Columbia Basin Ecological Province





The Columbia Basin Province in north-central Oregon includes the lower portion, generally below about 3,500 feet elevation, of several major watersheds that drain north into the Columbia River. It covers about 3.25 million acres and includes northwestern Umatilla County, the northern two-thirds of Morrow County, all of Gilliam County except for a very small area in the southeast corner, all of Sherman County, much of northeast and southeast Wasco County, a small area in northwest Wheeler County, and about 2,500 acres west of Willowdale in northern Jefferson County.

Milton-Freewater is in the northeast corner of the province; The Dalles is just west (outside) the northwest corner of the province in Oregon. This province is bisected by lower reaches of six major drainages: Deschutes River, John Day River, Rock Creek, Willow Creek, Butter Creek, and Umatilla River, all of which flow generally north. This province extends into Washington.



Elevations within Columbia Basin Province in Oregon range from about 100 feet near The Dalles to about 3,500 feet along the line of demarcation between Columbia Basin and Blue Mountain provinces. The physiography is mainly a hilly upland sloping up from north to south and dissected by numerous dendritic-pattern drainages. A sizable sandy basin lies west of the Hermiston area. There are no prominent mountains in the province in Oregon.

Columbia Basin Province in Oregon has two physiographic subdivisions which have significant ecological and management implications. These are the ancient lake basin, which is largely irrigated agriculture except on the Boardman Bombing Range Reserve, and the silty uplands, which are dryland agriculture and native rangelands.

The smaller of the two subdivisions is the ancient lake basin. Locally, this area is commonly called the Umatilla Basin, probably because early irrigation in the vicinity of Echo, Stanfield, Hermiston, Umatilla, Irrigon, and Boardman depended on water from Umatilla River stored in Cold Springs Reservoir west of Hermiston. More recently, pumping from wells and from the Columbia River has greatly expanded irrigation agriculture in the lake basin and likely would have expanded it a great deal more if it hadn’t been for the huge Boardman Bombing Range Reserve. The lake basin subdivision of Columbia Basin Province covers about 450,000 acres in Oregon.

The lake basin, with its underlying strata of gravel beds, hardpans, and other materials, is geologically related to the era of glacial melt farther north following the ice age. Geologists have reported that the glacial melt resulted in Missoula flood(s); the ice jam near The Dalles that backed up water (Lake Condon); and the ice floes containing sand, silt, gravel, and other glacial debris and erratics. When the floes melted, these erratics and debris were deposited over the landscape in the lake basin. Skeletal remains of mammoths and other ice-age artifacts have been uncovered in the lake basin.

The silty uplands portion of Columbia Basin Province is almost entirely dryland agriculture and rangeland. Nearly all arable acreage is being or has been farmed at one time, primarily for wheat production in a wheat summer-fallow alternate-year rotation (Fig. 13). All the silty upland rangeland is considered a natural grassland (less than 10% canopy cover of woody species in original ecological status).

Figure 13: Overview of extensive dryland farming area within Columbia Basin Province, Oregon. Washington state, north of the Columbia River, is in the background.

Throughout the natural grasslands that have not been cultivated in Columbia Basin Province in Oregon, silty grassland soils on most plateaus and ridges between about 1,700 and 3,500 feet elevation occur in a unique land pattern locally called “biscuit scabland.” The pattern consists of small mounds of grassland soils 5 to 20 or more feet in diameter and usually about 20 to 36 inches deep over basalt bedrock. Each mound, or biscuit, is surrounded by very shallow, very stony soils over basalt bedrock, which is locally called “scabland.” Hence the name “biscuit scabland” (Fig. 14).

Figure 14: Biscuit scabland is extensive in Columbia Basin Province, Oregon

The biscuits vary somewhat in shape and size, usually round but also oblong. They occupy about 5 to 30% or more of the area on which they occur. Where biscuits occupied more than about 40 to 50% of the area, some have been farmed, and cultivation has redistributed the soil mantle to the point that it is essentially continuous but of varying depth—deeper on former biscuits, shallower over what used to be scabland, with an undulate surface.

Biscuit scabland in Columbia Basin Province is common from just west of Pilot Rock in Umatilla County to Dufur in Wasco County; however, the soil of the biscuit component of this pattern changes from location to location. Soil series on biscuits include Condon, Morrow, Valby, Wapinitia, and Maupin silt loams; Condon series is the most extensive. The soil on the scabland component is Bakeoven. Biscuits do not occur generally below 1,700 feet elevation in aeolian soils such as Ritzville and Walla Walla silt loams.

During the past 150 years, more than 30 hypotheses have been advanced to explain the origin of biscuit scabland,according to John Eliot Allen, emeritus professor of geology at Portland State University. The possible reasons include moles, gophers, Indian burials, hut sites, buffalo wallows, ant hills, and mounds formed around water, gas, oil, or mud springs. Allen believed four of these were reasonable: erosion, during which the mounds were protected by vegetation; wind deposition, which posits that the mounds are coppice dunes formed around trees or shrubs; freezing and thawing, which, in the far north, produces polygonal ground on tundra; and gophers’ tunneling backward from their nest sites which, over the years, gradually built up the area around the nest (this concept was based on a situation in California). Undoubtedly, all such mounds are not alike, even though they look alike on the surface. Also, all four of Allen’s choices may apply somewhere but, conceptually, do not apply to the biscuit scablands of Columbia Basin Province in Oregon.

Here are some facts about Oregon’s biscuit scabland.

First, numerous roadcuts throughout the biscuit scabland in Columbia Basin Province prove that the material underlying the biscuits is solid, very thick basalt (Fig. 15). That fact alone discredits some hypotheses cited above, i.e., gophers, ants, trees, and freezing–thawing polygons.

Figure 15: A cross-section of a typical biscuit scabland in Columbia Basin Province, Oregon. It shows the massive basalt underlying both the biscuits of soil and the adjacent scablands.

Second, the soil profile of each Columbia Basin biscuit in any general area has essentially, if not exactly, the same sequence of horizons, colors, textures, and depth ranges as nearby continuous mantles of that soil. They are the same soil series whether in biscuits or in continuous mantle. This suggests that the geomorphology—the source of parent material and the conditions under which the soil was formed—is the same for the biscuits as for the soil mantle.

One previously unpublished concept of the origination of biscuit scabland is based on much personal experience with the biscuit scabland complex in the Columbia Basin and Blue Mountain provinces in Oregon and on various published articles, one of which lists 23 topical references.18

The concept originated with a point made in one of these articles about the era of glaciation in North America. The article dealt mainly with glaciers; however, it also noted that there were “sympathetic” ice sheets on areas at higher elevation south of the continental glaciers. This is credible and forms the basis for the following concept of how biscuit scabland originated.

Glaciers lay north of Oregon. Biscuit scabland is on ridges and plateaus between 1,700 and 3,500 feet in Columbia Basin Province and at about 3,500 to 4,500 feet elevation in Blue Mountain Province. Sympathetic ice sheets likely covered some of these higher areas at the time glaciers were receding. Elevations above 4,500 feet and other, lower areas could have been covered with thick ice or even local glaciers.

When a layer of ice melts, it melts uniformly over the surface but also melts faster in some spots than in others to form circular or elongated depressions which eventually become holes in the ice.

As glaciers were receding north, the prevailing winds were from the cold, glacial area to the warm, equatorial area, i.e., from north to south generally. As glaciers receded, glacial debris was exposed. Prevailing winds blew dust south where it was deposited to become the vast aeolian silty soils that typify the upland soils of Columbia Basin and Blue Mountain provinces.

Aeolian soil material was deposited as sympathetic ice sheets melted. Conceivably, the aeolian deposits on the ice washed into and collected in holes where the ice was melting. These collections of silt became mounds as the ice sheet melted. The final ice melt on the sympathetic ice sheet was where scabland now exists between the mounds.

Obviously, erosion from runoff would remove silts from the area that is now scabland as well as tend to round the top of each biscuit and slope the sides. Elongated biscuits are oriented generally with the slope of the land, and lines of biscuits oriented with the slope of the land are common. This might indicate that the final melt worked its way downslope in a concentrated stream, thereby flushing off the silt to form a linear pattern of scabland. This water action might also help explain why scabland without biscuits exists primarily along outer edges of ridges and on sloping areas in minor drainages where water had concentrated to wash the silt off basalt bedrock.

Some biscuits are deeper than others, which might be related to the thickness of the ice sheet—holes would have more time to fill with silt where the ice sheet was fairly thick and the process of melting was prolonged. Some areas do not have biscuit scabland but do have a solid mantle of silty soils. This may indicate the absence of an ice sheet at the time of aeolian deposition; many of these extensive soil areas are at relatively low elevations where ice sheets probably did not exist.

After the glaciers receded, the prevailing winds returned to the global pattern which is generally from southwest to northeast. These winds redistributed the aeolian silts over the landscape to leave shallow soil layers on south- and west-facing slopes and deep deposits like snowdrifts on north facing-slopes. This pattern is typical currently. After the glacial era, prevailing southerly winds deposited and redistributed volcanic ash over the area from sources such as Mt. Mazama. Consequently, volcanic ash is a prominent component of aeolian silty soils in both Columbia Basin and Blue Mountain provinces in Oregon.



Discharges of materials from prehistoric Umatilla River and wind redistribution of sandy and silty surface materials originating during glacial melt have likely influenced the mix of parent materials of the lake basin soils. Sandy soils of this basin include such series as Quincy, Royal, Taunton, and Koehler, which represent deep sands as well as sandy soils overlying hardpans or gravel beds and miscellaneous deposits related to the ice age. Stabilized sand dunes, usually oriented from southwest to northeast, account for some of the low ridges and valleys that typify portions of this basin. These dunes are represented generally by the Winchester soil series.

Bottomlands along lower Butter Creek and lower Umatilla River, which are used mainly for irrigated agriculture, are not sandy soils. They include such soil series as Onyx and Hermiston, which consist of alluvial silty materials originating in upper reaches of major drainages. Onyx silt loam is one of the best bottomland soils in eastern Oregon for irrigated agriculture. Extensive areas of strongly sodic bottomland and low terrace soils, such as Stanfield and Umapine series, are in the vicinity of Hermiston.

Around the perimeter of this ancient lake there is evidence of a lakeshore terrace at about 900 feet elevation where it butts against adjacent uplands. Generally, this terrace consists of laminated calcareous silty lacustrine deposits which are likely related to aeolian materials reportedly blown south during the era of receding glaciers farther north. Conceivably, aeolian materials that were deposited in the ancient lake became mud flats around the perimeter, and these now appear as a lakeshore terrace. Thin, laminated horizontal layers in this silty lakeshore terrace material support the concept that they were lacustrine deposits, i.e., deposited in still water over a long time.

Along the northeast perimeter of the lake basin, the lakeshore terrace is overlain by sandy materials of varying thickness. They likely are related to the prevailing northwesterly winds which redistributed sandy surface materials from the interior to the northeastern edge of the lake basin after the lake receded. Sagehill series soils illustrate this sand overlay of the silty lake terrace.

Along the southeastern and southern portion of the lake basin, the lakeshore terrace is overlain by aeolian silty material of varying thickness. These aeolian deposits likely were made after water receded and the lakeshore terrace was exposed; there is usually a fairly distinct line of demarcation between the aeolian deposit and the underlying lacustrine deposit. Furthermore, the aeolian deposit overlying the lakeshore terrace resembles nearby upland aeolian soils in parent material, structure, and texture. Silty soils overlying the lakeshore terrace around the perimeter of this lake basin include such series as Sagemoor, Warden, and Ellisforde.

Basalt underlies Columbia Basin uplands. Soils overlying this basalt bedrock have been formed primarily in silty aeolian deposits that reportedly originated and were blown south during the era glaciers receded in Washington. These silty soils are quite uniform throughout the province but vary somewhat from location to location due to climatic factors that influence soil profile development, such as color and subsoil texture.

For example, Ritzville series is a deep silt loam associated with precipitation of 9 to 11 inches. Walla Walla series is a deep silt loam associated with about 13 to 14 inches precipitation. Soils such as Condon and Valby are moderately deep silt loam over basalt, and Pilot Rock series is moderately deep silt loam over cemented hardpan; all are associated with 12 to 14 inches precipitation. Morrow silt loam is moderately deep over basalt and has a silty clay loam subsoil which reflects the 14-to-15-inch precipitation zone in which it occurs. All these soils in Columbia Basin Province are lighter in color than soils formed in comparable aeolian deposits in Blue Mountain Province.

Grassland soils on north-facing slopes are normally deeper than those on plateaus due to the snowdrift effect that prevailing southerly winds have had in redistributing aeolian deposits. Wrenthem soil series reflects this situation. Soils on south- and west-facing slopes are very stony and fairly shallow to bedrock. Lickskillet series is one example.

The upland soil that suports natural shrub–grassland vegetation in Columbia Basin Province lies on plateaus and ridgetops and is very shallow and very stony. It has a thin loamy surface layer and clayey subsoils and is usually less than 10 inches deep to basalt bedrock. The soil series is Bakeoven, and this land is locally called scabland.



Based on six official weather stations representing a cross-section of the ancient lake basin area in which farming usually requires irrigation, the average annual precipitation is 9 inches. Of this, 29% occurs during the growing season of native herbaceous plants, March through June. About 62% of total precipitation occurs October through February. Average January maximum and minimum temperatures are 39.4 and 24.2°F, respectively. Average March through June maximum and minimum temperatures are 71.3 and 43.7°F, respectively.

Based on 11 official weather stations representing a cross-section of the silty upland area in which dryland summer-fallow wheat farming is practiced, the average annual precipitation is 12.2 inches. Of this, 33% occurs during the growing season of dryland agricultural crops and herbaceous native plants, March through June. October through February precipitation is about 57% of total annual precipitation. Average January maximum and minimum temperatures are 38 and 22.8°F, respectively. Average March through June maximum and minimum temperatures are 66.2 and 39°F, respectively.

Precipitation and temperature data vary from locality to locality as shown in Table 11. 17

It should also be noted that this province, especially the ancient-lake basin portion, is noted for dense fog in winter. This is a source of precipitation that is not represented in weather-station data, yet likely it has a significant effect on the moisture available for plant growth.




The vegetation in the ancient lake basin portion of Columbia Basin Province in Oregon is a natural shrub–grassland (10% or more canopy cover of shrubs) based on soil and ecological site studies made during the 1950s and 1960s.38

Shrubs, such as big sagebrush, rubber rabbitbrush, green rabbitbrush, and broom snakeweed are prominent throughout the lake basin area on deep sandy soils, on sandy soils overlying buried sedimentary deposits, and on shallow silty soils overlying the sedimentary lakeshore terrace. Broom snakeweed was called “matchweed” by local old-timers because sheepherders wintering flocks on this “desert range” would crumple a handful of the resinous shrub as tinder to start a warming fire.

The 1937 Western Range Survey of USDI Grazing Service District 7, which extended more or less from Arlington east to the Stanfield area and from the Columbia River south to about the present southern boundary of the Bombing Range, mapped extensive areas of bitterbrush. This federal grazing district was, at that time, used primarily to winter sheep and cattle. Subsequent soil and site studies reveal that bitterbrush grows in this area on relatively deep sands and on stabilized sand dunes. In this location and climate, bitterbrush does not grow on sites where the sandy surface soil overlying sedimentary deposits is relatively thin nor on sites where the surface soil is loamy or silty. This suggests that there is a soil–moisture relationship.

The presence of vigorous bitterbrush on sandy soils in this location under a climate of less than 9 inches annual precipitation, over 60% of which falls in winter, is indeed an ecological oddity given the normal habitat of bitterbrush in Oregon. Ecological site studies in Oregon over many years indicate that bitterbrush requires effective moisture of 12 inches annual precipitation or more. 5 In the case of this lake basin, some combination of ecological factors obviously is producing the required effective environment. Possible factors might include the region’s fog patterns and soil characteristics.

This basin is noted for dense fog in winter which might augment precipitation records significantly because plant foliage intercepts fog moisture and directs it into the soil profile. Actual annual precipitation may exceed 9 inches in some years.

Sandy soils characteristically absorb water readily; a high percentage of water absorbed is readily available to plants; and absorbed water penetrates deeper in sandy soil during a given period than in soils of finer texture.

Sandy soils actually retain water at depths for longer than one might expect because sand itself acts as a mulch that protects deep soil moisture from excessive evaporation and desiccation. Capillary movement of water upward is less among coarse soil particles than among fine soil particles. Early soil and site studies in this basin revealed moist to wet sand below about 6 feet deep in July and August. However, the possibility that the moisture was wicked up from an underground aquifer is not improbable in those days before irrigation from wells lowered the water table. Bitterbrush is a very deep-rooted plant once established, and it can utilize such deep soil moisture.

The water-holding capacity of various soil textures (Table 12) indicates that there is a greater increment in moisture equivalent (field capacity) between several sandy-soil textures than there is between loamy soils or soils of finer texture.19, 20


An interesting factor in soil–plant relationships is presented by the incremental increases in field capacity—such as the ability of the soil profile to retain water, from coarse sand to fine sand and likely on to loamy sand—and then the decreasing increment between loamy sand and sandy loam textures. The relationship between the incremental increase of field capacity in various sandy textures and the occurrence of bitterbrush in this sandy basin is not clear.

However, a cursory study of relationships between bitterbrush stands and soil textures in this sandy lake basin was made during the 1950s or 1960s.38 Auger holes at most section corners and quarter-corners provided a uniform grid sampling in an area about 4 miles from east to west and 2 miles from north to south. The area had sporadic stands of vigorous bitterbrush, some 30 to 40 sample plots. Essentially all plots with soil described as loamy sand or fine sand produced bitterbrush. In virtually all locations where soil was described as sandy loam there was no bitterbrush.

The phenomenon of loamy sand plots with no bitterbrush suggests a wildfire pattern. The major habitat for bitterbrush in this lake basin portion of Columbia Basin Province are the Quincy loamy fine sand and Koehler loamy sand soil series.

The growing season of bitterbrush in the lake basin is very much earlier than any other bitterbrush site in eastern Oregon. Here, bitterbrush flowers at least 30 days earlier than in other locations where it is common. That indicates the annual growth cycle of bitterbrush in the basin has adapted to begin and end while adequate soil moisture is available.

The herbaceous vegetation on these sandy sites is dominated by needle-and-thread, Indian ricegrass, Sandberg bluegrass, and varying amounts of Columbia milkvetch, wormwood, pteryxia, Gorman lomatium, and Carey balsamroot. Cheatgrass is a strong winter annual which forms rosettes 3 to 6 inches in diameter over the course of the winter and provides excellent spring forage for various classes of grazing animals. These sandy sites also are well known as nesting habitat for curlews.

The herbaceous vegetation on stabilized sand dunes is dominated by Indian ricegrass, yellow wildrye, Sandberg bluegrass, squirreltail, and a variety of perennial forbs such as buckwheat, Gorman lomatium, wormwood, pteryxia, scurfpea, yellow spiderflower, veiny dock, and pricklypear.

Around the southern perimeter of the lake basin, the natural plant community on the silty lake terrace site is one of the simplest natural plant communities in Oregon. It consists of a shrub–grassland in which big sagebrush, gray rabbitbrush, and broom snakeweed constitute slightly more than 10% canopy cover. Bluebunch wheatgrass and Sandberg bluegrass strongly dominate the total cover, and perennial forbs, such as yarrow, snow buckwheat, woollypod loco, and spreading phlox are sparse in the canopy cover. A good example of this site in reasonably natural ecological status is The Nature Conservancy’s Boardman Research Natural Area in the southeast corner of the Boardman Bombing Range.

Based on 171 recordings representing examples of eight major ecological sites in this province in southern Umatilla, Morrow, Gilliam, and Sherman counties on which native vegetation was in high ecological status, the natural vegetation of silty uplands in Columbia Basin Province in Oregon is a natural grassland, i.e., less than 10% canopy cover of shrubs (Fig. 16). Data from these 171 recordings show that low gray rabbitbrush was on 87% of the plots but in very minor amounts. In contrast, big sagebrush was recorded on only 29% of the plots and only in minor amounts. It is typical of the rangelands throughout the silty uplands of this province in Oregon to have low gray rabbitbrush and broom snakeweed, rather than big sagebrush, as the shrub component of native plant communities in deteriorated ecological status. Natural revegetation of abandoned croplands in this province results in gray rabbitbrush as the primary shrub component. Other shrubs that are sparse in examples of high-ecological-status plant communities include rose, gray horsebrush, and green rabbitbrush.

Figure 16: Managed natural upland grasslands in Columbia Basin Province, Oregon

Obviously, the question of fire arises in discussions of whether these plant communities are natural grasslands. Fire can temporarily eliminate big sagebrush and may even aggravate growth of gray rabbitbrush. However, the area represented by the recordings is extensive, so it is not reasonable to assume that all the examples of high ecological status had uniform and comparable fire histories and that an original sagebrush cover had been uniformly and permanently altered by fire over the entire area.

Based on ecological site studies, the natural shrub pattern on silty upland ranges in Columbia Basin Province of Oregon changes from east to west. East of the southwest corner of Morrow County, big sagebrush is minimal, if present at all, in native plant communities in high ecological status. Big sagebrush is not a prominent shrub even under deteriorated ecological status. However, in the vicinity of southeastern Gilliam County and west into southern Sherman and Wasco counties, the likelihood of big sagebrush increases somewhat, especially under deteriorated ecological status. Still, shrubs collectively usually do not make up 10% canopy cover on silty upland ranges in that portion of the Columbia Basin Province.

Several broad ecological factors are related to this phenomenon.

1. The reason big sagebrush is minimal, if present at all, east of Gilliam County may be that Blue Mountain Province native plant communities form a buffer zone between that portion of the Columbia Basin and the John Day Province to the south. Big sagebrush is a prominent component of the natural shrub–grasslands that typify the John Day Province. Blue Mountain native plant communities have evolved under very favorable soil and climate conditions, and a wide variety of other shrubs typify this province. For this reason, big sagebrush has not been able to encroach north through the vigorous Blue Mountain plant communities. This concept is supported by the fact that, throughout Blue Mountain Province, big sagebrush is not a prominent shrub no matter what the ecological status of the plant community.

2. West of Kinzua, the westernmost extent of Blue Mountain Province, a huge topographic saddle extends from Kinzua west about 50 miles to the Mutton Mountains in the northeast corner of Warm Springs Indian Reservation. The Deschutes and John Day rivers flow north through this saddle; however, the rivers’ tributaries generally are entirely within John Day or Columbia Basin province. The demarcation line between the two provinces essentially is the division between watersheds draining north into Columbia Basin from those draining south into John Day Province.

Along this huge saddle, the Columbia Basin and John Day provinces are contiguous at a relatively low elevation of 3,000 to 3,500 feet. Presumably, this offers less resistance to encroachment of vigorous shrubs, such as big sagebrush, from John Day Province into Columbia Basin Province, unlike the situation in the Blue Mountain buffer zone. Nevertheless, native plant communities in high ecological status on silty aeolian Columbia Basin sites, such as near Shaniko, generally do not support a 10% canopy cover of shrubs. This is in spite of the likelihood that native vegetation in the vicinity of Shaniko was decimated in early 1900s when Shaniko was the largest inland shipping center for wool in United States.

Two reasonable conclusions are that the natural native vegetation on upland silty aeolian soils of Columbia Basin Province in Oregon was not a shrub–grassland, and that big sagebrush is not a vigorous component of deteriorated plant communities, such as those in John Day Province.

Specimens of big sagebrush that look very old are in western Columbia Basin Province. They are almost exclusively along drainageway bottoms in the Deschutes and John Day river canyons and their tributaries. In this part of the province, these drainageway bottoms consist of deep, very gravelly and stony colluvial materials originating from nearby basalt rimrocks and canyon walls. These kinds of soils provide ideal sites for big sagebrush. In the same area, where silty alluvium from upland silty soils has accumulated as meadows in drainages, the natural vegetation consists of basin wildrye and associated meadow species, depending upon degree and duration of soil moisture.

It seems reasonable to assume that the incidence of big sagebrush in the western part of the province likely is related to the avenues for encroachment that have been provided over time by the Deschutes and John Day river canyons and their dendritic tributaries, which have suitably deep, coarse, colluvial soils.

In the Columbia Basin Province of Oregon, the silty upland ranges include a ecological site locally called scabland. It occurs along the outer edges of ridgetops and on sloping areas in minor drainages between about 1,000 and 3,500 feet elevation. Slopes vary from 2 to 20%. The soil of this site is very stony loam and is very shallow (7 to 10 inches) over basalt bedrock. Natural vegetation is dominated by Sandberg bluegrass and stiff sagebrush, i.e., a shrub–grassland. Some areas, however, have no stiff sagebrush although the soil and other components of the plant community remain the same.

Overall, the natural plant communities of silty upland sites throughout the Columbia Basin Province of Oregon consist primarily of bluebunch wheatgrass, Sandberg bluegrass, Idaho fescue, and a wide variety of perennial forbs. Idaho fescue is absent on the most droughty sites, such as south exposures, but it strongly dominates the plant community on more moist sites such as steep north exposures.

The cultivar ‘Sherman’ big bluegrass was originally selected from a grassland site near Moro in Sherman County. Beardless bluebunch wheatgrass, from which the cultivar ‘Whitmar’ was developed, has been recorded in the silty-soil site west of Boardman and north of the sandy-soil area. However, the type location from which Whitmar was selected is in Whitman County, Washington.


Management Implications

The Columbia Basin Province in Oregon includes irrigated agriculture on bottomlands along major drainages such as Rock Creek in Gilliam County; Willow and Butter creeks in Morrow County; and Birch and McKay creeks and Umatilla River in Umatilla County. These sources of irrigation water are fairly reliable because they are renewed by runoff from prominent watersheds. However, water use likely will be more strictly controlled under fish-habitat and similar regional programs. These lands are used primarily to produce livestock forages.

Some farmers use sprinkler irrigation from wells in the vicinity of Athena to produce wheat, peas, alfalfa, and other crops. Irrigation near Milton-Freewater is primarily a gravity-flow system for orchards and a variety of food crops. Boardman-area irrigation involves both gravity-flow systems from water stored in Cold Springs Reservoir and pumping from wells and the Columbia River; production is of livestock forages, grains, and food crops. State regulations are limiting use of underground water to maintain aquifers. Other water sources are likely to come under stricter control because of regional fish habitat programs.

Wind erosion is a potential problem in the area of irrigated sandy soils if the water supply diminishes significantly. Cultivation has destroyed all native vegetation, so these soils are extremely erodible when exposed and dry. If it becomes necessary to take some lands out of irrigation due to limited water, it will be very difficult to revegetate with species that will establish and persist in the arid climate cnd shifting sandy soils. Massive sand dunes easily could inundate existing urban and rural homes and businesses and cover otherwise stable cgricultural lands.

From Pilot Rock west, arable soils are under a summer-fallow wheat cropping rotation with special effort to minimize erosion, especially in fallow years. Practices such as stubble mulch, cross-slope farming, diversions, terraces, contour strip cropping, and others have been tried with mixed success primarily because these practices often require special equipment and more time and mabor, which cuts net income.

The silty upland ranges of Columbia Basin Province in Oregon are not well suited for year-long livestock grazing because, generally speaking, their summer forage is not good quality . A well-balanced ranch in this area includes irrigated land for producing winter feeds, some higher elevation summer rangelands which are usually in Blue Mountain Province, and mower elevation rangelands in Columbia Basin Province for spring, late autumn, and, to some extent, winter grazing.

The ancient lakeshore terraces west of Butter Creek Junction on Highway 207 are covered by a mantle of silty aeolian soil.

Where the silty overlay is relatively thin, the underlying laminated calcareous lakeshore terrace is close enough to the surface to resist water and root penetration, making these soils generally unsuited for dryland agriculture. Examples of abandoned cropland are on the lakeshore terrace along the southern perimeter of the lake. Some of these abandoned croplands were later seeded to crested wheatgrass in an attempt to replace the destroyed native forage. However, the seedings were largely disappointing or totally unsuccessful, at least partly because the ancient lakeshore terrace is not suitable for cultivation or reseeding, even though it is a silty soil and easily worked. Management, or rather mismanagement, of the grass seedings likely is also a factor in the adverse results.


Province Demarcation

Ancient Lake Basin Demarcation

The eastern line of demarcation between the ancient lake basin and aeolian silty uplands in Columbia Basin Province begins at the Columbia River near the Oregon–Washington border. It meanders south and up Juniper Canyon to near where the north cnd south forks of Juniper Canyon merge. From there the line goes southwesterly to about 5 miles east of Cold Springs Reservoir and then generally south to cross Umatilla River about 3 miles southeast of Echo. In this area, the ancient lakeshore terrace comprises the substrata beneath overlying soils which are the Sagehill fine sandy loam and Ellisforde silt loam soils. The soil representing the aeolian silty uplands to the east of the lake basin is Ritzville silt loam.75

From about 1 mile south of Echo, the demarcation line follows southwest about 2 miles south of and nearly parallel to the highway between Echo and Highway 207 near Butter Creek. The line veers north to cross Butter Creek at about the junction of Echo Road and Highway 207 and then runs west.

The line between the ancient lake basin and the deep aeolian silty soil uplands to the south continues west to enter Morrow County in the middle of Township 3N R27E. Then it goes southwest to near Sand Hollow where it turns southeast to cross Highway 207 about 2 miles west of Butter Creek Junction. From this point west, the ancient lakeshore terrace that forms the perimeter of the lake basin is covered by a mantle of silty aeolian soil. Soils represented by this silty overlay include Warden, Ellisforde, and Sagemoor.85

From the vicinity of Butter Creek Junction, the line goes west across Sand Hollow to upper Juniper Canyon where it follows the canyon northwest to near the south boundary of the Bombing Range. From there it runs west at about 900 feet elevation along the upper level of the lakeshore terrace to Cecil where it crosses Willow Creek and enters Gilliam County.

The Oregon Trail traversed from east to west along the southern edge of this ancient lakeshore terrace on the way to Cecil on Willow Creek. Well Spring and the nearby historic site of a skirmish with Indians are on the lakeshore terrace.

From the vicinity of Cecil, where the demarcation line of the ancient lake basin crosses into Gilliam County, the line runs west to Fourmile Canyon. It follows the east canyon rim north cnd then along the Eightmile Canyon to cross Willow Creek at Rhea.78 This area of Warden soil in Gilliam County is the farthest west in the Columbia Basin Province in Oregon that a sizable crea of the calcareous laminated silty lakeshore terrace exists under a thin mantle of silty aeolian materials. However, small remnants of this lakeshore terrace occur about 4 miles up the Deschutes River from Highway 84, which indicates that this terrace may have been between Willow Creek and the Deschutes River while the ancient lake was intact. Conceivably, when the ice jam at The Dalles broke, water in the lake rushing down the Columbia Gorge could have scoured out all but remnants of the silty lakeshore terrace lying west of this remaining segment in northeastern Gilliam County.

Province Demarcation

Columbia and Palouse Demarcation

The line of demarcation between Columbia Basin and Palouse provinces begins at the Oregon–Washington border about 5 miles northwest of Milton-Freewater in Umatilla County. From that point the line runs southwest, passing just east of Milton-Freewater. About 2 miles south of Milton-Freewater the line turns west to about 3 miles north of Helix where it abruptly veers south across Wildhorse Creek to the Umatilla River about 1.5 miles northeast of Mission. It follows on the north side of the river east upstream to about Cayuse where the Columbia Basin, Palouse, and Blue Mountain provinces join. The demarcation line from Milton-Freewater to near Mission is the soil line between Walla Walla silt loam series, in Columbia Basin Province, and Walla Walla silt loam, high-precipitation phase, which is in Palouse Province.75

Columbia and Blue Mountain Demarcation

The line of demarcation between Columbia Basin and Blue Mountain provinces meanders southwest from near Cayuse along the mountain footslopes at about 2,000 feet elevation to south of Pilot Rock. It continues southwesterly up Owens Creek to cross Highway 395 about 8 miles south of Nye Junction, at an elevation of about 3,500 feet, and continues west across North Fork Butter Creek to the top of Franklin Hill. It lies just south of Lena on South Fork Butter Creek and travels southwesterly at about 3,500 feet elevation, passing about 6 miles southeast of Heppner and on to the vicinity of Hardman.66

From Hardman vicinity, the line runs west to cross into Gilliam County cnd then southwest to just north of Lonerock, which is in an isolated segment of John Day Province in southeast Gilliam County.78

About 1 mile west of Lonerock, the line of demarcation is between Columbia River and Blue Mountain provinces for about 3 miles cnd then, near the old Lost Valley School road junction, the line is between Columbia Basin to the north cnd John Day Province to the south. In this vicinity, the line of demarcation is a belt several miles wide in which differentiating features of Blue Mountain, Columbia Basin, cnd John Day provinces intermingle. The mapped line is a judgment call.78, 90

Columbia and John Day Demarcation

About 3 miles northwest of Kinzua, which is in Blue Mountain Province, the line of demarcation between Columbia River and John Day provinces goes westerly at about 4,000 feet elevation in the vicinity of Cummings Pass and across Butte Creek about 3 miles northwest of Fossil, which is in John Day Province. From that point, it runs southerly, westerly, cnd then north ct about 4,000 feet elevation to encompass a large rocky plateau that juts south into Wheeler County. The plateau is in Columbia Basin Province. The plateau’s southwest corner, where it overlooks John Day Province in the vicinity of Clarno, is a good place to see a very abrupt province boundary. Taking the photo from below, near Clarno, shows the plateau above the basalt rim, which is Columbia Basin, cnd the exposed light-color tuffaceous cnd sedimentary deposits below the rim which are typical of John Day Province.

The line between Columbia Basin and John Day provinces crosses the John Day River from Gilliam County into Wasco County ct the Wasco–Sherman county line on the river. The John Day River canyon north of this point is in Columbia Basin Province; south, it is in John Day Province. From this point the line of demarcation meanders southwest along the west-side breaks of John Day River canyon to about 2 miles north of Antelope and continues southwest along the north-side breaks of Antelope Creek. About 2 miles north of Antelope, c roadcut on Highway 218 displays the Columbia Basin basalt cap and silty aeolian soil overlying the light-color sedimentary materials of the John Day Province. This is another example of an abrupt line of demarcation that can be examined closely and easily.

From the vicinity of Kinzua westward, the demarcation line is represented by the soils line between Condon and Bakeoven soils in Columbia Basin and Simas and Tub soils in John Day Province.78, 86, 90

The line of demarcation crosses Highway 97 in Cow Canyon about 1 mile north of the confluence of Antelope and Trout creeks. From there it goes southwest into Jefferson County to encompass the small plateau area about 2 miles west of Willowdale.83

The line then follows north clong the east-side rimrocks of Deschutes River canyon to the railroad siding Nena, where it crosses to the west side of Deschutes River. The Deschutes River canyon north from Nena is in Columbia Basin Province; south, it is in John Day Province. Nena railroad siding is about 7 miles upriver from Maupin.

From Nena, the line of demarcation runs southwest up the ridge on the west side of Deschutes River to the top of the divide, which used to be the north boundary of Warm Springs Indian Reservation. Mutton Mountains are in John Day Province. The line follows the ridgetop northwest to the road between Wapinitia and Simnasho. It is near this ridgetop pass on the road that Columbia Basin, John Day, cnd The Dalles provinces adjoin. From this point, the line of demarcation between Columbia Basin and The Dalles provinces veers north to pass just east of Pine Grove, which is in The Dalles Province, and north in the vicinity of Smock Prairie School and Wamic, both of which also are in The Dalles Province. From east of Wamic, the line goes north. In this vicinity, the lower reaches of White River canyon, which heads near Mt. Hood, create c narrow strip of The Dalles Province extending into Columbia Basin Province southwest of Tygh Valley. Tygh Valley itself is a small island of John Day Province which extends about 10 miles east to west and about 1 mile north to south. The light-color clayey ancient sediments (Simas and Tub soils) that typify John Day Province are obvious in this location. North of Tygh Valley, Tygh Ridge and Postage Stamp Lookout are in Columbia Basin Province.

Columbia and The Dalles Demarcation

Where the road to Little Badger Forest Camp crosses Tygh Creek about 3 miles northwest of its junction with Highway 197, the line of demarcation between Columbia Basin and The Dalles provinces winds north to Dufur, which is barely inside Columbia Basin Province. From Dufur, the line travels toward The Dalles. However, the line passes about 2 miles east of The Dalles and ends at the Columbia River near the mouth of Fifteenmile Creek and The Dalles dam. The city of The Dalles is in The Dalles Province.

The line of demarcation between Columbia Basin and The Dalles provinces is based on soil lines between Condon, Walla Walla, Wapinitia, and Maupin soil series, which represent Columbia Basin Province in this vicinity, and Frailey, Wamic, and Skyline soil series, which are on the eastern footslopes of Cascade Mountains and represent The Dalles Province.86