Surface chemistry of calcium and phosphorus retention in selected acid tropical soils from the Republic of Vietnam

Material Information

Surface chemistry of calcium and phosphorus retention in selected acid tropical soils from the Republic of Vietnam
Tinsley, Richard Lee, 1943- ( Dissertant )
Popenoe, Hugh ( Thesis advisor )
Place of Publication:
Gainesville, Fla.
University of Florida
Publication Date:
Copyright Date:
Physical Description:
xiii, 160 leaves : ill. ; 28 cm.


Subjects / Keywords:
Acidity ( jstor )
Adsorption ( jstor )
Anions ( jstor )
Colloids ( jstor )
Crops ( jstor )
Ions ( jstor )
Isotherms ( jstor )
Minerals ( jstor )
pH ( jstor )
Soils ( jstor )
Dissertations, Academic -- Soil Science -- UF
Soil Science thesis Ph. D
Soil acidity ( lcsh ) ( lcsh )
Soils -- Vietnam ( lcsh )
bibliography ( marcgt )
non-fiction ( marcgt )


Soil colloids from six acid tropical pedons were identified and grouped as either predominately constant-charge or predominately constantpotential type colloids. The surface chemistry of Ca and P retention of the two colloidal types was contrasted and related to crop response to Ca amendments applied to the different pedons. The two Ultisol pedons contained predominately constant-charge phyllosilicate colloids similar to the colloids of most temperate soils. With this type of colloid, the surface charge results from ion substitution in the silicate minerals which provides a permanent negative charge on the colloidal surface. In this system, acidity results from H and Al3+ ions satisfying the negative charges, and inorganic P exists as Fe and Al phosphates. Liming the two soils that had constant-charge colloids neutralized the acidity, increased the base saturation, and increased the P availability by raising the solubility of the Fe and Al phosphates. These reactions provided a more favorable plant environment. The four Oxisol pedons contained predominately constant-potential colloids in the form of Fe and Al sesquioxides, non-crystalline material, and organic matter. In this system, the colloidal charge results from the adsorption of' potential-determining Jr and 0H~ ions. Since these soils are highly weathered, the soil reaction approaches that of maximum colloidal stability which is the zero point of charge. These soils, therefore, have a low surface charge for H and Al3+ td satisfy. Thus, these soils are low in exchangeable acidity even at low pH values. Also, in this colloidal system, P is tenaciously adsorbed directly to the colloidal surface by ligand exchange. The availability of adsorbed P is more dependent on competitive anions than pH. Thus, liming the pedons containing predominantly constant-potential colloids neither neutralized appreciable exchangeable acidity nor increased the availability of P. Instead, liming rapidly increased the surface charge density by adsorbing 0H~ ions, and this charge was satisfied by adsorbing Ca. The Ca so adsorbed was retained with greater tenacity than that on the exchange complex of pedons with predominantly constant-charge colloids. Greater crop response on the pedons dominated by constant-potential colloids occurred when competitive anions were added to reduce the P adsorption energy and increase the P desorption rate than when the acidity was neutralized with lime.
Thesis (Ph. D.)--University of Florida, 1974.
Includes bibliographical references (leaves 154-159).
General Note:
General Note:
Statement of Responsibility:
by Richard Lee Tinsley.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
000566499 ( alephbibnum )
37900555 ( oclc )
ACZ2931 ( notis )


This item has the following downloads:

Full Text








Phi Lan and Elaine


The author wishes first to express his sincere appreciation to

the members of his Supervisory Committee, all of whom individually

assisted in some portion of this project. Sincere appreciation is

extended to Dr. Hugh L. Popenoe, Chairman of the Supervisory Committee,

for his advice and counsel, his acute eye for grammatical detail in

the preparation of this manuscript, and for providing the assistantship

necessary to finance this research study; to Dr. Lucian W. Zelazny,

Co-Chairman of the Supervisory Committee, for his daily supervision of

the laboratory work, and his stimulating pedagogy from which many of

the basic concepts developed and evaluated in this research were

formulated; to Dr. Robert E. Caldwell for his assistance in the char-

acterization and classification of the soils involved; to Dr. Mason E.

IIMarvel who, as Chief-of-Party for the Viet Nam Advisory team, both

administratively and physically assisted in the field plots; and finally

to Dr. Gerald 0. Mott who, during the initial quarter the author

attended the University, asked the leading question which precipitated

the entire study. Additional appreciation is extended to Dr. Darell E.

M~cCloud for his efforts during the initial stages of this study. It is

regretted that his well deserved faculty development leave prohibited

his being able to actively review the final dissertation.

The author would also like to thank Drs. Charles F. Eno and

Donald F. Rothwell, Chairman and Graduate Coordinator, respectively,

for their administrative assistance.

Tlhe author also wishes to extend appreciation to Drs. Nguyen than

Hai, former Director of the National Agricultural Institute, and

Thai cong Tung, former Director of the Agricultural Research Institute

and their respective staffs for the opportunity to use their facilities

and their assistance in conducting the field experiments. In addition,

thanks is given to the Ha-Tien cement company of Thu Due for the

donation of 500 kg of gypsum for use in the field study.

And finally, the author would like to recognize his dearly beloved

wife, Ngoe Loan, whose efforts in this study began with the spreading of

the first increment of lime on the Thu Due plots and ended half a world

away with the typing of the final manuscript. Her efforts have made

this dissertation a truly joint family enterprise. W~e would also like

to mention our two daughters, Phi Lan and Elaine, who have had to do

without much of the love and attention two preschoolers richly deserve,

and to whom this work is dedicated.



ACKNOiLEDGEMENTS ........................................... iii

LIST OF TABLES ............................................ vii

LIST OF FIGURES ............................................ ix

ABSTRACT ................... ................... .......;......... dii

INTRODUCTION ............................................. 1

LITERATURE REVIEWJ ................... ................... ....... 3

Soil Colloids .......................................... 3
Phyllosilicate Minerals ............................. 3
Metallic Crystalline Oxides and Hydroxides .......... 4
Non-Crystalline Materials ........................... 5
Collaidal Systems ................................... 6
Distribution of Colloidal Systems ................... 9

Colloidal Charge Characteristics ......................,,. 10
Source of Colloidal Charge .......................... 10
Zero Point of Charge ................................ 11
Induced Surface Charge .............................. 11
Double-Layer Theory ................... .............. 13
Soil Acidity ............,.,,........................ 17
Liming ........................................... 19

Anion Reactions ......................................... 21
Constant-Charge Colloidal Systems ................... 21
Constant-Potential Colloidal Systems ................ 22
Evaluating Surface Adsorption ....................... 27

Soils Of The Republic Of Viet Nam ................... ....o 29
Mekong Delta ........................................ 30
Eastern Region ...................................... 30
Central Coastal Lowlands ................,,.......... 30
Central Highlands ................................... 31

MATERIALS AND METHODS ......................................... 32

Site Selection ............................o............. 32
Location of Soils ................................... 32
Field Trial Sites ................................... 33
Climatic Variables .................................. 33


Soil Characterization ..n................................ YC
Profile Descriptions ...........................***** Y1
Physical Analyses ................................*** Y,
Colloidal Analyses ................................. 36
Chemical Analyses ................................... 38

Field Trials ........................................... 41
Calcium Treatments .................................. 41
Plot Design ......................................**. 43
Fertilization .....................................** 43
Plot Operation ..................................**.. 44

Calcium And Phosphorus Studies ........................... 45
Calcium Treatments .................................. 45
Calcium Retention Experiment ...............'......... 47
Phosphorus Adsorption Study ......................... 49

RESULTS AND DISCUSSION~ ......................................... 52

Soil Characterization And Classification ................. 52
Profile Descriptions ................................ 52
Physical Analyses ................................... 59
Colloidal Analyses ......................,,,.,,...... 61
Chemical Analyses ................................... 78
Classification of the Six Profiles .................. 95
Additional Soil Properties .......................... 97

Field Trial Results ...................................... 105
Crop Response .......................,,.............. 105
Effect of pH .........................,,,,,.......... 109

Calcium Retention ........................................ 111
Hypotheses ......................................... 111
Results of Leaching Study ........................... 112
Evaluation ......................................... 129

Phosphorus Retention ..................................... 131
Langmuir Adsorption Isotherms ....................... 131
Desorption and Anion Competition .................... 138
Effect' or Ca Amendments on P Retention .............. 140
Fox Isotherms ....................................... 142

CONCLUSIONS .............................................. 150

LITERATURE CITED ........................................... 154~

BIOGIRAPICAL SKZETCH .......................................... 160


1. Climatological data for field trial locations .......... 35

2. Rates of Ca used in the field trials ..............,..... 42

3. Dates of each operation for individual field trials .... 46

4. Ca treatments used in laboratory Ca and P retention
studies .........................................., 48

5. Original concentration of P in adsorption isotherm
studies .***************** *****************. ........ So

6. Profile description of the six pedons examined ...,...... 53

7. Particle size distribution and organic carbon content
for the six soil pedons examined ....................... 60

8. Non-crystalline, vermiculite, and Fe209 analysis of
the six soil pedons examined ........,,,,, 73

9. Quantitative estimate of colloidal components for the
six soil pedons examined ............................... 76

10. Soil reaction for the six soil pedons examined .....;.... 79

11. Extractable and titratable acidity for the six soil
pedons examined .........................,....,,,,,,oo., 84

12. CEC and exchangeable bases for the six soil pedons
examined *************************************... .. .. 92

13. Comparison of ZPC with pH and computation of 0 for
selected horizons from the four principle pedons
examined **************.......... 102

14. Natural P-fractions and Bray extractable P of the six
soil pedons examined *****************.................. 100

15. Effect of applying Ca(0H)2 and CaS04 to five acid pedons
on okra or sorghum fodder **............................ 107


16. Effect of Ca amendments on pH during a 90 days growing
season for the six field trials ........................ 110

17. Effect of Ca amendments and leaching on soil reaction .. 114

18. Effect of Ca amendments on Ca concentration in the
leachate with time ..................................... 119

19. The recovery of Ca from the leaching study ............. 122

20, Effect of Ca amendments on conductivity in leachate with
time .............................................. 125

21. Effect of Ca amendments on K concentration in leachate
with time .....,..................................... 127

22. Effect of Ca amendments on KC1 extractable Al .......... 130

23. Langmuir monolayer coverage, V,, and energy, b, para-
meters for P adsorption isotherm of the four principle
pedons examined ........................................ 137

24. Desorption by CaC12 and CaSOS for adsorption values
closest to e Vm and Vm ********************************.. 139

25. Effects of Ca-amendments on Langmuir P adsorption
parameters of energy (b) and monolayer coverage (Vm) ... 141

26. Effect of Ca-amendments on resin extractable P ......... 143

27. Evaluation of Fox P sorption isotherms andlithe effect of
competitive anions for the Dalat and Eakmnat pedons ..... 148




1. Adsorption of potential determining ions on oxide
surfaces to provide either a positive or negative
surface charge ................... ................... ... 12

2. Helmholtz, Gouy-Chapman, and Stern models for the dis-
tribution of counter-ions with distance from a charged
colloidal surface ................... ................... 14

3. Various types of anion adsorption on oxide surfaces.
The dotted line represents the Stern layer and as equals
surface charge ......................................... 23

4. X-ray diffractographs for Mg-saturated, glycerol solvated
and heated K-saturated clay separates from selected
horizons of the Trang Bang and Thu Due pedons .......... 62

5. X-ray diffractographs for Mg-saturated, glycerol
solvated and heated K-saturated clay separates for
selected horizons of the Dalat pedons .................. 64

6. X-ray diffractographs for Mg-saturated, glycerol solvated
and heated K-saturated clay separates for selected
horizons of the Eakmnat pedons .......................... 65

7. DSC thermographs of clay separates from the Trang Bang
pedon ............................................. 67

8. DSC thermographs of clay separates from the Thu Due
pedon ............................................. 68

9. DSC thermographs of clay separates from the Dalat N
pedon ............................................. 69

10. DSC thermographs of clay separates from the Dalat T
pedon ............................................. 70

11. DSC thermographs of clay separates from the Eakmat I
pedon ............................................. 71

12. DSC thermographs of clay separates from the Eakmat II
pedon ............................................. 72


13. Potentiometric titration curves for the Trang Bang and
Thu Due pedons. iNeq of base are measured from the
beginning of individual curves with each unit equal to
1 meq ...,,,,.,,,............,,.......,..,.,,,,,,,,, 88

14. Potentiometric titration curves for the Dalat pedons.
Meq of base are measured from the beginning of individual
curves with each unit equal to 1 meq ,,,,.o.........,,.. 89

15. Potentiometric titration curves for the Eakmat pedons.
Meq of base are measured from the beginning of individual
curves with each unit equal to 1 meq ........,,.,,.. 90

16. Zero point of charge titrations for the Ap and B22
horizons of the Trang Bang pedon ..................... 98

17. Zero point of charge titrations for the Ap and B21
horizons of the Thu Due pedon .....................o.,,. 99

18. Zero point of charge titrations for the Ap and B21
horizons of the Dalat pedon ............................ 100

19. Zero point of charge titrations for the Ap and B23
horizons of the Eakmnat pedon ........................... 101

20. Ca concentration in the leachate from samples of the
Trang Bang surface horizon which had been treated with
different levels of Ca(0H)2 (L1-Lk) and CaS04 (G1-G3) as
compared to no Ca treatment (C) ....,,,...... 116

21. Ca concentration in the leachate from samples of the
Thu Due surface horizon which had been treated with
different levels of Ca(0H)2 (L1-Lk) and CaSC4 (G1-G3) .. 117

22. Ca concentration in the leachate from samples of the
Eakmat surface horizon which had been treated with
different levels of Ca(0HI)2 (L1-Lk) and CaS04 (G1-G3)
as compared to no Ca treatment (C) .......... 118

23, Ca concentrations in the leachate from samples of the
Dalat surface horizon which had been treated with
different levels of Ca(0H)2 (L1-Lk) and CaS04 (G1-G3)
as compared to no Ca treatment (C) ........... 121

24. Langmuir P adsorption isotherms for the Trang Bang pedon. 132

25. Langmuir P adsorption isotherms for the Thu Duc pedon .. 133

26. Langmuir P adsorption isotherms for the Dalat pedon .... 134

27. Langmuir P adsorption isotherms for the Eakmnat pedon ... 135


28, Techniques for determining P requirements and competitive
anion effects with Fox P sorption-desorption isotherms. 145

29. Fox sorption-desorption isotherm for the Dalat pedon.
Each unit on the Y axis represents 200 pg P/g soil as
measured from individual X axes ........................ 146

30. Fox sorption-desorption isotherms for the Eakm~at pedon
Each unit on.the Y axis represents 200 Clg P/g soil as
measured from individual I axes ........................ 147

31. Effect of Ca(0H)2 (Li-Lk) and CaS04 (G2) on crop growth
for the Trang Bang, Eakmnat, and Dalat pedons compared to
no Ca treatment (C) .................................... 152

Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy



Richard Lee Tinsley

December, 1974

Chairman: Hugh L. Popenoe
Co-Chairman: Lucian dJ. Zelazny
Major Department: Soil Science

Soil colloids from six acid tropical pedons were identified and

grouped as either; predominately constant-charge or predominately constant-

potential type colloids. The surface chemistry of Ca and P retention

of the two colloidal types was contrasted and related to crop response

to Ca amendments applied to the different pedons.

The two Ultisol pedons contained predominately constant-charge

phyllosilicate colloids similar to the colloids of most temperate soils.

With this type of colloid, the surface charge results from ion substitu-

tion in the silicate minerals which provides a permanent negative charge

on the colloidal surface. In this system, acidity results from H+ and

Al"+ ions satisfying the negative charges, and inorganic P exists as Fe

and Al phosphates. Liming the two soils that had constant-charge

colloids neutralized the acidity, increased the base saturation, and

increased the P availability by raising the solubility of the Fe and Al

phosphates. These reactions provided a more favorable plant environment.

The four Oxisol pedons contained predominately constant-potential

colloids in the form of Fe and Al sesquioxides, non-crystalline material,

and organic matter. In this system, the colloidal charge results from

the adsorption of potential-determining H~ and OH- ions. Since these

soils are highly weathered, the soil reaction approaches that of maximum

colloidal stability which is the zero point of charge. These soils,

therefore, have a low surface charge for H+ and Ala+ td satisfy. Thus,

these soils are low in exchangeable acidity even at low pH values. Also,

in this colloidal system, P is tenaciously adsorbed directly to the

colloidal surface by ligand exchange. The availability of adsorbed P

is more dependent on competitive anions than pH. Thus, liming the pedons

containing predominantly constant-potential colloids neither neutralized

appreciable exchangeable acidity nor increased the availability of P.

Instead, liming rapidly increased the surface charge density by adsorbing

OH ions, and this charge was satisfied by adsorbing Ca. The Ca so

adsorbed was retained with greater tenacity than that on the exchange

complex of pedons with predominantly constant-charge colloids. Greater

crop response on the pedons dominated by constant-potential colloids

occurred when competitive anions were added to reduce the P adsorption

energy and increase the P desorption rate than when the acidity was

neutralized with lime.



The humid tropics, in which most of Viet Nam is situated, should

be an ideal climate for agricultural production, Generally, there is

ample solar radiation for efficient photosynthesis, and sufficient

rainfall during some part of the year to assure production of at least

one major cultivated crop. Unfortunately, this combination of warmth

and moisture chemically reacts with the earth's surface to produce

soils that are predominantly severely weathered and strongly acid (69).

Maximum agronomic utilization of the humid tropics will depend on the

development of management techniques that are capable of handling these

adverse soil conditions.

The use of lime to correct strong acidity, as is commonly done in

temperate regions, has led to inconsistent results with soils in the

humid tropics. On many tropical soils, including two used in this study,

liming has been beneficial to crops production (1, 5, 43. 73). However,

on some soils liming has been beneficial to crops only in small incre-

ments well below the lime required to neutralize extractable acidity

(10, 43, 58); and on other soils, liming has adversely affected' crops

(17, 42, 43, 64). Finally, on many of these soils, including one used

in this study, neutral sources of Ca have been more beneficial for crop

growth than lime (571 72. 73). This problem of variations in crop

response to the liming of acid tropical soils has recently been reviewed

by McLean (46) and Kamprath (36).

Mcany acid tropical soils also retain large quantities of P in forms

unavailable to plants (23, 26. 57). This retention does not appear to

follow the pH-dependent Fe-, Al-, Ca-P precipitation-dissolution re-

actions associated with P availability in many temperate region soils.

Instead, it appears to involve specific adsorption to colloid surfaces.

These problems appear to this investigator to concern basic

chemical reactions involving the equilibrium between the soil solutions,

from which most plants must extract nutrients, and the surface of

chemically active soil colloids. Understanding these problems requires

the identification and quantification of the soil colloids that are most

important in these chemical reactions. Then, by relating these colloids

to the type, density and stability of colloidal charge, as well as by

relating these colloids to their capacity to adsorb and desorb various

ions might provide a means of estimating the best chemical treatments

for optimum productivity of acid tropical soils.


Soil Colloids

The colloidal fraction of most mineral soils is composed of large

amounts of phyllosilicate clay minerals, with smaller amounts of crys-

talline metallic oxides and hydroxides, non-crystallin'e inorganic gels,

and organic matter. Not all of these colloids exist as discrete parti-

cles, but some may interact with each other. Generally these inter-

actions involve the smaller non-crystalline particles which form coatings

on the more crystalline parts (74). This interaction can greatly modify

the chemically reactive surfaces of crystalline components until the

dominant reactive surfaces become those of the non-crystalline coatings

and gels.

Phyllosilicate M~inerals

The matrix of most soil colloidal systems is composed of a group of

platelike, crystalline, layered silicate minerals collectively referred

to as phyllosilicates or clay minerals. These minerals so dominate the

colloidal fraction of most soils that the basic concepts of soil

chemistry are essentially the chemistry of phyllosilicates (38). This

has led to the extensive study of these minerals, so that their struc-

ture and charge characteristics are an integral part of most soil texts

(8, 14). Some lengthy texts even deal specifically with the phyllo-

silicates and their relation to soils (28). Two good yet brief reviews

of these minerals are those by Jackson (33) and Zelazny and Calhoun (82),

Metallic Crystalline Oxides and Hiydroxides

In addition to the phyllosilicates, other crystalline components

exist in soil colloids. These minerals have not received as extensive

study as the phyllosilicates, but can be a chemically important portion

of the colloidal system of soils, particularly in the tropics, Generally

speaking, these minerals do not occupy as much volume as the silicates,

but are frequently smaller in size and, thus, have more surface area

per unit mass. This permits a greater influence on soil chemistry than

that indicated by their mass fraction. These crystalline minerals are

the oxides and hydroxides of Si, Fe, Al, and Ti. Those pertinent to

this investigation are quartz, geothite, hematite, gibbsite, and anatase.

Quartz.--This is probably the most common of the oxides (48). Its

formula is Si02 and is composed of tightly packed tetrahedrons with all

corners shared with adjacent tetrahedrons. This kind of packing elimi-

nates cleavage planes (82), making quartz one of the most resistant

minerals in the soil. It usually predominates in the coarse colloidal

fraction thereby resulting in a low surface area and a negligible CEC

in soils.

Geothite and hematite.--These are the most common hydroxide and oxide

of Fe respectively, found in well-drained soils. They are generally

responsible for the reddish-yellow or dark red color found in many soils

(48). Geothite has the chemical formula of aFe00H which implies a

structure where the octahedrons have hexagonal close-packed oxygens

aligned in double rows running parallel to the o axis (24, 41). The 08

radical fits into the hole between the double rows. In electronmicro-

graphs, the mineral appears needle shaped (41), Upon dehydration, the

staggered rows collapse to form aFe203 or hematite (24). In soils this

type of dehydration can occur in warm climates with a severe, well-

defined dry season (48, 66). These minerals and their non-crystalline

counterparts are generally associated with adsorption of large quanti-

ties of P and some pH-induced charges (30, 61).

Gibbsite.--This is the most common of the Al hydroxides. It has the

chemical formula of Al(0H)3 and consists of sheets of hexagonal closely

packed octahedrals, similar to the Al sheet in the phyllosilicate

minerals (24). The hydroxyls protrude above and below the sheets,

which are bound together by weak van-der-Wraals forces. Because of this

weak bonding, gibbsite exhibits perfect cleavage along the c axis (44).

On dehydration gibbsite goes to boeh~mite ~Al(0H)21 followed by Al203,

but these minerals are not as common in soils as gibbsite. Gibbsite is

generally involved in the same type of chemical reactions as goethite

and hematite (30, 61).

Anatase.--This is a Ti oxide with the formula Ti02, It is highly resis-

tant to weathering (32). It does not occur frequently in soils; but,

when it does, it serves as an index of the severity of the weathering

processes. In electronmicrographs, it appears as miniature barrels (35).

Non-Crystalline Materials

In addition to the crystalline materials, the soil frequently

contains colloidal material which is either too small or too disorga-

nized to be identifiable with X-ray diffraction. By definition, this

material is non-crystalline. For the same reason, it is structurally

and chemically very difficult to identify or formulate. Basically,

the inorganic non-cr- stalline material ; consists or' groups of tetrahe-

drals and/or octahedrals that are well-hydrated (4t, 66) but poorly

polymerized. They might be thought of as either dissolution products

of primary minerals or precursors to the more crystalline secondary

materials. Given time, most non-crystalline material will become

organized into secondary minerals,either oxides or phyllosilicates,

depending on chemical composition.

Allophane is not directly involved in this study, but frequently

confused with other inorganic non-crystalline material. It is a non-

crystalline Al-Si material composed of poorly organized Al octahedrals

and Si tetrahedrals characterized by the presence of Al-0-Si bonds (82).

Allophane is commonly found in the colloids of Andepts, which are soils

derived from volcanic ash (82).

A second type of non-crystalline material is organic matter. This

colloidal chemrically-active organic matter is the relatively stable,

ligneous decomposition product formed from the biodegradation of animal

and plant residues. Structurally, it is a disordered conglomerate of

organic rings and chains with various carboxy1, phenol, hydroxyrl, methoxy,

and aminyl functional groups (63). These functional groups are capable

of dissociating to form reactive negative point charges, giving organic

matter a relatively high CEC, whose magnitude depends largely on soil


Colloidal Systems

Not all the various colloidal oonstitutents of the soil occur as

discrete particles. Some tend to interact chemically and physically

with other colloidal particles to form the total soil colloidal system.

In general, the crystalline components with their distinct rigid char-

acteristic geometry do not possess surfaces compatible enough for

extensive chemical or physical interaction. Thus, they generally

remain mutually distinct (82). The non-crystalline material, however,

does not have such rigid geometry and can thus more easily interact

with itself and with the more crystalline materials (82). Some of

these more important interactions between various soil colloids are

discussed below.

Adsorption of non-crystalline Fe-oxides on kaolinite. --Fe-oxides will

be rapidly and specifically adsorbed by kaolinite if, during the process

of soil formation, positively charged colloidal Fe-oxides are formed

(21). The adsorption of Fe-oxides on kaolinite is a result of coulombic

attractive forces in which the oxides act as large, partly polymerized,

multivalent cations that neutralize the negative charges of the kaolinite

(21). The limited polymerization results in an average size of the

adsorbed Fe-oxcide being about 14- wide and 131 51 long (25). This is

too small for X-ray detection (25), but large enough to prevent dis-

placement by high concentrations of non-polymerized cations (21). The

removal of these adsorbed Fe-oxides requires strong oxidizing-reducing-

chelating extractions such as dithonite (21) or oxalate (26).

Since the adsorption of Fe-oxides on kaolinite occurs rapidly (21),

there is little opportunity for discrete crystals of Fe-oxides to form

until the kaolinite becomes saturated and all negative charges are

neutralized. This generally requires about 12j6 Fe203 in the soil (26,

52). Since most soils contain less than 12~ Fe203, most of the Fe-oxtides

found in soils occur as non-atystalline coatings adsorbed on the surface

of kaolinite or other phyllosilicates (52). In soils in which the

Fe209 content exceeds 12q6, polyrmerization of the oxides continues until

discrete crystals of the Fe-oxides or hydroxides are formed (52).

As the Fe209 content of the soil increases towards the saturation

level of 120, the chemical properties of the kaolinite become progres-

sively masked and the chemical properties of the oxide become more

dominant. Thus, a soil containing 8359 knolinite and 152# Fe203 could

reflect the chemical properties of the 150 Fe203 far more than the 85k


Amorphous gels coating crystalline materials.--Highly hydrated gelati-

nous material frequently appears as a film on crystalline surfaces (35).

Electronmicrographs indicate that these non-crystalline gels are not

specifically adsorbed to the surface, but are more likely in the water

film which surrounds most mineral particles (35). The effect of this

material on the chemical activity of the crystalline material is not

fully known, but it does not appear to completely mask the activity of

crystalline material to the same degree that the 120 Fe-oxide adsorbed

on kaolinite does.

Organic completing of metals.--The various functional groups on the

organic complex are capable of adsorbing free Fe and Al ions. W~hen the

organic molecule is small soluble and mobile, as with fulvic acid, this

completing and moving process called chelation can be significant,

Chelation helps move metal ions deeper into the profile, thus aiding in

the dissolution and transportation of these ions (52). If the organio

molecule is too large to be mobile, this completing can still be

significant since it may decrease the ability of microbes to degrade

the organic matter (15. 27). This will result in a build-up of organic

matter. This completing also reduces the potential charge sites, thus

reducing the chemical activity of organic matter (40).

Distribution of Colloidal Systems

Most soils are a combination of the different soil systems dis-

cussed above. The extent to which a given soil's colloids are dominated

by one system or another will depend on the genetic factors that affected

the soil's development. For several genetic reasons, there is distinct

latitudinal distribution of soils dominated by the various systems, The

tendency is for the temperate mineral soils to be dominated by a pure~

phyllosilicate colloidal system. The concentration of metallic oxides

in most temperate soils is relatively low, generally no more than a

mono-molecular layer adsorbed on the surface (19). This is indicated

by the degree to which organic matter is completed by metallic ions.

In the tropics, the phyllosilicates frequently become saturated with

oxides and other non-crystallinle materials which then dominate the

chemical properties of these soils. The organic material is frequently

completed by Fe and/or Al ions which render it of les's value than corre-

sponding amounts in temperate areas (15, 27). This is not to imply

that all tropical soil colloids are dominated by oxides. There are vast

areas of silicate-dominated soils throughout the tropics, but only

limited areas of oxide-dominated soils in subtropical and temperate areas.

Each of these colloidal systems have chemical properties which affect

the soil management techniques required for successful crop production.

Some of these chemical proper-ties, and their differences for the various

systems, are the subject of the next sections.

Colloidal Charge Characteristics

Source of Colloidal Charge

Soil colloids can be divided into two types depending on the source

of electrical charge. In the first type, the charge results from ionic

impurities within the crystalline structure of a valence different from

the ion replaced. The substitution of Al for Si in a tetrahedral, and

Fe, Mg, Ni, etc., for Al in an octahedral result in this type of charge.

Since the charge is internal to the structure of the colloid, it is

stable, constant, and virtually always negative. For this reason, these

colloids are referred to as 'constant-charge colloids.' In soils, the

phyllosilicate minerals are the best example of constant-charge colloids


The second type of charge results from the adsorption of potential-

determining ions on the colloid surface which cause an electrical poten-

tial between the colloid and surrounding electrolyte (78), The ions

which can be adsorbed to produce a potential are H+ and OH The magni-

tude of the induced potential is expressed by the Nernst equation:

-d RT [nrH 1]
F CHo]

in which: 0 = potential in volts.

R = the gas constant,

T = absolute temperature,

F = Faraday's constant,

CH1 = hydrogen ion concentration of the system, and

CHo] = hydrogen ion concentration when # is zero,
Since H ion concentration can be expressed as pH and the R, T, and F

terms are, or can be, constant, the Nernst equation can be rewritten

for 2500 as:

S= .059 (ZPC pH) C21

where ZPC is the H+ ion concentration when 0 = 0. Since the charge of

this type of colloid is based on an electrical potential, the colloid

is referred to as a 'constant-potential colloid' (78). In soils, the

constant-potential colloids include most of the oxides, hydroxides,

non-crystalline material, and any low-charged phyllosilicates where

edge effects outweigh interlayer effect such as kaolinite.

Zero Point of Charge

In order to evaluate constant-potential colloids, it is necessary

to solve the Nernst equation for potential. The solution of the Nernst

equation requires the pH of a reference point at which the net surface

charge and potential are zero, By definition this is 'zero point of

charge' (ZPC).

The ZPC of soil colloidal systems are not constant, but depend

on the type of electrolyte used and the presence of organic matter and

adsorbed ions. Generally, the organic matter and adsorbed P decreases

the ZPC as does a supporting electrolyte containing Ca. A supporting

electrolyte containing S04 will tend to increase the ZPC (38).

Induced Surface Charge

Starting with the ZPC, a potential and surface charge can be induced

on the neutral colloid by the adsorption of either H+ or OH- ions

(Figure 1). If a H ion is absorbed by a colloid such as goethite,

the H ion will be attached to a surface structural OH"- to form a struc-

tural water molecule. This will give the colloid a positive potential

Figure 1. Adsorption of potential determining ions on oxide surfaces to provide either a positive or
negative surface charge.

I ~
0 OH

0\ 0





/ E'+ HO 0

Fe f
HO 0


Fe Fe:
0 05 HO 0

Fe Fe

0 0 8 + OH -- O 0 0

Fe F

0 \OH HO 0

Fe Fe

Net Postive Charge

Net Zero Charge

Net Negative Charge

and a positive charge (38). In soil colloidal systems where the pH

is generally greater than 4.5, a net positive charge of this type is


If, instead of H an OH- ion was adsorbed by the goethite colloid,

the 08i would be attached to the H of a surface structural OH to form

an adsorbed H20 molecule. This would give the colloid a net negative

potential and a negative charge (38). This is the most common type of

induced colloid charge.

Double-Layer Theory

If a colloid is charged, the surface charge must be satisfied by

a set of counter-ions which are equal to, but opposite in sign from the

surface charge. The counter-ions are distributed Utrough an ionic

double-layer which has been mathematically modeled by three theories

(Figure 2).

Hemholts theory,--This is the simplest of the models. According to this

model, the counter-ions act as point charges that are adsorbed to the

surface of the colloid opposite the center of the charge they satisfy

(81). In a soil system, the Hemboltz theory would allow the exchangeable

cations which are counter-ions to be held on the surface of the phyllo-

silicate micelles.

Gauy-Chapman theory.--This is probably the most commonly accepted and

used double-layer theory. According to this approach, the counter-ions

and potential are distributed through a diffuse double-layer in which

concentration of counter-ions decreases while the concentration of co-

ions increases with distance from the surface until both are equal in

bulk solution (78). This double-layer is defined by the equation:

Holmb~oltz Gouy Stern

O ----X O X O- X----I

wit ditac fro a chage coliasre

G =2N~ sinh Si

a b_~~+ee change/cm*,

N d;cmj in bulk: solution

I ntial in volts, and

Z; nter /, ,! sre th-

~' "'.Qtric c

T = absolute tempic ? -e, and

e electron charc~e the cli : ts or potential constants.

The variables C&nbe relatec / mare coml nn, soil parameters. The

surface charge per, when drcc led by sur: e area and converted to

a 100 ba~ me~ rTniES CEC, -lc l jper cm3 crLtw~be directly converted to

normality. Fio(rj potentick1 related In ~I by the Nernst equation,

C1, 2 prevj Riscussed
Not as r~ indicate 'he double er thickn~ess which is

inversely rel t ro the squs, ?root term i i:ation [31 (78). Since

the primary v~ L"ie in this t. s N, the : ilaess of the double-

layer varies 4 :r! averse of !I a conceni t-on. The whole equation

expresses CE, I eLrms of thl root of c~ can! sntration, valence

of the counter-ions, and the sinh of pH.

For constant-charge colloidsethe CEC term is held constant (78).

Th~e ioanic concentration and valence of the counter-ions control the pH

of the system. When the ionic concentration or counter-ion valence

increases, the thickn~ess of the double-layer decreases, as does the

potential. This explains in part why the pH of soil suspensions

decrease when measured in N KC1 as compared to H20.

For constant-potential colloids, the situation is more complex.

N~ow the CEC becomes a function of three independent variables: ionic

concentration, counter-ion valence, and pH (78). VJariation in any one

variable can markedly change the surface charge of the system. This

becomes important in evaluating the acidity and exchange activities

of the system.

This increase in CEC with pH has been examined in ~some detail as

pH-dependent CEC (61). The dependence of induced (pH-dependent) CEC on

ionic concentration has been less extensively studied although some

data is available (6). The theory indicates that for a 100-fold decrease

in ionic strength there would be a 90h0 decrease in pH-induced CEC. This

difference is comparable to the solution used to measure CEC (N) vs the

true soil solution (.01N). Justification for this might be indicated

in the work of Barber and Rowell (6). If the permanent charge of their

soilvere assumed to be near 5,2 and the 100-fold ionic strength differen-

tialware chosen between 0.2 and 0.002N NH40A~c, then the CEC drop from

7.9 to 5.4 would approach the 900 predicted by the theory. This relation-

ship could become important in many Oxisols or Andepts common to the

tropics where constant-potential colloids predominate.

Stern theory.--One of the assumptions of the Gouy-Chapman theory is that

the counter-ions act like point charges that can become infinitely close

to the charged surface (78). Ions, however, do have definite size, thus

they can only get as close as one radius distance from a charged surface.

This discrepancy in the Gouy-Chapman theory can be accounted for in the

Stern theory. This latter theory actually combines the Gouy-Chapman

theory with the Helmholtz theory (78), According to this theory, a one

counter-ion thick molecular condenser of tightly adsorbed counter-ions is

formed next to the colloid surface. This layer resembles the Helmholtz

layer and is referred to as either the Stern layer or the outer

Helmholtz layer. The potential in the Stern layer drops linearly with

distance at a very rapid rate (78). Once across the Stern layer, the

remaining potential is dissipated according to the Gouy-Chapman theory


On constant-charge phyllosilicate minerals in which the charges are

generally low and widely spaced, the Gouy-Chapman theory is probably the

most accurate model. On constant-potential surfaces, however, the

surface charge increases as the sinh of pH,and might become high enough

for the Stern model to be more appropriate. If this happens, many of

the counter-ions could become too tightly adsorbed to be available for


Soil Acidity

Constant charge surfaces.--The traditional concept of soil acidity has

IE or Al3 ions retained as double layer counter-ions satisfying the

permanent negative charge of the phyllosilicate minerals and organic

natter (19, 62, 65). However, the Hi ion rarely occurs in the double

layer of natural soils unless the pHi is below 4 (19), The Al3 ion is

thus the major cause of soil acidity. The acidity of Al3 results from

a three step hydrolysis of trivalent Al"*, This is indicated by the

following equations:

Al" *6H20 +H20 -- Al(0H)a+.5H20 + H0 [

Al(0H)a+*5H20 + H20 -= Al(0H)2+*4H20 + HO(l 5]

Al(0HI)2 *4H20 + H20 ,- Al(0H)3.3H20 + H3Co [6]

The hydrolysis proceeds from equation C4]1 to equation C63 as the pH is

increased. A pH of 5.5 is generally considered the point at which most

monomeric Al and exchangeable H+ have been neutralized. Thus, equation

C43 would be near completion (19). This becomes important in that the
monomeric Al is the form that is most toxic to plants. Equation [6]

does not become completed until the pH reaches 8.5 or more (65). The

final theoretical product of equation [6] is the mineral gibbsite. In

soils, however, time and temperature modify this so that gibbsite can

occur under more acid conditions.

Constant-potential surfaces.--Acidity associated with constant-potential
surfaces has not been studied in as much detail as the acidity of

constant-charge colloids. The acidity associated with constant-potential

surfaces, which occur with greater probability in many tropical soils,

is generally thought to result from the dissociation of a H ion from
structural OH's to create the pH~-dependent charge (19, 62). However,

the Gouy-Chapman theory indicates that pH-induced charges result from

the adsorption of potential-determining OH ions (78). The distinction

between adsorption and dissociation might be more academic than real.

The adsorption of OH- results in an adsorbed H20 molecule that could be

coordinated with the nation in the double-layer and be indistinguishable

from the neutralization of dissociated H+ ion. It is conceivable that

the adsorption energy of the 0H could combine with the hydration energy

of the counter-ion to give the total adsorption energy of the Stern layer.

Organic matter.--A third source of acidity is the dissociation of H+ or
All ions from the functional groups in the organic fraction of the soil.

This release of He or Al"* ions is pH dependent, as is the constant-

potential acidity. However, this does appear to be a true dissociation

of H+ and not an adsorption of an OH- ion. The functional groups

involved are carboxyls, phenols, enols, and alcoholic hydroxyls, of

which the carbooxys are important contributors to soil acidity at low

pH values (19).


When lime (CaCOg) is applied to an acid soil, it reacts with the

soil water and dissolved CO2 according to the following equations (19):

CaC09 + H20 + CO2 -- Ca(HC03)2 C73
Ca(HCO9)2 --o- Caa+ + 2HC0- C8]

H20 + HCq- --,- H2C03 + OH- -- H20 + CO2 Cs
Kf 5+- H20 110]
08i +
Al **6H20 -- Al(0H)af*5H20 + H20

The H+ or Ala+ in equation C10] comes from the soil solution in equili-

brium with the colloid surfaces. The three equations can be summarized

as the single equation (65):

CaCO) + H20 --T Caa" + HCO9" + OH- C113

In this case, the OH- ion reacts to neutralize the acidic components.

In soils of temperate regions that have predominately constant-

charge colloids, liming to a soil pHq between 6 and 6.5 is generally
beneficial to the more commonly grown crops (1, 4. 46. 65). However,

with special crops such as alfalfa and other temperate legumes, higher

pa values are recommended (36). In other cases, such as potatoes, lower

pH values are more optimal. The low pH is to control potato scab rather
than achieve a better nutritional environment (2 When lime is used

to increase pH, the effect on the colloidal surface is to neutralize the

acidity by hydrolysing the Al" and increasing the Ca in the double-

layer; this results in increased base saturation (19).

In contrast, liming Oxisols which generally contain predominantly

constant-potential colloids is seldom beneficial above ph 5.5 (46, 57.

58), This increase is just high enough to hydrolyze the monometric Al,

thus eliminating Al toxicity (58). From a charge theory or acidity

perspective, there is no clear reason for either liming or not liming

above this value. There are no true surface charges for cations such

as H+ or Al"+ to satisfy. Additional lime only causes more OH- ions to

be adsorbed creating additional surface charge which must be satisfied

by adsorbing Ca from the lime into the double-layer. The result is that

both the cation and the anion from the lime are adsorbed by the colloids

without causing any major changes in electrolyte concentrations.

The reasons for not living constant-potential dominated soils above

pH 5.5 are related more to nutritional problems than charge theory. The

biggest problem appears to be P (22. 46, 57). This problem along with

other related anions will be discussed in the next section. Several

micronutrients, most noticably Fe, Mn, and Zn are known to be less avail-

able as the pH is increased (36). Some reports indicate increased Al

uptake immediately after liming (22, 46), which is thought to be related

either to increased electrolyte concentration or formation of Al anions

(22, 46). The increase in conductivity does not occur because of the

adsorption of both the cation and anion from lime. The formation of Al

anions requires at least isolated areas in which the pH must exceed 8,

which could happen only if Ca(0H)2 was the liming material. Such was

the case in the report eited (22),

knion Reactions

Constant-Charge Colloidal Systems

Double-layer exclusion.--When the exposed colloidal matrix has predomi-

nantly constant negative charges, anions will be electrostatically

excluded from the colloidal surfaces and excluded from the ion retaining

Gouy-Chapman double-layer. Therefore, these excluded anions can be

rapidly leached from the solum with sucessive downward moving wetting

fronts. This exclusion and leaching is readily demonstrated by miscible

displacement experiments (18, 39) in which the breakthrough peak of Cl"

and NO3 ions precedes the first pore volume (18) and the distribution

of Cl is concentrated in the lower portions of the wetting front in

which it is transported (39). This is attributed to the inability of

the anion to completely diffuse into the total pore volume because of

exclusion from the double-layer (18, 39).

From the above discussion, it would appear that anion nutrients

such as NO -, SOqa- and P043' could be rapidly depleted from the soil

unless there are other systems in the soil which retain them as is noted

by the continued reported slow movement of P in soils (14), These other

systems include incorporation in organic matter and precipitation as

inorganic minerals.

Incorporation in organic matter.--Mlost forms of living organisms can

utilize certain basic macronutrient anions such as S042-, P04"", and

possible NO3. Thus, plant roots and soil microbes of all sizes will

adsorb anions from the soil solution, utilizing them in their metabolism

or structural development. Later, the anions will be returned to the

soil as excreted by-products of metabolism or mineralizations from

organic residues. Most discussions of soil chemistry limit the treatment

of NO and 504'" in aerobic environments to biochemical reactions

related to the incorporation and release of the anions from organic

natter (8, 9, 14).

Precipitation-dissolution of inorganic minerals.--Some anions, most

noticeably the various forms of P, react with cations dissolved from

the soil minerals and precipitate as slightly soluble P-hydroxides.

Examples of this are the formation of strengite from the dissolution

of Fe in goethite; similarily, the formation of variscite from the

dissolution of Al in gibbsite or kaolinite; and, by a somewhat different

pathway, the formation of calcium phosphates from limestone (37). The

formation of all three of these secondary minerals depends on the solu-

bility products of the original soil minerals and on the concentration

of the OH- ion (37). Since the OH- ion is involved, the whole Fe-, Al ,

Ca-P precipitation process is highly pH dependent. When all systems are

operating there will be a minimum of precipitation, thus a maximum of

available P occurring near pH 7. One of the benefits of liming constant-

charge soils to near neutrality is this increased availability of P

(37, 46).

Constant-Potential Colloidal Systems

The negative charge in a constant-potential colloidal system is

not necessarily present, thus allowing the anions to react directly

with the colloidal surfaces. The anion-colloid surface reactions are

generally of two types: 1) non-specific adsorption, and 2) specific

adsorption (Figure 3).

Non-specific adsorption. --Non-specific adsorption occurs when the pH is

Figure 3. Various types of anion adsorption on oxide surfaces. The dotted line represents the Stern
layer and a equals surface charge.

0 > 0

0 =


0 0 :0

Fe :0

Fe :

0 050: H20


Specific (dissociated)


0 09 :0
O POii
Fe :0


0\ OPOH:O H20

Fe E

Specific (non-dissociated)

I Oi/ :
Fe +:

0 OH : C

Fe :

0 0 : 601

Fe :


below the ZPC of the soil. The colloids have a net positive charge

and the anions are the counter-ions in the double-layer (29, 30, 4~9).

All anions can be non-specifically adsorbed, regardless of their valence

or coordination groups, The adsorption equations of Gouy-Chapman or

Stern, previously discussed, also describes non-specific adsorption

except for changing the sign of 0. In non-specific adsorption, the

anions are retained in the Gouy-Chapman layer, never penetrating inside

the Stern layer.

Specific Adsorption. --Specific adsorption occurs when the anion is

adsorbed by ligand exchange inside the Stern layer (30). Anions so

retained are adsorbed out of proportion to the ionic strength of the

supporting electrolyte (30, 53). For specific adsorption to occur,

there must be both a proton donor and proton receiver (30, 53). The

proton donor can be either an undissociated weak acid such as phosphoric

acid or silicic acid, or a positively-charged colloidal surface (29, 30).

In either case, a water molecule is formed from the surface hydroxcyl and

the donated proton.

If the anion comes from a completely dissociated acid as with

sulfate, specific adsorption will only occur on surfaces with a net

positive charge, i.e. where there is already a proton or adsorbed water

molecule on the colloid surface (30). This does not mean that the total

soil must have a positive charge, just that the soil contains some

surfaces with positive charges. The amount of anions adsorbed cannot

exceed the extent of positive charges, and these, in turn, cannot exceed

the number of non-specifically adsorbed anions (30). Since the lower

the soil pH, the lower the negative charge and the higher the positive

charge; the specific adsorption of SO a- increases with decreasing soil

pH (7).

With weak incompletely dissociated acids, anion adsorption can

occur on surfaces with either a positive or negative charge. For

adsorption to occur on a negative surface, both the dissociated and

undissociated anions must be present. The undissociated acid donates

the proton by dissociating at the surface while the dissociated anion

is adsorbed (29, 30). Since the energy of dissociation is least near

the pKa value of weak acids, the maximum adsorption will occur near

this pK, value. For mono-basic anions, such as silicic or boric acid,

the maximum adsorption on goethite or gibbsite has been experimentally

demonstrated to occur just below the pK, value (29),

For polyprotic acids, such as phosphoric, there are a series of

pi~a values. Between these values, the acid shows various degrees of

dissociation. Thus, there are always both an undissociated species

capable of donating a proton existing as well as dissociated species

ready to be adsorbed (29, 30). In this case, the rate of adsorption

between the various pKa values would be linear and the total curve

would have a series of breaks occurring at the various pKa values (30).

This has been shown for the P system (29). The adsorption curve shows

breaks at 6.4 and 11.6, corresponding to the pits values of 7.2 and

12.7,respectively (53). In the pH range 4. 0 to 6.4 commonly found in

soils, the difference in maximum P adsorbed is small (7). This makes

the availability of P in soils relatively independent of pH, in sharp

contrast to constant-charge colloids but consistent with experiments

performed on Oxisols (57, 70).

Because there is basically a finite amount of anions which can be

adsorbed by a given colloidal system, the different anions available

must compete for the available adsorption positions. The degree of

preference of one anion compared to another will be related to the

affinity of the anions for the colloid surface as indicated by the

valence and the pH of the system relative to the pKa values of the

competing anions. In a phosphate-silica system, the Si does not reduce

the P adsorbed until the pH is above that point at which the adsorption

of the competing anions taken alone would be equal (53). Phosphate,

however, reduces Si adsorption at all pH values (53). In a phosphate-

sulfate system in which one anion is weakly dissociated while the other

is completely dissociated, the competiveness of the 3042- would be

expected to increase as the pH decreases. The amount of 3042- adsorbed

compared to PQ43 absorbed by a soil has been shown to increase with

decreasing pH in non-competitive systems (7), This would be related to

the decrease in negative charge as the pH decreases, providing more

positive areas for S~Q4" to be adsorbed.

Another effect of anion adsorption on soil properties is to

increase the negative charge of the soil by decreasing the pH of the

ZPC (30, 38, 53). This then causes an increase in cation retention of

the system by as much as 0,8 meq per male of P adsorbed (49). The use

of specific adsorbed anion instead of lime as a management practice

to increase CEC has been proposed by some investigators. However this

cation retention is from a constant-potential surface and affected by

both ionic strength and pH of the solution.

1 Uehara, G., L. D. Swindale, and R. C. Jones. Mineralogy and behavior
of tropical soils. University of Hawaii, unpublished manuscript.

Evaluating Surface Adsorption

Surface adsorption is usually examined by using adsorption

isotherms in which different amounts of adsorbate are equilibrated with

an absorber. After equilibration, the unadsorbed adsorbate is deter-

mined and subtracted from the original concentration to determine the

amount of adsorbed adsorbate. In soil science, this type of data can

be interpreted with either a thermodynamic or a practical type of iso-


Langmuir adsorption isotherms. --Langmuir isotherms are a thermodynamic

approach to surface adsorption, which provide a measure of the adsorp-

tion capacity of the adsorbing surface and a relative estimate of the

energy released during adsorption. The theory was originally developed

for the adsorption of a gas on a solid (3), but it has been readily

adapted for the adsorption of dissolved ions on a solid. This was done

by changing the pressure term, P, to concentration, C. The original

gas-solid adsorption equation is most conveniently written in the

linear form:

P/v = 1/bvm + P/vm C123
in which: P = pressure,

v = amount adsorbed per unit of mass,

b = energy term, and

va = amount adsorbed per unit mass at monolayer coverage (3).

This equation is converted to a solution-surface form by substituting

C = concentration of the adsording ion for P to become:

C/v = 1/bvm + C/vm C13]
The first term, C/v, represents the ratio of ions in solution to ions

adsorbed. A plot of this ratio versus equilibrium concentration should

give a straight line, with the slope equal to the inverse of monolayer

coverage and the intercept equal to the inverse product of the monalayer

coverage and the adsorption energy term, b. This energy term is not a

direct caloric value but contains several thermodynamic and systematic

constants, so that:

b = NoT(1/2 RTM)1 exp(Q/RT) C141

in which: Q = energy of adsorption,

R = thermodynamic gas constant,

T = absolute temperature,

N = Avogadro's number,

M = molecular weight,

(J = area of adsorbed molecule, and

7 = adsorption time (3),

With the exception of Q, all these terms will remain constant in any

given system making b exponentially proportional to the adsorption

energy. Since the Langmuir isotherm is theoretical, there are certain

assumptions made. These assumptions are:

1) The energy of adsorption, Q, is constant, independent of

extent of surface coverage and all surfaces are equivalent.

2) The adsorbed molecules are localized on sites and do not

interact with each other (3).

In the heterogenous soil system, these assumptions are not neces-

sarily valid. Soils contain a variety of colloidal minerals with

different types of reactive surfaces which will have different affinities

for adsorption. Furthermore, there is no reason to assume that adsorp-

tion will be complete or that it will be limited to monolayer coverage.

Non-surface adsorbate reactions such as the precipitation of Fe or Al

P-hydroxides can remove adsor~bate in a manner similar to surface adsorp-

tion and produce a linear Langmuir plot (54). However, an aggregate

Langmuir isotherm can be made with soil systems and some thermodynamically

valuable information obtained (31. 54).

Fox sorption isotherms.--The Fox isotherms have a statistically

agronomic derivation to provide direct application to land management

problems (23). W~ith the Fox isotherm, the amount of adsorbate adsorbed

is plotted on a linear scale against the amount remaining in solution

on a log scale. Plant response is statistically correlated with

equilibrium solution concentration of adsorbate. The amount of adsorbate

required to achieve that solution concentration is then determined from

the isotherm graph and becomes the recommended application rate of the

adsorbate. Fox recommends a solution concentration of 0,2 ppm P for

optimum growth of most agronomic crops (23).

Soils Of The Republic Of Viet Nam

The Republic of Viet Nam (RVN) provides a good opportunity to

compare tropical soils having constant-charge colloids with those having

constant-potential colloids. Most of RVN has a monsoon climate with an

excessively wet rainy season and a severe dry season of approximately

equal duration (51), Soils developed in this climate are generally

weathered until nearly void of weatherable minerals, have an acidic re-

action, and contain various proportions of constant-potential and constant-

charge colloids. The predominance of one or the other colloid species

is probably related mostly to parent material, redox conditions, and

landscape stability. The colloid system can be discussed in general

according to the four geophysical regions in RVN. These regions are the

Mekong Delta, Eastern Region, Central Coastal Lowlands, and the Central

Highlands (75).

Mekong Delta

The southern part of RVN comprises the Mekong Delta, Here the

soils are predominantly fine-textured recent alluvium, generally very

acid, with some areas containing acid sulfate soils (50, 77). These

soils are primarily of the constant-charge type, frequently containing

significant amounts of montmorillonite and vermiculite 322 the elay

fraction and small amounts of Fe203 (20, 56).

Eastern Region

North of the Delta proper is an extensive area of older alluvium

from the Dong Nai and Saigon rivers. This area comprises large portions

of Binh Duong, Tay Ninh, Binh Long, Binh Tuy, and Bien Hoa provinces

(77), Soils in this area have undergone post-depositional weathering,

which has resulted in an acid reaction, some clay mobilization, with

laterite occurring between 4 and 5 meters in some areas (77). Surface

texture is frequently sand or sandy loam. The soils were formally

classified as gray podsolic or low humic gley soils~depending on local

relief (50, 77), In the present taxonomic classification system (71),

these soil would be expected to be classified in the Ustult or Aquult

suborder of the Ultisols. These soils are low in Fe203 but high in Al

and Si (50). They would, therefore, be expected to have predominantly

constant-charge colloids.

Central Coastal Lowlands

Along the eastern edge of RVN are numerous small, sometimes

interconnected, alluvial areas where various rivers flow into the South

China Sea. The aggregate of these alluvial areas constitute the

Central Coastal Lowlands. The soils are generally similar to those

described for the Mekong Delta and Eastern Region.

Central Highlands

Northwest of the Eastern Region, the southern arm of the Annam

mountains forms the Central Highland of RVN. Soils in the Central

Highland are very complex due to difference in geologic formation from

which they were derived and the mountainous topography. In the steep-

sloped mountainous areas, medium to fine-textured red-yellow podzolic

soils are the most common (50, 77). These soils generally have an

acid reaction with distinct A, B, C horizons as well as an accumulation

of organic matter in the A horizon (77). In the present taxonomic

classification these soils might be Humults. Typical of these would

be the granite-derived soils around Dalat. The soils contain some

kaolinite and iron oxides (76),but not enough oxides to saturate the

colloid system. Soils like these would be expected to be a mixture of

constant-charge and constant-potential colloids.

Among the mountains are several level plateaus in which the soils

are deeply weathered, reddish-brown latosols derived from basalt (50,

77). These soils are generally high in clay but low in exchangeable

bases and pH. The clay minerals have been identified as knolinite

(50), but the 20j Fe203 would indicate that the oxide saturation level

has been exceeded. These soils would be expected to be in the Oxisol

order of the taxonomric classification system and have predominantly

constant-potential colloids.


Site Selection

Comparing the chemistry of acid soils with constant-charge

colloids to those with constant-potential colloids requires locating

areas in which the soils have varying amounts of the two types of

colloids. This can be estimated by monitoring the percent of free Fe203

contained in the soils of various areas. Thus, after examining the

general soils map of RVN prepared by Moormann (50) and discussing

available facilities with Drs. Nguyen than Hai, Director, National

Agriculture Institute (NAI) and Thai cong Tung, Director, Agricultural

Research Institute (ARI), three potential locations were selected.

These were: 1) the NAI's demonstration farm at Thu Due, 2) the ARI's

Dalat Agriculture Research Center just outside Dalat City, and 3) the

ARI's Eakmnat Agriculture Research Station near Ban Me Thuot.

Location of Soils

The Thu Due farm is located on an old alluvium terrace of the

Dong Nai river about 10 km north of Saigon. There are two main soil

series found on the farm. The Trang Bang series is found at the lower

elevations and the Thu Due series at higher elevations. Both series

have sandy surface horizons with clay increasing with depth. These

soils are similar to the low humic gley and gray podsolic soils dis-

cussed in the previous section. The colloids of both series have only

small quantities of free Fe203 and should be predominantly constant-

charge colloids.

The Dalat Research Center is on some of the higher mountains of the

Central Highlands some 300 kmn north of Saigon. The residual soils on

the Center are fine-textured red-yellow podsolic soils derived from

granite. The colloids of these soils contain significant amounts of

free Fe203 (76), but not sufficient to saturate the colloids. Therefore,

these soils should contain a mixture of both types of soil colloids.

The Eakmat Station is located on a mid-altitude basaltic plateau

in the Central Highlands 5 kmn east of Ban Me Thuot and 400 kmn north of

Saigon. The soils of the entire plateau are deep, well-drained fine-

textured reddish-brown latosols. The soil colloids should be Fe-oxide

saturated kaolinite with possibly some crystalline goethite or hemitite.

This would make the colloids predominantly constant-potential.

Field Trial Sites

A preliminary visit to each location was made to select two field

trial sites per location. At Thu Due, one site was selected on each

soil series. At Dalat, both sites were on the residual soils. One site

was on the natural 5-8;9 slope. The second on an area that had recently

been terraced according to the custom for vegetable production in this

area. At Eakmat, one site was in the main research area, the second was

in a field near the main entrance to the station about 1 km from the

first site.

Climatic Variables

Although this selection of locations potentially provides a good

group of soils to compare different types of colloids, the sites were

dispersed over a distance of 400 lau and 1500 m range in altitude. This

makes a significant difference in the local climate of each location

(Table 1). The mean temperature is 200 cooler in Ban Me Thuot and 6oC

cooler in Dalat than it is in Thu Due or Saigon (51), The distribution

of precipitation is also different for these locations,but not to the

same extent as the temperature (51). These climatic differences affect

the absolute value of any field trials, precluding any valid comparison

between sites. This forces the major comparisons on field trials to be

between treatments within a location and only relative treatment effects

can be compared between locations.

Soil Characterization

Profile Descriptions

Adjacent to each field trial site, a profile pit was dug. The soil

profile was described according to the Soil Survey Manual (70) and

sampled by horizons to a depth of 140 em. The exception to this sampling

procedure was the B2 horizons of the Eakmat profiles. In these profiles,

the B2 horizons were too thick and had to be sub-divided into three

samples. All samples were prepared for physical and chemical analyses

by air drying and grinding to pass a 10-mesh (2 mm) sieve.

Physical Analyses

Except for the Dalat profiles, particle size distribution of the

soil samples was measured with a Bouyoucos Hydrometer (12) following

soil dispersion with calgon. In the Dalat profiles, dispersiag with

calgon was not effective. In these samples, the clay content was

estimated by multiplying 15-bar soil moisture by 2.5 (71), the sand

fraction was determined by wet sieving through a 300-mesh sieve and the

Ban M~e Thuot (460)
Mean Mean
temp. precip.
OC mm

21.67 2.54e


26.11 25.40

26.11 121.92

25.56 261.62

25,00 271.78

24.44 302.26

25.00 261.30

24.44 340. 36

23.89 223.52

22,78 86,36

21.11 22.86

24. 00 1899. 92

Thu Due (12)** Dalat (1500)
Month Mean Mean Mean Mean
temp. precip, temp. precip.

* Adapted from Nuttonson (51).
** Numbers in parenthesis indicates altitude in meters.

Table 1. Climatological data for field trial locations













1 .22























269. 24








27 94


180. 34



254. 00



















Mean annual.

silt fraction by difference. Water-dispersable clay was determined on

the Dalat and Eakmat profiles after shaking a water-soil mixture over-

night (68). Organic carbon was measured by wet digestion with

K2Cr407 H2S04 (68).

Colloidal Analyses

Sample preparation.--Inorganic colloids were physically separated from

the silt and sand fraction by the general procedure of Jackson (32).

This included removing the multivalent cations and buffering the soil

by washing with Na0Ae adjusted to pH 5. 0. Organic matter was removed

by wet oxidation with H202 and the free Fe-oxides with Dithonite-

Citrate-Bicarbonate (DCB). After these cementing agents were removed,

the colloids were dispersed by raising the pH to 10 with Na2CO j. The

sand was separated by wet sieving through a 300-mesh sieve, and the clay

was separated by repeated centrifuging at 1000 rpm for 2 minutes in an

International No, 2 centrifuge. The residue was attributed to the silt


X-ray diffraction analysis,--X-ray diffraction analysis criteria were

used to identify the crystalline components of the inorganic colloids

separates. This was performed by plating duplicate 225-mg samples of

the appropriate clay suspension on unglazed ceramic tiles according to

the procedure of Rich (59). One sample was saturated with Mg and solvated

with glycerol, while the other sample was saturated with K. Samples so

prepared were analyzed with a General Electrio XRD 700 X-ray diffracto-

meter using Cu Ka radiation and scanning from 20 to 300 28. After dif-

fractographs were obtained for each sample, the K-saturated samples were

heated to 550oC for 4 hours and resoanned.

Thermal analysis.--Thermal properties of the inorganic colloid separates

were examined by heating approximately 2.5 mg of Mg-saturated oven-dried

samples in a Perkin Elmer S-2 Differential Scanning Calorimeter (DSC).

The samples were heated at 40oC per minute from 100 to 720oC. This

analysis provided dehydration endotherms for gibbsite and kaolinite at

280o and 53000, respectively. These endotherms are approximately 40oC

below the temperature they occur with DTA analysis (32). This difference

is due to the different methods of heating, heat sensing, and purge gas

used with the DSC compared to DTA.

Wet chemical analysis.--The Alexiades and Jackson (4) method was used

to estimate the amounts of non-crystalline material and vermiculite in

the colloid separates. With this technique, non-crystalline material

was measured by selective dissolution with hot Na0H,and vermiculite was

estimated from the difference in CEC determined by exchanging Ca with Mg,

compared to exchanging K with NHq after heating to 550oc. The Si, Al,

and Fe dissolved during selective dissolution and Ca exchanged from the

differential CEC were determined by atomic adsorption spectroscopy. The

K exchanged from the differential CEC was determined by flame emission


Fe-oxide analysis.--Since Fe-oxides contribute a significant portion of

the constant-potential surfaces in many tropical soils, more detailed

analysis was needed than provided by the DCB procedure alone. Thus the

method developed by McKeague et al. (45) was used. This procedure

involved separate extraction of the total soil with Na2 207 to extract

organic Fe; acid NHq-oxalate (Talmn's reagent) to extract the weakly

amorphous Fe; and DCB to extract total Fe. The latter, however, does

not dissolve hematite. The Fe and Al in all extracts were analyzed by

atomic adsorption spectroscopy. The DCB extracts were oxidized with

Eg02 prior to elemental analysis.

Quantitative analysis.--The total soil colloids include not only the

clay separates but also, most of the organic matter and Fe-oxides. Thus,

an estimation of the quantity of each component in the total colloidal

fraction was made using all analytical techniques discussed above. This

involved assuming the hydrometer accurately measured all the colloidal

fraction. The organic matter and Fe-oxides were normalized over this

total colloidal content and the clay separate normalized over the

remainder. The composition of the separates was estimated by using the

DSC analysis for gibbsite and knolinite, selective dissolution for non-

orystalline material, and X-ray for the intergrade phyllosilicates. The

gibbsite determined by DSC was subtracted from the selective dissolution

estimate of non-crystalline material. The kaolinite determined by DSC

was used as an internal standard in the X-ray patterns to compare relative

intensity of the vermiculite-chlorite diffractions. This relative

intensity was estimated from the area under the respective diffractions.

Finally, the differential CEC data was used to estimate what fraction of

the intergrade material was vermiculite.

Chemical Analyses

Soil acidity.--Soil reaction was determined by measuring the pH of a

1:1 soil-solution suspension using solutions of distilled H20, N KC1,

0.01M CaC12, and N K(2SC~c. The first three measurements are standard

characterization pH measurements (68). The fourth is a method used by

Mekaru and Uehara (49) to indicate soils with specific anion adsorption

potential. Differences between salt-pHI and H20-pH were calculated as

a pH values in which the H20-pH was subtracted from the salt-pH. All

measurements were made using a Fisher Accumet Model 220 pH meter with

a Sargent combination electrode No. 5-30072-15.

SExtractable acidity was determined by equilibrating the soil with

excess BaC12-Triethanolamine (TEA) adjusted to pH 8.2 and back titrating

the unreacted TEA with standard HC1 (68). Exchangeable acidity was

determined by overnight equilibration of 10 g soil with 100 ml NJ KC1.

A 50-m1 aliquot of extract was titrated to pH 8. 0 with 0.iN Ba(0H)2 to

determine the exchangeable acidity. The remainder was analyzed for Al

by atomic adsorption spectroscopy to determine the KC1 extractable Al.

Potentiometric titrations of soil-N KC1 suspensions with 0.1N Ba(0HI)2 '

using a modified Sargent Model D recording titrator and a Sargent combina-

tion electrode No. 3-30072-15,were made to determine the source of

extractable acidity according to the techniques of Rich (60) and Zelazny

and Fiskell (BS). The titrations were made in a C02-free environment

with titer values being determined at the low inflection near pH 5.0,

if present, and at pH 8.0.

SExchange properties,--Soil CEC was determined by saturating with N NH 0Ae

adjusted to pH 7.0, washing with C2H50H, exchanging the retained NHIC

with N NaC1, and determining the NHq released by distillation into boric

acid and back titrating with standard HC1 (68), Exchangeable bases

(Ca, Mg, K, and Na) were measured in the NH 0A0 extract from the CEC
determination (68), Calcium and Mg were measured by atomic adsorption

spectroscopy, while K and Na were measured by flame emission spectroscopy.

The sum of cations was calculated as the sum of extractable bases plus

the BaC12-TEA extractable acidity (68). Percent base saturation was

calculated by dividing the extractable bases times 100 by the sum of

cations (68). The CEC per 100 g clay was calculated by dividing the

CEC measured with pH 7,0 NH40Ae by percent olay (68),

SZero polint of charge.--The Ap and one B2 horizon from each of the four

major soil profiles was selected for zero point of charge (ZPC) analysis.

This analysis is based on the premise that at the ZPC, ionic strength

will have no affect on potential, i.e. pH. The basic procedure used was

that on; van Raij and Peech (79) except the soil pretreatment was omsitted

This procedure involves equilibrating several 4-g soil samples for 3

days in four concentrations of NaC1 (1.0, 0.1, 0.01, 0.0011) to which

various aliquots of Na0H or HC1 had been added. After equilibration,

the samples were centrifuged, the pH in the solution determined and

compared to the pH of a similar solution without the soil, From this

difference in pH, the meg of H~ or OHI adsorbed by the soil was detenrmned

and plotted against pH for each salt concentration. The common inter-

section of the four graphs was taken as the ZPC of the soil.

Phosphorus Analysis.--The amount of inorganic P in the soils was analyzed

by the Peterson and Corey modification of the Chang and Jackson P-frac-

tionation procedure (55). This is a sequential extraction using NHI4C1,

NH4F adjusted to pH 8.2, Na0H, DCB, and H2SO to extract water soluble,
Al, Fe, reductant soluble, and Ca forms of P. The organic-P was estimated

from the difference between P extracted with Na0H followed by H2504 from

samples ignited to 60000 for four hours compared with non-ign~ited samples,

The total P was calculated by adding all the P fractions. Available P

was estimated by extracting 2 g of soil for 5 minutes with 20 ml of

strong Bray extracting solution. The P in this and all subsequent P

experiments was determined colorimetrically using ammonium molyodate

with ascorbic acid as a reducing agent according to the procedures

developed by Watanabe and 01sen (80). The basic procedure was modified

to remove interference from various extracting reagents by adding boric

acid to F extracts, lowering the pH of Na0H extracts, and maintaining

the 01 concentrations below 1700 ppm. All P determinations were made

with a Bausch & Lomb Spectronie 20 spectrophotometer at a wavelength of

880 Cut

Field Trials

Calcium Treatments

During the preliminary site selection visit, soil samples were

collected from each potential field trial site at each facility. The

pH of these soil samples was determined in a Shoemaker, McLean, and Pratt

(SMP) buffer solution (67) in addition to a 1:1 soil-solution ratio with

H20 and N KC1. The SMP buffer is a simple means of rapidly approximating

the buffer capacity in terms of lime needed to raise the soil pH to

either 6.0, 6.4, or 6.8. The SMP buffer data was used to determine four

lime rates on the Dalat and Eakmnat sites. These rates were the equivalent

of 1/4, 1/2, 3/4, and 1 times the SMP requirement to raise the pH to 6.0,

With the Trang Bang and Thu Due sites, the buffer capacity was too low

to be measured with the SMP buffer. On these sites, the four lime rates

selected were equivalent to 1, 2, 3, and 4 T/ha CaCO Since the most

available liming material was Ca(0H)2, the rates determined above were

multiple by 0.74 to get the actual rate of liming material applied to

the field plots (Table 2). In addition to the lime treatments, a CaS04

treatment was used in an attempt to distinguish orop response to pH from

Table 2. Rates of Ca used in the field trials

Ca(0OH)2 Treatments Gypsum*
Soil site L1 L2 L3 L4 G2
----------------- T/ha ------------------

Trang Bang 0. 73 1.5 2.2 2,9 3.2
Thu Due 0. 73 1.5 2.2 2.9 3 .2

Dalat N 2.00O 4.0o 6. 0 B.0D 8. 6
Dalat T 2.00O 4.0O 6.0O 8.0O 8,6

Eakmat I 2. 00 4.0O 6. 0 8.0O 8. 6
Eakmnat II 2. 00 4. 0 6.0O 8. 0 8.6

* Ca in G2 = Ca in L2,

crop response to Ca fertility. The CaS0q treatment was equal in Ca to

the 2nd lime treatment; i.e. 1/2 buffer requirement or 2 T/ha. However,

since Ca(0H)2 is considerably more soluble than CaS~q and the Ca(0H)2

was ground into smaller particles than the Cas04; at equal rates more

Ca from Ca(0H)2 would be reacting with the soil than would Ca from CaS04.

A control treatment of no Ca made a total of six treatments.

Plot Design

In each field trial, the six treatments were replicated three times

to make a total of 18 plots. The plots were laid out in a randomized

complete block design with each field trial being individually randomized.

The plots were 3 x 10 m for a total of 30 ma or 0.009 ha/plot. A complete

field trial was 18 x 30 n4

Each 3 x 10 plot was split into two 3 x 5 subplots with one subplot

planted to okra (Hibiscus esculentus) and the other subplot planted to

sorghum (Sorghum valgare). These 2 plants were selected as indicators

of the response of dicot and monooot plants to the various treatments.

The 2 crops were planted in rows 0.5 m apart with 0.5 m between hills of

okr~a and 0.1 m between sorghum plants. Five seeds were placed in each

okra hill and 2seeds were planted in each sorghum hill. This planting

rate was based on germination trials in which the okra tested 45k viable

and the sorghum 80k viable. After thinning to 1 plant per hill this

allowed for 10 rows of okra with a total of 60 plants and 9 rows of

sorghum with a total of 270 plants. The space for the 10th row of sorghum

was used as a between block drainage ditch.


A uniform fertilizer application was applied to all plots knoluding

the control. The rate of fertilization of macro-nutrients was based

on previous sorghun fertilizer trials conducted in part at the Thu Due

location (13). The result of this fertilizer trial indicated that a

balanced 100-100-100 kg N-P205-K20/ha was optimum. This rate was obtained

by adding 5.4 kg of K2 94 to 33.75 kg of 16-16-8 mixed fertilizer for

each field trial. An additional 9.00 kg of MgSO4 was added to this mix

because of the observed Mg deficiency at each location. This made a

total of 48.15 kg of fertilizer per field trial with a final nutrient

content of 100-10)0-100-50-400 kg N-P205-Kgo-Mgo-s (100-44-83-30-40 kcg

N-P-K-Mg-S)/ha. The 40 kg/ha of S accounts for only that which came

from the K2SOS and MgS04 and does not account for S included in the

16-16-8 mixed fertilizer. This 40+ kg/ha of 3 was considered sufficient

to satisfy any S nutrient requirements of either crop. Thus no fertility

response to the 304 in the CaS04 treatment was anticipated and no S was

applied to the other plots to compensate for the S in the CaS04 treated


The 48.15 kg of fertilizer for each field trial was divided into

three nearly equal lots. The first was broadcast just prior to planting,

the second was banded 3 weeks after planting, and the third was banded

7 weeks after planting.

Plot Operation

The operation of the six field trials was a cooperative effort

between the University of Florida and the NAI or ARI. After the plots

were laid out and marked, the lime or gypsum treatments were broadcast

by hand as uniformly as possible. The Ca was incorporated into the top

15 cm of soil With a hoe. The plots were than individually raked

smooth and the first increment of fertilizer was broadcast and lightly

raked in. Drainage furrows were then placed around each plot and the

entire field trial. Finally, the plots were seeded across the entire

six treatments.

The cooperating institution managed the plots between visits of the

principle investigator, taking responsibility for weeding and applying

pesticides as necessary. Three weeks and 7 weeks after planting, the

second and third increments of fertilizer were applied, respectively,

and all crop plants from each subplot were harvested at maturity.

Soil samples were collected from each block prior to the Ca appli-

cations, from each plot at planting, just prior to the second increment

of fertilizer, and at the conclusion of the trial.

Since the six trials were separated by several hundred km and

required considerable logistic effort to move personnel and material

from one location to the next, all operations for the different trials

had to be done at different times. The dates of each operation at each

field trial is shown in Table 3.

Calcium and Phosphorus Studies

Calcium Treatments

The ability of the different soils to retain Ca from the different

sources and the affect of the different Ca sources on P adsorption and

availability were examined by treating subsamples from the Ap horizons

of the Trang Bang, Thu Due, Dalat N, and Eakmnat II profiles with either

Ca(0H)2 or CaS04. The soils were treated with reagent grade chemicals

at rates that approximated those used in the field trial plus two

additional gypsum treatments. The extra gypsum treatments were made

at rates corresponding to preliminary P adsorption estimates of the

2nd fertilizer 3rd fertilizer
Site Treated Planted increment increment Harvested

Trang Bang Aug. 12, 13 Aug. 13 Sept. 6 Oct. 3 Nov. 30

Thu Due Aug. 6, 7 Aug. 9, 10 Sept. 3 Oct. 2 Dec. 6

Dalat N Sept. 8, 9 Sept. 9, 10 Oct. 6 Nov. 10 Dec. 22

Dalat T Sept. 11 Sept. 11 Oct. 6 Nov. 10 Dec. 22

Eakmnat I Aug. 20 Aug. 21 Sept. 17 Oct, 22 Dec. 9

Eakmnat II Aug. 23 Aug. 24 Sept. 17 Oct. 22 Dec. 10

Table 3. Dates of each operation for individual field trials

amount of anion adsorption taking place in the various soils. The

Ca(0H)2 was applied to 500 g samples while the CaS0q was applied to

250 g samples. An additional untreated 500 g sample was also prepared.

The eight samples were labled C, L1, L2, L3, Lk, G1, G2, and G3

depending on source and amount of Ca applied. The actual applications

are shown in Table 4. All samples were air-dried and ground to pass

a 10-mesh sieve prior to treatment. After treating, the samples were

throughly mixed and moistened to a relatively uniform moisture content

(Table 4). The moistened samples were placed in plastic bags and

equilibrated for 30 days at ambient temperature, air-dried, remoistened,

and equilibrated for an additional 30 days before being air-dried and

ground to pass a 10-mesh sieve.

Calcium Retention Experiment

Leaching operation.--Duplicate 75-g subsamples of the Ca(0H)2 and CaSC4

treated and untreated samples were placed in Nalgene disposable micropore

filters. An initial application of 50 ml of distilled water, followed

each day by 40 ml additional distilled water, was gently poured on the

surface of the soil and allowed to gravimetrically percolate through the

soil into the collection cup. If some water remained ponded on the soil

after 24 hours, gentle suction was applied to the bottom of the filters

with rubber suction bulbs. If this was not sufficient to remove the

Ponded water, then a vacuum pump was attached to the filter units.

Each day the filtrate was collected and analyzed for conductivity, pHl,

Ca, and K.

Analytical procedures.--Alfter 14 days, the leaching was terminated and

the samples air-dried. The soil pH was determined in a 1:1 soil-solution

Table 4. Ca treatments used in laboratory Ca and P retention studies

Treatments Moisture during
C L1 L2 L3 Gk G1 G2 G7 euilibration
g Ca(0H)2/500 g soil g caso4/250 g soil

Trang Bang;
0 0, 20 0,40 0. 60 0. 80 0. 25 0. 50 1.0o 15
Thu Due
0 0. 20 0. 40 0. 60 0.80o 0.25 0. 50 1.0 10t

0 0.60 1,20 1,80 2.40 3.16 6.26 12.5 30
0 0.60 1.20 1.80 2.40 3.16 6.26 12.5 30

ratio using both distilled water and N KC1. The Ca retained by the soil

was extracted with separate 1:10 soil-extracts using N KC1 and 0.1N_ HC1.

The N KC1 was expected to extract exchangeable Ca retained in the

diffuse Gouy-Chapman double layer. The 0. 1N HC1 was expected to reduce

the surface potential, and extract both the exchangeable Ca from the

diffuse Gouy-Chapman double-layer, and adsorbed Ca from the Stern layer.

The N KC1 extraction was also analyzed for Al to evaluate possible

effects which the Ca treatments had on extractable Al.

Instru!en~taion.--The conductivity was determined on a Universal Interloo

Inc. Model 751 conductivity bridge equipped with an automatic temperature

compensating electrode having a cell constant of 1. The pH was measured

with an Orion Research Model 701 digital pH meter and a Sargent 3 30072-15

combination electrode. The concentration of K was determined with a

Beckman D flame spectrophotometer and Ca with a Perkin-Elmer 700 Atomic

Adsorption spectophotometer.

Phosphorus Adsorption Study

P-adsorption-desorption isotherms.--The capacity of the four principle

soils (Trang Bang, Thu Due, Dalat N, and Eakmat II) to adsorb P and the

influence of Cl- and SOqa- on the desorption of the adsorbed P was

examined by making adsorption-desorption isotherms on these four profiles.

The isotherms were made following the general procedure of Fox1. This

procedure involved equilibrating, for a period of 9 days, quadruple

5-g subsamples of each of the above soils in 50 ml of 0. 01N cacl2 with

six increasing concentrations of P from Ca(H2Pq))2 (Table 5). During

i Fox, R. L. Laboratory procedure, University of Hawaii. (Ditto).

Table 5. Original concentration of P in adsorption isotherm

Horizon Concentration of P

Trang Bang

Thu Due

60 a






the equilibrating period, the samples were reciprocally shaken twice

a day. After 9 days, the samples were centrituged and the P remaining

in solution analyzed. The excess solution was discarded, and duplicate

samples were refilled with 0.01N CaC12 or CaS04 and again equilibrated

for 9 days. The samples were again centrifuged and the P in solution

analyzed. The amount of P desorbed with CaC12 was compared to the

amount desorbed with CaS04. The results were evaluated using both the

Langmuir and Fox techniques.

Effect of Ca amendments.--In addition to the above study, the effect of

Ca(0H)2 and CaS04 applied to the soil on P adsorption was evaluated by

making similar adsorption isotherms on the samples used in the leaching

study of the previous section. The same solution concentrations and

equilibration time were employed, except desorption was not preformed.

These isotherms were evaluated with Langmnir techniques only.

To complete the study, the relative availability of natural P was

studied with an ion exchange extraction of the Ca-treated samples.

This involved grinding duplicate 5-g soil samples to pass through a

60-mesh sieve and adding 2 g of wet amberlite that was retained on a

60-mesh sieve and equilibrating the mixture overnight with 50 ml of

distilled H20. The P extracted from the soil and adsorbed on the amber-

lite was determined by wet sieving to recover the amberlite, then

expelling the P by leaching with 20 ml of 0.5N_ Na0H followed by 20 mi

0.5N; HC1. This technique was selected to prevent the interference with

the pHI effect of the Ca treatments which would have resulted from using

either strongly acidic or alkaline extractions.


Soil Characterization and Classification

Profile Descriptions

The profile descriptions (Table 6) indicated that the soils chosen

at each site were indicative of those previously identified as predomi-

nate at each location (50, 75. 77). The two pedons developed on the

alluvial terrace at Thu Due both have well-developed argillic horizons

underlying fine sandy loam or loamy fine sand surface horizons. The

Trang Bang pedon has a dark umbric epipedon that rapidly gives way to

grayish-brown well-mottled subsoil. These conditions indicate poor

drainage and a high watertable during at least some part of the year.

The poor drainage is the result of low relief in the area (11).

The Thu Due pedon has an ochric epipedon below which bright colors

persist through the entire profile except for a few fine mottles occurring
in the lowest horizon examined. These conditions indicate a better

drained profile resulting from the extra meter of elevation.

The pedon near Dalat developed from granite and has a deep dark

umbric epipedon in the natural phase. Below this the colors are a bright

yellowish-red which change to a brownish-yellow with mottles beginning

at approximately 88 om. At 111 em, there is a textural discontinuity

resulting in an increase in percent sand. This profile appears well-

drained in the upper half, but has impeded drainage at lower horizons

resulting both from compacted subsoil and the textural discontinuity.

Table 6. Profile description of the six pedons examined

TRANG BANG sandy loam
(Aeric Paleaquult fine loamy, isohyperthermio)

Location: Gia Dinh Province, Republic of Viet Nam; National Agricultural
Center Demonstration farm 15 m east of western property fence,
30 m southeast of water main access manhole and 70 m south of small
Vegetation and land use: Native pasture, with dikes for paddy.
Slope and land form: Nearly level old alluvial terrace.
Drainage and permeability: Poorly drained, slow runoff, moderately slow
permeability; some puddling lasts for up to one day following heavy
rain; water table rises to near the surface during the rainy season.
Parent material: Alluvium from the Dong Nai River.
Date sampled: January 17, 1972.

Horizon Degg9 Description
Ap 0-18 Dark grayish-brown (10YR 4/2) fine sandy loam;
moderate median subangular blocky structure;
friable when moist; strongly acid; abrupt smooth

Ai2 18-26 Dark brown (7.5YR 4/4) sandy clay loam; moderate
small to medium subangular blocky structure;
friable when moist; very strongly acid; alear
smooth boundary.

B2itg 26-48 Grayish-brown (10YR 5/2) sandy clay loam with
common medium distinct pale brown (10YR 6/3)
mottles and red (2.5YR 4/6) pore linings;
moderate medium angular blocky structure; friable
when moist; very strongly acid; clear way

BZ2tg 48-76 Grayrish-brown (10YR 5/2) sandy olay with many
small and medium distinct yellowish-red (5YR
5/6) mottles: moderate medium subangular blooky
structure; friable when moist; very strongly
acid; gradual smooth boundary.

B23tg 76-140t Pale brown (10YR 6/3) sandy clay with many
medium distinct strong brown (7.5YR 5/8)
mottles; moderate medium subangular blocky
structure; friable when moist; very strongly

Table 6, Continued

THU DUC loamy fine sand
(Typic Paleustult fine loamy, isohyperthermic)

Location: Gia Dinh Province, Republic of Viet Nam; National Agricultural
Center Demonstration farm, 100 m west of the new farm complex and
10 m east of ironstone grave site.
Vegetation and land use: Fallow with sparse sod; sometimes used for
student plots.
Slope and land form: Nearly level ( < 2k slope) on old alluvial terrace,
Drainag~e and permeability: Well-drained with rapid runoff and moderately
rapid permeability; water table never higher than bottom of sampled
Parent material: Alluvium from the Dong Nai river.
Date sampled: January 17, 1972.

Horizon Depth Description
Ap 0-20 Yellowish-brown (10YR 5/4) loam~y fine sand;
weak to moderate, fine to medium granular
structure; very friable when moist; very strongly
acid; scattered ironstone concretions; smooth
abrupt boundary.

B21t 20-50 Yellowish-red (5YR 5/6) sandy loam; moderate
fine blocky structure; friable when moist;
extremely acid; clear wavy boundary.

B22t 50-90 Yellowish-red (5YR 4/6) sandy loam; moderate
medium, blocky structure; firm when moist;
extremely acid; gradual smooth boundary,

B23t 90-140t Yellowish-red (5YR 5/6) sandy loam with few
fine distinct light gray (10YR 7/2) mottles;
moderate medium blocky structure; firm when
moist; extremely acid.

Table 6. Continued

DALAT clay (Natural phase)
(Aquic Haplohumox clayey, mixed, isothermic)

Location: Tuyen Due Province, Republic of Viet Nam; Dalat Agricultural
Research Station, 30 m upslope from a fence and 15 m west of
terrace field adjacent to the station.
Vegetation and land use: Recently plowed sparse sod, ordinarily used
for vegetables.
Slope and land form: Gentle southern slope (about 5%') in rolling terrain.
Drainage and permeability: Well-drained upper profile with impeded
drainage beneath; rapid surface runoff; moderate permeability;
intermittent perched water table within sampled profile during
rainy season.
Parent material: Granite weathered in-situ.
Date sailed: December 21, 1971.

Horizon Depth Description
Ap 0-18 Dark brown (7.5YR 3/2) clay; weak to moderate
and fine to medium crumb structure; friable
when moist; very strongly acid; alear smooth

A12 18-37 Yellowish-red (5YR 4/6) clay; weak to moderate,
fine to medium granular structure; firm when
moist; very strongly acid; scattered ironstone
concretions; 01ear wavy boundary.

B1 37-42 Yellowish-red (5YR 5/8) clay; moderate medium
subangular blocky structure; firm when moist;
very strongly acid; clear wavy boundary.

B21 42-65 Strong brown (7.5YR 5/6) 01ay; moderate to
strong medium angular blocky structure; firm
when moist; strongly acid; scattered ironstone
concretions; gradual smooth boundary.

B22 65-88 Brownish-yellow (toYR 6/6) and red (2.5YR 5/8)
clay; strong medium angular blocky structure;
very firm when moist; medium acid; gradual
smooth boundary.

B23 88-111 Light yellowish-brown (10YR 6/4) claywith many
coarse distinct yellowish-red (5YR 5/)mtls
strong medium angular blocky structure; very
firm when moist; slightly acid; clear smooth

C1 111-140+ Light yellowish-brown (2.5Y 6/4) sandy alay
with common to many coarse prominent red (7.5R
4/8) plinthic mottles; strong coarse subangular
blocky structures firm when moist, medium acid.

Table 6. Continued

DAL~AT clay (Terraced phase)
(Aquic Haplohumox clayey, mixed, isothermic)

Location: Tuyen Due Province, Republic of Viet Nam; Dalat Agricultural
Research Station, 30 m east of terrace bank and 15 m south of
corner of reactor property boundary.
Vegetation and land use: Fallow but being readied for vegetable
variety trials.
Slope and land form: Level to slightly sloping ( < 2;9 ), 1 year-old
terrace on normally rolling mountainous terrain.
Drainage and permeability: Moderately well-drained; slow runoff with
moderate permeability, possible intermittent perched water table
during rainy season.
Parent material: Granite weathered in-situ.
Date sampled: December 21, 1971.

Horizon Depth Description
Ap 0-20 Strong brown (7.5YR 5/8) and yellowish-red
(5YR 4/6) clay; strong medium subangular blocky
structure; very firm when moist; very strongly
acid; abrupt smooth boundary.

B21 20-65 Reddish-yellow (7.5YR 6/6) 01ay with many
medium to coarse distinct red (2.5YR 5/6)
mottles; strong medium subangular block~y struc-
ture; very firm when moist; medium acid; gradual
smooth boundary.

B22 65-100 Reddish-yellow (7.5YR 6/8) clay with many
medium to coarse prominent red (10R 4/8)
plinthic mottles; strong medium subangular
blocky structure; very firm when moist; medium
acid; gradual smooth boundary.

823 100-140+ Yellow (10YR 7/6) clay with many medium to
coarse distinct yellowish-red (5YR 5/6) mottles
and a few very dusky red (7.5R 2/4) plinthic
nodules; strong medium subangular blockyr strue-
ture; very firm when moist; strongly acid.

Remarks: A few ironstone pebbles are scattered throughout the profile.

Table 6. Continued

(Typic Haplustox very fine clayey,
kaolinitic, acid, isohyperthermic. )

Location: Darlac Province, Republic of Viet Nam; Eakmat Agricultural
Experiment Station, 50 m north of back boundary fence, and 20 m
west of access road through main plot area,
Vegetation and land use: Freshly plowed crop residue, formerly used
for variety trials.
Slope and land form: Level to nearly level ( < 2$ slope) on old
Drainage and permeability_: Excessively drained, little surface runoff,
very rapid permeability; water table always below sampled profile.
Parent material: Basalt weathered in-situ,
Date sampled: November 17, 1971

Horizon Dept1 Description
Ap 0-18 Very dusky red (10R 2.5/4) clay; moderate to
strong fine granular structure; firm when
moist, plastic and sticky when wet; very strongly
acid; smooth abrupt boundary; strong efferves-
conce with H202-

Bi 18-31 Very dusky red (7.5R 2/5/4) clay; moderate
medium subangular blocky breaking into moderate
fine granular structure; firm when moist,
plastic and sticky when wat; strongly acid;
small bits of charcoal scattered through
horizon; clear smooth boundary; strong effer-
vescence with H202*

B2* 31-140t Dusky red (10R 3/4) clay; moderate angular
blocky breaking into moderate fine granular
structure; firm when moist, plastic and sticky
when wet; strongly acid; gradual diffuse
boundary; strong effervescence with H202*

* No differences with depth could be distinguished in the B2,
Subdivisions of the B2 for laboratory analysis were made as follows:
821 from 31-60, B22 from 60-100 and B23 from 100-140+.

Table 6. Continued

(Typic H~aplustox very fine clayey,
knolinitic, acid, isohyperthermic, )

Location: Darlac Province, Republic of Viet N~am; Eakmat Agricultural
Experiment Station, 100 m from entrance and 30 m north entrance
Vegetation and land use: Peanuts, but formerly panted to mung beans.
Slope and land form: Level to nearly level ( < 20slope) on old
Drainage and permeability: Excessively drained with little surface
runoff and moderately rapid permeability; water table always below
sampled profile.
Parent material: Basalt weathered in-situ.
Date sampled: November 17, 1971.

Horizon Depth Description
Ap 0-18 Dark reddish-brown (2.5YR 2.5/4) clay; moderate
to strong fine granular structure; firm when
moist; very strongly acid; smooth abrupt bound-
ary; strong effervescence with H202-.

B2* 18-140+ Dark reddish-brown (2.5YR~ 2.5/4) clay; moderate
to strong medium subangular blocky breaking to
strong fine granular structure; firm when moist;
strongly acid; no boundary; strong effervrescence
with H202*

* No differences with depth could be distinguished in the B2,
Subdivisions for laboratory analysis weresmade as follows.
B21 from 18-20, B22 from 60-100 and B23 from 100-140+.

The terraced pedon has an oehric epipedon below which the color

becomes reddish-yellow similar to the natural phase and remains so for

i m before becoming yellow. Red plinthic mottles occur throughout the

profile but become more frequent in the lowest horizon. There was no

textural discontinuity observed in this profile.

The two pedons from Eakmaat have similar umbric epipedons below

which are deep undifferentiated subsoils. The only difference between

the profiles was the thin B1 horizon in the Eakmnat I profile in which

several small bits of charcoal were found.

Physical Analyses

Particle size distribution.--The physical analyses (Table 7) indicates

well-developed argillic horizons in the Trang Bang and Thu Due pedons.

The clay content of these argillic horizons does not decrease signifi-

cantly with depth within the sampled profile. This is evidence of

extensive development over a prolonged period of time (11).

In the Dalat natural phase pedon there is some evidence of clay

accumulation but not the necessary 1,2 times the overlying horizon to

qualify for an argillic horizon in the section of the profile that was

above the beginning of the oxic horizon. There is also no water-dis-

persible clay in any horizon. In the terraced Dalat pedon, truncation

prohibits any comparison which might indicate the presence of an argillic

horizon. As with the natural pedon, there is no measurable water-dis-

persible clay.

In the Eakmnat pedons the difference in elay content between surface

and subsurface horizons is insufficient for an argillic horizon. As with

the Dalat profiles there is no water-dispersible olay in either profile.

Table 7. Particle size distribution and organic carbon
content for the six soil pedons examined

Texture Organic
Horizon Depth Sand Silt Clay carbon
om - - - - - - -

Trang Bang
Ap 0-18 65 24 11 0.64
A12 18-26 60 22 18 0,29
B21 26-48 54 27 19 0.21
B22 48-76 59 18 23 0.20
B23 76-140+ 62 16 22 0.10

Thu Due
Ap 0-20 79 15 6 0.19
B21 20-50 70 12 18 0.18
B22 50-90 66 12 22 0.24
B23 90-140+ 63 9 28 0.14'

Dalat N
Ap 0-18 28 11 61 3.70
Ai2 18-37 29 16 55 2.34
Bi 37-42 25 13 62 1.63
B21 42-65 15 8 77 0.42
B22 65-88 9 9 82 0.31
B23 88-111 12 13 75 0.30
Ct 111-140+t So 9 41 0. 50

Dalat T
Ap 0-20 18 15 67 0.81
B21 20-65 7 0 93 0.27
B22 65-100 9 5 86 0.26
B23 100-140+ 9 16 75 0. 19

Eakmat I
Ap 0-18 2 12 86 2, 73
B1 18-31 0 12 88 1.59
B21 31-60 0 7 93 1.25
B22 60-100 0 7 93 0.96
B23 100-140 0 7 93 0, 92

Eakmnat II
Ap 0-18 0 12 88 2.67
B21 18-60 0 8 92 2.00
B22 60-100 0 7 93 1,04
B23 100-140+ 0 7 93 0. 79

Organic carbon.--The organic carbon in the Trang Bang and Thu Due

pedons is less than ik in the surface horizon with a general decrease

with depth (Table 7). This organic carbon content is consistent with

that accumulated in other more temperate soils of similar texture and

drainage (16, 47). The natural Dalat pedon and both Eakmat pe3ons,

however, accumulated considerable more organic carbon than found in

sails of comparable texture and drainage from more temperate areas such

as the southeastern U.S. (16). The accumulation of organic carbon in

these pedons persists to greater depth than in comparative temperate

soils, as noted by both Eakmat pedons containing nearly 10 organic carbon

in the lowest horizon sampled. This accumulation of organic matter

contradicts Jenny's hypothesis on the relationship between organic matter

accumulation and latitude (34), but is consistent with Jenny's observa-

tions of high organic matter accumulations in the tropics (34). Since

the native vegetation on these pedons, particularly at D~alat, was minimal,

Jenny' s explanation of massive bio-eycling and regeneration does not

explain the large accumulation of organic carbon in these pedons. The

accumulation of organic carbon in these pedons is therefore thought to

result from the interaction of Fe and Al with organic functional groups,

thus reducing the rate of decomposition as proposed by Greenland (27)

and Broadbent (15).

Colloidal Analyses

X-ra~y diffraction.--The X-ray diffractographs indicated that both the

Trang Bang and Thu Due soils have similar mineralogy in their clay

separates (Figure 4). Both contain predominantly well-orystallized

kaolinite as indicated by the sharp first and second order diffraction

peaks at 7.2 and 3.6 A. respectively, which completely disappeared upon

3,46 3450 7, 2 10 14 18 3.6 5.0 2 2 10 1* 18
Figure 4. X-ray diffractographs for Mg-saturated,
glycerol solvated and heated K-saturated
clay separates from selected horizons of the
Trang Bang and Thu Due pedons.

heating to 550oc. The second most common mineral was intergrade ver-

miculite-chlorite. This is indicated by the broad diffraction peak at

14 A which intensifies upon K saturation (not shown) and collapses to a

skewed 10 A diffraction peak upon heating to 5500C, The skewed 10 A

diffraction peak becomes less skewed with depth which indicates a general

progression toward pure vermiculite with depth. In the Trang Bang Ap

horizon, the 14 A peak extends to 18 A indicating the possible presence

of some montmorillonite. In addition to these minerals, there are smaller

amounts of quartz~and mica as indicated by the weak diffraction peaks

at 3.4 A and 10 A, respectively.

X-ray diffractographs of the clay separates from the Dalat pedons

(Figure 5) indicate large amounts of gibbsite in the upper horizons that

decrease with depth. This is shown by the decreasing intensity of the

diffraction peaks at 4.8 and 4.4 A and the disappearance of these peaks

upon heating. The rest of the mineralogy is similar to that of the

Trang Bang and Thu Duc pedons. There are strong kaolinite diffractions,

as well as the vermiculite-chlorite intergrade material. The latter is

perhaps slightly more chloritic in nature than in the previous pedons.

A small mica peak can also be discerned and a small quartz peak. The

quartz peak appears less intense than in the previous pedons. A possible

anatase peak at 3.5 A is shown after heating has decrystallized the

kaolinite to eliminated its second order diffraction.'

X-ray diffractographs of the clay separates from the Eakmat pedons

(Figure 6) indicate predominantly knolinite with a small amount of

gibbsite. The gibbsite content was enough for only the most intense

4.8 A peak to be shown. The skewed peak starting at 4.4 A is interpreted

as the 011 kaolinite diffraction from somewhat poorly crystallized and

Figure 5. X-ray diffractographs for Mg-saturated,
glycerol solvated and heated K-saturated
clay separates' for selected horizons of the
Dalat pedons.


Figure 6. X-ray diffractographs for Mg-saturated,
glycerol solvated and heated K-saturated
clay separates for selected horizons of the
Eakmnat pedons.

finely divided material. There is also a high angstrom diffraction

peak around 20 A which disappears upon heating. It is not positively

identified, but could be a super-structure diffraction from kaolinite.

Differential scanning calorimeter.--The thermographs (Figures 7. 8. 9,

10, 11, 12) provide the best quantitative estimate of gibbsite and

kaolinite in the clay separates. The results obtained were consistent

with the X-ray patterns discussed previously.

In the Trang Bang and Thu Due pedons, kaolinite accounts for

approximately half the separates, with some decrease in kaolinite with

depth within the argillic horizon. No measurable amounts of gibbsite

were found (Figures 7, 8).

In the Dalat pedons, kaolinite and gibbsite account for most of

the clay separates. As indicated by the X-ray patterns, the gibbsilte

decreases and the kaolinite increases with depth (Figures 9, 10).

In the Eakmaat pedons, kaolinite is the dominate mineral and uni-

formly accounts for approximately 5090 of the clay separates while

gibbsite contributes between 3 and 50 to the separates.

Wet chemical analysis.--The wet chemical analysis provides an estimate

of both the non-crystalline material and vermiculite (Table 8). The

results indicate the highest amounts of non-crystalline material occurs

in the Dalat profiles followed by the Eakmat profiles. However, the

Alexiades and Jackson procedure (4C) includes gibbsite in the non-crys-

talline analysis. The gibbsite content determined by DSC has to be

subtracted from the total non-orystalline analysis to obtain the correct

non-orystalline values.

The amount of vermiculite identified remained relatively constant

Degrees C

Figure 7. DSC thermography of clay separates from the Trang Bang pedon.


300 400 5ou0u
Degrees C
Figure 8. DSC thermographs of clay separates from the Thu Due pedon.

Degrees C
Figure 9. D)SC thermography of clay separates from the
Dalat N pedon.


200 300 400 500 son
Degrees C
Figure 10. DSfC thermographs of clay separates from the Dalat T pedon.

200 300 400 5ou ous
Degrees C
Figure 11, DSC thermographs of clay separates from the Eakmat I pedon.


200 300 400 500 600
Degrees C
Figure 12. DSC thermographs of slay separates from the Eakmat II pedon.


Table 8. Non-crystalline, vermiculite, and Fe203 analysis of the six soil pedons

Wet chemical analysis Fe203 extractions
Horizon Depth Non-crystalline Vermiculite NaqP207 N 4-0% DCB
cm ------ clay separate ------ ----- -$ whole soil -------

Trang Bang!
Ap 0-18 14.9 6.4 0.14 0.18 0.28
Ai2 18-26 6.6 4.4 0.14 0.38 0.44
B21 26-48 9.5 6.2 0.08 0.16 0.21
B22 48-76 7.7 5.0 0.04 0.15 0.21
B23 76-140+ 14.9 6.8 0.04~ 0.10 0.29

Thu Due
Ap 0-20 16.5 6.4 0.04 0.05 0.28
B21 20-50 6.7 5.3 0.16 0.15 0.46
B22 50-90 7.1 6.0 0.14 0.16 0.52
B23 90-140+~ 6.5 6.3 0.08 0.11 0.48

Dalat N
Ap 0-18 48.7 6.8 0.49 0.50 7.01
Ai2 18-37 45.1 6.7 0.48 0.45 7.00
Bi 37-42 62.0 5.6 0.33 0.38 7.26
B21 42-65 28.7 4.4 0.01 0.09 6.27
B22 65-88 31.4 4.8 0.00 0.07 4.58
B23 88-111 28.9 3.6 0.00 0.04 3.00
Ci 111-140+ 31.3 4.5 0.00 0.03 0.95

Wet chemical analysis Fe203 extractions
Horizon Depth Non-crystalline Vermiculite NaqP207 NH4-0X DCB
cm ------ k clay separate ------ -----B whole soil -------

Dalat T
Ap 0-20 27.7 5.3 0.56 0.26 8.83
B21 20-65 25.3 5.3 0.01 0.10 7.32
B22 .65-100 28.6 4.6 0.00 0.05 6.06
B23 100-140+ 23.1 4.5 0.00 0.11 4.23

Eakmat I
Ap 0-18 241.6 5.5 0.28 0.98 15.98
B1 18-31 29.9 5.9 0.04 1.18 17.92
B21 31-60 31.1 6.2 0.02 0.78 17.73
B22 60-100 26.8 6.3 0.01 0.84 16.06
B23 100-140+ 2164.8 0.01 1.01 18.16

Eakmnat II
Ap 0-1.8 28.6 5.8 0.27 0.74 17.23
B21 18-60 27.8 7.2 0.16 0.81 18.09
B22 60-100 20.4 5.4 0.01 16.92
B23 100-140+ 23.4 4.1 0.01 0.71 16.33

Table 8. Continued

for all the samples analyzed. In the Trang Bang, Thu Due, and Dalat

profiles, the vermiculite identified was considered to represent that

portion of the intergrade material that was vermiculite-like. The

trend of increasing vermiculite with depth observed in the X-ray pat-

terns of the Trang Bang, Thu Due, and Dalat soils failed to materialize

in the CMC analysis for vermiculite. Vermiculite is also identified in

the Eakmnat pedons, although none is apparent in the X-ray diffractographs.

In most Eakmaat samples, the vermiculite ranges from 4 to 6$ of the clay


The three Fe extractions indicate that most of the Fe required

DCB for solubilization (Table 8). This implies that the Fe is well-

polymerized, if not crystalline (45). Even in the Dalat and Eakmnat

pedons in which Fe accounts for 10 to 200 of the total colloids, better

than 900 of the Fe required DCB for dissolution. The only trends noticed

with the NaqP207 extract was that it followed the general trend of the

organic matter concentrations. The Tarmms reagent extract was able to

extract more Fe from the Dalat pedons than the Trang Bang or Thu Due

pedons, but the most Fe was extracted from the Eakmat samples which are

highest in total DCB-extractable Fe.

Quantitative analysis.--The final quantitative analysis indicates that

the colloidal fraction of the Trang Bang and Thu Due soils are comprised

of approximately 50 to 55P knolinite and 25 to 30k vermiculite-chlorite

intergrade (Table 9). Included in the intergrade is approximately 58

vermiculite. These values remain relatively constant throughout their

respective profiles. The remaining colloids include less than 10~ non-

orystalline material in all but the upper horizons, 100 or less quarter

Table 9, Quantitative estimate of colloidal components for the six soil pedons examined

Total Organic Non-crys- Kaoli- Gibb- Intergrade
Horizon Depth colloids matter Fe209 talline nite site Total Vermiculite QuaR :~ Total

Trang Bang
Ap 0-18 11 10. 0 2.5 12.9 46. 0
Ai2 18-26 18 2.8 2,4 10.3 45. 8
B21 26-48 19 :1i.8 1.1 9,2 50. 3
B22 48-76 23 1.5 0.9 7.5 48.2
B23 76-140+ 21 0. 8 0. 7 9.6 48 2

0 20.1 5.6
o 30.6 4.2
0 30.9 6.0
0 31.6 4.9
0 33.3 6.7

0 25. 8 5. 7
0 25.5 5.1
0 25.4 5.7
0 24.4 6. 1

9.9 101.4
8.4 100.3
85 101.8
10.2 ;9.9
9. 6 _jz2.2

9.8 101.6
9. 0 t oo.9q
7.6 101.5
6. 0 91.3

1.4 97.3
1.8 95.8
2.3 95.7
1.6 100.2
1.9 100.6

Thu Due
4.7 14.8 40.8
2.6 6.4 55.6
2,4 6,8 57,4
1.7 6.4 51.9

Ap 0-20 6
B21 20-50 18
B22 50-90 22
B23 90-140+ 28

Dalat N
Ap 0-18 61 10.3 115 10.2 161 34.9 12.4 5.3
A12 18-37 55 7.3 12. 7 10. 7 16. 5 38. 7 8.8 5.8
s1 37-42 62 4.5 11.7 23.2 20.5 31.0 5. 0 4. 7
B21 42-65 77 0.9 8,2 11.2 41.5 16,2 16.'0 4.0
822 65-88 82 0.7 5.6 17. 5 38. 6 13. 0 18. 0 4.5
B23 88-111 75 0.7 4. 1 20.6 50.39 2.8 20. 1 3.4
C 111-140+ 41 0,2 2.3 27.4 50. 5 1.0 17.3 3.4


Total Organic Non-crys- Kaoli- Gibb- Intergrade
Horizon Depth colloids matter Fe20) talline nite site Total Vermiculite Quartz Total

Table 9. Continued

Dalat T
2.1 13. 2 10. 5 40. 3 14. 8 17.3 4. 5
0. 5 7.9 14.4 51. 1 6. 7 19. 2 4. 8
0. 5 7. 0 16. 6 52, 8 5.4 20. 1 4.3
0. 5 5.6 19. 7 5~. 8 1. 0 20.8 4.2

Ap 0-20 67
821 20-65 93
B22 65-100 86
B23 100-140+ 75

Ap 0-18 86
Bi 18-31 88
821 31-60 93
B22 60-1 00 93
B23 100-140 93

Ap 0-18 88
B21 18-60 92
B22 60-100 93
B23 100-140+ 93

1.9 100.1
0 99.8
1.0 103.4
o 102.4e

Eakmnat I
5.5 182 16.4 37.8 5.0
2,8 20.4 20. 7 42. 9 4. 8
2.3 19.1 21.9 41.4 5.2
1.8 17.3 19.39 45. 5 4.8
1.7 19. 5 14.9 4c. 0 4. 5

Eakmnat II
5.2 19.6 18.3 36.3 5.8
3. 7 19. 6 18.2 40.3 5. 5
1.9 18.2 163 42.4 7.3
1.5 17,6 13. 2 40. 3 7. 8

o 4.2
0 4.5
0 4.9
0 5.1
0 3.8

o 4.4
o 5.5
o 4.3
0 3.3




and less than 30 Fe20 Since the intergrade minerals and kaolinite

are, or can be, constant-charge minerals, these two soils are considered

to have predominantly constant-charge colloidal surfaces.

The colloidal fraction of the Dalat pedons has decreasing quantities

of organic matter, Fe20 and gibbsite with depth; while the non-crystal-

line material, knolinite,and vermiculite-chlorite intergrade increases

with depth. Since the organic matter, Fe-oxides,and gibbsite are all

constant potential-colloids, while the intergrade minerals and kaolinite

are, or can be, constant charge colloids, these pedons change from a

constant-potential dominated colloidal system to a constant-charge

dominated colloidal system with depth. This unique property of these

pedons will be utilized extensively in future comparisons of the

chemistry of the two types of colloids. Since most crops are grown on

the surface horizon, agronomically, the Dalat soils would be considered

to have predominantly constant-potential colloids.

Except for the organic~matter which declines with depth, the

colloidal fraction of the Eakmat pedons are almost uniform. Each

horizon contains approximately 200 Fe20 20k non-crystalline material,

40P kaolinite, 6k gibbsite,and 4.58 vermiculite. If it is assumed that

20k Fe-oxcides will saturate the kaolinite, the constant-potential

surfaces of the Fe-oxides will dominate the colloidal surfaces in these

two pedons.

Chemical Analyses

Soil reaction.--The soil reaction (Table 10) measured in distilled H20

indicated all horizons of all pedons to be acidic; half the samples are

below pH 5.0 and only 4 subsurface samples are above pH 5.6. This amount

of acidity is enough to anticipate some orop response to reduction in

Table 10. Soil reaction for the six soil pedons examined

H20 KC1 CaC12 K2S04 KC1- CaC12- K2S04- K2S04-
Horizon Depth H20 H20 H20 KC1

Trang Bang
Ap 0-18 5. 28 4.11 4.60 4. 70 -1.17 -0.68 -0.58 -0.59
A 2 18-26 4.83 3.80 4. 08 4.38 -1.03 -0. 75 -0. 50 -0.58
B21 26-48 4. 56 3. 70 4. 06 4.395 -0. 86 -0. 50 -0. 21 -0.65
B22 48-76 4.49 3.61 3.93 4.22 -0. 88 -0. 56 -0. 27 -0.61
B23 76-140+~ 4.49 3. 61 3.94 4.391 -0. 88 -0. 55 -0. 18 -0. 70
Thu Due
Ap 0-20 4.80 4.26 b.35 4.70 -0.56 -0.45 -0.10 -0.44
B21 20-50 4.20 3.3 3.75 4.30 -0.67 -0.45 +0.10 -0.77
B22 50-90 4.27 3. 52 3. 80 4. 20 -0. 75 -0. 47 -0. 07 -0.68
B23 90-140t 4.40 3.52 3.85 4.18 -0.88 -0.55 -0.22 -0.66
Dalat N
Ap 0-18 3.98 3.91 3. 95 4.38 -0. 07 -0.03 +0.40 -0.47
A12 18-37 4. 90 4. 50 4. 60 5. 13 -0.40 -0.30 +0. 23 -0.63
Bi 37-42 4. 79 4.63 4. 68 5.396 -0. 16 -0. 11 +0. 57 -0.73
B21 42-65 5.41 4.50 4.68 5.51 -0.91 -0.73 +0.1 -.0
B22 65-88 5. 72 4. 13 4. 73 4. 90 -1. 59 -0. 99 -0. 82 -0.17
B23 88-111 6. 02 4. 08 4. 63 4. 58 -1. 94 -1. 39 -1.44 -0.50
C 111-140+ 5. 59 4.10 4. 50 4.63 -1.49 -1. 04 -0. 96 -0. 53

pHI __ pH_
H20 KC1 CaC12 K2SO4 KC1 CaC12- K2S04- K2Sq4-
Hoizn Depth H20 H20 H20 KC1

Table 10. Continued

Dalat T
4.33 5.03
4.98 5.o?
4.71 4.55
4.65 4.50

Eakmrat I
4.20 4.50
4.82 5.28
4.95 5.40
4.70 6,00
4, 79 5. 65

Eakmat II
4.31 4.58
4.58 4.95
4.80 5.45
5.00 5.70


-0. 64
-0.0 o

-0. 54
-0. 83

-0. 76

-0. 05
-0. 50

-0. 54


+0. 06


-0. 70
-0. 70

5. 86





4. 72




+0. 04 -0. 58
-0.17 -0.66
+0. 12 -0. 73
+0.38 -0.75


acidity by liming.

The addition of a non-specific electrolyte (KC1) decreased the ptI

by 0.01 to 1.94 units. This difference could in part be attributed to

the two different types of colloids, With constant-charge colloids,

the Gouy-Chapman equation [3] stipulates that the increase in ionic

strength produced by the addition of an electrolyte must be accompa-

nied by a reduction in potential in order for the surface charge to

remain constant. Thus, the added electrolyte not only reduces the pH

by mass action exchange with the B and Al*+ ions retained in the

double-layer, but also forces a reduction in surface potential and

depresses the double-layer thickn~ess. With constant-potential colloids

it is not necessary for the surface charge to remain constant. Thus

the Gouy-Chapman equation can balance the increase in ionic strength

by a combination of increasing surface charge and decreasing surface

potential. Thus under similar initial conditions, the potential drop,

i.e. pH drop, will be less with constant-potential colloids than with

constant-charge colloids.

The best indication for differentiating colloidal types from

L3 pH values is illustrated in the Dalat pedons, particularly the Dalat

N profile (Table 10). In these pedons, the colloids change from

constant-potential near the surface to constant-charge at the bottom

of the profile. This change in colloid type is accompanied by an

increase in a pH value as the colloids become progressively more the

constant-charge type. The other four pedons are either all predomi-

nantly constant-oharge as shown by Thu Due and Trang Bang or all

predominantly constant-potential as oeour with both Eakmnat pedons.

In comparing these four profiles, there is a greater average 6 pH

in the Trang Bang and Thu Due profiles (0.85) than in the Eakmat

profiles (0.60). In these four profiles the differences within the

profile can be attributed as much to differences in texture, organic

matter, or exchangeable bases as to colloid type.

The use of an anion reactive electrolyte (K2S04) compares the mass

action of the cation to replace exchangeable Al"+ and 8 with the

tendency of anions to be specifically adsorbed to the surface with

the release of OH- ions. When the anion adsorption exceeds the mass

action of the cation to exchange H+ and Al"* plus the reduction in

surface potential due to increased ions in solution, the pH increases

above the H20-pH. This is interpreted to mean that the anion colloid

surface interactions should be an important management consideration (49).

In these soils, a positive a pH was observed only in the Dalat and

Eakmat pedons. The Trang Bang and Thu Due pedons have a negative 6 pH.

This implies that the anion activity of the Trang Bang and Thu Due soils

are relatively low compared to the anion activity of the D~alat and

Eakmnat soils.

The difference between the pH with the non-specific and with the

anion reactive electrolyte is worthy of note. This value remains nearly

constant for all samples. The mean is 0.66 + 0.14. Theoretically, this

should measure the difference in adsorption potential at equal ionic

strength. It should thus show the greatest influence of colloid type.

The failure to do this requires additional research effort to develop

an adequate explanation. There is a trend toward more negative values

with depth. This is best shown in the Eakmat profiles and may be

related to adsorption energy and to the degree the anion adsorption

potential has been satisfied.

Extractable acidity.--BaC12-TEA adjusted to pH 8.2 extracted the maximum

measured acidity from all samples (Table 11). The values range from

1.35 meq/100g in the Ap horizon of the Thu Due profile to 32.13 meq/100g

in the Ap horizon of the Eakmat I profile. Although these values

represent the extreme, a sharp distinction may be seen between the small

amount extracted from the Trang Bang and Thu Due soils compared to the

much larger amounts extracted from the Dalat and Eakmat profiles. These

differences are thought to be related to adsorption of OH ions to form

pH-dependent charges on the constant-potential colloids that dominate

the Eakmnat and upper Dalat profiles, and not the result of neutralization

of H+ and Al3 from the exchange complex of phyllosilicates. If this

is the case, the use of BaC12-TEA to estimate acidity in soils with

predominantly constant-potential surfaces might grossly overestimate

active soil acidity.

In the case of constant-potential soil colloids, the KC1 extraction

might be a more reasonable estimation of soil acidity. Since KC1 is

a neutral unbuffered electrolyte, it will extract only exchangeable H+

and Al"+ ions on the exchange complex and not become involved in the

surface adsorption of potential-determining ions. In the soils studied,

the KC1 extractable acidity ranged from 0.25 meq in the lower horizons

of the Eakmat profiles to 3.88 meq in the lower horizons of the Dalat T

profile (Table 11). The trend of KC1 extractable acidity is for the

greater amounts to be extracted from the Trang Bang and Thu Due profiles

and the least to come from the Eakmat profiles, This represents a

complete reversal from the trend shown with the BaC12-TEA extraction.

This reversal of measured acidity between the two extracts supports the

contention that the BaC12-TSA extract measures more than acidity.

Table 11. Extractable and titratable acidity for the six soil pedons examined

KC1 extractable Ratio Ba(0H9)2 titration
Horizon Depth BaC12-TEA Acidity Al BaC12-TEA pHI 5 pH 8
/KC1 inflection
---------- meq/100g --------- ---- meq/100g ---

Trang Bans
0.42 0.11
0. 98 0. 84
1. 15 0. 89
1,42 1. 06
1.49 1,45







3. 09
2. 74


3. 09

0. 00

0. 00

0. 88

2. 08


Thu Due

Dalat N



0. 79

6, 74
2, 84

KC1 extractable Ratio Ba(0H)2 titration
Horizon Depth BaC12-TEA Acidity Al BaC12-TEA pH 5 pH 8
/KC1 inflection
---------- mceq/100g --------- ---- meq/100g ---

Dalat T
Ap 0-20 12.99 1,25 0. 41 10. 39 0. 77 6.43
B21 20-65 8.37 1.3 1.08 5.47 0.96 5.48
B22 65-100 9.54 3.58 2.94 2. 66 2. 52 5.95
B23 100-140+- 8.37 3. 88 3.399 2. 16 2. 90 5.8(e

Eakmnat I
Ap 0-18 32.19 2. 65 1.69 12. 12 1.97 17, 87
B1 18-31 19.46 0. 33 0. 12 58. 97 0.oo0 12. 90
B21 31-60 17.72 0. 56 0. 05 31.64 0. 00 10. 02
B22 60-100 15. 98 0. 25 0. 00 63.92 0. 00 9. 72
B23 100-140e 16.42 0. 76 0. 06 21. 61 0. 00 7. 38

Eakmat II
Ap 0-18 30.47 2.58 1.73 11.81 2,26 17.52
B21 18-60 23.56 0. 86 0. 46 27. 39 1.35 14. 81
B22 60-100 17. 04 0.25 0. 03 68. 16 0. 00 9. 32
B23 100-140t 14.47 0. 29 0. 00 49. 90 0. 00 7.64c

Table 11. Continued

Perhaps the best way to evaluate the two acidity extracting

procedures according to soil and colloid type is to look at the ratio

of BaC12-TEA acidity to KC1 acidity (Table 11). The lower the ratio,

the greater the percent of BaC12-TEA acidity which is extracted by KC1.

In the six profiles being studied, this ratio does appear to separate

the two colloid types. The Trang Bang, Thu Due, and lower horizons of

the Dalat profiles all have extractable acidity ratios less than 5.0.

These are the soils with predominantly constant-charge colloids in

which the acidity is associated with exchangeable H+ and Al* In the

upper horizons of the Dalat profiles and the Eakm~at profiles the extract-

able acidity ratios are greater than 10 which indicates a low level of

BaC12-TEA acidity is extracted by KCL. This implies that more OH- from

the buffered BaC12-TEA are being consumed by the soil than can be

accounted for by neutralization of active Kt or Al" These OH- would

appear to be involved more directly with colloidal surfaces. The Dalat

N profile best indicates this trend since it goes from a high ratio to

a low ratio as the nature of the colloidal surfaces change from constant-

potential surfaces of gibbsite and Fe-oxides to the constant-charges

surfaces of the vermiculite-chlorite intergrade.

The Al extracted with the KC1 and evaluated as tri-valent Al

ranges from 0.0 meq/100g to 3.39 meq/100g. These values are always

below the total K(C1-extracted acidity. This indicates that some of the

K(C1-extracted acidity is from exchangeable H This is unusual,although

perhaps justifiable,considering the low KC1 pH of many of these soils

and significant organic matter content,

Potentiometric titrations.--The entire problem of source and nature of

acidity in these soils can be illustrated by potentiometric titrations

with Ba(0H)2 (Figures 13, 14, 15). With these titration curves, the

acidity resulting from exchangeable H~ or Al3 appears graphically

similar to that of an acid base titration with an inflection at the

equivalence point. In contrast the acidity resulting from pH-dependent

sources appears as a gradual linear pH increase with added base. In

general, a soil sample will titrate with an inflection at approximately

pH 5.0. This inflection represents the exchangeable H+ and Al"* acidity

as shown by the correlation between the inflection value and the KC1-

extractable acidity. The regression equation and correlation coefficient

for all samples from the six pedons are:

KC1 acidity = 1.034 Ba(0H)2 int. + 0.4063 ra = 0.8992

The total acidity titrated to pH 8,0 is well-correlated with the total

extractable acidity as measured with BaC12-TEA. The regression equation

and correlation coefficient for these values from all samples are:

BaC12-TEA acidity = 1.7225 Ba(0H)2 pH 8.0 + 0.4138 ra = 0.9770

The difference between the two measurements of total acidity probably

reflects a difference in equilibration time and chelating ligands (83).

The Trang Bang and Thu Due profiles contain little or no titratable

acidity in their Ap horizons due to the very low buffer capacity of

their sandy surfaces (Figure 13). As the elay content increases in the

argillic horizons, the titration curves contain the distinct inflection

near pHi 5.0 after which the curves rise rapidly with additional base,

This is indicative of acidity from exchangeable H+ and Al3 with little

pH-dependent acidity.

The Dalat profiles indicate a progression from predominately pH-

dependent acidity to predominantly exchangeable Kr and Al3 acidity

with increasing depth (Figure 14). The curve for the Ap horizon contains