Summary of Recent Fieldwork

at the Hudson-Meng Site

L. C. Todd and D. J. Rapson

Click here for a  summary of our interpretations based on these data

| Introduction | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1998 | References |

Introduction

Since 1991 Drs. Larry Todd (lctodd@lamar.colostate.edu) and Dave Rapson (drapson@uwyo.edu) have been working at the Hudson-Meng site.  The work has been done with the cooperation and support of the the Nebraska National Forest. Over the last several years, Forest Archaeologist Doug Stephens  has been a key figure in the research and development at the site. Hudson-Meng is located in northwestern Nebraska in within walking distance of both the Pine Ridge escarpment and the White River Badlands.

This page  provides a very brief review of the previous 7 years field work. In general, each discrete excavation area was given a separate locality designationhmallex.jpg (89523 bytes).  More detailed presentations can be found in each of the annual reports on file with the Nebraska National Forest, Chadron, NE (Todd and Rapson 1991, 1992, 1995, 1996; Todd et al. 1994; Todd et al. 1997). There was no archaeological field work in 1997 since the bonebed enclosure was being constructed in that year.

In describing the locations of items from the   excavations, we often use two grid system designations: a metric grid based on the site datum coordinates and an alphanumeric system that defines 5x5 meter blocks subdivided into 1x1 meter units. This block-unit systems also serves as the basis for assigning individual item numbers to all excavated materials.

This two-grid system was begun in 1991. The first step was to relocate the Agenbroad (Agenbroad 1978) site datum and several of the old excavation walls and corners to ensure that the 1990s grid was as close as possible to the previous site coordinate systems. After reestablishing the original grid, the new block grid system of unit identification was imposed over the site area. As with the earlier grid system, the 1991 grid is aligned to magnetic north. The designation of site datum was changed from 0/0, which was used in the 1970s, to 1000.00 North/1000.00 East and assigned an arbitrary elevation of 100.00 m. There were several reasons for changing the datum designation. First, grids based on 0/0 coordinates result in potential confusion and error for computerized data entry of north versus south (which must be entered as a negative coordinate) and east versus west (also a negative coordinate). Second, using a different grid designation system allows easy differentiation between materials, and notes deriving from the present and the earlier excavations.

The block grid system hmgrid.jpg (46778 bytes)used at Hudson-Meng has been developed and refined during several recent archaeological excavations (Hofman et al. 1991; Rapson 1990; Todd and Rapson 1988). Under this system, each meter square at the Hudson-Meng site has been assigned a unique alphanumeric designation based on its location within a series of five-by-five meter blocks. Each meter unit within the five-by-five blocks is assigned a number with "1" at the northwest corner and meter unit "25" at the southeast . The lines of demarcation between each block are on 5 meter grid increments (i.e., 900, 905, 910, etc.). For example, the unit designation for the square whose southwest corner is at N960/E986 (would be "M82-6." This unit designation system can incorporate excavations in all potential areas of the site (i.e., all adjacent portions of the drainage system in where deposits of similar age may be found). Every item from a square is assigned a sequential number.

In the remainder of this page,  items and excavation areas will be referred to by both their metric and block-unit coordinates. Although initially somewhat confusing, by referring to the grid system map, it soon becomes fairly easy to determine which areas of the site are being discussed. Also, as reviewed in Section I, we also refer to each distinct excavation area by a descriptive Locality name, which helps to keep area distinctions clear.

1991

The 1991 field season concentrated on excavations of a north-south trench at the southwestern margins of the main bonebed as exposed in the 1970s excavations (Todd and Rapson 1991). This locality, referred to as the South Trench exposed a strip of bonebed that ranged from very sparse, heavily weathered, scattered and disarticulated bones at the southern end to a dense concentration of better preserved, often articulated bones in the north . Other than small pieces, which were mapped and removed, all bones from the 1991 excavations were recorded, covered in black landscaping fabric and left in situ. hmsouth.jpg (21448 bytes)At the end of the 1991 season the South Trench area was backfilled to the pre-excavation surface level. All crews during the 1991 field season were volunteers.

Additional limited excavations of more recent deposits (STRAT H), immediately below the modern ground surface were undertaken in a second locality: the North Block. This excavation area was not taken to the bonebed level elevation.  3492 mapped items were recorded during the 1991 season from both the North Block (Units M88-1, M88-2, M89-2, M89-3, M89-4, M89-5, M89-6, and M89-11) and South Trench localities (M80-6, M80-7, M80-8, M80-9, M80-10, M81-4, M81-6, M81-7, M81-8, M81-9, M81-10, M82-6, M82-7, M82-8, M82-13, M82-18).

Most of these documented materials were from the South Trench   with bison bones being the most commonly documented material class (N=853 or 26.3% of all mapped items). Chipped stone items, which were mostly small fragments of unmodified debitage were rare (N=19 or 0.6% of the mapped items). Of these chipped stone artifacts, 84.2% were found in the STRAT A deposits (above the main bonebed level).

The bonebed elevation, as defined by the mean elevations of all bison bones in STRAT B was highest at the southern end and sloped to the north, which was about 25 cm lower.  Within each meter square, bison bones are at a fairly uniform elevation, with standard deviations of bone elevations consistently being less than 5 cm.

For the five excavation units that had more than a single piece of chipped stone recorded, the relationship of average chipped stone elevations to average bison bone elevations are summarized in Table 1. hmtab1.jpg (16361 bytes)In order to avoid the possibility of artificially subdividing the materials based on the rather subjective distinction between STRAT A and STRAT B, all chipped stone and bison bones, regardless of STRAT designation were included in the calculation of values given in Table 1. Although the numbers of lithic artifacts is low, there is an unmistakable pattern of stone tool elevations being constantly higher than average bison bone elevations.

Based on the 1991 excavations, a preliminary formational model was developed that suggest that much of the variation in skeletal element frequencies, bone condition, and disarticulation and scattering could be a result of differential preservational histories across the bonebed, rather than being the unambiguous indications of human butchering/processing activities (Todd and Rapson 1991).

1992

In order to begin evaluation of the formational model developed after the 1991 field season, an excavation area was selected for excavation in 1992 that, since it was near the margins of the main bonebed concentration, were predicted to exhibit preservational attributes similar to the dispersed, scattered bone found at the southern end of the 1991 South Trench excavation area. The Southeast Block excavation was placed at the southeastern corner of the main bonebed area as previously defined during Agenbroad’s excavations. Although the bone was indeed sparse and heavily weathered, there were also a relatively high number of large cranial fragments that seemed to represent in situ destruction of skulls rather than intentional processing by humans. A number of the bones exhibit clear indications of having been modified by carnivores.

As with the 1991 excavations, the Southeast Block began at the edges of the previously excavated bonebed and extended to the south. hmse.jpg (107181 bytes)Excavations were conducted primarily with volunteer crews along with two students who received Field School credit from Chadron State College. Additional limited excavations were conducted in 1992 in a locality referred to as the East Bank. A permanent grid marker was placed near the East Bank excavation area at N875.000 E1110.000 Z107.658. The East Bank excavations were restricted only the late Holocene deposits and did not uncover any material suspect to be of the same age as the bonebed, although a layer of small bone fragments deeper in the cutbank is still of interest and should be investigated further in the future.

The materials recorded in the Southeast Block excavation area are listed by STRAT  and material class types. Bone fragments (N=545) and bison bones (N=467) were the most commonly reported materials. As in 1991, chipped stone artifacts were found in very low numbers (N=14 or 0.6% of the total mapped items).

The bonebed elevations in the Southeast Block are similar to the elevations at the northern end of the South Trench as uncovered in 1991. Again, the standard deviations of bison bone elevations are low.

Comparison of average chipped stone and bison bone elevations   show the same pattern as seen in 1991 – within a given meter square, the average elevation of the lithic artifacts is always higher than that of the bison bones. Other than small pieces, which were mapped and removed, and several bones removed for specialized analysis, all bones from the 1992 excavations were recorded, covered in black landscaping fabric and left in situ. The Southeast Block area was then backfilled to re-excavation surface level.

In general, the results of the 1992 excavations (Todd and Rapson 1992) tended to support the expectations of the basic formational model developed in 1991 with additional evidence provided to indicate that the skulls thought to be missing from the deposits during the 1970s excavations are in fact the result of differential preservation rather than human transport or processing decisions.

1993

In 1993, the goals of excavations at Hudson-Meng changed considerably. During the 1991-1992 seasons work was directed primarily at gaining a better idea of the issues involved in beginning research to come to a better understanding of the specific taphonomic formational history of the bonebed. The 1993 season, which began with the mechanical stripping of late Holocene sediments in two large trenches to the west and southwest of the main bonebed ( West Trench and New Trench respectively). These trenches were opened primarily to: 1) document the relationship of the bonebed paleosol to the bedrock in the areas to the west of the site that the original excavations had suggested was the primary kill location; 2) define the margins of the bonebed to allow planning of the bonebed enclosure to proceed; and 3) to collect a series of samples for paleoecological studies (Todd et al. 1995). Rather than concentrating on contiguous block excavations, the initial 1993 excavation units were laid out at 5 meter intervals along the length of each of the trenches.

As illustrated here, hmprof.jpg (28659 bytes)although a steep bedrock cliff was exposed at the western end of the West Trench, this potentially lethal drop was mantled with a distinct, well defined gently sloping soil surface at the time the bonebed was formed. This ‘bonebed paleosol’ was traced physically from excavation unit to excavation unit along the length of the trench. In addition, a series of AMS dates on charcoal chucks from the paleosol along the length of the trench confirm its age. The dense accumulation of bone that makes up the central bonebed stopped within the first 5-10 meters in each of the trenches, thus adding to a better definition of the edges of the main bonebed. Although scattered bones were found further along the trenches, no other massive bone accumulations were found in either trench. Studies conducted in 1993 included analysis of sediment samples for phytolith analysis, field and laboratory investigations of geology and geomorphology, submission of the first series of charcoal samples for AMS dating, and collection of sediment cubes from one of the burned areas for archaeomagnetic analysis.

Crews in 1993 included archaeological field school students from Colorado State University and the University of Nebraska-Lincoln, a small paid professional crew, and volunteers. The combined efforts of this large, hardy excavation team recorded  more items (N=7034) than had been documented during the first two seasons combined. Most of the materials (Table 2.11), however, came from STRAT A (above the main bonebed level). Since the very little dense bonebed was encountered, only 15% of the mapped items came from STRAT B during 1993.

The frequencies  of bison bones (N=166 or 2.4% of the mapped items) and chipped stone (N=9; 0.1% of mapped items) were both low. Although the bonebed paleosol could be followed along both trenches, only seven excavation units contained STRAT B bison bones. Therefore, the number of units for which we can calculate average bonebed elevations within the paleosol is much lower than the total number of units excavated.

At the end of 1993 excavations, in situ materials were recorded, covered in black landscaping fabric and the excavation units refilled with approximately 1.5 m of sediments. The two large trenches were left open, to be backfilled during the construction phase of the bonebed enclosure building project.

The 1993 excavations added considerably to our understanding of the site setting and to defining the limits of the main bonebed concentration.

1994

By the beginning of the 1994 season, preliminary plans for the building location had been completed, and the excavation area selected (Bone Block) was placed in one of the apparently densest concentrations of bone along what would be the west wall of the proposed enclosure. The southern end of this Bone Block locality is adjacent to the northern edge of the South Trench excavation area that had been excavated in 1991. Bone densities were indeed extremely high in this area, especially in the northwestern portions (Todd and Rapson 1995) where the margins of one of the burned areas described in the 1978 report (Agenbroad 1978). In fact, it took two more field seasons (1995 and 1996) to complete excavation of this area and several of the meter squares (M83-1, M83-2, and M83-3) each had over 2000 mapped items by the time the base of the burned bone was reached. During the 1994 season, however, only the upper surfaces of the bonebed were exposed in the Bone Block with most of the effort going to excavating the 50 cm of intact deposits that had been left in place above the bonebed with over 76% (N=5084) of 6643 items recorded coming from STRAT A. Since the main bonebed level was not completed in most of the units, the total number of bison bones was relatively low (N=438) and unidentified fragments made up over 60% of the materials. Average elevations of the bison bones in STRAT B indicate that the paleosurface sloped gently to the north.

Although only three pieces of chipped stone were recovered, one these was a nearly complete Alberta point uncovered slightly above the main bonebed level at the northern end of the Bone Block. At the end of the season, the in situ bone was covered in black landscaping fabric and the area reburied.

As in 1993, the field team included field school students from CSU and UNL, a small paid crew, and a series of volunteers. This was the first season that the entire excavation area was enclosed in a Weatherport cover, which served to protect the bonebed as it was being exposed. In addition to the main bonebed excavations, an auger testing program was begun in the area around the National Register of Historical Places marker on the eastern side of the stockpond.

1995

Excavations during the 1995 field season concentrated on removal of bone from a two meter wide strip of bone between E985-E987 along the proposed west wall of the bonebed enclosure building (Todd and Rapson 1996). Work continued in the Bone Block locality, as well as a locality composed entirely of bones that had been initially uncovered in the 1970s and reburied (North Bone). Additional limited hmnorth.jpg (142096 bytes)excavation also took place in the extreme southwestern end of the New Trench since, at that time it was still thought the trenches would be backfilled and it was felt that we should take to opportunity to collect additional paleoenvironmental samples from outside of the main bonebed. Crews consisted of CSU field school students, paid crew, and volunteers. Both areas were covered with Weatherport enclosures and a field lab was set up in a third enclosure to the east of the excavation area.

In both the North Bone and Bone Block areas, the number of complete bones, and bones that made up articulated skeletal units   was very high and over 11,000 items (80.1% of all mapped items) were recorded from STRAT B. Several relatively complete carcasses were recorded in both areas and the number of identified bison bones was high (N=5915). Chipped stone made up only 0.12% of the mapped items .

Elevations of the bonebed in the North Bone area (blocks M85, M86, and M87) were constantly lower than bonebed elevations to the south in the Bone Block. The North Bone area represents on of the lowest portions of the paleolandsurface excavated during the 1990s.

Unlike previous years, one of the units (M82-2) in which the average elevations of bones and chipped stone can both be calculated  indicates that both classes of material were found at the same average elevation. In both of the other two units, chipped stone was on the average higher than bison bone.

By the end of the season, bone had been recorded and removed from the entire North Bone area, but the northwestern portions of the Bone Block were not completed and were again covered in landscaping fabric and reburied for the second time. Late in the field season, given the high density of bones and the limited probability that a complete two-meter wide strip to accommodate a continuous foundation for the bonebed enclosure, the building support system was re-designed so that a series of separate support footings   would be used. Thus in planning for the 1996 field season, which was to be the last opportunity to excavate bone in danger of damage as a direct result of construction activities, we were able to target selected areas around the margins of the building where excavations had not previously taken place or where the bone had been uncovered and then reburied. The results of these targeted footings excavations are described  next.

1996

The 1996 investigations at the Hudson-Meng site entailed four general research objectivhmnbn.jpg (215641 bytes)es:

These objectives were designed to meet the immediate needs of. Forest Service project managers concerning the proposed construction of the bonebed enclosure (completed in 1997), while maintaining a viable framework for long term, multidisciplinary research within the project area. The results of the 1996 field work reported here must be viewed as preliminary in nature, rather than as a final synthesis, due to the formationally complex and spatially extensive nature of this deposit, as well as the on-going nature of the analysis program currently being conducted with these materials. Field work was conducted from June 4th until August 2nd, 1996. The field crew, under the supervision of Lawrence Todd and David Rapson, consisted of archaeological field school students from Colorado State University, professional crew members from Colorado State University, the University of Wisconsin-Madison, the University of New Mexico, Southern Methodist University, and the University of Iowa, plus a visiting professional archaeologist from the Russian Academy of Sciences in St. Petersburg (Olga Potapova), as well as student volunteers from Rutgers University, Hunter College, and other volunteers. Also involved were personnel from the U.S,D.A. Forest Service, Nebraska National Forest, including Regional Forest Service Archaeologist Doug Stephens, and Forest Service Passports in Time (PIT) volunteers under the direction of Forest Service Archaeologist Sherri Counce.

The first stage of field work consisted of establishing grid locations for the 11 construction footing areas  in preparation for their excavation. Most of these footing areas had been partially excavated in previous seasons, either during the 1970s (Agenbroad 1978a), or during the 1990s (Rapson and Todd 1992, 1995; Todd and Rapson 1991, 1992, 1995, 1996; Todd et al. 1994). Only footings J I, A3, A5, and A7 were previously unexcavated.

Once grid locations were identified, the overburden was removed with power equipment. Overburden was removed along the West, South, and East center lines of the proposed bonebed enclosure foundation. This included an area of approximately 39 m N-S by 25 m E-W centered on grid coordinates N991/E982 (Northwest Block area), N952/E982 (South Trench area), N952/EI007 (Southeast Block area), and N991/E1007 (Northeast Comer area, note that bone was removed here hmse2.jpg (199513 bytes)during the 1970s excavations). Within this area, a central block of unexcavated sediment located above the proposed bonebed viewing area was left intact. This intact block is centered near N960-N968/E990-E998. Depth of excavation with power equipment varied from only a few centimeters in the Northeast Corner, to a maximum of over 2 m in the South Trench area. The goal of this power stripping operation was to remove sediment to an elevation approximately 40 cm above the bonebed level, and to provide sufficient exposure at this elevation for the placement of three Weatherport portable structures over the excavation areas. This job required I and 1/2 days to complete, including the re-excavation of a 2 by 2 m sump pit for water screening located near the northwest edge of the stock pond. This sump hole has been excavated in the same location since 1991. It is intended to temporarily contain run-off and sediment from water screening activities in a holding pit away from the stock pond.

After completion of overburden removal, the coordinates of each planned footing center was located using an EDM, and the nearest corresponding 2 x 2 m excavation grid intersection point was identified. The 2 x 2 m excavation grid intersection points   reflect the nearest whole or half meter point on the master site grid system corresponding to the center of the planned footing locations. This conversion allowed excavation and hmswall.jpg (69602 bytes)documentation to be conducted within the established Hudson-Meng provenience system,  which employs either whole or partial 1 x 1 m units as the basis for excavation and documentation of all recovered materials.

Excavation was conducted by hand with trowels, in 5 cm arbitrary levels. As in 1991 through 1995 ), excavation levels were standardized throughout the site area, so that level 1 describes the sediment from 102.100 -102.050 m above the daturn plane regardless of where it occurs on the site. As a result, excavators may begin excavation in a given unit at any level (i.e. not necessarily "Level 1), depending on the elevation of the unit at the time hand excavation is begun. Each excavator is responsible for a given meter square, sub-divided into four 50 x 50 cm quadrants (although all four quadrants may not be excavated in each unit). Once the bonebed level is reached, excavators use brushes, tooth picks, and bamboo tools in order to minimize the damage inflicted to fragile bone specimens by metal tools. Plaster bandage jackets lined with paper towels are formed around most identifiable bone specimens before their removal since these "custom-made" containers provide a safe transportation and storage medium for the delicate bone specimens.

Excavations were completed in a total of 86 1 x 1 in units (either complete or partial) during 1996. All excavated sediment was water-screened by unit, level, and 50 x 50 cm quadrant, with the exception of the sediment samples removed from each level. Since 1996 represented the final season of archaeological excavations at Hudson-Meng before the construction activities in the Spring of 1997, it was essential that all archaeological materials be removed from the 10 designated footing areas (plus the unanticipated footing E7, discovered to still contain in situ bone near the end of the planned summer field season). Excavation and clearance of a 2 x 2 m area centered around each of these footings was designed to provide adequate space for footing construction, plus a 25 to 50 cm buffer area, in order to avoid unintended damage to the archaeological materials as a result of these activities. Excavation methods followed the procedures developed over the course of the first five field seasons (Todd and Rapson 1991, 1992, 1995; 1996; Todd et al. 1994). This included the collection of as much information as possible related to the formational history of the deposit in addition to the documentation and removal of all archaeological specimens.

In cases where large bones were securely positioned in unexcavated sidewalls and could not be easily removed, a determination was made as to whether the bone was outside the footing area and within the 25-50 cm buffer zone. Bones in this position (i.e. outside the footing area) were documented, mapped, and left in situ. Each was covered with a small square of landscaping fabric, to facilitate any eventual re-exposure and to indicate clearly to future excavators those specimens exposed and reburied by the 1991-1996 excavations. All fabric pieces were placed in an overlapping fashion over the bones. This technique allows the weight of the fill to be distributed more evenly across the bones, avoiding the creation of "high tension zones" across the tops of exposed items, which can result in damage to the more fragile elements.

Elevation data from all excavation units in the central area of the bonebed have been used to develop a topographic model of the landsurface on which the bison bonebed had been deposited. hmsurf.jpg (151431 bytes) As indicated in the image at the left, the bonebed clearly sloped to the north with the areas that had been excavated as the Southeast Block in 1992, the southern end of the South Trench excavated in 1991, and the South Wall excavated in 1996 being markedly higher than the central portions of the bonebed.   In both the South Trench and the South Wall areas, a clear elevational difference between the main bonebed and an apparent upper level was documented in 1996.  Nearly 100 square meters of this area of the site are included under the bonebed enclosure and will be the focus of additional investigations in the future.

1998

After having had a year's break in archaeological fieldwork while the bonebed enclosure was being constructed, excavations in 1998 were begun inside the building.  Only a 4 square meters of a portion of the bonebed that had been originally excavated in the 1970s and then reburied were uncovered in 1998. This limited area will be used to monitor bone condition before beginning larger scale exposure of the bonebed.   Working inside definitely has its benefits on the rainy days. The 1999 field school will open up approximately another 8 square meters of bone.

In 1998 we also started work on a survey program that for this first year focused on an area called Wolf Butte, which is about 20 miles east of the bonebed enclosure.  A summary of the initial reseach design can be viewed in Adobe Acrobat .pdf format by clicking here.  We anticipate expanding the scope of the survey project in 1999.

 

| AP460 Home | Site Location | 1991-1998 Fieldwork | Field School Overview | Hudson-Meng Photos |

| Student Projects | Grading | Application |