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

Paleoenvironmental Data Sources

Information on the paleoenvironments of Olympic National Park and the Olympic Peninsula region comes from two primary sources: palynology and geology. Most of these data are the result of direct paleoenvironmental reconstructions, although a certain number of inferences are drawn from the modern flora and ongoing geological processes. The following sections discuss these sources insofar as they influence Late Pleistocene and Holocene human occupation of the region.

Late Pleistocene Glacial History

The Olympic Peninsula lies at the southern extent of the Pleistocene Cordilleran ice sheet along the coast of western North America. Except for areas covered by alpine glaciers, the entire Peninsula was ice-free during the Olympic Interglacial, radiocarbon-dated to between 24,000 and 30,000 yr B.P. The Fraser glaciation extended between 24,000 and 13,000 yr B.P. and, at its maximum around 14,500 yr B.P. (the Vashon stade), much of the Peninsula was covered by glaciers (Thorson 1980b; Stuiver et al. 1978; Fulton 1971). The Juan de Fuca Lobe overrode the northern Peninsula, the Puget Lobe covered portions of the eastern Peninsula, and alpine glaciers centered on the Olympic massif extended to if not beyond the present coastline (see Figure 5.1). After this point in time, ice wastage was rapid and apparently only impeded the re-establishment of plant and animal life in the interior uplands of the Peninsula where alpine glaciers persisted until the early Holocene. Ice from the late-glacial Sumas stade of the Fraser glaciation (ca. 11,500 to 10,000 years B.P.) did not extend southward of the Fraser lowlands (Armstrong et al. 1965). By about 10,000 yr B.P., glacial ice was gone from the region (Heusser 1973).

Figure 5.1. Maximum extent of glaciation in northwestern Washington and southwestern British Columbia (redrawn from Armstrong et al. 1965). (click on image for a PDF version)

Alpine Glacial Activity during the Holocene

Studies of alpine glaciers in the mountains of the Northwest have revealed evidence of glacial deposits that occur upslope from the terminal Wisconsin glaciation moraines. It has been suggested that there was early Holocene glacial activity as well as multiple advances in the Cascade Range of Washington during the past 6,000 years—what is referred to as the Neoglacial (Burke and Birkeland 1983). A variety of dating techniques including dendochronology, tephrachronology [1], rates of lichen growth, soil development, rock weathering, and, of course, radiocarbon dating have been used in estimating the ages of the these glacial features (Burke and Birkeland 1983).

Evidence for the early Holocene advances is less widespread than for Neoglacial advances possibly because the latter were more extensive and have erased traces of the early Holocene advances (Beget 1981; Waitt et al. 1980). Early Holocene (8400-6700 BP) glacial deposits have been reported from the vicinity of Glacier Peak (Beget 1980). On Mount Rainier, one Neoglacial advance can be dated with tephras between 3500 and 2000 years old (Crandell and Miller 1964). Another appears to have occurred roughly in the 13th century (ibid). Based upon dendrochronology, the most recent advances occurred at about A.D. 1600 and A.D. 1850 (Miller 1969, cited in Burke and Birkeland 1983:9).

During the "Little Ice Age" (1600-1900), the extent of alpine glaciers in the Olympics may have been nearly double the 65 km² area these cover today (Heusser 1974:1552). This advance is estimated to have commenced in the middle of the 1600s. As recently as the early 1800s, glaciers such as the Hoh, Blue, and White extended from 150-130 m farther downslope than at present (Heusser 1974:1552). While geological evidence for two other and earlier Neoglacial advances has been observed in the Olympics (Hubley 1956), their age and extent is not known. [2]

The Holocene climatic changes represented in the activities of the alpine glaciers are also manifested in the sequence of terraces along the river valleys of the Olympic Peninsula. The Hoh River Valley and probably other valleys of the Peninsula exhibit four terrace levels which have been linked to climatic events (Fonda 1974). Patterns of vegetational succession are expressed sequentially from the lower and younger terraces through the higher older terraces. Ages of trees on the first three terraces of the Hoh River correlate with Neoglacial advances on Mt Rainier (Fonda 1974:928). The dates of these advances on Mt. Rainier have been estimated (on the basis of age of trees growing on end moraines) at A.D. 1217-1363, 1540-1635, and 1800-1850 (Crandell and Miller 1964). In other words, all but the highest of the four terraces appear to be associated with the last eight centuries of the Neoglacial.

Late Pleistocene and Holocene Paleoenvironments: Palynological Data

The composite palynological record from the Olympic Peninsula spans the period from ca. 47,000 yr B.P. to the present. Even earlier organic deposits are found in sea cliff exposures between Kalaloch and the Hoh River, dating to ca. 70,000 yr B.P. In this area, also, is a major exposure of interbedded peats and inorganic sediments that spans the late Pleistocene period of 47,000 to nearly 16,000 yr B.P. (Heusser 1977). Palynological work from these exposures suggests the existence of a refugium for various tree species including spruce, hemlock, fir, and pine throughout much of the late Pleistocene period. During this period (oxygen isotope stages 2-4), mean July temperature depressions are estimated to have been 5 to 7 degrees below modern values (ca. 15 degrees C); during interglacials, however, summer temperatures probably averaged only 2 degrees below modern values (Heusser 1977:Figure 16). The Vashon stade of the Fraser glaciation (ca. 24,000-13,000 yr B.P.), however, was among the coldest of the late Pleistocene, and was associated with a greater development of grasses and forbs as well as a proliferation of alder after the late-glacial maximum (ca. 18,000 yr BP).

The late Pleistocene and Holocene palynological record from the Olympic Peninsula derives primarily from bog and lake sections rather than from sea-cliff exposures. Research on such bogs began with Hansen (1941, 1944, 1947) and has continued with the modern work of Heusser (1977), particularly in the Hoh River area (1974), Bogachiel River area (1978), Humptulips River area (1964), and Soleduck River area (1973). Florer (1973) has shown that similar trends apply even to pollen assemblages from Holocene sediments deposited off the outer Washington coast. The basic trends from pollen cores, during the last 18,800 years, may be summarized as follows:

(1) Park Tundra: 18,800-10,000 years B.P. This interval is well represented in sections from the Hoh River valley (Heusser 1974) as well as from Wessler bog to the northeast of Ozette Lake (Heusser 1973). Shrubs and herbs dominate during the early part of this interval, accounting for at least 50% of the pollen. Mosses also predominate locally, and the only arboreal taxon found in significant percentages is lodgepole pine. The latter species is a pioneer which proliferates on recently deglaciated landscapes, and was already being challenged by spruce and pine after 14,500 years B.P. according to data from Wessler bog (Heusser 1973). By 12,500 yr B.P., arboreal pollen closely resembling that of modern subalpine areas east of the Cascades (including lodgepole pine, spruce, and mountain hemlock) dominates the pollen assemblage (Baker 1983:111).

The climate of this interval was significantly colder than at present but whether it was relatively wet or dry seems less certain. Research in progress seems to indicate that temperature and moisture are not so directly synchronized as paleoenvironmental interpretions have traditional assumed (Estella Leopold, personal communication, 1986) and, in fact, Barnowski (1981) and Hebda (1983) suggest that this interval was one of cool and dry climate in the Puget Lowlands and Vancouver Island respectively. Heusscr (1977:304; 1980:Figure 2) suggests that glacial periods may have had reduced precipitation.

(2) Early Holocene Warm-Dry Interval: 10,000-8,000 yr B.P. This interval is well represented from the Hoh River valley as well as from the Bogachiel River area. In general, this was a warm and dry interval and this is indicated by the replacement of arctic-alpine species by species such as Douglas fir, alder, and bracken fern—all shade-intolerant species. During this period arboreal pollen increased from 50% to ca. 81-93% of the total assemblage. Douglas fir, a tree favored by relative drought, began to dominate pollen assemblages, particularly on south-facing slopes. Bracken fern (Pteridium aquilinum), a sun-loving species, is strongly represented in the pollen assemblages of this interval reflecting the open character of the forest (Heusser 1974). Alder continued to increase in importance until the end of this period, as a result of increasingly warming temperatures. Thus, this period may be characterized as one of increasing warmth and aridity, culminating in a postglacial xerothermic maximum sometime around 7,000-8,000 yr B.P.

This sequence is also supported by an alder maximum in offshore sediment samples radiocarbon dated to 7,750 yr B.P. (Florer 1973), and a number of pollen samples from Vancouver Island (Hebda and Rouse 1979) and the adjacent region of southwestern British Columbia (Matthewes and Heusser 1981, Hebda 1983). At Bear Cove Bog on Vancouver Island, for example, Hebda (1983) interprets the pollen assemblage to reflect a shift from cool, dry conditions between 14,000-11,500 yr B.P. (assemblages dominated by pine, alder, and herbs) to cool, moist conditions (assemblages dominated by pine and hemlock) and finally to a postglacial xerothermic maximum culminating ca. 7,000-8,000 yr B.P. (assemblages dominated by Douglas fir). Mathewes and Heusser (1981) have suggested that the maximum postglacial temperature on Vancouver Island may have been ca. 16 degrees C, or about 2 degrees above modern July temperature values. Similarly, Heusser (1977:Figure 16) argues for a July temperature maximum at about 7,000 yr B.P. of ca. 16 to 17 degrees C, or about 1-2 degrees above modern values.

(3) Mid-Holocene Cooling: 8,000-3,000 yr B.P. The beginning of this interval is marked by the peak Holocene value for alder. During this interval, a noticeable change in forest composition begins to take place, as cool, moist conditions once again begin to return to the Olympic Peninsula region. Alder, Douglas fir, and bracken fern decrease throughout this interval. Simultaneous with these declines, pollen diagrams from the Hoh River, Bogachiel River, and Humptulips areas, indicate that hemlock and spruce begin a steady rise to postglacial maximum values. Cedar (Thuja) also begins its rise at about this time, presumably immigrating from a refugium zone outside the region. [3]

(4) Onset of Modern Conditions: 3,000 yr B.P. to Present. During this period, the character of the modern regional environment finally takes shape, dominated by a flow of Pacific maritime air onto the western Olympic Peninsula with accompanying cloudiness, high humidity, and abundant precipitation (Maunder 1968). Douglas fir and bracken fern values are low due to the cool, humid climate which prevails. Hemlock and spruce are the dominant species in the forest assemblage and cedar becomes important for the first time. In short, this interval represents the onset of climatic and vegetational conditions comparable to those of today. Probably the most profound effect associated with the beginning of this period as far as humans are concerned is that it marks the achievement of fully closed forests.

In comparing the various palynological sources for Northwestern Washington and surrounding regions, it is clear that the vegetation history varies at least in its details from area to area. For example, Heusser (1978:1576) notes that treeless conditions persisted longer after glacial retreat in the Hoh Valley than in the northwestern corner of the Olympic Peninsula. Similarly, pollen analyses carried out in the rainshadow portion of the Peninsula at the Manis site near Sequim revealed that herb and shrub communities including cactus persisted at 11,000-12,000 yr B.P. (Peterson et al. 1983). The arrival of coniferous trees in this area was apparently delayed by aridity until sometime between 11,000 and 9,600 yr B.P.

In the palynological records that have been analyzed for the Peninsula, Heusser (see especially Heusser 1978:1577; also Heusser 1974) concludes that the achievement of forest closure was reached about 3,000 B.P. In his Hoh valley pollen diagram (Heusser 1974), it is at about 3,000 yr B.P. that bracken fern pollen decreases to an insignificant proportion of total pollen (see Figure 5.2). It must be emphasized, however, that placement of forest closure at around 3,000 B.P. is based upon very few radiocarbon dates. The Hoh valley pollen record, for example, is apparently exceptional for the number of radiocarbon dates available for the Holocene portion of the record yet few pertain to the mid Holocene. Of the 11 dates of Holocene age in this record, only two fall between 8,660 and 2,510 yr B.P. (see Heusser 1974:Figure 7). With no finer chronological control than this, the temporal placement of forest closure on the Peninsula seems less certain. We are suggesting that the 3,000 date appears to be an estimate that may prove to have a very large margin of error—possibly amounting to 2-3,000 years even. This suggestion is made in the light of accumulating data from recent palynological research in surrounding areas.

Figure 5.2 Pollen Diagram from Hoh River bog (from Heusser 1974: Figure 7). (click on image for a PDF version)

The traditional Antevs (1948, 1955) three-stage Holocene climatic model that has gained such widespread application over much of western North America (Aikens 1983:239) is increasingly inconsistent with the pollen record of the coastal Northwest. This classic climatic framework postulated a cool-moist period from about 9,000 to 7,000 yr B.P. (the "Anathermal"), a warm-dry period from 7,000 to 4,500 yr B.P. (the "Altithermal"), and a final period with climate approximating modern conditions (the "Medithermal"). Pollen records from the Olympic Peninsula call into question the existence of an interval comparable to the "Anathermal" and, in fact, these records indicate that this was the warmest and dryest part of the Holocene in this region.

It is also evident that the earlier concept of an "Altithermal" (Hansen 1947) or "Hypsithermal" (Heusser 1973), a dry interval spanning roughly from 8,500 to 3000 yr B.P., is not being supported in some of the more recent palynological research. Based upon his research in southwestern British Columbia, Mathewes (1973:2085) concluded that there was "virtually no evidence for a classical Hypsithermal interval between 8500 B.P. and 3000 B.P." Admitting the validity of a Hypsithermal for interior areas of the Northwest (e.g. Columbia Basin or south-central British Columbia, Mathewes (1973) originally questioned the existence of a warm and dry interval during any part of the Holocene for the coast of British Columbia. Forest succession after deglaciation apparently was viewed by Mathewes as sufficient to account for the vegetation changes indicated for the Holocene.

Subsequently, Mathewes and L. Heusser (1981) applied transfer functions to obtain mean July temperatures and mean annual precipitation from pollen counts with quite interesting results. From these data, they found that the temperature maximum occurred between 10,370 B.P. and 6,000 B.P.—a very different time span than has generally been postulated. Rather than a mid-Holocene warm-dry period, they propose that the early Holocene was warm and dry whereas after 6,000 yr B.P. essentially modern conditions prevailed. The temperature and precipitation curves developed by Mathewes and Heusser are illustrated in Figure 5.3. Barnowski's (1981) finding that modern vegetation appears by 5,500 yr. B.P. in the southern Puget Lowland seems to lend further support to these new views on the "Hypsithermal." Hansen (1974) also seems to concur with a revision of the time scale for a Holocene "Thermal Maximum." Finally, Baker (1983:111) places the end of the warm, dry period at 7,000 B.P for western Washington and suggests that open Douglas fir forests were replaced by closed climax forests like those of modern times by 6,000 B.P.

Figure 5.3 Curves of reconstructed mean annual precipitation and mean July temperatures from Marion Lake sediment core (redrawn from Matthewes and Heusser 1981). (click on image for a PDF version)

In sum, the trend of the more recent palynological research seems to be leading in the direction of a fundamental revision of Antevs' Holocene climatic scheme. The three-stage Holocene climatic model seems less defensible than a two-stage model for western Washington and British Columbia. In other words, a warm-dry interval during the first part of the Holocene is followed by a cool-moist interval. Vegetational changes associated with these climatic intervals involve the transition from open to closed forests. What is less clear from the pollen records, however, is when this shift occurs. Although previous palynological research on the Olympic Peninsula places forest closure at ca. 3,000 yr B.P., there is a growing body of evidence from surrounding regions suggesting that closure may have been reached as early as 5,500-6,000 yr. B.P. Whether the Olympic Peninsula differed from these other areas or whether apparent differences are the result of inadequate chronological resolution in previous palynological studies remains to be seen.

In any case, the implications of forest closure are of considerable importance for productivity of terrestrial food resources and this subject is addressed in the following section.

Faunal Resources of the Late Pleistocene and Holocene

Ungulate Resources

Like the vegetation, animal life was profoundly influenced by cold climate and conditions that could be considered periglacial prior to 10,000 yr B.P. Although piedmont glaciers that bordered the Peninsula on the north and east had retreated by 14,000 B.P., the local effects of alpine glaciers in the Olympics persisted to about 10,000 yr B.P. Data on how animal resources responded to these influences are rather scarce. It is clear, however, that Rancholabrean fauna were present at least in the northeastern part of the Peninsula during the late Pleistocene.

Contrary to popular belief periglacial settings are not necessarily inhospitable for humans. Geist (1978) has described low-latitude, periglacial ecosystems as being exceptionally productive of large mammals. Besides resembling the African savannahs in terms of the abundance of large mammals, periglacial settings are characterized by high species diversity and individuals with exceptional body size.

There are major contrasts in the ecology of the windward and lee sides of a glacial system (Geist 1978:196). The unique and productive characteristics of periglacial ecosystems are distinctive to the leeward side of glacial systems. These are what Geist refers to as "pulse-stabilized ecosystems". The windward sides, by contrast, are moist, cloudy, and generally not productive settings for ungulates. Assuming that alpine glaciers in the Olympics were the dominant environmental influence during this interval, a northwest-southeast division of the Peninsula along the crest of the Olympics appears to have significance for food resource distributions.

In the light of these arguments, it may not be coincidence that most of the reported occurrences of extinct fauna on the Peninsula come from the area between Port Angeles and Port Townsend. Mastodon remains have been found east of Port Angeles in Pleistocene gravels and sands (Tabor 1975:75). In recent years there have been reported finds in a high terrace overlooking Admiralty Inlet at Fort Warden State Park near Port Townsend and at the Batelle Memorial Research Lab and John Wayne Marina near Sequim. Remains of mastodon, bison, and caribou have been recovered from the Manis site near Port Angeles and at present this site is the only deposit of Late Pleistocene age that is purported to have an archaeological association (Gustafson et al. 1979). The age of this deposit is documented with radiocarbon dates of 11,850 and 12,000 yrs B.P. (Gustafson et al. 1979). Elk, deer, moose, musk ox, bear, camel, mountain goat, ground sloth, and a number of other large mammals also may have been present on the Peninsula at this time although remains of these species have not been reported as yet. In general, food resources available would have been dominated by large land mammals.

There is very little direct evidence or scientific literature pertaining to changes in faunal resources in this region during the Holocene. Associated with the climatic/vegetational sequence that has been outlined earlier, however, it is possible to identify differences in the relative abundance of terrestrial game resources. These expectations may be listed:

1) Terrestrial game resources, particularly ungulates, would have been unusually abundant during the warm-dry interval between about 10,000 and 6,000 yr B.P. Deer and elk may have been the only ungulates present though in the absence of more evidence it would be premature to exclude bison. The open forests and dry conditions would have favored grazers. Conditions of this period would imply that ungulate densities would have been relatively high compared to modern densities. Assuming that Roosevelt elk were already present in the early Holocene (Guthrie 1966, cited in Starkey et al. 1982) their rather specialized feeding requirements may have kept their densities very low. Given the contrasts in deer and elk numbers today between the windward and leeward Olympics, (see Chapter 3), deer densities should have been quite high relative to elk throughout the Peninsula. Those ungulate species that most effectively used early successional vegetational communities would have constituted relatively high proportions of the ungulate biomass in this region. Ungulate carrying capacity in the rainshadow portion of the Peninsula must have always been high relative to more mesic areas but would have been especially so during the early Holocene.

2) With the onset of modern climatic and vegetational conditions, game densities would have been significantly depressed. The inverse relationship between primary and secondary biomass implies that overall carrying capacity for game is lowered by forest closure. Another consequence of the vegetational change would be that a greater proportion of the ungulate biomass would consist of animals (e.g. Roosevelt elk) that most effectively make use of old growth forests.

Anadromous Fish Resources

At the time of the glacial maximum, there can be little doubt that anadromous fish spawning habitat would have been largely if not completely eliminated on the Olympic Peninsula. It is less clear how quickly after deglaciation populations of these fish would have established themselves in the rivers of this region. Some archaeologists have suggested that anadromous fish resources would have been present in very limited numbers in Northwest Rivers after glacial retreat due to rapidly changing river channels and gradients (Fladmark 1975; Thompson 1978:32, 40). According to this argument, full productivity of salmon streams could not have been achieved until sea levels had stabilized at around 5,000 B.P. (Fladmark 1975).

Without denying the dynamic nature of sea levels and stream gradients following glaciation, it is also arguable that establishment of anadromous fish runs would have followed very rapidly upon the retreat of the ice. There are a number of points that can be made in support of this alternative. One point is that salmon and many other species of fish are, after all, "r-selected" species—they have reproductive strategies that evolved as adaptations to environmental instability. Such species typically produce eggs or young greatly in excess of the number that will actually survive to reproductive age. Although the losses from predation, habitat changes, climatic fluctuations, and a wide range of other conditions may be very high, a sufficient number of young "run the gauntlet" to become reproductive adults.

A second point here is that relatively little time is required for populations of anadromous fish to reclaim suitable spawning habitat. While admitting minor differences in homing ability among different species, it is highly unlikely that accessible habitat would go unutilized for more than a few years. Even if the process of reclaiming spawning streams occurred over a period of several decades following full deglaciation, development of strong fish runs would require mere "moments" of time from the temporal perspective of archaeology. [4]

A third point is that, contrary to popular belief, some species of anadromous fish including salmon migrate up rivers that carry substantial quantities of glacial silt. Evidence of this fact can be seen in Alaska today where rivers in rather close proximity to the glaciers of southeast Alaska are good producers of salmon (De Laguna 1972; see also Baillie-Grohman 1907). Similarly, glacially fed streams of the Olympic Peninsula can be good producers of salmon today.

A final point is that the various species of salmon have quite different spawning habitat requirements and some are specifically adapted to the harsh freshwater environments of the higher latitudes. Those species which are most abundant in the northerly areas of salmon distributions are the pink, chum, and sockeye salmon (Schalk 1977). Pink and chum salmon have adapted to the highly stressed freshwater environments of the far north by being more anadromous than other species such as coho or chinook. That is, they travel only short distances upstream to spawn and young pink and chum salmon migrate to the ocean quickly after hatching and are minimally dependent upon the freshwater environment as rearing habitat. Although young sockeye may spend up to two years in freshwater before migrating out to sea, the lakes that are invariably present in the drainages used by this species buffer their young against the rigors of cold-stressed environments. It would seem that species such as chum, sockeye, and pink which dominate the modern salmon catches in Alaska would have dominated areas to the south during the cooler periods of the late Pleistocene. Coho salmon and steelhead trout may have been restricted to rivers well to the south of the Olympic Peninsula during such cold climatic regimes. If present at all, chinook salmon are unlikely to have been differentiated into spring, summer, or fall runs like this species is today in the larger rivers of western North America. Rather, there would have been a single, brief run towards the end of a relatively short summer season.

Plant Resources

Edible plant resources are exceptionally scarce in a climax coniferous forest, and in this respect, they closely parallel terrestrial game resources. The productivity of plant food resources is directly correlated with length of growing season and the degree to which sunlight reaches the ground without being intercepted by a forest canopy. These basic ecological principles lead to the following expectations:

1) Edible plant resources would have been very limited in abundance, and therefore, of minimal dietary importance for humans in this region during the late Pleistocene. Plants with tuberous roots or seeds that can be digested by humans would be very scarce. Much of the primary production in a park tundra environment is only edible by grazers and browsers and then only for a brief period each year.

2) The availability of edible plants would have been higher during the early Holocene (10,000-6,000 B.P.) than at any other time during which this region was occupied by hunter-gatherers. The increased warmth associated with this interval would have meant longer growing seasons with increased primary production. Because open forests were characteristic of this period, much of the primary production would have occurred near the ground where it is accessible to humans. Owing to rainshadow effects, the northeastern part of the Olympic Peninsula should have been especially well-endowed with plant resources during the Holocene.

3) The cooler, more moist conditions of the late Holocene would have resulted in a substantial reduction in available plant resources. The most direct cause of this reduction would be development of closed canopy forests in which plant foods are quite limited. Important plant food resources would have largely restricted to openings or "prairies" in the forest.

Relative Sea Level in the Holocene

Changes in relative sea level are of considerable archaeological interest because these condition the formation and preservation of the archaeological record. Such changes also have potential to influence the distribution and abundance of important classes of aboriginal food resources.

With the withdrawal of the Late Pleistocene ice, climatic conditions in the Peninsula region were controlled both by world-wide climatic amelioration of the Holocene epoch and by air current phenomena restricted to the Northwest region (and, in some cases, the Olympic Peninsula). Associated with those climatic changes were eustatic changes in sea level resulting from the reductions of world-wide ice volumes. However, eustasy alone does not explain the observed Holocene changes in sea level along the Washington coast; isostatic and tectonic phenomena must also be considered.

After the recession of the Vashon ice, late Pleistocene relative sea level ranged well above present, with most values falling in the range of 100 to 150 m above present sea level (Armstrong et al. 1965; Larsen 1971). On the outer coast of the Peninsula, relative sea level rose less dramatically between 13,000 and 11,000 years B.P., with values of sea level rise ranging from ca. 15 to 50 m above present The difference was probably due to the greater ice loading on the Strait of Juan de Fuca and Puget Sound areas of the Peninsula (Thorson 1980a). Following deglaciation, however, a rapid emergence began to occur, probably as a result of the reduced thickness of the ice sheet and relative closeness of the region to the receding ice mass. By ca. 9,250 years B.P., the sea had fallen to its minimum postglacial level. Evidence from the Fraser Lowland and the inner coast of Puget Sound (based on radiocarbon dates on buried peats) suggests that sea level had fallen to as much as 11 m below present between 8,500 and 7,000 years BP (Larsen 1971), but on the outer coast it apparently fell only 1-2 m below present levels. The latter appears to be substantiated by the existence of wave-cut platforms (cf. Baldwin 1939). Again, the variation in sea level histories was probably due to differences in late glacial ice loading. There is some potential for the existence of archaeological sites dating to this time period, although they would likely be affected by subsequent sea-level rise, depending on the area in question. By 7,000 years B.P., however, the effects of isostatic rebound had greatly diminished. Continued eustatic sea-level rise caused rapid submergence of shorelines until ca. 4,500 years B.P. Subsequently, with both isostatic and eustatic factors largely diminished, tectonic factors came into play as the primary determinants of late Holocene sea levels. At this point, the sea level histories of the eastern and western Peninsula took on a very divergent character. This general model for relative sea level on the western Olympic Peninsula is depicted in Figure 5.4.

Figure 5.4 Relative sea level change model for the western Olympic Peninsula. (click on image for a PDF version)

The early Holocene period from ca. 9,000-4,000 yr B.P. is the most difficult to deal with in terms of reconstructing sea-level histories (Thompson 1978), but accumulating data from a variety of sources indicate the magnitude of the divergence between the outer and inner coasts (e.g., the east coast of Puget Sound). These data suggest that, following a higher-than-present mid-Holocene sea level stand, sea levels have continued to fall along the outer coast. That is, a mid-to-late Holocene pattern of submergence characterized the inner coast, while a pattern of emergence characterized the outer coast [5]

Active uplift, tilting, and faulting of the coastal margin is indicated in terrace formations along the outer coast, from Copalis north to Cape Flattery. Indications of this uplift have been known since the turn of the century. Uplift and tilting affects Tertiary and Quaternary deposits alike, and has probably been occurring since Tertiary times. The evidence, reviewed in the late 1930s by Baldwin (1939) and Glover (1940), suggests that the uplift has been continuous throughout this period, and is not solely the result of Pleistocene isostatic phenomena. The tilting and faulting of the sediments, in particular, suggests the importance of tectonic factors. Today we know that the direction of tilt follows the subduction of the Juan de Fuca Plate and is related to tilt occurring in the adjacent Olympic Mountains (Adams et al. 1980).

Unfortunately, there are few examples of raised beaches that would indicate the nature of maximum mid-Holocene sea level rise and subsequent emergence or uplift (falling sea level). Although Heusser (1972) and Florer (1972) provide evidence for raised beaches in sections between Kalaloch and the Hoh River, these sections yielded radiocarbon dates no younger than 16,700 years B.P. Recently, however, Stein (1984) has provided radiocarbon dates of 5,000-7,000 yr B.P. for wood from an upper section of the Kalaloch formation.

Western Vancouver Island, which appears to show a similar profile of Holocene emergence as the coast of the Olympic Peninsula (or the Queen Charlotte Islands), has also produced a number of other types of supportive data for this emergence (cf. Clague et al. 1982):

(1) A number of elevated beaches have been recorded, including a 7 m beach at Yuquot radiocarbon-dated to 4,230 yr B.P., a 3 m beach at Tofino dated to 1,540 yr B.P., and a 1 m beach at Hesquiat Harbor dated 680 yr B.P.

(2) Elevated marine/freshwater boundaries have been identified in lake cores, including one from Village Lake on the Hesquiat Peninsula measured at a 3 m elevation and dating to 2,760 yr B.P.

(3) Uplifted cave deposits have been identified, some of which include archaeological materials. The latter include two sites on Hesquiat Harbor (central west coast of Vancouver Island) ca. 2 m above sea-level which date to ca. 1,300 to 1,700 yr B.P.

Although few data on the ages of such terraces are available for the coast of the Olympic Peninsula, the data from Vancouver Island suggest that uplift has taken place throughout the Holocene, but that only in the mid-to-late Holocene period were its effects observable as first isostatic and then eustatic phenomena no longer came into play. This tends to contradict earlier models which suggested that tectonic factors were only locally important on the Southern Northwest Coast, and that east-west differences in Holocene sea level histories were best explained as a byproduct of isostatic phenomena. In order to demonstrate that tectonics are responsible, however, it is necessary to do two things: (1) to demonstrate that such processes are still occurring, and (2) to develop a general model to explain them.

A number of Canadian investigators (Clague et al. 1982; Vanicek and Nagy 1981; de Jong and Seibenhuener 1972) have demonstrated that tectonism is active on the west coast of Vancouver Island. Using data from precise relevelling surveys along the coast as well as from secular trends in tides, they have suggested a current sea level fall ranging from ca. 1.6-1.9 mm/yr, e.g., at Tofino. Interestingly, this is exactly the same range of values—1.6 to 1.9 mm/yr—that one gets if one plots recently falling sea levels based on the physical evidence from the raised beaches, elevated marine/freshwater boundaries in lake cores, and uplifted cave deposits. E. Balazs of the National Ocean Survey, NOAA, has recently published preliminary reports which suggest that for Neah Bay similar rates of emergence may have been characteristic of the recent past, based again both on precise relevelling surveys and secular tidal changes (Balazs and Holdahl 1974; Aldo and Balazs 1979). A complete profile of recent relevelling work on coastal benchmarks associated with tidal gauges for Neah Bay, Mukkaw Bay, and Toke Point (Willapa Bay) along the outer coast reveals interesting patterns. According to these data, benchmarks set in 1954 at Neah Bay demonstrate an average sea level rise of 1.94 mm/yr, while benchmarks set in 1959 show an average sea level rise of 1.97 mm/yr. Simultaneously, tidal data from the Neah Bay station at Fish Wharf show a secular change of 0.11 feet in 20 years or 1.67 mm/yr. In addition, tidal data for Toke Point show an even greater rise of 2.17 mm/yr, although these are based on a much shorter interval (10 yr). However, these data all fall well within the same range previously established by Clague et al. (1982) for the west coast of Vancouver Island.

According to Croes and Blinman (1980), there are also indications that uplift has taken place at the Hoko River site on the northwest corner of the Peninsula:

Modern counterparts of this surface [on which the Hoko site is constructed] lie approximately 1.5 m lower in elevation, although they are apparently identical to the laminae in other respects. A major change in river discharge could conceivably account for the elevation differences, but it is unlikely that the sedimentological resemblances would be maintained. A more parsimonious explanation is that the base level of the river (sea level) has fallen approximately 1.5 m since the site was formed. Independent evidence of recent relative sea level change (probably tectonic uplift) has been accumulated by Dr. Parke Snavely, Jr., USGS, at Clallam Bay, approximately 8 km to the east of the Hoko River mouth (personal communication). His preliminary estimate of 2 m of change in the past 2,000 years is in reasonable agreement with the Hoko River model (Croes and Blinman 1980:83).

These data suggest that late Holocene uplift in the northern Olympic Peninsula region has been occurring at the rate of 0.5-1.0 mm/yr. These data are very distinctive from the data for the inner (mainland) coast of eastern Puget Sound which shows a current rate of submergence of ca. 0.9 mm/yr (Balazs and Holdahl 1974; Aldo and Balazs 1979). Even greater rates of subsidence have been calculated for the inner British Columbia coast (e.g. at Port Atkinson which shows a rate of subsidence of 1.2-1.3 mm/yr; Clague et al. 1982; Andrews and Retherford 1978; Mathews et al. 1970; Fladmark 1975). The rates of inner coast submergence, furthermore, correlate well with the submergent Holocene curves produced by Larsen (1971), and Grabert and Larsen (1975) for the Bellingham area of Northern Puget Sound and with observations of submerged archaeological sites in the Puget Sound region (e.g. Old Man House; Gaston and Jermann 1975).

This observed emergence of the outer coast and submergence of the inner coast during the Holocene can be attributed to tectonic phenomena, specifically the pattern of subduction of the Juan de Fuca plate, which is subducting below Puget Sound and across the channel subdividing Vancouver Island from the adjacent inner coast (Riddihough 1979, 1982). Fault zones associated with the subduction of this plate have been traced in detail (Gower 1978), and seismic studies and other geophysical investigations have related these fault patterns to magnetic and gravity anomalies in the Strait of Juan de Fuca and Puget Sound (Danes et al. 1965; Rogers 1970; MacLeod et al. 1977).

On the northern portion of the outer coast, a more steeply sloping, higher energy coastline may have precluded noticeable changes in offshore sedimentation during much of the Holocene. In other areas, however, one could expect that longshore drift would have been an additional factor in altering coastlines during this period (Rau 1973, Tabor 1975). Recent research, for example, has demonstrated that the sandy spits of the Chehalis/Grays Harbor region are relatively recent, late Holocene phenomena that are constantly being altered through the process of longshore drift. Although less pronounced, this process may also have played an important role on the inner coast and in the region between the Hoh River and Point Grenville. This may have been particularly true during the late Holocene as sea levels stabilized to some degree.

An additional factor in sedimentation would have been the downcutting of rivers, forming terraces particularly toward the river mouths (Fonda 1974). Although most rivers on the Peninsula are relatively small, many can be expected to have formed terraces during the Holocene period. Downcutting would be expected to have increased both during periods of landscape uplift and during periods of greater water input, when water tables were higher and/or glacier melt was more rapid. This would most likely have taken place during both the early Holocene (when moisture was still high and glaciers were contracting) and the late Holocene (when moisture was again high and sea level was falling). In the latter case, delta progradation would be expected to occur as sediments were deposited on the coastal shelf. In fact, such processes are well known from the Hoh, Quillayute, Sooes, Waatch, Hoko, Soleduck, Elwha, and other river systems of the western and northern Peninsula region (Rau 1973; Tabor 1975). Unfortunately, few data are available on Holocene changes in glaciation at higher altitudes in the Olympic Mountains (Spicer 1985). The general picture that is available, however, suggests that a maximum glacial retreat occurred in the early to mid-Holocene. This corroborates a notion of warmer and dryer early Holocene conditions also suggested by the palynological data.

Summary

Maximum extent of the Cordilleran Ice Sheet occurred during the Vashon Stade of the Fraser Glaciation at about 14,500 B.P. The ice sheet extended south of what is now Olympia and out the Strait of Juan de Fuca. It covered the northern and eastern portions of the Peninsula while the Olympic Mountains were under extensive alpine glaciers. Areas of the southern Peninsula were beyond the southerly limits of the Juan de Fuca-Puget lobe of Vashon ice (Easterbrook 1963; Thorson 1980a). Thus, the coastal plain and the southern Peninsula region would have been available for human occupation by ca. 13,000 years B.P.

After 13,500 B.P. climate rapidly improved. Plants that invaded the deglaciated landscape of the northern Peninsula between 13,000 and 10,000 yr B.P. were primarily shrubs and herbs but pollen studies indicate that pioneering species such as lodgepole pine and later spruce and alder were increasingly important during this interval. The vegetation during the late Pleistocene might be described as park tundra.

After 10,000, the park tundra environment was replaced by open forests with a strong component of Douglas fir and alder. These forests were associated with a warm and dry interval that apparently continued until nearly 6,000 yr B.P. Although there is palynological evidence from southern Puget Sound and southwestern British Columbia that modern climatic conditions and forest closure were achieved by 6,000 to 5,500 yr B.P., interpretations of pollen records from the Peninsula place forest closure as late as 3,000 yr B.P. Western hemlock dominated these climax forest communities.

Although studies of Holocene glacial activity have been limited for the Olympics, it is known that some of the larger glaciers advanced as recently as the early 19th century. Geological evidence for earlier Neoglacial advances exists in the Olympics, but these features have not yet been dated. From various locations in the Cascades of Washington, advances have been dated for the early Holocene (8400-6700 B.P.), the early Neoglacial (ca. 5,000 yr B.P.), and the end of the Neoglacial (A.D. 1200, 1600 and 1850). These late Neoglacial advances seem to be correlated with the development of the lower three river terraces in the Cascades and the Olympics.

Relative sea level has passed through four general stages since the late Pleistocene according to the model proposed here. The first stage associated with deglaciation (13,000-11,000 yr B.P.) was a rapid eustatic rise in relative sea. Between about 11,000 and 9,250 yr B.P., isostatic rebound resulted in a rapid drop in relative sea level. Continued eustatic sea level rise after 7,000 yr B.P. produced submergence of shorelines until ca 4,500 yr B.P. Both isostatic and eustatic forces were attenuated during the last 4,500 years when techtonic factors became the major determinant of late Holocene sea levels. Sea level histories during the late Holocene were divergent for the eastern and western sides of the Olympic Peninsula with submergence the general trend for the former and emergence for the latter. It must be emphasized that this model is based largely on extrapolation from historic geodetic and tide-gage measurements or from prehistoric relative sea level records of surrounding areas. Data on relative sea level for the Olympic Peninsula are poor and, as more information is acquired, the model proposed here may prove to be far too simplified.

A rancholabrean fauna assemblage including bison, caribou, mastodon, and possibly man was present in the region during the late Pleistocene (13,000-10,000 yr B.P.) On the leeward side of the Olympics, environmental conditions may have been quite productive of large game. Such resources are likely to have been relatively abundant and diverse in that area during the late Pleistocene compared to other regions of the Peninsula. Anadromous fish resources would have been scarce if not entirely absent from the Peninsula during much or all of the late Pleistocene. Although distribution and abundance of marine resources must have been different in many respects from their modern patterns, many of the same species would have been present because they occur along a long latitudinal gradient today. Plant resources suitable as human food would have been extremely limited in a tundra parkland setting.

During the early Holocene (10,000-6,000 B.P.) warm-dry climate and open forests would have been conducive to relatively high densities of large herbivores. Species diversity was undoubtedly much reduced from late Pleistocene times and it is possible that game resources were already dominated by the two species that are dominant today—deer and elk. Anadromous fish resources were probably established in most of the rivers of the Peninsula by the beginning of this interval if not before. Like game resources, edible plants would have benefited significantly from the open forest conditions and longer growing seasons.

Possibly as early as 6,000 yr B.P. and certainly by 3,000 B.P. closure of the forests would have resulted in a substantial reduction in carrying capacity for large game resources and edible plants. Marine resources, however, are not likely to have changed dramatically in either their broad patterns of distribution or abundance throughout the Holocene. The picture which emerges is that an overall reduction in the productivity of terrestrial resources is the most outstanding feature of Holocene environmental change in the Olympic Peninsula Region. This trend probably applies to a much broader area including much of western Washington and beyond. Previous analyses of the paleoenvironmental data for the Northwest Coast have emphasized changes in marine resources during the Holocene as the driving force of cultural change (c.f. Fladmark 1975). It is argued here and elsewhere in this study that changes in the terrestrial rather than the marine ecosystem were likely to have been of greater importance in mid-Holocene cultural changes on the Olympic Peninsula.

Footnotes

¹Although there were several eruptions of volcanoes in the Cascade Range since the late Pleistocene, tephra from most of these eruptions did not drift into the Olympic Peninsula region. However, tephra from Mt Mazama (Crater Lake), estimated to have erupted about 6,700 yr B.P., has been found in the Hoh Valley (Heusser 1974). A tephra that is likely to be Mazama was identified in the Seven Lakes Hearth site (45CA274) in the upper Soleduck basin. A radiocarbon date from a charcoal layer above the ash produced a date of 4,990 ± 60 yr B.P. (Bergland 1984:45). If tephras other than Mazama are present in this region, they have not yet been reported. The prevailing westerly winds apparently tend to shelter the Olympics from such deposits under ordinary conditions.

²Hubley (1956:671) points out that the Olympics are unique for having glaciers that form at lower elevations than anywhere else in the Northern Hemisphere. The highly maritime climate is responsible for the formation of glaciers at elevations between 1000 and 2250 m. Although most glaciers are especially responsive to summer season climatic conditions, Hubley (1956:670-671) suggests that the low elevations of the Olympic Glaciers makes them unusually sensitive to variations in winter season temperatures.

³The latter phenomenon has become the basis for recent suggestions that the expansion of western red cedar in coastal forests led to the evolution of wood-working technology among Northwest Coast Indians, and had a massive impact on the evolution of those societies themselves (Hebda and Mathewes 1984). There are two serious defects with this argument that make it questionable as a model for the explanation of cultural change in this region. The first is that while Western red cedar may have been rare prior to 6,000 B.P., macrofossils of this species from the University of British Columbia Research Forest demonstrate that cedar was present at least in limited numbers before 10,000 yr B.P. (Mathewes 1973:2100). The species would not have to be present in great abundance to have served in the capacities that cedar did in this region prehistorically. The second defect in this argument is that maritime cultures have arisen in other regions of the world in the absence of western red cedar. Even admitting that cedar has some superior properties for the manufacture of boats or houses, very efficient boats were manufactured with animal hides by coastal Eskimo groups.

⁴An important historical example supporting this argument is provided by the range expansion of the American shad (Alosa sapidissima). This anadromous species is indigenous to the Atlantic but was introduced in the Sacramento River in the early 1870s. By 1891 it was present in the rivers of British Columbia and, by 1904, Alaska (Hart 1973:95-6). In general, Pacific salmon may have stronger homing instincts than shad and the different species of salmon are also variable in their homing abilities. Nonetheless, this historical example raises serious questions about the assumption that salmon would take hundreds or even thousands of years to reoccupy the rivers of the Northwest following the retreat of glaciers.

⁵The occurrence of uplifted marine deposits near Winslow appears to be in conflict with the general pattern described here. This occurrence underscores the complexity involved and perhaps how general patterns and local phenomena can diverge.