Introduction

This study is a field investigation of the geomorphological and sedimentological setting of prehistoric site 48LA277, located in the valley of Crow Creek. Approximately 10 days were spent in the field during July 1992. The area surrounding the site was geologically mapped using a plane table and alidade, with emphasis on delineating the geomorphological setting. Nine backhoe trenches were excavated in order to expose subsurface sediments. All descriptions of sediments and soils in this report are based solely on field examinations. No attempt was made to conduct laboratory analysis.

General Physiographic and Geological Setting

48LA277 is situated on the northeast side of the valley of Crow Creek (Figure A1), the master perennial drainage in this portion of Laramie County. Crow Creek originates 48 km to the northwest, within the metamorphic and granitic, pre-Cambrian core of the north-trending Laramie Mountains. The mountain front lies 24 km to the west of the site. On leaving the mountains, Crow Creek drains southeasterly across the northwest flank of the Julesburg Basin. In the vicinity of 48LA277, the stream flows southeasterly (155°), with a gradient of 0.00385 (0.22°) and a sinuosity of 1.2. Here, Crow Creek displays the 480-m-long segment of an approximately 25-m-wide, braided channel that lies 218 m due west of the site. A parallel, but now abandoned, former channel of Crow Creek lies 0-75 m to the northeast of the modern channel (Figure A2). This abandoned stream segment is 48 m long and contains pools of standing water. The 0-105-m-wide, "flat" area that lies between the modern and abandoned channels is the floodplain (T0) of the modern stream.

In the vicinity of site 48LA277, the valley of Crow Creek is approximately 460 m wide and 13 m deep. The floodplain (T0) of Crow Creek is bordered by two paired Holocene terraces (T1 and T2) and a paired Pleistocene terrace (P1). The lowest terrace tread (T1) slopes 0.7° to the southwest (240^o) and is ±68 m wide. It is separated from T0 by a steep, ±1.5-m-high riser. The next higher T2 terrace lies 1.6 m above T1, slopes 6^o to the southwest, and averages 35 m in width. A ±5.4-m-high riser, with a slope angle that varies between 9^o and 17^o , separates T2 from P1 (Figure A3). The bulk of site 48LA277 is situated on this riser surface. The Pleistocene terrace tread (P1) slopes ±2.5^o to the southwest (Figures A2-A5) and is underlain by a ±1-m- thick gravel deposit composed of quartz sand and pebble- and cobble-sized clasts (1.25-10 cm long) of granite, quartz, dark green metamorphic rocks, and sandstone. All of these rock fragments are derived from the Laramie Mountains to the west. The Pleistocene gravel deposits unconformably overlie Miocene bedrock, which is exposed in the riser that separates T2 and P1. Miocene bedrock, which underlies much of the surface within the Wyoming portion of the Julesburg Basin (Love and Christiansen 1985), is a clayey, sandy, tuffaceous silt interbedded with fluvial sands.

An investigation concerning soils and geoarchaeology has been conducted in eastern Laramie County in the vicinity of Pine Bluffs (Anderton 1990). The author identified three Holocene terraces (T1-T3). The highest (T3) and the lowest (T1) are classified as fill terraces, while the middle one (T2) is a cut terrace. The three terraces in ascending order stand 1-2 m, 3-4 m, and 5-6 m, above the modern floodplain (T0). At Pine Bluffs, the T3 terrace is underlain by more than 2 m of colluvium that ranges in age from late Pleistocene to late Holocene and contains cultural horizons ranging from Hell Gap (Paleoindian) to Historic. Anderton (1990:104) notes that:
 

Soil development is weak within the sediment fill that is present beneath all three terraces (Anderton 1990:32).

The lowest (T1) of the two Holocene terraces present along Crow Creek might well correlate with the lowest (T1) terrace in the Pine Bluffs area. The same cannot be said with certainty in regard to the T2 terrace in both areas. The T2 terrace and associated sediment at Pine Bluffs formed sometime between 1000 and 200 yrs B.P. The sediment beneath the T2 terrace at Crow Creek yielded five radiocarbon dates ranging between 1470 ± 60 B.P. and 970 ± 70 B.P., thus the ages of T2 terrace sediment in both areas is similar. However, the T2 sediment at Pine Bluffs lacks significant soil development, whereas a well-developed Mollisol is present at 48LA277. In addition, the T2 sediment fill at 48LA277 rests on bedrock, and here the T2 is not a cut terrace tread. The cut T2 surface that is present at Pine Bluffs is lacking at Crow Creek.

The T3 Holocene terrace present at Pine Bluffs is absent within the study area. This situation points out the fact that Holocene-age alluvial terraces cannot always be correlated or found to be equivalent between different areas (Albanese and Wilson 1974). In many cases, the number of Holocene alluvial terraces varies laterally within the same small drainage basin. This latter situation is particularly common in the Powder River basin of Wyoming (Albanese 1991).

In this portion of Francis E. Warren AFB, the bulk of the surface within the valley of Crow Creek has been severely impacted by historic human activity. The railroad tracks and associated corridor, border the study area on the northeast. The railroad tracks are bordered on the southwest by an approximately 10-m-wide firebreak that serves as a trail. Most of the original P1 terrace surface, which borders the railroad on the southwest (Figure A2), has been impacted by blading and other earthmoving activity. Whether this surface disturbance was associated with railroad construction is unknown. However, historic artifacts, identified by Douglas Scott as pre-1900 in age, were encountered ±30 cm beneath the surface in Backhoe Trench 6, which lies 30 m southwest of the railroad tracks (Figure A2; also refer to Figure 1, main text). An abandoned irrigation ditch is incised into the P1 terrace in the area located to the northwest of the site area (Figure A2).

All of the lower geomorphic surfaces within the Crow Creek Valley have been severely impacted by earthmoving machinery. Much of the T2 surface has been removed along a 200-m-long, northwest-trending area, the southeastern edge of which lies 133 m northwest of site 48LA277 (Figure A2). Much of the surface, within a 430-m-long area that lies between Backhoe Trenches 7 and 8 (Figure A2), has been "dug into," with resultant pits; bladed trails superimposed on shallow, ephemeral-stream channels; shallow elongated ditches; mounds of earth; hummocky topography; etc. Some of this disturbance is reflected in the "angular" topographic contours depicted on Figure A2. Past disturbance by blading is also evident along the bottoms of draws that are incised into the riser between the P1 and T2 surfaces. In Trench 7, only 40 cm of Quaternary sediment overlies bedrock. Here much of the original thickness was removed by blading. Surface disturbance on the T1 terrace tread, near site 48LA277, is not as obvious as in the previously cited areas; however, in this case "looks are deceiving." In the vicinity of Backhoe Trench 9, the T1 surface appears relatively smooth and planar, but in Trench 9 historical-period artifacts (metal scraps) were encountered to a depth of 46 cm. Nearby shovel tests dug into the T1 surface, encountered historic-period artifacts as deep as 60 cmbs (Melissa Connor, personal communication 1992). Obviously, there has been pronounced impact to the T1 surface by machinery. 

Site 48LA277 was nearly destroyed by earthmoving activity. As indicated on Figure A4, only a 4.6-6.0-m-wide by 25-m-long "strip" of the original site still remains. An approximately 58-x-58-m excavated pit borders the site on the northwest while a roughly 1-m-deep bladed trail abuts the site on the southeast. It would be interesting to search historical records and determine the reason for the "wholesale" surface impact within the valley. Aerial photos dating to 1948 indicate that the bulk of the disturbance occurred prior to that year. Most of the valley surface is now completely "healed" and covered by vegetation. It appears undisturbed to the casual observer. 

Quaternary Sediments and Soils

Sediments of both Pleistocene and Holocene age are present within the study area; both overlie Miocene bedrock. Information concerning sediments was secured from exposures in backhoe trenches and excavation pits within the site area. Nine backhoe trenches were excavated within the study area (Figures A4 and A5). Three penetrated the riser between the P1 and T2 terraces, five were dug into the T2 terrace tread, and one into the T1 terrace. The total length of all nine trenches was approximately 142 m. There is a relationship between sediment type and geomorphic position. For example, Pleistocene gravels underlie the P1 terrace surface, while colluvium/slope wash sediment (CS) covers the surface of the riser (step) between the P1 and T2 terraces. For this reason, the following discussion will focus on landforms and associated sediments and soils.

Sediment Present On The P1 Terrace

The P1 terrace tread is underlain by both Pleistocene gravels and overlying eolian sediment of Holocene age. These strata have been wholly or partially removed along much of the P1 tread as a result of machine blading. The eolian sediment contains buried prehistoric lithic artifacts, and where the eolian sediment has been totally or partially removed by historic activity, buried prehistoric resources have undoubtedly been affected. Many of the lithic artifacts encountered on the surface during cultural resource surveys may have been the "residue" from past earth-removal operations. 

Pleistocene Gravels. This stratum is approximately 1 m thick. It is composed of coarse, arkosic sand and rounded, pebble- to cobble-sized clasts of granite, white quartz, dark green metamorphic rocks and sandstone. This alluvial sediment, transported by Crow Creek during the Pleistocene, is derived from the Laramie Mountains located 30 km to the west. The gravels unconformably overlie beveled Miocene bedrock. 

Eolian Sediment (Holocene). This ±90-cm-thick eolian deposit also occurs on the P1 terrace, atop the Pleistocene gravels. It consists of 2 subunits, an upper 16-cm-thick interval that formed in historic time and unconformably overlies a lower 55-cm-thick subunit that at Stadia Station 89 (Figure A4) contains prehistoric debitage within the upper 20 cm. The upper (historic) subunit at Stadia Station 14 (Figure A4) consists of 6 cm of light brown, very-fine to coarse-grained, silty, clayey sand with a few pebbles (0.5-1.0 cm long) and an underlying 10 cm of similar sand that contains ±40 percent coal fragments that range in length from 0.2 to 1.0 cm. At Stadia Station 89, the upper "historic" subunit is a very fine to medium-grained silty sand with ±10 percent dark gray coal slag and cinder particles (0.3-0.5 cm long) and less than 1 percent coal fragments. The slag and coal fragments at both locales were obviously derived from the coal-burning trains that once traveled down the adjacent railroad tracks. The underlying 53-cm- thick, prehistoric portion of the eolian stratum is a massive, light brown, predominantly very-fine to fine-grained, well-sorted, silty sand that locally contains up to 10 percent medium-sized sand grains. The upper 36 cm of this interval displays a blocky pedogenic structure.

Sediment Present on the P1-T2 Riser

The 5.4-m-high riser (step) between the P1 and T2 terraces is characterized by the presence of short (45-75 m), first-order, southwest-draining, ephemeral-stream draws and intervening parallel ridges ("spurs"), the crestal axes of which slope 9-16^o to the southeast. The spacing between draws and ridges varies between 50 and 100 m. In the area that lies northwest of site 48LA277, much of the riser has been affected by earth removal by machinery. This is particularly true in regard to the bottoms of draws, most of which have been "bladed away" (Figures A4-A6).

Three backhoe trenches (2, 6, and 7) were dug into the P2-T2 riser. Trench 2 was located on a spur, and Trench 6 in the upper portion of a draw (Figures A4 and A5). Trenches 2 and 7 lie at the base of a draw (Figures A4 and A5). Trenches 2 and 7 turned out to be located in areas in which essentially all of the Quaternary sediment had been removed by machinery, as discussed in the "Decriptions of Sediments and Soils" section. 

Exposures in Trench 6 and excavation pits at site 48LA277 indicate that two main types of sediment are present on the riser-ephemeral-stream sediments in the bottoms of draws and CS on the crests and slopes of ridges.

The CS averages 1 m in thickness, is typically a massive, very poorly sorted, fine- to coarse-grained sand with an estimated 25-50 percent silt/clay content and 5-40 percent pebble- and cobble-sized clasts that are derived from the higher Pleistocene gravels. The percentage of pebbles and cobbles decreases in a downslope direction. The upper one-quarter to one-third of the CS displays very little soil development and is classified as an Entisol. The lower portion of the CS is overlain by a Mollisol that displays a prominent, grayish brown A horizon underlain by a Bk horizon that in places is also superimposed on the bedrock that underlies CS. The Mollisol thins in an upslope direction and fades out at the top of the P1-T2 riser. It resembles Mollisols present 60 km to the east, in the Pine Bluffs area, that are classified as Haplustolls by the Soil Conservation Service (Stevenson et al. 1983).

The second type of sediment present on the riser (ephemeral-stream sediments) was well exposed in Backhoe Trench 6, which was located approximately 120 m northwest of the main excavation area at site 48LA277. It lies at the bottom on a 3.7-m-deep, first-order, southwest-trending (260^o) draw that drains across the riser that separates the P1 and T2 surfaces (Figure A4). The trench is approximately 1.4 m deep and 20 m long. The bulk of the exposed sediment is sandy and was deposited by shallow ephemeral streams that drained the bottom of the draw, which is now covered by grass. The sediment was deposited in both historic (upper 60 cm) and prehistoric time (lower 80 cm). The upslope portion of the historical-period sediment is colluvium which changes to ephemeral-stream sediment in a downslope direction (Figure A7). Artifacts recovered from the historical-period sediment include complete and fragmentary bottles, tin cans, and a corset stay. A diligent search would undoubtedly turn up a much more varied collection of artifacts. The artifacts predate the year 1900 (Douglas Scott, personal communication 1992). In addition, this sediment contains coal and cinder fragments derived from trains that traversed the railroad tracks that lie approximately 30 m to the northeast of the trench.

The historic-age sediment in Trench 6 lacks significant soil development. A Mollisol is superimposed on the prehistoric sediment. This soil consists of cumulic A horizon (90+ cm thick) that overlies a Bk horizon. Collectively, the latter two pedogenic horizons are correlated with the aforementioned Mollisol (Haplustoll) developed on the CS that underlies ridges.

Sediment Present Beneath the T2 Terrace

Within the area covered by Figures A4-A6, four backhoe trenches (1, 3-5) penetrated the T2 terrace tread. With one exception-the occurrence of braided stream sediment in the southwestern portion of Trench 1-all of the Quaternary sediment observed in the backhoe trenches consisted of CS. This accumulation is part of the same, previously described sandy CS sequence that underlies site 48LA277 and other portions of the P1-T2 riser.

The same soil development (Entisol and Mollisol) present beneath the P1-T2 riser is also present beneath the T2 terrace in trenches 1, 3, 4, and 5 (for details concerning sediment and soils in these trenches, see the "Descriptions of Sediments and Soils" section).

Backhoe Trench 1 was the longest (43 m) excavated trench; its northeastern portion penetrated the southwest end of site 48LA277. It will be discussed more fully later in this paper. Surface disturbance by machinery was severe in the vicinity of Backhoe Trench 5, and most of the original Quaternary sediment has been removed. The ground surface has also been impacted in the vicinity of Trenches 3 and 4, but much of the original Quaternary sediment still remains.

The sandy, pebbly CS sequence present beneath the T2 surface near site 48LA277 grades to a sandy clay in the area northwest of the site. Here, Backhoe Trench 8 was excavated into the T2 terrace tread at a point 655 m northwest of the site (Figure 2). This "distant" trench was 7.3 m long and had a maximum depth of 210 cm. It contained approximately 175 cm of CS that consists of an upper ±1.2 m thick, sandy, silty, clay with a pebble content of less than 10 percent and a basal sand that thickens from 10 to 113 cm in an upslope direction (Figure A8). This latter situation indicates that the basal sand was originally a "thick" unit and later partially removed by erosion. The soil development displayed in Backhoe Trench 8 is also different from that present in the vicinity of site 48LA277 in that the CS sequence is overlain by an argillic Mollisol. The descending soil horizons are Abt, Btk, Bk and 2Bk. The pedogenic CaCO₃ development in the Bk horizon reaches Stages II and III. There is a sharp contact between the noncalcareous, Abt surface horizon (see below). The argillic Mollisol in Backhoe Trench 8 would be classified as an Argiustoll, in contrast to the Mollisol in the site area that lacks pedogenic clay and is classified as an Haplustoll.

As previously mentioned, an alluvial braided-stream sequence is present in the 43-m-long Backhoe Trench 1. This fluvial sequence overlies the Mollisol that was developed on CS as shown on Figure A9. The fluvial sequence interfingers to the northeast with the upper part of the CS sequence that overlies the Mollisol.

The former sedimentary interval displays an Entisol soil development. Much of the Entisol in the northeastern portion of Trench 1 has been removed by machinery. The alluvial sequence near the southwestern portion of Trench 1 (Stadia Station 38A) is 133 cm thick and consists mainly of thin interbedded lenses of sand and gravel. The percentage of gravel decreases to the northeast, and 7m away, in the vicinity of stadia station 39B, the sequence grades to a 5.4-m-wide zone of interbedded, weak Fluvent soils (Ak horizons) and thin sand lenses (Figure A9). In turn, this latter fluvial sequence interfingers to the northeast with CS, and is absent 4.4 m away. 

The aforementioned 5.4-m-wide zone of interbedded Fluvent soils and sand lenses marks the edge of a broad, braided-stream complex composed of shallow, anastomosing (vein-like) stream channels. The edge of this ancestral braided-stream channel complex of Crow Creek remained relatively stable for an appreciable period of time.

The bulk of the cultural resources at site 48LA277 lie within the sandy CS sequence that is superimposed by the Mollisol and overlain by the braided-stream sequence. Therefore the bulk of the braided-stream sequence was deposited after and possibly during the waning stages of the time of the main human occupation. At Stadia Station 38A in Trench 1, the sediment that comprises the basal Mollisol displays a change from the typical, poorly sorted CS to a laminated sand, suggestive of fluvial overbank deposition. This sediment is finer grained and less sandy than the overlying braided-stream sediment; however, its sedimentary structure is suggestive of possible proximity to an ancestral stream channel. Based on the evidence in Trench 1, one can postulate that ancestral Crow Creek was located much closer to the site area during the time of human occupation than at the present time. At the time of occupation approximately 1,500 years ago, the creek could have been located as close as 50 m to the southeast. Later, at some unspecified time, probably postoccupation, Crow Creek was located only ±25 m to the southeast of the main site area, whereas today it lies 210 m away. There is a problem of trying to squeeze in numerous geological events into a short time span. This will be addressed more fully later in this paper.

Sediment Present Beneath the T1 Terrace

Only one backhoe trench, Trench 9, was excavated into the T1 surface. This 5.7-m-long, ±1.1 m deep trench contained 30-46 cm of "modern" sediment that contained artifacts plus over 70 cm of underlying braided-stream deposits that are predominantly lenses of sand and gravel. One thin (±8 cm) Fluvent soil is present within the alluvial sequence (Figure A10 and soil description below). As mentioned in the preceding section concerning geomorphology, the T1 surface appears to be undisturbed at first glance, but is actually an artificial surface underlain by up to 60 cm of historical-period fill.

No dates are available from the alluvium present beneath the terrace surface, but based on geomorphic position and soil development it should be younger than the sediment present in pits excavated at site 48LA277.

Geological Setting at Site 48LA277

The main 1991-1992 excavation block at site 48LA277 lies on the P2-T2 riser, which at the site slopes ±10° to the southwest while the lower and contiguous T2 surface has an inclination of ±4°. In the site area, the same CS depositional unit underlies both geomorphic surfaces. The cultural features at site 48LA277 are encased within CS, which is approximately 1 m thick and rests on bedrock. Here, CS is a massive, very poorly sorted, fine- to coarse-grained sand with a 25-50 percent silt/clay content plus pebble- and cobble-sized clasts derived from the higher Pleistocene gravels. The CS sequence at site 48LA277 has been highly affected by rodent bioturbation, a process that is still continuing. Krotovina on excavation pit walls are 18-25 cm wide and 25-55 cm long. Their inclination varies from horizontal to near vertical. They can constitute up to 30 percent of the area displayed on the wall of a 1-m² pit. The intensity of bioturbation decreases away from the main site, and in areas only 15 m distant, bioturbation is minimal. One is tempted to postulate that the presence of cultural features, particularly hearths, attracts local rodents.

Four soil horizons that parallel the inclination of the modern surface can be recognized within the CS accumulation at the site. In descending order they are C, Ck, Ak, and Bk horizons. The C horizon (Unit 1) is a modern, loose, poorly sorted, surface sand that is 2.5-11 cm thick. The underlying weak Bk horizon (Unit 2) ranges between 5 and 30 cm in thickness and consists of sand with a high percentage of rounded pebbles. The A horizon (Unit 3) is grayish brown and is the most conspicuous of the horizons. It is a cumulic soil horizon that ranges between 33 and 76 cm in thickness. The basal Bk horizon (Unit 4) is 23-53 cm thick and displays Stage I CaCO₃. It consists of sand with local concentrations of pebbles (detailed lithologic descriptions of sediments begin on p. 75). Unit 4 overlies Miocene bedrock (Unit 5). A calcareous pedogenic horizon (Rbk) is also superimposed on bedrock in the site area. Due to pedogenesis and intense bioturbation, the contact between Unit 4 and bedrock is not always distinct.

Unit 2 (Ck horizon) comprises the Entisol mentioned earlier in this paper, while combined Units 3 and 4 (Ak and Bk horizons) constitute the previously cited Mollisol present on the P1-T2 riser. The Mollisol's physical characteristics resemble soils that have been classified as Haplustolls in the eastern portion of Laramie County near Pine Bluffs (Stevenson et al. 1983). 

Backhoe Trench 1 adjoins site 48LA277 on the southwest (Figure A3). The northeast end of the trench coincides with the contact between the P1-T2 riser and the T2 terrace tread. Buried cultural Features 9 and 10 were located at the northeast end of Trench 1 and mark the southwesternmost, subsurface portion of site 48LA277. Feature 10, dated 1160 ± 90 B.P. (Beta-56859), lies at the northeast end of the trench adjacent to Station 45, while Feature 9, dated 1470 ± 60 B.P., lies 6.7 m from the northeast end. Units 1 and 2 (C and Ak horizons) can be recognized in the northeasternmost 2 m of the trench; however, southwest of this point for a distance of 10 m, all of Units 1 and 2, along with portions of Unit 3 have been removed by machine blading. In the portions of Trench 1 where Unit 3 is relatively undisturbed, it is relatively uniform in thickness (73-80 cm). Within the trench, basal Unit 4 (B horizon) varies between 55 and 73 cm in thickness. Unit 4 "fades out" at a point located 21 m from the northeast end of Trench 1, where it is superimposed by Unit 3 (A horizon), which is present throughout the entire trench.

In Trench 1, the percentage of pebbles and cobbles within CS decreases to the southwest. This situation is depicted in Table A1, which tabulates the estimated percentage of clasts (> 0.5 cm) within Units 3 and 4. It is apparent that there is a rapid "drop off" in percentage of coarse clasts at a point located ±10 m from the northeast end of the trench. Southwest of this point, the total estimated percentage of clasts within CS averages ±5 percent.

In their 1987 description of site 48LA277, Tetra Tech describes six strata within the sedimentary section. In descending order they are labeled I, II (feature fill), IV, V, VI, and VII. Stratum I may correlate with Unit 2 of this report, Stratum IV with Unit 3, and Strata VI and VII with Units 4 and 5. 

Ten cultural features have been recognized at the site. Cultural Features 1-5 (principally hearths) appear to lie within Stratum IV, while Feature 6 lies within Stratum VI (Connor 1993). Feature 3 was dated at 1390 ± 150 B.P. (Beta-49836). Bulk radiocarbon dates of 1150 ± 80 B.P. (Beta-10722) and 970 ± 70 B.P. (Beta-10723) were secured from soil lying between Features 1 and 2. Feature 10, located within the upper part of the Mollisol (Unit 3), yielded a date of 1160 ± 90 B.P. (Beta-56859), while Feature 9, situated at the top of Unit 4 (Bk horizon), yielded a date of 1470 ± 60 B.P. (Beta-56858).

Based solely on radiocarbon dates, one would assume that the cumulic Mollisol present at 48LA277 began forming prior to 970 B.P. A puzzling aspect of this time frame is that Late Archaic-style corner-notched projectile points were also secured from the site, including one from the fill in Feature 3 (Connor 1993:14). This style artifact is usually assumed to predate 1500 B.P (Frison 1991:34). This problem will be discussed more fully in the following section.