This section highlights the map units (i.e., rocks and unconsolidated deposits) that occur in Theodore Roosevelt National Park and puts them in a geologic context in terms of the environment in which they were deposited and the timing of geologic events that ultimately created the present landscape.

Table 1. Geologic Time Scale

Pre-Paleocene
The geologic features displayed in Theodore Roosevelt National Park are primarily Paleocene and younger in age (=65 million years old). However, the geologic story of the area begins much earlier. The older rocks buried deep beneath the park contain significant natural resources such as oil and gas, which are extracted from Pennsylvanian, Mississippian, Devonian, and Ordovician rocks (see table 1) in the Williston Basin. The Williston Basin is a large, roughly circular depression on the North American craton covering several hundred thousand square miles across parts of North and South Dakota, Montana, and the Canadian provinces of Manitoba and Saskatchewan. The Williston Basin began subsiding during the Ordovician Period. Although the Williston Basin was subsiding, marine sediments were not continuously deposited in it. Nevertheless, the basin contains an unusually complete rock record compared to many basins; some rocks from each Phanerozoic period are preserved there. These strata record several cycles of marine transgressions that filled the basin, followed by marine regressions that drained it. In all, more than 16,000 feet (4,800 m) of sediments, mostly shallow marine but also some terrestrial deposits, settled into the Williston Basin.

The early structural history of the basin is poorly understood (Heck et al. 2002). The earliest rocks are difficult to study because the Lower Phanerozoic and Precambrian (see table 1) rocks are not exposed at the surface in North Dakota and only a few wells have penetrated these rocks. Present understanding of the early geologic history of the basin is pieced together from outcrops in adjacent states and provinces, seismic data, and limited well data (Heck et al. 2002).

During the Cretaceous, an inland sea covered the interior of the continent and more marine sediments collected. The Cretaceous Interior Seaway was hundreds of miles wide and divided North America into two separate land masses. The northwest- southeast oriented epicontinental seaway stretched from the Arctic Ocean to the Gulf of Mexico. The sea retreated from most of the continent by about 65 million years ago. Coincident with this marine regression, the Laramide Orogeny uplifted the Rocky Mountains hundreds of miles west of the park. Uplift produced hundreds of cubic miles of sediment that eroded from the newly formed mountains. Streams carried this sediment eastward and deposited it in great clastic wedges across the Great Plains. A thick sequence of terrestrial sediments, which ranges in age from Late Cretaceous through Oligocene, records this
event in many parts of the Dakotas and Montana.

Paleocene
In North Dakota and Theodore Roosevelt National Park, the Fort Union Group represents the Paleocene portion of the clastic wedge, which sloped eastward from the Rocky Mountains. The exact depositional setting of the Fort Union Group has been the subject of some debate. The sediments of the Fort Union Group were deposited on what various investigators have called an“alluvial plain” (Laird 1956), a “meandering fluvial channel system” (Fastovsky and McSweeney 1991), and a“broad sea- level delta” (Theodore Roosevelt National Park Web site, accessed October 25, 2004). Regardless of the specific geomorphic setting, water was apparently ubiquitous in time and space. Thus a reasonable interpretation for the general environment is a lowlying,
swampy region in which water- loving trees and other plants grew.

As leaves, branches, and entire dead trees fell into the stagnant waters of the swamps, they eventually became peat as bacteria only partially decomposed them and as additional sediments compressed the organic material. Streams continued to deposit more and more sediment in the area, creating lignite. If more pressure had been applied, bituminous (soft) coal would have eventually formed. A total thickness of about 1,350 feet (410 m) of sediment was deposited during the Paleocene Epoch, forming the 600- foot- (180 m) thick Bullion Creek Formation and the 750- foot- (230 m) thick Sentinel Butte Formation.

Also, during this time, volcanoes punctuated the landscape of the western United States. Rivers and winds transported the erupted ash into North Dakota. Along with other sediments and organic material, this ash accumulated in standing water. Silica (quartz) from these ash deposits played a major role in the petrifaction of cypress, sequoia, and other deciduous and cone- bearing trees that grew in the low- lying, subtropical environment. Few representative fossil- leaf locales are in the park, so investigators have interpreted the existence of various species from a fossil site called Wannagan Creek. This site is in the Bullion Creek Formation and is located just to the west of the South Unit (John Hoganson, North Dakota Geological Survey, written communication, December 7, 2004). Groundwater moving through the silica- rich volcanic ash dissolved the silica. When this silica- saturated water soaked into the trees, microcrystalline material replaced the organic compounds in the wood. In some cases, the internal structures of the trees including growth rings were
preserved in stone. In addition to allowing petrifaction of wood, the Paleocene ash layers formed the bentonite “blue beds,” which now contribute to the scenic badlands topography in the park.

Post-Paleocene
Scattered remnants of post- Paleocene sediments are present throughout western North Dakota as isolated caps on buttes. These units, including the Golden Valley Formation and the White River Group, were deposited during the Eocene and Oligocene epochs (56–23 million years ago). In the park, a lag deposit composed of silcrete called the Taylor bed (the upper portion of the Bear Den Member of the Golden Valley Formation), caps the Archenbach Hills of the North Unit. Geologists interpret this lag deposit as a weathering horizon developed on top of the Sentinel Butte Formation (Clayton et al. 1980).
The weathering horizon marks a hiatus in deposition, called an unconformity, between the Bear Den Member and the overlying Camels Butte Member (Biek and Gonzalez 2001). This silcrete lag contains only silica- rich chert and lacks lithic fragments such as feldspar and other readily weathered minerals. This composition suggests that resistant quartz was all that remained after a prolonged period of weathering (Hickey 1977).

As a result of intense long- term weathering and erosion, no bedrock units in western North Dakota are younger than Oligocene in age. Accumulations of gravel and sand (e.g., unit QTa on the geologic map) are difficult to date but provide the only evidence of deposition subsequent to Oligocene time. Post- Laramide regional uplift at the end of the Paleocene Epoch caused a change in regional base level, forcing streams to incise their channels. For millions of years, streams had been depositing sediment nearly continuously on the Great Plains. In western North Dakota the uplift caused the rivers ancestral to the
modern Little Missouri River system to dissect the plains, incising and eroding away much of the poorly consolidated upper rock layers. Huge volumes of
sediment from the northern Great Plains were carried towards Hudson Bay.

Figure 16. Pre- and Post-Glacial Drainages of North Dakota. A. Rivers flowed north into Canada and northeast to Hudson Bay before glaciers diverted them. B. Glacial diversion caused the rivers to change direction. Source: Murphy et al. (1999).

Pleistocene
After much incision, river channel courses were firmly established at the beginning of the Pleistocene Epoch (1.8 million years ago). At the same time global climate changes triggered the advance of great ice sheets from the north. These continental glaciers formed, advanced, and retreated many times during the Pleistocene Epoch. Glacial erratics in the North Unit record the farthest local extent of glacial advance onto the Missouri Plateau. The glacial effects in this area are not particularly pronounced; classic glacial features such as moraines do not appear in this landscape. However, glacial effects were significant with respect to the changes made to regional drainage patterns. These changes strongly influenced the recent geology of Theodore Roosevelt National Park (Laird 1956).

Before the initial advance of continental ice sheets, the Missouri River flowed northeastward into Canada and to Hudson Bay (figure 16). Its major tributaries, the Yellowstone and Little Missouri rivers, joined the Missouri in northwestern North Dakota. The eastflowing Knife, Heart, and Cannonball rivers in North Dakota also joined a stream that flowed northward to Hudson Bay (Trimble 1993). When continental ice sheets advanced southward from Canada and reached as far as the upper North Unit in the park, the ice blocked the courses of these north- flowing rivers. This forced them to create new routes eastward and southward, thereby emptying into the Mississippi River instead of Hudson Bay.

According to Biek and Gonzalez (2001), glacial diversion of the Little Missouri River occurred by mid- Pleistocene time, at least 640,000 years ago, though the exact timing is uncertain. Nevertheless, by the time the ice sheet retreated, the northern portions of both the Little Missouri and Missouri rivers were entrenched into new channels. The new route of the northern Little Missouri River followed a steeper course, causing the whole river to flow faster and begin downcutting rapidly into the layered sediments. Because the elevation of the mouth of the Little Missouri was now considerably lower than it had been before joining the Yellowstone River (just east of present- day Williston), it eroded quickly through the soft sedimentary rocks. As the river began rapid incision, its tributaries also began cutting gullies on a grand scale, carving the fantastically broken topography of the badlands. In addition, as streams eroded the poorly consolidated sediments, many valley walls became oversteepened and unstable. The presence of bentonite exacerbated this situation. These slopes often failed causing landslides. The most spectacular landslides are those in the North Unit, where two classic types of rotational slumps are present (see “Mass Wasting” section).

The rate of erosion and incision of the Little Missouri River was not constant. As intermittent erosion and deposition continued, the stream cut a series of four terraces, the remnants of which can still be seen in the park (Harris and Tuttle 1990). The complicated story recorded in these terraces is still unclear, and the lack of reliable radiometric dates has lead to various mapping styles and nomenclature for these deposits (Biek and Gonzalez 2001). Correct geomorphic interpretation of these terraces is significant for completely understanding (1) when the ancestral Little Missouri River occupied the highest terrace level, (2) the exact timing of drainage diversion of the ancestral Little Missouri River, (3) the timing of inception of incision of badlands erosion, and
(4) the rate of formation of badlands topography.

Holocene
In the past 11,800 years, various geologic processes that began during the Pleistocene have continued. This includes the formation of badlands topography, mass wasting, and the formation and mantling of pediments (see “Mass Wasting” and “Pediments” sections). However, three types of deposits in the park are exclusively Holocene in age: alluvial fans, various alluvium (units Qoal and Qal on the geologic map), and engineered fill. Alluvial fans are present at the mouths of nearly every small valley in the park where they mingle with stream deposits. The upstream portion of alluvial fans is gradational, interfingering with sheetwash and colluvial deposits that mantle valley margins. Other alluvial deposits of Holocene age are found along stream
channels show that most streams have been aggrading for the past 150 years or more, but the upper reaches have incised during this same period (Biek and Gonzalez 2001). Finally, some areas of engineered fill occur in the park: along U.S. Route 85 in the North Unit and along I- 94 in the South Unit. These mappable units show that humans are agents of landscape change.

References:

Biek, R. F., and M. A. Gonzalez. 2001. The geology of Theodore Roosevelt National Park, Billings and McKenzie counties, North Dakota. Miscellaneous Series
86, text (74 p.), 2 pls. (scale 1:24,000). Bismarck: North Dakota Geological Survey.

Clayton, L., with assistance from S. R. Moran, J. P. Bluemle, and C. G. Carlson. 1980. Geologic map of North Dakota. Special Geologic Map (scale 1:500,000).
Washington, D.C.: U.S. Geological Survey.

Fastovsky, D. E., and K. McSweeney. 1991. Paleocene paleosols of the petrified forest of Theodore Roosevelt National Park, North Dakota: A natural experiment in compound pedogenesis. Palaios 6:67–80.

Harris, A. G., and E. Tuttle. 1990. Geology of national parks, 4th ed. Dubuque, IA: Kendall/Hunt Publishing Company.

Heck, T. J., R. D. LeFever, D. W. Fischer, and J. LeFever. 2002. Overview of the petroleum geology of the North Dakota Williston Basin. Bismarck: North Dakota Geological Survey. http://www.nd.gov/ndgs/Resources/WBPetroleum_H. htm (accessed April 19, 2007), Web page modified February 7, 2007.

Hickey, L. J. 1977. Stratigraphy and paleobotany of the Golden Valley Formation (early Tertiary) of western North Dakota. Memoir 150, text (181 p.), 55 pls. (scale 1:250,000). Boulder, CO: Geological Society of America.

Hoganson, J. W., and J. Campbell. 1997. Paleontology of Theodore Roosevelt National Park. NDGS Newsletter 24(1):12–23.

Laird, W. M. 1956. Geology of the North Unit, Theodore Roosevelt National Memorial Park. Bulletin 32. Bismarck: North Dakota Geological Survey.

Trimble, D. E. 1993. The geologic story of the Great Plains. Medora, ND: Theodore Roosevelt Nature and History Association.