Late Pleistocene Palaeoenvironments of the Southern Lake Agassiz ...

JOURNAL OF QUATERNARY SCIENCE (2005) 20(3) 255–267 Copyright ß 2005 John Wiley & Sons, Ltd. Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jqs.905

Late Pleistocene Palaeoenvironments of the Southern Lake Agassiz Basin, USA CATHERINE H. YANSA1* and ALLAN C. ASHWORTH2 1 Department of Geography, Michigan State University, East Lansing, MI 48824, USA 2 Department of Geosciences, North Dakota State University, Fargo, ND 58105, USA Yansa, C. H. and Ashworth A. C. 2005. Late Pleistocene Palaeoenvironments of the Southern Lake Agassiz Basin, USA. J. Quaternary Sci., Vol. 20 pp. 255–267. ISSN 0267-8179. Received 8 September 2004; Revised 9 December 2004; Accepted 13 December 2004

ABSTRACT: Macroscopic plant remains, pollen, insect and mollusc fossils recovered from a cut bank on the Red River in North Dakota, USA, provide evidence that an extensive wetland occupied the southern basin of Lake Agassiz from 10 230 to 9900 14C yr BP. Marsh-dwelling plants and invertebrates had colonised the surface of a prograding delta during the low-water Moorhead Phase of Lake Agassiz. A species of Salix (willow) was abundant along distributary channels, and stands of Populus tremuloides (aspen), Ulmus sp. (elm), Betula sp. (birch), and Picea sp. (spruce) grew on the better-drained sand bars and beach ridges. Most of the species of plants, insects, and molluscs represented as fossils are within their existing geographic ranges. Based on a few species with more northerly distributions, mean summer temperature may have been about 1–2
C lower than the present day. No change in species composition occurred in the transition from the Younger Dryas to Preboreal. At the time that the wetland existed, Lake Agassiz was draining either eastward to the North Atlantic Ocean or northwestward to the Arctic Ocean. The wetland was drowned during the Emerson Phase transgression that resulted in meltwater draining southward to the Gulf of Mexico after 9900 14C yr BP. Copyright ß 2005 John Wiley & Sons, Ltd. KEYWORDS: Lake Agassiz; North Dakota; pollen; plant macrofossils; fossil beetles and molluscs; Younger Dryas–Holocene.

Introduction Lake Agassiz came into existence about 11 700 14C yr BP as meltwater accumulated south of the retreating margin of the Red River Lobe (Upham, 1895; Elson, 1967; Clayton and Moran, 1982; Teller et al., 1983; Teller, 2004). Although the lake was relatively short-lived, its drainage history was complex, controlled by advances and retreats of the southern margin of the Laurentide ice sheet, elevations of spillways, and isostatic rebound. At different times, the lake drained southward to the Gulf of Mexico, eastward to the North Atlantic Ocean, northwestward to the Arctic Ocean, and, in its final phase, northward to Hudson Bay (Teller and Thorleifson, 1983; Fisher and Souch, 1998; Leverington et al., 2000; Teller, 2001, 2004; Fisher, 2003). The eastward drainage of Lake Agassiz between 10 800 and 10 100 14C yr BP, via the Great Lakes and St. Lawrence seaway, has been cited as the trigger that shut down the thermohaline circulation of the North Atlantic Ocean and initiated the Younger Dryas Stade (Broecker et al., 1989; Teller, 1990, 2004; Clark et al., 2001). The plant and invertebrate fossils we report are from the last part of the low-water, Moorhead Phase of Lake Agassiz, depos-

* Correspondence to: Catherine H. Yansa, Department of Geography, Michigan State University, East Lansing, MI 48824, USA. E-mail: [email protected]

ited during a time when Lake Agassiz was draining through its northern outlets. The sedimentary- and bio-facies enable us to reconstruct the palaeoenvironment of the southern Lake Agassiz basin during the Pleistocene to Holocene transition. Additionally, we discuss the vegetation and invertebrate biota in its regional context, and especially in relationship to a cold ‘lake effect’ that reportedly modified the regional vegetation (i.e., Whitlock et al., 1993). We also discuss the deposits in terms of reconstructions of the drainage history of Lake Agassiz, lending a new insight into a controversial topic (e.g., Teller, 2001, 2002, 2004; Karrow, 2002; Fisher, 2003).

Location, stratigraphy and age of deposits The Trollwood Park site (46
510 N, 96
480 W) is located on the west bank of the Red River of the North within the northern city limits of Fargo, Cass County, North Dakota (Fig. 1). Since the site was excavated, the bank has been protected from further erosion by a boulder rip-rap, making the fossiliferous deposits less accessible than they were previously. The section measures 7 m from the top of the cut bank to the river water level (Fig. 2). Three formations are exposed (from bottom to top): the Brenna Formation (Unit A), Poplar River Formation

256

JOURNAL OF QUATERNARY SCIENCE

Figure 1 Location of fossil sites discussed in text in relation to modern vegetation zones and physiogeography (Missouri and Prairie Coteau uplands to the west and south of the Lake Agassiz basin, respectively). The southern portion of this lake basin covers parts of North Dakota (ND), Minnesota (MN) in the United States, and Ontario (ON) and Manitoba (MB) in Canada. Saskatchewan (SK) is also plotted for reference. Inset shows the sites situated in the Fargo–Moorhead area: (1) Trollwood Park (this study); (2) Seminary (McAndrews, 1967a); and (3) Moorhead (Rosendahl, 1948). The other sites shown include: (4) Big Stone Lake (Yansa, 2002); (5) Pickerel L. (Watts and Bright, 1968); (6) Medicine L. (Radle et al., 1989); (7) Wendel (Yansa, 2002); (8) Coldwater L. (Yansa, 2002); (9) Moon L. (Laird et al., 1996, 1998; Grimm, 2001); (10) Spiritwood (McAndrews et al., 1967); (11) Seibold (Cvancara et al., 1971; Ashworth and Brophy, 1972; Newbrey and Ashworth, 2004); (12) Rice L. (Laird et al., 1998; Grimm, 2001); (13) Kettle L. (Grimm, 2001); (14) Andrews (Yansa, 1998; Yansa and Basinger, 1999); (15) Elk L. (Whitlock et al., 1993); (16) Deep L. (Hu et al., 1997); (17) Snake Curve Section (Moran et al., 1971); (18) Mosbeck (Ashworth et al., 1972); and (19) Rainy River area (several localities; Bajc et al., 2000). Also shown is the location of the present-day wetland, (20) Delta Marsh, on the southern shore of Lake Manitoba (lake not shown)

(Unit B), and Sherack Formation (Unit C), as defined by Harris et al. (1974) and Arndt (1977). The sediments of Units A and C were deposited in deep water and those of Unit B in shallow water. Each of the 16 rhythmites of Unit B consists of a finingupwards sequence of fine sand, silty clay, and organic debris. Mollusc shells, including bivalves with both valves intact, occur in both the sand and silty clay beds. The rhythmite unit is capped by a 1.2 m thick, cross-laminated, fine sand containing abundant mollusc shells, but no organics. Three AMS 14C ages were obtained for the bottom, middle and top of the rhythmite unit (Table 1). In this paper, ages are stated in uncalibrated 14 C years.

(1973), Faegri and Iversen (1975), and Bates et al. (1978). Identifications of pre-Quaternary palynomorphs were based on descriptions and illustrations in Kapp (1969). A sum of more than 300 upland pollen types (including Cyperaceae) was counted per sample. These sums were used to calculate pollen percentages in a TiliaTM spreadsheet and plotted using a combination of Tilia.graphTM, TGViewTM, and Adobe IllustratorTM (Fig. 3). Taxonomic nomenclature, distribution, and autecology of plant species are based on Stevens (1963), the Great Plains Flora Association (1986), and Looman and Best (1987).

Plant macrofossils

Sampling, fossil preparation and analysis Sixteen silty-clay organic layers within the rhythmite facies (Unit B) were sampled for pollen, plant macrofossils, insects and molluscs. Sample sizes (wet weight) were 1 g for pollen, 250 g for plant macrofossils, and 5 kg for insects and molluscs.

Pollen Preparation, counting, and identification of the pollen samples followed standard procedures outlined in McAndrews et al. Copyright ß 2005 John Wiley & Sons, Ltd.

Plant macrofossils are the best preserved and most diverse of all the fossil groups at the Trollwood Park site. They occur throughout the rhythmite facies but are especially well preserved and abundant in the beds between 4.8 and 6 m depth. Sediment samples were sieved using nested screens with 425and 212-mm-mesh openings. Plant macrofossils were sorted and identified following standard procedures (Martin and Barkley, 1961; Montgomery, 1977; Birks, 1980, 2001). Total counts of seeds, fruits, buds and leaves are presented as numbers per 250 g of sediment and were plotted using the software listed above. Plant macrofossils were also inspected for wear to provide information about subaerial exposure, transport and redeposition within the local depositional environment (Glaser, 1981). J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

PALAEOENVIRONMENTS SOUTHERN LAKE AGASSIZ BASIN, USA

257

Figure 2 Lithostratigraphy and 14C ages of the Trollwood Park site sediments and their with associated phases of glacial Lake Agassiz and 14C ages

Table 1 Results of radiocarbon age determinations with reference to stratigraphic depth, lab sample number (#, 14C laboratory), AMS 14C age (with one standard deviation and corrected for
13C fractionation), calendar age (using the INTCAL98 calibration curve of Stuiver et al. (1998)), and the materials dated. 1University of Arizona AMS (AA) 14 C Laboratory, NSF Facility. 2Beta Analytic Inc. (Beta) Depth (m) Laboratory number

4.6 5.1 6.48

1

AA34343

AMS age Calibrated age 2 (14C yr BP) range/50% mean prob (cal yr BP) 9920  60

11 194–11 628 11 316 1 AA34344 10 040  120 11 201–12 259 11 575 2 Beta-121851 10 230  80 11 364–12 591 11 959

Materials dated

Populus wood Populus leaves and wood Populus wood

Identifications were made by comparison of the fossils with voucher specimens in reference collections. Detailed morphological descriptions for many of the fossil taxa are provided in Yansa and Basinger (1999). Identification of Potamogeton species was based on their distinctive seed morphologies exposed Copyright ß 2005 John Wiley & Sons, Ltd.

by longitudinal sectioning of the drupes. The fossils were reposited in the Pollen and Soils Laboratory, Michigan State University.

Invertebrates Each sample was wet sieved using a 300-mm mesh and the mollusc shells and larger fragments of chitin were collected from the plant matrix. The residue was then subjected to a standard kerosene flotation technique to concentrate insect remains (Ashworth, 1979). The fossils, after ultrasonic cleaning, were mounted on micropalaeontogical slides and reposited in the collections of the Quaternary Entomology Laboratory, North Dakota State University. Neither molluscs nor insects were as abundant as plant macrofossils. There did not appear to be any particular significance to the stratigraphic distribution of the taxa and in this report they are considered to belong to a single assemblage zone. This includes molluscs occurring in the sand immediately above the rhythmites. The number of individuals was too small for quantitative analyses. J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

Figure 3 Pollen percentage diagram for the fossiliferous rhythmite facies of the Trollwood Park site (most common taxa shown). See Fig. 2 for the legend of the lithology

258

Copyright ß 2005 John Wiley & Sons, Ltd.

JOURNAL OF QUATERNARY SCIENCE

J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

PALAEOENVIRONMENTS SOUTHERN LAKE AGASSIZ BASIN, USA

Interpretations Pollen The number of indeterminable palynomorphs exceeds that of identified pollen and spores (Fig. 3), and in this respect is similar to other deposits associated with the Moorhead Phase (McAndrews, 1967a; Bajc et al., 2000). Also, Cretaceousage palynomorphs, including Appendicisporites tricornatus Weyland & Greifeld, Reticuloidosporites dentatus Pfug. Coll., Hamulatisporis hamulatus Krutzsch., Aquillapollenites pulvinus Stanley, Hystrichosphaeridium tubiferum (Ehrbg.) Coll, and Hystrichosphaeridium recurvatum (White) Coll., are relatively abundant. Some of them have previously been reported from Lake Agassiz basin sediments in Manitoba (Nambudiri et al., 1980). The large number of degraded palynomorphs is consistent with the interpretation of a sedimentary facies in which deposits were continually being eroded and reworked by streams. Of the identifiable pollen, Salix sp. (willow; 8% to 59%), dominates the AP with the greatest values in the middle of the zone. High pollen values for Salix sp., species of Cyperaceae (sedges; 7–24%); Pediastrum sp. Meyen (green alga; 0–8%) and small amounts Potamogeton sp. (pondweed; not shown in Fig. 3) and Typha latifolia L. (broad-leaved cattail) support the interpretation of a wetland. The pollen values for coniferous and deciduous trees suggests that the uplands surrounding the Lake Agassiz basin were forested and that trees were able to disperse into the wetland and colonise river banks. Pollen of grassland vegetation, namely species of Poaceae, Artemisia, Ambrosia, Chenopodiaceae, and Asteraceae, is less well represented, suggesting that extensive grasslands were not present in the area. Picea sp. (spruce) pollen varies from 13% to 45% (Fig. 3). Some of the spruce pollen grains are degraded, suggesting reworking, but others are in better condition and probably are of local origin. Pinus sp. (pine; 0–10%) pollen, on the other hand, is probably all from long-distance transport. Even though represented by small percentages, the pollen of Quercus sp. (oak; 0–7%), Ulmus sp. (elm; 0–5%), Betula sp. (birch; 0–3%), Acer sp. (maple; 0–6%; some tentatively identified as A. negundo L. (box elder)), and Fraxinus sp. (ash; 0–10%, some identified as F. nigra Marsh. (black ash)), suggests that deciduous trees were growing on better-drained soils within the wetland. Populus sp. (poplar; 2–16%) pollen, a type that does not preserve well (MacDonald, 1990), suggests that trees of this taxon were present. Two taxa are tentatively identified, Populus tremuloides Michx.-type (aspen poplar) and Populus balsamifera L.-type (balsam poplar). Small amounts of pollen (  2%) of Shepherdia canadensis (L.) Nutt. (rabbitberry), and species of Alnus (alder), Carya (hickory), Ostrya/Carpinus (hop hornbeam/hornbeam), Corylus (hazel), Juglans (walnut), Ephedra (Mormon tea) and Rosaceae (rose family) are included together in the total arboreal pollen sum (AP). Some of these were probably of local origin, but others, such as Ephedra sp., were undoubtedly from long-distance transport (Maher, 1964). The site is located within what was once an extensive marsh. In addition to the aquatic and semiquatic taxa, the sand bars and channel levees may have been colonised by a vegetation that included spruce, aspen, balsam poplar, oak, elm, black ash, box elder, and some prairie herbs. Except for spruce, species of all of these taxa occur today in riparian habitats along the Red River. At present, the southernmost occurrence of native spruce is in the boreal forest, 400 km to the north. Copyright ß 2005 John Wiley & Sons, Ltd.

259

Plant macrofossils Well-preserved fossils considered to be contemporaneous with the deposition of the sediments are those of woodland, mudflat, emergent and aquatic vegetation, the most common of which are shown in Fig. 4. Illustrations of some of these taxa are shown in Fig. 5. Low numbers of Picea sp. (spruce) needles were recovered, and many were broken and worn, appearing to have been reworked. Some of these needles were also charred. Woodland or parkland is represented by macrofossils of deciduous trees and shade-intolerant shrubs. Populus tremuloides (aspen poplar) and Populus balsamifera (balsam poplar) were identified from terminal bud scales (a scale of the former is shown in Fig. 5G). They are plotted together as Populus sp. in Fig. 4. Catkins of Betula sp. (birch; Fig. 5D) and fruits of Rubus idaeus L. (red raspberry), Prunus cf. P. pensylvanica L. (pin cherry), and Viburnum edule (Mich.) Raf. (squashberry; Fig. 5H) were also identified. One seed of Salix sp. (willow) was also recovered, which is significant, since seeds of this taxon are rarely preserved (Birks, 1980). Fossils of plants of slightly alkaline mudflats and wet meadows are represented by several species. They include Lycopus americanus Muhl. ex Bart (water hoarhound), three species of Ranunculus (R. cf. R. gmelini DC. (small yellow buttercup), R. cf. R. pennsylvanicus L. (bristly buttercup) and R. cf. R. sceleratus L. (ditch buttercup)), Potentilla norvegica L. (rough cinquefoil), Caltha palustris L. (marsh marigold), Polygonum lapathifolium L. (pale smartweed; Fig. 5A), Cirsium cf. C. flodmani (Rydb.) Arthur (Flodman’s thistle), Grindelia squarrosa (Pursh) Dun. (curly-top gumweed), Rumex maritimus L. (golden dock; Fig. 5F), Mentha arvensis L. (field mint), Stachys sp. (hedge-nettle), Aster sp. (aster), a species of Apiaceae (parsley family), and Chenopodium spp. (goosefoot), some of which were identified as C. rubrum L. (red goosefoot). The low number of Chenopodium rubrum seeds and the absence of other brackish-water indicators, such as Zannichellia palustris L. (Yansa, 1998), suggest a relatively fresh-water wetland. In addition, fossils of emergent plants which grow in shallow water of less than 1 m depth, are also very abundant. They include Scirpus validus Vahl (common bulrush), S. microcarpus Presl (small-fruited bulrush; Fig. 5B), Eleocharis sp. (spike-rush), Typha latifolia L. (broad-leaved cattail), Sagittaria latifolia Willd. (broad-leaved arrowhead; Fig. 5C), Cicuta cf. C. maculata L. (water-hemlock), Sparganium sp. (bur-reed), and three species of Carex (sedge; Fig. 5E). The aquatic flora are also very diverse, and include Wolffia (watermeal), Myriophyllum verticillatum L. (whorled watermilfoil), Hippuris vulgaris L. (mare’s tail), and Lemna minor L. (ivy-leaved duckweed). Also eight species of Potamogeton (pondweed) were identified: P. filiformis Pers. (inland pondweed); P. gramineus L. (variable pondweed); P. natans L. (floating-leaved pondweed, Fig. 5I); P. obtusifolius Mert. & Koch (bluntleaf pondweed); P. pectinatus L. (sago pondweed); P. pusillus L. (small pondweed); P. richardsonii (Benn.) Rydb. (Richardson’s pondweed); and P. vaginatus Turcz. (sheathed pondweed). Oogonia of Chara sp. (green alga) and leaves and stems of Drepanocladus sp. (sickle-branch moss) were also recovered. These aquatic plants indicate eutrophic, fresh to slightly alkaline water, of depths of up to 1–2 m. All species represented by macrofossils at present occur within the region, except for Picea sp. and Potamogeton obtusifolius (blunt-leaved pondweed), which today occur in the boreal forest to the north.

Invertebrates The fossils are mostly those of aquatic species of molluscs, insects, crustaceans and a bryozoan. Shells of lymnaeid gastroJ. Quaternary Sci., Vol. 20(3) 255–267 (2005)

Figure 4 Plant macrofossil abundance diagram for the fossiliferous rhythmite facies of the Trollwood Park site (most common taxa shown). See Fig. 2 for the legend of the lithology

260

Copyright ß 2005 John Wiley & Sons, Ltd.

JOURNAL OF QUATERNARY SCIENCE

J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

PALAEOENVIRONMENTS SOUTHERN LAKE AGASSIZ BASIN, USA

261

Figure 5 SEM micrographs of nine plant macrofossil taxa identified at the Trollwood Park site. The scale bars for all micrographs are 1 mm. A: Polygonum lapathifolium (pale smartweed) achene with hooked perianth veins. B: Scirpus microcarpus (small-fruited bulrush) achene. C: Sagittaria latifolia (broad-leaved arrowhead) achene. D: Betula sp. (birch) catkin (the catkin is broken at the top right corner so the original specimen was longer). E: Carex cf. C. rostrata (beaked sedge) achenes (the specimen on the right is encased by a perigynium; this sheath is not preserved on the achene on the left). F: Rumex maritimus (golden dock) achene (barely visible) with attached perianth segments bearing marginal bristles. G: Populus tremuloides (aspen poplar) terminal bud scale. H: Viburnum edule (squashberry) drupe. I: Potamogeton natans (floating-leaved pondweed) drupe with an intact membranous sheath

pods, represented by Amnicola limosa (Say), Fossaria sp., Gyraulus deflectus (Say), Gyraulus parvus (Say), Planorbula armigera (Say) and Valvata tricarinata (Say), and the sphaeriid bivalve, Pisidium compressum Prime, occur throughout the rhythmite facies (Fig. 2). The habitat of P. armigera, which lives among vegetation in most kinds of perennial-water habitats, especially stagnant eutrophic habitats in which the substrate is mud, is typical for the assemblage (Clarke, 1981). The bivalve P. compressum is found in vegetation in shallow perennial water bodies with a wide range of substrates (Clarke, 1981). Insects are represented by larval head capsules of Chironomidae (midges), puparia of Diptera, mandibles of Odonata (dragonflies and damselflies), mandibles of Hymenoptera (ants), cases and larval schlerites of Trichoptera (caddisflies), and heads, pronota, elytra and sclerites of Coleoptera (beetles) (Table 2). The Coleoptera include several families of predaceous and herbivorous water beetles (Dytiscidae, Gyrinidae, Hydraenidae, Hydrophilidae, Haliplidae, Elmidae), families associated with aquatic and emergent vegetation (Helodidae, Chrysomelidae, Curculionidae), families associated with the mesic banks of riparian habitats (Carabidae, Staphylinidae, Heteroceridae), and Orthotomicus caelatus Eichoff, a bark beetle associated with conifers (Scolytidae). Lindroth (1963) described the habitat of the ground beetle Bembidion sordidum Kirby as being almost confined to shaded river banks; we have collected this species under alder Copyright ß 2005 John Wiley & Sons, Ltd.

and willow shrubs at the water’s margin. The assemblage is very similar in the ecology it represents to that of the Seminary site (Kompelian and Schwert, 1986), located 1.9 km east of the Trollwood Park site (Fig. 1, inset). Other fossils of invertebrates include hysterosomata of oribatid mites, ephippia of the cladoceran Daphnia sp. and statoblasts of the freshwater bryozoan, Cristatella mucedo Say, whose colonies attach themselves to the submerged stems of emergent vegetation. Collectively, the invertebrate fauna is one which would be associated with a nutrient-rich, eutrophic wetland. No fossils of vertebrates were found.

Discussion Comparison to other Moorhead Phase sites in the Fargo–Moorhead area Two other investigations of plant fossils have been reported from deposits of the Poplar River Formation in Fargo–Moorhead (Fig. 1, inset): the Moorhead site (Station 2) (Rosendahl, 1948); and the Seminary site (McAndrews, 1967a). Wood samples dated from these sites are comparable to the youngest J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

262

JOURNAL OF QUATERNARY SCIENCE

Table 2 List of fossil Coleoptera (beetles) identified at the Trollwood Park site, compared to those reported for the nearby Seminary site (Kompelian and Schwert, 1986). The habitat and author describing the association of beetles with their habitats is also listed. Skeletal parts include: H ¼ head(s); P ¼ pronotum(a); LE ¼ left elytron(a); RE ¼ right elytron(a); F ¼ fragment(s) Carabidae Carabidae Carabidae Carabidae Carabidae Gyrinidae Haliplidae Dytiscidae Dytiscidae Dytiscidae Hydrophilidae Hydrophilidae Hydrophilidae Hydrophilidae Hydrophilidae Hydraenidae Scydmaenidae Staphylinidae Staphylinidae Staphylinidae Staphylinidae Staphylinidae Staphylinidae Scarabaeidae Elmidae Heteroceridae Helodidae Coccinellidae Chrysomelidae Curculionidae Curculionidae Curculionidae Curculionidae

Agonum sp. Bembidion nigripes Kby. Bembidion pseudocautum Lth. Bembidion sordidum Kby. Dyschirius sp. Gyrinus sp. Haliplus sp. Agabus sp. Colymbetes sp. Hydroporus sp. Helophorus sp.(2) Paracymus sp. Enochrus sp. Cercyon sp. Hydrobius sp. Octhebebius sp. Sp. indet. Olophrum rotundicolle C.R. Sahlb. Olophrum consimile Aleocharini spp. indet. Stenus sp. Bledius sp. Pselaphini sp. indet. Aphodius cf. omissus LeC. Stenelmis sp. Heterocerus sp. Cyphon sp. Anisosticta bitriangularis Say Donaciini sp. indet. Bagous sp. Orthotomicus caelatus Eichoff Hylastes sp. Sp. indet.(2)

E P E P E H,E E P fE P,E fE E E E fE P,E E L P,L H,P,L P,L P E E E P,E H,E E E H,P E E H,P

14

C age for the Trollwood Park site; the Moorhead site has an age of 9930  280 14C yr BP (W-388) (Ruben and Alexander, 1958), and the Seminary site an age of 9900  400 14C yr BP (W-993) (McAndrews, 1967a). The fossiliferous deposits at the Seminary and Moorhead sites are only 0.15 m and 0.6 m thick, respectively, and thus provide less complete stratigraphic and fossil records than the 2 m thick deposit we studied. The Trollwood Park, Seminary and Moorhead sites share an almost identical assemblage of mudflat/wet prairie, emergent and aquatic herbaceous taxa, as identified from macrofossils. The relative abundance of Picea sp. (spruce) varies between sites, however. Spruce needles recovered from the Trollwood Park site are fewer than reported for the Moorhead site (Rosendahl, 1948), and many are also broken or charred, indiTable 3 Mean temperatures and precipitation values from meteorological stations at Fargo, North Dakota, and the University of Manitoba’s Delta Marsh Field Station on the southern shore of Lake Manitoba. Data based on the climate normal years 1961–1990 for Fargo (Owenby and Ezell, 1992) and 1967–1990 for Delta Marsh (McGinn, 1992) Meteorological station

Temperature (
C) Mean Mean Mean Jan. July annual

Fargo, North Dakota 46
550 N, 96
480 W (Trollwood Park site) Delta Marsh Field Station, Manitoba 50
110 N, 98
230 W

Precipitation (mm) Mean annual

14.5

21.7

5.0

494

19.8

19.1

1.5

499

Copyright ß 2005 John Wiley & Sons, Ltd.

Trollwood Trollwood Seminary Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Seminary Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood Trollwood

River banks Lake margins River banks River banks

Lindroth, 1963 Lindroth, 1963 Lindroth, 1963 Lindroth, 1963

Aquatic Aquatic Aquatic Aquatic Aquatic Aquatic Aquatic Aquatic Aquatic Aquatic Emergent veg. Emergent veg.

Campbell, 1983 Campbell, 1983

Aquatic margin Lake margin Running water Aquatic margin

Seminary

Emergent veg.

Trollwood Trollwood Seminary Trollwood

Aquatic On conifers On conifers

Helgesen and Post, 1967

Wood, 1982 Wood, 1982

cating that at least some were redeposited. Some of the spruce pollen grains at the Trollwood Park site are probably allochthonous, but values are high enough (13–45%) to indicate that pockets of spruce probably occupied the Agassiz basin. The presence of the bark beetle Orthotomicus caelatus Eichoff, which is associated with coniferous trees, supports this interpretation for the Trollwood Park site. Moreover, the pollen spectra from the Seminary site are dominated by Picea sp. (>60%) and the macrofossils include a single Larix laricina (Du Roi) K. Koch (tamarack) needle and wood of either Picea sp. or Larix sp. (McAndrews, 1967a). Rosendahl (1948) also identified macrofossils of Larix laricina, in smaller quantities than that of Picea glauca (Moench) Voss (white spruce), at the Moorhead site. McAndrews (1967a) interpreted the water-worn wood at the Seminary site to have been transported from the forested margins to the drained lake basin by rivers, but he considered the macrofossils of marsh vegetation to have been deposited in situ. Rosendahl (1948) interpreted the entire plant macrofossil assemblage of the Moorhead site to have been transported to the lake by the Buffalo River from forested moraines about 30 km to the east. However, his line drawings show several macrofossils with easily degraded appendages, including an achene of Polygonum lapathifolium (pale smartweed) with the hooked perianth veins still attached, similar to the Trollwood Park site specimens (one of which is shown in Fig. 5A). The preservation of these and other delicate tissues, such as membranous sheaths which encase some of the Potamogeton (pondweed) drupes (Fig. 5I) and Carex (sedge) achenes (i.e. perigynia; Fig. 5E), indicate that the plants were growing within the depositional environment. J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

PALAEOENVIRONMENTS SOUTHERN LAKE AGASSIZ BASIN, USA

We agree with both Rosendahl (1948) and McAndrews (1967a) that some of the fossils were transported fluvially from the morainic uplands surrounding the basin. Others, however, especially those of aquatic and emergent plants, and the majority of the invertebrates, inhabited the depositional environment. The Seminary site pollen spectrum is dominated by Picea sp. and in this respect is different from that of the Trollwood Park site. The differences can be explained in terms of the pollen assemblages being reworked to different degrees by fluvial processes active on a delta top.

Palaeoecology and palaeoclimate of the Moorhead Delta The productive wetland that we interpret was situated within the Moorhead Delta, which formed along the southern shore

263

of Lake Agassiz in the vicinity of Fargo–Moorhead during the late Moorhead Phase (Figs 1 and 6). This delta has been previously identified in Fargo–Moorhead from borehole data and is believed to have been formed by deposition from the ancestral Sheyenne, Red, Buffalo, and Maple Rivers when the lake was at its Ojata shoreline (Fig. 6B; Harris et al., 1974; Arndt, 1977; Brophy and Bluemle, 1983; Fenton et al., 1983). The Trollwood Park site record indicates that by 10 230 14 C yr BP, Lake Agassiz had transgressed southward to Fargo– Moorhead (Fig. 6A). This transgression followed a regression during the early Moorhead Phase when subaerial weathering and pedogenesis occurred in the southern basin (Moran et al., 1971; Arndt, 1977; Teller and Last, 1981; Fenton et al., 1983; Reid and Olson, 1996; Bajc et al., 2000; Teller et al., 2000). We propose that the Delta Marsh, a protected wetland and wildlife refuge in south-central Manitoba, Canada, located

Figure 6 A: schematic reconstruction of the Moorhead Delta, the palaeoenvironmental setting of the Trollwood Park site, during the late Moorhead Phase. The cities of Fargo and Grand Forks are plotted for reference. B: close-up of the ancestral streams that formed the Moorhead Delta along the southern shore of glacial Lake Agassiz. The physiography is adapted from Arndt (1977) and Fenton et al. (1983). C: diagrammatic cross-section of the Trollwood Park site area; transect from the shore of Lake Agassiz south to a lagoonal marsh. This reconstruction is based on Delta Marsh, which exists today along the southern shore of Lake Manitoba in southern Manitoba (adapted from Shay, 1999) Copyright ß 2005 John Wiley & Sons, Ltd.

J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

264

JOURNAL OF QUATERNARY SCIENCE

about 390 km north of the Trollwood Park site, is the modern analogue for the Moorhead Delta. Delta Marsh is situated within the middle of a 110-km wide belt of aspen parkland, which forms a transitional vegetation zone between grassland to the south and boreal forest to the north. The Delta Marsh complex of channels and lagoons forms where the Assiniboine River and its tributaries enter Lake Manitoba (Teller and Last, 1981). Over 75% of the plant species identified at the Trollwood Park site exist today at Delta Marsh (flora described by Shay, 1999), and the botanical associations were probably similar (Fig. 6C). The sand bars and beach ridges of the Moorhead Delta were probably occupied by Acer negundo (box elder), Populus tremuloides (aspen), Ulmus sp. (probably U. americanus L., American elm), Betula sp. (birch), Quercus sp. (probably Q. macrocarpa Michx., bur oak), Fraxinus nigra (black ash), and in some places Picea sp. (spruce). A zone of Salix sp. (willow) most likely existed along the shores of the lagoonal marshes within the delta. The shallow waters of these marshes were inhabited by abundant aquatic and emergent vegetation, molluscs and insects. Periodic flooding of the marshes in the Moorhead Delta would have killed the vegetation, resulting in a cyclic pattern of vegetation dieback and renewal, similar to that observed today at Delta Marsh (Shay, 1999). Periodic disturbance by flooding may account for the high species diversity (Collins et al., 1995). The majority of the plant, mollusc and insect species at the Trollwood Park site occur within their existing distributional limits. The exceptions are the aquatic plant Potamogeton obtusifolius (blunt-leaved pondweed) and Picea sp. (spruce). Investigations at the nearby Seminary site (Fig. 1, inset) provide more conclusive evidence for spruce forest with the occurrence of Picea sp. and the omaliine staphylinid beetle, Olophrum rotundicolle C. R. Sahlb (McAndrews, 1967a; Kompelian and Schwert, 1986), a species which we have collected from saturated mosses beneath willows on the margins of a spruce-lined lake in northern Minnesota. The occurrence of spruce, either on the margins of the basin, or on islands within the wetland, also suggests a climate that was slightly cooler and wetter than today. Based on the occurrence of a few taxa with more northerly distributions than they have today, we estimate that the average summer temperature was probably 1–2
C cooler at the Trollwood Park site during the late Moorhead Phase. Summer insolation was about 7% higher during the PleistoceneHolocene transition than today, but any increase in temperature that might have been expected was offset by regional cooling resulting from the Laurentide Ice Sheet to the north (Kutzbach et al., 1993). Ice still filled the Lake Superior basin to the east, and formed the northern and northeastern boundaries of Lake Agassiz in central Manitoba and western Ontario, respectively (Dyke and Prest, 1986).

Comparison to contemporaneous regional vegetation The pollen records of several lakes (Fig. 1) on adjacent morainic uplands provide a regional palaeoenvironmental context in which we can also interpret the palaeovegetation and palaeoclimate of the Trollwood Park site. At the time of deposition of the Moorhead Delta, the pollen spectra from several lakes (Moon, Rice, Kettle, Coldwater, and Spiritwood) and ponds (Wendel and Seibold) (Fig. 1), indicate that an extensive parkland of deciduous hardwoods and prairie had replaced the previous open white spruce forest (or parkland) (McAndrews et al., 1967; Barnosky et al., 1987; Laird et al., 1996, 1998; Grimm, Copyright ß 2005 John Wiley & Sons, Ltd.

2001; Yansa, 2002). Similar changes in the pollen profile during this time are also reported from Pickerel and Medicine Lakes (Fig. 1) on the Prairie Coteau of South Dakota (Watts and Bright, 1968; Radle et al., 1989). Grimm (2001) interpreted the regional vegetation during this time to be a mosaic in which the better-drained morainic hilltops supported prairie and the poorly drained low areas supported groves of trees, first Picea glauca (white spruce) and later species of Betula (birch), Populus (poplar), Ulmus (elm) and Quercus (oak). In contrast, the succession of postglacial plant communities on the upland east of the southern Agassiz basin was different. The pollen spectra of several lakes on the Itasca Moraine in northwestern Minnesota (Fig. 1) indicate that the spruce forest was replaced by a pine woodland (e.g. McAndrews, 1967b; Shay, 1967, 1971; Whitlock et al., 1993; Hu et al., 1997). This floristic transition is dated at about 9,100 14C yr BP at Elk Lake (Whitlock et al., 1993), and at 10 230 14C yr BP at nearby Deep Lake (Hu et al., 1997). Hu et al. (1997) considered the younger age for the transition at Elk Lake to be erroneous, because of a dating problem. The timing of the extirpation of spruce, however, may vary between lowland and upland sites. At the Mosbeck site, a lagoon on the eastern margin of Lake Agassiz situated at a similar latitude to Elk Lake, Picea mariana (Mill.) B.S.P. (black spruce) wood and cones, and a fossil beetle assemblage indicate the existence of a spruce forest at 9,940 14C yr BP before the site was drowned by the final high-water Emerson Phase of Lake Agassiz (Ashworth et al., 1972). This age is similar to that of a buried spruce log at the nearby Snake Curve Section site, which was deposited at the transition from the Moorhead to Emerson Phases (Moran et al., 1971). By 9600 14C yr BP, prairie herbs and deciduous trees and shrubs had colonised the area of Big Stone Lake, situated within what was earlier the southern outlet channel (River Warren) of Lake Agassiz, 160 km south of Fargo (Yansa, 2002). Farther north, in the Rainy River area of northwestern Ontario (Fig. 1), Bajc et al. (2000) described fossil plants, insects and molluscs from deltaic and fluvial deposits ranging in age from 10 800 to 9500 14C yr BP. Their data indicate that the vegetation during the Moorhead Phase was similar to existing boreal forest. The differences in floristic composition between the Trollwood Park site and other Moorhead Phase sites to the northeast imply a latitudinal vegetational zonation, similar to that of today. Considering that the Trollwood Park pollen assemblage has been modified by reworking, it shows remarkable similarities with the Moon Lake (Laird et al., 1996) and Pickerel Lake (Watts and Bright, 1968) pollen spectra, the closest records to the west and south, respectively (Fig. 1). In those records, a pioneering open spruce forest was being replaced by deciduous parkland at the time the Moorhead delta was being deposited.

Younger Dryas and lake effect Bartlein and Whitlock (1993) used various statistical techniques to convert pollen data from Elk Lake, Minnesota (Fig. 1), to climatic parameters, namely mean January and July temperatures and annual precipitation. Their results showed a marked climatic reversal towards cooler and wetter conditions dated from 10 500 to 9200 14C yr BP. While it is tempting to correlate the climate change with the Younger Dryas Stade, as Bartlein and Whitlock (1993) did, it begins and ends too late by several hundred years. Our data suggest that there were no J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

PALAEOENVIRONMENTS SOUTHERN LAKE AGASSIZ BASIN, USA

changes in biota during the transition from the Younger Dryas to the Holocene. Bartlein and Whitlock (1993) also discussed the possibility that Lake Agassiz modified the regional climate by exerting a lake effect, much as the Great Lakes do today; that is warmer winters, cooler summers, and greater precipitation on leeward shores. They proposed that when the lake returned to fill the basin during the Emerson Phase at 9900 14C yr BP, the winds coming across the cold lake waters produced downwind cooler and wetter conditions that effected vegetation development in the Itasca Moraine area of western Minnesota. As appealing as this idea may be, the climatic effect is not correlated especially well with major changes in lake levels. In the Elk Lake record (Bartlein and Whitlock, 1993), the colder and wetter conditions of the climatic reversal began during the low-water Moorhead Phase, several hundred years earlier than the Emerson Phase during which the lake reoccupied the southern basin. Hu et al. (1997) compared the
18O and pollen records of nearby Deep Lake, Minnesota (Fig. 1), to evaluate lake effect. They inferred a climatic reversal to cooler and wetter conditions, based on the isotopic data, which coincides with the refilling of the southern Lake Agassiz basin during the Emerson Phase (Preboreal not Younger Dryas). The climatic change associated with the lake effect, however, was too small to have produced any response in the vegetation at Deep Lake. The replacement of spruce forest by pine woodland, which had begun at about 10 500 14C yr BP at this lake, appeared to have continued unabated through the inferred climatic change. Our data from the late Moorhead Phase suggest that at least in the southern part of the basin the surface lake waters were warm. The plants and animals that inhabited the Moorhead Delta could not have existed if the winds coming from the lake were frigid. While it is improbable that there was a lake effect in the southern basin during the Moorhead Phase, we cannot rule out a lake effect later during the Emerson Phase. The absence of fossils, except for rare valves of the coldwater-adapted ostracode Candona subtriangulata Benson & Macdonald in Emerson Phase sediments, was interpreted by Curry (1997) to indicate cold unproductive water. Boulder-size dropstones suggest that ice bergs drifted south of Fargo during this time (Clayton et al., 1965; Bluemle, 1972; Thorleifson, 1996). Melting ice bergs could have chilled the surface waters and produced cooler summers on the eastern shores after 9900 14C yr BP. This cold surface water, according to Hostetler et al.’s (2000) regional climate model simulations, probably created anticyclonic air flow over the surface of the lake. This type of circulation would have blocked the penetration of moisture from the south and west, causing reduced precipitation eastwards. This climatic pattern may explain why on the Itasca Moraine, spruce forest was replaced by a drought-tolerant pine woodland rather than by deciduous forest.

265

According to Teller (2001, 2004), differential isostatic uplift raised the eastern outlet causing the lake to reoccupy the southern basin and once again overflow through its southern outlet between ca. 10 200 and 10 100 14C yr BP. Ice retreat then uncovered the northwestern outlet and the lake drained to the Arctic Ocean. Subsequent isostatic uplift and ice advance caused the northwestern outlet to become blocked and the lake once again reoccupied the southern basin, overflowing the southern outlet between ca. 9920 and 9830 yr BP. An alternative scenario is provided by Fisher (2003). Borehole data from the southern outlet indicate that this outlet was operating sometime between 9900 and 9400 yr BP, but for an unknown duration. The northwestern outlet opened at 9900 yr BP and was operational for a few hundred years (Fisher and Smith, 1994). Fisher (2003), consequently, suggested that drainage through the southern outlet occurred for a few hundred years before the final abandonment of this outlet at 9400 yr BP. The Trollwood Park site, in the axis of the southern Lake Agassiz basin, is in a critical position to examine the complex drainage history outlined by Teller (2001) and Fisher (2003). The Sherack Formation, which overlies the rhythmite facies of the Poplar River Formation (Fig. 2), was deposited at a time when the lake overflowed through the southern outlet. At this time, the water level was 30 m above the fossiliferous horizon (262 m a.s.l.). The flooding of the southern Agassiz basin at the onset of the Emerson Phase is dated at about 9900 yr BP based on 14C ages at the Trollwood Park and other sites (Moran et al., 1973). Our data are in agreement with Teller’s (2001, 2004) interpretation of drainage through the southern outlet after 9920 yr BP, but they do not support his interpretation for the southern outlet being active between 10 200 and 10 100 yr BP. The organisms in the Moorhead Delta wetland would be very sensitive to changes in water depth, and while they might have been able to survive a seasonal flooding of 1–2 m, it would be impossible for them to have survived flooding to a depth of 30 m associated with the refilling of the basin to the extent that it overflowed the southern outlet. The drowning of the Rainy River landscape by the rise in lake levels associated with the Emerson Phase occurred after 9920 yr BP in northwestern Ontario (Bajc et al., 2000), at about 10 000 yr BP in southern Manitoba (Teller et al., 2000), and at about 9920 yr BP at the Trollwood site, about 300 km to the south. The transgression appears to have been more or less synchronous over a very large area. This suggests to us that an ice advance blocking an outlet is a more probable cause for the Emerson phase transgression than isostatic uplift. Great advances have been made with respect to the sequence of events in the drainage history of Lake Agassiz, but the complex interaction between the southern, eastern and northwestern outlets is still not satisfactorily explained. At this time, our data fit the ‘conventional chronology’ for the outlets most recently revised by Fisher (2003) rather than the ‘new chronology’ proposed by Teller (2001, 2004).

Relationship of the site to the drainage history of Lake Agassiz The drainage of Lake Agassiz during most of the Moorhead low-water phase is considered to have been through the eastern outlets to the North Atlantic Ocean (Teller, 2001, 2004; Fisher, 2003). Karrow (2002) argued that evidence for an eastern outlet is weak and the lake was probably draining through the northwestern outlets to the Arctic Ocean. Teller (2002) refuted Karrow’s (2002) interpretation on the basis that the outlets in northern Saskatchewan were still covered by glacial ice during this time. Copyright ß 2005 John Wiley & Sons, Ltd.

Acknowledgements We thank Dr Barry Miller, Kent State University, for confirming the identifications of the molluscs. Mr Alex Wiedenhoeft of the U.S. Forest Products Laboratory, University of Wisconsin at Madison, confirmed identifications of the wood. Mr Michael Burton and Mrs Karie Trupka, science teachers, participated in the project through the NDSTEP programme. Two of the 14C ages were provided free of charge by the University of Arizona AMS (AA) 14C Laboratory, which are much appreciated. The palaeobotanical component of this paper is from the PhD dissertation of the primary author, who acknowledges the insightful comments of her committee: Vance T. J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

266

JOURNAL OF QUATERNARY SCIENCE

Holliday (advisor), Jim C. Knox, Lou Maher, Tom Vale, Lee Clayton and Marjorie Winkler.

References Arndt BM. 1977. Stratigraphy of offshore sediment, Lake Agassiz, North Dakota. North Dakota Geological Survey Report of Investigation no. 60; 1–58. Ashworth AC. 1979. Quaternary Coleoptera studies in North America: past and present. In Carabid Beetles: Their Evolution, Natural History and Classification, Erwin TL, Ball GE, Whitehead DR, Halpern A (eds). W. Junk: The Hague; 395–405. Ashworth AC, Brophy JA. 1972. Late Quaternary fossil beetle assemblage from the Missouri Coteau, North Dakota. Bulletin of the Geological Society of America 83: 2981–2988. Ashworth AC, Clayton L, Bickley WB. 1972. The Mosbeck site: a paleoenvironmental interpretation of the late Quaternary history of Lake Agassiz based on fossil insect and mollusk remains. Quaternary Research 2: 176–188. Bajc AF, Schwert DP, Warner BG, Williams NE. 2000. A reconstruction of Moorhead and Emerson Phase environments along the eastern margin of Glacial Lake Agassiz, Rainy River Basin, northwestern Ontario. Canadian Journal of Earth Science 37: 1335–1353. Barnosky CW, Grimm EC, Wright HE. 1987. Towards a postglacial history of the northern Great Plains: a review of the paleoecologic problems. Annals of Carnegie Museum 56: 259–273. Bartlein PJ, Whitlock CW. 1993. Paleoclimatic interpretation of the Elk Lake pollen record. In Elk Lake, Minnesota: Evidence for Rapid Climate Change in the North-Central United States, Bradbury JP, Dean WE (eds). Geological Society of America Special Paper no. 276; 275–294. Bates CD, Coxon P, Gibbard PL. 1978. A new method for the preparation of clay-rich sediment samples for palynological investigation. New Phytologist 81: 459–463. Birks HH. 1980. Plant macrofossils in Quaternary lake sediments. Archiv fu¨r Hydrobiologie—Advances in Limnology 15: 1–60. Birks HH. 2001. Plant macrofossils. In Tracking Environmental Change Using Lake Sediments: Volume 3, Terrestrial, Algal, and Siliceous Indicators, Smol JP, Birks HJB, Last WM (eds). Kluwer: Dordrecht; 49–74. Bluemle JP. 1972. Pleistocene drainage development in North Dakota. Geological Society of America Bulletin 83: 2189–2194. Broecker WS, Kennett JP, Flower BP, Teller JT, Trumbore S, Bonani G, Woelfli W. 1989. Routing of meltwater from the Laurentide ice sheet during the Younger Dryas cold episode. Nature 341: 318–321. Brophy JA, Bluemle JP. 1983. The Sheyenne River: its geological history and effects on Lake Agassiz. In Glacial Lake Agassiz, Teller JT, Clayton L (eds). The Geological Association of Canada, Special Paper no. 26; 173–186. Campbell JM. 1983. A revision of the North American Omaliinae (Coleoptera: Staphylinidae). 4. The genus Olophrum Erichson. Canadian Entomologist 115: 577–622. Clark PU, Marshall SJ, Clarke GKC, Hostetler SW, Licciardi JM, Teller JT. 2001. Freshwater forcing of abrupt climate change during the last glaciation. Science 293: 283–287. Clarke AH. 1981. The Freshwater Molluscs of Canada. National Museum of Canada, Ottawa. Clayton L, Moran SR. 1982. Chronology of late-Wisconsinan glaciation in middle North America. Quaternary Science Reviews 1: 55–82. Clayton L, Moran SR, Bluemle JP. 1965. Intersecting minor lineations on Lake Agassiz plain. Journal of Geology 73: 652–656. Collins SL, Glenn SM, Gibson DJ. 1995. Experimental analysis of intermediate disturbance and initial floristic composition: decoupling cause and effect. Ecology 76: 486–492. Curry BB. 1997. Paleochemistry of Lakes Agassiz and Manitoba based on ostracodes. Canadian Journal of Earth Science 34: 699–708. Cvancara AM, Clayton L, Bickley WB Jr., Jacobs AF, Ashworth AC, Brophy JA, Shay CT, Delorme LD, Lammers GE. 1971. Paleolimnology of late Quaternary deposits: Seibold site, North Dakota. Science 171: 172–174. Copyright ß 2005 John Wiley & Sons, Ltd.

Dyke AS, Prest VK. 1986. Late Wisconsinan and Holocene retreat of the Laurentide Ice Sheet. Geological Survey of Canada, Map 1703A. Elson JA. 1967. Geology of Glacial Lake Agassiz. In Life Land and Water, Mayer-Oakes WJ (ed.). University of Manitoba Press: Winnipeg; 37–95. Faegri K, Iverson J. 1975. Textbook of Pollen Analysis. Hafner Press: New York. Fenton MM, Moran SR, Teller JT, Clayton L. 1983. Quaternary stratigraphy and history in the southern part of the Lake Agassiz basin. In Glacial Lake Agassiz, Teller JT, Clayton L (eds). Geological Association of Canada Special Paper no. 26; 49–74. Fisher TG. 2003. Chronology of glacial Lake Agassiz meltwater routed to the Gulf of Mexico. Quaternary Research 59: 2711–276. Fisher TG, Smith DG. 1994. Glacial Lake Agassiz; its northwest maximum extent and outlet in Saskatchewan (Emerson Phase). Quaternary Science Reviews 13: 845–858. Fisher TG, Souch C. 1998. Northwest outlet channels of Lake Agassiz, isostatic tilting and a migrating continental drainage divide, Saskatchewan, Canada. Geomorphology 25: 57–73. Glaser PH. 1981. Transport and deposition of leaves and seeds on tundra: a late-glacial analog. Arctic and Alpine Research 13: 173–182. Great Plains Flora Association. 1986. Flora of the Great Plains. University of Kansas Press: Lawrence, KS. Grimm EC. 2001. Trends and paleoecological problems in the vegetation and climate history of the northern Great Plains, U.S.A. Proceedings of the Royal Irish Academy 101B(1–2): 47–64. Harris KL, Moran SR, Clayton L. 1974. Late Quaternary stratigraphic nomenclature Red River Valley, North Dakota and Minnesota. North Dakota Geological Survey Miscellaneous Series no. 52; 1–47. Helgesen RG, Post RL. 1967. Saprophagous Scarabaeidae (Coleoptera) of North Dakota. North Dakota Insects Publication Number 7, Agricultural Experiment Station, North Dakota State University; 1–60. Hostetler SW, Bartlein PJ, Clark PU, Small EE, Solomon AM. 2000. Simulated influences of Lake Agassiz on the climate of central North America 11 000 years ago. Nature 405: 334–337. Hu FS, Wright HE, Ito E, Lease K. 1997. Climatic effects of Glacial Lake Agassiz in the midwestern United States during the last deglaciation. Geology 25: 207–210. Kapp RO. 1969. How to Know Pollen and Spores. Wm. C. Brown: Dubuque, IA. Karrow PF. 2002. Discussion of ‘Formation of large beaches in an area of rapid differential isostatic rebound: the three-outlet control of Lake Agassiz’ by JT Teller. Quaternary Science Reviews 21: 2115–2118. Kompelian MG, Schwert DP. 1986. Paleoenvironmental implications of insect remains from the Seminary site, Cass County, North Dakota. Proceedings of the North Dakota Academy of Science 40: 131. Kutzbach JE, Guetter PJ, Behling PJ, Selin R. 1993. Simulated climatic changes: results of the COHMAP climate-model experiments. In Global Climates since the Last Glacial Maximum, Wright HE, Kutzbach JE, Webb T, Ruddiman WF, Street-Perrott FA, Bartlein PJ (eds). University of Minnesota Press: Minneapolis; 24–93. Laird KR, Fritz SC, Grimm EC, Mueller PG. 1996. Century-scale paleoclimatic reconstruction from Moon Lake, North Dakota, a closed-basin lake in the northern Great Plains. Limnology and Oceanography 41: 890–902. Laird KR, Fritz SC, Cumming BF, Grimm EC. 1998. Early-Holocene limnological and climatic variability in the Northern Great Plains. The Holocene 8: 275–285. Leverington DW, Mann JD, Teller JT. 2000. Changes in the bathymetry and volume of Glacial Lake Agassiz between 11 000 and 9300 14 C yr BP. Quaternary Research 54: 174–181. Lindroth CH. 1963. The ground-beetles (Carabidae, excl. Cicindelidae) of Canada and Alaska. Opuscula Entomologica Supplementum 3: 201–408. Looman J, Best KF. 1987. Budd’s Flora of the Canadian Prairie Provinces, revised edition. Agriculture Canada, Research Branch, Publication no. 1662. MacDonald GM. 1990. Palynology. In Methods in Quaternary Ecology, Warner BJ (ed.). Geological Association of Canada: Newfoundland. Geoscience Canada Reprint Series 5; 37–52. Maher LJ. 1964. Ephedra pollen in sediments of the Great Lakes region. Ecology 45: 391–395. J. Quaternary Sci., Vol. 20(3) 255–267 (2005)

PALAEOENVIRONMENTS SOUTHERN LAKE AGASSIZ BASIN, USA Martin AC, Barkley WD. 1961. Seed Identification Manual. University of California Press: Berkeley; 221 pp. McAndrews JH. 1967a. Paleoecology of the Seminary and Mirror Pool peat deposits. In Life, Land and Water, Mayer-Oakes WJ (ed.). University of Manitoba Press: Winnipeg; 253–269. McAndrews JH. 1967b. Pollen analysis and vegetational history of the Itasca region, Minnesota. In Quaternary Paleoecology, Cushing EJ, Wright HE (eds). Yale University Press: New Haven, CT; 219–236. McAndrews JH, Berti AA, Norris G. 1973. Key to the Quaternary pollen and spores of the Great Lakes region. Royal Ontario Museum: Toronto; Life Sciences Miscellaneous Publication; 1–61. McAndrews JH, Stewart RE, Bright RC. 1967. Paleoecology of a prairie pothole: a preliminary report. In Glacial Geology of the Missouri Coteau, Clayton L, Freers TF (eds). North Dakota Geological Survey Miscellaneous Series 30; 101–113. McGinn RA. 1992. Climatology of the Delta Marsh area. University of Manitoba Field Station (Delta Marsh) Annual Report 27: 65–77. Montgomery FH. 1977. Seeds and Fruits of Plants of Eastern Canada and Northeastern United States. University of Toronto Press: Toronto, Ontario; 232 pp. Moran SR, Clayton L, Cvancara AM. 1971. New sedimentological and paleontological evidence for the history of Lake Agassiz: Snake curve section, Red Lake County, Minnesota. Proceedings of the North Dakota Academy of Science 24: 61–73. Moran SR, Clayton L, Scott MW, Brophy JA. 1973. Catalog of North Dakota radiocarbon dates. North Dakota Geological Survey, Miscellaneous Series 53; 1–50. Nambudiri EMV, Teller JT, Last WM. 1980. Pre-Quaternary microfossils—a guide to errors in radiocarbon dating. Geology 8: 123–126. Newbrey MG, Ashworth AC. 2004. A fossil record of colonization and response of lacustrine fish populations to climate change. Canadian Journal of Fisheries and Aquatic Sciences 61: 1807–1816. Owenby JR, Ezell DS. 1992. Monthly Station Normals of Temperature, Precipitation, and Heating and Cooling Degree Days 1961–1990: North Dakota. National Oceanic and Atmospheric Administration, Climatography of the United States, no. 81. Radle N, Keister CM, Battarbee RW. 1989. Diatom, pollen, and geochemical evidence for the palaeosalinity of Medicine Lake, S. Dakota, during the Late Wisconsin and early Holocene. Journal of Paleolimnology 2: 159–172. Reid JR, Olson BL. 1996. Geomorphology/stratigraphy of the Halstad, Minnesota site. In Quaternary Geology of the Southern Lake Agassiz Basin, Harris KL, Luther MR, Reid JR (eds). North Dakota Geological Survey Miscellaneous Series no. 82; 84–94. Rosendahl CO. 1948. A contribution to the knowledge of the Pleistocene flora of Minnesota. Ecology 29: 284–315. Ruben M, Alexander C. 1958. U.S. Geological Survey Radiocarbon Dates IV. Science 127: 1476–1487. Shay JM. 1999. Annotated vascular plant species list for the Delta Marsh, Manitoba and surrounding area. University of Manitoba Field State (Delta Marsh) Occasional Publication No. 2: Winnipeg; 1–52. Shay TC. 1967. Vegetation history of the southern Lake Agassiz basin during the past 12 000 years. In Life, Land and Water, Mayer-Oakes WJ (ed.). University of Manitoba Press: Winnipeg; 231–252. Shay TC. 1971. The Itasca Bison Kill Site, an ecological analysis. Minnesota Prehistoric Series no. 197. Minnesota Historical Society: Minneapolis; 133 pp. Stevens OA. 1963. Handbook of North Dakota Plants. North Dakota Institute for Regional Studies: Fargo, ND.

Copyright ß 2005 John Wiley & Sons, Ltd.

267

Stuiver M, Reimer PJ, Bard E, Beck JW, Burr GS, Hughen KA, Kromer B, McCormac G, van der Plicht J, Spurk M. 1998. INTCAL98 radiocarbon age calibration, 24 000–0 cal BP. Radiocarbon 40: 1041–1083. Teller JT. 1990. Meltwater and precipitation runoff to the North Atlantic, Arctic and Gulf of Mexico from the Laurentide Ice Sheet and adjacent regions during the Younger Dryas. Paleoceanography 5: 897–905. Teller JT. 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 JT. 2002. Reply to the comment by P.F. Karrow on ‘Formation of large beaches in an area of rapid differential isostatic rebound: a three-outlet model for Lake Agassiz’. Quaternary Science Reviews 21: 2119–2122. Teller JT. 2004. Controls, history, outbursts, and impacts of large lateQuaternary proglacial lakes in North America. In The Quaternary Period in the United States, Gillespie AR, Porter SC, Atwater BF (eds). Elsevier: Amsterdam; 45–61. Teller JT, Last WM. 1981. Late Quaternary history of Lake Manitoba, Canada. Quaternary Research 16: 97–116. Teller JT, Thorleifson LH. 1983. The Lake Agassiz—Lake Superior Connection. In Glacial Lake Agassiz, Teller JT, Clayton L (eds). Geological Association of Canada Special Paper no. 26; 261–290. Teller JT, Risberg J, Matile G, Zoltai S. 2000. Postglacial history and paleoecology of Wampum, Manitoba, a former lagoon in the Lake Agassiz basin. Geological Society of America Bulletin 112: 943–958. Teller JT, Thorleifson LH, Dredge LA, Hobbs HC, Schreiner BT. 1983. Maximum extent and major features of Lake Agassiz. In Glacial Lake Agassiz, Teller JT, Clayton L (eds). Geological Association of Canada Special Paper no. 26; 43–45. Thorleifson LH. 1996. Review of Lake Agassiz History. In Sedimentology, Geomorphology and History of the Central Lake Agassiz Basin, Teller JT, Thorleifson LH, Matile G, Brisbin WC (eds). Geological Association of Canada Field Trip Guidebook for the 1996 GAC/ MAC Joint Annual Meeting; 55–84. Upham W. 1895. The Glacial Lake Agassiz. United States Geological Survey Monograph no. 25; 1–658. Watts WA, Bright RC. 1968. Pollen, seed, and mollusk analysis of a sediment core from Pickerel Lake, northeastern South Dakota. Geological Society of America Bulletin 79: 855–876. Whitlock CW, Bartlein PJ, Watts WA. 1993. Vegetation history of Elk Lake. In Elk Lake, Minnesota: Evidence for Rapid Climate Change in the North-Central United States, Bradbury JP, Dean WE (eds). Geological Society of America Special Paper no. 276; 251–274. Wood SL. 1982. The bark and ambrosia beetles of North and Central America (Coleoptera: Scolytidae), a taxonomic monograph. Great Basin Naturalist Memoirs no. 6; 1–1359. Yansa CH. 1998. Holocene paleovegetation and paleohydrology of a prairie pothole in southern Saskatchewan, Canada. Journal of Paleolimnology 19: 429–441. Yansa CH. 2002. Spatial and temporal vegetation changes in the northeastern Great Plains during the Late Pleistocene and Early Holocene. Unpublished Dissertation, Department of Geography, University of Wisconsin-Madison; 337 pp. Yansa CH, Basinger JF. 1999. A postglacial plant macrofossil record of vegetation and climate change in southern Saskatchewan. In Holocene Climate and Environmental Change in the Palliser Triangle, Southern Canadian Prairies, Lemmen DS, Vance RE (eds). Geological Survey of Canada Bulletin 534: 139–172.

J. Quaternary Sci., Vol. 20(3) 255–267 (2005)