An apparent late woodland boundary in Western Wisconsin

ABSTRACT

The Effigy Mound culture of southern Wisconsin has been examined for well over a century. Until recently, differential distributions of mound forms have been largely ignored. Investigations in the Driftless Area of western Wisconsin reveal a quantifiable distinction in selected effigy mound forms, with sharp spatial separation between the adjacent Bad Axe and Coon Creek drainages. This segregation is also reflected in projectile point styles, preferred lithic raw materials, and ceramics, suggesting a distinct boundary between Late Woodland groups of the southern (Eastman phase) and northern (provisional Lewis phase) portions of the Driftless Area. These data have implications for the emergence of nucleated Oneota groups at Red Wing and Apple River.

Introduction

During his career at the University of Wisconsin-Madison, Jim Stoltman devoted much of his research energies toward the unglaciated Driftless Area of southwestern Wisconsin. Beginning in the late 1970s, he began to concentrate his efforts in the Prairie du Chien locality, at the confluence of the Wisconsin and Mississippi rivers. Using diagnostic artifacts and radiocarbon dates from key stratified sites, including rockshelters, floodplain shell middens, and alluvial fans, he developed a chronological framework for Middle Archaic through Late Woodland cultures, presented in phase definitions for the Archaic (Stoltman 1997) and Woodland traditions (Stoltman 1979, 1986a, 1986b, 1990).

The Eastman phase represents the Late Woodland Effigy Mound culture in southwestern Wisconsin, and sites of this phase are common in the Prairie du Chien locality. It dates to ca. A.D. 750-1050 and is marked by effigy mounds, notched and unnotched arrow tips, and Madison ware ceramics (Stoltman 1990). This phase is also distinguished by intensive harvesting of Mississippi River mussels and the introduction of corn. Eastman phase sites are located on all landforms in the Prairie du Chien locality, representing the first culture to utilize the floodplain, terraces, and uplands since the Archaic.

In contrast, the archaeological record indicates only minor occupation of the Prairie du Chien locality by late prehistoric Oneota people, and it appears that the mouth of the Wisconsin River was largely abandoned for several centuries before French exploration in the mid-1600s. Nonetheless, Stoltman ventured into the never-ending debate on the emergence of the Oneota culture. Specifically, he stirred discussion at the 1983 Red Wing Oneota conference and again in a regional synthesis of Upper Mississippi River culture history (1983) by categorically rejecting all Oneota dates that pre-dated Cahokia's Stirling phase (cal A.D. 1150-1250). Based on this interpretation, Stoltman argued that Oneota developed from and replaced its Late Woodland Effigy Mound predecessor.

Stoltman directed several dissertations and master's theses on Driftless Area research. In the late 1970s, Constance Arzigian, Jeffery Behm, Robert Boszhardt, and James Theler worked under Stoltman at Prairie du Chien. Having obtained a small Survey and Planning grant from the State Historical Society of Wisconsin, the group adopted a formal name (Wisconsin Archaeology Program, Driftless Area Project, or "WAPDAP"). Three WAPDAPers have continued research in the portion of the Driftless Area north of the Prairie du Chien locality through the University of Wisconsin-La Crosse and the Mississippi Valley Archaeology Center. As a matter of practicality, all Archaic and Woodland finds in the northern Driftless Area have been compared to the standard phase sequence established by Stoltman at Prairie du Chien.

One of the first realizations working near La Crosse was that there are no shell middens in the adjacent Mississippi River floodplain. Indeed, there are relatively few prehistoric sites known in the Mississippi River floodplain above the mouth of the Bad Axe River in Navigation Pool 9. This fact led to the recognition that the character of the Mississippi Valley is quite different between La Crosse and Prairie du Chien (Boszhardt 1986). This difference is reflected in the width of the trench, which is up to 15 km at La Crosse where the trench walls are composed of relatively soft sandstone. In contrast, the trench is only 3-5 km wide at Prairie du Chien because the walls there are formed of more resistant dolomite. The constricted valley at Prairie du Chien results in an overall swifter river and an ideal habitat for massive mussel beds. Hence, floodplain surveys between the Bad Axe River and Prairie du Chien have identified over 50 shell middens, whereas none are recorded above the Bad Axe all the way to St. Paul (Boszhardt 1982, 1988, 1989, 1996a; Dobbs and Mooers 1991; Mooers 1994; Overstreet 1983, 1984; Pleger 1996, 1997; Pleger et al. 2000; Wahls 1989). In addition, the wider nature of the valley near La Crosse is reflected in expansive Pleistocene outwash terraces, while these sand and gravel landforms occupy far less area near Prairie du Chien.