118 7(3), June 1985 Managing Resources: The Universal Soil Loss Equation K. G. Renard and G. R. Foster

Someof the earliest soilerosion measurementsin the U.S. subsequently was the basis for mathemattUal rosuonpwuic- were made by A.W. Sampson and associates in 1912 on tion relationships likethe USLE. Cook (1936),in theearliest overgrazed rangelands in central Utah. These studies and effort to mathematically describe soil in the U.S., research by Chapline (1929) illustrated how overgrazing identified three major factors affecting erosion: (1) suscepti- allowed erosion to reduce soil fertility and water-holding bility of the soil to erosion, (2) the potential of rainfall and capacity. Unfortunately, erosion measurement/researchon runoff for causing erosion (erosivity), including the influ- rangeland languished since these early efforts until the ence of slopesteepness and length, and (3) the protection 1970's. Concern forthe ecological health of rangelandgrew afforded by vegetal cover. He described in detailhow other with thegeneral concern forthe environment that developed subfactors affect each of these major factors. His concepts during the late 60's and 70's, and excessive erosion was have been embodied in the string of erosion prediction again recognized as being detrimental to rangelands.As a methods that led to the USLE. consequence, managementplans for rangelandsfrequently By 1940, sufficient data had been collected for Zingg contained analyses on how managementalternatives would (1940)to developthe first erosion equation which calculated affecterosion. Since researchhas provided little information erosion as a function of slope length and degree of slope on erosion associated with rangeland, technology from (LS). In the following year, Smith (1941) added a crop factor othergeographic areaswas adapted to estimate erosion on (C) and supporting practice factor (P) to theequation, which rangeland. In particular, the Universal Soil Loss Equation was already beginning to resemblethe USLE. Thisequation, (USLE),which has been used successfullyon cropland since in contrast tothe USLE, was limited to a very specific region theearly 60'swas adaptedto estimate erosion on rangeland. and soiland specific crops in the vicinityof Missouri. Subse- Theobjective ofthis paper is to (1) familiarize rangescient- quent researchin the 1940's concerned refinementof predic- ists with the research which led to the USLE, (2) familiarize tion equation parametersbased on data from specific loca- range scientists/managers with the factors considered by tions;presented new data for crops, rotations, encoun- the USLE, and (3) discuss someof the problemswith extrap- tered in specific regions; and how the erosion hazard of olatingthe USLE researchfrom cropland torangeland areas. rainfall varies through the year at different locations in the U.S. History of Erosion Prediction By 1949, the concept of using erosion equations to help Early erosion research, started in 1917 at the Missouri design agronomic practicesto meet specific erosion hazards Agricultural Experiment Station, is the predecessorto mod- was recognized (Musgrave 1949). Concurrent with these ern (current) erosion research (Meyer 1984). Miller's 1/80- developmentswas Ellison's (1947) classic research on fun- acre plots (90.75 ft long by 6.0 ft wide) at Missouri greatly damental erosion processes. His research provided the influenced research initiated at the 10 experiment stations foundations forthe new process-orientederosion prediction established by Congress in 1929, during the crusades of methods that are just now beginning to be applied by user Hugh H. Bennett, the"father" of the soil conservation move- agencies. Had computers been available in the 40's, current ment.These stations were located at Guthrie, Okia.; Temple, erosion prediction methods might look a lot more like Elli- Texas; Tyler, Texas; Hays, Kans.; Bethan, Mo; Statesville, son's theorythan like the empirical form of the USLE. The N.C.; Pullman, Wash.; Clarinda, Ia; LaCrosse, Wis.; and USLE,and its predecessors,were very much structured tobe Zanesville,Ohio, and provided anextensive data base for the "user"friendly, becauseby theearly 50's, erosion equations decade or more that these stations operated. were accepted by the USDA-Soil Conservation Service as a The pre-World War II years were important for erosion powerful tool for tailoring practices to the research because the importance of soil conservation was needsof specific fields and farms. Unfortunately, duringthis recognized; key research procedures were established that period, a comparable erosion research program on range- are still used; fundamental research was encouraged and lands in thewestern U.S. was not underway,and thus, recent produced theorythat is just beginning tobe used in the more effortsto develop erosion methods for rangelands have not scientifically basederosion prediction methods;researchers had an extensivedata base. were enthusiastic about their endeavorsand many outstand- of the ing scientists were involved; and adequate funds were avail- Development liSLE able for staffing and facilities. The common experimental Priorto the development of the USLE, erosion equations design among thestations produced a wealth of data which had been developed from site specific data on soil losses, and were therefore limited to specific regions and soils. The Authors are research hydraulic engineers, Southwest Rarsgeland Watershed need for a single, widely applicable erosion equation was Research Center, 2000 East Allen road, Tucson, Ariz. 857)9, and USDA-National Laboratory,Purdue University, W. Lafayette, md.47907, respectively. recognized in the early 50's, but development of such an Thisarticle is a contributionof USDA, ARS Southwest Rangeland Watershed would require the collection and combination of Research Service,2000 E. Allen Rd., Tucson, Ariz. 85719. equation Rangelands7(3), June 1985 119 many data bases into a single data base. Thus, the National ticefactor expressed as a ratio of the soil loss with practices Runoffand Soil Loss Data Centerwas established by USDA- such as contouring, strip cropping, or terracing to that with Agricultural ResearchService (ARS)at Purdue University in farming up and down the slope. 1954, under the direction of W.H. Wischmeier, for the pur- Theterm 'universal' in the USLEwas given to theequation pose ofdeveloping an erosion prediction equation basedon to assist users who were accustomed to previous equations all thedata available throughoutthe U.S. Between 1956 and that applied to very specific regions in contrast to the USLE, 1970, additional plot-years and watershed-yearsof data from which applied, initially in 1965, to all of the U.S. east of the continuingstudies, and from about 20 new locations, were Rocky Mountains, and to the 1978revision, which applies to added to the data bank. Over 10,000 plot-years of data were all of the U.S. analyzed to develop the original USLE (Wischmeier and Wischmeier (1972) explained: Smith 1965). The name'universal' soil-loss equation originated as a means Because the costs of collecting data from plots under of distinguishing this prediction from the highly regionalized natural rainfall was ARS a - modelsthat precededit. None of its factors utilizes a reference rapidlyincreasing, developed thathas direct orientation. Inthe senseof the fall known as the rainulator and McCune point geographic simulator, (Meyer intended functions of the equation's six factors, the model 1958), to conduct erosion research on plots with artificial should have universal validity. However, its application is rainfall. By the 1970's, many of the natural runoffplot studies limited to states and countries where information is available were discontinued and replaced with studies using simu- for local evaluationsof the equation's individual factors. lated rainfall. WhenWischmeier and Smith revisedthe (1978) The USLE is sometimes referred to as a "midwest- USLE, they used rainfall simulator data todescribe soil erod- being ern" but the is much more based. ibility and toprovide values for the effectivenessof conserva- equation, equation broadly tion and Data used to developthe USLEcame from 48 locations listed tillage construction practices for controlling soil Handbook 537 and Smith erosion. in Agriculture (Wischmeier 1978). Out ofthe 48 locations, more than half, 27, are outside of the The USLE (Wischmeierand Smith, 1965, 1978) is: Midwest by the most liberal definition of the Midwest. If A=RXKXLXSXCXPwhere: locations like Zanesville and Coshocton, Ohio (representa- A is the estimated averageannual erosion rate per unit of tiveof eastern hill country), and Hastings, Neb., and Hayes, area computed by multiplying values for the other six fac- Kans. (representativeof the Great Plains) are taken outof the tors. It is an estimate of the average annual sheet and nIl Midwest count, the number on non-Midwest locations is 31 erosion from rainstorms on upland areas, and it does not out of 48 locations. If these locations are plotted on a U.S. includeerosion fromgullies or streambanks,snowmelt ero- map, they are reasonably well distributed across the U.S. sion, or wind erosion. It does includeeroded sediment that east ofthe Rocky Mountains. Datafrom the 48 locations were may subsequently be deposited on the toe of slopes and at principally usedto determine theeffects ofsoil, topography, otherplaces before runoffreaches streams or reservoirs. cover, and managementon erosion. Data used to calculate R is the rainfall and runoff factor for a specific location, the erosivity factor for the USLE came from 181 locations, usually expressedas averageannual erosion index units. with several, like Albuquerque, N. Mex.; Red Bluff, Calif.; K is the soil erodibilityfactor for a specific soil horizon, Billings, Mont.; and Casper, Wyo.; being from the West expressedas soil loss per unit of area per unit of R for a unit (Wischmeier and Smith, 1978). Therefore, a more correct plot (a unit plot is 72.6 feet long, with a uniform 9% slope representation ofthe USLE is thatit was primarily developed maintained in continuous fallowwith tillage when necessary from cropland data east of the Rocky Mountains. to break surface crusts and to control weeds).These dimen- In the early 70's, the USLE was beginning to be applied to sions were selected becausethe 1/100 ac erosion research noncropland applications like construction sites and undis- plots used in early erosion work in the U.S. were 72.6 feet turbed land, including rangelands. Since an extensive data long and had slopes near 9%. Continuous fallow was base was not available for these applications, Wischmeier selected as a base, because no cropping system is common (1975) developed the subfactor method to estimate values to all agricultural areas, and soil loss from any other plot for the C factor.The subfactor method usesrelationships for condition would be influenced by residual and currentcrop canopy, ground cover, and "within" soil effects to estimate a and management effects that vary from one location to composite value for C, the USLEcover-management factor. another. This development allowed the use of data collected from L is the dimensionless slope-length factor (not the actual more basic studies to be used in the USLE. Recognizingthe slope length) expressedas the ratio of soil loss froma given need for data, scientists began erosion experiments on slope length to that from a 72.6-ft length under the same rangeland todevelop USLEparameter values, and to evalu- conditions. atethe performanceof the USLEon rangelands.Table 1 lists S is the dimensionless slope-steepnes factor (not the some of this research, including some referencesshowing actual slope steepness) expressed as the ratio of soil loss problems with the use of the USLEon rangeland. from a given slope steepnessto that from a 9% slope under the same conditions. ParameterValues C isthe dimensionless cover and management,or cropping- Determinationof values ofthe individual USLEparameters management,factor expressed as aratio of soil loss from the for use on western rangelandspose some unique problems condition of interest to that from tilled continuous fallow. and conditions not encountered on cultivated cropland. P is the dimensionless supporting erosion-control prac- These conditions prevent the direct extrapolation of some 120 Rangelands 7(3), June 1985

Table 1. Examplesof research evaluating USLE or USLE parameterperformance on rangelands.

Area where Authors & Dates work was done Comments Dissmeyer, 1982 N. Mex. Usedsubfactor approach in evaluatingC on rangeland. Foster, et al. 1981 General Discussedapplicability ofUSLE to rangelands. Hart, 1982 Utah Measurederosion on sagebrushplots with a rainfall simulator. Hart, 1984 Utah Fair agreementof USLEwith simulated rainfall data. Slopefac- tor needs adjustment. Johnson et al. 1980 Ida. Usedcanopy and ground coverto compute potential erosion for sagebrushcontrol. Johnson et al. 1985 Ida., Nev. Used rainfall simulator and found interpretation of C on ungrazedareas neededrefinement. McCool, 1982 Wash. Analysis of LS factor. Osborn, Simanton, Renard, 1976 Ariz., N. Mex. Showed importance of stonesurface cover. Renard, Simanton, Osborn, 1974 Ariz. Used small watersheds;significant channel erosion. Renard, Simanton, 1975 Ariz., N. Mex. Explored estimation oferosion factor. Renard,1980 Ariz. Compared numeroussediment yield formulae. Renard,Stone, 1982 Ariz. Correlation of USLE estimateswith stockpond yields. Simanton, Osborn, Renard, 1977 Ariz. Showed effect of root plowing and reseeding on erosion control. Simanton, Osborn, Renard, 1980 Ariz. Applied to small watershedson storm basis. Simanton, Renard (a & b), 1982 Ariz., N. Mex. Evaluatederosivity of air-massthunderstorms. Simantonet al. 1984 Ariz. Measurederosion reduction causedby stone surfacecover. Smith etal. 1984 Texas,OkIa. Sedimentyield estimateswith modified USLE, watersheds <122 ha and on watershedswith mixed land uses. Tracy et al. 1984 Ariz. Measureddrop-size distribution of air-massthunderstorms for use in evaluating erosivity. Trieste,Gifford, 1980 Utah Used small plotswith rainfall simulator. SuggestedUSLE did not apply well to rangelands. Trott,Singer, 1983 Calif. USLE soil erodibilityfactor should consider soil mineralogy. Verma,Thames, Mills, 1977 Ariz. Measurederosion from disturbed and natural plots with artificialor simulated rainfall. Williams, 1982 Texas,OkIa., Estimatedsediment yield frommixed cover watershedswith Iowa, N. Mex. modified USLE.

valuesfrom cropland, and require caution in theextension of has a high correlation with El. Thus, although runoff might other values. intuitively have been a parameter to be included directly in The rainfall/runoff erosivity factor (R) is computed as the the USLE, the use of El servesas a surrogate forrunoff and productof the kineticenergy ofan individual storm times the precipitation-induced erosion. maximum 30-minute intensity forthe storm (El). The annual The erosivity factor R (remember, R summation of El for value then is the summation of all such storms in the course storms in a year) needs adjustment to account for erosion of a year. The equation used to compute kinetic energy for from runoff associated with thawing soils and snowmelt. each intensity period of the storm (time-intensity record) This adjustment was developedas 1.5 times the winter pre- was developed from data collected at the Bureau of Stand- cipitation (measuredas inches of water) whichis added toEl ards inthe late 1930's(Laws and Parsons1943). Other inves- erosivity for nonwinter storms. This adjustment is very tigators have developedspecific equations for algorithms in general, and data to support it are scarce. Since this type of other parts of the country, but the Bureau of Standards erosion can be appreciable on many rangelands,additional equation is generally used throughoutthe country (Tracy et research is needed on this problem. al. 1984). Thecover-management factor (C) ofthe USLErepresents The individual storm El is nearly proportional to the total the ratio ofsoil loss forland under specified conditions tothe precipitation times the maximum 30-minute intensity, rain- corresponding loss from clean-tilled, continuous fallow. fallparameters observedto be most important forestimating Obviously, the standard fallowplot uèedfor croplandsoils is runoff. Studies (e.g., at theSouthwest RangelandWatershed inappropriate for rangelands.An untilled bare plot, cleared Research Center) have shown that individual storm runoff of all surface vegetation and stonesand maintained through Rangelands 7(3), June 1985 121 chemical control of range vegetation, seems more approp- slope of a regression linethrough the origin for data on soil riate than thetilledfallow for the USLEunit plot on rangeland. loss (A) and El afteradjusting the ratios for C, LS, and P to The C-factor, like many otherUSLE terms, representsthe those of unit conditions. Thus, when the K value was deter- integratedeffects ofseveral conditions that affect erosion. In mined with natural storm data, it represented a range of the evaluation of the C-factor, one must have information storm sizes and antecedent soil conditions. Later, similar regarding the plant canopy (height and density) and basal experiments were performed using rainfall simulators, and area. The term thus reflects the interception of raindrops in produced a soil erodibility nomograph (Wischmeier and the canopy and, in turn,how drops reformed on the canopy Smith 1978) that gives K as a functionof a soil'spercent silt affect splash erosion. The term also reflects the binding and very fine sand, percent sand (0.10 to 2.0 mm), percent effect of plant roots and how the soil changes as it lies idle. organic matter,an index ofsoil structure, and a relativeindex Important, but ill defined, is how the grazing animal changes of infiltration. Values estimated with this nomograph for the value of the C-factor. Not only does the grazing animal bare, untilled fallow plots at the Southwest Rangeland remove some of the plant canopy which otherwise would WatershedResearch Center were comparable with experi- become cover (litter) in direct contact with the ground, but mental data. the hooves may roughen the surface, or even compact the soil, and thereby alterinfiltration, and thus, runoff (reflected DiscussIon in the K-term). Researchto betterdefine these cause-effect The USLE is a useful tool for estimating erosion, for relationshipsis needed tofill this major void in the technology. assessmentof the impact oferosion on productivity, and for Rock fragments,litter, and leavesin direct contact with the use as a guide in selecting erosion control practices on a soil surface are very effective ground cover affecting infiltra- variety of land uses, including rangelands. Its utility as a tion and erosion. Erosion rates, from simulator plots with planning tool has beenproven by over twodecades of useby rock fragment cover, were found to decreaseexponentially the USDA-Soil Conservation Service (SCS) on cropland. with increasing percent ground cover (Simanton et al., Furthermore, the USLE was developed by researchers in 1984). This relationship is considered in the C-factor. ARS and the State Agricultural Experiment Stations, along The topographic factors L and S describe the effect of with users in action agencies like the SCS. Thus, it repre- slopelength and steepness on erosion. Since the USLE is a sents the collective input of awide variety ofresearchers and sheet and nil erosion prediction equation, slope length refers users. to overland flow from where it originates to where runoff The USLEis a package oferosion information and know- reaches a defined channel, or to where deposition begins. ledge. To a major degree, the extent that the USLE inade- Thus, the USLEdoes not consider deposition likethat on the quately describes erosion representssignificant gaps in the toe of concave slopes; nor does it describe erosion. general knowledge about erosion. Fundamentally,the USLE Although slopes are usually treated as uniform landscape is scientifically sound, although clearly, its factor values can profiles, techniques are available for treating nonuniform be improved forwestern rangelands. Research is underway profiles (Wischmeier and Smith 1978). Maximum slope to make these improvements,although, at current funding lengths are seldom longerthan 600 ft. and the USLEdoes not levels, the answerswill not likely come very rapidly. applyto slope lengths shorter than 15 to 20 ft. Selection ofa The USLE provides a methodology and consistent means slope length requires judgment, and theinterpretation ofany for estimating erosion, something that is very important to topographic map complicates selection of the slope length federal agencies like SCS and BLM, dealing with large value. Accurate selection of a slope length often requires an regions. With the USLEas a convenient package of tech nol- on-site inspection. ogy, technicians without a complete knowledge of erosion The maximumsteepness of plotsused oncropland plotsto literature can effectively estimate erosion. derive the USLE S factorwas about 25 percent, flatter than The validityof an analytical method likethe USLEmust be any rangeland slopes. Data from rangelands (Hart 1984) judged. Such methods are judged to be valid if they serve suggest that the USLE may be overestimating the slope their intended purpose. Obviously, this criterion involves effect on rangeland, and the S factor will likely be adjusted more than just the accuracy of the method's estimates. For downward in the current USDA-BLM (Bureau of Land Man- example, resources required to use this method must be agement) revisions of the USLE based on analysis of new reasonable for the user, and experience with the USLE data. Use ofplots with simulated rainfall is providing apartial shows that it can be used by the field technican for in situ data base for rangeland erosion (Simanton and Renard planning of erosion control alternatives. 1982a). The accuracy issue can be addressed by considering By definition,the support practice factor(P) in theUSLE is whether the method leads to the desired managementdeci- the ratio of soil loss with a specific support practice to the sion which, in this case, relatesto erosion control. The issue corresponding loss with up-and-down slope culture. Unfor- here is notone ofwhether or notestimating erosion is agood tunately, experimental data to quantifythis termfor practi- method to estimate rangelandcondition. Given thaterosion ces on rangeland are notavailable, and thus, valuesfor Pare is a concern on rangeland because of the need to provide selected based on judgment and experience obtained on long-term protection to the soil resource, does the USLE cropiand. Practices generally reflected by P (e.g., terraces provide useful estimates oferosion on rangelands?We con- and strip cropping) are not typical on rangelands. tend that it does, but we recognize that many professionals The soil erodibilityterm (K) of the USLE is intended to do not agree with us. reflect thesusceptibility of soil to erosion. Basically, K is the 122 Rangelands 7(3), June 1985

In the end, each user ofthe USLEis obligated to decide if Renard,K.G., and J.J. Stone. 1982. Sedimentyield from small arid the USLE is valid for his application, and to inspect the rangeland watersheds. Proc. Workshop on Estimating Erosion and Sediment Yield on Rangelands. USDA-ARS, ARM-W-26, p. results heobtains with it. The user makesthe decision—not 129-144. the because is a tool that one. of USLE, it provides input Slmanton, J.R., H.B. Osborn, and K.G. Renard. 1977. Effects of information to go along with other inputs that the user may brush to grass conversion on the hydrology and erosion of a have available,such as specific data. Usedappropriately, the semiaridsouthwestern rangeland watershed. Hydrology andWa- USLEis auseful tool in thetoolbox of analytical methodsfor ter Res. in Arizona and the SouthwestV)l:249-256. Slmanton,J.R., H.B. Osborn, and K.G. Rnard. 1980. Application of guiding rangeland management. the USLE tosouthwestern rangelands. Hydrology and WaterRes. In Arizona and the SouthwestX:213-220. References Slmanton,J.R., andK.G. Renard. 1982a.Seasonal change in infiltra- tion anderosion from USLE in southeasternArizona. ChapHne, W.R. 1929. Erosion on rangeland. J. Amer. Soc. Agron. plots Hydrol. 21:423-429. &Water Resour.In Ariz. and the Southwest.University of Arizona Cook, H.L. 1936. The nature and controllingvariables of the water 12:37-46. erosion Soil Sd. Soc. Amer. Proc. 1:60-64. Simanton,J.R., and K.G. Renard. 1 982b.The USLErainfall factorfor process. U.S. Proc. on Diumeyer,G.E. 1982. Report on application of the USLEto desert southwestern rangelands. Workshop Estimating brush and conditions. Proc. on Erosion and Sedimentyield on Rangelands.USDA-ARS, ARM-W- rangeland Workshop Estimating 50-62. Erosion and SedimentYields on Rangelands.USDA-ARS, ARM- 26. p. W-26, 214-225. Simanton,J.R., E. Rawltz,and E.D. Shirley. 1984. The effects of rock p. on erosion of semiarid soils. 65-72. In: Elllson, W.D. 1947. Soil erosion studies. Agr. Eng 28:145-1 46, 197- fragment rangeland p. 201, 245-248, 297-300, 349-351, 402-405, 442-444. Erosion and Productivity of Soils Containing Rock Fragments, Soil Sci. Soc. Amer. Pub. 13. J.R. Simanton, K.G. L.J. and H.B. Chap. 7, Special Foster, G.R., Renard, Lane, Smith, D.D. 1941. Interpretation of soil conservationdata for field Osborn. 1981. Viewpoint: Application of the USLEto rangelands use. Agr. Eng. 22:173-175. on a per-storm basis. J. Range Manage. 34:161-166. J.R. R.G. G.A. Coleman. 1984. 1982. A test of the USLE on bareand Smith, S.J., Williams, Menzel, and Hart, G.E. sagebrushplots in Predictionof sediment from SouthernPlains with Utah. Proc.Workshop on Erosion and SedimentYield yield grasslands Estimating the Modified Universal Soil Loss Equation. J. Range Manage. on Rangelands.USDA-ARS, ARM-W-26, p. 101-105. G.E. 1984. Erosionfrom simulated rainfall on mountain 37:295-297. Hart, range- K.G. 1984. Rainfall land in Utah. J. Soil & Water Conserv. 39:330-334. Tracy, F.C., Renard, and N.M. Fogel. energy characteristics for southeastern Arizona. 559-566. In: Water- Johnson, C.W., G.A. Schumaker,and J.P. SmIth. 1980. Effects of p. and control on erosion. J. Today and Tomorow. Proc. Amer. Soc. Civil Eng, Irrig. & Drain. grazing sagebrush potential Range Div. Conf, Ariz. Manage. 33:451-454. Specialty Flagstaff, Johnson, C.W., M.R. Savabi, and S.A. LoomIs. 1985. Rangeland Trieste,D.J., andG.F. Gifford. 1980. Applicationof the UniversalSoil erosion measurementsfor the USLE. Accepted for Trans. ASAE. LossEquation to rangelands on a per-storm basis. J. Range Man- Laws,J.D., and0.A. Parsons.1943. The relation ofraindrop sizeto age. 33:66-70. intensity. Trans. Amer. Geophys. Union. 24:452-459. Trott,K.G., and N.J. Singer. 1983. Relativeerodibility of20 Califor- McCooi, D.K. 1982. Effects of slope length and steepnesson soil nia range and soils. Soil Sci. Soc. Amer.J. 47:753-759. erosion fromrangelands. Proc. Workshop on EstimatingErosion Verma, T.K., J.L Thames, and J.E. Mills. 1977. Soil erosion and and Sediment Yieldon Rangelands. USDA-ARS, ARM-W-26, p. sediment control on the reclaimed coal mine lands of semi-arid 73-95. Southwest.Hydrology and Water Res. In Arizona and the South- Meyer, L.O. 1984. Evolution of the Universal Soil Loss Equation.J. west. 7:41-47. Soil &Water Conserv. 39:99-104. Williams, J.R. 1982. Testing the Modified UniversalSoil LossEqua- Meyer, L.D., and D.L. McCune.1958. Rainfall simulator for Runoff tion. Proc. Workshop on EstimatingErosion and SedimentYield plots. Agr. Eng. 39:644-648. on Rangelands. USDA-ARS, ARM-W-26. p. 157-165. Musgrave,G.W. 1949. Designing agronomic practicesto meetspe- Wischmeier,W.H. 1915. Estimating the Soil Loss Equation's Cover cit ic erosion hazards. J. Soil &Water Conserv. 4:99-102. and Management Factor for Undisturbed Areas. p. 118-124. In: Osborn, H.B., J.R. Simanton, and K.G. Renard. 1976. Use of the Present and Prospective Technology for Predicting Sediment UniversalSoil Loss Equation in the semiarid Southwest.Soil Ero- Yields and Sources. USDA-ARS-40. sion: Prediction and Control. J. Soil & Water Conserv. Special Wlschmeier,W.H., and Smith. D.D. 1965. Predicting rainfall-erosion Pub. #21, p. 41-49. losses from cropland east of the Rocky Mountains. USDA Agr. Renard, K.G.,J.R. Simanton, and H.B. Osborn.1974. Applicability of Handbk.No. 282, 47 p. the UniversalSoil LossEquation to semiarid rangelandconditions Wlschmeier,W.H., and Smith, 0.0. 1978. Predicting rainfall-erosion in the Southwest. Hydrology and Water Res. in Arizona and the losses-aguide to conservationplanning. USDAAgr. Handbk.No. SouthwestIV:18-32. 537, 58 p. R.nard, K.G., and JR. Slmanton.1975. Thunderstormprecipitation Wlschmeler,W.H. 1972. Upslopeerosion analysis. In: Environmental effects on the rainfall-erosion index of the Universal Soil Loss Impacton Rivers. Water Res. Pub.,Fort Collins, Cob., p.15-1—15- Equation. Hydrology & Water Res. in Arizona and the Southwest 26. V:47-55. Zlngg, A.W.1940. Degreeand length of land slope as it affects soil Renard, K.G. 1980. Estimating erosion and sediment yields from loss in runoff. Agr. Eng. 21 :59-64. rangelands.Proc. Am. Soc. Civil Eng. Symp. on Watershed Man- agement, p. 164-175.

What's Cookin' on the Range? Find out what could be cooking in the Trait Boss's Cowboy Cookbook. The 330-page book sells for $13.50 postpaid ($14.50 if outside the US). Order from: Trait Boss's Cowboy Cookbook, 3506 Tilton Valley Drive, Fairfax, VA 22033.