The objective of this work is to provide a scientifically based, hierarchical, system of map units (sensu Bailey 1980) for both northwestern Wyoming and the Buffalo Resource Area (BRA) of northeastern Wyoming that incorporates ecological principles and processes across a range of scales. The work area includes all of Campbell County, and parts of Sheridan, Johnson, Big Horn, Washakie, Hot Springs, Park, Teton, Fremont and Sublette Counties as illustrated in Figure 1.

This report is part of a set of deliverables for the Wyoming State Office of the Bureau of Land Management as required by work orders K-910-P970194 and K-910-P970206. Other deliverables include the delineations as GIS coverages, a hard copy map and slides. Pertinent portions of the description of the work as set forth in these work orders include the following:

"...delineate landscape complexes ("landtype associations" sensu Bailey (1980) and higher order regional units for the BRA in the counties listed..." (K-910-P970194)

"...delineate landscape complexes ("landtype associations" sensu Bailey (1980) and higher order regional units for the northwestern quarter of Wyoming in the counties listed..." (K-910-P970206)

"Work performed under this contract will be "edge-matched" with completed landtype delineation work..." that was produced in earlier contracts with the BLM by the University of Wyoming.

"UW shall, insofar as practicable and possible, include and incorporate available NWWY US Forest Service (FS) ecological mapping data at the landtype association level into the NWWY project final report and mapping. This proviso is intended to make the NWWY final delineation report as comprehensive as possible without changing mapping philosophy and procedures UW has used in prior delineation projects for BLM."

In initial discussions regarding joining USFS map products, BLM and UW participants agreed that comparable National Park Service (NPS) products from Grand Teton National Park and Yellowstone National Park would also be included. UW personnel did communicate with relevant NPS personnel about this possibility but learned that such products, though in production, are not yet available.

The BLM and UW mutually agreed in a meeting on 21 May 1999 that the results for both the NWWY and Buffalo Resource Area work would be joined together in a single report. This decision, along with details on modifications of deliverables was formally authorized in an undated letter from Charles Padilla received some time after that meeting.

As a result of these modifications, this is a combined report for both northwestern Wyoming and the revised Buffalo Resource Area. This report refers to associated deliverables and to USFS map products that were available at the time of this writing.

This work is a continuation of related work with the Wyoming State Office of the Bureau of Land Management in which we delimited land units in northeastern, southwestern and southeastern Wyoming (Reiners and Thurston 1994, 1995, 1996, 1997). The products described in this report, together with the incorporation of land unit delineations provided by the various national forests and national parks of Wyoming, is meant to complete Landtype Association (LTA) mapping for the entire state.

The goal of this and our previous work for BLM has been to map terrain features having similar terrain configuration, ecological properties and management implications within which situational landscape-ecosystems can logically and conveniently be defined (Anonymous 1994). Within the formal structure of the ECOMAP (1993) national hierarchy of ecological units, we have mapped Landtype Associations as representing a practical means for delineating landscapes, or at least portions of terrain that can be decomposed into smaller landscapes having highly similar properties such as topography, vegetation and soil patterns. This is a terrestrial-centered approach--dividing terrain into map units in terms of terrain features rather than by hydrological units. This philosophy is consistent with the approach recommended by Godfrey and Cleaves (1991) and ECOMAP, and in contradistinction with an alternative philosophy in which drainage basins are the defining landscape unit (Omernik and Bailey 1997, Lotspeich 1998).

In previous work (Reiners and Thurston 1994, 1995, 1996, 1997), we took a subjective approach to delineating Landtype Associations (LTAs) in Wyoming. By using a variety of coverages, including 3 arc second and 30 m resolution digital elevation model data, geology, landcover, among others, we interactively digitized Landtype Associations. As land form was our primary criterion, relief was usually the final determinant of where delineations were ultimately digitized. Thus, each boundary between LTAs and higher order units was based on subjective opinion applied by different criteria along the boundary line. This approach has been termed by Bailey (1996) the "method of controlling variables," meaning that different variables dictating local delineation were the most prominent in different parts of a map or even along a single boundary line.

While we were very confident of the reality of the units previously delineated, we became increasingly concerned over two issues: the subjectivity of the process and the extent of the units we delineated. Typically, our LTAs ranged from about 750 acres to over 1,240,00 acres in the southeast. This is larger than the 100 to 1000s of acres suggested by ECOMAP (1993) and in many cases, considerably larger than LTAs mapped within national forests adjacent to our map area. We expected that terrain units would have larger extents in basins compared with mountains because of the greater simplicity of underlying geology and the character of dominant geomorphological processes occurring in basins. On the other hand, we were not satisfied with our previous differentiation of land forms within these basins and looked for a means to better tease apart variations in roughness and form.

Others have used statistical methods to classify land forms, usually involving several variables (Bourgeron et al. 1994, Davis and Dozier 1990, Host et al. 1996, Mann and Benwell 1996, Capen et al. 1999). We believe that the different variables used in these systems vary at different scales and therefore are not likely to bound at the same places or to lead to definition of integral systems. For example, vegetation varies at a smaller scale than do the terrain features scaled to fit ECOMAP guidelines for LTA sizes, at least in the landscapes mapped in this project. Such multivariate classification approaches depend on the assumption that ecosystem properties are co-located in landscapes. We do not subscribe to that assumption. We contend that such methods probably tend to "smooth over" prominent, differentiating features in local cases. Incorporating multiple variables leads to compromises in polygon boundaries. It is possible that those overall products make general scientific sense, but we question whether resultant polygon boundaries make sense at local scales of approximately hundreds to thousands of meters. Making sense in local cases is extremely important to our and BLM’s objectives. LTAs must be recognizable on the ground as well as on a map. In fact, in deciding whether geographic features should be mapped as an LTA, we asked ourselves the practical question:

This self-testing query forces us to see the landscape from the manager’s or operator’s points of view--on the ground. It also drives us toward use of terrain features as the primary determinant of LTAs. Terrain features may be based on underlying lithology or tectonics but more frequently they are products of fluvial geomorphological processes operating on the poorly consolidated materials filling Wyoming’s basins and plains.

In this project, we sought an approach for delineating LTAs other than multivariate classification for reasons given above. We wished to reduce subjectivity and to better partition different land forms in what we had formerly combined together as "rolling plains" typical of Wyoming’s basins. We sought to create smaller LTA units thereby creating a finer grain to the overall map. Terrain characterization systems exist, usually based on GIS processes (Fels and Matson 1996, Blaszczynski 1997, Medler and Yool 1998). We adopted an approach pioneered by Hammond (1964), who developed a map of "land-surface forms" of the United States from an empirical analysis of the land surface. Hammond (1964, 1970) used a selected group of surface characteristics as the basis of a simple classification scheme. Hammond’s criteria for choosing these characteristics were that characteristics should:

"1) be especially effective in conveying a visual image of surface form;

Hammond’s characteristics included:

We found Hammond’s system fit our scale of landform resolution reasonably well, and still allowed us a degree of latitude for subjective interpretation that we still found desirable. We modified his approach as is described in Methods (Section 3).

As in past reports, the hierarchical system containing our delineated units follows the structure of the ECOMAP national hierarchy of ecological units (ECOMAP 1993). We have mapped Landtype Associations (LTAs) as well as higher hierarchical units in which LTAs are imbedded. LTAs are defined by ECOMAP (1993) as units at the "landscape scale" as follows:

In addition to mapping LTAs , we have also mapped higher hierarchical levels of the National Hierarchy of Ecological Units for the defined map area. In this process, we have defined land units up to four hierarchical levels above the LTA. These definitions of higher hierarchical units usually conform with the nationally accepted nomenclature, but often differ with respect to location. Our rationale for delineations of higher hierarchical units are given in section 4 of this report.

While the principal objective of this work has been to produce a digital map of landscape-ecosystems for management purposes, the GIS coverages generated to do this, and provided along with the map as part of our deliverables, have a value in themselves. These coverages include information that can be used as decision support tools for regional planning and cumulative impact assessment. An array of accurate spatial data coverages can be used to provide distance, area, adjacency and proximity relations for features in a region of interest; GIS coverages can permit geographic analysis for queries and planning through joining and unioning coverages of thematic features for rapid assessment. For example, location of a facility with respect to bodies of water, geologic formations, or vegetation types can be done efficiently and accurately with GIS techniques. While we have not provided political boundary, road and other human-centered thematic data with this work, such data are readily available and can be added to the coverages delivered in this product to make queries and management decisions. For example, another data set that would be extremely valuable for land use managers would be the game range coverages undergoing refinement within the Wyoming Game and Fish Department.

The LTAs themselves have collective properties that are relevant to management decisions. Occurrences of a particular LTA should trigger an associated set of considerations appropriate to that LTA type. For example, LTAs in the glacial till and outwash hills class should signal consideration of swell and swale topography, scattered boulders, ice block ponds and wetlands. Alternatively, LTAs in the multiple cuesta and valley class signal the occurrence of linear, rocky outcrops interspersed with parallel valleys with deeper soils. In both cases, LTA class properties have relevance for managing wildlife, livestock and domestic or industrial development. Because LTAs have been mapped by consistent criteria across most of the state, understanding of one member of an LTA type can be transferred, with reparameterization, to another member of that type.

The hierarchical nature of this digital map product permits managers to scale up or down according to the scope of their concern. For example, managers may find that the Section or Subsection level of delineation is useful for strategic planning, whereas the LTA is most useful for evaluating specific projects.

Inasmuch as LTAs integrate basic environmental features (geology, geomorphology and geomorphic processes, soils, vegetation, wildlife, and habitats of rare and endangered species) in ways unique to that type of LTA, it may make sense for field data to be collected with respect to these map units, rather than by more artificial map units such as counties. For example, changes in management practices or aspects of the environment may produce trends in ecological properties that are characteristic of LTAs rather than politically or administratively defined units.

Managers are highly preoccupied with making judgments about resource use in the contemporary situation. In some cases, it may become mandatory to consider management tactics and rules in the context of possible climate change (e.g. Changnon and Demissie 1996). This certainly is the case for national park managers. The LTAs should be very valuable units for considering the implications of such changes.

In order for managers to make full use of LTA delineations, it will be necessary for them to understand what particular classes of LTAs represent in terms of patterns of environmental attributes, or simply the presence or absence of attributes. Delineating the LTAs will not be sufficient in itself. Much analysis of LTA types is needed to create an atlas of properties, both qualitative and quantitative.

In this and previous mapping projects for the BLM (Reiners and Thurston 1994, 1995, 1996, 1997), we have created a set of Landtype Association classes based as nearly as possible on geomorphic properties. Similarly, we have attempted to use names for these classes that match proper geomorphological terminology. The following are brief descriptions of the LTAs mapped in this work. Those LTAs that are based strictly on properties of slope and relief include the Hammond classification code in parentheses next to the LTA name.

Granite hills: A feature having mountain properties (according to Fairbridge (1968), a mountain is "a natural elevation of the earth surface rising more or less abruptly from the surrounding level, and attaining an altitude which, relatively to adjacent elevation, is impressive or notable") but of lesser relative relief (< 700 m) composed of crystalline metamorphic or magmatic rock. Some geomorphologists might describe these interesting hills as "bornhardts" (Fairbridge 1968).

Footslope: Many, but not all mountains have steep, upper slopes composed of bedrock above lower, less steep slopes of either weaker rocks, pediments or fans. Typically, there is a distinct break in slope between the steep upper portion and the less steep lower portion. Often the lower slope cantilevers gently into plains or basins away from the mountain. While the upper break in slope is relatively easy to delineate, often the delineation of the lower boundary of the lower slope is arbitrary. Because the lithology, hydrology, slope stability, soils and vegetation are so different between the upper and lower slopes, we have recognized a class of Landtype Association we term the "footslope." We recognize that these "footslopes" may be classified as pediments, benches, straths and fans, but without stratigraphic data and without broader consensus of these land forms, we are using the generic term "footslope." Generally, the base of the footslope was delineated at roughly the 4° break in slope, and the delineation between mountain and footslope at the 10° break in slope.

Breaks: This term is common in the American west but we have been unable to find a technical definition. Based on terrain designated as "breaks" on Wyoming maps, it appears that "breaks" most commonly refer to easily erodible, more or less horizontally bedded sedimentary materials that are highly incised by deeply cut, dendritic channel systems. Breaks often have more or less flat interfluves cut by a dense network of rapidly eroding, mostly ephemeral channels. This type is most often found in poorly consolidated Tertiary materials in the vicinities of rivers where base leveling is most advanced.

Mesa(s): ". . . flat-topped hills and mountains, cut off on one or more sides by steep escarpments or "breakaways" (Fairbridge 1968). Mesas mapped in this work are of "hill" proportions. In a classification based strictly on slope and relief, mesas would be divided into a flat plain surrounded by a steep escarpment. These two slope features are grouped together here to form a recognized landform.

Recessional escarpment: Steep, erosion fronts, often relatively linear in plan view, on poorly consolidated, sedimentary materials. Recessional escarpments often form prominent "rims"-- distinct boundaries occurring between lowlands excavated by tributary networks of major rivers and high, often flat or gently rolling uplands relatively untouched by the stream grade of the adjacent lowland. Headward erosion along the escarpment face is typically rapid and may be accompanied by badlands landforms.

Multiple cuesta and valley complex: As a complex, this is our own term. We refer here to a series of long narrow, more or less parallel ridges separated by valleys as wide or wider than the ridges. Generally, these ridges and valleys are composed of dipping alternating beds of relatively harder or softer rocks, respectively. Technically, if the ridges are steeper than 40-45°, they are termed "hogbacks;" ridges dipping from 5-40° are sometimes called homoclinal ridges (Fairbridge 1968). We use the less-well defined term "cuesta" for both types of ridges in this report.

Single cuesta: A single long, narrow ridge, either a hogback or monoclinal ridge (see above). Single cuestas often have interrupted, stairstep form to the strike slopes but are separated from multiple cuesta valleys by the absence of valleys between the resistant layers.

Glacial till and outwash hills: Low hills and dissected fans associated with terminal moraines and outwash adjacent to mountains. Drainage on these hills is often immature and the terrain can be dotted with wetlands, ponds and lakes. These are determined by the presence of glacial deposits on the bedrock geology coverage.

Low mountains (D5): Less than 20 percent of the area has a slope of less than 5% and the local relief is between 1000 and 3000 feet. This is a more specifically defined form of a "hill" as defined by Fairbridge (1968) as under "mountains" above.

High hills (D4): Less than 20 percent of the area has a slope of less than 5% and the local relief is between 500 and 1000 feet.

Hills (D3): Less than 20 percent of the area has a slope of less than 5% and the local relief is between 300 and 500 feet.

Low hills (D2): Less than 20 percent of the area has a slope of less than 5% and the local relief is between 100 and 300 feet.

Open high hills (C4): 20 to 50 percent of the area has a slope of less than 5% and the local relief is between 500 and 1000 feet.

Open hills (C3): 20 to 50 percent of the area has a slope of less than 5% and the local relief is between 300 and 500 feet.

Open low hills (C2): 20 to 50 percent of the area has a slope of less than 5% and the local relief is between 100 and 300 feet.

Irregular plains with hills (B3): 50 to 80% of the area has a slope of less than 5% and the local relief is between 300 and 500 feet.

Irregular plains (B2): 50 to 80% of the area has a slope of less than 5% and the local relief is between 100 and 300 feet.

Plains with hills (A3): More than 80% of the area has a slope of less than 5% and the local relief is between 300 and 500 feet.

Smooth plains (A2): More than 80% of the area has a slope of less than 5% and the local relief is between 100 and 300 feet.

Flat plains (A1): More than 80% of the area has a slope of less than 5% and the local relief is less than 100 feet.

Alluvial valley: These are river valleys containing a third order stream or higher and also contain mapped alluvium.

River valley: These are valleys with a distinct river channel, with a third order stream or higher, but do not contain mapped alluvium.

Arroyo/draw: These concave features are deeply incised drainage channels similar to those described under "breaks" but more or less isolated in their occurance on flat or rolling terrain. These are similar to "gullies" as described by Fairbridge (1968) but are not necessarily products of human influence or climate change.

Lake/reservoir: Lakes are bodies of water confined in basins that are or were natural but may today be regulated by dams and flumes. Reservoirs are bodies of water formed by damming naturally occurring running waters--rivers and streams.

In addition to our Landtype Association coverages, we are also providing coverages for the Bridger-Teton, Shoshone, and Bighorn National Forests. All three National Forests have used different criteria for mapping Landtype Associations, so we have not attempted to edge match across National Forest Boundaries. National Forest LTAs are shown in Figure 32, Figure 33 and Figure 34. Landtype Association types are explained in Tables 8, 9, and 10 of the Appendix.

We created the following GIS layers in Arc/Info format for northwest Wyoming and the Buffalo Resource Area:

Although the above coverages were all used interactively to identify LTAs, polygons were delineated primarily using a classification system of slope and relief based on the concepts of Hammond (1964). This required production of slope and relief coverages from 60 m DEM layers. The slope coverage consists of a neighborhood slope value for every 60 m cell. The neighborhood was defined by a circular window 13 DEM cells in radius for a total area of 1,911 ha. The window was positioned over every cell in the 60 m DEM coverage and the percent of area in the window having a slope of 5% or less calculated--the slope value for each cell. The resulting continuous slope values were broken into four slope classes as defined inTable 1.

Slope Class
A	More than 80 % of area is gently sloping
B	50 - 80 % of area is gently sloping
C	20 - 50 % of area is gently sloping
D	Less than 20 % of area is gently sloping

Note: "Gently sloping" is defined as a slope of 5 % or less.

The relief classification was created by calculating the maximum and minimum elevation changes occurring in the neighborhood surrounding the each cell using the Arc/Info Grid function FOCALRANGE. Continuous values for relief were divided into 6 relief classes as shown in Table 2.

The choice of 60 m DEM resolution and the size of the circular neighborhood window were based upon a series of trials. The neighborhood extent chosen seemed to afford optimal compromises between representing the grain of the terrain in most of the classified area, and data management efficiency.

Our methods involved several modifications of Hammond’s (1964) original method. He used a unit area of 6 square miles to classify land-surface forms, and a minimum mapping unit of 800 square miles. We reduced the window size, or neighborhood, to 13 cells, or approximately 1,911 ha and the minimum mapping unit from 800 mi² to 500 ha (1.9 mi²) to fit the finer mapping scale we desired for the terrain in question. Furthermore, we redefined "gentle inclination" as 5% instead of 8%, as we thought this better discriminated sub-elements of the rolling plains landscapes characteristic of the basins we mapped. Finally, we chose not to use Hammond’s third criteria (the percent of gently inclined surface in the lower half of relief), as this characteristic was used to define "tablelands" and "canyonlands." Because of our smaller neighborhood, hills and canyons were delineated as individual Landtype Associations, rather than being included in broad areas that had hills (tablelands) or canyons (canyonlands).

The overlay of slope and relief produced clusters of cells having similar combinations of these properties. The outlines of these clusters were digitized by hand, taking into account other coverages (shaded relief, hydrology and geology). It was at this point that we switched from a wholly objective method to a subjective one by making decisions leading to the exact boundaries of units delineated primarily by the Hammond criteria. Additionally, several important Landtype Associations that might have been split into multiple polygons using the Hammond classification were identified and digitized. These included multiple cuesta valley complexes and mesas, both of which tended to be broken into slope and plain components using the Hammond classification. Also included were glacial till and outwash hills and alluvial valleys, both of which were defined by their geological components.

After digitizing all LTA polygons, they were then classified using the modified Hammond criteria listed in Tables 1 and 2. Slope classification was done using a circular neighborhood of 3 cells rather than the 13-cell neighborhood used to guide the digitizing. This classification gave better resolution for flat polygons that were bordered by steeply sloping terrain, such as lakes. LTA codes were then assigned to each polygon. Many of these codes are based strictly on the slope and relief classification as prescribed by Hammond. However, some polygons were assigned to other LTA classes defined by recognizable features or other properties such as geology or presence of 3rd order or higher streams (see section 3.2).

For this report, our minimum mapping unit (MMU) guideline was 500 hectares (1,236 acres). Because we were constrained by adjacent US Forest Service boundaries and previous mapping (we maintained Scoria polygons, for example, in the Buffalo Resource Area), we mapped some Landtype Associations that were significantly smaller than our MMU. LTA polygons ranged from about 4.8 ha (12 acres) to over 226,316 ha (559,000 acres). Smaller polygons almost always occurred along the margins of the map areas, often where division, province, section or subsection boundaries intersected USFS boundaries. The mean polygon area in the northwest Wyoming coverage is approximately 12,200 acres, and in the Buffalo Resource Area is 11,740 acres. In general, LTA polygons were considerably smaller than polygons delineated in previous reports. For example, the mean polygon area we delineated in the southeast was approximately 101,700 acres.

The classified LTA polygon coverage was finally intersected with the landcover, soils, geology, surficial geology, precipitation and elevation coverages, and individual polygons attributed according to the properties of each intersected coverage. The result of these efforts are the Landtype Association coverages in Arc/Info format described in section 5.2. In addition, the arcs of the Landtype Association coverage have been attributed to indicate the source data and digitizing criteria (see section 5.4) .

This work benefited from numerous telephone and e-mail conversations with BLM personnel. In addition, individual meetings were held with BLM personnel on the following occasions.

21-23 October 1997 Rob Thurston and Bill Reiners attended an interagency workshop on ecological mapping held in Casper, WY. Reiners presented UW’s approach to modeling to this group and Thurston interacted throughout the meeting on seeking a common basis for mapping criteria and scale.

27-29 January 1998 Rob Thurston and Bill Reiners attended the Interagency Ecological Mapping Meeting held in Cheyenne, WY. This was a follow-up meeting to the previous Casper meeting at which representatives from involved agencies including BLM presented their map products and discussed development of a common product.

15 September 1998 Bill Daniels and Jon Johnson met with Bill Reiners and Rob Thurston in Laramie to review the status of the project and developing map products.

21 May 1999 A meeting between Jon Johnson, Jeff Carroll, Ellen Axtmann and Bill Reiners was held at the Dept. of Botany in Laramie during which progress was demonstrated and reviewed, changes in the product were described, and contract deliverables were defined. The outcome of this meeting is covered in a letter from Reiners to Johnson dated 27 May 1999.

Section 4 of this report contains rationales and descriptions of delineated land units at hierarchical levels above the Landtype Association level. Because of the finer scale mapping used in this report relative to previous reports and the large number of Landtype Associations, individual Landtype Association polygons are not described. Table 1 and Table 2 of the Appendix list all Landtype Association types and aggregated areas of each organized according to an ECOMAP-type hierarchy.

"Domain" is the highest hierarchical level in the US Forest Service Ecoregion system (Bailey 1995). The Dry Domain is characterized by potential evapotranspiration being greater than precipitation (Bailey 1995) and covers the entire state of Wyoming, although this condition does not prevail in parts of the mountains of the state. The Dry Domain is coded as "300" within this system so that all mapped units within Wyoming have a "3" as a prefix in their numerical notation. This hierarchical unit is important at the continental scale but is not relevant for purposes of ecosystem management in an area the size of Wyoming.

Within the "Dry Domain" of the US occur seven "Divisions," which are also defined primarily in terms of macroclimate. Three Divisions occur in this map area (Figure 24 and Figure 28):

a) Temperate Steppe,

b) Temperate Steppe Regime Mountains, and

c) Temperate Desert Divisions.

The Temperate Steppe Division (330) includes areas "with a semiarid continental climatic regime in which, despite maximum summer rainfall, evaporation usually exceeds precipitation" (Bailey 1995).

The Temperate Steppe Regime Mountains Division (M330) is characterized by the same macroclimate as the Temperate Steppe Division but having changes in temperature and precipitation with changes in elevation. This Division encompasses all the mountain systems of the current map area.

The Temperate Desert Division (340) is characterized as areas of low rainfall but strong contrasts in seasonal temperatures (Bailey 1995).

The next lower hierarchical level in the Ecoregion system is the "Province" defined by somewhat more specified macroclimatic properties. Within the area of the current map work, the Province boundaries happen to coincide with Divisional boundaries (Figure 24 and Figure 28). Thus, Provinces are the same as Divisions within the map area presented here. The single Province occurring in the Temperate Steppe Division is the "Great Plains-Palouse Dry Steppe" (331). According to Bailey (1995), it is

In the part of the Great Plains-Palouse Dry Steppe Province mapped in this map area, the dominant parent materials are weakly consolidated, Tertiary sediments (Lillegraven 1993) (Figure 5) although extensive beds of clinker (scoria) occur where coal bed fires have baked overlying materials into more indurated state. The predominating soils on unmodified Tertiary materials are Typic Torrifluvents, Typic Haplocambids, Ustic Haplargids, Ustic Haplocalcids and Aridic Haplustolls. In the clinker zones, soils are Typic Torriorthents and Rock Outcrop (Munn and Arenson 1998). Vegetative cover over this province is a complex mixture of mixed grass prairie interfingering with Wyoming big sagebrush. Some of the grassland is now in dryland agriculture. Particularly throughout the scoria-covered areas, ponderosa pine groves occur. For more details see Knight (1994) Merrill et al. (1996) and Driese et al. 1997.

The mapped boundary for this Province is set by the state borders with South Dakota and Nebraska to the east and Montana to the north, by the bases of the Black Hills, Laramie Range and Bighorn mountain systems, and by the irregular line marking the contact with the Intermountain Semidesert Province to the west (Figure 28). The rationale for the delineation between the plains and mountains is based on change in relief at the bases of the mountains. The rationale for the delineation between the Great Plains-Palouse Dry Steppe on the east and the Intermountain Semi-desert Province on the west was discussed in a previous report (Reiners and Thurston, 1997). A general geological discussion of this Province is provided by Osterkamp et al. (1987).

The sole Province within the Temperate Steppe Regime Mountains Division within this map area is the "Southern Rocky Mountain Steppe-Open Woodland-Coniferous Forest-Alpine Meadow Province (M331)" (Figure 24 and Figure 28). This is an area of

A general description of the geology and historical processes associated with this Province can be read in Madole et al. (1987).

With the exception of the Absaroka Range, the mountains in the map area are Laramide features (Brown 1993). Typically, these have arisen from deep-seated faulting and are cored by Precambrian rocks including granites, gneisses, other metamorphics, and limited areas of mafic outcrops. Small to extensive areas of Paleozoic sedimentary rock occur in our delineations of this Province. Among soils, Rock Outcrop is a common type in all of the mountains. Other soils of the Absaroka Range are mapped as Typic Cryorthents, Humic Dystrocryepts and Histic Cryaquepts. Other ranges in this province also have Typic Haplocryalfs, Typic Dystrocryepts, Typic Haplocryolls, and Lithic Cryorthents (Munn and Arneson 1998). Vegetation varies with elevation and slope aspect, ranging from mixed shrubs and grasses in the low foothills, to lodgepole pine, spruce-fir and alpine vegetation with increasing elevation (Knight 1994).

The single Province occurring within the Temperate Desert Division within this map area is the "Intermountain Semi-desert Province (342)."

In our map area, this Province has widely varying environments ranging from cool, moist mountains to rocky ridges to rolling plains, alluvial fans and playas. The small mountains in this Province are composed of Precambrian igneous and metamorphic rocks and Paleozoic sedimentary rocks. The ridge- and valley-forming rocks are usually Mesozoic sediments, mostly of Cretaceous age, and the plains are primarily composed of Cretaceous shales or poorly consolidated Tertiary sediments. Soils of the mountains are generally classified as Typic Hapludolls, Typic Hapludalfs and Rock outcrops. Soils of the cuestas are generally classified as Torriorthents. Dominant soils of the basins are Typic Haplargids, Typic Haplocalcids, Typic Natrargids (Munn and Arneson 1998). While the small mountains may be forested, mainly in lodgepole pine, the cuestas are often shrub-dominated as described by Knight (1994) for "escarpments and the foothill transition." The basins fall into his description of "sagebrush steppe" and "desert shrublands and playas." According to Merrill et al. (1996), dominant land cover types of the basins include greasewood, saltbush, mixed desert shrub, Wyoming big sagebrush with smaller amounts of black sagebrush and basin big sagebrush.

The Northwestern Great Plains Section is described but not delineated in McNab and Avers (1994). They say:

Our delineation of this Section is given in Figure 29.

Descriptions of the Subsections of this Section have not been published although some tentative Subsection delineations have been made on a map distributed by Freeouf (1996). Our Subsections (Figure 30) do not fully conform to theirs in name or organization, and definitely not in location of the delineating boundaries. We have rejected recognition of Powder River Basin as a subsection (Freeouf 1996). We have found structural basins in the eastern third of the state to not have much topographic expression and therefore, not have much practical use for ecoregion delineation. This is particularly true of the Power River Basin which doesn’t even have any hydrological control for this area. We recognize seven subsections in this entire section (Figure 30), three of which occur in this current mapping area.

In previous reports (Reiners and Thurston, 1994, 1995) we delineated Scoria Hills as a separate subsection. The Scoria Hills had been digitized from a map of scoria, or clinker, prepared by Mr. Ed Heffern (Heffern and Coates, 1997) but on the 30 M DEM it is clear that landscape features associated with mesas, scarps, talus slopes and fans go beyond the limit of the mapped scoria. Because the Hammond classification scheme we used did not conform readily to the previously mapped Scoria Hills, we have included Scoria or clinker as a polygon attribute for both the Upland Plains and Powder River Breaks Subsections as opposed to being a separate subsection in itself. Areas consisting of clinker have Typic Torriorthent and Rock Outcrop soils (Figure 9). Vegetation contains many of the same elements as for the other subsections but includes some xeric shrub (mainly mountain mahogany) and ponderosa pine (Figure 23).

4.4.1.1. Casper Arch Subsection (331Fa)

This Subsection is well delineated by a tectonic feature known geologically as the Casper Arch (Figure 30). This is a Laramide rise or anticline extending broadly from the northern end of the Laramie Range northwestward to the southeastern corner of the Bighorn Mountains (Figure 5), or alternatively, as a "sag" between the Laramie Range and Bighorn Mountains (Lageson and Spearing 1988). The northeastern and southwestern borders of this arch are bounded by deformed Cretaceous formations exposed at the surface as low to medium dip homoclinal ridges. Within the Arch occur some younger cuestas but much of the interior is occupied by a soft shale which has been eroded into gently rolling plains except for some incipient "breaks" near the rivers. Major soils of the eastern part of this subsection are Typic Torrifluvents, Typic Haplocambids, Ustic Haplargids, Ustic Haplocalcids and Aridic Haplostolls; soils of the western part of the subsection are Typic Hapludolls and Typic Hapludalfs (Figure 9). Primary vegetation throughout the subsection is a coarse-grained mixture Wyoming big sagebrush and mixed grass prairie, the type depending mainly on parent material.

Both the eastern and western margins of this subsection were digitized along a slope contour of roughly 10º. This subsection consists mainly of Mesozoic and Paleozoic limestone, dolomite and sandstone, although it include some plutonic rocks in the north, and gneissic rocks in the south (Figure 4 and Figure 5). Deep, steeply walled canyons cut through the dipping sedimentary rocks, particularly along the eastern flank. Rock Outcrop and Lithic Cryorthents are the most common soils (Figure 8 and Figure 9). Mixed grass prairie is mapped for much of this area although it has a distinct mountain meadow character associated with the higher elevations. Juniper woodland is found at lower elevations on rocky outcrops and Douglas-fir in the canyons (Figure 22 and Figure 23).

The majority of this subsection is included in the Bighorn National Forest LTA coverage (Figure 33). We have mapped 9 polygons in the northwest coverage within this subsection, seven of which are designated as "low mountains," one is "high hills," and one is "irregular plains."