THE ORIGIN AND CHARACTERISTICS OF SOIL MOUNDS
AND PATTERNED GROUND IN NORTH CENTRAL OREGON
by
CLARK A. NELSON
A Research Paper submitted to The Department of Geography
in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE
June 1977 ACKNOWLEDGEMENTS
The completion of this paper would not have been possible without the support and interest of many people. To my major professor, Dr.
Charles Rosenfeld, I wish to express a deep gratitude for initiating my interests in this topic and for supplying the inspiration to move when
I was standing still. His knowledge of physical geography and his ability to combine youthful insights with the wisdom of experience never ceases to amaze me. He also served a valuable role as liaison between myself and the National Guard of Oregon from whom much research material was obtained.
I would like to dedicate this paper to my parents and family who deserve a thanks which is beyond words. They have always believed in and encouraged me to pursue my own interests, and have given me unending support in various endeavors throughout my life.
A special tribute must be made to all my friends, who, often unbeknownst to themselves, supplied a continuing source of encouragement, and helped to clarify the realities which often become obscured through deep involve- ment.
Last but not least, special mention goes to the small but active band of coyotes residing near my research area. Their curiosity and nocturnal antics, though initially resulting in insomnia, soon were anticipated with relish upon subsequent returns to this area. TABLE OF CONTENTS
ACKNOWLEDGEMENTS
TABLE OF CONTENTS
LIST OF FIGURES
INTRODUCTION ...... 1
PHYSICAL SETTING...... 2
Geology...... 4
Soils...... 6
Climate...... 8
Vegetation...... 9
LandUse...... 11
METHODOLOGY ...... 11
PHYSICAL CHARACTERISTICS OF STONE RINGS AND SOIL MOUNDS
IN NORTH CENTRAL OREGON...... 13 Definitions...... 13
Problems Encountered in Studying Relict Forms...... 14 Morphology...... 15
Composition...... 21
ORIGIN OF PATTERNED GROUND AND SOIL MOUNDS IN NORTH
CENTRAL OREGON ...... 25 EarlyAccounts...... 25
Pleistocene Environmental Conditions ...... 27
Processes Involved in Formation...... 29
CONTEMPORARY GEOMORPHIC ACTIVITY...... 32 CONCLUSIONS .36
SUGGESTIONS FOR FURTHER RESEARCH...... 39
BIBLIOGRAPHY...... 43
LIST OF FIGURES
1. Location of the Deschutes-Umatilla Plateau...... 3
2. Geology of the Deschutes-Umatilla Plateau ...... 5
3. Soils of the Deschutes-Umatilla Plateau ...... 7
4. Relationship between station elevations and monthly frequency
of diurnal freeze-thaw cycles 1962-1971 ...... 10
5. Soil mound excavation ...... 12
6. Schematic diagram of patterned ground development and the
effectsof slope on shape ...... 17
7. Schematic diagram of the successive stages of the development
ofsorted stone nets...... 18
8. Aerial photograph of various soil mound configurations...... 19
9. Aerial photograph of soil mounds and stone stripes...... 20
10. Soil mounds and intermound drainage swale ...... 22
11. Cross-section of soil mound and stone ring...... 23
12. Stone stripe cross-section (two photos) ...... 24
13. Extent of glaciation and pluvial lake development in Oregon
duringQuaternary ...... 28
14. Sequence of the upfreezing of a stone within a. soil mass...... 31
15. Small stones heaved by contemporary frost action ...... 14
16. Frost boil formed by contemporary frost action...... 35
17. Contraction cracking in frost susceptible soil under current
climatic conditions ...... 36
18. Accumulation of frost shattering rubble ...... 38 THE ORIGINAND CHARACTERISTICSOF SOIL MOUNDS
AND PATTERNED GROUND INNORTHCENTRAL OREGON
ABSTRACT: Sections of the landscape in north central Oregon are dominated by numerous soil mounds surrounded by stone rings. Mounds vary in size and shape depending on slope and mound density. Inter- nal soil mass is free of stones to the bedrock level and is covered by successive layers of volcanic ash. Stone ring profiles exhibited strong vertical sorting of stones by size and a decrease in ring width with depth. Stone areas were generally void of soil except for a shallow layer on the bottom. Noticeable depressions in the bedrock layer were found beneath the rings. Inter-mound areas appear as large surface drainage swales with a very shallow soil layer contain- ing numerous stones. Formation of stone rings is attributed to the ejection of stones from the soil mass by intense frost action under a former periglacial climate. Increased subsurface drainage, piping, and erosion led to a lowering of the soil surface above the stone rings, thus segregating the high centered mounds. Small scale frost features continue to be formed though the climate is no longer of sufficient intensity to produce stone rings and soil mounds.
INTRODUCTION
There are a number of areas throughout the Pacific Northwest where
relict patterned ground features exist. Though each area is unique in
topography, soil conditions, and Pleistocene history, the patterned
ground has a similar appearance with stone stripes and steps, stone nets,
sorted stone rings,and their associated soil mounds. Landscapes composed
of these features are often referred to as pimpled plains or biscuit
scablands.
These phenomena occur in specific regions on the Columbia Intermontarie
Plateau as it extends into Washington, Idaho, and Oregon. My interests
are specifically with the sorted stone rings and soil mounds as they
exist on the Deschutes-Umatilla Plateau in Oregon. Upon initial in- vestigation, it became apparent that there is a dearth of information 2 concerning most aspects of these features. The patterned ground in this specific region has received only cursory treatment in relation to the
Pacific Northwest as a whole. Theories as to the origin of these micro- relief features are wide-ranging and quite controversial. Few of the theories consider the association between the mounds and stone rings, though each appears as an integral part of the other.
As these features are generally cold climate phenomena, I felt further study would be of importance in understanding the environmental! climatic realm of this region as it existed during the Pleistocene.
Hopefully it will also give insight into processes currently active, form- ing similar features in arctic and alpine areas.
The purpose of this research was to investigate and document the morphology, composition, distribution of, and relationship between the soil mounds and stone rings. I have used my research in conjunction with the previous work of other authors concerning similar features else- where, to develop a theory as to the origins of this biscuit-scabland topography. Finally, contemporary geomorphic processes active in my research area have been evaluated. Though these processes are comparable in their nature, the intensity with which they occur has greatly diminished since the Pleistocene, at which time the patterned ground is believed to have formed.
PHYSICAL SETTING
My research area is located on the Deschutes-Umatilla Plateau in north central Oregon (Figure 1). Though this broad region has a number of unifying, homogeneous characteristics, it contains a number of sub- regions, delineated on the basis of soil types, climate, and geology. DZCHUT5LJMAT1LLA OREGON PLATEAU
Cent1M.I.t &..nd.,I.. 0 10 20 30 40 20 0..cI.,.tJst;II. P1.,... hssd.,1 Figure 1. Location of the Deschutes-LJmatilla Plateau in north central Oregon. iI.t..Istt CA) 4
The Deschutes-Umatilla Plateau's northernborder extends along the
Columbia River for approximately 250 km. The westernmargin adjoins the
base of the Cascade Range and extends 11Ekm to the south. The southern
boundary is marked by the northern slopesof the Blue Mountair,while the
eastern margin extends south from the Coli.jmbia River approximately 15 km,
in the vicinity of Pendleton.
Geology
The Deschutes-Umatilla Plateau is composed ofa vast volcanic
sequence known as the Columbia River Basalts (Figure 2). At one time,
numerous lava flows emanated from large fissures, obscuring the original
topography with successive layers of basalt. Individual basalt flows
were from lm to over 60m thick and covered approximately 500,000sq km
in Washington, Idaho, and Oregon. Though the basalt flows decrease in
thickness toward the southern plateau edge, theyaverage over 6On in
depth throughout (Newcomb, 1969). Numerous ridges and exposed bedrock
outcrops, representing lava flows of varying resistances,are apparent
throughout the region.
The Plateau is on an asymmetrical syncline, dippingnorthwards from
to 3° from the Blue Mountains to the Columbia River (Newcomb, 1969).
Elevations range from lOOm near the Columbia toover ilOOm in the vicinity of Shaniko. A short distance south of Shaniko, the Plateau terminates with a large escarpment. My research was conducted near
Shaniko and northward on a broad, relatively undissectedplain, charac- terized by rolling hills with numerous bedrockoutcrops. Bar and ball on downthrown side Fault Syncline I icr Dashed where approximate Antici Dotted where covered nocline -. Hermi Pendlethn-.. -'-f icr ,--,tQT ; Q:IT( Dufu/3 ezQTs X"a1 ) S Moro icr: _____ d&*L Younger Tertiary and Quoternary se dimentciry rocks I - : Younger Tertiary arid Quoternory volcanic rocks Tcr 1 0 20 Kilometers 40 60 Columbia River Basalt ElIllIllIll Source: in M,nerol and WaterNewcomb, Resourses R. C., 1969, Geology of of Oregon, Bulletinthe 64, Deschutes.Umatilla p. 63. Plateau, Olderpre-Tertiary Tertiary and rocks Figure 2. Geology of the Deschutes-Umatilla Plateau. 6
Soils
A wide variety of soil types are found on the Deschutes-Lthatilla
Plateau, reflecting regional variability of precipitation, temperature,
and vegetation (Figure 3) (Franklin and Durness, 1969, p. 26). Most of
the soil is loessial in nature, originating from the flood plain and large
sand plains of the Columbia River, and the glacio-fluvial lake beds formed
during the Pleistocene (DyKsterhuis, etaL, 1969, p. 10).As the distance
from the Columbia River increases the depth of deposition decreases and
in some areas, a decrease in particle size is also observed.
There are four main soil regions encountered on the Plateau : 1) the
Columbia River lowlands; 2) the thick bess-mantled zone adjacent to the
lowlands; 3) the thin bess-mantled area covering the southern Plateau, and; 4) a narrow band coinciding with the lower, eastern slopes of the
Cascade Mountains.
The main area of concern in my research, the area where patterned ground is most prevalent, lies in the third and fourth regions. Within the third area, the soils exhibit a great deal of variability in depth, color, and composition. These soils of bess and mixed colluvium range from less than 40 cm to 2m in depth. Much of the soil is relatively stone free, however large areas of stone pavement may be encountered
(Norgren, et al., 1969). This latter phenomena occurs where the soil mantle is relatively thin and numerous well-rounded to angular basalt fragments are exposed in great concentrations at the surface. Better and deeper soils generally occur on the more level basalt uplands. Poorer soils and areas of stone pavement are encountered on gently sloping to steeply contoured areas. The better soils are of a medium silt loam I 25 50 kflom.t.,. 4!1Wt "1 Friend LMUI.pin Kent ondon Hordmon / 7ROCk c' 1 Shortiko tone rock 0/ Columbia River Lowlandi Mo ra ThinThick Loess-montled Loess-montleci Section Region _-; I: Western Margin j Requirem.nss For Woter, Appanthce 1.49 5,6.6 7. Figure 3. Generalized soil map of the Deschutes-Umatilla Plateau. texture, while the shallower, stony areas range from silt to clay loam.
The soils on the western margin of the Plateau are composed of
alluvium, bess, and volcanic ash. These well drained soils are from
50 cm to im deep and have a loam to clay or sandy loam texture. They
exhibit numerous characteristics similar to those soils in the third
region. Bedrock in the fourth region is composed of basaltic lava flows
interspersed with volcanic sediments from the Cascades, known as the
Dalles Formation (Norgren, et al., 1969).
In the areas where mounds were studied, soils were of the Bakeoven-
Condon complex. These soils are well-drained and occur on slopes of20
to200. Soil mounds are generally composed of the Condon series while
the surrounding scablands, or areas of stone pavement, are of the Bake-
oven series. Both soils are found on uplands in this region, where
elevations range from 45ii to ilOOm. Runoff is slow to rapid, while
permeability is only moderate. Erosional hazard is slight to moderate
(Green, 1970, p. 9). The waterholding capacity and texture of this type
of soil causes them to be particularly susceptible to frost action and
ice segregation within the soil mass (Embleton and King, 1968).
Climate
Marine climatic influences from the Pacific Ocean are seldom felt as the Deschutes-1natil1a Plateau lies in the rain shadow of the Cascade
Range. Mean annual precipitation ranges from 220 mm to 400 mm, varying greatly with year and locality. Less than 9% of the annual precipitation occurs during the summer months, characterizing the area as droughty to very dry. Convectional storms occur periodically during this time though
their contributions to annual precipitation amounts are of minor significance. The majority of the precipitation occurs in the fall, winter,
and spring. Snow during the winter months account for only 17% of the
yearly total.
Temperatures below freezing can occur throughout the year, however
the suniiier months are generally frost free. Mean annual air temperature
is 7°C to 9°C with typical diurnal fluctuations of 10°C to 16°C. Mean
temperature of the coldest month (January) currently ranges from -2°C to
1°C. Temperatures tend to decrease with an increase in elevation as one moves south away from the Columbia River. Consequently the total number
of frost days increases. In the vicinity of my research area, frost free
days number from 50 to 140 at 0°C and 100 to 170 at -2.2°C (Figure 4).
Vegetation
Steppe and shrub steppe vegetation characterizes much of the Plateau.
Trees appear only in areas of increased moisture, such as at higher ele- vations and in riparian sites. Where soil mounds and stone pavement occur, there are generally 2 distinct vegetation communities. The deeper and better drained soils of the mounds support a relatively luxuriant vegeta- tion association dominated by Big sagebrush (Artemisia tridentata), and
Bluebunch wheat grass (Agropyron spicatum). Other species commonly en- countered in this plant comunity include Idaho fescue (Festuca idahoensis), yarrow (Achillea millefolium), Sandberg bluegrass (Poa secunda), and a few other perennial forbs.
The adjacent scabland areas support a much sparser, association dominated by the low-lying Stiff sagebrush (Artemisia rigida) and Sandberg bluegrass (Poa secunda) with relatively few other perennial forbs. These abrupt changes in vegetation communities are very conspicuous on aerial > ,v .O q>, " 0 \Oq )onuO'y <,, ':> ,' w/ / ,/ //" 4" t/' cj' / Febru ory Ma,ch April 900 May 700800 JuniJuly 600 S.pt.mb.r August 400500 °0 October 200300 NovemDecember be, 100 Hti Numb., of Days riin which Freezing and 'homing Occurred Morr'bly 6.9I LoJ L:; 0 Figure 4. Pyrch,Relationshipfrequency 1973, of Figurebetween diurnal 5, station freeze-thawp. 14.) elevation cycles. and monthly (After 11 photos and serve to delineate readily the mounds from the scablands.
Land Use
The predominant land use in the southwestern section of the Deschutes-
Umatilla Plateau is cattle grazing. Shallow stony soils do not support the productive grain farming as found in other areas. In many areas the perennial bunch grasses have been eliminated due to overgrazing. Further north in this region, winter and spring wheat are commonly grown as the occurrence of biscuit scabland landscape decreases. In many areas where the percentage of scabland is small, soil mounds are plaflated to cover the scablands, thus increasing agriculturally productive land area.
METHODOLOGY
Four soil mounds and their adjacent stone rings were excavated in order to expose a cross-sectional profile (Figure 5). Field sketches and photographs of each cross-section were made for further reference.
From the profiles, numerous soil samples were taken from various areas to determine the extent, if any, of sorting of soil particles. Soil samples were analyzed by gravinimetric granulometry and the Bouyoucos hydrometer method.
The long axes of 50 contiguous rocks wer'e measured in the top, mid- dle, and bottom sections of each stone ring profile. This was done to determine and document the degree of vertical sorting within these stone features.
Slope angles were measured with a Suunto hand clinometer. Slope aspect and mound axis orientation were determined with a Brunton Pocket
Transit. Soil mound and stone ring dimensions were measured with a 3thi steel tape. 12
Figure 5. Cross-sectional excavation of a soil mound. 13
A field checklist was used to guide in the collection of pertinent
information concerning each mound area. This included descriptive, loca-
tional, and quantitative data. A great deal of information was obtained
from aerial photos, topographical maps, and soil surveys. Thermal in-
frared imagery was also used in an attempt to determine drainage patterns
and areas of relatively high soil moisture.
Trend surface analysis was used to analyze cirque floor elevations
in a transect across Oregon. From this, I determined the approximate
location of the Pleistocene climatic snowline as it increased in ele-
vation in relation to distance from the Pacific Coast.
PHYSICAL CHARACTERISTICS OF STONE RINGS AND SOIL MOUNDS
IN NORTH CENTRAL OREGON
flc,fini 14r,n
Patterned ground is defined by Washburn as a "group term for the more or less symetrical forms such as circles, polygons, nets, steps, and stripes, that are characteristic of, but not necessarily confined to mantle subject to intensive frost action" (Washburn, 1956,p. 824).
Contemporary formation of such features under arctic and alpine environ- ments has supplied an excellent opportunity to study these features.
However, the myriad of geomorphic processes resulting in a wide variety of forms, and the environmental heterogeneity under which these features are found, have resulted in a general lack of knowledge concerning these phenomena. Patterned ground forms, as defined by Washburn, are not re- stricted to cold climates, however it is here that they are most numerous and reach their greatest development. 14
Patterned ground features may be broadly classified as either
sorted or non-sorted, depending on the degree of spatial redistribution
of soil particles and rocks within the forms (Washburn, 1973, p. 101-103).
Numerous hypotheses have been formulated concerning the dominant processes
which result in patterned ground features. These range from the work of
biological agents such as pocket gophers and pre-historic man, to soli-
fluction, surface erosion, and frost action.
Washburn has stated the following premises dealing with the genesis
of patterned ground in general:
1) Patterned ground is polygenetic;
2) Similar forms may result from different processes;
3) The same process may produce different forms;
4) There are more processes than are currently recognized (Washburn,
1973, p. 137).
The dominant forms found in Oregon and the Pacific Northwest are fossil or relict forms, the expression of various geomorphic processes active under former climatic conditions of the Pleistocene. Such processes are currently either entirely absent or not of sufficient intensity to activate formational processes.
Problems Encountered in Studying Relict Forms
The study of these fossil forms presents a number of interesting problems. Accurate reconstruction of former environmental conditions, especially climate, must be made. Past climates are interpreted from present day landforms and vegetation, as well as interpolating from con- temporary climatic trends. From this one must make inferences as to for- merly active geomorphic processes. 15
Post-formation modifications of features,such as those caused by weathering, erosion, deposition, and land use, have undoubtedly somewhat obscured the original form of these features. The general lack of know-
ledge concerning the mechanics of formational processes, and the occurrence of similar forms under different environmental conditions have led to confusion and numerous, widely varying theories of origin.
According to Washburn's definition, soil mounds are not considered as patterned ground, as are the stone nets, stripes, and rings. The term stone net refers to the overall mesh of the surface pattern created by the stones. There has been a tendency in the past to consider the soil mounds and surrounding stone rings as separate features. However, when these 2 features are found in association with one another, they are closely related through both process and form.
Morphology
Mound fields appear to have 2 components which govern the distribution, shape, and size of individual mounds andassociated stone rings. The first defines a large scale region such as theDeschutes-Lthatilla Plateau, and is determined by climate, soil depth, ani bedrock. The second component controls local distributions within thislarger region. This control is related to slope, aspect, mound density,soil type, and relative soil depth.
Individual mounds are generally circular to ovoid in shape depending on their location relative to other mounds and the surrounding topography.
Often, the mounds are concentrically surrounded by a ring of rounded to sub-angular basalt fragments ranging from cobbles to boulders. Mound and stone ring shape can be described as a continuing succession related to 16
gradient, with general elongation being in the direction of slope (Figure 6).
Another important factor in determining mound shape is proximity to
other mounds. A number of stone rings may infringe upon one another during
formation, thus affecting the dimensions of the mounds and rings (Figure 7).
Partially formed mounds appear as lobes off of a larger soil mass (Figure 8).
Series of mounds and their accompanying stone nets appear to form runs in
relation to the general slope direction. This trend is correlated with
surface, and subsurface drainage patterns, and appears to be an integral part of the formation process.
Soil mounds and associated stone rings are not found on slopes greater than 10%. Stone stripes, exhibiting similar characteristics as
the nets, and formed by somewhat similar processes, are found on greater
slopes. Stone nets and soil mounds typically appear in groups on level plateaus and the broad, level to slightly sloping divides between stream valleys (Figure 9).
Soil mounds generally range from 1Cm to 20m in diameter, and rise from 70 cm to 120 cm above the surrounding ground. The heights of mounds located in the same area all rise to a common level, suggesting that at one time this was the soil depth for the entire region.
Mound density varies from a wide dispersion to very dense clusters in which mounds may actually coalesce. Stone rings may encircle mounds either partially or completely. They may range in width from 50 cm to 2m.
Often mounds arise out of large areas of stone pavement in which numerous basalt fragments, ranging from cobbles to boulders appear on the surface and throughout the thin layer of soil overlying bedrock. These latter areas appear as large surface drainage swales where the layers of soil have 17
Sorted Sorted Polygoni N.h ,,/ \ Strp.t 's...... --- ': - -. _u - .-. . ' ?k . i,_ ' -.'..- - .. . ,-,..- '. I. - - ... - ... ., - .. ; - a I '-
0 I --
Figure 6. Schematic diagram of patterned ground development. Mass wasting causes elongation of patterns as slope in- creases. (After Sharpe, 1938, Figure 5, p. 37.) iI3
B C O.
fines coarse / material r:r 7rrrv
Figure 7. Schematic diagram of the successive stages of stone net development. Notice how the develop- ment of forms affects shape. (After Dzulynski, 1963, Figure 1, p. 146.) Figure 8. soilmoundStereo mounds configurations. pair areof aerialstone rings.photos illustrating various soil The narrow dark areas surrounding Figure 9. stonedevelopmentStereo rings pair onareof a aerial elongatedplateau photos top. into illustrating stone stripes. soil mound As slope increases, 21
been eroded away, exposing a stony surface (Figure 10).
Composition
The 4 mounds dissected revealed similar internal characteristics.
The upper portions are composed of a number of layers of volcanic ash
interspersed with aeolian silts. These upper layers in turn rest upon a
more consolidated mass of silt loam. This extends down to a layer of
angularly fractured, in situ bedrock, which overlies the bedrock itself,
generally from 70 cm to lm beneath the top of the mound (Figure 11).
Small basalt fragments up to 3 cm in diameter were occasionally encountered
in random distribution within the soil mass. This is believed to be the
work of small burrowing animals or minor frost heaving.
Extensive soil sampling was done throughout each profile in both
vertical and horizontal arrays. This revealed no significant degree of
sorting of soil particles within the mound body.
Near the edge of the mounds, rocks, 6 cm and larger, begin to appear within the soil profile. As the soil depth above bedrock decreases, the rocks increase in number, density, and proximity to the surface, until what is essentially the edge of the mound is reached. In this area, stones are exposed on the surface, generally at the interface between the soil mound and its surrounding stone ring.
Stone ring profile characteristics are very clear in areas where soil is absent from between the stones. Stone ring cross-sections reveal a definite trend toward vertical sorting. Well rounded surface stones in the upper sections are approximately twice as large as those in the bot- toni of the profiles. Likewise, ring width decreases progressively with depth down to bedrock (Figure 12). 22
Figure 10. Grass topped soil mounds surrounded by a low lying drainage swale where numer- ous stones appear on the surface. a SQcm 1OOcn I.
-'' ):''-'<-_'', ..' Figure 11. Generalized cross-section of soil mound and stone ring. 'C'i''Z'- l", 24
Figure 12. Cross-sectional profile of two different stone rings. Knife and hand troll are 25 cm high. 25
Soil fines often fill the interstices between the smaller stone fragments near bedrock. This low-lying soil layer is found to have a relatively high water content due to the protective layer of rocks above and possibly a higher water holding capacity.
Depressions in the bedrock level are often found beneath stone rings.
This indicates greater access by erosional agents, and increased chemical and physical weathering of this area, as opposed to that beneath the mounds. Surface stones are generally rounded to sub-angular, grading down to sub-angular and angular with depth.
ORIGIN OF PATTERNED GROUND AND SOIL MOUNDS
IN NORTH CENTRAL OREGON
Much of the early research conducted in the biscuit scablands of
Oregon failed to consider the relationship between the soil mounds and the associated patterned ground. Generally the mounds were the focus of theorizing while the stone features were given only cursory treatment.
Early Accounts
The first written account of this unique landscape in Oregon came from Joseph LeConte in 1873. He considered the mounds, "the long hill- side furrows and ridges" and the "wide pebble-covered spaces" to be
"remnants of a general erosion of the surface soil." He felt the conditions which allowed this phenomena to occur were the finer more movable drift soils above a coarser, less mobile subsoil (LeConte, 1877).
Waters and Flagler, in 1929, essentially agreed with LeConte in attributing the formation of mound features on the Columbia Plateau to surface erosion resulting in the exposure of weathered basalt bedrock between the mounds. They analyzed the stone nets and stripes as "erosion
furrows" and believed "th intervening areas between the mounds form a definitely integrated and minutely adjusted drainage system. They did
not feel that different climatic conditions than those currently pre- vailing were necessary to cause formation of these features (Waters and
Flagler, 1929).
Knetchel saw erosion as the principal formation agent. His conclusion, upon examination of aerial photographs of the Shaniko area, was that the
intermound furrow pattern was "genetically associated with columnar joint- ing in the underlying basalt." He believed that the erosive force of surface runoff was concentrated in the bedrock joint patterns, thus leav- ing the high centered mounds relatively unaffected by this erosion (Knetchel,
1952).
Kaatz, in 1952, presented the theory that intensive frost action under a former periglacial climate was the formative agent for the mounds and associated patterned ground features. He was the first author to mention the association between these landscape features and the shallow aeolian deposits upon which they are found throughout the Pacific North- west. In addition to a formerly much colder climate, he believed the ba- salt bedrock layer served to restrict subsurface drainage, and thus allow for more severe frost action. Since mounds are not found where the soil mantle exceeds 2m, he speculated that the active freezing level penetrated only to this depth during the former climatic conditions. Stone nets were believed by him to have formed prior to the mounds, and scablands under- went intensive frost action subsequent to mound formation. He added that, although frost action and solifluction are currently occurring in 27
this area, they are not of sufficient magnitude to result in the formation
of these features (Kaatz, 1959).
In an article concerning the periglacial realm in North America during
the Wisconsin glaciation, Brunnschweiler stated a theory similar to
Kaatz's. However, he believed that the soil mounds and their associated
stone rings developed simultaneously. In addition to this, he postulated
the existence of a permafrost layer in the relatively thinly mantled areas, with an active layer at a depth of from l.5m to 2m (Brunnschweiler, 1962).
Pleistocene Environmental Conditions
The condition necessary for the formation of sorted nets and soil mounds are: 1) a relatively shallow soil mantly overlying an impermeable
layer, either permafrost or bedrock; 2) sufficient moisture for the work of intense frost action, the formation of ice lenses in the soil, and solifluction, and; 3) a temperature regime characterized by numerous freeze-thaw cycles, either seasonally or diurnally.
Evidence of a colder and wetter climate in this region during the
Pleistocene Era can be inferred from a number of factors. The Deschutes-
Umatilla Plateau is located between the southern edge of the formercon- tinental ice sheet as it extended into Washington, and the extensive ice cap which covered the Cascades (Figure 13).
The present mean annual temperature of Shaniko, Oregon is 9.1°C.
From the analysis of cirque elevation data in the Cascades, the Pleistocene climatic snowline is believed to have been in the vicinity of l67ci, for the western region of the Plateau. The current summer climatic snowline for this area is approximately 3350m. A difference of l680m and a lapse rate of 6°C/bOOm would have caused the temperature regime for this region to be 123' 121' 120' *19' lIe' 117' 124' 49.
"5 1
- _4Y :::____ - ( - ' ?J- 4$' 48' .'- -.- c?" r
I t. 47. ç
46'.
Dshjtes-lJr.aiIIc PIc!e.0 7.1-' '3.
4-4. 44.
43,
iIo..t.n
42 42'
$24. 123 122' $21' *20' II?' lIe' 117'
Glaciated Areac jPluvicI Lakes ? Uncertain Ss,.t WnLi ...J fe., 1963. TI. O..s..y I 'is. Us.I.4 Srn., j ILsI 1971 C1.c.l 4 G.sIsg
Figure 13. Maximum extent of glaciation and pluvial lake development in relation to the study area during the Quaternary. periglacial in nature, with a mean annual temperature below freezing. In
a somewhat similar comparison, Brunnschweiler has found the lower limit
of currently active ground frost to be at least 1600m above the level he
believed existed during the Pleistocene (Brunnschweiler, 1962).
It must be noted, however that the climatic factor of the number of
freeze-thaw cycles of the air temperature is not necessarily reflected by
a similar trend within the soil mass. Ambient air temperature's effect
on ground temperature is modified by the insulating effect of snow,
vegetation, and the characteristics of the soil mass itself. The effective-
ness and frequency of freeze-thaw cycles must also be viewed in terms of
the attenuation of temperature fluctuations with depth.
Pollen records from this area also indicate that a cooler, moister climate existed up to about 15,000 years ago, coinciding with the glacial maxima (Huesser, 1960). The moderate maritime clinatic influences and the development of pluvial lakes in the Great Basin during that time also indicate a previously moister environment (Hansen, 1946 and Morrison,
1965).
Similar sorted patterned ground features as those found here are currently forming in contemporary periglacial regions. This suggests that a similar climate at one time existed in the region of the Deschutes-
Umatilla Plateau. This latter fact has also helped in developing theories as to formational processes of these relict forms.
Processes Involved in Formation
The formation of sorted stone rings is the result of the ejection of stones from the soil mass by frost action. Upon freezing, water under- goes a volume expansion of approximately 9%. Water held within the soil 30 mass by the impermeable bedrock freezes and causes the entire soil mass to expand, carrying imbedded stones with it. It has been demonstrated that pressure generated by the growth of ice crystals is at right angles to the freezing isotherm (Taber, 1930). The heterogeneity of a soil mass in terms of particle size, water content, and varying thermal conductivi- ties, may influence the orientation of the cooling surface, and thus allow for expansion pressures in lateral, as well as horizontal directions
(Corte, 1962). Soil movement in an upward direction is referred to as frost heave, while that in a horizontal direction is referred to as frost thrust.
The processes involved in the mechanics of the upheaving of stones are numerous and many are little understood. The cavity left by dis- placement of the stone upon frost expansion is subsequently narrowed by lateral frost thrusting (Figure 14). In addition to this, during sub- sequent thawing and contraction of the soil mass, soil fines tend to adhere to one another and collapse around the base of the stone. These two processes, referred to as frost-pull mechanisms, prevent the stone from returning to its original position.
Another process involved in the sorting of stones from the soil mass is a result of the greater heat conductivity of stones in relation to soil fines. Due to this fact, ice tends to initially form around stones (Hogbom, 1914, p. 305) or at their bases (Washburn, 1973,p. 21).
Soil water is then preferentially drawn to the freezingarea around the stone, causing significant accretions of ice in this area, and consequently increasing frost expansion (Washburn, 1973,p. 76). Return of the stone to its original position is again prevented by slumping of adjacent soil 31
Figure 14. Sequence illustrating upheaving of stone (actual size, 3 cm in dia- meter). As frost line approaches stone (1), the soil above the stones heaves, creating a void (2). Stone is lifted by the grip of the frozen soil (3). Cavity be- neath stone has narrowed and filled with water, which has frozen (4). Entire sequence occurs in 79 hours. (After King, 1976, p. 45.) 32
material into the cavity left by the upheaving action. This entire pro-
cess is referred to as the frost-pull mechanism (Washburn, 1973,P. 76).
After a series of freeze-thaw cycles, whichactivate these processes
and initiate local differential heaving,stones are ejected radially out-
ward from the initial freezing nuclei. The maximum upfreezing of stones
was found to be related to the ston&s surface area,or more specifically
the "effective height", or vertical dimension (Washburn,1973, p. 71).
This explains the vertical sorting by sizefound in the stone rings.
The sorting action which separatesstones into concentrated areas
tends to encourage drainage in theseareas. Initially, the stone network
serves as a subsurface drainage area, draining the soil adjacentto the
rings. Eventually, through piping, erosion, and solifluction,the ground
surface is lowered in the stoneareas, and surface drainageways are
formed around the now high centered soilareas surrounded by stone rings.
The higher mound surfaces represent the originalsoil surface height prior
to erosion.
The large intermound areas composed ofstone pavement appear as sur-
face drainage swales around mounds and adjacentto the stone rings. The
soil within the stone pavement and in thelower levels of the stone rings
is believed to be a result of solifluctionand aeolian deposition. The melting of ice lenses within the soilmass produced a super-saturated
soil condition, particularly conduciveto solifluction. This resulted in
the progressive distortion of the originalcircular shape with an increase
in slope.
The present climate is not of sufficient intensityto produce the sorted stone patterns as are found in fossil formson the Deschutes- 33
Urnatilla Plateau. Evidence of the inactivity of these forms is apparent
in the lichen encrusted surface rocks, the undisturbed zones of vegetation encroachment in the stone rings, and the primarily rounded surfaces of exposed stones. This landscape phenomenon has been preserved in its present form by the establishment of low lying, rock filled drainageways and the relatively heavy vegetation covering, protecting the soil mounds from further erosion.
CONTEMPORARY GEOMORPHI C ACTIVITY
Diurnal freeze-thaw cycles are very frequent in the Plateau area, especially as the elevation increases away from the Columbia River.
Though the frequency of these cycles far exceeds that found in arctic and subarctic environments (Fraser, 1959), the intensity, compared to these latter areas, is much less. The excessive, and sometimes standing water in the low lying areas of stone pavement during periods of thaw, restricts frost action to low penetration and short duration.
On the Plateau, the contemporary frost action results in the heaving of small, angular stones, exposing them in areas of stone pavement. These appear in contrast to the larger, rounded, lichen encrusted stones pre- viously exposed (Figure 15).
During the fall, winter, and early spring, the ground is frequently frozen solid. Needle ice is often encountered in the surface soil layer and beneath surface Objects.Small frost boils form where soil is forced up by frost expansion through stony areas, or through the frozen crust
(Figure 16). These small areas, seldom more than 40 cm in diameter, are subject to small scale solifluction, erosion, and contraction cracking
(Figure 17) with the seasonal change in climatic conditions. These result 34
Figure 15. Small stones heaved by contemporary frost action in thin-mantled stone pavement area. 35
, k
Figure 16. Frost boil formed by contemporary frost action. Knife is 25 cm long. 36
Figure 17. Contraction cracking in frost susceptible soil under contemporary conditions. Knife is 25 cm long. 37 in the formation of debris islands within the matrix of the stone rings or stone pavement. Rubble produced by current frost shattering of surface stones is common also (Figure 18).
Current wind and water processes are particularly effective on the loose, friable surfacesoils which result from repeated frost action.
These processes resultin continued exposure of surface stones in some areas through erosion. In other areas, deposition of these materials results in filling theinterstices between stones, and partial burial of some features.
Wind and water acUon are largely seasonal in nature. During periods of thaw in the spring, soil in the intermound areas is saturated and sticky due to the relative abundance of water. Conditions are parti- cularly conducive to solifluction. During the summer, the lack of moisture precludes significant amounts of chemical weathering despite the warmer temperatures. However, the action of wind is relatively effective due to the dry, unprotected soil surface.
CONCLUSIONS
The evidence obtained from my research has led to a theory of origin based on the interaction of a number of geomorphic processes as previously discussed. The reconstruction of former climatic conditions indicated the existence of a much colder environment, with increased amounts of moistur2, especially during the interglacial periods of the Pleistocene. Climatic changes during this latter period, as evidenced by numerous cycles of glacial advance and retreat, have led to the suggestion that patterned ground formation processes may have been reactivated numerous times during the Pleistocene. 38
Figure 18.Accumulation of frost shattered rubble. The relatively undisturbed layers of bess within the mounds, and
the equal height of mound surfaces within specific areas suggests that
mounds represent remnants of the original soil surface, prior to erosion.
The stone rings encircling the mounds, the existence of numerous stories
in the thin soil mass of the inter-mound area, and the fact that these
latter areas serve as large surface drainage swales point to the importance
of the role of water induced erosion in the origin and maintenance of
these landscape formations.
The steps involved in the creation of stone rings and soil mounds
are as follows: 1) the sorting and ejection of stones from specific
areas within the soil mass as a result of frost action (frost heave and
thrust) initiated by numerous intense freeze thaw cycles; 2) the develop- ment of a subsurface drainage network related to the patterns produced by stone concentrations in the soil mass; 3) the lowering of the soil surface through piping and erosion in areas where subsurface drainage is encouraged by these concentrations of stones, and; 4) the relatively low rate of erosion in the areas from which stones were heaved, causing these areas to be high centered mounds surrounded by an extensive drain- age network.
The location of the soil mounds was determined by the presence of nuclei of intense frost action within the heterogeneous soil mass. The locations of these nuclei were controlled principally by factors affecting the microclimate of the surface such as aspect, elevation, and the in- sulating effects of vegetation and snow cover. Varying soil conditions allowed for a heterogeneous distribution of soil moisture, thus affecting the location and intensity of frost action within the soil mass. As these core areas increased in size, the stone rings they produced impinged upon one another, thus affecting their shape. On sloping land, stone features and soil mounds tended to be elongated, under the influence of gravity,
in the direction of slope. Soil mound size was controlled by the potential area for stone ring growth surrounding each frost action nucleus, and by the intensity and duration of frost action at specific sites within the soil mass.
Eventually, as environmental temperatures increased and moisture became less abundant, the geomorphic processes were no longer occurring at sufficient intensities to produce these large scale features. Current geomorphic processes operative on the Deschutes-Umatilla Plateau continue to produce small scale frost features as previously discussed. However, the large stone networks and soil mounds are merely representatives ofa single point of time on the environmental continuum. Such relict features as these represnt the morphoclimatic realm as it existed thousands of years ago, and from which our current polygenetic landscape has evolved.
SUGGESTIONS FOR FURTHER RESEARCH
Clearly there is a need for more research and the development ofa more comprehensive view of the factors responsible for the formation of all forms of patterned ground. This would be of value in studying both contemporary and relict forms and the processes of which they are a result.
Relict patterned ground has often been used as a "window,"a source of insight as to environmental conditions which existed at the time of formation. The ability to more precisely reconstruct former environments from current evidence would provide a stronger basis for continued research in this field. 41
The distribution of soil mounds on the Deschutes-Umatilla Plateau
appears to be strongly related to variation in soil conditions, most
notably depth. Further study of this relationship here and throughout
the Pacific Northwest is needed. Likewise, comprehensive environmental
comparisons, both historical and contemporary, or areas with similar
forms would help determine the extent to which the theory presented in
this paper is applicable.
In many areas on the Plateau, successive layers of volcanic ash
cover the soil mounds. Tephrochronology would yield valuable information
as to the relative time of deposition and mound formation, and any effects
the existence of these layers may have had on the formation of patterned
ground.
Trend surface and computer analysis of data from large areas could
be used to verify and determine new trends as they may relate to such
factors as size, aspect, shape, and environmental conditions. Further
exploration could yield valuable statistical information concerning dependent and independent variables related to the existence and distri- bution of mound fields.
A technique of great potential value in such studies is the use of
thermal infrared imagery. Images of patterned ground areas give valuable
insight into the distribution and movement of moisture within the soil mound and stone net complex. This moisture component of the environment
is intimately related to the theory of origin presented in this paper.
Current patterns of soil moisture distribution and movement are closely related to those in the past. Further knowledge of these processes would tend to lead to a greater understanding of the actual extent to which this 42 factor affected various patterned ground characteristics.
Finally, a detailed analysis of current rates of weathering and stone movement under varying climatic conditions is needed to help esta- blish a perspective on contemporary modification and the continuing evolution of the soil mounds and stone nets. 43
BIBLIOGRAPHY
Arkley, R. J. 1948. "The Mima Mounds," Science Monthly, Vol. 66, pp. 175-176.
Bretz, J. H., H. 1. U. Smith, and G. E. Neff. 1956. "Channeled Scab- lands of Washington: New Data and Interpretations," Bulletin, Geological Society of America, Vol. 67, pp. 957-1049.
Brunnschweiler, D. 1962. "Periglacial Realm in North America During the Wisconsin Glaciation," jyj Peryglacjalny, Vol. 11, pp. 15-27.
Bryan, K. 1928. "Glacial Climate in Non-glaciated Areas," American Journal of Science, Vol. l6, pp. 162-164.
Chambers, M. J. G. 1966. "Investigations of Patterned Ground At Signy Island, South Orkney Islands: Interpretations of Mechanical Analyses,t' British Antarctic Survey, Vol. 9, pp. 21-40.
Corte, A. E. 1962. "The Frost Behavior of Soils: Laboratory and Field Data for a New Concept, Part II. Horizontal Sorting," United States Arm Cold Regions Research and Engineering Laboratory Research Report 85, Part II, 20 pages.
Daiquest, W. W. and U. B. Scheffer. 1942. "Origin of Mima Mounds in Western Washington," Journal of Geology, Vol. 50, pp. 68-84.
Dyksterhuis, E. L., G. H. Simonson, J. A. Norgren, and R. E. Hosler. 1969. "John Day Drainage Basin, General Soil Map Report with Irrigable Areas," Oregon's Lon9-Range Requirements for Water, Appendix 1-6, Oregon State Water Resources Board, 101 pages.
Dzulynski, S. 1963. "Polygonal Structures in Experiments and Their Bearing on Some 'Periglacial Phenomena' ," Bulletin De L'Academie Polonaise Des Sciences, Vol. 11, No. 3, pp. 145-150.
Embleton, C. and C. A. M. King. 1968. Glacial and Periglacial Geo- morphology, Edward Arnold Ltd., London, 608 pages.
Fosberg, M. A. 1965. "Characteristics and Genesis of Patterned Ground in Wisconsin Time in a Chestnut Soil Zone of Southern Idaho," Soil Science, Vol. 99, No. 1, pp. 30-37.
Franklin, J.F. and C. 1. Dryness. 1969. "Vegetation of Oregon and Washington," U.S.D.A. Forest Service Research Paper PNW-80, Pacific Northwest Forest and Range Experiment Station, Portland, Oregon 216 pages.
Fraser, J.K. 1959. "Freeze Thaw Frequencies and Mechanical Weathering in Canada," Arctic, Vol. 12, pp. 40-53. 44
Green, G. L. 1970. "Soil Survey of Trout Creek-Shaniko Area, Oregon," U.S.D.A. Soil Conservation Service, U.S. Forest Service, and Oregon Agricultural Experiment Station, 83 pages.
Hansen, H. P. 1946. "Succession and Climate in the Oregon Cascades," American Journal of Science, Vol. 244, pp. 730-731.
Hays, 1. 1943. "Notes on Glacial Erosion and Stone Stripes," Geographical Journal, Vol. 102, pp. 13-20.
Henderson, E. P. 1968. "Patterned Ground in Southeast Newfoundland," Canadian Journal of Earth Sciences, Vol. 5, pp. 1443-1453.
Heusser, C. J. 1960. "Late Pleistocene Environments of North Pacific North America: An Evaluation of Late-Glaci&i and Post-Glacial Climatic, Physiographic, and Biotic Changes," American Geographic Society Special Publication No. 35, 308 pages.
Hogbom, B. 1914. "Uber Die Geologische Bedeutung des Frostes,' .2ja University Geol. Inst. Bull. No. 12, pp. 257-389.
Hunt, C. B. and A. L. Washburn. 1966. "Patterned Ground," U.S. Geological Survey Professional Paper No. 494B, pp. 104-133.
Hussey, K. M. and G. S. Anderson. 1963. Environment and Distribution of Thermal Relief Features in the Northern Foothills Section, Alaska, Dept. of Geology, Iowa State University, Ames, Iowa, 76 pages.
Kaatz, M. R. 1959. "Patterned Ground in Central Washington," Northwest Science, Vol. 33, No. 4, pp. 145-156.
King, C. A. M. (ed.). 1976. Periglacial Processes, Dowden, Hutchinson, and Ross, Inc., Stroudsburg, Pennsylvania, 445 pages.
Knetchel, M. M. 1952. "Pimpled Plains in Eastern Oklahoma," Bulletin, Geological Society of America, Vol. 63,pp. 689-699.
LeConte, J. 1877. "Hog Wallows or Prairie Mounds (of California and Oregon)," Nature, Vol. 15,pp. 530-531.
Malde, H. E. 1961. "Patterned Ground of Possible Solifluction Origin at Low Altitudes in Western Snake River Plain, Idaho," U.S. Geological Survey Professional Paper No. 424B, pp. 170-173.
1964. "Patterned Ground in the Western Snake River Plain and Its Possible Cold Climate Origin," Bulletin, Geological Society of America, Vol. 75, pp. 191-208.
Miller, R.R. and R. W. Galloway. 1954. "Stone Stripes and Other Surface Features of Tinto Hill," Geographical Journal, Vol. 120, pp. 216-219. 45
Newcomb, R. C. 1969. "Geology of the De3chutes-Umatilla Plateau,"pp. 60-66 in A. E. Weissenborn (ed.) ineral and Water Resources of Oregon, Bulletin 64, State of Oregon üepartmentof Geology and Mineral Industries, 462 pages.
Norgren, J. A. and C. H. Simonson. l9. "Deschutes Drainage Basin, General Soil Map Report with Irrigabie Areas,"Ors Long-Ran9e Reguiremer1ts for Water, Appendix 1-5, OregonState Water Resources Board, 116 pages.
Olmsted, R. K. 1963. "Silt lounds of Missouia Flood Surfaces," Bulletin, Geological Society of America, Vol. 74,Pp. 47-54.
Pyrch, J. B. 1973. The Characteristics and Genesis of Stone St es in North Central Oregon, Masters thesis,Department of Geography, Portland State University, Portland, Oregon,123 pages.
Ritchie, A. M. 1953. "Erosional Origin of Mima Mounds," Journalof Geology, Vol. 61,Pp. 41-50.
Ruhe, R. V. 1959. "Stone Lines in Soil," Soil Science, Vol. 87,pp. 223-231.
Sharpe, C. F. S. 1938. Landslides and Related to, Columbia University Press, New York, 137pages.
Smith, H. L U. 1949. "Physical Effects of Pleistocene Climatic Changes in Non-Glaciated Areas," Bulletin, GeologicalSociof America, Vol. 60,pp. 1485-1514.
1962. "Periglacial Frost Features and RelatedPhenomena in the United States," Biuletyn Peryglacjalny, Vol.11, pp. 325-342.
Taber, S. 1930. "The Mechanics of Frost Heaving," Journal ofGeo1og, Vol. 38, pp. 303-317.
Tricart, J. 1970. Geomorpholoqyof ColdEnvironments, (Translated by Edward Watson), St. Martins Press, New York, NY, 320 pages.
Washburn, A. L. 1956. "ClassificationofPatterned Ground and Reviewof Suggested Origins," Bulletin, GeologicalSociety ofAmerica, Vol. 67, pp. 823-866.
1969. "Patterned Ground in the Mesters Vig District,Northeast Greenland," Biuletyn Peryg1acjaln, Vol. 18,pp. 259-330.
1973. Periglacial Processes and Environments,St. Martin's Press, New York, NY, 320 pages.
Waters, A. C. and C. W. Flagler. 1929. "Origin of the Small Moundson the Columbia River Plateau," AmericanJournalofScience, Vol. 18, pp. 209-224.