7 Day SE0s10n and Deposition in the Loess—Mantled Great Plains, Medicine Creek Drainage Basin, Nebraska i U f} , GE EOLOGICAL SURVEY PROFESSIONAL PAPER 352—H Preparea’ as part of a program of t/ze Department of tfie Interior for development of tae Missouri River oasin and part of Me soil and moisture program JAN 20 1967 ' 4, e Minor “‘5” Erosion and Deposition in the Loess—Mantled Great Plains, Medicine Creek Drainage Basin, Nebraska By JAMES C. BRICE EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT GEOLOGICAL SURVEY PROFESSIONAL PAPER 352—H Preparea’ as part ofa program of toe Department of tae Interior for development oft/1e Missouri River oasm aaa’ part of t/ze soil arzaI moisture program UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1966 UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary GEOLOGICAL SURVEY William T. Pecora, Director For sale by the Superintendent of Documents, Washington, DC. 20402 —— US. Government Printing Ofiice Price 55 cents (paper cover) CONTENTS Abstract _ Introduction _ Climate, by Cloyd H. Scott ................................................ Vegetation and paleoclimate _ Climax plant communities ___________ Characteristics of native grasses ______________________________ Effects of drought ________________________________________________________ Effects of land use , Native trees and dendrochronology __________________________ Pre-Pleistocene rocks Unexposed rocks Exposed rocks Niobrara Formation ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Ogallala Formation ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Pleistocene and Recent deposits ________________________________________ Terraces ,, General stratigraphy of the Pleistocene and Recent Stratigraphy on Cedar Creek near Stockville ,,,,,,,, Stratigraphy on Cut Canyon near Curtis ,,,,,,,,,,,,,,,, Stratigraphy on Elkhorn Canyon near Maywood , . Loveland Formation and Sangamon soil ________________ Peorian Loess, Brady soil, Bignell Lo‘ess, and modern soil ,, Stockville terrace deposits __________________________________________ Mousel terrace deposits ______________________________________________ Recent alluvium 1 Significance of carbonized fragments of grass ,,,,,,,, Sand dunes Archeology and human occupation ____________________________________ Results of archeologic investigations Settlement and land use ______________________________________________ Page 255 255 258 262 262 262 263 263 264 264 264 264 264 265 265 265 268 270 271 272 274 275 278 278 27 8 280 280 280 280 281 Drainage system Evolution of the drainage system ____________________________ Morphometry Drainage transformation and variations in drain- age texture __ Gully erosion ., Definition and classification of gullies ____________________ Age and activity of erosional scarps ________ Channel scarps and valley-bottom gullies ________________ Development of valley—bottom gullies ,,,,,,,,,,,,,,,,,,,,,, History of late Recent gullying on Dry Creek _____ Valley-head and valley—side gullies __________________________ Measurement of valley-head and valley-side gullies on Dry Creek ____________________________________________________________ Areal distribution of gullies _________________________________ Formation of valley-head and valley—side gullies Gullying and land use Control of gullies ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Suspended sediment and the hydraulic geometry of channels, by Cloyd H. Scott ____________________________________________ Hydraulic relations at a section and in a down- stream direction Suspended-sediment discharge relations __________________ Particle-size distributions of suspended sediment and bed material ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Relative sediment contributions of the major tribu— taries Geomorphic properties in relation to water and sedi- ment discharge Summary and conclusions __________________________________________________ References Index . ILLUSTRATIONS Page FIGURE 175. Regional physiographic map 256 176. Location map ,,,,, , 257 177. Photograph of typical relief in southern part of basin __________________________________________________ 258 178. Graph of average monthly temperature and maximum, average, and minimum monthly precipitation at Curtis, Nebr., 1931—55 ____________________________________________________ 259 179—180. Graph of annual precipitation, 1895—1960— 179. Curtis, Nebr. ,,,,,, . 260 180. McCook, Nebr. .. 261 181. Graph of rainfall intensity—duration-frequency curves for Curtis, Nebr., 1951—58 262 182. Photographs showing effect of land use on native vegetation ______________________________________ 264 183. Geologic section across lower tip of basin 266 184. Geologic section across the basin A- 267 185. Section showing general topographic and stratigraphic relations of Pleistocene and Recent deposits 268 186. Channel profile, valley flat profile, and terrace profiles along the main course of Medicine Creek _ 270 III Page 282 282 284 288 290 290 291 291 295 300 301 304 307 309 313 314 315 315 321 IV FIGURES 187—189. 190—193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222—223. 224—225. 226—231. fl CONTENTS Profile and geologic section— 187. On Cedar Creek 1 188. Along road entering Cut Canyon 189. On Elkhorn Canyon Photographs showing— 190. Terrace sequence and valley fills along Elkhorn Canyon ________________________________ 191. Valley fill of the Peorian Loess set unconformably against Sangamon soil and Peorian Loess ,,,,,,,,, 192. Peorian Loess overlying A horizon of Sangamon soil ______________________________________ 193. Banding in Recent alluvium, Dry Creek Block diagram showing evolution of the relief since deposition of the Peorian Loess ,,,,,,,, Photographs of remnants of the side slopes and valley flats of former drainage systems 1111 ,,,,,,,,,,,,,, Map showing area of capture of North Plum Creek by Deer Creek ............................ Maps showing variations in drainage texture ,,,,,,,,, Graph showing mean channel length in relation to channel order .................................... Graph showing number of channels of each order in relation to channel order .............. Composite sketch showing varieties of valley-bottom gullies ________________________________________ Map of head scarps in major valley-bottom gullies 1 Profiles of valley and channel for some major tributaries ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Graph showing local valley slope in relation to drainage area ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Photographs of typical valley-bottom gullies 11 Profile showing changes in a large valley-bottom gully on Dry Creek ,,,,,,,,,,,,,,,,,,,,,,,,,,,, Profiles of a large valley-bottom gully in a tributary of Curtis Creek Canyon ............ Profile showing changes in a gullied tributary of Dry Creek ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Profiles of channel, valley flat, and terraces on Dry Creek ______________________________________________ Cross profiles of Dry Creek ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Photographs showing downstream changes in channel of Dry Creek ,,,,,,,,,,,,, _1 .............. Composite sketch showing varieties of valley-head and valley-side gullies ................... Photographs of valley-head and valley-side gullies Graph of cumulative frequency distribution of enlargements of valley-head and valley— side gullies on Dry Creek 1 Aerial photograph of a severely gullied area on upper Dry Creek .................................. Maplshowing areal frequency distribution of active valley-head and valley—side gu 1es ,,,,,, 1 ,,,,,, 1 Graph showing frequency of active valley-head and valley-side gullies in relation to area of upland and adjusted frequency of first-order channels ,,,,,,,,,,,,,,,,,,,,,,,,,,,, Profiles and plan View of a large valley-head gully Profile of a large valley-head gully ,,,,,,,,,,,, 11 Photograph of the head scarp of a small valley-head gully on Dry Creek ,,,,,,,,,,,,,,,,,,,,,, Hydrographs of daily mean flows equaled or exceeded about 1 percent of time ............ Graph showing change of width, depth, and velocity with increasing discharge at a section, Dry Creek near Curtis 11 1 Graph showing change of width, depth, and velocity in a downstream direction for channels of— 222. Perennial streams 1 223. Ephemeral streams ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Graph showing change of suspended-sediment load with increasing discharge at a section— 224. Perennial streams 1 1 225. Ephemeral streams ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Graph showing—- 226. Size distribution of suspended sediment 227. Size distribution of bed material 1 11 11 228. Cumulative water discharge and sediment discharge per square mile __________ 229. Mean channel slope in relation to channel order 111111111111111111111111111111111111111111111 230. Adjusted frequency of first-order channels in relation to mean valley-side slope, mean slope of first-order channels, and mean length of first- order channels 1 11111 1 1 231. Slope of first-order channels in lower Dry Creek subbasin in relation to length 111111111111111111111111111111111111111111111111 1 Trial graph showing multiple linear regression of sediment discharge in relation to relief ratio, percentage of upland, and adjusted frequency of first-order chan- nels 1 11 11 Page 271 272 273 273 274 276 279 283 284 285 287 289 290 292 293 295 296 297 298 299 300 301 302 303 304 305 306 308 310 311 312 313 314 316 317 319 320 323 324 325 326 327 328 329 330 331 —’— TABLE 1. 99°?“ CONTENTS TABLES Thickness and percentage of each rock type in seven exposed sections of the Ogallala Formation __________________________________________________________________________________________________________________________________ Correlation of stratigraphic units and terraces of the Pleistocene and Recent in Ne- braska and Wyoming _________________________________ _ ____________________________________________________________________________________ Measured and derived properties for the M dicine Creek basin and its major subbasins ...... Drainage system properties, listed according to channel order, for major subbasins of the Medicine Creek basin _____________________________________________________________________________________________________________ Measurements relating to the size and topographic setting of major valley-bottom gullies ,, Number of gullies according to type and size in the Medicine Creek basin and in major subbasins _______________________________________________________________________________________________________________________________________ Gully frequencies and related data for subbasins of the Medicine Creek basin _ Values of b, f, and m for streams in Medicine Creek basin ____________________________________________________ Areas in upland and valley flat in comparison with areas in two categories of land use Page 265 269 288 294 299 307 309 321 330 EROSION AND DEPOSITION IN THE LOESS-MANT LED GREAT PLAINS MEDICINE CREEK DRAINAGE BASIN, NEBRASKA By JAMES C. BRICE ABSTRACT The Medicine Creek basin is representative of the loess- mantled Great Plains in climate, surface materials, and relief forms, but its relief, degree of dissection, and rate of erosion are higher than average. In operation, if not in rate, processes of erosion and deposition in the basin are considered to be representative of loess-mantled regions in a semiarid or subhumid climate. About 12,000 years ago, the valley fill of the Peorian Loess was incised and the drainage system was ramified to approximately its present extent. During the subsequent episode of deposition (about 1,000—5,000 yr ago), the deposits of the Stockville terrace accumulated and the valley sides were graded. The formation of the Stockville terrace was followed by a minor episode of deposition, another of valley incision, and finally the accumulation of the late Recent alluvium, which has been intermittently incised during the past 500 years. The property of the drainage system that is most useful in accounting for the areal distribution of gullies is called adjusted channel frequency, in the derivation of which the drainage basin area is adjusted by subtracting the area of upland. The slope of the exponential curves, relating mean channel length to channel order, changes at the lower channel orders because of drainage transformations, in late Recent time, that affected only channels of low order. Gullies in the basin are widened, lengthened, and deepened mainly by scarp erosion. They are classified on the basis of topographic location as valley-bottom, valley-side, and valley— head gullies. Trenching of a valley reach to bedrock takes place not only by coalescence of discontinuous gullies but also by the successive upstream migration of several channel scarps. Many large valley-bottom gullies have evidently originated on locally steepened valley reaches, but the ratio of local slope to drainage area that is critical for the initia- tion of a gulley is not sharply defined. Valley-bottom gullies advance mainly because of plunge—pool action, but this mechanism is less important in the advance of valley-side and valley-head gullies. The areal frequency distribution of valley-head and valley-side gullies is correlated with two geomorphic properties: percentage of area in upland and adjusted frequency of first-order channels. Hydraulic geometry of the channels is expressed by the slope of curves relating width, depth, and velocity to dis- charges equaled or exceeded 1, 2, and 25 percent of the time. At the gaging stations on Dry, Mitchell, and Brushy Creeks, the flow is categorized as ephemeral; at the others, as peren- nial. In a downstream direction, an increase in discharge of ephemeral streams is accommodated by relatively large changes in depth and velocity, but the increase in discharge of perennial streams is accommodated by a relatively large change in width. Curves relating suspended—sediment load to water discharge for the perennial streams are characterized by a break in slope at water discharges above normal but below the bankfull stage. In general, the concentration of measured suspended sedi- ment is higher in wet years than in dry, and the ephemeral streams have higher concentrations than the perennial streams. Differences in runoff and sediment discharge among five subbasins are attributed mainly to differences in relief ratio, adjusted frequency of first-order channels, and per- centage of area in upland. Active valley-head and valley-side gullies, and all but a few of the active valley-bottom gullies, are attributed to land use since settlement rather than to climatic change. Restoration of native vegetation to the heads of valleys, together with conservation measures on the upland, would be effective in the control and prevention of valley—head gullies, which are the most numerous in the basin. INTRODUCTION The purpose of this report is to give an integrated account of the geologic and hydrologic factors that are pertinent to erosion and deposition in a part of the loess-mantled Great Plains and to evaluate the major factors involved in gully erosion, channel deposition, and the discharge of water and sediment through channels. Climatic change as a cause of modern erosion is evaluated in the light of ecologic, stratigraphic, and archeologic evidence from terrace deposits of late Pleistocene and Recent age. Proc- esses of scarp erosion are analyzed, gullies are classified, and the areal distribution of gullies is correlated with morphologic properties of the basin. Generalizations are made as to the hydraulic geome- try of the channels and the relations between dis- charge and suspended-sediment load. Differences in runoff and sediment discharge among five subbasins are attributed mainly to differences in specific mor- phologic properties among the subbasins. The Medicine Creek basin was selected by several cooperating agencies (U.S. Bur. of Reclamation, U.S. Agr. Research Service, and the Univ. of Nebraska) as a suitable area in which to investigate erosion and runoff. Collection of data, which was begun in 1951 and ended in 1958, was directed toward an 255 256 evaluation of soil and water conservation practices and also toward an understanding of the processes and causes of rapid erosion in this region, which is the major subject of the present report. The regional setting of the Medicine Creek basin, which is in the loess-mantled part of the Great Plains, is shown in figure 175. The processes of erosion and deposition in the basin are characteristic of loess-mantled regions in a semiarid or subhumid climate, and they apply specifically to an area of about 115,000 square miles that is within the Mis- souri River basin and is mantled with 8 feet or more of loess. Most of this area is within the Great Plains, although part is in the Interior Plains. The Medicine Creek basin is dissected to a much greater degree than many parts of the Great Plains, and the local relief is greater than average. Neverthe- less, nearly all the relief forms typical of the Great Plains occur within the basin, including areas of nearly flat upland and sharply dissected areas of narrow divides and flat-bottomed valleys. Gullies ——4_ EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT of all types and sizes occur, and the frequency of gullies ranges from high in some places to 10W in others. Medicine Creek, which is a tributary to the Republican River, has a drainage area of about 690 square miles that lies between the Platte and the Republican Rivers. Major subbasins and gen- eral geographical features are shown in figure 176. Remnants of the upland surface reach a maximum altitude of about 3,150 feet in the northern part of the basin and descend gradually to an altitude of about 2,400 feet at the bluffs of the Republican River. The maximum local relief, about 200 feet, is in the northern part of the basin. In general, the upland is more sharply dissected in the northern part, and the divides tend to be flat but narrow; in the central and southern parts, the divides are broader and slope gently toward the valleys. The general aspect of the relief in the southern part of the basin is shown in figure 177, which is an oblique aerial View southwest from a point above Medicine l .' SOUTH DAKOTA 5, 5 \R\ Kansas Cit Kai-““13 River EXPLANATION Mantled by 8—16 feet of Ioess Sand dunes Mantled by 16764 feet of loess Mantled by more than 64 feet of loess Q Medicine Creek basin FIGURE 175.—Map showing the regional physiographic setting of Medicine Creek basin. Distribution and thickness of loess generalized from Thorpe and others (1952). LOESS MANTLED GREAT PLAINS, MEDICINE CREEK BASIN NEBRASKA m . 10030 257 ); / NIL/KAY" \I‘):\ T. 11 N, KW yK \Q LK “Y “\K r W M < A, "' / :' :' _ : ‘\‘ \gi} z‘xl'z’ >/ (.13 . '. t '1’. 3 'IgY ._ . 3 WV\ Wellfleet Ca .\f\' . K' i243 :\\‘ \K V HAYES CO I: n . /' ’“a’c— A ._ f‘ gvfio /“ 2:) “9910.“ ‘ NJ \ 2 K a 33:23ch 2. 37-? Km" 33 W\ “H” %\ 3 .- \ \ WW‘Y’V ”3-) : Egg/f - .. 9) Curtis R.31 W. {\ (\ C '. '. | \ m \X \/~ — 40°30' R3OW. \: EXPLANATION Boundary of report area Boundary of subbasin 9 Stream-gaging station 5 MYLES |__1__;Fl__l———J FIGURE 176 ——Location map of Medicine Creek basm 258 FIGURE 177.——Oblique aerial photograph of typical relief in southern part of Medicine Creek basin. Medicine Creek at lower left. Photograph by River Basin Surveys, Smithsonian Institution, 1948. Creek and about 7 miles north of the town of Cambridge. Fieldwork for geomorphic purposes was begun in the summer of 1953 and continued during the sum- mers of 1954—56. Special hand-level and stadia—rod equipment was devised to permit the measurement of slopes without an assistant. Field studies, as well as office work, were facilitated by excellent topo- graphic map and aerial photographic coverage of the entire basin. On the 1 :24,000-scale topographic maps, prepared by photogrammetric methods and published by the Geological Survey in 1957—58, even minor topographic features are accurately repre- sented. The Munsell color system was used for field description of loess and soil. Data contributed by the cooperating agencies have been used in the preparation of this report. The detailed surveys necessary to measure rates of channel erosion on Dry Creek were made by the Bureau of Reclamation, who also contributed data on precipitation. Land use was tabulated by the Agricultural Research Service and the Soil Conser- vation Service. Runoff and suspended sediment were measured by the Geological Survey. Many persons contributed to this report, which was prepared under the successive supervision of P. C. Benedict, regional engineer, and D. M. Cul- bertson, district engineer, US. Geological Survey. The section on native vegetation was read by E. J. Dyksterhuis, range conservationist for the Soil Con- servation Service. Fossil snails were identified and information on snail ecology was given by W. J. Wayne of the Indiana Geological Survey. One car- bon-14 age determination was made by Meyer Rubin, geochemist of the US Geological Survey, and an— other by J. L. Kulp of Columbia University. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT CLIMATE By CLOYD H. SCOTT The climate of the Medicine Creek basin is typical of the interior of continents of the middle latitudes with rather light rainfall, hot summers, cold winters, large variations in precipitation and temperature from year to year, and frequent changes in weather from day to day. The mean annual temperature at Curtis, near the center of the basin, is about 52°F. Maximum and minimum temperatures above 100° and below 0°F occur occasionally. The coldest month is January when the average temperature is about 27°F, and the warmest month is July when the average tem- perature is about 78°. The monthly mean tempera- ture at the US. Weather Bureau Station at Curtis for the years 1931—55 is shown in figure 178. Much of the precipitation occurs during April through September, generally as thunderstorms. On the average, more than half the total annual pre- cipitation occurs from May through August. Pre- cipitation varies widely from the average (fig. 178) ; 7 months had no precipitation at least once during the period 1931—55, and some months had no pre- cipitation several times during that period. Maxi- mum amounts exceeded the averages by about two to seven times. The annual precipitation also varies widely from year to year (fig. 179). The minimum annual amount of slightly less than 11 inches was recorded in 1934, and the maximum of about 38.2 inches was recorded in 1915. The 66-year average (1895—60) was about 21.5 inches, but the 25—year average (1931—55) was only 19.1 inches. The 5-year moving averages (figs. 179 and 180) show the general trends of precipitation at Curtis and McCook from 1895 to 1960. The annual totals are highly variable from one year to another, but the 5-year moving averages show a somewhat cyclic trend. Even though the high— and low-precipitation years tend to be grouped somewhat, about half of the years have above-average precipitation amounts. However, during 1951—58 at Curtis, 5 years had precipitation below average; 4 of the years had more than 6 inches below average. The 5-year moving average for Curtis indicates a long-term trend toward decreasing annual pre- cipitation. The consistency of the precipitation data at Curtis was checked by plotting a double—mass curve of data at Curtis and at McCook. A pattern composed of data for McCook, Cambridge, and North Platte was first used; but the data for North Platte and Cambridge were inconsistent with the —>— LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 259 12 11 — Q - \ E 100 r- 10 — Q g - i \ «s Q E 90 - 9 — \ Q — \ \ \ ,_ \ \ E t \ Average monthly E 80 - 8 _ Q Q temperature _ Lu \ \ n: a \ \ I \ \ . . \ \ s L“ 70 - LZJ 7 —- \ \ \ _ u) _ \ \ \ LL] \ \ \ g; E i \ \ (D ‘ \ 3 60—3 6— 3 ~\ \ \ — F Annual average Q \ \ E < temperature \ \ \ \ ,_ l \ \ \ w s i \ g 50 " 8 \ \ _' ._ a \ \ \ \ \ < o. \ \ \ 3 \ E \ i gQ Q \ i n. 40 _ \ \ «A \ \ \ _ 2 \ \ ;\ \ \ \ L: t \ a t z t \ \ \ \ \ \ i 7: 7: t 2 t 30 - \ ¢\ /\ \ \ \ _ \ t t t t t \ \ \ Q /\ /\ 7\ Q \ Q 20 L \ /Q Q /§ \ \ _ \ \ E%\ /\ Q Q Q ¢\ =/\ /\ 7\ \ \ z E Q r: /: t t \ g \ %\ ¢\ \ x — \ \ ¢\ /\ 7\ \ \ \ /\ /\ \ \ \ i i /§ /§ \ . /\ /1\ \ I fix I g | \ 1 Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov Dec. FIGURE 178.—Average monthly temperature and maximum, pattern, whereas the data for McCook seemed to be consistent. The double-mass curve showed a trend toward more precipitation at Curtis beginning about 1908 but returned to about the original slope in 1914. There is no record that the gage was moved in 1908, although the location was changed in 1914 (US. Weather Bureau, 1955a). The pattern of the 5—year moving average for Curtis (fig. 179) is generally similar to the pattern for McCook (fig. 180) except that the precipitation for the period 1908—18 was higher at Curtis than at McCook and for the period 1940—52 was lower at Curtis than at McCook. The apparent secular de- average, and minimum monthly precipitation at Curtis, Nebr., 1931—55 crease in annual precipitation at Curtis is probably a result of uneven areal distribution and is not considered to be representative of the region. As shown by the double-mass curve, the data at Curtis may have been somewhat biased for the years 1908—14. Precipitation at Curtis averaged about 18.77 inches for the years 1951—58, or somewhat below the 66-year average of about 21.5 inches. Extremes of annual precipitation for this period ranged from 31.61 inches in 1951 to 12.36 inches in 1952. The occurrence of a high and a low in consecutive years is not unusual owing to the large variations in rain- fl 260 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 4O 38 '- _ 36 - _ 34 - _ 32 - — 3O — _ $3 I 28 - _ 2 - 5-year moving average (plotted at g 26 _ middle of 5-year period) — E z 24 _ 66-year average_ ’: E 8 22 ‘ V / g _.....-- ____ __.._ __ _-,_______ ______ _ scs _ _____L-: 20 — / _ fi/\ / 18 — ., I - 16 r _ 14 — — 12 _ l _ ,0 I l 5 25 1895 1900 05 10 1 20 30 35 4O 45 50 55 FIGURE 179.—-Annual precipitation at Curtis, Nebr., 1895—1960. fall over much of the central part of the continent. At Curtis, only 2 of 6 years since 1895 that had in excess of 30 inches of precipitation, were followed by years that had more than the long—term average amount. An extreme occurred in 1915—16 when the difference between precipitation amounts exceeded 20 inches. Although averages and extremes give some indi- cation of precipitation, a more meaningful descrip- tion of rainfall can be obtained through the use of rainfall intensity-duration-frequency curves. Such curves (fig. 181) were developed for Curtis for return periods of 2 and 5 years based on the annual series of 1-, 6-, and 24-hour amounts. The California method of plotting positions, T, : n/m, was used; a log normal distribution was assumed. T, is the return period in years of item having order number In in a decreasing series, and n is the period of record in years. An empirical. factor of 1.13 (US. Weather Bureau, 1957) was used to convert the clock-hour amounts of rainfall to maximum 60—minute values. An in- spection of the 6- and 24—hour amounts indicated that conversion for these periods was unnecessary. The plot of rainfall intensity versus return period, which was used to define the curves of figure 181, showed little scatter from the average curve, which indicated that no extreme storms had occurred dur- ing the period. The US. Weather Bureau (1955b) gives rainfall intensity-duration-frequency curves for about 200 stations, including North Platte, for the years 1906—51. Comparison of the curves for Curtis with those for North Platte shows that differences are minor between the two stations for both the 2- and 5-year return periods, even though only 8 years of record was used to define the curves for Curtis. The similarity between the short— and long-term curves and the fact that no extreme amounts of precipitation fell during the short period indicate that the rainfall regimen was well represented at Curtis during the years 1951—58, even though the —+ LOESS—MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 261 4O 38 P' 36 — 34— 30— 28— 24 - 66-year average -——- ”"4 i/ 22— PRECIPITATION, IN INCHES [I L 20 18— 16— 14— 12'- 5-year moving average — (plotted at middle of 5-year average) Tea—w-— ‘/‘ H“ml-“‘7 _ _1___._ i 1895 1900 05 10 1 20 25 ii Iii ‘ 3O 35 45 50 55 1960 FIGURE 180.—Annual precipitation at McCook, Nebr., 1895—1960. average precipitation for the period was somewhat lower than the long-term average. Evaporation is also variable from year to year, although the variations probably are not as extreme as those of rainfall. The only evaporation data available are for a few years at Medicine Creek Dam, and these data were used to check evaporation maps in a technical paper by the US. Weather Bureau (1959a). According to the maps the average annual evaporation from a class A pan in Medicine Creek basin is about 75 inches, or about 31/; times the average annual precipitation; and lake evapora- tion is about 50 inches, or nearly 21/; times the aver- age annual precipitation. Evaporation increases from north to south in the basin, but the average difference is only about 3 to 4 inches per year. Data on wind speed and direction nearest Medi- cine Creek basin are those at North Platte. Because North Platte is in the valley of the Platte River, the wind speeds and directions may not be entirely representative of conditions in the Medicine Creek basin. The average wind speeds at North Platte range from about 12 to 13 miles per hour March through July and from a little less than 10 to slightly more than 11 miles per hour for the re- mainder of the year. December and January have the lowest average speeds with 9.8 and 9.7 miles per hour, respectively, and April has the highest average with 13.1 miles per hour. The prevailing direction is from the southeast April through Au- gust; south-southeast in September and January; northwest in February, October, and November; north in March; and west-northwest in December. The annual prevailing direction at North Platte is considered to be from the southeast (US. Weather Bureau, 1959b). 262 10 I I I IIIII I II Illl I RAINFALL INTENSITY, IN INCHES PER HOUR o o’ lllll 1 l J 10 50 RAINFALL DURATION, IN HOURS 0.01 1 1 l 1 FIGURE 181.—Rainfall intensity-duration-frequency curves for Curtis, Nebr., 1951—58. VEGETATION AND PALEOCLIMATE CLIMAX PLANT COMMUNITIES The Medicine Creek basin lies in a transitional zone between the tall-grass region to the east (prairie grassland) and the short-grass region to the west (plains grassland). Trees are restricted to valley bottoms, except where they have been artificially planted. Some information regarding the history of plant communities appears in the notes of surveyors for the General Land Office who made brief reference to the grasses they saw while laying out the land- division grid in 1869—72.1 Their comments also indicate the conditionof the vegetal cover before white settlement. The following excerpts are from 1 Field notes of survey are on file at the Dept. of Educational Lands and Funds, State Capitol Bldg., Lincoln, Nebr. +— EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT the notes of Deputy Surveyor D. V. Stephenson, made in the Medicine Creek basin during the summer of 1872: “land broken by steep ravines running in almost every direction, in bottoms of some of which there is some blue grass. The upland is about three- fourths covered with short buffalo grass. * * * grass almost entirely short buffalo, growing in bunches.” Deputy Surveyor J. L. Slocum made the following notes, also in the summer of 1872: “Soil is dry and sandy, produces only short buffalo grass in scatter- ing bunches—unfit for cultivation or grazing. * * * Considerable good grass in bottom of ravines, upland is sparsely covered with short buffalo grass.” Dep- uty Surveyor Charles Wimpf wrote as follows, in August of 1869: “Bottoms of ravines filled with a thick very good grass which gives an excellent hay. The hills are covered with a short sweet grass.” Although the term “buffalograss” may have been used loosely by the early surveyors, their comments do indicate that during 1869—72 the uplands and valley—side slopes were covered with a rather sparse growth of short grass, and the valley bottoms with a thicker growth of tall grass. E. J. Dyksterhuis( writen commun. 1958), range conservationist for the Soil Conservation Service, considers that the climax community on the uplands in regions of climate and soils like those of the Medi- cine Creek basin contains a dominance of midgrasses with an understory of short grasses. He also says that the dominant midgrasses include western wheat- grass, needlegrasses, and little bluestem; little blue- stem, side—oats grama, and blue grama dominate on the valley sides, and big bluestem and switchgrass on the valley bottoms. CHARACTERISTICS OF NATIVE GRASSES Characteristics of native grasses are summarized in this section from Weaver and Albertson (1943, 1944), US. Department of Agriculture (1948), and Phillips Petroleum Co. (1956). Big bluestem is a coarse perennial native bunch- grass that reaches a height of 6 feet under favorable conditions. The root system is extensive; it may penetrate to a depth of about 8 feet and form a dense mat to a depth of 1 foot. Under favorable conditions it, together with other species of its community, will form a continuous sod; under less favorable conditions it forms separate, scattered bunches. Both rainfall and dust are largely inter- cepted by the foliage before reaching the ground. According to Clark (1937), a prairie covered by big bluestem may intercept as much as 53 tons of water per acre during a rainfall of 1 inch per hour, whereas a prairie covered by buffalograss will inter- LOESS—MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA cept about 28 tons per acre. Such intercepted rain- fall commonly evaporates from the leaves of the plant without reaching the ground. Buffalograss is native, sod forming, fine leaved, and perennial and typically grows to a height of 4 to 6 inches. The roots of well-developed buffalo- grass extend to a depth of 4 to 5 feet and form a very thick mat to a depth of 1 foot. Blue grama is a low native perennial grass whose flowering stems reach a height of about 18 inches. It is the most drought resistant of all the native prairie grasses, and it spreads, even during drought, mainly by means of seedlings. The root system is similar to that of buffalograss. All these native grasses may form a dense thick, continuous sod, although the bluestems and blue grama are more likely to form separate bunches than buffalograss. The taller grasses intercept a greater amount of rainfall, which is usually lost to the soil; but a corresponding decrease in erosion by raindrop impact is effected. Windblown dust is retained more effectively by the taller grasses. EFFECTS OF DROUGHT Some insight into the influence of ancient arid climatic regimes on native grasses and on the ability of these grasses to prevent erosion may be gained by study of the effects of the drought of 1933—40 in Kansas and Nebraska. Such study was made by Weaver and Albertson (1943, 1944). These authors took for a base station Hays, Kans., which has an average annual precipitation of about 22.9 inches and which is located about 120 miles south of the Medicine Creek basin. Total precipitation at Hays for the 6 years preceding 1933 was below normal. Precipitation during each of the 4 driest years was about 16 inches. According to Weaver and Alberston, decrease in grass cover as a result of drought varied with the intensity of grazing and of burial by dust and also with the type of grass cover, as grasses with shorter roots underwent greater destruction. The basal cover of short grasses in Kansas ranged between 80 and 95 percent in 1932, and this cover was com- monly reduced to 20 percent or less during the worst years of the drought. In some localities, particularly those that had been overgrazed or covered by blown dust, the native plant population was reduced almost to zero. Where the effects of drought were moderate, the sod was interrupted by patches of bare ground, a square foot or less in area, that formed an irregular but continuous network. Where the effects of drought were more severe, the patches of bare ground were larger, up to several square yards in 263 area; and in the most severely affected localities, the grass was so decimated that nearly all the ground was bare. These bare patches, described by Weaver and Albertson and noted by the writer (Brice, 1958) throughout the Medicine Creek basin in 1953—56, are important in slope erosion because erosion scarps commonly form at their edges. The influence of ancient droughts and ancient episodes 0f aridity on the grasses of the Medicine Creek basin can be inferred from Weaver and Albertson’s study. Short-term droughts would have reduced the cover of climax grasses, but erosion would not have been severe because these grasses are replaced by weeds during drought and recover rapidly after drought. Sheet-wash and raindrop impact would have removed soil from the irregular patches of bare ground between sod-covered areas, and low sod scarps would have formed on the slopes. During ancient episodes of aridity that lasted for hundreds or thousands of years, a climax community of grasses adapted to prevailing climate and rainfall distribution would have become established. Such a community would probably have consisted of an association of short grasses similar to the association now in eastern Colorado, where the average annual precipitation is about 15 inches. If this short-grass association were significantly different from that established before the arid episode, stream profiles and slopes would have been regarded by gullying and other erosional processes. EFFECTS OF LAND USE The pronounced effect of land use is strikingly shown by the contrast in vegetation on either side of a fence that follows the north-south section line between sections 15 and 16, T. 9 N., R. 27 W. (See-fig. 182, upper.) The heads of two tributaries to'East Fork of Dry Creek are isolated by this fence, and the aspect of the vegetation in these tributary heads is unlike that observed at any other place in the Medicine Creek basin. Although no information on the land-use history of the tributary heads was obtained, the contrast in vegetal cover at the fence is evident on both the 1937 and the 1952 aerial photographs. Therefore, the fence had divided areas of contrasting land use for at least 15 years, and the aspect of the vegetation in the tributary heads in 1953 suggested that it may have been disturbed in no important way since white settlement. Besides the much greater thickness of the sod and the domi- nance of tall grasses in the tributary heads, the thickness of brush such as Wild plum, sumac, buck- brush, and wild rose is exceptional (fig. 182, middle). A few ash trees were observed on the 264 FIGURE 182.—Photographs showing effect of land use on native vegetation, East Fort Dry Creek, August 1953. Upper, Areas separated by a fence and contrasting sharply in vegetal cover. Denser vegetation is at right of fence. Although sparse, the vegetation at left of fence was better than average for the Medicine Creek basin in 1953. Middle, Dense cover of grass and brush in ungrazed valley head. View looking west. Vegetation more dense on south side of valley. Lower, Sparse cover of short grass, yucca, and weeds in heavily grazed valley head, 1.5 miles southeast of valley head shown in middle. Valley is oriented in north-south direction. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT valley bottoms. The termination of the small valley- bottom gullies (fig. 182, upper) before reaching the fence may be fortuitous, but no gullies were observed in the tributary heads. By contrast, gullies are in most tributary heads in this part of Dry Creek (fig. 182, lower). Step scarps occur on the steep slopes of the heavily vegetated area, but they are much subdued and seem to be inactive. The upland drained by the tributary heads is planted in row crops or small grain, as elsewhere in the basin. NATIVE TREES AND DENDROCHRONOLOGY During the period 1869—72 General Land Ofl‘ice surveyors reported that trees in the Medicine Creek basin were confined to stream valleys, in which they identified cottonwood, ash, boxelder, elm, hackberry, and cedar. These same kinds of trees now grow from place to place in the valleys, but the Wider valley flats have been mostly cleared of trees. A tree-ring indication of precipitation during the past 400 years in western Nebraska, including the Vicinity of North Platte, was reported by Weakly (1940, 1943). According to his interpretation, the climate has been characterized by frequent dry years and by less frequent droughts that lasted 5 years or more. The 13 droughts (period of five or more successive dry years) had an average duration of 13 years, and the average interval between droughts was about 20 years. The longest drought lasted from 1539 to 1564 (26 yr), but other lengthy droughts were from 1587 to 1605 (19 yr) and from 1688 to 1707 (20 yr). PRE-PLEISTOCENE ROCKS UNEXPOSED ROCKS Neither the lithology nor the structure of unex— posed rocks in the Medicine Creek basin has any apparent direct influence on the present relief. The Paleozoic and Mesozoic formations that overlie the Precambrian basement dip gently to the west and are overlain unconformably by Tertiary formations that dip gently to the east. EXPOSED ROCKS NIOBRARA FORMATION The Niobrara Formation of Late Cretaceous age crops out along the banks of Harry Strunk Lake and from place to place along Medicine Creek down- stream from the lake, but it is not exposed elsewhere in the basin. Where exposed, the Niobrara consists mainly of orange or white chalk interbedded with thin layers of altered volcanic ash. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA O GALLALA FORMATION Outcrops of the Ogallala Formation of Tertiary (Pliocene) age are mainly restricted to valley sides and channels of larger streams in the Medicine Creek basin, and the best exposures are along Medi- cine Creek and Cedar Creek in the vicinity of Stock- ville. No outcrops were observed in the northern parts of the basin where the loess cover is thickest. Available well logs indicate that the Ogallala under— lies most of the basin and that its thickness increases from about 200 feet in the southern part to about 400 feet in the northern part. (See figs. 183 and 184.) Exposed sections of Ogallala at seven localities were measured and described in detail. Thicknesses of described sections ranged from 6 feet at one 10- cality on Cedar Creek to 73 feet at another locality downstream, and the total thickness of described sections was 290 feet. In general, the Ogallala con- sists of clay, silt, volcanic ash, sand, and gravel, poorly sorted into different beds that are cemented with (and partly replaced by) different amounts of carbonate. Five rock types were distinguished, and the total thickness of each type as represented in the seven measured sections is given in table 1. TABLE 1.—Thickn~ess and percentage of each rock type in seven exposed sections of the Ogallala Formation Thick- Percent ness of total (ft) thickness Gravel and sand 1 22 7.5 Sand, pebbly, and silt; loosely cemented, locally concretionary 156 54.0 Limestone, sandy, rather uniformly cemented ,, _ 62 21.4 Ash, volcanic, and 5111:, ,,,,,,,,,,,, 26 9.0 Clay; alternating beds of fine-grained limestone ........... 24 8.1 With regard to the origin of the Ogallala, Frye and others (1956) have presented convincing evi- dence that it was deposited by streams flowing east- ward from the Rocky Mountain region. In the earlier part of their history these streams occupied broad, relatively shallow valleys eroded into Cre- taceous bedrock. However, as alluviation proceeded, the deposits progressively overlapped the gentle valley sides, most divides were buried, and eventually a coalescent alluvial plain was formed. PLEISTOC‘ENE AND RECENT DEPOSITS TERRACES Three terraces were identified along Medicine Creek and its tributaries. The highest of these is named the Wellfleet terrace after the village of Wellfleet, Nebr., which is on Medicine Creek. A1- 265 though Wellfieet is not built on the terrace surface, conspicuous fiat-topped remnants of the terrace stand about 125 feet above the valley flat at intervals of a mile or more between Wellfieet and Maywood, Nebr. For example, the remnant in the SE14 sec. 8, T. 8 N,, R. 29 W., about 1 mile north of Maywood, is well defined both in the field and on the Curtis NW quadrangle sheet. There are similar fiat-topped remnants along Well Canyon and other major tribu- taries, but the originally flat surface of the terrace has been dissected in most places. The best preserved and most continuous terrace in the Medicine Creek basin is named the Stockville terrace after the town of Stockville, Nebr., which is built on a remnant of the terrace, along the west side of Medicine Creek. The Stockville terrace is represented along Medicine Creek by broad rem- nants that stand above the valley flat at a height of about 50 feet in the lower course of the creek and about 30 feet in its upper course. The Stockville terrace can be traced almost continuously along most tributary channels in the basin. Moreover, the sides and heads of valleys are broadly graded to the level of the Stockville terrace. The lowest terrace of general importance in the Medicine Creek basin is named the Mousel terrace after the Mousel School, which is in sec. 25, T. 5 N., R. 26 W., on the east side of the valley of Medicine Creek. (See fig. 176.) The Mousel terrace is repre— sented along Medicine Creek by widely scattered remnants that stand 15 to 20 feet above the valley fiat. Similar remnants are along most major tribu- taries, such as Well Canyon, Cedar Creek, and Fox Creek ;.but the Mousel terrace cannot be traced along most minor tributaries because its surface has been buried by late Recent alluvium. Even where the terrace is well defined, there has been little grading of the valley sides to the level of the terrace; the slope that forms the riser of the Stockville terrace rises abruptly above the tread of the Mousel terrace. A sequence of well-defined terraces appears along most streams in Nebraska, but general agreement has not been reached as to terrace nomenclature. Condra and others (1950) describe a sequence of six terraces in central Nebraska; and Schultz and others (1951) describe a sequence of five terraces that applies to Nebraska generally. Neither sequence would be expected to apply to all Nebraska streams because of complications, such as overlap of an older terrace deposit by a younger terrace deposit, that may be peculiar to a single stream. Schultz and others (1948, 1951) have specifically applied their terrace sequence to alluvial terraces EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 266 :30 $326 Ucw .iwwcoo 59696th .I_ A. Bt< .53: mo Eu .533 23 mwmmgo Rafi 35 553.53: a 953 non—own omwofiowUIfiwH .5505.” zOCUmm no ZOFED SHOES) & companion 93.522 3 4 V _ gofiwfihom 23m 9 J I 0 a Hm I a H. W newsflash Emzwmo IA §§ IE>M4 (mm Z_ m_ 23.20 snoaamsug Mddfl BF: A AHVNHELVDO 9149906270 k—r_t mm 9mm auaoond ;fi__1 ZO_I_.~ J . ‘3 §sm§e~ EEoSSNN \0 322m HEN wswm 3:: $.35 N353»: wu6§3~ 1a , $6334 .3 fiamfisN \c 3%. W; W :oEngom wnfigoq w w m d w “Ex: SQIS “3°33: was $5ng M so Swmfiowmfis 33 \e 3.5 m. mwmoq 538m m , u \ g u gum 95D mOHX ZOF W—a 1743993 bOHN boHN boNN I ISMN I boem I IOOmN I boom I benw I bowm I \oomm I J/Ix, ”I \\\o x \ 130,74; umyqndag 319mg aupoypaw //////// $1 . \Awm . . \s/fiml //J._§ &V\\\\\§\ a0 20 M 0 mm H.N N_L V I SM“ .ZVF 0— .Zm.._. O O 0 R V\ \ // ”‘VO‘W f A SN // // . / //, I x \ \\ ZNF _ .ZwF \/ 49 um / 7 \ 7 / I \\ \\‘ \\ Id \ § a O 00___H;IILN0_IM 00 NOSAAVG DP “r N :3 I .oomN I \ooem . 503 .MW .mfivfigmg$u~§nwm.wounmnflv.0..-755.13.. .‘\\\\\\\\\\\ ...%r’/.////Z/ - boom .0 W§§ / m muowe am§ 18% m w I ‘oowm A H _ H _ IEoz _vu .oomm _ 2:9 _ .zw: sum—0m _ _ O S . 0 L O_ H N H 267 LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA .325 05 $33 2:. £53.53: a use? nomauwm ammo—cowléma 852% was SsuEsSmZ J so .iamfiaN .fisgfifi \O fieZfiM was wcam 2.2m 8.85 E S n O 3 D — V13 H 2) snoaammg .mddn 3:: £33 fimaags was :3 -235»? so :smfiak Kc “Em. sosafiaom gauged VJ I ZOFUmm “.0 ZO_._.m:_ (mm Zcmo:ocm M 8% 1 W N \ /\‘ I 8% a 3 ‘ Z .\ / H a “m k w M 508 1 H 3 :3 x .1 m 388 m m u q m m w :58 m _ o o . m m. o “ :Eoz 88. N N D n . I88 m m u u _ o m m. \ _ Mm _ _ _ m _ m m: m _ _ m I" a o _ _ _ o _ m _ w _ V . D . O ‘ . . . O . zqk 0_ zmk _ zwk _ .259 _ .zmp _ .zmp _ zo: _ 2:9 _ .zw: 268 in the southern part of the Medicine Creek basin, in particular along Lime Creek. According to their nomenclature, alluvial terraces in Nebraska are num- bered upward from the youngest and lowest, which is called Terrace—0, to the highest and oldest, which is called Terrace-5. Although their numerical sys- tem of terrace nomenclature is not used, the terrace sequence in this report is based on, and in general agreement with, the work of Schultz and others. Correlation of terrace deposits with Pleistocene time units is uncertain, particularly in view of the present controversy about the units. A tentative correlation of the Medicine Creek terrace sequence with other terrace sequences and with the Pleistocene time units is given in table 2. The Stockville terrace of this report corresponds physiographically to Terrace-2 of Schultz and others (1948, 1951), and the Stockville terrace deposits correspond to fill A of Terrace-2. No terrace de- posits corresponding to fill B of Terrace-2 were observed in the Medicine Creek basin by the writer. The Wellfieet terrace, which is underlain by the upper part of the valley fill of the Peorian Loess, corresponds physiographically with the Terrace-3 of Schultz and others and with the Terrace no. 3 of Condra and others (1950, p. 35). No physio- graphic expression of the Terrace-4 of Schultz and others nor of the Terrace no. 4 of Condra and others was recognized in the Medicine Creek basin. At a locality on Medicine Creek (sec. 11, T. 5 N., R. 26 W.), Schultz and others (1948, p. 36) have identi- fied a surface as the tread of Terrace—3. This surface is interpreted by the writer as a long valley-side slope graded to the level of the Stockville terrace. At this same locality the surface identified as Ter- race-4 by Schultz and others is interpreted as the tread of Terrace-3, or the Wellfleet terrace. A ter- race equivalent to Terrace-4 was doubtless once EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT present in the Medicine Creek basin, but the physio- graphic expression has been obliterated by deposi- tion of the upper part of the Peorian Loess and by grading of the valley sides during the time that the Stockville terrace deposits were accumulating. GENERAL STRATIGRAPHY OF THE PLEISTOCENE AND RECENT Stratigraphy of the Pleistocene deposits in Ne- braska is well established, and satisfactory correla- tions have been made with the deposits of Kansas and other adjoining States, although nomenclature differs somewhat from State to State. For regional descriptions of the stratigraphic units the reader is referred to publications of the Nebraska Geological Survey (Condra and others, 1950; Lugn, 1935) and of the Kansas Geological Survey (Frye and Leonard, 1952). Leonard (1950, 1952) has established the ranges of land snails that are generally abundant in the Pleistocene deposits of Kansas. The general relations of Pleistocene and Recent stratigraphic units in the Medicine Creek basin are shown in figure 185, and in table 2 correlation is tentatively made with formally named units in Ne- braska and Wyoming. Long profiles of terrace de- posits along the course of Medicine Creek are shown in figure 186. Correlation of the upper Wisconsin and Recent units is based on carbon-14 age deter- minations and on physiographic expression as ter- races. Correlation of the older units is based on fossil evidence, on soil markers such as the Sangamon soil, and on lithology. Pleistocene and Recent de- posits in the Medicine Creek basin consist of dune sand, loess, and alluvium, overlying a pre—Pleistocene (or lower Pleistocene) surface cut into the Ogallala Formation. The deposits of Nebraskan or Kansan age are mainly of alluvial silt and sand; gravel, in lesser amounts, is confined to the bottoms of former Upland level _l L|J — a 0 -r Wellfleet terrace _| 2 < 25 ——~\\\\ , ///,, : Peorlan Lgess Peorian Loess 3 \ \ Stockville terrace Stockville terrace I / ’ \ r r E 50 — \\\ /// ““50“ T O \ Mousel // Saniam . "Ea / F Imam)“ d Laval a” \ \ terrace / Loveland o 50" m 75 —l and 0 "Tan \ \ Stockville / Varmolm‘ I _. ,_ 0” Valley flat ,,.--—' Lu terrace . ' Lu “- ' deposits WFormation _ Z 100 - Ogallala Formation V :5 Recent Mousel terrace Ogallala Formatlon l— alluvmm deposits 3-, 125 — _ D VERTICAL EXAGGERATION X8 400 0 400 FEET FIGURE 185.—General topographic and stratigraphic relations of Pleistocene and Recent deposits in the Medicine Creek basin. Differences in strati- graphy and relief on either side of the figure represent the range of differences among valleys in the basin. 269 LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA :32? ac: fl 8556 Swim: :o .m—acwsvwmnsm €35 393M can 3va an cwfiaou Sam. :0 find mhaww \m cofimoawv :3 334.53 .woatma vwmwonnxw mug—«oo— mno€a> Scum E35. :0 find muse? \fl EEQEQEMEQEQ a 9:359; 3:5 $5 memfiocm hue—onscoouunwv was E339 89G moawn— \m €32 £5 5 53M awe—3 :0 :md flaw? \m commmofi 3:: we mania .5me $3ch z<>2< 23: E23 min sfiifiw .3 03.892 macaw :0 rand Emu? A E emu 0 Eu onwpgom Em :SEMV—“zm :au—maunwz Em GEE—Em 5958—935 mow :a_a3«< Eu @920 «52 En ESE @520 Egg; Eu «gnaw $33 :ozafiuoh gnaw Eu “ESE 5:20 203:2 :3 3333A :umaum Ew annam Saab Em gnaw $8 552:5? =8 552:5“? :3 «gnaw 5:055“? gigs; I I §2§$§§§§§§ Eu 320 Em 39.0 :cfiaEhoh cum—954 «$3th Eu “Eu—9:5 EM cud—9’3 Samoan: =8 cofimmnuw :8 :oEuquw mom 6:333 :cancam Egg: Em SEEM 3.3m uaam ho=n> 258 M :8; grow» 2 E 33’ :5 Mummy :3 5.58 $33 533 W «$05.89 3 mom flanmfinoflng mom m |§§§2§§§§s§é§§ Egg; 9 En =¢3sz8 33m Egg Em nuiowm 2252 I =chst .Em =o>$nfla 83 mag :aioem m Em A :8 Annam— fl 22:33ch mow «-84..wa Em #30 2.3m E: n .c: Em 518m nan: 5.30 En 005:3 I/ “am #30 3.3 3.3.59 $50er 32.0sz I . can hodum 4 Eu #85...wa 9 :0» 332m— leuatmu. Mr? mom , :83 Egg»); Egg; ISNH «0 www :a umkusm mounv a33§>< .ou—Sm Em 35132 buum N .o: Em $255 -asm 363w £25va3 33 , , $3wa a 5m? moafiwtcu 335925 89:3 3682“. 36:3 w=w>xo3m Eu $950 5325);; Suififi ININIIT M Gomwmwonon— vat—93w coats SE 33132 35A , 893M , unwanE>0v mow ~25me F «5.5 as §§§§2§$§E§E§ g§§§2§§§§s§§ _ a A 2 SN. w a 2535;; 783wa w =em \§.Nn N m3 file: .530..— | Eu $93M H . @878: d4 55 83.8 NASNHSMNW 35:3 gig; I I m .8959» sultan .SQQD 930 w. $~+mw H €98 ham—Ewan no $83.10: 8 u N .558an . 2 3332. 35:3 $302 .502 .fllgflmonmcbwobuooz Em 26.330 3.5m .4 Egg; I Egg; >>>§>§>>§>>>>> Egg; I . . Babe» | En 23590 33 $5 Euaofimooma 83 D < :awom $33.78! wash—3 E5 Baa m Eu 255.35 938.58 N mow m aoéwaan 3:: 283” 533:“ 28$ 9:5: 5 E» L? £3833 Egg; I I gig: I I >§§§>§§2§>>I Egg; u w m Amman zany 83 €25: $5 32 .552 a; 2886 :3 £38 c5 5353 A89 £28 25 8288 w? 3.3m v3. D 3 nxwfinuz £020 2:3qu 9:8ng «23.562 «33.502 35:00 Ragga E333 39:30:00 §§§c§~$ ~33: afiséuz :0 ”5.3% 36 agnosamfim 2% xo 308:3 3.6 3E: cmdasaxska “8 :omuguiob .N HAM—<8 270 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT ‘ I I I I F I I I \ I I I I I I I I I I I I I I I I I I I I I I I I 1 \ _ \ K 2800 ‘\ \ \ K . _ \ \ \I - \ A \ . _ ”#41, 0., I\. _ ‘ 1 I \ \ \ 2700 \ \‘ ‘\\ l " I _, E \\‘~ I \I I Lu _ “I \ \ ' 7 5 o \\ \ I I\ A _ s \\ ‘ I s 2 \ \ I ‘ ' I :5 1 E flfAfl‘s ‘II\II A w s J ‘x I I I “x z 2600 \‘ x‘ I V I \ s _ | \ \ \\' I I I ‘-\ I E ‘3 \\ I .\I- W _ L; _ g ":’2: ‘ \‘I\ \9/07 - o 1 5; ”4‘ AW, QK \ I \eé‘t ._ E E . \ ~ \ \x “5’ 2500 I I , ff \ \‘ I ' --, LII—J é I I If—(fl ” \:\\41‘\ I S \.\’ E - o m ' s 0‘, 2‘ 8,. e gzg§ I 06. \§e%\Q,./l \. I i * 3 ~ 355 as I five»: , x g - ’ _ = x E x "/" \ \rao \er, ‘~\I a m I» o o g \\ \9 \‘90 a’ ‘ Q 3 “- m L \ \ x \e I\. n: :> 2400 '5 o I \ \ x I .c_ t ‘5 >. I ‘\\ \ \\ \ a) t 5 1 o 5 2 ,- ‘\\ l/\ \ I M 3 1 < = Harry 27‘ el/Q\ \ 7‘5 8 , 3: 8:7 \J’f/\ I\ n C — 8 0% \‘i’ \I- \ E “2’ — a ’7 \ \ \ m _ U) 4’(\ ~\ \ O a. g 6/, ~ \ C 2300 e\‘ . ~\ I 8.— I\ \\ \I m I T 1 ,5 x _ 2 3% , g 6 a _ 2200 l I I I I I l I l I l l I l I I I I I l I l I I I2 I 50 40 3O DISTANCE FROM CONFLUENCE, IN MILES FIGURE 186.—Channel profile, valley flat profile, and terrace profiles along the main course of Medicine Creek. valleys (figs. 183 and 184). Although loess may haVe been deposited during either Kansan or Ne- bra'Skan time, no loess of Kansan or Nebraskan age has been identified with certainty either on Medicine Creek or elsewhere in Nebraska. At two localities in the Medicine Creek basin, a paleosol was identi- fied as the Yarmouth soil on the basis of snail fauna and high content of volcanic ash. Silts beneath the Yarmouth soil are correlated with the Sappa For- mation. No deposits of ash of the Pearlette Ash Member of the Sappa Formation were seen in the Medicine Creek basin, but deposits up to 30 feet in thickness crop out in areas adjacent to the basin. No Pleistocene deposits older than the Sappa were recognized with certainty in the Medicine Creek basin. The thickness and areal distribution of pre-Wis— consin stratigraphic units cannot be determined without data from drilling. Moreover, the logs must include full and accurate descriptions of the rocks that are penetrated, because .the different Pleistocene stratigraphic units are similar in rock type to one another and to the Ogallala Formation. Two geo- logic sections (figs. 183 and 184) have been compiled from logs of test holes that were drilled by the Nebraska Conservation and Survey Division in co- operation with the U.S. Geological Survey. Figure 183 is simplified from an unpublished section by J. L. Deffenbaugh, based on his interpretation of logs of test holes drilled in 1948. On his original section, Deffenbaugh distinguishes the Loveland, Sappa, and Grand Island Formations. Figure 184 represents an interpretation, made by the writer, of logs of test holes drilled in 1960. Although the logs include full descriptions of the rocks penetrated, the writer was not able to distinguish with confi- dence any formational boundaries beneath the San- gamon soil and above the Ogallala Formation. To facilitate comparison of figures 183 and 184, some of the stratigraphic units distinguished by Defi‘en- baugh were combined in order that they correspond with the units shown on figure 184. STRATIGRAPHY ON CEDAR CREEK NEAR STOCKVILLE Along Cedar Creek, about 3 miles southwest of the village of Stockville (Bartley NW quad., NEMl, sec. 18, T. 6 N., R. 27 W.), several sections have been exposed by lateral cutting of the creek, and the terrace sequence is more complete and distinct than LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA NORTH 2580' 1 2570’ 2560' Mousel terrace 2550’ 2540' Stockville terrace deposits 1 /q- Valley flat A. 2530' / / terrace . depositsq./ / ". Cedar Creek 2520’ Recent /'\' 2510' — alluvium I“ Stockville terrace (projected) / . Flow line of Cedar Creek ’ , .— —-' Peorian Loess \ \ ’- Loveland Formation Yarmouth soil \mmmmwmfi s / 7 1" Ogallala Formation 2500’ O VERTICAL EXAGGERAT‘ON X3 10|O I DATUM IS MEAN SEA LEVEL 100 200 300 FEET FIGURE 187.—Profi1e and geologic section on Cedar Creek. usual for the Medicine Creek basin. The section described below applies to the right-hand (south) side of figure 187 : Depth . (ft) Peor1an Loess: Silt, pale-yellow (2.5Y 7/3 dry, 2.5Y 6/3 moist), calcareous. Surface soil not present ,,,,,,,,,,,,,,,, Loveland Formation: Silt, very pale brown (10YR 7/4 dry, 10YR 5/4 moist) ; a few streamers of fine sand near base Gravel, sand, and silt, crossbedded, calcareous, locally cemented with carbonate ,,,,,,,,,,,,,,,,,,,,,,,, Silt, pale-brown (10YR 6/3 dry, 10YR 4/2 moist), noncalcareous; indistinct sedimentary lamination Yarmouth soil: Silt, clayey, gray-brown (10YR 5/2 dry, 10YR 4/2 moist). Weak columnar structure, col- umns irregular, 2 to 5 in. wide; noncalcareous. Upper contact gradational ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Sappa Formation: Sand, fine, and ashy light-gray (2.5Y 7/2 moist) silt; a few thin white irregular seams of v01- canic ash. Upper 0.5 ft mottled with darker gray, gradational with unit above. Upper 2 ft is C... horizon of Yarmouth soil; contains meshwork of fine carbonate veinlets ,,,,,,,,,,,,,,,,,, Sand, fine, and ashy silt; a few streamers of sand and fine gravel near base; noncalcareous Gravel and other debris from Ogallala Forma- tion; crossbedded; a few streamers of fine sand Ogallala Formation: 0—15 15—20 20—22 22—26 26—28.9 28.9—32 32—35 35—37 37—53 A snail fauna was collected from the lower part of the soil identified as Yarmouth and from the upper part of the unit identified as Sappa, and identifi- cation was made by Dr. W. J. Wayne of the Indiana Geological Survey. In the following table, the num- ber of individuals of each species is given in the right-hand column. Gastrocopta tappam'ana (C. B. Adams) __________ 2 proarmifera Leonard ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 3 Gyraulus circumst'riatus (Tryon) _- I 21 Helicodiscus parallelus (Say _____________________________ 3 Physa anetina Lea ______________________________________________ 2 Retinella electrina (Gould) ________________________________ 3 Lymnaea [Fossaria] parva _________________________________ 2 Vallonia gracilicosta Reinhardt ________________________ 15 Valvuta tricarinata (Say) ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 4 According to Leonard (1950), G. proa'rm'ifem is restricted to the Yarmouth, and none of the species in the collection is restricted to deposits younger than the Yarmouth. Correlation of the soil with the Yarmouth and of the underlying silt with the Sappa is, therefore, supported by the faunal evidence. V. tricam'nata is aquatic, and the fauna as a whole indicates a moist environment such as would be afforded by a valley. The stratigraphic and topographic relations shown in figure 187 indicate that the valley of Cedar Creek was cut not later than Kansan time and that at least five episodes of valley alluviation, each fol- lowed by cutting, have taken place since Yarmouth time. STRATIGRAPHY ON CUT CANYON NEAR CURTIS Near the confluence of Cut Canyon with Fox Creek, a significant section of Pleistocene deposits is exposed along the county road that descends eastward from the upland and crosses Cut Canyon in the NW14 sec. 29, T. 9 N., R. 28 W. The presence of deposits of the Sappa indicates that the canyon was cut not later than Kansan time, and the Pleisto- cene sequence of pre-Wisconsin age seems to be complete. Both the Yarmouth and the Sangamon soils are well developed. Of the terrace deposits of late Wisconsin and Recent age, however, only the Stockville is exposed (fig. 188). A stratigraphic section along the road, beginning 1,500 feet west of Cut Canyon bridge and ending 1,150 feet west of the bridge, is as follows: Depth (ft) 0—30 Peorian Loess: Silt, yellowish-gray, massive, homogeneous ........ 272 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT WEST EAST 2300" Upland surface F 2780’ —\ “ 2760’ ~ _ 2740' - — I? Cut Canyon 0 2720' — h e/ Sangamon soil 2700 “ Stockville Peorian Loess — 2680’ _ terrace deposits Tr Stockville ' Lov';":”\ terrace deposits 2660’ — ~_ "d F°"nation 1 1 _ Yarmouth SOll \/\ \ 2640’ — Sappa Formation «.mw.‘ ’ ' 2620’ Faunal location h Ogallala Formation 2600’ VERTICAL EXAGGERATlON x3 200 0 l 1 200 400 FEET I l DATUM IS MEAN SEA LEVEL Fromm 188.——Profile and geologic section along road entering Cut Canyon. ”8”? t Sangamon soil and Loveland Formation: Silt, yellowish-brown (10YR 5/4 when moist), friable; no distinct structure ____________________________ Silt, light—yellowish-brown (10YR 6/4 when moist) ; weak prismatic structure, noncalcare- ous Silt, clayey, approximately same color as unit above, but has pinkish cast ____________________________ Silt, streaked and veined with abundant soft carbonate 30—32 32—38 38—38.5 38.5—43 Yarmouth soil: Silt, light-brownish-gray (5Y 7/2 when dry, 10YR.6/2 when moist), clayey; streaked with carbonate from unit above, but otherwise non- calcareous. Breaks into irregular small poly- hedrons (5 to 10 mm in diameter) that fit to- gether. Contact with unit above sharp but irregular Silt, yellowish-brown (10YR 5/4 when moist); contains abundant small iron concretions ________ 43—465 46.5-47 A stratigraphic section 800 feet west of the bridge is similar to that described above, except the Sappa Formation is exposed and the Yarmouth soil is somewhat different, as shown below: Depth (ft) Yarmouth soil: Silt, olive—gray (5YR 4/2 when moist) ______________ 0—1 Silt, light—gray (10YR 7/2 when moist) ; crumb structure 1—2.5 Silt, light-gray, mottled with rust color; struc- ture weakly prismatic, prisms have irregular sides, 1 to 3 cm Wide __________________________________________ 2.5—4.1 Silt, light—gray; no distinct structure _______________ 4.1—7.5 Sappa Formation: Sand, fine-grained, and silt; cross-laminated 7.5~9.5 No fossils were found at this section, but fossils were found in gray ashy silt, correlated with the Sappa, about 600 feet to the east. The faunas were identified by the writer, and the identification was checked and corrected by Dr. W. J. Wayne of the Indiana Geological Survey. The number of each species found is indicated in column at right in the following list: Carychium exile canadense Clapp ____________________ 2 perem’guum Baker ________________________________________ 2 Gastrocopta pentodon (Say) 6 proarmife'ra Leonard ___________________________________ 2 Gyraulus circumstriatus (Tryon) ____________________ 19 Pupilla musco'rum (Linné) ____________ 1 muscomm sinistm Franzen ________________________ 1 Retinella electrina. (Gould) ________________________________ 1 Vallom’a gracilicosta, Reinhardt . 10 Valvata tricam'nata (Say) ____________________________ 24 Vertigo nylande'ri Sterki ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 3 According to Leonard (1950) , G. proarmifera and P. muscorum sinistm are restricted to the Yarmouth and C. perexiguum is restricted to deposits of Ne- braskan and Yarmouth age; none of the species found is restricted to deposits younger than Yar- mouth age. Therefore, correlation of the ashy silt with the Sappa Formation is reasonable. STRATIGRAPHY ON ELKHORN CANYON NEAR MAYWOOD The Peorian Loess, Stockville terrace deposits, and late Recent alluvium are exposed along Elkhorn Canyon about 2 miles southwest of the Village of Maywood, near a bridge Where the county road crosses the canyon (Curtis SW quad., center of section line between secs. 31 and 32, T. 8 N., R. 29 W.). The exposure was studied after it had been thoroughly moistened by rainfall and runoff, because moistening greatly increases visibility of the sedi- mentary structures and textures. The relations be— tween the different units are shown in figure 189, and a general View of the locality is shown in figure 190. Evidence was found in the late Recent alluvium for two episodes of cutting and filling, which took LOESS—MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA SOUTH 2730’ "‘ Stockville terrace 2720’ — 2710' Stockville terrace Peonan \ deposns Loess l / 2700’ 2690' — \g 273 NORTH Numbers refer to parts of Recent deposits Layer from which \ \ ~ ‘ ~ _ _ ‘\ E13355: . C” sample W-754 \ \ \\ 2680 _ was collected \ 1 - 3 . \ "‘ Flow line of Elkhorn Canyon \ Recast deDOSItS Ogallala Formation 2670’ VERTICAL EXAGGERATION XZVZ 50 O 50 100 FEET I I I J DATUM lS MEAN SEA LEVEL FIGURE 189.—-Profile and geologic section on Elkhorn Canyon. place (according to a carbon-14 age determination) within the past 300 to 400 years. The late Recent alluvium consists of four parts, which are designated by the numbers 1 through 4 on figure 189. Part 1 is oldest, lightest in color, and most compacted, whereas the younger parts are successively darker in color and less compacted. The contact between part 1 and part 2, which lies at a depth of about 12 feet below the ground surface, is conformable and nearly horizontal. Beneath the contact, the deposits of part 1 show no evidence of erosion or of leaching. Part 2 is characterized by light brownish gray bands of fine sand and silt (2.5Y 6/2 moist) alternating with darker grayish brown bands of clayey silt (2.5Y 4/2 moist), whose darker colors result from a finer texture and a FIGURE 190.—View of terrace sequence and valley fills along Elkhorn Canyon. Prominent scarp marks front of Stockville terrace; lower scarp marks con- tact between parts 3 and 4 of late Recent alluvium. higher content of organic matter. Part 1 is also banded, but the banding is inconspicuous because of the relatively lighter color of the clay-rich layers. Part 2 is riddled with worm burrows, which are partly filled with castings and dark silt, and many of which terminate in a round cavity partly filled with castings. The darker clay-rich bands seem to have been preferred by the worms, for these bands contain castings in greatest abundance. Worm cast- ings are much less apparent in part 1, although dark vertical streaks, which probably represent the filled burrows of worms, were seen from place to place, and the bedding is apparently riddled by worm burrows. A lens of charcoal and wood ashes, mixed with silt and containing a single large fragment of charred bone, was found interbedded in part 2 at a depth of about 7.5 feet below the ground surface and 3 to 4 feet above the contact with part 1. Charcoal (Geol. Survey sample W—754; analysis by Meyer Rubin) from the lens yielded a carbon-14 date of 420:160 years B.P. (before present). Parts 1 and 2 of the late Recent alluvium are de- scribed here because the contact between these parts is considered to mark a distinct change in the rate of accumulation of the late Recent alluvium; the rate of accumulation of part 2 is considered to be much the more rapid. Furthermore, the more rapid accu- mulation of part 2 probably reflects an increase in rate of erosion on the uplands and valley sides. Part 1 is lighter in color because its slower rate of accumulation permitted a greater degree of oxida- tion of finely divided organic matter, which was originally intermixed with the silt ‘ and clay. A slower rate of accumulation of part 1 is also indi- 274 cated by the greater destruction of its sedimentary structures by earth worms. If the described changes from part 1 to part 2 do in fact reflect a change in rate of accumulation, this change took place about 500 years ago, as indicated by the position of the radiometrically dated charcoal. According to arche- ologic evidence, a period of drought may have begun about 500 years ago, when the region was apparently evacuated by people of the Upper Republican Culture. Parts 3 and 4 of the late Recent alluvium are dark gray brown (10YR 5/2 moist) and show rather dis- tinct banding. Carbonized fragments of grass stems, having an average length of about 0.5 mm, are scat- tered throughout, but they are more abundant in the darker, clay-rich bands. Dark laminae formed of ashes and charcoal, which probably represent the debris from grass fires, were seen from place to place. Much of the alluvium seems to have passed through the digestive tracts of earth worms, Whose castings dominate the texture and account for the loose consolidation. The steeply dipping erosional contact that sepa— rates parts 1 and 2 from part 3 is attributed to the passage of a major gully head scarp along the valley of Elkhorn Canyon. Part 3 of the late Recent allu- vium, which was deposited in the gully left by this head scarp, was in turn trenched by a second head scarp following consecutively behind the first. The former sides of the trench made by the second head scarp are represented by a pair of low terraces that were traced (on aerial photographs) to their ter- minus about 3 miles upvalley. Deposits formed behind the second head scarp (part 4) were in turn trenched by a third head scarp following consecu— tively behind the second. In 1952 the third head scarp was 13,300 feet upvalley from this locality, and it advanced about 500 feet during the period 1937—52. Relatively little time is required for the formation of a series of low terraces by the passage of consecu- tive gullies, because deposition is rapid on the floor of a gully and the head scarps travel rapidly as a result of concentration of flow on the gully bottom. The average rate of advance of the third head scarp during the period 1937—52 is probably conservative, partly because the head scarp has reached bedrock. An average advance of about 200 feet a year, at- tained by some head scarps in the Medicine Creek basin during the period 1937—52, would be adequate to account for passage of the events described during a period of about 350 years. On the other hand, these events would not likely be compressed into a period of much less than 350 years. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT LOVELAND FORMATION AND SANGAMON SOIL The most useful and distinctive soil marker within the Pleistocene deposits is the Sangamon soil, which is developed on the Loveland Formation. No other soil in the region, including the modern soil, ap- proaches the Sangamon in depth of profile or in degree of development of the CC, horizon. On the uplands the Loveland consists of loess, which mantles the valley sides and merges into contemporaneous valley-fill deposits of silt and sand. The maximum observed thickness of Loveland in the Medicine Creek basin is 35 feet (along Cedar Creek), and the aver- age thickness is about 20 feet. At some exposures in valleys, the Loveland is missing and the Sangamon soil is developed on the Ogallala Formation. Moist loess of the Loveland is typically light yel- lowish brown (10YR 6/4), but darker colors (10YR 5/ 3 and 10YR 5/4) were observed at some localities. The Sangamon soil, as observed from a distance, is a distinct dark-brown humus—rich band that sepa- rates loess of the Loveland from the distinctly lighter colored loess of the Peorian. (See fig. 191.) This band ranges from about 4.5 to about 2 feet in thick- ness and from yellowish brown (10YR 5/4) where the soil is weakly developed to very dark gray brown (10YR 4/3) where it is strongly developed. A Cm. horizon is nearly everywhere present in the soil, but its depth below the dark band ranges from about 2 feet in some localities to about 25 feet in others, and its thickness is also variable. The varia- tions indicate that development of the Sangamon in some localities was nearly continuous, whereas in other localities the developing soil was intermit- tently buried by fresh accumulations of silt. In spite of the conspicuous development of the humus—rich part of the Sangamon soil profile, it contains an abundance of fresh ferromagnesian min- FIGURE 191.—Valley fill of the Peorian Loess set unconformably against Sangamon soil and Peorian Loess. Steeply dipping unconformity, indi- cated by arrows, is exposed on either side of the small valley, and the valley fill of the Peorian is between the arrows. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA erals and hence has undergone little chemical weath- ering. About 20 species of heavy minerals were identified under the petrographic microscope; of these, clinozoisite, green hornblende, apatite, and brown hernblende are by far the most abundant. In addition, calcic feldspar (labradorite) was identified among the light minerals. PEORIAN LOESS, BRADY SOIL, 'BIGNELL LOESS, AND MODERN SOIL An account of the involved history of the terms “Peoria” and Peorian has been given by Leonard (1951, p. 323). In this report, the term Peorian Loess is applied to deposits that are stratigraphically above the Sangamon soil and terminate in the Brady soil. The absolute time at which the Sangamon Interglaciation ended has not been determined, but a reasonable estimate can be made by extrapolating from existing carbon-14 dates; the estimate of Frye and Willman (1960, p. 2) is between 50 and 70 thousand years B.P., as measured by carbon-14. The Brady soil has been dated, by means of a carbon- 14 age determination on organic carbon from the A horizon of the soil, at 9,160i250 years B.P. (Rubin and Suess, 1956, p. 443). This date is prob- ably too young because of root hairs in the sample. In general, the Peorian Loess mantles the whole of the Medicine Creek basin except the modern valley flats and the Stockville and Mouse] terraces. On the uplands the Peorian consists of massive silt, which grades laterally (at least at the surface) to fine sand in the vicinity of dune areas. This massive silt is called loess because its topographic position indicates that it is eolian and because the snails that it contains are not aquatic. In valley locations also, the greater part of the thickness of the Peorian consists of massive silt, except in the valleys in the vicinity of dune areas, where it consists of silt and fine sand. The lowermost 10 to 20 feet of the Peorian in the valleys commonly consists of silt interbedded with fine gravel, silty clay, or sand; and this lower- most part contains, in some localities, aquatic snails or pelecypods. The massive silt of the valleys is not called loess because its agent of deposition is not apparent. Where a vertical face cut in this silt is exposed to Wind and rainwash, it commonly shows well—defined lamination that could be attributed either to Wind or to water. Eolian deposition of silt on the uplands was probably accompanied by an equal or greater amount of eolian deposition in the valleys, but the silt in the valleys was probably reworked by water before final deposition. Deposits of the Peorian in valley locations, Whether of massive 275 silt or of some other sediment type, are called the valley—fill deposits of the Peorian. Leonard (1951) recognizes three biostratigraphic zones in the Peorian Loess of Kansas, but these zones are not marked by any physical discontinuities in the loess. The basal zone, which is devoid of molluscan fossils, is overlain by a lower molluscan faunal zone and an upper molluscan faunal zone. Between the upper and lower faunal zones is a transitional faunal zone, which bears elements of both the lower and upper faunal assemblages. A snail fauna was collected about 25 feet strati- graphically above the Sangamon soil at a locality on Dry Creek to ascertain the approximate position of these faunal zones relative to the total thickness of the loess of the Peorian in the Medicine Creek basin. The locality is in the NW% sec. 16, T. 9 N., R. 27 W. The stratigraphic position of the fauna is about midway of the total thickness of the loeSS on Dry Creek. A sample of loess weighing about 25 pounds was collected, and the contained snails were obtained by washing the loess through a sieve. Identification of the snails was made by Dr. W. J. Wayne of the Indiana Geological Survey, and the faunal list is as follows (the number of individuals of each species is given in right-hand column) : Columella alticola (Ingersoll) ,,,,,,,,,,,,,,,,,,,,,,,,,,,, 2 Discus cronkhitei (Newcomb) ,,,,,,,,,,,,,,,,,,,,,,,,,,, 2 shimeki (Pilsbry) ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 24 Euconulus ful’vus (Miiller) ,, ,,,,,,,,,,,,,,,,,,,,,,,,,,,, 2 Pupilla muscorum (Linné) ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 4 Succmca grosvenom’ Lea ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 16 Vallonia gracilicosta Reinhardt ,,,,,,,,,,,,,,,,,,,,,,,, 2 Vertigo modesta (Say) ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 3 tridentata Wolf ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 10 Of the nine species listed, five belong to the upper and transitional zones of Leonard, and four occur in all three zones. Because of the absence of S. (ward and the numerical dominance of D. shimeki and S. grosvenori, which are upper zone species, the assemblage is assigned by the writer to the upper zone. The fauna in this single locality indicates that at least half of the total thickness of the Peorian in the Medicine Creek basin was deposited during the time represented by the upper faunal zone of Leonard. Leonard correlated his upper zone With the Tazewell Stade of the Wisconsin Glaciation. During accumulation of the Peorian on the up- lands, a period of incision to bedrock took place in the valleys. Valley fill of the Peorian that accumu- lated after this incision is set unconformably against the Loveland Formation, the Sangamon soil, and the lower part of the Peorian. (See figs. 185 and 191, both of which Show valley fill of the Peorian set unconformably against the Sangamon soil and 276 the lower part of the Peorian.) From a distance the valley fill of the Peorian looks almost white, whereas the loess of the Peorian that overlies the Sangamon soil has a yellowish cast. On the uplands no physical discontinuity within the Peorian was discerned. Molluscan fauna collected from the base of the valley fill of the Peorian at two localities indicates that deposition of the valley fill began in Tazewell time. The first locality is on the bank of Dry Creek, about 700 feet upstream from the gaging station, in the SE14 sec. 24, T. 8 N., R. 28 W. Here the valley fill rests unconformably on the Ogallala For- mation. Identification of the fauna was made by the writer, with the assistance of Dr. W. J. Wayne. The faunal list is as follows: Columella alticola (Ingersoll) Discus cronkhitei (Newcomb) shimeki (Pilsbry) Gyraulus circumstriatus (Tryon) Physa anatina Lea Pupilla muscomm (Linné) Retinella electrina (Gould) Succinea grosvenom' Lea Vallon'ia gracilicosta Reinhardt Vertigo gouldi coloradensis Cockerell modesta (Say) Zonitoides arboreus (Say) According to Leonard (1952, p. 19), D. shimeki and D. cronkhitei are reliable indexes to the upper (Tazewell) faunal zone of the Peorian. None of the other species in the assemblage is restricted to de- posits either older or younger than this upper faunal zone. A second fauna from the valley fill of the Peorian was collected at the side of Mitchell Creek, near a county road in the NW14 sec. 9, T. 6 N., R. 26 W. Identification was made by the writer, with the as- sistance of Dr. W. J. Wayne. The faunal list is as follows: Discus cronkhitei (Newcomb) Euconulus ful'uus (Mfiller) Gastrocopta armifera (Say) Lymnaea (Stagm'cola) palustris (Miiller) Retinella electrina (Gould) Succi’nea grosvenom' Lea Vallom’a gracilicosta Reinhardt Vertigo tridentata Wolf Although this assemblage is different from the as- semblage collected from the valley fill of the Peorian on Dry Creek, a correlation with the upper mollus- can faunal zone of Leonard (1951, 1952) is indicated by the presence of D. cronkhitei and S. grosvenom'. The average thickness of the loess of the Peorian in the Medicine Creek basin is about 50 feet, and the maximum thickness is about 80 feet. Thicknesses were measured at expoSures in upland localities and EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT from the logs of test wells drilled by the Nebraska Conservation and Survey Division. (See figs. 183 and 184.) The thickness of the loess reaches a maxi- mum in the Vicinity of the Platte River, gradually decreases southward to a minimum in a belt which is about 30 miles south of the Platte River and 10 miles north of the Republican River, then again in- creases in thickness in the direction of the Republi- can. A well near the north side ofthe Republican River valley penetrated a thickness of 60 feet of loess, whereas a well near the south side of the valley penetrated a thickness of only 30 feet. About 50 feet of loess of the Peorian is exposed in the vertical wall of a deep pit, 1.5 miles west of Eustice, Nebr. (fig. 192). The valley fill of the Peorian is thicker than the loess. The exposed thickness of valley fill of the Peorian on the main stem of Medicine Creek, as measured from the surface of the Wellfieet terrace to the modern valley flat, is about 125 feet. On the assumption that the valley fill extends to bedrock, as it does on Dry Creek, the total thickness on Medicine Creek is about 200 feet. On the same assumption, the total thickness on Well Canyon, at a locality about 7 miles north of Curtis, is 180 feet. The exposed thickness in the upper reaches of Dry Creek is 42 feet, and the total thickness is about 65 feet. Loess of the Peorian in the Medicine Creek basin is a massive friable light—colored silt, similar to loess that has been described in other regions. The loess when dry is typically light gray (10YR 7/2) and when moist is light brownish gray or pale FIGURE 192.—Peorian Loess (50 ft thick) overlying a horizon of Sangamon soil (upper dark band, marked 8). Units beneath Peorian Loess are Loveland Formation (20 ft), Yarmouth soil (lowermost dark band; marked Y), Sappa Formation (33 ft), and Pearlette Ash Member of Sappa Formation (20 ft). Excavation, made for removal of Pearlette, is 1.5 miles west of Eustice, Nebr. Photographed by C. H. Hembree, 1956. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA brown (10YR 6/2 or 6/3). Although the loess ap- pears to be homogeneous when viewed from a dis- tance, closer inspection reveals abundant hollow tubules made by rootlets, worm burrows, and small botryoidal structures which are worm castings. Within 5 feet of the ground surface, the loess may show such an abundance of these botryoidal struc- tures as to indicate that all of it has passed through the digestive tracts of earth worms. Snail shells are generally present and are locally abundant. Lamination, although not usually observable, may appear where erosion by wind and raindrops has been sufficiently delicate to etch the laminae into relief. Thin sections of seven samples of loess of the Peorian, collected at different localities in the Medi- cine Creek basin, were studied under the petro- graphic microscope. The most distinctive textural features of loess, when viewed at high magnification, are the rather close packing arrangement and the presence of birefringent clay coatings around detrital silicate grains. Grains larger than 125 microns in maximum diameter constitute less than 1 percent by volume of the samples, and grains larger than 31 microns constitute 55 to 75 percent by volume. Particles larger than 31 microns may be regarded as the framework or skeleton of the loess, and par- ticles smaller than 31 microns may be regarded as the matrix. The clay grain coatings, which are evi- dently authigenic, may be regarded as cement. The framework grains are mainly separated by their clay coatings and by the matrix. Contacts between framework grains are mainly tangential or long, and an average of about two contacts per grain was observed. The birefringent clay coatings around detrital silicate grains consist of crystalline flakes oriented parallel with the surface of the coated grain. These coatings merge with the matrix and probably account in considerable part for the coherence of the loess and for the ability of loess to maintain a vertical face. Clay coatings on flat grains, such as mica flakes and glass shards, are in general thicker than coatings on equant grains. Clay coatings on grains are not characteristic of siltstones deposited under water. In loess the coatings are probably formed by wetting and drying of the surface layer during deposition. Kubiena (1938, p. 134) has attributed the formation of grain coatings in soils to such wetting and drying. After rainfall, the soil solution may fill all pore spaces in the soil, but evaporation at the ground surface causes the soil solution to retreat to the angles of the intergranular spaces and to the surfaces of grains. As the soil dries, sub- 277 stances that were peptized or dissolved in the soil solution are left as coatings on grains. Although the grain coatings described by Kubiena were of humus, coatings of clay also probably form by wetting and drying. In this region loess of the Peorian is generally calcareous, but the calcium carbonate is not uni- formly disseminated. Calcium carbonate in the loess occurs in the form of silt—sized grains, which are doubtless primary; as powdery streaks and vein fillings; as fine—grained aggregates that line or fill rootlet tubules, worm burrows, or other openings; and as concretions. Many loess outcrops show zones up to several feet thick that are noncalcareous. Frankel (1957) has reported vertical variations in the distribution and concentration of secondary calcareous concretions and in the abundance and state of preservation of fossil mollusks in the loess in Nebraska. He attributes these variations to an intermittent rate of loess deposition. Slow deposi- tion of loess is accompanied by solution of snail shells and by accumulation of secondary carbonate at depth. Rapid accumulation of loess is indicated by unaltered snail shells. Frankel suggests that many phantom soils are present. The writer has observed that massive silt, accumulated as valley fill and indistinguishable from upland wind-laid loess, is more commonly noncalcareous than upland loess. The Brady soil was named and described by Schultz and Stout (1948) from a locality on the steep south side of the Platte Valley near Bignell, Nebr. This locality is also the type section for the overlying Bignell Loess (Schultz and Stout, 1945). According to the writer’s observations, the humic zone of the type Brady soil is about 1.3 feet thick, and the soil is weakly calcareous throughout; soft streaks and films of carbonate appear to a depth of 2.2 feet from the top of the soil, and these are con- spicuous from 1.3 to 2.2 feet. Overlying the Brady soil is a thickness of 8 to 10 feet of Bignell Loess, on which the modern soil is developed. This modern soil is noncalcareous to a depth of 3.2 feet and shows a distinctly columnar structure to a depth of 1.3 feet. Above the modern soil is about 1 foot of lighter silt, evidently colluvial and related to cultivation of fields upslope from this locality. The modern soil appears to be more strongly developed than the Brady soil. In spite of the fact that Medicine Creek basin is adjacent to the type locality of both the Brady soil and the overlying Bignell Loess, the Brady soil, as separately developed from the modern soil, was ob- served at only three exposures, all in the northern part of the basin. 278 General deposition of a thin layer of Bignell Loess over the uplands of the Medicine Creek basin is indicated by abnormally thick A horizons that com- monly appear in the modern soil. The modern soils of Lincoln County, Nebr., have been mapped and described by Goke and others (1926) ; and the soils of Frontier County, Nebr., by Bacon and others (1939). The Medicine Creek basin lies within these two counties. Principal soils of the loess-mantled areas are the Holdredge very fine sandy loam and the broken phase of the Colby very fine sandy loam. The Holdredge soil, which is developed mainly on areas of flat or rolling upland, is mature. The B horizon of the Holdredge lies between 10 and 24 inches in depth and shows an imperfectly developed columnar structure; and the C0a horizon, which contains a concentration of carbonate in the form of streaks, splotches, or threads, lies between 3 and 4 feet in depth. A dark—brown zone, probably representing the Brady soil, lies between the B horizon and the C0,, horizon at many localities. STO CKVILLE TERRACE DEPO SITS Typical relations of the Stockville terrace deposits to other stratigraphic units are shown in figure 185. Along major valleys, the unconformity between the Stockville terrace deposits and older units is rarely marked by a scarp; if one were originally present, it has been obliterated by erosion on the valley-side slopes. Along minor valleys, the unconformity is usually marked by a distinct scarp. The Stockville terrace deposits consist mainly of calcareous silt. In minor valleys the silt is less compact, the color is darker, and the bedding is less well defined than in major valleys. Dark humus—rich bands, beneath which the silt is leached to a depth of a foot or two, are observable at most exposures of the Stockville terrace deposits. The depth of burial of the uppermost band ranges, in different parts of the basin, from about 2 to 20 feet; and the number of bands varies from one exposure to the next. These hands, together with the underlying leached zone, probably represent immature soils that formed during periods of slow deposition. The Stock— ville terrace deposits occur in nearly all valleys of the present drainage system, including valleys of. first order. The thickness, which depends on the size of the valley, is about 100 feet along Medicine Creek and Well Canyon, about 60 feet along Lime Creek, and about 30 feet along Dry Creek. A fauna of snails and pelecypods was collected‘ from the Stockville terrace deposits exposed along Medicine Creek in the NE cor. sec. 13, T. 6 N., R. 27 W. According to Dr. W. J. Wayne, who identified EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT the fauna, none of the species is extinct, and all are fresh-water species that now inhabit parts of central United States. Leonard (1952, p. 16) concludes, from his study of Bignellian molluscan faunas in Kansas, that the environment of Bignell Loess depo- sition was very much like present conditions in the Great Plains. The Bignell Loess is considered by the writer to be the upland counterpart of the Stockville terrace deposits. MOUSEL TERRACE DEPOSITS The Mousel terrace deposits consist mainly of silt and differ from the Stockville terrace deposits only by being somewhat darker in color and more com- pact. In some localities the Mousel terrace deposits contain fragments of carbonized grass stems, which are rare in the Stockville terrace deposits. Along the main course of Medicine Creek, two buried humus-rich bands, which are interpreted to be immature soils, were observed in deposits of the Mousel terrace. Charcoal suitable for carbon-14 age determination was collected from the Mousel terrace deposits on Dry Creek in the NWMLNEl/A, sec. 1, T. 7 N., R. 28 W. The charcoal was in a lens of wood ashes and silt, buried at a depth of 14.5 feet below the ground surface. The age was reported as 2,200i200 years (J. L. Kulp, Lamont No. 239C). RECENT ALLUVIUM Where valley trenching has exposed the full thick- ness of the late Recent alluvium—as at localities on Dry Creek, Cedar Creek, Elkhorn Canyon, and Lime Creek—the alluvium rests on the bedrock floor of the valley. Therefore, at these localities the valleys were incised to bedrock during the episode of incision that followed accumulation of the Mousel terrace deposits. Incision to bedrock may have been confined to localities where the bedrock is high beneath the valley floor. In some valleys, such as Dry Creek and Elkhorn Canyon, accumulation of the late Recent alluvium has been interrupted by one or more episodes of incision during which the valley was trenched for part of its length by the upstream migration of intermittently spaced channel scarps. In other valleys, such as Well Canyon and Fox Creek, accumulation of the late Recent alluvium has evi- ' dently proceeded without the interruption of valley incision. . In the upper reaches of Dry Creek, where the previously untrenched valley flat is being trenched by headward migration of a major channel scarp,‘ the silty late Recent alluvium consists of a banded LOESS—MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA FIGURE 193.-—-Banding in late Recent alluvium, Dry Creek. (SW14NE14 Sec. 29, T. 9 N., R. 27 W.) upper part, about 14 feet thick, and a lower part that is indistinctly banded and lighter in color. (See fig. 193.) The dark bands of the upper part are somewhat more clayey than the light bands, and the dark color is a result of a finer texture and a higher content of organic matter. Both light and dark bands are composed mostly of loosely consoli- dated silt, through which abundant fragments of carbonized grass stems are scattered. Earth-worm castings are a conspicuous feature of the upper banded part, but they are less apparent in the more compact silt of the lower part. A similar division of the late Recent alluvium of Elkhorn Canyon into an upper part that is banded and a lower part that is indistinctly banded has already been described. The banded fill is attributed to a rate of accumulation so rapid that time has not been available for oxidation of finely divided organic matter in the clayey bands. If the accumulation of banded fill began at the same time on Dry Creek as on Elkhorn Canyon, the banded fill shown in figure 193 has accumulated during the past 500 years, approximately. This corresponds to an average rate of accumulation of about 0.34 inch per year, al— though the appearance of the banding indicates that the rate of accumulation was not uniform. The banded alluvium seems to be characteristic of valleys that are being actively trenched, whereas the more compact unbanded alluvium is character- istic of more stable valleys. For example, the allu- vium beneath the valley flats of Well Canyon and Fox Creek, which are not being actively trenched, is unbanded. On the other hand, distinctly banded 279 alluvium is exposed in the walls of rapidly advancing trenches on Dry Creek and Curtis Creek Canyon. The present rate of accumulation of alluvium on valley flats must be known in order to relate sedi- ment discharge by streams from the drainage basin (sediment yield) to the amount of sediment that is eroded from uplands and valley-side slopes (gross erosion). In spite of a thorough search, exposures of the late Recent alluvium yielded no artifacts nor other materials that would give an accurate indication of the rate of accumulation. Valley flats in many localities were examined after runoff events, and the depth of accumulation of freshly deposited sediment was measured. The thickness of the freshly deposited silt layer ranges widely (from a fraction of an inch to about 6 in.) from one locality to an— other, and almost as great a range was observed from place to place at a single locality. Moreover, freshly deposited alluvium may either remain at its place of deposition or be removed during the next runoff event. Farmers were asked about the time required for fenceposts on the valley flat to become wholly or partly buried. In the upper part of Dry Creek, two farmers estimated independently that 5 or 6 feet of sediment had accumulated on the valley flat between 1920 and 1953. On Well Canyon one farmer estimated an accumulation of 2 feet on the valley flat in the past 40 years (1916—56) and an- other estimated an accumulation of 2 feet in the past 30 years (1926—56). On Medicine Creek, about 4.5 miles north of Cambridge, a farmer reported that the wire of a hogpen, last used before the 1935 flood on Medicine Creek, was buried to a depth of 6 feet by 1957. An average rate of accumulation of about 1 inch per year on the valley flats of the drainage basin is probably of the right order of magnitude. A range between 1%; inch and 3 inches is estimated, and this range would be found not only from valley to valley but also from place to place within the same valley. The rate of accumulation of alluvium within a trenched channel, downstream from an actively ad- vancing channel scarp, must be considered separately from the rate of accumulation on an untrenched valley flat. Examination of freshly deposited allu- vium on the floor of trenched channels, several hun- dred feet downstream from an actively advancing channel scarp, indicates that 0.5 foot or more of alluvium can be deposited during a single runofl" event. At several localities on Dry Creek, tin cans and wire were found buried to a depth of several feet in alluvium that had accumulated on the floor of the trench. Evidently, alluvium accumulates rapidly 280 in trenched channels but is susceptible to removal by a second episode of trenching. SIGNIFICANCE OF CARBONIZED FRAGMENTS 0F GRASS In this region the younger a fluvial deposit is, the darker is its color. The dark color is due in part to finely divided organic matter and in part to fragments of carbonized grass stems, which are more abundant in the younger deposits. The car- bonized fragments are attributed to grass fires, in- asmuch as uncarbonized plant fragments would decay in this environment. The scarcity of carbon- ized stems in the older deposits might mean that grass fires were uncommon during the deposition of these deposits, or it might mean that any frag- ments originally present have been destroyed by oxidation or by the activities of earth worms. In an effort to decide between these alternatives, five thin sections of the Stockville terrace deposits, one thin section of the Mousel terrace deposits, and two thin sections of the late Recent alluvium were examined. In mineralogy, degree of development of clay coat— ings around grains, and packing of particles, all these thin sections were similar to thin sections of loess. However, the upper Recent alluvium shows many open spaces, which are attributed mainly to the activity of earth worms. Although carbonized fragments in the Stockville terrace deposits are small, they show no signs of partial oxidation. The tentative conclusion is that grass fires were more common during deposition of the Mousel terrace deposits and the upper Recent alluvium than during deposition of the Stockville terrace deposits. A greater incidence of grass fires might reflect a drier climate, or it might reflect a greater use of fire drives as a method of hunting game. The extent to which fire drives were used by prehistoric man on the Great Plains is largely speculative (Wedel, 1961, p. 76). SAND DUNES The sand dunes within the Medicine Creek basin are part of a much larger dune area that is an outlier of the Sandhills of Nebraska. The geomorphology of the Sandhills has been briefly described by Smith (1955). In general, the dune relief in the Medicine Creek basin is irregular and hummocky, dominated by innumerable blowouts, most of which are stabil- ized by grass. However, an indistinct linear pattern was observed in aerial photographs of areas about 7 miles north and 10 miles northwest of Wellfleet. The pattern is made by low, discontinuous longi- tudinal dunes in subparallel arrangement. The dunes EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT have a maximum length of about one—fourth of a mile, an average width of about 200 feet, and heights ranging from 10 to 30 feet; they show the furrowed crest that is typical of longitudinal dunes in the Sandhills. They are probably relicts of a more ex- tensive dune system that has in most places been destroyed by blowouts but has in some places been modified into U-shaped dunes. Although the longi- tudinal dunes are oriented about N. 65° W., the ends of most of the U-shaped dunes are oriented more northerly. Elongated blowouts tend to be oriented at about N. 20°—30°W. Evidently, the prevailing wind direction has shifted since the formation of the longitudinal dunes. Most of the blowouts are clearly man induced, for they are associated with areas that formerly were or currently are cultivated. The stream-dissected relief of the loess-mantled part of the basin is separated from the sand-dune relief by a transitional belt of ground that is inter- mediate both in relief and in particle size of under- lying materials. The relief of this transitional belt is characterized by broad undrained depressions, generally elongated northward and separated by nearly level ground or by broad, rounded elevations. The underlying material, which is intermediate in particle size between dune sand and loess, is re— stricted to upland locations. The mixture of dune sand and loess, on which the Anselmo fine sandy loam and the Colby fine sandy loam are developed, is generally less than 3 feet thick and is underlain either by dune sand or by loess (Goke and others, 1926; Bacon and others, 1939). Because of seasonal shifts in wind direction, silt from the loess—mantled areas has become mixed with sand from the dune areas. In general, the dune sand does not seem to have transgressed very much over the loess-mantled areas since the end of deposition of the Peorian—that is, within the past 12,000 years. In areas bordering the Sandhills, the surface of the Wellfleet terrace is generally free of windblown sand. ARCHEOLOGY AND HUMAN OCCUPATION RESULTS OF ARCHEOLOGIC INVESTIGATIONS Archeologic investigations have been made in the Medicine Creek basin by the Nebraska State His- torical Society, by the Nebraska State Museum, and by the River Basin Surveys of the Smithsonian Institution. Three cultural aspects have been dis— tinguished. These are represented by the Early Lithic sites, which are buried about 35 feet beneath the surface of the Stockville terrace (Terrace—2); by Woodland sites on the Mousel terrace (Terrace- LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 1) ; and by Upper Republican sites on the Stockville terrace. Personnel of the University of Nebraska State Museum have excavated two sites on Lime Creek, designated Ft—41 and Ft—42, and one site on Medi- cine Creek about a half a mile downstream from the mouth of Lime Creek (Ft—50). All these sites are in the lower part of the Stockville terrace (Terrace- 2) at occupation levels marked by abundant artifacts and many hearths. No human remains were found. Schultz and others (1951), who investigated both the geologic and archeologic aspects of the sites, report dates of 9,167i600 years B.P. and 9,880: 670 years B.P. from two charcoal samples taken from the bone—artifact zone of site Ft—41. A char- coal sample from the lower occupation zone of Ft—50 yielded a date of 10,493i 1,500 years B.P. Two Woodland sites in the Medicine Creek Dam area are described by Kivett (1949). Habitation areas (marked by shallow basins, post molds, shallow pits, flint tools, and scant pottery fragments) were on the Mousel terrace (Terrace-1) and were covered by about 18 inches of silt. The remains give Wedel (1949) the impression of a rather uncertain hold on the region by a culturally simple group. Accord- ing to Kivett, calcite-tempered ware found at the site is a marker for the Keith focus, which is one of several western Woodland cultural complexes. Although no radiocarbon dates were obtained at the Woodland sites on Medicine Creek, a radiocarbon date of 1,343i240 years B.P. has been obtained from a Keith-focus site on Prairie Dog Creek in northern Kansas, about 60 miles southeast of the Medicine Creek sites (Wedel and Kivett, 1956). The Upper Republican Aspect represents the latest wholly prehistoric culture on Medicine Creek and is the most abundantly represented by archeologic re- mains, which are on the Stockville terrace and are buried under a silt mantle 6 to 18 inches thick (Kivett, 1949). The date of the culture is estimated to be about 500 to 600 years B.P. (Kivett, oral commun., 1957). The geologic significance of these archeologic re- searches may now be considered. Radiocarbon dates on charcoal from occupation levels in the lower part of the Stockville terrace indicate that deposition of the Stockville terrace deposits began about 10,000 years ago. If the Woodland sites on the Mousel ter- race do indeed belong to the Keith focus and if this focus is correctly dated, then the terrace is older than 1,300 years. The silt mantle ranging in thick— ness from 6 to 18 inches covers Woodland sites on the Mousel terrace and Upper Republican sites on the Stockville terrace. Most of the sites are not 281 adjacent to upland areas from which colluvium might be readily derived, and no fluvial sedimentary structures were observed. Probably most of the silt was deposited on both cultural sites by the wind after the passing of the Upper Republican Aspect— that is, during the past 500 years. Conditions under which the silt was deposited were of more than local magnitude, as the silt mantle appears on Upper Republican sites throughout the Republican River basin in southern Nebraska from Frontier County eastward to Webster County (Wedel, 1941). Prob- ably the Upper Republican people inhabited the region during a wet period and emigrated about 500 years ago at the onset of a dry period. During the wet period a humic zone developed, which was buried by windblown silt accumulated during one or more later dry periods. SETTLEMENT AND LAND USE One of the first settlers of the basin came to the Stockville area in 1860. His dwelling was used for the formal organization of Frontier County in 1872, when the population of the county consisted of a few stockraisers and only two permanent settlers. Through the later 1870’s, settlers gradually entered Frontier County, and a little village arose at Stock- ville. By 1880 farmers had begun to claim land and to settle on the divides. Until the Free Range Law was repealed in 1885, farming on the uplands was of little consequence. By 1900 most of the desirable land had been taken under the Homestead, Timber Claim, and Preemption Acts. Bacon and others (1939) note that corn has been the main crop since farming began. The first settlers did not understand how to adjust their farming methods to the highly variable precipitation, and their practices were crude and wasteful. A series of dry years, accompanied by plagues of grasshop- pers, culminated in the disastrous droughts of 1893 and 1894. Many of the early farmers were forced to leave the region, and agricultural development was delayed. The farmers who remained acquired larger holdings and gradually adjusted their farming methods to local conditions. Although corn remained the most important crop, increasing amounts of wheat, oats, rye, barley, and sorghums were grown. Census figures have not been published for the basin as a unit, but satisfactory approximations of the population may be made from the precinct census. The total population was 7,750 in 1930, 6,400 in 1940, and 5,900 in 1950. The total popula— tion of the basin decreased by 23 percent between 1930 and 1950, and the proportion of persons living in the towns has increased from 42 percent in 1930 282 to 51 percent in 1950. Population density of the basin in 1950 was 8.7 persons per square mile. The number of settlers and domestic animals in the basin before 1875 was too small to have any significant effect on erosion and deposition. The sparseness of the grass cover on the uplands during 1869—72 reported by surveyors cannot be attributed to grazing by livestock. According to the Federal census, cattle and horses in Frontier County in- creased from about 30,000 in 1890 to about 43,000 in 1935. Farming could have had little geologic importance before 1879, when, according to the Federal census, only about 600 acres in Frontier County had been plowed. Only about 104,000 acres of plowed land was reported in Frontier County in 1889, whereas about 425,000 acres was reported in 1929. DRAINAGE SYSTEM EVOLUTION OF THE DRAINAGE SYSTEM According to the stratigraphic evidence presented in this report, major valleys of the Medicine Creek basin, such as Cut Canyon and Cedar Creek, were in existence during the Kansan Glaciation, and their bedrock floors were at about the same depth as at present. In the Republican River valley also, the bedrock floor has not been deepened since Kansan time according to a geologic section of the river valley near McCook, Nebr. (Bradley and Johnson, 1957 pl. 38). Drilling along two traverses between the Republican and the Platte Valleys (figs. 183 and 184) revealed buried valleys, cut into the Ogallala Formation and filled with pre-Wisconsin deposits. These valleys, which evidently had an eastward trend, were probably filled with alluvium and aban- doned in Nebraskan or early Kansan time. Plum Creek, which flows eastward into the Platte River, is probably a relic of the early Pleistocene drainage. Deposits of Pearlette Ash Member of the Sappa For- mation of Kansan age are distributed along the valley of Plum Creek and its former westward ex- tension, which was captured by Deer Creek after deposition of the Peorian. The valleys of most major tributaries (fifth and higher order) were established by incision and prob- ably by extension of the drainage system during the episode of incision that followed deposition of the Sappa Formation. These incised valleys were subsequently filled with the Loveland Formation to levels that are generally 10 or 20 feet above the modern valley flats. Most valleys of lower order than fifth were probably not in existence in Sanga— mon time, for the Sangamon soil does not dip toward them. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT There is no evidence of extension of the drainage system between the end of Sangamon time and the initial deposition of loess of the Peorian, although the valleys may have been incised. After about half of the total thickness of the Peorian had accumulated on the uplands, the valleys were incised to bedrock. The incision was followed by an episode of deposition during which the upper part of the Peorian accumu- lated in the valleys and on the uplands. The landscape as it existed at the end of deposition of the Peorian can be reconstructed (fig. 194) from remnants such as those shown in figure 195 (upper). In a few places, as for example at the eastern tip of Dry Creek, the valley heads of Peorian time have been preserved. Deposition of the Peorian Loess and development of the Brady soil are tentatively assigned to the interval 60,000 to 12,000 years B.P. The drainage system was developed to approxi- mately its present pattern and extent during the episode of incision that followed deposition of the Peorian (fig. 1948). The incision is perhaps asso— ciated with a rapid retreat of late Wisconsin ice sheets and a rather abrupt climatic warming about 11,000 years B.P., the evidence of which is presented by Broecker and others (1960). Valleys were prob— ably incised to about the same bedrock altitudes as during previous erosional episodes, but drainage was gradually increased. The extent of incision and increase in drainage density is strikingly shown in an area just outside the Medicine Creek basin, where the head of North Plum Creek was captured by Deer Creek. Although the Deer Creek drainage was incised about 100 feet below the surface of the Peorian, the North Plum Creek drainage just east of the point of capture was not incised, and the surface of the Peorian has been preserved almost intact. (See fig. 196.) The lengthy duration of the episode during which the Stockville terrace deposits accumulated is indi- cated by extensive grading of valley-side slopes and valley heads to the level of the Stockville terrace. In profile, the valley-side slopes that join the Stock- ville terrace to the upland are straight for most of their length, gently convex upward at their inter- section with the upland and gently concave upward at their intersection with the terrace. Most of the slopes are graded across the Peorian, but some are graded across older rock units, including the Ogal— lala. Accumulation of the Stockville terrace deposits and grading of slopes are assigned tentatively to the interval between the climatic warming of about 11,000 years RP. and the middle of the Altithermal, about 5,000 B.P. The Stockville terrace and slopes graded to the terrace can be identified nearly every- LOESS—MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 283 ’ peorian Loe'ivS . - Loveland Formation 4 O allala Formation Stockville terrace deposns g D FIGURE 194.——Evolution of the relief since deposition of the Peorian Loess. A, Valleys were broad and valley sides gently sloping at end of deposition of the Peorian, about 12,000 years B.P. B, Drainage system was incised, extended, and branched during episode that lasted from about 12,000 to 11,000 years B.P. C, Valley sides and valley heads were graded during accumulation of Stockville terrace deposits, from about 11,000 to 5,000 years B.P. D. Relief was modified during the several episodes of incision and deposition that occurred during the past 5,000 years. 284 ‘1 ‘ I 3 ‘ -i- it Y ’. .51. FIGURE 195.—Remnants of the side slopes and valley flats of former drainage systems. Upper, Looking south from the Platte-Republican drainage divide (in the NEV; sec. 31, T. 11 N., R. 28 W.) along a fourth-order tributary to Fox Creek. Remnants of the valley sides of Peorian time are preserved on the divides between side tributaries, and these remnants show the general aspect of the former valley. Middle, Looking north along Well Canyon in the SW14 sec. 27, T. 10 N., R. 29 W. Gentle slope on hori- zon (left) is slope of valley side of Peorian time. steeper converging slopes on horizon (left center) are slopes of valley side of Stockville time. Remnant of the valley flat of Stockville time is in middle distance, and below this is a remnant of the valley flat of Mousel time. Narrow trench on modern valley flat showed little or no change from 1937 to 1952. Lower, Looking southeast from the upper end of a fifth-order tributary to Cut Canyon, in the NEJ/I; sec. 26, T. 10 N., R. 29 W. Crest of interstream divides between side tributaries in background indicate approximate level of valley flat of Peorian time. Valley head in foreground was graded in Stockville time, as were the steep side slopes leading to it. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT where in the basin and constitute a reference surface to which the incision of later erosional episodes can be related. The Stockville terrace was formed at some time before 2,200 years B.P., which is the carbon-14 date obtained from Mousel terrace deposits on Dry Creek. Tentatively, the Stockville terracing is placed at about 5,000 years B.P., and the cause is assigned to drought during the Altithermal, which was a well-established warm-dry climatic episode that lasted from about 6,000 to 4,000 B.P. In minor valleys (of fifth and lower order) the effects of this terracing cannot generally be distinguished from the effects of the Mousel terracing. During one or the other of these two episodes of erosion, trenching extended to the heads of many minor valleys. Other minor valleys were trenched for only part of their length, and still others escaped being trenched. Deposition of the Mousel terrace deposits began at some time before 2,200 years B.P. and ended at some time before 420 years B.P. (carbon-14 date from late Recent alluvium on Elkhorn Canyon). Tentatively, this deposition is assigned to the inter- val 4,000 to 1,000 years B.P., which corresponds roughly with the dates of the “Deposition 2” de- scribed by Miller (1958, p. 38) in his generalized alluvial sequence in Western United States. (See table 2.) The terracing of the Mousel terrace must have been brief, inasmuch as the thickness and prop- erties of the late Recent alluvium indicate that not much less than a thousand years would be required for its accumulation. A tentative chronology of post-Sanga‘mon deposi- tional and erosional episodes in the Medicine Creek basin is summarized below: Years B.P. Local incision of valley bottoms _______________________ Present to 500. Accumulation of late Recent alluvium .. , Present to 900. Terracing of Mousel terrace ............................ 900 to 1,000. Accumulation of deposits of Mousel terrace 1,000 to 4,000. Terracing of Stockville terrace ________________________ 4,000 to 5,000. Accumulation of deposits of Stockville ter- race. 5,000 to 11,000. Incision and extension of the drainage sys- tem to approximately its present pattern. 11,000 to 12,000. Accumulation of Peorian Loess and develop- ment of Brady soil. MORPHOMETRY Detailed measurements were made of the drainage basins of six major tributaries of the Medicine Creek basin, the locations of which are shown in figure 176. The purpose of the measurements was to de- termine the drainage-basin characteristics that cor- relate most closely with water and sediment yield and with the development of gullies. Where gaging 12,000 to 60,000. 40°35’20” LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 100°30’ 285 “m4 OWN gl 99J'D 5 MILES c 5%“ . (.__ (wvv ‘ W C? 1 '- eve /-‘ fl\ gm \1 \J 1 J MILE ,/ XN/fi v n z' N 1 J ' - g Q _ L\ meow % ° . ’- . m “k { .. \EgofirIER/ W95 T‘9N. FIGURE 196.—Capture of head of North Plum Creek by Deer Creek. In inset, which is reproduced from the Stockville NE quadrangle, relict topog- raphy of the depositional surface of Peorian time is shown at right. 286 stations (also shown on fig. 176) are not at the mouth of a basin, only that part of the basin up- stream from the station was measured. Ordering of the drainage system was carried out according to the method of Strahler (1952, p. 1,120), which is a modification of the method of Horton (1945, p. 281). Strahler’s method is not only more objective and convenient, but it is also more valu— able for many hydrologic purposes. A channel seg— ment of third or higher order obtained by Strahler’s method is of approximately uniform cross-sectional area throughout its length, but this is not true of channels obtained by Horton’s method. The word “channel” is used here in the same sense that the word “stream” is used by Horton and Strahler in connection with ordering. A channel is an established watercourse that may transmit water continuously, intermittently, or ephemerally. The flow of water in most channels of the Medicine Creek basin is ephemeral, and the term “stream” is inappropriate for them. Properties of fluvially eroded landforms and rec- ommended symbols for these properties are sum— marized by Strahler (1958, p. 282—283). The prop- erties discussed in this report and the symbols used are listed below. Measured properties are those that can be directly measured or counted, and de- rived properties are those that cannot be directly measured but must be computed from measurements. Measured properties: Total area of drainage basin, A Area of upland in drainage basin, Au Area of valley system in drainage basin, A* (A* = A — Aa) Channel order, u (for examples: ul, first order; ue, sec- ond order) Number of channels of order u, Nu Channel length, L Channel length, mean length of segments of order u; L. Basin relief, H Basin length, L», Derived properties: Channel slope, Sc Valley slope, S1, Bifurcation ratio, R!) (R. = 5.1) Channel length ratio, RL (R.= 5:.) Channel frequency, Fu (Fu = IX“) Adjusted channel frequency, F: (F? = 29:) EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT Drainage density, D ( v = 214-) Relief ratio, Rn A ( R» = Z Elongation ratio, Re i mete f circle (R=——— ) For practical purposes first-order channels in the Medicine Creek basin are defined as the lowest order of channels represented by V-shaped bends in con— tour lines on the available topographic maps. For- tunately, the accuracy of the topographic maps in representing small drainage channels is excellent, and the scale (1:24,000) and contour interval (10 ft) are also favorable for representation of small channels. Drainage patterns of several fifth—order basins were plotted both on aerial photographs and on topographic maps, and comparison showed that results obtained by the two methods were very similar. Nearly all the first-order channels represented by V-shaped bends in contour lines are channels that formed immediately before or during accumulation of the Stockville terrace deposits, and the valley sides leading to these channels show some degree of grading. In only a few places has modern gullying proceeded far enough to form channels where no channels had existed previously. Each of the six major subbasins was separately outlined on the topographic maps, and each subbasin area was measured with a polar planimeter. In addition, the area of upland in each subbasin was measured. Upland is defined as undissected remnants of the land surface that existed at the end of deposi- tion of the Peorian. For practical purposes upland is distinguished on topographic maps by smooth contour lines—that is, by contour lines that do not have V-shaped indentations. Where areas of upland are continuous around the periphery of a drainage system—as in basins I—1’, A’, 0—4’, and C—2’ of figure 197—the area of upland was determined by measuring with a polar planimeter the area actually occupied by the drainage system and subtracting this area from total basin area. For Dry Creek and Lime Creek, every channel of every order was drawn in color on topographic maps, counted, and measured individually. Sampling pro- cedures, similar to those described by Leopold and Miller (1956, p. 16), were tested for the approxi- mation of number, mean slope, and mean length of first- and second-order channels in the other sub- basins. These procedures were used for Mitchell Creek, Well Canyon, Fox Creek, and Brushy Creek. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 287 2000 0 2000 4000 FEET I I l 1 l l J CONTOUR INTERVALS 10 AND 20 FEET 4 DATUM Is MEAN SEA LEVEL FIGURE 197.—Variations in drainage texture in the Medicine Creek basin. Basins shown are fourth or fifth order. 288 Two derived properties—drainage density and channel frequency—commonly are used but are am— biguous for so-called immature or youthful basins in which areas of undissected upland remain. For such basins a distinction must be made between the properties of the drainage system and the properties of the drainage basin. If a particular drainage sys- tem were transferred without any change in prop- erties to a larger drainage basin, the values of drainage density and channel frequency would change regardless of the constancy of the drainage system. Drainage density is also ambiguous for another reason: a drainage system consisting of short channel segments may have the same total channel length as another drainage system that consists of fewer but longer segments. Melton (1958, p. 36) has proposed that the ratio of F/D2 is constant for mature drainage basins, but this is not true of Medicine Creek, which for the most part is immature. To compare the channel frequency of two immature drainage systems whose drainage basins differ in percentage of upland, drainage basin area must be adjusted by subtracting the area of upland. Channel frequency, expressed as number of channels per square mile of area actually occupied by the valley system, is here called adjusted channel frequency. Relief ratio was defined by Schumm (1956, p. 612) as the ratio between H, basin relief, and Lb, the longest dimension of the basin as measured parallel to the principal drainage channel. Basin relief is the difference in altitude between the highest and lowest points in the basin. A measure of basin EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT shape, also devised by Schumm, is the elongation ratio, which is derived by dividing the diameter of a circle having area equal to that of the basin by the basin length, Lb. A measure of mean valley—side slope is difficult to obtain because the side slopes leading to low-order channels are steeper than side slopes leading to high-order channels. For example, measurements on topographic maps indicate that slopes leading to first-order channels on the lower part of Dry Creek have a mean angle of about 25°, whereas slopes leading to the main channel (which is seventh order) have a mean value of about 12°. Slopes lead- ing to the main channel of lower Dry Creek were measured in the field during the preparation of 12 transverse valley profiles. Of 23 slopes, 13 ranged from 8 to 11°, only 2 were greater than 15°, and none were less than 6°. The mean angle was 11.5°; the mean deviation, 33°; and the mode, 10°. Mean side slopes of third- and fourth-order channels prob- ably represent the best approximation to mean values for the drainage system as a whole and are, therefore, quoted in table 3. DRAINAGE TRANSFORMATION AND VARIATIONS IN DRAINAGE TEXTURE Variations in drainage texture are accompanied by variations in most drainage-basin characteristics. The major variations in drainage texture within the Medicine Creek basin are illustrated by the repre- sentative fourth- and fifth-order drainage basins in figure 20. Basins 0—2’ and K—l’ are typical of the upper part of Medicine Creek; I—1’ and 0—4’, of the TABLE 3.—Measured and derived properties for the Medicine Creek basin and its major subbasins First-order channels Valley Mean . Area of Area of slope of valley Mean Relief Elonga- AdJUSted Drainage basin Total upland, valley main side upland ratio, tion Mean Mean freouency, area, A a flat segment slope slope Rh ratio, length, slope, F‘u (sq mi) (percent) (percent) (ft per ft) (ft per ft) (ft per ft) Re L Se (channels (ft) (ft per ft) per sq mi) Lime Creek....__.___.. ...... 11.6 37.0 14.0 ”......n... 0.179 0.026 0.0114 0.662 410 0.110 140 Mitchell Creek ....... 52.1 51.8 11.5 0.00299 .200 .013 .0050 .416 370 .100 165 Brushy Creek (above gage)_._......___..._......... 73.8 25.9 16.0 .00367 .412 .022 .0064 .734 192 .234 260 Dry Creek (above gage) : Total ........................ 21.1 34.8 14.2 ................... .360 .032 .0079 .475 267 .187 203 Upper only. -. 12.6 25.0 .................... 00550 260 .195 214 Lower only.___...._..___..._._...... 8.5 49.5 .................. __ 00366 287 .160 185 Fox Creek: Total ............................. - 72.5 16.5 27.2 00360 .363 .037 .0059 .445 250 .241 250 Upper only....___.._..__ 57.0 12.6 .................... 00386 Lower only 15.5 30.4 00810 Well Canyon: Total 53.3 20.4 22.6 00286 _______________ .036 .0052 .335 245 .251 235 Upper only 22.7 12.3 00290 260 Lower only ............ 30.6 26.5 ................. 00284 220 Medicine Creek above gage above raervoir........._... 549.0 30.0 .00435 ______________________________________________________ 200 LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA central part; and A’ and C—2’, of the lower part. Major differences in these small basins, as in the larger subbasins of which they are representative, lie in the percentage of upland, the length and fre- quency of first-order channels, and relief ratio. These differences can be conveniently related to differences in geomorphic history. In basins A’ and C—2’ valleys at the end of Peorian deposition were broad and shallow, and the gently sloping valley sides were graded all the way to the interstream divides. Divides were rounded, and no areas of flat upland remained. These valleys were deeply incised during the episode of erosion that preceded deposition of the Stockville terrace de- posits, and channel frequency was increased. During accumulation of the Stockville terrace deposits, the transformed drainage system was smoothly graded, and the gently sloping tips of first-order tributaries were extended nearly to the former drainage divides. Basin C—2’ represents a less advanced stage of grading than Basin A’, which is nearer the mouth of Medicine Creek; it was incised first, and conse- quently underwent a longer period of grading dur- ing accumulation of the Stockville terrace deposits. Basins I—1’ and 0—4’ are representative of the central part of the Medicine Creek basin, where the valleys at the end of deposition of the Peorian were more narrow than those in the lower part and were separated by broad, nearly flat uplands. Later, these valleys were deeply incised, and the drainage was extended into the upland; however, large areas were left undissected. The sides and heads of first— and second-order valleys were smoothed and some- what reduced in angle during deposition of the Stockville terrace deposits, but they remained rather steep. The surface into which basin 0—4’ was incised is unusually flat, because it represents the valley flat and gently sloping valley side of Medicine Creek at the end of deposition of the Peorian. In basin 0—2’, valleys at the end of deposition of the Peorian were narrow and steep sided (average slope angle about 14°). The channel frequency, al— ready considerably higher than that in the lower part of Medicine Creek, was greatly increased dur- ing the episode of incision preceding deposition of the Stockville terrace deposits. Moreover, the newly incised valleys, although somewhat graded during Stockville time, retained their steep sides and heads. Later episodes of incision have further increased the frequency of first- and second-order channels. Basin K—1’ is similar in history, but it formed on a less steeply sloping surface and underwent a greater degree of grading during the deposition of the Stockville terrace deposits. 289 500.000_ 11 111 111 1 1 1111 I 100,000 I lllllll I llllll 10,000 I I llllll llIllllll 1000 MEAN CHANNEL LENGTHC IN FEET [lllllll llllllll l1 l | l | l CHANNEL ORDER 100 " FIGURE 198.—Mean channel length in relation to channel order. In summary, variations in the modern drainage pattern are related to two conditions that existed at the end of deposition of the Peorian—the extent of drainage transformation that preceded deposition of the Stockville terrace deposits and the amount of grading that took place during deposition of the Stockville terrace dep0sits. To a minor extent, the drainage pattern has been modified by episodes of incision that followed deposition of the Stockville terrace deposits. At the end of Peorian deposition, valleys were broad and shallow in the lower part of the basin, narrow and separated by flat divides in the central part, and narrow but separated by nar- row divides in the upper part. Grading during deposition of the Stockville terrace deposits reached an advanced stage in the lower part of the basin, and in general the stage of grading decreases in an upstream direction. The curves representing the relation of mean chan- nel length to channel order (fig. 198) can be ar— ranged in a sequence that illustrates some of the major variations in drainage texture. The length of channel segments (of orders one through five) shows a consistent decrease from Lime Creek, in the lower part of the basin, to Brushy Creek, in the central part. Also, RL between orders one and two, as well as between orders two and three, is less than RL between higher orders. The change in slope at the lower end of the curves applies only to orders one and two for Lime Creek, but it applies to orders: one through four for Dry Creek and to orders one through three for Well Canyon, Fox Creek, and Brushy Creek. An upward concavity of the curves representing the relation of mean channel length to channel order, 290 20,000 If! 1|} ll Illll It 10,000 I lllllll 2... /°T m / 7cm / inn _/ I lllllll 1000 \L1 K, (i; 20 3X2?) 4 ‘1 “{y \ K\ K; K K 5 6 6 5K ( \ s 7 _ 1 ll [KIA61\i176\l16\lll\lli\78 CHANNEL ORDER FIGURE 199,—Number of channels of each order in relation to channel order. llllllll / N ’ w w (u ——o A [[111]]! J} J; 44» oweo 4% l & Jgfia 04 l .. o 0“ 04 “99’ U1 l i I 1 ll i 100 1 NUMBER OF CHANNELS I Illllll I Illlll I IIIIHI w A as plotted on semilogarithmic paper with channel order as independent variable, was noted by Strahler (1952) and Broscoe (1959, p. 5). Broscoe concluded that Horton’s (1945) “law of stream lengths” may not apply when Strahler’s method of ordering the drainage system is used. However, the basins ana- lyzed by Strahler and Broscoe were of fourth or smaller order; basins of larger order must be ana— lyzed for proper definition of the relation. In the Medicine Creek basin the relation is exponential, but the exponent changes at second or third order (fig. 21). If a drainage system having low density of chan- nels is transformed by the growth of new first— and second-order channels without extension of the higher order channels, a marked absolute shortening of lower order channels can take place with- out much shortening of higher order channels. The order of higher order channels may change, but the absolute length remains nearly constant. A trans- formation of this kind took place in the Medicine Creek drainage basin at the end of deposition of the Peorian. Furthermore, the transformation can take place without much change in the bifurcation ratio. The bifurcation ratios of the same subbasin, as well from one subbasin to another, are remarkably con- stant. (See fig. 199.) One result of a transforma- tion of the kind described is a change in slope, at lower channel orders, of the curve relating mean channel length to channel order. The major sub- basins that have undergone the greatest drainage transformation and, consequently, have the highest frequency of first-order channels are those that show the change in slope extending to third or even EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT fourth order. In a sense the change in slope of the curve relating mean channel length to channel order is a measure of disequilibrium of the drainage system. GULLY EROSION DEFINITION AND CLASSIFICATION OF GULLIES In a region such as the Medicine Creek basin, where many of the drainage channels have steep, unvegetated sides and are incised in massive silt, the distinction of a gully from other drainage chan— nels is difficult. According to general usage, the essential features of a gully are its size (which is larger than a rill), recency of extension in length, steepness of sides and head, incision into unconsoli- dated materials, and ephemeral transmission of flow. Gullies in the Medicine Creek basin are incised into the sides, head, or bottom of previously established drainageways, which are deepened or extended by development of the gully. The lower limit of depth for a gully is placed at about 2 feet because channel head scarps that are less than 2 feet in height show a very slow rate of advancement. Most of the actively advancing head scarps in the Medicine Creek basin are greater than 6 feet in height. The lower limit of width is arbitrarily placed at about 1 foot. The criterion of recency of extension requires that some evidence be obtained for extension within a period of a few years. Extension can be established by com- parison of aerial photographs, by comparison of field measurements made at intervals of a few years, or by indirect evidence observed in the field. If the sides and head of a drainage channel have a slope less than about 45°, the channel is probably not ac— tively advancing and hence would not be considered a gully. In this report the following definition of a gully is followed: A gully is a recently extended drainage channel that transmits ephemeral flow, has steep sides, a steeply sloping or vertical head scarp, a width greater than about 1 foot, and a depth greater than about 2 feet. The sides and heads of gullies in the Medicine Creek basin are- steep slopes, straight in profile, that are here called erosional scarps or simply scarps. In addition to the erosional scarps at the sides and heads of gullies, other erosional scarps are conspicuous elements of the landscape. These include step scarps on slopes (Brice, 1958), scarps that form the fronts .of terraces, and scarps at the base of valley-side slopes. The inclination of a scarp ranges from about 45° to nearly 90°, and the height ranges from a foot to about 40 feet. Generally, the top of a scarp meets the surface into which it is cut at a sharp angle, but the base of the scarp decreases in LOESS—MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA slope as it merges with the surface below. The term “channel scarp” is appropriate for a. scarp that forms a break in the long profile of a well-defined channel. If a particular channel scarp is the head- ward terminus of a gully and attention is to be drawn to this fact, the channel scarp may be called a head scarp. Not all gully head scarps are channel scarps. The term “head cut,” although commonly applied to the headward terminus of a gully, is un- satisfactory because it may refer either to the scarp at the head of a gully to to the whole gully head. Moreover, a channel may have many scarps along its length, and the designation of several of these as head cuts is confusing. The scarps that form the sides of a gully along its length are here called side scarps. Study of gullies in the field and on aerial photo- graphs of the Medicine Creek basin has shown that the depth of a gully, its areal pattern, and its rate of growth are more closely related to the topographic position of the gully head than to any other single factor. Of particular significance is the location of the gully head in relation to the previously estab- lished drainage system. On the basis of location, gullies are classified as valley-bottom gullies, valley— head gullies, and valley-side gullies. Inasmuch as valley bottoms grade smoothly into valley heads and valley sides, the distinction among the different classes of gullies is arbitrary. Moreover, a valley— bottom gully becomes a valley-head gully as its head scarp migrates into the valley head. The valley-head gullies are by far the most numerous kind, and nearly all of these are in steep valley heads that border areas of upland. The size of a gully depends on its depth and areal dimensions; but because of the irregular shape of gullies, the areal dimensions cannot be expressed simply and consistently. For complexly branching gullies, it is not clear whether or not the uncon- sumed area between the branches should be con- sidered part of the gully. The best approximation to an expression of gully size seems to be the maxi- mum width of the gully head. Inasmuch as gully depth increases roughly in proportion to width of gully head, the width of the gully head is an indirect expression of gully depth. In an arbitrary ranking of gullies according to width of gully head, valley— side and valley—bottom gullies must be considered separately from valley-head gullies, which are rela- tively wider in proportion to depth. AGE AND ACTIVITY OF EROSIONAL SCARPS Obviously, a scarp is younger than the surface into which it is cut. All the scarps in the Medicine 291 Greek basin are considered to be younger than the Stockville terrace deposits. Some of the slopes graded to the level of the Stockville terrace are steep, . especially those in the upper parts of the drainage basin; but these slopes are not regarded as scarps because they have an inclination less than 45°. Also, the steepest part of these slopes (in a particular area of the basin) is graded to about the same angle, and the upper parts of the slopes de- crease in angle as they merge with the older Peorian surfaces. ‘ Regardless of the age of the surface it cuts, a scarp may be either active or inactive. A scarp is considered to be active if it has migrated a distance that can be measured by comparing of aerial photo- graphs made in 1937 and in 1952 or if it shows field evidence of recent movement. In the field, inferences as to activity were made from the type of vegetation on the scarp, from the degree of inclination, or from evidence of recent or imminent slumping. A thick sod of native grass indicates an inactive scarp; a cover of brush (such as wild rose, buckbrush, or currant) indicates a moderately active scarp; and lack of vegetal cover, or a cover of weeds, indicates an active scarp. In general, the more nearly vertical a scarp, the more likely it is to be active. However, many scarps that were bare and nearly vertical showed little or no migration during the interval 1937—52. Even a bare and nearly Vertical scarp was not regarded as active unless recently slumped silt was observed at the scarp base or unless fissures were observed on the ground surface beyond the scarp. CHANNEL SCARPS AND VALLEY-BOTTOM GULLIES Varieties of valley-bottom gullies are distinguished on the basis of position with respect to one another and to different topographic levels produced by trenching of the drainage system. The main va- rieties are indicated by numerals in figure 200. The two gullies at upper left (1), which are in the same valley bottom but are separated by a reach of un- dissected valley, are of the sort called discontinuous by Leopold and Miller (1956, p. 29). When the head scarp of a downvalley gully reaches the tail of an upvalley gully, a stage of coalescence is reached; and the two gullies are integrated into a single trench as the downvalley scarp advances. Similarly, the advance of a major scarp in a large valley is commonly preceded by the advance of a minor scarp (2). A scarp may be initiated on the floor of a trench, where it may advance either into freshly deposited sediment (3) or into previously undis- turbed alluvium. Channel scarps on the floor of a 292 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT . ll] ). l -- (MAX .1 __ ‘Q , " “ ,; '._" _ '§4§‘.\.\\§\c‘ PI ifi “1‘1 \\\\/\:‘\‘3'—" W). i ‘3sz - \ . -\ ~ \*~_ (3‘. L’ {bill/5155f; 4’ ‘ a r‘ . . ' ~ I '7 l. . > I. ’ ~ ‘ V . ... -» ' 1”". J's-5‘"- :' -. .. W, ,. We «lav-'5 . . -\ -9t u 'a 121/0 ‘5 “ In“. ~ :. . -' ,l , ' AH", _I.;hl\.¢l,\n.- I xv, ‘1 \ "\\\u\ .3 . h. . . .-".(.:.‘-.:$,-:.-)g,"' FIGURE 200.—Composite sketch, based on field sketches and photographs, showing the different varieties of valley-bottom gullies. Actively eroding scarps are indicated by darker shading. Numbered features are described in text. trench are here described as inset. Remnants of the original gully floor, left along the gully sides after passage of an inset scarp, constitute an inset terrace (4) . Another variety of valley-bottom gully is illus- trated by the two gullies (5) in the middle ground of figure 200. Unlike the discontinuous gullies, which have no relation to the topographic level formed by trenching elsewhere in the drainage system, these gullies were initiated at the side scarps of the gully in the foreground and are accordant with its floor. . The location of the head scarps of major valley— bottom gullies in the Medicine Creek basin is shown in figure 201. Most of the gullies are in the size category of very large, but some are in the category of large. The scarps range in height from 10 to 25 feet and are in valleys of fourth or higher order. Advance Of the scarps during the period 1937—52 ranges from a maximum of about 3,400 feet (for gully B4 on fig; 201) to a minimum of about 200 feet. TWO significant" facts relevant to the origin of the scarps are indicated by their areal distribution: Scarps are most numerous in the central and lower part of the basin, and scarps in large valleys are commonly upstream from the confluence of a large tributary. The greater abundance of the scarps in the central and lower part of the basin is attributed to the relative narrowness of the valley bottoms there. (See table 4.) In general, the width of a valley bottom is much affected by the degree of trenching both before and after deposition of the Stockville terrace deposits. The branching valley at upper left (6) in figure 200 was represented by a relatively narrow and shallow trench before deposition of the Stockville terrace deposits. During deposition of the Stockville terrace deposits the valley sides were graded, and the valley has been little affected by post-Stockville trenching. By contrast, the valley at upper right (7) was represented by a relatively deep and wide trench before deposition of the Stockville terrace deposits. Post-Stockville trench- ing has extended to the valley head, and only narrow LOESS—MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 293 ‘n . ‘ J .. \ h f \\ f \ x W \x‘ 3 i \ _ 46A x L \ \/ f 202%" \ < \ B7 ‘ “ n \ \ \ ‘ \ r \ . a l ’ \ \J\\ \.\ \ \ BS I) «1) \ ‘ l\ \I 35‘ E} .‘ \ ”a; A\.’\\ \\ l\ (5:. 5 /I Q 32 E \A‘. ; \ \\ \ {<34 I 4/ z 1 \ \ ‘\ J E33 } \ 3 (9 \ x \ \ ( :l \ \ «9,» Ca, \ \ \ \ \ ; r I . w ‘1; \ \ \ l J BS ,\ \ \ J \ t 07-9916 ‘0 \ x \ \ \ '3. \ '.\ I We \\ \ egg \ l \ <<= ‘\ (“\J $99,. \ l\ \\ f g N g \\ g; N {PW Spring \‘ \ 0%: \‘ a / EXPLANATION L. >78” b / \ ‘ ' \ 62% \ \N \ Heavy transverse K 9 Cr ~ line indicates posi- \. ,v ‘ tion of head scarp \ /J ? 88 “ N‘ \ Numbered gully ‘ See text for description \ I, Xx FIGURE 201.—Head scarps of major valley-bottom gullies. 294 TABLE 4.—Drainage system properties, listed according to channel order, for major subbasins of the Medicine Creek basin Mean Chan- Number Mean Mean width Area of Subbasin nel of channel channel of val- valley order channels length slope ley flat flat (ft) (ft perft) (ft) (sqmi) Lime Creek ,,,,,,,,,,,,, 1 1,020 410 0.1100 25 0.37 2 233 670 .0517 50 .28 3 42 1,600 .0185 105 .25 4 9 4,110 .0111 120 .16 5 2 9,200 .00545 150 .10 6 1 26,000 .00384 500 .47 Mitchell Creek ________ 1 3,940 370 .100 35 1.83 2 950 537 55 1.00 3 172 1,170 1 90 .68 4 31 4,970 __ 140 .77 5 6 3,320 __ 160 .11 6 1 114,000 .................... 400 1.63 Dry Creek ................ 1 2,824 267 .187 35 .95 2 725 458 0673 60 .71 3 153 1,023 0343 96 .54 4 30 1,816 0155 125 .24 5 4 8,830 0088 160 .20 6 2 8,600 00576 200 .12 7 1 25,500 00296 260 .24 Well Canyon ........... 1 10,310 245 .251 35 3.17 2 2,460 480 70 2.85 3 545 772 W 110 1.75 4 104 2,195 140 1.23 5 24 5,375 .. 190 .88 6 1 126,000 545 2.18 Fox Creek ................ 1 15,900 250 35 4.98 2 4,250 460 .100 70 4.90 3 904 725 0465 120 2.82 4 177 2,040 0215 150 1.94 5 35 4,085 0101 190 .97 6 6 14,850 0055 270 1.71 7 2 32,000 00406 440 1.01 8 1 45,000 00274 880 1.40 Brushy Creek ________ 1 12,400 192 .234 35 2.99 2 3,500 370 .................... 60 3.39 3 729 674 95 1.67 4 147 2,040 135 1.44 5 29 5,450 165 .93 6 6 12,833 240 .66 7 2 40,500 300 .87 remnants of the Stockville terrace remain along the valley sides. The valley at right in figure 200 is typical of the upper part of the Medicine Creek basin; and the valley at left, of the central and lower part. The association of major channel scarps with the confluence of a large tributary is notable on Fox Creek, Cut Canyon, Curtis Creek Canyon, and Well Canyon. Although the present scarps are up— valley from the tributary confluence, the possibility is good that the scarps were initiated at the con- fluence, where the slope of the main valley may be locally steepened. After a valley floor has been trenched, its long profile cannot be accurately re- constructed; however, long profiles drawn from con- tours at 10-foot intervals on the topographic maps indicate that the main valley profile is generally, but not everywhere, steepened upstream from the EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT confluence of a large tributary. (See fig. 202.) On Curtis Creek Canyon a conspicuous steepening 0f the valley profile just upstream from the confluence of a large tributary is indicated on the topographic map. The steepening is probably due to deposition of a fan at the mouth of a tributary. This fan ponds the drainage in the main valley and leads to deposi- tion upstream from the fan. Schumm and Hadley (1957) have shown that, in small drainage basins in eastern Wyoming and northern New Mexico, discontinuous gullies can form on locally steep valley reaches. The proposal that gullies are initiated at a local steepening of the valley slope has as a corollary the proposal that local slope in an ungullied valley reach may be adjusted to water discharge. In the absence of discharge measurements for most valley reaches, the assumption may be made that the discharge in a reach is proportional to the upstream drainage area. The relation of local valley slope to drainage area for reaches along three valleys is shown in figure 203. Cut Canyon and Dry Creek are free of major channel scarps in their uppermost reaches, as shown in figure 202, but Coyote Creek has channel scarps throughout most of its length. In spite of the fact that the floors of all three valleys are well above bedrock, local slope is not constant in a given valley reach several hundred feet in length. The values of local slope plotted in figure 203 represent the best approximation obtainable from topographic maps for mean local slope in a reach bounded by contours and having a fall of 20 feet. A general relation between local valley slope and drainage area is apparent from figure 203. Slope values applying to gullied reaches tend to plot above the average curve representing this relation, but some slope values applying to ungullied reaches also plot above it. Aside from chance, factors other than local slope and drainage area seem to be involved in the initiation of gullies, and the most important A of these is probably valley width. The upstream reaches of Coyote Creek have a greater susceptibility to gullying than reaches of Cut Canyon because of the relative narrowness of the valley bottom of Coyote Creek. The main valley of Well Canyon is relatively free of gullies, and the few gullies present have grown slowly. The channel scarps in the long profile of Well Canyon (fig. 202) did not advance during the interval 1937—52 by an amount measurable on aerial photographs nor did the intricate small-scale mean— derings of the trenched channel reaches show any observable change. The aspect of a trenched channel on Well Canyon is shown in figure 195 (middle). LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA I l I I I I I I I I I I I I I I l VALLEY PROFILES _ 700 ~— — 600 "— — \ 500 —— 400 —— — Ix kg \ VERTICAL SCALE, IN FEET .1- 200 -— _ \ 100 —— — \ E l I l i L l l l l l J l l 25 15 10 5 O HORIZONTAL SCALE, IN MILES FIGURE 202.—Long profiles of valley and channel for some major tributaries to Medicine Creek. A break in the channel profile represents a channel scarp, and an arrow on the valley profile indicates the point of confluence of a large tributary. Profiles drawn from topographic maps. The general slope of the valley profile, although Somewhat lees than that of Dry and Curtis Creeks, is not significantly different from that of Mitchell and Fox Greeks. The tributaries received by Well Canyon are relatively short, however, and the valley of Well Canyon is relatively wide at the point of confluence of the two largest tributaries. Upstream from the confluence of the northernmost of these tributaries, the valley profile of Well Canyon shows lccal steepening and the presence of a channel scarp. Neither local steepening nor a channel scarp occurs at the confluence of the other tributary. The inac- tivity of channel scarps in the main valley of Well Canyon is attributed to the lack of large tributaries, to relatively great valley width, and to the generally low valley slope. DEVELOPMENT OF VALLEY-BOTTOM GULLIES‘ Photographs of typical valley—bottom gullies are reproduced in figure 204. After a gully has become large, neither the exact way in which it was initiated nor the point of initiation can be established. How- ever, study of small gullies in various stages of de- velopment indicates that valley-bottom gullies may begin as small depressions scoured by flowing water on the valley bottom. Breaks in the sod cover—such as might result from an animal burrow, a trail, or an excavation by man—are likely spots at which scour can take place. The upstream side of a de- pression can evolve into a scarp and advance up- Valley, meanwhile growing in height. Although most scarps in major valleys have apparently been ini- 296 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 0'1 I I Illlil I [Ill 1 l l l | I II I l l l I | l EXPLANATION Upper Cut Canyon 11111— A U pper Dry Creek O Coyote Creek Arrows indicate gullied reaches LOCAL SLOPE, IN FOOT PER FOOT o S lllllll l [>-‘— [Ail . <— <— I Illlill l Illllll 0.001 0.1 10 100 DRAINAGE AREA, IN SQUARE MILES FIGURE 203.—Local valley slope in relation to drainage area for reaches along upper Cut Canyon, upper Dry Creek, and Coyote Creek. tiated in a locally steepened valley reach, a scarp, once initiated, can continue to advance into reaches where the value of local slope is conservative in relation to drainage area (water discharge). More- over, many gullies occur in valleys of fourth or smaller order in which no evidence of local steepen- ing can be found. The value of the ratio of slope to drainage area that is associated with the initiation and rapid growth of a gully is not sharply defined. Advance of a head scarp takes place as the mas- sive silt at the base of the scarp becomes saturated and disintegrates. In this region the permeability of undisturbed dry loess in place is only about 0.8 foot per day (Holtz and Gibbs, 1952), but this value has no significance for a vertical face of loess that is immersed in water, as at the edges of a plunge pool. The loess in contact with the water continually sloughs along vertical joints and is removed by the turbulent water of the pool. In figure 204 (upper left) the recently collapsed base of a scarp may be seen at right of plunge pool. The cohesiveness of massive silt depends to a considerable extent on a cement formed by clay minerals that coat the silicate grains, and when wet the clay coatings have little or no cohesive effect. Gully side scarps regress by the same process. For a distance up to 12 feet from the edges of a gully, the ground is broken by cracks marking the boundaries of large blocks that have slumped because of saturation of the silt at their base. These blocks will eventually slump into the gully, as illustrated in figure 200. Head scarps can be maintained in massive silt whether or not they are capped by sod or some other resistant layer. Drainage is conveyed to many head scarps by a narrow channel, the bottom of Which is bare of vegetation and is cut below the soil profile. A deep notch is formed by the intersection of this channel with the head scarp. The notch in the head scarp of gully B1 on Dry Creek changed very little between 1953 and 1957. Downstream from a head scarp, long profiles of gullies change from year to year; the change de- pends on the amount of runoff. V. I. Dvorak (writ- ten commun., 1962) has compiled measurements of the long profiles of three major gullies on Dry Creek, which are designated gullies B1, B2, and B3 in figure 201. Changes in the long profile of gully B2 are typical and are represented in figure 205. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 297 Upper left, Gully in main channel of Dry Creek, designated gully B1 in figure 201. Photographed on July 5, 1956. Lower left, Oblique aerial photograph of gully B1, September 4, 1965. Upper right, Center, valley- bottom gully below large valley-head gully, which is designated by numeral 3 on figure 214. Note cowpath erosion on side slope of valley. Oblique aerial photograph, September 4, 1965. Lower Tight, Discontinuous valley- bottom gullies of a tributary to East Curtis Creek Canyon, SE14 sec. 31, T. 10 N., R. 27 W. Oblique aerial photograph, September 4, 1965. FIGURE 204.—TYPICAL VALLEY-BOTTOM GULLIES According to Dvorak, the relatively high runoif in 1951 was accompanied by gully slopes distinctly less than valley; slopes, but during the succeeding drier years the gully slopes increased. Gully slope was steeper than valley slope for gully B2 in 1956 and for gullies B1 and B3 in 1960. The maximum height attained by a head scarp is evidently controlled mainly by the steepness of the slope into which the head scarp is advancing. The highest head scarps in the basin, which reach a maximum of about 35 feet, are all in steep valley heads, whereas head scarps on the gently sloping valley bottoms all range from 10 to 25 feet in height. The height that a scarp can attain is also limited by the rate of deposition at, and downstream from, its base. The cross profile of a valley-bottom gully changes in a direction downstream from the head scarp, but the nature of the change is not consistent from gully to gully. Typically, gully depth decreases very gradually in a downstream direction, whereas gully width reaches a maximum within a thousand feet of the head scarp and then decreases in a down- stream direction. Downstream changes in the cross profile of a typical rapidly advancing valley—bottom gully (gully B4 in fig. 201) are shown in figure 206. The width of gully B4 remains nearly constant for about 5,000 feet downstream, but for other gullies the width decreases downstream at a much faster rate than depth. For example, in 1952 the width of a large valley-bottom gully on Curtis Canyon (gully B6) decreased from about 105 feet at a 298 EROSION AND SEDIMENTATION 90 IN A SEMIARID ENVIRONMENT I I I I I I l I I I I I 80 / / g \jaIIeY flat /——_/ ,_ _ E > // ’,,\//'/7—= I! 70 ‘r f < I I ' a: / I: e — ’ / < 1952/1I [‘1956 1960 — L; 1951 I / O 60 ’ I '2 ’ / __.E/—’ ’_ ____£_-=7/._#—/ ’ LIJ __ /__ _/ _//—/ I //l I11 / , / ,_—/ — _ __’ _ _— I '__, E I"! — “—I- J”? ui 50 __” .'——:r:——’-:=" D /-_ D I: I— — _ .1 < 40 30 I 1 | I I I I I I I I l I I o 500 1000 1500 DISTANCE FROM ARBITRARY FIGURE 205,—Changes in long profile of a large valley-bottom gully on Dry Creek, 1951—60. Gully is designated B2 in figure 201. From V. I. point 900 feet downstream from the head scarp to about 45 feet at a point 5,000 feet from the head scarp. Through this 5,000—foot distance, the depth of the gully changed little. Neither gully B4 nor B6 has a well-defined downstream terminus at which the gully floor merges with the valley floor; each merges downstream with a narrow but apparently stable channel that continues for miles along the valley bottom. From his study of valley—bottom gullies on Dry Creek, Dvorak (written commun., 1962) concluded that widening downstream from the head scarp effectively ceased when the width-depth ratio was in the range of 3.5 to 5. At great width, the depth and velocity of flowing water would be insufficient to remove material that accumulated from bank slumping, and the channel bottom would assume a parabolic shape. As is generally the case with channels in nature, the width-depth ratio that can be maintained for a gully probably depends on the magnitude of water discharge. Small gullies in the Medicine Creek basin have low values of the width-depth ratio, and large gullies have larger values. In 1956 none of the large valley-bottom gullies had attained a width-depth 2000 REFERENCE POINT, IN FEET Ovorak. ratio greater than 6, which was the value measured for gully B5 at a point 390 feet downstream from its head scarp. However, none of these gullies had a drainage area greater than about 16 square miles. Gully B9, the drainage area of which was about 1.5 square miles, had in 1956 a Width—depth ratio that did not exceed a value of 1 for a distance of 1,000 feet downstream from its head scarp. The advance of major valley—bottom gullies during the period 1937—52 is given in table 5, together with pertinent information as to gully size and topo- graphic setting. Gully advance and width of valley bottom were measured on aerial photographs; drain— age area was measured on topographic maps; and height of head scarp was measured inthe field. Slope of the valley bottom, which applies to the slope upvalley from the 1952 position of head scarps, was measured by field surveys for some of the gullies and on topographic maps for others. In general, the maximum depth of a gully is about equal to, or is greater than, the height of its head scarp. No cor- relation is apparent between the rate of advance and any of the other measurements. The estimate of the year at which active advance of the gully began is made by dividing gully length by average LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA TABLE 5,—Measurements relating to the size and topographic setting of major valley-bottom gullies [Gully locations are indicated on fig. 201] Width Slope Height Advance ‘ Esti- Drainage of of of head of head ~mated Gully area valley valley scarp scarp, year of in 1952 bottom bottom in 1956 1937—52 acti- (sq mi) (ft) (ft per ft) (ft) (ft) vation 6.1 215 0.0056 24 900 1850 5.6 210 .0055 22 750 7 .8 120 .0105 20 610 ‘.7 3.4 140 .0082 16 3,400 1900 9.7 200 .0058 18 2,400 1920 9.8 160 .0062 18 1,750 1920 15.9 180 .00415 18 2,800 1910 9.5 150 .0050 20 2,100 ? 1.4 120 .0125 23 2,100 1920 annual rate of advance during 1937—52. For Dry Creek, which has a continuous channel downstream from the head scarp, the point of origin is assumed to be at the confluence of East Fork. Valley-bottom gullies advance only during periods of runoff. As a result of rainfall on July 4 and 5, 1956, which totaled about 1.9 inches, gully B5 ad- vanced about 15 feet and gully B1 advanced 20 feet. The year 1937 was preceded by a period of relatively O C) O 299 low rainfall (fig. 179) and, as observed on aerial photographs, most of the gullies were inactive in 1937 and were partly filled with sediment. A sub- stantial part of the advance during the period 1937— 52 took place in 1951 as a result of high runoff. For example, of the 750-foot advance made by gully B2 during the period 1937—52, about 350 feet was made during 1951. Differences in rate of gully ad- vance (table 5) are probably in part due to unequal distribution of rainfall during 1951. The development of a locally continuous channel by the coalescence of two discontinuous valley—bot- tom gullies, as described by Leopold and Miller (1956, p. 31), is illustrated in figure 207. Only the upper reach of the gullied valley bottom is shown in the illustration. The shallow channel shown at left continues downvalley for 1,300 feet and termi- nates at the head scarp of gully B3. No evidence of downvalley extension of gullies during the period 1937—56 was observed at this locality, and in general downvalley extension of a gully is uncommon. The long profile of a single discontinuous gully is rarely smooth, but it is ordinarily broken by one or more inset scarps. 2O 4O 60 80 100 120 FEET I_|___;_I___l__l_._l VERTICAL AND HORIZONTAL SCALE OF CROSS PROFILES ALTITUDE, IN FEET ABOVE ARBITRARY DATUM as O l\) O so—\ l l l 0 0 1000 2000 3000 4000 5000 6000 DISTANCE, IN FEET FIGURE 206.—Long profile and cross profiles of a large valley-bottom gully in a tributary to Curtis Creek Canyon. Gully is designated B4 in, figure 201. Based on a. field survey, 1956. ’ 300 160 I l l | l I l EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT l l l l l I l l EXPLANATION Profile of valley in 1956 120 —— Profile of channel in 1956 — Probable profile of channel in 1937 (I) O 4: O ALTITUDE, lN FEET ABOVE ARBITRARY DATUM l l l l l l I i l l l l l l 1 L 2500 2000 1500 DISTANCE FROM VALLEY HEAD, IN FEET FIGURE 207.—Changes in long profile of a gullied tributary to Dry Creek, 1937—56. For the long profile in 1937, the position of head scarps was taken from aerial photographs, and other parts of the profile were estimated. Long and cross profiles in 1956 are based on a field survey. The gradient of a discontinuous gully is less steep than the gradient of the valley floor into which it is advancing, and Leopold and Miller attribute this to‘ the narrower Width of the gully. As gullies coalesce and Widen, a continuous channel is finally formed that has a gradient nearly parallel to the original valley floor. The gullies of the Medicine Creek basin, however, have not reached this final stage; channel scarps are advancing in succession along the valleys, and final regrading of a given reach will probably involve the passage of many channel scarps. HISTORY OF LATE RECENT GULLYING ON DRY CREEK Dry Creek was selected for detailed study of gully erosion by agencies involved in the Medicine Creek Watershed Investigations because the rate of erosion there seemed to be more severe than elsewhere in the basin. Periodic surveys were made of selected reaches of valley-bottom gullies to provide a basis for calculation of gross erosion by these gullies. A detailed geomorphic study of the Dry Creek channel and terraces was made by the writer, with the objective of deciphering the history of gullying in late Recent time and relating this history to White settlement and occupation. A detailed long profile of Dry Creek, based on a field survey made by personnel of the Bureau of Reclamation in 1951, is shown in figure 208. The terrace profiles in figure 208 and the cross profiles in figure 209 are also based on this survey, supple— mented with field surveys made by the writer. Station references used in the text and in figure 209 refer to distances in feet from the mouth of Dry Creek, are represented in figure 208. In general, the upper reaches of Dry Creek are ungullied (fig. 210, upper left), the middle reaches (as represented in fig. 210, lower left and upper right) have been twice gullied in late Recent time, and the lower reaches (as represented in fig. 210, lower right) were ungullied until about 1937. The inset terrace (figs. 209 and 210, lower left) provides evidence for the two episodes of gullying. The inset terrace deposits accumulated in the wake of a major channel scarp, and the inset terrace was formed as these deposits were trenched by the advance of a second scarp. The first channel scarp probably began at about station 10,000 for the inset terrace becomes indistinct at this point. A reasonable date for the beginning of this first scarp is about 350 years ago, which is the date suggested (on the basis of a carbon—14 age determination) for gullying on Elk— horn Canyon. The approximate age of the inset terrace deposits can be inferred from a cottonwood stump rooted on the terrace at station 43,300 (fig. 210, lower left) and from a tobacco tin buried to a depth of 5 feet in the deposits at station 43,700. The growth rings on the cottonwood stump indicate that it was about 40 years old at the time of cutting (1950) ; therefore, the inset terrace deposits had already accumulated in 1910. The tobacco tin, which is about 400 feet up- valley from the stump, was perhaps buried about this time or somewhat later. Because the inset terrace deposits accumulate in the wake of an ad- vancing head scarp, their age decreases in an upvalley LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 301 2900 I I I I I I EXPLANATION Stockville terrace 2800 — I l I I East Fork Dry Creek 2750 . / Valley flat . Inset terrace Flow line, 1951 channel 2700 - / , 2700 ’9/ f/, 11Tl'l'l'l'l' Bedrock outcrops in channel I I. I Position of channel scarp in 1937 VERTICAL SCALE IS EXAGGERATED 100 TIMES __ / 2600 ’ " ALTITUDE. lN FEET ABOVE MEAN SEA LEVEL 2500 / I I | l 2450 ' ‘ ' 1 o 10 20 30 40 50 60 7O 80 CHANNEL LENGTH, IN THOUSANDS OF FEET FIGURE 208.——Long profiles of channel, valley flat, and terraces on Dry Creek. direction. The channel scarp that is trenching the deposits and thus forming the inset terrace was at station 37,400 in 1937 and had advanced 8,300 feet by 1951. A new channel scarp began at or near the mouth of Dry Creek at a time not long before 1937, accord- ing to its position on the 1937 aerial photographs. This scarp had reached station 6,000 (fig. 210, lower right) by 1952, and it is moving rapidly upvalley at present. VALLEY-HEAD AND VALLEY-SIDE GULLIES Varieties of valley-head and valley-side gullies are distinguished mainly on the basis of shape of gully head in plan view. The main varieties in their characteristic topographic situations are illustrated in figure 211. A gully head that is on a relatively steep slope and receives a concentration of flow from a single direction, as from a narrow channel, tends to be narrow and pointed (gully 1, fig. 211). If the slope is less steep or the flow less concentrated, the gully head will be broadly lobed (2). If, on the other hand, the slope is gentle and the flow comes from diverse directions, the gully head will be com- plexly branching (3). Situations favorable for the initiation of a gully are related to valley shape (as determined by past erosional history of the valley), to the presence of cultivated upland at the valley head, and to the activities of man and livestock. The valley at right has steep sides and head because it was deeply trenched both before and after deposition of the Stockville terrace deposits. The steep valley head, which receives heavy runoff from a field planted in row crops, is a favorable place for the formation of a gully. A large valley-head gully (gully 3, fig. 211) , which began on the upper part of the valley head, has branched along the fence and into the field. The step scarps (4) on the steep valley sides began at animal trails along the slope contour and have grown in height as they slowly migrated upslope. The step scarp at extreme right has evolved into a valley-side gully (5). At the left side of the valley, two narrow valley-side gullies (1) are forming from cowpaths, and another valley-side gully (8) is advancing along a road ditch. The valley at center is shallow and has a relatively gently sloping head because it was trenched to shal- low depth before deposition of the Stockville terrace deposits, and post—Stockville trenching has not ex- tended far upvalley. In addition, the valley head borders a divide rather than a tract of upland and, therefore, has a small drainage area. The three gullies in the valley head (6) are lobed and have grown slowly. In the background beyond the center valley, two gullies (7) in another valley head are 302 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT STA 54.060 ___________ STA 39,000 I I I I l I I | I | I l I I I I 20 O 20 40 60 80 100 FEET |_ | I I I I I VERTICAL AND HORIZONTAL SCALE EXPLANATION Stockville terrace Valley flat Inset terrace Dashed lines represent valley profiles restored by projection of terrace slopes FIGURE 209.—Cross profiles of Dry Creek. Station numbers are distances in feet from mouth of creek. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA Upper left, Upper reach of East Fork Dry Creek, 1953. Narrow incised channel is shown in middle ground. Lower left, Channel and inset ter- race at station 43,300 in 1956. Cottonwood stump, near center, is rooted at outer edge of inset terrace. Upper right, at gaging station (sta. 17,660), August 17, 1953. Discharge about 300 cfs. Lower right, Channel at station 6,000 in 1953, looking downstream. 0n downstream side of cottonwood tree is a large channel scarp, advance of which has since felled the tree. 303 FIGUGE 210.~—DOWNSTREAM CHANGES IN CHANNEL OF DRY CREEK advancing rapidly—one in the direction of drainage from the road and the other in the direction of drain— age from the barnyard. The two valleys at extreme left are typical of many valleys in the upper part of the basin. Severe post-Stockville trenching has removed most of the deposits of Stockville age and left the valley sides raw and steep. The valley-side and valley-head gullies (2), which are lobed in plan View and have head scarps tens of feet in height, are advancing very slowly because their drainage area is small. Photographs of valley-head and valley-side gullies are shown in figure 212. The valley-head gullies in figure 212 (upper left and upper right) are among the largest in the basin. As is typical, these gullies border areas of cultivated upland. The gully in figure 212 (upper left) extends for about 900 feet along a fence line. A narrow but deep valley-head gully, whose head scarp attains a height of 31 feet, is shown in figure 212 (lower left). Valley-side gullies typical of the sharply dissected upper part of the basin are shown along the sides of the three valleys in figure 212 (upper right). Some are moderately active and others are inactive. The growth rate of such gullies is slow because they border narrow divides and hence have small drainage areas. The valley-side gullies in figure 212 (lower right) orig- inated from cowpaths and {mold field road. 304 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT .— . fie-s. ~ . " 4.x h ’ ;:\-, ”E 5/4“ ’K‘I‘E’fffl ’5‘" r". I» "A: ‘wu. law»: we murmur» Mvfiwéwmwéwtw ~.- -;.-- -. v -‘ . _. « z' , N. _ - - . a , I FIGURE 211.—Composite sketch, based on field sketches and photographs, showing difl'erent varieties of valley-head and valley-side gullies. Actively eroding seams are indicated by dark shading. Numbered features are described in text. MEASUREMENT OF VALLEY-HEAD AND VALLEY-SIDE GULLIES ON DRY CREEK In the summer of 1953 measurement of the volume increase of valley-head and valley-side gullies during the period 1937—52 was made by field surveys and comparison of aerial photographs. On aerial photo- graphs enlarged to a scale of 1 inch to 660 feet, reference lines across or near gully heads were drawn between fixed points, such as trees, fence corners, houses, or other landmarks. The position of gully head in 1937 was marked on a 1952 aerial photograph, which was then taken into the field to facilitate measurement of the volume increase. Changes in gully depth could not be determined from aerial photographs; but this is probably not a sig- nificant source of error, inasmuch as neither height of head scarp nor gully depth changes very much during a moderate advance. The horizontal dimen- sions of gullies were measured by pacing, and the depths were measured with a steel tape. Determina- tion of volume was complicated by the irregular shape of most gullies. No volume increases less than about 30 cubic yards are reported because increases less than this amount are considered to be unrecog- nizable by comparison of aerial photographs. Vol- ume increases, however, were measured for nearly all the gullies 0n Dry Creek that seem to be active, and erosion from gullies that seem to be inactive is probably not quantitatively important. According to these measurements, the enlargement of all active valley-head and valley-side gullies on Dry Creek (216 gullies) for the period 1937—52 totaled 106,500 cubic yards. A plot of the frequency distribution of gully enlargements on logarithmic probability paper indicates a logarithmically normal distribution, and the skewness at the lower end of the curve is attributed to the omission of perhaps 10 gully enlargements less than 30 cubic yards. (See fig. 213.) The median enlargement is about LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 305 FIGURE 212.——Valley-head and valley-side gullies. Upper left, Center, large valley-head gully in Curtis Creek subbasin, NEV; sec. 10, T. 8 N., R. 28 W. Oblique aerial view, September 4, 1965. Lower left, Valley-head gully in Dry Creek subbasin, center sec. 33, T. 9 N., R. 27 W., September, 1953. Upper right, Large valley-head gullies and smaller valley-side gullies in Curtis Creek subbasin, SW 1/1 sec. 30, T. 10 N., R. 27 W. Oblique aerial view, September 4, 1965. Lower Tight, Valley-side gullies in Well Canyon subbasin, SW 14 sec. 13, T. 9 N., R. 29 W. 200 cubic yards, and the range is from 28 to about 4,200 cubic yards. A satisfactory approximation to the measured gully enlargement was reached by another method, which is based on the premise that active gullies can be recognized on aerial photographs and also on the premise that small gullies have small volume increases and large gullies have large volume in— creases. Several years after the gully measure- ments were made on Dry Creek, active gullies in the Medicine Creek basin were studied on aerial photo- graphs and ranked into four categories according to size (not according to enlargement). On the basis of the measured gully enlargements on Dry Creek, each size category was assigned an enlarge- ment range, according to the following scheme: Geometric mean of Gully size Enlargement range, enlargement range 1937—52 (cu yd) (cu yd) Small ....................... 50—100 63 Medium .................. 100—600 245 Large ____________________ 600—3,600 1,470 Very large ............ 3,600—7,200 5,100 The number of gullies in each category on Dry Creek is shown in table 6. By multiplying the num— ber of gullies in each category by the geometric mean of the enlargement, a total enlargement of 97,300 cubic yards was obtained, which is a reason- able approximation to the measured enlargement. Although this method is obviously subjective, no means of applying rigorously objective sampling methods to such irregular objects as gullies is 306 PERCENTAGE ENLARGEMENT PERCENTAGE ENLARGEMENT EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 0.1 0.2 0.5 10 20— 30* 40— 50 60 '- 70 -— 80- 90 95- 98 — 99 1- 99.5 99.8 — 99.9 I I All 10 0.1 100 1000 GULLY ENLARGEMENT 1937-52, IN CUBIC YARDS 0.2 —- 10 20— 40— 50 60 - 7O - 90 / 99 — 99.5 I 99.8 99.9 l l l l l III B 11111111 10 100 1000 10,000 GULLY ENLARGEMENT 1937-52, IN CUBIC YARDS FIGURE 213.—Cumulative frequency distribution of 216 measured enlargements of valley-head and valley-side gullies on Dry Creek. A, Curve of 216 measured enlargements. B. Curve of 216 measur enlargements of less than 28 cubic yards. ed enlargements plus 7 hypothetical LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 307 TABLE 6.—Number of gullies according to type and size in the Medicine Creek basin and in major subbasins Number of gullies Width of gully head Medicine Gully type Size (range in Mitchell Dry Fox Brushy (subhasins feet) (subbasin C) (subbasin I) (subbasin K) (subbasin N) A—O) Valley head ____________ Small ____________________ 1—15 47 70 39 36 417 Medium ________________ 16—60 99 110 90 302 1,357 Large _-_____--_-____.. 61—180 28 35 16 90 360 Very large ___________ >180 1 2 ________________________ 3 20 Valley side ______________ Small ____________________ 1—10 22 3 7 10 112 Medium 11—25 21 14 17 66 279 Large __________________ 26—50 6 2 2 7 44 Very large ____________ >50 ,,,,,,,,,,, Valley bottom ,,,,,,,,, Small ____________________ 1—10 56 38 11 24 563 Medium ._____. 11—25 33 27 23 82 477 Large ____________________ 26—50 Very large ____________ >50 } 10 10 9 19 152 apparent. In the ranking of gullies into categories, a particularly difficult subjective decision lies in deciding whether the complex branches of a severely eroded valley head should be regarded as one gully or several. In the Dry Creek measurements, most of the branching valley heads were divided into several gullies, whereas in the later ranking of gullies according to size, the branching valley heads were regarded as one gully. For this reason the geometric mean of the largest size category (5,100 cu yd) is larger than the largest measurement re- ported (4,200 cu yd). The topographic situations of gullies, as well as most of the topographic features characteristic of the Medicine Creek basin, are illustrated by the vertical aerial photograph reproduced as figure 214. Sec. 17, T. 9 N., R. 27 W., is approximately in the center of the area, and the valley of Dry Creek crosses it from left to right. Gullies numbered 3 and 5 on the photograph are in the very large category and have complexly branching heads. Total enlargement of all branches amounted to about 6,950 cubic yards for gully 3 during the period 1937—52 and about 5,610 cubic yards for gully 5. Figure 204 (upper right) is a View of gully 3 as seen from the air. Gullies 1 and 4 are in the large category, and the lobed shape of their heads indicates that they have not advanced beyond the steeper parts of the valley heads, although the head of gully 4 is beginning to branch. Enlargement during the period 1937-52 was 1,320 cubic yards for gully 1 and 2,250 cubic yards for gully 4. The group of four gullies indi- cated by the numeral 2 are in the medium category, and their enlargements ranged from 107 to 225 cubic yards. Discontinuous valley-bottom gullies of small size are indicated by the two arrows at left of the nu- meral 4, and others of medium size are indicated by arrows in the lower left corner of the photograph. Small discontinuous valley-bottom gullies are also indicated by arrows on the valley flat of Dry Creek. The major valley—bottom gully on Dry Creek begins about 1 mile downstream (to the right) from this locality. A valley—side gully of medium size, which is advancing along a fence, is indicated by the numeral 6. Among the notable relief features shown on the photograph are the differences among valleys in flatness of bottom and steepness of side. The valleys at right of numeral 4 have flat bottoms and steep sides because post—Stockville trenching extended to the valley heads; the valleys at left and below numeral 4 have narrow bottoms and gentle sides because their trenching was incomplete. Also nota- ble, in the field at right center on the photograph, is a dark round spot that indicates a depression. Depressions of this sort, sometimes called buffalo wallows, are common on the Great Plains. Many hypotheses have been proposed as to their origin, but they probably originate in more than one way, and no hypothesis has been established ‘as generally valid. In the Medicine Creek basin, many of the depressions probably mark the positions of sink- holes in the underlying Ogallala Formation. AREAL DISTRIBUTION OF GULLIES Experience in the recognition of Dry Creek gullies on aerial photographs and in the estimation of their size and activity was used in the collection of infor- mation on the distribution of gullies in the Medicine 308 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT FIGURE 214.—Vertical aerial photograph of a severe] small- and medium-size discontinuous valley y gullied area on upper Dry Creek. Numbers indicate gullies described in text. Arrows indicate -bottom gullia described in text. Photograph by US. Commodity Stabilization Service. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA Creek basin as a whole. Aerial photographs of the Whole basin were studied stereoscopically, and all gullies in the basin that seemed to be active were plotted on topographic maps. Decisions as to the activity of a gully were checked by reference to 1937 aerial photographs. The gullies were grouped according to type (valley head, valley side, or valley bottom) and according to size (small, medium, large, very large). (See table 6.) A map showing the frequency distribution of gullies according to subbasin was prepared (fig. 215). The basin was divided into 24 subbasins, corresponding generally to subbasins used by the Agricultural Research Service in their land-use tab- ulations, that are reasonably homogeneous in texture of relief. The frequencies of valley-head, valley-side, and valley-bottom gullies (excepting large and very large valley-bottom gullies) in each subbasin are given in table 7. Gully frequency refers to number of gullies per square mile of valley system rather than to number per square mile of drainage area. Gullies are directly related to the valley system and are not present in the upland except at its edges. Area of upland and characteristics of the valley system should be regarded as independent variables in their relation to gullying. The frequency of active valley-head and valley- side gullies is low in the lower and upper part of the basin and ranges from moderate to high in the TABLE 7.—Gully frequencies and related data for subbasins of the Medicine Creek basin Fre- Fre- F're- fluency quency Drainage Area in Area of uuency of valley- of Sub— area upland vallev of first— head and valley- basin (sq mi) (percent) system order vallev-side bottom (sq mi) channels gullies gullies 27.2 45 15.0 113 3.2 5.3 43.4 38 26.4 150 2.9 4.2 11.6 37 7.3 140 3.7 5.5 22.3 50 10.8 190 16.0 4.3 29.8 52 14.4 150 3.5 3.4 74.4 45 40.9 160 5.7 2.6 10.9 42 6.3 165 5.2 6.1 58.9 34 35.3 170 6.9 4.1 23.2 42 13.4 195 12.4 5 0 16.5 48 7.4 185 11.7 4.2 12.6 40 9.5 214 18.5 5.1 8.5 49 4.3 185 14.2 5.3 17.6 12 15.5 275 3.3 3.0 22.6 45 12.6 200 16.4 4.6 57.0 12 49.8 260 1.5 4.2 15.5 30 10.8 200 9.1 1.6 22.7 12 19.9 260 .5 .2 30.6 26 22.5 220 5.2 .9 27.5 30 19.3 180 6.8 1.6 73.8 26 52.4 260 10.0 2 3 14.6 15.0 10 13.5 450 .7 .4 35.4 12 31.1 350 2.1 .4 22.0 55 9.9 350 10.8 1 8 309 central part (fig. 215). The frequency ratio of valley-head and valley-side gullies to valley-bottom gullies ranges from about 0.6 in the lower part of the basin (subbasin A) to about 6 in the central part of the basin (subbasin 0—4). The distribution of gullies according to type and frequency is con- trolled mainly by the steepness of valley heads, the narrowness of valley bottoms, and the amount of upland drained by a valley head. In general, the steepness of valley heads is directly proportional, and the narrowness of valley bottoms is inversely proportional, to first—order-channel frequency. Gully frequency in relation to percentage of area in upland and to adjusted frequency of first-order channels is shown in a semiquantitative way in figure 216. Moderate to high frequencies of valley- head and valley-side gullies are associated with values of first—order-channel frequency and percent- age of area in upland that plot to the right of the dashed line. On the other hand, values that plot to the left of the dashed line tend to have a higher frequency of small and medium valley-bottom gullies. Points representing high gully frequency would probably be more widely separated from points rep— resenting low gully frequency were it not for the fact that the subbasins are not entirely homogeneous in topographic character. The severity of post- Stockville trenching not only varies from the upper to the lower part of the basin, but also from one minor side tributary to the next (fig. 214) . The point indicated by a gully frequency of 10.8 and repre- senting subbasin 0.4, at upper right in figure 216, is most anomalous. In spite of the relatively low gully frequency in this subbasin, it includes an area of exceptionally severe erosion that is about 3 miles northeast of Maywood. This area is also exceptional in that its drainage was greatly extended by post- Stockville erosion. Several large gullies there made advances ranging from 60 to 180 feet during the period 1937—52. FORMATION OF VALLEY—HEAD AND VALLEY-SIDE GULLIES The head scarps of many valley—head gullies orig- inate on the steep upper part of the valley head and advance toward the upland. This is indicated by the position of short gullies and by the shape in plan View of long gullies. The width of many large valley- head gullies decreases sharply toward the lower part of the valley head, and the gullies are connected to the valley flat by a narrow trench. (See fig. 217 and gully 5 in fig. 214.) Gully 5 narrowed down- valley (in 1956) to a particularly deep trench about 25 feet deep, 15 feet wide at the top, and 5 feet wide 310 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT EXPLANATION Frequency of gullies per square mile of valley system 8-2 Subbasin See table 7 5 O 5 MILES L—I—_J_;L_J‘fi_l FIGURE 215.—Areal frequency distribution of active valley-head and valley-side gullies in the Medicine Creek basin. Subbasins illustrated in figure 197 are shown by heavy outlines. See table 7. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 311 60 l l l l \ 10.8 3.5 30.4 0 \ 016 50 - 14.2 — 8 2 5 7 \ 11.7 16 4 34's 3. . S . o o l A 5.2 12.4 0 \ O 18.5 40 O 02.9 ._ 3.7 2 O in o a: 6.9 N O u. GULLY FREQUENCY z MODERATE TO HIGH ‘5 6.8 9.1 2 30 — O Q _ < _l a. E GULLY FREQUENCY 5.2 10 0 LOW TO MODERATE O E a: \ 23.2 < A 20 \ A“ \ 13 3 \ \ EXPLANATION \ \ .\ 0.5 3.3 2.1 \ 2.1 O O o 7 10 — Subbasin 15 12 O ' _ 1.2 A Small basins less than 2 square miles in area Number beside symbol refers to gully frequency A2 per square mile of valley system 0 l l l i 100 150 200 250 300 350 400 450 500 ADJUSTED FREQUENCY OF FIRST-ORDER CHANNELS FIGURE 216.—-Frequency of active valley-head and valley-side gullies in relation to area of upland and adjusted frequency of first-order channels. at the bottom. Set in the bottom of this trench was a Winding notch, sections of which passed under— ground. Other valley-head gullies begin at an upvalley riser of the Stockville terrace, such as the one shown in the valley at center in figure 211, or as discon- tinuous valley-bottom gullies, such as the ones shown in the valley at upper left in figure 200. In the tabu— lation of all active gullies in the basin, valley-head gullies that are connected by a continuous, rather Wide trench to the valley flat were given a separate designation because they probably began on the valley bottom rather than in the valley head. The percentages of valley-head gullies that fall in this category are as follows for several subbasins: Dry Creek, .7 percent; Brushy Creek, 22 percent; Mitchell Creek, 9 percent; Fox Creek, 6 percent; Lime Creek, 45 percent; and subbasin A, 65 percent. Thus, in the narrow valleys of the lower part of the basin, a greater percentage of valley-head gullies begins on the valley bottoms. The head scarps of valley-head gullies reach a maximum height of about 35 feet, a value about 10 feet higher than the maximum height of head scarp measured for large valley-bottom gullies. This greater height is evidently related to the steeper slopes into which the valley-head and valley-side gullies are advancing and to the correspondingly steeper slopes at the base of their head scarps. (See figs. 217 and 218.) Material does not tend to accu- mulate at the base of the head scarp as it does in large valley-bottom gullies. In long profile, the head 312 190 180 170 I Ypproximate profik 160 " mockville valley fill \‘1 . ,1956 profile of slope \‘ \< adjacent to gully 150 \ \\ \\ 140 1 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 40 FEET Q 1956 profile ofW\\ 130 L\ 120 m Plan view ALTITUDE, IN FEET ABOVE ARBITRARY DATUM 110 a" \7\\ Long profile 100 0 100 200 300 400 500 600 700 800 900 DISTANCE ALONG GULLY, IN FEET WWW/wiry \G/V FIGURE 217.—-—Long profile, plan View, and cross profiles of a large valley-head gully, designated gully 4 in figure 214. 20 O 20 40 60 I_L1ll A j_ I 80 FEET I HORIZONTAL AND VERTICAL SCALE Cross profiles Shaded area on plan view indicates enlargement during the period 1954-57. Based of field surveys. Location of cross profile H is beyond limit of area shown in plan view. scarps are nearly vertical or even overhanging at the rim, and the channel profile downstream from the head scarp is typically broken by several channel scarps. The channel scarps deepen the gully as they advance and probably increase the height of the head scarp by merging with it. The modal value for height of head scarps of active valley-head and valley-side gullies 0n Dry Creek is about 10 feet. The mechanism of head—scarp advance is not identical with that described for large valley-bottom gullies. Most valley-head gullies are advancing into surfaces underlain by a well-developed soil profile and, except for gullies advancing into cultivated upland, protected by a sod cover. The rills or trenches that lead to the gully head have not been incised through this profile; hence, the head scarp has a resistant rim. On the other hand, plunge pools are not conspicuous at the base of the head scarps, and saturation of the massive silt by plunge pool action is less important than that for large valley-bottom gullies. The water flowing over the rim of the head scarp has relatively small volume and velocity, and silt beneath the resistant rim is disintegrated by back trickle of water. Underground drainage, which probably enters the ground through rodent burrows upslope from the head scarp, is con- veyed into the gully head at some point below the resistant rim. The abundance of rodent burrows intersected by some head scarps, as well as the over- hanging of the resistant rim, is shown in figure 219. Head scarps less than 4 feet high do not advance rapidly because the resistance of the soil profile and the sod cover extend to about this depth. Material that slumps from the head scarps and side scarps accumulates on the gully floor and is gradually removed by runoff. Deposition is uncom- mon in the trench that drains the gully head, but much deposition takes place where this trench enters the valley flat. The number of valley-head and valley-side gullies that discharge directly into the channels of major tributaries is insignificant, and LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 313 100 to O /-_ —« 0 -- | J I I ._-_l_-4lh=-—’—-h_- ___ -__ ___.__ "4:. 1‘ 0 2 4 6 8 10 12 14 16 18 20 22 24 TIME, IN HOURS FIGURE 220.—Typical hydrographs of large flows. bankfull discharge, these variables are simple power The product of mean velocity, width, and mean functions of the water discharge of the form depth must equal discharge; therefore, the sum of 22an the exponents b, f, and m must equal 1.00. (1an Points on graphs of width, depth, and velocity I 711046)” versus discharge for the gaging stations in the Medi- where 22 is mean velocity (fps), cine Creek basin scatter rather Widely from the d is mean depth (ft), mean, particularly for Brushy, Dry (fig. 221), and w is width (ft), and, Mitchell Creeks. Scatter at the lower values of Q is water discharge (cfs). discharge is caused, in part, by natural variations WIDTH, IN FEET DEPTH, IN FEET VELOCITY, IN FEET PER SECOND LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 317 60 10 III III ILIIIII I IIIIIII lllIIlll 1 10 100 DISCHARGE, IN CUBIC FEET PER SECOND FIGURE 221.—Change of width, depth, and velocity with increasing discharge at a section, Dry Creek near Curtis. 600 318 in channel shape at the different sections used for making discharge measurements by wading. Flows too deep to be waded were measured at a single section at each of the stations, and the scatter on the plots is somewhat reduced at the higher dis- charges. There would be some scatter on the plots even if all measurements were made at a single cross section. The slopes of curves relating width, depth, and velocity to discharge at each gaging station were fitted by eye and adjusted so that the values of b, f, and m summed to 1.00. The adjustments were small for all sections. The relations of width, depth, and velocity to water discharge at the individual sections have little meaning; therefore, the curves, without the defining points, were placed on one graph, and the discharges that were equaled or ex- ceeded 1, 2, and 25 percent of the time were indi— cated on each of the curves. Discharges equaled or exceeded 1, 2, and 25 percent of the time were selected because the 1-percent discharge represents approximately bankfull flow at the Fox Creek station but less than bankfull flow for the other stations. The 25—percent discharge is approximately the modal discharge for the Fox Creek and both Medicine Creek stations, and the 2-percent discharge is about the lower limit for which reliable measurements of Width, depth, and velocity are available for the Dry, Brushy, and Mitchell Creek stations. The slopes of width-discharge relations for Dry and Brushy Creeks are greater than those for the other sections, and lines joining points of equal- discharge frequency indicate that the sections should be divided into two groups (figs. 222 and 223). The fact that Mitchell Creek does not fit with either group is attributed to the presence of a concrete control at the section and to a low-water road cross- ing a short distance downstream, which causes back- water at the section. Therefore, the Mitchell Creek section will not be considered further. Lines that join points of equal frequency of dis- charge on the curves of the at-a-station relations represent the downstream relations of width, depth, and velocity to discharge, if the assumption is made that the relation at sections not on the same stream is representative of relations at different sections on the same stream. This assumption may be justi- fied for two reasons. First, the discharges of equal frequency for the two Medicine Creek stations and the Fox Creek station plot about on a straight line even though the uppermost section on Medicine Creek, at Maywood, is somewhat affected by reser- voir regulation. Second, the gaging sections are EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT on reaches where flows are over loess, which covers much of the basin. The assumption implies that the channels of two streams are similar at the points along their lengths where the discharges are the same at equal frequencies of occurrence. The values of b, f, and m for changes of Width, depth, and velocity in a downstream direction ob- tained from figures 222 and 223 cannot be considered reliable because too few stations are available to establish averages. The values can be used to show general differences between the types of streams represented by the two groups. The downstream relations of width, depth, and velocity to discharge could best be defined by meas— urements along a stream at several stations for which at-a—station relations of width, depth, and velocity to water discharge and flow-duration curves are available. However, gaging stations of the number required for such a study are found on few if any streams. When better data are lacking, discharge measurements might be made at several sections in a downstream direction at discharges of unknown frequency at each of the sections. Even though the frequency of the discharges were the same at all the sections measured, the plot of width depth, and velocity against discharge would scatter from the average line because of variations in shape that exist from place to place in a natural channel. Of course, it would not be possible to obtain measurements of discharges of the same frequency at all sections, especially for channels of ephemeral streams. Assume that a series of measurements are to be made in a downstream direction on Dry Creek at dis— charges that are equaled or exceeded about 2 percent of the time at each of the sections. Further assume that the downstream relation of width to discharge for Dry Creek is the same as that shown on figure 223 for discharges equaled or exceeded 2 percent of the time. If a measurement is made at the gaging station at a discharge assumed to be equaled or exceeded 2 percent of the time, but the discharge is actually the one equaled or exceeded 1 percent of the time, then the Width will be greater than that indicated by the downstream relation at discharges equaled or exceeded 2 percent of the time by about 40 percent plus or minus the deviation from the at-a-station relation (fig. 221). About two-thirds of the points that defined the at-a-station relation for the Dry Creek station were within plus 35 and minus 35 percent of the mean line. As the difference between the slopes of the lines for the at-a-station relations and the downstream relations of width, depth, and velocity to discharge becomes less, the scatter from the average line attributable to differ- LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 1°°_IIH llllll T lllllll — —< ._ _ _ LIJ L|J L E :5 +— 9 z 10 5 111 5 III '— UJ LLJ LL — _ E 35 p m U D l 0.5 1|] 5 Ill ’— LIJ Eg— — 20 78 Ew— — 035 80. LIJ > 1111i 1 Illlll I lllllll 5 10 100 1000 DISCHARGE, IN CUBIC FEET PER SECOND X Medicine Creek above Harry Strunk Lake A Medicine Creek at Maywood 0 Fox Creek at Curtis EXPLANATION Percentage of time daily flow equaled or exceeded: 25 FIGURE ZZZ—Change, of width, depth, and velocity in a downstream direction for channels of perennial streams. 319 320 WIDTH, IN FEET DEPTH, IN FEET VELOCITY, IN FEET PER SECOND EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 30 H O —— 20 10 Dry Creek near Curtis Brushy Creek near Maywood I I J l I I I J I I I I I I I I I I // .— I //!/ J I I l l I I I 0 100 DISCHARGE, IN CUBIC FEET PER SECOND EXPLANATION Percentage of time daily flow equaled or exceeded: 2 FIGURE 223.—Change of width, depth, and velocity in a. downstream direction for channels of ephemeral streams. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA ences in frequency of discharge at the various sta- tions becomes less. Establishing fairly reliable downstream relations of width, depth, and velocity to discharge is possible if a sufficient number of sec- tions are selected to provide a reliable average and if measurements of extremely high or low discharges are avoided. The slopes of the relations of width, depth, and velocity to discharge (table 8) indicate differences TABLE 8.——Values of b, f, and m for streams in Medicine Creek basin Perennial streams Ephemeral streams At a station 1 Downstream 2 At a station 1 Downstream 2 0.24 .56 .20 0.69 .12 .19 0.03 .48 .45 1 Unweighted average. 2 At discharge equaled or exceeded 1 percent of time. between the perennial and ephemeral streams in the Medicine Creek basin. The at-a-station slope values may be considered reliable, but slopes for the down- stream relations probably only indicate the correct order of magnitude. The differences between the at-a-station relations for the perennial and ephemeral streams reflect the differences in channel shapes. The channels of the perennial streams have fairly steep banks, and the width does not increase as rapidly as the depth. The channels of the ephemeral streams also have steep banks; but the bottoms of the channels are somewhat parabolic, and the width can increase rapidly as discharge increases until the limits of the steep confining banks is reached. How- ever, a relatively large discharge is required for the width of water surface to reach the confining limits of the steep banks in the broadly incised channels of ephemeral streams. The changes of velocity with increasing discharge are about the same for peren- nial and ephemeral streams; therefore, the change of width is less and the change of depth is greater with increasing discharge for perennial streams than for the broadly incised ephemeral streams. Depths of water in a channel normally do not increase very rapidly in a downstream direction. That is, it would not be expected that the depth at a cross section on a stream would be much greater than the depth at another section several tens of miles upstream. The downstream change in depth for perennial streams in Medicine Creek basin is relatively small, and the downstream change in velocity is about the same as the change in velocity 321 at a station. Because the discharge at a constant frequency increases downstream, the width must increase rapidly. However, the width for the broadly incised ephemeral streams increases very little down- stream. The increase in discharge is accommodated by relatively large changes in both depth and ve- locity in a downstream direction. The downstream change in width of the broadly incised ephemeral streams may be greater than that indicated on figure 223, but the change still would not be so large as the change of the perennial streams. Figure 204 (upper left) shows the main-stem channel of Dry Creek just below the gully scarp that forms the end of the broadly incised channel, and figure 210 (upper right) shows the main-stem channel at the gaging section. The top widths of the channel just below the gully scarp and in the vicinity of the gaging section are 40 to 50 feet. The distance is too short and the measurements are too rough to indicate the change in width along the Dry Creek channel; but if the width increased as rapidly as the indicated change of about 5 percent per mile on Medicine Creek between Maywood and the gage above Harry Strunk Lake, the width at the gage on Dry Creek should be more than 10 feet wider than the width in the vicinity of the head scarp. SUSPENDED-SEDIMENT DISCHARGE RELATIONS The curves relating suspended-sediment load to water discharge for the perennial streams (fig. 224) have breaks in slope at water discharges that are higher than normal but below bankfull stage. The curves applying to ephemeral streams show no cor- responding breaks in slope (fig. 225), and in value of slope they are generally similar to the upper part of the curves applying to perennial streams. The breaks in the suspended-sediment rating curves re- sult from a rather complex and uncertain set of conditions. The concentrations, peak and mean, for runoff periods are highly variable. Even two events, which may be very similar with respect to peak discharge and total discharge, may have very differ- peak suspended-sediment concentrations and mean concentrations. The peak concentration generally occurs before the peak water discharge for stations in the Medicine Creek basin, but the length of lead is variable. The length of lead may be from 1 to 3 or or more for the perennial streams and is generally an hour or less for the ephemeral streams. The shorter lead time, in general, occurs with a small peak discharge. The variability in concentration of suspended sediment and in length of lead results because the suspended-sediment load is made up principally of fine material derived from sources 322 other than the streambed. The amount of fine ma- terial that is transported by the streams during a given storm depends on intensity and duration of rainfall, soil conditions at the time of the storm, and other factors. The leading concentration indi- cates that the sediment rating curve should describe a loop, but because of the variability of concentration from storm to storm and the variability of the lead time of concentration, the loop would not be defined by an average of many measurements for many storms. Instead, the points defined by the measure- ments would scatter with increasing discharge. (See fig. 224.) The measurements that define the points of figures 224 and 225 are somewhat biased, however, because only a few measurements were obtained before peak water discharge and practically none were obtained near peak sediment concentration. The generally rapid rises and mud roads prevented arrival of workers at the stations much before the occurrence of the peak water discharge, particularly on the ephemeral streams. Nearly all the measure- ments at those stations were obtained after the peak water discharge; therefore, the scatter of points is fairly uniform throughout the range of water dis- charge. The lower part of the curves for' the perennial streams is represented mostly by measurements made during small rises. A small increase in water dis- charge is accompanied by a relatively large increase in sediment concentration, and the lead time for smaller rises is generally short. As a result, the slope of the relation of water discharge to sediment discharge increases more rapidly for low water dis- charges than for the high water discharges. The fact that the slope of the water-sediment dis- charge relation is slightly less than unity for the ephemeral streams indicates that the concentration remains about the same or decreases slightly in a downstream direction, but the fact that the slope of the water—sediment discharge relation is slightly greater than unity for perennial streams indicates that the concentration increases in a downstream direction. Because so few records are available to establish good average downstream relations, it is not possible to state that an actual difference exists between the slopes of the water-sediment discharge relations in a downstream direction for perennial and ephemeral streams. PARTICLE-SIZE DISTRIBUTIONS OF SUSPENDED SEDIMENT AND BED MATERIAL Average particle-size distributions of suspended sediment were obtained from unweighted averages of all size analyses at each of the stations. The size EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT analyses were not weighted because a plot of water discharge against percentage of suspended sediment finer than 0.004 mm and coarser than 0.062 mm indicated no relation between instantaneous water discharge and particle-size distribution. The average size distributions for each year for each of the stations did show an inverse relation to annual water discharge; that is, the percentage finer than a given size decreased as water discharge increased. The relation was not very well defined, and the indicated change was not large; but the relation of annual water discharge to the average percentage finer than a given size does indicate the possibility of a relation between the percentage finer than a given size and instantaneous water discharge. The large vari- ability of particle-size distributions for a given water discharge, however, prevent detection of a relation. Average particle-size distributions of suspended sediment and bed material (figs. 226 and 227) indi- cate little difference between size distributions of suspended sediment and a large difference between size distributions of bed material for Medicine Creek at Maywood and Medicine Creek above Harry Strunk Lake. Medicine Creek above Harry Strunk Lake has a small percentage of material in the 0.125- to 0.250-mm range in suspension; such material is normally not in suspension at Maywood. Although the size distributions of suspended sediment are similar at the Medicine Creek stations, the distribu- tions of bed material are very different. The bed material at Maywood has 36 percent of the material finer than 0.062 mm, 44 percent between 0.062 and 2.0 mm, and 20 percent coarser than 2.0 mm; at the station above Harry Strunk Lake only 8 percent is finer than 0.062 mm, 80 percent is between 0.062 and 2.0 mm, and only 12 percent is coarser than 2 mm. The increase of material in the 0.062- to 2.0- mm range is probably from the tributaries draining the area to the southwest of Medicine Creek. There are, however, no analyses of bed-material samples from any of these tributaries. The average distri— bution of only three samples, each of bed material from the low-flow channels of Brushy and Fox Creeks (fig. 50), indicates that the bed material at those stations is similar to bed material at Medicine Creek at Maywood, especially for the sizes finer than about 1.0 mm. Suspended-sediment size distributions for Brushy, Fox, Dry, and Mitchell Creeks (fig. 226) show that there is an apparent increase in fine material available for transport as suspended sedi- ment from the west to the east side of the basin. Brushy Creek, on the west side of the basin, has a median suspended-sediment particle size of about 0.014 mm, whereas Mitchell Creek, on the east side SUSPENDED—SEDIMENT DISCHARGE. IN TONS PER DAY LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA ,0 10° 0° _ I I | I I I I I I I I x I _ a 10,000 1000 EXPLANATION X——’——x 100 0 Medicine Creek above Harry Strunk Lake __ A— -— —5 fl — Medicine Creek at Maywood _' 0- ————— o ————— .0 _ Fox Creek at Curtis 10 I I I I I I I I I l I I 10 1000 10,000 WATER DISCHARGE, IN CUBIC FEET PER SECOND FIGURE 224.—Change of suspended-sediment load with increasing discharge at a section, perennial streams. 323 324 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 100,000_ I TIIIIII I Illllll I J'l|f|!_ _ . _ _ . O o 0 _ _ o O 0 j 0 0 °_ 0 10,000 )— < D _ n: Lu o. ._ u) z O +— _ E tn" (3 cc < 5 2 1000 a _ ._ _ Z —4 Lu E _ 0 Lu _ “P D _ Lu 0 2 Lu _. a. (I) D m 100 EXPLANATION — o———O————o _ Brushy Creek near Maywood _ (— I——-I _ Dry Creek near Curtis + ————— -\+ ————— J + Mitchell Creek above Harry Strunk Lake 10L l | l l l l I III I l l | l l I | 1 10 100 1000 WATER DISCHARGE, IN CUBIC FEET PER SECOND FIGURE 225.——Change of suspended-sediment load with increasing discharge at a section, ephemeral streams. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 325 50 -/ ‘5 /: ‘ /. I :7 Fox Cray/#9,, , Lr’z ’ , 99-9 lllllll l IIITIII l I 99.8 '- —. Lu 99.5 — — E 0) 99.0 " 4 S r/ z . . !— 98>—- ’ Medlcme Creekabove __ 5 Mitchell Creek>/ Harry Strunk Lake 5 E ’1 '9 Z 95 ,',A/. E / ’/// I— 90_ ,f/ / _ n: ' / in Z I 80 ‘— w _ m d _ 70—- _ a / n. 60— / ._ LL 0 m (5 < P z in o n: m n. 40 - - 30 Medicine Creek at Maywood 20 — _ 10 l l l l l I l l l l l J I l l l I 0.002 0.01 0.1 DIAMETER, IN MILLIMETERS FIGURE 226.—Size distribution of suspended sediment at gaging stations in Medicine Creek basin. of the basin, has a median particle size of about 0.004 mm. The decrease in median particle size from west to east reflects the increasing distance from the sand-dune area of the extreme western and northwestern parts of the basin. The measured loads for all the stations are nearly the total loads because the suspended sediment con- tains little sand and the fine suspended sediment should have fairly uniform distribution in the ver- tical. Computations of total load were made by the Colby method (Colby, 1957) for the two stations on Medicine Creek, both of which have sand beds. Not many sets of data are available for computation of total loads, but the average percentage of meas- ured load to total load is indicated. At low flow the measured load amounts to about 90 percent of the total load, but at high flow it amounts to more than 95 percent. Because all the suspended sediment at the Dry Creek gage results from storm runoff and a large part of the suspended sediment at the Brushy and Fox Creek gages results from storm runoff, the measured suspended sediment for those stations probably represents 95 percent or more of the total load for the period of record. RELATIVE SEDIMENT CONTRIBUTIONS OF THE MAJOR TRIBUTARIES In any basin or subbasin, erosion and deposition are taking place simultaneously, and sediment data collected at any point along a stream represent the net of erosion and deposition upstream from that point. A measure of the net erosion and deposition in a basin is the discharge-weighted mean concentra- tion of suspended sediment. This is the concentration that would result if all the water passing a point during some period of time were mixed with the suspended sediment passing during the same period. If the cumulative water volume is plotted against cumulative weight of sediment, the slope of a line that joins points of the graph defines the discharge- weighted mean concentration for the period of time for which the data are cumulated. The annual sedi- ment loads and storm runoff for the station were reduced to a per-square-mile basis and are shown 326 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 99" Till I I I I 99.8-— 99.5 - 99.0 I— u: 09 I 95 _ _4 II I I IIIIIF 90*— 70-— ’-"‘ ,Af’ k-H -f/ 60'— 50 Brushy Creek: 0-— 1.7%, / /, 40— / 20— PERCENTAGE OF PARTICLES FINER THAN INDICATED SIZE Medicine Creek above Harry Strunk Lake — 10 I / llll L I II ILII I I IIIIII 1.0 10 DIAMETER, IN MILLIMETERS FIGURE 227.-—Size distribution of bed material at gaging stations in Medicine Creek basin. on figure 228. Reducing the values to a per-square— mile basis does not change the constant of propor- tionality and presents a comparison of storm runoif and sediment yields. Such a comparison is valid if precipitation over the entire basin is assumed to be uniform in amount. Because of the relatively small size of the basin and the low relief, the assumption of uniform precipitation probably is valid if several years record are considered. The differences in total precipitation for the years 1951—58 recorded at the Wellfleet, Curtis, and Stockville rain gages were less than 10 percent. As the time period is shortened, however, the probability of uniform precipitation becomes less. Each of the stations shows a definite break in slope following the 1951 water year. The break in the curve probably occurs because the ratio between the variables is not constant at all rates of cumula- tion. If the ratio were constant at all rates of cumulation, the break would have to be explained by a physical change that would cause a greater reduction in sediment yield than in water yield (Searcy and Hardison, 1960). A break in the curve is drawn at the 1956 water year for the Dry Creek data; and breaks at the same year are indicative for Fox and Mitchell Creeks, but they are not drawn. For hydrologic data plotted in this manner, breaks that persist for less than about 5 years should be attributed to chance and ignored unless there is definite reason for believing that such a break should occur. The high rainfall and runoff during the 1951 water year explains the break at the end of that LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 327 24 ‘v, I I I I i Dr Creek d _ _ _ Y 7 r ‘r 20 e _ / ' Q 0 Q9 / ’/ _ / #K1956 _ .6000 j 18 —— _ / "L _ , / +’ /// X0900 I I, :4 _ _ _ 1951 ‘ E . . . i la: 16 Slopes of lines representing < discharge-weighted concentrations 3 I O I <0 _ ,’ _ n: / Lu :3. u: E 14 / '- Data for 1951-58 water years “- _ / except Mitchell Creek which _ O , is for 1952-57 water years ’10 2 / < 12 U) D O I __ __ ’- / z 1’ ". /’ DJ 10 3 / < I O — __ a / D [I I ’2 8 ’ E Fox Creek /x_ D X m — _ 0) Medicine Creek above Harry Strunk Lake 6 4 | _ V’AMitchell Creek _ | M ’0 :195. 0.. —— dicine Creek at Maywood — l I l l l 1 I l 60 80 100 120 140 160 180 200 WATER DISCHARGE, IN CUBIC FEET PER SECOND DAYS PER SQUARE MILE FIGURE 228.—Cumulative water discharge and sediment discharge per square mile at gaging stations in Medicine Creek basin. Year at which a curve changes in slope is indicated. 328 year. Data are available for only the part of the year when rainfall was extremely heavy. The reason for the break at the end of the 1956 water year is obscure. Runoff for the 1957 water year was gen- erally greater than normal, but the break is in the wrong direction on the basis of the experience for the 1951 water year. Land-conservation work was being actively carried on during the period of the in- vestigation, but there are too few data to evaluate the significance of the break. The ephemeral streams—Brushy, Dry, and Mitch- ell Creeks—have higher discharge—weighted mean concentrations and generally yield larger amounts of sediment per square mile for storm runoff than the perennial streams. GEOMORPHIC PROPERTIES IN RELATION TO WATER AND SEDIMENT DISCHARGE One major objective of quantitative geomorphic measurements is to select drainage-basin properties that have the highest possible degree of correlation with runoff and sediment yield. If runoff and sedi- ment yield are known for a large number of drainage basins in the same region, statistical methods can be used to evaluate the significance of different drain- age-basin properties. If, as in the Medicine Creek basin, runoff and sediment yield are known for only five subbasins, evaluation of the properties cannot be made rigorously by statistical methods. Ideally, the properties selected should be independent of one another, unambiguous, and not unduly time con— suming to obtain. For the Medicine Creek basin, properties that meet these qualifications and offer promise of good correlation with runoff and sediment yield are relief ratio, adjusted frequency of first- order streams, and percentage of area in upland. Relief ratio was shown to correlate with mean annual sediment accumulation for small drainage basins in the upper Cheyenne River basin by Hadley and Schumm (1961, p. 173). Maner (1958) reports that relief ratio correlated more closely with sedi- ment delivery rates than did size of sediment con- tributing area, drainage density, basin shape, or weighted—average land slopes. The drainage basins studied by Maner, which are in the Red Hills area of Oklahoma and Texas, range from 332 to 0.036 square miles. For these basins, relief ratio showed a correlation, as based on inspection of scatter dia- grams, with basin size, basin shape, average land slopes, and drainage density. Maner concludes that sediment delivery rate in the Red Hills area is a function of several drainage-basin properties that evidently are expressed adequately by relief ratio. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT For drainage basins that are incised in a level or gently sloping upland surface to approximately the same local relief, a correlation of relief ratio with size and shape properties is apparent. Basin relief increases at a much slower rate than basin length; therefore, relief ratio is reduced both by increase in basin size and by increase in basin length. Because of the decrease in valley-side slope with increasing channel order, the lower relief ratio of high—order basins is accompanied by a lower mean value of side slope. Similarly, because of the decrease in channel slope with channel order (fig. 229), the lower relief ratio of higher order basins is accompanied by a lower mean channel gradient. However, the rela— tion of channel slope to channel order may not be consistent; therefore, relief ratio may not express adequately the slopes of low-order channels. In a discussion of the interrelations of drainage- basin characteristics, Gray (1961) presents evidence to show that, for small basins, the properties of area, length of main stream, and length to center of area are highly correlated. In a region of homo— geneous relief, a correlation also exists between these properties and the slope of the main stream. Al- though not discussed by Gray, relief ratio probably will express adequately all the above interrelated properties for some regions. In addition, relief ratio will express differences in relief between basins of about the same size. The effectiveness of relief ratio as a property probably depends on, among other things, the constancy in shape of long profile among the basins being compared. A basin whose main 1.0 _. I I I I I I I I I I I I : F _ _ o _ _ 0 LI. .- K L|J I. k N I- 0.1 _ 1 _ o _ _ E “ 2 z - 2 4‘ - _ O “ _ 2 0 +0 _ E " :2 3 ‘0 _ o o 4 _ ‘f _ w 3 <1 0 a 0+ g 4 . 4 ‘ 0.01 _ z - ._ :‘E = 4 ~ _ 4 5 - o _ 5 _ E _ _ I; — 5 6 6 ‘ s 6 7 — 7 8 0.001 1 I l | l l l I l I I I CHANNEL ORDER FIGURE 229.—Mean channel slope in relation to channel order for Lime, Dry, and Fox Creek subbasins. LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA stream is strongly concave upwards in long profile may have the same relief ratio as another basin, of equal size, whose main stream is nearly straight in long profile. However, such differences in long pro- file surely afi'ect runoff and sediment yield. Frequency of first-order streams was found by Morisawa (1959) to give a high correlation with peak intensity of runoff for drainage basins in the Appalachian Plateaus. The properties used by Mori- sawa in multiple regression analysis were relief ratio, basin circularity, and first-order channel fre- quency. She considers that these three properties express the shape, relief, and network composition of a drainage basin, that they vary independently of each other and of area, and that they do not dupli- cate any other geomorphic factor. For the Medicine Creek basin, relief ratio is considered to be an ade- quate expression of basin shape, but first-order channel frequency is an important property not only in relation to runoff and sediment yield but also in relation to gully erosion. Frequency of first-order channels expresses the small—scale properties of a drainage basin and there— by complements relief ratio, which expresses the larger scale properties. Channel frequency must be adjusted for immature basins according to their differing percentages of upland area. In figure 230 scatter diagrams show the relation of adjusted fre- quency of first-order channels to mean valley-side slope, slope of first-order channels, and mean length of first-order channels. The number of points is too small to warrant the calculation of correlation co- efficients, but the existence of a relation is apparent not only from the scatter diagrams but also from examination of the topographic maps. In addition, the reason for the correlation is apparent from geo- morphic history. In subbasins where post-Peorian dissection was most intense, first-order channels are shorter, steeper, and more numerous. Much of the scatter of points in figure 230 is due to the fact that some of the subbasins are less homogeneous than others in texture of drainage. The relation of length to channel slope for first- order channels in the lower part of Dry Creek is shown in figure 231. Each point represents the mean slope and mean length of 10 channels in a particular length group. Noteworthy is the fact that the shorter channels have relatively steeper slopes than the longer channels. In addition, the shorter channels have steeper valley heads and more concave profiles. This is of particular importance in gully erosion because steep valley heads promote the formation of new gullies. In combination, relief ratio and frequency of first- 329 450 . . . l _ .\ _ 350 \\ 250 \_f_ AVERAGE LENGTH OF FIRST-ORDER CHANNELS IN FEET I 1 150 I l I 1 0-30 I I l I 0.25 0.15 SLOPE OF FIRST ORDER CHANNELS, IN FOOT PER FOOT 005 I I I I 0.45 I I l 0.25 ' / - // X - 0.15 /' l I I 120 160 200 240 280 ADJUSTED FREQUENCY OF FIRST-ORDER CHANNELS, IN CHANNELS PER SQUARE MILE EXPLANATION MEAN VALLEY—SIDE SLOPE, IN FOOT PER FOOT I O —O— A X Brushy Creek Fox Creek Dry Creek Mitchell Creek + A V O Well Canyon Upper Dry Creek Lower Dry Creek Lime Creek FIGURE 230.—Adjusted frequency of first-order channels in relation to mean valley-side slope, mean slope of first-order channels, and mean length of first-order channels. order streams should give indirect expression to another important variable—the percentage of area in valley flat. An unusual geomorphic feature of the Great Plains generally is the flatness of valley 330 0.240 | l 0.200 \ 0.160 \ \ \ 0.120 MEAN CHANNEL SLOPE, IN FOOT PER FOOT _ \KQ, _ 0.080 I I I 100 300 500 700 MEAN LENGTH OF FIRST-ORDER CHANNELS, IN FEET FIGURE 231.—Slope of first-order channels in lower Dry Creek subbasin in relation to length. bottoms along channels of both low and high order. The percentage of area in valley flat for each channel order in each major subbasin has been calculated from the product of average width of valley flat and average length of channel. (See table 4.) The per- centage distribution of valley flat according to chan- nel order differs considerably from one subbasin to the next and depends mainly on the distribution of channel length according to order. In Well Canyon subbasin, for example, about 18 percent of the total area of valley flat is in the wide bottom along the main channel, whereas in Dry Creek subbasin only 8 percent is along the main channel. The probability of deposition of sediment on the main valley flat seems greater for Well Canyon than for Dry Creek, and the probability of rapid runoff is less because of the lower slope angle of the Well Canyon valley. The relief ratio gives an indication of the percentage of valley flat along channels of fourth and higher order, whereas the frequency of first-order channels gives an indication of the percentage along channels of first, second, and third orders. (See fig. 198, in Which the slope of the curve relating channel length to channel order changes at about third order.) Because of the immaturity of drainage in the Medicine Creek basin, a variable is needed to express the difference among subbasins in percentage of upland. The slope of the upland, which shows vari— EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT ation from one subbasin to the next, is related to the slope of channels. In general, the slope of the upland is less than the slope of first- and second- order channels, about equal to the slope of third-order channels, and greater than the slope of fourth-order and higher order channels. (See tables 3 and 4.) The percentage of upland gives a reasonably good indication of land use. (See table 9.) Similarly, the percentage of area in side slope and valley flat shows a correspondence with the percentage of area used for hay and pasture. The probable effect of the selected properties (relief ratio, frequency of first-order streams, and percentage of area in upland) on runoff and sedi— ment yield can be inferred. On the premise that lower slopes, wider valley flats, and larger (or longer) drainage basins all tend toward reduction of runoff and sediment yield because of greater oppor- tunity for infiltration and for deposition of sediment, runoff and sediment yield should be positive func- tions of relief ratio. According to the graphs of Hadley and Schumm (1961), the function is expo- nential Where mean annual sediment accumulation in reservoirs is the dependent variable. According to the graph of Maner (1958, fig. 3), the function is exponential where sediment delivery rate is the dependent variable. Sediment delivery rate is the ratio, expressed as a percentage, between annual rate of sediment yield and annual gross erosion rate. An increase in frequency of first-order channels is associated with an increase in both valley slope and side slope for lower order channels; therefore, increase in water and sediment discharge should be a positive function of frequency of first-order chan- nels. For drainage basins whose main channel is very long and bordered by wide valley flats, a high yield of water and sediment from lower order chan- nels may be largely dissipated along the main channel. TABLE 9,—Areas in upland and valley flat in comparison with areas in two categories of land use [Land-use percentages are mean values from inventories taken in 1954, 1955, and 1957] Area in Area in Area in row crops, Ratio of side slope hay and Area in small row crops Subbasin and pasture upland grain, and to small valley flat (percent) (percent) fallow grain (percent) (percent) Mitchell Creek 48.2 51.3 51.8 46.6 1.72 Dry Creek ........... 65.2 69.7 34.8 28.7 1.43 Well Canyon ........ 79.6 83.4 20.4 15.6 1.34 Fox Creek ........... 83.5 86.2 16.5 12.7 1.57 Brushy Creek ..... 74.1 77.7 25.9 21.7 .98 LOESS-MANTLED GREAT PLAINS, MEDICINE CREEK BASIN, NEBRASKA 331 The effect of percentage of area in upland depends on the land use of the upland. If the upland is used mainly for row crops, runoff and sediment yield probably would be a positive function of percentage of area in upland. If, on the other hand, the upland were thickly sodded, the function probably would be negative. A positive function would be expected for the Medicine Creek basin. The preceding inferences as to the geomorphic properties most highly correlated with runoff and sediment yield cannot, unfortunately, be rigorously tested for Medicine Creek because of the small number of subbasins for which data are available. Nevertheless, a trial multiple linear regression of sediment discharge in relation to the three selected variables (fig. 232) is given in order to illustrate the numerical effects of these variables. The trial regression is based on data in table 3 and figure 228. The uncertainties involved in correlation are ap- parent from the graphs in figure 232, and clearly no quantitative significance can be attached to the results. In figure 232 the assumption is made that the sediment discharge of each subbasin for the period 1952—58 is correctly represented by the data. Data for 1951 were not used because no measurements for Mitchell Creek were made during that year. The 1958 suspended sediment for Mitchell Creek is esti— mated. Relief ratio is plotted against sediment dis- charge on semilogarithmic paper because previous work has indicated that this relation is exponential. Percentage of upland and adjusted frequency of first-order streams are plotted against sediment discharge on rectangular coordinate paper. When only relief ratio is considered (top graph, fig. 232), the sediment discharge of Fox Creek is very low and that of Mitchell Creek is high relative to the trial curve. This is perhaps accounted for by the fact that the percentage of upland on Fox Creek is lower than the average, and the percentage on Mitchell Creek is higher. When relief ratio and percentage of upland are held constant (bottom graph), a reas- onably good correlation with adjusted frequency of first-order channels can be obtained for all the sub- basins except Medicine Creek above Harry Strunk Lake, which shows a high sediment discharge not only for this trial regression curve but for the other two as well. The relatively high sediment discharge of Medi- cine Creek above Harry Strunk Lake is attributed to yet another factor—the presence of raw vertical banks along a continuous incised channel for a dis— tance of several miles upstream from the gaging station. In general, the banks along Mitchell, Fox, 10 | I I I I w — .4 z o — _ I- LL 0 — _ w a 3 — _ mm A 8:‘ I2 — _ I—uJ {I E< “53 08 (I '— .— 4. 76 SURFACE SAMPLES (TO 6-IN. DEPTH) FROM SLOPE TRANSECT 10 28 5 2 5 50 6 20 3 1 6 64 6 20 6 2 6 60 10 23 8 2 6 51 9 26 9 6 13 37 9 33 6 5 10 37 5 25 6 5 12 47 5 29 8 6 15 39 SOIL HORIZONS SALEIIPLED IN EXCAVATIONS ppel' 10 22 10 2 2 54 9 16 6 3 5 60 16 18 6 3 3 49 7 43 10 4 10 26 12 19 9 6 13 41 3 32 7 4 10 44 Dark brown 6 ................. 31 18 6 15 0 l2 ............... 10 41 17 6 8 18 Lower 7 _________________ 21 22 8 3 5 41 13- _______________ 2 33 6 5 11 43 The dark-brown zone is composed principally of aggregates of fine-grained calcite, but it contains a few dolomite pebbles and some fine-grained dolomite. Be- cause this zone is parallel to the ground surface even in areas where the mantle is being incised by stream channels, and because it is locally connected to the up— per dark zone by inclined layers of similar soil, it is thought to represent a true soil profile horizon that has developed in place, rather than a buried soil or accumu- lative layer. 349 The cement of the aggregates and casts and the coat— ing on pebbles seen throughout the soil is interstitially precipitated calcite. Calcite precipitation probably takes place in the late spring and summer, when the moisture content of the soil is reduced by evaporation and plant transpiration. The original source of the cal— cium carbonate is apparently the elastic dolomite mak- ing up most of the soil. Dolomite (CaMg (003).) dissolves congruently in water but rarely precipitates from dilute aqueous solutions under surface conditions (Garrels and others, 1960). Although seasonal precipi— tation of calcite alone from water originally containing dissolved dolomite would lead to progressively higher concentrations of magnesium in soil solutions, no evi- dence was found for the presence of secondary mag- nesian compounds in the soil. Perhaps periodic flushing by downward-moving water removes this dissolved magnesium. A textural feature common to nearly all the soil samples is the bimodal distribution of particle sizes. If secondary aggregates in the intermediate size classes are disregarded, half of a typical sample is composed of dolomite pebbles and cobbles and half of very fine sand-size and silt—size dolomite and silicate grains. The two principal size classes reflect two distinct modes of rock breakdown—the large multigranular particles, separation along joint and fracture surfaces, and the small dolomite fragments, dislodgement from individ- ual crystals along cleavage planes. The relative proportions of the two kinds of elastic particles are related to the kind of process that produces them. Frost shattering—the result of the constrained expansion of freezing water—is capable of dislodging particles in both size classes. The abundance of course debris, the meteorological evidence of frequent freeze- thaw cycles, and the widespread development of pat- terned ground suggest that frost action is a primary mechanical weathering process in the White Mountains. The growth of tree roots in fractures has clearly re— sulted in the dislodgement of large bedrock masses from cliffs. Similar wedging action by plant rootlets may contribute to the breakdown of soil particles. Inter- stitial crystal growth, colloid plucking, and other small- scale processes are probably active, especially within the soil. The processes of chemical alteration that are so active in the breakdown of silicate rocks are not operative except that of simple solution, the only such process that can affect the relatively pure dolomite. VEGETATION The broad pattern of plant distribution in the White Mountains is similar to that in other regions of high relief in the Southwest (Merriam, 1890). Four major 350 vegetational zones can be distinguished (Mooney, and others, 1962): desert scrub at altitudes below about 6,500 feet; pinyon—juniper woodlawn from 6,500 to 9,000 feet; subalpine coniferous forest from 9,000 feet to upper tree line, at about 11,500 feet; and alpine above tree line. In detail, plant distribution is related in both kind and amount to topography, rock type, and soil and slope characteristics. The subalpine coniferous forest, which includes both bristlecone pine and limber pine (Pinus flem'lz's James), is neither continuous nor homogeneous within its broad altitudinal limits. Only scattered stands of conifers are found in topographically favor— able locations underlain by limestone, shale, sandstone, or granite. These patches of forest, especially at lower altitudes, are composed mainly of limber pine. The dolomite areas support relatively dense continuous stands of bristlecone pine. BRISTLECONE PINE The bristlecone pine grows near upper tree line in many of the high mountain ranges of the Southwestern United States (Munns, 1938). Where geomorphic changes have been sufficiently rapid, the extreme longev— ity of many bristlecone pines makes possible the study of local degradational rates over the past several thou- sand years. The great age of some individual trees of this species was first discovered by Edmund Schulman in the White Mountains (Schulman, 1956), and by 1958, 17 specimens more than 4,000 years old were known in this area. Other bristlecone—pine stands in California, Nevada, and Utah are also known to contain very old trees. Currey (1965) recently described a 4,900-year— old bristlecone pine in eastern Nevada. The White Mountains support one of the largest known bristle— cone pine stands containing a large number of old trees, but relatively few trees have attained extreme ages (in excess of 4,000 years), and some areas contain many acres of only young trees (< 500 years old) . The oldest tree studied in this report is 3,100 years old, and the average age of those dated is only about 1,000 years. Most of the very old trees are found in restricted sites, near the lower forest border or on rocky exposed ridge crests. The great age attained by conifers apparently growing under the most severe local condi- tions has been discussed by Schulman (1954, 1956). The mature bristlecone pines have a great variety of sizes and forms. The trees in areas of high stand density are tall and straight. Each has a single stem that is circular in cross section and bark covered around the entire circumference. In contrast, the very old trees are isolated or are in more open stands. They are typi- cally squat and gnarled and have many dead branches EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT and large areas of exposed deadwood. The bristlecone pines are not large because they grow slowly, adding only 1/2—2 inches of wood along a given radius in 100 years. The tallest specimen reported in the White Moun- tains (Billings and Thompson, 1957) is only 60 feet high. The Patriarch, a multiple-stemmed tree near up- per tree line, is 37 feet in circumference, although it is only 1,500 years old (Schulman, 1958, p. 358). Living bristlecone pines are rarely overturned. There are many standing dead trees, however, and dating of the outermost growth rings shows that some died more than 1,000 years ago. These trees apparently fall only after the supporting roots have decayed or been under— mined by deep erosion. Some long-dead trees, firmly rooted in bedrock fractures, have weathered to mere stubs. The extent to which dead roots have been pre— served depends on the length of time that they have been exposed—small branch rootlets are still present on roots that have been rapidly and recently uncovered. At the other extreme, the root systems of a few long-dead trees have been reduced to formless stubs projecting a few inches outward from the base of the stem. The cool semiarid climate and the dense resinous na- ture of the wood seem to be responsible for the unusual persistence of the exposed deadwood. The stems and branches of standing trees and the roots lying above the ground surface are usually sound. Dead roots, fallen logs, and branches partly buried in the soil have rotted. Conditions seem to be most favorable for decay on the relatively moist north-facing slopes, which have denser vegetation and are littered with organic debris. TREE-RING- DATING Precise ages can be asigned to individual growth increments in the secondary xylem of bristlecone pines in the White Mountains. The dating method involves the counting and correlation of annual rings exposed in cross sections or in cores taken with an increment borer. The method is based on the number of rings and on year to year variations in ring width that are correlative among most of the trees in the area. During the summer growing season, new wood nor- mally forms in a concentric sheath around a. root or stem axis through activity of the cambial layer, which is immediately beneath the bark. Wood formed early in the season is light colored and possesses large thin- walled cells; wood formed toward the end of the growing season is much darker and has small thick-walled cells. A distinct annual layer is thereby defined, each layer appearing as a ring in transverse section. (See fig. 239.) The widths of rings formed in successive years differ. Certain years are characterized by narrow rings, not only at different points in the same tree, but also in most SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. nearby trees. Thus, a common response to some factor affecting the total seasonal growth is indicated. In semi- arid regions the availability of soil moisture is thought to be a determining factor; relatively thin rings may represent dry years (Fritts, 1966; Schulman, 1956). The “sensitive” growth records of some bristlecone pines show large year to year fluctuations in ring width. Trees with more uniform growth have “complacent” records. Sensitivity is associated with a low average growth rate and is characteristic of trees growing on rocky exposed sites or near the lower forest border; it is also typical of old trees. The relation of site to ring-width variability in bristlecone pines in the White Mountains was illus- trated by Fritts (1966). Locally absent or “missing” rings are also associated with slow growth and high sensitivity. Such rings occur only locally, if at all, in many trees and may not be present in a particular sample. Rings which are absent in the growth records of sensitive trees are found to correspond to relatively narrow rings in more com- placent records. F lse rings, representing more than one period of growth i: a calendar year, can be distinguished from true annual rings (Glock, 1937, p. 10) and are rare in the Whitefilountains (Schulman and Ferguson, 1956, p. 137) ; the have been noted only in the wood of very young trees. Cross dating (Douglass, 1914) is the correlation of distinctive sequences of wide and narrow rings (fig. 240). Simple ring counting yields precise dates only if no rings are missing from the sample, which must be from a. living tree and must include the outermost ring as a dating control. However ring sequences in a sam- ple can be cross dated with those in a dated sample from the same, or from a different, tree. This permits the dating of virtually any piece of wood from an area, provided that dated samples with overlapping or con- current growth records exist. To utilize the cross-dating properties of sensitive growth records, and yet retain the precise dating possible with complete, but com- placent, records, a chronology is made. This is a graph of the variation of average ring width with time and is constructed from the growth records of many trees in an area of homogenous ring-width variation. A chronology for the period from AD. 300 to A.D. 1954 was used in this study. It is based on the work (largely unpublished) of Edmund Schulman and C. W. Ferguson in the White Mountains. The chronology is similar to that published by Schulman (1956, p. 52), which is reproduced here in figure 239. Less precise con- trol in the period prior to A.D. 300 is provided by sam- ples from specimens with growth records extending back to about 2000 BC. Distinctive sequences in these samples were dated by ring count and used to cross-date 351 old samples that do not overlap the period covered by the chronology. Dates prior to AD. 300 obtained in this study are thus subject to an error due to the uncertainty in the number of locally absent rings in the control specimens. Experience in dating younger samples shows that 5—10 percent of the rings may be missing from growth records of highly sensitive trees. The older specimens can be more precisely dated when an ex- tended chronology becomes available. Through the use of cross dating and the building of a. local chronology, the tree-ring dating method can be made very precise in terms of the reproducibility of the results obtained by independent study. But the va- lidity of the calendar dates assigned to individual rings depends on the assumption that the growth increments represented in the chronology are annual rings. At least one line of evidence suggests that they are. Only one ring has been formed each year by most of the bristle- cone pines in the 10-year period since the first samples were collected by Schulman, as shown by comparison of samples collected in 1963 with the published chronology (Schulman, 1956, p. 52). In the Reed Flat area the out- ermost rings in some of the bark-covered stumps of cut trees have been dated in the mid~1860’s by cross dating with living trees in the vicinity; this date is corrobo- rated by the fact‘ that bristlecone timbers were used in a nearby mine first located in 1862 (Norman and Stew- art, 1951). Because the rings formed during this period are annual rings and do not differ qualitatively from those of earlier periods, it is felt that accurate dates can be assigned to growth rings in the wood of bristlecone pines. Although the potential accuracy of the ring-dating method is great, definite limitations are inherent in it. Some trees, during long periods of extremely slow growth, have added only one-half an inch of new wood in 100 years. Such intervals are difficult to date because the component rings are only a few cells in width. Reso- lution of individual rings is poor, and cross dating is al- most impossible. It is also difficult to cross date samples in which numerous rings are locally absent. These prob- lems can be partly overcome by sampling sectors of rel- atively rapid growth within a specimen and by cross dating the samples in intervals of maximum growth rate. The dating procedure did not include the actual measurement of ring width in wood samples. The razor— cut surface of the mounted sample core, daubed with tur— pentine, was first scanned under low magnification for distinctive ring patterns of known age. Because the main objective was the dating of the specimen itself rather than the study of its growth record, the older part of a sample or the oldest sample from a given speci- 352 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT o©©©ooooo OCDQCDQOOOO 060000000 888%8888% OQQQOQOOO ANNUAL 0 CDC) 0 0000 88988%% Q o 8C3c3000000 RING C30 C30 0000 0 GO GO 0000 ,0 :30 C30 0000 GOOD ©0000 OQQ D 00000 OIOO O 00 000' \ / \ / \\ // \ / \ / \ / \ / \CORE Bark Cambium 1963 1894 1899 1910 1923 19291934 1950 0.75 3g; I In E E 0.50 E E (D i 0.25 E 2 m 0.00 I l l I l I l | | | I l I I | I 1850 1900 1950 CHRONOLOGY FIGURI 239.—-Cross dating between portions of two cores, and chronology with diagrammatic enlargement of transverse section through annual ring. Chronology for White Mountains after Schulman (1956, p. 52). SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. men, was studied first. If no obvious correlation was possible and if the sample included the outermost ring of a living tree, the sample was approximately dated by ring count. Frequently this led to the recognition of cross-dating sequences in which one or more rings are microscopic or locally absent. If not such outer control were present, as with samples from dead trees or logs, a skeleton plot was made showing the relative spacing of narrow rings (Glock, 1937, p. 17). By comparing the skeleton plot with similar plots made from dated samples, or with the chronology, cross dates were often obtained. Wood samples from nearly 200 specimens with an average age of over 1,000 years were dated by these methods. AGE ESTIMATE Weathering, erosion, and decay have destroyed the older wood of the stems and roots of many of the bristle— cone pines. Therefore, determination of specimen age requires an estimate of the timespan represented by the missing wood as well as the dating of that wood which is still sound. This estimate is based on the probable amount of radial growth missing and on the inferred average growth rate during the period. The original center of secondary growth (stem or root axis) can be approximately located by inspection of the remaining wood. In the old trees with greatly reduced ratio of cambial area to total circumference, the growth layers formed after initial cambial reduc- tion are not continuous and are not concentric about the axis. However, as shown by well-preserved speci- mens, even these trees grew at a normal rate during an early period of up to several hundred years, forming an inner core 3—12 inches in diameter. Where portions of this early wood are preserved, the stem or root axis can be located at the intersection of projected branches or branch rootlets or at the intersection of the radii of curvature of concentric rings (fig. 240). Thus, the ap- proximate distance from the end of radially directed increment core to the axis can be estimated. Also, the average growth rate during the period rep- resented by the missing wood is a source of uncertainty in the age estimate. Samples of sound wood show that most of the trees have an early period of relatively rapid diametral growth. For example, along one radius at a height of 4 feet, the main stem of a 3,000-year-old speci- men added 3 inches of wood in the first 40 years of growth, but only 9 more inches in the succeeding 800- year period. However, this is an unusually rapid growth-rate decrease. The growth rate shown by the wood in the inner 1 or 2 inches of a sample was generally used to estimate the timespan represented by the miss- ing wood near the axis. 269—085 0—67—3 353 o/Stem axis z/(E' stimated FIGURE 240.-—Transverse section of eroded stem showing the geo- metrical basis for age estimation. The pattern of growth layers results from reduction of the ratio of cambial to total circum- ference. An early period of concentric diametral growth is indi- cated by the form of the inner rings. Many exposed roots show the same general features, The estimated uncertainty in the age assigned to each specimen used in the study is given in tables 2, 3, 4, and 5; it averages about 5 percent of the determined age and is greatest for very old trees with large amounts of wood missing. The‘uncertainty in the age determina- tions is comparable in magnitude to the uncertainties in the other measured quantities used in this work. ROOT EXPOSURE AND SLOPE DEGRADATION Exposed roots are direct evidence of degradation; unless a tree is overturned, its roots can become exposed only through removal of the enclosing'soil. However, this evidence has been little used to investigate degrada- tional rates except where wind erosion is involved. Sey— bold (1930) described the exposure of pine roots to a depth of 5 feet in 80 years by shifting of dune sand in Holland; Hueck (1951) used the exposure of the root systems of shrubs to estimate rates of aeolian denuda- tion in Patagonia. Deep root exposure is rarely seen on slopes degraded by mass-wasting and through erosion by surface runofl’. Degradation proceeds too slowly to cause significant lowering of the ground surface within the relatively short lifetimes of most trees. The un— covering and exposing of root systems, however, affects trees of several species in the White Mountains, includ- ing a 1,000-year-old limber pine and a 1,700-year-old juniper (Junipem sp.), as well as the old bristlecone pines. Root exposure is the direct consequence of shal- low root development and the great age of these trees. 354 ROOT SYSTEMS Bristlecone-pine roots can be seen in excavations and on overturned trees as well as exposed along the ground surface in the White Mountains. The root systems of mature trees are extensive but Shallow. Mapping of roots exposed on the wall of a pit Showed that more than 75 percent are concentrated in the uppermost foot of soil (Harold C. Fritts, written commun., 1965). Small roots penetrate to depths of several feet, but vertical taproot development is rare. In common with trees in other areas (Stout, 1956), rapid longitudinal growth apparently takes place early in the life of a bristlecone pine. Several major roots extend outward from a center at the base of the stem. Individual roots are largest at the stem junction, and most taper to a diameter of less than an inch within 10 feet; but some sparsely branched roots were seen that extend 20 feet or more with little change in size. The root systems of trees growing in coarse rubble or rooted in bedrock fractures are less regular. The close relationship of growing roots to the over— lying ground surface is also demonstrated by exposed root systems that parallel the profile of the topography that existed at the time of root development. Where individual roots crossed preexisting topographic ir- regularites, such as those along ridge crests or at the edges of cliffs and steep banks, the exposed roots retain the original irregular form. Conversely, where local re- lief has developed on a previously smooth slope, as adjacent to trees with asymmetrically exposed root sys- tems (described below), the root system shows the original planar form. EFFECTS OF EXPOSURE The roots of woody plants grow in two ways——longi- tudinal extension by activity of the apical meristem is soon followed by secondary growth around the primary axis through the addition of successive layers of second— ary xylem by the cam'bium (Esau, 1953). Only the ter- minal parts of the young branch rootlets absorb soil water. The sheaths of secondary wood that make up most of a mature root serve first for conduction of fluid and later as supporting tissue. The structure of mature roots is thus very similar to that of the stems and branches. Uncovering of a trunk root near the stem of a bristle- cone pine apparently has little immediate effect; how- ever the terminal, water-absorbing parts of the root system function only within the soil; they die when ex- posed, as can be seen along roadcuts and in excavations. Many of the naturally exposed roots dealt with in this study are also dead. The roots on the downhill side of a tree are uncovered more rapidly and more completely EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT than those projecting uphill or to one side. Many of the root systems in this downhill sector have not survived exposure. BUTTRESS ROOTS The development of a buttress form by individual lateral roots is a direct result of exposure (LaMarche, 1963). These roots are high but relatively narrow in transverse section, A buttress root is bark covered only on the bottom and owes its asymmetrical form to sec- ondary growth radially downward from the root axis. Only the narrow strip of bark along the base, with its underlying cambium and conductive tissue, connects vertical or inclined branch roots with the stem. Initial reduction of the cambial area follows the uncovering of the upper surface of the root. Continuous concentric growth rings can be seen around the axis of a well-pre- served buttress root, but the growth layers that formed after cambial reduction are limited to the lower side of the root and terminate at the sides. This discontinuity in the form of the growth layers marks the approxi— mate time of the initial root exposure. All stages of buttress root development are seen. The degree of asymmetry depends on the diameter of the root when it is first exposed and on the period of time since its initial exposure, as well as on the average growth rate. Uncovering of a shallow root normally takes place several hundred years after longitudinal growth. This interval is the time required for the re- moval of the overlying soil and is related to the original depth of root development and to the local degrada- tional rate. Rapid diametral growth is also a factor in early exposure of the upper surface of a root. The but- tress form can be developed only by living roots; the roots of trees that died before exposure do not show this feature (fig. 241). Uncovering of roots is not a recent phenomenon in the White Mountains, as is shown by the existence of buttress roots of different ages, stages of development, and depths of exposure. Root exposure and buttress root development have been regularly associated with increasing tree age during at least the past 3,000 years, as it will be shown subsequently. PROBLEMS OF MEASUREMENT CHOICE OF DATUM Only the axis, or center of radial growth, of an ex- posed root can be validly used to estimate the position of the ground surface as it existed at the time of root development. The top of a very shallow root may be uncovered simply ‘as the result of increase in diameter with time. Although most developing roots are buried to a certain depth in soil that must be removed before the roots are uncovered, this depth is not known for ex- posed roots. For any exposed root, all that is known is SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. FIGURE 241.—-—Stump of 350-year-old bristlecone pine that died about AD. 1350. Many such standing snags in the Reed Flat area have been cut for poles. Lack of buttress root development shows that root ex- posure has occurred since the tree died. that' its axis developed within the soil and that the position of the root axis marks the minimum possible level of the original surface. The depth of root exposure is the vertical distance from the root axis to the underlying ground surface (fig. 242). Use of an arbitrary figure representing the assumed initial depth of burial might improve the esti- mate of local slope degradation for a single tree, but this may be unnecessary in an area where a large num- ber of trees are studied. As shown in a subsequent sec— tion entitled “Degradational Rates,” data based on measurements of the depth of root exposure can be ex- trapolated to estimate the average depth of root de- velopment. In two separate areas this depth is shown to have been about one-half a foot, remarkably similar to observed depths of concentration of rootlets. The point of inflection of the basal flare of a tree (fig. 2420) is commonly misidentified as representing the original ground level at which root exposure oc- curred. This impression may be fortified by a pro- nounced change in the character of the wood or bark at this point. However, this typical feature of mature forest trees bears no relation to degradation. Clearly the inflection point will migrate upward with respect to even a static ground surface as a consequence of in- creasing stem and root diameter. If the inflection point is mistakenly used as the datum for measuring degrada- 355 tion, the degradational rates will be overestimated (fig. 243). ' ASYMMETRICAL EXPOSURE Asymmetrical root exposure causes large discrepan- cies between maximum depths of root exposure and actual slope degradation. The root systems of most bristlecone pines on debris-mantled slopes have not been uniformly around each tree. Typically (fig. 244), there A. ROOTS DEVELOP Stem Ground L surface s 7% Roots NM B. ROOTS UNCOVERED Origin_al g __g;oyfld_,_,__jflia£e _____ mm C. ROOTS EXPOSED lnflection point ground surface _____ O I'gl—r—‘afl— Basal flare § FIGURE 242.—Stages in development and exposure of a root system. Top‘of basal flare — -— _ _ _ Original level of — -— —- ground surface Original depth of burial Level of root axis — —— —— Total Depth of degradation rool exposure surface Ground FIGURE 243.—Cross section of exposed root near Item showing relation of root exposure to degradation. 366 EROSION AND SEDIMENTATION exists a gently sloping terrace on the uphill side of an old tree, and a concave hollow on the downhill side. Correspondingly, the roots upslope from the stem may be little exposed or may be buried, whereas those on the downslope side are deeply exposed. The ground sur- face may drop 6 feet in a horizontal distance of only 3 feet where the surrounding slope is at an angle of 35° or less. As shown by the data in table 4, the maxi- mum depth of root exposure is commonly three to four times greater than the slope degradation. The prominent topographic discontinuity is caused by the damming effect of the trees and develops where the stem and roots impede the downslope movement of surficial rock debris. Material accumulates above the barrier, forming the terrace. The hollow results from the net removal of surficial debris downslope from the tree, and in many places bedrock is exposed in the hollow. Microrelief features and the orientation of peb- bles in the surficial mantle show that the downslope flow lines are divided in the vicinity of such an obstruc- tion and that material moving downslope passes on either side of the tree. Such topographic changes are not associated with deep root exposure in two situa- tions: (1) where the root system extends only upslope or downslope, the. exposed roots do not interfere with debris movement (fig. 245), and (2) where the roots of a tree growing on a ridge crest extend down opposite slopes subparallel to the directions of movement and show correspondingly symmetrical exposure (fig. 246). st: FIGURE 244.—Terrace and hollow due to damming effect of old tree. Maximum depth of root exposure is about 4 feet. Note rock fragments spilling over top of root at stem base. IN A SEMIARID ENVIRONMENT FIGURE 245.—Root system symmetrically exposed because roots ex- tended directly upslope. Centers of roots are alined parallel to slope but lie about 2 feet above surface—a measure of degradation subse- quent to establishment of tree. FIGURE 246.—Root system exposed symmetrically because of location on ridge crest. Roots extend down opposite slopes subparallel to direc- tions of movement of rock debris; damming efifect is further reduced because volume of debris is small. SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. 35.7 E Maximum depth of root exposure D Minimum slope degradation 1 l Present ground surface Original ground surface Plane of root development FIGURI 247.—Slope profile in the vicinity of a tree with asymmetrically exposed roots. Shows terrace and hollow, approximate location of original ground surface, and basis for measurement of minimum slope degradation. Asymmetrically exposed root systems also can be used to estimate slope degradation if the surrounding slope is smooth and regular. In a study of two selected areas, a downslope profile at a scale of 1 inch equals 10 feet was made in the vicinity of each tree having an asymmetrically exposed root system associated with terrace and hollow development. Figure 247 shows that a line passing through the axis of the highest exposed roots lies above, but is parallel to, the line formed by projection of the present surface from points above and below the tree. The vertical distance (D) between the two lines approximates the minimum slope degradation (it does not include the original depth of burial) in the vicinity of the tree since the development of the root system. The vertical distance (E) between the root axis and the underlying ground surface is the maximum depth of root exposure. This depth is always found in the hollow on the downslope side and is always greater than the overall depth of degradation. CALCULATION OF DEGRADATIONAL RATES The study of an individual tree and its exposed roots can yield the maximum depth of root exposure and the period of time since the tree’s initial root develop- ment. A rate of exposure calculated from these data generally will not be equal to the local rate of degrada- tion during the same period. One source of error is the uncertainty as to the original depth of root develop- ment, for the vertical distance from the axis of an exposed root to the present ground surface is only a minimum estimate of the total depth of material removed. This error is fairly large for young trees that have only incipient roolt exposure; the calculated rate of root exposure will be less than the actual rate of degradation. Where the root systems of trees growing on slopes have been asymmetrically exposed, the max- imum depth of exposure may be'much greater than the actual slope degradation. The increase in degree of asymmetry with size and age of tree introduces a large discrepancy between local degradational rates and the maximum rates of root exposure of many older trees However, the study of individual old trees will give rates of root exposure that approximate degradational rates if the root exposure has been symmetrical and if 5the axes of the highest exposed roots are used. An improved estimate of the local degradational rate, as 358 indicated by the depth of root exposure of a single tree, can also be obtained, if the initial depth of root develop- ment can be inferred. DEGRADATIONAL RATES LOCAL BATES Evidence of the progressive change in ground-surface altitude provided by the exposed roots of bristlecone pines was used to estimate rates of degradation in areas underlain by the Reed Dolomite. The problems of measurement and interpretation initially met with led to the refinement of methods used during the rest of the investigation. Early work was concentrated in 1—square-mile area near Reed Flat (fig. 235). Attention EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT was focused on trees with deeply exposed roots and on those occupying special topographic sites. Either a direct determination of root age or an esti- mate based on stem age was made for each tree. In 1962, 99 trees were sampled in the Reed Flat area and elsewhere in the White Mountains; some of the speci— mens were not dated, and some are on other substrates and are not included in the results. The age determina- tion, depth of root exposure, and local slope of the ground surface are listed in table 2 for each of the speci— mens, and the results are graphically summarized in figure 248. The measured depths of root exposure range from 0 to 4.5 feet, and the estimated ages, from about 200 to 2,500 years. A general increase in the max- imum depth of exposure with increasing age can be , 5.0 l l A Mean values by 500—year age classes 3 O 0 Age from root sample :1 Age from stem sample 2 4'0 _ - Minimum age estimate E 5 LL! “- o E 3.0 — “ ‘uJ ‘ 1 a: D m 0 u. >< LIJ p.— o O 0: LL 0 2.0 — - E a. Lu 9 1.0 — El _ 0 I l l l “I”? l O 0.5 1 O 1.5 2.0 2.5 3.0 b USANDS OF YEARS FIGURE 248.—Relatlon of root exposure to age and probable range of local degradational rates. SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. seen, but there are great differences in the rates of root exposure indicated by these data. The differences reflect not only different local rates of degradation, but also the effects of differences in the symmetry of exposure and in the original depth of root development. TABLE 2.——Age, root exposure, and site data for trees in scattered localities in southern White Mountains Age estimate (centuries) Depth of Specimen root Slope ex osure (degrees) Age Uncertainty feet) Root age from root sample 9. 0 0. 5 l. 0 33 3. 8 . 1 1. 5 28 11. 7 . 1 1. 5 12 12. 1 . 3 1. 7 12 14. 0 . 3 2. 1 12 14. 5 . 2 1. 3 12 15. 5 . 5 2. 5 12 5 ......................... 7. 5 . 5 1. 4 26 6 ......................... 1 121+ ______________ 2. 9 32 6. 2 . 1 . 3 28 6. 2 . 1 . 3 28 25. 0 2. D 4. 5 36 10. 0 1. 5 . 2 2 8. 5 . 3 2. 0 24 4. 4 . 1 1. 6 24 5. 3 . 1 1. 5 17 2. 5 . 2 . 1 8 1 8. 0+ ______________ 1. 0 40 8. 1 0 2. 0 30 8. 5 . 4 1. 5 26 13. 5 1. 0 1. 7 20 5. 0 . 2 . 5 14 12. 0 2. 0 1. 3 25 2. 0 . 1 . 2 32 9. 5 1. 0 1. 5 25 1 8. 0+ .............. . 5 10 6. 5 . 5 . 3 8 8. 0 . 2 . 4 9 6. 4 0 . 9 16 11. 9 . 1 1. 1 20 9. 0 . 5 . 6 18 2 7. 1 . 1 1. 6 26 1 7. 8 0 1. 5 26 4. 5 1. 0 l. 0 19 2 9. 5 1. 0 . 6 20 6. 0 . 2 . 8 21 7. 0 . 1 . 3 20 6. 5 . 3 0 6 8. 0 . 5 . 1 6 5. 0 . 2 . 4 9 7. 0 . 4 . 2 9 19. 5 1. 0 1. 7 21 i 20. 0+ .............. 2. 3 24 5. 0 . 3 1. 5 16 24. 0 2. 0 3. 1 23 1 11. 0 ______________ ' 2.1 36 Tree age from stem sample 1. 8 0 ______________ 17 8. 8 0 0. 1 5 13. 0 . 5 . 2 8 6. 5 1. 0 ______________ 10 11. 5 . 1 1. 1 22 21. 0 1 0 1. 0 20 17.0 1 0 1. 5 16 3 20. 0+ ______________ 3.0 30 9. 2 0 1. 5 23 12. 0 . 2 1. 7 30 2 10. 5 .3 1. 6 25 1 Sample incomplete. 2 Dea tree. Age based on cross dating. 3 Dead tree. Minimum-age estimate equals total number of rings counted. .7 . SLOPE TRANSECT The possible effects of topographic position on root exposure were studied in trees along a line that extends directly up a north-facing slope from the base of the 359 knoll east of Reed Flat to the crest of the adjacent ridge. Each standing tree within 5 feet of the line was sampled for an age determination, and the depth of root exposure was observed; the results are listed in table 3. The slope profile, tree ages, and root exposure are shown graphi- cally in figure 249. TABLE 3.—Age, root exposure, and site data for trees on slope transect, listed in order of increasing distance from base of slope [Ages based on stem samples] Age estimate (centuries) Depth of Specimen root Slope ex osure (degrees) Age Uncertainty ee 4. 0 7 1. 9 11 .8 3 6. 2 10 1. 5 18 1. 9 18 . 5 25 2.0 28 10. 2 28 5. 5 30 7. 5 32 6. 5 29 1. 9 31 6. 5 32 5. 5 35 6.0 34 2 7. 0+ 32 15. 0 12. 5 34 10. 5 33 6. 1 30 14. 0 28 9.0 36 10. 0 25 17. o 29 7. 0 18 6. 5 16 1 Indicates no roots exposed. 2 Dead tree. Minimum age equals total number of rings counted. The lower part of the slope, extending about 500 feet to the base of the main slope, is gently rolling; the aver- age slope is about 10°. The line of profile crosses two depressions marking incised channels. The trees in this area are widely spaced, and the ground cover is rela- tively dense. The main slope, rising 300 feet in a hori- zontal distance of 500 feet, has a nearly linear profile with a slope of 30°. The surface of the ground is smooth, but the coarse soil is loose and readily dislodged. Litter, including the stems of fallen trees, is abundant. The upper slope is distinctively convex in profile. Its ground surface is much rougher than that of the main slope below. Numerous outcrops and small cliffs protrude through the thin patchy veneer of surficial debris. A striking relationship between tree age and location is shown in figure 249. The 6 trees on the lower slope have a mean age of 220 years; the 10 trees on the main slope, 520 years; and the 11 trees on the rocky upper slope, near the ridge crest, more than 1,000 years. Comparison of tree age with depth of exposure shows that the roots of trees less than 500 years old have not yet been exposed, owing to either slow degradation or to n, My; 360 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 600 l l l l l— 500 — |.|.l L|J LI. E Upper slope* '5" 400 — — z < |._ ‘2 o _J 300 — — < E SLOPE PROFILE 33 Main slope > 200 — __ — 20 100 \ 0 Lower slope O _ 15 m o — o E Q) E z - 0 O — 10 3 o TREE AGE 0 O E o O 0 o o o if; _ O O n 5 < O o O 0 _ O (9 O ._ 0 1.5 w . — .— I.|J E 1.0 * — z o __ o E 8 o 5 _ ROOT EXPOSURE 0 _ s - ° E o . o 0 1 I | J ° l l . 0 200 400 600 800 1000 1200 1400 HORIZONTAL DISTANCE. IN FEET FIGURE 249.—Slope profile, tree age, and depth of root exposure along slope transect. great depth of development. Incipient exposure, where the roots on the downhill side are partly uncovered, is characteristic of trees in the 500- to 1,000—year age range. Only the fairly old trees have deeply exposed roots; however, the oldest trees do not show the greatest expo— sure. Roots of two of the oldest trees, 1,400 and 1,700 years old, respectively, are not now exposed. These spec- imens, which are in the area of irregular topography immediately below the ridge crest, have been partly buried by lobate rock streams. Although local rates of degradation, suggested by the depth and symmetry of root exposure, are apparently greatest on the upper slope, the conclusions that can be drawn from observation of an individual specimen are valid only for the limited area around that tree. Fur— ther, the restricted distribution of trees old enough to show significant root exposure precludes strict compari- son of degradational rates at different points along the transect. The average rate of degradation on this slope is probably less than half a foot per 1,000 years. SELECTED AREAS The initial reconnaissance study of individual bristle- cone pines showed that deep root exposure is generally associated with great age. It suggested that the rate of exposure is proportional to the local degradational rate but also depends on the depth of root development and the degree of symmetry of exposure. The great range in rates of exposure seems to be related to differences in degradational rates in different parts of the landscape. SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. Inconsistencies in the results of study of nearby trees showed, however, that bias in specimen choice could lead to results that are not representative of an area much larger than that included within the root system of a single tree. The purposes of the second phase of the study were to convert measurements of root exposure to estimates of slope degradation, to characterize local topographic ' environment closely, and to obtain unbiased samples of the trees in selected areas as a basis for generalization of the results. The approach used included the large- scale planetable mapping of two areas that represent contrasting slope types. The maps (plates 10 and 11) show the general outlines of exposed roots and of fallen trees, in addition to the topography. All the standing trees in one area and a large random sample in the other were studied. Each of the areas was selected because it has old bristlecone pines; the average age of 71 trees in the two areas is 1,100 years. A third area, significant because it is incised by closely spaced drainage channels, was mapped at a smaller scale. The exposed root systems of six trees on low steeply sloping interfluvial ridges were mapped in great detail. These specimens were selected because they have deeply exposed roots; the trees have a mean age of 1,450 years. ' AREA 1 This area of about 2 acres is one-half a mile north- west of Blanco Mountain (fig. 235) and lies at an alti— tude of 10,500 feet. It is the west end of a hill of Reed Dolomite that rises abruptly from a gentle sagebrush- FIGURE 250.——Area 1, an viewed from the west; Blanco Mountain is in background. Note abrupt change in type of soil and vegetation at base of triangular slope. 361 covered slope (fig. 250). A broad swale at the base of the hill slopes to the south, where it deepens and con— tains a first-order drainage channel. The center of this topographic depression is also marked by a strong con— trast in both vegetation and soils: on one side of the de- pression is the dolomite slope 'with its sparse ground cover and scattered bristlecone pines, whereas on the other side is an unforested fiat whose soil is composed of sandstone and shale from the slopes of County Line Hill, to the west. The mapped area (pl. 10) includes the western slope of the hill and a broad north—sloping shoulder. The lin- ear rocky crest is 100 feet above the base of the slope. Below the irregular cliffs along the crestline, the slope has a fairly smooth concave profile and an average slope of 20°. The ground surface of the hillcrest and of the subsidiary spurs is dolomite bedrock that has local ac- cumulations of coarse rubble. Rock outcrops are nu— merous, and the mantle is thin over large areas of the shoulder and the upper slopes. The mantle is thicker and more continuous on the lower slopes. The mantle is fairly coarse textured and contains pebble- to cobble-sized angular dolomite fragments. The average surface-particle size, measured along each of four downslope transects, ranges from 14 mm, on a 10° slope on the shoulder in the northern part of the area, to 33 mm, on a 22° slope extending down from the main hillcrest (fig. 251). There is an irregular decrease of grain size in the swale at the base of the slope, where the soil is a dolomitic sandy loam with few pebbles. The surface of the debris mantle on the slopes is un- even. The steeper part of the west—facing slope has poorly Sorted stone-banked terraces, presumably re- lated to frost action, but there are no miniature pat- terned-ground features. Terrace and hollow develop— ment is locally prominent adjacent to standing trees and to some of the large wood fragments that litter the slopes. Three main groups of bristlecone pines can be dis— tinguished. One group grows on the north-facing slope on the shoulder of the hill; the second group grows on the west-facing slope; and the third group, which in- cludes the oldest trees in the area, grows on or immedi— ately below the cliffs west of the hillcrest. There are 49 standing trees in the area mapped. Stem core samples were taken from all specimens, and dating provided at least a minimum age estimate for all but four trees. The trees range in age from about 10 to 2,700 years and average nearly 1,000 years. The root systems of most of the trees are uncovered or are at least partly exposed. The deep symmetrical ex- posure of the roots of trees along the hillcrest (fig. 252) is especially striking. The minimum local slope deg— 3612 PARTICLE SIZE, IN MILLIMETERS 29153024 8 9 813 51 ”:50 Mean 14 mm PROFILE 1 16 25 46 213824152727 6 27 6 5 i PROFILE 2 [1:70 Mean 22 mm 12546 40 35 4114 34 69 39 23 PROFILE 3 24 52 32 22 7O 27 58 n: 120 Mean 32 mm 50415043 21 7181614121914 3 1 PROFILE 4 46 46 18 32 2O 23 47 [7:100 Mean 33 mm 19 20 13 41 16 32 FEET 20 40 FEET 0 FIGURE 251.—Slope profiles and results of particle-size measure- ments. Each number represents mean of five measurements of intermediate dimension of rock fragments made at 2—foot intervals along line. radation represented by the root exposure was meas- ured directly where root systems were symmetrically exposed. It was obtained from the individually con- structed downslope profiles for thosevspecimens show- ing asymmetrical exposure. These data, together with age determinations, are listed in table 4. The minimum local slope degradation is plotted against specimen age in figure 253. The points are scat- tered around a straight line representing a slope of 1.2 feet per thousand years, the average degradational rate. The extrapolated average depth of exposure at zero age, based on a least-squares analysis of data from 33 specimens, is one-half foot and represents the aver- age initial depth of development of the roots. The EROSION AND SEDIMEN'I‘ATION IN A SEMIARID ENVIRONMENT FIGURE 252.—Exposed root of specimen 110. Depth of root exposure shown by comparison with 18-inch hammer. Exposed bedrock is broken up. Note symmetrical exposure of root of this 2,500—year-old tree. analysis did not include data from 10 younger trees with no root exposure nor data from 6 specimens which were not reliably dated. Even though the trees are not distributed uniformly over the entire area, the results are thought to be representative of at least the crest and the upper slope areas. TABLE 4.—-—Data for trees in two selected areas [Ages from stem samples] Age data E . D (centuries) Maximum Minimum Specimen depth of local 510 9 root expo— degradat on Age Uncertainty sure (feet) (feet) Area 1 1 13. 0+ .............. 1. 6 1. 4 8. 5 0. 2 l. 2 . 7 4. 8 . 1 . 4 . 1 7. 4 . 1 1.2 . 4 (3) .............. 1. 8 l. 8 18. 7 . 5 2. 0 1. 5 8. 1 . 2 . 8 . 4 3. 7 . 1 . 8 0 6. 5 . 2 . 1 0 25. 0 2. 0 3. 2 3. 0 27. 0 2.0 3.2 3. 2 7. 6 . 1 . 8 . 6 7. 6 . 1 . 6 . 3 9. 5 1.0 . 8 . 7 7. 5 . 3 . 4 —. 1 7. 0 . 2 . 5 0 4. 5 .4 (3) 4. 5 . 3 (3) 7. 5 . 1 . 8 4. 8 . 2 . 9 See footnotes at end of table. SLOPE DEGRADATION, WHITE MOUNTAINS, CAL-IF. TABLE 4—Data for trees m two selected areas—Continued [Ages from stem samples] Age data E D (centuries) Maximum Minimum Specimen depth of local slope root expo- degradation Age Uncertainty sure (feet) (feet) Area l—Continued 1. 0 . 0 (3) .............. 20. 0 1. 0 1. 7 1. 7 21. 0 l. 0 . 6 . 6 9. 0 2. 0 . 5 . 5 18. 0 2. 0 l. 0 1. 0 1 9. 0+ .............. 1. 0 l. 0 20. 0 3. 0 2. 6 2. 3 15. 5 2. 0 1. 7 1. 7 . 1 0 (3) .............. 6. 5 . 5 . 7 . 7 7. 0 . 5 1. 5 l. 3 7. l . l . 2 . 2 1 22. 5+ ______________ 2. 2 l. 5 6. 0 1. 0 . 4 .4 (3) .............. 1. 3 l. 3 5. 0 . 5 . 8 0 4. 5 . 3 . 4 —. 1 13. 0 2. 0 1.3 . 9 . 1 0 (a) .............. 9. 0 3. 0 2. 2 1. 2 . 4 0 (3) .............. 5. 0 . 1 . 6 . 3 (3) .............. .3 . 3 13. 5 1. 0 1. 2 l. 1 17. 0 1. 0 2. 5 1. 3 . 6 0 (3) (3) .............. (3) . 1 0 (3) . 2 1 (3) Area 2 5. 5 0. 5 (3) .. 5. 0 . 4 (3) _ 5. 5 . 2 (3) __ 2. 5 . l (a) .. 17.0 2. 0 3 6 0 7 155 ....................... 4. 0 . 2 (3) .............. 156 ....................... 4 20. 0+ .............. 5. 4 l 2 ’ 6. 0 . 3 . 5 0 58 . 8 0 (3) ______________ 11. 0 2. 0 l. 6 l 1 l 22. 0+ ______________ (3) ______________ 9.0 . 5 1. 8 . 2 18.0 2. 0 3. 2 .4 31.0 1. 0 (8) .............. 4 10. 0+ .............. 4. 0 1. 3 30. 0 3. 0 5. 4 l. 7 8. 1 . 1 1. 2 0 1 30. 0 3. 0 4. 2 2. 0 20. 0 2. 0 4. 6 1. 3 17.0 2. 0 4. 0 1. 2 8. 5 1. 5 (3) ______________ 18.0 2. 0 4. 6 1. 2 10. 6 . 5 . 8 . 4 19. 0 l. 0 1. O . 5 7. 5 . 5 1.2 . 1 18. 0 2. 0 2. 0 1. 1 176 ....................... (2) ______________ . 5 . 2 177 _______________________ 1. 4 0 (3) ______________ 1 Dead tree. Age based on cross dating. 2 Age not determined. 3 No exposed roots. 4 Dead tree. Minimum age equals total ring count. Systematic deviations from the average degrada- tional rate might be expected owing to the topographic inhomogeneity of the area. Figure 254 shows the esti- mated rates of degradation, in feet per 1,000 years, at each sampling point. These values were obtained graph- ically from the scatter plot of tree age and minimum 363 slope degradation (fig. 254). Through the point repre- senting a given specimen, a line was drawn to the —O.5- foot mark on the ordinate (the approximate initial depth of root development). This line intersects the vertical line representing 1,000 years of elapsed time at some value, D, of minimum slope degradation. The ver- tical distance between this point and the —O.5-foot point represents the degradation taking place in 1,000 years at each point. Most of the values are within the range from 1.0 to 1.4 feet, regardless of specimen loca- tion. The highest values are at points along the rocky crest and upper slope; there is some suggestion of a downslope decrease in degradational rate on the west- facing slope. However, because the estimated degrada- tional rates vary Widely from point to point and because these variations are not clearly related to topo- graphic position, the author has concluded that no sig- nificant trends in slope development can be inferred from these data. AREA 2 Area 2 is a strip extending from a stream channel in a narrow alluvial flat to the crest of the adjacent ridge. This 1-acre area is representative of the steep side slopes of major canyons incised into the Reed Dolomite. It is 1 mile east of Reed Flat in a canyon tributary to the South Fork of Birch Creek (fig. 255). The narrow alluvial flat is at an altitude of about 9,700 feet. The crest, 350 feet above the base of the slope, is the end of a long dolomite ridge. There are no shrubs and only a few herbaceous ground cover plants on the slope. The only other vegetation is an open stand of old stunted bristle- cone pines (fig. 256 and pl. 11). The canyon has a “V- in-V” cross profile in this area, and the lower slopes are about 5° steeper than those above. The slopes are nearly linear in cross profile except at the narrow rounded crest. Coarse soil forms a continuous mantle over the lower two-thirds of the slope. Bedrock is locally exposed on the upper slope and almost continuously along the crest. The surface consists mainly of large angular dolomite fragments (fig. 257) which have been sorted into rock streams or less regular elongate patches of contrasting textures. No measurements of particle size were made, but the surface material is noticeably coarser than in area 1 and is very unstable. The slope surface has moderate local relief. On the lower part of the slope, microrelief features apparently reflect differences in the thickness of the mantle rather than bedrock irregularities. Upslope terraces and down- slope hollows have developed adjacent to many of the older bristlecone pines (fig. 244). The large maximum depths of root exposure (table 4) are related to this extreme asymmetry. 364 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 4.0 1 ! 0 Tree living 0 Tree dead—age from cross dating . 3'0 _ 0“ Tree dead—minimum age from ring count 0 N o I" O . MINIMUM SLOPE DEGRADATION (D), IN FEET 0.0 —1.0 1.5 2.0 2.5 3.0 AGE (7'), IN THOUSANDS OF YEARS FIGURE 253,—Age and minimum slope degradation for area 1. Average age of 10 trees with no exposed roots is about 125 years. Line fitted by least squares gives average degradational rate of about 1.2 feet per 1,000 years; intercept with vertical axis at T=0 gives average depth of root development of about one-half foot. All trees on the slope and the alluvial flat were in— cluded in the mapping. From a total of 80 standing trees, 28 (including 2 0n the flat) were selected for de- tailed study by means of an overlay of randomly plot- ted points. A minimum—age estimate was obtained for 26 trees; they range from 80 to 3,100 years in age. Measurements of the maximum depth of root exposure, estimated local degradation, and estimated tree age are given in table 3. The relation of local slope degradation to tree age is shown in figure 258; six trees with a mean age of only 380 years show no root exposure, and two are incom— pletely dated. A straight line fitted to the rest of the plotted points indicates an average degradational rate on the slope of about 0.8 foot per 1,000 years and an average depth of root development of one-half a foot. The distribution of estimated values of local degrada- tional rates was plotted and mapped for areas 2 and 3, which lie on opposite slopes of the same canyon (fig. 259). Most of the values for the local degradational rate at points in area 2 ranged from 0.5 to 1.0 foot per 1,000 years. Neither the crest nor the lower part of the slope are well represented by points, partly reflecting the actual distribution of old trees and partly owing to the random sampling method used. (Compare with fig. 256.) A stratified sampling procedure would have been a more efl’ective method of study. Little evidence was found to suggest a systematic downslope change SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. AREA 1 .1.1 4; 1.2. .1.2.1.2 r0 o g '1.4 0-8 0.7. 1.4. 01-1 .111 1.3. 00-6 .0.6 1-1. .12 1.4' O 1.4 .13 1-4..1.1 ‘ .1.8 1.10.3006 +4 - 1/1 .1.5 E’ .1.4 O .0.9 .o.6 2.5. .1 100 FEET FIGURE 254.———Distribution of values of estimated local degradational rates, in feet per 1,000 years, in area 1. FIGURE 255.—-—-Valley tributary to the South Fork of Birch Creek. View is toward the south. Area 2 extends from channel in center of photo- graph up the light-colored slope to the ridge crest at left; area 3 is 365 FIGURE 256,—Upper part of area 2. View is toward the southeast. Low density of trees and absence of ground cover are typical of stands of old bristlecone pines. FIGURE 257.—Surflcia1 mantle in area 2. Scale shown by 6-inch rule. Beneath surface, angular dolomitic rubble has matrix of finer soil. on lower part of opposite slope. 366 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 30 I I I 0 Tree living 0 Tree dead—age from cross dating O——Tree dead—minimum age from ring count N o MINIMUM SLOPE DEGRADATION (D), IN FEET .0 3" o o _1.0 I | I I I I 0 0.5 1.0 15 2.0 2.5 3.0 3.5 AGE (7'), IN THOUSANDS OF YEARS FIGURE 258.—Age and minimum slope degradation for area 2. Average age of six trees with no exposed roots is about 375 years; data. for two trees on alluvial flat at base of slope are not included. Line fitted by least squares gives average degradational rate of about 0.8 foot per 1,000 years; intercept with vertical axis at T=0 gives average depth of root development of about one-half a foot. A / A! —4OU I 200 FEET 1 I DATUM ASSUMED -300 —200 SLOPE PROFILE —.100' 0 FIGURE 259.——Distrlbutlon of values of estimated local degradational rates, in feet per 1,000 years, in areas 2 and 3, and slope profile of area 2. SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. in degradational rate in the upper two-thirds of the area. Three specimens in the lower third give relatively low values of local degradational rates, perhaps indicat- ing that the rates have been lowest in the lower part of the area, above the slightly oversteepened slope at the base. AREA 3 Area 3 is distinctly different from those previously dis0ussed. Although it is nearly opposite area 2, on the southeast-facing slope of the same canyon, area 3 does not have the same smooth to irregularly hummocked appearance. Instead, it appears to be corrugated be- cause it is incised by steep straight drainage channels (fig. 260) in smoothly rounded miniature valleys. The channels head higher up on the slope, about 200 feet from the ridge crest. A map of a 2—acre area on the lower slope, with an average slope of 28°, is shown in figure 261. Bristlecone pines and a sparse cover of shrubs and smaller plants grow on the slope. Some of the older trees growing on low, rounded interfluvial ridges show deep and extensive root exposure (fig. 262). The root systems of six of these trees, which range from 1,100 to 1,900 years in age (table 5), were mapped in detail (fig. 263), and the depth of root exposure was measured at many points on each. The average and maximum depths of exposure and the estimated ages are given in table 5. FIGURE 260.——Part of area 3, viewed upslope. Shallow swales and rounded interfluvlal ridges give slope a corrugated appearance. Trail in foreground. 367 TABLE 5.—-A ge and root-exposure data for trees in area 3 Age data (centuries) Root exposure Specimen Number of Mean Maximum Age Uncertainty measure- (it) (it) ments 13. 5 0. 14 0. 9 1. 9 19. 3 .3 l4 1. 3 2. 5 11. 0 2. 0 9 . 8 1. 7 13. 3 . 5 16 1. 1 2. 3 13. 0 1. 0 17 1. 2 2. 4 17. 5 . 5 15 l. 2 2. 4 The exposed root systems generally parallel the pres- ent ground surface; thus, the valley-and-ridge topog- raphy must have existed as much as 1,900 years ago, when the root system of specimen 184 was developing. The symmetry of exposure, however, is different from that shown by trees on unchanneled slopes. The depth of exposure here varies radially from the base of the stem, reflecting the bilateral symmetry in the directions of movement of material away from the sloping miniature crest. The depth of exposure is least immediately up- slope from the stem, where a narrow terrace exists; it is greatest on the downslope side. The results are sum- marized graphically in figure 264. For each specimen, the maximum depth of exposure is about twice the average depth of exposure, reflecting the asymmetry dis- cussed above. The average depth of root expoSure is a fair measure of the local degradation of the crest in the vicinity of each tree. As shown in figure 264, a trend line through paried values of age and average depth of ex- posure for these trees is flatter than one originating at zero time and zero exposure. This trend suggests that the degradation of this slope has not proceeded uniform— ly with time. COMPARISON Measurements of root exposure, minimum slope de— gradation, and elapsed time obtained by the same meth- ods and based on comparable samples can be directly compared for areas 1 and 2. The results have previously been reduced to estimates of local degradational rates for each area (figs. 254,259). The frequency distribution of this measure of degradational rate in the two areas is shown in figure 265. The modal value for area 1 is clearly greater than that for area 2, corresponding to the average degradational rates (obtained by least- squares analysis) of 1.2 and 0.8 feet per 1,000 years, respectively. The rocky crestal areas and relatively short gentle slopes of area 1 are being degraded at a significantly greater rate than the long steep valley side slope rep- resented by area 2. Confidence intervals can be com- puted (Wallis and Roberts, 1956) for the regression lines fitted to data from each area. In figure 266, the 95- percent confidence bands for the lines do not overlap EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT EXPLANATION X .183 Location only Mapped in detail Bristlecone pine Planetable map by V. C‘ LaMarche and O. F. Huffman, August 1963 A 100 FEET AI 280’— CONTOUR INTERVAL 5 FEET V DATUM ASSUMED - TOPOGRAPHIC MAP ~ 240'— — _ m _ l [D 200— SLOPE PROFILE _ 160’— — 120'— — 2-20’ CROSS-SLOPE PROFILE 200' Lu 5 \ u. 0 0: n. 180’ FIGURE 261.—.Location of trees studied in area 3, and profiles of area. 3. SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. Fromm 262.—Specimen 184 in area 3. Depth of exposure of root in fore- ground is about 2.5 feet, shown by comparison with stadia rod. Tree, which is 1,900 years old, is on creSt of low ridge that extends down the main slope. except near their point of convergence, at low values of the variables. RELATION TO TOPOGRAPHY Differences in the long-term degradational rate from place to place within an area underlain by a homogen- eous formation are clearly related to differences in to- pography and reflect variations in the intensity of degradational processes. The local rate of degration is not solely dependent on the slope of the ground surface but is also related to the geomorphic setting of the point of observation. Comparison of data obtained in two areas showed that a gently sloping knoll may be de— graded more rapidly than a long steep slope (table 6). Grouping of many local estimates of degrational rates by slope and site classes (table 7) substantiates these results. In general, rocky ridge crests, upper slopes, and steeply inclined spurs are being degraded more rapidly than the long valley side slopes. Furthermore, where degradational rates in crestal areas apparently increase with increase in slope angle, the main slopes are de- graded at fairly constant rates that do not vary system- atically with slope angle. 369 TABLE 6.—Summary of degradational rates and topographic characteristics in three selected areas Number of Average 1 Degrada- Area specimens Sampling slope tional rate Topographic dated method (degrees) (it per character 1,000 yr) 1 43 Total ____________ 20 2 l. 2 Rocky knoll. 2 26 Random. . . _ . 38 3 . 8 Long smooth slope. 3 6 Selective ......... 28 4 1.1 Gullied slope. 1 Mean of measurements, vicinity each specimen. 3 From least-squares analysis of data from 33 trees. 3 From least-squares analysis of data from 15 trees. 4 Estimated from mean 1,000-year degradation. TABLE 7.—Degradational rates for slope classes in crestal areas and on main slopes [Rates estimated from mean values of age and root exposure for specimens listed in tables 1 and 2] Age (thou- Estimated Slope Number 01 sands of Exposure degradationel observations years) (It) rate (ft per 1,000 yr) Crests] Areas Gentle _____________________ 9 0. 65 0. 2 0. 3 Moderate __________________ 16 1. 00 1. 0 1. 0 Steep ....................... 13 1. 13 1. 5 1. 3 Very steep _________________ 2 1. 70 2.8 1. 6 Main slopes Steep ______________________ 13 0.87 0. 7 0. 8 Very steep _________________ 6 . 86 . 4 . 5 The dolomite areas being degraded most slowly, if at all, are the gentle, lower slopes of high ridges, illus- trated in figure 249. Direct evidence of degradational rates in this topographic setting cannot be cited, be- cause the maximum ages of trees at these sites are low. The trees apparently do not survive long enough to show significant root exposure, although several bristle- cone pines on gently slopes (<10‘°) have attained ages of about 1,000 years. Low rates of slope degradation can also be inferred from the development of a soil profile in the surficial mantle of these areas. Of special geomorphic interest is the evidence of the breakdown of bedrock in progressive and relatively rapid degradation of the ridge-crest areas. The crests of high ridges on Reed Dolomite range in character from angular and irregular to gently rounded and fairly smooth. Bedrock outcrops are abundant, and the soil is coarse and patchy. Small cliffs—steep bedrock faces as much as 10 feet high and 50 feet long—have formed on the upper slopes of most ridges and along the crests of steeply inclined spurs. The shape and the appearance of such clifl's are controlled by the spacing and orienta- tion of joints. Overturning of trees has accompanied EROSION AND SEDILMENTATION IN A SEMIARID ENVIRONMENT 370 .82» 2: no 26 no 336? 5 3:83 ES .m «93 E 895 H? «0 Soon 33.58153 auburn EBw wuoom . Eccaco Emobm 33:62am / \ 3396 3:3 Lo uofioucou twmoaxm |.:.|...l...l 3001 83:5 959.0 58 E 65896 «09. E 580 Eye: té “a 8:03 320 Emaw %o g 2 § MAW—OWE Z 30°). Because root exposure tends to be symmetrical in crestal areas, this result probably reflects real differences in degradational rates rather than the effect of slope angle in increasing the asymmetry of ex- posure. The long comparatively steep (30°40°) slopes that extend from the valley floors to the crests of the main ridges are apparently being degraded more slowly (0.4—0.8 ft. per 1,000 yr.) than the crestal areas (1.2 ft. per 1,000 yr.). Because such slopes make up a major part of the terrain underlain by the Reed Dolomite, the average degradational rate for the entire area must also be less than 1 foot per 1,000 years. The most accurate estimate of the long-term rate of slope degradation based on study of exposed roots is obtained through unbiased (total or random) sampling of a fairly large number of specimens within, a small area. Measurements of root exposure arelreduced to estimates of minimum slope degradation, by‘co‘rrecting for the local topographic changes induced by the pres- ence of the tree itself. Results of study of two con- trasting areas, using this approach, reinforce conclu- sions based on analysis of data from individual trees scattered over a large area. A rocky knoll, representa- tive of crestal and upper slope areas, has been degraded at the average rate of 1.2 feet per 1,000 years during the past 2,500 years. Study of trees in a long narrow strip extending from a canyon bottom to the adjacent ridge crest showed that it has been degraded at a rate of only 0.8 foot per 1,000 years in the same period. The extent of slope degradation indicated by the wide- spread exposure of root systems of trees in the White Mountains requires the movement of large volumes of rock debris from the slopes to the stream channels. The accumulation of material behind logs and the pro— 376 nounced damming effect of standing trees show that such movement does take place. Relatively rapid move— ment is suggested by the fact that the appearance of an obstruction is reflected in the microtopography within a few hundred years. ' Removal of the debris produced from individual slopes, and perhaps a significant amount of slope ero- sion, may take place at infrequent intervals. Historical evidence shows that cloudbursts in White Mountain watersheds have generated floods that appeared as mud- flows on alluvial fans flanking the range. Small debris lobes and alluvial fans and large bouldery deposits in channels within the study area illustrate that large volumes of coarse dolomitic debris can be transported in single flood events. However, from the straight-line variation of degradation with time in each of two areas and from evidence from scattered observations of grad- ual and progressive root exposure provided by many buttress roots, it appears that degradational rates have not fluctuated greatly within the past 3,000 years. The landscape seems to be roughly adjusted to the nearly steady transportation of the products of rock weather- ing on this scale in time, even though major erosional events may recur only infrequently. A summary of the kind and scope of processes that are inferred to be sig— nificant in the degradation of the Reed Dolomite terrane in the White Mountains is given in table 8. TABLE 8.—Qualitative summary of major transport mechanisms Process Frequency Scope Rate of Associated Associated movement features events Creep ....... Continuous. Ridge-crests Slow--. . Rock Diurnal tem- and stream. perature slopes. change and freeze-thaw cycles. Solifluction.- Seasonal. _ _ . ________________________ Patterned ................ ground. Erosion ..... Occasional. . Slopes and Rapid. . Debris piles. Snowmelt and channels. cloutgaurst runo . The magnitude of degradational rates in the White Mountains generally corresponds to the results of cal» culations based on indirect evidence. of degradational rates in comparable areas. From data on sediment and dissolved loads of 17 streams draining mountain basins in semiarid and subalpine regions, Corbel (1959) calcu- lated an average denudational rate of 0.3 mm per year, or about 1 foot per 1,000 years. Schumm (1963), in a recent review of rates of denudation of small drainage basins, suggested a maximum long-term denudational rate of 3 feet per 1,000 years in the early stages of the erosion cycle in semiarid areas underlain mostly by sedimentary rocks. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT REFERENCES Anderson, H. W., 1957, Relating sediment yield to watershed variables: Am. Geophys. Union Trans, v. 38, p. 921—924. Billings, W. D., and Thompson, J. H., 1957, Composition of a stand of old bristlecone pine in the White Mountains of California : Ecology, v. 38, p. 158—160. Corbel, J ., 1959, Vitesse de l’erosion: Zeitschr. Geomorphologie, v. 3, p. 1—28. Currey, D. R., 1965, An ancient bristlecone pine stand in eastern Nevada : Ecology, v. 46, p. 564—566. D’Ooge, C. L., 1955, Five years of weather observations on the * White Mountain Range, California: American Meteorol. Soc. Bull., v.36, p. 172—175. -Doug1ass, A. E., 1914, A method of estimating rainfall by the growth of trees, in Huntington, Ellsworth, The climatic factor as illustrated in arid America: Carnegie Inst. Wash- ington Pub. 192, p. 101—138. Esau, Katherine, 1953, Plant anatomy: New York, John Wiley & Sons, 733 p. a‘Fritts, H. C., 1966, Growth-rings of trees; their correlation with climate : Science, v. 154, p. 973—979. Garrels, R. M., Thompson, M. E., and Siever, R., 1960, Stability of some carbonates at 25°C and one atmosphere total pres- sure: Am. J our. Sci., v. 258, p. 402418. .Glock, W. S., 1937, Principles and methods of tree-ring analysis : Carnegie Inst. Washington Pub. 486, 100 p. Glymph, L. M., Jr., 1954, Studies of sediment yield from water- sheds: Internat. Assoc. Sci. Hydrol., Gen. Assembly, Rome 1954, v. 1, p. 178—191. Hueck, K., 1951, Eine ibiologische Methode zum messen der erodieren Tatigkeit des Windes und des Wassers: Deutsche Bot. Gesell. Ben, v. 64, p. 53-56. Judson, Sheldon, and Ritter, D. F., 1964, Rates of regional de- nudation in the United States: Jour. Geophys. Research, v. 69, p. 3395—3401. bKesseli, J. E., and Beaty, C. B., 1959, Desert flood conditions in the White Mountains of California and Nevada: US Army Quartermaster Research and Eng. Tech. Rept. EP—108, 107 p. Knopf, Adolf, 1918, A geologic reconnaissance of the Inyo Range and the eastern slope of the southern Sierra Nevada, Cali- fornia; U.S. Geol. Survey Prof. Paper 110, 130 p. BLaMarche, V. 0., Jr., 1963, Origin and geologic significance of buttress root-s of bristlecone pines, White Mountains, Cali- fornia, in Geological Survey Research 1963: US Geol. Sur- vey Prof. Paper 475—0, p. 0149—0150. 'Leopold, L. B., 1956, Land use and sediment yield, in Thomas, W. L., ed., The role of Man in changing the face of the Earth: Univ. Chicago Press, p. 6394347. , p Lustig, L. K., 1965, C‘lastic sedimentation in Deep Springs Valley, California: US Geol. Survey Prof. Paper 352—F., p. 131—192. Merriam, C. H., 1890, Results of a biological survey of the San Francisco Mountain Region and the desert of the Little Colorado, Arizona : N. Am. Fauna, v. 3, p. 1—113. Mooney, H. A., St. Andre, G., and Wright, R. D., 1962, Alpine and subalpine vegetation patterns in the White Mountains of California: Am. Midland Naturalist, v. 68, p. 257—273. Munns, E. N., 1938, The distribution of important forest trees of the United States: US Dept. Agriculture Misc. Pub. 27, 176 p. Nelson, C. A., 1962, Lower Cambrian-Precambrian succession, White—Inyo Mountains, California : Geol. Soc. America Bull., v. 73, p. 139—144. ma SLOPE DEGRADATION, WHITE MOUNTAINS, CALIF. Nelson, C. A., 1963, Preliminary geologic map of the Blanco Mountain quadrangle, Inyo and Mono Counties, California: US. Geol. Survey Mineral Inv. Field Studies Map MF—256. Norman, L. A., Jr., and Stewart, R. M., 1951, Mines and mineral resources of Inyo County: California J our. Mines and Geol- ogy, v. 47, 11.,17—223. ' Pace, Nello, 1963, Climatomgical 'datasummary' for the decade 1 Jan. 1953 to 31 Dec. 1962 from the Crooked Creek Labora- tory and the Barcroft Laboratory : Berkeley, California Univ. White Mountain Research Station, duplicated report, ,‘ 52 p. . . Rice, C. M., 1940, Dictionary of Geological Terms: Ann Arbor, Mich., Edwards Bros, 364 p. Ruhe, R. V., and Daniels, R. B., 1965, Landscape erosion— geologic and historic: Jour. Soil and Water Conserv., v. 20, p. 52—57. Schulman, Edmund, 1954, Longevity under adversity in conifers: Science, v. 119, p. 396—399. 1956, Dendroclimatic changes in semiarid America: Tucson, Univ. Ariz. Press, 142 p. 1958, Bristlecone pine, oldest known living thing: Natl. Geog. Mag, v. 113, p. 355—372. Schulman, Edmund, and Ferguson, 0. W., 1956, Millenia old pine trees sampled in 1954 and 1955, m Schulman, Edmund, Dendroclimatic changes in semiarid America, Tucson, Univ. Ariz. Press, p. 136—138. 377 Schumm, S. A., 1963, Disparity between present rates of orogeny and denudation: U.S. Geol. Survey Prof. Paper 454—H, 13 p. 1964, Seasonal variation of erosion rates and processes on hillslopes in western Colorado : Zeitschr. Geomorphologie, v. 5, p. 215—238. , Seybold, A., 1930, Uber die Blosslegung des Wurzlesystems durch aolisch‘e und fluviatile erosion: Deutsche Bot. Gesell. 2 Ber., v. 48, p. 335—341. 3' Stewart, John H., 1966, Precambrian and Lower Cambrian for- mations in the Last Chance Range area, Inyo County, Cali- fornia, in Cohee, G. V., and West, W. 8., Changes in, stratigraphic nomenclature by the US. Geological Survey, 1964: US. Geol. Survey Bull. 1224—A, p. A60—A70. Stout, B. B., 1956, Studies of the root systems of deciduous trees: Black-Rock Forest Bull. 15, Harvard Black Rock ' Forest, Cornwall on the Hudson, N.Y., and Cambridge, Mass, 45 p. Troll, Karl, 1958, Structure soils, solifluction, and frost climates of the Earth: US. Army Snow, Ice, and Permafrost Re— search Establishment, Hanover, N .H., Translation 43, 121 p. Washburn, A. L., 1956, Classification, of patterned ground and review of suggested origins: GeoLSoc. America Bull., v. 67, p. 823—866. Wallis, W. A., and Roberts, H. V., 1956, Statistics—~11 new ap- proach: Glencoe, 111., Free Press, 646 p. I .U.S. GOVERNMENT PRINTING OFFICE : I967 0—269-085 ‘. c, UNITED STATES DEPARTMENT OF THE INTERIOR PROFESSISLITIAA—l-LE PQPER 552—1 GEOLOGICAL SURVEY ' O 100 FEET L._*___I EXPLANATION BEDROCK EXPOSURES AND PARTICLE-SIZE MEASUREMENTS Bedrock exposure, in percent A Line of particle-size measurements on / ./ Measurements shown figure 2 5 I Planetabie map by V. C. LaMarche and O. F‘ Huffman, June 1963 / EXPLANATION U ‘KI Stern cross section Exposed roots at 4-ft height 118 / Specimen No. FaIIen tree Bristlecone pine 50 I CONTOUR INTERVAL 2 FEET DATUM ASSUMED 100 FEET I TOPOGRAPHIC MAP OF AREA 1, SHOWING STANDING AND FALLEN TREES AND EXPOSED ROOTS, WHITE MOUNTAINS, CALIFORNIA 269.085 0—67 (In pocket) UNITED STATES DEPARTMENT OF THE INTERIOR PROFESSIONAL PAPER 352—1 GEOLOGICAL SURVEY PLATE 11 TOPOGRAPHIC MAP OF AREA 2, SHOWING STANDING AND FALLEN TREES AND EXPOSED ROOTS, WHITE MOUNTAINS, CALIFORNIA Q5113 , R F9 7 DAY g.3€a" T Chemical Weathering, Soil Development, and Geochemical Fractionation in a Part (\ of the White Mountains, Mono and W Inyo Counties, California GEOLOGICAL SURVEY PROFESSIONAL PAPER 352—] Chemical Weathering, Soil Development, and Geochemical Fractionation in a Part of the White Mountains, Mono and Inyo Counties, California By DENIS E. MARCHAND EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT GEOLOGICAL SURVEY PROFESSIONAL PAPER 352—J UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON: 1974 UNITED STATES DEPARTMENT OF THE INTERIOR ROGERS C. B. MORTON, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Director Library of Congress catalog—card N0. 73-600361 For sale by the Superintendent of Documents, US. Government Printing Office \Vashington, DC. 20402 — Price $135 (paper cover) Stock Number 2401-02484 CONTENTS Page Page Abstract __________________________________________________ 379 Chemical weathering and soil development—Continued Introduction ______________________________________________ 379 Chemical weathering—Continued Study area ________________________________________________ 381 Changes in the solid phases—Continued Location and description —————————————————————————————— 381 Chemical changes from bedrock to soil __________ 400 Previous work in the region ____________________________ 382 Reed Dolomite ____________________________ 401 Climate —————————————————————————————————————————————— 382 Adamellite of Sage Hen Flat ______________ 403 Vegetation ———————————————————————————————————————————— 384 Changes in the liquid phase ________________________ 405 Geology —————————————————————————————————————————————— 385 Chemical composition of precipitation __________ 405 Chemical weathering and soil development __________________ 388 Chemical composition of soil water ____________ 406 Method of approach ____________________________________ 388 Soil pH ______________________________________ 406 Field methods ———————————————————————————————————————— 388 Exchangeable cations __________________________ 406 Laboratory methods —————————————————————————————————— 388 Chemical composition of spring waters __________ 407 Description of soils ____________________________________ 389 Waters related in part to the Clarification of terms ______________________________ 389 Reed Dolomite __________________________ 407 Reed Dolomite SOilS ———————————————————————————————— 390 Waters related largely to the adamellite of Adamellite soils of Sage Hen Flat __________________ 390 Sage Hen Flat __________________________ 408 Basalt soils ______________________________________ 391 Analysis of chemical changes __________________ 408 Andrews Mountain sandstone soils ________________ 391 Relative mobilities ________________________ 408 SOil contamination ———————————————————————————————————— 392 Stability with respect to solid Physical weathering __________________________________ 392 and gas phases __________________________ 409 Reed Dolomite ———————————————————————————————————— 392 Relation of soil water to precipitation Adamellite 0f Sage Hen Flat ______________________ 393 and spring waters ______________________ 411 che$323;iiiffnii‘itffff‘ff‘fi:::::::::::::::::::::: 33: Chemical fractionation in the tedaccia-scii—water-aiaat Changes in the solid phases ________________________ 395 system """""""""""""""""""""""""" 413 Mineralogical changes ________________________ 395 Summary and discussion __________________________________ 416 Reed Dolomite ____________________________ 395 Weathering of Reed Dolomite __________________________ 417 Adamellite 0f Sage Hen Flat —————————————— 396 Weathering of adamellite of Sage Hen Flat ______________ 417 Electron microprobe studies of adamellite General weathering relations __________________________ 418 313135;?ifff‘fl‘ffi‘ii::::::::::::::::::: 332 References cited —————————————————————————————————————————— 418 Feldspars ________________________________ 400 Supplemental information—Methods, reproducibility, Other minerals ____________________________ 400 and accuracy ____________________________________________ 422 ILLUSTRATIONS Page FIGURE 269. Diagram showing reversible, irreversible, and cyclical processes associated with weathering and erosion __________ 380 270. Map of east-central California and adjacent part of Nevada, indicating location of the study area ____________________ 381 271. Photograph showing part of the upland surface in the White Mountains ____________________________________________ 381 272. Map of generalized topography of a part of the southern White Mountains __________________________________________ 381 273. Photograph showing North Sage Hen Flat auxiliary weather station and dust trap __________________________________ 383 274- Graph showing 7 maximum and minimum temperature and relative evaporation rates at Crooked Creek Laboratory and five auxiliary weather stations _________________________________________________________________________ 384 275. Graph showing precipitation recorded at Crooked Creek Laboratory and at five auxiliary weather stations ____________ 384 276. Photograph showing vegetational contrast between dolomite outcrops and colluvium and sandstone coliluvium ________ 385 277. Generalized geologic map of the study area ______________________________________________________________________ 386 278. Photograph showing view north across Sage Hen Flat, showing degraded White Mountain erosion surface and outcrop areas of adamellite of Sage Hen Flat, Reed Dolomite, and Andrews Mountain sandstone ________________________ 387 279. Photograph showing view east from Sage Hen Flat, showing resistant basalt capping less resistant m‘etasediments and granitic intrusive rock ______________________________________________________________________________________ 387 III IV CONTENTS Page FIGURE 280. Photograph showing exposures of reworked rhyolitic ash interbedded with terrace alluvium along Crooked Creek ______ 387 281 —284. Maps showing: 281. Depth of soil and colluvium at planetable map site R—l, Reed Dolomite, 10,600 feet ________________________ 339 282. Depth of soil and colluvium at planetable map site R—2, Reed Dolomite, 11,200 feet ________________________ 390 283. Depth of soiland colluvium at planetable map site S—l, adamellite of Sage Hen Flat ______________________ 390 284. Depth of soil and colluvium at planetable map site B—1. on basalt _________________________________________ 391 285. Photograph showing Reed Dolomite soil pit _____________________________________________________ . ________________ 391 286. Photograph showing soil pit in‘adamellite of Sage Hen Flat ________________________________________________________ 391 287. Graph showing cumulative grain size frequency curves for dolomite soils and comparison with ranges in bedrock grain 9 size ________________________________________________________________________________________________________ 3 2 288. Diagram showing gravel-sand-silt and clay distribution for soil above the C horizon ________________________________ 392 289. Diagram showing sand-silt-clay distribution for soil above the C horizon ____________________________________________ 393 290. Photograph of weathered adamellite outcrop on Sage Hen Flat ____________________________________________________ 393 291. Graph showing cumulative grain size frequency curves for adamellite soils and comparison with ranges of bedrock grain size ________________________________________________________________________________________________________ 394 292. Photomicrographs showing fresh, crushed bedrock dolomite and etched, embayed, altered soil dolomite grain ____________ 395 293. Photomicrographs showing fresh, crushed bedrock talc and soil talc grains __________________________________________ 396 294. Diagram showing mineral weathering sequence in Reed Dolomite soils, in order of decreasing resistance ____________ 396 295. Photomicrograph of weathered grains in adamellite soil thin section ________________________________ a _______________ 397 296. Photomicrograph of soil biotite grains ____________-______________.' ________________________________________________ 397 297. Diagram showing mineral weathering sequence in adamellite soils of Sage Hen Flat in order of decreasing resistance __________________________________________________________________________________________________ 398 298 —301. Graphs showing: 298. Electron rmicroprobe traverses normal to (001) cleavage of two soil biotite grains __________________________ 399 299. Electron microprobe traverses parallel to (001) cleavage of two soil biotite grains __________________________ 399 300. Electron microprobe analyses of 10 parts of a single plagioclase grain ____________________________________ 400 301. Percentage chemical losses for the indicated Reed Dolomite and adamellite soils of Sage Hen Flat with re- spect to bedrock ___________________________________________________________________________________ 402 302. Photograph of rock saw chips showing weathering rind on spheroidally weathered dolomite analyzed by electron microprobe ________________________________________________________________________________________________ 403 303 -308. Graphs showing: 303. Electron microprobe transect showing Ca, Mg, and Fe variation across an alteration find in spheroidally weathered Reed Dolomite ___________________________________________________________________________ 403 304. Electron microprobe transect for Ca, Mg, and Mn across same dolomite weathering rindl ________________ 403 305. Mineral percentage changes with respect to bedrock in two size fractions of a relatively uncontaminated adamellite soil _____________________________________________________________________________________ 404 306. Relative mobilities for eight elements in spring waters related to adamellite _____________________________ 409 307. Stability diagram for Na silicates ______________________________________________________________________ 411 308. Stability diagram for K silicates ________________________________________________________ . ________________ 411 309. Compositional diagram showing progressive cation changes from precipitation water to soil and spring waters related to Reed Dolomite __________________________________________________________________________________________ 412 310. Compositional diagram showing progressive cation changes from precipitation water to soil and spring waters related to adamellite ______________________________________________________________________________________________ 412 311 —314.. Graphs showing: 311. Mg/Fe and Ca/Fe ratio comparisons of waters related to adamellite ______________________________________ 413 312. Silica-cation ratio comparisons of adamellite-related waters ______________________________________________ 413 313. Comparison of elemental and total concentrations in soil, soil water, colloidal exchange, and plants ________ 415 314. Fractionation of nine elements in the adamellite bedrock-soil-water-plant system ___________ . ________________ 416 315. Diagram of flowsheet for laboratory treatment of bedrock samples __________________________________________________ 422 316. Diagram of flowsheet for laboratory treatment of soil samples ______________________________________________________ 422 TABLE NH—A ooqo'acnyzoa TABLES Page . Summary of climatic means and extremes from Crooked Creek and Mount Barcroft Laboratories ______________________ 383 . Types of information obtained at Crooked Creek Laboratory and five auxiliary weather stations during the period June 1966 to August 1967 ____________________________________________________________________________________ 383 . Estimated precision of measurement for some analytical quantities discussed in this paper ____________________________ 389 Mineral weight percentages in five Reed Dolomite bedrock samples __________________________________________________ 395 . Mineral weight percentages in five adamellite bedrock samples from Sage Hen Flat __________________________________ 396 . Chemical compositions of 10 major minerals in the adamellite of Sage Hen Flat ______________________________________ 397 . Layer silicates in silt and clay fractions of some adamellite soils ____________________________________________________ 398 . Averaged electron microprobe analyses of fresh and weathered biotites ______________________________________________ 399 TABLE 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. CONTENTS V Page Individual electron microprobe analyses of altered and relatively fresh parts of soil biotites ____________________________ 399 Averaged electron microprobe analyses of feldspars from fresh adamellite bedrock, grus, and soils ______ i ________________ 400 Electron microprobe analyses of four minerals from fresh adamellite of Sage Hen Flat and from derived sioil ______________ 400 Chemical analyses of Reed Dolomite bedrock and soils‘ ___________-_____1 ____________________________________________ 401 Chemical compositions of minerals and glass in soils used to correct chemical analyses ________________ a _______________ 402 Absolute losses of chemical constituents from bedrock to soil, assuming constant Zr percentage ________________________ 402 Averaged values for X-ray fluorescence chemical analyses of adamellite bedrock and soils from Sage Hem Flat __________ 404 Adamellite sample 94——bedrock chemical analysis and adjustment of <2-mm-soil analysis for ash and local contamination by biotite and hornblende _____________________________________________________________________________________ 405 Chemical analyses of precipitation collected near Crooked Creek Station and comparison with analyses of snow from the east slope of the Sierra Nevada ____________________________________________________________ , ________________ 406 Chemical constituents in water saturation extracts of two groups of White Mountains soils ____________ i ________________ 406 Mean values for total exchangeable cations, percentage exchangeable cations, and pH in Reed Dolomirte and adamellite soils of Sage Hen Flat ________________________________________________________________________________________ 407 Field and laboratory analytical data for two natural waters associated in part with the Reed Dolomite ______________ 407 Field and laboratory analytical data for two natural waters associated with the adamellite of Sage Hen Flat __________ 408 Relative mobilities for four elements in Cottonwood Spring waters __________________________________________________ 409 Degree of saturation with respect to calcite and dolomite for four spring waters ______________________________________ 410 Partial pressures of carbon dioxide in four spring waters and a comparison with atmosphericlPCOlz _____________________ 410 Chemical composition of some major plant species on Reed Dolomite and adamellite of Sage Hen Flat __________________ 414 Fractionation by plants with respect to soil water for seven elements in adamellite and dolomite terr'anes ____________ 416 Reproducibility of X-ray fluorescence analyses _______________________________________________________ . _______________ 423 Comparison of analytical results for some adamellite and Reed Dolomite samples, as analyzed by X-ray fluorescence, wet chemistry, and flame photometry __________________________________________________________ . ________________ 423 Variation in measured parameters of four White Mountains soil types ________________________________________________ 424 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT CHEMICAL WEATHERING, SOIL DEVELOPMENT, AND GEOCHEMICAL F RACTIONATION IN A PART OF THE WHITE MOUNTAINS, MONO AND INYO COUNTIES, CALIFORNIA By DENIS E. MARCHAND ABSTRACT The White Mountain erosion surface truncates a variety of late Precambrian and Cambrian metasedimentary units and several Mesozoic plutonic bodies; it is overlain by Tertiary basalts. Present climatic gradients in the 30-square-mile study area are not pronounced, and thus weathering of diverse lithologies under a reasonably uniform semiarid subalpine environment may be studied here. The local vegetation, consisting primarily of sagebrush, bristlecone and limber pine, and associated perennial herbs, is relatively sparse and shows marked discontinuities across geologic contacts. Soils in the region are immature lithosols, usually less than 1 foot deep, but better developed soils may have formed prior to or during late Tertiary uplift, only to be stripped by subsequent erosion. A reworked rhyolitic late Pleistocene ash and local windblown fragments are abundant soil contaminants and complicate efforts to decipher mineralogical and chemical changes due to weathering. Rock weathering in this region appears to be primarily a function of mineral composition and of the density and degree of physical rock flaws which serve as avenues for penetrating fluids. Physical breakdown, which tends to precede chemical weathering, seems as closely related to lithologic features as to climate. Since erosion is aided by rapid weathering, easily disintegrated rock types such as course-grained carbonate and adamellite and fissile shale have eroded to topographic lows while slow-weathering fine-grained carbonate, quartzose sandstone, basalt, aplitic dikes, and mafic inclusions have resisted erosion. Reed Dolomite, a Precambrian unit, weathers by frost riving to polycrystalline, cleavage-bounded grains, except where the presence of thermally recrystallized rock has led to production of single- crystal fragments. Mineral weathering occurs in the sequence dolo- mite >> tremolite, epidote > talc, K-feldspar, biotite 2 apatite > quartz and ilmenite. Percentage chemical losses from bedrock to soil are Mg > Ca > Sr > Mn z Fe, for carbonate-constituent elements. Authigenic calcite is apparently precipitated in the soil during dry seasons and partially dissolved during wet periods. Spring waters related in part to the dolomite show relative mobilities of Mg > Ca > Fe > Mn. Ion-activity product computations indicate that these waters are undersaturated with respect to both calcite and dolomite. PCO2 exceeds atmospheric values by over an order of magnitude and those of adamellite-derived waters by several times. Dolomite soil water extracts appear closer in pH to the spring waters than to rain or snow, although the soil water composition is obviously quite different from either of these fluids. Chemical changes in both solid and liquid phases related to the dolomite appear to begin along with physical breakdown in the early stages of rock weathering. Adamellite of Sage Hen Flat weathers by frost riving along intergranular weaknesses, causing boulder exfoliation and accumulation of grus. This coarse material, largely unaltered except for minor Fe oxidation, later undergoes important chemical breakdown during its transformation to finer sized particles. Primary minerals weather in the sequence plagioclase (An25_30) > hornblende > biotite, epidote > microcline, plagioclase (An10_15), allanite > apatite, chlorite, magnetite > ilmenite, muscovite, quartz, sphene > zircon, resulting in percentage chemical losses from fresh ., rock to soil in the sequence Rb '> Na = K z Mg > Sr > Mn 2 Ca > Ba > Si > Al >> Fe > Ti.1 Kaolinite, and possibly some vermiculite, is forming from feldspar, biotite, and other silicates. Microprobe 9 analyses of biotite indicate losses in the sequence Ba > K > Mg > Fe > Si > A1; cations having eightfold to twelvefold coordination are most readily lost, then cations in octahedral coordination, and finally ions in sixfold and fourfold coordination. Microclines reveal losses of Na and K, and plagioclases show changes implying removal of Na > Ca > Si > A1. Water saturation extracts from adamellite soils are distinct in composition from rain and snow water and from spring water in adamellite terrane. Relative mobilities of dissolved constituents in adamellite ground waters indicate changes with regard to bedrock of Ca > Mg 2: Na >> Si z K 2 Mn z Fe > A1. Owing to plant extraction, K and Al mobilities may be lower than bedrock-to-soil losses would suggest. High Ca and Mg mobilities could result from solution of carbonate grains blown into the soils. Chemical equilibrium does not exist in the natural waters studied, but steady-state conditions are closely approximated in spring waters. Contact with weathering rock over a considerable space-time interval is apparently necessary for the achievement of a steady state, but this interval may decrease for mobile constituents in readily weathered minerals. The chemistry of spring waters related to all lithologies studied appears very sensitive to differences in source material. Nine elements released by weathering appear to apportion themselves between vegetation, colloidal exchange, and solution as follows: Na, soln >> exch > veg; Si, soln >> veg; Ca, exch > soln > veg; Mg, exch > soln = veg; K, veg >> soln > exch; Fe, Mn, P, veg >> soln; Al, veg >>> soln. INTRODUCTION Bedrock, soil, vegetation, natural waters, and the atmosphere are interrelated by a highly complex sys- tem of reactions and processes, some of which are shown in figure 269. To adequately understand such a system, a broad investigation is necessary, including study of the many related and potentially significant aspects. It 1. A question mark is used in conjunction with greater than or less than symbols in this paper to indicate uncertain position in the sequence. 379 380 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT Photosynthesis V E G E T A T | O N1 Transpiration A T M O S P H E R E FOREIGN Np LITHIC “’00 g? MATERIAL J2 Littrfall5 ”1, 3'3 :2 7: .4" a. ,0 c .9 .9 :41 s," 4* % § ‘5 5.9” LITTER 95.0, «f egg 4‘ 5‘ “’9, E 3 2 “° 2% E m ::~ E ° ° V/ I .E c E o Oxidation e E g E g Humus Adsorption1 g 0 Solution Litter5 E? < F———_i Dissolved _ _ Solution . sclids1 Authlgenlc SOIL SOIL ““5”" d —— . Adsorption‘ an minerals1 WATER1 GASES denudati‘m gases Lithic Weathering‘ : fragments - .9 Weathered 80mm" E E particles1 8 E g '0 L" a SOIILL2 STREAM WATER‘ 3'3 M a) 3’ ‘6 Physical . \o 50° I- weathering summon ($061)? :90 Solution Dissolved solids and gases B E D R O C K1 I Deep percofilation FIcrkE 269—80me reversible, irreversible, and cyclical processes associated with weathering and erosion. Footnotes refer to data given in this and other papers: PRECIPITATES SPRING WATEFi1 1Total or partial chem- ical analysis; 2Physical analysis; 3Rate evaluation; 4Physical anaIySIS and rate evaluation (data in Marchand, 1970); 5Rate evaluation (data in Marchand. 1971). is impossible, for example, to clearly comprehend the transformation of bedrock into soil by studying changes in the mineral phases alone, for weathering involves important alteration of liquid and gas phases in contact with the minerals as well. This paper constitutes the fourth in a series of vegetational, erosional, and weath- ering studies in the southern White Mountains of east— ern California (Marchand, 1970, 1971, 1973). The focus here is primarily on the chemical weathering of two widely distributed rock types, dolomite and adamellite, utilizing chemical and physical data from all of the five system components mentioned. Chemical analyses of bedrock, soils, plants, and natural waters permit esti- mates of the direction and magnitude of chemical frac- tionation in parts of the geochemical system described in figure 269. The upland surface of the White Mountains, truncat- ing a diversity of sedimentary and plutonic bodies and overlain locally by basalt, provides an excellent area in which to study the extent, sequence, and processes of weathering and soil development applied to many lithologies in a semiarid, subalpine environment. Quantitative chemical weathering studies have previ- ously been largely confined to tropical or humid tem- perate regions, for example, Mead (1915), Goldich (1938), Cady (1951), Sherman and Uehera (1956), Hay (1959), and Wolfenden (1965). In such areas weathering has proceeded to such an extent that initial trends are difficult or impossible to recognize. Soils in the White Mountains are immature and clearly illustrate some of the early stages of the weathering process. Gerald Osborn and Francis H. Brown assisted with sample collection and planetable mapping during the summer of 1966. J. Ross Wagner aided in preparing X-ray fluorescence pellets of bedrock and soil samples, conducting mineral separations, and in determining many physical and chemical soil parameters. Electron microprobe studies were conducted by Lary K. Burns. J oaquim Hampel determined Na and K in eight bedrock and soil samples by flame photometer. Barbara Lewis and James Clayton analyzed plant materials for inor- ganic constituents. This paper represents a major portion of a Ph. D. dissertation carried out at the University of California, Berkeley, and financially supported by the US. Geolog- ical Survey. Many members of the Survey, especially WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA 381 Ivan Barnes, John Hem, and Edward J. Helley, ren- dered indispensable aid and advice. The manuscript in its various stages of preparation was aided by the criti- cal reading of F. J. Kleinhampl, John D. Hem, Richard L. Hay, Clyde Wahrhaftig, Peter W. Birkeland, and R. R. Tidball. STUDY AREA LOCATION AND DESCRIPTION The White Mountains, often termed the northern Inyo Range in older publications, extend from Westgard Pass in eastern California north to Montgomery Peak (north of map shown in fig. 270) in western Nevada. White Mountain Peak (14,246 ft.) is the highest point in the range. The area investigated encompasses approximately 30 square miles along the crest of the southern White FIGURE 271.—Part of the upland surface in the White Mountains. View north across Big , , _ _ Prospector Meadow. Outcrops of Andrews Mountain Member of Campito Formation in Mountains, east of Blshop, Callf., 1n Mono and Inyo foreground; white tree-covered exposures in middle and far distance are Reed Dolomite, Counties (fig. 270). It is characterized by a gently un- dulating summit upland (fig. 271) sloping southward from 11,500 feet near Sheep Mountain to about 10,000 feet east of Bucks Peak (fig. 272). Summits such as EXPLANATION T Basalt Sediments 37°40' ‘ 3‘- ' ' 1 2 MILES E— Sheep \9 § White Mountain Mountaln goo 1: LawPeak \Q 90° 1 i E N ' Mount Barcroft l , Laboratorv‘ \ 1 Big Prospector Meadow 37° 30' — Count?Line :33: 70 Hill ’¢00 Flat Blanco Mountain MONO COUNTY INYO COUNTY MONOCOLJNTY _ lNYO COUNTY ’0 Lu LU \‘ \/ '000 fi a“; c: n: .. s“ Base of . $39“ /_‘96‘o Tmountaln ‘s‘ ‘9 : /_ 37°20, _ gescarpment _\ 900 v. a d3 . o: .., . o g . o _ ‘ W “P, — a “5% l - O a. E o 2,. ‘a__ VALLEY _ 1’9 0 «00009 0 1 z a... ”h ‘ ““I .. Q, '19 90 /\/~ .‘l ' mil 9 9 $9, “9% -, H d 9: /\/ ee 0-7 ".7 3.“ Roads Flat 0 "a fill“; A surfaced $09 0 1 2 3 4 5M1, ES ” “‘ 9’3“" ’1 Jeep 37°10' — I E: - FIGL‘RY 272.—Genera1ized topography (contours in feet above mean sea level) ofa part ofthe 1 18° 20' 1 13° 10' southern White Mountains. Posterosion surface deposits shown locally. FIGURE 270.—East-central California and adjacent part ofNevada, indicating location ofthe study area and of the maps shown in figures 272 and 277. 382 Blanco Mountain, County Line Hill, and Campito Mountain stand 500 feet or more above this rolling surface. The steep western escarpment of the range, not included within the study area, descends 6,000—7,000 feet in 6 or 7 miles to the floor of the Owens Valley. The eastern flank slopes much more gradually, but is incised by a series of canyons as much as 1,000 feet deep, which drain southwestward to Deep Springs Valley and east- ward to Fish Lake Valley. PREVIOUS WORK IN THE REGION The brief description of Spurr’s (1903, p. 206—212) and Knopf ’s (1918) reconnaissance of the northern Inyo Range and eastern Sierra Nevada, which included a section on the stratigraphy of the Inyo Range by Edwin Kirk, marked the first geological studies of the White Mountains. Anderson and Maxson (1935) and Maxson (1935) described the physiography and Precambrian stratigraphy, respectively, of the northern Inyo Range. Recent work in the White Mountains led to publications on Precambrian and Cambrian stratigraphy, paleon- tology, and boundary problems by Nelson and Perry (1955), Nelson (1962), Taylor (1966), and Cloud and Nelson (1966). Anderson’s (1937) arguments for the granitization in the formation of the Inyo batholith were challenged by Hall (1964) and Emerson (1966). McKee and Nash (1967) dated most of the plutonic bodies and several metamorphic units in the area by K/Ar methods, and Krauskopf (1968) discussed the complex intrusive history of the region. Cenozoic vol- canism and tectonics were considered by Taylor (1965), and Dalrymple (1963) dated Tertiary volcanic rocks on the northwest edge of Deep Springs Valley. The Blanco Mountain 15-minute quadrangle (Nelson, 1963, 1966) and the Mount Barcroft (Krauskopf, 1971) 15-minute quadrangles were mapped. Hall (1964), Pittman (1958), and Gallick (1964) mapped parts of the White Mountain Peak, Mount Barcroft, and Blanco Mountain quad- rangles. A general summary of the area’s geologic his- tory was provided by Nicholls (1965). Geomorphic processes in the White Mountains were investigated by Powell (1963) and Kesseli and Beaty (1959), who studied desert flood conditions, and by Beaty (1959, 1960), who discussed slope evolution and gullying. LaMarche (1967, 1968) estimated erosion rates on the Reed Dolomite from tree-ring data and described spheroidal weathering in the carbonate rocks of the area. Several members of the US. Geological Survey dis- cussed dissolved constituents in snow (Feth, Rogers, and Roberson, 1964) and ground water (Feth, Roberson, and Polzer, 1964) in the eastern California region. Barnes (1965) studied the geochemistry of Birch Creek, which drains the southernmost part of the area and is EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT similar in some respects to spring waters discussed here. Pace (1963) and Nicholls (1965) summarized climatic data collected at the Mount Barcroft and Crooked Creek Laboratories of the White Mountain Research Station for the decade 1953—62. CLIMATE The White Mountains lie in the rain shadow of the Sierra Nevada and are generally characterized by a cold, dry climate (BWk to BSk of Koppen notation) with precipitation occurring largely in the form of snow and in sudden summer storms. Mean annual precipitation is about 13 inches at Crooked Creek and mean annual temperature is approximately 35°F. Climatic data from the Crooked Creek and Mount Barcroft Laboratories (fig. 270) of the White Mountain Research Station are summarized in table 1. The Crooked Creek Station (10,150 ft.) is located in a valley at the approximate center of the study area; the Mt. Barcroft Station (12,470 ft.) lies on the upland surface about 5—6 miles to the northwest. Both stations are situated about 1 mile east of the range crest. Therefore climatic data from the two laboratories indicate differ- ences due to elevation and exposure. Precipitation, rela- tive humidity, and wind velocity are higher at Barcroft. Temperature and diurnal temperature change are greater at Crooked Creek, and a greater proportion of the total precipitation occurs as rainfall there. Both stations show principal wind directional maximums from the west and southwest and a lesser maximum from the north or northeast throughout the year. Major (1967, p. 119—122) evaluated climatic data at the two stations in terms of evapotranspiration and water bal- ance. Because pronounced geographical climatic gradients might produce similar trends in weathering intensities or soil development, it was desired to determine whether significant east-west or north-south climatic gradients exist today in the study area. For this reason five auxiliary weather stations (AWS) on the upland surface were maintained from June 1966 to August 1967 (fig. 273). The stations chosen were the Patriarch AWS, just south of the Patriarch picnic area; the Cam— pito AWS, on the south flank of Campito Mountain; North Sage Hen AWS, on the north edge of Sage Hen Flat, above Crooked Creek Station; the Basalt AWS, on basalt cappings south of Crooked Creek Canyon; and the South Sage Hen AWS, on the southeast flank of County Line Hill. (See fig. 277 for locations.) Wind measurements recorded at Crooked Creek Laboratory were obtained from anemometers near the North Sage Hen AWS. All AWS sites were situated on relatively level topographic highs at spacings of 11/2—3 miles. WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA 383 TABLE 1.—Summary of climatic means and extremes from Crooked Creek ( J 0,1 50 ft) and Mount Barcroft Laboratories (12,470 ft) for the indicated periods Crooked Creek Laboratory (1949—1965) Crooked Creek Mount Barcroft Laboratory Laboratory Jan Feb. Mar Apr. May June July Aug. Sept Oct. Nov Dec. 1949—65 1953—65 Highest maximum temperature _______ °F__ 52 51 52 62 69 76 76 74 72 76 61 57 76 72 Mean maximum temperature_ _°F__ 30.1 31.7 32.7 40.7 46.9 58.3 65.2 63.9 58.4 50.3 40.6 34.3 46.1 35.5 Mean temperature ____________ _°F,, 19.6 21.4 22.5 30.3 36.2 46.0 52.6 51.4 45.7 38.4 29.3 23.5 34.8 27.4 Mean minimum temperature_ _°F_- 9.1 11.0 12.2 19.9 25.6 33.8 40.1 38.9 33,1 26.5 18.1 12.7 23,4 19.7" Lowest minimum temperature_ _°F__ ,21 —21 —16 —10 2 12 22 17 10 —7 —13 ~19 ~21 —35 Mean diurnal differenm ______ °F__ 21.0 20.7 20.5 20.8 21.2 24.5 25.1 25.0 25.3 23.8 22.5 21.6 22.7 16.1 Largest diurnal difference _ _“F__ 41 46 47 42 45 38 42 41 49 58 49 40 58 44 Smallest diurnal difference _°F_, 1 4 5 3 2 9 10 8 6 6 2 3 1 2 Mean clays 33°F or above ____ 13.8 13.8 17.1 24.0 28.4 29.9 30.9 31.0 30.0 29.4 23.9 19.4 291.6 per yr 204.5 per yr Mean 8 a.m. relative humidity- (percent)__ 57.5 61.2 59.1 55.4 51.5 40.1 40. 6 41 4 44.8 45.5 52.5 55.1 50.7 55.1 Mean 8 a.m. barometer ..... 517.7 516.4 518.1 519.2 522.5 526.1 525.6 524.3 523.1 521.3 520.1 1523.8 1481.3 Mean snowfall ________ 18.1 17.0 16.8 12.9 3. . 1.2 5.1 11.1 13.3 114.4 per yr 164.4 per yr Mean snow depth _ 15.5 16.5 11.0 2.4 .1 .0 0 .0 .3 3.4 5.7 ‘4.7 17.2 Highest snow depth _ 74 94 62 47 12 3 1 12 30 50 94 73 Mean snow water _ 1.62 1.38 1.63 1.11 .34 .04 .02 .12 .48 .92 .96 10.14 per yr 14.87 per yr Mean rainfall _______ .00 .00 .00 .10 .06 1. 30 .74 .48 .04 .00 .04 2. 73 per yr 2.11 per yr Mean total precipitation- 1.62 1.38 1.63 1.21 .40 1. 34 .76 .60 .52 .92 1.001287 per yr 16. 98 per yr Mean maximum Wind ___________ ( 19.0 18.8 16.8 15.3 13.5 12. 3 12. 2 13.0 13.8 15.8 17.9 15 8 22.5 Maximum wind direction (percent): 8.3 7.7 4.9 7.6 7.3 3.0 3.4 1.7 4.1 5.0 7.8 6.6 5. 8 12.7 10.3 16.8 11.4 10.3 14.2 10.3 8.0 7.2 5.8 11.8 13.6 12.3 11. 0 5.6 2.3 3.6 2.2 2.7 1.4 4.8 4.1 6.1 5.4 3.8 3.0 2.4 3. 5 5.4 3.5 1.7 3.5 4.9 5.4 9.8 15.2 13.7 9.8 10.9 5.6 5.1 7. 5 5.0 2.9 2.9 1.9 1.7 4.2 5.0 14.5 11.1 13.3 9.1 6.5 7.4 6.5 7.9 26.1 15.7 23.5 29.9 25.1 32.7 25.1 33.9 29.1 27.3 24.2 23.4 26. 2 17.0 32.4 36.0 33.8 33.1 31.5 27.7 26.1 22.3 26.9 25.8 29.4 27.1 29. 6 39.9 Northwest ___________________________ 14.5 16.0 18.8 9.9 11.1 6.5 3.4 4.4 5.8 6.6 9.8 15.7 10. 3 9.8 11953—62. FIGURE 27 3.—North Sage Hen Flat auxiliary weather station and dust trap. Evaporimeter to left of pole, thermometer in white container to right of pole, dust trap on top. Rain gage to left and rear of stand. View west, showing parts of the upland surface, with the Sierra Nevada in distance. Four-inch-diameter sections of stovepipe, 21/2 feet long and sealed at the bottom, were used as rain gages. Lightweight oil was added to minimize surface evapora- tion. Taylor maximum-minimum window thermome- ters, suspended inside white 1/2-gallon ice cream containers, were used for temperature readings. Rela- tive evaporation rates were obtained with modified Piche evaporimeters (Waring and Hermann, 1966). Temperature and precipitation readings were taken at bimonthly intervals from June 20 through September 1966, at the end of October 1966, and at monthly inter- vals from June to August 1967. Evaporation readings were obtained only during the summer of 1966. Table 2 TABLE 2:—Types of information obtained at Crooked Creek Laboratory and five auxiliary weather stations during the period June 1966 to A ugust 1 967 Position Maximum- Evapo- Abbre— relative to Eleva- minimum ration Name viation center of area tion (it) temperature Precipitation rate Center ______ 10,150 No. _ 10,200 Yes West ________ 10,440 No, Sa e Hen ___ NSH Center ,,,,,, 10,660 Yes Sout Sage Hen ____ SSH South ______ 10,740 Yes. Patriarch ,,,,,, P orth ______ 1 1 ,250 Yes. summarizes types of data collected at each weather station. A plot of temperature and evaporation rates for the auxiliary weather stations and Crooked Creek is shown in figure 274. The Campito and North Sage Hen AWS sites show nearly the same maximum and minimum temperatures; the Basalt commonly has lower max- imums and higher minimums. A11 Crooked Creek temperatures fall below those obtained on the upland surface, often by as much as 10—15°F. Evaporation data are quite spotty owing to malfunctioning at several sites, but the Basalt AWS values tend to be highest, followed by North Sage Hen, South Sage Hen, and the Patriarch. The comparison suggests a slight trend to— ward increasing evaporation rates from west to east, possibly owing to decreasing elevation. Since tempera- tures do not appear to increase eastward, the increased evaporation may be the result of a lower relative humid- ity. A comparison of precipitation at the five AWS sites and Crooked Creek (fig. 275) indicates highest values at 384 n+ EVAPORATION LOSS IGRAMS PER DAY) 6‘ I -+ I M. 1 0+. x x Maximum 0 temperature 09 TEMPERATURE I°FI Minimum temperature '10 I I | I I ,_§_, I L— 7-1 7-1-66 8-1 9—1 10-1 11-1 -67 8-1 EXPLANATI 0N Crooked Creek Laboratory North Sage Hen auxiliary weather station Campito auxiliary weather station Basalt auxiliary weather station South Sage Hen auxiliary weather station Patriarch auxiliary weather station naux+o FIGURE 274.~Maximum and minimum temperature and relative evaporation rates at Crooked Creek Laboratory and five auxiliary weather stations in the study area. Crooked Creek, closely followed by the Patriarch AWS. All the other sites received much less precipitation and tend to follow the order North Sage Hen > Basalt > Campito > South Sage Hen. There is a general trend of increasing precipitation toward the north, and higher values were recorded in the valley of Crooked Creek than on the surrounding upland surface, probably be— cause the AWS sites were exposed to high winds which would tend to decrease catchment of precipitation and to deflate snow accumulations. Crooked Creek data for wind velocity and direction, barometric pressure, and to some extent, relative humidity, approximate present-day climatic conditions for most of the study area. Absolute precipitation values at Crooked Creek are probably much more reliable than those obtained at the auxiliary stations, but the fact that the latter are much lower should not be dismissed altogether, for they suggest that Crooked Creek values may be high owing to drifted snow blown from adjacent slopes. To obtain representative temperature estimates for nearby upland surface areas, Crooked Creek max- imum temperatures should be increased by about 10°F and minimum temperatures by about 5°F. In the north- ern part of the study area, precipitation and relative humidity are probably somewhat higher and tempera- tures slightly lower than further south owing primarily to differences in elevation. With this exception, present climatic gradients in the study area are not marked. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 4O EXPLANATION CC = Crooked Creek Laboratory P = Patriarch auxiliary weather station NSH = North Sage Hen auxiliary weather station B = Basalt auxiliary weather station C = Campito auxiliary weather station SSH = South Sage Hen auxiliary weather station 1 =6/28/66 to 10/29/66 2 =10/29/66 to 6/28/67 3 = 6/28/67 to 7/22/67 4 = 7/22/67 to 8/17/67 5 = 6/28/66 to 8/17/67 35— 30— 25* PRECIPITATION, IN INCHES N o I NSI- B SSH RAIN FALL SNOW TOTAL PRECIPITATION FI(.l‘RiZ 275.—Precipitation recorded at Crooked Creek Laboratory and at five auxiliary weather stations for the indicated periods. VEGETATION The flora of the study area includes the Bristlecone Pine Forest and Sagebrush Scrub plant community types of Munz and Keck (1959, p. 11—18). Toward lower elevations the vegetation passes into Pinyon Juniper Woodland, and at higher elevations into Alpine Fell Fields. Stands of bristlecone pine (Pinus aristata), limber pine (Pinus flexilis), and aspen (Populus trem- uloides) make up a discontinuous tree canopy along the WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA crest of the range below about 1 1,500 feet. A continuous shrub understory on noncarbonate soils is dominated by sagebrush (Artemisia tridentata and A. arbuscula) and to a lesser extent by rabbitbrush (Chrysothamnus viscidiflorus), mountain mahogany (Cercocarpus ledifolius and C. intricatus), creambush (Holodiscus microphyllus), mountain misery (Chamaebatiaria mil- lefolium), and currant (Ribes velutinum). Beneath this two-story canopy of trees and sagebrush scrub is a nearly ubiquitous but sparse cover of grasses and her- baceous perennials, mostly low to the ground and often inconspicuous. Most areas are dry, but moister conditions occur 10- cally along streams and near seeps and springs. Such sites are characterized by annual grasses as well as Carex spp. (sedges), Cirsium drummondii (thistle), Artemisia ludoviciana (sagebrush), Castilleja miniata (paintbrush), and several species of Penstemon (beard- tongue), Mimulus (monkey flower), and Oenothera (evening primrose). Many plant species in the White Mountains are partly controlled by substrate in their distribution (Marchand, 1973). Edaphic restriction is evident in all vegetational strata, from bristlecone and limber pine to the smallest perennials. The principal botanical discon- tinuities occur across carbonate-noncarbonate bound- aries (fig. 276): dolomites and limestones support simi- lar vegetation as do, with a few notable exceptions, noncarbonate lithologies. Vegetation on sandstone and shale, granitic bodies, and basalt shows subtle differ- ences which are sufficiently consistent to permit identification of three noncarbonate plant com- munities, each characteristic of a given bedrock type. FIGURE 276,—Vegetational contrast between dolomite outcrops and colluvium (foreground and right rear) and sandstone colluvium (darker colored). Sandstone is covered by sage- brush and numerous other perennials, dolomite by certain small perennial herbs and bristlecone pine. 385 GEOLOGY A complex mixture of sedimentary, igneous, and metamorphic rocks constitute the bedrock of the south- ern White Mountains (fig. 277). The oldest strata in the region are a slightly metamorphosed Precambrian to Early Cambrian succession of sandstones, shales, limestones, and dolomites. The Precambrian Wyman Formation, consisting primarily of sandstone and shale with minor but conspicuous lenses of white limestone, is the oldest of this group. Overlying the Wyman is the Reed Dolomite, a massive and very pure carbonate rock. The Deep Spring Formation, composed of alternating members of carbonate rocks (both limestone and dolo- mite) and clastic sediments (quartzitic sandstone and shale) overlies the Reed. The next youngest formation is the Campito, consisting of a lower quartzose sandstone member, the Andrews Mountain, and an upper shale member, the Montenegro. The Precambrian-Cambrian boundary, coinciding with the base of the Fallotaspis Olenellid biozone (Taylor, 1966), occurs within the upper part of the Andrews Mountain Member. The youngest stratigraphic unit of the group is the Poleta Formation, a somewhat shaly carbonate. This succession was tilted toward the west and then intruded by the adamellites of Sage Hen Flat (130—138 m.y. (million years)) and Cottonwood area (151-170 m.y.) during the late Mesozoic (McKee and Nash, 1967, p. 672). Early to middle Tertiary erosion resulted in the beveling ofa surface oflow relief across the region (fig. 278), although monadnocks of Reed Dolomite and An- drews Mountain sandstone stood well above the erosion surface during its development. Parts of this surface were subsequently covered by thin local sediment and extensive basaltic flows (figs. 272, 279), K/Ar dated by Dalrymple (1963) at 10.8 m.y. In terms of late Cenozoic structure, the White Moun- tains are an eastward-tilted horst which exhibits rela- tively greater uplift on the west, where a steeply dip- ping fault zone bounding the range has undergone as much as 10,000 feet of normal displacement. The east margin is also faulted along the northwest side of Deep Springs Valley, but individual displacements here are about 1,000 feet or less. The basalts and underlying erosion surface are tilted and faulted up from 6,000 feet at the northeast end of Deep Springs Valley to 10,800 feet at Bucks Peak, showing that tectonic activity oc- curred later than 10.8 m.y. ago. Bateman and Wahrhaf- tig (1966) presented convincing evidence for regional uplift of east-central California and western Nevada between about 9 and 3.5 m.y. before present and forma- tion of Owens Valley and other collapse structures along the axis of uplift during a period of downfaulting beginning about 2.5 m.y. ago. The “Waucoba Lake 386 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT Cottonwood Spring X '68 Poison Creek \ Spring X‘ o .54 o‘. “.0.“ Q. at.» 37° 30' Blanco MOUJIQL o f. C - o“ 9 ° ° 5 ’OOked“ -" ° . . . a . EXPLANATION Olivine basalt .47 Bedrock, soil-sample site 0 Auxiliary weather station W4 W4 JU RAS- TERTI- Inyo batholith E31 Poleta Formation SIC AND AFlY CRETA‘ CEOUS X Spring sampled O Soil-map location CAMBRIAN Campito Formation, Monte— negro Member Campito Formation, Andrews Mountain Member V 2 Deep Spring Formation PRECAMBFIIAN Reed Dolomite 1MILE Wyman Formation “new ‘ o a» r a . . . / o oo'o.\ ..a- “u 7 . q .Q FIGURE 277.—General geology and sample locations in the study area. Beds,” exposed on the upthrown side of the frontal fault east of Big Pine, possess a predominantly Sierran mineralogy and apparently represent deposition in an early Pleistocene pluvial lake prior to the downfaulting of Owens Valley. The uppermost of several rhyolitic tuffs in the lakebeds has given a K/Ar age of 2.3 m.y. (Hay, 1966, p. 20). Coarse clastic debris from the White Mountains overlies and intertongues with the upper part of the lakebeds, which now dip westward at 5—7°, indicating some uplift of the range or downwarping of Owens Valley during the latter part of lakebed deposi- tion, about 2.3 m.y. ago. The range-front faulting may have begun at this time, but cannot have preceded the lakebed deposition. Regional arching therefore appears to have preceded faulting by at least 1 million years in this area, although some uplift in the White Mountains may have occurred as recently as 2.3 rn.y. before pres- ent. Pleistocene glaciation along the eastern slope of the White Mountains north of the study area has incised cirques and U-shaped valleys into the upland surface and has left morainal and outwash deposits along the valley walls and floors. According to D. R. Powell (1967, oral commun.), recognizable glacial advances in the White Mountains appear to correlate with those of the eastern Sierra Nevada. Quaternary alluvium and ter- race deposits are present along most of the major streams draining the range, especially on the east flank where gradients are gentler. Interbedded with terrace alluvium along Crooked Creek and its tributaries is a WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, FIGURE 278.—View north across Sage Hen Flat, showing degraded White Mountain erosion surface and outcrop areas of adamellite of Sage Hen Flat (light colored, foreground and middle distance), Reed Dolomite (white, tree covered), and Andrews Mountain sandstone (dark, left distance). FIGL‘RI: 279,—View east from Sage Hen Flat, showing resistant basalt capping less resistant metasediments and granitic intrusive rock. CALIFORNIA 387 FIGURE 280.—Exposures of reworked rhyoli’tic ash interbedded with terrace alluvium along Crooked Creek. White layers are largely composed of pumiceous material. white, reworked pumiceous ash (fig. 280), rhyolitic in composition and as much as 0.6 feet thick. Although glass shards and rhyolitic mineral fragments occur in soils throughout the area, no primary deposits have been seen. On the basis of major elements, trace ele- ments, refractive indices, mineralogy, and physical ap- pearance, the ash has been correlated with a source at Mono Craters (probable) or Mono Glass Mountain (pos- sible), both located on the eastern Sierra Nevada slope south of Mono Lake (Marchand, 1968, 1970). Those parts of the White Mountain erosion surface which have not been totally destroyed have been low- ered several hundred feet to the level of Sage Hen Flat (fig. 278), Reed Flat (fig. 272), or similar areas of sub- dued, rolling topography by late Tertiary or Quaternary erosion. Frost heaving, solifiuction (including altiplan- ation), nivation, and other periglacial phenomena con- tinue to affect the range, becoming increasingly impor- tant toward higher elevations. Rillwash channels as much as 2 inches deep and debris accumulations behind trees, fallen limbs, bushes, and large boulders testify to the effectiveness of slope runoff processes, particularly on the relatively bare surfaces beneath bristlecone pine forest. Talus cones and stone stripes are frequent, par- ticularly in the Campito Formation and other sand- stones or shales. Although eastward draining canyons are gradually eroding headward into degraded rem- nants of the erosion surface, fairly extensive parts of the lowered surface are still relatively undissected and 388 often serve as a local or temporary base level for peri- glacial processes and slopewash. The effect of these cir- cumstances is the accumulation of colluvium (consist- ing of weathered and unweathered rock, mineral, and organic material) on gentle slopes and in topographic lows on these surfaces. To put present climatic data into perspective, it is necessary to consider the regional climatic and tectonic changes that have occurred in the area during the late Cenozoic. Any attempt at backward extrapolation of climatic conditions is extremely hazardous, especially in areas such as this where faulting and uplift have complicated an already complex series of fluctuations, but a few conclusions appear to be valid: (1) Cold, dry, climatic conditions such as those which currently characterize the area, and the more severe climates of glacial periods, are not conducive to rapid soil formation (for example, cf. Morrison, 1964, p. 114); (2) the White Mountains area was at a lower and presumably warmer elevation and probably did not lie in the Sierran rain shadow prior to regional uparching, between 9 and 3.5 m.y. ago; (3) considering the total duration of warm, wet, or warm and wet intervals which promote rapid weathering, it would appear likely that most of the total amount of soil formed on the White Mountain surface occurred prior to or during the regional uplift, perhaps augmented during extended interglacial periods. The immaturity of present soils indicates that most of the weathered debris has been stripped by erosion, but the extreme weathering of some soil grains may date from an earlier period of soil formation. CHEMICAL WEATHERING AND SOIL DEVELOPMENT Chemical weathering is so closely interrelated with physical weathering that it is sometimes difficult to distinguish their separate effects. Taken together these two processes are responsible, along with various or- ganic effects, for the development of soil. Because of the close ties between physical and chemical processes in the evolution of soils, both aspects are considered here, although primary emphasis is placed on chemical weathering. METHOD OF APPROACH The immature soils in this area formed on four wide- spread parent materials. Because spring waters closely related to sandstone and basalt were not available and because changes in the liquid phase are important con- siderations in this study, detailed discussion is re- stricted to the Reed Dolomite and adamellite of Sage Hen Flat, two commonly recurring lithologies for which closely related spring waters could be obtained. For the dolomite and adamellite the sequence and degree of physical weathering is evaluated, and problems of soil EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT contamination are discussed. Chemical weathering is approached from the perspectives of chemical and mineralogical changes in dolomite and adamellite and of chemical changes in liquid phases associated with each lithology. FIELD METHODS After the geologic and soil mapping of the area (figs. 277, 281—284), 84 sample sites, about 20 to each lithol- ogy, were chosen at relatively even spacings on ‘Reed Dolomite, adamellite of Sage Hen Flat, basalt, and sandstone of the Andrews Mountain Member of the Campito Formation. (See fig. 277 for site locations.) Topographic highs were selected for sample locations to minimize contamination by slopewash, mass wasting, and aeolian processes. Soil samples encompassing the entire profile above the C horizon were collected at every site and fresh bedrock samples were also obtained on dolomite and adamellite sites. Eight surface soil pH measurements (determined using pHydrion paper and a Truog indicator kit) were made at each site within a 7-foot radius of the soil pit. The dimensions of the pit were measured as closely as possible, its volume was calculated, and the soil samples were weighed (air dry) after collection to obtain bulk density figures for the whole soil. Representative profiles were described, and field impregnations were made at several sites for dolomite, adamellite, and basalt soils. Over a 14-month period from June 1966 to August 1967, samples of rain, snow, and of spring waters related to the adamellite and dolomite were collected. LABORATORY METHODS Samples of both dolomite and adamellite were ran- domly selected for detailed chemical and mineralogical study in the laboratory. Minerals were identified primarily by petrographic microscope, aided by reflecting microscope for opaque minerals. Composi- tions of glass fragments were determined by immersion oils, universal stage, and electron microprobe. The microprobe was also used to assess chemical alteration within weathered minerals. Estimates of major mineral percentages in fresh adamellite were determined from point counts of stained thin sections. Percentages of minor mineral phases in adamellite, all dolomite con- stituents, and all minerals in sand fractions of soils were obtained from line counts of grain mounts, after two heavy-liquid separations. Silt- and clay-sized frac- tions were analyzed qualitatively and quantitatively by X-ray diffraction. Layer silicates were identified follow- ing the methods of Warshaw and Roy (1961). Con- taminative rhyolitic glass in the silt range of soils was estimated Visually in oil immersion mounts. X-ray fluorescence was employed in the chemical analysis of bedrock and soil samples. Standard sieve and pipette techniques were utilized in size analyses of 26 dolomite WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA and adamellite soils. Other analytical methods used are discussed at appropriate places in the report. For a more detailed account of procedures, precision, reproducibil- ity, and comparisons of methods, the reader is referred to the section “Supplemental Information.” Table 3 summarizes the estimated precision of meas- urement for some of the analyzed quantities discussed in the text. All percentages in the text, tables, and figures are weight percent, unless otherwise noted. TABLE 3.—Estimated precision of measurement for some analytical quantities discussed in this paper [See also "Supplemental Information”] Estimated range of precision Analytical method (percent of given value) X-ray fluorescence ______________________ :6 Grain counting: Major minerals ______________________ :5 Minor minerals ____________________ i 10—15 X-ray diffraction (silts) __________________ :15 Size distribution: Sieve range ________________________ i2 Pipette range (silt and clay) __________ :10 Plant ash analyses ______________________ Analyses of water-saturation extracts NH4Ac extract analyses __________________ :20 (usually < 10) __-_ :25 :20 DESCRIPTION OF SOILS CLARIFICATION OF TERMS In the White Mountains, as in other areas of high relief, absolute distinctions between residual soil and weathered colluvium are essentially arbitrary. Perigla- cial processes, as well as mass wasting, slopewash, and aeolian transport, are actively moving loose material into topographic lows. Many definitions of soil imply formation by in situ weathering alone (for example, cf. Brewer, 1964, p. 7—8), apart from the effects of erosion and deposition. For purposes of this discussion, the term “soil” will be expanded from this usage to include some colluvial and windblown material. In the present con- text, "soil" encompasses all weathered or slightly weathered surficial debris other than talus, stone stripe boulders, slopewash fans, alluvium, and sand dunes and necessarily includes debris that has undergone some slope transport, albeit of relatively short distance in many cases. The White Mountains soils are almost exclusively poorly developed lithosols having general All/A12 /C /R profiles less than 13 inches deep. Color, structural, or textural B horizons are faintly visible in a few profiles, but such occurrences are rare. Calcium carbonate crusts can be found on all lithologies, but are abundant only on dolomite soils. An irregular surface layer of relatively large fragments (A11 horizon), brought to the surface by frost heaving, is a common feature of most profiles, independent of parent material. Beneath this stratum is a layer of finer grained weathered material (A12 hori- zon), usually overlying large partly weathered frag- 389 ments in a matrix of silt and sand (C). Fresh bedrock (R) is generally encountered within 1 or 2 feet of the top of the C. Planetable maps of topography and maximum soil depth for representative areas Within dolomite, adamel- lite, and basalt terranes are shown in figures 281—284. Maximum soil depth was taken as the greatest penetra- tion of a l-inch-diameter soil auger in repeated borings at a given location. These depths correspond to the ap- proximate top of fresh bedrock, except in deep profiles where lateral friction on the auger made further pene- tration impossible. The map isochores indicate a gen- eral thinning over topographic highs and steep slopes and thickening in local depressions, especially on Sage Hen Flat. Soil thickness appears to be somewhat erratic on the basalt and dolomite. Two dolomite areas at dif- ferent elevations show no appreciable differences in soil depth. In the Patriarch area, the thickest soils occur in minor interfluves, and the thinnest are found in the intermittent stream channels, suggesting that the lat- ter are presently sites of erosion or slow weathering. This relation contrasts with that of Sage Hen Flat, where surficial material is thickest along topographic lows. The application of lithologic names to soil units (that is, Reed Dolomite soil) represent informal usage only; EXPLANATION Maximum soil depth (feet) Interval 0.2 feet Relative topography; number is feet above arbitrary datum. Interval 5 feet 9 Planetable site 0 1 00 F E E T [“1 FIGURE 281.—Depth of soil and colluvium at planetable map site R—l, Reed Dolomite, 10,600 feet. 390 on heap D9 \ g 14- East Peak/S ()9 v/ / / / EXPLANATION #— Maximum soil depth (feet) Interval 0.2 feet Relative topography; number is feet above arbitrary datum, Interval 5 feet (9 Planetable site +2 Soil descriptions and imprege nations FIGL‘RP 282,‘Depth ofsoil and colluvium at planetable map site m2, Reed Dolomite, 11,200 feet. no attempt has been made to formally designate or map soil series in this area. REED DOLOMITE SOILS Average soil profile thickness to Reed Dolomite bed- rock, about 7 inches at sampled sites, is slightly greater than for other soil groups. White calcium carbonate surface crusts, usually discontinuous and less than one-half inch thick, are one of the most conspicuous features of the Reed profiles. Calcium carbonate is ab- sent in lower parts of the A horizon, but occurs at the A—C interface and on nearly all rock fragments within the profile. Local 0 horizons, consisting of undecom— posed and partially humified plant litter together with a few mineral grains occur sporadically beneath trees. Where 0 horizons are present, carbonate crusts are absent. Rock fragments concentrated near the surface in Au horizons (fig. 285) usually appear quite fresh, except for minor solution pitting. Rock in the C horizon, however, is often so thoroughly decomposed that it is easily broken with a trowel. A12 horizons are occasion- ally quite red owing to oxidation of Fe—bearing miner- als. Subangular blocky peds are common here, but structure in other horizons is generally single grained or massive. The pH may be below 7.0 in or near 0 horizons, but the normal range is between 7.5 and 8.2, EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT Z 0 100 FEET EXPLANATION Maximum soil depth (feet) Interval 0.4 feet Relative topography; number is feet above arbitrary datum. Interval 5 feet (9 Planetable site +2 Soil descriptions and impreg— nadons FIGURE 283—Depth of soil and colluvium at planetable map site S—l, adamellite of Sage Hen Flat. appreciably higher than in noncarbonate soils. The pH tends to be high at the surface, lower in the A horizon, and high again near and within the C horizon, causing precipitation of calcium carbonate in localized areas Jf high pH within the soils. ADAMELLITE SOILS OF SAGE HEN FLAT The soils on adamellite of Sage Hen Flat are charac- terized by a general lack of large rock fragments above the C horizon. The surface horizon is commonly a grus, consisting of granules and coarse sand, and is usually less than 1 inch thick (fig. 286). Calcium carbonate crusts, although occasionally present, are not common and may result from solution of inblown carbonate grains. 0 horizons occur in forested areas. Sandy loam A12 horizons are infrequently underlain by weak color, structural, or textural B horizons, in which iron oxida- tion, subangular blocky structure, and very weak clay WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA EXPLANATION N Maximum Soil depth (feet) Interval 0.2 feet ——_— Relative topography; number is feet above arbitrary datum. Interval 2 feet 6) Planetable site + 2 Soil descriptions and impreg— nations 1oo FEET 0 |__i FIGURE 284.—Depth of soil and colluvium at planetable map site B—l, on basalt. FIGURE 285.—Reed Dolomite soil pit showing accumulation of large fragments (An horizon) overlying finer grained A12 horizon. films may be Visible. The transition to the C horizon is often abrupt on topographic highs, but gradual in areas where the parent material is largely colluvial. Frag- ments in the C horizon may bear only thin weathering 391 ad .p FIGURE 286.—Soil pit in adamellite of Sage Hen Flat. Thin grus layer (Au horizon) overlies A12 horizon, which tends toward a color and textural B at some sites. Lighter colored C horizon is transitional downward into unweathered colluvium (not shown). rinds or may be quite thoroughly decomposed. Depth to bedrock averages about 6 inches on topographic highs. BASALT SOILS Basalt soil profiles average about 6 inches in depth and are characterized by good crumb structure in most A horizons and by a sharp A—C contact. Calcium carbonate crusts are rare, the surface layer generally consisting of coarse fragments or basaltic grus. Beneath this pavement is a vesicular gravelly loam to clay loam A12 horizon, as much as 4 inches thick, with subangular blocky peds. The lower part of the A horizon is often oxidized, but rarely is there other evidence of an incip- ient B horizon. Surface cracking and the degree of ag- gregation suggest higher clay contents than in other soils. Roots are uncommon, and O horizons are never present. The pH does not vary consistently with depth in the profiles. ANDREWS MOUNTAIN SANDSTONE SOILS Soils formed on the Andrews Mountain sandstones 392 and shales show a striking bimodal size distribution: Large rock fragments are concentrated at the surface near and within the C horizon, while sand, silt, and minor clay occur primarily in the A12 horizon. Calcium carbonate crusts are present locally but are not com- mon. The A12 horizon is usually single grained or mas- sive, although aggregation into crumbs may occur. The C horizon consists of colluvium or slightly weathered bedrock. As in other noncarbonate soils, the pH is gen- erally 7.0 or slightly below and does not change sys- tematically within the profile. SOIL CONTAMINATION Mineralogical analyses of five soils derived from the Reed Dolomite reveal from 3 to 73 percent of material, primarily volcanic glass, quartz, feldspars, biotite, and hornblende, that cannot be accounted for by weathering changes. The principal sources of this contamination are the pumiceous ash from the Mono Craters or Mono Glass Mountain (as much as 30 percent) and local windblown fragments (as much as 50 percent). The ex— tent, nature, and implications of the soil contaminants were discussed elsewhere (Marchand, 1970). Their abundance necessitates considerable correction of soil data at most sites, as explained under “Chemical Weathering,” and renders many soil samples uninter— pretable in terms of weathering. PHYSICAL WEATHERING REED DOLOMITE The Reed Dolomite is strongly jointed in one direc- tion. Cross jointing is common, breaking the crystalline rock into angular fragments of about 3 feet to less than one-half inch in maximum dimension. Bedrock grain size varies considerably depending on degree of recrystallization due to thermal metamorph- ism. Micrometer eyepiece examination of 16 bedrock samples indicate a mean diameter range of from 0.002 to 2 mm. Dolomite displaying spheroidal weathering has a bedrock grain size range from 0.125 to 2 mm, but angular dolomite is much finer grained, ranging from 0.002 to 0.125 mm. Cumulative grain size frequency curves for 10 dolo— mite soil samples (fig. 287) show strong bimodal dis- tributions indicative of immature weathering. The two modes are in fragments above 8 mm (jointed blocks) and in finer materials derived from the large blocks. For comparison, the range of bedrock grain size has been superimposed on the diagram. Triangular plots of dolomite, adamellite, and Andrews Mountain sand- stone soils in terms of gravel, sand, and silt and clay, and of sand, silt, and clay are shown in figures 288 and 289, respectively. Gravel percentages vary greatly owing to extremely large fragments whose presence or EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 100 90 80 . From spheroudally weathered areas » 70 60 50 30 WEIGHT PERCENT GREATER THAN SIZE SHOWN ( Bedrock ranges: Spheroidally Angularly 10 weathered weathered 0 | l | l | - -2 0 2 4 6 8 10 16 4 1 0.25 0.0625 0.0156 0.0036 mm GRAIN SIZE (0) FIGURE 287.—Cumulative grain size frequency curves for dolomite soils and comparison with ranges in bedrock grain size based on samples from both spheroidally and angularly weathered sites. EXPLANATlON . Reed Dolomite soil 9 Adamellite soil A Andrews Mountain sandstone soil GRAVEL r30 SILT AND CLAY .~_ ;/ V P ‘90 °5 0 SAN D FIGI'RF 288.—Gravel-sand-silt and clay distribution for soil above the C horizon. WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA CLAY P o EXPLANATION 60 . Reed Dolomite sail a Adamellite soil A Andrews Mountain sandstone soil 95> Sandy loam ‘fi SAND 9x90 (9% A) $65 9% SILT 6 FIGURE 289.—Sand-silt-clay distribution for soil above the C horizon. absence strongly influences the results. Relative amounts of sand, silt, and clay are more consistent. The pattern of physical rock weathering after jointing and surface exposure appears to be a function of the earlier thermal history of the rock. Carbonates (both limestones and dolomites) adjacent to the Sage Hen and Cottonwood plutons tend to weather relatively rapidly to produce spheroidal boulders, whereas slower weath- ering of angular blocks characterizes carbonate terrane away from plutonic contacts. Most dolomite soils show maximums in the 2—4 phi interval, but several samples derived from thermally metamorphosed and spheroi- dally weathered dolomite have modes in the 1—2 phi range. Dolomite joint blocks have thus undergone two types of breakdown: (1) Fine-grained dolomite blocks have remained angular, and slow physical disintegra- tion has produced smaller polycrystalline aggregates; (2) intergranular stresses within coarse-grained recrys- tallized dolomite developed during cooling from metamorphic conditions to subaerial temperatures have caused breakdown to single-crystal grains. This second weathering pattern was first noted and ex- plained by LaMarche (1967), who contended that spheroidal forms are created by preferential attack at corners and edges of the more susceptible rock and are maintained by the greater porosity of the weathering block exterior. The author’s observations corroborate those of LaMarche except that unrecrystallized dolo- mite commonly weathers to polycrystalline rather than monocrystalline grains as suggested by LaMarche, at least in the initial phase of physical weathering. Cleav- ages apparently afford an easier avenue of parting than the tightly held grain boundaries, except where con- traction has created intergranular weaknesses. 393 ADAMELLITE OF SAGE HEN FLAT Joints spaced as closely as 2—6 inches, but more com- monly several feet apart, occur throughout the Sage Hen Flat pluton and tend to fall into two categories: A principal group trending north-northeast and dipping steeply to the southeast and a secondary group striking northwest and inclined to both northeast and south- west. Joint blocks are larger than in the dolomites and consequently have a lesser effect on weathering proces- ses. Some joint openings and adjacent wallrocks have been silicified, sericitized, and tightly cemented with iron oxides. Such alteration is almost certainly deuteric or hydrothermal, as large crystals of mica have grown within feldspars adjacent to joint planes, in contrast with the fine-grained feldspar weathering products ob- served in soil thin sections. The altered joint planes and adjacent rock weather in positive relief with respect to adjacent adamellite (fig. 290), as do fine-grained mafic inclusions and aplitic dikes. The medium-grained hypidiomorphic-granular tex- ture of the adamellite locally tends toward porphyritic, but microscopic examination of 17 thin sections indi- cates that mean grain size never exceeds 4 mm. Average grain diameters commonly range from —0.5 to 3.5 phi. In figure 291, cumulative grain size frequency curves for 10 adamellite soil samples and two grus samples are shown, along with the superimposed average and ex- treme ranges of bedrock grain size. With few exceptions, the adamellite soils are unimodal, the maximum fre- quency occurring between 2 and 9 phi in most samples. This mode falls within the mean bedrock range, sug- FIGL'RE 290.~Weathered adamellite outcrop on Sage Hen Flat. Note relative resistance of altered joint planes. Pencil gives scale. 394 100 90 80 70 60 50 40 30 WEIGHT PERCENT GREATER THAN SIZE SHOWN 20 Bedrock ranges: Mean 10 Extreme 8 0.0039 mm -4 *2 0 2 4 1O 1 0.25 0.0625 6 0.0156 GRAIN SlZE (0) Flf-l'RE 291.—Cumulative grain size frequency curves for adamellite soils and comparison with ranges of bedrock grain size. gesting a tendency toward production of single—crystal grains by physical weathering. Two grus samples, how- ever, collected beneath exfoliating boulders, show modes of -1—0 phi. This material is commonly composed of polycrystalline aggregates and occasional large grains. On textural triangular diagrams (figs. 288, 289), adamellite soils of Sage Hen Flat plot closer to the sand component and farther from the gravel and silt compo- nents than do the Reed soils. Large joint blocks are rare in the adamellite soils, unlike the dolomite soils. Silt- sized grains are not common in fresh adamellite and have not been produced to any notable degree by weath- ering. A two-stage process of physical weathering is sug- gested by the data. An initial breakdown into polycrys— talline grus, occurring at the weathering surface of rock masses, is followed by a size reduction to monomineralic grains within the soil profile. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT MECHANISMS OF DISINTEGRATION Of the many mechanisms suggested for physical weathering and exfoliation, repetitive freezing of inter- stitial water seems the most applicable to relatively cold regions such as the White Mountains. At Crooked Creek Laboratory, temperatures descend below freez- ing on an average of about 220 days per year, and most of the precipitation occurs during the winter. A striking feature of physical disintegration in the study area is that the same process applied to several lithologies has resulted in differing soil grain size dis- tributions. This observation and the similarity of grus development from granitic rock here with that in much wetter and warmer areas suggest that rates and pro- ducts of physical weathering may be as much a function of inherent lithologic properties as of climatic condi- tions. For frost riving to be effective, water must be able to penetrate the rock. The fact that coarse-grained lithologies weather more rapidly than fine-grained rocks of the same composition lends credence to LaMarche’s (1967) hypothesis that contraction of large mineral grains during cooling from elevated tempera- tures may result in weakened intergranular bonds, af- fording access to water during weathering. Grain size may thus be one factor causing plutonic rocks to weather more rapidly than basalt: A larger grain will contract more than will a smaller grain having cooled the same amount, creating wider intergranular spaces or, if no separation occurs, greater intergranular ten- sional stresses. Fissility, grain adhesion and cohesion, and jointing may also be important factors in determin- ing degree of water penetration. It is not the author’s intent to imply that the contrac- tion mechanism is the only likely cause of grus forma- tion. Wahrhaftig (1965, p. 1178—1179) suggested that alteration of biotite to swelling clays may aid in split- ting apart intrusive rock into granular fragments. Evi- dence presented in the following pages indicates that expandable clays are not an important biotite weather- ing product in this area, but Helley (1966) showed that biotite can swell without being chemically altered. The absence of biotite or similar alterable minerals in the dolomite, however, necessitates an explanation other than presence of expandable minerals in this case. An interesting aspect of erosion in the White Moun- tains is the striking correlation between erosional resis- tances and the manner of physical weathering in vari- ous lithologies. Fine-grained carbonate, quartzose sandstone, basalt, and aplitic dikes and mafic inclusions within the plutons are relatively resistant to erosion. These lithologies tend to produce immature soils having bimodal size distributions, large fragments represent- ing one of the modes. Physical breakdown of such mat- WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA erial into fine particles is obviously very slow and the larger blocks are not easily eroded. Recrystallized car- bonates, plutonic rock, and shales are much less resis- tant to physical weathering. These lithologies tend to yield submature soils lacking strong bimodal size dis- tributions and containing sand-, silt-, and clay-sized particles which can be removed by rillwash and sheet- wash. Thus erosion is faster in areas underlain by coarse-grained or fissile rock types (Marchand, 1971). CHEMICAL WEATHERING CHANGES IN THE SOLID PHASES MINERALOGICAL CHANGES REED DOLOMITE The Reed Dolomite is close to a pure carbonate, con- taining less than 0.5 percent of minerals other than dolomite (table 4). Talc comprises as much as 0.46 per- cent, but is generally less than 0.1 percent. Ilmenite, quartz, K-feldspar, plagioclase, apatite, epidote, and garnet are often present in trace amounts, and tremo- lite occurs in two of the five samples studied quantita- tively. Although never observed in thin sections or oil immersion mounts, a few grains of biotite, hornblende, and magnetite were found in several grain mounts, possibly as a result of sieve contamination. Traces of chlorite and sericite were identified by X-ray diffraction of dolomite residues. TABLE 4.—Mineral weight percentages in five Reed Dolomite bedrock samples, based on line counts of grain mounts 54 68 47 65 127 Dolomite 99.68 99.97 99.96 99.93 Talc ___ .090 .0336 .0067 .065 ’I‘remollt ______ .0005 .0010 ________ Ilmenite _ .0214 0143 .0117 006 Magnetite___ Trace ____________________ Quartz _______ .0001 .0119 .0008 .001 K-feldspar ___ _ ______ .001 .0013 .00003 Plagioclase I__ 1 .20 .004 ______ .001 Biotite _____ i .005 .0003 .0013 ________ Homblende _ _ .0003 .0005 .0004 00012 Apatite _____ _ .0058 .0002 .0004 00012 Epidote ___ ______ .0001 .0004 Garnet -i_ _ .0001 .0001 .0009 Chlorite--- _ ____________ .0014 Zircon __________________ .0002 ____________ Oil immersion and grain mounts of 3—8 phi (125—4 Mm) size fractions of five dolomite soils were examined for degree of etching or alteration of primary noncon- taminative grains. Dolomite is strongly etched in all samples and in all size fractions (fig. 292), although effects are increasingly evident in finer particles. Tre- molite is consistently etched, especially at the grain terminations, its degree of weathering being generally comparable to or slightly less than that of hornblende in adamellite soils. The weathering of talc is variable: In some samples it appears to be quite fresh, but in others has produced extremely irregular grain boundaries (fig. 293). K-feldspar, plagioclase, and biotite show minor 395 0 0.1mm Lg FIGURE 292,—Dolomite. A, Fresh, crushed bedrock (#47). Plane light. B, Etched, embayed, altered soil grain (#68 soil, fine sand). Plane light. Note tendency of etching to follow cleavage directions, weathering alteration. Apatite grains are often rounded, showing effects of some solution. Quartz and ilmenite show little or no alteration, with the exception of very minor oxidation around the edges of a few ilmen- ite fragments. A general four-category weathering se- quence, based on the preceding observations, is shown in figure 294. Attempts to quantify the sequence through grain counts proved unsuccessful owing to the abundance of contaminants. 0 0.1mm |_____l Fmi'ki, 293.—Talc. A, Fresh crushed bedrock talc (#47). Plane light. B, Soil talc grains (#47 soil, crushed, 50—230 mesh). Grains to upper left are not strongly etched, but lower right grain shows large embayment. Plane light. The only mineral that is definitely authigenic in Reed soils is calcite, which appears in both silt and clay frac- tions of about half of the analyzed samples. Several soil samples contain 7 A, 10 A, and 14 A layer silicates in both silt and clay fractions. These clay minerals are not EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 1 2 Quartz, ilmenite > apatitez K-feldspar, plagioclase, biotite, talc> tremolite, epidote >> dolomite l Flume 294.—Mineral weathering sequence in Reed Dolomite soils, in order of decreasing resistance. Based on visual comparison of bedrock and noncontaminative soil grains. Vertical lines separate major divisions of resistance. believed to be authigenic; they possibly represent the chlorite and sericite originally present in the dolomite, plus aeolian additions from the surrounding terrain. ADAMELLITE or SAGE HEN FLAT The mineralogy of the Sage Hen Flat pluton is typical of an adamellite. Quartz, microcline, and plagioclase make up more than 90 percent of the rock; biotite ranges from 3 to 7 percent; hornblende, muscovite, chlorite, epidote, sphene, ilmenite, and magnetite account for as much as 4 percent of the total; and allanite, zircon, apatite, and tourmaline are commonly present in trace amounts. A few pyrite grains were noted under the reflecting microscope.2 Mineral percentages in five adamellite bedrock samples are given in table 5, and chemical compositions of principal minerals in the plu- ton are given in table 6. Minerals in the silt and fine-sand fractions of Sage Hen soils and in soil thin sections were examined for size reduction and weathering modifications of mar- gins, as compared with fresh bedrock grains. Plagio- clase of composition An25_30 in the cores of zoned crys- tals is generally the most strongly weathered mineral in adamellite soils (fig. 295); fine-grained alteration products, chiefly kaolinite, sometimes make up much of the original plagioclase. Hornblende grains frequently possess no unmodified boundaries, the terminations TABLE 5,—Mineral weight percentages in five adamellite bedrock sam— ples from Sage Hen Flat based on point counts of stained thin sections 86 101 94 97 105 Quartz ____________________ 29.2 30.2 23.9 30.8 39.7 K-felds ar G v 25 8 19 1 25.4 25 7 23 4 Plagioc ase _ 37 9 41 0 38.8 33 5 26 9 Biotite ______ _ 6 7 5 4 5.5 3 2 7 1 Chlorite A- _ 022 24 .42 102 32 Muscovite ,_ _ 05 3 2,1 1 2 1 62 Hornblende , 13 ,,,,,, .61 008 ______ Epidote ___, _ 057 .75 .76 2.8 .39 Allanite _,,i , 0004 .088 .18 Trace(?) .018 Sphene "a 0081 1.10 .32 003 .0854 Zircon ______ 0064 .008 0005 ______ .0037 Ilmenite , i 0036 .60 0095 .105 .42 Magnetite .0016 1.27 1.98 2,65 Trace. Apatite ____ _ .015 .025 .01 .003 .025 Tourmaline ____________ .03 ____________ 2Sulfides are localized alongjoints and near the margin ofthe pluton and are consequently not common in the samples examined. All nonmagnetic opaque minerals are therefore combined under the term “ilmenite” for the purposes of this discussion, The weathering of local sulfide concentrations. however. is of definite importance. as evidenced by significant sulfate values in spring waters draining the pluton. WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA 397 TABLE 6,—Chemical compositions of 10 major minerals in the adamellite of Sage Hen Flat [Methods of determination as indicated. All Fe assumed to be in ferric state] Electron microprobe (percent) Stgge Oil immersion Indices of refraction Annite An Ab Or (Fe biotite) Phlogopite K Ca Mg Fe Al Ti An Nx’ Ny Nz’ Plagioclase: \lO’lm 5 5 Mean ____________________________________________ 9 Microcline: Maximum ________________________________________ .g 0 H Minimum Mean ______________________________ Biotite: Maximum ,c ”a Minimum a“ _ - _____ Mean ,_ _ - Chlorite: Maximum __ Minimum Mean ___________________ Hornblende: Maximum _________________________________________________________________________________________ Minimum __V_ __ Mean Epidote: Maximum ______________________________________________________________________________ 1 Minimum _, _ __ , __.A _- 1 1 ean _____________________________________________________________________________ Allanite: Maximum Minimum Mean Sphene: Maximum __________________________________________________ Minimum _A__ _, Mean _______________________________________ Magnetite: Maximum Minimum ___ ____________ . Mean __________________________________________________________________________________________________ .01 Ilmenite: Maximum __________________ Minimum . ________________________________ 1.756 137—3— Wifsbéiiéal >230“ ___________________ N Ne’ o Apatite 1.630 ....... Tremolite 1.624 1.610 0 '0.1mm FI(;L'RE 295i—Weathered grains in adamellite soil thin section Core of plagioclase (lower center to right center) is considerably altered, Microcline (upper center and upper right) has parted along cleavages and shows slight alteration. Quartz (left) is fresh. Crossed nicols. often displaying considerable etching. Many grains . , - , show no such effects, however, suggesting that the more 0 0.1mm severely etched fragments, and perhaps the dolomite Ixi soil talc grains as well, may owe their modifications to an earlier and more intense period of weathering. Along edges and Cleavages biOtite (fig 296) is altering t0 SGC- FICL‘RL 296,—- S‘oil biotite grains (#86 soil, fine sand), showing both almred (upper left and ondary products, Wthh are discussed in the section upper center) and relatively unaltered margins (lower center). Plane light 398 “Biotites.” Biotites and epidotes have lost appreciable portions of their original margins, but irregular grain boundaries are only occasionally found on soil allanites. Microcline (fig. 295) and An10_15 plagioclase in the rims of normally zoned fragments commonly show some al— teration to clay. Chlorite and apatite grains are fre— quently rounded, but are never deeply etched. Ilmenite and magnetite, especially the latter, may show minor external oxidation, but such changes are not common. Sphene, quartz (fig. 295), and muscovite display very little evidence of etching or alterations, but irregular boundaries not seen in bedrock grains are occasionally present. Soil zircons show perfectly euhedral margins. A seven—category weathering sequence for minerals in adamellite soils, based on the preceding evidence, is given in figure 297. In any given sample, a mineral may deviate from the sequence by one category, either above or below, but in general the order is consistent in all the samples studied. As in the case of the dolomite soils, quantitative mineral weathering studies were compli- cated by contaminants and yielded no substantial in- formation not given by figure 297. In figure 305, min- eral percentage changes with respect to bedrock are shown for the sand and silt fractions of #94 soil, a relatively uncontaminated site. Qualitative data for phyllosilicates in the silt and clay fractions of some adamellite soils are summarized in table 7. Sericite, normal chlorite, and septechlorite are present in the silt fraction as well as in the clay and may be metamorphic or hydrothermal. Kaolinite and 1 2 sphene > quartz muscovnte ilmenite 3 4 5 6 7 apatite microcline b' hlorite > plagioclase An > Iome >hornblende >plagioclase An > c . , lUAIS epidote 25* magnetite allannte Zircon 30 Flm R} 297.—Mineral weathering sequence in adamellite soils ofSage Hen Flat, in order of decreasing resistance. Based on visual comparison of bedrock and noncontaminative soil grains. Vertical lines separate major divisions of resistance. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT vermiculite are confined to clay fractions. As kaolinite occurs only in the adamellite soils, it is believed to be authigenic, but the vermiculite occurs widely and may have been partially or entirely added by aeolian proces- ses. A small amount of montmorillonite is present in the clay fraction of one sample. The presence of several coexisting authigenic clay minerals, perhaps forming from different primary phases, is not inconsistent with the immature nature of the White Mountains soils. ELECTRON MICROPROBE STUDIES or ADAMELLITE MINERAL VVEATHERING Seven minerals from adamellite bedrock, grus, and soil on Sage Hen Flat were chemically analyzed by electron microprobe. Biotite showed major weathering alteration; microcline, plagioclase, allanite, and some magnetite gave detectable changes; ilmenite and sphene yielded no measurable variations because of weathering. B IOTITES Biotite analyses reveal progressively lowered con- tents of K, Ba, Mg, and Si and increases in Fe and Al, from bedrock to grus to soil (table 8). Biotite in grus is chemically similar to fresh bedrock biotite, but both of these differ markedly from soil biotite grains (1—3 phi), suggesting that little chemical weathering of biotites occurs prior to or during grus formation. Decreases in grus Mg with respect to bedrock is apparently due to primary solid solution substitution of Fe for Mg rather than weathering losses. Some grus biotites, however, are obviously breaking apart along cleavages prior to chemical alteration. In terms of lattice sites, the biotite cations in 8-12-fold coordination appear to be the most vulnerable to weathering attack, followed by cations in octahedral coordination. (The values of table 8 assume no ferric iron is present and that all ferrous iron is in octahedral coordination.) Ions in sixfold and fourfold coordination taken together show little evidence of change. Within the 8—12-fold coordination sites, Ba appears to be more TABLE 7.——Layer silicates in silt and clay fractions of some adamellite soils [X : present; 0 : absent; n.d. = not determined. Clay is < 2;! in soil fragments] Sample number ,,,,,,,,,,,,,,,,,, 42 86 88 91 92 94 95 97 98 99 101 102 106 Kaolinite: Silt ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, n.d. 0 n d n.d n d 0 n d O 0 0 0 0 Clay ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, x 0 X x X n.d. X X X n d n.d. n d n d Chlorite and (or) septechlorite: Silt ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, n.d. >< n.d. n.d. n.d. x n.d. X 0 X 0) X X Clay X X 0 0 0 n.d. 0 0 0 n.d. n.d. n.d n.d. Vermiculite: ilt ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, n.d. 0 n.d. n.d. n.d, O n.d. 0 0 0 0 0 0 Clay ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 0 X 0 X 0 n.d. X X 0 n.d. n d n d n.d. Montmorillonite: Silt ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, n.d. 0 n.d. n d n.d. O n.d. 0 0 0 0 0 0 Clay 0 0 0 0 O 0 0 0 0 0 n.d. 0 X(?l Sericite: Silt ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, n.d. X n d n d n.d. >< n.d. >< n.d. X X X X ‘ x x x X X x X x X X n d X (7) n.d, X X X X X X X X x x x x X X x X x X x X 0 n.d. X (?) WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA 399 TABLE 8.—Averaged electron microprobe analyses of fresh and weathered biotites [CN=coordination number with respect to oxygen] Total of Source Elemental percentages analyzed Total: CN=8 to 12 CN=8 CN=4 to 6 CN=8 to 12 CN=8 CN=4 to 6 of elements Annite + analyzed as oxides‘ phlogopite1 grains K Ba Mg Fe A1 Si (percent) molecules K-I-Ba Mg+Fe Si+A1 K/Ba Mg/Fe Si/A] Fresh bedrock _- 7.5 0.79 6.9 14.9 7.5 16.7 90.4 81.6 8.3 1 24.2 9.5 0.46 2.22 Grus __________ 7.3 .51 5.8 15.8 6.7 16.5 87.3 81.6 7.8 21 6 23.2 14 .37 2.46 Soil (1—3¢) ______ 5.8 .10 4.4 15.7 13.7 10.5 82.9 73.3 5.9 2 1 24.2 58 .28 .77 1Assuming all Fe in ferrous state easily mobilized than K, but the extremely low Ba \ , . . . . _ - - - - _ ' '—7_: — ’::.:::.?'¢:—:=:=l§17'\ 010 counting rate precludes any definite conclusmns. Mg 1n 0-5 flq/E/ ‘ —-’ meavage/ K/Fe ‘ the octahedral site is obviously more mobile than Fe, 0'4 7 GRAIN 1 firmrmrse 0‘05 and Si, largely located in the sixfold to fourfold site, is 03 2'0 ' 6', ‘ .30 ' .1... ' 1800'” much more readily removed than Al, which occurs in both octahedral and tetrahedral positions. Microprobe analyses are usually grouped and aver- aged to compare changes in more than three elements, because only three elements may be analyzed simul- taneously and because it is virtually impossible to re- turn to the same location on the same grain in a subse- quent analysis. Soil grains may often contain appreci- able unaltered parts, however, and primary composi- tional differences between grains may obscure changes due to weathering. Comparisons were made between weathered and fresh parts of the same soil grain to eliminate those problems. Ratios of K/Fe and Mg/Fe along five typical transects across soil biotites, both parallel and perpendicular to the (001) cleavage, are reproduced in figures 298 and 299. Several conclusions seem apparent from the data: (1) Lower K/Fe and Mg/Fe ratios invariably occur either along edges or cleavages, but every edge and cleavage does not show changed ratios, (2) traverses perpendicular to the basal cleavage are more variable than transects parallel to cleavages, and (3) K losses with respect to Fe are gener - ally more frequent and larger than those of Mg. Exami- nation of the original data (not shown) reveals a rela- tively constant Fe percentage across the grains, indicat- ing that Fe losses are comparatively minor. Decreased K/Fe and Mg/Fe ratios therefore reflect absolute losses, an observation consistent with the data of table 8. Some red-brown to yellow-brown Fe oxidation, though, was observed in soil biotites examined under the petro- graphic microscope, especially along grain margins. To gain further information concerning the end pro- duct of biotite weathering, analyses of fresh and altered portions of soil biotites for Si, Al, and Ba were also conducted (table 9), using a finely focused electron beam. .The Si/Al ratio drops from above 2 in fresh por- tions to less than 1, and even approaches 0 in the fine- grained alteration products that commonly occur adja- cent to edges and cleavages. Ba also shows somewhat lower values in the alteration products. The Mg losses °-6/I\IAA.9IFeI-. I .'_._.L 1/.—1~.7«I?:—.°“5 0‘5 Weathqrfl ,M/:_—’\’ -/._-\,/'—K/Fe 0'10 . \_/ \/ \l/ Cleavage GRAIN 1 ‘Second traverse 0,4 I I 1 I I I I I 005 a, w 20 1 ' ”- h 0.7 I , “3° I 800 20 3 ¥ Weatheredrg: _ E 0.6 Anya/7'" _..:-\. if)?” 4:73: , _ "3' 0'15 05 . . K/FeX\/ ~ Cleavage fight 0.10 0.4 / \ 0.05 \. / \Cleavage 0.3 — . 0.00 0.2 — 0.1 - l _ GRAIN 2 - First traverse 0 I I I I I I I I 40 120 200 280 360 ,um from left edge Grain 1 ,right edge FIGURE 298.—Electron microprobe traverses normal to (001) cleavage of two soil biotite grains. 0.2 K/Fe GRAIN 1 m I I I I I I I I I 0'0 w & 40 80 120 160 200 240 280 320 360 & y D? 0.6 s._gI_ _I_ KI I I I I I I E 0.5 -"— '\_“ :Mg/F /\ ;:::::— 0.1 0.4 ; K/Fe 0.3 Mg/Fe 0.0 0'2 GRAIN 2 0_1 I I I I I I I I I I I 0 40 80 120 160 200 240 280 320 360 400 440 480 ,um from top Lower edge FIGURE 299.~Electron microprobe traverses parallel to (001) cleavage of two soil biotite grains. TABLE 9.—Individual electron microprobe analyses, in weight percent, of altered and relatively fresh parts of soil biotites Number of Ba Si Al Si /Al Description analyses 0.10 16.70 7.38 2.26 1 .30 16.57 7.55 2.19 4 00 7.44 9.17 .81 1 .16 11.54 14.80 .78 1 .04 10.33 15.10 .68 1 .36 9.91 18.57 .53 1 .00 1.88 16.88 .11 1 .00 .86 22.43 .04 1 .00 .67 29.91 .01 1 400 and Si/Al ratios rule out vermiculite and montmorillon- ite as possible weathering products. The Si/Al ratios suggest that complete alteration to gibbsite may have occurred, but microprobe analyses of porous and prob- ably hydrous materials such as these must be regarded with some caution. Four Si/Al values fall between 0.5 and 1.0, and 1:1 clays (but not gibbsite) were detected in X-ray diffraction patterns of soil and clay fractions; so, the formation of kaolinite from biotite, if only as a temporary weathering product, seems a more likely possibility. Presence of colloidal, noncrystalline aluminum hydroxide may account for the low Si/Al ratios. FELDSPARS Microscopic evidence indicates that plagioclase, espe- cially the more calcic zones, and microcline are altering to clays. The soil feldspars analyzed by the microprobe are much less severely altered than feldspars observed in some soil thin sections and grain mounts; so, chemi- cal differences between averaged microprobe analyses of fresh bedrock grains and soil fragments (table 10) are not striking. TABLE 10.-—Averaged electron microprobe analyses, in weight percent, of feldspars from fresh adamellite bedrock, grus, and soils Number of Total: grains K Ba Na Ca Fe An + Ab + Or analyzed Microcline: Fres ________ 13.2 0.0 0.53 0.01 0.2 99.8 1 Grus ________ 13.0 .0 .57 .01 .2 99.2 4 Soil ,,,,,,,,,, 13.0 .0 .31 .00 .2 96.4 2 Plagioclase: Fresh ,,,,,,,, .1 .0 6.9 2.9 .3 99.4 4 Grus ________ .2 .0 6.4 3.3 .0 98.0 2 Soil ,,,,,,,,,, .1 .0 6.5 3.4 .1 98.4 8 The analyzed microclines appear to have undergone some chemical losses during weathering, as evidenced by the 3.4 percent decrease in molecular totals (An+Ab+Or), from bedrock to soil. Microcline grains from grus samples, like biotites (table 8), do not appear to be appreciably weathered. Microprobe transverses across K-feldspars from seven grus and soil samples (data not shown) suggest that lattice deficiencies, indi- cated by low molecular totals, most commonly occur near grain edges and that low Na and K values tend to be associated with these deficiencies. Primary crystal zonation, however, often obscures weathering losses. Zoning is so prominent in the plagioclases that weathering changes are largely masked, but the molecular totals of table 10 indicate that losses due to weathering have occurred and that sodium may be the most mobile constituent. Plagioclases from both grus and soil appear to show detectable deficiencies of molecular totals. A very small electron beam was used to analyze fine—grained plagioclase alteration products within soil grains for Si and Al (fig. 300). The marked EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 30 .‘\‘ ' . Silicon — 3-0 \ . o Si/Al w o . A Aluminum 2 25 _\°\Q_‘ ' . — 2.5 l. E o ./ \ U \o '1 20 — \ ' 2 o — uJ ‘ 4 l : _l o / u) < A '2 l5 — °§A\ 1.5 E /‘\‘/‘\ o A LU 10 L Partly weathered Strongly weathered — 1‘0 Fresh 5 Physical appearance of selected microprobe spot locations 0 5 FIGURE BOO—Electron microprobe analyses of 10 parts of a single plagioclase grain. decrease in Si/Al ratios from 2.7 in fresh portions to about 1—1.5 in altered areas suggests that kaolinite may be the principal weathering product. As weathering proceeds, Si percentages appear to decrease more rapidly than Al percentages increase, such that Si + Al steadily decreases. OTHER MINERALS Analyses of fresh and weathered allanite, magnetite, ilmenite, and sphene are compiled in table 11. The one soil allanite examined is lower in Ca and Al and higher in Fe percentage than fresh grains. Analytical results: from 35 magnetite and ilmenite grains in both bedrock and soil gave little indication of weathering changes, but a few magnetite grains showed some visible altera- tion and less Fe (apparently due to oxidation) near their margins. Soil and bedrock sphenes gave virtually iden- tical chemical analyses, and traverses across soil grains showed no significant variations. TABLE 11.—Electron microprobe analyses, in weight percent, offour minerals fromfresh adamellite of Sage Hen Flat and from derived SOll Number of Ca Mg Fe Al Ti analyses Allanite: Fresh __________________ 8.0 ,,__ 11.5 8.2 12 Soil ____________________ 6.5 __,, 12.9 6.7 ,, 1 Magnetite: Fresh 0.00 70.8 ”A. 0.13 13 Soil ,,,,,,,,,,,,,,,,,,,, .03 72.6 __,_ .25 10 Ilmenite: Fresh ,,,,,,,,,,,,,,,,,, _,,_ .05 32.2 A» 27.2 11 Soil ____________________ “A- .26 30.0 ”,4 29.4 1 Sphene: Fresh ,,,,,,,,,,,,,,,,,, 23.3 .0 1.2 .7 19.1 11 Soil ____________________ 23.8 ,,-_ ,-_, .9 19.0 2 CHEMICAL CHANGES FROM BEDROCK T0 SOIL Chemical analyses of bedrock and soil samples were conducted by X-ray fluorescence techniques to assess the degree and sequence of bulk chemical changes due to weathering. For the adamellite and its derived soils, the US. Geological Survey samples G—l, G—2, GSP—l, AGV—l, BCR—l, and W—l were used as standards. The US. Bureau of Standards’ Dolomite #88 and five Reed WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA bedrock and soil samples analyzed by Ken-ichiro Aoki, Tohoku University, Japan, were used as standards for the dolomite group. REED DOLOMITE Arithmetic means and standard deviations for analyses of dolomite bedrock, soil greater than 2 mm, soil less than 2 mm, and soil less than 62 ,um are given in table 12. Standard deviations refer to differences be- tween samples collected at different locations rather than to analytical precision (replicate analysis). The principal reason for collection of numerous and widely distributed bedrock and soil samples was to ob- tain an indication of chemical and mineralogical varia- bility within each parent material and soil, such that changes due to weathering could be clearly differen- tiated from apparent changes caused by sampling deficiencies. Unfortunately, careful study of soil mineralogy shows most of the Reed soil samples to be contaminated by both rhyolitic ash and local materials to the extent that corrections for contamination would be meaningless. The effects of external aeolian addition are also indicated by the large increases in Na and K in the finer size fractions of the soils. Only one sample for which the mineralogy was determined, #65 soil less than 2 mm, appears to be little affected by contamina- tion. Corrections for the minor chemical effects of the tuff and local contaminants in this sample were made using weighted averages of glass and contaminative mineral percentages3 in the sand and silt fractions (per- centage clay was less than 1 percent and was therefore neglected in the calculations). Chemical composition of minerals and glass used to correct the original soil analysis are given, together with sources of informa- tion, in table 13. Local contaminants in this sample are assumed to be derived from the.nearby Sage Hen Flat “The methods by which percentages of ash and aeolian contaminants were determined has been discussed elsewhere (Marchand, 1970) in some detail. 401 pluton. Soil analytical values from which corrections had been subtracted were recalculated to the previous oxide percentage total and converted back to elemental percentages. Total estimated contamination in #65 soil less than 2 mm is only 3.9 percent, and thus any errors due to inaccurate amounts of correction are relatively small. Absolute losses (weight percent in bedrock minus weight percent in soil) are shown for the five carbonate-component elements in mean soil greater than 2 mm (hereafter termed “soil gravel”) and #65 soil less than 2 mm (corrected for both ash and local contam- inants) on the left side of table 14. To eliminate differ- ential effects of heavy and light elements, the absolute losses are divided by atomic weight in the adjacent columns. Percentage losses with respect to bedrock, de- termined as weight percent in bedrock-weight percentage loss = percent 1n $011 weight percent in bedrock are shown in bar graph form in the left side of figure 301. Both calculations were made holding constant the per- centage of Zr, the most stable element during the weathering process for the samples studied here. Un- doubtedly Zr percentage does change between bedrock and soil, but this assumption permits relative compari- son of all elements investigated, the principal concern of this study. The soil gravel fraction is assumed to be little affected by contamination, and #65 soil less than 2 mm has been corrected for all external additions; so, these two soil analyses should be chemically represen- tative of Reed terrane. Elemental absolute losses from dolomite bedrock to soil follow the sequence Ca > Mg >>> Fe >> Mn >> Sr, a trend strongly controlled by bedrock composition. When atomic weight is taken into account, the position of Ca and Mg in the sequence converge or trade places, but other relations remain TABLE 12.—Chemical analyses, in weight percent except as noted, of Reed Dolomite bedrock and soils [Data are from X-ray fluorescence techniques except as noted. n.d.=not determined; S.D.=standard deviation. Soil analyses are not corrected for contamination] Bedrock Soil > 2 mm Soil < 2 mm Soil < 62 [1.111 Number of Number of Number of Number of SD. samples Mean SD. samples Mean SD. samples Mean SD. samples 0.07 7 0.80 0 10 2 13.24 5.84 6 22.01 1.00 5 .0050 6 .022 .002 2 .20 .11 6 .40 .02 5 5.3 4 13 3 2 104 51 5 246 31 5 .059 6 .38 .06 2 4.15 1.70 6 7.16 .28 5 .09 12 .36 .005 2 1.83 .62 11 3.13 .06 5 .020 12 .08 .01 2 .11 .031 11 .12 .02 5 .43 7 11.01 .16 2 5.67 1.57 6 2.21 .38 5 .93 12 22.54 .43 2 12.93 4.16 11 8.38 .88 5 .00 6 .00 .00 2 .028 .012 5 .06 .004 5 7 6 62 2 124 27 5 222 17 5 .003 2 .0550 .003 2 .755 .395 2 n.d. n.d. 0 .006 2 .0545 0055 2 1.281 .335 3 n.d. n.d. 0 2.00 8 1.10 50 2 25.8 11.7 11 51.0 4.0 5 0085 2 n.d. n.d. 111111 .0545 ______ 2 n.d. n.d. 0 ...... 498.63 1,1.“ 1““- 497.69 ___,__ __,___ 5101.39 ,____, ___,V_ 694.63 lFlame photometer analyses by Joaquin Hampel. 2Wet chemical analyses by Ken-ichiro Aoki. aC02 taken as stoichiometric. Iron computed as Fe20. 4Includes H20+ and H20— analyzed by Ken-ichiro Aoki for three samples, 5Includes H20+ and 1120— analyzed by Ken—ichiro Aoki for two samples, 8Does not include H20+ and H20—. 402 TABLE 13.—Chemical compositions, in weight percent except as noted, [Probe analyses supplemented by data from Deer, Howie, and Zussman (1952a, EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT of minerals and glass in soils used to correct chemical analyses b; 1963 a, b), as indicated, except analyses for volcanic glass and biotite, which are supplemented by data from Carmichael (1967, p. 50 and 52, respectively) ] Volcanic Plutonic (local) 0 a) 4’ g ‘ a = 3 E as a "3 E E T: ‘3 % ‘E a? g be ‘5 E :c: o N P‘ "‘ 33 53 o a E to Cu 2 m a: I m o a 3 E A .H 00 a: 5 a) i E g g a: '5 ”E m G“ :: ”1 ' a: E a g 2 a g 9. EA .1 a A 55 £2 a 2°19 8% EE 3:83 .23 .a ;:: - ~— ga an is ca $3 ‘a as ”a l _ “ cog” V b . , co , H; b; H; m I!) Q Hg no a- 1—1 05 2% 2% 2% 3w as as 2% 28 no; no: 9.5, 9-2 as; no, no, no: E H a: a H E V as E H SH 5 H a“ o: as" N a‘“ "’ a“ m w to W m m (D to 30.7 30.2 16.8 30.1 29.6 16.3 22.5 17.4 12.0 ____________ 3.2 .01 2.4 .6 .05 .53 ‘ibfé‘ ‘16]? s. .0 .6 4. "‘ii' ”'Ti’ 6', .1 11.9 "'11— :2: 3.0 110.9 ‘Analyzed by electron microprobe. TABLE 14.——A bsolute losses, in weight percent, of chemical constituents from bedrock to soil, assuming constant Zr percentage {Figures in parentheses under mean dolomite column are losses expressed as percentages of #65 bedrock-to-soil losses for the given elements] Dolomite Adamellite #94 bedrock minus #94 soil < 2 mm1 #65 bedrock minus #65 soil < 2 mm‘ Mean dolomite bedrock minus mean dolomite soil > 2 mm Element Loss/ Loss/ Loss/ Ab501ute atomic Absolute atomic Absolute atomic loss weight loss weight loss weight ______________________ ~16.67 —0.5934 _________ ~4.33 <160 _________ -2.12 —.0542 ______________________ —1.77 *.0770 — 008(67) <98 r017 —10i25(57) <92 7.023 7696(65) —.43 <01 —.09 4.002 —.07 #0005 #028 4.00051 <05 7.0009 7.0038 #000043 —.0368 #000420 ,,,,,,,,,,,,,,,,,, #0041 4.000047 ‘Corrected for both ash and local contaminants, essentially unchanged. Percentage losses, a more significant indication of weathering mobility, occur in 9 the sequence Mg > Ca > Sr > MnéFe. In the percentage loss calculation, the atomic weight factor is automati- cally taken into account. Loss of Mg relative to Ca, as indicated by the percentage losses, is in agreement with the presence of secondary calcium carbonate in dolo- mite soils. Comparisons of both absolute and percentage losses indicate that, of the five elements shown (with the possible exception of Mn), well over half of the total elemental losses in bedrock to soil less than 2 mm occur during the transformation of bedrock into soil gravel. WEIGHT PERCENT IN BEDROCK ean soil >2mm WEIGHT PERCENTIN BEDROCK—WEIGHT PERCENT IN SOIL 00 X1 Mg Ca Sr Mn Fe Rb Na K Mg Sr Mn Ca Ba Si Al Fe Ti PERCEN‘II'AGE LOSS indicated Reed Dolomite and adamellite 1Corrected FIGURE 301 .—Percentage chemical losses for the soils of Sage Hen Flat with respect to bedrock, assuming constant Zr percentage. for both ash and local contamination. Obviously chemical breakdown of the dolomite begins almost simultaneously with physical disintegration in the early stages of the carbonate weathering process. Sections cut through spheroidally weathered dolo- mite fragments occurring at the surface or within the soil profile commonly show an abrupt change from fresh light cream to pink crystalline dolomite outward into a WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA porous gray or white concentrically layered weathering rind (fig. 302). Under the petrographic microscope, al- ternating layers of fresh dolomite and a fine-grained alteration product are evident within the rind. Electron microprobe analyses of one of the rinds for Ca, Mg, Mn, and Fe and microprobe transects from unaltered dolo- mite into the rind (figs. 303, 304) indicate that the weathering product is a very pure calcite, as shown in the following table. FIGURE 302.-——Rock saw chips showing weathering rind on spheroidally weathered dolomite (#68) analyzed by electron microprobe. Rind is about 0.2 inch thick. <———WEATHERED FllND——>+—FRESH DOLOMITE—> Fresh grains Boundary . 3000 i i i i _i — -\ ‘/- .—.’ v, 2500 '_' — C8 2000 — ' — Ca . ,. ‘ 1 .............. ~.—--—- I | l I l 1500 -\ /"-\,/'\,-/'\_,-—'_ 300 - Mg 250 — COUNTS PER SECOND 100 50 /'\ .—--\ ’._.— ............... \_—- .—.’ 'x. l l ' . l . '\ ..... /‘\. I I . 1 Horizontal distances not to scale, total width approximately 1cm FIGURE 303.—E1ectron microprobe transect showing Ca, Mg, ano .“e variation across an alteration rind in spheroidally weathered Reed Dolomite (#68). 403 CaO MgO MnO FeO Fresh dolomite __________ percent--__ 30.2 20.6 0.3 0.2 Alteration product ______ percent____ 54.1 .3 .2 .1 Leaching of Mg and deposition of calcite are undoubt- edly responsible for the development of carbonate hori- zons and other calcium carbonate accumulations in dolomite soils. Fe and Mn are more variable within the rind, but appear to occur in approximately equal con- centrations in both dolomite and calcite. ADAMELLITE or SAGE HEN FLAT Mean analytical values for bedrock, grus, soil less than 2 mm, and soil less than 2 Mm are given in table 15. The grus values are generally comparable to those of the bedrock, and where they differ, the direction of change suggests primary variability rather than weathering. Therefore, in the third column of table 15, means weighted on the basis of number of samples have been calculated for both bedrock and grus and were used as the bedrock values in all further computations. As in the case of the Reed, almost all adamellite soils are greatly contaminated by both rhyolitic ash and local debris. The first step in preparing the soil analyses for interpretation was to correct for the contaminating ash. After ash correction, only #94 < 2 mm gave a reasona- ble sequence of bedrock-to-soil mineralogical changes, on the basis of grain counts (fig. 305), suggesting ab- EWEATHERED RIND FRESH DOLOMITE—> Fresh grains Boundary l_—‘_I V V i 3000 — '_"-.\/—\/ — 2500 ' _ 2000 — _ Ca Ca ..-____.— 1500 D . g 300 — Mg \./ \ / U m to CC w 250 — D. U) l— E O 200 _ o 150 _ Mn 100 _ 50 _ , M l\. \. .\_\__~____./.'_._,_M_n.__.\__._ 0 l | ‘/ \ — "" — I Horizontal distances not to scale, total width approximately lcm FIGURE 304.—Electron microprobe transect for Ca, Mg, and Mn across dolomite weathering rind (#68). 4:04 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT TABLE 15.—Averaged values for X-ray fluorescence chemical analyses, in weight percent except as noted, of adamellite bedrock and soils of Sage Hen Flat [S.D.=standard deviation; n.d.:not determined. Soil analyses are uncorrected for contamination] "Bedrock” (= weighted average of bedrock and Bedrock Grus grus; see text) Soil < 2 mm Soil < 62 pm Mean SD. Mean SD. Mean SD. Mean SD. Mean SD. Si- 32.39 0.98 31.16 31.79 1.08 31.51 0.61 28.21 0.85 Ti"- __ .24 .04 .23 .24 .04 .32 .05 .51 .03 Zr __ppm__ 164 32 123 152 27 217 49 517 65 Al _______ _,,__ _, 8.36 141 8.94 8.53 .44 8.03 55 8.38 .22 2.16 .56 1.55 1.99 41 2.77 62 4.53 .50 .06 .02 .05 .06 03 .08 02 .12 .01 .46 .18 .31 . 1.32 22 1.44 .29 1.45 2.24 12 .12 .02 .15 .08 01 600 111 3 321 21 2.70 .24 2.53 1.57 .10 3.20 .45 4.41 2.97 .06 53.9 12.9 65.3 67.3 1.9 1.048 ______ n.d. n.d. ,"1" lnone ______ n.d. -, n.d. no“. 1.053 ______ n.d. n.d. ,H-.. Oxide total‘1 ______________ 100.20 ______ 97.73 ,,,,,, 99.32 777777 100.47 ______ 94.89 ...... Number of total analyse _ ___________________________ 10 ______ 4 AAAAAA 14 ,,,,,, 6 ______ 5 ______ Number of partial analy es __ _, ,,,,,,,,,,,,, 8 ,,,,,, 0 ______ 3 ______ 5 ,,,,,, 2 ______ 1Based on one wet chemical analysis by Ken»ichiro Aoki. 2Based on two wet chemical analyses by Ken-ichiro Aoki. 105 I I I I OI 10" — — o o X j— z 5 t”) 0 CE 2 A E 8 1o3 9 o "H _ :20 EXPLANATION I E OSand x Silt U I— . . a 2 Al Allanlte H g L” A Apatite 0 x3 O 1 II 2 B Biome _ _, 1o — — E c Chlorite 5M9 x0 0 . o u) P- E Epldote g g H Hornblende }— E I llmenite E E M Microcline \ . g ¥ 101 _ Mg MagnetI-te _ Lu 0 Ms Muscowte : g T P Plagioclase I D E' Q Quartz 9 ‘53 '5 S Sphene g Z 0 Z Zircon [L - 0 | I I I 0 0 I I I I < m 03 E 2 “OJ 3 OMS E l n. 05C OM 40' — o OE — XM O x P 2 I I ON I I 0.001 0.01 0.1 l 10 100 PERCENT IN BEDROCK FIGURE 305.-—Mineral percentage changes with respect to bedrock in two size fractions of a relatively uncontaminated adamellite soil (#94), after ash corrections. ”Fe calculated as Fe-an; 1120+, 1120—, P205, 002 not analyzed in grus and < 62 am soil samples. sence of significant local contamination. All major min- erals except biotite and hornblende appear in place in the #94 sand and #94 silt mineral sequences, based on the observed weathering sequence (cf. fig. 297). By using percentage changes consistent with those of quartz, microcline, and plagioclase and with the ob- served weathering sequence, corrections were com- puted for biotite and hornblende in the less-than-2-mm fraction of #94 soil, as shown in the left part of table 16. Chemical corrections using the values of table 13 as- sume that the biotite and hornblende came from the Sage Hen Flat pluton or do not differ greatly from that of the pluton. The adjustments, including those from the ash, were subtracted from the original analysis of #94 soil, followed by correction to the previous oxide total and reconversion to elemental values, as shown in the right part of table 16. Chemical changes due to weathering at site 94 may now be assessed from bedrock and corrected soil analy- ses. Absolute losses and losses divided by atomic weight from #94 bedrock to less-than-Z-mm soil are given on the right side of table 14. Percentage losses with respect to bedrock, computed from concentrations in weight percent and assuming constant Zr percentage, are pre- sented graphically in figure 301 (right side). Bedrock-t0— soil absolute losses occur in the sequence Si >> A1 > K > Na>Fe>Ca>Mg>> Ti>Ba>Mn>Sr>>Rb,in close accordance with bedrock abundances. Absolute losses adjusted for atomic weight yield a slightly dif- ferent sequence, Si >> Al > Na > K > Ca > Mg > WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA 405 TABLE 16.—Adamellite sample 94—bedrock chemical analysis and adjustment of < 2-mm-soil analysis for ash and local contamination by biotite and hornblende [All analytical data in weight percent, except as noted] a) E” 3 N r: ’5 E 44 U x h w: 7:: °’ m~ a i: 3 9’ ”3 T5 ‘5 .. 2 ‘3 m 132% "6" g g: > cu 0 53 x: “as 25“ 22 2 2 2 5 3‘22 ~. g = ‘1 8 2 4:!- “ cm on: 3‘5 3,: m mg,» 053 gm 353» a 8 a 22 a“? 725 55 '97? E: 353 “g g 3&5 $52 5% “V 5-4 ...2 mm "' - "‘ _, .— 5 22222" :2 a §°2 52 3 Ogge 5§3v8¢>£§ 94 sand: Biotite ____________ —1.8 —30 6.7 8.2 4.7 —3.5 __., ,-__ _,-_ ______________________ ___- ______________________________ Hornblende ________ +200 —40 .13 .42 .08 —.34 ____ 1-.- _--- ______________________ --__ ______________________________ 94 silt: Biotite ____________ +170 —35 6.7 15 4.4 ~10.6 ___, 1-.- _-,_ ______________________ ___, ______________________________ Hornblende ________ +1050 —45 .13 7 .07 —6.93 ____ _-._ __-, ______________________ ,___ ______________________________ 94 soil<2mm: Biotite __________________________ —4.7 —4.4 70.27 ______________________ -__, ______________________________ Hornblende V_ ______________ <97 —.97 V.-- ______________________ ___ ______________________________ Si _____________ __.- -__- -,__ 31.13 78.59 7094 21.60 46.21 65.66 30.70 31.72 ________ __-- ,.__ _,-_ .40 —.02 — 11 .27 . .64 .38 .28 ________ ___, ,___ -,__ 308 ~24 -1,‘ 284 384 546 404 198 ________ __-_ -__- 8.35 —1.62 —.39 6.34 11.98 17.02 9.01 8.75 ________ _--- -__- 3.75 —.26 —.75 2.74 3.92 5.57 3.90 2.89 ________ __-- .09 —.01 ~.02 .06 .08 .12 .09 .09 ________ ___- _-._ _.., .71 —.02 —.36 .33 .55 .78 .47 .66 ________ ____ --__ _.._ 1.29 —.10 ~11 1.08 1.51 2.15 1.54 1.67 ________ -_._ _,._ 0,- .09 7.005 -,,_ .085 .095 .135 .12 .13 ________ --._ -_.- -,__ 368 —1 __,_ 367 434 617 5 2 624 ________ __., --__ 2.07 >73 —.02 1.32 1.78 2.53 1.88 2.69 ______________ __-- ,.__ __-_ 2.95 ~99 7.34 1.62 1.95 2.77 2.30 3.25 ______________ ,__- _.-_ __-- 62.3 —43 _,,_ 19.3 42.2 60.0 27.4 54.0 Fe >> Ti > Mn > Ba > Sr >> Rb. Percentage losses, the most significant indicator of chemical behavior ‘7 during weathering, follow the sequence Rb $ Naz Kz Mg>Sr>Mn z Ca>Ba>Si>Al>>Fe>Ti. The placement of Rb within the latter sequence is somewhat uncertain owing to its low concentration in both bedrock and soil and to its susceptibility to con- tamination from biotite, feldspars, and glass. Examination of figure 301 shows that percentage chemical losses generally follow the sequence alkali metals > alkaline earths > Si > A1 > metals. Some notable differences in element mobility between the dolomite and adamellite are also evident. Greater abso- lute amounts of Sr, Fe, and Mn are lost from adamellite than from dolomite, whereas Ca, Mg, Sr, Mn, and Fe all undergo greater percentage losses from dolomite bed- rock to less-than-2-mm soil than from adamellite bed- rock to less-than-2-mm soil. The soluble dolomite obvi- ously releases all its constituent elements more readily than they can be weathered from silicate and oxide phases in the plutonic rock. The preceding interpretations are based on bedrock- to-soil changes recorded at only two sites, although mean dolomite soil gravel values and microprobe analyses of both dolomite and adamellite minerals give additional confidence to the patterns of chemical weathering. The extensive soil contamination rendered study of all other samples virtually meaningless. For confirmation of the chemical weathering sequences suggested by the limited data provided by bulk changes in the solid phases, it was necessary to examine the chemical composition of natural waters related to both dolomite and adamellite. CHANGES IN THE LIQUID PHASE Chemical weathering usually occurs only in the pres- ence of a liquid phase which is undersaturated with respect to the altering phases. A considerable amount of information concerning processes and sequences of de- composition can be obtained from observations of pro- gressive chemical changes in water as it passes from the form of precipitation into soil water and finally into ground water and springs. The following discussion considers the chemical nature of rain and snow water, soil water extracts, and spring waters related to both dolomite and adamellite, the saturation pH and ex- changeable cations of the soils, and finally the significance of geochemical differences between the var- ious fluids in terms of chemical weathering. CHEMICAL COMPOSITION OF PRECIPITATION Four rain samples and four snow samples, obtained during the period July 30, 1966, to June 29, 1967, were collected in plastic-lined pans and filtered through 0.45-,um filter membranes into airtight polyethylene containers. The analytical data4 for these samples, to- gether with mean values for Sierra Nevada snow, are given in table 1 7. The White Mountains samples tend to ‘The procedures of Barnes (1964) were followed for field determination of pH and alkalinity for all water samples. All other chemical analyses of water were performed by the U.S. Geological Survey. 406 TABLE 17.—Chemical analyses, in milligrams per liter except as noted, of precipitation collected near Crooked Creek Station and com- parison with analyses of snow from the east slope of the Sierra Nevada [Mean for snow from Feth, Rogers, and Robinson (1964, p. 18—26) ] Crooked Creek Station Mean __—————— for snow, ea t 510 Mean‘ : ofSSierrl;e standard Nevada deviation Maximum Minimum IStigiecific conductance. ______ ,umhos- 7 :3 11 4 ,,,,, ____________________________ 141.10 .31 .06 .41 .34 :25 .90 .16 .43 361.12 .57 .23 .84 .9 :07 .22 .034 .19 <.5 __________________ .003 :.002 .006 .000 2 .00 .014 1.010 .024 .000 2 .01 .012 1.011 .032 .001 2 .03 01 .02 .00 2 05 11 + 06 .2 .035 2 17 2 7 +1 3 4.3 1.2 3 59 2 t 0 .2 2 2 07 06 + 09 .2 0 1 14 00 + 00 .00 00 2 03 4 + 2 .7 47 ‘Mean for four rain and three snow samples. 2Value for entire Sierra Nevada. have lower mean concentrations of most constituents, especially sulfate, than do the Sierra samples. Mn, Fe, Cl, and nitrate are comparable or slightly higher in the White Mountains. CHEMICAL COMPOSITION OF SOIL WATER Obtaining representative soil water samples is a con- siderable problem that has been approached from sev- eral angles. Lysimeters and other elaborate devices probably collect the most typical fluids, but they are often difficult to install and maintain and give no indi- cation of geographical variation unless employed in large numbers. In the White Mountains, soils are either dry or frozen except for brief periods during spring snowmelt and after very heavy summer storms, making field collections difficult to obtain. For this study, water saturation pastes of 46 Reed Dolomite soils and adamel- lite soils of Sage Hen Flat were prepared, and after 1 hour water samples were vacuum-extracted by Buchner funnel in the laboratory. (See section "Supplemental Information” for methods). Saturation, rather than ex- cess water mixtures, was chosen for initial soil moisture content in the belief that this would more closely reflect actual soil water conditions. Distilled water was used in all extractions, the assumption being that any differ- ences between precipitation and distilled water, with regard to the analyzed species, would not be significant in comparison to the soil water concentrations. Analyses of these extracts, summarized in table 18, indicate that standard deviations within a given soil type are often nearly as large as or larger than differ- ences between soil types. Comparison between some chemical species of the two soil water groups, however, is still possible. The adamellite extracts are lower in Ca EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT TABLE 18.—Chemical constituents in water saturation extracts of two groups of White Mountains soils [All analytical values other than pH are in milligrams per liter] Standard deviation Number of Mean High Low analyses Reed Dolomite Field pH Laboratory pH ___________ Adamellite of Sage Hen Flat 22 :9 94 9.8 23 14 :4 26 7.4 23 61 +22 110 28 23 10 :3 17 5.6 22 .24 1.10 .56 .12 20 0.1 :1.3 14 8.0 21 .0150 .0040 .0190 .0110 2 .64 _________________________________ 6.1 _________________________________ Field pH ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Laboratory pH 77777777777777777 $3 } saturation paste and Mg and apparently higher in K and phosphate than the Reed samples. Na is surprisingly low in the adamel- lite soil water extracts. The large deviations in soil water composition, which exceed the standard deviations in total soil composition (compare table 19 with tables 12 and 15), imply consid- erable departures from equilibrium and steady state (rate of ion gains = rate of ion losses) conditions, perhaps accentuated by the relatively short duration of the extraction process. For this reason, soil water ex- tractions may differ to some extent, primarily in degree of variability between sites but probably in ionic pro- portions as well, from meteoric water that has been in contact with the soil for several days or weeks during snowmelt or after heavy rains. SOIL PH Table 19 gives a compilation of field and laboratory pH data for Reed Dolomite soils and adamellite soils of Sage Hen Flat. These data, together with similar in- formation from basalt and sandstone soils in the area, indicate that instrumental laboratory pH values and indicator-determined field pH values are comparable to pH’s about 8.0, where the field methods begin to give consistently higher readings. A comparison of the figures for the two soil types shows the Reed soils to have values 0.5—1.5 pH units higher than the adamel- lite soils. EXCHANGEABLE CATIONS Amounts of total extractable soil cations (in milli- equivalents per 100 grams of ovendry soil) were ob- WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA TABLE 19.—Mean values for total exchangeable cations, percentage exchangeable cations, and pH in Reed Dolomite and adamellite soils of Sage Hen Flat Standard Standard deviation Number of Mean deviation (% of mean) samples Reed Dolomite sails pH: Average mean field ________ 8.0 0.1 1 25 Average maximum field _____ 8.2 .1 1 25 Average minimum field _____ 7.9 .01 5 25 Laboratory _______________ 7 5 .2 3 25 Exchangeable cations (me/100 g ovendry soil): K. .58 .30 52 25 .27 .09 33 25 32.4 13.3 41 25 6.1 2.5 41 25 1.6 .9 56 21 .73 .27 37 21 80.7 7.6 9 21 16.9 6.7 40 21 Adamellite soils pH: . Average mean field I ________ 6.8 0.5 8 22 Average maximum field - 7.6 .6 8 22 Average minimum field _____ 6.2 .5 7 22 Laboratory _______________ 7.0 .4 5 22 Exchangeable cations (me/100 g ovendry soil): K 33 .13 4O 22 07 .05 69 22 7 7 2.5 31 22 1 09 .34 31 22 3 8 1.8 47 21 70 .25 33 21 83 2 4.2 5 21 12 2 3.3 27 21 tained by repetitive leaching with 1.0 N NH4Ac at pH=7.0. (See “Supplemental Information” for details.) Values for exchangeable cations (table 19) were calcu- lated from the total extractable and water-soluble cat- ions as follows: Exchangeable cations=total (NH4Ac) extractable cations—cations in water saturation extracts (all values in milliequivalents per 100 grams of dry soil). The exchangeable cations were then recalculated to 100 percent (table 19). The adamellite soils are much lower than Reed soils in amounts of all exchangeable cations; they are higher in percentage exchangeable K, slightly higher in percentage Ca, comparable in Na percentage, and somewhat lower in percentage Mg. CHEMICAL COMPOSITION OF SPRING WATERS Water samples from four springs related to Reed Dolomite and adamellite of Sage Hen Flat were col- lected during the summer and fall of 1966 and during the summer of 1967. Water samples were filtered through 0.45-Mm filters into airtight polyethylene con- tainers. The samples were usually collected near sun- rise or sunset to minimize air-water temperature differ- ences, which would affect pH measurements. A part of each sample collected during the summer of 1967 was 407 acidified to pH 2 to prevent precipitation of metals and was used for analysis of Al, Fe, and Mn. Al values are markedly higher for the 1967 samples, reflecting either precipitation of Al in previously collected samples or solution of particles less than 0.45 ,um in the acidified parts. There was relatively good agreement of Fe and Al values in the 1967 samples with those reported for granitic waters in the Sierra Nevada by Feth, Roberson, and Polzer (1964). Changes caused by orifice location, diurnal variation, and seasonal fluctuation at a given spring were found to be negligible compared with dif- ferences between springs. WATERS RELATED IN PART To THE REED DoLOMITE Poison Creek Spring emerges from the east side of a canyon wall about 400 feet east of the Wyman-Reed contact, which dips about 60° westward at this location. The spring water would appear to be chemically con- trolled in large part by the dolomite, which crops out continuously for several miles west of the Wyman con- tact, but the water is also influenced to a lesser degree by the sandstones, shales, and limestones near the orifice. Waters emerge from three orifices (herein des- ignated A, B, and C, from south to north) about 15 feet apart. Discharge from orifice C, the major outlet, aver- ages about 0.1 cfs (cubic feet per second), as measured with a Pygmy current meter. Cottonwood Spring, the principal source of the south fork of Cottonwood Creek, is in a valley that partially separates the Reed Dolomite from the Cottonwood plu- ton. The orifice lies within the Reed, though, and the apparent catchment basin is dominated by dolomite. Discharge is quite constant at about 1.8—2.0 cfs. Analytical data for these two sampling locations are summarized in table 20. The predominant ions are Ca, TABLE 20.—-Field and laboratory analytical data for two natural waters associated in part with the Reed Dolomite [All concentrations in milligram per liter] Poison Creek Spring Cottonwood Spring Standard Standard Mean deviation Mean deviation Field pH __________________ 7. 0.14 Specific conductance K. _______ l286 ________ K t _______________ .99 03 Na+ _____________________ 3.35 09 29 2 14.9 3 1<5 1.002 1.001 ‘.019 13 {02 1.8 175 1.3 7.8 _____ 1.00 Water temperatur 4°C) , 5.8 Cation ________________ ean 2.84 ________ Anion ________________ epm. 3.06 ________ Number of analyses __________ 7 ________ 1Based on one analysis. 408 Mg, and bicarbonate; Na and sulfate are quite low. Poison Creek Spring shows the influence of the Wyman beds in its increased Na, K, Si, and sulfate with respect to the Cottonwood samples, and its higher Ca/Mg ratio reflects the presence of limestones near the orifice. The pH of Poison Creek is the same as that of Cottonwood Spring, however, and in most respects the waters of the two springs are similar. WATERS RELATED LARGELY To THE ADAMELLITE OF SAGE HEN FLAT Sage Hen Spring and Crooked Creek Spring appear to be almost entirely related to the Sage Hen Flat pluton. Formations adjacent to the adamellites, such as the Wyman Formation, Reed Dolomite, Deep Spring For- mation, or Campito Formation could have some slight effect on the chemical composition of these waters. Dis- charge at Sage Hen Spring is about 0.1 cfs, but fluctuates appreciably. Discharge at Crooked Creek Spring is even smaller, its mean about 0.05 cfs or less. Analytical data for the two sampling sites related to the adamellite are given in table 21. Crooked Creek Spring has a higher pH than Sage Hen Spring and contains less Mg and slightly less Na. The fact that the two waters, draining the same plutonic body, can be consistently distinguished chemically is an indication of the sensitivity of natural water to variation in bed- rock composition and weathering conditions. The adamellite-derived spring waters, however, are virtu- ally identical in most other respects and contrast sharply with those related to dolomite. Waters draining the adamellite are much higher in Na, sulfate, and chloride than are the dolomite waters and much lower in Ca, Mg, and bicarbonate. The pH values of the four spring waters are almost the same except for Crooked TABLE 21.—Field and laboratory analytical data for two natural waters associated with the adamellite of Sage Hen Flat [All concentrations in milligrams per liter] Crooked Creek Spring Sage Hen Spring Standard Standard Mean deviation Mean deviation Field H. __________________ 7.96 0.15 7.72 0.11 Speci ic conductance K. _______ 1129 ________ l127 ________ K+ _______________________ .79 .05 .64 02 6.23 .19 6.56 16 17 1 16 5 .93 08 1.55 08 2<.5 ________ 2<.5 ________ ‘.Ol4 ________ 1.000 ________ 1.02 ________ 1.019 ________ ‘.014 ________ [033 ________ 15.4 .8 16.4 5 1.015 ________ 1.01 ________ ‘.9 ________ {85 ________ 61 3 60 1 ‘3.3 ________ ‘.8 ________ 9.8 .5 9.4 1.1 1.145 ________ ‘.11 ________ 5.6 1 6.5 1 1.22 ________ 1.24 ________ ion. , 1.29 ________ 1.22 ________ Number of analyses __________ 13 ________ 7 ________ 1Based on two analyses. 2Based on one analysrs. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT Creek Spring, which is more alkaline than either of the springs related to the dolomite. ANALYSIS OF CHEMICAL CHANGES For purposes of analysis and interpretation, the chemical data for natural waters related to both dolo- mite and adamellite are viewed in the following discus- sion from three perspectives: (1) Relative mobilities, showing the degree to which the various chemical species are mobilized from bedrock by weathering and released into solution, (2) stability with respect to some important solid and gas phases, including a discussion of water chemistry in relation to several equilibrium reactions, and (3) a chemical comparison of soil water with precipitation and spring waters. RELA'I‘IVE MOBILITIES A common means of evaluating the ease with which various chemical constituents are released by weather- ing from bedrock to ground water is through the calcu- lation of relative mobility: Percent X of N-element Relative mobility_total in water (element X) —Percent X of N-element total in bedrock where N is the number of elements considered. This computation assumes that a given water can be definitely correlated with a given lithology, that the mean analytical values for that lithology are represen— tative, and that the composition of the water with re- gard to the investigated constituents is entirely a func- tion of contact with the bedrock material. It is likely that the waters sampled in the White Mountains have been affected to some extent by windblown contamin- ants and the rhyolitic ash, but evidence presented here (see “Relationship of Soil Water to Precipitation and Spring Waters”) suggests that major influences on ground—water composition occur after percolation through the soil. Relative mobilities calculated for Cottonwood Spring with respect to dolomite bedrock (table 22) fall into the sequence Mg > Ca >> Fe > Mn. Mobilities of other elements, for example, Si, Al, Ti, Na, and K, are more susceptible to influence by weathering of contaminants than of the bedrock, which contains only trace amounts of these constituents. Interpretation is therefore confined to the carbonate-component elements. The rel- ative mobility sequence of table 22 agrees well with the order of percentage losses from dolomite bedrock to soil (cf. fig. 301), except that the mobility of Fe in dolomite is clearly greater than that of Mn, a relation not evident from solid—phase considerations. In figure 306, ranges of relative mobilities for eight elements, based on analytical means and standard de- WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA TABLE 22.—Relative mobilities for four elements in Cottonwood Spring waters Mg Ca Fe Mn Concentration in bedrock ______________ percent“ 12.46 21.52 0.26 0.055 Concentration in Cottonwood Spring water ______________ mg/l __ 15.9 24 .043 .004 Percent of four~element total (bedrock) __ is. 36.33 62.75 .76 .16 Percent of four-element total (spring) __ , L“ 39.79 60.08 .108 .010 Relative mobility ____________________________ 1.095 .957 .142 .063 30 I | 1 Ca 10 — M9 — its 1 — ._. Mn >- K S! p— : f E E O E LU > ,: S 0.1 — _ LLI o: E 0.01 — —- EXPLANATION A Sage Hen Spring 0 Crooked Creek Spring Note: Overlapping ranges are offset for clarity Al Mn 0.001 ' | 1 0.05 0.1 1 10 50 MEAN PERCENTAGE IN BEDROCK FIGURE 306—Relative mobilities for eight elements in spring waters related to adamellite. Vertical bars indicate ranges of values. viations at Crooked Creek Spring and Sage Hen Spring, are plotted against mean bedrock percentage on a log- log scale. The horizontal scale serves to emphasize the importance of the various constituents with regard to bedrock weathering and to their net contribution to the ground water. Elements plotting in the upper right of the diagram (Ca, Na, and Si in this case) are mobile constituents present in appreciable quantities in the bedrock and will be liberated in relatively large total amounts to the ground water. Elements that plot to- ward the lower left (such as Mn) are comparatively immobile, are not abundant in the adamellite, and are 409 therefore released to the ground water in trace amounts. The apparent sequence of mobilities for the ada- mellite-draining springs is Ca > Mg z Na >> Si z K 2 Mn z Fe >Al. Values for Mn, and to some extent for Mg, vary considerably between the two springs, making the placement of these elements within the order somewhat uncertain. The results as a whole are in good agreement with those of Feth, Roberson, and Polzer (1964) for quartz monzonite (adamellite) terrane in the Sierra Nevada and are generally similar to losses from bedrock to soil (table 14; fig. 301). The relative positions of Ca, Mg, and Na in the weathering sequence differ, however, between the solid phase and liquid phase perspectives. Considering the problems of soil contamination, one possibility is that carbonate grains may be continually blown into adamellite soils and quickly dissolved, re- sulting in increased Ca and Mg concentrations in wa- ters draining the pluton. Only a few carbonate frag- ments were observed in adamellite soils; however, solu- ble grains (especially those reactive fragments of silt or clay size) would not be expected to persist in nonsoluble soils, and both dolomite and calcite were found in dust trap residues collected within the area (Marchand, 1970). Ruhe (1967, p. 57—59) suggested solution of windblown contaminants to explain the formation of thick caliche horizons in soils formed on low-Ca parent materials in New Mexico. STABILITY WITH RESPECT To SOLID AND GAS PHASES5 By using means and standard deviations for ion con- centrations, pH, and temperature from Cottonwood Spring and Poison Creek Spring, maximum, minimum, and most probable values of ion-activity products (IAP) were calculated for both calcite and dolomite as follows: IAPc = [aca++][aco:] and IAPd = [aCa++][aMg++][acoa-2]2 where [a] is the ionic activity of the solutions (approxi- mately equal to concentration). A comparison of the spring waters with regard to carbonate equilibria is presented in table 23. All waters are undersaturated with regard to both calcite and dolomite, yet minor amounts of calcium carbonate have precipitated on ex- posed rocks near all the springs sampled. Calcite lacks saturation by a factor of about 2.5—10 in the dolomite- 5Sources for physical constants used in calculations: K1 first carbonic acid dissociation; Harned and Davis (1943. p. 2030). K2 second carbonic acid dissociation; Harned and Scholes (1941, p. 1708). KC02 graphical solution, using data of Markham and Kobe (1941, p. 449). A first constant, Debye-Hucke equation; Manov, Bates, Hamer, and Acree (1943, p. 1765). Kc calcite equilibrium constant; Larson and Buswell (1942, p. 1667). Kd dolomite equilibrium constant; (Hsu, 1964). 410 TABLE 23.—Degree of saturation with respect to calcite and dolomite for four spring waters [Ranges of values reflect temperature and compositional extremes. MP. 2 most probable value; IAP : ion-activity product] Cottonwood Poison Creek Crooked Creek Sage Hen Spring Spring Spring Spring IAPc: Maximum ,,,,,,,,,,,,, 2.83x10-9 2.51X10'9 ,,,,,,,,,,,,,,,,,,,,,,,, MP. ,,,,,,,,,,,,,,,,, 1.37X10'9 1.67><10'9 0,746X10'9 0.401><10'9 Minimum ,,,,,,,,,,,,, ,682X10‘9 .672><10'9 ________________________ Kc‘: Maximum _____________ 7.44x 10'9 7.92><10‘9 ________________________ M.P. _________________ 7.69X10‘9 7r'98><10'9 8.02X10'9 7,79X10’9 Minimum ,,,,,,,,,,,,, 7.92X10'9 8,02><10'9 ,,,,,,,,,,,,,,,,,,,,,,,, IAPc/Kc: Maximum ,,,,,,,,,,,,, .381 .317 ________________________ M.P. ,,,,,,,,,,,,,,,,, .178 .209 .093 .052 Minimum ,,,,,,,,,,,,, .086 .083 ________________________ IAPd: Maximum ",1 8,38x10'” 5.09X10"a ,,,,,,,,,,,,,,,,,,,,,,,, M.P. ,u- 2,03X10'" 2.34X10‘18 4.99X10'20 2.56><10'm ________________________ Kd2 ,,,,,,,,,,,,,, ,1 2.0><1()'17 2.0X10‘.” 2.0X10'17 2.0X10'17 IAPd/Kd: Maximum ,,,,,,, A, .419 .254 ________________________ 1 .P. ,,,,,,,,,,, __, .103 .117 2.5><10'a 1.28><10‘a Minimum ,,,,,,,,,,,,, .026 .020 ________________________ 1From Larson and Buswell (1942, p, 1667). 2From Hsu (1964). draining waters and by more than 10 in adamellite- derived waters. Cottonwood Spring and Poison Creek Spring are also much closer to dolomite saturation than are their counterparts in the pluton, but all waters are even more undersaturated with respect to dolomite than to calcite. The undersaturation of calcium carbonate in the Crooked Creek and Cottonwood drainage basins con- trasts with the fourfold supersaturation reported by Barnes (1965, p. 92) in the headwater springs of Birch Creek, just south of the study area. The latter drains a mixed terrane in which solution of Wyrnan limestones, as well as the Reed Dolomite, may have a profound influence on the state of the ground-water calcite reac— tion. Partial pressures of carbon dioxide in the four spring waters (table 24) were computed from the following relation: (H+)(H003-) PC02(aq) = KiKC02 where K1 2 (H+)(H003 —) (H2003) and PCOZ All calculated values for the waters are at least twice that of the atmosphere, and Cottonwood Spring shows C02 partial pressures as much as 22 times greater than the air. The two springs related to the dolomite terrain have partial pressures much greater than the springs in EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT TABLE 24.—Partial pressures of carbon dioxide in four spring waters and a comparison with atmospheric PCO2 (MP. = most probable value] Cottonwood Poison Creek Crooked Creek Sage Hen Spring Spring Spring Spring PCO2 of water (atm): Maximum ,,,,,,,,,,, 5.05X10'3 4,10X10‘5 ______________________ . _______________ 2.66X10'3 2.86X10‘3 5.70X10‘4 9.79X10" Minimum ,,,,,,,,,,, 1,44X10'3 1.98><10'3 ______________________ PCO2 of atmospherehatm __- 2.3X10'4 2.3X10" 2.3X10" 2.3X10" PCOz(w) /PCOz(a): Maximum ___________ 22 18 ______________________ M.P _______________ 12 12 2.5 4.2 Minimum ,,,,,,,,,,, 6,3 8.6 ______________________ the adamellite, yet the pH of all springs is similar. Solution of dolomite would tend to increase PC02 but would also increase pH, the net effect being to decrease free C02. The explanation for the striking contrast be- tween the two groups of springs must partly lie in the marked vegetational differences between the two lithologies: The adamellites are dominated by rela— tively dense growth of sagebrush, limber pine, grasses, and perennials, whereas bristlecone pine and sparse low perennials cover most of the Reed terrane. Perhaps photosynthetic rates, and hence root respiration, differ greatly between the two plant communities. In any event, it is apparent that local vegetational changes may play an important role in controlling the pH of ground water and hence in regulating rates of chemical degradation. For the adamellite-draining springs and adamellite soil water, means and standard deviations were used to compute molalities and activities for Na+, K+, H+, and H48i04 and to plot the range of each fluid on low- temperature stability diagrams for Na silicates (fig. 307) and K silicates (fig. 308). Plots based on activities did not differ significantly from those based on molality and are not reproduced on the diagrams. Sage Hen Flat pluton water plots within the kaolinite field in both figures, as do most stream and spring waters. The posi- tion of the water plots, toward the silica-rich side of the kaolinite field and away from the gibbsite field, tends to support the X-ray diffraction and microprobe evidence that kaolinite is forming as a stable authigenic phase in the adamellite soils. The waters are well above quartz saturation, probably owing to the weathering of Si from feldspars, biotite, and other silicates, in addition to quartz, but are significantly below saturation with amorphous silica. Major departures from equilibrium are evidenced in adamellite-related waters by oversaturation with re- spect to quartz and undersaturation with respect to colloidal silica and in waters draining the Reed by un- dersaturation with regard to dolomite. Such depar- tures, however, do not preclude the existence of steady- state conditions in the spring waters. 8 I | I I | I I I I 7 _ Albite _I Montmorillonite I I I I e — I I I I I I I 5 ’ I u I A 0" 7 s e g I 3 a g I: I ‘ I — Gibbsite Kaolinite :‘ ' E I = | I i1 o 2 _ I E- E I E C .9 I ‘5 2 | a 1 — I _ Means: x Soil Water extracts I . Crooked Creek Spring A Sage Hen Spring I Ranges: D I 0 | I I I —6 —5 —4 —3 Log [H4 SIO4] FIGURE 307I—Stability relations ofphases in the system NazO-Alz 03-Si02-H20 at 25°C and 1 atmosphere total pressure as functions of(Na+)/(H+) and (H4Si04). After Feth, Roberson, and Polzer (1964, p 65). RELATION OF SOIL WATER To PRECIPITATION AND SPRING WATERS An important question in the chemistry of natural waters is the space-time dimension of chemical weath- ering changes. Do most of the chemical additions to rain and snow water occur quickly within the soil, or does the 411 8 | I ‘I I l Potassium mica Potassium feldspar I 7 — I I I I I 6 _ I I I I I I I I 5 ‘ I u I I : U I e I § E r 4 — .5 I - fi ‘5 ‘a s I 3 [05 I | A | 9‘ 3 — Gibbsite Kaolinite H :s— ' a I :— l I 2 — I g _ .C I i 8 I I 52° 1 — I _ Means: x SoiI water extracts I . Crooked Creek Spring A Sage Hen Spring I Rangeszfl I 0 L . I I I 4 vs —5 —4 —3 Log [H4 SIOdl FIGURE 308.—Stability relations ofphases in the system KgO-Alg 03-Si02-H20 at 25°C and 1 atmosphere total pressure as functions of (K+)/(H+) and (H4Si0‘). After Garrels and Christ (1965, p, 361). composition of percolating ground waters continue to change, even if the lithology remains the same? Al- though absolute concentrations of chemical species in the soil water extracts are much higher than those in precipitation and spring water, comparisons are possi- ble through ratios of dissolved constituents. 412 Figure 309 compares the composition of waters re- lated to the Reed Dolomite in the ternary system Na + K — Ca— Mg.6 The dolomite soil water extracts are more or less intermediate in composition between precipitation and spring water values, but somewhat higher in percentage Ca than either of the other fluids. A progressive increase in the divalent/monovalent cat- ion ratio occurs from precipitation to soil water to springs. The Ca/Mg ratio increases slightly from about 1.8 in precipitation to a value of nearly 5 (the ratio in bedrock is about 1.6) in soil water, but then decreases substantially from soil water to spring water. This de- crease may be in response not only to bedrock composi- tion, but to selective adsorption and removal of Ca by the soil exchange complex. The fact that the spring waters do not plot closer to the bedrock is probably attributable to the weathering of ash and local soil contaminants and to differing relative mobilities of the plotted cations. Waters related to the adamellite are compared for the same ternary system in figure 310 (bottom). Soil water plots over a wide range and shows a definite increase in Ca/Na + K and Ca/Mg with respect to both precipita- tion and bedrock. It appears to be closer in composition to spring water than precipitation in terms of Ca con- centration. The hydrolysis of plagioclase, solution of secondary calcium carbonate in the soil, and the high mobility of Ca is probably responsible for this trend, which proceeds away from bedrock composition. Also in figure 310 (top), it appears that the Na/K ratio in the soil water extracts has decreased relative to precipita- tion in response to bedrock and perhaps ash composi- tion. The large increase in Na/K from soil water to M9 \50 u Na+K EXPLANATION 0 Bedrock mean E Volcanic glass 1:) Distribution + Precipitation mean I Soil water mean xct Cottonwood Spring mean A Exchangeable cations mean FlGL'RE 309.‘Pr0gressive cation changes from precipitation water to soil and spring waters related to Reed Dolomite. Plots of bedrock, volcanic glass, and exchangeable cations suggest relative influence of these factors. 6Values plotted on compositional diagrams in this paper were calculated on a parts per million (milligrams per liter) rather than an equivalents per million basis because the oxidation states of Fe and Mn were not known and a uniform basis between diagrams was desired, EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT 10,0 / .1. I 4 / \ { /\ \l I / .,,__, , t“, / / §¥1.0— / '\ 2-- o / ...s--— / \ \ \\ \ I \ EXPLANATI 0N + Precipitation mean A Exchangeable cations mean 0 Soil water mean 0 Bedrock mean xc Crooked Creek Spring mean E Volcanic glass x5 Sage Hen Spring mean 1:» Distribution FIGURE 310.———Progressive cation changes from precipitation water to soil and spring waters related to adamellite. ground water is interpreted as the result of K fixation by colloidal adsorption and selective plant uptake, and of high Na mobility. Soil water has a Na/K ratio much closer to precipitation than to spring water, but is quite distinct from both. Figure 311 shows regular increases in both Mg/Fe and Ca/Fe from precipitation to soil water to spring water, with soil water plotting approximately inter- mediate in position between the other two fluids. Both trends are away from bedrock composition and appar- ently result from low Fe mobility due to precipitation of Fe oxides within the soil. From precipitation water to soil water to ground water, increases are evident in Si relative to Na+K and Mg+Fe (fig. 312), presumably in response to bedrock composition and possibly also to the rhyolitic ash. Soil water extracts on this diagram plot much closer to pre— cipitation than to spring waters. The pH value is probably the best single indicator of water chemistry. The median pH of snow samples from the Sierra Nevada is reported as 5.8, close to the value for pure water in equilibrium with atmospheric carbon dioxide (Feth, Rogers, Roberson, 1964, p. 24—25). All soil and ground waters in the White Mountains have much WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA 100 r EXPLANATION + Precipitation mean WM=White Mountains 1’35 SN =Sierra Nevada ~\ '\ . Soil water mean 0.1 - Xs Sage Hen Spring mean Xc Crooked Creek Spring mean 0 Bedrock mean a Volcanic glass ‘ Distribution Mg/Fe E I 0.01 ' I l I I 0.01 0.1 1.0 10 100 1000 Ca/Fe FIGURE 311,—Mg/Fe and Ca/Fe ratio comparisons of waters related to adamellite. 10 1‘0 xs xc EXPLANATION _ + Precipitation mean '13 if I {ifs-l WM=White Mountains m / . a. . I _ z I 1' SN =Sierra Nevada \ .i . Soil water mean 0 1 + W“, x; Sage Hen Spring mean SN Xc Crooked Creek Spring mean 0 Bedrock mean 0 Volcanic glass 1:; Distribution 00] l I l | I 0.01 0.1 1.0 10 100 1000 Si Mg + Fe FIGURE 312.—Silica-cation ratio comparisons of adamellite-related waters. higher pH’s (tables 18, 20, 21) than this figure, owing principally to solution and hydrolysis reactions. For the dolomites, soil saturation pH is comparable to spring water pH, but for. the adamellite, soil saturation values are intermediate between the pH of precipitation and that of the springs. Most of the chemical trends just described appear to be controlled by bedrock composition or relative mobil- ity, but the ground waters plot in tight groups distinct from the bedrock, suggesting that a steady-state condi- tion, rather than chemical equilibrium, exists in spring waters that have passed through both lithologies. In the soils, absence of both equilibrium and steady state is manifested by the scatter of soil water points on the compositional diagrams. Soil water extracts from both dolomite and adamel- lite terrain are notably distinct in composition from precipitation and spring waters draining the respective lithologies. Total concentrations are much greater in soil saturation extracts than in spring waters owing to 413 the diluting effects of percolating precipitation water on the latter. The contrasting ionic composition of these two groups of fluids imply the existence of weathering reactions at considerable depth and preferential chan- neling of ground-water flow from certain localized areas of high recharge and possibly of different lithology. Most of the soil water extracts plot in an intermediate position between precipitation and ground water, but some (N a/K for adamellite-related water, for example) soil water plots do not fall in a position between the other two fluids. The pH of dolomite soil water extracts is much closer to spring water than to precipitation, however, suggesting that chemical adjustments in sys- tems having soluble phases occur more rapidly than in those involving silicate or oxide phases. Precipitation of calcium carbonate, probably sporadic in response to wet and dry seasons, and the contaminat- ing influence of the rhyolitic ash, local aeolian debris, and foreign waters complicate the interpretation of soil water compositions. Also, the preceding conclusions as- sume that saturation extracts approximate actual soil water, an assumption which may not be entirely valid. A longer period of water contact with the soil might be expected to shift soil water composition closer to that of spring waters and to decrease the variability of extract compositions. A need exists for methods of soil water extraction which will produce fluids clearly representa- tive of the actual soil solution under field conditions. CHEMICAL FRACTIONATION IN THE BEDROCK-SOIL-WATER-PLANT SYSTEM Elements released by solution or hydrolysis of min- eral grains follow a variety of courses. They may dis- solve in the soil water and be flushed out of the system, precipitate as a solid soil phase, become adsorbed on colloidal soil particles, or be utilized by organisms, especially vegetation. Chemical analyses of representa- tive plants from the area, together with data already presented, make possible the location and apportion- ment of a number of chemical species within the bedrock-soil-water-plant systems of the Reed Dolomite and the adamellite of Sage Hen Flat. Soil water and exchangeable cation concentrations were converted from milliequivalents per 100 g of soil to grams per cubic centimeter of soil, using milliequiv- alent weights and mean bulk densities of the total soil. Since the laboratory data apply to only the less-than-Z-mm fraction, two sets of figures were calcu- lated for the soil water, one assuming no contributions from the soil gravel fraction (minimum) and a second assuming a contribution from the coarse material pro- portional to its weight percentage of the whole soil (maximum). For practical reasons, only a few of the hundreds of 414 White Mountains plant species could be analyzed. Aboveground parts of sagebrush (A rtemisia arbuscula), junegrass (Koeleria cristata), and sandwort (Arenaria kingii), and twigs, needles, and cones of bristlecone pine (Pinus aristata and limber pine (P. flexilis) were col- lected on both olomite and adamellite. Aboveground parts of flax (L um perenne) were obtained from dolo- mite terrain. T ese six species were chosen to represent vegetational oups as follows: Pinus aristata and P. flexilis—trees; Artemisia arbuscula—sagebrush and shrubs; Koeler'a cristata—grasses; Arenaria kingii— —noncarbona e-associated herbs; and Linum perenne— car onate-associated herbs. Plant samples were washed ith distilled water and ovendried at 104°C prior to ash analysis (table 25; see section "Sup— plemental Information” for analytical procedures.) Percentage cover for the preceding groups, estimated at about 20 randomly selected sites on each lithology and then averaged for dolomite and for adamellite, is given in column 12 of table 25. These figures are recal- culated to 100 percent in the last column. The percent— ages of total vegetation (in terms of cover) were then used to obtain weighted averages for each element in plants on each substrate, as follows: Weighted average Si=(percent Si in species X) (per- cent X in total vegetation) + (percent Si in species y) (percent y in total vegetation) + . . . + (percent Si in species N) (percent N in total vegetation), whereN is the total number of species considered. Vege- tational concentrations from table 25 were adjusted to a grams per cubic centimeter basis using the weight and percent cover of each plant sampled (giving grams per square centimeter) and dividing by the mean depth of each soil type (giving grams per cubic centimeter of EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT soil). Calculations were made for the individual species and for the weighted averages applicable to each soil type. Since these averages assume continuous vegeta- tive cover, a second pair of weighted averages were computed, adjusting for the actual plant cover. Plots of total soil (uncorrected for contamination), exchange- able cation, soil water, and plant concentrations for adamellite and dolomite terrane are shown in figure 313. The adamellite comparisons, expressed on a per- centage basis, are extended to precipitation, bedrock, and spring waters in figure 314. Vegetational concen— trations are computed as percentage of soil water plus plant concentration in table 26. An examination of element totals in figure 313 indi- cates that constituents released by weathering are con- centrated primarily in exchangeable positions (cations only) and secondarily in soil water and plants. Ex- changeable cations are adsorbed in the sequence Ca > Mg > K > Na. Cation exchange in soils therefore tends to decrease the Ca/Mg and K/Na ratios with re- spect to ratios of cations initially made available by weathering. Cation exchange is less important as a fractionation mechanism than these illustrations would imply, because after the exchange complex be- comes saturated, no further removal by exchange can occur unless the exchange capacity increases. Although trends vary to some extent between species, the overall tendency of the White Mountains flora, with respect to soil water (table 26), is to withdraw elements in the sequence P > Fe >> K > Ca; Mg > Si >> Na. The ratio of plant concentration to soil water concentration is less for the dolomite, but the sequence is the same regardless of lithology. As the result of plant uptake, the soil K/Na ratio is further decreased, and Fe and P are extracted in relatively large quantities. Si is sufficiently soluble at the moderate to high pH’s of the adamellite and dolomite soils to become concentrated in the liquid phase, although plant concentrations consti- TABLE 25.—Chemical composition (dry weight percentage basis) of some major plant species on Reed Dolomite and adamellite of Sage Hen Flat and weighted averages used for geochemical comparisons Number of Number of Mean percent Percent of total plants duplicate cover vegetation Si Al Fe Mn Mg Ca Na K P analyses (see text) (see text) Dolomite: Pinus aristata ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 0.0080 0.073 0.015 0.0016 0.17 0.58 0.0053 0.20 0.0480 2 1 17.60 64.9 Pinus flexilis ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, .0140 .071 .013 .0061 .09 .57 .0133 .23 .0630 1 1 .77 2.8 Artemisia arbuscula .0300 .60 .062 .0063 .23 71 .0081 77 .0874 3 3 1.45 5.4 Koeleria crislata ,_ .0980 .82 .10 .0165 .26 81 .0066 27 .0560 3 1 2.74 10.1 Arenaria kingii "0 0760 1.1 108 .0100 .47 3 39 .0219 69 0530 4 1 2.43 9.0 Linum perenne "a .0549 , .42 040 .0102 .26 85 .0034 82 0745 6 1 2.11 7.8 Weighted average ,,,,,,,,,,,,,,,,,,,,,,,,,,, .0282 .29 036 .0049 .21 88 .0072 33 .0539 ____________________ (27.10) (100.0) Adamellite: Firms aristata 0120 .051 010 .0049 .08 55 .0078 31 .0690 1 1 .05 .2 Firms flexilis ,,,,,,,,,,,,,,,,,, 0120 .034 0068 .0044 .08 44 .0083 45 .1100 1 1 2.61 9.6 Artemisza arbuscula 1., ,0, 0600 1.07 147 .0131 . 57 .0118 80 .0829 4 4 9.85 36.0 Koeleria cristata ,,,,,,,,,,,,,,,, 0300 .47 064 .0064 .09 34 .0058 27 .0385 2 1 6.47 23.7 Arenaria kingii ,,,,,,,,,,,,,,,,,,,,,,,,,,,, 1200 5.5 43 .0267 .33 2 52 .0353 92 .0555 2 1 8.32 30.5 Weighted average ,,,,,,,,,,,,,,,,,,,,,,,,,,, 0665 2.18 20 .0148 . 1.10 .0172 68 .0666 ,,,,,,,,,,,,,,,,,,,, (27.30) (100.0) o 10-1 10‘2 10-3 _l 0 fl) LL 0 u: 10’4 Lu )— U.| E l- 2 LLI U 2 m 10‘5 D O D: LU L (I) E E o 10‘6 10—7 10‘8 10—9 FI(.l'RI-: 313. WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA 415 I I | | | I I I I I I I I I A A A A _ ‘ _. A A A A _ X ‘ X _ A A X X A _ X _ A X x X X _ 'Ak X .Ak _ 182 .Ak :2: -Ak ‘9 4’ ,Aa $ Aa Ak- Pf' w: .Aa ’Pa .Pf X W Pf:w 58’ w k:Aa Pa' Pa Pf‘ A Pf: BI ,w~c 181 .Pa I81 _ 'Aa ? .Ak ¢ 'Aa Pf .W-C _ 'Ak Pa “4; 'Kc Bl .w-c 'K 3A3 .W-C 'PS $ g I81w .w-c 'W 'Pf c 'P3 'K0 'Ak Aa ”*3 'P‘ Pf ch Ipf .Pa . -L A Pa' oKC Aa. nKC .w~c . a .w . .Pa W w-c ' Ak' .w-c Kc .W _ ‘ Ak. ' .L .Kc .L Pf K .W ;w«c ' c .w-c gAk K Pf 8A3 Pa .Pa . Pf .L ° c - . Kc 'w-c P3 'K Pf c .Aa Aa _ ,w~c 181 'pa Ak: _ 69 L . KC. Iw-c Kc ‘5 _ (U E -L § E E Q) 3 1’ E I» E 2 _ E 3 _ L a e E I I I I I I I I I I I I I I Na Si Ca Mg K Fe P + Na Mg Ca K P f ADAMELLITE OF SAGE HEN FLAT REED DOLOMITE EXPLANATION 4* Soil water minimum Aa Arremisia arbuscula Pf Pinus flexilz‘s 81 Soil water maximum Ak Arenaria kingil' w Weighted average X Exchangeable Kc Koeleria Cristata w-c Weighted average A Soil (<2mm) L Linum perenne (adjusted for ac- . Plants: Pa Pinus arismta tual coverI —Comparison ofelemental and total concentrations in soil, soil water, colloidal exchange, and plants, expressed in terms ofgrams per cubic centi— meter of soil, 416 ‘00 I | I I I I I I I X A 3* (9 + + (9 o 2 ' g e X + G) e 10 — + _ e e D A + A D g G) _J < 6 ._ O I .— O E E] 0 X u'J + E 1 — Cl _ 2 LL 0 X LLI O < .— 2 Lu 3 © Lu + 0. <1) E . O 0.1 — o A _ c O . C 0000 $000 001 I I I I I I Y ' I _l' ' Na K Ca Mg Si Fe Mn Al P ADAMELLITE OF SAGE HEN FLAT EXPLANATION A Mean bedrock + Precipitation A #94 bedrock 9} Soil water [J #94 soil <2mm, corrected for contamination X Exchangeable cations 0 Sage Hen Spring 0 Crooked Creek Spring 5) Vegetation FIGURE 314,—Fractionation of nine elements in the adamellite bedrock-soil-water-plant system based on percentage concentrations. tute as much as 30 percent of the nonexchangeable total for areas of continuous vegetative cover. Fe and P tend to be quite insoluble under these oxidizing and alkaline soil conditions. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT TABLE 26—Fractionation by plants with respect to soil waterfor seven elements in adamellite and dolomite terranes Plant concentration (g /cc soil) [Elemental percentage : Plant concentration + soil water concentration (g /cc soil) X 100] Si Ca Mg Na Adamellite: Continuous cover____ 25—30 46—53 47—54 5.9 —7.4 Actual cover ________ 8.3—11 19—23 19—24 1.7 —2.1 Dolomite: Continuous cover____ ______ 17—30 17—30 1.1 -2.3 Actual cover ______________ 5.1—10.0 5.2—102 0.31—0.65 K Fe P Total Adamellite: Continuous cover --_ 51—58 98 ~100 47—54 Actual cover ________ 22—27 92 99 20—24 Dolomite: Continuous cover ___ 34—51 __ 100 18—32 Actual cover ________ 12—22 __ 99—~100 5.7—11 The actual extent to which vegetational fractionation occurs is of course dependent on the rate of plant uptake with respect to the rate of leaching. It has been demon- strated (Marchand, 1971) that chemical denudation by plant uptake and litter erosion in this area, although of much less significance than leaching in terms of total chemical removal, may be of considerable importance in the extraction of certain elements, especially metals such as Al, Fe, and Mn. It should also be remembered that elemental concentrations in vegetation represent large static quantities which are unavailable to leach- ing and which are potentially removable by sporadic accelerated erosion owing to fires or flooding, as well as by normal seasonal denudation. SUMMARY AND DISCUSSION The influences of erosion, deposition, vegetation, and microbial activity combine to further complicate an al- ready complex series of chemical and physical altera- tions (fig. 269) in the weathering processes by which unstable minerals are converted to stable forms under the prevailing surficial conditions, which may them- selves change with time. A complete analysis of all aspects of the weathering process is beyond the scope of this paper (the influences of microorganisms, for exam- ple, is not evaluated), but the data of the preceding pages suggest the direction and approximate mag- nitude of some physical, mineralogical, and chemical changes resulting from the weathering of dolomite and biotite adamellite in a present-day semiarid, subalpine climate. These changes involve alterations not only in the solid phases, but in the associated liquid and gas phases as well. Elements released by weathering are fractionated through the effects of plants, colloidal ex- change, and leaching. WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA WEATHERING 0F REED DOLOMITE Field observations and size analyses of soils and bed- rock indicate that the initial step of dolomite weather- ing in the White Mountains involves both physical and chemical breakdown of angular blocks produced by jointing. Polycrystalline soil grains, commonly bounded by cleavages, slowly accumulate from freeze-and-thaw disintegration of these blocks, except in areas of recrys- tallization near plutonic bodies; here postthermal con- traction may have created intergranular stresses, lead- ing to accelerated weathering and spheroidal forms. Chemical weathering is evident throughout the process of grain diminution, but appears to exert a greater effect in the interval between bedrock and soil gravel than between soil gravel and less-than-2-mm soil. Mineralogical decomposition, occurring in the sequence dolomite >> tremolite, epidote > talc, K-feldspar, plagioclase, biotite 2 apatite > quartz and ilmenite (fig. 294), resulted in percentage chemical losses, from 7 bedrock to soil, of Mg > Ca > Sr > Mn z' Fe (fig. 301). Pure calcite, apparently precipitated during periods of desiccation and partially dissolved during seasonal influx of meteoric waters, occurs as the single verifiably authigenic soil phase, in rinds on spheroidally weather- ing boulders, on surfaces of soil fragments, and as a discontinuous calcium carbonate soil crust. Relative mobilities of spring waters with respect to the dolomite bedrock yield the sequence Mg > Ca >> Fe > Mn (table 22). These findings agree with losses from bedrock to soil, except that they show Fe to be more easily mobilized from dolomite than Mn. Dolomite soil water extracts differ notably in composition from both spring and precipitation waters, probably owing in part to secondary calcite, ash fragments, and local contam- inants in the soils. The soil water extracts have pH’s much closer to the spring waters than to rain or snow, suggesting a fairly rapid adjustment to soil composition in these soluble soils. Comparisons of ion-activity prod- ucts with solubility constants indicate that waters draining the terrane are undersaturated with respect to both calcite and dolomite. Calculated carbon dioxide partial pressures for dolo- mite-derived waters exceed atmospheric values by more than an order of magnitude and those of waters from the adamellites by several times. The excesses above at- mospheric PC02 are presumably created by a combina- tion of root respiration, humus oxidation, and other organic processes; the greater PCOz values in the dolo- mite waters must primarily reflect the change in vege- tation with substrate, not solution of carbonate. WEATHERING OF ADAMELLITE OF SAGE HEN FLAT Weathering in the Sage Hen pluton occurs in several stages. Physical disintegration proceeds by frost riving 417 along intergranular weaknesses caused by cooling and contraction of the rock or possibly by swelling of biotite during initial weathering. Boulders exfoliate and grus accumulates in the A11 soil horizon. This coarse mater- ial is largely unaltered chemically or mineralogically except for minor Fe oxidation; the principal chemical attack occurs during the transformation of grus to finer sized particles. Primary minerals weather in the se- quence plagioclase (An2g30) > hornblende > biotite, epidote > microcline, plagioclase (AnuH5), allanite > apatite, chlorite, magnetite > sphene, quartz, musco- Vite, ilmenite > zircon (fig. 297), resulting in percen- tage chemical losses, from fresh rock to soil, in the ., sequence Rb $ Naz Kz Mg > Sr > an Ca > Ba > Si > A1 >> Fe > Ti (fig. 301). X-ray diffractometer and electron microprobe results suggest that kaolinite, and possibly small amounts of vermiculite and montmoril- lonite, are forming as authigenic alteration products of feldspars, biotite, and other silicates. Microprobe ‘7 analyses of biotite indicate losses in the sequence Ba > K > Mg > Fe > Si > A1 (table 8); cations in eightfold to twelvefold coordination appear to be lost most readily, followed by octahedral cations, and finally elements in sixfold and fourfold coordination. Microcline shows some losses of both Na and K, and plagioclase (Ammo) shows changes that imply removal in the sequence Na > Ca > Si > A1 (table 10; fig. 300). Relative mobilities of chemical species in adamellite spring waters indicate that changes with regard to bed- rock occur in the sequence Ca > Mg 2 Na >> Si: K 2 Mn: Fe > A1 (fig. 306). Although this order shows some agreement with bedrock-to-soil losses, it is not entirely consistent with conclusions based on the solid phases alone. The low K mobility is probably the result of plant extraction and fixation within the soil. Plant uptake and litter erosion may explain the low Al con- tent of spring waters. It is suggested that Mg and espe- cially Ca are augmented in waters draining the pluton owing to the solution of fine-grained carbonate particles continually blown into the adamellite terrane. If this latter hypothesis is correct, the chemistry of ground waters is extremely sensitive to the presence of small quantities of soluble materials, including those foreign to the area. The fact that waters from two springs drain- ing the adamellite can be chemically distinguished is another indication of such sensitivity. Both soil and spring waters associated with the pluton plot within the kaolinite stability field (figs. 307, 308) and are super- saturated with quartz but undersaturated with respect to colloidal silica. Proportions of solutes in adamellite soil water extracts appear to be roughly intermediate between precipitation and spring water in the adamel- lite terrane. 418 GENERAL WEATHERING RELATIONSHIPS Observations in the White Mountains, as well as in warmer and wetter regions, suggest that physical weathering of a given lithology may be chiefly a func- tion of the nature and spacing of initial physical weak- nesses, particularly intergranular openings, cleavages, fissility, and planar weaknesses, especially joints, which may serve as avenues for fluids. The process of physical breakdown, which tends to precede chemical weathering, may be as closely related to lithologic fea— tures as to external conditions. Thermal history or in- herent properties such as grain size, grain adhesion or cohesion, crystallographic properties of mineral con- stituents, bedding characteristics, or jointing may be more critical in affecting rates of physical weathering than climate. . Chemical weathering of any mineral is a function of the chemical composition and crystalline structure of the mineral and of the chemical environment in which weathering occurs. Ca, Mg, Fe, Mn, and Sr, for example, are all lost more readily from dolomite than from sili- cates and oxides in the adamellite. Dolomite-to-soil per- centage losses of Fe are about the same as for Mn, but adamellite Mn losses are much greater than those of Fe. In nonsoluble lithologies such as the adamellite of Sage Hen Flat, chemical losses tend to adhere to the general pattern of alkali metals > alkaline earth metals > nonmetals > metals. Precipitation water undergoes major chemical changes upon entering the soil, but significant addi- tional modifications occur during percolation to springs. The state of both soil and spring waters with respect to calcite, dolomite, and silica saturation preclude chemi- cal equilibrium in the natural waters studied. This re- sult is not surprising since chemical weathering, which supplies most of the constituents to the waters, is itself an open-system disequilibrium process. It is note- worthy, however, that although steady-state conditions are definitely not attained in soil waters, spring waters closely approximate constant composition. Contact with weathering materials over a considerable space- time interval is apparently necessary for the achieve- ment of a steady state, but this interval may be de- creased somewhat in the case of mobile constituents in readily weatherable minerals. The existence of steady- state spring water compositions indicates another area of applicability of the “dynamic equilibrium” concept suggested by Hack (1960) and others in regard to surficial processes. Elements removed from lattice sites in primary min- erals undergo fractionation by (1) formation of secon- dary phases—primarily clay minerals, calcium carbon- ate, and various insoluble precipitates, (2) adsorption on colloidal mineral or organic particles, (3) plant up- EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT take, and (4) solution. Quantitative information con- cerning fractionation of elements into secondary phases is not available for the two lithologies studied here (Al, Si, Fe, Mn, P, and Ca probably undergo fixation to some extent, in this form), but nine principal elements appear to apportion themselves between vegetation, colloidal exchange, and solution as follows (figs. 313, 314; table 25): Na, soln >> exch > veg Si, soln >> veg Ca, exch > soln > veg ‘7 Mg, exch > soln '= veg K, veg >> soln > exch Fe, Mn, P, veg >> soln Al, veg >>> soln Constituents entering authigenic phases or ex- changeable positions are retained within the soil and are removed only through soil erosion or change in chemical environment. Elements taken up by plants are recycled through the soil to the extent that decom- position takes precedence over litter erosion. Evidence presented elsewhere (Marchand, 1971) suggests that, in this area at least, most elements are eventually dis- solved and carried out of the soil by deep percolation. Results from the preceding study indicate that removal by plant uptake and litter erosion, although probably not significant in terms of total chemical denudation, may play some part in the total extraction of P and K from the bedrock and soil and undoubtedly accounts for appreciable parts of the chemical denudation of A1, Fe, and Mn. Another effect of vegetation in the weathering process may lie in the conversion of relatively insoluble constituents to more soluble chemical forms in which they could be carried out of the system. Plant uptake, litter fall, and decomposition may thus facilitate leach- ing, but data are lacking to evaluate the quantitative importance of this aspect of the geochemical cycle. REFERENCES CITED Anderson, G. H., 1937, Granitization, albitization, and related phenomena in the northern Inyo Range of California-Nevada: Geol. Soc. America Bull, v. 48, no. 1, p. 1—74. Anderson, G. H., and Maxson, J. H., 1935, Physiography of the north— ern Inyo Range [abs]: Geol. Soc. America Proc., 33d, Berkeley 1934, Cordilleran Sec., p. 318. Barnes, Ivan, 1964, Field measurement of alkalinity and pH: U.S. Geol. Survey Water-Supply Paper 1535—H, 17 p. 1965, Geochemistry of Birch Creek, Inyo County, California, a travertine depositing creek in an arid climate: Geochim. et Cos— mochim. Acta, v. 29, no. 2, p. 85—112. Bateman, P. C., and Wahrhaftig, Clyde, 1966, Geology of the Sierra Nevada, in Geology of northern California: California Div. Mines and Geol. Bull. 190, p. 107-172. Beaty, C. B., 1959, Slope retreat by gullying [California-Nevada]: Geol. Soc. America Bull., v. 70, p. 1479—1482. WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA 1960, Gradational processes in the White Mountains of California and Nevada: Ph.D. thesis, Univ. California, Berkeley, 260 p. Bower, C. A., Reitemeier, R. F., and Fireman, M., 1952, Exchangeable cation analysis of saline and alkali soils: Soil Sci., v. 73, no. 4, p. 251—261. Bower, C. A., and Wilcox, L. V., 1965, Soluble salts, in Methods of soil analysis; Part 2, Chemical and microbiological properties: Madi- son, Wis., Am. Soc. Agronomy (Agronomy, no. 9), p. 933—951. Brewer, Roy, 1964, Fabric and mineral analysis of soils: New York, John Wiley and Sons, Inc., 470 p. Cady, J. G., 1951, Rock weathering and soil formation in the North Carolina Piedmont region: Soil Sci. Soc. America Proc., v. 15, p. 337—342. Carmichael, I. S. E., 1967, The iron-titanium oxides of salic volcanic rocks and their associated ferromagnesian silicates: Contr. Mineralogy and Petrology, v. 14, no. 1, p. 36—64. Cloud, P. E., Jr., and Nelson, C. A., 1966, Phanerozoic-cryptozoic and related transitions—new evidence: Science, v. 154, no. 3750, p. 766—770. Dalrymple, G. B., 1963, Potassium-argon dates of some Cenozoic volcanic rocks of the Sierra Nevada, California: Geol. Soc. America Bull., v. 74, no. 4, p. 379—390. Deer, W. A., Howie, R. A., and Zussman, J., 1962a, Rock-forming minerals; Volume 3, Sheet silicates: New York, John Wiley and Sons, Inc., 270 p. 1962b, Rock-forming minerals; Volume 1, Ortho- and ring- silicates: New York, John Wiley and Sons, Inc., 333 p. 1963a, Rock-forming minerals; Volume 2, Chain silicates: New York, John Wiley and Sons, Inc., 379 p. 1963b, Rock-forming minerals; Volume 4, Framework sili- cates: New York, John Wiley and Sons, Inc., 435 p. Emerson, D. 0., 1966, Granitic rocks of the Mt. Barcroft quadrangle, Inyo batholith, California-Nevada: Geol. Soc. America Bull., v. 77, no. 2, p. 127—152. Feth, J. H., Roberson, C. E., and Polzer, W. L., 1964, Sources of mineral constituents in water from granitic rocks, Sierra Nevada, California and Nevada: US. Geol. Survey Water-Supply Paper 1535—I, 70. p. Feth, J. H., Rogers, S. M., and Roberson, C. E., 1964, Chemical com- position of snow in the northern Sierra Nevada and other areas: US. Geol. Survey Water-Supply Paper 1535—J, 39. p. Folk, R. L., 1965, Petrology of sedimentary rocks: Austin, Texas, Hemphill’s, 159 p. Gallick, C. M., 1964, Geology of a part of the Blanco Mountain quad- rangle, Inyo County, California: M. A. thesis (unpub.), Univ. California, Los Angeles. Garrels, R. M., and Christ, C. L., 1965, Solutions, minerals, and equilibria: New York, Harper and Row, 450 p. Goldich, S. S., 1938, A study in rock-weathering: J our. Geology, v. 46, p. 17—58. Hack, J. T., 1960, Interpretation of erosional topography in humid temperate regions (Bradley volume): Am. Jour. Sci., v. 258—A, p. 80—97. Hall, M. L., 1964, Intrusive truncation of the Precambrian-Cambrian succession in the White Mountains, California: M. A. thesis, Univ. California, Berkeley, 90 p. Harned, H. S., and Davis, Raymond, J r., 1943, The ionization constant of carbonic acid in water and the solubility of carbon dioxide in water and aqueous salt solutions from 0° to 50° 0.: Am. Chem. Soc. Jour., v. 65, p. 2030—2037. Harned, H. S., and Scholes, S. R., Jr., 1941, The ionization constant of H003 from 0° to 50° G: Am. Chem. Soc. Jour., v. 63, p. 1706—1709. Hay, R. L., 1959, Origin and weathering of late Pleistocene ash de- posits on St. Vincent, B. W. I.: Jour. Geology, v. 67, no. 1, p. 65—87. 419 1966, Zeolites and zeolitic reactions in sedimentary rocks: Geol. Soc. America Spec. Paper 85, 130 p. Helley, E. J ., 1966, Sediment transport in the Chowchilla River basin: Mariposa, Madera, and Merced Counties, California: Ph. D. thesis (unpub.), Univ. California, Berkeley, 153 p. Hsu, K. J ., 1964, Solubility of dolomite estimated on the basis of the chemical composition of Florida ground waters [abs]: Geol. Soc. America Spec. Paper 76, p. 8485; Ground Water, V. 2, no. 1, p. 56. Johnson, C. M., and Ulrich, Albert, 1959, Analytical methods for use in plant analyses: California Agr. Expt. Sta. Bull. 766, Part 2, p. 26—7 7. Jones, L. H., and Thurman, D. A., 1957, The determination of aluminum in soil, ash, and plant materials using Eriochrome Cyanine R. A.: Plant and Soil, v. 9, p. 131—142. Kesseli, J. E., and Beaty, C. B., 1959, Desert flood conditions in the White Mountains of California and Nevada: US Army Quar- termaster Research and Eng. Center, Tech. Rept. EP—108, 107 p. Kilmer, V. J., 1965, Silicon, in Methods of soil analysis; Part 2, Chemical and microbiological properties: Madison, Wis., Am. Soc. Agronomy (Agronomy, no. 9), p. 959—962. Knopf, Adolf, 1918, A geologic reconnaissance of the Inyo Range and the eastern slope of the southern Sierra Nevada, California: US. Geol. Survey Prof. Paper 110, 130 p. Krauskopf, K. B., 1968, A tale often plutons: Geol. Soc. America Bull., v. 79, no. 1, p. 1—18. 1971, Geologic map of the Mount Barcroft quadrangle, California-Nevada: U.S. Geol. Survey Geol. Quad. Map GQ—960. LaMarche, V. C., Jr., 1967, Spheroidal weathering of thermally metamorphosed limestone and dolomite, White Mountains, California, in Geological Survey research, 1967: US. Geol. Sur- vey Prof. Paper 575—C, p. 032—037. 1968, Rates of slope degradation as determined from botanical evidence, White Mountains, California: US. Geol. Survey Prof. Paper 352—1, p. 341—377. Larson, T. E., and Buswell, A. M., 1942, Calcium carbonate saturation index and alkalinity interpretations: Am. Water Works Assoc. Jour., v. 34, no. 11, p. 1667—1684. McKee, E. H., and Nash, D. B., 1967, Potassium-argon ages of granitic rocks in the Inyo batholith, east-central California: Geol. Soc. America Bull., v. 78, no. 5, p. 669—680. Major, Jack, 1967 , Potential evapotranspiration and plant distribu- tion in Western States with emphasis on California, in Shaw, R. H., ed., Symposium on ground level climatology, 1965, Berkeley, Calif: Am. Assoc. Adv. Sci. Pub. 86, p. 93—126. Manov, G. G., Bates, R. G., Hamer, W. J., and Acree, S. F., 1943, Values of the constants in the Debye-Huckel equation for activity coefficients: Am. Chem. Soc. Jour., v. 65, p. 1765. Marchand, D. E., 1968, Chemical weathering, soil formation, and geobotanical correlations in a portion of the White Mountains, Mono and Inyo Counties, California: Ph. D. thesis, Univ. Califor- nia, Berkeley, 376 p. 1970, Soil contamination in the White Mountains, eastern California: Geol. Soc. America Bull., v. 81, no. 8, p. 2497—2505. 1971, Rates and modes of denudation, White Mountains, east- ern California: Am. Jour. Sci., v. 270, p. 109—135. 1973, Edaphic control of plant distribution in the White Moun- tains, eastern California: Ecology, v. 54, no. 2, p. 233—250. Markam, A. A., and Kobe, K. A., 1941, The solubility of carbon dioxide and nitrous oxide in aqueous salt solutions: Am. Chem. Soc. J our., v. 63, p. 449. Maxson, J. H., 1935, Pre-Cambrian stratigraphy of the Inyo Range [abs]: Geol. Soc. America Proc., 33d, Berkeley 1934, Cordilleran Sec., p. 314. Mead, W. J ., 1915, Occurrence and origin of the bauxite deposits of Arkansas: Econ. Geology, v. 10, p. 28—54. 420 Morrison, R. B., 1964, Lake Lahontan: Geology of southern Carson Desert, Nevada: US. Geol. Survey Prof. Paper 401, 156 p. Munz, P. A., and Keck, D. D., 1959, A California flora: Berkeley, Univ. California Press, 1681 p. Nelson, C. A., 1962, Lower Cambrian-Precambrian succession, White-Inyo Mountains, California: Geol. Soc. America Bull., v. 73, p. 139—144. 1963, Preliminary geologic map of the Blanco Mountain quad- rangle, Inyo and Mono Counties, California: US. Geol. Survey Mineral Inv. Field Studies Map MF—256. 1966, Geologic map of the Blanco Mountain quadrangle, Inyo and Mono Counties, California: US. Geol. Survey Geol. Quad. Map GQ—529. Nelson, C. A., and Perry, L. J ., 1955, Late Precambrian-early Cam- brian strata, White-Inyo Mountains, California [abs]: Geol. Soc. America Bull., v. 66, no. 12, Part 2, p. 1657—1658. Nicholle, J. W., 1965, Summary of the climatic conditions and geologic history of the White Mountains region, California: Space Sci. Lab. Pub., Univ. California, Berkeley. Olsen, S. R., and Dean, L. A., 1965, Phosphorus, in Methods of soil analysis; Part 2, Chemical and microbiological properties: Madi- son, Wis., Am. Soc. Agronomy (Agronomy, no. 9), p. 1035—1049. Pace, Nello, 1963, Climatological data summary for the decade 1 January 1953 through 31 December 1962 from the Crooked Creek Laboratory (10,150 ft) and the Barcroft Laboratory (12,470 ft): Berkeley, Univ. California, White Mountain Research Sta. rept., 52 p. Pittman, E. D., 1958, Geology of the northwestern portion of the Blanco Mountain quadrangle, California: M. S. thesis (unpub.), Univ. California, Los Angeles, 103 p. Powell, D. R., 1963, The physical geography of the White Mountains, EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT California-Nevada: M. A. thesis (unpub.), Univ. California, Ber- keley. Rosenblum, Samuel, 1956, Improved techniques for staining potash feldspars: Am. Mineralogist, v. 41, nos. 7—8, p. 662—664. Ruhe, R. V., 1967, Geomorphic surfaces and surficial deposits in southern New Mexico: New Mexico Bur. Mines and Mineral Re- sources Mem. 18, 65 p. Sherman, G. D., and Uehera, Goro, 1956, The weathering of olivine basalt in Hawaii and its pedogenic significance: Soil Sci. Soc. America Proc., v. 20, p. 337—340. Spurr, J. E., 1903, Descriptive geology of Nevada south of the fortieth parallel and adjacent portions of California: US. Geol. Survey Bull. 208, p. 206—212. Taylor, M. E., 1966, Precambrian mollusc-like fossils from Inyo County, California: Science, v. 153, no. 3732, p. 198—201. Taylor, R. L., 1965, Cenozoic volcanism, block faulting, and erosion in the northern White Mountains, Nevada: M, A. thesis (unpub.), Univ. California, Berkeley, 95 p. Wahrhaftig, Clyde, 1965, Stepped topography of the southern Sierra Nevada, California: Geol. Soc. America Bull., v. 76, no. 10, p. 1165—1190. Waring, R. H., and Hermann, R. K., 1966, A modified Piche evaporimeter: Ecology, v. 47, no. 2, p. 308—310. Warshaw, C. M., and Roy, Rustum, 1961, Classification and a scheme for the identification of layer silicates: Geol. Soc. America Bull., v. 72, p. 1455—1492. Wolfenden, E. B., 1965, Geochemical behavior of trace elements dur- ing bauxite formation in Sarawak, Malaysia: Geochim. et Cos- mochim. Acta, v. 29, p. 1051—1062. Wooley, J. T., and Johnson, C. M., 1957, Silicon determination in ashed plant material: Agr. and Food Chemistry, v. 5, no. 11, 872. p. SUPPLEMENTAL INFORMATION 422 METHODS, REPRODUCIBILITY, AND ACCURACY Field procedures for the collection of data and samples have been discussed in the text. Several temporary field impregnations were obtained from soil pits with Quickmont (available from E. V. Roberts and Associates, Culver City, Calif.) and later fully impregnated under vacuum with leucite. The advantage of the field impregnation is that natural orientation and textures are maintained. Bedrock and soil samples were treated in the laboratory as outlined in figures 315 and 316.7 Fresh bedrock chips and soil samples (soil gravel, less than 2 mm) were crushed with a tungsten carbide mortar and pestle and ground to —50 mesh in a tungsten carbide ball mill. Each fraction was then split and made into an X-ray fluorescence pellet. The values of table 27 (top), representing chips of the same sample, ground in different mortars and separately prepared, indicate no significant source of error in the grinding, splitting, or pellet-making operations. The replicate analyses of the table suggest precision of about : 0—6 percent of the analyzed value and generally below 2 percent for values above 2.0 percent. Table 28 compares X-ray fluorescence results with those of wet chemistry (unpub. data of Ken-ichiro Aoki) and those for Na and K by flame photometer (by Joaquin Hampel; see “Introduction”). Agreement is good in some cases and very poor in others. Lighter elements such as Na, Mg, A1, and Si are not as easily excited by X—rays as heavier elements and are most susceptible to error. Where samples approximate the composition of well-analyzed standards, however, X—ray results appear to be both precise and accurate. The values of 7The following parameters were determined for the less-than-Z-mm fraction of many soil samples for purposes outside the scope of this paper: Phosphorus in 10:1 water extracts, permanent wilting point (soil moisture at 15-atm tension), moisture equivalent (soil moisture at 1 /3-atm tension), available water (mositure equivalent minus permanent wilting point), percentage carbonate, and total carbon. EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT BEDROCK SAMPLE Removal of weathered portions (rock saw) I Crushed,ground lAdamellite only) I (Tungsten carbide, agate mortars, <50 mesh) Sieved Th’ . (SO-230' mesh) In selctions Separated X-ray Staln|2j§ft$smm (two heavlyliquids) fluorescence Ip Grain mounts pellets Pornt Countlng (line counted) . I Chemical . . Mineralogic analyses Iadamel- analyses MmeraIOQ'C ana'YSES (quartz, potassium feldspar,plagioclase. biotite) lite: accessory minerals; do- lomite: all minerals Grains for microprobe, im- mersion oils, U stage FxGURE 315.—Flowsheet for laboratory treatment of bedrock samples. Hampel were used for Na and K, but all other figures were taken from X-ray determinations (except values of Aoki used as X-ray standards). Thin sections of five adamellite bedrock samples were stained for K-feldspar (Rosenblum, 1956) and point counted to determine quartz, K-feldspar, plagioclase, and biotite percentages. Mineralogical analyses of crushed bedrock (50—230 mesh) and less-than-Z-mm-sand (greater than 62 um, crushed to -50 mesh), fine sand and very fine sand fractions of soils were obtained from line counts of grain mounts after removal of organic matter with 2 percent hydrogen peroxide (soils only) and two successive heavy liquid separations. For the dolomite samples, liquids of specific gravity 2.69 (bromoform plus alcohol) and 2.95 (tetrabromoethane) were used; liquids having den- sities of 2.88 (bromoform) and 3.3 (methylene iodide) were used for adamellite samples. A minimum of 900 grains, and usually more than SOIL SAMPLE l r t Impregnated Sieved (<2mm) thin section V I I I I | I i i Crushed,ground (tung- NH, Ac Saturation 10:1 extraction Furnace Removal of Sieve and pipette 1/3-atmosphere 15-atmosphere sten carbide, agate extraction paste (water) oxidation organic matter analyses moisture moisture mortars, <50 mesh) * t + (H202) Extractable cations P Total C SIZE ‘ distribution ‘ Percentage pH Water-soluble carbonate cations Sieved Sieved I H202 I I I >2mm <2mm <62pm >62,um Fine sand Very fine sand 2-62,um X-ray fluorescence pellets Two heavy-quuid separations, grain mounts, line counting Quantitative X-ray I chemical I Total analyses diffraction I I Mineralogical analyses Grains for microprobe analyses, oil immersion, Ustage FIGL'RE 316,—Flowsheet for laboratory treatment of soil samples. WEATHERING, SOILS, GEOCHEMICAL FRACTIONATION, WHITE MOUNTAINS, CALIFORNIA TABLE 27 .—Reproducibility of X-ray fluorescence analyses [Values yielded by opposite sides of the same pellet, except sample 58] Fe Mn Ca Sr Rb (weight percent) 58 (agate) ______________ 58 (tungsten carbide)____ 47 soil 47 soil __ 52 soil 52 soil 64 soil 64 soil 66 soil 66 soil 68 soil 1,500 grains, per sample was counted. The statistical analysis of Brewer (1964, p. 46—50) indicates that counts of 1,500 grains yield results within about 5 percent probable error for major constituents and less than 15 percent for minor minerals. Mineral densities used to compute weight percentages from grain mounts were obtained by direct measurement or from chemical composition information. Line-counted samples were used as standards for quantitative X-ray diffraction analyses of silt fractions. Both standards and sam- ples were finely ground in a mortar, mounted on glass slides with the aid of acetone, and X-rayed under identical conditions on the same day. Abundances were based on peak heights above background. Glass was assumed to make up the difference between totals and 100 percent, an assumption verified by many visual estimations of silt- glass percentage under the polarizing microscope. Probable error is estimated at less than 15 percent of the given value. Laboratory soil pH measurements were made with a Beckman Zeromatic pH meter on water-saturation pastes. Replicate analyses of samples yielded results within 0.1 pH unit. Similar pastes were ex- tracted by the Buchner funnel method of Bower and Wilcox (1965, p. 423 935). Replicate analyses of 19 saturation extract samples indicate reproducibility at the level of :25 percent of the stated value. Soil samples were extracted with NH4Ac solutions at pH 7.0, using the centrifuging method of Bower, Reitemeier, and Fireman (1952, p. 253), and analyzed for total extractable Na, K, Ca, and Mg. Results are reproducible to within 20 percent of the stated values. Na and K in water and N I-LAc extracts were analyzed by flame photometry. Ca, Mg, and Fe in soil extracts were determined by atomic absorption spectrophotometry. Si was analyzed using colorimetric procedures described by Kilmer (1965, p. 959—962). P in water-saturation ex- tracts was determined by the methods of Olsen and Dean (1965, p. 104—107). Plant samples from the field and greenhouse were washed with distilled water and ovendried at 104°C. Into 30-milliliter digestion flasks, 02000—05000 grams of each dry sample was weighed. Five milliliters of concentrated HNO3 and HCLO4 were added, and the samples were digested overnight in a sand bath. The samples were then heated on Kjeldahl burners until they appeared colorless and HCLO4 fumes were evident. The solutions were cooled and diluted to 100 milliliters in volumetric flasks with 10 milliequivalents per liter SrClz (to mask interferences). Aliquots were taken for subsequent analysis. K, Na, Ca, Mg, Fe, and Mn were analyzed by atomic absorp- tion spectrophotometry. P was determined colorimetrically by the molybdenum blue method described by Johnson and Ulrich (1959), using stannous chloride as a reducing agent. Al was determined by the Eriochrome Cyanine R. A. method of J ones and Thurman (1957). The methods of Wooley and Johnson (1957, p. 872) were employed in the dry ashing and Si analysis of plants. Absolute limits of precision are :20 percent of the given value, but analytical results are within :10 percent in all but a few cases. The preceding discussion has been confined to methods and preci- sion of measurements for any given sample. Statistical computer data (table 29) permit a comparison of various physical and chemical soil parameters with regard to variation within a given soil type. Geo- graphic patterns of variation within and between soil types were not assessed. The figures of table 29, based on samplings of from 12 to 25 sites per soil type, indicate seven general groups of soil parameter variation for this area, decreasing in reliability as follows: TABLE 28.—Comparison of analytical results for some adamellite of Sage H en Flat (samples 92, 98) and Reed Dolomite samples, as analyzed by -ray fluorescence (XRF), wet chemistry (Ken-ichior Aoki), and flame photometry (Joaquim Hampel) 52 bedrock 52 soil, < 2 mm 54 bedrock 54 soil, < 2 mm 54 soil, > 2 mm XRF Aoki Hampel XRF Aoki Hampel XRF Aoki Hampel XRF Aoki Hampel XRF Hampel 0.16 ______ 41.19 23.01 ,,,,,, 0.55 0.81 111111 135.74 35.74 ______ 1.95 ______ Trace ____________ .44 __ .03 .03 ______ .47 .50 ______ .01 ,,,,,, .16 ______ 11.48 7.87 ,_ .40 .47 ______ ‘9.89 9.89 ...... .08 ,,,,,, .23 ______ 3.78 3.56 ______ .39 .39 ______ 13.14 3.14 ______ .53 ,,,,,, .04 ______ .21 .17 ______ .07 .05 ______ 1.16 .16 .09 21.29 ______ 6.96 8.86 ______ 20.75 21.76 ______ 18.44 8.44 18.54 31.24 ______ 9.16 22.09 ______ 29.74 30.29 ______ 114.14 14.14 30.92 .03 0.04 1.00 .46 0.48 <.01 .08 0.04 .72 .16 <.01 02 .02 1.81 .82 .98 .01 .07 .03 ______ 1.63 .07 68 bedrock 68 soil, > 2 mm 64 bedrock 92 bedrock 92 soil, < 2 mm 98 soil, < 2 mm XRF Hampel XRF Hampel XRF Aoki XRF Aoki XRF Aoki XRF Aoki Hampel ______ 1.49 “”1, 0.75 0.59 65.83 67.20 65.93 65.40 66.93 69.65 "mm ______ .04 "c"- ______ 0.03 ""11 .53 ______ .63 .50 .43 ,__,,_ ______ .61 ”1,1- .19 .12 _,__-_ 16.33 ,v___ 15.77 ______ 14.57 “.1“ ,,,,,, .51 ""11 .33 .46 3.96 3.12 4.73 3.54 2.67 2.50 ___,__ ______ .12 .03 .03 .10 .06 .11 .05 .04 .04 ___,,_ ______ 19.83 7.20 7.42 1.05 .77 11.15 1.15 .79 .65 ",1" ______ 3041 ,v__-_ 3969 46.87 2.37 2.86 2.14 2.32 1.60 1.63 ____,_ 03 <01 .08 <.01 .07 3.44 3.73 3.31 3.35 2.84 3.24 3.39 01 .08 .07 .05 .06 3.88 .11 3.77 3.82 3.69 3.53 3.63 ‘Valve of Aoki used for standard. 424 EROSION AND SEDIMENTATION IN A SEMIARID ENVIRONMENT TABLE 29.-——Variation in measured parameters of four White Mountains soil types Adamellite of Andrews Mountain Reed Dolomite Basalt Sage Hen Flat Sandstone Range of percent standard Percent Percent Percent Percent devration standard standard standard standard Number of deviation Number of deviation Number of deviation Number of deviation samples from mean samples from mean samples from mean samples from mean Total extractable K _____________________________________ 25 5] 18 20 22 37 19 26 20—51 Total extractable Na _ 25 31 18 20 22 62 19 23 20—62 Total extractable Ca _ 25 41 18 16 22 32 19 22 16—14 Total extractable Mg ______________ 25 40 18 23 22 31 19 18 1&40 K in water-saturation extracts ._ 21 78 17 42 21 83 12 55 42—83 Na in water-saturation extracts__ 21 35 17 53 21 28 12 59 28—59 Ca in water—saturation extracts _, 21 39 17 49 21 4O 12 51 39—51 Mg in water-saturation extracts__ 21 36 17 42 21 36 12 37 36—42 Percent exchangeable K ______ — 21 56 17 19 21 47 12 31 19—56 Percent exchangeable Na___1 21 37 17 19 21 33 12 17 17—37 Percent exchangeable Ca ,1 21 9 17 2 21 5 12 3 2- 9 Percent exchangeable Mgdn 21 40 17 12 21 27 12 13 12-40 Sum of exchangeable cations 21 32 17 15 21 29 12 14 14-32 Pin 10:1 water extracts ___- 18 73 17 67 21 59 15 111 59—111 High field pH ................................................... 25 1 18 s 22 s 19 s 1- 8 LOW field H _____________________________________________________ 25 5 18 5 22 7 19 8 5— 8 Mean fiel pH ______ 7 25 1 18 7 22 8 19 8 1— 8 Laboratory pH __________ _ 25 3 18 4 22 5 17 7 3- 7 Soil moisture at Va-atm tenswn ,,,,,, _ 17 28 16 9 22 16 16 14 9—28 Soil moisture at 15-atm tension _____________ , 23 45 17 24 22 25 17 30 24—45 Available water _______________________ _ 17 28 16 21 22 12 15 18 12-28 Soil depth on topographic highs _ ....... 24 27 17 22 21 22 17 14 14—27 Bulk density _________________ .— ________________ 24 34 17 20 21 17 __ __ __ _, 17—34 Percent fragments greater than 2 mm ________________________ 14 20 A___ ____ 14 61 19 16 16-61 . Laboratory and field pH [Or—I graphic highs . Exchangeable cations, bulk density . Percentage of soil gravel fragments . Cations in water-saturation extracts . 10:1 available phosphorus \IQUIACD . Total extractable cations, 15-atmosphere moisture presented here would suggest that although a few pH and . Available water, 1/3-atmosphere moisture, soil depth on topo- 1xii-atmosphere moisture determinations may suffice to characterize properties of a soil series as a whole, most other parameters range widely within a given soil type and cannot be accurately evaluated by a few analyses. The White Mountains soils are, of course, lithosols and therefore may show more variation than finer textured soils formed on alluvium or strongly developed soils on almost any parent material. Nonetheless it is unlikely that analyses at one type locality are Many soil surveys and soil descriptions include only one or two sufficient to describe accurately properties of a soil series having any analyses of these soil properties for each soil series. Information appreciable lateral extent. 1" U-S. GOVERNMENT PRINTING OFFICE: 1974—543—583/ 93