* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Factors Influencing Infiltration and Sediment Productionof Semiarid
Survey
Document related concepts
Transcript
WATER RESOURCES RESEARCH VOL. II, NO.6 DECEMBER 1975 Factors Influencing Infiltration and Sediment Production of Semiarid Rangelands in Nevada W. H. BLACKBURN Range Science Department. Texas A & M University. College Station. Texas 77843 Simulated rainfall was used to study infiltration rates and sediment production of28 plant communities and soils of five watershed areas in central and eastern Nevada. Two antecedent moisture conditions were used: soil initially air dry and initially at field capacity. InfiltratIOn rates and sediment production of the various soils are largely controlled by extent and surface morphology of dune interspace soils. Vesicular horizons are unstable in dune interspace surface soils. These horizons seldom occur in coppice dunes or in well-aggregated dune interspace soils. Infiltration rate is negatively related, and sediment production positively related to the occurrence and morphology of vesicular horizons. More sediment is produced from soils with antecedent moisture initially at field capacity than from'initially dry soil because of the instability of vesicular horizons when the soils were saturated. High-intensity summer thunderstorms in the Great Basin account for most of the runoff and sediment production from rangelands. Runoff is the major force initiating soil movement and transporting sediment. In order to manage rangeland watersheds effectively, factors influencing infiltration and sediment production need to be understood. Infiltration and erosion on rangelands have been studied for some time. Review of the literature includes Chapline [1929], Forsling [1932], U.S. Department of Agriculture [1940], Harper [1953], Gifford [1968], and Branson et al. [1972]. However, for Nevada rangeland there is practically no informatIOn on infiltration and sediment production or on the influence of dune interspace areas and associated vesicular horizons on infiltration and sediment production. The objective of this study was to determine which ground cover and soil parameters most influence infiltration and sediment production. METHODS Twenty-eight study sites were selected within five watersheds (Table I). Nine sites were located in the Duckwater Watershed, six sites in the Coils Creek Watershed, six sites in the Steptoe Watershed, and seven sites in the Pine and Mathews Canyon Watersheds. The watersheds were sampled at the following times: Duckwater and Steptoe, June I-July 15; Coils Creek and Pine and Mathews Canyon, July 16-September I. Study sites were selected for their accessibility, repetition over large areas in the Great Basin, and vegetation and soil properties. Each site was located on a typical area with a different plant community and/or soil. Coppice dune is the area of accum ulation of litter and soil under shrubs and bunch grasses. The infiltrometer and methods of application are described by Blackburn et al. [1974]. In the Great Basin, summer thunderstorms usually occur on dry soil, although some occur on soil that is at or near field capacity. An application rate of 3 in./h for a duration of onehalf hour was applied to two antecedent moisture conditions, soil surface horizon initially air dry and soil surface horizon initially at field capacity. Basically, infiltration was defined for any point in time as the difference between total water applied and total runoff. Two types of runoff plots were used, i.e., 3 X 3 ft and variable. Regular plots were situated so the same mean Copyright © 1975 by the American Geophysical Union. percent coppice dune that occurred on the site also occurred in the plot. Variable plots were located to contain approximately 100% dune interspace area or coppice dune area. Variable plots were used only on sites that had obvious dune interspace and coppice differences. Each treatment was replicated 6 or 8 times. Sediment production was determined from a 30-fluid-oz runoff sample and other sediment trapped in the collection apparatus. Suspended sediment was allowed to settle in the laboratory and the water drained off. Samples were ovendried, weighed, and converted to sediment in pounds per acre. Plant cover and composition, litter, rock (1-2 in. and >2 in. in largest dimension), and bare ground on each plot were measured with a point frame [Levy and Madden. 1933]. Coppice dunes w~thin the plots were measured in two directions, and the ellipse formula used to compute cover. Dune interspace areas were computed as the complement of coppice dune. A soil profile description was made at each study site according to the procedures outlined by U.S. Department of Agriculture [1951, 1960]. Family-level identifications were made in accordance with the Soil Conservation Service [U.S. Department of Agriculture. 1960, 1970]. The surface horizon (from the surface to some diagnostic subsurface horizon) was rated from I to 5 on the basis of structure and vesicular pores (Table 2). Likewise, each soil was classified as to hydrologic group [U.S. Department of Agriculture. 1964]. Bulk density and initial soil moisture content were determined by using the Troxler surface moisture-density gage. Bulk density was determined for surface 2-, 4-, and 6-in. depths of dune interspace areas and coppice dunes. Soil moisture before the dry infiltration test was taken concurrent with bulk density readings. To prevent plot disturbance, these readings were taken on areas adjacent to plots. In order to obtain soil moisture before the wet run an additional plot was sprinkled, covered with polyethylene plastic, and left for 24 hours before the moisture reading was taken. Particle size distribution of soil from each horizon, including surface dune interspace areas and coppice dunes, was measured by the hydrometer method [Bouyoucos. 1962]. Organic carbon content of soil from each horizon including surface dune interspace areas and coppice dunes was determined by a hightemperature induction furnace [Black. 1965]. Each study site was classified as to its physiographic place- 929 \C) ..... o TABLE I. Location Elevation 30 airline mi south and east of Eureka, mostly in White Pine County, Nevada. Site, 100 mi' [Blackburn et al., 1968]. Highest peak is around 7300 ft and the basin outlet is 4800 ft. Description of Study Sites Climate Geology and Soils Dominant Vegetation Duckwater Watershed Annual precipitation 7.8-13.7 in., mostly as snow. Temperature at the lower elevations ranges from a low of - 34 ° F to a high of99°F, with a mean annual temperature of 43° F. Volcanics and sedimentaries, i.e., tuff, basalt, andesites and limestone. Aridisols and Entisols, i.e., Torrifluvents, Durorthids, Haplargids, Durargids, and Natragids. Black sagebrush· (A rtemisia nova), big sagebrush (Artemisia triden/ata), shadscale (Atriplex conferti/olia), winter fat (Eurotia lanata), green rabbit brush (Chrysothamnus viscidiflorus), Utah juniper (Juniperus osteosperma), single leaf pinyon (Pinus monophylla). t:I:l ,.... ;l> (l :>: til c: ;>; " Coils Creek Watershed 32 airline mi northwest Highest peak is around of Eureka in Eureka 8400 ft and the basin outCounty, Nevada. Site, 48 let is 6500 ft. mi' [Blackburn et ai, 1969a]. 24 airline mi southeast of Ely in White Pine County, Nevada. Site, 45 mi' [Heinze et al., 1966]. Highest peak is around 9081 ft and the basin outlet is 7100 ft. Annual precipitation 8.8-14.8 in., mostly as snow. Temperature ranges from a low of - 26° F to a high of 110°F, with a mean annual temperature of47°F. Volcanics and sedimentaTles, i.e., basalt, shale, sandstone and limestone. Entisols and Aridisols or Mollisols, i.e., Torriorthents, Camborthids, Haplustolls, Haploxerolls, Durixerolls, and Argixerolls. Steptoe Watershed Annual precipitation is 12 in., mostly as Limestone. Aridisols or Mollisols, i.e., snow. Temperature ranges from a low of Camborthids, Haplargids, Durargids, -26°F to a high of9r F, with a mean and Argixerolls. annual temperature of44°F. Low sagebrush (A rtemisia arbuscula), big sagebrush, snow berry (Symphoricarpos longiflorus), Utah juniper, single leaf pinyon, Sandberg bluegrass (Poa secunda), bluebunch wheatgrass (Agropyron spicatum), woolly wyethia (Wyethia mol/is), squirreltail (Sitanion hys/rix), arrowleafbalsamroot (Balsamorhiza sagil/a/a), diffused phlox (Phlox dijJusa). Highest peak is around 6700 ft and the basin outlets are approximately 5600 ft. Annual precipitation 11.9-21.8 in., mostly as snow during winter or rain in late summer. Temperature of the basins ranges from a low of OaF to a high of 101 0 F, with a mean annual temperature of50°F. Volcanics and old lake bed sediments, i.e., andesite, tuff, ignibrite, tuffaceous clay, sand, and silt. Aridisols of Mollisols with a few Entisols, i.e., Torriorthents, Durargids, Haplargids, Argixerolls, and Haploxerolls. -i "-i ;l> oZ ;l> Z " '" "3: en '"-iZ Big sagebrush, bitter brush (Purshia /riden/ata), Utah juniper, single leaf pinyon, bluebunch wheatgrass, crested wheatgrass (Agropyron desertorum). Pine and Mathews Canyon Watersheds 18 airline mi southeast of Caliente in Lincoln County, Nevada. Site, 66 mi' [Blackburn et al., 1969b]. Z :!l ,.... Big sagebrush, black sagebrush, rubber rabbitbrush (Chrysothamnus nauseosus), serviceberry (Amelanchier alni/olia), Utah juniper, single leaf pinyon, intermediate wheatgrass (Agropyron intermedium), squirreltail. '"1:1 "o " c: ~ oz 931 BLACKBURN: INFILTRATION AND SEDIMENT PRODUCTION TABLE 2. Rating Soil Surface Horizon Morphological Rating ment, i.e., mountain, smooth or dissected alluvial fan, or floodplain. Percent slope was obtained from an Abney level, and aspect from an 8-point compass. Surface roughness of each plot was determined with a microrelief meter as described by Kincaid and Williams [1966]. Soil Surface Horizon Morphological Description Single-grain sandy texture or well-aggregated granular structure without vesicular pores· Granular or massive structure having few vesicular po rest Massive or weak platy structure having common vesicular pores Moderate platy structure having common vesicular pores, or massive structure having many vesicular pores, or clayey and weakly structured Strong platy structure baving many vesicular pores, or clayey and massive 2 3 4 5 RESULTS AND DISCUSSION Infiltration The following discussion is based on regression equations (Table 3) where the dependent variables are average infiltration rates by treatment at the end of 30 min. Dependent and independent variables and the mean and standard deviation of the variables appearing in the regression equations are given in Table 4. Duckwater Watershed. The regression equations for this watershed are (I) and (2) in Table 3. Weighted percent silt is the most important variable explaining variation in infiltration. Plant cover and dune interspace areas are least important, the other variables being intermediate in their importance. ·Structures are defined as being granular (approximately spherical with no accommodation of faces to surrounding peds), platy (with vertical dimension small with regard to horizontal dimensions and faces accommodated with those of adjacent peds), or massive (no aggregation ). tIn classes of pore abundance the number of units per area of surface was 1-3 for few pores, 4-14 for common pores, and more than 14 for many pores. A unit is defined as a square inch for fine, very fine, and micropores and as a square yard for medium and coarse pores. TABLE 3. Equation Number Infiltration and Sediment Multiple-Regression Equations by Watershed and Treatment Watershed Antecedent Moisture Regression Equation Coefficient of Determination R2 I nfl/lration Duckwater dry Yl 2 Duckwater field capacity Y2 3 Coils Creek dry YI 4 Coils Creek field capacity Y2 5 Steptoe dry Y1 6 Steptoe field capacity Y2 7 Pine and Mathews Canyons dry Yl 8 Pine and Mathews Canyons field capacity Y2 9 Duckwater, Coils Creek, Steptoe, Pine and Mathews Canyons Duckwater, Coils Creek, Steptoe, Pine and Mathews Canyons dry YI field capacity Y2 10 0.00470(X9) + 0.00738(XI3) -0.128(XI8) - 0.033(X27) + O.l23(X36) = 2.865 - 0.00994(X9) - 0.OO213(XI0) + 0.OO258(XI3) - 0.147(XI8) - 0.0341(X27) + 0.118(X36) = 3.243 - 0.OOI21(X9) - 0.00835(XIO) -0.144(XI8) - 0.OO304(X27) = 3.386 - 0.OO230(X9) - 0.0144(XIO) -0.170(XI8) - 0.OO913(X27) = 1.948 + 0.951(X4) - 0.00367(XI0) -0.116(XI8) = -0.542 + 2.42(X4) - 0.011(X9) - 8.107(XIO) - 0.251(XI8) + 0.192(X36) = 3.987 - 0.00404(X9) - 0.OOI69(XI0) + 0.OO392(XI2) + 0.219(XI5) - 0.0457(X25) = 2.863 - 0.0255(X7) - 0.OO582(X9) - 0.00822(XIO) + 0.00390(XI3) = 3.121 - 0.00593(XIO) - 0.163(XI8) + 0.175(X20) - 0.OO931(X30) 0.723 = 3.074 - 0.0119(XIO) - 0.176(XI8) + 0.184(X20) - 0.00958(X30) 0.630 = 2.808 - 0.748 0.719 0.812 0.310 0.605 0.661 0.799 0.505 Sediment = 3.096 + 2.829(X4) + 1.274(X9) - 0.413(XI8) - 2.109(X24) + 3.555(X27) Y4 = 6.715 - 0.402(XI8) + 3.473(X27) II Duckwater dry 12 Duckwater field capacity 13 Coils Creek dry Y3 = 16.827 - 2.094(X4) + 0.000409(X9) + 0.OO75(XIO) - 0.37(X21) - 0.0874(X24) 14 Coils Creek field capacity Y4 = 9.415 - 15 Steptoe dry Y3 16 Steptoe field capacity Y4 17 Pine and Mathews Canyons dry Y3 18 Pine and Mathews Canyons field capacity Y4 Y3 - O.l35(X27) 1.266(X4) + 0.00783(XIO) - 0.398(X21 ) - 0.020(X24) = 6.222 - 1.636(XI0) - 1.012(XI2) - 0.853(X21) + 0.074(X27) = 4.535 - 2.954(X4) - 5.918(XI0) - 2.519(XI2) - 1.214(X21) + 0.IOI(X24), = 17.143 - 1.278(X4) - 0.506(XI8) - 1.00(X21) - 0.100(X24) - 8.895(X27) = 10.577 - 1.I09(X4) - 0.0098(X9) + 0.0042(XI0) - 0.537(XI8) - 0.064(X24) + 0.0489(X27) 0.258 0.256 0.267 0.376 0.203 0.416 0.235 0.241 BLACKBURN: INFILTRATION AND SEDIMENT PRODUCTION 932 TABLE 4. Mean Standard DeviatIOn and Units of Measure of the Independent Variables (X) and Units of Measure on the Dependent Variables (Y) Appearing in Regression Equations by Watershed Number X4 X9 XIO XI3 XI8 X24 X27 X36 X4 X9 XIO XI8 X21 X24 X27 X4 X9 XIO XI2 XI8 X21 X24 X27 X36 X4 X7 X9 XIO XI2 XI3 XI5 XIS X21 X24 X25 X27 XIO XI8 X20 X30 Variable Mean Duckwater Watershed soil bulk density in surface 0-4 in., dune 1.42 interspace bare ground 32.82 dune interspace 62.63 plant cover 23.25 soil surface horizon morphological 3.04 rating, weIghted sand fraction of soil surface horizon, 56.54 weighted 23.95 silt fraction in soil surface horizon, weighted 4.21 depth of surface horizon, weighted Coils Creek Watershed soil bulk density in surface 0-4 in., 1.35 dune interspace 22.41 bare ground 65.93 dune interspace 1.38 soil surface horizon morphological rating, weighted 1.57 carbon, weighted 6.28 sand fraction in soil surface horizon, weighted 27.14 silt fraction in soil surface horizon, weighted Steptoe Watershed soil bulk density in surface 0-4 in., 1.29 dune interspace 22.41 bare ground 65.93 dune interspace 53.99 litter soil surface horizon morphological 1.38 rating, weighted 2.06 carbon, weighted sand fraction in soil surface horizon, 51.20 weighted silt fraction in soil surface horizon, 32.86 weighted 4.58 depth of surface horizon, weighted Pine and Mathews Canyon Watersheds 1.57 soIl bulk density in surface 0-4 in., dune interspace 10.30 rock 0-2 in.) 37.01 bare ground dune interspace 60.57 43.28 litter plant cover 31.78 roughness factor 0.42 soil surface horizon morphological 2.09 rating, weigh ted 0.92 carbon, weighted 3S.33 sand fraction in soil surface horizon, weighted silt fraction in soil surface horizon, 33.95 coppice dune silt fraction in soil surface horizon, 38.39 weighted Watersheds Combined 60.41 dune interspace 2.56 soil surface horizon morphological rating, weighted US carbon, dune interspace 21.74 clay fraction in soil surface horizon, weighted Standard DeviatIOn 0.08 Unit of Measure Ib/ft' 19.45 33.02 22.06 1.25 percent percent percent 1-5 15.81 percent 9.01 percent 1.49 in. 0.15 Ib/ft' 13.67 26.17 0.65 percent percent 1-5 0.51 3.94 percent percent 8.12 percent 0.24 Ib/ft' 13.67 26.17 24.25 0.65 percent percent percent 1-5 0.64 8.61 percent percent 8.59 percent 1.03 in. 0.39 Ib/ft' 4.63 27.20 38.11 38.35 24.96 0.27 1.16 percent percent percent percent percent in.' 1-5 0.44 S.36 percent percent 4.61 percent 5.52 percent 34.91 2.64 percent 1-5 2.27 8.67 percent percent All Watersheds YI mean infiltration rate after 30 min, soil surface initially dry in./h 933 BLACKBURN: INFILTRATION AND SEDIMENT PRODUCTION TABLE 4. Number Y2 Y3 Y4 (continued) Variable mean infiltration rate after 30 min, soil surface initially at field capacity log total sediment production, soil surface initially dry log total sediment production, soil surface initially at field capacity Coils Creek Watershed. Bare ground, dune interspace, weighted soil surface horizon morphological rating, and weighted percent silt of the surface horizon are the important variables in (2) and (3) (Table 3). Dune interspace, however, is the most important variable. Weighted soil surface horizon morphological rating and percent silt of the surface horizon are intermediate in importance, and bare ground is the least important variable. Steptoe Watershed. In (5) and (6) (Table 3), bulk density, bare ground, dune interspace, weighted soil surface horizon morphological rating, and weighted depth of soil surface horizon are the important variables. Dune interspace and weighted soil surface horizon morphological rating are the most important variables in determining infiltration. Pine and Mathews Canyon Watersheds. Pavement, bare ground, dune interspace, litter, plant cover, roughness factor, and silt of the coppice dune differ in importance with the various treatments. In (7) (Table 3), silt of the surface horizon, coppice dune, and litter are the most important. In (9), dune interspace and pavement become the most important in determining infiltration. Combined analysis. Data from the five watersheds were combined in order to develop some uniform equations that might be more generally applied (Table 3). In both equations, weighted soil surface horizon morphological rating is the most important variable in determining infiltration; dune interspace, percent organic carbon, and dune interspace are intermediate in importance, weighted clay fraction in the soil surface horizon being least important. Many plant communities in the Great Basin are characterized by a shrub overstory and very sparse vegetation between shrubs. According to Duley and Kelly [1939], infiltration rates of a site should be the same or very similar for different rainfall intensities as long as the terminal infiltration is exceeded. To demonstrate this, simulated rainfall was applied to dune interspace areas of the winter fat community at a rate of 3, Ii, 0.97, 0.57, 0.44, and 0.22 in./h. It was not until rainfall intensities of 0.57, 0.44, and 0.22 for air-dry soil and 0.22 for soil initially at field capacity occurred that infiltration rates dropped below a fairly constant terminal rate. This was not the case when coppice dunes and dune interspace areas were included in the study plots. Infiltration rate for the 3-in./h simulated rainfall was almost twice that of the li-in./h rainfall. On variable plots coppice dunes displayed an exceedingly high infiltration rate, i.e., sometimes 3-4 times greater than the dune interspace areas and near or exceeding the application rate. Dune interspace areas exhibit a relative low infiltration rate (Figure 1). This relationship held true over a wide variety of soils, except for a few Utah juniper coppice dunes where hydrophobic soils were encountered. Infiltration rates for shrub or grass (big sagebrush/crested Unit of Measure Standard Deviation Mean All Watersheds (continued) in./h Ib/acre Ib/acre wheatgrass only) coppice dunes are significantly higher than those for corresponding dune interspace areas. Highest infiltration rates are in big sagebrush/crested wheatgrass and low sagebrush Sandberg bluegrass/squirreltail coppice dunes, and lowest rates occur in big sagebrush and low sagebrush/Sandberg bluegrass/squirreltail dune interspace areas. The only grass coppice dune sampled (big sagebrush/crested wheatgrass) had a significantly higher infiltration rate than its untreated big sagebrush coppice dune counterpart (big sagebrush/rubber rabbitbrush) (Table 5). Rate at which water will enter a soil landscape is governed mainly by the extent and soil surface horizon morphology of the dune interspace areas (Figure 2). The big sagebrush community, Duckwater Watershed, is used to demonstrate this soil difference (Table 6). Dune interspace areas have a shallower surface horizon, a lower percent carbon, a higher pH, a higher bulk density, and a higher percent silt than the coppice dunes. Soil structure in dune interspace areas is moderate to fine platy as compared to weak fine granular in the coppice dunes. Dune interspace areas also have larger and many more vesicular pores in their surface horizon than the coppice dunes. These conditions account for more than 3 times higher infiltration rates on coppice dunes. This relationship is further attested by personal observation after a high-intensity thunderstorm at Duckwater Watershed, July 22, 1970. This storm had an average intensity for the basin of 1.2 in./h, mostly as hail and rain. However, intensities within the watershed exceeded 4 in./h for short periods (John Trimmer, personal communication from records of Bureau of Land Management, Nevada State Office, Reno, Nevada, \970). Runoff was large, causing major downstream damage to the Duckwater Indian Reservation and to the watershed. After the storm, moisture had penetrated more than 8 in. in the coppice dunes and only 2 in. in dune interspace soil. Vesicular horizon. Infiltration rates are negatively related to vesicular horizons (Figure 3), and the strength of this .... -.... ... .5 ~ .... • .... - ....... • .. - - .. - - · - - - - - - . - - - - - ... - - ... _1.7 ,.: ,.,.." .... 2 ]co ~ ~ ..._,.,:.... .., ................., .............., ' , 1 ..! - 5011 lur'.ca InltloUy .t "oW c.~.dty ...... - .... 5011 turfoco Inltlolty II,,. O~----~----~~----~----~~----~----~ o 10 15 ::10 :u 30 Time (min) FIg. I. Infiltration curves for the big sagebrush community, Duckwater Watershed. 934 BLACKBURN: INFILTRATION AND SEDIMENT PRODUCTION TABLE 5. Mean Infiltration Rate (Inches per Hour) and Sediment Production (Pounds per Acre) for Coppice Dune and Dune Interspace Areas of Six Plant Communities in Three Watersheds Sediment Production Infiltration Rate Dry Plant Comm unity and Watershed· Coppice Big sagebrush/crested wheatgrass (M) Big sagebrush/Sandberg bluegrass/diffused phlox (C) Big sagebrush (D) Big sagebrush/bluebunch wheatgrass/arrowleaf balsam root (C) Low sagebrush/Sandberg bluegrass/squirreltail (C) Big sagebrush/rubber rabbitbrush (M) Field Capacity Interspace Coppice Interspace Dry Coppice Field Capacity Interspace Coppice Interspace 2.890 1.79bc 2.720 1.16cd 3000 5200 1200b 4800 2.870 1.56c 2.710b 0.84cd 10 4600 60b 6200 2.870 2.820 1.03 1.99b 2.600b 2.780 0.82cd 1.26bc IOOob 20 200ab 3400 2200b 60b 3200b 6400 2.800 1.54c 2.840 0.65d 40b 3000 60b 6400 2.54 1.78bc 2.29b 0.71d 3600 3200 320ab 300ab Means followed by the same letter are not significantly different (0.05) as determined by Duncan's multiple-range test. All comparisons are made within the two columns under each antecedent moisture condition (dry or field capacity). * The letters C, D, and M represent, respectively, Coils Creek, Duckwater, and Mathews Canyon. relationship is dependent on vesicular horizon morphology. Sudicial vesicular horizons develop in arid and semiarid areas of sparse vegetation cover. Yolk and Geyger [1970] observed in many warm-arid areas of the earth that there are patches free of vegetation distributed in a mosaiclike pattern, although the precipitation is sufficient for plant growth. These 'scalded areas' are not necessarily caused by overgrazing or soil salinity. They made extensive observations in southern Spain, Morocco, and southwest Africa and found near the soil surface a structure formation that they designated as 'foam structure' (vesicular porosity). Miller [1971] studied vesicular pore formation under furrow irrigation in Washington and speculated that a platy structure develops first. Then, with continued wetting and drying, the pores between platelets became spherical and the spheres Fig. 2. became larger. He also noted that all of the soil exhibited considerable swelling and shrinking with wetting and drying. The bulk density of the surface decreased with a number of wetting cycles. Miller concluded that soils involved in vesicular pore formation are very unstable when they are nearly saturated and the air pressure is sufficient to form the cavity between platelets into a sphere, thus achieving the smallest surface area per unit volume. More air is entrapped in the soil with each wetting and drying cycle. Then in wet, fluid soil, small vesicles merge into larger ones because the surface area per unit volume decreases as they enlarge. Vesicles remain air-filled during water application. This horizon develops in surface 2 or 3 in. of the soil. Soils involved in vesicular development are classified as Aridisols, Torrifluvents or Torriorthents. They have low organic matter and high percent silt-sized particles. Soil profile showing coppice dune and dune interspace soil surface horizon. BLACKBURN: INFILTRATION AND SEDIMENT PRODUCTION TABLE 6. Soil Surface Horizon Parameters and Infiltration Rates for Coppice Dune and Dune Interspace Areas of the Big Sagebrush Community, Duckwater Watershed Parameter Dune Interspace 2 Depth, in. Carbon, % 0.5 pH 7.7 Bulk density, Ib/ftS Silt, % Structure Pores 0.12 26 Infiltration rate, in./h medium fine platy . many very fine and fine vesicular 0.74 Coppice Dune 5 1.0 7.4 0.10 18 weak fine granular few fine vesicular and interstitial 2.54 These vesicular horizons are poorly aggregated and very unstable when they are nearly saturated. Organic matter. Organic matter in the soil is the major cause of aggregation. It not only binds soil particles into aggregates but lightens and expands the soil, thus increasing the porosity and decreasing vesicular horizon formation. Infiltration is positively related to organic matter except for a few juniper coppice dunes where hydrophobic soils occur. Bulk density. Coppice dunes consistently have a lower bulk density than dune interspace soils. However, dune interspace soils with vesicular porosity exhibit a lower bulk density than dune interspace soils without vesicular porosity. Infiltration rates of soils without vesicular horizons decrease as bulk density increases. On soils with a vesicular horizon, infiltration rates decrease as bulk density decreases. Texture. Silt and clay-sized particles are negatively correlated with infiltration rates. Sand-sized particles are positively related to infiltration. Generally, the coarser the surface texture, the higher the infiltration rate. Of the three particle sizes, percent silt of the soil surface horizon has the strongest influence on infiltration because of vesicular horizon formation in: soils high in silt. Surface horizon depth. Thick surface horizons are found in highly productive soils and in coppice dunes; i.e., coppice Fig. 3. 935 dunes are 1-4 in. thicker than interspace soils. Infiltration rates increase as the surface horizon increases in thickness. Moisture. Infiltration rate was lower on soils that were initially at field capacity than on those that were initially dry. The higher the initial moisture content, the more micropores that are filled and the lower the infiltration rate. Infiltration rates are usually negatively related to antecedent moisture content. Plant and litter cover. Plant and litter cover are both positively correlated with infiltration; however, their influence is not as strong as has been shown in other studies [Dortignac and Love, 1966; Rauzi et al., 1968; M eeuwig, 1970]. In this study, vesicular horizon and dune interspace areas were more important in explaining infiltration than plant and litter cover probably because of the sparse cover of vegetation in the interspace areas and the kind of soils. Rock. Study sites were characterized by a very low cover of large rocks (>2 in. in largest diameter); thus rock showed a poor correlation with infiltration. However, as small rocks (1-2 in. in largest diameter) increased on the sites, infiltration rates decreased. This latter relationship is explained by the large percent of small rocks associated on the surface of soils with vesicular horizons. Bare ground. Infiltration rates were strongly negatively correlated with bare ground. This is consistent with findings of Duley and Domingo [1949], Branson and Owen [1970], and many others. Slope. Percent slope shows a positive correlation with infiltration rates and indicates that as slope increases, so does infiltration rate. Actually in this study, slope had very little influence on infiltration rates. Most study sites were characterized by very gentle slopes except in the mountains where sites on steeper slopes have high infiltration rates. The correlation, then, is due to different soils and not to slope. Surface roughness. Surface roughness shows a weak positive correlation with infiltration and indicates that an irregular surface may increase infiltration rates slightly. This poor correlation is probably due to the fairly level relief of the study area. Vesicular horizon showing platy structure and vesicular porosity; scale division is 0.04 in. 936 BLACKBURN: I NFILTRATION Sediment Production The following discussion is based on multiple regression equations (Table 3) where the dependent variables are common logarithms of sediment production at the end of 30 min. The mean and standard deviation of the independent variables in the regression equations are given in Table 4. Equations developed from data where the soils were initially at field capacity usually had the lowest standard error of estimate and the highest R2 values. Runoff, as the only independent variable, is highly correlated with sediment production. However, when runoff was included in the multiple-regression equations, it failed to improve the standard error of estimates or R2 values. The reason for this relationship is that the same factors often influence infiltration and sediment production. Thus runoff does not appear in regression equations. Duckwater Watershed. Bulk density, bare ground, soil surface horizon morphological rating, weighted sand, and weighted silt fraction of the surface horizon are the important variables explaining the variation in sediment production in (II) and (12) (Table 3). Soil surface horizon morphological rating is the most important variable explaining sediment production, and weighted silt fraction the least important. Coils Creek Watershed. Bulk density, bare ground, dune interspace, weighted organic carbon, weighted sand, and weighted silt fraction of the soil surface horizon are the important variables (Table 3). Dune interspace is the most important variable in determining sediment production, and weighted sand fraction the least important. Steptoe Watershed. Bulk density, dune interspace, litter, weighted organic carbon, weighted sand, and weighted silt fraction of the surface horizon are the most important variables in (15) and (16) (Table 3). Weighted organic carbon and weighted silt fraction in surface horizon are the most important variables explaining the variation in sediment production, and dune interspace and litter the least important. Pine and Mathews Canyon Watersheds. Bulk density, bare ground, dune interspace, soil surface horizon morphological rating, weighted sand, and weighted silt fraction of the surface horizon are the important variables for these watersheds. Bulk density and soil surface horizon morphological rating are the most important variables explaining sediment production, and dune interspace and weighted silt fraction of the surface horizon the least important. Combined analysis. Because of the high standard error of estimates and low R2 values for the combined data analyses the equations are not presented. Sediment is usually considerably higher from dune interspace areas than from coppice dunes. This relationship held true except for the big sagebrush/rubber rabbitbrush community. Sediment produced from dune interspace areas of the three communities sampled at. Coils Creek Watershed is significantly higher than it is from corresponding coppice dunes (Table 5). Variations in sediment production from coppice dunes and dune interspace areas range from none to as much as 46 times more sediment from big sagebrush/Sandberg bluegrass/diffused phlox community dune interspace areas than from their corresponding coppice dunes. As dune interspace areas increase, sediment production increases. Similar relationships also are observed for percent bare ground and percen t sil t. Significantly more sediment is produced from soils initially at field capacity than from those initially dry because of the in- AND SEDIMENT PRODUCTION stability of the surface horizon when it is saturated. The soil surface reaches saturation quicker on the wet test; thus there is a longer time to erode the dispersed soil particl.es. In this way, initial moisture content shows a positive correfation with sediment. This relationship is most marked for soils with a vesicular surface horizon. As organic matter, sand-sized particles, coppice dunes, and litter increase on the study sites, sediment production decreases. Plant cover was poorly correlated with sediment production and does not appear in regression equations. Roughness factor is slightly negatively correlated with sediment production. As bulk density ofthe surface 4 in. increases, sediment production usually increases except on sites with vesicular horizons where bulk density is negatively related to sediment production. The large variation in sediment production can be explained by the gentle slopes which allow detached and suspended soil particles to settle out in the small depressions before they reach the collection trough and the location of coppice dunes and dune interspace areas. Coppice dunes have a lower rate of sediment production than dune interspace areas. Thus the amount of coppice dune in the plot, if not constant, would influence sediment production. Likewise, suspended sediment was seen to settle out along the coppice dune edge. This effect was magnified the closer the coppice dune was to the collection trough. Acknowledgment. This study was a cooperative project between the Burea.u of Land Management and Nevada Agricultural Experiment Station, Journal Series 268. The financial support of the bureau is acknowledged. REFERENCES Black, C. A. (Ed.), Methods of Soil Analysis, Amer. Soc. Agron. Ser. 9, 1572 pp., American Society of Agronomy, Madison, Wis., 1965. Blackburn, W. H., P. T. Tueller, and R. E. Eckert, Jr., Vegetation and soils of the Duckwater Watershed, Rep. R-40, 80 pp., Coli. of Agr., Univ. of Nev., Reno, 1968. Blackburn, W. H., R. E. Eckert, Jr., and P. T. Tueller, Vegetation and soils of the Coils Creek Watershed, Rep. R-48, 80 pp., Coli. of Agr., Univ. of Nev., Reno 1969a. Blackburn, W. H., P. T. Tueller, and R. E. Eckert, Jr., Vegetation and soils of the Pine and Mathews Canyon Watersheds, Rep. 46, 109 pp., Coli. of Agr., Univ. of Nev., Reno, 1969b. Blackburn, W. H., R. O. Meeuwig, and C. M. Skau, A mobile infiltrometer for use on rangelands, J. Range Manage., 27, 322-323, 1974. Bouyoucos, G. J., Hydrometer method improved for making particle size analysis of soil, Agron. J., 54,464-465, 1962. Branson, F. A., and J. B. Owen, Plant cover, runoff, and sedimet yield relationships on Mancos Shale in western Colorado, Water Resour. Res., 6,783-790, 1970. Branson, F. A., G. F. Gifford, and .J. R. Owen, Rangeland hydrology, Range Sci. Ser. 1, 84 pp., Soc. for Range Manage., Denver, Colo., 1972. Chapline, W. R., Erosion on rangelands, A mer. Soc. Agr. J., 21, 423-429, 1929. Dortignac, E. J., and L. D. Love, Infiltration studies on ponderosa pine ranges of Colorado, U.S. Forest Servo Rocky Mt. Forest Range Exp. Sta. Pap. 59, 1-34, 1961. Duley, F. L., and C. E. Domingo, Effect of grass on intake of water, Neb. Agr. Exp. Sta. Res. Bull., 159, 1-15, 1949. Duley, F. L., and L. L. Kelley, Effect of soil type, slope and surface conditions on intake of water, Neb. Agr. Exp. Sta. Res. Bull., 112, 1-16,1939. Forsling, C. L., Erosion on uncultivated lands in the intermountain region, Sci. Monogr., 34,311-321, 1932. Giffo};d, G. F., Rangeland watershed management-A review, Nev. Agr. Exp. Sta. Pap. R-52, I-50, 1968. BLACKBURN' I NFILTRATION Harper, V.L., Watershed management-Forest and range aspects in the United States, Unasylva, 3, 105-114, 1953. Heinze, D. H., R. E. Eckert, Jr., and P. T. Tueller, The vegetation and soils of the Steptoe Watershed, 40 pp., Coll. of Agr., Univ. of Nev., Reno, 1966. \ Kincaid, D. R., and G. Williams, Rainfall effects on soil surface characteristics following range improvement treatments, J. Range Manage., 19,346-351, 1966. Levy, E. B., and E. A. Madden, The point method of pasture analysis, N. Z. J. Agr., 46, 267-279, 1933. Meeuwig, R. 0., Infiltration and soil erosion as influenced by vegetation and soil in northern Utah, J. Range Manage., 23, 185-188, 1970. Miller, D. E., Formation of vesicular structure in soil, Soil Sci. Soc. Amer. Proc., 35,635-637, 1971. Rauzi, F., C. L. Fly, and E. J. Dyksterhuis, Water intake on midcontinental rangelands as influenced by soil and plant cover, U.S. Dep. Agr. Tech. Bull., 1390, 1-58, 1968. U.S. Department of Agriculture, Influence of vegetation and AND SEDIMENT PRODUCTION 937 watershed treatments on runoff, silting and stream flow, U.S. Dep. Agr. Misc. Publ. 397, 1-80, 1940. U.S. Department of Agriculture, Soil Survey Manual, Handb. 18, 503 pp., Washington, D. C., 1951. U.S. Department of Agriculture, A Comprehensive System, The Seventh Approximation, 265 pp., Washington, D. C., 1970. U.S. Department of Agriculture, Hydrology, Part 1, Watershed Planning, Washington, D. C., 1964. U.S. Department of Agriculture, Soil Taxonomy of the National Cooperatwe Satl Survey, 265 pp., Washington, D. C., 1960. Volk, O. H., and E. Geyger, 'Schaumboden' als Ursache der Vegetationslosigkeit in ariden Gebieten ('Foam' soils as the cause of absence of vegetation in arid lands), Z. Geomorphol., 14, 79-95, 1970. (Received August 10, 1973; revised June 24, 1975; accepted July 23, 1975,)