HomeMy WebLinkAboutAPA1990ffB 2 i '%5
THE CARLO CREEK SITE: GEOLOGY AND ARCHEOLOGY OF AN
EARLY HOLOCENE SITE IN THE CENTRAL ALASKA RANGE
Peter Michael Bowers
ARLIS
Alaska Resources
Library & Information Services
Ancho~·age Alaska
Anthropology and Historic Preservation
Cooperative Park Studies Unit
University of Alaska
Fairbanks, Alaska
October 1980
Occasional Paper No. 27
ARLIS
, d:, Resources Library & Information Services
>..ibrary Building, Suite 111
321 1 Providence Drive
f',Jtchorage, .A~ 99508-4614
16
cover· by Susan Steinache
PREFACE
This monograph is virtually identical to my unpublished Master's thesis
of the same title, which was written during the winter of 1978-79, and defended
at Washington State University in 1980. Fieldwork for this study, which focuses
on a locality adjacent to Mount Mckinley (Denali} National Park, was conducted
in 1976 and 1977.
i i
I wish to add a note of appreciition to Zarro Bradley, Cooperative Park
Studies Unit, University of Alaska, who has provided funding for publication of
this study. He is to be commended for providing in this Occasional Papers_5eries
a forum for the rapid dissemination of information pertaining to Anthropology
in Alaska. Additional thanks are also due to Jim Dixon, University of Alaska
Museum, who in 1975 encouraged me to undertake this project .
. , ... , : ~
.. •
Peter M. Bowers
Fairbanks, Alaska
•• ••
ACKNm~LEDGHENTS
The Carlo Creek project was funded by the Geist Fund, University of
Alaska Museum, Sigma Xi, and the Washington State University Graduate Student
Association. Radiocarbon dates were provided by the Cooperative Park Studies
Unit, University of Alaska, and the Washington Archeological Research Center.
This study was made possible through the combined efforts. of many
people. I am particularly grateful for the services of the principal members .
of the field crew during the 1976 and 1977 field seasons: David Hoch, Mary
Whelan, and Edward Wick. Other persons who donated their time and efforts
during parts of those field seasons include Jane Bryant, E. James Dixon, Jr.,
Virgeen Hanna, Mike Schuster, Susan Todd, and Tom Waite. The fieldwork
benefitted greatly by the logistic support and/or equipment loans provided
by Doug Bowers, Zarro Bradley, John Cook, David Plaskett, Anne Shinkwin,
Bill Schneider, Kris Hoy-Thorson, and Bill and Karen Workman.
I am also indebted to a number of individuals for their help during
the laboratory and report preparation phases of this project. Thoughtful
critiques of portions of a February 1979 draft of this manuscript were
provided by Thomas Ager, J. Jeffrey Flenniken, E. James Dixon, Jr., and
Robert Thorson. Thomas Hamilton contributed numerous insights concerning
the Quaternary geology of the upper Nenana River Valley throughout the .
course of this project. This thesis has benefitted considerably by the
reviews of committee members Robert Ackerman, William Lipe, and Fekri Hassan.
My father, William Bowers, generously assisted with the preparation of the .
i i;
photographs; typing was done by Terri Jordan, Debbie Huyler, Kathy Cox,
# # ,
Cindy Bosley, and Marcia Gross. Eric Blinman, Ricky Hoff, Howard Smith, Bob
Wilkinson, and Susan Will provided invaluable support during the final
preparation of this manuscript. Numerous graduate school-related details
were attended to in my absence from Pullman (1978-1979) by Eric Blinman and
Marcia Kelley. To all of these people, as well as the others who have assisted
in one way or another, I express my sincere gratitude. I alone accept respon-
sibilities for possible inaccuracies in the interpretatton of the data pre-
sented here. Last, but not least, my hat is off to Mount Saint Helens, which
made my return to eastern Washington in May 1980 a most memorable experienc~.
This thesis is dedicated to the individual who first sparked my
interest in arctic archeology, through her scholarship, teaching, and stories
iv
about the Alaska and Greenland of yore : Frederica de Laguna, Bryn Mawr College.
THE CARLO CREEK SITE: GEOLOGY AND ARCHEOLOGY OF AN
EARLY HOLOCENE SITE IN THE CENTRAL ALASKA RANGE
ABSTRACT
by Peter Michael Bowers, M.A.
Washington State University, 1980
Chairman: Robert E. Ackerman
During the 1976 and 1977 field seasons, excavations were undertaken
at the deeply buried and frozen Carlo Creek site, located in the upper
Nenana River Valley, Central Alaska Range. The lowest of two cultural
components (Component I) contained well preserved faunal remains, indicating
human exploitation of caribou (Rangifer sp.), sheep (Ovis sp.), and ground
squirrels (Citellus sp.). The lithic assemblage from four workshop loci
within Component I includes percussion-flaked elongate bifaces, a large
prismatic blade-like flake, biface fragments, several possible bone tools,
retouched flakes, and more than 8000 waste flakes. There is no evidence for
an on-site core and blade industry in the Component I cultural horizon.
Several lines of evidence suggest that some of the locally-derived argillite/
hornfels used during the Component I occupation was thermally treated at the
site, presumably to improve the flaking characteristics of the lithic raw
material. Charcoal from two hearth areas in the lower cultural level has
been radiocarbon dated at circa 3500 years B.P. Although the lithic remains
are largely undiagnostic in terms of cultural affiliation, dat i ng and limited
v
typological comparisons suggest a possible affinity with an early Holocene
phase of the Denali complex.
vi
On the basis of modern day observations of faunal distributions,
seasonal hibernation patterns of arctic ground squirrels, and ethnographically-
documented aboriginal subsistence patterns, a late summer or fall human
occupation is inferred for Component I. Interpretation of the faunal data
suggests that the site functioned as a secondary kill site/butchering
station, where at least one sheep, one caribou, and nine ground squirrels
were brought to the site and butchered for transport to a larger regional
"base" camp--perhaps a site such as Dry Creek. The presence of bones,
representing carcass portions of relatively low dietary yield, suggest tha~
a selective "culling" strategy was employed by the site's inhabitants. The
fragmentary nature of the faunal remains and presence of an anvilstone (?)
implies that the bones were smashed for marrow extraction and/or bone grease
preparation.
An upper cultural level (Component II) consisted of an undiagnostic
assemblage of 637 rhyolite biface reduction flakes, and occurred stratigraph-
ically about one meter above Component I. Based on relative stratigraphy
and bracketing C-14 dates, Component II is inferred to be between about
6000 and 7500 years old.
Aboriginal selection of the site during both the Component I and
Component II cultural occupations was probably influenced by its strategic
mid-valley location ~tithin a major mountain pass, availability of lithic and
food resources, and presence of a freshwater spring.
The locality 1 S geology suggests that the Carlo Creek site may have
fanned as a result of alluvial deposition from the Nenana River, or from a
tributary thereof. This deposition sequence probably took place at least
1500 to 2000 years after the valley had been deglaciated, and occurred after
more than 75% of post-glacial valley incision had taken place. Limited
paleoecological data ·suggest that spruce forests may have been present in
the upper Nanana Valley by this time.
On the basis of interpretaion of the four major geologic units
identified at the site, the Component I cultural occupation probably occurred
on a stabilized sandbar of the Nenana River; the brief Component II habita-
tion most likely took place within an active floodplain. Beginning approxi-
mately 3800 to 4000 years ago, the alluvial deposits which enclose both
cultural levels were capped by an eolian unit. Contained within this unit
is a discontinuous tephra horizon, which has been correlated with the
Cantv1ell Ash Bed.
.·
vii
TABLE OF CONTENTS
ACKNOWLEDGMENTS
ABSTRACT ...
LIST OF TABLES
LIST OF ILLUSTRATIONS
Chapter
1. THE CARLO CREEK SITE
Introduction ..
Location . . . . . . . .
History of Research/Field Methods
Research Design
2. ENVIRONMENTAL SETTING .....
Physiography and Topography
Climate ..... .
Flora . . . . . . .
Fauna . . . . . . .
Permafrost and Soils
Cultural Environment
3. GEOLOGY
Page
i i i
v
X
xi
1
2
2
14
23
23
27
28
32
36
37
42
Bedrock Geology . 42
Quaternary Geology 43
Methodology . 43
Glacial Geology 45
Holocene Geology and Stratigraphy of the Car l o Creek
Site . . . 56
4. CULTURAL COMPONENTS
Component I
Component I:
Component I:
Component II
.....
Lithic Artifacts
Possible Bone Tools
; 77
77
83
91
93
Chapter
5. DATING . . . . . . . . . . . . . . . . . .
Radiocarbon Dates Relative to Cultural Occupation
Carlo Creek Region:. Geologic Dates
Carlo Creek Component II
6. LITHIC ANALYSIS
Introduction .
Methodology . . . . . . ... . .
The Carlo Creek Lithic Reduction System: Component I
Stage 1: Raw Material Selection
Stage 2: Heat Treatment .
Stage 3: Cortex Removal .
Stage 4: Biface Thinning
Stage 5: Finished Tools
Stage 6: Tool Use . . . . .•...
Stage 7: Disposal Mode: Introduction into the
Archaeological Record
..
Page
95
95
98
100
103
103
105
117
117
119
121
122
123
124
129
The Carlo Creek Lithic Reduction System: Component II 129
Stage 1 :
Stage 2:
Stage 3:
Stage 4:
Stage 5:
Stage 6:
Stage 7:
Summary
7. FAUNAL ANALYSIS
Raw Material Selection
Heat Treatment .
Cortex Removal ..
Biface Thinning
Finished Tools
Tool Use
Disposal
Introduction . . . ............. .
Faunal Analysis: Methodology and Interpretation
Seasonality and Aboriginal Procurement Strategies
8. PALEOECOLOGY .......... .
Regional Paleoecology ....
Local Paleoecological Setting
9. PREHISTORIC CULTURAL RELATIONSHIPS
10. SUMMARY AND CONCLUSIONS
BIBLIOGRAPHY ......... .
130
130
131
131
131
132
132
132
134
134
140
152
156
156
159
165
174
181
ix
LIST OF TABLES
Table Page
1. Pre 7000 B.P. Radiocarbon Dated Archeological Sites in
Alaska and Yukon Territory . . . . . . . . . 19
2. Generalized Description of Stratigraphic Units, Carol Creek Site 60
3. Carlo Creek Site Granulometric Analysis 61
4. Component I Artifacts
5. Summary of Radiocarbon Dates from the Carlo Creek Area
6. Carlo Creek Debitage Analysis
7. Component I Faunal Data
8. Frequencies of Selected Characteristics of the Faunal
Assemblage from the Carlo Creek Site ..
9. Calculation of Potential Available Meat Weights
84
96
108
136
137
147
X
LIST OF ILLUSTRATIONS
Figure
1. Location Map, Carlo Creek Study Area
2. Map of Upper Nenana River Valley, Central Alaska Range
3. Nenana River Valley
4. Topographic Map of Carlo Creek Site
5. Carlo Creek Site
Page
3
4
5
8
12
xi
6. Excavations at the Carlo Creek Site, 1977 12
7. Central Alaska, Showing Selected Localities Mentioned in Text ~24
8. Upper Nenana River Valley, Showing Probable Terminal Positions
of Late Pleistocene Glaciers . . . . . . . . . . . . . 47
9. Sketch Map of Terraces in the Nenana River Valley in the
Vicinity of Carlo Creek and Profile Along Line A-A 1 52
10. Generalized Stratigraphy, Carlo Creek Site 54
11. Stratigraphic Section Along Line 0 West _59
12. Carlo Creek Site Granulometric Analysis 62
13. Grain Size Distributions, Carlo Creek Site 63
14. C/M Diagram, Carlo Creek Site Sediment Analysis 65
15. Cumulative Grain Size Curves, Carlo Creek Site 66
16. Correlation of Dated Stratigraphic Sections, Upper Nenana
Va 11 ey . . . . . . . . . . . . . . . . . . . . . . . . 72
17. Limits of Excavation and Flake Distributions for Component I 79
18. Component I: Distribution of Artifacts, Faunal Remains,
and Hearths . . . . 80
19. Component I Artifacts . . . . . . . . . . . . . 85
Figure
20. Cobble Manuport, Component I .
21. Large Tabular Flake, Component I
22. Component I: Possible Bone Tools.
23. Component II: Distribution of Lithic Debitage
24. Determination of Component II Inferred Age
25. Component I Lithic Reduction System.
26. Component II Lithic Reduction System
27. The Fallacy Generated by Using Length x Width x Thickness =
Volume . . . . . . . . . . . . . . . . . . . . . .
28 . Templates Used in Determining Flake Size Categories.
29. Representative Examples of 3 Mammalian Species Found in
Component I, . . . . . . . . . . . . . . . .
Xi i
Page
90
9()
92
94
102
lOG
107
11 0
llO -
135
CHAPTER 1
THE CARLO CREEK SITE
Introduction
The Carlo Creek site is an early Holocene-age archeological site
located in the upper Nenana River Valley, central Alaska: The site contains
two deeply buried cultural components, both representing brief occupations
by small groups of prehistoric hunters. Although the artifactual assemblag2
from the site is small, the site is significant in its deep stratification,
good organic preservation, and alpine-floodplain geologic setting. The site
provides new data which are brought to bear on the problem of early Holocene
subsistence/settlement patterns and technological strategies employed among
hunter-gatherers in central Alaska. It represents one of less than a dozen
known late Pleistocene/early Holocene sites in Alaska which contain in situ
cultural material, and which have been systematically excavated. Among
pre-Hypsithermal age sites in interior Alaska/Yukon, only the Carlo Creek
and Dry Creek sites (Powers and Hamilton 1978) contain well-preserved
faunal remains in unambiguous association with lithic artifacts. On the
basis of detailed study of the Carlo Creek assemblage, no firm cultural or
technological affinities can be established with other early interior
sites; however, several lines of evidence suggest possible affinities with
Component II of the Dry Creek Site, and possibly with part of an early
Holocene phase of the Denali Complex (cf. West 1967, 1975).
The purpose of this thesis will be to provide a data base for
interpreting the site's setting and human occupations. The majority of
these data are derived from studies of the site's geology, lithic assemblage,
and faunal remains. Additional interpretation about the site is drawn from
available paleoecological data and limited comparisons with other sites in
interior Alaska.
Location
The Carlo Creek site is located in the central Alaska Range, within
a narrow constriction in the upper Nenana River Valley (see Figs. 1, 2,
and 3). The Nenana River is fed in part by meltwaters of the Nenana gl aci_er,
which heads some 75 km upstream from Carlo Creek. The site's valley-floor
setting lies at an elevation of 620 m above sea level. The mountains adja-
cent to the site rise t~elevations of 1480 to 2000 m a.s.l ., whi1e some of
the highest elevations in North America, notably Mt. McKinley (6194 m a.s.l .)
and Mt. Deborah (3700 m a.s.l.) are found within 80 km of the site.
The Carlo Creek site is situated within fluvial sediments which were
exposed by highway construction activities in the early 1970s. It is
located at mile 223.5 on the George Parks (Anchorage to Fairbanks) Highway;
NW l/4 NE l/4 Sec. 1, T16S, R7W, Healy quadrangle. The site occurs at a
latitude of 63° 33' 28" and longitude of 148° 49' 03". It lies roughly 150
air-kilometers SW of Fairbanks, and about 220 km NE of Anchorage.
History of Research/Field Methods
The Carlo Creek site was discovered during the summer of 1975 by
the geologists Robert M. Thorson and Thomas D. Hamilton, who were then
investigating recently-exposed roadcuts along the new highway from Cantwell
to Healy. At that time, non-diagnostic lithics and bone fragments were
2
(_ (. '/' .1
'
..
Fig . 1.--Location Map, Carlo Creek StudY Area.
CARLO CREEK
STUO'i AREA
.. ,.
IXI'lANAIION
m OIA<tll ,, '""''
A\ ....
Ml N IIOHf• IN II X I
.,·.J·-,,
10 II
U"JO' ·
Fig. 2.--Map of the Upper Nenana River Valley, Central
Alaska Range . 1--Carlo Creek Site; 2--Cantwell Ash type site (see
Fig. 16, locality 2); 3--tephra sampling locality; 4--tephra sampling
locality; 5--Nenana River Gorge Site; 6--Dry C~eek Site.
Fig. 3.--Nenana River Valley (view to south). Carlo Creek site is indicated by the arrow.
noted eroding out of what appeared to be post-glacial fluvial deposits.
Preliminary assessment of the s ite•s geomorphic setting suggested an age
for the cultural layer of between 6000 and 10,000 B.P.
The existence of the site was called to the attention of archeo-
logists David Plaskett and E. James Dixon, Jr. Because well preserved
condition of organic remains were present, it was felt that the site might
yield important data bearing on the aboriginal use of organic tools during
the early Holocene. Artifactual materials collected from the site by Thorson
and Hamilton were turned over to the University of Alaska Museum. A later
inspection of the site was ~ade that fall by Alaska State Archeologist
Douglas R. Reger, as well as Thomas Hamilton, Jim Dixon, and a representa~1ve
of the Alaska State Division of Highways.
In September of 1975, R. Thorson re-investigated the area, and
collected additional data concerning the glacial history of the upper
Nenana Valley. Thorson was concerned with the regional geomorphology and
early post glacial incision and alluviation within the upper Nenana River
Valley. On the basis of his fieldwork, a more conservative geologic age
assessment of between 3000 and 10,000 B.P. was sugg(~ted (Thorson, personal
communication, December, 1975).
The site was visited by archeologists Charles E. Holmes and D.
Plaskett during the late summer 1975, and briefly mentioned in a manuscript
report submitted to the Alaska Division of Parks in December of that year
(Holmes 1975). On the basis of the apparent association of faunal remains
and cultural lithics at Carlo Creek, it was recommended that further investi-
gations be undertaken. In the fall of 1975, the Carlo Creek site was
assigned the designation of HEA-031 by the State of Alaska Heritage Resource
Survey.
6
In November 1975, the author visited the site with David Plaskett.
At that time, the initial strategies were formulated for preliminary testing
to assess the site 1 s significance. However, due to a temperature of -30 .5°C
(-23°F) and an accompanying 40 kph wind, little fieldwork was undertaken at
that time.
Test excavations at the site were first undertaken by the writer
during July and August of 1976. This research was made possible by small
grants from the Geist Fund, University of Alaska Museum, Sigma Xi, and
limited personal funds. The 1976 fieldwork was conducted under a State of
Alaska Field A~cheology Permit issued to the University of Alaska Museum.
The primary objectives of the 1976 testing program were to: (1) locate and_
determine the extent of the archeological components,(2) map the site, and
(3) collect data necessary to make preliminary interpretations about the
site 1 S archeology, age, and paleoenvironments.
During the first season of fieldwork, a total area of approximately
30m 2 was excavated. The 1976 field crew consisted of the author and two
volunteer assistants. Total research time in the Nenana Valley amounted to
about 45 days.
Mapping of the site was completed by the end of the first week.
Coordinates used in generating the map presented in Fig. 4 were obtained by
use of an engineering transit and stadia rod. Datum was established at an
existing concrete highway right-of-way marker, located 48 m southeast of the
main site area. A subdatum was established on-site at a point 340° (true
north) north-northwest of the main datum. All horizontal and vertical
control measurements during excavation were determined from this point. An
approximation of absolute elevation was determined from the USGS Healy (C-4)
1:63,360 series map, 2000 ft. elevation contour line. Subsequent to the
r
7
HEA 031 . CARLO CREEK SITE
CON TOUR INHRVAl · I lUter ELEVATION 111 m•t•u 0 •. I.
LEGEND :
-MAIN EXCAVAT ION ARE A
-TEST PIT
L _-=] VEGETATION
I: . ·1 ROADCUT AR EA
j',:,::;,::::t] BERM
1 creek Site 4.--Tonographic ~~ap of the Car o Fig. t
--10
establishment of the subdatum, (point ON-OW), a grid of one-meter squares
was laid out over the main site area . The north-south axis was extended to
20 m north, while the east-west axis was extended downslope to 24 m west,
and upslope to 4 m east. The point 0 north/24 west, located within the
berm of the highway, was used as an elevation-control point (610 m a.s.l .).
Following the surficial mapping, test excavations were begun. The
major initial problem of the fieldwork was to determine the probable source
location of cultural materials. Observations were hampered somewhat by the
fact that, in places, as much as 30 em of colluvium had mantled the exposure
face, and in other places, several slump-blocks had displaced part of the
upper strata . No additional artifacts were located during these salvage
operations.
The original excavation strategy was to excavate a trench perpendic-
ular to the exposure face, in order to reveal the site microstratigraphy.
This plan was soon abandoned because of frozen ground and relatively non-
cohesive sediments. Frozen ground was encountered about 1 m below the
surface, and about 2.5 m back from the exposure face. Thaw rate varied
from approximately 5 em per day to about 20 em per day when squares east of
the lW line were excavated. At a depth of about 1.5 m, there were a number
of difficulties with cave-ins, slumpage, and general deformation of the
excavation walls, despite numerous attempts to shore the walls. It was
soon realized that a different strategy would be necessary, considering the
limitations of funding, manpower, field season, and frozen ground.
As a result, excavat i on strategy was altered to permit excavating
on a parallel to the roadcut face (parallel to the highway). This was
found to be quite productive--as a testing device, it made the best use of
available time, avoided frozen ground . (at this time of yea ~, the relatively
9
high sun angle permitted thawing back to about 2.5 m from the exposure
surface), and permitted us to obtain a continuous north-south transect
sample of the site. Excavation was comparatively easier in these areas of
non-frozen, loosely-consolidated sands and silts.
All vertical measurements were recorded with respect to site
subdatum (point ON/OW). Artifacts, faunal remains, and lithic debitage
were mapped three-dimensionally, to the nearest centimeter.
Because one objective of this research was to collect a sample of
terrestrial mollusks, considerable care was taken while digging in the zone
in which land snails were abundant, or approximately from a level about
50 em above the Unit 2-3 contact, down to the lower cultural level (see
Fig. 11, Chapter 3). To maximize recovery of mollusks, hand-sieves were
used; this aspect of the fieldwork was extremely time-consuming. Although
we had planned to use these snails for dating, we later found they had little
value for radiocarbon dating (J. Sheppard, personal communication, 1976).
In addition to excavations of the "main area" (refer to Fig. 4), a
number of peripheral test-pits and stratigraphic-control trenches were dug.
None yielded cultural or faunal remains, however.
Excavations in the main site were relatively productive, in that
precise stratigraphic location of the cultural materials was determined,
and an unexpected second component was uncovered. Although the recovered
artifacts were not clearly diagnostic of time .period or cultural affiliation,
a larger than expected number (more than 2800) waste flakes were recovered
during the 1976 season.
In addition to intrasite archeological and stratigraphic excavations,
a number of extra-site studies were made. These included examining roadcuts
10
in the valley, correlating terraces, and tracing out a number of suspected
volcanic ash horizons which occur at the site and to the south of the site
as far as Cantwell (cf. Bowers 1979a ).
Unfortunately, the 1976 excavations did not hit the primary archeo-
logical activity loci until late in the season. Consequently, the first
priority of the following field season was to expand the sample from the
southern sector of the excavations, particularly in the vicinity of Hearth
(see Chapter 4) .
Following a winter of data analysis and tentative interpretation of
the previous summer•s research (Bowers 1977a), excavations at the site were
continued from July to September of 1977, for a total of 60 working days.
This research was again assisted by grants from the Geist Fund, University
of Alaska Museum, and Sigma Xi . Additional support was obtained from the
National Park Service. The size of the 1977 field crew was highly variable,
as might be expected in a volunteer situation, and ranged from as few as
two to as many as nine persons at one time.
Fieldwork during the second season included expansion of site
excavations, regional geomorphological investigations, and limited archeo-
logical survey in the upper Nenana River Valley (Bowers 1979b). Work at
the site was again hampered by the presence of a considerable volume of
overburden, estimated at about 45 cubic meters. Depending on deptn of the
lowermost cultural level, the excavation depths ranged from 1 .5 to 3.5 m
below surface. In all areas of the 1977 investigations, the lower cultural
component occurred in frozen strata. A total area of 15 m2 of the lower
cultural level was exposed during the 1977 fieldwork (see Figs. 5 and 6).
During most of the 1977 season, all artifacts and faunal remains
were mapped in to the nearest centimeter. During the waning days of the
11
Fig. 5.--Carlo Creek Site (view to south). The location of the
site is indicated by the arrow.
'C -·' --... -.:.
Fig. 6.--Excavat i ons at the Carlo Creek Site, 1977 (view
to south).
r
field season, this strategy was shifted to one of recording and collecting
debitage by decimeter units. 1977 excavations were tied into the previously
established grid of one meter squares. During both field seasons, cultural
materials from the lower excavation levels was screened through a 1/4 in.
mesh screen. The site was dug entirely in natural stratigraphic units.
Five specific objectives were realized during the second field
season: (1) expansion of the main excavation area, and recovery of additional
cultural and faunal materials, (2) collection and dating of additional
radiocarbon samples, (3) collection and analysis of additional sediment
samples, (4) terrace profiling and related geomorphological investigations,
and (5) limited regional site surveys ( cf .. Bowers 1978b).
At the end of the 1977 field season, the site was backfilled, with
plastic sheeting placed at the base and eastern wall of the excavation
areas .. The remaining unexcavated portions of the site are now protected by
a minimum of two meters of fill. Arrangements were made with the Alaska
Division of Highways to have the site area and roadcut revegetated.
13
Laboratory analysis of the artifacts and faunal remains were under-
taken primarily at the Laboratory of Anthropology, Washington State University,
and completed at the NPR-A Project Archeology Lab of the Bureau of Land
Management, Fairbanks. Sediment analysis was completed at the Geoarcheology
Laboratory, WSU Department of Anthropology. Tephra analysis was undertaken
at the Tephrochronology Laboratory of the WSU Department of Soils and
Agronomy. Electron microprobe analysis of one volcanic ash sample was
performed at the Idaho Bureau of Mines and Geology, University of Idaho.
Radiocarbon dating of six samples was completed by the Radiocarbon Laboratory,
Department of Chemical and Nuclear Engineering, WSU, and Geochron laboratories
of Cambridge, Massachusetts. One lithic sample was identified by the X-ray
diffraction technique at the WSU Department of Geology.
Research Design
As in most multi-year archeological studies, the research design and
methodology for the Carlo Creek project evolved through several phases and
focused on a number of different aspects of problem orientation. As new
data became available from the site and from ancillary studies, research
strategies changed. The basic research framework, however, remained
unaltered throughout this study: it was designed simply to test, evaluate,
and interpret the site's archeology, age, and paleoecology, in an attempt
to add new data to the poorly-understood cultural-historical framework of
interior Alaskan prehistory.
Before detailing the rationale behind this study, it is necessary
to briefly familiarize the _uninitiated reader with the present status of
interior Alaska's prehistory. "Interior" Alaska, as used here, includes
the south slope of the Brooks Range, southward to and including. the south
slope of the Alaska Range, westward to and including the Koyukuk River and
upper reaches of the Kuskokwim River, and eastward to the Canadian border.
This area roughly encompasses the ethnographic range of the Athapaskan-
speaking Indians, as defined by McKennan (1969) and Krauss (1973, 1974)
--
At the outset, it should be ~phasized that only within the past two decades
has there been much intensive archeological investigations carried out in
interior Alaska. Consequently, work in this area (interior -Alaska-Yukon) is
still in its infancy, and is at a stage of research which may be considered
as "classificatory-historical" (Willey and Sabloff 1974:88). Similar senti-
ments have recently been expressed by Cook in his introduction to Pipeline
r
14
Archeology (1977:1): 11 Although ''new 11 archeological methods and concepts are
eminently suitable, the culture history of interior Alaska is still in the
historiography stage of archeological development .''
Clearly, there is still an overriding concern with developing
chronologies and filling sequence gaps, despite attempts of workers to keep
current with, and contribute to, the mainstream of archeological theory.
Due to great distances involved, difficulty of access, and the
relatively sparse nature of the resources themselves (reflecting the very
nature of subarctic/interior subsistence and settlement patterns), workers
in the subarctic have only recently begun to establish a basic cultural-
historical framework within which to place archeological data. The more
complex tasks of explaining changes in subsistence, settlement, or technology
as by-products of environmental and cultural process have thus scarcely
begun. A point made by one veteran of subarctic research is well taken:
The western subarctic is one area whose pioneer archeologists
are still among us; they are not even safely retired ... Today
archeology in the western subarctic is quantitatively different
but not different in kind. Success has been qualified. So many
sites are so small that often an assemblage of 20 implements
is a significant highlight, and sometimes I wonder if not our
southern colleagues view us as the 'northwest microcephalic
tradition'. I think I can fairly epitomize the current status
and future prospects as 'dogged does it.' Dogged is doing it
and dogged will do it. We still have our bits and patches,
but they are bigger and more exiting. They are still patches
because they do not fit to9ether or perhaps they do, and we
have not seen the obvious (D. Clark 1975:76-77).
As recently as 1975, a current statement about interior Alaskan
prehistory was summarized as follows:
The known sites in interior Alaska may be placed into three
broad categories of culture history: 1) historic or late
prehistoric occupations, rather definately Athapaskan in nature,
2) an older cultural stratum which may, or may not be early or
ancestral Athapaskan, and 3) a vaguely defined early period
(Cook 1975:125).
15
Because of the limited knowledge of interior prehistory, the
information potential of test excavations at the Carlo Creek site was
considerable. While some of my colleagues regarded as imprudent (foolhardy?)
this shoestring-budgeted excavation of a site which had yielded only a few
flakes, one biface fragment, and bone fragments (all buried under some 3m
of frozen overburden), I nevertheless felt the site held promise of much
important data. It was, in a large part, the uniqueness of HEA 031 which
provided impetus (and funding rationale) for more than one field season's
efforts. The research design also focused on the site's unique charac-
teristics:
1. The Carlo Creek site was known to contain well preserved organics,
and it was 0riginally thought to be between 3000 and 10,000 years old. The
fact that only two other known dated and in situ early Holocene archeological
sites in interior Alaska (Dry Creek and Healy Lake) contained any preserved
organics suggested that the Carlo Creek data would be able to provide a
much-needed comparative data set. Specifically, HEA 031 was regarded as a
prime candidate for providing data bearing on the organic fraction of early
man's toolkit. We know virtually nothing about the use of bone, wood, or
antler by early man in the . arctic and subarctic. If one examines some of
the available ethnographic literature for interior Alaska (e.g., Allen 1887;
McKennan 1959, 1965; Michael 1967; Osgood 1936a, 1936b, 1937, 1970; Schwatka
1885; Van Stone 1974; Whymper 1869), it becomes readily apparent that only
a small proportion of the total available aboriginal material culture--
perhaps as small as 5 to 10%--consisted of lithics. Yet lithic remains
provide the vast majority of the data base from which we formulate our
interpretation of ancient lifestyles, subsistence-settlement patterns (cf.
Streuver 1968:136), cultural-adaptive systems, etc. Among all known late
--
16
Pleistocene or early Holocene in situ, dated, and excavated arctic or sub-
arctic sites (including non-interior sites) only the Trail Creek Caves
(Larsen 1968:52-56) and the Canyon site (Workman 1974, 1978) contain preserved
organic artifacts. Several other northern early man sites have yielded
sparing quantities of bone implements, however, these have all been subjected
to major post-depositional disturbance in one form or another (cf. Bonnichsen
1978, 1979; Irving and Harington 1973; Porter 1978; Rainey 1939). In
contrast, the Carlo Creek Site appeared to be a locality where well-
preserved artifacts of bone, wood, or antler might be recovered in situ.
2. Because of its well-preserved faunal remains, it was felt that
HEA 031 might be able to contribute new paleoecological data bearing on
early Holocene subsistence-settlement strategies and mammalian ecology. .·
3. The geologic/physiographic setting of the site was regarded as
unique among early Alaska sites (interior and non-interior alike), in that
it suggested a floodplain -occupation (not redeposited alluvium) within ·a
restricted high-alpine valley. Only one other early man site/complex--the
Kobuk complex assemblage from Onion Portage, band 8 (levels 1 and 3)--
suggested a similar floodplain setting (cf. Anderson 1970:2-4), and no
other known early sites indicated an equivalent mountain pass physiographic
setting. Thus,~ new data from HEA 031 would add substantially to existing
knowledge of Holocene settlement patterns. One question that begged resolu-
tion was: is it reasonable or prudent to apply analogies of resource
exploitation, types of settlements, and seasonal rounds (Binford 1978b;
Campbell 1968; Chang 1962) from the ethnographic present back to the early
Holocene?
4. During analysis of the results of the first season's excavations,
the data suggested that the original occupants of the site may have been
intentionally altering the texture, lustre, and molecular structure of
1 7
certain lithic raw materials through thermal pretreatment (I am indebted to
Jeff Flenniken for first calling this possibility to my attention). Sub-
sequent field research and replicative laboratory analysis has tended to
strengthen this argument. If this supposition is correct, the Carlo Creek
site may represent the earliest documented case for intentional thermal
pretreatment of lithics in North America.
5. The original broad estimate of the site 1 S date suggested it
might fall within the period from 6000 to 8000 B.P.--a time from which we
have virtually no dated archeological sites in interior Alaska or Yukon
(see list of pre-7000 B.P. radiocarbon dated sites in Table 1). Among
interior sites, only the Tuktu site (Campbell 1962), mesa site (Kunz,
personal communication, 1980), and the Canyon site (Workman 1978) have been
radiocarbon dated to this crucial 2000 year span. At each of the three
major non-coastal stratified sites--Onion Portage, Dry Creek, and Healy Lake--
there appears a major occupation hiatus between at least 8000 and 6000 B.P.
(cf. Anderson 1968a; Cook 1969; Cook, unpublished data; Powers and Hamilton
1978). It is significant that, at each site, the hiatus appears to mask a
change from a 11 paleo 11 techno-cultural tradition (e.g., American Paleo
Arctic: Denali, Chindaden, Kobuk, Akmak) to an apparent boreal forest
adaptation, with concomittant changes in toolkits (e.g., Northern Archaic:
Palisades, Portage and Tuktu). Whether or not such changes reflect adaptive
responses to major post-glacial reforestation of Alaska, as suggested by
Anderson (1968b), Dumond (1969), and Bacon (1970), is uncertain. We clearly
need data from this time period to demonstrate a change in technologies (if
any) and (if possible) to explain such a transition. It was hoped that
data from the Carlo Creek site might help resolve these problems.
18
Site
Old Cr ·ow 29,100 t )(100
loc . l4N -2000
25 ,7 50 i 1800
-1500
27,000 t 3000
llluefish Cave 1£,900 1 100
Cd nyon Site 7£95 100
Hoose Creek llluff 11,730 I £~0
Oo ·y Creek 11,120 ll!>
10,690 £50
Cdr·lo to ·eek 11400 1 lOO
8690 ! 330
10,040 4 35
lle•ly lake 8960 !>0
(Vill aye Site) 10,250 I 380
ll680 240
11655 1 280
wl it [ 9895 2 10
Sdllt!Jit! 10,150 l 210
11,090 I 170
SIJ I i L [ 6645 2!10
samp l e 8210 155
spl iL [ 11465 I 360
Silml>le 10,041) £10
10,500 I 21l 0
Tdnyle lakes
Hll Ill IO,J!>O I £[10
Tdnyle lakes 9720 l 320
llliiO I 200
Clticken Hininy £/,OOO -J3,000
llepos its (I )
[itst -Ce ntr·a l
Int e rior·
TAillE 1
Pill 7000 ll.P. RAD IOCAHUON OArED AII CII EOI.OGICAl SITES IN AlASKA AND ¥UKOH
GX 1567 mdnono th bone
GX 1568 man oliO th IJOne
GX 1640 cdril>ou Lit.la
GSC 2881 horse femur
Sf 117 char·coa I
GX 6281 c harcod I
S f £8110 Ch arC Od J
Sl 1561 ch•o-coa l
wsu 1 700 Chd rcoal
GX 5132 c hao"Coa I
GX 5131 11 chdrcoa 1
GX 1340 bone
GX 2 173 Ch4rc oal
GX 2 170 dtdi'COa J
GX 2 171 c h •o·coa I
GX 2114 charcod I
Sf 737 Chdi'COd I
GX ll9lb !June
GX 2 1 59 Chdo ·coa I
S f (1) chdrcoa 1
GX 2 17 !. Clt dO 'COd J
S f 739 dldo ·coa l
GX 1944 c hJr ·coa 1
UI>A 572 soil ('I)
(14 .. 975 ) wuod
(1-4 567) wooJ
houe
Con on eots
Yukon
secondary deiJoS Its; fldkes
removed (1)
second.ry deposIts ; fldkes
removed
serrated edge (flesher)
Unit VII; may date occupation
lilt l e Anu phase
Refer·ence
lrvlny & lldrington 1973 :336
Ronnl c hsen 197!1, 1979
Bonnlchsen 197 8, 1979
Cin!j -Har·s 1979 :24
Wor ·k man l974b:94, 1977 :54
------·------------.
Cent r·a l Alaska
bifaces, lanceolate IJOiuts
Contponen t I ; Chi nda dn ( 1 )
COwlponenl II ; !lena II
lomj)OIIen t I ; !lena It -r·e I a Led ( 1)
level 4 "Qua r·t zite" llof'i zon
levt:l 6 "fdrly" tlortzon
l evel 7 "fady" Horizon
level 7 "Early" llol'i zon
level 7 "Early" llol'i zon
level 7 "Ec1rly" Horizon
level 8 "Early" llorizon
lt:ve 1 10 "Early" llor i zon
l eve l 10 "Ear·ly " llorl zon
Le v e l 10 "Larly" llurlzon
level 10 "Early" llof'izon
Level 10 "Ear ly " llod zon
llend I I Comp 1 ex
liu1iting dates fo1 · Oenali Comvlex
S ites; relc~le to " ... cl hiyh beach
s t rand with whi c h Denali Complex
Sites are consbtent ly associated."
bone ap11arenlly altered by uoan;
co II ec ted from sec ond4ry
deposits
llo ffccker· 1979
Thorson & lldluil ton 1977:166
Powers & llami I ton 1978
Bowers 1971lb; lhts report
Coo~ 1969
Pewe 197 5c~:26 ; Cook unpubli s hed Jc~t4
West 19/5:711
Current Rcse4rc h; Autc!·)ca •~ ~.!~~'!.1.~ Vof.'-j6 (4 f:49il -·
Porter 1971J
-. ,_, __ -·---.. ·---·----------·-· -------------------------·----------~-'-------
Site
Onion Portage
Tr·a i I Creek
Cdves
81J7 IOJ
8070 82
8210 t 84
8440 188
9857 155
9010 t 150
13,070 2110
15,7~0 i 350
p 1076
p 91l4
p 9115
GX 1508
I( 1583
I( 980
I( 1327
I( lliO
Halcl'ial
Udlcd
bone
bone
bone
bone
bone
bone
Lone
bone
TAUU !--(Continued)
Cououcnls
Northwcs t AI aska
Uand II, level I (Kobuk Coml'lexi
Band 8, level I (Kobuk Coo~>lex
Sand 'S, level I (Kobuk Coml'lex
Sand 8, level 3 (Knbuk Complex)
Oelow barul 8 (Akmilk Complex)
Scwar. · Penlnsu Ia
Layer Ill, section 2
dates bone I'Oints and
associated rnlcr·oblades
fractured bison calcaneous
outside cave 9
fr ·actured ho rse scapu Ia
outside cave 9 -------------·--·---·--· _____ , ____________ , ______________ _
North Slol'e
Reference
Andc•·sun 1970 :4
Anderson 1970:4
Lancn l961J:~4
Lar s en 1968:62
Lar ·sen 1968:63
.... ·----· --·--------____ , _______ , ___ .. ______ , __ -· .. ------------------------------------------------------·
Gal layer
flint Stat ion '10,!>40 1 150 SJ \174
Putu Site 60'10 150 liAK 4'140
8454 130 wsu 1318
11,470 1 500 51 23021>
Mesa Site Ji:\p 1 95 OJC 15U'I
-··-·---------------------------
charcoal
charcoa I
charloa I
c lodrco" I
drarcoa I
Loca I lty I ; cor·es & blades
fnlln soil In uncertain association
wl th fluted l'llints
Directly associated with lanceolate,
concave -based projectile points
llixun 1975 :69
Alexander 1974 :25
Morl ilfl 1977 ; 99
Kunz, unpublished data
-----------------____ , ______________ , _______________ ...
Aleutian Islands ---------____________________________________ .. _____________________ _
Ananyula
Uyashik Narrows
Nakn ek Rcylon
Ave,· aye of 33 da tcs :
7785 1 230
kanye :
69'12 --8400
7615 • 2b()
8425 115
8995 1 295
776~ 95
7890 1 90
SJ 1'198
SJ 2641
51 2492
S I 195~
51 1956
mo s tly on
dMCOdJ (?)
Anangula core & blade Industry
Alaska Peninsula
Pa leodrC tit lrad ilion; Dunoond' s
stage I
Ugashik Narrows j.~hase
Paleoarctlc traditlll(l;
Koygiung comple•
Lauyhl in 1975:510
(wml'are ioler,retation of
diltiuy with Aigner· 1976:51 -62)
llumond ct "'-197b: 19
ll enn I97U :l 2
Oumoud el d I . 1976 :19
lleun 1978 :12
Site Dale (8. P . )a lab. I
Hater! a I
Dated
TABlf !--(Continued)
Cooments
Southeast Alaska
--·--··-·--------------·-------------------------·--
Groundhog Day 88()0 125 I 7057 charcoa I nolcroblade & core
7545 I 185 I 7058 charcoa I Industry; Cotuponenl II
8 23 0 130 I 6395 charcoa I ranges to 4180 D.P.
10,180 t BOO wsu 412b charcoa I Cotuponen t Ill ; flakes,
9220 ! 80 Sl 2112 charcoal I scraper, 2 b I face fragtnents
9130 130 I 6304 charcoa I
Ill dden Fa II s 9860 75 wood Cotnponeh t I ; core & blade Industry
9410 ! 80 wood
7175! 155 wood
Reference
Ackerman, llano II ton, and S tuckenra th
1979:200
Ackennan, llam I It on, and Stuckenra lh
1979 :200
Davis 1979, 1980
--------------·-----·-. ------·---····--------------------------------------
aDates not adjusted for differences between l.ll>by and Penn half lives .
bOale questiondble .
In short, the Carlo Creek site research strategies were formulated
with the primary aim of helping to fill a major lacunae in the extant body
of archeological and paleoecological data for interior Alaska; test excava-
tions were not expected to resolve fully the above questions. It was felt
that the site probably was a small campsite, reflecting one aspect of an
''extinct" seasonal round (Binford 1962:217). Subsequent research has
indicated that it does represent a unique phenomenon among known early
Holocene Alaskan sites: a high altitude, riverine floodplain killsite/
campsite, which combines excellent organic preservation, good stratification,
·relatively minor post-depositional changes, and readily datable materials.
22
CHAPTER 2
ENVIRONMENTAL SETTING
Physiography and Topography
The Carlo Creek area (Figs. 1, 2, and 7) is located in the central
portion of the Alaska Range, which is part of the Pacific Mountain system.
The Alaska Range forms a mountainous arc some 960 km long and 80 to 170 km
wide, stretching from the Canadian border in the east, southwesterly to
the Alaska Peninsula and Aleutian Islands. The mountain system ranges in
elevation from relatively low passes such as the 600 m elevation Nenana
River Valley to massifs as high as Mt. McKinley .(6194 m elevation), located
120 km southwest of Carlo Creek. Mountains within the Alaska Range are
generally less than 3000 min elevation (Wahrhaftig 1965).
To the south of the Alaska Range lies the Broad Pass depression,
which is a flat-floored trench between the Alaska Range and Talkeetna
Mountains. Summit Lake, which marks the drainage divide between the Pacific _
and Bering Sea drainages, is located within Broad Pass, at an elevation of
690 m above sea level. The pass drains to the north via the Jack and Nenana
Rivers to the Tanana-Yukon drainages, and to the south via the Chulitna and
Susitna Rivers to Cook Inlet. Although Broad Pass is generally character-
ized as a north-south trending pass through the Alaska Range, it has been
described by Moffit (1915) as an east-west trending depression which connects
the heads of the Chulitna and Susitna Rivers.
r
23
0 200 ----KM
Fig. ?.--Central Alaska, Showing Selected Localities Mentioned
in Text. Also included are key localities in Northwest Alaska (7), Brooks
Range (8), and Yukon (9). 1: Carlo Creek site, 2: Dry Creek site,
3: Healy Lake site, 4: Tangle Lakes district, 5: Yardang Flint station,
6: Lake Minchumina, 7: Onion Portage, 8: Anaktuvuk Pass, 9: Old Crow,
10: Ne dana River, 11: Teklanika River, 12: Tok1at River, 13: Kantishna
River, 14: Delta River, 15: Broad Pass, 16:· Chulitna River.
24
North of the Alaska Range lies the northern foothills province,
which comprises a series of roughly parallel east-west trending ridges,
approximately 600 to 1350 m in altitude. The foothills are generally
between 4.8 and 11.2 km in width (USOI 1976). North of the foothills are
the Tanana-Kuskokwim lowlands of the Intermontane Plateau. This area of
low mountains and generally rolling topography drains the Nenana, Delta,
Tanana, Koyukuk, Yukon, and Kuskokwim Rivers. Elevations in this region
are generally between about 180 and 480 m above sea level (Wahrhaftig
1965). It should be noted that the northern foothills province has been
the focus of considerable archeological and geo-archeological research
since the mid-1970s (e.g., Thorson and Hamilton 1977).
The Alaska Range in the Carlo Creek Study area is snow-clad at
elevations above 2100 m on north facing slopes and 1800 m on the south side
of the range (USDI 1976). Canyons and gorges above these elevations are
filled by hundreds of cirque and alpine glaciers (USDI 1976).
One of the major glacier complexes in the Central Alaska Range forms
on the flanks of the 3700 m high Mt. Oeborah/Mt. Hess massif. Glaciers
originating in this area include the Susitna and West Fork glaciers, which
drain into the Susitna River, the Yanert glacier, which drains into the
river of the same name, and the Nenana Glacier, whose meltwaters form in
part, the Nenana River.
In the vicinity of the Carlo Creek study area, the most obvious
physiographic feature is the Nenana River Valley (Figs. 2 and 3). The
Nenana River begins some 75 km upstream from Carlo Creek. This braided
river then flows south and west through a wide valley between the Alaska
and Talkeetna Mountains. In the vicinity of the Broad Pass depression,
25
after its confluence with the Jack River, the Nenana turns ab ~ptly northward,
r
and cuts a deep U-shaped valley through 15-20 km . of the core of the Alaska
Range. Within this valley, the Nenana is joined by major tributary streams,
including Carlo Creek, Yanert Fork, and Riley Creek (Wahrhaftig 1958). The
mean annual high water discharge of the river, recorded 5.6 km downstream
from the Moody Bridge, is 7.0 cubic m per second (USDI 1976:22).
After joining with its major tributary, Yanert Fork, the Nenana
River then flows for some 13 km through the narrow Nenana River Gorge
(Plaskett 1977). Once through the gorge, and into the northern foothills
province, the river•s gradient is decreased gradually until its eventual
confluence with the Tanana River, located some 120 km north of Carlo Creek.
In the northern foothills region, the Nenana is fed by tributaries such a~:
Healy Creek, Dry Creek, and Lignite Creek.
Carlo Creek, which flows 16 km northwesterly into the Nenana River,
drains from the northern flanks of the rugged Panorama Mountain. This
clearwater stream empties into the Nenana at right angles at a point roughly
1.0 km northwest of the Carlo Creek archeological site. The site itself is
situated within terrace alluvial sediments formed along the eastern wall of
a relatively narrow constriction in the glacially-scoured valley. At this
point, the Nenana Valley is about 2.0 km in width. Immediately across from
the site is a prominent hanging valley, whose rim is approximately 200 m
above the valley floor. Major topographic features in the area include Mt.
Carlo (1482 m a.s.l .), Panorama Mountain (1933 m a.s.l .), and an unnamed,
1777 m high peak directly across the valley from Carlo Creek.
The locality 11 Carlo 11 was first reported by the Alaska Railroad on a
1923 manuscript map, and was later the name given to the section house at
Milepost 334.4 on the Alaska Railroad (Orth 1967). The Alaska Railroad
•
26
follows a sinuous route along the western bank of the Nenana River.
The more recently completed Anchorage to Fairbanks highway (George Parks
Highway) is located on the eastern side of the valley.
Climate
The Alaska Range in the Nenana Valley study area forms a transition
between the two major climatic zones in interior Alaska. To the north of
the range, a Continental climate predominates (hotter summer, colder
winters). To the south is found the more temperate transitional zone (USDI
1976).
Climatological data have been only sporadically collected in the
study area. There are detailed records for only the past few decades at
Nenana, Summit, and McKinley Park Stations. Of these three recording
stations, the latter is environmentally most analogous to the Carlo Creek
area.
Weather data collected at the Park for the past 40 years indicate an
extreme range of 32°C to -46°C (89.6 to -50.3°F). The coldest month, January,
has mean temperature extremes of -l0°C (l4°F) and -46°C(-50.8°F), ~vhile the
warmest month, July, has mean maximum and minimum temperatures of 20°C
68 °F) and 7°C (44.6°F) (USDI 1976).
Precipitation at the McKinley Park recording station averages approx-
imately 38 em per year. In general, winters are drier than summers .
November has the lowest average precipitation (0.78 em), while July has the
highest average with 9.6 em. Average snowfall is 192.2 em per year, with
drifts of greater than 6.10 m commonly occurring at higher elevations. For
forested areas, snow accumulation may typically average about 90 em (USDI
1976).
27
The prevailing wind in the Carlo Creek study area is l argely deter-
mined by the valley's orientation. Prevailing winds and the strongest winds
are from the south, especially during the summer months. North winds may
occur during the winter, but are less severe than their southern counter-
parts. The constriction in the valley between Cantwell and Carlo Creek
causes a natural 11 funneling 11 effect, resulting in extremely high-velocity
winds. (The locality 11 Windy 11
, 11 km south of Carlo Creek, was aptly named
by Alaska Railroad personnel in 1922). Although wind velocity records have
not been kept in this locality, average windspeeds of 32-40 kph would not
be an unreasonable estimate, with velocities as great as 160 kph at Windy
Pass. The locally strong winds in the Carlo Creek study area are importan~
factors in our consideration of: (1) eolian sedimentation in the valley and
(2) creation of wind-swept fall and winter pasture for grazers such as sheep
and caribou.
Flora
Floral communities in the region may be divided into two major
biomes, and subdivided into four local ecosystems (Viereck and Little 1972;
USDI 1976). Biomes include taiga and tundra, while the ecosystems include:
(1) bottomland spruce-poplar, (2) upland spruce-hardwood,(3) high brush and
(4) alpine tundra (USDI 1976).
Characterization of a specific area's vegetation is dependent on a
variety of complex and interrelated factors .. In general, the net annual
productivity of the subarctic environment is low. Factors such as short
growing season, low temperatures, precipitation, availability of nutrients,
and permafrost all affect plant growth. Slope direction and drainage are
critical factors. In addition , forest or tundra fires may be a major
contributor to the type and stage of succession of a given plant community.
28
In the Central Alaska Range, the boreal forest biome occurs in areas
of lower elevation, generally below the 700 m level. Included within this
biome are the two dominant sub-systems, bottomland spruce-poplar and upland
spruce-hardwood forest, and one transitional/early successional sub-system,
high brush (USDI 1976; Viereck and Little 1972).
Bottomland spruce-poplar forest contains dense to open stands of
white spruce (Picea glauca) and balsam poplar (Populus balsamifera). This
system is common along floodplains, where Populus sp. is a major constituent
of the early-successional forest. Climax stands of white spruce (Picea
glauca) may later develop as floodplain areas become more stable. On north
facing slopes with developed permafrost, black spruce (Picea mariana) may ..
become the predominant tree. Dominant shrubs in this ecosystem include:
American green alder (Alnus crispa), thin leaf alder (Alnus tennifolia),
little tree willow (Salix arbusculoides), feltleaf willow (Salix alaxensis)
and high bush cranberry (Viburnum edule) (USDI 1976; Viereck and Little
1972).
Upland spruce-hardwood forest occurs in dense to open stands, and
consists of paper birch (Betula papyrifera), and quaking aspen (Populus
tremuloides). Depending on drainage, aspect, and underlying soils and
permafrost, balsam poplar (Populus balsamifera) and black spruce (Picea
mariana) may also occur. Upland forest generally occupies areas of higher
elevation, with good drainage and no permafrost. White spruce (Picea
glauca) is the dominant species in the climax stage. Principal shrubs in
this system include the following: crowberry (Empetrum nigrum), narrow-leaf
labrador tea (Ledum decumbens), prickly rose (Rosa acicularis), bebb willow
(Salix bebbiana), mountain cranberry (Vaccinium vitis-idaea), feltleaf
r
29
willow (Salix alaxensis), littletree willow (Salix arbusculoides), and high
bush cranberry (Viburnum edule) (USDI 1976; Viereck and Little 1972).
The high brush ecosystem occurs as a transitional zone between the
forest and tundra zones, or as an earlier successional stage in the river
floodplain areas. Vegetative cover may vary considerably, ranging from
dense shrub thickets to comparatively sparse patches of shrubs. The
following species are found within this system, depending on moisture,
slope, aspect, and permafrost: alpine bearberry (Arctostaphylos alpina),
resin birch (Betula glandulosa), bush cinquefoil (Potentilla fruticosa),
Barclay willow (Salix barclayi), Alaskan bog willow (Salix fuscescens),
diamond leaf willow (Salix planifolia), Richardson willow (Salix lanata), _
Beauverd spirea (Spiraea beauverdiana), American green alder (Alnus crisoa),
cranberry (Viburnum sp.), narrowleaf laborador tea (Ledum decumbens), bog
blueberry (Vaccinium uliginosum), and mountain cranberry (Vaccinium
vitis-idae). Also occurring in this vegetation zone, particularly in areas
of poor drainage are: labrador tea (Ledum groenlandicum), sweetgale (Myrica
~), low blueberry willow (Salix myrtillifolia), bog cranberry (Vaccinium
oxycoscos), resin birch (Betula glandulosa), narrow leaf labrador tea (Ledum
decumbens), Barclay willow (Salix barclayi), Alaskan bog willow (Salix
fuscesens), diamond leaf willow (Salix planifolia pulchra), bog blueberry
(Vaccinium uliginosum), and mountain cranberry (Vaccinium vitis-idae). In
addition to the above species, numerous varieties of grasses (Arctagrostis
sp.), sedges (Carex sp.) and mosses (Sphagnum sp.) are common to both phases
of high brush ecosystems (USDI 1976; Viereck and Little 1972).
Alpine tundra occurs at elevations greater than 700 m in the Central
Alaska Range, and is characterized by low mat plants (both herbaceous and
shrubby) and areas devoid of vegetation. Large continuous areas may be
30
covered by White mountain-avens (Dryas octopetala), along v1ith grasses
(Arctagrostis sp.), sedges (Carex sp.), and lichens (Cladonia sp.). Common
shrubs and herbs in alpine tundra include: arctic willow (Salix arctica),
alpine bearberry (Arctostaphylos alpina), resin birch (Betula glandulosa),
mountain-avens (Eryas octopetala), crowberry (Empetrum niarum), narrow-leaf
labrador tea (Ledum decumbens), moss-campion (Silene acaulis), black
Oxytrope (Oxytropis nigrascens), arctic sandwort (Minvartia arctica) (USOI
1976; Viereck and Little 1972).
The immediate vicinity of the Carlo Creek archeological site is
classified as a bottomland spruce-poplar ecosystem. It occurs on a west-
facing slope, with moderate to poor drainage, and is underlain by discontinu-
ous permafrost. During August of 1976, the following species were identified
in the site area by S. Todd, then of the Institute of Northern Forest,
Fairbanks: Black spruce (Picea mariana) (dominant), narrow-leaf labrador
tea (Ledum decumbens), willow (Salix sp.), alder (Alnus sp.), mountain
cranberry (Vaccinium vitis-idaea), bog blueberry (Vaccinium uliginosum),
crowberry (Empetrum nigrum), horsetail (Eguisetum sp.), cottongrass
(Eriophorum vaginatum), paper birch (Betula papyrifera), prickly rose
(Rosa acicularis), grasses (Arctagrostis sp.), blue joint grass
(Calamagrostis canadensis), dwarf fireweed (Epilobium latifolium), rhubarb
(Polygonum alaskanan), sedges (Carex sp.), mosses (Sphagnum sp.), and
lichens (Cladonia sp.). In disturbed areas along the highway and river
floodplain, the following were observed: wild rhubarb (Polygonum alaskanum),
horsetail (Equisetum sp.), willow (Salix sp.), grasses (Arctagrostis sp.),
and dwarf fireweed (Epilobium latifolium).
:\.L.!.. ~ _;_-, .;. I _, f .n.. y 1 ~ . ; • ~~=~ .. /·
U.i.. DE~:·!'. OP IS'i i:.l~lOR
r
31
Fauna
A diverse mammalian fauna is present in the study area today, and
has probably been extant in the area for a considerable length of time. It
should be emphasized that most of the large mammal species in the Nenana
Valley are transient, following seasonal patterns of movement through a
diversity of topographic and vegetative zones. It has been postulated
(Whitten 1975; restated by Guthrie and Powers 1977) that such movements are
directly related to the seasonal nutritional "peaks'' in the maturation of
key plant species, which enable the animals to optimize available nutrients.
In the Carlo Creek study area, three major ungulates are present
today: moose (Alces alces), barrenground caribou (Rangifer arcticus), and_
Dall sheep (Ovis dalli). The following discussion will examine in detail
only those species directly re1evant to the Carlo Creek archeological
assemblage: caribou, Dall sheep, and ground squirrels (Citellus sp.).
Caribou, a gregarious and grazing mammal, occur seasonally over a
wide range in the Central Alaska Range (see Figs. 1, ~ and 7). The McKinley
herd, although presently quite small, was estimated at ca. 30,000 animals as
recently as 1941 (Hemming 1971). At present, the herd ranges over 9600 km 2
(5965 square miles), extending from the Nenana Valley westward to the
north fork of the Kuskokwim River. According to Hemming (1971 :45), one of
the major wintering grounds for the herd in recent history has been the
Broad Pass region, and the hills between the Nenana and Kantishna Rivers.
Calving generally takes place during June in the rolling tundra regions
between the upper Savage and Teklanika Rivers (HeiTD'Tling 1971; :1urie 194~.;
Skoog 19G8). After calving, the herd then moves over the Alaska Range to
the area between Cantwell and the west fork of the Chulitna River, which
" ... has a long history of early summer use" (Hemming 1971 :45). An additional
32
major summer range is the area north of Carlo Creek -Panorama Mountain,
where abundant lichens and sedges provide ample food resources (USDI 1976).
During the summer, a general western movement occurs, with the herd again
crossing the main range. In late August and September, herds begin to drift
back toward winter ranges, dispersing enroute (Hemming 1971 :4). On the
basis of available data, the Carlo Creek region is a primary caribou
habitat from mid-summer to spring, and could have been so in the early
Holocene.
Dall sheep occur on steep tundra slopes and rocky areas on both
sides of the Nenana River Valley, and are restricted primarily to alpine
tundra ecosystems. They are gregarious animals, occupying higher elevations --
during the summer, and lower elevations during the winter. Wint.er distribu-
tions are largely controlled by snow conditions. Although sheep rarely
range far from the safety of precipitous terrain, they do occasionally cross
valleys to get from one mountain slope to another (Whitten 1975). Major
lambing and breeding grounds are found on the eastern side of the Nenana
Canyon between Sugar Loaf and Panorama Mountains (USDI 1976). Guthrie
and Powers (1977:20) have observed that:
... Sheep require windblown pastures for winter range, and
travel great distances to mountain pass areas which are more
consistently exposed to winter winds. The upper Nenana Valley
is a major sheep winter range today because of this phenomenon
and undoubtedly was in the past for sheep and a number of other
large mammal grazers.
The Carlo Creek vicinity, particularly elevations above 700 m, thus
may have provided good sheep habitat during virtually any season of the
year.
Moose, which are also common in this area, are browsers, and are
generally associated with the forest or high brush ecosystems. Willows and
33
.------
various kinds of lacustrine aquatic vegetation are major food sources for
these animals. They are found throughout the study area and throughout
most of the interior Alaskan taiga.
Other large mammalian species found in the area are wolf (Canis
lupus), Toklat grizzlies (Ursus arctos), Black bears (Ursus americanus),
coyotes (Canis latrans), red fox (Vulpes fulva), and lynx (Lynx canadensis).
Lesser mammals in the region include the following: arctic gound squirrel
(Citellus parryi), marmot (Marmota sp.), porcupine (Erthyzon canadensis),
wolverine (Gulo loscus), flying squirrel (Glaucomys sabrinus), red squirrel
(Tamiasciurus hudsonicus), various mice and shrews (Sorex sp.), mink
34
(Mustela vison), otters (Lutra canaden~is), and red-backed vole (Clethrion9mys
rutilus) (USDI 1974).
Of the above-mentioned minor species, only one, the arctic ground
squirrel, will be discussed in any greater detail. Its presence in the
faunal remains at the Carlo Creek si t e make it of special significance
for paleoecological and archeological interpretations.
The arctic ground squirrel is generally regarded as a highland form
(either latitude or elevation) (Tikhomirov 1959:33). They are found in
colonies , frequently along river banks or abandoned river terraces.
Melchior (1964:41) states that the topographic/edaphic features of burrow
site location include: (1) well drained soils, (2) topographic prominence and
(3) high proportions of phosphorus and potassium in the burrow area. Carl
(1962:46) suggested that grasses and willows are the most frequently-observed
plant associated with squirrel mounds, and later stated (1971 :4 00) that
" ... all of these [burrow] locations have a comparatively deep permafrost
level."
35
The ground squirrel goes into hibernation in burrows in late September
to early November, depending on temperature. Bee and Hall (1956:55) suggest
that the "disappearance of ground squirrels in autumn and their reappearance
in spring, after hibernation, was thought to be governed by altitude and
surface exposure." Tikhomirov (1959:33) observed that temperatures of
approximately -10 to -20°C are required to induce hibernation. Similar
observations were made by Carl (1962:51) who observed no squirrels above
ground at temperatures less than -l5°C.
A ground squirrel colony was observed by the author on the west bank
of the Nenana River, near the Carlo section house, in August of 1977.
Burrows at this location were dug into well drained sandy sediments
overlying a bedrock outcrop, at an elevation of about 605 m a.s.l.
Vegetation in the surrounding area was open canopy black spruce-poplar
forest, with an open grass and willow association in the immediate vicinity
of the burrow.
In addition to mammals, numerous species of birds are found in the
Carlo Creek area. Most of these are migratory, and are found in the area
only in summer. The Nenana River Valley is one of the most important mi-
gration routes through the Alaska Range.
Several species of fish are also found in the study area, although
these are not of significant economic or subsistence importance. The Alaska
Department of Fish and Game regards the Nenana River as an anadromous fish
stream, but this applies only to the river's lower reaches, near its
confluence with the Tanana River (USDI 1976). Arctic grayling (Thymallus
articus) are probably the most abundant species in the area; these are found
in the Nenana and nearly all tributary streams. Carlo Creek provides a good
grayling habitat during the warmer months . Lake trout (Salvelinus
nomaycush) are found in some lakes in the area.
Permafrost and Soils
The Carlo Creek area presently lies with a zone of discontinuous
permafrost. Local conditions such as slope angle and direction, drainage,
permeability, and vegetation to a large degree determine the presence/
absence of permafrost. At the Carlo Creek archeological site, permanently
frozen ground was consistently encountered at a depth of 1.0 m below
surface. In areas of denser vegetation, or of drainage poorer than at the
site locality, frozen ground was encountered at depths as shallow as 0.25 m
in mid-August.
Soils in the study area can be divided into three orders: histosols,
inceptisols, and spodosols; and may be subdivided into five suborders:
fibrists, cryaquepts, aquepts, ochrepts and orthods (terminology based on
USDA 1975). There has been no detailed soil association mapping done in the
Carlo Creek area.
Fibrists (histosols) are fibrous or woody peats, largely
undecomposed, and develop in areas of poor drainage and high permafrost
table. They are formed primarily from organic matter and are associated
with wet tundra.
Cryaguepts (low humic gley inceptisols) develop in areas of poor
drainage, and are associated with black spruce (muskeg vegetation). Perma-
frost is at or near the surface, and drainage is poor.
Aguepts (inceptisols) may frequently be encountered on alluvial
floodplains and recent alluvial terraces. They are seasonally saturated and
poorly developed soils with little vegetative covering.
36
Ochrepts (inceptisols) may develop in association with early stages
of forest soil formation, or may be associated with tundra vegetation.
Qchrepts characteristically have thin or light colored surface horizons with
little organic matter accumulation. Drainage is relatively good, with
little or no permafrost.
Orthods (Spodosols or subarctic brown forest soils) form in fairly
well drained areas, and represent developed forest soils. These soils
generally contain a thin elluvial (E) horizon, and a developed B illuvial
horizon, with subsurface accumulations of iron, aluminum, and organic
matter. Soils in the immediate vicinity of the Carlo Creek archeological
site have been classified by the author as Orthods.
Cultural Environment
Present day population centers in the region include Healy (pop. 79)
located off mile 252.5, Parks Highway, and Cantwell (pop. 62), located
about 22 km (14 miles) south of Carlo Creek. Healy was established as a
coal mining camp in 1905, and later became a railroad supply station when
the Alaska Railroad was built in the 1920 1 S. Healy is still an active coal
mining town. Cantwell, established during construction of the Alaska Rail-
road, is inhabited primarily by Athapaskan Indians who moved to the area
from the Ahtna/Copper Center area (USDI 1976:40). The Carlo Creek locality
has a permanent population of six (USDI 1976:42).
Mount McKinley National Park, the entrance and headquarters of
which lie roughly 20 km north of Carlo Creek, has a year around population
of 15, with a seasonal population of about 250 (USOI 1976:43). The park
was established in 1917 as an area for public recreation and wildlife
preservation. The eastern boundary of the Park lies directly across the
Nenana River from the Carlo Creek site, a distance of about 0.4 km.
37
Prior to white exploration and settlement in the early twentieth
century, the Carlo Creek/upper Nenana Valley was only sporadically occupied,
and only on a seasonal basis. The area was probably used by several
different native groups for late summer to winter hunting of caribou,
sheep, and moose. The area was within the probable seasonal round of
activities for the Ahtna (de Laguna, personal communication, 1975; Osgood
1936a), Tanaina (Osgood 1937), and Tanana (McKennan 1959). All of these
groups are of northern Athapaskan linguistic stock (Krauss 1973, 1974) and
can be characterized aboriginally as opportunistic hunters and gathers
whose lifestyle and relationship to the environment shifted throughout the
year.
The following is a highly generalized description of historic
period Athapaskan subsistence strategies and seasonal round. Summertime was
preoccupied with fishing, particularly along the major salmon-spawning
rivers. Fishing camps were established and traditionally re-used year
after year. Summer also involved moose hunting. During the fall and
earlier winter, the emphasis shifted to hunting sheep and caribou in the
upland areas, especially during the seasonal migrations of the caribou
herds. Depending on extent and availability of cached reserves, the winter
and spring was quite likely a time of scarcity and possibly starvation.
Groups may have split into family or extended family units and scattered
into the highland areas in search of game. Springtime might have been
involved with the taking of waterfowl, and, in early summer, the pattern
may have again emphasized fishing and the taking of caribou (abstracted
from descriptions in de Laguana, personal communication [1975], McKennan
[19591, Osgood [1971], and VanStone [1974]).
38
The Nenana River Valley mountain pass, which runs through the core
of the Alaska range, could have played a key role in virtually any part of
the above pattern. The most likely use of the valley, however, would
probably have been in the late summer to early winter. It should be noted
that recent settlement patterns may be biased somewhat in favor of riverine
seasonal settlements, particularly along anadromous fish streams (cf .
Andrews 1977; Bowers and Hoch 1978; Morlan 1973). The introduction of the
fish wheel and white trade undoubtedly resulted in more intensive and semi-
permanent settlements along the rivers.
39
Although the area is today claimed by the Ahtna group, it is by no
means certain that this group was in control of the area prior to White
contact. The natives of Cantwell moved to that location during construction
of the Alaska railroad, probably from the Copper River, Susitna River, and
Valdez Creek mining areas. Historically, they maintained cultural and kinship
ties with the Tanana of interior Alaska as well as with the Ahtna groups to
the southeast (Ahtna, Inc. 1973).
The early historic literature for the area is noticeably lacking in
detail concerning the region's inhabitants. The few early accounts which
describe the upper Nenana Valley underscore the region's importance as an
aboriginal seasonal hunting ground and as a major transportation corridor
through the Alaska Range. William Yanert, in his report entitled "A Trip to
the Tanana River" (1900:736) observed that ·
The divide between the west fork of the Sushitna and the
river flowing northward [Nenana] is low, 2600 feet, and has the
appearance of a valley from 7 to 9 miles in width. A belt of
spruce extends across from north to south. A well-worn footpath
leads over the divide, which the guide assured me was made and
used by indians going to and from the Tanana. He also pointed
out the frames of two indian houses north of the divide, stating
that these were used by the Tanana Indians during the hunting
season. Moose and caribou signs were plentif ul in this locality
and it appeared to be the wintering place of game and other animals.
Yanert's Indian guide, who declined to continue northward beyond
this point indicated to him that the Tanana River could be reached in three
and a half days journey from the Cantwell area: ''As concerns the Indian's
refusal to remain in service as a guide, I feel certain that his
unwillingness to do so was prompted by fear of the Tanana Indians, who, he
frequently assured me, were very numerous and bad" (Yanert 1900:736).
Van Schoonhoven (1900:736) reported a similar aboriginal utilization
of the high country in the upper Nenana Valley: " ... The Jack River country
is a good one for game, and the Sushitna [Tanaina] Indians make this their
hunting ground for caribou and sheep." He later continued northward,
through the Nenana Valley Pass, and proceeded toward the Tanana (1900:
737).
It is thus quite likely that the region has been used extensively
in the prehistoric past. Cultural boundaries in this transitional area
doubtlessly shifted back and forth, permitting an influx of traits and
material goods from the south central coastal area, as well as the deep
interior. Beyond the past several hundred years in the area, however,
cultural and linguistic identification of the region's inhabitants is
difficult (cf. McKennan 1969; Krauss 1973, 1974). Plaskett's (1977)
detailed research at the Nenana Gorge site, located 30.0 km north of Carlo
Creek, provides probably the best view of the late prehistoric use of the
area. He notes many material similarities with other Athapaskan sites from
both sides of the Alaska Range, and points out the importance of the area
for hunting and as a trade route (1977:216). Plaskett also provides an
excellent summary of the ethnohistoric period in the Nenana River Valley.
Thus, the availabie ethnographic literature, though limited,
indicates the possibility that the Carlo Creek area was used seasonally by
40
groups from either the south or north of the Alaska Range. Of importance
to our consideration of early Holocene use of the area are: (1) the extent
of deglaciation in the upper valley, (2) the availability of faunal resources
and(3) faunal resource procurement systems. All three of these factors
either have been, or will be, discussed in greater detail elsewhere in this
thesis.
..
41
CHAPTER J
GEOLOGY
Bedrock Geology
The bedrock in the study area consists of east-west trending rocks
which range in age from pre-Cambrian to Tertiary. The older group of
rocks, pre-Cambrian to Cretaceous, consists of schist, phylite, gneiss,
chert, argillite, hornfels, limestone, conglomerate, slate, shale, and -·
coal. These are generally metamorphosed and well consolidated. A younger
group consists of low-grade metamorphic rocks which were formed by intrusion
and alteration of basic lava flows. These two grouRs are separated by a
major unconformity, and form two contrasting groups, with respect to age,
consolidation, and tectonism (Wahrhaftig 1958).
The structure of the Alaska Range is a broad synclinal complex,
with indications for early Cenozoic orogeny (W. Gilbert 1976:5 ). This
synclinorium contains Cretaceous rocks in the center and Paleozoic rocks on
the flanks (Wahrhaftig 1965). Paralleling the major east-west axis of the
range are a number of faults, including the major Denali Fault. Rocks
located north of the Denali Fault are primarily Paleozoic and older meta-
morphosed sediments, while rocks south of the fault are " ... characterized
by a monotonous sequence of predominantely dark gray argillite, slate,
graywacke, and a few intervals of limestone of late Paleozoic(?) and
Mesozoic age 11 (USDI 1974:67).
42
The Paleocene age Cantwell Formation, the major formation under-
lying the study area, was deposited in a large continental basin and consists
of metamorphosed conglomerate, sandstone, siltstone, shale, coaly shale,
coal, and argillite (Wahrhaftig 1958; Wolfe and Wahrhaftig 1970). Approx-
imately 60% of the Cantwell Formation is comprised of sandstone and conglom-
erates (Wahrhaftig 1958:9). The upper portion of this formation contains a
number of volcanic flows and tuffs (USDI 1974). The Cantwell formation
extends over an area of some 3500 km 2 ; within the Nenana River Valley, from
Clear Creek to the McKinley Park Station.
Quaternary Geology
Methodology
The recent geologic history of the Carlo Creek region was studied
through field, laboratory, and library research. The glacial geology .and
selected aspects of the regional geomorphology have been summarized from
previous published work (Wahrhaftig 1958; Thorson and Hamilton 1977) and
through informative communications with the geologists T. D. Hamilton, R.
D. Reger, N. Ten-Brink, and R. M. Thorson. Although additional work has
been done in the area since 1978 as part of the North Alaska Range Early
Man Project, these data have not been made available to the writer and have
not yet been published.
Reconnaissance mapping of the Carlo Creek study area was made
possible through the use of USGS and Alaska Department of Highways aerial
photos, as well as USGS 1:63,360 topographic maps. A total of eleven
Brunton compass traverses were made of the post-glacial terraces which are
found on both sides of the river.
43
Most of the geologic data for this report were obtained from the
detailed examinati~n of the Carlo Creek site. Seventeen sediment samples
from the three main areas tested in 1976-77 were analyzed. Samples were
collected in approximately 5 em increments, within each of the coarse and
fine textured laminae at the site. The samples which were later selected
for analysis represent increments of approximately 20 em and are fully
representative of the inter-site variability and range of grain-size,
sorting, and bed morphology. The results of the laboratory data are
presented in Figures 12, 13, 14, 15, and Tables 2 and 3.
44
Laboratory analysis of the samples first involved drying at 45°C,
weighing out of 85-100 gm split samples, then pretreating to remove carbon~tes
and organic matter. A 10% dilute Hcl solution was used to remove carbonates;
this was followed by treatment with a 30% H2o2 solution to remove organic
matter. After each step, the sample was dried at 45°C, then weighed to
record weight loss. Proportions of carbonate and organic matter in the
sample are expressed in terms of the original dry sample weight.
Following pretreatment, each sample was wet-sieved through a 4 phi
(0) screen (0.0625 mm), to reduce the total volume of silt-clay fraction in
the dry sieve procedure, as well as to separate the fine fraction for later
hydrometer analysis. The amount of silt-clay fraction was determined by
drying and weighing the sand fraction remaining in the 4 0 screen, then
subtracting from the total after pretreatment.
Each sample was then disaggregated with a calgon solution, dried at
45°C, and weighed. The fraction < 4 phi (obtained from the earlier wet
sieving), plus the pan fraction, were then combined for hydrometer analysis
of the silt-clay portion of each sample . Samples were analyzed in 1000 ml
cylinders with sample sizes restricted to the 20-30 gm range. The silt-
clay boundary, as employed here, is 8 phi (0.002 mm). Percentages of silt,
clay, sand, and gravel represent the total weight after removal of carbonates
and organics.
The analytical methods used here are based on those in use at the
Geoarcheology Laboratory, Washington State University (Hassan n.d.).
Calculations for mean, sorting, skewness, and kurtosis are based on Folk
and Ward (1957).
Glacial Geology
The surficial geology in the Carlo Creek region is the result of a
complex history of repeated glacial advances, followed by post-glacial
downcutting and aggradation. Interpreting the Quaternary glacial geology
is further complicated by the considerable amount of uplift that has occurred
in the tectonically-active Alaska Range (Wahrhaftig 1958).
Wahrhaftig (1958) first outlined the glacial sequence in the valley,
and recognized four major glacial advances: the Browne, Dry Creek, Healy,
and Riley Creek events. These deposits extended well to the north of the
present day limits of the northern foothills province of the Alaska Range.
The earliest of these is thought to be late Pliocene or early Pleistocene
(Thorson and Hamilton 1977).
The subsequent Dry Creek glaciation, regarded as broadly middle
Pleistocene in age (Thorson and Hamilton 1977), was less extensive than its
predecessor. An uplift of some 150m probably occurred in the vicinity of
Healy during and slightly after this episode (Wahrhaftig 1958).
The following Healy glaciation, regarded as broadly Illinioan to
early Wisconsinan in age (Pewe et al. 1965; Thorson and Hamilton 1977), re-
sulted in the formation of a large preglacial lake in the Nenana Canyon
45
46
(Wahrhaftig 1958), approximately 30 km downstream from Carlo Creek. Lacustrine
and alluvial fan deposits eventually filled this lake, and caused an eastward
displacement of the Nenana River. Coupled with this displacement also was
extensive downcutting into schist bedrock and formation of the present-day
Nenana Canyon. The Healy glacier terminated approximately 1.0 km south of
the Dry Creek archeological site (38 km north of Carlo Creek); the outwash
from the Healy moraine forms the base on which that site's eolian deposits
later accumulated (Thorson and Hamilton 1977:152-153).
The most recent major expansion of glacial ice in the Nenana Valley
was the Riley Creek event, first defined by Wahrhaftig (1958). On the
basis of outwash terrace profiles and drift limits, three advances of the.·
Riley Creek glaciation have been suggested (Wahrhaftig 1958; Thorson and
Hami 1 ton 1977) .
The Riley Creek I advance (see Fig. 8) probably extended into the
Nenana River canyon, and terminated approximately 2 km north of the McKinley
Park Headquarters, about 20 km north of Carlo Creek (Thorson and Hamilton
1977:168). The Riley Creek I glaciation also resulted in the deposition of
a kame terrace in a low pass 746 m a.s.l. (2450') in an area about 4.8 km
northwest of Carlo Creek. These deposits lie at an elevation about 152m
above the present level of the Nenana River. Outwash gravel at this point
is greater than 76 m thick. A "high terrace," traced by Wahrhaftig upstream
to the Carlo vicinity, is composed of till an.d gravel. The elevation of
this feature suggests that Riley Creek glacial ice was at least 1036 m
a.s.l. near the Carlo Creek area, and at least 416 m thick (Wahrhaftig
1958).
It is probable that the Riley Creek I glacial retreat resulted in
the deposition of large amounts of meltwater deposits. There is some
... ~
., .. ,. -
•1· 1o·
fXrlANAIION
f.fj ou.cua
II lltOIYf4AfiAIIOit
RC I Rlt £Y CMfiK 1 rHt:IAIIIIIt
RC II Mil IY L•tH II IHCIAIIOII
(I (ARI U Nl I\IIVI\Ht:l
-6]" JO'
t41.
Fig. 8.--Upper Nenana River Valley, Showing Probable Terminal
Positions of Late Pleistocene Glaciers. ADAPTED FROM: Thorson and
Hamilton (1977) and Wahrhaftig (1958).
•,
evidence for the formation of a large preglacial lake in the vicinity of
the Nenana River's confluence with the Yanert Fork (Wahrhaftig 1958). The
Riley Creek I event probably represents the maximum advance of late
Wisconsinan glaciation in the Nenana River Valley, suggesting a date of
approximately 18,000 B.P.
The major Wisconsinan event recognized in the research area is the
Riley Creek II glaciation (Thorson and Hamilton 1977), which terminated
approximately 2 krn south of the Nenana Canyon, or about 20 krn north of
Carlo Creek (Fig.8). Thorson and Hamilton (1977) and Thorson (personal
communication) have suggested that the Riley Creek II deposits, which
terminate in a SO m-high moraine near the entrance to McKinley Park, may .
correlate with the more firmly dated Naptowne drift of the Kenai Peninsula
(Karlstrom 1965; Schmoll et al. 1972) and the Itkillik II drift of the
Brooks Range (Hamilton and Porter 1975). If such postulated correlations
are correct, then the Riley Creek II maximum may date to about 13,000 to
14,000 B.P.
The Riley Creek I and II glaciations (Fig. 8) were probably both
fed by extensive firn fields in the Broad Pass-Susitna drainage areas,
48
which forced massive quantities of ice and meltwater northward, down the
Nenana River Valley. Due to the tremendous amounts of ice in the accumulation
zone, these glacial advances may have persisted well past the generally-
accepted dates of glacial retreat elsewhere in Alaska. Thorson (personal
communication) has suggested that Riley Creek II drift in the Nenana Valley
may actually be related to a glacial advance which is controlled more by
dynamic, rather than climatic causes.
This factor is important in interpreting the Carlo Creek regional
geomorphology. It may well be that stagnant ice and meltwater channels
continued to exist in the valley long after Holocene amelioration of climate.
Thorson and Hamilton (1977:168) suggest that the Riley Creek II glacier
stagnated across a 11
••• broad zone, that includes the lower 4-6 km of Yanert
Fork and adjacent parts of the Nenana Va 11 ey 11
• In these regions today,
fresh kame and kettle topography is common; such a landscape may have been
present for early Holocene hunters and their big-game food resources.
In order to determine the earliest possible date for human use of
the upper Nenana River Valley, it will be necessary to discuss the known
radiocarbon assays relative to deglaciation from the area. First of all,
it should be stated that precise dating of deglaciation in the valley is
equivocal at this time. Two radiocarbon dates for deglaciation are repre-
sented by assays of 10,560 ± 200 years: 8610 B.C. (W-49), and 9060 ± 160
.·
years: 7110 B.C. (AU-94), (Wahrhaftig 1958; Thorson and Hamilton 1977).
Hamilton, who collected the latter date, cautions that this type of limiting
date (kettle fill) may actually be thousands of years younger than actual
date of deglaciation, (Hamilton, personal communication, 1976). AU-94 is
thought to represent a minimum age for 11 Carlo 11 outwash (Thorson and Hamilton
1977:169). The older of the two dates indicates a period of ice wastage
after Riley Creek II advance but before Carlo time (Wahrhaftig 1958; Thorson
and Hamilton 1977). However, it should be cautioned that the 10,560 year
old sample was analyzed during the early years of the C-14 process, and
hence is based on the solid carbon method of dating. As a result, the date
may be somewhat suspect.
Further minimum limiting dates for deglaciation in the upper Nenana
River Valley are indicated by the author's C-14 dates from the Carlo Creek
site, located approximately 20 km up valley from the Riley Creek II terminal
moraine. The dates obtained from the Carlo Creek component I occupation
49
suggest that by at least 8500 B.P., more than 75 % of post-glacial valley
incision had already taken place. Since it is quite likely that initial
post glacial downcutting was fairly rapid, these dates may offer a good
estimation for the age of late glacial outwash in the upper valley. They
do offer nearly incontrovertible evidence that Riley II glacial advances
were completed by at least 8500 years ago, and probably 1500 to 2000 years
earlier. Hamilton (personal communication, 1977) has cautioned that " ... in
areas of unusual thick drift and massive stagnant ice, such as probably the
upper Nenana Valley, inactive buried ice probably would have persisted for one
to several thousand years longer, under probably mild climatic conditions."
Further supporting data bearing on the age of the Riley Creek II
retreat may be suggested by a C-14 date of 12,500 ± 150 years: 10,550 B.C.
(W-161), obtained from a basal kettle deposit in the adjacent Teklanika
Valley, located approximately 40 km west of Carlo Creek (Hamilton, personal
communication, 1977). This date appears to be somewhat more compatible, in
general, with the more firmly dated deglaciation record in other parts of
the Alaska and Brooks Ranges. If this date does in fact correlate with the
late glacial record in the Nenana Valley, then an age of between 12,000 to
14,000 B.P. may be ascribed to the Riley Creek II glacial maximum.
A third, minor, late Pleistocene-early Holocene event will be
discussed here in an effort to describe the regional geology in the Carlo
Creek area. The so-ca 11 ed "Carlo re-advance," however, remains somewhat
enigmatic. (For the purposes of this reptirt, the Pleistocene-Holocene
boundary is set at 10,000 B.P., as suggested by Hopkins [1975:10]). This
event was originally postulated by Wahrhaftig (1958) to account for a belt
of ice-marginal topography located between 0.5 and 5.0 km north of Carlo
Creek. According to his interpretation, the waning Riley Creek glacier had
50
retreated to its present position --nearly 75 km upstream from Carlo Creek--
before it again advanced. It then readvanced to a position just north of
Carlo Creek, fanning what he called the Carlo "moraine". This position
would have been 14.5 km south of the Riley Creek II moraine. As it again
retreated, meltwaters from the Nenana glacier excavated an 80 km long
canyon through Riley Creek morainal deposits and outwash, and then refilled
the canyon with gravel to the elevation of the original Carlo outwash
plain--a thickness of ca. 76 min an area about 4.8 km north of Carlo
(Wahrhaftig 1958:54). Thorson and Hamilton (1977:168) state that Carlo
drift is morphologically similar to Riley Creek II drift, and observe that
it fonns a wide tenninal zone about 3 km north of Carlo Creek. ..
As recession of the "Carlo" glacier continued, it was hypothesized
(Wahrhaftig 1958) that a moraine-dammed lake fanned, which extended southward
for a distance of some 20 km (see Fig. 8). Also fanned at this time was a
prominent outwash terrace, which lies at an approximate elevation of 750 m
a.s.l ., about 100m above the level of the Nenana River . This terrace is
composed entirely of gravel and interstitial sand (Wahrhaftig 1958) and is
capped with approximately 1.5 m of massive sand and silt of probable eolian
origin (Fig. 9; terrace 4).
Wahrhaftig (1958:54) suggested that the hypothesized preglacial
lake in the Carlo area could be explained by: (1) the flat valley floor in
the Nenana Valley between Clear Creek and Carlo Creek,(2) the steep valley
walls, (3) the "low gradient and gentle meandering flow of the river in this
stretch," and (t.'l,) (lacustrine) sand deposits. However, on the basis of the
51
Carlo Creek geoarcheological research, further supporting evidence for this
postulated lake has not materialized. The sand deposits reported by Wahrhaftig
(1958:53-55) as lacustrine appear to this investigate ~ to be fluvial in
EXP~NAT!ON
l-4 rerraces
A Alluvial Fan
F Floodplain
:!!i Bedrock
Data com~iied from Alas~a Deot. of
Hignways and USGS Air~hotos. Fi eia
cnecked via :hain ~ comoass survey.
Locations of t~rrac~s approxi~at~;
dashed where inferred.
~RS ABOVE FLOODPLAIN
0 ~ 0
~ .... ""
< ' '
<
<.:1 z: s <
'""' = ..... :a
:i:
~ -·-~
8-· ..
8 ,...,
"" = '""' ,_
~
g
"'
g -
Fig. 9.--Sketch Map of Terraces in the Nenana River Valley in the
Vicinity of Carlo Creek and Profile Along Line A-A'. Area enclosed by
rectangle shown in Fig. 4; locality "1-C" refers to C-14 sampling site
referred to in text, Fig. 16, and Table 5.
origin, as they show considerable cross bedding, ripple laminations, scour
and fill features, and have generally coarse textures. As an alternative
explanation, the features originally described by Wahrhaftig (1958) are
interpreted here as evidence for an early post-glacial period of fluvial
aggradation, occurring during minor stillstands in valley incision. The
basal Unit of the Carlo Creek site (Fig. 10; Unit 1) may have been deposited
as glacio-fluvial sediments during or soon after the ••carlo 11 event.
At the time of Wahrhaftig•s pioneering research in the Alaska
Range, no similar 11 Carlo 11 features could be found in any of the other major
valleys in the Central Alaska Range (1958:52-55). Subsequent research by
Thorson and Hamilton (1977:168) has also indicated that deposits correlati~e
with the 11 Carlo 11 event probably do not exist elsewhere in the Alaska Range.
Hamilton (1973:37) suggested one possible correlation between Carlo age
drift and deposits of the Donnelly III glaciation in the Delta drainage.
If such a correlation is valid, then an approximate age of 10,000 ± 500
B.P. may be posited for the 11 Car.lo 11 event (Thorson and Hamilton 1977:167-168).
Investigations carried out north of the Carlo Creek area by
Thorson (personal communication, 1976) have raised several questions concern-
ing the significance of the 11 Carlo 11 deposits. Possible differences with
Wahrhaftig•s interpretation are suggested by several tentative observations:
(T) similarities inmorphology between 11 Carlo 11 and Riley II deposits,(2) the
absence of a clearly-defined terminal moraine associated •.vith 11 Carlo 11
deposits, and (3) the apparent continuity between Riley Creek II and ucarlo 11
features (Thorson personal communication, 1976). To these points I would
add the previous observation concerning the lack of definite evidence for
lacustrine-preglacial deposits.
53
_.
w >
~ 622
~ 620
(/)
~ 618
0
~ 616
~ 614
w
._ 612 w
:.:i:
610
N CANTWEll ASH----..
C-< 3780180 B.P.)
VERTICAL EXAGGERATION
C=:J .... ~~~~~,
0 8 16 24 30
METERS
X 2
PM8
fig. 10.--Generalized Stratigraphy, Carlo Creek Site. Major stratigraphic units indicated
by numbers 1-4; granulometric samples indicated by "G~l" through 11 G-19". View to east from road
1 eve 1.
However, some mechanism must be set forth to account for the gravel
outwash terraces that can be traced downstream from the Riley II moraine to
as far as 23 km down valley from Dry Creek (Thorson and Hamilton 1977:168;
Wahrhaftig 1958:52-55). Thorson and Hamilton (1977:168) report that " ... near
the mouth of Dry Creek, the [Carlo] terraces stand about 7 m above the modern
floodplain of the Nenana River and are about 6 m lower than the Riley Creek
II terrace."
Alternative explanations for formation of the various "Carlo"
deposits may be: (1) the Carlo "moraine" may represent valley floor Riley
Creek II ablation drift (Thorson, personal communication, 1978), (2) these
mid-valley deposits may represent a minor stillstand in waning Riley II
glaciation, and (3) "Carlo" drift may have been the result of minor late-··
glacial surges of glacial activity, limited to the upper portions of cirque-
headed valleys feeding the Nenana system (Hamilton, personal communication,
1978). In addition, Hamilton (personal communication, 1978) has suggested
that a slight readvance in the Yanert Fork glacial system could have
-
caused aggradation of that stream, resulting in a rise in the local base
level of the Upper Nenana River. One final possibility could be a land-
slide, which could have temporarily blocked the canyon, raised base level,
and caused a lake to form (Thorson, personal communication, 1978).
Regardless of whether or not the "Carlo readvance" was a real
event, it appears that the major glacial retreat in the upper valley must
have been completed by at least 10,000 B.P. This does not preclude the
possibility for remnant quantities of massive stagnant ice which could have
prevented human use of the valley. The series of C-14 dates from the Carlo
Creek area sediments does indicate that the Nenana glacier had retreated
sufficiently to enable access to the valley floor by herbivores and human
55
hunting bands by at least 8500 B.P. By the time of the Carlo Creek I
occupation, it is probable that the Nenana Valley was sufficiently clear of
ice and multi-braided outwash channels to permit north-south movement
through the Alaska Range, by way of the Nenana Valley, Broad Pass, and
Susitna drainages (see Figs. 7, 8, and 9).
Holocene Geology and Stratigraphy
of the Carlo Creek Site
Following the final retreat of late Wisconsinan glaciers in the Yanert
Fork and Nenana Valleys, the valleys were subjected to episodes of extensive
downcutting, marked by minor stillstands and periods of aggradation. One
of the best records of these events is preserved as a set of four depositi~nal
terraces directly across (to the west of) the Carlo Creek site (see Fig. 9).
These terraces form a series of discontinuous benches that can be traced
for several kilometers on both sides of the valley.
The highest surface (not indicated in Fig. 9). represents an erosional
terrace which was probably scoured by Riley Creek II ice; this surface is
about 1000 m a.s.l. Below this surface is a prominent outwash terrace
(Fig. 9; terrace 4), which is approximately 700 m a.s.l., and about 80 m
above the level of the Carlo Creek site. On this "Carlo" age surface
(cf. Wahrhaftig 1958:53-54), the author has observed collapse structures,
kettles, and erratics as large as 5 m in diameter. These observations would
suggest a late Riley Creek II or "Carlo" age f~r the 700 m surface, or about
10,000 ± 500 B.P. The presence of glacier-margin features such as kettles
or erratics would indicate that stagnant ice may have been present during
the early stages of postglacial downcutting.
Downcutting probably would have initially taken place fairly rapidly;
however, the cause of the temporary halt in downcutting at ca. 8500 B.P .--
56
57
the age of the Carlo Creek site Unit 2 and 3 sediments--is not clear. Hamilton
(personal communication, 1973) has suggested t\-10 possible causes for
such a temporary halt in valley incision: (1) a minor readvance of cirque
glaciers, perhaps those at the head of Carlo Creek, Clear Creek, or Slime
Creek, could have supplied extensive meltwater sediments to the Nenana
River system, thus increasing the sediment yield and floodplain deposition
in the Carlo Creek region. (2) A minor readvance of the Yanert Fork glacier
could have caused significant aggradation of Yanert Fork, thereby raising
the local base level for the Nenana River.
Such postulated causes for the 8500 B.P. halt in downcutting may be
_supported by evidence from the Copper River Valley, where a date of 8800
B.P. was obtained in association with cirque glacier marginal moraines
(Sirkin et al. 1971 :708), and from the North Slope, where Detterman
(1964:130-131) has reported similar evidence for an early Holocene episode
of alluviation.
--
The Carlo Creek archeological site (HEA 031) occurs within sediments
comprising the third of the four post-glacial depositional terrace remnants
(Fig. 9; terrace 2). HEA 031 is about 623 m a.s.l ., and is located 20m
above the level of the Nenana River. The terrace surface below which the
Carlo Creek site occurs (terrace 2) can be traced upstream for nearly 3 km.
On both sides of the river this terrace, as well as one 6 m below jt (terrace
1), are preserved as a ·series of benches. In some places, these surfaces
have been dissected laterally by small tributary streams, and have in
places been mantled by alluvial fan sediments. The terrace surface at the
Carlo Creek site may correlate with a terrace 11 about 60 feet high 11
, located
upstream from Carlo, as reported by Wahrhaftig (1958:54).
The stratigraphy of the Carlo Creek site (Fig. 10) is made up of four
major sedimentary units: Unit 1, the bottom of the profile, contains a
minimum of 8 m of sandy gravel; Unit 2 consists of m of bedded fluvial sand;
Unit 3 consists of about 3-4m of floodplain silt and fine sand; and Unit 4
contains ca. 30 em of sand/silt--probably due to primary loess deposition,
reworked alluvium, and volcanic ash (see Figs. 10, 11, and Table 2).
A brief discussion of the grain size parameters for each of these
four major sedimentary units will be necessary to more completely interpret
the site 1 s past geologic history of the site. Granulometric analysis is an
invaluable tool in determining relative percentages of gravel, sand, silt,
and clay, and can ultimately provide important data bearing on the deposi:
tional environment of a site. Data relevant to these analyses are presented
in Tables 2 and 3, and Figures 12, 13, 14 and 15. Interpretation of
granulometric data are based on previous studies (Allen 1965, 1970; Folk
1966; Hassan, n.d.; Kukal 1971; Reineck and Singh 1965; Visher 1965,
1969).
The lowermost unit in the site, Unit 1, is extremely poorly sorted,
very negatively skewed, and mesokurtic, and has a median grain size of
0.7 phi (0). The poor sorting, and the presence of large clasts (up to
.75 m diameter) suggests that it was deposited as part of the traction load
within a major channel of the Nenana river or a tributary stream~ The
histogram (Fig. 13; Number 17) representing this unit tends toward bimodality,
a characteristic of channel bed deposits.
Unit 2, comprised of laminated silt and sand, shows some cross
bedding in places. Median grain sizes of the coarse and fine lamianae
samples range from 3.4 phi to 1.2 phi. Samples Number 11 to 16, collected
58
from a depth of 185 to 592 em, are generally poorly to moderately sorted, {
... CARLO CREEK SITE SECTION ALONG OW
0
...
-I
""'
-z
hearth ..... ,. td:l ,,
Fig. 11.--Stratigraphic Section Along Line 0 West.
LEGEND :
B A,_llfAC1S
i!!l CHA.COAL
liiil PALtOSOL
18] OAANULOM( JIUC IAMP'l(
CfAOM SOU1H WALL.
IQUAft( O -IN, 0 -IW~
<omponenl II
(.J1
\.0
Unit
a
4 b
c
3
2
TA BLE 2
GENERALIZED DESCRIP TI ON OF STRATIGRAPHIC UNITS
CARLO CREEK SITE
Apprcx. depth
below surface
0-30 em
30-200 em
200-600 em
600 + em
Descriction
Modern organic mat and acti ve pedogenic horizons.
Silt to fine sand with many very fine to medium
rootlets; contains charcoal and burned roots i n
and above A2 hori zen. Severa 1 diffuse pa 1 eo sols
evident. Unit 4b reoresents a 5-15 mm thick dis-
continuous lens of volcanic ash, found about
20 em be 1 ow surface. Abrupt, 'Navy, uncomfortab 1 e
boundary.
Laminated fine sand and silty fine sand. Fi ne
l aminae consist of very dark gray (10YR3 /l, mo i st )
s il t /clay . Coarse laminae consist of very dark
grayish brown (10YR3/2, moist) f in e sand. Bedding
i s near ly horizontal, with an in creasing dip (max.
of 25°) toward the exposure face (west ). Lamina-
t i ons vary between 2-5 em and tend to thin upwards.
Two archeological ccmoo"nents occur within this
unit: The upper cultural level, Component I I
(lithics) occurs 100 to 110 em below surface, and
about 100 em above lower contact. Component l
(l ower ) cultural materials (lithics, bones and
charcoal) are found between 0-5 em above the l ower
boundary. Fairly uniform scattering of t erres-
trial mollusca occur in lower 50 em of unit 3.
Or9anic stains and/or weakly developed paleosol
(?)occur between 0-5 em above l ower boundary .
Minor car bonate accumulat i on occurs on the under-
side of Com conent l artifacts . "Paleosol" exhibits
'Neak, fine c:-um b structure, 'llith moderate cem enta-
tion; and soft, very fr i able, very sti cky and very
plastic consistence. Paleosol appears as 3 wavy
bands not more than 5 em thick (total ), and ranees
from dark reddish brown (2.5YR 3/4 , moist ) to dark
brown (1 0YR3/3, moist ). Boundary: abrupt, wavy,
uncomfortable.
Bedded sand with occasional fi ne gravel near base.
Fluvial cross-bedding ev id ent i n places. La minae
are nearl y parallel. Grains are ;ubrounded to
rounded and are very dark gra yi sh brown (2.5 YR3/2,
mo i st ) t o dark grayish brown (lOY RS /2, moist ).
Thi ckness of l aminat i ons· range f rom 5-10 em near
base of unit, to 2-5 c~ near top. Boundary:
abrupt, smooth, uncomfortable.
Gravel and sand. Rounded to subrounded cobbles
to s~all boulders, i n sandy matrix, ~oorly sorteM.
Lenses of sand and gravell y sand occur th rougnout .
60
.·
TABI.E 3
CARLO CREEK SITE GHA NULOHETRI C ANA LYSIS
. ----------------------------------------------·-------------------------------------------------·---------------------
Oepth 1 Organic Median
lin it (en~) Conouents l.ab I % Gravel % Sand % Sll t % Clay Matter 1. Caco 3 IJ Sorting Skewness Kurto s is
-----------------------------------------------------------------_____________________ .. ____ ----·-------
4c 9 -12 AI horizon 0 26.43 74.15 0 28.61 0 4.4 0 .59 0 .16 1.07
4b 19-20 Ash l ens 2 0 24 .48 72 .37 3. 71 19.50 0 4 .5 1 .27 0 .29 1.46
4a 22 -25 8s hol"lzon 3 0 40.52 59 .22 0.28 6 .79 0 4 .I 0 .62 -.16 I .27
3 42 -26 C hort zon 0 73.86 22.79 2.65 1.19 0 3 .4 1.12 0.21 1. 79
62-65 5 0 53 .70 44.76 1.35 0 .22 0 4 .0 0 .80 -. 21 1.00
81 -115 6 0 42 .72 55 .49 0.99 2 . 011 0 4 .I 0 .80 -.14 1.1 5
100 -104 f.omp. II 0 44 .69 55.12 0.28 0 .16 0 4 .0 0 . 72 0 .02 1.14
121 -124 8 0 62 .00 35 .21 3 .20 0.24 0 3 .8 1.06 0 .22 2 .14
141 -144 9 0 .06 34 .39 64 .58 1 .20 3 .53 0.04 4.2 1.05 0 .13 1.19
168 -171 Comp . I 10 0 57.42 41.77 0.80 4 .06 3 .64 2 .6 1 .28 0 . 73 1. 52
2 181 -18 5 11 0 .17 70.04 24 .29 4.87 0 .311 0.4 2 3 .1 2 .05 0 .23 1 .19
197 -200 12 0.17 75.00 21 .37 2.52 0 .68 0 2 .8 1.73 0.17 1 .20
248-252 13 0 .14 80.67 19 .70 0 .09 0.01 0 2 .2 1. 79 0 .26 1.10
273 -277 14 1. 2 5 92 .96 5 .81 0.05 0 .14 0 1 .2 l .H 0 .26 I .25
533 -537 15 0 97 .53 2 .41 0 0.03 0 2.3 0 .66 0.04 1. 23
588 -592 16 0.05 71 .09 27 .74 1.12 0.27 0 3 .4 1 .44 0 .14 1.72
618 -622 17 27.39 66 .94 5.43 0 0 0 0 . 7 2. 71! 0.20 1.09
----------------------------------------------------------------------------'------·---------------------------------•
m
~
-.:,.., ...... _.,
ow
IW
4 os/1
0
3
compii -
3
conp. I -
MAJOR SJAATiflRAPHIC UNITS
0 10 zo
% OIIAV(L
0 2 0 10 60 80 100
"4 SANO
I . .._ : }_1
•,... f ·, ~ ... ~· ,. ' --------
I I I
40 10 10
% llll
0 II
%CLAY
'•
--~ '> .( ___ -----__ L __
.II 1.0 15 2 -2 0 4
MfOIAN , SORJINO Skl Wll£55
10 11 I 8
~UR TOSIS
% CARIONAff
~ PMB · 78
Fig. l2.~~Carlo Creek Site Granulometric Analys1s.
-....
..... z:
'""' :...> ::: ....
~
ao Sample l: 9-12 c:n ao Samp 1 e 5 : 62-65 em
60 60
40 40
20 20
0 0
8 3 10
80 Sc!J!lple 2: 19-20 c:n (asn) ao Samo i e 6 : 81-85 ~~
60 60
40 40'
20 20
0 0
3 4 5 6 7 8 9 10 1 2 3 4 5 5 7 s 9 10
ao Sample 3: 22-25 em ao Sample 7: 10 0-104 c:n
60 60
40 40
20 20
0 0
2 6 7 8 9 10 2 3 <1 5 5 7 3 9 10
so Sample 4 : 42 -46 c:n 30 Sample 3: l21 -124 c:n
50 50
40 ~0
20 20
0 0
I ' '
-4 -3-2 -1 0 1 2 3 4 5 6 7 a 9 10 --4 -3-2 -1 0 1 2 3 4 5 6 7 3 ') l'J
DIAI11ETE:::! :N PHI UNiiS
Fig. 13.--Grain Size Distributions, Carlo Creek Site.
Histograms represent diameters in phi units vs. percent of total
sample after removal of carbonates and organic matter. Refer to
sample locations in Figs. 10. 11, and 12. Sample depths are in
reference to subdatum.
63
--
64
dO Sample 9: 141-144 em 30 S<impl e 13: 250 ::m
50 60
40 40
20 20
0 0
-4 -3-2 -i. 0 1 2 3 4 5 5 a 9 10 -4-3 -2-1 0 1 2 3 4 5 6 7 3 g 10
ao Sample 10: 167-171 (component l) 80 Sample 14: 275 em
60 60
40 40
-M
20 20 ---0 u z:
""" -4 -3-7 8 9 10 --+ -3-2-1 0 1 2 3 ~ 5 6 7 a 3 10 c:::
""" • 30 Sample 11: 181-185 Cll1 30 Sample 15 : 535 em
60 60
40 ~0
20 20
0 0
--+ -3-2 -1 0 2 3 4 5 6 7 3 9 10 -J. -3-2 -1 J l 2 3 4 5 5 7 3 9 10
-30 Sample 12 : 197-200 c:n 30 Same i e 17: 5~~ em
50 60
40 40
20 20
0 0
-4 -3 -i:-1 0 1 2 3 4 5 6 7 8 9 10 ..J. -3-2 -1 J 2 3 4 5 3 :1 :o
DIAM ETER [N PHI UN I iS
Fig. 13. --Continued .
u
.
U,,
~ -4 -1-z:
;j -5
!::::
U,,
Q..
1--6
V')
!:::: -1.1..
1
7 / J ~r,,
10\ ,...,.'; a. ~· 'I '~2 / • • 6
, --~' II I 5 I
, , \ 12 ' I 3
,. , , 15 \ • • 11 \ 16 I
, 13. • ' Y I
(.14 I.I ' .... ~-/
' --------..
KEY:
III = STRATIGRAPHIC UNIT
(see fig. 10 and table 3).
17 = GRANULOMETRIC SAMPLE
(see fig. TO and table 3).
1 2 3 4
MEDIAN DIAMETER fN PHI U ~IT S (M)
Fig. 14.--C/M Diagram, Carlo Creek Site Sediment
Analysis. Refer to cumulative grain size curves, Fig. 15.
65
99
95
~ U4
~ 75 -
1:>! w n.
1-ffi !iO .... w ::c
5
I -
16 8
--,
-5 . -4
OIAHHER IN ~It
4 2 0 0 .25 0 .125 .0625 .0313 .0156 .0078 .OCI39 .0020
]
4
-]
IIIAHEHil IN Pill UNITS
f1g. 15.--Cumulative Grain Size Curves. Carlo Creek Site. Samples l-6.
OIAHETER IN Pft
10
99 99
95 95
!z 84 ll4
..... 75 u 75 ~ .....
<>-
1-
50 X 50 l!J .....
~
w 25 25 > .....
1-16 16 :5
::>
~ u
5 5
OIAHETER IN Pill UNITS
'•
99
95
!Z 84 w u 75 0::
lot
Q.
1-
X
f.!) 60 ~ u.
3 ....
> ..... 25 -1-
:3 16 => ::l: => LJ
011\t-IH(R IN 1fl
16 8 4 2 0 0.25 0 .125 0.0625 0 .0313 .0156 .0078 .0039 .0020
t-_ _J_ __ L __ ~_I __ J __ _J_ _ __. ___ _J_._ _ _.._ _ ___.~___t __ J __ ...__ _ _._ _ __. _ ___,
--,--,
-5 -4
I
-3
,-
-2 -1 -01 2 3 4
011\HHER IN Pill UNITS
Fig. 15.--Continued. Samples 13-17.
16
5 6 7 ll
-99
-!\!1
84
75
!)0
25
16
5
m co
are nearly symetrically skewed, and are mesokurtic to platykurtic. On the
basis of field observations and the granulometric data, it can be inferred
that these bedded sediments were probably laid down as channel bar deposits.
The field data do not indicate the three dimensional geometry characteristic
of levee deposits (Allen 1965:146). They do, however, show some interbedded
and alternating coarse and fine laminae, another characteristic of levee
deposits (Allen 1965). The probably-rapid transition from channel (Unit 1)
and bar deposits (Unit 2) in the Carlo Creek may be attributable to westward
migration and/or minor downcutting of the channel, away from the site area.
Unit 3 consists of laminated silts and fine sands, which indicates
a rather complex micro-stratigraphic sequence of considerable variability. :
(Cultural Component I occurs at the base of Unit 3; Component II was found
within the upper 50 em of this unit). Median grain size of samples 4
through 10 ranges from 2.6 phi to 4.2 phi (Fig. 12). All samples were
either moderately sorted or moderately well sorted. Measures of skewness
indicate positively skewed samples (3), negatively skewed samples (2), and
nearly symmetrical samples (2). The samples are generally leptokurtic to
very leptokurtic. These data, together with the field observations, would
tend to indicate an actively aggrading floodplain as the Unit 3 depositional
environment. The variability of the samples is attributed to: (1) variation
in stream flow velocity during high water stages and (2) eolian reworking of
exposed alluvium. A number of authors (Allen 1965, 1970; Leopold et al.
1964) have discussed the complex nature of floodplain environments, where
readily available, seasonally replenished overbank sediments are reworked
through the action of wind and water.
Also evident in Unit 3 are zones of gleying, mottling and organic
accumulations (see Fig. 12 and Table 3). These postdepositional alterations
69
within Unit 3 offer further evidence for a floodplain depositional environ-
ment. These features indicate that the early stages of Unit 3 development
were marked by long seasonal periods of standing water (pending) and anaerobic
conditions . As indicated on Table 3 and Figure 12, the lower 30 em of
Unit 3 also contains a small percentage of Caco 3 , which is probably derived
from fossil snail shells occurring in the lower 50 em of Unit 3.
The uppermost unit at the site, Unit 4, is primarily eolian in
origin, although it has also been affected slightly by downslope movement.
Unit 4 sediments, with the exception of the tephra unit (Unit 4b), have a
median range of 4.1 to 4.5 phi, are moderately to moderately well sorted,
are symmetrical to positvely skewed, and are mesokurtic to leptukurtic.
The pronounced increase in silt, combined with the good sorting _, indicates
quite clearly that these deposits are of eolian origin. The cumulative-
frequency grain size curve agrees well with curves for Alaskan loess sum-
marized by Pewe (1975b:39).
In interpreting the site stratigraphy, it appears the sedimentary
record can be best interpreted in terms of a preglacial braided river
sequence, which reflects a change from a channel bed, to a channel bar, to
overbank flood deposits. A thin-eolian cap derived from a local source was
later deposited on top of the fluvial sequence. Median grain size shows a
fining trend upward, with a concommitant improvement in sorting. Based on
field observations, granulometric analysis., and C-14 dating, the following
explanation is offered for the depositional history of the site. It is
recognized, however, that alternate explanations may be equally valid,
especially as new data are collected and interpreted.
By approximately 8500 radiocarbon years ago, the Nenana River had
incised its valley floor, from a level of approximately 700 m to 620 m.
70
This incision must have required about 2000 ± 500 years to complete. At
least one intermediate terrace (Fig. 9; terrace 3) was formed between the
Carlo Creek terrace (terrace 2) and the Riley Il (700 m) level surface
(terrace 4). By 8500 B.P., an active channel of the Nenana River was
depositing channel bed material (Unit 1) at the site locality. These large
clasts--some as large as 0 .75 min diameter --were deposited while the
active channel was at least 18 m above its present level, and while it was
at least 0.8 km to the east of its present position.
As the river continued to downcut and migrate westward, the depo-
sition of coarse gravels and sand was replaced by Unit 2 cross-bedded
sands. Observations of bed morphology, texture, structure, and statistical.
parameters of this unit suggest it was laid down as a channel deposit
(braided stream), within the preglacial Nenana River system, or possibly as
part of a tributary stream flowing from the east.
Eventually, the downcutting and lateral shift of the river reached
a point where major bar deposition ceased at the site . A period of relative
stability followed --perhaps as long as 500 years--which saw incipient
soil development (inceptisols) on this exposed surface. At some point
during this probably-brief hiatus in downcutting and deposition, the first
cultural occupation of the site took place, as represented by Component I
(Figs. 10 and 11). Remains of this 1 ower occupation--1 it hi cs, bone and
charcoal--are consistently found to occur within a 5 em zone above the
abrupt unconformity which separates the Unit 2 sands from the overlying
Unit 3. This unconformity has been traced for 1.5 km up the valley in
roadcut sections morphologically similar to HEA 031 (see Fig. 16). The
alteration of the old surface at this level probably represents a paleosol,
but may alternatively be explained as drifted-in detrital or ganics, deposited
r
71
lA
8400.200
osgot :no~~R~"ii
10,040*440
IB IC 2
79.40t 130
Fluvial
Fluvial
Fluvial
Eolian
Moll led
3 -----
tXPLANATION
grovel -
tond
alii
tilt
alit ond cloy
4 ---------e:J Om
.:0"·
-I
r .. .. ..
2 E ..
0
0 ..-
3 ..
::J ..
• 0
4 li
.0
&. ... e
5 0
Ttphro -6
Organic horlaon
Cultural horl1on ,
Talut
Fig. 16.--Correlation of Dated Stratigraphic Sections, Upper Nenana River
Valley. Refer to locations in figs. 2 and 9. ADAPTED FROM: Bowers (1979a) .
• •
during the early phases of floodplain sedimentation. On the basis of
radiocarbon dates and stratigraphic similarities between the Carlo Creek
site and Section 1-C (see Fig . 16), located 0.8 km south of the site, it is
inferred that the river's halt in downcutting lasted at least 500 years.
During much of that time, the main channels of the river were probably in
the eastern half of the valley (refer also to discussion in Chapter 5).
73
After initial human occupation and abandonment of the site, deposition
of Unit 3 sand/silt occurred. Unit 3 is a complex micro-stratigraphic
sequence, consisting of 3 to 4 m of finely laminated silt, fine sand, and
clay. These bedded laminae appear to represent deposition primarily by
fluvial processes (aggrading overbank sediments, derived from either the
Nenana River or a tributary). To a lesser extent, wind-reworked alluv i um
is found.
:
Within the lower 50 em of Unit 3 is a fairly high concentration of
two species of terrestrial molluscs: Succinea sp. and Discus cronkhitei;
which are the most likely explanation for the presence of carbonates in the
lower 30 em of Unit 3 (see Table 3 and Fig. 12). Also contained within
Unit 3 sediments is evidence of a second, albeit brief, human occupation.
Component II of the Carlo Creek site is located about 1 m below surface,
1 m above the Component I level, and 80 em below the top of Unit 3. This
horizon consists of a small amount of lithic debris, and suggests a short
period of human occupation. At the time of Component II occupation, t he
site was probably within an actively-aggrading floodplain.
Between Units 3 and 4 is a distinct unconformity. Unit 4 represents
the uppermost 30 em of the profile, and was most likely formed by eolian
actively, through the reworking of locally available alluvium. The eolian
origin of this unit is suggested by its relatively massive s tructure (as
opposed to the laminated silt/sand of the underlying Unit 3), by its
relatively better sorting, and by a noticeable increase in the percentage
74
of silt in the sediments. Unit 4 includes the zone of modern soil development
at the site, as well as several diffuse paleosols. Modern soils at the
site are Subarctic Brown Forest Soils (orthods) which are common in the
boreal forests of interior Alaska.
Unit 4 has been subdivided into three subunits (a, b, and c), to include
a discontinuous lens of volcanic ash (Unit 4b) which is found about 20 em
below surface. Unit 4b, which has been correlated (Fig. 16) with the recently
reported "Cantwell Ash Bed 11
, (Bowers 1979a) is radiocarbon dated to slightly
less than 3780 ± 80 years: 1830 B.C. (WSU 1747). The Cantwell Ash has beftn
previously described in terms of distribution, age, morphology, petrology,
mineralogical composition, and refractive index of glass shards (Bowers
1979a).
The date of 3780 B.P. for the deposition of Unit 4b provides an
approximate age for the beginning of Unit 4a deposition. Such a date
accords well with similar basal dates on loess sections and peat caps in
central Alaska (Hamilton and Robinson 1977; Hamilton 1973). Extensive
loess accumulation and peat formation may be due to the effects of late
Holocene Neoglaciation and cooling, which has been widely recognized in
alpine regions of Alaska. Hamilton and Robinson (1977:1003) suggest that
the interval of about 4500 to 3500 B.P., climaxing at about 3000-2500 B.P.,
was'' ... marked by alluviation of glacier-fed streams, widespread cccumulation
of peat, intensified ice-wedge growth, and solifluction .... " Hamilton (per-
sonal corranunication, 1978)has suggested that, in the Carlo Creek area, large
amounts of coarse-grained gravel fed into the Nenana River system would
tend to decrease or stop the river's downcutting. This in turn could cause
a cessation of slope steepening and of erosion, as well as renew deposition
on the site through the accretion of readily-available sand and silt from
the exposed floodplain. This period of slow cliff-head loess deposition has
continued up to the present time.
In addition, the fact that the tephra increases in abundance as one
digs into the hillside (away from the bluff exposure) indicates that the
hillslope morphology at the site was established in more or less its present
configuration by at least 3700 radiocarbon years ago. The thickening of
the ash into the hillside suggests that ash deposited nearer the bluff edge
was subsequently translocated by eolian activity, whereas the ash farther
in the hill was somewhat better protected from wind.
The ensuing 3000-4000 years at the Carlo Creek site have probably been
marked by minor deposition of airborne particles derived from the Nenana
75
River floodplain, lateral incision of the bluff to the north by Carlo Creek
and to the south by gulleying (see Fig. 10), and erosion of the western part of
the site by the Nenana River. Podzolization, humus accumulation, and other
soil processes have undoubtedly been active.
The major postdepositional alteration which has affected the site appears
to have been a westward dipping of the deposits due to mass movement. In
places, (e.g., closest to the bluff edge) this dip is as great as 20 to 25
degrees. A likely explanation for this deformation is that the northward-
flowing Nenana River truncated the hillside at Carlo Creek, which initiated
a gradual downslope movement of the sediments. An important observation to
be made in this regard concerns the three dimensional geometry of the site
stratigraphy: as one digs eastward into the hillside, away from the river,
the dip angle of the laminae decreases. This, coupled with the fact that
microfaulting increases in a westward direction toward the bl~ff edge,
suggests that the 20 to 25 degree dip can be best explained as a secondary
deformation, rather than as a primary deposit such as an alluvial fan.
There is no evidence to suggest that these deformations have significantly
altered the relative intrasite spatial relationships of cultural and faunal
remains. There were no other major observable deformations due to frost
activity, such as cyroturbation features, ice wedges, or solifluction lobes,
nor did it appear that forest activity (e.g., tree fall) significantly altered
the deposits.
The final major alteration which affected these strata was the
construction work of the early 1970s which first exposed these sections.
It is not known how much of the site 1 s sedimentary and cultural record was
destroyed as a result of these activities.
76
CHAPTER 4
CULTURAL COMPONENTS
Component I
As I have discussed in the preceding chapter, two distinct cultural
components were discovered during the excavations of the Carlo Creek site.
Of these two, the lowermost (Component I) has produced the most significant
cultural data . ..
On the basis of the known distribution of cultural materials, the
Carlo Creek Component I occupation appears to have been that of a small
group. The entire known horizontal area covered by Component I human
debris is 10 m x 5 m. This does not necessarily represent the boundary of
the Component I level, because the western part of the site was destroyed
during highway construction, and excavation into the eastern portion of the
hill was restricted by manpower limitations, frozen ground, difficulty in
disposing of backdirt, and problems of shoring unconsolidated sediments.
The vertical range from which materials were recovered in the lower
component was less than 5.0 em. Materials were found to occur either
directly on the surface of Unit 2 sands, or within the paleosol which makes
up the basal 5 em of Unit 3 floodplain silts and fine sands (see Fig. 11).
77
It is evident that the cultural materials were deposited directly on the
surface of Unit 2 sands and later incorporated into the organic layer which
formed on this stabilized surface. The horizontal distribution of cultural
materials--bone, charcoal, and lithics --indicates they ar ~ clearly associated
with two hearth areas (see Figs. 17 and 18). Based on stratigraphic and
radiocarbon dating evidence, it is believed that these two features are
closely contemporaneous, most likely related to the same occupation.
The first hearth encountered during the 1976 excavation season is
designated as Hearth 1 (Fig. 18). Lithic debitage, ground squirrel bones,
caribou and sheep bones are scattered in and around this feature, covering
an area of about 4 x 5 m, and centered at a point about 2 N, 0.5 W. The
exact limits of the hearth were difficult to define; charcoal was concentrated
in an oval configuration measuring approximately 1.5 m x 0.75 m. Most of
the charcoal appeared as small pieces, with a few well preserved twig-sized
pieces (willows?) measuring as large as 4 x em. Most of the Hearth 1 wa~
apparently built directly on the ground surface, with little evidence of
intrusive pits. However, there is one part of the hearth, designated
Feature 1 (see stratigraphic section, Fig. 11), which was unquestionably dug
into the underlying sediments, causing local disturbance of the sedimentary
contact between Units 2 and 3. This feature will be dealt with in Chapter6,
in the discussion of possible thermal alteration of lithics. Whether or not
the observed scattering of charcoal in the Hearth 1 area reflects aboriginal
activities is equivocal; it is possible that post-occupation seasonal flooding
or even wind could have caused a minor scattering of charcoal. Judging by
the northwest to southeast trend in the axis of the charcoal, the expected
direction of a paleo-flood current, this is a definite possibility. The
amount of charcoal displacement, however, was probably less than 0.5 m. Two
charcoal samples have been dated from Hearth one: 8400 ± 200 radiocarbon
years: 6450 B.C. (WSU -1700), and 10,040 ± 440 radiocarbon years: 8090
B • C • ( GX-51 31 ) .
78
2W
3~1
511 4H JH 2N IN 011
1£
l3N 12N liN lOH ... ,, 114 .•1
"·"' SUUIJATUH (0/0)
41 l /11 II kl 14.
I!>N 14H ( .... l) IS 2S
9H 011 7tl 6H IU II ... ~
------------------·---·--+-~ IW
14M ( 11·1'11 •~• I 1111 !•) "" ( .... IU (~ •I (II •I Ill
·--. -----· ·----1---~ ---1----o 2W
2H ,., l l U/ II
I" II (H .~l (!i .l) .... J~., ao •I
----
~-----------· --------4 (!.I 41
·------------L--·---
J CIU 1 1 Ill (•4;·.l) lUi (14 .1•1 --------
3W
Fig. 17.--Limits of Excavation and Flake Distributions for Component I. Upper numbers
within squares indicate total count per one meter square; numbers in parentheses indicate
\'Ieight in grams per one meter square. Refer to site topographic map, Fig. 4 .
• •
9 N IN
+ +
+ +
+ +
N-«:1
7N
+
+
15.---
•
+
6N
+
+
[:}. +
5
SN
+
+
0
COMPONENT 1
LEGEND:
JN 2N lN ON 4N
+ ~--L-------~------~------4-lE
~,--+,,
, I
.,.,.. I I
1 // • ,' '•
I \
I • / ,.\ •
• ,+ _____ ,/~-• ~',
,.. .... -... ,~ ', I ., ' \ \
I "V \ , •A '
onvlleton'~ _ \ ·~/'· ',,_
I \ \ ~ '
+ '+ • L '--•+ \ ._ ' .. ., -'
+
.... \ . .. \ -,, [:}. / \ . .. , __
' . ' ',....._ I
.._ ____ ---
+
M 3
' '
lS
Datum+
2S
tow
L-------L-------, 1 w
+ .--------__....~ 2W
' JW
975 collection
Artifacts & Faunal Remains
• · Large mammal bone • . Cite II us bone
6. · liface
0 • Blade fragment
@ • Cranial frag1.
~ · Pouible bone tool
r-... \ •
'' .... } Flake concentration
..
Fig. 18.--Component I: Distribution of Artifacts, Faunal Remains,
and Hearths. Refer to site topographic map, Fig. 4. en
0
The scattering of faunal and lithic remains about Hearth 1 can be
interpreted as indicating an activity-specific locus, representing two
major activities: bone processing/butchering, and biface reduction. As
illustrated in Figures 17 and 18, lithic debitage and associated biface
fragments appear to cluster into at least four groups. Judging by the
quantities of flake debris and broken, unutilized bifaces scattered in this
area, these remains probably represent several separate biface -reduction
"events", closely spaced in time, as opposed to a single flintknapping
activity. As discussed in Chapter 6, these prehistoric actions were
probably aimed at producing "tools of the moment", which could be quickly
manufactured from local materials, and which could have quickly been put to
use in butchering/bone processing. Several biface fragments, as well as 3
fragments of a large prismatic blade, were found within the limits of
Hearth 1, suggesting they may have been "tossed" in, perhaps in a manner
:
similar to that described by Binford (1978b). Also located in the activity
area surrounding Hearth 1 are two possible bone points (see Fig. 22), which
may have been brought into the site within the carcass of a caribou or sheep
that had been killed nearby.
There are two major concentrations of fragmentary and splintered
81
bone remains (see Fig. 18) in the Hearth area, one located in squares 1-2 N,
0-1 W and 2-3 N, 0-1 W, and the other in 1-2 N, 0-1 E. It is suggested
that these represent butchering, marrow extraction, and/or bone grease
preparation areas (see discussion in Chapter 7).
A cobble manuport (Fig. 20) located in square 2-3 N, 1-2 W, is
interpreted as having served as an anvilstone on which some of these bones
were broken. (The reader is referred to excellent discussions of bone
alterations by human agencies in Bonnichsen 1973, 1978, 1979 j
Binford 1978a; and Sadek-Koors 1972). This specimen does not exhibit any
battering on the margins, which would alternatively suggest its use as a
hammerstone; this had been the original interpretation (Bowers 1977a:12).
It does not lie in direct association with the nearest splintered faunal
remains. It is thus quite probable that this manuport was discarded into
the "toss zone 11 of site use in a manner described ethnographically by
Binford (l978b).
A second hearth area (Fig. 18) was partially excavated near the end
of the 1977 field season. This feature, located primarily in the western
half of square 6-7 N, 1-2 W, appears to be of roughly the same size as
Hearth 1. The eastern portion of the feature was not excavated. There are
no indications that Hearth 2 was excavated into Unit 2 sediments; conversely,
it was probably built directly on the surface of Unit 2 sand surface.
There are no major bone-processing areas associated with this feature,
except for a few scattered ground squirrel teeth and mandible fragments.
The heaviest concentration of lithic debris in the site, 1487 flakes per
m2 , occurs within 1 m to the north of Hearth 2 (Fig. 17). The distance from
Hearth 2 to Hearth 1 is 4.5 m. The one radiocarbon date from this feature
indicates that it is closely contemporaneous with Hearth 1: 8690 ± 330
radiocarbon years: 6740 B.C. (GX-5132).
There is no evidence in Component I for structural features, burials,
or cache pits. Except for the suspected heat-treatment pit in Hearth 1,
there is little evidence for surface (i.e~ top of Unit 2) disturbance by
the site's prior occupants.
82
Component I: Lithic Artifacts
The lithic artifacts from the lowermost cultural level at Carlo
Creek are described individually in the following section, and are summarized
in Table 4. Observations concerning possible cultural affiliations with
these specimens are reserved for Chapter 9.
UA 76 212 1 (Fig. 19, H), is a biface fragment of very dark grey
(Munsell value= 3) argillite. The specimen is broken by a transverse
fracture, possibly due to endshock. Considerable stepping and hinging is
evident on both faces. There is no discernible use-wear or intentional
platform preparation on the lateral margins. Flaking technology was most
likely by direct freehand percussion, as indicated by the non-patterned
flake scars. The artifact is bi-convex in cross-section, with fairly
symmetrical lateral margins. Length: 912 em (broken); width: 6.0 em;
thickness: 1 .9 em; weight: 119.2 gm; provenience: 6.98 N, 2.22 W; 230 em
below subdatum (B.D.).
UA 75 11 9 (Fig. 19, G) is a biface fragment of very dark grey
(Munsell value= 3) hornfels. It was broken transversely, by a perverse
(helical) fratture. The blow which caused breakage was probably delivered
to the right-hand margin. Stepping and hinging is evident, but not pro-
nounced. Only minor flaking is evident on the ventral surface, giving the
specimen a nearly-unifacial appearance. Technology was probably direct
freehand percussion. There is no indication of intentional platform prepara-
tion, and no indications of use wear. Morphologically, this artifact is
asymmetrically bi-convex in cross-section, with an asymmetrical outline.
Length: 8.5 em (broken); width: 6.7 em; thickness: 2.1 em; weight:
121.1 gm; provenience: found in roadcut at level of unconformity (unit 2/3
contact); approximate coordinates: 1-4 N, 3-4 W.
r
83
84
TABLE 4
COMPONENT I ARTIFACTS
Catalog # Description Inferred Materia 1 Figure
Function Reference
UA-76-212-1 biface fragment preform argi 11 i te 19-H
UA-75-11-9 biface fragment preform hornfe 1 s 19-G
UA-76-121-2 cobble manu port anvil stone gabbro 20
UA-77-61+62+63 prismatic knife ( ?) hornfels 19-A
blade-like-flake
UA-77 -64-125 bi face fragment biface argillite 19-F
trilTilling
flake
UA-77 -3+4 biface knife (?) argillite 19-8.
UA-77 -64-2 biface preform hornfels 19-0
UA-77 -64-1 +317 bi face handaxe/ argillite 19-I
chopper ( ?)
UA-76-212-105 bi face fragment preform argillite 19-E
UA-76-212-400 biface fragment preform argillite not
i 11 ustrated
UA-77-64-79 bi face fragment scraper (?) argillite 19-C
UA-76-212-385 retouched scraper (?) argillite not
flake i 11 ustrated
UA-76-212-172 retouched scraper (?) argillite not
flake illustrated
UA-77 -64-5 tabular flake argillite 21
flake core ( ?)
UA-77-64-7 caribou awl ( ?) bone 22-C
vestigial
metacarpal
UA-77 -64-33-A bone projecti 1 e bone 22-A
point point
tip (?)
UA-77-64-33-8 bone projectile bone 22-8
point point
tip (?)
A c
E F
Fig. 19--Component I Artifacts (refer to Table 4).
E u
~
0
UA 76 121 2 (Fig. 20), is a cobble manuport of dark grey (value=
3) gabbro. This specimen was found within the main concentration of Locus
A argillite debitage. It was probably brought into the site for use as an
anvil stone, on which bone could be smashed. There is little indication of
any use wear anywhere on this rock. Length: 10.2 em; width: 9.1 em;
thickness: 4.2 em; weight: 630.7 gm; provenience: 254 M, 1.64 W; 223 B.D.
UA 77 61+62+63 (Fig. 19, A), is a prismatic blade or blade-like
flake of dark olive grey (5YR3/2) hornfels. It exhibits possible use wear
or both edges, in the form of medium sized step and feather fractures. The
distal end is of particular interest, in that it may have served as a
"burin-like implement" (cf. Mauger 1970:46). The manufacturing technique; ..
which produced this facet is impossible to determine, although it was most
likely formed as a "snap 11 fracture rather than by a deliberate burin blow.
The distal end of this implement terminates abruptly, forming a transverse
facet at right angles to both faces. The surface of this facet is heavily
polished, with considerable blunting and polishing of the juncture between
the ventral surface and "burin" facet. The artifact was found broken in
three pieces, all within a 1.5 m diameter; the fractures which separated
the three pieces suggest an extreme 11 bending 11 force was applied while the
implement was held by its proximal end. Most of the apparent wear is on
the ventral side edge of the "burin like 11 facet. The length of this facet
is 3.5 em, with a width of 0.8 em. The specimen has nearly parallel lateral
margins, and is prismatic in cross section. It may have been struck off a
prepared core, although this is impossible to tell from the piece itself,
and there is no debitage in the Component I level which would indicate a
86
maeroeore and blade industry. Length: 10.75 em; width: 4.9 em; thickness:
1.5 em; weight: 87.0 gm; provenience: 1.75 N, 0.45 W, 193 B.D.; 1.95 N,
0.68 W, 187 B.D.; 2.35 N, 0.8 W, 186 B.D .
UA-77-64-125 (Fig. 19, F), is a biface fragment of very dark grey
(value= 3) argillite. It was not utilized, and represents a large biface
trimming flake. The specimen has a fresh-looking ventral surface, and
possesses considerable stack-step fracture on the margin (due to manufacture).
Length: 7.5 em; width: 4.6 em; thickness: 1.55 em; weight: 51.5 gm;
provenience: 8.35 N, 1.97 W, 192 B.D.
UA-77-3+4 (Fig. 19, B), is a biface of very dark grey (va.lue = 3)
argillite. It is bi-convex in outline and nearly lenticular in cross
section (asymmetrical). This specimen exhibits minute step fractures along
one lateral margin, and could have been utilized briefly as a knife. There
is possible edge-damage on the tip, due to either manufacturing or use. It
is only slightly modified on the ventral surface, giving it an almost
unifaeial appearance. It is broken in mid section by a transverse fracture.
Length: 11.0 em; width: 3.85 em; thickness: 1.25 em; weight: 50.5 gm;
provenience: 1.67 N, 0.81 W, 192 B.D.; 1.55 N, 0 .72 W, 192 B.D.
UA-77-64-2 (Fig. 19, D), is a biface/preform of hornfels. Color
varies from black (value= 0) to greyish brown (2.5 YR 5/2). It is ovate
in outline, with a thick asymmetrical cross section. This particular
material type is the finest grained, hardest, and densest material found in
the Component I level of the site. It was found in Feature 1, within
Hearth 1, and may have been heat treated. It is crudely flaked by direct
percussion, and possesses some remnants of the original river-cobble cortex
surface. Length: 8.7 em; width: 5.2 em; thickness: 2.7 em; weight:
122.5 em; provenience: 1.92 N, 0.45 W, 184 B.D.
87
UA-77-64-1 + 317 (Fig. 19, I), is a large biface of very dark grey
(value= 3) argillite. It has a hand-axe like appearance, although that
does not necessarily imply function. It is bi-convex in form, with a thick
diamond shaped cross-section. Large, wide percussion flakes have been
removed from both faces. The distal end has considerable minute stack-step
fracturing; however, no striations can be seen at 70X magnification. This
implement could have been utilized briefly, perhaps in bone processing
activities. Length: 13.15 em; width: 6.8 em; thickness: 4.3 em; weight:
384.8 gm; provenience: 2.15 N, 0.38 W, 184 B.D.; 2.35 N, 0.15 W, 186 B.D.
UA-76-212-105 (Fig. 19, E), is a biface fragment of very dark grey
(value= 3) argillite. It was broken along cleavage planes on all but one
side. It does not show indications of edge-damage, intentional or otherwise.
One small segment of reddened cortex is present. Length: 55.0 em (broken);
width: 37.1 em; thickness: 1.6 em; weight: 39.9 gm; provenience: 1.47 N,
1 • 63 W, 22 B . 0.
UA-76-212-400 (not illustrated), is a small biface fragment of very
dark grey (value= 3) argillite. It was broken via a helical (perverse)
fracture. Length: 6.55 em (broken); width: 2.1 em; thickness: 1.3 em;
weight: 28 gm; provenience: 7.60 N, 2.22 W, 237 B.D.
UA-77-64-79 (Fig. 19, C), is a biface fragment of very dark grey
(value= 3) argillite. It is bi-convex in outline, with a thick triangular
cross section (broken) The specimen exhibits miniature step fractures
along a 2.8 em portion of one margin, beginning . at a point 1.0 em down from
the tip. Although it was broken longitudinally when it was detached from
the parent piece, it does show some edge damage on the dorsal side margin
(right hand side in Fig. 19, C), which would have had to be created after
the fragment had been detached. It may have been utilized briefly as a
88
scraping implement; this is further suggested by minor rounding and polishing
of edges. At a ?OX magnification, no polish or striations are observable.
Length: 9.79 em; width: 2.62 em; thickness: 1 .35; provenience: 8.12 N,
1 . 95 W, 200 B.D.
UA-76-212-385 (not illustrated), is a possibly-retouched flake of
very dark grey (value= 3) argillite . It is roughly ovate in outline with
an asymmetrical cross section. Edge damage occurs along 3.1 em of the
distal end, which may indicate use as an end scraper. At ?OX magnification,
no polish or striations are visible; however, there is a slight rounding of
the point of juncture between adjoining flake scars . Length: 6.7 em;
width: 5.4 em; thickness: 1.1 em; weight: 39.0 gm; provenience: 6.27 N, --
3.10 W, 216 B.D.
UA-76-212-172 (not illustrated), is a possibly retouched flake of
very dark grey (value= 3) argillite. It is roughly ovate in form, nearly
prismatic in cross section, and possess one dorsal ridge. There is some
minor edge damage on a 1.7 em segment of the distal end of this specimen,
suggesting possible function as an end scraper. Microscopic examination
revealed no striations or polish, and only minor rounding of the edge.
~ength: 4.2 em; width 3.1 em; thichness: 0.5 em; weight: 7 .8 gm; proveni-
ence: 2 .05 N, 1.50 W, 226 B.D.
UA-77-64-5 (Fig. 21), is a large, unmodified tabular flake.. It is
composed of very dark grey (value= 3) argillite or hornfels. It is
illustrated primarily to describe the nature of some of the quarried raw
material brought into the site from the probably quarry source located
approximately 1.0 km away from the site. It has not been shaped into any
preform or quarry blank form. It was found within Feature 1, the suspected
89
.·
Fig. 20.--Cobble Manuport, Component I.
0 5 c m
Fig. 21 .--Large Tabular Flake, Component I .
heat-treatment pit, and was directly associated with UA-77-64-2. Length:
15.1 em, 12.4 em, 5.4 em; weight: 586.7 gm; provenience: 2.18 N, 0.35 W,
182 B.D.
Component I: Possible Bone Tools
It is not possible to state with certainty that any of the specimens
listed below actually functioned as organic artifacts. There is
little that is diagnostic about any of these specimens to suggest they were
used by man, except for fonn and possible indications of ·intentional shaping.
However, each of them exhibits one or more morphological char_acteristics
which suggests they c6uld have served as bone artifacts d~ring the Component
I human occupations of the Carlo Creek site.
UA-77-64-7 (Fig. 22, C), is a caribou vestigial (second) meta-
carpal. It is slightly damaged on the distal end, a feature which could be
due to either post-depositidnal damage or utilization as an awl. Length:
7.81; width: 0.93 em, 0.64 em; weight: 2.4 gm; provenience: 1.32 N,
0.65 W, 195 B.D.
UA-77-64-33-A (Fig. 22, A), is a small bone point (species unknown),
probably derived from a long bone. It exhibits a slight polishing on the
pointed end, which may be due to intentional modification. Length: 3.42 em;
width: 0.86 em; thickness: 3.6 em; weight: 0.5 gm; provenience: 1.34 N,
0.37 E, 176 B.D.
UA-77-33-B (Fig. 22, B), is a small bone point, virtually identical
to the one described above. It was found within 1 em of the pointed bone
object described above. Length: 4.2 em; width: 0.71 em; thickness:
0.43 em; weight: 0.6 gm.
UA-76-212-358 (not illustrated) is a small pointed frasment of wood
(species unknown). It exhibits no evidence of modification, and may be
91
0
I I
c
em
I I
·4jfft1'Kii!tt;
I
B
5
I
Fig. 22.--Component 1: Possible Bone Tools (refer to
Table4).
..
intrusive into the Component I level. It was found during the screening of
Component I backdirt, hence its occurrence within the Component I level is
suspect. Length: 3.0 em; width: 0.41 em; thickness: 0.41 em; weight:
0.3 gm; provenience: backdirt from square 3-4 N, 0-1 W.
Component II
As previously discussed, a small lithic component (Component II)
was located in the upper part of Unit 3 (see Figs. 10 and 11). An oval-
shaped distribution of rhyolite waste flakes approximately 1.5 m by 1m in
size constitutes this cultural level. The total assemblage from the upper
component consists of 637 lithic waste flakes. Based on the small size of
this concentration, these appear to have been the result of the activities--
of a single flintknapper for one brief point in time (see Fig. 23).
There is little doubt that Component I and Component II are the
93
result of distinct occupations. Component II occurs stratigraphically
approximately 1 m above the Component I level and it is separated horizontally
from the nearest cluster of Component I materials by about 3m. No organic
materials of any kind were recovered from Component II, nor are there any
indications of any features. Lithic materials from Component II are of
white rhyolite, as opposed to the dark grey argillite/hornfels of Component I.
As illustrated in Fig. 23, these flakes are tightly clustered, and occur in
only on~ small locus within the upper cultural level.
The physical setting at the time of the second Carlo Creek occupation
was probably similar to that of the Component I occupation, except that the
Component II flintknapper was probably situated further back on the flood-
plain (relative to the river) than were the hunters of the Component I
occupation. The Component II assemblage occurs stratigraphically entirely
within Unit 3 laminated silts and fine sands.
r
1
lW
2W
JW
5N 4N JN 2N lN ON
1E
l3N l2N llN JON
SUBOATUM
5N l4N JS
(
2
0/0)
s
9N 8N 7N 6N
4 !iU2
(l. 7) (HO)
l 19
(1.4) t lUi)
----L... ___ JW -----------· -----· --------------· ------------------------
Fig. 23.--Component II: Distribution of Lithic Oebitage. Upper numbers within squares refer
to total count per one meter square; numbers in parentheses refer to total weight in grams per one
meter square. Refer to site topographic map, Fig. 4.
• .
lW
2W
CHAPTER 5
DATING
Radiocarbon Dates Relative to Cultural Occupation
Pursuant to the research of the Carlo Creek site, a total of six
radiocarbon samples were dated. Two apply to geological contexts away from
the site proper. The four dates that pertain to the cultural occupation
will be discussed first. A summary of radiocarbon dates from the Carlo
Creek Site and Carlo Creek region is presented in Table 5; the stratigraphic
relationships of these samples are illustrated in Figure 16.
All four of the culturally-related samples were collected from the
Component I level. All of these are from within 5 em (at or above) the
contact between natural stratigraphic Units 2 and 3. Two of these four
dates, obtained from two different radiocarbon laboratories, match within
one standard deviation, and offer good evidence for an approximate date of
8500 B.P. for the age of the Component I level. Except for WSU 1727, the
four Component I radiocarbon samples are all derived from culturally-
deposited charcoal.
Two of these four samples appear to be anomalous. Sample WSU 1727,
which is approximately 3380 years younger than the inferred 8500 year age
of Component I, was obtained through the combination of eight separate soil
humic acid samples from the Component I level. The sample was run primarily
as a 11 back up 11 and cross-check on the charcoal samples from the same level . .
The young date is not too surprising in light of the 2-3m of floodplain
95
Location
Carlo Creek
Site, Component
1 eve 1 , 1 ower
paleosol at base
of Unit 3.
Roadcu t, l oc.
0. 5 mile south
of Carlo Creek
site
I
Cantwell Ash Bed
type locality, mile
218.3 Parks Highway
TABLE 5
SUMMARY OF RADIOCARBON OATES FROM THE CARLO CREEK AREA
Oa te ( 8. P. ) a Lab. No. Material
8400 ± 200 wsu 1700 Charcoal
8690 ± 330 GX 5132 Charcoal
10,040 ± 435 GX 5131 Charcoal
5120 ± 265 wsu 1727 Soil humic acid
7940 ± 130 WSU 2148 Wood and peat
3780 ± 80 WSU 1747 Wood
Hearth #1.
Hearth #2.
Hearth #1.
Significance
Oates Component I occupation
Oates Component I occupation
Date appears too old--see
discussion in text
Average of many soil samples. Date is
anomalously young--probably contam-
inated by groundwater humic acids.
Sampling locality 1-C (see Fig. 16).
Stratigraphy identical to HEA 031.
Sample dates transition from
Unit 1 channel gravels to Unit 2
braidbar sands. Sample collected
45 em above Unit 1-2 contact.
Sample collected 0.5 em below tephra
layer, at a depth of 73.5 em below
surface (Bowers l979a).
aAll dates expressed in terms of Libby half life (5570 ± 30 years) .
• .
sediments which overlie the level. The possibility is quite good that the
organic (humic acid) fraction of these samples was contaminated by more
recent additions of humic acids transported by groundwater. Numerous
studies have documented these types of problems associated with soil dating
(Campbell et al. 1967; Davies 1971; Goh and Malloy 1972; and Goh et al.
1977: 1 77-1 96) .
The 10,040 year date derived from sample GX 5131 is somewhat more
puzzling to interpret. According to H. Krueger of Geochron Laboratories
(personal communication, 1977) this sample was quite small, amounting to
less than 0.5 gm after combustion. It was necessary to dilute it and was
subsequently counted in Geochron•s smallest counter. The sample size, in --
and of itself, should not have thrown the age off by the apparent 1540 years
indicated. The sample was charcoal, and from the same hearth that yielded
the site•s 11 best 11 date: WSU 1700. One factor which should be considered is
the possible contamination of the sample by detrital fragments of coal or
lignite, both of which are present in small amounts in the Cantwell
Formation (Wahrhaftig 1958). A similar problem was apparently encountered
'Hith the dating of the Dry Creek Site, located 35 km down valley from Carlo
Creek (Thorson and Hamilton 1977; Powers and Hamilton 1978).
As pointed out by both Stuckenrath (Adovasio et al. 1978:153) and
Sheppard (1975, 1977:5) it would require a significant volume of contaminant
in a sample to significantly alter its true age. However, the fact that
this was a small sample, with resultant high counting error, would make the
sample more susceptible to contamination.
97
Another, but less likely explanation in these apparent inconsistencies
may be differences in laboratory analytical techniques. An example of this
problem occurred when a series of three paired samples (see Table 1) from
the Healy Lake site, east-central Alaska, yielded paired dates that differed
by 1575, 255, and 1565 years between determinations made by Geochron
Laboratories and the Smithsonian, with the Geochron dates consistently
younger (Pewe 1975a, Table 4; Cook, personal communication, 1979).
Taken together, however, these four samples indicate an approximate
age of 8500 B.P. for the Component I occupation level. In an earlier report
(Bowers 1978a:7) I attempted a weighed average of all samples, following a
technique developed in Long and Rippeteau (1974:208). The result of this
average was 8487 ± 260 B.P. However, I now feel that it would be most
'prudent to reject altogether samples WSU 1727 and GX 5131, the youngest and
oldest dates. The fact that samples WSU 1700 and GX 5132 fall within 290
years of each other is probably the best single _line of evidence in support
of our inferred dating of the site. The contemporaneity of WSU 1700 and GX
5132 is readily apparent; an average of these two samples yields the
following date:
N= 8545 ± 265 B.P.
sigma range: 8280 to 8810 B.P.
2 sigma range: 8015 to 9075 B.P.
Carlo Creek Region: Geologic Dates
Sample number WSU 2148, dated at 7940 = 130 years:5990 B.C., was
obtained from 45 em above the Unit 1 and Unit 2 contact at a roadcut located
at mile 223.0, Parks Highway (see Fig. 16; Section 1-C). The exposed
section at this point appears identical to the stratigraphy of the Carlo
Creek site in overall morphology elevation, sediment structure, unit
thickness, and apparent sorting and texture. It is quite probable that the
deposition of these sediments was roughly coeval with the deposition of the
98
Carlo Creek site sediments . The date obtained from this sample is from a
large wood and peat specimen (species unknown) embedded within a bar deposit
similar to Unit 2 at HEA 031. This date, when viewed in light of the
inferred depositional regime, and physical differences between Units 1 and
2 at the Carlo Creek site, suggests that the aggradational period which
marked a temporary halt in the river•s post glacial downcutting lasted for
at least 500 years. Although sample WSU 2148 appears to occur stratigraph-
ically at the same level as Component I at Carlo Creek, it is not too
surprising that this sample is younger than Carlo Creek Component I: in
this case, the rock stratigraphic boundary is not necessarily a time-
transgressive boundary. Section 1-C probably represents an identical
99
fluviatile sedimentary cycle as the one exposed at the Carlo Creek site. It
indicates a similar channel sequence at approximately the same elevation,
which occurred a few hundred years later. As the braided river cut numerous
channels, and meandered within its floodplain, there were undoubtedly numerous
composite sedimentary deposits which are similar to the Carlo Creek sequence.
As pointed out by Allen (1970:136)
The floodplain of a well-developed braided stream is
not a continuous region, for it consists of the many alluvial
islands or braid bars which divide up the flow. Moreover,
the elements of the floodplain have little permanency for the
sediment bars experience a continual and rapid modification
by the flow passing round them.
The final date reported in this study is sample WSU 1747, dated at
3780 ± 80 radiocarbon years:1830 B.C., which provides a maximum age for the
Cantwell Ash Bed (Bowers 1979a). Collected from the Cantwell Ash type
locality at mile 218.3 Parks Highway, this sample dates the Cantwell area
tephra to between 3800 and 3600 B.P. (Bowers 1979a). It should be noted
that the Cantwell Ash Bed is present within the upper Unit 4 sediments at
the Carlo Creek site, and thus provides a minimum age for the Component II
occupation. The utility of this tephra unit as a stratigraphic marker
horizon is illustrated in Figure 16.
Carlo Creek Component II
During excavation of the Component II occupation level, no radio-
carbon dateable materials were recovered. In an effort to estimate the age
of this occupation, I have attempted to interpolate the strata's age on the
basis of maximum and minimum limiting dates for the Component I level and
the Cantwell Ash Bed.
..
Age estimation for Component II 'r'las determined by two different
methods. First, I assumed that the materials occur in the strata roughly
half way between the Component I level and the ash level. On the basis of
this assumption, the following age was determined:
8500-3780 = 4720
~ = 2360
2360 + 3780 = 6140
Age = 6140 B.P.
This date represents a midpoint between the early and later dated
strata from the site.
The second method, which is considered to be the more accurate means
of interpolation, has assumed a logarithmic relationship between rate of
Unit 3 deposition and time. Various studies of floodplain sedimentation
(e.g., Allen 1965; Leopold et al. 1964) have ind.icated that deposition rate
decreases as distance from channel increases. The fact that Unit 3 sediments
generally tend to fine upward would lend support to this notion: as the
1 oo
Nenana River shifted laterally and/or downcut, it carried less and less
sediment with it and deposited gradually smaller amounts of overbank deposit.
It is thus felt that this rate of deposition will more closely approach a
logarithmic relationship than it would a linear one. Figure 24 illustrates
this relationship.
From Figure 24 an age of 7250 B.P. is obtained. Estimating conserva-
tively, I have thus inferred that the date of Component II occupation falls
somewhere between 7500 to 6000 B.P. This suggests that the second brief
occupation of the Carlo Creek Site occurred between 1000 and 2500 years
after the initial utilization of the site, or at about 6700 ± 750 years B.P.
r
1 01
....
~
~ ...,
~
~ < U-
~
~
V1
·~
~
~
~
~
Q.
~
~
0
20 CANTWELL ASH (3780 6P .
40
60
80
100
COMPONENT II
120
140
160
180
COMPONENT I (8500 B.P.)
200
8500 7875 7250 6625 6000 4750 3500
5375 4125 2875
YEARS BEFORE PRESENT (625 YR. INTERVALS )
Fig. 24.--Determination of Component II
Inferred Age. Note semi-logarithmic relationship
between rate of sedimentation (depth) and intervals
between bracketing radiocarbon dates (years before
present). Sediment deposition probably was rapid at
first, then tapered off gradually as Nenana River
channel moved away from site. See discussion in text,
Chapters 3 and 4.
CHAPTER 6
LITHIC ANALYSIS
Introduction
Analysis of the Carlo Creek lithic assemblage proceeded along three
lines of inquiry: (1) morphological attributes, (2) technological stages,
and (3) function. Morphology is defined as the physical traits which
enable distinctions to be made among a number of specimens. This includes
such phenomena as length, width, size class, weight, shape, edge angle,
or form of hafting element. (Morphological attributes of tools from the
site have been discussed previously in Chapter 4). Technology is the
technique employed in the manufacture of lithic implements, including
techniques of obtaining raw materials, pretreatment of raw materials prior
to flintknapping, sequential stages of tool manufacture, selection and
modification of fabricators, and reworking of tools. Function refers to
the use, intended or accidental, to which a stone tool is put after its
manufacture and prior to its disposal into the archaeological "record"
(cf. Schiffer 1976):
This research was directed by two fundamental questions, which
invariably face any archeologist involved with lithic studies (Gummerman
1975:7);
1. What types of information are potentially available in lithic
debris and tools? -
2. How can this information be retrieved from archeological
specimens?
103
The purpose of this analysis is primarily to place the lithic
remains from the Carlo Creek Site within an anthropological framework. As
products of past human activity, it is important to understand how these
materials came to be part of the archeological record, what relationship
they have to the past human inhabitants ·Of the site, and, finally, what
significance they hold in an attempt to 11 reconstructing 11 or interpreting a
record of human activity. I do not hold the opinion that the study of
lithic technology is an end in itself; it is, however, an extremely valuable
analytical tool which can be applied to archeological interpretation. It is
especially significant in understanding the early prehistory in central
Alaska, where it is evident that a major proportion of the preserved items·
of material culture are lithics.
In reviewing available literature on the methodology of lithic
analysis, it is readily apparent that a wide variety of techniques have been
employed. The procedure utilized here has attempted to extract the maximum
amount of interpretive data from this particular lithic assemblage. It has
been influenced by a number of diverse techniques and ideas from a number of
sources, including: Brink (1978); Brose (1975); Crabtree (1964, 1966, 1972,
1976); Flenniken (1975, 1977, 1978); Frison (1968); Hassan (1976); Mauger
(1970); Mute (1971);'Semonov (1964); Sheets (1973); Solberger and Hester
(1972); Tringham et al. (1974); Weymouth and Mandeville (1974); Wilmsen
(1970); and Womack (1977).
A basic premise underlying this approach is that a lithic assemblage
represents part of a larger system of resource procurement, alterations,
useage, and disposal (cf. Schiffer 1976). A lithic 11 System 11
, as defined by
Flenniken (n.d.:4) is
1 04
the entire life of a stone tool from its inception to its deposition
in archeological context. In other words, a lithic system involves
the selection of raw material, reduction of the raw material into
preconceived tools, hafting of the selected stone tools, and func-
tions of these stone tools--every aspect of a stone tool while it
is in systemic context or until it is rejected by its original
owner.
The theoretical approach of this analysis was somewhat more deduc-
tive in nature than were the other analyses of the Carlo Creek data. During
and after excavation of the site, but prior to the lithic analysis, several
hypotheses were developed concerning the systemic context of the lithic
remains:
Hypothesis 1. The Carlo Creek Component I occupants used locally
available raw materials.
Hy~othesis 2~ Component I materials were heat treated prior to
re uction.
Hypothesis 3. Technology employed at the site was biface reduction,
using direct freehand percussion.
The flow charts in Figures 25 and 26 show the lithic systems thought
to be operative at HEA-031. These diagrams will form the basis for later
discussion of the Carlo Creek lithic systems.
Methodology
Study of the Carlo Creek lithics followed a reduction stage concept,
as developed variously in Flenniken (1975, 1977, 1978), Holmes (1919), Muto
(1971), Sharrock (1966), and Womack (1977). As a first step in sequence
determination, debitage was examined, wi.th data recorded for the following
categories: (1) provenience, (2) size class, (3) stage of manufacture,
(4) platform, (5) cortex, (6) lipping, (7) termination, (8) use, (9) mate-
rial, (10) weight, and (11) blade-like-flakes. Each of these categories
will be briefly discussed below, and are summarized in Table 6.
105
SELECTION OF RAW MATERI AL
AT QUARRY SOURCE (1)
DEBITAGE AT REDUCTION AT QUARRY PRIOR
QUARRY SITE ~ TO TRANSPORT TO SITE
I
TRANSPORT TO SITE I
HEAT TREATMENT AT SITE NO HEAT TREAT-
( 2) MENT AT SITE
J 1CORTEX KEMOVAL (3) I
-............
JBIFACE THINNING l 4) 1---USEABLE FL AKES
( 5 )
WASTE FLAKES ~USEABLE TOOLS l S) I /
I I·· AND /
BROKEN TOOLS USED NOT
(7) ( 6) USED
(7) --r j
I REMOVAL FROM sITE I
Fig. 25.--Component I Lithic Reduction System. Numbers
in parentheses refer to technological stage represented at the
Carlo Creek Site. Stage 7, disposal, occurs at most points in
the sequence. Refer to discussion in text.
106
SELECTION OF RAW MATERIAL
AT QUARRY SOURCE (1)
DEBITAGE AT REDUCTION AT QUARRY PRIOR
QUARRY SITE 1-TO TRANSPORT TO SITE
TRANSPORT TO SITE -I
I l
HEAT TREATMENT AT SITE NO HEAT TREAT-
(?) MENT AT SITE
( 2)
~CORTEX REMOVAL (3: l
BIFACE THINNING (4 ) :
WASTE FLAKES USEABLE TOOL (?)
(7)
REMOVAL FROM SITE (?)
Fig. 26.--Component II Lithic Reduction System. Numbers
in parentheses refer to technological stage represented at the
Carlo Creek Site. Stage 7, disposal, occurs at most points in
the sequence. Refer to discussion in text.
107
Category
1 (Largest )
2
3
4
( Smi 11 est)
Primary decortication
Secondary oe<:ortication
rninninq
Biface thinninq
fo\11 tipl e remova 1
Shatter
Chipoed olatform
Abradea olatform
I noetermi nate
QuaM"y
Incipient cone
No cortex
a1 ade-11 ke
Lioped
Hinge
Steo
=eu~er
I noetenni na ~~
Outrecasse
Used
Maybe
~n-used
~nyo1 ite
Hornfels
Chttl"t
Argillite
r:.sL:: 6
CARLO CREEK OEB!rAGE ANAL~SiS
Com oonen:
:..ocus ~d
No.
28
56
100
186
260
698
Size :lass·
2.1
4.2
i.S
14.0
19.5
52.5
LO CYS 3~
110.
12
30
90
108
96
148
2.4
6.2
18.6
22.3
19.5
30.6
Stage of Manufacture
10
74
34
206
68
946
226
0
1102
150
42
1136
0.8
5.5
2.5
15 .s
5 .1
71.2
Plat form
17.0
0
83 .0
Cortex
11 .3
3 .1
85.5
10
24
18
74
10
348
116
0
368
28
2
454
2.1
5.0
3.7
15.4
2.1
71 . 9
23.9
0
76.0
5 .8
0 .4
93.8
Misce11aneaus Attributes
21
138
1.6
10.4
Terminat i on
180
78
1114
65c5
0
13 .5 :.a
31.2
49.4
0
Util izat 1o n
2
3
1323
0.2
0.2
99.6
'Ia teri a 1 rype
0
18
1
1308
0
1. a.
98.
so
c58
70
12 8
218
0
0
1
483
0.8
10 .3
14.0
14.4 ze.4
45.0 c
0
0.2
99.8
0 0
0 0
0 0
484 100
Comoonen: ::c
No.
0
6
12
24
46
54
2
16
8
36
10
72
12
48
84
2
0
142
4
16
47
4
46
50
0
0
0
1<14
1<14
0
0
0
0
~.2
8.3
16.5
32.0
37.5
1.4
11 .1
5. 5
25.5
7.0 so.o
3.3
33.3
58.3
1 .4
0
98.6
2.7
11.1
32.6
2.i
31 . g
J4.i
0
0
0
100
100
·J
c
aThel"e we~ 5460 flakes (weight: 3236.2 gm ) i n Lo cus A; 1328 cf tne
flakes (24 .3 ~) were anal y z~.
llrhere werl! 2262 fialc es (~ignt: 2453.54 gm) in Lo cus 3; 484 of :ne
flakes (21.4 :) wei"! analyzed .
'Thtl"e wer! 637 flues (weignt: U2.i gm) in Comoonent 2; 1<14 of :he
flakes (22.6 ~) were analyzeo.
0Refer to Fig. 28.
1 08
..
An attempt at acheiving a random sample for this analysis was made
by selecting flakes in 11 grab 11 fashion from level bags. (Each flake or flake
cluster is in a separate coin envelope within a level bag; selecting an
envelope instead of an actual flake reduces the potential bias due to touching
or 11 feel 11 of a certain size or shape of flake.) Flakes were selected _ in
this manner until a 20 to 25% sample had been analyzed from each 1 x 1 meter
square. The total number of flakes thus analyzed was 1956, or 23.4 % of the
total collection. For purposes of analysis, proveniences were grouped into
three spatially-discrete units within the site: (1) Component I, locus A,
(2) Component I, locus B, and(3) Component II.
During excavation of HEA 031, flake provenience was recorded by
square and by intra-square coordinates, for each of the 8359 flakes.
109
Although not done in this study, it should be possible for a future researcher
to use these data for distributional analysis, according to flake 11 type 11
•
This could theoretically enable a nearly complete reconstruction of the
lithic reduction activities at the site.
Each of the 1956 flakes analyzed was categorized according to its
size class rather than by traditional length-width-thickness measurements.
The problems in standard measurements of flake morphology have been previously
pointed out by Brauner (1968:12), and are illustrated in Figure 27.
I
For the purposes of presenting flake-size date, it was felt that a
graphic representation of size is more useful than standard numerical
length, width, and thickness measurements. I believe that it is too hard
for most people to visualize sizes of flakes if only metric attributes or
weight are presented. In this study, 60° ellipses were used as templates on
which flakes were 11 Sized 11 (see Fig. 28). A 60° ellipse is regarded as a
c:::::; I Th c ' I Th
L
w ~~~
W ~ Th
Fig. 27.--The Fallacy Generated by Using Length x Width x
Thickness = Volume. ADAPTED FROM: Brauner (1968:12).
Flakes
Larger
Than
Size 2
L
.·
6
Flakes
Sma 11 er
Than
Size 5
Fig. 28.--Templates Used in Determining Flake Size Categories. Refer
1
.
to debitage analysis data in Table 6 and discussion in text.
i
good approximation for the average morphology of a flake, even though it
does not completely avoid the pitfalls discussed by Brauner (1968:12).
Stage of manufacture (unless otherwise stated, lithic technology-
related definitions follow Crabtree [1972]) was subdivided into classes of:
(1) primary decortication (White 1963:5), (2) secondary decortication
(White 1963:5), (3) thinning (Womack 1977:65), (4) biface thinning (Womack
1977:65; Flenniken 1977:76), (5) multiple removal (Womack 1977:70), and
(6) shatter (Binford and Quimby 1972:347).
Determination of platform preparation was made through visual
inspection with a 4x hand lens, and was restricted to categories of chipped
vs. chipped and abraded. Abraded platforms are those which show subs .tantial
bevelling of the striking platform through grinding or polishing. No atte~t
was made to record data for platform angles.
Flakes from both cultural components were divided into three cate-
gories with respect to cortex: (1) quarry, (2) incipient cone, and (3) no
cortex. Quarry cortex, as used here, is a weathered surface on tabular
stone, that is broken along natural cleavage planes. The color of Component I
argillite quarry cortex, red or reddish brown, formed a sharp contrast with
the grey unweathered rock. Component II cortex was a reddish hue, which
was distinct from the light colored unweathered rhyolite. Incipient cone
(cobble) cortex is that which occurs as a result of rounding and battering
in a riverine environment. Cobble cortex was found only in the Component I
assemblage.
Presence or absence of a pronounced 11 lip" on the vertical surface
of a flake, immediately below the striking platform, was considered at
the time of this analysis to be another potentially useful observation.
111
It was suggested by Crabtree (1972) that such criteria may indicate the
type of fabricator employed; i.e., hard hammer, soft hammer, or antler
baton. However, Flenniken (personal communication, 1979) no longer regards
this relationship as valid, and does not believe that one can determine
fabricator on the basis of lipping.
Nature of terminations (i.e., hinge, step, feather, outrepasse
[Tixier 1974])were other criteria utilized in this analysis. Relative
proportions of these termination types may be useful as indicators of the
.-relative ease or difficulty with which a raw material was worked.
Flakes judged to have been utilized were set aside during analysis
and later restudied along with edge-damage analysis of major artifacts.··
These were first given cursory examination under a 4x hand lens, then
examined under a binocular microscope at various magnifications ranging
from 20x to 70x.
Material types from HEA 031 are restricted to argillite/hornfels
(Component I), rhyolite (only in Component II), and chert (1 flake only).
Lithic raw materials utilized in the lower level of occupation were all
argillite/hornfels, with two exceptions: (1) a single black chert flake,
and (2) a single large prismatic blade composed of a different color-
grade-texture of stone than any of the other materials. This has not been
unequivocally identified, although it probably represents a different grade
and composition of contact-metamorphosed argillite.
Although not detailed in Table 6, weight was recorded separately
for each flake in the 23.4% sample, down to 0.1 gm. The majority of the
smallest-sized flakes (size class 6) were 0.1 gm or less in weight. Total
flake weights per meter square are presented in Figs. 17 and 23.
112
Finally, observations were made on flakes which appeared, morpho-
logically to be blade-like. Following Tixier (1974:7), a specimen that
is called a blade must be at least 1.2 em wide and 5 em long, and its
width must be at least twice its width. Microblades are less than 1.2 em
wide, and at least twice as long (See Table 6).
All of the above data were recorded on IBM Fortran coding forms.
The technique employed here was judged to be fairly time-efficient, with
a maximum amount of data recorded in a fairly short time. Analytical
units were grouped within the site according to each of the three clusters
of flaking debris: Component I, 1 ocus A; Component I, 1 ocus B; and
Component II. Finally, total counts and weights were recorded for 100% ,
of the sampling universe. These data have been presented in Figures 17
and 23.
After completing the analysis of debitage, all biface fragments,
blade-like flakes, manuports, and possibly edge-modified flakes were ex-
amined, with observations recorded for morphology, technology, and inferred
function. The results of these observations and measurements have been
presented in Chapter 4.
In addition to all of the above, two additional analytical tech-
niques were employed in the analysis of the Carlo Creek lithic collection.
These are both considered 11 replicative experimentation 11
, defined by
Flenniken (1978:3; and further modified from Crabtree 1976:106). I
would caution, however, that any given replication sequence only suggests
one possible means of achieving a desired goal, and does not necessarily
exhaust the range of possible techniques for producing a certain stone
tool.
113
... Replication, in its strictest sense, is reproducing stone
tools, using the aboriginal artifacts as controls, by aboriginal
stone working fabricators, ~playing the same raw materials, and
following ... similar reduction technology. Therefore, the end
products as well as the debitage, sequential stages of manufacture
and rejuvenated tools are the same as the aboriginal controls in
terms of technical category percentages, morphologies, and
technologies (Flenniken 1978:3).
As employed in the Carlo Creek analysis, this methodology was
applied in less than its "strictest sense" and was largely subjective in
approach. I did not attempt rigorously-controlled experiments which could
control for the myriad of mechanical or physical variables in replicative
studies (e.g~ Bonnichsen 1977; Speth 1972; Tringham et al. 1974). The
major purpose of this part of the Carlo Creek lithic analysis was to (1).
produce replicas of the lithic artifacts, in an effort to gain personal
insights into the technology employed aboriginally, (2) utilize these
replicated tools in a variety of bone processing functions, to provide
edge-wear controls which could be directly compared to the aboriginal
specimens, and (3) determine the extent, if any, of intentional thermal
pretreatment of the Carlo Creek Component I sample. The major goal was
to replicate a specific artifact~. rather than a specific attribute.
Altogether, ten lithic replications were produced by myself and
J. Jeffrey Flenniken. Raw materials used included: (1) unmodified Carlo
Creek argillite/hornfels (non-aboriginal), (2) basalt obtained from the
Stockhoff quarry, LaGrande, Oregon (Womack 1977), and (3) experimentally
heat-treated Carlo Creek argillite/hornfels (non-aboriginal). Stockhoff
basalt was used because it was available in large quantities and because
it is very similar to Carlo Creek argillite/hornfels in terms of texture,
lustre, hardness and "flakeabiliti' (flakeability, as used here, follows
the definition in Flenniken and Garrison 1975).
114
115
The controlled heat-treatment experiment was designed to determine
whether or not lithic raw materials from the site were heat-treated aborigin-
ally. Ever since Crabtree•s (1964) observations concerning the possibili t y
of intentional thermal alteration of lithics by prehistoric flintknappers,
a growing body of data has been collected which suggests that such procedure
is archaeologically widespread (Crabtree 1964; Flenniken and Garrison 1975;
r~ndeville 1971; Mandeville and Flenniken 1974; Purdy 1971; Shippie 1963;
Solberger and Hester 1972; \~eymouth and Mandeville 1974). It has been
demonstrated rather convincingly that numerous changes occur during heat
alteration, including fracture strength, lustre, color, tensile strength,
and flakeability (Flenniken and Garrison 1975) of raw material.
To test this possibility in the Carlo Creek lithic sample, a series
of rhyolite and argillite specimens from the site and from nearby outcrops
were heated in a temperature-controlled furnace at 100°C increments, ranging
from 100° to 600°C. Each sample was heated in a sand bath for 10 hours,
allowed to gradually cool, then removed. The samples were then examined
for color and texture change, after which they were flaked to notice any
changes in workability. On the basis of these experiments, it is evident
that the argillite/hornfels from Component I was heat-treated. These
observations will be discussed in greater detail later in this chapter
under the heading of 11 Stage 2: Heat Treatment••.
As a final step in the analysis of lithics from the Carlo Creek
site, microscopic edge wear studies were conducted on all "utilized flakes"
or tools. A variety of magnifications were used, ranging from a 4x hand
lens, up to ?Ox with a binocular microscope. Edge damage criteria previously
established by Semenov (1964), Sheets (1978), and Tringham et al. (1974)
were used to identify possible edge damage. As a control, j our flakes
116
from the aboriginal collection and three replicated specimens of argillite
and Stockhoff Basalt were subjected to bone cutting, smashing, and scraping
tasks to experimentally determine edge damage criteria for these raw
materials. These were then compared side-by-side with suspected aboriginally
-utilized specimens to attempt to determine function. The replications were
conducted on fresh articulated cow bones obtained from a local supermarket.
Two flakes were used in a scraping motion for 20 and 50 strokes, and two
flakes were used in a cutting motion for 20 and 50 strokes. One biface was
used as a chopping implement (20 strokes), one was used as a knife (20
strokes) and one was used as a scraper (20 strokes). The major purpose
of this experiment was not to rigorously control all variables in stone
manipulation, as in Tringham et al. (1974), but rather to ,gain personal
insights as to the functions which may have been operative at the Carlo
Creek Site. The results of these observations are presented in the
section entitled "Stage 6: Tool Use".
Before turning to : the interpretation of the Carlo Creek lithic
reduction sequence, it will be useful to examine briefly the results of
the lithic debitage analysis presented in Table 6. The major purpose
of analyzing the 23.4% sample was to attempt to "fingerprint" the
debitage assemblage, to enable the making of quantifiable comparisons,
both among intra-site activity areas, and with other Alaskan site
assemblages.
The two major loci in the Component I level are remarkedly similar
in overall characterization of debitage attributes. Both show similar
distributions of flake sizes, manufacturing "stages'', percent of platform
preparation, percent of "lipped" flakes, distribution of termination
attributes, and relative percentage of blade-like flakes (See Table 6).
The only noticeable difference was in the relative amounts of quarry cortex
and incipient cone cortex: 11.3% quarry cortex and 3.1 % cone cortex in
locus A as compared to 5.8% quarry cortex and 0.4 cone cortex in l ocus B.
There also was a lower percentage of step terminations in locus B than in
A. This suggests that the sequence of activities in locus A involved the
reduction of material having more quarry cortex on it than did the material
in locus B. Evidence from both loci indicates the same technique of re-
duction was used after cortex had been removed.
By way of contrast, the 11 fingerprint 11 of the Component II debitage
(See Table 6), is strikingly different from Component I. In Component II,
there are higher percentages of small (Size 5 and 6) flakes, more evidence
of bifacial thinning, more abraded platforms, and more hinge fractures.
117
This probably reflects differences in material types, (argi11 ite vs. rhyolite),
and in flaking techniques (percussion in Component I vs. pressure in Com-
ponent II).
The Carlo Creek Lithic Reduction System: Component I
On the basis of the procedures outlined above, the following
system of procurement, pretreatment, reduction, and usage appears to have
been operative at HEA 031 during its occupation of circa 8500 B.P. As
an aid to the organization of the discussion, I will follow the flow
ch~rt in Figure 25, discussing each stage in light of field observations
and laboratory analyses. Discussion of the Component II lithic system
will be reserved for the last part of this chapter (See Fig. 26).
Stage 1: Raw Material Selection
The major raw material present in Component I at HEA 031 has
been characterized, on the basis of X-ray diffraction anal j sis of one
sample, as an argillite/hornfels (Fait, personal communication, 1978). It
is composed of fine-grained opaque mineral bands, with relict bedding,
and contains 20-30% of anhedral quartz grains, 65-75% mica and other layer
silicates, and about 5% opaque oxides. According to Or. F.F. Fait of the
vJ.S.U. Geology Department, this rock is 11 definitely metamorphic". It
could be 11 either an argillite or hornfels", but because of its density,
may be more like the latter, It very closely resembles basalt in texture,
lustre, and color (Fait, personal communication, 1978).
Didier (1975) has discussed this problem previously:
The argillite problem is extremely complex, and difficult, both
archaeologically and geologically. Argillite is an aphanitic mudstone,
a sedimentary deposit that was deeply buried, compacted, and harderred
at relatively low temperatures and low grade metamorphism. It occurs
in ranges of colors and textures, and is really not a single unique
composition but a continuous gradation of related, yet distinguishable
materials ... Fresh hornfels (altered argillite) is black, with a
crystalline sheen, dense, strong, and flakes with a good conchoidal
fracture. Hornfels is a high temperature recrystallized rock; normal
argillite is a non-high temperature and pressure 11 metamorphosed 11 rock.
The material utilized in Component I activities appears to have
been derived from two sources: (1) a minor amount, represented by 3.16%
of the analyzed debitage, was derived from [a] stream-rolled cobble[s].
This is posited on the basis of incipient cone cortex (Flenniken, personal
communication, 1978), and smooth polished surface. The total amount of
this material in the site is estimated at about 112 gm, an amount which
could easily represent a single fluvial cobble. (2) Most of the stone
in this level was probably derived from a large and quite prominent
argillite outcrop located about 0.8 km east of the site. Several large
chunks of tabular argillite (e.g., Fig. 21) found in the vicinity of
Hearth appear to be similar to material from this source, in cortex
texture, size of naturally-occurring spalls, and morphology.
11 (
The suspected quarry source was investigated during both the 1976 and 1977
field seasons. However, no definite indications of aboriginal utilization
were discovered. The base of the outcrop/cliff was heavily vegetated,
with extensive recent frost-cracked rock fragments and large boulders
covering the surface (see location in Fig. 9).
The fact that a fairly small proportion of flakes from both Locus A
(0.75%) and Locus B (2.06%) were primary decortication flakes suggests that
part of the decortication processing was performed at or near the quarry
source. The raw materials were probably brought into the site as partially
decertified tabular spalls. Based on the total known weight and distribution
of materials--in Component I, these probably weighed about 500-1000 gm per
piece.
Stage 2: Heat Treatment
On the basis of available data, a good argument can be made for
intentional thermal alteration of lithics at Carlo Creek. Solberger and
Hester (1972:18) have stated that:
A major advantage in thermal treatment is that it allows an improved
conductivity of force in flake removal, i. e., a cleavage which
terminates with a thin, feathered edge. This eliminates many of the
snapped-off flakes and hinge fractures that frequently ruin bifaces
during manufacture. This improved conductivity often allows the
developing split or flake cleavage to pass through flaws or inclusions
within the mass being flaked. Thermal treatment also permits the
removal of much larger flakes by pressure methods than is possible
on the same material which is not heat-treated. ·
Although fully conclusive evidence of thermal treatment at HEA 031
must await the results of X-ray diffraction analysis (cf. Weymouth and
Mandeville 1974), the following observations can be marshalled in support
of the claim that this technique was used at Carlo Creek:
1. The cortex on the Carlo Creek debitage is identical in hue .
(reddish-brown) to that from experimentally heated non-aboriginal samples
119
from the argillite outcrop located near the site. It was found that the
closest color match occurred between the aboriginal control and the non-
aboriginal samples heated to between 400°C and 500°C. In contrast, unheated
non-aboriginal samples show a yellowish-brown cortex.
2. According to the opinion of an expert flintknapper, the unheated
raw material from the outcrop is virtually um>~orkable (Flenniken, personal
communication, 1978). Hhen struck by soft hammer, using direct, freehand
percussion, the margin crumbles and a series of stack-step fractures result.
It is nearly impossible to achieve a controlled conchoidal fracture with the
unaltered stone. However, \'then the same stone has been heated in a furnace
to at least 400°C, then gradually cooled, it can be flaked quite easily,
with greatly reduced stepping and hinging. The aboriginal collection
indicates a step and hinge rate of between about 5 and 15%. Flenniken
(personal communication, 1978) has suggested that the Carlo Creek Component I
material would have had to have been heat treated in order to be knapped
successfully into the bifaces found in Component I.
3. An aboriginally-excavated pit (Feature l; Fig. 11), located in
the northeast end of Hearth 1, may represent an aboriginal oven which could
have been used to alter rock. Sand was evidently excavated from the under-
lying stratum (Unit 2), then piled on top of several quarry blanks and a
large tabular chunk of argillite. The contact between the undisturbed
sediments and those lying above the blanks exhibited a redder hue than did
other sediments in the Component I level. This pit-like feature showed up
clearly on the side-wall profile (Fig. 11). Other similar heat-treatment
units have been observed archeological1y (Shipee 1963; Solberger and Hester
1972).
l 20
In addition to the excavated pit and possible "blanks 11 for thennal
treatment, the underlying sediments were significantly redder in hue than
wereothersediments in the vicinity of the hearth. An excellent comparative
replicative study by Mandeville and Flenniken (1974:146-148) describes a
heat-treatment pit which was utilized in January with ambient air temperatures
of -27°C (-16°F). In that experiment, it was necessary to dig a pit of only
60 em depth. It was found that a temperature of greaterthan 200°C could be
maintained for up to 14 hours during a single firing, resulting in altered
stone which was significantly easier to flintknap (Mandeville and Flenniken
1974). In addition, Harner (1956:40) reports that " ... an ordinary wood-
burning campfire is easily capable of heating stones (flint) ... " to a
temperature of 1000° fahrenheit (537.8° celcius) if they are on the ground
beneath the fire."
It cannot yet be stated unequivocally that the CarJo Creek Component I
materials were thermally treated . However, the available evidence from the
lithic analysis and an archeological feature strongly supports such an
argument.
Stage 3: Cortex Removal
As previously stated, the majority of cortex removal probably took
place at or near the original quarry site. This i s based on the observation
that secondary cortex is more prevalent than primary cortex in both Locus A
and Locus B. (In this study, data for shatter with cortex has been recorded
as percentage of cortex on all analyzed flakes.) A total of 14.4 % of flakes
from Locus A and only 6.2% from Locus B contained cortex, which is regarded
as a low percentage. In contrast, Flenniken (1977:75) reports a figure of
52.0 % of decortication flakes from a riverine/quarry setting at the Miller
Site in Central ~~ashington, while Womack (1977 :59) reports a t otal of 14 %
121
decortication flakes from the Stockhoff Basalt quarry in eastern Oregon.
Such correlations are difficult to quantify, however, because of the unknown
variables of original cobble size and percentage of cortex on the original
rock.
Stage 4: Biface Thinning
122
After removal of most of the cortex, the Carlo Creek artisans ap-
parently worked tabular pieces of argillite/hornfels down bifacialiy, utiliz-
ing direct freehand percussion. Biface thinning flakes exhibit flake scars
on their dorsal surface; only the original detachment flake scar is evident
on the ventral surface. As used here, thi ·s broadly-defined term can encom-
pass several different biface-reduction techniques. Approximately 15% ·of the
analyzed sample from Carlo Creek level 1 represent biface thinning flakes.
Platform preparation was evidently accomplished through chipping,
as there is little apparent evidence for heavy grinding on prepared-platform
flakes. Included in this stage of manufacture are multip1e removal flakes,
defined by Womack (1977:70) as flakes which " ... exhibit a positive bulb
of force on the ventral surface on the flake, and a negative bulb of force
superimposed on the dorsal surface of the flake."
Throughout this stage, as well as those previous, a large amount
of shatter resulted: 71.2% and 71.9% for the two Component 1 loci. Shatter
has been defined (Binford and Quimby 1972:346) as "cubical and irregularly
shaped chunks that frequently lack any well defined bulbs of percussion or
systematic alignment of cleavage scars on the various faces." The high fre-
quency of shatter at HEA 031 reflects: (1) the coarseness of the raw mater-
ial, and (2) the relatively early stages of reduction via direct freehand
percussion.
Stage 5: Finished Tools
Finished tools from the Carlo Creek Component 1 level are represen-
ted by only four complete or broken specimens. An additional six biface
fragments were recovered in addition to these specimens, all of which were
apparently broken in manufacture. The finished tools, along with larger
biface fragments, have been described individually in Chapter 4; finished
tools represent 9.02% of the total lithic assemblage from Component I.
One specimen, a prismatic blade (UA-77-64-61+62+63), is of a no-
ticeably different material type than other materials from Component I; it
is also apparently the product of a different technological sequence than
that described above. However, this material is not represented in larg~
enough quantities in the debitage (1.3% from locus A, 0% from locus B) to
suggest that it was reduced at the site. This specimen apparently was
brought into the site as a finished tool, was used on site, and was subse-
quently broken during usage or resharpening activities.
123
In addition to the bifaces and one 11 prismatic blade 11
, two moderately
sized flakes were judged to have been utilized. These are discussed in the
following section under function.
It should be noted that, in the process of flake analysis, a total
of 25 (1.5%) 11 blade-1ike 11 or "linear" flakes were culled out of the debitage.
In the early stages of analysis, I was uncertain as to whether or not there
was a definite core and blade technology present at the site. In view of
the artifactual data, coarse-grained raw material, and technological se-
quence as revealed through debitage analysis, I am of the firm opinion these
11 bladesll are fortuitous, and are not the result of an intentional core and
blade manufacturing technique.
1 24
Stage 6: Tool Use (Function)
As discussed in the lithics methodology section, determination of
function was established through direct comparison with replicated "controls".
I do not claim that these observations offer conclusive "proof 11 that a parti-
cular specimen was utilized for a given function; I only suggest possible
uses to which it may have been put.
Semenov (1964:14), in his excellent pioneering study of use wear,
determined three basic kinds of observed wear patterns: (1) polishing (fine
abrasion), (2) coarse abrasion (grinding and striations), and (3) rasping
(minute chipping of the edge). The utility of these three distinctions have
been supported variously by Tringham et al. (1974) and Keller (1966). lhese
three basic criteria were applied to microscopic observations of the Carlo
Creek specimens.
On the basis of microscopic analysis of the edge-wear of replicated
specimens, it was not possible to identify striations on the margin or face
of any piece. This is believed to be a function of the extremely durable
edges of the hard argillite/hornfels, and control basalt samples. Lack of
visible striations on worked pieces have been observed previously in use-
wear studies, as in, for example, Tringham et al. (1974:175):
It was found that, even with the addition of earth and other
abrasives, striations form very slowly, sometimes not at all, and
that their use as a criterion of variation in wear patterns is pos-
sible only with the aid of high magnifications (frequently higher
than those possible with a binocular 100~ microscope).
Discernable modifications occurred on experimentally modified
aboriginal flakes only after considerable scraping or cutting directed to
a fresh bone surface. It was experimentally observed that after 20 strokes,
a number of minute 11 rasping" flakes were removed. Only at 50 strokes of
these flakes was it possible to see even incipient polishing. Two of the
experimental bifaces, when subjected to these actions, were affected in a
similar way, except that observable modifications were even less pronounced.
Striations, even after 50 strokes on fresh bone, were not visible. On the
biface subjected to chopping actions, considerable blunting of the margin
was noted, with a high degree of stack-step fractures observed after 50
strokes.
125
In order to compare these observations to the Carlo Creek Site data,
it was necessary to consider three factors, which further complic~te inter-
pretation. The first factor is in attempting to distinguish edge abrasion
during manufacture from use-wear (Sheets 1973). During normal biface reduc-
tion, it is necessary to 11 Strengthen 11 the edge by decreasing the edge-angle;
this can be achieved by abrading the edge by either grinding or fine chip-
ping (cf. Crabtree 1972; Mute 1971). These minute flake scars have, in the
past, been misinterpreted as indications of use-wear (cf. Nance 1971).
Sheets (1973:218) has suggested that the best means of differentiating be-
tween manufacturing abrasion and use abrasion is through microscopic exami-
nation:
A clean, crisp juncture between an abraded area and a flake
scar indicates an edge that was strengthened and then flaked, but
not used. Use tends to nick the sharp boundaries and blur the
cr.isp, fresh edges. In other words, abrasion occurring during
manufacture is cross-cut by flaking, whereas abrasion from use
crosscut or blurs the edges of the flake scars.
He further emphasizes the importance of debitage analysis in dis-
tinguishing use-wear from manufacturing abrasion (1973:218): "If the plat-
forms of biface trimming flakes exhibit the same abrasions as found on por-
tions of the bifacially flaked tool, then abrasion as a manufacturing pro-
cedure is indicated.''
A second major limiting factor in the production of use-wear damage
on a butchering tool is in the accumulation of animal fat on the edge, as
suggested by Brose (1975). His analyses suggest that" ... a major factor
inhibiting the creation of striations during butchering is the accumulation
of fat along the working edge" (Brose 1975:93). In the case of a durable
material such as Carlo Creek argil~ite/hornfels, this might have been an
especially significant factor, which would substantially reduce the apparent
edge damage caused by usage. Brose (1975:93) further observes that signifi-
cant patterned wear may not be noticed on flakes unless they are used for
longer than about 3-4 minutes, or up to several hundred strokes. Since he
was using flints and cherts, which may have been of a lower hardness value
than the carlo Creek material, it is quite possible that argillite/hornfels
could have been utilized briefly for butchering activities with virtuallY
no identifiable butchering damage observable.
In the case of the experimentally-replicated bifaces used in this
analysis, it was not possible to distinguish manufacturing abrasion on one
edge from use-wear abrasion on the opposite edge of the same specimen. At
a magnification of 70x, there were no use-induced striations or similar
edge-damage. Similar results were obtained in replicative studies by
Faulkner (1977). It ~'las only after a tool had been subjected to heavy
usage on bare (non-fleshy) bone that it was possible to detect changes .
Thus, material used aboriginally at HEA 031 for only a few strokes may go
completely unnoticed.
A third factor which must be considered here is the effect of post-
depositional alterations such as trampling by the site's occupants, mis-
handling during shipment (this factor may be ruled out in this case, due to
individual wrapping of specimens during shipping and in the lab) or trowel/
screen retouch. Several investigators (e.g., Tringham et al. 1964) have
pointed out various types of "retouch" which can occur to flakes either dur-
ing detachment or during occupations of a refuse floor.
126
Taking into consideration all of the above, several tentative
inferences can be made about use of lithics during the Component I occupation
of HEA 031. I do not make any certain claims as to utilization; I only
suggest possible uses for selected specimens (refer to Table 4 and Fig. 19),
based on limited experimental and observed data:
1. Two specimens, UA-77-64-3+4, and UA-77-64-79, may have been
utilized as cutting tools. They both exhibit a series of minute step
fractures along segments at least 3.0 em long along one or both lateral
margins. Under ?Ox magnification, these segments have rounded edges, and
appear slightly polished. There are no visible striations.
2. The single hornfels .. prismatic blade 11 (UA-77-64-61+62+63) .·
exhibits possible wear patterns on both lateral edges, in the form of medium
sized step and feather fractures, and on the distal end. The latter is of
particular interest, in that it may have served as a "burin-like implement"
(terminology based on Mauger 1970:46). The distal end of this implement
terminates abruptly, forming a transverse facet at right angles to both
faces. The manufacturing technique which produced this facet is impossible
to determine, although it was most likely formed as a "snap" fracture rather
than by a deliberate burin blow. The surface of the facet is heavily
polished, with considerable blunting and polishing of the juncture between
the ventral surface and "burin"facet. When found, the artifact was in three
pieces, all within a 1.5 m of one another (See Fig. 18); the fractures which
had separated the three pieces suggest an extreme "bending" force while the
implement was being held by its proximal end.
3. A large biface, UA-77-64-1, could have been used in a hand-axe
fashion, as inferred from the slight edge damage on its distal end. The
nature of this damage is similar to stack-step fractures obse rved on an
127
experimental specimen used in a battering motion for 20 strokes. If it was
in fact utilized aboriginally, this specimen was used for only a short
period of time before being discarded.
4. Two "retouched" flakes, UA-76-212-385, and UA-76-212-172, both
have systematic minute step fractures on their distal ends. Modified margins
are 3.1 em long on one, and 1.7 em long on the other. Both appear similar
to flakes that were experimentally-modified by being used in a cutting mode
for less than 20 strokes.
It is thus suggested that several of the Carlo Creek lithics may
have been utilized briefly, probably in connection with animal dismemberment.
Due to the durable natura of the hornfels and argillite, it was not possible, ..
even on experimental controls, to develop "fail-safe" criteria for detennin-
ing sources of edge-damage. Even where controls were used for SO strokes,
it was difficult to observe changes. The additional complicating factors
such as animal fats (Brose 1975), "spontaneous" retouch (Brink 1978), edge
abrasion during manufacture (Sheets, 1973), and secondary alteration after
discarding make any statements concerning utilization somewhat equivocal.
The fact that primary butchering remains are associated with these
lithics suggests that these lithic materials were brought into the site for
the express purpose of dismembering the carcass of at least one caribou, one
sheep, and possibly nine ground squirrels (see Chapter 7). Tha relationship
of the lithic debitage and regional geology further suggests that these
materials were brought into the site from close by, and were modified and
used on-site. I regard these as "tools of the moment", \-Jhich achieved a
potentially-functiona1 stage quite early in t1e reduction process.
12
Stage 7: Disposal Mode: Introduction into
the Archeological Record
The final "stage" of the Component I lithics sequence is disposal,
i.e. the output of archeological remains from their systemic context into the
archeological context (cf. Schiffer 1976). For the sake of discussion, this
wi11 be viewed as a "stage", a1though in reality, discarding occurs at all
stages including quarrying, heat treatment, cortex removal, biface reduction,
and breaking of finished tools during resharpening or use. (See also the
discussion of spatial relationships in Chapter 4, as this relates to disposal
of lithics). Included in this fina1 ·11 stagen is some debitagefrom all prior
stages. The disposal mode is the only "stage" readily observed during exca-
vation of a site, and is in effect, the archeological 11 record". In the Ci.rlo
Creek Component 1 assemblage, discarded tools, broken tools, and products of
tool production mainly occurred as primary disposal. The only apparent ex-
ception to this is the single "prismatic blade", which was probably manu-
factured at another location, and discarded on site.
The Carlo Creek Lithic Reduction System: Component II
129
The small flake cluster found in the site 1 S upper level (see Fig. 23)
is clearly different from the flake scatters found in the lower level. Aside
from an obvious difference in material type, the lithic reduction technology
was noticeably different. In particular, differences can be noted 1n flake
size, percent of decortication flakes, and overall proportion of biface
thinning and shatter flakes. Especially significant is the presence in
Component II of platform abrasion, which may be taken as supporting evidence
for pressure flaking (not the only criterion). The following briefly sum-
marizes the Component II reduction sequence, as indicated at Carlo Creek.
It is not a complete sequence, for the complete sequence is n t represented
at the site. As previously mentioned, the total number of Component II
waste flakes is 637.
Stage 1: Raw Material Selection
The only rock type found in the Carlo Creek Component II assemblage
is white rhyolite. This lithic material was probably obtained locally, from
one of the numerous outcrops of rhyolite in the Cantwell Formation and in
the Central Alaska Range in general. Wahrhaftig (1958:14): reports that a:
" ... layer of white rhyolite 100ft. wide crops out on the crest of the
ridge between Riley Creek and the Nenana River about 1 3/4 miles due west of
the lagoon section house and 3 miles due south of McKinley Park Station."
:
(The latter is 14 km north of Carlo Creek). In addition, one of his rock
sampling localities for soda rhyolite is described as a " ..• hill (alti-
tude 2200 ft.) about 2 miles N 20° E of the mouth of Carlo Creek" (1958:15).
However, no aboriginal rhyolite quarries have yet been recognized in the
A 1 aska Range.
It should be noted that this material type is found widely, both
temporally and spatially, in archeological collections from interior Alaska.
Among collections I have personally observed this material type in are the
campus site, Dry Creek Component II, Teklanika east and west, Dragonfly
Creek site, and the Nenana Gorge Site.
Stage 2: Heat Treatment
It is not known whether or not the Component II rhyolite was
thermally heated. Although I did experimentally heat Carlo Creek aboriginal
rhyolite to 600° C, I did not have the benefit of non-aboriginal control
samples as in the case of the Component I argillite/hornfels. Until such
controls can also be heated, it will not be possible to determine the
l3Q
!
j
presence of heat treating in the Component II assemblage. I did, however,
observe a slight color change (increase in purple hue) and change in lustre
(glassier).
Stage 3: Cortex Removal
After a quarried piece of unknown size was brought to the site, the
remaining cortex was removed by direct percussion. The amount (12.5%) of
cortex flakes present indicates that probably most cortex was removed at a
local quarry source.
Stage 4: Biface Thinning
Analysis of a 22.6% sample of the 637 flakes in the Component II
cluster indicates that a direct freehand percussion technique was probably
employed by the aboriginal artisans during at least part of the Component II
reduction sequence. Platforms were prepared primarily by light abrasion
(33%), after which biface thinning flakes were detached. A number of small
flakes in the Component II collection suggest removal by pressure (multiple
removal flakes; 17% of total). These generally have pronounced lips, are
ovate to linear in form, and have a prominent bulb of force.
Stage 5: Finished Tools
No finished tools were recovered during the Component II excavations.
It thus can only be inferred what end products are represented in the
Component II assemblage. Observations of the rhyolite debitage from Compo-
nent II suggest that the tools being worked on that this locus were bifaces;
there is no evidence of core and blade production.
131
Stage 6: Tool Use
On the basis of the extremely limited data from the Component II
level (i.e., no artifacts), it is not possible to make any statement con-
cerning tool use. Also, because no faunal remains were recovered from
this upper level, we cannot infer lithic tool use as reflected by butchering
practices.
Stage 7: Disposal
As in the case of Component I, discarding of detached flakes may have
occurred at any stage of biface reduction. Since this entire system is not
represented in the Component II debitage, most Component II disposal probably .-
occurred during Stages 3 and 4. The disposal of debitage in the Component II
level appears to have occurred at the same locus as its locus of manufacture;
it is thus considered primary disposal.
Surrrnary
The primary purpose of this chapter has been to attempt to transcend
the static record of the lithic remains, and to place thm in their systemic
context of resource procurement, modification, and discard. While it can
never be stated with 100% certainty precisely what activities took place at
the site, we have at least attempted to explore the range of possibilities
as borne out by the data.
In essence, the Component I lithic system entailed the procurement
of argillite/hornfels from a local quarry source and perhaps a cobble or two
from the Nenana River gravel, probable heat-treatment on-site to improve
flaking characteristics of the raw material, biface reduction, and possible
use of several implements and flakes in the dismembering of several animals.
132
133
The Component II system was limited to secondary biface (?) reduction,
, probably utilizing both direct percussion and pressure flaking. The l imited
data from Component II do not permit a more precise statement than this,
however. As in the case of Component I, the Component II lithic raw materials
were probably obtained locally, within the Nenana Valley.
--
CHAPTER 7
FAUNAL ANALYSIS
Introduction
Faunal analysis, as defined here, is minimally the identification,
analysis and interpretation of bone remains from an archeological site.
Through faunal analysis, much important site interpretative data may be
realized, including aboriginal subsistence strategies, settlement patterns,
butchering practices, seasonality, and relative dietary importance of
animal species. The Carlo Creek research strategy included faunal data as
an integral part of the interpretation, from the perspective of both the
systemic and archeological contexts.
During the excavation of HEA 031, bone or shell remains representing
five different taxa were recovered. These can be placed into two general
categories: (1) artifactual and (2) non-artifactual. Artifactual faunal
remains are those associated with, or a result of, past human activity at
the site. Such remains occur non-randomly due to actions of human behavior.
They are purposely brought to a locus, and are discarded and/or modified
for one purpose or another. At the Carlo Creek Site, (Component I), arti-
factual faunal remains (N=306) represent three mammalian species: caribou
(Rangifer sp.), mountain sheep (Ovis dalli), and arctic ground squirrel
(Citellus sp.), (see Fig. 29, and Tables 7 and 8).
Non-artifactual faunal remains found at Carlo Creek include two
species of pulmunate snails: Succinea sp. and Discus cronkhitei. Both of
134
I
l
j
A
0
I I
B
em
I I
5
I I
.. ~r:il ~~
c
Fig. 29.--Representative Examples of 3 Mammalian Species
Found in Component I. A= Ovis; B = Rangifer; C = Citellus .
..
TABLE 7
COMPONENT I FAUNAL DATA
Anatomical Large Mammal Small Ma01nal All Remains
Element
Cariboua a Citellusa Spiral Total Sheep Unassigned Unassigned Burned Cut Fracture Elements
Skull 31 31 31
Mandible 9 9
Maxi 11 a 4 4
Metacarpal 2 l 2
Talus 0
Carpal l l 1
Metapodial 7 1 l 4 7
Phalange l 2 2
Phalange 2 1 1
Phalange 3 l l
Metacarpal
(vestigial) l l
Carpa 1 /tarsa 1 20 10 20
Long bone 16 23 l 4 39
Indeterminate 55 5 1 60
Incisor, upper 7 7
Incisor, lower 14 1 14
Incisor,
indeterminate 32 32
Premolar 7 1 7
Molar .... 68 7 68
TOTAL 7 1 129 141 28 51 2 10 306
a M.N.I. for caribou and sheep is l • M.N. I. for Citellus is 9. • .
LJ
a>
TABLE 8
FREQUENCIES OF SELECTED CHARACTERISTICS
OF THE FAUNAL ASSEMBLAGE FROM THE CARLO CREEK SITE
Characters Proportion of
Large Mammal Bonesa
Proportion of
Small Mammal Bonesb
Proportion of
Total Bones
Burned 31 .38% 5.33% 17.00%
Spiral fracture 7.29% 0 3.28%
Dentition 0 75.74% 41 .83%
Unidentifiabl ec 94.16% 16.57% 51 .30%
aLarge mammal bones comprise 44.77% of the total number of bones.
bSmall mammal bones comprise 55.22% of the total number of bones.
cThe frequency of unidentified bone is a relative index of frag-
mentation.
--
137
these snail species occur from 5 em to 50 em above the cultural level,
within the basal portion of Unit 3 (refer to Fig. 11). They are fairly
evenly distributed throughout this zone, and show no apparent relationship
with the main cultural areas of artifact and bone clusters. Snail samples
were collected from unit 3 sediment samples; no attempt was made to recover
all snails from that level. Discussion of snails at the Carlo Creek site is
reserved for a later section on paleoecology (Chapter 7).
Bone remains excavated from the site occurred only in the Component I
level. There were no indications of food storage pits, nor were there any
observable non-human features such as anima .l burrows. It is posited that
the animals represented by these bone remains •r~ere brought into the site,_
where they were discarded, modified, or utilized in one form or another by
the site 1 s prehistoric inhabitants. Faunal remains at HEA 031 are scattered
in and around the main area of lithic modification activities, and appear
to be most closely associated with Hearth 1 (see Fig. 18) in the Component I
level.
In discussing faunal analysis, Bonnichsen (1973:9) has aptly stated
that, "Faunal remains are a primary source material for interpreting
aboriginal subsistence patterns as well as identifying man 1 S presence. It
is more important than ever to be able to distinguish cultural from natural
systems of bone alteration." He further states that " ... before bone data
can be understood, the kinds of cultural and natural filters (cf. Reed
1963) through which they have passed must first be understood" (1973:13).
The following discussion will deal with these various "filters"
through which they passed in the transformation from living animal, to prey
species, to butchering, and finally, to the transformation from cultural-
ecological systemic context to the archeological record. Among other
1 38
considerations, it will be necessary to examine the ecology of mammals,
possible or potential aboriginal hunting strategies, decision-making
necessary to harvest these resources, the "schlepping" (cf. Daley 1969:149)
of animal butchering units from kill site to campsite, further "culling"
_(cf. Binford 1978a:460) of butchering units out of the site context, on-
site processing (cf. Binford 1978a:460) and utilization, recognizable
effects of human alteration of bone, disposal, and finally, post-depositional
alterations, preservation, and archeological recovery.
For a number of reasons, there is little doubt that the Component I
level bones are artifactual; this includes the ground squirrel remains as
well as the caribou and sheep. This argument is supported by the following
points:
1. The bones, together with lithic remains and charcoal, occur in a
stratigraphic/geomorphic setting which would make it virtually impossible
for them to have been transported and deposited in this area by natural
processes. The predominately sandy matrix in which they occur indicates
they were deposited by human agency. In addition, there is no sorting of
the bones according to weight or size, as one might expect if they had been
transported by water. There is no apparent sorting of lighter bone and
charcoal fragments relative to massive lithic specimens.
2. Ma!11T1alian faunal remains at HEA 031 are spatially related to
other aspects of human occupation-hearths and lithic concentrations (see
Fig. 18). This association is clearly not a random one.
3. As previously stated, there is no evidence anywhere in the site
for krotovinas or active animal burrows. This fact is especially important
in the consideration of the Citellus sp. remains. There is little doubt
139
that they occur at the site as a result of human action, and probably
represent a food resource for early Holocene hunters. Small mammal remains
were found in close proximity to the hearths, and at least nine bones were
charred.
4. Some of the bones have been altered in such a way as to indicate
human manipulation. Such alterations include charring (17.0% of total), cut
marks (0.65% of total), and "spiral fractures" (3.28% of total) directed to
the medial segments of long bones (cf. Sadek-Kooros 1972; Bonnichsen 1973,
1977' 1978' 1979) .
5. There are no suggestions for carnivore-derived bone alterations,
such as tooth perforation marks, gnawing, spiral fractures directed from -·
the epiphyseal ends, or partial digestion (Bonnichsen 1973:24). Many of
the bones at HEA 031 are splintered and fragmentary, which is taken as a
further indication of human actions. Such actions may have included bone
smashing to extract marrow, preparation of bone "grease" (Binford 1978a) of
various segments of the carcass.
Faunal Analysis: Methodology and Interpretation
In an attempt to establish tight spatial controls, faunal remains
excavated from HEA 031 were mapped in to the nearest centimeter. All in
situ mammalian faunal material was collected from the site, while. non-
artifactual gastropods were collected randomly, as part of Unit 3 sediment
samples. Bone remains were stabilized, where necessary, with vinylite
resin and/or polyethylene glycol. The majority of the bones, anced thawed
from the frozen Component I level, needed little or no stabilization, and
were bagged without further treatment.
140
Preservation of faunal remains at the Carlo Creek site is regarded
as relatively good, despite the small sample recovered from the site. A
large proportion of the recovered pieces were fragmentary; the size of these
fragments is interpreted as having resulted from human activity, rather than
from post-depositional changes at the site. The fact that small and compara-
tively delicate bones such as ground squirrel mandibles (N=9) were preserved
attests to the site 1 S quality of organic preservation. In addition to the
presence of frozen ground, preservation in the site may also be attributable
to the base-rich sediments in the vicinity of Unit 2 and Unit 3 contact.
In discussing the excavation of bone at Carlo Creek, mention should
be made of a probable recovery bias in favor of large-mammal and large
fragment remains. Guthrie (1968:224) has previously addressed this question:
Small mammal remains are more subject to long distance transport
by either physical or biotic forces ... because of their small size,
the bones of small mammals would be less resistant to destruction by
exposure and crushing than large mammal bones, but might be more
easily covered by sediment and preserved.
Similar observations have been made by Plaskett (1977), Repenning
et al. (1964), and by Ziegler (1965). The first study (1977:114) suggests
that a significant recovery bias may result from screening backdirt through
l/4 in. mesh screens. (Although a number of bulk soil samples were collected
from Component I, most of the excavations employed the use of a l/4 in. mesh
screen.) Thus, for these reasons, any MNI figures (minimum number of
individuals) developed for the Carlo Creek small mammal data must be regarded
as conservate estimates.
Laboratory analysis of the faunal collection proceeded as follows:
Individual bones or bone fragements were first cleaned, cataloged, and if
necessary, further stabilized with vinylite resin. Individual fragements or
bones were then classified with data recorded for: (1) cata ~og number,
141
(2) provenience, (3) anatomical element, (4) side, (5) species, (6) age, and
(7) presence or absence of modifications (charring, cut marks, or spiral
fracture). The systematic variable list utilized for this project was
modified and simplified from Bonnichsen and Sanger (1977:115 ), with influ-
ences from Binford (1978a), Chaplin (1971), and Grayson (1973).
Identification relied upon comparative specimens from the faunal
laboratory of the Washington State University Anthropology Department,
Conner Museum, Washington State University, and the vertebrate paleontology
collection of the University of Alaska Museum. Although all identifications
were made by the author, several specimens were examined for confirmation by
faunal experts Carl Gustafson, Chris Brown , S. Kent Harkins (all of WSU), ..
and R. Dale Guthrie (University of Alaska). In addition, several published
and unpublished keys were used (Glass 1951; Gustafson n.d.; Hall and Kelson
1959; Lawrence 1951; MacDonald 1978; and Schmid 1972).
For the purposes of classification, the analytical concept of size
working group was evoked. Bones were divided up first into two major and
largely subjective categories based on relative size. After each fragment
or complete bone was classified according to the above procedures, the data
were entered in Table 7.
As indicated in Tables 7 and 8, most of the 306 bones from the site
are small fragments, which precluded positive identification of more than
51% of the assemblage . The most commonly-represented elements in ~he
collection are tooth fragments (n = 128; 42 % of total), all of which are
of Citellus species. Dentition was the best-preserved portion of the
faunal assemblage. Molars and premolars were, for the most part, intact.
In several cases, they were still articulated within a maxilla or mandible
fragment.
142
A number of cranial fragments (31) were excavated from within the
limits of Hearth 1 (see Fig. 18). These were all in a fragmentary and
charred condition, and were identifiable as cranial fragments on the basis
of unfused sutures, and presence of trabeculated bone between two compact
bone layers. On the basis of relative size, they appear to be large mammal
remains, probably caribou. I cannot, however, rule out the possibility
that they are human. The largest fragment measured 25 x 19 mm. None of
these specimens were identifiable according to species, although their
proximity to other caribou bones leads me to suspect they are of that
taxon. The position of the cranial fragments within the hearth, their
fragmentary condition, and the fact that they were found in an ''inverted'' ~
position (inside of skull facing upwards) suggests the cranium was purposely
smashed, perhaps to extract brains. The non-closed sutures suggest a young
adult or immature animal.
Identifiable large mammal remains, with the exception of the cranial
fragments, are restricted to either fore or hind limb bones: metacarpals (2),
carpal (1), unassigned metapodials (7), and phalanges (4). Seven of the
above bones appear to be from a single caribou limb. The presence of sheep
is indicated only by a single 3rd phalange.
The one sheep bone is worthy of special note. It was found out of
context, during the initial 1975 investigation of the site. Found protruding
from the Component I level, it was associated with argillite flakes and one
biface fragment. I have estimated its position within the excavation grid
as square 1-2 N, 3-4 W. There is little doubt, however, that this represen-
tative of the taxon Ovis was associated with the Component I occupation.
The identification of this specimen was cross-checked a number of times,
using a minimum of six comparative specimens. The fact that ~heep is
143
represented by only one phalange is probably due to ·the highway construction
activities which first exposed the site. It is quite likely that there are
fragmentary remains of at least one early Holocene-age mountain sheep
scattered somewhere in the fill of the George Parks Highway.
Because of the fragmentary nature of the Carlo Creek large mammal
remains, it was not possible to determine the sex of those specimens.
Determining age of individuals represented by the bones likewise was not
possible. In several instances, it was possible to make general statements
regarding the apparent degree of epiphyseal fusion or suture closure; these
observations were not sufficient, however, for making definitive statements
on the age of the animal. No attempt was made to determine sex or age of:
_the small mammal remains.
Intentional human-induced (?) modifications of bone were identified
by several criteria (see Table 7). Charring or calcification of bone was
determined simply by color observations (charred black or calcified white,
as opposed to unaltered brownish color) and/or hardness (e.g., denseness and
good preservation of bone resulting from thermal alteration). Cut marks
were indicated unambiguously on only t\-.10 samples; these both show tiny
grooves transverse to the main axis of the bone. Neither of these specimens,
a metapodial and long bone, were identifiable as to species. Presence or
absence of spiral fractures was determined through application of previously-
establi ~hed comparative critera (Sadek-Kooros 1972; Bonnichsen 1973, 1978,
\
and personal cOJTTtlUnication, 1977). A total of 10 specimens possessed
spiral fractures: 1 caribou metapodial, 4 unassigned metapodial fragments ,
4 unassigned long bone fragments, and one completely undiagnostic bone
fragment. It should be noted that presence or absence of ~of these
traits does not in and of itself imply human alteration. However, when
144
viewed in terms of the total assemblage and intrasite spatial relationships,
the presence of certain traits may be of interpretative importance.
Following classification of bone fragments and human modifications,
the faunal data were manipulated to derive estimates of minimum number of
individuals (MNI). MNI is here defined as the lowest number of individual
animals present from an archeological locus, which are necessary to account
for all the elements (or fragments thereof) of a particular species (modified
from Shotwell 1955). Although a variety of formulas have been proposed for
MNI calculation (e.g., Chaplin 1971; Grayson 1973) the simplified expression
of the concepts, as used here, appear to me to most accurately estimate true
MNI.
One problem encountered in interpreting these data has resulted
from the fragmentary condition of the bones, which necessitated a slight
revision of the above formulations. For example, if ten metacarpal fragments
were identified, we cannot assume automatically that ten metacarpals, and
accordingly, five individuals are represented. The faunal analyst must do
some anatomical reconstruction of the fragments of each identified bone to
obtain the smallest number of complete elements represented by the fragments.
If we were to assume that our hypothetical ten fragments were all from the
same metacarpal, then we would have identified only one individual from the
site. One can readily see the potential for error in this mis-appJication
of these formulae. Similar problems in MNI determination are discussed in
greater detail by Binford (1978a) and by Grayson (1973).
Given the above parameters, calculation for MNI in the Carlo Creek
Component I assemblage indicates l caribou, represented by 7 elements, 1
sheep, represented by l element, and at least 9 ground squirrels, represented
14 5
by 141 elements. The figures for the caribou and sheep are regarded as
fairly accurate assessments of the data. The estimate of Citellus sp. MNI
is quite definitely a minimum one, due to the previously discussed sampling
bias in favor of large mammal remains.
One final level of abstraction from the data will be mentioned here.
For a number of reasons, however, its utility in interpretation of the Carlo
Creek data is questionable.
A number of authors have attempted to calculate the weight of meat
which may have been available for human use at a given site (e.g., Hall
'1971; Lyman 1979; Plaskett 1977; Thomas 1969; White 1953a). In an earlier
preliminary presentation of the Carlo Creek faunal data, I attempted a
similar calculation (Bowers 1977a:l3). The major aims of these kinds of
analyses have been to try to determine potential diet, carrying capacity,
and, ultimately, population sizes for a given site.
.·
Using standard formulas (Lyman (1979; Thomas 1969; White 1953a) it
can be estimated that the large mammal bones found in Component I represent
8.49 lbs. (3.82 kg.) of meat (Table 9). While other studies have attempted
to extract dietary and population inferences from such a figure, it is felt
that this kind of projection would be inappropriate for the Carlo Creek site
data.
The first problem encountered is the fact that the sites boundaries
(i.e., sampling universe) are ill defined, due to destruction of part of the
site, and/or the lack of definition of the eastern boundaries due to excessive
overburden and manpower limitations. These factors preclude any exact
statements concerning diet and nutritional "reconstruction" for the site as
a whole based on faunal remains.
146
TABLE 9
CALCULATION OF POTENTIAL AVAILABLE MEAT WEIGHTS
Portion Caribou Da 11 Sheep
( 1 bs) ( 1 bs)
Live weight 240 180
Carcass wei ghta 168 131
Butchering unitb 8.4 6.5
Meat weight/butchering unite 4.8 3.7
ADAPTED FROM: White 1953a; Lyman 1979.
aCarcass weight is 70% of live weight for cari-
bou, 73% of live weight for sheep.
bin this case, a butchering unit is assumed to
be either a hindshank or foreshank; weight is about 5%
of carcass weight.
cMeat weight is about 57% of butchering unit
weight.
147
A second problem in determining dietary potential from faunal
remains is the assumption that only fresh meat was used at a site. In
addition to this potential use, meat obtained from the lower limb of a
caribou or sheep may also have been used for dried meat, sinew, marrow, bone
"grease 11 (cf. Binford 1978a:32-39; Hhite 1953b), dog food, or trap bait.
Additionally, such bone remains may represent raw materials for artifact
manufacture. One of the key questions, given these possible uses, is how
much meat was actually left on the bone and thus available for human use?
Furthermore, the assumption that the meat was used ~site is
perhaps the major source of error in faunal interpretation. It is not a
148
reasonable a priori assumption that an entire carcass was used at a given··
site. Modern ethnographic observations (cf. Binford 1978a) suggest that,
more often than not, an animal is butchered into units for the purposes of
transport and/or further modi fi cation. If such ethnographi ca 11 y-observed
"culling" (Binford 1978a:460) of selected anatomical units is correct, we
should be able to predict the anatomical portions represented at functionally-
specific types of sites.
The Carlo Creek data suggest that most of the meat obtained from the
caribou and sheep was probably used elsewhere, and that HEA 031 was a killsite
and/or primary butchering site. The fact that bones were cracked and
splintered tends to compound this interpretation, suggesting utilization of
marginal-value bones for marrow or bone great extraction. Only lower limb
bones (phalanges up to metapodials) are represented at the site, in addition
to the few cranial fragments . Such an assemblage is highly suggestive of a
kill station or "field" butchering site, where a dispatched an i mal was
processed, and the low value meat units (e.g., lower legs, skull) were
selectively culled from the dismembered carcass. Presumably, the high-yield
\
1
1
portions (higher caloric potential) were transported to a "central base "
camp (terminology based on Chang 1962 and Campbell 1968).
A number of ethnographic sources demonstrate analogous selectivity
in faunal assemblages. For example, White (1953b:162-163) discussed bison
bone remains from two archeological sites near Pierre, South Dakota:
If the kill was some distance from the village, the lower legs
were removed by chopping through the wrist and ankle ... Since the
lower limb does not carry any usable meat, it is conceivable that it
was chopped off, either through the distal end of the radius or
through the carpus, and left at the place of kill in order to reduce
the load .
Bonnichsen (1973:11) has described a similar situation in his
observations of butchering practices among the Calling Lake Cree in Alberta:
''The rather nonproductive carpals, tarsals, and phlanges are not broken for
marrow. In fact, they are not represented in the Calling Lake assemblage,
a fact which suggests that they were left in the field with the animal's
intestines."
One of the best interpretative sources for the Carlo Creek faunal
data is Binford's recently published Nuniamuit Ethnoarcheology, which
describes in detail the economic anatomy of sheep and caribou, butchering
practices, types of sites, subsistence/procurement strategies, and
subsistence/seasonal patterns of a Brooks Range (Alaska) inland Eskimo
group .. His study, based on observations of 277 caribou dissections at 411
separate kill-butchering locations, provides comparative da t a which one can
apply cautiously to interpret the archeological record:
The faunal remains in hunting camps have a characteristic
signature in that some culling strategy is invariable practiced ...
the relative success and the attendant variations in transport
problems may condition the particular forms of culling strategies
employed (1978a:342).
149
The "culling" strategy as applied to kill locations is the frequently-
observed discarding of the lower limb bones, by smashing through the
metapodials, leaving behind distal metapodial and phalanges. The skull is
likewise discarded at the killsite, to enable easier transport of the high-
yield meat-bearing anatomical elements back to the habitation site.
Binford noted significantly higher percentages of distal leg and skull
portions of caribou remaining in dispersed kill/butchering locations
(l978a:76-77).
The fact that all of the Carlo Creek large mammal remains have been
cracked and/or splintered may be interpreted as indicating extraction of
marrow or bone grease (Binford 1978a:32) preparation. As suggested by
several of the above-cited ethnographic sources (Binford l978a; Bonnichsen
1973; White 1S53a), it was common practice to make use of lower yield
bones at the killsite, rather than squander high-yield meat or carry low
value portions back to camp: " ... thus the bones were left behind. The
marrow and sometimes the brain were eaten as the butchering went on "
(Wissler 1910:41-42).
In discussing the calorific-maximizing strategy employed by the
Nunamiut, Binford makes the following observations, suggesting that intensive
use of low-value bone sets could indicate hardship conditions. These
observations offer a possible glimpse into the bone processing that took
place some 8500 years ago at Carlo Creek:
If food is in short supply, or if marrow bones from the legs are
not available at the meal, consumers may break the phalanges for
marrow. Today this is accomplished by placing the articulated first
and second phalanges on an anvil stone and striking them a transverse
blow with the back of a knife blade in the middle of each shaft.
This results in breaks in the middle of each phalange. These are
picked up and separated midshaft. The units recovered are (a)
proximal first phalange, (b) articulated distal. first phalange and
proximal second phalange, and (c) distal second phalange. These are
150
i
l
i
sucked and the marrow licked out of the small cavities. Once
finished they are tossed in the bone dish for subsequent dumping by
the woman of the house. Articulations remaining from such a meal
are (a) vestigial metapodials and vestigial phalanges and (b) distal
first phalange and proximal second phalange. Since this is a--
11maximizing11 meal--· that is, only prepared when stores are low the
remains from such meals are rarely mixed with other anatomical
parts. The resulting dump is almost always discrete resultin9 in a
pile of almost nothing but phalanges (Binford 1978a:148-149).
On the basis of such ethnographic analogies, the Carlo Creek faunal
data may best be interpreted as the remnants of a kill site or primary
butchering site. One may speculate that the sheep and caribou represented
at the site were killed fairly close by, then brought into the site for
processing. Concomitant with carcass dismemberment and preparation of high-
value portions for further transport might have been the on-site use of lo~.
value anatomical elements (e.g. lower leg and skull) for marrow extraction,
bone grease preparation, and brain extraction.
As pointed out earlier in this chapter, it is difficult, if not
impossible, to make projections concerning occupancy or population size;
particularly as based on MNI or MT. WT. calculations:
At kill sites and processing locations, the number of animals
present in the faunal asse~blages bears little if any relationship
to the number of consumer days of occupation. In all these
logistical locations, the relationship is systematically in favor of
the presence of many more animals than are needed to sustain the
occupants during the use of the location (Binford 1978a:448).
Thus, on the basis of the Component I faunal data and ethnographic
comparisons with contemporary hunters in northern Alaska, we have been able
to posit a number of statements concerning site function and activities.
Although the large-mammal faunal assemblage from Carlo Creek is small, it
is distinctive, and clearly suggests the nature and extent of human activites
which occurred at the .site some 8500 years ago.
151
Seasonality and Aboriginal Procurement Strategies
The faunal data from HEA 031, in addition to offering insights into
site utilization and duration of occupancy, may also be useful in inferring
seasonality. The fact that no antler, horn or other season-specific
epiphysial elements were present in the large mammal remains from the site
precludes the use of actual bone for determination of seasonality. However,
a number of inferences can be made, based on the present-day ecology of the
three mammalian species present at the site: ground squirrels, sheep, and
caribou.
On the basis of available data, a late summer or fall utilization
is suggested. The following observations may be marshalled in support of
.·
this interpretation:
1. The presence of numerous ground squirrel bones at the site, in
association with other occupational debris, point towards use of the site
during the warmer months, before winter freeze-up. Ethnographically, it is
reported that Citellus parryi was taken by a variety of methods, almost
exclusively during the late spring, summer, and early fall (e.g., Bee and
Hall 1956; Campbell, 1968; Gubser 1965; Ingstad 1954; McKennan 1959, 1965;
Nelson 1973; Spencer 1959). There is little evidence in the literature
that these animals were harvested during hibernation periods in the late
fall and winter, nor would it seem feasible. Indeed, such cold weather
procurement of ground squirrels would present difficulties for the hunter,
in that extracting the sleeping animals from snow-covered and frozen
burrows might have required expending more net energy than could be realized
in return.
2. Based on the observations of Bee and Hall (1956:55), Carl (1962:
51), and Tikhomirov (1959:33), daily average temperature of about -15°C
15 2
1
to -zooc are necessary before ground squirrels hibernate. If it is possible
to project such patterns back to the early Holocene, we might surmise that
the use of local ground squirrels--and hence occupation of HEA 031--took
place while temperatures were higher than this. Given present climatological
data for the Nenana Valley, this would suggest that the site was occupied
after early April, and before early November. However, this assumes that
(l)thc ground squirrel bones at the site were deposited as freshly-killed
animals, not as dried meat, and (2) ground squirrels were not dug out of
their burrows during hibernation.
3. Such a seasonal pattern may be further supported by ethnographic
data. Campbell, in a study of the North Alaska Brooks Range Eskimo, pro-
vides the following excellent observation:
With the exception of two ungulates, the arctic ground squirrel
was the most important food mammal. Adult arctic ground squirrels
typically weigh only two or three pounds, but they are usually fat,
are easily captured~ and are abundant north of the forest. They
hibernate in winter, and are then very difficult to obtain, but in
summer, which is traditionally a time of hunger ... a few trappers may
take a dozen or more day after day for several weeks in an area of
only a few square miles. For this reason, during the summers the
Tuluaqmuit depended heavily on them. Arctic ground squirrels are so
common and so widely distributed north of the tree line, that they
had little to do with the precise locations of summer settlements
within the tundra zone. However, the fact that this species is an
open lands mammal very probably had much to do with the Tuluaqmuit
dwelling on the tundra rather than in the forest during the warm
season (Campbell 1968:13).
..
Carl (1962:83) reports that Arctic Ground squirrels were frequently
caught by Eskimos of northwestern Alaska, by use of traps, sticks, hands,
rifle, and were frequently taken by children with stones. He states that
11
••• most of the squirrels are caught near the village or near seasonal
camps.11
McKennan (1959:60-62; 1965:32), Nelson (1973:144), and Osgood
(1970:111) all report the use of ground squirrels by inhabita ~ts of the
153
taiga zone. They were harvested in the past by snares and deadfalls. It
should be noted that, during the historic period, ground squirrels were
frequently looked upon as "famine" food, taken primarily when large game
was not plentiful. In addition, squirrels are today, and probably were in
the past, regarded as sources of high quality pelts. The Inupiat word for
these animals is Siksrik, translated roughly as "parka squirrel" (Webster
and Zibell 1970:76).
4. As stated in Chapter 2 of this thesis, the occurrence of caribou
and sheep in the Nenana Valley area is presently most frequent during late
summer to early winter (Alaska Department of Fish and Game 197J; Guthrie and
Powers 1977; Hemming 1971; USOI 1976; Van Schoonhoven 1900; Yanert 1900).-
Whitten (1975:71) states that the main Alaska Range supports many sheep in
the summer, but that winter habitats are generally restricted to the outer
ranges. A similar seasonal pattern may have been operative during the
early Holocene. Sheep reportedly are taken most frequently in the Brooks
Range during the period July through October, and begin coming down into
the lower elevation mountain valleys in late summer and fall (Binford
1978a; Bee and Hall 1956). Binford (1978a) observes that distal leg remains
of caribou (of the sort found at HEA 031 ), tend to occur in greatest
proportions in archeological sites in the fall.
Taken together, the above observations suggest that the Carlo Creek
site was most likely occupied during the late summer or early fall. The
presence of a sizeable number of ground squirrels at one locus argues that
they were not harvested during the colder months. Present patterns of
seasonal movements of caribou and sheep suggest these animals were in the
Carlo Creek area in greatest densities during the late summer to early
154
winter. Ethnographic patterns of seasonality and subsistence strategies
indicate the most intensive aboriginal use of the area occurred in the fall
and early winter.
In terms of aboriginal procurement techniques, it is probable that the
caribou represented in the fossil record at Carlo Creek was obtained through
any one of four ethnographically-documented hunting strategies: intercept,
pursuit, lure, and drive (Burch 1972:339-368; Plaskett 1977:179). Sheep
may have been obtained by all of the above methods except for lure (Plaskett
1977:179). Ground squirrels may have been collected by snaring, deadfall,
trapping, lure, or pursuit.
To summarize, the mid-valley location of the Carlo Creek site was ....
probably a strategic one for hunting. The probable aboriginal use of the
site was in late summer or fall. During Component I occupation, at least
nine ground squirrels were taken locally, in addition to one sheep and one
caribou. The large m&nmals were butchered at the site; the carcass
portions of relatively high dietary value were removed from the site,
perhaps to some "regional base" type of camp. The left-over large mammal
faunal remains at the site were utilized for marrow, and possibly in bone
grease preparation. The bones found at the site are all representative of
relatively lower dietary value butchering segments. On the basis of
available data, aboriginal group size cannot be determined; however,
several hunters, a small hunting band, or an extended family would be a
reasonable estimate for the nature of the social unit exploiting the site.
155
CHAPTER 8
PALEOCOLOGY
Regional Paleocology
Precise determination of early Holocene vegetation in the upper
Nenana Valley is not possible at this juncture. One of two pollen cores
collected from Otto Lake in connection with the Dry Creek Early Man Project
has yielded a maximum age of 5620 ± 250 years: 3670 B.C. (W-3888) (Ager,
personal communication, 1979). The cores from Otto Lake indicate that
spruce was established in the northern foothills of the Alaska Range by at
least the mid-Holocene. However, they tell us little about regional pre-
Hypsithermal vegetational changes.
The vegetation sequence for interior (unglaciated) central Alaska
is somewhat better understood. Based primarily on the work of Ager (1975
and personal communication, 1969), the following sequence is indicated for
the Tanana Valley lowlands.
From at least as early as 16,000 B.P. to 14,000 B.P. a late glacial
"tundra steppe" environment zone is indicated, consisting of
... grasses, Artemisia, some sedges, and a number of opportun-
istic herbs such as Plantago, Taraxacum, and various Compositae and
Cruciferae. Climate during that full-glacial interval was extremely
continental with very severe winters and dry, warm summers of short
duration. Mean annual temperature was perhaps in the range of -7°C
to -l2°C, whereas at present it is about -4°C in the lowland.
Pollen of spruce and alder are nearly absent in sediments of late
Wisconsin age from the region ... (Ager 1975:87). (Note: Ager
[personal communication, 1979] suggests that his earlier figures
for full glacial temperatures should be viewed extremely conserva-
tively).
156
At approximately 14,000 B.P., a sudden climatic change evidently
occurred, resulting in a change from xeric steppe tundra to more mesic
shrub tundra. Evidence from this pollen zone (2) indicates that shrub
birch, willows, sedges, grasses, and heaths were common in the area (Ager
1975:87).
The first indications of spruce migration into the lowlands occurred
as early as 11,000 B.P. and most likely, at about 9500 B.P. (Ager, personal
communication, 1979). By 8000 B.P., spruce-birch forests had more or less
dominated the shrub tundra (Ager 1975:87).
The past 9400 years, as represented by Ager's subzone 3B, indicates
increasingly high percentages of alder, with a noticeable decline in spruce
in the interval from 8400 to 7000 B.P. (1975 :88). Ager has suggested that
this decline could be explained either by an increase in frequency of
forest fires, brought about by warm dry conditions, or alternately, by
forest pathogens (Ager 1975:88).
The final 6500 years of the Tanana Valley pollen record shows
little change. However, it is unlikely that tree-line fluctuations would
have been recorded in these lower elevation sampling locations (Ager
1975:88).
Of significance to the Carlo Creek study is the question of the
time of spruce appearance in the Tanana Valley, and more importantly, its
appearance in the higher elevation mountain passes of the Alaska Range.
Ager's sampling localities are at elevations of approximately 400 m above
sea level (1975:3); in contrast, the Carlo Creek area lies at an elevation
of 600-700 m a.s.l.
-
Supporting evidence for the presence of spruce in the Tanana Valley
is indicated in the work of Matthews (1970, 1974). He reports that spruce
157
158
forest was probably established in the vicinity of Tofty, central Alaska,
by ca. 7300 B.P., while a fairly contemporaneous forest presence is documented
for the Fairbanks area by at least 8080 B.P. (Matthews 1970:249).
A more precise dating of the beginning of the taiga forest in the (
interior is indicated by Matthew's (1974) pollen percentages for Picea and
Betula in the Isabella basin near Fairbanks. There, his zone B-C boundary
indicates the occurrence of a major vegetation change between about 7800 to
9200 years B.P. (Matthews 1974:838). Matthews (1974:838) has followed
Pewe's suggestion that the .. grassland .. to forest change had occurred by at
least 8500 B.P.
West (1972:15) reports that the pollen work of Schweger in the ..
Tangle Lakes vicinity indicates that spruce was present at that relatively
high altitude (ca. 879 m a.s.l.) by about 9100 B.P. The present day ecosystem
in the Tangle Lakes region has been described as 11 Shrub tundra, low sedge,
and grass tundra, and low matted tundra with occasional open stands of
b 1 ack and white spruce .. (West 1974 :217).
The work of Denton and Karlen (1977) may offer perhaps the best
data for inferring the elevation of the upper Nenana Valley early Holocene
treeline. They report (1977:103) a date of 8020 = 120 years: 6070 B.C.
(Y-2302) on a buried spruce stump which occurs at an elevation of 1067 m
a.s.l. (183m below present treeline) in the White River Valley, northern
St. Elias Mountains. They interpret this date as a minimum age for early
Holocene forestation in the valley. This age also appears to correlate
fairly well with Rampton's (1971) date of about 8700 B.P. for spruce migration
into the northern St. Elias Mountains.
Taken together, the above data would indicate that spruce could
have been established in the upper Nenana Valley by the time of Carlo
Creek I occupation. Obviously, the details of this interpretation are
obscured by the lack of good pollen data from the site itself. It is
probable that the constant shifts in the course of the braided Nenana River
would have resulted in a series of early successional stages of vegetational
communities, as is described below. Whether the vegetation in the area was
a sparse scattering of spruce in a shrub tundra or herbaceous tundra associa-
tion, or a denser closed spruce forest is conjectional at present. However,
dating from other parts of the interior Alaska do suggest that the treeline
may have risen to at least 600 m in the Central Alaska Range by at least
8500 B.P. Ager (personal communication, 1979), has stated that 11
••• it is
very difficult to confidently differentiate between closed spruce forest
and open spruce-shrubby tundra on the basis of pollen data alone. It is
159
also difficult to differentiate treeless shrub tundra from open spruce
forest, if the shrub site is within 20-30 km or so of spruce-alder sources 11
•
Local Paleocological Setting
As discussed in the geology chapter, an actively aggradating flood-
plain has been inferred for the time at which the Component I human occupation
took place. On the basis of mammalian faunal analysis, a late summer-fall
utilization of the site has been suggested. It may be postulated that the
site was fairly dry, at least during the brief periods of human occupation.
Without detailed modern studies of the Nenana River floodplain for comparison,
it is not possible to reconstruct the exact setting of the camp in relation
to the river. One might surmise, however, that the hearths were utilized
within a few tens of meters from the river, on a dry portion of an abandoned
river braid bar.
It is possible, too, that a spring which is present at the site
(presently used extensively by the local inhabitants), was flowing in
roughly its present pattern. Based on the site's present position in the
post-glacial terrace sequence, it is fairly certain that the locality would
have commanded a fairly good view of the floodplain, for approximately
3.0 km to the south, 1.0 km to the west, and 2.0 km to the north. The
position of the site at an inflection point in the river's course (see
160
Fig. 9) may indicate a fairly wide sandy braid-bar deposit on which occupation
occurred. It is thus probable that a trek of only a few kilometers would
enable the inhabitants of the site to obtain necessary water, lithic resources,
and probably, willow or alder for firewood and shelter. Within 0.5 km to··
the east, a higher terrace would have provided an excellent vantage point,
with a good view of the entire valley.
It has not been possible to determine the vegetation at the site
during the time of occupation. Pollen is not well enough preserved at the
site to permit detailed palynological studies. It is quite possible that,
despite the frozen strata overlying the early cultural deposit, episodes of
oxidation/reduction during the early phases of Unit 3 deposition would have
been sufficient to destroy much of the fossil pollen.
During recovery of sedimentological samples at the site a number of
pollen samples were systematically collected. A single sample from the
Unit 2/3 contact "paleosol" was submitted for analysis to Thomas Ager, USGS
palynologist. His report on the sample indicated that pollen preservation
was poor: "The sample from the site itself is nearly barren of pollen and
spores. There are a few grains of resistant spores like Lycopodium,
monolete fern spores, and occasional pollen grains of spruce, and alder"
(Ager, written communication, March 1978).
l
I
[ .
On the basis of the local pollen data it is thus not possible to
reconstruct early Holocene vegetation. It is probably that spruce may have
been present in the valley at ca. 8500 B.P., along with alder and willow.
This assumption is by no means unreasonable, based on the limited pollen
preservation at the site, and given the supporting data from other parts of
the interior Alaska, especially Denton and Karlen 1 S 8020 year old dated
spruce from the White River Valley (1977:103), and Schweger 1 S 9100 B.P.
evidence for spruce migration into the Tangle Lakes area (West 1972:15).
Data on terrestrial invertebrates collected from the site may offer
additional evidence relating to the site 1 S paleo-environment. Two species
of gastropods occur stratigraphically from 5 em to approximately 50 em
above the Component I occupational level, within Unit 3 floodplain silts.
The occurrence of two species of terrestrial mollusca, Succinia sp., and
Discus cronkhitei, are generally associated with moist habitat, as in a
..
river floodplain (Dall 1905; Henderson 1931; Pilsbry 1939-48; Pevo~e 1975a;
and Taylor (1965). Kalas (Matthews 1974:835) has suggested an open woodland
type of habitat for Discus cronkhitei, while McCulloch et al. (1965) have
posited that these snails also inhabited 11 hypoarctic 11 tundra sites. Matthews
(1974:835) uses the association of Discus cronkhitei macrofossils, Picea,
and Rubus idaeus, in zone Aa of the Isabella Basin near Fairbanks to argue
for a probable 11
••• open, sedge-dominated environment with scattered spruces
and practically no alders". This, he suggests, may be roughly analogous to
the contemporary forest-tundra ecotone (1974:336). In addition, Repenning
et al. (1964: 183-185) report an apparent association of Succinea sp. in
sediments bearing plant macro-fossil remains tentatively identified as
Labrador tea (Ledum sp.) and cranberry (Vaccinium vitis-idaea).
161
I must add a note of caution, however, in the use of such species
as index fossils or in suggesting their possible utility as sensitive cli-
matic indicators. Pewe (1975a:88) summarizes Succinea occurrences in
central Alaska, and notes that they occur in loess less than 1000 years old
as 1vell as in strata of Wisconsin age. J. ~1ead (personal corrrnunication, 1976)
has indicated to the writer that Succinea sp. in particular, may be a
rather unreliable indicator of microenvironment as they are known from
throughout North America and have a wide ecologic amplitude.
On the basis of available data, however, we may posit a locally
moist environment for the period following Carlo Creek I occupation
predictable in an actively aggrading floodplain and may tentatively .·
162
speculate that a riparian open-canopy forest may have been in the valley.
Obviously, detailed pollen studies from the upper valley will be needed to
firmly establish early Holocene movements of spruce into areas of successively
higher elevations.
Viereck (1966) has summarized the modern day vegetational succession
on gravel outwash from the Muldrow Glacier, Central Alaska Range, located
at elevations of 730 m to 760 m a.s.l. His observations may provide a
useful analogy in interpreting the paleoenvironments of the Carlo Creek
site cultural occupations. His five stages (vegetation stands) of succession
are thought to span the intervals 25 to 30 years, 30 to 100 years, 150 to
200 years, 200 to 300 years, and 5000 to 9000 years, after initial subaerial
exposure of glacial outwash sediments (Viereck 1966:198):
The vegetation development progresses from that of scattered
mat plants with isolated willow and Shepherdia shrubs to a closed
grassy meadow interspersed with small but dense clumps of willow
shrubs and eventually replaces them. The low shrub birch forms
a continuous and even stratum underlain by a thick moss layer
of Hylocomium splendens and Pleurozinium schreberi. The final
stage, the climax tundra, consists of low shrub birch and erica-
ceous shrubs interspersed with Eriophorum vaginatum tussocks growing
through Sphagnum and other mosses.
1
J
l
l
On the basis of these observations (Viereck 1966), the Carlo Creek
Component I data may be comparable with Viereck 1 S stage I and/or stage II
vegetation stands. Relatively minor proportions of organic matter in the
paleosol at the contact between Units 2 and 3 at Carlo Creek suggest that
the development of the soil (inceptisol) was in its early stages of develop-
ment; hence vegetation was probably restricted to early successional stages.
The Stage I vegetation described by Viereck (1966:186-187) is composed
primarily of 11
••• mat and clump species which attain a height of only a few
centimeters above the surface ... visually estimated cover is 50%11 • He
indicates that Dryas sp. and legumes are a major constituent of these
communities, even though their total cover is less than 20 %. --
In addition, a few shrubs in 11 Stage I 11 as tall as 1.5 m represent
approximately 1% of the total cover; these include Salix spp . and Populus
spp. (Viereck 1966: 187). His suggested age of 25-30 years for development
of Stand I would probably be a fairly good estimate for the amount of time
it took the Component I soil to develop. Floodplain alluviation would in
all likelihood have occurred fairly rapidly after this region 1 S human
occupation, truncating the soil ~evelopment processes.
Stand II from the Muldrow Glacier area, as described by Viereck,
was found to consist of large groups of Elymus innovatus, Festuca altaica,
163
and Poa sp. (Viereck 1966:187). These species vi rtually covered trye abandoned
braid bar, he observed, and included a number of herbs, other grasses,
willows, and low shrub birch (1966 :187).
Using a model for vegetation development such as Viereck 1 S (1966)
study, it is postulated that as many as four of his five stands may have
been represented in the early Holocene Nenana valley 1 s various floodplains
and terraces . These communities were probably dominated by ea rly successional
r
phases of plant growth, which would have provided attractive grazing areas
for large mammal herbivores. A similar idea was previously suggested by
Ager (1975:85) who pointed out that well-drained outwash and eolian deposits
may have provided good habitat for large grazing mammals, and may have been
predominately " ... vegetated by pioneer plant types such as grasses and
herbs" (1975:85).
This hypothesis also has been stated by Pewe (1975a:l02), who
pointed out that floodplains of braided rivers such as this were probably
major refugia for late Pleistocene herbivores due to the "disclimax" vegeta-
tion which favored grasslands. He observed (1975a:l02) that such areas
have a low permafrost table, are well drained, and have little muskeg. .·
Particularly in the winter, such environments would have provided ideal
habitat for herbivores, particularly in the wind-swept, snow-free floodplain
areas. Thus, it might be inferred that a variety of local plant communities
were available on the Nenana River floodplain for grazing herbivores and
concominant human exploitation. Early successional plants such as grasses,
sedges, Salix sp., Alnus sp. are common today on abandoned floodplain and
braid bars in the Nenana Valley; these observations might provide a useful
analogy with the early Holocene situation. This mountain pass is today an
environmentally-attractive zone for many game animals, both in terms of
vegetation, and as a natural "arteri' for funneling animals through the
Alaska Range. A site such as Carlo Creek would thus have been ideally
located to intercept animals exploiting the area's rich and diverse
vegetation.
164
I
J
)
I
[
•'
l
1
CHAPTER 9
PREHISTORIC CULTURAL RELATIONSHIPS
Data from the Carlo Creek Site indicate that initial human
utilization of the upper Nenana River Valley occurred not later than circa
8500 B.P. It is presumed that by this time, the valley would have been
sufficiently free of ice and outwash to permit human and animal movements
in a north-south direction through the Alaska Range. HEA 031 represents a ..
high mountain pass floodplain site, which was probably utilized briefly (one
to several days) by a small hunting group, perhaps by several males or
family sized unit. The faunal remains preserved at the site are consistent
of species extant in the valley today. Comparisons with ethnographic data
suggest aboriginal use of the area occurred in late summer to early winter
for hunting of sheep and caribou, with additional exploitation of ground
squirrels. The nature of the faunal remains, combined with ethnographic
data and the strategic mid-valley location of the site, suggest its use as
a kill site, where the major portions of the large mammal carcasses were
probably dismembered, then carried off to a regional 11 Central base 11 camp.
Campbell's (1968) model of Nuniamuit settlement types may offer a
useful modern day analogy for the probable pattern of subsistence and
settlement for the early Holocene. His type III settlements are (1968:15)
hunting or fishing camps. These are occupied most intensively during the
leanest parts of the year (in the case of the Nunamuit, February through
165
March and June through July) by one to five males "· ... for periods of two to
five days for the purpose of securing food." (1968:15). He Observed that:
Caribou and Oall sheep were the game most sought after, but some
Type III settlements were placed specifically for purposes of securing
other vertebrates ... [such as ground squirrels?] ... The importance of
caribou and Dall sheep explains why so many settlements of this type
were situated well up toward the heads of creeks or elsewhere high in
the mountains, where the sheep dwell the year around and where caribou
live in the summer.
Campbell suggests that a Type III settlement would have probably
contained a makeshift shelter, most likely consisting of" ... a few caribou
hides stretched over boulders or supported by willow sticks '' (1968:17).
He states (1968:9) that the location of such settlements was largely
dependent on available wood supplies (willow, in the case of the Nunamuit)··
and that they frequently were situated on " ... well-drained, gravelly
ground, usually on old stream bars 11 (1968:9).
What we can thus infer from the Carlo Creek data is a season-
specific, task-specific site, which is representative of just one portion
of an overall seasonal-settlement pattern (cf. Binford 1962). The limited
variety of artifacts from the site supports this interpretation. The Carlo
Creek data thus offer an excellent glimpse of one specific aspect of the
early Holocene subsistence settlement pattern in interior Alaska.
Campbell (1968:18) has eloquently stated the importance of such
factors as seasonality, ecologic setting, and limitations of site function:
... [If] we can assume that wandering hu~ters of the more distant
past were obliged, like the Tuluaqmiut, to adopt to their habitats
to secure the necessities of life, then we must recognize that the
archeological record is likely to be selective not only in respect
to the kinds of artifacts that survive, but also in respect to the
kinds of sites that can be recognized. Tuluaqmiut settlements of
Types II, III, IV and VI are fundamental components of the way of
life, and information derived from Types I and Valone gives a biased
picture. However, only the latter types normally possess sufficient
166
cultural debris to permit discovery by archeologists centuries or
millenia after their abandonment. This fact is not always fully
recognized by specialists in Paleo-Indian or Paleolithic cultures;
on the contrary, known sites are likely to be taken as representative
of settlement patterns of these early hunting groups.
To this statement, a more general comment by Binford (1978a:342) may
be added: " ... hunting camps are accumulation points for foods in an
extended logistical system. Animals introduced and culled at such
locations are being accumulated not directly in terms of the consumer needs
of the occupants but instead in terms of the consumer needs of a much
larger group of consumers in the residential locations". It is apparent
that prehistorians must lo ok more closely at the details of seasonality,
function, and setting in interpreting the portion of the total available
tool kit which may or may not be represented.
The objective of these observations has been to place the Carlo
Creek site within a larger framework of seasonality, function, and
environmental setting, in order to make some cultural inferences about the
site. To place the site in diachronic perspective, however, it will be
necessary to go beyond the site itself, and compare the artifactual data
with relevant data from other early Holocene age sites in central Alaska.
.-
First of all, the lithic artifacts represented at Carlo Creek are
primarily bifaces or biface fragments, with the single exception of a large
"prismatic blade" or blade-like flake. These large percussion-flaked
bifaces, manufactured from locally-available raw materials, represent what
I would term "tools of the moment". As such, they would have achieved a
functional level in the manufacturing process long before other implement
types would, such as finely prepared microblade cores or projectile points.
Many of the large, coarse-grained bifaces from interior Alaskan sites may
represent quarry blanks carried to a site for further reducti on as the need
167
arose . It is thus extremely difficult to make definitive comparison s until
it is known from what stage in the manufacturing process a set of spec i mens
is derived and what function they ultimately served. Even with this
knowledge of morphology, technology, and function, comparisons may have
little cultural-historical value.
With the above cautionary remarks in mind, the following
observations are made concerning the techno-cultural placement of HEA 031 .
These are not intended as an exhaustive review of the available data; such
a review would be of marginal utility in light of the largely undiagnostic
materials from Carlo Creek.
Artifacts from the Component II level at the Dry Creek Site (Holm&&
1974; Powers and Hamilton 1978), located 35 km downvalley from Carlo Creek,
suggest perhaps the closest similarities with those from HEA 031. Based on
similar morphologies and some edge-wear comparisons, affinities are
suggested between the biface industries at the two sites. Guthr.ie and
Powers (1977:11-13) report that the 1977 excavations at Dry Creek revealed
a number of spatially discrete biface reduction loci wit hin the
Component I I level at the site; these suggest areas analogous to Loci A and
B at Carlo Creek. It is significant to note that these biface "workshops"
appear to occupy spatially separate locations in the site, apart from the
core and blade activity loci (Powers. personal communication, 1978). One
might speculate about what the interpretation of that site might have been
if only the biface reduction areas . has been tested. Assuming
contemporaneity between these loci and the core and blade production areas,
it would appear that the biface loci represent just one portion of a much
more complex lithic technology. A number of specimens from the Dry Creek
168
collection show strong morphological similarity with Carlo Creek bifaces,
and particularly with the forms represented in Figure 18; B and I of this
report.
In addition to bifaces, the occurrence of large retouched prismatic
blades or blade-like flakes (e.g., Fig. 18; A) may suggest a toolkit
affinity between the two sites. My examination of the Dry Creek collection
indicates that there is a limited macroblade and core technology at Dry
Creek II (loess 3). These large blades and blade-like flakes
characteristically were struck from large blocky cores by direct
percussion. The cores have poly-directional platforms, with little
intentional platform preparation. There does not appear to have been any __
specific selection for blades per se, as both flakes and blades are
represented. Many of the large flakes and blade-like flakes from Dry
Creek II were further reduced as bifaces, or were at least edge modified in
a bifacial fashion.
In both sites, there was a tendency for coarse-grained rocks to be
selected as raw materials for both biface and macroblades. In the case of
the Carlo Creek I materials, hornfels and argillite were used, whereas at
Dry Creek, rhyolites, rhyodacites, and dacites were selected (Guthrie and
Powers 1977). In terms of texture, lustre, and overall "look", the latter
are difficult to distinguish from the argillite/hornfels from Carlo Creek.
These raw materials represent a logical choice with respect to their
durability and strong cutting edge; but to what extent they represent a
culturally-patterned choice from among available options we will never
know.
Probable use of manuports as anvil stones for bone breaking is a
trait apparently common to both assemblages (Guthrie and Pow~rs 1977).
169
However, no further significance can be attached to .a largely
time-transgressive trait such as this.
Finally, the temporal placement of the two occupations must be
considered. Dry Creek II is radiocarbon dated at 10,690 ± 250 years:
8740 B.C. (SI-1561) (Thorson and Hamilton 1977:166), and Carlo Creek I
is about 8500 years old. If these dates are correct, it would indicate that
Dry Creek II pre-dates Carlo Creek I by approximately 2190 years. On the
basis of the suggested dating for the interior Alaska Denali Complex, as
originally defined by West (1967, 1975), both of these occupations could be
culturally-related.
The toolkit represented at Dry Creek II appears to contain all the ..
defined categories of the Denali complex, as previously observed by Holmes
(1974), and Powers and Hamilton (1978). The Carlo Creek I materials,
conversely, would appear to meet these criteria only for(l) bifacial convex
knives (West 1967:365, 372) and(2) large 11 blades 11 and blade-like flakes
(West 1967:366, 372). However, these largely undiagnostic implements could
be just as easily included in later cultural traditions, as I will detail
below.
Given the activity specific nature and probably short duration of
occupation at HEA 031, the lack of other such 11 diagnostic 11 tools such as
microblades, cores, endscrapers, and burins, is not surprising. It seems
unlikely that a group of hunters would have carried their entire lithic
tool inventory with them, and even less likely that they would have left it
behind, at every site they utilized. The ethnoarcheological work of
Binford among present day hunters at Anaktuvuk Pass has suggested that
archeologists should not assume a patterned and static model of disposal
into the archeological record ( cf. Binford 1978a, 1978b). We must begin
170
J
to become more cognizant of the factors which affect the transformation of
material culture from its systemic context into the archeological record
(cf. Bowers and Bonnichsen 1980; Schiffer 1976).
The differences in material remains at the two sites clearly
reflect a different range of activities. While the Carlo Creek Site was
probably used solely for limited butchering, (using "tools of the moment"),
as inferred from a small assemblage of bone mashing and cutting tools, the
activities at Dry Creek II component additionally included use of bone
incising tools (burins), composite tool production (microblades, cores,
core tablets), and skin working (scrapers of various sorts). In short, the
Carlo Creek I occupation was limited in duration and activity, while Dry .-
Creek II was probably used much more extensively, perhaps over several
seasons, with a wider variety of activities represented. In a sense, we
might view Carlo Creek I as a type of satellite site of a Dry Creek-like
"regional base" camp.
Returning to the temporal placement of the Carlo Creek I
occupation, I should also briefly mention its possible affiliation with
later cultural manifestations such as the Northern Archaic tradition
(Anderson 1968b). At each of the three major non-coastal stratified
sites--Onion Portage, Dry Creek, and Healy Lake--there appears a major
occupational hiatus between about 8000 and 6000 B.P. (cf. Anderscrn 1968a;
Powers and Hamilton 1978; Cook unpublished data). At each site, this
apparent hiatus appears to mask a change from a "pa 1 eo" tradition (American
Paleo-Arctic: Denali, Akmak, Kobuk, Chindadn), to an apparent boreal
forest-adapted culture (Northern Archaic: Palisades Tuktu, Portage). We
171
have very 1 i ttl e data which can be brought to bear on the problem of
understanding this apparent hiatus and adaptive change. The tools represented
at Carlo Creek, especially coarse-grained bifaces, are well documented in
Northern Archaic assemblages (e.g., Anderson 1968a and b; Skarland and Keirn
1958 ), and thus could have been t he result of a Northern Archaic tradi t ion
occupation. Due to our lack of knowledge about this crucial time period, we
cannot state with certainty when these postulated forest adapted cultures
first appeared in Alaska; they may well predate the currently-accepted date
of about 6200 B.P. (Anderson 1968a and 1968b).
Going somewhat farther afield from Carlo Creek and the Nenana
Valley, limited similarities in artifact morphology, technology, and
apparent age may also be indicated with the Teklanika East and West sites
(West 1965, 1967). These two undated sites, located about 40-45 km north-
west of Carlo Creek, were included by West (1967 ) in his original definition
of the late Pleistocene/early Holocene interior Alaskan Denali Complex.
The sites, both which lie at an elevation of about 790 m a.s.l, were
originally interpreted ~(West 1965) as game lookout and flaking stations
(see Fig. 7).
A number of other widely scattered surface sites within McKinley
Park reported by Morgan (1965) and Treganza (1964 ) may also indicate
possible affinity with the occupation at Carlo Creek I. However, these are
all small, undated, and all lack faunal remains . These sites differ from
the Carlo Creek site in that they represent chipping and game l ookout /
intercept stations, located on relatively high r i dges of elevations
between 760 m and 915 m a.s.l.
Several more recent surveys along the north flank of the Alaska
Range and within the Nenana Valley have located sites which may have
affinities in site function or artifact technology with Carlo Creek
172
J
I
j
(Holmes 1975; Guthrie and Powers 1977; Plaskett 1976 ). Again, however, all
of these sites are poorly preserved, undated, and most were surfacial in
nature.
Attempting to make any but the most general statements concerning
artifact or site affinities between Carlo Creek and these undated and
largely "undiagnostic" assemblages is of little utility. The specific
tasks represented by the Carlo Creek data, together with the relative
"crudeness" of tool fonn, makes any detailed trait-by-trait interpretive
statements difficult. Suffice it to say at this point that there has
undoubtedly been intensive utilization of the upper Nenana Valley by semi-
nomadic hunting groups ever since the valley was free of ice and extensive
outwash channels at the end of the Pleistocene. Based largely on the age·
of the Carlo Creek site, it is speculated that the site may be part of an
early Holocene phase of the Denali Complex toolkit, as originally defined
(West 1967).
173
CHAPTER 10
SUMMARY AND CONCLUSIONS
Geological, ecological, radiometric and archeological data from the
surrounding area have enabled a fairly clear picture to be drawn of the
Carlo Creek Site 1 s function and setting. There are, obviously, a myriad of
~nsolved questions concerning occupational group size, techno-cultural
affiliations, etc. In general, it is felt that the task-specific nature of
this site effectively limited the diversity of materials left behind by the
site 1 s occupants.
In summary, the site 1 S geological and archeological history may be
stated as follows:
1. The site 1 S geomorphology appears to be the result of a fluvial
sequence attributable to downcutting and lateral shift of the glacial
Nenana River. The earliest evidence of human use of the site occurred as
soon as 1500-2000 years after deglaciation of the valley, and may have
occurred at a time while minor amounts of stagnant ice were still wasting
in the upper valley. At the time of Component I occupation (circa 8500
B.P.), a temporary cessation of downcutting had taken place; by this time,
at least 75% of all Holocene downcutting had already occurred.
2. Stabilization of this surface was probably a short lived
phenomenon lasting perhaps no longer than 500 years. Limited soil
development (inceptisols) from this surface did begin, but was halted by
later overbank alluviation. Component I utilization of the site took place
174
directly on the surface of a stabilized braid bar on the eastern banks of
the Nenana River; this early occupation of the site is radiocarbon dated to
approximately 8500 B.P.
3. Based on faunal data, a late-summer/fall seasonal use is
inferred for Component I. Functionally, the site appears to represent a
secondary kill site and brief (perhaps overnight?) encampment, where one
sheep and one caribou were brought into the site and dismembered for
transport to a larger regional 11 base camp 11
--perhaps a site such as Dry
Creek. The fact that only low-value bone remains are present supports a
notion of 11 Culling 11 of the carcass portions with greater dietary
importance. The fragmentary remains of the bone and the presence of a
possible anvil stone suggests that it was smashed and processed for marrow
and/or bone grease. On the basis of comparisons with recent Nuniamuit
Eskimo hunting practices, the Carlo Creek Site could be interpreted as a
season and task-specific upland hunting site, located within modern ranges
of both sheep and caribou. Additional locally-procured resources included
ground squirrels, argillite for lithic raw materials, and probably, alder
or willow (for fire and shelter?). The site location was probably
determined, in part, by its strategic intercept position in this valley,
and possibly by the presence of a freshwater spring which flows near the
site today. Ethnographic data for the area support these interpretations
of seasonality and site function.
4. The activities in the lower component appear to be spatially
separated into at least three activity areas, centered around two hearths.
Activities indicated include biface production, possible thermal heat
treatment(?) of lithics, carcass dismemberment, and bone smashing for
marrow/bone grease extraction.
175
5. Lithic technology at the Carlo Creek I occupation was almost
exclusively devoted to biface reduction. Data suggest that lithic raw
materials were locally procured from a quarry source within the valley, and
a lesser proportion of rock was obtained from the stream bed of the Nenana
River. There is fairly good evidence that at least part of the raw
argillite brought into the site was intentionally heat-treated in a crude
sand "furnace", presumably to improve the workability of the rock for
flaking and edge durability. In addition to argillite, a single large
"prismatic blade" or blade-like flake of hornfels is present at the site.
The lack of chipping debris associated with this specimen suggests it was
manufactured elsewhere, then brought into the site and discarded following:
breakage.
6. Due to a small and largely undiagnostic lithic sample, it has
not been possible to make definitive technocultural comparisons between
Carlo Creek and other known late Pleistocene-early Holocene sites in
Alaska-Yukon. On the basis of temporal similarities, general tool
morphology, inferred site function, and raw material selection,
similarities are noted primarily with the biface-reduction workshop loci in
Component II of the nearby Dry Creek site. On a more speculative level,
the Carlo Creek Component I level can be assigned to an undesignated early
Holocene phase that may or may not be analogous to the Denali Complex, as
originally defined by West (1967, 1975). However, any such provisional
assignment must take into account the site 1 S inferred function and
restricted tool inventory.
7. The Carlo Creek Site, Yardang Flint Station (Reger et al. 1964),
and the Canyon Site (Workman 1974, 1978), all represent relatively rare
occurrences in the subarctic: firmly dated sites within clear
176
stratigraphic contexts. Unfortunately, all three of these sites appear to
contain relatively few artifactual remains. Perhaps this fact in and of
itself has much to tell us about the nature of the relatively meager
subarctic archeological record, or perhaps it is entirely a site sampling
problem. It is increasingly evident that human subsistence in the
subarctic was possible through the use of rudimentary material culture
inventories. One major drawback to this interpretation is, however, our
lack of knowledge concerning the organic portion of the toolkit of
aboriginal occupants in Interior Alaska. This question, unanswered during
the Carlo Creek research program, will be a necessary next step in the
understanding of Alaska's prehistory.
8. Environment at the time of Component I occupation is inferred as
being not too different from that of today, although the extent of spruce
forest colonization into the alpine areas is not known. By 8500 B.P. a
number of distinct early successional plant associations were probably
established in the Nenana Valley's floodplains, terraces, and valley
slopes. These most likely reflect a pattern of succession similar to those
found today in subarctic alpine regions. Probably included in these
communities were grasses, alder(?), willow, open spruce/shrubby tundra
and closed spruce(?) forest. The site utilization most likely occurred
within an early successional phase of grasses, alders(?) and/or willows.
9. Following abandonment of the site after the brief Component I
occupation at 8500 B.P., the locality was subjected to repeated episodes of
overbank flooding, which resulted in a net accumulation of 3-4m of
floodplain laminated silts and fine sands. Within the lower 50 em of this
overbank unit are two species of terrestrial mollusca, which offer further
evidence for locally-moist conditions and possibly an open spruce
177
ecosystem. Throughout this overbank unit are the poorly preserved remains
of detrital organics, which were probably drifted in during high water
stages. The overbank unit (3) shows alternate periods of wetting and
drying, as evidenced by oxidation/reduction sequences.
10. After approximately one meter of overbank sediments had been
deposited, a second extremely brief human occupation occurred. This later
use of the site appears to represent the activity of only a single (?)
individual, and reflects only stone working activities. The lithic raw
material used at the Component II occupation was different from that of the
lower occupation (rhyolite vs. argillite) and appears to reflect different
stoneworking technique (pressure vs. percussion). It is not known whether;
or not the Component II rhyolite was heat treated. The presence of a
spring, as well as strategic mid-valley location, may have again influenced
selection of this site as a chipping/lookout station. On the basis of
inferred rates of sediment deposition, and bracketing C-14 dates, this
occupation is estimated to have occurred sometime between 6000 and 7500
B.P.
178
11. Cessation of overbank deposition at the site was probably the I
result of a gradual resumption in the Nenana River downcutting and/or f
lateral shift to the west, away from the site area. Floodplain
sedimentation was replaced at approximately 3000 to 4000 B.P. by eolian
deposition. The cause of this depositional change is unknown, although it
is thought to relate to neoglaciation in the central Alaska Range.
12. These eolian deposits, which cap the site and mantle the
hillside, are thought to be of local origin, probably derived from locally
available floodplain alluvium. Contained within this uppermost unit, at an
average depth of about 20 em below surface, is a discontinuous lens of
volcanic ash, which has been correlated by comparison of refractive index,
shard morphologies, and phenocrysts with at least four other tephra
horizons in the upper Nenana Valley. This ash has been assigned a maximum
C-14 age of 3780 ± 80 years: 1830 B.C. (WSU-1747), on the basis of a
dated wood sample collected from 5 em below the ash at a locality approxi-
mately 8 km up valley from the Carlo Creek site. It has not been possible
to correlate this ash bed with others reported in central Alaska, although
its vent source may be one of three sources in the southwest Alaska Range
dated to about 3600-3800 B.P. (cf. Bowers 1979a).
13. The geologic setting of the Carlo Creek site may provide
.· northern Quaternary specialists with a useful model for predicting site
locations and directing future 11 early man 11 surveys. Although the maximum
possible age for the Carlo Creek site is itself limited by former glacial
activity in the Nenena Valley, the site's stratigraphy and strategic mid-
valley location may be applicable to a variety of models of late
Pleistocene or early Holocene settlement patterns and site locations. The
fact that well defined terraces occur on both sides of the Nenana River in
this part of the valley means that one could closely formulate a future
sampling strategy directed towards the location of a functionally-specific
or temporally-specific type of site. Also, the fact that we are in reality
sampling space, rather than area demands that we concentrate our limited
funding and manpower resources in areas that combine the geologic
conditions for site preservation (e.g., alluvial sediments) with a means of
observing the third dimension (e.g.,roadcut, exposed riverbank, etc.).
14. In sum, the Carlo Creek Site represents an early Holocene
floodplain occupation (Component I) within a high subarctic alpine valley.
Activities at the site appear to reflect a number of special i zed activities
179
dictated by seasonality, available resources, and function. The site is
noted particularly for its good organic preservation and possible
indications of thermal pre-treatment of lithics. It served first as a
secondary kill/butchering site, with marrow extraction, lithic
pre-treatment and biface reduction, and later (Component II) functioned
briefly as a lookout/chipping site.
180
BIBLIOGRAPHY
Ackerman, R. E., T. D. Hamilton, and R. Stuckenrath
1979 Early culture complexes on the northern Northwest Coast. Canadian
Journal of Archaeology 3:195-209.
Adovasio, J. M., J. D. Gunn, J. Donahue, and R. Stuckenrath
1978 Meadowcraft Rockshelter, 1977: An Overview. American Antiquity
43:632-651.
Ager, T. A.
1975 Late Quaternary environmental history of the Tanana Valley,
Alaska. Ohio State University Institute of Polar Studies Report 54,
Columbus, Ohio.
Ahtna, Inc.
1973 The Ahtna Region. Copper Center, Alaska.
Aigner, J. S.
1976 Dating the early Holocene maritime village of Anangula. Anthro-
pology Papers of the University of Alaska 18(1):51-62.
Alaska Department of Fish and Game
1973 Alaska 1 S Wildlife and Habitat. Juneau.
Alexander, H. L.
1974 The association of Aurignacoid elements with fluted point complexes
in North America. In International Conference on the Prehistory and
Paleoecology of Western North American Arctic and Subarctic, edited . -
by R. Scott and P. Schledermann, pp. 21-32. Archaeological Association,
University of Calgary.
181
Allen, H. T.
1887 Report of an expedition to the Copper, Tanana, and Koyukuk Rivers,
in the Territory of Alaska, in the year 1885. U.S. Army, U. S.
Government Printing Office, Washington, D. C.
Allen, J. R. L.
1965 A review of the origin and characteristics of recent alluvial
sediments. Sedimentology 5 :89-191.
1970 Physical Processes of Sedimentation, An Introduction. G. Allen
and Unwin, London.
Anderson, D. D.
1968a A stone age campsite at the gateway to America. Scientific
American 218(6) :24-33.
1968b Early notched point and related assemblages in the Western
American Arctic. Paper prepared for the 67th Annual Meeting of the
American Anthropological Association, Seattle.
1970 Microblade traditions in northwestern Alaska. Arctic Anthro-
pology 7 (2):2-16.
Andre•.vs, E. F.
1977 Report on the cultural resources of the Doyon Region, Central
Alaska. Occasional Paper 5, Cooperative Park Studies Unit, Uni versity
of Alaska, Fairbanks.
Bacon, G.
1977 The prehistory of Alaska: A speculative alternative. In
Problems in the Prehistory of the North American Subarctic: The
Athapaskan Question, edi terl by J. 1tJ. Helmer, S. Van Dyke and F. J.
Kense, pp. 1-10. Archaelogical Association, University of Calgary.
182
Bee, J. W. and E. R. Hall
1956 Mammals of Northern Alaska. University of Kansas Museum of
Natural History, Miscellaneous Publication 8.
Binford, C. R.
1962 Archaeology as anthropology. American Antiquity 28(2):217-225.
1978a Nunamiut Ethnoarchaeology. Academic Press.
1978b Dimensional analysis of behavior and site structure: Learning
from an Eskimo hunting stand. American Antiquity 43(3):330-361.
Binford, L. R. and G. I. Quimby
1972 Indian sites and chipped stone materials in the northern Lake
Michigan area. In An Archaeological Perspective, edited by L. Binford,
pp. 346-372. Seminar Press, New York.
Bonnichsen, R.
1973 Some operational aspects of human and animal bone alteration.
In Mammalian Osteo-Archaeology: North America, edited by B. M. Gilbert.
Special Publication of the Missouri Archaeological Society, Columbia,
Missouri.
1977 Models for deriving cultural information from stone tools.
Archaeological Survey of Canada, Mercury Series 60, Ottawa.
1978 Critical arguments for Pleistocene artifacts from the Old Crow
Basin, Yukon: A preliminary statement. In Early Man in America
from a Circum-Pacific Perspective, edited by A. L. Bryan, pp. 102-108.
Occasional Papers No. 1 of the Department of Anthropology, University
of Alberta.
r
1979 Pleistocene bone technology in the Beringian Refugium. Archae-
ological Survey of Canada, Mercury Series 89. Ottawa.
183
Bonnichsen, R. and D. Sanger
1977 Integrating faunal analysis. Canadian Journal of Archaeology
1 : 1 09-135.
Bowers, P. M.
1977a Preliminary Investigations of the Carlo Creek Site, Upper Nenana
Valley, Central Alaska. Paper presented at the Fourth Annual Meeting
of the Alaska Anthropology Conference, Fairbanks. April 8-9, 1977.
1977b Field report: Archeological survey of a proposed electric
distribution line extension to Mount McKinley National Park, Alaska.
Report submitted to the National Park Service, Seattle.
1978a Research summary: 1977 investigations of the Carlo Creek Arche-.·
ological Site, Central Alaska. Report submitted to the University of
Alaska Museum, Fairbanks.
1978b Geology and archaeology of the Carlo Creek Site, an early Holocene
Campsite in the central Alaska Range. Abstracts of the Fifth
Biennial Conference, American Quaternary Association, p. 188. Edmonton,
Alberta, September 2-4, 1978.
1979a The Cantwell Ash Bed, a Holocene Tephra in the central Alaska
Range. Short Notes on Alaskan Geology--1978. pp. 19-24. Geologic
Report No. 61, Alaska Division of Geologic and Geophysical Surveys.
College, Alaska.
1979b Final report: Archeological survey of a proposed electrical
distribution line extension to Mount McKi~ley National Park, Alaska.
Report submitted to the National Park Service under contract CX-9000-
7-0027. University of Alaska, Fairbanks.
184
Bowers, P. M. and R. Bonnichsen
1980 Displacement of surface lithic artifacts due to natural phenomena:
An experimental study from selected subarctic and arctic localities.
Paper presented at the 7th Annual Meeting of the Alaska Anthropological
Association, Anchorage.
Bowers, P. M. and D. Hoch.
1978 An archeological reconnaissance of the Copper Creek Drainage,
Upper Charley River area, East-Central Alaska. Anthropology and
Historic Preservation, Cooperative Park Studies Unit, Occasional Paper
11, pp. 1-63. University of Alaska, Fairbanks.
Brauner, D. R.
1968 Space, time, and debitage: A study of lithic debris, Chagvon
Bay Alaska. Unpublished M.A. Research Paper, Washington State
University.
Brink, J. \~.
1978 An experimental study of microwear formation on endscrapers.
Archeological Survey of Canada Mercury Series 33. Ottawa.
Brose, D. S.
1975 Functional analysis of stone tools: A cautionary note on the
role of animal fats. American Antiquity 40:86-93.
Burch, E. S., Jr.
1972 The caribou/wild reindeer as a human resource. American Antiquity
37:339-368.
185
Ca mpbe11, C.A ., E. A. Pau l , O.A. Rennie, and K.J. t'v1cCa11um
1967 Factors affecting the accuracy of the carbon dating method in soil
humus studies. Soil Science 104(2):81-85.
Campbell, J. M.
1962 Cultural succession at Anaktuvuk Pass, Arctic Alaska. In
Prehistoric Cultural Relations Between the Arctic and Temperate
Zones of North America, edited by J.M. Campbell, pp. 39-54. Arctic
Institute of North America, Technical Paper No. 11.
1968 Territoriality among ancient hunters: Interpretations from
186
ethnography and nature. In Anthropological Archaeoloqy in the Americas,
edited by B. Meggers, pp. 1-21. The Anthropological Society of 1.-Jash}f'lgton.
Carl , E. A.
1962 Ecology of the arctic ground squirrel, Citellus parryi. In
Terrestrial Mammals Investigation, Ogbturuk Creek, Cape Thompson and
Vicinity. Part B, Project Chariot, Final Report, U.S. Atomic Energy
Commission, pp. 47-91. A. E. C., !tJashington.
1971 Population control of arctic ground squirrels. Ecology 52(3):395-
413.
Chang, K.
1962 A typology of settlement and community patterns in some circumpolar
societies. Arctic Anthropology 1(1) :28-42.
Chaplin, R. E.
1971 The Study of Animal Sones from Archaeological Sites. New York,
Seminar Press.
~
I
)
I
I
Cinq-1•1ars, J.
1979 Bluefish Cave I: A late Pleistocene eastern Beringian Cave
deposit in the Northern Yukon. Canadian J ournal of Archaeology
(3):1-32.
Clark, D. W.
1975 Prehistory of the Western subarctic. Canadian Archaeological
Bulletin (7):76-95.
Cook, J. P.
1969 The Early Prehistory of Healy Lake, Alaska. Unpublished
Ph.D. Dissertation, University of Wisconsin, Madison.
1975 Archeology of interior Alaska. The Western Canadian Journal
of Anthropology 3-4:125-133.
1977 Pipeline Archeology. University of Alaska, Institute of
Arctic Biology.
Crabtree, D. E.
1966 Notes on experiments in flintknapping: 2, A stone workers
approach to analyzing and replicating the Lindenmeir Folsom.
Tebiwa 9(1):3-39.
1972 An introduction to flintworking. Occasional Papers of the
Idaho State University Museum 28. Pocatello, Idaho.
1976. The potential of lithic technology. In Primitive Art and
Technology, edited by J. S. Raymond, B. Lovesetn, C. Arnold,
and G. Reardon, pp. 1-6. Archeological Association, University
of Calgary.
Crabtree, D. E. and B. R. Butler
1964 Notes on experiments in flintknapping: 1, heat tr~atment
of silica minerals. Tebiwa 7(2):1-6.
187
Da 11 , W. H.
1905 Land and fresh water molluscs of Alaska and adjoining regions.
Harriman Alaska Expedition 13:1-171.
Daly, P.
1969 Approaches to faunal analysis in archaeology. American
Antiquity 34:146-153.
Davies, R. I.
1971 Relation of polyphenols to decomposition of organic matter
to pedogenic processes. Soil Science 111:80.
Davis, S.
1979 Hidden Falls: A Stratified site in Southeast Alaska. Paper
presented at the 6th Annual Alaskan Anthropology Conference,
Fairbanks.
1980 Hidden Falls: A multicomponent site in the Alexander
Archipelago of the Northwest Coast. Paper presented at the 7th
Annual Alaskan Anthropology Conference, Anchorage.
Denton, G. H., and '..J. Karlen
1977 Holocene glacial and tree line variations in the White River
Valley and Skolai Pass, Alaska and Yukon Territory. Quaternary
Research 7:63-111.
Oettennan, R. L.
1964 Early Holocene warm interval in northern Alaska. Arctic
23: 130-131 .
188
I
l
Didier, M. E.
1975 The argillite problem revisited: An archaeological and geological
approach to a classical archaeological problem. Archaeology of
Eastern North America 3:90-100.
Dumond, E. E.
1969 Toward a prehistory of the Na-Dene, with a general co!11Tlent on
population movements among nomadic hunters. American Anthropologist
75:857-863.
Dumond, E. E., W. Henn, and R. Stuckenrath
1976 Archeology and Prehistory on the Alaska Peninsula. Anthropological
Papers of the University of Alaska 18(1):17-29.
Faulkner, A.
1977 Radial striations: Clues to interpreting wear on stone tools.
Paper presented at the Conference on Lithic Use-Wear, Vancouver,
B.C., March 1977.
Flenniken, J. J.
1975 Test excavations of three archeological sites in Desarc Canyon
watershed, White County, Arkansas. Report on file with the Arkansas
Archeological Survey, Fayetteville.
1977 Analysis of lithic tools, 1976 sample. In Preliminary Archae-
ological Investigations at the Miller Site, Strawberry Island 1976:
A Late Prehistoric Village Near Burbank, Franklin County, Washington,
pp. 69-101. Project Report No. 46, Washington Archaeology Research
Center, Pullman.
1978 Reevaluation of the Lindenmeir Folsom: A replication experiment
in lithic technology. American Antiquity 43:473-479.
189
Flenniken, J. J.
n.d. Replicative systems analysis of the lithic artifacts from the
Hoko River site. Unpublished manuscript, Department of Anthropology,
Washington State University.
Flenniken, J. J. and E. Garrison
1975 Thermally altered novaculite and stone tool manufacturing
techniques. Journal of Field Archaeology 2:125-131.
Folk, R. L.
1966 A review of grain size parameters. Sedimentology 6:73-93.
' Folk, R. L. and W. Ward
1957 Brazos River bar: A study in the significance of grain size
parameters. Sedimentary Petrology 27:3-36.
Frison, G. C.
1968 A functional analysis of certain chipped stone tools. American
Antiquity 33:149-155.
Gey, M. A., J. H. Banzler, and G. Roeschamann
1971 Problems of dating Holocene soils by radiometric methods. In
Paleopedology: Origin, Nature and Dating of Paleosols, edited by
D. H. Yaalon, pp. 63-75. University Press, Jerusalem, Israel.
Gilbert, B. M.
1973 Mammalian Osteo-archaeology: North America. Missouri
Archaeological Society, Columbia, Missouri.
Gilbert, W. G.
.·
1976 Evidence for early Cenozoic orogeny in the Central Alaska Range.
Short Notes on Alaskan Geology--1976. Geologic Report 51, Alaska
Division of Geological and Geophysical Surveys, College, Alaska.
190
l
Glass, 6. P.
1951 A Key to the Skulls of North American r·1arrmals. Burgess Publi-
cations, Minneapolis.
Goh, K. M and 6. P. J. Molloy
1972 Reliability of radiocarbon dates from buried charcoals.
Proceedings of Eiahth International Conference on Radiocarbon
Dating, p. 565. Lower Hutt, New Zealand, October 1972.
Goh, K. t·1., 6. P. Molloy, and T. A. Rafter
1977 Radiocarbon dating Quaternary loess deposits, Banks Peninsula,
Canterbury, New Zealand. Quaternary Research 7(2):177-196.
Grayson, 0. K.
1973 On the methodology of faunal analysis. American Antiquity
38:432-439.
Gubser, N. J.
1965 The ~unamiut Eskimos: Hunters of Caribou. Yale University
Press, New Haven.
Gummerman, M.
1975 Mechanical models for the analysis of lithic assemblages:
Preconditions necessary for use. Newsletter of Lithic Technology
1-2:7-10.
Gustafson, C.
--
n.d. Unpublished faunal key for deer, sheep, and pronghorn. Washington
State University, Department of Anthropology. Mimeographed.
Gu t h r i e , R . 0 .
1968 Paleoecology of a late Pleistocene small mammal community from
interior Alaska. Arctic 21:223-244.
191
Guthrie, R. 0., and W. R. Powers
1977 Early man studies, 1977: A Preliminary report. A report to the
National Park Service and National Geographic Society. pp. 1-32.
Hall, E. R., and K. R. Nelson
1959 Mammals of North America. Ronald Press, New York (2 volumes).
Ha 11 , E. S. , Jr.
1971 Kangiguksuk: A cultural reconstruction of a sixteenth century
Eskimo site in northern Alaska. Arctic Anthropology 8(1):1-101.
Hamilton, T. D.
1973 Late Quaternary glacial history, Delta-Johnson Rivers Region,
Northeastern Alaska Range. Manuscript on file, Geology Department,
University of Alaska.
Hamilton, T. D., and S. C. Porter
1975 Itkillik glaciation in the Brooks Range, Alaska. Quaternary
Research 5:471-487.
Hamilton, T. 0., and S. Robinson
1977 Late Holocene (neoglacial) environmental changes in central
Alaska. Abstracts with Programs, Geological Society of America
9:1003.
Harner, M. J.
1956 Thermo-facts vs. artifacts: An experimental study of the
Malpais Industry. University of California Archeological Survey
Reports 33:39-43.
192
j
(
I
1
I
f
f
r
I
f
f
Hassan, F. A.
1976 The study of lithic artifacts: An analytical model and two
case studies. In Primitive Art and Technology, edited by J. S.
Raymond, B. Loveseth, C. Arnold and G. Reardon, pp. 27-46.
Archeological Association, University of Calgary.
n.d. Unpublished laboratory manual. Geoarcheology Laboratory,
Washington State University, Pullman.
Hemming, J. E.
1971 The distribution and movement patterns of caribou in Alaska.
Alaska Department of Fish and Game Technical Bulletin 1.
Henderson, L.
1931 Molluscan provinces in the western United States. University
of Colorado Studies 18:177-186.
Henn, W.
1978 Archaeology on the Alaska Peninsula: The Ugashik Drainage,
1973-1975. University of Oregon Anthropological Papers No. 14,
Eugene.
Hoffecker, J. F.
1979 The Search for early man in Alaska: Results and
recorrrnendations of the North Alaska Range project. Report
to the National Geographic Society and the National Park Service.
Holmes, C. E.
1974 ~ew Evidence for a Late Pleistocene Culture in Central Alaska:
Preliminary Investigations at Dry Creek. Paper presented at the
Seventh Annual Meeting of the Canadian Archaeological Association,
~~hitehorse. r~arch, 1974.
193
Holmes, C. E.
1975 Archeological survey in the Nenana Valley, 1975. Report
submitted to the Alaska Division of Parks in fulfillment of
contract PL 89-665 Matching Grant. Unpublished manuscript,
Anchorage.
Holmes, W. J.
1919 Handbook of aboriginal American antiquities: Part I. Bureau
of American Ethnology Bulletin 60. Washington.
Hopkins, D. M.
1975 Time-stratigraphic nomenclature for the Holocene epoch. Geology
3:10.
Ingstad, H.
1954 Nunamiut: Among Alaska's Inland Eskimos. Norton and Company,
New York.
Irving, W. N. and C. R. Harington
1973 Upper Pleistocene radiocarbon-dated artifacts from the northern
Yukon. Science 179:335-340.
Karlstrom, T. N. V.
1965 Upper Cook Inlet area and Matanuska River Valley. In Guidebook
for Field Conference F, Central and South Central Alaska, edited by
T. L. P~w~, 0. J. Ferrians, D. R. Nichols, and T. N. V. Karlstrom.
International Association for Quaternary Research, Seventh Congress,
U.S.A., 1965. Nebraska Academy of Sciences.
Keller, C. M.
1966 The development of edge damage patterns on stone tools. Man.
1(4):501-511
1~
J
l
f
I .
Krauss, M. E.
1973 Na-Oene. Current Trends in Linguistics 10(1):903-978.
1974 Native peoples and languages of Alaska (map). Alaska Native
Language Center, Fairbanks.
Kukal, Z.
1971 The application of knowledge of recent sediments to ancient
sediments. Geology of Recent Sediments, ch. 23. Prague.
Larsen, H.
1968 Trail Creek: Final report on the excavations of two caves on
Seward Peninsula, Alaska. Acta Arctica 15:7-79.
Laughlin, W. S.
1975 Aleuts: Ecosystem, Holocene history, and Siberian origin.
Science 189:507-515.
Lawrence, B.
1951 Post-cranial skeletal characters of deer, pronghorn, and
sheep-goat with notes on Bas and bison. Report No. 4 of the
Awatovi Expedition Papers of the Peabody Museum, Harvard
Univers ity XXXV(3):1-43.
Leopold, L. B., M. G. Wolman, and J. P. Miller
1964 Fluvial Process in Geomorohology. Freeman Press, San Francisco
and London.
Long, A. and B. Rippeteau
1974 Testing contemporaneity and averaging radiocarbon dates.
A~erican Antiquity 39:205-215.
Lyman, R. L.
1979 Available meat from faunal remains: A consideration of
techniques. American Antiquity 44:536-545.
195
·.
MacDonald, S. 0.
1978 Keys to the voles, mice, and rats of Alaska. University
of Alaska Museum, Fairbanks.
Mandeville, M. 0.
1971 The Baked and the Half-baked: A Consideration of the
Thermal Pretreatment of Chert. Unpublished M.A. Thesis, University
of Missouri.
Mandeville, M. D. and J. J. Flenniken
1974 A comparison of the flaking qualities of Nehawka chert before
and after thermal pretreatment. Plains Anthropologist 19:146-148.
Matthews, J. V., Jr.
1970 Quaternary environmental history of interior Alaska: Pollen
samples from organic colluvium and peats. Arctic and Alpine
Research 2(4):241-251.
1974 Wisconsin environment of interior Alaska: Pollen and macrofossil
analysis of a 27 meter core from the Isabella Basin (Fairbanks
Alaska). Canadian Journal of Earth Sciences 11:828-841.
Mauger, J. E.
1970 A Study of Donnelly Burins in the Campus Archeological Collection.
Unpublished M.A. Thesis, Washington State University.
McCulloch, E. S., D. W. Taylor, and M. Rubin
1965 Stratigraphy, non-marine molluscs, and radiometric dates from
Quaternary deposits in the Kotzebue Sound area, western Alaska.
Journal of Geology 73:242-253.
McKennan, R. A.
1959 The Upper Tanana Indians. Yale University Publications in
Anthropology 55. New Haven.
196
I
I
I
I
I
I
McKennan, R. A.
1965 The Chandalar Kutchin. Arctic Institute of North America
Technical Paper 17.
1969 Athapaskan groupings in central Alaska. In Contributions to
Anthropology: Band Societies, edited by D. Damas. National Museums
of Canada Bulletin 228, Anthropological Series 24, pp. 93-115. Ottawa.
Melchoir, H. R.
1964 Ecological relationships of arctic ground squirrels to burrow
site features. In Proceedings of the 15th Alaska Science Conference,
pp. 41-42. College, Alaska.
Michael, H. N. (ed.)
1967 Lieutenant Zagoskin's travels in Russian America, 1842-1844:
The first ethnographic and geographic investigations in the Yukon
and Kuskokwim Valleys for Alaska. In Anthropology of the North:
Translations from Russian Sources No. 7. Arctic Institute of
North America, University of Toronto Press.
Moffit, F. H.
1915 The Broad Pass Region, Alaska. U.S. Geological Survey
Bulletin 608.
Morgan, H. M.
1965 An archeological Survey of Mt. McKinley National Park. Report
to the National Park Service, M.S.
Morlan, R. E.
1973 The later prehistory of the Middle Porcupine Drainage, northern
Yukon Territory. Archaeological Survey of Canada Paper 11.
Mercury Series. National Museum of Man., Ottawa.
197
Mar 1 an, R. E.
1977 Fluted point makers and the extinction of the arctic-steppe
biome in eastern Beringia. Canadian Journal of Archaeology 1:95-109.
Murie, A.
1944 The wolves of Mt. Mckinley. U.S. National Park Service, Fauna
of the National Parks, Fauna Series 5:1:238.
~1uto, G. R.
1971 A Technological Analysis of the Early Stages in the Manufacture
of Lithic Artifacts. Unpublished M.A. Thesis, Department of
Anthropology, Idaho State University.
Nance, J. D.
1971 Functional interpretations from microscopic analysis. American
Antiquity 36:361-366.
Ne 1 son, R. K.
1973 Hunters of the Northern Forest: Designs for Survival Among
the Alaskan Kutchin. University of Chicago Press, Chicago.
Orth, D. J .
1967 Dictionary of Alaska Place Names. U.S. Geological Survey
Professional Paper 567. U.S. Government Printing Office, Washington.
Osgood, C. B.
1936a The distribution of the northern Athapaskan Indians. Yale
University Publications in Anthropology 7. New Haven.
1936b Contributions to the ethnography of the Kutchin. Yale
University Publications in Anthropology 14. New Haven.
1937 The ethnography of the Tanaina. Yale University Publications
in Anthropology 16. New Haven.
198
Osgood, C. B.
1970 Ingalik material culture. Yale University Publications in
Anthropology 22. New Haven.
/ / Pewe, T. L.
1975a Quaternary stratigraphic nomenclature in central Alaska.
U.S. Geological Survey Professional Paper 862. Washington.
1975b Quaternary geology of Alaska. U.S. Geological Survey Professional
Paper 835. Washington.
"' / Pewe, T. L., 0. J. Ferrians, D. R. Nichols, and T. N. V. Karl strom
1965 Guidebook for field conference F--Central and south central
I
Alaska. International Association of Quaternary Research Seventh ·.
Congress U.S. A. 1965. Nebraska Academy of Sciences, Lincoln.
Pilsbry, H. A.
1939-Land Molluscs of North America (North of Mexico). Monograph
48 of the Academy of Natural Sciences, Philadelphia.
Plaskett, D. C.
1976 A cultural resource survey in an area of the Nenana and
Teklanika Rivers of central Alaska. Ms. on file with the Alaska
Division of Parks, Anchorage.
1977 The Nenana River Gorge Site: A Late Prehistoric Athapaskan
Campsite in Central Alaska. Unpublished M.A. Thesis, University
of Alaska.
Porter, L.
1978 Evidence for Late Pleistocene Human and Animal Life in the
Alaskan Yukon. Paper presented at the 31st Annual Northwest
Anthropological Conference, Pullman, Washington.
199
Powers, W. R. and T. 0. Hamilton
1978 Dry Creek: A late Pleistocene human occupation in central
Alaska. In Early Man in America From a Circum-Pacific Perspective,
edited by A. L. Bryan, pp. 72-77. Occasional Papers ~o. 1 of the
Department of Anthropology, University of Alberta.
Purdy, B. A.
1971 Thermal alteration of silica minerals: An archeological
approach. Science 173:322-325.
Rainey, F. G.
1939 Archaeology in central Alaska. Anthropological Papers of
the American 1·1useum of Natural History 36:351-405.
Ramp ton, V.
1971 Late Quaternary vegetational and climatic history of the Snag-
Klutlan area, southwestern Yukon Territory, Canada. Geological
Society of America Bulletin 82:959-978.
Reed, C. A.
1963 Osteo-archaeology. In Science in Archaeolooy, edited by
Brothwell and Higgs, pp. 204-216. Basic Books, Inc., New York.
Reger, R. D., T. L. Pewe, F. H. West, and I. Skarland.
1964 Geology and archeology of the Yardang Flint Station. Anthro-
pological Papers of the University of Alaska 12(2):92-100.
Reineck, H. E. and I. B. Singh
1975 Depositional Sedimentary Environments. Springer-Verlag,
Berlin, Heidelberg, New York.
200
.·
Repenning, C. A., D. M. Hopkins, and M. Rubin
1964 Tundra rodents in a late Pleistocene fauna from the Tofty
Placer District, central Alaska. Arctic 17(3):176-197.
Sadek-Koors, H.
1972 Primitive bone fracturing: A method of research. American
Antiquity 37:369-382.
Schiffer, M. B.
1976 Behavioral Archeology. Academic Press, New York.
Schmid, E.
1972 Atlas of Animal Bones. Elsevier Publishing Co., Amsterdam
Schmoll, H. R., B. J. Szabo, M. Rubin, and E. Dobrovolny
1972 Radiometric dating of marine shells from the Bootlegger Cove
clay, Anchorage area, Alaska. Geological Society of America
Bulletin 83:1107-1113.
Schwatka, F.
1885 Report of a military reconnaissance in Alaska, made in 1883.
U.S. Government Printing Office, Washington, D.C.
Semenov, S. A.
1964 Prehistoric Technology. Cory, Adams, and McKay, London.
Sharrock, F. W.
1966 Prehistoric occupation patterns in southwest Wyoming and
cultural relationships with the Great Basin and Plains Culture
area. Anthropological Pacers 77, University of Utah, Department
of Anthropology.
Sheets, P. D.
1973 Edge abrasion during biface manufacture. American Antiquity
38:215-218.
201
Sheppard, J. C.
1975 A radiocarbon dating primer. Bulletin 338, Washington State
University Radiocarbon Laboratory, Department of Chemical and
Nuclear Engineering, Pullman.
1977 Influence of contaminants and preservatives on radiocarbon
dates. Paper prepared for the 30th Annual Northwest Anthropological
Conference, Victoria, B.C., April, 1977.
Shippee, J. M.
1963 Was flint annealed before flaking? Plains Anthropologist
8:271-272.
S ho twe 11 , A .
1955 An approach to the paleoecology of mammals~ Ecology 36:327-
337.
Sirkin, L. A., S. J. Tuthill, and L. S. Clayton
1971 Late Pleistocene history of the lower Copper River Valley,
Alaska. Abstracts with Programs, Geological Society of America
3 ( 7): 708.
Skarland, I. and C. Keirn
1958 Archaeological Discoveries on the Denali Highway, Alaska.
Anthropological Papers of the University of Alaska 6(2):79-88.
Skoog, R. 0.
1968 Ecology of the Caribou (Rangifer tarandus granti) in Alaska.
Ph.D. Dissertation, University of California, Berkeley.
Solberger, J. B. and T. R. Hester
1972 Some additional data on the thermal alteration of siliceous
stone. Bulletin of the Oklahoma Anthropological Society 21:181-
185.
202
Spencer, R. F.
1959 The north Alaskan Eskimo: A study in ecology and society.
Smithsonian Institution Bureau of American Ethnology Bulletin 171.
Speth, J. D.
1972 Mechanical basis of percussion flaking. American Antiquity
37:34-60.
Struever, S.
1968 Problems, methods and organization: A disparity in the growth
of archaeology. In Anthropological Archaeology in the Americas,
edited by B. J. Meggers, pp. 131-151. The Anthropological Society
of Washington.
Taylor, D. W.
1965 The study of Pleistocene non-marine mullusks in North America.
In The Quaternary of the United States, edited by H. E. Wright and
D. G. Frey, pp. 597-612. Princeton University Press.
Thomas, D. H.
1969 Great Basin hunting patterns: A quantitative method for
treating faunal remains. American Antiquity 34:392-401.
Thorson, R. M. and T. D. Hamilton
1977 Geology of the Dry Creek Site: A stratified early man site in
interior Alaska. Quaternary Research 7:149-176.
Tikhomirov, B. A.
1959 The interrelationships of the animal life and vegetational
cover of the tundra. Translated from Russian by A. Merado.
Israel Program for Scientific Translations, 1966.
203
Tixier, J.
1964 Glossary for the description of stone tools. Newsletter of
Lithic Technology: Special Publication No. 1.
Treganza, A. E.
1964 Archeological Survey in Mt. McKinley National Park, 1964.
Report to the National Park Service, Unpublished Manuscript.
Tringham, R., G. Cooper, G. Odell, B. Voytek, and A. Whitman
1974 Experimentation in the formation of edge damage: A new
approach to lithic analysis. Journal of Field Archaeology 1(1-
2 ) : 1 71 -1 96 .
U.S. Department of Agriculture
1975 Revised Soil Survey Manual. Agricultural Handbook 18.
Washington, D.C. Manuscript copy.
U.S. Oepart~ent of the Interior
1974 Final environmental impact statement for proposed Mt. McKinley
National Park additions, Alaska. U.S. Government Printing
Office, Washington.
1976 Proposed electric distribution line extension to McKinley
Park, Mount McKinley National Park, Alaska. Final Environmental
Impact Statement, Pacific Northwest Region, National Park Service,
Seattle.
Van Schoonhoven, G. W.
1900 A quest for the Tanana River. In Compilations of Narratives
of Sxplorations in Alaska, pp. 736-737. U. S. Government Printing
Office, Washington.
204
f
Van Stone, J.
1974 Athapaskan Adaptations. Aldine, Chicago.
'/i ereck, L. A.
1966 Succession and soil development on gravel outwash of the
Muldrow Glacier, Alaska. Ecological Monographs 36:181-199.
Viereck, L. A., and E. Little
1972 Alaska Trees and Shrubs. U.S. Department of Agriculture,
Handbook No . 18.
Visher, G. S.
1965 Fluvial processes as interpreted from ancient and recent
fluvial deposits. In Primary Sedimentary Structures and their
Hydrodynamic Interpretation, edited by G. V. Middleton, pp. 116-
132. Soc. Econ. Paleotologists and Mineralologists Special Publication
12.
1969 Grain size distributions and depositional processes. Journal
of Sedimentary Petrology 39:1074-1106.
Wahrhaftig, C.
1958 Quaternary geology of the Nenana River Valley and adjacent
parts of the Alaska Range. U.S. Geological Survey Professional
Paper 293a. U. S. Government Printing Office, Washington, D.C.
1965 Physiographic divisions of Alaska. U.S. Geological Survey
Professional Paper 482. U. S. Government Printing Office, Washington,
D.C.
Webster, D. H. and W. Zibell
1970 Inupiut Eskimo Dictionary. Summer Institute of Linguistics,
Fairbanks.
205
West, F. H.
1965 Excavations at two sites on the Teklanika River, Mt. Mckinley
National Park, Alaska. Report to the National Park Service.
Unpublished manuscript on file, Mt. Mckinley National Park.
1967 The Donnelly Ridge site and the definition of an early core and
blade complex in central Alaska. American Antiquity 32:360-382.
1972 Archeological and paleoecological research in the Tangle Lakes,
Central Alaska, 1966-1972. A report of progress. Unpublished
manuscript.
1974 The significance of typologically early site collections in
the Tangle Lakes, Central Alaska: A preliminary consideration. In·
International Conference on the Prehistory and Paleoecology of
Western North American Arctic and Subarctic, edited by R. Scott and
P. Schledermann, pp. 217-238. Archaeological Association, University
of Calgary.
1975 Dating the Denali Complex. Arctic Anthropology 12(1):76-81.
l..leymouth, J. W., and M. D. Man de vi 11 e
1974 An X-ray diffraction study of heat-treated chert and its
archaeological implications. Archaeometry 17 (1 ):61-75.
White, A.
1963 Analytic description of the chipped-stone industry from Snyders
Site, Calhoun Illinois. In Miscellaneous ·Studies in Typo1ogy and
Classification, pp. 1-70. University of Michigan, Anthropological
Papers No. 19.
White, T. E.
1953a A method of calculating the dietary percentages of various food
animals utilized by aboriginal peoples. American Antiquity 18:396-398.
206
f
White, T. E.
1953b Observations on the butchering technique of some aboriginal peoples,
No. 2 American Antiquity 19:160-164.
Whitten, K. R.
1975 Habitat relationships and population dynamics of Dall sheep
(Ovis dalli dalli) in Mt. Mckinley National Park, Alaska. Unpublished
M.S. thesis, University of Alaska, Fairbanks.
Whymper, F.
1869 Travel and Adventure in the Territory of Alaska. New York.
Willey, G., and J. Sabloff
1974 A History of American Archaeology. Thames and Hudson, London .. ·
Wilmsen, E. N.
1970 Lithic analysis and cultural inference: A paleo-Indian case.
Anthropological Papers of the University of Arizona 16. Tucson.
Wissler, C.
1910 Material culture of the Blackfoot Indians. Anthropological
Papers of the American Museum of Natural History 5(1).
Wolfe, J. A., and C. Wahrhaftig
1970 The Cantwell fonnation of the Central Alaska Range. In
Changes in Stratigraphic Nomenclature by the U.S. Geological
Survey, 1968, pp. A41-A49. U.S. Geological Survey Bulletin
1294-A.
207
l·lomack, B. R.
1977 An Archeological Investigation and Technological Analysis
of the Stockhoff Basalt Quarry, Northeastern Oregon. Unpublished
:'1.A. Thesis, 1tJashington State University.
Workman, W. B.
1974 First dated traces of early Holocene man in the southwest
Yukon Territory, Canada. Arctic Anthropology 11 (Supplement): 94-
1 03.
1978 Prehistory of the Aishihik-Kluane area, southwest Yukon
Territory. Archaeological Survey of Canada Mercury Series 74.
Ottawa.
Yanert, It/.
1900 Account of an expedition from the middle fork of the Susitna
to the Talkeetna. In Compilations of Narratives of Explorations
in Alaska. U.S. Government Printing Office, Washington.
Ziegler, A. C.
1965 The role of faunal remains in archaeological investigations.
In Symposium on Central California .A.rchaeology, edited by F.
Curtis, pp. 47-75. Sacramento Anthropological Society Papers 3.
Sacramento.
208
209
POSTSCRIPT
Since the completion of this research, much additional work has been
undertaken in the upper Nenana Valley, primarily by the geologist Norman Ten-
Brink and colleagues associated with the North Alaska Range Project. Ten-Brink 1 S
research included a re-examination of the geology of the Carlo Creek site; his
preliminary interpretation of the site 1 S sediments suggested to him an alluvial
fan origin, as opposed to the fluvial model proposed herein, Although I still
regard as valid the fluvial interpretation, I do not feel that an alternative
geologic explanation would alter my interpretations of the site\s strati~yaphy,
technological or faunal analysis, or cultural-historical placement of the
artifacts recovered at this small early holocene butchering station. I remain
open to any new data which may favor other explanations, and hope that eventual
publication of the North Alaska Range Project \s research will serve to clarify
the various interpretations. My major interest at this point is to see that the
available data are disseminated in a timely fashion.
Peter M. Bowers
Fairbanks, Alaska
r