If so, implications are that:

a) True quarry sites will be found in this region that are as yet unknown to archaeologists or geologists.

b) Lithic reduction techniques will reflect the local availability of raw material through the occurrence of less economical lithic reduction techniques (e.g. biface reduction). However, as distance from the source(s) increases, less wasteful techniques may be of greater importance.

c) Procurement from a quarry source implies that early reduction stages should typically lack evidence of cobble cortex.

Hypothesis 3: This material was glacially rafted from outside the region and deposited as "drift" within the Park.

a) Source areas will be found either along the northern or eastern boundaries of the Park (or the Peninsula) that were penetrated by Pleistocene ice lobes. Those lobes moved southward up the Elwha from the Strait of Juan de Fuca and westward up the valleys of the Hood Canal drainages so these areas are where glacially rafted materials should occur.

b) Early stages of lithic reduction should occur on sites located in proximity to the sources of drift basalt and decline in frequency with distance from those sources. A major contrast should exist between the northeast and southwest half of the Park. This contrast should involve fewer early lithic reduction byproducts in the southwestern section and possibly greater dependence upon other lithic raw materials.

c) Lithic reduction byproducts should include a significant proportion of primary cobble cortex as would be expected in rafted materials.

Shellfish Studies

The models that have been developed in this study suggest that the role of shellfish in prehistoric subsistence systems of this region changed through time. Simply stated, three different roles for shellfish are postulated in the changing land use systems of this region through the Holocene. For the early Holocene and possibly well into the mid-Holocene, shellfish seem to have played little if any role in subsistence systems of Western Washington. The initial systematic use of shellfish is postulated to have occurred in the context of late winter/early spring nutritional stresses related to diets high in lean meat. Shellfish usage would have been limited to in-season/immediate consumption. In an ecosystem with such limited carbohydrate sources as the Olympic Peninsula, shellfish can take on a dietary role out of proportion to their subsistence value in any other cost/benefit sense. This emergent use of shellfish is viewed as a result of a population/resource imbalance caused either by population increase, reduction in terrestrial resource productivity at about 6,000 yrs. B.P., or a combination of the two. After the appearance of winter sedentary systems of land use of the collector type, the role of shellfish use is postulated to have changed again. With winter sedentism, the potential for storage of dryed shellfish developed. Along with various oil-rich marine resources, dryed shellfish would have provided an additional means for protein-sparing during late winter/early spring (Speth and Spielmann 1983). The potential for storage of shellfish that developed as an effect of winter sedentism based upon major food staples (e.g. salmon, halibut), resulted in a substantial expansion in the season of shellfish exploitation and probably the species of shellfish exploited as well. The general processes associated with the foraging/collecting transition are likely to be pivotal in explaining recurrent patterns in the succession of shellfish species in shellmiddens along the Pacific Coast.

Few archaeologists have been willing to accept the relatively recent ages of shell middens on the Southern Northwest Coast at face value and, consequently, the demographic view of shellfish as marginal return resources (e.g. Osborn 1977) has not gained widespread acceptance. The tendency has been to explain away the absence of early and mid-Holocene shellmiddens as a result of fluctuating sea levels, deterioration of the shellfish remains, the lack of archaeological survey of the appropriate landforms (e.g. raised beach terraces), or the absence of well-developed shellfish habitats prior to the mid-Holocene. In contrast, the demographic model for shellfish intensification best accounts for the available archaeological data.

Surrounding the whole question of shellfish use are fundamental issues about the productivity, procurement costs, seasonal changes in tidal amplitude and nutritional content, depth distribution in the intertidal zone, replacement rates, and other variables relative to these resources. Many facets of this entire subject offer important directions for future research too numerous to address in depth here.

One point about the archaeological investigation of aboriginal shellfish usage deserves particular emphasis. When viewed from an ecological perspective, many ideas about the nature of shellfish usage in the past tend to be overly normative. One example mentioned earlier would be the notion that shell middens along the entire Northwest Coast can be linked to the appearance of winter sedentary systems of land use much like those of the ethnographic horizon (e.g. Fladmark 1975; Alkens et al. 1986). Although most shell midden deposits may have been created by such systems, there is reason to believe that the circumstances under which such deposits are created are variable and not necessarily associated exclusively with any particular type of land use system. Another example here is the suggestion that shellfish exploitation by hunter-gatherers throughout the temperate zone should be approached from the perspective that the primary role of these resources was to satisfy protein requirements (Erlandson 1988:106-107). Although Southern California and the Northwest Coast both are temperate environments, the structure of energy and the relative abundance of carbohydrate versus protein food resources in these two settings are markedly different. For the aboriginal subsistence systems of the Northwest Coast's moist, densely forested areas, where carbohydrates are relatively scarce, the dietary role of shellfish is likely to have been quite different from that in the more arid regions of the temperate zone. I am suggesting that shellfish use in the Olympic Peninsula region and throughout most of the Northwest Coast can only be understood as a protein sparing strategy.

Analyses of seasonality using shellfish growth rings or other techniques are of considerable significance for gaining a basic understanding of the prehistoric role of shellfish resources in subsistence. It is important that such analyses consider the great variability that exists in the ecology of individual species before the results of analyses of one species are extrapolated to all shellfish. To date, such analyses for the Olympic Peninsula region have focused mainly upon the native littleneck clam (Wessen 1982). This particular clam is distinctive from other hardshell clams in being shallowly buried in the substrate but also in its high position in the intertidal zone. Other factors being equal, these characteristics would favor a wider range of seasonal usage for this species relative to other major species of clams. In general, determining the seasonal patterning of shellfish usage for the major species represented in archaeological deposits of different ages and function will be a necessary first step in evaluating basic models of prehistoric shellfish usage. To the extent that shellfish were dryed for delayed consumption at certain times and places, inferring the nature of shellfish usage from archaeological remains will have to consider the fact that there is not necessarily a simple relationship between season of harvest and season of consumption. [6]

Other Research Topics

A number of other research topics of relevance to Olympic National Park have been identified by Bergland (1984a:Table 6). To these, the following questions can be added:

1) What were the effects of the the massive population reductions from Old World infectious diseases on regional subsistence systems?

2) What are the archaeological correlates of a possible southward expansion of Nootkans onto the Olympic Peninsula from Vancouver Island? What relationship might such an expansion have to the development of maritime collecting systems of land use on the Olympic Peninsula?

3) What is the significance of the virtual disappearance of chipped stone artifacts in the last 1-2,000 years of the prehistoric record of this region? Is this disappearance associated mainly with maritime collecting systems of land use or does it also occur in areas where riverine collecting systems persisted into the historic period?

4) Is there a direct relationship between the development of maritime collecting systems and the appearance of archaeological evidence for warfare (e.g. fortifications)?

Notes

¹When Lieutenant O'Neills' expedition was traveling through the valleys of the Olympics in 1890, the men frequently came upon what O'Neill (Wood 1976:38) referred to as "elk yards, the winter home of elk." These yards sometimes covered hundreds of acres and were "always found on the southern slope of a ridge or mountain," but were so hemmed in by forest and terrain as to be "protected on all sides (ibid)." The description given by O'Neill for the elk winter yarding areas conforms closely with a very broadly observed pattern in ungulate distributions in mountainous areas (Schalk and Mierendorf 1984:24-34). These kinds of settings are likely to be particularly prime locations for early and mid-Holocene archaeological remains on the Olympic Peninsula and throughout the Northwest cordillera.

²On a very generalized map of the northwestern Olympic Peninsula, Reagan (1917:22) shows approximate locations of two midden sites—one along the Soleduck River to the south of Lake Pleasant and the other on the Calawah River roughly three miles west of Forks.

³Wessen (1982:57) states that 50 localities were surveyed. Only 45 such localities are actually identified in his report (ibid:Appendix).

⁴Daugherty et al. (1983) state that the mountain goat "could be obtained in the nearby Olympic Mountains (1982:3)". Although this statement seems to suggest that mountain goats were an indigenous species in the Olympic Mountains, these authors offer no supportive citation so that it is unclear what the basis for such a conclusion might be.

⁵Other species that are identified by Reagan as being present in archaeological middens of the Quillayute Region cast some doubt on the reliability of his faunal identifications. He reports the presence of bones from a list of "fish" that includes squid and octopus as well as five species of Pacific salmon (Reagan 1917:15). Although the presence of salmon remains in these deposits seems more plausible than those of squid, it is extremely unlikely that species level distinctions were made from the faunal remains. In other words, Reagan does not make a clear distinction between species actually identified from midden collections and those thought to be present on the basis of independent information.

⁶Wessen (1982) suggests that between 10 and 20% of the diet of the occupants of the Ozette site was obtained from shellfish and that shellfish were exploited throughout the year. This view runs counter to that which maintains that shellfish were widely used as marginal return resources during seasons of resource stress (Osborn 1977). Because of the potential seasonal disparity between collection and consumption, archaeological evidence (i.e. growth rings) for summer season harvesting does not in itself disprove this second view.