At the beginning of the 1995 archaeological investigation at the Sugar Creek site, there was no ground surface visibility. The eastern half was in hay crop and the abandoned campground in the western half was covered in high weeds. The initial survey consisted of excavating 53 SVS units across the site (Figure 4.4), as described in the methodology section.

Artifacts from the 50.0 cm-by-50.0 cm units were taken to the GBL and sorted into four major artifact classes: ceramic, lithic, faunal, and historic. Counts for each of these major artifact classes were used to construct proportionally scaled plot density maps (Figures 4.5, 4.6, 4.7, and 4.8). In addition, weights were recorded in grams for the large quantities of waterworn pebbles present at the site. These weights were standardized in grams per cubic meter to determine whether geomorphological conditions were consistent across the site (Figure 4.9). Table 4.1 presents the totals for all the SVS units.

The most prominent characteristic of this site was the extremely high density of small waterworn pebbles (referred to in our cataloging system as river gravel), which appear to have been deposited as part of a glacial outwash plain, although differences in the density of the river gravel were observable over the course of the excavation. These pebbles, ranging in size from about 2.0 cm in diameter to coarse gravel, were present in both the matrix and the feature fill; because of the loose unconsolidated nature of the sediments, the pebbles had a tendency to obscure feature and soil boundaries. River gravel recovered from this site was weighed, not counted, and discarded. A one-liter sample of the gravel larger than one-quarter inch, or rocks that did not pass through a 1/4-inch screen, was weighed (1497.1 grams) and counted (1,516) to give an average weight of about 1 gram per pebble. Figure 4.9 shows the distribution of the river gravel in grams per cubic meter, after correcting for the depth of each SVS unit. What is apparent from the distribution is that density of river gravel near the surface falls off dramatically toward the south or southeast. The site drops in elevation toward an old waterway channel with the lower density of river gravel indicating the beginning of alluvial deposits, while gravels associated with an outwash plain remain on the higher, flat elevations.

While the clusters of artifact densities recovered from the SVS units were successful in identifying the locations of subsurface features, the distribution of materials was only somewhat helpful in tentative identification of site structure. Artifact distributions appeared to reveal fairly discrete clusters of artifacts, rather than what is thought to be the typical Oliver Phase circular distribution, found at two recently investigated Oliver Phase sites, Cox's Woods (Redmond and McCullough 1996) and Clampitt (Redmond 1994b). These sites had midden deposits with high densities of pottery sherds and faunal remains in a circular pattern, and, at both sites, the higher frequencies of pottery ended abruptly near the outer perimeter of the village, immediately outside the location of stockade walls, and decreased toward the center of the village.

The distributions of prehistoric artifacts from the SVS units (see Figures 4.5, 4.6, and 4.7) roughly coincide with each other. The density areas for lithic distribution show wider variation than do the ceramic distributions, but areas of higher lithic density do overlap with the same clusters in the ceramic distribution. The recovery of Archaic points indicates that the lithics distribution cannot be attributed entirely to the Oliver Phase occupation, and wider distributional variation would be expected. The faunal distribution shows only one area of high density, which is also associated with areas of higher frequencies of lithic and ceramic artifacts. The scatter of historic artifacts (see Figure 4.8 and and Table 4.2), most of which are recent, appears to be associated with the recently abandoned campground. Subsequent investigations in the western portion of the site revealed that clusters of high artifact density were associated with subsurface features. For examples, a feature with an extremely high density of faunal material was close to the faunal cluster identified at about N200, E195 (see Figure 4.7), and the SVS unit with the highest density of ceramic material (count, 48) was immediately adjacent to Feature 1.

The SVS units, however, were not very accurate indicators of the overall site structure, even though they were able to detect some clusters of material. As mentioned, after the excavation at 12 Jo 289 was completed, the eastern half of the site was plowed, making it possible to test the validity of the SVS distributions. Twelve surface features (Figure 4.10) of varying sizes were identified on the eastern portion of the site. Some of these were fairly small in size and probably represented isolated features. However, most were fairly sizable and had surface scatters of domestic refuse. (For an example, see Plate 4.1a where the people are standing on M8, viewed to the north.) Table 4.2 demonstrates material recovered from each midden stain during a survey when the surface was deeply furrowed and poorly rainwashed. None of the 29 SVS units placed on the eastern half of the site fell within the boundaries of any of the twelve midden stains shown in Figure 4.10. Further, the SVS units did not indicate that there was a roughly oval-to-circular pattern to the midden distributions. Although the distribution of materials does not appear to delineate a stockade wall or tightly nucleated settlement like the ones at the Clampitt site (Redmond 1994b) and Cox's Woods site (Redmond and McCullough 1996), it does appear to be a loose arrangement of middens or features related to groups of domestic structures scattered around a possible central plaza area, rather than the discrete, dispersed features first indicated by the SVS units.