Spatial variability of soil redistribution processes in a small agricultural watershed by John Cornelius Pings A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Earth Sciences Montana State University © Copyright by John Cornelius Pings (1990) Abstract: Physical soil redistribution processes were studied in a small (61 ha) watershed in a region of dryland winter wheat agriculture in north-central Montana. Two approaches were used, a model approach using the Universal Soil Loss Equation (USLE) and Wind Erosion Equation (WEE) soil erosion models, and a field sampling approach using 137Cs. The 137Cs was used as a tracer of erosion and deposition from upland sites (hilltops, midslopes and footslopes) to depositional zones (channels and a pond reservoir bottom). Cesium-137, a fallout product of atmospheric nuclear testing, is strongly adsorbed to clay and has been proven to trace sediment movement. A volumetric approach, developed by De Jong and associates of the University of Saskatchewan, was used to estimate erosion rates of eroding sites and deposition rates of the depositional sites. Landscape units, labelled topographic positions and depositional zones, were defined from a 1:2500 scale plane table topographic contour map, and analyzed for areal concentration of 137Cs to attain erosion rate estimates. A 125 m random sample grid was used to generate USLE and WEE erosion rate estimates. USLE estimates were calculated using the point method of Griffin et al. (1988). WEE estimates were calculated using equations of Skidmore (1988) which were developed to fit the nomographs conventionally used in WEE applications. The model approach yielded an erosion estimate of approximately 9.0 Mg ha-1 yr-1; combining a USLE average estimate of 4.5 Mg ha-1 yr-1 with a WEE average of 4.5 Mg ha-1 yr-1. A site by site comparison of combined model and 137Cs estimates for the 137Cs sample sites yielded a regression output of .07, possibly indicating poor model performance. However, problems in assessing the spatial and temporal variability of soil redistribution indicate a need to further refine the cesium method to reduce variances. Using the 137Cs method, hilltops, midslopes and footslopes were found to be eroding at 22.9, 28.1, and 0.5 Mg ha-1 yr-1, respectively, for a total net erosion rate of 10.4 Mg ha-1 yr-1. Ponds and channels were found to have deposition rates of 243.9 and 43.0 Mg-1 ha-1 yr, respectively, for a total net deposition rate of 5.0 Mg-1 ha-1 yr. The USLE estimated 90 % of the measured value while the WEE predicted only 44 % of the measured wind erosion. The poor model performance and low precision of the cesium method suggests that the use of the models needs to be considered carefully, especially with regard to watershed scale soil erosion assessments. SPATIAL VARIABILITY OF SOIL REDISTRIBUTION PROCESSES IN A SMALL AGRICULTURAL WATERSHED by John Cornelius Pings A th e s is submitted in p a r tia l f u lf illm e n t o f the requirements fo r the degree of Master o f Science in Earth Sciences MONTANA STATE UNIVERSITY Bozeman, Montana June, 1990 ^7 ^ il APPROVAL o f a thesis submitted by John Cornelius Pings This th e s is has been read by each member o f the th e s is committee and has been found to be s a tis fa c to ry regarding content, English usage, form at, c ita tio n s , b ib lio g ra p h ic s ty le , and consistency and is ready fo r submission to the College o f Graduate Studies. ^ Date Chairperson, Graduate Committee Approved fo r the Major Department C -/S ' Ia jo r Department Date Approved fo r the College o f Graduate Studies Z jT 7 Date ^ /??& Graduate Jean iii STATEMENT OF PERMISSION TO USE In presenting th is thesis in p a rtia l fu lfillm e n t o f the fo r a m aster's degree a t Montana State U n iv e rs ity , I agree Library shall make i t a v a ila b le to borrowers under rules o f B rie f quotations from th is thesis are allowable requirements th a t the the Library. without special permission, provided th a t accurate acknowledgement o f source is Permission fo r extensive quotation from or made. reproduction of th is thesis may be granted by my major professor, or in his absence, by the Dean o f L ib ra rie s when, in the opinion of e ith e r, the proposed use m aterial is fo r scholarly purposes. Any copying or use o f the in th is thesis fo r fin a n c ia l gain shall not be allowed without permission. Signature. Date of the m aterial my w ritte n iv I dedicate th is thesis to my w ife , Laura, to my daughter, Lauren and to my great uncle Robert W. 0 ' lo ughlin . M ichelle V ACKNOWLEDGEMENTS I would lik e to thank the members o f my thesis committee fo r th e ir assistance and guidance. My committee included Dr. Katherine Hansen- B ristow, Dr. Joseph Ashley, Dr. W illiam Locke I I I and Dr. Gerald I would e s p e c ia lly lik e to thank my thesis advisor and chairperson, Dr. John Wilson, fo r his guidance and fo r me the p rin c ip le s of sound s c ie n tific research. Drs. Douglas Sherman and Bernard Bauer fo r I Nielsen. committee impressing upon would lik e to thank th e ir help, advice and frie n d s h ip . I would lik e Montana State to acknowldge the Department o f U n iversity research assistantships. fo r t h e ir support Earth including Sciences at teaching I also want to thank Dr. Wilson fo r supporting my research with funds from his National Science Foundation would lik e to thank my f ie ld a ssistan ts, grant. Loretta Thomas and Sanderson; Artem Vartanian and Lee Murray fo r th e ir d ra ftin g ; fe llo w graduate students, in p a rtic u la r Jon Aspie, and I Forrest and my Chet Clarke and Mike Trombetta fo r t h e ir help and comradery. I would lik e to express my appeciation and g ra titu d e to Mr. and Mrs. Norman W. Jackson fo r allowing would lik e to thank Mr. the Choteau County s u ita b le study area Nadwornick, SCS State me to use th e ir farm fo r my research. I Raymond McPhail, SCS D is tr ic t Conservationist of Conservation D is t r ic t , and fo r fo r his help in fin ding access to government data and Mr. a Ron Agronomist, fo r estim ating wheat residue values. vi TABLE OF CONTENTS Page LIST OF TABLES .............................................................................................. LIST OF FIGURES ........................................................................................... ABSTRACT ..................................................... 1. 3. x x ii INTRODUCTION ......................................................................................... I Scope and Purpose ...................................................... Previous Watershed-scale Soil Erosion Studies Model Studies and Applications ................... Previous Cesium-137 Studies ......................... Description of Study Area ........................................................ Thesis Organization .......................... 12 16 METHODS AND DATA SOURCES ...................... r—I CO CO LO 2. v iii 17 Topographic Map Generation ...................................................... Model Estimates ............................................................................. Universal Soil Loss Equation ................................. Wind Erosion Equation ........................................................ Cesium-137 Erosion Estimates .................................................. Cesium-137 Sample S ite Selection ................................. Cesium-137 Sample C ollection ......................................... Cesium-137 Sample Increments ......................................... Cesium-137 Sample Preparation ....................................... Cesium-137 Laboratory Analysis ...................................... Bulk Density Sample C ollection ...................................... Method of Areal Cesium-137 Analysis ........................... Erosion and Deposition Rate and Mass Balance Estimation Method ............................................ 17 20 20 25 33 33 35 37 38 38 39 39 RESULTS .................................................................................................... 42 USLE Erosion Estimates .............................................................. WEE Erosion Estimates ................................................................ USLE and WEE Erosion Rate Estimates Combined ................. Cesium-137 Results .............................. 42 44 48 50 41 v ii TABLE OF CONTENTS--Continued Page 4. DISCUSSION ................................................................... USLE and WEE Soil Loss Estimates ............................................ V a lid a tio n o f USLE and WEE Soil Loss Estimates ............... Soil Erosion/Deposition Rates In fe rre d from Cesium-137 Gains/Losses .......................................................... Conclusions ....................................................................................... 60 60 64 66 70 REFERENCES C IT E D ......................................................................................... 71 APPENDICES........................................................................................................ 78 APPENDIX A. EROSION MODEL RESULTS ........................................ APPENDIX B. CESIUM-137 LABORATORY DATA AND AREAL CONCENTRATIONS .......................................... APPENDIX C. SIEVING RESULTS ..................................... 79 86 91 v iii x LIST OF TABLES Table 1. Page Erosive Wind Energy Occurring by Month a t Great F a lls , MT ................................................................................. 15 Confidence o f Table S tation B Elevation Determined by D iffe re n t Sightings .................................................. 21 Computation o f Average Annual C Factor fo r T ille d Soils ..................................................................................... 24 4. Soil E r o d ib ility by Soil Mapping Unit .......................................... 29 5. Estimates of K Factor, Vegetation Weight and Erosive Wind Energy fo r Cropstage Periods and T illa g e Operations, Jackson Farm ............................................ 30 6. Sample WEE Computation fo r Sample Point ID ................................ 33 7. USLE Estimates ......................................................................................... 43 8. WEE Soil Loss Estimates ....................................................................... 47 9. Cesium-137 Erosion and Deposition Rate Estimates ................... 55 10. Average Cesium-137 Erosion and Deposition Mass Rates by Topographic Position .......................................................... 59 Comparison of USLE Factor Estimates Used by Author and USDA-SCS ......................................................... 61 Comparison of WEE Factor Estimates Used by Author and WEQ Factor Estimates Used by USDA-SCS ................................. 62 S ite by S ite Comparison of Model and 137Cs Erosion/Deposition Rate Estimates ............................................ 67 Summary o f Watershed Average 137Cs Erosion and Deposition Mass Estimates .......................................................... 69 15. USLE Factor and Soil Loss Point Estimates ................................. 80 16. WEE Factor and Soil Loss Point Estimates ....................... -........... 83 17. Cesium-137 Laboratory Data and Areal Concentrations ............. 87 2. 3. 11. 12. 13. 14. ix LIST OF TABLES--Continued Table ,18. Page Sieving Results ............... ....................................................................... 92 :x LIST OF FIGURES Figure L Page Cesium-137 in a drainage basin Tfrom Campbell e t a l . , 1982) ........................................ 8 2. Study area and lo c a tio n map . . . ................... .......... ............................. 13 3. Soil series map showing s o il mapping units as defined by the s o il survey team, SCS Choteau County Conservation D is t r ic t ............................................................................. 15 Map showing lo c a tio n o f baseline AB and plane ta b le s ta tio n s A, B, C, D and E w ith in study watershed ................... 18 Map showing lo c a tio n o f randomly selected 125 m g rid ,points and la b e ls used fo r USLE and WEE c a lc u la tio n s ........... 22 6. Diagram showing topographic po sition s ........................................... 35 7. Study area map showing lo catio ns o f 137Cs samples w ith in study watershed and lo catio n s o f control samples on the Jackson Farm ................................................................... 36 Areas o f the topographic positions used fo r 137Cs e x tra p o la tio n ................................................................................... 37 D is trib u tio n o f USLE s o il loss ra te s fo r C arter watershed (n = 81) ........................ 43 S p a tia l v a r i a b ili t y o f USLE s o il loss rates fo r C arte r watershed ....................................................................................... 45 D is trib u tio n o f s o il loss rates fo r C arter watershed using WEE po int method and 125 m g rid (n = 80) ........................ 47 S p a tia l v a r i a b ili t y o f WEE s o il loss rates fo r C arte r watershed ...................................................................................... 49 D is trib u tio n o f to ta l s o il loss ra te s fo r C arter watershed using USLE/WEE point methods and 125 m g rid (n = 81) ........................................................................ 50 S p atial v a r i a b ili t y o f s o il loss rates fo r C arter watershed using USLE and WEE point methods .............................. 51 4. 5. 8. 9. 10. 11. 12. 13. 14. xi LIST OF 'FISORES--CCTtanued Figure 15. 16. 17. Page S ite areal 137Cs a c t iv it y fo r h illt o p s , midslopes, foots!opes, t i l l e d channels, incised channel, pond and control s ite s ..................................................................................... 52 Range and average 137Cs areal concentrations by topographic p o s itio n ........................................................................ 54 S c a tte rp lo t diagram o f p redicted and measured values generated fo r s it e by s ite comparison o f 137Cs sample s ite s ................................................................................................ 67 x ii ABSTRACT Physical s o il r e d is trib u tio n processes were studied in a small (61 ha) watershed in a region o f dryland w in te r wheat a g ric u ltu re in northc e n tra l Montana. Two approaches were used, a model approach using the Universal Soil Loss Equation (USLE) and Wind Erosion Equation (WEE) s o il erosion models, and a f i e l d sampling approach using 137Cs. The 137ICs was used as a tra c e r o f erosion and deposition from upland s ite s ( h illt o p s , midslopes and foots!opes) to depositional zones (channels and a pond re s e rv o ir bottom ). Cesium-137, a f a llo u t product o f atmospheric nuclear te s tin g , is stro n g ly adsorbed to c lay and has been proven to tra c e sediment movement. A volum etric approach, developed by De Jong and associates o f the U n iv e rs ity o f Saskatchewan, was used to estim ate erosion ra te s o f eroding s ite s and deposition ra te s o f the depositional s ite s . Landscape u n its , la b e lle d topographic positions and depositional zones, were defined from a 1:2500 scale plane ta b le topographic contour map, and analyzed fo r areal concentration o f 137Cs to a tta in erosion ra te estim ates. A 125 m random sample g rid was used to generate USLE and WEE erosion ra te estim ates. USLE estim ates were c a lc u la te d using the po in t method o f G r if f in e t a l . (1 9 8 8 ). WEE estim ates were c a lcu lated using equations o f Skidmore (1988) which were developed to f i t the nomographs conventionally used in WEE a p p lic a tio n s . The model approach y ie ld ed an erosion estim ate o f approxim ately 9 .0 Mg ha'1 y r'1; combining a USLE average estim ate o f 4 .5 Mg ha"1 y r'1 w ith a WEE average o f 4 .5 Mg ha'1 yr"1. A s it e by s ite comparison o f combined model and 137Cs estim ates fo r the 137Cs sample s ite s y ie ld e d a regression output o f .0 7 , possibly in d ic a tin g poor model performance. However, problems in assessing the s p a tia l and temporal v a r i a b ili t y o f s o il r e d is trib u tio n in d ic a te a need to fu r th e r re fin e the cesium method to reduce variances. Using the 137Cs method, h illt o p s , midslopes and foots!opes were found to be eroding a t 2 2 .9 , 2 8 .1 , and 0 .5 Mg ha'1 y r'1, re s p e c tiv e ly , fo r a to ta l net erosion ra te o f 10.4 Mg ha"1 y r'1. Ponds and channels were found to have deposition ra te s o f 243.9 and 43.0 Mg'1 ha"1 y r , re s p e c tiv e ly , fo r a to ta l net deposition ra te o f 5 .0 Mg"1 ha"1 y r . The USLE estim ated 90 % o f the measured value w h ile the WEE predicted only 44 % o f the measured wind erosion. The poor model performance and low p recision o f the cesium method suggests th a t the use o f the models needs to be considered c a r e fu lly , e s p e c ia lly w ith regard to watershed scale s o il erosion assessments. I CHAPTER I INTRODUCTION Scope and Purpose The increasing scale of human impact in a g ric u ltu ra l areas and a desire to reduce soil erosion problems have provided the impetus fo r erosion assessments in recent years. (USDA) Soil Conservation the Universal Soil Loss Equation (WEE) to and recent work inputs to State soil The U. S. Department of A gricultu re Service (SCS) re g u la rly u t iliz e s models such as Equation (USLE) and a version o f the Wind Erosion guide the implementation of conservation procedures, has focused on improving the methods o f estim ating model produce erosion estimates th a t are s p a tia lly v a ria b le . In the of Montana there is a need fo r large scale so il erosion studies to assist in the development of soil erosion/crop p ro d u c tiv ity S im ila rly , there is a need to c o lle c t so il erosion data these modeling e ffo r ts so th a t the contribution of the assessments. independent of modeling e ffo r ts , themselves, can be evaluated. This study addresses both needs by quantifying deposition in a small a g ric u ltu ra l catchment. used. s o il erosion and Two approaches have been The f i r s t approach uses the USLE and WEE to estim ate soil losses from water and wind, re sp ec tiv e ly , in the watershed. The second approach uses the s p a tia l v a r ia b ilit y of Cesium-137 ( 137Cs) detected in so il samples 2 to q u a n tify erosion from both water and wind and deposition by w ater. The USLE and WEE have emerged as the most w idely used s o il erosion models in North s ta te /p ro v in c ia l America. It assessments has been used fo r in the United States many national and and Canada and fo r watershed- and p lo t-s c a le studies in both countries (T rim b le, 1974, 1977, 1983; van V li e t and W all, 1979, 1981; Coleman, 1982; S n e ll, 1984, 1985; Wilson, 1989). SCS to USLE and WEE s o il loss estim ates are also used by the USDA determine q u a lific a tio n and maintenance requirements fo r Conservation Reserve Programs (C .R .P .) in Montana. The techniques used in th is study to estim ate erosion and deposition ra te s from 137Cs areal associates (1981a, (1 9 8 3 ). concentrations were developed 1981b) and la t e r re fin e d by De Jong and The 137Cs isotope acts as a tra c e r o f physical s o il processes by i t s by adhesion to fin e s o il g ra in s . It Brown and associates re d is trib u tio n is a by-product o f atmospheric nuclear te s tin g and is d e liv e re d through p r e c ip ita tio n and w in d -c a rried sediments. depo sition al processes Pennington e t a l . , It s has use been as an in d ic a to r widespread 1976; McHenry and R itc h ie , 1980; Wise, 1980; Brown e t a l . , 1984; Arnalds e t a l . , 1981b; De Jong of (R itc h ie erosional et a l., and 1974; 1977; McCallan e t a l . , et a l., 1983; Arnalds, 1989; Dibb, 1989). Using these approaches, the o b je ctiv es o f th is study were tw ofold. The f i r s t o b je c tiv e was to estim ate s o il losses with the USLE and WEE fo r a small a g ric u ltu ra l watershed. The second o b je c tiv e was to q u an tify s o il erosion and deposition using 137Cs and to use these re s u lts to evaluate USLE and WEE performance in the same watershed. The fo llo w in g kinds o f data were generated and analyzed to answer these research questions: I ) 3 analysis o f c lim a tic , s o il, topographic and vegetative cover factors to produce USLE and WEE so il loss estimates at m u ltip le s ite s ; 2) analysis o f 137Cs samples to define to ta l and incremental 137Cs areal concentrations fo r several s ite s , and determine r e la tiv e erosion and deposition; and 3) extra p o la tio n of USLE and WEE estimates and 137Cs areal concentration averages by landscape units to obtain erosion and deposition rates fo r those units and the e n tire watershed. Previous Watershed Scale Soil Erosion Studies Model Studies and Applications The USLE was derived from 10,000 plot-years o f data at locations throughout the United States. q u a n tific a tio n of fa c to r This model estimates erosion through the values fo r r a in f a ll- e r o s iv it y (R ), so il c r e d ib ilit y (K ), topographic factors defined by slope length (L) and slope steepness (S ), cover management (C) and supporting practices (P ). The USLE is freq u en tly used by the SCS as a tool to determine conservation p ractices fo r strip-cropping the control and te rra c in g of flu v ia l erosion, (Wischmeier, including contour 1976; Wischmeier and Smith, 1978). Some recent studies have tr ie d to improve the USLE by developing new methods o f estim ating fa c to r values, p a r tic u la r ly the topographic fa c to rs , L and S. W illiams and Berndt (1977), fo r example, proposed a method of generating slope frequency data by using a th ird -o rd e r natural spline function developed by G reenville (1967) fo r points defined by horizontal distances and elevations of contours th a t cross grid lin e s on topographic maps. Slope in the d ire c tio n of the g rid lin e s was determined by 4 d if fe r e n t ia t in g the s p lin e function at each g rid in te rs e c tio n p o in t. Wilson (1986a, 1986b) developed a d if fe r e n t approach using topographic map input data and GreenviT ie 's .(1967) s p lin e function to estim ate slope le n g th , shape and steepness fo r slope p r o file s th a t cross the elevatio n contours perpendicularly.. to d iv id e ir r e g u la r computer slope generated method o f estim ate LS values. o f estim ating S t a t is t ic a l analysis o f slope segments was used slope Foster p r o file s in to and Wischmeier segments (1974) and the was used to G r if f in and his associates (1988) developed a method LS fa c to r values fo r a series o f random p o in ts. T h eir method used the distance downslope from the top o f the slope p r o f ile as w e ll as cum ulative and slope segment gradients to estim ate LS values fo r s p e c ific points in a landscape. A more comprehensive re v is io n o f several fa c to rs w ill re s u lt in the p u b lic a tio n o f a computer program and manual fo r the Revised Universal Soil Loss Equation (RUSLE) in 1990 (Renard, 1989, personal communication). The o rig in a l USLE and RUSLE w ill most l i k e l y be replaced by a completely new physically-based modeling technology in the mid-1990s (F o s te r, 1989, personal communication). The Wind Erosion Equation (WEE) was developed a t about the same time as the USLE to assess the s u s c e p tib ility o f f ie ld s o ils to wind erosion and to help p ra c tic es w ith (Chepil the et s e le c tio n a l., 1962; and design o f wind erosion Woodruff and Siddoway, 1965). control This equation estim ates erosion as a function o f magnitude and d ire c tio n o f wind as w ell as s o il c r e d ib i lit y ( I ) , s o il ridge roughness (K ), vegetation o rie n ta tio n and cover ( V ) , and f ie ld fe tch length ( L ) . was used in e a rly conservation a p p lic a tio n s . The WEE equation Skidmore and Woodruff (1968) 5 la t e r compiled p e rtin e n t c lim a tic data fo r many s ta tio n s from e x is tin g sources to a s s is t in the a p p lic a tio n o f the WEE throughout the country. The model and c lim a tic data were published fo r farmers and conservation program workers. In one o f the most innovative a p p lic a tio n s o f the WEE to d a te , Bondy e t a ! . (1980) estimated wind erosion by cropstage period as a function o f wind energy d is tr ib u tio n . T h e ir method used tem porally v a ria b le v e g e ta tiv e residue e q u iva len ts, s o il t i l l a g e conditions and s o il c r e d ib i lit y to estim ate wind erosion fo r 10 d if fe r e n t cropstage periods in a w in te r wheat/summer fa llo w system in Kansas and a spring wheat/spring w h e a t/fa llo w system in North "Dakota. Skidmore (1988) substitu ted equations to estim ate WEE fa c to r values fo r the tab les and nomographs o f Woodruff and Siddoway (1 9 6 5 ). These equations, f i r s t proposed by W illiam s e t a l . (1 9 8 4 ), e lim in a te the need to in te rp o la te fa c to r inputs and reduce the time and e f f o r t needed to apply the WEE. The SCS A g ric u ltu ra l Research Service developed a version o f the WEE, the WEQ, fo r use by SCS personnel in f i e l d a p p lic a tio n s . Previous Cesium-137 Studies Cesium-137 is a w idely dispersed ra d io a c tiv e isotope th a t is a by-product o f atmospheric nuclear te s tin g which the United S ta te s, Soviet Union and United Kingdom began on a frequent basis in 1954 (Campbell e t a l . , 1982). These te s ts are s t i l l c a rrie d out on a much sm aller scale by France China and (Anonymous, 1989). The period of most intense atmospheric dispersion occurred between 1962 and 1965 immediately before the United S ta te s , Soviet Union and 41 other nations signed a tr e a ty suspending atmospheric te s tin g o f nuclear weapons in 1966 (Campbell e t 6 a l., 1982). Cesium-137 has a regional d is tr ib u tio n p re c ip ita tio n sources and q u a n titie s . the stratosphere d e liv e re d to s o ils and troposphere th a t is lin k e d to The isotope is transported through by global through two methods. c irc u la tio n patterns and One method involves d e liv e ry through p r e c ip ita tio n , and the second method involves dry deposition o f 137Cs w ith atmospheric p a r tic u la te m atter from the atmosphere. I t s lo cal concentration is dependent upon the a v a ila b le amount o f the isotope in the atmosphere at the tim e of p r e c ip ita tio n events, its a lt it u d e , and p re v a ilin g regional and lo cal m eteorological conditions (McCallan e t a l . , 1980). Several assumptions govern the use o f th is isotope as a tra c e r o f s o il r e d is tr ib u tio n process a n a ly s is . I t is assumed th a t: I ) the isotope has been d e liv e re d to the watershed uniform ly; 2) the 137Cs becomes adsorbed to the c la y - and s ilt - s iz e d s o il p a r tic le s when i t reaches e a rth ; 3) these sediments have not undergone any s o rtin g ; and 4) the re d is tr ib u tio n o f 137Cs by w in te r winds, p la n t and animal l i f e has been minor (< 5 percent) and/or a t le a s t uniform throughout an in d iv id u a l watershed (Brown e t a l . , 1981a, 1981b; De Jong e t a l . , 1982, 1983; Arnalds, 1984). Cesium-137 can be used to q u an tify s o il erosion and deposition rates because its (dispersion h a lf-life (3 0 .2 since 1954) years) and a d d itio n a l v a ria b le s topographic p o s itio n , period of allow estim ates o f average annual deposition ra te s to be made (Brown e t a l . , known, known 1981a). such as in te n s ity existence erosion and Once these rates are o f areal concentration, and maximum depth o f 137Cs a c t iv it y in the s o il p r o f ile f a c i l i t a t e the d e lin e a tio n o f erosional and depo sition al zones in watersheds because the 137Cs moves only w ith sediment th a t is transported 7 (De Jong e t a l . , 1982, 1983). A conceptual model o f 137Cs in p u t, a c t iv it y and tra n s p o rt in a drainage basin is reproduced in Figure I . Most studies use th is type o f model to q u a n tify erosion and deposition ra te s . The bar diagrams o f ty p ic a l 137Cs areal concentrations fo r various landscape un its in Figure I il lu s t r a t e the v a ria b le nature o f 137Cs a c t iv it y a t d if fe r e n t lo ca tio n s in a watershed. Several studies have examined 137Cs in te s t p lo ts and watersheds to determine regional concentrations ( e . g . , Rogowski and Tamura, 1970; Lance et a l., 1986; Kachanoski, average 137Cs areal 1987; Dibb, 1989). concentrations between 3 .6 Arnalds (1984) measured and 20.2 pi cocuries per square cm (pCi cm'2) a t 12 s ite s throughout Montana. These concentrations were stro n g ly re la te d to p re c ip ita tio n (R2 = 0 .9 2 ), in d ic a tin g th a t lo cal p r e c ip ita tio n to ta ls w ill provide a good f i r s t estim ate o f expected 137Cs le v e ls (A rnalds, 1984; Arnalds et a l., 1989). Given the previously mentioned regional sources and controls and the assumptions noted above, several researchers have constructed 137Cs mass balances to in fe r s o il erosion and deposition ra te s (De Jong e t a l . , 1983; Arnalds, 1984; Pennock and De Jong, 1987). Cesium-137 has been applied to erosion and sedimentation studies in a v a rie ty o f contexts. Several studies have shown th a t concentrations have increased in areas o f sediment accumulation such as v a lle y flo o r s , la k e s , re s e rv o irs and s a lt marshes ( e . g . , McHenry e t a l . , et a l., 1974, 1975; Pennington et a l., 1976; Delaune 1973; R itc h ie et a l., 1978; McCallan e t a l . , 1980; Brown e t a l . , 1981a, 1981b; Campbell e t a l . , 1982; De Jong et a l., 1983; Arnalds, 1984; Pennock Cesium-137 has been used to determine erosion o f and De Jong, natural and 1987). c le a rc u t 8 O t a ’ Ce AND P R E C IP IT A T IO N V O IN P U T V Figure I . Cesium-137 in a drainage basin (from Campbell e t a l . , 1982). fo re s ts and f o r e s t / f ie ld systems at a v a rie ty o f s p a tia l scales (Brown e t a l., 1981a, 1981b; Campbell e t a l . , studies have a g ric u ltu ra l et a l . , 137Cs to evaluate s o ils (Brown e t a l . , s o il erosion and 1981b; Campbell e t a l . , 1988). Other deposition on 1982; De Jong 1983; Arnalds, 1984; Pennock and De Jong, 1987). Recent e n tir e used 1982; Lowrance e t a l . , studies watersheds have q u a n tifie d by using weighted d iffe r e n t landscape u n its Arnalds, and deposition average areal (Brown e t a l . , 1984; Pennock and De Jong, developed to d is tin g u is h erosion fo r concentrations fo r 1981b; De Jong e t a l . , 1987). landscape u n its . rates 1983; Several methods have been For example, (1981b) divided t h e ir watershed in to erosional Brown e t and depositional a l. zones, 9 w ith h illto p s and midslppes representing erosional zones and foots!opes and an a llu v ia ! fan representing depositiona! zones. De Jong e t a ! . (1 9 8 3 ) , on the other hand, distinguished three landscape u n its m iddle, and lower slopes - determined in the f ie ld by pacing. (1984) added two fu rth e r un its to th is c la s s ific a tio n , between h illto p s - and midslopes, and toes!opes upper, Arnalds shoulder slopes beneath lower slopes (fo o ts lo p e s ) but defined his tran sects by s e le c tin g s o il samples fo r 137Cs analysis 50 m a p a rt. to De Jong e t a l . Pennock and De Jong (1987) used a s im ila r approach (1983) in an attempt to decipher the influence p r o f ile and plan curvature on s o il r e d is tr ib u tio n processes. of A d ig it a l te r r a in model was created from an e x is tin g d ig itiz e d topographic survey and used to d is tin g u is h convergent and divergent backs!opes, convergent and divergent shoulders, convergent and divergent foo tslo p es, and le v e l areas (Pennock and De Jong, 1987). Care must be exercised when using th is term inology because the landscape units, are defined and labeled using d iffe r e n t methods and d e fin itio n s general in d if fe r e n t s tu d ies . approach o f e x tra p o la tin g average 137Cs areal However, the concentrations to landscape u n its is now w idely used to estim ate average annual erosion and deposition ra te s . Most recent studies have undertaken 137Cs d etection w ith labo ratory analysis using a lith iu m -d r ifte d germanium semi-conductor gamma ray d e te c to r coupled to a nuclear data m ulti-channel analyzer (Cutshall and Larsen, 1980; Larsen and C u ts h a ll, 1981; Brown e t a l . , Jong et a l., 1983; Arnalds, 1984). Cesium-137 1981a, 1981b; De concentrations are c a lc u la te d by m u ltip ly in g the 137Cs a c t iv it y in core samples (pCi g"1) by the s o il mineral bulk density (g cm"3) and depth o f the s o il sample (cm). 10 Results are expressed as r a d io a c tiv ity per u n it surface area of soil surface or the "areal concentration" ( e . g . , Brown e t a l 1981b; De Jong e t a l . , 1983; Arnalds, 1984; Pennock and De Jong, 1987). Campbell e t a l . (1982) used a d iffe r e n t approach to determine 137Cs areal concentrations. Their approach r e lie s on the fa c t th a t 137Cs adheres to fin e s o il fra c tio n s , so th a t the areal concentrations are calculated using only the s i l t and clay fra c tio n s . S i l t and clay frac tio n s were estimated using the hydrometer method and th e ir bulk d e n s itie s were used instead o f th a t o f the to ta l sample. This study used an experimental watershed in New South Wales, A u s tra lia , which was found to have 137Cs d is trib u te d evenly in upper so il layers due to plowing. Areal concentrations were q u an tifie d fo r a v a rie ty of topographic positions in upland and lowland areas although no erosion or deposition rates were estimated (Campbell e t a l . , 1982). In co ntrast, many North American studies have tr ie d to estimate erosion and deposition zones and rates from 137Cs an alysis. Brown and associates (1981b) computed erosion estimates ranging from 3 to 27 t ha"1 yr"1 fo r erosional zones based upon detected 137Cs ranging from 3 .5 to 15.2 pCi cm'2 in two W illam ette V a lle y , OR watersheds. foots!opes tested were found to be eroding Six out of eight (Brown e t a l . , 1981b). In another study, De Jong e t a l . (1983) found upper slopes to have lo s t 200 to 600 t ha"1 o f s o il while lower slopes were found to have gained 250 to 800 t ha"1. However, middle slopes were found to be both depositional and erosional over 20 to 25 years in eight small Saskatchewan, Canada basins with g la c ia l s o ils (De Jong et a l . , 1982; 1983). 41 Arnalds (1984) examined erosion in a small watershed near Power, MT in which he detected 137Cs a rea l concentrations ranging from 2 .4 to 24.6 pCi cm"2. H illto p s and foots!opes both were found to be eroding a t 16.5 Mg ha'1 y r " \ w h ile shoulders and midslopes were eroding a t 20.9 and 45.1 Mg ha"1 yr"1, re s p e c tiv e ly . Xoeslopes were found to have deposition occurring a t a ra te o f 9 .9 Mg ha"1 yr"1. and foots!opes The estim ates fo r the shoulders, mid slopes compared favorab ly w ith USDA SCS s o il loss estimates produced w ith the USLE and WEE models. Lance et a l. (1986) c o lle c te d data on 137Cs a c t iv it ie s in the southwestern United States as w ell as in adjacent c u ltiv a te d and grassed watersheds in Oklahoma. A major conclusion o f th is study based upon the re s u lts o f a s o il mass balance was th a t 137Cs a c t iv it y might be a more s e n s itiv e in d ic a to r o f s o il p ro d u c tiv ity losses than measurements o f to ta l mass o f s o il removed from a f ie ld caused by highly lo c a liz e d erosion and d e p o sitio n . M cIntyre e t a l . (1987) found a c le a rc u t fo re s t to be eroding a t only 0 .2 t ha"1 yr"1 and a ttrib u te d the erosion to s o il compaction by liv e s to c k grazing o f n a tiv e grasses. In a re la te d study Lowrance e t a l . (1988) estim ated erosion and deposition ra te s o f 63 and 256 Mg ha"1 y r'1, re s p e c tiv e ly , fo r a rip a ria n f o r e s t / f ie ld system watershed on the southern Georgia coastal p la in . The unusually high deposition was a ttrib u te d to sediment th a t was transported from upstream lo catio ns and deposited during flo od events. Three d if fe r e n t c a lc u la tio n methods produced n early id e n tic a l re s u lts . Pennock and De Jong (1987) examined more landform elements than the other studies and found four o f seven to be erosive, from most to le a s t erosive: convergent shoulders, divergent backs!opes, convergent backs!opes 12 and divergent shoulders. The depositional units were, in order from le a s t depositional to most: divergent foots!opes, level foots!opes. areas and convergent Soil loss predictions fo r these same areas using the USLE were two to nine times lower than those indicated by the 137Cs method (Pennock and De Jong, 1987). Description of Study Area The study watershed covers approximately 61 ha and is located in Choteau County near C arter, Montana at 47° 53'30" N and 110° SZzOO" W (Figure 2 ). It is located on the U.S.G.S. C arter N.E. 7.5 minute quadrangle in the SW and SE 1 /4 's of section 30, Township 25 North, Range 7 East. The watershed is located w ithin a mapped u n it of the Colorado Shale and is located approximately 40 km north of the southernmost extent o f the Laurentide ice sheet. The watershed consists o f hummocky te rra in o f g la c ia l o rig in . Local r e l i e f is 22 m, with the highest elevations along the western boundary approaching 922 m. The watershed is drained by an unnamed trib u ta ry of the Frank G ilb e rt Coulee, which drains to the Teton River and eventually to the Missouri R iver. The area is a mixture of gentle slopes in the west and steeper slopes along the eastern boundary. A well defined ephemeral channel system drains in to an incised channel and a 0.25 ha pond situated behind an earth dam constructed in 1973. The annual average p re c ip ita tio n measured in nearby Great F alls is 390 mm yr"1, with a May-June maximum. Average annual r a in f a ll varied from 186 to 475 mm yr"1 between 1954 and 1986. F a lls between Analysis o f wind data fo r Great 1950 and 1955 indicates a predominant annual wind d ire c tio n 13 • J a c k s o n Form, C arter, M ontana _ — Ephemeral Channels W atershed Boundary Pond ------- - I M eter Contours LOCATION MAP NO SCALE 100 200 Meters N orth Figure 2. Study area and lo catio n map. 47° 53' 30" N and 110° 52' 00" W. The Jackson Farm is located at o f west-southwest (Skidmore and Woodruff, 1968). Of p a rtic u la r in te re s t to th is study is the percent of cumulative erosive wind energy (EWE) th at occurs by a p a rtic u la r month (see Table I ) . c r it ic a l According to the data, the period fo r erosive wind is between October and A p r il, during which 83 % of the erosive wind occurs. 14 The s o ils A rg ib o ro lls of (E thridge m ontm orillonit i c (unpublished the study area s e rie s , are a f in e , s ilty m ontm orillonit i c clay loam) and fin e , f r i g i d , Typic Albaqualfs (Nishon s e rie s , a s i l t y c la y ) s o il survey maps, Mr. Raymond M cPhail, C o n servatio n ist, Choteau County, MT, 1987) (Figure 3 ) . p a rt o f the Mr. Norman W. Jackson farm which is c u ltiv a tio n o f w in te r wheat. fa llo w A rid ic system. SCS D is t r ic t The watershed is used mostly fo r the The farm is managed in a w in te r wheat/summer The watershed is almost completely tille d with the exception o f the incised channel and pond areas and t h e ir margins. The watershed h a lf-c e n tu ry . has undergone considerable change in the la s t The middle portion o f the study area was form erly a sm all, seasonal la k e . In the 1940's, a channel was incised from the lake to the coulee bottom to drain the lake to allow t i l l a g e . The incised channel remains an a c tiv e component o f the f lu v ia l system, channeling overland flow from the upland areas o f the watershed. An earth dam was constructed across the mouth o f the coulee to create a pond in 1973. The landowner had planned to stock th is pond with fis h but the pond has always emptied a fte r p r e c ip ita tio n (Norman Jackson, events C a rte r, and spring MT, personal ru n o ff because o f a slow leak communication, 1987). This has prevented permanent f i l l i n g o f the re s e rv o ir, but i t drains slowly enough (a period o f months) to be an e f f ic ie n t sediment tr a p . nature o f the pond was b e n e fic ia l to the present provided access to re s e rv o ir bottom sediments. The study ephemeral because i t 15 Table I . Erosive Wind Energy Occurring by Month at Great F a lls , MT.3 J F M A M J J A S O N D Monthly Total (%) 15 14 9 9 5 4 2 2 4 9 12 15 Cumulative Total (%) 15 29 38 47 52 56 58 60 64 73 85 100 Month aErosive wind g re a te r than 8 m sec"1. Agronomy Manual, Source : USDA SCS 1988 National 385B LEGEND: WATERSHED BOUNDARY 200 M eters North Figure 3. Soil series map showing s o il mapping un its as defined by the s o il survey team, SCS Choteau County Conservation D i s t r i c t . Ethridge s o ils are 38A, 38B and 3858, and Nishon s o il is 28. S tip p led area was undefined by SCS. 16 Thesis Organization This f i r s t chapter has described the scope and purpose o f the study, previous s o il erosion studies which use the USLE and WEE models and 137Cs techniques, and the study watershed. Chapter 2 describes the methods o f data c o lle c tio n , assumptions used and the computations required to use both the so il erosion models and 137Cs method. I t describes the method used fo r the topographic survey and the d iv is io n o f the watershed into landscape u n its . Data sources and methods used to estim ate USLE and WEE fa c to r values are also described. sampling procedures and laboratory analysis fo r 137Cs detection Soil in the watershed are outlined as are the methods of c a lc u la tio n of the areal concentrations, erosion and deposition ra te s , and watershed mass balance. Chapter 3 is a substantive review of a ll the re s u lts from both the model and 137Cs methods. I t includes average annual erosion ra te estimates from the USLE and WEE, as well as an analysis of the s p a tia l v a r ia b ilit y of those estim ates. each s ite , The 137Cs re su lts included are I ) the 137Cs a c tiv ity fo r 2) the so il average erosion average areal erosion and deposition rates fo r each s it e , 3) and deposition rates by topographic p o sitio n , concentrations by topographic po sition s, 4) the and 5) a to ta l watershed erosion and deposition mass balance. Chapter 4 is a discussion of the re su lts and the im plications of the performance o f these methods of estim ating so il erosion and deposition to s o il erosion p ro d u c tiv ity n o rth -cen tral Montana. assessments of other s im ila r watersheds in Possible sources of e rro r are also discussed. The fin a l p art of Chapter 4 is the conclusions section. 17 CHAPTER 2 METHODS AND DATA SOURCES Topographic Map Generation A The study area was mapped using a plane ta b le , telesco pic alidade and the beaman arc procedure described by Compton (1962). The topographic map was drawn at a scale o f 1:2500 using a i m contour in t e r v a l. The s ta rtin g elevation was determined by lo catin g the highest point in the watershed and assigning i t the elevation o f the point o f highest r e l i e f on the U.S.G.S. C arter NE 15 minute quadrangle (3023 fe e t or 921.4 m) (Table Station A in measurements were Figure 4 ). A ll elevations and lin e a r distance rounded to the nearest .1 m. A baseline was established between two central h illto p s (Figure 4 ). The baseline was hand measured twice (632.2 m and 632.6 m) using a 100 m tape, giving an average distance of 632.4 m. Plane ta b le measurements of the base lin e f e l l w ith in ± 4 m of th is length, in d ic a tin g errors of less than .7% when measuring lin e a r distances. No check on the accuracy of elevation measurements is possible using the method described by Compton (1962). Following establishment of the baseline, flags were set at to f a c i l i t a t e telescopic sighting from anywhere in each end the watershed. This was not possible in several instances, so two additional flags were set 18 Baseline North 200 LEGEND: Meters WATERSHED BOUNDARY Figure 4. Map showing lo c a tio n o f baseline AB and plane ta b le stations A, B, C, D and E w ith in study watershed. up to perm it the sig h tin g of at le a s t tr ia n g u la r resection from any point in the three fla g s watershed. which allowed a The elevations and lo catio n s o f the fla g s ite s were established with the a lid ad e r e la tiv e to the known e le v a tio n a t one end o f the baseline and it s le n g th . points became plane ta b le s ta tio n s A, B, C and D. These fla g 19 One a d d itio n a l ta b le s ta tio n (E) was added to allow coverage o f extreme northern p a rt o f the watershed. map using a tr ia n g u la r ta b le s ta tio n s was used -the I t was located on the plane ta b le resection method. Hence, a to ta l o f fiv e plane and those established a t fla g s ta tio n s were located e x a c tly I m east o f each fla g . Four s ta tio n s could be relocated from sig h tin g anywhere with the The mapping was accomplished by sig h tin g the s ta d ia rod through the te le s c o p ic in the watershed by t h e ir fla g s a lid a d e . te le s c o p ic alid ade to reveal a s ta d ia in te rc e p t, d ire c tio n h a ir reading. Trigonom etric c a lc u la tio n s y ie ld e d distance and d iffe re n c e s and d ire c tio n was established by marking the w ith a ray drawn along the fid u c ia l edge o f the Topographic fe atu res were id e n tifie d s ta d ia rod d ir e c t ly on them. in the was drawn in the f ie ld to allow fo r d e fin itio n o f lo ca l topography points were The cross ele v a tio n plane ta b le map te le s c o p ic a lid ad e. watershed by placing the Two hundred and on the plane ta b le map in th is manner. Data and eigh ty points were marked f i r s t d r a ft o f a contour map c o lle c tio n o f more points wherever became d i f f i c u l t . elim inated if the e levatio n s could not be reproduced using the mathematical formulae described in Compton (1962). Four data points (1.4% o f those mapped in the f i e l d ) were elim inated because there was a d iffe re n c e between the elevations ca lc u la te d in the f i e l d and those generated during subsequent rechecking o f computations. D ifferences between the e le v a tio n computed fo r ta b le s ta tio n B from ta b le s ta tio n A and from la t e r , shorter sightings o f data points A28 from ta b le s ta tio n A27 and B were discovered when the computations rechecked. The e le v a tio n a t ta b le s ta tio n B was determined from were Table 2, 20 which shows the r e la tiv e confidence values associated w ith the individual telesco pic alidade sigh ting s. Hence, an elevation o f 918.7 m was chosen fo r ta b le s ta tio n B due to the high confidence values associated with shorter sigh ting s. The elevations determined a fte r the sighting of s ta tio n B were adjusted to r e fle c t th is change in elevation (since ta b le stations C, D and E re lie d on th is elevation e levatio n s) and a ll data points were rechecked. were redrawn and a fin a l to determine th e ir A ll elevation contours revision o f the topographic contour map was drawn. Model Estimates Universal Soil Loss Equation The modified version of the USLE developed by G r if f in e t a l . (1988) to estim ate sheet and r i l l erosion at d iffe r e n t points in a landscape was used in th is study. This equation, which combines the method of Foster and Wischmeier (1974) fo r measuring the L and S facto rs fo r irre g u la r slopes with the o rig in a l Smith, model (Wischmeier and 1978), can be w ritte n as follow s: D = (m + I ) R K ( x / l u) m S C P (I) where D is so il loss in tonnes per hectare per year ( t ha"1 yr"1) , m is slope length exponent, R is the r a in f a ll- e r o s iv it y fa c to r in the megajoules per m illim e te r per hectare per hour per year (Md mm ha'1 hr"1 yr"1) , K is the s o il c r e d ib ilit y fa c to r in tonnes per hectare per per megajoules per hour per hectare m illim e te r ( t ha hr ha"1 MJ"1 mm"1) ; x is the distance o f the sample point from the top o f the slope p r o file in meters (m), I u 21 Table 2. Confidence of table station B elevation determined by d iffe re n t sightings. Known Route Unknown Elevation Confidence Aa----------------------------â–º B (921.4 m) 632 m 918.1 m Moderate ------------------------ B 630 m 918.7 m Moderate --------- B 700 m 918.7 m High A ------—â–º A 28 -*--------- B 700 m 918.7 m High 1,080 m 918.3 m Low A A ------—â–º A27 Ir X t A ------ CO Distance aElevation of highest h illto p from U.S.G .S. 15 minute Carter N.E. quadrangle. Table station A is located on the highest point of the same h i l l . See Figure 3 fo r location. is the length of the USLE un it plot (22.13 m), S is the slope gradient factor, C is the cropping management factor practices fa cto r. and P is the supporting A grid with perpendicular lines drawn 125 m apart was placed randomly over the study watershed to determine the locations at which soil losses were to be estimated. This grid was then shifted 62.5 m in both southerly and easterly directions to generate additional sample points. The sample grid points generated and th e ir labels are shown in Figure 5. Although th is approach produced 81 locations marked by grid lin e intersection points, soil losses from sheet and r i l l estimated at only 60 of these locations. were eliminated as follow s: Twenty-one erosion were points (locations) I) nine points located on or near h illto p s were eliminated because slope lengths and related USLE erosion estimates 22 15 G 15 F 13 G I I O I I C 10 O 10 N 12 Q I I H I I G 11 E 10 M 10 K 12 P 12 O 12 N 12 M 12 K 14 P 14 O 13 D 13 C I I B 14 N 14 M 14 L 10 P LEGEND: SAMPLE POINT WATERSHED BOUNDARY 200 Meters North Figure 5. Map showing location of randomly selected 125 m g rid points and labels used fo r USLE and WEE calculations. are zero in these circumstances (Wischmeier and Smith, 1978); 2) six points located immediately adjacent to and/or in channels were eliminated because the USLE does not apply to channel erosion (Wischmeier and Smith, 1978); and 3) six points located on concave slope segments with slope gradients less than one h a lf o f the cumulative average slope gradient from the drainage divide to th is location were eliminated based on the 23 assumption th a t these areas represent depositiona-1 zones (to which the USLE does not apply.) (Wilson,, lS86b.; JG riffin e t a ] . , 1 9 8 8 ). o f s o il losses a t the remaining 60 s ite s ,Determination involved estim ation o f USLE fa c to r values a t these lo c a tio n s . The r a i n f a l l -e r o s iv it y and s o il e r o d ib ilit y values used by the USDA S oil Conservation Service in th is watershed (R = 4-76.6 MJ mm ha'1 hr"1 yr"1 and K = .049 tonnes ha hr ha"1 MJ"1 mm"1 fo r both the s e rie s ) were used as w ell (M cPhail, R value includes the Rs adjustment Slope length (L) and slope Ethridge and Nishon personal communication, 1987). This fo r snowmelt. g ra d ie n t (S) fa c to r values were determined from th e topographic map constructed fo r th is p ro je c t. The slope length (x ) was measured by extending a lin e from each sample p o in t (g rid in te rs e c tio n ) up the slope to the drainage d iv id e , perpendicular to the contours. The remaining inputs required to taken from Wischmeier and Smith (1 9 78 ). The estim ated a t each po int using the method fo r estim ate L ( I and m) were slope g ra d ie n t fa c to r was ir r e g u la r slopes f i r s t proposed by Foster and Wischmeier (1 9 74 ). The crop management (C) fa c to rs were computed using the procedure of Wischmeier and Smith (1978, 2 8 -3 4 ). Table 3 summarizes methodology and in term ediate re s u lts o f computing an annual the average C fa c to r value fo r the w in te r wheat/summer fa llo w system employed on the Jackson farm. Cropstage C values were estim ated fo r each tim e period the two year w in te r wheat/summer fa llo w cycle by using a weighted of average o f C according to the amount o f p lan t cover and EI ( r a i n f a ll- e r o s iv it y ) in each p erio d . EI in the period was taken from Table 7 and the s o il loss r a tio s were taken from Table 5 o f Wischmeier and Smith (1 9 78 ). Time 24 Table 3. Computation o f Average Annual C Factor fo r T ille d S o ils ." Dateb A c t iv it y ' Year I 10/1 4/15 5/15 6/15 8/15 Cumulative % o f E Id EI in Period" 98 4 20 63 79 .06 .16 .43 .16 .25 Pl I 2 3 Harvest Soil Loss R atio’ .17 .14 .12 .07 .04 Cropstage C Value Crop Year Totals .0102 .0224 .0516 .0112 .0100 .1054 Year 2 4/15 C u ltiv a tio n I 5/1 C u ltiv a tio n 2 10/1 Rotation Totals 4 6 98 .23 .39 .02 .92 .0046 .3588 .3634 .4688 2.00 Average C Value Year"1 .2344 "Annual C value c a lcu lated fo r residue - 2500 kg ha"1. bDate by which cropstage growth or t i l l a g e operation is completed. cCropstage a b b reviatio n s, Pl = plan t crop; I = 10 % crop canopy cover; 2 = 50 % crop canopy cover; 3 = 75 % crop canopy cover. ^Cumulative EI fo r Great F a lls , MT from Wischmeier and Smith, 1978, Table 7. "El fo r period ending at date in f i r s t column, from Wischmeier and Smith, 1978, Table 7. ’Soil loss r a tio fo r period ending a t dates shown in f i r s t column, from Wischmeier and Smith, 1978, Table 5, 5-D. periods were distingu ished by a c t iv it y and cropstage w ith o f the Soil Conservation Service and the landowner. values were obtained by m u ltip ly in g the EI by the the assistance The s o il cropstage C loss Cropstage (C) values were then added together and divided by to obtain the average annual C value fo r the w inter r a t io . two years w h ea t/fa llo w system 25 o f 0.2344 reported in Table 3 . A value o f 0.053 Table 10 o f Wischmeier and Smith (1978) fo r the o f the watershed. no conservation watershed. was in te rp o la te d from u n t ille d , grassed parts The supporting p ra c tic es (P) fa c to r was ignored since p ra c tic es fo r water erosion are used in the study A ll USLE fa c to r estim ates were converted to m etric using the method o f Foster e t a l . (1981) and combined using the p o in t method o f G r if f in e t a l . (1 9 88 ). Wind Erosion Equation The m odified version o f the WEE proposed by Skidmore (1988) estim ate wind erosion a t d iffe r e n t points in a landscape was used in study. This version of the WEE s u b s titu tes a series of to th is equations (c a lc u la te d in stages) fo r the o rig in a l nomographs and can be w ritte n : El = r (2) where El is the f i r s t stage erosion estim ate in tonnes per hectare per year ( t ha'1 yr"1) , and I ' is the s o il c r e d ib i lit y fa c to r value in tonnes per hectare per year ( t ha"1 yr"1) . E2 = T K (3) where E2 is the second stage erosion estim ate in tonnes per hectare per year ( t ha"1 y r'1) , and K is the s o il ridg e roughness E3 = I'KC fa c to r . (4) where E3 is the th ir d stage erosion estim ate in tonnes per hectare per year ( t ha"1 yr"1) and C is the c lim a tic fa c to r. 26 E4 = (WF0-348 + E3°'348 - EZb34aF i87 (5) where E4 is the fo u rth stage erosion estim ate in tonnes per hectare per year ( t ha"1 yr"1) , and WF is a f i e l d length weighting parameter defined by equation 5a. WF = E 2 ( l.0 - 0.122(L/Lo)-°-383 e x p (-3 .3 3 L /L o )) (5a) where L is the s t r ip f i e l d length in meters (m) and Lo is a f ie ld length parameter defined by equation 5b. Lo = 1.56 x 106(E2)"1'26 exp(-0 .0 0 15 6 E2) (5b) ES = P1 E4 (P2) (6) where ES is the f in a l erosion estim ate in tonnes per hectare per year ha"1 y r'1) and P1 and P2 are v e g etative parameters defined by (t equations 6a and 6b. P1 = exp(- 0.759V - 4.74 x IO 2V2 + 2.95 x IQ-4V3) (6a) P2 = I + 8.93 x IO 2V + 8.51 x IO 3V2 - 1.5 x IO 5V3 (6b) where V is the v e g e tative fa c to r input in tonnes per hectare ( t ha"1 yr"1) defined by equation 7a. The v e g e ta tiv e residue was estimated using the equations o f Armbrust and Lyles (1985) reported in Skidmore (1988) as a function o f small grain e q u ivalen ts. which compute The equations the residue used fo r 27 w in te r wheat are: V = 0.2533 (SGe) 1363 where V is the v e g e tative kilograms per hectare (kg fa c to r ha:1) , input (7a) fo r equations 6a and 6b in and SGe is the small grain equivalent c a lc u la te d using equations 7b, 7c and 7d fo r standing stubble, fla tte n e d stubb le, and growing crop in fla tte n e d stubble, re s p e c tiv e ly . SGe= 4 .3 (Rws)"97 (7b) where SGe is the small grain e q u iva len t, and Rws is the above-ground weight o f the standing stubble residue in kilograms per hectare (kg SGe= 7.3 (Rwf) 0 dry ha'1) . (7c) where SGe is the small grain e q u iv a le n t, and Rwf is the above-ground dry weight o f the fla tte n e d (kg stubble residue in kilograms per hectare ha'1) . SGe= (8 .9 )-172(7 .3 )-82a(Rwg) (-9)(-172)+(-8)(-020) (7d) where SGe is the small grain e q u iva len t, and Rwg is the above-ground weight o f the crop growing in fla tte n e d residue in kilograms (kg The I , L, Lo and V fa c to r values were then determined fo r the points generated fo r the USLE a p p lic a tio n (see Figure 5) dry ha'1) . same sample and used in equations I through 7d to estim ate WEE s o il losses. Soil c r e d ib i lit y ( I ) was estimated by determining the non-erodible fra c tio n o f the surface s o il. The non-erodible fra c tio n ( i . e . , g re a te r than 0.84 mm in diam eter) was determined fo r 50 p a rtic le s s u r f ic ia l s o il 28 samples b y standard dry sieving using s ie v e . c ir c u la r f l a t screen , I n it i a l I values were c a lc u la te d fo r these samples using Table I o f Woodruff and Siddoway (1965) by s o il mapping u n it. a a 0.84 mm s e ries of d iffe re n c e s (see Table 18, Appendix C) and grouped Average values were computed fo r each d iffe re n c e of means te s ts was used between means were s t a t i s t i c a l l y to group and determine s ig n ific a n t if (Table 4 ) . S t a t is t ic a lly s ig n ific a n t d iffe re n c e s were found between the means fo r the (#38A) Ethridge Ethridge map u n its s ig n ific a n c e . map u n it (#388 and #3858) Nishon model was used here to series (#28) and other a t the 5 % and 10 % le v e ls The four means were used s o il u n its were not sampled losses. and the of to estim ate I values because the p ro p o rtio n a lly and because the wind erosion assess s p a tia l v a r i a b ili t y o f wind-generated s o il One po in t (2F) th a t f e l l w ith in the incised channel was excluded from the WEE analysis because wind does not erode th is area. A fte r the i n i t i a l computation o f I , the points located on windward slopes s h o rter than 152 m were adjusted using Figure I o f Woodruff and Siddoway (1 9 6 5 ). This diagram computes knoll c r e d ib i lit y ( I s) function o f slope grad ien t fo r two d if fe r e n t landscape p o s itio n s . c r e d ib i lit y ( I s) was expressed as a percentage (> 100 %) fo r points and I ' (th e fa c to r combining I and I s) was estim ated as a Knoll q u a lify in g by computing the product o f I and I s fo r these points (Chepil e t a l . , 1962; Woodruff and Siddoway, 1965). I ' was reduced by 50 % u n til Following periods o f c u ltiv a tio n , seeding follow ing the advice o f Bondy and associates (1 9 8 0 ). K fa c to r values were obtained from the equations o f W illiam s e t a l. (1984) in Skidmore (1 9 88 ). K values were estim ated fo r each 29 Table 4. S oil E r o d ib ilit y by Soil Mapping U n it.* Number o f Samples Average I Valuec 28 Nishon(28) 3 280.9 - E th r idge(38A) 7 180.2 LO O - E th r idge(385B) 27 253.3 .# .05 Ethridge(38B) 13 234.2 # .10 Soi I sb 38A 385B 38B # d iffe r e n c e o f means te s t re s u lts in d ic a tin g le v e l o f s ig n ific a n c e o f d iffe re n c e o f means in proportions. The # symbol in d ic a te s no s ig n ific a n t d iffe re n c e . bSoil numbers are s o il mapping un its used by the s o il survey team, SCS, Fort Benton (see also Figure 3 ) . 0See Appendix C fo r ta b le showing map i d e n t i f i cation codes, s o il mapping u n it, percentages o f s o il aggregates > 0.84 mm and s o il c r e d ib ilit y ( I ) index values fo r each o f the 50 s u r f ic ia l s o il samples. cropstage period or t i l l a g e op eration. A ridg e height o f 42 mm and a ridg e spacing o f 356 mm was used to estim ate K fo r periods o f ( .9 0 ) ; w h ile a ridge height o f 102 mm and a ridge used fo r periods immediately follow ing seeding c u ltiv a tio n spacing o f 356 mm was ( .4 9 ) ; and a ridge height o f 25 mm and a ridg e spacing o f 610 mm was used fo r the periods follow ing c u ltiv a tio n w ith a to o lb a r and duckfoot implements (.6 6 ) At th is stage, each p o in t's lo c a tio n w ith in a s o il mapping u n it recorded (see Figure 3 and Table 18, Appendix C). average I ' ( E l, equation # 2 ), (see Table 5 ). Using the was appropriate E2 (equation #3) was computed fo r each period o f the w in te r wheat/summer fa llo w system as the product o f I ' and K. Next, the annual C value, 0 .9 0 , was m u ltip lie d by E2 to (equation # 4 ). compute E3 This C value is used by the SCS to represent the v ic in it y 30 Estimates o f K Factor, Vegetation Weight and Erosive Wind Energy fo r Cropstage Periods and T illa g e O perations, Jackson Farm, C a rte r, MT. A c tiv ity * 8 /1 5 -1 0 /1 Harvest 1 0 /1 -4 /1 Winter 4 /1 -5 /1 K EWE in Periodb - 3035 10270 .49 2430 8270 .66 .74 1190 2110 .90 .09 950 1765 .90 .17 .30d O CXJ Summer Fallow Vegetation Small Grain Weight Equivalent kg ha'1 Seeding .30d 760 1705 .49 .09 1 1 /1 -4 /1 Winter 180 520 .49 .65 390 980 .49 .09 1100 2310 .49 .12 4 /1 -5 /1 5 /1 -8 /1 5 3 - 4 weeks growth Mature growth - °n 1 0 /1 -1 1 /1 ro 5 /1 -1 0 /1 C u ltiv a tio n (Z ) Loss of Residue % O CXJ Period O CJl Table 5. “A c tiv ity or cropstage growth completed during period in Column I . ^Proportion o f erosive wind energy. cLoss o f dead residue fo r summer and w in te r due to biomass decomposition. dEstimates o f residue loss due to t i l l a g e operations from National Agronomy Manual, Montana Supplement, Table I , (USDA-SCS, 1988); 60 kg ha"1 assumed to remain a f t e r deep furrow d r i l l seedings (Nadwornick, personal communication, 1990). o f C a rte r, MT and was in te rp o la te d from a statewide contour map o f C values (M cPhail, personal communication, 1987). The next stage o f the computation fa c to rs , L and Lo. incorporated the f ie ld length The f ie ld length o f the study area was a function the unsheltered distance ( i . e . , distance in which wind b a r r ie r of erosion 31 control measures were not used) o f the f ie ld s tr ip s . On the th is was simply the width o f the f ie ld s trip s from the Jackson farm western edge to the lo c a tio n o f the sample po int along the wind erosion d ire ctio n to the fa c t th a t no b a r r ie r and/or other wind control p ractices were employed. F a lls , Data c h a ra c te rizin g wind MT from the Montana Manual (SCS, 1988) and WEE used to determine the ,due d ire c tio n and in te n s itie s a t Great Supplement o f the 1988 National Agronomy handbook (Skidmore and Woodruff, 1968) were predominant wind erosion d ire c tio n . The minimum windspeed necessary fo r wind erosion is approximately 29 km hr"1 (8 m sec"1) d ire c tio n ( Bondy e t a l., 1980). The eval uation o f wind and in te n s ity only considered winds above th is threshold and revealed two predominant wind erosion d ire c tio n s o f SW and WSW. d ire c tio n applied to the periods o f January to December. The SW d ire c tio n applied to the period o f A p ril to The WSW March and October to September. The unsheltered distance fo r each po int was obtained by measuring lin e s drawn p a r a lle l to th is predominant wind d ire c tio n (WSW) from the western edge o f the f i e l d s t r ip to the sample p o in t. This value was used fo r January to March and also fo r October to December. distance along the SW d ire c tio n was estim ated by WSW by the cosine o f 22.5° (th e angle o f the d ir e c tio n s ). This L value was applied These L values were then d iv id in g the L along d iffe re n c e between the two between A p ril used to compute the fa c to rs WF and Lo (equations #5a and 5 b ). The unsheltered f ie ld and September. length weighting E4 was then c a lc u la te d fo r each period o f the cycle (equation # 5 ). The fin a l fa c to r o f the WEE, the v e g etative cover fa c to r ( V ) , was needed to complete the computation o f erosion estim ates. The V fa c to r 32 was a function o f cropstage period, t i l l a g e operation and stubble above-ground weight converted to a small (Table 5 ). grain The V fa c to r equivalen t fo r each using equations residue period 7b, was 7c or 7d. Equation 7a was then used to determine the fa c to r value in p u t, V, which was used in equations #6a and #6b to estim ate s o il losses a t each sample p o in t. Standing stubble weight was estim ated by m u ltip ly in g the average y ie ld fo r the Jackson Farm, 25 bushels acre'1 (Jackson, personal communication, 1987; M cPhail, SCS, personal communication, 1990), by HO pounds bushel"1 to obtain an above-ground weight estim ate o f residue o f 3025 kg ha'1 (USDA-SCS, 1988). points th a t f e l l (John Siddoway, The small grain equivalen t o f the s ix in the grassed area was estimated to be 5600 SCS-Fort Benton, MT, personal kg ha'1 communication, 1990). S ix ty kilograms per hectare o f dead residue was assumed to remain a f t e r seeding and fo u r weeks crop growth, weighing approxim ately 120 kg ha'1, was assumed before dormancy in w in te r, fo r a to ta l o f kg ha'1 (Nadwornick, SCS S tate Agronomist, Bozeman, approximately 180 MT, personal communication, 1990). A s o il loss estim ate fo r each period o f the c ro p /fa llo w cycle was c a lc u la te d by using ES (equation # 6 ). The annual erosion ra te estimates fo r each period were then m u ltip lie d by a weighted average o f wind energy (see Table I fo r monthly to t a ls ) estim ate period wind erosion ra te s . in the same period The period to ta ls were then to estim ate annual wind erosion rates fo r the crop and fa llo w the c y c le . erosive Table 6 shows these computations fo r sample po in t to added years o f ID. 33 Table 6. Sample WEE Computation fo r Sample Point 2M.a Period K Residue SGe kg ha"1 Fallow year 8/15-10/1 10/1-4/1 4 /1 -5 /r 5/1-10/1° Crop Year 10/1-11/1 11/1-4/1 4 /1- 5 /1 5/1-8/15 V EWEb Mg ha"1 ES Period Erosion t Ii a"1 yr"1 .49 .66 .90 .90 3035 2430 1190 950 10270 8270 2110 1765 74.40 55.39 8.60 6.74 .05 .74 .09 .17 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.0 .49 .49 .49 .49 759 180 390 1100 1705 520 985 2310 6.43 1.27 3.04 9.78 .09 .65 .09 .12 0.1 5.4 2.0 0.0 0.0 3.5 0.2 0.0 Total Cycle 2.0 3.7 Average 1.0 1.8 aWEE computed using I ' of 234.2 t ha1, C of .90, I. of 85 m. Proportion of cumulative erosive wind energy (EWE) in period -in Column cPeriods in which soil c re d ib ility ( I ') plowing ( Bondy et aI . 1980). is reduced by I50 % following Cesium-137 Erosion Estimates Cesium-137 Sample Site Selection Slopes were paced in the fie ld to determine the location of 137Cs transects. The highest point on the transect was chosen sample s ite and the lowest point was chosen as a as a h illto p channel sample s ite . Following the method of Brown et a l. (1981b) slopes were segmented by pacing u p h ill, following the path of greatest slope, dividing slope lengths in to th ird s with footslope and midslope sample sites chosen at .34 distances one th ir d and two th ird s up 6 il lu s t r a t e s how t h is other h a lf selected s ta rtin g a t the same sample s ite s channel) (two were the slope, re s p e c tiv e ly . Figure s e le c tio n was made fo r each tra n s e c t, with the by using the channel s it e . h illt o p s , d e lin e a ted in same method on the opposite slope A to ta l o f s ix tran sects w ith seven two midslopes, the fie ld . two Due to footslopes the high and one cost of la b o ra to ry a n a ly s is , only a portion o f the samples were analyzed fo r 137Cs a c t iv it y . The samples c o lle c te d during the f i r s t sample c o lle c tio n were analyzed f i r s t . U n fo rtu n a te ly , enough funds were not a v a ila b le to analyze the samples c o lle c te d during the second c o lle c tio n p e rio d , one year l a t e r . T h erefore, the 137Cs re s u lts are biased towards the steeper sloping, eastern po rtion o f the study area, where the f i r s t samples were c o lle c te d (see Figure 7 ). Figure 7 also in d ic a te s the s ite s th a t were sampled in the incised channel leading from the drained lake to the pond (n = 3 ) , bottom (n = 4 ), and along watershed (n = 4 ) . (uneroded) control a fence lin e 250 meters in the pond from the study The samples taken along the fe n c e lin e represent s ite s because the fence acted as a b a r r ie r to wind erosion and a t i l l a g e berm fe n c e lin e . a to ta l O v e ra ll, prevented overland flow from crossing th is o f 53 s ite s was used and s o il samples representing 261 increments were c o lle c te d , although only 83 increments from 25 s ite s were analyzed fo r 137Cs a c t iv it y . Following the method used by Brown e t a l . (1981b), a strateg y to defin e areas o f landscape u n its , termed topographic p o s itio n s , was used. These areas were computed using slope p r o file s which were divided in to segments by assigning boundaries to points halfway between each 137Cs 35 H ILLTO P SAM PLE M ID S LO P E S A M PLE FO O TSLO PE SAM PLE CHANNEL SAM PLE Figure 6. Diagram showing topographic positions. sample lo ca tio n . The p ro file s were drawn 2 cm (50 m) apart on a copy of the topographic map and areas fo r each topographic position by connecting each boundary point to it s counterpart on p ro file s . were defined adjacent slope These areas were calculated by adding up 2.5 m2 squares w ithin the boundary lin e s using metric grid paper. The areas corresponding to each o f the topographic positions are shown in Figure 8. Cesium-137 Sample Collection The 137Cs samples were gathered from I m3 hand excavated so il p its . Care was taken to preserve the soil pedon in a natural, uncompacted state. A fte r the s o il p it was dug, the 137Cs samples were removed in a la te ra l method to minimize the effects of compaction during sampling. A 5.1-cm diameter PVC tube with a cap affixed to one end to prevent loss was used to c o lle c t the samples. Holes were d r ille d in to the cap to allow a ir to 36 C O D E KEY — 2T1 North ----4 T I 5 ----- 5 T I QTI q 7—7T1/7T2 8 ———- — QT2 9 -—— 5T 2/ / ----- ------ 200 Meters L O C A T IO N MAP Control sites â–¡â–¡ LEGEND; EPHEMERAL CHANNELS Legend Cs Sample Site Not To Scalel WATERSHED BOUNDARY Q Buildings -H- Fence Figure 7. Study area map showing locations o f 137Cs samples w ithin study watershed and locations o f control samples on the Jackson Farm. escape so i t would not displace potential sample volume. The tube was driven h o rizo n ta lly in to the wall of the so il p it a m allet. The tube was 7.6 cm long and was completely f i l l e d with sample to obtain standard sample volumes. p it wall w ith a large kn ife . with each The tube was removed from the Care was taken to preserve the in te g rity o f the sample by preventing the entry of topsoil and organic matter into the tube during extra ction . The samples were then placed in p la s tic lined 37 Figure 8. Areas corresponding to each of the topographic positions used fo r 137Cs extrapolation. sample bags and transported to Bozeman. Cesium-137 Sample Increments The samples were taken in increments whose depths corresponded to the width of the sampling device, 5.1 cm, or m ultiples o f i t , depending on t illa g e practices at d iffe re n t sampling s ite s . Plowed sites were assumed to have uniform 137Cs a c tiv ity in th e ir plow la yer, or Ap horizons 38 (Brown e t a l . , 1981b; Campbell e t a 15 cm (Jackson, personal communication, 1987). l 1982) . communication, Below these increments, samples bottom and the control s ite s ) in 5.1 cm increments. was tre a te d as were gathered as 30 cm fo r the in the incised were sampled from the top A minimum o f f iv e samples were personal successive 5.1 depth o f about The n o n -tille d s ite s ( i . e . , was set as M cPhail, From th a t point i t s in g le increments o f 5 .1 cm to a to ta l plowed s ite s . 1987; This sample was c o lle c te d as three cm deep increments and then bulked. a homogeneous sample. The plow la y e r channel, pond o f the s o il p it c o lle c te d fo r each n o n -tille d s it e . Cesium-137 Sample Preparation In the la b , the 137Cs samples were dried at 105° C fo r a minimum of 72 hours, and then m echanically ground and sieved through a 2 mm screen. Each sample was weighed and a r i f f l e re p re s e n ta tiv e 100 g sample. to standardize the s p l i t t e r was used to This uniform 100 g sample s ize counting times (D r. Ingevar Larsen, remove a was required Environmental Sciences D iv is io n , Oak Ridge National Laboratory, Oak Ridge, TM, personal communication, 1988). The samples were then tra n s fe rre d to dry s o il sample bags, closed and wrapped fo r shipment. Cesium-137 Laboratory Analysis The samples were analyzed fo r 137Cs concentrations Larsen a t by Dr. Ingevar Oak Ridge National Laboratory w ith a Canberra lith iu m d r ifte d germanium (G e (L i)) gamma ray detecto r coupled to a Nuclear Data 4096 channel data analyzer system (C utshall and Larsen, 1980; CutshalI , 1981). The procedure required a counting tim e o f No. 6700 Larsen and 60 to 100 39 minutes per sample. 137Cs a c t iv it y , c u rie E ig h ty -th ree sample increments were expressed in pi cocuries analyzed fo r per gram o f s o il (pCi g"1) . is a measure o f r a d io a c tiv ity such th a t one c u rie A is 3 .7 x IO10 d is in te g ra tio n s per second per gram o f radium and one pi cocurie is IO*12 curies (C utshall concentrations and Larsen, are reported 1980; in A rnalds, un its o f pCi 1984). g"1. The 137Cs mass The 137Cs data were transformed from pCi g'1 to pCi cm"3 by m u ltip ly in g mass concentrations by average bulk d e n s itie s (obtained from a weighted average bulk density by depth o f the horizon in which the 137Cs sample a c t iv it y was converted to areal (pCi cm"3) by the depth o f was c o lle c te d ). This 137Cs concentration by m u ltip ly in g the a c t iv it y the sample la y e r (cm). . Bulk Density Sample C o llec tio n Bulk density measurements concentration c a lc u la tio n s . were required fo r the 137Cs areal Bulk density samples were taken a t the tim e as the 137Cs samples using s o il sampling tin s o f known volume. s o il horizon was sampled in each s o il p it. same Every The g e n e ra lly dry s o il samples were d rie d and weighed to determine dry bulk d e n s ity . It possible to determine water content or moist bulk d ensity due to the lack o f control o f evaporation in the storage area a t the f i e l d long time between sampling c o lle c tio n (two 4-week s ite periods, was not and the one year a p a r t). Method o f Areal Cesium-137 Analysis Areal analysis r e lie s on the fa c t th a t 137Cs loss has been found be proportional to s o il loss and to vary a t depth ( Rogowski and 1970; McHenry e t a l . , 1973; McHenry and R itc h ie , 1977). The to Tamura, method used 40 fo r areal analysis o f 137Cs in th is study was taken from associates (,19.83). This method computes s o il loss using De Jong and the follow ing equations: 137Cs loss = [0 .9 5 X - Y] [0 .9 5 X ] 1 (8) where 137Cs loss is the amount o f 137Cs lo s t (p e rc e n t), X is average present a t control s it e (pCi cm'2) , Y is average 137Cs present a t 137Cs eroded s ite s (pCi cm"2) ; and s o il loss = 137Cs loss x d x BD (9) where s o il loss is the amount o f s o il (g cm'2) , d is the thickness (cm) the la y e r in which 137Cs is present, and BD is the average bulk cm"3) o f the la y e r in which 137Cs is present (De Jong e t a l., of density (g 1983). A value o f 95 % o f the 137Cs areal concentrations was used to make the c a lc u la tio n s because previous studies had found th a t removal o f 137Cs by crops, animals and snow d r if t in g et a l., 1982; Arnalds, 1984). is approximately 5% (De Jong The concentrations o f 137Cs were cm'2 and were generated by m u ltip ly in g the la b o ra to ry (pCi g'1) by the bulk density (g cm'3) and the Jong e t a l . , expressed in pCi concentration data increment depth (cm) (De 1983). In t i l l e d s ite s , gains o f 137Cs by deposition were ca lc u la te d 137Cs d is tr ib u tio n a t depth. The depth to which 137Cs was areas was estim ated by subtracting 15 cm (plow la y e r maximum depth in which 137Cs was present. The using present in these depth) from the thickness o f s o il deposited by s o il r e d is tr ib u tio n was found by: D1 = D2 - 15 cm (10) 41 where D1 is the thickness (cm) o f s o il deposited, D2 is the maximum depth (cm) to which 13^Cs is present, and 15 cm is the depth o f c u ltiv a tio n . S oil gain was estim ated as s o il gain =^= D1 x BD where BD is s o il bulk density (g cm"3) , cm"2 (De Jong e t a l., 1983). (11) and s o il gain is expressed in g For u n tille d s ite s , gains o f 137Cs were estim ated by m u ltip ly in g the maximum depth o f 137Cs a c t iv it y detected the s it e 's average bulk d e n s ity . Soil gains were only computed i f s it e 's to ta l areal concentration was g re a te r than th a t o f the by the average o f the control s ite s (11.10 pCi cm"2) . Erosion and Deposition Rate and Mass Balance Estimation Method Annual to ta l s o il erosion and deposition ra te s were determined by d iv id in g lo s s /g a in by the time th a t 137Cs has been deposited in watershed (since 1954) except fo r the pond s ite s . Deposition a t s ite s was computed using the time period since dam construction 1973 to 1987, or 14 y e a rs ). and areas of landscape Averages based upon the areal un its (topographic determine the erosion or deposition ra te p o s itio n . topographic yr"1. These values were then p o sitio n s) the those (i.e ., concentrations were used to associated w ith each topographic m u ltip lie d by the area of each p o sitio n to obtain mass erosion and deposition rates in Mg H illt o p , midslope, footslope and incised channel erosion rates were combined to estim ate a net mass erosion ra te and t i l l e d channel and pond deposition ra te s were combined to obtain a net deposition r a te . 42 CHAPTER 3 RESULTS USLE Erosion Estimates USLE so il loss estimates were computed at 81 points on a randomly located 125 m grid using constant R, K and P values and s p a tia lly variable C, L and S values in th is study. Most of the v a r ia b ility in the erosion estimates was due to variations in the topographic factors, L and S, because one C was used with the exception of six points which f e ll w ithin the grassed margin of the pond area. The fin a l 60-sample set was determined using the method of G riffin et a l. (1988) and Wilson (1986b) to eliminate environments. sample grid points near h illto p s and in depositional The USLE factor values were combined using the point method of G riffin et a l. (1988). The topographic factor and soil loss estimates are summarized in Table 7. The slope gradients and slope lengths varied considerably from point to point and account fo r most of the spread in so il loss estimates. Watershed so il losses averaged 4.5 t ha'1 yr"1. A ll 81 are reproduced in Figure 9 as a frequency histogram points where erosion equals 0.0 t ha"1 yr"1. This point estimates (including the 21 diagram indicates how the d is trib u tio n of point estimates is p o sitive ly skewed with large numbers of estimates concentrated in the 0 - 6 t ha"1 yr"1 range and smaller 43 Table 7. USLE Soil Loss Estimates.® Slope Length (m) Slope Gradient (%) Soil Loss ( t ha"1 yr"1) Minimum Maximum Average Standard Deviation 19.0 0.3 0.0 373.0 16.0 14.9 92.2 4.2 4.5 71.3 3.0 3.6 aUSLE estimates using sample generated from a random 125 m g rid (n = 81). EROSION ESTIMATES (T /H A /Y R ) Figure 9. (n = 81). D is trib u tio n of USLE so il loss rates fo r Carter watershed 44 numbers spread throughout the 7 - 18 t ha"1 yr"1 range. s ta tis tic a l analysis was conducted to determine of the watershed USLE so il loss average. Using A nonparametric the level of confidence a sample size of 60, the watershed so il loss average of 4.5 t ha"1 yr"1 and standard deviation of 3.6 t ha"1 yr"1, and a desired 2.13, corresponding to an formation rates commonly precision of + 1.0 t ha"1 yr"1, Z equaled area under a normal curve o f 0.4834. Soil assumed fo r so ils developed in glacial t i l l s exceed 1.0 t ha"1 yr"1 (Dr. Gerald Nielsen, Department of Plant and Soil Science, Montana State University, Bozeman, MT, personal 1990). communication, This analysis indicates that the sample mean f e ll w ithin + 1.0 t ha"1 yr"1 of the true mean with a confidence level of 96.7 % (level of significance of 0.033). â– The spatial v a r ia b ility of USLE soil loss estimates is depicted by Figure 10. The squares, triangles and circle s denote points that represent h illto p s , ephemeral channels and concave slopes, These points were excluded from the original 81 point erosion was estimated to be 0.0 t ha"1 yr"1 fo r them. diagram shows an area of high erosion due to of the watershed. respectively. sample set and Examination of th is water in the eastern h a lf This f lu v ia lIy erosive area corresponds to steeper sloping areas of the eastern h a lf of the watershed (compare Figures 2 and 10) . WEE Erosion Estimates Eighty points were sampled fo r the WEE calculations. calculations were based upon s p a tia lly variable I ' variable I ' , K and V, and constant C factor values. The WEE and L, temporally The I ' and L factor 45 LEG END: SAMPLE POINT WATERSHED BOUNDARY 200 Meters North Figure 10. Spatial v a r ia b ility of USLE s o il loss rates ( t ha"1 yr"1) fo r Carter watershed. Squares, trian gles and c irc le s represent points where the USLE estimated erosion as 0.0 t ha"1 y r'1. Shaded area represents area o f high water erosion. values were obtained using topographic and so ils maps. L was to re fle c t a seasonal s h ift in the wind erosion d ire c tio n . Bondy et a l . (1980), K and V factors were estimated using National Agronomy Manual and the equations reported in The C value was held constant over time but erosive I adjusted Sim ilar to the 1988 Skidmore (1988). wind energy (EWE) 46 values, which varied by period, were incorporated. estim ated fo r each cropstage and tilla g e period Wind erosion was of crop/summer fa llo w system fo r a ll 80 p o in ts . A two year was computed based upon period estim ates o f s o il m u ltip ly in g the period s o il loss estim ate by the the two year weighted average lo ss , obtained by cumulative erosive wind energy in th e p e rio d . Table 8 summarizes standard deviatio n s estim ates. a ll fo r the I' watershed minima, and L fa c to rs maxima, averages and watershed WEE s o il The watershed WEE average was 4 .5 t ha"1 yr"1. Figure 11 80 s o il loss estim ates as a frequency histogram. This reveals a p o s itiv e ly skewed d is tr ib u tio n , influenced by the o f estim ates th a t were 0 .0 t ha"1 yr"1. was used to determine the le v e l average. A nonparametric yr"1) , and standard d e v ia tio n (4 .6 t ha"1 y r'1) and + 1.0 t ha'1 yr"1, Z equaled 0 .6 5 , corresponding curve of 0.2422. in te rs e c tio n s Therefore, w ith o f a 125 m g rid , a sample loss shows diagram la rg e number s t a t is t ic a l te s t o f confidence o f the Using a sample s ize o f 80, the watershed and WEE watershed average (4 .5 t ha"1 a desired precision o f to an area under a normal s ize the watershed WEE determined by the average estimates erosion fo r the watershed w ith in ± 1.0 t ha"1 y r'1 o f the tru e mean with a confidence le v e l o f o f 0 .5 1 6 ). 48.4 % (le v e l o f s ig n ific a n c e This low le v e l o f confidence occurred because o f the high standard d eviatio n th a t is due to the in fluence estim ate (See Figure 1 1 ). o f the o u tly in g p o in t's (2P) extremely high This po int had a high wind erosion estim ate due to i t s high s o il c r e d ib i lit y caused by a high I s adjustment o f 671 %. 47 Table 8. WEE Soil Loss Estimates.® Minimum Maximum I ' Factor Values ( t ha"1 yr"1) Field Length L (m) Soil Loss ( t ha'1 yr"1) 180.2 0 0 670.3 155 34.8 Average 256.0 41 4.5 Standard Deviation 56.0 43 4.6 aWEE so il loss rate estimates using a sample set generated from a random 125 m g rid (n = 80). EROSION ESTIMATES (T /H A /Y R ) Figure 11. D is trib u tio n of so il loss rates fo r Carter watershed WEE point method and 125 m g rid (n = 80). using 48 The spatial v a r ia b ility of the WEE results is depicted in 12. Figure The highest estimated soil losses were located on or near h illto p s (compare Figures 2 and 12). There was also a group of high estimates the central portion, near the old lakebed. Another trend is located in in the grassed area near the pond, where the high small grain equivalent assigned to crested wheatgrass resulted in estimates of 0.0 t ha"1 yr"1. USLE and WEE Erosion Rate Estimates Combined Though not customarily done, USLE and WEE results were combined to determine i f any spatial patterns existed in the estimated soil losses. This was done to permit a rough comparison to the 137Cs results, which include erosion from both water and wind, even though the models do not estimate g u lly erosion. This allowed fo r the id e n tific a tio n of highly erosive areas and th e ir spatial relationship to the location of the 137Cs sample s ite s. erosion Combining the USLE and WEE watershed averages, a predicted rate of. 9.0 t ha"1 yr"1, based on a range from 0 to 36.5 t ha"1 yr"1, was computed. The 81 point estimates were combined and depicted as a frequency histogram (Figure 13). The spatial v a ria b ility of the USLE/WEE combined results is shown in Figure 14. The diagram shows the combined resu lt from the two models fo r each grid point, with the squares, triangles and c irc le s the points were water erosion equaled zero fo r the USLE. representing The point that was excluded from the WEE analysis was assumed to have erosion equal to 0.0 t ha"1 yr"1 due to its location in the bottom of the 4-feet deep incised channel (excluding i t from both the USLE and WEE). These data indicate very high so il loss rates along the eastern margin because of 49 LEGEND: SAMPLE POINT WATERSHED BOUNDARY 200 Meters North Figure 12. Spatial v a r ia b ility of so il loss rates ( t ha"1 yr"1) fo r Carter watershed using WEE point method and 125 m random g rid . Shaded areas represent areas of high wind erosion. Excluded point (40) not shown. both high flu v ia l erosion rates and high rates fo r h illto p s . This re s u lt shows how the 137Cs transect located in high so il loss areas. wind erosion on the sample sites were 50 EROSION ESTIMATES (T/H A /Y R ) Figure 13. D istrib u tio n of to ta l soil loss rates fo r Carter watershed using USLE/WEE point methods and 125 m grid (n = 81). Cesium-137 Results The results of the laboratory analysis fo r 137Cs a c tiv ity are depicted in Figure 15. The 137Cs areal a c tiv ity at each sample s ite depth and is reproduced in Figure 15 as bar graphs increment's 137Cs a c tiv ity . The 137Cs a c tiv itie s calculated by m ultiplying the mass concentrations varied with representing each (pCi cm"3) shown were of 137Cs a c tiv ity (pCi g"1 by the increment's bulk density (g cm3). Note that 3 of 4 pond sites and sites 3C and SC did not reach a depth of zero detection, causing estimates of deposition rates to be underestimated. The to ta l sample 51 1 3 .0 6 a0 4 .0 8 .2 1 4 .2 1 6 .4 9 .8 LEGEND: SAMPLE POINT 200 WATERSHED Meters North BOUNDARY Figure 14. Spatial v a r ia b ility of so il loss rates fo r Carter watershed using USLE and WEE point methods. Squares, tria n g le s and c irc le s represent points where the USLE estimated erosion as 0.0 t ha'1 yr'1. Shaded areas represent areas of high net erosion. depth was inadequate fo r these samples. Site concentration (pCi cm'2) was lower than the control even though it s depth p ro file is s im ila r depositional s ite , i t was treated as an in 3C's to ta l areal site s average, so, shape to eroding s ite . that of a The changes in a c tiv ity fo r the con trol, pond and SC site s are important to notice as they re fle c t much greater 137Cs a c tiv ity than the transects and three of 52 ACTIVITY (pCI cm ~3) DEPTH (cm) 137C3 O LEGEND; 15 T TT = Trace detection of 7T1 / 7T2 )ND = No detection of 137 Cs 13^Cs 30 J Figure 15. Site areal 137Cs a c tiv itie s (pCi cm"3) fo r Carter watershed. 53 the fo u r incised channel s ite s ( s it e SC is a depo sition al s ite and th e re fo re an exception).. The diagram o f s it e 7P stands out in comparison to the other pond s ite s . It s p r o f ile is ty p ic a l o f an eroding s it e , which is unusual because i t was located on the margin o f the pond. In g eneral, however, the diagrams fo r d iffe r e n t landscape po sition s are s im ila r to those in the general model reproduced in Figure I . The 137Cs sample s ite s were grouped and averaged, w ith the re s u lts shown in Figure 16. This graphical illu s t r a t io n shows the re la tio n s h ip o f the topographic po sition s to the control samples by d iv id in g the in to two sections. graph Those topographic po sition s with averages to the l e f t o f the average areal concentration o f the control s ite s , 11.10 pCi cm"2, represent erosional s ite s , w hile those topographic po sition s w ith averages to the r ig h t o f the lin e represent depositional s ite s . concentration is marked on the bar by a t ic k mark in The average areal the middle portion o f the bar, w ith the whole bar representing the range o f values. im portant to po int out th a t several bars range o f d is trib u tio n s I t is overlap, in d ic a tin g the wide and great s p a tia l v a r i a b ili t y o f the 137Cs in sediments. The equation p r e c ip ita tio n based o f Arnolds on a strong c o rre la tio n (1984) and Arnolds e t a l . predicted the amount o f 137Cs deposition a t C a rte r. concentration, of 137Cs and annual (1989) was used to This p re d icts an areal 7 .9 pCi cm'2, 71 % o f the amount found in the control average o f th is study, 11.1 pCi cm"2. Table 9 shows the re s u lts c a lc u la tio n s fo r each s it e . of the erosion and deposition ra te These ra te s characterized the s o il erosion or deposition r e la t iv e to the control s ite s . The fo u rth column o f the 54 DEPOSITION EROSION INCISED CHANNELS • â– ------------ H MIDSLOPES 1— I------------ 1 HILLTOPS .------------- 1------------------ FOOTSLOPES , 1 I I CONTROL TILLED CHANNELS .------------------ 1-------------------------------- â– POND __________I_____________ , I ( ) 12 6 pCi 18 24 cm Figure 16. Range and average 137Cs areal concentrations by topographic po sitio n . table li s t s the 137Cs depletion (a percent) fo r each was made only i f a s ite 's to ta l areal control s ite average. s ite . This estimate concentration was less than the The average 137Cs a c tiv itie s used fo r the control s ite 's plow (0 to 15 cm) and 16 to 20 cm layers were 11.03 and 0.07 pCi cm"2, respectively, or a to ta l increments collected from the 0.07 pCi cm"2 of the comparison of 11.10 pCi cm"2. Some o f the 16 to 20 cm transect sites had s lig h tly greater 137Cs than control s ite average. This would cause the to re s u lt in a negative 137Cs depletion value, although these resu lts were disregarded follow ing the advice of De Jong et a l. (1983). Total s o il loss or is gain shown in the f if th and sixth columns. 55 Table 9. Cesium-137 Erosion and Deposition Rate Estim ates. S ite # Topographic Position Bulk Cesium Density D epletion6 g cm"3 Total Total Soil (o r) Soil Loss Gain g cm"2 S ite Rate3 S ite E rror (+) Mg ha"1 yr"1 IT l H illto p 7T1/7T2 1T2 1.27 1.23 1.25 0.08 0.76 0.42 1.61 13.44 7.67 --— -- 4 .9 40.7 23.3 0 .6 3 .2 4.1 2T1 6T1 2T2 6T2 Midslope 1.32 1.27 1.19 1.32 0.40 0.61 0.36 0.65 7.78 10.92 5.77 12.56 —-— -- 23.6 33.1 17.5 38.1 3 .6 6.6 2.1 7.4 3T1 5T1 3T2 5T2 Footslope 1.32 1.37 1.30 1.30 0.17 0.32 0.22 3.29 6.11 3 .98 ——— 12.70 10.0 18.5 12.1 -3 8 .5 1.1 2.5 1.8 3.4 13.73 12.50 16.35 -4 1 .6 -3 7 .9 -4 9 .6 3.3 4.2 3 .7 51.5 34.2 26.5 9 .0 5.3 6.2 -232.1 -2 5 7.5 -2 3 7.5 32.2 24.5 21.2 41.1 5.9 Il Il Il Il Il M Il Il - - " 4T1 T ille d Channel 4T2 SC Il IC Il U n tille d Channel Il 2C 3C IP 4P 6P 7P Il Pond Bottom H Il Pond Margin 1.39 1.35 1.23 1.40 1.40 1.40 1.30 1.44 1.33 1.40 - — - - - - 0.81 0.54 0.42 —— — 17.00 11.27 8.74 — - - — — — 0.82 — — - - 5.76 — — - — 32.50 36.00 33.25 “P o s itiv e s ite rates in d ic a te erosion and negative rates d epo sition . bCesium-137 loss as proportion as computed using equation # 8. in d ic a te The s ite erosion ra te s are shown in the next column. shows the counting e rro r. The la s t column For erosional s ite s the e rro r was expressed as a proportion o f the 137Cs depletio n and then put loss equations (#8 and #9) shown in Chapter 2. For the e rro r was c a lc u la te d as a s o il gain in equation T n itia T ly in to the s o il depositional s ite s , 11 and then expressed as a proportion o f the s it e 's s o il gain. The re s u lts in Table 9 show several in te re s tin g in sig h ts about the s p a tia l v a r i a b ili t y o f 137Cs a c t iv it y and erosion/deposition ra te s . tran s ec t s it e w ith the g re a tes t erosion located on a h illt o p The was s ite 7T1 (4 0 .7 t ha"1 y f 1) , in the steeper eastern portion o f the However, two other h illt o p s ite s have suffered much low er - - 4 .9 t ha"1 yr"1 fo r IT l and 23.3 t ha"1 yr"1 fo r 1T2 watershed. erosion rates - - in d ic a tin g th a t as a group the h illto p s vary g re a tly in ra te o f erosion. The midslopes have erosion ra te s th a t are less v a ria b le . midslopes th a t were on the north ends o f the lower erosion ra te s (2 3 .6 and 17.5 t ends, 6T1 and 6T2, (33.1 and note th a t the midslope same tra n s e c t d ir e c t ly The two tra n s e c ts , 2T1 and 2T2, had ha'1 yr"1) , than those on the southern 38.1 t ha"1 yr"1) . I t is also in te re s tin g to w ith the highest erosion ra te was located on the u p h ill from a foots!ope s it e (5T2) th a t experienced 137Cs enrichment. This s it e w hile the showed a net gain o f 137Cs a c t iv it y , in d ic a tin g deposition, other foots!opes (3T1, 5T1 and 3T2) had 137Cs losses, r e fle c tin g erosional cases. A possible explanation fo r th is d if fe r e n t r e s u lt be the fa c t th a t 5T2's 137Cs a c t iv it y r e fle c ts sediment tra n s p o rt. could It s p o sitio n on a sho rt, steep slope could r e la te to sediment tran s p o rt from s ite 6T2, and possibly from 7T2, which was also on the same tran sect and 57 had the highest h illt o p erosion ra te . The were located on slopes th a t were not as the sediment has already been has not y e t reached them fa c t th a t the other foots!opes steep could in d ic a te th at e ith e r transported through those s ite s or th a t i t (see Figures 2 and 7 ). The t i l l e d channel s ite s , 4T1, 4T2 and SC, a ll have deposition th a t are closely grouped, ranging from 37.9 to 49.6 t standard deviation fo r th is group, 4 .9 t ha'1 yr"1, was e n tire watershed. This indicated th a t th is s ite would lik e ly those sampled. underestimated. Therefore, it s This is very already the highest in the ha'1 yr'1. The the lowest in the The other ephemeral channel s it e , amount o f 137Cs a c t iv it y in the deepest sample rates SC, has a large increment (see Figure 15). have more 137Cs at depths below deposition ra te is almost c e rta in ly s ig n ific a n t since SC's deposition ra te was watershed. The s ite with the highest erosion in the watershed (51.5 t ha"1 yr"1) was 1C, located in the incised channel. The other incised channel 2C and SC, have erosion rates of 34.2 and 26.5 t ha"1 yr"1, S ite 3C's re s u lt is misleading because i t did not However, it was trea te d as an concentration was much less than erosional re sp ec tiv e ly . reach "zero" a c t iv it y . s ite the c o n tro l. s ite s , because it s to ta l The p o s s ib ility th a t rece n tly deposited sediments have buried the n a tu ra lly deposited 137Cs remote because the sample was collected from between large boulders were placed in the incised channel bottom to prevent channel Despite th is p ro te c tiv e measure th is channel was a c tiv e ly is th a t scour. eroding at an average erosion ra te o f 37.4 t ha"1 yr"1. The re s u lts fo r the pond s ite s indicated the same problem with sample depth. Referring to Figure 15, s ite s IP , 4P and 6P a ll in d icate th a t the 58 sample depth did not reach an increment w ith "zero" 137Cs a c t iv it y , which suggests th a t pond 137Cs deposition was almost c e r ta in ly underestimated. Pond s ite 7P, located on the pond margin, exh ib ited 137Cs a c t iv it y th a t was c h a r a c te r is tic o f an erosional s it e . This s ite was located on the margin o f the pond near the dam, which was b u ilt in 1973 w ith excavated s o ils from other areas. It s 137Cs a c t iv it y and erosion ra te probably in d ic a te a s ite th a t was s o il pedon). average areal modified during dam construction ( i . e . i t is not a natural This s ite was elim inated from the c a lc u la tio n o f the concentration fo r the pond s ite s . Using the erosion and deposition ra te s , i t was possible to determine averages o f erosion or deposition ra te s fo r each topographic po sition (Table 1 0 ). ra te was The topographic p o sitio n w ith the highest erosion the incised channel and the lowest was the foots!opes. deposition were 243.0 and re s p e c tiv e ly . midslopes The Mass rates o f 61.0 Mg yr"1 fo r the t i l l e d channels and pond, highest mass ra te o f erosion was a ttrib u te d to the a t 609.8 Mg yr"1, topographic p o s itio n s , le v e l o f confidence which also 21.7 ha. had the la rg e s t area o f the This group o f samples also had a high and, considering the problems p e rta in in g to sample depth fo r the channel and pond s ite s , is li k e l y the most accurate o f a ll the topographic p o s itio n s . This is re in fo rced by i t s low standard d e v ia tio n (8 .0 t ha"1 yr"1) . The mass ra te estim ates in dicated a c tiv e ly eroding midslopes and h illt o p s , both erosional and depositional footslopes, s ig n ific a n t o f 137Cs laden sediments in the ephemeral channel system and d e p o s itio n in the small pond. storage substantial The watershed net mass erosion ra te is approxim ately 631 Mg yr"1, which is the sum o f the mass ra te s in Table 10. 59 Table 10. Average Cesium-137 Erosion and Deposition Mass Rates by Topographic P ositio n . Topographic Position Sample Size Area ha H illto p s Midslopes Footslopes T ille d Channels Incised Channels Pond 3 4 4 13.5 21.7 19.6 Erosion/ Standard Confidence Total Deposition D eviation Level Erosion/ Rate8 Deposition t ha"1 y r'1 % Mg year"1 23.0 28.1 0.5 14.6 8 .0 22.7 44.4 78.9 34.0 310.5 609.8 9 .8 4 5.65 -4 3 .0 4.9 95.5 -243.0 4 4 0.15 0.25 37.4 -243.9 10.5 9.3 65.8 72.9 5.6 -6 1 .0 “Erosion and deposition ra te s , negative rates in d ic a te depo sition . Confidence values from nonparametric s t a t is t ic a l a n a ly s is . 60 CHAPTER 4 DISCUSSION USLE and WEE Soil Loss Estimates The USLE and WEE are the most widely used models fo r evaluating erosion hazard in Montana. The Soil Conservation Service applies the USLE and a version o f the WEE, the WEQ, to determine q u a lific a tio n fo r and maintenance in the Conservation Reserve Program (CRP) program. SCS estimates fo r th is watershed are 0.7 and 10.10 t ha"1 yr"1 fo r the USLE and WEQ, re s p e c tiv e ly . The underlying ob jective of SCS a p p lic a tio n 's these models is to determine average conditions. This study sought estimate average conditions as well as the spatial v a r ia b ilit y of losses by using s p a tia lly v a riab le fa c to r values (Tables 11 and in to the USLE while the SCS's application incorporated only one to so il 12). The study incorporated s p a tia lly v a ria b le topographic fa c to r the e n tire watershed. of values value fo r Both studies used the same R, K and P factors and the C facto rs were s im ila r (0.234, th is study, and 0.242, SCS) with the exception o f six points in the grass area fo r which a d iffe r e n t value was used (0 .0 5 3 ). d iffe r e n t . The topographic facto rs of the two The SCS's estimates r e fle c t a slope length times the longest slope length measured in th is gradient studies s lig h tly la rg e r than the sm allest study are very g reater than ten and gradient measured a slope in th is 61 Table 11. Comparison o f USLE Factor Estimates Used by Author and USDA-SCS. Factor By Author R a in fa ll -e r o s iv ity , R 476.6 Soil c r e d ib i lit y , K By USDA-SCS 476.6 0.049 Slope le n g th , L 0.049 1,600 m 19.0 m to 92.2 m 0 .4 % 0.30 % to 4.17 % Slope g ra d ie n t, S 0.12 0.01 to 0.56 LS Cover management, C 0.242 0.2344 or 0.053 Supporting p ra c tic e s , P 1.0 1.0 Soil lo ss , A 4.47 t ha"1 yr"1 0.67 t ha"1 y r'1 study. N eith er o f these values is close to watershed from a series o f 81 points and are th e re fo re not study watershed. the USLE g re a tly The accuracy of the topographic a ffe c ts the model's a b il it y re p res e n ta tiv e o f the fa c to rs ' estim ation fo r to v a ria b le erosion rates in the watershed (Wischmeier, Smith, 1978; Wilson, 1986b). and deposition areas. ap p lic a tio n s are very This study also The average re s u lts averages computed p re d ic t s p a tia lly 1976; Wischmeier and distingu ishes net erosion produced from these two USLE d iffe r e n t and the point re s u lts produced in th is study were highly v a ria b le . 62 Table 12. Comparison o f WEE Factor Estimates Factor Estimates Used by USDA-SCS. Factor Used By Author Soil c r e d ib ility ,! ' 107.6 t ha"1 yr"1 0 .4 9 , 0.66 and 0.90 1.00 0.90 0.90 0 to 155 m 142 m C lim ate, C F ie ld le n g th , L Vegetation Residue, V By USDA-SCSa 180.2 to 670.3 t ha"1 yr"1 Soil ridge roughness, K by Author and WEQ 1200 kg ha"1 130 to 3025 kg ha"1 10.10 t ha"1 yr"1 4 .7 t ha"1 yr"1 Soil lo ss, E aWEQ estim ates made by SCS f ie ld personnel, Fort Benton, MT, fo r Jackson Farm fo r use in Conservation Reserve Program (CRP) co n tra c ts. The WEE a p p lic a tio n method a p p lic a tio n in th is study d iffe r s from the SCS WEQ because i t computes wind erosion by periods where the SCS u t iliz e s sin g le fa c to r estimates th a t are n e ith e r tem porally nor s p a tia lly v a ria b le (See Table 12). study is The mean o f the I ' values used in th is more than twice as large as the SCS value. This study also adjusted I ' fo r periods follow ing c u ltiv a tio n , and also incorporates the I s, knoll varied c r e d ib i lit y , fa c to r at fiv e p o in ts . a t each point according to s t r ip Eighty-one L facto rs were fetch le n g th . Many of the L estim ates were less than the SCS L estim ate o f 142 m, accounting fo r many low wind erosion estimates and a low average erosion estim ate (4 .7 t ha"1 yr"1) in comparison to the SCS (WEQ) estim ate (10.1 t ha1 y r 1) . The C 63 fa c to rs used are the same but th is study determines period s o il losses as a function o f the year c ro p /fa llo w maximum as a at c y cle . The SCS estim ate o f K was held constant a t it s possible value, whereas, th is study c a lc u la te d period K values function o f t i l l a g e operations. varied SCS. d is tr ib u tio n o f erosive wind energy (EWE) over the two The V fa c to r in th is study also by period , as opposed to the s in g le , low estim ate p referred by However, the SCS a p p lic a tio n suggested th a t the watershed is eroding 10.1 t ha"1 yr"1 w hile th is study found an average o f only 4 .5 t ha"1 yr"1. Figure confirmed by 12 shows th a t these re s u lts also were hig h ly v a ria b le , the standard d eviatio n o f the WEE average (4 .6 t ha"1 yr"1j . This is s im ila r to the v a r i a b ili t y o f th is study's USLE ha"1 yr"1) , confirm ing the fa c t th a t modeled wind and estim ate (3 .6 t water erosion ra te s in the watershed are highly v a ria b le . It is possible to make several general observations based on these r e s u lts . The SCS appears to apply these models a t too coarse a scale (a t in le a s t landscapes with hummocky r e l i e f ) . They also have used n o n-rep resentative L and S fa c to rs in the USLE; where the L is too long and the S is near the gentle end o f the measured d is tr ib u tio n . This study d e lin e a ted erosional and depositional zones, w h ile the SCS made attempt to do th is and consequently characterizes the e n tir e as e ro s iv e . This is obviously not tru e fo r the ephemeral pond where deposition was indicated by the 137Cs method. th a t the SCS's estim ates are not a d d itiv e because they w ith a p a r tic u la r po int in mind and because o f the erosion and deposition areas. no watershed channel and This suggests were not produced f a ilu r e to d efin e net 64 The model re su lts show a great deal o f v a r ia b ilit y (see Figures 12 and 14). The model re su lts are combined in Figure p o in t's erosion. net depositional This zones were step was id e n tifie d based 14 to depict each on the by a rb itra ry 10, fa c t rules th a t fo r the the USLE (w a te r), but th is was not done (nor could i t be done) fo r the WEE (w in d ). This alludes to the question at issue, "How accurately do the USLE and WEE estim ate so il loss and sp atial v a r ia b ilit y ?". V a lid a tio n o f USLE and WEE Soil Loss Estimates A p o s s ib ility exists th a t the models or the 137Cs method, or both are simply wrong, at le a s t fo r estim ating so il re d is trib u tio n processes in the northern Great Plains and Montana. v a lid it y o f model In the past, an assessment re s u lts was d i f f i c u l t a v a ila b le to answer th is (above) question. because there were no However in the la s t years, 137Cs has offered an opportunity here and elsewhere to data to address the question and assess s p a tia l (De Jong e t a l., 1983; Arnalds, 1984; of the several a tta in f ie ld v a r ia b ilit y in soil loss Pennock However, the app licatio n of th is method in th is data and De Jong, 1987). study revealed two major lim ita tio n s which warrant fu rth e r explanation. The in a b ilit y o f the s t r a t if ie d sampling scheme used in th is and other studies to capture the spatial v a r ia b ilit y of 137Cs a c tiv ity is one lim ita tio n . Although a la rg e r number o f data points might improve the confidence le v e ls of the erosion estimates d iffe r e n t landscape inclusion o f samples collected from u n its , the the gently sloping, western portion o f the watershed may make matters worse. The transect samples used were concentrated portion in the steeper, eastern of the watershed and 65 e x tra p o la tin g t h e ir average 137Cs whole watershed might have occurred. Even if in dicated more net erosion than has r e a lly both areas p ro p o rtio n a lly , the s t r a t i f i e d major lim it a t io n because i t v a r i a b i l i t y o f 137Cs approach could be flow in to the a c t iv it ie s to the eroding portion o f the of the watershed were sampled sample scheme employed would s t i l l be a might not capture a ll aspects o f the s p a tia l a c t iv it y and s o il e ro s io n /d e p o s itio n . An a lte r n a tiv e to incorporate divergence and convergence o f overland landscape un its such as th a t used by Pennock and De Jong (1 9 87 ). Temporal v a r i a b ili t y is a second lim ita tio n and may be ju s t d i f f i c u l t to account fo r in s o il samples analyzed fo r 137Cs a c t iv it y . frequency and magnitude of wind and water erosion im portant ro le in the dynamics o f s o il lo ss . between October and March in th is region . events play as The an Most wind erosion occurs Water erosion has freq u en t, small magnitude events moving sediments to foots!opes and the ephemeral channel system, and, in fre q u e n t, la rg e magnitude events moving from these s ite s through the incised channel to the pond. sediments It is im portant to know the stage o f s o il re d is tr ib u tio n represented by the 137Cs a c t iv it y . This a c t iv it y depends on the recent h is to ry o f f lu v ia l erosion events ( i . e . whether or not a la rg e magnitude event has occurred in the short-term p a s t). Th erefore, careful designation o f landscape u n its is c r i t i c a l to the accurate s o il g a in s/lo s s e s . This study used sample sample locations and q u a n tific a tio n o f 137Cs and lo catio ns based on equal areas o f slope delin eated by f i e l d pacing and s o il samples c o lle c te d over a four week period in mid-summer. adequately c h a ra c te rize the This sample d e lin e a tio n method f a ile d to tem porally v a ria b le dimensions o f 137Cs and 66 s o il re d is trib u tio n on the Jackson Farm. Despite these problems some data were collected th a t can be used qu an tify and in te rp re t so il erosion/deposition rates a t s ite s from so il samples were c o lle c te d . A s ite by s ite comparison which of so il erosion/deposition rates estimated by the models and 137Cs approaches performed to evaluate the model re s u lts . to was The individual s ite s ' predicted and measured rates were analyzed with lin e a r regression. Predicted rates were computed by adding together USLE and WEE estimates fo r each of the eroding 137Cs sample points (Table 13). The 137Cs depositional s ite s were not used because the models do not apply to them. A lin e a r regression of ra te estimates fo r the 10 remaining s ite s yielded an R2 o f 0.07. I f the three s ite s fo r which the USLE did not apply ( i . e . , based on the h illto p s , channels, and/or concave slope exclusion ru les) are removed from the sample set the re s u lt improves only m arginally (R2 = s c a tte rp lo t is shown in Figure 17. These re su lts not work fo r estim ating to ta l erosion in the 0 .1 4 , n = 7 ). A in d ic a te the models may C arter watershed. Soil Erosion/Deposition Rates In ferre d from Cesium-137 Gains/Losses Based on the comparison of measured versus predicted ra te s , the USLE and WEE do not appear to work very well However, it p a r tic u la r ly appears with th a t respect the to 137Cs method understanding is in the study watershed. not very sp a tia l good e ith e r, v a r ia b ilit y and estim ating a sediment budget unless many more data points are collected and analyzed (le v e ls of confidence range from 34.0 to 95.5 % by landscape u n it). 67 Table 13. Site by Site Comparison of Model and 137Cs Erosion/Deposition Rate Estimates. S ite USLE IT l 2T1 3T1 4T1 5T1 6T1 7T1 1T2 2T2 3T2 4T2 5T2 6T2 0.0 10.4 15.0 0.0 45.6 9.3 0.0 0.0 13.2 1.9 0.0 11.5 5.1 Models Total WEE t ha"1 yr"1 10.2 9.9 9.2 8.5 9.0 9.7 10.1 10.4 12.7 12.7 12.7 5.1 4.5 10.2* 20.3 24.2 8.5 44.6 19.0 10.I a 10.4* 25.9 14.6 12.7 16.6 9.6 137Cs 4.9 23.6 10.0 -41.6" 18.5 33.1 40.7 23.2 17.5 12.1 -37.9" -38.5" 38.1 aSite predicted value attained from WEE estimate only (shown as empty symbols in Figure 17). 6Depositional s ite excluded from regression analysis. M EASUR ED ER O SIO N (T /H A /Y R ) Figure 17. S catterplot diagram of measured and predicted values generated fo r s ite by s ite comparison of 137Cs sample sites (R2 = 0.07, n = 10). Empty symbols are sites with WEE estimate only, with these sites removed, R2 = 0.14 (n = 7). 68 However, i t is possible to compute a f i r s t estim ate o f the -watershed sediment budget (Table 14) with the data th a t were collected,. ..Several problems, including the Tow (le s s than 60 %) le v e ls o f confidence fo r of the landscape u n its , the i n a b ili t y to separate wind and water processes, overlapping erosion and depo sition areas, the inadequate sample depth and the fa c t th a t the 137Cs a c t iv it ie s ( p a r tic u la r ly the t i l l e d two pond are time-dependent channels.) in d ic a te the need to in te r p r e t such re s u lts c a u tio u s ly . The 137Cs a c t iv it ie s in dicated a net mass erosion r a te fo r the portions o f the watershed (about 54 ha) o f 9336 Mg y r'1.. The deposition ra te a ttr ib u ta b le to f lu v ia l deposition (304 Mg yr"1) in te rp re te d to represent f lu v ia l erosion. erosion ra te from the to ta l erosion ra te Subtracting th is eroding mass could be mass water y ie ld ed a mass erosion estim ate fo r wind (a net value not counting deposition) o f 632 Mg yr"1. The mass wind erosion ra te was used to . compute the net wind erosion ra te o f 10.4 t ha'1 yr"1 and the mass erosion ra te o f 5 .0 water erosion ra te was used to compute a net water t ha"1 yr"1 fo r a to ta l o f 15.4 t ha"1 yr"1. The USLE and WEE estimates o f s o il loss predicted a net s o il (exclu sive o f several loss of 9 t (4 .5 t ha"1 yr"1 fo r ha"! yr"1 in the study b o th ), watershed USLE points representing deposition and one WEE po in t where wind erosion does not a p p ly ). This r e s u lt suggests th a t the models predicted only 58 % o f the t o ta l erosion estim ated w ith the 137Cs method. A comparison o f the net f lu v ia l erosion ra te (5 .0 t ha"1 yr"1) and the USLE watershed average showed th a t the USLE predicted 90 % o f the measured water erosion even though th is USLE average was known to because i t did not include channel erosion. The WEE estim ate be low (4 .5 t ha"1 69 Table 14. Summary of Watershed Average 137Cs Erosion and Deposition Mass Estimates. Topographic P osition Area ha Mass Estimates3 Mg y r'1 13.5 21.7 19.6 0.15 311 610 10 6 Erosion ( f lu v i a l and a eo lian ) H illto p s Midslopes Footslopes Incised Channel Measured Erosion Mass Estimate 936 T ille d Channels Pond -.243 -61 Deposition ( f lu v ia l and aeolian ) 5.65 0.25 Measured Deposition Mass Estimate -304 aMass estim ates were c a lcu lated by m u ltip ly in g the topographic p o s itio n s ' average e rosion/deposition ra te by i t s ' area. Negative values r e f le c t dep o sitio n . yr"1) was much less than the watershed net wind erosion ra te (1 0 .4 t ha"1 yr"1) and accounted fo r only 44 % o f the WEQ estim ate (10.1 t measured wind erosion. The SCS's ha"1 yr"1) is s im ila r to the net wind erosion r a te , but d i f f i c u l t to explain because they did not use re p re s e n ta tiv e s o il c r e d ib i lit y or f i e l d length fa c to r estim ates. T h erefore, if the 137Cs re s u lts were t r u ly in d ic a tiv e o f the s o il re d is tr ib u tio n processes in the study watershed, then the USLE has performed much b e tte r than the WEE in p re d ic tin g s p a t ia lly v a ria b le erosion ra te s . However, we probably cannot t o t a l l y r e je c t the WEE re s u lts unless more 137Cs data are gathered in the hope o f reducing the variances. 70 Conclusions Both the USLE and WEE model and 137Cs methods indicated a substantial erosion hazard and a great deal of sp a tia l v a r ia b ilit y . is p e rfe ct as applied here v a r ia b ilit y o f the hazard. but s u ffic ie n t to N either approach show the extent and C le a rly , the re su lts in d ic a te th a t there has been s ig n ific a n t f lu v ia l and aeolian erosion of h illto p s , midslopes, some foots!opes and an incised channel over 33 deposition on one footslope, in the pond years, and s ig n ific a n t flu v ia l and in the ephemeral channel system over 14 years. Despite problems associated with capturing temporal and sp atial v a r ia b ilit y , the 137Cs data suggests th a t the USLE point method used much more e ffe c tiv e than the point WEE method used. Measured flu v ia l erosion rates were 10.4 and 5.0 t ha'1 yr"1, the modeled erosion rates were 4.5 t ha"1 yr"1 fo r was aeolian and re sp ec tiv e ly ; while both wind and water. These re su lts have important consequences fo r crop p ro d u c tiv ity now and in the fu tu re (despite low p re c is io n ). At present, 50 years of erosion and deposition has occurred, ju s t one of many facto rs influencing and explaining current crop p ro d u c tiv ity . For the fu tu re , there is s t i l l a need to compute s p a tia lly v a riab le s o il erosion rates in order to b e tte r understand so il scales. erosion/crop p ro d u c tiv ity re la tio n s h ip s at watershed REFERENCES CITED 72 References C ited Anonymous., 1989, "Known Nuclear Tests: 1945 to December 31, B u lle tin o f Atomic S c ie n tis ts , 4 5 (1 ), A p r il, 1989, 48. 1988", Armbrust, D .V ., and Lyles, L ., 1985, "Equivalent Wind Erosion P rotection from Selected Growing Crops", Agronomy Journal. 77, 703-707. Arnolds, 0 . G. , 1984, "Radiocesium in Montana S oils and A pplications fo r S oil Erosion Measurement", Unpublished M.Sc. Thesis, Montana S tate U n iv e rs ity , 58p. Arnalds, O .G ., C u ts h a ll, N .H ., and N ielsen, G .A ., 1989, "Cesium-137 In Montana S o ils " . Health Physics. 5 7 (6 ), December, 1989, 955-958. Bondy, £ . , Lyles, L.., and Hayes, W.A., 1980, "Computing Soil Erosion by Periods Using Wind Energy D is trib u tio n " , Journal o f Soil and Water Conservation, 3 5 (4 ), 173-176. Brown, R .B ., C u ts h a ll, N .H ., and K lin g , E .F ., 1981a, " A g ric u ltu ra l Erosion In dicated by Cesium-137 R e d is trib u tio n : I , Levels and D is trib u tio n o f Cs-137 A c tiv ty in S o ils " , Soil Science Society o f America Journal, 45, 1184-1190. Brown, R .B ., K lin g , E .F ., and C u ts h a ll, N .H ., 1981b, " A g ric u ltu ra l Erosion In dicated by Cesium-137 R ed is trib u tio n : I I , Estimates Of Erosion Rates", Soil Science Society o f America Journal. 45, 1191-1197. Campbell, B .L ., Loughran, R .J ., and E l l i o t t , G .L ., 1982, "Cesium -137 as an In d ic a to r o f Geomorphic Processes in a Drainage Basin System", A u s tra lia n Geographical S tudies, 20, 49-64. C h e p il, W .S., Siddoway, F .H ., and Armbrust, D .V ., 1962, "C lim atic Factor fo r Estim ating Wind E r o d ib ility o f Farm F ie ld s ", Journal o f Soil and Water Conservation, 1 7 (4 ), 162-165. Coleman, D. E ., 1982, " Id e n tific a tio n o f P r io r it y Management Areas in the Avon R iver Watershed", Upper Thames R iver Conserv. A u th ., London, O ntario. Compton, R .R ., 1962, "Manual o f F ie ld Geology", John W iley and Sons, In c ., New York, NY, 162p. 73 C utshalI , N .H ., and Larsen, 1. 1 . , 1980., "Peak-By-Peak C orrection o f G e(Li) Gamma-Ray Spectra fo r Photopeaks from Background", Nuclear Instruments and Methods. 178, 233-236. De Jong., £ . , V illa r ., J i,, and Bettany., J.R . , 1982, "P relim inary In v e s tig a tio n s on the Use o f Cesium-137 to Estimate Erosion in Saskatchewan", Canadian Journal o f Soil Science. 62, 673-683. De JoPg, E ., Begg, C.B ;M ., and Kachanoski, R .G ., 1983, "Estimates Of Soil Erosion and Deposition fo r Some Saskatchewan S o ils , Canadian Journal o f Soil Science, 63, 607-617. Delaune, R .D ., Patrick,, VLH. J r . , Buresh, R ..J., 1978, "Sedimentation Rates Determined by Cs-137 Dating in a Rapidly Accreting S a lt Marsh", N ature, 275, 532-533. D ibb, J . , 1989, "S c ie n tis ts Discover Problems Using Chernobyl F a llo u t fo r Dating the Greenland Ice Sheet", Earth in Space. 2 ( 3 ) , November, 1989, 8 -9 . Dyer, E .B ., 1977, "Use o f the Universal Soil Loss Equation: A River Basin Experience", In "Soil Erosion: P redictio n and C o n tro l", Foster, G. R ., e d ., S oil Conservation Society o f America, Ankeny, Iowa, 292-297. Fo ster, G .R ., McKool, D .K ., Renard, K .G ., and Moldenhauer, W.C., 1981, "Conversion o f the Universal Soil Loss Equation to SI M etric U n its ", Journal o f Soil and Water Conservation, 3 6 (6 ), 355-359. Foster, G .R ., and Wischmeier, W.H., 1974, "Evaluating Ir r e g u la r Slopes fo r S oil Loss P re d ic tio n ", Transactions o f the American Society of A g ric u ltu ra l Engineers. 1 7 (2 ), 305-309. G re e n v ille , T .N .E ., 1967, "Spline Functions, In te rp o la tio n , and Numerical Quadrature", In "Mathematical Methods fo r D ig ita l Computers", Ralston, A ., and W olf, H. S ., e d s ., John W iley, New York, NY, 156-168. G r if f i n , M .L ., Beasley, D .B ., F le tc h e r, J . J . , and Fo ster, G .R ., 1988, "Estim ating Soil Loss on Topographically Nonuniform F ie ld and Farm U n its ", Journal o f Soil and Water Conservation, 4 3 (4 ), 326-331. 74 H i l l i a r d , R .L ., 1977, "The New York Erosion and Sediment In vento ry", In "Soil Erosion: P red ictio n and C o n tro l" , Foster, 6. R. , e d ., S o il Conservation Society o f America, Ankeny, Iowa, 273-276. Kachanoski, R .G ., 1987, "Comparison o f Measured Soil 137-Cesium Losses and Erosion Rates", Canadian Journal o f Soil Science. 67, 199-203. Lance, J .C ., M cIntyre, S .C ., Nancy, J .W ., and Rousseva=, S .S ., 1986, "Measuring Sediment Movement a t low Erosion Rates Using Cesium-137", Soil Science Society o f America Journal. 50, 1303-1309. Larsen, I . L . , and CutshalI , N .H ., 1981, "D ire c t Determ ination o f 7Be in Sediments", Earth and P lanetary Science L e tte rs . 54, 379-384. Lowrance, R ., M cIntyre, S ., la n c e , C ., 1988, "Erosion and Deposition in a F ie ld /F o re s t System Estimated Using Cesium-137 A c t iv it y " , Journal o f Soil and Water Conservation. 4 3 (2 ), 195-199. McCall an, M .E ., O'Leary, B.M ., and Rose, C.W., 1980, "R e d istrib u tio n o f Cesium-137 by Erosion and Deposition on an A u s tra lia n S o il" , A u stralian Journal o f S oil Research. 18, 119-128. McHenry, J .R ., R itc h ie , J .C ., and G i l l , A .C ., /1 9 7 3 , "Accumulation Of F a llo u t Cesium-137 in S o ils and Sediments in Selected Watersheds", Water Resources Research, 9 ( 3 ) , 676-686. McHenry, J .R ., and R itc h ie , J .C ., 1977, "Estim ating F ie ld Erosion Losses from F a llo u t Cesium-137 Measurements", In "Erosion and S olid M atter Transport in Inland Waters. Proc. Symp.", J u ly , 1977. IAHS P ub lication #122, 26-32. M cIntyre, S .C ., Lance, J .C ., Campbell, B .L ., and M il le r , R .L ., 1987, "Using Cesium-137 to Estimate Soil Erosion on a C learcut H ills id e " , Journal o f Soil and Water Conservation. 4 2 (2 ), 117-120. Pennington. W., Cambray, R .S ., Eakins, J .D ., and Harkness, D .D ., 1976, "Radionuclide Dating o f Recent Sediments o f Blelham Tarn", Freshwater B iology. 6, 317-331. Pennock, D .J ., and De Jong, E ., 1987, "The.Influence o f Slope Curvature on Soil Erosion and Deposition in Hummock T e rra in " , S oil Science, 1 4 4 (3 ), 209-217. 75 R itc h ie , J .C ., Spraberry, J .A ., and McHenry, J .R ., 1974, "Estim ating Soil Erosion from the R e d is trib u tio n o f FaTiout Cesium-137", Soil Science Society o f America Proceedings. 38, 137-139. R itc h ie , J .C ., Hawks, P .H ., and McHenry, J .R ., 1975, "Deposition Rates in V alleys Determined Using F a llo u t Cesium-137", Geological Society o f America B u lle t in . 86, 1128-1130. R itc h ie , J .C ., and McHenry, J .R ., 1975,, "F a llo u t Cesiumr 137: A Tool in Conservation Research", Journal o f Soil and Water Conservation. 30, 283-286. Rogowski, A .S ., and Tamura, T ., 1970, "Erosional Behavior o f Cesium- 137", Health Physics. 18, 467-477. Skidmore, E .L ., and Woodruff. N .P ., 1968, "Wind Erosion Forces in the United States and T h e ir Use in P redictin g Soil Loss", A g ric u ltu re Handbook #346, A g ric u ltu re Research S ervice, United States Dpartment of A g ric u ltu re , Washington, D. C. Skidmore, E .L ., 1988, "Wind Erosion" In "Soil Erosion Research Methods", L a l, e d ., Soil and Water Conservation S ociety, Ankey, Iowa, 244p. S n e ll, E .A ., 1984, "A Manual fo r Regional Targeting o f A g ric u ltu ra l Soil Erosion and Sediment Loading to Streams", Evironment Canada, Environment Conservation S ervice, Lands D ir . Working Paper # 36, Ottowa, O n ta ria. S n e ll, E .A ., 1985, "Regional Targeting o f P o ten tial Soil Erosion and Nonpoint-source Sediment Loading", Journal o f Soil and Water Conservation, 40, 521-524. Trim ble, S.W ., 1974, "Man-induced Soil Erosion on the Southern Piedmont, 1700-1970", Soil Conservation Society o f America, Ankeny, Iowa. Trim ble, S.W ., 1977, "The Fallacy o f Stream E quilibrium in Contemporary Denudation S tudies", American Journal o f Science. 277, 876-887. Trim ble, S.W ., 1983, "A Sediment Budget fo r Coon Creek, the D r iftle s s Area, Wisconsin, 1852-1977", American Journal o f Science, 283, 454-474. 76 United States Department o f A g ric u ltu re , S oil Conservation S ervice, 1988, "1988 National Agronomy Manual", Part 502, "P redicting Wind Erosion", U.S. Government P rin tin g O ffic e , Washington D. C. United States Department o f A g ric u ltu re , Soil Conservation S ervice, 1988, "1988 National Agronomy Manual", Montana Supplement, U.S. Government P rin tin g O ffic e , Washington D. C. Van V l i e t , L .J .P ., and W all, G .J ., 1979, "Comparison o f Predicted and Measured Sheet and R ill Erosion Losses in Southern O n tario ", Canadian Journal o f Soil Science. 59, 211-213. Van V l i e t , L .J .P . and W all, G .J ., 1981, "Soil Erosion Losses from Winter Runoff in Southern O n tario ", Canadian Journal o f S oil Science, 61, 451-454. W a ll, G .J ., Van V l i e t , L .J .P ., and Dickinson, W .T., 1982, "A g ricu ltu re and Water Q u a lity in the Canadian Great Lakes: 2, F lu v ia l Sediment", Journal o f Environmental Q u a lity . 11, 482-486. W illiam s , J .R ., and Berndt, H .D ., 1977, "Determining the Universal Loss Equation's Length-slope Factor fo r Watersheds", In "Soil Erosion: P red ictio n and C o n tro l", Foster, G. R ., e d ., Soil Conservation Society o f America, Ankeny, Iowa, 217-225. W illiam s, J .R ., Jones, C .A ., and Dyke, P .T ., 1984, "A Modeling Approach to Determining the R elationship between Erosion and P ro d u c tiv ity ", Transactions o f the American Society o f A g ric u ltu ra l Engineers, 27, 129144. Wilson, J .P ., 1986a, "The Use o f S t a t is t ic a l Models to Document Environmental Change in the Lake Simcoe-Couchiching Basin", Unpublished Ph.D. D is s e rta tio n , Department o f Geography, U n iv e rs ity o f Toronto, 358p. Wilson, J .P ., 1986b, "Estim ating the Topographic Factor in the Universal S oil Loss Equation fo r Watersheds", Journal o f Soil and Water Conservation. 4 1 (3 ), 179-184. Wilson, J .P ., 1989, "Soil Erosion from A g ric u ltu ra l Land in the Lake Simcoe-Couchiching Basin, 1800-1981", Canadian Journal o f Soil Science, 69, 137-151. 77 Wischmeier, W.H., and Smith, D .D ., 1978, "P redicting R a in fa ll Erosion Losses- A Guide to Conservation Planning", A g ric u ltu re Handbook #537, United States Department o f A g ric u ltu re , Washington, D. C .. Wise, S .M ., 1980, "Caesium-137 and Lead-210: A reveiw o f the Techniques and Some A pplications in Geomorphology", In "Timescales in Geomorphology", C u llin g fo rd , R .A ., Davidson, D .A ., and Lewin, J . , e d s ., John Wiley & Sons L t d ., London, England. Woodruff, N .P ., and Siddoway, F .H ., 1965, A Wind Erosion Equation", Soil Science Society Proceedings, 29, 602-609. 78 APPENDICES APPENDIX A EROSION MODEL RESULTS 80 Table 15. USLE Factor and Soil Loss Point Estimates.® Point Slope Gradient Slope Length # % m IF IG 2M 2N 20 2P 3E 3F 3G 4N 40 4P SE 5F SG SH SM SN 60 SP 7B 7C 70 7E 7F 7G 8J 8K 8L SM SN 80 8P 9A 9B 9C 90 9E 9F 9G 9H 91 10.7 6.7 0.2 1.7 16.0 8.0 0.8 6.7 7.3 0.0 0.0 0.0 0.9 1.8 7.3 4.3 0.3 2.4 7.3 4.4 0.3 0.2 2.1 0.0 5.7 8.0 2.5 1.4 1.1 2.8 1.7 6.3 6.7 0.7 3.6 1.8 3.1 1.3 4.7 0.8 7.3 10.0 105 120 0 118 55 26 33 43 50 10 0 18 33 60 19 50 25 13 135 53 45 2 80 0 24 90 80 108 20 123 83 38 63 13 123 213 55 153 168 190 60 28 USLE USLE C Valueb S Valuec Exclusion Ruled Erosion Estimate6 t ha"1 yr"1 0.053 0.053 0.234 0.234 0.053 0.053 0.234 0.234 0.234 0.234 0.053 0.234 0.234 0.234 0.053 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 1.295 0.661 0.074 0.161 2.455 0.844 0.105 0.661 0.743 0.065 0.065 0.065 0.111 0.168 0.743 0.380 0.079 0.211 0.743 0.391 0.079 0.074 0.189 0.065 0.535 0.844 0.219 0.141 0.123 0.243 0.161 0.609 0.661 0.100 0.313 0.168 0.268 0.135 0.422 0.105 0.743 1.169 0 0 I 0 0 0 0 0 0 I I I 2 0 0 0 0 I 0 0 0 I 0 2 0 0 0 2 0 0 0 0 0 I 0 0 0 0 0 3 0 0 5.20 2.84 N/A 1.88 7.14 1.69 0.74 7.50 9.09 N/A N/A N/A N/A 1.60 1.27 4.01 0.53 N/A 14.94 4.21 0.59 N/A 1.96 N/A 4.54 13.87 2.27 N/A 0.84 2.87 1.69 6.50 9.08 N/A 4.72 2.34 2.93 1.70 7.22 N/A 9.96 10.71 81 Table 15. USLE Factor and Soil Loss Point Estimates (Continued).a Point Slope Gradient Slope Length # % m IOJ IOK IOL IOM ION 100 IOP 10Q IlB IlC IlD HE HF HG IlH 111 12K 12L 12M 12N 120 12P 12Q 13C 130 13E 13F 13G 13H 14K 14L 14M 14N 140 14P 1.4 5.2 2.7 0.4 0.0 3.1 0.8 5.3 1.5 3.6 1.9 1.9 4.0 5.5 6.5 10.0 5.0 1.9 2.0 LI 3.6 8.0 0.0 3.1 3.0 1.4 1.4 1.2 1.2 1.6 3.1 2.0 1.3 1.4 1.8 30 120 200 380 0 112 80 90 30 125 285 373 60 83 165 40 38 163 288 153 58 25 0 44 80 198 138 125 23 5 104 38 95 125 68 USLE USLE C Valueb S Value0 Exclusion Ruled Erosion Estimate' t ha"1 y r 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.234 0.141 0.477 0.235 0.084 0.065 0.268 0.105 0.488 0.148 0.313 0.175 0.175 0.351 0.511 0.635 1.169 0.454 0.175 0.182 0.123 0.313 0.844 0.065 0.268 0.260 0.141 0.141 0.129 0.129 0.154 0.268 0.182 0.135 0.141 0.168 0 0 0 3 2 0 3 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 2 0 0 0 0 3 0 I 0 0 0 3 0 1.09 9.04 3.21 N/A N/A 3.90 N/A 8.02 1.14 4.75 2.65 2.88 3.97 8.07 14.12 12.80 4.85 2.25 2.77 N/A 3.49 7.31 N/A 2.68 3.30 1.92 1.73 N/A 0.92 N/A 3.79 1.51 1.48 N/A 1.66 82 Table 15. Point USLE Factor and Soil Loss Point Estimates (Continued).a Slope Gradient Slope Length # % m 15E 15F 15G 160 1.3 5.0 1.3 0.0 15 45 75 8 USLE USLE C Valueb S Valuec Exclusion Ruled Erosion Estimate® t ha"1 yr"1 0.234 0.234 0.234 0.234 0.135 0.454 0.135 0.065 I 0 0 I N/A 5.28 1.37 N/A aUSLE soil loss estimates calculated using R = 476.6, K = .049, L as calculated using equation # I, page 21 and P = 1.0. bUSLE C factor from Table 2. cUSLE S factor using formula of Foster and Wischmeier (1984). dExclusion code; O = apply USLE, I = exclude due to h i ll crest rule, 2 = channel rule, and 3 = concave slope rule. eN/A = USLE does not apply, erosion = 0.0 t ha"1 yr"1. 83 Table 16. Site WEE Factor and Soil Loss Point Estimates Soil Groupb r # IF IG 2M 2N 20 2P SE 3F SG 4N 40 4P SE SF SG SH SM SN 60 SP 7B 7C 70 7E 7F 7G 8J 8K 8L SM SN 80 8P 9A 9B 9C 90 9E 9F 9G Is rc (%) 38B 38B 38B 38B 38B 38A 38B 38B 38A 38B d 38A 38B 38B 38A 38A 38B 38B 38B 385B 38B 38B 38B 38B 38B 385B 38B 38B 38B 38B 38B 28 385B 385B 385B 38B 38B 38B 28 28 234.2 234.2 234.2 234.2 234.2 180.2 234.2 234.2 180.2 234.2 - 180.2 234.2 234.2 180.2 180.2 234.2 234.2 234.2 253.3 234.2 234.2 234.2 234.2 234.2 253.3 234.2 234.2 234.2 234.2 234.2 280.9 253.3 253.3 253.3 234.2 234.2 234.2 280.9 280.9 100 100 100 100 100 372 100 100 100 100 - 215 100 100 100 100 100 102 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 160 100 100 100 100 100 100 234.2 234.2 234.2 234.2 234.2 670.3 234.2 234.2 180.2 234.2 - 387.4 234.2 234.2 180.2 180.2 234.2 238.9 234.2 253.3 234.2 234.2 234.2 234.2 234.2 253.3 234.2 234.2 234.2 234.2 234.2 280.9 253.3 405.3 253.3 234.2 234.2 234.2 280.9 280.9 a Fetch Length Soil Loss Estimates Fallow Crop Total Cycle m Mg ha"1 yr"1 38 113 28 5 58 80 43 S3 28 148 5 80 80 25 18 S3 13 140 8 70 50 88 38 73 55 35 30 25 25 8 133 28 155 5 130 83 85 68 50 90 0.0 0.0 0.0 0.0 0.0 1.2 0.0 0.0 0.0 0.0 - 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.1 5.8 13.0 3.7 0.0 8.7 68.4 6.6 9.2 1.0 14.6 2.9 6.5 1.9 0.0 4.4 34.8 3.3 4.6 0.5 7.3 - - 27.4 10.8 2.9 0.0 5.0 0.0 14.8 0.0 11.7 7.6 11.4 5.8 10.2 8.3 6.5 4.1 2.9 2.9 0.0 14.0 6.7 17.0 0.0 15.9 11.0 11.2 9.7 11.6 16.2 13.8 5.4 1.5 0.0 2.5 0.0 7.4 0.0 5.9 3.8 5.7 2.9 5.1 4.2 3.3 2.1 1.5 1.5 0.0 7.0 3.4 8.6 0.0 8.0 5.5 5.6 4.9 5.8 8.2 84 Table 16. Site WEE Factor and Soil Loss Point Estimates Soil Groupb r # 9H 91 IO J IOK IOL I OM ION 100 IOP IOQ IlB IlC IlD HE H F HG IlH 111 12K 12L 12M 12N 120 12P 12Q 13C 13D 13E 13F 13G 13H 14K 14L 14M 14N 140 14P Is r c (%) 385B 385B 385B 385B 38B 38B 28 28 28 385B 385B 385B 385B 38B 28 28 28 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 385B 253.3 253.3 253.3 253.3 234.2 234.2 280.9 280.9 280.9 253.3 253.3 253.3 253.3 234.2 280.9 280.9 280.9 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 100 100 100 100 100 100 100 100 100 100 100 100 100 100 113 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 Fetch Length Soil Loss Estimates Fallow Crop Total Cycle m 253.3 253.3 253.3 253.3 234.2 234.2 280.9 280.9 280.9 253.3 253.3 253.3 253.3 234.2 317.4 280.9 280.9 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 253.3 45 68 65 18 20 I 128 23 150 15 123 78 80 60 45 83 40 63 10 15 140 25 86 143 8 73 73 55 40 8 33 5 8 133 115 10 138 a Mg ha"1 yr"1 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.1 8.4 11.4 11.1 1.7 1.5 0.0 18.8 5.7 20.0 0.7 15.5 12.6 12.6 8.9 14.1 15.6 9.8 10.9 0.0 0.7 16.3 4.0 13.1 16.5 0.0 11.9 11.9 9.9 7.5 0.0 6.1 0.0 0.0 16.0 15.1 0.0 16.3 4.2 5.7 5.6 0.9 0.8 0.0 9.5 2.9 10.0 0.4 7.8 6.3 6.3 4.5 7.1 7.9 4.9 5.5 0.0 0.4 8.2 2.0 6.6 8.3 0.0 6.0 6.0 5.0 3.8 0.0 3.1 0.0 0.0 8.0 7.6 0.0 8.2 85 Table 16. Site WEE Factor and Soil Loss Point Estimates (Continued).a Soil Groupb r # 15E 15F 15G 160 Is I'C (%) 385B 385B 385B 385B 253.3 253.3 253.3 253.3 100 100 100 100 Fetch Length Soil Loss Estimates Fallow Crop Total Cycle m 253.3 253.3 253.3 253.3 48 33 70 5 Mg ha"1 yr"1 0.0 0.0 0.0 0.0 8.9 6.1 11.7 0.0 4.5 3.1 5.9 0.0 aWEE soil loss estimates calculated using K = 0.49, 0.66 and 0.9, C = 0.90, and V varying by periods (see Table 5). bSoil mapping un it numbers; 28, Nishon series, 38A, 38B and 385B, Ethridge series (see Figure 3). cSoil e ro d ib ility (I and I ' ) in t ha"1 yr"1. dPoint excluded due to its location in the incised channel. APPENDIX B CESIUM-137 LABORATORY DATA AND AREAL CONCENTRATIONS 87 Table 17. Site3 # Cesium-137 Laboratory Data and Areal Concentrations Depth oi: Increment cm ITl 0 16 - 15 20 Bul k Density g cm"3 1.27 1.28 Mass Concerntra tio n Counting Error (±) Increment Areal A c tiv ity pCi g-1 0.53 0.05 Error (+) Total Site A c tiv ity pCi cm"2 0.05 0.01 10.10 0.32 0.95 0.06 10.42 2T1 0 15 16 - 20 1.30 1.34 0.34 0.05 0.04 0.01 6.63 0.34 0.78 0.07 6.97 3T1 0 16 - 15 20 1.30 1.34 0.47 0.00 0.04 0.00 9.17 0.00 0.78 0.00 9.17 4T1 0 16 21 26 - - 15 20 25 30 1.35 1.44 1.44 1.44 0.51 0.33 0.07 0.02 0.04 0.03 0.03 0.03 10.33 2.38 0.50 0.14 0.81 0.22 0.22 0.22 13.35 5T1 0 16 - 15 20 1.29 1.41 0.39 0.03 0.04 0.01 7.55 0.21 0.77 0.07 7.76 6T1 7T1 0 16 15 20 1.19 1.36 0 15 16 - 20 1.18 1.27 - 0.24 0.02 0.04 0.01 4.28 0.14 0.71 0.07 4.42 0.15 0.05 0.01 0.01 2.66 0.32 0.18 0.06 2.98 1T2 0 16 - 15 20 1.22 1.27 0.35 0.00 0.05 0.00 6.41 0.00 0.92 0.00 6.41 2T2 0 16 21 - 15 20 25 1.07 1.39 1.39 0.44 0.25 0.00 0.05 0.01 0.00 7.06 1.74 0.00 0.80 0.07 0.00 8.80 88 Table 17. S ite3 # Cesium-137 Laboratory Data and Areal Concentrations (Continued). Depth Bul k of Density Increment cm 3T2 0 - 15 16 - 20 Mass Concentra tio n Increment Areal A c tiv ity pCi g'1 g cm"3 1.19 1.41 Counting Error (±) 0.48 0.00 0.06 0.00 Error (+) Total Site A c tiv ity pCi cm"2 8.57 0.00 1.07 0.00 8.57 4T2 0 - 15 16 - 20 21 - 25 1.25 1.46 1.46 0.54 0.26 0.06 0.06 0.06 0.02 10.13 1.90 0.44 1.13 0.44 0.15 12.47 5T2 0 16 21 26 - 15 20 25 30 1.27 1.33 1.33 1.33 0.45 0.38 0.19 0.00 0.04 0.05 0.01 0.00 8.57 2.53 1.26 0.00 0.76 0.33 0.07 0.00 12.36 6T2 0 - 15 16 - 20 1.29 1.35 0.20 0.03 0.03 0.02 3.87 0.20 0.58 0.14 4.07 7T2 0 - 15 16 - 20 1.18 1.27 0.15 0.05 0.01 0.01 2.66 0.32 0.18 0.06 2.98 IC 2C 3C 0 - 5 6 - 10 11 - 15 1.40 1.40 1.40 0 6 11 16 - 5 - 10 - 15 - 20 1.40 1.40 1.40 1.40 0 6 11 16 21 - 1.40 1.40 1.40 1.40 1.40 5 10 15 20 25 0.24 0.06 0.00 0.03 0.01 0.00 1.68 0.42 0.00 0.21 0.07 0.00 2.10 0.27 0.28 0.18 0.02 0.03 0.03 0.04 0.04 1.89 1.96 1.26 0.14 0.21 0.21 0.28 0.28 5.25 0.30 0.12 0.08 0.26 0.16 0.05 0.01 0.03 0.04 0.16 2.10 0.84 0.56 1.82 1.12 0.35 0.07 0.21 0.28 0.07 6.44 89 Table 17. Site" # Cesium-137 Laboratory Data and Areal Concentrations (Continued). Depth Bul k oiF Density Increment cm SC 0 16 21 26 - 15 20 25 30 Mass Concentra tio n g cm'3 1.09 1.36 1.36 1.36 pCi 0.54 0.59 0.65 0.93 Counting Error (±) Increment Areal A c tiv ity Error (+) pCi cm'2 g '1 0.04 0.03 0.05 0.06 8.83 4.01 4.42 6.32 0.65 0.20 0.34 0.41 23.58 ' IP 0 6 11 16 21 Total Site A c tiv ity _ - 5 10 15 20 25 1.30 1.30 1.30 1.30 1.30 0.36 0.55 0.71 0.69 0.71 0.05 0.02 0.01 0.08 0.04 2.34 3.58 4.62 4.49 4.62 0.33 0.13 0.07 0.52 0.26 19.65 4P 0 6 11 16 21 - 5 10 15 20 25 1.44 1.44 1.44 1.44 1.44 0.42 0.38 1.01 0.59 0.32 0.04 0.01 0.09 0.06 0.02 3.02 2.74 7.27 4.25 2.30 0.29 0.07 0.65 0.43 0.14 19.58 6P 7P IW 0 6 11 16 21 - 0 6 11 16 - 0 6 11 16 5 10 15 20 25 1.33 1.33 1.33 1.33 1.33 0.56 1.16 0.70 0.51 0.06 0.05 0.02 0.05 0.06 0.01 3.72 7.71 4.66 3.39 0.40 0.33 0.13 0.33 0.40 0.07 19.88 - 5 10 15 20 1.40 1.40 1.40 1.40 5 10 15 20 1.24 1.24 1.24 1.24 0.28 0.12 0.04 0.00 0.04 0.03 0.01 0.00 1.96 0.84 0.28 0.00 0.28 0.21 0.07 0.00 3.08 1.14 0.03 0.00 0.00 0.11 0.03 0.00 0.00 7.07 0.19 0.00 0.00 0.68 0.19 0.00 0.00 7.26 90 Table 17. Site8 # 2W Cesium-137 Laboratory Data and Areal Concentrations (Continued). Depth Bul k of Density Increment 0 6 11 16 cm g cm"3 - 5 - 10 - 15 - 20 1.26 1.26 1.26 1.26 Mass Concentra tio n pCi 1.59 0.16 0.02 0.00 Counting Error (±) Increment Areal A c tiv ity Total Site A c tiv ity pCi cm'2 g 1 0.10 0.04 0.04 0.00 Error (±) 10.02 1.01 0.13 0.00 0.63 0.25 0.25 0.00 11.16 3W 0 6 11 16 - 5 - 10 - 15 - 20 1.13 1.13 1.13 1.13 2.07 0.29 0.06 0.00 0.12 0.06 0.04 0.00 11.70 1.64 0.34 0.00 0.68 0.34 0.23 0.00 13.68 4W 0 6 11 16 21 - 5 10 15 20 25 1.13 1.13 1.13 1.13 1.13 1.68 0.40 0.05 0.05 0.00 0.09 0.05 0.02 0.03 0.00 9.49 2.26 0.28 0.28 0.00 0.51 0.28 0.11 0.17 0.00 12.31 aSample s ite code; Tl = transect 1,T2 = transect 2, C = channel, P = pond, W = c o n tro l. m APPENDIX C SIEVING RESULTS 92 Table 18. Sieving Results. Soil Mapping Unit Site Non-erodible Particles > 0.84 mm % WEE I Factor t ha"1 yr"1 Nishon Series (28) 4T2 5T2 4C 36.5 43.5 40.5 314 253 275 53.0 45.0 53.1 48.2 37.4 52.6 36.5 49.0 41.5 28.5 52.2 49.5 51.2 32.3 40.0 40.2 44.8 50.8 48.8 35.5 44.8 39.5 40.0 33.2 35.5 41.5 157 242 155 Ethridge Series (3858) ITl 2T1 3T1 4T1 5T1 7T1 1T2 6T2 ITA 2TA 3TA 4TA 5TA 6TA 7TA ITB 2TB 3TB 4TB 5TB 6TB 2TC 2TD 3TD 2TE 3TE 212 307 161 314 206 267 392 166 195 177 354 282 280 244 182 204 321 244 379 282 345 320 267 93 Table 18. Soil Mapping Unit Sieving Results (Continued). Site Non-erodible Particles > 0.84 mm % WEE I Factor t ha"1 yr"1 Ethridge Series (38A) 19T3 25T4 26T4 27T4 28T4 29T4 30T4 51.0 62.0 52.8 52.2 54.8 44.8 44.8 179 96 170 188 139 244 244 45.6 49.9 52.2 39.0 49.5 50.8 35.8 51.7 40.4 48.0 40.6 46.7 39.5 238 193 166 291 202 182 318 170 278 215 276 229 287 Ethridge Series (388) 22T3 23T3 24T3 3TC 4TC 5TC 6TC 7TC 4TD 4TE 5TE 6TE 7TE MONTANA STATE UNIVERSITY LIBRARIES 762 10069392 6