Citation
Influence of soil organic matter on phosphorus and oxalate sorption and desorption in a spodoic horizon

Material Information

Title:
Influence of soil organic matter on phosphorus and oxalate sorption and desorption in a spodoic horizon
Creator:
Bhatti, Jagtar S., 1958-
Publication Date:
Language:
English
Physical Description:
xii, 147 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Anions ( jstor )
Carbon ( jstor )
Clay fraction ( jstor )
Desorption ( jstor )
Oxalates ( jstor )
pH ( jstor )
Phosphates ( jstor )
Soil organic matter ( jstor )
Soils ( jstor )
Sorption ( jstor )
Dissertations, Academic -- Soil and Water Science -- UF
Soil and Water Science thesis Ph. D
City of Madison ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1995.
Bibliography:
Includes bibliographical references (leaves 136-146)
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Jagtar S. Bhatti.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
021777449 ( ALEPH )
33436664 ( OCLC )

Downloads

This item has the following downloads:


Full Text











INFLUENCE OF SOIL ORGANIC MATTER ON PHOSPHORUS AND
OXALATE SORPTION AND DESORPTION IN
A SPODOIC HORIZON














-By

JAGTAR S. BHATTI


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1995















ACKNOWLEDGMENTS


It is my privilege to express my sincere gratitude, appreciation and heartfelt thanks to Nick B. Comerford, my advisory committee chairman, for his able guidance, keen interest, constructive criticism and constant encouragement during the course of this investigation. I feel pleasure also in expressing high regards to other members of my advisory committee, Earl Stone, Brian McNeal, Cliff Johnston, P. Surash Rao, H. Gholz and Eric Jokela, for my scientific development.

The financial support provided by the National Science Foundation for funding my assistantship and the experimentation is greatly acknowledged. I appreciate also the valuable technical assistance of Mary McLeod, Randy, and Dr. Cliff Johnston at various phases of the work. Mary's friendship was invaluable as was her advice on a wide range of issues. My sincere thanks are also due to everyone in the Forest Soils Laboratory for their cooperative attitude and discussions during the various phases of this investigation. I enjoyed all of my new friendships, especially those of my fellow graduate students.

Finally, I want to thank my wife Gurmeet and my son Kulraj for their

patience, love and understanding during this adventure. Without their tolerance and moral support, I would have not done it.


ii















TABLE OF CONTENTS


ACKNOWLEDGMENTS.. ......... ii

LIST OF TABLES............. ... . . v

LIST OF FIGURES. .................... viii

ABSTRACT . .................... xi


CHAPTERS . ........ ..............

1 GENERAL INTRODUCTION........... ...
Phosphorus in flatwood Spodosols . . . . . 1 Spodic Horizons... . . . . .. 2
Phosphorus in Spodic Horizons. . . . . .3
Organic Anions in Soil . . 5
Influence of Organic anions on Phosphorus Availability 6

2 INFLUENCED OF SOIL ORGANIC MATTER AND pH ON
OXALATE SORPTION INTO A SPODIC HORIZON . 12
Introduction 12
Material and Methods 15
Results . . 22
D iscussion . . . . . . . 50

3 INFLUENCED OF SOIL ORGANIC MATTER AND OXALATE
ON SORPTION OF PHOSPHORUS INTO A SPODIC
HORIZON . 55
Introduction.......... ... . 55
Material and Methods....... .... . 59
Results . . 64
Discussion . . 78


iii









4 INFLUENCED OF SOIL ORGANIC MATTER ON DESORPTION OF


PHOSPHORUS AND OXALATE FROM A SPODIC
HORIZON ...
Introduction ...
Material and Methods ..............
Results ............
Discussion ...................


5 CONCLUSIONS ..
Conclusions from this Study ......
Influence of Oxalate on the P Nutrition of Trees Future Research ........ ........

APPENDICES

A TITRATION CURVES FOR THE WHOLE SOIL
AND THE CLAY FRACTION

B SORPTION OF OXALATE IN THE PRESENCE OF P
BY THE CLAY FRACTION
Introduction ......
Material and Methods .........
Results ........ ...........


C MEASUREMENT OF P DESORPTION BY DIFFERENT
METHODS AT VARIABLE AND
CONSTANT pH .
Introduction ............. ......
Material and Methods .
Results. ...


. . . . .85 . . . .8 5 . . . 8 7 . . . .9 2
104


108 108
.110 ...... 112


114 114


118 118 118 . 120


. . . 123
. . . 123
. . . 126
. . . 129


D DESORPTION OF P AND OXALATE .


REFERENCES.


BIOGRAPHYCAL SKETCH


iv


134 136

147















LIST OF TABLES


Table Page

2.1. Parameters of the linear regression models for oxalate sorption
by the clay fractions and the whole-soil samples . 25

2.2. Parameters of the linear regression models relating
release of OH ions and organic carbon for the
clay fractions and whole-soil samples. 34

2.3. Ratios of OH- ions released to oxalate sorbed
for the clay fractions and whole-soil samples 37

2.4. Parameters of the linear regression models relating release of
aluminum and iron with oxalate sorption for the clay
fractions and whole-soil samples 40

2.5. E4/E6 ratio of organic carbon released into solution
during the sorption of oxalate by the clay fractions .. 45

2.6. Intensity of the absorption bands at 1610 cm'
(aromatic C=C and/or H-bonded C=O stretching
of COOH) and 1460 cm1 (OH deformation,
C-O stretching of phenolic-OH, and/or C-H deformations
of CH, and CH2) with standard deviations
for clay with organic matter at different pH values for
different amount of oxalate added. . 49

3.1. Langmuir sorption isotherm parameters for P sorption onto
the whole-soil samples and clay fractions with and
without organic matter 66
3.2. P sorption onto the whole soil material compared to P
that of sorption onto the clay fraction
expressed on a whole soil basis... 69


V









3.3. Amount of Al Fe and organic carbon released
into solution with P + oxalate sorption onto
the whole-soil samples and the clay fractions 70

3.4 Effect of oxalate in reducing the sorption of P onto
whole-soil samplesand the clay fractions. 71

3.5. Parameters of the linear regression models for
the release of OH- ions and organic carbon from soil and clay . 75

3.6 Ratios of OH- ions released to P sorbed for P
only and P + oxalate for the whole-soil
samples and the clay fractions . 77

3.7 E4/E6 ratio of organic carbon released into
solution during the sorption of P only, and
P + oxalate by the clay fractions 79

3.8 Parameters of the linear regression models relating release of
aluminum and iron with P sorption for P only and
P + oxalate by the whole-soil and clay fractions. . 80

4.1 Parameters of Freundlich models for the desorption of
P from the whole-soil samples and clay fractions in the
presence of 5 mM oxalate.... . 94

4.2 Quadratic equation coefficients for the sorption of anions
by the whole-soil samples and the clay fractions. 97

4.3 Influence of oxalate sorption by clay fractions and whole-soil samples
on the ratio of OH ions released, and OH + P
released to oxalate sorbed 98

4.4. Influence of P sorption by clay fractions and whole-soil samples
on the ratio of OH ions released, and OH + oxalate
released to P sorbed. . 103

B-i Percent reduction in oxalate sorption for the
whole-soil samples and clay fractions.. . 121

C-I Freundlich isotherm parameters for the desorption of P from
soil using the dilution and sequential extraction methods
at variable and constant pH....... . 131


vi









D-1 Influence of previously sorbed P by clay and whole-soil
samples on the oxalate sorption and release of P,
Al, Fe, OH, and organic matter.. . 134

D-2 Influence of oxalate previously present on the P
sorption and release of oxalate, Al, Fe, OH,
and organic carbon by clay fractions and whole-soil samples.. . . . 135


vii
















LIST OF FIGURES


Figure

1.1. Three possible phosphate surface complexes .

2.1 FTIR spectra of clay (a) with organic matter and
(b) without organic matter. .

2.2 Isotherms for oxalate sorption by clay fractions at
different pH values. .

2.3 FTIR spectra of clay without organic matter after oxalate
sorption at pH 3.5 (A), 4.5 (B), and 5.5 (C). .

2.4 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 3.5 .

2.5 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 4.5 .

2.6 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 5.5 2.7 Isotherms for the sorption of oxalate by the whole-soil sam

2.8 Release of OH ions by clay fractions at varing pH as
a function of oxalate sorption .

2.9 Release of OH ions into solution during the sorption of
oxalate by the whole-soil samples .

2.10 Kinetics of OH ions released during the sorption of oxalate
by the clay fractions at pH 4.5 2.11 Release of Al from the clay fractions at varing pH as a
function of oxalate sorption. . . .


Page





. 22 . 24 . 26 . 27 . 28


ples.


29 31 32 33 35


38


viii









2.12 Release of Al from the whole-soil samples as a
function of oxalate sorption. ............... 39

2.13 Concentrations of aluminum species at different pH levels as
influenced by oxalate 41

2.14 Release of organic carbon by the clay fractions at varing pH as a
function of oxalate sorption. ..... 43

2.15 Release of organic carbon from the whole-soil samples
as a function of oxalate sorption. ..... 44

2.16 Relationship between Al release and organic carbon release
by the clay fraction at varing pH............ . 46

2.17 Relationship between Al release and organic carbon release
from the whole-soil samples. 47

3.1 Isotherms for the sorption of P by the clay fraction ........ 65 3.2 Isotherms for the sorption of P by the whole-soil samples .. 67 3.3 Release of OH- ions during sorption of P by the clay fraction 73 3.4 Release of OH- ions during sorption of P by the whole-soil samples 74

3.5 Kinetics of OH- ions released during the sorption of anions
by the clay fraction 76

4.1 P desorption curve for whole-soil samples and clay fractions. 93

4.2 Influence of previously sorbed P on oxalate sorption and
P desorption by clay fractions. 95

4.3 Influence of previously sorbed P on oxalate sorption and
P desorption by whole-soil samples .. 96

4.4 Influence of previously sorbed oxalate on P sorption and
oxalate desorption for clay fractions. 101

4.5 Influence of previously sorbed oxalate on P sorption and
oxalate desorption by the whole-soil samples 102


ix









A-i Titration curves for the clay fractions: A) with organic matter,
B) without organic matter .......... 114

A-2 Buffer curve for the clay fractions, used to calculate the
amount of OH ions released into solution: A) with organic matter,
B) without organic matter . 115

A-3 Titration curve for the whole-soil samples: A) with organic matter,
B) without organic matter . . 116

A-4 Buffer curve for the whole-soil samples, used to calculate the
amount of OH ions released into solution: A) with organic matter,
B) without organic matter.. . ..... 117

C-i P desorption curves by dilution and extraction methods
at both variable and constant pH.. . 130

C-2 Relationship between Kd and the equilibrium concentration
of P as obtained by different methods .. . 133


x















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


INFLUENCE OF SOIL ORGANIC MATTER ON PHOSPHORUS AND
OXALATE SORPTION AND DESORPTION IN SPODOSOL By

JAGTAR S BHATTI

May, 1995


Chairperson Dr. N. B. Comerford Major Department: Soil and Water Science

Phosphorus (P) deficiencies are common for poorly-drained Spodosols of the flatwoods region of the Lower Coastal Plain of the Southeastern United States. Large quantities of total P may be present in spodic horizons, but the level of water-soluble P tends to be quite low. The goal of this investigation was to study P and oxalate sorption and desorption by the clay fraction and whole-soil material of a spodic horizon as influenced by soil organic matter. Understanding the sorption mechanisms for P in the presence of oxalate and soil organic matter should form the basis for explaining P extraction by roots from soil surfaces, and the buffering of P levels in soil solutions.

The spodic horizon had a higher capacity to sorb P compared to oxalate,


xi









with soil organic matter significantly reducing the sorption of both P and oxalate. Maximum reduction in P sorption (about 50%) was observed when both organic carbon and oxalate were present in the system. Ligand exchange apparently was the dominant mechanism for P and oxalate sorption. Some of the sorption sites were common for both the anions. Oxalate and P formed different types of surface complexes, with pH determining the type of surface complex formed by oxalate. Oxalate formed monodentate and/or binuclear surface complexes at pH

3.5, while bidentate complexes at pH 4.5 and 5.5. P formed monodentate and/or binuclear surface complexes at pH 4.5.

Significant amounts of organic carbon, Al and Fe were released into

solution during the sorption of oxalate. This study provided the first experimental evidence for Stumm's theory of congruent dissolution of minerals by oxalate.

Oxalate desorbed large amounts of P into solution. The presence of organic matter further increased the amount of P desorbed. Oxalate appeared to desorb the P both through ligand exchange and surface dissolution. This could increase the initial P concentration in solution and effect the P buffer power. Since oxalate formed stronger complexes on the mineral surfaces P failed to desorb the oxalate into solution.


xii















CHAPTER 1

GENERAL INTRODUCTION



Phosphorus in Flatwood Spodosols

Phosphorus (P) deficiencies are common for the poorly-drained Spodosols of the flatwoods of the lower Coastal Plain of the Southeastern United States (Comerford et al., 1984). Cycling of P, which occurs through mineralization, immobilization and redistribution of P in soil, depends on its physico-chemical properties. These include P sorption by colloidal surfaces, as well as microbial, mycorrhizal and plant uptake of P (Stewart and Tiessen, 1987). Soil phosphorus can be divided into three general categories: i) ions and compounds in the soil solution, ii) ions sorbed onto or incorporated into the surfaces of inorganic constituents, and iii) ions that are components of soil organic matter.

Inorganic P in the soil solution of a Spodosol's A horizon is replenished by mineralization of organic P, phosphorus leaching from the forest floor, and sometime, by fertilizer application (Polglase et al., 1992). Mineralogy of the surface soil and bleached E horizon is dominated by quartz (Carlisle et al., 1988). Therefore, mineralization of organic matter, plays an important role in P availability. In many of these soils, mineral components have virtually no P retention capacity


I









2

in their surface horizons (A and E), either due to low clay contents or the nature of the clay fractions (Ballard and Fiskell, 1974; Fox et al., 1990b; Yuan, 1992). For the surface soils, a clay content as low as 10 g kg-1 is common. Thus P, along with organic matter, can be leached from the surface horizon and accumulate in the underlying spodic horizon.



Spodic Horizons

Organic matter migrates through the A horizon of Spodosols in soluble and colloidal forms and is adsorbed or precipitated, immobilized in the Bh horizon. Current concepts of the formation of spodic horizons are based on the formation of Al and or Fe humic complexes. Stability and mobility of these compounds depend on the metal concentration in the soil solution. If the amount of Al and/or Fe available for organo-metal complex formation is low in the A horizon, complexes will be formed in the A horizons with low metal/organic ratios. In this case the amount of Al and/or Fe chelated is insufficient to cause immobilization of metal organic compounds, and may then move down in the pedon. During the downward migration, these metal-organic compounds (De Coninck, 1980), concurrently sorbed more polyvalent cations, which results in a progressive decrease of their net negative charge. The presence of higher concentrations of Al and Fe in the subsoil and/or at pH values different from that of the surface horizon may eventually neutralize the remaining charge. This results in the precipitation of metals along with organic matter in subsurface horizons and leads to the









3


development of a spodic horizon (De Coninck, 1980; Farmer et al., 1983; Buurman, 1984; Tan, 1986; Ugolini et al., 1988). Therefore, the complexation and translocation of Al and Fe by organic acids is a primary mechanism during the podzolization process. This can result in large accumulations of organic carbon (which is mostly as humic and fulvic acid), Al, and Fe in spodic horizons.



Phosphorus in Spodic Horizons

Spodic horizons contain elevated levels of amorphous or poorly crystalline Al oxides, which can sorb P. Not all of this sorbed P is in plant-available form (Ballard and Fiskell, 1974). Large quantities of total P may be present in spodic horizons, but the level of water-soluble P tends to be quite low. The physico-chemical sorption of P is as an inner-sphere complex at the surface of Al and Fe hydroxides and at the broken edges of silicate clay minerals (Sposito, 1984). An inner-sphere P complex refers to a surface complex resulting from ligand exchange between a surface Lewis acid site (S) and the adsorbed ion (Parfitt, 1978; Goldberg and Sposito, 1985). Such complexes are quite stable, showing mainly covalent or ionic bonding character.

Three types of inner-sphere P complexes have been postulated: monodentate, bidentate and binuclear (Fig. 1; Parfitt et al., 1977). Tejedor-Tejedor and Anderson (1990) used fourier transform infrared spectroscopy (FTIR) to study the sorption of orthophosphate onto goethite particles in an aqueous suspension. They observed the formation of three different types of complexes: protonated and













Monodentate


OH


S-OH+ H2PO4 S-0-P--OH + OHB identate 0


OH S +
OH


H2PO4 -*


S


P 4


0


20H


OH


Binuclear S "'*

OH + H2PO4

S


+ OH-+H20


OH


Figure 1. 1. Three possible phosphate surface complexes. "S" could represent
either Al or Fe.


4


S-O s -0O









5

nonproptonated bidentate bridging, and a nonprotonated monodentate complex. The speciation of these complexes was a function of pH, as pH determines the species of phosphate in solution. They also identified a monodentate complex on the geothite surfaces when the iron/phosphate ratio was less than unity and pH was between 3.5 and 6.0. At higher pH (6.0 to 8.3) and higher concentrations of phosphate in the system, surface complexes were bidentate. Therefore, the type of inner-sphere complex in different soils may be determined by the pH and the concentration of available P in soil solution. Nanzyo (1987) studied sorbed phosphate species using diffuse reflectance infrared spectroscopy (DRIR) and found that phosphate reacts not only with surface sites but also with the structural aluminum of allophanic soils. He reported that P can also precipitate as noncrystalline aluminum phosphate. Most of the studies, however, that have been carried out on the mechanism of P retention and release have used pure Fe & Al oxides as the solid matrix. None of these studies were accomplished using soil materials, where different soil components may coexist.



Organic Anions in Soil

A wide variety of low molecular weight organic anions have been identified in forest and agricultural soils (Stevenson 1982). The commonly found organic anions include oxalate, citrate, formate, acetate, malate, maleate, lactate and fumarate (Gardner et al., 1982; Hue et al., 1986; Pohlman and McColl, 1988). Fox and Comerford (1990) reported that oxalate comprised 60 to 80 percent of









6


the identified, free, low molecular weight, organic anions in a group of Spodosols from north Florida. They also reported that the oxalate levels in soil solutions averaged an order of magnitude higher in spodic horizons than in the corresponding surface A horizon. These data show that oxalate is present and, if it also affects P sorption, could play a significant role in forest productivity.



Influence of Organic Anions on Phosphorus Availability

Fox et al. (1990a) proposed that, in lower Coastal Plain Spodosols, low molecular weight organic anions, whether leached from the surface horizon or produced in situ, might stimulate the release of phosphorus from mineral surfaces. Organic anions acting as ligands are known to release P by i) replacing P sorbed at surfaces of Al or Fe oxides through ligand-exchange reactions (Huang and Schnitzer, 1986); ii) dissolving metal oxide surfaces and releasing sorbed P (Martell et al., 1988); and iii) complexing Al and Fe in solution, thus preventing the reprecipitation of metal-P compounds (Ng Kee and Huang, 1977). It has also been observed that organic anions may block sites on mineral surfaces and reduce P sorption (Kafkafi et al., 1988).

Huang and Violante (1986) studied aluminum-citrate complexes in aqueous solution. They reported that Al and citrate form a 1:1 complex where the citrate ligand occupies three of the six coordination sites around each Al while each of the other three sites are occupied by a water molecule. Occupation of coordination sites by citrate instead of water imposes a restriction on the









7


subsequent hydrolysis of hydroxy-aluminum polymers. The greater the concentration of organic anions in the system, the greater should be the replacement of water molecules and the blocking of Al coordination sites. Therefore, the occupation of coordination sites by organic ligands would also block sites which would otherwise be available for the sorption of P. Thus, organic anions, through their reactions with Fe and Al both in solution and at soil surfaces should increase the availability of P in soils (Martell et al., 1988). The activity of these surface sites available for coordination with organic ligands, and the activity of ligands in the solution, strongly depend upon pH. The role of pH would be through i) its effect on the relative fractions of various species of organic anions in solution (and these species differ in their affinity for the adsorbent); ii) the variation in charge density of the solid surfaces with pH; and/or iii) competition between OH- and the organic anion for common sorption sites.

Release of P is possible only if the stability of the metal-organic complex is higher than that of the solid-phase metal-P complex. The stability of soluble metal-organic complexes depends on the presence and molecular arrangement of carboxylic and phenolic functional groups on the organic anion (Pohlman and McColl, 1986; Martell et al., 1988). Formation constants can be used to predict the relative effectiveness of different organic acids for releasing P into solution. Organic anions such as oxalate and malate have been classified as having high complexing ability for Al and Fe (Fox et al., 1990b; Pohlman et al., 1990) when they tend to form five- and six-membered rings between the anion and metal.









8


Therefore, a low molecular weight organic anion like oxalate could play an important role in solubilizing soil Al, therefore indirectly affecting the availability of phosphorus in forest soils.

Among 16 organic anions studied, a threshold value for log K, between 4.0 and 4.5 was required before substantial amounts of P were released from a spodic horizon in Florida (Fox et al., 1990a). Beyond this log K, value, release of Al and inorganic P increased with increasing formation-constant values. Violante et al. (1991) and Violante and Gianfreda (1993) studied the competitive sorption of phosphate and oxalate on aluminum oxide and montmorillonite. They reported that oxalate and P ions competed strongly for the same adsorption sites on Al oxides under both acidic and neutral conditions. Lan et al. (1994) studied the release of P by oxalate from different spodic horizons of different soils. These soils desorbed variable amounts of P, although each contained about the same amount of total P. The differences in P release may have been related to mechanisms by which the P was held by different soil components (crystalline minerals, an amorphous fraction, occluded P, or organo-mineral complexes) and/or the interaction of oxalate with these components.

Oxalate may influence P availability through its reaction with Al both in solution and at the surfaces of Al oxides as it forms stable complexes with Al (Martell et al., 1988). Release of P by oxalate could be by direct ligand exchange and/or dissolution of Al oxide mineral surfaces (Stumm and Morgan, 1981; Huang and Violante, 1986). Stumm (1986) proposed a theory of congruent dissolution of









9


mineral surfaces by low molecular organic anions. He suggested that organic ligands form complexes with Al and Fe both in solution and on the surfaces of oxides. The formation of complexes in solution by organic ligands increases the rate of dissolution of Al-oxalate complexes on surfaces. Ligand exchange at the surface decreases the strength of Al-OH-Al bonds and helps bring Al-oxalate complexes into solution. This rupture is thought to be induced by the competition of oxalate with the hydroxyls linking the aluminum atoms. Stumm (1986) further proposed that surface complexation is a prerequisite for nucleation for most solid crystals formed in solution as well. However, there is no experimental evidence in the literature to support the above theory, even with the use of pure clay minerals or Al and Fe oxides.

The complexation and sorption reactions of humic and fulvic acids with Al and Fe are similar in magnitude to those for the corresponding monomeric units (e.g. carboxylate, salicylate, dihydroxybenzoate) believed to contain functional groups similar to those in fulvic and humic acid (Stumm and Morgan, 1981; Stevenson and Fitch, 1986, Tan, 1986). Inoue and Wada (1971) concluded that the adsorption of fulvic and humic acids by allophane and imogolite resulted from the ligand exchange of oxygen in carboxylic groups with structural oxygen in the coordination shells of Al and Fe atoms. This suggests that organic carbon may block some of the adsorption sites which might otherwise be available for the sorption of P and organic anions. There are no studies to my knowledge which









10

examine the influence of natural soil organic carbon on the retention and release of organic anions and P in Spodosols.

A significant portion of the total P in a spodic horizon is inorganic, but not readily available for plant use. Phosphorus uptake by plants is influenced by the amount of root surface area, root growth rate, and P concentration in solution at the root surface (Barber, 1984). Phosphorus supplying capacity of a soil is determined not only by the amount of P in the soil solution but also by the soil's ability to replenish P lost from soil solution and/or during transport of P to plant roots via diffusion and/or mass flow. The soil's ability to replenish P lost from the soil solution depends partially upon release of P from sorbed and insoluble P. Thus, it is important to understand how P is being held by different colloidal surfaces as well as the influence of various chemical processed on the release of previously sorbed and insoluble P.

From the aforementioned, it can be concluded that there is a need to

understand how forest trees acquire nutrients from the soil profile in lower Coastal Plains. We can start to address this question by i) understanding the types of sites available and binding mechanisms for oxalate via spectroscopic studies; ii) investigating P sorption and desorption in the presence of oxalate in spodic horizons; and iii) defining the influence of soil organic colloids on the retention and release of P and oxalate. Understanding the sorption mechanisms for P in the presence of oxalate should form the basis for explaining P extraction by roots from soil surfaces, and the buffering of P levels in soil solutions.















CHAPTER 2

INFLUENCE OF SOIL ORGANIC MATTER AND pH ON OXALATE SORPTION ONTO A SPODIC HORIZON Introduction

Natural organic anions in soils and freshwater environments are derived from plant and animal residues, microbial metabolism and canopy drip. Simple aliphatic anions such as citrate and oxalate are continuously produced through the activities of microorganisms (Flaig, 1971), aqueous extraction of leaves (Bruckert, 1971; Stevenson, 1982), and activities of microorganism and/or roots in the rhizosphere (Reddy et al., 1977). The type of vegetation, pH, oxidation potential, water potential and temperature each affect the kinds and amounts of low molecular weight organic anions that are produced (Stevenson, 1982; Goh and Haung, 1986; Hue et al., 1986; Pohlman and McColl, 1988).

Among the aliphatic anions detected in the extracts of litter and soils,

oxalate occupies an important place in many ecosystems, both in well-, and poorlydrained soils (Vedy and Bruckert, 1982). For many organisms, oxalate is a metabolic end product of low residual nutrient energy (Cromack et al., 1979). Smith (1969) and Fox and Comerford (1990) measured significant concentrations of


11









12

oxalate (up to 2 mM) and other carboxylic anions in a variety of sediments, in the forest floor, in bulk soil, and in the rhizosphere of forest trees.

Oxalate alters chemical processes in soils through complexation reactions

with Al and Fe that occur in the soil solution and on the surfaces of soil particles (Stumm, 1986; Martell et al., 1988). It has acidic characteristics due to the presence of -COOH groups and reacts with mineral soil surfaces through i) electrostatic attractions, ii) complex or chelate formation, and iii) water bridging (Tate and Theng, 1981). Three types of inner-sphere oxalate complexes with soil surfaces are possible: monodentate, bidentate, and binuclear (Parfitt et al., 1977). Each requires the displacement of OH2 or OH- that had been coordinated to Al and Fe atoms on the surfaces. Such a reaction would increase the pH of the system. Therefore, the amount of OH- released could be used to identify possible reactive sites and possible mechanisms of oxalate sorption at different concentrations of oxalate and/or pH. A change in pH would alter the soil surface charge and, possibily, the species of organic ligand present in the system.

Spodosols of the southeastern Coastal Plain have an accumulation of

organic matter in their spodic horizons. As organic anions originate in the forest canopy or forest floor, they form mobile complexes with Fe and Al, and move to the Bh horizon where they are constrained as organo-minerals. Researchers have shown that organic matter may be immobilized through complex interactions with mineral surfaces (Greenland, 1971; Davis and Glour, 1981; Sibanda and Young,









13


1986). The reactivity of labile soil organic matter is generally attributed to the presence of carboxyl (-COOH) and phenolic hydroxyl (-OH) groups.

The major mechanisms by which soil organic matter is sorbed to mineral surfaces include i) anion exchange, ii) ligand exchange-surface complexation, iii) hydrophobic interaction, iv) hydrogen bonding, and vi) cation bridging (Sposito, 1984). Parfitt et al. (1977) suggested that carboxyl groups of fulvic and humic acid replace surface OH- from gibbsite, goethite and imogolite. Ligand exchange of surface-coordinated OH- and OH, from Fe oxides by humic substances had also been noted by Tipping (1981b). Jardine et al. (1989) suggested that the primary mechanism of dissolved organic carbon sorption is physical adsorption followed by an anion exchange mechanism. This suggests that organic carbon should alter the surfaces charge, surface area and retention properties of inorganic colloids in spodic horizons.

According to Stumm (1986), organic ligands form complexes with Al and Fe in solution and on the surfaces of Fe and Al oxides. The formation of such complexes in solution increases the rate of dissolution of surface complexes. Huang and Violante (1986), Miller et al. (1986), Tan (1986), and Pohlman and McColl (1986) also reported that organic anions which formed complexes with Al and Fe stimulated the dissolution of Al and Fe solid phases. Ohman and Sjoberg (1988) proposed that polycarboxylic acids are effective in solubilization of Al surfaces. They proposed this as a series of complexation-hydrolysis reactions:









14

Al(OH)3(S) Alq(OH)pL,(S) ;- Alq(OH)pL,(aq) z AlL.(aq) 7 Al3,(aq) (2.1) where S represents the mineral surface, L is the organic ligand, aq is the solution phase and p, q and r are the number of atoms in each molecule. The calculations of soil solution speculation with kaolinite as the solid phase, showed that, under acidic conditions, solution concentration of Al would increase in the presence of oxalate. This increase was attributed to the formation of [AIL]' and [AlL2] species in solution. Fox et al. (1990a) observed higher release of Al into solution with the adsorption of oxalate, and attributed this to the dissolution of solid-phase Al surfaces in Spodosols. Bilinski et al. (1986) found that the binuclear complex Al2(OH)2(C204)4- formed in solution when oxalate concentration was high, and then precipitated as Al oxalate with the oxalate ion acting as a bridging ligand between two aluminum atoms.

The objectives of the present study were to describe i) oxalate sorption on the clay fraction of a spodic horizon, ii) the effect of soil pH on oxalate sorption capacity, and iii) the influence of soil organic matter on oxalate sorption using fourier transform infrared (FTIR) spectroscopy.



Materials and Methods

Soil Material

Soil was collected from the spodic horizon (Bh) in a soil pit of a Pomona series (sandy, siliceous, hyperthermic Ultic Alaquod) at the Gator Nationals Forest site located in Alachua County, 10 km northeast of Gainesville, Florida (Swindel et









15


al., 1988). The soil material was air dried, passed through a 2-mm sieve and stored in plastic bags.



Preparation of Whole Soil

Two sets of soil samples were prepared: one with organic matter and the other without organic matter. One portion was treated with hot, 30% H20, to remove organic matter as outlined by Kunze and Dixon (1986). This sample was centrifuged and washed several times with distilled water, while the other portion was not treated. Both soil samples (with and without organic matter) then were saturated with Na using 0.01 M NaCl Excess Na was removed with four or five centrifugal washings using distilled water. The samples were oven dried at 110 0C.



Preparation of the Soil Clay Fraction

The water-dispersible clay fraction was separated by ultrasonification at a 1:5 soil to water ratio (Genrich and Bremner, 1974). After ultrasonification, wetsieving was used to remove the sands. The clay fraction (< 0.2 Pim diameter) was separated by centrifugal sedimentation (Jackson, 1979). The resulting clay was divided into two portions. One portion was treated with hot, 30 % H202 to remove organic matter as outlined by Kunze and Dixon (1986). The other portion was not treated. Both clay samples (with and without organic matter) then were









16


saturated with Na using 0.01 M NaCl Excess Na was removed with four or five washings using distilled water. The samples were oven dried at 110 "C.



Mineralogy of the Clay Fraction

Mineralogy of the clay was determined by X-ray diffraction (XRD) on

parallel-oriented specimens, using Cu-Ka radiation. Approximately 250 mg of clay were deposited on a ceramic tile under suction followed by K and Mg saturations on the tiles by washing with the respective chloride salt and rinsing. Glycerol then was added to the Mg saturated samples. Samples were scanned at 2' 20 per minute, at 25 C for both the K- and Mg-saturated samples. Minerals were identified from XRD peak positions, making use of differentiating responses to ion saturation. Detectable minerals in the clay sample included quartz, kaolinite, hydroxy interlayered vermiculite, and gibbsite.



Oxalate Sorption Studies

Sorption experiments were carried out with different concentrations of

oxalate over a pH range (3.5, 4.5 and 5.5). Triplicate samples of clay (250 mg with and without organic matter) were placed in 30 ml bottles with 25 ml of solution having an oxalate concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM adjusted initially to pH 3.5, 4.5 or 5.5 (using 0.1 M HCl or NaOH). Triplicate five-gram whole-soil samples from each treatment (with or without organic matter) were placed in 100 ml bottles with 50 ml of solution having an oxalate









17

concentrations of 0 mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM or 10 mM adjusted initially to pH 4.5 (using 0.1 M HCl or NaOH).

Two drops of toluene were added to inhibit microbial growth and all samples were placed in a reciprocating shaker for 24 hrs. The pH of the suspension was periodically (every three hours) adjusted with 0.1 M HCl or NaOH over the 24 hrs of the experiment, to maintain the initial pH. At the end of the reaction period, each suspension was centrifuged at 12,000 g for 20 minutes. The adsorbate was used for FTIR analysis and the filtrate was analyzed for oxalate, Al, Fe, and organic carbon. Sorbed oxalate was calculated from the difference between the initial and final oxalate concentration in solution.



Chemical Analysis

Solution pH was measured using a combination glass electrode and an

Orion pH meter. Aluminum was determined via flame emission spectrophotometer with a N20-CH2 flame. Fe was also determined, with an acetylene flame. Total organic carbon content of the clay suspension (with organic matter) was measured via persulfate oxidation and IR analysis of CO2 produced on a TOC apparatus (College Station TX). The E4 (X=465) to E6 (X=665) ratio of organic carbon released in solution by clay with organic matter was measured using a Shimadzu UV/Vis spectrophotometer.

Oxalate in the extract was determined by HPLC (Fox and Comerford,

1990a) using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton









18


Co., Reno, Nevada) with a Gilson single piston high pressure pump along with a Pheodyne model 7125 injection valve fitted with a 20 pL injection loop. The HPLC system used a Gilson Holochrom variable wavelength UV detector in conjunction with a Gilson computerized integrator. The eluent was 0.005 M HS04 at a flow rate of 2 mL min-' Oxalate concentration was calculated from the calibration curve obtained with standard solutions of 0.1 to 10 mM.



Titration Curves for the Clay Fractions and the Whole Soil Material

To determine the amount of OH released into solution during the sorption of oxalate, titration curves were prepared (Duquette and Hendershot, 1993) for each of the clay and whole-soil samples (i. e., with and without organic matter). Duplicate one-gram samples of clay and 10-gram samples of soil were equilibrated with 100 mL of water adjusted to pH 3.0. The suspension was allowed to equilibrate for 24 hrs. Titration was carried out with 0. 1 N NaOH from pH 3.0 to 7.0. With each addition of 0.1 ml NaOH, the solution was allowed to equilibrate for 2 to 5 minutes or until pH stabilized. The titration curves, for each type of clay and soil are presented in Appendix A-i and Appendix A-3. pH buffer curves were calculated which relate the change in pH with unit additions of NaOH for three different ranges of pH. These ranges were 3.44 to 4.44, 4.44 to

5.65, and 5.45 to 6.75. Models were fit to these curves and used to calculate the change in OH concentration required to change pH of the system to specific values (Appendix A-2 and Appendix A-4).









19


FTIR Analysis

FTIR spectra were obtained by placing small amounts of clay suspension on AgCl windows and allowing the deposit to dry. Spectra then were collected on a Bomem DA3.10 spectrophotometer equipped with a MCT detector and a KBr beam splitter, operating at 2.0 cm-' resolution. The Bomem DA3.10 spectrophotometer was controlled through a general-purpose interface bus (IEEE488) interfaced to a DEC Vaxstation-II computer.



Adsorption Isotherms

The Langmuir adsorption model was fitted to all sorption data. The following equation was used:



S KC
S. '" (2.2)



Where S is the amount of oxalate taken up per unit mass of soil (mM kg-1), Sm is the maximum amount of oxalate that was bound, C is the equilibrium concentration of oxalate (mM), and K is a constant related to the binding energy for oxalate. The parameters (K and Sm) were calculated by a least squares fit to a linear form of the equation:


C 1 C
KSM (2.3)









20


Speciation Calculations

The metals speculation model MINTEQA2 version 3.2 was used to calculate the species of Al and oxalate in solution, assuming that the system was under equilibrium conditions. Calculations were carried out at solution concentrations of oxalate and Al for the highest loading rate of oxalate (10-2 M oxalate) at different pH values. Log K values for soluble Al-oxalate complexes were obtained from Thomas et al. (1991). Log K was assumed to be 6.12 for AlC204 and 11.15 for Al(C204);. Two solid phase species of Al-oxalate were considered in the speciation model. The formation constants were taken from Bilinski et al. 1986. The values of constants used were 21.87 for Al3(OH)7(C204).3H20 and 5.61 for NaAl(OH)2(C204).3H20. Formation constants used for calculations of other species were those of the MINTEQA2 data base taken from Martell and Smith (1982).



Statistical Analysis

Statistical differences between different regression lines at each pH value were tested using the General Linear Models procedure of the SAS framework (SAS Institute, 1985). The model used was:



Yij- a X E (2.4)



where a is the intercept, P is slope of the line, and E is the error term. To compare two regression lines, we compared the a and P values of the respective









21

lines. The ANOVA and t tests were used to test the differences between a, and a2 as well as between P1 and P2. Significance of the difference between means was determined by the t-test (Snedecor and Cochran, 1980). Standard deviations or standard errors of means are given as well, when appropriate.


Results

Characterization of the Clay Fraction using FTIR

FTIR spectra of clay samples with and without organic matter are presented in Figure 2.1. Characteristic infrared bands for quartz, kaolinite, HIV and gibbsite minerals are identified. The broad OH-stretching band in the region 3000 to 3600 cm-1, along with well defined OH-stretching is characteristic of kaolinite, HIV and gibbsite, while the broad deformation band for water in the 1640 cm' region suggests the presence of amorphous clay. The spectra of clay with organic matter shows major absorption bands as well in the regions of 3300 cm-1 (0-H and N-H stretching), 2960/2925 cm' (aliphatic C-H stretching of CH, and -CH, groups), 1610 cm' (aromatic C=C and/or H-bonded C=O stretching of COOH), and 1400 (OH deformation, C-O stretching of phenolic OH, and C-H deformations of CH3 and CH2) together with the characteristic infrared bands for quartz, kaolinite, HIV and gibbsite minerals. Intensity of the broad OH-stretching band in the region 2700 to 3600 cm' for clay with organic matter was greater, reflecting the presence of OH-stretching associated with organic matter.










22


1.5

(A)
1.25


1


0.75 0.5


0.25


0
500 1000 1500 2000 2500 3000 3500 4000
Wavenumbers



1.5


1.25 (B)





0.75


0.5


0.25

v

500 1000 1500 2000 2500 3000 3500 4000
Wavenumbers




Fig. 2.1 FTIR spectra of clay (a) without organic matter and (b) with organic
matter.









23


Sorption of Oxalate by the Clay Fractions

Sorption isotherms for oxalate on clay samples with organic matter and without organic matter at constant pH are presented in Figure 2.2. Linear isotherms provided best fit for the data within the range of oxalate concentrations used. Parameters of the model at different pH values are presented in Table 2.1, with oxalate sorption being strongly dependent on pH. As pH increased, the oxalate sorption decreased. Soil organic carbon also had a significant influence on oxalate sorption at all pH values, with approximately a 2 to 3 fold reduction in oxalate sorption in the presence of organic carbon (Table 2.1).

FTIR spectra at the highest oxalate loading are presented in Figure 2.3 for clay without organic matter and in Figures 2.4, 2.5, and 2.6 for clay with organic matter A small shoulder was seen at 1710 cm-1 for the FTIR spectra of clay with organic matter for the 10-2 M oxalate loading at both pH 3.5 and 4.5. Parfitt et al. (1977) also observed the same bands for oxalate sorption by goethite and proposed the formation of a binuclear complex between oxalate and two Fe" ions. If inner-sphere complexes form between oxalate and clay surfaces, they must involve the coordination of carboxylic groups from the oxalate ion with Al and/or Fe atoms of the clay surface, through oxygen atoms. However, the absorption bands linking Al-oxalate vibrations, which would normally provide the most direct information on complexation, are too weak to be definitive. It was not possible to isolate these bands from the FTIR spectra of the clay due to the low surface concentrations of the oxalate.










1,000 300

100 E 30

-101 0 3

-E
01
CA


0.3 0.1


IA
~~~ -.... -.
-. . . . . . .


L


+ OM 3.5

- OM 3.5


0


2


+ OM 4.5 + OM 5.5

- OM 4.5 OM 5.5


6


8


4


Equilibrium concentration of oxalate (mM)



Isotherms for the sorption of oxalate by clay fractions at different
pH values.


24


10


Fig. 2.2









25


Table 2.1 Parameters of the linear regression models
clay fractions and the whole-soil samples.


for oxalate sorption by the


Organic matter pH Oxalate sorption

Slope Intercept R2


Clay Fractions
With Organic Matter 3.5 64.3c 21.3a 0.99
4.5 30.9e 17.8a 0.97
5.5 15.9' 13.6a 0.96
Without Organic Matter 3.5 209.Oa 20.4a 0.99
4.5 101.4 11.9a 0.99
5.5 46.5d 9.2a 0.99


With Organic Matterq 36.8b 17.6a 0.79
Without Organic Matter 1 19.1 13.9a 0.77


Whole Soil Samples
With Organic Matter 4.5 0.070b 0.044a 0.94
Without Organic Matter 4.5 0.095a 0.036a 0.95


a lack of significance samples.


The same letter in a column at different pH values indicates at the 5 % level within the clay fractions and the whole-soil @ Includes the combined samples at pH 3.5, 4.5, and 5.5;













1

08

06

0.4 0.2

0 13


2200 2650 3100
Wavenumbers (cm-i)


! ---- 10-2 MOXamsb --0.00 MOiMis


I

o0.6 ~0.6

0.4 02

0





Fig. 2.3


170


22 2M 3100
Wavenumben (cm-1)


3"0


FTIR spectra of clay without organic matter after oxalate .sorption at pH 3 5 (A), 4 5 (B), and 5 5 (C).


2200 2650 3100
Wavenumbers (cm-I)


1750


0
U


0.6


0.4 0.2


0
13


3550


3550


00


4000


4000


1750


I


30


0


-A


-


--





FTIR spectra of clay with organic matter for different concentrations of oxalate sorbed at pH 3 5.


Fig. 2.4


102 M Oxalate 10" M Oxalate 1e M Oxalae 0.00 M Oxate

















LI


FTIR spectra of clay with organic matter for different copcentrations of oxalate sorbed at pH 4. 5.


102 M Oxalate 10" M Oxalate 10i M Oxalate 0. 00 M Oxalaze


Fig. 2.5
















































'.4.

Fg 2.6


lO-2 M Oxalaie io- M Oxaiate O4 MOxaate


0.00 M oxalate


FTIR spa of CdY WM orSmac la for ditrsK cotcfltions
of oxalate sorbed at pH 5.5.









30


Oxalate Sorption by Whole Soil Samples

Oxalate sorption isotherms for the whole-soil samples are presented in Figure 2.7. The Langmuir model gave the better fit. The parameters of the linearlized form of the Langmuir models are presented in Table 2.1. The sorption maxima (S.) by the whole soil material was 10.57 for soil with organic matter and 14.20 for soil without organic matter The presence of soil organic matter significantly reduced oxalate sorption.



Release of OH

The sorption of oxalate onto the clay fractions and whole-soil samples

released large amounts of OH ions (Fig.2.8 and Fig. 2.9). Oxalate sorption was significantly and positively correlated with OH ions released (Table 2.2). The OHrelease was also significantly influenced by soil organic matter (Table 2.2). The amount of OH ions released at pH 3.5 into solution by the clay fractions was significantly different than the amount of OH ions released into solution at pH 4.5 and 5.5. The FTIR spectra of clay without organic matter showed considerable decrease in the intensity of OH-bands in the region 3000 to 3800 cm-' following oxalate sorption (Fig. 2.3).

The OH release in solution follows first order kinetics (Figure 2.10) for oxalate sorption by clay surfaces at pH 4.5. Release of OH was initially rapid during the sorption of oxalate (Fig. 2.10), and remained high even after 24 hrs when organic matter was not present. The ratios of moles of OH- released per












16


14 00%12



E 0
8
0
6
Nii
0 4


2


v I I II


2


4


6


8


Equilibrium Oxalate Concentration (mM)


Isotherms for the sorption of oxalate by the whole-soil samples.


31


,-'

+ OM-OM


10


0
0


Fig. 2.7








32


1,000

500


200


E 100

50
"C

202

OM 3.5- OM 4.5- OM 5.5
10__ ---

5 + OM 3.6- OM 4.6- OM 5.5

0
2

1
0 200 400 600 800
-1
Oxalate sorbed (mmol kg)
Fig. 2.8 Release of OH ions by the clay fractions at varing pH as a function
of oxalate sorption.











33


50 ,




IMI
40AA



E
E 30




20
0



T +' +OM -OM

0 10




0
0 5 10 15 20
I-1






Oxalate sorbed (mmol kG )



Fig. 2.9 Release of OH ions into solution during the sorption of oxalate by
the whole-soil samples.










34


Table 2.2 Parameters of the linear regression models
and organic carbon for the clay fractions


Material OM/pH


OH ions released


Slope Intercept


relating release of OH ions and whole-soil samples. Organic carbon released


Slope Intercept


R2


Clay fractions


With organic matter


3.5 0.65

4.5 1.65b

5.5 1.88b


Without organic matter
3.5 0.71a


13.73a 0.99

- 4.07a 0.99


1.30 a


58.4b 73.8 b 40.6 b


0.98


0.97

0.99 0.99


0.081a 2.52a

0.131b 6.34b

0.249' 5.85b


Whole-Soil


With organic matter
4.5 1. 99b
Without organic matter
4.5 2.483


0.73


-0.07b 0.95

-0.143 0.96


significance


0.99 0.99

0.99


-0.43


0.93


The same letter in a column at different pH values indicates a lack of at the 5 % level within the clay fractions or the whole-soil samples. @ Includes all samples at pH 3.5, 4.5, and 5.5;


4.5 1.82 b

5.5 1.73 b













800



IoM
-600
0
E


0


(0
2400

0
E


EE

C


)


-OM +ONM


5


10


15


20


Time (hrs)


Fig. 2.10 Kinetics of OH- ions release during the sorption
clay fractions at pH 4.5.


of oxalate by the


35


25


30









36


mole of oxalate sorbed are presented in Table 2.3. Average molar ratio at the higher sorption rates was not significantly different from 1.0 at pH 3.5, nor was it significantly different from 2.0 at pH 4.5 and 5.5 for both the clay and whole-soil surfaces.



Release of Aluminum and Iron

The influence of oxalate sorption on the release of Al is presented in Figures 2.11 and 2.12. Aluminum release increased linearly with sorption of oxalate (Table 2.4). The release of Al from whole-soil samples followed a sigmoidal shape, both for samples with and without organic matter (Figures 2.12). Soil organic matter significantly increased Al release, but the release of Al from the clay fractions without organic matter at pH 3.5, 4.5, and 5.5 was not significantly different (Fig. 2.11). Speciation of Al solution in the presence of oxalate was calculated using MINTEQA2. The majority of the Al in solution was present as [AlL]- and [AIL2]- species at pH 4.0 and above (Figure 2.13). The system was undersaturated with respect to both NaAl(OH)2(C204).3H20 and Al3(OH)7(C204).3H20. The concentration of [AlL2]- species increased with pH.

The Spodosol used in the study had very low concentrations of total Fe Iron released from the clay fractions and the whole-soil samples is presented in Table 2.4. The clay fractions and the whole-soil samples with organic matter released a significantly greater amount of Fe then materials without organic matter.









37


Table 2.3 Ratios of OH~ ions released to oxalate sorbed for the clay fractions
and whole-soil samples (mean SD; n=3).

pH OH released/Oxalate sorbed

Clay fractions
With Organic Matter 3.5 0.98 0.03a
4.5 1.73 0.11b
5.5 1.88 0.22b

Without Organic Matter 3.5 0.86 0.06a
4.5 1.96 0.11b
5.5 1.85 0.08b


Whole-Soil Samples
With Organic Matter 4.5 1.92 0.21b
Without Organic Matter 4.5 1.86 0.17b


Superscript "a" and "b" or 2 respectively at the


indicate that values are not significantly different from 1.0 5% level.









38


1,000



300







30




+ OM 3.5 + OM 4.5 + OM 5.5 3 OM 3.5 OM 4.5 OM 5.5



1


0 200 400 600 800

-1
Oxalate sorbed (mmol kg )





Fig. 2.11 Release of Al from the clay fractions at varing pH as a function of
oxalate sorption.








39


IAI


+ OM


-OM


I


/
I
A
I


/
/2
I


-


I I I I I I


0 2 4 6 8 10 12


14 16


-1
Oxalate sorbed (mmol kg )





Fig. 2.12 Release of Al from the whole-soil samples as a function of oxalate
sorption.


40


'0)



E
E

"C


30





20 10


0


I










40


Table 2.4 Parameters of the linear regression models relating release of aluminum
and iron with oxalate sorption for the clay fractions and whole-soil
samples.

Material Al release Fe release

OM/pH Slope Intercept R2 Slope Intercept R2

Clay Fractions


Without organic matter


0 573 40.62a


0.59"'


8.62a


0.74a -1.03 a


With Organic Matter


b.83b


7.11


2.77c -15.30a 4.27d -28.66a


0.99 0.99 0.99


0.97 0.99 0.98


0.007a 0. 009b 0.012'


0.0083 0.01 9d

0.03 6e


-0.213 0.99

-0.22a 0.99

-0.15a 0.99


-0.19a 0.98

-0.27a 0.97

-0.13a 0.98


With organic matter


2.96a -12.27a


Without organic matter

0.63b 16.07b


0.93 0.98


0.021a 0.010 b


-0.19a 0.73


-0.19 0.96


Whole-Soil Samples


With organic matter


4.5


Without organic matt
4.5


0.76b 2.16a

er
3.07a -2.75'


0.96 0.95


0 042a


0.025b 0.96


-0.061a 0.89


3.5

4.5 5.5


3.5

4.5

5.5


@ Includes all samples at pH 3.5, 4.5, and 5.5; The same letter in a column at different pH values indicates a lack of significance at the 5 % level within the clay fractions and the whole-soil samples.

















Al 3a

]Al- (OXalate)2


4.0 4.5


120



100


pH of the system


Fig. 2. 13 Concentrations of aluminum species in solution at different pH levels
in the presence of oxalate.


41


SAl-(oxalate)



























5.0 6.0


75





0 CL
0


80 60



40 20



0


3.5









42


Release of Organic Carbon

Organic carbon was released from the clay fractions and whole-soil samples in a linear fashion relative to oxalate sorption (Fig. 2.14 and Fig. 2.15). The amount of organic carbon desorption is presented in table 2.2. Seven to ten times more organic carbon was desorbed by 10 mM oxalate as compared to water. The pH of the system also greatly influenced organic carbon desorption, with significantly more organic carbon released at pH 5.5 than at pH 3.5 (0.08 mg g-' versus 0.249 mg g' ). E4/E6 ratios of the organic carbon released into solution during sorption of oxalate are presented in Table 2.5. The E4/E6 ratio, at the highest loading rate of oxalate (10 mM), decreased to 1.88. There was a linear relationship (R2 = 0.99) between organic carbon release and Al released at pH 3.5, pH 4.5 and 5.5 for the clay fractions (Figs. 2.16 and 2.17).

Spectra for the clay fractions with organic matter showed major absorption bands representing organic carbon in the regions of 3300 cm' (0-H and N-H stretching), 2960/2925 cm1 (aliphatic C-H stretching of CH3 and -CH2 groups), 1610 cm-1 (aromatic C=C and/or H-bonded C=O stretching of COOH), and 1400 cm-1 (OH deformation, C-O stretching of phenolic OH, and C-H deformations of CH3 and CH2,). Intensities of the COOH and CH3 bands in the spectral region 1610 cm-1 and 1400 cm' were calculated at all pH values and at all equilibrium










50

30 20


*10


0
3


2


Ae






3.5 4.5 5.5
0


0


100


200


300


-1
Oxalate sorbed (mmol kg )

Fig. 2.14 Release of organic carbon by the clay fractions at varing pH as a
function of oxalate sorption.


43


400


500








44


10




8










0
0


0 2
L


0)
LO
0

0




0 2 4 68 10 12 14
Oxalate sorbed (mmol kg )




Fig. 2.15 Release of organic carbon from the whole-soil samples as a function
of oxalate sorption.









45


Table 2.5. E4/E, ratio of organic carbon released into solution during the
sorption of oxalate by the clay fractions (Mean SD; n=3).

Treatment Concentration of E4/E6 ratio
oxalate added (mM)



3.5 10* 3.15 0.10
1.0 7.54+0.25

0.1 8.24 0.56

0.0 9.26 0.07


4.5 10 1.88 0.13
1.0 7.13 0.04

0.1 8.08 0.13

0.0 9.07 0.15


5.5 10 2.14 0.06
1.0 6.93 0.05

0.1 7.65 0.21

0.0 9.07 0.15


* Concentration of oxalate added into solution







46


1,000


800
3.5 4.5 5.5

2 600
2


400


200


0
0 10 20 30 40 50
-1
Organic carbon released (mg g )





Fig. 2.16 Relationship between Al release and organic carbon release by the
clay fraction at varing pH.








47


40





0%30

'E

E E
20
-3




10





0
0 2 4 6 8 10
-1
Organic carbon released (g kg )



Fig. 2.17 Relationship between Al release and organic carbon release from the
whole-soil samples.









48

concentrations of oxalate. The integrated absorption intensities was normalized with respect to the weight of the clay sample deposited on the AgCl window. Intensity of the 1610 cm' (aromatic C=C and/or H-bonded C=0 stretching of COOH and CH3) bands decreased as progressively more oxalate was sorbed by the clay surfaces (Table 2.6, Fig. 2.4, 2.5. and 2.6). The greatest decrease in the intensity of bands at 1610 cm1 (aromatic C=C and/or H-bonded C=O stretching of COOH) and 1460 cm- (OH deformation, C-O stretching of phenolic OH, and C-H deformations of CH, and CH2) corresponded with the maximum oxalate sorption. Decrease in the intensity of these bands was higher at pH 5.5, again corresponding to the pattern of organic carbon release seen in the batch studies. Release of organic carbon was significantly correlated to the intensity of these bands, with R2 values of 0.82 to 0.96.



Discussion

Surface Characteristics

Mineralogical analysis using XRD and FTIR data showed that the dominant crystalline clay minerals in the spodic horizon of this Spodosol were kaolinite, HIV and gibbsite. FTIR data established that the clay fraction also contained large amounts of noncrystalline Al oxide. Fox et al., (1990a) reported high amounts of oxalate-extractable (1357 mg/kg) and pyrophosphate-extractable (1379 mg/kg) Al for the same soil. Therefore, both crystalline and amorphous mineral surfaces were available for the sorption of oxalate.









49


Table 2.6. Intensity of the absorption bands at 1610 cm-' (aromatic C=C and/or
H-bonded C=O stretching of COOH) and 1460 cm-1 (OH
deformation, C-O stretching of phenolic-OH, and/or C-H
deformations of CH3 and CH2) with standard deviations, for clay with organic matter at varing pH values for different amount of
oxalate added (Means SD; n=3).

Oxalate added Organic carbon
pH mM released mg g-' V 1610 cm-' V 1460 cm-1


35.65a 0.60A


6.67 0.44 3.93 0.29

3.34 0.22


60.74 4.93 74.95 3.08 79.68 2.37

80.40 4.57


r = 0.93e


38.13 0.56

15.44 0.85 7.26 0.45

5.42 0.25


54.39 1.04 72.0 1.69


79.18 2.95 83.76 8.86

r 0.934


35.05 3.54

40.32 2.04 43.83 1.65

46.85 4.42


r = 0.938


32.14 1.19 42.15 0.22


43.38 2.19 43.65 3.67 r = 0.95'


42.09 0.74 18.95 1.04 6.21 0.63 5.35 0.61


39.49 0.19 61.78 2.07 67.93 1.79

70.40 1.74


r = 0.984


17.08 0.45 32.78 3.56 37.57 2.39 39.68 0.60 r = 0.978


@ Correlation coefficient between organic carbon released and intensity of the absorption bands at 1610 cm-1 at pH 3.5, 4.5 and 5.5. & Correlation coefficient between organic carbon released and intensity of the absorption bands at 1460 cm-' at pH 3.5, 4.5 and 5.5.


3.5


10.0 1.0 0.1 0.0


4.5


10.0 1.0 0.1 0.0


5.5


10.0


1.0 0.1 0.0









50

FTIR data also indicated the presence of a large number of OH functional groups. Since this soil has a very low Fe concentration, we assume that the OH groups were primarily attached to Al. The terminal Al on the mineral surfaces should be Al-OH- (aqua), Al-OH (hydroxy), or Al-0- (oxo) groups, depending on pH (Sposito, 1984). Since these spodic horizons have no anion retention capacity, the number of Al-OH,' (aqua) functional groups is insignificant. In the pH range of this study rules out the significance presence of Al-O-, therefore Al-OH (hydroxy) was assumed to be the dominant surface group.

Solution pH determines the oxalate species available for a surface reaction. The dissociation of oxalic acid is:


HOOC-COOH HOOC-COO- + H- pKI = 1.2 (2.4)

HOOC-COO- -OOC-COO- + H+ pK, = 4.2 (2.5)


These pK values would result in dissociation of only one COO- group at pH 3.5, while at pH 4.5 and 5.5 both COO- groups of oxalate would be dissociated. Therefore, at pH 3.5, oxalate should only form monodentate and/or binuclear surface complexes, while at pH 4.5 and 5.5, bidentate surface complexes can be formed.



Oxalate Sorption

Oxalate sorption onto the mineral surfaces may be through ligand exchange and/or precipitation. If the sorption process were dominated by ligand exchange









51


then we would expect to see several things from these data. First, there should be a release of OH ions accompanying oxalate sorption (Goldberg and Sposito, 1985). This would not be the case if precipitation of oxalate dominated. Second, there should be a linear relationship between the amount of oxalate sorbed and OH ions released into solution. Third, the molar ratio of OH ions released to oxalate sorbed should be approximately I or 2. Depending upon pH, a ratio of one would suggest a monodentate or binuclear inner-sphere complex, while a ratio of 2 would suggest a bidentate inner-sphere complex. We would expect a ratio of I at pH 3.5, because only one functional group of oxalate is dissociated at this pH and the formation of monodentate and/or binuclear surface complexes is expected At pH 4.5 and 5.5 we would expect ratios around 2, suggesting that inner-sphere complexes are almost exclusively bidentate.

If precipitation of oxalate were the dominant process we would not expect to see a strong relationship between Al and Fe release into solution and the amount of oxalate sorbed as well as a continuous release of OH- ions into solution with oxalate sorption, even after 24 hrs. The MINTEQA2 speciation calculations indicate that the amount of solid phase species of polynuclear aluminum hydroxyoxalate complexes (A13(OH),(C204).3H2O and NaAl(OH)7(C2O4).3H2O) were insignificant. This implies that precipitation was not the process controlling oxalate sorption in this system. However, this criterion has to be used with caution, as these calculations are based on the assumption that system was at equilibrium which has not been shown to exist in this study.









52


The data presented here overwhelmingly support the process of ligand

exchange. The OH was released in proportion to oxalate sorbed, as shown with both the batch sorption and FTIR measurements. The molar ratio was not statistically different from I at pH 3.5 or from 2 at pH 4.5 and 5.5 (Table 2.3). We further conclude that precipitation cannot be a significant factor in oxalate sorption in this study. Al was released into solution in relation to oxalate sorption, showing that oxalate sorption caused Al dissolution rather then precipitation. Also, MINTEQA2 calculations suggested that the oxalate formed soluble Al-oxalate complex in solution and keep it from precipitating. It must be remembered that this was a Na-saturated system without any Ca present on the exchange complex to form insoluble CaC204 species. Under field conditions, spodic horizon pH is about 3.8 to 4.2, with little Ca present on the exchange complex to interact with native levels of oxalate.

These data further support the conclusion that oxalate is sorbed as

monodentate and/or binuclear surface complexes at pH 3.5 ([mmol OH / HC20;] 1), but forms predominantly a bidentate complex at higher pH values ([mmol OH / C2042] = 2). These data support the conclusions of Parfitt et al. (1977), who also showed monodentate inner-sphere complex formation of oxalate on goethite surfaces at pH 3.4, at high oxalate concentrations, using infrared spectroscopy. They also suggested formation of a binuclear complex on goethite when oxalate loading rates were low.









53

Large releases of Al, Fe, and organic carbon during oxalate sorption further suggest a concurrent dissolution reaction. Fox et al. (1990a) studied the kinetics of oxalate sorption onto the same soil and reported that oxalate was sorbed during the first 6 hrs while release of Al continued for at least 48 hrs. These data, taken together, support Stumm's (1986) theory of congruent dissolution of mineral surfaces by oxalate following ligand exchange. This constitutes the first such experimental evidence to support Stumm's hypothesis. Oxalate on the clay and whole-soil surfaces forms stable bidentate and binuclear complexes. These complexes involve five- and six-membered rings between the oxalate and Al, which would decrease the strength of the Al-OH-Al bonds and bring Al-oxalate complexes into solution.

The clay fractions and whole-soil samples with organic matter released more Al, respectively than did the clay or soil without organic matter. Another source of Al released from the clay and soil with organic matter was the Al present as metal-organic complexes. Lee et al., (1988b) studied the forms of aluminum in selected Florida Spodosols and found that more than 75 % of the Al in spodic horizons was present as Al-fulvate. Aluminum can act as a bridge between soil particles and organic matter. If oxalate solubilizes Al through the formation of stable, soluble, Al-oxalate complexes in solution, then this could result in increased Al release.









54


Effect of Organic Carbon on Oxalate Sorption.

The removal of organic carbon increased the sorption of oxalate. This can be attributed to either competition between organic carbon and oxalate for same sorption sites or formation of new sorption sites on mineral surfaces when H202 was used to oxidize the organic carbon. These data do not differentiate between these possibilities; however, either option would enhance the surface area for oxalate sorption. Zelazny and Quresih (1973) reported that H,02 treatment of clay material from Florida soils enhanced surface area and decreased surface charge. This evidence suggests that, with the removal of organic matter, surface area would increase somewhat resulting in greater sorption of oxalate by clay and soil surfaces without organic matter.


l















CHAPTER 3

INFLUENCE OF OXALATE AND SOIL ORGANIC MATTER ON SORPTION
OF PHOSPHORUS ONTO A SPODIC HORIZON Introduction

Phosphorus deficiencies are common on the poorly-drained Spodosols of the flatwoods region of the lower Coastal Plain of the Southeastern United States (Pritchett and Comerford, 1983; Comerford et al, 1984). P availability depends on physico-chemical properties such as P sorption by colloidal surfaces. Organic ligands are continuously released into the rhizosphere by decaying plants and animals, through microbial processes, and as root exudates (Stevenson, 1982; Fox and Comerford, 1990). The supply of P to plants is strongly influenced by the presence of these organic ligands. Different organic anions have been reported to modify the sorption of phosphate by soils and soil components (Deb and Datta, 1967, Earl et al., 1979; Kafkafi et al., 1988).

Aluminum and iron, either in solution or as crystalline/amorphous soil

constituents, are the principal agents responsible for chemical fixation of phosphate in acid soils (Yuan and Lavkulich, 1994). Organic anions which are capable of forming stable complexes with aluminum and iron in solution (Appelt et al., 1975; Viotante and Huang, 1985; Traina et al., 1986a, 1986b; Huang and Schnitzer,


55









56

1986; Comerford and Skinner, 1989) or on mineral surfaces (Pohlman and McColl, 1986; Stumm, 1986; Kafkafi et al., 1988; Martell et al., 1988; Fox et al., 1990b) can be effective in reducing the P sorption capacity of soils. Phosphate sorption was reduced in the presence of humic acids, fulvic acids, and low molecular organic acids while each of these compounds were specifically sorbed onto pure mineral surfaces (Nagarajah et al., 1968, 1970; Sibanda and Young, 1986; Ryden and Syers 1987; Moore et al., 1992; and Violante and Gianfreda, 1993). However, little work has documented the effect of oxalate and soil organic matter on the sorption of P using soil materials.

The ability of different organic anions to compete with P for sorption sites on the surfaces of soil components was reported to be greatest at a pH equivalent to the pKa of the organic acid (Hingston et al., 1967; Hingston et al., 1972). Soil organic acids which contain carboxylic (-COOH) and/or phenolic (-OH) functional groups can bind to oxide surfaces, thereby reducing the number of surface sites available for P sorption (Yuan, 1980). This also alters electrostatic charge at the solid surface. Both of these interactions of organic anions are influenced by the solution pH, relative concentrations of different anions which may be present and intrinsic affinities of these anions for the mineral surfaces. The pH of the system is important in surface and solution complexation reactions as it will i) regulate the concentrations of various P and organic anion species which differ in their affinity for the solid surface, ii) affect the charge density of solid









57


surfaces, and iii) control the competition between OH-, P and organic anions for common adsorption sites (Barrow, 1987; Kafkafi et al., 1988).

Poorly-drained Spodosols are the dominant soil type in the flatwoods of Florida's lower Coastal Plain. Among the low-molecular weight organic anions present in these soils, the most abundant is oxalate (Fox and Comerford, 1990). Earlier work by various researchers indicated that the specific sorption of phosphate and organic ligands like oxalate is through ligand exchange (Goldberg and Sposito, 1985). This process should result in the release of OH. The amount of OH released during a sorption reaction in turn depends upon the characteristics of the surfaces, concentrations of the adsorbing species, and solution pH. The change in solution concentration of OH- ions can be used to identify the types of complexes formed on the various solid surfaces.

Organo-mineral complexes have an important influence on the physical and chemical properties, and reactivity of soil particles (Viotante and Huang, 1984; Viotante and Huang, 1985; Huang and Schnitzer, 1986). The main clay minerals present in flatwoods soils are quartz, kaolinite, gibbsite and hydroxy-interlayered vermiculite (HIV) (Harris et al., 1987a). Since Florida soils are sandy, these minerals are present as coatings on the sand particles (Harris et al., 1987a; 1987b) The binding materials for these clay-sized mineral particles to sand grains is reported to be an Al dominated gel-like substance (Lee et al., 1988a). Lee et al. (1988b) further indicated that Al acts as a cementing material, probably as Alfulvate. The Spodosols of the lower Coastal Plain are known to have high









58


concentrations of amorphous Al oxides and are classified as Alaquods. Organic substances which are present as surface coatings and/or acting as cementing material can significantly affect the retention of P by modifing the specific surface area and surface charge of crystalline and noncrystalline minerals in the finer fractions of the Spodosols. Therefore, it is important to investigate the sorption of P by this highly reactive component.

Recently, studies were carried out to investigate the sorption of P by

organo-mineral complexes. Violante and Huang (1989) studied the sorption of P on precipitated-Al products formed in the presence of organic ligands (citrate, tartrate, malate, aspartate and tannate). They found that the amount, the nature and the size of organic ligands coprecipitated with Al, as well as the surface properties of Al-organic ligand complexes, strongly influenced P sorption capacity. Haynes and Swift (1989) investigated the effect of pH on P sorption onto Alorganic matter complexes. They reported that increase in pH greatly increased P sorption capacity. These studies were performed on pure minerals. However in soils, many minerals coexist, and were formed in the presence of different organic coumpounds. Therefore, P sorption properties of these minerals could be considerably different from those of the pure minerals.

The purpose of the present investigation was to study: i) P sorption by the clay fraction and the whole-soil samples of a spodic horizon, and ii) the influence of oxalate and soil organic matter on P sorption in a such system.


1









59


Materials and Methods

Soil Material

Soil was collected from the spodic horizon (Bh) of a single soil pit of a Pomona series (sandy, siliceous, hyperthermic, Ultic Alaquod) at the Gator Nationals Forest site located in Alachua County, Florida. The soil material was air-dried, passed through a 2-mm sieve and stored in plastic bags. Particle-size analysis of the sample was carried out using standard methods (Page et al., 1986). The soil contained 91 % sand, 8 % silt and I % clay.



Preparation of Whole-Soil Sample and the Clay Fractions

The prepartion of whole-soil samples for the sorption experiment was previously described in Chapter 2. The water-dispersible clay fraction was separated (Genrich and Bremner, 1974) as reported in Chapter 2. Both the wholesoil and clay samples were dried at 110 C.



Sorption Studies on Clay Fraction and Whole Soil Sorption of Phosphate Alone

Sorption experiments were carried out over a range of phosphate

concentrations at pH 4.5 using a batch procedure. Triplicate 250 mg samples of clay (with and without organic matter) were placed in 30 ml bottles with 25 ml of solution having a phosphate concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM adjusted initially to pH 4.5 (using 0.1 M HCl or NaOH). Triplicate five-gram









60


sample of whole soil from each treatment (with or without organic matter) were placed in 100 ml bottles with 50 ml of solution having a phosphate concentration of 0 mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM or 10 mM adjusted initially to pH 4.5 (using 0.1 M HCl or NaOH).



Sorption of Phosphate and Oxalate Added as a Mixture

Sorption experiments were carried out at different oxalate to phosphate molar ratios. The concentration of oxalate used was 1.0 mM with a phophate concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM for the clay samples. For the whole-soil sample with and without organic carbon, an oxalate concentration of

1.0 mM was used, with a phosphate concentration of 0 mM, 0.1 mM, 0.5 mM,

1.0 mM, 3.0 mM, 7.0 mM or 10 mM. pH of the clay suspensions and the whole-soil suspensions was initially adjusted to 4.5 using 0.1 M HCl or NaOH.

Two drops of toluene were added to inhibit microbial growth, and the samples were placed in a reciprocating shaker for 24 hrs. pH values of the suspensions were periodically (every three hours) adjusted to 4.5 with 0.1 M HCI or NaOH over the 24 hrs of the experiment. At the end of the reaction period, each suspension was centrifuged at 12,000 g for 20 minutes. The resultant supernatant solutions were used for analysis of oxalate, Al, Fe, inorganic P, and organic carbon. Sorbed P and oxalate were calculated from the difference between the initial and final concentrations in solution. The adsorbate was saved for further experiments, as described in Chapter 4.









61


Chemical Analysis

Solution pH was measured using a combination glass electrode in

conjunction with an Orion pH meter. Aluminum was determined using a flame emission spectrophotometer with N2O-C2H2 flame whereas Fe was detected with a C2H, flame. Total organic carbon content of the clay sample with organic matter was measured via persulfate oxidation and IR anaylasis of the resultant CO, using on a TOC apparatus (College Station, TX). E4 (X=465) to E6 (k=665) ratios of the organic carbon released into solution from the clay with organic matter were measured using a Shimadzu UV/Vis spectrophotometer.

Oxalate in the extract was determined by HPLC (Fox and Comerford,

1990), using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton Co., Reno, Navada) along with a Gilson single piston high pressure pump and a Pheodyne model 7125 injection valve fitted with a 20 pL injection loop. The HPLC system used a Gilson Holochrom variable wavelength UV detector in conjunction with a Gilson computerized integrator. Eluent was 0.005 M H2S04 at a flow rate of 2 mL min-' Oxalate concentration was calculated from the calibration curve obtained with standard solutions of 0.1 to 10 mmol L-'.

Inorganic P in the filtrate was determined by a molybdenum-blue

colorimetric procedure using ascorbic acid as the reductant (Murphy and Riley, 1962). This is an operational definition of inorganic P, for molybdenum may also hydrolyze some organic P (Stainton, 1980).









62


Titration Curves for the Clay and Soil Samples

To determine the amount of OH- ions released into solution during the sorption of P and/or oxalate, titration curves were prepared for each clay and whole-soil sample (with and without organic matter). Duplicate one-gram samples of clay and 10-gram samples of soil were equilibrated with 100 mL of water adjusted to pH 3.0. The suspension was allowed to equilibrate for 24 hrs, titration then being carried out (with 0.1 N NaOH) from pH 3.0 to 7.0. With each addition of 0.1 ml NaOH, the solution was allowed to equilibrate for 2 to 5 minutes or until the pH stabilized. pH buffer curves, relating the change in pH with unit addition of NaOH over the pH range 4.40 to 6.00, were calculated Curves were fit to the data and used to calculated the amount of OH- ions required to change pH of the soil and clay systems to specific values.



Adsorption Isotherms

The Langmuir adsorption model was fit to the sorption data. The following equation was used:

S KC
Sm (3.1)
1+KC


where S is the amount of P or oxalate taken up per unit mass of clay or soil (mM kg-), Sm is the maximum amount of P or oxalate that was bound, C is the equilibrium concentration of P/oxalate (mM), and K is a constant related to the









63


binding energy of P or oxalate sorption. The parameters (K and Sj were calculated by a least square fit of the linear form of the equation: C 1 C
[ 3.2
S KS, SM


Speciation Calculations

The metal speciation model MINTEQA2 version 3.2 was used to calculate the species of Al and P in the solution. It was assumed that the system had achieved equilibrium conditions. Although the total activity of Al ions was not known, the assumed activity of Al from the congruent dissolution by oxalate was used (Chapter 2) to estimate the saturation level with respect to an amorphous aluminum phosphate solid phase. Calculations were carried out at solution P concentrations corresponding to the highest loading rate of P (102 M). The formation constants used for calculations of species were those of the MINTEQA2 data base taken from Martell and Smith (1982).



Statistical Analysis

The linear form of the Langmuir equation was used to test the differences between parameters for P sorption. Statistical differences among regression lines for the desorption of Al, Fe OH, and organic carbon during P sorption were tested using the General Linear Models procedure of the SAS framework (SAS Institute, 1985). The model used was:









64


Y pa.X.e .. (3.3)



where a is the intercept, P is the slope of the line and E is the error term. To compare two regression lines, the a and P values of the two lines were compared. ANOVA and t-test were used to test the difference between a, and a2 well as between P1 and P2. The significance of the difference between two means was determined using a t-test (Snedecor and Cochran, 1980). Standard deviations or standard errors of means are given, when appropriate.



Results

Sorption of P

Clay Fractions

Sorption isotherms for phosphate sorption onto clay samples with and

without organic matter are presented in Fig. 3.1. The presence of oxalate and soil organic matter significantly reduced P sorption. Maximum reduction in P sorption (about 50 %) was observed when both organic carbon and oxalate were present in the system. The P sorption data, for the range of concentrations studied, conformed to the Langmuir equation. Sorption parameters for the Langmuir equation are presented in Table 3.1. Whole-Soil Samples

Phosphate sorption isotherms are presented in Figure 3.2. Sorption








65


100 50



20


10


0


1,000 500



200


8


Isotherms for the sorption of P by the clay fractions.


---.---------------/*


T7 M

.9

E





0 0


5



2


P P+ox P P+ox
-OM -OM +OM +OM E A


2 4 6
Equilibrium P concentration (mM)


Fig. 3. 1.









66


Table 3. 1. Langmuir sorption isotherm parameters for P sorption onto the wholesoil samples and clay fractions with and without organic matter.

Langmuir Model

Treatments Anion Sorption Intercept Sorption
Maxima energy
Sm. (1/b) 1/SmK K R2


Soil without organic matter


29. 33 c


P

P


0.051 a


20.08b 0.042a


Soil With Organic Matter
17.39b 0.0353


P

P


13.83a


0.052a


0.66 1.18


1.62 1.39


0.88 0.97


0.98 0 98


Clay without organic matter


588.0d

454.0c


0.0005 0.0022a


Clay with organic matter


379.0b

328.a0a


0.00133

0.0047a


P


P+X


P

P+OX


P


P+X


P

P


3.32 2.42


P


P+OX


0.97 0.99


P

P


1.90

0.64


0.98 0.99


* Values within columns followed by same letter for the whole-soil or clay samples are not significantly different at the 5% level








67



30




25




20


E
E o15


0
0





5'
P only P + Ox P only P + Ox
+OM +OM -OM -OM

0
0 2 4 6 8 10
Equilibrium P concentration (mM)


Isotherms for the sorption of P by the whole-soil samples.


Fig. 3.2









68

parameters for the Langmuir model are presented in Table 3. 1. In the case of the whole-soil samples, P sorption maxima were significantly reduced by soil organic matter. The presence of soil organic matter and oxalate significantly reduced P sorption by both types of soil material.

P sorption parameters of the clay samples also were recalculated on a

whole-soil basis (assuming a clay % for the whole soil = 1.3 %), and compared to the P sorption parameters of whole-soil samples. Significantly more P was sorbed by the whole-soil samples than by the clay samples when reported on a whole-soil basis (Table 3.2). P sorption by whole-soil samples was 3 to 10 times higher than that of the clay samples.



Sorption of Phosphate in the Presence of Oxalate

Less P was sorbed by the clay fractions and whole-soil samples in the presence of oxalate (Table 3.3). The percent efficiency of oxalate in reducing P sorption was calculated according to the expression of Deb and Datta (1967): Pox (3.4)
Oxei-[1---xlOO
P

where Oxe is the efficiency of oxalate in reducing P sorption (%), P,. is the P sorbed in the presence of oxalate, and P0 is the P sorbed alone. The efficiency of oxalate in reducing P sorption increased as the concentration of P in solution increased (Table 3.4).









69


Table 3.2. P sorption onto the whole soil material compared to P sorbed onto
the clay fraction expressed on a whole soil basis (Mean SD; n=3).

Conc of P sorption (mmol kg-')
anions
added + OM OM
Treatment (mM) Whole soil Clay Fraction Whole soil Clay fraction



P alone 10 16.7 0.27a 4.58 0.05b 24.6 0.5a 7.09 0.07b
1.0 5.9 0.13a 0.94 0.01b 7.1 0.06a 1.17g0.01b
0.1 0.9 0.02a 0.12t0.01b 0.8 0.01a 0.12 0.01b


P + Ox 10+1 13.3 0.253 3.52+0.04b 19.3 0.37a 5.95 0.04b
1.0+1 5.3 0.14a 0.79 0.01b 5.8 0.05a 0.96 0.01b
0.1+1 0.78 0.02a 0.090.01b 0.8 0.01a 0.08 0.001b


+ OM = with organic matter and OM is without organic matter @ Values within each row for each material with organic matter and without organic matter are significantly different (P = 0.05) if followed by a different letter









70


Table 3.3. Amount of Al, Fe and organic carbon released into solution with P +
oxalate sorption onto the whole-soil samples and the clay fractiions
(Mean SD; n=3).

With Organic Matter Without Organic matter

Sorbed Released Sorbed __Released
P + Ox Oxalate Al Fe 0. Carbon Oxalate Al Fe
Ratio --- (mmol kg-1)--- (mg g',) -------- (mmol kg-1)------Whole-Soil samples

10:1 4.2 0.11 8.72 0.044 1.44 5.0 0.10 5.01 0.079
7:1 4.8 0.15 6.99 0.034 1.30 5.4 0.08 5.40 0.082
3:1 5.2 0.06 7.33 0.027 1.21 5.9 0.15 5.44 0.087
1:1 5.4 0.10 7.60 0.029 1.12 6.0 0.04 6.26 0.101

0.5:1 5.5 0.11 7.83 0.037 0.97 6.1 0.05 6.49 0.101
0.1:1 5.5 0.08 7.87 0.034 0.62 6.1 0.02 6.64 0.104



Clay fractions

10:1 32.9 1.88 24.45 0.44 166 22.4 2.05 14.19 0.23
1:1 38.8 0.83 18.91 0.31 125 36.7 3.50 14.46 0.31
0.1:1 42.2 0.52 19.45 0.12 117 50.42 1.65 14.59 0.30









71


Table 3.4. Effect of oxalate in reducing the sorption of P onto whole-soil samples
and the clay fractions (Mean SD; n=3).

% reduction in P sorption


With OM


Whole soil


20.7 3.5"


22.5 4.7aa 16.7 5.7a A 10. 8 4.4aa 1.2 1 0. ba

1.2 0.8b


Without OM


21.3 3.5ba 26.6 2. 9a 30.2 3.7" 18.4 0.4ba

-5.3 0.5ed

-0.9 0.3da


Clay fraction


23 2aa 16 1 ba

2 2ca


P:Oxalate ratio


10:1


7:1

3:1

1:1


0.5:1

0.1:1


10:1


1:1


0.1:1


22 0.5a 5 0.5 bb 1c


The first superscript letter indicates significant differences within a column and material, while the second letter indicates significant differences between rows.









72


Release of OH

OH ions were released into the system with the sorption of P alone or P + oxalate (Fig 3.3 and Fig. 3.4 ). For the clay samples, P and oxalate together released more OH into solution than P alone (Table 3.5). The kinetics of OH ions release are presented in Figure 3.5 for the P alone, and P + oxalate, treatments with the clay. Approximately 80% of the OH- release accompanying P sorption was detected in the first 8 hrs. For clay with organic matter, the release of OH- accompanying P sorption was initially rapid and reached equilibrium after 14 hrs.

The molar ratio of OH- released per mole of P sorbed was between 0.85

and 1.38 for the clay fractions (Table 3.6). A similar range was observed for the whole-soil samples. On average, the ratio was not significantly different from I for P sorption alone but was significantly higher than I for sorption of P + oxalate.



Release of Organic Carbon

There was a linear relationship between the amount of organic carbon released and the sorption of anions by the whole-soil samples and the clay fractions (Table 3.5). During sorption of P by the clay or soil, a small amount of organic carbon was released into solution. The presence of oxalate with the P did not significantly affect organic carbon release (Table 3.5). The E4/E6 ratio of released organic matter was approximatly 9.0 during P sorption at different P








73


500 --..- --
500




200


0 100
E
E

50


P P+OX P P+OX 20 -OM -OM +OM +OM


10


5
0 100 200 300 400 500 600
-1
P sorbed (mmol kg )


Release of OH- ions during sorption of P by the clay fractions.


Fig. 3.3













/
/
/


A


/


A-


35 30 25


E

E 0

15



10
0



5 0


5
P or


10 15 20 25
oxalate sorbed (mmol kgl)


Release of OH- ions during sorption of P by the whole-soil samples.


74


A A A A
'A*.


0,0 ..0




*OM -OM+OM +Ox


0


30


Fig. 3.4










75


Table 3.5. Parameters of the linear regression models for the release of OH- ions
and organic carbon from the whole-soil samples and clay fractions.

OH- ions released Organic Carbon released

Treatments Slope Constant R2 Slope Constant R2

Soil without organic matter


P


P + Ox


0.663 0.59"


2.

16.1"


0.89 0.80


Soil with organic matter


P


P + Ox


1.06" 1.19"


2.7'

5.8a


0.95

0.94


0.05 a 0.063


0.5 9a 0.71 a


0.98

0.94


Clay without organic matter 66b 0.99


116'


0.99


Clay with organic matter


6 a


P


0.99 0.96


0.23a

0.22 a


or clay samples are not


P


0.74a


P + Ox


P + Ox


1.08c


67.3a 128.8


Values within columns followed by the same letter for soil significantly different at the 5% level


0.96 0.98

















600




500


E

400 3:
0
0
0 300




0200



E 100


0


Fig. 3.5


P P+OX P P+OX
-OM -OM +OM +OM


5


10


15


20


Time (hrs)

Kinetics of OH- release during sorption of anions by the clay fraction.


76


25









77


Table 3.6. Ratios of OH- released to P sorbed for P only and P + oxalate for
the whole-soil samples and the clay fractions (Mean SD; n=3).

OH released/Anion sorbed

W/OM or WO/OM Treatment Whole soil Clay fraction


With OM


P


P + Ox


Without OM


P


P + Ox


Superscript "a" indicates that values are 5% level.


*1.03 0.05a 1.61 0.10


0.83 0.05a 1.53 0.08


not significantly different from 1.0 at the


0.92 0.03a 1.38 0.08


0.85 0.07a
1.24 0.11









78


loading rates (Table 3.7).



Release of Al and Fe.

A small amount of Al and Fe was released during the sorption of

phosphate and for the P + oxalate sorption treatment, these amounts were even greater (Tables 3.3 and 3.8). Ion-activity product calculations using MINTQEA2 V3.11 showed that the system was supersaturated with respect to amorphous aluminum phosphate at the higher loading rates for P.



Discussion

The dissociation constants (log K, and log K2) of oxalic acid are 1.27 and 4.20. At pH 4.5, both COOH groups of the oxalate will be dissociated, making the C20;2 species dominant. Assuming ligand exchange as the surface reaction this oxalate species can form either monodentate, binuclear, or bidentate surface complexes.

The dissociation of phosphoric acid is described by.


H'p0 FH' 0-1 -2p-3
H4PO H2PO ~HPo P 3.5



where pK, = 2.2, pK2 = 7.2 and pK3 = 12.3. The predominant species of phosphate at pH 4.5 is H,P04. Thus, H2PO41 should only form monodentate or binuclear surface complexes.









79


Table 3.7. E4/E, ratio of organic carbon released into solution during the sorption
of P only, and P + oxalate by the clay fractions (Means SD; n=3).

Concentration of
Treatment Anions added (mM) E4/E, ratio


P 10 9.04 0.10a
1.0 9.31+0.253
0.1 9.24 0.56a

0.0 9.26 0.07a


P + Ox 10 + 1 9.14 0.06a
1.0 + I 8.93 0.05a
0.1 + 1 7.65+0.21b


Values within columns followed by same letter are not 5% level


significantly different at the









80


Table 3.8 Parameters of the linear regression model relating release of aluminum
and iron with P sorption for P only and P + oxalate by the whole-soil
and clay fractions.

Al release Fe release

Treatments Slope Constant R2 Slope Constant R2

Soil with organic matter
P 0.08a 2.05a 0.93 0.001a 0.015a 0.87
P + Ox 0.01a 7.66b 0.07 0.0004a 0.031a 0.24



Soil without organic matter
P 0.01a 2.52a 0.69 0.0001a -0.008a 0.31
P + OX -0.09a 6.72b -0.94 -0.002a 0.106a -0.84



Clay without organic matter
P 0.03a 34.273 0.94 0.0003a -0.01a 0.97
P + OX 0.05a 69.01c 0.94 0.0005a -0.21l 0.96



Clay With Organic Matter
P 0.02a 13.20a 0.96 0.001a 0.19a 0.98
P + ox -0.01a 54.01b -0.61 0.001a 0.52 0.97


Values within columns followed by the same letter for soil or clay samples are not significantly different at the 5 % level









81


P Sorption

Phosphate can be sorbed by the clay and whole-soil surfaces either by

ligand exchange or precipitation. Ligand exchange should result in the release of OH- ions into solution (Goldberg and Sposito, 1985). In fact, large amounts of OH- were released in this study. The amount of OH- ions released varied with the amount of P and/or of oxalate sorbed. The average ratio of OH- released to phosphate sorbed was close to unity. When oxalate was also sorbed this ratio increased to 1.5. Given that the ratio was unity and that the dominant species of phosphate was the H2P04 ion one could argue that P was forming a monodendate and/or binuclear inner-sphere complex. The ratio of OH- ions released to P sorbed was similar for both the whole-soil and clay samples. Recently Tejedor-Tejedor and Anderson (1990) using CIR-FTIR studied the surface complexation of phosphate by goethite surfaces between pH 3.5 to 8.0. They suggested that between pH 3.5 and 5.5, phosphate formed both binuclear and monodendate complexes with surface Fe (III) Earlier work by Parfitt et al. (1977), using infrared spectroscopy, also reported the presence of both binuclear and monodendate surface complexes by P on goethite at pH 4.0.

The molar ratio of OH ions released to P sorbed supports ligand exchange as the dominant P sorption process onto both the clay and whole-soil samples. The formation of measureable crystalline aluminum phosphate was not possible, given the reaction time of this experiment. Veith and Sposito (1977) and Sposito (1984) reported that reaction times longer than 140 hrs are required for significant









82

discrete crystal growth to occur. This is much longer than the 24 hrs represented by this study. Sposito (1984) further suggested that the adsorption-dominated stage of phosphate sorption takes less than about 50 hrs. Calculated ion-activity product constants using MINTQEA2 V3.11 suggested that, at equilibrium, this system was supersaturated with respect to amorphous aluminum phosphate, Al(OH)2H2PO4. The formation or precipitation of amorphous aluminum phosphate thus could be possible. Sorption of phosphate through ligand exchange is known to serve as a nucleation site for the precipitation of amorphous aluminum phosphate onto the clay surface. However, formation or precipitation of amorphous aluminum phosphate (Al(OH)2H2PO4) consumes OH- instead releasing them. Our kinetic data on the release of OH- ions showed no evidence of this. Up to 14 hrs, OH- ions were continuously released into the solution. This suggests a ligand-exchange reaction. After 14 hrs, there was no further release of OH- ions into solution from the clay surfaces with organic matter present while the clay surfaces without organic matter were still releasing OH-. Thus, for clay surfaces, there was no absolute decrease in OH- concentration in solution. This suggests that precipiation of amorphous aluminum phosphate was unlikely. Therefore, these results suggest that ligand exchange was the dominant reaction for P sorption.



Effect of Oxalate and Organic Carbon on P Sorption

Phosphate sorption significantly decreased in the presence of oxalate.

Likewise, P was effective in reducing oxalate sorption (Appendix B), indicating that









83


some of the sorption sites were common for either P or oxalate. These observations suggest that between 20 to 30 % of the sorption sites on the wholesoil material were common. Similar results were reported by other researchers. Competitive sorption studies for P and oxalate on tropical soils (Lopez-Hernandez et al., 1986) and montmorillonite (Kafkafi et al., 1988) showed that oxalate was effective in reducing P sorption. Each observed that oxalate masked about 20% of the sites otherwise available for P sorption.

Soil organic carbon significantly reduced P sorption capacity of both the

clay and whole-soil samples. Only a small amount of organic carbon was released during P sorption, and the E4/E6 values for organic carbon released were high. E4/E6 ratios between 8 to 10 are generally indicative of fulvic acid, and ratios from 2 to 5 represent humic acid (Thurman, 1985). The E4/E6 ratio of released organic carbon following P sorption fit the range for fulvic acid and was similar to that for control samples. Since the ratio of OH released to P sorbed was close to I and the amount of organic carbon released into solution during P sorption was small. This suggests that P was not effective in replacing organic carbon sorbed onto the surfaces. The significant reduction in P sorption by mineral surfaces in the presence of organic matter indicates that organic carbon masked and/or occupied the active sites for P sorption. Characterization of the clay fractions using FTIR (Chapter 2) revealed the presence of -COOH surface groups associated with organic matter. Competitive sorption of soil organic matter and boron by soil material was studied by Marzadori et al. (1991). They observed that soil organic









84

matter appears to be responsible for occluding important adsorption sites on the surfaces of soil particles which otherwise would be available for boron sorption.















CHAPTER 4

INFLUENCE OF SOIL ORGANIC MATTER ON DESORPTION OF
PHOSPHORUS AND OXALATE FROM A SPODIC HORIZON Introduction

In Chapters 2 and 3, the sorption of oxalate and P by the clay fractions and the whole-soil material of a spodic horizon from northcentral Florida were studied. Phosphate and oxalate are specifically sorbed by replacing the coordinated

-OH groups of Al present on the surfaces of oxides and clay minerals. As discussed in Chapter 3, oxalate reduced P sorption onto clay fraction and wholesoil surfaces. This investigation studied P and oxalate desorption from the clay fraction and the whole-soil material. Phosphorus sorbed to the soil surfaces can be taken up by plants after it has been desorbed into the soil solution.

The competitive sorption of organic anions and P on mineral surfaces has received attention because the presence of organic ligands in the rhizosphere is thought to influence P fixation and, therefore, the supply of P to the plant (Deb and Datta, 1967; Lopaz (1974); Appelt et al., 1975; Earl et al., 1979; LopezHernandez et al., 1986, Kafkafi et al., 1988; Martell et al., 1988; Fox et al., 1990a: Fox et al., 1990b; Violante et al., 1991; Violante and Gianfreda, 1993). Oxalate in forest soils originates from exudates of plant roots and through the


85









86

activities of fungi and bacteria (Stevenson, 1982). Although oxalate is abundant in the rhizosphere (Fox and Comerford, 1990), only a few investigations have studied the competitive sorption of oxalate and phosphate. These include studies in tropical soils (Lopez-Hernandez et al., 1986); and montmorillonite (Violante and Gianfreda, 1993). It has been proposed that oxalate can release P from Al- and Fe-hydroxide surfaces through ligand-exchange reactions. Recently, Fox et al. (1990a) studied the kinetics of P desorption by oxalate from spodic and argillic horizons of a Spodosol. They suggested that P released by oxalate was through ligand exchange, but did not provided conclusive evidence.

Phosphate desorption studies (Barrow, 1983; Kuo and Pan, 1988; Bakheit Said and Dakermanji, 1993; Raven and Hossner, 1993) have indicated that a large portion of the retained P is irreversibly sorbed. In all such desorption studies, researchers used either 0.01 M CaC2, Ca(N03)2 or anion exchange resins. Phosphate desorption isotherms do not normally coincide with P sorption isotherms (Nye and Tinker, 1977; Kuo and Pan, 1988; Bakheit Said and Dakermanji, 1993). Lopaz (1974) studied the desorption of P using citrate and found that desorption could only occur when the displacing anion was specifically sorbed and present in sufficient concentration in the soil solution. Organic anions like citrate and oxalate are continuously produced in soil (Smith, 1969) and can be present in high concentrations (Fox and Comerford, 1990). However, desorption isotherms using oxalate as an extractant have not been reported to our knowledge.









87

The influence of soil organic matter on P sorption has been investigated by many researchers. Many have suggested that sorption of P and organic matter occurs on the same sorption sites. Yuan (1980) studied the sorption of P and water-extractable soil organic matter by soil and Al oxides. He reported that most of the sorption sites for organic matter and P sorption may be different, though some sites are common for both P and organic matter. Fulvic acid was found to reduce P adsorption considerably. This effect of fulvic acid is apparently due to the chelating ability of fulvic acid's -COOH and -OH functional groups for Al and Fe (Parfitt, 1978; Sibanda and Young, 1986). The results in Chapter 3 showed that soil organic matter significantly reduced P sorption. However, the mechanism for P desorption in the presence of organic matter apparently has not be studied.

The objectives of this study were to investigate: i) P or oxalate desorption from the clay fraction and whole soil of a spodic horizon in the presence of oxalate or P, respectively, and ii) the influence of soil organic matter on the desorption of P and oxalate.


Materials and Methods

Whole-Soil Material and the Clay Fraction

The whole-soil and clay-fraction samples that were used for sorption studies of oxalate (Chapter 2) and phosphate (Chapter 3) were further used to study the desorption of anions at pH 4.5.









88


Phosphate Desorption by Oxalate

P sorption experiments described in chapter 3 formed the basis for these samples. In Chapter 3, the solutions used for P sorption by the clay samples had P concentrations of 0 mM, 0.1 mM, 1.0 mM, and 10 mM. The P sorption solutions used for the whole-soil samples had P concentrations of 0 mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM and 10 mM. At the end of the reaction period, each suspension was centrifuged at 12,000 g for 20 minutes. Filtrates then were equilibrated with solution containing 5 mM of oxalate per 25 ml of solution for the clay fractions and per 50 ml of solution for the whole-soil samples. The suspensions were then shaken for another 24 hrs. The suspension pH was periodically (every four hours) adjusted to pH 4.5 with 0.1 M HCl or NaOH over the 24 hrs of the experiment. Two drops of toluene were added to inhibit microbial growth, and samples were placed in a reciprocating shaker for 24 hrs. At the end of the reaction period, each suspension was centrifuged at 12,000 g for 20 minutes. The filtrate then was analyzed for oxalate, Al, Fe, inorganic P and organic carbon. Sorbed oxalate was calculated from the difference between the initial and final oxalate concentration in solution.



Oxalate Desorption by Phosphate

The oxalate sorption experiments of Chapter 2 served as the basis for these samples. The solutions used for the initial sorption of oxalate in Chapter 2 had oxalate concentrations of 0 mM, 0.1 mM, 1.0 mM, and 10 mM for the clay




Full Text
139
Harris, W G, V W Carlisle, and K C. J Van Rees 1987a Pedon zonation
of hydroxy-interlayered minerals in Ultic Haplaquods. Soil Sci. Soc. Am. J.
51:1367-1372.
Harris, W. G, V. W. Carlisle, and S. L. Chesser. 1987b. Clay mineralogy as
related to morphology of Florida soils with sandy edipedons. Soil Sci. Soc.
Am J. 51:1673-1677.
Harris, W G., R D Rhue, G. Kidder, R B Brown and R Littell 1994
Rapid assessment of relative phosphorus retention capacity for sandy coastal
plain soil materials. Ecological Engineering (In press)
Haynes, R J and R S. Swift 1989 The effect of pH and drying on
adsorption of phosphate retention by active aluminum and iron of Ando
soils. J. Soil Sci. 40:773-781.
Hingston, F J R J Atkinson, A M Posner, and J P Quick. 1967. Specific
adsorption of anions. Nature 215:1459-1461.
Hingston, F. J., A M Posner, and J P Quick 1972 Anion adsorption by
goethite and gibbsite. I. The role of the proton in determining adsorption
envelopes J Soil Sci. 23:177-192.
Hingston, F J A. M Posner, and J P Quick 1974. Anion adsorption by
geothite and gibsite. II. Desorption of anions from hydrous oxide surfaces.
J Soil Sci. 25:16-26.
Huang, P. M. and M Schnitzer (ed.) 1986. Interactions of Soil Minerals with
Natural Organics and Microbes. Soil Sci. Soc. Am Spec. Publ. 17,
Madison, WI
Huang, P M and A Violante 1986 Influence of organic acids on
crystallization and surface properties of precipitation products of aluminum
p 139-221 In P M Huang and M Schnitzer (ed ) Interactions of Soil
Minerals with Natural Organics and Microbes Soil Sci. Soc. Am Spec
Publ 17, Madison, WI
Hue, N V., G R Craddock, and F Adams 1986 Effect of organic acids on
aluminum toxicity in subsoils. Soil Sci. Soc. Am. J. 50:28-34.
Inoue, T. and K. Wada 1971. Reactions between humified clover extract and
imogolite as a model of humus-clay interactions: Part I and II. Clay Sci.
4:61-70.


7
subsequent hydrolysis of hydroxy-aluminum polymers. The greater the
concentration of organic anions in the system, the greater should be the
replacement of water molecules and the blocking of A1 coordination sites.
Therefore, the occupation of coordination sites by organic ligands would also block
sites which would otherwise be available for the sorption of P Thus, organic
anions, through their reactions with Fe and A1 both in solution and at soil surfaces
should increase the availability of P in soils (Martell et al.. 1988). The activity of
these surface sites available for coordination with organic ligands, and the activity
of ligands in the solution, strongly depend upon pH. The role of pH would be
through i) its effect on the relative fractions of various species of organic anions in
solution (and these species differ in their affinity for the adsorbent); ii) the
variation in charge density of the solid surfaces with pH; and/or iii) competition
between OH' and the organic anion for common sorption sites.
Release of P is possible only if the stability of the metal-organic complex is
higher than that of the solid-phase metal-P complex. The stability of soluble
metal-organic complexes depends on the presence and molecular arrangement of
carboxylic and phenolic functional groups on the organic anion (Pohlman and
McColl, 1986; Martell et al.. 1988). Formation constants can be used to predict
the relative effectiveness of different organic acids for releasing P into solution.
Organic anions such as oxalate and malate have been classified as having high
complexing ability for Al and Fe (Fox et al.. 1990b; Pohlman et al.. 1990) when
they tend to form five- and six-membered rings between the anion and metal.


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy. .
0 -
Eric J. Jofca *
Associate Professor of Forest
Resources and Conservation
This dissertation was submitted to the Graduate Faculty of the College of
Agriculture and to the Graduate School and was accepted as partial fulfillment of
the requirements for the degree of Doctor of Philosophy.
May, 1995
%ac y-4
Dean, College of Agriculture
A^icull
Dean, Graduate School


Al released (mmol kg )
38
Oxalate sorbed (mmol kg )
Fig. 2.11 Release of A1 from the clay fractions at varing pH as a function of
oxalate sorption.


133
Equilibrium concentration of P in solution (mM)
Fig. C-2 Relationship between Kd and the equilibrium concentration of P as
obtained by different methods.
A


84
matter appears to be responsible for occluding important adsorption sites on the
surfaces of soil particles which otherwise would be available for boron sorption.


8
Therefore, a low molecular weight organic anion like oxalate could play an
important role in solubilizing soil Al, therefore indirectly affecting the availability of
phosphorus in forest soils.
Among 16 organic anions studied, a threshold value for log between
4.0 and 4.5 was required before substantial amounts of P were released from a
spodic horizon in Florida (Fox et al.. 1990a). Beyond this log value, release
of Al and inorganic P increased with increasing formation-constant values.
Violante et al (1991) and Violante and Gianfreda (1993) studied the competitive
sorption of phosphate and oxalate on aluminum oxide and montmorillonite. They
reported that oxalate and P ions competed strongly for the same adsorption sites
on Al oxides under both acidic and neutral conditions. Lan et al. (1994) studied
the release of P by oxalate from different spodic horizons of different soils. These
soils desorbed variable amounts of P, although each contained about the same
amount of total P. The differences in P release may have been related to
mechanisms by which the P was held by different soil components (crystalline
minerals, an amorphous fraction, occluded P, or organo-mineral complexes) and/or
the interaction of oxalate with these components.
Oxalate may influence P availability through its reaction with Al both in
solution and at the surfaces of Al oxides as it forms stable complexes with Al
(Martell et al.. 1988). Release of P by oxalate could be by direct ligand exchange
and/or dissolution of Al oxide mineral surfaces (Stumm and Morgan, 1981; Huang
and Violante, 1986). Stumm (1986) proposed a theory of congruent dissolution of


P sorbed (mmol kg )
67
Fig. 3.2
Isotherms for the sorption of P by the whole-soil samples.


90
Oxalate in the extract was determined by HPLC (Fox and Comerford,
1990a), using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton
Co., Reno, Nevada) with a Gilson single piston high pressure pump along with a
Rheodyne model 7125 injection valve fitted with a 20 /uL injection loop. This
HPLC system uses a Gilson Holochrom variable wavelength UV detector in
conjunction with a Gilson computerized integrator. The eluent was 0.005 M
H2S04 at a flow rate of 2 mL min'1. Oxalate concentration was calculated from
acalibration curve obtained with standard solutions of 0.1 to 10 mM.
To determine the amount of OH' released with the sorption of oxalate,
titration curves were prepared for the whole-soil and clay fractions samples (with
and without organic matter) as described in Chapter 3. Models were fitted to
these curves and used to calculate the changes in OH concentration required to
change the pH of the system to specific values.
Adsorption Isotherms
Freundlich model was fitted to the desorption data. The Freundlich model
was obtained by:
S=kC (4
where S and C are defined as previously and k and n are empirical constants.
The linear form of the above equation was used to determine sorption parameters.


119
mm sieve, and stored in plastic bags. Samples preparation for the sorption studies
was described in detail in chapter 2.
Sorption of Oxalate and Phosphate Added as a Mixture
Sorption experiments were carried out at different oxalate to phosphate
molar ratios. The concentration of oxalate used was 1.0 mM, with a phosphate
concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM for the clay samples.
Triplicate samples consisting of 250 mg of clay (with and without organic matter)
were placed in 30 ml bottles along with 25 ml of solution. For the whole-soil
samples, both with and without organic matter, an oxalate concentration of 1.0
mM was used along with phosphate concentrations of 0 mM, 0.1 mM, 0.5 mM,
1.0 mM, 3.0 mM, 7.0 mM or 10 mM. Triplicate five-gram samples of whole-soil
from each treatment (with or without organic matter) were placed in 100 ml
bottles along with 50 ml of solution. pH of the clay suspension and whole-soil
was adjusted initially to 4.5, using 0.1 M HC1 or NaOH
Two drops of toluene were added to inhibit microbial growth and samples
were placed in a reciprocating shaker for 24 hrs. The pH of the suspension was
periodically (every three hours) adjusted with 0.1 M HC1 or NaOH over the 24
hrs of the experiment, to maintain the initial pH. At the end of the reaction
period, each suspension was centrifuged at 12,000 g for 20 minutes. The filtrate
was used for analysis of oxalate. Sorbed oxalate was calculated from the
difference between initial and final oxalate concentrations in solution.


20
Speciation Calculations
The metals speciation model MINTEQA2 version 3.2 was used to calculate
the species of A1 and oxalate in solution, assuming that the system was under
equilibrium conditions. Calculations were carried out at solution concentrations of
oxalate and A1 for the highest loading rate of oxalate (10"2 M oxalate) at different
pH values. Log K values for soluble Al-oxalate complexes were obtained from
Thomas et al. (1991). Log K was assumed to be 6.12 for A1C204+ and 11.15 for
A1(C204)2\ Two solid phase species of Al-oxalate were considered in the
speciation model. The formation constants were taken from Bilinski et al. 1986.
The values of constants used were 21.87 for A13(0H)7(C204).3H20 and 5.61 for
NaAl(0H)2(C204).3H20. Formation constants used for calculations of other species
were those of the MINTEQA2 data base taken from Martell and Smith (1982).
Statistical Analysis
Statistical differences between different regression lines at each pH value
were tested using the General Linear Models procedure of the SAS framework
(SAS Institute, 1985). The model used was:
7,= + pX+e.. (2,4)
where a is the intercept, P is slope of the line, and e is the error term To
compare two regression lines, we compared the a and P values of the respective


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presen: ition and is fully adequate, in scope and
quality, as a dissertation for the degree of Doc: of Philosophy
CgY
Nicholas B Comerford, Chair
Professor of Soil and Water Science
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy
Brian L. McNeal
Professor of Soil and Water Science
I certify that I have read
acceptable standards of scholarly
quality, as a dissertation for the
I certify that I have read
acceptable standards of scholarly
quality, as a dissertation for the
I certify that I have read
acceptable standards of scholarly
quality, as a dissertation for the
this study and that in my opinion it conforms to
presentation and is fully adequate, in scope and
degree of Doctor of Philosophy
Water Science
/A -JaL
^arl L. Stone
Professor of Soil and
this study and that in my opinion it conforms to
presentation and is fully adequate, in scope and
degree of Doctor of Philosophy
tor or Knuosoony
ifWd^T Johnston
ssociate Professor of/Soil and Water
Cliffi
Associate
Science
this study and that in my opinion it conforms to
presentation and is fully adequate, in scope and
degree of Doctor
P S
Graduate Reserach Professor of Soil
and Water Science
of Philosophy.
'Ih>
C Rao


137
Bruckert, S. 1970. Influence des composes organiques solubles sur la
pedogenese en milieu acide. Ann. Agron. 21:421-252.
Buurman, P. 1984. Podzols. Van Nostrand Reinhold Co., New York.
Carlisle, V. W., F. Sodek III, M E. Collins, L C. Hammond, and W G
Harris. 1988. Characterization data for selected Florida soils. Univ. of
Florida, Soil Sci. Dep Res Rep 88-1, Gainesville, FL.
Comerford, N. B and M. F. Skinner. 1989. Residual phosphorus solubility for
an acid, clayey, forested soil in the presence of oxalate and citrate. Can. J.
Soil Sci. 69:111-117.
Comerford N. B G. Kidder and A. V. Mollitor. 1984. Importance of subsoil
fertility to forest and non-forest plant nutrition, p 381-404. In E. L.
Stone (ed.) Forest soils and Treatment Impacts. University of Tennessee.
Knoxville TN
Cooperband, L. R. and T. J. Logan. 1994. Measuring in situ changes in labile
soil phosphorus with anion-exchange membrabes. Soil Sci. Soc. Am. J.
58:105-115.
Cromack, K P Sollins, W. C. Graustein, C. Y Li, and R. L. Todd 1979.
Calcium oxalate accumulation and soil weathering in mats of the hypogeous
fungus. Soil Biol. Biochem. 11:463-469.
Davis, J. A. and R Glour. 1981. Adsorption of dissolved organic in lake water
by aluminum oxide. Environ. Sci. Technol. 15:1223-1229.
De Coninck, F. 1980. Major mechanism in formation of spodic horizons.
Geoderma 24:101-128.
Deb, D L. and N. P Datta 1967. Effect of associating anions on phosphate
retention in soil: I Under variable phosphorus concentrations. Plant Soil
26:303-316.
Duquette, M. and W. Hendershot. 1993. Soil surface charge evaluation by back-
titration: I. Theory and method development. Soil Sci. Soc. Am. J.
57:1222-1228.
Earl, K. D J. K. Syers, and J. R McLaughlin 1979. Origin of the effects of
citrate, tartrate and acetate on phosphate sorption by soils and synthetic
gels. Soil Sci. Soc. Am. J. 43:674-678.


33
Fig. 2.9 Release of OH ions into solution during the sorption of oxalate by
the whole-soil samples.


LIST OF FIGURES
Figure Page
1.1. Three possible phosphate surface complexes 4
2.1 FTIR spectra of clay (a) with organic matter and
(b) without organic matter 22
2.2 Isotherms for oxalate sorption by clay fractions at
different pH values 24
2.3 FTIR spectra of clay without organic matter after oxalate
sorption at pH 3.5 (A), 4 5 (B), and 5.5 (C) 26
2.4 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 3.5 27
2.5 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 4.5 28
2.6 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 5.5 29
2.7 Isotherms for the sorption of oxalate by the whole-soil samples. .31
2.8 Release of OH ions by clay fractions at varing pH as
a function of oxalate sorption 32
2.9 Release of OH ions into solution during the sorption of
oxalate by the whole-soil samples 33
2.10 Kinetics of OH ions released during the sorption of oxalate
by the clay fractions at pH 4.5 35
2.11 Release of A1 from the clay fractions at varing pH as a
function of oxalate sorption 38
viii


66
Table 3.1. Langmuir sorption isotherm parameters for P sorption onto the whole-
soil samples and clay fractions with and without organic matter.
Treatments
Anion
Langmuir Model
Sorption Intercept
Maxima
SMax (1/b) 1/SmK
Sorption
energy
K
R2
Soil without organic
matter
P
P
29.33
0.051a
0.66
0.88
P+OX
P
20.08b
0.042a
1.18
0.97
Soil With Organic Matter
P
P
17.39b
0.035a
1.62
0.98
P+OX
P
13 83a
0.052a
1.39
0.98
Clay without organic
matter
p
P
588.0d
0.0005a
3.32
0.97
P+OX
P
454.0
0.0022a
2.42
0.99
Clay with organic i
natter
p
P
379.0b
0.0013a
1.90
0.98
P+OX
P
328.0a
0.0047a
0.64
0.99
* Values within columns followed by same letter for the whole-soil or clay
samples are not significantly different at the 5% level


132
Using the Freundlich isotherms, the partition coefficients (Kd) also were
calculated. The plot of Kd vs solution P is presented in Fig C-2. The results of
this study show that the two desorption methods differed significantly. The
amount of P desorbed by the dilution method under constant and variable pH was
less than for the sequential method. This resulted in higher partition coefficient
(Kd) values for P being calculated by the dilution method.


10
examine the influence of natural soil organic carbon on the retention and release of
organic anions and P in Spodosols.
A significant portion of the total P in a spodic horizon is inorganic, but not
readily available for plant use. Phosphorus uptake by plants is influenced by the
amount of root surface area, root growth rate, and P concentration in solution at
the root surface (Barber, 1984). Phosphorus supplying capacity of a soil is
determined not only by the amount of P in the soil solution but also by the soil's
ability to replenish P lost from soil solution and/or during transport of P to plant
roots via diffusion and/or mass flow. The soil's ability to replenish P lost from
the soil solution depends partially upon release of P from sorbed and insoluble P
Thus, it is important to understand how P is being held by different colloidal
surfaces as well as the influence of various chemical processed on the release of
previously sorbed and insoluble P.
From the aforementioned, it can be concluded that there is a need to
understand how forest trees acquire nutrients from the soil profile in lower Coastal
Plains. We can start to address this question by i) understanding the types of
sites available and binding mechanisms for oxalate via spectroscopic studies; ii)
investigating P sorption and desorption in the presence of oxalate in spodic
horizons; and iii) defining the influence of soil organic colloids on the retention
and release of P and oxalate. Understanding the sorption mechanisms for P in the
presence of oxalate should form the basis for explaining P extraction by roots from
soil surfaces, and the buffering of P levels in soil solutions.


72
Release of OH'
OH' ions were released into the system with the sorption of P alone or P
+ oxalate (Fig 3.3 and Fig. 3.4 ). For the clay samples, P and oxalate together
released more OH' into solution than P alone (Table 3.5). The kinetics of OH'
ions release are presented in Figure 3.5 for the P alone, and P + oxalate,
treatments with the clay. Approximately 80% of the OH' release accompanying P
sorption was detected in the first 8 hrs. For clay with organic matter, the release
of OH" accompanying P sorption was initially rapid and reached equilibrium after
14 hrs.
The molar ratio of OH' released per mole of P sorbed was between 0.85
and 1.38 for the clay fractions (Table 3.6). A similar range was observed for the
whole-soil samples. On average, the ratio was not significantly different from 1
for P sorption alone but was significantly higher than 1 for sorption of P +
oxalate
Release of Organic Carbon
There was a linear relationship between the amount of organic carbon
released and the sorption of anions by the whole-soil samples and the clay
fractions (Table 3.5). During sorption of P by the clay or soil, a small amount of
organic carbon was released into solution. The presence of oxalate with the P did
not significantly affect organic carbon release (Table 3.5). The E4/E6 ratio of
released organic matter was approximatly 9.0 during P sorption at different P


125
removal of P by anion exchange resin; i) release from the soil; ii) transport
through the soil solution; and iii) sorption by the resin. In a recent study of P
desorption in the field by Cooperband and Logan (1994) with an anion exchange
membrane (AEM), it was shown that the AEM measured the net change and not
the total change in P for the time period during which the resin is in contact with
the soil. They stated that P was continuously adsorbed and desorbed from the
membrane surface as a function of the soil chemical-biological microenvironment.
Soil pH has a strong effect on P sorption and desorption (Barrow, 1983)
and P desorption leads in turn to a decrease in soil pH Small differences in pH
can have a significant influence on the P species present in solution. An
equilibrium between the particular P species in solution and solid phase P controls
the quantity of phosphate desorbed into solution. In most of the studies relating
to P desorption, pH was not measured nor maintained constant (Logan, 1982;
Menon et al., 1989; Graetz and Nair 1994; Harris et al.. 1994; Raven and
Hossner, 1994).
The objectives of this investigation were to: i) compare the P desorption
isotherm obtained by a sequential extraction technique with that by a dilution
method; and ii) compare P desorption underboth constant and variable pH
conditions by both methods.


113
3. Since oxalate significantly influenced the sorption and desorption of
inorganic P, its influence on the sorption and desorption reactions of
organic P compounds also needed to be investigated.
4. Characterization of the organic carbon released into solution, using NMR
and FTIR techniques to understand its nature and influence on nutrient
availability.


25
Table 2.1 Parameters of the linear regression models for oxalate sorption by the
clay fractions and the whole-soil samples.
Organic matter
pH
Oxalate sorption
Slope
Intercept
R2
With Organic Matter
3.5
Clav Fractions
64.3
21.3a
0.99
4.5
30.9e
17.8a
0.97
5.5
15.9f
13.6a
0.96
Without Organic Matter
3.5
209.0a
20.4a
0.99
4.5
101 4b
11.9a
0.99
5.5
46.5d
9.2a
0.99
With Organic Matter@
36.8b
17.6a
0.79
Without Organic Matter1
119 Ia
13.9a
0.77
Whole Soil Samples
With Organic Matter
4.5
0.070b
0.044a
0 94
Without Organic Matter
4.5
0.095a
0.036a
0.95
The same letter in a column at different pH values indicates a lack of significance
at the 5 % level within the clay fractions and the whole-soil samples.
@ Includes the combined samples at pH 3.5, 4.5, and 5.5;


88
Phosphate Desorption by Oxalate
P sorption experiments described in chapter 3 formed the basis for these
samples. In Chapter 3, the solutions used for P sorption by the clay samples had
P concentrations of 0 mM, 0.1 mM, 1.0 mM, and 10 mM. The P sorption
solutions used for the whole-soil samples had P concentrations of 0 mM, 0.1
mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM and 10 mM. At the end of the
reaction period, each suspension was centrifuged at 12,000 g for 20 minutes.
Filtrates then were equilibrated with solution containing 5 mM of oxalate per 25
ml of solution for the clay fractions and per 50 ml of solution for the whole-soil
samples. The suspensions were then shaken for another 24 hrs. The suspension
pH was periodically (every four hours) adjusted to pH 4.5 with 0.1 M HC1 or
NaOH over the 24 hrs of the experiment. Two drops of toluene were added to
inhibit microbial growth, and samples were placed in a reciprocating shaker for 24
hrs. At the end of the reaction period, each suspension was centrifuged at 12,000
g for 20 minutes. The filtrate then was analyzed for oxalate, Al, Fe, inorganic P
and organic carbon. Sorbed oxalate was calculated from the difference between
the initial and final oxalate concentration in solution.
Oxalate Desorption by Phosphate
The oxalate sorption experiments of Chapter 2 served as the basis for these
samples. The solutions used for the initial sorption of oxalate in Chapter 2 had
oxalate concentrations of 0 mM, 0.1 mM, 1.0 mM, and 10 mM for the clay


145
Trana, S J, G Sposito, D Hesterberg, and U Khafkafi. 1986a Effect of
pH and organic acids on orthophosphate solubility in an acidic,
monmorillonitic soil. Soil Sci. Soc. Am. J. 50:45-52.
Trana, S I, G Sposito, D Hesterberg, and U. Khafkafi 1986b Effect of
ionic strength, calcium, and citrate on orthophosphate solubility in an acidic,
monmorillonitic soil. Soil Sci. Soc. Am. J. 50:623-627.
Ugolini, F.C., R Dahlgren, S Shoji and T. Ito. 1988 An expample of
andosolozation and podzolization as revealed by soil solution studies,
southern Hakkoda, Northeastern Japan. Soil Sci. 145:11-125.
Van Rees, K.C.J. and N. B. Comerford. 1986. Vertical root distribution and
strontium uptake of a slash pine stand on a Florida Spodosol. Soil Sci.
Soc. Am J. 50:1042-1046.
Van Rees, K.C.J., N B Comerford, and W McFee 1990 Modeling
potassium uptake by slash pine seedling from low K supplying soils of the
Southeastern Coastal Plain. Soil Sci. Soc. Am. J 49:1413-1421.
Vedy, J. C. and S. Bruckert. 1982. Soil solution: Composition and pedogenic
significance, p 184-213 In Bonneau, M. and Souchier (eds): Constituents
and Properties of Soils. Acdemic Press, London, .
Veith, J. A. and G. Sposito. 1977. Reactions of aluminum-solicates, aluminum
hydrous oxides, and aluminum oxides with o-phosphate: The formation of x-
ray amorphous analogs of variscite and montebrasite. Soil Sci. Soc. Am J.
41:870-876.
Violante, A. and L. Gianfreda. 1993. Competition in adsorption between
phosphate and oxalate on an aluminum hydroxide montmorillonite complex
Soil Sci. Soc. Am. J. 57:1235-1241
Violante, A., C. Colmbo and A. Buondonno. 1991. Competitive adsorption of
phosphate and oxalate by aluminum oxides. Soil Sci Soc. Am. J 55:65-
70.
Viotante, A. and P M. Huang. 1984 Nature and properties of pseudoboehmites
formed in the presence of organic and inorganic ligands Soil Sci. Soc.
Am. J. 48 1193-1203


CHAPTER 1
GENERAL INTRODUCTION
Phosphorus in Flatwood Spodosols
Phosphorus (P) deficiencies are common for the poorly-drained Spodosols of
the flatwoods of the lower Coastal Plain of the Southeastern United States
(Comerford et al., 1984). Cycling of P, which occurs through mineralization,
immobilization and redistribution of P in soil, depends on its physico-chemical
properties. These include P sorption by colloidal surfaces, as well as microbial,
mycorrhizal and plant uptake of P (Stewart and Tiessen, 1987). Soil phosphorus
can be divided into three general categories: i) ions and compounds in the soil
solution, ii) ions sorbed onto or incorporated into the surfaces of inorganic
constituents, and iii) ions that are components of soil organic matter
Inorganic P in the soil solution of a Spodosol's A horizon is replenished by
mineralization of organic P, phosphorus leaching from the forest floor, and
sometime, by fertilizer application (Polglase et aL 1992). Mineralogy of the
surface soil and bleached E horizon is dominated by quartz (Carlisle et al.. 1988).
Therefore, mineralization of organic matter, plays an important role in P availability.
In many of these soils, mineral components have virtually no P retention capacity
1


17
concentrations of 0 mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM or 10 mM
adjusted initially to pH 4.5 (using 0.1 M HC1 or NaOH).
Two drops of toluene were added to inhibit microbial growth and all
samples were placed in a reciprocating shaker for 24 hrs. The pH of the
suspension was periodically (every three hours) adjusted with 0.1 M HC1 or
NaOH over the 24 hrs of the experiment, to maintain the initial pH. At the end
of the reaction period, each suspension was centrifuged at 12,000 g for 20
minutes. The adsorbate was used for FTIR analysis and the filtrate was analyzed
for oxalate, Al, Fe, and organic carbon. Sorbed oxalate was calculated from the
difference between the initial and final oxalate concentration in solution.
Chemical Analysis
Solution pH was measured using a combination glass electrode and an
Orion pH meter. Aluminum was determined via flame emission spectrophotometer
with a N20-C,H, flame. Fe was also determined, with an acetylene flame. Total
organic carbon content of the clay suspension (with organic matter) was measured
via persulfate oxidation and IR analysis of CO, produced on a TOC apparatus
(College Station TX). The E4 (A=465) to E6 (A=665) ratio of organic carbon
released in solution by clay with organic matter was measured using a Shimadzu
UV/Vis spectrophotometer.
Oxalate in the extract was determined by HPLC (Fox and Comerford,
1990a) using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton


5
nonproptonated bidentate bridging, and a nonprotonated monodentate complex. The
speciation of these complexes was a function of pH, as pH determines the species
of phosphate in solution. They also identified a monodentate complex on the
geothite surfaces when the iron/phosphate ratio was less than unity and pH was
between 3.5 and 6.0. At higher pH (6.0 to 8.3) and higher concentrations of
phosphate in the system, surface complexes were bidentate. Therefore, the type of
inner-sphere complex in different soils may be determined by the pH and the
concentration of available P in soil solution. Nanzyo (1987) studied sorbed
phosphate species using diffuse reflectance infrared spectroscopy (DRIR) and found
that phosphate reacts not only with surface sites but also with the structural
aluminum of allophanic soils. He reported that P can also precipitate as
noncrystalline aluminum phosphate. Most of the studies, however, that have been
carried out on the mechanism of P retention and release have used pure Fe & A1
oxides as the solid matrix. None of these studies were accomplished using soil
materials, where different soil components may coexist.
Organic Anions in Soil
A wide variety of low molecular weight organic anions have been identified
in forest and agricultural soils (Stevenson 1982). The commonly found organic
anions include oxalate, citrate, formate, acetate, malate, maleate, lactate and
fumarate (Gardner et al.. 1982; Hue et al.. 1986; Pohlman and McColl, 1988).
Fox and Comerford (1990) reported that oxalate comprised 60 to 80 percent of


9
mineral surfaces by low molecular organic anions. He suggested that organic
ligands form complexes with A1 and Fe both in solution and on the surfaces of
oxides. The formation of complexes in solution by organic ligands increases the
rate of dissolution of Al-oxalate complexes on surfaces. Ligand exchange at the
surface decreases the strength of A1-OH-A1 bonds and helps bring Al-oxalate
complexes into solution. This rupture is thought to be induced by the competition
of oxalate with the hydroxyls linking the aluminum atoms. Stumm (1986) further
proposed that surface complexation is a prerequisite for nucleation for most solid
crystals formed in solution as well However, there is no experimental evidence in
the literature to support the above theory, even with the use of pure clay minerals
or A1 and Fe oxides
The complexation and sorption reactions of humic and fulvic acids with A1
and Fe are similar in magnitude to those for the corresponding monomeric units
(e g. carboxylate, salicylate, dihydroxybenzoate) believed to contain functional
groups similar to those in fulvic and humic acid (Stumm and Morgan, 1981;
Stevenson and Fitch, 1986; Tan, 1986). Inoue and Wada (1971) concluded that
the adsorption of fulvic and humic acids by allophane and imogolite resulted from
the ligand exchange of oxygen in carboxylic groups with structural oxygen in the
coordination shells of A1 and Fe atoms. This suggests that organic carbon may
block some of the adsorption sites which might otherwise be available for the
sorption of P and organic anions. There are no studies to my knowledge which


14
A1(0H)3(S) 4= Alq(OH)pLr(S) Alq(OH)pLr(aq) 4= AlLn(aq) 4= Al3+(aq) (2.1)
where S represents the mineral surface, L is the organic ligand, aq is the solution
phase and p, q and r are the number of atoms in each molecule. The calculations
of soil solution speciation with kaolinite as the solid phase, showed that, under
acidic conditions, solution concentration of A1 would increase in the presence of
oxalate. This increase was attributed to the formation of [A1L]T and [A1L2]' species
in solution. Fox et al. (1990a) observed higher release of A1 into solution with
the adsorption of oxalate, and attributed this to the dissolution of solid-phase Al
surfaces in Spodosols. Bilinski et al. (1986) found that the binuclear complex
A12(0H)2(C204)44' formed in solution when oxalate concentration was high, and then
precipitated as Al oxalate with the oxalate ion acting as a bridging ligand between
two aluminum atoms.
The objectives of the present study were to describe i) oxalate sorption on
the clay fraction of a spodic horizon, ii) the effect of soil pH on oxalate sorption
capacity, and iii) the influence of soil organic matter on oxalate sorption using
fourier transform infrared (FTIR) spectroscopy.
Materials and Methods
Soil Material
Soil was collected from the spodic horizon (Bh) in a soil pit of a Pomona
series (sandy, siliceous, hyperthermic Ultic Alaquod) at the Gator Nationals Forest
site located in Alachua County, 10 km northeast of Gainesville, Florida (Swindel et


:9
10J M Oxalate
1CT1 M Oxalate
Fig. 2 6
FTTR spectra of day with organic matter for different concentrations
of oxalate sorbed at pH 3.3.


97
Table 4.2. Quadratic equation coefficients for the sorption of anions by whole-soil
samples and clay fractions.
Anion
Sorbed
Materials
a
P,
P2
R2
OX
Clay + OMa
160
76.5
-39.4
0.94
Clay OMb
378
-24.9
2.8
0.95
Soil + OMa
11.4
5.63
-10.7
0.92
Soil OMb
14.5
3.67
-7 66
0.97
P
Clay + OMa
283
-76.6
17.5
0.85
Clay OMb
213
-3.89
-3.7
0.74
Soil + OMa
7.1
1.74
-4.76
0.87
Soil OMb
12.6
1.32
-2.41
0 82
+ OM= With organic matter and OM = Without organic matter
Different letters in the same column are significantly different at the 5% level for
the clay and soil samples.


59
Materials and Methods
Soil Material
Soil was collected from the spodic horizon (Bh) of a single soil pit of a
Pomona series (sandy, siliceous, hyperthermic, Ultic Alaquod) at the Gator
Nationals Forest site located in Alachua County, Florida. The soil material was
air-dried, passed through a 2-mm sieve and stored in plastic bags. Particle-size
analysis of the sample was carried out using standard methods (Page et al.. 1986).
The soil contained 91 % sand, 8 % silt and 1 % clay.
Preparation of Whole-Soil Sample and the Clay Fractions
The prepartion of whole-soil samples for the sorption experiment was
previously described in Chapter 2. The water-dispersible clay fraction was
separated (Genrich and Bremner, 1974) as reported in Chapter 2. Both the whole-
soil and clay samples were dried at 110 C.
Sorption Studies on Clay Fraction and Whole Soil
Sorption of Phosphate Alone
Sorption experiments were carried out over a range of phosphate
concentrations at pH 4.5 using a batch procedure. Triplicate 250 mg samples of
clay (with and without organic matter) were placed in 30 ml bottles with 25 ml of
solution having a phosphate concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM
adjusted initially to pH 4.5 (using 0.1 M HC1 or NaOH). Triplicate five-gram


94
Table 4.1. Parameters of Freundlich models for the desorption of P from whole-
soil samples and clay fraction in
the presence of 5
mM oxalate.
Type of Surfaces Desorption Constant Bonding Energy
K N
R2
Whole Soil
With Organic matter
10.59a
0 98a
0.85
Without Organic matter
18 08a
1 23a
0.86
Clav Fractions
With organic matter
354a
0.94a
091
Without organic matter
301a
1.03a
090
Letters in column are significantly different at the 5% level for clay and soil
materials.


P remained sorbed (mmol kg )
130
Fig. C-l. P desorption curves by dilution and extraction methods at both
variable and constant pH.


106
nucleei for subsequent precipitation.. However, the log K of soluble aluminum
phosphate species in solution (A1H2P04(0H)2 log K =3.0) is low compared those
to the possible aluminum oxalate species in solution [A1C204" (6.5) or A1(C204)2'
(13.7)]. Therefore, oxalate should form stable soluble complexes with A1 as
A1C204 or A1(C204)'. The speciation calculations also suggested that less than
10% of the P in solution was present as aluminum oxalate phosphate complexes (
A1H2P04C204), with the majority of the P being in the form of H2P04 species.
Thus, one would expect P desorbed into solution by oxalate to be in the H2P04
form.
Oxalate Desorption by Phosphate
The small amount of oxalate that was desorbed by P indicates that oxalate
ions were held very strongly by the clay particle and whole-soil surfaces. Oxalate
forms bidentate surface complexes ( Chapter 2, Parfitt et al.. 1977). P forms
monodentate or binuclear complexes (Chapter 3; Parfitt, 1978; Tejador Tejador and
Anderson, 1990). Since oxalate forms a stronger complex, P could not replace the
sorbed oxalate. Our data suggest that the ratio of OH' released to P sorbed by
the whole-soil samples and the clay fractions in this study was not significantly
different from unity. This confirmed that P was sorbed onto the clay fractions and
whole-soil surfaces as monodentate and/or binuclear surface complexes, and that P
was sorbed by sites on the clay and soil surfaces not already occupied by oxalate.


with soil organic matter significantly reducing the sorption of both P and oxalate.
Maximum reduction in P sorption (about 50%) was observed when both organic
carbon and oxalate were present in the system. Ligand exchange apparently was
the dominant mechanism for P and oxalate sorption. Some of the sorption sites
were common for both the anions. Oxalate and P formed different types of
surface complexes, with pH determining the type of surface complex formed by
oxalate. Oxalate formed monodentate and/or binuclear surface complexes at pH
3.5, while bidentate complexes at pH 4.5 and 5.5. P formed monodentate and/or
binuclear surface complexes at pH 4.5.
Significant amounts of organic carbon, A1 and Fe were released into
solution during the sorption of oxalate. This study provided the first experimental
evidence for Stumm's theory of congruent dissolution of minerals by oxalate.
Oxalate desorbed large amounts of P into solution. The presence of organic
matter further increased the amount of P desorbed. Oxalate appeared to desorb
the P both through ligand exchange and surface dissolution. This could increase
the initial P concentration in solution and effect the P buffer power. Since oxalate
formed stronger complexes on the mineral surfaces P failed to desorb the oxalate
xii
into solution.


110
P and Oxalate desorption
11. P desorption from the spodic horizon by oxalate appeared to be through
two processes: i) ligand exchange replacing P from the mineral surfaces;
and ii) oxalate forming surface complexes through ligand exchange
(replacing OH ions), dissolving the mineral surfaces, and releasing P into
solution.
12. Soil organic matter and oxalate significantly increased P desorption.
13. Once P was released into solution, oxalate did not allow P to
reprecipitate by forming stable and soluble complexes with A1 and Fe.
14 Oxalate was not released into the solution from spodic horizon material
during P sorption.
15. Some of the sites were highly specific for P sorption.
Influence of Oxalate on the P Nutrition of Trees
P availability often limits forest productivity in flatwood soils. These
results provide information about the availability of P in a spodic horizon.
Phosphorus is absorbed by plants largely as soluble H2P04' or HP042' species.
A soil's ability to supply P is determined by the concentration of P in the soil
solution, along with its ability to replenish any P lost from solution. Our
results confirm that oxalate significantly increases the solution concentration of P
by: i) reducing P sorption; ii) releasing P into solution from sorbed and
insoluble pools; and iii) complexing Al, Fe or Ca in solution, thereby reducing


P retained by soil (mmol kg ) P retained by clay (mmol kg )
93
Fig. 4.1
P desorption curves for whole-soil samples and clay fraction.


98
sorption was significantly higher for materials without organic matter than with
organic matter.
Release of Hydroxyls
Hydroxyls were released into solution during the sorption of oxalate by both
the whole-soil and clay fraction (Appendix D-l). The ratios of OH' desorbed to
the amount of oxalate sorbed and are presented in Table 4.3. This ratio ranged
between 1.68 and 1.90 for the clay fraction and between 1.64 and 2.01 for the
whole-soil samples. The ratio of OH' desorbed to oxalate sorbed increased as the
amount of P desorbed in solution decreased.
For the clay fraction, ratios of OH + P released into solution to oxalate
sorbed were not significantly different from 2. Ratio was significantly higher than
2 for the whole-soil samples, however, with higher amounts of P previously present
on these surfaces.
Release of Al, Fe, and Organic Carbon
Considerable amounts of aluminum, iron and organic carbon were released
into the solution during oxalate sorption, both for the whole-soil samples and the
clay (Appendix D-l). There was a significant correlation between the amounts of
A1 and Fe released and oxalate sorbed.


71
Table 3.4. Effect of oxalate in reducing the sorption of P onto whole-soil samples
and the clay fractions (Mean SD; n=3).
P: Oxalate
ratio
% reduction
in P sorption
With OM
Without OM
Whole soil
10:1
20.73.5aa
21.33.5ba
7:1
22.54.7aa
26.62.9ba
3:1
16.75.7ab
30.2*3.7
1:1
10.8*4.4
18.40.4ba
0.5:1
1.21.0ba
-5.30.5cb
0.1:1
1.20.8ba
-O.OiOJ1*3
Clav fraction
10:1
2 3 2
22*0.5
1:1
16lba
5*0.5bb
0.1:1
22ca
llca
The first superscript letter indicates significant differences within a column and
material, while the second letter indicates significant differences between rows.


42
Release of Organic Carbon
Organic carbon was released from the clay fractions and whole-soil samples
in a linear fashion relative to oxalate sorption (Fig. 2.14 and Fig. 2.15). The
amount of organic carbon desorption is presented in table 2.2. Seven to ten times
more organic carbon was desorbed by 10 mM oxalate as compared to water. The
pH of the system also greatly influenced organic carbon desorption, with
significantly more organic carbon released at pH 5.5 than at pH 3.5 (0.08 mg g'1
versus 0.249 mg g'1 ). E4/E6 ratios of the organic carbon released into solution
during sorption of oxalate are presented in Table 2.5. The E4/E6 ratio, at the
highest loading rate of oxalate (10 mM), decreased to 1.88. There was a linear
relationship (R2 = 0.99) between organic carbon release and A1 released at pH 3.5,
pH 4.5 and 5.5 for the clay fractions (Figs. 2.16 and 2.17).
Spectra for the clay fractions with organic matter showed major absorption
bands representing organic carbon in the regions of 3300 cm'1 (O-H and N-H
stretching), 2960/2925 cm'1 (aliphatic C-H stretching of CH3 and -CH2 groups),
1610 cm"1 (aromatic C=C and/or H-bonded OO stretching of COOH), and 1400
cm'1 (OH deformation, C-0 stretching of phenolic OH, and C-H deformations of
CH3 and CH2,) Intensities of the COOH and CH3 bands in the spectral region
1610 cm'1 and 1400 cm'1 were calculated at all pH values and at all equilibrium


Amount of 0.1 N NaOH used
APPENDIX A
TITRATION CURVES FOR THE CLAY FRACTION AND WHOLE-SOIL
MATERIAL
pH
Fig. A-l Titration curves for the clay fraction: A) With organic matter, B)
Without organic matter.
114


Al released (mmol kg )
39
Oxalate sorbed (mmol kg^)
Fig. 2.12 Release of A1 from the whole-soil samples as a function of oxalate
sorption


51
then we would expect to see several things from these data. First, there should
be a release of OH ions accompanying oxalate sorption (Goldberg and Sposito,
1985). This would not be the case if precipitation of oxalate dominated. Second,
there should be a linear relationship between the amount of oxalate sorbed and OH
ions released into solution. Third, the molar ratio of OH ions released to oxalate
sorbed should be approximately 1 or 2. Depending upon pH, a ratio of one
would suggest a monodentate or binuclear inner-sphere complex, while a ratio of 2
would suggest a bidentate inner-sphere complex. We would expect a ratio of 1 at
pH 3.5, because only one functional group of oxalate is dissociated at this pH and
the formation of monodentate and/or binuclear surface complexes is expected At
pH 4.5 and 5.5 we would expect ratios around 2, suggesting that inner-sphere
complexes are almost exclusively bidentate.
If precipitation of oxalate were the dominant process we would not expect
to see a strong relationship between A1 and Fe release into solution and the
amount of oxalate sorbed as well as a continuous release of OH' ions into solution
with oxalate sorption, even after 24 hrs The MINTEQA2 speciation calculations
indicate that the amount of solid phase species of polynuclear aluminum hydroxy-
oxalate complexes (A13(0H)7(C204) 3H20 and NaAl(0H)7(C204) 3H20) were
insignificant. This implies that precipitation was not the process controlling oxalate
sorption in this system. However, this criterion has to be used with caution, as
these calculations are based on the assumption that system was at equilibrium
which has not been shown to exist in this study.


70
Table 3.3. Amount of Al, Fe and organic carbon released into solution with P +
oxalate sorption onto the whole-soil samples and the clay fractiions
(Mean SD; n=3).
With Organic Matter
Without Organic matter
Sorbed
Released
Sorbed
Released
P + Ox Oxalate
A1
Fe
O Carbon
Oxalate
A1
Fe
Ratio
(mmol kg"1)
(mg g1)
(mmol kg'1)-
Whole-Soil samples
10:1
4.20.11
8.72
0.044
1.44
5.00.10
5.01
0.079
7:1
4.80.15
6.99
0.034
1.30
5.40.08
5.40
0.082
3:1
5.20.06
7.33
0.027
1.21
5.90.15
5.44
0.087
1:1
5.40.10
7.60
0.029
1.12
6.00.04
6.26
0.101
0.5:1
5.50.11
7.83
0.037
0.97
6.10.05
6.49
0.101
0.1:1
5.50.08
7.87
0.034
0.62
6.10.02
6.64
0.104
Clav fractions
10:1
32.91.88
24.45
0.44
166
22.42.05
14.19
0.23
1:1
38.80.83
18.91
0.31
125
36.73.50
14.46
0.31
0.1:1
42.20.52
1945
0.12
117
50.421.65
14.59
0.30


83
some of the sorption sites were common for either P or oxalate. These
observations suggest that between 20 to 30 % of the sorption sites on the whole-
soil material were common. Similar results were reported by other researchers.
Competitive sorption studies for P and oxalate on tropical soils (Lopez-Hernandez
et al., 1986) and montmorillonite (Kafkafi et al., 1988) showed that oxalate was
effective in reducing P sorption. Each observed that oxalate masked about 20% of
the sites otherwise available for P sorption.
Soil organic carbon significantly reduced P sorption capacity of both the
clay and whole-soil samples. Only a small amount of organic carbon was released
during P sorption, and the E4/E6 values for organic carbon released were high.
E4/E6 ratios between 8 to 10 are generally indicative of fulvic acid, and ratios from
2 to 5 represent humic acid (Thurman, 1985). The E4/E6 ratio of released organic
carbon following P sorption fit the range for fulvic acid and was similar to that
for control samples. Since the ratio of OH released to P sorbed was close to 1
and the amount of organic carbon released into solution during P sorption was
small. This suggests that P was not effective in replacing organic carbon sorbed
onto the surfaces. The significant reduction in P sorption by mineral surfaces in
the presence of organic matter indicates that organic carbon masked and/or
occupied the active sites for P sorption Characterization of the clay fractions using
FTIR (Chapter 2) revealed the presence of -COOH surface groups associated with
organic matter. Competitive sorption of soil organic matter and boron by soil
material was studied by Marzadori et al. (1991). They observed that soil organic


Sorbed oxalate (mmol kg )
24
3
+ OM 3.5
+ OM 4.5
+ OM 5.5
B
--A--
-G--
- OM 3.5
- OM 4.5
- OM 5.5
- -
A
0.3
0 2 4 6 8 10
Equilibrium concentration of oxalate (mM)
Fig. 2.2
Isotherms for the sorption of oxalate by clay fractions at different
pH values.


44
Fig. 2.15
Release of organic carbon from the whole-soil samples as a function
of oxalate sorption.


Oxalate sorbed (mM kg )
31
Equilibrium Oxalate Concentration (mM)
Fig. 2.7
Isotherms for the sorption of oxalate by the whole-soil samples.


124
Fiskell, 1974). While relatively large amounts of total P can be present in spodic
horizons, water-soluble P tends to be low. According to Hingston et al. (1974),
chemisorption and bi- and multidentate complex formation can decrease P
desorption by increasing the irreversibility of sorbed P. Van Rees and Comerford
(1986) observed that roots are present in the spodic horizon, and Neary et al
(1990), using a mechanistic nutrient uptake model, suggested that subsurface
horizons could contribute a significant amount of the P required by southern pine.
Soil buffering capacity used in this model was calculated using a sorption isotherm.
To predict P uptake accurately, it may be important instead to determine and work
with the P desorption isotherm.
Desorption of P is generally determined by one of three different methods:
i) sequential extraction of soil with CaCl2, Ca(N03)2 or other salt solution at a
constant soil to solution ratio for a constant equilibration time; ii) equilibration of
soil with Ca(N03)2 or other salt solution at different soil-to-solution ratio for a
constant time period; and iii) anion exchange-resin extraction, in which the soil is
treated with different amounts of anion exchange resin in water or salt solutions
for a constant time period The first two methods study the chemical release of P
from soil relating to the movement of P in the soil profile including plant uptake.
The third method is more related to the overall extraction of P by plant roots.
The anion-exchange resin instead proved inadequate for precise characterization of
labile P, because it did not account for rate phenomena (Yang et al.. 1991;
Abrams and Jarrell, 1992). This is because three steps are involved during


112
the trees. Using Ca(N03)2 as a extractant instead, the estimated contribution is
approximately 0.20 kg of P ha'V1 As pine trees remove roughly 3 kg P
ha'V1 form the soil, this is about 20 % of the tree's demand. Given this
example, it shows that oxalate could significantly increase the availability of P in
spodic-horizon material. These results, in combination with earlier work (Fox et
af, 1990a; Fox et al.. 1990b; Neary et al.. 1990) highlight the potential
importance of subsoil fertility to the productivity of pine plantations in
flatwoods.
Future Research
Results from this study have defined the mechanism of P release into the
soil solution by oxalate, and its influence on initial solution P concentrations and
buffer power for this soil. The next step could be to incorporate this
information into a mechanistic nutrient uptake model in order to predict P
uptake by plants. This should support the conclusion that oxalate released into
the soil solution increases P availability in these soils.
To enhance our understanding of P cycling in spodic horizons,
fiirtherwork should address the following issues:
1. Influence of oxalate on P sorption and desorption in the presence of
additional cations and anions.
2. Identification and characterization of organic P compounds.


Amount of NaOH used (mL) Amount of NAOH used (mL)
116
pH
Fig. A-3.
Titration curves for the whole-soil: A) With organic matter, B)
Without organic matter


103
Table 4.4 Influence of P sorption by clay fractions and whole-soil samples on the
ratio of OH released, and OH + oxalate released, to P sorbed (Mean
SD; n=3).
Without Organic Matter
With Organic Matter
Ox PrevA
Present
(mmol kg'1)
Ratio
Ox Prev
Present
(mmol kg'1)
Ratio
OH/P
(OH + Ox)/P
OH/P
(OH + Ox)/P
Clav Fractions
45617
0.94*0. laa
1.04*0.1
27514
o.spo.i33
0.92*0.1
48*2.4
1.08*0.1
1.03*0.1
430.2
0.980.r
0.98*0.1
30.5
0.98*0. laa
0.99*0. laa
50.2
0.950. laa
0.95*0.1
Whole Soil
Material
14.9*0.7
1.08*0.1
2.990.2ab
11.61.0
Timor
2.900.3ab
12.40.4
1.01*0.1
1.250.2ba
8.80.6
1.09*0.2
2.050.2bb
9.6*0.5
1.150. Ia3
1 350.2ba
7.40.1
1.06*0.1
1.26*0.2ca
6.2*07
0.96*0.1
1.160.2ba
5.50.2
1.07*0.1
1.07*0. lca
4.0*0.1
0.850.1
0.99*0. lba
4.00.1
0.98*0.1
0.98*0.1
0.80.1
0.990. laa
1.04*0. lba
0.80.1
1.04*0.1
1.04*0.1
AOx Prev is Oxalate previously sorbed;
The first superscript letter indicates significant differences within a column and
material, and the second letter indicates significant differences between rows.


2.12 Release of Al from the whole-soil samples as a
function of oxalate sorption 39
2.13 Concentrations of aluminum species at different pH levels as
influenced by oxalate 41
2.14 Release of organic carbon by the clay fractions at varing pH as a
function of oxalate sorption 43
2.15 Release of organic carbon from the whole-soil samples
as a function of oxalate sorption 44
2.16 Relationship between A1 release and organic carbon release
by the clay fraction at varing pH 46
2.17 Relationship between A1 release and organic carbon release
from the whole-soil samples 47
3.1 Isotherms for the sorption of P by the clay fraction 65
3.2 Isotherms for the sorption of P by the whole-soil samples 67
3.3 Release of OH' ions during sorption of P by the clay fraction ... 73
3.4 Release of OH' ions during sorption of P by the whole-soil samples 74
3.5 Kinetics of OH" ions released during the sorption of anions
by the clay fraction 76
4.1 P desorption curve for whole-soil samples and clay fractions 93
4.2 Influence of previously sorbed P on oxalate sorption and
P desorption by clay fractions 95
4.3 Influence of previously sorbed P on oxalate sorption and
P desorption by whole-soil samples 96
4.4 Influence of previously sorbed oxalate on P sorption and
oxalate desorption for clay fractions 101
4.5 Influence of previously sorbed oxalate on P sorption and
oxalate desorption by the whole-soil samples 102
IX


60
sample of whole soil from each treatment (with or without organic matter) were
placed in 100 ml bottles with 50 ml of solution having a phosphate concentration
of 0 mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM or 10 mM adjusted
initially to pH 4.5 (using 0.1 M HC1 or NaOH).
Sorption of Phosphate and Oxalate Added as a Mixture
Sorption experiments were carried out at different oxalate to phosphate
molar ratios. The concentration of oxalate used was 1.0 mM with a phophate
concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM for the clay samples. For
the whole-soil sample with and without organic carbon, an oxalate concentration of
1.0 mM was used, with a phosphate concentration of 0 mM, 0.1 mM, 0.5 mM,
1.0 mM, 3.0 mM, 7.0 mM or 10 mM. pH of the clay suspensions and the
whole-soil suspensions was initially adjusted to 4.5 using 0.1 M HC1 or NaOH.
Two drops of toluene were added to inhibit microbial growth, and the
samples were placed in a reciprocating shaker for 24 hrs. pH values of the
suspensions were periodically (every three hours) adjusted to 4.5 with 0 1 M HC1
or NaOH over the 24 hrs of the experiment. At the end of the reaction period,
each suspension was centrifuged at 12,000 g for 20 minutes. The resultant
supernatant solutions were used for analysis of oxalate, Al, Fe, inorganic P, and
organic carbon. Sorbed P and oxalate were calculated from the difference between
the initial and final concentrations in solution. The adsorbate was saved for further
experiments, as described in Chapter 4.


13
1986). The reactivity of labile soil organic matter is generally attributed to the
presence of carboxyl (-COOH) and phenolic hydroxyl (-OH) groups.
The major mechanisms by which soil organic matter is sorbed to mineral
surfaces include i) anion exchange, ii) ligand exchange-surface complexation, iii)
hydrophobic interaction, iv) hydrogen bonding, and vi) cation bridging (Sposito,
1984). Parfitt et al. (1977) suggested that carboxyl groups of iulvic and humic
acid replace surface OH' from gibbsite, goethite and imogolite. Ligand exchange
of surface-coordinated OH' and OH2 from Fe oxides by humic substances had also
been noted by Tipping (1981b) Jardine et al. (1989) suggested that the primary
mechanism of dissolved organic carbon sorption is physical adsorption followed by
an anion exchange mechanism. This suggests that organic carbon should alter the
surfaces charge, surface area and retention properties of inorganic colloids in spodic
horizons.
According to Stumm (1986), organic ligands form complexes with Al and
Fe in solution and on the surfaces of Fe and Al oxides. The formation of such
complexes in solution increases the rate of dissolution of surface complexes.
Huang and Violante (1986), Miller et al. (1986), Tan (1986), and Pohlman and
McColl (1986) also reported that organic anions which formed complexes with Al
and Fe stimulated the dissolution of Al and Fe solid phases. Ohman and Sjoberg
(1988) proposed that polycarboxylic acids are effective in solubilization of Al
surfaces They proposed this as a series of complexation-hydrolysis reactions:


P sorbed (mmol kg )
65
Equilibrium P concentration (mM)
Fig. 3.1.
Isotherms for the sorption of P by the clay fractions.


BIOGRAPHICAL SKETCH
Jagtar Singh Bhatti was born on March 9, 1958 in Jalandhar, Punjab, India.
He attended high school in Ludhiana and graduated from Malwa Khalsa High
School in 1974. He went to Punjab Agricultural University Ludhiana, Punjab and
received his Bachelor of Science (1979) and Master of Science in Soil Science
(1981). He joined the faculty of Soil Science at Punjab Agricultural University
Ludhiana, Punjab and taught undergraduate courses in soil science. In December
1984, he met Gurmeet and got married. In June 1985, they decided to migrate to
Canada. After coming to the cold country, Jagar entered another Master's program
in soil chemistry at University of Saskatchewan, Saskatoon and graduated in 1988.
Jagtar worked as Research Assistant at University of Washington, Seattle for one
and a half years. After traveling into Far Eastern countries and Australia for
about four months, in May 1991, he joined University of Florida for Doctor of
Philosophy in Forest Soils in the Department of Soil and Water Science. He is
currently employed as an Ecological Modeler at Sault Ste. Marie, Ontario, Canada
for Forestry Canada Ontario Region
147


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EWOK818YT_JDPK20 INGEST_TIME 2015-03-27T18:55:40Z PACKAGE AA00029814_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


86
activities of fungi and bacteria (Stevenson, 1982). Although oxalate is abundant in
the rhizosphere (Fox and Comerford, 1990), only a few investigations have studied
the competitive sorption of oxalate and phosphate. These include studies in
tropical soils (Lopez-Hernandez et al., 1986); and montmorillonite (Violante and
Gianfreda, 1993). It has been proposed that oxalate can release P from Al- and
Fe-hydroxide surfaces through ligand-exchange reactions. Recently, Fox et al
(1990a) studied the kinetics of P desorption by oxalate from spodic and argillic
horizons of a Spodosol. They suggested that P released by oxalate was through
ligand exchange, but did not provided conclusive evidence.
Phosphate desorption studies (Barrow, 1983; Kuo and Pan, 1988; Bakheit
Said and Dakermanji, 1993; Raven and Hossner, 1993) have indicated that a large
portion of the retained P is irreversibly sorbed. In all such desorption studies,
researchers used either 0.01 M CaCl2, Ca(N03)2 or anion exchange resins.
Phosphate desorption isotherms do not normally coincide with P sorption isotherms
(Nye and Tinker, 1977; Kuo and Pan, 1988; Bakheit Said and Dakermanji, 1993).
Lopaz (1974) studied the desorption of P using citrate and found that desorption
could only occur when the displacing anion was specifically sorbed and present in
sufficient concentration in the soil solution. Organic anions like citrate and oxalate
are continuously produced in soil (Smith, 1969) and can be present in high
concentrations (Fox and Comerford, 1990). However, desorption isotherms using
oxalate as an extractant have not been reported to our knowledge.


122
Table B-l. Percent reduction in oxalate sorption for the whole-soil samples and
clay fractions (Mean SD. n=3).
P: Oxalate
Percent reduction
in oxalate sorption
molar ratio
+ OM
- OM
Whole-Soil
10:1
23.51.2aa
18.50.9ab
7:1
13.90.7ba
12.30.7ba
3:1
5.20.7ca
5.50.4ca
1:1
1.70.4d
2.40.4da
0.5:1
0.21.0da
1^0.5"
0.1:1
-0.90 8da
1.60.3'b
Clay Fractions
10:1
232aa
562ab
1:1
10lba
284bb
0.1:1
22ca
llca
In each superscript first letter indicates significance among means within the column
and second letter indicates significance among the rows, within the clay fraction
and the whole-soil samples.


REFERENCES
Abrams, M. M. and W M. Jarrell. 1992. Bioavailability index for phosphorus
using ion exchange resin-impregnated membranes. Soil Sci. Soc. Am. J.
56:1532-1537.
Appelt, H N. T. Coleman, and P F Pratt 1975. Interactions between organic
compounds, minerals and ions in volcanic-ash-derived soils: II. Effect of
organic compounds on the adsorption of phosphate. Soil Sci. Soc. Am.
Proc 39:628-630.
Bakheit Said, M. and A. Dakermanji 1993. Phosphate adsorption and
desorption by calcareous soils of Syria. Commun Soil Sci. Plant Anal.
24:197-210
Baldwin, J.P., P B Tinker, and P H. Nye 1972 Uptake of solutes by
multiple root systems from soil: II. The theoretical effects of rooting
density and pattern on uptake of nutrients from soil. Plant and Soil
36:693-708.
Ballard, R, and J. G. A. Fiskell. 1974. Phosphorus retention in Coastal Plain
forest soils: I Relationship to soil properties. Soil Sci. Soc. Am. Proc.
38:250-255.
Barber, S.A. 1984 Soil Nutrient Bioavailability: A Mechanistic Approach
Wiley-Interscience, New York
Barrow, N. J. 1987. Reactions with Variable Charge Soils. Martinus Nijhoff Pub.,
Dordrech, The Netherlands.
Barrow, N. J. 1983. On the reversibility of phosphate sorption by soil. J. Soil
Sci. 34:751-758.
Bilinski, H., L Horvath, N. Ingri and S. Sjoberg. 1986. Equilibrium aluminium
hydroxy-oxalate phase during initial clay formation; H+, Al3", oxalic acid-Na*
system. Geochim. Cosmochim. Acta. 50:1911-1922.
136


52
The data presented here overwhelmingly support the process of ligand
exchange. The OH was released in proportion to oxalate sorbed, as shown with
both the batch sorption and FTIR measurements. The molar ratio was not
statistically different from 1 at pH 3.5 or from 2 at pH 4.5 and 5.5 (Table 2.3).
We further conclude that precipitation cannot be a significant factor in oxalate
sorption in this study. A1 was released into solution in relation to oxalate
sorption, showing that oxalate sorption caused A1 dissolution rather then
precipitation. Also, MINTEQA2 calculations suggested that the oxalate formed
soluble Al-oxalate complex in solution and keep it from precipitating. It must be
remembered that this was a Na-saturated system without any Ca present on the
exchange complex to form insoluble CaC204 species. Under field conditions,
spodic horizon pH is about 3 8 to 4.2, with little Ca present on the exchange
complex to interact with native levels of oxalate
These data further support the conclusion that oxalate is sorbed as
monodentate and/or binuclear surface complexes at pH 3.5 ([mmol OH / HC204] =
1), but forms predominantly a bidentate complex at higher pH values ([mmol OH /
C204'2] = 2). These data support the conclusions of Parfitt et al. (1977), who
also showed monodentate inner-sphere complex formation of oxalate on goethite
surfaces at pH 3.4, at high oxalate concentrations, using infrared spectroscopy.
They also suggested formation of a binuclear complex on goethite when oxalate
loading rates were low.


109
a coordinated complex with A1 at the mineral surface, with sufficient
strength to break the Al-0 bond and solubilize the metals into solution.
5. Considerable amounts of organic carbon were released into solution during
the sorption of oxalate. Aluminum ions can form cation bridges between
the soil organic carbon and soil particles. Oxalate can detach this bridge to
form soluble complexes with A1 in solution, resulting in the release of
large amounts of organic carbon.
6. The organic carbon released into solution during oxalate sorption had a
lower E4 to E6 ratio, suggesting that it was humic in nature.
P Sorption
7. P appeared to be sorbed by the mineral surfaces through a ligand-exchange
process. P formed monodendate and/or binuclear complexes on the mineral
surfaces.
8.Organic carbon reduced P sorption. Maximum reduction in P sorption
(about 50 %) was observed when both organic carbon and oxalate were
present
9. Some of the sites for sorption onto the surfaces of soil and clay particles
were common for oxalate, P and organic carbon.
10. P sorption released small amounts of Al, Fe and organic carbon into
solution.


Ill
P precipitation. This could increase the initial P concentration in solution and
affect the P buffer power of these soils resulting in a higher concentration of P
in solution.
The oxalate concentration in the rhizosphere of a spodic horizon range
from 0.0 to 2.8 mM (Fox and Comerford, 1990). This reported concentration
of oxalate in bulk soil solution tends to be much lower then in the vicinity of
roots and fungal hyphae. It seems reasonable to assume that actual
concentration near roots may be one or two orders of magnitude higher than
the concentration in bulk soil solution. The oxalate concentration of 5 mM
used in this study to desorb P exists under field conditions. At this oxalate
concentration, the whole-soil samples released 0.1 mM P into solution.
Assuming a gravimetric moisture content of 20 % and a 0.1 mM concentration
of P, the total amount of P in the soil solution at one time would be 0.6
mg kg'1. However, slash pine roots occupy less than 1 % of the soil volume
(Van Rees and Comerford, 1986), Therefore, high concentrations of oxalate
would exist in the zone immediately adjacent to the roots, which is only a few
millimeters wide and coincides with the rhizosphere region where microorganism
are particularly active. To include the influence of fungal hyphae along with
roots, one can further assume that these high concentration of P (0.6 mg kg'1)
would be present in approximately 2 % of soil volume. Assuming that the
entire pool of oxalate was being replenished on a weekly basis (Fox et al..
1990b), the spodic horizon would contribute approximately 0.6 kg P ha'V'1 to


CHAPTER 2
INFLUENCE OF SOIL ORGANIC MATTER AND pH ON OXALATE
SORPTION ONTO A SPODIC HORIZON
Introduction
Natural organic anions in soils and freshwater environments are derived from
plant and animal residues, microbial metabolism and canopy drip. Simple aliphatic
anions such as citrate and oxalate are continuously produced through the activities
of microorganisms (Flaig, 1971), aqueous extraction of leaves (Bruckert, 1971;
Stevenson, 1982), and activities of microorganism and/or roots in the rhizosphere
(Reddy et ah. 1977). The type of vegetation, pH, oxidation potential, water
potential and temperature each affect the kinds and amounts of low molecular
weight organic anions that are produced (Stevenson, 1982; Goh and Haung, 1986;
Hue et al. 1986; Pohlman and McColl, 1988).
Among the aliphatic anions detected in the extracts of litter and soils,
oxalate occupies an important place in many ecosystems, both in well-, and poorly-
drained soils (Vedy and Bruckert, 1982). For many organisms, oxalate is a
metabolic end product of low residual nutrient energy (Cromack et al.. 1979).
Smith (1969) and Fox and Comerford (1990) measured significant concentrations of
11


92
difference between means was determined by a t-test (Snedecor and Cochran,
1980). Standard deviations or standard errors of means are given as appropriate.
Results
Desorption of Phosphate in the Presence of Oxalate
Desorption of Phosphate
A considerable amount of phosphate was desorbed during oxalate sorption,
and the presence of organic matter did not significantly influence the desorption of
P by oxalate. P desorption isotherms are presented in Figure 4.1 both for the
clay fractions and the whole-soil samples. Corresponding parameters for the
Freundlich models are presented in Table 4.1.
Sorption of Oxalate
Oxalate sorption by the clay fraction and the whole-soil samples is presented
in Figure 4.2 and 4.3, respectively. The amount of oxalate sorbed increased as
the amount of P previously present on the mineral surfaces decreased both for the
whole soil and the corresponding clay fractions. The relationship between release of
previously sorbed P and amount of oxalate sorbed is presented in Appendix D-l.
As low amounts of oxalate were sorbed in the present experiment, large amounts
of P were desorbed. The amount of oxalate sorption followed a quadratic model.
Parameters of the model are presented in Table 4.2. The amount of oxalate


79
Table 3.7. E4/E6 ratio of organic carbon released into solution during the sorption
of P only, and P + oxalate by the clay fractions (Means SD; n=3).
Treatment
Concentration of
Anions added (mM)
E4/E6 ratio
P
10
9.04+0.10a
1.0
9.310.25a
0.1
9.240.56a
0.0
9.260.07a
P + Ox
10 + 1
9.14+0.06a
1.0 + 1
8.930.05a
0.1 + 1
7.650.21b
Values within columns followed by same letter are not significantly different at the
5% level


61
Chemical Analysis
Solution pH was measured using a combination glass electrode in
conjunction with an Orion pH meter. Aluminum was determined using a flame
emission spectrophotometer with N20-C2H2 flame whereas Fe was detected with a
C2H2 flame. Total organic carbon content of the clay sample with organic matter
was measured via persulfate oxidation and IR anaylasis of the resultant CO, using
on a TOC apparatus (College Station, TX). E4 (A=465) to E6 (A=665) ratios
of the organic carbon released into solution from the clay with organic matter
were measured using a Shimadzu UV/Vis spectrophotometer.
Oxalate in the extract was determined by HPLC (Fox and Comerford,
1990), using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton
Co., Reno, Navada) along with a Gilson single piston high pressure pump and a
Pheodyne model 7125 injection valve fitted with a 20 /uL injection loop. The
HPLC system used a Gilson Holochrom variable wavelength UV detector in
conjunction with a Gilson computerized integrator. Eluent was 0.005 M H,S04 at
a flow rate of 2 mL min'1 Oxalate concentration was calculated from the
calibration curve obtained with standard solutions of 0.1 to 10 mmol L'1.
Inorganic P in the filtrate was determined by a molybdenum-blue
colorimetric procedure using ascorbic acid as the reductant (Murphy and Riley,
1962). This is an operational definition of inorganic P, for molybdenum may also
hydrolyze some organic P (Stainton, 1980).


58
concentrations of amorphous A1 oxides and are classified as Alaquods. Organic
substances which are present as surface coatings and/or acting as cementing
material can significantly affect the retention of P by modifmg the specific surface
area and surface charge of crystalline and noncrystalline minerals in the finer
fractions of the Spodosols. Therefore, it is important to investigate the sorption of
P by this highly reactive component.
Recently, studies were carried out to investigate the sorption of P by
organo-mineral complexes. Violante and Huang (1989) studied the sorption of P
on precipitated-Al products formed in the presence of organic ligands (citrate,
tartrate, malate, aspartate and tannate). They found that the amount, the nature
and the size of organic ligands coprecipitated with Al, as well as the surface
properties of AJ-organic ligand complexes, strongly influenced P sorption capacity.
Haynes and Swift (1989) investigated the effect of pH on P sorption onto Al-
organic matter complexes. They reported that increase in pH greatly increased P
sorption capacity. These studies were performed on pure minerals. However in
soils, many minerals coexist, and were formed in the presence of different organic
coumpounds. Therefore, P sorption properties of these minerals could be
considerably different from those of the pure minerals.
The purpose of the present investigation was to study: i) P sorption by the
clay fraction and the whole-soil samples of a spodic horizon, and ii) the influence
of oxalate and soil organic matter on P sorption in a such system.


91
Speciation Calculations
The metal speciation model MINTEQA2 version 3.2 was used to calculate
the species of Al, P and oxalate in the solution. It was assumed that the system
had achieved equilibrium conditions Log K values for Al-oxalate complexes were
obtained from Thomas et al. (1991). Assigned Log K values were 6.12 for
A1C204 1 and 11.15 for A1(C204)2'. The formation constants used for calculations
of other species were those of the MINTEQA2 data base, as taken from Martell
and Smith (1982).
Statistical Analysis
Statistical differences in desorption of oxalate or phosphate using different
regression lines were tested using the General Linear Models procedure of the SAS
framework (SAS Institute, 1985). Observed desorption of oxalate or phosphate
was related to the amount of P and oxalate previously present on the surfaces,
respectively. The model employed was:
7a + p.(logV.)+ P2(logA/+ e. (4.2)
where a is the intercept, P, and P2 are 1st and 2nd order coefficients of the line
and e is the error term. To compare two regression lines, we compared the a
and p values of the two lines with ANOVA and t-test used to test the differences
between 0^ and a2 as well as between p, and P2. The significance of the


OH released (mmol kg )
73
Fig 3.3
Release of OH' ions during sorption of P by the clay fractions


37
Table 2.3 Ratios of OH' ions released to oxalate sorbed for the clay fractions
and whole-soil samples (mean SD, n=3).
pH
OH released/Oxalate sorbed
With Organic Matter
3.5
Clav fractions
0.980.03a
4.5
1.730.11b
5.5
1 880.22b
Without Organic Matter
3.5
0 860.06a
4.5
1.960.1 lb
5.5
1 850.08b
With Organic Matter
4.5
Whole-Soil Samples
1.920.21b
Without Organic Matter
4.5
l,860.17b
Superscript "a" and "b" indicate that values are not significantly different from 1.0
or 2 respectively at the 5% level.


16
saturated with Na using 0.01 M NaCl Excess Na was removed with four or
five washings using distilled water. The samples were oven dried at 110 C.
Mineralogy of the Clay Fraction
Mineralogy of the clay was determined by X-ray diffraction (XRD) on
parallel-oriented specimens, using Cu-Ka radiation. Approximately 250 mg of clay
were deposited on a ceramic tile under suction followed by K and Mg saturations
on the tiles by washing with the respective chloride salt and rinsing. Glycerol then
was added to the Mg saturated samples. Samples were scanned at 2 20 per
minute, at 25 C for both the K- and Mg-saturated samples. Minerals were
identified from XRD peak positions, making use of differentiating responses to ion
saturation. Detectable minerals in the clay sample included quartz, kaolinite,
hydroxy interlayered vermiculite, and gibbsite.
Oxalate Sorption Studies
Sorption experiments were carried out with different concentrations of
oxalate over a pH range (3.5, 4.5 and 5.5). Triplicate samples of clay (250 mg
with and without organic matter) were placed in 30 ml bottles with 25 ml of
solution having an oxalate concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM
adjusted initially to pH 3.5, 4.5 or 5.5 (using 0.1 M HC1 or NaOH). Triplicate
five-gram whole-soil samples from each treatment (with or without organic matter)
were placed in 100 ml bottles with 50 ml of solution having an oxalate


36
mole of oxalate sorbed are presented in Table 2.3. Average molar ratio at the
higher sorption rates was not significantly different from 1.0 at pH 3.5, nor was it
significantly different from 2.0 at pH 4.5 and 5.5 for both the clay and whole-soil
surfaces.
Release of Aluminum and Iron
The influence of oxalate sorption on the release of A1 is presented in
Figures 2.11 and 2.12. Aluminum release increased linearly with sorption of
oxalate (Table 2.4). The release of A1 from whole-soil samples followed a
sigmoidal shape, both for samples with and without organic matter (Figures 2.12).
Soil organic matter significantly increased A1 release, but the release of A1 from
the clay fractions without organic matter at pH 3.5, 4.5, and 5.5 was not
significantly different (Fig. 2.11). Speciation of A1 solution in the presence of
oxalate was calculated using MINTEQA2. The majority of the A1 in solution was
present as [A1L]+ and [A1L2]' species at pH 4.0 and above (Figure 2.13). The
system was undersaturated with respect to both NaAl(0H)2(C204).3H20 and
A13(0H)7(C204).3H20 The concentration of [A1L2] species increased with pH.
The Spodosol used in the study had very low concentrations of total Fe
Iron released from the clay fractions and the whole-soil samples is presented in
Table 2.4. The clay fractions and the whole-soil samples with organic matter
released a significantly greater amount of Fe then materials without organic matter.


82
discrete crystal growth to occur. This is much longer than the 24 hrs represented
by this study. Sposito (1984) further suggested that the adsorption-dominated stage
of phosphate sorption takes less than about 50 hrs. Calculated ion-activity product
constants using MINTQEA2 V3.ll suggested that, at equilibrium, this system was
supersaturated with respect to amorphous aluminum phosphate, A1(0H)2H2P04. The
formation or precipitation of amorphous aluminum phosphate thus could be
possible. Sorption of phosphate through ligand exchange is known to serve as a
nucleation site for the precipitation of amorphous aluminum phosphate onto the
clay surface. However, formation or precipitation of amorphous aluminum phosphate
(A1(0H)2H2P04) consumes OH' instead releasing them. Our kinetic data on the
release of OH' ions showed no evidence of this. Up to 14 hrs, OH' ions were
continuously released into the solution. This suggests a ligand-exchange reaction.
After 14 hrs, there was no further release of OH' ions into solution from the clay
surfaces with organic matter present while the clay surfaces without organic matter
were still releasing OH'. Thus, for clay surfaces, there was no absolute decrease in
OH' concentration in solution. This suggests that precipiation of amorphous
aluminum phosphate was unlikely. Therefore, these results suggest that ligand
exchange was the dominant reaction for P sorption
Effect of Oxalate and Organic Carbon on P Sorption
Phosphate sorption significantly decreased in the presence of oxalate.
Likewise, P was effective in reducing oxalate sorption (Appendix B), indicating that


APPENDIX D
DESORPTION OF P AND OXALATE
Table D-l Influence of previously sorbed P by clay and whole-soil samples on
oxalate sorption and release of P, Al, Fe, OH, and organic carbon
(Mean SD, n = 3).
P Previously
Present
(mmol kg'1)
Oxalate
Sorbed
(mmol kg"1)
P
Amount Released
(mmol kg"1)
A1 Fe
OH
OCA
(mg g"1)
Clay without Organic Matter
5459
1757
70.11
1885
2.60.07
32611
910.3
1911
25.51
1864
2.70.07
36418
9.50.5
2056
4.31
1851
2.90.09
39014
Clay with Organic Matter
3535
2054
490.5
2251
2.50.15
34414
2676
720.7
2042
180.1
2252
2.90.07
3817
2777
9.20.8
2273
50.1
2303
3.10.03
4159
2676
Soil without Organic
Matter
24.50.7
9.40.3
9.70.1
4.80.1
0.120.01
15I
21 70.6
10.50.2
7.40.1
4.20.1
O.lliO.Ol
181
16.20.2
10.70.1
2.50.1
4.00.1
0.120.01
181
7.10.1
12.0.6
1.50.1
3.90.1
O.lliO.Ol
202
3.70.1
12.40.1
0.90.1
4.10.1
0.090.00
221
0.80.1
12.50.4
0.50.1
4.40.1
0.110.01
252
Soil With
Organic Matter
16.70.5
2.40.1
9.20.2
271
0.130.01
4.20.2
3.40.1
15.51.4
2.70.1
7.80.1
351
0.140.01
4.80.2
3.80.1
10.20.7
2.80.3
4.60.1
351
0.150.01
5.20.7
3.90.1
5.40.2
5.90.4
2.30.1
361
0.170.01
10.50.2
4.90.1
4.10.1
6.80.5
1.10.1
371
0.170.01
12.70.9
4.40.1
0.90.1
6.90.3
0.60.1
371
0.190.01
12.70.2
4.50.1
A OC = Organic carbon
134


INFLUENCE OF SOIL ORGANIC MATTER ON PHOSPHORUS AND
OXALATE SORPTION AND DESORPTION IN
A SPODOIC HORIZON
^ By
JAGTAR S BHATTI
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1995

ACKNOWLEDGMENTS
It is my privilege to express my sincere gratitude, appreciation and heartfelt
thanks to Nick B Comerford, my advisory committee chairman, for his able
guidance, keen interest, constructive criticism and constant encouragement during
the course of this investigation. I feel pleasure also in expressing high regards to
other members of my advisory committee, Earl Stone, Brian McNeal, Cliff
Johnston, P Surash Rao, H Gholz and Eric Jokela, for my scientific development
The financial support provided by the National Science Foundation for
funding my assistantship and the experimentation is greatly acknowledged. I
appreciate also the valuable technical assistance of Mary McLeod, Randy, and Dr
Cliff Johnston at various phases of the work. Mary's friendship was invaluable as
was her advice on a wide range of issues. My sincere thanks are also due to
everyone in the Forest Soils Laboratory for their cooperative attitude and
discussions during the various phases of this investigation. I enjoyed all of my
new friendships, especially those of my fellow graduate students.
Finally, I want to thank my wife Gurmeet and my son Kulraj for their
patience, love and understanding during this adventure. Without their tolerance and
moral support, I would have not done it
11

TABLE OF CONTENTS
ACKNOWLEDGMENTS ii
LIST OF TABLES v
LIST OF FIGURES viii
ABSTRACT xi
CHAPTERS 1
1 GENERAL INTRODUCTION 1
Phosphorus in flatwood Spodosols 1
Spodic Horizons 2
Phosphorus in Spodic Horizons 3
Organic Anions in Soil 5
Influence of Organic anions on Phosphorus Availability .6
2 INFLUENCED OF SOIL ORGANIC MATTER AND pH ON
OXALATE SORPTION INTO A SPODIC HORIZON 12
Introduction 12
Material and Methods 15
Results 22
Discussion 50
3 INFLUENCED OF SOIL ORGANIC MATTER AND OXALATE
ON SORPTION OF PHOSPHORUS INTO A SPODIC
HORIZON 55
Introduction 55
Material and Methods 59
Results 64
Discussion 78
iii

4 INFLUENCED OF SOIL ORGANIC MATTER ON DESORPTION OF
PHOSPHORUS AND OXALATE FROM A SPODIC
HORIZON 85
Introduction 85
Material and Methods 87
Results 92
Discussion 104
5 CONCLUSIONS 108
Conclusions from this Study 108
Influence of Oxalate on the P Nutrition of Trees 110
Future Research 112
APPENDICES 114
A TITRATION CURVES FOR THE WHOLE SOIL
AND THE CLAY FRACTION 114
B SORPTION OF OXALATE IN THE PRESENCE OF P
BY THE CLAY FRACTION 118
Introduction 118
Material and Methods 118
Results 120
C MEASUREMENT OF P DESORPTION BY DIFFERENT
METHODS AT VARIABLE AND
CONSTANT pH 123
Introduction 123
Material and Methods 126
Results 129
D DESORPTION OF P AND OXALATE 134
REFERENCES 136
BIOGRAPHYCAL SKETCH 147
IV

LIST OF TABLES
Table Page
2.1. Parameters of the linear regression models for oxalate sorption
by the clay fractions and the whole-soil samples 25
2.2. Parameters of the linear regression models relating
release of OH ions and organic carbon for the
clay fractions and whole-soil samples 34
2.3. Ratios of OH' ions released to oxalate sorbed
for the clay fractions and whole-soil samples 37
2.4. Parameters of the linear regression models relating release of
aluminum and iron with oxalate sorption for the clay
fractions and whole-soil samples 40
2.5. E4/E6 ratio of organic carbon released into solution
during the sorption of oxalate by the clay fractions 45
2.6. Intensity of the absorption bands at 1610 cm'1
(aromatic C=C and/or H-bonded C=0 stretching
of COOH) and 1460 cm'1 (OH deformation,
C-0 stretching of phenolic-OH, and/or C-H deformations
of CH3 and CH2) with standard deviations
for clay with organic matter at different pH values for
different amount of oxalate added 49
3.1. Langmuir sorption isotherm parameters for P sorption onto
the whole-soil samples and clay fractions with and
without organic matter 66
3.2. P sorption onto the whole soil material compared to P
that of sorption onto the clay fraction
expressed on a whole soil basis 69
v

3.3. Amount of Al Fe and organic carbon released
into solution with P + oxalate sorption onto
the whole-soil samples and the clay fractions 70
3.4 Effect of oxalate in reducing the sorption of P onto
whole-soil samplesand the clay fractions 71
3.5. Parameters of the linear regression models for
the release of OH' ions and organic carbon from soil and clay. .75
3.6 Ratios of OH' ions released to P sorbed for P
only and P + oxalate for the whole-soil
samples and the clay fractions 77
3.7 E4/E6 ratio of organic carbon released into
solution during the sorption of P only, and
P + oxalate by the clay fractions 79
3.8 Parameters of the linear regression models relating release of
aluminum and iron with P sorption for P only and
P + oxalate by the whole-soil and clay fractions 80
4.1 Parameters of Freundlich models for the desorption of
P from the whole-soil samples and clay fractions in the
presence of 5 mM oxalate 94
4.2 Quadratic equation coefficients for the sorption of anions
by the whole-soil samples and the clay fractions 97
4.3 Influence of oxalate sorption by clay fractions and whole-soil samples
on the ratio of OH ions released, and OH + P
released to oxalate sorbed 98
4.4. Influence of P sorption by clay fractions and whole-soil samples
on the ratio of OH ions released, and OH + oxalate
released to P sorbed 103
B-l Percent reduction in oxalate sorption for the
whole-soil samples and clay fractions 121
C-l Freundlich isotherm parameters for the desorption of P from
soil using the dilution and sequential extraction methods
at variable and constant pH 131
vi

D-l Influence of previously sorbed P by clay and whole-soil
samples on the oxalate sorption and release of P,
Al, Fe, OH, and organic matter 134
D-2 Influence of oxalate previously present on the P
sorption and release of oxalate, Al, Fe, OH,
and organic carbon by clay fractions and whole-soil samples 135
Vll

LIST OF FIGURES
Figure Page
1.1. Three possible phosphate surface complexes 4
2.1 FTIR spectra of clay (a) with organic matter and
(b) without organic matter 22
2.2 Isotherms for oxalate sorption by clay fractions at
different pH values 24
2.3 FTIR spectra of clay without organic matter after oxalate
sorption at pH 3.5 (A), 4 5 (B), and 5.5 (C) 26
2.4 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 3.5 27
2.5 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 4.5 28
2.6 FTIR spectra of clay with organic matter for different
concentrations of oxalate sorbed at pH 5.5 29
2.7 Isotherms for the sorption of oxalate by the whole-soil samples. .31
2.8 Release of OH ions by clay fractions at varing pH as
a function of oxalate sorption 32
2.9 Release of OH ions into solution during the sorption of
oxalate by the whole-soil samples 33
2.10 Kinetics of OH ions released during the sorption of oxalate
by the clay fractions at pH 4.5 35
2.11 Release of A1 from the clay fractions at varing pH as a
function of oxalate sorption 38
viii

2.12 Release of Al from the whole-soil samples as a
function of oxalate sorption 39
2.13 Concentrations of aluminum species at different pH levels as
influenced by oxalate 41
2.14 Release of organic carbon by the clay fractions at varing pH as a
function of oxalate sorption 43
2.15 Release of organic carbon from the whole-soil samples
as a function of oxalate sorption 44
2.16 Relationship between A1 release and organic carbon release
by the clay fraction at varing pH 46
2.17 Relationship between A1 release and organic carbon release
from the whole-soil samples 47
3.1 Isotherms for the sorption of P by the clay fraction 65
3.2 Isotherms for the sorption of P by the whole-soil samples 67
3.3 Release of OH' ions during sorption of P by the clay fraction ... 73
3.4 Release of OH' ions during sorption of P by the whole-soil samples 74
3.5 Kinetics of OH" ions released during the sorption of anions
by the clay fraction 76
4.1 P desorption curve for whole-soil samples and clay fractions 93
4.2 Influence of previously sorbed P on oxalate sorption and
P desorption by clay fractions 95
4.3 Influence of previously sorbed P on oxalate sorption and
P desorption by whole-soil samples 96
4.4 Influence of previously sorbed oxalate on P sorption and
oxalate desorption for clay fractions 101
4.5 Influence of previously sorbed oxalate on P sorption and
oxalate desorption by the whole-soil samples 102
IX

A-l Titration curves for the clay fractions: A) with organic matter,
B) without organic matter 114
A-2 Buffer curve for the clay fractions, used to calculate the
amount of OH ions released into solution: A) with organic matter,
B) without organic matter 115
A-3 Titration curve for the whole-soil samples: A) with organic matter,
B) without organic matter 116
A-4 Buffer curve for the whole-soil samples, used to calculate the
amount of OH ions released into solution: A) with organic matter,
B) without organic matter 117
C-l P desorption curves by dilution and extraction methods
at both variable and constant pH 130
C-2 Relationship between Kd and the equilibrium concentration
of P as obtained by different methods 133
x

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
INFLUENCE OF SOIL ORGANIC MATTER ON PHOSPHORUS AND
OXALATE SORPTION AND DESORPTION IN SPODOSOL
By
JAGTAR S BHATTI
May, 1995
Chairperson: Dr. N. B Comerford
Major Department: Soil and Water Science
Phosphorus (P) deficiencies are common for poorly-drained Spodosols of the
flatwoods region of the Lower Coastal Plain of the Southeastern United States.
Large quantities of total P may be present in spodic horizons, but the level of
water-soluble P tends to be quite low. The goal of this investigation was to study
P and oxalate sorption and desorption by the clay fraction and whole-soil material
of a spodic horizon as influenced by soil organic matter. Understanding the
sorption mechanisms for P in the presence of oxalate and soil organic matter
should form the basis for explaining P extraction by roots from soil surfaces, and
the buffering of P levels in soil solutions.
The spodic horizon had a higher capacity to sorb P compared to oxalate,
xi

with soil organic matter significantly reducing the sorption of both P and oxalate.
Maximum reduction in P sorption (about 50%) was observed when both organic
carbon and oxalate were present in the system. Ligand exchange apparently was
the dominant mechanism for P and oxalate sorption. Some of the sorption sites
were common for both the anions. Oxalate and P formed different types of
surface complexes, with pH determining the type of surface complex formed by
oxalate. Oxalate formed monodentate and/or binuclear surface complexes at pH
3.5, while bidentate complexes at pH 4.5 and 5.5. P formed monodentate and/or
binuclear surface complexes at pH 4.5.
Significant amounts of organic carbon, A1 and Fe were released into
solution during the sorption of oxalate. This study provided the first experimental
evidence for Stumm's theory of congruent dissolution of minerals by oxalate.
Oxalate desorbed large amounts of P into solution. The presence of organic
matter further increased the amount of P desorbed. Oxalate appeared to desorb
the P both through ligand exchange and surface dissolution. This could increase
the initial P concentration in solution and effect the P buffer power. Since oxalate
formed stronger complexes on the mineral surfaces P failed to desorb the oxalate
xii
into solution.

CHAPTER 1
GENERAL INTRODUCTION
Phosphorus in Flatwood Spodosols
Phosphorus (P) deficiencies are common for the poorly-drained Spodosols of
the flatwoods of the lower Coastal Plain of the Southeastern United States
(Comerford et al., 1984). Cycling of P, which occurs through mineralization,
immobilization and redistribution of P in soil, depends on its physico-chemical
properties. These include P sorption by colloidal surfaces, as well as microbial,
mycorrhizal and plant uptake of P (Stewart and Tiessen, 1987). Soil phosphorus
can be divided into three general categories: i) ions and compounds in the soil
solution, ii) ions sorbed onto or incorporated into the surfaces of inorganic
constituents, and iii) ions that are components of soil organic matter
Inorganic P in the soil solution of a Spodosol's A horizon is replenished by
mineralization of organic P, phosphorus leaching from the forest floor, and
sometime, by fertilizer application (Polglase et aL 1992). Mineralogy of the
surface soil and bleached E horizon is dominated by quartz (Carlisle et al.. 1988).
Therefore, mineralization of organic matter, plays an important role in P availability.
In many of these soils, mineral components have virtually no P retention capacity
1

2
in their surface horizons (A and E), either due to low clay contents or the nature
of the clay fractions (Ballard and Fiskell, 1974; Fox et al., 1990b; Yuan, 1992).
For the surface soils, a clay content as low as 10 g kg'1 is common. Thus P,
along with organic matter, can be leached from the surface horizon and accumulate
in the underlying spodic horizon.
Spodic Horizons
Organic matter migrates through the A horizon of Spodosols in soluble and
colloidal forms and is adsorbed or precipitated, immobilized in the Bh horizon.
Current concepts of the formation of spodic horizons are based on the formation
of Al and or Fe humic complexes. Stability and mobility of these compounds
depend on the metal concentration in the soil solution. If the amount of Al
and/or Fe available for organo-metal complex formation is low in the A horizon,
complexes will be formed in the A horizons with low metal/organic ratios. In this
case the amount of Al and/or Fe chelated is insufficient to cause immobilization of
metal organic compounds, and may then move down in the pedon. During the
downward migration, these metal-organic compounds (De Coninck, 1980),
concurrently sorbed more polyvalent cations, which results in a progressive decrease
of their net negative charge. The presence of higher concentrations of Al and Fe
in the subsoil and/or at pH values different from that of the surface horizon may
eventually neutralize the remaining charge. This results in the precipitation of
metals along with organic matter in subsurface horizons and leads to the

3
development of a spodic horizon (De Coninck, 1980; Farmer et al.. 1983;
Buurman, 1984; Tan, 1986; Ugolini et al., 1988). Therefore, the complexation and
translocation of Al and Fe by organic acids is a primary mechanism during the
podzolization process. This can result in large accumulations of organic carbon
(which is mostly as humic and fiilvic acid), Al, and Fe in spodic horizons.
Phosphorus in Spodic Horizons
Spodic horizons contain elevated levels of amorphous or poorly crystalline
Al oxides, which can sorb P Not all of this sorbed P is in plant-available form
(Ballard and Fiskell, 1974). Large quantities of total P may be present in spodic
horizons, but the level of water-soluble P tends to be quite low. The
physico-chemical sorption of P is as an inner-sphere complex at the surface of Al
and Fe hydroxides and at the broken edges of silicate clay minerals (Sposito,
1984). An inner-sphere P complex refers to a surface complex resulting from
ligand exchange between a surface Lewis acid site (S) and the adsorbed ion
(Parfitt, 1978; Goldberg and Sposito, 1985). Such complexes are quite stable,
showing mainly covalent or ionic bonding character.
Three types of inner-sphere P complexes have been postulated: monodentate,
bidentate and binuclear (Fig. 1; Parfitt et al.. 1977). Tejedor-Tejedor and
Anderson (1990) used fourier transform infrared spectroscopy (FTIR) to study the
sorption of orthophosphate onto goethite particles in an aqueous suspension. They
observed the formation of three different types of complexes: protonated and

4
Monodentate
sOH + H2PO4 s
Bidentate
OH
OPOH + 0H
II
O
s
H2P04
,0
/p\ +
o
OH
S
S
Binuclear
\
OH + H2PO4
Figure 1.1. Three possible phosphate surface complexes. "S" could represent
either A1 or Fe.

5
nonproptonated bidentate bridging, and a nonprotonated monodentate complex. The
speciation of these complexes was a function of pH, as pH determines the species
of phosphate in solution. They also identified a monodentate complex on the
geothite surfaces when the iron/phosphate ratio was less than unity and pH was
between 3.5 and 6.0. At higher pH (6.0 to 8.3) and higher concentrations of
phosphate in the system, surface complexes were bidentate. Therefore, the type of
inner-sphere complex in different soils may be determined by the pH and the
concentration of available P in soil solution. Nanzyo (1987) studied sorbed
phosphate species using diffuse reflectance infrared spectroscopy (DRIR) and found
that phosphate reacts not only with surface sites but also with the structural
aluminum of allophanic soils. He reported that P can also precipitate as
noncrystalline aluminum phosphate. Most of the studies, however, that have been
carried out on the mechanism of P retention and release have used pure Fe & A1
oxides as the solid matrix. None of these studies were accomplished using soil
materials, where different soil components may coexist.
Organic Anions in Soil
A wide variety of low molecular weight organic anions have been identified
in forest and agricultural soils (Stevenson 1982). The commonly found organic
anions include oxalate, citrate, formate, acetate, malate, maleate, lactate and
fumarate (Gardner et al.. 1982; Hue et al.. 1986; Pohlman and McColl, 1988).
Fox and Comerford (1990) reported that oxalate comprised 60 to 80 percent of

6
the identified, free, low molecular weight, organic anions in a group of Spodosols
from north Florida. They also reported that the oxalate levels in soil solutions
averaged an order of magnitude higher in spodic horizons than in the
corresponding surface A horizon. These data show that oxalate is present and, if
it also affects P sorption, could play a significant role in forest productivity.
Influence of Organic Anions on Phosphorus Availability
Fox et al (1990a) proposed that, in lower Coastal Plain Spodosols, low
molecular weight organic anions, whether leached from the surface horizon or
produced in situ, might stimulate the release of phosphorus from mineral surfaces.
Organic anions acting as ligands are known to release P by i) replacing P sorbed
at surfaces of Al or Fe oxides through ligand-exchange reactions (Huang and
Schnitzer, 1986); ii) dissolving metal oxide surfaces and releasing sorbed P (Martell
et al.. 1988); and iii) complexing Al and Fe in solution, thus preventing the re
precipitation of metal-P compounds (Ng Kee and Huang, 1977). It has also been
observed that organic anions may block sites on mineral surfaces and reduce P
sorption (Kafkafi et al.. 1988).
Huang and Violante (1986) studied aluminum-citrate complexes in aqueous
solution They reported that Al and citrate form a 1:1 complex where the citrate
ligand occupies three of the six coordination sites around each Al while each of
the other three sites are occupied by a water molecule. Occupation of
coordination sites by citrate instead of water imposes a restriction on the

7
subsequent hydrolysis of hydroxy-aluminum polymers. The greater the
concentration of organic anions in the system, the greater should be the
replacement of water molecules and the blocking of A1 coordination sites.
Therefore, the occupation of coordination sites by organic ligands would also block
sites which would otherwise be available for the sorption of P Thus, organic
anions, through their reactions with Fe and A1 both in solution and at soil surfaces
should increase the availability of P in soils (Martell et al.. 1988). The activity of
these surface sites available for coordination with organic ligands, and the activity
of ligands in the solution, strongly depend upon pH. The role of pH would be
through i) its effect on the relative fractions of various species of organic anions in
solution (and these species differ in their affinity for the adsorbent); ii) the
variation in charge density of the solid surfaces with pH; and/or iii) competition
between OH' and the organic anion for common sorption sites.
Release of P is possible only if the stability of the metal-organic complex is
higher than that of the solid-phase metal-P complex. The stability of soluble
metal-organic complexes depends on the presence and molecular arrangement of
carboxylic and phenolic functional groups on the organic anion (Pohlman and
McColl, 1986; Martell et al.. 1988). Formation constants can be used to predict
the relative effectiveness of different organic acids for releasing P into solution.
Organic anions such as oxalate and malate have been classified as having high
complexing ability for Al and Fe (Fox et al.. 1990b; Pohlman et al.. 1990) when
they tend to form five- and six-membered rings between the anion and metal.

8
Therefore, a low molecular weight organic anion like oxalate could play an
important role in solubilizing soil Al, therefore indirectly affecting the availability of
phosphorus in forest soils.
Among 16 organic anions studied, a threshold value for log between
4.0 and 4.5 was required before substantial amounts of P were released from a
spodic horizon in Florida (Fox et al.. 1990a). Beyond this log value, release
of Al and inorganic P increased with increasing formation-constant values.
Violante et al (1991) and Violante and Gianfreda (1993) studied the competitive
sorption of phosphate and oxalate on aluminum oxide and montmorillonite. They
reported that oxalate and P ions competed strongly for the same adsorption sites
on Al oxides under both acidic and neutral conditions. Lan et al. (1994) studied
the release of P by oxalate from different spodic horizons of different soils. These
soils desorbed variable amounts of P, although each contained about the same
amount of total P. The differences in P release may have been related to
mechanisms by which the P was held by different soil components (crystalline
minerals, an amorphous fraction, occluded P, or organo-mineral complexes) and/or
the interaction of oxalate with these components.
Oxalate may influence P availability through its reaction with Al both in
solution and at the surfaces of Al oxides as it forms stable complexes with Al
(Martell et al.. 1988). Release of P by oxalate could be by direct ligand exchange
and/or dissolution of Al oxide mineral surfaces (Stumm and Morgan, 1981; Huang
and Violante, 1986). Stumm (1986) proposed a theory of congruent dissolution of

9
mineral surfaces by low molecular organic anions. He suggested that organic
ligands form complexes with A1 and Fe both in solution and on the surfaces of
oxides. The formation of complexes in solution by organic ligands increases the
rate of dissolution of Al-oxalate complexes on surfaces. Ligand exchange at the
surface decreases the strength of A1-OH-A1 bonds and helps bring Al-oxalate
complexes into solution. This rupture is thought to be induced by the competition
of oxalate with the hydroxyls linking the aluminum atoms. Stumm (1986) further
proposed that surface complexation is a prerequisite for nucleation for most solid
crystals formed in solution as well However, there is no experimental evidence in
the literature to support the above theory, even with the use of pure clay minerals
or A1 and Fe oxides
The complexation and sorption reactions of humic and fulvic acids with A1
and Fe are similar in magnitude to those for the corresponding monomeric units
(e g. carboxylate, salicylate, dihydroxybenzoate) believed to contain functional
groups similar to those in fulvic and humic acid (Stumm and Morgan, 1981;
Stevenson and Fitch, 1986; Tan, 1986). Inoue and Wada (1971) concluded that
the adsorption of fulvic and humic acids by allophane and imogolite resulted from
the ligand exchange of oxygen in carboxylic groups with structural oxygen in the
coordination shells of A1 and Fe atoms. This suggests that organic carbon may
block some of the adsorption sites which might otherwise be available for the
sorption of P and organic anions. There are no studies to my knowledge which

10
examine the influence of natural soil organic carbon on the retention and release of
organic anions and P in Spodosols.
A significant portion of the total P in a spodic horizon is inorganic, but not
readily available for plant use. Phosphorus uptake by plants is influenced by the
amount of root surface area, root growth rate, and P concentration in solution at
the root surface (Barber, 1984). Phosphorus supplying capacity of a soil is
determined not only by the amount of P in the soil solution but also by the soil's
ability to replenish P lost from soil solution and/or during transport of P to plant
roots via diffusion and/or mass flow. The soil's ability to replenish P lost from
the soil solution depends partially upon release of P from sorbed and insoluble P
Thus, it is important to understand how P is being held by different colloidal
surfaces as well as the influence of various chemical processed on the release of
previously sorbed and insoluble P.
From the aforementioned, it can be concluded that there is a need to
understand how forest trees acquire nutrients from the soil profile in lower Coastal
Plains. We can start to address this question by i) understanding the types of
sites available and binding mechanisms for oxalate via spectroscopic studies; ii)
investigating P sorption and desorption in the presence of oxalate in spodic
horizons; and iii) defining the influence of soil organic colloids on the retention
and release of P and oxalate. Understanding the sorption mechanisms for P in the
presence of oxalate should form the basis for explaining P extraction by roots from
soil surfaces, and the buffering of P levels in soil solutions.

CHAPTER 2
INFLUENCE OF SOIL ORGANIC MATTER AND pH ON OXALATE
SORPTION ONTO A SPODIC HORIZON
Introduction
Natural organic anions in soils and freshwater environments are derived from
plant and animal residues, microbial metabolism and canopy drip. Simple aliphatic
anions such as citrate and oxalate are continuously produced through the activities
of microorganisms (Flaig, 1971), aqueous extraction of leaves (Bruckert, 1971;
Stevenson, 1982), and activities of microorganism and/or roots in the rhizosphere
(Reddy et ah. 1977). The type of vegetation, pH, oxidation potential, water
potential and temperature each affect the kinds and amounts of low molecular
weight organic anions that are produced (Stevenson, 1982; Goh and Haung, 1986;
Hue et al. 1986; Pohlman and McColl, 1988).
Among the aliphatic anions detected in the extracts of litter and soils,
oxalate occupies an important place in many ecosystems, both in well-, and poorly-
drained soils (Vedy and Bruckert, 1982). For many organisms, oxalate is a
metabolic end product of low residual nutrient energy (Cromack et al.. 1979).
Smith (1969) and Fox and Comerford (1990) measured significant concentrations of
11

12
oxalate (up to 2 mM) and other carboxylic anions in a variety of sediments, in the
forest floor, in bulk soil, and in the rhizosphere of forest trees.
Oxalate alters chemical processes in soils through complexation reactions
with A1 and Fe that occur in the soil solution and on the surfaces of soil particles
(Stumm, 1986; Martell et al.. 1988). It has acidic characteristics due to the
presence of -COOH groups and reacts with mineral soil surfaces through i)
electrostatic attractions, ii) complex or chelate formation, and iii) water bridging
(Tate and Theng, 1981). Three types of inner-sphere oxalate complexes with soil
surfaces are possible: monodentate, bidentate, and binuclear (Parfitt et al.. 1977).
Each requires the displacement of OH2 or OH' that had been coordinated to Al
and Fe atoms on the surfaces. Such a reaction would increase the pH of the
system. Therefore, the amount of OH' released could be used to identify possible
reactive sites and possible mechanisms of oxalate sorption at different
concentrations of oxalate and/or pH. A change in pH would alter the soil surface
charge and, possibily, the species of organic ligand present in the system.
Spodosols of the southeastern Coastal Plain have an accumulation of
organic matter in their spodic horizons. As organic anions originate in the forest
canopy or forest floor, they form mobile complexes with Fe and Al, and move to
the Bh horizon where they are constrained as organo-minerals. Researchers have
shown that organic matter may be immobilized through complex interactions with
mineral surfaces (Greenland, 1971; Davis and Glour, 1981; Sibanda and Young,

13
1986). The reactivity of labile soil organic matter is generally attributed to the
presence of carboxyl (-COOH) and phenolic hydroxyl (-OH) groups.
The major mechanisms by which soil organic matter is sorbed to mineral
surfaces include i) anion exchange, ii) ligand exchange-surface complexation, iii)
hydrophobic interaction, iv) hydrogen bonding, and vi) cation bridging (Sposito,
1984). Parfitt et al. (1977) suggested that carboxyl groups of iulvic and humic
acid replace surface OH' from gibbsite, goethite and imogolite. Ligand exchange
of surface-coordinated OH' and OH2 from Fe oxides by humic substances had also
been noted by Tipping (1981b) Jardine et al. (1989) suggested that the primary
mechanism of dissolved organic carbon sorption is physical adsorption followed by
an anion exchange mechanism. This suggests that organic carbon should alter the
surfaces charge, surface area and retention properties of inorganic colloids in spodic
horizons.
According to Stumm (1986), organic ligands form complexes with Al and
Fe in solution and on the surfaces of Fe and Al oxides. The formation of such
complexes in solution increases the rate of dissolution of surface complexes.
Huang and Violante (1986), Miller et al. (1986), Tan (1986), and Pohlman and
McColl (1986) also reported that organic anions which formed complexes with Al
and Fe stimulated the dissolution of Al and Fe solid phases. Ohman and Sjoberg
(1988) proposed that polycarboxylic acids are effective in solubilization of Al
surfaces They proposed this as a series of complexation-hydrolysis reactions:

14
A1(0H)3(S) 4= Alq(OH)pLr(S) Alq(OH)pLr(aq) 4= AlLn(aq) 4= Al3+(aq) (2.1)
where S represents the mineral surface, L is the organic ligand, aq is the solution
phase and p, q and r are the number of atoms in each molecule. The calculations
of soil solution speciation with kaolinite as the solid phase, showed that, under
acidic conditions, solution concentration of A1 would increase in the presence of
oxalate. This increase was attributed to the formation of [A1L]T and [A1L2]' species
in solution. Fox et al. (1990a) observed higher release of A1 into solution with
the adsorption of oxalate, and attributed this to the dissolution of solid-phase Al
surfaces in Spodosols. Bilinski et al. (1986) found that the binuclear complex
A12(0H)2(C204)44' formed in solution when oxalate concentration was high, and then
precipitated as Al oxalate with the oxalate ion acting as a bridging ligand between
two aluminum atoms.
The objectives of the present study were to describe i) oxalate sorption on
the clay fraction of a spodic horizon, ii) the effect of soil pH on oxalate sorption
capacity, and iii) the influence of soil organic matter on oxalate sorption using
fourier transform infrared (FTIR) spectroscopy.
Materials and Methods
Soil Material
Soil was collected from the spodic horizon (Bh) in a soil pit of a Pomona
series (sandy, siliceous, hyperthermic Ultic Alaquod) at the Gator Nationals Forest
site located in Alachua County, 10 km northeast of Gainesville, Florida (Swindel et

15
al., 1988). The soil material was air dried, passed through a 2-mm sieve and
stored in plastic bags.
Preparation of Whole Soil
Two sets of soil samples were prepared: one with organic matter and the
other without organic matter. One portion was treated with hot, 30% H202 to
remove organic matter as outlined by Kunze and Dixon (1986). This sample was
centrifuged and washed several times with distilled water, while the other portion
was not treated Both soil samples (with and without organic matter) then were
saturated with Na using 0.01 M NaCl Excess Na was removed with four or
five centrifugal washings using distilled water. The samples were oven dried at
110 C.
Preparation of the Soil Clay Fraction
The water-dispersible clay fraction was separated by ultrasonification at a
1:5 soil to water ratio (Genrich and Bremner, 1974). After ultrasonification, wet-
sieving was used to remove the sands. The clay fraction (< 0.2 pm diameter)
was separated by centrifugal sedimentation (Jackson, 1979). The resulting clay was
divided into two portions. One portion was treated with hot, 30 % H20, to
remove organic matter as outlined by Kunze and Dixon (1986). The other portion
was not treated. Both clay samples (with and without organic matter) then were

16
saturated with Na using 0.01 M NaCl Excess Na was removed with four or
five washings using distilled water. The samples were oven dried at 110 C.
Mineralogy of the Clay Fraction
Mineralogy of the clay was determined by X-ray diffraction (XRD) on
parallel-oriented specimens, using Cu-Ka radiation. Approximately 250 mg of clay
were deposited on a ceramic tile under suction followed by K and Mg saturations
on the tiles by washing with the respective chloride salt and rinsing. Glycerol then
was added to the Mg saturated samples. Samples were scanned at 2 20 per
minute, at 25 C for both the K- and Mg-saturated samples. Minerals were
identified from XRD peak positions, making use of differentiating responses to ion
saturation. Detectable minerals in the clay sample included quartz, kaolinite,
hydroxy interlayered vermiculite, and gibbsite.
Oxalate Sorption Studies
Sorption experiments were carried out with different concentrations of
oxalate over a pH range (3.5, 4.5 and 5.5). Triplicate samples of clay (250 mg
with and without organic matter) were placed in 30 ml bottles with 25 ml of
solution having an oxalate concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM
adjusted initially to pH 3.5, 4.5 or 5.5 (using 0.1 M HC1 or NaOH). Triplicate
five-gram whole-soil samples from each treatment (with or without organic matter)
were placed in 100 ml bottles with 50 ml of solution having an oxalate

17
concentrations of 0 mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM or 10 mM
adjusted initially to pH 4.5 (using 0.1 M HC1 or NaOH).
Two drops of toluene were added to inhibit microbial growth and all
samples were placed in a reciprocating shaker for 24 hrs. The pH of the
suspension was periodically (every three hours) adjusted with 0.1 M HC1 or
NaOH over the 24 hrs of the experiment, to maintain the initial pH. At the end
of the reaction period, each suspension was centrifuged at 12,000 g for 20
minutes. The adsorbate was used for FTIR analysis and the filtrate was analyzed
for oxalate, Al, Fe, and organic carbon. Sorbed oxalate was calculated from the
difference between the initial and final oxalate concentration in solution.
Chemical Analysis
Solution pH was measured using a combination glass electrode and an
Orion pH meter. Aluminum was determined via flame emission spectrophotometer
with a N20-C,H, flame. Fe was also determined, with an acetylene flame. Total
organic carbon content of the clay suspension (with organic matter) was measured
via persulfate oxidation and IR analysis of CO, produced on a TOC apparatus
(College Station TX). The E4 (A=465) to E6 (A=665) ratio of organic carbon
released in solution by clay with organic matter was measured using a Shimadzu
UV/Vis spectrophotometer.
Oxalate in the extract was determined by HPLC (Fox and Comerford,
1990a) using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton

18
Co., Reno, Nevada) with a Gilson single piston high pressure pump along with a
Pheodyne model 7125 injection valve fitted with a 20 ¡uL injection loop. The
HPLC system used a Gilson Holochrom variable wavelength UV detector in
conjunction with a Gilson computerized integrator. The eluent was 0.005 M
H2S04 at a flow rate of 2 mL min"1 Oxalate concentration was calculated from
the calibration curve obtained with standard solutions of 0.1 to 10 mM.
Titration Curves for the Clay Fractions and the Whole Soil Material
To determine the amount of OH released into solution during the sorption
of oxalate, titration curves were prepared (Duquette and Hendershot, 1993) for
each of the clay and whole-soil samples (i. e., with and without organic matter).
Duplicate one-gram samples of clay and 10-gram samples of soil were equilibrated
with 100 mL of water adjusted to pH 3.0. The suspension was allowed to
equilibrate for 24 hrs. Titration was carried out with 0.1 N NaOH from pH 3.0
to 7.0. With each addition of 0.1 ml NaOH, the solution was allowed to
equilibrate for 2 to 5 minutes or until pH stabilized. The titration curves, for
each type of clay and soil are presented in Appendix A-l and Appendix A-3. pH
buffer curves were calculated which relate the change in pH with unit additions of
NaOH for three different ranges of pH These ranges were 3.44 to 4.44, 4.44 to
5.65, and 5.45 to 6.75. Models were fit to these curves and used to calculate the
change in OH concentration required to change pH of the system to specific values
(Appendix A-2 and Appendix A-4).

19
FTIR Analysis
FTIR spectra were obtained by placing small amounts of clay suspension on
AgCl windows and allowing the deposit to dry. Spectra then were collected on a
Bomem DA3.10 spectrophotometer equipped with a MCT detector and a KBr
beam splitter, operating at 2.0 cm'1 resolution. The Bomem DA3.10
spectrophotometer was controlled through a general-purpose interface bus (IEEE-
488) interfaced to a DEC Vaxstation-II computer.
Adsorption Isotherms
The Langmuir adsorption model was fitted to all sorption data. The
following equation was used:
S=
S KC
m
1+ KC
(2.2)
Where S is the amount of oxalate taken up per unit mass of soil (mM kg"1), Sm
is the maximum amount of oxalate that was bound, C is the equilibrium
concentration of oxalate (mM), and K is a constant related to the binding energy
for oxalate. The parameters (K and Sm) were calculated by a least squares fit to
a linear form of the equation:
KS.
M
(2.3)

20
Speciation Calculations
The metals speciation model MINTEQA2 version 3.2 was used to calculate
the species of A1 and oxalate in solution, assuming that the system was under
equilibrium conditions. Calculations were carried out at solution concentrations of
oxalate and A1 for the highest loading rate of oxalate (10"2 M oxalate) at different
pH values. Log K values for soluble Al-oxalate complexes were obtained from
Thomas et al. (1991). Log K was assumed to be 6.12 for A1C204+ and 11.15 for
A1(C204)2\ Two solid phase species of Al-oxalate were considered in the
speciation model. The formation constants were taken from Bilinski et al. 1986.
The values of constants used were 21.87 for A13(0H)7(C204).3H20 and 5.61 for
NaAl(0H)2(C204).3H20. Formation constants used for calculations of other species
were those of the MINTEQA2 data base taken from Martell and Smith (1982).
Statistical Analysis
Statistical differences between different regression lines at each pH value
were tested using the General Linear Models procedure of the SAS framework
(SAS Institute, 1985). The model used was:
7,= + pX+e.. (2,4)
where a is the intercept, P is slope of the line, and e is the error term To
compare two regression lines, we compared the a and P values of the respective

21
lines. The ANOVA and t tests were used to test the differences between a, and
a2 as well as between p, and P2. Significance of the difference between means
was determined by the t-test (Snedecor and Cochran, 1980). Standard deviations
or standard errors of means are given as well, when appropriate.
Results
Characterization of the Clay Fraction using FTIR
FTIR spectra of clay samples with and without organic matter are presented
in Figure 2.1. Characteristic infrared bands for quartz, kaolinite, HIV and gibbsite
minerals are identified. The broad OH-stretching band in the region 3000 to 3600
cm'1, along with well defined OH-stretching is characteristic of kaolinite, HIV and
gibbsite, while the broad deformation band for water in the 1640 cm'1 region
suggests the presence of amorphous clay. The spectra of clay with organic
matter shows major absorption bands as well in the regions of 3300 cm"1 (O-H
and N-H stretching), 2960/2925 cm'1 (aliphatic C-H stretching of CH3 and -CH2
groups), 1610 cm'1 (aromatic C=C and/or H-bonded C=0 stretching of COOH),
and 1400 (OH deformation, C-0 stretching of phenolic OH, and C-H deformations
of CH3 and CH2) together with the characteristic infrared bands for quartz,
kaolinite, HIV and gibbsite minerals. Intensity of the broad OH-stretching band in
the region 2700 to 3600 cm'1 for clay with organic matter was greater, reflecting
the presence of OH-stretching associated with organic matter.

Absorbance
22
Wavenumbers
Wavenumbers
Fig. 2.1 FTIR spectra of clay (a) without organic matter and (b) with organic
matter

23
Sorption of Oxalate by the Clay Fractions
Sorption isotherms for oxalate on clay samples with organic matter and
without organic matter at constant pH are presented in Figure 2.2. Linear
isotherms provided best fit for the data within the range of oxalate concentrations
used. Parameters of the model at different pH values are presented in Table 2.1,
with oxalate sorption being strongly dependent on pH. As pH increased, the
oxalate sorption decreased. Soil organic carbon also had a significant influence on
oxalate sorption at all pH values, with approximately a 2 to 3 fold reduction in
oxalate sorption in the presence of organic carbon (Table 2.1).
FTIR spectra at the highest oxalate loading are presented in Figure 2.3 for
clay without organic matter and in Figures 2.4, 2.5, and 2.6 for clay with organic
matter A small shoulder was seen at 1710 cm"1 for the FTIR spectra of clay
with organic matter for the 10'2 M oxalate loading at both pH 3.5 and 4.5.
Parfitt et al. (1977) also observed the same bands for oxalate sorption by goethite
and proposed the formation of a binuclear complex between oxalate and two Fe34
ions. If inner-sphere complexes form between oxalate and clay surfaces, they must
involve the coordination of carboxylic groups from the oxalate ion with Al and/or
Fe atoms of the clay surface, through oxygen atoms. However, the absorption
bands linking Al-oxalate vibrations, which would normally provide the most direct
information on complexation, are too weak to be definitive. It was not possible to
isolate these bands from the FTIR spectra of the clay due to the low surface
concentrations of the oxalate.

Sorbed oxalate (mmol kg )
24
3
+ OM 3.5
+ OM 4.5
+ OM 5.5
B
--A--
-G--
- OM 3.5
- OM 4.5
- OM 5.5
- -
A
0.3
0 2 4 6 8 10
Equilibrium concentration of oxalate (mM)
Fig. 2.2
Isotherms for the sorption of oxalate by clay fractions at different
pH values.

25
Table 2.1 Parameters of the linear regression models for oxalate sorption by the
clay fractions and the whole-soil samples.
Organic matter
pH
Oxalate sorption
Slope
Intercept
R2
With Organic Matter
3.5
Clav Fractions
64.3
21.3a
0.99
4.5
30.9e
17.8a
0.97
5.5
15.9f
13.6a
0.96
Without Organic Matter
3.5
209.0a
20.4a
0.99
4.5
101 4b
11.9a
0.99
5.5
46.5d
9.2a
0.99
With Organic Matter@
36.8b
17.6a
0.79
Without Organic Matter1
119 Ia
13.9a
0.77
Whole Soil Samples
With Organic Matter
4.5
0.070b
0.044a
0 94
Without Organic Matter
4.5
0.095a
0.036a
0.95
The same letter in a column at different pH values indicates a lack of significance
at the 5 % level within the clay fractions and the whole-soil samples.
@ Includes the combined samples at pH 3.5, 4.5, and 5.5;

Absorbance Absorbance
26
Wavenumbers (cm-1)
-10-2 M Oxalate 0.00 M Oxalate
Wavenumbers (cm-1)
Fig. 2 3
F I'IR spectra of clay without organic matter after oxalate sorption at
pH 3 5 (A), 4 5 (B), and 5 5 (C).

10'J M Oxalate
KT* M Oxalate
0 00 M Oxalate
FTIR spectra of clay with organic matter for different concentrations
of oxalate sorbed at pH 3 3.

:s
0 00 M Oxalate
Fig. 2 5 FTIR spectra of clay with organic matter for different concentrations
of oxalate sorbed at pH 4 5.

:9
10J M Oxalate
1CT1 M Oxalate
Fig. 2 6
FTTR spectra of day with organic matter for different concentrations
of oxalate sorbed at pH 3.3.

30
Oxalate Sorption by Whole Soil Samples
Oxalate sorption isotherms for the whole-soil samples are presented in
Figure 2.7. The Langmuir model gave the better fit. The parameters of the
linearlized form of the Langmuir models are presented in Table 2.1. The sorption
maxima (Smax) by the whole soil material was 10.57 for soil with organic matter
and 14.20 for soil without organic matter The presence of soil organic matter
significantly reduced oxalate sorption.
Release of OH
The sorption of oxalate onto the clay fractions and whole-soil samples
released large amounts of OH ions (Fig.2.8 and Fig. 2.9). Oxalate sorption was
significantly and positively correlated with OH ions released (Table 2.2). The OH'
release was also significantly influenced by soil organic matter (Table 2.2). The
amount of OH ions released at pH 3.5 into solution by the clay fractions was
significantly different than the amount of OH ions released into solution at pH 4.5
and 5.5. The FTIR spectra of clay without organic matter showed considerable
decrease in the intensity of OH-bands in the region 3000 to 3800 cm'1 following
oxalate sorption (Fig. 2.3).
The OH release in solution follows first order kinetics (Figure 2.10) for
oxalate sorption by clay surfaces at pH 4.5. Release of OH was initially rapid
during the sorption of oxalate (Fig. 2.10), and remained high even after 24 hrs
when organic matter was not present. The ratios of moles of OH' released per

Oxalate sorbed (mM kg )
31
Equilibrium Oxalate Concentration (mM)
Fig. 2.7
Isotherms for the sorption of oxalate by the whole-soil samples.

OH-ions released (mmol kg )
32
Release of OH ions by the clay fractions at varing pH as a function
of oxalate sorption.

33
Fig. 2.9 Release of OH ions into solution during the sorption of oxalate by
the whole-soil samples.

34
Table 2.2 Parameters of the linear regression models relating release of OH ions
and organic carbon for the clay fractions and whole-soil samples.
Material
OM/pH
OH
ions released
Organic carbon released
Slope
Intercept
R2
Slope
Intercept
R2
Clav fractions
With organic
matter
3.5
0.65a
13.73a
0.99
0.081a
2.52a
0.99
4.5
1.65b
- 4.07a
0.99
0 13 lb
6.34b
0.99
5.5
1.88b
1.30a
0.98
0.249c
5.85b
0.99
Without organic matter
3.5
0.7 Ia
58 4b
0.97
4.5
1 82b
73.8b
0.99
5.5
1 73b
40.6b
0.99
Whole-Soil
With organic
matter
4.5
1 99b
-0.07b
0.95
0.73
-0.43
0.93
Without organic matter
4.5
2.48a
-0.14a
0.96
The same letter in a column at different pH values indicates a lack of significance
at the 5 % level within the clay fractions or the whole-soil samples.
@ Includes all samples at pH 3.5, 4.5, and 5.5;

35
Fig
10 Kinetics of OH' ions release during the sorption of oxalate by the
clay fractions at pH 4 5.

36
mole of oxalate sorbed are presented in Table 2.3. Average molar ratio at the
higher sorption rates was not significantly different from 1.0 at pH 3.5, nor was it
significantly different from 2.0 at pH 4.5 and 5.5 for both the clay and whole-soil
surfaces.
Release of Aluminum and Iron
The influence of oxalate sorption on the release of A1 is presented in
Figures 2.11 and 2.12. Aluminum release increased linearly with sorption of
oxalate (Table 2.4). The release of A1 from whole-soil samples followed a
sigmoidal shape, both for samples with and without organic matter (Figures 2.12).
Soil organic matter significantly increased A1 release, but the release of A1 from
the clay fractions without organic matter at pH 3.5, 4.5, and 5.5 was not
significantly different (Fig. 2.11). Speciation of A1 solution in the presence of
oxalate was calculated using MINTEQA2. The majority of the A1 in solution was
present as [A1L]+ and [A1L2]' species at pH 4.0 and above (Figure 2.13). The
system was undersaturated with respect to both NaAl(0H)2(C204).3H20 and
A13(0H)7(C204).3H20 The concentration of [A1L2] species increased with pH.
The Spodosol used in the study had very low concentrations of total Fe
Iron released from the clay fractions and the whole-soil samples is presented in
Table 2.4. The clay fractions and the whole-soil samples with organic matter
released a significantly greater amount of Fe then materials without organic matter.

37
Table 2.3 Ratios of OH' ions released to oxalate sorbed for the clay fractions
and whole-soil samples (mean SD, n=3).
pH
OH released/Oxalate sorbed
With Organic Matter
3.5
Clav fractions
0.980.03a
4.5
1.730.11b
5.5
1 880.22b
Without Organic Matter
3.5
0 860.06a
4.5
1.960.1 lb
5.5
1 850.08b
With Organic Matter
4.5
Whole-Soil Samples
1.920.21b
Without Organic Matter
4.5
l,860.17b
Superscript "a" and "b" indicate that values are not significantly different from 1.0
or 2 respectively at the 5% level.

Al released (mmol kg )
38
Oxalate sorbed (mmol kg )
Fig. 2.11 Release of A1 from the clay fractions at varing pH as a function of
oxalate sorption.

Al released (mmol kg )
39
Oxalate sorbed (mmol kg^)
Fig. 2.12 Release of A1 from the whole-soil samples as a function of oxalate
sorption

40
Table 2.4 Parameters of the linear regression models relating release of aluminum
and iron with oxalate sorption for the clay fractions and whole-soil
samples.
Material
Al release
Fe release
OM/pH
Slope
Intercept
R2
Slope
Intercept
R2
Clav Fractions
Without organic
matter
3.5
0.57a
40.62a
0.99
0.007a
-0.21a
0.99
4.5
0.59a
8 62a
0.99
0.009b
-0.22a
0.99
5.5
0.74a
-1.03a
0.99
0.012
-0.15a
0.99
With Organic Matter
3.5
1.83b
7.11a
0.97
0.008a
-0.19a
0.98
4.5
2.77
-15.30a
0.99
0.019d
-0.27a
0.97
5.5
4.27d
-28 66a
0.98
0.036
-0.13a
0.98
With organic matter
2.96a
-12.27a
0.93
0.021a
-0.19a
0.73
Without organic
matter
0.63b
16.07b
0.98
0.010b
-0.19a
0.96
Whole-Soil Samples
With organic matter
4.5
0.76b
2.16a
0.96
0.012b
0.025b
0.96
Without organic
4.5
matter
3.07a
-2.75
0.95
0.042a
-0.061a
0.89
@ Includes all samples at pH 3.5, 4.5, and 5.5;
The same letter in a column at different pH values indicates a lack of significance
at the 5 % level within the clay fractions and the whole-soil samples.

41
120
100
C
.2 80
3
O
(0
- 60
(0
a> 40
o
a>
a
20
Al
3+
Al-(oxalate)
Al-(oxalate),
1
3.5
4.0
4.5
5.0
pH of the system
xxjx:x
6.0
Fig. 2.13
Concentrations of aluminum species in solution at different pH levels
in the presence of oxalate.

42
Release of Organic Carbon
Organic carbon was released from the clay fractions and whole-soil samples
in a linear fashion relative to oxalate sorption (Fig. 2.14 and Fig. 2.15). The
amount of organic carbon desorption is presented in table 2.2. Seven to ten times
more organic carbon was desorbed by 10 mM oxalate as compared to water. The
pH of the system also greatly influenced organic carbon desorption, with
significantly more organic carbon released at pH 5.5 than at pH 3.5 (0.08 mg g'1
versus 0.249 mg g'1 ). E4/E6 ratios of the organic carbon released into solution
during sorption of oxalate are presented in Table 2.5. The E4/E6 ratio, at the
highest loading rate of oxalate (10 mM), decreased to 1.88. There was a linear
relationship (R2 = 0.99) between organic carbon release and A1 released at pH 3.5,
pH 4.5 and 5.5 for the clay fractions (Figs. 2.16 and 2.17).
Spectra for the clay fractions with organic matter showed major absorption
bands representing organic carbon in the regions of 3300 cm'1 (O-H and N-H
stretching), 2960/2925 cm'1 (aliphatic C-H stretching of CH3 and -CH2 groups),
1610 cm"1 (aromatic C=C and/or H-bonded OO stretching of COOH), and 1400
cm'1 (OH deformation, C-0 stretching of phenolic OH, and C-H deformations of
CH3 and CH2,) Intensities of the COOH and CH3 bands in the spectral region
1610 cm'1 and 1400 cm'1 were calculated at all pH values and at all equilibrium

Organic Carbon Release (mg g )
43
Release of organic carbon by the clay fractions at varing pH as a
function of oxalate sorption

44
Fig. 2.15
Release of organic carbon from the whole-soil samples as a function
of oxalate sorption.

45
Table 2.5. E4/E6 ratio of organic carbon released into solution during the
sorption of oxalate by the clay fractions (Mean SD; n=3).
Treatment
Concentration of
oxalate added (mM)
E4/E6 ratio
3.5
10*
3.150.10
1.0
7.540.25
0.1
8.240.56
0.0
9.260.07
4.5
10
1 880.13
1.0
7.130.04
0.1
8.080.13
0.0
9.070.15
5.5
10
2.140.06
1.0
6.930.05
0.1
7.650.21
0.0
9.070.15
* Concentration of oxalate added into solution

Al release (mM kg )
46
Fig. 2.16
Relationship between A1 release and organic carbon release by the
clay fraction at varing pH.

47
Fig. 2.17
Relationship between A1 release and organic carbon release from the
whole-soil samples.

48
concentrations of oxalate. The integrated absorption intensities was normalized with
respect to the weight of the clay sample deposited on the AgCl window. Intensity
of the 1610 cm'1 (aromatic C=C and/or H-bonded C=0 stretching of COOH and
CH3) bands decreased as progressively more oxalate was sorbed by the clay
surfaces (Table 2.6, Fig. 2.4, 2.5. and 2.6). The greatest decrease in the intensity
of bands at 1610 cm'1 (aromatic C=C and/or H-bonded C=0 stretching of
COOH) and 1460 cm'1 (OH deformation, C-0 stretching of phenolic OH, and C-H
deformations of CH3 and CH2) corresponded with the maximum oxalate sorption.
Decrease in the intensity of these bands was higher at pH 5.5, again
corresponding to the pattern of organic carbon release seen in the batch studies.
Release of organic carbon was significantly correlated to the intensity of these
bands, with R2 values of 0.82 to 0.96.
Discussion
Surface Characteristics
Mineralogical analysis using XRD and FTIR data showed that the dominant
crystalline clay minerals in the spodic horizon of this Spodosol were kaolinite, HIV
and gibbsite. FTIR data established that the clay fraction also contained large
amounts of noncrystalline A1 oxide. Fox et al.. (1990a) reported high amounts of
oxalate-extractable (1357 mg/kg) and pyrophosphate-extractable (1379 mg/kg) Al for
the same soil. Therefore, both crystalline and amorphous mineral surfaces were
available for the sorption of oxalate.

49
Table 2.6. Intensity of the absorption bands at 1610 cm'1 (aromatic C=C and/or
H-bonded OO stretching of COOH) and 1460 cm'1 (OH
deformation, C-0 stretching of phenolic-OH, and/or C-H
deformations of CH3 and CH2) with standard deviations, for clay
with organic matter at varing pH values for different amount of
oxalate added (Means SD; n=3).
pH
Oxalate added
mM
Organic carbon
released mg g'1
V 1610 cm'1
V 1460 cm"1
3.5
10.0
35.65a0.60A
60.74i4.93
35.05i3.54
1.0
6.670.44
74.95i3.08
40.32i2.04
0.1
3.930.29
79.68i2.37
43.83il.65
0.0
3.340.22
80.40i4.57
46.85i4.42
r =
- 0.93 r =
- 0.93&
4.5
10.0
38.130.56
54.39il.04
32.14 1.19
1.0
15.44i0.85
72.0il.69
42.15i0.22
0.1
7.260.45
79.18i2.95
43.38i2.19
0.0
5.420.25
83.76i8.86
43.65i3.67
r =
- 0.93 r =
- 0.95&
5.5
10.0
42.09i0.74
39.49i0.19
17.08i0.45
1.0
18.95il.04
61.78i2.07
32.78i3.56
0.1
6.20.63
67.93il.79
37.57i2.39
0.0
5.35i0.61
70.40il.74
39.68i0.60
r =
- 0.98
r = 0.97&
@ Correlation coefficient between organic carbon released and intensity of the
absorption bands at 1610 cm'1 at pH 3.5, 4.5 and 5.5.
& Correlation coefficient between organic carbon released and intensity of the
absorption bands at 1460 cm"1 at pH 3.5, 4.5 and 5.5.

50
FTIR data also indicated the presence of a large number of OH functional
groups. Since this soil has a very low Fe concentration, we assume that the OH
groups were primarily attached to Al. The terminal A1 on the mineral surfaces
should be Al-OH2+ (aqua), Al-OH (hydroxy), or Al-O" (oxo) groups, depending on
pH (Sposito, 1984). Since these spodic horizons have no anion retention capacity,
the number of Al-OH2+ (aqua) functional groups is insignificant. In the pH range
of this study rules out the significance presence of Al-O', therefore Al-OH
(hydroxy) was assumed to be the dominant surface group.
Solution pH determines the oxalate species available for a surface reaction.
The dissociation of oxalic acid is:
HOOC-COOH ^ HOOC-COO' + H* pKt = 1.2 (2 4)
HOOC-COO ^ OOC-COO + H+ pK2 = 4 2 (2.5)
These pK values would result in dissociation of only one COO' group at pH 3.5,
while at pH 4.5 and 5.5 both COO' groups of oxalate would be dissociated.
Therefore, at pH 3.5, oxalate should only form monodentate and/or binuclear
surface complexes, while at pH 4.5 and 5.5, bidentate surface complexes can be
formed.
Oxalate Sorption
Oxalate sorption onto the mineral surfaces may be through ligand exchange
and/or precipitation. If the sorption process were dominated by ligand exchange

51
then we would expect to see several things from these data. First, there should
be a release of OH ions accompanying oxalate sorption (Goldberg and Sposito,
1985). This would not be the case if precipitation of oxalate dominated. Second,
there should be a linear relationship between the amount of oxalate sorbed and OH
ions released into solution. Third, the molar ratio of OH ions released to oxalate
sorbed should be approximately 1 or 2. Depending upon pH, a ratio of one
would suggest a monodentate or binuclear inner-sphere complex, while a ratio of 2
would suggest a bidentate inner-sphere complex. We would expect a ratio of 1 at
pH 3.5, because only one functional group of oxalate is dissociated at this pH and
the formation of monodentate and/or binuclear surface complexes is expected At
pH 4.5 and 5.5 we would expect ratios around 2, suggesting that inner-sphere
complexes are almost exclusively bidentate.
If precipitation of oxalate were the dominant process we would not expect
to see a strong relationship between A1 and Fe release into solution and the
amount of oxalate sorbed as well as a continuous release of OH' ions into solution
with oxalate sorption, even after 24 hrs The MINTEQA2 speciation calculations
indicate that the amount of solid phase species of polynuclear aluminum hydroxy-
oxalate complexes (A13(0H)7(C204) 3H20 and NaAl(0H)7(C204) 3H20) were
insignificant. This implies that precipitation was not the process controlling oxalate
sorption in this system. However, this criterion has to be used with caution, as
these calculations are based on the assumption that system was at equilibrium
which has not been shown to exist in this study.

52
The data presented here overwhelmingly support the process of ligand
exchange. The OH was released in proportion to oxalate sorbed, as shown with
both the batch sorption and FTIR measurements. The molar ratio was not
statistically different from 1 at pH 3.5 or from 2 at pH 4.5 and 5.5 (Table 2.3).
We further conclude that precipitation cannot be a significant factor in oxalate
sorption in this study. A1 was released into solution in relation to oxalate
sorption, showing that oxalate sorption caused A1 dissolution rather then
precipitation. Also, MINTEQA2 calculations suggested that the oxalate formed
soluble Al-oxalate complex in solution and keep it from precipitating. It must be
remembered that this was a Na-saturated system without any Ca present on the
exchange complex to form insoluble CaC204 species. Under field conditions,
spodic horizon pH is about 3 8 to 4.2, with little Ca present on the exchange
complex to interact with native levels of oxalate
These data further support the conclusion that oxalate is sorbed as
monodentate and/or binuclear surface complexes at pH 3.5 ([mmol OH / HC204] =
1), but forms predominantly a bidentate complex at higher pH values ([mmol OH /
C204'2] = 2). These data support the conclusions of Parfitt et al. (1977), who
also showed monodentate inner-sphere complex formation of oxalate on goethite
surfaces at pH 3.4, at high oxalate concentrations, using infrared spectroscopy.
They also suggested formation of a binuclear complex on goethite when oxalate
loading rates were low.

53
Large releases of Al, Fe, and organic carbon during oxalate sorption further
suggest a concurrent dissolution reaction. Fox et al. (1990a) studied the kinetics
of oxalate sorption onto the same soil and reported that oxalate was sorbed during
the first 6 hrs while release of Al continued for at least 48 hrs. These data,
taken together, support Stumm's (1986) theory of congruent dissolution of mineral
surfaces by oxalate following ligand exchange. This constitutes the first such
experimental evidence to support Stumm's hypothesis. Oxalate on the clay and
whole-soil surfaces forms stable bidentate and binuclear complexes. These
complexes involve five- and six-membered rings between the oxalate and Al, which
would decrease the strength of the A1-OH-A1 bonds and bring Al-oxalate
complexes into solution
The clay fractions and whole-soil samples with organic matter released more
Al, respectively than did the clay or soil without organic matter. Another source
of Al released from the clay and soil with organic matter was the Al present as
metal-organic complexes. Lee et al.. (1988b) studied the forms of aluminum in
selected Florida Spodosols and found that more than 75 % of the Al in spodic
horizons was present as Al-fulvate. Aluminum can act as a bridge between soil
particles and organic matter. If oxalate solubilizes Al through the formation of
stable, soluble, Al-oxalate complexes in solution, then this could result in increased
Al release.

54
Effect of Organic Carbon on Oxalate Sorption.
The removal of organic carbon increased the sorption of oxalate. This can
be attributed to either competition between organic carbon and oxalate for same
sorption sites or formation of new sorption sites on mineral surfaces when H202
was used to oxidize the organic carbon These data do not differentiate between
these possibilities; however, either option would enhance the surface area for
oxalate sorption. Zelazny and Quresih (1973) reported that H202 treatment of clay
material from Florida soils enhanced surface area and decreased surface charge.
This evidence suggests that, with the removal of organic matter, surface area
would increase somewhat resulting in greater sorption of oxalate by clay and soil
surfaces without organic matter.

CHAPTER 3
INFLUENCE OF OXALATE AND SOIL ORGANIC MATTER ON SORPTION
OF PHOSPHORUS ONTO A SPODIC HORIZON
Introduction
Phosphorus deficiencies are common on the poorly-drained Spodosols of the
flatwoods region of the lower Coastal Plain of the Southeastern United States
(Pritchett and Comerford, 1983, Comerford et ah 1984). P availability depends on
physico-chemical properties such as P sorption by colloidal surfaces. Organic
ligands are continuously released into the rhizosphere by decaying plants and
animals, through microbial processes, and as root exudates (Stevenson, 1982, Fox
and Comerford, 1990). The supply of P to plants is strongly influenced by the
presence of these organic ligands. Different organic anions have been reported to
modify the sorption of phosphate by soils and soil components (Deb and Datta,
1967; Earl et al.. 1979; Kafkafi et al.. 1988).
Aluminum and iron, either in solution or as crystalline/amorphous soil
constituents, are the principal agents responsible for chemical fixation of phosphate
in acid soils (Yuan and Lavkulich, 1994) Organic anions which are capable of
forming stable complexes with aluminum and iron in solution (Appelt et al.. 1975;
Viotante and Huang, 1985; Trana et al.. 1986a, 1986b; Huang and Schnitzer,
55

56
1986; Comerford and Skinner, 1989) or on mineral surfaces (Pohlman and McColl,
1986; Stumm, 1986; Kafkafi et al.. 1988; Martell et al.. 1988; Fox et al.. 1990b)
can be effective in reducing the P sorption capacity of soils. Phosphate sorption
was reduced in the presence of humic acids, fulvic acids, and low molecular
organic acids while each of these compounds were specifically sorbed onto pure
mineral surfaces (Nagarajah et al., 1968, 1970; Sibanda and Young, 1986; Ryden
and Syers 1987; Moore et al.. 1992; and Violante and Gianfreda, 1993).
However, little work has documented the effect of oxalate and soil organic matter
on the sorption of P using soil materials
The ability of different organic anions to compete with P for sorption sites
on the surfaces of soil components was reported to be greatest at a pH equivalent
to the pKa2 of the organic acid (Hingston et al.. 1967; Hingston et al.. 1972).
Soil organic acids which contain carboxylic (-COOH) and/or phenolic (-OH)
functional groups can bind to oxide surfaces, thereby reducing the number of
surface sites available for P sorption (Yuan, 1980). This also alters electrostatic
charge at the solid surface. Both of these interactions of organic anions are
influenced by the solution pH, relative concentrations of different anions which may
be present and intrinsic affinities of these anions for the mineral surfaces The
pH of the system is important in surface and solution complexation reactions as it
will i) regulate the concentrations of various P and organic anion species which
differ in their affinity for the solid surface, ii) affect the charge density of solid

57
surfaces, and iii) control the competition between OH', P and organic anions for
common adsorption sites (Barrow, 1987; Kafkafi et al.. 1988).
Poorly-drained Spodosols are the dominant soil type in the flatwoods of
Florida's lower Coastal Plain. Among the low-molecular weight organic anions
present in these soils, the most abundant is oxalate (Fox and Comerford, 1990).
Earlier work by various researchers indicated that the specific sorption of phosphate
and organic ligands like oxalate is through ligand exchange (Goldberg and Sposito,
1985). This process should result in the release of OH'. The amount of OH'
released during a sorption reaction in turn depends upon the characteristics of the
surfaces, concentrations of the adsorbing species, and solution pH. The change in
solution concentration of OH' ions can be used to identify the types of complexes
formed on the various solid surfaces.
Organo-mineral complexes have an important influence on the physical and
chemical properties, and reactivity of soil particles (Viotante and Huang, 1984,
Viotante and Huang, 1985; Huang and Schnitzer, 1986). The main clay minerals
present in flatwoods soils are quartz, kaolinite, gibbsite and hydroxy-interlayered
vermiculite (HIV) (Harris et al.. 1987a). Since Florida soils are sandy, these
minerals are present as coatings on the sand particles (Harris et al.. 1987a; 1987b)
The binding materials for these clay-sized mineral particles to sand grains is
reported to be an Al dominated gel-like substance (Lee et al.. 1988a). Lee et al.
(1988b) further indicated that Al acts as a cementing material, probably as Al-
fulvate. The Spodosols of the lower Coastal Plain are known to have high

58
concentrations of amorphous A1 oxides and are classified as Alaquods. Organic
substances which are present as surface coatings and/or acting as cementing
material can significantly affect the retention of P by modifmg the specific surface
area and surface charge of crystalline and noncrystalline minerals in the finer
fractions of the Spodosols. Therefore, it is important to investigate the sorption of
P by this highly reactive component.
Recently, studies were carried out to investigate the sorption of P by
organo-mineral complexes. Violante and Huang (1989) studied the sorption of P
on precipitated-Al products formed in the presence of organic ligands (citrate,
tartrate, malate, aspartate and tannate). They found that the amount, the nature
and the size of organic ligands coprecipitated with Al, as well as the surface
properties of AJ-organic ligand complexes, strongly influenced P sorption capacity.
Haynes and Swift (1989) investigated the effect of pH on P sorption onto Al-
organic matter complexes. They reported that increase in pH greatly increased P
sorption capacity. These studies were performed on pure minerals. However in
soils, many minerals coexist, and were formed in the presence of different organic
coumpounds. Therefore, P sorption properties of these minerals could be
considerably different from those of the pure minerals.
The purpose of the present investigation was to study: i) P sorption by the
clay fraction and the whole-soil samples of a spodic horizon, and ii) the influence
of oxalate and soil organic matter on P sorption in a such system.

59
Materials and Methods
Soil Material
Soil was collected from the spodic horizon (Bh) of a single soil pit of a
Pomona series (sandy, siliceous, hyperthermic, Ultic Alaquod) at the Gator
Nationals Forest site located in Alachua County, Florida. The soil material was
air-dried, passed through a 2-mm sieve and stored in plastic bags. Particle-size
analysis of the sample was carried out using standard methods (Page et al.. 1986).
The soil contained 91 % sand, 8 % silt and 1 % clay.
Preparation of Whole-Soil Sample and the Clay Fractions
The prepartion of whole-soil samples for the sorption experiment was
previously described in Chapter 2. The water-dispersible clay fraction was
separated (Genrich and Bremner, 1974) as reported in Chapter 2. Both the whole-
soil and clay samples were dried at 110 C.
Sorption Studies on Clay Fraction and Whole Soil
Sorption of Phosphate Alone
Sorption experiments were carried out over a range of phosphate
concentrations at pH 4.5 using a batch procedure. Triplicate 250 mg samples of
clay (with and without organic matter) were placed in 30 ml bottles with 25 ml of
solution having a phosphate concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM
adjusted initially to pH 4.5 (using 0.1 M HC1 or NaOH). Triplicate five-gram

60
sample of whole soil from each treatment (with or without organic matter) were
placed in 100 ml bottles with 50 ml of solution having a phosphate concentration
of 0 mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM or 10 mM adjusted
initially to pH 4.5 (using 0.1 M HC1 or NaOH).
Sorption of Phosphate and Oxalate Added as a Mixture
Sorption experiments were carried out at different oxalate to phosphate
molar ratios. The concentration of oxalate used was 1.0 mM with a phophate
concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM for the clay samples. For
the whole-soil sample with and without organic carbon, an oxalate concentration of
1.0 mM was used, with a phosphate concentration of 0 mM, 0.1 mM, 0.5 mM,
1.0 mM, 3.0 mM, 7.0 mM or 10 mM. pH of the clay suspensions and the
whole-soil suspensions was initially adjusted to 4.5 using 0.1 M HC1 or NaOH.
Two drops of toluene were added to inhibit microbial growth, and the
samples were placed in a reciprocating shaker for 24 hrs. pH values of the
suspensions were periodically (every three hours) adjusted to 4.5 with 0 1 M HC1
or NaOH over the 24 hrs of the experiment. At the end of the reaction period,
each suspension was centrifuged at 12,000 g for 20 minutes. The resultant
supernatant solutions were used for analysis of oxalate, Al, Fe, inorganic P, and
organic carbon. Sorbed P and oxalate were calculated from the difference between
the initial and final concentrations in solution. The adsorbate was saved for further
experiments, as described in Chapter 4.

61
Chemical Analysis
Solution pH was measured using a combination glass electrode in
conjunction with an Orion pH meter. Aluminum was determined using a flame
emission spectrophotometer with N20-C2H2 flame whereas Fe was detected with a
C2H2 flame. Total organic carbon content of the clay sample with organic matter
was measured via persulfate oxidation and IR anaylasis of the resultant CO, using
on a TOC apparatus (College Station, TX). E4 (A=465) to E6 (A=665) ratios
of the organic carbon released into solution from the clay with organic matter
were measured using a Shimadzu UV/Vis spectrophotometer.
Oxalate in the extract was determined by HPLC (Fox and Comerford,
1990), using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton
Co., Reno, Navada) along with a Gilson single piston high pressure pump and a
Pheodyne model 7125 injection valve fitted with a 20 /uL injection loop. The
HPLC system used a Gilson Holochrom variable wavelength UV detector in
conjunction with a Gilson computerized integrator. Eluent was 0.005 M H,S04 at
a flow rate of 2 mL min'1 Oxalate concentration was calculated from the
calibration curve obtained with standard solutions of 0.1 to 10 mmol L'1.
Inorganic P in the filtrate was determined by a molybdenum-blue
colorimetric procedure using ascorbic acid as the reductant (Murphy and Riley,
1962). This is an operational definition of inorganic P, for molybdenum may also
hydrolyze some organic P (Stainton, 1980).

62
Titration Curves for the Clay and Soil Samples
To determine the amount of OH' ions released into solution during the
sorption of P and/or oxalate, titration curves were prepared for each clay and
whole-soil sample (with and without organic matter). Duplicate one-gram samples
of clay and 10-gram samples of soil were equilibrated with 100 mL of water
adjusted to pH 3.0. The suspension was allowed to equilibrate for 24 hrs,
titration then being carried out (with 0.1 N NaOH) from pH 3.0 to 7.0. With
each addition of 0.1 ml NaOH, the solution was allowed to equilibrate for 2 to 5
minutes or until the pH stabilized. pH buffer curves, relating the change in pH
with unit addition of NaOH over the pH range 4.40 to 6.00, were calculated
Curves were fit to the data and used to calculated the amount of OH' ions
required to change pH of the soil and clay systems to specific values.
Adsorption Isotherms
The Langmuir adsorption model was fit to the sorption data The following
equation was used:
5 KC
m
1 + KC
(3.1)
where S is the amount of P or oxalate taken up per unit mass of clay or soil
(mM kg'1), Sm is the maximum amount of P or oxalate that was bound, C is the
equilibrium concentration of P/oxalate (mM), and K is a constant related to the

63
binding energy of P or oxalate sorption The parameters (K and Sm) were
calculated by a least square fit of the linear form of the equation:
r 1 c
-=[]+
S KS S
m m
3.2
Speciation Calculations
The metal speciation model MINTEQA2 version 3.2 was used to calculate
the species of A1 and P in the solution It was assumed that the system had
achieved equilibrium conditions. Although the total activity of A1 ions was not
known, the assumed activity of A1 from the congruent dissolution by oxalate was
used (Chapter 2) to estimate the saturation level with respect to an amorphous
aluminum phosphate solid phase. Calculations were carried out at solution P
concentrations corresponding to the highest loading rate of P (10'2 M). The
formation constants used for calculations of species were those of the MINTEQA2
data base taken from Martell and Smith (1982).
Statistical Analysis
The linear form of the Langmuir equation was used to test the differences
between parameters for P sorption. Statistical differences among regression lines
for the desorption of Al, Fe OH, and organic carbon during P sorption were
tested using the General Linear Models procedure of the SAS framework (SAS
Institute, 1985). The model used was:

64
(3-3)
where a is the intercept, P is the slope of the line and e is the error term. To
compare two regression lines, the a and P values of the two lines were
compared. ANOVA and t-test were used to test the difference between a, and
a2 well as between p, and P2. The significance of the difference between two
means was determined using a t-test (Snedecor and Cochran, 1980). Standard
deviations or standard errors of means are given, when appropriate.
Results
Sorption of P
Clay Fractions
Sorption isotherms for phosphate sorption onto clay samples with and
without organic matter are presented in Fig. 3.1. The presence of oxalate and soil
organic matter significantly reduced P sorption. Maximum reduction in P sorption
(about 50 %) was observed when both organic carbon and oxalate were present in
the system. The P sorption data, for the range of concentrations studied,
conformed to the Langmuir equation. Sorption parameters for the Langmuir
equation are presented in Table 3.1.
Whole-Soil Samples
Phosphate sorption isotherms are presented in Figure 3.2. Sorption

P sorbed (mmol kg )
65
Equilibrium P concentration (mM)
Fig. 3.1.
Isotherms for the sorption of P by the clay fractions.

66
Table 3.1. Langmuir sorption isotherm parameters for P sorption onto the whole-
soil samples and clay fractions with and without organic matter.
Treatments
Anion
Langmuir Model
Sorption Intercept
Maxima
SMax (1/b) 1/SmK
Sorption
energy
K
R2
Soil without organic
matter
P
P
29.33
0.051a
0.66
0.88
P+OX
P
20.08b
0.042a
1.18
0.97
Soil With Organic Matter
P
P
17.39b
0.035a
1.62
0.98
P+OX
P
13 83a
0.052a
1.39
0.98
Clay without organic
matter
p
P
588.0d
0.0005a
3.32
0.97
P+OX
P
454.0
0.0022a
2.42
0.99
Clay with organic i
natter
p
P
379.0b
0.0013a
1.90
0.98
P+OX
P
328.0a
0.0047a
0.64
0.99
* Values within columns followed by same letter for the whole-soil or clay
samples are not significantly different at the 5% level

P sorbed (mmol kg )
67
Fig. 3.2
Isotherms for the sorption of P by the whole-soil samples.

68
parameters for the Langmuir model are presented in Table 3.1. In the case of the
whole-soil samples, P sorption maxima were significantly reduced by soil organic
matter. The presence of soil organic matter and oxalate significantly reduced P
sorption by both types of soil material.
P sorption parameters of the clay samples also were recalculated on a
whole-soil basis (assuming a clay % for the whole soil = 1.3 %), and compared to
the P sorption parameters of whole-soil samples. Significantly more P was sorbed
by the whole-soil samples than by the clay samples when reported on a whole-soil
basis (Table 3.2). P sorption by whole-soil samples was 3 to 10 times higher than
that of the clay samples.
Sorption of Phosphate in the Presence of Oxalate
Less P was sorbed by the clay fractions and whole-soil samples in the
presence of oxalate (Table 3.3). The percent efficiency of oxalate in reducing P
sorption was calculated according to the expression of Deb and Datta (1967):
Ox =[l~]*100 (3 4)
* P
o
where Oxe is the efficiency of oxalate in reducing P sorption (%), Pox is the P
sorbed in the presence of oxalate, and P0 is the P sorbed alone. The efficiency of
oxalate in reducing P sorption increased as the concentration of P in solution
increased (Table 3.4).

69
Table 3.2. P sorption onto the whole soil material compared to P sorbed onto
the clay fraction expressed on a whole soil basis (Mean SD; n=3).
Cone of
anions
P sorption (mmol kg'1)
added
+ OM
- OM
Treatment
(mM)
Whole soil
Clay Fraction
Whole soil
Clay fraction
P alone
10
16.7+0.27
4.580.05b
24.6+0.5
7.090.07b
1.0
5.9+0.13a
0.940.01b
7.1+0.06
1.170.01b
0.1
0.9+0.02
0.120.01b
0.8+0.01
0.120.01b
P + Ox
10+1
13.3+0.25
3.520.04b
19.3+0.37
5.950.04b
1.0+1
5.3+0.14
0.79+0.01b
5.8+0.05
0.960.01b
0.1+1
0.78+0.02
0.09+0.01b
0.8+0.01
0.08+0.001'
+ OM = with organic matter and OM is without organic matter
@ Values within each row for each material with organic matter and without
organic matter are significantly different (P = 0.05) if followed by a different letter

70
Table 3.3. Amount of Al, Fe and organic carbon released into solution with P +
oxalate sorption onto the whole-soil samples and the clay fractiions
(Mean SD; n=3).
With Organic Matter
Without Organic matter
Sorbed
Released
Sorbed
Released
P + Ox Oxalate
A1
Fe
O Carbon
Oxalate
A1
Fe
Ratio
(mmol kg"1)
(mg g1)
(mmol kg'1)-
Whole-Soil samples
10:1
4.20.11
8.72
0.044
1.44
5.00.10
5.01
0.079
7:1
4.80.15
6.99
0.034
1.30
5.40.08
5.40
0.082
3:1
5.20.06
7.33
0.027
1.21
5.90.15
5.44
0.087
1:1
5.40.10
7.60
0.029
1.12
6.00.04
6.26
0.101
0.5:1
5.50.11
7.83
0.037
0.97
6.10.05
6.49
0.101
0.1:1
5.50.08
7.87
0.034
0.62
6.10.02
6.64
0.104
Clav fractions
10:1
32.91.88
24.45
0.44
166
22.42.05
14.19
0.23
1:1
38.80.83
18.91
0.31
125
36.73.50
14.46
0.31
0.1:1
42.20.52
1945
0.12
117
50.421.65
14.59
0.30

71
Table 3.4. Effect of oxalate in reducing the sorption of P onto whole-soil samples
and the clay fractions (Mean SD; n=3).
P: Oxalate
ratio
% reduction
in P sorption
With OM
Without OM
Whole soil
10:1
20.73.5aa
21.33.5ba
7:1
22.54.7aa
26.62.9ba
3:1
16.75.7ab
30.2*3.7
1:1
10.8*4.4
18.40.4ba
0.5:1
1.21.0ba
-5.30.5cb
0.1:1
1.20.8ba
-O.OiOJ1*3
Clav fraction
10:1
2 3 2
22*0.5
1:1
16lba
5*0.5bb
0.1:1
22ca
llca
The first superscript letter indicates significant differences within a column and
material, while the second letter indicates significant differences between rows.

72
Release of OH'
OH' ions were released into the system with the sorption of P alone or P
+ oxalate (Fig 3.3 and Fig. 3.4 ). For the clay samples, P and oxalate together
released more OH' into solution than P alone (Table 3.5). The kinetics of OH'
ions release are presented in Figure 3.5 for the P alone, and P + oxalate,
treatments with the clay. Approximately 80% of the OH' release accompanying P
sorption was detected in the first 8 hrs. For clay with organic matter, the release
of OH" accompanying P sorption was initially rapid and reached equilibrium after
14 hrs.
The molar ratio of OH' released per mole of P sorbed was between 0.85
and 1.38 for the clay fractions (Table 3.6). A similar range was observed for the
whole-soil samples. On average, the ratio was not significantly different from 1
for P sorption alone but was significantly higher than 1 for sorption of P +
oxalate
Release of Organic Carbon
There was a linear relationship between the amount of organic carbon
released and the sorption of anions by the whole-soil samples and the clay
fractions (Table 3.5). During sorption of P by the clay or soil, a small amount of
organic carbon was released into solution. The presence of oxalate with the P did
not significantly affect organic carbon release (Table 3.5). The E4/E6 ratio of
released organic matter was approximatly 9.0 during P sorption at different P

OH released (mmol kg )
73
Fig 3.3
Release of OH' ions during sorption of P by the clay fractions

OH released (mmol kg )
74
P or oxalate sorbed (mmol kg1)
Fig. 3.4
Release of OH' ions during sorption of P by the whole-soil samples.

75
Table 3.5. Parameters of the linear regression models for the release of OH' ions
and organic carbon from the whole-soil samples and clay fractions
Treatments
OH' ions released
R2
Organic Carbon released
Slope
Constant
Slope
Constant
R2
Soil without
organic
matter
P
0.66a
2.9a
0.89
P + OX
0.59a
16 lb
0.80
Soil with organic matter
P
1.06b
2.7a
0.95
0.05a
0.59a
0.98
P + OX
1.19b
5.8a
0.94
0.06a
0.71a
0.94
Clay without
organic
matter
P
0.74a
66b
0.99
P + OX
0.9 lb
116c
0.99
Clay with organic matter
P
0.90b
6a
0.99
0.23a
67.3a
0.96
P + OX
1.08
72b
0.96
0.22a
128.8b
0.98
Values within columns followed by the same letter for soil or clay samples are not
significantly different at the 5% level

76
Fig. 3.5
Kinetics of OH release during sorption of anions by the clay
fraction.

77
Table 3 6 Ratios of OH' released to P sorbed for P only and P + oxalate for
the whole-soil samples and the clay fractions (Mean SD, n=3).
OH released/Anion
sorbed
W/OM or WO/OM
Treatment
Whole soil
Clay fraction
With OM
P
P + Ox
*1.030.05a
1610.10
0.920.03a
1 380.08
Without OM
P
P + Ox
0.830.05a
1 530.08
0 850.07a
1240.11
Superscript "a" indicates that values are not signficantly different from 1.0 at the
5% level.

78
loading rates (Table 3.7).
Release of A1 and Fe
A small amount of A1 and Fe was released during the sorption of
phosphate and for the P + oxalate sorption treatment, these amounts were even
greater (Tables 3.3 and 3.8). Ion-activity product calculations using MINTQEA2
V3 11 showed that the system was supersaturated with respect to amorphous
aluminum phosphate at the higher loading rates for P.
Discussion
The dissociation constants (log K, and log K2) of oxalic acid are 1.27 and
4.20. At pH 4.5, both COOH groups of the oxalate will be dissociated, making
the C2042 species dominant. Assuming ligand exchange as the surface reaction this
oxalate species can form either monodentate, binuclear, or bidentate surface
complexes.
The dissociation of phosphoric acid is described by:
hjo^h2po; 1^hpo; 2^po; 3 3.5
where pK, = 2.2, pK2 = 7.2 and pK3 = 12.3. The predominant species of
phosphate at pH 4 5 is H2P04''. Thus, H2P041_ should only form monodentate or
binuclear surface complexes

79
Table 3.7. E4/E6 ratio of organic carbon released into solution during the sorption
of P only, and P + oxalate by the clay fractions (Means SD; n=3).
Treatment
Concentration of
Anions added (mM)
E4/E6 ratio
P
10
9.04+0.10a
1.0
9.310.25a
0.1
9.240.56a
0.0
9.260.07a
P + Ox
10 + 1
9.14+0.06a
1.0 + 1
8.930.05a
0.1 + 1
7.650.21b
Values within columns followed by same letter are not significantly different at the
5% level

80
Table 3.8 Parameters of the linear regression model relating release of aluminum
and iron with P sorption for P only and P + oxalate by the whole-soil
and clay fractions.
Treatments
A1 release
Fe release
Slope
Constant
R2
Slope
Constant
R2
Soil
with organic
matter
P
0.08a
2.05a
0.93
o.oor
0.015a
0.87
P + OX
0.01a
7.66b
0.07
0.0004a
0.031a
0.24
Soil without organic matter
P
0.01a
2.52a
0.69
0.0001a
-0.008a
0.31
P + OX
-0.09a
6.72b
-0.94
-0.002a
0.106a
-0.84
Clay without organic matter
p
0.03a
34.27a
0.94
0.0003a
-0.01a
0.97
P + ox
0.05a
69.01c
0.94
0.0005a
-0.21a
0.96
Clay With Organic Matter
p
0.02a
13 20a
0.96
0.001a
0.19a
0.98
P + ox
-0.01a
54.01b
-0.61
0.001a
0.52b
0.97
Values within columns followed by the same letter for soil or clay samples are not
significantly different at the 5 % level

81
P Sorption
Phosphate can be sorbed by the clay and whole-soil surfaces either by
ligand exchange or precipitation. Ligand exchange should result in the release of
OH' ions into solution (Goldberg and Sposito, 1985). In fact, large amounts of
OH' were released in this study The amount of OH' ions released varied with
the amount of P and/or of oxalate sorbed The average ratio of OH released to
phosphate sorbed was close to unity. When oxalate was also sorbed this ratio
increased to 1.5. Given that the ratio was unity and that the dominant species of
phosphate was the H2P04' ion one could argue that P was forming a
monodendate and/or binuclear inner-sphere complex. The ratio of OH' ions released
to P sorbed was similar for both the whole-soil and clay samples. Recently
Tejedor-Tejedor and Anderson (1990) using CIR-FTIR studied the surface
complexation of phosphate by goethite surfaces between pH 3.5 to 8.0. They
suggested that between pH 3.5 and 5.5, phosphate formed both binuclear and
monodendate complexes with surface Fe (III) Earlier work by Parfitt et al.
(1977), using infrared spectroscopy, also reported the presence of both binuclear
and monodendate surface complexes by P on goethite at pH 4.0.
The molar ratio of OH ions released to P sorbed supports ligand exchange
as the dominant P sorption process onto both the clay and whole-soil samples.
The formation of measureable crystalline aluminum phosphate was not possible,
given the reaction time of this experiment. Veith and Sposito (1977) and Sposito
(1984) reported that reaction times longer than 140 hrs are required for significant

82
discrete crystal growth to occur. This is much longer than the 24 hrs represented
by this study. Sposito (1984) further suggested that the adsorption-dominated stage
of phosphate sorption takes less than about 50 hrs. Calculated ion-activity product
constants using MINTQEA2 V3.ll suggested that, at equilibrium, this system was
supersaturated with respect to amorphous aluminum phosphate, A1(0H)2H2P04. The
formation or precipitation of amorphous aluminum phosphate thus could be
possible. Sorption of phosphate through ligand exchange is known to serve as a
nucleation site for the precipitation of amorphous aluminum phosphate onto the
clay surface. However, formation or precipitation of amorphous aluminum phosphate
(A1(0H)2H2P04) consumes OH' instead releasing them. Our kinetic data on the
release of OH' ions showed no evidence of this. Up to 14 hrs, OH' ions were
continuously released into the solution. This suggests a ligand-exchange reaction.
After 14 hrs, there was no further release of OH' ions into solution from the clay
surfaces with organic matter present while the clay surfaces without organic matter
were still releasing OH'. Thus, for clay surfaces, there was no absolute decrease in
OH' concentration in solution. This suggests that precipiation of amorphous
aluminum phosphate was unlikely. Therefore, these results suggest that ligand
exchange was the dominant reaction for P sorption
Effect of Oxalate and Organic Carbon on P Sorption
Phosphate sorption significantly decreased in the presence of oxalate.
Likewise, P was effective in reducing oxalate sorption (Appendix B), indicating that

83
some of the sorption sites were common for either P or oxalate. These
observations suggest that between 20 to 30 % of the sorption sites on the whole-
soil material were common. Similar results were reported by other researchers.
Competitive sorption studies for P and oxalate on tropical soils (Lopez-Hernandez
et al., 1986) and montmorillonite (Kafkafi et al., 1988) showed that oxalate was
effective in reducing P sorption. Each observed that oxalate masked about 20% of
the sites otherwise available for P sorption.
Soil organic carbon significantly reduced P sorption capacity of both the
clay and whole-soil samples. Only a small amount of organic carbon was released
during P sorption, and the E4/E6 values for organic carbon released were high.
E4/E6 ratios between 8 to 10 are generally indicative of fulvic acid, and ratios from
2 to 5 represent humic acid (Thurman, 1985). The E4/E6 ratio of released organic
carbon following P sorption fit the range for fulvic acid and was similar to that
for control samples. Since the ratio of OH released to P sorbed was close to 1
and the amount of organic carbon released into solution during P sorption was
small. This suggests that P was not effective in replacing organic carbon sorbed
onto the surfaces. The significant reduction in P sorption by mineral surfaces in
the presence of organic matter indicates that organic carbon masked and/or
occupied the active sites for P sorption Characterization of the clay fractions using
FTIR (Chapter 2) revealed the presence of -COOH surface groups associated with
organic matter. Competitive sorption of soil organic matter and boron by soil
material was studied by Marzadori et al. (1991). They observed that soil organic

84
matter appears to be responsible for occluding important adsorption sites on the
surfaces of soil particles which otherwise would be available for boron sorption.

CHAPTER 4
INFLUENCE OF SOIL ORGANIC MATTER ON DESORPTION OF
PHOSPHORUS AND OXALATE FROM A SPODIC HORIZON
Introduction
In Chapters 2 and 3, the sorption of oxalate and P by the clay fractions
and the whole-soil material of a spodic horizon from northcentral Florida were
studied. Phosphate and oxalate are specifically sorbed by replacing the coordinated
-OH groups of A1 present on the surfaces of oxides and clay minerals. As
discussed in Chapter 3, oxalate reduced P sorption onto clay fraction and whole-
soil surfaces. This investigation studied P and oxalate desorption from the clay
fraction and the whole-soil material. Phosphorus sorbed to the soil surfaces can be
taken up by plants after it has been desorbed into the soil solution.
The competitive sorption of organic anions and P on mineral surfaces has
received attention because the presence of organic ligands in the rhizosphere is
thought to influence P fixation and, therefore, the supply of P to the plant (Deb
and Datta, 1967; Lopaz (1974); Appelt et al., 1975; Earl et al.. 1979, Lopez-
Hernandez et al., 1986; Kafkafi et al.. 1988, Martell et al.. 1988; Fox et al..
1990a: Fox et al.. 1990b; Violante et al.. 1991; Violante and Gianfreda, 1993).
Oxalate in forest soils originates from exudates of plant roots and through the
85

86
activities of fungi and bacteria (Stevenson, 1982). Although oxalate is abundant in
the rhizosphere (Fox and Comerford, 1990), only a few investigations have studied
the competitive sorption of oxalate and phosphate. These include studies in
tropical soils (Lopez-Hernandez et al., 1986); and montmorillonite (Violante and
Gianfreda, 1993). It has been proposed that oxalate can release P from Al- and
Fe-hydroxide surfaces through ligand-exchange reactions. Recently, Fox et al
(1990a) studied the kinetics of P desorption by oxalate from spodic and argillic
horizons of a Spodosol. They suggested that P released by oxalate was through
ligand exchange, but did not provided conclusive evidence.
Phosphate desorption studies (Barrow, 1983; Kuo and Pan, 1988; Bakheit
Said and Dakermanji, 1993; Raven and Hossner, 1993) have indicated that a large
portion of the retained P is irreversibly sorbed. In all such desorption studies,
researchers used either 0.01 M CaCl2, Ca(N03)2 or anion exchange resins.
Phosphate desorption isotherms do not normally coincide with P sorption isotherms
(Nye and Tinker, 1977; Kuo and Pan, 1988; Bakheit Said and Dakermanji, 1993).
Lopaz (1974) studied the desorption of P using citrate and found that desorption
could only occur when the displacing anion was specifically sorbed and present in
sufficient concentration in the soil solution. Organic anions like citrate and oxalate
are continuously produced in soil (Smith, 1969) and can be present in high
concentrations (Fox and Comerford, 1990). However, desorption isotherms using
oxalate as an extractant have not been reported to our knowledge.

87
The influence of soil organic matter on P sorption has been investigated by
many researchers. Many have suggested that sorption of P and organic matter
occurs on the same sorption sites. Yuan (1980) studied the sorption of P and
water-extractable soil organic matter by soil and A1 oxides. He reported that most
of the sorption sites for organic matter and P sorption may be different, though
some sites are common for both P and organic matter Fulvic acid was found to
reduce P adsorption considerably. This effect of fulvic acid is apparently due to
the chelating ability of fulvic acid's -COOH and -OH functional groups for A1 and
Fe (Parfitt, 1978; Sibanda and Young, 1986). The results in Chapter 3 showed
that soil organic matter significantly reduced P sorption. However, the mechanism
for P desorption in the presence of organic matter apparently has not be studied
The objectives of this study were to investigate: i) P or oxalate desorption
from the clay fraction and whole soil of a spodic horizon in the presence of
oxalate or P, respectively, and ii) the influence of soil organic matter on the
desorption of P and oxalate
Materials and Methods
Whole-Soil Material and the Clay Fraction
The whole-soil and clay-fraction samples that were used for sorption studies
of oxalate (Chapter 2) and phosphate (Chapter 3) were further used to study the
desorption of anions at pH 4 5

88
Phosphate Desorption by Oxalate
P sorption experiments described in chapter 3 formed the basis for these
samples. In Chapter 3, the solutions used for P sorption by the clay samples had
P concentrations of 0 mM, 0.1 mM, 1.0 mM, and 10 mM. The P sorption
solutions used for the whole-soil samples had P concentrations of 0 mM, 0.1
mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM and 10 mM. At the end of the
reaction period, each suspension was centrifuged at 12,000 g for 20 minutes.
Filtrates then were equilibrated with solution containing 5 mM of oxalate per 25
ml of solution for the clay fractions and per 50 ml of solution for the whole-soil
samples. The suspensions were then shaken for another 24 hrs. The suspension
pH was periodically (every four hours) adjusted to pH 4.5 with 0.1 M HC1 or
NaOH over the 24 hrs of the experiment. Two drops of toluene were added to
inhibit microbial growth, and samples were placed in a reciprocating shaker for 24
hrs. At the end of the reaction period, each suspension was centrifuged at 12,000
g for 20 minutes. The filtrate then was analyzed for oxalate, Al, Fe, inorganic P
and organic carbon. Sorbed oxalate was calculated from the difference between
the initial and final oxalate concentration in solution.
Oxalate Desorption by Phosphate
The oxalate sorption experiments of Chapter 2 served as the basis for these
samples. The solutions used for the initial sorption of oxalate in Chapter 2 had
oxalate concentrations of 0 mM, 0.1 mM, 1.0 mM, and 10 mM for the clay

89
fractions. Oxalate concentrations used for the whole-soil sorption isotherms were 0
mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM and 10.0 mM. At the end of
the reaction period, each suspension was centrifuged at 12,000 g for 20 minutes
Filtrates then were equilibrated with 25 ml of a 5 mM P solution for the clay
fraction or with 50 ml of solution for the whole-soil samples. The suspension
then was shaken for another 24 hrs. pH of the suspension was periodically (every
four hours) adjusted to pH 4.5 with 0.1 M HC1 or NaOH over the 24 hrs of the
experiment Two drops of toluene were added to inhibit microbial growth, and
samples were placed in a reciprocating shaker for 24 hrs. At the end of the
reaction period, each suspension was centrifuged at 12,000 g for 20 minutes. The
filtrate was used for analysis of oxalate, Al, Fe, inorganic P and organic carbon.
Sorbed phosphate was calculated from the difference between the initial and final
phosphate solution concentrations.
Chemical Analysis
Solution pH was measured using a combination glass electrode on an Orion
pH meter. Aluminum was determined using a flame emission spectrophotometer
with N20-C2H2 flame, and Fe was determined using a C2H2 flame. Total organic
carbon of the clay suspension with organic matter was measured using a persulfate
oxidation followed by IR analysis of the CO, produced on a TOC apparatus
(College Station, TX).

90
Oxalate in the extract was determined by HPLC (Fox and Comerford,
1990a), using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton
Co., Reno, Nevada) with a Gilson single piston high pressure pump along with a
Rheodyne model 7125 injection valve fitted with a 20 /uL injection loop. This
HPLC system uses a Gilson Holochrom variable wavelength UV detector in
conjunction with a Gilson computerized integrator. The eluent was 0.005 M
H2S04 at a flow rate of 2 mL min'1. Oxalate concentration was calculated from
acalibration curve obtained with standard solutions of 0.1 to 10 mM.
To determine the amount of OH' released with the sorption of oxalate,
titration curves were prepared for the whole-soil and clay fractions samples (with
and without organic matter) as described in Chapter 3. Models were fitted to
these curves and used to calculate the changes in OH concentration required to
change the pH of the system to specific values.
Adsorption Isotherms
Freundlich model was fitted to the desorption data. The Freundlich model
was obtained by:
S=kC (4
where S and C are defined as previously and k and n are empirical constants.
The linear form of the above equation was used to determine sorption parameters.

91
Speciation Calculations
The metal speciation model MINTEQA2 version 3.2 was used to calculate
the species of Al, P and oxalate in the solution. It was assumed that the system
had achieved equilibrium conditions Log K values for Al-oxalate complexes were
obtained from Thomas et al. (1991). Assigned Log K values were 6.12 for
A1C204 1 and 11.15 for A1(C204)2'. The formation constants used for calculations
of other species were those of the MINTEQA2 data base, as taken from Martell
and Smith (1982).
Statistical Analysis
Statistical differences in desorption of oxalate or phosphate using different
regression lines were tested using the General Linear Models procedure of the SAS
framework (SAS Institute, 1985). Observed desorption of oxalate or phosphate
was related to the amount of P and oxalate previously present on the surfaces,
respectively. The model employed was:
7a + p.(logV.)+ P2(logA/+ e. (4.2)
where a is the intercept, P, and P2 are 1st and 2nd order coefficients of the line
and e is the error term. To compare two regression lines, we compared the a
and p values of the two lines with ANOVA and t-test used to test the differences
between 0^ and a2 as well as between p, and P2. The significance of the

92
difference between means was determined by a t-test (Snedecor and Cochran,
1980). Standard deviations or standard errors of means are given as appropriate.
Results
Desorption of Phosphate in the Presence of Oxalate
Desorption of Phosphate
A considerable amount of phosphate was desorbed during oxalate sorption,
and the presence of organic matter did not significantly influence the desorption of
P by oxalate. P desorption isotherms are presented in Figure 4.1 both for the
clay fractions and the whole-soil samples. Corresponding parameters for the
Freundlich models are presented in Table 4.1.
Sorption of Oxalate
Oxalate sorption by the clay fraction and the whole-soil samples is presented
in Figure 4.2 and 4.3, respectively. The amount of oxalate sorbed increased as
the amount of P previously present on the mineral surfaces decreased both for the
whole soil and the corresponding clay fractions. The relationship between release of
previously sorbed P and amount of oxalate sorbed is presented in Appendix D-l.
As low amounts of oxalate were sorbed in the present experiment, large amounts
of P were desorbed. The amount of oxalate sorption followed a quadratic model.
Parameters of the model are presented in Table 4.2. The amount of oxalate

P retained by soil (mmol kg ) P retained by clay (mmol kg )
93
Fig. 4.1
P desorption curves for whole-soil samples and clay fraction.

94
Table 4.1. Parameters of Freundlich models for the desorption of P from whole-
soil samples and clay fraction in
the presence of 5
mM oxalate.
Type of Surfaces Desorption Constant Bonding Energy
K N
R2
Whole Soil
With Organic matter
10.59a
0 98a
0.85
Without Organic matter
18 08a
1 23a
0.86
Clav Fractions
With organic matter
354a
0.94a
091
Without organic matter
301a
1.03a
090
Letters in column are significantly different at the 5% level for clay and soil
materials.

95
- OM + OM OM + OM
Fig. 4.2
Influence of previously sorbed P on oxalate sorption and P
desorption for clay fraction.
Phosphate desorbed (mmol kg )

Oxalate sorbed (mmol kg
96
-OM + OM -OM + OM
+OX +OX -P -P
O -A-
Phosphate previously sorbed (mmol kg )
Fig. 4.3 Influence of previously sorbed P on oxalate sorption and P
desorption by whole-soil samples.
Phosphate desorbed (mmol kg

97
Table 4.2. Quadratic equation coefficients for the sorption of anions by whole-soil
samples and clay fractions.
Anion
Sorbed
Materials
a
P,
P2
R2
OX
Clay + OMa
160
76.5
-39.4
0.94
Clay OMb
378
-24.9
2.8
0.95
Soil + OMa
11.4
5.63
-10.7
0.92
Soil OMb
14.5
3.67
-7 66
0.97
P
Clay + OMa
283
-76.6
17.5
0.85
Clay OMb
213
-3.89
-3.7
0.74
Soil + OMa
7.1
1.74
-4.76
0.87
Soil OMb
12.6
1.32
-2.41
0 82
+ OM= With organic matter and OM = Without organic matter
Different letters in the same column are significantly different at the 5% level for
the clay and soil samples.

98
sorption was significantly higher for materials without organic matter than with
organic matter.
Release of Hydroxyls
Hydroxyls were released into solution during the sorption of oxalate by both
the whole-soil and clay fraction (Appendix D-l). The ratios of OH' desorbed to
the amount of oxalate sorbed and are presented in Table 4.3. This ratio ranged
between 1.68 and 1.90 for the clay fraction and between 1.64 and 2.01 for the
whole-soil samples. The ratio of OH' desorbed to oxalate sorbed increased as the
amount of P desorbed in solution decreased.
For the clay fraction, ratios of OH + P released into solution to oxalate
sorbed were not significantly different from 2. Ratio was significantly higher than
2 for the whole-soil samples, however, with higher amounts of P previously present
on these surfaces.
Release of Al, Fe, and Organic Carbon
Considerable amounts of aluminum, iron and organic carbon were released
into the solution during oxalate sorption, both for the whole-soil samples and the
clay (Appendix D-l). There was a significant correlation between the amounts of
A1 and Fe released and oxalate sorbed.

99
Table 4.3 Influence of oxalate sorption by clay fractions and whole-soil samples
on the ratio of OH released, and OH + P released, to oxalate
sorbed (Mean SD, n=3).
Without Organic Matter With Organic Matter
P Prev Ratio P Prev Ratio
Present Sorbed
(mmol kg'1)
OH/OX
(OH + P)/OX
(mmol kg'1) OH/OX
(OH +P)/OX
Clav fraction
5459
1.87*0.1
2.12*0. lab
3535
.SiO.r
1.92*0. lab
91*0.3
1.860. laa
2.050.1aa
721
1.83*0.1
1.96*0.06
9.50.5
1.90*0.1
1.91*0.1
91
1.89*0.0
1.85*0.10
Whole Soil
Material
24.50.7
1.64*0.1
3.220.2ab
16.70.5
1.75*0.1
4.210.2ab
21.7*0.6
1.73*0.1
3.11*0 lab
15.51.4
1.79*0.1
3.140.2bb
16.2*0.2
i.76o.i
3 040.2ab
10.20.7
1.84*0.1
2.83*0. lbb
7.10.1
i.69o.i
2.16*0. lbb
5.40.2
1.81*0.2
2.450.4cb
3.70.1
1.83*0. lab
2.06*0. lba
4.10.1
1.85*0.1
2.320.3ca
0.80.1
2.01*0.lba
1.930. lba
0.90.1
1.85*0.1
1.88*0.1
The first superscript letter indicates significant differences within a column and
material, while the second letter indicates significant differences between rows.

100
Desorption of Oxalate in the Presence of Phosphate
Desorption of Oxalate and Sorption of Phosphate
Small amounts of oxalate were desorbed by the whole-soil samples and the
clay fraction with the sorption of phosphate (Figs. 4.4 & 4.5). Higher amounts of
oxalate were released from the soil and clay containing organic matter. The
sorption of P by clay and soil surfaces is presented in Figures 4.4 and 4.5. The
presence of previously sorbed oxalate onto the clay surface had no significant
influence on P sorption in the present experiment.
Release of Hydroxyls
The amounts of hydroxyl released to solution during P sorption are
presented in Appendix D-2. The ratio of OH' released to amount of P sorbed
was calculated and shown to vary between 0.89 and 1.08 for the clay fractions
and between 0.85 to 1.15 for Values were not significantly different from 1 (Table
4.4), and ratios were not influenced by the amount of oxalate desorbed into
solution
The ratio of OH + Ox released into solution to P sorbed was also
calculated (Table 4 4) The two ratios (OH/P and (OH + Ox)/P) were not
significantly different for the clay samples. However, at high amounts of oxalate
previously present on whole-soil surfaces, OH + Ox/ P ratio were significantly
greater than 1.

101
-OM + OM -OM + OM
Fig. 4.4
Influence of previously sorbed oxalate on P sorption and oxalate
desorption for clay fraction
Oxalate desorbed (mmol kg )

102
2.5
2 T
1.5
1
0.5
0
Fig. 4.5.
Influence of previously sorbed oxalate on P sorption and oxalate
desorption by whole-soil samples.
Oxalate desorbed (mmol kg )

103
Table 4.4 Influence of P sorption by clay fractions and whole-soil samples on the
ratio of OH released, and OH + oxalate released, to P sorbed (Mean
SD; n=3).
Without Organic Matter
With Organic Matter
Ox PrevA
Present
(mmol kg'1)
Ratio
Ox Prev
Present
(mmol kg'1)
Ratio
OH/P
(OH + Ox)/P
OH/P
(OH + Ox)/P
Clav Fractions
45617
0.94*0. laa
1.04*0.1
27514
o.spo.i33
0.92*0.1
48*2.4
1.08*0.1
1.03*0.1
430.2
0.980.r
0.98*0.1
30.5
0.98*0. laa
0.99*0. laa
50.2
0.950. laa
0.95*0.1
Whole Soil
Material
14.9*0.7
1.08*0.1
2.990.2ab
11.61.0
Timor
2.900.3ab
12.40.4
1.01*0.1
1.250.2ba
8.80.6
1.09*0.2
2.050.2bb
9.6*0.5
1.150. Ia3
1 350.2ba
7.40.1
1.06*0.1
1.26*0.2ca
6.2*07
0.96*0.1
1.160.2ba
5.50.2
1.07*0.1
1.07*0. lca
4.0*0.1
0.850.1
0.99*0. lba
4.00.1
0.98*0.1
0.98*0.1
0.80.1
0.990. laa
1.04*0. lba
0.80.1
1.04*0.1
1.04*0.1
AOx Prev is Oxalate previously sorbed;
The first superscript letter indicates significant differences within a column and
material, and the second letter indicates significant differences between rows.

104
Release of Al. Fe. and Organic Carbon
Small amounts of Al, Fe and organic carbon were released into solution
during P sorption. The amounts of A1 and Fe desorbed by clay and soil with and
without organic matter are presented in Appendix D-2 The amounts of A1 and
Fe released to solution were higher for the clay fraction and for the whole-soil
samples having organic matter.
Discussion
Phosphate Desorption by Oxalate
It has been previously shown that i) oxalate and P are retained on
surfaces of clay particles and whole-soil samples through ligand exchange (Chapter
2; Chapter 3; Parfitt et al.. 1977; Sposito, 1984; Goldberg and Sposito, 1985); ii)
oxalate sorption causes dissolution of the mineral surfaces (Chapter 2; Stumm,
1986); and iii) at pH 4.5, oxalate formed a bidentate surface complex (Chapter 2)
while P was sorbed as monodentate and/or binuclear surface complexes (Chapter
3).
We postulated that oxalate released P from the mineral surfaces through
two processes: i) ligand exchange replacing P from the mineral surfaces; and ii)
dissolution of the mineral surfaces and release of P into solution.
The data indicated that, when high amounts of P were present on the
mineral surfaces, P was released into solution during oxalate sorption. At pH
4.5, oxalate would form bidentate complexes with the mineral surface and the ratio

105
of OH released to oxalate sorbed would be 2 (Chapter 2). If oxalate sorption
released previously-sorbed P into solution through ligand exchange, than one would
not expect the ratio of OH' released to oxalate sorbed to be close to 2. Data
from this study support this observation. The ratio of OH' released to oxalate
sorbed was significantly less than 2 when higher amounts of P were present on the
surfaces. However, the ratio of OH + P released into solution to oxalate sorbed
is equal to 2 for the clay fraction samples. This confirmed that oxalate released
some P directly through ligand exchange, with some sites for P and oxalate
sorption being common.
For whole-soil samples, the ratio of OH + P released to oxalate sorbed
was significantly higher than 2 when higher amounts of P were present on the
surfaces. The higher ratio suggests that more P was desorbed into the solution
with each unit of oxalate sorbed A second process which can bring P into
solution is dissolution of the soil surfaces. Release of A1 and desorption of P
each support the observation that oxalate sorption caused some surface dissolution.
Oxalate dissolution of mineral surfaces was also observed in earlier studies (Chapter
2; Stumm, 1986). Therefore, one can argue that some P also was released into
the soil solution through the dissolution of mineral surfaces.
Once P was released into solution along with Al, it could form various
solution species as A1H2P04(0H)2 and A1H2P04C204, or reprecipitate as amorphous
aluminum phosphate. For P to reprecipitate on the mineral surfaces, soluble
species of aluminum phosphate should first form in solution. This would serve as

106
nucleei for subsequent precipitation.. However, the log K of soluble aluminum
phosphate species in solution (A1H2P04(0H)2 log K =3.0) is low compared those
to the possible aluminum oxalate species in solution [A1C204" (6.5) or A1(C204)2'
(13.7)]. Therefore, oxalate should form stable soluble complexes with A1 as
A1C204 or A1(C204)'. The speciation calculations also suggested that less than
10% of the P in solution was present as aluminum oxalate phosphate complexes (
A1H2P04C204), with the majority of the P being in the form of H2P04 species.
Thus, one would expect P desorbed into solution by oxalate to be in the H2P04
form.
Oxalate Desorption by Phosphate
The small amount of oxalate that was desorbed by P indicates that oxalate
ions were held very strongly by the clay particle and whole-soil surfaces. Oxalate
forms bidentate surface complexes ( Chapter 2, Parfitt et al.. 1977). P forms
monodentate or binuclear complexes (Chapter 3; Parfitt, 1978; Tejador Tejador and
Anderson, 1990). Since oxalate forms a stronger complex, P could not replace the
sorbed oxalate. Our data suggest that the ratio of OH' released to P sorbed by
the whole-soil samples and the clay fractions in this study was not significantly
different from unity. This confirmed that P was sorbed onto the clay fractions and
whole-soil surfaces as monodentate and/or binuclear surface complexes, and that P
was sorbed by sites on the clay and soil surfaces not already occupied by oxalate.

107
As P sorption was not affected by the presence of previously sorbed oxalate, one
can conclude that many sites are highly specific for P.
Conclusions
This investigation elucidate the mechanism of P and oxalate desorption from
spodic horizon materials. Results from this study confirmed that oxalate released
P from the mineral surfaces through two processes: i) ligand exchange replacing P
from the mineral surfaces, and ii) dissolving the mineral surface, thereby bringing
sorbed P into solution. Results from the preceding chapter and this study
establish that the presence of soil organic matter and oxalate increases the amount
of P desorbed. Soil organic matter and oxalate significantly increase the P
concentration in solution.

CHAPTER 5
CONCLUSIONS
Following are the major conclusions from this research, its significance with
respect to the P nutrition of plants, and an outline of priorities for future studies.
Conclusions from this Study
Oxalate Sorption
1. Oxalate was sorbed by the clay fractions and whole-soil samples through
ligand exchange.
2. Sample pH determined the charge characteristics of clay surfaces and the
species of oxalate in solution, which in turn controled the sorption process.
Oxalate formed monodentate and/or binuclear complexes at pH 3.5, while
bidentate surface complexes were formed at pH 4.5 and 5.5.
3. Soil organic matter significantly reduced the amount of oxalate sorbed by
the mineral surfaces.
4. Data from this study provided experimental evidence for the theory of
congruent dissolution as proposed by Stumm (1986). Sorption is the first
step in the dissolution of mineral surfaces. Through sorption, oxalate forms
108

109
a coordinated complex with A1 at the mineral surface, with sufficient
strength to break the Al-0 bond and solubilize the metals into solution.
5. Considerable amounts of organic carbon were released into solution during
the sorption of oxalate. Aluminum ions can form cation bridges between
the soil organic carbon and soil particles. Oxalate can detach this bridge to
form soluble complexes with A1 in solution, resulting in the release of
large amounts of organic carbon.
6. The organic carbon released into solution during oxalate sorption had a
lower E4 to E6 ratio, suggesting that it was humic in nature.
P Sorption
7. P appeared to be sorbed by the mineral surfaces through a ligand-exchange
process. P formed monodendate and/or binuclear complexes on the mineral
surfaces.
8.Organic carbon reduced P sorption. Maximum reduction in P sorption
(about 50 %) was observed when both organic carbon and oxalate were
present
9. Some of the sites for sorption onto the surfaces of soil and clay particles
were common for oxalate, P and organic carbon.
10. P sorption released small amounts of Al, Fe and organic carbon into
solution.

110
P and Oxalate desorption
11. P desorption from the spodic horizon by oxalate appeared to be through
two processes: i) ligand exchange replacing P from the mineral surfaces;
and ii) oxalate forming surface complexes through ligand exchange
(replacing OH ions), dissolving the mineral surfaces, and releasing P into
solution.
12. Soil organic matter and oxalate significantly increased P desorption.
13. Once P was released into solution, oxalate did not allow P to
reprecipitate by forming stable and soluble complexes with A1 and Fe.
14 Oxalate was not released into the solution from spodic horizon material
during P sorption.
15. Some of the sites were highly specific for P sorption.
Influence of Oxalate on the P Nutrition of Trees
P availability often limits forest productivity in flatwood soils. These
results provide information about the availability of P in a spodic horizon.
Phosphorus is absorbed by plants largely as soluble H2P04' or HP042' species.
A soil's ability to supply P is determined by the concentration of P in the soil
solution, along with its ability to replenish any P lost from solution. Our
results confirm that oxalate significantly increases the solution concentration of P
by: i) reducing P sorption; ii) releasing P into solution from sorbed and
insoluble pools; and iii) complexing Al, Fe or Ca in solution, thereby reducing

Ill
P precipitation. This could increase the initial P concentration in solution and
affect the P buffer power of these soils resulting in a higher concentration of P
in solution.
The oxalate concentration in the rhizosphere of a spodic horizon range
from 0.0 to 2.8 mM (Fox and Comerford, 1990). This reported concentration
of oxalate in bulk soil solution tends to be much lower then in the vicinity of
roots and fungal hyphae. It seems reasonable to assume that actual
concentration near roots may be one or two orders of magnitude higher than
the concentration in bulk soil solution. The oxalate concentration of 5 mM
used in this study to desorb P exists under field conditions. At this oxalate
concentration, the whole-soil samples released 0.1 mM P into solution.
Assuming a gravimetric moisture content of 20 % and a 0.1 mM concentration
of P, the total amount of P in the soil solution at one time would be 0.6
mg kg'1. However, slash pine roots occupy less than 1 % of the soil volume
(Van Rees and Comerford, 1986), Therefore, high concentrations of oxalate
would exist in the zone immediately adjacent to the roots, which is only a few
millimeters wide and coincides with the rhizosphere region where microorganism
are particularly active. To include the influence of fungal hyphae along with
roots, one can further assume that these high concentration of P (0.6 mg kg'1)
would be present in approximately 2 % of soil volume. Assuming that the
entire pool of oxalate was being replenished on a weekly basis (Fox et al..
1990b), the spodic horizon would contribute approximately 0.6 kg P ha'V'1 to

112
the trees. Using Ca(N03)2 as a extractant instead, the estimated contribution is
approximately 0.20 kg of P ha'V1 As pine trees remove roughly 3 kg P
ha'V1 form the soil, this is about 20 % of the tree's demand. Given this
example, it shows that oxalate could significantly increase the availability of P in
spodic-horizon material. These results, in combination with earlier work (Fox et
af, 1990a; Fox et al.. 1990b; Neary et al.. 1990) highlight the potential
importance of subsoil fertility to the productivity of pine plantations in
flatwoods.
Future Research
Results from this study have defined the mechanism of P release into the
soil solution by oxalate, and its influence on initial solution P concentrations and
buffer power for this soil. The next step could be to incorporate this
information into a mechanistic nutrient uptake model in order to predict P
uptake by plants. This should support the conclusion that oxalate released into
the soil solution increases P availability in these soils.
To enhance our understanding of P cycling in spodic horizons,
fiirtherwork should address the following issues:
1. Influence of oxalate on P sorption and desorption in the presence of
additional cations and anions.
2. Identification and characterization of organic P compounds.

113
3. Since oxalate significantly influenced the sorption and desorption of
inorganic P, its influence on the sorption and desorption reactions of
organic P compounds also needed to be investigated.
4. Characterization of the organic carbon released into solution, using NMR
and FTIR techniques to understand its nature and influence on nutrient
availability.

Amount of 0.1 N NaOH used
APPENDIX A
TITRATION CURVES FOR THE CLAY FRACTION AND WHOLE-SOIL
MATERIAL
pH
Fig. A-l Titration curves for the clay fraction: A) With organic matter, B)
Without organic matter.
114

115
a* Y = 0.079 + 1.84 X R2 = 0.98
3.4^4.4 4.4 5.6 5.5 6.7
Change In pH
Fig. A-2
Buffer curves for the clay fraction, used to calculate the amount of
OH released into solution: A) With organic matter, B) Without
organic matter.

Amount of NaOH used (mL) Amount of NAOH used (mL)
116
pH
Fig. A-3.
Titration curves for the whole-soil: A) With organic matter, B)
Without organic matter

117
3.5 to 4.4 4.5 to 5.5 5.5 to 6.7
B -A ...0..
Fig A-4
Buffer curves for the whole-soil, used to calculate the amount of
OH' released into solution: A) With organic matter, B) Without
organic matter.

APPENDIX B
SORPTION OF OXALATE IN THE PRESENCE OF PHOSPHATE ONTO A
SPODIC HORIZON
Introduction
Oxalate significantly reduced P sorption capacity (Chapter 3). It has been
observed that oxalate blocks sites on soil materials and thus reduces P sorption.
Oxalate is continuously released into solution and competes with P for sorption
sites. To understand the nature of competition between these two anions, it is
important to study the influence of P on oxalate sorption capacity as well. The
aim of this work was to investigate the sorption of oxalate in the presence of P
as influenced by soil organic matter.
Materials and Methods
Soil Material
Soil was collected from the spodic horizon (Bh) in a single soil pit of a
Pomona series (sandy, siliceous, hyperthermic Ultic Alaquod) at the Gator
Nationals Forest site located in Alachua County, 10 km northeast of Gainesville,
Florida (Swindel et al.. 1988). The soil material was air dried, passed through 2-
118

119
mm sieve, and stored in plastic bags. Samples preparation for the sorption studies
was described in detail in chapter 2.
Sorption of Oxalate and Phosphate Added as a Mixture
Sorption experiments were carried out at different oxalate to phosphate
molar ratios. The concentration of oxalate used was 1.0 mM, with a phosphate
concentration of 0 mM, 0.1 mM, 1.0 mM, or 10 mM for the clay samples.
Triplicate samples consisting of 250 mg of clay (with and without organic matter)
were placed in 30 ml bottles along with 25 ml of solution. For the whole-soil
samples, both with and without organic matter, an oxalate concentration of 1.0
mM was used along with phosphate concentrations of 0 mM, 0.1 mM, 0.5 mM,
1.0 mM, 3.0 mM, 7.0 mM or 10 mM. Triplicate five-gram samples of whole-soil
from each treatment (with or without organic matter) were placed in 100 ml
bottles along with 50 ml of solution. pH of the clay suspension and whole-soil
was adjusted initially to 4.5, using 0.1 M HC1 or NaOH
Two drops of toluene were added to inhibit microbial growth and samples
were placed in a reciprocating shaker for 24 hrs. The pH of the suspension was
periodically (every three hours) adjusted with 0.1 M HC1 or NaOH over the 24
hrs of the experiment, to maintain the initial pH. At the end of the reaction
period, each suspension was centrifuged at 12,000 g for 20 minutes. The filtrate
was used for analysis of oxalate. Sorbed oxalate was calculated from the
difference between initial and final oxalate concentrations in solution.

120
Oxalate in the extract was determined by HPLC (Fox and Comerford,
1990a) using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton
Co., Reno, Nevada), with a Gilson single piston high pressure pump along with a
Pheodyne model 7125 injection valve fitted with a 20 fuL injection loop. This
1TPLC system uses a Gilson Holochrom variable wavelength UV detector in
conjunction with a Gilson computerized integrator. The eluent was 0.005 M
H2S04 at a flow rate of 2 mL min'1 Oxalate concentration was calculated from
the calibration curve obtained with standard solutions of 0.1 to 10 mM.
Statistical Analysis
Statistical differences in the sorption of oxalate in the presence of P were
compared. Significance of the difference between two means were determine by a
t-test (Snedecor and Cochran, 1980). Standard errors of means are given as well.
Results
Less oxalate was sorbed by the clay fractions and by the whole-soil samples
at high P concentrations. To evaluate the ability of P to depress oxalate sorption
when both P and oxalate were added together, the percent efficiency of P in
reducing the oxalate sorption was calculated according to the expression of Deb
and Datta (1967):
P -
1~
Ox
p
Ox
O
xioo
(B-l)

121
where Pe is efficiency of P to reduce oxalate sorption (%), OXp is oxalate sorbed
in the presence of P; and Ox0 is oxalate sorbed alone. Efficiency of phosphate in
reducing the sorption of oxalate in the P + oxalate system for the clay fractions is
presented in Table B-l. P was more effective in reducing oxalate sorption onto
the clay fraction when organic matter was not present.

122
Table B-l. Percent reduction in oxalate sorption for the whole-soil samples and
clay fractions (Mean SD. n=3).
P: Oxalate
Percent reduction
in oxalate sorption
molar ratio
+ OM
- OM
Whole-Soil
10:1
23.51.2aa
18.50.9ab
7:1
13.90.7ba
12.30.7ba
3:1
5.20.7ca
5.50.4ca
1:1
1.70.4d
2.40.4da
0.5:1
0.21.0da
1^0.5"
0.1:1
-0.90 8da
1.60.3'b
Clay Fractions
10:1
232aa
562ab
1:1
10lba
284bb
0.1:1
22ca
llca
In each superscript first letter indicates significance among means within the column
and second letter indicates significance among the rows, within the clay fraction
and the whole-soil samples.

APPENDIX C
MEASUREMENT OF P DESORPTION FROM A SPODIC HORIZON USING
DIFFERENT METHODS AT CONSTANT AND VARIABLE pH
Introduction
Inorganic soil P can be present in one of three different fractions: P in
the soil solution; P on the mineral surfaces in the labile pool, or P in the
nonlabile pool. The relationship between sorbed and solution P is often described
by adsorption/desorption isotherms. Adsorption and desorption commonly exhibit
hysteresis phenomena. Raven and Hossner (1993) showed a steeper isotherm slope
for sorption than for desorption. To predict P uptake by plants, however,
desorption isotherms are more useful since they evaluate the release of P into soil
solution. The availability of P is often controlled by the desorption isotherm as a
plant progressively extracts P from the soil around its roots (Yang et al.. 1991;
Abrams and Jarrell, 1992; Graetz and Nair 1994; Harris et al.. 1994; Raven and
Hossner, 1994). The desorption of soil P is a function of various soil parameters
mainly: a) the concentration of sorbed P; b) the concentration of P in the soil
solution; and c) the rate of P desorption into solution.
Spodic horizons can contain high amounts of amorphous Al oxides, which
can sorb P. Some of the sorbed P is in a form unavailable to plants (Ballard and
123

124
Fiskell, 1974). While relatively large amounts of total P can be present in spodic
horizons, water-soluble P tends to be low. According to Hingston et al. (1974),
chemisorption and bi- and multidentate complex formation can decrease P
desorption by increasing the irreversibility of sorbed P. Van Rees and Comerford
(1986) observed that roots are present in the spodic horizon, and Neary et al
(1990), using a mechanistic nutrient uptake model, suggested that subsurface
horizons could contribute a significant amount of the P required by southern pine.
Soil buffering capacity used in this model was calculated using a sorption isotherm.
To predict P uptake accurately, it may be important instead to determine and work
with the P desorption isotherm.
Desorption of P is generally determined by one of three different methods:
i) sequential extraction of soil with CaCl2, Ca(N03)2 or other salt solution at a
constant soil to solution ratio for a constant equilibration time; ii) equilibration of
soil with Ca(N03)2 or other salt solution at different soil-to-solution ratio for a
constant time period; and iii) anion exchange-resin extraction, in which the soil is
treated with different amounts of anion exchange resin in water or salt solutions
for a constant time period The first two methods study the chemical release of P
from soil relating to the movement of P in the soil profile including plant uptake.
The third method is more related to the overall extraction of P by plant roots.
The anion-exchange resin instead proved inadequate for precise characterization of
labile P, because it did not account for rate phenomena (Yang et al.. 1991;
Abrams and Jarrell, 1992). This is because three steps are involved during

125
removal of P by anion exchange resin; i) release from the soil; ii) transport
through the soil solution; and iii) sorption by the resin. In a recent study of P
desorption in the field by Cooperband and Logan (1994) with an anion exchange
membrane (AEM), it was shown that the AEM measured the net change and not
the total change in P for the time period during which the resin is in contact with
the soil. They stated that P was continuously adsorbed and desorbed from the
membrane surface as a function of the soil chemical-biological microenvironment.
Soil pH has a strong effect on P sorption and desorption (Barrow, 1983)
and P desorption leads in turn to a decrease in soil pH Small differences in pH
can have a significant influence on the P species present in solution. An
equilibrium between the particular P species in solution and solid phase P controls
the quantity of phosphate desorbed into solution. In most of the studies relating
to P desorption, pH was not measured nor maintained constant (Logan, 1982;
Menon et al., 1989; Graetz and Nair 1994; Harris et al.. 1994; Raven and
Hossner, 1994).
The objectives of this investigation were to: i) compare the P desorption
isotherm obtained by a sequential extraction technique with that by a dilution
method; and ii) compare P desorption underboth constant and variable pH
conditions by both methods.

126
Materials and Methods
Soil Material
Soil was collected from the spodic horizon (Bh) in a single soil pit of a
Pomona series (sandy, siliceous, hyperthermic Ultic Alaquod) at the Gator
Nationals Forest site located in Alachua County, 10 km northeast of Gainesville,
Florida (Swindel et al. 1988). The material was air dried, passed through a 2-mm
sieve, and stored in plastic bags.
Preparation of Phosphorus Enriched Soil Material
Soil was enriched with 1000 ppm solution of P as Ca(H2P04) and incubated
for 7 days at room temperature. At the end of this reaction period, samples
were centrifuged, the soil samples were oven dried to constant weight at 110 C,
and the supernatant was analyzed for inorganic P. The difference between the
amount of P added in the beginning and the amount of P in solution at the end
of the reaction period was taken as the amount of P sorbed by the soil.
Desorption by a Dilution Method
This experiment was conducted both under constant and variable pH
conditions. For one set of samples pH was kept constant, while for the other pH
was allowed to change. For each set, a varying amount of a 0.01M calcium
nitrate solution was added to three replicates of 5 gram soil samples. The soil to
solution ratios used were 1:10, 1:20, 1:30, 1:40, 1:50, 1:75, and 1:100. Two

127
drops of toluene were added to inhibit microbial growth, and samples were placed
in a reciprocating shaker for 24 hrs. For variable pH conditions, pH was
measured at the begining and at the end of the reaction period. For constant pH
conditions, pH of the suspension was periodically (every six hours) adjusted with
0 1 M HC1 or NaOH over the entire period of time, to the initial pH value. At
the end of the reaction period, each suspension was centrifuged at 12,000 g for 20
minutes. The supernatant solutions were analyzed for inorganic P and aluminum.
Desorption by a Sequential Extraction Method
Five grams of soil were repeatedly desorbed by 50 ml of 0.01 M calcium
nitrate added in successive 24 hr periods for 8 days. The other experimental
conditions were the same as described above, for the dilution method. The P
concentrations in successive volumes were measured, and the total P desorbed from
the soil was calculated by summation.
Chemical Analysis
Solution pH was measured using a combination glass electrode in conjuction
with an Orion pH meter Inorganic P in the filtrate was determined by a
molybdenum-blue colorimetric procedure, using ascorbic acid as a reductant
(Murphy and Riley, 1962)

128
Adsorption Isotherms
The Freundlich model was fitted to the desorption data. The Freundlich
model has been found to be useful in describing desorption of ions into solution
from from soil particles (Sposito, 1984). The following equation was used:
S=K[Cb]
where S is the amount of P retained per unit mass of soil (mM kg _1), C is the
equilibrium concentration of P (mM), and K and b are empirical constants for the
model K and b were calculated by a least squares fit to a linear form of the
Freundlich equation.
LogS=LogK+ LogC (C-2)
b
The derivative of this function with respect to C equals the Kd (solid to solution
partition coefficient) value which can be expressed by:
K l~'
Kd= x[C.]6 (C-3)
b
where K and b are the empirical constants calculated from the Freundlich equation.
Statistical Analysis
The linearized form of the Freundlich equation was used to test the
statistical differences between different regression lines using the General Linear

129
Models procedure in the SAS framework (SAS Institute, 1985). The model used
was:
Va + PV6ff (c-4)
where a is the intercept, P is the slope of the line and e is the error term. To
compare two regression lines, we have compared the Pj and P2 values of the two
lines. The ANOVA and t tests were used to test the difference between a, and
a2, and between Pj and P2.
Results
Desorption Isotherms
The desorption isotherms obtained by dilution and sequential extraction
methods under conditions of constant and variable pH are presented in Figure C-l.
Desorption isotherms followed the Freundlich model. With parameters of the
isotherms as given in Table C-l. The slopes of the regression lines for the
different methods illustrate hat the change in surface-sorbed P relative to solution P
were significantly different for the two methods. The variable vs constant pH
approaches did not significantly influence the amount of P released. The dilution
method, both at variable and constant pH, had a statistically lower release of P
into solution compared to the sequential extraction method at the same level of P
present on soil surfaces.

P remained sorbed (mmol kg )
130
Fig. C-l. P desorption curves by dilution and extraction methods at both
variable and constant pH.

131
Table C-l Freundlich isotherm parameters for the desorption of P from soil using
the dilution and sequential extraction methods at variable and
constant pH
Desorption
pH
Parameters of the Freundlich isotherm
Method
Conditions
Slope
Intercept
R2
Dilution
Constant
0.121a
2.20a
0.96
Variable
0.114a
2.12b
0.94
Sequential
Constant
0.088b
2.08b
0.95
Variable
0.083b
2.14b
0.94
Dilution
0.117a
2.16a
0.99
Sequential
0.086b
2.1 lb
0.98
@Includes all samples at variable and constant pH
Values within columns for different methods followed by the same letter are not
significantly different at the 5 % level.

132
Using the Freundlich isotherms, the partition coefficients (Kd) also were
calculated. The plot of Kd vs solution P is presented in Fig C-2. The results of
this study show that the two desorption methods differed significantly. The
amount of P desorbed by the dilution method under constant and variable pH was
less than for the sequential method. This resulted in higher partition coefficient
(Kd) values for P being calculated by the dilution method.

133
Equilibrium concentration of P in solution (mM)
Fig. C-2 Relationship between Kd and the equilibrium concentration of P as
obtained by different methods.
A

APPENDIX D
DESORPTION OF P AND OXALATE
Table D-l Influence of previously sorbed P by clay and whole-soil samples on
oxalate sorption and release of P, Al, Fe, OH, and organic carbon
(Mean SD, n = 3).
P Previously
Present
(mmol kg'1)
Oxalate
Sorbed
(mmol kg"1)
P
Amount Released
(mmol kg"1)
A1 Fe
OH
OCA
(mg g"1)
Clay without Organic Matter
5459
1757
70.11
1885
2.60.07
32611
910.3
1911
25.51
1864
2.70.07
36418
9.50.5
2056
4.31
1851
2.90.09
39014
Clay with Organic Matter
3535
2054
490.5
2251
2.50.15
34414
2676
720.7
2042
180.1
2252
2.90.07
3817
2777
9.20.8
2273
50.1
2303
3.10.03
4159
2676
Soil without Organic
Matter
24.50.7
9.40.3
9.70.1
4.80.1
0.120.01
15I
21 70.6
10.50.2
7.40.1
4.20.1
O.lliO.Ol
181
16.20.2
10.70.1
2.50.1
4.00.1
0.120.01
181
7.10.1
12.0.6
1.50.1
3.90.1
O.lliO.Ol
202
3.70.1
12.40.1
0.90.1
4.10.1
0.090.00
221
0.80.1
12.50.4
0.50.1
4.40.1
0.110.01
252
Soil With
Organic Matter
16.70.5
2.40.1
9.20.2
271
0.130.01
4.20.2
3.40.1
15.51.4
2.70.1
7.80.1
351
0.140.01
4.80.2
3.80.1
10.20.7
2.80.3
4.60.1
351
0.150.01
5.20.7
3.90.1
5.40.2
5.90.4
2.30.1
361
0.170.01
10.50.2
4.90.1
4.10.1
6.80.5
1.10.1
371
0.170.01
12.70.9
4.40.1
0.90.1
6.90.3
0.60.1
371
0.190.01
12.70.2
4.50.1
A OC = Organic carbon
134

135
Table D-2 Influence of oxalate previously present on P sorption and release of
oxalate, Al, Fe, OH, and organic carbon by clay fractions and whole-
soil samples (Mean SD, n = 3).
Oxalate
P
Amount Released
Previously
Sorbed
(mmol kg"1)
O
o
>
Present
(mmol kg"1)
(mmol kg'1)
Ox
A1
Fe
OH
(mg g'1)
Clay without Organic Matter
45617
2711
29.30.5
361
0.050.02
2566
482.4
2831
9.20.4
330
0.010.01
28613
2.70.5
3051
1.10.2
211
0.080.0
30020
Clay with Organic Matter
27514
1192
0.490.02
210.8
0.810.03
1075
211
430.2
1793
0.240.1
80.5
0.210.03
17518
461
4.80.2
984
0.090.06
80.5
0.150.02
1887
461
14.90.7
8.30.1
Soil Without Organic
1.10.06 3.40.1
matter
0.030.01
9.01
12.40.4
9.20.2
0.20.01
2.10.1
0.030.01
9.31
9.60.5
10.80.3
0.20.01
1.70.1
0.030.01
12.52
6.20.7
12.90.1
0.20.01
1.50.1
0.040.01
12.52
4.00.1
13.80.1
0.10.01
1,60.1
0.020.01
11.71
0.80.1
14.10.1
0.10.01
1.10.1
0.010.01
14.12
11.61.0
5.70.1
Soil With
1 90.1
Organic Matter
4.60.1 0.060.01
5.91
l.liO.l
8.80.6
7.80.2
0.90.1
4.10.1
0.040.01
8 62
1.00.1
7.40.1
9.10.4
0.20.1
1,60.1
0.020.01
9.61
0.80.0
5.50.2
10.30.8
traces
1 40.1
0.010.01
11.11
0.70.1
4.00.1
10.90.3
traces
1.20.1
0.010.01
10.61
0.70.1
0.80.1
10.90.1
traces
0.70.1
O.OliO.Ol
11.41
0.60.1
A
OC = Organic Carbon

REFERENCES
Abrams, M. M. and W M. Jarrell. 1992. Bioavailability index for phosphorus
using ion exchange resin-impregnated membranes. Soil Sci. Soc. Am. J.
56:1532-1537.
Appelt, H N. T. Coleman, and P F Pratt 1975. Interactions between organic
compounds, minerals and ions in volcanic-ash-derived soils: II. Effect of
organic compounds on the adsorption of phosphate. Soil Sci. Soc. Am.
Proc 39:628-630.
Bakheit Said, M. and A. Dakermanji 1993. Phosphate adsorption and
desorption by calcareous soils of Syria. Commun Soil Sci. Plant Anal.
24:197-210
Baldwin, J.P., P B Tinker, and P H. Nye 1972 Uptake of solutes by
multiple root systems from soil: II. The theoretical effects of rooting
density and pattern on uptake of nutrients from soil. Plant and Soil
36:693-708.
Ballard, R, and J. G. A. Fiskell. 1974. Phosphorus retention in Coastal Plain
forest soils: I Relationship to soil properties. Soil Sci. Soc. Am. Proc.
38:250-255.
Barber, S.A. 1984 Soil Nutrient Bioavailability: A Mechanistic Approach
Wiley-Interscience, New York
Barrow, N. J. 1987. Reactions with Variable Charge Soils. Martinus Nijhoff Pub.,
Dordrech, The Netherlands.
Barrow, N. J. 1983. On the reversibility of phosphate sorption by soil. J. Soil
Sci. 34:751-758.
Bilinski, H., L Horvath, N. Ingri and S. Sjoberg. 1986. Equilibrium aluminium
hydroxy-oxalate phase during initial clay formation; H+, Al3", oxalic acid-Na*
system. Geochim. Cosmochim. Acta. 50:1911-1922.
136

137
Bruckert, S. 1970. Influence des composes organiques solubles sur la
pedogenese en milieu acide. Ann. Agron. 21:421-252.
Buurman, P. 1984. Podzols. Van Nostrand Reinhold Co., New York.
Carlisle, V. W., F. Sodek III, M E. Collins, L C. Hammond, and W G
Harris. 1988. Characterization data for selected Florida soils. Univ. of
Florida, Soil Sci. Dep Res Rep 88-1, Gainesville, FL.
Comerford, N. B and M. F. Skinner. 1989. Residual phosphorus solubility for
an acid, clayey, forested soil in the presence of oxalate and citrate. Can. J.
Soil Sci. 69:111-117.
Comerford N. B G. Kidder and A. V. Mollitor. 1984. Importance of subsoil
fertility to forest and non-forest plant nutrition, p 381-404. In E. L.
Stone (ed.) Forest soils and Treatment Impacts. University of Tennessee.
Knoxville TN
Cooperband, L. R. and T. J. Logan. 1994. Measuring in situ changes in labile
soil phosphorus with anion-exchange membrabes. Soil Sci. Soc. Am. J.
58:105-115.
Cromack, K P Sollins, W. C. Graustein, C. Y Li, and R. L. Todd 1979.
Calcium oxalate accumulation and soil weathering in mats of the hypogeous
fungus. Soil Biol. Biochem. 11:463-469.
Davis, J. A. and R Glour. 1981. Adsorption of dissolved organic in lake water
by aluminum oxide. Environ. Sci. Technol. 15:1223-1229.
De Coninck, F. 1980. Major mechanism in formation of spodic horizons.
Geoderma 24:101-128.
Deb, D L. and N. P Datta 1967. Effect of associating anions on phosphate
retention in soil: I Under variable phosphorus concentrations. Plant Soil
26:303-316.
Duquette, M. and W. Hendershot. 1993. Soil surface charge evaluation by back-
titration: I. Theory and method development. Soil Sci. Soc. Am. J.
57:1222-1228.
Earl, K. D J. K. Syers, and J. R McLaughlin 1979. Origin of the effects of
citrate, tartrate and acetate on phosphate sorption by soils and synthetic
gels. Soil Sci. Soc. Am. J. 43:674-678.

138
Farmer, V C J O Skjemsted, and C H Thompson, 1983 Genesis of humus
B horizons in hydromorphic humus podzols. Nature 304:342-344.
Flaig, W. 1971. Organic compounds in soil. Soil Sci. 111:19-33.
Fox, T R and N. B Comerford 1990. Low-molecular-weight organic acids in
selected forest soils of southeastern USA. Soil Sci. Soc. Am. J. 54:1139-
1144.
Fox, T R N B Comerford, and W W McFee 1990a Kinetics of
phosphorus release from Spodosols: Effect of oxalate and formate. Soil
Sci. Soc. Am. J. 54:1441-1447.
Fox, T R N B Comerford, and W W McFee 1990b Phosphorus and
aluminum release from a spodic horizon mediated by organic acids. Soil
Sci. Soc. Am. J. 54:1763-1767.
Fusseder, A., M. Kraus and E. Beck. 1988 Reassessment of root competition
for P in field- grown maize in pure and mixed cropping. Plant Soil
106:299-301.
Gardner, W K D G Parbery, and D A Barber 1982 The acquisition of
phosphorus by Lupinus albus I Some characteristics of the soil/root
interface Plant Soil 68:19-32.
Genrich, D A. and J M Bremner 1974 Isolation of soil particle size
fractions. Soil Sci. Soc. Am. Proc. 38:222-225.
Goh, T B and P M Huang 1986 Comparison of the influence of citric and
tannic acids on hydroxy-Al interlayering of monmorillonite. Clays Clay
Miner. 33:210-215
Goldberg, S. and G. Sposito. 1985. On the mechanism of specific phosphate
adsorption by hydroxylated mineral surfaces: A review. Commun. Soil Sci.
Plant Anal. 16:801-821.
Graetz, D A. and V D Nair 1994. Fate of phosphorus in Florida Spodosols
contaminated with cattle manure. Ecological Engineering (In press)
Greenland, D J 1971. Interaction between humic and fiilvic acids and clay. Soil
Sci 111:34-41

139
Harris, W G, V W Carlisle, and K C. J Van Rees 1987a Pedon zonation
of hydroxy-interlayered minerals in Ultic Haplaquods. Soil Sci. Soc. Am. J.
51:1367-1372.
Harris, W. G, V. W. Carlisle, and S. L. Chesser. 1987b. Clay mineralogy as
related to morphology of Florida soils with sandy edipedons. Soil Sci. Soc.
Am J. 51:1673-1677.
Harris, W G., R D Rhue, G. Kidder, R B Brown and R Littell 1994
Rapid assessment of relative phosphorus retention capacity for sandy coastal
plain soil materials. Ecological Engineering (In press)
Haynes, R J and R S. Swift 1989 The effect of pH and drying on
adsorption of phosphate retention by active aluminum and iron of Ando
soils. J. Soil Sci. 40:773-781.
Hingston, F J R J Atkinson, A M Posner, and J P Quick. 1967. Specific
adsorption of anions. Nature 215:1459-1461.
Hingston, F. J., A M Posner, and J P Quick 1972 Anion adsorption by
goethite and gibbsite. I. The role of the proton in determining adsorption
envelopes J Soil Sci. 23:177-192.
Hingston, F J A. M Posner, and J P Quick 1974. Anion adsorption by
geothite and gibsite. II. Desorption of anions from hydrous oxide surfaces.
J Soil Sci. 25:16-26.
Huang, P. M. and M Schnitzer (ed.) 1986. Interactions of Soil Minerals with
Natural Organics and Microbes. Soil Sci. Soc. Am Spec. Publ. 17,
Madison, WI
Huang, P M and A Violante 1986 Influence of organic acids on
crystallization and surface properties of precipitation products of aluminum
p 139-221 In P M Huang and M Schnitzer (ed ) Interactions of Soil
Minerals with Natural Organics and Microbes Soil Sci. Soc. Am Spec
Publ 17, Madison, WI
Hue, N V., G R Craddock, and F Adams 1986 Effect of organic acids on
aluminum toxicity in subsoils. Soil Sci. Soc. Am. J. 50:28-34.
Inoue, T. and K. Wada 1971. Reactions between humified clover extract and
imogolite as a model of humus-clay interactions: Part I and II. Clay Sci.
4:61-70.

140
Jackson, M. L 1979. Soil Chemical Analysis Advanced Course Madison,
WI.
Jardine, P. M N.L. Weber, and J F. McCarthy. 1989. Mechanism of dissolved
organic carbon adsorption on soil. Soil Sci. Soc. Am. J. 53: 1378-1385.
Jersak, J. M and J. G. McColl 1989. Aluminum release from solid-phase
components of forest soil leached with citric acid Soil Sci. Soc. Am. J.
53:550-555.
Kafkafi, U., B Bar-Yosef, R Rosenberg, and G Sposito. 1988 Phosphorus
adsorption by kaolinite and montmorillonite:II. Organic anion competition.
Soil Sci. Soc. Am. J. 52:1585-1589.
Kunze, G. W. and J B. Dixon. 1986. Pretreatment for mineralogical analysis, p
91-100 In Page, A L., R H Miller, and D R Keeney (ed.) 1982.
Methods of Soil Analysis. Part 2. 2nd ed. Agronomy 9. Am. Soc.
Agron & Soil Sci Soc Am Madison, WI
Kuo, E J and W L Pan 1988 Influence of phosphate sorption parameters of
soils on the desorption of phosphate by various extractants. Soil Sci. Soc.
Am. J 52:974-979.
Lan, M N. B Comerford and T R Fox. 1994 Effect of organic anions of
varying complexation power on the release of phosphorus from 10 spodic
horizons. (Submitted to Soil Sci. Soc. Am. J ).
Lee, F. Y., T. L. Yuan, and V. W Carlisle. 1988a Nature of cementing
material in ortstein horizons of selected Florida Spodosols: I. Constituents
of cementing materials. Soil Sci. Soc. Am. J. 52:1411-1418.
Lee, F. Y., T. L. Yuan, and V. W Carlisle. 1988b, Nature of cementing
matreial in ortstein horizons of selected Florida spodosols: II. Soil
properties and chenical forms of aluminum. Soil Sci. Soc. Am. J. 52:1766-
1801.
Logan, T. J. 1982. Mechanism of release of sediment-bound phosphate to water
and the effects of agriculture land management on fluvial transport of
particulate and dissolved phosphate. Hydrobiologia 92:519-530.
Lopez-Herandez, D 1974. Phosphate desorption isotherms in four selected
tropical soils and one temperate soil. Commun. Soil Sci. Plant Anal
5:145-154.

141
Lopez-Hernandez, D, G. Siegert, and J.V. Rodriguez. 1986. Competitive
adsorption of phosphate with malate and oxalate by tropical soils. Soil Sci.
Soc. Am. J. 50:1460-1462.
Martell, A. E., R.J Motekaitis and R M. Smith. 1988. Structural stability
relationships of metal complexes and metal speciation in envirionmental
aqueous solutions. Environ. Toxicol. Chem.7:417-434.
Martell, A. E. and R M. Smith. 1982. Critical Stability Constants, Vol. 5: First
supplement. Plenum Press. New York.
Marzadori, C, L. V. Antisari, C. Ciavatta, and P. Sequi 1991. Soil organic
matter influence on adsorption and desorption of boron. Soil Sci. Soc Am.
J. 55:1582-1585.
Menon, R G, I I Hammond, and H Sissingh 1989. Determination of plant-
available phosphorus by iron hydroxide impregnated paper. Soil Sci. Soc.
Am. J. 53:110-115.
Miller, W P., L W Zelazny, and D C. Martens 1986. Dissolution of
synthetic crystalline and noncrystalline iron oxide by organic acids.
Geoderma 37:1-13.
/vloore, T R W Souze, and J F Koprivnjak 1992 Controls on the sorption
of dissloved organic carbon by soils. Soil Sci. 154:120-129.
Murphy, J. and J P Riley. 1962. A modified single solution method for the
determination of phosphate in natural waters. Anal. Chim. Acta 27:31-36.
Nagarajah, S., A M Posner, and J. P Quirk 1968. Desorption of phosphate
from kaolinite by citrate and bicarbonate. Soil Sci. Soc. Am. Proc. 32:507-
510.
Nagarajah, S., A M Posner, and J P Quick. 1970 Compatitive adsorption of
phosphate with polygalacturonate and other organic anions on kaolinite and
oxide surfaces. Nature 228:83-85.
Nanzyo, M. 1987. Formation of noncrystalline aluminum phosphate through
phosphate sorption on allophanic Ando soils. Commun Soil Sci. Plant
Anal. 18:735-742.

142
Neary, D G, E. J Jokela, N B Comerford, S R Colbert, and T E. Cooksey.
1990. Understanding competition for soil nutrients the key to site
productivity on southeastern coastal plain Spodosols. p 432-450. In: S. P.
Gessel, D S. Lacate, G F. Weetman, and R F Powers (eds.) Sustained
Productivity of Forest Soils. Proc Am. For Soil Conf 7th, Uni of
British Columbia Faculty of Forestry Publ., Vancouver, BC, Canada
Ng Kee, K. F. and P. M. Huang. 1977. Sorption of phosphate by hydrolytic
reaction product of aluminum Nature 271:336-337.
Nye, P. H. and P. B. Tinker. 1977. Solute Movement in the Soil-Root System.
Univ. of California Press, Berkeley.
Ohman, L. and S. Sjoberg. 1988. Thermodynamic calculations with special
reference to the aqueous aluminum system PP 1-33 In J R Kramer and
H. E. Allen, (eds) Metal Speciation Theory, Analysis and application. Lewis
Pub Inc., Ann-Arbor, MI USA
Page, A L, R H Miller, and D R Keeney (ed.) 1982 Methods of Soil
Analysis, part 2. 2nd ed Agronomy 9. Am. Soc. Agron & Soil Sci.
Soc. Am., Madison, WI
Parftt, R. L. 1978. Anion adsorption by soils and soil materials. Adv. Agron.
30:1-50.
Parftt, R. L., V.C. Farmer and J.D Rusell 1977. Adsorption on hydrous
oxides I Oxalate and Benjoate on goethite J Soil Sci. 28:29-39.
Pholman, A. A., J. G. McColl, J. M. Jersak, S. C. Tan, and R. R. Northup.
1990. Organics and metal solubility in California forest soils, p 178-195.
In: S P Gessel, D S. Lacate, G F Weetman, and R F Powers (eds)
Sustained Productivity of Forest Soils. Proc. Am. For. Soil Conf. 7th,
Uni of British Columbia Faculty of Forestry Publ., Vancouver, BC,
Canada
Pohlman, A. A. and J. G. McColl. 1988. Soluble organics from forest litter and
their role in metal dissolution Soil Sci. Soc. Am J. 52:265-271.
Pohlman, A. A and J G McColl 1986 Kinetics of metal dissolution from
forest soils by soluble organic acids. J. Environ. Qual 14:86-92.

143
Polglase, P. J., N. B Comerford, and E. J. Jokela. 1992. Mineralization of
nitrogen and phosphorus from soil organic matter in Southern pine
plantation. Soil Sci. Soc. Am. J. 56:921-927.
Pritchett, W L and N B Comerford 1983. Nutrition and fertilization of slash
pine p 69-90 In E.L Stone (ed.) The Managed Slash Pine Ecosystem
Symp Proc., Gainesville, FL. 9-11 June 1981 Univ Florida, Gainesville
Raven, K P and L. R Hossner 1993. Phosphorus desorption quantity-intensity
relationships in soils. Soil Sci. Soc. Am. J 57:1501-1508
Raven, K P and L R Hossner 1994 Soil phosphrus desorption kinetics and
its relationship with plant growth. Soil Sci. Soc. Am. J. 58:416-423.
Reddy, M. N., G. Ramagopal, and A. S. Rao 1977. Phenolic acids in
groundnut seed exudates. Plant Soil 46:655-658.
/Ryden, J C. and J K Syers. 1987 Origion of the labile phosphate pool in
soils. Soil Sci. 123:353-361.
SAS Institute. 1985. SAS Users Guide. Statistics. Version 5 ed. SAS Inst.,
Cary, NC.
/Sibanda, H. M., and S. D. Young. 1986. Competitive adsorption of humus
acids and phosphate on geothite, gibsite and two tropical soils. J. Soil Sci.
38:211-217.
Smith, W H. 1969. Release of organic materials from the roots of tree seedlings.
Forest Sci 15:138-143
Snedecor, G. W. and W. G. Cochran. 1980. Statistical Methods. The Iowa St
University Press. Ames, IA.
Sposito, G 1984 The Surface Chemistry of Soils. Oxford Univ Press, New
York
Stainton, M. P. 1980. Error in molybdenum blue methods for determining
orthophosphate in freshwater. Can. J. Fish. Aquat. Sci. 37:472-478.
Stevenson, F J 1982. Humus Chemistry John Wiley and Sons, New York

144
Stevenson, F. J. and A. Fitch. 1986. Chemistry of complexation of metal ions
with soil solution organics, p 29-58. In P. M. Huang and M. Schnitzer
(eds.) Interactions of Soil Minerals with Natural Organics and Microbes.
Soil Sci. Soc. Am. Spec. Publ., 17, Madison, WI.
Stewart, J. W. B. and H. Tiessen. 1987. Dynamics of organic phosphorus.
Biogeochemistry. 4:41-60.
Stumm, W. 1986. Coordinative interactions between soil solids and water -An
aquatic chemist's point of view. Geoderma, 38:19-30.
Stumm, W and J. J Morgan. 1981. Aquatic Chemistry. 2nd ed. John Wiley
& Sons, New York.
Swindel, B F D G Neary, N. B Comerford, D L Rockwood, and G M
Blakeslee. 1988 Fertilization and competition control accelerate early
southern pine growth on flatwoods. South J. Appl. For. 12:116-121
Tan, K. H 1986. Degradation of soil minerals by organic acids, p. 1-28. In P.
M. Huang and M. Schnitzer (ed.) Interactions of Soil Minerals with Natural
Organics and Microbes Soil Sci Soc Am Spec Publ 17, Madison, WI
Tate, K R and B K G Theng. 1981. Organic matter and its interactions
with inorganic soil constituents. PP 225-249. In B K. G. Theng (ed)
Soils with Variable Charge New Zealand Soc. Soil Sci., Lower Hutt
New Zealand
Tejedor-Tejedor, M I and M. A. Anderson. 1990. Protonation of phosphate on
the surface of geothite as studied by CIR-FTIR and elecrophoretic mobility.
Langmuir 6:602-611.
Thomas, F., A Masion, J Y Bottero, J Rouiller, F. Genevrier and D
Boudot. 1991. Aluminum (III) speciation with acetate and oxalate
Environ. Sci. Technol. 25:1553-1565.
Thurman, E. M. 1985. Organic Geochemistry of Natural Waters. Martinus
Nijhoff/Dr. W. Junk Pub, Boston.
Tipping, E. 1981a Adsorption by geothite (a-FeOOH) of humic substances
from three different lakes. Chem. Geol. 33:81-89.
Tipping, E. 1981b Adsorption of aquatic humic substances by iron oxides
Geochim. Cosmochim. Acta 45:191-199.

145
Trana, S J, G Sposito, D Hesterberg, and U Khafkafi. 1986a Effect of
pH and organic acids on orthophosphate solubility in an acidic,
monmorillonitic soil. Soil Sci. Soc. Am. J. 50:45-52.
Trana, S I, G Sposito, D Hesterberg, and U. Khafkafi 1986b Effect of
ionic strength, calcium, and citrate on orthophosphate solubility in an acidic,
monmorillonitic soil. Soil Sci. Soc. Am. J. 50:623-627.
Ugolini, F.C., R Dahlgren, S Shoji and T. Ito. 1988 An expample of
andosolozation and podzolization as revealed by soil solution studies,
southern Hakkoda, Northeastern Japan. Soil Sci. 145:11-125.
Van Rees, K.C.J. and N. B. Comerford. 1986. Vertical root distribution and
strontium uptake of a slash pine stand on a Florida Spodosol. Soil Sci.
Soc. Am J. 50:1042-1046.
Van Rees, K.C.J., N B Comerford, and W McFee 1990 Modeling
potassium uptake by slash pine seedling from low K supplying soils of the
Southeastern Coastal Plain. Soil Sci. Soc. Am. J 49:1413-1421.
Vedy, J. C. and S. Bruckert. 1982. Soil solution: Composition and pedogenic
significance, p 184-213 In Bonneau, M. and Souchier (eds): Constituents
and Properties of Soils. Acdemic Press, London, .
Veith, J. A. and G. Sposito. 1977. Reactions of aluminum-solicates, aluminum
hydrous oxides, and aluminum oxides with o-phosphate: The formation of x-
ray amorphous analogs of variscite and montebrasite. Soil Sci. Soc. Am J.
41:870-876.
Violante, A. and L. Gianfreda. 1993. Competition in adsorption between
phosphate and oxalate on an aluminum hydroxide montmorillonite complex
Soil Sci. Soc. Am. J. 57:1235-1241
Violante, A., C. Colmbo and A. Buondonno. 1991. Competitive adsorption of
phosphate and oxalate by aluminum oxides. Soil Sci Soc. Am. J 55:65-
70.
Viotante, A. and P M. Huang. 1984 Nature and properties of pseudoboehmites
formed in the presence of organic and inorganic ligands Soil Sci. Soc.
Am. J. 48 1193-1203

146
Viotante, A and P M. Huang 1985 Influence of inorganic and organic ligands
on the formation of aluminum hydroxides and oxyhydroxides. Clays Clay
Miner. 33:181-192.
Viotante, A and P M Huang. 1989. Influence of oxidation treatement on
surface properties and reactivities of short-range ordered precipition products
of aluminum. Soil Sci. Soc. Am. J. 53:1402-1407.
Yang, J E., E. O Skogley, and B E. Schaff 1991. Nutrient flux to mixed-bed
exchange resin: Temparature effects. Soil Sci. Soc. Am. J. 55:762-767.
yYuan, T. L. 1980 Adsorption of phosphate and water-extractable soil organic
matter material by synthetic aluminum silicates and acid soils. Soil Sci
Soc. Am J. 44:951-955.
Yuan, T. L. 1992. Three potential amendements for better fertilizer utilization in
sandy soils Soil Crop Sci. Soc. Florida Proc 51:49-53.
Yuan, G. and L. M. Lavkulich. 1994. Phosphate sorption in raltion to
extractable iron and aluminum in Spodosols. Soil Sci. Soc. Am. J. 58:343-
346.
Zelazny, I. W. and J. N. Quresih. 1973. Chemical pretreatment effect on clay
separation and mineralogical analysis of selected Florida soils. Soil and
Crop Sci. Soc. Florida Proc. 32:117-121.

BIOGRAPHICAL SKETCH
Jagtar Singh Bhatti was born on March 9, 1958 in Jalandhar, Punjab, India.
He attended high school in Ludhiana and graduated from Malwa Khalsa High
School in 1974. He went to Punjab Agricultural University Ludhiana, Punjab and
received his Bachelor of Science (1979) and Master of Science in Soil Science
(1981). He joined the faculty of Soil Science at Punjab Agricultural University
Ludhiana, Punjab and taught undergraduate courses in soil science. In December
1984, he met Gurmeet and got married. In June 1985, they decided to migrate to
Canada. After coming to the cold country, Jagar entered another Master's program
in soil chemistry at University of Saskatchewan, Saskatoon and graduated in 1988.
Jagtar worked as Research Assistant at University of Washington, Seattle for one
and a half years. After traveling into Far Eastern countries and Australia for
about four months, in May 1991, he joined University of Florida for Doctor of
Philosophy in Forest Soils in the Department of Soil and Water Science. He is
currently employed as an Ecological Modeler at Sault Ste. Marie, Ontario, Canada
for Forestry Canada Ontario Region
147

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presen: ition and is fully adequate, in scope and
quality, as a dissertation for the degree of Doc: of Philosophy
CgY
Nicholas B Comerford, Chair
Professor of Soil and Water Science
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy
Brian L. McNeal
Professor of Soil and Water Science
I certify that I have read
acceptable standards of scholarly
quality, as a dissertation for the
I certify that I have read
acceptable standards of scholarly
quality, as a dissertation for the
I certify that I have read
acceptable standards of scholarly
quality, as a dissertation for the
this study and that in my opinion it conforms to
presentation and is fully adequate, in scope and
degree of Doctor of Philosophy
Water Science
/A -JaL
^arl L. Stone
Professor of Soil and
this study and that in my opinion it conforms to
presentation and is fully adequate, in scope and
degree of Doctor of Philosophy
tor or Knuosoony
ifWd^T Johnston
ssociate Professor of/Soil and Water
Cliffi
Associate
Science
this study and that in my opinion it conforms to
presentation and is fully adequate, in scope and
degree of Doctor
P S
Graduate Reserach Professor of Soil
and Water Science
of Philosophy.
'Ih>
C Rao

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy. .
0 -
Eric J. Jofca *
Associate Professor of Forest
Resources and Conservation
This dissertation was submitted to the Graduate Faculty of the College of
Agriculture and to the Graduate School and was accepted as partial fulfillment of
the requirements for the degree of Doctor of Philosophy.
May, 1995
%ac y-4
Dean, College of Agriculture
A^icull
Dean, Graduate School



75
Table 3.5. Parameters of the linear regression models for the release of OH' ions
and organic carbon from the whole-soil samples and clay fractions
Treatments
OH' ions released
R2
Organic Carbon released
Slope
Constant
Slope
Constant
R2
Soil without
organic
matter
P
0.66a
2.9a
0.89
P + OX
0.59a
16 lb
0.80
Soil with organic matter
P
1.06b
2.7a
0.95
0.05a
0.59a
0.98
P + OX
1.19b
5.8a
0.94
0.06a
0.71a
0.94
Clay without
organic
matter
P
0.74a
66b
0.99
P + OX
0.9 lb
116c
0.99
Clay with organic matter
P
0.90b
6a
0.99
0.23a
67.3a
0.96
P + OX
1.08
72b
0.96
0.22a
128.8b
0.98
Values within columns followed by the same letter for soil or clay samples are not
significantly different at the 5% level


INFLUENCE OF SOIL ORGANIC MATTER ON PHOSPHORUS AND
OXALATE SORPTION AND DESORPTION IN
A SPODOIC HORIZON
^ By
JAGTAR S BHATTI
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1995


63
binding energy of P or oxalate sorption The parameters (K and Sm) were
calculated by a least square fit of the linear form of the equation:
r 1 c
-=[]+
S KS S
m m
3.2
Speciation Calculations
The metal speciation model MINTEQA2 version 3.2 was used to calculate
the species of A1 and P in the solution It was assumed that the system had
achieved equilibrium conditions. Although the total activity of A1 ions was not
known, the assumed activity of A1 from the congruent dissolution by oxalate was
used (Chapter 2) to estimate the saturation level with respect to an amorphous
aluminum phosphate solid phase. Calculations were carried out at solution P
concentrations corresponding to the highest loading rate of P (10'2 M). The
formation constants used for calculations of species were those of the MINTEQA2
data base taken from Martell and Smith (1982).
Statistical Analysis
The linear form of the Langmuir equation was used to test the differences
between parameters for P sorption. Statistical differences among regression lines
for the desorption of Al, Fe OH, and organic carbon during P sorption were
tested using the General Linear Models procedure of the SAS framework (SAS
Institute, 1985). The model used was:


144
Stevenson, F. J. and A. Fitch. 1986. Chemistry of complexation of metal ions
with soil solution organics, p 29-58. In P. M. Huang and M. Schnitzer
(eds.) Interactions of Soil Minerals with Natural Organics and Microbes.
Soil Sci. Soc. Am. Spec. Publ., 17, Madison, WI.
Stewart, J. W. B. and H. Tiessen. 1987. Dynamics of organic phosphorus.
Biogeochemistry. 4:41-60.
Stumm, W. 1986. Coordinative interactions between soil solids and water -An
aquatic chemist's point of view. Geoderma, 38:19-30.
Stumm, W and J. J Morgan. 1981. Aquatic Chemistry. 2nd ed. John Wiley
& Sons, New York.
Swindel, B F D G Neary, N. B Comerford, D L Rockwood, and G M
Blakeslee. 1988 Fertilization and competition control accelerate early
southern pine growth on flatwoods. South J. Appl. For. 12:116-121
Tan, K. H 1986. Degradation of soil minerals by organic acids, p. 1-28. In P.
M. Huang and M. Schnitzer (ed.) Interactions of Soil Minerals with Natural
Organics and Microbes Soil Sci Soc Am Spec Publ 17, Madison, WI
Tate, K R and B K G Theng. 1981. Organic matter and its interactions
with inorganic soil constituents. PP 225-249. In B K. G. Theng (ed)
Soils with Variable Charge New Zealand Soc. Soil Sci., Lower Hutt
New Zealand
Tejedor-Tejedor, M I and M. A. Anderson. 1990. Protonation of phosphate on
the surface of geothite as studied by CIR-FTIR and elecrophoretic mobility.
Langmuir 6:602-611.
Thomas, F., A Masion, J Y Bottero, J Rouiller, F. Genevrier and D
Boudot. 1991. Aluminum (III) speciation with acetate and oxalate
Environ. Sci. Technol. 25:1553-1565.
Thurman, E. M. 1985. Organic Geochemistry of Natural Waters. Martinus
Nijhoff/Dr. W. Junk Pub, Boston.
Tipping, E. 1981a Adsorption by geothite (a-FeOOH) of humic substances
from three different lakes. Chem. Geol. 33:81-89.
Tipping, E. 1981b Adsorption of aquatic humic substances by iron oxides
Geochim. Cosmochim. Acta 45:191-199.


Absorbance Absorbance
26
Wavenumbers (cm-1)
-10-2 M Oxalate 0.00 M Oxalate
Wavenumbers (cm-1)
Fig. 2 3
F I'IR spectra of clay without organic matter after oxalate sorption at
pH 3 5 (A), 4 5 (B), and 5 5 (C).


105
of OH released to oxalate sorbed would be 2 (Chapter 2). If oxalate sorption
released previously-sorbed P into solution through ligand exchange, than one would
not expect the ratio of OH' released to oxalate sorbed to be close to 2. Data
from this study support this observation. The ratio of OH' released to oxalate
sorbed was significantly less than 2 when higher amounts of P were present on the
surfaces. However, the ratio of OH + P released into solution to oxalate sorbed
is equal to 2 for the clay fraction samples. This confirmed that oxalate released
some P directly through ligand exchange, with some sites for P and oxalate
sorption being common.
For whole-soil samples, the ratio of OH + P released to oxalate sorbed
was significantly higher than 2 when higher amounts of P were present on the
surfaces. The higher ratio suggests that more P was desorbed into the solution
with each unit of oxalate sorbed A second process which can bring P into
solution is dissolution of the soil surfaces. Release of A1 and desorption of P
each support the observation that oxalate sorption caused some surface dissolution.
Oxalate dissolution of mineral surfaces was also observed in earlier studies (Chapter
2; Stumm, 1986). Therefore, one can argue that some P also was released into
the soil solution through the dissolution of mineral surfaces.
Once P was released into solution along with Al, it could form various
solution species as A1H2P04(0H)2 and A1H2P04C204, or reprecipitate as amorphous
aluminum phosphate. For P to reprecipitate on the mineral surfaces, soluble
species of aluminum phosphate should first form in solution. This would serve as


138
Farmer, V C J O Skjemsted, and C H Thompson, 1983 Genesis of humus
B horizons in hydromorphic humus podzols. Nature 304:342-344.
Flaig, W. 1971. Organic compounds in soil. Soil Sci. 111:19-33.
Fox, T R and N. B Comerford 1990. Low-molecular-weight organic acids in
selected forest soils of southeastern USA. Soil Sci. Soc. Am. J. 54:1139-
1144.
Fox, T R N B Comerford, and W W McFee 1990a Kinetics of
phosphorus release from Spodosols: Effect of oxalate and formate. Soil
Sci. Soc. Am. J. 54:1441-1447.
Fox, T R N B Comerford, and W W McFee 1990b Phosphorus and
aluminum release from a spodic horizon mediated by organic acids. Soil
Sci. Soc. Am. J. 54:1763-1767.
Fusseder, A., M. Kraus and E. Beck. 1988 Reassessment of root competition
for P in field- grown maize in pure and mixed cropping. Plant Soil
106:299-301.
Gardner, W K D G Parbery, and D A Barber 1982 The acquisition of
phosphorus by Lupinus albus I Some characteristics of the soil/root
interface Plant Soil 68:19-32.
Genrich, D A. and J M Bremner 1974 Isolation of soil particle size
fractions. Soil Sci. Soc. Am. Proc. 38:222-225.
Goh, T B and P M Huang 1986 Comparison of the influence of citric and
tannic acids on hydroxy-Al interlayering of monmorillonite. Clays Clay
Miner. 33:210-215
Goldberg, S. and G. Sposito. 1985. On the mechanism of specific phosphate
adsorption by hydroxylated mineral surfaces: A review. Commun. Soil Sci.
Plant Anal. 16:801-821.
Graetz, D A. and V D Nair 1994. Fate of phosphorus in Florida Spodosols
contaminated with cattle manure. Ecological Engineering (In press)
Greenland, D J 1971. Interaction between humic and fiilvic acids and clay. Soil
Sci 111:34-41


53
Large releases of Al, Fe, and organic carbon during oxalate sorption further
suggest a concurrent dissolution reaction. Fox et al. (1990a) studied the kinetics
of oxalate sorption onto the same soil and reported that oxalate was sorbed during
the first 6 hrs while release of Al continued for at least 48 hrs. These data,
taken together, support Stumm's (1986) theory of congruent dissolution of mineral
surfaces by oxalate following ligand exchange. This constitutes the first such
experimental evidence to support Stumm's hypothesis. Oxalate on the clay and
whole-soil surfaces forms stable bidentate and binuclear complexes. These
complexes involve five- and six-membered rings between the oxalate and Al, which
would decrease the strength of the A1-OH-A1 bonds and bring Al-oxalate
complexes into solution
The clay fractions and whole-soil samples with organic matter released more
Al, respectively than did the clay or soil without organic matter. Another source
of Al released from the clay and soil with organic matter was the Al present as
metal-organic complexes. Lee et al.. (1988b) studied the forms of aluminum in
selected Florida Spodosols and found that more than 75 % of the Al in spodic
horizons was present as Al-fulvate. Aluminum can act as a bridge between soil
particles and organic matter. If oxalate solubilizes Al through the formation of
stable, soluble, Al-oxalate complexes in solution, then this could result in increased
Al release.


78
loading rates (Table 3.7).
Release of A1 and Fe
A small amount of A1 and Fe was released during the sorption of
phosphate and for the P + oxalate sorption treatment, these amounts were even
greater (Tables 3.3 and 3.8). Ion-activity product calculations using MINTQEA2
V3 11 showed that the system was supersaturated with respect to amorphous
aluminum phosphate at the higher loading rates for P.
Discussion
The dissociation constants (log K, and log K2) of oxalic acid are 1.27 and
4.20. At pH 4.5, both COOH groups of the oxalate will be dissociated, making
the C2042 species dominant. Assuming ligand exchange as the surface reaction this
oxalate species can form either monodentate, binuclear, or bidentate surface
complexes.
The dissociation of phosphoric acid is described by:
hjo^h2po; 1^hpo; 2^po; 3 3.5
where pK, = 2.2, pK2 = 7.2 and pK3 = 12.3. The predominant species of
phosphate at pH 4 5 is H2P04''. Thus, H2P041_ should only form monodentate or
binuclear surface complexes


:s
0 00 M Oxalate
Fig. 2 5 FTIR spectra of clay with organic matter for different concentrations
of oxalate sorbed at pH 4 5.


117
3.5 to 4.4 4.5 to 5.5 5.5 to 6.7
B -A ...0..
Fig A-4
Buffer curves for the whole-soil, used to calculate the amount of
OH' released into solution: A) With organic matter, B) Without
organic matter.


TABLE OF CONTENTS
ACKNOWLEDGMENTS ii
LIST OF TABLES v
LIST OF FIGURES viii
ABSTRACT xi
CHAPTERS 1
1 GENERAL INTRODUCTION 1
Phosphorus in flatwood Spodosols 1
Spodic Horizons 2
Phosphorus in Spodic Horizons 3
Organic Anions in Soil 5
Influence of Organic anions on Phosphorus Availability .6
2 INFLUENCED OF SOIL ORGANIC MATTER AND pH ON
OXALATE SORPTION INTO A SPODIC HORIZON 12
Introduction 12
Material and Methods 15
Results 22
Discussion 50
3 INFLUENCED OF SOIL ORGANIC MATTER AND OXALATE
ON SORPTION OF PHOSPHORUS INTO A SPODIC
HORIZON 55
Introduction 55
Material and Methods 59
Results 64
Discussion 78
iii


4
Monodentate
sOH + H2PO4 s
Bidentate
OH
OPOH + 0H
II
O
s
H2P04
,0
/p\ +
o
OH
S
S
Binuclear
\
OH + H2PO4
Figure 1.1. Three possible phosphate surface complexes. "S" could represent
either A1 or Fe.


128
Adsorption Isotherms
The Freundlich model was fitted to the desorption data. The Freundlich
model has been found to be useful in describing desorption of ions into solution
from from soil particles (Sposito, 1984). The following equation was used:
S=K[Cb]
where S is the amount of P retained per unit mass of soil (mM kg _1), C is the
equilibrium concentration of P (mM), and K and b are empirical constants for the
model K and b were calculated by a least squares fit to a linear form of the
Freundlich equation.
LogS=LogK+ LogC (C-2)
b
The derivative of this function with respect to C equals the Kd (solid to solution
partition coefficient) value which can be expressed by:
K l~'
Kd= x[C.]6 (C-3)
b
where K and b are the empirical constants calculated from the Freundlich equation.
Statistical Analysis
The linearized form of the Freundlich equation was used to test the
statistical differences between different regression lines using the General Linear


4 INFLUENCED OF SOIL ORGANIC MATTER ON DESORPTION OF
PHOSPHORUS AND OXALATE FROM A SPODIC
HORIZON 85
Introduction 85
Material and Methods 87
Results 92
Discussion 104
5 CONCLUSIONS 108
Conclusions from this Study 108
Influence of Oxalate on the P Nutrition of Trees 110
Future Research 112
APPENDICES 114
A TITRATION CURVES FOR THE WHOLE SOIL
AND THE CLAY FRACTION 114
B SORPTION OF OXALATE IN THE PRESENCE OF P
BY THE CLAY FRACTION 118
Introduction 118
Material and Methods 118
Results 120
C MEASUREMENT OF P DESORPTION BY DIFFERENT
METHODS AT VARIABLE AND
CONSTANT pH 123
Introduction 123
Material and Methods 126
Results 129
D DESORPTION OF P AND OXALATE 134
REFERENCES 136
BIOGRAPHYCAL SKETCH 147
IV


146
Viotante, A and P M. Huang 1985 Influence of inorganic and organic ligands
on the formation of aluminum hydroxides and oxyhydroxides. Clays Clay
Miner. 33:181-192.
Viotante, A and P M Huang. 1989. Influence of oxidation treatement on
surface properties and reactivities of short-range ordered precipition products
of aluminum. Soil Sci. Soc. Am. J. 53:1402-1407.
Yang, J E., E. O Skogley, and B E. Schaff 1991. Nutrient flux to mixed-bed
exchange resin: Temparature effects. Soil Sci. Soc. Am. J. 55:762-767.
yYuan, T. L. 1980 Adsorption of phosphate and water-extractable soil organic
matter material by synthetic aluminum silicates and acid soils. Soil Sci
Soc. Am J. 44:951-955.
Yuan, T. L. 1992. Three potential amendements for better fertilizer utilization in
sandy soils Soil Crop Sci. Soc. Florida Proc 51:49-53.
Yuan, G. and L. M. Lavkulich. 1994. Phosphate sorption in raltion to
extractable iron and aluminum in Spodosols. Soil Sci. Soc. Am. J. 58:343-
346.
Zelazny, I. W. and J. N. Quresih. 1973. Chemical pretreatment effect on clay
separation and mineralogical analysis of selected Florida soils. Soil and
Crop Sci. Soc. Florida Proc. 32:117-121.


APPENDIX B
SORPTION OF OXALATE IN THE PRESENCE OF PHOSPHATE ONTO A
SPODIC HORIZON
Introduction
Oxalate significantly reduced P sorption capacity (Chapter 3). It has been
observed that oxalate blocks sites on soil materials and thus reduces P sorption.
Oxalate is continuously released into solution and competes with P for sorption
sites. To understand the nature of competition between these two anions, it is
important to study the influence of P on oxalate sorption capacity as well. The
aim of this work was to investigate the sorption of oxalate in the presence of P
as influenced by soil organic matter.
Materials and Methods
Soil Material
Soil was collected from the spodic horizon (Bh) in a single soil pit of a
Pomona series (sandy, siliceous, hyperthermic Ultic Alaquod) at the Gator
Nationals Forest site located in Alachua County, 10 km northeast of Gainesville,
Florida (Swindel et al.. 1988). The soil material was air dried, passed through 2-
118


10'J M Oxalate
KT* M Oxalate
0 00 M Oxalate
FTIR spectra of clay with organic matter for different concentrations
of oxalate sorbed at pH 3 3.


142
Neary, D G, E. J Jokela, N B Comerford, S R Colbert, and T E. Cooksey.
1990. Understanding competition for soil nutrients the key to site
productivity on southeastern coastal plain Spodosols. p 432-450. In: S. P.
Gessel, D S. Lacate, G F. Weetman, and R F Powers (eds.) Sustained
Productivity of Forest Soils. Proc Am. For Soil Conf 7th, Uni of
British Columbia Faculty of Forestry Publ., Vancouver, BC, Canada
Ng Kee, K. F. and P. M. Huang. 1977. Sorption of phosphate by hydrolytic
reaction product of aluminum Nature 271:336-337.
Nye, P. H. and P. B. Tinker. 1977. Solute Movement in the Soil-Root System.
Univ. of California Press, Berkeley.
Ohman, L. and S. Sjoberg. 1988. Thermodynamic calculations with special
reference to the aqueous aluminum system PP 1-33 In J R Kramer and
H. E. Allen, (eds) Metal Speciation Theory, Analysis and application. Lewis
Pub Inc., Ann-Arbor, MI USA
Page, A L, R H Miller, and D R Keeney (ed.) 1982 Methods of Soil
Analysis, part 2. 2nd ed Agronomy 9. Am. Soc. Agron & Soil Sci.
Soc. Am., Madison, WI
Parftt, R. L. 1978. Anion adsorption by soils and soil materials. Adv. Agron.
30:1-50.
Parftt, R. L., V.C. Farmer and J.D Rusell 1977. Adsorption on hydrous
oxides I Oxalate and Benjoate on goethite J Soil Sci. 28:29-39.
Pholman, A. A., J. G. McColl, J. M. Jersak, S. C. Tan, and R. R. Northup.
1990. Organics and metal solubility in California forest soils, p 178-195.
In: S P Gessel, D S. Lacate, G F Weetman, and R F Powers (eds)
Sustained Productivity of Forest Soils. Proc. Am. For. Soil Conf. 7th,
Uni of British Columbia Faculty of Forestry Publ., Vancouver, BC,
Canada
Pohlman, A. A. and J. G. McColl. 1988. Soluble organics from forest litter and
their role in metal dissolution Soil Sci. Soc. Am J. 52:265-271.
Pohlman, A. A and J G McColl 1986 Kinetics of metal dissolution from
forest soils by soluble organic acids. J. Environ. Qual 14:86-92.


APPENDIX C
MEASUREMENT OF P DESORPTION FROM A SPODIC HORIZON USING
DIFFERENT METHODS AT CONSTANT AND VARIABLE pH
Introduction
Inorganic soil P can be present in one of three different fractions: P in
the soil solution; P on the mineral surfaces in the labile pool, or P in the
nonlabile pool. The relationship between sorbed and solution P is often described
by adsorption/desorption isotherms. Adsorption and desorption commonly exhibit
hysteresis phenomena. Raven and Hossner (1993) showed a steeper isotherm slope
for sorption than for desorption. To predict P uptake by plants, however,
desorption isotherms are more useful since they evaluate the release of P into soil
solution. The availability of P is often controlled by the desorption isotherm as a
plant progressively extracts P from the soil around its roots (Yang et al.. 1991;
Abrams and Jarrell, 1992; Graetz and Nair 1994; Harris et al.. 1994; Raven and
Hossner, 1994). The desorption of soil P is a function of various soil parameters
mainly: a) the concentration of sorbed P; b) the concentration of P in the soil
solution; and c) the rate of P desorption into solution.
Spodic horizons can contain high amounts of amorphous Al oxides, which
can sorb P. Some of the sorbed P is in a form unavailable to plants (Ballard and
123


Absorbance
22
Wavenumbers
Wavenumbers
Fig. 2.1 FTIR spectra of clay (a) without organic matter and (b) with organic
matter


15
al., 1988). The soil material was air dried, passed through a 2-mm sieve and
stored in plastic bags.
Preparation of Whole Soil
Two sets of soil samples were prepared: one with organic matter and the
other without organic matter. One portion was treated with hot, 30% H202 to
remove organic matter as outlined by Kunze and Dixon (1986). This sample was
centrifuged and washed several times with distilled water, while the other portion
was not treated Both soil samples (with and without organic matter) then were
saturated with Na using 0.01 M NaCl Excess Na was removed with four or
five centrifugal washings using distilled water. The samples were oven dried at
110 C.
Preparation of the Soil Clay Fraction
The water-dispersible clay fraction was separated by ultrasonification at a
1:5 soil to water ratio (Genrich and Bremner, 1974). After ultrasonification, wet-
sieving was used to remove the sands. The clay fraction (< 0.2 pm diameter)
was separated by centrifugal sedimentation (Jackson, 1979). The resulting clay was
divided into two portions. One portion was treated with hot, 30 % H20, to
remove organic matter as outlined by Kunze and Dixon (1986). The other portion
was not treated. Both clay samples (with and without organic matter) then were


87
The influence of soil organic matter on P sorption has been investigated by
many researchers. Many have suggested that sorption of P and organic matter
occurs on the same sorption sites. Yuan (1980) studied the sorption of P and
water-extractable soil organic matter by soil and A1 oxides. He reported that most
of the sorption sites for organic matter and P sorption may be different, though
some sites are common for both P and organic matter Fulvic acid was found to
reduce P adsorption considerably. This effect of fulvic acid is apparently due to
the chelating ability of fulvic acid's -COOH and -OH functional groups for A1 and
Fe (Parfitt, 1978; Sibanda and Young, 1986). The results in Chapter 3 showed
that soil organic matter significantly reduced P sorption. However, the mechanism
for P desorption in the presence of organic matter apparently has not be studied
The objectives of this study were to investigate: i) P or oxalate desorption
from the clay fraction and whole soil of a spodic horizon in the presence of
oxalate or P, respectively, and ii) the influence of soil organic matter on the
desorption of P and oxalate
Materials and Methods
Whole-Soil Material and the Clay Fraction
The whole-soil and clay-fraction samples that were used for sorption studies
of oxalate (Chapter 2) and phosphate (Chapter 3) were further used to study the
desorption of anions at pH 4 5


LIST OF TABLES
Table Page
2.1. Parameters of the linear regression models for oxalate sorption
by the clay fractions and the whole-soil samples 25
2.2. Parameters of the linear regression models relating
release of OH ions and organic carbon for the
clay fractions and whole-soil samples 34
2.3. Ratios of OH' ions released to oxalate sorbed
for the clay fractions and whole-soil samples 37
2.4. Parameters of the linear regression models relating release of
aluminum and iron with oxalate sorption for the clay
fractions and whole-soil samples 40
2.5. E4/E6 ratio of organic carbon released into solution
during the sorption of oxalate by the clay fractions 45
2.6. Intensity of the absorption bands at 1610 cm'1
(aromatic C=C and/or H-bonded C=0 stretching
of COOH) and 1460 cm'1 (OH deformation,
C-0 stretching of phenolic-OH, and/or C-H deformations
of CH3 and CH2) with standard deviations
for clay with organic matter at different pH values for
different amount of oxalate added 49
3.1. Langmuir sorption isotherm parameters for P sorption onto
the whole-soil samples and clay fractions with and
without organic matter 66
3.2. P sorption onto the whole soil material compared to P
that of sorption onto the clay fraction
expressed on a whole soil basis 69
v


12
oxalate (up to 2 mM) and other carboxylic anions in a variety of sediments, in the
forest floor, in bulk soil, and in the rhizosphere of forest trees.
Oxalate alters chemical processes in soils through complexation reactions
with A1 and Fe that occur in the soil solution and on the surfaces of soil particles
(Stumm, 1986; Martell et al.. 1988). It has acidic characteristics due to the
presence of -COOH groups and reacts with mineral soil surfaces through i)
electrostatic attractions, ii) complex or chelate formation, and iii) water bridging
(Tate and Theng, 1981). Three types of inner-sphere oxalate complexes with soil
surfaces are possible: monodentate, bidentate, and binuclear (Parfitt et al.. 1977).
Each requires the displacement of OH2 or OH' that had been coordinated to Al
and Fe atoms on the surfaces. Such a reaction would increase the pH of the
system. Therefore, the amount of OH' released could be used to identify possible
reactive sites and possible mechanisms of oxalate sorption at different
concentrations of oxalate and/or pH. A change in pH would alter the soil surface
charge and, possibily, the species of organic ligand present in the system.
Spodosols of the southeastern Coastal Plain have an accumulation of
organic matter in their spodic horizons. As organic anions originate in the forest
canopy or forest floor, they form mobile complexes with Fe and Al, and move to
the Bh horizon where they are constrained as organo-minerals. Researchers have
shown that organic matter may be immobilized through complex interactions with
mineral surfaces (Greenland, 1971; Davis and Glour, 1981; Sibanda and Young,


121
where Pe is efficiency of P to reduce oxalate sorption (%), OXp is oxalate sorbed
in the presence of P; and Ox0 is oxalate sorbed alone. Efficiency of phosphate in
reducing the sorption of oxalate in the P + oxalate system for the clay fractions is
presented in Table B-l. P was more effective in reducing oxalate sorption onto
the clay fraction when organic matter was not present.


23
Sorption of Oxalate by the Clay Fractions
Sorption isotherms for oxalate on clay samples with organic matter and
without organic matter at constant pH are presented in Figure 2.2. Linear
isotherms provided best fit for the data within the range of oxalate concentrations
used. Parameters of the model at different pH values are presented in Table 2.1,
with oxalate sorption being strongly dependent on pH. As pH increased, the
oxalate sorption decreased. Soil organic carbon also had a significant influence on
oxalate sorption at all pH values, with approximately a 2 to 3 fold reduction in
oxalate sorption in the presence of organic carbon (Table 2.1).
FTIR spectra at the highest oxalate loading are presented in Figure 2.3 for
clay without organic matter and in Figures 2.4, 2.5, and 2.6 for clay with organic
matter A small shoulder was seen at 1710 cm"1 for the FTIR spectra of clay
with organic matter for the 10'2 M oxalate loading at both pH 3.5 and 4.5.
Parfitt et al. (1977) also observed the same bands for oxalate sorption by goethite
and proposed the formation of a binuclear complex between oxalate and two Fe34
ions. If inner-sphere complexes form between oxalate and clay surfaces, they must
involve the coordination of carboxylic groups from the oxalate ion with Al and/or
Fe atoms of the clay surface, through oxygen atoms. However, the absorption
bands linking Al-oxalate vibrations, which would normally provide the most direct
information on complexation, are too weak to be definitive. It was not possible to
isolate these bands from the FTIR spectra of the clay due to the low surface
concentrations of the oxalate.


35
Fig
10 Kinetics of OH' ions release during the sorption of oxalate by the
clay fractions at pH 4 5.


Organic Carbon Release (mg g )
43
Release of organic carbon by the clay fractions at varing pH as a
function of oxalate sorption


6
the identified, free, low molecular weight, organic anions in a group of Spodosols
from north Florida. They also reported that the oxalate levels in soil solutions
averaged an order of magnitude higher in spodic horizons than in the
corresponding surface A horizon. These data show that oxalate is present and, if
it also affects P sorption, could play a significant role in forest productivity.
Influence of Organic Anions on Phosphorus Availability
Fox et al (1990a) proposed that, in lower Coastal Plain Spodosols, low
molecular weight organic anions, whether leached from the surface horizon or
produced in situ, might stimulate the release of phosphorus from mineral surfaces.
Organic anions acting as ligands are known to release P by i) replacing P sorbed
at surfaces of Al or Fe oxides through ligand-exchange reactions (Huang and
Schnitzer, 1986); ii) dissolving metal oxide surfaces and releasing sorbed P (Martell
et al.. 1988); and iii) complexing Al and Fe in solution, thus preventing the re
precipitation of metal-P compounds (Ng Kee and Huang, 1977). It has also been
observed that organic anions may block sites on mineral surfaces and reduce P
sorption (Kafkafi et al.. 1988).
Huang and Violante (1986) studied aluminum-citrate complexes in aqueous
solution They reported that Al and citrate form a 1:1 complex where the citrate
ligand occupies three of the six coordination sites around each Al while each of
the other three sites are occupied by a water molecule. Occupation of
coordination sites by citrate instead of water imposes a restriction on the


3
development of a spodic horizon (De Coninck, 1980; Farmer et al.. 1983;
Buurman, 1984; Tan, 1986; Ugolini et al., 1988). Therefore, the complexation and
translocation of Al and Fe by organic acids is a primary mechanism during the
podzolization process. This can result in large accumulations of organic carbon
(which is mostly as humic and fiilvic acid), Al, and Fe in spodic horizons.
Phosphorus in Spodic Horizons
Spodic horizons contain elevated levels of amorphous or poorly crystalline
Al oxides, which can sorb P Not all of this sorbed P is in plant-available form
(Ballard and Fiskell, 1974). Large quantities of total P may be present in spodic
horizons, but the level of water-soluble P tends to be quite low. The
physico-chemical sorption of P is as an inner-sphere complex at the surface of Al
and Fe hydroxides and at the broken edges of silicate clay minerals (Sposito,
1984). An inner-sphere P complex refers to a surface complex resulting from
ligand exchange between a surface Lewis acid site (S) and the adsorbed ion
(Parfitt, 1978; Goldberg and Sposito, 1985). Such complexes are quite stable,
showing mainly covalent or ionic bonding character.
Three types of inner-sphere P complexes have been postulated: monodentate,
bidentate and binuclear (Fig. 1; Parfitt et al.. 1977). Tejedor-Tejedor and
Anderson (1990) used fourier transform infrared spectroscopy (FTIR) to study the
sorption of orthophosphate onto goethite particles in an aqueous suspension. They
observed the formation of three different types of complexes: protonated and


141
Lopez-Hernandez, D, G. Siegert, and J.V. Rodriguez. 1986. Competitive
adsorption of phosphate with malate and oxalate by tropical soils. Soil Sci.
Soc. Am. J. 50:1460-1462.
Martell, A. E., R.J Motekaitis and R M. Smith. 1988. Structural stability
relationships of metal complexes and metal speciation in envirionmental
aqueous solutions. Environ. Toxicol. Chem.7:417-434.
Martell, A. E. and R M. Smith. 1982. Critical Stability Constants, Vol. 5: First
supplement. Plenum Press. New York.
Marzadori, C, L. V. Antisari, C. Ciavatta, and P. Sequi 1991. Soil organic
matter influence on adsorption and desorption of boron. Soil Sci. Soc Am.
J. 55:1582-1585.
Menon, R G, I I Hammond, and H Sissingh 1989. Determination of plant-
available phosphorus by iron hydroxide impregnated paper. Soil Sci. Soc.
Am. J. 53:110-115.
Miller, W P., L W Zelazny, and D C. Martens 1986. Dissolution of
synthetic crystalline and noncrystalline iron oxide by organic acids.
Geoderma 37:1-13.
/vloore, T R W Souze, and J F Koprivnjak 1992 Controls on the sorption
of dissloved organic carbon by soils. Soil Sci. 154:120-129.
Murphy, J. and J P Riley. 1962. A modified single solution method for the
determination of phosphate in natural waters. Anal. Chim. Acta 27:31-36.
Nagarajah, S., A M Posner, and J. P Quirk 1968. Desorption of phosphate
from kaolinite by citrate and bicarbonate. Soil Sci. Soc. Am. Proc. 32:507-
510.
Nagarajah, S., A M Posner, and J P Quick. 1970 Compatitive adsorption of
phosphate with polygalacturonate and other organic anions on kaolinite and
oxide surfaces. Nature 228:83-85.
Nanzyo, M. 1987. Formation of noncrystalline aluminum phosphate through
phosphate sorption on allophanic Ando soils. Commun Soil Sci. Plant
Anal. 18:735-742.


95
- OM + OM OM + OM
Fig. 4.2
Influence of previously sorbed P on oxalate sorption and P
desorption for clay fraction.
Phosphate desorbed (mmol kg )


45
Table 2.5. E4/E6 ratio of organic carbon released into solution during the
sorption of oxalate by the clay fractions (Mean SD; n=3).
Treatment
Concentration of
oxalate added (mM)
E4/E6 ratio
3.5
10*
3.150.10
1.0
7.540.25
0.1
8.240.56
0.0
9.260.07
4.5
10
1 880.13
1.0
7.130.04
0.1
8.080.13
0.0
9.070.15
5.5
10
2.140.06
1.0
6.930.05
0.1
7.650.21
0.0
9.070.15
* Concentration of oxalate added into solution


135
Table D-2 Influence of oxalate previously present on P sorption and release of
oxalate, Al, Fe, OH, and organic carbon by clay fractions and whole-
soil samples (Mean SD, n = 3).
Oxalate
P
Amount Released
Previously
Sorbed
(mmol kg"1)
O
o
>
Present
(mmol kg"1)
(mmol kg'1)
Ox
A1
Fe
OH
(mg g'1)
Clay without Organic Matter
45617
2711
29.30.5
361
0.050.02
2566
482.4
2831
9.20.4
330
0.010.01
28613
2.70.5
3051
1.10.2
211
0.080.0
30020
Clay with Organic Matter
27514
1192
0.490.02
210.8
0.810.03
1075
211
430.2
1793
0.240.1
80.5
0.210.03
17518
461
4.80.2
984
0.090.06
80.5
0.150.02
1887
461
14.90.7
8.30.1
Soil Without Organic
1.10.06 3.40.1
matter
0.030.01
9.01
12.40.4
9.20.2
0.20.01
2.10.1
0.030.01
9.31
9.60.5
10.80.3
0.20.01
1.70.1
0.030.01
12.52
6.20.7
12.90.1
0.20.01
1.50.1
0.040.01
12.52
4.00.1
13.80.1
0.10.01
1,60.1
0.020.01
11.71
0.80.1
14.10.1
0.10.01
1.10.1
0.010.01
14.12
11.61.0
5.70.1
Soil With
1 90.1
Organic Matter
4.60.1 0.060.01
5.91
l.liO.l
8.80.6
7.80.2
0.90.1
4.10.1
0.040.01
8 62
1.00.1
7.40.1
9.10.4
0.20.1
1,60.1
0.020.01
9.61
0.80.0
5.50.2
10.30.8
traces
1 40.1
0.010.01
11.11
0.70.1
4.00.1
10.90.3
traces
1.20.1
0.010.01
10.61
0.70.1
0.80.1
10.90.1
traces
0.70.1
O.OliO.Ol
11.41
0.60.1
A
OC = Organic Carbon


30
Oxalate Sorption by Whole Soil Samples
Oxalate sorption isotherms for the whole-soil samples are presented in
Figure 2.7. The Langmuir model gave the better fit. The parameters of the
linearlized form of the Langmuir models are presented in Table 2.1. The sorption
maxima (Smax) by the whole soil material was 10.57 for soil with organic matter
and 14.20 for soil without organic matter The presence of soil organic matter
significantly reduced oxalate sorption.
Release of OH
The sorption of oxalate onto the clay fractions and whole-soil samples
released large amounts of OH ions (Fig.2.8 and Fig. 2.9). Oxalate sorption was
significantly and positively correlated with OH ions released (Table 2.2). The OH'
release was also significantly influenced by soil organic matter (Table 2.2). The
amount of OH ions released at pH 3.5 into solution by the clay fractions was
significantly different than the amount of OH ions released into solution at pH 4.5
and 5.5. The FTIR spectra of clay without organic matter showed considerable
decrease in the intensity of OH-bands in the region 3000 to 3800 cm'1 following
oxalate sorption (Fig. 2.3).
The OH release in solution follows first order kinetics (Figure 2.10) for
oxalate sorption by clay surfaces at pH 4.5. Release of OH was initially rapid
during the sorption of oxalate (Fig. 2.10), and remained high even after 24 hrs
when organic matter was not present. The ratios of moles of OH' released per


115
a* Y = 0.079 + 1.84 X R2 = 0.98
3.4^4.4 4.4 5.6 5.5 6.7
Change In pH
Fig. A-2
Buffer curves for the clay fraction, used to calculate the amount of
OH released into solution: A) With organic matter, B) Without
organic matter.


34
Table 2.2 Parameters of the linear regression models relating release of OH ions
and organic carbon for the clay fractions and whole-soil samples.
Material
OM/pH
OH
ions released
Organic carbon released
Slope
Intercept
R2
Slope
Intercept
R2
Clav fractions
With organic
matter
3.5
0.65a
13.73a
0.99
0.081a
2.52a
0.99
4.5
1.65b
- 4.07a
0.99
0 13 lb
6.34b
0.99
5.5
1.88b
1.30a
0.98
0.249c
5.85b
0.99
Without organic matter
3.5
0.7 Ia
58 4b
0.97
4.5
1 82b
73.8b
0.99
5.5
1 73b
40.6b
0.99
Whole-Soil
With organic
matter
4.5
1 99b
-0.07b
0.95
0.73
-0.43
0.93
Without organic matter
4.5
2.48a
-0.14a
0.96
The same letter in a column at different pH values indicates a lack of significance
at the 5 % level within the clay fractions or the whole-soil samples.
@ Includes all samples at pH 3.5, 4.5, and 5.5;


ACKNOWLEDGMENTS
It is my privilege to express my sincere gratitude, appreciation and heartfelt
thanks to Nick B Comerford, my advisory committee chairman, for his able
guidance, keen interest, constructive criticism and constant encouragement during
the course of this investigation. I feel pleasure also in expressing high regards to
other members of my advisory committee, Earl Stone, Brian McNeal, Cliff
Johnston, P Surash Rao, H Gholz and Eric Jokela, for my scientific development
The financial support provided by the National Science Foundation for
funding my assistantship and the experimentation is greatly acknowledged. I
appreciate also the valuable technical assistance of Mary McLeod, Randy, and Dr
Cliff Johnston at various phases of the work. Mary's friendship was invaluable as
was her advice on a wide range of issues. My sincere thanks are also due to
everyone in the Forest Soils Laboratory for their cooperative attitude and
discussions during the various phases of this investigation. I enjoyed all of my
new friendships, especially those of my fellow graduate students.
Finally, I want to thank my wife Gurmeet and my son Kulraj for their
patience, love and understanding during this adventure. Without their tolerance and
moral support, I would have not done it
11


120
Oxalate in the extract was determined by HPLC (Fox and Comerford,
1990a) using a Hamilton PRP-X300 150 x 4.1 mm organic acid column (Hamilton
Co., Reno, Nevada), with a Gilson single piston high pressure pump along with a
Pheodyne model 7125 injection valve fitted with a 20 fuL injection loop. This
1TPLC system uses a Gilson Holochrom variable wavelength UV detector in
conjunction with a Gilson computerized integrator. The eluent was 0.005 M
H2S04 at a flow rate of 2 mL min'1 Oxalate concentration was calculated from
the calibration curve obtained with standard solutions of 0.1 to 10 mM.
Statistical Analysis
Statistical differences in the sorption of oxalate in the presence of P were
compared. Significance of the difference between two means were determine by a
t-test (Snedecor and Cochran, 1980). Standard errors of means are given as well.
Results
Less oxalate was sorbed by the clay fractions and by the whole-soil samples
at high P concentrations. To evaluate the ability of P to depress oxalate sorption
when both P and oxalate were added together, the percent efficiency of P in
reducing the oxalate sorption was calculated according to the expression of Deb
and Datta (1967):
P -
1~
Ox
p
Ox
O
xioo
(B-l)


143
Polglase, P. J., N. B Comerford, and E. J. Jokela. 1992. Mineralization of
nitrogen and phosphorus from soil organic matter in Southern pine
plantation. Soil Sci. Soc. Am. J. 56:921-927.
Pritchett, W L and N B Comerford 1983. Nutrition and fertilization of slash
pine p 69-90 In E.L Stone (ed.) The Managed Slash Pine Ecosystem
Symp Proc., Gainesville, FL. 9-11 June 1981 Univ Florida, Gainesville
Raven, K P and L. R Hossner 1993. Phosphorus desorption quantity-intensity
relationships in soils. Soil Sci. Soc. Am. J 57:1501-1508
Raven, K P and L R Hossner 1994 Soil phosphrus desorption kinetics and
its relationship with plant growth. Soil Sci. Soc. Am. J. 58:416-423.
Reddy, M. N., G. Ramagopal, and A. S. Rao 1977. Phenolic acids in
groundnut seed exudates. Plant Soil 46:655-658.
/Ryden, J C. and J K Syers. 1987 Origion of the labile phosphate pool in
soils. Soil Sci. 123:353-361.
SAS Institute. 1985. SAS Users Guide. Statistics. Version 5 ed. SAS Inst.,
Cary, NC.
/Sibanda, H. M., and S. D. Young. 1986. Competitive adsorption of humus
acids and phosphate on geothite, gibsite and two tropical soils. J. Soil Sci.
38:211-217.
Smith, W H. 1969. Release of organic materials from the roots of tree seedlings.
Forest Sci 15:138-143
Snedecor, G. W. and W. G. Cochran. 1980. Statistical Methods. The Iowa St
University Press. Ames, IA.
Sposito, G 1984 The Surface Chemistry of Soils. Oxford Univ Press, New
York
Stainton, M. P. 1980. Error in molybdenum blue methods for determining
orthophosphate in freshwater. Can. J. Fish. Aquat. Sci. 37:472-478.
Stevenson, F J 1982. Humus Chemistry John Wiley and Sons, New York


107
As P sorption was not affected by the presence of previously sorbed oxalate, one
can conclude that many sites are highly specific for P.
Conclusions
This investigation elucidate the mechanism of P and oxalate desorption from
spodic horizon materials. Results from this study confirmed that oxalate released
P from the mineral surfaces through two processes: i) ligand exchange replacing P
from the mineral surfaces, and ii) dissolving the mineral surface, thereby bringing
sorbed P into solution. Results from the preceding chapter and this study
establish that the presence of soil organic matter and oxalate increases the amount
of P desorbed. Soil organic matter and oxalate significantly increase the P
concentration in solution.


49
Table 2.6. Intensity of the absorption bands at 1610 cm'1 (aromatic C=C and/or
H-bonded OO stretching of COOH) and 1460 cm'1 (OH
deformation, C-0 stretching of phenolic-OH, and/or C-H
deformations of CH3 and CH2) with standard deviations, for clay
with organic matter at varing pH values for different amount of
oxalate added (Means SD; n=3).
pH
Oxalate added
mM
Organic carbon
released mg g'1
V 1610 cm'1
V 1460 cm"1
3.5
10.0
35.65a0.60A
60.74i4.93
35.05i3.54
1.0
6.670.44
74.95i3.08
40.32i2.04
0.1
3.930.29
79.68i2.37
43.83il.65
0.0
3.340.22
80.40i4.57
46.85i4.42
r =
- 0.93 r =
- 0.93&
4.5
10.0
38.130.56
54.39il.04
32.14 1.19
1.0
15.44i0.85
72.0il.69
42.15i0.22
0.1
7.260.45
79.18i2.95
43.38i2.19
0.0
5.420.25
83.76i8.86
43.65i3.67
r =
- 0.93 r =
- 0.95&
5.5
10.0
42.09i0.74
39.49i0.19
17.08i0.45
1.0
18.95il.04
61.78i2.07
32.78i3.56
0.1
6.20.63
67.93il.79
37.57i2.39
0.0
5.35i0.61
70.40il.74
39.68i0.60
r =
- 0.98
r = 0.97&
@ Correlation coefficient between organic carbon released and intensity of the
absorption bands at 1610 cm'1 at pH 3.5, 4.5 and 5.5.
& Correlation coefficient between organic carbon released and intensity of the
absorption bands at 1460 cm"1 at pH 3.5, 4.5 and 5.5.


D-l Influence of previously sorbed P by clay and whole-soil
samples on the oxalate sorption and release of P,
Al, Fe, OH, and organic matter 134
D-2 Influence of oxalate previously present on the P
sorption and release of oxalate, Al, Fe, OH,
and organic carbon by clay fractions and whole-soil samples 135
Vll


41
120
100
C
.2 80
3
O
(0
- 60
(0
a> 40
o
a>
a
20
Al
3+
Al-(oxalate)
Al-(oxalate),
1
3.5
4.0
4.5
5.0
pH of the system
xxjx:x
6.0
Fig. 2.13
Concentrations of aluminum species in solution at different pH levels
in the presence of oxalate.


99
Table 4.3 Influence of oxalate sorption by clay fractions and whole-soil samples
on the ratio of OH released, and OH + P released, to oxalate
sorbed (Mean SD, n=3).
Without Organic Matter With Organic Matter
P Prev Ratio P Prev Ratio
Present Sorbed
(mmol kg'1)
OH/OX
(OH + P)/OX
(mmol kg'1) OH/OX
(OH +P)/OX
Clav fraction
5459
1.87*0.1
2.12*0. lab
3535
.SiO.r
1.92*0. lab
91*0.3
1.860. laa
2.050.1aa
721
1.83*0.1
1.96*0.06
9.50.5
1.90*0.1
1.91*0.1
91
1.89*0.0
1.85*0.10
Whole Soil
Material
24.50.7
1.64*0.1
3.220.2ab
16.70.5
1.75*0.1
4.210.2ab
21.7*0.6
1.73*0.1
3.11*0 lab
15.51.4
1.79*0.1
3.140.2bb
16.2*0.2
i.76o.i
3 040.2ab
10.20.7
1.84*0.1
2.83*0. lbb
7.10.1
i.69o.i
2.16*0. lbb
5.40.2
1.81*0.2
2.450.4cb
3.70.1
1.83*0. lab
2.06*0. lba
4.10.1
1.85*0.1
2.320.3ca
0.80.1
2.01*0.lba
1.930. lba
0.90.1
1.85*0.1
1.88*0.1
The first superscript letter indicates significant differences within a column and
material, while the second letter indicates significant differences between rows.


131
Table C-l Freundlich isotherm parameters for the desorption of P from soil using
the dilution and sequential extraction methods at variable and
constant pH
Desorption
pH
Parameters of the Freundlich isotherm
Method
Conditions
Slope
Intercept
R2
Dilution
Constant
0.121a
2.20a
0.96
Variable
0.114a
2.12b
0.94
Sequential
Constant
0.088b
2.08b
0.95
Variable
0.083b
2.14b
0.94
Dilution
0.117a
2.16a
0.99
Sequential
0.086b
2.1 lb
0.98
@Includes all samples at variable and constant pH
Values within columns for different methods followed by the same letter are not
significantly different at the 5 % level.


76
Fig. 3.5
Kinetics of OH release during sorption of anions by the clay
fraction.


54
Effect of Organic Carbon on Oxalate Sorption.
The removal of organic carbon increased the sorption of oxalate. This can
be attributed to either competition between organic carbon and oxalate for same
sorption sites or formation of new sorption sites on mineral surfaces when H202
was used to oxidize the organic carbon These data do not differentiate between
these possibilities; however, either option would enhance the surface area for
oxalate sorption. Zelazny and Quresih (1973) reported that H202 treatment of clay
material from Florida soils enhanced surface area and decreased surface charge.
This evidence suggests that, with the removal of organic matter, surface area
would increase somewhat resulting in greater sorption of oxalate by clay and soil
surfaces without organic matter.


2
in their surface horizons (A and E), either due to low clay contents or the nature
of the clay fractions (Ballard and Fiskell, 1974; Fox et al., 1990b; Yuan, 1992).
For the surface soils, a clay content as low as 10 g kg'1 is common. Thus P,
along with organic matter, can be leached from the surface horizon and accumulate
in the underlying spodic horizon.
Spodic Horizons
Organic matter migrates through the A horizon of Spodosols in soluble and
colloidal forms and is adsorbed or precipitated, immobilized in the Bh horizon.
Current concepts of the formation of spodic horizons are based on the formation
of Al and or Fe humic complexes. Stability and mobility of these compounds
depend on the metal concentration in the soil solution. If the amount of Al
and/or Fe available for organo-metal complex formation is low in the A horizon,
complexes will be formed in the A horizons with low metal/organic ratios. In this
case the amount of Al and/or Fe chelated is insufficient to cause immobilization of
metal organic compounds, and may then move down in the pedon. During the
downward migration, these metal-organic compounds (De Coninck, 1980),
concurrently sorbed more polyvalent cations, which results in a progressive decrease
of their net negative charge. The presence of higher concentrations of Al and Fe
in the subsoil and/or at pH values different from that of the surface horizon may
eventually neutralize the remaining charge. This results in the precipitation of
metals along with organic matter in subsurface horizons and leads to the


Oxalate sorbed (mmol kg
96
-OM + OM -OM + OM
+OX +OX -P -P
O -A-
Phosphate previously sorbed (mmol kg )
Fig. 4.3 Influence of previously sorbed P on oxalate sorption and P
desorption by whole-soil samples.
Phosphate desorbed (mmol kg


129
Models procedure in the SAS framework (SAS Institute, 1985). The model used
was:
Va + PV6ff (c-4)
where a is the intercept, P is the slope of the line and e is the error term. To
compare two regression lines, we have compared the Pj and P2 values of the two
lines. The ANOVA and t tests were used to test the difference between a, and
a2, and between Pj and P2.
Results
Desorption Isotherms
The desorption isotherms obtained by dilution and sequential extraction
methods under conditions of constant and variable pH are presented in Figure C-l.
Desorption isotherms followed the Freundlich model. With parameters of the
isotherms as given in Table C-l. The slopes of the regression lines for the
different methods illustrate hat the change in surface-sorbed P relative to solution P
were significantly different for the two methods. The variable vs constant pH
approaches did not significantly influence the amount of P released. The dilution
method, both at variable and constant pH, had a statistically lower release of P
into solution compared to the sequential extraction method at the same level of P
present on soil surfaces.


47
Fig. 2.17
Relationship between A1 release and organic carbon release from the
whole-soil samples.


CHAPTER 3
INFLUENCE OF OXALATE AND SOIL ORGANIC MATTER ON SORPTION
OF PHOSPHORUS ONTO A SPODIC HORIZON
Introduction
Phosphorus deficiencies are common on the poorly-drained Spodosols of the
flatwoods region of the lower Coastal Plain of the Southeastern United States
(Pritchett and Comerford, 1983, Comerford et ah 1984). P availability depends on
physico-chemical properties such as P sorption by colloidal surfaces. Organic
ligands are continuously released into the rhizosphere by decaying plants and
animals, through microbial processes, and as root exudates (Stevenson, 1982, Fox
and Comerford, 1990). The supply of P to plants is strongly influenced by the
presence of these organic ligands. Different organic anions have been reported to
modify the sorption of phosphate by soils and soil components (Deb and Datta,
1967; Earl et al.. 1979; Kafkafi et al.. 1988).
Aluminum and iron, either in solution or as crystalline/amorphous soil
constituents, are the principal agents responsible for chemical fixation of phosphate
in acid soils (Yuan and Lavkulich, 1994) Organic anions which are capable of
forming stable complexes with aluminum and iron in solution (Appelt et al.. 1975;
Viotante and Huang, 1985; Trana et al.. 1986a, 1986b; Huang and Schnitzer,
55


104
Release of Al. Fe. and Organic Carbon
Small amounts of Al, Fe and organic carbon were released into solution
during P sorption. The amounts of A1 and Fe desorbed by clay and soil with and
without organic matter are presented in Appendix D-2 The amounts of A1 and
Fe released to solution were higher for the clay fraction and for the whole-soil
samples having organic matter.
Discussion
Phosphate Desorption by Oxalate
It has been previously shown that i) oxalate and P are retained on
surfaces of clay particles and whole-soil samples through ligand exchange (Chapter
2; Chapter 3; Parfitt et al.. 1977; Sposito, 1984; Goldberg and Sposito, 1985); ii)
oxalate sorption causes dissolution of the mineral surfaces (Chapter 2; Stumm,
1986); and iii) at pH 4.5, oxalate formed a bidentate surface complex (Chapter 2)
while P was sorbed as monodentate and/or binuclear surface complexes (Chapter
3).
We postulated that oxalate released P from the mineral surfaces through
two processes: i) ligand exchange replacing P from the mineral surfaces; and ii)
dissolution of the mineral surfaces and release of P into solution.
The data indicated that, when high amounts of P were present on the
mineral surfaces, P was released into solution during oxalate sorption. At pH
4.5, oxalate would form bidentate complexes with the mineral surface and the ratio


89
fractions. Oxalate concentrations used for the whole-soil sorption isotherms were 0
mM, 0.1 mM, 0.5 mM, 1.0 mM, 3.0 mM, 7.0 mM and 10.0 mM. At the end of
the reaction period, each suspension was centrifuged at 12,000 g for 20 minutes
Filtrates then were equilibrated with 25 ml of a 5 mM P solution for the clay
fraction or with 50 ml of solution for the whole-soil samples. The suspension
then was shaken for another 24 hrs. pH of the suspension was periodically (every
four hours) adjusted to pH 4.5 with 0.1 M HC1 or NaOH over the 24 hrs of the
experiment Two drops of toluene were added to inhibit microbial growth, and
samples were placed in a reciprocating shaker for 24 hrs. At the end of the
reaction period, each suspension was centrifuged at 12,000 g for 20 minutes. The
filtrate was used for analysis of oxalate, Al, Fe, inorganic P and organic carbon.
Sorbed phosphate was calculated from the difference between the initial and final
phosphate solution concentrations.
Chemical Analysis
Solution pH was measured using a combination glass electrode on an Orion
pH meter. Aluminum was determined using a flame emission spectrophotometer
with N20-C2H2 flame, and Fe was determined using a C2H2 flame. Total organic
carbon of the clay suspension with organic matter was measured using a persulfate
oxidation followed by IR analysis of the CO, produced on a TOC apparatus
(College Station, TX).


100
Desorption of Oxalate in the Presence of Phosphate
Desorption of Oxalate and Sorption of Phosphate
Small amounts of oxalate were desorbed by the whole-soil samples and the
clay fraction with the sorption of phosphate (Figs. 4.4 & 4.5). Higher amounts of
oxalate were released from the soil and clay containing organic matter. The
sorption of P by clay and soil surfaces is presented in Figures 4.4 and 4.5. The
presence of previously sorbed oxalate onto the clay surface had no significant
influence on P sorption in the present experiment.
Release of Hydroxyls
The amounts of hydroxyl released to solution during P sorption are
presented in Appendix D-2. The ratio of OH' released to amount of P sorbed
was calculated and shown to vary between 0.89 and 1.08 for the clay fractions
and between 0.85 to 1.15 for Values were not significantly different from 1 (Table
4.4), and ratios were not influenced by the amount of oxalate desorbed into
solution
The ratio of OH + Ox released into solution to P sorbed was also
calculated (Table 4 4) The two ratios (OH/P and (OH + Ox)/P) were not
significantly different for the clay samples. However, at high amounts of oxalate
previously present on whole-soil surfaces, OH + Ox/ P ratio were significantly
greater than 1.


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
INFLUENCE OF SOIL ORGANIC MATTER ON PHOSPHORUS AND
OXALATE SORPTION AND DESORPTION IN SPODOSOL
By
JAGTAR S BHATTI
May, 1995
Chairperson: Dr. N. B Comerford
Major Department: Soil and Water Science
Phosphorus (P) deficiencies are common for poorly-drained Spodosols of the
flatwoods region of the Lower Coastal Plain of the Southeastern United States.
Large quantities of total P may be present in spodic horizons, but the level of
water-soluble P tends to be quite low. The goal of this investigation was to study
P and oxalate sorption and desorption by the clay fraction and whole-soil material
of a spodic horizon as influenced by soil organic matter. Understanding the
sorption mechanisms for P in the presence of oxalate and soil organic matter
should form the basis for explaining P extraction by roots from soil surfaces, and
the buffering of P levels in soil solutions.
The spodic horizon had a higher capacity to sorb P compared to oxalate,
xi


50
FTIR data also indicated the presence of a large number of OH functional
groups. Since this soil has a very low Fe concentration, we assume that the OH
groups were primarily attached to Al. The terminal A1 on the mineral surfaces
should be Al-OH2+ (aqua), Al-OH (hydroxy), or Al-O" (oxo) groups, depending on
pH (Sposito, 1984). Since these spodic horizons have no anion retention capacity,
the number of Al-OH2+ (aqua) functional groups is insignificant. In the pH range
of this study rules out the significance presence of Al-O', therefore Al-OH
(hydroxy) was assumed to be the dominant surface group.
Solution pH determines the oxalate species available for a surface reaction.
The dissociation of oxalic acid is:
HOOC-COOH ^ HOOC-COO' + H* pKt = 1.2 (2 4)
HOOC-COO ^ OOC-COO + H+ pK2 = 4 2 (2.5)
These pK values would result in dissociation of only one COO' group at pH 3.5,
while at pH 4.5 and 5.5 both COO' groups of oxalate would be dissociated.
Therefore, at pH 3.5, oxalate should only form monodentate and/or binuclear
surface complexes, while at pH 4.5 and 5.5, bidentate surface complexes can be
formed.
Oxalate Sorption
Oxalate sorption onto the mineral surfaces may be through ligand exchange
and/or precipitation. If the sorption process were dominated by ligand exchange


101
-OM + OM -OM + OM
Fig. 4.4
Influence of previously sorbed oxalate on P sorption and oxalate
desorption for clay fraction
Oxalate desorbed (mmol kg )


18
Co., Reno, Nevada) with a Gilson single piston high pressure pump along with a
Pheodyne model 7125 injection valve fitted with a 20 ¡uL injection loop. The
HPLC system used a Gilson Holochrom variable wavelength UV detector in
conjunction with a Gilson computerized integrator. The eluent was 0.005 M
H2S04 at a flow rate of 2 mL min"1 Oxalate concentration was calculated from
the calibration curve obtained with standard solutions of 0.1 to 10 mM.
Titration Curves for the Clay Fractions and the Whole Soil Material
To determine the amount of OH released into solution during the sorption
of oxalate, titration curves were prepared (Duquette and Hendershot, 1993) for
each of the clay and whole-soil samples (i. e., with and without organic matter).
Duplicate one-gram samples of clay and 10-gram samples of soil were equilibrated
with 100 mL of water adjusted to pH 3.0. The suspension was allowed to
equilibrate for 24 hrs. Titration was carried out with 0.1 N NaOH from pH 3.0
to 7.0. With each addition of 0.1 ml NaOH, the solution was allowed to
equilibrate for 2 to 5 minutes or until pH stabilized. The titration curves, for
each type of clay and soil are presented in Appendix A-l and Appendix A-3. pH
buffer curves were calculated which relate the change in pH with unit additions of
NaOH for three different ranges of pH These ranges were 3.44 to 4.44, 4.44 to
5.65, and 5.45 to 6.75. Models were fit to these curves and used to calculate the
change in OH concentration required to change pH of the system to specific values
(Appendix A-2 and Appendix A-4).


CHAPTER 5
CONCLUSIONS
Following are the major conclusions from this research, its significance with
respect to the P nutrition of plants, and an outline of priorities for future studies.
Conclusions from this Study
Oxalate Sorption
1. Oxalate was sorbed by the clay fractions and whole-soil samples through
ligand exchange.
2. Sample pH determined the charge characteristics of clay surfaces and the
species of oxalate in solution, which in turn controled the sorption process.
Oxalate formed monodentate and/or binuclear complexes at pH 3.5, while
bidentate surface complexes were formed at pH 4.5 and 5.5.
3. Soil organic matter significantly reduced the amount of oxalate sorbed by
the mineral surfaces.
4. Data from this study provided experimental evidence for the theory of
congruent dissolution as proposed by Stumm (1986). Sorption is the first
step in the dissolution of mineral surfaces. Through sorption, oxalate forms
108


21
lines. The ANOVA and t tests were used to test the differences between a, and
a2 as well as between p, and P2. Significance of the difference between means
was determined by the t-test (Snedecor and Cochran, 1980). Standard deviations
or standard errors of means are given as well, when appropriate.
Results
Characterization of the Clay Fraction using FTIR
FTIR spectra of clay samples with and without organic matter are presented
in Figure 2.1. Characteristic infrared bands for quartz, kaolinite, HIV and gibbsite
minerals are identified. The broad OH-stretching band in the region 3000 to 3600
cm'1, along with well defined OH-stretching is characteristic of kaolinite, HIV and
gibbsite, while the broad deformation band for water in the 1640 cm'1 region
suggests the presence of amorphous clay. The spectra of clay with organic
matter shows major absorption bands as well in the regions of 3300 cm"1 (O-H
and N-H stretching), 2960/2925 cm'1 (aliphatic C-H stretching of CH3 and -CH2
groups), 1610 cm'1 (aromatic C=C and/or H-bonded C=0 stretching of COOH),
and 1400 (OH deformation, C-0 stretching of phenolic OH, and C-H deformations
of CH3 and CH2) together with the characteristic infrared bands for quartz,
kaolinite, HIV and gibbsite minerals. Intensity of the broad OH-stretching band in
the region 2700 to 3600 cm'1 for clay with organic matter was greater, reflecting
the presence of OH-stretching associated with organic matter.


A-l Titration curves for the clay fractions: A) with organic matter,
B) without organic matter 114
A-2 Buffer curve for the clay fractions, used to calculate the
amount of OH ions released into solution: A) with organic matter,
B) without organic matter 115
A-3 Titration curve for the whole-soil samples: A) with organic matter,
B) without organic matter 116
A-4 Buffer curve for the whole-soil samples, used to calculate the
amount of OH ions released into solution: A) with organic matter,
B) without organic matter 117
C-l P desorption curves by dilution and extraction methods
at both variable and constant pH 130
C-2 Relationship between Kd and the equilibrium concentration
of P as obtained by different methods 133
x


68
parameters for the Langmuir model are presented in Table 3.1. In the case of the
whole-soil samples, P sorption maxima were significantly reduced by soil organic
matter. The presence of soil organic matter and oxalate significantly reduced P
sorption by both types of soil material.
P sorption parameters of the clay samples also were recalculated on a
whole-soil basis (assuming a clay % for the whole soil = 1.3 %), and compared to
the P sorption parameters of whole-soil samples. Significantly more P was sorbed
by the whole-soil samples than by the clay samples when reported on a whole-soil
basis (Table 3.2). P sorption by whole-soil samples was 3 to 10 times higher than
that of the clay samples.
Sorption of Phosphate in the Presence of Oxalate
Less P was sorbed by the clay fractions and whole-soil samples in the
presence of oxalate (Table 3.3). The percent efficiency of oxalate in reducing P
sorption was calculated according to the expression of Deb and Datta (1967):
Ox =[l~]*100 (3 4)
* P
o
where Oxe is the efficiency of oxalate in reducing P sorption (%), Pox is the P
sorbed in the presence of oxalate, and P0 is the P sorbed alone. The efficiency of
oxalate in reducing P sorption increased as the concentration of P in solution
increased (Table 3.4).


19
FTIR Analysis
FTIR spectra were obtained by placing small amounts of clay suspension on
AgCl windows and allowing the deposit to dry. Spectra then were collected on a
Bomem DA3.10 spectrophotometer equipped with a MCT detector and a KBr
beam splitter, operating at 2.0 cm'1 resolution. The Bomem DA3.10
spectrophotometer was controlled through a general-purpose interface bus (IEEE-
488) interfaced to a DEC Vaxstation-II computer.
Adsorption Isotherms
The Langmuir adsorption model was fitted to all sorption data. The
following equation was used:
S=
S KC
m
1+ KC
(2.2)
Where S is the amount of oxalate taken up per unit mass of soil (mM kg"1), Sm
is the maximum amount of oxalate that was bound, C is the equilibrium
concentration of oxalate (mM), and K is a constant related to the binding energy
for oxalate. The parameters (K and Sm) were calculated by a least squares fit to
a linear form of the equation:
KS.
M
(2.3)


77
Table 3 6 Ratios of OH' released to P sorbed for P only and P + oxalate for
the whole-soil samples and the clay fractions (Mean SD, n=3).
OH released/Anion
sorbed
W/OM or WO/OM
Treatment
Whole soil
Clay fraction
With OM
P
P + Ox
*1.030.05a
1610.10
0.920.03a
1 380.08
Without OM
P
P + Ox
0.830.05a
1 530.08
0 850.07a
1240.11
Superscript "a" indicates that values are not signficantly different from 1.0 at the
5% level.


OH released (mmol kg )
74
P or oxalate sorbed (mmol kg1)
Fig. 3.4
Release of OH' ions during sorption of P by the whole-soil samples.


126
Materials and Methods
Soil Material
Soil was collected from the spodic horizon (Bh) in a single soil pit of a
Pomona series (sandy, siliceous, hyperthermic Ultic Alaquod) at the Gator
Nationals Forest site located in Alachua County, 10 km northeast of Gainesville,
Florida (Swindel et al. 1988). The material was air dried, passed through a 2-mm
sieve, and stored in plastic bags.
Preparation of Phosphorus Enriched Soil Material
Soil was enriched with 1000 ppm solution of P as Ca(H2P04) and incubated
for 7 days at room temperature. At the end of this reaction period, samples
were centrifuged, the soil samples were oven dried to constant weight at 110 C,
and the supernatant was analyzed for inorganic P. The difference between the
amount of P added in the beginning and the amount of P in solution at the end
of the reaction period was taken as the amount of P sorbed by the soil.
Desorption by a Dilution Method
This experiment was conducted both under constant and variable pH
conditions. For one set of samples pH was kept constant, while for the other pH
was allowed to change. For each set, a varying amount of a 0.01M calcium
nitrate solution was added to three replicates of 5 gram soil samples. The soil to
solution ratios used were 1:10, 1:20, 1:30, 1:40, 1:50, 1:75, and 1:100. Two


CHAPTER 4
INFLUENCE OF SOIL ORGANIC MATTER ON DESORPTION OF
PHOSPHORUS AND OXALATE FROM A SPODIC HORIZON
Introduction
In Chapters 2 and 3, the sorption of oxalate and P by the clay fractions
and the whole-soil material of a spodic horizon from northcentral Florida were
studied. Phosphate and oxalate are specifically sorbed by replacing the coordinated
-OH groups of A1 present on the surfaces of oxides and clay minerals. As
discussed in Chapter 3, oxalate reduced P sorption onto clay fraction and whole-
soil surfaces. This investigation studied P and oxalate desorption from the clay
fraction and the whole-soil material. Phosphorus sorbed to the soil surfaces can be
taken up by plants after it has been desorbed into the soil solution.
The competitive sorption of organic anions and P on mineral surfaces has
received attention because the presence of organic ligands in the rhizosphere is
thought to influence P fixation and, therefore, the supply of P to the plant (Deb
and Datta, 1967; Lopaz (1974); Appelt et al., 1975; Earl et al.. 1979, Lopez-
Hernandez et al., 1986; Kafkafi et al.. 1988, Martell et al.. 1988; Fox et al..
1990a: Fox et al.. 1990b; Violante et al.. 1991; Violante and Gianfreda, 1993).
Oxalate in forest soils originates from exudates of plant roots and through the
85


102
2.5
2 T
1.5
1
0.5
0
Fig. 4.5.
Influence of previously sorbed oxalate on P sorption and oxalate
desorption by whole-soil samples.
Oxalate desorbed (mmol kg )


3.3. Amount of Al Fe and organic carbon released
into solution with P + oxalate sorption onto
the whole-soil samples and the clay fractions 70
3.4 Effect of oxalate in reducing the sorption of P onto
whole-soil samplesand the clay fractions 71
3.5. Parameters of the linear regression models for
the release of OH' ions and organic carbon from soil and clay. .75
3.6 Ratios of OH' ions released to P sorbed for P
only and P + oxalate for the whole-soil
samples and the clay fractions 77
3.7 E4/E6 ratio of organic carbon released into
solution during the sorption of P only, and
P + oxalate by the clay fractions 79
3.8 Parameters of the linear regression models relating release of
aluminum and iron with P sorption for P only and
P + oxalate by the whole-soil and clay fractions 80
4.1 Parameters of Freundlich models for the desorption of
P from the whole-soil samples and clay fractions in the
presence of 5 mM oxalate 94
4.2 Quadratic equation coefficients for the sorption of anions
by the whole-soil samples and the clay fractions 97
4.3 Influence of oxalate sorption by clay fractions and whole-soil samples
on the ratio of OH ions released, and OH + P
released to oxalate sorbed 98
4.4. Influence of P sorption by clay fractions and whole-soil samples
on the ratio of OH ions released, and OH + oxalate
released to P sorbed 103
B-l Percent reduction in oxalate sorption for the
whole-soil samples and clay fractions 121
C-l Freundlich isotherm parameters for the desorption of P from
soil using the dilution and sequential extraction methods
at variable and constant pH 131
vi


57
surfaces, and iii) control the competition between OH', P and organic anions for
common adsorption sites (Barrow, 1987; Kafkafi et al.. 1988).
Poorly-drained Spodosols are the dominant soil type in the flatwoods of
Florida's lower Coastal Plain. Among the low-molecular weight organic anions
present in these soils, the most abundant is oxalate (Fox and Comerford, 1990).
Earlier work by various researchers indicated that the specific sorption of phosphate
and organic ligands like oxalate is through ligand exchange (Goldberg and Sposito,
1985). This process should result in the release of OH'. The amount of OH'
released during a sorption reaction in turn depends upon the characteristics of the
surfaces, concentrations of the adsorbing species, and solution pH. The change in
solution concentration of OH' ions can be used to identify the types of complexes
formed on the various solid surfaces.
Organo-mineral complexes have an important influence on the physical and
chemical properties, and reactivity of soil particles (Viotante and Huang, 1984,
Viotante and Huang, 1985; Huang and Schnitzer, 1986). The main clay minerals
present in flatwoods soils are quartz, kaolinite, gibbsite and hydroxy-interlayered
vermiculite (HIV) (Harris et al.. 1987a). Since Florida soils are sandy, these
minerals are present as coatings on the sand particles (Harris et al.. 1987a; 1987b)
The binding materials for these clay-sized mineral particles to sand grains is
reported to be an Al dominated gel-like substance (Lee et al.. 1988a). Lee et al.
(1988b) further indicated that Al acts as a cementing material, probably as Al-
fulvate. The Spodosols of the lower Coastal Plain are known to have high


127
drops of toluene were added to inhibit microbial growth, and samples were placed
in a reciprocating shaker for 24 hrs. For variable pH conditions, pH was
measured at the begining and at the end of the reaction period. For constant pH
conditions, pH of the suspension was periodically (every six hours) adjusted with
0 1 M HC1 or NaOH over the entire period of time, to the initial pH value. At
the end of the reaction period, each suspension was centrifuged at 12,000 g for 20
minutes. The supernatant solutions were analyzed for inorganic P and aluminum.
Desorption by a Sequential Extraction Method
Five grams of soil were repeatedly desorbed by 50 ml of 0.01 M calcium
nitrate added in successive 24 hr periods for 8 days. The other experimental
conditions were the same as described above, for the dilution method. The P
concentrations in successive volumes were measured, and the total P desorbed from
the soil was calculated by summation.
Chemical Analysis
Solution pH was measured using a combination glass electrode in conjuction
with an Orion pH meter Inorganic P in the filtrate was determined by a
molybdenum-blue colorimetric procedure, using ascorbic acid as a reductant
(Murphy and Riley, 1962)


64
(3-3)
where a is the intercept, P is the slope of the line and e is the error term. To
compare two regression lines, the a and P values of the two lines were
compared. ANOVA and t-test were used to test the difference between a, and
a2 well as between p, and P2. The significance of the difference between two
means was determined using a t-test (Snedecor and Cochran, 1980). Standard
deviations or standard errors of means are given, when appropriate.
Results
Sorption of P
Clay Fractions
Sorption isotherms for phosphate sorption onto clay samples with and
without organic matter are presented in Fig. 3.1. The presence of oxalate and soil
organic matter significantly reduced P sorption. Maximum reduction in P sorption
(about 50 %) was observed when both organic carbon and oxalate were present in
the system. The P sorption data, for the range of concentrations studied,
conformed to the Langmuir equation. Sorption parameters for the Langmuir
equation are presented in Table 3.1.
Whole-Soil Samples
Phosphate sorption isotherms are presented in Figure 3.2. Sorption


56
1986; Comerford and Skinner, 1989) or on mineral surfaces (Pohlman and McColl,
1986; Stumm, 1986; Kafkafi et al.. 1988; Martell et al.. 1988; Fox et al.. 1990b)
can be effective in reducing the P sorption capacity of soils. Phosphate sorption
was reduced in the presence of humic acids, fulvic acids, and low molecular
organic acids while each of these compounds were specifically sorbed onto pure
mineral surfaces (Nagarajah et al., 1968, 1970; Sibanda and Young, 1986; Ryden
and Syers 1987; Moore et al.. 1992; and Violante and Gianfreda, 1993).
However, little work has documented the effect of oxalate and soil organic matter
on the sorption of P using soil materials
The ability of different organic anions to compete with P for sorption sites
on the surfaces of soil components was reported to be greatest at a pH equivalent
to the pKa2 of the organic acid (Hingston et al.. 1967; Hingston et al.. 1972).
Soil organic acids which contain carboxylic (-COOH) and/or phenolic (-OH)
functional groups can bind to oxide surfaces, thereby reducing the number of
surface sites available for P sorption (Yuan, 1980). This also alters electrostatic
charge at the solid surface. Both of these interactions of organic anions are
influenced by the solution pH, relative concentrations of different anions which may
be present and intrinsic affinities of these anions for the mineral surfaces The
pH of the system is important in surface and solution complexation reactions as it
will i) regulate the concentrations of various P and organic anion species which
differ in their affinity for the solid surface, ii) affect the charge density of solid


81
P Sorption
Phosphate can be sorbed by the clay and whole-soil surfaces either by
ligand exchange or precipitation. Ligand exchange should result in the release of
OH' ions into solution (Goldberg and Sposito, 1985). In fact, large amounts of
OH' were released in this study The amount of OH' ions released varied with
the amount of P and/or of oxalate sorbed The average ratio of OH released to
phosphate sorbed was close to unity. When oxalate was also sorbed this ratio
increased to 1.5. Given that the ratio was unity and that the dominant species of
phosphate was the H2P04' ion one could argue that P was forming a
monodendate and/or binuclear inner-sphere complex. The ratio of OH' ions released
to P sorbed was similar for both the whole-soil and clay samples. Recently
Tejedor-Tejedor and Anderson (1990) using CIR-FTIR studied the surface
complexation of phosphate by goethite surfaces between pH 3.5 to 8.0. They
suggested that between pH 3.5 and 5.5, phosphate formed both binuclear and
monodendate complexes with surface Fe (III) Earlier work by Parfitt et al.
(1977), using infrared spectroscopy, also reported the presence of both binuclear
and monodendate surface complexes by P on goethite at pH 4.0.
The molar ratio of OH ions released to P sorbed supports ligand exchange
as the dominant P sorption process onto both the clay and whole-soil samples.
The formation of measureable crystalline aluminum phosphate was not possible,
given the reaction time of this experiment. Veith and Sposito (1977) and Sposito
(1984) reported that reaction times longer than 140 hrs are required for significant


69
Table 3.2. P sorption onto the whole soil material compared to P sorbed onto
the clay fraction expressed on a whole soil basis (Mean SD; n=3).
Cone of
anions
P sorption (mmol kg'1)
added
+ OM
- OM
Treatment
(mM)
Whole soil
Clay Fraction
Whole soil
Clay fraction
P alone
10
16.7+0.27
4.580.05b
24.6+0.5
7.090.07b
1.0
5.9+0.13a
0.940.01b
7.1+0.06
1.170.01b
0.1
0.9+0.02
0.120.01b
0.8+0.01
0.120.01b
P + Ox
10+1
13.3+0.25
3.520.04b
19.3+0.37
5.950.04b
1.0+1
5.3+0.14
0.79+0.01b
5.8+0.05
0.960.01b
0.1+1
0.78+0.02
0.09+0.01b
0.8+0.01
0.08+0.001'
+ OM = with organic matter and OM is without organic matter
@ Values within each row for each material with organic matter and without
organic matter are significantly different (P = 0.05) if followed by a different letter


Al release (mM kg )
46
Fig. 2.16
Relationship between A1 release and organic carbon release by the
clay fraction at varing pH.


62
Titration Curves for the Clay and Soil Samples
To determine the amount of OH' ions released into solution during the
sorption of P and/or oxalate, titration curves were prepared for each clay and
whole-soil sample (with and without organic matter). Duplicate one-gram samples
of clay and 10-gram samples of soil were equilibrated with 100 mL of water
adjusted to pH 3.0. The suspension was allowed to equilibrate for 24 hrs,
titration then being carried out (with 0.1 N NaOH) from pH 3.0 to 7.0. With
each addition of 0.1 ml NaOH, the solution was allowed to equilibrate for 2 to 5
minutes or until the pH stabilized. pH buffer curves, relating the change in pH
with unit addition of NaOH over the pH range 4.40 to 6.00, were calculated
Curves were fit to the data and used to calculated the amount of OH' ions
required to change pH of the soil and clay systems to specific values.
Adsorption Isotherms
The Langmuir adsorption model was fit to the sorption data The following
equation was used:
5 KC
m
1 + KC
(3.1)
where S is the amount of P or oxalate taken up per unit mass of clay or soil
(mM kg'1), Sm is the maximum amount of P or oxalate that was bound, C is the
equilibrium concentration of P/oxalate (mM), and K is a constant related to the


80
Table 3.8 Parameters of the linear regression model relating release of aluminum
and iron with P sorption for P only and P + oxalate by the whole-soil
and clay fractions.
Treatments
A1 release
Fe release
Slope
Constant
R2
Slope
Constant
R2
Soil
with organic
matter
P
0.08a
2.05a
0.93
o.oor
0.015a
0.87
P + OX
0.01a
7.66b
0.07
0.0004a
0.031a
0.24
Soil without organic matter
P
0.01a
2.52a
0.69
0.0001a
-0.008a
0.31
P + OX
-0.09a
6.72b
-0.94
-0.002a
0.106a
-0.84
Clay without organic matter
p
0.03a
34.27a
0.94
0.0003a
-0.01a
0.97
P + ox
0.05a
69.01c
0.94
0.0005a
-0.21a
0.96
Clay With Organic Matter
p
0.02a
13 20a
0.96
0.001a
0.19a
0.98
P + ox
-0.01a
54.01b
-0.61
0.001a
0.52b
0.97
Values within columns followed by the same letter for soil or clay samples are not
significantly different at the 5 % level


48
concentrations of oxalate. The integrated absorption intensities was normalized with
respect to the weight of the clay sample deposited on the AgCl window. Intensity
of the 1610 cm'1 (aromatic C=C and/or H-bonded C=0 stretching of COOH and
CH3) bands decreased as progressively more oxalate was sorbed by the clay
surfaces (Table 2.6, Fig. 2.4, 2.5. and 2.6). The greatest decrease in the intensity
of bands at 1610 cm'1 (aromatic C=C and/or H-bonded C=0 stretching of
COOH) and 1460 cm'1 (OH deformation, C-0 stretching of phenolic OH, and C-H
deformations of CH3 and CH2) corresponded with the maximum oxalate sorption.
Decrease in the intensity of these bands was higher at pH 5.5, again
corresponding to the pattern of organic carbon release seen in the batch studies.
Release of organic carbon was significantly correlated to the intensity of these
bands, with R2 values of 0.82 to 0.96.
Discussion
Surface Characteristics
Mineralogical analysis using XRD and FTIR data showed that the dominant
crystalline clay minerals in the spodic horizon of this Spodosol were kaolinite, HIV
and gibbsite. FTIR data established that the clay fraction also contained large
amounts of noncrystalline A1 oxide. Fox et al.. (1990a) reported high amounts of
oxalate-extractable (1357 mg/kg) and pyrophosphate-extractable (1379 mg/kg) Al for
the same soil. Therefore, both crystalline and amorphous mineral surfaces were
available for the sorption of oxalate.


OH-ions released (mmol kg )
32
Release of OH ions by the clay fractions at varing pH as a function
of oxalate sorption.


140
Jackson, M. L 1979. Soil Chemical Analysis Advanced Course Madison,
WI.
Jardine, P. M N.L. Weber, and J F. McCarthy. 1989. Mechanism of dissolved
organic carbon adsorption on soil. Soil Sci. Soc. Am. J. 53: 1378-1385.
Jersak, J. M and J. G. McColl 1989. Aluminum release from solid-phase
components of forest soil leached with citric acid Soil Sci. Soc. Am. J.
53:550-555.
Kafkafi, U., B Bar-Yosef, R Rosenberg, and G Sposito. 1988 Phosphorus
adsorption by kaolinite and montmorillonite:II. Organic anion competition.
Soil Sci. Soc. Am. J. 52:1585-1589.
Kunze, G. W. and J B. Dixon. 1986. Pretreatment for mineralogical analysis, p
91-100 In Page, A L., R H Miller, and D R Keeney (ed.) 1982.
Methods of Soil Analysis. Part 2. 2nd ed. Agronomy 9. Am. Soc.
Agron & Soil Sci Soc Am Madison, WI
Kuo, E J and W L Pan 1988 Influence of phosphate sorption parameters of
soils on the desorption of phosphate by various extractants. Soil Sci. Soc.
Am. J 52:974-979.
Lan, M N. B Comerford and T R Fox. 1994 Effect of organic anions of
varying complexation power on the release of phosphorus from 10 spodic
horizons. (Submitted to Soil Sci. Soc. Am. J ).
Lee, F. Y., T. L. Yuan, and V. W Carlisle. 1988a Nature of cementing
material in ortstein horizons of selected Florida Spodosols: I. Constituents
of cementing materials. Soil Sci. Soc. Am. J. 52:1411-1418.
Lee, F. Y., T. L. Yuan, and V. W Carlisle. 1988b, Nature of cementing
matreial in ortstein horizons of selected Florida spodosols: II. Soil
properties and chenical forms of aluminum. Soil Sci. Soc. Am. J. 52:1766-
1801.
Logan, T. J. 1982. Mechanism of release of sediment-bound phosphate to water
and the effects of agriculture land management on fluvial transport of
particulate and dissolved phosphate. Hydrobiologia 92:519-530.
Lopez-Herandez, D 1974. Phosphate desorption isotherms in four selected
tropical soils and one temperate soil. Commun. Soil Sci. Plant Anal
5:145-154.


40
Table 2.4 Parameters of the linear regression models relating release of aluminum
and iron with oxalate sorption for the clay fractions and whole-soil
samples.
Material
Al release
Fe release
OM/pH
Slope
Intercept
R2
Slope
Intercept
R2
Clav Fractions
Without organic
matter
3.5
0.57a
40.62a
0.99
0.007a
-0.21a
0.99
4.5
0.59a
8 62a
0.99
0.009b
-0.22a
0.99
5.5
0.74a
-1.03a
0.99
0.012
-0.15a
0.99
With Organic Matter
3.5
1.83b
7.11a
0.97
0.008a
-0.19a
0.98
4.5
2.77
-15.30a
0.99
0.019d
-0.27a
0.97
5.5
4.27d
-28 66a
0.98
0.036
-0.13a
0.98
With organic matter
2.96a
-12.27a
0.93
0.021a
-0.19a
0.73
Without organic
matter
0.63b
16.07b
0.98
0.010b
-0.19a
0.96
Whole-Soil Samples
With organic matter
4.5
0.76b
2.16a
0.96
0.012b
0.025b
0.96
Without organic
4.5
matter
3.07a
-2.75
0.95
0.042a
-0.061a
0.89
@ Includes all samples at pH 3.5, 4.5, and 5.5;
The same letter in a column at different pH values indicates a lack of significance
at the 5 % level within the clay fractions and the whole-soil samples.