Geoarchaeology and Late Glacial Landscapes in the Western Lake … · 2019-04-30 · Geoarchaeology and Late Glacial Landscapes in the Western Lake Superior Region, Central North

Geoarchaeology and Late Glacial

Landscapes in the Western Lake

Superior Region, Central North America

Christopher L. Hill*

Department of Anthropology, Boise State University, Boise, ID 83725-1950

The transition from full glacial to interglacial conditions along the southern margin of theLaurentide ice sheet resulted in dramatic changes in landscapes and biotic habitats. Strataand landforms resulting from the Wisconsin Episode of glaciation in the area directly west ofLake Superior indicate a context for late Pleistocene biota (including human populations)connected to ice margins, proglacial lakes, and postglacial drainage systems. Late Glacial land-scape features that have the potential for revealing the presence of Paleoindian artifacts includeabandoned shorelines of proglacial lakes in the Superior and Agassiz basins and interiordrainages on deglaciated terrains. The linkage between Late Pleistocene human populationsand Rancholabrean fauna has yet to be demonstrated in the western Lake Superior region,although isolated remains of mammoth (Mammuthus) have been documented, as well asfluted points assigned to Clovis, Folsom, and Holcombe-like artifact forms. Agate Basin andHell Gap (Plano-type) artifacts also imply the presence of human groups in Late Glacial land-scapes associated with the Agassiz and Superior basins. © 2007 Wiley Periodicals, Inc.

INTRODUCTION

Late Pleistocene landscape habitats of the western Lake Superior region aredirectly linked to changing glacial conditions. The glacial chronology can serve asa framework for understanding the potential effects of landscape evolution on latePleistocene biotic populations, including prehistoric human groups, entering ice-margin and recently deglaciated landscapes stretching from glacial Lake Agassiz inthe west to the Superior basin in the east. The present geomorphology, along withthe stratigraphic relationships of the deposits that form this landscape, provides ameans to evaluate the physical context for the Late Glacial and earliest post-Glacialpaleobiotic communities of the region. The geomorphic processes occurring duringthe late Quaternary affected the age, distribution, and visibility of paleontological andarchaeological sites in northeastern Minnesota. Their study provides an opportu-nity to highlight the potential of an earth-science approach to integrate various facetsof the prehistoric record (Rapp and Hill, 2006).

The remains of extinct mammals and the presence of fluted and basally thinnedartifacts seems to argue for the existence of Late Pleistocene habitats that could

Geoarchaeology: An International Journal, Vol. 22, No. 1, 17–49 (2007)© 2007 Wiley Periodicals, Inc.Published online in Wiley Interscience (www.interscience.wiley.com). DOI:10.1002/gea.20145

*E-mail: [email protected].

sustain biotic communities associated with the end of the Rancholabrean land-mammal age in the Great Lakes region (Fisher, 1996; Haynes, 2002a, 2002b). Geologiccontexts associated with deglaciated landscapes contain the potential for the dis-covery of associations between extinct Rancholabrean fauna and Late Pleistocenehumans. For instance, intriguing discoveries of mastodon and mammoth remainspotentially connected with Late Pleistocene humans have been reported for thesouthern Great Lakes (Fisher, 1984, 1987; Shipman et al., 1984; Overstreet and Kolb,2003). Although rare, Clovis artifacts or other fluted points are known from the west-ern Great Lakes region (Shane, 1989; Romano and Johnson, 1990; Mulholland et al.,1997; Mulholland, 2000). If these fluted points can be considered reliable indicatorsof the time of prehistoric human presence, they imply inhabitable landscapes start-ing at least by the Pleistocene/Holocene boundary, which is marked by the end of theYounger Dryas chronozone. Here, the phrase “Late Glacial” is applied to the timeinterval associated with the Late Pleistocene (“Woodfordian,” including the LastGlacial Maximum and ending with the Younger Dryas chronozone) and early Holocene(Preboreal chronozone, ~10,000–9000 14C yr B.P.) time-transgressive deglaciation ofthe western Lake Superior region.

This article provides a review of landscape evolution in the western Lake Superiorregion (Figures 1–3) and examines its potential relevance to prehistoric humangroups during Late Glacial events. The chronologic framework for these events is alsoappraised by integrating information based on stratigraphic and geomorphic rela-tionships (Tables I and II) with radiocarbon ages of regional paleoecological recordsand their correlation with time-diagnostic Paleoindian artifact forms. Radiocarbondates are used to evaluate the chronologic framework for environmental change inthe region and to infer the potential temporal ranges for artifact assemblages, basedon diagnostic point forms.

Although relatively common in southern Minnesota and in the Great Lakes region(Stauffer, 1945; Cleland et al., 1998; Holman, 2001), Rancholabrean fossils are rareimmediately west of Lake Superior. Mammuthus jeffersoni (Jefferson’s mammoth)has been recovered in gravels along the Cloquet River drainage (Figure 1), andMammuthus primigenius (woolly mammoth) and possibly Mammut americanum

(American mastodon) are also documented from the area west of Lake Superior andthe Agassiz basins (Stauffer, 1945; Harington and Ashworth, 1986). Thus, isolateddiscoveries of extinct Pleistocene fauna and artifacts apparently dating to the latePleistocene both hint at the presence of inhabitable Late Glacial landscapes in theregion. Based on studies elsewhere in the Great Lakes region, it should also be keptin mind that there is potential evidence for links between Late Pleistocene humanpopulations and Rancholabrean fauna without the presence of diagnostic artifactforms (cf. Fisher, 1984; Holman, 2001; Overstreet and Kolb, 2003). Here, a review ofthe landscape evolution is used to highlight the potential for using an integratedgeoarchaeological approach for investigating the Late Glacial (generally the end ofthe Wisconsin Episode; Karrow et al., 2000) paleoecology and human prehistory ofthe area west of Lake Superior.

Two radiocarbon data sets are used to evaluate the potential geoarchaeologicalrelationships in the region: radiocarbon ages from geologic and paleoecologic

HILL

GEOARCHAEOLOGY: AN INTERNATIONAL JOURNAL, VOL. 22, NO. 1 DOI: 10.1002/GEA18

DOI: 10.1002/GEA GEOARCHAEOLOGY: AN INTERNATIONAL JOURNAL, VOL. 22, NO. 1

LATE GLACIAL LANDSCAPES IN THE WESTERN LAKE SUPERIOR REGION

19

Table I. Radiocarbon dates used to constrain western Lake Superior region glacial events.

Locality Age (yr B.P.a) Lab No. Reference Material

Minnesota

Myrtle Lake 11,120 � 250 Y-1782 Janssen, 1968 Wood

Heikkilla Lake 12,100 � 150 Lu-2556 Bjork, 1990 �3% organicsb

Heikkilla Lake 12,070 � 170 Lu-2496 Bjork, 1990 �4% organicsb

Heikkilla Lake 11,320 � 130 LuB2497 Bjork, 1990 �4% organicsb

Heikkilla Lake 10,690 � 120 Lu-2498 Bjork, 1990 25% organicsb

Heikkilla Lake 10,110 � 150 Lu-2499 Bjork, 1990 43% organicsb

Lempia 1 12,050 � 240 Lu-2555 Bjork, 1990 �3% organicsb




Sabin 1 10,230 � 120 Lu-2506 Bjork, 1990 13% organicsb

Sabin 2 10,320 � 170 Lu-2507 Bjork, 1990 14% organicsb

Swamp 9510 � 90 Lu-2504 Bjork, 1990 �4% organicsb

Swamp 8890 � 80 LuB2505 Bjork, 1990 27% organicsb

Big Rice Lake 12,040 � 540/570 DIC-3322 Huber, 1988 Organic clay

Mariska-Gilbert 11,330 � 350 W-827 Winter, 1971 Wood

Mariska-Gilbert 11,100 � 400 W-1140 Winter, 1971 Wood

Glatsch Lake 9720 � 120 Y-1983 Wright and Watts, 1969 Gyttja?

Weber Lake 14,690 � 390 W-1763 Florin and Wright, 1969 Moss

Weber Lake 10,550 � 300 W-873 Fries, 1962 Organics in sediments

Weber Lake 10,180 � 160 U-176 Fries, 1962 Organics in sediments

Weber Lake 9150 � 130 U-164 Fries, 1962 Organics in sediments?

Kylen Lake 15,850 � 240 I-5048 Birks, 1981 Sedimentsc

Kylen Lake 13,450 � 500 Q-1503 Birks, 1981 Sedimentsc

Kylen Lake 12,390 � 120 WIS-682 Birks, 1981 Sedimentsc

Kylen Lake 10,820 � 120 Q-1502 Birks, 1981 Sedimentsc



Kylen Lake 9320 � 110 Q-1501 Birks, 1981 Sedimentsc

Kylen Lake 8575 � 80 WIS-690 Birks, 1981 Sedimentsc

Kylen Lake 8410 � 85 WIS-694 Birks, 1981 Sedimentsc

Kylen Lake 13,900 � 300 UW1234 Lund and Banerjee, 1985 Aquatic moss

Kylen Lake 16,050 � 230 UW1233 Lund and Banerjee, 1985 Silty gyttja


Klyen Lake 16,760 � 220 UW1231 Lund and Banerjee, 1985 Silty gyttja



Kylen Lake 10,470 � 40 UW1228 Lund and Banerjee, 1985 Organic gyttja

Kylen Lake 9840 � 40 UW1227 Lund and Banerjee, 1985 Organic gyttja

Cloquet Lake 16,450 � 410 PITT-1191 Huber, 2001 Organic clay

Aitkin Main Site 11,710 � 325 W-502 Farnham et al., 1964 Wood

Aitkin Main Site 11,560 � 400 W-1141 Farnham et al., 1964 Wood

Rossberg Bog 10,100 � 140 Y-2460 Wright and Watts, 1969 Gyttja

Ball Bluff 9080 � 200 GX-6777 Hobbs, 1983 Shells in marl

(continued)

HILL


Table I. (Continued).


Bog 2, Reservoir Lakes 9270 � 190 UCR-1825 Personal communication,

Taylor to Rapp, 1985 Peat

Bog 2, Reservoir Lakes 9420 � 180 UCR-1826 Personal communication,

Taylor to Rapp, 1985 Peat

Spider Creek 22,000 � 600 W-1233 Baker, 1965 Organic sedimentsd

Spider Creek 13,000 � 400 W-1234 Baker, 1965 Organic sediments

Kotiranta Lake 16,150 � 550 W-1973 Wright and Watts, 1969 Sediment



White Lily Lake 15,250 � 220 I-5051 Wright et al., 1973 Organic sediments

Jacobsen Lake 10,400 � 300 L-794-F Florin and Wright, 1969 Plant debris

Jacobsen Lake 10,400 � 300 L-794-1 Wright and Watts, 1969 Silty gyttja

Andersen Lake 10,800 � 300 L-794 A-D Florin and Wright, 1969 Plant debris

Andersen Lake 10,500 � 200 Y-1362 Wright and Watts, 1969 Gyttja

Andersen Lake 10,200 � 200 L-794 E Wright and Watts, 1969 Gyttja

Blackhoof 10,630 � 500 W-1677 Wright and Watts, 1969 Peat

Blackhoof 10,420 � 300 W-1714 Wright and Watts, 1969 Peat

Wolf Creek 20,500 � 400 I-5443 Birks, 1976 Noncalcareousc

Wolf Creek 14,720 � 210 I-5442 Birks, 1976 Sideritic sediments

Wolf Creek 13,600 � 200 I-5046 Birks, 1976 Sideritic sediments

Wolf Creek 13,670 � 210 I-5910 Birks, 1976 Sideritic copropel

Wolf Creek 13,775 � 310 Q-1084 Birks, 1976 Marly copropel

Wolf Creek 13,510 � 190 I-5441 Birks, 1976 Marly copropel


Wolf Creek 12,240 � 170 I-5440 Birks, 1976 Marly copropel

Wolf Creek 11,230 � 165 I-5439 Birks, 1976 Marly coprepel


Wolf Creek 11,270 � 165 I-5438 Birks, 1976 Copropelic marl

Wolf Creek 11,040 � 165 I-5437 Birks, 1976 Copropelic marl

Wolf Creek 10,534 � 200 Q-1081 Birks, 1976 Detritus copropel

Wolf Creek 10,080 � 170 I-5436 Birks, 1976 Detritus copropel

Wolf Creek 9150 � 150 I-5435 Birks, 1976 Detritus copropel

Swift 10,050 � 100 WIS-1325 Bjork and Keister, 1983 Wood and peat

Swift 9350 � 100 WIS-1324 Bjork and Keister, 1983 Wood and peat

Lake of the Clouds 9630 � 110 Y-2183 Stuiver, 1975 Organics

Lake of the Clouds 9410 � 100 Y-2182 Stuiver, 1975 Organics

Mosbeck 9940 � 160 I-3880 Ashworth et al., 1972 Wood

Ontario

Cummins Pond 11,080 � 190 DIC-190 Julig et al., 1990 Marl

Cummins Pond 9260 � 70 TO-547 Julig et al., 1990 Wood

Cummins Pond 8110 � 135 DIC-2579 Julig et al., 1990 Wood

Cummins Pond 8110 � 110 BETA-4486 Julig et al., 1990 Wood

Cummins 8480 � 390 NMC-1216 Ross, 1997 Burned bone

Oliver Pond 8210 � 70 WAT-1572 Julig et al., 1990 Marl

(continued)



21



Oliver Pond 11,950 � 70 WAT-1584 Julig et al., 1990 Marl

Rattle Lake 11,110 � 100 WIS-1327 Bjorck, 1985 �6% organicsb



Rattle Lake 8420 � 90 WIS-1398 Bjorck, 1985 21% organicsb

Sioux Pond 9740 � 100 WIS-1329 Bjorck, 1985 19% organicsb

Indian Lake 9140 � 100 WIS-1331 Bjorck, 1985 15% organicsb

Rainy River Basin 10,500 � 200 WAT-1760 Bajc et al., 2000 Wood

Rainy River Basin 9750 � 170 BGS-1304 Bajc et al., 2000 Wood

Rainy River Basin 9530 � 140 WAT-1934 Bajc et al., 2000 Wood

Rainy River Basin 9920 � 110 BGS-1305 Bajc et al., 2000 Plant detritus


Rainy River Basin 10,100 � 180 WAT-1936 Bajc et al., 2000 Charcoal


Rainy River Basin 10,700 � 140 WAT-1910 Bajc et al., 2000 Peat

Wisconsin and Michigan (Southern Lake Superior Region)

Brule Channel 9050 � 60 B-112985 Engseth, 1998 Plant material

Saxon Moraine 10,100 � 100 WIS-409 Black, 1976 Wood

Saxon Moraine 9730 � 140 I-5082 Black, 1976 Wood

Northern Wisconsin 10,880 � 160 I-3651 Black, 1976 Basal lake organic mud

Northern Wisconsin 12,900 � 300 I-3780 Black, 1976 Basal lake organic mud

Northern Wisconsin 9460 � 100 WIS-371 Black, 1976 Basal lake organic mud

Porcupine Mountain Moraine 10,230 � 500 W-1414 Levin et al., 1965 Wood

Porcupine Mountain Moraine 10,100 � 250 W-1540 Peterson, 1982 Wood

Porcupine Mountain Moraine 10,250 � 250 W-1541 Peterson, 1982 Wood

Porcupine Mountain Moraine 10,420 � 220 OX-1 Black, 1976 Wood

White Pine, Caribou Creek 10,200 � 500 M-359 Crane, 1956 Wood

White Pine, Caribou Creek 10,230 � 280 W-964 Hack, 1965 Wood

White Pine, Caribou Creek 9600 � 350 W-965 Hack, 1965 Wood

White Pine, Caribou Creek 9500 � 350 W-1150 Hack, 1965 Gyttja

Camp 11 Lake 10,200 � 150 unknown Brubaker, 1975 Unknown

Lost Lake 11,945 � 180 unknown Brubaker, 1975 Unknown

Yellow Dog 9100 � 150 unknown Brubaker, 1975 Unknown

Lake Gribben forest bed 9895 � 55 A-7878 Lowell, 1999 Wood

Lake Gribben 9910 � 55 A-7876 Lowell, 1999 Wood

Lake Gribben 9965 � 55 A-7877 Lowell, 1999 Wood

Lake Gribben 10,040 � 55 A-7879 Lowell, 1999 Wood



Lake Gribben 10,075 � 95/– 90 A-7880 Lowell, 1999 Wood



Lake Gribben 10,330 � 300 W-3896 Hughes and Merry, 1978 Needles

Lake Gribben 10,220 � 215 DAL-338 Hughes and Merry, 1978 Wood

(continued)

contexts in the western Lake Superior region (Tables I and II), and radiocarbon agesthat help to estimate the temporal range of time diagnostic artifact forms (Holliday,2000). Besides the stratigraphic and taphonomic contexts, interpretation of the radio-carbon ages takes into account factors such as the material dated and the size ofthe standard deviation (1- and 2-sigma confidence interval ranges). Figure 2 showsthe location of radiocarbon sample sites mentioned in the text. Stratigraphic andgeomorphic relationships and radiocarbon ages are used to assess the framework ofLate Glacial landscape change in the western Lake Superior region (Table III). Thesechronologic relationships are ultimately compared and integrated with temporalinformation based on time-diagnostic artifact forms. Thus, an integrative geoar-chaeological approach is used to evaluate temporal relationships associated withthe deglaciation chronology, as well as specific landscape settings that can be linkedto Late Glacial human populations and Rancholabrean biotic communities.

BACKGROUND

Linkages between environmental conditions and Late Glacial human populationshave been the subject of studies in both the Great Lakes and Lake Agassiz areas.Some examples of the use of a geoarchaeological perspective in the Agassiz basininclude Pettipas and Buchner (1983), Buchner and Pettipas (1990), and Ross (1997).Studies by Farrand (1977), Storck (1988), Julig (1991), Phillips (1993), Cleland et al.(1998), Kuehn (1998), Anderton (1999), Jackson et al. (2000), and Julig andMcAndrews (1993) illustrate the potential of linking environmental change andPaleoindian populations in the Great Lakes region.

The basis for this review is research conducted under the auspices of the Universityof Minnesota’s Archaeometry Laboratory, founded and directed by George (Rip) Rapp.This research evaluates potential Late Pleistocene archaeological and vertebrate set-tings by integrating geomorphic and ecological dynamics, which are partly based oninterpretations of the regional glacial geology. In fact, early in his career, Rip Rappserved as a field assistant for H. E. Wright, Jr., conducting research on the history ofvarious glacial lobes and their regional direction of advances using pebble counts

HILL




Lake Gribben 9850 � 300 W-3866 Hughes and Merry, 1978 Wood


Lake Gribben 9545 � 225 DAL-340 Hughes and Merry, 1978 Wood


Lake Gribben 9830 � 70 B-91835 Pregitzer et al., 2000 Wood


Lake Gribben 10,290 � 90 B-91832 Pregitzer et al., 2000 Wood


aRadiocarbon years before present, 1-sigma (68% confidence interval). bMeasured percentage of organics or organic mat-ter in sediments. cSediments with up to 5% pre-Quaternary microfossils, but no lignite observed. dContains pre-Pleistocenemicrofossils, and lignite.



23

Table II. Contexts for radiocarbon ages used to constrain western Lake Superior region glacial eventsand potential geoarchaeological settings.

Locality Contextual information

Minnesota

Myrtle Lake Overlies Koochiching Lobe, proposed temporal correlation is with Alborn phaseof St. Louis sublobe.

Heikkilla Lake Within Vermilion moraine complex (Rainy Lobe). Lempia Lake Within Vermilion moraine complex (in front of Wahlsten moraine, Rainy Lobe).Sabin Lake Within Embarrass channel (Giants Ridge area).Swamp Lake Within Vermilion moraine complex (in front of Big Rice moraine, Rainy Lobe).Big Rice Lake Within Vermilion moraine complex (in front of Wahlsten moraine, north of Big

Rice moraine).Mariska Mine (Gilbert) In Alborn phase till, St. Louis sublobe (laterally correlative deposits are over-

lain by glacial Lake Upham sediments).Glatsch Lake Overlies Alborn phase deposits of St. Louis sublobe.Weber Lake Overlies moraine (drumlin) of St. Croix phase, Rainy Lobe.Kylen Lake Overlies moraine of St. Croix phase, Rainy Lobe.Cloquet Lake Overlies Highland moraine (Automba phase, Superior Lobe).Aitkin Main In Lake Aitkin II clays, overlies paleosols in sands, overlies Alborn phase till of

St. Louis sublobe.Rossberg Bog Overlies Alborn phase till, St. Louis sublobe.Ball Bluff Marl formed in lagoon, correlative with strandline of Lake Aitkin-Upham after

Alborn phase St. Louis sublobe.Bog 2, Reservoir Lakes Overlies moraine of St. Croix phase Rainy Lobe, and possible Automba phase

outwash.Spider Creek Overlies Alborn phase, St. Louis sublobe.Kotiranta Lake Overlies Cloquet outwash, on Automba/Split Rock phase, in front of Thompson

moraine (Nickerson moraine), Superior Lobe.White Lily Lake Overlies Automba and Split Rock phase, Superior Lobe.Jacobsen Lake Overlies Nickerson moraine, Superior Lobe.Andersen Lake Overlies Nickerson moraine, Superior Lobe.Blackhoof Channel Fill above last diversion channel around Thompson moraine (Nickerson phase,

Superior Lobe), channel is eroded through Thompson moraine, last channelbefore Lake Nemadji, after Nickerson phase.

Wolf Creek Overlies St. Croix phase moraine (Pierz drumlins).Swift On Campbell beach of Lake Agassiz, wood and peat above Moorhead low and

below Emerson phase transgression gravel.Lake of the Clouds Between recessional moraines of Rainy Lobe, north of Vermillion moraine.Mosbeck Spruce and tamarack incorporated into Emersion transgression deposits (plants

from Moorhead low).

Ontario

Cummins Pond South of Dog Lake moraine (Rainy Lobe), within region covered by Marquettephase (Superior Lobe) advance to Marks moraine, adjacent to Minong beach.

Oliver Pond South of Dog Lake moraine (Rainy Lobe), within region covered by Marquettephase (Superior Lobe) advance to Marks moraine, near Minong beach.

Rattle Lake In front of Eagle-Finlayson-Brule Creek Moraine, Rainy Lobe.(continued)

and till fabric studies. The data Rapp helped to gather was included in the landmarkarticle by Arneman and Wright (1959) and he was acknowledged for his assistance byWright (1957, 1962). The linkage of glacial geology and geomorphology with studiesof the Paleoindian archaeology of northern Minnesota was part of several researchprojects initiated by the Archaeometry Laboratory starting in the 1970s. Some exam-ples include studies of the ecologic, geomorphic, and paleogeographic settings asso-ciated with Paleoindian localities along the Cloquet River (Hill et al., 1985; Huber andHill, 1987; Hill, 1994, 1995a, 1995b; Harrison et al., 1995); Late Quaternary contexts ofthe Big Fork River valley and Agassiz basin (Hill et al., 1995; Rapp et al., 1995; Hill andHuber, 1996); and the potential for predicting the location of Paleoindian sites alongabandoned shorelines within the Lake Superior basin (Hill, 1994; Phillips and Hill,1993, 1994a, 1994b, 1994c, 2001; 2004).

STUDY AREA

The focus of this article is the region immediately west of Lake Superior in north-ern Minnesota and Ontario (Figure 1). Some discussion of adjacent areas, including

HILL


Table II. (Continued).

Locality Contextual information

Sioux Pond Overlies Sioux Lookout moraine, Rainy Lobe.Indian Lake In front of Whitewater moraine, Rainy Lobe.Rainy River Basin Wood incorporated into Lake Agassiz Moorhead phase alluvium and lacustrine

deposits and Emerson beach deposits.

Wisconsin and Michigan (Southern Lake Superior Region)

Brule Channel Fill within channel postdates glacial Lake Duluth shorelines (channel inter-preted as outlet for glacial Lake Duluth).

Saxon Moraine Wood incorporated into red till of Saxon moraine (correlated with Grand Mariasmoraine and Marquette advance); wood is older than advance. Till is belowshorelines of glacial Lake Duluth.

Northern Wisconsin Basal lake organic muds.Porcupine Mountain Wood incorporated into red till of Porcupine Mountain moraine (correlated Moraine with Grand Marias moraine and Marquette advance); wood is older than

advance. Till is below shorelines of glacial Lake Duluth.White Pine Wood incorporated into Porcupine till correlated with Grand Marias moraine Caribou Creek and Marquette advance; wood is older than advance. Also log and gyttja above

Marquette phase till. Below shorelines of glacial Lake Duluth; log and gyttjadeposited after withdrawal from high shorelines of Lake Duluth, LakeOntonagon (contemporary with Lake Nemadji).

Camp 11 Overlies Yellow Dog outwash.Lost Lake Overlies Michigamme outwash.Yellow Dog Overlies Michigamme till.Lake Gribben Trees date North Bay Interstade, overlain by ice-margin lake deposits and distal

forest bed outwash formed by Marquette advance to Grand Marais moraine.



25

Figure 1. Location of western Lake Superior region, with major rivers and lakes. Straight dashed lies arecounty boundaries. Inset shows Upper Great Lakes area (after Figure 3 in Hill, 1995b).

Figure 2. Location of sites mentioned in text. Inset shows location of western Lake Superior region.

<zaq;11>

the Lake Agassiz basin and the area south of Lake Superior, is included becausestratigraphic and geochronologic data from these areas are crucial to the interpre-tation of the glacial chronology and paleoecology of the Pleistocene/Holocene bound-ary in the western Lake Superior region (Figure 2). Figure 3 shows the approximateposition of morainal features discussed in the text.

LATE GLACIAL ARCHAEOLOGY

Clovis and Folsom fluted points and related artifact assemblages form the oldestclearly defined archaeological entities in North America (Lepper, 1999; Holliday,2000). Human groups using Clovis artifacts existed as part of a Rancholabrean biota,including mammoth (Mammuthus), while Folsom artifacts have been recoveredwith extinct forms of bison (Bison antiquus). Plano-type Paleoindian point formsare generally assigned to latest Glacial and postglacial (but pre-Altithermal) pale-oenvironmental contexts. Paleoindian artifact assemblages are generally consideredto reflect human populations living in small nomadic bands with a subsistence strat-egy based on hunting and foraging. Plano-type Paleoindian artifact assemblages inthe Great Lakes region are believed to be younger than fluted-forms and to reflect asubsistence strategy partially based on the procurement of caribou and generalizedforaging (Cleland, 1965; Jackson, 1988, 1989; Kuehn, 1998).

The chronologic status of Late Glacial artifact forms in the Agassiz-western GreatLakes region is based primarily on comparisons with Paleoindian artifact formsdated on the Great Plains (Irwin and Wormington, 1970; Holliday, 2000; Sellet, 2001).Paleoindian artifacts have been placed into two groups—an “early” set consisting offluted forms and a “late” set consisting of nonfluted lanceolate forms. Fluted pointsrecovered from the region west of the Agassiz-Superior basins include a Clovis pointfrom Island Lake within the Cloquet River drainage (Figures 1 and 2; Romano andJohnson, 1990) and a fluted point from Round Lake (cf. Mulholland et al., 1997;Mulholland, 2000; Figure 2). A fluted point from northern Wisconsin has beendescribed as either Clovis (Mires, 1989) or Holcombe (basally thinned; Dudzik, 1991),

HILL


Table III. Proposed relative chronology, key to major phase depicted in Figure 3. See text for references.

Relative chronology Koochiching Lobe- Rainy Superior (see Figure 3) Agassiz Basin area Lobe basin

7 Whitewater Moraine Lakes6 Emerson-Agassiz Sioux Lookout

Hartman-Dog Lake Marks-Marquette5 Moorhead-Agassiz Eagle-Finlayson-Brule Lakes

Lockhart-Agassiz Lakes4 Cass/Lake Koochiching-Agassiz Steep Rock Lakes3 Alborn Vermilion Complex Nickerson2 Vermilion Complex Automba1 St. Croix St. Croix

and other Holcombe-like points have been reported for northeastern Minnesota(Mulholland et al., 1997). Nonfluted lanceolate artifacts (Plano-type, “Late”Paleoindian) are also distributed over the region (Harrison, 1995; Mulholland et al.,1997; LeVasseur, 2000; Mulholland and Shafer, 2000; Okstad et al., 2000). The occur-rence and distribution of Paleoindian artifacts associated with the Agassiz basin insouthern Manitoba and south of Lake Superior in Wisconsin are described by Buchnerand Pettipas (1990), Cleland et al. (1998), Dudzik (1991), and Kuhen (1998).

The western Lake Superior region contains an artifactual record that, by typo-logical correlation, can be assigned to a chronological position associated with thePleistocene/Holocene boundary (approximately the interval of the Allerød–Preborealchronozones, 12,000–9000 yr B.P., cf. Eriksen, 1996; Haynes, 2002a). For example, onemodel for the temporal relationships between Paleoindian artifacts for the GreatLakes region postulates a boundary between the Clovis and Folsom fluted forms at11,000 yr B.P., and a boundary between Folsom–Holcombe forms and unfluted points(the Agate Basin variant of Plano) at about 10,200 yr B.P. (Deller and Ellis, 1992).However, it may be useful to estimate the temporal range of artifact forms by taking



27

Figure 3. Major end moraines connected to glacial and lake phases. Numbers refer to proposed relativeage of ice advance, see Table III. See text for references.

into consideration the standard deviations associated with the radiocarbon agesassigned to these diagnostic artifacts. Although sometimes depicted as a develop-mental sequence with artifact forms occurring during discrete intervals, the radio-carbon dates appear to indicate a chronologic overlap of the diagnostic artifact types(Holliday, 2000; Sellet, 2001). For example, northern Plains Paleoindian assemblages,based on 1-sigma standard deviations, show an overlap of fluted and unfluted pointforms across the Pleistocene/Holocene boundary. These ranges are approximately:Clovis, 11,700–9600 yr B.P.; Folsom, 11,000–9600 yr B.P.; and Agate Basin–Hell Gap,11,400–9500 yr B.P. (Holliday, 2000). At a 2-sigma age range (standard deviation, 95%confidence interval), the chronologic overlap would be even larger. This is illus-trated by sites in the Great Lakes region where radiocarbon dates appear to linkRancholabrean fauna with human populations from before 12,000 yr B.P. to possi-bly as late as 9200 yr B.P., although the younger ages are considered to be anom-alous (Fisher, 1996; Laub and Haynes, 1998; Overstreet and Kolb, 2003). For exam-ple, northern Plains Paleoindian assemblages, based on 1-sigma standard deviations,show an overlap of the fluted and unfluted forms for the Pleistocene/Holocene bound-ary. Although the general ranges are approximately 11,200–10,900 yr B.P. for Clovis,10,900–10,200 yr B.P. for Folsom, and 10,500–9500 yr B.P. for Agate Basin-Hell Gap(Holliday, 2000), the standard deviations indicate the potential for notable overlapbetween these sets of radiocarbon ages. Clovis components have been found instratigraphic contexts, indicating they are older than Folsom; however, they have a1-sigma range of about 11,700–10,200 yr B.P. Folsom assemblages appear to gener-ally coincide with the Younger Dryas chronozone; on the northern Plains, this set hasa 1-sigma range from about 11,000 to 9600 yr B.P. Most of the radiocarbon ages forAgate Basin and Hell Gap assemblages range from about 10,500 to 9500 yr B.P.; how-ever, at 1-sigma, an Agate Basin component at the Hell Gap locality might even beolder than 11,000 (the 1-sigma standard deviation indicates a range from between11,400 and 10,300 yr B.P.). At a 2-sigma (standard deviation, 95% confidence interval),the chronologic overlap would be even larger between the sets of radiocarbon agesassociated with Clovis, Folsom, and Agate Basin-Hell Gap assemblages. This pat-tern may be reflected by sites in the Great Lakes region that appear to linkRancholabrean fauna with human populations prior to 12,000 yr B.P. to possibly aslate as about 9200 yr B.P., although the younger ages are considered to be anomalies(Fisher, 1996; Laub and Haynes, 1998; Overstreet and Kolb, 2003).

GLACIAL DYNAMICS AND CHRONOLOGY

Late Glacial Landscape Evolution

The landforms west of Lake Superior reflect the advance and melting of icelobes of the Laurentide ice sheet as modified by postglacial events. Glaciers haveadvanced over the region from three directions (Figure 3 and Table III; cf. Wrightand Watts, 1969; Richmond and Fullerton, 1983; Attig et al., 1985; Matsch andSchneider, 1986; Wright, 1998). Ice originating from the northeast within the

HILL


Superior basin is designated the Superior Lobe. Ice that advanced from southernCanada and through northeastern Minnesota, generally from the northeast, hasbeen designated the Rainy Lobe, whereas the Koochiching Lobe (including the St.Louis sublobe) generally extended from southern Manitoba and northwesternMinnesota southeastward toward the Superior basin.

Superior Lobe

There were several major phases of the late-Wisconsin glacial episode (Wrightand Watts, 1969; Lannon and Matsch, 1987; Leher and Hobbs, 1992; Johnson andMooers, 1998; Karrow et al., 2000; Figure 3, Table III). The relative order for themajor advances from the Superior basin is St. Croix, Automba, Nickerson, andMarquette. The St. Croix phase has been linked to the Woodfordian terminal morainein the western Great Lakes (Peterson, 1986). The age of the St. Croix phase (Figure3, Table III) can be estimated by radiocarbon measurements from Wolf Creek.Noncalcareous deposits overlying St. Croix phase deposits at Wolf Creek (Figure 2)date to about 20,500 yr B.P. (Table I). Although lignite was not observed (Birks, 1976),the presence of pre-Quaternary microfossils may indicate some contamination ofthis sample (Clayton and Moran, 1982). An overlying radiocarbon age implies theabsence of St. Croix phase ice by about 15,000 yr B.P. (Birks, 1976; Table I).

The Superior Lobe advanced to the Mille Lacs and Highland moraines (Figure 4)during the Automba phase (Wright and Watts, 1969; Matsch and Schneider, 1986;Leher and Hobbs, 1992). Deposits in Koitranta and White Lily Lakes (Figure 2, TableII) overlie a moraine attributed to the Split Rock phase, which is interpreted as afacies of the melting Automba phase ice (Lannon and Matsch, 1987). Radiocarbon-dated basal organic sediments from Koitranta Lake have a 2-sigma (95% confidenceinterval) range of about 17,250–15,500 yr B.P., while organic sediments from WhiteLily Lake have a 2-sigma radiocarbon age of 15,700–14,800 yr B.P. (Table I). A radio-carbon age from basal silty clay overlying the Highland moraine at Cloquet Lake(Figure 2) has a 2-sigma range of 17,260–15,640 yr B.P. (Huber, 2001). Thus, assum-ing the potential for some contamination by older carbon, the Automba phase mayhave been over by about 15,000 yr B.P.

Age estimates for the Nickerson phase are based on materials recovered from sed-iments overlying Nickerson phase till and correlations with radiocarbon dated woodoverlying St. Louis sublobe-Alborn phase sediments (Figure 5). The 2-sigma rangefor radiocarbon dates on plant debris and organic sediments from Jacobsen andAndersen Lakes (Figure 2; Wright and Watts, 1969) is about 11,400–9800 yr B.P., withan overlap interval from about 11,000–10,200 yr B.P. (Table I). Based on the correla-tion with the wood dates above Alborn phase till at Aitkin (Figure 2, Table I), the 2-sigma range is about 12,400–10,800 yr B.P (Farnham et al., 1964; see also Figure 6).

The last advance of the Superior Lobe into the region has been linked to theMarquette phase (Figure 7). This advance blocked the eastward drainage of theAgassiz basin by about 10,000 yr B.P. (LeVasseur, 2000). South of the Lake Superiorbasin, the Marquette advance buried the Lake Gribben forest (Figure 2; Hughes andMerry, 1978; Lowell et al., 1999; Pregitzer et al., 2000). Radiocarbon ages from wood



29

and needles indicate that the spruce forest was present from perhaps 10,200–9900yr B.P. (although the 2-sigma range is about 10,930–9250 yr B.P.), until being buriedby the Marquette advance (Table I). Radiocarbon ages from the southwestern mar-gin of the basin have been used to trace the Marquette advance westward from theMarquette Gribben forest area westward towards Duluth (Hack, 1965; Brubaker,1975; Black, 1976; Peterson, 1982; Clayton, 1984; Attig et al., 1985; Lowell et al., 1999;Table I, Figure 7). The Marquette phase southwest of the Superior basin may havereached the inner margins of the Nickerson and Thomson moraines (Mooers, Larson,et al., 2005).

Rainy Lobe

The Superior and Rainy lobes merged to form one large ice sheet during the St. Croixphase (Figure 3, Table III). During this phase, the combined Superior and Rainy lobesadvanced to the St. Croix moraine (Wright and Watts, 1969; Wright et al., 1973; Matschand Schneider, 1986). Age estimates for the advance of the St. Croix phase RainyLobe are in the range of 20,000–15,000 yr B.P., based on the Wolf Creek sequence (seeabove) and radiocarbon-dated sediments at Weber and Kylen Lakes (Tables I and II,

HILL


Figure 4. Paleogeographic model associated with Automba phase of Superior Lobe, with Rainy Lobemargin depicted within the Vermilion moraine complex at the Big Rice moraine. Lake Aitkin-Upham I rep-resents pre-Alborn phase lacustrine features. See text for references used to develop the interpretationof landscape evolution presented in Figures 5–7 (after Figure 5 in Hill, 1995b).

Figure 2). Pre-Quaternary microfossils in the Kylen Lake sediments may indicatesome contamination by older materials, although no lignite was observed (Birks,1981). Perhaps the most reliable constraint on the earliest deglaciated habitats afterthe melting of the Rainy Lobe is an age on aquatic moss from Kylen Lake with a 2-sigmarange of 14,500–13,300 yr B.P. (Lund and Banerjee, 1985; Table I), and an age on mossfrom Weber Lake with a 2-sigma range of 15,560–13,900 yr B.P. (Florin and Wright, 1969;Table I). The Rainy Lobe melted back to the Giants Range–Vermilion Lake region,forming the Vermilion moraine complex (Figure 4). This complex consists of theAllen, Wampas, Big Rice, Wahlsten, and Vermilion moraines (Leher and Hobbs, 1992).Radiocarbon dates from localities within the moraine complex from Heikklla andLempia Lakes (Bjork, 1990; Figure 2), and Big Rice Lake (Huber, 1988; Figure 2), sug-gest deglaciated terrain by before 12,500 yr B.P., although the dates are on sediments



31

Figure 5a. Paleogeographic model of ice margins associated with Nickerson phase of Superior Lobeand Alborn phase of St. Louis sublobe: Major advance of St. Louis sublobe. See Hobbs (1983) for config-uration of St. Louis sublobe and proglacial lakes (after Figures 7, 8, and 10 in Hill, 1995b).

low in organics or organic rich clays (Table I). Based on the presence of KoochichingLobe deposits overlying the Vermilion moraine, Leher and Hobbs (1992) have pro-posed that the Vermilion moraine is older than the St. Louis sublobe (see alsoRichmond and Fullerton, 1983). If this relationship is valid, the Vermilion phase couldbe partly contemporary with the Automba phase of the Superior Lobe, as depicted inTable III. Deposits from near the Unites States–Canada boundary at Lake of the Clouds(Stuvier, 1975) indicate deglaciation before 9600 yr B.P. (Table I, Figures 5a and 5b).

Continued retreat of the Rainy Lobe formed several moraines in southern Canada(Figure 5c). Based on radiocarbon dates on low organic sediments from Rattle Lake(Bjork, 1985; Figure 2), the Rainy Lobe was near the Eagle-Finlayson-Brule morainesometime around 11,310–10,910 yr B.P. (2-sigma range; Table I, Figure 5c).Radiocarbon ages of 10,250 (ETH-31430) and 10,190 (Beta-195959) yr B. P. north ofthe Brule moraine may provide an indication of the timing of deglaciation (Loope etal., 2006). By the time of the Superior Lobe Marquette advance (which formed theMarks ice margin in Ontario), the Rainy Lobe may have been in the vicinity of theHartman-Dog Lake moraine (Figure 7), although Bjork (1985) has proposed that the

HILL


Figure 5b. Paleogeographic model of ice margins associated with Nickerson phase of Superior Lobeand Alborn phase of St. Louis sublobe: Initial melting of St. Louis sublobe, formation of Lake Koochiching.See Hobbs (1983) for configuration of St. Louis sublobe and proglacial lakes (after Figures 7, 8, and 10in Hill, 1995b).

Sioux Lookout moraine is correlated with the Marquette advance, based on radio-carbon dates from Sioux Pond (Figure 2, Table I). A radiocarbon age on wood fromCummins pond (Julig et al., 1990; Figure 2) indicates this region was deglaciatedsometime before 9400–9120 yr B.P. (2-sigma range; Table I). There are older agesfrom Cummins and Oliver Ponds, but they are on marls (Table I). A series of reces-sional moraines show continued retreat to the Whitewater moraine (Figure 3) per-haps slightly before 9000 yr B.P. (based on radiocarbon dated organic matter fromIndian Lake; Table I; Bjork, 1985; Dredge and Cowan, 1989; Klassen, 1989).

Koochiching Lobe and St. Louis Sublobe

The Koochiching Lobe expanded eastward into the region west of Lake Superior.During the Alborn phase, the St. Louis sublobe advanced from the northwest (Figure5a, 5b), incorporating sediments from the Lake Aitkin and Upham basins that were



33

Figure 5c. Paleogeographic model of ice margins associated with Nickerson phase of Superior Lobe andAlborn phase of St. Louis sublobe: Landscape model with remnants of St. Louis sublobe and proglaciallakes along Koochiching and Superior lobes. See Hobbs (1983) for configuration of St. Louis sublobe andproglacial lakes (after Figures 7, 8, and 10 in Hill, 1995b).

deposited after the melting of the Rainy Lobe (Figure 4). Radiocarbon dates fromMariska Mine (Figure 2) on wood from sediments identified as St. Louis sublobe-Alborn phase till (Winter, 1971) have a combined 2-sigma range of about 12,000–10,300yr B.P. (Table I). Wood from deposits overlying Alborn phase till at Aitkin (Farnhamet al., 1964) has a 2-sigma range of about 12,400–10,800 yr B.P (Table I). One hypo-thetical chronology would thus include an advance of the St. Louis sublobe sometimeafter 12,000 yr B.P. with melting and the presence of some unglaciated terrain in theregion by about 11,000 yr B.P. and perhaps earlier. There are several buried soils strati-graphically between the Alborn phase till and the lake deposits with the radiocarbondated wood at Aitkin (Farnham et al., 1964). These paleosols reflect landscapes thatwould have been available after the melting of the St. Louis sublobe and before the for-mation of the lake. Melting of the St. Louis sublobe and other Koochiching Lobe iceled to the presence of the lake in the Upham basin (Hobbs, 1983).

LATE GLACIAL LANDSCAPES

Regional Late Glacial deglaciated terrains and ice margin environments includedabandoned lake shorelines in the Superior basin and interior lake and fluvial set-tings (such as the Cloquet River drainage and settings associated with transgres-

HILL


Figure 6. Paleogeographic model for western Lake Superior area during Moorhead phase regression ofLake Agassiz (after Figure 12 in Hill, 1995b).

sions and regressions in the glacial Lake Agassiz basin). Summaries of the exten-sive literature pertaining to changing lake levels in the Superior basin (cf., Farrandand Drexler, 1986) are presented by Phillips and Hill (1993, 1994a, 1994b, 1994c,2001, 2004) and Phillips et al. (1994). Pro-glacial lake sediments were likely depositedduring the melting of the St. Croix phase ice sheet (Figure 3) because they wereincorporated into the younger Automba phase deposits (Figure 4). High abandonedlake shorelines may indicate separate glacial lakes or a single extensive lake alongthe north and southwest shores of the Superior basin prior to and during theMarquette phase (Phillips and Hill, 2001, 2004; Figure 7). Shorelines thought to be asso-ciated with isolated high lakes along the Superior basin, such as Lake Nemadji andOntonagon, may have been present sometime after 11,220 yr B.P. (Figure 5c). Forinstance, a log in lake clay with a 2-sigma radiocarbon age range of 11,220–9220 yrB.P. is associated with glacial Lake Ontonagan in Michigan (Figure 2) and has beencorrelated with an early stage of Lake Duluth (Lake Nemadji; Table I; Crane, 1965).The older shorelines would not have been destroyed by the Marquette advance ifthe glacial margin did not reach their locations—thus, they are surviving features



35

Figure 7. Marquette phase of Superior Lobe, and Emerson phase of Lake Agassiz (after Figure 13 in Hill,1995b).

potentially dating to before 10,000 yr B.P. For example, south of Lake Superior, Saxonmoraine till associated with the Marquette advance and located below glacial LakeDuluth shorelines contains wood with a 2-sigma range of 10,300–9900 yr B.P. (TableI, Figure 2).

High (older) abandoned shorelines are preserved southwest of Duluth and extendalong the north and south shores of Lake Superior. For example, shoreline featuresat Duluth are present at elevations around 338–348 masl (meters above sea level).Some of these features are present directly north of the Math-Geology Building atthe University of Minnesota–Duluth and other buildings that, at times, housed theArchaeometry Laboratory (Figure 8). Lake elevations were likely higher than or at theDuluth level (~327–330 masl) in the southwestern part of the basin as water drainedthrough the Portage channel, and at slightly lower elevations during drainage throughthe Brule channel (Figure 2). A radiocarbon date on plant material from Brule chan-nel fill has a 2-sigma range of about 9170–8930 yr B.P. (Engseth, 1998), providing a min-imum age for its drainage (Table I). An older age is available for fill in the Blackhoofdiversion channel (Figure 2, Tables I and II); the measurement has a large standarddeviation with a 2-sigma range of about 11,600–9600 yr B.P. (Wright and Watts, 1969).The Marquette advance filled much of the Superior basin with ice, but perhaps not onthe extreme southwest margin or the Minnesota section of the north shore (Figure 7).Lake elevations in these areas were at Lake Duluth levels or higher.

Melting of the Marquette phase ice enlarged the area of the lake in the Superiorbasin and began to lower water to the Minong level again (Phillips and Hill, 2001,2004). The earliest of these recessional shoreline features is at about 307 masl in the

HILL


Figure 8. University of Minnesota, Duluth campus. Lake clay over sand; glacial Lake Duluth. Thesedeposits are adjacent to buildings used by the Archaeometry laboratory. Photograph courtesy of CharlesL. Matsch; photograph taken in 1972.

Duluth area. This shoreline extends along the entire Minnesota north shore reach-ing (because of tilt caused by postglacial rebound of the land) a present elevation of386 masl at Grand Portage, near the Canadian border. Three less developed shore-lines, designated Highbridge, Moquah, and Washburn, are between the sub-Duluthlevel and the well-developed Beaver Bay shoreline. Several other lake margins havealso been recognized between the Beaver Bay and Minong levels (Phillips and Hill,1993, 1994a). By about 9500 yr B.P., Lake Minong shorelines formed around the entiremargin of the Superior basin. Lake levels continued to drop from 9500 to 8000 yrB.P. The extreme low around 8000 yr B.P is designated the Houghton level (Phillipsand Hill, 2001; 2004).

Interior Lakes and Rivers

The Cloquet River drainage provides an example of the interior settings formedafter the melting of St. Croix phase ice (Figures 1 and 2; Hill, 1995a). Meltwatersfrom the Automba phase (Superior Lobe, Highland moraine) and the RainyLobe–deposited outwash over St. Croix phase deposits along the Cloquet River val-ley and then into Lake Aitkin-Upham basins (pre-Alborn phase Aitkin-Upham; Figure4). This region remained unglaciated throughout the Late Glacial period, although gla-cial advances occurred to the east in the Superior basin and to the west associatedwith the St. Louis sublobe (Figure 5). Radiocarbon ages indicate that the depositionof peat had begun by about 9800–8900 yr B.P. (Bog 2; Table I, Figure 2), althoughthe landscape was presumably inhabitable after the melting of St. Croix phase ice.

Two lake episodes in the Upham basin are separated by the Alborn phase advanceof the St. Louis sublobe (Figures 4 and 5). The extent and age of the lakes is con-strained by the retreat of the Rainy Lobe (the end of the St. Croix phase) and theadvance of the St. Louis sublobe (Alborn phase). Melting of the St. Louis sublobebegan in the area south of the Giants Range, creating a lake in the Upham basin (gla-cial Lake Upham; Figure 5b, 5c). As more ice melted, the Lake Aitkin basin was filledwith water (Upham, 1899; Winchell, 1899, 1901; Hobbs, 1983; Mooers, Marlow, et al.,2005). Remnants of these lakes may have existed to about 9000 yr B.P. (Hobbs, 1983),although the age is on marl (Tables I and II).

A series of lakes formed to the north and west of the Aitkin and Upham plains,eventually leading to the development of glacial Lake Agassiz. As the KoochichingLobe melted to the west, Lakes Norwood and Koochiching formed (Figure 5a, 5b).Further melting led to the expansion of Lake Koochiching (early Cass Agassiz; Fentonet al., 1983) dating to around 11,600 yr B.P. At this time, the Rainy Lobe may have beenin the vicinity of the Steep Rock moraine in Canada (Antevs, 1951; Nielsen et al.,1982; Teller and Thorleifson, 1983; Fisher et al., 2005; Lowell et al., 2005). When theRainy Lobe was between the Eagle-Finlayson-Brule and Hartman-Dog Lake moraines,Lakes Koochiching and Climax coalesced to form the Lockhart stage of glacial LakeAgassiz, around 11,000 yr B.P. (Figure 5c). Retreat of ice from the Superior basinopened the previously blocked eastern outlets for Lake Agassiz around 10,800 yrB.P., lowering the water level to the Moorhead stage of Lake Agassiz (Ashworth etal., 1972; Figure 6). Wood and charcoal radiocarbon ages associated with the



37

Moorhead regression have a combined 2-sigma range of 10,720–9700 yr B.P. (Bajc etal., 2000; Fisher and Lowell, 2006; Table I).

By about 10,000–9500 yr B.P., the Rainy Lobe was near the Hartman-Dog Lakemoraine, and the Superior Lobe had reached the Marks-Mackenzie moraine in Ontarioand the Gribben forest region south of Lake Superior (Figure 7). The Marquetteadvance of the Superior Lobe blocked the eastern outlet of Agassiz and led to the startof the Emerson stage transgression of Lake Agassiz (Bajc et al., 2000; Teller, 2001).The timing of the Emerson transgression is constrained by radiocarbon dates fromSwift (Bjork and Keister, 1983; Figure 2), where wood and peat underlying beachgravels have a combined 2-sigma radiocarbon age range of 10,250–9150 yr B.P., as wellas other Rainy River basin–ages on wood (Bajc et al., 2000) from Emerson phasedeposits (Tables I and II). The Lake Agassiz plain was exposed and presumably avail-able for habitation in northern Minnesota by about 9300 yr B.P. (Heinselman, 1963;Severson et al., 1980; Leverington et al., 2000).

PALEOBIOTIC AND ARCHAEOLOGICAL CONTEXTS

Paleoindian occurrences are known to be associated with both streams that flowdirectly into the Superior basin and interior lake and river settings (Figure 1).Geomorphic contexts within the Superior basin are summarized by Phillips and Hill(2001) and Phillips et al. (1994). Archaeological sites in the internal drainages havealso been documented (Schneider, 1982; Romano and Johnson, 1990; Harrison, 1995;Peters et al., 1995; Ross, 1997; Mulholland et al., 1997; Mulholland and Shafer, 2000).

In terms of site visibility and site-formation processes, some records of Paleoindianpresence prior to about 10,000 yr B.P. could have been eroded or buried by theMarquette ice margin advance, if the advance actually reached the north shore of thebasin (Phillips and Hill, 2001). Sites on higher shoreline features situated above the levelof the ice sheet would not have been affected by this ice advance (Figure 7). Examplesof these landscape conditions are present along the south side of the margin (Crane,1965; Hack, 1965). Oscillating water levels in the Superior basin resulting from dis-charge events from Lake Agassiz would have caused redeposition, as has been inter-preted at the Cummins site (Julig et al., 1990; Figure 2). Any post-Minong occupationsassociated with the recessional beaches preceding the Houghton low stand at around8000 yr B.P. would potentially have been inundated or eroded by the later transgres-sion. The archaeological discontinuity between Plano-type Paleoindian and Archaic siteshas been addressed by Phillips (1993). The shoreline features associated with variousglacial lake levels have the most potential for containing Paleoindian sites in the areasthat directly drain into the Superior basin. If the Marquette ice margin was separatedfrom the northwest side of the basin by a glacial lake (Figure 7), it is possible that theshoreline was occupied immediately after melting of the Automba phase ice (Figure4). Paleoindian sites could also be located on higher shoreline features of Lake Duluth.For example, unfluted lanceolate points similar to Agate Basin forms have been linkedto a glacial Lake Duluth beach (Mulholland and Dahl, 1988); the range of radiocarbonages (1-sigma) for Agate Basin points on the northern Plains is about 11,400–9500 yrB.P. (Holliday, 2000; Sellet, 2001).

HILL


Based on the availability of older deglaciated landscapes, the region west of theSuperior basin (west of the Highland moraine) and south of the Rainy Lobe margin(Vermilion moraine complex area) has the potential for containing early Paleoindiansites (Figure 4). Outwash from melting ice was deposited along the Cloquet Riverdrainage. Mammoth remains found in gravels within the Cloquet River drainage(Stauffer, 1945) may be associated with these settings, although detailed chrono-logic information is unavailable for these fossils. Unless they reflect pre-Last GlacialMaximum materials that have been incorporated into St. Croix phase till, the Cloquetmammoth fossils are probably Late Glacial in age (contemporary with the Autombaphase or younger).

In the internal drainages removed from the Superior basin (Figure 1), archaeo-logical visibility has been both helped and hindered by the 20th century creation ofreservoir lakes on some of the major tributaries (Harrison et al., 1995). These reser-voirs have inundated portions of major rivers along which sites are situated (Hill,1995a). Oscillating water levels have made artifact occurrences more visible (Harrisonet al., 1995), but they have also produced lag accumulations. The effect of postoc-cupational erosion and redistribution on the spatial patterning and composition ofthe cultural assemblages could be high in some cases. Also, aggradation and peatformation have buried available terrains after the initial deglaciation (Hill, 1995a).

If the Clovis point recovered from the Island Lake area (Figure 2; Mulholland etal., 1997; Romano and Johnson, 1990) can be used as a time-marker, it implies humanpresence between 11,700 and 9600 yr B.P. (Holliday, 2000). The geomorphic associ-ation of Agate Basin and Hell Gap (Plano-type) artifacts, such as those found in theCloquet River drainage (Harrison, 1995; Hill, 1995a), appear to indicate a humanpresence between about 11,400 and 9500 yr B.P. (Holliday, 2000; Sellet, 2001).Paleoindian sites in the vicinity of Island Lake (Hill, 1995a) were situated alongstreams and lakes within Automba phase outwash (overlying St. Croix phase till). Theupper reaches of the Cloquet River drainage also contain a site with a Paleoindiancomponent (at Cloquet Lake; Figure 2; Mulholland et al., 1997); the site is in an areacomposed of Automba phase till of the Superior Lobe. The site is situated in a geo-morphic terrain similar to the Paleoindian localities in the Island Lake area. West ofthe Cloquet River, a series of Plano-type Paleoindian sites are located in the WhitefaceRiver drainage (Figures 6 and 7; Mulholland et al., 1997; Mulholland and Shafer,2000). In this region, deposits of the Rainy Lobe are overlain by outwash deposits andtill from the St. Louis sublobe (Alborn phase). Thus, while the area to the east mayhave been a deglaciated landscape, the area to the west was directly associated withthe Alborn phase ice margin.

Melting of the St. Louis sublobe and Rainy Lobe resulted in the creation ofpreglacial<zaq;1> and postglacial shorelines and drainages (Figures 5 and 6).Rancholabrean mammal communities, including human populations, could haveinhabited the region between the melting St. Louis sublobe and the ice within theSuperior basin (Figure 5). Biotic communities could have been established by taxamigrating up the Cloquet River drainage and then northward toward the ice margin.Melting of Alborn phase ice would have led to the creation of shorelines within theLake Upham basin; these would have been available to Paleoindian populations.



39

Paleosols under radiocarbon-dated wood at Aitkin (Figure 2) would have been avail-able for habitation sometime before 10,800 yr B.P. (based on the earliest date of the2-sigma range; Table I). Later deglaciation provided accessible landscapes associatedwith the Agassiz basin.

Melting of the St. Louis sublobe resulted in an expansion of Lake Aitkin (see pale-ogeographic reconstructions in Hobbs, 1983), and these new shorelines may be asso-ciated with late Clovis and early Folsom artifacts (based on the Aitkin wood date;Table I). By about 10,000 yr B.P., Lake Upham drained through the St. Louis River andinto a lake in the Superior basin that discharged southward into the St. Croix Rivervia the Portage and Kettle channels. At about the same time, Lake Aitkin first drainedthrough the Snake River channel and eventually the Mississippi River (Hobbs, 1983).The remnants of these lakes were probably gone by the time human populationsassociated with later Plano artifacts were present in the region (~9000–8000 yr B.P.).Studies of the buried soils, lake deposits, and recessional lake margin features mighthelp to clarify the lake chronology and provide information on Paleoindian groupsin the region.

The Agassiz basin, including the Rainy River drainage (Figures 2, 6, and 7), alsocontains Paleoindian artifacts (Steinbring, 1974; Schneider, 1982; Pettipas andBuchner, 1983; Buchner and Pettipas, 1990). Finds are also situated along lakes andrivers in the vicinity of the Vermilion moraine complex and near the border withCanada (Peters et al., 1995; Phillips and Hill, 2001). Several Paleoindian sites areassociated with the Vermilion moraine complex (Peters et al., 1995; Mulholland etal., 1997), including a site near Big Rice Lake (Figure 2) situated in an area of proglaciallake sediments. The location could have been part of Glacial LakeNorwood/Koochiching until about 12,500 yr B.P. (Leher and Hobbs, 1992). Big RiceLake contains basal organic sediments with a 1-sigma age range of about12,600–11,500 (Huber, 1988), thus indicating that the glacial margin was north of thisregion prior to this time interval.

Other Paleoindian sites lie between the Vermilion moraine and the Eagle-Finlayson-Brule moraine, which marks the ice margin at perhaps about 11,000 yr B.P. (Figure5c). For example, a Holcombe-like (basally thinned) artifact was recovered fromthe Bearskin Lake area (Okstad et al., 2000). If the age estimates are reasonable, theentire area was potentially habitable by late Rancholabrean faunal communities andhuman populations using Clovis artifact forms, as well as later Paleoindian popula-tions. For instance, McLeod (2001) has proposed that artifacts west of Thunder Baymay indicate the presence of human populations in that region prior to the Marquetteadvance.

The chronology of Lake Agassiz is important in terms of potential early humanoccupation of the Rainy River system (Buchner and Pettipas, 1990). The Lockhartphase, which may date to before 12,000–11,000 yr B.P., is associated with Hermanbeach lines, as well as Norcross, Tintah, and Campbell levels (Teller, 2001). Woollymammoth (Mammuthus primigenius) remains have been recovered from depositsforming the Herman strandline (Harington and Ashworth, 1986). The beginning of theMoorhead phase is marked by a low water stand starting at about 11,000 yr B.P.(Figure 6). A return to about the high Campbell strandline is associated with the

HILL


Emerson phase (sometime after 10,000 yr B.P.; Fenton et al., 1983; Leher and Hobbs,1992; Leverington et al., 2000). This transgression flooded the previously exposed land-scape that had been available for habitation during the Moorhead phase (Figure 6).The lake abandoned the Campbell beach line and withdrew into Canada betweenabout 9500 and 7000 yr B.P. (Johnston, 1946; Klassen, 1989). Based on this chronol-ogy, any Paleoindian presence dating to 11,000 yr B.P. would be associated with fea-tures associated with the final part of the Lockhart phase (Table III). Artifacts linkedto Paleoindian groups inhabiting the area exposed during the Moorhead-phase regres-sion (Figure 6) would have been eroded or buried by the flooding associated withthe Emerson=phase transgression (Figure 7, Table III). Potential Paleoindian habi-tation between about 10,000 and 9500 yr B.P. could be associated with Emerson-phase Lake Agassiz features. Within this context, the fluted point found north ofLake Winnibigoshish (Figure 1) near Round Lake (Peterson in Ross, 1997) is in aregion containing Lake Agassiz sediments and Koochiching Lobe till (Meyer, 1993;Richmond and Fullerton, 1983).

Archaeological sites are situated close to major tributaries of the Rainy River orin contexts related to glacial Lake Agassiz. There are two sites where Plano-typePaleoindian artifacts have been found: the Pelland site and the Plummer site(Stoltman, 1971). Stratigraphic sequences situated north of the Pelland site (Figure6) suggest that the low-water Moorhead stage of Lake Agassiz started around 11,000yr B.P. and ended with a rise to at least Campbell levels around 10,000 yr B.P.(Emerson phase transgression; Bajc et al., 2000; Nielsen et al., 1982). The Plummersite (Figure 6) may be associated with outwash deposits on the Lake Agassiz plain.It seems to be situated on a terrace of a previous channel of the Rainy or BlackRivers. Further to the west, the Greenbush site contains Plano-type (Scottsbluff vari-ant) artifacts and is situated on the Campbell shoreline of Lake Agassiz (Nystuenand Peterson, 1972; Peterson, 1972).

SUMMARY

An integrative geoarchaeological approach can help to evaluate the chronologic rela-tionship between deglaciation dynamics and time-diagnostic artifacts forms, as wellas the habitats available to Late Glacial Rancholabrean fauna and Paleoindian groups.Geomorphic and stratigraphic relationships of Quaternary deposits can be used toreconstruct the physical settings potentially available to late Rancholabrean-age fau-nal communities and Paleoindian populations in the western region of Lake Superior.To the east, in the Superior basin, lakes fluctuated in elevation and extent. Shorelinesassociated with the Duluth level and higher shorelines could have been occupied byhuman populations using either fluted or lanceolate artifact forms. Artifacts depositedon Minong levels may have been buried or reworked as a result of a lake transgres-sion associated with the Marquette ice advance, or by the advance of this ice sheet ifit reached the north shore of the basin. This transgression started around 10,000 yrB.P. and lasted until about 9500 yr B.P. Thus, artifacts associated with occupations priorto or during this interval may be associated with Duluth or higher lake levels. Afterabout 9500 yr B.P., occupations would have been initially associated with shoreline



41

features of Lake Minong. Continued lowering of the lake formed recessional beachesthat would contain occupations dating to about 8000 yr B.P.

While glacial activity persisted to the east in the Superior basin and to the westin the Agassiz basin, the Rainy Lobe melted away, leaving a deglaciated terrain in thecentral portion of the area. An ice-free region extended north to the Vermilion morainecomplex probably by about 12,000 yr B.P. By about 11,000 yr B.P., the ice margin ofthe Rainy Lobe was in southern Canada, leaving the central portion of northeasternMinnesota available for Rancholabrean faunal communities and Paleoindian popu-lations associated with Clovis artifacts, as well as other fluted and unflutedPaleoindian artifact forms. These deglaciated landscapes are partly contemporarywith ice margin habitats elsewhere in the Great Lakes that contain radiocarbon-dated Rancholabrean fauna, sometimes in association with nontime diagnostic arti-facts (cf. Overstreet and Kolb, 2003).

Melting of the St. Louis sublobe starting ~11,000–9000 yr B.P. created landscapesconnected with changing configurations of the lake shorelines within the Uphamand Aitkin basins. These landscape features may contain a suite of fluted and Plano-type artifact forms. To the north and west, in the Agassiz basin, Paleoindian artifactsmay be associated with higher shoreline features associated with the Lockhart phase(~11,000 yr B.P.). Recessional beaches of the low water Moorhead stage may beassociated with later fluted and unfluted artifact forms. Artifacts at these sites wouldhave been buried or possibly eroded and redeposited when Lake Agassiz returnedto higher levels during the Emerson stage. Early Plano artifacts might be expectedon these higher beaches, as well as later recessional beaches.

The western region of the Lake Superior-Agassiz basin provides a record of land-scape evolution connected with deglaciation dynamics from the Last GlacialMaximum through the Preboreal chronozone. This time interval, containing thePleistocene/Holocene boundary, was associated with the extinction of Rancholabreanmegafauna and the presence of time diagnostic fluted and unfluted points. Remainsof extinct Late Rancholabrean mammals, including Mammuthus, have been recov-ered from the Agassiz basin (Herman beach), the upper Mississippi drainage, and partsof the Cloquet River drainage, although they have yet to be found in direct associa-tion with artifacts. Clovis, Folsom, Holcombe, Agate Basin, and Hell Gap artifactforms, elsewhere dated to the Late Glacial interval, have been found within theregion, and landscapes were available at this time that could contain older nondiag-nostic artifact forms associated with Rancholabrean fauna.

Geomorphic and stratigraphic relationships of Quaternary deposits in the west-ern Lake Superior region can be used to construct interpretations of the physicalsettings potentially available to late Rancholabrean-age faunal communities andPaleoindian populations. Between about 20,000 and 14,000 yr B.P., the Rainy Lobemelted away from the region between the Agassiz and Superior basins. By perhaps12,500 yr B.P., an ice-free region extended to about the location of the Vermilionmoraine, and by 11,000 yr B.P. the ice margin was in southern Canada. This left thearea immediately west of the Superior basin available for Rancholabrean faunalcommunities and Late Glacial human populations using fluted and unfluted points.Isolated ice-margin lakes may have been present within the Superior basin as a result

HILL


of the Marquette advance, which also resulted in the Emerson transgression of LakeAgassiz. Geoarchaeological models that integrate interpretations of the spatial andtemporal landscapes and assessments of the temporal range of Paleoindian artifactscan help provide a more complete understanding of Late Glacial biotic communitiesin the western Lake Superior region.

It is a special pleasure to dedicate this paper to Regents Professor of Geoarchaeology George (Rip) Rapp.I am especially grateful to Zhichun Jing for the opportunity to contribute this essay in honor of Rip, andthank him for his hard work and excellent humor. Special thanks to Rip and the many members of Rip’sArchaeometry Laboratory who have shared in the ongoing integration of geology, paleobiology, andarchaeology; I especially appreciate the kindness of Susan Mulholland, James Huber, and Seppo Valppu.Research on the glacial chronology of the Lake Superior region was initiated as part of Rip’s project onartifacts collected from the Island Lake area of Minnesota and was undertaken as a senior thesis underhis supervision. My introduction to Rip and interest in Quaternary geology was the result of a long-timefriendship with Charlie Matsch; he kindly provided background information on Rip’s early career in gla-cial geology and the photograph used in Figure 8. The enjoyable collaborative research along the LakeSuperior basin was directed by Brian Phillips; he has conducted extensive research on the connectionsbetween Late Glacial paleogeographic contexts and Paleoindian archaeology. Research in the glacialLake Agassiz basin was undertaken as part of a project directed by Rip focused on the archaeology of theRainy River area. Parts of this paper were initially prepared for the symposium entitled “Archaeology,Geomorphology, and Paleoenvironment: Paleoindian Occupations in the Western Great Lakes” at the60th Annual Meeting of the Society for American Archaeology in 1995, organized by Susan Mulholland.I am very grateful to three anonymous reviewers for their suggestions and constructive criticisms.

REFERENCES<<zzaaqq;;1100>>

Anderton, J.B. (1999). The soil-artifact context model: A geoarchaeological approach to paleoshorelinesite dating in the Upper Peninsula of Michigan, USA. Geoarchaeology, 14, 265–288.

Antevs, E. (1951). Glacial clays in Steep Rock Lake, Ontario, Canada. Geological Society of AmericaBulletin, 62, 1223–1262.

Arneman, H.F., & Wright, H.E., Jr. (1959). Petrography of some Minnesota tills. Journal of SedimentaryPetrology, 29, 540–554.

Ashworth, A.C., Clayton, L., & Bickley, W.B. (1972). The Mosbeck Site: A paleoenvironmental interpreta-tion of Late Quaternary History of Lake Agassiz based on fossil insect and mollusk remains. QuaternaryResearch, 2<zaq;3>.

Attig, J.W.E., Clayton, L., & Mickelson, D.M. (1985). Correlation of Late Wisconsin Glacial phases in thewestern Great Lakes area. Geological Society of America Bulletin, 96, 1585–1593.

Bajc, A.F., Schwert, D.P., Warner, B.G., & Williams, N.E. (2000). A reconstruction of Moorhead and EmersonPhase environments along the eastern margin of Glacial Lake Agassiz, Rainy River Basin, northwest-ern Ontario. Canadian Journal of Earth Sciences, 37, 1335–1353.

Baker, R.G. (1965). Late-Glacial pollen and plant macrofossils from Spider Creek, southern St. LouisCounty, Minnesota. Geological Society of America Bulletin, 76, 601–610.

Birks, H.J.B. (1976). Late-Wisconsinan vegetational history at Wolf Creek, central Minnesota. EcologicalMonographs, 46, 395–429.

Birks, H.J.B. (1981). Late Wisconsin vegetational and climatic history at Kylen Lake, northeasternMinnesota. Quaternary Research, 16, 322–355.

Bjork, S. (1985). Deglaciation chronology and revegetation in northwestern Ontario. Canadian Journal ofEarth Sciences, 22, 850–871.

Bjork, S. (1990). Late Wisconsin history north of the Giants Range, northern Minnesota, inferred from com-plex stratigraphy. Quaternary Research, 33, 18–36.

Bjork, S., & Keister, C.M. (1983). The Emerson phase of Lake Agassiz, independently registered in north-western Minnesota and northwestern Ontario. Canadian Journal of Earth Sciences, 20, 1536–1542.



43

Black, R.F. (1976). Quaternary geology of Wisconsin and contiguous Upper Michigan. In Mahaney, W.C. (Ed.),Quaternary stratigraphy of North America (pp. 93–117). Stroudsburg, PA: Dowden, Hutchinson and Ross.

Brubaker, L.B. (1975). Postglacial forest patterns associated with till and outwash in northcentral UpperMichigan. Quaternary Research, 5, 499–526.

Buchner, A.P., & Pettipas, L.F. (1990). The early occupations of the Glacial Lake Agassiz Basin in Manitoba;11,500 to 7,700 B.P.. In N.P. Lasca & J. Donahue (Eds.), Archaeological Geology of North America (pp.51–59). Boulder, CO: Geological Society of America.

Clayton, L. (1984). Pleistocene geology of the Superior region, Wisconsin. Information Circular No. 46.Madison, WI: Wisconsin Geological and Natural History Survey.

Clayton, L., & Moran, S. (1982). Chronology of Late Wisconsin glaciation in middle North America.Quaternary Science Reviews, 1, 55–82.

Cleland, C.E. (1965). Barren ground caribou (rangifer arcticus) from an early man site in southeasternMichigan. American Antiquity, 30, 350–351.

Cleland, C.E., Holman, M.B., & Holman, J.A. (1998). The Mason-Quimby line revisited. The WisconsinArchaeologist, 79, 8–27.

Crane, H.R. (1965). University of Michigan radiocarbon dates I. Science 124, 664–672.Deller, D.B., & Ellis, C.J. (1992). Thedford II: A Paleo-indian site in the Ausable River watershed of south-

eastern Ontario. Museum of Anthropology, Memoirs, No. 24. Ann Arbor, MI: University of Michigan.Dredge, L.A., & Cowan, W.R. (1989). Quaternary geology of the southwestern Canadian Shield. In Fulton,

R.J. (Ed.), Quaternary geology of Canada and Greenland (pp. 189–214). Geology of North America, K-1. Boulder, CO: The Geological Society of America.

Dudzik, M.J. (1991). First people: The Paleoindian tradition in northwestern Wisconsin. The WisconsinArchaeologist, 72, 137–154.

Engseth, M.F. (1998). A model of Paleoindian site location along the Brule/St. Croix Spillway, Wisconsin.Unpublished master’s thesis, University of Minnesota, Minneapolis, MN.

Eriksen, B.V. (1996). Resource exploitation, subsistence strategies, and adaptiveness in Late Pleistocene-Early Holocene northwest Europe. In L.G. Straus, B.V. Eriksen, J.M. Erlandson, & D.R. Yesner (Eds.),Humans at the end of the Ice Age: The archaeology of the Pleistocene–Holocene transition (pp.101–128). New York: Plenum.

Farnham, R.S., McAndrews, J.H., & Wright, H.E., Jr. (1964). A Late-Wisconsin buried soil near Aitkin,Minnesota, and its paleobotanical setting. American Journal of Science, 262, 393–412.

Farrand, W.R. (1977). Revision of the Two Rivers (“Valders”) drift border and the age of fluted points inMichigan. In C.E. Cleland (Ed.), For the director: Research essays in honor of James B. Griffin (pp.74–84). Museum of Anthropology Anthropology Papers No. 61. Ann Arbor, MI: Michigan: Universityof Michigan.

Farrand, W.R., & Drexler, C.W. (1986). Late Wisconsinan and Holocene history of the Lake Superior Basin.In P.F. Karrow & P.E. Calkin (Eds.), Quaternary evolution of the Great Lakes (pp. 33–37). The GeologicalAssociation of Canada Special Paper 30. St. Johns, Newfoundland: Geological Association of Canada.

Fenton, M.M., Moran, S.R., Teller, J.T., & Clayton, L. (1983). Quaternary stratigraphy and history in the south-ern part of the Lake Agassiz Basin. In J.T. Teller & L. Clayton (Eds.), Glacial Lake Agassiz (pp. 49–74).The Geological Association of Canada, Special Paper 26. St. Johns, Newfoundland: GeologicalAssociation of Canada.

Fisher, D.C. (1984). Mastodon butchery by North American Paleoindians. Nature, 308, 271–272.Fisher, D.C. (1987). Mastodont procurement by Paleoindians of the Great Lakes Region: Hunting or scav-

enging? In M.H. Nitecki & D.V. Nitecki (Eds.), The evolution of human hunting (pp. 309–421). New York:Plenum Press.

Fisher, D.C. (1996). Extinction of proboscideans in North America. In J. Shoshani & P. Tassy (Eds.), Theproboscidea: Evolution and paleoecology of elephants and their relatives (pp. 296–315). Oxford:Oxford University Press.

Fisher, T.G., & Lowell, T.V. (2006). When was the Moorhead low-water phase of Glacial Lake Agassiz?Geological Society of America Abstracts with Programs, 38(4), 57<zaq;3>.

Fisher, T.G., Lowell, T.V., Glover, K., & Hajdas, I. (2005). A later deglaciation of northwestern Ontario:Implications for eastern Lake Agassiz drainage. Geological Society of America Abstracts with Programs,37(5), 21<zaq;3>.

HILL


Florin, M., & Wright, H.E., Jr. (1969). Diatom evidence for the persistence of stagnant glacial ice inMinnesota. Geological Society of America Bulletin, 80, 695–704.

Fries, M. (1962). Pollen profiles of Late Pleistocene and recent sediments from Weber Lake, Minnesota.Ecology, 43, 295–308.

Glover, K.C., Lwell, T.V., Fisher, T., Hadjas, I., Waterson, N., Larson, P., et al. (2006<zaq;4>). Additionalradiocarbon control for the deglaciation and subsequent landscape development of northeasternMinnesota. Paper presented at the Tenth International Paleolimnology Symposium, Duluth, MN.<zaq;2>

Hack, J.T. (1965). Postglacial drainage evolution and stream geometry in the Ontonagon area, Michigan.Professional Paper 504-C. Washington, DC: United States Geological Survey.

Harington, C.R., & Ashworth, A.C. (1986). A mammoth (Mammuthus Primigenius) tooth from LateWisconsin deposits near Embden, North Dakota, and comments on the distribution of woolly mam-moths south of the Wisconsin Ice Sheets. Canadian Journal of Earth Sciences, 23, 909–918.

Harrison, C. (1995). Late Paleo-indian manifestations around the western Superior Basin. In C. Harrison,E. Redepenning, C.L. Hill, G.J. Rapp, S.E. Aschenbrenner, J.K. Huber, et al. (Eds.), The Paleo-indianof southern St. Louis County, Minnesota: The Reservoir Lakes complex (pp. 6–16). Dubuque, IA:Kendall/Hunt Publishing.

Harrison, C., Redepennning, E., & Hill, C.L. (1995). Site description. In C. Harrison, E. Redepenning, C.L.Hill, G.J. Rapp, S.E. Aschenbrenner, Huber, J.K., et al. (Eds.), the Paleo-indian of southern St. LouisCounty, Minnesota: The Reservoir Lakes complex (pp. 101–135). Dubuque, IA: Kendall/Hunt Publishing.

Haynes, G. (2002a). The catastrophic extinction of North American mammoths and mastodonts. WorldArchaeology, 33, 391–416.

Haynes, G. (2002b). The early settlement of North America: The Clovis Era. Cambridge/New York:Cambridge University Press.

Heinselman, M.L. (1963). Forest sites, bog processes, and peatland types in the Glacial Lake Agassizregion, Minnesota. Ecological Monographs, 33, 327–374.

Hill, C.L. (1994<zaq;4>). Quaternary stratigraphy and geomorphology in east-central and north-eastMinnesota. In Guidebook for field trip E, 13th Biennial Meeting of the American Quaternary Association(pp. 5–18), Minneapolis, MN.

Hill, C.L. (1995a). Geological context of Paleo-indian occurrences. In C. Harrison, E. Redepenning, C.L.Hill, G.J. Rapp, S.E. Aschenbrenner, J.K. Huber, et al. (Eds.), The Paleo-indian of southern St. LouisCounty, Minnesota: The Reservoir Lakes complex (pp. 89–100). Dubuque, IA: Kendall/Hunt Publishing.

Hill, C.L. (1995b<zaq;4>). Quaternary stratigraphy and geomorphic context for Paleoindian occurrencesin Northeastern Minnesota. Paper presented at the 60th Annual Meeting of the Society for AmericanArchaeology, Minneapolis, MN.

Hill, C.L., & Huber, J.K. (1996). Geo-ecologic dynamics in the Glacial Lakes Duluth-Agassiz region (north-ern Minnesota) during the Late Pleistocene and Early Holocene. Current Research in the Pleistocene,13, 119–121.

Hill, C.L., Rapp, G., Jr., & Huber, J.K. (1985). Final Pleistocene and Holocene pollen stratigraphic sequencefrom the Cloquet River area, St. Louis County, northeastern Minnesota. Geological Society of AmericaAbstracts with Programs, 17, 610<zaq;3>.

Hill, C.L., Rapp, G., Jr., Valppu, S., & Jing, Z. (1995). Geoarchaeology and geochronology of the Hannafordsite. Archaeometry Laboratory Report Number 95–7. Duluth, MN: Archaeometry Laboratory, Universityof Minnesota.

Hobbs, H.C. (1983). Drainage relationship of Glacial Lakes Aitkin and Upham and Early Lake Agassiz innortheastern Minnesota. In J.T. Teller & L. Clayton (Eds.), Glacial Lake Agassiz (pp. 245–259). GeologicalAssociation of Canada Special Paper 26. St. Johns, Newfoundland: Geological Association of Canada.

Holliday, V.T. (2000). The evolution of Paleoindian geochronology and typology on the Great Plains.Geoarchaeology, 14, 227–290.

Holman, J.A. (2001). In quest of Great Lakes Ice Age vertebrates. East Lansing, MI: Michigan StateUniversity Press.

Huber, J.K. (1988). Use of green algae abundances for augmenting palynological interpretation. CurrentResearch in the Pleistocene, 5, 48–50.

Huber, J.K. (2001). Palynological investigations related to archaeological sites and the expansion of wildrice (Zizaniz Aquatica L.) in northeast Minnesota. Duluth, MN: University of Minnesota.



45

Huber, J.K., & Hill, C.L. (1987). A pollen sequence associated with Paleoindian presence in northeasternMinnesota. Current Research in the Pleistocene, 4, 89–91.

Hughes, J.D., & Merry, W.J. (1978<zaq;4>). Marquette buried forest 9,850 years old. Paper presented atthe 144th Annual Meeting of the American Association for the Advancement of Science, Washington,DC.

Irwin, H.T., & Wormington, H.M. (1970). Paleo-Indian tool types in the Great Plains. American Antiquity,35, 24–34.

Jackson, L.J. (1988). Fossil cervids and fluted point hunters: A review for southern Ontario. OntarioArchaeology, 48, 27–41.

Jackson, L.J. (1989). Late Pleistocene caribou from northern Ontario. Current Research in the Pleistocene,6, 72–74.

Jackson, L.J., Ellis, C., Morgan, A.V., & McAndrews, J.H. (2000). Glacial Lake levels and eastern GreatLakes Palaeo-indians. Geoarchaeology, 15, 415–440.

Janssen, C.R. (1968). Myrtle Lake: A Late- and Post-Glacial pollen diagram from northern Minnesota.Canadian Journal of Botany, 46, 1397–1408.

Johnson, M.D., & Mooers, H.D. (1998). Ice-margin positions of the Superior lobe during Late Wisconsinandeglaciation. In C.J. Patterson & H.E. Wright, Jr. (Eds.), Contributions to Quaternary studies inMinnesota (pp. 7–14). Minnesota Geological Survey Report of Investigations 49. St. Paul, MN: MinnesotaGeological Survey.

Johnston, W.A. (1946). Glacial Lake Agassiz, with special reference to the mode of deformation of thebeaches. Canada Geologic Survey Bulletin no. 7. Ottawa: Kings Printer and Controller of Stationery.

Julig, P.J. (1991). Late Pleistocene archaeology in the Great Lakes region of North America: Current prob-lems and prospects. Revista de Arqueologia Americana, 3, 7–30.

Julig, P.J., & McAndrews, J.H. (1993). Paleoindian cultures in the Great Lakes region of North America:Paleoclimatic, geomorphic, and stratigraphic context. L’Anthropologie, 97, 623–650.

Julig, P.J., McAndrews, J.H., & Mahaney, W.C. (1990). Geoaracheology of the Cummins site on the beachof proglacial Lake Minong, Lake Superior Basin, Canada. In N.P. Lasca & J. Donahue (Eds.),Archaeological geology of North America (pp. 21–50). Boulder, CO: Geological Society of America.

Karrow, P.F., Dreimanis, A., & Barnett, P.J. (2000). A proposed diachronic revision of Late Quaternarytime-stratigraphic classification in the eastern and northern Great Lakes Area. Quaternary Research,54, 1–12.

Klassen, R.W. (1989). Quaternary geology of the southern Canadian interior plains. In R.J. Fulton (Ed.),Quaternary geology of Canada and Greenland (pp. 138–173). The Geology of North America, K-1.Boulder, CO: Geological Society of America.

Kuehn, S.R. (1998). New evidence for Late Paleoindian–Early Archaic subsistence behavior in the west-ern Great Lakes. American Antiquity, 63, 457–476.

Lannon, P.M., & Matsch, C.L. (1987). Quaternary stratigraphy and Glacial history of the Late WisconsinSuperior lobe, Duluth-Superior area. Geological Society of America Abstracts with Programs, 19,230<zaq;3>.

Laub, R.S., & Haynes, G. (1998). Fluted points, mastodons, and evidence of Late Pleistocene drought atthe Hiscock site, western New York State. Current Research in the Pleistocene, 15, 32–34.

Leher, J.D., & Hobbs, H.C. (1992). Glacial geology of the Laurentian divide area, St. Louis and LakeCounties, Minnesota. Minnesota geological survey field trip guidebook 18, prepared for the 39thMidwest Friends of the Pleistocene Field Trip, St. Paul, MN.

Lepper, B.T. (1999). Pleistocene peoples of midcontinental North America. In R. Bonnichsen & K.L.Turnmire (Eds.), Ice age people of North America: Environments, origins, and adaptations (pp.362–394). Corvallis, OR: Oregon State University Press.

LeVasseur, A.K. (2000). 10,000 Years in the headwaters: Archeology on the Chippewa National Forest.The Minnesota Archaeologist, 59, 11–21.

Leverington, D.W., Mann, J.D., & Teller, J.T. (2000). Changes in the bathymetry and volume of GlacialLake Agassiz between 11,000 and 9300 14C Yr B.P. Quaternary Research, 54, 174–181.

Loope, H.M., Fisher, T.G., Lowell, T.V., & Hadjas, I. (2006). Glaciological and chronological constraints ofLaurentide Ice Sheet occupation of the Superior Basin during the Younger Dryas. Geological Societyof America Abstracts with Programs, 38(4), 57<zaq;3>.

HILL


Lowell, T.V., Fisher, T.G., Comer, G.C., Hajdas, I., Waterson, N., Glover, K., et al. (2005). Testing the LakeAgassiz meltwater trigger for the Younger Dryas. EOS, Transactions of the American GeophysicalUnion, 86, 365-373.

Lowell, T.V., Larson, G.L., Hughes, J.D., & Denton, G.H. (1999). Age verification of the Lake Gribben for-est bed and the Younger Dryas advance of the Laurentide ice sheet. Canadian Journal of Earth Sciences,36, 383–393.

Lund, S.P., & Banerjee, S.K. (1985). Late Quaternary paleomagnetic field secular variation from twoMinnesota lakes. Journal of Geophysical Research, 90(B1), 803–825.

Matsch, C.L., & Schneider, A.F. (1986). Stratigraphy and correlation of the glacial deposits of the glaciallobe complex in Minnesota and Northwest Wisconsin. Quaternary Science Reviews, 5, 59–64.

McLeod, M. (2001). The Paleo-indian tradition in the Thunder Bay area, earliest evidence: Periglacial peo-ples. In B.A.M. Phillips & M. McLeod (Eds.), The deglaciation and geoarchaeology of theMinnesota–Ontario borderlands, a field guide to the excursions for the Midwest Friends of thePleistocene 47th Conference, Thunder Bay, Ontario.

Meyer, G.N. (1993). Surficial geologic map of parts of Koochiching, Itasca, and Beltrami Counties, north-central Minnesota, 1:250,000. Minnesota Geological Survey Miscellaneous Map Series, Map M-76.<zaq;5>

Mires, P.B. (1989). A fluted point find from Iron County, Wisconsin. The Wisconsin Archaeologist, 70,463–468.

Mooers, H.D., Larson, P.C., & Marlow, L.R. (2005). Ice advances in the western Lake Superior region: Areevaluation of the St. Louis sublobe and the Marquette phase of the Superior lobe. Geological Societyof America Abstracts with Programs, 37(5), 92<zaq;3>.

Mooers, H.D., Marlow, L.R., & Phillip, P.C. (2005). Glacial lakes in northern Minnesota: Drainage rela-tionships, chronology, and isotopic composition of meltwater. Geological Society of America Abstractswith Programs, 37(5), 22<zaq;3>.

Mulholland, S.C. (2000). The arrowhead since the glaciers: The prehistory of northeastern Minnesota.The Minnesota Archaeologist, 59, 1–10.

Mulholland, S.C., & Dahl, E. (1988). Two Late Paleoindian projectile points associated with a Glacial LakeDuluth beach, Carlton County, Minnesota. The Minnesota Archaeologist, 47, 41–50.

Mulholland, S.C., Mulholland, S.L., Peters, G.R., Huber, J.K., & Mooers, H.D. (1997). Paleo-Indian occu-pations in northeastern Minnesota: How early? North American Archaeologist, 18, 371–400.

Mulholland, S.C., & Shafer, J.R. (2000). The St. Louis River project: Precontact contexts in the Cloquet andWhiteface River drainages. The Minnesota Archaeologist, 59, 29–43.

Nielsen, E., McKillop, W.B., & McCoy, J.P. (1982). The age of the Hartman Moraine and the Campbellbeach of Lake Agassiz in northwestern Ontario. Canadian Journal of Earth Sciences, 19, 1933–1937.

Nystuen, D.W., & Peterson, L.D. (1972). The archaeological salvage of the Greenbush “Barrow Pit” site (21-R0-11). Report to the Commissioner of Minnesota Highways.<zaq;5>

Okstad, W., Woodward, D., & Crump, S. (2000). Overview of the Superior National Forest heritage pro-gram 1979–1999. The Minnesota Archaeologist, 59, 22–28.

Overstreet, D.F., & Kolb, M.F. (2003). Geoarchaeological contexts for Late Pleistocene archaeologicalsites with human-modified woolly mammoth remains in southeastern Wisconsin, U.S.A.Geoarchaeology, 18, 91–114.

Peters, G.R., Mulholland, S.C., Mulholland, S., & Larimer, G. (1995<zaq;4>). Paleoindian research innortheastern Minnesota: The land between complexes. Paper presented at the 60th Annual Meetingof the Society for American Archaeology, Minneapolis, MN.

Peterson, L. (1972). An early prehistoric stone workshop site in northwestern Minnesota. The MinnesotaArchaeologist, 32, 1–52.

Peterson, W.L. (1982). Preliminary surficial geologic map of the Iron River 1 degree by 2 degree Quadrangle,Michigan and Wisconsin. United States Geological Survey, Open-file Report, 82–301.<zaq;5>

Peterson, W.L. (1986). Late Wisconsinan glacial history of northeastern Wisconsin and western UpperMichigan. United States Geological Survey Bulletin 1652.<zaq;5>

Pettipas, L.F., & Buchner, A.P. (1983). Paleo-indian prehistory of the Glacial Lake Agassiz region in south-ern Manitoba, 11,500 to 6,500 B.P. In J.T. Teller, & L. Clayton (Eds.), Glacial Lake Agassiz (pp. 421–451).Geological Association of Canada Special Paper. St. Johns, Newfoundland: Geological Association ofCanada.



47

Phillips, B.A.M. (1993). A time-space model for the distribution of archaeological sites in the Lake SuperiorBasin. Geoarchaeology, 8, 87–107.

Phillips, B.A.M., & Hill, C.L. (1993). Judge C.R. Magney State Park: Geomorphic features and potentialarchaeological sites [Technical Report]. St. Paul, MN: Minnesota Department of Natural Resources,Division of Parks and Recreation.

Phillips, B.A.M., & Hill, C.L. (1994a). The geology, glacial and shoreline history and archaeological poten-tial of the Minnesota north shore of Lake Superior [Technical Report]. St. Paul, MN: MinnesotaDepartment of Natural Resources, Division of Parks and Recreation.

Phillips, B.A.M., & Hill, C.L. (1994b<zaq;4>). Implications of abandoned shoreline features above GlacialLake Duluth levels along the north shore of the Superior Basin in the vicinity of the Brule River. Paperpresented at the 13th Biennial Meeting of the American Quaternary Association, University ofMinnesota, Minneapolis.

Phillips, B.A.M., & Hill, C.L. (1994c). Tettagouche State Park: Geomorphological features and potentialarchaeological sites [Technical Report]. St. Paul, MN: Minnesota Department of Natural Resources,Division of Parks and Recreation.

Phillips, B.A.M., & Hill, C.L. (2001). Deglaciation history and geomorphological character of the regionbetween the Agassiz and Superior basins, associated with the “interlakes composite” of Minnesota andOntario. In B.A.M. Phillips & M. McLeod (Eds.), The deglaciation and geoarchaeology of theMinnesota–Ontario Borderlands: A field guide to the excursions for the Midwest Friends of thePleistocene, 47th Field Conference<zaq;6>, Thunder Bay, ON.<zaq;7>

Phillips, B.A.M., & Hill, C.L. (2004). Deglaciation history and geomorphological character of the regionbetween the Agassiz and Superior basins, associated with the “interlakes composite” of Minnesota andOntario. In <zaq;8>The Late Paleo-Indian Great Lakes, geological and archaeological investigationsof Late Pleistocene and Early Holocene Environments, Mercury Series Archaeology Paper 165 (pp.275–301). Ottawa, ON: Canadian Museum of Civilization.

Phillips, B.A.M., Hill, C.L., Fralick, P.W., & Ross, B. (1994). Post-Glacial Lake shorelines and Paleoindianmigration along the northwestern shore of Lake Superior. In Guidebook for field trip E, prepared forthe 13th Biennial Meeting of the American Quaternary Association, University of Minnesota, Duluth,MN.<zaq;9>

Pregitzer, K.S., Reed, D.D., Bornhorst, T.J., Foster, D.R., Mroz, G.D., McLachlan, J.S., et al. (2000). A buriedspruce forest provides evidence at the stand and landscape scale for the effects of environment on veg-etation at the Pleistocene/Holocene boundary. Journal of Ecology, 88, 45–53.

Rapp, G., Jr., & Hill, C.L. (1998). Geoarchaeology: The earth-science approach to archaeological inter-pretation. New Haven, CT: Yale University Press.<zaq;2>

Rapp, G., Jr., & Hill, C.L. (2006). Geoarchaeology: The earth-science approach to archaeological inter-pretation (2nd ed.). New Haven, CT: Yale University Press.

Rapp, G., Jr., Mulholland, S.C., Mulholland, S.L., Jing, Z., Stoessel, D.E., Hill, C.L., et al. (1995). Final report:Hannaford data recovery project, Koochiching County, Minnesota. Archaeometry Laboratory ReportNo. 95–31. Duluth, MN: University of Minnesota.

Richmond, G.M., & Fullerton, D.S. (1983). Qaternary geologic map of the Minneapolis 4 degree x 6 degreeQuadrangle, Unites States, Quaternary geologic atlas of the United States. U.S.G.S. MiscellaneousInvestigation Series, Map I-1420.<zaq;5>

Romano, A.D., & Johnson, E. (1990). A Northern Minnesota gunflint silica Clovis point. The WisconsinArchaeologist, 71, 204–215.

Ross, W. (1997). The Interlakes composite: A re-definition of the initial settlement of the Agassiz-MinongPeninsula. The Wisconsin Archaeologist, 76, 244–268.

Schneider, F. (1982). The Pelland and Moe blades: Paleo-Indian culture history in the Upper Midwest.Plains Anthropologist, 27, 125–135.

Sellet, F. (2001). A changing perspective on Paleoindian chronology and typology: A view from theNorthwestern Plains. Arctic Anthropology, 38, 48–63.

Severson, L.S., Mooers, H.D., & Malterer, T.J. (1980). Inventory of peat resources Koochiching County,Minnesota. Hibbing, MN: Minnesota Department of Natural Resources, Division of Minerals.

Shane, O.C., III. (1989). Ice Age hunters in Minnesota. Institute for Minnesota Archaeology QuarterlyNewsletter, 4<zaq;3>.

HILL


Shipman, P., Fisher, D.C., & Rose, J.J. (1984). Mastodon butchery: Microscopic evidence of carcass pro-cessing and bone tool use. Paleobiology, 10, 358–365.

Stauffer, C.R. (1945). Some Pleistocene mammalian inhabitants of Minnesota. Proceedings of the MinnesotaAcademy of Science, 13, 21–43.

Steinbring, J. (1974). The preceramic archaeology of northern Minnesota. In E. Johnson (Ed.), Aspectsof Upper Great Lakes anthropology (pp. 64–73). St. Paul, MN: Minnesota Historical Society.

Stoltman, J.B. (1971). Prismatic blades from northern Minnesota. Plains Anthropologist, 16, 105–110.Storck, P.L. (1988). The Late Wisconsinan ice margin and Early Paleo-Indian occupation in the mid-con-

tinent region. Midcontinental Journal of Archaeology, 13, 259–272.Stuvier, M. (1975). Climatic versus changes in 13C content of the organic component of lake sediments dur-

ing the Late Quaternary. Quaternary Research, 5, 251–262.Teller, J.T. (2001). Formation of large beaches in an area of rapid differential isostatic rebound: The three-

outlet control of Lake Agassiz. Quaternary Science Reviews, 20, 1649–1659.Teller, J.T., & Thorleifson, L.H. (1983). The Lake Agassiz–Lake Superior connection. In J.T. Teller & L.

Clayton (Eds.), Glacial Lake Agassiz (pp. 261–290). Geological Association of Canada, Special Paper26. St. Johns, Newfoundland: Geological Association of Canada.

Upham, W. (1899). Report on Aitkin County. Minnesota Geological Survey Final Report, 4, 46–47.<zaq;5>

Winchell, W.H. (1899). The geology of the southern portion of St. Louis County. Minnesota Geologicaland Natural History Survey Final Report, 4, 212–221.<zaq;5>

Winchell, W.H. (1901). Glacial lakes of Minnesota. Geological Society of America Bulletin, 12, 108–128.Winter, T.C. (1971). Sequence of glaciation in the Mesabi-Vermilion Iron Range area, northeastern Minnesota.

United States Geological Survey Professional Paper, 750-C, 82–88.<zaq;5>

Wright, H.E., Jr. (1957). Stone orientation in Wadena drumlin field, Minnesota. Geografiska Annaler, 39,19–31.

Wright, H.E., Jr. (1962). Role of the Wadena lobe in the Wisconsin glaciation of Minnesota. GeologicalSociety of America Bulletin, 73, 73–100.

Wright, H.E., Jr. (1998). A sketch of the history of glacial investigations in Minnesota. In C.J. Patterson &H.E.J. Wright (Eds.), Contributions to Quaternary studies in Minnesota (pp. 1–6). Minnesota GeologicalSurvey Report of Investigations, 49. St. Paul, MN: Minnesota Geological Survey.

Wright, H.E., Jr., Matsch, C.L., & Cushing, E.J. (1973). Superior and Des Moines lobe. Geological Societyof America Memoir, 136, 153–185.

Wright, H.E., Jr., & Watts, W.A. (1969). Glacial and vegetational history of northeastern Minnesota. SpecialPublication 11. St. Paul, MN: Minnesota Geological Survey.

Received May 31, 2001Accepted for publication August 21, 2006



49

QUERIES

<<enote>>AQ1: AU: You meant preglacial not pro-glacial here, correct?<<enote>>AQ2: AU: Please cite in text or omit from the Reference List.<<enote>>AQ3: AU: Please provide inclusive page numbers for the article.<<enote>>AQ4: Au: Please provide the month the meeting was held.<<enote>>AQ5: AU: Please note the Department and the city where this can be accessed.<<enote>>AQ6: AU: Please provide inclusive page numbers for the chapter.<<enote>>AQ7: AU: Please provide the city, state/country: name of publisher.<<enote>>AQ8: AU: Any editor(s) for this?<<enote>>AQ9: AU: Duluth location correct?<<enote>>AQ10: AU: There were a number of places in the reference list where I was unsure if the pro-

ceedings of meetings were formally published or if you were simply citing a paper presented at themeeting. Please review your references carefully and:

1. If a published proceedings, give the title of the paper. In (name of editor(s), title of book (inclu-sive page numbers). city, state: name of publisher.2. If the paper was presented at a meeting give the month the meeting was held after the year.note “Paper presented at (name of meeting), city meeting was held, state/country. Many associa-tions do have their proceedings available. Check their Web site for availability and follow citationas for a published book noting the Society/Association as the Publisher.3. Provide city, state: name of department/division for all state/federal reports, maps, etc.

<<enote>>AQ11: AU: There is no inset in Figure 2 but there is in Figure 1. Are the captions/figures con-fused?

HILL


Documents

Geoarchaeology and Late Glacial Landscapes in the Western Lake … · 2019-04-30 · Geoarchaeology and Late Glacial Landscapes in the Western Lake Superior Region, Central North