20
EI\IVIRONMENTAL CHANGEANI) ARCTIC PALEOINDIANS DANIEL H. MANN, RICIIARD E. REAMER, DOROTITYM. PETEET, MICHAEL L. KUNZ, and MARK JOHNSON Abstract. Intensive Paleoindian occupation of the Mesa site occurred between 10,300 and 9700 '4C yr B.P. Was the timing of this occupation controlled by environmental changes? We investigated stratigraphic records of geomorphology and vegetation to describe how landscapes in the Arctic Foothills changed between 13,000 and 8000 'nC yr B.P. Paleoindians were present during a time of rapid and sweeping changes in vegetation, slope erosion, floodplain dlmamics, permafrost stability, and soil type. We speculate that Paleoindians moved into arctic Alaska to exploit population highs in large ungulates that in turn were triggered by landscape-scale, geomorphic disturbances caused by climate changes. Environmental changes slowed and rising sea levels flooded Alaska's continen- tal shelves during the Early Holocene, causing summer temperatures to fall and precipitation to in- crease.A more stable climate and the prevalence of maritime air masses allowed the spread of tussock-tundra vegetation, which probably drove both the Paleoindians and their prey species out of the region. Introduction Paleoindians occupied arctic Alaska during a time of sweeping environmental changes at the Pleisto- cene-Holocene boundary. Did some of these envi- ronmental changes determine the timing of their occupation of Alaska? For events occurring 10,000 years ago, we can only infer cause and effect from temporal coincidences between environmental and human history. Elsewhere (Mann et al. 2002), we describe in detail how ecosystems responded to the rapid shifts in climate that occurred between 13,000 and 8000 'nC yr B.P.l in the Arctic Foothills. Here we infer the possible effects that these environ- mental changes had on the timing of Paleoindian occupations. Rising sea level was one of the most important factors driving paleoenvironmental change in Beringia during post-glacial times was ris- ing sea level. Preliminary analyses, presented here, describe how the flooding of Alaska's continental shelves by rising sea levels may have altered the re- gional climate. Our conclusions are speculative. We think that the Paleoindian occupations of Alaska, north of the Brooks Range, represent intermittent forays into an otherwise inhospitable territory by an opportunistic and mobile cultural group during brief intervals when its prey species reached high population den- Daniel H. Mann, Institute of Arctic Biology and ALaska Quaternary Center, Universitv of Alaska. Fairbanks.AK 99775 Richard E. Reanier,Reanier and Assobiaies, 1807 Thirty SecondAvenue, Seattle, Washington98122 Dorothy M. Peteet, NASA Goddard Institute for Space Siudies and lnmont Doherty Earth Observatory, Palisades,New York 10964 Michael L. Kunz, Bureau of Land Management,1150 university Avenue, Fairbanks,AK 99708 Mark lohnson, Institute of Marine Sciences, University of Alaska, Fairbanks, AK 99775 ARCTIC ANTHROPOLOGY Vol. 38, No. 2, pp. 119-138,zoo't ISSN0066-6939 O 2001by the Board of Regents of the University of WisconsinSystem

Date: 10/7/2007 9:17:16 PM

  • Upload
    hatu

  • View
    216

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Date: 10/7/2007 9:17:16 PM

EI\IVIRONMENTAL CHANGE ANI)ARCTIC PALEOINDIANS

DANIEL H. MANN, RICIIARD E. REAMER, DOROTITY M. PETEET,MICHAEL L. KUNZ, and MARK JOHNSON

Abstract. Intensive Paleoindian occupation of the Mesa site occurred between 10,300 and 9700 '4Cyr B.P. Was the timing of this occupation controlled by environmental changes? We investigated

stratigraphic records of geomorphology and vegetation to describe how landscapes in the Arctic

Foothills changed between 13,000 and 8000 'nC yr B.P. Paleoindians were present during a time of

rapid and sweeping changes in vegetation, slope erosion, floodplain dlmamics, permafrost stability,

and soil type. We speculate that Paleoindians moved into arctic Alaska to exploit population highs

in large ungulates that in turn were triggered by landscape-scale, geomorphic disturbances caused

by climate changes. Environmental changes slowed and rising sea levels flooded Alaska's continen-

tal shelves during the Early Holocene, causing summer temperatures to fall and precipitation to in-

crease. A more stable climate and the prevalence of maritime air masses allowed the spread of

tussock-tundra vegetation, which probably drove both the Paleoindians and their prey species out

of the region.

IntroductionPaleoindians occupied arctic Alaska during a timeof sweeping environmental changes at the Pleisto-cene-Holocene boundary. Did some of these envi-ronmental changes determine the timing of theiroccupation of Alaska? For events occurring 10,000years ago, we can only infer cause and effect fromtemporal coincidences between environmental andhuman history. Elsewhere (Mann et al. 2002), wedescribe in detail how ecosystems responded to therapid shifts in climate that occurred between 13,000and 8000 'nC yr B.P.l in the Arctic Foothills. Herewe infer the possible effects that these environ-

mental changes had on the timing of Paleoindianoccupations. Rising sea level was one of the mostimportant factors driving paleoenvironmentalchange in Beringia during post-glacial times was ris-ing sea level. Preliminary analyses, presented here,describe how the flooding of Alaska's continentalshelves by rising sea levels may have altered the re-gional climate.

Our conclusions are speculative. We think thatthe Paleoindian occupations of Alaska, north of theBrooks Range, represent intermittent forays into anotherwise inhospitable territory by an opportunisticand mobile cultural group during brief intervalswhen its prey species reached high population den-

Daniel H. Mann, Institute of Arctic Biology and ALaska Quaternary Center,Universitv of Alaska. Fairbanks. AK 99775

Richard E. Reanier, Reanier and Assobiaies, 1807 Thirty Second Avenue, Seattle, Washington 98122Dorothy M. Peteet, NASA Goddard Institute for Space Siudies and lnmont Doherty Earth Observatory,

Palisades, New York 10964Michael L. Kunz, Bureau of Land Management, 1150 university Avenue, Fairbanks, AK 99708

Mark lohnson, Institute of Marine Sciences, University of Alaska, Fairbanks, AK 99775

ARCTIC ANTHROPOLOGY Vol. 38, No. 2, pp. 119-138,zoo't ISSN 0066-6939O 2001 by the Board of Regents of the University of Wisconsin System

Page 2: Date: 10/7/2007 9:17:16 PM

120

sities. Similarities in timing and lithic technologybetween the Mesa and Agate Basin complexes of theHigh Plains (Bever, this volume) suggest these siteswere occupied by a related, if not single people,who consequently must have been mobile acrosscontinental distances. Whether Paleoindians movednorth into Alaska from the Great Plains or first de-veloped their cultural traits in the Arctic, their sub-sistence strategy probably was tied closely to bison.Bison may have flourished in northern Alaska dur-ing the Pleistocene-Holocene transition because ofrapid climatic shifts that maintained immatrue, rela-tively dry soils dominated by early successionalvegetation containing significant amounts of grass.Both bison and Paleoindians probably departedAlaska because of the spread of tussock-tundra veg-etation that was caused by an increasingly moistsununer climate, a change driven in part by theflooding of the Bering Land Bridge. The Paleoindi-an's intermittent and probably prey-specialized oc-cupation of northern Alaska may have been typicalof the opportunistic way humans utilized northernlandscapes during the periods of climatic instabilitythat characterized much of the Late Pleistocene.

Study AreaMany of the processes that organize the modernlandscapes of arctic Alaska are peculiar to this re-gion, and understanding them is necessary to inter-pret events in prehistory. Alaska's North Slope hastwo major physiographic divisions, the Arctic Foot-hills flanking the north side of the Brooks Rangeand the Arctic Coastal Plain lying between the Arc-tic Foothills and the Chukchi and Beaufort Seas(Fig. f ). Our 100 x 200 km study area is in the Arc-tic Foothills and crosses from the Brooks Range tothe southern edge of the Arctic Coastal Plain. Bothphysiographic divisions of the North Slope are de-scribed here because their environmental historiesare closely interrelated.

The Arctic Foothills ale east-west trendingridges of carbonate bedrock that protrude from tun-dra-covered plains (Grantz et al. 1994). Permafrost iscontinuous north ofthe Brooks Range and reachesa thickness of several hundred meters (Ferrians1994). Much of the Arctic Foothills reqion has neverbeen glaciated. Areas near the Brooks -Runge

and in-cluding the Mesa archaeological site (Kunz and Re-anier 1994) were last glaciated in the Tertiary andEarly Pleistocene. At the Iast glacial maximum(LGM), glaciers in the Brooks Range terminated atthe northern range front (Hamilton 1986).

The Arctic Coastal Plain is underlain by abroad, low relief bedrock surface that dips seawardfrom the Arctic Foothills. Over the last 3 m.v.. thesea has repeatedly tralsgressed and receded acrossthis surface, leaving a veneer of unconsolidated andinterfingering marine ald non-marine deposits

Arctic Anthropology 3B:2

(Dinter et al, 1990). Among the non-marine depositsare those of sandy river chalnels and deltas. Duringdry intervals in the Late Pleistocene, these sandysediments formed extensive dune fields and loessbelts (Carter et al. 1987). Immediatelv north of thestudy area, the Arctic Coastal Plain i! underlain bythe now-stable Ikpikpuk Dunes, which formed a

-

12,000-km2 sand sea during the LGM and was partlyreactivated during post-gla-ial times (Carter, rg-8f ,

-

1993; Dinter et al. 1990; Galloway and Carter1993).

Marked north-south gradients in climate occuracross the North Slope (Walker 1987, 2000). fulymean temperature increases from 4"C at Barrow to'1.2"C at Toolik Lake near the Brooks Range front(Zhang et al. 1996). Mean annual precipitation in-creases inland from 200 mm at Barrow to 320 mmat Toolik Lake. About half of the precipitation issnow, which persists on the ground for more thaneight months of the year. Rainfall increases over thecourse of the summer with maxima during cyclonicstorms in fuly, August, and September (Kane et al.1992). Most of these storms cross the Brooks Rangefrom the Bering Sea (Moritz 1,575).

During summer, much of the North Slope ex-ists in a state of waterlogged aridity. Potentialevapotranspiration equals or exceeds total annualprecipitation (Rovansek et al. 1996), actual evapo-transpiration exceeds annual runoff (Hinzman et al.1996), and the region is classified as semi-arid bythe Thomthwaite method (Patric and Black tg68;Newman and Branton '1,972). Nonetheless, soils atmany sites remain saturated throughout the summerbecause water tables are perched on the frozen, ice-rich substrate, and water is concentrated at theground surface. This situation can persist becauseevapotranspiration is greatest in early summer be-fore active layers fully thaw, and precipitation inlate summer recharges soil moistrue at a time whenevapotranspiration is low (Hinzman et al. 1996;Zhang et al. 1992).

Tundra covers the entire North Slope; how-ever, a major vegetation boundary lies along thenorthern edge of the Arctic Foothills (Walker etal. rSgA). In the Arctic Foothills, most vegetationis moist acidic tundra (Sphagno-Eriophoretum)dominated by dwarf shrubs (Betula nana, Ledumpalustre, Salix planifolia pulchra), tussock sedges(Eriophorum vaginatum), and acidophilous mosses,among which Sphagnum species are prominent. Animportant point for paleoecology is that ericaceousshrubs, Sphagnum moss, and Rubus chqmaemorus(cloudberry) are characteristic of moist acidic tundravegetation today, which is everywhere underlain bypeaty organic horizons (Bockheim et al. 1998). Onthe Coastal Plain, most vegetation is moist, non-acidic tundra (Dryado-integrifoliae-Caricetum bigelo-mi), which is dominated by non-tussock sedges(Carex bigelowii, C. membranacea, and Eriophorum

Page 3: Date: 10/7/2007 9:17:16 PM

Mann et al.: Environmental Chanse and Arctic Paleoindians 121

ChukchiSea

I

Arct lc Coeste, PIatn

r#tr

slopeo '

, nlllly .y,* L:E

;,"G*{gX$

r.' I

studyarea

- . \

BeringSea

Frgure r. The Arctic Foothills lie between the Brooks Range and the Arctic Coastal Plain on Alaska's North Slope. Theright panel shows the locations of stratigraphic sections described in Mann et al. (2002) and reviewed here. The squaresmark fluvial sections. NI is Nigu 1, N2 is Nigu 2; EF is on the East Fork of the Etivluk River; MC is Mesa Creek section;and fTr is on lteriak Creek just downsbeam of the Mesa archaeological site. Along the Ikpikpuk River, CT is Cotton-wood Bend; IKI is Ikpikpuk 1; ISB is Liftle Supreme Bluff; II(2 is lkpikpuk 2; and IK3 is lkpikpuk 3. LOP is the "Lakeof the Pleistocene." Triangles indicate solifluction sections; C is Cobra Gulch. TUK is Tukuto Lake. The portion of theArctic Coastal Plain shown here includes part of the stabilized rkpikpuk dunes.

30O km

-_

triste), prostrate shrubs (Dryas integrifolia, Salix arc-tica, S. reticulata, and Arctuous rubra), and minero-trophic moss taxa (Walker et al. 1998). Trees areabsent from the North Slope except for severalwidely separated stands of Populus balsamifera(balsam poplar) growing on floodplains in the ArcticFoothills (Murray 1980; Edwards and Dunwiddie1985). With a mean |uly temperature of 10-12oC,the Arctic Foothills lie near the Iatitudinal treelinetoday (Hopkins 1959; MacDonald et aI. 2000).

The general trend of soil development on theNorth Slope is towards paludification, the accumu-lation of waterlogged organic material on previouslywell-drained terrain (Bockheim et al. 1998). In theArctic Foothills today, much of the landscape isblanketed by 5-40 cm of peat (Everett and Brown1982; Ping et al. 1998), which is soil material con-taining >30% organic matter (Heathwaite et al.1993). The water logging that accompanies peat ac-cumulation influences vegetation distributionthroughout the region (Walker and Walker 1996;Walker et al. 1998). Sphagnum mosses play a key

role in paludification on the North Slope by acidifying soils, lowering soil-nutrient levels, and re-taining water. As the result of insulation by organicsurface horizons and their high water content, activelayers are typically only 25 cm thick in poorlydrained areas (Everett and Brovyn 1982; Kane 1996).In contrast, summer thaw on well-drained slopeswith thin or disturbed organic surface horizons canpenetrate to depths of 1 m (Kane 1996). Organic soilhorizons affect floodplain dynamics by retardingslope erosion and limiting the input of sediments tostreams (Hinzman et al. 1996).

Floodplains play key roles in North Slope eco-systems. In contrast with much of the surroundingterrain, floodplain soils have deep active layers, lackthick organic horizons, and are often well drained(Ping et al. 1998). They are sites of high primaryproductivity and plant-species diversity (Shaver etal. 19go). Unvegetated parts of floodplains aresorrrces of loess, which has pervasive effects onsoils and vegetation located dor,rmwind and beyondfloodplain margins (Walker and Everett L991). Loess

Page 4: Date: 10/7/2007 9:17:16 PM

122

and sand input retards soil acidification, which lim-its organic-matter accumulation and maintains rela-tively deep and well-drained active layers.

The major runoff event of the year on theNorth Slope is the snowmelt flood (Arnborg et al.1966; Carter et al. 1987; Kane 1996); however, fro-zen soils restrict erosion during snowmelt, and sum-mer rainstorms probably cause most erosion onslopes, This also is true for the beds of smallerstreams where bottom-fast ice and frozen sedimentsarmor channels during snowmelt (Scott 1978).Where channels impinge on higher terrain, a largepart of lateral channel migration is due to thermalerosion of ice-rich permafrost (Carter et al. rgAz).Most streams have wandering, gravel- and cobble-bedded channels today, though braided reaches ex-ist in areas of aufeis accumulation, and low-orderstreams often are beaded or straight-channeled be-tween balks of neat.

Results and DiscussionChronology of Human Occupation atthe Mesa Site

When did Paleoindians live in the Arctic Foothills?The answer to this ouestion defines the time ueriodof interest. Fifty-one radiocarbon dates have beenobtained fiom the Mesa site (see Bever, this vol-ume), providing a degree of age control that isunique for a Paleoindian archaeoloeical site. Sevenof these ages come flom standard, idiometric dateson multiple grams of wood charcoal, while theother 44 are accelerator mass spectrometry (AMS)dates on milligrams of wood charcoal. Paired radio-metric and AMS samples reveal that the radiometricages from the Mesa site are anomalous and usuallytoo young. Only the ages of the 44 AMS samplesare considered further.

The 1-sigma errors of 42 of the Mesa AMSdates faII between 12,4oo and to,+oo cal yr B.P. Un-fortunately for chronological precision, the C-14time scale is distorted by global fluctuations in theconcentration of radiocarbon that occurred duringthe Younger Dryas (YD) chronozone (ca. 11,000-10,000 'nC yr B.P.) (Goslar et al. 1995; Goslar et al.2000; Biorck et al. 1996; Haidas et al. 1998; Kita-gawa and van der Plicht 1998). As a result, radiocar-bon ages in this time range correspond to anunusually wide range of calendar ages (Stuiver et al.2000).

Forty-one of the Mesa AMS dates lie between10,300 and 9700 'aC yr B.P. The combined, 1-sigmaerrors of these dates span the interval from 12,350to 11,100 cal yr B.P. An outlier date of g33O 14C yrB.P. has a 1-sigma error range between 10,640 andL0,430 cal yr B.P. Unlike the other 41 dates, the 2-sigma error of this young outlier does not overlapwith the 2-sigma error of the next older date.

Arctic Anthropology 38 :2

The age distribution of thirteen AMS datesfrom Gordon's Hearth provides insights into the pre-cision of age control at the Mesa site. The radiocar-bon ages from Gordon's Hearth encompass the ageranges of the remaining 28, post-11,000 'nC yr B.P.dates from the Mesa, with the exception of theyoung outlier just described (Fig. 3). Gordon'sHearth probably was used only briefly, probably forseveral hours to several days. How could the ages ofcharcoal fragments in this one hearth span most ofthe history of the entire site? As noted earlier, largeand ranid fluctuations occurred in atmosoheric 'nClevels during the YD and earliest Holocene causingradiocarbon ages in this time range to correspond toan unusually wide range of c;alendar ages. Perhapscontributing to this imprecision is the fact that Pa-leoindians were burning wood of varying ages. Theoldest willow bushes growing along Iteriak Creek to-day are <50 years old; however, dead wood couldconceivably be preserved cln gravel bars for severaladditional decades. In light of the wide range ofages from Gordon's Hearth, we cannot eliminate thepossibility that the Paleoindian occupation of north-ern Alaska was very brief. At one extreme, their oc-cupation of the Mesa site might have lasted only thehours or days that a fire burned in Gordon's Hearth.The large number of Paleoindian hearths at theMesa (n > 30) indicates this is unlikely and sr.rg-gests multi-year occupation of this site (Kunz andReanier, 1994). The Mesa site probably was occu-pied seasclnally over an intervil ranging from de-cades to centuries.

Most radiocarbon dates fiom other sites wherenorthern Paleoindian artifacts occur fall within therange of ages from Gordon's Hearth at the Mesa site.Charcoal fragments associated with edge-ground,lanceolate projectile points at two other BrooksRange sites, Hilltop and the nearby Bedwell site(Reanier, 1995), produced dates whose 1-sigma er-rors range from 12,810 to 11,960 cal yr B.P. (Fig. Z),making them slightly older than the 42 Mesa datesjust described, though their 2-sigma age rangesstill overlap with ages from Gordon's Hearth. Atthe Engigstciak site on the Beaufort Sea coast innorthwestern Canada, three bones of extinct Bisonpriscus yielded radiometric dates of 9400 -1_ 230,9770 | 180, and 9870 + 180 14C yr B.P. (Cinq-Marset al. 1991; MacNeish, 2000). The calibrated ages ofthese samples overlap with ages from Gordon'sHearth (Figs. 2, 3), and collections from the En-gigstciak site seem to include Mesa-type projectilepoints. In southwestern Alaska at Spein Mountain(Ackerman, 1996), an AMS date on charcoal associ-ated with Mesa-type projectile points has a 1-sigmaerror ranging between 11,900 and 11,500 cal yr B.P.,which overlaps with ages from Gordon's Hearth(Fig. z). An outlier whose age does not match theages from Gordon's Hearth comes from the Putu sitein the northern Brooks Range (Reanier, 1995). Char-

Page 5: Date: 10/7/2007 9:17:16 PM

123Mann et al.: Environmental Change and Arctic Paleoindians

-32

-34

-36

-34

-4o

-42

Error in estimate of YD's end in GRIP core

Error in estimate of YD's end in GISP2 core

YD

Glacial

)Engigstciak 15,000

Hilltop

10,000

40

30

o

;eo

P(ao-&.C9

ag(UoO:@

o-oE=z

| 1 sigma error

! 2 sigma enor

010,000 11 ,000 13,000 14,000 15,000

coal fragments associated with edge-ground' lance-olate projectile points at Putu yielded an AMS date

of aeio -r 69 t+f yr B.P, (Reanier, 1995), whichwhen calibrated yields a 2-sigma age range between10,150 and gS0O cal Yr B.P.

In the Noatak River drainage southwest of the

Mesa site, reconnaissance work has yielded AMSdates ranging from 10,L00 to 95OO 'nC )ry-B.P. on

wood charcoal associated with oblanceolate proiec-

tile points (R. Gal, personal communication, 1998;

Rasii zooo). In addition, three AMS dates on wood

charcoal aL the Tuluaq Hill site range in age from

Holocene

[;l;;* -

]I o ]{v1niry1 l

Time (cal yr BP)aska and the Yukon. Hilltop and Bedwell sites are describedI the Engigstciak site by MacNeish (2000)' The delta 'nO;ecore/gie6ntand/summit/gripiisotopg:/gripdl so'Et; Johnscnffiates of the timing of the end of

of the YD has an error of t250 years, and estimates of its ti:the Mesa site are either contemporaneous with or younger tJthat was suggested by Alley et al' (1.993' 1997)' Pre-13,000 ciform a distfulct group ofoutliers. The present discussion focrthat the older ages cbincide with the Allerod chronozone, wGreenland.

Page 6: Date: 10/7/2007 9:17:16 PM

Gordon's HT

other MesaT

124

10,000 11,000 12.000

Time (cal yr B.P.)Figure 3. Radiocarbon ages ofcharcoal fragments fromGordon's Hearth superimposed over other post-13,000cal yr B.P. ages from the Mesa site. The histograms de-pict the number of AMS ages whose calibrated l-sigmaranges (Stuiver et al. 1ggB, zo0o) fall within a given de-cade. The 13 radiocarbon ages from this single hearthencompass the same 1.300-year interval spanned by mostof the other radiocarbon ages from the site. It is knor,vnthat rapid flucfuations in C-14 concentrations occurredin the atunosphere during the YD chronozone. Theseflucfuations cause numerous reversals and plateauswhen the radiocarbon ages of wood that was alive dur-ing the YD are plofted against the calibrated ages ofthesame samples. Radiocarbon ages from Gordon's Hearthindicate that the main occupation of the Mesa site couldhave been much shorter than 13OO vears.

1.1.,1,L0 to 11,200 'aC y' B.P. (Rasic 2000). The arti-facts from Tuluaq Hill and other sites in the Noatakvalley probably belong to a Mesa-type assemblage,though further work is required to confirm this rela-tionship.

Two AMS dates on charcoal from the Mesasite are distinctly older than the other 42 dates(Beta-57430, 11,190 -r 7014C )T B.P.and Beta-55286, 11,660 + 80 14C yr B.P.). The charcoal pro-ducing these two dates came from a single hearth ina locality that also contained other nearby hearthsdating to ca, 10,000 1aC yrs. B.P. (Kunz and Reanier1995:20), No distinctly different artifact types wereassociated with these older two radiocarbon dates.The technological uniformity of the site's artifacts(Kunz and Reanier 1994, 1995) suggests occupationby a single cultural goup. Several possible explana-tions exist for the presence of these two older dates.They could represent "old wood," wood interred infrozerr, fluvial deposits for many centuries beforeit was gathered by the Mesa people for firewood(Hamilton and Goebel 1999). Indeed, in geologicalstudies along the Ikpikpuk River, we encounteredburnable wood >10,000 years old that was pre-served in fluvial sections (Mann et al. 2002). A sec-ond explanation is that the two older dates record

Arctic Anthrcpologlr 3B:2

occupation by an earlier cultural group whose diag-nostic artifacts have not yet been found. Finallv, theolder dates could represent a pre-13,000 cal yr B.P.occupation of the Mesa site by people using thesame tool assemblage that they did almost 1500 cal-endar years later. At present, we ire unable to elimi-nate any of these hypotheses.

In summary, the available radiocarbon chro-nology suggests that Paleoindians occupied northemAlaska and the Yukon intermittently between13,800 and SZOO cal yr B.P. Evidence for the pre-13,000 cal yr B.P. ociupation of the region is frag-mentary in comparison with the period of occupa-tion between 12,4OO and 11,100 cal yr B.P. that is sowell-documented at the Mesa site. Except for thetwo pre-13,000 cal )T B.P. dates and the lone,younger outlier, the 1-sigma errors of AMS datesfrom the Mesa site fall within this 1300-year inter-val, as do the dates from Spein Mountain (Acker-man 1996), Enigstciak (MacNeish 2000), and theBedwell and Hilltop sites (Reanier 1995). The dateseries from Gordon's Hearth suggests that AMS ageswhose l-sigma errors fall between 12,4OO and11,100 cal y'r B.P. could represent Paleoindian occu-pation during a much shorter period of real time. Inthis paper, we ignore the hints of a pre-13,000 cal yrB.P. occupation of northern Alaska by Paleoindianpeople and concentrate on their well-dated use ofthe Arctic Foothills between 1.2,4OO and 11,100 calyr B.P. (Fig. Z). We suggest that Paleoindians disap-peared from northern Alaska before 9560 cal yr 8.P.,which is the lower limit of the Z-sigma, calibratedage range of the 8800 -f 60 14C

)'r B.P. date from thePutu site (Reanier 1995). The time interval of inter-est here for paleoenvironmental reconstructions isroughly 15,000 to 8000 cal y'r B.P.

Comparisons between Archaeology andIce-Core Records of Paleoclimate

Isotope, dust, snow-accumulation, and acidity rec-ords from the Greenland ice cores provide paleo-environmental type sections against which climaticevents in other parts of the world can be compared.The most intensive Paleoindian occupation of theMesa site occuned around the close of the YD, a1200-1300-yearJong climatic reversal with a strik-ingly rapid initiation and end (Alley, 2O0O). The ex-act relationship between the Paleoindian occupationof the Arctic and the cold, dry climatic conditionstypically ascribed to the YD is obscured by the pre-viously described difficulties in the radiocarbontime scale during this period and by inaccuracies inthe dating of the Greenland ice cores (Fig. 2). Datinguncertainties in the ice cores are around 2o/" aLIev-els corresponding to the Pleistocene to Holocenetransition (Alley et al. 1995; Meese et al. 1992;|ohnsen et al. 1997). Hence, even though the end of

fi20oo1s\t

d10b€5E=z0

Page 7: Date: 10/7/2007 9:17:16 PM

Mann et al.: Environmental Chanse and Arctic Paleoindians

the YD is knor,m to have been abrupt (Alley, 2000),age estimates for its end have errors of about +250calendar years. Perhaps the best estimate for the endof the YD is 11,640 -t 250 cal yr B.P. (Alley et al.1993; Alley et al. 1997). In Figure 2, the shaded barsshow the range of age estimates for the end of theYD in the GISP2 and GRIP ice cores, lntensive Pa-leoindian use of the Mesa site may have occunedimmediately after the YD as recorded in the GISP2core. The 1-sigma errors of all but the two youngestof the <11,000 14C

)T B.P. dates from the Mesa siteare compatible with this interpretation (Fig. 2).However, given the imprecision of radiocarbon agesduring the YD chronozone, it is also possible thatPaleoindians were using the Mesa site during thelatter half of the YD as well. Whether they werethere before, during, or after the YD's termination,the important point remains that Paleoindian occu-pation of the Arctic occurred during a period ofrapid changes in global climate.

Synthesis of Arctic FoothillsPaleoenvironmental HistoryWhat environmental changes were occurring innorthern Alaska when Paleoindians occupied theArctic Foothills? To answer this question, we usedthe stratigraphic records contained in fluvial, lacus-trine, hillslope, and peat deposits to infer how sur-ficial geology, soils, permafrost, and vegetationresponded to rapid climatic changes occurring be-tween 13,000 and 8000 laC g B.P. during the Pleis-tocene to Holocene (P-H) transition in the ArcticFoothills. Results of these studies are reported indetail elsewhere (Mann et al. 2002) and are onlysummarized here. They show that a series of sweep-ing changes occurred in land-surface processes andvegetation characteristics in parallel with the gen-eral trends in northern hemisphere climate de-scribed by the Greenland ice cores.

Central to our reconstructions of paleoenviron-mental changes in the Arctic Foothills are the de-posits of the Lake of the Pleistocene (LOP), aninformally named drained lake 20 km west of theMesa site in the Arctic Foothills (Fig. r). The sedi-ments of the LOP are exposed in a cutbank of theEtivluk River and contain lengthy records of lake-level arld vegetation changes (Mann et al. 2002). Weused sedimentary structures to infer lakelevel fluc-tuations, and pollen and spores to reconstruct vege-tation history, The history of lake-level fluctuationsindicated by the LOP sections coincides withchanges in vegetation, paludification, floodplain dy-namics, and solifluction in the surrounding region(Fig. +). The history of paludification was recon-structed by radiocarbon dating basal peats recoveredby augering through permafrost and by examiningstratigraphic sections in stream cuts. Floodplain dy-

12q

namics were described by studying the chronologyof terrace aggradation and erosion as revealed instratigraphic sections along streams draining theArctic Foothills. The history of solifluction, the slowdov,rrslope movement of water-saturated sedimentsresulting from the thaw of frozen ground, also wasdocumented in stratigraphic sections exposed instleam cuts.

Rising water levels in the LOP around 12,5OO1aC

I'r B.P. marked the beginning of a period ofraoid alluviation bv braided streams which lasteduntil -11,000 'nC yr B.P. Increased alluviation prob-ably was caused by a combination of increased per-mafrost melting and heightened hillslope erosion,which perhaps was triggered by increased summerrainfall on a landscape with discontinuous vegeta-tion cover (cf. Cogley and McCann 1986; Edlund etal . rsag).

Rising water levels in LOP around 12,5000 14C

)T B.P. also coincide with the spread of shrub tun-dra in the Arctic Foothills as indicated by pollenrecords from the LOP and Tukuto Lake (Oswald etal. rgg9). At the same time, peat deposition beganall across the North Slope (Fig. a), probably also inresponse to increasing effective moisture, a well-known trigger for paludification (Gorham 1991;Ovenden 1990). Effective moisture equals annualprecipitation minus annual evapotranspiration. Priorto 10,000 'aC y'r 8.P., peat deposition was limited totopographic low points in the Arctic Foothills, sug-gesting that effective moisture was lower than dur-ing the Holocene. Populus trees expanded theirrange between 12,000 and tt,OOO 'nC yr B.P. in re-sponse to warmer summers and the wider availabil-ity of recently deposited alluvium.

Elsewhere in northern and interior Alaska,there also is evidence for a major increase in effec-tive moisture -1,2,5OOlaC yr B.P. Glaciers in theBrooks Range underwent a minor re-advance be-tween -13,000 and tt,SOO 1aC years 8.P., possiblyin response to increasing winter snowfall (Hamilton,1986). On the Arctic Coastal Plain, the Ikpikpukdunes became inactive between -12,5OO and f f ,OOO1oC yr B.P., most likely in response to increased soilmoisture (Carter, 1993). During this same interval ininterior Alaska, wetter conditions led to the erosionof extensive gully systems on loess slopes (Hamiltonet al, 1988), and water levels at Birch Lake rose )18m between 12,700 and \2,200 1aC vr B.P. (Abbott etal. 2000).

During the YD, water levels fell in LOP andpaludification slowed in the Arctic Foothills (Fig. a).Streams incised their floodplains, probably becausea decline in effective precinitation reduced the in-put of slope sedimenti. Sediment input may havedeclined because of reduced solifluction activityand fewer landslides caused by permafrost melting.The distribution of Populus in the Arctic Foothillsseems to have retracted, perhaps in response to

Page 8: Date: 10/7/2007 9:17:16 PM

1.26

$oEaoEE+o-

E$,zo

FiSure 1' Synthesis of environrnental changes in the Arctic Foothills during the Pleistocene-Holocene transition (Mann etal.2o0z). Time scale in calendar years to allow comparison with the GISPZ, Greenland 6rsO record liom Grootes andStuiver.(1997). Histograms_depict_numbers of 'nC datbs whose calibrated 16 age range fatls within a given decade. Dateson stabilization of the Ikpikpuk dunes from Carter (1993). Data on stream geo'morpf,ology, solifluctio"n, f.rut f""t

"g*",fake level in Lake of the Pleistocene (LoP), and ages of Populus wood comi flo-'ruan11"J aL (2002).

cooler summers and to the shrinkage of its flood-plain habitat as streams became enlrenched. On theArctic Coastal Plain, dunes were reactivated after11,000 'nC p B.P. in response to decreased soilmoisture (Carter 1993).

Around 10,000 1aC yr B.P. water levels in LOpagain rose, and Popu.lus again expanded its distribu-tion in the Arctic Foothills. A brief episode of wide-spread solifluction occurred, probably in response rothe melting of ground ice that had accumulaled dur-

rains, deeper thawing of soils, and widespread soli-fluction. In northwestern Canada, numerous thaw

Nelson and Carter 1987; Ritchie et al. 1983).The LOP sections suggest that water levels

rose further after -8600 laC yr B.P. This observationis consistent with paleoclimatic findings elsewhere

in the region. On the Arctic Coastal Plain, increas-ing soil moisture again stabilized the Ikpikpukdunes -8500 '"C yr B.P. (Carter 1993). In interiorAlaska, water levels at Birch Lake rose markedly be-tween 8800 and 8000 14C

)T B.P. (Abbott et al. 2000;Edwards et al. 2001).

The LOP pollen record suggests that the mod-ern vegetation of the Arctic Foothills, along with itspoorly drained, peaty soils, probably *as

"--rt"b-Iished between 9000 and 8500 r4C r,n B.P. Other nol-len records from the North Slope (Eisner aldPeterson 1998; Oswald et al. 1999) are consistentwith this interpretation. The widespread establish-ment of organic soil horizons was a major turningpoint for ecosystem history in the Arctic Foothilli.Soil temperatures must have fallen and soil mois-ture risen because of the insulating and water-hold-ing properties of the organic horizons (Bockheim etal. 1998). These same organic surface horizons re-stricted the frost heaving of mineral material to theground surface and reduced soil erosion, thus de-priving streams of sediment inputs and forcing theminto a long-term trend of floodplain incision. Asfloodplains narrowed and became vegetated, the rateof loess deposition in dovrrnwind areas slowed,probably enhancing soil acidification and pro-moting further plaudification (Walker and Everett1991).

Arcl ic Anlhropology 38:2

10,000 12,000 14,000 16, 10,000 12,00a 14,000 16,Time (cal vr

Page 9: Date: 10/7/2007 9:17:16 PM

Mann et al.: Environmental Change and Arctic Paleoindians

Changes in Continentality Caused byFlooding of Alaska's Continental ShelvesShifts in regional and global climate drove many ofthe landscape changes just described from the Arc-tic Foothills. One of the most important of thesebackground changes was the flooding of Alaska'scontinental shelves by post-glacial, sea level rise.Today, Alaska is a peninsuli surrounded. by seas,and maritime air masses exert strong influences ontemperature and precipitation throughout the state.During the LGM, the present Alaska was embeddedin a much larger land mass created by the exposueof continental shelves by lower sea level lncreaseddistance from the ocean affected climate by chang-ing the continentality of many sites. The post-glacialflooding of the Bering Strait has long been identifiedas a probable cause for major climatic changes inBeringia (Anderson and Brubaker 1994; Cwynar1982; Elias et al. 1996; Guthrie 1990, 2001; Hopkins1.967, \982; Lozhkin et al. 1993). To date, no at-tempt has been made to quantify the effects ofchanging continentality on Alaska's climate duringthe flooding of its continental shelves.

Here we present a preliminary analysis of theeffects of post-glacial sea level rise on climate nearthe Mesa site. This analysis consists of six stepsstarting with the construction of depth/area crrrvesderived from modern bathymetric data (Fig. s).First, bathymetric data from the National Geophysi-cal Data Center's global bathymetry files were inter-polated within 5 X 5 minute areas in the Bering,Chukchi, East Siberian, and Beaufort Seas. Areas

127

were calculated using spherical coordinates. Bathy-metric data for depths <150 m then were sortedinto 5-m elevation bins for the entire Bering andChukchi Seas, the East Siberian Sea east of longi-tude 170' E, and the Beaufort Sea west of longitude140" W. We excluded from consideration the Pacificcontinental shelf of Alaska because of extensive LateWisconsin glaciation of the Aleutian Islands andAlaska Peninsula (Mann and Peteet 1994).

Second, we applied well-dated, eustatic, sealevel curves from tectonically stable and ice-sheetdistant parts of the globe to infer how post-glacialsea level changed on Alaska's continental shelves.There is no widelv aoolicable. eustatic curve forAlaska (Mann arrd ffa-itton 1995). In lieu of locallyderived sea-level histories, post-glacial sea level his-tories from tropical regions (Bard et al. 1990, 1996;Fairbanks 1989) are the next best thing, especially ifthey can be constrained locally by limiting dates.Fortunately, work in western Alaska and northwest-ern Canada (Elias et al. tggZ, 1996; Faas 1966; PeI-letier 1987; Jordan and Mason 1999) providelimiting dates that check the general applicability oftropical sea level curves to the Bering Strait region.The local dating control, though spa-rse, verifies thatthe tropical sea-level curves are indeed applicable tothe coastline of northem Alaska (Fig. 6).

The third sten involved combinins the eustaticsea-level curve witir the bathymetric daia from Alas-ka's marginal seas to reconstruct the history offlooding of the continental shelves. Similar resultsare presented in PALE (2000) at different time anddepth intervals. Fourth, we estimated the changing

0.2Area (km' x 10u)

o.4 0.6 0.8

-100

-120

0.80.6o.4

-60

E=ocoC)(JL

oE-c.$d)

-1.25

-20

-40

-80

-100

-120Figure 5. Hypsometry of the Alaskancontinental shelves today. The Arcticshelfincludes the Chukchi Sea, theBeaufort Sea west of 140' W, and theEast Siberian Sea east oflongitude170" E.

Bering shelf

Arctic shelf

Page 10: Date: 10/7/2007 9:17:16 PM

128 Arctic Anthropology 3 B :2

8-76676 8-80205

-80

-100

-1205000 10,000 15,000 20,000

Time ('oC yr B.P.)Frgure 6. The Barbados sea level curve of Fairbanks (1g8g) and Bard et d. (1996) compared with maximum-limiting ra-diocarbon ages on sea level changes on the continental shelves offnorthern and western Alaska (Faas 1966; Elias et al.1996; [ordan and Mason 1999). Data from the Mackenzie River delta from Pelletier (1ss7). Dates produced by Beta-Analytic Incorporated are abbreviated as "B-."

0E

-20oal 40

J

E -ooao

(Uot

0

distances between the Mesa site and the Chukchiand Beaufort Sea coasts as sea level rose (Fig. z).

The fifth step was to establish modern rela-tionships between climatic factors and distancefrom the sea. Regressions between luly mean tem-perature and distance to the ocean show that inte-rior sites (high continentality) are warmer in

7000 9000 11,000 13,000 15,000

Time (1aC yr B.P.)Figure 7. Changes in the area of the emergent continen-tal shelves bordering northern Alaska during the latePleistocene and early Holocene.

summer than coastal ones (Figs. 8, 9). July tempera-tures increase rapidly in the first 50 km from thecoastline then more slowly for 100s of km inland.Transects inland from the southwestern Bering Sea,the Chukchi Sea, and the Beaufort Sea reveal simi-lar relationships between distance inland and ]ulytemperature.

The relationship between distance from thesea and mean annual precipitation is more compli-cated. At weather stations located along a broadtransect from Sand Point in the western Gulf ofAlaska to Fort Yukon in the eastern interior (Fig.10), precipitation declines rapidly over the first sev-eral hundred km from the coastline, then decreasesmore slowly for 1800 km inland. In contrast, on theNorth Slope, precipitation increases markedly in-Iand from the Beaufort Sea over a distance of ZSOkm to the northern front of the Brooks Ranse(Zhang et al. 1996). Most of the precipitatio"n on theNorth Slope, however, falls in July, August, andSeptember and comes from maritime air masses ar-riving from the Bering Sea region (Moritz 1979).Hence the Beaufort Sea is not acting as the primarymoisture source; rather sites on the North Slope areat the extreme interior end of a moisture-transportsystem originating in the Bering Sea and North Pa-cific.

The final step in our analysis was to applythese modem relationships to the changing dis-

ococco^obf roxO)N( t )Ec-xtJJ =sO0-ctt u)oL

usGs-758Y

AA.

AB-3034 vAA-14676

843953v t B+ggsz

y maximum limiting date (tenestrial peats' on Alaskan continental shelves)

A minimum limiting date (detrital organicsdeposited ofi Mackenzie delta ftont)

a peats near mean higher high wateron Seward Peninsula (Jordan and

sea level

:*";:.T'

Arctic shelf O

Bering SheF f

Page 11: Date: 10/7/2007 9:17:16 PM

Mann et aL: Environmental Chans.e and Arctic Paleoindians

Nunivak-Ft Yukon stations

600

rChukchi-Beaufort coast and N Slope stations tlt t;t

129

.stR

A11

^18o3-E14=(E

8. roE.oc6(Uo

=

-

10 l*

y = 6.6x0 12

12 = O.g2 |

y = 2.gxo'25

r2=0.89

tances to the coastline after 18,000 yearc ago, and toestimate the contribution that the flooding of AIas-ka's continental shelves had on climatic change nearthe Mesa site. In this very preliminary analysis, wecalculate changing distance to the sea along straightIines drawn northwestward and southwestwardfrom the Mesa site. The bathymetric and sea levelhistory data indicate that flooding of the continentalshelves was most rapid between 10,000 and ASOO"C yr B.P. (Fig. z). Flooding was especially rapidduring this interval in the Chukchi, Beaufort, andEast Siberian basins due to the occurrence of exten-sive shelf areas there that Iie between -60 and -40m (Fig. 5). Regression equations relating modernfuly temperature to distance from the sea along a

line stretching northwest from the Mesa site into theChukchi Sea suggest fuly temperatures at the Mesasite fell 1-1.soc between 10,000 and 9000 'nC yrB.P. in response to rising sea level (Fig. 11). After9000 14C yr 8.P., cooling rates decreased as the ap-proach of the ocean slowed. When distance to thesea is calculated along a line stretching southwestfrom the Mesa site into the Bering Sea, coolingstarts earlier, ca. 12,000 taC yr B.P. and achievesmaximum rates 11,000-8500 'nC yr B.P., with a to-tal of 1-3o C of cooling predicted between 12,000and 6000 laC yr B.P. In contrast, the decline in sum-mer temperature caused by marine transgressionalong the Beaufort Sea coast was probably minor be-cause of the narrowness of this continental shelf.

16 18

t1s '17Figure 9. The modern gradientin |uly mean temperafure fromcoastal Alaska inland along awest to east hansect of stations.1, Wales; I Shishmaref;3, Ki-valina;4 Kotzebue; 5, Candle;4 Noorvik; Z Selawik; B, Am-bler;9, Shungnak; 14 Kobuk;11, Hughes; 12, Allakaket; 13,Bettles; 14, Rampart; 15, Vene-tie; fO, Fort Yukon; 1Z Cenhal;1& Circle. Climate records fromLeslie (198s).

6ra

Ea14EoCLE10ptr6oo=->q2

8..9

200 400 600 800 1000

Distance (km) east of 169o 30'W

Distance (km)Figure B. The upper curve shows the gradient in luly mean temperature at Alaskan stations along a coast to interiorhansect stretching northeastwards from Nunivak Island. A, Nunivak Island; fl Hooper Bay; C Emmonak; D, Bethel; 4Mountain Village; fl St. Mary's; G Russian Mission; II, Aniak; d Holy Cross; /, Sleehnute; K McGrath; .L, Lake Min-chumina; M, Manley Hot Springs; N, Nenana, O, Fairbanks; P, Central; Q, Venetie; R, Circle; S, Fort Yukon. Climaterecords from Leslie (1989). Bottom line (triangles) shows stations at different distances inland from Beaufort and Chuk-chi Seas. 1, Barrow; 2, Point Hope; 3, Point Lay; 4, Barter Island; 5, Wainwright; 6, Kuparuk; 7, Lonely; B, FranklinBluffs; 9, Umiat, IQ Happy Valley; 11, lmnavait Creek; 72, Toolik Lakei 73, Galbraith Lake. Climate records from Leslie(1eBe).

1200

Page 12: Date: 10/7/2007 9:17:16 PM

L30

2.a

4

y = 307,997 x-oser2 = 0.74

Some of the greatest impacts of lowered sealevel on the climate of Beringia probably camethrough its effects on storm tracks, and this prelimi-nary analysis describes these effects only indirectly.Today, most precipitation in Alaska north of theAlaska Range comes in summer and autumn fromcyclonic storms originating in the northwest PacificOcean (Hare and Hay 1,974; Kane 1996; Moritz1979), The penetration of Pacific stoms into Alaskais markedly seasonal today and probably was alsoin the past (Bryson and Hare 1,974).In winter andspring when sea ice covers much of the Chukchiand Bering Seas, strongly zonal circulation steersPacific stoms south of mainland Alaska across theGulf of Alaska (NASA/GISS, 1999). During theseseasons, high pressure over the Arctic Basin causesnortheasterly flow across Alaska's North Slope. Cir-culation tends to reverse in summer and autumnwhen ice-free conditions permit meridional circula-tion to develop over the Bering Sea. Pacific cyclonesmove northeastward along this trough to penehatethe interior of the state (Hare and Hay 1974; Mock

0 50 100 150 200 250

Distance (km) inland from Beaufort Sea

Arctic Anthropology 3 I :2

p ssoE

a1

EE- 12ooco(U=o.'6E Boo(L

I7.r t )

(!:fCc

I-(Uo

29 t33

600 1000 1400 1800

Distance (km) northeastward of Sand Point, Gulf of Alaska

Figure 1O. The modern precipitation gradient between Bristol Bay in southwest Alaska and the Beaufort Sea. 1, CapeNewenham; 2, Dillingham;3, Dillingham 17 NW;4, Naknek; 5, Aleknagik; 6, Aniak; Z Sleetnute; B, Stony River; 9,McGrath; lQ Farewell; 11, Galena, 12, Ruby, 13, Rampart 2;1.4, Clear;15, Tanana; 16, Manley Hot Springs; 17, Hughes;14 Indian Mountain; 19, Fairbanks; 20, Allakaket;27,Livengood;22, Bettles; 23, Central; 24, Circle Hot Springs; ZS,Circle; 26, Cold,Foot; 27, Wiseman, 28, Venetie, 29, Fott Yukon; 34 Anaktuvuk; 31, Dietrich; 32, Chandalar Lake; 33,Urniat; 34, Arctic Village; 35, Coleen River; 36, Barrow; 3Z Barter Island. Data from Leslie (1g8g). Inset shows precipita-tion on North Slope redrawn from Zhang et al. (1996). B, Barrow; P, Prudhoe nay; E Sagwon; T, Toolik; A, Atigun

-

Camp. Climate records from Leslie (1S89).

at36 37

et al. 1998; Moritz 1979), North of the BrooksRange, additional surruner precipitation originatesfrom rather weak cyclones generated along the Arc-tic front over Russian sectors of the Arctic Ocean.

Exposure of the continental shelves during theice age probably affected storm tracks in severalways. Dry land, like sea ice, deprives cyclonicstorms of the thermal energy otherwise availablefrom the underlying water (Lamb 1972). The EastAsian Trough was probably less developed whenthe continental shelves were exposed, and cyclo-genesis in Russian sectors of the Arctic basin wouldhave been lessened. It is likely that the Arctic front(Krebs and Barry 1970) was shifted south of theBering Strait in surnmer to lie along the southernedge of the exposed Bering Sea shelf where thetemperature gradient between sea and land wassteepest. The Arctic front would have steered Pa-cific storms eastwards and away from higher lati-tudes,

By ignoring the effects of changing stormtracks, our results undoubtedly miss some of the

24128

Page 13: Date: 10/7/2007 9:17:16 PM

'g'acoE;.cl 5;sEOGCt3EEEoFEO

"9=.9>! r=E5o-

Mann et al.: Envfuonmental Chanse and Arctic Paleoindians 131

A Beaufort Sea effect using modern North Slope gradient

B Bering Sea effect using Nunivak-Fort Yukon gradient

G Chukchi Sea effect using Nunivak-Fort Yukon gradient

D Chukchi Sea effect using North Slope gradient

E Bering Sea effect using \Ahles-Fort Yukon gradient

F Chukchi Sea effect using modern Wales-Fort Yukongradient

2000 4000 6000 8000 10,000 12,000 14,000 16,000

Time ('oC yr B.P.)

Figure 11. Cumulative changes in fuly temperafure near the Mesa site that could have occurred in response to floodingof the continental shelves. Predictions are based on regressions between modern temperafure and distance-to-coastshown in Figures B and g.

-2

most dramatic charges in climate caused by chang-ing continentality. Though similar in sign, the pre-cipitation changes we predict here for the NorthSlope probably underestimate the total impacts ofchanging continentality on precipitation.

This said, we use the modern relationship be-tween distance to coast and mean annual precipita-tion to predict changes in precipitation at the Mesasite caused by the approach of the Bering Sea. Thechanging distance to the sea is measured along aIine drawn southwest across the central SewardPeninsula to St. Matthew Island in the centralBering Sea. Precipitation first increased ca. 12,5001aC yr 8.P., then increased very rapidly 10,000-80001aC yr B.P. (Fig. r2). The total predicted increase inmean annual precipitation is 300 mm between theLGM and the mid-Holocene. This figure is a crudeestimate. Nonetheless, remembering that mean an-nual precipitation is only 200-300 mm over muchof the North Slope today, it suggests that changingcontinentality had a major impact on the moistureregime of the North Slope during the Pleistocene toHolocene transition.

Speculations and DiscussionWere Arctic Paleoindians Dependent onEnvironmental Disturbance?

It would be difficult to find a time period containingmore radical shifts in climate and biota than the

several millennia years spanning the Pleistocene toHolocene transition; the remainder of the Holocenewas uneventful by comparison. Global climatejumped from interglacial conditions to glacial con-ditions during the YD (fohnsen et al. 1997; Alley2000; Elias 2000), then back to interglacial condi-tions within less than a decade around 10,OOO 14C

)T B.P. (Alley et al. tggg; Alley 2000; Isarin andBohncke 1999; Mayewski et al. 1993). Widespreadextinctions affected the world's megafauna duringthis interval (Martin and Klein 1984; Guthrie 1990).As described earlier, the flooding of Alaska's conti-nental shelves, especially after 10,000 1aC

1'r 8.P.,probably had a great impact on Alaskan climate.Vegetation chalges repeatedly swept through north-ern Alaska between ca. 14,000 and O0O0 1aC yr B.P.(Anderson and Brubaker 1994). Shifts in climateand vegetation disturbed soils by altering water ero-sion, solifluction rates, floodplain extent, and loessdeposition.

Why did the Paleoindians utilize arctic Alaskaduring an interval of repeated, widespread, andrapid environmental disturbances? The answer maybe that their prey species flourished intermittentlyduring alternations in ecological disturbance re-gimes that occurred within this period of rapid cli-matic change. Any discussion about the ways inwhich environmental factors controlled human his-tory in northern Alaska 10,000 years ago devolvesto a discussion about the effects of the environmenton the large-mammal prey species. The climate of

Page 14: Date: 10/7/2007 9:17:16 PM

132

8000 12,000 16,000

Time (1oC yr B P.)

Arctic Anthropology 3 B :2

Figure 1.2. Cumulative changesin precipitation near the Mesasite that could have occurred inresponse to flooding ofthe conti-nental shelves. These predictionsare based on the modern rela-tionship between distance-to-coast and precipitation shown inFigure 10.

eE

ovE)Ccon' 'F-cGoEEB€dEEo-dotr.z<6c=ola5>o.s

350300

200

100

0

northern Alaska is harsh, and potential food sourcesare geographically and seasonally limited, especiallyfor people not using marine resources. Ethnographicand archaeological records show unequivocally thatin interior (non-coastal) arctic landscapes of NorthAmerica, aboriginal people relied on the hunting oflarge ungulate prey species to survive (Meltzer andSmith 1986; Dumond and Bland 1995). In north-ern Alaska and northern Canada, caribou were themainstay of interior Inuit populations (Spencer,1959; McGhee, 1996), and musk oxen were vital tothe episodic occupations of the world's northern-most point, Peary Land in northern Greenland, dur-ing the Middle and Late Holocene (Knuth 1,s66167).

During the transition from Pleistocene to Holo-cene, bison was the most likely primary prey spe-cies of Paleoindians ranging north of the BrooksRange. Bison priscus were butchered 9800-9400 14C

)T B.P. at the Engigstciak site 500 km east of theMesa site (Cinq-Mars et al. 1991) by people using atool assemblage similar to that used at the Mesa site.The radiocarbon ages of a large series of bones fromthe Ikpikpuk River indicate that bison were pres-ent on the North Slope during the interval 13,000-10,000 1aC p 8.P., but that caribou were rale duringthis time (P. Matheus, personal communication). Onthe Great Plains, the people of the Agate BasinPaleoindian complex hunted bison (Frison andStanford 1982), utilizing a stone tool assemblagevery similar to that employed at the Mesa site (Kunzand Reanier 1994; Bever, this volume). The appar-ent chronological synchrony between Agate Basinand Mesa complexes (Holliday 2000; Kunz andReanier 1994, 19S5) suggests these people dispersedvery rapidly across continental distances, whichalso argues for a hunting economy specialized on asingle widespread species (Kelly and Todd 19BB).Bison, with their far-ranging migratory behavior andsubsequent intercontinental distribution (Guthrie

20,000

1990), seems the most likely candidate for this preyspecies. Natural populations of bison are absentfrom Alaska today, though several small, introducedpopulations survive on the geomorphologically dis-turbed floodplains of the Delta, Kuskokwim, andChitina Rivers in the interior, the most climaticallycontinental region of Alaska.

The very climatic instability that characterizedthe Pleistocene-Holocene transition may have al-Iowed bison populations to flourish in northernAlaska during this period. Steppe bison ate mainlygraminoids, especially grasses (Guthrie 1990, 2001).In boreal regions where forests are the climax vege-tation today, graminoid species capable of support-ing bison often are abundant in early successionalplant communities. Tree arrival into northern re-gions lagged behind deglaciation by anywhere fromcenturies to millennia (MacDonald and Cwynar1985), and grasses and sedges would have beenabundant components of the interim vegetation. To-day in the Arclic Foothills, grasses atu ib,-darrtwhere soils are disturbed frequently by animal dig-ging, solifluction, stream erosion, or loess deposition(Walker et al. 2001). The species diversitv and pri-mary produclivity of grasses are highest where ioildisturbance prevents organic accumulation andmaintains thick active lavers. Todav in the ArcticFoothills, grasses are important components of vege-tation on streamside bluffs, active sand dunes, andrecently stabilized floodplains (Walker and Everett1991). By repeatedly destabilizing soils and inter-rupting ecological succession, climatic changes mayhave favored bison habitat in northern Alaska, espe-cially under the continental climate regimes thatpreceded flooding of the continental shelves.

It also is possible that musk oxen were an im-portant prey species for the Mesa people. Musk oxbones are abundant in Pleistocene- and Holocene-aged deposits along the Ikpikpuk River; though

4000

Page 15: Date: 10/7/2007 9:17:16 PM

Mann et al.: Environmental Change and Arctic Paleoindians

these bones are currently undated. Like bison, muskoxen undergo explosive population growth duringbrief periods of favorable climate (Jingfors and Klein1982). AIso like bison, the optimal forage of muskoxen is provided by early successional plants grow-ing in geomorphically disturbed sites (Forchhammerand Boomsma 1995; Klein and Bay 1984).

Why did Paleoindians abandon northernAlaska? The a-nswer probably lies with the increas-ing climatic stability that accompanied the onset ofthe present interglacial, decreasing continentality,and the resultant spread of tussock-tundra (moist

ally and spatially heterogenous landscape, highmobility is important both for large ungulates (flein1999) and for hunter-gatherer groups (Mandryk1993). Human foot travel across tussock tundra isvery difficult in summer. Bison are small-footed,dry-ground grazers who cope with seasonal changesin grazing conditions by making long migrations.The location of the Mesa, Putu, Bedwell, and Hill-top sites on lookout points along the northern flankof the Brooks Range may be related to seasonal bi-son migration routes through these mountains. Be-

Another attribute of tussock tundra that is hos-tile to both bison and humans is the huge popula-tions of mosquitoes that breed there in numerous

as they certainly do on humans today. The swarmsof mosquitoes that plague the Arctic Foothills insummer today probably accompanied the establish-ment of the tussock tundra.

Acknowledgmenfs. We thank Charles and ConnieAdkins, L. Plug, K. Tae, P. Heiser, E. Dillingham, J.Chase, J. Dub6t, T. Reanier, S. Durand, ald H. Mac-donald for their help in the field. A. Krumhardt ard

P ft_ood provided invaluable laboratory support. Wethank D. Murray, C. Parker, and A. Ba[ten lor helpwith macrofossil identifications; O. Mason for refer-ences; and D. Meltzer, M. Bever, P. Groves, D. Hop-kins, and S. Walker for valuable discussions. The-Bureau of Land Management provided funding for

133

support of paleoenvironmental research at the Mesaarchaeological site.

End Note1. Radiocarbon and calendar years diverge widelyin the early Holocene and Late Glacial. To describethe interrelationships between the chronologies thatare the subject of this study, it is necessary to citeboth C-f+ and calendar years. We express uncali-brated C-14 ages as "laC )T 8.P." Dates that are cali-brated to calendar ages ile expressed as "cal yrB.P."

ReferencesAbbott, M. B., B. P. Finney, M. E. Edwards, and K. R.Kelts2000 Lake-level Reconstructions and Paleohydrology of

Birch Lake, Central Alaska, Based on Seismic Re-flection Profiles and Core Transects. QuaternaryResearch 53:154-166.

Ackerman, R. E.1996 Spein Mountain. In Anre.ncan Beginnings, edited

by F. H. West, pp. 456-460. University of Chi-r;ago Press, Chicago.

Alley, R. 8., A. I .Gow, S. f . fohnsen, f . Kipfstuhl, D. A.Meese, and T. Thorsteinsson1995 Comparison of Deep Ice Cores. Nature 323:393-

394.

Alley, R. 8., D. A. Meese, C. A. Shuman, A. J. Gow, K. C.Taylor, P. M. Grootes, ] . W.C. White, M. Ram, E. D. Wad-dington, P. A. Mayewski, and G. A. Zielinski1993 Abrupt Increase in Greenland Snow Accumlrla-

tjon at the End of the Younger Dryas Event. No-ture 362:527-529.

Alley, R, B., C. A. Shuman, D. A. Meese, A. f . Gow, K. C.Taylor, K. M. Cuffey, J. f. Fitzpatrick, P. M. Grootes, G. A.Zielinski, M. Ram, G. Spinelli, and B. C. ElderL937 Visual-Stratigraphic Dating of the GISP2 Ice Core:

Basis, Reproducibility, and Application. lournalof Geophysical Research i.o2:26,367 -26,381..

Alley, R. B.2000 The Younger Dryas Cold Interval as Viewed from

Central Greenland. Quaternary Science Reviews'1.9:213-226.

Anderson, P. M.1988 Late Quaternary Pollen Records from the Kobuk

and Noatak River Drainages, NorthwesternAlaska. Quaternary Research 29:263 -27 6.

Arderson, P. M. and L. B. Brubaker1994 Vegetation History of Northcentral Alaska: A

Mapped Summary of Late-Quaternary PollenData. Quofernary Science Reviews 1-371-92.

Page 16: Date: 10/7/2007 9:17:16 PM

134

Arnborg, L., H. I . Walker. and ]. Peippo1966 Water Discharge in the Colville River, 1962. C,eo-

grafi ska Annaler 4BA1,95 -21,O.

Bard, E., B Hamelin, arld R. G. Fairbanks1990 U-Th Ages Obtained by Mass Spectrometry in

Corals from Barbados: Sea Level during the Past130,000 Years. Nolure 346:456-458.

Bard, E., B Hamelin, M. Arnold, L. Montaggioni, G.Cabioch, G. Faure, and F. Rougerie1996 Deglacial Sea-level Record from Tahiti Corals and

the Timing of Global Meltwater Discharge. Na-ture 382:241-244.

Bjorck, S., B. Kromer, S. fohnsen, O. Bennike, D. Ham-marlund, G. Lendahl, G. Possnert, T. L. Rasmussen, B.Wohlfarth, C. U. Hammer, and M. Spurk1996 SynchronizedTerrestrial-AtmosphericDeglacial

Records Around the North Atlantic. Science 274:1 1 55-1 160.

Bockheim, I. G., D. A. Walker, L. R. Everett, F. E. Nelson,and N. I. Shiklomanov1998 Soils and Cryoturbation in Moist Nonacidic and

Acidic Tundra in the Kuparuk River Basin, Arc-tic Alaska, LISA. Arcfic and Alpine Research 3O:1.66--174.

Bryson, R. A. and F. K. Hare1,974 The Climates of North America. ln World Suruey

ol Climatology, edited by R. A. Bryson and F. K.Hare, pp. 1-48. Climates of North America, Vol.11. Elsevier Scientific. New York.

Burn, C. R.1.597 Cryostratigraphy,Paleogeography,andClimate

Change during the Early Holocene Warm Interval,Western Arctic Coast, Canada. Canadian lournalol Earth Sciences 34:91,2-925.

Burn, C. R., F. A. Michel, and M. W. Smith19BG Stratigraphic, Isotopic, and Mineralogical Evi-

dence for an Early Holor;ene Thaw Unconformityat Mayo, Yukon Territory. Canadian lournal ofEorth Sciences 23:794-8O3.

Carter, L. D.1981 A Pleistocene Sand Sea on the Alaskan Arctic

Coastal Plain. Science 211:381-383.

1993 Late Pleistocene Stabilization and Reactivation ofEolian Sand in Northern Alaska: Imolications forthe Effects of Future Climatic Warming on anAeolian Landscaoe in Continuous Permafrost. InProceedings, Siih lnternational Conference onPermoJrost. Beijing:, China, YoI. 1, pp. 78-83.South China University Technology Press,Wushan.

Carter, L. D., L. A. Heginbottom, and M-K. Woo1987 Arctic Lowlands. In Geomorphic Systems ol

North Americd, edited by W. L. Graf, pp. 583-628. Geoiogy of North America, Yol. Z, Geologi-cal Society of America, Boulder.

Arctic Anthropology 3 B : 2

Cinq-Mars, ]., C. R.Harington, D. E. Nelson, and R. S.MacNeish199L Engigstciak Revisited: A Note on Eaily Holocene

AMS Dates from the "Buffalo Pif. " Occasional Pa-per No. 1, Canadian Archaeological Association.

Cogley, l. G. and S. B. McCarn"1976 An Exceptional Storm and Its Effects in the Cana-

dian High Arctic. Arcfi'c and Alpine Research B:105-110.

Cwynar, L. C.1.982 A Late-Quaternary Vegetation History fiom Hang-

ing Lake, Northern Yukon. Ecological Mono-gruphs 52:'1,-24.

Dinter, D. A., L. D. Carter, and /. Brigham-GretteL990 Late Cenozoic Geologic Evolution of the Alaskan

North Slope and Adjacent Continental Shelves.In The Arctic Ocean Region, edited by A. Grantz,L. ]ohnson, and J. F. Sweeney, pp. 459-490. Geo-logical Society of America, Boulder.

Dumond, D. E. and R. L. Bland1995 Holocene Prehistory of the Northernmost North

Pacific. lournal of World Prehistory 9:401-450.

Edlund, S. A., B. T. Alt , and K. Young1989 Interaction of Climate, Vegetation, and Soil Hy-

drology at Hot Weather Creek, Fosheim Penin-sula, Ellesmere Island, Northwest Territories.Geological Sttrvey of Canada, Paper B9-1D, pp.' t25-1,33.

Edwards, M. E. ard P. W. Dunwiddie1985 DendrochronologicalandPalynological Observa-

tions on Populus balsanifera in Northern Alaska.Arctic and Alpine Research 17:271.-278.

Edwards, M. E., C. J. Mock, B. P. Finney, V. A. Barber,and P. J. Bartlein2OO'1. Potential Analogues for Paleoclimatic Variations

in Eastern Interior Alaska During the Past 14,000Years. Quof e.rn a ry S c ie n c e R evi ew s 2O :1,85 -2O2.

Eisner, W. R. and K. M. Peterson1998 High-resolution Pollen Analysis of Tundra Poly-

gons from the North Slope of Alaska. lournal ofGeophysical Research 1,o3:28,929-28,937.

Elias, S. A.2000 Late Pleistocene Climates of Beringia, Based on

Analysis of Fossil Beetles. Quatet'nary Research53:229-235.

Elias, S. A., S. K. Short, and R. L. Phillipslggz Paleoecology of Late-Glacial Peats from the

Bering Land Bridge, Chukchi Sea Shelf Region,Northwestern Alaska. Quaternary Research 38:371,-378.

Elias, S. A., S. K. Short, C. H. Nelson, and H. H. Birks19S6 Life and Times of the Bering Land Bridge. Nature

382:60-63.

Everett, K. R. and I. Brown1,SBZ Some Recent Trends in the Physical ald Chemi-

Page 17: Date: 10/7/2007 9:17:16 PM

Mann et aI.: Environmental Change and Arctic Paleoindians

cal Characterization and Mapping of TundraSoils, Arctic Slope of Alaska. Soil Science \33:264-280.

Faas, R. W.1S66 Paleoecology of an Arctic Estuary. Arctic 1,9:343-

348.

Fairbanks, R. G.19Bg A 17,000-year Glacio-eustatic Sea Level Record:

Influence of Glacial Melting Rates on YoungerDryas Event and Deep-ocean Circulation. Nature342:637-642.

Ferrians, O. J.1994 Permafrost in Alaska. h The Geology of Alaska,

edited by G. Plafker and H. C. Berg, pp. 845-854.Geological Society of America, Boulder.

Forchhammer, M. C. and I. ]. Boomsma1995 Foraging Strategies and Seasonal Diet Optimiza-

tion of Muskoxen in West Greenland. Oecolosia104:169-180.

Frison, G. C. and D. |. Stanford1982 The Agate Basin Site, A Record of the Paleo-

indion Occupation of the Northwestern HighPlains. Academic Press, New York.

Galloway, f. P. and L. D. Carter1993 Late Holocene Longitudinal and Parabolic Dunes

in Northern Alaska: Preliminarv Interoretations ofAge and Paleoclimatic Significince. tlnited StatesGeological Survey Bulletin 2068:3-11.

Gorham, E.1991 Northern Peatlands: Role in the Carbon Cycle and

Probable Responses to Climatic Warming. Ecolog-;^^t a ̂ ^ t : -^t :^rc .L:82_.Lg' ,lour t rPPl lvuLtul

Goslar, T., M. Arnold, E. Bard, T. Kuc, M. F. Pazdur, M.Ralska-lasiewiczowa, K. Rozanski, N. Tisnerat, A. Wala-nus. B. Wicik. and K. Wieckowski1995 High Concentrations of Atmospheric 'oC during

the Younger Dryas Cold Episode. Nature 377:414-41.7.

Goslar, T., M. Arnold, N. Tisnerat-Laborde, J. Czernik, andK. Wieckowski2000 Variations of Younger Dryas Atmospheric Radio-

carbon without Ocear Circulation Changes. ly'o-ture 4O3:877-BBO.

Grantz, A., S. D. May, and P. E. Hart1994 Geology of the Arctic Continentai Margin of

Alaska. ln The Geology of Alaska, edited by G.Plafker and H. C. Berg, pp. 1,7-48. Geological So-ciety of America, Boulder.

Grootes, P. M. and M. Stuiver1.9s7 Oxygen 18116 Variability in Greenland Snow and

Ice with lOs to 105-Year Time Resolution. lournalof Geophysical Research 1.o2:26,455-26,47 o

Guthrie, R. D.1990 Frozen Fauna of the Mammoth Steppe. Univer-

sity of Chicago Press, Chicago, Illinois.

L35

zoo'J. Origin and Causes of the Mammoth Steppe: AStory of Cloud Cover, Wooly Mammoth ToothPits, Buckles, and Inside-out Beringia. QuaternaryScience Reuiews 2O:549-57 4.

Hajdas, L, G. Bonani, P. Bod6n, D. M. Peteet, and D. H.Mann1998 Cold Reversal on Kodiak Island, Alaska, Corre-

Iated with the European Younger Dryas UsingVariations of Atmospheric 'nC Contenl. Geology26:'1,O47-1,O5O.

Hamilton, T. D.1986 Late Cenozoic Glaciation of the Central Brooks

Range. ln Glaciation in Alaska, edited by T. D.Hamilton, K. M. Reed, and R. M. Thorson, pp. 9-50. Alaska Geological Society, Anchorage.

Hamilton, T. D. and T. Goebel1999 Late Pleistocene Peoplings of Alaska. In lce Age

People of Nofih America, edited by R. Bonnich-sen and K. L. Turnmire, pp. 156-199. OregonState University Press, Corvallis.

Hamilton, T. D., I. L. Craig, and P. V. Sellman19BB The Fox Perma-frost Tunnel: A Late Quaternary

Geologic Record in Central Alaska. Geological So-ciety of Ameilca Bulletin 100:948-969.

Hare, F. K. and j. E. HayL974 The Climate of Canada and Alaska, In Climates

of North America, Vol rr, World Suwey of CIi-matologlr, edited by R. A. Bryson and F. K. Hare,pp. 49-192. Elsevier Scientific, New York.

Heathwaite, A. L., E. G. Gottlich, E. G. Brumeister, G.Kaule, and T. Grospietsch1993 Mires: Definitions and Forms. In Mires: Process,

Exploitation, and Consentadon, edited by A. L.Heathwaite, pp. 1-76. John Wiley, New York.

Hinzman, L. D., D. L. Kane, C. S. Benson, and K. R.Everett1996 Energy Balance and Hydrological Processes rn an

Arctic Watershed. In Landscaoe Function andDisturbance in Arctic Tundra', edited by f. F.Reynolds and J. D. Tenhunen, pp. 1,31,-1,52.Springer Verlag, New York.

Holliday, V. T.2000 The Evolution of Paleoindian Geochronology and

Typology on the Great Plains. Geoarchaeology 1-5227-25O.

Hopkins. D. M.1959 Some Characteristics of the Climate in Forest and

Tundra Regions of Alaska. Arctic 12:214-22O.

1967 The Cenozoic History of Beringia-A Synthesis.ln The Bering Land Bridge, edited by D. M. Hop-kins, pp. 451,-4A4. Stanford University Press,Stanford.

L982 Aspects of the Paleogeography of Beringia duringthe Late Pleistocene. In Paleoecology of Beringia,edited by D. M. Hopkins, J. V. Mathews, C. E.

Page 18: Date: 10/7/2007 9:17:16 PM

136

Schweger, and S. B. Young, pp. 3-28. AcademicPress, New York,

Isarin, R. F. B. and S. J. P. Bohncke1999 Mean JuIy Temperatures during the Younger

Dryas in Northwestern and Central Europe as In-ferred from Climate Indicator Plant Species. Quo-ternary Research 51:158-1 73.

Jingfors, K. and D. Klein7SB2 Productivity in Recently Established Muskox

Populations in Alaska. lourna) of Wildlife Man-agement 46:1092-1096.

fohnsen, S, I . , H.B. Clausen, W. Dansgaard, N. S. Gunde-strup, C. U. Hammer, U. Andersen, K. K. Andersen, C. S.Hvidberg, D. Dahl-lensen, |. P. Steffensen, H. Shoji, A. E.Sveinbiornsdottir, J, White, f. louzel, and D. FisherL997 The Delta O-18 Record along the Greenland Ice

Core Project Deep Ice Core and the Problem ofPossible Eemian Climatic Instability. lournal ofGeophysica) Research 1o2:26,397 -426,41o.

Jordan, J. W. and O. K. Mason1999 A 5000-Year Record of Intertidal Peat Stratigra-

phy and Sea Level Change from NorthwestAlaska. Quaternary International 6o:37 -47 .

Kane, D.L.1996 The Impact of Hydrologic Perturbations on Arctir;

Ecosystems lnduced by Climate Change. InGlobd Change and Arctic Terrestrial Ecosystr:ms,edited by W. C. Oechel, T. Callaghan, T. Gilma-nov, f. I. Holten, B. Muwell, U. Molau, and B.Sveinbjornsson, pp. 63-81, Springer Verlag, NewYork.

Kane, D. L., L. D. Hinzman, M-K. Woo, and K. R. Everett1392 Arctic Hydrology and Climate Change. ln Arctic

Ecosystems in a Changing Climate, an Ecophysio-)ogical Perspective, edited by F. S. Chapin, R. L.

feflries, J. F. Reynolds, G. R. Shaver, and f. Svo-boda, pp. 35-57. Academic Press, New York.

Kelly, R. L. and L. C. Todd19BB Coming into the Country: Early Paleoindian Mo-

bility and Hunting. American Antiquity 53:231-244.

Kitigawa, H. and f. van der Plicht1998 Atmospheric Radiocarbon Calibrations to 45,000

Years B.P.: Late Glacial Fluctuations and Cosmo-genic Isotope Production. Science 279:1.^lB7-1190.

Klein, D. R.1996 Arctic Ungulates at the Northern Edge of Terres-

trial Life. Rangifer 16:51-56.

19sg The Roles of Climate and Insularity in Establish-ment and Persistence of Rangifer torondus Popu-lations in the High Arctic. Ecological Bulletins47:96-1,O4.

Klein, D. R. and C. Bay1994 Resource Partitioning by Mamma-lian Herbivores

in the High Arctic. Oecologia 97:439-450.

Arctic An Lhropology 38 :2

Knuth, E.1,566/7 The Ruins of Musk-ox Way. Folk 8-9:191-219,

IGebs, f. S. and R. G. Barry"1970 The Arctic Front and the Tundra-Taiga Bor-rndary

in Eurasia. Geographical Review tio:548-554.

Kunz, M, L. and R. E. Reanier1994 Paleoindians in Beringia: Evidence {iom Arctic

Alaska: Science 263:660-662.

1995 The Mesa Site: A Paleoindian Hunting Lookoutin Arctic Alaska. Arctic Anthropology 32:5-30.

Lamb, H. H.1.972 Fundanentals of Clintate Now' Methuen,

London.

Lesl ie, L. D.1989 Alaska Climate Summaries: A Compilation of

Long-Term Means and Extremes at 478 AlaskanStations. Alaska Climate Technical Note Number3. Arctic Environmental Information and DataCenter, University of Alaska, Fairbanks.

Lozhkin, A. V., P M. Anderson, W. R. Eisner, L G. Ra-vako, D. M. Hopkins, L. B. Bmbaker, P. A. Colinvaux, andM. C. Millel1993 Late Quaternary Lacustrine Polien Records from

SotLthwestern Beringia. Quaterncry Researc h 3931,4-324.

MacDonald, G. M., A. A. Veiichko, C. V. Kremenetski,O. K. Borisova, A. A. Goleva, A. A. Andreev, L. C. Cwy-nar, R. T. Riding, S. L. Forman, T. W. D. Edwards, R. Ara-vena, D. Hammarlund, |. M. Szeicz, and V. N. Gattaulin2000 Holocene Treeline History and Climate Change

Across Northern Eurasia. Quaternary Research53(3):302-311.

MacDonald, G. M. and L. C. Cr,tynar1985 A Fossil-Based Reconstruction of the Late Quater-

nary History of Lodgepole Pine (Prnu.s contortassp. Iatifolia) in the Western Interior of Canada.Canadian fournal of Forest Research 15:1039-' to44.

MacNeish, R. S,2OO"l The Significance of Early Radiocarbon Dates on

Extinct Bison Priscus Bones from Engigstciak,Yukon Arctic Coast. In Mounds, Modoc, andMesoamerica: Papers in Honor of Melvin L.Fowler, edited by S. R. Ahler, pp. 1-24. IllinoisState Museum Scientific Papers, Vol. 28. Spring-field.

Mandryk, C. A. S.1993 Hunter-Gatherer Social Costs and the Nonviabil-

ity of Submarginal Environmenls. lournal of An-thropological Research 49:39-7 1,.

Mann, D. H. and D. M. PeteetL994 Extent and Timing of the Last Glacial Maximum

in Southwest Alaska. Quaternary Research 42'.136-148.

Page 19: Date: 10/7/2007 9:17:16 PM

Mann et al.: Environmental Chanpe and Arctic Paleoindians

Mann, D. H. and T. D. Hamilton1995 Late Pleistocene and Holocene Paleoenvuon-

ments of the North Pacific Coast. Quafern ary Sci-ence Reviews 14:449-47 1,.

Mann, D. H., D. M. Peteet, R. E. Reanier, and M. L. Kunz2OOZ Responses of an Arctic Landscape to Late Glacial

and Early Holocene Climatic Changes: The Im-pofiarrce of Moisture. Quaternary Science Re-views 21,:997-1,O21,.

Martin, P. S. and R. G. Klein1984 Quaternary Extinctions, A Prehistoric Revolution.

lJniversity of Arizona Press, Tucson.

Mayewski, P. A., L. D. Meeker, S. Whitlow, M. S.Twickler, M. C. Morrison, R. B. Alley, and K. Taylor1993 The Atmosphere during the Younger Dryas. Sci-

ence 26'1,:195-197.

McGhee, R.1996 Ancient People of the Arctic. University of British

Columbia Press, Vanc;ouver.

Meese, D. A., A. | . Gow, R. ts. Al ley, G. A. Ziel inski, P. M.Grootes, M. Ram, K. C. Taylor, P. A. Mayewski, and J. F.Bolzan1.997 The Greenland ke Sheet Project 2 Depth-Age

Scale: Methods and Resuits, lournal of Geophysi-ca I Re seo rc h "102:26,41 1 -26,423.

Meltzer, D. f. and B. D. Smith1S86 Paleoindian and Early Arr;haic Subsistence St_r'at-

egies in Eastern North America. ln Foraging,Collecting, and Harvesting: Archaic Period Sub-sistence and Settlentent in the Eastern Wood-,lands, edited by S. Neusis, pp 3031. Center forArchaeological Investigations Occasional PaperNumber 6, Southern Illinclis University, Carbon-dale.

Mock, C. |., P. J. Bartlein, and P. M. Anderson1998 Atmospheric Circulation Patterns and Spatial Cli-

matic Variability in Beringia. Intemational lour-nal of Climatology 18:1,O85-L1 04.

Moritz, R. E.1979 Slrnoptic Climatdogy oJ the Beaufoft Sea Coast

of Alaska.Institute of Arctic ald Alpine ResearchOccasional Paper Nrlmber 30, 176 pp.

Murray, D. F.1980 Balsam Poplar in Alaska. Canadian lournd of

Anthropology 1,:29-32.

Nelson, R. E. and L. D. Carter1,987 Paleoenvironmental Analysis of Insects and

Extralimital Populus from an Early Holocene Siteon the Arctic Slope of Alaska. rJSA. Arctic andAI pi ne Research tgtzzo-zql

NASA/GISS1999 Storm Track Atlas. URL: http://icp.giss.nasa.gov/

cgi-bin/ stormsl. cgi. NASA/ Goddard Institute forSpace Studies. New York.

L37

Ne'nt'rnan, |. E. and C. L Branton1.972 Annual Water Balance and Agricultural Develop-

ment in Alaska. Ecology 53:513-519.

Oswald, W. W., L. B. Brubaker, P. M. Anderson1999 Late Quaternary History of the Howard Pass

Area, Northwestern Alaska. Canadion lournal ofBotany 77:57O-58'1..

Ovenden. L.1990 Peat Accumulation in Northern Wetlands. Quo-

ternary Research 33:32 7 -386.

PALE2000 Paleoenvironmental Atlas of Beringia, National

Geophysical Data Center URL: http://ngdc.noaa.gov/ paleo / pale / atlas /beringia / index.html.

Patric, f. H. ard P. E. Black1968 Potential Evaporation and Climate in Alaska by

Thornthwaite's Classilication. U nited States De-partment of Agricultu'e Forest Serice ResearchPoper PNW-71. Par;ific; Northwest Forest anclRange Experimental Station, Institr.rte of NorthemForestry, luneau.

Pellet ier, B. R.1987 Marine Science Aths of the Beauloft Sea: Geol-

ogy and Geophysir:s. Geological Survey of []anadaMiscellaneous lleport 40, Energy, Mines, ancl Re-sources, Ottawar.

Ping, C. L., ] . C.Bockheim, f . M. Kimble, G. J. Michael-son, and D. A. Walker1S98 Characteristics of Cryogenic Soils along a Latitu-

dinal Transect in Arctir; Alaska. Jountcil oJ'Geo-physical Reseorch 1o3:28,917 -28,928.

Rasic, l . T.2000 Prehistoric Lithic Technology at the Tuluaq Hill

Site, Northwest Alaska. M.A. thesis, Departmentof Anthropology, Washington State University,Pullman.

Reanier, R. E.19S5 The Antiquity of Paleoindian Materials in North-

ern Alaska. Arctic Anthropology 32:31.-5O.

Ritchie, I. C., L. C. Cwynar, and R. W. Spear1gtt3 Evidence flont North-west Canada for an Earlv

Holocene Milankovitch Thermal Maximum. l/r.r-ture 3O5:1,26-1,28.

Rovansek, R. I., L. D. Hinzman, and D. L. Kare1996 Hydroiogy of a Tundra Wetiand Complex on the

Alaska Arctic Coastal Plain, IJSA. Arctic and Al-pine Research 28:31,'1,-317.

Scott, K. M.],S7B Effects of Permafrost on Stream Channel Behavicir

in Arctic Alaska. United States Geologicd SurveyProfessional Poper 1068:18 pp.

Shaver, G. R., I. A. Laundre, A. E. Giblin, and K. l. Nadel-hoffer1996 Changes in Live Plant Biomass, Primary Produc-

tion, and Species Composition along a Riverside

Page 20: Date: 10/7/2007 9:17:16 PM

1s8

Toposequence in Arctic Alaska, USA. Arctic andAlpine Reseorch 28:363-3 79.

Spencer, R. F.1S59 The North Alaskan Eskimo: A Study in Ecology

and Society. Bureau of American Ethology, Bulle-tin 171,, United States Government Printins Of-fice, Washington, D.C.

Stuiver, M., P. I.Reimer, and R. Reimer2000 CALIB 4.3 Radiocarbon Calibration. htto://

depts.washineton.edu/qil /calib/calib-hfr l

Stuiver, M., P,]. Reimer, E. Bard, J. W. Beck, G. S. Burr,K. A. Hughen, B. Kromer, G. McCormac, I. Van der Plicht,and M. Spurk1998 INTCAL9B Radiocarbon Age Calibration, 24,000-

0 cal B.P. Radiocorbon 40:104L-1084.

Walker, D. A.1.9a7 Height and Growth-Ring Response of Salix lanata

ssp. .Ricfiordsonii along the Coastal TemperatureGradient of Northern Alaska. Canadian lournal ofBotany 65:988-993.

2000 Hierarchical Subdivision of Arctic Tundra Basedon Vegetation Response to Climate, Parent Mate-rial, and Topography. Global Change Biology 6:19-34.

Walker, D. A. and K. R. Everett1991 Loess Ecosysyems of Northern Alaska: Regional

Gradient and Toposequence at Prudhoe Bay. Eco-logical Monographs 6:437 -464.

Arctic Anthropologlr 3 B :2

Walker, D. A. and M. D. Walker1996 Terrain and Vegetation of the Imnavait Creek

Watershed. ht Landscape Function and Distur-bance in Arctic Tundra, edited by I. F. Reynoldsand J. D. Tenhunen, pp. 73-108. Springer Verlag,New York.

Walker, D,A., J. G. Bockheim, F. S. Chapin, W. Eugster,f. Y. King, J. McFadden, G. l. Michaelson, F. E. Nelson,W. C. Oechel, C. L. Ping, W. S. Reeburgh, S. Regli, N. LShiklomanov, and G. I. Vourlitis1998 Energy and Trace-Gas Fluxes across a Soil pH

Boundary in the Arctic. Nature 394:469-472.

Walker, D. A., I. G. Bockheim, F. S. Chapin, W. Eugster,F. E. Nelson, and C. L. Ping200L Calcium-Rich Tundra, Wildlife, and the "Mam-

moth Steppe." Quaternary Science Reviews 2O:149-163.

Zhang, T., T. E. Osterkamp, and K. Stamnes1996 Some Characteristics of the Climate in Northern

Alaska, USA. Arctic and ALpine Research 28:509-518.

'Lgs7 Effects of Climate on the Active Layer andPermafrost on the North Slope of Alaska, U.S.A.Permafrost and Periglacial Pt:ocesses 8:45-67.