The Archaeological Heritage of Alberta’s Lower Athabasca Basin
Brian M. Ronaghan
Part One Postglacial Environments
1 A Tale of Two Floods
How the End of the Ice Age Enhanced Oil Sands Recovery—and Decimated the Fossil Record
James A. Burns and Robert R. Young
2 Glacial Geology and Land-Forming Events in the Fort McMurray Region
Timothy G. Fisher and Thomas V. Lowell
3 Raised Landforms in the East-Central Oil Sands Region
Origin, Age, and Archaeological Implications
Robin J. Woywitka, Duane G. Froese, and Stephen A. Wolfe
4 Kearl Lake
A Palynological Study and Postglacial Palaeoenvironmental Reconstruction of Alberta’s Oil Sands Region
Luc Bouchet and Alwynne B. Beaudoin
Part Two Human History
5 The Early Prehistoric Use of a Flood-Scoured Landscape in Northeastern Alberta
Grant M. Clarke, Brian M. Ronaghan, and Luc Bouchet
6 A Chronological Outline for the Athabasca Lowlands and Adjacent Areas
Brian O. K. Reeves, Janet Blakey, and Murray Lobb
8 The Early Human History of the Birch Mountains Uplands
John W. Ives
Part Three Lithic Resource Use
9 Beaver River Sandstone
Characteristics and Use, with Results of Heat Treatment Experiments
Eugene M. Gryba
10 The Organization of Lithic Technology at the Quarry of the Ancestors
Nancy Saxberg and Elizabeth C. Robertson
11 Microblade Technology in the Oil Sands Region
Distinctive Features and Possible Cultural Associations
Angela M. Younie, Raymond J. Le Blanc, and Robin J. Woywitka
Part Four Archaeological Methods
Investigative Approaches in the Athabasca Oil Sands
Gloria J. Fedirchuk, Jennifer C. Tischer, and Laura Roskowski
13 Cumulative Effects Assessment
Evaluating the Long-Term Impact of Oil Sands Development on Archaeological Resources
Brian M. Ronaghan
List of Contributors
Index of Sites
From its headwaters in Jasper National Park, Alberta’s Athabasca River travels some 1,500 miles northeastward across the province, its waters ultimately flowing into Lake Athabasca. The river drains a vast region—roughly 159,000 square kilometres—along the southern margin of Canada’s Boreal Plains ecozone, which encompasses much of central and northern Alberta. In contrast to the grasslands further south, the Boreal Plains ecozone is heavily forested, forming part of Canada’s more broadly defined boreal forest.1 The Lower Athabasca basin lies wholly within the Boreal Plains ecozone, and, although some notable uplands occur, the area consists for the most part of glacial lake bed or till plain on which large tracts of muskeg and fen have developed between areas of modest elevation. Major fish-bearing lakes—Lake Athabasca and Lac La Biche—are situated along its northern and southwestern margins, respectively, and smaller lakes and ponds are scattered throughout the area. As with similar regions across the northern hemisphere, food resources are relatively meagre and widely distributed, traditionally supporting only comparatively small populations of hunter-gatherers. With the exception of the access afforded by the river, travel throughout the region is difficult.
Although the Lower Athabasca basin features a broad spectrum of natural resources, two classes of resource have proved to be of major commercial interest to outsiders. In 1778, explorer and trader Peter Pond arrived in the Athabasca region and established a fur trade post, thereby inaugurating an activity that would serve as the basis of the region’s commercial economy for a century and a half. Pond noted the occurrence of a fluid tar-like substance (Stringham 2012, 21), which explorer and trader Alexander Mackenzie described in his journal a decade later, in 1788:
At about 24 miles from the fork [of the Athabasca and Clearwater Rivers] are some bituminous fountains into which a pole of 20 feet long may be inserted without the least resistance. The bitumen is in a fluid state and when mixed with gum, the resinous substance collected from the spruce fir, it serves to gum the Indians’ canoes. In its heated state it emits a smell like that of sea coal. (Mackenzie 1970, 129)
As Mackenzie noted, the Athabasca River and some of its tributaries had cut down through the surrounding bedrock, exposing bitumen along river banks in various locations to the north of modern Fort McMurray.
Aboriginal people had known of these bitumen deposits for millennia. Long before Pond arrived in the oil sands area, a Cree chief, known in historical records as “the Swan,” had brought them to the attention of the Hudson’s Bay Company traders. In 1715, the Swan visited York Factory, where he described the Athabasca region, telling of a river on whose banks could be found “Gum or pitch.” Four years later, the Swan returned to York Factory, bearing a sample of “that Gum or pitch that flows out of the Banks of that River,” which he presented to HBC governor Henry Kelsey.2 Other resources in the region were, however, of greater value to the indigenous inhabitants. As the chapters in this volume demonstrate, a conjunction of natural and human factors has resulted in an unusually rich record of landscape development and human use that makes the Lower Athabasca basin exceptionally valuable for understanding the early history of Canada’s north. The archaeological record currently being revealed represents perhaps the most intense pattern of prehistoric human landscape use yet identified in Canada’s boreal forest region.3
Investigation into the character, extent, and value of Alberta’s bitumen resources began in the 1890s. Early in the twentieth century, in an account of an expedition to the Mackenzie basin in connection with Treaty 8 (signed in 1899), Charles Mair wrote: “The tar, whatever it may be otherwise, is a fuel, and burned in our camp-fires like coal. That this region is stored with a substance of great economic value is beyond all doubt, and, when the hour of development comes, it will, I believe, prove to be one of the wonders of Northern Canada” (1908, 121). Mair’s instinct would prove to be correct. Research was accelerated by the demand for fuel during both world wars, and in the early 1960s a commercially viable process for recovering bitumen was industrially implemented.4 Only in the late 1990s, however, did technical advances and high oil prices combine to make large-scale investment in recovery attractive. Interest in understanding and developing bitumen resources originated among academic researchers and in government circles, although the role of these sectors has since evolved into providing research support for, as well as oversight of, the industrial activities now underway. Information relating to the effects of oil sands development on both the natural and the social environment continues to be collected by industry for compliance and planning purposes, but this mass of information has not been synthesized or widely disseminated. It is the intent of this volume to provide an overview of the information available on a subject that bears significantly on some of the current issues in the region: the historic value of the landscapes that have already been, or may still be, irrevocably altered by development activities.
Oil sands are deposits of bitumen, a molasses-like, viscous oil that will not flow unless heated or diluted with lighter hydrocarbons. For reasons discussed in this volume, bitumen deposits occur close to the surface only in a limited area within the oil sands region.5 Owing to its viscosity and the depth at which it is usually buried, extracting the bitumen from the sands in which is embedded requires large-scale industrial operations. In addition to on-site separation facilities, the process demands the creation of an infrastructure to supply power, water, and materiel, and some of the by-products, including overburden and tailings, require large storage areas prior to reclamation. In addition, the bitumen must undergo further processing and then be transported before it becomes a marketable product. Not only do these activities require major capital investment, massive amounts of equipment, and substantial workforces, but they transform large areas of formerly forested land.
Like other natural resources in Canada, petroleum reserves are owned by the people of the province in which they occur, and virtually all of the land that contains oil sands resources is provincial Crown land. Consequently, the public has a vested interest in how development proceeds. Since the onset of modern oil sands development in the 1970s, environmental and historical resources legislation has established comprehensive procedures to determine whether specific development projects are in the public interest and to licence and monitor such projects as they progress through various stages. These procedures typically include efforts to address the impact of development on the natural and cultural heritage.
As a result of the review and approval processes now in place and the unprecedented levels of development in the oil sands region over the past two decades, many studies have been undertaken to meet the conservation requirements laid out in the Alberta Historical Resources Act (2000). These studies have produced a remarkably abundant and detailed record of prehistoric land use. It is ironic to realize that much of the information presented in this volume might very well never have been collected were it not for the processes involved in planning for regional oil sands development and assessing its environmental effects.
To date, close to 3,400 archaeological sites have been recorded in the oil sands area, and numerous major excavations have recovered evidence of intensive prehistoric human use of the region’s resources. In addition, key geological and palaeoenvironmental studies have provided important contextual information that sheds crucial light on the reasons underlying changes in the prehistoric use of this landscape, as well as the nature of those changes. In the decades to come, if the large-scale surface mines and related developments currently planned are completed, they will erase a critical portion of this irreplaceable record. Although these archaeological resources are non-renewable, and the impact of industrial development on them is permanent, mitigation measures implemented in advance may help to offset these losses. While challenges exist, the expanding information base produced by ongoing studies enhances our ability to limit the effects of industrial development on the region’s natural and cultural heritage.
THE ARCHAEOLOGICAL STUDY OF THE LOWER ATHABASCA BASIN: KEY ISSUES
The record of prehistoric human use of the forested landscapes of northern Alberta consists of the remains of materials lost, discarded, or abandoned by the small groups of hunter-gatherers that lived in these environments for more than ten thousand years. Archaeological resources can range from a single artifact lost along a trail to dense concentrations of materials that represent a complex series of tasks undertaken by large groups of people, perhaps during repeated use of a specific resource-rich location over long periods of time. As important as the physical remains of these activities are, the relationships among them can provide equally valuable information on such matters as group structure and linkages between the activities required to complete certain tasks. The range in scientific and historic values reflected in this archaeological evidence is considerable, and because it represents past activity in response to past conditions, it cannot be replicated.
Determining the scientific significance of the archaeological resources recovered in northern Alberta poses certain challenges, however. Archaeological resources identified in boreal forest environments typically occur in shallow soil horizons that are acidic as a result of the decomposition of the coniferous vegetation that covers most upland sites. Given such conditions, many of the organic components of the archaeological record originally present—including any wood, hide, feather, grass, or reed artifacts, all bones, and most residues from animals or plants consumed or processed—have long since decayed. With the exception of microscopic residues that may be present on stone tools or fragments of ceramic vessels, and possibly mineralized remnants of organic materials that were subjected to fire (calcined bone or carbonized seeds, for example), there is little direct evidence of the resources on which prehistoric groups subsisted, and many classes of artifacts that were employed in economic, social, and cultural activities are missing from the record. What most frequently remains is only the evidence of stone tool manufacture and use, along with occasional remnants of hearths and cooking fires.
Although many of these conditions also apply in regions that possess more neutral soil conditions, and although stone artifacts can reveal much about ancient cultural practices, the absence of bone and other organic materials places significant limitations on the interpretive potential of the archaeological record of the boreal forest. For the most part, subsistence strategies can only be inferred, and ascribing absolute dates to archaeological assemblages is virtually impossible given the analytical techniques currently available to us. Dating boreal forest site occupations is almost entirely a matter of inference, one that depends on the presence of diagnostic artifacts—projectile points, ceramics, and occasionally other tool types—the style of which of which corresponds to the style of artifacts recovered elsewhere, the age of which has been established by radiocarbon methods. Diagnostic artifacts typically make up only a small proportion of the materials present at any site, however, and may not be represented at all in the remains of small, task-specific activities.
As will become clear in the course of this volume, despite these frustrations, we have learned an enormous amount about the prehistory of the oil sands region since archaeological investigations began there in the early 1970s. In the course of these investigations, attention has focused on several topics that come up repeatedly throughout the book. One concerns the origins and nature of the landscape that became available for human occupation at the end of the Pleistocene Epoch, following the retreat of glacial ice.6 As the ice receded, large lakes (termed proglacial lakes) formed along its edges. We know that somewhere between 9,800 and 9,600 years ago, a catastrophic discharge of water from one such proglacial lake—Glacial Lake Agassiz—scoured the deglaciated Lower Athabasca basin. Possibly in combination with an earlier flood event, this massive deluge of water created a landscape that fundamentally influenced prehistoric human use of the region.
Those familiar with glacial floods will realize that the Glacial Lake Agassiz event reflects processes that have occurred in many other situations associated with retreat of glacial ice. These events varied in the extent and degree of their effects, and each was conditioned by the specific environmental and geological circumstances in which it occurred. Given the intricacies of these location-specific variations, a comparative analysis of the Agassiz flood is beyond the scope of this book. The Glacial Lake Agassiz flood is, however, crucial to understanding many of the analyses presented in the volume, particularly those that highlight geological and palaeoenvironmental information. A number of the chapters thus include discussions of the Lake Agassiz flood, offering various perspectives on both the timing and significance of this event.
For reasons mentioned above, the majority of the evidence of prehistoric occupation of the Lower Athabasca basin exists in the form of remnants of the manufacture and use of stone tools. These artifacts were produced in vast quantities almost exclusively from a single type of locally obtained stone, originally named Beaver River Quartzite but now most commonly called Beaver River Sandstone (BRS). Issues surrounding the origin and use of this ubiquitous stone material constitute another major theme in oil sands archaeology. BRS, which occurs within the Cretaceous-age McMurray Formation, largely consists of sand grains fused in a matrix. It ranges in granularity from exceedingly coarse material that looks like a variety of quartzite or sandstone to very fine-grained material in which the original grains have been subsumed in a silica matrix—variations that reflect the degree of post-depositional heat and pressure applied to the formation. For the most part, the artifacts that have been recovered are made of relatively fine-grained material, which is better suited to the manufacture of stone tools.
The question thus arose as to the source of this material, given that the raw BRS found in the area, whether in situ (notably at a site known as the Beaver River Quarry) or in the form of boulders or cobbles, was typically quite coarse in grain. In 2003, the discovery of a complex of archaeological sites now called the Quarry of the Ancestors shed new light on this issue, as the stone available there is generally of finer grain.7 However, numerous questions surround the role of the Quarry of the Ancestors in prehistoric patterns of land use in the area and in the distribution of BRS throughout the region. In addition, the question remains whether the fine-grained BRS found in artifacts reflects human intervention in the form of heat treatment, a process that can transform even relatively coarse-grained stone into a material more suitable for tool manufacture. Fortunately, in 2012, the Quarry of the Ancestors was designated a Provincial Historic Resource and will thus be permanently preserved for future study.
TIME AND PLACE: SOME CONVENTIONS
For the benefit of readers for whom archaeology is relatively new terrain, some basic background information may be in order. It is sometimes possible to assign an age to prehistoric materials on the basis of radiocarbon dating of directly associated organic residues. This technique measures the proportional decay of a radioactive isotope of carbon (carbon 14, or 14C) subsequent to the death of living entity, be it plant or animal. It is most often applied to bone or wood but is suitable for a wide range of organic substances, all of which contain carbon. Originally developed in 1949 at the University of Chicago’s Institute for Nuclear Studies, radiocarbon analysis has since been refined through the use of accelerator mass spectrometry (AMS), which has increased its accuracy and decreased the amount of material required. Laboratories that conduct radiocarbon analysis report ages in radiocarbon years “before present” (BP), with the “present” defined as 1950 (reflecting, of course, the time at which the technique originated). From an archaeological perspective, the relatively short span of time between 1950 and the present day is for all practical purposes insignificant. Because levels of atmospheric carbon have not remained constant over time, however, ages reported in radiocarbon years BP begin to deviate from ages expressed in calendar years as one moves back in time from 1950: 9,850 radiocarbons years BP is, for example, roughly 11,250 calendar years BP. Calibration curves have thus been developed that, by taking into account past variations in the levels of atmospheric carbon, allow radiocarbon ages to be converted into calendar dates.
Archaeologists generally express dates in radiocarbon years rather than in calendar years. When laboratories report radiocarbon dates, they do so giving the age yielded by the analysis, the uncertainty range, and the sample number, which identifies the specimen on which the analysis was conducted: 9,410 ± 280 14C yr BP (UCR-2430B), for example. In general discussion, such dates are typically abbreviated by omitting both the uncertainty range and the sample number: 9,410 14C yr BP. As the original uncertainty range indicates, radiocarbon dates are not absolutely precise—nor is radiocarbon analysis perfectly reliable, as samples may occasionally be contaminated with hydrocarbons from an external source. All the same, dates founded on the scientific analysis of organic remains are regarded as relatively firm.
Given that boreal forest settings are characterized a paucity of organic remains, radiocarbon analysis is rarely possible. Instead, archaeologists must rely on stylistic comparisons of the sort described above. Dates derived through comparative methods are necessarily in the nature of estimates, and they depend for their reliability on the observational skills of the person examining the artifacts, as well as on the depth and breadth of the knowledge base that this person brings to the analysis. Such dates are not, however, inherently subjective. Assuming an experienced analyst, dates generated by comparative methods can in fact be credited with a fair degree of reliability. In this volume, dates that are in the nature of estimates founded on some form of comparative analysis are labelled with a simple “BP,” while those that are grounded in radiocarbon analysis of a specific specimen are labelled “14C yr BP.” When dates in calendar years are included (usually in addition to dates in radiocarbon years), these are labelled “cal yr BP.”
An archaeological site may represent only a single occupation, whether relatively brief or of long duration, or it may reflect successive occupations by different cultural groups, stretching over many millennia. In other words, archaeological sites cannot meaningfully be assigned a single, discrete date but rather occupy a span of time. Sites do, however, exist in a specific place. In Canada, archaeological sites are identified using the Borden system—an alpha-numeric system developed in 1952 by Charles Borden at the University of British Columbia. The system establishes a grid based on longitude and latitude that extends across the entire country and divides it into major and minor blocks. Major blocks, which are designated by capital letters, correspond to areas of 2 degrees of latitude by 4 degrees of longitude. Minor blocks, designated by lowercase letters, represent areas of 10 minutes of latitude by 10 minutes of longitude. Each minor Borden block thus represents an area of approximately 16 square kilometres. Within each minor block, archaeological sites are then numbered consecutively as they are recorded. Thus, for example, a specific site might be designated HhOv-73, with “Hh” representing the major and minor blocks of latitude and “Ov” the major and minor blocks of longitude. This system makes it possible to know the approximate location of a site purely on the basis of its designation.
The contributions to this volume reflect two fundamental approaches to the study of the past, the first relating to palaeoenvironmental conditions and the second to prehistoric human adaptations to these conditions. The chapters in part 1 discuss changes in the postglacial landscapes and environments of the Lower Athabasca basin, factors that are essential to our understanding of past human use of the area. This discussion is followed, in part 2, by chapters that focus on the patterns of prehistoric human occupation that emerge from the archaeological evidence so far recovered. Given that most of this evidence consists of stone tools and remnants of their manufacture and use, the chapters in part 3 discuss aspects of the origin, processing, and distribution of the region’s lithic resources. The volume concludes, in part 4, with two chapters that examine the effectiveness of the field and analytical methods presently in use, including current approaches to cumulative effects assessment, which seeks to forecast the impact of regional development on archaeological resources.
The volume opens with a consideration of the postglacial landscape in the Athabasca region. In the first chapter, “A Tale of Two Floods,” James Burns and Robert Young propose an alternative explanation for the extreme dearth of Pleistocene mammalian fossils in northern Alberta. In the oil sands area, only two vertebrate fossils have so far been recovered: pelvic bones from a mammoth, which appear to date to roughly 32,150 14C yr BP, and the skull of a wapiti, or elk, which has been dated to 5,550 14C yr BP and thus to the early Holocene Epoch. Burns and Young describe these specimens and consider the depositional contexts in which they were found. With regard to the question of why so few such remains have survived, they introduce the idea that, as during the most recent period of deglaciation, the Lower Athabasca basin may have been extensively modified not once but twice by large-scale floods. In addition to the well-known Glacial Lake Agassiz event, which occurred around 9,800 to 9,600 14C yr BP and flooded recently deglaciated terrain, a less well-defined event is believed to have taken place some millennia earlier. This flood would have occurred beneath the retreating ice cap, when the pressure of the meltwater that had built up under the ice reached the point that the water escaped and a massive subglacial flood ensued. By washing away massive amounts of Pleistocene sediments across a broad swath of the Lower Athabasca valley, these floods left bitumen deposits relatively close to the surface, thereby bequeathing to us ready access to the oil sands. But these floods would also have washed away virtually all existing megafaunal remains.
The second chapter, by Timothy Fisher and Thomas Lowell—“Glacial Geology and Land-Forming Events in the Fort McMurray Region”—discusses the timing and geomorphic effects of the retreat of the Laurentide Ice Sheet at the end of the Wisconsinan glaciation, more than 10,000 years ago. Drawing on digital elevation data from the Shuttle Radar Topography Mission and on radiocarbon dates obtained from sediment at the bottom of lakes in the vicinity of newly defined moraines, Fisher and Lowell offer new insights into the palaeogeography of the proglacial lakes that formed in the Churchill and Lower Athabasca valleys along the edges of the retreating ice sheet. As they suggest, the immediate source of the flood waters that scoured the Lower Athabasca valley was Glacial Lake Churchill, a proglacial lake that formed temporarily in the Churchill River valley as deglaciation proceeded. Building on earlier research, they review the geomorphic and sedimentological evidence for the massive discharge of water along the Clearwater–Lower Athabasca spillway, concluding that the spillway was first occupied sometime between 9,800 and 9,600 14C yr BP, with water continuing to flow along the channel for several hundred years afterward as Glacial Lake Agassiz drained through its northwestern outlet. In combination, they argue, evidence suggests that the processes of deglaciation that shaped the Lower Athabasca landscape took place somewhat later than had previously been thought.
In “Raised Landforms in the East-Central Oil Sands Region,” Robin Woywitka, Duane Froese, and Stephen Wolfe shed further light on the landscape that supported intense prehistoric human activity by examining the formation and character of the elevated landforms within the flood-modified Cree Burn Lake–Kearl Lake lowland. An analysis of landform shape and orientation, as revealed by LiDAR imaging, coupled with sedimentary observations, confirms that a majority of these features were formed as gravel bedforms related to the catastrophic flooding during deglaciation. These features are frequently mantled with windblown sand, indicating that windy, dry conditions prevailed following the deposition of sediments by the flood, and the occurrence of archaeological materials in these aeolian sands points to a human presence during this period and/or shortly thereafter. Subsequent to aeolian deposition, peat began to accumulate in the intervening lowlands, suggesting that, by this time, the surfaces of raised landforms had been stabilized by vegetation. The combination of a burgeoning wetland community and stable uplands would have provided an attractive habitat for human occupation. In addition, these well-drained landforms would have differed sharply from the surrounding till-based plain, with its dense, silty soils.
The final chapter in part 1, Luc Bouchet and Alwynne Beaudoin’s “Kearl Lake: A Palynological Study and Postglacial Palaeoenvironmental Reconstruction of Alberta’s Oil Sands Region,” presents the results of an analysis of pollen recovered from a sample core taken from Kearl Lake, located within the Athabasca-Clearwater plain along the margins of the Glacial Lake Agassiz outwash zone. As this pollen record indicates, spruce-dominated woodland had become established in the area prior to the flood, perhaps as early as 10,250 14C yr BP, during the terminal Pleistocene. Between about 9,820 and 7,580 14C yr BP, this landscape gave way to relatively more open deciduous woodland, characterized by birch, a shift that reflects the advent of the warmer, dryer conditions that prevailed during the Hypsithermal interval.8 The proliferation of birch might additionally have been encouraged by the destructive impact of the flood, given that, as the authors note, birch often dominates the new tree cover in areas that have been disturbed. The shift in climate is also visible in the increased presence of non-arboreal pollen types, variously indicative of reduced lake levels and of greater openness in surrounding upland vegetation. As the Hypsithermal waned, jack pine became a more prominent member of the vegetation community, as did spruce, while the growing abundance of peat moss spores probably signals the development of muskeg in lowland areas—alterations that essentially represent the establishment of modern boreal forest in the region. The pollen record thus provides crucial evidence of the shifting environmental conditions to which the region’s human inhabitants reacted.
In part 2, the focus shifts to prehistoric human occupation of the Lower Athabasca basin. The opening chapter, “The Early Prehistoric Use of a Flood-Scoured Landscape in Northeastern Alberta,” written by Grant Clarke, Luc Bouchet, and myself, offers an interpretive model of the Early Prehistoric human occupation of the Lower Athabasca valley in the wake of the Glacial Lake Agassiz flood. Drawing on palaeoenvironmental data, we argue that the scouring effects of the flood, in combination with the Hypsithermal climatic conditions that prevailed during the early postglacial period, created a highly productive regional microenvironment that stood in striking contrast to the surrounding higher-elevation forest. This parkland-like landscape, dominated by grasses and herbs, with open deciduous forest along ridge tops and meadows in intervening channels, would have been attractive to caribou and bison—grazing species that, because of their herding behaviour, are well suited to communal hunting. A review of the chronological sequence of human occupations developed in archaeological studies to date appears to support the contention that the most intense prehistoric use of the Athabasca oil sands region coincided with the period during which this microenvironment existed. As temperatures cooled and forests began to close in, animal populations would have shifted, with browsers such as moose increasing in number, and a pattern more typical of the boreal forest would have emerged, one characterized by smaller, more dispersed human settlements.
In the following chapter, “A Chronological Outline for the Athabasca Lowlands and Adjacent Areas,” Brian Reeves, Janet Blakey, and Murray Lobb lay out a sequence of occupations in the Lower Athabasca lowlands region. Drawing on Alberta’s database of archaeological sites, in tandem with a comprehensive review of the existing literature and collections, the authors offer detailed descriptions of the series of cultural complexes that, in their analysis, characterize the history of human occupation in the region, from its beginnings some 10,000 years ago through to the arrival of Euro-Canadian fur traders toward the end of the eighteenth century. Given that the paucity of organic remains typical of boreal forest settings largely rules out radiocarbon dating, the authors rely instead on established archaeological methods of comparative stylistic analysis, focusing on the identification of chronologically diagnostic artifacts, notably projectile points. As is standard practice, the individual archaeological complexes within the proposed sequence are given local names, but relationships with more widely distributed cultural traditions are both identified and explored. The resulting framework, the product of an extraordinary exercise in synthesis, will serve as a basis for establishing the relative age and/or cultural affiliation of assemblages recovered in the future, at least until techniques for the absolute dating of archaeological materials recovered from boreal forest settings can be developed.
Robin Woywitka’s “Lower Athabasca Archaeology: A View from the Fort Hills” provides an introduction to the prehistory of the area surrounding the Fort Hills, a moderately elevated series of uplands situated to the northeast of Fort McKay, on the northern periphery of the Lower Athabasca archaeological “heartland.” During the Lake Agassiz flood, outwash waters surrounded this uplands area but left its flanks untouched, producing a landscape that offers something of a contrast to the flood zone immediately to the south. The area now encompasses five major geographic features, which include, in addition to the uplands, the Late Pleistocene Athabasca braid delta and two prominent wetlands. Archaeological sites in the Fort Hills region, while less densely concentrated than those to the south, have yielded assemblages that are again dominated by Beaver River Sandstone. As is the case in the Athabasca lowlands, assemblages consist primarily of debitage but do include both formed and expedient tools. Woywitka offers a detailed consideration of the diagnostic projectile points recovered from sites in the Fort Hills area and of the evidence for microblade technology found at the Little Pond site, together with a review of the very few radiocarbon dates currently available. Drawing on these discussions, he provides a chronological and palaeoenvironmental account of the Fort Hills region, as well as an analysis of subsistence and land use patterns, and situates the Fort Hills in a broader regional context. As he points out, although the Fort Hills archaeological record is closely tied to that of the heartland area, it also exhibits characteristics more akin to boreal forest assemblages recovered throughout the southern Canadian subarctic.
The final chapter in part 2, “The Early Human History of the Birch Mountains Uplands,” by Jack Ives, discusses the prehistoric human history of a major contrasting ecosystem adjacent to the Lower Athabasca valley. Rising some 525 to 850 metres above the plain below, the Birch Mountains were among the first landforms to be exposed in the Athabasca region as the Laurentide Ice Sheet retreated. The ecology differs from that of the lowlands, featuring vegetation communities that developed under colder, drier conditions, as well as significant fish-bearing lakes. Drainage forms a radial pattern, linking the Birch Mountains not only to the Lower Athabasca region but also to the Peace River area and the Wabasca River drainage, to the northwest and southwest, respectively. Evidence of continued human occupation is present from the earliest postglacial times down to the historic period, with groups resident elsewhere travelling to the Birch Mountains as part of their seasonal round. As Ives notes, Beaver River Sandstone—so ubiquitous in the Lower Athabasca valley—is relatively rare in Birch Mountains assemblages, with the notable exception of a cache of tools discovered at the Eaglenest Portage site that appear to be the contents of a container used for transport. Ives reviews the key archaeological findings for the Early, Middle, and Late Prehistoric periods, including several radiocarbon dates, while cautioning against the temptation to define specific phases or complexes on the basis of scant and/or potentially ambiguous evidence. Our understanding of prehistoric land use in the Athabasca lowlands, he argues, would benefit from further research conducted in the surrounding regions, as this would allow us to investigate crucial questions pertaining to patterns of human movement and variations in the use of lithic resources.
The vast bulk of the archaeological record throughout Canada’s boreal forest region consists of the remnants of stone tool manufacture and use. Understanding how this evidence is distributed across the landscape and what kinds of human activity it represents is thus a critical component of regional archaeological study. The third part of this volume is accordingly devoted to the lithic record as it has emerged in the Lower Athabasca basin.
Part 3 opens with a chapter by Eugene Gryba, “Beaver River Sandstone: Characteristics and Use, with Results of Heat Treatment Experiments.” In it, Gryba reviews the physical and chemical characteristics of Beaver River Sandstone, as well as its origins and stratigraphic position within the regional geological sequence. In addition, he considers both known and potential sources of the stone and the frequency of its occurrence in archaeological assemblages that lie at some distance from source locations. As is well known, the material properties of a particular stone have a significant influence on the uses to which the stone can be put. These properties can to some extent be altered, however, by the application of heat. The heat treatment of raw stone material to improve its workability has been attested in a wide range of archaeological contexts, in North America and elsewhere. Gryba reports the results of experiments in which BRS was heated to temperatures in the range of 400°C to 450°C, which prehistoric peoples would also have been able to achieve. At such temperatures, BRS recrystallizes, often developing a thin, rust-coloured rind on its cortex and a smoother, more lustrous fracture surface. It also becomes considerably easier to work by both percussive and pressure methods. The possible application of this technique to BRS clearly has important implications for our understanding of lithic technology in the Lower Athabasca region, while it also provides a new angle on the question of why the BRS in artifacts sometimes appears to be of higher quality than the naturally occurring stone.
In the following chapter, “The Organization of Lithic Technology at the Quarry of the Ancestors,” Nancy Saxberg and Elizabeth Robertson present the first detailed discussion of an archaeological site of such significance that it has been set aside for permanent preservation. Discovered in 2003, the Quarry of the Ancestors is a complex of archaeological sites centred around two exposures of Beaver River Sandstone in the Muskeg River basin, to the east of the Athabasca River. The quarry was used intensively for several millennia roughly 9,000 to 6,000 years ago, when the climate was comparatively warm and dry, and data collected during initial archaeological investigations reveal that the inhabitants practiced a flexible, opportunistic approach to lithic reduction. Saxberg and Robertson surmise that the physical characteristics of BRS in its natural state facilitated extraction and reduction and that easily transportable packages of stone were then removed to sandy uplands, where the stone may have been heat-treated to improve its workability. The discovery of the Quarry of the Ancestors raises a series of questions—about the relationships among specific components of the site complex, about the lithic reduction methods employed at the quarry, about the variety of tools that occur and their distribution across the site, about the relative dearth of finished tools thus far recovered, and about the role of the quarry in regional patterns of occupation and resource use. Analyzing the evidence, Saxberg and Robertson propose that, rather than representing a place where groups out on their seasonal round stopped to procure lithic materials, the Quarry of the Ancestors served as a home base for peoples of relatively low residential mobility. A more complete understanding of the significance of the quarry will have to await future excavations, which may also shed light on the intriguing question of why the quarry apparently fell into disuse.
Microblade technology—that is, the manufacture of tiny stone blades, typically designed to be inset into projectile points made of bone, antler, or ivory but sometimes hafted onto handles—is highly characteristic of adaptations to arctic and subarctic environments. Microblades are common in Alaska and the Yukon, and they also occur in assemblages from sites along the northern Pacific coast and in the interior of British Columbia. They are, however, seldom seen in Alberta. In “Microblade Technology in the Oil Sands Region: Distinctive Features and Possible Cultural Assoications,” Angela Younie, Raymond Le Blanc, and Robin Woywitka examine the evidence for microblade technology at sites in the Lower Athabasca region. The technology was first identified in the early 1980s at the Bezya site, located in the Lower Athabasca valley not far northeast of Cree Burn Lake and, more than two decades later, at the Little Pond site to the north of the Fort Hills. However, with the onset of intensive oil sands development and the consequent upsurge in archaeological impact assessments, reports of the discovery of microblades and other evidence of microblade technology proliferated. On the basis of a re-examination of many of the specimens recently identified, including several from the Quarry of the Ancestors, the authors develop a critical approach that turns on the need to distinguish genuine microblades from blade-like flakes. As the authors point out, because microblades can be produced in a number of different ways, the analysis of microblades in Asian, Alaskan, and Beringian studies has focused on identifying the underlying technology—that is, the sequence of reduction of a microblade core—rather than on the simple presence or absence of microblades themselves. The authors argue that microblade technology is defined by an entire suite of distinctive features, not by only a few of these features in isolation. Evidence of true microblade technology in northern Alberta remains very scant, but careful analysis of this evidence in the light of traditions of microblade production recognized elsewhere allows us to trace potential cultural relationships between peoples in the Lower Athabasca and those in the far northwest of North America.
Historic resource management in Alberta is based on principles enshrined in legislation, which requires that commercial developers undertake archaeological assessment studies in advance of a proposed project (unless the impact of the project will clearly be negligible). If the initial assessment indicates that significant archaeological resources exist in the area slated for development, developers are further required to arrange for mitigative excavations, which are intended to recover a representative sample of the site in question. The goal of the assessment process is thus to determine the relative importance of the archaeological materials that exist within an area proposed for development, prior to making decisions about whether a proposed project will be allowed to proceed and, if so, on what conditions.
Because assessment studies are typically prompted by specific plans for development, they are fundamentally reactive, focusing on the immediate and relatively localized effects of a proposed project. Moreover, archeological assessment of these plans must often proceed without the benefit of a broader regional perspective founded on extensive prior archaeological study, of the sort that would provide contextual knowledge. As a result, the methods employed to assess the actual and potential effects of development are designed to be applied in a broad set of circumstances. As the knowledge base within a given region builds through continued study, these methods often prove to fall short of their intended goal, namely, to strike an optimal balance between the economic and social needs that drive development and the historical and cultural interest in preserving the archaeological record. These failures can, however, be instructive, as they often point the way to new, more effective assessment strategies. The final section of this volume offers two discussions of some of the issues surrounding current approaches to assessment.
“Quarries: Investigative Approaches in the Athabasca Oil Sands,” by Gloria Fedirchuk, Jennifer Tischer, and Laura Roskowski, examines the effectiveness of some of the archaeological techniques currently employed in impact assessment studies. One critical question concerns the fragmentation of data—that is, the difficulty inherent in efforts to construct an accurate picture of prehistoric land use patterns on the basis of widely dispersed studies undertaken in connection with specific project proposals, which typically stop at the borders of the area slated for development. Another concerns the philosophy underlying mitigative excavations, which privileges strategies designed to maximize the number of artifacts recovered, to the neglect of areas of lesser density that could well contain artifacts of greater interpretive significance. In addition, the authors critically evaluate the effectiveness of predictive models, such as those generated for the Quarry of the Ancestors, that attempt to determine the most promising locations for excavations by forecasting archaeological potential on the basis of landforms. In response to the perceived shortcomings of existing approaches, the authors propose a shift toward a more interpretive approach that would place greater emphasis on context, on identifying relationships among sites, and on developing specific research questions to guide archaeological study. By way of illustration, they explore three main areas of theoretical inquiry regarding ancient quarry sites: technology and mining, economic interactions, and social organization. In the light of this analysis, the authors suggest a number of ways in which these new approaches could be integrated to existing impact assessment procedures.
Development projects are additive, and so, of course, is their impact. In the closing chapter, “Cumulative Effects Assessment,” I examine the issues surrounding our efforts to assess the combined effects of regional oil sands development on archaeological resources. In Alberta, cumulative effects assessments (CEAs)—which, as in most jurisdictions, are mandated by regulatory processes—are incorporated within the framework of Environmental Impact Assessments, embedded in which are requirements that the effects of a proposed project on historical resources be described. CEA procedures are rooted in ecological studies, and, in addition to a number of more general shortcomings, the effectiveness of these procedures is significantly reduced when they are applied to archaeological resources, the character of which clearly differs from that of natural resources. In boreal forest settings, the situation is complicated by the nature of surviving archaeological materials and by the methods employed to find them and to evaluate their significance. Turning to issues of impact, I describe the long-term effects, both direct and indirect, that mining operations and planned in situ projects are likely to have on archaeological resources and also review our current understanding of the distribution and significance of known archaeological resources within the three officially demarcated oil sands regions. Using the categories developed for the CEA process, I then evaluate the significance of the combined effects of development activities on archaeological resources, focusing on the degree of confidence with which effective predictions can be made. As I point out, in its present form, the CEA process is initiated by developers, on a project-specific basis, for purposes of gaining approval from regulatory agencies. While this does not necessarily imply deficiencies in the quality of information provided, the restricted scope and objectives of CEAs have arguably limited their utility on a regional scale, as well as their capacity to guide regulators who consider the public interest in matters of development. Effective assessment is also limited, however, by our baseline knowledge, which accumulates slowly and is constantly evolving. As I suggest in closing, ongoing review and synthesis of archaeological information may therefore prove to be one of the most powerful methods available to offset the cumulative effects of development, whether in the oil sands region or elsewhere.
As I write, development of Alberta’s bitumen reserves has slowed considerably, largely in response to the worldwide oil economy, and, especially given global imperatives to reduce our dependence on petroleum products, the extent to which the pace will pick up again remains to be seen. All the same, the Lower Athabasca basin continues to be the locus of widespread industrial activity, and, in one form or another, further commercial development in the region seems likely. In addition to its other effects, this activity will disturb intact boreal forest, including the as yet undiscovered archaeological resources that lie not far below the surface. These resources are fragile, and they cannot be reclaimed, much less replaced. Archaeological research in the Lower Athabasca valley to date—the overwhelming majority of it occasioned by oil sands development—has revealed an exceptional record of intense prehistoric human use, one that appears unparalleled in the Canadian boreal forest. The information generated by these studies and the physical materials recovered have already immeasurably enriched our understanding of the early human presence in the region and hold significant value for future scientific research.
What is less often recognized is the value of this information for public education. Thus far, the findings of archaeological studies are for the most part scattered across unpublished compliance reports on specific research permits, while the results of geophysical and palaeoenvironmental studies typically appear as articles in scientific journals. This volume represents an effort to draw some of this information together, to take stock of the current state of our knowledge, to offer some provisional interpretations of the evidence, and to discuss some of the issues that complicate our efforts to develop a more comprehensive picture of prehistoric lifeways in the Lower Athabasca region. It is my hope that the chapters in this book will contribute to a deeper and more nuanced public understanding of the history embedded in this landscape and, in so doing, will help to build interest in the rich prehistoric heritage of northern Canada, while also serving to illustrate how archaeological knowledge evolves. I also hope that, by encouraging a more complex appreciation of the origins and archaeological significance of the Lower Athabasca basin, this volume will provide the impetus for improved conservation of the resources, both archaeological and natural, of the oil sands region.
1 Canada’s boreal forest covers more than a third of the country, spanning upwards of 1.5 million square miles, or some 3.9 million square kilometres (Henry 2002, xiii), roughly 38.9% of Canada’s total area. On the Boreal Plains ecozone, see “Boreal Plains,” Canadian Biodiversity Web Site, n.d., https://canadianbiodiversity.mcgill.ca/english/ecozones/borealplains/borealplains.htm; and, on the Athabasca River drainage, “About the Athabasca River Basin,” Athabasca Basin River Research Institute, Athabasca University, n.d., https://arbri.athabascau.ca/About-the-Athabasca-River-basin/Index.php.
3 The term prehistoric has come under criticism for tending to imply that oral cultures have no history (or no sense of history)—which is, of course, manifestly untrue. Some authors thus prefer the term precontact, and both terms occur in this volume. Although the latter term avoids reinforcing stale images of oral cultures as timeless and static, it also tends to suggest that the arrival of Europeans was an event of pivotal importance, one that fundamentally transformed these cultures. While no one would debate the destructive effects of colonization, these cultures in fact possess a resilience that is visible in the continuity of language and traditions despite the depredations wrought by the Euro-Canadian presence. In short, neither term is ideal. We understand prehistory simply as history preceding written history, with no value judgment implied.
4 The history of oil sands development has been covered in numerous publications. See, for example, Chastko (2005); Ferguson (1986); and Hein (2000).
5 Strictly speaking, the term oil sands region refers to three administrative areas created by the Energy Resources Conservation Board (now the Alberta Energy Regulator), an independent quasi-judicial agency of the Government of Alberta, in order to manage applications for the development of heavy oil deposits under the provisions of the Oil Sands Conservation Act and its attendant regulations. Under this legislation, three administrative areas—Athabasca, Cold Lake, and Peace River—were created in 1984, by independent orders, each specifying both the geological provenance of the relevant deposits and the area of land within which applications will be considered. Since then, the boundaries of the areas have been extended somewhat, such that the three areas now span a total of 142,200 square kilometres. See “Facts and Statistics,” Alberta Energy, 2016, https://www.energy.alberta.ca/oilsands/791.asp, and, for the three orders, “Rules and Directives: Oil Sands,” Alberta Energy Regulator, 2016, https://www.aer.ca/rules-and-regulations/by-topic/oil-sands.
6 The Pleistocene Epoch extended from about 2.58 million to 11,700 years ago. During this time, several episodes of glacial advance and retreat occurred, separated by warmer, dryer interglacial periods. The last of these glacial episodes, the Wisconsinan, which began roughly 110,000 years ago, is of the greatest relevance for our understanding current landscapes and subsequent human use. The Pleistocene Epoch was succeeded by the Holocene Epoch, which extends down to the present day; together, the two constitute what geologists define as the Quaternary Period. For an outline of geological time, see International Chronostratigraphic Chart, International Commission on Stratigraphy, 2015, https://www.stratigraphy.org/ICSchart/ChronostratChart2015-01.pdf.
7 Following the discovery of the Quarry of the Ancestors, the term Muskeg Valley Microquartzite (MVMq) was coined to describe the relatively fine-grained material that occurs at the quarry (which is situated in the Muskeg River valley). BRS is also sometimes called Beaver River Silicified Sandstone, with reference to the process whereby it was formed. These variations in nomenclature reflect geological analyses of the structure of the stone, which give rise to differing opinions as to how it should be classified. Several contributors to this volume thus speak of Muskeg Valley Microquartzite rather than Beaver River Sandstone, although the two are essentially the same material.
8 The Hypsithermal interval, also known as the Holocene Climatic Optimum or the Holocene Thermal Maximum (as well as by several other names), began early in the Holocene and persisted for several millennia, into the mid-Holocene, ending by roughly 5,000 BP. In the high Arctic, temperatures rose by several degrees centigrade, but the increase declined rapidly with latitude. As Bouchet and Beaudoin indicate in chapter 4, evidence for Hypsithermal conditions also varies with altitude, such that both the timing and the effects of this period of global warming differ depending on the location under consideration.
2004 Developing Alberta’s Oil Sands: From Karl Clark to Kyoto. University of Calgary Press, Calgary.
Ferguson, Barry G.
1986 Athabasca Oil Sands: Northern Resource Exploration, 1875 to 1951. Canadian Plains Research Center, Regina.
Hein, Francis J.
2000 Historical Overview of the Fort McMurray Area and Oil Sands Industry in Northeast Alberta. Earth Sciences Report 2000–05. Alberta Energy and Utilities Board and Alberta Geological Survey, Edmonton.
Henry, J. David
2002 Canada’s Boreal Forest. Smithsonian Natural History Series. Washington DC: Smithsonian Institution Scholarly Press.
Mackenzie, Sir Alexander
1970 The Journals and Letters of Sir Alexander Mackenzie. Edited by W. Kaye Lamb. Cambridge University Press for the Hakluyt Society, Cambridge.
1908 Through the Mackenzie Basin: A Narrative of the Athabasca and Peace River Treaty Expedition of 1899. William Briggs, Toronto.
2012 “Energy Developments in Canada’s Oil Sands.” In Alberta Oil Sands: Energy, Industry and the Environment, edited by Kevin E. Percy, pp. 19–34. Elsevier, Oxford and Amsterdam.
1 A Tale of Two Floods | How the End of the Ice Age Enhanced Oil Sands Recovery—and Decimated the Fossil Record
Although Ice Age vertebrate fossils occur in abundance in some regions of Alberta, the oil sands area, to the north of Fort McMurray, has yielded only a few remains of extinct mammals.1 Historically, the dearth of such fossils in northern Alberta has been attributed to the slower rate of settlement and development of natural resources—in-ground and above-ground—in comparison with the more southerly parts of the province. While Ice Age fossils from the northern portion of the province are indeed demonstrably rare in museum collections, we propose an alternative explanation for their scarcity.
The Athabasca oil sands have been known and used by local inhabitants for centuries, and their potential for commercial development has been an object of interest for decades. Until recently, factors limiting access and extraction kept the oil locked up in the sands, but over the past several decades extraction techniques have made mining of the sands an economically supportable enterprise. However, the oil sands would not have been accessible at all without the extraordinary events that took place during the latter part of the Wisconsinan glaciation, the most recent of the glacial episodes that occurred over the course of the Pleistocene Epoch. During that period, we argue, two major floods, one subglacial and one proglacial, discharged in different directions, scouring the landscape and eating through Pleistocene deposits and into the Mesozoic bedrock. Massive removal of these sediments by flood waters rendered underlying oil sands deposits in the Athabasca River valley readily accessible, but the action of these floods also offers a cogent explanation for the dearth of late Pleistocene vertebrate fossils.2
The past extent of glaciers and their paths of flow can be reconstructed by mapping the occurrence of rocks in areas where they are not normally found, as well as by the presence of characteristically glacial landforms, such as drumlins, flutings, hummocky topography, and tunnel channels, that is, channels thought to have been cut beneath a glacier (Rains et al. 2002). During the most recent glaciation, which reached its maximum extent approximately 25,000 years ago, the massive Laurentide Ice Sheet became so thick in northeastern Alberta that it was able to push its way up the slope of land to the west and southwest of present-day Fort McMurray, ultimately terminating at positions over 1,000 metres higher in the Porcupine Hills, southwest of Calgary, and all the way down into Montana (Young et al. 1994, 1999; Rains et al. 2002). In the Porcupine Hills, specimens of rock from the Northwest Territories and an array of subglacially eroded valleys occur at approximately 1,500 metres above sea level, along the margin of the Laurentide Ice Sheet (Rains et al. 1993). From this we can infer that the ice reached a depth of at least 1,500 metres in the southern part of the province. Given that the Lower Athabasca valley, in which Fort McMurray now lies, is situated at approximately 245 metres above sea level and the surrounding plains, above the valley, at about 350 metres, the city and region would have been submerged under a minimum of 1,255 and 1,150 metres of ice, respectively. It is very unlikely the ice sheet was flat, however: it would have been thicker toward the dispersal centre. Theoretically, then, the ice over the Fort McMurray region could have been as much as twice as thick—greater than 2,000 metres.
During the early stages of deglaciation, as the ice sheet began to melt, reservoirs formed on the surface of the ice, as a result of solar heating, as well as at its base, owing to internal friction and geothermal heat. Both reservoirs would have possessed a large amount of potential energy. A basal reservoir, or a series of such reservoirs, would have been confined and pressurized by the overlying kilometre or two of ice, while the water bodies on top of the ice would have contained tremendous amounts of potential energy by virtue of their degree of elevation above the land surface. The solar-heated water in surface reservoirs would gradually have eroded tunnels and crevasses, allowing it to flow into pre-existing subglacial reservoirs. Eventually, the pressure of water beneath the ice would have exceeded the pressure of the overlying ice, causing the water to escape and drain catastrophically, eroding enormous channels across Alberta, from the Fort McMurray region as far south as Montana (fig 1.1; and see Rains et al. 1993).
As these regions drained, the landforms and sediments characteristic of large turbulent water flows were left behind across the landscapes of Alberta (Rains et al. 1993; Sjogren and Rains 1995), while the outflow caused global sea levels to rise by 5 to 8 metres. Currently, similar but much smaller floods occur in Alaska, as well as in Iceland, where they are called jökulhlaups. The result in the Fort McMurray region was the formation of a wide erosional swath that stripped off some of the Ice Age deposits, leaving behind a level plain in the areas surrounding the Athabasca valley (fig 1.2, the black arrows). That swath can be traced to the south (see fig 1.1), where its elevation rises nearly a kilometre before it exits from Alberta into Montana. The flowing water was able to follow a path uphill because it was pressurized and flowed under confinement beneath the ice.
Figure 1.1. Digital elevation model (DEM) of Alberta, showing the channels cut by Late Wisconsinan subglacial megafloods
Several thousand years later, during deglaciation, large lakes were impounded against the margins of the retreating Laurentide Ice Sheet. One of the largest, and later, of these lakes—Glacial Lake Agassiz—covered most of the Hudson Bay drainage at one point or another in its existence. Rather than spanning a relatively fixed area, however, the lake occupied a series of areas, its shorelines shifting, and drained in a number of directions over the course of its lifetime, causing its water levels to rise and fall during specific phases in its history. At one early period in the evolution of Glacial Lake Agassiz, its water was diverted into the Missouri-Mississippi drainage, where it flowed into the Gulf of Mexico (Kennett and Shackleton 1975). Later, the failure of the ice dam along the eastern edge of the lake allowed a large discharge of meltwater through what is now the St. Lawrence Seaway, where the frigid water cooled the North Atlantic and may have caused a temporary relapse into ice age conditions (see, for example, Lowell et al. 2005; but see also Fisher, Lowell, and Loope 2006).
Figure 1.2. The flow direction of the two flood events in the Fort McMurray area. The black arrows pointing south indicate the local subglacial flood paths, while the arrow pointing north shows the subsequent path of Lake Agassiz flood waters entering the region through the Clearwater River.
At a subsequent high stand, known as the Emerson Phase, the lake drained to the northwest through the Fort McMurray area, into Glacial Lake McConnell (see fig 1.2, the white arrow), and then through the Mackenzie River drainage into the Arctic Ocean (Smith and Fisher 1993). This flood, originally thought to have occurred around 9,900 14C yr BP (about 11,335 cal yr BP), lowered the level of Lake Agassiz some 52 metres (Fisher and Smith 1994). Although it raised global sea levels by only about 6 centimetres, this flow would have caused the Arctic Ocean to rise 6 metres and would also have effected a freshening of the water. The lighter fresh water would have flowed over the denser ocean water, and since the fresh water was more likely to freeze, it would have served to increase the thickness of pack ice in the Arctic. The thickened pack ice and somewhat altered ocean circulation in the North Atlantic might have been responsible for a brief, 100- to 150-year cooling trend at that time, known as the Preboreal oscillation (Fisher, Smith, and Andrews 2002).
Figure 1.3. The northwestern drainage path of Glacial Lake Agassiz, about 9,800 14C yr BP. Flood waters carved a broad, deep valley along the Clearwater River, joined the Athabasca River, and then headed north to Glacial Lake McConnell and on to the Arctic Ocean. Two vertebrate fossils, one dating to the late Pleistocene Epoch and the other to the early Holocene, were recovered in 1976 and 1997, respectively, from the Suncor mine north of Fort McMurray.
This last flood also formed the valley of the Clearwater River, which rose in Saskatchewan, flowed westward into Alberta, and joined the Athabasca River on its way north (fig 1.3). The valley is much larger than it should be, given the size of the Clearwater River today. It is much wider than the Athabasca River valley upstream of Fort McMurray, even though the Athabasca channel carries a great deal more water. Downstream of Fort McMurray, where many of the current oil sands projects are located (including the Syncrude and Suncor operations), the valley retains the broad dimensions inherited from the Clearwater flood. The widest parts of the valley experienced the initial decreases in velocity during the waning stages of the flood and became sediment traps for large amounts of flood gravel and sand. Deposition of thick gravel-and-sand sequences and the establishment of roughly modern flow conditions set the final stage for the current geological conditions in the valley. Several thousand years later, a wetter and cooler climate saw the growth and expansion of the boreal forest (MacDonald and McLeod 1996), as well as the formation of the bogs that preserved the mid-Holocene wapiti antlers, described below, at the Suncor site.
THE GREAT CANADIAN OIL SANDS MAMMOTH
To date, the oil sands area has yielded only two significant Quaternary vertebrate fossils. The first was discovered in July 1976, when Steve Vayda and Dennis Olson, two employees of the Great Canadian Oil Sands Company (now part of Suncor), recovered three large pieces of bone from overburden removed during mining operations. Although the bones were not found in situ, a site supervisor estimated that they had come from a gravel pit about 50 feet (15.2 m) below the surface, where they had apparently been deposited by fluvial action.3 The bones are from the pelvis of a mammoth, most probably a woolly mammoth (Mammuthus primigenius), and consist of the nearly complete right innominate, as well as the pubic and iliac portions of the left innominate (fig 1.4). Dark mottling, ranging from mid-brown to black, suggests impregnation by hydrocarbons from the tar. The fossils are retained by the Royal Alberta Museum (accession no. P97.6.1).
The first radiocarbon date obtained for the specimen positioned it squarely in the mid-Wisconsinan interstadial: 32,150 ± 1,950 14C yr BP (S-3005). From the outset, though, this date was questioned by University of Calgary geomorphologist Derald Smith. In the early 1990s, Smith and Timothy Fisher (then Smith’s graduate student) had been studying postglacial landscape modification in the Fort McMurray region. They posited a flood event, brought about by the catastrophic draining of Glacial Lake Agassiz westward through the Clearwater River from Saskatchewan (see fig 1.3), which had inundated the area at the confluence of the Clearwater and Athabasca rivers, not far from present-day Fort McMurray. In the space of only a few months to a year, the flood dumped an estimated 21,000 cubic kilometres of water into the Athabasca River, which flowed into Glacial Lake McConnell (which survives today as Lake Athabasca) to the north (Fisher and Smith 1994; Smith and Fisher 1993).
Figure 1.4. The Great Canadian Oil Sands mammoth. The nearly complete right innominate (pelvic) bone, from what was probably a woolly mammoth, found in 1976 on the site of the Great Canadian Oil Sands project (now part of Suncor). The bone was initially dated to roughly 32,150 14C yr BP.
At the time, Smith and Fisher argued that the Lake Agassiz flood occurred at approximately 9,900 14C yr BP, during the lake’s Emerson Phase, when water levels were relatively high. Their dating of the event was based on radiocarbon dates obtained from eleven samples of wood and peat deposited in areas scoured by the flood, which yielded an average age of 9,869 14C yr BP. The dates for these samples coincide with dates for wood samples from elsewhere on the Emerson Phase shorelines of Glacial Lake Agassiz (Bajc et al. 2000; Smith and Fisher 1993). But in order to have been deposited by the Agassiz flood in the fluvial sediments of the Athabasca River, Smith and Fisher argued, the mammoth bones must necessarily date to this postglacial period. Hence, Smith ordered a second radiocarbon assay.
Smith may also have rejected the initial date because he suspected contamination, as the sample used to obtain that date had not been pre-treated for what appeared to be hydrocarbons introduced by bitumen from the surrounding oil sands. The second assay employed a method of hydrocarbon extraction developed at the University of California at Riverside, in which benzene was used to remove hydrocarbon contamination (in this case caused by tar from the La Brea tar pits in downtown Los Angeles). Using only the organic fraction of the Fort McMurray mammoth bone, the lab performed hydrocarbon extraction, with negligible residue, and reported a date of 9,410 ± 280 14C yr BP (UCR-2430B; in litt., R. E. Taylor to D. G. Smith, August 1990). For Smith, this younger date for the specimen was acceptable, as it indicated a postglacial deposition that coincided with the other available dates, from the wood and peat samples. All the same, the date should give one pause, as it suggests the occurrence of a very late-surviving mammoth in northern Alberta. This is not impossible, but it is unlikely, and so far unmatched.
In contrast, the date of 32,150 14C yr BP is not unreasonable given the suite of dates that were subsequently derived from mammalian megafaunal fossils recovered in various parts of the province (see, for example, Burns 1991, 1996a, 1996b, 2004; Burns and Young 1994; Matheus et al. 2004; Wilson and Burns 1999; Young et al. 1994, 1999). If the earlier date can be accepted (and we have no problem with that estimate), then, at the very least, it serves as a datum point, in both time and space, in the late Pleistocene fossil record of Alberta. Such a date is in fact nothing remarkable in Alberta, as mammoths have been dated from sites spanning an interval from roughly 43,000 14C yr BP in the Edmonton area (Burns, unpublished data) to 10,240 ± 325 14C yr BP from a site near Sundre, 95 kilometres north-northwest of Calgary (Burns 1996a).
In 2006, a third sample was removed from the mammoth pelvis and submitted for AMS radiocarbon dating to the Oxford University Accelerator Unit. Pre-treatment included hydrocarbon decontamination that, again, yielded negligible residue. The date returned—32,140 ± 230 14C yr BP (OxA-16322)—handily validates the initial date of 32,150 ± 1,950 14C yr BP and suggests that the process of tar extraction carried out at Riverside may have introduced younger carbon into the sample.
THE SUNCOR WAPITI SKULL
In March 1997, an impressive skull specimen, complete with a full rack of antlers (fig 1.5), was recovered from overburden at a site on Suncor’s property to the south of Mildred Lake.4 Loader operator Joe Revol spotted the large, creamy-white specimen lying next to the place from which he had taken his previous load of dirt, some 29 feet (8.8 m) below the current surface. We subsequently visited the site to collect the skull, which was identified as that of a wapiti, or elk (Cervus elaphus), and now resides at the Royal Alberta Museum (accession no. P97.3.1). Some minor desiccation cracks were evident on the bone, and the nasal cavity was blocked with a marly deposit containing tiny mollusc shells.
Stratigraphy of the site. Although we initially assumed that the marl on (and in) the skull meant that it had come from marly deposits, we reconsidered the assumption upon seeing the stratigraphy of the find’s locality. The actual locality had been obliterated by subsequent overburden removal, and we had to settle for examining a section 50 metres east of the find site. Even then, the depth of the face of this “proxy” site had been reduced by overburden removal. Working from the current top of the proxy section, which consisted of a layer of relatively dry golden peat, we determined that the skull had probably come from about 1.8 metres below the surface, toward the bottom of a progression of decomposed peat (muskeg), with some deep roots extending into the underlying sediment, that gave way to a dark grey marl from 1.84 to 1.96 metres below the surface (fig 1.6). From 1.96 to roughly 3.00 metres lay a stratum of medium planar-bedded sand with some cross-bedding in the lower portion; this layer also contained low-grade, friable, sandstone rip-up clasts and bitumen. Below that, there was a layer of extremely poorly sorted clastic fragments of local bedrock.