Historic, archived document
Do not assume content reflects current
scientific knowledge, policies, or practices.
UNITED STATES DEPARTMENT OF AGRICULTURE
SOIL CONSERVATION SERVICE
H. H. Bennett, Chief
EROSION OF TOPSOIL REDUCES
J. H. Stallings
Soil Conservation Service - Research
Washington 25, D. C.
EROSION OF TOPSOIL REDUCES PRODUCTIVITY
By J. H. Stallings, Research Specialist, Soil Conservation Service
Acre yields have not kept pace with advances in the science of production 2
Large quantities of plant nutrients are removed 3
Erosion depletes soil fertility 3
Water sorts out the most fertile soil constituents 7
Organic matter, nitrogen and fine materials usually the first to be removed 9
Runoff greater on eroded soils 9
Depth of soil an important factor in crop yields 14
Wind erosion most harmful in early stages 17
A-horizon more productive than B- and C -horizons 17
Crop yields higher on uneroded than on eroded soils 18
Terracing affects yields 26
Crop yields higher on topsoil than on subsoil 28
Erosion reduces efficiency of .fertilizer s 29
Conclusion -. . • 29
Literature cited 29
The potential increase in agricultural output resulting from the vast improvements made
in the science of crop production in past years has been offset in large measure by the damage
to the soil resulting from the action of wind and water in the erosion process. The sifting and
sorting action of wind and water separates the organic matter and silt and clay fractions from
the soil mass and carries them from the field. The parts lost contain most of the essential plant
nutrients and other life-giving substances.
Large quantities of these valuable materials may be removed from a field by the erosion
process without entailing a corresponding loss of soil volume from the surface of the land. In
extreme cases the soil is removed bodily. These life-giving substances usually constitute the
first portion of the soil to be removed by erosion. The removal of the soil and'the accompanying
organic matter and plant nutrients by either wind or water erosion results in lowering the pro-
duction potential of the soil.
Vast improvements have been made in the science of crop production during the last 60 to
75 years. Inorganic fertilizing materials have been and are being constantly improved both as
to the content and availability of nutrients and the physical properties of the product. Rates of
fertilizer application have been increased enormously in many areas and the number of farmers
using fertilizer has increased many times over. Better adapted and higher -yielding varieties of
crops have been developed through plant breeding and other work in the plant sciences. Improved
methods of crop and soil management have been made a v a id a b 1 e . Better materials and equip-
ment for control of insect pests and diseases have been developed.
- 1 -
GPO O SCS---196
TABLE 1. — National average annual acre yield of cotton and corn by 10-year periods beginning with
1869 and ending with 19-48
Average annual acre yield
Average for the 9-year period ending with 1947.
ACRE YIELDS HAVE NOT KEPT PACE WITH ADVANCES
IN THE SCIENCE OF PRODUCTION
These and other related developments would have brought about large increases in per-
acre crop production if the fertility of the soil had been maintained. If the productivity of the
Nation's soils that prevailed when the soils were first put under the plow had been maintained
at or near its former level, the average per-acrecrop yields should have shown steady and con-
sistent increases estimated at as much as 40 to 60 percent during the last 50 years for most of
our Drincipal crops. Actually, the increase of some crops has been much less than this. The
national Average acre - yield of corn and cotton, for example, increased but very little up to
1938 as shown in table 1. Since then there has been some increase in the average acre-yields of
both these crops.
The recent increases coincide with a combination of a g r i c u 1 1 u r a 1 advances which were
greatly stimulated as a result of the war effort- -increased spread of conservation farming and,
for corn, the advent of hybrid varieties.
Cotton acreage was greatly curtailed in the old cotton-growing area and expanded into new
territory in the Southwest. In the Southeast, better lands were selected for cotton production,
the rate of fertilization was stepped up appreciably along with a tremendous expansion in the
planting of legume cover-crops and in other improved soil-conserving practices. Perhaps the
advent of hybrid. corn was the greatest single factor contributing to the recent increase in corn
yield. As with cotton, the amount of fertilizer used on the crop during this period was greatly
increased and efforts toward the adoption of better farming practices in general were intensified.
A more or less steady decline in soil fertility has largely, sometimes entirely, offset the
many advances and improvements made along the line of agricultural production. This is illus-
trated by the average per -acre yields of eight of the more important crops grown in Michigan
during the period 1871 to 1940 (table 2).
TABLE 2. — Average annual yields of 8 crops in Michigan for 70 years by 10-year periods, in
bushels per acre 1
Tame hay 2
x Data from L. M. Turke, Department of Soil Science, Michigan State College.
2 Yield expressed in tons per acre.
- 2 -
With the exception of wheat and potatoes, there was no appreciable trend of increase in the
yield of any of these crops in the 70-year period, whereas an actual decline occurred in some
LARGE QUANTITIES OF PLANT NUTRIENTS ARE REMOVED
Analyses of old cropped or eroded Willamette Valley soils in Oregon, when compared with
native sod land, have shown definite reductions in plant nutrients and an increase in soil acidity
(27) 1 . The decreases in nitrogen, calcium and sulphur were as much as 50 percent, and the lime
requirement increased about one-half to three -fourths ton an acre. It is estimated that the soils
of the Willamette Valley sustain a net annual loss of 29, 000 tons of nitrogen and a heavy loss of
potassium, sulphur, calcium and magnesium, ranging from 2, 500 to 106, 000 tons annually. The
annual loss of nutrients is from 2 to 17 times the amounts returned to the land, depending on
the kind of nutrient.
The immensity of plant nutrient losses due to erosion is indicated by the amount of silt and
nutrients carried in the water of the Tennessee River system. For example, on the assumption
that this silt came entirely from the row crop, idle and other land subject to severe erosion, it
has been estimated that the loss from each acre of such land during 1939 would average 5. 2 tons
of silt, 84. 6 pounds of CaO, 97. 9 pounds of MgO, 212. 2 pounds of K 2 0, 1 3. pounds of P 2 5 ,
and 23. 8 pounds of nitrogen (10). Calculated on the basis of the total acreage of the watershed,
the average acre losses of the three bases, as oxides, carried in solution were 167.0 pounds of
calcium, 31. 7 pounds of magnesium and 7. 1 pounds of potassium.
The estimated losses from the land in the entire Mississippi River basin during the same
period averaged 1. 9 tons of silt, 43. 6 pounds of CaO, 53. 8 pounds of MgO, 55. 6 pounds of K ? 0,
5.08 pounds of P,0 , and 6. 64 pounds of nitrogen per acre of open farm land. On the basis of the
total acreage in the watershed the average acre losses were .6 ton of silt, 13.6 pounds of CaO,
16. 9 pounds of MgO, 17. 4 pounds of K 2 0, 1. 59 pounds of P 2 5> and 2. 03 pounds of nitrogen.
The amount of mineral nutrients contained in the drainage waters of the Tennessee River
watershed varies with the nature of the stratum from which the waters flow. The water draining
the limestone areas contains the greatest amountof total mineral matter and that from the sand-
stone areas the least. However, the drainage waters from the sandstone areas contain more than
twice as much potassium as the drainage waters from the limestone areas (20). The amount of
phosphorus in the drainage waters varied with the amount contained in the soils from which the
water flowed. The waters draining the highly phosphatic soils of the bluegrass area contained
the greatest amount of phosphorus. The greatest amount of nitrate -nitrogen was found in drain-
age waters containing the largest amount of soluble phosphorus, thus indicating the close relation
between soluble -phosphorus and nitrate -nitrogen content.
The amounts of nutrients carried annually in solution by the Mississippi and Ohio Rive r s
are shown in table 3 (20).
TABLE 3. — Amounts of plant nutrients carried in solution annually in the Ohio and Mississippi
Ohio River 1
Mississippi River 2
At confluence with the Mississippi River.
2 At Baton Rouge, La.
In addition to the mineral nutrients carried by the Mississippi River in solution, this stream
carried in suspension 7, 469 million cubic feet of soil annually.
l Figures in parenthesis refer to literature cited.
- 3 -
Leaching removes large quantities of plant nutrients from the soil, and much larger
amounts of nutrients are lost by this means from certain productive soils than from less pro-
ductive ones (33). For example, Muscatine soil in Illinois lost 311.4 pounds of calcium per acre
in 3 years and 8 months, or a little more than 27 times as much as Cowden soil which lost 11.5
pounds in the same time. The nitrogen loss during this period was 280.9 pounds per acre on
Muscatine soil as compared with 14. 2 for Cisne soil. Magne sium losses were not as high as
those of nitrogen or calcium but showed the same general relation to soil types. There was only
a slight los s of potash from any of the 10 soil types studied. The greatest lcs s was 5.0 pounds
from Muscatine soil and the smallest was I. 8 pounds from Osceola soil.
EROSION DEPLETES SOIL FERTILITY
Soil erosion has played, and continues to play, a major role in impairing the productive
capacity of the nation's soils. Organic matter, nitrogen, and the clay and silt f r a c t i o n s of the
soil, which contain the life -producing nutrients, are removed by the erosive action of wind and
water. The selecting and sorting action of these agencies separates the lighter materials from
the coarser and heavier sand particles and carries them off, leaving the more inert and less
productive material behind. If erosion is severe, the body of the soil itself is carried off (figs.
1 and 2).
Depletion of fertility in crop land is brought about by the combined action of many factors-
(23). Annual cropping removes large amounts of nutrient materials. Soluble constituents are
lost through leaching processes. Organic -matter decomposition as a result of microbial activity
proceeds at a rapid rate in cultivated soils. In addition to these and many other factors, the
process of erosion is now recognized as one of the most serious forces in the rapid depletion of
fertility and productivity of cultivated lands.
Much experimental evidence is available to show the extent to which erosion carries away
the life -producing part of the soil--the part that contains the nitrogen and mineral plant nutrients
essential to plant growth.
WIND REMOVES THE MOST PRODUCTIVE PART OF THE SOIL
The light-weight particles of soil are the important ones in the great new dry-land winter-
wheat belt of the Southern High Plains (9) . It was the loss of such particles during the dust
storms of the 1930'sthat opened the way for serious inroads on the fertility reserves of the
soils in this area (fig. 3). The first soil drift of 1933 to lodge in a fence row on the Panhandle
Experiment Station at Goodwell, Okla. contained 24.6 percent of organic matter. The drifted
soil had been separated by wind from the surface of topsoil averaging less than 2 percent organic
matter. Removal of the rich topsoil lowered crop yields 4. 5 times as fast as did later removals
of surface and subsoil material.
Each shift of soil by the wind serves to remove more plant nutrients. After the soil is
moved a large number of times, the remaining soil that forms the dunes is mainly sand, regard-
less of the original texture (7). In Oklahoma, after the heavy wind storms of the early 1930's,
the organic matter-nitrogen ratio in the cropped soil was 22.47, that in the virgin soil 23. 30;
and the average of the drifts was 24.44. As a result of cropping and wind erosion, the organic
matter in the cultivated soils was decreased 18 percent and the nitrogen was decreased 15 per-
cent. Very little difference occurred in the nitrogen and organic matter content of the cropped
and virgin subsurface soils.
The wind tends to change the soil texture through removal of the silt fraction and may de-
plete the total fertility of the soil by sifting out the lighter and more fertile portion and carrying
it away (11). Samples of dust collected in Oklahoma during the dust storms of 1930 contained on
the average 62. 5 percent silt and 14. 3 percent sand. The original soil, Richfield silt loam, con-
tained 42 percent silt and 35. 4 percent sand, whereas the drift soil contained 58. 2 percent sand
and only 15 percent" silt. The effect of the sorting action of the wind on the silt and sand content
of the resulting products as compared with that of the original Richfield silt loam is illustrated
in figure 4.
The dust contained 1.77 times as much combustible matter as the field soil and 1 . 47
times as much as the drift soil. The total nitrogen content of the dust was 2.15 times that of field
soil and 1. 88 times that of drift soil. The dust contained 1. 95 times as much phosphorus as the
field soil and 2.04 times as much as the drift soil, and contained 1. 99 times as much base -ex-
change calcium as the field soil.
- 5 -
Figure 2.— Effect of wind erosion. The organic matter, silt and clay particles in the soil
were blown away, leaving the coarser sand to form drifts. The soil beyond the drifts
was blown out to plow depth.
- 6 -
Samples of dust laid down on snow and ice in Iowa by a dust storm originating in the Texas-
Oklahoma Panhandle early in 1937 were collected and compared with samples taken from a
small dune formed by the same wind d i s tu r b a n c e at Dalhart, Texas (4). The dust contained
rouohly 10 times as much organic matter, 9 times as much nitrogen, 19 times as much phos-
phoric acid, and about one and one -half times as much potash as the dune material. Analyses
indicated a similar sorting effect with respect to r e m o v a 1 of both soil particles and chemical
constituents. The unaffected grass-covered soil contained 79. 2 percent coarse materials (total
sands) as compared with no sand in the dust, and 19. 6 percent of fine material (silt and clay) as
compared with 97 percent in the dust. The dust contained more than three times as much organic
matter and nitrogen, respectively, as the virgin soil; nearly five times as much phosphoric
acid; and one and one-quarter times as much potash.
Samples of the dust originating in the 1937 Panhandle storm and deposited at Hays, Kans.
and Clarinda, Iowa were compared with samples of soil of unplowed grassland and with dune
sand collected near Dalhart, Texas (4). The results of the analysis of the samples are shown in
TABLE 4. — Organic matter and partial chemical content of soil of unplowed grassland, dune sand
'Jr.: - •*•• : "r . .
near Dalhart, Tex.
Dune sand, Dalhart,
Hays , Kans .
The data show that the original unplowed soil was much higher in essential plant nutrients
and organic matter than the dune sand but much lower in these materials than the dust collected
at Hays and Clarinda.
WATER SORTS OUT THE MOST FERTILE SOIL CONSTITUENTS
Much fertile material in the soil is lost through erosion (fig. 5). Material eroded from
Collington sandy loam in New Jersey from June 12, 1938 to December 31, 1941 contained four
times as much organic matter, 1.5 times as much phosphorus, 1.4 times as much potassium
and 2. 3 times as much calcium as there was in the soil before erosion occurred (24). The loss
per acre due to erosion was 1, 149 pounds of organic matter, 67 pounds of nitrogen, 154 pounds
of P 2 O 5 , 575 pounds of K 2 and 141 pounds of CaO. There were more than 3. 5 times as many
particles averaging less than 50 microns in diameter in the eroded material as in the surface
soil from which the material came. The eroded material contained 58 percent of materials of
this size-class as compared with slightly less than 16 percent in the original soil.
The material eroded from Dunmore silt loam cropped to corn was 16 percent richer in
total nitrogen and 11 percent richer in phosphorus than the original soil (28). Water-soluble
phosphorus in water extracts of eroded material from corn land contained six to eight times as
much organic phosphorus as was contained in extracts of the parent soil.
A study of 48 depleted soils and the corresponding virgin soils in Michigan showed that the
virgin soil had a greater rate of solubility, as measured by the freezing point method with mois-
ture content somewhat above saturation (21). A decrease in rate of solubility is one of the im-
portant changes a soil undergoes in passing from a virgin to a more or less depleted condition.
This is important since most crop plants feed primarily in the surface or plowed stratum of the
soil and the solubility of subsoils is usually very low compared with that of the surface soils.
The total amount of salts in runoff water from soil erosion plots at Columbia, Mo. during
the year May 1, 1924 to April 30, 1925 ranged from 166.8 pounds per acre for a plot in wheat
and clover to 380. 1 pounds per acre for a plot that was spaded 4 inches deep in the spring and
fallowed throughout the season (8). Calcium and sulphur were lost in larger amounts than any of
the other elements determined. Although the loss of potassium was rather small, the loss of this
element from several plots was much greater than the amount that would ordinarily be applied
in commercial fertilizer.
- 7 -
Figure 3. — Fertility removal in action. The dust carried aloft in the wind erosion pr
contains the most fertile part of the soil.
- 8 -
Soil type and cover had a marked effect on both the amounts and concentrations of the solu-
bles lost in runoff at Geneva, N. Y. during the 13-month period, March, 1938 to March, 1939,
inclusive (5). These effects appeared to be related to variations in soluble concentrations at the
soil surface and to the relative rates of infiltration and runoff. The concentrations tended to be
higher in the summer fnonths. The proportional losses of the separate soluble constituents in
runoff varied considerably. Although losses of solubles reported in runoff were small, an analy-
sis of the factors that produce variability in runoff losses indicates that appreciable losses may
be incurred where poor soil management practices are employed.
ORGANIC MATTER, NITROGEN AND FINE MATERIALS
USUALLY THE FIRST TO BE REMOVED
The annual nitrogen losses from land in Missouri planted to intertilled crops on slopes
averaging 200 feet in length have been found to range from 3. 8 percent of the total amount con-
tained in the surface 7 inches of soil for a 2-percent slope to 11.1 percent for a 12-percent
slope (38). The annual losses on 2-percent slopes ranged from 3. 8 percent for slopes averaging
200 feet in length to 10.9 percent for slopes that averaged 1. 200 feet in length. Corresponding
losses on a 12-percent slope were 11. 1 percent for the 200-foot slope and 18. 1 percent for the
1, 200 -foot slope.
The loss of nitrogen declined with the introduction of sod-producing crops into the rotation
and disappeared altogether on well-sodded meadows or pastures.
Numerous reported depletions of organic matter, formerly attributed to oxidation, mayhave
resulted from erosion (30). Depletion of organic matter appeared to be a linear function of
erosion. The calculated organic matter percentage of the soil dropped 0.002 percent at both
Clarinda, la. and Bethany, Mo. for each ton of soil lost by erosion. The amount of organic
matter removed by erosion is greater than the corresponding depletion indicated by analyses of
the plot soils; consequently, restoration to the original organic matter level does not compensate
for losses of "reserve" organic matter.
It was estimated that erosion removed the organic matter 18 times as fast as did oxidation
from a fallow plot on which the greatest erosion occurred, and that to have maintained the or-
ganic matter at the original level it would have been necessary to apply as much as 9. 2 tons of
clover hay annually.
Marked and significant differences in erodibility occurred in New York under a uniform
treatment, following treatments which permitted great differences in the rate of erosion (12).
The calculated percentage of organic matter in the soil to plow depth was found to have dropped
about 0. 002 percent for each ton of soil lost by erosion.
The losses of organic matter caused by erosion in New York vary both in amount and
character. They tend to be high in proportion to the total amount of soil and the proportion of
fines that are lost (31).
The depth of topsoil has been found to be less important than the selective removal of cer-
tain parts of the soil by the raindrop splash process (18). An 1 1 -year study on four soil types at
Ithaca, N. Y. showed that soil was lost at rates varying from a trace to 138 tons per acre. Only
29 percent of the remaining plow layer passed through a 2-millimeter screen, whereas approxi-
mately 95 percent of the soil that was washed off passed through such a screen.
RUNOFF GREATER ON ERODED SOILS
Studies in Wisconsin showed that a severely eroded soil was not only lower in organic
matter and nitrogen, but lost more rainfall by runoff than did less eroded soils (14). Severely
eroded Fayette silt loam had only one-third as much organic matter and one-half as much nitro-
gen as moderately eroded Fayette silt loam. The severely eroded soil, when planted to grain,
lost through runoff about twice as much rainfall during the growing season as moderately eroded
soil. A severely eroded soil planted to corn allowed 1.3 times as much runoff as moderately
eroded soil, and severely eroded soil planted to hay allowed 2.8 times as much runoff as
moderately eroded soil.
Under severe erosion, eroded materials tend to approximate the composition of theuneroded
soil, and the process is in effect "removal layer by layer" (31). With more moderate runoffs
there is a selective removal of the finer particles. Small local deposits of sand on the soil sur-
face maybe swept off by later rains, but if frequent cultivation constantly p r e s e n t s afresh
surface to the sorting action of running water, a continued removal of the finer particles may be
- 9 -
Original Drift Dust
Figure 4. — Changes in silt and sand content of Richfield silt loam caused
by the sifting and sorting action of wind in the erosion process.
- 10 -
Figure 6.— Effect of past erosion on the yield of corn and oats on Cecil soil,
Watkinsville, Georgia. Average annual yield per acre for the period 1943-47.
- 12 -
Figure 7.— Correlation between depth of topsoil and cotton yield on Cecil soil at WatkinsvilL
DEPTH OF SOIL AN IMPORTANT FACTOR IN CROP YIELDS
A study was conducted with corn in Iowa in 1936 and with corn and oats in 1937 to deter-
mine the relation of soil type to corn and oat yields (22). As a result it was found that yields
vary not only with soil types but also, and more importantly, with depth of soil on the same soil
type. Corn yield increased from 31 to 53 bushels per acre on Tama silt loam in 1936 with in-
crease in depth of surface soil from to 1 2 inches. An even more impressive correlation of
yield and soil depth was obtained in 1937, both with corn and oats. Corn yield in 1937 increased
from 47 to 88 bushels per acre as the depth of topsoil increased from 0-2 to 12 inches. The
average acre-yields of corn on Tama silt loam, as related to depth of topsoil during the 2-year
period 1936-37, and of oats in 1937, are shown in table 5.
TABLE 5. — Effect of depth of topsoil on yield of corn and oats on Tama silt loam
Depth of surface soil,
Bu . per
These data indicate that the depth of topsoil, particularly from 7 inches, has a pronounced
effect on corn yield on Tama silt loam. The results with corn grown on this soil were substantiated
by data obtained on Clarion loam and Clarion fine sandy loam during 1937. Clarion loam which,
after erosion, had less than 7 inches of topsoil remaining, produced 51 bushels of corn per
acre, whereas the same type of soil with over 7 inches of topsoil produced 67 bushels. Corres-
ponding yields for Clarion fine sandy loam were 39 bushels per acre with less than 7 inches of
topsoil and 54 with more than 7 inches of topsoil remaining.
A close correlation existed between the yield of corn, cotton and oats and the depth of top-
soil under certain cropping practices on three A-horizon depths in the 2- to 8-inch range depths
of Cecil soil in Georgia during a 5-year period (16). The yield of cotton was 615.9 pounds per
ac re at the 2 -inch depth, 889 . 1 pounds at the 5 -inch depth and 1 , 019. 6 pounds at the 8-inch
depth. The yields of corn were 18.8, 24.9 and 32.2 bushels per acre for the 2-, 5-, and 8-inch
depths, respectively. The yield of oats was 49.6 bushels per acre for the 2-inch depth, 61.3
bushels for the 5-inch depth, and 69. 7 bushels for the 8-inch depth.
These yields represent an average of a number of crop rotations. The data show an average
annual decline of 64 pounds of seed cotton per acre for each inch of topsoil lost from the 2-inch
to the 8-inch depth. The yield of corn declined an average of 2. 2 bushels per acre for each inch
of topsoil lost and the yield of oats declined 3. 3 bushels per acre for each inch of topsoil lost
within the 2- to 8-inch depth range of topsoil. The yields of corn and oats as correlated with the
depth of topsoil are shown graphically in figure 6.
The yield of cotton grown continuously during a 5-year period at Athens, Ga. ranged from
405 pounds of seed cotton per acre where there was no'topsoil to 985 pounds for the 8.5-inch
depth. The yield for the 6. 5-inch depth of topsoil was 894 pounds and the estimated yield at the
3.5-inch depth was 650 pounds. This gave practically a straight-line average increment of 70
pounds of seed cotton per acre for each inch of topsoil depth. This straight-line relationship is
illustrated in figure 7.
A remarkably close association of yield with soil depth and rotation system was demon-
strated by this study. Cotton grown in the same rotation pattern on land in capability classes
- 14 -
II and III yielded 1, 217 and 1, Z13 pounds of seed cotton per acre, respectively, on the two series
of plots, each of which averaged exactly 7. 6 inches in topsoil depth.
Yields of corn, oats, wheat and second-year hay were obtained over a 2-year period on
various depths of "topsoil for one residual soil group and three glacial soil groups in the Mus-
kingum River watershed in Ohio (3). These yields represent the average of approximately 200
wheat plots selected at random on 24 different farms, 35 hay plots on 8 different farms, 74 oat
plots on 12 different farms, and 81 corn plots on 10 different farms. The fields selected for
study had received uniform treatment, and the crop history of each field during the 3- or 4-year
period prior to sampling was known. Areas were selected within the fields which had different
depths of topsoil, and yields were obtained for these areas.
The data indicate that a reduction in crop yield can be expected as erosion reduces the
depth of the topsoil. This reduction occurs whether a relatively high or low scale of productivity
is maintained. The reductions in yields per acre which can be anticipated for each inch of top-
soil removed by erosion are reported in table 6.
TABLE 6. — Average annual reduction in acre yields of corn, oats, wheat and hay that can be
anticipated for each inch of topsoil lost on glacial and residual soils, Muskingum River
A pronounced relationship between depth of surface soil and grape production was observed
on Chenango soil in Pennsylvania (1). Production varied with depth of surface soil, depth being
the factor which determined the amount of organic matter, nitrogen, and available moisture
contained in the surface layer. By allowing an additional amount of soil per vine, more of these
essential factors for plant growth apparently became available to each plant, and production was
thus increased. These findings are summarized in table 7.
TABLE 7. — Relation between depth of surface soil, pruning weight and yield of grape vines -
average for 1946-47
Depth of surface soil in inches
pounds per vine
Yield tons per acre
Studies were conducted on three watersheds at Coshocton, Ohio in 1941 to determine tne
relation of the depth of topsoil to corn yields. Yield measurements made on 4-, 5-, 6-, 7-, 8-,
and 9-inch depths of topsoil on each watershed, expressed in bushels per acre, were 33. 7, 41. 2,
46.4, 59.9, 51. 1 and 59.5, respectively. These yield data are shown graphically in figure 8.
Erosion has lowered crop yields throughout the country generally, and has resulted in
abandonment of both large and small areas. Some of the abandoned land may have been too shal-
low for cultivation at the time it was broken out, but much of it was reduced in depth by erosion
to the point where it became too shallow for cultivation. Once reduced to a depth insufficient for
adequate water storage for crop growth, such lands are virtually lost to the growing of cultivated
crops except under irrigation, regardless of the inherent productivity of the soil material.
Information was obtained on 989 soil observations and wheat samples taken from the Wild
Horse and Rock Creek sample areas in Oregon in 1939, 1940 and 1941 (34).
On the Wild Horse area the average yield of wheat on soils 48 inches or more in depth was
46 bushels per acre as compared with 22 bushels on soils less than 24 inches in depth. The yield
was 43 bushels per acre where the topsoil was 14 to 18 inches in depth as compared with 13
bushels where the topsoil was less than 6 inches in depth. The results on the Rock Creek area
- 15 -
2 4 6 8 10 12
Figure 8. --Corn yields expressed in bushels per acre for different
depths of topsoil in Iowa and Ohio.
- 16 -
show that yields of 27 bushels were reduced to 20 bushels as the total soil depth changed from
more than 48 inches to less than 24 inches, and that yields of 31 bushels declined to 21 bushels
with a change from 1 5 to less than 10 inches in depth of topsoil.
The loss of an inch of topsoil caused a reduction of 0.9 bushels per acre in the yield of
wheat in the Wild Horse area, and 0.8 bushels in the Rock Creek area. The change in yield due
to erosion was most pronounced on shallow soils, amounting to as much as 1 . 7 bushels per inch
where the topsoil was less than 6 inches deep.
The corn yield for 1943 on non-desurfaced Austin clay at Temple, Texas was 27.6 bushels
per acre , whereas that for de surfaced soil, representing the marly C horizon, was but 16.4
bushels (36). The desurfaced plot had been cropped to row crops during the period 1932-43 with
a green manure crop that was turned under each year. In spite of this treatment, the corn yield
on the desurfaced plot was but 59.4 percent of the yield from the non-desurfaced plot where no
green manure crops were grown or turned under.
WIND EROSION MOST HARMFUL IN EARLY STAGES
Wind erosion has reduced the productive capacity of High Plains wheat lands an average of
2. 2 bushels per acre annually on land in continuous wheat culture and 4. 29 bushels on land in
summer fallow (9) . The studies show that erosion during the cultivated life of a field was 4. 5
times as destructive in the early stages of erosion as in the later stages. Erosion reduced the
yield 0. 52 bushels per acre per year for the first 4 years the land was subject to erosion, and
at the rate of 0. 11 bushels annually for the next 21 -year period. Serious erosion began ordi-
narily 2 to 4 years after the land was first put in cultivation.
Erosion lowered the potential productive capacity of fallow land in line with the annual
rate of production loss. As a result, the land capability was affected in such a way as to forecast
a step-by-step loss of the productive resources of the High Plains if erosion in this area should
continue at substantially the same rate that has prevailed since cultivation began.
A-HORIZON MORE PRODUCTIVE THAN B- AND C -HORIZONS
The A-horizon of naturally eroded Cecil sandy loam produced an average annual acre yield
over a 4-year period of 939 pounds of seed cotton compared with a yield of 304 pounds for the
B-horizon and 81 pounds for the C -horizon (19). The A-horizon plots produced an average annual
yield of 635 pounds more seed cotton per acre than the B-horizon and 858 pounds more than the
C-horizon. The A-horizon was more than 3 times as productive as the B-horizon and 11 times
as productive as the C-horizon.
Manure was applied at the rate of 4 tons per acre to half of the plots on each horizon in
1939, the last year of the 4-year period of study. The yield of seed cotton as influenced by the
addition of the manure is shown in table 8.
TABLE 8. — Effect of the addition of 4- tons of manure per acre on the yield of seed cotton on A-,
B,- and C-horizons of Cecil soil
Yield of seed
cotton per acre
The addition of manure resulted in increased yields on all horizons, but its beneficial ef-
fect was relatively greater on the C-horizon than on the A- and B -horizons . The addition of
manure lifted the yield on the C-horizon plot above that for the unmanured B-horizon plot, but
did not raise the yield of the B-horizon plot to that of the unmanured A-horizon plot.
- 17 -
CROP YIELDS HIGHER ON UNERODED THAN ON ERODED SOILS
Experiments conducted over a 4-year period on Fayette silt loam at La Crosse, Wis. and
on other soil types at other conservation experiment stations show that crop production was sig-
nificantly higher on uneroded than on eroded soil (14). Moderately eroded soil with 5 to 6 inches
of surface soil yielded significantly more corn, small grain and hay than severely eroded soil
which had lost all but 2 inches of its surface soil. This difference in yield was evident even
though all plots were brought up to a similar potash and phosphorus level and planted to a 5-year
rotation including 3 years of hay. The average yield of corn on moderately eroded soil was 92
bushels per acre as compared with 65 bushels on severely eroded soil. Barley produced 21
bushels per acre on moderately eroded soil as compared with 12 bushels per acre on severely
eroded soil. Hay produced 3.5 tons on moderately eroded soil as compared with 3.2 tons on
severely eroded soil. The severely eroded soil produced only 70 percent as much corn, 57 per-
cent as much barley and 91 percent as much hay as the moderately eroded soil.
Field tests were made in Wisconsin on 45 farms located as far north as Barron County
and as far south as Grant County in order to determine if erosion affected production on other
soil types and under other climatic conditions in that State. Slightly eroded, moderately eroded
and severely eroded plots were selected in each field where all three stages of erosion were
present. At least four plots selected at random were harvested for each soil depth, and on most
fields two or three locations of similar soil depth were harvested.
Small-grain yields were obtained on 18 farms of which 15 showed a significantly higher
yield on the less severely eroded soil. Three soil depths were found in the same field on 11
farms. Small-grain on these fields yielded an average of 65 bushels per acre where there was 6
or more inches of surface soil, 50 bushels where there was 4 to 6 inches of surface soil and 43
bushels per acre where there was three inches or less of topsoil. In other words, there was a
reduction of 22 bushels per acre as a result of the loss of 3 inches of surface soil.
Corn yields were obtained on different soil depths on 12 farms. In each instance the yield
of corn was higher on the slightly eroded than on the more severely eroded soil. On 8 of the
farms where all the soil depths were present in the same field, the average yield of corn was
80 bushels per acre where the soil was only slightly eroded, 67 bushels per acre where the soil
was moderately eroded, and 60 bushels per acre where the soil was severely eroded.
These data indicate that the yield of a crop is significantly decreased with loss of topsoil
through erosion, and that depth of surface soil is a major factor influencing yields of crops.
With few exceptions, the yield was correlated directly with the depth of surface soil regardless
of soil type, variety of crop, cultural treatment, fertilizer practice or climatic condition.
A study was conducted in Wisconsin over the 6-year period 1939-44, using 45 farms classi-
fied on the basis of their susceptibility to erosion losses (2). It was found that the shallower the
topsoil, the smaller was the yield of crops. In general, the greater the erosion hazard on a
given soil type, the less was the amount of topsoil remaining. Crop yields of the farms with the
greatest probability of losing their surface soil were as low as 89 percent of the average for all
farms, whereas for the farms with the least probability the yields were as high as 109 percent
of the all-farm average.
The records show that the differences in corn and oat yields among these groups of farms
were highly significant. The differences in hay yields were somewhat smaller. The reduction of
yield on the fields with the greater erosion hazards was only about 1/7 for hay in comparison
with a reduction of from 1 /4 to 1 /3 for the small grain and corn crops. The indexes of crop
yields of farm lands with different erosion hazards for the 6-year period of study are recorded
in table 9.
TABLE 9. — Index of crop yields of farm lands with different erosion hazards
Bu . per
- 18 -
The average annual acre yield of corn on certain soils in Missouri declined from 64 bush-
els with 12 inches of topsoil to 16 bushels where all the topsoil had been eroded away (32).
Yields were obtained on Shelby, Grundy and Mexico soils over the 5-year period 1940-44 on
soils with 0, 2, 4, 6, 8, 10, and 12 inches of topsoil. The average annual acre yields in bushels
for the three soil types studied were 16 bushels where topsoil was lacking, 25 for 2 inches of
topsoil, 38 for 4 inches of topsoil, 46 for 6 inches of topsoil, 54 for 8 inches of topsoil, 60 for
10 inches of topsoil and 64 bushels per acre for 12 inches of topsoil. These yields are shown in
figure 9. Past erosion reduced the yield of soybeans on an average of about 1.79 bushels per
acre per inch of topsoil lost on Shelby soil and 0. 73 bushel on Marshall soil.
Data secured in 1938 by the technical staff of the Greenfield, Iowa Soil C o n s e r v a t i o n
Service demonstration project showed that the average corn yield was lowered about 5 bushels
per acre for each inch that the surface soil was lowered below 8 inches by erosion (35).
The yield of corn in 1939 at Bethany, Mo. ranged from 17. 1 bushels per acre where only
3 inches of topsoil remained to 47. 3 bushels where 13 or more inches of topsoil remained (37).
These were the average yields of four fields where yields were determined for different depths
of topsoil in the same field. A similar study of seven fields during 1940 showed a corn yield of
15.2 bushels per acre where all the topsoil had been removed by erosion as compared with a
yield of 67. 5 bushels per acre where 13 or more inches of topsoil remained.
Both of these studies showed that there was a constant and substantial decline in corn yield
with the loss of each inch of topsoil. The yields are shown graphically in figure 10. The data for
1939 show the yield for each inch of topsoil depth from 3 inches through 13 inches, and those
during 1940 show the yield for each 2-inch depth.
Similar studies were conducted during 1939 and 1940 in Indiana, Iowa and Ohio. In the
studies carried out near Fowler, Ind. , as at Bethany, Mo. the corn yields were determined for
different soil depths in the same field. The studies were made on 16 fields in 1939 and on 18 in
1940. The depth-classes of topsoil ranged from 1 to 13 inches in 1939 and from to 13+ inches
in 1940. There was a consistent decline in the yield of corn for each inch of topsoil lost. The
average yield for the 16 fields in 1939 ranged from 49. 2 bushels per acre for topsoil 1 to 2
inches deep, to 93.4 bushels for the 13+ inch depth. The corresponding yields for the 18 fields
in 1940 ranged from 19. 8 bushels per acre where all the topsoil had been removed by erosion to
69. 5 bushels where 13+ inches of topsoil still remained. In 1939, the 1 - to 2-inch depth of top-
soil produced only 52. 6 percent as much corn as did the soil that was 1 3 inches or more in
depth. In 1940, land without topsoil produced only 28. 4 percent as much as did the soil that was
13 inches or more in depth. These yields are shown graphically in figure 11.
The 1939 yield data for each of the three important soil types mapped in the 16 fields at
Fowler are presented graphically in figure 12. The data show that for comparable depths the
corn yields for the three soil types were fairly consistent, except for the 11. 5- and 13. 5-inch
depths of the Parr silt loam. However, in general there was a substantial decline in the corn
yield on each soil as the depth of topsoil decreased. The Fowler silt loam averaged 51. 9 bushels
of corn per acre on areas where the topsoil was from 3 to 4 inches deep compared with 92.9
bushels for areas having 13 inches or more of topsoil remaining. Similarly, the Parr silt loam
produced an average of 49. 2 bushels of corn per acre where the topsoil averaged 1 to 2 inches
in depth and 112. 5 bushels where the topsoil was 13 or more inches deep. Corresponding yields
for Brookston silt loam were 50. 6 bushels per acre where the topsoil was 3 to 4 inches in depth
and 92.4 bushels where there was 13 or more inches of topsoil.
The 13+ inches of topsoil produced 1.78 times as much corn as the 3- to 4-inch depth of
topsoil on the Fowler silt loam, 1. 82 times as much on the Brookston silt loam, and 2. 28 times
as much as the 1 - to 2-inch depth on the Parr silt loam.
Corn yields obtained on two soil types at Greenfield, Iowa in 1938 and on 1 fields at
Shenandoah, Iowa in 1939 also showed a consistent relation between depth of topsoil and corn
yield. The Tama silt loam produced 31. 8 bushels of corn per acre where the depth of topsoil
was to 2 inches compared with 63. 9 bushels with 11 to 12 inches of topsoil. Corresponding
yields for the Shelby silt loam were 24.7 bushels per acre on the to 2 inches of topsoil and
49. 8 bushels for the 9- to 10-inch depth of topsoil. The average yield for the 10 fields at
Shenandoah was 52. 3 bushels per acre for the 1 - to 2-inch depth of topsoil and 119. 9 bushels for
the 1 1 - to 12-inch depth. The results of the study conducted at Shenandoah are shown graphically
in figure 8.
- 20 -
10 L — ■ 1 — — 1 1 1 — -J — 1 »
2 4 6 8 10 12 14
Figure 11. — The average yield of corn in bushels per acre for different depths of topsoil near
Fowler, Indiana in 1939 and 1940.
- 22 -
2 4 6 8 10 12 14
Figure 12.— Corn yields expressed in bushels per acre for different
depths of topsoii for three soil types near Fowler, Indiana, in 1939.
- 23 -
- 24 -
Approximate evaluations of the principal soil characteristics such as profile, surface
texture, slope and erosion were made for the soils of Smokey Hill and Lyon townships of Geary
County, Kans. , (26). These evaluations were tested and revised by the use of crop-yield data
collected by two methods : (a) a survey of farms to obtain yield and practice data for fields of
small grains, and (b) harvesting samples of corn from the different soils in fields farmed by the
same operator under identical cultural practices.
The results of this evaluation showed that productivity was inversely proportional to the
degree of erosion. The productivity rating ranged from 95 percent for none or slight erosion to
a minimum of 40 percent for very severe erosion, whereas moderate erosion had a rating of 85
percent and severe erosion 70 percent.
It is of interest to consider the composite soil rating in relation to depth of topsoil alone.
This relationship for the 46 samples comprising the basic data is shown in figure 13. It will be
noted that on the average the composite soil rating is about 20 for a topsoil depth of 2 inches.
Where the depth of topsoil is 12 inches the average composite soil rating averages approximately
80, or four times as great.
The basic data on corn yields from the upland soils in the study indicated that in 1936 the
yields averaged about 38 bushels per acre where the topsoil depth was in excess of 6 inches. An
average decrease in yield of approximately 3. 5 bushels per acre appeared to be associated with
a loss of each inch of topsoil after the 6-inch level was reached.
Yields of wheat for 1946, as calculated from the yield and practice survey in Geary County,
Kans. are shown in table 10.
TABLE 10. — Effect of erosion on the yield of wheat in Geary County, Kansas, 1946
Topography and degree of erosion
Level to gently sloping - none to slight erosion
Undulating to sloping - none to slight erosion
do - severe erosion
do - very severe erosion
Bu. per acre
In Yates County, N. Y. , moderately eroded soil produced three times as many grapes as
severely eroded soil, and slightly eroded soil produced five times as many (13). The slightly
eroded soil averaged 14 inches in depth, the moderately eroded 11 inches , and the severely
eroded 5 inches.
The yields of both corn and cabbage were higher on uneroded Dunkirk soil at Geneva, N. Y.
in 1943 than on severely eroded soil of the same type despite the fact that 126 pounds of readily
available nitrogen was applied per acre to the severely eroded soil and none was applied to the
slightly eroded soil. The uneroded soil produced 4. 7 tons of sweet corn and 16. 7 tons of cabbage
per acre in contrast with yields of 3. 2 tons of sweet corn and 13. 8 tons of cabbage per acre for
the severely eroded soil.
Moderately eroded soil (33 inches of soil over shale) in Ontario County, N. Y. produced
38 No. 1 rose stock per hundred feet of row space, whereas severely eroded soil (14 inches of
soil over shale) of the same type and in the same field produced none. The moderately eroded
soil produced 18 No. 11/2 grade stock per hundred feet of row space against 4 for the severely
eroded soil, and 11 No. 2 grade rose stock as compared with 34 for the severely eroded soil.
The detrimental effect of erosion on the yield of rose stock is accentuated by the low commercial
value of the No. 2 rose stock.
Yields of oats were obtained on five widely scattered farms in southeastern Minnesota from
areas that were as nearly uniform as possible with the exception of depth of topsoil (15). Slightly
eroded or uneroded, moderately eroded, and severely eroded plots were selected in each field.
The yield of oats on uneroded soil with 10 or more inches of surface soil was 36. 1 bushels per
acre. The yield on moderately eroded soils, with 5 to 10 inches of surface soil remaining, was
30.0 bushels; and the yield on severely eroded soils, with 5 or less inches of surface soil re-
maining, was 22. 7 bushels per acre. In other words, the uneroded soil produced 6. 1 bushels
more oats per acre than the moderately eroded soil, and 13.4 bushels more per acre than the
- 25 -
severely eroded; and the moderately eroded soil produced 7. 3 bushels more per acre than the
severely eroded soil. Yield data obtained on Bellefontaine silt loam during 1944 from 10 farms
with severe and moderate erosion and from 3 farms with slight erosion show an average yield
of 92. 8 bushels of oats per acre for slight erosion, 74. 6 bushels for moderate erosion and 66. 8
bushels for severe erosion.
Wheat yields on the South Fork of the Palouse Project in Washington ranged from 15.3
bushels per acre where all the topsoil had been removed by erosion to 50 bushels where an av-
erage of 24 inches of topsoil remained (17). The yield was 35. 1 bushels of wheat per acre where
the topsoil depth averaged 10.7 inches and 22.7 bushels for an average topsoil depth of 4.7
inches. These yields were all on land with a south slope and represent the average for a number
of sites over a 3-year period. Corresponding yields for land with a north slope were 18.5 bush-
els of wheat per acre where the topsoil averaged 18. 5 inches in depth as compared with 42.6
bushels where the topsoil depth averaged 27.4 inches. The data are shown graphically in figure
Runoff plots were established at Pullman, Wash, in 1935 on a tract of newly cultivated vir-
gin land and on adjacent land that had been under cultivation for about 50 years. The old culti-
vated land had lost about one-third of the original topsoil by 1934. A cropping system consisting
of winter wheat grown in alternate years with summer fallow was initiated on both sites in 1934.
The average annual yield per acre for the 1 1 -year period, 1936-46, was 38.7 bushels for the
virgin land put in cultivation in 1935 and 22. 8 bushels for the land that had been put in cultivation
some 50 years or more earlier. In other words, the productive capacity of the land put in cul-
tivation about 1885 had declined 15.9 bushels. This decline in production averaged about 0. 33
bushel per acre annually over the 50-year period.
In New Jersey, yields of nine crops were measured over varying periods of time on areas
with less than 6 inches of topsoil and on other areas with more than 6 inches to determine the
relation between yields and the depth of topsoil (25). There was a significant difference in yields
in every case in favor of the greater depth of topsoil. The results are shown in table 11.
TABLE 11. — Crop yields per acre for various depths of topsoil, New Jersey
Depth of topsoil
Studies conducted during each of 8 years with corn, soybeans, oats, rye, barley and al-
falfa in New Jersey (28) showed that where all factors other than erosion were constant, the
yields of all of these crops from eroded areas were less than those from comparatively uneroded
areas. Data collected during 1948 for wheat and potatoes show the same results. Wheat yields
from eroded areas averaged 25 bushels per acre in contrast with 46 bushels for uneroded areas
in the same fields. Differences in potato yields were even greater. Eroded areas averaged 124
bushels while uneroded portions of the same fields averaged 383 bushels per acre.
TERRACING AFFECTS YIELDS
Corn yields at Bethany, Mo. were affected by the unequal distribution of the mantle of top-
soil as a result of field terracing (39). Using the yield obtained on the top of the terrace ridge
as a basis of comparison, the yield increased to a maximum at a point just off the terrace back-
slope, or approximately 15 feet down the slope from the ridge center. It declined progressively
from this point down the slope to the center of the terrace channel. The yield increased from the
channel to a point approximately one -half the distance to the ridge top of the front slope of the
terrace berm. It again showed a decline where the surface soil on the remainder of the front
slope of the terrace was more than 1 foot deep.
- 26 -
Figure 14.— Wheat yields correlated with the depth of topsoil,
South Fork, Palouse Demonstration Project. 1936-39.
- 27 -
The yield of corn increased with depth of surface soil, regardless of its location on the
terrace, to surface soil depths up to 1 foot. Soil depths above this figure occur on the terrace
ridge, and a slight decline in yield occurred on these extra depths of filled-in surface soil.
CROP YIELDS HIGHER ON TOPSOIL THAN ON SUBSOIL
A study was initiated in 1936 near Wooster, Ohio to determine the relative crop production
of topsoil and subsoil (37). Three series of plots were used. The topsoil was stripped off down
to the subsoil on one series of plots. This topsoil was spread over a second series of plots where
the original topsoil was undisturbed, thus doubling the depth of topsoil. On a third series of
plots the original topsoil was left undisturbed. Corresponding plots in each of the three series
were planted to the same rotations and given identical treatment otherwise.
The average annual yield of corn, oats, wheat and hay per acre on these plots for all treat-
ments for the 9-year period 1937-45 are shown in table 12.
TABLE 12. — Effect of depth of topsoil on the yield of corn, oats, wheat and hay at Wooster, Ohio
over a 9-year period
Average annual plot yield
1 Average for 1937-47.
The yield of each kind of crop except oats was higher on the double topsoil than on the vir-
gin topsoil. For oats, the average annual yield was some 5 bushels per acre higher on the
virgin-topsoil plots than on the double-topsoil plots. The yields were much lower in every case
on the subsoil plots than on the virgin-topsoil plots. On the subsoil the corn yield was only 48
percent of that for the virgin topsoil. Corresponding percentages for the other crops were 37
for oats, 42 for wheat and 74 for hay.
Untreated soil from which about one-half the topsoil had been removed by erosion yielded
more than four times as much grain as untreated subsoil at Bethany, Mo. over a 7-year period
(37). Similar soils limed at the rate of 3 tons per acre in the beginning and fertilized with 200
pounds of superphosphate per acre produced 1.8 times as much grain where one-half of the top-
soil still remained as the subsoil that received the same treatment. The addition of 8 t o n s of
manure per acre in addition to the fertilizer produced 1.82 times as much corn on the area
which had about one-half of the topsoil remaining as was produced on the subsoil.
A study similar to that reported at Wooster was started at Columbus, Ohio in 1938. The
average annual acre yield of corn for all treatments in the three series of plots for the 8-year
period 1938-45 was 42.9 bushels for normal topsoil, 18.2 bushels for subsoil and 67.0 bushels
for the double topsoil. The addition of 500 pounds of 0-14-6 fertilizer per acre and enough lime-
stone to bring the pH to 6. 5 changed the yields to 48 bushels per acre for normal topsoil, 20. 1
for subsoil and 65. 2 for double topsoil.
A study was made in California of the influence of soil horizons A, B, and C on the rate
of growth of certain annual plants which dominate early successional stages, compared with
certain perennial herbs recognized as stable or climax in grassland communities, and on the
comparative plant development in horizons A, B, and C (29).
The growth rate of both the annual and the perennial species was appreciably greater in
horizon A than in lower horizons, regardless of the soil series or the species used. Likewise,
the amount of plant material produced in horizon A was consistently greater than in horizon C,
regardless of species or soil series. This held true, also, in horizon B with the exception of
Olympic soil, in which the B-horizon proved nearly as productive as the A-horizon for two
annual species. This deviation from the average trend may be accounted for by the fact that the
soil used to represent the Olympic series was produced in grassland formation; hence, the B-
horizon may have been subject to greater accumulation of materials from the upper horizon than
in the forest soil.
- 28 -
These studies indicate the importance of keeping intact the A-horizon soil layer, which
according to preliminary studies of nutrient displacement, is richer in both total nitrogen and
nitrate nitrogen than the underlying horizons. The removal of the A-horizon tends not only to
decrease the luxuriance of growth of the vegetation but to greatly retard, if not actually prevent,
the reestablishment of the climax and subclimax plant cover.
Greenhouse studies of lettuce grown in topsoils and subsoils of some of the prevalent soil
types in California gave irrefutable evidence that the topsoils consistently out-yielded the sub-
EROSION REDUCES EFFICIENCY OF FERTILIZERS
Studies in New York showed that cropping and cultural practices that resulted in sheet
erosion rapidly led to inefficient use of fertilizer (18). Corn was grown on four of the leading
soil types of the State which were cropped uniformly and fertilized with 1,000 pounds of 10-10-10
fertilizer per acre. The per-acre yields ranged from 17 to 88 bushels on Bath flaggy silt loam
at Ithaca; 62 to 100 bushels and 54 to 82 bushels, respectively, on Ontario sandy clay loam and
Dunkirk silty clay loam at Geneva; and from 49 to 69 bushels on Honeoye silt loam at Marcellus
even after two years of alfalfa-clover-timothy hay.
Along with the decline of productivity chargeable to removal of soil by erosion has gone a
lesser but more steady decline of fertility from the land due to crop removals. The decline of
productivity in the Southern High Plains due to erosion was five times as great as that due to
crop removals during the first 4 years of cultivation (9). It was only 1. 5 times as great during
the next 21 years of cultivation. The total decline in productivity of land that had been in culti-
vation 15 years or more, due to both causes, amounted to 7. 01 bushels of wheat per acre.
Soil erosion has played, and continues to play, a major role in destroying the productive
capacity of the Nation's soil. In fact, the erosion process is now recognized as the most wide-
spread and destructive agent involved in bringing about the rapid depletion of the fertility and
productive capacity of cultivated land. This is accomplished through the removal of the organic
matter, nitrogen, and clay and silt fractions, which are the life of the soil, as well as through
the physical removal of the soil body itself.
The selective and sorting action of wind and water in the erosion process under some con-
ditions may remove organic matter and essential plant nutrients without producing a significant
net loss of soil from the surface of the land. This process leaves the soil in an infertile state.
In order to restore the soil to its original state of fertility the plant nutrients and organic matter
thus removed must be replenished. This, if physically possible of accomplishment, would in-
volve tremendous expense. In many instances, replacing the lost material would be completely
out of the question.
Wind and water action in its most violent forms not only results in the loss of the plant
nutrients and organic matter but in the physical removal of the soil mass. This is important
since on most soils the thickness of the surface soil, or A-horizon, determines the amount and
quality of growing space available for root development for most cultivated crops. Thi-s in turn
determines the potentialities of the remaining soil for crop production.
1. Aldefer, R. B. and Fleming, H. K. Some factors influencing grape production on well-
drained lake terrace areas. Pa. Agr. Exp. Sta. Bui. 495. 1948.
2. Anderson, H. O. Crop yields vary with erosion conditions. USDA-SCS and Wis. Agr. Exp.
Sta. , La Crosse, Wis. Mimeo. 1946.
3. Barre, R. D. Effect of erosion on crop yields. Supplement to Muskingum Watershed Ohio
Survey Report, runoff and water retardation and soil erosion prevention for flood control
purposes. Appendix X. 1939.
4. Bennett, H. H. Soil conservation. McGraw-Hill Book Co. , Inc. , 1939.
- 29 -
5. Bryant, Jay C. ana Slater, C. S. Runoff water as an agent in the loss of soluble materials
from certain soils. ' Iowa State Coll. Jour, of Sci. , 22:3, 260-297. 1948.
6. Cosby, S. W. Nutrient level determinations by greenhouse pot studies. Soil Conserv.
Service-190, 191 and 192. July 15, 1947.
7. Daniel, Harley A. and Langhan, Wright H. The effect of winu ciuaion and cultivation on
the nitrogen and organic matter content of soils in the Southern High Plains. Amer. Soc.
Agron. Jour. 28:587-596. 1936.
8. Duley, F. L. The loss of soluble salts in runoff water. Soil Sci. , 21:401-409. 1926.
9. Finnell, H. H. Monthly Progress report, Soil Conserv. Serv. Oct. 1949.
10. Fippin, Elmer O. Plant nutrient losses in silt and water in the Tennessee River system.
Soil Sci. , 60:223-239. 1945.
11. Fly, Claude. A preliminary report of the chemical and mechanical analyses of dust de-
posited by wind at Goodwell, Oklahoma. Panhandle Agr. Exp. Sta. Bui. 57. 1935.
12. Free, George R. Evidence of the effect of erosion on the organic matter and erodibility of
Honeoye soil. Amer. Soc. Agron. Jour., 38:207-217. 1946.
13. , Carleton, E. A. , Lamb, John, Jr. , and Gustafson, A. F. Experiments in
the control of soil erosion in central New York, N. Y. Agr. Exp. Sta. (Cornell) Bui. 831.
14. Hays, O. E. and Muckenhirn, R. J. The effect of depth of surface soil and contouring on
crop yields. Soil Conserv. Serv. and Minn. Agr. Exp. Sta. Mimeo. ,1944.
15. , and Rost, C. O. The effect of surface soil and contouring on crop yields in
Minnesota. Soil Conserv. Serv. and Minn. Agr. Exp. Sta. Mimeo. 1944.
16. Hendrickson, B. H. Review of Prinicpal Results - 1947. U. S. Dept. Agr. , Soil Conserv.
Serv. Watkinsville, Ga. Multilithed. 1948.
17. Kaiser, Verle G. Wheat yields as affected by soils, slope and erosion. Soil Conserv. Serv.
Handbook of Colfax Work Group, Oregon.
18. Lamb, John, Jr. , Carleton, Everett A. , and Free, George R. Effect of past management
and erosion on fertilizer efficiency. Paper presented at Soil Sci. Soc. Amer. Meeting.
November 17-20, 1947.
19. Latham, Earle E. Relative productivity of the A horizon of Cecil sandy loam and the B and
C horizons exposed by erosion. Amer. Soc. Agron. Jour. 34:12. 1940.
20. McHargue, J. S. and Peter, A. M. The removal of mineral plant-food by natural drainage
waters. Ky. Agr. Exp. Sta. Res. Bui. No. 237. 1921.
21. Millar, C. E. Studies on virgin and depleted soils. Soil Sci. 16:433-448. 1923.
22. Murray, William G. , Englehorn, A. J. and Griffin, R. A. Yield tests and land valuation.
Iowa Agr. Exp. Sta. Res. Bui. 262. 1939.
23. Neal, O. R. Erosion control aids fertility maintenance. N. H. Agr. Exp. Sta. , Crops and
Pastures, '6:4. 1942.
24. . The influence of soil erosion on fertility losses and on potato yields. Amer.
Potato Jour. , 20:57-64. 1943.
25. , and Brill, B. D. Annual report of research in methods of soil and water
conservation in New Jersey. Soil Conserv. Serv. 1948.
26. Pine, Wildred H. Methods of classifying Kansas land according to economic productivity.
Thesis submitted in partial fulfillment of the requirements for a doctor's degree. Univ. of
- 30 -
27. Powers, W. L. , Jones, J. S. and Ruzek, C. V. Composition, rating and conservation of
Willamette Vallev soils. Ore. Agr. Exp. Sta. Bui. 365. 1939.
28. Rogers, H. T. Plant nutrient losses from a corn, wheat, clover rotation on Dunmore silt
loam. Soil Sci. Soc. Amer. Proc. , 6:263-271. 1941.
29. Sinclair, Jesse D. and Sampson, Arthur W. Establishment and succession of vegetation on
different soil horizons. Univ. of Calif. Hilgardia, 5:7. 1931.
30. Slater, C. S. and Carleton, E. A. The effect of erosion on losses of soil organic matter.
Soil Sci. Amer. Proc, 3:123-128. 1938.
31. . Variability of eroded material. Jour. Agr. Res. 65:209-219 1942.
32. Smith, Dwight D. The effect of contour planting on crop yields and erosion. Amer. Soc.
Agron. Jour. i3:810-819. 1946.
33. Stauffer, R. S. Runoff, percolate and leaching losses from some Illinois soils. Amer.
Soc. Agron. Jour. 34:830-835. 1942.
34. Thomas, H. L. , Stephenson, R. E. , Freese, Carl R. , Chapin, Roy W. , and Huggins,
W. W. The economic effect of soil erosion on wheat yields in eastern Oregon. Ore. Agr.
Exp. Sta. Cir. 157. 1943.
35. Uhland, R. E. Field method of evaluating effects of physical factors and farm management
practices on soil erosion and crop yields. Soil Sci. Soc. Amer. Proc. , 5:373-376. 1940.
36. Rebuilding eroded soil is a slow process. Soil Conservation. 9:276-279. 1944.
37. . Crop yields lowered by erosion. U. S. Dept. Agr. SCS-TP-75. 1949.
38. Wooley, J. C. Fertility losses as a basis for erosion control planning. Agr. Engin. 24:377-
39. Zingg, A. W. and Whitt, D. M. Distribution of corn yields on farm terraces on the Shelby
soil. Agr. Engin. 23:126-178. 1942.
- 31 -
NATIONAL AGRICULTURAL LIBRARY