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

There are two fundamentally different ways of looking at the archaeological record of the last 3,000 years in northwestern Washington. One view is that the basic elements of settlement and subsistence were already present much earlier and that cultural change during this interval is largely elaboration of existing patterns (Fladmark 1975; 1979; Aikens et al. 1986). The second view is that basic systemic changes occurred during this interval and, in particular, the transition from foraging to collecting systems took place. The latter viewpoint underlies the discussions of this chapter.

Proponents of the first view interpret the appearance of shellmiddens as heralding the rise of winter villages. Since shellmiddens appear in a number of areas of the Northwest Coast between 4000 and 5000 years ago, it has been argued that winter villages emerged at about this time. There are two problems with this approach that deserve consideration. The first is that shellmiddens are questionable as indicators of winter sedentism. The fact that shellfish can be an important resource to hunter-gatherers practicing forager type land use strategies is amply demonstrated among recent hunter-gatherers around the world. The Yahgan and Ona of Tierra del Fuego represent but two examples of groups who made use of shellfish without practicing a collector strategy (Stuart 1972). Similar examples can be found from coastal Australia. There is no necessary relationship between shellmiddens and a particular form of settlement system.

The second point is that shellmiddens do not appear all along the Northwest Coast at 4-5,000 years B.P. In the state of Washington, nearly all radiocarbon dates on shellmiddens have yielded age estimates less than 2,000 years B.P. A date of ca. 2700 B.P. from the Hoko River site comes from a deposit containing limited numbers of mussel shells and this apparently represents the earliest claim for a shell midden in western Washington. If there are 5,000 year old shellmiddens in this area, the first has yet to be reported.

If the appearance of shell middens can not be equated with winter villages and the collector system, then what features of the archaeological record are diagnostic of such adaptive changes? Some features of potential relevance in this regard include permanent house remains, evidence for food storage, mass harvest technology, and diversification of settlement types.

Since heavy investments in facilities are generally taken to be characteristic of collectors, the appearance of plank houses would seem to be an important indicator of winter sedentary land use systems. The oldest examples of house remains, however, in northwestern Washington are less than 1,000 years old. Immediately to the north in the Gulf of Georgia region, plank house remains extend back to about 2000 B.P. (Marpole Phase; Mitchell 1971:52). At the St Mungo site on the Fraser River delta, dates of 4480 B.P. have been obtained from a post frame house floor; a series of younger floors from the site date between 3300 to 4300 B.P. (Ham et al. 1984). Interestingly, however, the seasons of occupation at this site seem to be spring and late summer rather than winter season. An equally intriguing fact is that there are as yet no known house remains associated with winter village settlement types for the interval 2000 to 3000 B.P. ("Locarno Beach Culture Type") in the Fraser River/Gulf of Georgia region (Stiefel 1980:188). The oldest house remains from the Olympic Peninsula are those from the Ozette site dating to the last four centuries (Huelsbeck 1983a). The picture which emerges regarding the appearance of houses is that the earliest houses that have been dated are not associated with winter villages but in spring and summer season fishing camps. In other words, the architectural remains that would provide the most direct evidence for winter sedentism have not yet been reported.

Other indicators of winter sedentism can be identified and these would include archaeological evidence for significant dependence upon food storage. Evidence for food storage generally has less archaeological visibility in this region than in some areas due to infrequent use of below ground pits. High precipitation, high humidity, and mild winter temperatures make below ground storage difficult. Historically, above ground storage of smoked/dried foods was the standard practice. Indirect evidence for food storage comes from stake moulds located on the beach of the St Mungo site that are believed to be fish drying racks and these have been dated to 3280 to 3485 B.P. (Ham et al. 1984:147).

The abundance of ground slate knives at fishing stations has been interpreted as evidence for efficient processing of fish for storage (Mitchell 1971:52-3). Chipped stone knives, hafted microliths, and basket fragments from 2200 to 2700 year old deposits at the Hoko River site have been interpreted in similar fashion by Croes and Blinman (1980:315). The most obvious artifacts associated with mass harvest techniques are those related to net fishing. There are pieces of preserved nets that have been found at the Hoko Site dated between 2200 and 2700 B.P. (Croes and Blinman 1980:319).

Possibly the evidence which would provide the most decisive case for the emergence of the collector system are settlement data. One of the basic premises of the forager-collector model is that a collector strategy involves more diversified settlement types with a higher degree of intersite assemblage variability. As mentioned earlier, there have been few large scale archaeological surveys west of the Cascade Range. There has really only been one study that approaches this subject and it compared assemblage data from a number of published northern Puget Sound site reports with assemblage data from a site survey on the Skagit River delta (Thompson 1978). The conclusion of this study was that specialized site types appeared only after ca 1500 B.P. and that before that time there was but one generalized site type. Interpreted in terms of the forager-collector model, these data indicate a rather late appearance of the collector system.

Modeling the Emergence of Winter-Sedentary Land Use Systems

The model of Old Cordilleran land use that has been proposed—maintains that the systems of the early Holocene were forager-like in their organization. Ethnographic land use systems that were discussed in Chapter 4, were all of the collector type system although considerable variation within this type was suggested for the Olympic Peninsula region. The important questions that arise at this juncture then are these:

1) What processes led to the widespread development of collector system types throughout the Olympic Peninsula region?

2) What processes resulted in the variations in ethnographic land use systems within the Olympic Peninsula region?

The answers proposed for these two questions have already been foreshadowed in previous portions of the report and will be discussed in order.

It is postulated that cultural change in the past 3,000 years must have been largely driven by factors other than environmental change. Undoubtedly low order climatic changes have occurred during this interval and may have had at least minor impacts on some food resources, but the magnitude of these changes was definitely much less than those of the early Holocene. In discussing Old Cordilleran land use, it was argued that the late substage (6,000-3,000 B.P.) of this interval involved a trend toward increased use of fish and marine resources during the spring and summer seasons. This trend, motivated by decreasing terrestrial game and plant resources after 6,000 B.P. and possibly slow population increase, was accomodated by non-systemic adjustments such as reductions in home range sizes and residential mobility. But these accomodations would have been accomplished by what remained basicly foraging systems.

Continued population growth would have placed increasing strain on those resources relied upon during the winter season when few fresh foods were available. During the late stage of the Old Cordilleran, substitution of some fish and marine resources for game during the spring through fall months would have reduced pressure on the terrestrial game resources critical to winter survival. However, this response would only offer a temporary solution to an ever increasing demand for game resources resulting from still further population growth. The carrying capacity of large ungulates would establish the human population ceiling through the operation of the winter season "bottleneck". Given the relatively low carrying capacity of this region for large game, sizable population fluctuations that are typical of most ungulates, and relatively low reproductive rates in the elk of this region [1], the population density threshold at which cultural stresses came into play would not be very high. These stresses on the human populations would require alternative strategies to immediate consumption—i.e. food storage.

There are few feasible alternatives in this region to terrestrial game for winter season provisioning a human population by means of immediate consumption. Most of the more productive marine resources are only seasonally available and tend to be restricted in abundance or entirely unavailable during the winter months. Added to this are the accessibility problems associated with exploitation of marine resources during the winter. The weather can be stormy for days on end making travel by watercraft difficult or impossible. Many shellfish are only available at low tide and these occur for only a few days out of each month. Tides during the winter months are not as low as during the spring and early summer and they occur at night time. Added to these potential obstacles to dependable subsistence, is the fact that storms can override astronomical variables by keeping the tide from going out even when it is supposed to. In general, the structure of marine resources militates against their intensive use during the winter months in systems of immediate return.

Storage of spatially concentrated and seasonally available resources is the inevitable outcome of population increase in this environment. While many resources can be preserved for delayed consumption, most are not suitable as storage staples in the sense that they can not be mass harvested. Salmonid resources are generally the best suited class of resources for storage in this region and the importance of these fish in systems of delayed consumption throughout extensive areas of North America is well known (Kroeber 1939).

It is postulated here that the initial collector systems were riverine rather than coastal in their regional settlement pattern and that intensive use of marine resources for delayed consumption (e.g. Makah use of halibut) was initiated in the past 1000-1500 yrs or less. In addition, it is proposed that the initial shift from winter hunting in a foraging strategy to delayed consumption based upon storage of salmon occurred along the major regional drainages or major salmon migration routes and not along the smaller streams or the saltwater coasts. This proposition may be argued on both theoretical and empirical grounds.

The environmental basis of this argument is that the larger rivers support not only larger, more diverse, and more reliable populations of anadromous fish but importantly they offer fish runs for longer portions of the year (Schalk 1977). This point was made in Chapter 3 with respect to differences between larger rivers such as the Quinault and the smaller rivers such as the Hamma Hamma or even the Sooes. The additional months of availability of salmon in the larger rivers would provide two important advantages for populations making the transition from primary reliance upon winter hunting to primary reliance upon storage for delayed consumption. The first advantage is that these locations would provide the longest period during which storage could be effectively carried out. The second is that the lengthy season of availability of salmonids in these rivers would reduce the portion of the year in which delayed consumption was a necessity. Both of these characteristics would make the transition from foraging to collecting an easier one.

There is yet another reason for the expectation that the initial occurrence of winter sedentary land use systems would have a riverine rather than a marine focus in this environment. In Chapter 2, arguments were presented to the effect that the transition from a system of immediate consumption to one of delayed consumption is, in an organizational sense, an abrupt one. Although such a change would inevitably be rather profound and there is apparently no real middle ground between foraging and collecting, there are differences in the degree of dependence upon delayed consumption and the extent of reliance upon logistic mobility as a means for matching resources to consumers. The evolutionary pathway from foraging to collecting that offers the least discontinuity is that in which immediate consumption would continue to play a major role. In other words, for hunter-gatherers whose primary over-wintering strategy involved the hunting of ungulates, initial sedentism is likely to have occurred in those places where winter hunting offered an important supplement to delayed consumption during the period of dependence upon stored foods. A similar argument has been proposed to account for the geographic distribution of the earliest archaeological evidence for winter-sedentism on the Columbia Plateau (Schalk 1987).

A riverine rather than a coastal origin for the initial appearance of winter villages would have permitted an ongoing and important role for the winter hunting of large ungulates. In the Olympic Peninsula region, ungulate resources would be most concentrated in three kinds of settings 1) natural or cultural prairies, 2) rainshadow areas or areas of lesser precipitation where secondary biomass is higher due to more open vegetation conditions, 3) areas flanking mountains where the animals which forage over broad areas in the summer time are highly concentrated onto winter yarding areas. It is noteworthy that important elk winter range areas on the Olympic Peninsula are generally located at some distance away from saltwater (see Figure 3.3) and this pattern probably obtained throughout the Holocene. Along these same lines, Olson's (1936:42) ethnographic informants observed that "elk rarely frequented the ocean shore." The saltwater coastline is generally not well situated relative to the regional distribution of ungulates and typically does not offer good intercept potential for anadromous fish. [2] The river valleys, on the other hand, would have provided the locations where positioning with respect to both ungulates and salmon could be optimized.

Empirical evidence for an initial riverine appearance of winter villages along the major river drainages of Puget Sound Region is mostly indirect. Given the disproportionate amount of archaeological investigation that has occurred in saltwater shellmiddens, the most significant observation seems to hinge on what is missing from these sites. In western Washington, there is a remarkable scarcity of house remains older than a millenium from archaeological sites situated on saltwater. The occurrence of houses at Ozette appears, for example, to be associated with the much more recent appearance of full-blown maritime collector systems. Similarly, the house remains from Old Man House on the Port Madison Reservation were historic in age (Snyder 1956). Despite the fact that shellmidden sites have been more extensively investigated than any other type of site in western Washington, and have provided most of the radiocarbon dates, most do not appear to contain evidence of being winter villages.

The majority of saltwater shell middens appear to be seasonally occupied field camps where subsistence activities focused on marine resources (e.g. marine fish, sea mammals, or molluscs). Most of the shellmidden age estimates from western Washington fall within the last 2000 years or roughly a millenium after the initial appearance of mass harvest technology and diversified artifact assemblages (Mitchell 1971; Croes and Blinman 1980; Croes and Hackenburger 1984). In areas well endowed with salmon resources, most saltwater shellmiddens are likely to be seasonal resource procurement locations associated with riverine-based settlement systems. This is the expectable archaeological pattern if regional populations first established riverine villages roughly 3000 years ago, increasingly employed logistic procurement strategies thereafter, and intensified subsistence as a result of population growth. In those areas such as the northern Olympic Peninsula and especially the northwestern Peninsula where marine resources were unusually productive, fully maritime collector systems of land use emerged. Most of the shell middens associated with these maritime systems would continue to be seasonal resource procurement sites (i.e. "field camps" and "locations") but a few would be places that were occupied on a more permanent basis and, in particular throughout the winter (i.e. "villages"). The point to be emphasized is that shell middens of at least superficially similar physical character might be associated with rather different systems of land use. Furthermore, the existence of shell middens need not imply any particular level of dependence upon marine resources.

Following these lines of argument, rivers such as the Quinault, Queets, Hoh, Elwha, and Skokomish are the areas expected to have seen the earliest collector type land use systems on the Olympic Peninsula. These are the rivers with the largest and most diverse runs of anadromous fish. They are also rivers which have their headwaters in the mountains and have extensive areas of winter range for ungulates below the winter snowline (see Figure 3.3). In addition, most if not all had prairies that would have provided game and plant resources to supplement the riparian resources. Although the age of the historically recorded prairies has not been the subject of any systematic research (Greengo 1983:83), arguments presented in previous chapters support the conclusion that the historic distribution of prairies on the Olympic Peninsula probably represents a reasonable approximation of their distribution during the past several millenia. However, forest encroachment undoubtedly claimed some areas that were formerly prairies (Smith 1951:53).

Diversification and Intensification of the Collector Systems

In the preceding section, a model has been proposed for the emergence of riverine land use systems. In some areas of the Olympic Peninsula and western Washington, such systems persisted into historic times. The Quinault of the early and mid-19th century represent this type of land use system. To account for the variability known historically on the Olympic Peninsula, however, it is also necessary to explain the appearance of fully maritime land use strategies as well.

Accelerated population growth is generally considered to be one of the consequences of settling down (Binford and Chasko 1976; Harris 1978). With a more sedentary settlement system established in the river valleys, the developmental sequence thereafter is largely expected to be a process of diversification and subsistence intensification within the basic collector system. There would be a general growth in the degree of dependence upon delayed consumption, probably manifested in the form of greater quantities of particular resources being stored as well as a wider range of resources stored. In addition to this elaboration in the temporal control of energy flow, another way in which a collector system can intensify subsistence is through more logistically organized resource procurement. This would be reflected in the extension of the "catchment area" from which resources are returned to central locations such as winter villages and, to cope with rising transport costs, a greater degree of resource processing at special purpose resource procurement sites.

A growing regional population would be expected to progressively establish winter villages in less optimal settings and make increasingly greater use of marine resources—especially those that are more costly to procure (e.g. sea mammals, shellfish). Because many of the smaller rivers are neither productive or reliable enough to provide adequate supplies of salmon for winter survival by themselves, certain adjustments would be required. One adjustment expected here would be a more saltwater-oriented pattern of exploiting salmon in which numerous smaller and more dispersed salmon resources are exploited at or near river mouths out of villages located on saltwater. Another adjustment might be trolling for salmon on saltwater, a technique that is effective where there are salmon migration paths and feeding areas. Local groups based at saltwater villages and relying upon a more logistically organized settlement system (made possible by water transport) could exploit multiple small streams along a stretch of coastline, none of which was itself capable of providing the entire annual demand for salmon for a productive unit (household group). This, of course, is the pattern described ethnographically for the Makah and some of the Klallam and Twana (see Chapter 4).

Perhaps the ultimate divergence from the riverine collector system comes with the substitution of other resources for salmon as the primary storage staple. The exploitation of halibut by the Makah represents this final step. Combined with intensive use of marine mammals, this strategy permitted occupation of a region of the Peninsula that would probably not have supported a truly riverine economy with salmon as the main resource used for winter survival.

Because the logistic potentials of locating winter villages in riverine versus coastal settings have not been fully appreciated in the archaeological and ethnographic literature for the Northwest Coast, it is instructive to consider how different these potentials are. Following Binford's (1982; see also Thomas 1983:Figure 11), terminology for the economic zonation surrounding a residential base, there are concentratic zones established by the foraging radius and the logistic radius. For pedestrian hunter-gatherers like the !Kung Bushman or the Great Basin Paiute, the foraging radius is the distance to which people engaged in food procurement can walk to and from in a day-trip. This radius is generally estimated at about 10 km and it establishes a roughly circular foraging zone with the residential base as its focal point. (The shape of this zone will actually vary substantially from a circular pattern due to terrain features that make travel costs in different directions non-uniform). Surrounding the foraging zone is the logistic zone—the area beyond the foraging radius which is exploited by task groups involved in food procurement activities taking more than a single day. Exploitation of the logistic zone requires the use of field camps. In other words, this economic zone lies far enough from the residential base that overnight stays are necessary for its exploitation.

Ethnographic land use systems of the Northwest Coast require a somewhat more complex model of economic zonation than that applied to strictly pedestrian hunter-gatherers. The character of economic zonation surrounding a winter village in this region requires considerations not addressed in the generalized inland hunter-gatherer model just described. The difference derives largely from the additional logistic potential offered by water transport and the effects of a site positioning strategy at the interface of marine and terrestrial ecosystems. In Figure 7.1, economic zones surrounding riverine and coastal are compared. A residential base located on a river approximates the generalized inland hunter-gatherer logistic zonation model. The use of watercraft on the river probably results in at least minor increases in the size of the foraging and logistic zones adjacent to the river channel. Given the swift currents characteristic of Peninsula streams and the single-directional nature of stream flow, the increment in logistic efficiency from watercraft in this region would not be impressive. The utility of watercraft travel on a river undoubtedly is greater where the currents are weakest and this will generally mean that watercraft transport is most efficient along the lower reaches of rivers.

Figure 7.1 Economic zonation surrounding a riverine village compared to that for a Coastal Village.

With the increased mobility and load-carrying capacity of watercraft on the ocean, economic zonation surrounding a residential base or winter village is qualitatively different from that associated with riverine settlements. Figure 7.1 illustrates the nature of these differences. Positioning a winter village on saltwater reduces the area of land that is within the foraging radius (up to half the foraging zone on a straight coastline). Greatly offsetting this potential loss of access to terrestrial habitat is a large increase in the area of marine habitat that falls within the daily foraging radius. The foraging radius for watercraft used on saltwater is considerably larger than that for pedestrian hunter-gatherers and, judging from ethnographic accounts of distances travelled out to sea (see Chapter 4), may have been as much as 20 km or more. Croes and Hackenberger (1988:Figure 7) suggest that the daily travel distance over water is more than double that over land. The fact that marine resources tend to be highly "clumped" in their spatial distributions is a factor that would favor greater use of logistic organization for hunter-gathers with greater reliance upon marine resources (Schalk 1981). This is still another reason for expecting that the logistic zone will be much larger for coastal villages than for riverine villages.

To the extent that round-trip resource procurement forays on the ocean are accomplished within a day, the foraging radius and the logistic radius would encompass identical areas. However, there are three conditions under which forays in watercraft would not have been limited to an area within one day's round trip travel of a residential base. One such circumstance occurs when field camps are established on offshore islands outside the foraging radius. A second situation that would contribute to an expansion of the logistic zone beyond the foraging zone on saltwater would be when overnight trips are conducted without a landfall. In effect, sleeping in a canoe amounts to using a canoe as a sort of portable field camp. Some of the longer forays mentioned in the ethnographies (e.g. up to 50 mi at sea, Pettitt 1950:8) must have involved "camping" in the canoes. The third situation would be those cases when a logistic trip involved paralleling the shoreline and establishment of a field camp on the mainland. In these examples, a logistic zone may substantially exceed the foraging zone when watercraft are used on saltwater.

It may be concluded that the economic zonation surrounding coastal residential bases is significantly different from that around riverine residential bases. The overall size of both the foraging and logistic zones is greatly increased due largely to the increased travel and transport efficiency on the ocean. Coastal residential bases imply a reduction in the amount of terrestrial habitat within the foraging and logistic zones, but for adaptations making intensive use of marine resources this reduction is of little consequence. The increased access to marine resources within the foraging and logistical zones of fully maritime systems of land use would easily offset any disadvantage of reduced access to terrestrial resources within these concentric zones.

The preceding discussions provide the background for postulating the developmental sequence for regional land use systems. In Figure 7.2 a schematic model for the evolution and diversification of prehistoric land use systems of the Olympic Peninsula is depicted. Old Cordilleran foraging systems are represented as being initially replaced by riverine collecting systems and these systems would have developed first on the larger, more productive drainages of the Olympic Peninsula. Subsequently, areas that were more marginal in terms of salmon resources were occupied but this radiation required a different kind of collector system. Effective use of these salmon-poor areas would have required a saltwater rather than a riverine winter village settlement pattern. By locating on saltwater, villages could serve as the bases for the more extensive logistic networks necessary to exploit smaller, more dispersed salmon resource loci and also to exploit marine mammals and fish. The ethnographic Quinault, Queets and Skokomish closely approximate the riverine collector model and it is argued that these land use systems will prove to be older than those associated with the fully maritime groups of the Olympic Peninsula. The Makah and the Klallam are good examples of full-blown maritime collector systems. These maritime collector systems dominate historically along the coast northward from the Olympic Peninsula and are represented by the Nootka, Kwakiutl, Haida, and Tlingit of the Alexander Archipelago.

Figure 7.2 Schematic model for the evolution and diversification of land use systems of the Olympic Peninsula. (click on image for a PDF version)

While there do appear to be general correspondences between these two strategies and the linguistically defined tribal groupings on the Peninsula, it should also be noted that both land use strategies are represented within individual tribal groupings too and there is a sort of gradient from mostly maritime to mostly riverine/terrestrial resource orientations moving upriver along the larger drainages like the Quinault. There were, for example, Quinault and Queets villages at the mouths of these rivers and their settlement strategy was apparently quite maritime compared to their upriver neighbors. Within the Quileute, the Klallam, and the Twana there are examples of both riverine and coastal village settlement strategies. These within-group variations suggest that there is no simple correspondence between ethnic groups and the different organizational forms discussed here.

The diagram of Figure 7.2 leads to the impression that a linear evolutionary progression is being postulated. However, the arguments that have been presented do not require in-place linear developments to account for the cultural sequence represented in the archaeological record of the Olympic Peninsula region. Actually, it should be recognized that the emergence of the maritime collector system may have occurred through a process of colonization from the north. The Makah are, of course, linguistically related to the Nootka of the west coast of Vancouver. They shared with the Nootka a maritime collector system and environmental settings that were not well endowed with salmon. Initial development of a maritime collector type land use system on the Olympic Peninsula may have come about by a process of extension of a new land use strategy that first developed to the north and subsequently expanded southward to the limits of the habitat that it evolved to exploit. In short, the transition from riverine to maritime collector type systems need not have occurred in situ on the Olympic Peninsula but may have been first established through an expansion of a land use strategy that originated in another region. In this sense, the spread of the maritime land use strategy onto the Olympic Peninsula would have filled in areas that were either minimally used or very marginal habitat for the riverine-based settlement strategy.

It should be clear from this example that Figure 7.2 should not be interpreted in cultural historical terms. There is no intent to suggest a phylogeny of cultural or ethnic groups but only a developmental sequence of land use systems. Whether or not there was biological continuity in the various populations that occupied the Olympic Peninsula throughout the Holocene is an interesting question but not one which can be adequately addressed at this time.

The patterns identified here within the aboriginal groups of the Olympic Peninsula at the time of contact correspond with the broader environmental and social gradients along the Northwest Coast (Schalk 1981; Richardson 1981; Ames 1985). In moving from south to north along the Northwest Coast, there is a general increase in dependence on marine resources, decreasing dependence on terrestrial resources, and a tendency toward greater resource clumping (Schalk 1981). Associated with this clinal distribution there is an increase in home range size, village size and logistic mobility (Schalk 1981). It has been suggested that the size of the resource owning group (Richardson 1981) and social hierarchies also increase northward (Ames 1985). Most of these broad north-south trends seem to apply even at the level of that small section of the Northwest Coast that comprises the Olympic Peninsula. In terms of the land use models that have been presented in this chapter, the Olympic Peninsula region encompasses a transitional zone between aboriginal land use systems that were dominantly riverine to the south and marine to the north. This transition appears to be a regional manifestation of a broader pattern in the relative productivity of terrestrial and marine resources along a latitudinal gradient of Western North America.

Alternative Models

Before summarizing this chapter, it is appropriate to place the arguments that have been presented into the context of two other explanatory models for the development of the cultural systems described in ethnographies of the Northwest Coast. These two models differ fundamentally from the perspective developed in this study in that they both view demography and culture as entirely dependent variables responding passively to changes in the natural environment.

At least since Kroeber's (1939) landmark study entitled Cultural and Natural Areas of Native North America, the Northwest Coast has been viewed as derivative of the interior land-based hunting societies. In this light, it is not surprising that the first monograph-length archaeological site report for the Northwest Coast was presented from this general perspective. This study by Arden King (1950) was a report on excavations in 1946 and 1947 at Cattle Point in the San Juan Islands. In this large, stratified deposit, King defined a series of four phases interpreted as spanning the transition from hunting-and-gathering based primarily on land resources to fully maritime economies indistinguishable from the historic cultures of the region. The earliest cultural levels at Cattle Point contained chipped stone tool assemblages dominated by leaf-shaped points and faunal remains indicative of a land-based economy. Shellfish were entirely lacking, fish and sea-mammals were scarcely represented and deer bones were most abundant in these levels. For this level, King (1950:79) noted "affinities with the interior of western North America" (1950:79) and "a hunting and gathering economy with some fishing as an adjunct to the food supply." In the overlying stratum, shell remains (especially mussel) appeared, along with increasing quantities of seal bone, a greater variety of bone tools and, in general, evidence for a growing dependence upon resources of the sea. This level is characterized as the "Developmental phase," suggesting economic changes leading up to a "Maritime phase". Deposits assigned to the latter period contained abundant, uncompacted shellfish remains, large numbers of fish and sea mammal remains, and greater importance of bone and ground stone tools. Overlying this stratum were deposits containing only a few tool types which King termed the "Late phase".

This study and others carried out subsequently on the Southern Northwest Coast have served as the basis for what might simply be referred to as the climatic change model. I have previously summarized this model as follows (Schalk 1985:114):

The first model relates directly to the three main environmental regimes that prevailed during the interval when humans are likely to have been present in this region. These climatic regimes include (1) a period between 14,000 and 8,000 B.P. when periglacial influences were present in the form of alpine glaciers in the Olympics, (2) a second interval between 8,000 and 3,000 B.P. (Hylisithermal) when dryer and warmer conditions prevailed, and (3) the last 3,000 years with climate much like today's (Heusser 1973). This model maintains that the major intervals of culture change were direct responses to climatically induced environmental changes.

Given the differences in resources associated with each of these climatic regimes, this model predicts a sequence of three adaptations. The first is a big game hunting adaptation that would have prevailed prior to the wasting of alpine glaciers before 8,000 B.P. This adaptation would have been characterized by primary dependence the year around upon herds of large herbivores, some of which are now extinct. With the onset of dry and warm conditions of the Hypsithermal, the diversity of large ungulates would be reduced and deer and elk would have become the primary game resources. Carrying capacity of these species would have been quite high due to open forest conditions and milder winters. Mobile hunting adaptations that focused upon deer and elk would be expected to have prevailed from about 8,000 to 3,000 BP. A major land use change would be expected at ca. 3,000 BP. when cooler and wetter conditions resulted in a reduction in the overall productivity of land resources. This model predicts that land hunting would have been largely replaced by systematic and intensive marine resource use as the dominant subsistence strategy. Since no significant environmental changes have been documented for the last 3,000 years of the prehistoric record, this model predicts that settlement and subsistence systems basically like those of the historic period were established from the beginning of this interval.

Elements of this model can be found in many different studies (cf. Tuohy and Bryan 1959; Bryan 1963; Kidd 1964; Mitchell 1971; Grabert and Larson 1975; Thompson 1978; Wessen 1978).

The second model, alluded to above is one which was developed by Knut Fladmark (1975; 1979; 1982). It is noteworthy that this model has as its empirical base archaeological investigations along the Northern Northwest Coast. Referred to here as the ancient mariners model, it has been the most dominant general explanation proposed for the evolution of Northwest Coast economic systems in recent years.

Fladmark contends that Northwest Coast subsistence systems have always been maritime in orientation and that the initial populations entering the New World may have exploited marine resources along the outer margins of the Coast during the Pleistocene (Fladmark 1979). The scarcity of archaeological evidence for maritime subsistence prior to 5,000 BP is attributed to inundation or destruction of the coastal archaeological record due to fluctuations in postglacial sea-levels. Archaeological remains that might be associated with terrestrial hunting adaptations (e.g. "Old Cordilleran" or "Olcott" materials) are postulated as inland dispersals of a "single basal coastal culture" (Fladmark 1982:117). The only significant economic variability recognized within the entire Northwest Coast before 5,000 BP is that "Terrestrial and littoral resources dominated the southern-inner coast, while marine species were the most significant northern fauna." This broad distinction is attributed to sea levels that were significantly higher than those of today north of Johnston Straits and considerably lower than today's south of Johnston Straits (Fladmark 1975:293).

Out of a basal adaptation that was already maritime in orientation, economic systems like those of the ethnographic horizon are postulated to have emerged in the mid-Holocene. This emergence is considered to have been relatively synchronous throughout the Northwest Coast as a response to environmental changes that were area wide. Fladmark (1975) argues that environmental instability in the form of fluctuating sea levels, river gradients, and climate during the first half of the Holocene had profound effects on food resources—especially anadromous fish.

As integral components of the coastal ecosystem, salmon require a stable environment to achieve biological equilibrium and climax productivity. Geomorphological and paleotemperature data indicate that productivity of all salmon species was reduced considerably below that of the present prior to 8,000 to 9,000 B.P. Although pink, chum, and chinook populations may have been present in glacial refugia throughout the Wisconsin, followed by increasing productivity proportional to the extent of deglaciation, it is difficult to conceive of any salmon, particularly sockeye and coho, attaining full productivity prior to the complete stabilization of stream gradients about 5,000 B.P. After sea-level stabilization and maturation of stream gradients, sockeye and coho probably rapidly attained a high level of productivity, producing a marked increase in energy resources available to aboriginal occupants of the Northwest Coast (Fladmark 1975:207).

Fladmark interprets the archaeological appearance of shell middens as symptomatic of the emergence of semi-sedentary salmon-based economies basically similar to those documented ethnographically. Increased shellfish use is interpreted to be a consequence of larger, more sedentary populations. This transition is postulated as a coast-wide phenomena at about 5,000 BP. Most importantly, this emergence is viewed as mainly an increase in social complexity without fundamental change in subsistence:

It is worth reiterating, however, that all or most of the fundamental technological and economic basis of Northwest Coast culture may always have been present, from the very first occupation of the region. This would most likely have been the case if the initial occupants infiltrated the region from other coastal areas of the North Pacific—an hypothesis more supportable at the moment than any notion that the bulk of early coastal peoples penetrated out to the Pacific from various interior homelands. If coastal peoples have always been maritime adapted it becomes difficult to interpret later social elaboration as simply the result of fundamental evolution in subsistence related technologies or strategies (e.g., Burley 1979; Schalk 1977). However, such social elaboration might result from some significant changes in the natural environment, affecting the quality and/or quality of resources effectively obtained by an essentially unchanging resource exploitation strategy. For this reason, I have contended that the apparent synchroneity of initial archaeological indications of cultural complexity (the early Developmental Substage) with initial stabilization of relative sea levels is significant (Fladmark 1982: 132).

While Fladmark is recognized as the major proponent of this model, many other archaeologists maintain that the lack of earlier evidence for marine-oriented subsistence is an understandable consequence of changing sea-levels (cf. Mitchell 1971; Grabert and Larsen 1975; Thompson 1978; Wessen 1978).

To reiterate, both of these models view human demography as a totally dependent variable in the evolution of Northwest Coast cultural systems. Both identify gross environmental changes as the major driving force behind cultural changes. Although the geographic boundaries to the regions over which either of these models are supposed to apply do not seem to have been clearly identified, one was developed primarily with Southern Northwest Coast archaeological data and the other mostly Northern Northwest Coast data. The models developed in the present study are intended to apply specifically to the Olympic Peninsula Region. It has already been emphasized that this region seems to straddle a boundary or transitional zone between two subareas of the Northwest Coast that differ ethnographically, archaeologically, and environmentally.

Summary

Population growth is viewed as the major driving variable behind culture change in the direction of greater reliance upon marine resources. Climatically induced environmental change is not excluded altogether but emphasis upon population growth, particularly during the late prehistoric times, provides a non-cultural mechanism for explaining cultural change during an interval of no significant climatic change. Climatic changes in this model are reduced to a catalytic role in cultural change that is driven primarily by demographic conditions. This model maintains that the major land use changes represented in the archaeological record of the Northwest Coast involved (1) a seasonal shift from hunting as the major year round economic pursuit to use of fish and some marine resources during spring, summer, and fall, (2) a seasonal shift from hunting to reliance upon stored food as the primary winter season subsistence activity, and (3) emergence in some areas of the Peninsula of fully maritime adaptations based upon highly logistic land use strategies.

The first change is postulated to have occurred between about 6,000 and 3,000 yrs B.P. and amounts to a trajectory of increasing marine resource dependence within systems that were forager-like in character. The second change involves crossing the threshold between systems that are basicly foragers in their organization to those that more closely approximate the collector system type. This change was arguably rather abrupt and should be marked by the most pronounced changes in the character of the archaeological record throughout the Holocene prior to major impacts of European contact. The time frame for this development is estimated at roughly 3,000 yrs. B.P. The third change was restricted to those areas of the region that were particularly well endowed with marine resources. The emergence of fully maritime adaptations is again viewed as a developmental trajectory of growing dependence upon marine resources but, in this case, the trajectory is reflected in variations in the organization of land use systems that fall under the collector rubric. These maritime systems are postulated to be the most recent systems to emerge prehistorically and this development is estimated to have occurred at about 1000-1500 yrs. B.P.

The major changes in land use systems during the Holocene are viewed as providing the energy base necessary to support larger populations through time. Such changes can be viewed as intensification in the sense that they solve the basic problem of how to maintain larger numbers of people when the amount of land per capita is shrinking. Inasmuch as the major changes proposed also involve longer spatial and temporal pathways for energy flow through cultural systems, the proposed changes imply progressively more complex economic and social systems.

The suggested time for the emergence of winter-sedentary systems of land use on the Olympic Peninsula is highly tentative and is based upon very limited archaeological evidence from this region. [3] Given the presently available radiocarbon evidence for Peninsula sites (see Figure 6.1), it seems clear that there are no reported archaeological manifestations that hint at the existence of winter sedentary land use systems in this region any earlier than the 3,000 yrs B.P. estimate. On the other hand, it is entirely possible that the timing of this development may have been considerably later than 3,000 yrs B.P.. Admittedly, the 3,000 yr B.P. estimate is based partially upon the better known archaeological record of the Fraser River Delta (e.g. Mitchell 1971) where specialized, seasonally occupied resource procurement sites date to 3,000-3,500 yrs B.P. (Ham 1982:367). Although structural remains are reported from sites on the Fraser Delta that may be as much as 4,500 to 5,000 yrs old, none of these has been clearly identified as a winter season occupation. Most appear to be early spring herring and shellfish gathering sites (Ham 1982). Obviously, the approximate timing for the initial appearance of winter sedentism even in areas as well known as the Fraser Delta is not altogether clear. This seems to be the result of a lack of agreement among archaeologists about what the archaeological correlates of a winter sedentary land use system are.

Given the substantial environmental differences that exist from one region to the next within the southern Northwest Coast, it seems unrealistic to expect that major changes in land use such as the emergence of semi-sedentism were synchronized over the entire area. More probable is an intricate mosaic in land use systems in which earlier forms existed in certain areas much longer than in neighboring areas. Such a pattern has begun to emerge for the initiation of winter sedentism on the Columbia Plateau (Schalk 1987). In this same sense, forager-like land use systems may have persisted considerably longer over much if not all of the Olympic Peninsula than these did in other areas of the Southern Northwest Coast.

Other models for the emergence of the cultural systems recorded ethnographically on the Northwest Coast have been formulated with particular emphasis upon the northern Northwest Coast (see Fladmark 1975; Aikens et al. 1986). By contrast, the models proposed in the present study have been formulated to account for the archaeological record of a specific region. If, however, it is accurate that the Olympic Peninsula straddles a transitional zone between two provinces that had rather different land use systems historically, the archaeological record here may reflect a more complex developmental history than is reflected either in regions to the north or the south. Some of the implications of this observation are considered in the next chapter.

The models that have been developed in this chapter and others in this report can be used to develop predictions for the regional archaeological structure of the Olympic National Park. As a means of spatially differentiating important environmental variations that are likely to have significance for archaeological variability, it is necessary to divide the Park into research units. The next chapter proposes a set of research units that would seem most useful in approaching the kinds of archaeological problems that have been addressed in this study. For each research unit, expectations that flow out of the aboriginal land use models are discussed.

ENDNOTES

¹According to Trainer (1971; cited in Starkey et al. 1982:355), only 50 percent of the Roosevelt elk cows become pregnant each year compared to 90 percent for Rocky Mountain elk. Recruitment rates are similarly depressed for Roosevelt elk (30-40%) compared to Rocky Mountain elk (70-90%; ibid). Forage quality is identified as a possible cause for the lowered reproductive rates of Roosevelt elk, especially selenium deficiency—an endemic problem in the Northwest today. Whether the same patterns held in the past is uncertain, however (Starkey et al. 1982).

²The various saltwater "passes" or approaches to the Fraser River are the equivalent of large rivers and, therefore, exceptions to the argument proposed here. These exceptions would include the salmon migration routes between certain of the San Juan Islands and a few other islands in the Gulf of Georgia. They would also include places like Birch and Semiahmoo bay areas where the Fraser River formerly had distributaries or locations where salmon are funnelled around coastal features as they approach this large and very productive salmon river. All of these exceptions are places with unusual intercept potential for mass-harvesting anadromous fish and, therefore, they have a "riverine" character. Adding to the exceptional character of some of these locations is that they also would have had rather high ungulate densities. The San Juan and Gulf Islands, for example, enjoyed the combined advantages of being saltwater settings with relatively high carrying capacities for ungulates (due to rainshadow effects) and excellent intercept points for the Fraser River runs along the passes between the islands.

³There has been no cultural sequence or phase chronology previously proposed for the Olympic Peninsula region. This is not altogether a negative circumstance, however, because the existence of such chronologies often tends to encourage cultural historical approaches. The "type, style, and phase fever" associated with the theoretical perspective of the cultural historian almost invariably becomes a substitute for useful contributions to knowledge about economic systems (Ham 1982:363). Given the kinds of questions that are being asked of the archaeological record, archaeological research on the Olympic Peninsula may actually advance more efficiently if it is unencumbered by the typical cultural historical framework.

Croes and Hackenberger (1988) place the Hoko River sites as well as the Ozette Village site within the framework of the phase chronology of the Fraser Delta. The Hoko River archaeological deposits are viewed as encompassing the Locarno Beach Phase (3,000 B.P.-2700 B.P.), and the Gulf of Georgia Phase (1400 B.P. to Historic). The interval between these two phases (Marpole Phase) is a depositional hiatus at the Hoko sites.