At the Intersection of Geology, Archaeology, History, and ArtEngineer Cantonment (1819–1820), Washington County, Nebraska
In the fall of 1819, members of the Stephen Long Expedition constructed two log and limestone buildings along the west bank of the Missouri River north of modern Omaha, Nebraska, to serve as their winter quarters (Engineer Cantonment). A detailed watercolor and sketches of the site by expedition member Titian Ramsay Peale were key to the rediscovery of Engineer Cantonment. Subsequent excavations of the site identified three distinct archaeological components: two Native American hearths that date to the late Holocene, about 870 and 810 14C years before present (BP), the Engineer Cantonment from AD 1819–20, and a farmstead from the late 19th to mid-20th century. The site is a low terrace-fan complex composed of four major depositional units: a late Holocene Missouri River alluvial fill, a late Holocene through modern alluvial fan, a younger historic alluvial fill, and 19th-to 20th-century Missouri River floodplain deposits. The late Holocene fill is a fining-upward sequence of channel sand and gravel overlain by silt and clay overbank deposits with redoximorphic features. The fan includes multiple beds of yellowish brown, brown, and gray alluvial fan sediments with thin, weakly developed buried soils. The historic alluvial fill truncates the distal portions of the alluvial fan and the upper portions of the late Holocene fill. The modern floodplain deposits consist of a thin bed of silty clay overlying cross-bedded silty sand. Our study demonstrates two things: the significance of the margins of the valley floor of the Missouri River as a record of late Pleistocene through modern fluvial activity, and a potentially long record of human occupation of the valley during the Holocene. This project also demonstrates the potential usefulness of 19th-century art in the study of cultural and environmental change during the last two centuries.
alluvial fan, Engineer Cantonment, late Holocene, Long Expedition, Missouri River, Nebraska, Titian Ramsay Peale
On September 19, 1819, members of a western exploration party led by Major Stephen H. Long reached a point along the Missouri River approximately 8 km north of present-day Omaha, Nebraska (Fig. 1). The exploration party, commonly known as the Long Expedition, was a scientific branch of the larger “Yellowstone Expedition” (Goodwin 1917; Nichols 1971). The site provided timber, protection from north winds, and what was described as a “sharp bend in the river bank. . . . There, a sort of cove in the river offered protection for the [steamship] Western Engineer” (Nichols and Halley 1980, 101). Soon after landing, the expedition party began construction of cabins that would serve as their winter quarters, named Engineer Cantonment (James 1823). On June 6, 1820, the expedition party left Engineer Cantonment to begin the overland portion of their journey west (James 1823). The abandoned structures subsequently fell to ruins, possibly dismantled by soldiers working on the construction of nearby Fort Atkinson (NSHS 2004; Bozell et al., forthcoming), and the remains of the structures were buried by sediments eroded from the adjacent uplands. Engineer Cantonment was lost to history for the next 182 years (NSHS 2004; Bozell et al., forthcoming).
During late winter 2003, Nebraska State Historical Society (NSHS) archaeologists made a preliminary identification of the site location by matching the modern bluff line with that shown in a watercolor painting of the site by expedition member Titian Ramsay Peale, dated February 1820 (Fig. 2A). Subsequent NSHS excavations uncovered remnants of the limestone foundation of a building, including a large central fireplace referenced [End Page 1]
[End Page 2]
in the party members’ journals. The artifacts recovered included buttons from military uniforms and civilian clothing, knives, lead shot, musket shot, gunflints, portions of fine chinaware diagnostic of the early 19th century, and a variety of everyday personal items such as razors, pipes, and eating utensils (NSHS 2004; Bozell et al., forthcoming). Based on the similarities between the painting and the modern landscape (Fig. 3), and the assemblage of age-diagnostic artifacts, NSHS confirmed the site as Engineer Cantonment (NSHS 2004).
The site is important for a number of reasons. The landing is historically significant because it is the oldest Euroamerican archaeological site discovered in Nebraska (NSHS 2004; Bozell et al., forthcoming). Additionally, the expeditionary party traveled by steamship, thus marking the farthest west any steamship had reached on the Missouri River at the time. The site is believed to be the first, and perhaps only, example of 19th-century American government-sponsored exploration in which trained, field-experienced scientists gathered data within a single area over an extended period of time (Bozell et al., forthcoming). The extended stay by trained scientists yielded new ethnographic, biological, and environmental information that is still of value today (e.g., Evans 1997; Genoways and Ratcliffe 2008; Bozell et al., forthcoming). The site also contains two deeply buried hearths that demonstrate an earlier prehistoric (Central Plains Tradition or possibly Oneota) occupation. Both the morphology of the hearths (rectangular rather than round or oval) and their position in the landscape (on an active floodplain) are uncommon for sites of their age within the middle Missouri River valley (NSHS 2004; Bozell et al., forthcoming). Floodplain Oneota habitations and camps have been recorded in the Des Moines River valley of central Iowa and the Mississippi River alluvial plain in southeast Iowa (see Gradwohl 1967; Henning 1998, 373–75; Alex 2000, 198) but not in the Missouri valley.
Finally, the site is situated on a low, late Holocene terrace-fan complex on the margins of the Missouri River valley bottom (Dillon 2013). This position in the middle Missouri River landscape has received little attention by previous investigators (e.g., Guccione 2008).
The archaeological excavations at Engineer Cantonment are described in detail by NSHS (2004) and Bozell et al. (forthcoming). In this paper we present the results of the soil-geomorphic investigation conducted in support of the NSHS archaeological excavations at Engineer Cantonment. We characterize the late Holocene alluvial landforms, geology, and soils of the site. We found many detailed studies of Holocene alluvium, terraces, and alluvial fans in the tributary valleys along the Missouri River in western Iowa and eastern Nebraska (e.g., Daniels et al. 1963; Hoyer 1980; Bettis 1990; Mandel 1995; Mandel and Bettis 2001a; Bettis and Mandel 2002; Dillon and Mandel 2008), but few detailed studies of Holocene alluvium within the Missouri River valley itself. Most previous investigations of the middle Missouri River valley bottom have taken a regional approach and provide only brief descriptions of Holocene-age deposits (e.g., Dahl 1961; Miller 1964; Shroba et al. 2001). Here we provide detailed descriptions of late Holocene through historic fluvial and alluvial fan deposits and associated buried soils. We also explain the human and geomorphic processes that led to such a well-preserved archaeological site. Finally, we present additional research questions and strategies.
The site is located at the foot of the bluffs on the western margin of the Missouri River valley in eastern Nebraska (Fig. 1). The Missouri River valley is 6.25 km wide at the site. The modern, regulated Missouri River channel is 1.75 km to the east of the site. The mean annual temperature at nearby Blair, Nebraska, is approximately 10.4°C, with an average annual precipitation of approximately 73 cm. However, during the period of record (1893–2001) yearly rainfall totals have ranged from 38.7 cm up to 105.4 cm.
The Missouri River valley in eastern Nebraska forms the eastern margin of the Great Plains (Lavin et al. 2011). In general, the valley floor includes Pleistocene through modern alluvium and slope wash with textures ranging from sand and gravel to clay (e.g., Dahl 1961; Miller 1964; Mason 2001). The pre-channelization floodplain comprises the bulk of the valley floor, and is marked by a complex of abandoned channels, meander scars, splays, and hummocky topography (Dahl 1961; Miller 1964; Mason 2001). Remnants of low (~2 m relief) alluvial surfaces occur throughout the valley floor, while remnants of higher, older terraces occur along valley margins (Dahl 1961; Miller 1964; Mason 2001). The higher terraces, such as the Fort Calhoun Terrace to the north, are composed of older Pleistocene alluvium capped by late Pleistocene loess. Alluvial fans occur along the valley margins where small streams enter the main valley of the Missouri River. The fans range in size from less than 50 m2 near gullies [End Page 3]
[End Page 4]
to greater than 2 km2 where larger tributaries supply the sediment. The valley margins also include colluvial aprons composed of slope wash. The adjacent uplands include a thick sequence of Pleistocene loess, glacial till, and glacio-lacustrine and/or alluvial deposits with buried soils (Miller 1964; Dreeszen 1970; Bettis et al. 1986; Mason 2001). The Pleistocene and Holocene sediments overlie Pennsylvanian limestone and shale of the Kansas City and Lansing groups (Burchett 1970).
The geoarchaeological field investigation included detailed topographic mapping of the site with a Total Station, excavation of four backhoe trenches, completion of 18 soil borings with a Giddings hydraulic soil probe, two hand-auger borings, and a large shovel test (Fig. 3). Soils and sediments were described in the field and laboratory using standard terminology and procedures (Soil Survey Staff 1993; Schoeneberger et al. 2002). We use informal names (Units I through IV) to identify the major deposits that comprise the site. Where possible, we correlate these deposits with sediments and stratigraphic units described by other authors.
We submitted oil and sediment samples collected from profiles T4P1, T4P3, and T4P4 from Trench 4 to the University of Nebraska–Lincoln Soil and Plant Analytical Laboratory for analysis. Particle size was determined using a Coulter Counter. Total inorganic carbon was measured by chemical oxidation via titration (Walkley-Black method). Carbon-14 (14C) age determinations on three charcoal samples were completed at Beta Analytic, Inc., using conventional techniques. All ages are reported as uncalibrated radiocarbon years before present (BP). More recently, we submitted samples from eight soil borings (EC3 through EC8, EC13, and EC14) to the Kansas Geological Survey Geoarchaeology and Paleoenvironment Laboratory for particle-size analyses using a pipette method (Gee and Bauder 1986). [End Page 5]
[End Page 6]
[End Page 7]
[End Page 8]
The site extends across two alluvial landforms. A portion of the site is on a low fluvial terrace remnant approximately 200 m long by 35 m wide, bounded by bluffs to the west and the modern Missouri River valley bottom on the east. Local relief from the site to the adjacent bluff top is about 72 m . The terrace stands approximately 1.2 m above the modern floodplain. The site also extends onto a small alluvial fan at the mouth of a narrow gully that issues from the bluffs. The gully drains approximately 2.3 ha. At its apex the alluvial fan stands approximately 2 m above the surrounding land surface. The land surface slopes gently to the northeast, away from the bluffs. The land east of the adjacent county road is the modern Missouri River floodplain, as evidenced by multiple meander scars and splay deposits identified on topographic maps and aerial photographs.
We identified four basic depositional units at the site on the basis of texture, soil and weathering features, and stratigraphic position (Figs. 4, 5, and 6). We refer to these informally, from oldest to youngest, as Unit I through Unit IV. Unit I is a late Holocene Missouri River alluvial fill. Unit II includes multiple beds of alluvial fan sediments that overlie Unit I. Unit III is an alluvial fill that is cut into the upper portions of Unit I and truncates and/or interfingers with the distal portions of Unit II. Unit IV consists of modern Missouri River channel and overbank sediments. We included representative sections with soil profile descriptions for each unit in Appendix Tables 1–3.
Unit I (Late Holocene Missouri River Alluvium)
Unit I consists of sand and gravel channel deposits that fine upward to sandy silt, silt, and clay overbank sediment. Soil borings EC3 and EC4 demonstrate that Unit I is at least 7 m thick at the site. The deepest boring, EC3, encountered coarse sand and fine gravel at a depth of 7.16–7.31 m below ground surface (bgs). The fine gravel is composed of limestone, pink feldspar, quartz, and opaque dark gray silicates. The base of the core includes tabular limestone clasts that are only slightly rounded. The core barrel could not advance below 7.31 m, thus we could not determine if the limestone clasts are a channel deposit or the top of the bedrock. The upper 1.5–2.0 m of Unit I consists of gray, dark gray, and grayish brown fine silt and silty clay with common redoximorphic features such as mottles, Fe-Mn oxide masses and channel linings, and reduced colors (Fig. 7).
The more clayey portions of Unit I are very firm, with fine angular blocky structure and few slickensides. We found little evidence of soil formation in the upper portion of Unit I at this site. The top of the unit does not have a high organic matter content or darkened color typical of soil A horizons (Fig. 8). However, it does have biogenic features that suggest some degree of subaerial exposure, such as root channels and scattered voids filled with fine worm casts. Also, its structure and redoximorphic features are commonly observed in hydric soils (e.g., Vepraskas 2001; Vasilas et al. 2010). However, the upper portions of Unit I also include thin beds of laminated silt and very fine sand that represent brief episodes of higher energy flow—possibly distal splay deposits (Fig. 7B). These thin beds have not been disturbed by pedogenic or other processes, yet the clayey sediments immediately above and below include redoximorphic features and angular structure. Thus, we interpret the angular blocky structure and redoximorphic features as products of the high clay content of the alluvium and a fluctuating water table, respectively, rather than traditional, surface-downward pedogenic processes such as illuviation.
F0402 is a rectangular fire pit or hearth that was encountered at a depth of 228 cm bgs in the western portion of Trench 4 (Fig. 3). This feature is 210 cm long (east–west) by 54 cm wide (north–south) and 15 cm thick. F0402 was filled with silty sediment and contained abundant ash and charcoal (elm, ash, and black walnut) as well as 40 small animal bone fragments, including identifiable beaver (Castor canadensis) and vole (Microtus pennsylvanicus) elements. The base of the feature mostly consisted of wood charcoal and burned earth. Charcoal from this feature yielded a radiocarbon age of 870 ± 40 yr BP. A second feature, F0403, is located approximately 7.5 m southeast of F0402 in Trench 4. F0403 also is a rectangular feature and was encountered at a depth of 180 cm bgs. It is 134 cm long (east–west) by 34 cm wide (north–south) and 10 cm thick. Charcoal from F0403 yielded an age of 820 ± 40 yr BP. The third sample was a charcoal fragment recovered from the upper 15 cm of Unit I on the eastern end of Trench 4 (Fig. 6). This sample yielded a radiocarbon age of 720 ± 60 yr BP.
These three 14C ages demonstrate that Unit I is a late Holocene fill. Based upon its lithology and position in [End Page 9] the landscape Unit I is equivalent to the late Holocene terrace fill (Qam3) described by Mason (2001). The age of the lower portion of Unit I at Engineer Cantonment is unknown, and Mason (2001) did not recover radiocarbon-datable materials from Qam3. However, Mason reported a radiocarbon age of 890 ± 60 yr BP from the lower portion of a younger fill (Qam2). At Engineer Cantonment, the two radiocarbon ages from hearths indicate that the upper portion of Unit I dates to ca. 850 yr BP. Thus, the underlying Unit I deposits are at least that old.
A 20-to 40-cm-thick mantle of Missouri River alluvium overlies the hearths in Unit I. Thus, aggradation of Unit I continued until sometime after ca. 850 14C yr BP at this site. The 14C age of 720 ± 60 yr BP from sedimentary charcoal in the eastern portion of the site suggests the presence of a younger late Holocene fill within Unit I. Mason (2001) described multiple, laterally discontinuous cut-and-fill sequences within the Holocene-age Missouri River fills. However, he utilized extensive sections exposed at an active limestone quarry. At Engineer Cantonment, exposures are limited to shallow backhoe pits and borings, and thus we cannot confirm the presence of a younger fill within Unit I.
Unit II (Alluvial Fan Deposits)
Unit II includes multiple beds of light yellowish brown (10YR 6/4) to brown (10YR 5/3) and gray (2.5Y 6/2) silt and silt loam. Individual fan beds are separated by buried soils or abrupt contacts. The beds are generally thickest (up to 81 cm) and well oxidized near the fan apex, and become darker, finer grained, and thinner in the distal portions of the fan. Also, Unit II includes a prominent buried soil that was the stable land surface when Engineer Cantonment was constructed in AD 1819–20. We refer to this buried soil informally as the Cantonment Soil. Unit II is the most internally complex unit at the site.
The surface soil and most buried soils within Unit II are weakly developed Entisols with A-C horizonation. The A horizons are typically dark gray (10YR 4/1) or grayish brown (10YR 5/2) with weak granular or platy structure. The platy structure is a remnant of lamination rather than true pedogenic structure. The parent materials are massive or laminated and commonly include scattered fine charcoal fragments.
The Cantonment Soil is buried by approximately 1.2 m of fan deposits near the apex of the fan and less than 30 cm in the distal portions of the site. Near the apex the Cantonment Soil exhibits A-Bw-C horizonation. The A horizon is approximately 15 cm thick and composed of dark gray (10YR 4/1) silt loam with granular structure, common pores, and earthworm casts. The Bw horizon is thin (13 cm) but discernible by its brown (10YR 5/3) color, subangular blocky structure, and slightly higher clay content than the A horizon above and the parent material below. In the medial portions of the fan the parent material for the Cantonment Soil is gray (2.5Y 6/1–6/2) silt loam. In this portion of the site the A and Bw horizons include thin, discontinuous silt and clay films along fractures and pores. The thin, dark flows are probably the result of silt and clay in suspension deposited on the surface during times of flooding (e.g., Brammer 1971; Mandel and Bettis 2001a). Trench 4 and the soil borings demonstrate that the Cantonment Soil extends across most of the site (Figs. 3 and 6). [End Page 10]
Previous archaeological investigations have recovered many historic artifacts from the Cantonment Soil and overlying fan sediments (NSHS 2004; Bozell et al., forthcoming). The top of the Cantonment Soil is the AD 1819–20 land surface. The sediments above the Cantonment Soil yield diagnostic late 19th-to mid-20th-century former farmstead materials such as brown glass, flat nails, barbed wire, etc. The fan sediments below the Cantonment Soil did not yield any cultural materials, but they must be younger than approximately 870–720 14C yr BP because they bury Unit I where the hearths are located (Fig. 6). Based upon its position in the landscape and age we correlate this alluvial fan with mapping unit Qaf2 of Mason (2001).
Unit III (Historic Alluvium)
Unit III occurs beneath the low terrace surface at the eastern end of the alluvial fan (Fig. 6). Unit III includes a weak surface soil formed in approximately 1.5 m of thin-bedded and laminated, dark grayish brown (10YR 4/2) and light brownish gray (10YR 6/2) alluvium. The texture of Unit III is similar to that of Unit II (Fig. 4). However, Unit III is slightly darker than the alluvial fan sediments to the west and is recognized as a separate fill by an abrupt contact at the top of the underlying fill. Also, Unit III is thicker than the adjacent fan deposits (Fig. 6). The lateral contact between Unit III and the fan is obscured by intense bioturbation (mostly medium and coarse tree roots) and recent farming and road construction activity. The lower contact with Unit I is marked by an abrupt textural change from laminated, sandy silt loam (24% very fine sand) at the base of Unit III to massive silty clay loam with <1% sand in Unit I. Unit III was exposed in the eastern portions of Trenches 1 and 4, in soil borings EC-13, EC-16, and EC-17, and in hand-auger boring HA-1 (Figs. 3 and 6). We could not determine the eastern extent of Unit III because it is buried by the paved county road.
We did not find datable organic or cultural materials from Unit III. However, Unit III is likely a historic alluvial fill due to its position above Unit II and the modern floodplain deposits (Fig. 6). It could represent a gully fill associated with late 19th-to 20th-century farming activity and the county road, or it might comprise the western edge of Unit IV.
Unit IV (Modern Missouri River alluvium)
Unit IV consists of a thin mantle of fine-grained over-bank deposits that overlie sandy lateral accretion deposits. We placed Trench 3 (Fig. 9) on the valley bottom to the east of the county road and exposed a thin surface soil developed in 28 cm of dark grayish brown (2.5Y4/2) silt loam with weak platy structure. An abrupt boundary separates the silt loam from the underlying laminated and cross-bedded light olive brown (2.5Y 5/3) and gray (2.5Y 4/1) fine sandy loam. Trench 3 was about 2.15 m deep, but additional shovel-testing at the bottom of the pit demonstrated that the laminated silty sand extends to at least 3.5 m below ground surface.
We correlate Unit IV with map unit Qam2 (Missouri River floodplain alluvium) described by Mason (2001). Unit IV is an alluvial fill that predates the Missouri River’s late 19th-to early 20th-century transition from a highly sinuous meandering channel to a lower sinuosity, semi-braided channel (e.g., Hallberg et al. 1979; Mason 2001). The Peale painting (Fig. 2A) shows a sandy floodplain and meander cutoff in the foreground. [End Page 11] We interpret the bedded silty sand of Unit IV as near-channel facies and the silty clay as overbank deposits that subsequently filled the meander cutoff.
Chronology and Site Interpretation
Sometime before ca. 900 14C yr BP, the Missouri River channel was at or very near the site locality. The basal gravel, sand, and beds of sandy silt in the lower portions of Unit I were deposited within the Missouri River channel or in a very near-channel environment. The river then shifted to the east, away from the bluffs. The abrupt change from sandy silt to silty clay and clay (Fig. 5) indicates a rapid shift away from the site—likely a meander cutoff event. At the time of the late prehistoric Native American occupations, the channel had moved far enough to the east that the site was now a distal floodplain, or backwater environment. The clayey beds of Unit I aggraded during times of flood when the site was inundated by clay-and silt-laden water. The brief occupations occurred when the ground was dry enough for small encampments.
The alluvial fan (Unit II) began to aggrade shortly after the valley bottom became stable enough for the accumulation of sediment at the mouth of the gully that issues from the bluffs. At least one thin bed of fan sediment interfingers with the clay-rich portions of Unit I at the western edge of backhoe Trench 4 (Fig. 6). Deposition was episodic, with at least five beds separated by abrupt lower contacts and weak soil development. As the fan grew, the floodplain surface was progressively buried. At the time of occupation by the Long Expedition, the site was no longer part of the active floodplain. Instead, it was a low terrace-fan complex standing approximately 1–2 m above the active channel. The Cantonment structures were built on the fan.
After AD 1820, sediment comprising Unit III was deposited on the eastern portion of the site. Unit III appears to be a shallow gully fill associated with late 19th-and 20th-century farming activity and the adjacent county road. Unit IV aggraded sometime after ca. 720 14C yr BP. Its upper surface represents the active floodplain at the time of Engineer Cantonment. In the decades following the abandonment of Engineer Cantonment, the Missouri River continued to migrate on its floodplain, eventually settling in its current position where it was levied and stabilized during the early to mid-20th century (Sayre and Kennedy 1978; Hallberg et al. 1979).
The excavation of Engineer Cantonment yielded a rich assemblage of artifacts (NSHS 2004). The site owes its high level of preservation to four basic factors. First, the site locality is within a depositional environment. Engineer Cantonment was built on a small alluvial fan that merges with a terrace underlain by late Holocene alluvium. Since its abandonment in June 1820, the site has been subject to episodic deposition of sediment from the adjacent uplands and also from high-magnitude Missouri River flood events. By the time of the late 19th-century farmstead occupation, the Engineer Cantonment ruins were buried by approximately 10 to 30 cm of sediment.
The second factor contributing to preservation is the nature of soil development at the site. The upper portion of Unit I exhibits minimal evidence of pedogenesis. Thus, even though the sediments are clay-rich, they have not been subject to significant shrink-swell processes that are common to soils formed in fine-grained parent materials. The soils in the fan, including the Cantonment Soil, are formed in silt-rich parent materials with minimal or no B horizon development. Thus, the soil processes associated with clay-rich B horizons and sediments (argilliturbation) were not a factor (see Holliday 2004, 275–77).
The third factor favoring preservation is that this site has not been plowed. Portions of the site were impacted by farming practices, particularly animal pens, which may have destroyed evidence of a second Cantonment structure, but the remainder of the site was undisturbed (Bozell et al., forthcoming). Additionally, aerial photographs dating to the 1940s indicate that dense stands of mature trees have grown only in the southernmost portion of the site and along the eastern boundary near the county road. The central portion of the site, where the Cantonment structure was located, has remained clear or mostly clear of trees.
Finally, the site has not been eroded by the extensive late 19th-and 20th-century migration of the Missouri River channel. Thus, the alluvial fan at the site has been able to prograde across the Cantonment ruins. At the time of excavation, the structure was mantled by 50–70 cm of fan sediment.
Our results point to some strategies that can be employed in future geological and archaeological investigations [End Page 12] in the middle Missouri River valley. For example, the valley bottom along the Nebraska-Iowa border is dominated by 19th-and 20th-century alluvium (Mason 2001). However, the valley margins are a tapestry of landforms and deposits that span the late Pleistocene through modern times. Of particular interest are the alluvial fans of various sizes that occur where tributary streams enter the Missouri valley. The age of each fan is a function of when the Missouri River channel was last flowing against the bluffs at that location. At some localities, arcuate or linear bluff lines with steep slopes rise from the valley bottom with no alluvial fans or terrace remnants. Along other stretches there are small alluvial fans that mantle older alluvium (such as at Engineer Cantonment), larger fans that cover more than 2 km2, or multiple fans that have coalesced to form a bajada (a broad alluvial slope) along the base of the bluffs. Alluvial fans also occur on the higher, late Pleistocene terrace remnants, such as on the Fort Calhoun terrace to the north of Engineer Cantonment (Mason 2001).
Alluvial fans often contain a rich record of environmental and cultural change. For example, buried soils and charcoal within the fans can yield radiocarbon chronologies that document landscape evolution in various portions of the valley. Also, various proxies, such as plant macro-and microfossils and the δ13C values of soil organic matter can be used to assess bioclimatic changes (e.g., Baker at al. 2000; Gorynski and Mandel 2009; Cordova et al. 2011). In addition, detrital charcoal within the fan deposits are a potential record of fire on the uplands and within the basins that feed sediments to the fan (e.g., Meyer et al. 1995; Pierce and Meyer 2008; Nelson and Pierce 2010). Finally, the fans along the middle Missouri River valley margin may contain an archaeological record that spans most if not all of the Holocene (e.g., Hoyer 1980; Mandel 1995; Bettis 2003).
Additionally, our findings demonstrate that the floodplain alluvium beneath the fans can also contain archaeological features and deposits. The two hearths from the upper portions of Unit I were likely created by people associated with either the Central Plains or Oneota traditions and cannot be linked with any specific modern tribe. Most recorded sites associated with these cultural traditions in eastern Nebraska are on high terraces, bluffs, and ridges overlooking valleys (e.g., Steinaker and Carlson 1998; Bozell and Ludwickson 1999). The presence of a Central Plains or Oneota camp located on a floodplain is a very poorly known site type in the middle Missouri River valley. Also, rectangular pits and hearths of this age are extremely rare, most being round or oval. Thus, our discovery of these sites has potential for a better understanding of late Holocene, special-purpose floodplain camps likely associated with subsistence procurement activities along the Missouri River in eastern Nebraska (Bozell et al., forthcoming).
We also suggest that 19th-century western art may be useful for identifying archaeological sites and assessing environmental change. The configuration of the bluff line shown in the Peale sketches and watercolor of Engineer Cantonment was critical to locating and confirming the site. An interesting aspect of the painting is that Peale portrays the bluffs with very few trees, while the modern bluffs are heavily wooded (Fig. 2B). Indeed it was a view of the modern bluffs during winter (with leaves fallen) that led to the recognition and rediscovery of the site (NSHS 2004). Later paintings of this area by Karl Bodmer, for example, also show the bluffs with relatively few trees (see Missouri in the Morning below Council Bluffs: Bodmer 1984). If these depictions are accurate, there has been an increase in forest density on the bluffs since the early 19th century. Members of the Long Expedition noted the lack of trees along the bluffs near Saint Louis, Missouri. They described Native Americans setting fire to the uplands, and speculated that the purpose was to promote the growth of grass and attract game (James 1823, 405). Another possibility is that late 19th-and 20th-century upland agriculture and fire suppression may have favored the development of forest along the bluffs. However, the change from few trees to dense forest (if accurate) might also suggest a subtle change in climatic influences during the last 150–200 years. There are few proxy climate records for this time period in eastern Nebraska. Thus, paintings by Peale, Bodmer, and others might be useful tools in the development of climate reconstructions.
The February 1820 painting of the site by Titian Ramsay Peale (Fig. 2A) also shows the Cantonment structures sitting on a low rise. Our excavations have demonstrated that the structures were indeed built on a low alluvial fan and subsequently buried by younger alluvial fan deposits. Bull (1977, 223) suggests that the first published description of alluvial fans was in 1841, and the term “alluvial fan” first appears in the geologic literature in 1873. We would like to think that Peale was expressing his “inner geomorphologist,” but more likely he was simply portraying the landscape as accurately as he could. Another possibility is that Peale was careful to show the Cantonment structures on a low rise in order [End Page 13] to contrast their site selection from that of the military branch of the expedition. The military established their encampment (Cantonment Missouri) on the lower ground of the floodplain. During the late winter and spring Cantonment Missouri was flooded; Engineer Cantonment was not (Bozell et al., forthcoming).
Also note the prominent channel bend shown in the painting. This is the “cove” or “harbor” described in the expedition members’ journals. Figure 10 is a portion of a digital orthoquadrangle of the study area (image dated June 16, 2010). Channel and meander scars are shown by darker tones. The meander scar indicated on the image is in the same position as the channel bend shown in the painting. Figure 2B is a recent photograph of the site from a point to the east of the meander scar. Recent flooding left this low area under water until approximately July 14, 2014. In the photo taken on July 21, 2014, the channel bend is marked by the dead vegetation in the foreground. Note the similarity to the painting. The “channel” in the painting clearly is a meander cutoff. We cannot say if this was the main channel of the Missouri River or a smaller branch of the channel, similar to the nearby Boyer Chute on the modern landscape. Based upon the size of the channel in the painting, and the channel scar, it probably was the latter.
Western art, particularly that of Karl Bodmer and George Catlin, has been used to aid archaeological interpretations of features at Fort Clark and the adjacent Mandan village, called Mih-tutta-hang-kusch, along the Missouri River in North Dakota (Wood et al. 2011, 58–71; [End Page 14] Mitchell 2014, 2–8). Paintings by Bodmer and Catlin also proved useful to the excavations at the Fontenelle and Cabanne trading posts near Omaha, Nebraska (Jensen 1998). Wagner (1988, 1989) discusses the movement of 19th-century landscape artists from idealized, pastoral landscapes toward geologically and ecologically accurate portrayals. This was in part related to the increasing popularity of geology and natural science at the time (Bedell 2001). Comparisons of modern landscapes with sketches and paintings by 19th-century artists and scientists (e.g., Murphy 2011; Lindholm et al. 2013) have illustrated how detailed these works can be. Thus, they represent a potentially important research tool that should not be overlooked.
Engineer Cantonment is a stratified, late Holocene through modern archaeological site on the western margin of the Missouri River valley north of Omaha, Nebraska. The late Holocene Native American occupations are somewhat unique with their rectangular-shaped hearths and location on the valley bottom rather than on terrace or upland surfaces. The historical component, Engineer Cantonment, is significant because it is the oldest Euroamerican site discovered in Nebraska. It is also unique in that it represents the first, if not only, example of government-sponsored western exploration in which trained scientists made an extended stay at a single locality.
The site is only about 35 m wide, occupying less than 1% of the Missouri River valley bottom, and yet it contains a complex record of late Holocene Missouri River channel cutting, filling, and alluvial fan development. Most of the Missouri River valley bottom is underlain by a complex assemblage of historic-age channel and over-bank deposits. However, terrace remnants and alluvial fans occur along the valley margins that preserve much of the late Wisconsin through late Holocene record. Engineer Cantonment is an example of the rich geological and archaeological record preserved on the margins of the Missouri River valley floor, and also an example of the value of 18th-century landscape art as a research tool for geoscientists.
Jeremy S. Dillon (firstname.lastname@example.org), Department of Geography and Earth Science, University of Nebraska at Kearney, Kearney, NE 68849
This project was made possible through funding from the Nebraska State Historical Society, Nebraska State Historical Society Foundation, Nebraska Department of Roads, and the Federal Highway Administration. We wish to thank the landowners, Herb and Gloria Gibreal, and also Mr. Gayle Carlson and Mr. Robert Pepperl for their assistance. Finally, we wish to thank the three reviewers whose comments and suggestions greatly improved this manuscript. [End Page 15]