The understanding of human lifeways dur-ing the late Pleistocene into the earliestHolocene in the arid west of North Amer-
ica has benefited recently from important finds atseveral locales across the Great Basin; these in-clude Bonneville Estates Rockshelter in easternNevada (various but see Goebel 2007; Graf 2007),the Sunshine Locality in central Nevada (Beckand Jones 2009, 2010), and Paisley Five MilePoint Caves in south central Oregon (Jenkins2007), among others (Figure 1). This researchhas resulted in several publications describing ashift in lifeways from a Paleoarchaic to an Archaic
pattern (e.g., Beck and Jones 2009; Geib andJolie 2008; Graf and Schmitt 2007; Jenkins 2007;Pitblado 2003; Rhode et al. 2006). In short, schol-ars have described the Paleoarchaic lifeway asmore residentially mobile with a somewhatbroader diet, especially in terms of fauna ex-ploited, as compared to a less mobile Archaicpattern with a greater emphasis on large game,and this shift has been characterized as an “adap-tive discontinuity” (Pinson 2007). The questionthat has remained unanswerable until recently iswhether the shift from a Paleoarchaic to an Ar-chaic pattern as described for the Great Basin ex-
THE PALEOARCHAIC TO EARLY ARCHAIC TRANSITION ON THE COLORADO PLATEAU: THE ARCHAEOLOGY
OF NORTH CREEK SHELTER
Joel C. Janetski, Mark L. Bodily, Bradley A. Newbold, and David T. Yoder
Recent literature on Paleoarchaic and Early Archaic strategies in the arid west of North America has characterized the shiftfrom the earlier to later period as constituting an adaptive discontinuity. The empirical bases for this conclusion are shiftsin mobility, subsistence, and technology. Paleoarchaic peoples are described as highly mobile, focused on a diversity of ani-mal resources, including birds and small game, and using more exotic toolstone, an aspect of mobility. In contrast, EarlyArchaic strategies are less mobile, yet more focused on large game, using more local toolstone, and adopting grinding toolspresumably to process small seeds. Accompanying this transition is the shift from stemmed to notched projectile points.Data bearing on this transition on the Colorado Plateau have been scarce. Excavations of finely stratified deposits at NorthCreek Shelter on the Colorado Plateau in southern Utah have yielded support for the adaptive discontinuity position, althoughqualitative differences between the Great Basin and Colorado Plateau are apparent in mobility and faunal use.
Literatura reciente sobre las estrategias Paleo Arcaico y Arcaico Temprano en el oeste árido de Norte América (principal-mente el Great Basin) ha caracterizado el cambio de la época temprano a más tarde como constituyendo una discontinuidadadoptiva. El base empírico para esta conclusión son cambios en la movilidad, la subsistencia, y la tecnología. Gente PaleoArcaico son representados como altamente móvil, enfocado en una diversidad de recursos animales, incluyendo aves, y ani-males pequeños de caza, y confiando en toolstone más exóticas, un aspecto de movilidad. Al contrario, estrategias del ArcaicoTemprano son menos móvil, pero más enfocados en animales grades de caza, usando más herramientas de piedra local, yadoptando herramientas de moler, presumiblemente para preparar semillas pequeñas. Acompañando a esta transición es elcambio de pie a mella de puntos proyéctales. Datos apoyando esta transición en el Colorado Plateau han sido muy escasos.Las excavaciones de depósitos vena-fina a North Creek Shelter en el sur de Utah han producido apoyo por la posición de dis-continuidad adoptiva, sin embargo diferencias cualitativas entre el Great Basin y el Colorado Plateau son evidentes en lamovilidad y en el uso de fauna.
Joel C. Janetski � Department of Anthropology, Brigham Young University, Provo, Utah 84602Mark L. Bodily � US Forest Service, Humboldt-Toiyabe National Forest, Sparks, Nevada 89431Bradley A. Newbold � Department of Anthropology, Washington State University, Pullman, Washington 99164David T. Yoder � William Self Associates, Cedar City, Utah 84720
tends to the Colorado Plateau and whether such ashift constitutes an adaptive discontinuity as de-fined by Pinson (2007:203). Documenting abroader geographical pattern for such a shift pro-vides a basis for arguing that such changes werenot restricted to specific localities with specific,more historical causes, but demand more globalexplanations. New data from North Creek Shelter,a tightly stratified site on the Colorado Plateau of
southern Utah, bridge the Paleoarchaic and EarlyArchaic periods and provide for the first time anopportunity to broaden the discussion of this tran-sition beyond the Great Basin to include the Col-orado Plateau and the greater Southwest.We use the term “Paleoarchaic” to refer to the
earliest occupation in North Creek Shelter, al-though we recognize the debate among someGreat Basin scholars over the appropriate term for
UT
NV
CA
CO
AZ NM
WY
IDOR
Great Basin
ColoradoPlateau
Paisley Five MilePoint Cave
SunshineLocality
The Pits
North CreekRockshelter
Old ManCave
Joes ValleyAlcove
AspenShelter
Cowboy&
WaltersCaves
SuddenShelter
Dust Devil & Sand Dune CavesTsosieShelter
Danger Cave BonnevilleEstates
Rockshelter
Ventana CavePacif ic Ocean
N
S
EW
Figure 1. Location of select early archaeological sites discussed in the text.
pre-Archaic time periods. The recent text titledPaleoindian or Paleoarchaic?: Great Basin Hu-man Ecology at the Pleistocene-Holocene Tran-sition (Graf and Schmitt 2007) is evidence of thatdiscussion. Goebel (2007), for example, uses theterm “Pre-Archaic” in his chapter but opts for“Paleoindian” in his conclusions based on thehigh residential mobility reflected in the earliestchipped stone assemblage from Bonneville Es-tates Rockshelter. Haynes (2007) also concludesthat the earliest peoples differed from those in thelater Archaic in terms of mobility and technology:they were more mobile, made stemmed ratherthan notched points, and did not intensively utilizemilling technology (Haynes 2007:252). He ac-knowledges that early period diets remain poorlydefined but refutes the notion that the absence ofa “narrow prey-choice strategy of hunting biggame animals by people using fluted points” pre-cludes use of the Paleoindian label in the GreatBasin. Madsen (2007:15), however, maintainsthat the key technological difference between thePaleoarchaic and Archaic is the appearance ofgrinding stones in the latter and characterizes thestrategy shift from early to late as from broadspectrum to very broad spectrum, and for thatreason prefers “Paleoarchaic” to “Paleoindian.”Others in the Graf and Schmitt (2007) text use“Paleoarchaic” but do not engage explicitly inthe debate. Our use of “Paleoarchaic” rather than“Paleoindian” is based on the absence of extinctmegafauna in the earliest North Creek Shelterdeposits.1
Changes in human strategies during the LatePleistocene/Early Holocene transition were ac-companied by significant shifts in climates. Pale-oenvironmental modeling using data from Col-orado Plateau packrat middens (Betancourt 1984,1990; Betancourt and Davis 1984; Cole 1995,1990; Cole and Arundel 2005), for example, de-scribe the depression of late Pleistocene (~12,000to 9000 B.P.) vegetative belts by some 762 m inelevation, suggesting a cooler and wetter regimefor the region similar to that proposed for theGreat Basin prior to about 8300 B.P. (Bensen etal. 2002;.Grayson 1993, 2002; Huckleberry et al.2001; Madsen 2002; Madsen et al. 2001; Wigandand Rhode 2002; Rhode and Louderback 2007;Schmitt and Madsen 2005). Faunal evidence fromBechan and Cowboy Caves on the northern Col-
orado Plateau of Utah also suggest that theseearly conditions were preferable for mesic-lovingorganisms and now-extinct fauna (Agenbroad etal. 1989; Jennings 1980). A warmer, drier cli-mate, however, was present at Cowboy Cave byabout 8700 B.P., as evidenced by the expansion ofjuniper and oak (Quercus sp.) and the contractionof sagebrush (Artemisia sp.) (Spaulding and Pe-terson 1980). These trends are not unlike that forthe North American Southwest where the shiftaway from a winter dominant precipitation patternresulted in the replacement of more mesic vege-tation by more xeric plant communities by 8900B.P. at the earliest and 8000 B.P. at the latest(Huckell 1996:316).2
In a useful discussion of the Paleoarchaic toArchaic transition in the Great Basin, Haynes(2007) maintains that a fundamental characteris-tic of this transition is a record of initial dispersalof peoples whose stays were short but length-ened as they “settled in” to a region. Sites and pre-sumably populations are few and artifacts usuallyscarce in known Paleoarchaic levels (e.g., Elstonand Zeanah 2002:109). The process of settling inis often perceived as abrupt rather than gradualgiven the sometimes punctuated nature of siteuse and the corresponding archaeological record(however, see Madsen 2007:19). Visitation atBonneville Estates Rockshelter, for example, wasfairly regular in the earliest levels, beginning asearly as 11,000 B.P. and continuing perhaps atshort intervals until ~9400 B.P. based on abundantradiocarbon dates and stratigraphic interpretation(Graf 2007:97). Following this early occupation,there was a 2,000-year hiatus in site use untilabout 7400 B.P. by which time considerablechange had occurred in stone tool technology andsubsistence. Like Haynes, Goebel (2007:184) hascharacterized the earlier strategy as consisting ofrelatively short stays by people practicing highresidential mobility and who were tied to an “ex-tensive lithic conveyance system” as evidenced bymore exotic toolstone and more formal tools rel-ative to the later Archaic, although projectilepoints are rare at Bonneville Estates Rockshelteruntil about 6500 B.P. (Goebel 2007:173). Thisview is consistent with that of Jones et al. (2003),who have argued for large foraging territories forPaleoarchaic peoples in the Great Basin. By 7400B.P. Archaic peoples were practicing longer-term
rePOrtS 127
site use, using more local toolstone, and groundstone was common, a technology absent in the Pa-leoarchaic levels.Faunal data for the Great Basin are also indica-
tive of change from Paleoarchaic to Archaic times.In his summary of faunal data from Bonneville Es-tates Rockshelter, Hockett (2007:211–212) reportsa rich archaeofauna in the lowest levels with artio-dactyls, especially pronghorn, hares, and sagegrouse well represented. Early Archaic assem-blages, on the other hand, are less rich and tend tocontain more artiodactyls as a percentage of the as-semblage. Relying on faunal data from the north-ern Great Basin, Pinson (2004, 2007) has suc-cinctly characterized the Paleoarchaic to EarlyArchaic transition in that region in terms of climate,projectile points, and hunting focus. Pinson’s cli-matic discussion concurs with the pattern describedabove; that is, a transition from a cooler and wet-ter pattern in the Paleoarchaic to a warmer and drierpattern in the Archaic. Projectile points changefrom stemmed styles early to notched forms late.Contrary to subsistence expectations, but somewhatconsistent with Hockett’s (2007) findings, she notesthat artiodactyl use is low in the Paleoarchaic andrises in the Archaic. Grayson’s (1988:16) faunaldata from Danger Cave appear to support this viewas well since large game is scarce until DII times(~8500 B.P.), and bird bones diminish after DItimes (~10,100 B.P.) as well (Parmalee 1980:38;date estimates from Rhode et al. 2006). This im-portant issue is confronted in more depth later inthis paper.Additional subsistence insights for the transi-
tion come from botanical studies. Small seed useis evident in both the Paleoarchaic as well as theEarly Archaic levels at Bonneville Estates Rock-shelter, but intensive use of those resources, basedon the presence of ground stone, does not de-velop until 7400 B.P. (Rhode and Louderback2007:240). A similar transition, although datedroughly 2,000 years earlier (~8500 B.P.), is pre-sent in Danger Cave (Rhode et al. 2006). As notedabove, Madsen (2007:15) has pointed to the crit-ical addition of ground stone at about 8500 B.P.as evidence of a subsistence shift from the Pale-oarchaic to the Archaic.Reported human occupations on the Colorado
Plateau are later than those just mentioned for theGreat Basin, and few are earlier than 8000 B.P.
The paucity of data relating to pre–8000 B.P. oc-cupations has hindered attempts to even describethe transition on the Plateau much less under-stand it (but see Pitblado 2003). The earliest siteon the Colorado Plateau previous to the work atNorth Creek Shelter is The Pits on the RainbowPlateau in northern Arizona, which dates to 9780± 80 B.P., although recovered cultural data arelimited (Geib and Spurr 2002; Geib et al. 2007).Dust Devil Cave, also on the Rainbow Plateau,was tested in 1961 as part of the Glen CanyonProject with full excavation occurring in 1970; thelatter work was done by Northern Arizona Uni-versity (Ambler 1996). It yielded a basal date of8830 ± 160 B.P. from Stratum IV (Ambler1996:Table 7; Geib 1996:Table 5; Lindsay et al.1968:Table 22). This earliest level (~8800 to 6700B.P.) contained abundant ground stone and co-prolites containing both Sporobolus cryptandrus(dropseed) and Chenopodium sp. seeds (Van Ness1986; Van Ness and Hansen 1996) and appar-ently good numbers of faunal remains (Lindsay etal. 1968:120). The exact age of the artifacts is un-known, although one of the coprolites containingseeds was directly dated to 7630 ± 120 B.P. (Geib1996:Table 5). Tsosie Shelter on Black Mesa con-tains very deep deposits (~6 m) with the earliestlevels dating to about 8130 ± 80 B.P. (Smiley2002:19). A single piece of ground stone was as-sociated with the earliest occupation of the site,yet very little botanical or faunal information is re-ported for this early level.
Further to the north in the arid Canyonlandsregion of south-central Utah are Jim Walters andCowboy Caves. Both are large dry alcoves inves-tigated by the University of Utah in 1975 (Jen-nings 1980). Limited work was done at Jim Wal-ters Cave at the time, but recent reanalysis ofboth the site stratigraphy and additional dates ona twined sandal and fecal material suggests theearliest occupation of the cave is about 7800 B.P.(Geib and Jolie 2008). Cowboy Cave, on the otherhand, was the scene of extensive excavations. Thedeepest levels were noncultural but containeddramatic evidences of late Pleistocene fauna (bi-son, elk, mammoth, and sloth) and flora (spruce,Douglas fir, birch, and aquatic plants) document-ing a more mesic regime between 13,000 and11,000 B.P. (Hanson 1980; Spaulding and Peter-son 1980). The earliest cultural level, which con-
128 american antiquity [Vol. 77, no. 1, 2012]
tained ground stone as well as abundant seeds anda few rabbit bones, was previously dated to 8275± 80 B.P. (various in Jennings 1980; Schroedland Coulam 1994), but numerous new dates sug-gest no human occupation predates 8000 B.P.(Geib and Jolie 2008). The human occupationhere clearly occurred after the onset of a morexeric climate in the region.Joes Valley Alcove is the northernmost site on
the Colorado Plateau with human occupation dat-ing prior to 8,000 years ago. This site is in mixedconifer woodland on the northwestern edge of theColorado Plateau. Excavated by the U.S. ForestService in 1972, 1973, 1974, and 1978 (DeBloois1983; DeBloois et al. 1979) as well as by the Uni-versity of Utah in 1990 (Barlow and Metcalfe1993), the site yielded a date of 8940 ± 180 B.P.from the lower levels. Associated with this earlydate are Elko Corner-notched and Pinto Shoul-dered projectile points, abundant large animal bone,and plant remains. One of the points illustrated issimilar to the Pinto points from North Creek Shel-ter (Barlow and Metcalfe 1993:Figure 25e). Amano fragment came from Stratum 3 bracketed bythe just-mentioned date and 8510 ± 130 B.P.In sum, the Archaic is well documented on the
Colorado Plateau by 8000 B.P. with grindingtechnology, basketry, and Pinto points serving asdiagnostics (Geib and Jolie 2008; Matson 1991).Minimal information is available about earlyHolocene human occupation prior to 8,000 yearsago, however. Reasons for this may be the focuson dry caves in resource-poor regions (i.e., awayfrom permanent water) or the low populationsgenerally at this time. For example, surface findsof late Paleoindian projectile points (Pitblado2003) hint of early human presence, but the rar-ity of those early points suggest Paleoarchaic orPaleoindian age (~pre–9000; see various tables inPitblado 2003) sites are also scarce. Good num-bers of Early Archaic sites, on the other hand,have provided excellent insights into humanstrategies from about 8,500 years ago into themore recent past.The American Southwest stands apart from
both the Great Basin and Colorado Plateau giventhe well-documented presence of Paleoindian oc-cupations associated with extinct megafauna insouthern Arizona (Cordell 1997; Stanford 1999).The transition to an Archaic pattern is not well
documented, although both Huckell (1996) andCordell (1997) describe a northern Oshara Tradi-tion and a southern Cochise Tradition with the lat-ter dated somewhat earlier than the former, about9500 B.P. Ventana Cave (Haury 1950; Huckelland Haynes 2003) remains the best source of di-agnostic material for the period in question. TheArchaic pattern with ground stone is in place in theSouthwest by 8000 to 8500 B.P., including the useof grinding technology, but with either Jay or Ba-jada points (Huckell 1996:332; Matson 1991:129).
Research Issues
The primary issues emerging from the aboveoverview are subsistence, technology, mobility,and paleoenvironments. The received wisdom isthat there was a gradual warming and drying inplace in the Great Basin, Colorado Plateau, andSouthwest by 8500 B.P. Concomitant with thisclimatic transition was a broadening of the humandiet marking the onset of the Archaic throughoutthe region (Broughton et al. 2008; Grayson 1993;Huckell 1996; Matson 1991; O’Connell et al.1982; Zeanah and Simms 1999). This conclusionis based largely on the adoption of grinding tech-nology presumably for processing small seedsinto flour to make gruel or mushes (various, butsee Madsen 2007; Rhode and Louderback 2007;Rhode et al. 2006). The dietary broadening doesnot appear to pertain to targeted fauna, however.As noted above, Paleoarchaic archaeofaunas aremore diverse than the subsequent Archaic sug-gesting broader prey choice early and a more fo-cused hunting strategy in the Early Archaic. Thesepatterns fly in the face of foraging theory expec-tations (e.g., Elston and Zeanah 2002; Pinson2007). If the diet is broadening, as evidenced bythe appearance of grinding tools, one would ex-pect lower-ranked (i.e., smaller) prey to enter thediet as well; clearly that is not the case at manytransitional sites. According to Elston and Zeanah(2002), explanations for the Paleoarchaic and Ar-chaic contrasts are complex and must consider en-vironmental changes as well as differing strategiesfor men and women if we are to explain thisseeming counterintuitive pattern in prey choice.Pinson (2007) has explored this contradic-
tion in some depth and found no easy explana-tion, although she points to the possibility that
rePOrtS 129
different patterns may occur in contrasting en-vironmental settings. In any case, she notes theexistence of an “adaptive discontinuity across thePaleoarchaic/Archaic transition in the northernGreat Basin” (Pinson 2007:203). She, like anumber of researchers (e.g., Haynes 2007), alsonotes technological inconsistencies. Paleoar-chaic projectile points, for example, do notmatch the extant faunal assemblages given thatearly points are large, presumably for huntinglarge game, while Archaic points are smaller, yetearly faunal assemblages include significantnumbers of small prey while later Archaic sitesseem focused on large game.Mobility, on the other hand, seems less com-
plicated. As noted above, both Goebel (2007) andJones et al. (2003) suggest Paleoarchaic peopleswere wide-ranging based on the characteristics ofchipped stone assemblages (however, see Dukeand Young 2007 for some cautionary notes andBeck and Jones 2010 for additional insights).Later Archaic peoples were less residentially mo-bile with increasingly longer stays at locales suchas Bonneville Estates Rockshelter resulting in abroadening of the recovered artifact array includ-ing, for example, ground stone, but also the fullrange of biface production stages (Goebel2007:173).Below we present findings from North Creek
Shelter and discuss the implications for under-standing this transition and related issues by con-trasting those findings with patterns in the GreatBasin and, to a lesser extent, the southern RockyMountains and Basin and Range physiographicregions.
North Creek Shelter Setting and Description
North Creek Shelter lies on the west edge of Es-calante Valley, a small basin surrounded on thenorth, west, and southwest by highlands and well-watered by several streams draining these up-lands. Three streams come together within a quar-ter mile of North Creek Shelter— Upper ValleyCreek, Birch Creek, and North Creek— to formthe Escalante River. North Creek is by far thelargest of these streams and provides a reliablewater source today as it most likely did in the past.Elevations of these uplands vary from 3,444 m onthe Aquarius Plateau on the north to 3,081 m on
Table Cliff Plateau to the west, and 2,195 to 2,438m on the Kaiparowits Plateau on the southwest.Escalante Valley vegetation is Upper Sonoranwith greasewood and sagebrush/ grasslands on theflats and pigmy forest on hills surrounding thevalley. Cottonwoods and willows border the Es-calante River. Rising elevations characterize allthe contributing drainages and within a fairlyshort distance upslope are scattered ponderosapine and scrub oak with some Douglas fir onnorth-facing slopes. Higher elevations containdense ponderosa forest and, higher yet, stands ofaspen and blue spruce. Early European Pioneersreferred to the area as “Potato Valley,” a referenceto Solanum jamesii, a small wild “potato” that ap-parently grew in abundance here at the time ofAnglo settlement (Bamberg et al. 2003).The shelter lies at 1895 m elevation at the base
of a south-to-southwest-facing sandstone cliff ofthe Straight Cliffs formation. Scattered juniperand pinyon trees grow on the slope below the sitealong with a typical Upper Sonoran understory.Several Fremont-age granaries are tucked intothe cliff face overhead, and rock art, both paintedand pecked, is abundant. Rock art panels includehistoric inscriptions from Native American andEuropean visitors as well as Fremont (~1500 B.P.to 700 B.P.) and Archaic-style elements. A layerof historic dung overlies much of the flat area atthe cliff base, and chipped stone debris, ceramics,bone fragments, and ground stone are commonthere as well as on the slope below the cliff.Smaller rock art panels and scatters of cultural de-bris continue several hundred meters along thecliff to the east and north. In sum, the site is largeand clearly multicomponent, all of which testifiesto the attraction of the locale due to the availabil-ity of reliable water, passive solar heating, and, forlater farmers, arable land between the slope baseand the creeks.Crews from Brigham Young University and
University of Nevada, Las Vegas, excavated thesite from 2004 to 2008. The maximum horizon-tal exposure to date is 26 m2 and maximum depthof cultural deposits is 3.5 m below ground surface.A goal of the excavations was to obtain compa-rable samples of all early levels given the ex-pressed concerns that deep excavations often pre-clude adequate exposures resulting in a restrictedsample size (Huckell 1996:325). Upper strata ex-
130 american antiquity [Vol. 77, no. 1, 2012]
posures are, as a practical matter, broader than thedeeper levels; however, Strata II through V are allat least 2 m by 6 m in extent. We feel the samplesobtained are adequate to make plausible state-ments about the time periods represented. Thework revealed numerous discrete Paleoarchaic(10,000 to 9000 B.P.), Early Archaic (9000 to8000 B.P.), as well as Fremont (~1000 B.P.) oc-cupations, although the focus here is on the earlylevels. Numerous pits and hearths as well as abun-dant tools and faunal bone are present in all de-posits. All sediments were screened with .318 cm(one-eighth inch) mesh and artifacts greater than5 cm in size were point plotted on Early Archaic
and Paleoarchaic use surfaces; fill sediments wereremoved by 5 cm increments in most cases.
Site Stratigraphy
Site stratigraphy at North Creek Shelter is dra-matic due to a sharp contrast between lower(pre–8000 B.P.) and upper (post–8000 B.P.) de-posits (Figure 3). Geological study has concludedthat all sediments, early and late, derived initiallyfrom colluvial action brought on by precipitationor snow melt that carried silts and sands down ver-tical cracks in the cliff face to the west and east ofthe primary occupation area (Morris and Hicks2009). These waterborne sediments then were
rePOrtS 131
100 N100 E
Drip line
Rock faceat ground level
Rock face at -4.40 mbelow primary datum
0 1 2 m
NORTH CREEK SHELTER42GA5863Excavation Area N MN (15½º)
S
EW
Extent of Excavat ion
Area of Maximum
Excavat ion
Figure 2. Plan of site showing extent of excavations and area of maximum or deepest excavation
132 american antiquity [Vol. 77, no. 1, 2012]
Roc
kR
oden
t bur
row
Cha
rcoa
l-ric
hG
ray
sedi
men
tsR
edde
ned
m
Une
xcav
ated
Une
xcav
ated
0.00
1.00
2.00
3.00
4.00
5.00
100
E11
7 N
113
N11
4 N
115
N11
6 N
112
N11
1 N
110
N10
9 N
108
N10
7 N
106
N
IVV III
VII
I - L
ate
Preh
isto
ric 7
00 B
.P. –
His
toric
III
Early
Arc
haic
8,00
0 –
9,00
0 B
.P.
VII
- Fr
emon
t 110
0 –
700
B.P
.
VI -
Mix
ed A
rcha
ic94
0±40
B.P
.
7990
±30
/ 770
0±50
B.P
.79
70±8
0 B
.P.
8320
±120
B.P
.
8310
±70
B.P
.90
20±7
0 B
.P.
890±
40 B
.P.
9510
±80
B.P
.
9690
±60
B.P
.
9960
±30
B.P
. / 9
800±
50 B
.P.
Pale
oarc
haic
9,00
0 –
10,0
00 B
.P.
Ster
ile
Nor
th C
reek
She
lter
(100
Eas
t Pro
file)
dc
s
gl
e
n
o
pq
rb
c a
eg
ibfa
b
fa
acc
h
bd
tt
f
ae
e
gcc
d
gb
di
c/f
io
q
h
p
a
gh
nra
d
b
a
j
kl
m
ub
b
Cli ff F
ace
100
E10
6 N
107
N10
8 N
st
c c a
8320
±120
B.P
.
940±
40 B
.P.
109
N10
N1
b
b
d
a rp
q79
70±8
0 B
.P.
VII
- Fr
emon
t 1
VII
I - L
ate
Preh
isto
ric 7
00 B
.P
1 N
1112
N1
13 N
1
bbu
a
t
Red
dene
dG
ray
sedi
men
tsC
harc
oal-r
ich
Rod
ent b
urro
wR
ock
(100
Eas
t Pro
file)
Nor
th C
reek
She
lter
Arc
haic
VI -
Mix
ed
.10
0 –
700
B.P
VII
- Fr
emon
t 1
. – H
isto
ricV
III -
Lat
e Pr
ehis
toric
700
B.P
15 N
114
N1
Cli ff F
ace
Red
dene
dG
ray
sedi
men
tsC
harc
oal-r
ich
Rod
ent b
urro
wR
ock
m
(100
Eas
t Pro
file)
Nor
th C
reek
She
lter
16 N
117
N110
0 E
2.00
1.00
0.00
Ster
ile
.9,
000
– 10
,000
B.P
Pale
oarc
haic
.8,
000
– 9,
000
B.P
Arc
haic
Early
III
II IV IV
a hc/f
gc c g e a
d
9960
±30
B.P
. / 9
800±
50 B
.P.
9690
±60
B.P
.
9510
±80
B.P
.
890±
40 B
.P.
9020
±70
B.P
.83
10±7
0 B
.P.
n
hg
o i
c/f i d
b
de
f
h
af
ba fb i
ge
nl
g
9510
±80
B.P
.7970
±80
B.P
.79
90±3
0 / 7
700±
50 B
.P.
ml
jb
rq
p
o
e
c
Une
xcav
ated
d
k
Cli ff F
ace
4.00
3.00
Une
xcav
ated
fd b
ac5.
00
Figure 3. Stratigraphic profile of North Creek Shelter on the 100E line with selected radiocarbon dates.
carried via alluvial processes over the occupationarea resulting in horizontal bedding and somepooling, a process that continues today. The vary-ing energy level of these processes is visible incontrasting sediment sizes by level with alternat-ing deposits of coarse sands, finer sands, and silts,some of which are marked by desiccation poly-gons suggesting runoff ponding and drying. Thedifferences in the upper and lower strata are a re-sult of the rate of deposition. The upper one me-ter of fill is darkly stained from human activitiesthat occurred here over the 8,000 years duringwhich those sediments were deposited. As a con-sequence of this slow deposition, the upper layershave been subject to massive bioturbation fromrodent activity as well as human occupations thatmay have manipulated the site surface to accom-modate site use at various times. The nearly 2.5 mconstituting the lower levels, on the other hand,consist of discrete bands of tan sands and silts al-ternating with darkened cultural layers laid downover about 2,000 years. This relatively rapid de-position resulted in the sealing of occupation lev-els or zones marked by increases in material re-mains along with charcoal stained sediments. Theculturally stained darker bands are often imme-diately underlain by reddened sediments marking
burning episodes that oxidized underlying siltysands. These reddened bands are horizontally dis-continuous suggesting they are a consequence ofcultural activity rather than natural range fires. Asa result of these stratigraphic characteristics, ex-cavators were able to separate, document, andmap 15 occupation surfaces dating prior to 8,000B.P., although other less-extensive use episodeswere also noted. The sediments between the dis-crete surfaces were not sterile suggesting lighter,but more sporadic use. Further insight into site formation comes from
a geological assessment. Relying on dates andsediment analysis, the calculated rate of deposi-tion during the initial two millennia of site use is.18 cm/year; in contrast, the upper deposits accu-mulated at the rate of .017 cm/year, a ten-fold de-crease in the upper levels (Morris and Hicks2009). The implications for climatic shifts is clearfrom these data as well as from the absence of bio-turbation and the diminished charcoal staining inthe lower strata. The rapid deposition rate stronglysuggests climatic conditions during the earliestHolocene were more mesic with increased pre-cipitation. This, in turn, generated greater sedi-ment accumulation; after ~8,000 years ago pre-cipitation apparently slowed significantly
rePOrtS 133
Table 1. Radiocarbon Dates from North Creek Shelter.
Conventional Age 13C/12C Cal yrs ± 2� Cal yrs ± 2�Sample No Stratum Material 14C ± � yrs B.P. Ratio A.D./B.C B.P.
resulting in more gradual buildup of site deposits.Alternative explanations could include a decreasein mesa-top vegetation or greater rainfall intensityduring the early Holocene. The former seems un-likely in this case given that early Holocene veg-etative belts were lower in elevation and perhapsmore rather than less dense (Betancourt 1984).The latter explanation is a possibility and de-serves additional consideration. Paleoenviron-ments are discussed in greater detail below.Although bioturbation from rodents and hu-
man disturbance has disrupted regular horizontalbedding in the later Holocene, important featuresare present. Fremont age (~1000 B.P.) clay- andslab-lined hearths associated with remnants of aprepared use surface occur about 55 cm belowground surface in the more western excavatedgrids. Three of the hearths are on the same level.Site surface leveling to accommodate this inten-sive occupation may be responsible for the ab-sence of sediments separating Level VIc and VIdwhich yielded dates of ~6000 B.P. and ~1000B.P. respectively (Table 1). An additional clay-rimmed hearth with an associated compacted sur-face was present in the more easterly grids, al-though the hearth was not intact and the surfacewas discontinuous.
Dating
The 20 radiometric dates from North Creek Shel-ter span the past 10,000 years and, in general, arein proper sequence from the bottom to the top(Table 1). The exception is Beta sample 195226,which is clearly out of sequence. We obtained thissample from deep in the initial test in 2004, andit likely represents an unfortunate choice of char-coal that had fallen into the excavation from ahigher level during initial testing of the site byfield school students. That initial excavation wasa one by one meter test that extended to at leasttwo meters deep and which, by the time excava-tors had reached that depth, the size of the test wasroughly 70 cm square making sediment removalas well as exiting the test difficult without brush-ing the walls. We prevented subsequent contam-ination by expanding our excavation and main-taining absolutely vertical profiles. Differencesbetween Beta dates and Paleo Research Institute(UCIAMS) dates are explained as the result ofmore rigorous pretreatment.3
An examination of the dates reveals a signifi-cant break in site occupation between about 6,000and 900 years ago. The absence of mid-Holocenedates is not totally unexpected as there tends to bea mid-Holocene break or reduction in human useat a number of sheltered sites on this portion of theColorado Plateau, although the break here isgreater than seen at other sites. For example, OldMan Shelter (Geib and Davidson 1994), AtlatlRock Cave (Geib et al. 2007), Broken ArrowCave (Talbot et al. 1999), Cowboy Cave (Jen-nings 1980), and Sand Dune Cave (Lindsay et al.1968) contain few dates between ~7,000 and4,000 years ago (Geib 1996:31). The reasons forthis reduction in human use may be related to cli-matic warming and drying at this time which mayhave been especially severe in more arid reachesof southern Utah (see Geib 1996:32 for an ex-tended discussion of this demographic fluctua-tion). We hasten to point out that neither the Es-calante Valley nor the larger region wasabandoned during the mid-Holocene as BackyardAlcove, a site a few miles to the east of NorthCreek Shelter yielded a date of 3150 ± 40 B.P.(Beta 189342), which confirms human presencein the region at a time not yet documented (or atleast not dated) at North Creek Shelter. In addi-tion, middle Holocene sites are known in nearbyCapitol Reef National Park (Janetski et al. 2005),and upland areas of east central Utah such as As-pen Shelter and Sudden Shelter on the OldWoman Plateau yielded dates that fall in the mid-Holocene (Janetski et al. 1991; Jennings et al.1980:21 respectively). These dates, along withthat from Backyard Alcove, suggest that the ab-sence of mid-Holocene dates at North Creek Shel-ter is perhaps sampling error. This possibility isreinforced by the presence of Rocker Side-notched and Elko series points that suggest siteuse during the middle Archaic (Holmer 1986).These points are from Stratum VI, the heavily bio-turbated levels that have been avoided for datingsamples.
Use Surfaces
The identification of use surfaces raises long-standing concerns regarding the assumption thatartifacts found in close vertical or horizontal prox-imity can be used to reconstruct ancient activities.Such assumptions are fraught with pitfalls as de-
134 american antiquity [Vol. 77, no. 1, 2012]
tailed by O’Connell (1993; see also Kroll andPrice 1991 and references therein), especiallywith only modest exposures (O’Connell 1987). Asnoted above, excavators uncovered numerous usesurfaces in the early Holocene levels at NorthCreek Shelter. The assumption that these levelsrepresent activity areas used for a short period oftime, perhaps a few weeks at most, and that somebehavioral or functional interpretation is possible
was based on several criteria: (1) the reddened andcompacted sediments visible in the profile of theinitial test and which immediately underlay arti-fact and feature concentrations, (2) artifacts lyingflat on the presumed surface, (3) the existence ofpits or hearths originating at the same level, (4) ar-tifact point plots at depths within a few centime-ters of the reddened surface, (5) patterned distri-butions of artifacts and features sealed by rapid
rePOrtS 135
110 N
111 N
112 N
113 N
114 N
115 N
111 N
112 N
113 N
114 N
115 N
116 N
98 E
A’
A
A A’
Bone bead
Reddened surface Rock Bone FlakeCB CobbleCO CoreFL FlakeSC Scraper
0 1 m
NORTH CREEK SHELTER42GA5863Plan Map of Level II g
NMN (15½º)
S
EW
Baulk
Rock face
Light-colored ash
Extent of Hearth after excavation
Extent of Hearth prior to excavationAntler
CBSC
CO
Figure 4. Plan of Paleoarchaic Level IIg.
deposition. In addition, refitting of chipped stonedebitage to date have found numerous matches,and refit flakes occurred within discrete levels,typically within a 5 cm vertical range for the gridin which they were found. Refitting efforts thusfar have focused on spatially limited flake con-centrations and clearly demonstrate biface pro-duction occurred on site during the Paleoarchaic.Stratigraphic levels ranged in spatial extent
from ~9 m2 to 13 m2. This variation in exposureis a function of the excavation depth; that is, thedeeper we excavated the size of the sheltered areaincreased as the cliff face was undercut to thenorth (see Figure 3). Twelve of the 15 levels areidentified as activity surfaces or heavy use zonesmarked by varying quantities of artifacts and bonescrap and, in some cases, numbers of pits and/orhearths. Three of the 15 levels consist of a singlelayer of sandstone slabs and little or no culturaldebris, pits, or hearths. All three slab levels are inthe Paleoarchaic deposits. With the exception ofLevel Vt, the strata are quite horizontal, althoughsome slope up slightly toward the rear of the shel-ter as can be seen in the profile. The cultural lev-els were readily followed between grids 110 and113 North as some compaction and reddeningare present. Those attributes diminish toward therear of the shelter, which made following sur-faces difficult at times. Artifact and bone countspresented below are those within 5 cm or less ver-tical distance of the compacted/reddened sur-faces. Here we describe four activity areas andone of the enigmatic slab stone layers. Artifact andfaunal summaries for the Paleoarchaic and EarlyArchaic follow these Level descriptions. We chosethese levels, two each from the Paleoarchaic and
Early Archaic, as they represent extensive artifactscatters or contained discrete features that we feltbest illustrate the contrasts in the occupations.
Paleoarchaic Levels (10,000 to 9,000 B.P.)
Level IIg
This level lies about 3 m below ground surface. Asingle date from Level IIIa just above this surfaceplaces the occupation prior to 9690 ± 60 B.P.(Beta 221415). The dominant feature is a very dis-crete, oval, ash-filled hearth located somewhattoward the back of the shelter (Figure 4). Items onthis surface include a North Creek Stemmed pro-jectile point base (see discussion of this point be-low) (not plotted on Figure 4), a core, two steep-edged scrapers (one plotted), two unmodifiedquartzite cobbles, a few bone fragments, severalsmall unworked sandstone slabs, and two bonebead or tube fragments. These are the only orna-ments found in either the Early Archaic or the Pa-leoarchaic levels at the shelter. The two fragmentsmay be from the same bead but could not be re-fit. Both are decorated with scored incisingsplaced at regular intervals at right angles to thelong axis (Figure 5). Debitage is common with 83small internal flakes recovered. Deer bone wascommon on this level along with some leporids,a single canid element, and several rodents.
Level IIIb
Lying only a few centimeters below Level IIIe(see below), this is one of three levels character-ized by a layer of sandstone slabs (Figure 6). Eachof these rock levels was clearly visible in the pro-
136 american antiquity [Vol. 77, no. 1, 2012]
Figure 5. Bone bead or tube from Level IIg.
rePOrtS 137
Rock face
100 E 101 E108 N 108 N
109 N
98 E
99 E
109 N
110 N
111 N
112 N
113 N
114 N
110 N
111 N
112 N
113 N
114 N
115 N 115 N
116 N
Reddened surface Rock Bone Flake 0 1 m
NORTH CREEK SHELTER42GA5863Plan Map of Level III b
NMN (15½º)
S
EW
Instrusion
Extent of intrusion
Figure 6. Plan of Paleoarchaic Level IIIb.
file due to the strong reddening of sediments onwhich the slabs lay. The other two such levels areIIIg and IVb. The stones in Level IIIb form themost discrete pattern— an arcing array or pave-ment that extends into the east profile— of the
three such layers encountered and was, therefore,selected for illustration. All three distributionstend to extend to the east. The sandstone slabs ineach case were all lying flat, were not stacked, andwere relatively small (mostly less than 25 cm on
138 american antiquity [Vol. 77, no. 1, 2012]
Maxilla
Mandible
Cobble with ochre
Bonewith ochre TT
A A’
B B’
A
A’
B
CB
BF
CB
CB
CB
B’
Pit
Pit
Asymmetrical biface0 1 m
Reddened surface Rock Bone Lithic tool*BF BifaceCB Cobble
NORTH CREEK SHELTER42GA5863Plan Map of Level III e
NMN (15½º)
S
EW
*Lithic flakes not included on map
Area where surfacewas cut through
Polished stone
Unexcavated
Unexcavated
Mandible
Rock Face
109 N 109 N
108 N
110 N
111 N
112 N 112 N
113 N 113 N
114 N 114 N
115 N
116 N
115 N
116 N 98 E
111 N
110 N
99 E
98 E
108N 100 E 101E
101E
Ashyarea
Ashyarea
Figure 7. Plan of Paleoarchaic Level IIIe.
maximum dimension) and thin (less than 10 cm).All slabs were fire reddened and were immediatelyunderlain by abundant small diameter twig char-coal fragments that lay on the reddened sediment.Each appears to be a consequence of a singleburning episode. The relationship between thestones, the charred vegetative material and the un-derlying level suggest the site occupants laid downa mat of small twigs or other plant material, over-laid that layer with the sandstone slabs, and, at
some point, fired the plant layer. Artifacts andbone were very scarce on all such levels, althougha single flake and four bone fragments (just oneplotted on Figure 6) were present on Level IIIb.Due to the vertical proximity to IIIe, the heartharea from that level intruded into this level.Several facts support the conclusion that these
levels are a consequence of human behaviors/ac-tivities rather than the result of wholly naturalprocesses such as roof spalling. First, all three are
rePOrtS 139
111 N 111 N
110 N 110 N
109 N
108 N108 N
109 N
112 N 112 N
113 N 113 N
114 N
115 N
98 E
101 E
101 E100 E
99 E
102 E
102 E
MA
Hearth0 1 m
NMN (15½º)
S
EW
MA
GSMT
GS
MT
GS
Coprolite Coprolite
DR (Jimmy Allen point)
PP (Pinto point)
Reddened surface Rock Bone FlakeDR DrillGS GroundstoneMA ManoMT MetatePP Projectile pointR Rodent burrow
Pit
R
NORTH CREEK SHELTER42GA5863Plan map of Level V a
Rock face
Figure 8. Plan of Early Archaic Level Va.
restricted to the Paleoarchaic suggesting they areunique to this early period. In addition, the sand-stone slabs were lying flat rather than askew orotherwise jumbled suggesting intentional arrange-ment of the layer. Finally, in each case the sand-stone slabs form a rough circular pattern and areunderlain by burned material indicating redundantbehavior. Taken together, these patterns suggestthese rock levels represent human activity forsome unknown purpose.
Level IIIe
Level IIIe is about 2.7 m below ground surface. Ithas not been directly dated as of this writing, butis tightly sandwiched between dates of 9690 ± 60B.P. and 9510 ± 80 B.P. from levels IIIa and IVarespectively. This surface was the most inten-sively used Paleoarchaic level encountered or atleast the level yielding the richest artifact assem-blage. The exposure revealed an activity or dumparea on the east consisting of abundant tools andbone and an ashy, fire-reddened thermal area inthe west grids (Figure 7). Two small pits are pre-sent, one on the north and one on the south edgeof the amorphous ashy hearth area. The pits do notappear to have been thermal features as little red-dening is present; they may have served for smallitem storage. Associated with this level wereabundant chipped stone tools including a smallfragment of a North Creek Stemmed projectilepoint base (not plotted), a hafted asymmetrical bi-face (see below), bifaces, cores, and hammer-stones. Also present were numerous unifaciallyflaked stone tools (mostly steep-edged scrapersand utilized flakes), debitage (over 2,500 pieces),a small, black, highly polished chert cobblestone,several quartzite cobbles, and abundant faunalbone. The latter is dominated by deer bone frag-ments, although good numbers of cottontail rab-bit were also recovered at this level. Neither thedebitage nor bone could all be piece-plotted giventhe density of material, but the plan map reflectsthe tools and bone that were plotted. One of thecobbles has traces of hematite and subtle smooth-ing on one face. This area represents butcheringand processing activities as supported by the pres-ence of the many unifacial tools, tool sharpeningdebris, and the cobbles, which were perhaps usedas anvils for cracking marrow-rich bone. Charredgrass (unidentified), cheno-am, and Asteraceae
seeds were present on the surface as well.
Early Archaic (9,000 to 8,000 B.P.)
Level Va
Level Va, lying just over 2 meters below groundsurface, is significant as this marks the appearanceof grinding stones and notched points. Five sand-stone slab metate fragments and two manos— one complete and one fragmented, both one- handed— lay on this surface. In addition a singlePinto point basal fragment, a reworked JamesAllen projectile point base, a biface, a core, over200 pieces of debitage, and faunal bone were re-covered from this level. Again deer is most com-mon, although a single mountain sheep bone anda few leporids were present here. The surfacewas extensively reddened, more so than seen inthe Paleoarchaic levels. An amorphous hearth layin the central portion of the exposure and a cir-cular pit was present but not fully exposed on thesoutheast edge of the excavation (Figure 8). A sin-gle date of 9020 ± 70 B.P. (Beta 194030) camefrom level Vc just above this surface. A date of9510 ± 80 B.P. is from Level IVa about 60 cm be-low Va which suggests the onset of the Early Ar-chaic is close to the 9020 ± 70 BP date. Charredchenopod and Asteraceae seeds were present aswere uncharred Celtis seeds.
Level Vt
This surface was cut through by our initial ex-ploratory trench, and the resulting profile and sub-sequent horizontal exposure demonstrates that theLevel slopes up quite sharply on the north andmore gradually on the south (see Figure 3). Thisprofile is in decided contrast with the earlier lev-els, which are typically horizontal. The flattestportion of the surface lies at about 1.2 meters be-low ground surface and was hard packed and heav-ily reddened in places. The exposed area con-tained two hearths and numerous pits and in situartifacts (Figure 9). One circular, small diameter(10 cm), shallow (5 cm deep) pit with verticalsides and considerable charcoal in the fill may bea post mold, although no clear post remnants werepresent. These characteristics suggest that Level Vtmay have been covered by a superstructure whichburned. Several dates place this surface at about
140 american antiquity [Vol. 77, no. 1, 2012]
8000 B.P. At least 10 pits originated on this sur-face, some overlapping and most with reddenedrims and sidewalls suggesting they were used forroasting foods. Associated stone tools includethree metate fragments, a mano, a Pinto point,
several unifacially flaked tools, hammerstones,169 pieces of debitage, and a unique sharpeningstone. The latter was blocky sandstone with severalincised grooves, perhaps used for shaping orsharpening bone or wooden tools. Faunal remains
rePOrtS 141
Pit 10
Pit 11
Pit 9
Pit 8
Pit 7Pit 6
Pit 5
Pit 4
Pit 3Pit 2
Pit 1MT
MA
GS
Reddened edge Reddened surface Rock Bone FlakeGS GroundstoneMA ManoMT MetateR Rodent burrowUN Uniface
Post Hole
Niche
Hearth
Hearth
AntlerManible fragment
Shaped stoneUNUN
MT
Incised rock
Rock
Face
112 N 112 N
113 N
109 N
110 N
111 N
108 N
109 N
108 N 108 N
107 N
106 N
107 N
106 N
111 N
110 N
99 E98 E
101 E100 E
102 E
103 E 104 E
Estimated edge of Use Surface
0 1 m
Eroded and burrowed
NORTH CREEK SHELTER42GA5863Plan Map of Level V t
NMN (15½º)
S
EW
Figure 9. Plan of Early Archaic Level Vt.
142 american antiquity [Vol. 77, no. 1, 2012]
Figure 10. Selected projectile points and tools from the Paleoarchaic levels: (a–i) North Creek Stemmed, (j) asymmetri-cal biface, (k) reworked Jimmy Allen point (from Early Archaic level), (l) reworked Scottsbluff.
rePOrtS 143
Figure 11. Pinto projectile points from the Early Archaic levels.
consist of deer, mountain sheep, leporids, and ro-dents, including one porcupine.
Discussion of Paleoarchaic and EarlyArchaic Occupations at North Creek Shelter
Use Surfaces/Activity Areas
Paleoarchaic and Early Archaic activity areas dif-fered in several ways. For example, pits weremore common in the later levels with at least 20such features present in the Early Archaic levels,while only two were encountered in the Paleoar-chaic. The Early Archaic pits were very similar insize and shape, measuring 25 to 35 cm in diame-ter and up to 30 cm deep. Most were straight-sided with rounded bottoms, and a number dis-played heavily reddened rims and sides. The twoPaleoarchaic pits were smaller and shallower thanthose in the Early Archaic and lacked reddenedrims. Early Archaic pit function was likely roast-ing based on the reddening, although some mayhave served for storage. In fact, one straight-sidedpit on Level Vg contained a large quartzite flakeand a mano. Only one pit was bell-shaped and itlacked side or rim reddening clearly suggesting
storage (Level Vh).Contrasts were also present in hearth size and
depth. Hearths were present in all levels, but in thePaleoarchaic these tended to be rather large, shal-low, irregular hearth areas. The hearth in Level IIgis perhaps the best and most discrete example. Itwas oval and measured about 90 cm long by 60cm wide and up to 5 cm deep (Figure 4). A largehearth area was also present on Level Va, the ear-liest stratum containing ground stone. Above thislevel, however, hearths tended to be smaller(25–30 cm diameter) and round with straight sidesand somewhat deeper, ~10 cm. The best examplesof Early Archaic hearths were in Level Vt wheretwo were documented, one on the east edge of theexposed area and one intruding into the west pro-file (Figure 9).We found no slab rock levels in the Early Ar-
chaic. As noted earlier, we documented three suchlevels in the Paleoarchaic deposits.
Technology: Chipped and Ground StoneTools/Toolstone
The most dramatic difference in stone tools is inthe projectile point assemblage: stemmed pointsare present in the earliest Paleoarchaic levels
144 american antiquity [Vol. 77, no. 1, 2012]
Table 2. Paleoarchaic and Early Archaic Chipped Stone Tools.
while notched points appear at Level Va (the firstEarly Archaic occupation) along with groundstone and continue through Level Vt (Figures 10and 11) (Table 2). The North Creek Stemmedprojectile points are strongly shouldered withshort, parallel-sided stems and a rounded base(Bodily 2009). Stem margins and bases arelightly, but clearly, edge ground. Stemmed pointtoolstone is predominantly local (Morrison petri-fied wood, Boulder jasper, and Paradise chert)with the exception of obsidian (one reworked ob-sidian base, Figure 10c) and a greenish rhyolitefrom an unknown source (Figure 10i). Thesepoints are confined to Strata II and III, and, al-though not rare (n = 9), they are not as abundantas notched points from Stratum V. These stemmedpoints from North Creek Shelter are unique on theColorado Plateau and unlike Great BasinStemmed types, which tend to have long, taper-ing, variably ground stems, and are weakly shoul-dered, if at all (e.g., Beck and Jones2009:167–170). Comparisons well to the south re-veal morphological similarities of North CreekStemmed with Ventana-Amargosa I points fromthe Red Sand layer at Ventana Cave (Haury1950:202). Ventana Cave was revisited by Huck-ell and Haynes (2003), who obtained several dateson the Volcanic Debris layer just under the RedSand. These dates suggest the Red Sand layermost likely postdates 9000 B.P., which suggeststhe Ventana Cave stemmed points are later intime than the North Creek Stemmed. Other pos-sible affinities are with Jay and Bajada points ofthe Oshara Complex, although these are typicallyassociated with grinding tools and appear to dateafter 8500 B.P. suggesting an Early Archaic as-
sociation (Huckell 1996:332). Given the dissim-ilarity with Great Basin styles and in the absenceof comparable points on the Colorado Plateau, wefeel justified calling these points North CreekStemmed.Other projectile points from the Paleoarchaic
levels include a reworked stemmed point with asquared and heavily ground base from Stratum IV(Figure 10l). This point most resembles an Eden orScottsbluff point diagnostic of the Late Paleoin-dian Cody Complex on the Plains and RockyMountains that dates roughly to the same time pe-riod as the North Creek deposits that yielded thispoint (Frison and Todd 1987:98; Kornfeld et al.2010:90–91). The point is also similar to Windustprojectiles from the northern Great Basin (Beckand Jones 2009:196; Connelly and Jenkins1999:113). An unusual hafted biface from LevelIIIe is also reminiscent of the Cody Complex (Fig-ure 10j) as the curving leading edge is not unlikean asymmetrical biface recovered from the LarsenCache (Ingbar and Frison 1987:462; Kornfeld etal. 2010:384, Figure 6.36m). Another artifact withLate Paleoindian affinities is a James Allen pointbase reworked into a drill with the distal portionmissing (Figure 10k). It was found on Level Va,the earliest Early Archaic surface which dates justprior to 9000 B.P., which is consistent with recentdates on James Allen points from the Norton Sitein Kansas (Kornfeld et al. 2010:92). The lowermargins and basal concavity on this heavily pati-nated tool are strongly ground; the reworking hasremoved the patination, revealing what appears tobe Paradise chert, a local toolstone.Projectiles from the Early Archaic levels are
characterized by shallow, broad notches and
rePOrtS 145
Table 3. Toolstone from Paleoarchaic and Early Archaic Components.
deeply concave bases (Figure 11). There was noedge or basal grinding noted. One (Figure 11t) isremarkable for exuberant serrations, althoughseveral others are also lightly serrated. As with thestemmed points, these are made of local tool-stone, although not all material sources areknown. These points are abundant in Stratum Vwith 25 recovered, most of which are complete.Comparable points are illustrated by Holmer(1980:69, 1986:98) and Geib et al. (2001:195,Figure 6.12; see also Beck and Jones 2009:209).It is worth noting that two projectile points fromBurial 2 in Sand Dune Cave (Lindsay et al. 1968:Figure 23v) are very similar morphologically tothe North Creek Shelter notched assemblage, al-though the Sand Dune Cave burial dates to 4480± 60 B.P. (CAMS 10353, Coltrain et al.2007:316), considerably later than the NorthCreek Shelter points suggesting the Sand Dunespecimens were perhaps curated; Geib and Am-bler (1991:20) labels one Pinto (Figure 2, 5) andthe other as possible Elko-eared (Figure 2, 6).Deciding what to call these points was difficult
as they tend to overlap morphologically with
Pinto and Elko-eared types as described for theGreat Basin and Colorado Plateau (Geib et al.2001; Holmer 1980, 1986; Thomas 1981;Vaughan and Warren 1987). We tend to favorPinto as these points fall into the Pinto time frameprovided by Holmer for similar points from Sud-den Shelter and from Joes Valley Alcove, also onthe Colorado Plateau (Holmer 1980; see also Bar-low and Metcalfe 1993; Holmer 1986). This ty-pology issue is discussed at greater length underTechnology below.As noted, local toolstones were preferred in
both time periods with rather similar percentages ofMorrison petrified wood, Paradise chert, and Boul-der jasper (Table 3). Obsidian, however, is moreabundant in the Paleoarchaic. Obsidian sourcesused were all in the eastern Great Basin: MineralMountains, Black Mountain, Panaca Summit, andBlack Rock. All sources are represented in bothtime periods, although Mineral Mountain obsidianwas preferred in the Paleoarchaic with 77 percentvs. 44 percent in the Early Archaic.Other chipped stone tool contrasts include a
preference for unifacial tools in the Paleoarchaic,
146 american antiquity [Vol. 77, no. 1, 2012]
Figure 12. Notched obsidian flakes from Stratum II.
especially steep-edged or high-angle scrapers(Table 2). As noted, these were abundant on LevelIIIe and imply hide processing and other butcher-ing activities were more common in the Paleoar-chaic, despite the tendency for a greater focus onartiodactyls in the Early Archaic. These unifa-cially flaked tools were typically made by re-touching the distal ends of thick, chunky flakes,although some lateral margin retouch is present.Bifaces, both early and late stages, are common inboth components as are cores, hammerstones,and gravers. A final difference between the lithicassemblages is the recovery of eight notched ob-sidian flakes from Stratum II of the Paleoarchaicoccupation (Figure 12); these do not occur insubsequent strata. Seven of the eight notchedflakes exhibit opposable notching on the proximalend while the eighth is notched on the distal end.These are small tools; seven range in size from18-to-24-mm long by 13-to-17-mm wide withrounded notches measuring 4-to-8-mm in diam-eter (Bodily 2009:52). The eighth is larger (42-mm long by 24-mm wide) and with wider notches(13-to-15 mm in diameter). Similar notched ob-sidian flakes were recovered from Smith CreekCave on the Nevada/Utah border to the west
(Bryan 1979:208–210). The Smith Creek Cavespecimens were called “nosed” flakes and wererecovered from the Mount Moriah OccupationZone that dated between ~11,000 and 9,000 B.P.,making them contemporary with the North CreekShelter specimens. Light crushing wear in thenotches suggests use as spokeshaves.
Formal ground stone appears on Level Va asdescribed for that level above. The two manos areboth one-handed types. The slabs are all sand-stone and mostly thin with clear evidence of peck-ing and grinding, including parallel striations,and all but one are fragmented. The completespecimen is a lightly pecked and smoothed metatewith a subtle basin. Six small (less than 6.5 cm onmaximum dimension) fragments of thin (5-mmthick) sandstone slabs with light, but undeniablegrinding were recovered from the PaleoarchaicLevel IVr, directly below Va. Two of these frag-ments refit, and others are possibly from the sameartifact given similarities in size and smoothingpatterns. A possible grinding tool is present inLevel IVm as it has a central depression and someadjacent smoothing, but the depression looksmore eroded than ground and cannot be con-firmed as a tool. Also, Level IIIe yielded a cobble
rePOrtS 147
0 10 20 30 40 50
II
III
IV
V
VI
VII
VIII
Grinding Tool Frequency
Stra
ta
Slabs/Other Manos
Figure 13. Ground Stone frequency by stratum.
148 american antiquity [Vol. 77, no. 1, 2012]
Table 4. Mammalian Fauna from the Early Levels at North Creek Shelter.
used to process hematite (see Level IIIe above)demonstrating that not all grinding tools wereused to process foodstuffs. Considering the above,Level Va, dated to roughly 9000 B.P., is inter-preted as evidence of more intensive use of grind-ing technology, presumably for small seed pro-cessing. This is several centuries earlier than asimilar technology appeared in the Great Basin(see Yoder et al. 2010 and Rhode and Louderback2007 for more on this issue including a discussionof seeds). Grinding tools are common throughoutStratum V (Figure 13).
Fauna/Subsistence
Faunal analysis relied on comparative collectionsfor taxonomic identification. Taphonomic issuesnoted include breakage patterns, burning, butcher-ing marks, gnawing, density-mediated attrition,weathering, and evidence of carnivore damage. Inthe following discussion, large and small artio-dactyls are used synonymously with large gameunless otherwise noted. The unidentified cervidspresented in Table 4 are represented by antlerfrom either deer or elk. Additional study of themany artiodactyl specimens to clarify taxonomicidentification is ongoing.In general, bone preservation was fair to good
in all levels with 16.5 percent of the assemblage
identifiable to taxon and small, fragile, and frag-mented bone well represented throughout. Burn-ing was more common in the Early Archaic (28percent of all bone) than in the Paleoarchaic (11.6percent). Fragmentation was greatest in the EarlyArchaic with Paleoarchaic specimens averaging2.1 g per fragment and Early Archaic .6 g per frag-ment. Paleoarchaic specimens are more weatheredthan Early Archaic (11.8 and 6.3 percent respec-tively), although specimens from the deepest Pa-leoarchaic levels exhibited the least amount ofweathering. The high incidence (nearly 80 per-cent) of spiral and oblique fractures as well as nu-merous impact cones in both time periods suggestbreakage was due primarily to cultural activity(Lyman 1994). Evidence of carnivore damage tothe bones themselves was minimal. However, 24probable large carnivore coprolites were recov-ered during excavation; all but two are from thePaleoarchaic levels. These often contained bonebits and suggest carnivore visits to the site. Preser-vation of artiodactyl remains does not appear to bestrongly correlated with bone density for eitherthe Paleoarchaic (r2 = .029) or Early Archaic (r2
= .073) period as determined by density valuesprovided by Lyman (1982, 1984, 1994). Artio-dactyl age profiles constructed for each period re-flect some preference for juveniles by Paleoar-
rePOrtS 149
Table 5. Non-mammalian Fauna from the Early Levels at North Creek Shelter.
chaic hunters while Early Archaic hunters clearlytargeted adults (Newbold 2009).Taxonomic differences between the faunal as-
semblages of the two periods in question are moresubtle than those of other artifact categories, dif-fering more in degree than form. Unlike what hasbeen reported for other sites in the arid west dur-ing the Paleoarchaic/Early Archaic transition (var-ious in Graf and Schmitt 2007; Grayson 2002;Schmitt and Madsen 2005), a broadening of thediet does not appear to occur at North Creek Shel-ter during these times, or at least not to a great de-gree. This shift in subsistence focus is usually sig-naled by an increase in the diversity of low-rankedanimals (e.g., leporids, rodents, and birds) in re-sponse to the depletion of high-ranked resources(i.e., artiodactyls) due to Early Holocene environ-mental degradation (Broughton 1999; Ugan 2005).At North Creek Shelter artiodactyls dominate bothcomponents, especially mule deer (Table 4), whichsuggests these large animals were apparently moreabundant in the well-watered uplands of the Col-orado Plateau than in the Great Basin lowlandsduring the same period (see Hockett 2007; Pinson2007). Yet, while large game was the focus, faunaldata indicate relatively high prey diversity at thesite. Both genus counts and diversity values (ascalculated by Shannon and Simpson indices) arerelatively high during the two periods and actuallydecrease slightly over time (Shannon: .673 to .636;Simpson: .797 to .747). Significant contributionsby smaller mammals and birds to the Paleoar-chaic diet made it quite broad, while small gamediversity decreases between the Paleoarchaic andEarly Archaic periods (Shannon: .492 to .399;Simpson .858 to .781 respectively) (see Tables 4and 5).Seasonal indicators in the North Creek Shelter
faunal assemblage including perinatal artiodactylbones, deer skulls with shed or articulated antlers,and bones of hibernating animals and migratorybirds suggest redundant multiseasonal occupa-tion through most, if not all, times of the year dur-ing both the Paleoarchaic and Early Archaic. Anequal distribution of artiodactyl skeletal elementswithout relation to their economic utility in bothcomponents (Paleoarchaic: r2 =.049; Early Ar-chaic: r2 =.063) implies large-game harvestingin the immediate vicinity (Binford 1978, Met-calfe and Jones 1988). There are no signs of de-
pletion of high-ranked resources during theseearly periods, however. Rather, we see a sustainedabundance of large game across the Paleoar-chaic/Early Archaic boundary. The artiodactyl in-dex values for both periods are high and there isa slight increase in the proportion of high-rankedgame taken through time (.670 to .735), even asthe number of identifiable specimens remainscomparable for each period (Tables 4 and 5). Pin-son (2007) also observes an increase in artio-dactyl representation in the northern Great Basinthrough time, although the NISP counts fromNorth Creek Shelter are much higher, resulting inmore robust conclusions.While it appears that Early Holocene hunters
using North Creek Shelter were successful in cap-turing large game in abundance and a broad se-lection of other prey, some changes in speciespresence exist across the Paleoarchaic/Early Ar-chaic boundary. These are more than likely due toenvironmental transformation than to human pre-dation, as the declining species all seem to favorwetter and cooler upland habitats. For example,while mule deer remain the prey of choicethroughout, bighorn sheep move from a rarespecies in the Paleoarchaic to a secondary pref-erence in the Early Archaic, as evidenced by a sig-nificant reduction of the deer-to-sheep MNI ratio(18.3:1 to 2.4:1). Other smaller species (primar-ily rodents and birds) also undergo changes inabundance. Significant reductions are noted inbeaver, porcupine, rabbits and hares, groundsquirrels, woodrats, mice and voles, turkey,4 duskygrouse, and ducks, while pocket gophers, alongwith bighorn, increase in numbers (Tables 4 and5). It is apparent that the conditions at NorthCreek Shelter during the Paleoarchaic were bet-ter suited for montane and/or mesic-loving or-ganisms, but as the climate and environmentshifted toward more xeric conditions during theEarly Archaic, these organisms went from beingpresent or even abundant to rare or absent withinthe faunal assemblage.Insights into the dietary role of plants come
from the recovery of small charred seeds, whichinclude specimens representing Chenopodiaceae,Compositae, and Poaceae. These occur in mostearly levels and suggest some degree of plant usethroughout both Paleoarchaic and Early Archaicperiods. Grinding tools (manos and metates) are
150 american antiquity [Vol. 77, no. 1, 2012]
not present until the Early Archaic, however (seeYoder et al. 2010 for more on this topic).
Paleoenvironments
Today, North Creek Shelter sits near the upperlimit of the pinyon-juniper woodland, but faunaland floral evidence suggest the site was sur-rounded by a mixed conifer forest, possibly tran-sitioning into spruce during the early Holocene(~10,000 to 7500 B.P.). Supporting this notionare remains of montane and/or mesic-loving or-ganisms (e.g., beaver, diving ducks, yellow-belliedmarmot, dusky grouse, mustelids, voles, etc.) nowrare or absent from the site locale but present in theearliest deposits. Charcoal identification from Stra-tum II found Douglas fir (Pseudotsuga menziesii),netleaf hackberry (Celtis reticulata), mountainmahogany (Cercocarpus sp), Mexican cliffrose(Cowania mexicana), probable aspen (Populus cftremuloides), juniper (Juniperus sp.), and pine(Pinus spp.) (Puseman 2007). Douglas fir charcoalalso came from Strata III, IV, and V but was notfound Stratum VI. Dave Rhode (personal com-munication 2010) has tentatively identified at leasttwo Pinus species in submitted charcoal samplesincluding ponderosa (P. ponderosa), and possiblepinyon (P. edulis). The latter identification is ten-uous since pinyon was slow to arrive on the cen-tral Colorado Plateau (Betancourt 1990), althoughpackrat midden analyses suggest it is well-estab-lished on Navajo Mountain by 8000 B.P. (Kohler2007). With the exception of pinyon, these plantsdo not grow near the site today, suggesting thatvegetative belts were depressed in the very earlyHolocene. Additionally, as noted above, dates fromthe basal site sediments imply rapid depositionduring the Early Holocene, as more than 2.5 m ofsediments accumulated over about 2,000 years, be-tween Levels IIa and Vn, while about 1 m was de-posited during the subsequent 8,000 years or aboveStratum V.This climatic regime change is also observed
to some extent within the North Creek Shelter fau-nal record as evidenced by a decrease in or loss ofthe same numerous upland and mesic-loving an-imals and an increase in others suited to a drier en-vironment. The site is strategically located inwhat was likely a forest-wetland ecotone with itsaccompanying biotic diversity during the earliest
Holocene. Nearby was the confluence of threemajor stream systems fed by remnant glacialstreams and upland precipitation. The presence ofdiving ducks and especially beaver in the Pale-oarchaic is significant, as both require perma-nent, stable, slow-moving bodies of water typicalof a wetland ecotone. The Aquarius Plateau, di-rectly to the north of the site and the highestplateau in North America (Flint and Denny 1958),was the final refuge of Utah glaciation at the endof the Pleistocene (Stokes 1986). Boulder Moun-tain itself (the northeastern section of the plateau)supported an ice cap of more than 80 km2 drap-ing over the edges of the plateau in several broadlobes and outlet glaciers as long as 14.5 km thatdescended as low as 2012 m in elevation— a mere122 m above the shelter (Stokes 1986). The Boul-der Mountain ice cap did not begin its final retreatuntil sometime after ~15,000 B.P. (Marchetti et al.2005), and the flow from the recently meltedglaciers and their remains would have been sub-stantial at the Terminal Pleistocene/EarlyHolocene transition. In fact, the immediate vicin-ity of the shelter within the upper Escalante Val-ley at that time probably would have resembledthe environment of the higher regions of the Col-orado Plateau today, including rich forests of as-pen, spruce, and fir, as well as subalpine grass-lands and wet meadows— a decided contrast tothe more arid landscape today.
North Creek Shelter in a Broader Context
The fundamental issues identified earlier relativeto the Paleoarchaic/Early Archaic transition inthe arid west are subsistence, technology, mobil-ity, and paleoenvironments. We consider thesebelow and emphasize at the outset that the Col-orado Plateau Paleoarchaic is represented only byNorth Creek Shelter, a fact that makes general-izations for the Plateau tenuous. Below we com-pare North Creek Shelter primarily with GreatBasin sites given the absence of assemblages ofcomparable age reported for the Colorado Plateauand the greater Southwest.
Subsistence
The Paleoarchaic diet throughout much of theGreat Basin appears to have been taxonomicallyrich with large game animals present in low num-
rePOrtS 151
bers (Pinson 2007) with smaller prey more com-mon including birds, lagomorphs, and even in-sects (Grayson 1988; Hockett 2007; Pinson 2007).Early assemblages are still rare, however, and, insome cases, not yet reported in detail. A somewhatsimilar pattern is seen in the North Creek ShelterPaleoarchaic faunal assemblage—it is richer thanthe Early Archaic, yet it is clearly dominated bysmall artiodactyls, especially deer. This contrastswith reported Paleoarchaic sites in the Great Basin(Hockett 2007; Pinson 2007). Reasons for thiscould be differences in site function, preservation,or environments. The first two are readily dis-missed as all sites with assemblages for compar-ison appear to be short term residential bases,and preservation appears to have been good to ex-cellent in all cases. This leaves environmentalcontext as the most plausible reason for the dif-ferences. Although climates were somewhatcooler and moister than today in both areas, habi-tat, and therefore prey type and abundance, werequite different in Escalante Valley from those onthe western edge of the Great Salt Lake or insouthern Oregon 9,000 to 10,000 years ago.Despite these environmental differences, sev-
eral relative similarities exist between Early Ar-chaic subsistence patterns at North Creek Shelterand those in the Great Basin. In both areas faunalrichness decreases and small artiodactyls rise inrelative abundance through time; this is despitethe appearance of abundant grinding tools andthermal features that imply a broadening of thediet and increasing importance of plant foods af-ter ~9000 B.P., or ~8600 B.P. in the Great Basin.Foraging theory would predict an increase in dietbreadth and fewer small artiodactyls as low-returnresources such as small seeds enter the diet. Thediet does appear to broaden at later Archaic siteson the Colorado Plateau, however. At CowboyCave faunal remains are scarce in all levels, es-pecially Stratum II, the earliest cultural level,which contained only a few rabbit and prairiedog elements whereas plant remains and millingstones are abundant throughout (various in Jen-nings 1980). The heightened dietary importanceof plants relative to hunted game is also evident atother Colorado Plateau Archaic sites such as OldMan Cave, Sand Dune Cave, and Dust Devil Cave(see Van Ness and Hansen 1996 for a review).One explanation for this may be that these sites
represent complementary stops on an annualround with North Creek Shelter occupied at adifferent time of the year and in a contrasting en-vironmental context than those just mentioned(see also Van Ness and Hansen 1996:125).The above draws attention to a fundamental is-
sue discussed earlier: the findings at North CreekShelter and at early sites spanning the Paleoar-chaic/Archaic transition in the Great Basin appearanomalous within a foraging theory framework(Elston and Zeanah 2002; Grayson and Cannon1999; Simms 1987). Again, the received wisdomfrom that perspective is perhaps best expressed bythe earlier quote from Madsen (2007:15) whostates that the Paleoarchaic/Archaic transition ismarked by “a shift from broad-spectrum to verybroad-spectrum foraging” with the most impor-tant technological change at the boundary beingthe addition of grinding stones to process smallseeds. With a shift toward high-cost, low-returnresources such as small seeds one would alsopredict an expansion of the diet and the con-comitant reduction in high ranked prey such asdeer and mountain sheep in faunal assemblages(Elson and Zeanah 2002). Clearly at North CreekShelter the addition of grinding technology is notassociated with a decrease in large game; rather,we see greater emphasis on large game during theEarly Archaic and are left to explore why suchwould be the case.Pinson (2007:203) explains differences in Pa-
leoarchaic and Archaic faunal assemblages bysuggesting that foragers practiced contrastingland-use patterns and hints at increased Archaicefficiency in hunting large game. She suggeststhat Paleoarchaic peoples focused on small gameand associated resources in lowlands and marsheswhile Archaic groups foraged in the uplands, con-clusions buttressed by site distributions in thenorthern Great Basin (Pinson 2007:203). This ex-planation invokes a patch choice model towardunderstanding settlement and subsistence choices,an approach also embraced by Elston and Zeanah(2002). The patch choice model refines our un-derstanding of forager choices by recognizingthat resources are not randomly distributed acrossthe landscape, and that foragers shift from patchto patch as caloric returns from a currently ex-ploited patch fall below anticipated returns fortraveling to and foraging in a distant patch (Kelly
152 american antiquity [Vol. 77, no. 1, 2012]
1995). Elston and Zeanah (2002) consider genderroles in understanding shifts in Paleoarchaic andArchaic foraging patterns and assume a tradi-tional sexual division of labor; that is, men huntand women gather. Further, they propose that res-idences were located so as to insure access topatches exploited by women, who, in light of po-tentially high variability in hunting success, pro-vided resources essential to avoid food shortages.A similar approach may offer insights in the
North Creek Shelter case as well. The smaller prey— rabbits, marmots, beaver, ducks, and other birds— represent a more mesic patch exploitedduring the Paleoarchaic by women, and perhapschildren. That patch, which during the earliestHolocene would have been close by, provided atleast a supplemental dietary contribution for resi-dential groups spending time at the shelter. Withincreasing aridity, that patch diminished in pro-ductivity in the vicinity, and women shifted focusto grasses and other small seeded plants for theircontribution to family provisioning, although useof some smaller animals continued. Men contin-ued targeting large game, perhaps bypassing pre-viously exploited rodents and birds whose habitathad moved up in elevation and therefore furtherfrom the site and less attractive due to increasingtransport costs. The appearance of grinding stonesnot only makes women clearly visible for the firsttime, they, along with the constructed features onEarly Archaic levels, hint at longer stays to ac-commodate harvests and perhaps storage of earlyripening seeds as well as fall hunts. Insights frombotanical analyses now underway may provideuseful data to support these scenarios.
Technology
Technological shifts across the Paleoarchaic/Ar-chaic transition in the Great Basin occurred inboth ground and chipped stone technologies.Grinding tools become much more abundant re-flecting increasing importance of small seeds,and projectile points shift from large stemmed andlanceolate forms to smaller notched types (Pinson2007). Similar patterns are present at North CreekShelter with ground stone and notched points ap-pearing together at about 9000 B.P. Importantly,there does not appear to be a break in occupationof any duration at North Creek Shelter at the timeof this transition. Further, the array of Paleoar-
chaic chipped stone projectiles and butcheringtools seems appropriate given the abundant artio-dactyl remains, unlike the uneasy fit of largepoints associated with small prey seen at earlyGreat Basin sites. Other differences include de-creasing use of steep-edged scrapers through timeand the disappearance of the unique notched ob-sidian flakes found in the Paleoarchaic.As noted earlier, there is a question about the ty-
pological assignment of the notched points thatappear along with ground stone at 9000 B.P. atNorth Creek Shelter. The points can be designatedas either Elko-eared or Pinto depending on whetherprimacy is given to specific criteria such as bladethickness or overall size (see Bodily 2009 for a de-tailed discussion). We recognize that typing anddating Pinto points in the Great Basin has beenchallenging (e.g., Beck and Jones 2009:206–213;Holmer 1986:97–99; Vaughn and Warren 1987),and, although clearly important, a thorough con-sideration would be too lengthy to pursue here. Thesequential relationship of what we have calledPinto points and stemmed styles seen at NorthCreek Shelter holds some interest as it stands in ap-parent contrast with that seen on Old River Bedsites in western Utah (Schmitt et al. 2007) andeast central Nevada (Beck and Jones 2009). Inboth cases, Pinto points co-occur as surface findswith Paleoarchaic diagnostics such as Great BasinStemmed projectiles, crescents, and fluted points;ground stone is absent on those surfaces. Groundstone co-occurs with the Pinto points at NorthCreek Shelter, however. A critical difference, then,between the Great Basin and North Creek Sheltercases is that in the former the tools are surface findsand likely palimpsests; in the latter the sequence isstratigraphic. Regardless of the type names em-ployed, the fact remains that stemmed styles pre-cede rather than occur with notched types at NorthCreek Shelter, and the notched styles appear withformal grinding stones at 9000 B.P.
Mobility
Paleoarchaic residential mobility in the GreatBasin is described as high, both in the number ofmoves and the distance moved with the length ofstay at any one location correspondingly short.Residential mobility in the Archaic decreasedwith a concomitant increase in lengths of stay.This characterization is largely based on lithic
rePOrtS 153
studies demonstrating that Paleoarchaic toolstoneis often of high quality and is found long distancesfrom its source, while Early Archaic toolstone isoften of varying quality and tends to be local. Al-though obvious exceptions occur (e.g., Schmitt etal. 2007), this scenario suggests Paleoarchaic peo-ples were moving long distances to obtain theseresources directly whereas Archaic groups usedmore of what was on hand (e.g., Jones et al.2003), likely because reduced residential and lo-gistic mobility did not allow direct access to moreexotic materials. In addition, Goebel (2007) hasdescribed differences between Pre-Archaic andEarly Archaic stone tools and flaking debris atBonneville Estates Rockshelter, which argue forincreased length of stay by Early Archaic peoples.In contrast, Paleoarchaic toolstone at North CreekShelter is primarily of local origin with the ex-ception of obsidian, which occurs in higher pro-portions in the Paleoarchaic (Table 3).As mentioned above, some chipped stone tools
from the upper levels of the Paleoarchaic occupa-tion at North Creek Shelter are reminiscent ofLate Paleoindian Cody Complex artifacts sug-gesting affinities with the Rocky Mountains andPlains to the east. The North Creek Stemmedpoints are not contemporary with the Plains-liketools just mentioned; rather, they are restricted tothe earliest Paleoarchaic levels, and most stronglyresemble projectiles from the deep Southwest (i.e.,Ventana Cave in southern Arizona). These appar-ent affinities are clearly very tenuous due to thelow numbers, but they could imply widely rangingconnections through time and add to the charac-terization of high mobility in the Paleoarchaic.Factors other than chipped stone assemblages
are useful in assessing length of stay. Paleoarchaicliving surfaces at North Creek Shelter, for example,tended to have few formal features suggesting shortstays. Early Archaic levels, on the other hand, of-ten contained numerous pits perhaps used for on-site storage and resource processing, which sug-gests that Early Archaic groups were using the sitemore intensively and for longer periods of timethan Paleoarchaic groups. These changes could beevidence of a shift in site function with earlier oc-cupations more logistical in nature and later onesincreasingly residential. Such a shift should berepresented in the faunal remains in the form of dif-ferences in body-part representation, but Newbold
(2009) found that complete animals were likely be-ing transported to the site in both periods. The ap-pearance of pits and ground stone suggest increas-ing importance of plants that were typicallygathered by women as noted the ethnography of theKaiparowits band of Southern Paiute (Kelly1964:154). Such facilities also are evidence of anexpanded array of activities; therefore, site occu-pations were likely longer and more intensive.
Paleoenvironments
Paleoenvironmental insights from North CreekShelter complement the work of Betancourt and hiscolleagues on the Colorado Plateau (Betancourt1984, 1990; Betancourt and Davis 1984; Betan-court et al. 1990), by various scholars in the GreatBasin (Bensen et al. 2002; Grayson 1993, 2002;Huckleberry et al. 2001; Madsen 2002; Madsen etal. 2001; Schmitt and Madsen 2005; Wigand andRhode 2002), and those studying the central andsouthern Rocky Mountains (see Graham 2006 andPitblado 2003 for summaries). There is generalagreement that western North America experi-enced increasing aridity after 10,000 B.P. withcooler and wetter conditions lingering until about8500 B.P. (Pitblado 2003:41; Schmitt and Madsen2005:236), after which conditions were warmerand drier than previously. Climatic patterns werenot static, however, as periods of cooler weather aredocumented in the Northern Plains during the mid-Holocene (Kornfeld et al. 2010).The period of rapid sediment deposition at
North Creek Shelter appears to have ceased atabout 8000 B.P. given the visually dramaticchange in sediment color from tan below Level Vtto dark above that level. Geological analysis ofsediments concluded that aggradational processesat about 8000 B.P. were replaced by degrada-tional processes as a consequence of a combina-tion of climate change and site disturbance due tohuman and animal activity (Morris and Hicks2009). These findings are in general agreementwith the brief regional summary provided above.
Summary and Conclusions
North Creek Shelter on the Colorado Plateau ofsouthern Utah contains a finely stratified record ofmultiple occupations spanning the Paleoarchaic toEarly Archaic transition. The high-resolution de-
154 american antiquity [Vol. 77, no. 1, 2012]
posits at the site allow us to track changes intechnology, subsistence, and other aspects of earlylife at a scale that is rarely possible. Aspects of thearchaeological assemblage document some be-havioral continuity with chipped stone debrisnearly identical in terms of kinds of flakes,amount of cortex, and toolstone used. These pat-terns suggest similar activities occurred in bothperiods. Other remains suggest lack of continuityover the transition. Specifically, projectile-pointstyles shift from stemmed in the Paleoarchaic tonotched in the Early Archaic; projectile pointsare more abundant in the Early Archaic reflectinga greater focus on small artiodactyl hunting at thattime; access to obsidian diminishes through time;and, importantly, formal grinding tools appear atthe onset of the Early Archaic. The much greaterabundance of butchering tools in the Paleoarchaic(especially Level IIIe) is evidence that carcassprocessing was more likely to occur on site dur-ing this earlier period. The numerous Early Ar-chaic roasting pits suggest more intensive pro-cessing of foodstuffs, most likely plants. Thefinely stratified sequence at the site reveals thatthe transition was relatively rapid (Table 1). Thereason(s) for the rapid change are elusive but maybe related to climatic warming and drying as wellas more residential use of the location. Nonethe-less, sediment analysis and stratigraphic obser-vations suggest that, although precipitation ratesgradually slowed between 9000 and 8000 B.P.,conditions were still more mesic than the laterHolocene as evidenced by continued rapid depo-sition and the separation of occupation episodesduring this period. Those accretional processes di-minished significantly after 8000 B.P.Comparisons of the Paleoarchaic occupations
at North Creek Shelter with sites of similar age inthe Great Basin reveal some differences both interms of specific technologies and in subsistenceemphasis, although residential mobility appearshigh in both areas relative to the subsequent EarlyArchaic. We here repeat the question asked byPinson (2007:203): Is there adaptive discontinu-ity at the Paleoarchaic/Early Archaic transition atNorth Creek Shelter? We believe the answer tothat question is “yes.” Stemmed points give wayto notched types and ground stone appears at thetransition as seen in the Great Basin. In addition,formal features such as pits and discrete hearths
increase significantly in the Early Archaic. Thereasons for this shift are best understood by in-voking a strategy adjustment similar to that de-scribed by Pinson (2007) for the northern GreatBasin and refined by Elston and Zeanah (2002);that is, Paleoarchaic peoples tended to exploit awider range of prey and habitats while Early Ar-chaic foragers were more focused on large gamein the uplands and plant foods. Those changes areperhaps best understood by invoking a patchchoice model that includes a consideration ofgender roles in foraging societies. Despite differ-ences in specifics, the broader pattern of changein the arid west seems to hold for the ColoradoPlateau as well as the Great Basin. The ultimatecauses for this strategy shift are most likely relatedto changes in the climatic regime felt across thearid west of North America. Proximate causesmay be more localized and may include the move-ment of people and/or ideas across the landscape.We expect refinement of this scenario for the Col-orado Plateau as more early sites are discoveredand explored.
Acknowledgments. We recognize with gratitude the generos-ity of the Rex family, not only for allowing us to excavateNorth Creek Shelter, which is on their lands, but also for al-lowing us to camp and the free use of both electricity and wa-ter. We are grateful to the Grand Staircase-Escalante NationalMonument staff who involved us in regional research whicheventually led to the work at North Creek and who providedin-kind support for the work. This manuscript has benefittedby comments on the paper from Richard Talbot, David Mad-sen, and Charlotte Beck as well as useful discussions with JimWilde, Rick Holmer, and Phil Geib. Additional commentsand suggestions from the several anonymous reviewers werealso very helpful. Funding for the project was generouslyprovided by the College of Home and Family Sciences, theDepartment of Anthropology, and the Charles Redd Center forWestern Studies, all at Brigham Young University. Additionalfinancial support has come from the National Science Foun-dation, Award No. BCS-0818971, and grants from the Grad-uate and Professional Student Association, University ofNevada, Las Vegas.
References Cited
Agenbroad, Larry D., James I. Mead, E. M. Mead, and D. Elder1989 Archaeology, Alluvium, and Cave Stratigraphy: TheRecord from Bechan Cave, Utah. Kiva 54:335–351.
Ambler, J. Richard 1996 Dust Devil Cave and Archaic Complexes of the GlenCanyon Area. In Glen Canyon Revisited, by P. R. Geib, pp.40–52. University of Utah Anthropological Paper No. 119.University of Utah Press, Salt Lake City.
rePOrtS 155
Bamberg, John, A. Del Rio, A. Huaman, M. Martin, A. Salas,J. Pavek, S. Kiru, C. Fernandez, and David Spooner
2003 A Decade of Collecting and Research on Wild Pota-toes of the Southwest USA. American Journal of PotatoResearch, 80:159–172.
Barlow, K. Renee, and Duncan Metcalfe1993 Excavations at Joes Valley Alcove. University ofUtah Archaeological Center Reports of Investigations93–1. Salt Lake City, Utah.
Beck, Charlotte, and George T. Jones2009 The Archaeology of the Eastern Nevada Paleoarcha-ic, Part I: The Sunshine Locality. Anthropological PapersNo. 126. University of Utah Press, Salt Lake City.
2010 The Role of Mobility and Exchange in the Conveyanceof Toolstone During the Great Basin Paleoarchaic. In In-vestigating Prehistoric Trade and Exchange in WesternNorth American, edited by Richard M. Hughes. Universi-ty of Utah Press, Salt Lake City, in press.
Benson, Larry, Michaele Kashgarian, Robert Rye, Steve Lund,Fred Paillet, Joseph Smoot, Cynthia Ketser, Scott Mensing,David Meko, Susan Lindstrom
2002 Holocene Multidecadal and Multicentennial DroughtsAffecting Northern California and Nevada. Quaternary Sci-ence Review 21:659–682.
Betancourt, Julio L.1984 Late Quaternary Plant Zonation and Climate inSoutheastern Utah. Great Basin Naturalist 44(1):1–35.
1990 Late Quaternary Biogeography of the ColoradoPlateau. In Packrat Middens: the Last 40,000 Years of Bi-otic Change, edited by Julio L. Betancourt, T. R. Van De-vender, and Paul S. Martin, pp. 259–292. University of Ari-zona Press, Tucson.
Betancourt, Julio L., and O. K. Davis1984 Packrat Middens from Canyon de Chelly. QuaternaryResearch 21:56–64.
Betancourt, Julio L., T. R. Van Devender, and Paul S. Martin1990 Packrat Middens: The Last 40,000 Years of BioticChange. University of Arizona Press, Tucson.
Binford, Lewis R.1978 Nunamiut Ethnoarchaeology. Academic Press, NewYork.
Bodily, Mark K.2009 Residential Mobility of Paleoarchaic and Early ArchaicOccupations at North Creek Shelter (42GA5893): An Analy-sis of Chipped Stone Artifacts. Unpublished Master’s the-sis, Department of Anthropology, Brigham Young University,Provo, Utah.
Broughton, Jack M.1999 Resource Depression and Intensification during the LateHolocene, San Francisco Bay: Evidence from the EmeryvilleShellmound Vertebrate Fauna.Anthropological Records Vol.32. University of California Press, Berkeley.
Broughton, Jack M., David A. Byers, Richard A. Bryson, WilliamEckerle, and David B. Madsen
2008 Did Climatic Seasonality Control Late Quaternary Ar-tiodactyl Densities in Western North America? QuaternaryScience Reviews 27:1916–1937.
Bryan, Alan L.1979 Smith Creek Cave. In The Archaeology of Smith CreekCanyon, Eastern Nevada, edited by Donald R. Tuohy andDoris Rendall, pp. 163–251. Nevada State Museum An-thropological Papers No. 17. Carson City, Nevada.
Cole, Keith L.1990 Reconstruction of Past Desert Vegetation along the Col-orado River Using Packrat Middens. Palaeogeography,Palaeoclimatology, and Palaeoecology 76:349–366.
1995 Current Rates of Vegetation Change in National
Parks. In Our Living Resources 1994: A Report to the Na-tion on the Distribution, Abundance, and Health of U.S.Plants, Animals, and Ecosystems, edited by E. T. LaRow,G. S. Farris, C. E. Pucket, and P. B. Doran. National Bio-logical Service, Washington, DC.
Cole, Keith L., and S. T. Arundel2005 Carbon 13 Isotopes from Fossil Packrat Pellets and El-evational Movements of Utah Agave Reveal YoungerDryas Cold Period in Grand Canyon, Arizona. Geology33:713–716.
Coltrain, Joan B., Joel C. Janetski, and Shawn W. Carlyle2007 The Stable-and Radio-Isotope Chemistry of WesternBasketmaker Burials: Implications for Early Puebloan Di-ets and Origins. American Antiquity 72:301–321.
Connolly, Thomas J., and Dennis L. Jenkins1999 The Paulina Lake Site (35DS34). In Newberry Crater.A Ten-Thousand-Year Record of Human Occupation andEnvironmental Change in the Basin-Plateau Borderlands,by Thomas J. Connolly, pp.86–127. University of Utah An-thropological Paper No. 121. University of Utah Press, SaltLake City.
Cordell, Linda S.1997 Archaeology of the Southwest. Academic Press, SanDiego, California.
DeBloois, Evan I.1983 High Altitude Sites in Utah. In High Altitude Adap-tations in the Southwest, edited by Joseph C. Winter, pp.53–68. Cultural Resource Management Report No. 2. USDAForest Service, Southwestern Region, Albuquerque, NewMexico.
DeBloois, Evan I., Dee F. Green, and Jerry Wylie1979 Joes Valley Alcove: An Archaic-Fremont Site in Cen-tral Utah. Manuscript on file, Manti-LaSal National For-est, Price, Utah.
Duke, Daron G., and D. Craig Young2007 Episodic Permanence in Paleoarchaic Basin Selectionand Settlement. In Paleoindian or Paleoarchaic? GreatBasin Human Ecology at the Pleistocene-Holocene Tran-sition, edited by K. E. Graff and D. N. Schmitt, pp.123–138University of Utah Press, Salt Lake City.
Elston, Robert G., and David W. Zeanah2002 Thinking Outside the Box: A New Perspective on DietBreadth and Sexual Division of Labor in the PrearchaicGreat Basin. World Archaeology 34:103–130.
Flint, Richard Foster, and C. S. Denny1958 Quaternary Geology of Boulder Mountain, AquariusPlateau, Utah. US Geological Survey Bulletin 1061-D, pp.103–164. U.S. Government Printing Office, Washington,D.C.
Frison, George C., and L. C. Todd (editors)1987 The Horner Site: The Type Site of the Cody CulturalComplex. Academic Press, New York.
Geib, Phil R.1996 Glen Canyon Revisited. University of Utah Anthro-pological Paper No. 119. University of Utah Press, Salt LakeCity.
Geib, Phil R., and J. Richard Ambler1991 Sand Dune Side-notched: An Early Archaic ProjectilePoint Type of the Northern Colorado Plateau. Utah Ar-chaeology 4:17–31.
Geib, Phil R., and D. Davidson1994 Anasazi Origins: A Perspective from Preliminary Workat Old Man Cave. Kiva 60:191–202.
Geib, Phil R., and Edward A. Jolie2008 The Role of Basketry in Early Holocene Small SeedExploitation: Implications of a ca. 9,000 Year-Old Basketfrom Cowboy Cave, Utah. American Antiquity 73:83–102.
156 american antiquity [Vol. 77, no. 1, 2012]
Geib, Phil R., and Kimberly Spurr2002 The Forager to Farmer Transition on the RainbowPlateau. In Traditions, Transitions, and Technologies:Theses in Southwestern Archaeology, edited by S. H.Schlanger, pp. 224–244. University Press of Colorado, Boul-der.
Geib, Phil R., James H. Collette, and Kimberly Spurr2001 Kaibabitsinugwu: An Archaeological Sample Surveyof the Kaipariwits Plateau. Cultural Resource Series No.25. Grand Staircase-Escalante National Monument SpecialPublication No. 1. Bureau of Land Management, Salt LakeCity, Utah.
Geib, Phil R., Kimberly Spurr, and James H. Collette (editors)2007 Prehistory of the Northern Kayenta Anasazi Region:Archaeological Investigations along the Navajo MountainRoad, Volume II Archaic Site Descriptions. Navajo NationArchaeology Department Report No. 02-48. Flagstaff, Ari-zona.
Goebel, Ted2007 Pre-Archaic and Early Archaic Technological Activ-ities at Bonneville Estates Rockshelter: A First Look at theLithic Artifact Record. In Paleoindian or Paleoarchaic?Great Basin Human Ecology at the Pleistocene-HoloceneTransition, edited by K. E. Graff and D. N. Schmitt,pp.156–184. University of Utah Press, Salt Lake City.
Graf, Kelly E.2007 Stratigraphy and Chronology of the Pleistocene toHolocene Transition at Bonneville Estates Rockshelter, East-ern Great Basin. In Paleoindian or Paleoarchaic? GreatBasin Human Ecology at the Pleistocene-Holocene Tran-sition, edited by K. E. Graff and D. N. Schmitt, pp. 82–104.University of Utah Press, Salt Lake City.
Graf, Kelly E., and David N. Schmidt (editors)2007 Paleoindian or Paleoarchaic? Great Basin Human Ecol-ogy at the Pleistocene-Holocene Transition. University ofUtah Press, Salt Lake City.
Graham, Russell, W.2006 Climate and Biota of Western North America. In En-vironment, Origins, and Population, edited by Douglas H.Ubelaker, pp. 61–66. Handbook of North American Indi-ans, Vol 3, William C. Sturtevant, general editor, SmithsonianInstitution, Washington D.C.
Grayson, Donald K.1988 Danger Cave, Last Supper Cave, and Hanging RockShelter: The Faunas. American Museum of Natural HistoryAnthropological Papers Vol. 66, No. 1. New York.
1993 The Desert’s Past. A Natural Prehistory of the GreatBasin. Smithsonian Institution Press, Washington, D.C.
2002 Great Basin Mammals and Late Quaternary ClimateHistory. In Great Basin Aquatic Systems History, edited byR. Hershler, D. B. Madsen, and D. R. Currey, pp. 369–386.Smithsonian Contributions to the Earth Sciences No. 33.Smithsonian Institution, Washington, D.C.
Hansen, Richard M.1980 Late Pleistocene Plant Fragments in the Dungs of Her-bivores at Cowboy Cave. Appendix III, in Cowboy Cave,by Jesse D. Jennings, pp. 179–190. Anthropological PaperNo. 104. University of Utah Press, Salt Lake City.
Haury, Emil W.1950 The Stratigraphy and Archaeology of Ventana Cave.University of Arizona Press, Tucson.
Haynes, Gary2007 Paleoindian or Paleoarchaic? In Paleoindian or Pale-oarchaic? Great Basin Human Ecology at the Pleis-tocene-Holocene Transition, edited by K. E. Graff and D.N. Schmitt, pp. 251–258. University of Utah Press: Salt LakeCity.
Hockett, Bryan2007 Nutritional Ecology of Late Pleistocene to MiddleHolocene subsistence in the Great Basin: Zooarchaeolog-ical Evidence from Bonneville Estates Rockshelter. In Pa-leoindian or Paleoarchaic? Great Basin Human Ecologyat the Pleistocene-Holocene Transition, edited by Kelly E.Graff and David N. Schmitt, pp. 204–230. University of UtahPress, Salt Lake City.
Holmer, Richard N.1980 Projectile Points. In Sudden Shelter, by Jesse D. Jen-nings, Alan R. Schroedl, and Richard N. Holmer, pp. 66–84.Anthropological Papers No. 103. University of UtahPress, Salt Lake City.
1986 Common Projectile Points of the Desert West. In. An-thropology of the Desert West: Essays in Honor of Jesse D.Jennings, edited by Carol J. Condie and Don D. Fowler, pp.89–115. Anthropological Papers No. 110. University of UtahPress, Salt Lake City.
Huckell, Bruce B.1996 The Archaic Prehistory of the North American South-west. Journal of World Prehistory 10:305–373.
Huckell, Bruce B., and C. Vance Haynes, Jr.2003 The Ventana Complex: New Dates and New Ideas onit place in Early Holocene Western Prehistory. AmericanAntiquity 68:353–371.
Huckleberry, Gary, Charlotte Beck, George. T. Jones, A.Holmes, Michael. Cannon, Stephanie Livingston, and JackM. Broughton
2001 Terminal Pleistocene/Early Holocene EnvironmentalChange at the Sunshine Locality, North-Central Nevada,USA. Quaternary Research 55:303–312.
Ingbar, Eric E., and George C. Frison1987 The Larson Cache. Appendix 6, In The Horner Site: TheType Site of the Cody Cultural Complex, by G. C. Frison andL. C. Todd, pp. 461–474. Academic Press, New York.
Janetski Joel C., Lee Kreutzer, Richard K. Talbot, Lane D.Richens, and Shane A. Baker
2005 Life on the Edge: Archaeology in Capitol Reef NationalPark. Museum of Peoples and Cultures Occasional PapersNo. 11. Brigham Young University, Provo.
Janetski, Joel C., Richard I. Crosland, and James D. Wilde1991 Preliminary Report on Excavations at Aspen Shelter:An Upland Deer Hunting Camp on the Old WomanPlateau. Utah Archaeology 1991 4:32–43.
Jenkins, Dennis L.20007 Distribution and Dating of Cultural and PaleontologicalRemains at the Paisley Five Mile Point Caves in the North-ern Great Basin: An Early Assessment. In Paleoindian orPaleoarchaic? Great Basin Human Ecology at the Pleis-tocene-Holocene Transition, edited by Kelly E. Graff andDavid N. Schmitt, pp. 57–81. University of Utah Press, SaltLake City.
Jennings, Jesse D.1980 Cowboy Cave.Anthropological Papers No. 104. Uni-versity of Utah Press, Salt Lake City.
Jones, George T., Charlotte Beck, Eric E. Jones, and Richard E.Hughes
2003 Lithic Source Use and Paleoarchaic Foraging Territoriesin the Great Basin. American Antiquity 68:5–38.
Kelly, Isabel1964 Southern Paiute Ethnography. Anthropological PapersNo. 69. University of Utah Press, Salt Lake City.
Kelly, Robert L.1995 The Foraging Spectrum: Diversity in Hunter-GathererLifeways. Smithsonian Institution Press, Washington D.C.
Kohler, P. A.2007 Vegetation Changes on the Rainbow Plateau since the
rePOrtS 157
Last Glacial Stage. In Prehistory of the Northern Kayen-ta Anasazi Region: Archaeological Investigations along theNavajo Mountain Road, edited by Phil R. Geib and Kim-berly Spurr, pp. I.6.1–I.6.25. Navajo Nation Archaeolog-ical Department Report No. 02–48. Flagstaff, Arizona.
Kornfeld, Marcel , George C. Frison, and Mary L. Larson2010 Prehistoric Hunter-Gatherers of the High Plains andRockies. Left Coast Press, Walnut Creek, California.
Kroll, Ellen M. and T. Douglas Price (editors)1991 The Interpretation of Archaeological Spatial Patterning.Plenum Press, New York.
Lindsay, Alexander J., J. Richard Ambler, Mary A. Stein, andPhilip M. Hobler
1968 Survey and Excavations North and East of NavajoMountain, Utah, 1959–1962. Glen Canyon Series 8, Mu-seum of Northern Arizona Bulletin 45. Museum of North-ern Arizona, Flagstaff.
Lyman, R. Lee1982 The Taphonomy of Vertebrate Archaeofaunas: BoneDensity and Differential Survivorship of Fossil Classes. Un-published Ph.D. dissertation, Department of Anthropolo-gy, University of Washington, Seattle.
1984 Bone Density and Differential Survivorship of FossilClasses. Journal of Anthropological Archaeology 3:259–299.
1994 Vertebrate Taphonomy. Cambridge University Press,Cambridge, Massachusetts.
Madsen, David B.2002 Great Basin Peoples and Late Quaternary Aquatic His-tory. In Great Basin Aquatic Systems History, edited by R.Hershler, D. B. Madsen and D. R. Currey, pp. 387–405.Smithsonian Contributions to the Earth Sciences No. 33.Smithsonian Institution, Washington, D.C.
2007 The Paleoarchaic to Archaic Transition in the GreatBasin. In Paleoindian or Paleoarchaic? Great Basin Hu-man Ecology at the Pleistocene-Holocene Transition,edited by Kelly E. Graff and David N. Schmitt, pp.3–20.University of Utah Press, Salt Lake City.
Madsen, David B., David E. Rhode, Donald K. Grayson, JamesM. Broughton, Stephanie D. Livingston, J. M. Hunt, J.Quade, David N. Schmitt, and M. William Shaver III
2001 Late Quaternary Environmental Change in the Bon-neville Basin, Western USA. Palaeogeography, Palaeo-climatology, Palaeoecology 167:243–271.
Marchetti, D. W., T. E. Cerling, and E. W. Lips2005 A Glacial Chronology for the Fish Creek Drainage of Boul-der Mountain, Utah, USA. Quaternary Research 64:263–271.
Matson, R. G.1991 The Origins of Southwestern Agriculture. Universityof Arizona Press, Tucson.
Metcalfe, Duncan, and Kevin T. Jones1988 A Reconsideration of Animal Body Part Utility Indices.American Antiquity 53:496–504.
Morris, Thomas H., and Tanner C. Hicks2009 Sedimentology and Depositional History of the NorthCreek Shelter, Escalante, Utah. Ms on file, Department ofAnthropology, Brigham Young University, Provo, Utah.
Newbold, Bradley A.2009 Paleoindian Lifeways of Paleoarchaic Peoples: A Fau-nal Analysis of Early Occupations at North Creek Shelter,Utah. Unpublished Master’s thesis, Brigham Young Uni-versity, Provo, Utah.
O’Connell, James F.1987 Alywara Site Structure and Its Archaeological Impli-cations. American Antiquity 52:74–108.
1993 What Can Great Basin Archaeologists Learn from theStudy of Site Structure? An Ethnoarchaeological Perspective.Utah Archaeology 1993:7–26.
O’Connell, James F., Kevin T. Jones, and Steven R. Simms1982 Some Thoughts on Prehistoric Archaeology in the GreatBasin. In Man and Environment in the Great Basin, edit-ed by David B. Madsen and James F. O’Connell, pp.227–240. SAA Papers 2, Society for American Archaeol-ogy, Washington D. C.
Parmalee, Paul W.1980 Utilization of Birds by the Archaic and Fremont Cul-tural Groups of Utah. In Paper in Avian Paleontology Hon-oring Hildegarde Howard, edited by K. E. Campbell, Jr.,pp. 237–250. Contributions in Science No. 330. Natural His-tory Museum of Los Angeles County, Los Angeles, Cali-fornia.
Pinson, Ariane O. 2004 Of Lakeshores and Dry Basin Floors: A Regional Per-spective on the Early Holocene Record of EnvironmentalChange and Human Adaptation at the Tucker Site. In Ear-ly and Middle Holocene Archaeology of the Northern GreatBasin, edited by Dennis L. Jenkins, Thomas J. Connolly,and C. Melvin Aikens, pp.53–76. University of Oregon An-thropological Papers No. 62. Museum of Natural Historyand Department of Anthropology, Eugene, Oregon.
2007 Artiodactyl Use and Adaptive Discontinuity across thePaleoarchaic/Archaic Transition in the Northern Great Basin.In Paleoindian or Paleoarchaic? Great Basin Human Ecol-ogy at the Pleistocene-Holocene Transition, edited by Kel-ly E. Graff and David N. Schmitt, pp. 287–203. Universi-ty of Utah Press, Salt Lake City.
Pitblado, Bonnie L.2003 Late Paleoindian Occupation of the Southern RockyMountains: Early Holocene Projectile Points and Land Usein the High Country. University Press of Colorado, Boulder.
Puseman, Kathryn2007 Identification of Charcoal from the North CreekShelter Site 42GA5863, in the Escalante Valley, Utah. Tech-nical Report 05–59/06–81. PaleoResearch Institute, Gold-en, Colorado.
Rhode, David E., and Lisbeth A. Louderback2007 Dietary Plant Use in the Bonneville Basin During theTerminal Pleistocene/Early Holocene Transition. In Pale-oindian or Paleoarchaic? Great Basin Human Ecology atthe Pleistocene-Holocene Transition, edited by Kelly E.Graff and David N. Schmitt, pp. 231–247. University of UtahPress: Salt Lake City, Utah.
Rhode, David E., David B. Madsen, and Kevin T. Jones2006 Antiquity of Early Holocene Small-Seed Consump-tion and Processing at Danger Cave. Antiquity 80:328–339.
Schmitt, David N., and David B. Madsen2005 Camels Back Cave. Anthropological Papers No. 125.University of Utah Press, Salt Lake City, Utah.
Schmitt, David N., David B. Madsen, Charles G. Oviatt, andRachel Quist
2007 Late Pleistocene/Early Holocene Geomorphologyand Human Occupation of the Old River Bed Delta, West-ern Utah. In Paleoindian or Paleoarchaic? Great Basin Hu-man Ecology at the Pleistocene-Holocene Transition,edited by Kelly E. Graff and David N. Schmitt, pp.105–119.University of Utah Press, Salt Lake City.
Schroedl, Alan R., and Nancy J. Coulam1994 Cowboy Cave Revisited. Utah Archaeology 7(1):1–34.
Smiley, Francis E.2002 Black Mesa before Agriculture. In Prehistoric CultureChange on the Colorado Plateau: Ten Thousand Years onBlack Mesa, edited by Shirley Powell and Francis E. Smi-ley, pp. 15–34. University of Arizona Press, Tucson.
Spaulding, W. Geoffrey, and Kenneth L. Petersen1980 Late Pleistocene and Early Holocene Paleoecology of
158 american antiquity [Vol. 77, no. 1, 2012]
Cowboy Cave. Appendix II, in Cowboy Cave, by Jesse D.Jennings, pp. 163–178. Anthropological Paper No. 104. Uni-versity of Utah Press, Salt Lake City.
Stanford, Dennis1999 Paleoindian Archaeology and Late Pleistocene Envi-ronments in the Plains and Southwestern United States. InIce Age Peoples of North America: Environments, Origins,and Adaptations of the First Americans, edited by R. Bon-nichsen and K. L. Turnmire, pp. 281–339. Center for theStudy of the First Americans, Oregon State University Press,Corvallis.
Stokes, William L.1986 Geology of Utah. Utah Museum of Natural History Oc-casional Paper No. 6, University of Utah, Salt Lake City.
Talbot, Richard K., Lane D. Richens, Shane A. Baker, and JoelC. Janetski
1999 Broken Arrow Cave (42Ka4356): 1997 Testing Results.Museum of Peoples and Cultures Technical Series 99–5.Brigham Young University, Provo, Utah.
Thomas, David H.1981 How to Classify the Projectile Points from MonitorValley, Nevada. Journal of California and Great Basin An-thropology 3(1):7–43.
Ugan, Andrew S.2005 Does Size Matter? Body Size, Mass Collecting, andTheir Implications for Understanding Prehistoric ForagingBehavior. American Antiquity 70:75–90.
Van Ness, Margaret A.1986 Desha Complex Macrobotanical Fecal Remains: AnArchaic Diet in the American Southwest. Unpublished Mas-ter’s thesis, Department of Anthropology, Northern ArizonaUniversity, Flagstaff.
Van Ness, Margaret A., and Eric Hansen1996 Archaic Subsistence in the Glen Canyon Region. InGlen Canyon Revisited, by Phil R. Geib, pp.117–125. Uni-versity of Utah Anthropological Paper No. 119. Universi-ty of Utah Press, Salt Lake City.
Vaughn, S., and Claude N. Warren1987 Toward a Definition of Pinto Points. Journal of Cal-ifornia and Great Basin Anthropology 9:199–213
Wigand, Peter E., and David E. Rhode2002 Great Basin Vegetation History and Aquatic Systems:The Last 150,000 Years. In Great Basin Aquatic SystemsHistory, edited by Robert Hershler, David B. Madsen, andDonald R. Currey, pp. 309–368. Smithsonian Contributionsto the Earth Sciences No. 33. Smithsonian Institution, Wash-ington, D.C.
Yoder, David T., Mark L. Bodily, Joel C. Janetski, Sara Hill, andBradley A. Newbold
2010 The Onset of Small Seed Processing on the ColoradoPlateau. Kiva 76:425–446.
Zeanah, David W., and Steven R. Simms1999 Modeling the Gastric: Great Basin Subsistence Stud-ies Since 1982 and the Evolution of General Theory. In Mod-els for the Millennium: Great Basin Anthropology Today,edited by Charlotte Beck, pp. 118–140. University of UtahPress, Salt Lake City.
Notes
1. We hasten to add, however, that one of us (Newbold2009) titled his master’s thesis “Paleoindian Lifeways of Pa-leoarchaic Peoples: A Faunal Analysis of Early Occupations at
North Creek Shelter, Utah,” based on the abundance of largegame, primarily deer, in the earliest deposits. This title reflectssome considerable debate within the excavation team on thisissue.
2. All text dates are expressed in conventional radiocarbonages. Calibrated dates from North Creek Shelter are presentedin Table 1.
3. The following is a brief explanation for the differencesin the PaleoResearch Institute dates and those from Beta An-alytic. A more detailed comparison of the differences in datesis available from Linda Scott Cummings at the PaleoResearchInstitute in Golden, Colorado.
The two most likely contributors to the return of olderdates by PRI are more complete removal of humates (andbase-soluble organics) and the use of nitric acid to remove ni-tric acid-soluble organics, specifically the minute fungal re-mains, rootlets, and bacteria. It is interesting to note that thesamples that contained the largest quantities of humates, as ob-served at PRI, are the samples that yielded the largest spreadbetween dates by these two labs. This suggests that the partic-ular pieces of charcoal that were dated contained humates andprobably minute fungal remains, very fine rootlets, and/orbacteria. Addition of this younger carbon while the charcoalwas buried, and subsequent incomplete removal in the lab,would certainly have had the net result of making samples ap-pear younger than they are.
Charcoal sample 4029 was recovered from the fill of ahearth. Hearths usually contain more organic matter availableto nourish plants, fungus, and micro-organisms than strati-graphic sediments. These organisms that thrive in the nutrient-rich hearths, and the organic matter on which they feed, are pre-sent as nitric acid-soluble organics.
Finally, comparison of the calendar dates and associatedcalibration ranges indicates that for samples 4029 and 4364, theradiocarbon dates reported by PRI and Beta Analytic do notoverlap at the two sigma level (Table 1). The radiocarboncurve is very steep between approximately 9900 and 9600RCYBP, reflecting a probable drop in the production of ra-diocarbon in the atmosphere during this interval. The dates forsample 3716 do overlap at both the one and two sigma levels,based on the shape of the calibration curve.
4. Four bones (a sternum, matching right and left coracoids,and a probable rib) from the Paleoarchaic assemblage wereidentified as modern turkey (Meleagris gallopavo) and date be-tween 9000 and 9500 BP. If they indeed represent the modernand not an extinct species, which is likely given morphomet-ric comparisons, these specimens would be the earliest knownexamples of the M. gallopavo on the Colorado Plateau and thefarthest north and west within the prehistoric record. Further ex-ploration into this issue is ongoing.
Submitted March 29, 2010; Revised August 25, 2010;Accepted September 12, 2010.
