SECTION 2.0

PREHISTORIC SETTLEMENT PATTERNS IN UPLAND SETTINGS:
AN ANALYSIS OF SITE DATA IN WATERSHED D
(CONEMUGH RIVER-BLACKLICK CREEK)

Beverly A. Chiarulli

2.1 Introduction and the Physical Setting

2.2 Research Issues

2.3 The PASS File Database for the Conemaugh-Blacklick Watershed
(assisted by Michaelynn Harle, Jamie Rich, Barbara Malinky, Emily Griffin, Kalynn Yastro, Rebecca Tollman and Eli Long (GIS))

The PASS file database sent to IUP by the BHP lists 341 sites in the Conemaugh-Blacklick watershed (Figure 12). The original study (BHP 1996) listed 241 sites, so it appears that additional sites have been identified in the watershed since 1997. Of these, 50 are single component historic sites and were eliminated from the investigation, leaving a total of 291 prehistoric and unidentified sites. As shown on Table 12, there are a total of 835 archaeological components identified for the 291 sites. Sixty-one (30%) of the sites are identified as single component prehistoric sites, 144 (70%) are listed as multicomponent occupations.

Table 12: Frequency of Chronological Components in the Conemaugh - Blacklick Watershed

However, these numbers include sites that are in non-upland settings. As defined by the BHP (1996), five topographic settings-beaches, floodplains, floodplain rises, terraces, and islands are identified as riverine or floodplain settings (Table 13). The prehistoric sites in the Blacklick-Conemaugh Watershed are located in 12 topographic settings. One site is identified as located on a beach, however, this clearly seems to be a misidentification, since the site (36IN3) is located along the Conemaugh River on the edge of a flood control area.

A total of 153 (53%) of the sites are located in floodplain or riverine settings; 138 sites are located in upland settings. The overall density of sites is 1 site per 2.4 square miles (700 square miles/291 sites). According to BHP (1996), the watershed has a density of 1 site per 1.3 square miles. This discrepancy may result from the elimination of historic sites from our sample. We tried to calculate the percentage of the watershed classified as riverine versus upland, but were not successful. We tried to create a GIS layer that put a 100-foot buffer along the floodplains of all streams and rivers. The layer provided to us by the Conemaugh-Blacklick watershed association did not distinguish between first- and second-order streams, which are upland settings and third- order stream floodplain which are riverine settings. Locating these specific environments proved to be difficult given the scale of maps used in the investigation.

Table 13: Topographic Settings of Sites in the Conemaugh-Blacklick Watershed

One initial step involved an examination of the distribution of sites by USGS quadrangle. A preliminary inspection found that there are significant differences in the distribution of sites in the watershed. This watershed covers parts of 14 USGS quadrangles. Only twelve of the quads contain any recorded sites. Considering both riverine and upland sites, most (68%) are located in the quadrangles at the western edge of the drainage including the Blairsville, Bolivar, Brush Valley, and Derry quadrangles plus the New Florence quadrangle (6%), which is also along the Conemaugh River. There are 122 sites identified in upland settings. Three quadrangles, the Blairsville, Bolivar, and Derry quads, contain 76% of the upland sites. The only other quadrangle with significant numbers of sites in general or upland sites is the Indiana quadrangle, which has 14% of the total sites and 7% of the upland sites. This information is shown on Table 14.

Table 14: Number and Percentage of Sites by USGS Quad In the Conemaugh - Blacklick Watershed

Each of the USGS quads covers an area of approximately 56 square miles. Only the Blairsville quadrangle has a site density approaching one site per 0.5 square miles. For this quad, the density of the total site database is one site per 0.63 square miles.

Even quadrangles with the highest site densities tend to have sites distributed in clusters. The Westmoreland County portion of the Blairsville quadrangle contains 66 recorded sites of the 89 for the complete quadrangle, whereas only 23 sites are in the Indiana County portion of the quadrangle. Only one of the Westmoreland County sites on the quadrangle is located in a floodplain/riverine setting. The other 65 are upland sites and represent the entire upland site sample for the Blairsville quadrangle. It seems unlikely that there are no upland sites in the Indiana County portion of the quadrangle, since the Indiana County section of the Bolivar quadrangle immediately to the east contains 9 of the 14 upland sites found on that quadrangle.

The sites in the Westmoreland County portion of the Blairsville quadrangle are also clustered in their distribution. Thirty of the sites are located along an approximately 5-mile long section of Stony Creek, from the southern edge of the quadrangle to the point where the creek joins the Conemaugh River. A single collector who systematically investigated the drainage identified almost all of these sites.

The types of recorded sites in the watershed was also examined. As a first step in addressing this question, the distribution of site types found in the floodplains was compared to those from the uplands in the Conemaugh - Blacklick watershed (Table 15).

Table 15: Distribution of Site Types for the Total Watershed, Upland and Floodplain Settings

The majority of sites from both the floodplain and upland settings are open habitation sites. Small surface scatters are found in much higher percentage in the floodplain settings than in the uplands, so that open habitations and small scatters are approximately 79% of the total site sample. Significant rare site types such as villages are found in similar proportions. Other site types such as isolated fluted point loci, petroglyphs, rockshelters, and quarries are found only in the uplands.

This preliminary view of the distribution of sites led to the investigation of site distribution through a more powerful approach. Because IUP is located within the watershed, the university Spatial Science Research Center had a number of Geographic Information System map layers available in ARC View format. The BHP provided us with an ARC View shape file of the site locations that could be used as an overlay of the watershed data. The first attempt to match the geographic projections of the two data sets was not entirely successful. We then contacted the Conemaugh Watershed Association, which also has GIS shape files for a number of environmental layers for the watershed. Using that information, a series of maps showing the distribution of archaeological sites in relation to other environmental variables was created. The most useful base maps were those in which the soil and geologic bedrock formations were shown along with the river and tributary locations. The Catawba Path was also added to the base maps and site locations were connected to other variables from the PASS site database such as cultural affiliation, site type, number of recovered diagnostics, and topographic settings.

We also examined the relationship between soil type, bedrock geology, and site distribution. Figure 13 shows the topographic categorization of the sites overlaid on the soil associations. It appears that there is a stronger association between the type of surface geology and site distribution than the association between soil types and site location. As can be seen on Figure 14, which overlays the site topographic settings on the geologic base map, there are two main clusters of upland sites. The first cluster (Upland Site Cluster A) is in the southwestern corner of the drainage between Blairsville and Derry (Figure 15). The second is located at the eastern edge of the drainage between the Conemaugh River and the headwaters of Blacklick Creek just to the east of the crest of Laurel Ridge (Figure 16).

While it is not clear if the cluster on the summit of Laurel Ridge is a result of collector bias or reflective of a prehistoric pattern, Site Cluster A seems to be associated with exploitation of the Loyalhanna chert source. Most of these sites are on the Monongahela Group geologic formation, which is the source of the chert. The types of sites that form this cluster include a quarry site and several village sites (Figure 17). In the discussion in the previous section on lithic sources, we considered the information available from these sites. As shown in Figure 17, the sites associated with the Loyalhanna chert are not only found on the Monongahela Group bedrock, but are also found on the surrounding Casselman Formation.

The cultural affiliation of sites was then examined to determine the kind of settlement pattern data that is available in the watershed and whether these patterns changed through time. Only those sites that contained diagnostic artifacts and could be assigned to a cultural period were included in this analysis. Although it seems as though the site locations are more strongly correlated with surface geology, these are also shown with the soil association base map. As an example of the site location layers, all of the Archaic sites were mapped on a background of either the soil type or geological formation and are shown on Figure 18 and Figure 19.

Only three Paleo-Indian sites are recorded in the watershed. As described above, these are located in both floodplain and upland settings and were identified based on the recovery of a single fluted point. All three of these are similar to the short-term camps identified in the Meyersdale Bypass project. The sites are associated with either the Monongahela Group geologic formation or with the Catawba Path, which suggests that this path represents a very ancient route from the Carolinas to New York (Figure 20).

Early Archaic sites show a similar pattern in location to that of the Paleo-Indian sites. Some of these are associated with the Monongahela Group geologic formation (three sites) and the Catawba Path (two sites), although others are in new locations either along the Conemaugh River or in the uplands along the watershed boundary with the Crooked Creek drainage (Figure 21). The large base camp, 36WM278, located along Stony Run was first occupied during this period, although the other sites are probably short-term camps.

There is a marked increase in the number of sites in the watershed identified as Middle Archaic sites, which is somewhat unusual since many investigators view the Middle Archaic as a missing period (Custer 1996). Thirty-two Middle Archaic sites have been identified. As in earlier periods these are most commonly associated with the Monongahela Group. In addition, a number of sites are found along the facies between the Casselman and Glenshaw Formation, which is also the channel of Blacklick Creek (Figure 22). The 32 sites show no major differences in the number of artifacts recovered. The PASS forms classify sites by the number of diagnostics recovered. From Paleo-Indian through the Middle Archaic, all of the sites are classified as having less than 25 artifacts or diagnostics that are present in an unspecified number. Since the only information for most of these sites is whatever is recorded on the archaeological site survey form, it is difficult to classify sites by occupation type.

In contrast, during the Late Archaic, two sites are identified as containing from 25 to 50 diagnostic artifacts (Figure 23). One of these is 36WM278, investigated in the Stony Run Mine permit area, which can be classified as a base camp. The final phase of the Archaic period, the Terminal Archaic or Transitional Period, shows a decrease in the total number of sites, but an increase in the number of sites along major drainages (Figure 24).

During the Woodland Period, several new site types have been identified in the drainage including a burial mound and several village sites (Figure 25). There is also an increase in the number of sites, although many of these are re-occupations of locations used during the Archaic period. This is in part is due to a pattern in which many locations are occupied throughout the cultural sequence (Figure 26, Figure 27, and Figure 28). Late Woodland sites are often thought to be associated with well-drained soils. Figure 29 shows the association of Late Woodland sites with soil types.

In addition, there are several distinct types which cannot be assigned to a cultural period. Of particular note are two petroglyph sites on Laurel Ridge.

Table of Contents

Top of Page

Single Component Sites Listed in the PASS Files from the Conemaugh-Blacklick Watershed

It is also useful to examine the distribution of different site types by cultural period and topographic setting and to compare the information in the PASS files for riverine and upland sites. Table 16 lists the single component sites along with this information.

Table 16: List of Site Types by Topographic Setting and Cultural Period for Single Component Sites

For the Paleo-Indian period, a single site is recorded on a stream bench. During the Early Archaic, three sites are found in both upland and riverine settings, on stream benches and floodplains. Only open habitations are recorded. The number of sites dramatically increases in the Middle Archaic to 14 upland sites, and 5 riverine sites. All of the upland sites are described as open habitations, although sites are now found in most of the upland environments including a hilltop, hillslopes, stream benches and upland flats. A greater variety of Middle Archaic site types are found on riverine terraces including a village, an open habitation, two small lithic scatters, and an isolated find. Almost all of the Late Archaic sites in both upland and non-upland settings are identified as open habitation sites. Two small surface scatters are listed in non-upland settings. A new site type is identified with a general Woodland designation. A burial mound located on a non-upland terrace setting. Early Woodland occupations are primarily open habitations as well in both upland and non-upland settings. Middle Woodland sites include both open habitations and small scatters and a rock shelter. Only nine Late Woodland single component sites are listed in the database. These include villages, surface scatters, open habitations, and rockshelters in both settings.

Multicomponent Sites Listed in the PASS Files from the Conemaugh-Blacklick Watershed

Tables 17 and 18 list the multicomponent sites from the watershed for upland and non-upland topographic settings, respectively. Thirty sites are recorded in upland settings. Almost all of these are described as containing occupations dated from the Archaic through Late Woodland or Protohistoric periods. None of these are situated in upland saddles, most are on stream benches (n=15), upland flats (n=1), hillslopes (n=10), or hilltops (n=2). The hilltop sites also are the most unexpected settlement types and are identified as a village and rockshelter. Almost all the others are described as open habitation sites. Similar site types are listed for the non-upland multicomponent sites. However, there are three villages in this sample, and 10 small unknown- function lithic scatters.

Table 17: Multicomponent Sites in the Blacklick Drainage Area - Non-Upland Setting

Table 18: Multicomponent Sites in the Blacklick Drainage Area - Upland Settings

For the Paleo-Indian period, a single site is recorded on a stream bench. During the Early Archaic, three sites are found in both upland and riverine settings, on stream benches and floodplains. Only open habitations are recorded. The number of sites dramatically increases in the Middle Archaic to 14 upland sites, and 5 riverine sites. All of the upland sites are described as open habitations, although sites are now found in most of the upland environments including a hilltop, hillslopes, stream benches and upland flats. A greater variety of Middle Archaic site types are found on riverine terraces including a village, an open habitation, two small lithic scatters, and an isolated find. Almost all of the Late Archaic sites in both upland and non-upland settings are identified as open habitation sites. Two small surface scatters are listed in non-upland settings. A new site type is identified with a general Woodland designation. A burial mound located on a non-upland terrace setting. Early Woodland occupations are primarily open habitations as well in both upland and non-upland settings. Middle Woodland sites include both open habitations and small scatters and a rock

Table of Contents

Top of Page

Condition of the Site Database

One assumption in the survey priorities policy is that the sample of sites is adequate for research questions that we may want to investigate now and in the future. This assumption is valid only if the sites remain intact. Table 19 presents a breakdown from the PASS file database of the condition of upland sites in the watershed. This variable describes the current condition of a site. Only 1 site is described as 100% destroyed, 19 are described as from 90-100% undisturbed. Relatively few of the sites have been evaluated, and from examining the CRM reports, it is clear that sites, which have been destroyed as a result of a project, are still listed as in good condition in the files. This is certainly important information to record and to update when new information becomes available. Table 20 lists the types of impacts, which have affected archaeological sites.

Previous Cultural Resource Management Investigations in the Conemaugh-Blacklick Watershed and Adjacent Watersheds (assisted by Rebecca Tollman, Jamie Rich, Emily Griffin, and Michaelynn Harle)

Forty-eight cultural resource investigations have been conducted within or adjacent to the watershed (Table 21). This information was collected by Michaelynn Harle, an IUP Anthropology Student, from the CRM report library at the Bureau for Historic Preservation in Harrisburg. It is not a complete list of all the projects and several significant projects are not listed, including the investigations for the Rt. 119 road widening project and the Senate Coal Company Tipple along Blacklick Creek. This list does provide a representative sample of the kinds of projects that have been conducted and the results of those investigations. Figure 30 shows the locations of these projects and Table 22 summarizes the number of sites found in the surveys. Appendix 2 contains a brief summary of each of these projects.

Table 19: Condition of Sites in the Conemaugh-Blacklick Watershed

Table 20: Causes of Site Destruction Recorded in the PASS Files

Table 21: Summary of Area covered by CRM Projects and Number of Sites Discovered

Forty-eight projects have been conducted. Archaeological sites were found in only 17 of these. The number of sites ranged from 1 to 15 per project. It was not possible to determine the size of a number of the projects. The 27 for which the extent was recorded covered 595 hectares. Ten sites were found in these projects leading to a projection that the average density would be approximately 1 site per 0.59 square kilometers (59.5 hectares) which is not that different from the site estimates projected from the tests of the Crooked Creek Model.

Eight projects were identified in the part of Cambria County that included the Conemaugh - Blacklick watershed, although only six of these are actually in the watershed. Four sites were discovered in two of the projects; of these only one of the sites produced diagnostic artifacts, a Meadowood projectile point, which dates to the Early Woodland period. Additional investigations were recommended for only one of the sites, 36CB142, which was located just outside the watershed in the headwaters of a tributary that flows into the Susquehanna drainage. This site was identified through the recovery of 20 flakes from a single shovel test placed on a hillside bench in a forested area. It was probably a small, undisturbed special-activity site. It was not investigated further, because the site was to be avoided by Pennsylvania Department of Transportation.

Twenty-five projects were located in Indiana County, one project was located in Indiana and Westmoreland counties, and one was located in Indiana and Cambria counties. Archaeological sites were discovered in 10 of the projects. Twenty-two of the projects were Phase I surveys. Only one Phase III data recovery excavation was undertaken in Indiana County. A total of 23 sites were discovered of which 9 were found in one sewer line survey in Burrell Township and 4 were located in the survey for the Indiana Bypass in White, Center, and Armstrong townships. The projects were located in either the western third of the county, in White Township surrounding the town of Indiana, or in the townships located between the Conemaugh River and Blacklick Creek.

Investigations in the report files in the BHP offices in Harrisburg identified ten Phase I archaeological surveys that have been conducted in the Conemaugh-Blacklick watershed in Westmoreland County or in the townships just outside the watershed boundary. In only two of the projects were archaeological sites identified. However, the survey of one project, the Phase I survey of the Stony Run Surface Mine permit area, resulted in the discovery of 15 sites and in important Phase II and III investigations of some of these sites. The only other project in which sites were located was a reconnaissance survey of Loyalhanna Lake, which is located just outside the watershed under investigation.

Investigations of National Register Eligible Archaeological Sites in the Conemaugh-Blacklick Watershed

Sites found in three of the CRM projects conducted in the Conemaugh-Blacklick drainage have been evaluated through Phase II investigations. The projects include the Stony Run Mine investigation, the proposed Indiana Bypass and the investigation of the Blacklick Site during the replacement of the Buena Vista Bridge. Two data recovery investigations were conducted. However, these reports were not reviewed for this investigation.

Table 22: Cultural Resource Projects in the Conemaugh - Blacklick Watershed and Adjacent Areas

The Stony Run Mine Project

The Phase II and Phase III Investigations of four sites in the Stony Run Mine surface mine permit area found that these sites represented a cluster of a base camp and a number of small short-term occupations, probably located in this area to exploit the Loyalhanna chert. Phase II Investigations were conducted at four of the sites in 1988 by NPW Consultants, Inc. The sites are located along the western edge of Chestnut Ridge, which is a noted source of chert including Loyalhanna chert.

The investigations included plowing and disking of the sites, controlled surface collections, and the mechanical stripping of 5-10% of the plow zone from areas where both artifact concentrations and dispersed artifacts were identified. Robert Oshnock originally identified all of these sites and his collections were also re-analyzed as part of the Phase II investigation. Prehistoric cultural features and diagnostic artifacts were found at three of the sites, 36WM277, 36WM278, and 36WM283. Sites 36WM277 and 36WM283 were interpreted as single component occupations dated to the Archaic and Woodland periods. Site 36WM278 appears to be a multicomponent site used throughout the occupation sequence. Site 36WM277 contained two features dated to the Late Archaic and Late Woodland periods. Because the site boundaries extended into a wooded area that had not been disturbed by plowing, the site was identified as potentially eligible for the National Register of Historic Places and additional investigations were conducted of the site. More detailed descriptions of each of the sites follows.

The Durika I Site (36WM277)

This is one of four sites investigated during a Phase II evaluation of a proposed Surface Mine Permit by NPW Consultants, Inc. (1988). The sites were originally recorded by Robert Oshnock, a local avocational archaeologist, and investigated during a Phase I survey by GAI Consultants, Inc (McHugh 1982). The investigation of this site included a controlled surface collection and stripping of from 5-10% of the topsoil to expose the base of the subsoil in a strategy designed to locate pre-European features. The investigations recovered a limited number of lithic artifacts and debitage including 2 Brewerton points, 1 Perkioman Point, and 1 Levanna Point and 78 pieces of debitage. Over 90% of the debitage was identified as Loyalhanna chert. Thirty percent of the flakes were primary flakes. Two features were located during the Phase II investigation. One of these, a bark- lined pit, produced charcoal that was dated to 2445 +/- 85 BC; the second feature produced a Late Woodland date of AD 950+/-35.

The Durika II Site (36WM278)

This is a large multicomponent site that contained occupations dating from the Late Archaic through the Late Woodland (NPW Consultants, Inc.1988). The majority of artifacts suggested that the primary occupation was during the Late Archaic period, which is supported by four radiocarbon dates of 2915+/- 35 BC; 1775 +/- 125 BC; AD 1055+/- 100; 2650 +/- 80 BC; 2490 +/- 70 BC; 1190 +/- 80 BC. Charred hickory nutshell was recovered from 7 features. From one of these features a charred hawthorn seed was also recovered. Two features contained charred bark. The site revealed evidence for the use of Loyalhanna chert for lithic reduction. Seventy-four percent of the recovered lithic debitage was of this chert type. Recovered projectile point types include: Brewerton, Normanskill, Vosburg, Lamoka and Bare Island from the Late Archaic; Perkiomen points from the Terminal Archaic period; Meadowood points from the Early Woodland; Snyders, and Green Points from the Middle Woodland; and Jack's Reef, Levanna, and Madison Points from the Late Woodland; and Madison from the Late Prehistoric. The site was estimated to cover approximately 30,800 square meters in area.

The Durika III Site (36WM282)

This is the smallest of the sites investigated in the proposed Stony Run Surface Mine (NPW Consultants, Inc. 1988). Only two projectile point fragments were recovered along with a small number of pieces of lithic debitage. Eleven of the sixteen recovered artifacts were Loyalhanna chert, indicating that the occupants of this site, like the others in this area, were exploiting the local chert sources.

The Durika Site IV (36WM283)

This site yielded only two diagnostic artifacts indicative of a Late Archaic occupation NPW Consultants, Inc. 1988). One small conical feature did produce a radiocarbon date of AD 1285 +/- 50, suggesting that it was also occupied during the Late Prehistoric Monongahela period.

The Proposed Indiana Bypass

Four sites were found within the project area located in White, Center, and Armstrong townships, Indiana County, on the Ernest, Elderton, Indiana, and Whitesburg USGS topographic quadrangles by Archaeological and Historical Consultants, Inc. Most of the project was located on an undulating upland terrace, although a small section is in the floodplain of Curry Run. The sites included the Anderson Knoll (36IN331), a Mid-Late Archaic site located just outside of the project area. Late Archaic Lamoka points were recovered from the site. The second site, 36IN105, was also outside of the right of way. The third site, 36IN330, produced only debitage. The artifacts recovered at this third site were debitage. The fourth site, the Simpson Site (36IN57), was identified as a Late Archaic through Early Woodland hunting camp. Although all four sites were recommended for further testing, only the Simpson Site and one other were investigated.

The Simpson Site (36IN57) is named after the major drainage in the area, the Simpson Spring. The site is situated in a relatively level upland saddle, adjacent to the Simpson Spring. The site was found to be eligible for listing on the National Register and additional Phase III investigations were conducted. The Simpson Site was investigated through surface collections during both the Phase II and Phase III investigations, and the excavation of 1 meter squares and the removal of the plow zone during the Phase III investigation. During the Phase II Investigation, 306 flaked stone artifacts, including 13 projectile points, were recovered. Three cultural features identified as hearths were also identified during this work. An additional 416 artifacts were recovered during the Phase III project. These included 15 projectile points, 21 non-diagnostic bifaces, and 7 unifacial tools. Although the site was occupied from the Middle Archaic period through the Late Prehistoric, the majority of the diagnostic artifacts dated to the Late Archaic period. A list of the points recovered from the site is included on Table 11.

The data recovery investigations conducted at the Simpson Site were designed to address five research topics. These were developed from a review of the regional literature and focused on questions of the transition from the Late Archaic to the Terminal Archaic (or Transitional) period, since almost all of the recovered diagnostic points dated to these periods. Five research areas were to be investigated at the site and included (AHCS 1987):

1. Refinement of the chronology of the site occupation: This topic was to be investigated through the analysis of recovered projectile points and radiocarbon samples to provide an accurate date range for point types. Most of the points recovered in the Phase II investigation were dated to the Late and Terminal Archaic periods, so the expectation was that features would be found associated with these occupations. While additional points were recovered, charcoal samples produced dates associated with the Late Prehistoric period, so data needed to address this question was not recovered.

2. An investigation of the exploitation strategies of lithic sources, especially from the Late Archaic to the Terminal Archaic periods: Previous research suggested that the pattern of lithic resource procurement changed from the Paleo-Indian to the Late Archaic periods from one in which exotic or non-local raw materials were used to one in which there was increased reliance on local sources. In most of the region this pattern continued into the Terminal Archaic. This was a different pattern than that in the Susquehanna Basin, where the Terminal Archaic (or Transitional) period is associated with the use of non-local high quality jaspers. Twenty percent of the Simpson Site Phase II assemblage was identified as jasper. The Phase III investigation was then proposed to examine changes in the patterns of raw material usage.

3. An analysis of intra-site activity areas: This topic was to be investigated through the spatial analysis of artifacts, raw materials, and features to determine whether horizontal patterning could be identified at the site.

4. Identification of site function was to be based on an analysis of the types and relative numbers of artifacts and features at the site. Data generated from specialized studies like use-wear analysis was also projected to provide insight into the function of the site.

5. The investigation of subsistence strategies was also planned through the analysis of botanical remains from features and through the possible recovery of faunal remains.

The Phase III investigation of the Simpson Site found little data that could be used to address these topics. Most of the recovered projectile points dated only to the Late Archaic period, so there was little opportunity to address questions of the transition to the Terminal Archaic. Only two hearth features were recovered. These dated to very late in the Late Prehistoric period and were associated with few artifacts. The site did contain a large sample of Late Archaic point types and provides some insight into the technology used during that period. The site was interpreted as a seasonally occupied, small open camp, although the site seems to have a high number of projectile points (n=28) in the assemblage (n=722).

Of the six sites evaluated through Phase II investigations in the watershed, only two of the sites were found to be eligible for listing on the National Register (Table 23). All six of these sites were multiple component, unstratified sites. Very few sites in the watershed have been evaluated for eligibility for the National Register. One site has been determined eligible by the SHPO. Five other sites have been recommended as eligible by either a consultant or agency. Of these, three are located in upland areas, including two on hillslope and one in a hilltop setting. Two sites in non-upland sites were also recommended as potentially eligible. Ten sites have been determined to be not eligible, including six in upland settings and four in riverine settings.

Informant Interviews

The small number of sites identified through cultural resource investigations in this watershed and the skewed distribution of sites in which most are located in the southwestern part of the watershed led to the question of how representative the currently recorded sites are of the archaeological record in this watershed. Just prior to the start of this project, the author was contacted by a local avocational archaeologist who was familiar with a number of sites in the Indiana and Westmoreland counties, specifically along Blacklick Creek. This individual, Butch Laney, had provided information to the Carnegie Museum of Natural History since the 1970s. Figure 31 shows the location of sites identified during several days by Mr. Laney. Most of these have not been previously been recorded and some are unique site types. One site is a large boulder on the upland flat above the Blacklick Creek floodplain; the site has a large pestle ground in the top of the stone (Photo 1). No diagnostic artifacts have been found in association with this specialized activity area, but flakes have been observed in the backdirt of a nearby groundhog hole.

Table 23: National Register Eligibility Assessments Recorded in the PASS File Database

A second relatively rare site type is a rockshelter also along the edge of the uplands along Blacklick Creek. Photo 2 provides a view of the rockshelter. Mr. Laney provided information on 30 sites in this small part of the watershed, including some unusual site types in the watershed that have not been recorded in the PASS files. Mr. Laney also provided some observations and anecdotes about the distribution of sites in the watershed that suggest avenues for future research:

He noted that the styles of projectile points from sites north of Homer City are shorter and thicker than those from the south. He claimed that this pattern holds for points that are classified as the same type through the chronologic sequence. He suggested that there was a cultural boundary in the Homer City area that corresponds to this stylistic difference. This hypothesis could be easily tested. It seems to support William Johnson's suggestion that the boundary between the Monongahela influenced groups and the Upper Allegheny Valley groups is between the Kiskimientas River and Crooked Creek.

Mr. Laney also suggested that Blacklick Creek functioned as a pathway to Chestnut Ridge and the divide between the Ohio and Susquehanna drainages. This divide is found in the uplands just west of Ebensburg. According to local folklore, historic tribes followed the Blacklick to lead deposits. There are lead deposits east of the Allegheny Front outside Altoona that were exploited during the American Revolution at Ft. Robideau. These deposits might be reached through this route. Again, this suggests a topic for additional investigation.

Mr. Laney also indicated that large Late Woodland/Late Prehistoric villages are located at regular intervals from one another. He suggests that the distance is approximately 5 miles between villages.

Summary of Information from the PASS Files and CRM Reports

We began this project with the question "Do we know enough about the upland sites and have enough of them recorded that we do not need to continue looking for more in the Conemaugh-Blacklick watershed?" Before trying to answer the question, it is useful to summarize what we have learned about upland sites and site distribution in the Conemaugh-Blacklick watershed and western Pennsylvania, in general. We have found that:

There are sites in the uplands in the Conemaugh-Blacklick watershed and throughout Western Pennsylvania;

These sites vary in size and density and include small scatters of a few non-temporally diagnostic flakes that may represent specialized, though unidentifiable activity loci, multicomponent lithic scatters, petroglyphs and mortars carved in boulders, base camps with large numbers of artifacts and features, and even villages with large numbers of artifacts and features.

Most of the site types and cultural periods found in the uplands are also found in floodplain settings, although several types are only found in one setting or the other. For example, petroglyph sites are only found in the uplands; the only recorded burial mound is on a terrace.

Most of the sites in both floodplain and upland settings are short-term camps or short-term encampments. In both cases, sites are identified as these types through the recovery of a small number of artifacts.

Except in one small part of the watershed, associated with the Monongahela Group geologic formation, only 33% of the Phase I surveys conducted in the watershed in either floodplain or upland areas discovered prehistoric archaeological sites.

This review has demonstrated that upland sites contain features and sufficient numbers of artifacts to enhance our understanding of the prehistory of the area.

Finally, the review of previous surveys found that when systematic surveys are conducted using the same test intervals as those in riverine areas, that the actual upland site densities are much higher than those projected in the survey priorities policy for the high-site-density watersheds.

The most significant upland environments that are not currently listed as an area to be investigated are hilltops and stream benches. Village sites are found in hilltops. Sites that have been determined eligible for the listing on the National Register are found on stream benches.

2.4 Site Significance and Survey Priorities

Table of Contents

Top of Page