Minnesota Map
Institute for Minnesota Archaeology logo From Site to Story logo

Sources - Papers








Vol. 50, No. 2 1991                                1989

The Application of Remote Sensing Techniques to Settlement Pattern Analysis at the Red Wing Locality

by Clark A. Dobbs, Institute for Minnesota Archaeology

© 1989 Minnesota Archaeological Society


Mound Relocation and Photogrammetric Mapping
Energy Park Site
Silvernale Site
Controlled Surface Collection
Soil Resistivity
Discussion and Conclusions
References Cited
  (big thumbnails)     (medium)     (small)


The Red Wing Locality is a group of Middle Mississippi-related mound and village sites in southeastern Minnesota. Analysis of settlement patterns within the Locality is being used to evaluate the interaction between and influence of Middle Mississippian and local populations during the 12th and 13th centuries AD. A variety of techniques, including photogrammetric mapping, soil resistivity, and controlled surface collection are being employed to delineate the settlement plans of sites in the region. Investigations at the Energy Park Site (21GD52/GD158) and the Silvernale Site (21GD3/GD17) illustrate the strengths and weaknesses of these methods at sites in the Upper Midwest.


For the last forty years, archaeologists have recognized the importance of the study of settlement patterns to gain a better understanding of ancient societies. Complex societies, with their varied forms of architecture, are particularly amenable to settlement studies. In the New World, archaeologists first concentrated on the complex societies of South America, but similar studies quickly followed in North America, particularly the Midwestern United States (e.g. Wiley 1953, 1956; Roper 1979). In the Midwest, however, the application of settlement pattern analysis is made more difficult by a variety of factors. Sites typically contain a lower density of artifacts and other debris of archaeological interest; many ancient earthworks have been obliterated by a century or more of cultivation; and structures were generally made of bark, wood, and animal hide which are perishable in eastern North America.

The emergence of remote sensing methods during the last several decades has provided archaeologists with a suite of tools that may be used to obtain information about archaeological properties without excavation. In this paper, we review the application of remote sensing methods to the problem of developing models of inter- and intra-site settlement patterns at the Red Wing Locality in the Upper Mississippi River Valley.

Located some 500 miles upriver from Cahokia and the American Bottom (Figure 1), the Red Wing Locality is situated on the Mississippi River at the confluence of the Cannon and Trimbelle Rivers (Dobbs 1985a; Dobbs 1986; Dobbs and Breakey 1987) in Goodhue County, Minnesota and Pierce County, Wisconsin. The Locality encompasses an area of some 58 square miles and contains several thousand earthworks, eight major village sites, and dozens of smaller secondary sites (Figures 2, 3). The Locality is the most northern center of Mississippian interaction in North America and is perhaps the largest cluster of Mississippian-related sites north of Illinois.

In 1983, the IMA began a long-term research program at the Locality. This program was designed to examine three broad questions: the transition from hunting and gathering to maize horticulture in the region, the character of the trade and interaction between the Red Wing Locality and Middle Mississippian cultures farther to the south, and the nature of the human/land interactions during a period of climatic instability between ca. AD. 900 and 1300.

Addressing these broad research questions has been complicated by the fact that the culture history and interrelationships between the sites in the Locality remain murky. Earlier studies (Gibbon 1974) relied on limited fieldwork conducted the 1940s and 1950s. Despite subsequent research, the chronology of the region remains poorly known because radiocarbon dating has not been particularly successful in developing a culture-historical sequence of the relatively short period of time during which the Locality was occupied (ca. AD. 1000 - 1300). These limitations make it difficult to address the more sophisticated questions in which we are interested. We have therefore structured the first stages of our investigations to address two basic questions which must be resolved before we can move forward.

The first question involves the relative contemporaneity of the sites in the area. Were these sites occupied at approximately the same time, reflecting a brief and very intense episode of habitation? Or, do they represent a sequence of occupation which documents the transition from Late Woodland to Mississippian to the subsequent Oneota cultures?

A second problem involves the interaction between the Locality and Middle Mississippian groups to the south. Some scholars have argued that there was a movement of Middle Mississippian people to the north, resulting in 'site-unit intrusion' at the Locality (Grifffin 1961) or at least that there was very strong influence on Red Wing from the American Bottom (Stoltman 1986). Others propose that the sequence at Red Wing was an 'in-situ' development and that the Mississippian sites in the area were more the consequence of the diffusion of Middle Mississippian traits into a local population (Gibbon 1974,1979; Gibbon and Dobbs n.d.).

Much of the discussion of the Red Wing Locality has been based on analysis of ceramics and other artifacts (e.g. Wilford 1945, 1952; Gibbon 1978,1979). However, the ceramic assemblages from the Locality are complex and some of the sites appear to have been occupied over an extended period of time. We felt that a continued emphasis solely on comparative artifact studies would not resolve the questions we were interested in. We hypothesized that if there was an in-situ developmental sequence in the region over a period of time, the settlement plan and physical setting of the major village sites should be different from one another. To test this hypothesis, we decided to examine the inter- and intra- site settlement patterns of the major villages in the region.

The major village sites within the Locality are quite large. It is not possible (and probably not desirable) to completely excavate each one. Therefore, we have applied a variety of methods, including soil resistivity, photogrammetric mapping, and controlled surface collection, to discern internal site patterning. This paper reviews the effectiveness of these methods in the study of two sites, the Energy Park Site (21GD52/GD158) and the Silvernale Site (21GD3/GD17).

The Energy Park Site comprises a mound group (21GD52) and village site (21GD158) complex situated on a glacial outwash terrace overlooking the confluence of the Cannon and Mississippi Rivers (Figure 4). The surficial sediments on the terrace consist of interbedded sands and gravels. The village site covers approximately 5 acres and the mound area, 20 acres. The site is presently interpreted as a seasonal village possibly contemporary with the Moorehead Phase at Cahokia.

The mound group surrounds the village site on the landward side and was first mapped in 1885 (Figure 5) by Theodore Hayes Lewis of the Northwestern Archaeological Survey (cf. Dobbs 1991). Lewis identified 64 mounds, including a flat-topped pyramidal mound (Figure 6). Since 1885, the area has been continually cultivated and only a few mound remnants are readily evident.

The village portion of the site was unknown until it was discovered in 1984 by IMA archaeologists (Figure 7). Controlled surface collections were made at the site in 1986 and 1988. Soil resistivity surveying was conducted in 1986, 1987, and 1988. Excavations were conducted in 1987, 1988, and 1990.

The Silvernale Site is a large mound (21GD17) and village (21GD3) site located on a glacial terrace immediately adjacent to the delta of the Cannon River near its confluence with the Mississippi. The site is roughly one mile east of the Energy Park Site. The mound portion of the site encompasses roughly 40 acres and the village covers some 10 acres (Figure 8). We suspect that Silvernale is roughly contemporary with the Sterling Phase at Cahokia and is therefore earlier than the Energy Park Site.

The mounds were originally mapped in 1885 by T. H. Lewis, who reported 226 mounds and observed that there were fifty to seventy-five more mounds which he did not map in a cornfield (Figure 9).

The village portion of the site was first identified by Lloyd A. Wilford who excavated there in the 1940s (Wilford 1945). Salvage excavations were also conducted by Minnesota archaeologists in the 1970s. The site itself has been severely affected by cultivation and a large portion of the mound group and village site was destroyed by construction of an industrial park in the mid-1970s. However, because the site is so large, is probably early in the Red Wing sequence, and is the type site for the Silvernale Phase, it remains particularly important.


More than 2,000 earthworks have been documented within the Red Wing Locality. The majority have been obliterated by cultivation and urban development. Fortunately, all of these groups were mapped by Theodore Hayes Lewis of the Northwestern Archaeological Survey at the end of the 19th century before the worst damage was done. Lewis was trained both as an archaeologist and surveyor and his field notebooks provide detailed information on the size, form, and height of each individual mound within each group (Figure 5). He also conducted open traverse surveys providing a bearing and distance from each mound in the group to the next. Lewis' data makes it possible to recreate the configuration of individual mound groups even though the mounds themselves are no longer readily visible.

Lewis' data are an invaluable source for studying the earthworks within the Locality. However, there are two significant problems with his material. First, his surveying instruments were considerably less accurate than those available today. He used an open surveyor's compass which had an accuracy of perhaps plus or minus one degree. Therefore, there is the potential for considerable error to accumulate over the course of one of his traverses. Second, he apparently did not routinely close or backsight his traverses. Therefore, we have no way of confirming the accuracy of mound location maps computed from his bearings and distances, unless enough of the mound group survives to provide some independent verification. Of course, for groups that have been destroyed or badly damaged, independent verification of Lewis' information is difficult or impossible.

The patterning of mounds both across the landscape and at individual sites is an important part of the settlement pattern at the Locality. Further, we hypothesize that the mounds are an integral part of the large village sites. Without a detailed map precisely showing the locations of mounds relative to village materials, it is impossible to rigorously evaluate the interconnections between these types of cultural features.

For many years, it was generally assumed that these mounds were associated with the Woodland Tradition. Investigations by Maxwell (1950) and Johnson, Peterson and Streiff (1968) provided direct evidence that at least some of these mounds were associated with the Oneota and/or Mississippian cultures. More recent evaluation of mound patterning by Dobbs (1985) indicates that there is distinct patterning of groups across the landscape and that each major village site has a large mound group associated with it. Typically the village is on a bluff or high terrace overlooking a river and the mound group surrounds the village on the landward side. Examination of the layout of several of these groups suggests that there is also intentional patterning within each individual group. However, the structure and meaning of this patterning remains unclear.


Thanks to the efforts of the Goodhue-Pierce Archaeological Society, the Goodhue County Historical Society, the Minnesota Parks and Trails Foundation, and the State of Minnesota, the village and a portion of the mounds at the Energy Park Site have been purchased and preserved. In 1989, we initiated a planning process to outline the development of this site as an archaeological park. The planner involved had prepared a detailed contour map of the site using aerial photography and photogrammetric mapping techniques. This method requires large-scale aerial photography (ca. 1: 1000) and precisely-located ground control points. Using this information, a rectified high-resolution stereoscopic model can be prepared from which maps can be generated. To our surprise, probable mound remnants were visible in the stereoscopic model that are not discernable on the ground surface. The end product was a scaled map that met National Map Accuracy Standards, showing topography and locations of probable mounds remnants. It proved possible correlate these remnants with a map prepared from Lewis' data (Figure 10).


Although the Silvernale Site is particularly important, a modern, comprehensive map of the site has never been prepared. Because so much of the site has been badly disturbed, we had assumed that it would be very difficult to obtain information about the settlement plan at Silvernale. The success of the Energy Park Site map, however, led us to attempt a similar project at Silvernale.

High-quality aerial photography was available for the Silvernale Site (Figure 11). In 1967, the Minnesota Dept. of Transportation took a series of aerial photographs that included the site prior to the time when most construction at the site occurred. This photography clearly shows possible mound remnants and we hoped that we could again create a detailed map linking mounds to actual on-the-ground features. Although the ground surface had changed since 1967, we were able to obtain appropriate ground-control points and Horizons, Inc. prepared a stereoscopic model and map for us. We obtained this map in digital format and transferred it to our CAD system (Figure 12). We then created a map of the mounds and ground features using Lewis notes (Figure 13). Lewis initially mapped the site in August of 1885. After examining the Horizons, Inc. map and Lewis' map, we were able to determine that Mound 15 corresponded with a specific mound remnant visible in the aerial photography. We then overlaid the Lewis map on the Horizons map and rotated them to obtain a fit between his mounds and the mound remnants visible in the photography.

We were able to fit Lewis' map to the Horizons map by using the mound remnants as control points (Figure 14). We observed that mounds taller than 4 feet were particularly apparent on the Horizons map and these were also used as control points. Further, Lewis had provided measurements from several of the mounds to the bluff edge and to a railroad which is still present. These points provided an additional way to align the Lewis map with the Horizons map.

It quickly became apparent that there was considerable divergence between Lewis' map and the Horizons map. Although mounds shown by Lewis in the westernmost portion of the site aligned very neatly with the Horizons map, this correspondence deteriorated for the central and eastern parts of the site. For example, Lewis indicated that Mound 206 was 25 feet from the center of the nearby railroad tracks. In aligning his map to the Horizons map, this mound was more seen 90 feet from the tracks and none of the other remnants in the area aligned with his map.

After a careful examination of both maps and his data, we identified the probable sources of this error. First, Lewis' measurements were consistently about 50 feet short per 1000 feet measured. Second, he measured the angles between Mounds 100 and 111 in two readings over a distance of more than 1000 feet. These types of long surveying 'shots' are an obvious source of potential inaccuracy and in this instance it appears that Lewis' readings deviated from the actual angle by about 3.5 degrees.

Using this information on Lewis' error, we were able to realign blocks of mounds into positions where they corresponded with mound remnants and other control points. The resulting map shows in detail where mounds actually are located on the ground surface (Figure 15). Even though many of the mounds mapped by Lewis are no longer visible, their locations can be plotted with confidence. Further, Lewis noted that there were 50 to 75 mounds in the cornfield that he did not map. We had assumed that the location of these mounds would never be known. However, many of these mounds were clearly visible in the 1967 photography and we were able to plot them. The end result is a base map, in digital form, that can be used to delineate and evaluate the internal settlement pattern at the Silvernale Site. It will also be possible to incorporate past excavation information onto the map as well as any future excavation that may take place. Obviously, ground truthing of results is essential, but it has not been possible to conduct this task on the massive scale that would be required for these sites.


It has been repeatedly demonstrated that meaningful patterns of artifact distribution often survive in sites that have been extensively cultivated (O'Brien and Lewarch 1981) and that an effective way to delineate these patterns is by controlled surface collection of artifacts. Controlled surface collection is not normally subsumed under the term 'remote sensing' but is one important technique for retrieving information from extensively disturbed sites.

As part of its ongoing research program, the Institute is conducting controlled surface collections at all of the major sites in the Red Wing Locality. Various methods of horizontal control are being employed depending on the size and density of the site. At sites where artifact density is low, individual artifacts are piece-plotted. For sites of medium density and area, collection is done using either five meter squares or two meter 'dog leash' units spaced at five meter intervals. At very large, dense sites we employ two-meter 'dog leash' units on a ten-meter grid.

The Energy Park Site had been in pasture since the 1920s and had not been subjected to the intense collecting by amateur archaeologists that some other sites in the region have experienced. After plowing the site to a shallow depth, we conducted a complete controlled surface collection of the village portion of the site in 1986 and again in 1988. The site was gridded into 5 meter squares and the contents of each square were collected. The principal artifacts recovered were chipped stone tools, debitage, and pottery.

Debitage was separated by raw material category and size graded. The assemblage is dominated by three specific raw materials: Hixton Silicified Sandstone (imported from Wisconsin), Cedar Valley Chert (imported from quarries in Mower County, Minnesota), and Prairie du Chien chert (a local raw material). In general, the assemblage suggests that the principal tool making activity appears to have involved resharpening and manufacture of small non-patterned tools and that primary tool manufacture did not occur at the site.

Stone tools were identified by functional class. Ceramics were fragmentary and were sorted by surface treatment and temper. All ceramics fit generally into the Oneota/Mississippian category. Only a very few rim sherds were recovered and no further identification was normally possible. Ground stone tools and fire-cracked rock were conspicuous by their absence.

Artifact totals for each collection unit were entered into a contouring program to generate contour maps of artifact density for the site. Figure 16 is a composite artifact density map using both the 1986 and 1988 collections and shows the distinctive patterning occurring within the site. Figure 17 illustrates eleven selected artifact concentrations.

Table I contains information on the debris profiles of the total assemblage from the site and each individual artifact concentration. Although there are some broad similarities between the concentrations, there are distinct differences among them, suggesting that they represent different functional or activity areas. Areas A and B, for example, probably represent places where tool resharpening and small non-patterned tool manufacture was the dominant activity. In contrast, Area H probably represents the debris from domestic or household activities.


Soil resistivity has a long history in archaeological research and its strengths as a discovery technique on historic sites in North America are unquestioned (see Weymouth and Huggins 1985, Wynn 1990). Its utility on prehistoric sites, however, is not always as apparent, in large measure because the signatures of middens and earth-filled pits at prehistoric sites are typically more subtle than those generated by stone walls and other structural features found on Euro-American sites.

At the Energy Park Site, we have conducted extensive soil resistivity surveying in 1986,1987, and 1988. At this writing, approximately 50% of the habitation area has been surveyed. The targets of interest are deep pit features that are typically .75 to 1.25 meters in width and 1 to 2 meters in depth. These features contain deposits of rich, highly organic materials. This pit fill is distinctive and contrasts markedly with the interbedded sands and gravels of the surrounding parent material.

The resistivity surveys were conducted using a Geohm 3 resistivity unit with a Wemmer array. Given the size of the target pit features, probes were spaced at l meter intervals. The surveying was tied to the master coordinate system for the site, which is the same system used for the controlled surface collection and excavation.

Because the resistivity survey was conducted over a period of days, weeks, and years, control measurements were taken over the same line every day prior to commencing survey. The importance of control measurements quickly became apparent. Soil moisture has a profound effect on resistivity readings and the actual readings varied significantly between days and weeks. Figure 18 shows the variation of resistivity readings at different times along the control line.

The resistivity readings taken during the fail of 1986 were obtained during a period when soil moisture was particularly high. They show very little variation and are not included in this analysis. The readings taken during 1987 are much more helpful, but the readings taken during the drought year of 1988 are notable for the high contrast between pit features and the surrounding soil matrix. The remaining resistivity work at the site is scheduled for July and early August, a time of maximum dryness in this part of the Midwest.

Analysis and interpretation of the resistivity results was at first a difficult problem. Various methods for analyzing such data have been used in the past, including analysis of individual resistivity lines, complex search programs looking for particular signatures, and so on. However, because we wished to produce a map of large, contiguous areas, none of these methods seemed appropriate.

Ultimately, we used a contouring program to generate maps of the resistivity data. We first attempted to generate a complete resistivity map of the site. However, it quickly became apparent that contouring readings for a six acre area resulted in a significant loss of resolution in the analysis because of the gridding algorithms used by the software. We discovered that contouring a 20 by 20 meter block (400 readings) produced very good results. By setting the parameters in the contouring software, we were able to hachure distinct resistivity lows so that they would be readily distinguished from surrounding high readings. Since we believed that pit features would be characterized by particularly low readings, surrounded by highs characteristic of the parent materials, this approach seemed to be both straightforward and easily interpreted. A composite map of soil resistivity for the portions of the site we have completed is shown in Figure 19.

The results of the resistivity analysis using the contouring method appeared to be very informative. Areas of resistivity highs were clearly delineated and anomalies which we interpreted as deep pit features were obvious. However, it was unclear whether our interpretations of the analysis were correct. Therefore, during the 1987 and 1988 excavations, we used the resistivity data to guide the placement of two excavation blocks. These blocks were excavated to compare what was actually in the ground with the resistivity data.

In 1987, we placed Excavation Block Five over a large resistivity anomaly that conformed to a noticeable change in the vegetation (Figure 20). We suspected that the anomaly might represent a house floor with one or more pit features within it. Figures 21 and 22 are horizontal plan maps of Excavation Block Five at two different levels. What we discovered during excavation was that there was an extensive area of black, greasy staining with large amounts of fire-cracked rock in the center of the block. This conformed to the area shown as a modest resistivity anomaly in Figure 20. This area of staining (defined as Feature 53a) was shallow and varied between 5 and 10 cm in depth beneath the base of the plow zone. Underneath this area of staining were two overlapping pit features. Feature 70 was 21 cm in depth and Feature 53b was 95 cm in depth (Figures 23,24). When we compared the resistivity data with the excavated data, it was clear that resistivity was providing good horizontal and vertical delineation of deep pit features and other types of features in this block. The one problem we discovered was that it appeared that the resistivity anomalies were displaced roughly 1.5 meters from the actual horizontal location of the features uncovered in the excavation. It is not clear if this is a distortion introduced during the interpretation of the resistivity data or whether the resistivity survey technique "sees" phenomena that do not necessarily exactly correspond to the features visible to the archaeologists. Figure 25 contains the resistivity data and labels indicating the features discovered during excavation.

In 1988, we placed Excavation Block Six in an area of low artifact density which contained numerous resistivity anomalies, one of which appeared to be particularly large and deep. The resistivity data for Block 6 is shown in Figure 26. Excavation of Block Six confirmed that the resistivity data were consistently detecting the subsurface deposits. The large resistivity anomaly turned out to be two overlapping pit features, one of which (Feature 85) was among the largest pit features ever found at the site. Feature 85 measures 1.6 meters in width and 2.4 meters in depth. Eight other features were identified in the Block, again corresponding to resistivity anomalies. Figure 27 is a horizontal plan map of Excavation Block Six and Figure 28 contains the resistivity data with the features they represent labeled. Figure 29 is a vertical profile of Feature 85.

Based on the 1987 and 1988 excavations, we conclude that resistivity surveying is a highly effective technique for locating pit features and other types of sub-surface remains on Oneota and Mississippian sites at the Red Wing Locality. The superb results we have obtained are due to the sharp contrast between the coarse, relatively droughty, inorganic parent material of the terrace and the highly organic pit fill and staining which retains moisture far longer than the parent material.

We also conclude that our admittedly simple contouring approach to analyzing the resistivity data is effective. Not only does the contouring approach delineate the location of deep pit features, it often records shallow deposits and staining as well. In fact, evaluation of the resistivity maps provides an important complement to excavation since, in some cases, it shows cultural deposits at the base of the plow zone that may not be readily apparent during excavation. The only weakness to the approach is that the resistivity anomalies are sometimes delineated as much as a meter and a half away from the actual horizontal locations of the pit features.


One tenet of our research at the Red Wing Locality is that the prehistoric sites of the region must be treated as the material remains of coherent, functioning communities. We have chosen the analysis of settlement patterns as one tool to use in this study. By combining several different remote sensing techniques, we are able to generate models of internal site plan that may be compared to one another and used to guide problem-oriented excavation at each site.

Figure 30 is a composite map of feature locations indicated by soil resistivity and artifact density determined from controlled surface collection at the Energy Park Site. At this scale, the map is somewhat difficult in interpret. However, several patterns are apparent. We know from excavations at the Bryan Site (Dobbs 1985b) that structures often are surrounded by pits, but that their interiors are generally devoid of features.

Figure 31 is a detail view of the area surrounding Artifact Concentrations A and B. These concentrations are interpreted as activity areas where tool resharpening and the manufacture of small non-patterned tools took place. The resistivity data suggest that features in this area are relatively small, co-occur with areas of high artifact density, and are not distributed in a fashion consistent with structures we have excavated elsewhere. Further, the resistivity data suggest that there are individual pits present without the associated staining and widespread deposits we observed in Excavation Block 5.

Figure 32 shows the distribution of artifacts and resistivity anomalies around Artifact Concentration C, which presents a markedly different picture from Concentrations A and B. There is a large central area where there are no features and where artifact density is very low. Areas of high artifact density occur to the west and east of this central area, and a series of pit features surrounds the central zone. Further, many of the pit features are within larger resistivity anomalies that are similar to those identified in Excavation Block 5. We interpret these as areas where there has been extensive and perhaps repeated disturbance associated with refuse disposal, pit use, and so on. In sum, Figure 35 has most of the characteristics that we would associate with some type of fairly large structure.

Figure 33 is a composite map showing the relationship of artifact concentrations, mounds, and site topography at the Energy Park Site. This map represents a picture of the overall settlement plan of a 12th century community which probably was a seasonal village contemporary with the Moorehead Phase at Cahokia. As we complete our resistivity surveying, refine our mapping of the mound group, and expand our excavations, we will be able to interpret the entire site without excavating the entire area. Further, the settlement plan at Energy Park can be compared with that of other sites to test our hypothesis about the sequence of development within the Red Wing Locality.

Based on the examples discussed in this paper, we conclude that these methods are effective research tools.Scholars should no longer be willing to accept interpretations of large sites in the Upper Midwest based on data collected from limited number of excavation units. Furthermore, we can no longer responsibly or economically excavate large areas of complex village sites for a limited amount of money. Thus, remote sensing allows us to more effectively cope with our responsibilities for stewardship, preservation, and management of archaeological properties in the region.

The work at Cahokia and numerous other sites in the American Bottom have clearly demonstrated the value of intensive, regional approaches to the study of Mississippian culture in the Midwest. Our work at the Red Wing Locality is designed to provide a similar regional framework for this northern complex of Mississippian and Oneota sites. Our interpretation and study of individual sites is enhanced by the development of settlement pattern models for each site. Each site may thus be interpreted on a community, not simply an artifactual, level. Further, when these models are combined and examined with other data from the Locality, we can generate and test meaningful hypotheses about cultural process, site function, and other aspects of our broad research objectives.

This process is already underway. Controlled surface collections and mapping at a variety of sites indicate that the internal plan of villages at the Locality varies significantly. Although some of this variability may be related to the function of individual sites (e.g. seasonal village, permanent village, etc.), we believe that a significant portion of this variability is related to changes occurring through time. We see a similar type of variability in the composition and distribution of mound groups as well, particularly those associated with the large village sites. We suspect that sites with large, loosely organized mound groups like Silvernale and the Mero I Site (47PI02) represent communities that are relatively early in the sequence. Sites like Adams (47PI 12), Mero 2, and Energy Park have tightly constrained groups of mounds that surround the village in an arc may be later in the sequence. Debris profiles and artifact patterning at sites like Bryan (21GD4) and Mero I may represent permanent or semipermanent villages. The patterning at Energy Park suggests that it was a rather different type of community, while the habitation portion of the Adams Site obviously represents a series of periodic, seasonal occupations episodes. A detailed analysis of this model of Mississippian development, growing out of our settlement pattern analysis, will be the subject of a future paper.


Dobbs, Clark A.
1985a An Archaeological Survey of the City of Red Wing, Minnesota (2 Volumes).Reports of Investigations Number 2. institute for Minnesota Archaeology, Minneapolis.

1985b A preliminary report on the analysis of materials recovered during excavations at the Bryan site (21 GD4), Goodhue County, Minnesota. Reports of Investigations Number 3. Institute for Minnesota Archaeology, Minneapolis.

1986 Archaeological excavations at the Bryan Site (21GD4), Goodhue County, Minnesota: 1983 and 1984 (2 volumes). Reports of Investigations Number 8. Institute for Minnesota Archaeology, Minneapolis.

1991 The Northwestern Archaeological Survey: An appreciation and guide to the Field Notebooks. Institute for Minnesota Archaeology Reports of Investigations Number 135. Minneapolis.

Dobbs, C. A. and K. Breakey
1987 A preliminary report on investigations at the Energy Park site (21GD158): a Silvernale Phase village at the Red Wing Locality. 1987 Midwest Archaeological Conference, Milwaukee.

Gibbon, G. E.
1974 A model of Mississippian development and its implications for the Red Wing area. In Aspects of Under Great Lakes archaeology. Papers in Honor of Lloyd A. Wilford edited by E. Johnson. Minnesota Prehistoric Archaeology Series, No. 11, Minnesota Historical Society, St. Paul.

1978 A simplified algorithm model for the classification of Silvernale and Blue Earth phase ceramic vessels. In Some studies of Minnesota prehistoric ceramics: papers presented at the first Council for Minnesota Archaeology Symposium-1976. Edited by A. R. Woolworth and M. A. Hall.

1979 The Mississippian occupation of the Red wing area. Minnesota Historical Society, St. Paul. Gibbon, G. E. (Ed.).

1982 Oneota Studies. Publications in Anthropology No. 1, University of Minnesota, Minneapolis.

Gibbon, G.E. and C. A. Dobbs
n.d. The Mississippian presence in the Red Wing area of Minnesota. Paper in volume submitted for publication, J. V. Stoltman, editor.

Griffin J. B.
1961 Some correlations of climatic and cultural change in eastern North American prehistory. Annals, New York Academy of Science,95:710-7 17.

Johnson, E., M. Q. Peterson and J. Streiff
1968 Birch Lake burial mound group. Journal of the Minnesota Academy of Science, 36(1): 3-8.

Maxwell, M. S.
1950 A change in the interpretation of Wisconsin's prehistory. Wisconsin Archeologist, 33: 427-443.

O'Brien, M. J. and D. E. Lewarch
1981 Plowzone Archeology: contributions to theory and technique. Vanderbilt University Publications in Anthropology Number 27. Nashville, Tennessee.

Roper, D. C.
1979 Archaeological survey and settlement pattern models in central Illinois. Midcontinental Journal of Archaeology Special Paper #2.

Stoltman, J. B.
1986 The appearance of the Mississippian Cultural Tradition in the Upper Mississippi Valley. In J. B. Stoltman (ed.), Prehistoric Mound Builders of the Mississippi Valley. The Putnam Museum, Davenport, Iowa.

Weymouth, J. W. and R. Huggins
1985 Geophysical surveying of archaeological sites. In Rapp, G. and Gifford, J. A. (eds.), Archaeological Geology. Yale University Press, New Haven, CT.

Wilford, L. A.
1945 Three village sites of the Mississippi Pattern in Minnesota. American Antiquity 11 :32-40.

1952 The Silvernale Village and Mounds Site. Unpublished manuscript on file, Department of Anthropology, University of Minnesota, Minneapolis.

Willey, G. R.
1953 Prehistoric settlement patterns in the Viru Valley, Peru. Bureau of American Ethnology Bulletin Number 155. Washington, D. C.

1956 Prehistoric settlement patterns in the New World. Viking Fund Publications in Anthropology # 10. New York.

Wynn, J. C.
1990 Applications of high resolution geophysical methods to archaeology. In Archaeological Geology of North America, N. P.. Lasca and J. Donahue, (eds.). Geological Society of American, Centennial Special Volume 4. Boulder, Colorado.


The research on which this paper is based was funded by gifts from the City of Red Wing, the Red Wing Area Fund, Johnson and Scofield Land Surveyors, numerous other Red Wing area businesses, and gifts from private individuals. Analysis and preparation of this paper was supported by the Friends of the Institute for Minnesota Archaeology and the IMA Sponsored Research Program.

Orrin C. Shane III (The Science Museum of Minnesota) has been a consistent guide and advocate for controlled surface collection and resistivity surveying. Orrin, Jack Davies, and other Science Museum staff have been exceptionally gracious in providing equipment, technical assistance, and numerous critical comments in implementing the research described here.

The Goodhue County Historical Society has been consistently helpful and supportive of our efforts at the Energy Park Site, as has the City of Red Wing and the Red Wing Port Authority. We particularly wish to thank Mr. Dean Massett (City Council Administrator) and Mr. Brian Peterson (Planning Coordinator) for their continuing interest and assistance.

Funds for purchasing the Energy Park Site were provided by several private individuals and a special appropriation from the Legislature of the State of Minnesota. We particularly acknowledge the assistance of the Minnesota Parks and Trails Council and Foundation, the Goodhue County Archaeological Site Management Board, the Goodhue-Pierce Archaeological Society, and Chuck Richardson, Jean Chesley, Elizabeth Hedin, and others in this effort.

Curator of Collections Kim Breakey provided invaluable editorial assistance and support, John Redmann drafted many of the figures that illustrate this paper, and George Christiensen provided critical reading of the manuscript.

An earlier version of this paper was presented in 1991 at the 56th Annual Meeting of the Society for American Archaeology, New Orleans, as part of the symposium "An Archaeological Perspective on Geological Prospection, or when not to dig."


Sources Stories Credits Search Contents Links
Northern Headwaters Twin Cities Metro Area Red Wing Locality

From Site to Story web address
© 1999 The Institute for Minnesota Archaeology
Email us: feedback@fromsitetostory.org
Updated 29 Jun 1999