In his recently published paper summarizing the stratigraphic and morphologic record of southern Saskatchewan and adjacent areas, Christiansen (1979) proposed a chronology of deglaciation that is in error because it fails to integrate relevant data from outside of Saskatchewan. Christiansen's interpretation shows a progressive northward retreat of the glacial margin from Montana and North Dakota beginning before 16 000 years BP. By 15 500 years ago, the ice is shown to have retreated north of the Missouri River watershed (which previously had channelled all meltwater well south of the glaciated region) with runoff discharging eastward into Manitoba via the Souris and Pembina channels (Phase 3, Fig. 12 in Christiansen). At 14 000 years BP (Phase 4, Fig. 13), the glacial margin of Christiansen has retreated still farther north, with icefree conditions having developed over most of southern Saskatchewan; all rivers of the deglaciated region of Saskatchewan and Alberta are shown draining eastward into Lake Agassiz in Manitoba. The 14 000 year BP margin of Christiansen (1979) brings his interpretation, as well as that of Klassen (1972), into apparently irresolvable conflict with evidence to the south and east of Saskatchewan. In one of the most closely and certainly dated events in the glacial chronology of central North America, ice advanced southward over a forest bed near Des Moines, Iowa, at 14 000 years BP to deposit the Bemis Moraine (Ruhe 1969). Radiocarbon dates from six samples of wood, including two trees in growth position in loess beneath till, three trees rooted in the top of the loess and pushed over by the advancing ice, and one log incorporated in till, record this advance (Table 1). Lobate patterns and flow indicators confirm that, to reach Iowa, this ice must have flowed through the Lake Agassiz basin (Fig. 1) and over the only southern outlet from this basin, the Minnesota River Valley. In short, it does not seem possible to have water discharging eastward into the Lake Agassiz basin at 14 000 years BP, nor probably at 15 500 years BP, as Christiansen (1979) has shown because the basin was filled with ice, making eastward and southeastward drainage impossible. This point also has been made by Teller (1976) and Teller and Fenton (1980). On the basis of the evidence available in North Dakota, South Dakota, and Iowa, the last major glacial advance across the region reached its maximum western position about 14000 years ago (Clayton et al., in press). We suggest that Christiansen's Phase 2 (16 500 years BP) glacial boundary (Fig. 11) may more closely reflect conditions at 14 000 years BP. While the pattern of deglaciation in southwestern Saskatchewan and Alberta may have occurred as depicted by Christiansen, drainage from these western regions must have been southward into the Missouri River system until sometime after 14 000 years BP, rather than eastward through Lake Agassiz and into the upper Mississippi River system. Although it is not the main purpose of this discussion to explore the reason for Christiansen's unacceptably early chronology of deglaciation, we feel that a comment about the radiocarbon dates he used is pertinent. Nearly all of the dates that are used to document the deglaciation of southern Saskatchewan before 14 000 years BP are from finelydisseminated organic matter in fine-grained sediment (S-522 to S-526, GSC-1369, and S-
Read full abstract