G. W. LANGDALE AND W. D. SHRADER2
Soil erosion always increases the cost of crop production and causespotential environmental hazards as well as human suffering. Erosion of soils by water reduces crop fields principally through the loss of nutrients andavailable water. Exposed subsoils caused by severe soil erosion also exhibitmany adverse properties with respect to soil management for economic cropproduction.
Agronomic implications of soil erosion by water in the United States havebeen derived mainly from limited research on Mollisols, Alfisols, and Ultisols.Because cultivated Ultisols of the southeastern USA are thinner and sufferproblems associated with subsoil acidity, crop yield reductions appear morepermanent and difficult to restore. The permanency of soil erosion on cropyield reductions on many Mollisols soils appears ephemeral, because onlyadditional quantities of N. occasionally P. and micronutrients are required torestore crop yields.
Additional research is urgently needed to quantify crop yield lossesassociated with soil erosion and reduce the cost of restoring crop productionto an economic competitive level on eroded landscapes. Research of this naturecould also provide insights for controlling unacceptable soil erosionlevels.
Soil erosion affects crop production principally by reducing nutrientsupply, water infiltration, and soil water-holding capacity. Adverse soil tilthand aeration may accompany progressive soil erosion. Surface configuration ofthe landscape is also objectionably altered with severe soil erosion. Agronomicimplications of soil erosion research in the United States have been derivedmainly from lands used for the production of corn. soybean (Glycine max L.),cotton (Gossypium hirsutum L.), and small grains, which are all annualcrops.
Early studies on soil productivity and soil erosion concentrated ondiminished nutrient supply, because almost all of the plant available N in thesoil is in the form of soil organic matter, usually concentrated in the surface4 to 12 inches (10 to 30 cm). Also, 50% of the plant-available P is usually inthe organic form (Black. 1968).
Numerous long-term studies in the United States from about 1935 to 1950amply document a trend of reducing crop yield (Adams, 1949: Copley et al.,1944; Hays et al., 1948: Latham, 1940; Musgrave and Norton 1937: Smith et al.,1945; Whitney et al., 1950). Yields of row crops studied during this perioddeclined drastically on all soils as the surface soil was lost by erosionunless soil nutrients, organic matter, and occasionally water were intensivelysupplied. Universally, these studies suggest that when the soil surface waslost, the supply of N and P was drastically reduced and crop yields declined.Fertility inputs and crop yields reported for these studies were much lowerthan current crop production even on eroded soils. Where topsoil depths wereless than 30 cm. crop yields on severely eroded surfaces were reduced 20 to50%.
Studies with only a goal to provide vegetative cover for drasticallydisturbed or eroded nonagricultural lands support findings related toagricultural lands (Bennett, 1939; Parks et al., 1967; Peperzak et al., 1959:Richardson and Diseker. 1961: Vanderlip, 19623). Numerousexperimental sites on roadside backslopes which exposed calcareous loess andglacial till subsoils suggested that nitrifiable N and available P exerted thelargest positive influence on plant growth. A high percentage of either sand orclay in the backslopes in Iowa reduced vegetative cover: possibly because ofadverse water relations and supply. When rates of complete fertilizers wereused, plant growth was achievable to control soil erosion. On drasticallydisturbed kaolinitic soils of the southeastern USA, mulches, lime, fertilizers,and rhizosphere activity are often required to achieve sufficient vegetativecover to control soil erosion. Kudzu (Pueraria thunbergiana) and native pines (Pinus echinata L.) are usually grown more successfully on these lands (Bennett, 1939).
Some rangelands and Pacific Northwest croplands are susceptible to rapidirreversible soil erosion conditions. Attention will be given to these lands inother chapters of this publication.
Shrader et al. (1963) stated that the onsite effects of soil erosion varywith the remedial measure as well as the variation in soil properties. Wintersand Simonson (1951) provide an excellent treatise of desirable and undesirableproperties of exposed subsoil as they affect plant growth. Some knowledge ofexposed subsoil properties is necessary even to suggest remedial treatment forvegetative cover purpose. Individual crops have been known to responddifferently to eroded soils for many decades in the United States (Adams, 1949;Baver, 1950; Buntley and Bell, 1976). The objective in this paper is toevaluate the current yield reduction extent and permanency of soil erosion eastof the Rocky Mountains
Land forming, generalized functions relating plant response to measuredsoil factors, or some combination of these studies provide the best insights todescribe how soil erosion affects crop yields. Soils described in these studiesare not necessarily subject to severe erosion. Henao4 evaluated theeffects of 95 soil factors on corn yields with multiple regression techniquesover a wide range of soil conditions in 17 Iowa counties. Plant-availablewater-holding capacity of the soil was highly correlated with corn yields. Soilerosion estimates were also needed to explain corn yield variation. Soilerosion had an overall negative effect on yields, but there was nodetermination of specific soil effects, such as drainage characteristics, depthof solum, or clay content. This study included such soils as Ida (fine-silty,mixed, mesic Typic Udorthents) and Monona and Marshall (fine-silty, mixed,mesic Typic Hapludolls) which form in the medium textured, deep loess deposits,where only a small decrease in yields due to erosion would be expected undermodern conditions of high fertilizer use Included also were such soils asShelby (fine-loamy, mixed, mesic Typic Argiudolls) or Seymour (fine,montmorillonitic, mesic Aquic Argiudolls) where dense subsoils would markedlydecrease yields.
Thomas and Cassel (1979) measured changes in physical and chemical factorsof the root zone caused by land forming Atlantic Coastal Plain soils.Regression methods were used to account for corn yields produced by these soilfactors. The soil variables ranked in order of declining importance were (1) Ahorizon thickness, (2) plant available P, (3) bulk density, (4) availablewater-holding capacity, (5) organic matter content, and (6) plant-available K.This study included Goldsboro (fine-loamy siliceous thermic Aquic Paleudults),Lynchburg (fine-loamy, siliceous thermic Aeric Paleaquults) and Rains(fine-loamy, siliceous, thermic Typic Paleaquults) soils. Surface soilthickness was important in this study because the subsoil (sandy loam to sandyclay loam textures) was high enough in exchangeable A1 to limit root growth ofmost plants (Table 1). Webb and Bear5 show that soil thickness ofEdina silt loam (fine, montmorillonitic mesic Typic Argialbolls) is of lessimportance in southern Iowa.
In Texas soils, Eck et al (1965), Heilman and Thomas (1961), and Thomas etal (1974) showed that low mineralizable N and available P limited plant growthin land-leveling studies. Their study soil sites were Pullman silty clay loam(fine, mixed, thermic Torrertic Paleustolls) and Hidalgo sandy clay loam(fine-loamy, mixed hyperthermic Typic Calciustolls) with clay and clay loamsubsoil textures, respectively. These soils are not necessarily subject tosevere soil erosion. However, relationships associated with mechanicallytruncated soil properties to plant growth are important to elucidate theeffects of soil erosion on soil productivity.
Deleterious effects of plant nutrient imbalances, soil texture, and bulkdensities are not the only adverse conditions produced by exposing subsoil.Black and Creb (1968) indicated that exposing a Weld silt loam (fine,montmorillonitic, mesic Aridic Paleustolls) subsoil changes the reflectance(color) which causes a temperature interaction that may indirectly affect soilwater storage and ultimately crop yields. Exposed fine textured subsoils alsodecrease water infiltration (Batchelder and Jones 1972). Almost all soilresearchers and farmers elude to tilth, surface configuration, and poor plantpopulation problems on eroded soils. Organic matter, cations, and kind andamount of clay are the principal factors affecting soil aggregation or tilth.Olson (1977) discussed the soil tilth problem as it affected machineryperformance and sparse plant stands on exposed glacial till.
AVAILABILITY OF SOIL EROSION-SOIL PRODUCTIVITY DATA
Reliable crop yield data on eroded soils with a noneroded comparisontreatment is difficult to find in the literature. Randomized statisticalfield-plot designs are not useful tools for measuring crop yield variations oneroded landscape because the random-spatial nature of water erosion. Because ofincreasing crop yields caused by new technologies during the past 30 years inthe United States, intensive soil erosion-soil productivity researchaccomplished between 1935 to 1950 is of little value for predicting cropresponses on eroded soils today. Technology advancements have masked soilproductivity declines due to erosion. Although there is practically noquantification available, it is logical to expect significant crop andlandscape effects. Bennett (1939) and Baver's (1950) classical reviewsadequately describe the gross effects of soil erosion on crop yield during thisearly period.
Since 1950, only two research methods have been used extensively to measurethe effects of soil erosion on soil productivity. The most frequently usedtechnique is the cut and fill method that followed extensive land formingprocesses in the United States since 1950. The other approach is multipleregression analyses applied to random samples of associated crop yield andmeasured soil erosion. Much of this data is available only in dissertationform.3, 4 These approaches are less than desirable to assessadequately the effects of soil erosion on crop yields.
SOIL EROSION-PRODUCTIVITY ASSOCIATED WITH SOIL CLASSIFICATION
A summary of the effects of crop yield reduction due to soil erosion isshown by soil series, texture, and family classification in Tables 2 and 3.These estimates assume that adequate levels of plant nutrients and water weresupplied for optimum plant growth. Further, all of these soils have historiesof erosion problems by natural rainfall. They generally separate with respectto soil productivity as follows: Mollisols and Alfisols vs. Ultisols (SoilSurvey Staff, AH436, 1975).
Except for the Grenada soil of west Tennessee, which is a fragipan soil, itis safe to say the initial crop yield reductions on exposed subsoils ofMollisols-Alfisols are <= 50% those of the Ultisols. These reductions on theformer vary from 5 to 30%. Exceptions among these soils may occur. Some earlyevidence suggests the Shelby loam and associated soils may be the most erosiveof the Mollisols (Bennett, 1939: Smith et al., 1945). Smith et al. (1945)reported small grain and corn yield reductions of 41 and 52% respectively on aShelby loam. No recent data, with high fertilization and improved germplasm,appears available to confirm the effects of soil erosion on the Shelby loam.Equivalent yield reductions on the Ultisols vary from 22 to 47%.
Because cultivated Ultisols of the southeastern USA are thinner and sufferproblems associated with subsoil acidity, there is more recent research ontheir productivity related to soil erosion. Langdale et al. (1979b) used awatershed study to demonstrate the seriousness of a few centimeters of soilerosion on an Ultisol in the Southern Piedmont. At current nonirrigated cornproduction levels, each centimeter of eroded topsoil costs the producer 150kg/ha per year (2.34 bu/acre per year) of corn grain. Soil deposition (localalluvium) did not significantly improve grain yields on the watershed. Some ofthese yields as well as a few reported by Adams (1949), Batchelder and Jones(1972), and Buntley and Bell (1976) are summarized in Table 4a to illustratethe expected yields of variable eroded Ultisols and a problem Alfisol. Thesedata suggest that grain crops (corn, soybeans, and wheat) are more drasticallyaffected by erosion than are cotton and tall fescue (Festuca arundinacea L.).Adams (1949) suggests that vetch (Vicia Suhu L.) growth response exceeded thatof grain and cotton crops on low fertility eroded plots. Baver (1950) alsodiscussed some research on North Carolina soils (Ultisols) that suggests asimple shift from cultivated row crops to forage legumes masks the effects ofsoil erosion. However, the tolerance specificity of agronomic plant species toacid soils is well known today (Foy, 1974). A crop yield survey by Fenton etal. (1971) in Iowa shows that only the most erosive Mollisols and Entisolspossess similar yield reduction trends as the Ultisols (Table 4b).
PERMANENCY OF SOIL EROSION
Bennett (1939) and Baver (1950) both noted that productivity damage by soilerosion was persistent between 1930 to 1950. Soil management technologies suchas high fertilization, improved plant germplasm, and minimum tillage withmulched surfaces were not used in studies of this era. Recently, researchershave frequently noted that crop yields on denuded soils were lower initiallyand improved after several years. Further, only increased quantities of N(mineral or organic) were required for corn growth to equal that on normaltopsoil (Engelstad et al., 1961a, 1961b: Hays et al., 1948; Moldenhauer andOnstad, 1970; Spomer et al., 1973). Moldenhauer and Onstad (1975) used barnyardmanure on 2.1-m deep construction cuts in a deep medium textured loess ofwestern Iowa. After 2 years, their corn yield (100 q/ha) on these cuts equaledthose on the normal surrounding areas. They further state that barnyard manureis desirable but not necessary to restore crop yields on desurfaced loess soilsof Iowa. Engelstad and Shrader (1961a, 1961b) showed that 200 kg N/ha wasrequired initially to produce a 75-q/ha corn yield on a Marshall silt loamsubsoil. Only 125 kg of N was required to obtain the same yield on an unerodedtopsoil. Rosenberry's et al. (1980) energy inputs and soil erosion estimatesindicate that current cost of erosion control to Iowa farmers is greater thaneconomic return from controlling erosion. However, only short-term aggregateeconomic impacts were considered.
Researchers located in Montana (Ruess and Campbell, 1961) and North Dakota(Carlson et al., 1961) all showed that N as well as P and occasionallymicronutrients such as Zn were deficient in soils desurfaced by land leveling.These soils were Keiser clay loam (fine-silty, mixed, mesic UstollicHaplargids) and Gardena fine sandy loam (pachic Udic Haploborolls),respectively. Their nutrient deficiencies were correctable without greatdifficulty.
Batchelder and Jones (1972) found that lime, mulch, and irrigationtreatment were also necessary to restore corn yields on a desurfaced Grosecloseclay loam (clayey, mixed, mesic Typic Hapludults) in western Virginia. Yieldson the desurfaced and normal soils of this study were not equivalent untilafter the 4th year. The research of Phillips and Kamprath (1973) and Thomas andCassel (1979) involving land leveling of Atlantic Coastal Plain soils in NorthCarolina strongly supports the findings of Batchelder and Jones.
Soybean yields were improved with minimum tillage (inrow chisel 23 cm deepthrough a surface wheat mulch) on Southern Piedmont soils (Table 4a b) 6. However, these yields were far below those produced by the sametillage procedure on moderately eroded soil sites. Furthermore, this tillagemethod appears to be of little value where saprolite is within the plowlayer.
On the distant horizon, agricultural technologies appear to be gaining alittle on the permanency of soil erosion. No-tillage (fluted coulter) isbeginning to drastically halt soil erosion and in some cases improve cropyields on lands that are highly erodible (Langdale et al., 1979a: Phillips etal., 1980).
The extreme but not uncommon example on a world scale where erosion revealsbarren rock is so obvious as to require no documentation. On deep mediumtextured soils only additional N and P fertilizers and occasionalmicronutrients are necessary to produce crop yields on eroded soils equivalentto those of uneroded. In most other cases between the above extremes, however,the crop, the soil, and the level of technology to be applied must be specifiedbefore an accurate appraisal of the effect of erosion on crop yield can bemade. In the southeast USA, loss of a few centimeters of surface soil may bevery important because of thin surface topsoils and phytotoxic levels ofexchangeable A1 associated with acid subsoils. Some evidence in the literatureindicates forage crops are more tolerant to eroded southeastern soils thangrain or cotton species.
On some eroded soils, a shift from row crops to legume forages minimizesthe effects of soil erosion. This is mainly due to N fixation during periods oflow evapotranspiration. Minimum tillage also enhances the recovery of some ofthe grain crop yield potential lost to soil erosion through water conservation.In most cases, additional fertilizers, lime, mulches, and irrigation may berequired to restore crop yields to a competitive economic level on erodedsoils. Current resource-product price ratios associated with availabletechnologies probably would inhibit reclamation of most severely eroded soilsfor row crop production. All of these factors impinge on the time periodrequired for agronomic reclamation of eroded soils. Because of therandom-spatial nature of soil erosion, randomize plot designs cannot give anaccurate measure of crop yields lost on eroded soils. This research is alsoexpensive and unrewarding to the soil scientist. To a large extent, data fromcut and fill as well as land resource survey type studies were used to estimatecrop yield reduction due to soil erosion on cultivated lands. These methodsprobably bias our estimates. Winters and Simonson (1951) stated that thescarcity of precise data bearing on the relationships of exposed subsoilproperties to plant growth precludes comprehensive discussion. To a largeextent this is still true almost 30 years later.
Adams, W. E. 1949. Loss of topsoil reduced crop yield. J. Soil WaterConserv. 4:130.
Batchelder, A. R., and J. N. Jones, Jr. 1972. Soil management factors andgrowth of Zea mays L. on topsoil and exposed subsoil. Agron. J.64:648-652.
Baver, L. D. 1950. How serious is soil erosion. Soil Sci. Soc. Am. Proc.42:1-5.
Bennett, H. H. 1939. Soil conservation. McGraw-Hill Book Co., Inc.
Black, A. L. 1968. Conservation bench terraces in Montana. Trans. ASAE11:393-395.
-----, and B. W. Greb. 1968. Soil reflectance, temperature, and fallowwater storage on exposed subsoils of a Brown soil. Soil Sci. Soc. Am. Proc.32:105-109.
Black, C. A. 1968. Soil-plant relationships. John Wiley and Sons, Inc., NewYork.
Buntley, G. J., and F. F. Bell. 1976. Yield estimates for the major cropsgrown on the soils of west Tennessee. Tenn. Ayric. Exp. Stn. Bull. 561. 124p.
Carlson. C. W., D. L. Grumes, J. Alessi, and G. A. Reichman. 1961. CornGrowth on Gardena surface and subsoil as affected by application of fertilizerand manure. Soil Sci. Soc. Am. Prnc. 25:44-47.
Copley, T. L., L. A. Forrest, A. G. McCall, and F. C. Bell. 1944.Investigations in erosion, control and reclamation of eroded land at theCentral Piedmont Conservation Experiment Station. Statesville. N.C. U.S. Dep.Agric. Tech. Bull. No. 873.
Eck, H. V., V. L. Hauser, and R. H. Ford. 1965. Fertilizer needs forrestoring productivity on Pullman silty clay loam after variable degrees ofsoil removal. Soil Sci. Soc. Am. Proc. 29:209-213.
Engelstad, O. P., and W. D. Shrader. 1961a. The effect of surface soilthickness on corn yields: II. As determined by an experiment using normalsurface soil and artificially-exposed subsoil. Soil Sci. Soc. Am. Proc.25:497-499.
---------, O. P., W. D. Shrader. and L. C. Dumenil. 1961b. The effects ofsurface soil thickness on corn yields: I. As determined by a series of fieldexperiments in farmer operated fields. Soil Sci. Soc. Am. Proc.25:494-497.
Fenton, T. E., E. R. Duncan, W. D. Shrader, and L. C. Dumenil. 1971.Productivity levels of some Iowa soils. Spec. Rep. No. 66. Iowa Agric. Exp.Stn. 23 p.
Foy, C. D. 1974. Effects of aluminum on plant growth. p. 601-642. In E. W.Carson (ed.) Plant root and its environment. Univ. Press of Va.,Charlottesville.
Hays, O. E., C. E. Bay, and H. H. Hull. 1948. Increasing production on aneroded loess-derived soil. J. Am. Soc. Agron. 40:1061-1069.
Heilman, M. D, and J. R. Thomas. 1961. Land leveling can adversely affectsoil fertility. J. Soil and Water Conserv. 16:71-72.
Langdale, G. W., A. P. Barnett, R. A. Leonard. and W. G. Fleming. 1979a.Reduction of soil erosion by the no-till system in the Southern Piedmont.Trans. ASAE. 2:82-86. and 92.
---------, J. E. Box, Jr., R. A. Leonard, A. P. Barnett, and W. G. Fleming.1979b. Corn yield reduction on eroded Southern Piedmont Soils. J Soil and WaterConserv. 34:226-228.
Latham, E. E. 1940. Relative productivity of the A horizon of the acidsandy loam and the B & C horizons exposed by erosion. J. Am. Soc. Agron.32:950-954.
Moldenhauer, W. C., and C. A. Onstad. 1975. Achieving specified soil losslevels. J Soil and Water Conserv. 30:166-168.
Musgrave, G. W., and R. A. Norton. 1937. Soil and water conservationinvestigations at the Soil Conservation Experiment Station, Missouri ValleyLoess Region. Clarinda, Iowa. U.S. Dep. Agr. Tech. Bull. 558.
Olson. T. C. 1977. Restoring the productivity of a glacial till soil aftertopsoil removal. J. Soil and Water Conserv. 32:130-132.
Parks, C. L., H. F. Perkins, and J. T. Mays. 1967. A greenhouse study of Pand K requirements for ladino clover establishment on Kaolin strip mine spoil.Georgia Agr. Res. 9:8-10.
Peperzak, P., W. D. Shrader, and O. Kempthorne. 1959. Correlation ofselected soil indices with plant growth on highway backslopes in Iowa. HighwayRes. Board Proc. 38:622-637.
Phillips, J. A., and E. J. Kamprath. 1973. Soil fertility problemsassociated with land-forming in the Coastal Plain. J. Soil and Water Conserv.28:69-73.
Phillips, R. E., R. L. Blevins, G. W. Thomas. W. W. Frye, and S. H.Phillips. 1980. No-tillage agriculture. Sci. 208:1108-1113.
Reuss, J. O., and R. E. Campbell. 1961. Restoring productivity to leveledland. Soil Sci. Soc. Am. Proc. 25:302-304.
Richardson, E.C., and E.G. Disker. 1961. Control of roadbank erosion in thesouthern Piedmont. Agron. J. 53:292-294.
Rosenberry, P., R. Knutson, and L. Harmon. 1980. Predicting the effects ofsoil depletion from erosion. J. Soil Water Conserv. 35:131-134.
Shrader, W.D., H.P. Johnson, and J.F. Timmons, 1963. Applying erosioncontrol principals. J. Soil Water Conserv. 18:195-199.
Smith, D.D., D.M. Whitt, A.W. Zingg, A.G. McCall, and F.G. Bell. 1949.Investigations in erosion control and reclamation of eroded Shelby and relatedsoils at the Conservation Experiment Station, Bethany, Missouri. USDA Tech.Bull. 883.
Soil Survey Staff. 1975. Soil Taxonomy. USDA, Soil Conserv. Ser. AH436. 754p.
Spomer, R., W.D. Shrader, P. Rosenberry, and E.L. Miller. 1973. Levelterraces with stabilized backslopes on loessial cropland in the MissouriValley--A cost-effectiveness stydy. J. Soil Water Conserv. 28:127-131.
Thomas, J.R., M.D. Heilman, and L. Lyles, 1974. Predicting nitrogenfertilizer requirements after land leveling. Agron. J. 66:371-374.
Whitney, R.S., R. Gardner, and D.W. Robertson. 1950. The effectiveness ofmanure and commercial fertilizers in restoring the productivity of subsoilsexposed by leveling. Agron. J. 42:239-245.
Winters, E., and T.W. Simonson. 1951. The subsoil. Adv. Agron.3:1-92.
1 Contribution from Southern Piedmont Conservation Research Center,USDA-ARS. Watkinsvile, GA 30677 and the Dep. of Agronomy: Journal Paper No.J-9600 of the Iowa AGric, and Home Econ. Exp. Stn., Ames, IA 50011.
2 Soil scientist, Southern Piedmont Conservation rEsearch Center, USDA-ARS,Watkinsville, GA 30677: and profesor of agronomy, Iowa State University.
3 Vanderlip, R.L. 1962. Interrelationships of mulches and fertilizers inerosion control on highway backslopes. Unpublished M.S. Thesis, Library, IowaState University, Ames, Iowa.
4 Henao, J. 1976. Soil variables for regresing Iowa corn yields on soilmanagement and climatic variables. Unpublished Ph.D. Dissertation, Library,Iowa State University, Ames, Iowa.
5 Webb, J.R., and C. Bear. 1972. Unpublished data from Southern IowaExperimental Farm, Iowa State University, Ames, Iowa.
6 Langdale, G.W. 1987. Unpublished dtat from the Southern PiedmontConservation Research Center, USDA-ARS, Watkinsville, GA.