660/2-74-043
1974
Environmental Protection Technology Series
Prediction of Subsoil Erodibility Using
Chemical, Mineralogical and Physical
Parameters
I
55
\
S3SZ
ro
-z.
LU
C3
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development,
Environmental Protection Agency, have been grouped into five
series. These five broad categories were established to
facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface
in related fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed
to develop and demonstrate instrumentation, equipment and
methodology to repair or prevent environmental degradation
from point and non-point sources of pollution. This work
provides the new or improved technology required for the
control and treatment of pollution sou'rces to meet environmental
quality standards.
This report has been reviewed by the Office of Research and
Development. Approval does not signify that the contents
necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
-------
June
PREDICTION OF SUBSOIL ERODIBILITY
USING CHEMICAL, MINERALOGICAL
AND PHYSICAL PARAMETERS
by
Charles B. Roth
Darrell W. Nelson
Mathias J. M. Romkens
Project No. 15030 HIX
Program Element 1BB042
Reap/Task PEMP 03
Project Officer
Ronald D. Hill
Industrial Waste Research Laboratory
National Environmental Research Center
Cincinnati, Ohio 45268
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price $1.90
-------
ABSTRACT
This report presents evidence that the surface soil erodibility pre-
diction nomograph (Wischmeier et al., 1971) which uses terms involv-
ing soil particle size, organic matter, structure and permeability,
could not be improved upon by consideration of other mineralogical and
chemical parameters. However, the surface soil erodibility nomograph
did not adequately predict the soil erodibility factor, K, of high
clay subsoils studied in the field under simulated rainfall conditions
as a part of this project. A multiple linear regression equation and
nomograph were developed which can be used to estimate the erodibility
factor, K, of many high clay subsoils. The subsoil erodibility nomo-
graph uses terms involving soil particle size distribution and the
amount of amorphous hydrous oxides of iron, aluminum, and silicon in
the soil. Multiple regression analysis revealed that amorphous iron,
aluminum and silicon hydrous oxides serve as soil stabilizers in sub-
soils, whereas, organic matter is the major stabilizer in surface soils.
Evidence is presented to show that soil erodibility from semi-compacted
fill and scalped subsoil surface conditions were essentially identical.
It is reported that the scalped condition is the best standard soil
surface to base the calculation of the erodibility factor for subsoils.
It is suggested that a soil-management factor should replace the
cropping-management factor in the Universal Soil-Loss Equation when the
Equation is used to predict subsoil erosion.
This report, number 5460 of the Agricultural Experiment Station, Purdue
University, is submitted in fulfillment of Project Number 15030 HIX,
Contract Number 6709, by Purdue Research Foundation under the sponsor-
ship of the Environmental Protection Agency. Work was completed as
of December 1973.
ii
-------
CONTENTS
Page
Abstract ii
List of Figures iv
List of Tables ix
Acknowledgments ...... xi
Sections
I Conclusions 1
II Recommendations ..... 2
III Introduction 3
IV Field Experiments 6
V Laboratory Characterization of Reference Soils
for Physical, Chemical, and Mineralogical Prop-
erties 54
VI Statistical Analysis of Data Obtained in Field
and Laboratory Experiments 65
VII A Nomograph for Estimating the Erodibility
Factor, K, of High Clay Subsoils 74
VIII References 80
IX Appendices 82
iii
-------
FIGURES
No. Page
1 Plot view of the semi-compacted fill treatment on the
McGary subsoil before rainstorm tests 18
2 Plot view of the scalped treatment on the McGary sub-
soil before rainstorm tests 18
3 Plot view of the tilled treatment on the McGary sub-
soil before rainstorm tests 18
4 Close-up view of the semi-compacted fill treatment on
McGary subsoil before rainstorm tests 19
5 Close-up view of scalped treatment on McGary subsoil
before rainstorm tests 19
6 Close-up view of tilled treatment on the McGary sub-
soil before rainstorm tests 19
7 Surface condition of tilled treatment on McGary sub-
soil after 63.5 mm (2.5 inches) of artificial rain 20
8 Close-up view of scalped treatment on Portageville
subsoil after 127mm (5 inches) of artificial rain 20
9 Close-up view of tilled treatment on Portageville
subsoil after 127mm (5 inches) of artificial rain 22
10 Plot view of tilled treatment on Portageville subsoil
after 127mm (5 inches) of artificial rain 22
11 Plot .view of scalped treatment on Portageville subsoil
after 127mm (5 inches) of artificial rain 22
12 Plot view of scalped treatment on St. Clair subsoil
before rainstorm tests 23
13 Plot view of tilled treatment on St. Clair subsoil
before rainstorm tests 23
iv
-------
FIGURES (continued)
No. Page
14 Plot view of semi-compacted fill treatment on St. Clair
subsoil before rainstorm tests 23
15 Close-up view of the scalped treatment on St. Clair
subsoil before rainstorm tests 24
16 Close-up view of the tilled treatment on St. Clair
subsoil before rainstorm tests 24
17 Close-up view of the semi-compacted fill treatment
on St. Clair subsoil before rainstorm tests 24
18 Sediment load of runoff from the compacted fill
treatment on the St. Clair subsoil 26
19 Close-up view of the tilled treatment on St. Clair
subsoil 17 hours after the initial 63.5 mm (2.5 inch)
artificial rainstorm 27
20 Plot view of the tilled treatment on St. Clair sub-
soil after 63.5 mm (2.5 inches) of artificial rain 27
21 Plot view of the tilled treatment on St. Clair sub-
soil after 127mn (5 inches) of artificial rain 27
22 Site area of Wymore subsoil during site preparation 28
23 Plot view of the tilled treatment on Wymore subsoil
before application of artificial rain 29
24 Close-up view of the tilled treatment on Wymore
subsoil before application of artificial rain 29
25 Plot view of the tilled treatment on Wymore subsoil
immediately after 63.5mm (2.5 inches) of artificial rain 31
26 Plot view of the tilled treatment on Wymore subsoil
24 hours after the initial 63.5mm (2.5 inch) artificial
rainstorm 31
v
-------
FIGURES (continued)
No.
27 Close-up view of the tilled treatment on Wyraore sub-
soil 24 hours after the initial 63.5mm (2.5 inch)
artificial rainstorm 31
28 Plot view of the scalped treatment on Wymore subsoil
before the rainstorm tests 32
29 Close-up view of the scalped treatment on Wymore sub-
soil before the rainstorm tests 32
30 Sediment load of runoff from the scalped treatment on
the Wymore subsoil 33
31 Close-up view of the tilled treatment on Wymore sub-
soil one week after rainulator tests 34
32 Tilled treatment area on Pawnee subsoil during site
preparation 35
33 Scalped treatment area on Pawnee subsoil during site
preparation 35
34 Close-up view of the tilled treatment on Pawnee sub-
soil before rainstorm tests 37
35 Plot view of the tilled treatment on Pawnee subsoil
before rainstorm tests 37
36 Close-up view of the tilled treatment on Pawnee sub-
soil after 18mm (0.71 inches) of artificial rain 37
37 Plot view of the tilled treatment on Pawnee subsoil
after 63.5mm (2.5 inches) rainstorm 38
38 Plot view of the scalped treatment on Pawnee subsoil
before rainstorm tests 38
39 Close-up view of the scalped treatment on Pawnee
subsoil before rainstorm tests 38
vi
-------
FIGURES (Continued)
No. Page
40 Plot view of the scalped treatment on Pawnee subsoil
17 hours after 63.5 mm (2.5 inches) of artificial rain 38
41 Close-up view of the tilled treatment on Pawnee sub-
soil about 2 weeks after rainulator tests 39
42 Tilled treatment area on Mayberry subsoil during
site preparation 40
43 Scalped treatment area on Mayberry subsoil during
site preparation time 40
44 Scalped treatment area on Mayberry subsoil after
8 weeks of natural weathering 40
45 Plot view of the scalped treatment on Mayberry
subsoil before rainstorm tests 41
46 Close-up view of the scalped treatment on Mayberry
subsoil before rainstorm tests 41
47 Aggregate-size distribution for aggregates on Wymore
and Mayberry subsoil before rainstorm tests 42
48 Sediment load of runoff from the scalped treatment on
the Mayberry subsoil 44
49 Plot view of the tilled treatment on Mayberry subsoil
before rainstorm tests 45
50 Close-up view of the tilled treatment on Mayberry sub-
soil before rainstorm tests 45
51 Plot view of the tilled treatment on Mayberry subsoil
after 188mm (7.4 inches) of rain 46
52 Close-up view of the tilled treatment on Mayberry
subsoil after 188mm (7.4 inches) of rain 46
vii
-------
FIGURES (continued)
No.
53 Plot view of the scalped treatment on Mayberry subsoil
after 188mm (7.4 inches) of rain 47
54 Close-up view of the scalped treatment on Mayberry sub-
soil after 188mm (7.4 inches) of rain 47
55 Nomograph for estimating the credibility factor, K, of
high clay subsoils 75
viii
-------
TABLES
No.
1 Summary of Dates of Site Preparation and Rainulator Test 9
2 Observed Soil Loss and Infiltration on Subsoils During
Successive Rainstorms 10
3 Observed Soil Loss and Infiltration on Nebraska Surface
Soils During Successive Rainstorms 15
4 Observed and Predicted Soil Erodibility Factors for
Tilled Subsoils 49
5 Observed and Predicted Soil Erodibility Factors for
Scalped Subsoils 50
6 Observed and Predicted Soil Erodibility Factors for
Surface Soils 52
7 Physical Characteristics of Soils Used in This Study 58
8 Chemical Composition of Soils Used in This Study 60
9 Clay Mineralogical Composition of Soils Used in This
Study (percent of the clay fraction) 62
10 Estimates of Parameters in the Simple Linear Regression
Equation of Kobs with Analyzed Soils Variables 66
11 Estimates of Partial Regression Coefficients in the
Multiple Linear Regression Equation of Kobs with
Soil Variables 68
12 Estimates of Partial Regression Coefficients in the
Multiple Linear Regression Equation of Kobs with
Surface Soil Variables 70
13 Estimates of Partial Regression Coefficients in the
Weighted Multiple Linear Regression Equation of Kobs
with Surface and Subsoil Variables 72
ix
-------
TABLES (continued)
No. Page
14 Estimates of Partial Regression Coefficients in the
Multiple Linear Regression Equation of Kobs with Sub-
soil Variables 73
15 Comparison of the Soil Erodibility Factor, K, Deter-
mined in Field Experiments and Those Computed from
the Subsoil Nomograph 78
16 Correlation Matrix of Surface Soil Variables (co-
efficient of correlation - r) 106
17 Correlation Matrix of Subsoil Variables (coefficient
of correlation - r) 109
-------
TABLES
No. Page
1 Summary of Dates of Site Preparation and Rainulator Test 9
2 Observed Soil Loss and Infiltration on Subsoils During
Successive Rainstorms 10
3 Observed Soil Loss and Infiltration on Nebraska Surface
Soils During Successive Rainstorms 15
4 Observed and Predicted Soil Erodibility Factors for
Tilled Subsoils 49
5 Observed and Predicted Soil Erodibility Factors for
Scalped Subsoils 50
6 Observed and Predicted Soil Erodibility Factors for
Surface Soils 52
7 Physical Characteristics of Soils Used in This Study 58
8 Chemical Composition of Soils Used in This Study 60
9 Clay Mineralogical Composition of Soils Used in This
Study (percent of the clay fraction) 62
10 Estimates of Parameters in the Simple Linear Regression
Equation of Kobs with Analyzed Soils Variables 66
11 Estimates of Partial Regression Coefficients in the
Multiple Linear Regression Equation of Kobs with
Soil Variables 68
12 Estimates of Partial Regression Coefficients in the
Multiple Linear Regression Equation of Kobs with
Surface Soil Variables 70
13 Estimates of Partial Regression Coefficients in the
Weighted Multiple Linear Regression Equation of Kobs
with Surface and Subsoil Variables 72
ix
-------
I. CONCLUSIONS
The nomograph developed by Wischmeier et al. (1971) to estimate credibility
factor, K, of surface soils could not be statistically improved upon even
considering the large number of chemical and mineralogical parameters
investigated as a part of this study.
The nomograph developed by Wischmeier et al. (1971) to predict the soil
erodibility factor, K, of surface soils does not adequately predict the
erodibility factor of high clay subsoils.
The number of observations on subsoils was too small to develop a univer-
sally applicable model for predicting subsoil erodibility. However, the
observations made as a part of this study did allow the development of a
nomograph which can be used to predict the erodibility factor, K, of high
clay subsoils with very slow permeability and blocky or massive structure
containing amorphous iron and aluminum hydrous oxides.
The average soil losses and infiltration rates were similar on the scalped
and semi-compacted fill treatments on subsoils.
The scalped subsoil surface was chosen as the standard soil surface condi-
tion to be used for determining soil erodibility factors on subsoils.
A soil-management factor was introduced to describe the scalped subsoil
treatment. This factor should replace the cropping-management factor in
the Universal Soil-Loss Equation when used to predict subsoil erosion.
-------
II. RECOMMENDATIONS
The soil erodibility prediction nomograph developed for high clay sub-
soils is restricted in that it can be used only on subsoils with massive
or blocky structure and very slow permeability. Since Wischraeier et al.
(1971) has shown structure and permeability to be important factors in
soil erodibility prediction of surface soils, it is highly probable that
these factors are important in subsoils. It is recommended that more
soil erodibility measurements be made on subsoils having structures other
than massive or blocky and permeabilities other than very slow.
An important term in the proposed model to predict high clay subsoil
erodibility is the amount of citrate-dithionite extractable iron and
aluminum. The iron and aluminum removed by this reagent from the sub-
soils investigated in this study is thought to have existed in the sub-
soils as amorphous hydrous oxides. The amorphous hydrous oxides of iron
and aluminum can serve as binding agents in soils, thereby, increasing
aggregate stability, but the crystalline forms of iron and aluminum
hydrous oxides do not affect aggregate stability. Since the citrate-
dithionite reagent can extract both crystalline and amorphous hydrous
oxides or iron, it is recommended that a procedure be developed which
will distinguish between crystalline and amorphous forms of iron and
aluminum hydrous oxides in soils.
Field observation indicated that initial moisture content of the subsoil
could affect erodibility determinations. Additional work in this area is
recommended.
Fulfillment of the above recommendations would allow a more universal
application of a subsoil erodibility model such as developed in this study.
-------
Ill. INTRODUCTION
Sediment is a major pollutant of surface water in the United States.
Much of the sediment is derived from agricultural land. However, with
the extensive and rapid conversion of agricultural land to other uses
such as housing, road, school, business, and industry, an increasing
amount of sediment currently has its source in urban areas. Sediment
yield from areas in intensive suburban developments is often appreci-
able larger than that of cultivated land in rural areas. Thus, tech-
niques are needed for minimizing soil losses in urban situations, where
soil erosion has received little attention in the past.
Sediment yield from agricultural land has been successfully described
by the Universal Soil Loss Equation, which combines the principal
factors that influence surface soil erosion by water. The equation
takes the form:
A - RKLSCP
where
A is the soil loss expressed in the units selected for K and
for the time period covered by factor R, short tons/acre.
R is the rainfall factor, usually expressed in units of rain-
fall-erosivity index, El, ft-short tons/acre times the
maximum 30-min intensity in inches/hour time 10~2.
K is the soil-erodibility factor, commonly expressed in short
tons per acre per El unit.
L is a slope-length factor, dimensionless ratio.
S is the slope-steepness factor, dimensionless ratio.
3
-------
C is the cropping and management factor, dimensionless ratio.
P is the erosion control practice factor, dimensionless ratio.
The above equation can be expressed in metric units by multiplying the
English El units by 1.735 to arrive at the storm energy in tm/ha times
the maximum 30-minute intensity in cm/hr. The factor for direct
conversion of K to t/ha per metric El unit is 1.292 (Wischmeier, 1972).
The most difficult parameter to be specified in the equation is the
soil erodibility factor. Although the Universal Soil Loss Equation has
been used successfully by the Soil Conservation Service for predicting
soil losses and conservation measures on agricultural land, little
attempt has been made to adapt the equation for use with urban construc-
tion site soil loss. A primary difficulty in using the Universal Soil
Loss Equation for predicting soil erosion on construction areas is the
evaluation of the erodibilities of subsoils, which are commonly heavier
in texture than the surface soils for which existing relations have
been derived. In addition, subsoils likely have aggregating agents
that are very much different than those found in surface soils and the
degree of soil aggregation is known to have a profound influence on soil
erosion by water.
To allow accurate prediction of soil losses from urban construction sites,
an improved method of relating the soil erodibility to basic soil para-
meters must be developed. The first step for such an improvement was
made with the development of a nomograph from which soil erodibilities
of predominant light-textured soils can be determined (Wischmeier et al.
1971). Soil parameters used in predicting the soil erodibility for
surface soils are silt content, sand content, organic matter content,
structure and permeability of the soil profile. Since the actual cohe-
sion between soil particles is determined by chemical and mineralogical
constituents, any effort to improve the existing procedure for predicting
erodibilities of surface soils and develop a technique for estimating
the erodibility of subsoils will likely utilize basic chemical and
physical parameters.
-------
The objectives of this study were: (1) to test the soil credibility
model on soils having textural extremes, which are commonly found in
subsoils at construction sites but were not present in the surface
soils from which the model was developed; (2) to determine various
chemical, physical, and mineralogical characteristics of selected sur-
face and subsoils and to relate these parameters to the credibility
factor, K; and (3) to attempt to improve the soil erodibility factor
model so that subsoils are included, or to develop a separate model for
use with subsoils, by utilization of data produced during accomplishment
of the first two objectives.
Objective 1 was accomplished by determining the erodibility of six sub-
soils from the Midwestern part of the country by use of a field rain-
fall simulator. Subsoils were selected with variation in texture, iron
oxide, and organic matter content. The observed erodibilities of sub-
soils were compared with the erodibility predicted by the erodibility
nomograph to judge the accuracy of the nomograph for subsoils. Objec-
tive 2 was accomplished by determining a. variety of chemical, mineralog-
ical, and physical parameters of surface and subsoils for which the
erodibilities had been measured. The erodibilities of the soils were
than related statistically to the other parameters measured to determine
those important in influencing intrinsic soil erodibility.
Objective 3 was met by multiple regression analyses of the data collected
under Objectives 1 and 2 to produce models which successfully predict
erodibility of surface and subsoils as functions of their chemical,
mineralogical and physical properties. From the model for subsoil
erodibility a nomograph was constructed which allows estimation of the
erodibility factor, K, for high clay subsoils.
After the erodibility factor, K, has been determined, the Universal
Soil Loss Equation may be used to predict soil loss from subsoils for
a given rainfall pattern, slope steepness, and slope length. Once the
potential soil loss from a site is established, soil erosion control
measures may be recommended which are effective in maintaining soil
loss from the area within established tolerances.
-------
IV. FIELD EXPERIMENTS
The soil-erodibility factor, K, in the Universal Soil Loss Equation
(A - RKLSCP) is defined as the rate of soil erosion per unit of
erosion-index (El) from unit plots on that soil. An unit plot is
defined as being 22.1 m (72.6 ft) long with a uniform length-wise
slope of 9 percent in continuous fallow, tilled up and down the slope
(Wischmeier and Smith, 1965). A further refinement of this definition
is given by specifying that continuous fallow represents a condition
in which land is tilled and kept free of vegetation for a period of
two years or until prior crop residues have decomposed. Before con-
ducting soil-loss measurements, the plot is plowed and placed in
conventional corn seedbed each spring and is tilled as needed to pre-
vent vegetal growth or crusting. Under these conditions, the factors
L, S, C, and P in the universal soil loss equation each have a value
of 1 and the soil erodibility factor, K, equals A/R, where A is the
measured soil loss, while R represents the erosion-index or the rain-
fall factor (Wischmeier and Smith, 1965).
From the definition of the soil erodibility factor it may 'be inferred
that determinations of K are most readily performed on surface soils.
More recently, a soil erodibility nomograph was developed by Wischmeier
et al. (1971) which permits predictions of the soil erodibility factor
from routine laboratory determinations on soil and standard soil profile
descriptions. The accuracy and validity of the nomograph was confirmed
for 13 benchmark surface soils by comparing the predicted K values with
those measured in long term, natural-rain plot studies. Actually, the
nomograph's derivation was based on soil-loss measurements of mostly
medium textured surface soils, causing some uncertainty as to its
6
-------
accuracy for high clay subsoils. To establish a greater accuracy in
predictions of subsoil credibility, soil losses were measured on
selected sites of high clay subsoils using established simulated rain-
fall procedures. Two surface conditions, commonly found at subsoil
sites, were evaluated. The credibility factor of the subsoils for each
of these treatments was computed by standard procedures. The accuracy
of the nomograph was then checked for these soils by comparing measured
and predicted values. Then, a selection was made as to which soil treat-
ment should represent the standard surface conditions for soil credi-
bility determinations on subsoils.
PROCEDURE
The subsoils selected were located in a wide geographical area in the
middle West. The selections, primarily based on the clay content, were:
McGary silty clay near Bloomington, Monroe Co., Indiana
Portageville clay near Portageville, New Madrid Co., Missouri
St. Clair silty clay near Woodburn, Allen Co., Indiana
Wymore silty clay near Burr, Otoe Co., Nebraska
Pawnee clay loam near Burr, Otoe Co., Nebraska
Mayberry clay loam near Burr, Otoe Co., Nebraska
The soil profile descriptions of the above soils are given in Appendix A.
The credibility factors, K, were measured and calculated for the surface
soils on the Nebraska sites. These K values will be used in the sta-
tistical analysis phase of this project.
The general site areas were selected in cooperation with University
(Purdue University and the University of Missouri) and Soil Conserva-
tion Service personnel using high clay content of the subsoil as the
principal criterion. The clay contents of these subsoils, as determined
by procedures outlined in Chapter V, are: McGary 39.8%, Portageville
66.5%, St. Clair 38.7%, Wymore 38.5%, Pawnee 35.4%, and Mayberry 33.9%.
The specific locations were largely determined by practical considera-
tions such as accessibility, proximity to a water source, cooperation
of landowners and tenants, and natural topography.
7
-------
Site preparation was in accordance with the procedures outlined in the
Project Plan of the Research Proposal. The overburden soil was removed
with a bulldozer, and the site area was sloped to 9 percent steepness.
Three treatments were conducted on the McGary, Portageville, and St.
Clair sites: scalped, tilled and semi-compacted fill. A third of the
area that was sloped to 9 percent steepness with a bulldozer was used
without farther manipulations for the scalped treatment. The tilled
treatment consisted of plowing and disking to approximately 127mm
(5 inches) another area similar to that used for the scalped treatment.
The tilled treatment on the McGary, Portageville, and St. Clair sub-
soils received two thorough diskings (2-3 passes each time), one
immediately following plowing, the other just before rain tests. The
Nebraska sites received three thorough diskings, the third one performed
about two weeks after site preparation. The semi-compacted fill treat-
ment consisted of excavating the soil from another area of the sloped
site to a depth of approximately 300 mm (12 inches), returning the
removed soil back to the plot area and compacting the soil with the
bulldozer tread. Only the scalped and tilled treatments were tested
on the Nebraska sites. The deletion of the semi-compacted fill treat-
ment from the Nebraska sites, which were prepared under the supplemental
part of the project, was suggested by the similarity in results of soil
loss and runoff observed earlier on the McGary and Portageville sites.
Site preparation and rainfall test dates are summarized in Table 1.
Artificial rainstorms of about 63.5 mm (2.5 in) per hour were applied
to replicated treatments on 1.8 by 10.7 m (6 by 35 ft) plots, using
the rainfall simulator (rainulator) described by Meyer and McCune
(1958). Each rainulator test series consisted of an initial run of 60
minutes followed 24 hours later by two 30-minute runs on the-wet soil.
On the Nebraska sites, the two 30-minute runs were combined in a single
60-minute storm. However, the data were divided into two 30-minute
storms by using basic information such as sediment load and runoff data
from this storm and an analytical procedure for generating hydrographs
described by Foster et al. (1968). Runoff from each plot was collected
by gutters which extended across the lower plot end and emptied into
8
-------
Table 1. SUMMARY OF DATES OF SITE PREPARATION AND RAINULATOR TEST
a
Soil Site Preparation Rainulator Tests
McGary 16 August 1971 16-17 September 1971
Portageville 1 September 1971 15-16 October 1971
St. Clair 10 June 1972 5-6 September 1972
Wymore S 28 June 1972 14-15 August 1972
Pawnee S 28 June 1972 16-17 August 1972
Mayberry S 28 June 1972 24-25 August 1972
Wymore T 28 June 1972 22-23 August 1972
Pawnee T 28 June 1972 14-15 August 1972
Mayberry T 28 June 1972 25-26 August 1972
a S = subsoil, T « topsoil
the approach of a 18.3 cm (0.6 ft), calibrated, HS-flume. The flume
has a stilling well and stage recorder to measure the rate and amount
of runoff (Meyer, I960)- Approximately one percent of the runoff was
collected by a sampling slot located on an electrically powered rotat-
ing wheel placed in the runoff stream between the gutter and flume.
When large soil particles (aggregates) were transported, 100 percent
runoff samples were collected to insure that none of the large
aggregates were missed because of non-passage through the sampling
slot. Runoff samples were taken at 3-5 minute intervals during the
runoff period. Soil loss and runoff were computed by integrating the
measured hydrograph and acquired sediment content values of collected
runoff samples. Computations were performed on the Purdue University
CDC 6500 computer.
RESULTS
The observed soil losses and infiltration rates for the subsoils tested
are summarized in Table 2. Due to larger infiltration rates and/or
-------
Table 2. OBSERVED SOIL LOSS AND INFILTRATION OF SUBSOILS DURING
SUCCESSIVE RAINSTORMS
Storm
dura- Soil loss,
Treat- tion t /ha (short tons /acre) Slope,
Soil ment min Plot 1 Plot 2 Ave. %
McGary Scalped
60
24
91° .
38
.11
(11.11)° (17.00)
Semi-
compacted
fill
Tilled
Port- Scalped
ageville
Semi-
compacted
fill
Tilled
30
30
60
12
13
33
.35
(5.51)
.72
(6.12)
.33
(14.87)
30
30
60
30
30
60
30
30
60
30
30
60
30
30
12
17
15
11
11
3
-1
1
3
2
1
0
0
.71
(5.67)
.60
(7.85)
.11
(6.74)
.25
(5.02)
.66
(5.20)
.38
(1.51)
.37
(0.61)
.21
(0.54)
.27
(1.46)
.13
(0.95)
.84
(0*82)
.47
(0.21)
.90
(0.40)
0.61
(0.27)
13
16
30
.61
(6.07)
.03
(7.15)
.04
(13.40)
16
15
14
9
10
4
4
3
6
2
3
0
0
0
.10
(7.18)
.87
(7.08)
.39
(6.42)
.68
(4.32)
.54
(4.70)
.64
(2.07)
.17
(1.86)
.18
(1.42)
.10
(2.72)
.51
(1.12)
.47
(1.55)
.96
(0.43)
.13
(0.06)
.16
(0.07)
38
.11 9.0
(17.00)
12
14
31
.98
(5.79)
.88
(6.64)
.69 9.0
(14.14)
14
16
14
10
11
4
2
2
4
2
2
.41
(6.43)
.74
(7.47)
.75 9.0
(6.58)
.47
(4.67)
.10
(4.95)
.01 9.0
(1.79)
.77
(1.24)
.20
(0.98)
.69 9.0
(2.09)
.32
(1.03)
.66
(1.19)
0.72 9.0
0
(0.32)
.52
(0.23)
0.38
(0.17)
Infil- Adjusted soila
tration loss, t/ha
mm(in) (short tons/a)
17.02
(0.67)
4.32
(0.17)
1.78
(0.07)
22.10
(0.87)
4.57
(0.18)
2.03
(0.08)
27.43
(1.08)
4.06
(0.16)
2.29
(0.09)
16.26
(0.64)
5.08
(0.20)
2.54
(0.10)
13.21
(0.52)
4.32
(0.17)
1.52
(0.06)
32.26
(1.27)
16.26
(0.64)
8.13
(0.32)
54
.88
(24.48)
18
21
45
.70
(8.
.43
(9.
.64
(20.
20
.74
34)
56)
36)
(9.25)
24
.10
(10.75)
21
.25
(9.48)
15
15
5
3
3
6
3
3
1
0
0
.09
(6.
.98
(7.
.78
(2.
.99
(1.
.16
(1.
.75
(3.
.34
(1.
.83
(1.
73)
13)
58)
78)
41)
01)
49)
71)
.03
(0.
.74
(0.
46)
33)
.54
(0.
24)
10
-------
Table 2 (continued).
OBSERVED SOIL LOSS AND INFILTRATION OF SUBSOILS
DURING SUCCESSIVE RAINSTORMS
Treat--
Storm
dura- Soil loss,
tion t/ha(short) tons/acre)
Infil-
Slope, tration
Adjusted soil
loss, t/ha
Soil merit min
St. Scalped 60
Clair
30
30
Semi- 60
compacted
fill 30
30
Tilled 60
Plot 1
52.37
(23.
23.47
(10.
20
46
.98
(9.
.25
(20.
16
12
31
.79
(7.
.71
(5.
.52
(14.
30
30
Wymore Scalped 60
14
14
45
.68
(6.
.57
(6.
.95
(20.
30
30
Tilled 60
30
30
22
16
0
4
7
.28
(9.
.88
(7.
.00
(0.
.75
(2.
.20
(3.
36)
47)
36)
63)
49)
67)
06)
55)
50)
50)
94)
53)
00)
12)
21)
Plot
44.58
(19.
19.77
(8.
16.43
(7.
42.03
(18.
16.41
(7.
14.32
(6.
31.56
(14.
9.66
(4.
10.81
(4.
49.21
(21.
20.80
(9.
13.98
(6.
0.00
(0.
3.38
(1.
7.15
(3.
2 Ave. %
89)
82)
33)
75)
32)
39)
08)
31)
82)
95)
28)
24)
00)
51)
19)
48.48
(21.
21.63
(9.
18.61
(8.
44.14
(19.
16.60
(7.
13.52
(6.
31.54
(14.
12.17
(5.
12.69
(5.
47.58
(21.
21.54
(9.
15.43
(6.
0.00
(0.
4.07
(1.
7.17
(3.
9.0
63)
65)
30)
9.0
69)
41)
03)
9.0
07)
43)
66)
8.7
23)
61)
89)
9.1
00)
82)
20)
mm(in) (short tons/a)
8.89
(0.35)
1.78
(0.07)
0.
(0
10.
(0
3.
(0
1.
(0
21.
(0
6.
(0
3.
(0
20.
(0
7.
(0
4.
(0
63.
(2
25.
(1
20.
(0
76
.03)
16
.40)
56
.14)
78
.07)
84
.86)
35
.25)
81
.15)
57
.81)
37
.29)
83
.19)
50
.50)
40
.00)
32
.80)
69.83
(31.15)
31.14
(13.89)
26.95
(12.02)
63.57
(28.
23.90
(10.
19.45
(8.
45.42
(20.
17.53
(7.
18.27
(8.
71.60
(31.
32.30
(14.
23.09
(10.
36)
66)
68)
26)
82)
15)
94)
41)
30)
0.00
(0.
5.85
(2.
00)
61)
10.24
(4.
57)
11
-------
Table 2 (continued).
OBSERVED SOIL LOSS AND INFILTRATION OF SUBSOILS
DURING SUCCESSIVE RAINSTORMS
Storm
dura- Soil loss, Infil- Adjusted soil3
Treat-,, tion t/ha(short tons/acre) Slope, tration. loss, t/ha
Soil ment min Plot 1 Plot 2 Ave. % mm (in) (short tons/a)
Pawnee
May-
berry
Scalped 60
30
30
Tilled 60
30
30
Scalped 60
30
30
Tilled 60
30
30
47.39
(21.14)
23.38
(10.43)
27.01
(12.05)
11.21
(5.00)
16.03
(7.15)
15.56
(6.94)
70.68
(31.53)
30.82
(13.75)
23.45
(10.46)
4.77
(2.13)
2.35
(1.05)
3.52
(1.57)
38.24
(17.06)
16.78
(7.48)
17.17
(7.66)
27.98
(12.48)
15.40
(6.87)
18.88
(8.42)
66.15
(29.51)
c
c
2.31
(1.03)
c
c
42.82 8.9
(19.10)
20.08
(8.96)
22.09
(9.86)
19.59 8.9
(8.74)
15.71
(7.01)
17.22
(7.68)
68.42 9.0
(30.52)
30.82
(13.75)
23.45
(10.46)
3.54 9.1
(1.58)
2.35
(1.05)
3.52
(1.57)
19.56
(0.77)
5.59
(0.22)
2.54
(0.10)
43.69
(1.72)
7.11
(0.28)
5.33
(0.21)
10.67
(0.42)
1.27
(0.05)
0.00
(0.00)
44.96
(1.77)
15.24
(0.60)
10.16
(0.40)
63.26
(28.22)
29.68
(13.24)
32.66
(14.57)
28.83
(12.86)
23.20
(10.35)
25.40
(11.33)
98.70
(44.03)
43.00
(19.18)
32.71
(14.59)
5.10
(2.24)
3.45
(1.54)
5.16
(2.30)
Soil loss values were adjusted for slope steepness and slope length to
unit plots (Wischmeier and Smith, 1965).
This value was deleted in subsequent computations due to reduced soil
loss resulting.from a severe concavity at the plot end. At the end of
the 60-minute storm, sedimentation had eliminated the irregularity in
slope.
These values could not be determined because a natural rainstorm, which
occurred after the 60-minute storm, destroyed the plot.
12
-------
storage capacities and reduced runoff velocities, soil losses from the
tilled treatment were consistently smaller than those from the scalped
treatment. The combined soil losses from the two 30-minute storms on
the scalped plots were, for all but the Portageville subsoil, less than
the soil losses from the 60-minute initial storm. This tendency may
be explained by the mass removal of loose soil material as was evident
by the larger soil content in runoff samples collected during the first
storm in comparison to those collected during the 30-minute storms.
On the other hand, the tilled treatment for all soils showed larger
soil losses for the combined 30-minute storms than for the 60-minute
storm. These latter findings reflect decreased infiltration rates and
void storage as well as increased runoff velocities following slaking
of clods during the course of the experiment.
Average soil losses and infiltration rates of the semi-compacted fill
treatment of the McGary, Portageville, and St. Clair subsoils were
generally similar to those for the scalped treatment during correspond-
ing storms. Infiltration rates during corresponding storms tended to
be slightly larger on the filled treatment than on the scalped treat-
ment for the McGary and St. Clair soils; the reverse trend was
observed for the Portageville soil. However, the observed differences
between these two treatments are relatively small and are presumably
within experimental error of determinations. Because of the similarity
in data between the filled and scalped treatments, the semi-compacted
fill treatment was deleted from further consideration on the Nebraska
sites.
Appreciable differences in soil loss within a treatment were obtained
between subsoils. The largest soil erosion rate was obtained on the
scalped treatment of the Mayberry subsoil during the 60-minute storm,
98.7 t/ha (44.0 short tons/acre); this was followed by the Wymore,
71.6 t/ha (31.9 short tons/acre) and St. Clair, 69.8 t/ha (31.2 short
tons/acre). Next in this sequence were Pawnee, 63.3 t/ha (28.2 short
tons/acre), McGary, 54.9 t/ha (24.5 short tons/acre) and, a significant
last, Portageville, 5.8 t/ha (2.6 short tons/acre). This same sequence
13
-------
also resulted if soil losses of the two 30-minute storms were added,
except that the positions of Pawnee and Wymore were interchanged.
The Pawnee subsoil showed a nearly constant soil erosion rate during
successive rainstorms. A similar tendency was observed on the
Portageville subsoil, whereas all other subsoils tested showed an
appreciable decrease in soil erosion rates when progressing from storm
1 to storm 3. Some of the reasons for these differences will be
discussed later.
The tilled treatment showed the opposite trend for soil erosion rates
from that of the scalped treatment. On all subsoils except St. Clair,
soil erosion rates during the two 30-minute storms were larger than
those of the 60-minute storm. Soil loss from the second 30-minute
storm was generally larger than that from the first 30-minute storm.
No satisfactory explanation can'be given for the deviant response of
the St. Clair subsoil, except that a rapid breakdown of clods upon
wetting, followed by sealing and a subsequent massJremoval of small and
readily transportable soil particles, led to'initial high soil losses.
The observed increase in soil erosion rates from the tilled treatment
with successive storms can be attributed to a reduction in surface
roughness and infiltration rates, leading to larger runoff velocities
and ipso facto larger detachment and transport rates.
A summary of the observed soil losses and infiltration rates for the
Nebraska surface soils is given in Table 3.
FIELD OBSERVATIONS ON INDIVIDUAL SUBSOILS
McGary subsoil
The McGary subsoil, a lake bed deposit, is a very heterogeneous soil.
Appreciable variation (nearly linear) in clay content was evident from
the upper to the lower end of the plots. Also, textural, chemical,
and mineralogical variation could be visually discerned at any given
location on the plots. The impact of these variations on soil loss is
difficult to assess, but is thought to have increased soil erosion
rates because of reduced structural homogeneity. A view of the plots
14
-------
Table 3. OBSERVED SOIL LOSS AND INFILTRATION OF NEBRASKA SURFACE SOILS
DURING SUCCESSIVE RAINSTORMS
Storm
dura- Soil loss ,
Treat- tion t/ha(short tons/acre) Slope,
Soil ment min
Wymore Tilled
Pawnee Tilled
May- Tilled
berry
60
30
30
60
30
30
60
30
30
Plot 1
11.03
(A. 92)
7.40
(3.30)
8.02
(3.58)
18.09
(8.07)
10.56
(A. 71)
9.A6
(A. 22)
18. 3A
(8.18)
12.55
(5.60)
13. A3
(5.99)
Plot 2 Ave. %
10.98
(A. 90)
7.89
(3.52)
7.29
(3.25)
17.01
(7.59)
11.66
(5.20)
11. A6
(5.11)
17. OA
(7.60)
9.73
(A.3A)
10.20
(A. 55)
11.01 5.6
(A.91)
7.6A
(3.A1)
7.66
(3.A2)
17.55 7.3
(7.83)
11.11
(A. 96)
10. A6
(A. 67)
17.69 8.5
(7.89)
11. 1A
(A. 97)
11.81
(5.27)
Infil- Adjusted soiia
tration loss, t/ha
ram (in) (short tons/a)
38.35
(1.51)
7.37
(0.29)
5.8A
(0.23)
A2.A2
(1.67)
12. A5
(O.A9)
11. 9A
(O.A7)
29.21
(1.15)
6.35
(0.25)
5.08
(0.20)
30.31
(13.52)
21.05
(9.39)
21.07
(9.AO)
33. 7A
(15.05)
21.32
(9.51)
20.06
(8.95)
27.66
(12. 3A)
17. AO
(7.76)
18. A5
(8.23)
Soil loss was adjusted for slope steepness and slope length to unit
plots (Wischmeier and Smith, 1965).
of semi-compacted fill, scalped and tilled treatments is shown in Figures
1, 2 and 3, respectively, with close-ups of the soil surface for these
treatments in Figure A, 5, and 6.
Plot preparation on this site differed from those of other sites in
that polyethylene sheets were used to cover the plot area between
site preparation and tests. This may have affected the weathering pro-
cess in two respects. Wetting by natural rainfall in the intervening
period was prevented, thereby retarding the attainment of an adequately
weathered surface condition on the filled and scalped treatment.
Secondly, loose soil on the scalped plots was not removed by runoff
15
-------
from natural rainstorms before the commencement of rainulator tests.
Hence, a larger than average amount of sediment may have resulted from
the removal of loose soil material during the 60-minute storm. On the
other hand, this effect was compensated, at least in part, by the
absence of loose soil material produced during natural weathering
processes.
To obtain uniform slopes on the scalped plots, a final reshaping of
the surface with a 1-m wide improvised blade appeared necessary one
day before the scheduled rainulator test. One plot, however, retained
a severe concavity. Therefore, soil-loss values during the 60-minute
storm from this plot were excluded from analysis.
The tilled treatment represented a condition with appreciable void
storage. However, the surface clods broke down rapidly thereby
filling up voids (Figure 7).
Rills on this subsoil did not appear to be an important source of soil
in the first 127 mm (5 in) of artificial rain under the soil surface
conditions prevailing at this site. Only the upper plot end of the
tilled treatment showed the presence of some minor rills.
Portageville clay
The study on Portageville clay differed in several respects from all
other experiments: -
1. The study was conducted in the fall.
2. Clods in the tilled treatment had a high moisture content at the
time of the rainulator tests.
3. This subsoil had the highest clay percentage (66.5%) of all sub-
soils studied.
4. This subsoil had an unusually high organic carbon content, which
is known to be an important parameter in the soil credibility
factor.
It is not clear which of these factors might have contributed the most
to the relatively low rates of soil erosion.
16
-------
The weathering period led to very similar soil conditions between the
semi-compacted fill and scalped treatment. The plots had distinct
rills, which occurred mostly in residual bulldozer tracks. The soil
surface consisted of an agglomeration of aggregates, 56% of which were
in the 4-7 mm (0.16-0.28 in) size fraction while most of the remaining
aggregates were larger than 7 ran (0.28 in). The aggregates appeared
to be quite stable due to a high moisture level. However, upon air
drying and submersion in water under laboratory conditions the aggre-
gates disintegrated in a matter of seconds. This suggests that soil
erosion rates from these treatments might have been more severe if pro-
longed weathering, especially drying, had taken place before the rainu-
lator tests. This drying would have to have been longer than 3 weeks
since the last natural rain in the vicinity of this site occurred 3
weeks before the rainulator test.
Plot observations during and after rainstorms suggested that most soil
in runoff originated from rills where sufficient concentration of flow
enabled transport of large aggregates. The importance of this effect
in soil erosion can be seen in Figure 8, where the rill portion in the
photograph shows mostly coarse aggregates embedded or still attached
to the soil mass, while the interrill region exhibits a much more
uniform texture consisting of finer aggregates. Also, a 3- to 4-
fold increase in measured soil content upon full sampling of runoff
over that obtained by sampling 1% of runoff with the rotating wheel
with sampling slot (Meyer, 1960) indicated the importance of soil
erosion by aggregate detachment and transport.
The surface roughness of the interrill region on the scalped (Figure
8) and semi-compacted fill treatment indicated a large degree of
stability of individual aggregates. Also, it appeared that aggregates
resisted detachment and transport into existing rills by splash, which
is considered to be an important mode of soil erosion from upland areas.
The tilled treatment yielded large clods, which were still very wet at
the time of the rainulator tests. The stability of these clods after
127 mm (5 in) of artificial rain at 63.5 mm (2.5 in) per hour was not
17
-------
3»k.
**i
Figure 1. Plot view of the semi-
compacted fill treatment
on the McGary subsoil
before rainstorm tests.
Figure 3. Plot view of the
tilled treatment on
the McGary subsoil
before rainstorm
tests.
Figure 2. Plot view of the scalped treatment on the
McGary subsoil before rainstorm tests.
18
-------
*** \
Figure 4. Close-up view of the semi-compacted fill treatment on McGary
subsoil before rainstorm tests.
4: "!S! r6, . Si 9j :! I,c.- I
Figure 5. Close-up view of scalped treatment on McGary subsoil before
rainstorm tests.
Figure 6. Close-up view of tilled treatment on the McGary subsoil
before rainstorm tests.
19
-------
>
Figure 7. Surface condition of tilled treatment on
McCary subsoil after 63.5 mm (2.5 Inches)
of artificial rain.
Figure 8. Close-up view of scalped treatment on
Portageville subsoil after 127 mm (5 inches)
of artificial rain.
20
-------
appreciably affected (Figure 9) by impacting raindrops. Consequently,
the large voids between clods stored sizable quantities of water, there-
by reducing runoff. Also, the tortuous pathway of flow to the lower
plot end reduced runoff velocities and, thus, soil detachment by shear
flow. Although some disintegration of clods was visible after the
rain tests (Figure 9) soil particles detached from clods were deposited
in the void spaces between clods.
Figures 10 and 11 show the plot condition of the tilled and scalped
treatment, respectively, after 127 mm (5 in) of artificial rain.
St. Clair silty clay
Like the McGary subsoil, the St. Clair subsoil was a lake bed deposit
of non-uniform composition. The non-uniformity in this subsoil resulted
mainly from a cut into a 16% natural slope. As a consequence, the
lower 3.0 m (10 ft) of the plot area consisted of soil material from
the B-horizon. However, within a single horizon the subsoil appeared
to be quite uniform in contasts to the McGary subsoil. A second
difference was the presence of finer cracks in the St. Clair subsoil
than in the McGary (Figures 15 and 17). These cracks may explain the
lower infiltration rate during the 60-minute test on the St. Clair
subsoil for both the scalped and semi-compacted fill treatments.
Figures 12, 13, and 14 show the surface condition of the scalped,
tilled, and semi-compacted fill treatments, respectively, with close-
ups of each treatment in Figures 15, 16, and 17. In contrast to the
subsoil on the McGary site, the St. Clair subsoil was exposed to natural
weathering for at least 10 weeks. In fact, this site was exposed longer
to weathering than any other site in this project. Rills were in
strong evidence at the time of rainulator tests on both the scalped and
filled treatment. Rilling became more severe during the course of the
experiments. The effect of weathering is apparent in Figure 18 showing
a relatively high soil content in runoff samples during the initial
portion of the 60-minute storm.
21
-------
Figure 9. Close-up view of tilled treatment on
Portageville subsoil after 127 mm
(5 inches) of artificial rain.
Figure 10. Plot view of tilled
treatment on Portageville
subsoil after 127mm (5.0
inches) of artificial rain.
22
Figure 11. Plot view of scalped
treatment on Port-
ageville subsoil
after 127mm (5.0
inches of artificial
rain.
-------
N>
UJ
Figure 12. Plot view of scalped Figure 13. Plot view of tilled Figure 14,
treatment on St. Clair treatment on St.
subsoil before rain- Clair subsoil before
storm tests. rainstorm tests.
Plot view of semi-
compacted fill treat-
ment on St. Clair
subsoil before rain-
storm tests.
-------
Figure 15.
Close-up view of the scalped treatment on St. Clair
subsoil before rainstorm tests.
Figure 16.
Close-up view of the tilled treatment on St. Clair
subsoil before rainstorm tests.
Figure 17.
I
Close-up view of the semi-compacted fill treatment
on St. Clair subsoil before rainstorm tests.
24
-------
14
10
e
0)
c
o
_ 6
°o
(A
4
St. Clair Subsoil
treatment - Semi-Compacted Fill
60
Time , min.
90
120
Figure 18. Soil content in runoff from the compacted fill treatment
on the St. Clair subsoil.
25
-------
The tilled treatment was disked immediately before the application of
artificial rains. During the 60-minute storm, clods in this treatment
slaked relatively fast thereby filling up voids (Figure 19). At 20-
minutes into this storm, runoff from this treatment had reached
appreciable levels. Rills were very apparent at the end of the 60-
minute storm (Figure 20) and were mostly concentrated in the lower
3.0 m (10 ft) section of the plot area representing the B-horizon.
Rills did not extend beyond 4.6 m (15 ft) from the plot end even after
63.5 mm (2.5 in) of additional rain (Figure 21). It should be noted
that the tilled treatment on the St. Clair soil produced high soil
losses during the first 60-minutes of rain.
Wymore silty clay
Site preparation on the Wymore subsoil is shown in Figure 22. The
Wymore subsoil was an extremely friable material which, after plowing
followed by two diskings (Figure 22) and two weeks of natural weather-
ing, readily broke into small aggregates during successive diskings.
Figures 23 and 24 show the surface condition of the tilled treatment
plot area after a final disking immediately before the application of
artificial rain. The tilled treatment on this subsoil most nearly
resembled the plot condition of surface soil studied for erodibility
factor determinations. The tilled treatment produced a condition that
was able to absorb-almost all rain applied during the first rainstorm.
Slaking of soil clods and aggregates was almost non-existent, though
60-minutes of rain did produce some soil consolidation. The soil sur-
face retained a large degree of roughness due to stable aggregates
adhering to the soil surface. The low runoff rate and "spongy" nature
of this soil in the tilled treatment can be seen in Figure 25, taken
at the end of the 60-minute storm. The stability of individual
aggregates prevented effective sealing of the soil surface. Consequent-
ly, the tilled soil retained a high water absorptive capacity. Infil-
tration during the combined 30-minute storms was large on the tilled
trsatment plots and far exceeded infiltration observed on the scalped
treatment plots during corresponding storms (Table 2). The surface
26
-------
^^Sg^'.-'^^SSf^
^^^^3£^5p?5|
-.i>r_ ;<* JA- - L'-^tf : -k j.- -T«J
Figure 19. Close-up view of the tilled treatment on St.
Clair subsoil 17 hours after the initial 63.5
mm (2.5 inch) artificial rainstorm.
f
Figure 20. Plot view of the tilled
treatment on St. Clair
subsoil after 63.5 mm
(2.5 inches) of artifi-
cial rain.
Figure 21. Plot view of the
tilled treatment on
St. Clair subsoil
after 127mm (5 inches)
of artificial rain.
27
-------
^'^^^^^'^..:^'^^-i^^f^ii'"^/. *£'£: -,^*&-f*
Figure 22. Site area of Wymore subsoil during
site preparation.
28
-------
Figure 23. Plot view of the tilled treatment
on Wymore subsoil before application
of artificial rain.
Figure 24. Close-up view of the tilled treatment
on Wymore subsoil before application of
artificial rain.
29
-------
condition of the tilled soil after 24 hours of drying (Figures 26 and
27), however, was very similar to that of the scalped-treatment soils
before rain tests (Figures 28 and 29). Apparently, the loose, yet
stable, intergranular matrix produced by the tilled treatment provided
ample storage for infiltrating water, while the weathered surface re-
presented an agglomerate of aggregates which was similar to that
produced by the scalped treatment. The reason for the stability of
individual aggregates is not clear but is probably related to the
chemical composition of the soil.
The scalped-treatment soil showed a hexagonal crack pattern with about
50 mm (2 in) diameter hexagons overlaid with numerous aggregates. These
aggregates, of which 2/3 were in the 0.5-5 mm (0.02-0.20 in) size
fraction, were readily detachable. Consequently, mass removal of
aggregates in rills, fed by interrill sources through sheet flow and
splash, led to initially large soil content in the runoff. As the
application of rain continued, a gradual decrease in the soil content
of runoff was observed (Figure 30). Rain applied during the first 12
to 15 minutes of the first storm was largely absorbed in soil cracks
which subsequently were closed by the swelling soil.
Observations at this site suggest that following a rainstorm and sub-
sequent drying, the original condition of a weathered surface is quickly
re-established. Figure 31 shows the surface condition of this subsoil
one week after the completion of rainulator tests.
Pawnee clay loam
The preparational phase of the Pawnee site is shown in Figures 32 and 33
for the tilled and scalped treatments, respectively. The Pawnee subsoil
was extremely difficult to till especially in a dry state. Plowing
followed by three thorough'diskings, each disking consisting of two to
three passes, did not appreciably reduce clod sizes. The final clod
size achieved before application of artificial rain varied from about
25 to 76 mm (1 to 3 in) (Figure 34). Figure 35 shows the tilled plot
before the application of rain. A rain of less than 30-minutes was
30
-------
Figure 25. Plot view of the tilled Figure 26. Plot view of the
treatment on Wymore sub-
soil immediately after
63.5 mm (2.5 inches) of
artificial rain.
tilled treatment on
Wymore subsoil 2A
hours after the ini-
tial 63.5 mm (2.5
inches) of artificial
rainstorm.
Figure 27. Close-up view of the tilled treatment on Wymore
subsoil 24 hours after the initial 63.5 mm (2.5 inch)
artificial rainstorm.
31
-------
Figure 28. Plot view of the scalped
treatment on Wymore subsoil
before the rainstorm tests.
Figure 29. Close-up view of the scalped treatment
on Wymore subsoil before the rainstorm
tests.
32
-------
12
1O
c
9)
= 6
o ,
u
O
Wymore Subsoil
treatment - Scalped
9O
120
6O
Time, mi n.
Figure 30. Soil content in runoff from the scalped treatment on the
Wymore subsoil.
33
-------
Figure 31. Close-up view of the tilled treatment
on Wymore subsoil one week after rainulator
tests.
34
-------
%&{?-* "4
*tjf- ** .. , A.
Figure 32. Tilled treatment area on Pawnee
subsoil during site preparation.
Figure 33. Scalped treatment area on Pawnee
subsoil during site preparation.
35
-------
Figure 34. Close-up view of the
tilled treatment on
Pawnee subsoil before
rainstorm tests.
Figure 35. Plot view of the
tilled treatment on
Pawnee subsoil before
rainstorm tests.
Figure 36. Close-up view of the tilled treatment on
Pawnee subsoil after 18 mm (0.71 inches)
of artificial rain.
36
-------
needed to produce an effective surface seal caused by slaking of the
clods. Figure 36 shows the surface condition of the tilled plot after
17 minutes of rain, while Figure 37 gives an overview of this plot
following the 60-minute storm. Some rill development was evident at
the end of the 60-minute storm.
The scalped treatment morphology prior to the rain tests is shown in
Figures 38 and 39. Selective soil erosion on this subsoil is in strong
evidence as demonstrated by the deposition of white colored silt and
fine sand in the rills or depressions of the plot. The readily disper-
sable nature of this soil is not fully understood, but may be related,
at least in part, to the relatively low levels of amorphous hydrous
oxides of iron and aluminum. The uniformity in soil erosion rates for
the 60-minute and 30-minute rainstorms on this treatment is probably
related to the dispersable nature of this soil, although sediment
content of runoff samples declined somewhat during the 30-minute storms,
About 127 mm (5 in) of artificial rain and 61 mm (2.4 in) of natural
rainfall eliminated any visible difference in surface condition between
the scalped and tilled treatment. A plot view of the scalped treatment
24 hours after the 60-minute rainstorm is shown in Figure 40, while a
close up view of the soil surface 2 weeks after all rainstorms is shown
in Figure 41.
Mayberry clay loam
The preparational phase on the Mayberry site is shown in Figures 42
and 43 for the tilled and scalped treatments, respectively. The two-
month period between site preparation and rainulator tests yielded a
highly weathered subsoil with numerous aggregates loosely bound to the
soil surface (Figure 44). Figure 45 shows the surface condition of
the scalped treatment before rainulator tests, while a close-up view
is shown in Figure 46. The largest frequency of aggregates was in the
1 to 2 mm (0.04 to 0.08 in) size fraction (Figure 47). The maximum
frequency in aggregate size on this soil was appreciably smaller than
that for the Wymore subsoil. A possible explanation for the increase
37
-------
Figure 37. Plot view of the tilled
treatment on Pawnee subsoil after
63.5 mm (2.5 inch) rainstorm.
Figure 39. Close-up view of the
scalped treatment on Pawnee sub-
soil before rainstorm tests.
Figure 38. Plot view of the
scalped treatment on Pawnee sub-
soil before rainstorm tests.
Figure 40. Plot view of the scalped
treatment on Pawnee subsoil 17 hours
after 63.5mm (2.5 inches) of artifi-
cial rain.
38
-------
Figure 41. Close-up view of the tilled treatment
on Pawnee subsoil about 2 weeks after
rainulator tests.
39
-------
Figure 42. Tilled treatment area on Mayberry subsoil
during site preparation.
Figure 43. Scalped treatment area on Mayberry subsoil
during site preparation time.
Figure 44. Scalped treatment area on Mayberry subsoil
after 8 weeks of natural weathering.
40
-------
"Si
Figure 45. Plot view of the scalped Figure 46. Close-up view of the
-. treatment on Mayberry scalped treatment on
subsoil before rainstorm Mayberry subsoil
tests. before rainstorm tests,
41
-------
30
20
10
a
i .
X .
JB
30
w
k
10
Mayberry Subsoil
Wymore Subsoil
0.01
0.1
Aggregate
Size
1.0
mm
10
Figure 47. Aggregate-Size Distribution for Aggregates on
Wymore and Mayberry Subsoil Before Rainstorm Tests
42
-------
in soil loss from the Mayberry subsoil over that of the Wymore subsoil
during the 60-minute storm (Table 2) may be the more readily detachable
and transportable nature of aggregates in the Mayberry subsoil. Also,
the soil content trend for runoff from the scalped treatment of the
Wymore and Mayberry subsoils were very similar (Figures 30 and 48).
It should be noted, however, that the data obtained from the Mayberry
subsoil for the two 30-minute storms had to be synthesized because of
the confounding influence of a natural rainstorm of 61 mm (2.4 in)
following the 60-minute storm. The synthesis, based on soil content
and infiltration data (Figure 48) from the simulated rainstorms,
introduced some uncertainty in the acquired soil-loss values. On the
other hand, the consistency in these basic data and good reproduci-
bility between replicates for the 60-minute storm gave credence to the
derived soil loss values for the 30-minute storms. Furthermore, it
should be noted that in computations of soil erodibility, the influence
of the soil loss measurements of the 60-rainute storm appreciably out-
weighed those of the 30-minute storms.
The tilled treatment of the Mayberry subsoil differed from that of the
Wymore subsoil in that the soil was less friable. In fact, for the
tilled treatment, even following a final disking,, a large degree of
non-uniformity in clod sizes was retained (Figures 49 and 50). The
soil did not readily slake or break down as was observed with the
Pawnee subsoil. In fact, 188 mm (7.4 inches) of rain left a soil
surface with a considerable degree of roughness (Figures 51 and 52)
causing reduced runoff velocities and soil loss. Clods disintegrated
to some extent into small aggregates (Figure 52), which were either
deposited in voids or carried by the runoff water.
The tremendous stability of the individual aggregates was also apparent
in the scalped treatment (Figures 53 and 54). One day after the con-
clusion of the rainulator tests, the soil still looked like an agglom-
erate of stable aggregates (Figute 54). Again, the stability of these
aggregates is not well understood, but is presumably related to their
chemical composition.
43
-------
16
14
12
e
o
u
o
(A
Computer
Gersereted
IVIciy berry Subsoil
treatment - Scalped
60
120
180
Time, min.
48. Soil content in runoff from the scalped treatment on the Mayberry subsoil.
-------
Figure 49. Plot view of the tilled treatment on
Mayberry subsoil before rainstorm tests,
Figure SO. Close-up view of the tilled treatment on
Mayberry subsoil before rainstorm tests.
-------
Figure 51. Plot view of the tilled treatment
on Mayberry subsoil after 188 mm
(7.4 inches) of rain.
Figure 52. Close-up view of the tilled treatment
on Mayberry subsoil after 188 mm
(7.4 inches) of rain.
46
-------
Figure 53. Plot view of the scalped treatment
on Mayberry subsoil after 188 ram
(7.4 inches) of rain.
,^Ejf- ' >f. ^ *-
&tM$A
Figure 54. Close-up view of the scalped treatment
on' Mayberry subsoil after 188 mm
(7.4 inches) of rain.
47
-------
Most of the soil removed from the Mayberry scalped-treatment plot
appears to have originated from rills (Figure 53). The bulk of runoff
sediment consisted of aggregates. The significance of this observa-
tion is that high soil erosion rates from a weathered and well aggre-
gated subsoil like the Mayberry, and to some extent also the Wymore
subsoil, is a recurrent phenomenon following drying. It appears that
particle size and the degree of interaggregate bonding may be important
parameters in soil erosion problems of high clay subsoils.
COMPUTATION OF SOIL ERODIBILITY FACTOR, K
In computing the soil credibility factor, K, soil loss measurements
for each storm and treatment were adjusted to standard conditions of
9-percent slope steepness and 72.6 feet slope length using relation-
ships given by Wischmeier and Smith (1965). The cropping-management
factor, C, and erosion-control practice factor, F, were taken to be 1.
K-values could then be computed as the average soil-loss per unit of
R, where R represents the number of erosion-index units for a given
storm as defined in the Universal Soil Loss Equation. So that the K
values computed from the simulated rainstorms would more nearly repre-
sent the storm size distribution of natural rain, average soil losses
per unit of R were computed for combinations of thirteen 63.5 mm (2.5
in) rains on moderately dry soil, four 31.75 mm (1.25 in) rains on wet
soil, and three 63.5 mm (2.5 in) rains on wet soil (Wischmeier et al.,
1971). This approach minimized the influence of variations in antece-
dent moisture and reflected annual rainfall pattern in the geographical
area where tests were conducted. To further adjust for differences in
rainfall energy between a natural rainstorm and the simulated rain-
storm, both of 63.5 mm (2.5 in) per hour rainfall intensity, the weighted
soil erodibility factor was multiplied by a factor 0.8 (Meyer and
McCune, 1958). A summary,of the observed erodibility factors, K , , and
the nomograph derived erodibility factors, K (Wischmeier, et al.,
Tionto
1971), using physical and chemical parameters determined from soil pro-
file descriptions and laboratory analyes, is given in Tables 4 and 5
for the tilled and scalped conditions, respectively.
48
-------
Table 4. OBSERVED AND PREDICTED SOIL ERODIBILITY FACTORS FOR TILLED SUBSOILS
Soil
Total
Soil
ID carbon
number Kobs
organic Organic Sand Silt
matter (>100y), (2-100pm), Struc-C Perme-C
% % % ture ability
Knomo
McGary
Portageville
St. Clair
Wymore
Pawnee
Mayberry
191E
192E
212E
210E
206E
208E
0.17
0.01
0.31
0.03
0.24
0.04
0,34
1.23
0.75
0.92
0,82
0.82
0.54
2.12
1.29
1.58
1.41
1.41
0.46
0.00
9.80
2.01
20.56
7.56
59.05
32.15
48.82
53.45
38.68
54.18
4
4
4
4
4
4
6
6
6
6
6
6 '
0.43
0.19
0.34
0.34
Oo29
0.37
Determination according to Mebius (1960). Conversion factor for organic carbon to organic
matter was taken to be 1.72.
Procedures described in section V.
Evaluations were made from soil profile descriptions.
Soil erodibility factor, K, as determined from the nomograph of Wischmeier et al., 1971.
-------
Table 5. OBSERVED AND PREDICTED SOIL ERODIBILITY FACTORS FOR SCALPED SUBSOILS
Ol
o
Soil
McGary
Portageville
St. Clair
Wymore
Pawnee
Mayberry
Soil
ID
number
191S
192S
212S
210S
206S
208S
Kobs
0.36
0.05
0.48
0,49
Oo45
0.67
a
Determinations according
, matter was taken to be 1»
Totala
organic
carbon
7,
0.34
1.23
0.75
0,92
0,82
0,82
to Mebius
72.
a
Organic
matter
0.58
2.12
1.29
1.58
1.41
1.41
(1960).
Sandb
(>100y),
0.46
0,00
9.80
2.01
20.56
7.56
Conversion
Siltb
(2-100ym)8
7,
59.05
32.15
48.82
53.45
38.68
54.18
factor for
Struc-C
ture
4
4
4
4
4
4
organic
c
Perme-
ability
6
6
6
6
6
6
carbon to
d
Knomo
0.43
0.19
0,34
0.34
0.29
0.37
organic
Procedures described in section V.
Evaluations were made from soil profile descriptions.
Soil erodibility factor, K, as determined from the nomograph of Wischmeier et al., 1971.
-------
Significant differences were obtained between K , and K for both
obs nomo
the tilled and scalped subsoils. The K , values derived from the
obs
tilled treatments for all subsoils were consistently smaller than the
Knomo values» wnile K bs from the scalped treatment for all but the
McGary and Portageville subsoils were appreciably larger than the K
The magnitude of the differences raises doubt about the accuracy of
the nomograph as derived by Wischmeier et al. (1971) in soil erosion
predictions from subsoils - at least from high clay subsoils. Some
uncertainty exists, however, concerning the proper interpretation of
the structure and permeability parameters for subsoils. By equating
the tilled treatments with permeable surface soils underlain by massive
clay (storage taken to be similar to high intake rates) it may be
argued that the permeability factor for some subsoils should be assigned
a value 5 or perhaps even 4. In that case the K values should be
nomo
reduced by 0.03 or 0.06 units. Such a correction would improve the
erodibility factor prediction but significant discrepancies with the
K values would remain. Only the predictions for the Pawnee and St.
Clair subsoils would approach the K values for those subsoils.
Similarly, the soil structure of the weathered scalped subsoils could,
in some cases (Mayberry), be assigned the value 3. Again, no signifi-
cant improvement in erodibility predictions would be obtained In fact,
the new predictions might enhance the discrepancy (Mayberry).
A summary of the observed and predicted erodibility factors for surface
soils, including those of the Nebraska sites studied in this project8
has been given in Table 60 The procedures used for these determinations
have been outlined by Wischmeier et al., (1971). The cropping manage-
ment factors used for the Wymore, Pawnee and Mayberry topsoils were
based on the cropping history of the sites.
SELECTION OF STANDARD CONDITION FOR SUBSOIL ERODIBILITY FACTOR DETER-
MINATION
Subsoil erodibility determinations often involve soil conditions differ-
ent than those on surface soils. Subsoils are usually free of plant
residues, which would relax, if not eliminate, the suggested two-year
51
-------
Table 6. OBSERVED AND PREDICTED SOIL-ERODIBILITY FACTORS
FOR SURFACE SOILS
Soil name
Bedford
Bewleyville
Cincinnati
Muren
Russell
Rossmoyne
Switzerland
Parr
Morley
Miami
Miami
Fox
Princeton
Princeton
Princeton
Pembroke
Morley
Elkinsville
Varna
Frederick
Morley
Russell
Ockley
Grayford
Miami
Warsaw
Zanesville
Mar love
Mark land
Zanesville
Celina
Celina
Morley
Wea
Parr
Fox
Morley
Avonburg
Pawnee Topsoil
Mayberry Topsoil
Wymore Topsoil
Sample ID
number
101
103
104
105
106
112
114
115
117
119
121
123
125
126
128
131
133
135
140
144
145
147
149
150
152
154
155
157
160
162
164
166
168
169
170
171
172
174
176
178
179
180
182
207
209
211
Soil credibility factor
Knomo
.46
.39
.36
.52
.42
.44
.51
.41
.47
.45
.30
.30
.26
.24
.28
.24
.41
.08
.50
.08
.53
.31
.41
.29
.43
.38
.44
.39
.51
.32
.13
.52
.34
.22
.40
.26
.38
.48
.24
.24
.09
.37
.54
.28
.31
.32
Kobs
.46
.39
.39
.54
.43
.42
.55
.40
.51
.43
.33
.26
.22
.25
.28
.25
.42
.07
.39
.07
.54
.25
.42
.27
.39
.37
.48
.41
.58
,36
.11
.52
.36
.20
.36
.24
.34
.47
.26
.25
.09
.38
.55
.37
.31
.34
Knomo is the K factor derived
developed by Wischmeier et al.
measured.
from the soil erodibility nomograph
(1971) and Kobs is the K factor actually
52
-------
requirement for decomposition of plant material before credibility
measurements can be made. On the other hand, subsoils have had much
less exposure to processes of natural weathering, that is drying and
wetting, than surface soils. This consideration suggests an extended
fallow period upon removal of the overburden soil. Hence, a straight-
forward application of techniques developed for surface soils may not
be appropriate for determining the credibility factor of subsoils.
The poor predictions for the erodibility factor of high clay subsoils
from both tilled and scalped surface conditions seems to confirm this
notion. Therefore, it would seem justified to select standard condi-
tions for subsoil erodibility measurements which may constitute a
better basis for future erodibility predictions. The following reasons
led to the selection of the scalped surface as a more suitable standard
surface condition on high clay subsoils for erodibility measurements:
1. Scalped surfaces are more commonly found on construction sites.
2. Scalped surfaces are more reproducible than tilled soil.
3. Variations in void volume and thus storage capacity at differ-
ent locations and within replicates of a given location are
minimized.
Because of these considerations, the observed soil erodibility factors,
K , chosen for further analysis in the statistical phase of this
obs
project are those derived from the scalped treatment. It could be
argued then that the ratio K (tilled)/K (scalped) represents in
° obs obs
fact the soil-management factor. This factor is smaller than 1 for all
subsoils tested. Assuming that the basic format of the Universal Soil-
Loss Equation also holds for subsoils, the cropping-management factor
in this equation should then be replaced by the soil-management factor.
In arriving at the soil erodibility factor for subsoils it should be
stressed that a further refinement for the standard plot condition will
be needed especially in regard to the degree of weathering. Also, the
above considerations should not be interpreted as negating the existing
nomograph (Wischmeier et al., 1971) for all subsoils. Medium textured
subsoils may follow the soil erodibility factor predictions of the
nomograph developed by Wischmeier et al. (1971).
53
-------
V. LABORATORY CHARACTERIZATION OF REFERENCE SOILS FOR
PHYSICAL, CHEMICAL, AND MINERALOGICAL PROPERTIES
To arrive at relationships between the erodibility factor, K, determined
in field experiments, and the specific physical, chemical, and mineral-
ogical properties of soils, it was necessary to carry out extensive
laboratory analyses of reference soils. Wischmeier and Mannering (1969)
had previously used organic matter content, a variety of particle size
parameters, and other physical properties to empirically predict the
erodibility factor for surface soils. Later Wischmeier et al. (1971)
developed a nomograph utilizing five parameters (% silt plus very fine
sand, % sand, organic matter content, structure, and permeability) to
estimate the soil erodibility factor of surface and subsoils. Their
model was based on analyses of 55 surface soils.
Intuitively, intrinsic soil properties must control soil erodibility.
However, selection of the soil properties to be quantitatively measured
is difficult due to the large number of soil factors which may affect
erodibility either directly or indirectly. We decided that measurement
of basic soil constituents and properties was the best hope for improve-
ment in the prediction of the soil erodibility factor and further that
nonquantitative, subjective parameters should be avoided if possible.
The soil parameters measured were those which are known to influence
soil erodibility or have been reported to play some role in soil aggre-
gation processes.
SAMPLING AND SOURCE OF SOIL SAMPLES
Forty-three of the surface soil samples considered in this study were
part of the 55 soil samples used by Wischmeier et al. (1971 ) in
54
-------
development of the nomograph to predict erodibility of surface soils.
The other twelve soils used by Wischmeier could not be used in this
study because of insufficient amount of sample. When received, the
soil samples had been air-dried, ground to 2-ram and stored in paper
bags since they were collected between 1961 and 1965. Three additional
surface soils were collected as a part of this project during field
studies near Syracuse, Nebraska.
One subsoil sample was obtained from Agricultural Research Service Soil
Erosion Group at Purdue University. The remaining six subsoil samples
were collected at the site on which field studies of soil erodibility
were conducted as a part of this project.
The soil samples collected as a part of this study represented a com-
posite of several individual samples (0-15 cm depth) taken from the
actual rainulator plots. The composited samples were air-dried at
room temperature (25 C), ground to pass a 2-mm screen, and stored in
sealed plastic bags. A subsample was removed and ground to pass a 80
mesh sieve for use in certain chemical analyses. Finely ground samples
were stored in glass vials. All analyses given are the average of at
least duplicate determinations and are reported on a moisture-free basis.
DETERMINATION OF PHYSICAL PROPERTIES
Particle size analysis of samples was performed using the basic proce-
dure outlined by Jackson (1956). The soil samples were dispersed with
a sodium carbonate solution after removal of carbonates, soluble salts,
and divalent cations with sodium acetate buffer (pH 5), removal of
organic matter with hydrogen peroxide, and removal of iron oxides with
sodium citrate-bicarbonate-dithionite solution. The sand and coarse
silt (> 20ym diameter) particles were separated from the dispersed soil
sample by gravity sedimentation. The sand and coarse silt material was
then separated into standard size fractions by dry-sieving in a nest of
sieves. The fine silt (2-20ym diameter) was separated from the clay
(<2ym) by centrifuge sedimentation.
55
-------
The amounts of fine and coarse silt were combined to give a value for
total silt as were the fine sand fractions to give total sand. Summa-
tion of the values for sand, silt, and clay does not yield 100% due
to removal of carbonates, iron oxides, and organic matter in the dis-
persion process. A "new" silt parameter was also calculated by adding
the amount of very fine sand (50 to lOOym diameter) to the silt value
to give a value for the total soil particles having a mean diameter
between 2 and lOOym. Computation of a "new" silt parameter also
resulted in a "new" sand parameter involving particles with diameters
between 100 and 2000pm. These calculations merely involved transfer of
the very fine sand component from the sand fraction to the silt frac-
tion. A particle size factor, M, which Wischmeier et al. (1971) found
to be highly related to soil erodibility, was also calculated. The M
factor was computed by multiplying the "new" silt percentage by the
sum of "new" silt and "new" sand.
Soil structure and permeability classes were coded from information in
soil profile descriptions made at the rainulator sites as described by
Wischmeier et al. (1971). The permeability classes refer to the soil
profile as a whole. Both soil structure and permeability are somewhat
subjective parameters which depend upon the accuracy of the soil pro-
file description.
Table 7 provides data on the physical parameters determined for each
soil in the study. It is evident that the soils studied represent a
wide range in sand, silt, and clay contents and vary widely in struc-
ture, permeability and textural classification. Six of the seven sub-
soils analyzed contain in excess of 33% clay and as a group the sub-
soils have a good range in sand and silt contents. However, all of
the subsoils were considered to have blocky or massive structure
(coded 4), and very slow permeability (coded 6).
DETERMINATION OF-CHEMICAL COMPONENTS
The reference soil samples were analyzed for a variety of chemical con-
stituents. A sodium cltrate-bicarbonate-dithionite (CDB) extraction
56
-------
to remove iron oxides was performed on soil samples as outlined in
Appendix B. The CDB extraction procedure removes crystalline and non-
crystalline iron oxides except highly crystalline hematite and magne-
tite and also removes aluminum hydrous oxides and hydrous silica
associated with iron oxides in soils (Roth et al., 1969). Iron, alu-
minum, and silicon in the CDB extract were determined colorimetrically
as outlined in Appendix B and are reported as citrate-dithionite
extractable iron oxide, aluminum oxide, and silica.
Organic carbon was determined in soils by the procedure of Mebius (1960)
using < 80-mesh samples. Sodium pyrophosphate extractable organic
carbon and hot water extractable organic carbon were determined by the
procedures given in Appendix B. An index of the amounts of polysac-
charides present in soils was obtained by reacting the samples with
sodium periodate and measuring the decrease in periodate concentration
after 24 hours as described in Appendix B. Total organic carbon and
extractable organic carbon are reported as percent of soil on a weight
basis and the index of polysaccharides is given as millimoles of
periodate consumed per gram of soil in 24 hours of reaction at room
temperature. Total N in soils was estimated by the procedure of Nelson
and Sommers (1972) and total P by the method of Sommers and Nelson
(1972). Total N and total P are reported as ppm of soil on a weight
basis.
Table 8 gives a listing of the chemical constituents of the 46 surface
soils and 7 subsoils used in this study. Fairly wide ranges in Fe20.j,
Al-0., and SiO- content were observed, however, the surface soils were
somewhat uniform in nitrogen, phosphorus, and carbon content. A wide
range in periodate-oxidizable polysaccharides was evident with both
surface and subsoils. The subsoils studied had a reasonable range in
all chemical parameters measured.
DETERMINATION OF CLAY MINERALOGICAL COMPONENTS
The clay-sized material originally separated in the particle size
analysis procedure was subjected to x-ray diffraction analysis according
57
-------
Table 7. PHYSICAL CHARACTERISTICS OF SOILS USED IN THIS STUDY
00
Soil
Name
Surface soils:
Bedford
Bewleyville
Cincinnati
Muren
Russell
Rosstnoyne
Switzerland
Parr
Morley
Miami
Miami
Fox
Princeton
Pembroke
Morley
Elkinsville
Varna
Frederick
Morley
Russell
Ockley
Grayford
Miami
Warsaw
Zanesville
Type3
SIL
SIL
SIL
SIL
SIL
SIL
SIL
SIL
SIL
SIL
L
CL
SL
L
SIL
SICL
SIL
LS
SL
LS
SIL
CL
SIL
L
SIL
L
SIL
SIL
SIL
SL
SL
SIL
ID No.
101
103
104
105
106
112
114
115
117
119
121
123
125
126
128
131
133
135
140
144
145
147
149
150
152
154
155
157
160
162
164
166
Particle size
distribution
clay
15.0
20.3
14.4
18.6
19.7
13.5
16.8
19.9
14.5
18.2
19.2
24.3
6.1
19.7
18.8
36.8
19.4
3.9
8.3
2.6
15.6
25.4
12.6
15.5
21.8
14.5
10.7
11.9
15.5
7.8
13.2
17.2
silt
69.9
67.4
76.6
73.5
71.5
56.2
71.2
20.2
52.8
61.4
46.3
39.6
24.0
36.4
65.0
44.1
68.8
8.2
35.1
10.1
76.8
29.4
67.5
48.8
61.7
41.2
69.4
54.8
75.9
27.5
22.5
72.5
sand
10.1
7.9
4.8
2.6
4.0
27.1
8.1
4.2
28.1
16.2
28.7
28.8
66.9
38.4
10.3
9.1
7.0
86.4
54.4
85.6
3.3
28.4
16.5
31.5
11.6
41.1
15.9
29.6
3.7
61.9
59.3
6.9
1-2
5.7
5.4
18.8
15.8
9.0
2.2
6.6
7.2
1.6
4.6
6.8
5.6
3.1
6.0
6.2
2.7
0
0.1
0.1
0.7
6.2
8.3
2.5
4.1
1.5
3.0
5.4
12.5
3.4
3.9
9.8
4.0
Sand fractions (mm)
.5-1
6.9
14.9
26.6
18.4
7.9
10.2
15.8
14.2
1.9
8.2
15.7
12.9
10.7
12.3
14.4
10.8
7.8
5.8
0.6
23.4
12.1
15.9
1.2
11.0
3.9
10.6
15.8
29.0
15.5
11.5
20.4
9.0
0.5-
0.25
0.25-
0.1
_j
0.1-
0.05
5.7
29.6
10.1
15.8
5.7
22.0
19.7
19.4
4.5
10.4
22.2
16.4
20.1
18.9
16.5
12.0
7.8
37.9
5.0
39.0
5.9
15.6
1.2
27.6
3.9
19.9
24.5
29.6
22.4
21.9
31.9
7.9
38.8
20.3
15.9
18.4
7.9
35.8
27.6
30.6
40.6
46.4
32.1
33.4
42.0
35.2
13.4
22.8
34.1
47.8
74.7
29.0
17.9
32.0
25.2
39.5
42.8
39.3
34.2
21.6
29.3
40.5
30.4
22.0
42.8
41.9
30.4
31.6
69.6
29.8
30.3
28.6
51.4
30.5
23.2
31.7
24.6
27.5
49.5
51.7
50.4
8.5
19.8
7.9
57.9
28.2
69.9
17.8
47.8
27.2
20.1
7.2
29.3
22.2
7.5
57.1
New
silt
2-100 Vn
ft
New
sand
i .l-2mm
t
. Struc-
Mb turec
Perme- .
ability
74.2
70.7
78.0
74.3
74.3
64.2
73.6
71.4
67.2
66.3
53.0
48.7
40.5
48.4
70.1
48.8
72.3
15.5
45.8
16.9
78.7
47.4
79.1
54.4
67.3
52.4
72.6
57.0
76.9
41.3
27.0
76.4
5.8
4.6
3.3
1.8
1.2
19.0
5.7
3.0
13.7
11.2
22.0
19.7
50.4
26.4
5.2
4.4
3.5
79.1
43.6
78.9
1.4
20.4
5.0
25.9
6.1
29.9
12.7
27.4
2.6
48.1
54.8
2.9
5938
5328
6347
5649
5607
5347
5839
5314
5439
5138
3974
3333
3677
3616
5284
2597
5478
1470
4101
1615
6300
3211
6641
4372
4939
4309
6186
4804
6118
3691
2208
6057
3
2
2
3
2
3
3
3
3
3
3
3
3
3
2
3
3
2
3
2
4
3
2
3
3
4
3
3
4
2
3
3
4
4
3
6
3
3
6
3
4
4
3
4
3
3
3
4
3
2
3
2
5
4
3
4
3
4
3
3
5
4
2
6
-------
Table 7 (continued). PHYSICAL CHARACTERISTICS OF SOILS USED IN THIS STUDY
Ul
V£>
Soil
Name
Mar love
Markland
Zanesville
Celina
Celina
Morley
Wea
Parr
Fox
Morley
Avonburg
Pawnee
Mayberry
Wymore
Subsoils:
Wingate
McGary
Portageville
Pawnee
Mayberry
Wymore
St. Clair
Type3
SIL
SICL
SIL
SL
L
SIL
SIL
L
GSL
L
SIL
SICL
SICL
SICL
SL
SIC
C
CL
SICL
SIC
SIC
ID No.
168
169
170
171
172
174
176
178
179
180
182
207
209
211
188
191S
192S
206S
208S
210S
212S
Particle size
distribution
clay
13.4
34.9
17.3
10.0
19.8
11.7
17.3
13.2
6.9
16.8
10.3
36.3
29.4
37.5
9.9
39.8
66.5
35.4
33.9
38.5
38.7
silt
55.5
54.1
64.9
25.6
49.4
59.8
62.1
41.4
12.5
44.3
66.1
41.8
46.9
53.3
29.2
58.7
32.2
28.6
47.4
50.8
43.8
sand
26.9
1.8
13.8
61.1
27.6
24.4
15.7
40.1
78.1
35.0
20.3
16.5
18.8
6.3
54.0
0.8
0.3
30.7
14.4
4.7
14.8
1-2
2.4
9.1
0.8
2.7
2.7
8.9
3.2
5.2
12.9
1.9
5.3
3.1
0.5
4.9
8.8
20.0
e
2.7
2.2
4.9
2.6
Sand fractions (mm)
.5-1
15.4
9.1
1.6
12.7
6.1
13.0
12.7
14.1
29.8
7.7
14.1
6.1
6.0
11.1
14.8
0
e
15.0
10.7
16.8
16.8
0.5-
0.25
of aa
0.25-
0.1
'30.4 32.0
9.1
0.8
25.0
15.8
16.8
22.1
22.3
28.9
24.5
20.6
17.5
16.3
7.6
18.1
20.0
e
19.7
14.4
11.9
21.2
18.2
28.7
40.3
34.6
32.4
29.7
25.7
23.0
43.4
32.9
30.1
28.3
8.3
34.0
20.0
e
29.6
25.3
8.9
25.7
0.1-
0.05
19.8
54.5
68.0
19.2
40.8
28.9
32.3
32.6
5.4
22.6
27.2
38.8
48.9
68.1
24.3
40.0
e
33.0
47.4
57.4
33.7
New
silt
2-100 uin
_______ 7 _
60.8
55.1
74.3
37.3
60.7
66.9
67.2
54.5
16.7
52.2
71.6
48.2
56.1
57.5
42.3
59.1
32.2
38.7
54.2
53.5
48.8
New
sand
. l-2mia
21.6
0.8
4.4
49.4
16.3
17.3
10.6
27.0
73.9
27.1
14.8
10.1
9.6
2.0
40.9
0.5
0.3
20.6
7.6
2.0
9.8
Mb
5011
3077
5847
3238
4676
5633
5225
4440
1515
4138
6186
2809
3685
3425
3514
3514
1045
2291
3345
2965
2862
Struc-
ture
2
3
2
4
4
3
2
2
2
3
3
3
3
3
4
4
4
4
4
4
4
Perme- ,
ability
3
3
3
3
4
4
3
3
3
5
6
5
5
5
6
6
6
6
6
6
6
SIL, silty loam; L, loam; CL, clay loam; SL, sandy loam; SICL, silty clay loam; LS, loamy sand; CSL, gravely sandy
loam; C, clay; SIC, silty clay.
M=% new silt (7. new silt + 7, new sand).
1, very fine granular; 2, fine granular; 3, med. or coarse granular, 4, blocky, platy or massive.
1, rapid; 2, mod. to rapid; 3, moderate; 4, slow to med.; 5, slow; 6, very slow.
Insufficient sand in sample for fractionation.
-------
Table 8. CHEMICAL COMPOSITION OF SOILS USED IN THIS STUDY
Sample
ID
Number
Citrate-Dithionite
Extractable
Fe2°3
A1203
sio2
Total P
Total N
Total C
Surface
101
103
104
105
106
112
114
115
117
119
121
123
125
126
128
131
133
135
140
144
145
147
149
150
152
154
155
157
Soils:
1.53
1.61
1.58
1.81
2.31
1.06
1.65
1.97
1.55
1.94
1.50
2.08
0.69
1.83
1.65
3.01
1.96
0.73
0.96
0.58
1.60
2.02
1.31
1.20
2.07
1.33
1.09
1.22
.304
.246
.278
.324
.333
.210
.290
.305
.272
.318
.254
.288
.157
.257
.241
.398
.324
.135
.139
.088
.260
.246
.252
.242
.377
.226
.198
.189
.051
.099
.054
.088
.097
.048
.043
.058
.048
.066
.097
.112
.033
.100
.095
.114
.094
.046
.053
.030
.067
.129
.065
.081
.071
.071
.061
.061
305
352
315
298
433
482
450
290
273
485
420
750
287
446
482
676
529
375
303
305
447
425
371
381
891
257
358
471
1120
980
1350
1050
950
1120
990
970
1470
1190
1630
1720
1050
1130
1330
1230
1330
490
680
460
1150
1170
970
1440
1400
1010
1300
1350
1.06
0.77
1.25
0.93
0.84
1.20
1.28
1.09
0.94
0.76
1.86
1.72
1.26
1.29
1.38
1.05
1.25
0.78
0.49
0.31
0.96
1.03
0.88
1.38
1.01
1.16
1.38
1.17
Pyrophosphate
extractable
carbon
%..___.
.27
.26
.34
.22
.26
.30
.34
.24
.22
.22
.38
.31
.35
.31
.35
.26
.27
.19
.21
.18
.28
.24
.27
.47
.29
.32
.41
.36
Hot H-0
extractable
carbon
.044
.029
.064
.041
.039
.036
.032
.031
.029
.032
.067
.053
.053
.051
.057
.043
.047
.020
.033
.029
.041
.038
.043
.064
.048
.055
.065
.044
Periodate
consumed
(nmoles/gm)
.257
.195
.261
.313
.262
.209
.228
.212
.214
.191
.338
.365
.194
.284
.339
.516
.394
.116
.147
.098
.319
,564
.277
.277
.356
.213
.276
.323
-------
Table 8 (continued). CHEMICAL COMPOSITION OF SOILS USED IN THIS STUDY
Sample
ID
Number F
Citrate-Dithionite
Extractable
'e2°3
A12°3
Si02
Total P
Total N
Total C
~
160
162
164
166
168
169
170
171
172
174
176
178
179
180
182
207
209
211
Subsoils:
188
191S
192S
206S
208S
210S
212S
1.42
1.05
1.39
2.19
1.20
2.32
1.76
1.02
1.81
1.15
1.39
1.22
1.08
1.43
1.06
1.08
1.00
0.93
2.09
3.52
1.42
.85
1.15
0.80
2.34
.248
.173
.260
.305
.198
.385
.304
.160
.255
.213
.236
.175
.141
.216
.226
.232
.220
.178
.194
.242
.145
.121
.229
.293
.250
.070
.058
.080
.094
.086
.095
.067
.054
.090
.072
.112
.090
.054
.073
.053
.185
.135
.255
.156
.121
.265
.264
.127
.319
.155
566
257
427
483
387
733
308
231
303
303
590
887
300
335
300
762
593
1016
372
516
767
800
412
1018
1130
1900
830
1590
830
1290
2240
1090
920
940
1310
1670
1360
630
850
900
620
500
650
410
540
970
140
300
200
180
1.50
0.71
1.61
0.63
1.24
2.09
0.98
0.90
0.83
1.39
1.75
1.72
0.69
0.69
0.83
2.04
1.31
1.80
0.53
0.34
1.23
0.82
0.82
0.92
0.75
Pyrophosphate
extractable
carbon
.39
.27
.53
.25
.37
.45
.26
.28
.30
.40
.48
.38
.23
.24
.26
.41
.31
.38
a
.04
.38
.12
.18
.22
.06
Hot HO
extractable
carbon
.059
.049
.090
.036
.054
.073
.037
.066
.044
.062
.081
.057
.045
.028
.036
.141
.129
.102
a
.020
.018
.117
.056
.097
.080
Periodate
consumed
(mmbles/gm)
.444
.187
.294
.241
.281
.644
.295
.213
.197
.289
.358
.577
.184
.168
,241
.356
.275
.367
a
.124
.259
.305
.220
.249
1.056
Insufficient sample was available to run these analyses.
-------
Table 9. CLAY MINERALOGICAL COMPOSITION OF SOILS USED IN THIS STUDY
(percent of the clay fraction)
Sample
ID
No.
Vermi-
culite
Surface soils:
101
103
104
105
106
112
114
115
117
119
121
123
125
126
128
131
133
135
140
144
145
147
149
150
152
154
155
157
160
162
164
166
168
169
170
171
172
174
176
178
179
1.5
2.0
2.1
1.2
2.6
0.0
0.0
0.1
0.7
Io2
0.8
2.9
3.1
0.5
0.0
3.2
0.3
0.1
0.0
0.0
1.7
4.4
1.6
1.1
2.3
5.2
3.2
9.7
10.0
9.6
10.7
16.9
9.4
8.2
11.4
8.8
7.4
3.7
10.2
1.9
6.5
Kaolinite
plus
Mica halloysite
19.4
23.2
6.7
21.1
21.9
24.8
19.8
20.1
19.2
17.3
29.2
30.4
36.1
31.1
23.4
41.7
18.8
23.3
23.6
23.1
22.1
43.2
21.8
37.2
17.2
36.4
24.0
24.2
24.3
31.5
25.3
22.0
20.0
50.4
15.5
29.6
37.7
44.1
27.5
37.1
18.4
20.3
11.6
14.9
18.8
13.6
16.9
17.3
17.7
23.5
18.4
9.2
10.0
12.5
12.6
14.7
8.3
18.5
17.1
15.0
12.9
14.5
6.0
16.8
8.4
20.0
10.0
11.8
14.2
18.3
11.1
13.6
15.7
12.1
9.6
16.5
11.9
10.4
8.6
12.2
11.7
13.4
Amorphous
material
10.6
9.9
9.1
12.0
7.7
13.4
10.5
11.8
10.0
10.4
11.7
7.0
8.8
9.2
10.6
5.0
17.3
17.6
16.7
11.7
18.5
5.8
18.3
14.8
27.6
13.7
15.1
18.8
20.5
12.3
15.7
15.6
12.6
7.0
17.7
12.7
11.3
9.9
19.0
15.8
18.9
Montmo-
rillonite
11.0
21.3
9.9
16.6
23.0
12.5
12.3
13.2
10.3
12.1
9.2
14.9
4.5
15.6
18.3
11.0
23.3
18.2
16.9
1.4
20.1
12.9
18.8
3.4
20.9
15.7
19.0
10.5
3.7
6.4
8.1
4.4
10.8
2.1
7*6
6.7
10.6
11.3
14.2
19.9
19.0
Quartz
plus
feldspar
12.5
9.3
13.0
9.1
6.3
13.5
11.8
9.9
9.7
9.0
13.5
8.7
16.1
6.9
8.3
6.4
8.5
13.7
12.6
26.7
11.0
9.3
10.2
15.8
8.4
10.0
14.8
13.2
11.2
11.4
16.5
6.1
11.8
8.6
8.8
11.3
11.1
16.2
11.9
11.1
14.2
Chlo-
rite
24.7
22.7
44.3
21.2
24.9
18.9
28.3
27.2
26.6
31.7
26.3
26.1
19.0
24.1
24.7
24.4
13.3
9.9
15.1
24.2
12.1
18.4
12.5
19.3
3.7
8.9
12.1
9.4
12.1
17.8
10.2
19.3
23.4
14.1
22.5
19.1
11.5
6.4
5.0
2.5
9.5
62
-------
Table 9 (continued). CLAY MINERALOGICAL COMPOSITION OF SOILS USED IN
THIS STUDY
(percent of the clay fraction)
Sample
ID Vermi-
No. culite
180
182
207
209
211
Subsoils:
188
191S
192S
206S
208S
210S
212S
4.7
2.4
17.3
12.7
4.1
4.3
5.3
7.7
12.2
7.1
12.4
5.2
Kaolinite
plus
Mica halloysite
35.0
15.8
18.6
21.6
24.8
53.9
37.2
23.0
16.5
20.1
22.0
45.0
6.9
17.7
12.4
14.4
18.8
7.1
8.9
9.8
12.9
14.8
9.4
9.1
Amorphous
material
11.8
15.9
17.6
18.9
16.6
6.2
9.2
9.3
13.4
15.9
11.1
10.4
Montmo-
rillonite
14.6
16.7
25.6
14.1
22.3
11.4
19.4
22.6
30.0
28.6
30.4
5.7
Quartz
plus
feldspar
13.5
13.2
6.6
6.2
10.3
11.9
6.3
7.9
8.7
6.0
7.0
13.9
Chlo-
rite
13.5
16.7
2.0
12.1
3.2
5.3
13.7
19.8
6.3
7.5
7.8
10.7
63
-------
to procedures outlined by Jackson (1956) to provide a qualitative esti-
mate of clay mineral composition. The x-ray diffraction studies reveal-
ed that the clay fraction of almost every soil contained montmorillonite,
kaolinite, mica (illite), vermiculite, quartz, feldspar, and chlorite.
In addition, previous studies have shown that the <2jjm fraction of soils
contain variable amounts of amorphous inorganic material, e.g. allo-
phane, which is not detected by x-ray diffraction techniques. Accord-
ingly, the clay-sized fraction (<2um mean diameter) of soils in this
study was analyzed quantitatively for amorphous and crystalline compo-
nents. Montmorillonite and vermiculite were determined by the procedure
of Alexiades and Jackson (1965) as modified by Chapman (1970). Amor-
phous material and kaolinite plus halloysite were estimated by the
procedures described by Hashimoto and Jackson (1960). Mica and quartz
plus feldspar were determined by the procedures given in Appendix B
which was modified from Jackson, 1956. Chlorite was estimating by
summing the percent of the clay fraction composed of montmorillonite,
vermiculite, kaolinite plus halloysite, mica, amorphous materials, and
quartz plus feldspar and subtracting the sum from 100. Clay mineral
components are reported in Table 9 as percentages of the clay fraction,
however, in statistical analyses the clay mineral components were used
as percentages of the whole soil on a weight basis.
Table 9 shows that substantial variation exists in the clay mineralog-
ical composition of the surface soils studied. Subsoils tended to
contain somewhat higher amounts of vermiculite and montmorillonite and
lower amounts of chlorite as compared to surface soils. However, a
reasonable range in clay mineralogical composition was observed in sub-
soil samples.
64
-------
VI. STATISTICAL ANALYSIS OF DATA OBTAINED IN
FIELD AND LABORATORY EXPERIMENTS
STATISTICAL PROGRAMS USED
The majority of the statistical analyses on this project were performed
using the procedures found in the SPSS (Statistical Package for the
Social Sciences) manual (Nie, Bent and Hull, 1970). In particular the
REGRESSION procedure was used from the SPSS package of statistical
programs. An option in this procedure allows the user to specify that
the entrance of variables into the multiple linear regression model
will follow the forward selection technique (Draper and Smith, 1966)
such that the variable, not already included in the model, which
exhibits the highest partial correlation, will be the next forced into
the model.
The "weighted regression analysis program" (WRAP) was valuable in cer-
tain phases of the statistical analysis portion of this project. This
program performs multiple linear regression analysis using the back-
ward elimination technique (Draper and Smith, 1966) in which all the
independent variables are forced into the model and subsequently delet-
ed at each state until all variables remaining are significant at the
user defined probability level. WRAP differs from the SPSS program in
that the latter uses a forward selection technique whereas the former
uses a backward elimination technique. The WRAP contains a useful
option which allows each observation or case to be weighted by a func-
tion of its variance, whereas, the SPSS programs do not contain this
option.
Several other techniques and programs were evaluated as to usefulness
in the statistical analysis phase of this project but none of these
65
-------
Table 10. ESTIMATES OF PARAMETERS IN THE SIMPLE LINEAR REGRESSION
EQUATION OF K WITH ANALYZED SOILS VARIABLES
Independent
Variable
Name
Knomo
Z clay
Z sand (old)
Z silt (old)
2.0-1.0 mm sand
1.0-0.5 mm sand
0.5-0.25 mm sand
0.25-0.10 mm sand
0.10-0.05 mm sand
Z sand (new)
Z silt (new)
vr_Y (Y 4.Y ^
11^ 11 10'
Structure
Permeability
Z C-total
Z C-Na pyro.
Z C-Hot H20
Periodate Consumed
Fe203
A1203
Si02
ppm P
Z N
Z vermiculite
Z mica
Z kaolinite
amor, material
Z montmorillonite
Z quartz
Z chlorite
x__+x_0
(XO2
(X,.) (X,)
(xfj)
-------
was found to be any large benefit to the analysis.
STATISTICAL ANALYSIS OF SURFACE SOIL DATA
Simple linear correlation analysis of the observed K, K , , values for
obs
the surface soils with each of the independent variables reported in
Tables 7, 8 and 9 reveals that about half of the variables are posi-
tively correlated with K whereas the other one-half are negatively
correlated (Table 10). The values used for all of the independent
variables considered in Table 10, and for all subsequent statistical
analysis reported later, have been converted to a whole soil base.
That is, all of the values for the sand size fractions were converted
to percent whole soil in lieu of percent of sand as reported in Table
7 and the clay mineralogical composition values were converted to per-
cent whole soil in lieu of percent of the clay fraction as reported in
Table 9. There are several combinations of variables calculated and
reported in Table 10, such as the % Fe + Al as X and the percent silt
(new) squared as X.q. The combinations of variables that proved to be
of significant help in the statistical analysis were those represented
by X and X_ . The specific surface variable, X_7, was an attempt to
estimate the specific surface area of the soil from the amount of the
various clay minerals present and the specific surface areas reported
in the literature for the various clay minerals. The correlation
matrix for the surface soil variables is reported in Table 16, Appendix C.
Multiple linear regression analysis was performed on the surface soil
data with K , as the dependent variable using the WRAP. The partial
obs
regression coefficients of the multiple linear regression equations
are reported in Table 11. Table 11 contains the last 6 steps in the
backward elimination of variables in an initial model containing 22
independent variables. The final model at step 6 contains 4 variables
which are significant at the 0.05 probability level. The initial model
considered in Table 11 contained all of the single variables analyzed
(29) except for % sand-new (X,Q), % silt^-new (X,,) and the five sand
fractions (X - X_, inclusive). Later analysis including the combina-
tion variable % Fe + % Al (X ) did not alter the deletion process in
67
-------
Table 11. ESTIMATES OF PARTIAL REGRESSION COEFFICIENTS IN THE MULTIPLE LINEAR REGRESSION
.a
EQUATION OF KQbs WITH SURFACE SOIL VARIABLES
00
Independent
Variable
Name
Constant
M x 10"5
Permeability
Structure
% C-Na pyro.
7o mica
% clay
7o chlorite
7° amor. mat.
7, N
Coefficient of
determination (r )
Partial regression coefficients in the multiple linear regression
equation
x.b
Xo
X12
X14
X13
X16
X25
X2
X30
X27
X23
Step 1
-0 . 1344
6.3932
0.03780
0.03704
-0.3255
-0.01779
0.01253
-0.02060
-0.03592
0.5645
0,921
Step 2
-0.1382
6.9180
0.03291
0.04083
-0.1879
-0.01213
0.00949
-0.01438
-0.02857
0.912
Step 3
-0.1287
6.7004
0.03540
0.03624
-0.1992
-0.00366
0.00206
-0.00366
0.905
Step 4
-0.1395
6.5201
0.03639
0.03770
-0.1673
-0.00461
0.00177
0.903
Step 5
-0.1357
6 . 6044
0.03938
0.03532
-0.1603
-0.00109
0.899
Step 6
-0.1357
6.7097
0.03847
0.03448
-0.1732
0.899
Backward elimination of variables (WRAP) until all variables remaining are significant at the
0.05 probability level.
As defined in Table 10.
-------
the last six steps. Therefore, the model as proposed in step 6 would
be the best model to use to predict K for surface soils. The model
proposed in step 6 might best be written as:
Kpred " °-1357 + (6-710) (10) X12 + 0.03448X13 + 0.03847X14 - 0.1732X16
where the X are those reported in Table 10,
With the use of the SPSS regression program we arrive at the same con-
clusions as reported above. That is, with the forward selection of
independent variables to be added to the regression model on the basis
of highest partial correlation we arrive at the models reported in
Table 12. The variables included in the models up to and including
those of step 4 are statistically necessary, at the 0.05 probability
level, to adequately predict the dependent variable, K. The slight
differences in the partial regression coefficients in the models
arrived at by the two methods when the same independent variables are
included, such as step 6, Table 11 and step 4, Table 12, is attributed
to numerical accuracy differences of the two programs.
It is very interesting to note that the four most significant variables
to be included in the surface soil model to predict K are essentially
the same four variables that Wischmeier, Johnson and Cross (1971) use
in their nomograph to predict soil erodibility. The only difference
being the use of sodium pyrophosphate extractable carbon in the former
and percent organic matter in the latter. The sodium pyrophosphate
extractable carbon is highly correlated with the percent organic mat-
ter, r = 0.866. It is thought by the authors that the sodium pyro-
phosphate extractable carbon is more representative of that organic
material in soils which can affect intra-particle bonding in soils.
STATISTICAL ANALYSIS OF SUBSOIL DATA
The main problem in establishing a model for the prediction of subsoil
erodibility is the low number of K , values from subsoils. The corre-
lation matrix for the subsoil variables is reported in Table 17,
Appendix C. With only 7 sets of observations for the subsoils, any
69
-------
Table 12. ESTIMATES OF PARTIAL REGRESSION COEFFICIENTS IN THE MULTIPLE LINEAR REGRESSION
a
EQUATION OF
WITH SURFACE SOIL VARIABLES
Variable
Constant
M x 10"5
Permeability
Structure
% C-Na pyro.
% mont.
0.25-Q.l mm
sand
0.1-0,05 mm
sand
% N
% vermiculite
Coefficient of
determination
x.b
X12
X14
X13
X16
X28
X8
X9
X23
A-o /
/ (\
(r2)
Partial regression coefficients in
Step 1 Step 2 Step 3 Step 4
-0.00753 -0.1270 -0.1894 -0.1447
7.9616 6.7259 6.8147 6.8711
0.0472 0.0399 0.0376
0.0303 0.0338
-0.1571
0.736 0.871 0.888 0.897
the multiple linear regression equation
Step 5
-0.1503
6.9787
0.0343
0.0358
-0.1692
0.0048
0.902
Step 6
-0.2184
7.5961
0.0367
0.0337
-0.1232
0.0081
0,0016
0.908
Step 7
-0.2226
7.8378
0.0367
0.0343
-0.1204
0.0095
0.0027
-0.0037
0.915
Step 8
-0.2561
7.8661
0.0405
0.0297
-0.1850
0.0122
0.0035
-0.0039
0.3377
0.919
Step 9
-0.2613
8.1960
0.0382
0.0292
-0.2390
0.0118
0.0040
-0.0043
0.4167
0.0085
0.922
Forward selection of variables on basis of highest partial correlation of the variables that are
not subsequently in the model.
As defined in Table 10.
-------
model containing 6 independent variables will completely predict the
dependent variable from the 7 observations. Therefore, weighted
regression analysis was used in which each of the 46 surface soil
observations were weighted 1/46 or 0.021739 and the 7 subsoil were
weighted 1/7 or 0.142857.
The weighting and subsequent regression analysis was performed using
the weighted regression analysis programs (WRAP). The partial regres-
sion coefficients and the coefficients of determination for this
weighted model are reported in Table 13. It was hoped that by using
this type of approach we could determine those basic independent
variables that affect the prediction of K in both surface soil and
subsoils.
Comparison of the variables necessary to predict K in the combined
surface soil and subsoil weighted regression model of Table 13, with
those in the model developed in Tables 11 and 12 for surface soils
reveals that the amount of sodium citrate-bicarbonate-dithionite (CDB)
extractable Fe and Al, X, , and CDB extractable Si, X?., might be
important in the former model and are not even considered in the latter.
Therefore, the amount of CDB extractable Fe, Al and Si must be impor-
tant in predicting the K factor of subsoils. Such a model is shown
in Table 14 along with some additional models. On the basis of F-test
analysis the amount of sodium pyrophosphate extractable carbon is not
statistically necessary, at the 5 percent level, to predict K for high
clay subsoils. However, if the percent silicon released by CDB
extraction is included with M and the % Fe + % Al there is a statisti-
cal improvement in the prediction model, Table 14-Equation 7, at the
5 percent level. Therefore, the model
K , - 0.32114 + 20.167X10"5 X.. -0.14440X.. -0.83686X
pred 12 31 21
can be used to statistically predict the credibility of the high clay
subsoils considered in this study.
71
-------
N3
Table 13. ESTIMATES OF PARTIAL REGRESSION COEFFICIENTS IN THE WEIGHTED MULTIPLE LINEAR
REGRESSION EQUATION OF K WITH SURFACE AND SUBSOIL VARIABLES3
Independent
Variable
Name
Constant
% silt-old
% C-Na pyro.
% Fe203+%Al203
0.1-0.05 mm sand
Structure
1.0-0.5 mm sand
Permeability
% S102
% montmorillonite
0.25-0.1 mm sand
Coefficient of _
determination (r )
Partial regression coefficients in the multiple linear regression
equation
x.b
X4
X16
X31
X9
X13
X6
X14
X21
X28
X8
Step 1
-0.05513
0.00840
-0.61251
-0.10516
0.01290
0.06545
0,00442
0.03096
-0.29004
-0.00424
-0.00148
0.841
Step 2
-0.13859
0.00903
-0.58304
-0.09887
0.01288
0.06494
0.00547
0.03305
-0.24154
-0.00304
0.840
Step 3
-0.16775
0.00931
-0.58278
-0.09508
0.01401
0.06466
0.00676
0.03175
-0.31414
0.838
Step 4
-0.24337
0.01007
-0.57060
-0.08272
0.01590
0.06278
0.00902
0.02328
0.830
Step 5
-0.15458
0.01002
-0.67752
-0.09160
0.01441
0.08661
0.00724
0.820
Step 6
-0.00526
0.00889
-0.72662
-0.09438
0.01448
0.06929
0.805
Backward elimination of
0.05 probability level.
As defined in Table 10.
variables (WRAP) until all variables remaining are significant at the
-------
Table 14. ESTIMATES OF PARTIAL REGRESSION COEFFICIENTS IN THE MULTIPLE LINEAR REGRESSION
EQUATION OF K , WITH SUBSOIL VARIABLES3
obs
Independent
Variable b
Name i
Constant
M x 10"5 X12
% Fe2 3+, 2 3 x31
% C-Na pyro. X16
% C-total X15
% clay X2
7» montmorillonite X00
Zo
% amor, mat, X??
% si02 x21
Coefficient of
determination (r )
Partial
Equation
1
0.00367
23.001
-0.10839
0.904
regression
Equation
2
0.17565
19.496
-0.12085
-0.34917
0.921
coefficients
Equation
3
-0.09575
24.271
-0.10020
0.06213
0.905
in the multiple linear
Equation
4
0.05255
21.907
-0.10550
-0.00064
0.905
Equation
5
0 , 18608
20.204
-0.12241
-0.00929
0.933
regression
Equation
6
-0.07558
24.161
-0.10562
0.01006
0.909
equation
Equation
7
0.32114
20.167
-0 . 14440
-0.83686
0.950
The WRAP were used to obtain these values without using the backward elimination option.
As defined in Table 10,
-------
VII. A NOMOGRAPH FOR ESTIMATING THE ERODIBILITY
FACTOR, K, OF HIGH CLAY SUBSOILS
A nomograph which can be used to predict the erodibility factor, K, of
subsoils is shown in Figure 55. The nomograph was developed from the
multiple linear regression equation (Section VI) relating the erodi-
bility factor to the soil texture factor, M, the amount of CDB
extractable iron and aluminum oxides, and the amount of CDB extract-
able silica. The nomograph is similar to that developed for surface
soils (Wischmeier et al., 1971). The equation used to derive the
nomograph was:
K = 0.32114 + 2.0167 x 10~4 M - 0.14440 (% Fe00_ + % A100,)
pred / j f- J
- 0.83686 (% Si02)
The soil texture factor, M, is calculated arithmetically by summing
the percent "new" silt (2 - 100 ym mean diameter) and "new" sand (100 -
2000 ym mean diameter) then multiplying the sum by the percent "new"
silt.
Inspection of the nomograph reveals that for given levels of CDB
extractable iron plus aluminum oxide and silica, the erodibility factor,
K, increases with an increase in M. The soil texture factor, M, is in
turn influenced by the relative proportions of sand, silt and clay in
the sample. In high clay soils, M is generally low and a function of
the ratio of silt to sand. In sandy soils, M is also lower than in
more silty soils and behaves as a function of the clay to silt ratio.
The value of M increases as the square of the silt content when soils
grade from loamy sands to sandy loams. The same trend is evident when
soil textures grade from clays to silty clays to silty loams to silt
loams and to silts.
74
-------
E 0
o
I
o
o
0
z
<
l/>
40
>- 50
60
- 7O
Z 80
UJ
«J
Of
uj 90
10O
PERCENT
PERCE
(0.1-
+AI
'
VlT S
2 miVi
WD
0.5
20
15
10
-±
0 100O 2000 3000 4000
Factor M
A
71
T
ERCE
NT SiO,
0.4
O.3
0.2
O.I
o
0 O.2 O.4 0.6 0.8
SOIL EROOIBILITY FACTOR, K
Figure 55. Nomograph for Estimating the Erodibility Factor, K, of High Clay Subsoils
-------
It is also evident from the nomograph that at a given value of M there
is a decrease in K with an increase in the amount of iron and aluminum
oxides and silica in the sample. This finding is expected because
amorphous iron and aluminum hydrous oxides and hydrous silica are known
to bind soil particles into stable aggregates. Although the citrate-
dithionite-bicarbonate (CDB) reagent used to extract these soils removes
t
some of the crystalline iron oxides, it is most likely that almost all
of the iron extracted by CDB from the subsoil samples used in this study
was in the amorphous form. The CDB reagent will extract a very small
amount of aluminum and silicon from crystalline clay minerals. However,
the amounts of aluminum and silicon extracted from the subsoils used
in this study, suggest that the Al and Si is combined with the amorphous
iron oxide and is released when the iron is solubilized by the CDB
reagent.
The chemistry of iron and aluminum suggest that their reactions with
primary soil particles should be similar. This suggestion is supported
by the observation that the summation of CDB extractable iron and
aluminum oxides leads to a factor that serves as a better predictor of
K than does iron and aluminum oxides when used separately in the model.
The inclusion of CDB extractable silicon in this summation does not
statistically improve the prediction of K; however, when silica is
included as a separate term in the model along with Fe.O.. + Al-0, there
is a statistical improvement in the resulting regression model. This
suggests that CDB extractable iron and aluminum act similarly in serv-
ing as binding agents in subsoils and that CDB extractable .silica,
important in predicting the erodibility of high clay subsoils, reacts
differently with the primary soil particles than do iron and aluminum
oxides. Evidently the amorphous iron and aluminum oxides and silica
are the primary binding agents in subsoils, much as organic matter is
the primary binding agent in surface soils, which promotes the forma-
tion of soil aggregates that are resistant to raindrop detachability.
The nomograph is used by entering the sum of new silt (2 - 100 ym
diameter) on the left-most vertical axis and proceeding horizontally
76
-------
to intercept the curve for the appropriate sand content (100 - 2000 jam
diameter). From this point proceed vertically to intercept the
appropriate curve for the percent Fe20 + Al.O . Then proceed horizon-
tally to intercept the appropriate curve for the percent SiO» and from
there proceed vertically to the bottom horizontal axis. At the point
where the vertical line intersects the bottom right-hand horizontal
axis the value predicted for K may be read directly. The value for M
is not precomputed nor used directly, but rather may be calculated
indirectly in using the nomograph by drawing a vertical line from the
point of intersection of the "new" silt (2 - 100 ym diameter) horizon-
tal with the appropriate curve for "new" sand (100 - 2000 ym diameter)
to the bottom, left-hand, horizontal axis.
Table 15 gives a comparison of K values observed in field experiments
with seven subsoils and similar values computed with the nomograph.
A close relationship exists between the K values predicted by the nomo-
graph and those measured in field experiments. The difference in K
values between observed and predicted by the nomograph exceeds 0.03
only in the case of the St. Clair subsoil where the nomograph under-
estimates the observed K by 0.09.
The nomograph may be used with some degree of confidence with subsoils
similar to those used in this study because the study soils represented
a wide range in K values, textural composition, chemical composition,
clay mineralogy and pedogenic origin. It should be emphasized that
the nomograph may not be applicable to subsoils having large amounts
of iron and aluminum in crystalline forms which are extractable by the
CDS reagent. The iron and aluminum in such subsoils may not be active
in binding soil particles into aggregates which resist raindrop dis-
ruption. The nomograph would therefore underestimate the K for sub-
soils containing large amounts of crystalline iron and aluminum com-
pounds since the nomograph is based on the finding that CDB extractable
iron, aluminum and silicon content of soils is inversely related to the
soil erodibility factor. The nomograph may not be applicable to sub-
soils that have a structure other than blocky or massive and a permea-
bility other than very slow. It would be necessary to investigate
77
-------
Table 15. COMPARISON OF THE SOIL ERODIBILITY FACTOR, K,
DETERMINED IN FIELD EXPERIMENTS AND THOSE
COMPUTED FROM THE SUBSOIL NOMOGRAPH
Soil
Dayton
McGary
Portageville
St. Clair
Pawnee
Mayberry
Wymore
Sample ID
number
188
191S
192S
212S
206S
20 8S
210S
Soil erodibility factor, K
Observed
.54
.36
.05
.48
.45
.67
.49
nomograph*
.57
.39
.08
.39
.42
.69
.49
aSoil erodibility factor, K, as determined from Figure 55.
78
-------
subsoils that have other structures and permeabilities to determine
the exact effect of these two variables on the erodibility of subsoils.
The use of this nomograph to predict erosion at a proposed construc-
tion site should include the following steps:
1. Composite subsoil samples should be collected which represent
each subsoil horizon and/or topographical characteristic pro-
posed for prolonged exposure during the construction phase.
2. These subsoil samples should be submitted to a soil characteriza-
tion laboratory for particle size analysis (silt, 2 - 100 pm and
sand, 100 - 2000 pm) and determination of iron, aluminum and
silicon released by citrate-dithionite-bicarbonate (CDB) extrac-
tion. The particle size analysis can be performed on the same
sample used for the CDB extraction following the procedure of
Jackson, 1956.
3. The results from these analyses can then be used to predict the
soil erodibility factor, K, for each of the areas involved by
using the nomograph presented in Section VII of this report.
4. Soil loss values can then be estimated using the Universal
Soil-Loss Equation using procedures outlined by Wischmeier and
Smith, 1965.
79
-------
VIII. REFERENCES
1. Alexides, C. A. and M. L. Jackson. Quantitative Determination of
Vermiculite in Soils. Soil Sci. Soc. Amer. Proc. 29:522-527, 1965.
2. Chapman, S. L. and M. L. Jackson. Quantitative Clay Mineralogy of
Selected Soils from North Central United States. Agronomy Abstracts.
p. 131, 1969.
3. Draper, N. R. and H. Smith. Applied Regression Analysis. New
York, Wiley, 1966. 407 p.
4. Foster, G. R., L. F. Huggins and L. D. Meyer. Simulation of Over-
land Flow on Short Field Plots. Water Resour. Res. 4:1179-1187,
1968.
5. Hashimoto, I. and M. L. Jackson. Rapid Dissolution of Allophane
and Kaolinite-Halloysite After Dehydration. Clays Clay Minerals.
7:102-113, 1960.
6. Jackson, M. L. Soil Chemical Analysis-Advanced Course (1st Edition,
3rd Printing 1967) Published by the author. Department of Soil
Science, University of Wisconsin, Madison, WI, 53706, 1956.
7. Mebius, L. J. A Rapid Method for Determination of Organic Carbon
in Soils. Anal. Chim. Acta. (Amsterdam). 22:120-124, 1960.
8. Meyer, L. D. Use of the Rainulator for Runoff Plot Research.
Soil Sci. Soc. Amer. Proc. 24:319-322, 1960.
9. Meyer, L. D. and D. L. McCune. Rainfall Simulator for Runoff
Plots. Agr. Eng. 39:644-648, 1958.
10. Nelson, D. W. and L. E. Sommers. A Simple Digestion Procedure
80
-------
for Estimation of Total Nitrogen in Soils and Sediments. J.
Environ. Quality. 1:423-425, 1972.
11. Nie, N. H., D. H. Bent, and C. H. Hull. SPSS-Statistical Package
for the Social Sciences, New York, McGraw-Hill, 1970. 343 p.
12. Roth, C. B., M. L. Jackson, and J. K. Syers. Deferration Effect
on Structural Ferrous-Ferric Iron Ratio and CEC of Vermiculities
and Soils. Clays Clay Minerals. 17:253-264, 1969.
13. Sommers, L. E. and D. W. Nelson. Determination of Total Phos-
phorous in Soils: A Rapid Perchloric Acid Digestion Procedure.
Soil Sci. Soc. Amer. Proc. 36:902-904, 1972.
14. Wischmeier, W. H. Upslope Erosion Control In: Environmental
Impact on Rivers. Fort Collins, Colorado, H. W. Shen, 1972.
p. 15-1 to 15-26.
15. Wischmeier, W. H. and J. V. Mannering. Relation of Soil Proper-
ties to Its Erodibility. Soil Sci. Soc. Amer. Proc. 33:131-137,
1969.
16. Wischmeier, W. H. and D. D. Smith. Predicting Rainfall-Erosion
Losses from Cropland East of the Rocky Mountains. Agr. Handbook
No. 282. U.S. Gov't Printing Office, Washington, D.C. 1965.
17. Wischmeier, W. H., C. B. Johnson and B. V. Cross. A Soil Erodi-
bility Nomograph for Farmland and Construction Sites. J. Soil
Water Conserv. 26:189-193, 1971.
81
-------
IX. APPENDICES
Page
A. Profile Descriptions of Soils Used in This Study 83
B. Methods Modified or Developed and Used During the
Course of This Study 97
C. Correlation Matrix of Surface Soil and Subsoil
Variables 105
82
-------
APPENDIX A
PROFILE DESCRIPTIONS OF SOILS USED IN THIS STUDY
83
-------
Soil type; McGary silt loam
Location: 3.9 mi. north of Bloomington, Indiana, along highway 37,
approximately 150 feet west (TW 9, N R 1, W sect. S4, NW 1/4)
Land Use: pasture
Parent Material: calcareous lacustrine
Physiography: lake bed
Slope: 2%
Salt or Alkali: none Permeability: slow to very slow
Stoniness: none Drainage: somewhat poorly
Erosion: none Ground water: deep
Aspect: south
Root Distrib.: none
Profile description:
Ap -- 0-11" -- Gray (10YR 5/1-6/1) silt loam; fine cloddy somewhat
compact and massive; friable; neutral; abrupt
smooth boundary. (6 to 11 inches thick.)
B21gt -- 11-15" -- Grayish-brown (10YR 5/2) light silty clay, common,
fine, faint gray (10YR 6/1) mottles; moderate
medium subangular blocky structure; firm; medium
acid; clear smooth boundary. (2 to 5 inches thick.)
B22gt -- 15-22" -- Grayish-brown (10YR 5/2) silty clay, common fine
faint yellowish-brown (10YR 5/2) mottles; weak,
fine soft black (10YR 2/1) manganese and iron
oxide concretions; thin gray (10YR 5/1) clay
films on ped faces; firm; neutral; clear smooth
boundary. (5 to 10 inches thick.)
B23gt -- 22-27" -- Grayish-brown (10YR 5/2) silty clay, common fine
faint yellowish-brown (10YR 5/4) mottles; moderate
medium prismatic structure, breaking to moderate
to strong medium angular blocky structure; thin
gray (10YR 5/1) clay films on ped faces; tongues
a
Profile description taken from Soil Survey.
84
-------
McGary silt Loam (cont'd.)
of gray (10YR 5/1) silty clay, 1 to 2 inches thick
and 6 to 10 inches apart, firm; mildly alkaline;
gradual irregular boundary. (4 to 10 inches thick.)
B3gt -- 27-39" Gray (10YR 5/1) light silty clay, common fine
distinct light yellowish-brown (10YR 6/4) mottles;
moderate to strong, fine to medium prismatic
structure, breaking to moderate to strong, medium
blocky structure; firm; mildly alkaline; clear
irregular boundary. (9 to 15 inches thick.)
Cg -- 39-50" -- Gray (10YR 6/1) stratified silty clay loam and
clay, common fine distinct yellowish-brown (10YR
5/6) mottles; moderate coarse blocky structure;
firm; thin gray (10YR 5/1) clay films on ped
faces; calcareous; tongues of the B3 horizon pro-
ject into this horizon.
85
-------
Soil type: Portageville clay
Location: Delta Experiment Station, Portageville, Mo.
Land Use: Cultivated
Parent Material: alluvial deposit
Physiography: backwater deposit
Slope: 0%
Salt or Alkali: none Permeability: very slow
Stoniness: none Drainage: poor
Erosion: none Ground Water: shallow
Aspect: none
Root distrib.: none
a
Profile description;
Ap __ o- 6" Very dark grayish brown (10YR 3/2) clay with few,
fine, dark reddish-brown (5YR 3/3) mottles; weak,
fine to medium, granular structure; very firm;
pH 7.0; abrupt, smooth boundary.
Al -- 6-15" -- Very dark gray (10YR 3/1) clay with common, medium,
dark reddish-brown (SYR 3/4) mottles; moderate, fine
to medium, subangular blocky structure; very firm;
few fine roots and pores; Ph 7.5, weakly calcareous;
gradual, smooth boundary.
Bl -- 15-25" -- Dark-gray (5Y 4/1) clay with common, medium, dark
reddish-brown (SYR 4/3) mottles and few, coarse,
yellowish-red (SYR 4/6) mottles; moderate, medium,
angular blocky structure; very firm; few polished-
surfaces 1 to 4 inches across in angular position
(slickensides); pH 8.0, weakly calcareous; gradual,
smooth boundary.
B2 -- 25-47" -- Gray (5Y 5/1) clay with few, fine, dark reddish-brown
(2.5YR 2/4) mottles and common, fine, dark reddish-
brown (SYR 3/4) and yellowish-red (SYR 4/8) mottles;
weak, medium, angular blocky structure; very firm;
a
Profile description taken from Soil Survey, Pemiscot County, Missouri,
USDA-SCS, 1971.
86
-------
Portageville clay, Po (cont'd.)
many ped interiors of dark grayish brown (2.5Y 4/2);
few polished surfaces 1 to 4 inches across in angular
position (slickensides); pH 8.0, strongly calcareous;
gradual boundary.
C -- 47-60" -- Dark gray (5Y 5/1) and dark grayish-brown (10YR 4/2)
clay with common, fine, dark reddish-brown (SYR 3/4)
and dark-brown (7.SYR 4/4) mottles; massive, strat-
ified; strata are 1/2 inch to 2 inches thick; dark
reddish-brown (SYR 3/4) and yellowish-red (SYR 4/8)
colors appear as thin lines between the strata, and
there are occasional thin lenses (about 1/2 inch
thick) of very fine sandy loam; very firm; pH 8.0,
strongly calcareous.
At a depth between 10 and 40 inches, these soils have a clay content of
40 to 65 percent or more.
The A horizon ranges mainly from 10 to 20 inches in thickness, but is as
much as 24 inches in places. Color ranges from 10 YR to 2.5Y in hue,
2 to 3 in value, and 0 to 2 in chroma. The texture is commonly clay
and silty clay loam, but there are areas of sandy loam or loamy sand
overwash.
The Bl horizon ranges from 10YR to yellower in hue, has a value of 3.5
to 4.5, and has a chroma of 1. The B2 and C horizons range from 10YR to
yellower in hue and have a value of 4 to 6 and a chroma of 1. Ped in-
teriors have a chroma of 1 or 2.
Reaction ranges from slightly acid to moderately alkaline in the top 10
inches, and from neutral to moderately alkaline below. The soil mater-
ialin the top 10 inches is calcareous in some places, and it ranges to
violently calcareous below a depth of 10 inches.
NOTES: This soil occupies level and depressional areas on lakebeds and
meanders of former channels in the recent Mississippi River
87
-------
Portageville clay, Po (cont'd.)
flood plain. These areas are like oxbows in shape, are several
hundred acres in size, and are in the eastern part of the
countyo
Classification: fine, montmorillonitic, calcareous thermic, vertic,
haplapuoll
88
-------
Soil type; Mayberry clay loam, F2G2D
Location: .16 mi. E and 350' N from SW corner, Sec. 35, T 7 N, R 10 E
Otoe County
Land use; cultivated
Parent material: stratified clay loam, sandy loam and clay (Fullerton
formation?)
Physiography: upland
Slope: 18%
Salt or Alkali: none Permeability: slow
Stoniness: none Drainage: moderately well
Erosion: moderate Ground water: very deep
Slope: north
Root distrib.: none
Profile description:
Alp -- 0- 8" -- Dark brown (7.5YR 3/2) moist; clay loam; weak
S72Neb-66-l-l coarse blocky parting to weak coarse to fine granu-
lar structure; hard, friable; medium acid (pH 5.6);
clear smooth boundary.
SB21t -- 8-20" -- 50% dark brown (7,SYR 3/2) and 50% reddish-brown
S72Neb-66-l-2 (5YR 4/4) moist; clay; weak coarse blocky parting
to strong medium and fine angular blocky structure;
extremely hard, very firm; slightly acid (pH 6.5);
gradual smooth boundary.
B22t --20-40" -- Brown (10YR 5/3) moist; silty clay to clay; moderate
S72Neb-66-l-3 medium angular blocky structure; extremely hard,
very firm; neutral (pH 7,0); gradual smooth boundary.
°B23t 40-53" Dark brown (7.SYR 4/4) moist; clay; moderate medium
S72Neb-66-l-4 and coarse angular blocky structure; extremely hard,
very firm; clear smooth boundary.
dC --53-65" -- Yellowish brown (10YR 5/4) moist; stratified clay
S72Neb-66-l-5 loam and sandy loam; weak coarse blocky structure;
hard, friable.
89
-------
Mayberry clay loam, F2G2D (cont'd.)
NOTES: This soil thought to be formed in Paleosol or clayey sediments
of Fullerton formation. 100' of crest has reddish brown color,
heavy clay loam or silty clay loam in upper 2'. Moved down-
slope for erosion study because of slope and testure. This
profile possibly taxadjunct because of brown color and is part
of Mayberry mapping unit. Thick clay films in lower B and C
could be characteristic of the parent material, as this has
been observed elsewhere.
Classification: fine, montmorillonitic, mesic family of Aquic
ArgiudolIs.
Few dark organic coatings, common, small, hard, round dark
accumulations, thin patchy clay film.
Few large dark organic coatings along cracks, few faint dark
accumulations, common distinct brown mottles (7.5YR 4/4) thin patchy
clay films.
c
Common soft, diffuse dark accumulations, thick discontinuous dark
brown clay films fill pores and channels, few hard lime concretions,
few gravel.
d
Many thick dark brown clay films fill channels.
e
Courtesy of Mr. H. Sauter, District Soil Scientist, USDA-SCS, Otoe
County, Syracuse, Nebraska.
90
-------
Soil type: Pawnee clay loam, F2GD
Location: .5 mi. N, .3 mi. W and 580' S from SE corner, Sec. 27-7-9
Otoe County
Land use: cultivated
Parent material: calcareous glacial drift
Physiography: upland
Slope: 8%
Salt or Alkali: none Permeability: slow
Stoniness: none Drainage: moderately well
Erosion: moderate Ground water: very deep
Aspect: east
Root distrib.: none
g
Profile description:
Alp -- 0- 7" -- Very dark brown (10YR 2/2) moist; clay loam, weak
S72Neb-66-2-l coarse blocky parting to weak coarse to fine
grandular structure; hard, friable; slightly acid
(pH 6.4); clear smooth boundary.
B21t 7-18" Very dark grayish-brown (10YR 3/2) moist; clay;
S72Neb-66-2-2 moderate, medium angular blocky structure; extremely
hard, very firm; neutral (pH 6.8); gradual smooth
boundary.,
B22t --18-25"
S72Neb-66-2-3
Dark brown (10YR 4/3) moist; clay; moderate coarse
and medium angular blocky structure; extremely
hard, very firm; neutral (pH 7.2); gradual smooth
boundary.
"B23t __25-40" -- Grayish-brown (10YR 5/2) moist; clay; moderate
S72Neb-66-2-4 coarse and medium angular blocky structure; extremely
hard; very firm; gradual smooth boundary.
JB3 --40-60" -- Grayish-brown (2.5Y 5/2) moist; heavy clay loam;
S72Neb-66-2-5 moderate medium angular blocky structure; very hard;
firm.
91
-------
Pawnee clay loam, F2Gd (cont'd.)
NOTES: This is moderately eroded soil (about 50 x 50') included in
mapping unit, Pawnee soils, severely eroded. (Slope in this
general area of mapping unit averages about 7%.)
Classification: fine, montmorillonitic, mesic family of Aquic
Arguidolls.
ft
Few brown (7.SYR 4/4) mottles, thin discontinuous clay films.
Few brown (7.5YR 4/4) mottles, many very dark brown (10YR 2/2)
organic coatings along cracks, thin discontinuous clay films.
Common yellowish brown (10YR 5/4) mottles, common organic coatings,
common hard lime concretions, thin discontinuous clay films.
Many prominent dark accumulations (manganese), few brown accumulations
(iron), few soft lime concretions, thin patchy clay films.
Courtesy of Mr. H. Sauter, District Soil Scientist USDA-SCS, Otoe
County, Syracuse, Nebraska.
92
-------
Soil type; Wymore silty clay loam, F2DC
Location: .5 mi. N, .45 mi. W and 130' S from SE corner, Sec. 27,
T 7 N, R 9 E, Otoe County
Land use: cultivated
Parent material: silty clay loam, loess
Physiography: upland
Relief: 5 1/2%
Salt or Alkali: none Permeability: slow
Stoniness: none Drainage: moderately well
Erosion: moderate Ground water: very deep
Aspect: east
Root Distrib.: none
Profile description:
Alp -- 0- 8" -- Very dark brown (10YR 2/2) moist; silty clay loam;
S72Neb-66-3-l weak coarse blocky parting to weak coarse to fine
granular structure; hard, friable; slightly acid
(pH 6.1); clear smooth boundary.
aB21t -- 8-19"-- Mixed very dark brown (10YR 2/2), dark grayish-
S72Neb-66-3-2 brown (10YR 4/2), and very dark grayish-brown
(10YR 3/2) crushed, moist; silty clny; strong
medium and fine angular blocky structure; extremely
hard, very firm; neutral (pH 6.6); gradual smooth
boundary.
B22t 19-35" Dark grayish-brown (2.5Y 4/2) moist; silty clay;
S72Keb-66-3-3 strong medium and fine angular blocky structure;
extremely hard; very firm; neutral (pH 7»0);
gradual smooth boundary.
°B3t 35-48" Mixed grayish-brown (2.5Y 4/2) and grayish-brown
S72Neb-66-3-4 (2.5Y 5/2) moist; heavy silty clay loam; moderate
medium angular blocky structure; very hard, firm;
gradual smooth boundary.
93
-------
Wymore silty clay loam, (cont'd.)
C 48-60" -- Grayish-brown (2.5Y 5/2) moist; silty clay loam;
S72Nev-66-3-5 weak coarse and medium blocky structure; hard,
friable.
Classification: fine montmorillonitic, mesic family of Aquic
Argiudolls.
o
Thin patchy clay films.
Few faint small brown (7.SYR 4/4) mottles, common organic coatings,
thin discontinuous clay films.
Many prominent large brown (7.SYR 4/4) mottles, few pinhead-size
dark accumulations, many small pores, common lime concretions,
,patchy clay films.
Many prominent diffuse strong brown (7.SYR 5/6) mottles, many large
and small pores.
6Courtesy of Mr. H. Sauter, District Soil Scientist USDA-SCS, Otoe
County, Syracuse, Nebraska.
94
-------
Soil type; St. Clair silty clay loam
Location: S13, T 31 N, R 14 E, Allen County, Indiana
Land use: cultivated
Parent material: calcareous till
Physiography: upland
Slope: 16%
Salt or Alkali: none Permeability: slow to very slow
Stoniness: none Drainage; moderately well
Erosion: moderate Ground water: deep
Aspect: east
Root Distrib.: none
pi
Profile description:
Ap -- 0- 5" Dark grayish-brown (10YR 4/2) silt loam; weak, fine,
granular structure; friable when moist; neutral;
abrupt, smooth boundary.
Bl -- 5- 7" -- Brown (10YR 5/3) light silty clay loam; moderate,
fine, subangular blocky structure; firm when moist;
medium acid; clear, wavy boundary.
B21t -- 7-14" -- Brown (10YR 5/3) silty clay; weak, medium, prismatic
structure breaking to strong, medium, angular blocky
structure; very firm when moist; thin clay films on
many ped faces; very strongly acid; clear, wavy
boundary.
B22t -- 14-24" Brown (10YR 5/3) clay; few, fine, faint mottles of
yellowish brown (10YR 5/8) in the upper part and
common, medium, distinct mottles of yelloxtfish brown
in the lower part; weak, medium, prismatic structure;
extremely firm when moist; thin to thick clay films
on many ped faces; strongly acid; clear, wavy
boundary.
Profile description taken from Soil Survey, Allen County, Indiana,
USDA-SCS, 1969.
95
-------
St. Glair silty clay loam (cont'd.)
C -- 24-40" -- Dark grayish-brown (10YR 4/2) heavy silty clay loam
to silty clay; common, medium, prominent mottles
of light gray (10YR 7/2); weak, medium, angular
blocky structure; very firm when moist; calcareous.
The A horizon ranges from 5 to 10 inches in thickness and from silt
loam to silty clay loam in texture. Soil material from the Bl horizon
is mixed with that of the A horizon in many cultivated areas. In some
places there is no Bl horizon. The depth to calcareous material
ranges from 18 to 28 inches.
Classification: fine, illitic, mesic, typic hapludalf
96
-------
APPENDIX B
METHODS MODIFIED OR DEVELOPED AND USED
DURING THE COURSE OF THIS STUDY
97
-------
SODIUM CITRATE - SODIUM BICARBONATE - SODIUM DITHIONITE EXTRACTION
OF SOILS
Free iron oxides are extracted from soils by a two step procedure in
which organic matter is first removed with hydrogen peroxide and iron
oxides are extracted with the CDB reagent.
A. Removal of organic matter with hydrogen peroxide
1. Add 10 ml increments of 30$ HO to a 5 g soil sample in a
600-ml beaker and digest until organic matter is destroyed.
Evaporate the sample to a thin paste.
2. Transfer the sample to a 100-ml centrifuge tube using 0.5 N
NaOAc (pH5).
3. Wash twice with 50 ml portions of .5 N NaOAc (pH 5),
centrifuge at 2000 rpm, and decant.
k. Wash once with 0.5 N NaOAc (pH 7), centrifuge at 2000 rpm and
decant.
5. Wash once with 0.5 N NaCl, centrifuge and decant.
B. Removal of iron oxides
1. Add kd ml of sodium citrate-bicarbonate solution to each tube.
2. Heat the suspension in the tube to 75 - 80 C in a water
bath. (Do not exceed 80 C or ferrous sulfide will form.
Suspension should be a light gray. Black indicates ferrous
sulfide.)
3. Add approximately 1 g of solid sodium dithonite (Na S 0, ) to
each tube. Stir constantly for 1 minute and allow to react for
5 minutes. Keep the temperature between 75 - 80° C.
h. After 5 minutes add an additional gram of sodium dithionite,
stir and allow to react for five minutes.
5. Repeat step k once more.
6. Centrifuge at 1600 - 2000 rpm for 5 minutes and decant.
IMPORTANT. Place supernatant liquid in a 250-ml volumetric
flask and save for iron determination.
98
-------
7. Wash twic£ with 0.5 N NaCl, centrifuge and decant the liquid
into the 250-ml volumetric flask.
8. Wash once with 30 ml of water, centrifuge and decant liquid
into volumetric flask.
Determination of Fe in Na S 0, -Na Citrate-NaHCO~ extract
1. Place appropriate aliquot of an air-oxidized sample (containing
less than UOO mg of Fe) in a 100-ml volumetric flask.
2. Add 2 ml 10$ NH2OH-HC1 (hydroxylamine hydrochloride) .
3. Add 5 ml of NaOAc-HCl buffer.
k. Dilute to about 80 ml and mix.
5. Add 3 ml °f 0.4$ orthophenanthroline in 95$ ethanol and mix well
(There must be at least 8 mg of orthophenanthroline present in
the final color development flask).
6. Dilute to volume with water.
7. Read at 510 my after 10 min. Color is stable at least 20 hours.
NaOAc-HCl buffer
81.6 g NaOAc
369 ml glacial HOAc
41.8 ml cone. HC1
Make to 1 liter with water
D. Determination of low concentrations of Al in Na S 0^-Na citrate-
NaHCO,, extracts
1. Place an appropriate aliquot of air-oxidized extract containing
2-1*0 Ug Al in a 50 ml beaker.
2. Add 10 ml concentrated
».
3. Add 3 ml concentrated H SO. .
U. Mix.
5. Heat to dryness on hot plate then ignite at UOO C for 3 hours.
6. When cool add 1 ml 6NHC1 and 20 ml water.
7. Digest on steam plate 30 min. then allow to cool.
99
-------
8. Neutralize to pH U.O with IN NaOH (determine with pH meter).
9. Add 2 ml 1% thioglycolic acid.
10. Digest 30 min. on steam plate.
11. Cool and add 10 ml aluminon-acetate buffer, transfer to 50 ml
volumetric flask.
12. Allow to stand 1 hour and read at 530 my
Aluminon-acetate buffer
120 ml glacial HOAc - 900 ml ly)
2U g NaOH
Mix
0.35 g aluminon
Dilute to 1 liter with water
E. Determination of Si in Dithionite-Citrate-Bicarbonate Extracts*
1. Place 5 to 10 ml of an air-oxidized extract (containing 5 to 25
g Si) in a 50 ml volumetric flask.
2. Add 10 ml of 1 N H SO, and mix.
-2
3. Add 10 ml of 0.3 M MoO, and mix.
k. After 2 min. add 5 ml of 20$ tartaric acid.
5. Add 1 ml of l-amino-2naphthol-^-sulfonic acid reductant reagent.
6. Dilute to volume with distilled water.
7. Measure absorbance at 820m V after 30 min.
1 N HSO
Dilute 29 ml concentrated H So, to 1 liter.
0.3 M molybdate
Dissolve 54.0 g of ammonium molybdate, (ML ),-MoO 0 , * h HO
in approximately 800 ml of HO, adjust to pH 7.0 with
5 N NaOH and make to 1 liter volume.
tartaric acid
Dissolve 100 g of tartaric acid in 500 ml of H 0.
*The citrate dithionite extract, diluted to 500 ml, is assumed to contain
12 mmoles of Na citrate, 5 mmoles of NaHCO_ and 17.5 mmoles of oxidized
Wa dithionite. -*
100
-------
l-ajnino-2-naphthol-^-sulfonic acid reductant
Dissolve 25 g of NaHSO in 200 ml of HO. Dissolve 2 g of
anhydrous Na SO and O.k g of l-amino-2-naphthol-U-sulfonic
acid in 25 ml HO. Combine the two solutions, dilute to
250 ml and store in a refrigerator in plastic.
F. Determination of hot water extractable carbon in soils
1. Add 10 g of <2 mm soil to 125 ml 2k/kd Standard Taper Erlen-
meyer flask.
2. Add 20 ml distilled water.
3. Boil under reflux for 30 min.
k. Rinse down condenser with small amount of HO.
5. Transfer contents of Erlenmeyer flask to centrifuge tube (50 ml)
with small amount of HO.
6. Centrifuge at approximately 10,000 rpm for 10 min.
7. Decant supernatant into 50 ml volumetric flask.
8. 'Wash residue with 10 ml of hot water.
9. Centrifuge as before, decant and add washing to 50 ml volumetric
flask.
10. Make to 50 ml final volume.
11. Take 5 ml aliquot for carbon determination add to 125 ml
Standard Taper Erlenmeyer flask.
12. Add 5 ml of 0.2 N K_Cr 0 solution.
c. f- I
13. Add 15 ml concentrated H SO..
ih. Boil under reflux for 30 min.
15. Remove from heat and cool sample.
16. Titrate sample with approximately 0.05 W f errous ammonium
sulfate solution using N-phenylanthanilic acid as an indicator.
17. Have boiled and unboiled blanks using 5 ml distilled HO + 5 ml
0.5 N K2Cr20? + 15 ml H^O^.
18. Calculate data as described by. Mebius (1960).
G. Determination of sodium pyrophosphate extractable carbon in soils
1. Weigh 5 g samples of < 2 mm soil into 50 ml centrifuge tubes.
101
-------
2. Add exactly 25 ml of 0.15 M sodium pyrophosphate adjusted to
pH 8.0.
3. Stopper tube and shake overnight approximately 16 hours.
U. Remove stoppers and centrifuge at 10,000 rpm for 10 minutes.
5. Decant supernatant.
6. Remove 5 ml aliquots of supernatant and place into 125 ml
Standard Taper Erlenmeyer flasks.
7. Add 5 ml of 0.5 N KgCr^ + 15 ml concentrated H^O^.
8. Boil under reflux for 30 min.
9. Remove from heat and cool sample.
10. Titrate sample with 0.1 N Ferrous ammonium sulfate solution
using N-phenylanthranillic acid as redox indicator.
11. Have boiled and unboiled blanks using 5 ml distilled HO plus
5 ml 0.5 W K Cr 0 plus 15 ml concentrated H SO, .
12. Calculate results as described by Mebius (1960).
H. Periodate oxidizable polysaccharides in soil
Reagents:
1. 0.05 M (0.1 N) sodium periodate - Dissolve 10.695 g of UalO. in
distilled HO and dilute to 1 liter.
2. 0.1 M (0.2 N) sodium arsenite - Dissolve 12.990 g of NaAs00 and
(L.
k g of NaHCO_ in water and dilute to 1 liter.
3. 0.05 M ().05 N) iodine solution - Dissolve 20 g of iodate-free
KI in 40 ml of water. Add 6.35 g of sublimed iodine to the KI
solution, stopper, shake and bring to 1 liter after iodine
dissolves.
h. Starch solution - Make a paste of 1 g of soluble starch and a
small amount of water. Pour the paste into 100 ml of boiling
water and boil for 1 minute. Allow the solution to cool and
add 3 g of KI. Use two ml aliquots for solution for titration.
Procedure:
Weigh out 2.5 g soil samples and place into small plastic bottles or
50 ml Erlenmeyer flasks. Add exactly 25 ml of 0.05 M periodate
102
-------
solution and place on a wrist action shaker for 2k hours (use very
show speed). Centrifuge sample and place a 15 ml aliquot of
supernatant into a 125 ml Erlenmeyer flask. Add 1 g of NaHCO_ and
10 ml of 0.1 M NaAsOp solution. Wash down sides of flask and add
2 ml of starch solution. Titrate the mixture with 0.05 M iodine
solution. Standardize the iodine solution by titration against
10 ml of arsenite solution (titration about Uo ml) and standardize
periodate solution by adding 10 ml of arsenite solution to 15 ml of
periodate and titrating the resultant mixture with iodine to starch
end point. Always use a blank having 25 ml of periodate with your
sample set and take 15 aliquot of this blank for subsequent titra-
tion.
Report results as millimoles of periodate consumed per gram of soil.
I. Procedure for the determination of mica in soil clays
1. A 100 mg clay sample that has been freshly treated with NaOAc
buffer pH 53 H?0~ and NaOAc pH 7 is placed in a 15-ml centrifuge
tube.
2. Wash five times with 1 M (ffi^) CO .
3. Transfer to a tared platinum crucible with water and dry
overnight at 110 C.
k. Weigh crucibles after cooling 5 minutes in desiccator.
5. To the sample add several drops of water, .5 ml 60% perchloric
acid and 5-10 ml hydrofluoric acid.
6. Cover 9/10 of crucible with platinum lid and place on 200-2ifO°C
sand bath and evaporate to dryness.
7. Repeat steps 5 and 6.
8. Remove and cool crucible.
9. Add 5 ml of 6 N HC1 to the crucible and place on a porcelain
plate and return to the 2^0° C hot plate for 1/2 hour.
10. Remove and cool for 10 minutes.
11. Transfer contents of crucible to 100-ml volumetric flask and
dilute to volume with distilled water.
103
-------
12. Determine K with flame photometer -
13. % mica = % K 0 x 10.
J. Determination of quartz plus feldspar content of soil clays
1. Weigh 0.200 g of dried clay sample into a 50-ml vitreous
silica crucible containing about 5 g NaHSO. (fused).
2. Mix sample with salt then add 7-10 g of NaHSO, (fused).
3. Fuse under hood.
h. Transfer cake to 150-ml beaker with 3 N HC1 (approximately 60 ml)
boil gently then transfer suspension to centrifuge tubes.
Centrifuge and discard solution.
5. Wash 2 times with 3 N HC1.
6. Transfer residue from tube with 0.5 N NaOH to Ni-beaker. Make
up to 100 - 150 ml of 0.5 N NaOH.
7. Bring suspension rapidly to boiling, boil for 25 minutes and
cool in ice bath.
8. Transfer to plastic centrifuge tubes with 3 N HC1 and centrifuge.
9. Transfer residue to weighed l8-ml centrifuge tubes.
10. Wash 3 times with 3 N HC1.
Dry at 110° C and weigh
feldspar in the sample.
11. Dry at 110 C and weigh. Residue is amount of quartz plus
104
-------
APPENDIX C
CORRELATION MATRIX OF SURFACE
SOIL AND SUBSOIL VARIABLES
105
-------
Table 16. CORRELATION MATRIX OF SURFACE SOIL VARIABLES
(coefficient of correlation - r)
X.
i
2
3
4
5
6
7
8
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
x~
2
1.0000
xio
-.6232
.9876
- . 9045
,5540
.8276
,9497
.8813
.4068
1.0000
X18
.6151
-.4523
-.2231
-,,0653
-.2685
-.4107
-.4960
-.2250
-.4507
.1937
-.0081
o0876
.0552
.6437
.4332
.3275
x^
3
-.6184
1.0000
Xll
.2331
-.8901
.9800
-.5377
-.7684
-.8693
-.7698
-.3527
-.9039
1.0000
X19
.5657
-.6096
.4422
-.2097
-.4410
-.5689
-.5718
-.3552
-.5977
.4051
.2078
.1183
.1521
.0556
-.1072
-.2184
X,
4
.2684
-.9214
1.0000
X12
-.0843
-.7035
.9081
-.4417
-.6327
-.7138
-.5857
-.2415
-.7214
.9467
1.0000
X20
.5883
-.7437
.6013
-.3271
-.5911
-.6825
-.6775
-.4157
-.7325
.5668
.3685
.1671
.2305
.1935
.0381
-.0600
X
5
-.2998
.5187
-.4994
1.0000
X13
.1999
-.1635
.1054
-.1226
-.2258
-.2403
-.0622
-.0015
-.1779
.1161
.0774
1.0000
X21
.8069
-.3642
.0703
-.1330
-.2754
-.3275
-.3904
-.1335
-.3719
.0463
-.1993
.0486
.2896
.5230
.3063
.6464
X,
6
-.4893
.7836
-.7253
.7851
1.0000
X14
.3375
-.4647
.4214
-.2484
-.3973
-.4643
-.4039
-.1849
-.4718
.4222
.3402
.4021
1.0000
X22
.6728
-.3405
.0903
-.1676
-.2371
-.2722
-.3641
-.1771
-.3378
.0582
-.1614
.0066
.0994
.5967
.3558
.4703
X-,
7
-.5713
.9162
-.8421
.4693
.8265
1.0000
X15
.5205
-.3174
.1208
-.0012
-.2176
-.2701
-.3633
- . 1484
-.3181
.0986
-.0573
.1175
-.0230
1.0000
X23
.1885
-.3752
.3275
-.0144
-.2357
-.3530
-.3810
-.2315
-.3655
.3075
.2441
.1541
-.1624
.5934
.5867
.0713
X0
8
-.5901
.9085
-.8200
.2565
.4904
.7501
1.0000
X16
.2249
-.1698
.0886
.1502
-.0474
-.1212
-.2394
-.1611
-.1548
.0601
.0097
.0989
-.1018
.8223
1.0000
X24
.5779
-.2022
-.0214
-.0871
-.1490
-.1705
-.2307
-.0624
-.2086
-.0383
-.2082
.1213
.2872
.3975
.3251
.7140
XQ
9
-.2701
.5452
-.5317
.0602
.1402
.2599
.5803
X17
.4995
-.1256
-.0795
.0639
-.0483
-.1151
-.1886
-.0174
-.1336
-.0919
-.2315
.1215
.0884
.7023
.6584
X25
.8138
-.3488
.0103
- . 1544
-.2855
-.3329
-.3370
-.1142
-.3587
-.0155
-.2805
.2462
.1603
.4623
.2641
o2915
106
-------
Table 16 (continued). CORRELATION MATRIX OF SURFACE SOIL VARIABLES
(coefficient of correlation - r)
X.
1
i
18
19
20
21
22
23
24
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
X18
1.0000
X26
.7750
-.6665
.4570
-.3766
-.5384
-.6121
-.6118
-.3208
-.6662
.4296
.1817
.1632
.3762
.3641
.0710
.4016
.3380
.3666
.5279
.6702
.6302
.0409
.3906
.3640
1.0000
X19
.4827
1.0000
^7
.6714
-.4717
.2768
-.2534
-.3800
-.4303
-.4514
-.2030
-.4761
.2582
.0566
.1915
.3130
.4350
.2804
.6505
.3333
.0920
.2727
.7022
.6899
-.0253
.6313
.2740
.7908
1.0000
X20
.4557
.8995
1.0000
X28
.7344
-.4418
.2029
-.2271
-.3583
-.4120
-o4254
-.1641
-.4507
.1857
-.0384
.0280
.2815
.3421
.0533
.4805
.2931
.2106
.2420
.8114
.6162
-.2035
.4207
.3505
.7262
.7303
1.0000
X21
.4686
.1664
.1428
1.0000
X29
.7782
-.5266
.2664
-.1413
-.3615
-.4646
-.5239
-.3322
-.5115
.2163
-.0187
.2740
.2606
.6492
.5085
.4686
.5395
.3066
.3916
.6916
.5431
.3790
.3237
.6972
.5841
.5285
.5025
1.0000
X22
.6553
.2488
.2738
.6783
1.0000
X30
.4822
-.4881
.3345
- . 2400
-.3743
-.4518
-.4409
-.2714
-.4811
.3058
.1372
-.0167
.1331
.0591
-.1621
-.1726
.2456
.7468
.6648
.1004
.0261
.2781
-.0755
.4384
.2768
-.1327
.0987
-.2950
1.0000
X23
.6244
.4142
.4808
-.0405
.2553
1.0000
X31
.5746
-.6326
.4664
-.2263
-.4642
-.5890
-.5909
-.3665
-.6207
.4293
.2297
.1256
.1633
.0731
-.0906
-.2014
.4846
.9986
.9216
.1653
.2545
.4269
.0583
.5539
.3903
.1151
.2167
.3203
.7449
1.0000
X24
.3043
.0494
.1179
.5725
,3988
-,0626
1.0000
X32
-.5021
.7889
-.7166
.2780
.7144
.9382
.6512
.1086
.8393
-.7662
-.6472
-.3032
-.4419
-.3385
-.2572
-.1878
-.4017
-.4731
-.5644
-.3101
-.2173
-.3976
-.1688
-.3259
-.4690
-.3672
-.3441
-.4686
-.3792
-.4894
1.0000
X25
.7113
.5586
.4703
.5755
.5131
.3530
.3937
X33
.8431
-.5678
.2608
-.2733
-.4281
-.5106
-.5345
-.3147
-.5598
.2142
-.0712
.1483
.1836
.2855
.0351
.1108
.6353
.8625
.7945
.4657
.5019
.3363
..3223
.8378
.5328
.3188
.4387
.5528
.6953
.8636
-.4113
1.0000
107
-------
Table 16 (continued). CORRELATION MATRIX OF SURFACE SOIL VARIABLES
(coefficient of correlation - r)
X.
i
2
3 -
4
5 -
6 -
7 -
8 -
9 -
10 -
11
12
13
14
15
16 -
17
18
19
20
21
22
23 -
24
25
26
27
28
29
30
31
32 -
33
34 1
35
36
37
38
39
Knomo
Kobs
X34
.7314
.5186
.2772
.2623
.4011
.4757
.4933
.2405
.5200
.2498
.0105
.0234
.1878
.1829
.1015
.1825
.3837
.6027
.5579
.6083
.5416
.0247
.2203
.4528
.6274
.5199
.8589
.4002
.4033
.6038
.3784
.7045
.0000
X35
.8966
-.5005
.1750
-.1907
-.3631
-.4291
-.5050
-.2840
-.4922
.1267
-.1544
.1651
.2042
.7705
.5793
.7103
.6491
.3262
.4150
.7969
.7175
.3414
.6575
.7685
.6650
.6800
.6216
.8293
.2449
o3406
-.4048
.6588
.5007
1.0000
X36
.7048
-.3747
.1372
-.1763
-.2906
-.3368
-.3824
-.1405
-.3820
.1186
-.0907
.0215
.2398
.5037
.2615
.6432
.3117
.0615
.1243
.8459
.6601
-.1479
.5316
.3334
.6878
.7643
.9591
.5471
-.0322
.0698
-.2973
.3410
.7278
.7068
1.0000
X37
.8648
-.5343
.2510
-.2741
-.4297
-.4934
-.5141
-.2142
-.5421
.2271
-.0388
.1333
.3382
.4609
.1907
.6130
.4290
.2837
.3641
.8606
.7202
-.0632
.6168
.4943
.8244
.8877
,9351
.6283
.1247
.2966
-.4231
.5554
.7899
.7891
.9237
1.0000
X38
-.5688
.9196
-.8497
.6158
.8922
.9847
.7197
.2435
.9565
-.8816
-.7239
-.2385
-.4634
-.2413
-.0787
-.0903
-.3795
-.5488
-.6734
-.3186
-.2761
-.3179
-.1695
-.3275
-.6203
-.4340
-.4124
-.4424
-.4503
-.5700
.8920
-.5094
-.4761
-.4205
-.3353
-.4943
1.0000
X39
.1371
-.8474
.9761
-.5044
-.6968
-.7876
-.7499
-.4167
-.8454
.9807
.9503
.0507
.3887
.0142
-.0067
-.1582
.1218
.3712
.5363
-.0353
-.0032
.2532
-.0876
-.1215
.3882
.2072
.1298
.1187
.2661
.3954
-.6467
.1469
.2012
.0324
.0583
.1554
-.8021
1.0000
Knomo
-.0215
-.6229
.7828
-.4414
-.6397
-.7103
-.4210
-.1878
-.6436
.8208
.8592
.3632
.5615
-.2149
-.2000
-.3094
-.0781
.2374
.3653
-.1839
-.1782
.0969
-.1674
-.1890
.2216
.0733
-.0018
-.0056
.1394
.2556
-.6540
.0002
.0595
-.1821
-.1023
.0035
-.7208
.7991
1.0000
Kobs
.0564
-.6801
.8179
-.4408
-.6393
-.7167
-.5220
-.2606
-.6923
.8425
.8577
.3324
.6379
-.0906
-.1074
-.1658
-.0357
.1930
.3516
-.0700
-.0980
.0916
-.0459
-.1839
.3150
,1790
.1199
.0570
.1112
.2145
-.6553
.0062
.1167
-.0671
.0548
.1226
-.7264
.8251
.9555
1.0000
108
-------
Table 17. CORRELATION MATRIX OF SUBSOIL VARIABLES
(coefficient of correlation - r)
X.
1
i
2
3
4
5
6
7
8
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
X2
1.0000
X10
-.8077
.9957
-.6552
.9373
.9949
.9918
.9991
.9320
1.0000
X18
.2299
-.2092
.0748
-.3739
-.1582
-.1468
-.2558
- . 1444
-.2274
.0175
-.1515
a
a
.1754
-.0689
.4510
X3
-.8208
1.0000
Xll
-.2879
-.3032
.9127
-.2321
-.3238
-.3316
-.3187
-.2282
-.3232
1.0000
X19
-.1142
-.0967
.4353
.1029
-.0826
-.0722
-.0408
-.2545
-.0421
.4021
.4028
a
a
-.7333
-.5639
-.5731
X4
.1080
-.6513
1.0000
X12
-.7977
.3257
.5030
.3759
.3014
.2915
.3165
.3530
.3109
.7971
1.0000
X20
-.1872
-.3392
.8099
-.1748
-.3306
-.3710
-.3527
-.3157
-.3412
.8291
.6107
a
a
-.3058
-.2223
.0708
X5
-.7942
.9098
-.5174
1.0000
X13
a
a
a
a
a
a
a
a
a
a
a
a
X21
.3952
-.1860
-.3006
-.2320
-.1473
-.1737
-.2071
-.1658
-.1895
-.4533
.6098
a
a
.6850
.6195
.4969
X6
-.8072
.9956
-.6619
.9073
1.0000
X14
a
a
a
a
a
a
a
a
a
a
a
a
a
x22
.4175
-.3784
.0581
-.4878
-.3050
-.3185
-.4233
-.3347
-.3866
-.1009
-.3901
a
a
.4344
.2496
.6598
X7
-.7957
.9950
-.6711
.8871
.9975
1.0000
X15
.6722
-.3275
-.3885
-.3959
-.3218
-.3311
-.3477
-.2660
-.3425
-.6123
-.8386
a
a
1.0000
X23
.5925
-.3463
-.1274
-.0863
-.3571
-.3598
-.2683
-.5156
-.2834
-.4209
-.5581
a
a
.3150
.5454
-.7284
X8
-.8087
.9955
-.6523
.9384
.9910
.9890
1.0000
X16
.8139
-.5760
-.1299
-.5405
-.5858
-.5954
-.5727
-.5404
-.5779
-.4368
-.7990
a
a
.9093
1.0000
X24
.7584
-.5708
-.0849
-.6532
-.5534
-.5552
-.5917
-.4975
-.5855
-.3596
-.7571
a
a
.7789
.8304
.4855
X9
-.8177
.9615
-.6062
.7812
.9466
.9534
.9338
X17
.0668
-.1109
.0225
-.4380
-.0852
-.0734
-.1970
.0952
-.1781
.0764
.0925
a
a
.2669
.0711
X25
.6367
-.6496
.3594
-.5250
-.6058
-.5997
-.6214
-.7441
-.6068
.0577
-.3171
a
a
.0838
.1679
-.1839
109
-------
Table 17 (continued). CORRELATION MATRIX OF SUBSOIL VARIABLES
(coefficient of correlation - r)
X.
1
i
18
19
20
21
22
23
24
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
X18
1.0000
X26
.8904
-.6985
.0231
-.8104
-.7168
-.6884
-.7126
-.5899
-.7229
-.2747
-.7329
a
a
.7177
.8225
.2362
.1540
-.3614
-.3198
.3300
.2462
.4052
.7588
.3349
1.0000
X19
.0692
1.0000
X27
.8554
-.7438
.1354
-.8803
-.7605
-.7349
-.7715
-.6013
-.7789
-.1431
-.6433
a
a
.6975
.7877
.3672
.2335
-.4003
-.1784
.3429
.3352
.2564
.7671
.3219
.9820
1.0000
X20
.2065
.2402
1.0000
X28
.7571
-.6262
.0310
-.6492
-.6457
-.6404
-.6287
-.5506
-.6407
-.2450
-.6561
a
a
.6853
.8729
.2226
-.2808
-.5192
-.2604
.6317
.1335
.4082
.9007
.0106
.8330
.8139
1.0000
X21
-.0588
-.7033
-.1739
1.0000
X29
.7660
-.5124
-.1001
-.5687
-.4610
-.4492
-.5191
-.5286
-.4981
-.3940
-.7210
a
a
.5706
.4717
.1974
.7394
-o0007
-.1563
.2589
.7209
.2968
.4639
.7890
.5911
.5765
.2449
1.0000
X22
.7571
-.2438
.2244
.5613
1.0000
X30
.9109
-.6491
-.0424
-.5055
-.6367
-.6320
-.6030
-.7451
-.6057
-.4348
-.7979
a
a
.5813
.7455
-.3181
.0725
.0785
-.2928
.2503
.1915
.8549
.5458
.6669
,7188
.6212
.5997
.7003
1.0000
X23
-o3335
.2427
-.3458
.0170
-.2811
1.0000
X31
-.1237
-.1158
.4776
.0906
-.1015
-.0936
-.0616
-.2696
-.0622
.4461
.4335
a
a
-.7399
-.5682
-.5594
.0807
.9982
.2973
-.7024
-.2261
.2176
-.6157
.5608
-.3750
-.4047
-.5266
-.0102
.0593
1.0000
X24
.0165
-.6116
-.2327
.8505
.5173
.2107
1.0000
X32
-.7768
.9674
-.6428
.9717
.9690
.9601
.9851
.8644
.9854
-.3428
.2866
a
a
-.3563
-.5465
-.2926
-.3265
.0189
-.3396
-.1604
-.4295
-.1624
-.5726
-.5758
-.7471
-.8208
-.6085
-.5186
-,5293
-.0022
1.0000
X25
.5931
,5581
.1940
.1584
.4589
.4213
.0973
X33
.5699
-.6883
.5353
-.5319
-.6689
-.6508
-.6436
-.7860
-.6437
.2514
-.1369
a
a
-.1870
.0537
-,3282
.2541
.7380
.1569
-.3239
.1222
.4967
.0198
.8836
.3158
.2852
.0907
.4974
.6148
.7355
-.5864
110
-------
Table 17 (continued).
CORRELATION MATRIX OF SUBSOIL VARIABLES
(coefficient of correlation - r)
X.
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Knomo
Kobs
X34
.6723
-.7675
.4886
-.6072
-.7846
-.7690
-.7202
-.8165
-.7378
.1785
-.2673
a
a
.0277
.4085
-.2943
-.3000
.3515
-.0205
-.0053
-.1511
.6385
.3828
.4653
.5581
.5088
.6118
.1785
.6696
.3445
-.6494
.7513
1.0000
X35
.8424
-.5173
-.2385
-.4252
-.5179
-.5257
-.4917
-.5543
-.4560
-.5768
-.8782
a
a
.8570
.9536
-.1435
-.0700
-.3772
-.3453
.5356
.1946
.7390
.7329
.3090
.7660
.6849
.7754
.5586
.8803
-.3921
-.4371
.1914
.4544
1.0000
X36
.8040
-.5136
-.2132
-.4345
-.5199
-.5309
-.4948
-.5254
-.5007
-.5310
-.8392
a
a
.8749
.9819
-.0655
-.1557
-.4863
-.3113
.6131
.1775
.6673
.7927
.1820
.7661
.6999
.8461
.4568
.8048
-.4973
-.4444
.0872
.4425
.9854
1.0000
X37
.8831
-.7386
.0758
-.7716
-.7503
-.7397
-.7425
-.6633
-.7511
-.2481
-.7283
a
a
.7270
.8976
.2299
-.0768
-.4313
-.2384
.5643
.2541
.4475
.9010
.2233
.9262
.9100
.9703
.4497
.7163
-.4387
-.7307
.2505
.6469
.8226
.8613
1.0000
Knomo
-.7682
.3271
.4850
.4431
.3058
.2966
.3397
.2951
.3322
.7453
.9703
a
a
-.9177
-.8277
-.2867
-.2401
.5728
.5413
-.6740
-.4856
-.3796
-.8160
-.2266
-.7719
-.7185
-.6825
-.7159
-.6874
.5965
.3407
-.0001
-.1473
-.8419
-.8261
-.7523
1.0000
Kobs
-.7842
.4579
.2054
.3101
.4238
.4216
.4026
.6108
.3986
.5653
.8071
a
a
-.4557
-.5917
.3037
.0253
-.1317
.4297
-.3824
-.2204
-.8177
-.5390
-.5823
-.5183
-.3969
-.5148
-.6279
-.9161
-.1033
.2954
-.5543
-.6065
-.7743
-.6991
-.5908
.6556
1.0000
Coefficient of correlation could not be computed because of missing
data.
Ill
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-660/2-7lt-Ql+3
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Prediction of subsoil erodibility using chemical,
mineralogical and physical parameters
5. REPORT DATE
June
6. PERFORMING ORGANIZATION CODE
Purdue Research Foundation
7. AUTHOR(S)
Charles B. Roth, Darrell W. Nelson,
Mathias J. M. RSmkens
8. PERFORMING ORGANIZATION REPORT NO
5460
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Purdue Research Foundation
Executive Building
West Lafayette, Indiana 47907
10. PROGRAM ELEMENT NO.
1BB042
11. CONTRACT/GRANT NO.
15030 HIX
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report presents evidence that the surface soil erodibility prediction nomograp
(Wischmeier et al., 1971) which uses terms involving soil particle size, organic matte
structure and permeability, could not be improved upon by consideration of other min-
eralogical and chemical parameters. However, the surface soil erodibility nomograph
did not adequately predict the soil erodibility factor, K, of high clay subsoils
studied in the field under simulated rainfall conditions as a part of this project. A
multiple linear regression equation and nomograph were developed which can be used to
estimate the erodibility factor, K, of many high clay subsoils. The subsoil erodibil-
ity nomograph uses terms involving soil particle size distribution and the amount of
amorphous hydrous oxides or iron, aluminum, and silicon in the soil. Multiple regres-
sion analysis revealed that amorphous iron, aluminum and silicon hydrous oxides serve
as soil stabilizers in subsoils, whereas, organic matter is the major stabilizer in
surface soils.
Evidence is presented to show that soil erodibility from semi-compacted fill and
scalped subsoil surface conditions were essentially identical. It is reported that
the scalped condition is the best standard soil surface to base the calculation of the
erodibility factor for subsoils.
It is suggested that a soil-management factor should replace the cropping-managemen
fart-nr In the Unii
iaf--irm
iBrosion.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Agronomy, Agricultural Engineering, Alumi-
num inorganic compounds, clay minerals,
erosion, erosion control, iron inorganic
compounds, regression analysis, silicate
minerals, size separation, soil erosion,
soil aggregates, soil analysis, soil
structure
13. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
21. NO. OF PAGES
123
20. SECURITY CLASS (This page)
22. PRICE
EPA Form 2220-1 (9-73)
------- |