The Broken Mammoth Archaeological Project

Broken Mammoth Site
Excavations, 1991. Photo by Dr. Charles Holmes.

The Broken Mammoth Archaeological Project began in 1990 following the discovery of the Broken Mammoth site in 1989. The project was directed by Dr. Charles E. Holmes, Office of History & Archaeology. Investigations at the Broken Mammoth, Mead, and Swan Point sites have been sponsored by the Office of History & Archaeology and the University of Alaska Anchorage with support from the National Geographic Society and a 1991 National Science Foundation Research Grant (DPP-9112174). Special thanks to the Alaska Department of Transportation & Public Facilities (Fairbanks) and the Alaska Division of Forestry (Delta Station) for logistical support, and to the many students and volunteers who have helped over the years.

Swan Point Archaeological Site

*All Archaeology Projects on State of Alaska Land requires a State Archaeology Permit*

Broken Mammoth Site Excavation. Student excavating elk (Cervus sp.) bones at the Broken Mammoth archaeological site in 1991. This layer of the site was occupied by hunters around 11,500 B.C. Photo by Dr. Charles Holmes.

  • The Broken Mammoth Site Introduction

    • Why and where sites occur in the landscape continue to be crucial questions for archaeologists. The problem of finding sites becomes increasingly difficult the farther back in time one goes. The oldest sites may be deeply buried in loess or flood deposits, or they could be shallow surface scatters that lack context and perhaps have become mixed with more recent material. Or they simply may no longer exist because of natural erosion. In the Alaskan interior, boreal forest, taiga, and muskeg vegetation further hinder our ability to locate archaeological sites.

      The oldest archaeological sites yet known in Alaska are found in the Tanana Valley between the Alaska Range and the Tanana-Yukon Upland. Radiocarbon determinations for these sites are between 11,000 and 12,000 yr. BP using both conventional and Atomic Mass Spectrometer dates. Late Pleistocene and Early Holocene loess deposits, rich with calcium carbonate, have preserved butchered remains of birds and mammals in unequivocal stratigraphic context with human-made stone tools and hearth charcoal at three sites near Shaw Creek in the middle Tanana Valley, the Broken Mammoth, Mead, and Swan Point sites. Investigations at these three archaeological sites have yielded data to show that a broad-based hunting and foraging economy was practiced in Eastern Beringia at the end of the last ice age. These sites are the oldest in Alaska to contain evidence of artifacts directly associated with extinct mammals such as wapiti, bison, and mammoth. The co-occurrence of human tools and animal bones in these three sites provide a rare opportunity to refine our knowledge about human adaptation to environmental change.

  • Tanana Valley Prehistory

    • The traditional terminology developed in the 1960s and early 1970s is not completely satisfactory to address the archaeological record for the Tanana Valley. Archaeologists, while tending to retain traditional terms, nevertheless have modified them to specific needs, e.g, adding projectile points to the American Paleo-Arctic tradition and Denali Complex or including microblade technology in the Northern Archaic tradition. We also see terms like "Denali Phase" and "Late Denali Complex" which add to the confusion. To lessen this confusion and avoid forcing local data into chronological pigeonholes established elsewhere, I have divided time into five broad periods.

      BERINGIAN PERIOD (> 11,000 yr. BP) - The earliest yet discovered archaeological remains in the region include assemblages with microblades (e.g., Swan Point and Healy Lake) and assemblages lacking microblades (Nenana complex). Although this phenomena appears confusing, differences in assemblage compositions may reflect differences in site habitat, function, or seasonality. The term American Paleo-Arctic tradition seems inappropriate for this material as does use of Denali Complex or Nenana Complex. The term East Beringian tradition is proposed to include these earliest assemblages. This was before the forest was established, and while a land connection still existed between Alaska and Siberia.

      TRANSITIONAL PERIOD (11,000 TO 8,500 yr. BP) - During this period the American Paleo-Arctic tradition (which is conceived of as geographically widespread with technological variation) develops as an Alaskan phenomenon. It is manifest as a local variant that includes projectile points, and as such is restricted in time and space. Included within this lumping are aspects of both the Denali and Chindadn complexes. The land connection with Siberia disappears and the region undergoes changes, such as, animal extinctions, climatic change, and the beginning of forestation.

      EARLY TAIGA PERIOD (8,500 to 5,000 BP) - The boreal forest becomes fully established during this time. The American Paleo-Arctic tradition gives way over time to the Northern Archaic tradition. This is referred to as the Transitional Northern Archaic and is only poorly known for the region. There is continued debate on the issue of migration of new peoples into Alaska with the recognition of an "Archaic" tradition.

      MIDDLE TAIGA PERIOD (5,000 to ca. 2,500 yr. BP) - The local Tanana variant of the Northern Archaic tradition includes continuation of microblade and burin technology from the previous period. The use of the term Northern Archaic is somewhat problematic and should possibly be discarded in favor of a neutral term.

      LATE TAIGA PERIOD (ca. 2,500 yr. BP to modern) - This period represents a continuation of basic Tanana variant of the Northern Archaic tradition traits, but with the eventual loss of microblade technology. Around 1,500 yr. BP technological change marks the beginning of the Athapaskan tradition, which leads to ethnically recognizable Athapaskan groups.

  • Stratigraphy and Chronology

    • The stratigraphy at the Broken Mammoth and other sites in the Shaw Creek region consists of a series of aeolian sediments (sand and loess) overlying a frost-shattered and weathered felsic gneiss bedrock of the Yukon-Tanana crystalline terrane. The sediments and soils at the sites contain a record of past environmental changes in east-central Alaska. For example, the Broken Mammoth, Mead, and Swan Point stratigraphy documents the change from a full glacial, deflational environment at the close of the Pleistocene to one of alternating episodes of aeolian deposition and soil formation during the Holocene.

      Paleosol complexes associated with the two earliest human occupations (c.11,500 yr. BP and c. 10,400 yr. BP), are indicative of stable occupation surfaces. This interpretation correlates with the regional pollen record, specifically the latter part of the "Birch Period" (an open shrubland that included dwarf birch and willow) and the "Populus-Salix Zone" (described as open Populus woodland and birch-shrub tundra). After about 9,000 years ago, this shrubland changed to a woodland that included spruce and alder species. Faunal remains of Red squirrel and porcupine from the Broken Mammoth Site, suggest the process of forestation may have begun slightly prior to 9,500 years ago. Shortly thereafter, windier conditions may have become reestablished, as loess accumulation appears to have accelerated until about 7,000 years ago. After which, more modern conditions appear to have prevailed, with the modern forest soil established sometime after 5,000 years ago.

  • Interpretation

    • Although many early interior sites have some common characteristics, e.g., topography and aspect, there are also significant differences. Evidence suggests that most sites can be classified as either seasonal campsites, which served as headquarters, or "spike" camps from which small hunting groups operated. Tool manufacture and maintenance, along with domestic chores such as food processing and skin working, were routine activities. Large animals, e.g., bison, elk, and even mammoth were hunted, killed (or scavenged), and butchered away from camp. Hides, meat, and raw material, such as antler, bone and ivory would have been transported back to the campsite for further use.

      "Mammoth" ivory tusk fragments and artifacts found at the three Shaw Creek sites indicate that such material was obviously available to Late Pleistocene/Early Holocene human populations in the region. Thus far, no post cranial remains of mammoth have been found at the sites. It may be that ivory, along with hide and meat, was obtained at kill sites located away from the bluff top campsites and brought back to camp for use as raw material. Radiocarbon dates on ivory from the Broken Mammoth and Mead sites indicates that "fossil" ivory in the range of 15,000 to 17,000 years old was collected for use as tools. However, dates of 11,500 and 12,000 yr. BP obtained on ivory collagen from the Broken Mammoth and Swan Point sites is close in age to hearth charcoal dates to strongly suggest that mammoth and humans coexisted.

      The data provide a picture of small, highly mobile groups whose strategy for survival was to move around seasonally to exploit various resources such as caribou in the Fall or migratory waterfowl in the Spring. These people were highly knowledgeable about the environment and the animals they hunted. The evidence suggests that mammoths were a part of their resource base. The resources being exploited were changing during that period, and the human populations were adapting with new technologies such as changes to projectile point forms.

      What is almost completely missing from the archaeological record are sites interpreted as "kill" sites. By kill site I mean a location at or next to an ambush or capture location, where successful hunting parties may have momentarily rested while they butchered their prey, cooked a meal, and repaired their equipment. One example of such an early "kill" site may be the Erode away Site in the upper Nenana Valley. A single radiocarbon date of c. 8,500 yr. BP was obtained on charcoal from one of two small hearths. The artifact inventory consists of numerous biface projectile point fragments, apparently broken during the hunt and discarded around the two hearths. The rather small amount of flaking detritus was indicative of biface manufacture, repair, and re-sharpening. Unfortunately, very little faunal material was preserved in the shallow silt matrix. What is notable about the Erode away Site is that the site is not located on a bluff or promontory, but in a rather mundane setting along a glacial out wash feature of low topographic relief. Without extensive stripping of the ground surface in preparation for highway construction, it is unlikely that this site would have been discovered.

      Research during the past two decades in the Nenana Valley and more recently in the Middle Tanana Valley has led some archaeologists to conclude that the earliest well-documented Late Pleistocene culture in Alaska consisted of a non-microblade industry. This further led to speculation about deriving mid-latitude North American Paleoindian lithic technology and populations from Alaskan populations that did not have a microblade industry, i.e., the Nenana Complex. The 1993 test excavations at Swan Point have clearly shown that microblades belong to the earliest lithic assemblage for the region.

      For years there has been debate over how the early material from Healy Lake, defined by Cook as the "Chindadn Complex", fits or doesn't fit into the non-microblade Nenana Complex and the Denali Complex which relies on microblade technology. The inclusion of Chindadn into Nenana accepts the characteristic "Chindadn" points but denies the inclusion of microblade technology. On the other hand some Denali Complex proponents are reluctant to accept the Chindadn points as belonging to Denali. At the same time others are content to include both Nenana and Chindadn complexes within a greater Denali Complex, although a Denali Complex more ancient than originally thought. John Cook and others who have examined the bifacial material from Broken Mammoth and Swan Point have noted similarities to some "Chindadn" point forms. If it turns out to the case that there is a Chindadn component at Swan Point, then it should be accepted that the Chindadn Complex includes microblade technology. Evidence now supports the presence of microblade as well as non-microblade assemblages in central Alaska 11,000 to 12,000 yr. BP. Archaeologists now need to reevaluate the hypothesis that Clovis and other mid-latitude Paleoindian cultures were closely related to eastern Beringian cultures. It no longer appears tenable to propose that microblade technologies, exemplified by the Denali Complex, arose in Alaska following a pre-11,000 year BP non-microblade using culture like the Nenana Complex. It seems more plausible for microblade technology to have come into Alaska early from western Beringia. Once this technology became established here it appears to have stayed throughout almost the entire Holocene, at least in central Alaska.

      Although there is ample evidence to show that humans were present in central Alaska before 11,500 yr. BP, there is circumstantial evidence that people were here even earlier. The fact that many of these oldest sites, e.g., Broken Mammoth and Swan Point, contain obsidian artifacts identified as coming from the Batza tena source and from John Cook's Group A, implies the following: (1) First, these obsidian sources had to have been discovered and exploited. It could have required some time for the first emigrants to find these sources. (2) Next, it also must have taken time for the obsidian raw material to have become widely known and distributed among geographically separated groups through some form of social interaction. This too must have taken time for effective trading relationships to become operational between small populations throughout central Alaska.

      What this leaves us with is a vast, geographically diverse land mass - Eastern Beringia that has an archaeological time depth of at least 12,000 yr. BP. But with only a handful of identified archaeological sites for this early period and an even smaller handful of fully excavated sites. Reconstruction of ancient landscapes (including an understanding of paleoenvironmental parameters and prey animal behavior) are essential in developing testable predictive models of site locations that will better delineate the full range of human settlement at the terminal Pleistocene.

  • Further Reading

    • Dixon, E. James. 1993 Quest for the Origins of the First Americans. Academic Press, New York.

      Hoffecker, John F., W. Roger Powers, and Ted Goebel. 1993 The Colonization of Beringia and the Peopling of the New World. Science 259:46-52.

      Holmes, Charles E. 1991 The Broken Mammoth Archaeological Project. Heritage Quarterly Newsletter of the Office of History and Archaeology, No. 48.

      1992 Ancient Skin Sewing Needle found Near Delta Junction. Heritage Quarterly Newsletter of the Office of History and Archaeology, No. 53.

      1996 Broken Mammoth. In: American Beginnings: the Prehistory and Paleoecology of Beringia, edited by Frederick H. West, pp. 312-318. The University of Chicago Press, Chicago.

      Holmes, Charles E., Richard VanderHoek and Thomas E. Dilley. 1996 Swan Point. In: American Beginnings: the Prehistory and Paleoecology of Beringia, edited by Frederick H. West, pp. 319-323. The University of Chicago Press, Chicago.

      Holmes, Charles E. and David R. Yesner. 1992 Investigating the Earliest Alaskans: the Broken Mammoth Archaeological Project. Arctic Research of the United States, Vol. 6, pp. 6-9.

      Holmes, Chuck.

      Yesner, David R., Charles E. Holmes, and Kris J. Crossen. 1992 Archaeology and Paleoecology of the Broken Mammoth Site, Central Tanana Valley, Interior Alaska. Current Research in the Pleistocene 9:53-57.