THE USE OF REDOX POTENTIALS
IN STUDIES OF SOIL GENESIS

Thai: for Ike Degree OF M. S.
MICHIGAN STATE COLLEGE
Lloyd J. McKenzie
I954

THESIS

This is to certify that the

thesis entitled

The Use of Redo: Potentials in Studies of Soil Genesis

presented by

Lloyd J. McKenzie

has been accepted towards fulfillment
of the requirements for

Master om degree in Meme

URLEKE

Major professor

 

Date M81 203 19514

0-169

THE USE OF REDOX POTENTIALS
IN STUDIES OF SOIL GENESIS

By

Lloyd J. ggXenzie

A THESIS

Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Soil Science
Year 1954

,/
, J.<O
4

x; )

ACKNOWLEDGEMENT

The author is grateful to Dr. A. E. Erickson and
Dr. E. P. Whiteside for their assistance in the selection
of the sites, setting out the electrodes, and for their
continued interest and willingness to discuss the work as
it was carried out. Special thanks is also due to Dr. Kirk
Lawton for his interest and assistance in the chemical
problems that arose. For the Opportunity to do the graduate
work, the author is grateful to Dr. L. M. Turk. To Dr. Ba-
ten, Experiment Station Mathematician, thanks are also due
for invaluable assistance with the statistical analysis.
Dr. Mb; Mortland and Dr. R. L. Cook made many helpful sug-

gestions and comments on the manuscript.

11

TABLE OF CONTENTS

INDEX OF FIGURES AND TABLES . . . . . .
INTRODUCTION . . . . . . . . . . . . . .
REVIEW OF LITERATURE . . . . . . . . . .
THEORETICAL CONSIDERATIONS . . . . . . .
Redox Potentials . . . . . . . . .
Factors Affecting Redox Potentials
Significance of Eh . . . . . . . .
The Podzolization Process . . . . .
THE FIELDZEXPERIEENT . . . . . . . . . .
McBride fine sandy loam . . . . . .
Mariette loam . . . . . . . . . . .
Parkhill loam . . . . . . . . . . .
Installation of Electrodes . . . .
Measurement of Redox Potentials . .
LABORATORX'EXPEHIMENTS . . . . . . . . .
The Eh-pH Relationship . . . . . .
The Laboratory Experiment . . . . .
Wears Sand . . . . . . . . . . . .
RESULTS AND DISCUSSION . . . . . . . . .
SUMMARY . . . . . . . . . . . . . . . .
BIBLIOGRAPHY . . . . . . . . . . . . . .
APPENDIX . . . . . . . . . . . . . . . .

iv

E.)

16
16
19
20
21
24
24
26
27
28
29
56

58
59
45
52

59

Figure 1.

Figure 2.
Figure 5.

Figure 4.

Figure 5.

Figure 6.

Figure 7.
Figure 8.
Figure 9.
Figure 10.

Figure 11.

Table I.

INDEX OF FIGURES AND TABLES

Text

The Relationship between Eh and the relative

amount of oxidation or reduction for a system.

Layout of electrodes ig_situ. . . . . . . .
Method of making redox measurements in situ.

The effect of season and horizon on the redox
potentials in McBride fine sandy loam. . . .

The effect of season and horizon on the redox
potentials in Marlette loam. . . . . . . . .

Effect of season and horizon on the redox
potentials for Parkhill loam.. . . . . . . .

Nitrogen cell for measuring Eh-pH relationships.

A replicate of the sand column experiment. .
The sand column experiment. . . . . . . . ,

Effect of depth on the redox potentials for
treatment A of the sand column experiment.xt

Effect of depth on the redox potential for
treatment B of the sand column experiment. .

Ferrous iron content of the extractable iron
for treatments A and B. . . . . . . . . . .

iv

.18
.28
.51

.55

.54

.55
.56
.41
.42

.47

.50

Appendix

Figure l. Eh - pH relationship for McBride
fine sandy loam . . . . . . . . . .

Table I. Average values or five replicates
of treatment A by depth and dates .

Table II. Table of means for treatment B of
the sand column experiment . . . . .

Table III. Analysis of variance for days,
depths and replicates for treat-
ment B of the sand column experiment

Table IV. Instantaneous and diffusion readings
for oxygen compared with Eh readings
for depth in treatments A and B. . .

Table V. Moisture variation in the sand col-
umns with depth for treatments
A and B O O O O O O O O O O O O O O

INTRODUCTION

oxidation and solution are two chemical processes that
are involved in the deveIOpment of the soil body. There is
a differential solubility in the oxidized and reduced state
of such metallic ions as iron and manganese. For example,
ferrous hydroxide is a million times more soluble than fer-
ric hydroxide. This variation in the state of oxidation
within the soil body, may be a means whereby such elements
are brought into solution, translocated, and precipitated I
in horizons of different oxidation-reduction status. There-
fore a knowledge of the oxidation-reduction potentials in
soils may give information usable in explaining soil forma-
tion phenomena.

Recently Quispel (29) has described a method whereby
the oxidation state of the soil body can be measured in
pgitg. An adaptation of his method has been used in this
study to measure the oxidation state of the soil body at
depths corresponding to key soil horizons and to relate
these findings to soil genesis in the podzol region. A lab-
oratory eXperiment was also performed to see if a redox
profile would form in a sand column and what effect this

might have on the status of the iron present.

REVIEW OF LITERATURE

Gillespie (10) was one of the first investigators to
work with oxidation-reduction potentials in soils. In lab-
oratory studies he found that soils became highly reduced
when subjected to waterlogging. This reducing condition
was emphasized by the addition of dextrose. The increase
in the intensity of the reducing conditions was accompanied
by a foul odor. He discussed the significance of reducing
conditions and of reduction potentials in soil study. The‘
toxic effect of freshly incorporated green manure may possi-
bly be the result of extreme reduction.

Neller (22), in 1982, found that growing green plants
accelerated the oxidation processes in the soil. Using car-
bon dioxide as the index of the effect of the plants, he
found that the average total carbon dioxide recovery from
planted cultures was 12.1 percent higher than that from un-
planted cultures. Buckwheat, field beans, and soybeans gave
respectively 116.5 percent, 70.8 percent, 60.0 percent, in-
crease in carbon dioxide production. He also found that the
second crap gave greater increases than the first. All evi-
dence obtained showed that growing plants of buckwheat, field

beans, and soybeans had a beneficial effect on the oxidative

5.

processes in the substrata in which they were grown and
suggested a symbiotic relationship between the soil oxie
dizing organisms and the growing green plants.

Bouyoucos (2), in 1925, found that different chemical
agents have a decidedly different effect on the oxidation
of iron both as to rate and extent. The non-oxidizing
effect of some reagents may dominate the oxidizing effect
of others; and the oxidation can take place even in the
absence of oxygen accessible from the air.

In 1952, Willis (59), using Dunbar fine sandy loam,
determined the Eh—pH relationship in the soil. He used a
special apparatus to provide a nitrogen atmosphere. The
soil was treated with lime at rates of from 1/2 ton to four
tons per acre. Eh readings and pH measurements were taken
and curves prepared. There was an inverse relationship be-
tween Eh and pH. Extracts of the limed soils were made and
titrated with hydrochloric acid. The same Eh-pH relation-
ship was observed in the extracts, i.e. a decrease in Eh of
about sixty millivolts for each unit increase in pH.

In 1954, Bradfield, Batjer, and Oskamp (5) used redox
potentials to evaluate soils for orchard purposes. They
found that maximum differences occurred in soils in April
and May, and that they could eliminate 80 to 90 percent of

the poor orchard sites on the basis of redox potentials.

At the same time Brown (4) reviewed the principles
involved in the study of soils by means of oxidation-reduc-
tion potentials. He stated that by means of data obtained
in this way, it may be possible to determine the more exact
nature of certain soil processes which are important in
plant growth. Brown used a rapid method of measuring Eh in
the laboratory, and develOped the method of preparing elec-
trodes which were later used by Quispel (29) and Lemon (18).

In the same year Heihtze (11) used the glass instead
of the calomel electrode as the reference electrode in order
to minimize polarization and to facilitate instantaneous I
Eh-pH measurements. She found that the ratio of soil to
water was of little importance in a suspension; and that the
presence of air or nitrogen did not affect the potential.

She also found that soils, known to contain readily decom-
posible organic material, on being waterlogged, showed a
marked drOp in potential as soon as conditions of tempera-
ture and pH are favorable; while soils that contained little
organic matter tended to resist any drOp in Eh. She ob-
served that oxidation-reduction potentials of soils depended
so largely on pH that the values should not be considered
separately. Highly contrasting soil types with equal pH val-
ues gave similar oxidatiOn-reduction potentials. She found
that interlogging brought about a greater change in potential

in the Chernozem soils of Russia than in any other soil studied.

In 1955, Kohnke (15) found that the oxygen supply of
a soil determines the oxidation-reduction potential. A
high potential is indicative of a.well drained and oxidized
soil and a low potential is indicative of a poorly drained
de-aerated soil. The principle substances entering into
the oxidation-reduction reactions of soils are colloidal
clay, organic substances and humus, and compounds of iron
and manganese. He found the Eh change per unit change in.)
pH to be fifty-nine millivolts in the pH range of two to
seven. Colloidal clays showed Eh-pH relationships compara- ‘
ble to those of soils. Aeration increased the oxidation-
reduction potential of a soil and.waterlogging lowered the
potential. The changes took place within two to three days.
Seasonal variations in the years 1955-54 were small due to
unusually dry weather. Kohnke also found that the oxidation-
reduction potentials of surface soils are lower than those
of the deeper horizons during the Spring. Later on those
differences tended to disappear. Quantitative determinations
of oxidized and.reduced substances in the soil gave large
significant differences between well and poorly aerated soils.
The experimental error was rather large.

In the same year, Peech and Batjer (24) used 0.1 N sul-
phuric acid as the suspension medium. They found that the
previous use Of the electrodes had little effect on the read-

ings and that the suspension was better poised. Some soils

appeared to be less susceptible to oxidation in H2804.
They found that no appreciable reduction took place in
soils below temperature of 55°F.

In 1956, Sturgis (54) found that low potentials in
waterlogged soils were largely caused by the decomposition
of fresh organic matter. The solubility of phosphorus
was reduced under low potential conditions. In the absence
of actively decomposing organic matter, large amounts of
iron compounds precipitated on and around the roots of rice
plants. Adding gypsum to a waterlogged soil caused the prof
duction of sulphides wnich reduced the yield of rice. The
application of leguminous organic matter increased the yield.
Sturgis also found that the addition of fresh organic matter
to waterlogged soils resulted in potentials as low as eighty
millivolts as compared with two hundred millivolts in a wa-
terlogged soil low in organic matter.

At the same time, Darnell and Eisenmenger (8) found that
there was little or no change in potential with the addition
of fertilizers carrying nitrogen. The changes correlated
better with the change in pH. The rapid decomposition of
fresh organic matter brought about a marked fall in potential.
They eXplained this fall as being due to oxygen depletion.

Burrows and Gordon (S), in 1956, found that the type of
decomposable organic matter in a soil was an important fac-

tor in the determination of the reducing intensity that

prevailed. They found that the decomposition of casein re-
sulted in highly positive potential levels, while the de-
composition of carbohydrate resulted in negative potentials
which did not appear to differ greatly from those produced
in soil to which organic matter had not been added. They
found that moisture did not appreciably affect potentials.

Willis (40) showed that manganese deficiency caused by
liming could be explained by consideration of Eh as well as
pH. He also stated that the two soil components, organic
matter, and oxygen, appeared to govern the oxidation-reduction
equilibrium in the soil. Both required activation, the former
by micro-organisms, and the latter by catalysis. He further
discussed the role of oxidation and reduction in soil ferti-
lity, and suggested evidence that phosphorus, potassium, cop-
per, manganese, boron, silicic acid, organic matter, pH,
aeration, and temperature are to_some degree interdependent
variables.

In 1959, Buehrer et a1 (5) found that the Eh-pH relation-
ship could be determined by simply diluting a soil suspension
with water. They showed that the Eh change amounted to about
sixty-eight to seventy millivolts per unit change in pH; and
they showed that bubbling nitrogen through the soil tended to
lower the Eh value. They found that puddling caused a drop
in soil Eh which they believed was due to factors other than

oxygen depletion, that the presence of abundant oxygen Caused

an increase in Eh. The addition of alfalfa to a soil sus-
pension caused a marked decrease in Eh, which was taken to
be indicative of the nature of the reduced compounds formed
during its decomposition.

Stevenson et a1 (55) working with Oregon soils found
that different soil types under field conditions showed little
variation in oxidation-reduction potential with the methods
used. There was little variation in potential between hori-
zons even when there was a tight subsoil. They also found
Ithat fresh organic matter alone in a moist soil did not cause
a fall in potential, but waterlogging a wet soil caused a ra-
pid fall in potential. Anaerobic conditions in the soil were
not indicated by the oxidation reduction potential in a relies
ble manner.

Volk (57), working with Alabama soils, did more than any-
one to establish a standard technique for the measurement of
oxidation-reduction potentials in the laboratory. Batjer,
Bradfield, Kohnke, Peech, and Stevenson had all used 0.1 N
H3804 as the suspension medium to inhibit bacterial activity
and thus prevent the development of reducing conditions in
the soil, and to add poise to the system. Volk claimed that
results obtained in this way were not comparable to water
suspensions since the sulphuric aCid dissolved materials that
were not usually active in the soil. Only Kohnke and Willis

censidered the eXpulsion of air with nitrogen gas; while Brown,

Darnell, Heintze, Sturgis and others disregarded the possi-
bility of a change in potential during the analysis. Volk
found that arable soils in a high state of oxidation changed
little, but those that were in a reduced state became rapid-
ly oxidized unless air was eXpelled with nitrogen. He also
found that all preservatives used to inhibit bacterial action
changed the Eh of the soil. As a substitute for a preserva-
tive, Volk cooled all his samples to a point Just above freez-
ing. He found this procedure to be effective. Oxidation

was prevented by use of water from which the free oxygen was
removed 1; by boiling and.replaced with nitrogen. All analy-
ses were performed in a nitrogen atmosphere. Volk also

found that the Eh—pH relationship was variable between 59 and
101 millivolts for each pH unit. Potential drift was found
to be due to the past history of the electrode. Wire elec-
trodes were found to be superior to foil electrodes. He used
a battery of thirty glass electrodes and 120 blank electrodes
and was able to run fifteen to twenty samples per hour.

Also, about the same time, Volk (55) made extensive stu-
dies of Alabama soils. He found that cultivated soils, in
general, had a higher Eh value than did virgin soils in the
zero to eight inch depth; but there was little difference in
the subsoils. He found that differences resulting from soil
type variations rarely exceeded fifty millivolts. These he

did not think important since they could have been due to

lO

differences in soil material rather than to state of oxida-
tion. Swampy soils, for example, frequently had higher Eh
values than did.well drained upland soils. After a rain the
Eh rose for a time which Volk eXplained by suggesting that
the rainwater carried oxygen into the soil. When the excess
oxygen was removed and conditions were once again stable,
the Eh of the soil returned to normal. Seasonal variations
in Alabama soils did not exceed sixty millivolts. Volk does
not think that Eh is an indicater of the oxidation state of
the soil since it is dependent, not only on the relative
amounts and kinds of the ions present but also on the ratio
of oxidized to reduced substances present. In another paper
in 1939 (56), Volk made a study of the effect of Eh on the
growth of plants. He controlled conditions so that no other
factor was limiting, and was able to adjust the Eh of the nu-
trient solution from 585 to 525 millivolts by use of hydro-
quinone. The hydroquinone had no toxic effect on young seed-
lings in concentrations up to fifteen ppm. and in later sta-
ges of growth, he could use concentrations up to sixty to
seventy ppm. There was no limiting effect of Eh on plants in
the range of 525 to 525 millivolts. He concluded that Eh was
not limiting to plant growth on Alabama soils.

Keaton and Kardos (14), in 1940, concluded that the
oxidative character of the soil was reflected in the redox

potential. Because of the complexity of the soil medium,

11

however, they did not consider potential to be that of an
individual system; but regarded it as the function of many
complex interlocking systems. With this in mind, they sug-
gested that redox potential may be used in the study and in-
terpretation of the general chemical processes taking place
in the soil. They studied the effect of various materials

on the arsenite-arsenate system. They were able to show

that ferric oxide raised the potential and caused arsenite

to be oxidized to arsenate. Alumina produced no effect on
the system. Clay colloids tended to shift the equilibrium a
in the opposite direction, i.e., arsenate was reduced to ark
senite; and at the same time the potential was lowered.

Their interpretations were based on calculations using the
Nernst equation. They further 00ncluded that the addition

of iron in the oxidized form caused increases in plant growth
in soils poisoned with arsenical spray residues; and that the
beneficial effect was primarily due to increased fixation of
arsenate as it was oxidized from arsenite to arsenate.

The next significant advance occured in 1947 when Quie-
pel (29) measured redox.potentials in gitg. He studied the
soils in Holland that were inundated during World War II,
and found potentials as low as -250 millivolts. He used
platinized, platinum electrodes, and found the results more
reproducible. Due to the great variance between measure-

ments ;Q.situ, Quispel recommended that a great many readings

12

be taken. He found that oxygen also affected the pctential
greatly; thus the Eh of the soil was determined in part by
the state of aeration of the soil. When the potential of a
soil was determined in the field, it was possible to define
the aerobic or anaerobic character of the soil in terms of Eh.

In the same year Roberts (50) working with micro-organisms
found that B. polymyxa, in the presence of glucose, reduced
ferric oxides or ferric hydroxide. The fermentation of glu-
cose was more rapid in the presence or ferric hydroxide and
the production of hydrogen was reduced. Roberts tested a
great number of cultures isolated from the soil and he found
that only B. polymyxa and One culture of Clostridium sp. re-
duced iron in appreciable amounts. These were facultative aero-
bic bacilli.

Bertramson and White (1) give an excellent review of they
literature dealing with oxidation-reduction potentials in re-
cent years. They concluded that the Eh of the soil represen-
ted the sum total or the oxidizing and reducing tendencies
in the soil and depended on the nature of the system. The
redox system varied with the nature of the soil, and there-
fore varied from soil to soil. They were critical of present
laboratory techniques since there are no suitable methods for
measuring Eh in the soil in a manner similar to the measure-
ment of pH. According to them the present field quick test
_for ferrous iron may be more suitable for the measurement of

reducing conditions in the soil.

15

In 1949, LaFonde (16) investigated the redox character-
istics of several types of forest humus from Wisconsin and
Quebec. He measured pH, redox potential, and the total quan-
tity of reduced substances by titration with permanganate
solution. His results showed that mull humus was character-
ized by a positive redox potential and strong acidity. La
Fonde concluded that redox potentials could be used to charac-
terize forest humus types.

In the same year, Wilde and others (58) studied the elec-
tro—chemical properties of the ground water in the four major
types of organic soils in Wisconsin. They measured pH, spe-
cific conductivity and redox potential in moss peat, wood
peat, sedge pest, and muck. The ground water underlying the
four soils possessed specific electro-chemical prOperties.
All results were highly significant. They concluded that
electro-chemical analysis of ground water was important in
dealing with problems in drainage of muck or peat soils and
regulation of the water table.

In 1950, Pierce (26) studied the prairie—like mull of
Downs silt loam, one of the most conspicuous forms of mull in
the prairie forest region. Measurement of pH, Specific con-
ductance and redox potential indicated an intermediate posi-
tion between Carrington silt loam and Miami silt loam which
was taken to indicate that the soil was originally a prairie
soil which had undergone a slight modification as a result of

becoming forested.

14

In 1955, Pierce (25) measured pH, Specific conductance
and redox potentials ip_§itg in swamps supporting different
types of forest vegetation. There was a close correlation
between redox potential and the oxygen content of the ground
water and with the rate of tree growth. The Specific con-
ductance, a measure of the electrolyte content, was found to
influence the forest type rather than the rate of growth.
These studies were carried out in Wisconsin. Further studies
in the Hearst region of Ontario gave similar results.

From the literature it is apparent that the measurement
of the redox potential of the soil has been hampered in the .
past by lack of consideration of all the factors involved.
From the scattered studies of the nature of these factors it
is suggested that their normal effect on the soil can best
be studied in the field, or if laboratory studies are carried
out, field conditions should be similated as closely as possi-
ble. The redox potential of the soil appears to be tied up
very closely with the oxygen content; and also with some of
the materials within the soil such as iron, organic matter‘and
manganese. There may also be other materials involved. Plants
seem to indicate that a deficiency of oxygen exists in the soil
before the deficiency can be observed through change in the
redox potential. This is true because the redox system in the

soil is more or less well poised or buffered. Nevertheless,

15

the study of the redox system in the soil may be useful in
the study of soil genesis since those factors that determine

the redox state are also important in soil genesis.

THEORETICAL CONSIDERATIONS
Redox Potentials

When a metal is placed in contact with a solution of
its ions, it acquires a charge; and if two such electrodes
are connected to form a cell, a current will flow which is
an expression of the difference in potential between the two
electrodes. The positive cell becomes reduced, and the nega-
tive one becomes oxidized. Thus, where ever oxidation is
taking place, reduction is also taking place. The system is
spoken of as an oxidation-reduction system. In this paper
the contraction, redox, is therefore used to replace the
longer term, oxidation-reduction.

When a cell is reversible, it is possible to relate the
electrical phenomena observed to the chemical activity taking
place by means of the Nernst equation.

Eh - E0 -/- BILog 29$)
nF Red)

To date, no one has been able to establish the actual value
of a half cell in volts; but two half cells together have a
potential which can be measured. When one of the half cells
such as the hydrogen electrode or the calomel electrode is

selected as a standard all cells can be given definite rela-

tive values. The Eh value is taken as the potential difference

17

between two half cells when they have been connected. From
the Nernst equation, it is evident that when (95).: 1, then
Eh - Eo. This makes it possible to express ré§:gzve oxida-
tion-reduction intensities in terms of electrode potentials.

It should be made clear that redox potentials are not a
measure of the amount of oxidation or reduction that takes
place. They are a measure of the oxidizing or reducing in-
tensity or tendency. This is quite a different matter from
the capacity of a system to oxidize or reduce. Some idea of
the capacity of a system to oxidize or reduce can be obtained
by titration with a strong oxidizing or reducing agent.

Eo can be determined for a system. It represents the
potential at fifty percent oxidation and fifty percent reduc-
tion of the system; and depending on the flatness of the Eh
curve as the system becomes relatively more oxidized or re-
duced the system will be well or poorly poised. If a system
is well poised at a certain Eo value, it will tend to be
oxidized by any other such system that has a higher Eo value;
and it will tend to be reduced by any other such system that
has a lower Eo value. Therefore redox potentials can be
plotted on a scale and compared in their ability to oxidize
or reduce other systems.

Another important consideration is that any particular
system can be measured if it is reversible and the Nernst

equation is applicable to that system.

H
U:

”r: ‘v‘ V!" ‘nn ‘fifl Y‘ ‘L V! ~r1‘9 :3...‘ Q‘Rq‘ ‘“ «\‘1-‘fil

4....-8 U... 0-: “:5 ”:90; 55-0‘ OnK-‘fi‘ \s-b&‘:::s- -0. 5- \s—.:\
n 0 A“”‘”‘ “1‘ AF“; ‘0‘" +r~ --V\"f A: :‘fi-;s c ‘8‘ ‘ o“
tic-c O- PC~$ -.EUS Uu~ ‘C--C~. LJ L.--9 v.¢-\.:u hr ..".\..t . ..---\.

'. ‘. - o— a. n ----- o‘ ‘ \ V‘ ‘- ’ 3

health Service (7). Figire 1. She‘s a .i.cer e- t”::-:.-:ll
O O “‘n -fl‘ _ ‘.. .fi . p: l‘ - a. 0" ‘ 3 ‘ .‘ r 0

curves oi a“ values which indicate the xezistiou o- -h $1.“

the.relative percentage of oxidation and red:
the different svstens can be cozparei.
curve is the most poorly poised.

Percent Oxidation

loo 1: 50 15 o
035' v T I

- volts
“\>

Eh

 

 

 

00° 1 l A

 

o 25 50 75' no
Percent Feéuction

Figure 1. The re ationship between Eh
and the relative amount of oxidation
or reduction for several'theoretical systems.

19

as a natural dynamic system, it is inferred that some of
the processes involved in soil develOpment may be redox
precesses. If such is the case, then the Eh of the soil
will be indicative of the overall state of oxidation of
the soil.
Factors Affecting Redox
Potentials of Soils

Since Gillespie's time much work has been done; and
apart from the contribution made in the field of oxidation
and reduction in biology, a number of factors which affect
redox potentials in the soil have/been observed. These may
be considered to be of three main types. First, there are
the environmental factors, climate, aeration, and soil dis-
turbances such as cultivation and compaction. Second, there
are factors active in the soil medium such as mineral ele-
ments, clay colloids, organic matter, micro-organisms and
growing plants. Third, there are variations in the redox
potential which are caused by the treatment of the soil prior
to the measurement of the potential and by the method used
in obtaining the measurement.

Volk found that reduced soils quickly become oxidized
when they are eXposed to the air. For this reason he placed
his samples in nitrOgen saturated water and froze them before
storage for analysis. Other investigators placed the samples

in water and allowed them to stand for a number of hours be-

fore measuring the redox potential. Some have used sulphuric

acid instead of water as the medium for their measurements.
In all cases the relationship between the soil sample and its
environment is destroyed and the results obtained are not
representative of the true oxidative state of the soil. It
is maintained that a mucn more representative picture of the
oxidation state or the soil can be obtained by measuring

the redox pctentials of the soil ig'§333;as some investiga-
tors have done (16, 25, 29, 58). All investigators except
QuiSpel and those who followed him have neglected to consider

this point.
Significance of Eh

From the discussion above, it can be seen that the fac-
tors affecting the redox potential or Eh of the soil are
numerous and varied. Because of the complexity in the soil
medium, the oxidation-reduction system should be regarded as
a number of individual systems Operating simultaneously.

The redox potential of the soil is a measure of the oxidation-
reduction status of all the component systems. Variations

in any one of such factors as degree of aeration, moisture
content, temperature, microbial population, or stage of de-
composition of organic matter will cause changes in the
oxidation state of the soil and will be reflected in the re-

dox potential. Such factors are also considered among the

21

factors of soil formation (12). On this basis, the redox
system in the soil may be considered as related to soil
genesis. The measurement of redox potentials may therefore

serve as a useful tool in the study of soil forming processes.
The Podzolization Process

In general, Podzolization is the group of soil formation
processes that result in develOpment of Podzol soils. This
involves the leaching of bases and translocation of sesqui-
oxides and organic matter from the A to the B horizon. In
the B horizon the sesqui-oxides and organic matter are deposi-
ted, often in layers with the organic matter above, but often
in one layer with the sesqui-oxides and organic matter inti-
mately mixed.

Morphologically speaking, the profile produced is very
striking. The gray, ashy A2 horizon, sometimes called blei-
cherde, contrasts greatly with the reddish brown to dark
brown ortstein or orterde B horizon. The overlying A1 hori-
zon may be lacking, in which case, the A0 and A00 horizons
make up the surface layers. There is then a very sharp boun-
dary between the surface horizons and the underlying A3 hori-
zon. This may indicate a very complete conversion of organic
matter into materials which become mobile in the A0 and A1
horizons and are transported through the A2 horizon into the

B horizon where precipitation takes place. Not all materials

82

are precipitated in the B horizon however (15, 51). Puust—
Jarvi et al (28), in Sweden, noted the accumulation of rust
precipitates in drainage lines and ditches. This trouble
occured in recently glaciated areas. Joffe (15) noted the
aluminum in the drainage waters of New Jersey streams.

A most commonly accepted theory of podzolization is
that the sesqui-oxides and organic matter are translocated
in the colloidal state, and precipitated in the B horizon
due to the presence of electrolyte. In Europe, Dr. Albert
found a relationship between the depth of penetration of sums
mer rains and depth to the B horizon or ortstein. He also
investigated the possibility of the oxides being precipitated
by oxygen in the B horizon, and found neither a deficiency of
oxygen in the A horizon nor any excess in the B horizon. He
concluded that oxygen played little part in the deve10pment
of the profile. In Sweden, Aarnio conducted kboratory experi-
ments in connection with the precipitation of oxides and or-
ganic matter in the B horizon. He was able to show that orb
ganic matter and iron in colloidal solution could be preci-
pitated due to mutual effects at certain concentrations. In
addition to the iron oxides and organic matter, manganese and
phOSphorus are deposited in the B horizon.

An excellent discussion of these works and their impor-

tance is given in lectures by Marbut (79).

25

Most theories on podzolization are connected with or-
ganic matter in some manner. This serves to indicate the
importance of organic matter in the process. Podzols occur
in a cool humid climate, and are commonly found in Michigan
under hardwood or coniferous types of vegetation, although
they may occur under a grass type as well. Podzols also
develop on a variety of parent materials ranging from alka-
line to acid in reaction and from sands to loams or finer
textured materials.

There seems to be no doubt that organic matter and its.
decomposition products play a very important part in the
podzolization process as do iron, aluminum, and manganese.
Organic matter and its decomposition products have been shown
to be effective in lowering the redox potential of soils (29).
Iron and manganese also play an important part in the oxida-
tion-reduction processes that Operate in the soil medium, and
there are also other mineral elements that could play a part
as well. Therefore it is logical to expect that oxidation
and reduction may also play an important part in the deve10p-
ment of podzols. Since podzols have been shown to form in a
relatively short time, they offer possibilities for redox

studies in relation to soil deve10pment.

THE FIELD E XPERIMENT

For the field experiment, two sites were selected in
Sanilac County, Michigan. They showed deve10pment of the
podzol type profile commonly found in the transition zone
between the Podzol region in northern Michigan and the Gray
Brown Podzolic region in southern Michigan. A discussion
of these soils may be found in a publication by Gardner and
Whiteside (9). The soils at both sites develOped on glacial
drift of Cary Age. McBride fine sandy loam is a well drained
soil developed on sandy loam parent material. Marlette loam
is another well drained soil deve10ped on loam parent material.
A third soil, Parkhill loam, is a poorly drained soil, whidh
occurred in association with Marlette loam. It also is devel-
Oped on loam parent material. The selection of these soils
permitted the comparison.of soils that differed in both tex~
ture and drainage. The following profile descriptions are

of these soils.

McBride Fine Sandy Loam

Location: T 12 N - R 15 E Section 27, SE&, ofithe NEé,
Sanilac County, Michigan.
Vegetative The present cover consists of a mixture
cover: of White Birch (Betula papyrifera), and

Trembling Aspen, (POpulus tremuloides).
The original cover was White Pine (Pinus
strobus), and Hemlock (Tsuga canadensis).

Physiography:

SlOpe:
Profile:
Horizon

A0, A00

A1

Ages

Ba GB

 

25

Rolling till plain of Cary Age at the
western edge of the Whittlesy lake bed
in Sanilac County, Michigan.

Six percent.

 

Depth Description
l-O' Leaf litter and partially decomposed

organic matter black in color.

O-l" Dark grayish yellowish brown, (10 YR
2/2)“, fine sandy loam with a crumb
structure, high in organic matter,
friable when moist, pH 5.5.

1-5" Light yellowish brown (10 YR 7/5),
loamy sand, weak crumb structure, '
pH 5.50

5-9" Strong yellowish brown (7.5 YR 5/6),

loam, crumb structure, friable when
moist, pH 4.5.

9-15" Light yellowish brown (10 YR 6/5),
sandy loam, weak platy structure,
friable, pH 5.2.“

15-48" Moderate yellowish brown (10 YR.5/4),
silt loam, weak blocky structure,
friable when moist, pH 5.1.

48-72“ Light yellowish brown (10 YR 6/4),
silt, massive, friable when moist,

pH 5.0.

72” Light yellowish brown (10 YR 6/4),
silt, stone free glacial drift,
pH 6.5.

* Munsell color notation. and ISCC é NBS color names.

Location:

Vegetation:

Physiography:

Slape:
P rofile:
Horizon

A1

Base

a)
O)

Harlette Loam

T'll N - R 12 E Section 50 NWi of NEi, Sanilac
County, Michigan.

The present cover is American Elm (Ulmus
americana), and White Birch. The original
cover was sugar maple, elm, hemlock, and birch.

The site is located at the western extremity
of the Yale moraine, of Cary age.

Six percent.

Depth
0-5'

5-6'

6-11'

11-15"

15-50"

50 “

Description

 

Brownish gray (10 YR 4/1), loam,
fine granular structure, friable
when moist, pH 6.0.

Light grayish yellowish brown (10
YR 6/2), loam, fine granular struc-
ture, friable wnen moist, pH 5.0.

Moderate yellowish brown (10 YR 5/4),
loam, fine granular structure, fria-
ble when moist, pH 5.0.

Grayish yellowish brown (10 YR 5/5),
light clay loam, weak medium angular
blocky structure, sticky when wet,
pH 5.2.

Moderate yellowish brown (10 YR 4/4),
clay loam, plastic wnen wet, pH 6.8,
medium to coarse angular blocky
structure.

Grayish yellowish brown (10 YR 5/2),
light clay loam, with dark grayish
(10 YR 4/2) mottles, medium to coarse
angular blocky structure, glacial
till, calcareous.

Location:

Vegetation:

Physiography:

SlOpe:
Profile:
Horizon

A1

As

27

Parkhill Loam

T 11 N - R 12 E Section 50 NW} offNEi,iSanilac
County, Michigan.

The present cover is American Elm, and White

Birch.

The original cover was sugar maple,

elm, hemlock, and birch.

The site is located at the western extremity
of the Yale moraine, of Cary age.

Level to depressional.

Depth

6-10"

10-16“

16-42”

42"

Description

 

Brownish black (10 YR 2/1), loam,
fine granular structure, friable
when moist, pH 6.5.

Light grayish yellowish brown

(10 YR 6/2), light loam, weak fine
angular blocky structure, friable
when moist, pH 6.5.

Grayish yellowish brown (10 YR 5/2),
clay loam with light yellowish brown
(10 YR 6/4) mottles, medium angular
blocky structure, sticky when wet,
pH 6.5.

Light grayiSh yellowish brown

(10 YR 6/2), mottled, medium angular
blocky structure, mottled brownish
yellow, light clay loam, sticky when
wet, pH 7.5.

Light grayish yellowish brown (10 YR
6/2), mottled light yellowish brown
(10 YR 6/4), loam to light cla loam
till, pH 7.80

28

Installation of Electrodes

Electrodes, prepared in the manner described by Guis-
pel (29), were installed in groups. COpper connections were
made of different lengths so that the electrodes could be
installed at the selected depths. The most suitable time
for installation of the electrodes was in early Spring after
the frost was gone but while the soil was still saturated.

At that time there was little difficulty in forcing the elec-
trodes into the soil and little breakage occurred. After

the soil became dry the electrodes were checked to make sure‘
that the soil had not shrunk away to allow an excess of air
to enter the soil.

After a general examination of each site to select a
suitable profile, three auger holes were made three feet
apart in a triangular pattern as shown in Figure 2. The
depth and thickness of each horizon in the profile was mea-
sured. Holes were then punched in the wet soil within the

triangle to the required depth with a sharpened steel rod.

o<———— 5Pt———->o

\ “2w. /
SST. 8PT.

Figure 2. Layout of Electrodes gp situ.

29

The probe was selected of a dianeter slightly smaller than
the electrodes to insure a tight fit and thus prevent the
passage of excess air to the electrode. Electrodes of the
prOper length to reach the desired soil horizon were then
forced carefully into the soil through the holes made with
the probe.

Three sets of electrodes were installed in the Marlette
loam at the Kenneth Knight site and two sets were installed
in McBride fine sandy loam at the Dan Jurn site. Each of
these sets consisted of five electrodes. An additional two,
sets consisting of three electrodes each were installed in

the Parkhill loam at the Kenneth Knight site.
Measurement of Redox Potentials

The electrodes were installed in the latter part of
April and readings with a Beckman vacuum tube potentiometer
were taken weekly during the first month. Later they were
taken only monthly. The saturated calomel electrode served
as the Standard, or reference electrode. The cell was as
follows:

Pt / HgCl, KCl // Soil / Pt

The KCl bridge was formed by wetting the soil with a
saturated KCl solution, and inserting the calomel electrode
a few inches into the soil. The potentiometer could then

be connected to the calomel and platinum electrodes, and the

50

reading taken. Polarization of the platinum electrode was
prevented by making an approximate adjustment of the poten-
tiometer before completing the connection. Then only small
corrections were required to get the reading and a minimum
of current flow took place. Several precautions were necessary
to insure satisfactory readings. The electrode leads had to
be perfectly dry to prevent shorting of the circuit when the
connection was made. The most suitable time for taking read—
ings was in the early afternoon. The readings were easily
affected by plants in close proximity, If the plant stems
touched the leads they affected the readings. In the early
spring when the soil was saturated and there was a breeze,
the trees rocked; and the r00ts in the vicinity of the elec-
trode disturbed the soil enough to cause very unsteady read-
ings. At the same time when the soil was saturated, the soil
solution was very dilute and therefore poorly poised. This

resulted in some very unsteady readings. Later when the soil

was drier, the readings were quite stable and reproducible.

51

Saturated calomel
El ctrode Potentiometer

 

i_ If OE
:§A\\\\\\\\<S>\ >\‘§§§f;:--§$$$SS>§§$6QQQQS§§SS§§§

\ Surface -

A
w
A

   

 

  

l

H
I

t Soaked with saturated
KCl solution to insure

Sub8011 good contact

Platinized Platinum
Electrode

Figure 5. Method of making redox
measurements in _s__i_t_u.

The electrodes were set out on the 25th of April and
readings taken regularly until the 10th of November. The
electrode sets at the Dan Jurn site were then dug out, pro-
file descriptions taken, and samples collected for laboratory
analysis. In November the weather was too cold for working
out of doors. The Beckman pH meter would not function pro-
perly below 10°C, and samples for pH measurement had to be
taken to a warm place before pH readings could be taken. Due

to these difficulties the remaining electrode sets were left

32

in the soil over winter to observe what occurred. Little
change was observed during the early part of the winter;

and considerable difficulty was experienced getting readings
through the snow. Frost heaved the electrodes; and they had
to be reset a number of times. ioisture shorted out the
readings, and great care was required to get readings. Spe-
cial equipment is needed to measure redox potentials during
winter weather; and to prevent heaving of the electrodes by
the frost.

Figures 4, 5, 6 show the data for McBride fine sandy
loam, Marlette loam and Parkhill loam. All results were
corrected to pH 5.0 to make them comparable and at the same
time to represent the natural soil condition as closely as

possible.

- .. —-'-——«..‘

- ., ‘qr-~—-—~ q... -

‘ ‘a--- .b— vqm.’ now-L... .fi '
.

r
>
1 "
r
l , g’
, v
I
' a
l
I
' '2
‘ u
!_ ‘
‘ D
‘ .
. f V‘
‘ u.
_a
s
I '1
§ 1
t A 2
I t
O

,-{-.:v-t--h-‘~—‘*'~'- ' ‘ ‘
3

i :-
Il .
t b .
l I .
,; , g
’ 'I

Figure 4. The effect of season and

, ; horizon on the redox potentials in
§ icBride fine sandy loam.

 

_‘ r
at: £1051

.fr‘
‘-\J

+

a;
a-

mm._.<a

 

 

 

 

tmmzmkmmm 332‘ .33. 3.3» .35
1 - q q 4 A

-oom-
U
Joo¢-m
moan ..v~ o
lull co? .m. x
.1111... can .b Joomnw
III II mud :n «In...
N
llllllll _( z. 0 .II._
2350: Emma w
4.0m moompouau .
..|I . w
3'3... 323 2:... mean 0: loom w
1
w

00
v n
s
\ 00m

 

 

 

OOO

54

m v.11
e o
a. I
It
‘4 ..
m
. n
N «.1.
III (JR. .lrha‘v. L
at. .0- I 4
I 5...-
.I»
‘ ii-.. 0... t 0 80‘ 4|.
‘ hr O
I}! ‘ Iv ‘ .5
a J U.v
a.."vi ‘l.l-nn.nl.vN3l
392‘ . a: g . ‘0.
I. I." .1 0‘ t
‘-
fiillntv.“ ‘&
- "
‘9’! a: I
I O 1 a. I.“ I I i 'Il ( II.
' I» z -
’ ('9..-
ll.

..' I}
. ..

t) c

o.-

The effect of season and

orizon on the redox potentials in

5.

gure

-W Marlette loam.

Pin s
l. . | I! ..V
1'

‘

. .... . N
. and girl! I i 4.. 4 .. \‘(H I“!

R‘ .
.n‘.a .cI .\
9‘. t
.0. 3|
.5! 6‘ ‘ Ir- ..
.‘v‘fl

nu!“ n.

 

.3

. _

.. . n .

.. a

I I n .

as. C. . A».

-. . . x 9

.r to \\ to‘ r» QI .

m. 5 I‘m. a: . .. ‘
t . . \r.

.t 1! u ‘v .
(I " lbw-DIV- ‘ I L“ \‘b 7. A‘ {M
In \I... e. . . .N i\.
.. V ‘ o .9) a
II . JP 1....
I I .3..\ kih \
.N 21:: III: F
v I

‘5 .k r 3“... . C .
CL 4

x‘ I" -

0V

fin» noaaae To Jaazta 95? ” smugiq

HI airtdnejoa rotai SuJ no noaiwod

é r n P
."Jf‘uCI Q + 'N'

vwv .A .. ..-&

mm._.<o

 

 

 

 

 

tumtmkmmm «mass .33. 3.3» .3).
q! q 4 q - J
Loom-
a
o .3 . oov- u
I I ll DQND ON 0
.. x
II I | mGN< ..N.
I ll mum gm - 084M
IIIIII lN< :n o m
11111- .< .._ \\I O U
23:3: :53 \ w
4.0m moocsouau . .
N
ll 03 I
x<oa upsu4m<z \ a n
\ M
cow m
x I.
\o 8

 

 

 

_.- n.~. «-

I
‘\
‘0
',I§
«.
a. -.. ;
‘1 u
\
.. I
.
3.
. —
I
t.‘
.‘,
f .
. 2'
t? ~1-
b .
‘ if. ‘
w
.9
P
d
.'
‘ "3:.

Figure 6.d Effect of season and horizon
on the redox potentials for Parkhill loam.

i

'u

1

9

~.‘
\.
I
X

3

N

\

5‘ .5
'I _ '9
. ‘x
\
.“
i ""1 p
I
I
g
."I ,
‘5
s‘
‘t
\\
2‘
y" '

"It

Iv

.A
Ta

9

4 a

‘a"

mmPCO

 

 

khmsswk {mm .5 «Smxyv. x Q§5 N§§S x ‘Q o
. . . . . A
a..-
O
\o x
\ N/o\ . 00¢...
a m :3 o \.
I III o< :N. I /
aaaaa _ < .3 \ /.\ .. com-
ZON E0: :hauo
4.0m moothMJw

 

3(04 4...:x14u

 

 

 

\ . / oom
. z
x o
\ .
\\
\ xx 00..
0’ 9" \
\ o \ com
\
”VGA 10 \o\\\
0"--. ||||||| O lllll 0 ||||||

000

SL'IOM'I'IIN - 'IVILNSLOd X0038

LABORATORY EXPERIMENTS
The Eh—pH Relationship

The data collected in the field could not be considered
comparable until it had'been corrected for pH. A.modifica-
tion of Willis' method (59) was used to determine the EhrpH
relationship. The apparatus, shown in Figure 7, was set up
as follows: a glass electrode, a calomel electrode and two
platinum electrodes were fitted tightly for support into a
large rubber stOpper which would fit a heavy glass tumbler.

 

 

 

 

 

   
   
 

 

N itrogen\ To n. Platinum
Tank cell Electrodes
,, Glass
1 Electrode “ '
_! Stepper
Calomél
Electrode Glass Tumbler
Pyrogallol
Scrubber

 

 

Suspension

 

 

Stirrer

 

 

 

 

 

Figure 7. Nitrogen cell for measuring
EhrpH relationships.
Inlet and outlet holes for nitrogen gas were also cut in the
stapper. A glass stirring rod was fitted into the center of

the stapper through a glass sleeve and adjusted sochat it

57

could be rotated with an electric motor. Nitrogen gas was
introduced through a small glass tube so that it was distri-
buted over the surface of the soil suspension and the air
was excluded. The nitrogen gas used was first conducted
through an alkaline pyrogallol scrubber to remove all traces
of oxygen. Boiled water was used for preparation of the soil
suspensions. A 1:1 suspension could be used and this made it
possible to measure both Eh and pH at the same time. The pH
values of the soil suspension were varied by the addition of
different amounts of 0.1 N sulphuric acid and 0.1 normal so-_
dium hydroxide. The sample was stirred for five minutes and
the Eh measured as soon as the streaming potential had sub-
sided and the reading became steady. The pH was measured at
the same time. Beckman equipment was used for both Eh and pH
measurements.

In Figure l of the appendix it can be seen that there
is a straight line relationship between Eh and pH. This
agrees with the findings of Willis (59). However, results
also indicate that the theoretical values of fifty-nine milli-
volts is not a constant for all soils but may be a value which
is characteristic for each soil. This result agrees with the
findings of Volk (53). There was not a significant difference
between values for the individual horizons in this case.
Therefore the Eh-pH relationship should be determined for each

scil studied.

58

Using these data, all horizons were corrected to pH 5.0
in order to make the readings most representative of the
natural soil condition. For the Dan Jurn Site this required
the least amount of correction. For the Kenneth Knight Site
the samplesvvere not taken until the following summer. For
this reason, Willis' theoretical value of 59 millivolts was
used to correct the readings for the Marlette and the Park?

hill loams.
The Laboratory Experiment

All results obtained in the field were extremely variar
ble. Therefore, there was a large experimental error. For
this reason, it was decided to set up a laboratory experiment
to try and control some of the variables in such a way as to
reduce some of the variability and still simulate natural
field conditions as closely as possible. Enough replicates
were used so that the data could.be analyzed statistically
for significance. By this :methodht, it was hOped that the
experimental error could be evaluated. For this experiment,
a sample of the Cl horizon taken from a Weare sand profile
on an old horse shoe dune in Sanilac County was used. At
the same time, a sample of an A0, A1 mixture was obtained.

The profile description follows:

59

Wears Sand

Location: T 15 N - R 15 E Section 22 BWi of the NWi
Sanilac County, Michigan. The site is loca-
ted lfi miles south of the village of Argyle
on the east side of M—lS.

Vegetation: The site supported a moderately dense cover
of Trembling Aspen with a ground cover of
Wintergreen and Bracken.

Physiographic
position: The site was located on an old sand dune in
the interlobate region of the Thumb region
of Michigan.

Profile:

A1 0-5 inches Grayish brown (7.5 YR 3/2),
loamy sand, high in organic
matter, pH 5.0.

A3 5-12 inches Grayish yellowish pink (7.5
YR 7/2), sand, pH 4.5.

B2 12-50 inches Strong yellowish brown (7.5
YR.5/6), loamy sand, discon-
tinuous ortstein, pH 4.5.

Cl 50 inches Light yellowish brown (10 YR

7/6), sand, pH 5.0.

A large sample was collected from the Cl horizon and
allowed to dry. The sample was then thoroughly mixed to
eliminate variability of the material and to get a reasonably
homogeneous sample particularly with respect to pH. This elim-
inated the need for determination of the Eh-pH relationship.

At the same time a sample of the A0 and A1 horizon was taken,
dried, and thoroughly mixed. Using these materials, the
following experiment was performed to simulate, as closely as

possible, natural field conditions.

A tension table was constructed using the purest obtain-
able plaster of paris with an arrangement of glass beakers
for adjustment of the water tension. A cylinder of three
inch metal downspouting twenty-four inches long was painted
on the inside with Tygon paint, placed on the tension table
and supported by a ring,stand. The Joint between the table
and cylinder was sealed with wax to make it water tight. The
cylinder was then filled with the sand from the Weare sample
to within three inches of the tOp, and four platinized plat-
inum electrodes prepared as in the field experiment were
placed in the sand at selected depths for redox measurements.
An additional four shiny platinum electrodes were placed at
the same depths, and used to measure oxygen diffusion accord-
ing to the method of Lemon (18). The cylinders were then
filled with about 2% inches of the A0, A1 mixture. Two methods
of watering were then compared, with each treatment replicated
five times. A third treatment, also replicated five times,
provided for only enough water to moisten the organic layer.
Due to difficulties encountered with this treatment it was
discontinued and the results are not shown. The apparatus
for a single sand column and the entire green house set-up

are shown in Figures 8 and 9.

41

:‘fi:

 

 

 

c‘te of the land

_ L
l
w
.J... . .
ALrbpl
. rinentj,

 

 

 

...}::j
l

 

 

 

 

 

 

1
.1”.

 

column Q

"fill ' II .. | 'I I‘I Ill- 1“

. «a.

 

I
Figure 8

 

 

(3’ 'rfil":'7”l. 0 W

._‘1"

"x
#4.

.J

. z . ,. .. r-
33:Ia91 A .9 9Iu,lu

.Jnemiiaaxe nthioo

ELECTRODES

 

A l I
11‘ La
IIITI
I I

ORGAN/6 LAYER

11

 

 

 

 

A v A; I A
4' [III

 

 

' I 1.: SAND LAYER

 

 

 

 

 

 

 

 

   

.. :°.'SEALIN6 WAX
% mwumnmmwy E R 0F PA R15
CA KE

 

Ll}.-- __.. WATER LEVEL
avmnow]

 

 

 

——
’7

 

42

 

(I

"P

(1’

P:f(:‘ ‘I‘ "
.... - fl.-. 3

males

,5

r
A

v
A

PI
f\

S)
..I

913'?

.8

Sinai?

 

45

Treatment A was continuous leaching with distilled water.
Treatment B was leaching every three to four days, each time
with 100 cc. of distilled water. Due to the continuous water
flow, in treatment A, and the solubility of the Plaster of
Paris used, it was very difficult to maintain the tension on
the tables. In Spite of these difficulties some results were
obtained. The tension tables in treatment B worked very well
throughout the course of the experiment. Redox potentials
were measured every three to four days for six weeks using
the Beckman vacuum tube potentiometer and the saturated calf
omel electrode in the same manner as for the field experiment.
Results are shown in Tables II and III of the appendix and
Figures 10 and 11 of the text.

At the same time, using the shiny platinum electrodes,
and a specially prepared saturated calomel electrode, oxygen
diffusion readings were obtained by the method of Lemon (18).
The instantaneous readings are the initial swing of the
ammeter needle and are a measure of the oxygen concentration
at the electrode. The diffusion readings are a measure of
the rate at which oxygen diffuses to the electrode. The re-
sults obtained are shown in TablelV of the appendix.

At the conclusion of the experiment, the sand columns
were Opened and examined. Samples of the sand taken at se-
lected depths, were extracted with normal ammonium acetate

solution buffered at pH 4.8 with acetic acid. The extracts

44

were tested for ferrous iron using a modification of Olson's
method (25) which uses o-phenanthroline as the color reagent.
The intensity of color deve10pment was measured on the Cole-
man Junior SpectrOphotometer set at a wavelength of 480
millimicrons. The extracts were then treated with hydroxyl
amine hydrochloride to reduce all the iron to the ferrous
form, and the iron was again determined. This gave a measure
of the proportion of the total extractable iron that was
ferrous in form. The moisture content of the samples was
also determined at the same time. Results of the iron deter-
minations are shown in Table I of the text and the moisture
content of the sand columns at the time of Opening in Table

TV'of the appendix.

RESULTS AND DISCUSSION

There is considerable variability between readings of
electrodes in any horizon $3 gitg due to the small portion
of the soil body sampled. Variations in texture, structure,
organic matter distribution, aeration, or moisture content
of the soil as well as the proximity of the electrode to
r00ts or worm channels cause variations in the redox poten-
tials. Replacement of the electrodes during the course of .
the eXperiment indicated that the error was not due to effects
on the electrodes themselves. The error caused by this varia-
bility could be evaluated statistically by the use of many
replicates. Further error may be introduced unless care is
used in the measurement of the potential in the field.

Large junction potentials can be introduced unless care is
used in making the contact between the calomel electrode and
the soil.

In Spite of the limited number of electrodes used in this
exploratory field eXperiment the trends are very interesting.
The results of the field experiment are shown in Figures 4,

5 and 6. The redox potentials for McBride fine sandy loam,
Marlette loam and Parkhill loam are shown plotted against
dates. The initial readings were taken immediately after the

electrodes were inserted on April 28 and probably before they

0
5

Figure 10. Effect of depth on the
redox potentials for treatment A of
the sand column eXperiment.

46

DEPTH—INCHES

a

-
(I
T

22.

 

 

 

j l j A J

 

I00

200 300 400 500 600

REDOX POTENTIAL -M|LLIVOLT$

Figure 11. Effect Of depth on the
redox potential for treatment B of
the sand column experiment.

47

'I
Am F + ~ -v w r
O! “J C") Xk ~31: I

we"...
0..
‘Ar..
..
no.
?. .L
‘1

.. . f.
T.
.11.. .q..
.IwJ .isf.
m .9
_ . .
I. 4
r
C . .
C n
3..
0.4
. 3

e

v ;. 'V 4..

'V'
17"..

O
.\I

300 400 500 600

200

 

 

- . »

 

IOO

9 5

mmroz. I rho-ma

O‘-
3
22

REDOX POTENTIAL -M|LLIVOLTS

48

came to equilibrium with the soil. After a week the soil
had settled sufficiently that equilibrium was established.
The redox potentials for McBride fine sandy loam Figure 4,
had risen very high by the second week in May with little
variations between horizons. In the Marlette loam, Figure
5, the redox potentials had not reached a level comparable
to those in McBride fine sandy loam until the third week in
May. They were also more variable between horizons than
were those for the McBride in the first weeks.

Since both soils were well drained, the differences in
redox potential were mainly associated with differences in
texture throughout the profiles. McBride fine sandy loam
dried out and warmed more quickly than did the Marlette loam.
During the summer, the redox potentials for both soils were
high, probably because of improved aeration. There was also
a tendency for higher readings in the lower horizons than
in the surface horizons.

It has been generally observed (6, 8, ll, 15, 16, 29,
54, 55) that organic matter tends to lower the redox poten-
tial of the soil. Since the surface layers of the soil, in
general, contain more organic matter than the subsoil layers,
these observations are substantiated in McBride fine sandy
loam and Marlette loam. Figure 6 shows the results obtained

in the poorly drained Parkhill loam. These results are of

49

interest since this soil was waterlogged throughout the
Spring and the soil had not completely recovered from the
effects of this condition until July let. The early read-
ings were extremely low.

The readings were lowest in the twelve inch depth,
indicating the zone of highest reduction in the profile.
After July let, the redox potentials compared favorably with
those of the well drained soils. These results indicate
outstanding seasonal differences in the Parkhill loam as
compared to the well drained soils; the difference being
associated with variations in drainage.

In the sand column experiment, the redox potentials were
treated statistically and checked by Nair's quick test (21).
Treatment A was not considered a good test since the tension
tables gave so much difficulty. However the means for both
treatments are shown in Tables II and III in the appendix
together with the statistical analysis for treatment B.
Results for both treatments are similar. There is little
difference between the general means for both treatments.
The average redox potentials plotted against depths for five
replicates are shown in Figures 10 and 11 for both treatments
A and B. The potentials in the three inch depth were signi—
ficantly lower (at the five percent level) than they were at
lower depths. Oxygen diffusion readings by Lemon's method

(18), shown in TableIV in the appendix, were also higher in

50

the deeper portions of the profile. Examination of Table
I shows that by far the greatest amount of ferrous iron was
contained in the three inch layer. Generally the proportion
decreased with depth in the column. This would indicate

that iron was being reduced in the three inch layer. In

 

this state iron is more soluble than in the ferric state. It 5'
could therefore move downward in the leaching waters to the
lower levels. Since there existed in the soil column a state

of higher oxidation in the lower levels, the ferrous iron

 

could be oxidized to ferric iron and partially precipitated. {an
Although certainly not conclusive, these data are considered
as evidence of a possible method whereby iron may be translo-

oated in the soil and precipitated in the B horizon.

 

 

 

TABLE I. FERROUS IRON CONTENT OF THE EXTRACTABLE IRON FOR
ThEAkaNTS A AND B.
BEpth Treatment
Inches A2 A5 A4 Bl B2 B5
5 52.5 55.0 47.5 58.5 42.5 52.0
9 14.1 40.0 57.5 18.4 16.8 14.2
15 15.0 15.2 26.5 15.5 24.5 28.0
22 10.6 15.2 18.0 12.6 52.0 50.5

51

This is not due to the redox potential itself, but to the
combination of factors which results in the differential
redox potentials.
If the process involved in the deve10pment of Podzols
is defined as the translocation of sesqui-oxides from the
A horizons and their precipitation in the B horizon, then
the sand column eXperiment would indicate a possible mech-
anism important, among others, in the deve10pment of Podzols.
The field results indicate that the season of the year when
this process would be most active on well drained soils is
in the early spring. More field studies are now underway.
The sand column experiment has shown that the use of
sufficient replicates and a statistical treatment of the da-
ta is feasible when some of the variables are controlled.
It may also be feasible to use statistical methods in the
field and thus make it possible to overcome the large experi-

mental error involved in the use of this tool.

SUMFARY

Redox potentials have been measured with variations in
time and space in McBride fine sandy loam, Marlette loam, and
Parkhill loam. In addition, redox potentials were studied
in sand columns in a greenhouse eXperiment. The following
observations were made:

(1) Extremely low redox potentials persisted until
July 1 in poorly drained Parkhill loam. .

(2) The potentials observed in the well drained McBride
fine sandy loam, and Mariette loam, were high in early spring,
indicating that the state of oxidation of these two soils
would not limit plant growth, even early in the growing
season.

(5) The Marlette loam showed a tendency toward a lower
state of oxidation than McBride fine sandy loam until about
May 12th.

(4) The protracted measurements of redox potentials of
the various soil horizons in the field gives a more complete
picture of the true organic conditions in the soil body.

(5) Redox potentials measured ig_§itg_may be useful as
a tool for checking soil conditions affecting plant growth,

particularly in the case of poorly drained soils.

55

(6) A redox profile has been observed in a laboratory
study on sand columns which is submitted as evidence of a
possible mechanism of Podzol formation. Reduced iron in
solution was produced in the upper part of a well drained
sand column. This could be precipitated lower in the sand
column by oxidation. Redox potentials were significantly
higher, oxygen concentrations were greater, and the propor-
tion of soluble iron in the ferrous state was less lower in
the sand columns.

(7) The sand column experiment indicates that many
replicates and the use of statistical methods might make
the use of redox potentials feasible in the field.

(8). It is suggested that the measurement of redox
potentials may be useful in studies of soil formation, par-
ticularly of Podzols, Ground Water Podzols, and Humic Gley

soils.

(l)

(2)

(3)

(4)

(6)

(7)

BIBLIOGHAPHY

Bertramson, B.R. and White, J.L.
Soil Chemistry Notes.
Student Book Corporation
Wasnington State College
Pullman, Washington

Bouyoucos, G.J.
The effect of chemical agents on oxidation
in soil forming rocks and minerals.
Soil Sci. 15: 19-22 (1925).

Bradfield, R., Batjer, L.P. and Oskamp, J.
Soils in relation to fruit growing
in New York. Part :Z.
The significance of oxidation-reduction
potential in evaluating soils for orchard
purposes.
Cornell Univ. Agr. Exp. Sta. Bull. 592.
Ithaca, New York. 1954.

Brown, L.A. Oxidation-reduction potentials in soils:
principles and electometric determinations.
Soil Sci. 57: 65-76 (1954).

Buehrer, T.F. et. al.
The oxidation-reduction potentials of
alkaline calcareous soils in relation to
puddling and organic matter content.
Jour. Am. Soc. of Agron. 51: 905-914. (1959).

Burrows,W. and Cordon, T.C.
The influence of the decomposition of
organic matter in the oxidation-reduction
potentials of soils.
Soil Sci. 42: 1-10 (1956).

Clark, W. Mansfield, et a1.
Studies on oxidation and reduction.
Hygienic Laboratory Bulletin No. 151.
United States Public Health Service.
(1928).

 

(8)

(9)

(.10)

(11)

(12)

(15)

(14)

(15)

55

Darnell, M.C. and Eisenmenger, W.S.
Oxidation and reduCtion potentials of
soil suspension in relation to acidity
and nitrification.
Jour. Agr. Research 55: 75-80 (1956).

Gardner, D.R. and Whiteside E.P.
Zonal soils in the transition region
between the Podzol and Gray Brown Pods
zolic regions in Michigan.
Soil Sci. Soc. Am. Proc. Vol. 16, No. 2
Page 157 (1952).

Fillespie, L. i
Reduction potentials of bacterial cultures :
and or waterlogged soils.
Soil Sci. 9: 199-216 (1920).

 

Heintze, G.C.
The use of the glass electrode in soil
reaction and the oxidation reduction
potential measurements.
Jour. of Agr. Sci. 24: 28-41 (1954).

Jenny, Hans.
Factors of Soil Formation.
MoGraw Hill Book Co. Inc.
New York and Landon.

Joffre, Jacob, S.
Pedology, 2nd Ed.
Pedology Publications, New Brunswick,
New Jersey.

Keaton, C.M. and Kardos, L.T.
Oxidation reduCtion potentials cf arsenate
arsenite systems in sand and soils.
Soil Sci. 50: 189-208 (1940).

Konnke, H.E.L.
Some factors afiecting the oxidation
reduction pctentials or soils.
Reprint from Abstract of Doctors'
DissertatiOn, No. 15
The Ohio State University Press, 1955.

(16)

(17)

(18)

(19)

(20)

(81)

(22)

(25)

(24)

56

LaFond, A.
Oxidation reduction potential as a
characteristic of forest humus types.
Soil Sci. Soc. Am. Proc. 14: 557-540.
(1949).

Leeper, G.W.
Factors affecting availability of inor-
ganic nutrients in soils with special
reference to micro—nutrient metals.
Annual Review of Plant Physiology.
Vol. 5 (1952).

Lemon, E.R., Erickson, A.E.
The measurement of oxygen diffusion in
the soil with a platinum micro-electrode.
Soil Sci. Soc. Am. Proc. 16: (2) 160-65
(1952).

MEP'CU‘L, COFO
Soils: Their Genesis and Classification.
Published by the Soil Sci. Soc. of Am.
1951. .

Millar, C.E. and Turk, L.M.
Fundamentals of Soil Science.
John Wiley and Sons, Inc.
New York.

Nair, K.R.
The distribution of the extreme deviate
from the sample mean and its studentized
form.
Biometrics 55: 118-144 (1948).

Neller, J.R.
The influence of growing plants on oxida-
tion processes in the soil.
Soil Sci. 15: 159-155 (1922).

Olson, R.V.
The use of o-phenanthroline in the deter-
mination of small amounts of ferrous and
ferric iron in soils.

Soil Sci. Soc. Am. Proc. 12: 155-157 (1947).

Peecn, M. & Batjer, L.P.

A critical study of the methods for measur-
ing oxidation-reduction potentials of soils

with Special reference to orchard soils.
Cornell Univ. Agr. Exp. Sta. Bull. 625.
Ithaca, New York 1955.

 

(25)

(26)

(27)

(28)

(29)

(50)

(51)

(52)

(35)

57

Pierce, R.S.

Oxidation-reduction potential and specific
conductance of ground water: Their influence
on natural forest distribution. '

Soil Sci. Soc. Am. Proc. 17: 61-65 (1955).

 

;—

Puri, A.N.

Puustjarvi,

Quispel, A.

Roberts, J.

Prairie like mull humus, its physio-chemical
and micro-biological prOperties.
Soil Sci. Soc. Proc. 15: 562-564. (1950).

and SarUp, A.
Oxidation-reduction potentials in soils.
Soil Sci. 46: 525-529 (1959).

VilJo and Juusela, Taneli.

Rust precipitates present in drainage pipes
and the prevention of their formation.
Dept. of Agr. Chem.

Helsinki Univ. Sweden (1952).

 

4’:

Measurement of oxidation-reduction potentials
of normal and inundated soils.
Soil Sci. 65: 265-275 (1947).

L.

Reduction of ferric hydroxide by strains of
bacillus polymyxa.

Soil Sci. 65: 135-140 (1947).

Robinson, C.W.

Soils: Their Origin, Constitution, and
Classification.

John Wiley & Sons Inc.

440 Fourth Avenue, New York. (1949).

Russel, E.J.

Stevenson,

Soil Conditions and Plant Growth. 8th Ed.
Revised by E.W. Russel,

Longmans Green and Co.

London, New York, and Toronto.

R.E. et a1.

Oxidation-reduction potentials in orchard
soil.

Jour. Am. Soc. of Agron. 50: 91-96 (1958).

(:55)

(36)

(37)

(58)

(40)

58

Sturgis, M.B.

Volk, N.J.

 

Changes in redox equilibrium in soils as
related to the physical prOperties of the
soil and the growth of rice.

La. Agr. Exp. Sta. Bull. 271. (1956).

The oxidation—reduction potentials of

Alabama soils as affected by soil types,

soil moisture, cultivation, and vegetation.
Jour. of Am. Soc. of Agr. 51: 577-589. (1959).

The effect or oxidation—reductiOn potential
on plant growth.
Jour. of Am. Soc. Agron. 51: 665-670 (1959).

The determination of redox potentials of
soils.
Jour. of Am. Soc. of Agron. 51: 544—551 (1959).

Wilde, S.A., Trach, J. and Peterson, S.F.

Electro-chemical prOperties of ground water
in major types of organic soils.
Soil Sci. Soc. Am. Proc. 14: 279-281 (1949).

Willis, L.G.

Oxidation-reduction potentials and the hydro-
gen ion concentration of a soil.
Jour. of Agr. Research 45: 571—575 (1952)

 

Evidences of the Significance of oxidation
reduction equilibrium in soils.
S.S.S.A. Proc. 1: 291-297 (1956).

 

APPENDIX

 

 

 

60

 

(
700
6CH)
0
2:
O
.2
EE‘QOC)
E
1
.:
Ill
ZCHJ
2 3
FIGURE 1.
TABLE I.

pH

Eh - pH Relationship for McBride
fine sandy loam.

TREATMENT A BY DEPTH AND DATES

.-
‘

AVERAGE VALUES OF FIVE REPLICATES OF

—_
__

.— -" "u‘.’ 11
.]

 

 

 

depths General
dates 1 2 5 4 5 6 7 8 9 Mean
5 420 409 588 567 569 412 221 491 522 578
9 459 450 455 465 490 488 565 444 592 458
15 554 555 559 608 601 572 520 544 575 565
22 482 498 495 585 575 508 485 489 557 497
Means 474 475 464 505 509 495 598 492 412 469

 

469

 

61
TABLE II. TABLE OF MEANS FOR TREATMENT E OF
THE SAND COLUMN EXPERIMENT.
depths Means

dates 1 2 5 4 5 6 7 8 9 10

 

5 in. 554 581 574 420 474 542 556 511 512 506 561
9 in. 477 499 477 515 549 477 479 470 487 475 490
15 in. 556 551 541 576 577 541 557 501 492 511 556

 

22 in. 416 420 446 448 460 478 488 456 457 455 452

 

446 465 460 489 515 460 460 455 457 456 460*

 

*By using K.R. Nair's test (21) we find that the average for
the three inch depth is significantly different from the
general average and of course is significantly different from
all the other averages.

 

 

62

 

 

 

 

TABLE III. ANALYSIS OF VARIANCE FOR DAYS, DEPTHSV
AND REPLICATES FOR TREATKENT s 09 THE
SAND COLUMN EXPERIMENT
D.F. 5.5. M.Sq. F. Nec.F.
Total 199' 5,850,589
Rep. (R) 4 1,595,499
Depth (D) 5 829,499 276,500 5.08 2.87
Days (d) 9 '117,170 15,019
d . D 27 97,415 5,608
D - R 12 1,078,077 89,875*
d - R 58 109,512 5,042
Rod.D 108 227,417 2,106
Err. 158 1,415,006 8,584

 

”Using the largest possible error term a difference of 85

millivolts or greater is significant at the five percent level.

 

63

 

 

Nv.wH mm¢

 

 

mm.p ms.m m¢.m mme mm s
so.ma mm.p~ mom on.n ma.m omm ma n
nm.m m.mH new 98.” ms.n men m m
en.m mm.e . mom em.m 40.9 4mm n H
Amnsw.av A mad.av A>2v Amnaa.a~ Ammad.av A>EV monocH .02
o .Haa o .afim gm 0 .ean o .adm .nm apnea
55m
m uncapwmhe 4 pcmspwmhe ouchpooam

 

 

 

 

 

 

.m Qz«_4 mazmfie¢mma 2H
mama IO...“ mGZHodmm Em EH3 Qmmdngao zmwwxo
mom m¢ZHQ4mm ZOHmDthQ Q24 mDOH24Hz<emZH

.>H mqm¢e

 

64

TABLE V. MOISTURE VARIATION IN THE SAND COLUMNS
WITH DEPTH FOR TREATKENTS A AND B.

 

r*

Electrode 1 Moisture Content in Percent
Set

 

Depth
NO. inches A2 A5 A4 B1 B2 B5

 

1 5 8. 7 9. 8. 7. 12.
2 9 12. 10. 11. 12. 9. 10.
5 15 14. 15. 15. 18.. 14. 10.
4 22 18. 15. 17. 15. 15. 14.