Abstract Paleoindian sites (approximately 7,500 to 11,500 B.P.) on “raised beach terraces” and “strandlines” in the Great Lakes region are a well-documented aspect of early settlement patterns. In the lower Great Lakes region of Ontario, research on the early (fluted-point) cultures (approximately 9,500 to 11,500 B.P.) has been particularly productive over the past two decades, with survey and excavations concentrated on proglacial Lake Algonquin strandlines (Deller, 1976; Storck, 1982, 1984). Radiometrically dated materials from these beaches were used to establish maximum geological dates for Lake Algonquin strandlines; for sites along the north shore of Lake Ontario there are bracketing geological dates (Roberts, 1984). In the northwestern Lake Superior Basin there is a similar pattern of numerous late Paleoindian (Piano) sites of the Lakehead Paleoindian complex (Fox, 1975, 1980; Dawson, 1983; Julig, 1984,1988) found on or near beaches of an earlier, deeper stage of Lake Superior: Lake Minong (Fig. 1). Prehistoric groups often favored shorelines for settlement sites due to their high biological productivity (and subsistence resources), as well as a ready supply of fresh water. Along the north shore of the upper Great Lakes there are regional loci of Paleoindian sites at coastal locations where there was also abundant lithic material for stonetool manufacture. Several such loci of large coastal quarry/workshop sites, surrounded by numerous smaller sites, include the Cummins site and the Lakehead Paleoindian Complex in the northwestern Lake Superior Basin (Fox, 1975; Julig, 1988), and the Sheguiandah and George Lake Complex of sites in northern Lake Huron (Lee, 1954,1955

Use of modern pollen spectra as a basis for interpreting diagrams of fossil spectra requires compilation of the modern spectra in readily accessible form, such as contoured maps of percentage values of individual pollen types. Maps are presented that show the distribution of modern pollen based on 606 samples from central North America (lat 35°N to 70°N, long 75°W to 110°W). Only data published after 1960 are included, and data from 69 sites are presented for the first time. The maps show differences in the pollen percentages among vegetational regions. For example, peak pollen values of Cyperaceae and Betula occur in the tundra; high values of Picea appear in the northern boreal forest; high values of Pinus appear in the southern boreal forest and the adjacent conifer-hardwood forest; high values of Tsuga , Fagus , and Acer occur eastward in the conifer-hardwood forest; high values of Quercus , Ambrosia , Fraxinus , and Carya occur in the deciduous forest; and high values of nontree pollen (Gramineae, Artemisia , Chenopodiineae, and Compositae) appear in the prairie. Trend-surface analysis and principal components analysis summarize the regional trends of each pollen type and illustrate the patterns of covarying pollen types within the data. Although these data provide a basis for interpreting the major fossil pollen zones for Holocene time in central North America, additional sampling and more detailed examination of the data are required for description of the fine-scale changes within fossil zones.