Late Wisconsinan Port Eliza Cave deposits and ... - Wiley Online Library

Late Wisconsinan Port Eliza Cave Deposits and Their Implications for Human Coastal Migration, Vancouver Island, Canada M. Al-Suwaidi,1 B.C. Ward,2,* M.C. Wilson,3 R.J. Hebda,4 D.W. Nagorsen,5 D. Marshall,2 B. Ghaleb,6 R.J. Wigen,7 and R.J. Enkin8 1

Sub-Surface Engineering – Geology, Zakum Development Company, P.O. Box 46808, Abu Dhabi, UAE 2 Earth Sciences Department, Simon Fraser University, Burnaby, British Columbia, Canada, V5A 1S6 3 Department of Geology and Department of Anthropology/Sociology, Douglas College, P.O. Box 2503, New Westminster, British Columbia, Canada,V3L 5B2 4 Royal British Columbia Museum, 675 Belleville Street, Victoria, British Columbia, Canada, V8W 9W2 5 Mammalia Biological Consulting, 4268 Metchosin Road Victoria, British Columbia, Canada, V9C 3Z4 6 Université du Québec à Montréal, Centre GEOTOP-UQAM-McGill, CP 8888, Succursale Centre-Ville, Montréal, Québec, Canada, H3C 3P8 7 Department of Anthropology, University of Victoria, Victoria, British Columbia, Canada, V8W 3P5 8 Geological Survey of Canada – Pacific, 9860 West Saanich Road, POB 6000, Sidney, British Columbia, Canada, V8L 4B2

Sediments of Port Eliza Cave provide a record of the Last Glacial Maximum (LGM) on Vancouver Island that has important implications for human migration along the debated coastal migration route. Lithofacies changes from nonglacial diamict to glacial laminated silt and clay and till, then a return to nonglacial conditions with oxidized clay, colluvial block beds, and speleothems, along with radiocarbon and U/Th dates, define glacial–nonglacial transitions. Scanning electron microscope studies and clay mineralogy confirm that the laminated fines represent glaciation. Preglacial faunal evidence shows a diverse range from small species, including birds, fish, vole, and marmot, to larger species, such as mountain goat. Pollen data from the same unit show a cold, dry tundra environment with sparse trees. Deglaciation occurred prior to an age of 12.3 ka B.P. based on dated mountain goat bone. These data support the viability of the coastal migration route for humans prior to ~16 ka B.P. and then as early as ~13 ka B.P. © 2006 Wiley Periodicals, Inc.

INTRODUCTION There has been significant recent debate concerning the viability of a coastal migration route as an alternative to the inland “Ice-Free Corridor” hypothesis (e.g., *Corresponding author; E-mail: [email protected]. Geoarchaeology: An International Journal, Vol. 21, No. 4, 307–332 (2006) © 2006 Wiley Periodicals, Inc. Published online in Wiley Interscience (www.interscience.wiley.com). DOI:10.1002/gea.20106

AL-SUWAIDI ET AL.

Mandryk et al., 2001). Acceptance of early human settlement at Monte Verde, Chile dated to ~12.5 ka B.P. (unless otherwise stated, all ages are in 14C years before present) (Dillehay, 1997; Adovasio and Pedler, 1997; Meltzer, 1997), necessarily pushes the date of human arrival in North America further back to allow time for travel from Beringia to South America (Mandryk et al., 2001). The “Ice-free Corridor” hypothesis was developed in the 1930s following the discovery of late Pleistocene Clovis culture sites east of the Cordillera. This postulated route consisted of a corridor between the Laurentide and Cordilleran ice sheets, extending from the Yukon Valley, into the Mackenzie watershed and down the eastern edge of the Rockies (Johnston, 1933; Wilson and Burns, 1999; Mandryk et al., 2001; Dyke, 2004; Jackson and Wilson, 2004). The alternative “Coastal Migration route” hypothesis envisions early Paleoindian groups traveling along an emergent continental shelf when sea level was lower than present, likely assisted by watercraft. This proposed route extended from the Bering Strait, along the southern coast of Alaska and down the British Columbia coast (Fladmark, 1979; Dixon, 1999) (Figure 1). Until recently, an evaluation of the viability of either route has suffered from a paucity of direct evidence. To help address this gap in knowledge, we present the results of our field studies carried out in Port Eliza Cave, Vancouver Island, which lies along the proposed coastal route. This research shows that glacial ice did not cover the outer coast of the island until after 16.3 ka B.P., and that paleoenvironmental conditions immediately before this were viable for human migration. BACKGROUND AND SETTING The Cordilleran ice sheet periodically dominated the Quaternary landscape of British Columbia, Yukon, western Alberta, and portions of the northwestern United States. It comprised valley and piedmont glaciers, as well as mountain ice sheets (Clague et al., 2004). The last ice sheet, present during the Late Wisconsinan Fraser Glaciation, is well represented and has been extensively studied in southern British Columbia (Figure 2). Two advance stades, the Coquitlam and Vashon, separated by the Port Moody Interstade, have been recognized in the Fraser Lowland. Recent work in the eastern Fraser lowland indicates the Coquitlam Stade was not as extensive as previously thought (Ward and Thomson, 2004). The Sumas Stade is a late glacial re-advance. The impetus for the investigation of cave sites in this area came from seminal research carried out in Norway that established raised sea caves along glaciated coastlines as valuable sources of direct data for paleoenvironmental studies of past glaciations (Larsen et al., 1987; Valen et al., 1996). Raised sea caves form by wave action along exposed coastlines during periods of higher relative sea level. If the caves lie above the postglacial marine limit, they can preserve a distinctive stratigraphy that records glacial and nonglacial conditions. During glaciations, the cave entrance is blocked by ice and the cave fills with glacial meltwater, forming a subglacial lake, in which stratified silt and clay sediments are deposited (Figure 3). In some cases, ice may extend into the cave depositing till. During nonglacial conditions, freeze–thaw action leads to roof-fall accumulation on the cave floor pos308

GEOARCHAEOLOGY: AN INTERNATIONAL JOURNAL, VOL. 21, NO. 4

DOI: 10.1002/GEA

PORT ELIZA CAVE DEPOSITS

Figure 1. Location of Port Eliza Cave along the hypothesized “coastal migration route.”

sibly accompanied by speleothem formation (precipitated from groundwater). Bone material is commonly found in association with these units, and coupled with the speleothem deposits, can provide both chronological control and paleoenvironmental data. Port Eliza Cave is situated on the west coast of Vancouver Island (Figure 1) and was formed by wave action exploiting a fault in Jurassic volcanic rocks of the Bonanza Group. The cave is approximately 85 m above modern sea level, well above the local postglacial marine limit of 35 m. The cave is 60 m long with considerable relief, varying in height from less than 1 to 15 m (Figure 4). The floor is covered with large angular blocks near the front and has standing water above fine-grained material at the back with areas of dripstone. METHODS Fieldwork was carried out at the cave from 2000 to 2002. Results from the 2000 and 2001 field seasons were published by Ward et al. (2003). Here, we synthesize these results and add new data collected during the summer of 2002, during which a more DOI: 10.1002/GEA


309

AL-SUWAIDI ET AL.

Figure 2. Late Quaternary geologic, climate units, and events of southwestern British Columbia. Adapted from “The Sister Creek Formation: Pleistocene Sediments Representing a Nonglacial Interval in Southwestern British Columbia at About 18 ka,” by S.R. Hicock and O.B. Lian, 1995. Canadian Journal of Earth Sciences, 32, 758–767. © 1995 by Natural Resources Canada. Adapted with permission.

comprehensive study of the cave sediments was made. The previous excavation was widened and deepened and sediments described in greater detail. Large bulk samples of the basal diamicton were collected using hammer and chisel, in 10-cm increments to a depth of 80 cm. These samples were examined for macrofossils and pollen, as well as textural characteristics. Block samples were collected from Unit I-2 (Figure 4) for detailed analysis. All color descriptions were made using a Munsell chart. 310


DOI: 10.1002/GEA


Figure 3. Schematic model of glacial-nonglacial cave sedimentation based on Larsen et al. (1987). (a) nonglacial conditions—accumulation of rock fall and organic material; (b) glacial conditions—formation of subglacial lake, deposition of till and laminated fines; and (c) return to nonglacial conditions.

Granulometric analysis, based on the Wentworth scale, of bulk samples from Unit I-1 was carried out using the hydrometer method defined by Gee and Bauder (1986), while samples from Unit I-2 were processed using a Micromeritics sedigraph 5100 (Micromeritics Instruments Corp., Norcross, GA) at the University of Alberta. Quartz-grain morphologies from selected intervals within the diamicton of Unit I-1 were analyzed with a Bausch and Lomb NANOLAB LE 2100 (Rochester, NY) scanning electron microscope (SEM). Subsamples of block samples from Unit I-2 were measured for organic and inorganic carbon according to a procedure defined by Smith (2003) for clay-rich glaciolacustrine sediments. Unit I-2 sediments were mounted in resin for thin-section analysis at the University of British Columbia. Slabs 1 cm 12 cm were cut, frozen in liquid nitrogen, desiccated, and impregnated with resin. The hardened samples were then cut into thin sections for examination under a petrographic microscope. Subsamples of selected clay sediment from Unit I-2 were analyzed using a Philips model PW1730 x-ray diffractometer (Philips Electronics NV, Amsterdam, The Netherlands) to determine mineralogy. The subsamples were chosen based on dominant colors of laminations within the section. Mineral peaks were identified using the JCPDS Database (JCPDS, 1993). Single bone fragments of known species were selected for AMS radiocarbon dating. Collagen of high-molecular weight was extracted and burned to CO2 at Simon DOI: 10.1002/GEA


311

AL-SUWAIDI ET AL.

Figure 4. Cross-section and stratigraphy of Port Eliza Cave.

Fraser University and then sent to Lawrence Livermore National Laboratory (Livermore, CA) for analysis (Table I). Extraction yields and C/N ratios indicate the collagen was well preserved. All measured radiocarbon ages (given at 1 sigma) were converted to calendric ages (cal B.P. at 1 sigma) using the program CALIB v4.0 (Stuiver and Reimer, 1993; Stuiver et al., 1998). Uranium-series dating was carried out at GEOTOP-Université du Québec à Montréal-McGill, Montreal. Analysis was carried out both by thermal ionization mass spectrometry (TIMS) and alpha spectrometry. Processing was carried out using protocols defined by Edwards et al. (1987). Mass fractionation was corrected for by normalizing to the known 236U/233U ratio (1.14). The alpha spectrometry results of 232Th, 230 Th/232Th,238U/232Th, and 234U/232Th from the diamicton samples are considered to represent the detrital end member and are used to correct the dripstone samples, which showed evidence of contamination, indicated by the presence of 232Th. For paleomagnetic sampling, plastic cylinders were gently tapped into a clean vertical face in Unit I-2 and their orientations measured using a Brunton compass. Paleomagnetic remanence was measured at the Geological Survey of Canada, Pacific Division (Sidney, British Columbia, Canada) on an AGICO JR5-A spinner magnetometer (AGICO Advanced Geoscience Instruments Company, Brno, Czech Republic). The remanence was strong (0.2 to 5 A/m) and stable (median destructive field between 60 and 100 mT), indicating that fine-grained magnetite is the magnetic carrier. 312


DOI: 10.1002/GEA

PORT ELIZA CAVE DEPOSITS Table I. Radiocarbon ages from Port Eliza Cave. Unit

Materialb

Weight (mg)

13C (‰ vPDB)

I-3 I-3

Charcoal Mountain Goat

15.3 79.1

88275 I-1 Sparrow 45.0 74625 I-1 Vole 43.3 102798 I-1 Mountain Goat 51.3 88274 I-1 Marmot 46.2 77.5 97341c 102797 I-1 Vole 67.6 74624 I-1 Vole 41.1 a Lawrence Livermore Laboratory. b All bone dates were obtained on collagen. c Duplicate analysis of same bone fragment.

CAMS#a 74626 97342

14

C age (BP)

Calendar age (calBP)

–25.5 –19.8

9540 40 12,340 50

–18.6 –20.9 –19.5 –20.5

16,270 170 16,340 60 16,340 60 16,460 170 16,965 45 17,100 70 18,010 100

11,065–10,940 15,035–14,615 and 14,405–14,120 19,720–19,010 19,790–19,180 19,790–19,180 19,970–19,250 20,520–19,885 20,685–20,030 21,750–21,060

–19.7 –20.2

Sediments for pollen analysis were prepared by a standard protocol (Faegri and Iversen, 1989). Identification was made by comparison to standard reference works, an unpublished key to pollen and spores of British Columbia (Hebda, North, & Rouse, 2002), and the pollen and spore reference collection at the Royal British Columbia Museum (Victoria, British Columbia, Canada). RESULTS Lithostratigraphy Lithostratigraphic units are presented in terms of sediments of the “interior cave complex” (I-1 through 4) and those of the “cave entrance complex” (E-1 and 2) (Figure 4). Interior Cave Complex Unit I-1 is a well-indurated, largely unsorted, matrix-supported diamicton (Figures 5 and 6). It was excavated to a depth of ~80 cm, but the lower contact was not observed. Overall, the upper surface is concave with relief of 15 cm over the 1.5 m described but is also irregular with relief of 4-5 cm over distances of 10–15 cm. The unit contains ~30% clasts of varying lithology, including local cave rocks, as well as exotic clasts of porphyritic granite, metamorphic rocks, siltstone, and sandstone. Local clasts are commonly angular to subangular, whereas exotic clasts are subangular to rounded and display the full range of sphericity. Clasts range from 1 cm to 19 cm (clasts 19 were not collected). For clasts separated from the bulk samples, the average size was 1.9 cm and modal size was 0.4 cm. Dripstone fragments up to 6 cm were also found within this unit. The matrix of Unit I-1 is dominantly sand-sized, with most samples averaging 60% sand and 14% silt (Figure 6). There is little observed structural or textural DOI: 10.1002/GEA


313

AL-SUWAIDI ET AL.

Figure 5. Typical sediments from Port Eliza Cave. (a) Sample of diamicton from Unit I-1, (b) selection of bones from Unit I-1, (c) thin-section of Unit I-2 laminations with some convolutions, (d) distinctive dark beds in massive clay of Unit I-2 (trowel is 26 cm long and point is on lower contact), (e) diamicton of Unit E-1, rock hammer is 35 cm long, and (f) faulting in laminated silt and clay of Unit I-2.

variation in the sediment package as a whole, except for the upper 10–15 cm. In this portion, sediments are unsorted and fissile, with faint vestiges of bedding. It is light gray and has a finer matrix (only 20% sand and silt) with fewer granules than the sediments below. The lower portion of the diamicton is massive and dark brown. 314


DOI: 10.1002/GEA


Figure 6. Ages, grain-size, and bone concentrations of Unit I-1.

Unit I-1 contains numerous small bones, many intact and with little wear (Figure 5, Table II). They occur randomly throughout the thickness sampled, with no vertical variation in concentration (Figure 6). They range in color from white to gray and light to dark brown, and their surfaces from fresh to weathered. All variations were evident throughout the diamicton with a slight tendency for the dark brown-stained specimens to be more common at depth. Shell fragments are also present in the unit and are most common in the lowest 10 cm sampled. Radiocarbon ages on bones from this unit range from 18.0 to 16.3 ka B.P. (n 6), while U-series ages on two dripstone fragments are 16 and 14 ka (Tables I and III). A SEM analysis of quartz and feldspar grains and their surface morphologies from the matrix showed little variation throughout the excavated material. Grains are angular and conchoidally fractured, with some showing evidence of crushing (Figure 7, cf Van Hoesen & Orndorff, 2004) Unit I-1 records ice-free conditions from 18.0 to 16.3 ka B.P. (Tables I and III; Figure 6). It represents either the paleo-floor of the cave or sediments from the floor of the cave that were remobilized during the initial stages of ice cover. The range in ages, as well as the bone colors and their weathering throughout the unit, indicates that there has been some mixing, through either bioturbation or colluviation. We favor colluviation for the genesis of this unit; bone concentrations do not decrease with depth, no burrowing structures were observed, and faint bedding is present. Quartz grains show little sign of transport. Although some of the grains are subangular with slight rounding of the edges, they bear none of the common characteristics of grains deposited in subaqueous or aeolian environments, notably rounding and smooth to frosted surfaces (Krinsley and Doornkamp, 1973). This sediment may have been emplaced by a previous glaciation close to the entrance to the cave, which DOI: 10.1002/GEA


315

AL-SUWAIDI ET AL. Table II. Animal species from Port Eliza Cave. Taxon Pre-LGM Fauna Fish Onchorhynchus sp. Oncorhynchus ?clarkii Gasterosteus sp., cf. G. aculeatus Hexagrammidae, indet. Theragra chalcogramma Pleuronectiformes, indet. Hemilepidotus sp. ?Cottidae, indet. Microgadus proximus Amphibians Bufo boreas Birds Gavia sp., cf. G. stellata Alcidae, indet. cf. Phalacrocorax sp. cf. Anas sp. cf. Eremophila alpestris Passerculus sandwichensis additional bird species Mammals Microtus townsendii M. longicaudus Phenacomys intermedius Marmota sp., cf. M. caligata Martes americana ssp., cf. M. a. nobilis Oreamnos americanus Carnivora, indet. Mollusks Mytilus sp. Balanus sp. Postglacial Fauna Mammals Peromyscus sp. Oreamnos americanus

Common name

salmon ?cutthroat trout threespine stickleback greenling pollock flatfish Irish lord ?sculpin tomcod western toad small loon (?red-throated) small alcid cormorant ?small duck ?horned lark savannah sparrow

Townsend’s vole long-tailed vole heather vole marmot, alpine group large (?noble) marten mountain goat ?canid (tooth mark evidence) mussel barnacle

mouse mountain goat

would explain the presence of striated and glacially crushed clasts. With the onset of late Wisconsinan glaciation, the sediment and the bones on its surface were transported a short distance into the cave via small sediment gravity flows. This hypothesis explains the crude bedding and finer-grained matrix of the upper 15 cm of the unit and is supported by the intact character of many of the bones. Unit I-2 is well stratified, comprising predominantly thinly bedded to finely laminated silt and clay, with interbeds of massive and convoluted clay and rare sand partings (Figure 5). The stratification varies from a massive blue-gray clay, approxi316


DOI: 10.1002/GEA

DOI: 10.1002/GEA

Unit I-4 Stalagmite Unit I-3 Stalactite Unit I-1 Dripstone fraction Unit I-1 Diamicton fraction Unit I-1 Dripstone fraction Unit I-1 Diamicton fraction 55.844 0.303 540.26 0.01

b

451.11 3.31

13.837 0.078 17.436 0.119 11.091 0.079

TIMS

b

TIMS

TIMS

TIMS

Method

U (ppb)

238

987.10 67.49

11.885 0.094

764.80 58.62

897.284 6.703 265.108 1.896 8.965 0.072

Th (ppt)

232

1.20 0.05

3.081 0.033

1.28 0.05

5.033 0.0287 4.903 0.038 3.786 0.106

U/238U activity ratio

234

0.90 0.05

0.159 0.006

0.95 0.06

0.047 0.002 0.029 0.001 0.274 0.019

Th/234U activity ratio

230

1.80 0.15

7.014 0.250

2.19 0.20

11.023 0.354 28.383 0.512 3.921 0.256

Th/232Th activity ratio

230

12.80 0.93

44.239 0.616

8.70 0.70

237.198 2.517 985.592 11.301 14.313 0.417

U/232Th activity ratio

234

10.66 0.79

14.36 0.138

6.79 0.56

47.130 0.441 201 1.985 3.781 0.040

18.43 0.71

5.15 0.06 3.17 0.06 33.48 2.69

14.09 1.15

16.50 1.75

U/232Th Age Age activity (ka) (ka) ratio Uncorrected Corrected

238

The dripstone subsamples were corrected for the detrital contribution for U and Th isotopes using the measured values of U and Th in the diamicton. Alpha spectrometry.

b

a

MHA02-M7

MHA02-M7

MHA02-M1

MHA02-M1

BCW04-01

MHAOS-01

Sample

Subsample and location

Table III. Uranium series ages from dripstone from the Interior Cave Complex.



317

AL-SUWAIDI ET AL.

Figure 7. SEM images of sand grains from Unit I-1. Note the conchoidal fracturing and angular surface textures.

mately 18 cm thick, to light brown laminations of less than 1 mm. Individual laminations vary spatially from normal to reverse grading and locally display evidence of soft-sediment deformation. The unit is 2 m thick (cumulative thickness of 3.3 m), and its lower contact drapes the irregular surface of Unit I-1 and contains isolated larger clasts. The grain size of samples averages 83% clay. Unit I-2 contains numerous small faults along with a larger fault parallel to that of the cave bedrock, and two ice-wedge pseudomorphs. Unit I-2 can be divided into two packages of fine sediments, one in the lower 25 cm, dominated by massive clay, and above this, predominantly very pale brown interlaminated clay and silt. The lower one consists primarily of massive, dark gray clay (92% clay, modal diameter 0.64 µm) that drapes irregular clasts at the top of Unit I1. Although zones of diffuse, discontinuous laminations were noted, they are more weakly expressed than in the rest of the unit. In the lower 15 cm, two distinctive, bluish-black clay layers, each ~1–2 cm thick, are in sharp contact with the more massive clay. These two layers are ~50% clay with 32% of very fine silt (modal diameter 4.59 µm). A fault plane, which trends approximately parallel to the trend of the fault along which the cave formed, cuts the unit. The fault plane strikes roughly north–south and displays slickensides. Above the massive clay, interlaminated fine silt and clay (97% clay) predominate, consisting primarily of creamy light brown to blue-gray laminations, and lesser dark gray laminae 1 mm in thickness. Rare laminations display reddish-brown mottling. There are microscale deformation structures, such as flame structures, rare ripup clasts, and small- to large-scale convolute bedding. There is no evidence of ripples or other large-scale physical structures. In places, the laminations appear to grade from light gray to dark gray terminating at a sharp contact and then a return 318


DOI: 10.1002/GEA


to light gray. This was confirmed in thin-section analysis, which showed a recurring pattern of normally graded laminations characterized by light brown fine silt or clay passing upward into dark gray clay and capped by a sharp upper contact. These beds contained numerous normal faults with displacements of about a few cm to tens of cm, most evident within the upper portion of the unit in a package of distinct dark gray and dark brown laminations (Figure 5). In general, these faults dip toward the back of the cave and their dip and displacement decrease down-fault. The lower half of the unit displays rare faults. Unit I-2 contains two ice-wedge pseudomorph structures that occur at the same level. The larger and better developed of the two occurs in the lower 90 cm of the section. This structure tapers sharply from 110 cm at the top to only 17–25 cm below, and then more gradually to 4 cm wide near the base where it terminates at the contact with Unit I-1 (Figure 8). Contacts with the adjacent, normally faulted, laminated silt and clay are sharp and distinct. The lower half of the wedge is dominantly light brown fine-grained sand and silt with minor amounts of contorted laminated fines. Thin (1–2 mm) fractures comprising the same light brown fine sand extend outward and downward, up to 6 cm from this zone. Near the base of the wedge is a vertical lens of medium sand that is 3–4 cm wide at the contact with Unit I-1 and pinches and swells. It gradually narrows upward, terminating 35 cm from the base. Irregular lenses of similar fine sand are present up to approximately 10 cm above this but are not evident beyond this point. Above this, the light brown fine-grained sand and silt is increasingly mixed with finer-grained materials from adjacent to the wedge. These mixed beds are convoluted and brecciated, but include competent, normally faultbounded blocks, especially higher in the structure. The lower 20 cm of laminated sediments that drape the wedge form a graben-like structure. Three samples from Unit I-2 were analyzed by x-ray diffraction. They were selected to test for characteristics of distinctive colors noted. Sample M3 came from the lower massive gray clay; sample GS-9 is very dark grayish-brown clay; sample GS-8 is from one of the lower bluish-black beds (Figure 5). The samples yielded similar diffraction patterns, with peaks representing quartz, feldspar, and chlorite or corrensite. Samples GS-8 and GS-9 also contained calcite (Figure 9). Loss-on-ignition analysis showed low amounts of organic and inorganic carbon. Loss of organic carbon ranged from ~7 to ~2%, with an overall decline up through the section. This was also true of inorganic carbon values, which were generally 1%, but were as high as 2% in the lower 20 cm. Paleomagnetic analysis of Unit I-2 revealed high levels of remanent magnetization: 1–6 A/m, the highest for sediments ever recorded in the GSC–Pacific paleomagnetism laboratory and are normally only from outcrops hit by lightning. The paleomagnetic directions show smooth swings typical of a few thousand years’ geomagnetic secular variation, corroborating the limiting radiocarbon ages from the bounding units that indicate that the unit spans a considerable length of time. This secular variation record will be correlated with other raised sea-cave records. The laminated silt and clay of Unit I-2 represent deposition from suspension under very low-energy subaqueous conditions, such as a subglacial lake, as indicated by the clay-rich nature of the sediment, the strong remanent magnetization, DOI: 10.1002/GEA


319

AL-SUWAIDI ET AL.

Figure 8. Large ice wedge pseudomorph in Unit I-2. Trowel in all views is 26 cm. (a) View of majority of wedge. A graben structure is present in sediments that are draping wedge feature (g). (b) Middle portion of wedge with mix of fine-grained material, some convoluted, from Unit I-2 with a large lens of silty fine sand (l). Thin fractures of fine sand extend outward and downward from the margin of the wedge (arrowed). (c) Lower portion of wedge and the lower contact with the diamicton of Unit I-1. A vertical bed of fine sand extends from this contact to the top of the view. Thick dark clay layers deflect downward adjacent to the wedge. Thin fractures of fine sand extend outward and downward from the margin of the wedge (arrowed).

the clay mineralogy, and the detailed sedimentology. Grain-size analysis shows little variation between samples, and thin-section analysis shows no current-generated structures (consistent with deposition under calm conditions). Mineralogically, these sediments are predominantly made up of quartz and feldspar, and the lack of clay minerals suggests that the source was glaciogenic rock flour. Low organic and inorganic carbon values strengthen the argument for a glacial setting. The higher values of carbon noted in the lower 20 cm are likely due to the larger amounts 320


DOI: 10.1002/GEA


Figure 9. XRD spectrum of clay samples from Unit I-2 with major peaks of quartz, calcite, chlorite (or corrensite), and feldspar.

of carbon in the system during the initial stages of glaciation as the ice overrode soils and vegetation. The origin of the different colored laminations remains unclear, as they do not appear to reflect differences in grain size or clay mineralogy. The wedge features are interpreted to represent ice-wedge casts based on their characteristic tapering shape, the thin fractures extending out and down from the wedges, the associated normal faulting of adjacent sediments, and the variable character of material infilling the features. This interpretation suggests that the subglacial lake drained and sediments were subaerially exposed. Normal faulting of material draping the structure indicates melting of at least a portion of the wedge after burial. Two scenarios are proposed to explain periodic drainage: (a) oscillation of the ice front, and (b) changes in subglacial drainage. We favor the first explanation. The fault that parallels the cave and cuts the bulk of Unit I-2 could represent reactivation, following deglaciation, of the fault along which the cave formed. Reactivation could be caused by glacioisostatic adjustments or may have occurred during one or more earthquakes. The deformation evidenced by the thick zone of convolute material near the top of the sequence may have been the result of this event. The smaller normal faults either are related to the same event or are the result of differential compaction of sediments. Unit I-3 consists of overall massive oxidized clays with some convolute bedding (Figure 4). This forms the floor of the interior cave chambers. It has a sharp lower contact with the laminated silt and clay of Unit I-2. The partially disarticulated skeleton of a mountain goat was found at a depth of ~30 cm along with dripstone and DOI: 10.1002/GEA


321

AL-SUWAIDI ET AL.

charcoal fragments. Ages on the mountain goat and an overlying charcoal fragment are 12.3 ka B.P. and 9.5 ka B.P., respectively (Table I). Unit I-3 is interpreted to be a postglacial deposit formed as a result of resedimentation of fines from Unit I-2 in ponded water. The sediments are oxidized due to periodic drying of the cave floor. Dripstone fragments and soda straws have accumulated from the roof. Unit I-4 consists of modern cave sediments, including stalagmites, stalactites, other flowstones and includes rubble near the back of the cave entrance complex. The base of the dripstone unit has a lower sharp contact with the oxidized clay of Unit I-3. Several stalagmites of this unit were found around the area of the excavation. U-series ages on one of these stalagmites and on a stalactite fragment from base of the unit yielded ages of 5.1 ka and 3.2 ka, respectively. Sediments of the Cave Entrance Complex Unit E-1 is a matrix-supported diamicton that occurs at the base of the 5-m vertical section near the cave entrance (Figures 4 and 5). It is 0.5 m thick and extends an unknown distance into the cave. The lower contact is not visible. The diamicton contains large subrounded to rounded boulders, the largest 1.5 m. The largest clasts consist of cave rock, whereas pebbles and cobbles are dominantly exotic lithologies. The matrix is 88% coarse silt or finer. An SEM analysis of sand grains showed angular clasts with conchoidal fracturing. The matrix and surface contains calcium carbonate precipitation. Unit E-1 is interpreted to be till emplaced by ice that blocked the cave entrance. The larger blocks of cave material were likely pushed ahead or incorporated into the ice. Conchoidally fractured, angular grains are similar in morphology to those within Unit I-1 (Figure 7). The presence of this ice would have allowed the cave to fill with water and, therefore, the unit is likely correlative with the laminated silt and clay of Unit I-2. Unit E-2 is a clast-supported diamicton that forms the floor of the modern cave entrance. It comprises the upper 4 m of the vertical section at the cave entrance (Figure 4) and is exposed at the surface in the outer cave complex. It has an irregular, sharp basal contact with Unit E-1. Unit E-2 largely comprises openwork angular cave-rock blocks, some more than 1.5 m across. Fine sand and silt with rare bone fragments form the matrix at the surface. Unit E-2 is interpreted to be a postglacial deposit, probably the result of rockfall from the roof of the cave caused by freeze thaw and seismic events. It is, therefore, correlative with the oxidized clays of Unit I-3 and the speleothems of Unit I-4. Faunal Assemblage More than 4000 bones were recovered from Unit I-1 and represent a diverse vertebrate fauna (Table II). Although there is evidence for redeposition with limited transport, the fauna is interpreted to indicate environments near the cave between 18 and 16 ka B.P. The assemblage is dominated by small mammals and birds. The mammals include three species of vole and an alpine marmot; but there are also 322


DOI: 10.1002/GEA


larger fragments from at least one larger species, notably a mountain goat. Various fish species occur throughout the sequence along with mussel and barnacle shells, which appear more common near the base. Several mammal species strongly suggest an open landscape. Among these are three species extirpated from Vancouver Island (heather vole, long-tailed vole, mountain goat), and possibly an extinct species (noble marten?). Of the three voles, only Townsend’s vole occurs on Vancouver Island today, principally in nonforested habitats from sea level to 1740 m (Nagorsen, 2004). The long-tailed vole ranges widely from wet meadows through shrub thickets to forests. The heather vole typically inhabits open coniferous forests but also occurs in the alpine environments that are also inhabited by marmots. The alpine marmot group typically occurs in a treeless habitat. The modern Vancouver Island marmot is the most endangered mammal in Canada, restricted to small areas of grass-forb meadows above 700 m (Bryant and Janz, 1996; Nagorsen et al., 1996). Mountain goats similarly favor meadows and cliffs in the alpine zone, and though they occurred on Vancouver Island after the LGM, they were extirpated by the end of the Pleistocene, likely because of habitat restriction (Nagorsen and Keddie, 2000). The passerine birds (savannah sparrow and possible horned lark) are also consistent with open conditions and even disturbed landscapes, as would be expected along a storm-prone coastline. The Port Eliza species most suggestive of forest is the marten, which today is strongly associated with dense coniferous forest. However, late Pleistocene marten remains in Beringia occur in settings interpreted to include mixed taiga and steppe (Graham and Graham, 1994). Notably, the Port Eliza specimen is larger than most modern marten and is similar in size to the extinct noble marten (M. nobilis) and the modern marten on the Queen Charlotte Islands (M. a. nesophila). Taxonomy of the noble marten is unclear. Some authors view it as a subspecies of Martes americanus (Youngman and Schueler, 1991), while others interpret it as a separate species of wide habitat tolerance. Grayson (1984; personal communication, 27 Feb. 2003) reported M. nobilis (which he took as a valid species) from Holocene cave deposits in Nevada and Idaho that lacked boreal elements in the fauna; consequently, the marten remains cannot be taken as diagnostic of forested conditions. On balance, the fauna indicates an open landscape, possibly with patches of trees. Fish remains indicate that sea level was close to the cave entrance and document a varied fauna capable of supporting terrestrial predators and scavengers. Except for salmon (Oncorhynchus sp.) and trout, the taxa are completely marine (greenling, pollock, flatfish, Irish lord, tomcod, and sculpin), suggesting that the shoreline was close enough for a predator to bring such material back to the cave. The presence of abundant shell fragments, including mussels and barnacles, confirms this. Salmon and trout could also represent capture or scavenging from local streams. Pollen and Spores Pollen and spores were only recovered from Units I-1 and I-3 (Figure 10). Pollen abundance is low with scarcely 100 palynomorphs in several of the samples. All samples are dominated ( 70%) by non-arboreal pollen (NAP). The NAP consists mainly of DOI: 10.1002/GEA


323

Figure 10. Pollen from Unit I-1 and Unit I-3.

AL-SUWAIDI ET AL.

324


DOI: 10.1002/GEA


grasses and Selaginella (a diminutive spikemoss), with lesser amounts of Asteraceae, Apiaceae (mostly Heracleum lanatum—cow parsnip), and Campanula (harebell). The lowest sample in Unit I-1 has abundant Cyperaceae (sedges), and the highest sample in the sequence (Unit I-3) has relatively abundant Ericaceae (heath). Arboreal pollen (AP) consists mainly of Pinus (pine) and Tsuga heterophylla (western hemlock). The assemblages indicate open landscapes at the time of deposition of Units I-1 and I-3. No modern samples from the region, not even those from moss polsters, have such high NAP values dominated by the taxa identified in this study (Hebda, 1983; Hebda and Allen, 1993, Allen et al., 1999 and references therein). The abundance of grasses, Caryophyllaceae (pink family) and Selaginella indicates a landscape covered largely in low-growing, herbaceous species. The Selaginella spore-type is like that of S. densa or S. wallacei, both likely indicators of dry conditions, as is Campanula. More productive and diverse plant communities on the landscape are indicated by pollen of the Apiaceae. Although much of the pollen indicates aridity, Heracleum lanatum and Cyperaceae, especially near the base of Unit I-1 and the top of I-3, suggest that more mesic communities were also present. The low values of shrubs, such as Alnus (alder) and Salix (willow), suggests that few woody plants grew in the area during these intervals. Paleoenvironmental Reconstruction The faunal and floral assemblages of Unit I-1 provide a picture of the pre-LGM environment ca 18 to 16 ka B.P. The dominantly herbaceous landscape suggests a cold, dry climate. The arboreal pollen, mainly from pine, can easily be accounted for by long-distance dispersal. The occurrence of some trees, however, should not be ruled out as pine is present at Tofino at 16.7 ka B.P. (Figure 1, Blaise et al., 1990) and a recent date from the same site indicates the presence of western fir at 16,430 40 (CAMS 111667). This landscape supported a wide variety of small mammals and birds and at least one larger species, mountain goat. Local streams contained trout and possibly runs of salmon. The sea was close to the cave, and contained salmon and other shallow-water fish, as well as shellfish. The fact that relative sea level was close to the cave at a time when eustatic sea level would have placed the shoreline ~15 km away points to significant isostatic depression. Thus, Vancouver Island supported large glaciers that would likely have been visible from the shore. The cold, dry environment represented at Port Eliza has not been recognized in the region during the interval represented by the sediments. Unit I-1 assemblages are contemporaneous with to slightly younger than sediments of the Port Moody Interstade of the Fraser Lowland of British Columbia (Figures 1 and 2), yet the pollen assemblages of the two sites are not at all similar. The Fraser Lowland supported an open spruce-fir forest representing a cool to cold but moist environment (Hicock et al., 1982); whereas the Port Eliza area was cold and probably had only scattered trees. If the two sites are contemporaneous, a steep climatic gradient existed between them. One would expect a more temperate climate at Port Eliza than in the Fraser Lowland, because the cave is much nearer the moderating effects of the open Pacific Ocean. Based on the range in the radiocarbon ages, the pollen spectra could reflect conditions closer to ~16 ka DOI: 10.1002/GEA


325

AL-SUWAIDI ET AL.

B.P., which would be after the Port Moody Interstade and closer to the LGM. Postglacial vegetation on the Queen Charlotte Islands at ~15 ka B.P. consisted of treeless tundra dominated by grasses, sedges, Caryophyllaceae, and other herbaceous types (Mathewes, 1989). The assemblage differs from Port Eliza where Selaginella was abundant as well. The laminated silt and clay (Unit I-2) along with the till (Unit E-1) noted at the entrance of the cave mark the local onset of the Fraser Glaciation after 16 ka B.P. XRD and SEM data indicate a glacial source for these sediments. Paleomagnetic data show secular variation pointing to a long period of deposition in quiet water. These observations confirm the outer coast was glaciated, and the bounding dates constrain the local start and end of the LGM. The age on the mountain goat in Unit I-3 provides a minimum age for local deglaciation and is similar to ages on two other mountain goats of 12.3 and 12.5 ka B.P. from northern Vancouver Island (Nagorsen and Keddie, 2000). Older ages of 13.1 and 13.6 ka B.P. on basal bulk sediments from lakes suggest deglaciation could have been earlier (Hebda, 1983). Similar to Unit I-1, the I-3 pollen suggests open parkland with dry herbaceous shrub cover with rare stands of pine and alder. Northern Vancouver Island would have been viable for mountain goats by ~12.5 ka B.P., which either migrated from refugia to the north on the now submerged continental shelf (Hetherington et al., 2004) or from the south (Nagorsen and Keddie, 2000). CONCLUSION Implications for Human Migration Data presented here provide important clues into the viability of the “coastal migration” hypothesis. Until recently, the earliest known British Columbia coastal archaeological site was Namu, on the central coast alongside Fitzhugh Sound, dated at 9.7 ka B.P. (Carlson, 1996a, 1996b). More recent work at Hunter Island, 20 km to the west, revealed cultural deposits dated to 9.9 ka B.P. (Cannon, 2000). The oldest securely dated archaeological site on Haida Gwaii is Kilgii Gwaay, dated to 9.4 ka B.P. (Fedje et al., 2001). A karst cave on Prince of Wales Island, southeastern Alaska, contained a culturally modified bone from a terrestrial mammal that yielded an age of 10.3 ka B.P. (Dixon, 1999). Isotopic studies of human remains from the same site indicate a dominantly marine diet at 9.2 ka B.P. (corrected reservoir age). Older postglacial human occupations are likely to be found in all of these areas, especially on Prince of Wales Island where trace element analysis of obsidian artifacts indicates that an extensive trade network had already been established by ~10 ka (Fedje et al., 2004). People were present to the north, in eastern Beringia by 12 ka B.P., if not much earlier, though questions of context and association remain (Cinq-Mars, 1979; Cinq-Mars and Morlan, 1999; Hamilton and Goebel, 1999; Dixon, 1999). In adjacent U.S. states to the south of British Columbia, the earliest well-dated human occupation is at the Cooper’s Ferry site in the lower Salmon River, Idaho, with a cultural stratum dated on charcoal to 11.4 ka B.P. A cache pit extending down from this level gave ages of 11.4 and 12.0 ka B.P. and contained four stemmed projectile points (Davis and Sisson, 1998). Several other sites of the stemmed-point tradition in the Great Basin and Snake 326


DOI: 10.1002/GEA


River Plain appear to be of the same age (Bryan and Tuohy, 1999). Occurrences associated with the last Glacier Peak tephra (~11.25 ka B.P.) include a stemmed point embedded in redeposited but pure tephra at Pilcher Creek, northeast Oregon (Brauner, 1985) and a cache of Clovis fluted projectile points (one resting on the surface of the tephra) at the Richey–Roberts site near Wenatchee, Washington (Mehringer and Foit, 1990). Each occurrence is interpreted to immediately postdate the ashfall event. The Indian Sands site on the Oregon coast includes a human occupation associated with a paleosol dated to 10.4 ka B.P. (Davis, 2003; Davis et al., 2004). If the evidence for human activity is valid, the 12 ka B.P. Manis mastodon from Sequim, on the Olympic Peninsula of Washington, would be the earliest record of human occupation in the Pacific Northwest (Gustafson et al., 1979; Dixon, 1999). Dixon (1999) has argued that North America was first colonized by people who used watercraft to travel down the west coast ~13.5 ka B.P., supporting the coastal hypothesis of Fladmark (1983). An initial criticism of the coastal-migration theory is that the majority of the route was barred by the extensive Cordilleran Ice Sheet, which advanced to the shelf edge (Prest, 1969). However, this study and recent work (Blaise et al., 1990; Mann and Peteet, 1994; Heaton et al., 1996; Josenhans et al., 1997; Barrie and Conway, 1999; Fedje and Josenhans, 2000; Hetherington et al., 2003; Ward et al. 2003; Hetherington et al., 2004) indicate that the LGM was less extensive and diachronous along the coast, allowing the possibility of human migration. Evidence suggests that glaciers in southern Alaskan coastal areas were retreating by or before ~16 ka B.P., following an earlier expansion at ~23 ka B.P. (Mann and Peteet, 1994). In contrast, our data show that the LGM on northern Vancouver Island occurred after ~16 ka B.P., raising the possibility that there was a brief window of opportunity for migration along a relatively ice-free Northwest Coast just before this time. Refugia likely existed in portions of southeast Alaska, supporting terrestrial mammals as well as seals (Heaton and Grady, 2003). Glaciers on Haida Gwaii (Queen Charlotte Islands) were at their maximum prior to ~16 ka B.P. and were retreating by ~15 ka B.P. (Clague et al., 2004). Ice-free areas may have existed along the west coast of Haida Gwaii and on the now submerged continental shelf (Fedje et al., 2004; Lacourse et al., 2003; Ramsey et al., 2004). The presence of brown bear at 14.3 ka B.P. (Ramsey et al., 2004) implies adjacent coastal refugia, because glaciers would have blocked migration from the south along Vancouver Island. Human migrants would still have had to deal with extensive sea ice (Brigham-Grette et al., 2004) and would have had to skirt major ice lobes along the way (e.g., Hecate Strait, Dixon Entrance), but Dixon’s (1999) strong argument for use of watercraft by the earliest colonizing populations applies here. If, as he argues, they were using watercraft by ~13.5 ka B.P., they also could have done so 2500 years earlier. Humans would have found a habitable environment on the west coast of Vancouver Island before ~16 ka B.P. The fauna appears to have been dominated by small species, but the presence of mountain goat indicates an environment viable for ungulates. The diverse assemblage of fish also signifies a significant source of protein from a nearby rich marine environment. The presence of salmon suggests seasonal anadromous fish runs in local streams. There were likely sea bird colonies along steeplands on the coast. This suggests that the environment, although less productive than today, could DOI: 10.1002/GEA


327

AL-SUWAIDI ET AL.

have supported small groups of humans migrating down the coast, who would have been able to exploit both marine and terrestrial resources. SUMMARY Port Eliza Cave deposits provide multiple lines of evidence that elucidate the character of the LGM on the west coast of Vancouver Island. The sediments that accumulate in raised sea caves can preserve faunal and floral remains and record the sedimentological evidence of glaciation and deglaciation. Data sets such as these are critical in any discussion of coastal human migration as they establish a glacial chronology and ascertain whether the environment was viable for humans. Here, we show that the site was glaciated after ~16 ka B.P., and could have supported humans before this date. The presence of a diverse flora and fauna near the cave shows that food resources were adequate to support humans. Pollen indicates a cold, dry, open landscape. The presence of mountain goat indicates an environment capable of sustaining large herbivores. Fish, such as salmon, and shellfish offered an important source of protein for migrants. The onset of glaciation is signaled by deposition of laminated silt and clay. Deglaciation occurred before ~12.5 ka B.P. The implication, therefore, is that this section of the coastal migration route would have been unavailable to humans for no more than 3000 years; during this time, glaciers presumably covered the entire ( 400 km) west coast of Vancouver Island, likely out to the shelf edge, forming a significant impediment, even for people with watercraft. Findings at Port Eliza Cave are thus of great importance to hypotheses of human entry into the New World. Clearly, the west coast was blocked by Cordilleran ice only briefly ( 3000 years), and the glacial maximum was diachronous along the coast. If the humans who occupied Monte Verde, Chile, by ~12.5 ka B.P., came from northeast Asia, then there are three possible scenarios for west coast travel: after 13 ka B.P, shortly before 16 ka B.P., and well prior to the LGM (as recently championed by Madsen, 2004). The post-13 ka B.P. scenario requires rapid southward movement, which is plausible given the likelihood that watercraft could be used for part of the journey. The longer timescales of the pre-16 ka B.P. window of opportunity and the earlier scenario allow several millennia or more for people to reach Chile. Many coastal sites from such an early southward movement either would have been overridden by ice at the LGM, destroyed by wave action, or now lie below sea level, still awaiting discovery. Even south of the Cordilleran ice sheet, any record of the movements of marine-adapted peoples would suffer from the last two problems, although progress is being made (Davis, this issue; Punke and Davis, this issue). Given that people apparently migrated across open ocean to Australia by at least ~38 ka B.P. and perhaps much earlier (Allen and Holloway, 1995; Mulvaney and Kamminga, 1999), the difference between 16 and 13 ka B.P. is trivial in terms of the use of watercraft and associated technologies; one time is just as plausible as the other. What is needed now is a detailed reconstruction and summary of coastal paleogeography from Alaska to Washington 17 to 16 ka B.P., a time once ignored as icebound and inapplicable to the question of human migration. 328


DOI: 10.1002/GEA

PORT ELIZA CAVE DEPOSITS This paper is a partial requirement for the M.Sc. thesis of Majid Al-Suwaidi supervised by B.C. Ward in the Earth Sciences Department of Simon Fraser University. The authors extend special thanks to the Ehattesaht Nation, specifically Sharon Elshaw, Lyle Billy, and Victoria Wells, for allowing this research to be conducted on their traditional territories. Western Forest Products and Taylor Contracting kindly provided logistical support. Funding was provided by a NSERC grant to Dr. B.C. Ward. Faunal material was identified through comparison with faunal collections at Simon Fraser University, Royal British Columbia Museum, and the University of Victoria. Dr. Paul Matheus, University of Alaska Fairbanks, confirmed the Townsend’s vole identification. S. Bains, L. Billy, S. Elshaw, C. Kowalchuk, A. Miskovic, S. Todd, and S. Villeneuve provided able assistance in the field. Thanks are also due to Mathew Plotnikoff, at Simon Fraser University, for guidance with grain size analysis, Dr. Duane Froese at the University of Alberta for processing of samples through the sedigraph, and Dr. Brian Menounos for help with thin-section preparation at University of British Columbia. The manuscript benefited greatly from comments from Loren G. Davis and an anonymous journal reviewer, as well as from thesis committee members Drs. John Clague, Lionel Jackson, and James MacEachern.

REFERENCES Adovasio, J.M., & Pedler, D.R. (1997). Monte Verde and the antiquity of humankind in the Americas. Antiquity, 71, 573–580. Allen, J., & Holloway, J.S. (1995). The continuity of Pleistocene radiocarbon determinations in Australia. Antiquity, 69, 101–112. Allen, G.B., Brown, K.J., & Hebda, R.J. (1999). Surface pollen spectra from South Vancouver Island, British Columbia, Canada. Canadian Journal of Botany, 77, 786–789. Barrie, J.V., & Conway, K.W. (1999). Late Quaternary glaciation and postglacial stratigraphy of northern Pacific margin of Canada. Quaternary Research, 51, 113–123. Blaise, B., Clague, J.J., & Mathewes, R.W. (1990). Time of maximum late Wisconsinan glaciation, West Coast of Canada. Quaternary Research, 47, 140–146. Brauner, D. (1985). Early human occupation in the uplands of the Southern Plateau: Archaeological excavations at the Pilcher Creek Site, Union County, Oregon. Corvallis, OR: Department of Anthropology, Oregon State University. Brigham-Grette, J., Lozhkin, A.V., Anderson, P.M., & Glushkova, O.Y. (2004). Environmental conditions in Northeast Asia and Northwestern North America. In D.B. Madsen (Ed.), Entering America: Northeast Asia and Beringia before the Last Glacial Maximum (pp. 29–61). Salt Lake City, UT: The University of Utah Press. Bryan, A.L., & Tuohy, D.R. (1999). Prehistory of the Great Basin/Snake River Plain to about 8,500 years ago. In R.A. Bonnichsen & K.L. Turnmire (Eds.), Ice Age peoples of North America: Environments, origins, and adaptations of the first Americans (pp. 249–263). Corvallis, OR: Oregon State University Press. Bryant, A.A., & Janz, D.W. (1996). Distribution and abundance of Vancouver Island marmots (Marmota vancouverensis). Canadian Journal of Zoology, 74, 667–677. Cannon, A. (2000). Settlements and sea-levels on the central coast of British Columbia: Evidence from shell midden cores. American Antiquity, 65, 67–77. Carlson, R.L. (1996a). Introduction to early human occupation in British Columbia. In R.L. Carlson & L. Dalla Bona (Eds.), Early human occupation in British Columbia (pp. 3–10). Vancouver, BC: University of British Columbia Press. Carlson, R.L. (1996b). Early Namu. In R.L. Carlson & L. Dalla Bona (Eds.), Early human occupation in British Columbia (pp. 83–102). Vancouver, BC: University of British Columbia Press. Cinq-Mars, J. (1979). Bluefish Cave I: A late Pleistocene eastern Beringian cave deposit in the northern Yukon. Canadian Journal of Archaeology, 3, 1–32. Cinq-Mars, J., & Morlan, R.E. (1999). Bluefish Caves and Old Crow Basin: A new rapport. In R.A. Bonnichsen & K.L. Turnmire (Eds.), Ice Age peoples of North America: Environments, origins, and adaptations of the first Americans (pp. 200–212). Corvallis, OR: Oregon State University Press. Clague, J.J., Mathewes, R.W., & Agar, T.A. (2004). Environments of Northwestern North America before the Last Glacial Maximum. In D.B. Madsen (Ed.), Entering America: Northeast Asia and Beringia before the Last Glacial Maximum (pp. 63–94). Salt Lake City, UT: The University of Utah Press.

DOI: 10.1002/GEA


329

AL-SUWAIDI ET AL. Davis, L.G. (Nov. 2003). Geoarchaeology of late Pleistocene occupation at the Indian Sands Site, southern Northwest Coast. Paper presented at the Geological Society of America, 2003 Annual Meeting, Seattle, WA. Davis, L.G. (2006). Geoarchaeological insights from Indian Sands, a late Pleistocene site on the southern Northwest coast, USA. Geoarchaeology, 21, 351–361. Davis, L.G., & Sisson, D.A. (1998). An early stemmed point cache from the lower Salmon River Canyon of westcentral Idaho. Current Research in the Pleistocene, 15, 12–13. Davis, L.G., Punke, M.L., Hall, R.L., Fillmore, M., & Willis, S.C. (2004). Evidence for a late Pleistocene occupation of the southern Northwest Coast. Journal of Field Archaeology, 29, 1–10. Dillehay, T.D. (1997). Monte Verde: A late Pleistocene settlement in Chile. Volume 2: The archaeological context and interpretation. Washington, DC: Smithsonian Institution Press. Dixon, E.J. (1999). Bones, boats, & bison: Archeology and the first colonization of Western North America. Albuquerque, NM: University of New Mexico Press. Dyke, A.S. (2004). An outline of North American deglaciation with emphasis on central and northern Canada. In J. Ehlers & P.J. Gibbard (Eds.), Quaternary glaciation—Extent and chronology. Part II: North America (pp. 373–424). Amsterdam, The Netherlands: Elsevier. Edwards, R.L., Chen, J.H., & Wasserburg, J.G. (1987). 238U-234U-230Th-232Th systematics and precise measurement of time over the last 500,000 years. Earth and Planetary Science Letters, 81, 175–192. Faegri, K., & Iversen, J. (1989). Textbook of pollen analysis (4th ed.). Chichester, UK: Wiley. Fedje, D.W., & Josenhans, H.W. (2000). Drowned forest and archaeology on the continental shelf of British Columbia, Canada. Geology, 28, 99–102. Fedje, D.W., Wigen, R.J., Mackie, Q., Lake, C., & Sumpter, I. (2001). Preliminary results from investigations at Kilgii Gwaay: An Early Holocene archaeological site on Ellen Island, Haida Gwaii, British Columbia. Canadian Journal of Archaeology, 25, 98–120. Fedje, D.W., Mackie, Q., Dixon, E.J., & Heaton, T.H. (2004). Late Wisconsin environments and archaeological visibility on the Northern Northwest Coast. In D.B. Madsen (Ed.), Entering America: Northeast Asia and Beringia before the Last Glacial Maximum (pp. 97–138). Salt Lake City, UT: The University of Utah Press. Fladmark, K.R. (1979). Routes; alternate migration corridors for early man in North America. American Antiquity, 44, 55–69. Fladmark, K.R. (1983). Times and places: Environmental correlates of mid- to late Wisconsinan human population expansion in North America. In R. Shutler (Ed.), Early man in the New World (pp. 13–41). Beverly Hills, CA: Sage. Gee, G.W., & Bauder, J.W. (1986). Particle-size analysis. In A. Klute (Ed.), Methods of soil analysis, Part 1: Physical and mineralogical methods. Madison, WI: Society of Agronomy/Soil Science Society of America. Graham, R.W., & Graham, M.A. (1994). Late Quaternary distribution of Martes in North America. In S.W. Buskirk, A.S. Harestad, M.G. Raphael, & R.A. Powell (Eds.), Martens, sables, and fishers. Biology and conservation (pp. 26–58). Ithaca, NY: Cornell University Press/Comstock Publishing Associates. Grayson, D.K. (1984). Time of extinction and nature of adaptation of the noble marten, Martes nobilis. In H.H. Genoways & M.R. Dawson (Eds.), Contributions in Quaternary vertebrate paleontology: A volume in memorial to John E. Guilday (pp. 233–240). Pittsburgh, PA: Carnegie Museum of Natural History. Gustafson, C.E., Gilbow, D., & Daugherty, R.D. (1979). The Manis mastodon site: Early man on the Olympic Peninsula. Canadian Journal of Archaeology, 3, 157–164. Hamilton, T.D., & Goebel, T. (1999). Late Pleistocene peopling of Alaska. In R.A. Bonnichsen & K.L. Turnmire (Eds.), Ice Age peoples of North America: Environments, origins, and adaptations of the first Americans (pp. 156–199). Corvallis, OR: Oregon State University Press. Heaton, T.H., & Grady, F. (2003). The late Wisconsin vertebrate history of Prince of Wales Island, Southeast Alaska. In B.W. Schubert, J.I. Mead, & R.W. Graham (Eds.), Ice Age cave faunas of North America (pp. 17–53). Bloomington, IN: Indiana University Press. Heaton, T.H., Talbot, S.L., & Shields, G.F. (1996). An ice refugium for large mammals in the Alexander Archipelago, Southwestern Alaska. Quaternary Research, 46, 186–192. Hebda, R.J. (1983). Lateglacial and postglacial vegetation history at Bear Cove Bog, northeast Vancouver Island, British Columbia. Canadian Journal of Botany, 61, 3172–3192. Hebda, R.J., & Allen, G.B. (1993). Modern pollen spectra from west central British Columbia. Canadian Journal of Botany, 71, 1486–1495.

330


DOI: 10.1002/GEA

PORT ELIZA CAVE DEPOSITS Hebda, R.J. North, M.E.A., & Rouse, G.E. (2002). Pollen and Spores of British Columbia. Unpublished Manuscript. 60 p. at Royal British Columbia Museum, Victoria, B.C. Canada. Hetherington, R., Barrie, J.V., Reid, R.G.B, Macleod, R., Smith, D.J., James, T.S., & Kung, R. (2003). Late Pleistocene coastal paleogeography of the Queen Charlotte Islands, British Columbia, Canada, and its implications for terrestrial biogeography and early post glacial human occupation. Canadian Journal of Earth Sciences, 40, 1755–1766. Hetherington, R., Barrie, J.V., Reid, R.G.B, Macleod, R., & Smith D.J. (2004). Paleogeography, glacially induced crustal displacement, and late Quaternary coastlines on the continental Shelf of British Columbia, Canada. Quaternary Science Reviews, 23, 295–318. Hicock, S.R., Hebda, R.J., & Armstrong, R.J. (1982). Lag of the Fraser glacial maximum in the Pacific Northwest: Pollen and macrofossil evidence from western Fraser Lowland, British Columbia. Canadian Journal of Earth Sciences, 19, 2288–2296. Hicock, S.R., & Lian, O.B. (1995). The Sister Creek Formation: Pleistocene sediments representing a nonglacial interval in southwestern British Columbia at about 18 ka. Canadian Journal of Earth Sciences, 32, 758–767. Jackson, L.E., Jr., & Wilson, M.C. (2004). The ice-free corridor revisited. Geotimes, 49, 16–19. JCPDS International Center for Diffraction Data. (1993). Mineral powder diffraction file data book. Swarthmore, PA: Author. Johnston, W.A. (1933). Quaternary geology of North America in relation to the migration of man. In D. Jenness (Ed.), The American aborigines, their origin and antiquity (pp. 9–45). Toronto, ON: University of Toronto Press. Josenhans H., Fedje D.W., Pientz, R., & Southon J.R. (1997). Early humans and rapidly changing Holocene sea levels in the Queen Charlotte Islands—Hecate Strait, British Columbia, Canada. Science, 277, 71–74. Krinsley D.H., & Doornkamp J.C. (1973). Atlas of quartz sand surface textures. Cambridge, UK: Cambridge University Press. Lacourse, T., Mathewes, R.W., & Fedje, D.W. (2003). Paleoecology of late-glacial terrestrial deposits with in situ conifers from the submerged continental shelf of western Canada. Quaternary Research, 60, 180–188. Larsen, E., Gulliksen, S., Lauritzen, S., Lie, R., Lovlie, R., & Mangerud, J. (1987). Cave stratigraphy in Western Norway; multiple Weichselian glaciations and interstadial vertebrate fauna. Boreas, 16, 267–292. Madsen, D.B. (2004). Colonization of the Americas before the Last Glacial Maximum: Issues and problems. In D.B. Madsen (Ed.), Entering America: Northeast Asia and Beringia before the Last Glacial Maximum (pp. 1–26). Salt Lake City: The University of Utah Press. Mandryk, C.A.S., Josenhans, H., Fedje, D.W., & Mathews, R.W. (2001). Late Quaternary paleoenvironments of Northwestern North America: Implications for inland and coastal migration routes. Quaternary Science Reviews, 20, 301–314. Mann D.H., & Peteet, D.M. (1994). Extent and the timing of the Last Glacial Maximum in Southwestern Alaska. Quaternary Research, 42, 136–148. Mathewes, R.W. (1989). Paleobotany of the Queen Charlotte Islands. In G.G.E. Scudder & N. Gessler (Eds.), The outer shores (pp. 75–90). Vancouver, BC: University of British Columbia. Mehringer, P.J., Jr., & Foit, F., Jr. (1990). Volcanic ash dating of the Clovis cache at East Wenatchee, Washington. National Geographic Research, 6, 495–503. Meltzer, D.J. (1997). Monte Verde and the Pleistocene peopling of the Americas. Science, 276, 754–755. Mulvaney, J., & Kamminga, J. (1999). The prehistory of Australia. Washington, DC: Smithsonian Institution Press. Nagorsen, D.W. (2004). The rodents and lagomorphs of British Columbia. Royal British Columbia handbook. Vancouver, BC: University of British Columbia Press. Nagorsen, D.W., & Keddie, G. (2000). Late Pleistocene mountain goats (Oreamnos americanus) from Vancouver Island: Biogeographic implications. Journal of Mammalogy, 81, 666–675. Nagorsen, D.W., Keddie, G., & Luszcz, T. (1996). Vancouver Island marmot bones in subalpine caves: Archaeological and biological implications (BC Parks, Occasional Paper No. 4). Victoria, BC: Ministry of Environment, Lands and Parks.

DOI: 10.1002/GEA


331

AL-SUWAIDI ET AL. Prest V.K. (1969). Retreat of Wisconsinan and recent ice in North America, Map 1257 A. Ottawa, Ontario: Geological Survey of Canada. Punke, M.L., & Davis, L.G. (2006). Problems and prospects in the preservation of late Pleistocene cultural sites in Southern Oregon coastal river valleys: Implications for evaluating coastal migration routes. Geoarchaeology, 21, 333–350. Ramsey, C.L., Griffiths, P.A., Fedje, D.W., Wigen, R.J., & Mackie, Q. (2004). Preliminary investigations of a late Wisconsinan fauna from K1 cave, Queen Charlotte Islands (Haida Gwaii), Canada. Quaternary Research, 62, 105–109. Smith, J.G. (2003). Aspects of the loss-on-ignition (LOI) technique in the context of clay-rich, glaciolacustrine sediments. Geografiska Annaler, 85, 91–97. Stuiver, M., & Reimer, P.J. (1993). Extended 14C database and revised CALIB radiocarbon calibration program. Radiocarbon, 35, 215–230. Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Burr, G.S., Hughen, K.A., Kromer, B., MacCormac, F.G., Plicht, J., & Spurk M. (1998). INTCAL98 radiocarbon age calibration 24,000–0 cal BP. Radiocarbon, 40, 1041–1083. Valen, V., Mangerud, J., Larsen, E., & Hufthammer, A.K. (1996). Sedimentology and stratigraphy in the cave Hamnsunhelleren, western Norway. Journal of Quaternary Science, 11, 185–201. Van Hoesen, J.G., & Orndorff, R.L. (2004). A comparative study on the micromorphology of glacial and nonglacial clasts with varying age and lithology. Canadian Journal of Earth Sciences, 41, 1123–1139. Ward, B.C., & Thomson, B. (2004). Late Pleistocene stratigraphy and chronology of lower Chehalis River valley, southwestern British Columbia: Evidence for a restricted Coquitlam Stade. Canadian Journal of Earth Sciences, 41, 881–895. Ward, B.C., Wilson, M.C., Nagorsen, D.W., Nelson D.E., Driver, J.C., & Wigen, B. (2003). Port Eliza cave: North American West Coast interstadial environment and implications for human migrations. Quaternary Science Reviews, 22, 1383–1388. Wilson, M.C., & Burns, J.A. (1999). Searching for the earliest Canadians: Wide corridors, narrow doorways, small windows. In R.A. Bonnichsen & K.L. Turnmire (Eds.), Ice Age peoples of North America: Environments, origins, and adaptations of the first Americans (pp. 213–248). Corvallis, OR: Oregon State University Press. Youngman, P.M., & Schueler, F.W. (1991). Martes nobilis is a synonym of Martes americana, not an extinct Pleistocene-Holocene species. Journal of Mammalogy, 72, 567–577.

Received May 18, 2004 Accepted for publication October 21, 2005

332


DOI: 10.1002/GEA