TURFGRASS ROOT-GROWTH RESPONSE TO SUBSOIL ACIDITY AND

AMELIORATION OF ACID SUBSOILS WITH GYPSUM IN ULTISOLS

by

JOHN SAYLER KRUSE

(Under the Direction of William P. Miller)

ABSTRACT

Highly weathered Ultisols typically have subsoils with little exchangeable

calcium and high exchangeable acidity, often restricting root growth into depths that offer

a potential reservoir of water. Research has not quantified the extent to which turfgrass

root growth will respond to acid subsoils ameliorated by gypsum. Hydroponic, column,

and field studies were utilized to determine gypsum efficacy.

Tall fescue was grown in solutions containing levels of Al and/or CaSO4·2H2O at

pH 4.5 for 7 d. Root growth was significantly reduced in low Al concentrations absent

Ca. The addition of CaSO4·2H2O significantly increased root growth in the presence of

Al. A simple logistic model adequately explained the relationship between Al and

CaSO4·2H2O and relative root growth. Mechanistically, a computer-aided predictor of Al

species (VMINTEQ) demonstrated that the addition of CaSO4·2H2O in Al solution

reduced Al activity and the quantity of Al3+ as a percentage of total Al, and increased the

percentage of relatively non-toxic AlSO4+ and solution ionic strength. A previously

published calcium-aluminum balance equation did not adequately predict root growth or

root-aluminum concentrations.

Tall fescue was grown 85 d in columns filled with acid subsoils treated with lime

or gypsum. After an initial period, the tall fescue was physiologically stressed by a

column dry-down procedure. At experiment termination, soil columns were sectioned for

chemical analysis and root density determination. Lime and gypsum ameliorated soils

significantly increased exchangeable Ca and reduced exchangeable acidity. Lime and

gypsum treated soils significantly increased root density over control treatments, with

lime-treated soil producing a higher root-growth response than gypsum. Column

evapotranspiration did not differ significantly between treatments.

A field study utilizing lime, gypsum, and combinations of the two was conducted

growing bermudagrass and zoysiagrass on an acid Cecil soil. Soil moisture was measured

by time domain reflectometry. After 15 months, soils were analyzed by depth for

chemical properties and root density. Differences in root density due to the addition of

lime and gypsum were significant (α=0.10) in zoysiagrass but not bermudagrass. Despite

a wet growing season, slight differences in volumetric water content were detected for

zoysiagrass between 20-58 cm, but were nonexistent for bermudagrass.

INDEX WORDS: Subsoil acidity, gypsum, amelioration, lime, turfgrass,

bermudagrass, zoysiagrass, tall fescue, exchangeable aluminum, calcium sulfate, CaSO4·2H2O

TURFGRASS ROOT-GROWTH RESPONSE TO SUBSOIL ACIDITY AND

AMELIORATION OF ACID SUBSOILS WITH GYPSUM IN ULTISOLS

by

JOHN SAYLER KRUSE

B.S.A., The University of Georgia, 1986

M.S., The University of Georgia, 2002

A Dissertation Submitted to the Graduate Faculty of The University of Georgia in Partial

Fulfillment of the Requirements for the Degree

DOCTOR OF PHILOSOPHY

ATHENS, GEORGIA

2005

© 2005

John Sayler Kruse

All Rights Reserved

TURFGRASS ROOT-GROWTH RESPONSE TO SUBSOIL ACIDITY AND

AMELIORATION OF ACID SUBSOILS WITH GYPSUM IN ULTISOLS

by

JOHN SAYLER KRUSE

Major Professor: William Miller

Committee: Miguel Cabrera Ronald Hendrick David Kissel Malcolm Sumner Clint Waltz

Electronic Version Approved: Maureen Grasso Dean of the Graduate School The University of Georgia December 2005

DEDICATION

This work is dedicated to my wife, Lyn, and my children, Johnny, Katie, Noah,

Philip, and Annie. Your sacrifice and patience through this challenging time was

extraordinary. This experience has deeply shaped all of our lives, and it is my hope that

blessings will flow from it. I love you and deeply appreciate all that you have done to

make this possible. I also dedicate this work to my Lord and Savior Jesus Christ. It has

been a challenge and a joy to discover God’s mind after Him.

iv

ACKNOWLEDGEMENTS

I wish to express my genuine gratitude to Dr. Bill Miller, who has provided me

with great support throughout this work and encouraged me to pursue ideas and methods

that led to the eventual success of this endeavor. His patience and understanding of my

distinct situation is a debt I could not repay and I am very grateful for his countless hours

of support and encouragement. His willingness to help even with the backbreaking

physical labor involved in this project shows a unique dedication to his high calling as a

teacher, advisor, and friend.

I would also like to extend my gratitude to the members who served on my

committee and who took the time to advise me throughout my research and the review of

this document; including Dr. Miguel Cabrera, who has been a great mentor, Dr. Ron

Hendrick, Dr. David Kissel, and Dr. Clint Waltz. A special debt of thanks is owed to Dr.

Malcolm Sumner who agreed to participate on my committee in his retirement, and

whose ideas helped steer the direction of this research.

I am grateful to Lamar Larrimore and the research staff of The Southern

Company for the financial support of this work, and it is my hope that the information

gleaned from this work will prove useful. I also wish to express appreciation to the

members of the Department of Crop and Soil Sciences, who showed me great kindness

and encouragement. A special note of thanks goes to Greg Pillar, who was a great friend

as we took classes and progressed through our dissertation journeys together.

v

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS.................................................................................................v

LIST OF TABLES............................................................................................................. ix

LIST OF FIGURES ........................................................................................................... xi

CHAPTER

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

References Cited............................................................................................5

2 LITERATURE REVIEW ..................................................................................8

Subsoil acidity ...............................................................................................9

Amelioration techniques..............................................................................13

Gypsum effects on agronomic crops ...........................................................23

Gypsum and subsoil acidity effects on grasses ...........................................27

Sources of gypsum ......................................................................................32

Objectives....................................................................................................33

References Cited..........................................................................................34

3 A SIMPLE LOGISTIC MODEL TO DESCRIBE ROOT-GROWTH

EFFECTS OF CALCIUM AS GYPSUM ON TALL FESCUE

VARIETIES IN ALUMINUM-RICH SOLUTIONS..................................49

Abstract .......................................................................................................50

Introduction .................................................................................................51

vi

Materials and Methods ................................................................................53

Results and Discussion................................................................................57

Conclusions .................................................................................................62

References Cited..........................................................................................64

4 MECHANISTIC ANALYSIS OF TALL FESCUE ROOT GROWTH IN

ALUMINUM SOLUTIONS AFFECTED BY ADDITION OF

CaSO4·2H2O................................................................................................78

Abstract .......................................................................................................79

Introduction .................................................................................................80

Materials and Methods ................................................................................82

Results and Discussion................................................................................85

Conclusions .................................................................................................91

References Cited..........................................................................................93

5 ROOT-GROWTH AND WATER USE RESPONSE OF TALL FESCUE TO

GYPSUM AND LIME-AMENDED ACID SUBSOILS..........................107

Abstract .....................................................................................................108

Introduction ...............................................................................................109

Materials and Methods ..............................................................................110

Results and Discussion..............................................................................114

Conclusions ...............................................................................................122

References Cited........................................................................................123

6 WARM-SEASON TURFGRASS RESPONSE TO AMELIORATION OF

SUBSOIL ACIDITY IN THE FIELD USING LIME AND GYPSUM ...140

vii

Abstract .....................................................................................................141

Introduction ...............................................................................................142

Materials and Methods ..............................................................................144

Results and Discussion..............................................................................148

Conclusions ...............................................................................................153

References Cited........................................................................................155

7 SUMMARY AND CONCLUSIONS ............................................................165

APPENDICES .................................................................................................................169

Appendix A. Griffin field study list of maintenance materials applied to plots ...169

viii

LIST OF TABLES

Page

Table 3.1: Definitions of symbols in the modified simple logistic model used to

characterize tall fescue root growth in response to Al and CaSO4 in solution,

and predict CaSO4 levels necessary to overcome Al toxicity ...........................69

Table 3.2: Model fit statistics for simple logistic model per variety .................................70

Table 3.3: Predicted optimal levels and 80% of fescue root growth (mg), and CaSO4

estimates (µM) for 80% optimal root growth by level of aluminum and tall

fescue variety.....................................................................................................71

Table 4.1: Root mass of tall fescue varieties in the presence of zero total Al in solution .97

Table 4.2: Al speciation in various levels of Ca as dissolved gypsum as predicted by

Visual MINTEQ2 ..............................................................................................98

Table 4.3: Linear regression goodness of fit for tall fescue varieties modeling root growth

as a function of the CAB equation ....................................................................99

Table 5.1: Initial soil physical and chemical characteristics............................................127

Table 5.2: Soil column Ca, Mg, and K after conclusion of experiment by depth and

treatment. Tukey’s mean separation of significant treatment differences

(α=0.05) designated by letters for each depth .................................................128

Table 5.3: Mean root density values (mg kg-1) for treatment and soil types. Letter

designations indicate significant differences between treatments or soils at each

depth based on Tukey’s mean separation (α=0.05) for each depth.................129

ix

Table 6.1: Soil chemical conditions 15 months after treatment application....................159

Table 6.2: Exchangeable acidity values as a percent of CEC by depth...........................160

Table 6.3: Root density values by treatment and depth. Letters signify significant

differences (α=0.10) between treatments for each depth ................................161

Table 6.4: Volumetric water content (expressed as %) of treatment plots by date and

turfgrass type at various depths. Mean comparisons using Fisher’s lsd (α=0.10)

are between treatments at each depth. Columns 7 and 8 were sampled after

dormancy .........................................................................................................162

x

LIST OF FIGURES

Page

Figure 3.1: The simple logistic model showing the parameters Φ1, the horizontal

asymptote as x → ∞; Φ2, the value of x for which y = Φ1*0.5; and Φ3, the value

of x for which y = Φ1*0.8 ..................................................................................72

Figure 3.2: Comparison of calculated root length to root mass of tall fescue seedlings

grown in hydroponic solution containing various levels of CaSO4 and Al.......73

Figure 3.3: Root scan for digital analysis for root surface area using ARC INFO (left),

and fescue root growth in 0 Al and 10,000 μmol CaSO4 (right) showing

uniformity of root diameters in solution............................................................74

Figure 3.4: Relative tall fescue root growth sensitivity to total Al in an acid (pH 4.5)

hydroponic solution. Error bars indicate standard deviation of mean values ...75

Figure 3.5: Effect of increasing CaSO4 in solution on relative root growth in the presence

of Al. Error bars indicate standard deviation of mean values ...........................76

Figure 3.6: An example of fitting the simple logistic model: K31 root response to

increasing levels of CaSO4 at various fixed levels of Al. Error bars indicate

standard deviation of mean values ....................................................................77

Figure 4.1: Tall fescue relative root growth sensitivity to Al in solution (pH 4.5). Error

bars indicate standard deviation of the mean ..................................................100

Figure 4.2: Tall fescue relative root growth in Al solutions at various levels of CaSO4.

Error bars indicate standard deviation of the mean.........................................101

xi

Figure 4.3: Effect of increasing CaSO4 concentration (in mM) on ionic strength (IS);

activity coefficient of Al3+; complexation of Al by SO4 (expressed as %,

averaged over Altot = 3-70 μM); and activity of Al3+ in a solution of Altot = 100

μM containing levels of CaSO4 as shown on abscissa (x-axis). All parameters

generated using VMINTEQ (Davies equation, default parameters) ...............102

Figure 4.4: Tall fescue relative root growth in response to total Al activity in solution

with nonlinear regression model equation and goodness of fit. Error bars

indicate standard deviation of the mean ..........................................................103

Figure 4.5: CAB analysis: Effect of increasing calcium-aluminum balance (CAB) on the

relative root growth of tall fescue cultivars in Ca/Al solutions. Error bars

indicate standard deviation of the mean ..........................................................104

Figure 4.6: Effect of increasing calcium-aluminum balance on the concentration of Al in

tall fescue roots, including linear regression analysis. Error bars indicate

standard deviation of the mean........................................................................105

Figure 4.7: Effect of increasing total Al activity on root Al concentrations in tall fescue

varieties grown in various CaSO4/Al solutions, including nonlinear regression

analysis. Error bars indicate standard deviation of the mean ..........................106

Figure 5.1: Profile of column compartment for maintaining 15o soil temperature..........130

Figure 5.2: Electrical conductivity (EC) of soils (2.5:1 water: soil) from Cecil and

Cunningham soils as a function treatment at conclusion of experiment Error

bars indicate standard error of mean values ....................................................131

xii

Figure 5.3: soil pHH2O values for Cecil and Cunningham soils from columns sampled

after conclusion of experiment. Error bars indicate standard error of mean

values...............................................................................................................132

Figure 5.4: Soil pHKCl values from Cecil and Cunningham soils from columns sampled

after conclusion of experiment. Error bars indicate standard error of mean

values...............................................................................................................133

Figure 5.5: Exchangeable acidity of sandy loam and sandy clay soils from columns

sampled after conclusion of experiment. Error bars indicate standard error of

mean values .....................................................................................................134

Figure 5.6: Extractable soil Ca of sandy loam and sandy clay soils from columns sampled

after conclusion of experiment. Error bars indicate standard error of mean

values...............................................................................................................135

Figure 5.7: Soil Ca, Mg, and K leachate concentrations (mg L-1) collected from columns

30 d after planting. Error bars indicate standard error of mean values ...........136

Figure 5.8: Root density values for Cecil and Cunningham soils at conclusion of

experiment. Letter differences indicate significant differences by treatment for

each depth based on Tukey’s mean separation (α=0.05) ................................137

Figure 5.9: Gravimetric water loss from soil columns during drydown procedure. Error

bars indicate standard error of mean values ....................................................138

Figure 5.10: Column water loss on a per day basis by treatment. Error bars indicate

standard error of mean values .........................................................................139

Figure 6.1a: Zoysiagrass: Exchangeable Ca levels for treatments at 0-10, 10-20, and 20-

40 cm depths. Error bars indicate standard error of mean values ...................163

xiii

Figure 6.1b: Bermudagrass: Exchangeable Ca levels for treatments at 0-10, 10-20, and

20-40 cm depths. Error bars indicate standard error of mean values ..............164

xiv

CHAPTER 1

INTRODUCTION

1

Vast areas of arable land are acidic or have acidic subsoils, to the extent that plant

root growth is substantially limited. At least one quarter of the world’s soils are classified

as predominantly acidic, including most of the southeastern United States, tropical

Central and South America, equatorial Africa, Southeast Asia, and northeastern Australia.

Most of these soils are classified in the Ultisol and Oxisol soil orders. Ultisols are

considered to be highly weathered and strongly leached, and were originally forest soils

with relatively low inherent fertility. They usually have a subsurface horizon in which

clays have accumulated that is often acidic. Found primarily in humid temperate and

tropical regions on older, more stable landscapes in which annual rainfall exceeds

evapotranspiration, they occupy about 8.5% of the global ice-free land area and contain

approximately 18% of the earth’s human population (McDaniel, 2005). Oxisols are

described as even more intensively weathered than Ultisols and contain very few primary

minerals. They usually have low base-cation saturation and in their native state have high

phosphorus retention and little cation exchange capacity, but are rich in iron and Al

oxides. Oxisols occupy roughly 7.5% of the earth’s ice-free land area and are found in

tropical regions of the world (McDaniel, 2005).

Acid soils are frequently defined as those soils affected in some part of the profile

by acid conditions leading to Al solubility to the extent that it impacts plant root growth

(Sumner, 1995). These acid soils are often utilized extensively for agricultural

production. Furthermore, as population density increases, so too does the conversion of

land from crop production to urban landscapes, especially turfgrass-dominated

environments. In both agricultural crop production and turfgrass culture, the presence of

acid soils, especially subsoils, can prevent the proliferation of roots lower into the soil

2

profile leaving plants with relatively shallow root systems that result in reduced plant

vigor, yield, and the ability to withstand drought stress (Adams, 1984; Kamprath and

Foy, 1985; Helyar, 1991; Foy, 1992; Liu et al., 1995; Huang et al., 1997a,b; Rout et al.,

2001; Vasconcelos et al., 2002). The primary reason roots are often unable to proliferate

in an acid soil environment, and thus access subsoil water, is due to aluminum (Al)

toxicity on the root cells. This toxicity is due to root absorption of Al, which severely

inhibits root elongation within hours, interferes with cell division, prevents the absorption

of Ca, and causes cell death due to the destruction of the plasma membrane of younger

cells near the root tip (Sasaki et al., 1995; Wagatsuma et al., 1995; Matsumoto, 2000).

There is some evidence that cell death occurs due to intracellular peroxidase-mediated

hydrogen peroxide production (death by free radicals) (Simonovicova et al., 2004). Plants

subjected to acid soil/drought syndrome often require frequent irrigation and are

subjected to secondary pathogenic attack by fungi and insects (Sumner and Hylton, 1994;

Duncan and Carrow, 1997). The result for the grower is increased labor and irrigation

costs, more frequent pesticide and fertilizer applications, lower yields and reduced plant

longevity.

Acid subsoils often contain low levels of exchangeable Ca and high levels of

exchangeable Al. A high percentage of the Al in these low-ionic strength soil solutions is

monomeric, which often has toxic effects on roots by interfering with cell division and

elongation. Techniques employed in acid-soil amelioration usually include agricultural

lime due to its relative abundance, low cost, and ease of application. Limestone is very

effective in topsoil acid-amelioration when it can be incorporated, but because it is

sparingly soluble, does not move readily through the soil profile. Several other techniques

3

and materials have been investigated by researchers for their effectiveness in subsoil

acidity amelioration, including the use of gypsum (CaSO4·2H2O). Examinations of crop

responses to subsoil amelioration have to date been focused on agronomic crops, and

little is known about turfgrass response to gypsum-ameliorated acid subsoils.

Improved management techniques for turfgrass promise to reduce energy use,

conserve water and other resources, and produce a high-quality turfgrass with reduced

environmental impacts by providing more deeply rooted plants that are better able to

withstand stresses. A better understanding of processes regulating root growth in

turfgrasses, particularly with respect to subsoil acidity and its amelioration on root

growth, is an important aspect of developing new management strategies to this end. The

studies described in this dissertation are aimed at that goal.

4

References Cited

Adams, F. 1984. Crop response to lime in the southern USA. p. 211-266. In F. Adams

(Ed.) Soil acidity and liming, 2nd Ed. Am. Soc. Agron., Madison, WI.

Duncan, R.R., and R.N. Carrow. 1997. Stress resistant turf-type tall fescue (Festuca

arundinacea Schreb.): Developing multiple abiotic stress tolerance. Int. Turfgrass

Soc. Res. J. 8: 653-662.

Foy, C.D. 1992. Soil chemical factors limiting plant root growth. Adv. Soil Sci. 19: 97-

149.

Helyar, K.R. 1991. The management of acid soils. p. 365-382. In R.J. Wright et al. (Eds.)

Plant-soil interactions at low pH. Kluwer Academic Publishers, Dordrecht, The

Netherlands.

Huang, B., R.R. Duncan, and R.N. Carrow. 1997a. Drought-resistant mechanisms of

seven warm-season turfgrasses under surface soil drying: I. Shoot response. Crop

Sci. 37: 1858-1863.

Huang, B., R.R. Duncan, and R.N. Carrow. 1997b. Drought-resistant mechanisms of

seven warm-season turfgrasses under surface soil drying: II. Root aspects. Crop

Sci. 37: 1863-1869.

Kamprath, E.J., and C.D. Foy. 1985. Lime-fertilizer-plant interactions in acid soils. p. 91-

151. In O.P. Englestad (ed.) Fertilizer technology and use. Soil Sci. Soc. Am.,

Madison, WI.

McDaniel, P. 2005. http://soils.ag.uidaho.edu/soilorders/orders.htm. (Verified

24OCT2005).

5

Liu, H., J.R. Heckman, and J.A. Murphy. 1995. Screening Kentucky bluegrass for

aluminum tolerance. J. Plant Nutr. 18: 1797-1814.

Matsumoto, H. 2000. Cell biology of aluminum toxicity and tolerance in higher plants.

Int. Review of Cytology 200: 1-46.

Rout, G.R., S. Samantaray, and P. Das. 2001. Aluminum toxicity in plants: A review.

Agronomie 21: 3-21.

Sasaki. M., M. Kasai, Y. Yamamoto, and H. Matsumoto. 1995. Involvement of plasma

membrane potential in the tolerance mechanism of plant roots to aluminum

toxicity. Plant and Soil 171: 119-124.

Simonovicova, M., J. Huttova, I. Mistrik, B. Siroka, and L. Tamas. 2004. Root growth

inhibition by aluminum is probably caused by cell death due to peroxidase-

mediated hydrogen peroxide production. Protoplasma 224: 91-98.

Sumner, M.E. 1995. Amelioration of subsoil acidity with minimum disturbance.

p. 147-185. In Jayawardane and Stewart (eds.) Subsoil management techniques.

Adv. in Soil Sci. CRC Press, Boca Raton, FL.

Sumner, M.E., and K. Hylton. 1994. A diagnostic approach to solving soil fertility

problems in the tropics. p. 215-234. In J.K. Syers, and D.L. Rimmer (Eds.) Soil

science and sustainable land management in the tropics. CAB International,

Wallingford, England.

Vasconcelos, S.S., J. Jacob-Neto, and R.O.P. Rosiello. 2002. Differential root responses

to aluminum stress among Brazilian rice genotypes. J. Plant Nutr. 25: 655-669.

6

Wagatsuma, T., S. Ishikawa, H. Obata, K. Tawaraya, and S. Katohda. 1995. Plasma

membrane of younger and outer cells is the primary specific site for aluminum

toxicity in roots. Plant and Soil 171: 105-112.

7

CHAPTER 2

LITERATURE REVIEW

8

Subsoil acidity

Definition of the problem

At least one quarter of the world’s soils are classified as predominantly acidic,

including most of the southeastern United States, tropical Central and South America,

equatorial Africa, Southeast Asia, and northeastern Australia. Acid soils are defined here

as those soils affected in some part of the profile by acid conditions leading to Al

solubility to the extent that it impacts plant root growth (Sumner, 1995). These acid soils

are often utilized extensively for agricultural production. Furthermore as population

density increases, so too does the conversion of land from crop production to urban

landscapes, especially turfgrass-dominated environments. In both agricultural crop

production and turfgrass maintenance, the presence of acid soils, especially subsoils, can

prevent the proliferation of roots lower into the soil profile leaving plants with relatively

shallow root systems that result in reduced plant vigor, yield, and the ability to withstand

drought stress (Adams, 1984; Kamprath and Foy, 1985; Helyar, 1991; Foy, 1992; Liu et

al., 1995; Huang et al., 1997a,b; Rout et al., 2001; Vasconcelos et al., 2002). The primary

reason roots are often unable to proliferate in an acid soil environment, and thus access

subsoil waters, is due to aluminum (Al) toxicity on the root cells. This toxicity is due to

root absorption of Al, which inhibits root elongation within hours, interferes with cell

division, prevents the absorption of Ca, and causes cell death due to the destruction of the

plasma membrane of younger cells near the root tip (Sasaki et al., 1995; Wagatsuma et

al., 1995; Matsumoto, 2000). There is some evidence that cell death occurs due to

intracellular peroxidase-mediated hydrogen peroxide production (death by free radicals)

(Simonovicova et al., 2004). Plants subjected to acid soil/drought syndrome often require

9

frequent irrigation and are subjected to secondary pathogenic attack by fungi and insects

(Sumner and Hylton, 1994; Duncan and Carrow, 1997). The result for the grower is

increased labor and irrigation costs, more frequent pesticide and fertilizer applications,

lower yields and reduced plant longevity.

Description of acid-soil chemistry

The causes of soil acidity are both natural and anthropogenic, with natural

processes occurring over decades to millennia and anthropogenically-induced processes

taking place within a single to a few years. Regardless of the cause, acid soils have poor

fertility that is usually due to a combination of aluminum toxicity, manganese toxicity,

iron toxicity in reduced soils, and phosphorus, calcium, magnesium, potassium and

molybdenum deficiencies. Additionally, many acid soils are physically affected by

factors that include low water holding capacity, and susceptibility to crusting, erosion,

and compaction (von Uexküll and Mutert, 1995)

The natural cause of soil acidity is primarily weathering and to a lesser extent

decomposition of organic residues. Most acid soils are found in humid (udic or ustic)

moisture regimes in which rainfall exceeds evapotranspiration throughout the year or

during certain seasons. The rainfall is slightly acidic due to the water being in equilibrium

with the CO2 content of the atmosphere and thus forming carbonic acid (H2CO3).

Moreover, CO2 production (at higher than atmospheric pressure) by plant root and

microbial respiration increase and maintain elevated carbonic acid levels of soil water

(Bloom et al., 2005). The acidified water percolates into the subsoil and hydrolyzes

primary minerals, enabling the formation of secondary clay minerals. Basic cations are

10

leached over time, leaving clay minerals that are highly saturated with Al and Fe. Sumner

(2001) illustrated the reaction thus:

CaAl2Si2O8↓ + 2NaAlSi3O8↓ + KMgFe2AlSi3O10(OH)2↓ + 7CO2 + 20H2O + ½O2

→ Al2Si2O5(OH)4 + Al2Si4O10(OH)2↓ + Al(OH)3↓ + 2Fe(OH)3↓ + Ca2+ + 2Na+ + K+ +

Mg2+ + 7HCO3- + 5H4SiO4 (↓ indicates insoluble) (1)

As base cations are released and leach, the ionic strength of the soil solution decreases,

accelerating the dissolution of clay minerals. Solubility product considerations cause the

loss of monovalent and divalent cations first (e.g. Na+, K+, Ca2+, Mg2+) through leaching

by excess rainfall. In the absence of these cations on cation exchange sites,

soluble/exchangeable Al accumulates. The strong adsorption of Al to soil cation

exchange sites tends to out-compete basic cations for those sites, with consequences for

overall soil fertility. Thomas and Hargrove (1984) state “Exchange sites on clays and

organic matter that are saturated with Al do not readily exchange with other cations in the

soil solution. Thus, the cation exchange capacity of these soils is reduced, often to a

significant degree.” Another critical factor in the natural formation of acid soils is that in

most cases, the parent material of these soils is often felsic, with low base cation content

in the primary and secondary minerals.

Anthropogenic causes of soil acidity are primarily nitrification from the use of

ammoniated fertilizers, acid rainfall deposition, base cation removal during harvest, and

leaching of base cations replaced by H+ and subsequently Al3+. Most commercial

fertilizers are ammoniacal sources of N, and once applied to the soil are subject to the

microbial-mediated process of nitrification that oxidizes NH4+ to NO3

- and generates soil

acidity.

11

2NH4+ + 3O2 → 2NO2

- + 2H2O + 4H+ (2)

2NO2- + O2 → 2NO3

- (3)

Ammonium can also be taken up directly by plant roots that will exchange a H+ back to

the soil in order to maintain electrical neutrality. The amount of acidity actually generated

by the application of ammoniacal fertilizer depends on the N source, plant uptake of

NH4+ and NO3

-, nitrate leaching (discussed below), and denitrification, which can reverse

acidity to some extent.

Adams (1984) described the potential acidity that can be generated by different

fertilizer sources:

Anhydrous ammonia NH3(g) + 2O2 → H+ + NO3- + H2O (4)

Urea CO(NH2)2 + 4O2 + urease → 2H+ + 2NO3- + CO2 + H2O (5)

Ammonium salts 2NH4+ + X2- + 4O2 → 2H+ + X2- + 2H+ + 2NO3

- + 2H2O (6)

where X = SO42-, 2NO3

-, 2Cl-, 2H2PO4-, or HPO4

2- (after Sumner, 2005).

Note that anhydrous ammonia and urea generate one mole of protons for one mole of

NH4-N, whereas ammonium salts generate 2 moles of protons per mole of NH4-N, which

must be taken into consideration when devising soil acidity amelioration schemes.

Leaching of base cations can occur when NO3- applied to crops exceeds plant

uptake. The anion migrates down through the soil profile as a salt, ionically associated

with Ca, Mg, K, or Na. Crop regimes that focus on maximum yield often over-apply N,

since it is such a critical nutrient. The loss of base cations from the soil solution causes

base cations held on exchange sites to come into solution in order to maintain

12

equilibrium. Over time, the loss of base cations from soil exchange sites leaves those

exchange sites increasingly Al-saturated. Sod growers and turfgrass installation projects

in the Southeastern U.S. are particularly vulnerable to acidity from fertilization, due to

the rates of N applied as commercial fertilizer in order to maximize growth over a short

time period (often 50 kg N ha-1 wk-1 for 8 to 12 weeks).

Harvest of crop biomass removes base cations to the extent that these cations are

lost to the system. Roots exchange protons in order to assimilate the base cations, and the

base cations are not returned to the soil due to harvest or removal. This loss of alkalinity

is the same as a gain in acidity (Sumner, 2001). Turfgrass clippings are often returned to

soil at cutting, but if they are removed, this can represent a substantial loss of base cations

over time, increasing soil acidity. Sod growers, because they remove almost all of the

crop biomass, must be especially attentive to this potential form of acidity.

Amelioration techniques

Nitrogen reduction

Nitrogen applied in excess of plant uptake is available to leach, with concomitant

acidification of the soil. Judicious applications of N by amounts and timing that match

crop need can prevent the accumulation of NO3-N in the soil, thus reducing the

probability of cation leaching (Adams and Moore, 1983). The advent and widespread use

of slow-release N fertilizers in the turfgrass-maintenance industry has helped alleviate

this problem to some degree. Nevertheless, pressure on turfgrass managers to maintain

turf with a “dark green” appearance can lead to surplus N applications, increasing the

potential for rapid soil acidification.

13

Manure applications

Some studies have shown long-term manure applications, particularly chicken

litter, have met with success in slightly raising subsoil pH (Hue, 1992; Kingery et al.,

1993; Kingery et al., 1994; Whalen et al., 2000) or slowing pH reduction compared to

inorganic fertilizer applications (Franzleubbers et al., 2004). Poultry litter appears to have

sufficient base cations to counteract the mineralized and nitrified N in the litter, thus

replacing base cations lost by leaching. The utility of any manure in slowing acidification

or possessing a potential liming effect is based on the ratio of cations to available-N and

S in the product. Pocknee and Sumner (1997) found that although some manures such as

beef manure have the potential for acidifying soils, it only occurs occasionally due to the

volatilization of N as ammonia gas or humus formation, reducing the amount of N

available for mineralization and subsequent nitrification. Hue and Licudine (1999)

conducted a column experiment with chicken manure and CaSO4·2H2O and found Ca

movement through the soil column was facilitated by manure-derived organic molecules

that acted as ligands. The Ca-organic complexes reacted with Al, releasing Ca and

forming Al-organic complexes thought to be non-phytotoxic.

Green manures

Although green manure crops have other primary purposes, the potential exists for

certain green manures to ameliorate subsoil acidity. Black oats (Avena strigosa Schreb.)

and other plant residues were investigated in Brazil to determine the effect of these

residues on the mobility of surface applied lime in the soil profile. Ziglio et al. (1999) and

Miyazawa et al. (2002) found that surface applied lime alone affected the pH and

14

exchangeable Ca, Mg, K, and Al of the upper 10 cm, and black oats affected these same

factors throughout the 60-cm columns. They placed the efficiency of residues on lime

mobility in the following order: black oats>rye>mucuna>leucaena, and wheat residue had

no effect. They speculated the mechanism can be attributed to the formation of metal-

organic complexation followed by leaching.

Adams and Pearson (1969) investigated the use of calcium gluconate and found

that it leached well into the subsoil, where microbial activity consumed the gluconate in

excess of the Ca, raising the subsoil pH. They found that it raised the pH and

exchangeable bases to a depth of 45 cm substantially more than dolomitic limestone

applied at three times the rate of calcium gluconate. Black (1993) asserted that calcium

acetate would perform in a similar manner as calcium gluconate, but that both would be

prohibitively expensive on a commercial scale. Hern et al. (1982) and Wright et al.

(1985) leached an acid soil with EDTA in the presence of lime with nearly all the

exchangeable Al removed. In contrast, in soils treated with lime alone, alkalinity failed to

move through the profile or ameliorate exchangeable acidity. The authors speculated that,

as with gluconate and acetate, the detoxification of Al could be attributed to

complexation of Al3+ with soluble organic ligands, although no direct evidence was

presented. Additionally, it is unclear whether the Al complexes leach from the soil or

remain in the soil profile in a less toxic state.

Liming

The addition of a liming material can significantly raise soil pH and reduce Al

toxicity. Calcitic lime (CaCO3) or dolomitic lime (CaMg(CO3)2) are most often used in

15

agriculture and the turfgrass industry due to their relatively low cost of production and

transport, ease of spreading, and the fact that they are not particularly caustic. Other

liming materials have potential use but have not gained widespread acceptance, usually

because of one of the noted factors. The liming reaction that takes place in the soil begins

with the dissolution of lime (e.g. CaCO3) in the soil solution forming the Ca2+ cation and

the CO32- anion. The CO3

2- species does not exist in abundance at the pH typically found

in soils due to thermodynamic principles, and instantaneously reacts with H+ from the

soil solution, to form bicarbonates and hydroxides.

CaCO3 + H2O → Ca + HCO3- + OH- (7)

The Ca-bicarbonate and –hydroxide compounds react with active and exchangeable

acidity in the soil to ultimately consume excess H+, replace exchangeable Al3+, and

precipitate Al(OH)3. (Thomas and Hargrove, 1984; Sumner, 2005). This raises the pH of

the soil solution and neutralizes exchangeable Al in a form that does not react with plant

roots (amorphous Al(OH)3 and gibbsite).

2Soil-Al3+ + 3Ca2+ + 3HCO3- + 3OH → 3Soil-Ca2+ + 2Al(OH)3↓ + 3CO2↑. (8)

The rate-limiting factor in this series of reactions is the dissolution of the liming material,

which is a function of surface area, and thus particle size. Powdered limes will generally

react more quickly than pelletized lime; however they are more difficult to handle and

spread.

The Al content and form found in the soil can also have an impact on the rate at

which lime affects soil pH. Coleman and Thomas (1964) and Kissel et al. (1971) found

that the concentration of H+ ions in the soil solution is a function of the hydrolysis rate of

16

Al3+ or hydroxy-Al, which is influenced by their adsorption to clay or organic matter, and

that the rate of hydrolysis is increased by the introduction of salts into the soil solution.

Thomas and Hargrove (1984) note that, conversely, strong adsorption of intermediate

reaction compounds such as Al(OH)2+ reduces the rate of hydrolysis, causing the

neutralization reaction to occur over a much longer period of time. They found in data

from Moschler et al. (1962) that 5 yr after an application of 18 Mg dolomitic limestone

ha-1 to several Virginia soils, the pH was still rising, appreciable titratable acidity

remained, and that none of the soil pH levels had exceeded 7.0.

Calcitic or dolomitic lime is an appropriate product to ameliorate topsoil,

especially under conditions where it can be incorporated into the soil. Each, however, are

sparingly soluble and thus do not move readily through the soil profile. Therefore these

products are poor choices to ameliorate subsoil. Despite its limitations, researchers have

investigated the use of lime in ameliorating subsoils, with varying degrees of success.

Lathwell and Peech (1964), Bradford and Blanchar (1977), Perez-Escolar and Lugo-

Lopez (1978), McKenzie and Nyborg (1984), Hammel et al. (1985), and Sumner et al.

(1986) conducted studies that mechanically incorporated lime deep into the soil profile

(0.5-1.0 m) to neutralize subsoil acidity. Sumner (1995) reported that in all cases, this

method was highly effective in promoting deep rooting and resulted in higher yields. He

noted the main limitation to these studies was the heavy equipment and high energy cost

necessary to mix the lime at great depths. Other researchers (Anderson and Hendrick,

1983; Farina and Channon 1988; Jayawardane et al., 1995; Kirchof et al., 1995; and

Richards et al., 1995) used a lime-slotting technique in bands behind a deep-tine ripper

and also found improved root growth, water extraction, and yield over surface-applied

17

lime. Pearson et al. (1973) found that some crops such as cotton had stronger geotropic

tendencies than chemotropic, concluding that these types of crops would benefit little

from lime-slotting techniques unless the slots were placed in close proximity. Kauffman

and Gardner (1978) found wheat roots had a strong chemotropic response to subsoil

liming, with almost all visible roots following the limed slots.

Gascho and Parker (2001) participated in a long-term (31 yr) study monitoring the

effects of surface applied limestone on subsoil acidity. They found that regular

applications of dolomitic lime significantly increased pH, Ca, and Mg to a depth of 90

cm, and increased cotton yields. Sumner (2001) noted the necessity of maintaining

topsoil pH above 5.6 in this case so that surface-horizon exchangeable acidity does not

prevent the eventual downward movement of the lime. Obviously, the major limitation to

adopting this technique is the extensive time (and adequate rainfall over long time

periods) required to achieve success.

In an experiment focusing on the unique constraints of turfgrass, Shim and

Carrow (1997) found limestone slurry plus Turf Conditioner™ applied with a high-

pressure liquid injector (Verti-Drain®) resulted in modest increases in subsoil pH and

extractable Ca. They cautioned that further research was necessary to establish more

definitive conclusions as to whether injected lime or other materials would have a long-

term impact on soil chemical parameters such as base saturation and pH.

Liming with fertilizer (Ca(NO3)2; lime and NH4NO3)

The surface application of lime in conjunction with nitrogenous fertilizer has

proved to be an effective technique in ameliorating subsoil acidity (Percival et al., 1955;

18

Pearson and Abruña, 1961; Pearson et al., 1962; Abruña et al., 1964; Weir, 1975; and

Adams, 1981). The rapid nitrification of ammoniacal-N fertilizer in a well-limed topsoil

creates conditions conducive to Ca(NO3)2 leaching into the subsoil (Adams et al., 1967).

When roots of a high-N consuming crop are present in sufficient quantities in the subsoil,

anion uptake by the plant roots far in excess of cation uptake causes the subsoil pH to

increase. Adams and Pearson (1969) found similar results with high application rates of

Ca(NO3)2 on forages, although lower rates (224 kg N ha-1) of Ca(NO3)2 showed little

effectiveness in increasing subsoil pH (Perez-Escolar and Lugo-Lopez, 1978). The

presence of a crop with roots that are acid-tolerant enough to be present in the subsoil to

take up NO3-N is essential to the process. Kotze and Deist (1975) and Pleysier and Juo

(1981) found only slight shifts in exchangeable Ca and Al in subsoils over clean fallow

from the application of Ca(NO3)2. The primary difficulty in the adoption of this process is

the increasing cost of nitrogenous fertilizers, and the high rates necessary to leach

significant quantities of Ca(NO3)2 into the subsoil.

Gypsum

Gypsum (CaSO4·2H2O) is a salt and does not affect soil pH in the same manner as

a liming agent. In water, gypsum dissolves much more readily than lime, a characteristic

that gives it much of its utility (Bolan et al., 1991). However, dissolved gypsum does not

raise the pH of deionized water, and in soils the pH often drops slightly due to the salt

effect. Gypsum affects the soil solution and growth of acid-sensitive roots in three

primary ways. As dissolved gypsum leaches into the acid subsoil by excess moisture it

supplies Ca2+, which is often deficient at depth in acid soils. Gypsum also increases the

19

ionic strength of the soil solution, changing the distribution of Al species present and

decreasing Al activity. Finally, specific adsorption of SO42- ions may release OH- ions

and increase the negative charge on soils with high variable charge components, or SO42-

may react with Al to form aluminum sulfate precipitates or AlSO4+.

It is often observed that the combined effects of low Ca and high exchangeable Al

make many acid subsoils a hostile environment for many crops. Adams and Lund (1966)

stated that based on the work of Ragland and Coleman (1959) and Howard and Adams

(1965) with cotton and peanut roots, it seems that except for sandy soils, poor root

growth in acid subsoils is generally not the result of Ca deficiency, but more likely Al

toxicity. Some crops, such as corn (Souza and Ritchie, 1986) and alfalfa (Sumner and

Carter, 1988), tend to have high Ca requirements, and thus respond well to increased soil

Ca applications.

In an early experiment that established the efficacy of gypsum as a material

suitable for subsoil amelioration, Reeve and Sumner (1972) demonstrated that compared

to Ca(OH)2, gypsum did not dramatically raise exchangeable Ca nor eliminate

exchangeable Al in a topsoil. Nonetheless, because of the presence of SO42- as a counter-

ion enhancing its ability to move down the soil profile, it was more effective than

Ca(OH)2 in increasing subsoil Ca and reducing exchangeable Al.

Pavan et al. (1982) studied the treatment of acid soil with gypsum, finding that

where gypsum was added in amounts equivalent to the aluminum removed by molar KCl,

about 40% of the solution aluminum was present as the sulfate complex. This is

significant because Al complexed with SO42- was found by Tanaka et al. (1987), Kinraide

and Parker (1987), and Noble et al., (1988) to have little or no toxicity on plant roots.

20

Moreover, O’Brien and Sumner (1988) and Courchesne and Hendershot (1990) found

that with gypsum treatment Al is retained in the soil to the degree that Al measured in

column leachate is quite low. They suggested that some Al may be precipitated as

aluminum hydroxy sulfates that are sparingly soluble. Noble et al. (1988) combined a

Ca/Al solution study of soybean root response with computer-aided predictions of Al

species activity and found that the addition of Ca in solution decreased overall Al

activity, and especially the Al3+ species activity. They formulated an equation that

accounted for the ionic charge and activity of Ca and certain Al species in solution and

obtained a high degree of goodness of fit to the relative taproot length of soybean in

hydroponic solutions (r2= 0.900). The equation was expressed thus:

CAB = [2log(Ca2+)] – {3log(Al3+) +2log[Al(OH)2+] + log[Al(OH)2+]} (9)

where CAB is calcium-aluminum balance. The coefficients in the equation were

introduced to accommodate the assumed importance of valence in ion interactions at the

root surface. The Al species selected for inclusion in the equation were those with the

highest correlation between species concentration (as predicted by computer-aided

models) and relative taproot length.

In certain cases it appears that gypsum has a slight, but measurable effect on

raising the pH of acid soils. Reeve and Sumner (1972) proposed a mechanism that

involves the ligand exchange of SO42- for OH- ions on iron- and aluminum-sesquioxides,

that neutralize soluble Al:

21

2[Fe,Al] [Fe,Al] + Ca2+ + SO42- 2[Fe,Al] [Fe,Al] + Ca(OH)2

OH OH

OH

OH

SO4-

2Al3+ + 3Ca(OH)2 → 2Al(OH)3 + 3Ca2+ (10)

Adams and Rawajfih (1977) and Nordstrom (1982) demonstrated that under acid

soil and elevated SO42- levels, aluminum sulfate minerals would precipitate:

3Al3+ + K+ + 2Ca2+ + SO42- + 3H2O → KAl3(OH)6(SO4)2 + 3H+ + Ca2+ (11)

or 4Al3+ + 10OH- + SO42-→ Al4(OH)10SO4 (12)

or Al3+ + OH- + SO42- → AlOHSO4 (13)

Although gypsum-treated soils have elevated SO42- levels compared to untreated soils,

they are still often undersaturated with respect to SO42- as CaSO4. Alva et al. (1991)

proposed salt sorption in which cation and anion are adsorbed by the subsoil phases in

equivalent amounts and can act as a control on the level of Al3+ in solution probably by

the co-immobilization of Ca2+ and Al3+ together with the SO42- (Sumner, 1990, 1995).

Sumner (1995) proposed measuring electrical conductivity (EC) and the difference in pH

between a soil saturated with CaCl2 or CaSO4 as a test for the potential utility of a

gypsum application to a given soil. By plotting the difference in pH between the two Ca-

salts vs. gypsum sorbed (calculated from EC), a determination could be made as to

whether the soil would respond to a gypsum application.

22

Gypsum effects on agronomic crops

The application of gypsum in field trials has resulted in increased depth of rooting

and higher yields for some agronomic crops on a range of acid soils. Due to the fact that

most soils do not have yield-limiting deficiencies of Ca in their topsoils, improved yields

in gypsum-amended soils over untreated soils can be likely attributed to successful

subsoil-acidity amelioration, leading to increased root proliferation deeper in the soil

profile. Under field conditions in which surface soils become dry due to

evapotranspiration exceeding rainfall for extended periods, deeper roots are able to

exploit sub-surface soil water, providing a measure of drought resistance, resulting in

comparatively improved yields. Conversely, the lack of yield response to gypsum on acid

soils may be attributed to several factors, including 1) genetic resistance of crop or

variety to soil acidity/Al toxicity; 2) variation in subsoil properties i.e. some subsoils are

more acidic than others, including within-field variation, and/or soil physical properties

that limit root growth regardless of soil chemical properties; 3) rainfall quantities and

patterns that may either limit sufficient gypsum movement through the soil profile or

provide continuously adequate soil water near the surface such that water is not a yield-

limiting factor; and 4) time since application, especially if insufficient time for gypsum

leaching to the subsoil profile has been granted.

Maize (Zea mays L.) has demonstrated a positive yield response to gypsum

applications ranging from 4 to 35 Mg ha-1on Ultisols and Oxisols from several

geographic regions, ranging from 19% yield response on a South African Ultisol (Farina

and Channon, 1988) to 47-77 and 82% yield response on Brazilian Oxisols and Ultisols,

respectively (Malavolta, 1992: Souza et al., 1992). Some studies have indicated increased

23

shoot or yield response to soil-gypsum applications without a corollary increase in root

mass (Punshon et al., 2001; Caires et al., 2004). In Caires et al. (2004) the increased yield

was attributed to an increase in the Ca soil saturation of the surface soil layer, implying

that Ca as a nutrient was yield-limiting in this case.

Wheat (Triticum aestivum L.) has also shown positive gains in yield from gypsum

applications, although applications rates and yield responses between studies are not well

correlated. For example, McLay et al. (1994) realized a 45% yield response after the

application of 9 Mg ha-1, and Souza et al. (1992) obtained a 76% yield response on a

Brazilian Oxisol by applying 6 Mg ha-1. Guimares (1986) reported a 63% yield response

on wheat after applying 5.4 Mg ha-1 on a Brazilian Oxisol, using phosphogypsum.

Alfalfa (Medicago sativa L.) has exhibited positive growth responses to gypsum-

ameliorated acid subsoils when sufficient time (at least one year) was given for gypsum

movement through the soil profile. Sumner and Carter (1988) applied gypsum at 10 Mg

ha-1 to a Georgia Ultisol and realized a 6% yield response on alfalfa the first year,

increasing to 26% the second year, 27% the third year, and a dramatic 117%, 64%, and

213% in years 4, 5, and 6, respectively. The researchers noted that years 4 and 5 were

especially dry, possibly highlighting the benefit to alfalfa, a naturally deep-rooting plant,

being able to exploit subsoil water more effectively than in the control treatment. The

combination of lime-amended topsoil and gypsum-amended subsoil optimized the

probability SO42- from gypsum would not react with topsoil oxides but was “free” to

migrate lower into the soil profile (Sumner and Carter, 1988; Farina and Channon, 1988).

Alfalfa has a high Ca content that may be a confounding factor in its response to gypsum

ameliorated acid soils. Clark and Baligar (2003) grew alfalfa in pots with an acid soil

24

(80% Al saturation) and observed poor root growth in unamended soil, but significant

improvement in gypsum-amended soils. Dry matter (DM) yield increased from 8.3 mg

plant-1 in the control to 1615 mg plant-1 in flue-derived gypsum amended soil, a 62%

increase over the optimal lime-amended soil and an identical root:shoot ratio. In contrast,

Chen at al. (2005) focused on S deficiencies in Ohio soils, and observed a 40% increase

over control soils in new alfalfa plantings followed by 5 to 6% overall yield

improvements in subsequent years. These soils all had pH readings of 5.7 or higher and

high base saturation levels, so the likelihood of a response due to increased pH or Ca was

remote.

Several investigators studied the effect of gypsum on soybean (Glycine max L.)

crop yields and found favorable results (Hammel et al., 1985; Carvalho et al., 1986;

Sousa et al., 1992b). The experiment conducted by Carvalho et al. (1986) only

demonstrated a 2% yield response to gypsum applied at 4 Mg ha-1, but they noted the

very wet growing season prevented any drought stress on the control treatments. In a

recent study, Caires et al. (2003) conducted a split-plot field experiment on a Brazilian

Oxisol for 3 yr under no-till soybeans, and observed no improvement in grain yields.

They concluded that even though gypsum improved subsoil pH, Ca, and S-SO42-, it was

not an “interesting” strategy for no-till soybeans. However, no rainfall data was provided

and it is unclear what effect this may have had on results.

Studies of clover (Trifolium spp.) response to gypsum applications have not

sufficiently documented a significant correlation between gypsum-ameliorated subsoils

and increased clover yield. Ritchey et al. (2004) investigated clover growth response to

gypsum or lime and found in pastures interseeded with red clover or white clover, the

25

percentage clover in the sward tripled as pH increased from 4.3 to 5.0 and herbage mass

increased 75% as clover percentage increased. However, the herbage growth response

was more closely correlated with soil properties in the top 2.5cm soil than from deeper

samples, indicating clover improvements were likely a response to lime-improved topsoil

and not due to subsoil amelioration. Similarly, Brauer et al. (2002) had earlier noted in

clover trials of root nodulation, that after comparable applications of lime or gypsum, the

number of nodules per primary root was more closely associated with changes in topsoil

pH from lime than soil Ca alone from gypsum. Clark and Baligar (2003) found

insignificant (2.5%) increases in white clover DM grown in pots with gypsum-amended

soil over control pots.

Cotton (Gossypium hirsutum L.) results have been mixed, with some research

showing a lint yield response that was significant (16%) in only one of three years after

applying phosphogypsum at 10 Mg ha-1 (McCray et al., 1991), and seed yield showing a

14% positive response on a 3.64 Mg ha-1 application, but a –10% response on a 1.85 Mg

ha-1 application (Rosolem and Machado, 1984). Punshon et al. (2001) observed cotton

root growth in a mesocosm study, and recorded a slight but significant improvement in

root growth in gypsum-treated soil compared to the control soil, but wide variability in

the study results made interpretation difficult.

A 3-yr investigation by McLay et al. (1994) on a sandy Australian Aquic

Hapludox revealed substantial yield suppression of lupins (Lupinus angustifolius L.)

(important in Australian wheat rotations) by gypsum, but that the negative gypsum effect

was not evident when applied in conjunction with lime on the soil surface. A gypsum

application was not recommended if lupins were to be included in the crop rotation

26

within two years. A potentially excessive K:Mg ratio was postulated for decreased plant

size, although the authors strongly urged further research on the mechanisms involved.

Long-term studies have illuminated the importance of the time factor in a gypsum

influence on agronomic performance. Farina et al., (2000a,b) conducted a study over 11

seasons investigating the effects of gypsum and lime applications on subsoil acidity and

found that significant reductions in subsoil acidity did not occur until the 6th year in the

0.6-0.75 m depth range, but that the application increased pH by 0.4 units, decreased

exchangeable acidity by 1.5 cmol(+) kg-1, and decreased acid saturation by 30%. Soil

amelioration was accompanied by a 25% increase in maize yield, which was superior to

all calcite lime incorporation strategies. Toma et al. (1999) published results of long-term

effects on maize and alfalfa grown 16 yr after a gypsum application. They reported

reductions in exchangeable Al to a depth of 80 cm, but little change in pH. Maize yields

were 29-50% greater in gypsum plots versus control plots, and alfalfa yields were 50%

higher. The authors note that initial gypsum material and application costs amortized over

this many years makes this ameliorant highly economical.

Gypsum and subsoil acidity effects on grasses

Turfgrasses and forages are grouped according to the climatic zone in which they

are best adapted. Cool-season grasses include fescue spp. (Festuca spp.), bentgrass

(Agrostis spp.), bluegrass (Poa spp.) and ryegrass (Lolium spp.). Grasses adapted to

maximum growth under high summer temperatures are classified as warm-season grasses

and include bermudagrass (Cynodon spp.), zoysiagrass (Zoysia spp.), centipedegrass

(Eremochloa ophiuroides) and St. Augustinegrass (Stenotaphrum secundatum). The

27

Piedmont region of the Southeastern United States is located in the northern end of the

humid, warm-season zone and lies near the transition zone (between the cool-season and

warm-season zones) that is characterized by occasional hot, dry summers and/or

significantly cold winters. Sporadic, seemingly random shifts in temperature and rainfall

patterns often add significant stresses on turfgrass viability. Additionally, area soils are

typically characterized as having high soil strength and acid subsoils, creating a complex

of stresses Foy (1992) termed an ‘acid soil complex.’

In early work that characterized grass tolerance to Al toxicity, Adams et al.,

(1967) considered the effects of lime sources and rates on Coastal bermudagrass on a

Cecil sandy loam under high NH4NO3-N rates over 7 yr. They found maximum

production was associated with a soil pH of 4.8 (1:1 soil:water), suggesting bermudagrass

has a high exchangeable Al tolerance. Although there was a marked increase in forage

production from pH 4.3 to 4.8 (1,592 kg ha-1 higher in the last year of the study), no

significant increase was measured above pH 4.8. A high lime rate of 20.2 Mg ha-1

prevented the development of soil acidity over the study, even though a total of 5,605 kg

N ha-1 cumulative, was applied, half of which was NH4-N.

More recent research on turfgrass cultivar improvements to inhibit or minimize

the effects of Al toxicity or characterize drought resistance has been conducted over the

past decade (Murray and Foy, 1978; Carrow, 1995, 1996a,b; Liu et al., 1995,1996a,b;

Huang et al., 1997a, b; Duncan and Carrow, 2001; Carrow and Duncan, 2003; Liu, 2005).

A common methodology employed is screening many cultivars of a turfgrass species

with hydroponic or sand/solution cultures that have one high rate of solution Al (e.g.

640µM Al). Liu (2005) and coworkers (Liu et al., 1997a,b,c; 1996a,b; 1995) compared

28

seeded bermudagrass, bentgrass, fescue, and bluegrass cultivars and found significant

differences in resistance to Al toxicity. Soil-column and field studies have also been

employed to better characterize plant drought resistance/avoidance among grass cultivars

(Murray and Foy, 1978; Carrow, 1996; Huang et al., 1997a,b; Qian et al., 1997). One

example of a column and field study combination to characterize zoysiagrass root growth

under moisture stress is Marcum, et al. (1995), who found a high correlation between the

average maximum rooting depth, total root weight, and root numbers at different depths

with percent green plot cover in the field under deficit irrigation (0 and 35 % ET). They

indicated that rooting weight, depth, and branching at lower depths are important

drought-resistance mechanisms for this species. Putting zoysiagrass in context with other

grass species, Qian et al. (1997) found zoysiagrass to be the least able to root deeply

compared to tall fescue, hybrid bermudagrass, and buffalograss, even though the study

was conducted in the field on Aquic Argiudolls.

Several studies have been conducted on the response of grasses, either as forages

or turfgrass, to soil gypsum amendments. Passos et al. (1997) compared lime, gypsum,

and lime + gypsum applications on two tropical forages grown on an acid latossol and

found the highest dry matter yields taken over 6 cuttings (258 days after treatment

application) occurred on soils treated with lime + gypsum. Mora et al. (2002) observed

significant shifts in species distribution on acid Andisols treated with limestone,

dolomite, and gypsum versus control plots from predominantly weeds to annual ryegrass

and white clover as Al saturation dropped from 20 to less than 1%. Control plot soils

contained 87% weed species and combined-amendment soils had only 17% weed species

present, indicating lime and gypsum inputs gave the ryegrass and clover a competitive

29

advantage over weed species. Clark and Baligar (2003) grew K-31 tall fescue and

orchardgrass on flue-derived gypsum-treated acid soils applying extraordinarily high

rates (up to 75% of gravimetric soil content) and found significant increases in tall fescue

root and shoot growth over the control treatment and root and shoot yields that were

comparative to incorporated lime. Orchardgrass, in contrast, grew considerably better in

control soils and did not respond to gypsum or lime amended soils to the degree that tall

fescue responded.

Gypsum applications on acid soils growing turfgrasses such as perennial ryegrass

and bentgrass have resulted in significant increases in nutrient uptake of NH4+and NO3

-

(Bailey, 1992) and Ca and PO4 (Kuo, 1993). However, a golf course study on calcareous

sand-based greens by St. John et al. (2003) found no increase in creeping bentgrass leaf

tissue-Ca from gypsum over control, limed, chelated Ca, or Ca(NO3)2 treatments. The

researchers also measured visual quality, density, uniformity, root dry mass, and clipping

yield and reported no differences between treatments. The gypsum-treated bentgrass had

an 11% drop in leaf tissue Mg over the control plots, but this was not considered serious

based on the work of Mills and Jones (1996). It is notable that the concentration of

extractable Ca in the soil did not increase in Ca-amended plots over the control,

indicating that the porous sand-based media was not amenable to retaining additional Ca

inputs. This raises the question as to whether a lack of leaf-tissue Ca accumulation was

due to the amendment or whether the amendment washed out of the growing media.

Several forage studies that included gypsum as a treatment have explored yield

responses to SO4-S inputs. In a sulfur-deficient soil of the Gulf Coastal Plain, Phillips and

Sabbe (1994) found a significant interaction exists between S and N sources and N rates

30

to affect N and S recovery on Coastal bermudagrass. They found %N recovery declined

with increasing N rates, but were significantly higher with S fertilizer. They also

observed higher S recovery with gypsum than with wettable S, although no explanation

was offered for the response difference. Dorsett et al. (1995) conducted a similar study on

a slightly acid, montmorillonitic Texas soil and found by-product gypsum provided from

6 to 30% higher Coastal bermudagrass forage production than control plots, depending

on the cutting time and year of production. Overall mean forage production on the

gypsum treated soil increased 20% over the control soil plots.

Bowman et al. (2002) conducted a greenhouse study that analyzed NO3-N

leaching through sandy soil columns and noted zoysiagrass only recovered 63% of

applied N (applied as NH4NO3) and hybrid bermudagrass recovered 84% of N. Most N

lost through the column occurred at the initial leaching event following fertilizer

application. This has significant implications for subsoil acidity under turfgrass that is

maintained on a high N regimen. Golf courses and athletic fields on sandy soils such as

those found in Florida and the Gulf Coast also have the potential to create conditions

favoring NO3-N leaching into groundwater, which is often at shallow depths in these

regions. Sartain (1985) compared nutrient retention in the soil by depth after application

of different fertilizer-N sources on bermudagrass and found the highest pH values, Ca,

Mg, K, and P (Mehlich I extractable) in the order

IBDU>(NH4)2SO4 + Ca(OH)2 > (NH4)2SO4 >>(NH4)2SO4 + Elemental S

which suggets that N- and S-causing soil acidification contributed to the loss of CEC in

the soil. Nevertheless, turfgrass-clipping yield was not affected by the drop in pH, and in

fact, maximum growth rates occurred at pH<4.0.

31

Very little research is currently available on turfgrass response to gypsum

applications on acidic, low Ca, low CEC soils. The purpose of this series of experiments

was to build a body of knowledge on the root growth of selected turfgrass species and

cultivars in the presence of exchangeable aluminum and Ca from flue-gas derived

gypsum.

Sources of gypsum

A long-established source of high-purity gypsum is mined gypsum imported from

Novia Scotia, Canada, and is used primarily in industrial and construction applications.

Other sources of gypsum include by-product gypsum. Phosphogypsum is a by-product of

the phosphorus mining industry, located primarily in Florida. There are currently over

500 million metric tons stockpiled and 30 million metric tons produced each year, but the

material has variable quantities of radionuclides and is therefore highly restricted in its

use (Miller, 1995). Flue-gas desulfurization gypsum is increasingly produced by coal-

fired electric utilities to remove SO2 from stack gases as a result of the Federal Clean Air

Act (last amended 1990). Approximately 12 million metric tons were produced in 1994

and is increasing annually (Miller et al., 2000). By-product gypsum is greater than 90%

pure CaSO4·2H2O, is as soluble as mined material, and contains only low,

environmentally safe levels of contaminants such as trace inorganics (Miller, 1995).

Availability of by-product gypsum varies by location, but if close to point-of-use, can be

very economically obtained. At present, mined gypsum can be purchased for $50-100 per

metric ton, compared to $5-10 at point of sale for by products (Miller, personal

communication, 2005).

32

Objectives

Limited basic research is currently available describing turfgrass root response to

Ca and Al in solution, information which is needed to gauge potential response of these

plants to gypsum applications in the field on acidic, low Ca, low CEC soils. The purpose

of this series of experiments was to build a body of knowledge on the root growth of

selected turfgrass species and cultivars in the presence of exchangeable Al and Ca from

flue-gas derived gypsum. Specific objectives were to 1) investigate the interaction of

solution concentrations of Ca and Al on root growth of seedling turfgrass species and

varieties using hydroponic culture techniques; 2) evaluate the impact of CaSO4·2H2O

applications on soil properties and rooting patterns of turfgrasses grown in soil columns

under control conditions in the laboratory; and 3) evaluate field responses of selected

turfgrasses to gypsum and lime applications in terms of root depth and water extraction

within the soil profile.

33

References Cited

Abruña, F., J. Vicente-Chandler, and R.W. Pearson. 1964. Effects of liming on yields and

composition of heavily fertilized grasses and on soil properties under humid

tropical conditions. Soil Sci. Soc. Am. Proc. 28: 657-661.

Adams, F. 1981. Alleviating chemical toxicities: liming acid soils. p. 269-301. In G.F.

Arkin and H.M. Taylor (Eds.) Modifying the root environment to reduce crop

stress. Am. Soc. Agr. Eng., St. Joseph, MI.

Adams, F. 1984. Crop response to lime in the southern USA. p. 211-266. In F. Adams

(Ed.) Soil acidity and liming, 2nd Ed. Am. Soc. Agron., Madison, WI.

Adams, F., and Z.F. Lund. 1966. Effect of chemical activity of soil solution aluminum on

cotton root penetration of acid subsoils. Soil Sci. 101: 193-198.

Adams, F., and B.L. Moore. 1983. Chemical factors affecting root growth in subsoil

horizons of Coastal Plain soils. Soil Sci. Soc. Am. J. 47: 99-102.

Adams, F., and R.W. Pearson. 1969. Neutralizing soil acidity under bermudagrass sod.

Soil Sci. Soc. Am. Proc. 33: 737-742.

Adams, F., and Z. Rawajfih. 1977. Basaluminite and alunite: A possible cause of sulfate

retention by acid soils. Soil Sci. Soc. Am. J. 41: 686-692.

Adams, W.E., A.W. White, Jr., and R.N. Dawson. 1967. Influence of lime sources and

rates on ‘Coastal’ bermudagrass production, soil profile reaction, exchangeable

Ca and Mg. Agron. J. 59: 147-149.

Alva, A.K., M.E. Sumner, and W.P. Miller. 1990. Reactions of gypsum or

phosphogypsum in highly weathered acid subsoils. Soil Sci. Soc. Am. J. 54: 993-

998.

34

Alva, A.K., M.E. Sumner, and W.P. Miller. 1991. Salt absorption in gypsum amended

acid soils. p. 93-97. In R.J. Wright et al. (Eds.) Plant-soil interactions at low pH.

Kluwer Acad. Pub., Netherlands.

Anderson, D.L., and J.G. Hendrick. 1983. Subsoil lime injector. Soil Sci. Soc. Am. J. 47:

337-339.

Bailey, J.S. 1992. Effects of gypsum on the uptake, assimilation and cycling of N-15-

labeled ammonium and nitrate-N by perennial ryegrass. Plant and Soil 143: 19-

31.

Black, C.A. 1993. Soil fertility evaluation and control. CRC Press, Inc., Boca Raton, FL.

p. 712-728.

Bloom, P.R., U.L. Skyllberg, and M.E. Sumner. 2005. Soil acidity. p. 411-459. In L. Al-

Amoodi et al. (Eds.) Chemical processes in soils. Soil Sci. Soc. Am. Series 8.

Madison, WI.

Bolan, N.S., J.K. Syers, and M.E. Sumner. 1991. Dissolution of various sources of

gypsum in aqueous solutions and in soil. J. Sci. Food Agric. 57: 527-541.

Bowman, D.C., C.T. Cherney, and T.W. Rufty. 2002. Fate and transport of nitrogen

applied to six warm-season turfgrasses. Crop Sci. 42: 833-841.

Bradford, J.M., and R.W. Blanchar. 1977. Profile modification of a Fagiudalf to increase

crop production. Soil Sci. Soc. Am. J. 41: 127-131.

Brauer, D., D. Ritchey, and D. Belesky. 2002. Effects of lime and calcium on root

development and nodulation of clovers. Crop Sci. 42: 1640-1646.

35

Caires, E.F., J, Blum, and G. Barth. 2003. Changes in chemical soil characteristics and

soybean response to lime and gypsum applications in a no-tillage system. Rev.

Bras. Ciêc. Solo 27: 275-286.

Caires, E.F., M.T. Kusman, and G. Barth. 2004. Changes in soil chemical properties and

corn response to lime and gypsum applications. Rev. Bras. Ciêc. Solo 28: 125-

136.

Carrow, R.N. 1995. Drought resistance aspects of turfgrasses in the Southeast:

Evapotranspiration and crop coefficients. Crop Sci. 35: 1685-1690.

Carrow, R.N. 1996a. Drought avoidance characteristics of diverse tall fescue cultivars.

Crop Sci. 36: 371-377.

Carrow, R.N. 1996b. Drought resistance aspects of turfgrasses in the Southeast: Root-

shoot responses. Crop Sci. 36: 687-694.

Carrow, R.N., and R.R. Duncan. 2003. Improving drought resistance and persistence in

turf-type tall fescues. Crop Sci. 43: 978-984.

Carvalho, L.J.C.B., R.L. Gomide, G.C. Rodriguez, D.M.G. Souza, and E. de Freitas.

1986. Reposta do milho a aplicacao de gesso e deficit hidrico em solo de

Cerrados. An. Sem. Uso. Fosfogesso Agricultura 61-83.

Chen, L., W.A. Dick, and S. Nelson, Jr. 2005. Flu gas desulfurization products as sulfur

sources for alfalfa and soybean. Agron. J. 97: 265-271.

Clark, R.B., and V.C. Baligar. 2003. Growth of forage legumes and grasses in acidic soil

amended with flue-gas desulfurization products. Comm. Soil Sci. Plant Anal. 34:

157-180.

36

Coleman, N.T., and G.W. Thomas. 1964. Buffer curves of acid clays as affected by the

presence of ferric iron and aluminum. Soil Sci. Soc. Am. Proc. 28: 187-190.

Courchesne, F., and W.H. Hendershot. 1990. The role of basic aluminum sulfate minerals

in controlling sulfate retention in the mineral horizons of two Spodosols. Soil Sci.

150: 571-578.

Dorsett, D.J., L Nickel, and H.D. Pennington. 1995. Evaluation of coal generated by-

product gypsum as a soil amendment on improved hybrid bermudagrass. Am.

Forage Grassland Council Proceedings 4: 136&.

Duncan, R.R., and R.N. Carrow. 1997. Stress resistant turf-type tall fescue (Festuca

arundinacea Schreb.): Developing multiple abiotic stress tolerance. Int. Turfgrass

Soc. Res. J. 8: 653-662.

Duncan, R.R., and R.N. Carrow. 2001. Molecular breeding for tolerance to

abiotic/edaphic stresses in forage and turfgrass. p. 251-260. In G. Spangenberg

(Ed.) Molecular breeding of forage crops. Kluwer Acad. Pub., Netherlands.

Edmeades, D.C., and A.M. Ridley. 2001. Diagnosis and amelioration of surface and

subsoil acidity. p. 161-185. In M. Farina et al. (Eds.) Proc. 5th Plant-soil

interactions at low pH symposium. Inst. For Soil, Climate and Water, Pretoria,

South Africa.

Farina, M.P., and P. Channon. 1988. Acid-subsoil amelioration: I. A comparison of

several mechanical procedures. Soil Sci. Soc. Am. J. 52: 169-175.

Farina, M.P., P. Channon, and G.R. Thibaud. 2000a. A comparison of strategies for

ameliorating subsoil acidity:I. Long-term growth effects. Soil Sci. Soc. Am. J. 64:

646-651.

37

Farina, M.P., P. Channon, and G.R. Thibaud. 2000b. A comparison of strategies for

ameliorating subsoil acidity:II. Long-term soil effects. Soil Sci. Soc. Am. J. 64:

652-658.

Foy, C.D. 1992. Soil chemical factors limiting plant root growth. p. 97-149. In Hatfield,

J.L., and B.A. Steward (Eds.) Advances in Soil Science 19. Springer-Verlag, Inc.,

New York.

Franzluebbers, A.J., S.R. Wilkinson, and J.A. Stuedemann. 2004. Pasture management in

the Southern Piedmont USA: VIII. Soil pH and nutrient cations. Agron. J. 96:

1390-1399.

Gascho, G.J., and M.B. Parker. 2001. Long-term liming effects on Coastal Plain soils and

crops. Agron. J. 93: 1305-1315.

Guimaraes, P.T.C. 1986. O. gesso agricola n neutralizacao do aluminio nas camadas

subsuperficias do solo: Aplicacoes as cultivars anuais e Perenes. P. 145-168. In

An. Sem. Uso Fosfogesso Agricultura, EMBRAPA, Brasilia D.F., Brazil.

Hammel, J.E., M.E. Sumner, and H. Shahandeh. 1985. Effect of physical and chemical

profile modification on soybean and corn production. Soil Sci. Soc. Am. J. 49:

1508-1512.

Hargrove, W.L., and G.W. Thomas. 1982. Conditional formation-constants for

aluminum-organic matter complexes. Can. J. Soil Sci. 62: 571-575.

Helyar, K.R. 1991. The management of acid soils. p. 365-382. In R.J. Wright et al. (Eds.)

Plant-soil interactions at low pH. Kluwer Academic Publishers, Dordrecht, The

Netherlands.

38

Hern, J.L., A. Menser, R. Sidle, R.L. Wright, and O.L. Bennett. 1982. The effects of

organic complexing materials on ion movement and rooting environments in acid

soils. Agron. Abs. 174.

Howard, D.D., and F. Adams. 1965. Calcium requirement for penetration of subsoils by

primary cotton roots. Soil Sci. Soc. Am. Proc. 29: 558-562.

Huang, B., R.R. Duncan, and R.N. Carrow. 1997a. Drought-resistant mechanisms of

seven warm-season turfgrasses under surface soil drying: I. Shoot response. Crop

Sci. 37: 1858-1863.

Huang, B., R.R. Duncan, and R.N. Carrow. 1997b. Drought-resistant mechanisms of

seven warm-season turfgrasses under surface soil drying: II. Root aspects. Crop

Sci. 37: 1863-1869.

Hue, N.V. 1992. Correcting soil acidity of a highly weathered Ultisol with chicken

manure and sewage sludge. Commun. Soil Sci. Plant Anal. 23: 241-264.

Hue, N.V., and D.L. Licudine. 1999. Amelioration of subsoil acidity through surface

application of organic manures. J. Environ. Qual. 28: 623-632.

Jayawardane, N.S., H.D. Barrs, W.A. Muirhead, J. Blackwell, E. Murray, and G. Kirchof.

1995. Lime-slotting technique to ameliorate subsoil acidity in a clay soil. II.

Effects on Medic root growth, water extraction, and yield. Aust. J. Soil Res. 33:

443-459.

Kamprath, E.J., and C.D. Foy. 1985. Lime-fertilizer-plant interactions in acid soils. p. 91-

151. In O.P. Englestad (ed.) Fertilizer technology and use. Soil Sci. Soc. Am.,

Madison, WI.

39

Kauffman, M.D., and E.H. Gardner. 1978. Segmental liming of soil and its effect on the

growth of wheat. Agron. J. 70: 331-336.

Kingery, W.L., D.P. Delaney, J.C. Williams, and G.L. Mullins. 1994. Impact of long-

term land application of broiler litter on environmentally related soil properties. J.

Environ. Qual. 23: 139-147.

Kingery, W.L., D.P. Delaney, J.C. Williams, G.L. Mullins, and E. Vansanten. 1993.

Implications of long-term land application of poultry litter on tall fescue pastures.

J. Prod. Agric. 6: 390-395.

Kinraide, T.B. and D.R. Parker. 1987. Non-phytotoxicity of the aluminum sulfate ion,

AlSO4+. Physiologia Plantarum 71: 207-212.

Kirchof, G., N.S. Jayawardane, J. Blackwell, and E. Murray. 1995. Lime-slotting

technique to ameliorate subsoil acidity in a clay soil. I. Effects on soil pH and

physical characteristics. Aust. J. Soil Res. 33: 425-441.

Kissel, D.E., E.P. Gentzsch, and G.W. Thomas. 1971. Hydrolysis of non-exchangeable

acidity in soils during salt extractions of exchangeable acidity. Soil Sci. 111:

293&.

Kotze, W.A.G., and J. Deist. 1975. Amelioration of subsurface acidity by leaching of

surface applied amendments. A laboratory study. Agrochemophysica 7: 39-46.

Kuo, S. 1993. Calcium and phosphorus influence creeping bentgrass and annual

bluegrass growth in acidic soils. Hortscience 28: 713-716.

Lathwell, D.J., and M. Peech. 1964. Interpretation of chemical soil tests. Corn. Univ.

Agric. Exp. Sta. Bul. 995.

40

Liu, H., 2005. Aluminum resistance among seeded bermudagrasses. Hortscience 40: 221-

223.

Liu, H., J.R. Heckman, and J.A. Murphy. 1995. Screening Kentucky bluegrass for

aluminum tolerance. J. Plant Nutr. 18: 1797-1814.

Liu, H., J.R. Heckman, and J.A. Murphy. 1996a. Screening fine fescues for

aluminum tolerance. J. Plant Nutr. 19: 677-688.

Liu, H., J.R. Heckman, and J.A. Murphy. 1996b. Screening tall fescues genotypes for

aluminum tolerance. p. 149. In Agronomy Abstracts. ASA, CSSA, SSSA,

Madison, WI.

Liu, H., J.R. Heckman, and J.A. Murphy. 1997a. Greenhouse screening of turfgrasses for

aluminum tolerance. Int. Turfgrass Soc. Res. J. 8: 719-728.

Liu, H., J.R. Heckman, and J.A. Murphy. 1997b. Aluminum tolerance among genotypes

of Agrostis species. Int. Turfgrass Soc. Res. J. 8: 729-734.

Liu, H., J.R. Heckman, and J.A. Murphy. 1997c. Screening bermudagrasses for

aluminum tolerances. p. 135. In Agronomy Abstracts. ASA, CSSA, SSSA,

Madison, WI.

Malavolta, E. 1992. O gesso agricola no ambiente e na nutricao da planta: Perguntas e

respostas. II. Sem. Uso Gesso Agricultura 41-66.

Marcum, K.B., M.C. Engelke, S.J. Morton, and R.H. White. 1995. Rooting

characteristics and associated drought resistance of zoysiagrasses. Agron. J. 87:

534-538.

Matsumoto, H. 2000. Cell biology of aluminum toxicity and tolerance in higher plants.

Int. Review of Cytology 200: 1-46.

41

McKenzie, R.C., and M. Nyborg. 1984. Influence of subsoil acidity on root development

and crop growth in soils of Alberta and northeastern British Columbia. Can. J.

Soil Sci. 64: 681-697.

McCray, J.M., M.E. Sumner, D.E. Radcliffe, and R.L. Clark. 1991. Soil Ca, Al, acidity

and penetration resistance with subsoiling, lime and gypsum treatments. Soil Use

and Management 7: 193-199.

McLay, C.D.A., G.S.P. Ritchie, and W.M. Porter. 1994. Amelioration of subsurface

acidity in sandy soils in low rainfall regions. I. Response of wheat and lupins to

surface-applied gypsum and lime. Aust. J. Soil Res. 32: 835-46.

Miller, W.P. 1995. Environmental considerations in land application of by-product

gypsums. p. 183-208. In Agricultural utilization of urban and industrial-

byproducts. Am. Soc. Agron., Madison, WI.

Miller, W.P. 2005. Personal communication.

Miller, D.M., W.P. Miller, S. Dudka, and M.E. Sumner. 2000. Characterization of

industrial by-products. p. 107-126. In Land application of agricultural, industrial,

and municipal by-products. Soil Sci. Soc. Am., Madison, WI.

Mills, H.A., and J.B. Jones. 1996. Plant analysis handbook II. MicrMacro Publ., Athens,

GA.

Miyazawa, M., M.A. Pavan, and J.C. Franchini. 2002. Evaluation of plant residues on the

mobility of surface applied lime. Brazilian Archives. Biol. Tech. 45: 251-256.

Mora, M.L., P. Cartes, R. Demanet, and I.S. Cornforth. 2002. Effects of lime and gypsum

on pasture growth and composition on an acid Andisol in Chile, South America.

Comm. Soil Sci. Plant Anal. 33: 2069-2081.

42

Moschler, W.W., R.K. Stivers, D.L. Hallock, R.D. Sears, H.M. Camper, M.J. Rogers,

G.D. Jones, M.T. Carter, and F.S. McClaugherty. 1962. Lime effects on soil

reaction and base content of eleven soil types in Virginia. Virginia Agric. Exp.

Sta. Tech. Bul. 159.

Murray, J.J., and C.D. Foy. 1978. Differential tolerances of turfgrass cultivars to an acid

soil high in exchangeable aluminum. Agron. J. 70: 769-774.

Noble, A.D., M.V. Fey, and M.E. Sumner. 1988. Calcium-aluminum balance and the

growth of soybean roots in nutrient solutions. Soil Sci. Soc. Am. J. 52: 1651-

1656.

Nordstrom, D.K. 1982. The effect of sulfate on aluminum concentrations in natural

waters: Some stability relations in the system Al2O3SO3H2O at 298 K. Georchim.

Cosmochim. Acta. 46: 681-692.

O’Brien, L.A., and M.E. Sumner. 1988. Effects of phosphogypsum on leachate and soil

chemical composition. Comm. Soil Sci. Plant Anal. 19: 1319-1329.

Passos, R.R., V. Faquin, N. Curi, A.R. Evangelista, and M.R. Villa. 1997. Effect of

phosphorus sources, lime and gypsum on dry matter production and tillering of

two forage grasses in samples of an acid latossol. J. Brazilian Soc. Animal Sci.

26: 227-233.

Pavan, M.A., F.T. Bingham, and P.F. Pratt. 1982. Toxicity of aluminum to coffee in

Ultisols and Oxisols amended with CaCO3, MgCO3, and CaSO4·2H2O. Soil Sci.

Soc. Am. J. 46: 1201-1207.

Pearson, R.W., and F. Abruña, 1961. Nitrogen speeds lime movement in subsoil. Crops

and Soils 13: No. 9:23.

43

Pearson, R.W., F. Abruña, J. Vicente-Chandler. 1962. Effect of lime and nitrogen

applications on downward movement of calcium and magnesium in two tropical

soils of Puerto Rico. Soil Sci. 93: 77-82.

Pearson, R.W., J. Childs, and Z.F. Lund. 1973. Uniformity of limestone mixing in acid

subsoil as a factor in cotton root penetration. Soil Sci. Soc. Am. Proc. 37: 727-

732.

Percival, G.P., D. Josselyn, and K.C. Beason. 1955. Factors affecting the micronutrient

element content of some forages in New Hampshire. New Hamp. Agric. Exp. Sta.

93.

Perez-Escolar, R., and M.A. Lugo-Lopez. 1978. Effect of depth of lime application on

yield of two corn hybrids grown on a typical ultisol of Puerto Rico. J. Agric.

Univ. Puerto Rico 62: 203-213.

Phillips, J.M., and W.E. Sabbe. 1994. Effects of sulfur on nitrogen recovery in coastal

bermudagrass as influenced by sulfur fertilization. Comm. Soil Sci. Plant Anal.

25: 1763-1769.

Pleysier, J.L., and A.S.R. Juo. 1981. Leaching of fertilizer ions in a ultisol from the high

rainfall tropics: :Leaching through undisturbed soil columns. Soil Sci. Soc. Am. J.

45: 754-760.

Pocknee S., and M.E. Sumner. 1997. Cation and nitrogen content of organic matter

determine its soil liming potential. Soil Sci. Soc. Am. J. 61: 86-92.

Punshon, T., D.C. Adriano, and J.T. Weber. 2001. Effect of flue gas desulfurization

residue on plant establishment and soil and leachate quality. J. Environ. Qual. 30:

1071-1080.

44

Qian, Y.L., J.D. Fry, and W.S. Upham. 1997. Rooting and drought avoidance of warm-

season turfgrasses an tall fescue in Kansas. Crop Sci. 37: 905-910.

Ragland, J.L., and N.T. Coleman. 1959. The effect of soil solution aluminum and calcium

on root growth. Soil Sci. Soc. Am. Proc. 23: 355-357.

Reeve, N.G., and M.E. Sumner. 1972. Amelioration of subsoil acidity in Natal Oxisols by

leaching of surface-applied amendments. Agrochemophysica 4: 1-6.

Richards, J.E., G.C. Misener, P. Milburn, and L.P. McMillian. 1995. Incorporation of

limestone into naturally compacted subsoil during deep-ripping. Soil Till. Res. 36:

21-32.

Riley, W.R. 1991. Yield, physiology and soil moisture extraction of alfalfa and Sericea

lespedeza as affected by gypsum and irrigation. Ph.D. diss., University of

Georgia, Athens, GA.

Ritchey, K.D., D.P. Belesky, and J.J Halvorson. 2004. Soil properties and clover

establishment six years after surface application of calcium-rich by-products.

Agron. J. 96: 1531-1539.

Rosolem, C.A., and J.R. Machado. 1984. Efeitos de calagem e gessagem no producao de

algodao e na lixiviacao de bases em dois latossolos. R. Bras. Ci. Solo. 8: 103-109.

Rout, G.R., S. Samantaray, and P. Das. 2001. Aluminum toxicity in plants: A review.

Agronomie 21: 3-21.

Sartain, J.B., 1985. Effect of acidity and N-source on the growth and thatch accumulation

of tifgreen bermudagrass and on soil nutrient retention. Agron. J. 77: 33-36.

45

Sasaki. M., M. Kasai, Y. Yamamoto, and H. Matsumoto. 1995. Involvement of plasma

membrane potential in the tolerance mechanism of plant roots to aluminum

toxicity. Plant and Soil 171: 119-124.

Shim, S.R., and R.N. Carrow. 1997. Cultivation and chemical injection: Influence on soil

physical and chemical properties. Int. Turfgrass Soc. J. 8: 533-540.

Simonovicova, M., J. Huttova, I. Mistrik, B. Siroka, and L. Tamas. 2004. Root growth

inhibition by aluminum is probably caused by cell death due to peroxidase-

mediated hydrogen peroxide production. Protoplasma 224: 91-98.

Souza, D.M.G, and K.D. Ritchie. 1986. Uso do gesso no solo de Cerrado. An. Sem. Uso.

Fosfogesso Agricultura 119-144.

Souza, D.M.G., E. Lobato, K.D. Ritchey, and T.A. Rein. 1992. Resposta de culturas

anuais e leucena a gesso no Cerrado. II. Sem. Uso Gesso Agricultura 277-306.

St. John, R.A., N.E. Christians, and H.G. Taber. 2003. Supplemental calcium

applications to creeping bentgrass established on calcareous sand. Crop Sci. 43:

967-972.

Sumner, M.E. 1990. Gypsum as an ameliorant for the subsoil acidity syndrome. Florida

Inst. of Phosphate Res., Bartow, FL.

Sumner, M.E. 1995. Amelioration of subsoil acidity with minimum disturbance. p. 147-

185. In N.S. Jayawardane and B.A. Stewart (Eds.) Subsoil management

techniques. CRC Press, Inc., Boca Raton, FL.

Sumner, M.E. 2001. Global extent, development and economic impact of acid soils. p. 3-

32. In M. Farina et al. (Eds.) Proc. 5th Plant-soil interactions at low pH

symposium. Inst. For Soil, Climate and Water, Pretoria, South Africa.

46

Sumner, M.E. 2005. Soil acidity and impact on soil fertility. p. 13-24. In C.P.

Christiansen and D.R. Smart (Eds.) Proceedings of soil environment and vine

mineral nutrition symposium, San Diego. Am. Soc. Enology Viticulture, Davis,

CA.

Sumner, M.E., and E. Carter. 1988. Amelioration of subsoil acidity. Comm. Soil Sci.

Plant Anal. 19: 1309-1318.

Sumner, M.E., and K. Hylton. 1994. A diagnostic approach to solving soil fertility

problems in the tropics. p. 215-234. In J.K. Syers, and D.L. Rimmer (Eds.) Soil

science and sustainable land management in the tropics. CAB International,

Wallingford, England.

Sumner, M.E., H. Shahandeh, J. Bouton, and J.E. Hammel. 1986. Amelioration of an acid

soil profile through deep liming and surface application of gypsum. Soil Sci. Soc.

Am. J. 50: 1254-1258.

Tanaka, A., T. Tanado, K. Yamamoto, and N. Kanamura. 1987. Comparison of toxicity

to plants among Al3+, AlSO4+, and Al-F complex ions. Soil Sci. Plant Nutri. 33:

43-55.

Thomas, G.W., and W.L. Hargrove. 1984. The chemistry of soil acidity. p. 3-56. In F.

Adams (Ed.) Soil acidity and liming, 2nd ed., Am. Soc. Agron. Monograph 12,

Madison, WI.

Toma, M., M.E. Sumner, G. Weeks, and M Saigusa. 1999. Long-term effects of gypsum

on crop yield and subsoil chemical properties. Soil Sci. Soc. Am. J. 63: 891-895.

Vasconcelos, S.S., J. Jacob-Neto, and R.O.P. Rosiello. 2002. Differential root responses

to aluminum stress among Brazilian rice genotypes. J. Plant Nutr. 25: 655-669.

47

von Uexküll, H.R., and E. Mutert. 1995. Global extent, development and impact of acid

soils. Plant and Soil 171: 1-15.

Wagatsuma, T., S. Ishikawa, H. Obata, K. Tawaraya, and S. Katohda. 1995. Plasma

membrane of younger and outer cells is the primary specific site for aluminum

toxicity in roots. Plant and Soil 171: 105-112.

Weir, C.C. 1975. Effect of lime and nitrogen application on citrus yields and on the

downward movement of calcium and magnesium in a soil. Trop. Agric. (Trinidad)

51: 230-234.

Whalen, J.K., C. Chaing, G.W. Clayton, and J.P. Carefoot. 2000. Cattle manure

amendments can increase the pH of acid soils. Soil Sci. Soc. Am. J. 64: 962-966.

Wright, R.J., J.L. Hern, V.C. Baligar, and O.L. Bennett. 1985. The effect of surface

applied soil amendments on barley root growth in an acid subsoil. Comm. Soil

Sci. Plant Anal. 16: 179-192.

Ziglio, C.M., M. Miyazawa, and M.A. Pavan. 1999. Organic and inorganic forms

of calcium mobilization in soil. Brazilian Arch. Biol. Tech. 42: 257-262.

48

CHAPTER 3

A SIMPLE LOGISTIC MODEL TO DESCRIBE ROOT-GROWTH EFFECTS OF

CALCIUM AS GYPSUM ON TALL FESCUE VARIETIES IN

ALUMINUM-RICH SOLUTIONS

______________________________________________________ Kruse, J.S., W.P. Miller, and M.E. Sumner, E.G. To be submitted to Soil Science Society

of America Journal.

49

Abstract

A study was conducted to quantify root-growth effects of Ca2+ and SO4

2- (as

gypsum) in hydroponic solutions containing root-damaging aluminum on three varieties

(K31, Rendition, and Dynasty) of tall fescue (Festuca arundinacea) seedlings. An

incomplete factorial experiment was performed in a growth chamber using 1-L pots

containing levels of Al from 0 to74 μM with CaSO4 levels from 0 to10 mM. Tall fescue

seedlings were grown for 7 d, harvested, air-dried, scanned and weighed for treatment

comparisons. Total root length as calculated from digital scanning was highly correlated

with root mass because of root-diameter uniformity in hydroponic solutions. Significant

differences existed between varieties in Al-only solutions at low Al concentrations. All

varieties showed less than 15% relative root growth in 37 μM Al and higher

concentrations in the absence of Ca. An increase in Ca2+ and SO42- at a given

concentration of Al provided a protective effect from Al toxicity, resulting in increased

root growth. The greatest relative root response to increased Ca levels occurred at Al

levels of 37 and 74 μM Al. Relative root growth increased from about 30% to >80% at 37

μM Al as Ca increased from 2.5 to 10 mM. A simple logistic model adequately described

the effects of Al and Ca on root growth (R2 = 0.86, 0.95, and 0.96 for the three varieties,

with an RMSE of 19.061, 8.119, and 9.553, respectively) and may be useful as a

predictor of the quantity of gypsum needed to overcome Al toxicity in acid subsoils.

50

Introduction

Acid soils inhibit plant root growth primarily because of toxic levels of Al

(Adams and Lund, 1966; Foy and Fleming, 1978; Pavan and Bingham, 1982; Blamey et

al., 1983; M.E. Sumner, 1995; Kim et al., 2001; Sumner and Yamada, 2002). Monocot

plant species demonstrate a range of responses to aluminum tolerance, with differences

also found between cultivars of the same species (Wenzl et al., 2002; Giaveno and

Miranda, 2000; Gallardo et al., 1999). Turfgrass species have different levels of tolerance

to Al, and also show differences between varieties of the same species (Liu et al., 1996;

Liu et al., 1997; Liu 2005). Vasconcelos et al. (2002) reported differences in Al-tolerance

in rice varieties (Oryza sativa L.) grown in solution. Kim et al. (2001) studied Brazilian

wheat (Triticum aestivum L. em Thell), triticale (X Triticosecale Wittmack), and rye

(Secale cereale L.) and found greater Al tolerance in newer varieties of triticale, but the

greatest Al tolerance in wheat in a 50-year old variety.

The presence of Ca is known to affect root growth response to plants growing in

acidic, Al-rich hydroponic solutions (Pintro and Taylor, 2004; Ryan and Kochian, 1993;

Noble et al., 1988; Alva et al., 1986a,b). The addition of Ca to a solution culture increases

the ionic strength of the solution and affects the activity of toxic Al species (Noble et al.,

1988). The interactive effects of Ca and Al (Simpson et al, 1977) affect the uptake of Ca

in Al-sensitive plants (Ryan and Kochian, 1993), and other mechanisms may exist.

The logistic model has been useful in modeling the response of living organisms to

increasingly toxic doses of a substrate, such as pesticide residuals on the growth of soil

mesofauna (Folker-Hansen et al., 1996), or Collembola populations feeding on metal-

polluted fungi (Bengtsson et al., 1983) or atrazine (Badejo and Van Straalen, 1992).

51

Several researchers have successfully utilized the logistic model to investigate crop

response to environmental factors such as ambient temperature and soil moisture

(Panozzo and Eagles, 1999), and rates of nitrogen application on forage grasses under

various conditions (Overman and Brock, 2003a,b; Overman and Scholtz 2003a,b;

Overman et al., 1994). Man and Cheung (2002) used a simple logistic model to interpret

results obtained for a study of 137Cs sorption to soils from Hong Kong. The model

successfully described certain trends but failed to describe the quantitative aspects of the

findings. Agreement between theory and experimental data was good for small soil

masses, but not large masses. Parameters were derived from the data using a simple

logistic model. A modified simple-logistic model (Pinheiro and Bates, 2000) was deemed

appropriate for use because it has a sigmoidal shape that was believed to fit the growth

curve of roots as a function of Al and CaSO4, and because of interpretability of model

parameters.

Computer-aided methods of determining root surface areas have been established

and tested, resulting in standard protocols for determining specific root length (m g-1),

which measures a plant’s capacity to acquire water and nutrients per unit carbon

expenditure (Bouma et al., 2000). Root biomass (mg) is cited as a measure of the costs

associated with root development and root maintenance (Bouma et al., 1996). Derner et

al. (2001) investigated root growth response of tall-grass prairie species for 8 weeks in

soil-filled pots and found the majority of roots widths were between 0.3 and 0.6 mm, but

the greatest proportional response to increased ambient CO2 was in fine roots (0.0 to 0.3

mm diameter).

52

The objectives of this study were to 1) determine tall fescue seedling root

response to various concentrations of Al and Ca in solution, and 2) model the root growth

to determine at what concentration of Al in solution did root growth differ significantly

from optimal and for a given concentration of Al, how much Ca as CaSO4 was required

to offset 80% of the toxic effects of Al on roots.

Materials and Methods

Hydroponic experiments

Two hydroponic studies were conducted utilizing an incomplete factorial design.

In both experiments, 1-L pots (Ziploc®) were fitted with stainless steel wire mesh capable

of suspending seed at the mouth of the pot and covered with a layer of cheesecloth. The

surface area of the wire mesh was 156.25 cm2. Three varieties of turf-type tall fescue

(Festuca arundinacea) – ‘Rendition’, ‘Dynasty’, and ‘K31’, were selected based on their

history and marketing emphasis. ‘K31’ is the oldest commercial variety still marketed,

and is the genetic base of turf-type tall fescues grown and marketed in the United States

for consumer lawns. ‘Dynasty’ is marketed by a large seed and lawn-fertilizer company

as drought resistant and adapted to high heat, requiring less watering. ‘Rendition’ is sold

as a low-cost, widely available turf-type tall fescue with typical drought resistance for

cool season grasses grown in the southern U.S. Each pot received 1.60 g of a single seed

variety. After hand-distribution of the seed on the cheesecloth, pots were filled with

deionized water to the point that the cheesecloth became saturated, and placed randomly

in a growth chamber under fluorescent lights. Light intensity was measured to be 95 W

m-2. The chamber temperature was maintained at 25o C, with lights kept on continuously.

53

Pot water levels were maintained on a daily basis to preserve cheesecloth saturation. As

soon as roots became visible through the wire mesh - after 5 d - the deionized water was

removed and replaced with an appropriate hydroponic solution.

In the first experiment investigating varietal root-sensitivity to Al toxicity, the

three fescue varieties were grown in AlCl3·6H2O solutions of 0, 3.7, 7.4, 14.8, 37.0, or

74.1 μmol L-1. All solutions were brought to pH 4.5, in order to simulate an acid soil

solution, with 0.01 M HCl dripped under constant stirring during measurement.

Treatments were grown for 7 d, at the end of which solution subsamples were preserved

in 20-mL scintillation vials, and roots were air-dried for 24 h. Roots were cut from the

stainless steel mesh and measured on Mettler AM50 scales (Mettler Instrument Corp.,

Hightstown, NJ) to the nearest 0.1 mg.

In the second experiment, the fescue was grown in solutions containing various

levels of Al from AlCl3·6H2O and/or Ca from CaSO4·2H2O. Varying numbers of

experimental replicates were assigned to each treatment combination to maximize the

range of levels of Al and CaSO4 evaluated. The combined Al and CaSO4 solutions were

created by diluting saturated CaSO4·2H2O (14 mM) with deionized water to the

appropriate molarity, addition of 6.67 mM Al from a buret as the solution was stirred,

followed by addition of 0.01 M HCl from a buret under constant stirring until a stable pH

reading of 4.5 was achieved. No other nutrients were added to the solution. Treatments

were grown for 7 d after solution replacement, with periodic inputs of deionized water to

replace water lost through evapotranspiration. The pH of each solution was adjusted

periodically to maintain a pH of 4.5. After termination of the experiment, roots were air-

54

dried for 24 h, then cut and weighed. Solution subsamples were saved in 20-mL

scintillation vials. Roots were weighed and placed in ceramic crucibles, then ashed in a

Thermolyne muffle furnace (Sybron Corp., Dubuque, IA) at 500o C for 8 hr. After

cooling to room temperature, 1mL of 6M HCl was added to dissolve the ash. Extracts

were filtered through Whatman #1 filter paper, with crucibles rinsed three times with

deionized water, pouring the rinsate into the filter paper. Extracts were brought to volume

in 50-mL volumetric flasks with deionized water, then analyzed for Ca on a Perkin-Elmer

Aanalyst 200 Flame Atomic Absorption Spectrometer (Perkin-Elmer, Inc., Wellesley,

MA) and Al on a Perkin-Elmer Elan 9000 Intermittently Coupled Plasma-Mass

Spectrometer.

Statistical analysis

Statistical analysis was performed using the 2-Way ANOVA procedure (α = 0.05)

utilizing SAS (SAS Institute, 1991) in order to assess the presence or absence of an

interaction between the tall fescue variety and Al factors in the model. Moreover, contrast

analyses were conducted assuming that optimal root growth occurred at Al=0 level, for

each of the three varieties and each level of Al. Dunnett’s Method was used in order to

control the family-wise error rate (the combined error rate of multiple tests, in this case

multiple contrasts) for all contrasts, which avoids inflated Type I error rates (where the

null hypothesis is incorrectly rejected).

CaSO4 and Al interactions were investigated using a factorial ANOVA (GLM),

with differing numbers of observations at each Variety/CaSO4/Al level, creating an

unbalanced design. After analysis of the three- and two-way interactions, separate 2-Way

55

ANOVAs were run for differing varieties. Profile plots were generated to display the

nature of any three-and two-way interactions.

In order to estimate a CaSO4 threshold concentration required for 80% root

growth recovery for a given level of Al, a reparameterization of the simple logistic model

(Pinheiro and Bates, 2000) was employed. A re-parameterization of the classical simple

logistic model was developed to estimate root growth recovery at the 80% of the optimal

level. The model took the following form (see Figure 3.1 and Table 3.1).

The simple logistic model was used to evaluate three varieties of tall fescue

separately. In the case where Al is absent (Al = 0), the equation (Φ1i – βi*γiCa(k))*D1

models the initial curve and horizontal asymptote of the predicted Al = 0 growth curve

for each variety. In the case where Al is present, the remaining portion of the equation

models the predicted growth curves for each variety as a function of Ca and Al. D1 and

D2 act as toggle switches to activate or deactivate appropriate portions of the equation,

depending on the presence or absence of Al in solution. Optimal levels of root growth

were then estimated by interpreting the Φ1 parameter of the model for each tall fescue

variety. These estimates are to be interpreted as the optimal root growth of a given variety

at a given level of Al. Predicted values of Ca that would achieve 80% of optimal root

growth were obtained by interpreting the Φ3 parameter for each tall fescue variety.

Root measurement evaluation

In addition to root mass measurements, image analysis was used to determine root

areas of harvested roots. A method to scan and digitize root length was created using

ARC INFO (Esri, Inc., Redlands, CA) to count root surface area.

56

Scanning calibration was determined by using segments of white nylon

monofilament similar in diameter to root thickness of tall fescue seedlings. Monofilament

sections were cut to 10-cm lengths, placed under a microscope at 43x and their diameter

was measured. The segments were placed on a flat black background and scanned at 300

dpi resolution to create a bitmap image. The digital-image data of black and white pixels

was analyzed by ARC INFO, which produced separate counts of black and white pixels.

The total surface area of the monofilament segments was divided by the measured

diameter of the segments to obtain the scanned total segment length. The total length as

determined by the scanning procedure was compared to the total length of the six

segments. This procedure was repeated to evaluate the coefficient of variation of the

method. The same background was used in all procedures, including those measuring

roots. The predicted pixel count was compared to actual pixel counts of monofilament

and goodness of fit was determined from the linear regression. The regression equation

resulting from the comparison was used as a correction factor after scanning roots to

obtain a corrected pixel count.

After growing 36 experimental units of ‘Rendition’ tall fescue seedlings in

various levels of CaSO4 and Al according to the method previously described, roots were

cut, air-dried for 24 h, and weighed to the nearest 0.1 mg. Subsamples of roots were

weighed, placed under a microscope at 43x to measure root width, then computer-

scanned for surface area. The actual scanned surface areas of the subsamples were used

to create estimated surface areas of the whole root mass by proportion.

57

Results and Discussion

Root measurement evaluation

Repeated pixel counts of the monofilament segments in ARC INFO produced

consistent values. The total monofilament length as determined by the scanning

procedure was compared to the total length of the six segments as measured under the

microscope, resulting in an average difference of 1.75%. Replicate counts of several trials

of segments of the same total length produced a coefficient of variation (CV) of 6.59%.

Actual pixel counts of monofilament segments compared to predicted pixel counts based

on 300 dpi resolution gave a large degree of goodness of fit (r2=0.995), and an equation

for correcting actual root pixel counts (y = 0.7078x – 2728.3) unique to the background

plate used. Because the same background was used for all calibration and root

measurements, this factor could be incorporated into the comparison.

Root mass was highly correlated with the calculated root length based on root

scanning (Figure 3.2). Based on the observation that roots growing from seedlings in the

hydroponic solutions were highly consistent in root diameter, regardless of CaSO4 or Al

concentrations in solution (e.g. Figure 3.3) and only differed significantly in length, root

mass was determined to be an adequate measurement of root growth response to the

treatments.

Root response to Aluminum

All three fescue varieties demonstrated similar sensitivity to increasing levels of

Al in solution (Figure 3.4). Initial results of the 2-way ANOVA showed that Al and

variety were statistically significant predictors of root mass. Because aluminum and

58

variety had a significant interaction, interpretation of the main effects was deemed

inappropriate. Contrast analysis of Al and variety was performed to ascertain the level of

Al at which significant reduction in root mass occurred, with the result that comparisons

of Al were statistically significant for all varieties at 7.4 μmol L-1. Despite initial

assumptions that turf varieties would differ greatly in Al sensitivity, all varieties showed

less than 15% relative root growth at 37.0 μmol Al (1.0 mg L-1), which held true for Al

concentrations greater than 37.0. Relative root mass dropped off precipitously from 3.7 to

37.0 μmol Al, with Dynasty showing somewhat more sensitivity to Al than the other two

varieties in the range of 7.4 to 37 μmol Al.

Root response to Ca and Al in solution

The addition of CaSO4 to Al-containing solutions had a significant impact on root

growth. Increasing Ca (as CaSO4·2H2O) increased relative root growth for all three

varieties at 7.4 μmol Al and higher (Figure 3.5). Greater CaSO4 levels resulted in

significantly greater relative root growth especially at 14.8 μmol Al and greater

concentration. For example, at 37.0 μmol Al (about 1 mg L-1), 2500 μmol CaSO4

increased relative root growth from less than 15% to about 30%. At 5000 μmol CaSO4,

relative root growth at the same level of Al increased 60 to70% relative to the non-

treated, and at 10,000 μmol CaSO4 (about 70% saturated) relative root growth was 80%

of optimal. Significant differences were observed between varieties at smaller levels of

CaSO4 and greater levels of Al. ‘K31’ had slight but significantly more relative root

growth compared to the other varieties at 37.0 μmol Al at 2500 and 5000, but not 10,000

59

μmol CaSO4. At the greatest level of Al measured (74.1 μmol Al or 2 mg L-1) and at

2500 μmol CaSO4, ‘Rendition’ was superior to ‘K31’ which was superior to ‘Dynasty’.

When CaSO4 was increased to 5000 μmol, ‘K31’ was significantly superior to

‘Rendition’ and ‘Dynasty’, but all three showed greater relative root growth compared to

the smaller CaSO4 level. At the 10,000 μmol CaSO4 level and largest Al level, Dynasty

demonstrated the greatest root recovery, increasing from roughly 20% to 80% relative

root growth. This may imply ‘Dynasty’ has a greater Ca requirement than the other

species tested to overcome Al toxicity. Huang et al. (1992) demonstrated Ca uptake in an

Al-sensitive cultivar of wheat (Triticum aestivum L.Thell) was more inhibited than in a

tolerant cultivar.

The factorial ANOVA analysis demonstrated no Al/CaSO4/Variety 3-way

interaction. A significant 2-way interaction existed between CaSO4 and Al, probably

because of the effect of ionic strength and shifts in Al speciation (Noble et al., 1988, Alva

et al., 1986a). Although there was a significant interaction between Al and Variety (P>F

= 0.0228), no such interaction occurred with CaSO4 and Variety (P>F = 0.9338),

suggesting much less root sensitivity to changes in Ca solution concentration compared to

Al. This makes sense in the soil environment where Ca uptake occurs by mass flow and

Ca concentrations are typically orders of magnitude larger compared to soil-solution Al.

Simple logistic model

The reparameterized simple logistic model demonstrated slightly varying levels of

fitness for each of the three varieties (Table 3.2). The model fit statistic should be

interpreted as the ratio of the variation explained by the model over the total variation in

60

the data. Residual plots were generated and inspection of the plots suggested no

violations of assumptions for any of the models. Overall goodness of fit was satisfactory

for the model to the data, with an R2 = 0.86, 0.95, and 0.96 for ‘Rendition’, ‘K31’, and

‘Dynasty’, respectively.

Predicted optimal levels of root growth were calculated for each level of Al based

on the Φ1 parameter for each variety and are reported in Table 3.3. There were significant

differences in optimal root mass for each variety, indicating differences in overall vigor

between cultivars. At some levels of Ca, the varieties had greater optimal root growth in

small concentrations of Al than in Al=0.This phenomenon has been repeatedly observed

in the literature (Pintro and Taylor, 2004; Ryan and Kochian, 1993; Noble et al., 1988).

Parameter estimates were generated for each variety, including the molar concentration of

Ca necessary to achieve 80% of optimal root growth.

Model regression curves fit the data well at all the levels of Al and CaSO4.

Graphic displays of selected model-fit exercises are presented in Figure 3.6. In the

absence of Al, an initial response to added CaSO4 is observed. As Al concentration is

increased, recovery to an 80% of optimal root growth occurs at greater rates of CaSO4,

creating an increasingly sigmoidal shape and shifting the curve to the right. At the largest

level of Al (74.1 μmol), root growth does not recover to the 80% level until the largest

CaSO4 level is present (10,000 μmol).

Model parameters suggest ‘Rendition’ required more CaSO4 in solution to

achieve 80% root growth recovery than the other varieties at all levels tested. ‘Rendition’

also had the largest predicted optimal level of root growth, suggesting vigorous rooting

varieties may require greater levels of Ca. ‘K31’ required greater levels of CaSO4

61

compared to ‘Dynasty’ at low Al, but at greater levels of Al the opposite was true.

Although solution Al in soils is difficult to measure, most acid soils in the Southeastern

U.S. probably have solution Al that falls into the smaller range of Al tested in this

experiment (Sposito and Mattigod, 1979; Sumner et al., 1986).

Gypsum (CaSO4·2H2O) offers the potential to ameliorate subsoil acidity under

established turfgrass. It is much more soluble than limestone (2.41 g L-1 at 25o C for

CaSO4·2H2O versus 0.0153 g L-1 at 25o C for CaCO3) and can move through the soil

profile with rainfall and irrigation water. Extensive research has been conducted on soil

chemical changes induced by surface applications of gypsum and resulting improvements

in various crop yields (Shainberg et al., 1989; Sumner, 1995; Ritchey et al., 2004). The

simple logistic model fit to this hydroponic investigation of tall fescue indicates gypsum

may elicit a positive root-growth response in the field through a reduction in soil-

exchangeable acidity.

Conclusions

Acid soils predominate in the southeast and in many parts of the world, and

research indicates that in some crops and some soils, gypsum offers an effective

ameliorant of soil – especially subsoil – acidity (Shainberg et al., 1989; Sumner, 1995;

Farina et al., 2000). A major limitation to the use of gypsum as an ameliorant of soil

acidity is accurately determining the quantity of material necessary for application.

Although this study was performed under hydroponic conditions, and cannot take into

account exchangeable and reserve acidity, the statistical modeling approach taken does

offer clues as to the quantity of Ca as gypsum required to overcome toxic Al in solution

62

under acid conditions. Some varieties of tall fescue are more vigorous in rooting than

others but require more CaSO4 to recover from the effects of Al toxicity. As Al in

solution increases, more CaSO4 is required for root recovery. CaSO4 levels required to

overcome Al are orders of magnitude higher than Al, suggesting severe sensitivity to

slight increases in soil-solution Al. A simple logistic model adequately describes the

relationship between CaSO4 and root growth for a given level of Al, and could potentially

be useful as a component in a more sophisticated model predicting soil-solution Al as a

component of exchangeable and reserve acidity. A mechanistic approach to

understanding tall fescue root growth response to Al and CaSO4 in solution is discussed

in the following chapter.

63

References Cited

Adams, F., and Z.F. Lund. 1966. Effect of chemical activity of soil solution aluminum on

cotton root penetration of acid subsoils. Soil Sci. 101:193-198.

Alva, A.K., D.G. Edwards, C.J. Asher, and F.P.C. Blamey. 1986a. Effects of

phosphorus/aluminum molar ratio and calcium concentration on plant response to

aluminum toxicity. Soil Sci. Soc. Am. J. 50: 133-137.

Alva, A.K., D.G. Edwards, C.J. Asher, and F.P.C. Blamey. 1986b. Relationship between

root length of soybean and calculated activities of aluminum monomers in

nutrient solution. Soil Sci. Soc. Am. J. 50: 959-962.

Badejo, M.A., and N.M. Van Straalen. 1992. Effects of atrazine on growth and

reproduction of Orchesella cincta (Collembola). Pedobiologia 36: 221-230.

Bengtsson, G., T. Gunnarsson, and S. Rundgren. 1983. Growth changes caused by metal

uptake in a population of Onychiurus armatus (Collembola) feeding on metal

polluted fungi. Oikos 40: 216-225.

Blamey, F.P.C., D.G. Edwards, and C.J. Asher. 1983. Effects of aluminum, OH:Al and

P:Al molar ratios, and ionic strength on soybean root elongation in solution

culture. Soil Sci. 136: 197-207.

Bouma, T.J., K.L. Nelson, and B. Koutstall. 2000. Sample preparation and scanning

protocol for computerized analysis of root length and diameter. Plant Soil 218:

195-196.

Bouma, T.J., A.G.M. Broekhuysen, and B.W. Veen. 1996. Analysis of root respiration of

Solanum tuberosum as related to root growth, ion uptake and maintenance of biomass.

Plant Phys. Biochem. 34: 795-806.

64

Derner, J.D., H.W. Polley, H.B. Johnson, and C.R. Tischler. 2001. Root system response

of C4 grass seedlings to CO2 and soil water. Plant Soil 231: 97-104.

Farina, M..P.W., P. Channon, and G.R. Thibaud. 2000. A comparison of strategies for

ameliorating subsoil acidity: I. Long-term growth effects. Soil Sci. Soc. Am. J.

64: 646-651.

Folker-Hansen. P., P.H. Krogh, and M. Holmstrup. 1996. Effect of dimethoate on body

growth of representatives of the soil living mesofauna. Ecotox. Environ. Safety

33: 207-216.

Foy, C.D., and A.L. Fleming. 1978. The physiology of crop tolerance to excess available

aluminum and manganese in acid soils. P. 301-328. In G.A. Jung (ed.) Crop

tolerance to sub-optimal land conditions. Spec. Pub. 32. ASA, Madison, WI.

Gallardo, F., F. Borie, M. Alvear, and E. vonBaer. 1999. Evaluation of aluminum

tolerance of three barley cultivars by two short-term screening methods and field

experiments. Soil Sci. and Plant Nutr. 45: 713-719.

Giaveno, C.D., and J.B. Miranda. 2000. Rapid screening for aluminum tolerance in maize

(Zea mays L.). Genetics and Molecular Biology 23: 847-850.

Huang, J.W., D.L. Grunes, and L.V. Kochian. 1992a. Aluminum effects on the kinetics of

calcium uptake into the cells of wheat root apex: Quantitation of calcium fluxes

using a calcium selective vibrating microelectrode. Planta 188: 414-421.

Kim, B.Y., A.C. Baier, D.J. Somers, and J.P. Gustafson. 2001. Aluminum tolerance in

triticale, wheat, and rye. Euphytica 120: 329-337.

Liu, H. 2005. Aluminum resistance among seeded bermudagrasses. Hortscience 40: 221-

223.

65

Liu, H., J.R. Heckman, and J.A. Murphy. 1996. Screening fine fescues for aluminum

tolerance. J. of Plant Nutr. 19: 677-688.

Liu, H., J.R. Heckman, and J.A. Murphy. 1997. Greenhouse screening of turfgrasses for

aluminum tolerance. Int. Turfgrass Soc. Res. J. 8: 719-728.

Man, C.K., and Y.P. Cheung. 2002. Determination and analysis of sorption of 137Cs to

soils in a Hong Kong reservoir. Environmental Pollution 117: 357-364.

Noble, A.D., M.V. Fey, and M.E. Sumner. 1988. Calcium-aluminum balance and the

growth of soybean roots in nutrient solutions. Soil Sci. Soc. Am. J. 52: 1651-

1656.

Overman, A.R., and K.H. Brock. 2003a. Model comparison of coastal bermudagrass and

Pensacola bahiagrass response to applied nitrogen. Comm. Soil Sci. Plant Anal.

34: 2163-2176.

Overman, A.R., and K.H. Brock. 2003b. Model analysis of corn response to applied

nitrogen and tillage. Comm. Soil Sci. Plant Anal. 34: 2177-2191.

Overman, A.R., and R.V. Scholtz. 2003a. Model analysis of response of Pensacola

bahiagrass to applied nitrogen on two soils. Comm. Soil Sci. Plant Anal. 34:

1465-1479.

Overman, A.R., and R.V. Scholtz. 2003b. Model comparison for three forage grasses at

the same location. Comm. Soil Sci. Plant Anal. 34: 735-745.

Overman, A.R., S.R. Wilkinson, and D.M. Wilson. 1994. An extended model of forage

grass response to applied nitrogen. Agron. J. 86: 617-620.

66

Panozzo, J.F., and H.A. Eagles. 1999. Rate and duration of grain filling and grain

nitrogen accumulation of wheat cultivars grown in different environments. Aust.

J. Agric. Res. 50: 1007-1015.

Pavan, M.A., and F.T. Bingham. 1982. Toxicity of aluminum to coffee seedlings grown

in nutrient solution. Soil Sci. Soc. Am. J. 46: 993-997.

Pinheiro, J.C., and D.M. Bates. 2000. Mixed effect models in S and S-PLUS. p. 519-525.

Springer, New York.

Pintro, J.C., and G.J. Taylor. 2004. Effects of aluminum toxicity on wheat plants

cultivated under conditions of varying ionic strength. J. Plant Nutr. 27: 907-919.

Ritchey, K.D., D.P. Belesky, and J.J. Halvorson. 2004. Soil properties and clover

establishment six years after surface application of calcium-rich by-products.

Agron. J. 96: 1531-1539.

Ryan, P.R., and L.V. Kochian. 1993. Interaction between aluminum toxicity and calcium

uptake at the root apex in near-isogenic lines of wheat (Triticum aestivum L.)

differing in aluminum tolerance. Plant Physiol. 102: 975-982.

SAS Institute. 1991. SAS/STAT users guide. Ver. 7. 4th ed. SAS Institute, Cary, NC.

Shainberg, I., M.E. Sumner, W.P. Miller, M.P.W. Farina, M.A. Pavan, and M.V. Fey.

1989. Use of gypsum on soils: A review. p. 1-111. In Advances in Soil Science,

Vol. 9. Springer-Verlag, New York.

Simpson, J.R., A. Pinkerton, and T. Lazdovskis. 1977. Effects of subsoil calcium on the

root growth of some Lucerne genotypes (Medicago sativa L.) in acidic soil

profiles. Aust. J. Agric. Res. 28: 629-638.

67

Sposito, G., and S.V. Mattigod. 1979. Geochem: A computer program for the calculation

of chemical equilibria in soil solutions and other natural water systems. Kearney

Foundation of Soil Science, Univ. of California, Riverside, CA.

Sumner, M.E. 1995. Amelioration of subsoil acidity with minimum disturbance. p. 147-

185. In Jayawardane and Stewart (eds.) Subsoil management techniques. Adv. In

Soil Sci. CRC Press, Boca Raton, FL.

Sumner, M.E., H. Shahandeh, J. Bouton, and J.E. Hammel. 1986. Amelioration of an acid

soil profile through deep liming and surface application of gypsum. Soil Sci. Soc.

Am. J. 50: 1254-1258.

Sumner, M.E., and T. Yamada. 2002. Farming with acidity. Commun. Soil Sci. Plant

Anal. 33: 2467-2496.

Vasconcelos, S.S., J. Jacob-Neto, and R.O.P. Rossiello. 2002. Differential root responses

to aluminum stress among Brazilian genotypes. J. Plant Nutr. 25: 655-669.

Wenzl, P., J.E. Mayer, and I.M. Rao. 2002. Aluminum stress inhibits accumulation of

phosphorus in root apices of aluminum-sensitive but not aluminum-resistant

Brachiaria cultivars. J. Plant Nutr. 25: 1821-1828.

68

Table 3.1. Definitions of symbols in the modified Simple Logistic Model used to

characterize tall fescue root growth in response to Al and CaSO4 in solution, and predict CaSO4 levels necessary to overcome Al toxicity.

Symbol Definition

i 1 to 3 fescue varieties j 1 to m(i) levels of aluminum k 1 to 6 levels of calcium l 1 to n(i) replicates of each variety of turf m(i) Where m(1) = 8 for Rendition, m(2) = 5 for K31, and m(3) = 5 for Dynasty n(i) Where n(1) = 90 for Rendition, n(2) = 48 for K31, and n(3) = 48 for Dynasty Φ1ij A parameter determining the horizontal asymptote of the predicted growth

curves for each variety of fescue at each level of aluminum Φ2ij A parameter determining the value of calcium at which ½ of the horizontal

asymptote is achieved for each variety of fescue at each level of aluminum Φ3ij A parameter determining the value of calcium at which 0.80 of the horizontal

asymptote is achieved for each variety of fescue at each level of aluminum yijkl The mean predicted root mass (mg) for the ith fescue variety, measured at the

jth level of aluminum, at the kth level of calcium βi A parameter for the aluminum = 0 curve for the ith fescue variety determining

the initial intercept of that curve γi A parameter for the aluminum = 0 curve for the ith fescue variety Ca(k) The value of the kth level of calcium D1 Dummy variable coded 1 if Al = 0, 0 otherwise D2 Dummy variable coded 1 if Al > 0, 0 otherwise εijkl Residual error assumed to be N(0,σ2)

69

Table 3.2. Model fit statistics for simple logistic model per variety.

Model Error Variety ----------DF----------

SSE MSE Root MSE Model Fit Statistic

‘Rendition’ 19 62 22525.8 363.3 19.1 0.8589 ‘K31’ 12 31 2084.2 67.2 8.2 0.9495 ‘Dynasty’ 12 31 2828.9 91.3 9.6 0.9586

70

Table 3.3. Predicted optimal levels and 80% levels of fescue root growth (mg), and

CaSO4 estimates (μM) for 80% of optimal root growth by level of aluminum and tall fescue variety.

Predicted optimal

level

80% of optimal

level

-------------------- Al (μM) -------------------- 3.7 7.4 14.8 37.0 74.1 Variety

--Root mass (mg)-- -------------------- Ca (μM) -------------------- ‘Rendition’ 119.9 95.9 739 1438 3710 6285 8938 ‘K31’ 89.8 71.8 * 731 1581 4328 6076 ‘Dynasty’ 101.2 81.0 * 470 1430 5504 7153 (*) Data not collected for this variety at this level of Al.

71

Φ1j

0.721(Φ3j – Φ2j)

Φ2j – Ca(k)( ) 1 + expyijkl = (Φ1-β*γCa(k))*D1 + * D2 + εjkl

( )

Φ1

Roo

t gro

wth

Calcium

Optimal root growth

Φ3

Φ2

Figure 3.1. The simple logistic model showing the parameters Φ1, the horizontal asymptote as x → ∞; Φ2 the value of x for which y = Φ1*0.5; and Φ3, the value of x for which y = Φ1*0.8.

72

Root mass (g)0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16

Roo

t len

gth

(cm

)

0

500

1000

1500

2000

2500

3000

Coefficients:b[0]=-34.2725b[1]=15233.8856r ²=0.907n=36

Figure 3.2. Comparison of calculated root length to root mass of tall fescue seedlings grown in hydroponic solution containing various levels of CaSO4 and Al.

73

Figure 3.3. Root scan for digital analysis of root surface area using ARC INFO (left), and fescue root growth in 0 Al and 10,000 μM CaSO4 (right) showing uniformity of root diameters in solution.

74

Total Al (μM)

0 20 40 60 80

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0

1.2

'Rendition''K31''Dynasty'

Figure 3.4. Relative tall fescue root growth sensitivity to total Al in an acid (pH 4.5) hydroponic solution. Error bars indicate standard deviation of mean values.

75

10,000 μM CaSO4

Total Al (μM)

0 20 40 60 80

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0

'Rendition''K31''Dynasty'

5000 μM CaSO4

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0'Rendition''K31''Dynasty'

2500μM CaSO4

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0'Rendition''K31''Dynasty'

Figure 3.5. Effect of increasing CaSO4 in solution on relative root growth in the presence of Al. Error bars indicate standard deviation of mean values.

76

Al = 0 μM

0 2000 4000 6000 8000 10000R

oot m

ass (

mg)

0

20

40

60

80

100

Al = 37.0 μM

CaSO4 (μM)

0 2000 4000 6000 8000 10000

Roo

t mas

s (m

g)

0

20

40

60

80

100

Al = 7.4 μM

0 2000 4000 6000 8000 10000

0

20

40

60

80

100

Al = 74.1 μM

CaSO4 (μM)

0 2000 4000 6000 8000 100000

20

40

60

80

100

Figure 3.6. An example of fitting the simple logistic model: K31 root response to

increasing levels of CaSO4 at various fixed levels of Al. Error bars indicate standard deviation of mean values.

77

CHAPTER 4

MECHANISTIC ANALYSIS OF TALL FESCUE ROOT GROWTH IN

ALUMINUM SOLUTIONS AFFECTED BY ADDITION OF CaSO4

______________________________________________________ Kruse, J.S., W.P. Miller, and M.E. Sumner, E.G. To be submitted to Soil Science Society

of America Journal.

78

Abstract

An experiment was conducted in hydroponic solutions to evaluate

mechanistic/computer-aided explanations for tall fescue seedling root response to Al in

the presence of various levels of Ca as CaSO4·2H2O. An incomplete factorial experiment

was performed in a growth chamber using 1-L pots containing levels of Al from 0 to 74

μM with Ca levels from 0 to10 mM, at a solution pH of 4.5. Three varieties of tall fescue

seedlings were grown for 7 d, harvested, air-dried, scanned and weighed for treatment

comparisons. Roots were analyzed for Al concentration. Relative root growth was

reduced to less than 20% in total Al concentrations of 37 μM in the absence of Ca. The

addition of Ca as CaSO4 significantly increased relative root growth in the presence of Al

concentrations exceeding 7.4 μM. Computer-aided analysis of Al activity and speciation

using VMINTEQ demonstrated the addition of CaSO4 in Al solution substantially

reduced Al activity and the quantity of Al3+ as a percentage of total Al. The percentage of

Al as AlSO4+, a relatively non-toxic form of ionic Al, increased and ionic strength

increased. Relative root growth followed an exponential decay equation as a function of

monomeric-Al activity in solution (R2=0.76). The Calcium-aluminum balance (CAB)

equation did not adequately predict relative root growth (R2=0.64) or Al concentrations in

the roots (R2=0.62). The sum of monomeric-Al activity in solution adequately predicted

root Al concentrations (R2=0.93).

79

Introduction

Acid soils inhibit plant root growth primarily because of toxic levels of Al

(Adams and Lund, 1966; Clarkson, 1969; Foy and Fleming, 1978; Pavan and Bingham,

1982; Blamey et al., 1983; Alva et al., 1986a,b; Kim et al., 2001; Sumner, 1995; Sumner

and Yamada, 2002). Monocots demonstrate a range of responses to aluminum tolerance,

with differences also found between cultivars of the same species (Wenzl et al., 2002;

Giaveno and Miranda, 2000; Gallardo et al., 1999). Turfgrass species have different

levels of tolerance to Al, and show differences between varieties of the same species (Liu

et al., 1996; Liu et al., 1997). Computer-aided speciation models have allowed

researchers to narrow the range of Al species that are toxic to many plants. Pavan and

Bingham (1982) described reductions in coffee (Coffea arabica L.) seedling root growth

related to Al3+ activity. Blamey et al. (1983) and Alva et al. (1986a,c) showed that taproot

growth reduction in soybean (Glycine max (L.) Merr.) was best correlated with the sum

of the concentrations or activities of monomeric Al species, based on a computer-aided

speciation model. Vasconcelos et al. (2002) reported differences in Al-tolerance in rice

varieties (Oryza sativa L.) grown in solution. Kim et al. (2001) studied Brazilian wheat

(Triticum aestivum L. em Thell), triticale (X Triticosecale Wittmack), and rye (Secale

cereale L.) and found higher Al tolerance in newer varieties of triticale, but the highest

Al tolerance in wheat in a 50-year old variety.

In an experiment that included both Al and Ca (as CaSO4), Noble et al. (1988)

found that total solution Al did not describe soybean (Glycine max L. Merr.) seedling root

response to Al very well, but an equation they termed CAB (calcium-aluminum balance)

that includes activity and charge of Ca and select Al species, satisfactorily described the

80

effects of Ca and Al on root growth (R2 = 0.900). Cameron et al. (1986) examined the

effects of Al3+ and its complexes with F- and SO42- on root elongation of barley

(Hordeum vulgare) in nutrient solution. The study revealed seedling root elongation was

correlated with Al3+ concentrations, but not with total soluble Al or Al complexed with F-

or SO42-, and the authors speculated that this could be one reason why measurements of

labile Al using complexing agents have not been satisfactory in distinguishing Al-toxic

and nontoxic soils. Ryan and Kochian (1993) measured root growth and Ca uptake in

near-isogenic lines of wheat (Triticum aestivum L. Thell) and found Ca uptake in Al-

sensitive plants was inhibited after 40 minutes exposure to 50 μM Al, whereas in Al-

tolerant plants, Ca uptake was either unaffected or stimulated. They noted that in some

treatments Al was able to inhibit root growth significantly without affecting net Ca

influx. Ownby and Popham (1990) also found root exposure for only 3 h to Al in solution

inhibited root growth in wheat. More recently, rapid screening methods were developed

to detect Al-tolerant plants by utilizing hematoxylin staining (Giaveno and Miranda,

2000).

The objectives of this study were to determine seedling root response to Al in

three varieties of tall fescue (Festuca arundinacea L.) in the presence of various levels of

Ca supplied as CaSO4. Computer-aided, model-based predictions of Al speciation were

conducted in order to test the CAB equation (Noble et al., 1988) on tall fescue and to

explore alternative models for Al-root sensitivity in tall fescue.

81

Materials and Methods

Hydroponic experiments

Two hydroponic studies were conducted utilizing an incomplete factorial design.

In both experiments, 1-L pots (Ziploc®) were fitted with stainless steel wire mesh capable

of suspending seed at the mouth of the pot and covered with a layer of cheesecloth. The

surface area was 156.25 cm2. Three varieties of turf-type tall fescue (Festuca

arundinacea) – ‘K31’, ‘Rendition’, and ‘Dynasty’ - were selected. ‘K31’ was chosen

because it is the oldest variety of turf-type tall fescue still in commercial use and is the

genetic ancestor of most commercial varieties used presently. ‘Dynasty’ was selected

based its advertised drought resistance and suitability for hot summer climates such as

those experienced in the Southeastern U.S. ‘Rendition’ was selected as an intermediate

variety with unremarkable drought resistance or acid soil tolerance, yet wide availability

in the marketplace. Each pot received 1.60 g of a single seed variety. After hand-

distribution of the seed on the cheesecloth, pots were filled with deionized water to the

point that the cheesecloth became saturated, then placed randomly in a growth chamber

under fluorescent light. Light intensity was maintained at 95 W m-2. The chamber

temperature was maintained at 25o C, with continuous lighting. Pot water levels were

maintained on a daily basis in order to preserve cheesecloth saturation. As soon as roots

became visible through the wire mesh (after 5 d) the deionized water was poured out and

replaced with the appropriate hydroponic solution.

In the first experiment investigating varietal root-sensitivity to Al toxicity,

‘Rendition’, ‘K-31’ and ‘Dynasty’ tall fescue were grown in deionized water for 5 d. The

solution was then replaced with aluminum (Al) as AlCl3·6H2O in solutions of 0, 3.7, 7.4,

82

14.8, 37.0, or 74.1 μmol L-1. All solutions were brought to pH 4.5 with 0.01 M HCl

dripped under constant stirring during measurement. Treatments were grown for an

additional 7 d, at the end of which solution subsamples were preserved in 20-ml

scintillation vials, and roots were air-dried for 24 hr. Roots were cut from the stainless

steel mesh and measured on Mettler AM50 scales (Mettler Instrument Corp., Hightstown,

NJ) to the nearest 0.1 mg.

In the second experiment, ‘Rendition’, ‘K-31’ and ‘Dynasty’ tall fescue were

grown in deionized water for 5 d, with solutions, then replaced with various levels of Al

from AlCl3·6H2O and/or Ca from CaSO4·2H2O. The combined Al and Ca solutions

were created at 22o C by diluting saturated CaSO4·2H2O (14,000 μM) with deionized

water to the appropriate molarity, adding 0.01M Al from a buret as the solution was

stirred, followed by adding 0.01 M HCl from a buret under constant stirring until a stable

pH reading of 4.5 was achieved. Treatments were grown for an additional 7 d after

solution replacement, with periodic inputs of deionized water to replace water lost due to

evaporation underneath the growth chamber lights. The pH of each solution was adjusted

periodically to maintain a pH of 4.5 in order to simulate an acid soil solution. Upon

termination of the experiment, roots were air-dried for 24 h, then cut and weighed.

Solution subsamples were saved in 20-ml scintillation vials. ‘Dynasty’ and ‘K31’ roots

were weighed and placed in ceramic crucibles, ashed in a Thermolyne muffle furnace

(Sybron Corp., Dubuque, IA) at 500o C overnight, cooled to room temperature, and one

ml of 6M HCl was applied to dissolve the ash. Extracts were filtered through Whatman

#1 filter paper, with crucibles rinsed three times with deionized water, pouring the rinsate

into the filter paper. Extracts were brought to volume with deionized water, then analyzed

83

for Ca on a Perkin-Elmer Aanalyst 200 Flame Atomic Absorption Spectrometer (Perkin-

Elmer, Inc., Wellesley, MA) and Al on a Perkin-Elmer Elan 9000 Inductively Coupled

Plasma-Mass Spectrometer.

Al speciation

The distribution of specific Al and Ca species and the activities of their ionic

species in solution were calculated using VISUAL MINTEQ 2.30 (EPA, 1999;

Gustafsson, 2004). Calculated concentrations of Al, Ca, and H (pH) were used as inputs.

Mean activities of Al and Ca species across replications of treatments were calculated.

Default values of thermochemical constants of the various species in VISUAL MINTEQ

were used in calculations. The sum of monomeric Al species was calculated from the

predicted activities of individual monomeric Al species according to the equation: ΣaAl

mono = (Al3+) + (Al(OH)2+) + (Al(OH)2+) + (Al(OH)3

0) + (Al(OH)4-). The species

designated as AlSO4 species were AlSO4+ and Al(SO4)2

-.

The CAB equation (calcium-aluminum balance) from Noble et al. (1988) was

applied to root-growth results and root Al concentrations in this experiment. The equation

incorporates the activities of Ca and certain Al species as follows:

CAB = 2 log (Ca2+) – [3 log (Al3+) + 2 log (AlOH2+) + log (Al(OH)2+]

84

The purpose of Noble et al’s (1988) equation was to evaluate the interaction

between Ca and Al in determining root growth by taking the activities of free Ca2+ and

Al3+ into account and to consider the effect of other monomeric Al species and weighting

the effect of all of them in terms of valence.

Statistical analysis

Statistical analysis was performed using the 2-Way ANOVA procedure (α = 0.05)

and regression analysis utilizing SAS (SAS Institute, 1991). Comparisons were made

between experiments using Student’s t test for repeatability, with insignificant differences

(α = 0.05).

Results and Discussion

Aluminum only in solution – root response

There were significant differences between tall fescue varieties in sensitivity of

root response to aluminum concentration in solution. Total root mass for each variety was

also significantly different at the Al=0 μM level (Table 4.1), so root mass data was

normalized for comparison between varieties (Figure 4.1). ‘Rendition’, ‘Dynasty’ and

‘K31’ showed similar sensitivity to increasing levels of Al in solution, although

‘Dynasty’ was more sensitive at 7.4 and 14.8 μM. ‘Rendition’ was significantly less

sensitive to Al in solution than the other two varieties at 37.0 μM Al. Increased total Al in

solution caused a steep decline in relative root growth between 0 and 14.8 μM Al. At the

14.8 μM concentration level the least sensitive variety (‘Rendition’) produced only 40%

of optimal root growth, the second most sensitive (‘K31’) produced 37% of optimal

85

growth, and the most sensitive variety (‘Dynasty’) produced only 17% of optimal growth,

demonstrating that within varieties of tall fescue, the potential exists for agronomically

significant differences in soil-solution Al sensitivity, leading to differences in root growth

and root exploration of subsoils with high exchangeable acidity. Shainberg et al. (1989)

reported 8.9 μM soil-solution Al in a South African Oxisol, and Sumner et al. (1986) and

Buyeye and Fey (1987) recorded 11 and 34 μM soil-solution Al in a Georgia (USA) and

South African Ultisol, respectively. Thus, the range of naturally occurring soluble Al

corresponds closely to concentrations used in this experiment.

Aluminum and Calcium in solution – root response

Seedling root response to Al in solution was significantly affected by the quantity

of Ca as CaSO4 in solution. Root masses of the fescue varieties at 2500, 5000, and 10000

μM CaSO4 are affected by increasing Al concentration (Figure 4.2). Sumner et al. (1986)

reported soil-solution Ca concentrations in an acid Appling series soil of 40 μM at 0.90 to

1.05 m to 1780 μM at 0 to 0.15 m depth in control plots; and 480 μM at 0.90 to 1.05 m to

7850 μM at 0 to 0.15 m depth in gypsum-treated plots two years after application. Thus

the range of Ca concentrations used in this experiment reflect Ca concentrations in

certain field situations.

At 2500 μM Ca, an increase from 0 to 7.4 μM Al caused a slight decrease in root

growth for ‘Rendition’, no change in root growth for ‘Dynasty’, and an increase in root

growth for ‘K31’. However, at 14.8 μM Al, all three varieties showed a decreasing trend

in root growth, with severe root injury at 37.0 μM Al, and higher. ‘Rendition’ showed the

greatest tolerance to the highest level of Al (74.1 μM), and ‘Dynasty’ the least tolerance,

86

at the 2500μM CaSO4 in solution. Relative root growth at this level of CaSO4 and lower

levels of Al (7.4 and 14.8 μM) was much greater compared to zero CaSO4 at the same

levels of Al.

At the 5000 μM CaSO4 level, ‘Dynasty’ had the highest root mass at 7.4 μM Al,

but the lowest root mass at the 74.1 μM Al. ‘K31’ also had the highest root mass at 7.4

μM Al, and had significantly higher root mass than the other two varieties at the highest

concentrations of Al. It is of interest to note that ‘Rendition’ root mass increased from an

Al concentration of 7.4μM to 14.8 μM, in a manner similar to ‘K31’ in 2500 μM CaSO4,

as it increased from 0 to 7.4μM Al. In each case, these particular treatments were not

replicated due to spacing constraints, so it is difficult to ascertain whether these increases

were significant or not. Observable detrimental effects of Al toxicity on relative root

growth shifted from relatively low Al concentrations to higher Al concentrations as

CaSO4 increased from 2500 to 5000 and 10000 μM CaSO4.

At 10000 μM CaSO4, or roughly 71% saturation of CaSO4·2H2O, root mass

improved over lower levels of CaSO4 at all levels of Al. Root mass for ‘Rendition’

increased 37% at the same Al level (37.0 μM) as CaSO4 increased from 5000 to 10,000

μM. Increases for ‘K31’ and ‘Dynasty’ were not as dramatic at the same level of Al, but

significant increases were observed as CaSO4 increased. The greatest increases in root

growth as CaSO4 increased from 5000 to 10,000 μM occurred at the highest Al level

(74.1 μM), where ‘K31’ relative root growth increased from approximately 0.5 to 0.8,

‘Rendition’ increased from 0.25 to 0.5, and ‘Dynasty’ increased from 0.2 to 0.8 relative

to the optimal root growth. There were significant differences between the varieties at the

highest level of Al when CaSO4 was present at 10,000 μM, with ‘Rendition’

87

demonstrating more Al sensitivity; however, there were no significant differences

between varieties at 37.0μM Al or less. Overall, Al toxicity was nearly completely

ameliorated at this high level of CaSO4 addition.

Mechanistic effect of SO42- ion

An increase in Ca as gypsum (CaSO4·2H2O) increases the ionic strength of the

solution in a linear manner, but causes a more dramatic initial decrease in the Al3+

activity coefficient (Figure 4.3). Moreover, Al activity initially decreases sharply with

increasing CaSO4, with a corollary increase in AlSO4+. The addition of the SO4

2- ions

complex active Al, rendering them relatively harmless to roots (Noble et al., 1988). Alva

et al. (1991) proposed outer-sphere complexation of Al with SO42- as the mechanism by

which Al is still in an ionic form yet rendered less toxic to plant roots. In their study, an

increase in the SO4:Al ratio from 1 to 5 increased the percentage of Al complexed with

SO42- from 1.7 to 91.5 according to size exclusion chromatography, and from 29.0 to

61.3% as calculated by MINTEQ.

This study had a SO42- ion concentration several orders of magnitude greater than

the largest Al concentration, effectively creating an infinite supply of SO42- ions relative

to Al. Alva et al. (1991) stated that excess SO42- ions in solution had a shielding effect in

that the outer-sphere complexation of SO42- to aquo-Al species makes the complex

behave like a negatively charged ion. This would make binding to root meristematic

tissue or in the zone of elongation less likely, reducing the overall Al concentration in

roots.

88

Aluminum and Calcium in solution – Aluminum speciation

The activities of Al predicted by VMINTEQ in the various hydroponic solutions

are listed in Table 4.2. The model predicts that the addition of CaSO4 as gypsum

significantly reduces the activity of Al in solution compared to the total concentration.

The toxic aluminum species selected as relevant and summed were Al3+, AlOH2+,

Al(OH)2+, Al(OH)3

o, and Al(OH)4- (Noble et al., 1988; Kinraide, 1991). Species not

included due to their extremely low concentration levels were AlCl2+, Al2(OH)24+, and

Al3(OH)45+. Due to the relatively low levels of Al in solution, the addition of CaSO4

increased the ionic strength of the solutions by orders of magnitude. Increased ionic

strength had a significant impact on reducing the activity of Al in solution. The activity of

7.4 μM Al in solution with no CaSO4 added was 94% of the concentration of AlTot. The

addition of just 625 μM CaSO4 decreased Al activity at the same concentration, to 25%

of AlTot concentration. When CaSO4 was included at 2500 μM, Al activity dropped to less

than 10% of the AlTot concentration, and at 10,000 μM, Al activity was less than 4% of

the AlTot concentration in solution. CaSO4 is saturated at approximately 14,000 μM.

Within a given concentration of CaSO4, an increase in Al decreased the CaSO4:Al ratio,

increasing the probability of Al intercepting root tips.

Predicting root growth based on Ca and Al activities

An increase in Al activity - whether caused by an increase in Al at a given CaSO4

level or a decrease in CaSO4 at a given Al level - in solution at very low levels of Al had

an impact on relative root growth (Figure 4.4) causing an exponential loss of relative root

growth as Al activity increased. Most of the decrease occurred between 0 and 8 μmol

89

(Insert in Figure 4.4). Relative root growth was reduced to less than 40% at 14 μM and

10% or less at Al activity levels higher than 30μM. No single variety appeared to have

greater tolerance to Al activity over the others, and all variety data were combined for

nonlinear regression analysis. The relationship fit a non-linear, exponential decay

regression (R2=0.787), with less overall goodness of fit caused by under-predicting

relative root growth at Al activity levels above 10 μM.

Noble et al. (1988) studied the growth of soybean roots in nutrient solutions that

were subjected to various levels of Ca and Al, and improved the goodness of fit for

relative root length as a function of Ca and Al in solution by utilizing the CAB equation,

producing an R2 = 0.900 when plotted against relative root length. Analysis of tall fescue

root mass was conducted using the CAB equation with less goodness of fit (Table 4.3) as

a function of relative root growth. The adjusted R2 for all fescue varieties correlated to

the CAB equation was 0.64. After separating the results by variety, ‘Dynasty’ had a

slightly greater coefficient of variation (0.66), while ‘K31’ and ‘Rendition’ had similar

coefficients of variation, at 0.64. In general, a positive correlation existed between fescue

root mass and CAB, however no further root growth was observed as CAB increased

above 6.5 for all varieties (Figure 4.5).

Root Aluminum

The concentration of Al found in roots was related to the CAB equation (Figure

4.6). The data were combined and regressed with a linear equation giving a goodness of

fit of 0.62. In general, an inverse relationship existed between the concentration of root

Al and CAB value. Since the CAB equation provides a difference of the log of Ca

90

activity and Al activity, this can be interpreted as a greater Ca activity in solution will

result in lower root Al. Excess Ca ions may provide a root-protective effect in solution in

the sense that the Ca (and SO4) addition reduces Al activity, so the root tip has less

contact with active Al.

The concentration of Al found in roots was exponentially related to the activity of

Al in solution (R2=0.93). A steep rise in root-Al concentrations occurred between 0 and

14 μM active solution-Al, with no significant difference in root Al when solution-Al

activity was between 34 and 62 μM (Figure 4.7). At these high levels of solution-Al

activity, root growth was damaged to such an extent that no increase in root Al was a

function of a lack of root mass (Figure 4.4), suggesting Al saturation on the small root

mass remaining.

Conclusions

Aluminum toxicity represents a potentially serious obstacle to adequate root growth for

turf-type tall fescues in acid soils. Aluminum toxicity in the soil can potentially be

overcome by the surface application of Ca as gypsum, because of its greater solubility

relative to lime. In hydroponic solutions, the addition of soluble CaSO4 increased root

growth for three varieties of tall fescue subjected to various levels of Al in solution. The

CAB equation did not provide an adequate model for explaining the mechanistic effects

of Al and Ca on tall fescue roots in this study. Computer models such as MINTEQ

predicted that the addition of CaSO4 changed the Al species in solution from toxic Al3+ to

relatively non-toxic species such as AlSO4+ and that the increased ionic strength of the

solution significantly reduced Al activity. The CAB equation was not a particularly good

91

predictor of tall fescue root growth as a function of Ca and Al in solution. The CaSO4

root-protection mechanism appears to be a function of reduced monomeric Al activity

caused by Al ions forming outer-sphere complexes with SO42- ions in solution from the

addition of soluble CaSO4, rendering Al ions less toxic. Predicted levels of CaSO4

required to overcome a given level of Al in solution can be drawn from the observation

that Al activity is reduced to less than 10% by the addition of 2.5 mM CaSO4 (Figure

4.3), most of which has been converted to AlSO4+. This relationship may be useful to

construct estimates of how much gypsum would be required to overcome Al toxicity for

an acid soil.

92

References Cited

Adams, F., and Z.F. Lund. 1966. Effect of chemical activity of soil solution aluminum on

cotton root penetration of acid subsoils. Soil Sci. 101:193-198.

Alva, A.K., D.G. Edwards, C.J. Asher, and F.P.C. Blamey. 1986a. Effects of

phosphorus/aluminum molar ratio and calcium concentration on plant response to

aluminum toxicity. Soil Sci. Soc. Am. J. 50: 133-137.

Alva, A.K., D.G. Edwards, C.J. Asher, and F.P.C. Blamey. 1986b. Relationship between

root length of soybean and calculated activities of aluminum monomers in

nutrient solution. Soil Sci. Soc. Am. J. 50: 959-962.

Alva, A.K., F.P.C. Blamey, D.G. Edwards, and C.J. Asher. 1986c. An evaluation of

aluminum indices to predict aluminum toxicity to plants grown in nutrient

solutions. Commun. Soil Sci. Plant Anal. 17: 1271-1280.

Alva, A.K., G.L. Kerven, D.G. Edwards, and C.J. Asher. 1991. Reduction in toxic

aluminum to plants by sulfate complexation. Soil Sci. 152: 351-359.

Blamey, F.P.C., D.G. Edwards, and C.J. Asher. 1983. Effects of aluminum, OH:Al and

P:Al molar ratios, and ionic strength on soybean root elongation in solution

culture. Soil Sci. 136: 197-207.

Buyeye, S.M., and M.V. Fey. 1987. Several explanations for the amelioration of acid,

sesquioxidic soils by gypsum application. Proc. Soil Sci. Soc. S. Afr. Conf.

Cameron, R.S., G.S.P. Ritchie, and A.D. Robson. 1986. Relative toxicities of inorganic

aluminum complexes to barley. Soil Sci. Soc. Am. J. 50: 1231-1236.

93

Clarkson, D.T. 1969. Metabolic aspects of aluminum toxicity and some possible

mechanisms of resistance. p. 381-397. In I. H. Rorison (ed.) Ecological aspects of

the mineral nutrition of plants. Br. Ecol. Symp. No. 9. Sheffield, England. 1-5

April 1968. Blackwell Scientific Publications, Edinburgh.

EPA. 1999. MINTEQA2/PRODEFA2, a geochemical assessment model for

environmental systems. http://www.epa.gov/ceampubl/mmedia/minteq/index.htm

(verified 16MAY2005).

Foy, C.D., and A.L. Fleming. 1978. The physiology of crop tolerance to excess available

aluminum and manganese in acid soils. P. 301-328. In G.A. Jung (ed.) Crop

tolerance to sub-optimal land conditions. Spec. Pub. 32. ASA, Madison, WI.

Gallardo, F., F. Borie, M. Alvear, and E. vonBaer. 1999. Evaluation of aluminum

tolerance of three barley cultivars by two short-term screening methods and field

experiments. Soil Sci. and Plant Nutr. 45: 713-719.

Giaveno, C.D., and J.B. Miranda. 2000. Rapid screening for aluminum tolerance in maize

(Zea mays L.). Genetics and Molecular Biology 23: 847-850.

Gustafsson, J.P. 2004. Visual Minteq ver. 2.30: Windows version of MINTEQA2 ver.

4.0. Royal Inst. Tech. in Stockholm, Stockholm, Sweden.

Kachurina, O.M., H. Zhang, W.R. Raun, and E.G. Krenzer. 2000. Simultaneous

determination of soil aluminum, ammonium- and nitrate-nitrogen using 1 M

potassium chloride extraction. Commun. Soil Sci. Plant Anal. 31: 893-903.

Kim, B.Y., A.C. Baier, D.J. Somers, and J.P. Gustafsson. 2001. Aluminum tolerance in

triticale, wheat, and rye. Euphytica 120: 329-337.

94

Kinraide, T.B. 1991. Identity of the rhizotoxic aluminum species. Plant and Soil 134:

167-178.

Liu, H., J.R. Heckman, and J.A. Murphy. 1996. Screening fine fescues for aluminum

tolerance. J. of Plant Nutr. 19: 677-688.

Liu, H., J.R. Heckman, and J.A. Murphy. 1997. Greenhouse screening of turfgrasses for

aluminum tolerance. Int. Turfgrass Soc. Res. J. 8: 719-728.

Liu, H. 2005. Aluminum resistance among seeded bermudagrasses. Hortscience 40: 221-

223.

McElreath, D.L., R.L. Westerman, and G.V. Johnson. 1992. Comparison of soil

extraction parameters for Al, Mn, and Ca determinations. Commun. Soil Sci.

Plant Anal. 23: 2493-2510.

Noble, A.D., M.V. Fey, and M.E. Sumner. 1988. Calcium-aluminum balance and the

growth of soybean roots in nutrient solutions. Soil Sci. Soc. Am. J. 52: 1651-

1656.

Ownby, J.D., and H.R. Popham. 1989. Citrate reverses the inhibition of wheat root

growth caused by aluminum. J. Plant Physiol. 135: 588-591.

Ownby, J.D., and H.R. Popham. 1990. Citrate reverses the inhibition of wheat root-

growth caused by aluminum. J. Plant Phys. 135: 588-591.

Pavan, M.A., and F.T. Bingham. 1982. Toxicity of aluminum to coffee seedlings grown

in nutrient solution. Soil Sci. Soc. Am. J. 46: 993-997.

Pinheiro, J.C., and D.M. Bates. 2000. Mixed effect models in S and S-PLUS. p. 519-525.

Springer, New York.

95

Ryan, P.R., and L.V. Kochian. 1993. Interaction between aluminum toxicity and calcium

uptake at the root apex in near-isogenic lines of wheat (Triticum aestivum L.)

differing in aluminum tolerance. Plant Physiol. 102: 975-982.

SAS Institute. 1991. SAS/STAT users guide. Ver. 6. 4th ed. SAS Institute, Cary, NC.

Shainberg, I., M.E. Sumner, W.P. Miller, M.P.W. Farina, M.A. Pavan, and M.V. Fey.

1989. Use of gypsum on soils: A review. Adv. Soil Sci. 9: 1-111.

Sumner, M.E. 1995. Amelioration of subsoil acidity with minimum disturbance. p. 147-

185. In Jayawardane and Stewart (eds.) Subsoil management techniques. Adv. In

Soil Sci. CRC Press, Boca Raton, FL.

Sumner, M.E., H. Shahandeh, J. Bouton, and J.E. Hammel. 1986. Amelioration of an

acid soil profile through deep liming and surface application of gypsum. Soil Sci.

Soc. Am. J. 50: 1254-1258.

Sumner, M.E., and T. Yamada. 2002. Farming with acidity. Commun. Soil Sci. Plant

Anal. 33: 2467-2496.

Vasconcelos, S.S., J. Jacob-Neto, and R.O.P. Rossiello. 2002. Differential root responses

to aluminum stress among Brazilian genotypes. J. Plant Nutr. 25: 655-669.

Wenzl, P., J.E. Mayer, and I.M. Rao. 2002. Aluminum stress inhibits accumulation of

phosphorus in root apices of aluminum-sensitive but not aluminum-resistant

Brachiaria cultivars. J. Plant Nutr. 25: 1821-1828.

96

Table 4.1. Root mass of tall fescue varieties in the presence of zero total Al in solution.

Mean Standard error Variety --------mg pot-1--------

‘Rendition’ 106.3 8.56 ‘K31’ 67.9 1.83 ‘Dynasty’ 92.7 3.97

97

Table 4.2. Al speciation in various levels of Ca as dissolved gypsum as predicted by Visual MINTEQ2.

CaSO4 Al ΣAlmonoactivity

Ionic Strength

a† Ca:Al

-----------------μM----------------- (M) (%) molar ratio 0 7.4 6.93 0.0001 93.7 0.0 14.8 13.58 0.0001 91.8 0.0 37.0 32.53 0.0002 88.0 0.0 74.0 61.98 0.0004 83.8 0.0

625 7.4 1.84 0.0023 24.9 252.8 14.8 3.70 0.0023 25.0 125.8 37.0 9.38 0.0023 25.4 49.6 74.0 19.23 0.0023 26.0 24.2

1250 7.4 1.17 0.0044 15.8 706.6 14.8 2.33 0.0044 15.8 353.3 37.0 5.83 0.0044 15.8 141.3 74.0 11.67 0.0044 15.8 70.6

2500 7.4 0.71 0.0082 9.6 1977.7 14.8 1.42 0.0082 9.6 988.9 37.0 3.55 0.0082 9.6 395.5 74.0 7.11 0.0082 9.6 197.5

5000 7.4 0.42 0.0151 5.7 5442.8 14.8 0.85 0.0151 5.7 2721.4 37.0 2.12 0.0151 5.7 1088.6 74.0 4.25 0.0151 5.7 543.5

10000 7.4 0.25 0.0277 3.4 14716.4 14.8 0.50 0.0277 3.4 7351.8 37.0 1.25 0.0277 3.4 2933.1 74.0 2.52 0.0277 3.4 1458.2

†a: (Al activity)/(Al concentration)*100.

98

Table 4.3. Linear regression goodness of fit for tall fescue varieties modeling root growth as a function of the CAB equation.

All varieties ‘Dynasty’ ‘K31’ ‘Rendition’ Adj. R2 0.64 0.66 0.64 0.64 RMSE 24.48 22.26 14.90 27.36 F ratio 113.82 45.29 47.75 41.11 Prob > F <0.0001 <0.0001 <0.0001 <0.0001

99

Total Al (μM)

0 20 40 60 80

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0

1.2

'Rendition''K31''Dynasty'

Figure 4.1. Tall fescue relative root growth sensitivity to Al in solution (pH 4.5). Error bars indicate standard deviation of the mean.

100

10,000 μM CaSO4

Total Al (μM)

0 20 40 60 80

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0

'Rendition''K31''Dynasty'

5000 μM CaSO4

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0'Rendition''K31''Dynasty'

2500μM CaSO4R

elat

ive

root

gro

wth

0.0

0.2

0.4

0.6

0.8

1.0'Rendition''K31''Dynasty'

Figure 4.2. Tall fescue relative root growth in Al solutions at various levels of CaSO4. Error bars indicate standard deviation of the mean.

101

mM CaSO4

0 2 4 6 8 10IS, A

l3+ a

ct. c

oeff

., A

l act

ivity

, or %

Al a

s SO

4 com

plex

0

20

40

60

80

100

% Al as SO4 complex

activity coefficient of Al3+

Al activityIS (mM)

Figure 4.3. Effect of increasing CaSO4 concentration (in mM) on ionic strength (IS); activity coefficient of Al3+; complexation of Al by SO4 (expressed as %, averaged over AlTot = 3-70 μM); and activity of Al3+ in a solution of AlTot = 100 μM containing levels of CaSO4 as shown on abscissa (x-axis). All parameters generated using VMINTEQ (Davies equation, default parameters).

102

ΣAlactivity (μM)

0 20 40 60

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0

RenditionK31Dynastynon-linear regression:R2=0.787: y = 0.9473*e-0.2209x

Insert

ΣAlactivity (μM)

0 2 4 6

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0

Figure 4.4. Tall fescue relative root growth in response to total Al activity in solution with nonlinear regression model equation and goodness of fit. Error bars indicate standard deviation of the mean.

103

CAB

4 5 6 7 8 9 10

Rel

ativ

e ro

ot g

row

th

0.0

0.2

0.4

0.6

0.8

1.0

RenditionK31DynastyR2=0.640: y = -0.3063 + 0.1548x

Figure 4.5. CAB analysis: Effect of increasing calcium-aluminum balance (CAB) on the

relative root growth of tall fescue cultivars in Ca/Al solutions. Error bars indicate standard deviation of the mean.

104

CAB

4 5 6 7 8 9

Roo

t Al

con

cent

ratio

n (μ

g g-1

)

0

500

1000

1500

2000

2500

DynastyK31

R2=0.620: y = 1894 + -218.4x

Figure 4.6. Effect of increasing calcium-aluminum balance (CAB) on the concentration of Al in tall fescue roots, including linear regression analysis. Error bars indicate standard deviation of the mean.

105

ΣAlactivity in solution (μM)

0 20 40 60

Roo

t Al c

once

ntra

tion

(μg

g-1)

0

1000

2000

3000

4000

5000

6000DynastyK31

R2=0.926: y = 3993(1-e-0.05764x)

Insert

0 2 4 6 8

Roo

t Al c

once

ntra

tion

(μg

g-1)

0

500

1000

1500

2000

2500

Figure 4.7. Effect of increasing total Al activity on root-Al concentrations in tall fescue varieties grown in various CaSO4/Al solutions, including non-linear regression analysis. Error bars indicate standard deviation of the mean.

106

CHAPTER 5

ROOT-GROWTH AND WATER-USE RESPONSE OF TALL FESCUE TO

GYPSUM AND LIME-AMENDED ACID SUBSOILS

______________________________________________________ Kruse, J.S. and W.P. Miller, E.G. To be submitted to Journal of Plant Nutrition

107

Abstract

A factorial experiment was conducted in the laboratory to determine the effects of

gypsum compared to lime on the soil chemical properties of acid subsoils, and

subsequent root growth and water-use patterns of tall fescue (Festuca arundinacea v.

‘Rendition’) using repacked soil columns. Tall fescue was grown for 85 d on two acid

subsoils with similar chemical characteristics but differing in texture. Treatments

included surface-applied CaSO4·2H2O and soil-incorporated CaCO3. Columns were

placed under lights in a plywood container to maintain the soil temperature at 15o C.

After an initial growth period, the tall fescue was physiologically stressed by a dry-down

procedure of 33 and 30 d. Upon termination of the experiment, columns were divided

into three sections for soil chemical analysis and root density determination. Lime and

gypsum significantly increased soil extractable Ca through the column profile, indicating

surface-applied gypsum moved well through the soil. Gypsum significantly reduced Mg,

had little effect on exchangeable K, and some effect on CEC and exchangeable acidity.

Lime-treated soils produced a significantly larger root-density growth response (59%)

over control treatments, while gypsum had a significant, though lesser effect (32%

growth response). There were significant differences in root densities between soils

across treatments. Root distribution by depth was not affected by the gypsum treatment in

a Cecil soil, but highly affected in the Cunningham soil. Differences in root density based

on treatment did not result in significant differences in column water loss during the dry-

down phase, indicating water loss was a poor measure of root response to changes in soil

chemical conditions.

108

Introduction

Subsoil acidity is a significant problem in maintaining vigor and quality of

turfgrasses on highly weathered soils of the southeastern United States (Murray and Foy,

1978). Strongly acid subsoils, especially those below pH 5.5, create conditions that lead

to poor root growth, vigor, nutrient acquisition, and the ability to compete with acid-

tolerant weed species. Research has demonstrated that aluminum toxicity plays a key role

in limiting growth in many acid soils because of its detrimental effects on root cell

division and elongation (Matsumoto, 2000).

Optimal ranges in soil pH have been established for many turfgrass species

(Beard, 1973), with tall fescue (Festuca arundinacea L.) demonstrated to be relatively

sensitive to Al (Palazzo and Duell, 1974; Liu et al., 1997). Transition areas of the

southern United States are characterized by very hot summers with concomitant rates of

evapotranspiration that result in dry surface soils. Because of the presence of toxic Al,

these cool-season grasses often have shallow root systems vulnerable to desiccation or

damage from diseases or insects. Acid subsoils often contain large reservoirs of available

soil water that offer the potential for increasing turfgrass survival if soil chemical and

physical conditions could be suitably modified to make the environment compatible for

root development.

Turfgrass presents a unique limitation to amelioration of subsoil acidity because

once established, turfgrasses are not deep-plowed in order to incorporate limestone.

Although the benefits of limestone are well recognized for improved crop production

(Adams, 1984), surface applications of limestone often offer less-than-satisfactory results

109

due to its relative insolubility and slow movement through the soil profile (Gascho and

Parker, 2001; Farina and Channon, 1988).

Numerous studies have shown that increased Ca levels in soils are able to at least

partially offset toxic Al effects in acid subsoils (Pavan et al., 1982; Stehouwer et al.,

1996; Wang et al., 1999) Calcium movement in the soil is dependent upon excess rainfall

and the abundance of compatible anions such as NO3-, SO4

2- (Sumner, 1995) or certain

plant or animal waste residues (Miyazawa et al., 2002; Hue and Licudine, 1999). Gypsum

(CaSO4·2H2O) offers an alternative to limestone in ameliorating acid subsoils under

turfgrass. Although a salt and not a liming agent, gypsum is more soluble than limestone

(CaCO3) (2.41 g L-1 at 25o C for gypsum versus 0.0153 g L-1 at 25o C for CaCO3), and

has the potential to reduce exchangeable acidity and improve crop performance on many

acidic subsoils (Shainberg et al., 1989; Levy and Sumner, 1998). Little information exists

in the literature about the effects of gypsum on turfgrass growing on acid soils (Murray

and Foy, 1978; Clark and Baligar, 2003). The objectives of this experiment were to 1)

determine the effects of surface-applied gypsum on soil chemical properties of differing

origin and 2) evaluate subsequent root growth and water use patterns of tall fescue.

Materials and Methods

Bulk soil was collected from a field on a mountain plateau in Walker County,

Georgia, at a depth of 0.5 to 1.0-m and from a field with a 35-yr pine forest at the

University of Georgia Whitehall Forest in Clarke County, Georgia at a depth of 0.8 to

1.2-m. The soils were a Cunningham fine sandy loam (fine, mixed, semiactive, thermic

Typic Hapludults) and a Cecil sandy clay (fine kaolinitic, thermic Typic Kanhapludult),

110

respectively. Soils were air dried, crushed, homogenized and passed through a 2-mm

sieve. The water content of the soils at field capacity (-0.01MPa) and wilting point (-1.5

MPa) was determined in a pressure chamber using methods described by Klute (1986).

The experiment was a factorial combination of the two soils and three treatments

(lime, gypsum, and control) with three replicates for each treatment. Soil properties are

shown in Table 5.1. Reagent grade CaCO3 was incorporated into soils at a rate sufficient

to attain a target pHH2O of 6.5, by thoroughly mixing soil and lime in a plastic container.

Columns were prepared from 10-cm diameter, schedule-40 polyvinyl chloride tubes cut

in 34-cm lengths. The interior wall of the column was painted with a copper oxide paint

to prevent root whorling in the soil/tube interface. Plastic end caps were secured to the

bottom of the column, with 3x3-cm screen and cheesecloth placed over a 1-cm diameter

hole cut in the center of the end cap to allow for drainage. Approximately 150 g pea

gravel was placed in the bottom of the column, followed by 200 g of acid-washed, course

sand, to ensure adequate drainage after leaching. The columns were packed with soil to a

soil thickness of 27.5 cm by pre-weighing 3300 g of Cunningham soil and 2864 g of

Cecil soil, then pouring the soil through a funnel while tapping vigorously on column

sides with a metal rod, followed by gently tamping the column on the base. The packed

columns had an initial bulk density of 1.02 g cm-3 for the Cunningham soil and 1.18 g

cm-3 for the Cecil soil.

Based on previously published rates (Sumner, 1990) columns receiving gypsum

treatments (reagent grade CaSO4⋅2H2O) mixed with 75 g of an acid-washed sand was

placed on top of the packed soil column, to provide a rate equivalent to 11 Mg ha-1. The

111

Ca equivalent applied as lime was 40.4% of the Ca equivalent applied as gypsum in the

Cecil sandy clay, and 31.1% in the Cunningham sandy loam.

A second, empty column was placed on top of the soil column and secured with electrical

tape, and 4000-ml deionized water was slowly poured into all columns. Columns were

allowed to drain for 48 h, weighed, seeded with 0.32-g of a turf-type tall fescue (Festuca

arundinacea v. ‘Rendition’) mixed in 75 g washed sand and placed in a greenhouse with

intermittent misters. Columns were watered by misters for 6 s every 10 m for 11 h d-1.

The rate of water applied to each column was calculated to be 0.81 cm d-1 based on

collection containers placed adjacent to the columns. Fertilizer was applied at the rate of

18.5 kg N, 25 kg P2O5, and 4.6 kg K2O ha-1 24 h after seed application. The fertilizer

components consisted of urea, methylene urea, monoammonium phosphate, and

potassium chloride. Columns were left in the greenhouse 17 d, then transferred to an

insulated unit in which all but the top 2 cm of the column were in air maintained at 15o C

(± 1o C) to simulate soil temperatures (Figure 5.1). Column tops were placed under metal

halide growth lights illuminated 12 h d-1, creating column surface temperatures of 28o C.

Light output was measured at 102.3 W m-2 on a Li-cor LI 189 (Li-cor, Inc., Lincoln, NE).

Columns were weighed and brought to field capacity gravimetrically with deionized

water every 48 h and rotated periodically.

Thirty days after germination, 200 ml deionized water was added to columns and

leachate was collected from the bottom of the column using plastic containers. Leachate

was filtered and analyzed for electrical conductivity (EC) and pH. Calcium, Mg, and K

were determined by flame atomic absorption spectroscopy (Aanalyst 200 atomic

absorption spectrometer, Perkin Elmer, Inc., Wellesley, MA). Shoot growth was

112

quantified by collecting herbage at 2-cm height every 15 d, air-drying, and weighing

cuttings.

Forty-five days after germination columns were brought to field capacity and the

weight was recorded. Columns were then allowed to dry down, weighed daily, and half

the water lost the previous day by evapotranspiration was replaced gravimetrically. This

dry-down procedure continued for 19 d, after which columns were refilled

gravimetrically to field capacity, then dried down with no water replacement for 21 d.

Periodic gravimetric measurements of water loss were recorded.

Upon termination of the experiment at 85 d after germination, final shoot weights

were made. Columns were laid on their sides and the bottom cover, screen, cheesecloth,

sand and pea gravel were removed. Column soil was pushed out of the top of the column

utilizing a push rod against a flat metal surface cut to fit the interior dimensions of the

column. The columns were sectioned with a serrated knife at 0-8, 8-16, and 16-27.5 cm

and placed in open, liter-sized plastic bags. Bags were tared and weighed. Subsamples of

soil (30.0 g) were removed randomly from each bag for chemical analysis. Remaining

soil was weighed, and washed in a 2-mm sieve to capture roots. Roots were air-dried and

weighed to the nearest 0.1 mg. Root density of each subsection was calculated as root

mass per soil mass.

Subsamples removed for chemical analysis were measured for pHH2O, pHKCl, EC,

CEC, exchangeable acidity, and extractable Ca, Mg, and K. The pHKCL was determined

according to the methods of McLean (1982) substituting 1M KCl for 0.01M CaCl2. CEC

was determined according to the methods of Sumner and Miller (1996) for acid soils. A

soil:water paste of 2.5:1 was allowed to equilibrate overnight, with the supernatant

113

measured for EC on a CDM 80 conductivity meter (Radiometer Copenhagen, GmbH,

Willich-Schiefbahn, Germany) . Exchangeable acidity was measured by titration with

0.01 M NaOH after extraction with 1 M KCl in a 2.5:1 solution:soil paste. Extractable

Ca, Mg and K were extracted with 1M NH4OAc and measured by atomic absorption

spectroscopy.

Results and Discussion

Soil chemistry

The two soils that were selected for this experiment were chemically similar in

that they were both acidic, with a large exchangeable acidity, and with similar cation

exchange capacity (Table 5.1). The Cunningham sandy loam had small concentration of

bases and less buffer capacity than the Cecil sandy clay. Texture differed mainly in silt

and clay content. The siltier Cunningham soil contained primarily fine sand and had a

lower porosity that resulted in lower permeability during the initial leaching. The

Cunningham sandy loam had a yellow color, indicating predominantly Fe and Al

hydroxides and oxyhydroxides. The Cecil sandy clay was reddish in appearance,

indicating predominantly Fe and Al oxides.

Soil chemical analysis indicates that gypsum was successfully watered in through

the full length of the columns. Measurement of electrical conductivity (EC) of both the

sandy loam and sandy clay soils revealed much greater EC values in gypsum-treated soils

than either lime-treated or control columns, as would be expected due to the addition of a

salt (Figure 5.2). The EC values for the control treatments were small and typical of

highly weathered soils. Lime-treated soils showed some increase in EC in the lower

114

column sections. Electrical conductivity levels of gypsum-treated columns were

significantly greater than lime-treated columns, below gypsum saturation levels, and

evenly distributed through the column profile, indicating effective movement of gypsum

through the column-soil profiles.

The pH of the different column treatments indicated that although lime was

incorporated in the lime-treated columns to reach a target pHH2O of 6.5, end-of-

experiment pHH2O values were below 6.0 for the sandy loam and 6.0 for the sandy clay

(Figure 5.3). The small addition of ammonium-based nitrogen fertilizer at the rate of 18.5

kg ha-1 was thought unlikely to account for this. The pH values in the gypsum-treated

columns were a full pH unit below the lime, and just significantly greater than the control

columns. Analysis of the pHKCl values (Figure 5.4) provides a more accurate comparison

of soil pH values between treatments by eliminating the salt effect of added gypsum. This

method indicates pH values for lime-treated soils still significantly greater than the other

two treatments, and little difference in pH between gypsum-treated and control columns,

as was expected.

Control columns exhibited the greatest exchangeable acidity (Figure 5.5), with

gypsum treated soils having significantly less exchangeable acidity than the controls.

However, there was significantly more exchangeable acidity than the lime-treated

columns. This is further indication that gypsum, although not a liming agent, can reduce

exchangeable acidity, either by specific adsorption of SO42- (the self-liming effect: Reeve

and Sumner, 1972), complexing monomeric Al as aluminum sulfate and/or increasing the

ionic strength of the soil solution, which can accelerate aluminum polymerization (Alva

et al., 1991). Extractable Ca concentrations provide an indication that lime was well

115

mixed in the lime-treated columns, and that gypsum moved all the way through the

columns during the initial (watering-in) phase of the experiment. Calcium levels were

highest in the lime-treated columns, significantly less in the gypsum columns, and

significantly lower yet in the control columns (Figure 5.6). Calcium concentrations

increased with depth in the control and lime-treated columns of both soils, probably

caused by depletion by plant uptake in the soil surface, but was fairly uniform with depth

in the gypsum-treated columns.

Gypsum exchanged and leached Mg in significant amounts, especially in the

Cunningham sandy loam soil that had low initial exchangeable Mg (Table 5.2). Gypsum

did not leach K. O’Brian and Sumner (1988) found significant leaching of Mg and K in

packed-soil columns after the addition of phosphogypsum at similar rates. The

extractable Ca in the Cecil and Cunningham soils treated with gypsum were similar in the

top, middle and lower sections of the columns (Table 5.2). Column leachate results taken

30 d after planting indicate, that gypsum moved well through the soil by leaching.

Calcium was collected in concentrations exceeding 130 mg L-1 after total column

watering reached 6.2 L, indicating gypsum movement through the entire profile. Other

cation leaching was slight, with Mg and K collected in concentration of 5 mg L-1 or less

(Figure 5.7). Liu and Hue (2001) reported similar results on column gypsum movement

in lime and gypsum-amended Ultisols. Columns were watered for 27 d and only 7.6% of

the Ca from surface-applied lime moved beyond the surface 10 cm, while 60% of Ca as

gypsum moved past the surface 10 cm, and 6.4% moved beyond 45 cm. Aluminum

saturation was reduced from approximately 47% in all profile segments in the control

columns to less than 20% in all profile segments of the gypsum treatments.

116

Root density

There were significant differences (α=0.05) in column root densities between

treatments averaged over soil type and between soil types averaged over treatments

(Table 5.3). In treatment comparisons averaged over the two soil types, lime and gypsum

treatments had significantly more root mass in the total columns than the control

treatment, and the lime treatment had significantly more root density than the gypsum

treatment. Lime-treated soil improved root density by 59% over control soils, indicating

the untreated acid soils were root-limiting. Gypsum-induced improvements in root

density were less dramatic than the lime treatment, with an average 32% growth response

over the control for both soils. Statistical analysis of root density growth response over

control treatments by individual soil type increases the variability within treatments,

reducing significant differences between treatments for the total column (Figure 5.8). The

gypsum results are less than those of Clark and Baligar (2003) where ‘K31’ tall fescue

was grown in flue-derived gypsum amended acid-soils and reported a 29 to 41% root

response over optimal CaCO3-treated columns. This was a greater response than observed

with orchardgrass (Dactylis glomerata L.) that showed a –22 to 28% root response

compared to CaCO3-treated columns.

The Cunningham sandy loam and Cecil sandy clay soils produced mean root

densities of 2248.1 and 3233.1 mg kg-1, respectively, across all treatments. Tukey’s

minimum significant difference for soil type was 665.3, indicating a significant

difference between soil types (Figure 5.8). Although both soils had similar soil chemical

properties, it is interesting to note that the sandy loam produced 30% less root mass than

117

the sandy clay. Soil physical characteristics such as poor structure may have played a role

in the mechanical impedance of roots (Radcliffe et al., 1986; McCray et al., 1991).

Root density results from the gypsum-treated sandy loam contrast with those of

the sandy clay. 17% of the sandy clay column root density was measured in the 8-16 cm

section of the column, and 21% of the root density in the 16-27 cm depth (Figure 5.8). By

comparison, only 8% and 2% of the root densities were found in the middle and bottom

sections, respectively, of the sandy loam. Ninety percent of the mean root density for

gypsum treatment in the sandy loam was measured in the top third of the soil column.

Root distribution was significantly affected by soil type. The distribution pattern

in the Cecil sandy clay was consist across treatments, with roughly 60% of roots in the

top 8 cm, 20% in the middle 8 cm, and 20% in the bottom 8 cm (Figure 5.8). This pattern

did hold true for the Cunningham sandy loam, where the gypsum-treated soil had 90% of

the root density in the top 8 cm (Figure 5.8). Statistical analysis of the columns by

individual depth revealed a significant interaction between soil type and treatment, even

though no such interaction was detected in analysis of the whole column. This may be

caused by the way roots were distributed in the columns, with the majority of roots

growing in the top 8 cm of column soil for both soil types. Means testing the columns by

depth indicated that in the top 8 cm of the column, the sandy clay had significantly more

(28%) roots than the sandy loam across treatments Furthurmore, the gypsum and lime

treatments had significantly more root density than the control soils. The mean root

density of the gypsum treated soils across soil type in the top 8-cm was 50% greater than

the mean root density of the control soils and the same as lime-treated soils (Table 5.3).

118

In the 8 to 16 cm depth an overall reduction in root mass across soils and

treatments compared to the upper section of soil was measured in the lime-incorporated

columns, indicating that some root limitations observed in experimental columns and in

the field, are partly a function of grass morphology, in that grasses typically grow most of

their roots near the surface of the soil profile regardless of soil chemical conditions. In 8

to 16 cm section of the columns, lime-treated soil had significantly greater root density

than the gypsum-treated and control soils, and the gypsum-treated soil was not

significantly greater than the control columns. Root density in the sandy clay soils was

almost 50% larger than the sandy loam soils at this depth (Table 5.3).

Results from the 18 to 27 cm depth were similar to the middle section (Table 5.3).

Lime-treated soils had significantly greater root density than gypsum-treated and control

soils, and gypsum treated root density was not significantly more than control column

root density. Soil types were significantly different, with the sandy clay root density

having 237% of the sandy loam root density.

Drydown procedure

An attempt was made to determine whether drying down the columns and

periodically measuring water loss gravimetrically would reveal differences in water use

based on treatment. The hypothesis was that acid-ameliorated soils would allow increased

root density that would lead to more rapid water use in the total column that could be

detected by periodic measurement. Despite significant differences in total root density

between treatments (Table 5.3), differences between treatments in soil water extraction

during the drydown were not observed. ANOVA analysis of water evapotranspiration per

119

day over the 90-d period found no significant interaction between soil type and treatment,

and no significant differences in the main effects. On average, 21.2 g of column water

were lost per d, with a linear water loss, even when the soils were becoming dry (Figures

5.9 and 5.10). All tall fescue plants were observed to dessicate at roughly the same rate,

regardless of treatment. The procedure to allow the columns to dry down, then replace

half the water lost during the previous drydown period, can be termed a “conditioning

drydown” that allows the plant to physiologically adapt to drier soil conditions (Lo

Bianco et al., 2000). Earlier column drydown studies with zoysiagrass (Zoysia joponica),

creeping bentgrass (Agrostis palustris), and bermudagrass (Cynodon dactylon) also

revealed no differences in water extraction by treatment (data not shown), although these

studies were conducted in columns in a greenhouse with no attempt at maintaining cool

“soil-like” temperatures for the below-surface portion of the columns.

Water is lost from the soil through evaporation from the soil surface, transpiration

through the plant, and percolation below the root zone. Field research (Tisdale et al.,

1999) indicates that a lower leaf area index (LAI) increases the percentage of soil

exposed to direct sunlight, and that this condition allows a considerable amount of water

to evaporate from a moist soil. Conversely, a high LAI produces a dense canopy over the

soil that shades the soil surface, allowing less evaporative energy to reach the soil

surface. Soil temperatures are reduced and the crop provides insulation to maintain higher

humidity just above the soil due to less air movement. The observation that more LAI is

correlated with less water loss is dependent on the fact that transpiration per unit surface

area is a fraction of evaporation. The authors state, “Close rows and adequate stands,

along with adequate nutrition, provide a heavy crop canopy. For example, water use

120

would be less in 21-in rows than in 42-in rows.” Thus, in column studies with similar

LAI, differences in evapotranspiration may not be observable through gravimetric

measurements of water loss even though there may be significant differences in root

density or overall root mass. ANOVA analysis of column shoot mass and root:shoot

ratios revealed no significant difference in the total shoot mass grown over the period of

the study between treatments or between soils. Since total shoot biomass was not

significantly different between treatments it can be assumed that no real differences in

LAI existed between treatments. Furthermore, there was no significant difference in

root:shoot ratio between treatments or soils, and no significant interaction between these

main effects. The lack of a significant difference in LAI may help explain the lack of

difference in column water extraction between treatments.

Soil exchangeable Al and water stress are known to have a significant interaction

on root growth of some forage crops. Krizek and Foy (1988) grew barley (Hordeum

vulgare) cultivars known to be either Al-sensitive or Al-resistant in an acid soil and found

optimal root growth for both varieties under low Al and low water stress, and lower but

not significantly different between cultivars root growth in low Al and high water stress.

Under high Al and low water stress the Al-sensitive variety had significantly reduced root

matter, and under combined high Al and water stress both cultivars had significantly

reduced root dry matter, indicating a synergistic effect between the two stress treatments.

121

Conclusions

Soil-incorporated lime significantly improved tall fescue root density in acid soils

that were originally subsoils compared to untreated columns, as expected. The

improvement in overall root density in gypsum-treated soils was not as great as soil-

incorporated lime, however lime incorporation into acid subsoils under established

turfgrass is highly problematic. There were significant differences in root density values

between soils, indicating potential differences in tall fescue response to gypsum-

ameliorated acid soils based on soil type. Calcium and sulfate as gypsum moved well

through the soil column reducing exchangeable acidity and increasing exchangeable Ca.

While gypsum offers a practical method for improving acid subsoil chemistry and

marginal improvements in tall fescue root density, gypsum or lime-treated soils had no

significant effect on rates of evapotranspiration, probably due to similar leaf area indexes

of the tall fescue plants. Root density improvements in treated soils did not impact

drought resistance in a measurable way for column-grown tall fescue.

122

References Cited

Adams, F. 1984. Crop responses to lime in the Southern United States. P. 211-265. In F.

Adams (ed.) Soil acidity and liming. Agron. Monogr. 12. 2nd ed. ASA, CSSA,

SSSA, Madison, WI.

Alva, A.K., G.L. Kerven, D.G. Edwards, and C.J. Asher. 1991. Reduction in toxic

aluminum to plants by sulfate complexation. Soil Sci. 152: 351-359.

Beard, J.B. 1973. Turfgrass: Science and culture. p. 340. Prentice Hall, Inc. Engelwood

Cliffs, NJ.

Clark, R.B., and V.C. Baligar. 2003. Growth of forage legumes and grasses in acidic soil

amended with flue gas desulfurization products. Comm. Soil Sci. Plant Anal. 34:

157-180.

Farina, M.P.W., and P. Channon. 1988. Acid-subsoil amelioration: I. A comparison of

several mechanical procedures. Soil Sci. Soc. Am. J. 52: 169-175.

Gascho, G.J., and M.B. Parker. 2001. Long-term liming effects on Coastal Plain soils and

crops. Agron. J. 93: 1305-1315.

Hue, N.V., and D.L. Licudine. 1999. Amelioration of subsoil acidity through surface

application of organic manures. J. Environ. Qual. 28: 628-632.

Klute, A. 1986. Water retention, Laboratory Methods. p. 643. In Physical and

Mineralogical Methods. Am. Soc. Agron., Madison, WI.

Krizek, D.T., and C.D. Foy. 1988. Role of water stress in differential aluminum tolerance

of two barley cultivars grown in an acid soil. J. Plant Nutr. 11: 351-367.

123

Levy, G.J., and M.E. Sumner. 1998. Mined and by-product gypsum as soil amendments

and conditioners. p. 187-215. In A. Wallace and R.E. Terry (eds.) Handbook of

soil conditioners. Marcel Dekker, Inc. New York.

Liu, H., J.R. Heckman, and J.A. Murphy. 1997. Greenhouse screening of turfgrasses for

aluminum tolerance. Int. Turfgrass Soc. Res. J. 8: 719-728.

Liu, J., and N.V. Hue. 2001. Amending subsoil acidity by surface applications of

gypsum, lime, and composts. Comm. Soil Sci. Plant Anal. 32: 2117-2132.

Lo Bianco, R., M. Rieger, and S.S. Sung. 2000. Effect of drought on sorbitol and sucrose

metabolism in sinks and sources of peach. Physiologia Plantarum 108: 71-78.

Matsumoto, H. 2000. Cell biology of aluminum toxicity and tolerance in higher plants.

Int. Review of Cytology 200: 1-46.

McCray, J.M., M.E. Sumner, D.E. Radcliffe, and R.L. Clark. 1991. Soil Ca, Al, acidity

and penetration resistance with subsoiling, lime and gypsum treatments. Soil Use

and Management 7: 193-199.

McLean, E.O. 1982. Soil pH and lime requirement. In (A.L. Page et al., ed.) Methods of

Soil Analysis, Part 2. Agronomy Monograph 9(2): 199-209, ASA-SSSA,

Madison, WI.

Miyazawa, M., M.A. Pavan, and J.C. Franchini. 2002. Evaluation of plant residues on the

mobility of surface applied lime. Brazil. Arch. Biol. Tech. 45: 251-256.

Murray, J.J., and C.D. Foy. 1978. Differential tolerances of turfgrass cultivars to

an acid soil high in exchangeable aluminum. Agron. J. 70: 769-774.

O’Brian, L.O., and M.E. Sumner. 1988. Effects of phosphogypsum on leachate and soil

chemical composition. Comm. Soil Sci. Plant Anal. 19: 1319-1329.

124

Palazzo, A.J., and R.W. Duell. 1974. Response of grasses and legumes to soil pH. Agron.

J. 66: 678-682.

Pavan, M., F. Bingham, and P.F. Prat. 1982. Toxicity of aluminum to coffee in Ultisols

and Oxisols amended with CaCO3, MgCO3, and CaSO4·2H2O. Soil Sci. Am. J.

46: 1201-1207.

Radcliffe, D.E., R.L. Clark, and M.E. Sumner. 1986. Effect of gypsum and deep-rooting

perennials on subsoil mechanical impedance. Soil Sci. Soc. Am. J. 50: 1566-1570.

Reeve, N.G., and M.E. Sumner. 1972. Amelioration of subsoil acidity in Natal Oxisols by

leaching of surface-applied amendments. Agrochemophysica 4: 1-6.

SAS Institute. 1991. SAS/STAT users guide. Ver 6. 4th ed. SAS Institute, Cary, NC.

Shainberg, I., M.E. Sumnerm W.P. Miller, M.P.W. Farina, M.A. Pavan, and M.V. Fey.

1989. Use of gypsum on soils: A review. p. 1-111. In B.A. Stewart (ed.)

Advances in soil science Vol. 9. Springer-Verlag, New York.

Stehouwer, R.C., P. Sutton, and W.A. Dick. 1996. Transport and plant uptake of soil-

applied dry flue gas desulfurization by-products. Soil Sci. 161: 562-574.

Sumner, M.E. 1990. Gypsum as an ameliorant for the subsoil acidity syndrome. Florida

Institute of Phosphate Research, Bartow, FL.

Sumner, M.E., 1995. Amelioration of subsoil acidity with minimum disturbance. p. 147-

185. In N.S. Jayawardane and B.A. Stewart (eds.) Subsoil management

techniques. Lewis Publishers, Boca Raton, FL.

125

Sumner, M.E., and W.P. Miller. 1996. Cation exchange capacity and exchange

coefficients. p. 1201-1230. In Sparks et al. (Eds.) Methods of soil analysis: Part 3

– Chemical methods. Soil Sci. Soc. Am. Series No. 5. Soil Sci. Soc. Am.,

Madison, WI.

Tisdale, S.L., J.L. Havlin, J.D. Beaton, and W.L. Nelson. 1999. Soil fertility and

fertilizers. Prentice Hall, Inc., Upper Saddle River, NJ. p. 408.

Wang, H.L., M.J. Hedley, N.S. Bolan, and D.J. Horne. 1999. The influence of surface

incorporated lime and gypsiferous by-products on surface and subsurface soil

acidity. II. Root growth and agronomic implications. Aust. J. Soil Res. 37: 181-

190.

126

Table 5.1. Initial soil physical and chemical characteristics.

Soil Cunningham f. sandy loam

Cecil sandy clay

% sand 54 46 % silt 26 10 % clay 20 44 CEC (cmol(+) kg-1) 3.66 3.26 Exch. acidity (cmol(+) kg-1) 2.84 1.01 % Base saturation 22 62 pHH2O 4.6 4.8 Buffer capacity (mg CaCO3 kg-1 pH-1) 2214 3255

Munsell color 2.5Y 7/4 5YR 5/8

127

Table 5.2. Soil column Ca, Mg, and K after conclusion of experiment by depth and treatment. Tukey’s mean separation of significant treatment differences (α=0.05) designated by letters for each depth. Treatment 0-8 cm 8-16 cm 16-27 cm Cunningham sandy loam -------- Ca (cmol(+) kg-1) -------- Lime 3.54 a 3.59 a 4.26 a Gypsum 1.83 b 1.83 b 1.67 b Control 0.28 c 0.71 c 0.98 c Cecil sandy clay Lime 3.47 a 4.16 a 4.37 a Gypsum 3.12 b 2.80 b 3.22 b Control 1.16 c 1.79 c 2.37 c Cunningham sandy loam -------- Mg (cmol(+) kg-1) -------- Lime 0.07 a 0.06 a 0.07 ab Gypsum 0.01 b 0.01 b 0.01 b Control 0.08 a 0.09 a 0.08 a Cecil sandy clay Lime 0.14 a 0.16 a 0.17 ab Gypsum 0.02 b 0.07 b 0.14 b Control 0.27 a 0.17 a 0.25 a Cunningham sandy loam --------- K (cmol(+) kg-1) --------- Lime 0.06 a 0.06 a 0.08 a Gypsum 0.05 a 0.07 a 0.07 a Control 0.07 a 0.08 a 0.08 a Cecil sandy clay Lime 0.04 a 0.05 a 0.06 a Gypsum 0.04 a 0.05 a 0.06 a Control 0.05 a 0.05 a 0.05 a Cunningham sandy loam ----- Exch. acidity (cmol(+) kg-1) ----- Lime 0.13 c 0.33 b 0.06 b Gypsum 2.37 b 2.37a 2.33 a Control 3.26 a 2.92 a 2.38 a Cecil sandy clay Lime 0.03 c 0.03 b 0.03 b Gypsum 0.56 b 0.62 a 0.46 b Control 1.45 a 0.95 a 1.21 a Cunningham sandy loam ----- CEC (cmol(+) kg-1) ----- Lime 3.79 b 4.05 a 4.47 a Gypsum 4.26 a 4.28 a 4.08 ab Control 3.69 b 3.79 a 3.52 b Cecil sandy clay Lime 3.69 ab 4.40 a 4.62 a Gypsum 3.74 a 3.55 ab 3.87 b Control 2.92 b 2.96 b 3.88 b

128

Table 5.3. Mean root density values (mg kg-1) for treatment and soil types. Letter designations indicate significant differences between treatments or soils at each depth based on Tukey’s mean separation (α=0.05) for each depth.

Treatment 0-8 cm 8-16 cm 16-27 cm Total column Lime 2017 a 745 a 576 a 3338 a Gypsum 2046 a 364 b 372 b 2782 b Control 1353 b 419 b 330 b 2102 c Soil 0-8 cm 8-16 cm 16-27 cm Total column Sandy loam 1585 b 410 b 253 b 2248 b Sandy clay 2025 a 609 a 599 a 3233 a

129

Column height: 34 cm

soil

sand

sandgravel

A/C unit

Figure 5.1. Profile of column compartment for maintaining 15o C soil temperature.

130

EC (μS cm-1)

Cunningham sandy loam

0 50 100 150 200 250 300

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Cecil sandy clay

0 50 100 150 200 250 300

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Figure 5.2. Electrical conductivity (EC) of soils (2.5:1 water:soil) from Cecil and Cunningham soils as a function of treatment at conclusion of experiment. Error bars indicate standard error of mean values.

131

pHH2O

Cunningham sandy loam

4.5 5.0 5.5 6.0

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Cecil sandy clay

4.5 5.0 5.5 6.0

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Figure 5.3. Soil pHH2O values for Cecil and Cunningham soils from columns sampled after conclusion of experiment. Error bars indicate standard error of mean values.

132

pHKCl

Cunningham sandy loam

4.0 4.5 5.0 5.5 6.0

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Cecil sandy clay

4.0 4.5 5.0 5.5 6.0

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Figure 5.4. Soil pHKCl values for Cecil and Cunningham soils from columns sampled after conclusion of experiment. Error bars indicate standard error of mean values.

133

Exchangeable acidity (cmol(+) kg-1)

Cunningham sandy loam

0 1 2 3 4

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Cecil sandy clay

0 1 2 3 4

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Figure 5.5. Exchangeable acidity of sandy loam and sandy clay soils from columns sampled after conclusion of experiment. Error bars indicate standard error of mean values.

134

Extractable Ca (cmol(+) kg-1)

Cunningham sandy loam

0 1 2 3 4 5

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Cecil sandy clay

0 1 2 3 4 5

Dep

th (c

m)

0

5

10

15

20

limegypsumcontrol

Figure 5.6. Extractable soil Ca of sandy loam and sandy clay soils from columns sampled after conclusion of experiment. Error bars indicate standard error of mean values.

135

sandy loam sandy clay

Ca

leac

hate

(mg

L-1)

0

20

40

60

80

100

120

140

160

180

limegypsumcontrol

Mg

leac

hate

(mg

L-1)

0

1

2

3

4

5

6

limegypsumcontrol

K le

acha

te (m

g L-1

)

1

2

3

4

5

6

limegypsumcontrol

Figure 5.7. Soil Ca, Mg, and K leachate concentrations (mg L-1) collected from columns 30 d after planting. Error bars indicate standard error of mean values.

136

Cecil sandy clay

0 1000 2000 3000 4000

Col

umn

dept

h (c

m) total column

0-8

8-16

16-27

limegypsumcontrol

a

b

aab

ab* (α=0.10)

a

b

bb

aa

a

Cunningham sandy loam

Root density (mg kg-1)

0 1000 2000 3000 4000

Col

umn

dept

h (c

m) total column

0-8

8-16

16-27limegypsumcontrol

aab

b

ca

b

a

bc

aab

b

% of total 63 62 63

% of total 57 90 66

21 17 19

16 21 18

25 8 22

18 2 12

Figure 5.8. Root density values for Cecil and Cunningham soils at conclusion of experiment. Letter differences indicate significant differences by treatment for each depth based on Tukey’s mean separation (α=0.05).

137

0 20 40 60 80 1g H

2O g

-1so

il00

0.00

0.05

0.10

0.15

0.20

limegypsumcontrol

Days0 20 40 60 80 100

g H2O

g-1

soil

0.00

0.05

0.10

0.15

0.20

limegypsumcontrol

Cecil sandy clay

Cunningham sandy loam

Figure 5.9. Gravimetric water loss from soil columns during drydown procedure. Error bars indicate standard error of mean values.

138

Cecil sandy clayco

lum

n w

ater

loss

(gH

2O d

-1)

10

20

30

40

limegypsumcontrol

Cunningham sandy loam

Measurement day20 30 40 50 60 70 80 90 100

Col

umn

wat

er lo

ss (g

H2O

d-1

)

0

10

20

30

40

limegypsumcontrol

Figure 5.10. Column water loss on a per day basis by treatment. Error bars indicate standard error of mean values.

139

CHAPTER 6

WARM-SEASON TURFGRASS RESPONSE TO AMELIORATION OF SUBSOIL

ACIDITY IN THE FIELD USING LIME AND GYPSUM

______________________________________________________ Kruse, J.S., W.P. Miller, and M.E. Sumner, E.G. To be submitted to Hortscience

140

Abstract

A field study was conducted to quantify the effects of lime, gypsum, and

combinations of both on the amelioration of subsoil acidity, and determine the effect

these amendments would have on turfgrass root growth and subsoil-water use. A

complete-factorial arrangemetn was used for zoysiagrass (Zoysia japonica v. ‘Meyer’)

and hybrid bermudagrass (Cynodon dactylon v. ‘Tifsport’) grown on an acid Cecil soil

(fine kaolinitic, thermic Typic Kanhapludult). Lime, gypsum, lime + gypsum at two rates

were applied to the soil followed by sod establishment. Soil moisture was monitored at 3

depths using in-situ time domain reflectometry (TDR) probes. Soil samples for root

density and soil chemical analysis were extracted after 15 months. Gypsum moved

through the soil profile to a depth of 40 cm. Combinations of lime + gypsum had the

greatest effect on EC, pH, CEC, exchangeable Ca and exchangeable acidity. Significant

differences in root density by treatment were more pronounced in zoysiagrass than

bermudagrass, with the lime + low gypsum treatment indicating highest overall increases

over the control treatment: 213% root-growth response for zoysiagrass, and an

insignificant 24% response in bermudagrass. Due to a wet growing season, differences in

volumetric water content were slight, although greater in zoysiagrass than bermudagrass.

The largest differences in soil water content for zoysiagrass were between control plots

and lime + gypsum treatments at 20-58 cm depth, indicating improved root colonization

in the amended subsoils.

.

141

Introduction

Highly weathered soils that have elevated levels of exchangeable Al and are low

in CEC, pH, and base saturation are known to restrict the root growth and yield of many

agronomic crops (Adams and Pearson, 1970; Barszczak et al., 1993; Foy, 1997; Carter

and Richards, 2000). Turfgrass resistance to subsoil acidity presents a unique situation in

agronomy because of different management regimes and grower expectations than

agricultural crops. Turfgrass managers do not value yield (mower clippings) as much as

color, density, and uniformity. An emphasis on deep-rooting cultivars has been subdued

in the past, due to the relatively low cost and reliable supply of irrigation water. Turfgrass

species, and cultivars within species, have different levels of tolerance to Al toxicity

(Murray and Foy, 1978; Andersson, 1988; Foy, 1992; Kochian et al., 2004; Liu, 2005),

leaving many shallow-rooted varieties vulnerable to dry soil conditions. Increased

demands on water supplies driven by rapid urbanization has led some researchers to

focus on plant breeding that emphasizes deeper rooting, drought tolerance, and/or acid

resistance (Qian et al., 1997; Duncan and Carrow, 2001).

Subsoils in the Southeastern U.S. are often highly acidic, making a hostile rooting

environment for many plant species that restricts the uptake of water and nutrients

(Sumner et al., 1990). Turfgrass species susceptible to soil acidity have shallow root

systems that fail to penetrate acid subsoils high in exchangeable Al and therefore fail to

utilize soil water available at lower depths (Huang et al., 1997). This leaves many species

susceptible to drought, as surface soils dry in warmer months as evapotranspiration

exceeds precipitation.

142

Amelioration of subsoil acidity has rarely been investigated on grasses, but when

it has been studied it was mostly in the context of forages. Following work by Adams et

al. (1967) on bermudagrass that investigated the effects of lime applications to topsoil,

Adams and Pearson (1969) conducted a landmark study in which they attempted to

ameliorate subsoil acidity under bermudagrass sod with various applications of lime,

sodium nitrate, calcium nitrate, or calcium gluconate. They demonstrated several

important mechanisms, including subsoil liming by the application of lime at rates that

exceed N application rates, the utility of NO3- fertilizer to boost subsoil pH, and the use

of chelated-Ca to effect the downward mobility of Ca2+ ions in the soil. Mora et al.

(2002) amended pastures containing ryegrass and white clover with lime and gypsum and

found the combination of lime and gypsum improved yield by 50% over the control.

Most of the improvement was attributed to the response of ryegrass coverage increasing

from 13 to 75% at the expense of pasture weeds (unidentified). Al saturation in the top 20

cm decreased from 20 to less than 1%, and the soil pH was slightly raised. The use of

only 3 exclusion cages per treatment created much variability in the study. Hoveland

(2000) reviewed major achievements in grassland management and included fertilizer

and lime use as one of the keys to forage management improvements.

Gypsum has played a limited role in supplying nutrients to forage- and turf-

grasses. The application of gypsum on S-deficient soils increases N uptake in

bermudagrass and ryegrass (Bailey, 1992; Phillips and Sabbe, 1994), although the studies

do not conclusively demonstrate a mechanism by which this occurs. Possible options

include increased soil Ca2+ allowing roots to take up larger quantities of NO3-N and/or

SO42- performing the same role for NH4-N; or increased root exploration due to soil

143

acidity amelioration by gypsum may increase total N uptake. Using gypsum as a Ca-

source, Kuo (1993) demonstrated the importance of sufficient soil Ca for improving

bentgrass growth in acid soil containing low initial Ca.

The purpose of this study was to quantify the effect of lime and/or gypsum on the

amelioration of subsoil acidity, and determine the effect these amendments would have

on turfgrass subsoil-water use in dry periods and turfgrass root growth.

Materials and Methods

Location and materials

A field experiment was conducted to examine the effects of gypsum on warm-

season turfgrass rooting in an acid subsoil. The site was located at the University of

Georgia Experiment Station at Griffin, in the Piedmont region, approximately 40 miles

south of Atlanta, Georgia. The soil was previously described as a Cecil sandy clay loam

(fine kaolinitic, thermic Typic Kanhapludult) (Shim and Carrow, 1997). Selected soil

characteristics were measured for this experiment and are listed in Table 6.1. Each plot

was 5 x 5 m, with irrigation sprinklers at the corners of each plot. The study was designed

as a complete factorial with two turf species: ‘Meyer’ zoysiagrass and ‘Tifsport’ hybrid

bermudagrass. Treatments included a non-treated control, dolomitic lime-only applied at

a rate of 3.25 Mg ha-1, gypsum-only applied at a rate of 5.5 Mg ha-1, gypsum-only applied

at a rate of 11 Mg ha-1, dolomitic lime plus gypsum at the small rate, and dolomitic lime

plus gypsum at the large rate. All treatments were randomly assigned to plots and

replicated three times.

144

The gypsum source was flue-gas desulfurization gypsum from a coal-fired

electrical generating station southwest of Atlanta, Georgia. In the forced oxidation wet

scrubber, SO2-containing stack gas from the burning of coal are mixed with a CaCO3

slurry resulting in a CaSO4·2H2O by-product. The gypsum was analyzed using EPA

Method 6020 (EPA, 1994) on a Perkin-Elmer Elan 9000 ICP-MS (Perkin Elmer, Inc.,

Wellesley, MA), obtaining results of 97.5% CaSO4·2H2O, and less than 1% CaCO3.

Site preparation and maintenance

Plots were tilled twice with a walk-behind rototiller, to a depth of 10 cm.

Dolomitic lime was applied to the appropriate plots with a calibrated rotary push-

spreader, and all plots rototilled a third time. Pre-weighed gypsum was spread by hand on

the appropriate plots, followed by an application of 20-27-5 fertilizer (urea, methylene

urea, MAP, potassium chloride) at the rate of 14 g m-2. In August, 2003 sod was placed

by hand on each plot area and watered immediately. Sod survival was 100%.

Turfgrass plots were maintained over 15 months by regular, periodic mowing at a height

of 2.5 cm and periodic irrigation until dormancy and after initial spring green-up.

Cumulative rainfall was recorded by an automated system located 120 m from the site

(www.Georgiaweather.net; Hoogenboom et al., 2003). Pesticides were applied as needed,

with all plots receiving the same treatments. A 40-0-0 slow-release nitrogen fertilizer

(urea, methylene urea) was applied at 3 to 8-week intervals during the growing season, at

the rate of 49 kg N ha-1 application-1.

145

TDR installation and measurement

Access boxes made of treated plywood and measuring 0.6 m long, 0.3 m wide,

and 0.3 m deep, were inserted into each plot such that the top surface of the box was

coplanar with the soil surface, to allow access to the soil profile to a depth of 0.3 m. One

vertical box wall was removed to allow access to the soil profile at one end of the box. A

fitted wooden lid was placed over the box when access to the site was not needed. Two

parallel aluminum rods with a diameter of 6.35 mm and 38.1 cm in length were inserted

into the soil horizontally at 5 and 15 cm depths leaving the ends exposed. Identical rods

that were 61 cm in length were inserted beginning at 20.3 cm at a 35o angle to the soil

profile to a depth of 61 cm, leaving 5 cm of rod exposed for reading access. Soil moisture

measurements were made by TDR (Topp et al., 1980) using a Tektronix 1502C Metallic

TDR Cable Tester (Tektronix, Inc., Beaverton, OR). Signal output was converted to soil

moisture content by setting the velocity of propagation relative to the speed of light in a

vacuum (Vp) to 0.66, moving the cursor to the top of the first peak and recording distance

displayed, then moving the cursor to the point where the signal starts to rise, and

recording the distance. The first distance was subtracted from the second distance to get

the apparent wave-guide length (La), and inserted into the following equation to calculate

the dialectric (ka)

ka = (La / Vp*L)2

where L is the actual length of the waveguide. The dialectric was converted to volumetric

water content using Topp’s equation (Topp et al., 1980).

θ = -5.3*10-2 + 2.92*10-2*ka – 5.5*10-4*ka2 + 4.3*10-6*ka

3

146

Measurements were taken on six occasions throughout the summer and fall of 2004, prior

to turfgrass dormancy and after periods of drought, again after dormancy to establish

whether soil moisture was consistent across all treatments and replicates during a period

of little evapotranspiration, and again during the growing season in summer of 2005.

Moisture content values were compared on a relative basis to each of the treatments using

the PROC GLM procedure (SAS, 1998).

Soil sampling

Soil samples were taken in November with a Giddings 5-cm diameter soil core

probe (Giddings Machine Co., Windsor, CO) at 0-15, 15-30, 30-45, and 45-71 cm depths.

Three cores were taken from each plot and composited, weighed, and maintained at 0o C

until wet sieving to catch root mass on a 2-mm sieve. Root masses were air-dried and

weighed to ± 0.1 mg. Analysis was performed on soil subsamples to determine pHKCl,

CEC, base saturation, and extractable Al. The pHKCl and extractable Al and H were

determined using 1 M KCl, with quantification of exchangeable acidity by titration with

0.005 M NaOH to a phenolphthalein endpoint. Base cations were extracted with 1 M

NH4OAc and measured by Flame Atomic Absorption Spectroscopy (Aanalyst 200 atomic

absorption spectrometer, Perkin Elmer, Inc., Wellesley, MA). CEC and base saturation

were calculated from the summation of extractable cations. The water content of the soils

at field capacity (-0.01 MPa) and wilting point (-1.5 MPa) was determined in a pressure

chamber using methods described by Klute (1986).

147

Results and Discussion

Soil chemical changes

Gypsum moved down the soil profile to 40-cm depth over a 15-month period

(Table 6.1). Zoysiagrass electrical conductivity (EC) measurements indicated

significantly greater salt levels in the high gypsum and both lime + gypsum applications

at the 0-10 cm, 10-20 cm, and 20-40 cm depths compared to the control and lime

applications. Measured EC under bermudagrass did not show as clear a contrast between

treatments. The high gypsum and lime + high gypsum applications had elevated EC

levels at the surface and bottom depths – but not at 10 to 20 cm - compared to the control,

lime, low gypsum, and lime + low gypsum applications.

Soil pH levels rose in the lime and gypsum applications (Table 6.1), indicating

some evidence of the self-liming effect (Reeve and Sumner, 1972). The pH values

measured in 1 M KCl (to eliminate salt effect from gypsum on pH readings) showed

almost a full unit increase in pH at the surface from lime applications, with greater values

in soils treated with lime and gypsum than lime or gypsum alone. Lime alone improved

soil pH at the 10-20 cm depth for bermudagrass but not zoysiagrass, whereas lime +

gypsum applications or gypsum alone provided a 0.1-0.3 unit pH change or none at all

over the control. Means separation using Fisher’s least significant difference revealed the

lime + gypsum applications had overall significantly higher pHKCl values than other

treatments, and the lime treatment was significantly higher than the control and gypsum-

only treatments. The surface depth for all treatments was significantly higher than the

lower depths, which did not differ significantly from each other. There was no significant

difference in pHKCl between turf species.

148

Exchangeable acidity was low in all treatments at the surface 0-10 cm, ranging

from 1 to 4.9% of CEC (Table 6.2.) but significant differences occurred between the

control treatment and all other treatments except low-gypsum. At 10-20 cm the control

treatment had the highest average exchangeable acidity and was significantly greater as a

% of CEC compared to the lime + gypsum applications. At 20-40 cm, field variability

masked any potential differences in exchangeable acidity between treatments.

Under zoysiagrass plots at 10-20 cm, the low gypsum application was not

significantly different from the control plots, but the high gypsum application

significantly increased exchangeable acidity to 35% of CEC, and to 36% at 20-40 cm.

Conversely, lime + gypsum applications significantly reduced exchangeable acidity at

10-20 and 20-40 cm compared to control and lime-only treatments.

Near-surface exchangeable acidity was also low in bermudagrass plots (<5% of

CEC), but all non-control treatments reduced exchangeable acidity by half or more at the

10-20 cm depth (Table 6.2). There were no significant reductions in exchangeable acidity

at 20-40 cm compared to the control treatment, indicating lime had not moved in

significant enough quantities to that depth to affect levels of exchangeable acidity.

Soil Ca levels in zoysiagrass plots indicate both lime + gypsum treatments

significantly increased Ca over control, lime-only, and gypsum-only treatments at 10-20

and 20-40 cm depths (Table 6.2). All treatments were significantly higher in Ca in the top

10 cm compared to control plots (Figure 6.1a). There was no difference in Ca levels

between the control and lime-only treatments at 10-20 and 20-40, but Ca in the lime +

gypsum application was double the control and lime-only levels.

149

Bermudagrass soil Ca levels increased in all treatments, except low gypsum, over

the control in the top 10 cm (Table 6.2). All treatments were higher than the control at

10-20 and 20-40 cm depth, although not nearly to the extent as the same treatments under

zoysiagrass. Figure 6.1b shows Ca concentrations at various depths compared to the

control plots, indicating the minor increase in Ca that these treatments had at the lower

depths.

Turfgrass rooting

Total root density summed over the three depths for both grass species showed no

significant differences between treatments; however root density by depth and species

revealed a shift in root distribution for some treatments compared to the control (Table

6.3). Total root density was analyzed using 2-Way ANOVA (α=0.10) and no significant

interaction existed between grass-type and treatment, thus allowing for investigation of

main effects. As expected, there was a significant difference in root density between

bermudagrass and zoysiagrass, with mean root values for all treatments of 661 mg kg-1

and 306 mg kg-1 respectively. Zoysiagrass plots with lime, gypsum, or a combination all

had significantly improved root density over the control plots (α=0.05), and there were

significant differences between treatment applications at the α=0.10 level. The lime plus

low gypsum application had the highest overall mean root density in zoysiagrass and was

significantly better than the control, lime, and lime plus high gypsum applications.

Bermudagrass overall root density did not respond to treatments (Pr>F=0.8307). Other

treatments not mentioned in the previous analyses showed no significant improvement in

overall root density compared to the control treatment in either turfgrass species.

150

Differences in root density at various depth ranges indicated zoysiagrass

responded positively to the combination of lime plus low rate of gypsum (Table 6.3). At

0-15 cm depth, root density for this treatment produced twice the root density of the lime-

only treatment, and quadruple the root density over the control. Even at 15-30 cm depth,

mean root density doubled over the control and lime-only treatments, even though this

represented a small fraction of the root density in the surface horizon. Root densities were

greater for the lime plus low gypsum application than the lime only and control

treatments at 30-45 cm and 45-70 cm, although overall root mass was slight at these

lower depths. Bermudagrass root density improved 33 and 39% over the lime-only and

control applications in the top 15 cm (Table 6.3), and was greater than control or lime-

only treatments at 15-30 cm depth, although field variability negated statistically

significant differences. From 30-45 cm, no significant difference in root density between

treatments was measured, and at 45-70 cm no treatment had significantly higher root

density than the control.

Soil moisture results

There were significant differences in soil volumetric water content in zoysiagrass

plots at the 20-58 cm depth (Table 6.4) on two dates. Six separate readings taken during

the summer and autumn registered less water content at these lower depths on the lime +

low gypsum treatment compared to control plots (Table 6.4). Differences in water content

correlated with significantly increased root density for this treatment, and indicated more

rapid transpiration occurred in plots where more of the soil volume was exploited by

roots. As zoysiagrass became dormant, differences in water content between treatments

151

disappeared and by late winter total water content had increased, indicating dormant

zoysiagrass was not transpiring water. It is not clear why the soil was significantly more

dry at 20-58 cm depth for the lime + low gypsum treatment and yet most of the increased

root mass occurred in the top ten cm for that treatment.

No significant differences in soil water content were detected in bermudagrass

plots at 0-10, 10-20, or 20-58 cm. This lack of difference agrees to some extent with less

difference in root density between treatments, although the lime + low gypsum treatment

had significantly more roots than the control plots at the 15-30 depth. Overall root density

may play a larger role than root density at individual depths.

ANOVA analysis (SAS, 1998) revealed zoysiagrass had significantly drier soil

than bermudagrass at all three depths during the growing season, indicating more rapid

transpiration took place in the zoysiagrass plots. Treatments were not significantly

different in the 0-10 and 10-20 cm depths for either grass type. When data from both

turfgrasses were combined for analysis, differences at the 20-58 cm depth were

measured. The two lime + gypsum plots were significantly more dry than the control,

lime-only, and low gypsum treatments. The lsmean separation signified most of this

difference occurred between treatments in the zoysiagrass plots.

Conclusions

Treatments that contained gypsum without lime did not increase soil Ca or lower

exchangeable acidity at lower depths. Root density was not appreciably greater for these

treatments over control plots. Lime-only applications increased soil Ca in the surface 10

cm, but a root-growth response occurred only in the zoysiagrass. Calcium levels did not

152

increase to an appreciable extent in the lower depths, and exchangeable acidity was

generally the same as the control plots, except for the bermudagrass plots at 10-20 cm,

which cut acidity in half. Concurrently, root density below 15 cm was no different than

the controls.

Lime + gypsum applications showed a marked improvement in soil chemical

conditions, with higher soil Ca and lower exchangeable acidity at lower depths, with

improved topsoil conditions, including higher CEC values compared to the control

treatments. The lime + low gypsum applications had higher root densities than control

plots in zoysiagrass and insignificant differences in bermudagrass. The lime + low

gypsum application was superior to the lime + high gypsum application in producing a

higher root density that correlated with improved water extraction.

The combination of lime incorporated into topsoil and surface application of

gypsum may be a method of subsoil acidity amelioration for zoysiagrass, with an overall

214% root growth response over the control, and 100% root growth response over the

lime-only application. Root growth response in these treatments correlated well with

increased water extraction under dry conditions. Bermudagrass did not respond to

gypsum or lime + gypsum treatments to the extent of zoysiagrass, with a 24% overall

increase in root density compared to the control, and a 19% growth response compared to

lime, both of which were statistically insignificant at α = 0.10. No significant differences

(α = 0.10) in water extraction were detected in bermudagrass plots. This cultivar of

bermudagrass, with its high tolerance for exchangeable acidity and ability to extract

exchangeable Ca from the soil, appears an unlikely candidate for root growth

improvement utilizing gypsum-ameliorated subsoils in the Piedmont. The lack of

153

response to the treatments, however, may be an artifact of this particular cultivar. Hays et

al. (1991) simulated drought stress in 10 bermudagrass cultivars by using vertical,

sectioned PVC pipes in a greenhouse and gradually lowering watering depth in the pipes,

and found significant differences in root distribution. Recently, Liu (2005) reported

significant differences in resistance to Al toxicity between 16 bermudagrass cultivars

grown in solution culture and on an acid Tatum soil (Clayey, mixed, thermic, typic

Hapludult). Because bermudagrass is the predominant turfgrass in golf course fairways

and athletic fields in the Southeastern U.S., further study of the potential to ameliorate

subsoil acidity on a wide selection of bermudagrass cultivars is warranted. Moreover, the

decision to use zoysiagrass, known for its shallow-rooted traits on acid soils, may

increase as turfgrass managers become aware of its positive response to a combination

lime and gypsum application.

154

References Cited

Adams, F., and R.W. Pearson. 1969. Neutralizing soil acidity under bermudagrass sod.

Soil Sci. Soc. Am. Proc. 33: 737-742.

Adams, F., and R.W. Pearson. 1970. Differential response of cotton and peanuts to

subsoil acidity. Agron. J. 62: 9-12.

Adams, W.E., A.W. White, Jr., and R.N. Dawson. 1967. Influence of lime sources and

rates on ‘Coastal’ bermudagrass production, soil profile reaction, exchangeable

Ca and Mg. Agron. J. 59: 147-149.

Andersson, M. 1988. Toxicity and tolerance of aluminum in vascular plants. Water, Air,

and Soil Pollution 39: 449-462.

Bailey, J.S. 1992. Effects of gypsum on the uptake, assimilation and cycling of N-15

labeled ammonium and nitrate-N by perennial ryegrass. Plant Soil 143: 19-31.

Barnhisel, R., and P.M. Bertsch. 1982. In (A.L. Page et al., ed.) Methods of

Soil Analysis, Part 2. Agronomy Monograph 9(2): 281-283, ASA-SSSA,

Madison, WI.

Barszczak, Z., T. Barszczak, and C.D. Foy. 1993. Effect of moisture, nitrogen rates, and

soil acidity on seed yield and chemical composition of winter oilseed rape

cultivars. J. Plant Nutr. 16: 85-96.

Carter, M.R., and J.E. Richards. 2000. Soil and alfalfa response after amelioration of

subsoil acidity in a fine sandy loam Podzol in Prince Edward Island. Can. J. Soil

Sci. 80: 607-615.

155

Duncan, R.R., and R.N. Carrow. 2001. Molecular breeding for tolerance to

abiotic/edaphic stresses in forage and turfgrass. p. 251-260. In G. Spangenberg

(ed.) Molecular breeding of forage crops. Kluwer Acad. Publ., Netherlands.

EPA. 1994. Method 6020: Inductively coupled mass spectrometry. http://www.epa.gov/

epaoswer/hazwaste/test/pdfs/6020.pdf. (Verified 8AUG2005).

Foy, C.D. 1992. Soil chemical factors limiting plant growth. p. 97-149. In J.L. Hatfield

and B.A. Stewart (Eds.) Limitations to plant growth. Adv. Soil Sci. 19.

Foy, C.D. 1997. Tolerance of durum wheat lines to an acid, aluminum toxic subsoil. J.

Plant Nutr. 20: 417-418.

Gillman, G.P. 1979. A proposed method for the measurement of exchange properties of

highly weathered soils. Australian Journal of Soil Research 17: 129-139.

Hays, K.L., J.F. Barber, M.P. Kenna, and T.G. McCollum. 1991. Drought avoidance

mechanisms of selected bermudagrass genotypes. Hortscience 26: 180-182.

Hoogenboom, G., D.D. Coker, J.M. Edenfield, D.M. Evans, and C. Fang. 2003. The

Georgia Automated Environmental Monitoring Network: 10 years of weather

information for water resources management. p. 896-900. In (K.J. Hatcher, ed.)

Proceedings of the 2003 Georgia Water Resources Conference. Institute of

Ecology, The University of Georgia, Athens, Georgia.

Hoveland, C.S. 2000. Achievements in management and utilization of southern

grasslands. J. Range Manage. 53: 17-22.

Huang, B., R.R. Duncan, and R.N. Carrow. 1997. Drought-resistance mechanisms of

seven warm-season turfgrasses under surface soil drying: II. Root aspects. Crop

Sci. 37: 1863-1869.

156

Klute, A. 1986. Water retention, Laboratory Methods. p. 643. In Physical and

Mineralogical Methods. Am. Soc. Agron., Madison, WI.

Kochian, L.V., O.A. Hoekenga, and M.A. Pineros. 2004. How do crop plants tolerate

acid soils? Mechanisms of aluminum tolerance and phosphorus efficiency. Annu.

Rev. Plant Biol. 55: 459-493.

Kuo, S. 1993. Calcium and phosphorus influence creeping bentgrass and annual

bluegrass growth in acidic soils. Hortscience 28: 713-716.

Liu, H. 2005. Aluminum resistance among seeded bermudagrasses. Hortscience 40: 221-

223.

McLean, E.O. 1982. Soil pH and lime requirement. In (A.L. Page et al., ed.) Methods of

Soil Analysis, Part 2. Agronomy Monograph 9(2): 199-209, ASA-SSSA,

Madison, WI.

Mora, M.L., P Cartes, R. Demanet, and I.S. Cornforth. 2002. Effects of lime and gypsum

on pasture growth and composition on an acid Andisol in Chile, South America.

Comm. Soil Sci. Plant Anal. 33: 2069-2081.

Murray, J.J., and C.D. Foy. 1978. Differential tolerances of Turfgrass cultivars to an acid

soil high in exchangeable aluminum. Agron. J. 70: 769-774.

Phillips, J.M., and W.E. Sabbe. 1994. Effects of sulfur on nitrogen recovery in Coastal

bermudagrass as influenced by sulfur fertilization. Comm. Soil Sci. Plant Anal.

25: 1763-1769.

Qian, Y.L., J.D. Fry, and W.S. Upham. 1997. Rooting and drought avoidance of warm-

season turfgrasses and tall fescue in Kansas. Crop Sci. 37: 905-910.

157

Reeve, N.G., and M.E. Sumner. 1972. Amelioration of subsoil acidity in Natal Oxisols by

leaching of surface-applied amendments. Agrochemophysica 4: 1-6.

SAS Institute. 1998. SAS User’s guide: Statistics. Version 7. SAS Institute, Inc., Cary,

North Carolina, USA.

Shim, S.R., and R.N. Carrow. 1997. Cultivation and chemical injection: Influence of soil

physical and chemical properties. Int. Turfgrass Soc. Res. J. 8: 533-540.

Stein, J.K. 1986. Coring archeological sites. American Antiquity 51(3): 505-527.

Sumner, M.E., M.V. Fey, and A.D. Noble. 1990. Nutrient status and toxicity problems in

acid soils. p. 149-182. In B. Ulrich and M.E. Sumner (Eds.) Soil acidity. Springer-

Verlag. Berlin, Germany.

Topp, G. C., J. L. Davis, and A. P. Annan. 1980. Electromagnetic determination of soil

water content: Measurments in coaxial transmission lines. Water Resources

Research 16: 574-582.

www.Georgiaweather.net (Verified 15AUG2005)

158

Table 6.1. Soil chemical conditions 15 months after treatment application.

Zoysia Depth CEC Ca2+ Ex. acidity

0-10 cm EC (dS m-1) pHH2O pHKCl --------cmol+ kg-1-------- Control 110 5.68 5.22 2.72 2.10 0.06 Low gypsum 112 5.80 5.00 3.08 2.82 0.08 High gypsum 137 5.84 5.28 3.53 3.04 0.06 Lime 114 6.04 5.65 3.33 3.28 0.04 L. + Low gyp 152 6.29 5.99 4.00 3.31 0.06 L. + high gyp 133 6.23 6.03 3.57 3.35 0.05

10-20 cm Control 96 4.60 4.34 1.89 1.06 0.43 Low gypsum 99 4.69 4.33 1.99 2.82 0.41 High gypsum 116 4.71 4.12 1.86 1.03 0.65 Lime 96 4.67 4.30 1.67 0.90 0.47 L. + Low gyp 108 4.82 4.60 2.79 2.20 0.17 L. + high gyp 115 4.93 4.49 2.17 1.64 0.28

20-40 cm Control 97 4.75 4.43 1.70 0.99 0.28 Low gypsum 115 5.03 4.34 1.94 1.22 0.35 High gypsum 136 4.46 4.21 2.02 1.09 0.73 Lime 93 4.89 4.28 1.98 1.06 0.52 L. + Low gyp 97 5.04 4.49 2.68 2.05 0.22 L. + high gyp 129 4.88 4.42 3.57 1.58 0.30

Bermuda Depth CEC Ca2+ Ex. acidity

0-10 cm EC (dS m-1) pHH2O pHKCl --------cmol+ kg-1-------- Control 107 5.37 4.80 2.24 1.60 0.11 Low gypsum 113 5.76 5.11 1.85 1.76 0.04 High gypsum 121 5.91 5.33 3.57 3.17 0.05 Lime 120 6.12 5.65 4.55 3.86 0.05 L. + Low gyp 139 6.33 6.06 4.00 3.60 0.04 L. + high gyp 123 6.13 5.93 4.01 3.20 0.06

10-20 cm Control 95 4.53 4.15 1.60 0.61 0.66 Low gypsum 104 4.85 4.41 1.45 1.19 0.23 High gypsum 118 4.96 4.25 2.24 1.44 0.48 Lime 98 4.73 4.44 1.64 0.95 0.33 L. + Low gyp 101 4.91 4.48 1.85 0.97 0.23 L. + high gyp 93 4.77 4.36 1.70 1.15 0.31

20-40 cm Control 98 5.05 4.27 1.58 0.80 0.44 Low gypsum 110 5.05 4.46 2.21 1.65 0.52 High gypsum 125 4.75 4.29 2.27 1.47 0.48 Lime 113 4.68 4.22 2.14 1.01 0.63 L. + Low gyp 112 4.74 4.23 1.96 0.97 0.52 L. + high gyp 125 4.82 4.39 2.18 1.36 0.48

159

Table 6.2. Exchangeable acidity values as a percent of CEC, by depth.

Depth (cm) Control Lime Low gyp High gyp

Lime + low gyp

Lime + high gyp

Zoysiagrass ---------------- Exchangeable acidity (% of CEC) ---------------- 0-10 2.2 1.2 2.6 1.7 1.7 1.4 10-20 22.8 28.2 20.6 35.0 6.1 12.9 20-40 16.5 26.2 18.0 36.2 8.2 13.2 Bermudagrass 0-10 4.9 1.1 2.1 1.4 1.0 1.5 10-20 41.3 20.1 15.9 21.4 12.4 18.2 20-40 27.8 29.5 23.5 21.1 26.5 22.0

160

Table 6.3. Root density values by treatment and depth. Letters signify significant differences (α=0.10) between treatments for each depth.

---------------Depth (cm)--------------- Zoysiagrass root density (mg kg-1) 0-15 15-30 30-45 45-70 Total Control 113 b 27 bc 15 a 1 c 157 c Low gypsum 228 ab 22 c 28 a 4 bc 282 abc High gypsum 309 ab 71 a 35 a 5 b 420 ab Lime 202 ab 28 bc 12 a 3 bc 245 bc Lime + low gypsum 400 a 59 ab 24 a 9 a 492 a Lime + high gypsum 156 b 48 abc 30 a 5 b 239 bc

---------------Depth (cm)--------------- Bermudagrass root density (mg kg-1) 0-15 15-30 30-45 45-70 Total Control 365 ab 87 a 122 a 65 ab 638 a Low gypsum 340 b 96 a 98 a 83 a 617 a High gypsum 321 b 125 a 105 a 58 ab 609 a Lime 384 ab 104 a 114 a 61 ab 663 a Lime + low gypsum 508 a 148 a 99 a 37 b 792 a Lime + high gypsum 380 ab 150 a 69 a 50 ab 649 a

161

Table 6.4. Volumetric water content (expressed as %) of treatment plots by date and turfgrass type at various depths. Mean comparisons using Fisher’s lsd (α=0.10) are between treatments at each depth. Columns 7 and 8 were sampled after dormancy.

Zoysiagrass 1 2 3 4 5 6 7 8 Date 7-28-04 9-13-04 9-23-04 10-7-04 11-5-04 11-16-04 12-7-04 3-1-05

0-10 cm Control 22.8 a 22.9 a 22.8 a 21.1 a 28.9 ab 27.4 a 30.9 ab 28.9 a Low gypsum 24.3 a 22.6 a 21.5 ab 19.2 a 27.8 b 26.3 a 29.8 b 28.7 a High gypsum 23.9 a 22.1 a 21.5 ab 19.7 a 28.7 ab 26.3 a 30.2 b 30.5 a Lime 22.8 a 20.9 a 20.0 b 18.7 a 28.0 b 26.5 a 30.0 b 29.9 a L. + low gyp 24.6 a 21.4 a 20.8 ab 19.7 a 28.3 b 26.7 a 29.6 b 28.5 a L. + high gyp 25.4 a 21.6 a 20.4 ab 19.3 a 30.4 a 28.3 a 32.5 a 29.5 a

10-20 cm Control 21.9 a 23.9 a 24.2 a 23.0 a 25.9 ab 25.6 ab 27.4 ab 25.5 b Low gypsum 24.0 a 23.8 a 24.3 a 22.8 a 26.8 ab 26.0 ab 27.5 ab 27.2 ab High gypsum 22.0 a 22.7 a 22.5 a 21.4 a 25.5 b 24.7 b 25.9 b 25.7 b Lime 24.8 a 24.5 a 24.2 a 23.1 a 28.0 a 27.2 b 28.5 a 28.1 a L. + low gyp 24.5 a 24.3 a 24.4 a 22.6 a 27.5 ab 26.9 ab 28.1 a 27.3 ab L. + high gyp 24.3 a 23.9 a 23.7 a 22.4 a 27.3 ab 26.6 ab 27.8 ab 27.3 ab

20-58 cm Control 30.5 a 26.5 a 27.1 a 26.4 a 27.2 ab 27.2 ab 28.0 a 31.3 a Low gypsum 30.1 ab 26.8 a 27.3 a 26.5 a 27.7 a 27.7 a 27.8 a 30.8 a High gypsum 28.7 abc 25.6 ab 25.9 ab 25.1 ab 26.7 ab 26.7 ab 26.5 a 30.6 a Lime 28.0 abc 25.0 ab 25.0 ab 24.0 ab 25.8 ab 25.8 ab 26.3 a 31.6 a L. + low gyp 26.0 bc 21.8 b 22.2 b 21.8 ab 23.0 b 23.0 b 26.5 a 31.0 a L. + high gyp 25.7 c 21.9 b 22.2 b 21.4 b 23.6 ab 23.6 ab 25.1 a 30.6 a Bermudagrass

Date 7-28-04 9-13-04 9-23-04 10-7-04 11-5-04 11-16-04 12-7-04 3-1-05 0-10 cm

Control 26.5 a 26.7 ab 25.6 a 23.1 ab 32.2 a 30.7 a 31.0 a 31.3 a Low gypsum 27.6 a 26.1 ab 24.9 a 23.5 ab 31.4 a 29.9 ab 30.5 a 31.5 a High gypsum 27.6 a 25.5 ab 24.9 a 23.0 ab 30.8 a 29.7 ab 30.2 a 31.2 a Lime 26.8 a 25.7 ab 25.5 a 23.5 ab 30.7 a 28.9 ab 29.9 a 30.8 a L. + low gyp 27.6 a 25.2 b 24.3 a 22.6 b 30.3 a 28.3 b 29.6 a 31.6 a L. + high gyp 27.8 a 27.5 a 26.4 a 25.0 a 31.8 a 30.2 ab 30.1 a 31.9 a

10-20 cm Control 25.5 a 26.1 a 26.1 a 24.1 a 28.7 a 28.4 a 28.3 a 28.1 a Low gypsum 26.1 a 26.2 a 26.0 a 24.7 a 28.8 a 28.4 a 28.2 a 28.5 a High gypsum 25.7 a 25.9 a 25.2 ab 23.5 ab 28.1 a 27.6 a 28.0 a 27.7 a Lime 24.2 a 24.8 ab 25.1 ab 23.5 ab 27.3 a 26.8 a 26.9 a 27.0 a L.+ low gyp 24.8 a 24.0 b 22.9 b 21.1 b 26.6 a 27.1 a 27.2 a 26.7 a L. + high gyp 25.5 a 25.9 a 26.0 ab 24.4 a 28.4 a 27.3 a 27.6 a 27.4 a

20-58 cm Control 24.5 a 25.3 a 25.7 a 24.6 a 26.6 a 26.3 a 27.0 a 30.3 a Low gypsum 25.7 a 26.3 a 26.7 a 25.9 a 27.8 a 27.2 a 27.3 a 30.9 a High gypsum 25.0 a 24.8 a 24.9 a 24.0 a 25.7 a 25.5 a 26.9 a 30.8 a Lime 26.4 a 27.3 a 27.6 a 26.8 a 28.4 a 27.8 a 28.2 a 30.1 a L. + low gyp 24.8 a 25.0 a 25.5 a 24.6 a 26.3 a 26.1 a 26.9 a 29.5 a L. + high gyp 26.3 a 26.4 a 27.0 a 26.1 a 27.7 a 27.6 a 27.5 a 30.8 a

162

Zoysiagrass soil Ca

Soil Ca (cmol+ kg-1)

0 1 2 3 4 5 6D

epth

(cm

)

0

10

20

30

controllimelow gypsumhigh gypsumlime + low gyplime + high gyp

Figure 6.1a. Zoysiagrass: Exchangeable Ca levels for treatments, at 0-10, 10-20, and 20-

40 cm depths. Error bars indicate standard error of mean value.

163

Bermudagrass soil Ca

Soil Ca (cmol+ kg-1)

0 1 2 3 4

Dep

th (c

m)

50

10

20

30

controllimelow gypsumhigh gypsumlime + low Gyplime + high gyp

Figure 6.1b. Bermudagrass: Exchangeable Ca levels for treatments, at 0-10, 10-20, and

20-40 cm depths. Error bars indicate standard error of mean value.

164

CHAPTER 7

SUMMARY AND CONCLUSIONS

165

The series of experiments conducted for these studies demonstrated that surface-

applied gypsum is effective in modifying and ameliorating subsoil acidity in several

Southeastern U.S. Ultisols, and that the mechanism appears to be predominantly related

to a sharp decrease in monomeric Al activity after the introduction of CaSO4 into the soil

solution. This work also indicated that not all turfgrass species respond to subsoil

amelioration by increasing root growth. In general, cool-season grasses responded to a

greater extent than warm-season grasses, and bermudagrass did not demonstrate

increased root growth in acid-ameliorated subsoil.

The hydroponic studies of tall fescue seedling roots showed that several tall

fescue varieties are sensitive to Al in an acid solution, probably owing to the dominant Al

species being monomeric in nature. The seedlings also responded in a significant way by

dramatically increasing root growth when CaSO4 was added to the solution, strongly

indicating gypsum would be an effective ameliorant of subsoil acidity for tall fescue

grass. The computer aided prediction of Al species in the hydroponic solutions using

VMINTEQ indicated that gypsum is effective in allowing increased root growth in the

presence of high Al by substantially reducing overall Al activity, decreasing the amount

of Al3+ as a percentage of total Al, increasing the quantity of the relatively non-toxic

AlSO4+ species in solution, and increasing the ionic strength of the solution. In general,

for a given level of Al in solution, increased solution CaSO4 resulted in increased root

growth. The levels of Al and CaSO4 used in this experiment were at or near levels

reported in the literature for Ultisol soil solutions with and without gypsum amelioration,

and indicate that this treatment would be effective for increased root growth from tall

fescue on acid Ultisol subsoils. The calcium-aluminum balance (CAB) equation did not

166

describe the relationship between relative root mass and CAB as well as expected, and

may be due to inherent differences in the calcium requirements of legumes and grasses.

The soil column study highlighted differences in root-growth response between

soils, as well as between treatments. The fine sandy loam had relatively poor percolation

compared to the sandy clay, which could have influenced root growth distribution

patterns in the columns. Overall, tall fescue responded to acid subsoil amelioration with

significant increases in root growth, confirming results obtained in the hydroponic study.

The incorporated lime treatment produced greater root growth than surface-applied

gypsum, probably due to greater reductions in exchangeable acidity and increased pH

levels making other essential nutrients more available.

The fact that there were no detectable differences in water content between

treatments during the drydown was surprising, since there were significant differences in

root density between treatments. Earlier drydown investigations using short columns (14

cm height) with bermudagrass and zoysiagrass also demonstrated no detectable

differences in plant water use, although these studies were conducted in a greenhouse

with warm temperatures and no attempt to keep the columns cool. This lack of detectable

differences in gravimetric water use may be an artifact of column studies. The high outer

soil-surface to soil volume ratio may promote rapid water loss through evaporation to

such an extent that differences due to transpiration are effectively masked. This potential

problem may be overcome with larger-diameter columns or deeper columns.

The field experiment demonstrated that amelioration of subsoil acidity induced by

applications of lime and gypsum do not necessarily result in significant increases in deep-

profile root densities. Zoysiagrass, which has been described as relatively shallow-rooted

167

and acid sensitive, did show increased root growth in soils treated with lime and gypsum.

Bermudagrass, however, had no root-growth response to lime and gypsum treatments.

Bermudagrass is generally described as drought tolerant and has shown adaptability to a

wide range of soils, including those with highly acid subsoils, and may have the genetic

capability to overcome low Ca and high soil exchangeable acidity. The difficulty in

concluding that bermudagrass does not need and probably would not respond to acid

subsoil amelioration with gypsum is that the field experiment conducted in this study had

a high degree of variation that may be overcome in the future with more treatment

replication, and the use of more than one variety of bermudagrass. Clearly, further

research on this grass species that represents a large percentage of the turfgrass, hay, and

pasture use in the Southeastern United States is needed.

168

APPENDICES

Appendix A. Griffin field study list of maintenance materials applied to turf plots.

Halosulfuron-methyl was applied September 3rd to control emerged yellow and

purple nutsedge at a rate of 0.05 kg active ingredient (a.i.) per ha-1. A complete fertilizer

with an analysis of 15-3-30 (urea, DAP, potassium chloride) was applied September 15th

at a rate of 160 kg ha-1. Oxadiazon was applied October 1st at a rate of 2.24 kg a.i. ha-1 as

a pre-emergent herbicide to control winter annual weeds. Dithiopyr was applied March

1st at a rate of 0.56 kg a.i. ha-1 as a pre-emergent herbicide to control summer annual

monocot weeds. 2,4-D herbicide was applied March 15th at a rate of 0.56 kg a.i. ha-1 to

control late winter and early summer broadleaf weeds. A 40-0-0 slow release nitrogen

fertilizer (urea, methylene urea) was applied May 3rd, June 14th, and August 2nd, at the

rate of 119 kg ha-1 application-1. Dithiopyr was applied October1st at a rate of 0.56 kg a.i.

ha-1 to suppress winter annual weeds.

169