Citation
Soil testing as a guide to phosphorus fertilization of slash pine (Pinus elliottii var. elliottii Engelm.)

Material Information

Title:
Soil testing as a guide to phosphorus fertilization of slash pine (Pinus elliottii var. elliottii Engelm.)
Creator:
Ballard, Russell
Publication Date:
Language:
English
Physical Description:
xviii, 273 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Fertilization ( jstor )
Fertilizers ( jstor )
Forest soils ( jstor )
Forests ( jstor )
pH ( jstor )
Phosphates ( jstor )
Phosphorus ( jstor )
Seedlings ( jstor )
Soil science ( jstor )
Soils ( jstor )
Dissertations, Academic -- Soil Science -- UF
Phosphatic fertilizers ( lcsh )
Slash pine ( lcsh )
Soil Science thesis Ph. D
Soils -- Phosphorus content ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1974.
Bibliography:
Includes bibliographical references (leaves 259-272).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Russell Ballard.

Record Information

Source Institution:
University of Florida
Rights Management:
This item is presumed in the public domain according to the terms of the Retrospective Dissertation Scanning (RDS) policy, which may be viewed at http://ufdc.ufl.edu/AA00007596/00001. The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator(ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
030019314 ( ALEPH )
37902058 ( OCLC )

Downloads

This item has the following downloads:


Full Text












SOIL TESTING AS A GUIDE TO PHOSPHORUS FERTILIZATION OF SLASH PINE (Pinus elliottii var. elliottii Engelm.)






By

Russell Ballard









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
















UNIVERSITY OF FLORIDA 1974














DEDICATION
















To

Phillipa













ACKNOWLEDGMENTS


I would like to thank Dr. W. L. Pritchett, the chairman of my

supervisory committee for his valuable guidance and counsel during the tenure of this study. The interest and assistance given by the other members of my committee, Dr. J.G.A. Fiskell, Dr. F.G. Martin, Dr. W.H. Smith and Dr. C.A. Hollis are very much appreciated.

I would like to acknowledge the encouragement given by Dr. D.F. Eno, chairman of the Soil Science Department and other members of the faculty. In particular I would like to thank Dr. L.W. Zelazny and Dr. J.G.A. Fiskell for their contribution to my academic enrichment and Dr. D.F. Rothwell for his administrative work on my behalf.

Thanks are also due to the CRIFF laboratory personnel, in particular to Mary McLeod for her help in the laboratory phase of the work, and to Dr. H.L. Breland and the staff of the Soils Analytical laboratory and to Dr. J. NeSmith and the staff of the Soil Testing laboratory for their expert analysis of numerous samples.

I have enjoyed the companionship of the graduate students of the Soil Science Department who provided necessary light relief and much intellectual stimulation. Specifically I would like to thank my fellow graduate students in the CRIFF program, Terry Sarigumba and Roy Voss, who willingly helped with some of the heavy work.

My most sincere thanks to my wife Pip, who has been actively involved in all phases of this work from initial collection of samples through to the final typing.

iii










Financial assistance from the CRIFF program, the National Research Advisory Council of New Zealand, and the Fulbright-Hays Foundation is gratefully acknowledged.

Finally, I would like to thank the Forest Research Institute, Rotorua, New Zealand for granting me leave of absence to pursue this study.











































iv

















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . iii

LIST OF TABLES . . . . . . . . . . . . . . . . . . . . .. .. . viii

LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . xiv

ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi

INTRODUCTION .. . . . . . . . . . . . . . . . . . . . . . . .

LITERATURE REVIEW . .................. .... 4

Forest Fertilization . . . . . . . . . . . . . . . . . . . 4
Historical . . . . . . . . . . . . . . . . . . . . . . 4
N and P Fertilization . . . . . . . . . . . . . . 5
P Sources Used in Forestry . . . . . . . . . . . . . . . . 9
Reactions in Soils . . . . . .. . . . . . . . . . . . 9
P Effectiveness in Forestry . . . . . ........... . . 15
/Diagnostic Methods . . ............ . . . . ... 20
Foliar Analysis . . . . .. . . . . . . . . . . . . . . 20
Soil Analysis . . . . . . . . . . . . . . . . . . . . 24
Sampling procedures . . . . . . . . . . . . . . . 25
Extraction methods . ...... ...... .. .. 28
Interpreting soil test results . . . . . . . . . . 33
Prediction of productivity . . . . . . . . . . 33 Prediction of fertilizer response . . . . . . . 34

MATERIALS AND METHODS . .................. . . 39

Introduction . . . . . . . . . . . . . . . . . . . . . . . 39
Field Trials . . . . . . . . . . . . . . . . . . . . . . . 40
Soil Sampling . . . . . . . . . . . . . . . . . . . . 40
Foliage Sampling . . . . . . . . . . . . . . . . . . . 42
Growth and Response Parameters . . . . . . . . . . . .. 42
Greenhouse Trial I . . . . . . . . . . . . . . . . . . . . 45
Establishment . . . . . . . . . . . . . . . . . . . . 45
Harvesting ..................... . 46
Growth and Response Parameters ............ 47
Greenhouse Trial 2 . . . . . . . . . . . . . . . . . . . . 48
P Compounds . . . . . . . . . . . . . . . . . . . . . 48
Establishment . ....... .......... . . . 49










TABLE OF CONTENTS (Continued)


Page

Harvesting . . . . . . . . . . . . . . . . . . . . . 51
Extraction of P Compounds ....... ...... . 51
Phosphorus-Retention Study . ..... ... . . . . . . . . 54
Soil and Foliage Samples ...... . . . . . . 54
Determination of P Retention . ........... 54
Sample Analysis . . .. .. . . . . . . . . . . . . . . . 57
Soil Characterization . . . . . . . . . . . . . . . 57
Soil P Analysis . . . . . . . . . . . . . . . . . 57
Soil Al and Fe Analysis .. . . . . . . . . . . 65
Plant Tissue Analysis . . . . . . . . . . . . . . . 65
Statistical Analysis . . . . . . . . . . . . . .. . . . 67

RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . 68

Preliminary Screening of Soil-Test Methods . . . .. . . 68
Relationships Between Soil-Test Values and Relative
Height . . . . . . . . . . . . . . .. . . . . . . . 72
Relationships Between Soil-Test Values and P Uptake
and Tissue P . . . . . . . . . . . . . . . . . ... . 79
Relationships Between Soil-Test Values and
Fertilizer Requirements . . . . . . . . . . . . . . . 81
Relationships Between Soil-Test Values and Height . . 83 General Discussion . . . . . . . . . . . . . . . . . 85
Phosphorus Compounds . .. ... .. . . . . . . . . . . . 88
Solubility of P Compounds in Chemical Extractants . . 88
Monocalcium phosphate (MCP) . . . . . . . . . . . 88
Dicalcium phosphate (DCP) . . . . . . . . . . . . 90
Fluorapatite (FA) . . . . . . . . . . . . . . . . 91
Colloidal aluminum phosphate (CAIP) . . . . . . . 92 Potassium taranakite (KTK) . . . . . . . . . . . 92
Wavellite (WA) . . . . . . . . . . . . . . . . . 93
Colloidal ferric phosphate (CFeP) . .. . . . . . 93 Strengite (STR) . ................ 94
Utilization of P Compounds by Slash Pine Seedlings 94
Relationships Between Seedling Utilization and
Solubility of P Compounds . ............. 99
General Discussion . . . . . . . . . . . . . . . . . 102
Field Calibration of Selected Soil-Test Methods . . . . . 105
Relationships Between Soil-Test Values and Relative
Height .................... . . . . . 105
Effect of soil sampling position and depth 115
Relationships Between Other Soil and Site
Parameters and Relative Height . ...... . . . . . 117
Relationships Between Soil-Test Values and PFertilizer Requirements . .. . .. . . . . . . . . . 126
Effect of soil sampling position and depth . . . 131


vi









TABLE OF CONTENTS (Continued)


Page

Relationships Between Other Soil and Site Parameters
and P-Fertilizer Requirements ............. 133
Relationships Between Soil-Test Values and Height . 136
Relationships Between Other Soil and Site Parameters
and Height .............. . . . . . . ..... 140
Relationships Between Foliar-P Concentrations and
Tree Parameters . . . . . . . . . . . . . . . . . . . 145
Relationships Between Foliar P and Soil-Test Values . 150 General Discussion . . . . . . . . . . . . . . . . . . 150
Phosphorus-Retention Study . ............... 155
Relationships Between P Retention and Soil Properties . 155
Physical and chemical properties . ........ 157 Extractable Al and Fe . . . .. . . . . . . . .. 161
Relative contribution of Al and Fe to P retention 165 Aluminum and Fe extracted by soil-test methods . 167
Relationships Between P Retention and Foliar Nutrient
Concentrations . . . . . . . . . . . . . . . . . . . . 169
Calibration of Extractable Al Against Field-P Retention 172

SUMMARY AND CONCLUSIONS . ................. . . 179

APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . . . 259

BIOGRAPHICAL SKETCH .................... .. . 273























vii















LIST OF TABLES


Table Page


1 FOLIAR P CONCENTRATIONS PRIOR TO FERTILIZATION IN RELATION
TO RESPONSE OF SOUTHERN PINES TO P FERTILIZER . ...... 21

2 SOIL-P VALUES (SURFACE HORIZON) OF UNFERTILIZED SOILS IN
RELATION TO RESPONSE OF SOUTHERN PINES TO P FERTILIZER. . . 37

3 PROPERTIES OF PHOSPHORUS SOURCES USED IN GREENHOUSE TRIAL 2 . 50

4 PHYSICAL AND CHEMICAL PROPERTIES OF SOILS USED IN
GREENHOUSE TRIAL 2. . .................. .. 52

5 PHOSPHORUS EXTRACTION METHODS . ............... 59

6 CLASSIFICATION AND SELECTED PROPERTIES OF 10 SOILS USED
IN GREENHOUSE STUDY I . . . . . . . . . . . . . . . . . . . 69

7 HEIGHT, RELATIVE HEIGHT, P-FERTILIZER REQUIREMENTS, AND
P CONCENTRATION AND UPTAKE OF SLASH PINE SEEDLINGS AFTER
I AND 2 YEARS' GROWTH ON 10 SOILS IN THE GREENHOUSE . . . . 70

8 HEIGHT, RELATIVE HEIGHT AND P-FERTILIZER REQUIREMENTS OF
SLASH PINE AFTER 1, 3, AND 5 YEARS' GROWTH, AND FOLIAR P
CONCENTRATION AFTER 4 YEARS' GROWTH IN THE FIELD ON
10 SOILS. . . . . . . . . . . . . . . . . . . . ...... 71

9 COMPARISON OF THE GOODNESS OF FIT OF FOUR STATISTICAL
MODELS, AS INDICATED BY THE SQUARE OF THE MULTIPLE
CORRELATION COEFFICIENT (R2), RELATING SELECTED TREE PARAMETERS (DEPENDENT VARIABLE) AND SOIL-TEST VALUES
(INDEPENDENT VARIABLE). . ................. . 73

10 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND
RELATIVE HEIGHT GROWTH OF SLASH PINE IN FIELD AND
GREENHOUSE EXPERIMENTS ON 10 SOILS. . ........ . . . . 77

11 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND
TISSUE P PARAMETERS OF GREENHOUSE AND FIELD SLASH
PINE GROWN ON 10 SOILS. . ... ............... . 80

12 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND PFERTILIZER REQUIREMENTS OF SLASH PINE IN FIELD AND
GREENHOUSE EXPERIMENTS ON 10 SOILS. . ........... . 82


viii










LIST OF TABLES (continued)


Table Page


13 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND HEIGHT
GROWTH IN THE ABSENCE OF P FERTILIZER OF SLASH PINE IN
FIELD AND GREENHOUSE EXPERIMENTS ON 10 SOILS. . ....... 84

14 SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS IN THE
PRESENCE AND ABSENCE OF TWO SOILS . ............ 89

15 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS
BETWEEN SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS
AND THE UPTAKE OF P FROM THESE COMPOUNDS BY SLASH PINE
SEEDLINGS GROWN ON TWO SOILS IN THE GREENHOUSE. . ..... 100

16 .SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS FOLLOWING
2 MONTHS' INCUBATION IN TWO SOILS . ........... . 101

17 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL
DEPTHS AND RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3,
AND 5 YEARS' GROWTH ON 72 FIELD SITES ........... 107

18 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN AMOUNTS OF
P EXTRACTED FROM THE SURFACE 20 cm OF SOIL (BEDDED
AREA) BY FIVE SOIL-TEST METHODS . ............. 108

19 SEPARATION OF 72 FIELD SITES INTO RESPONSE QUADRANTS
USING THE TECHNIQUE OF CATE AND NELSON (1965) AND A
CRITICAL HCI-H2SO4-EXTRACTABLE P VALUE OF 5 ppm . ..... 108

20 REGRESSION EQUATIONS RELATING RELATIVE HEIGHT OF SLASH PINE AT AGE 1, 3, AND 5 YEARS ( YI, Y3, AND Y5) TO
THE LOG TRANSFORMED P EXTRACTED FROM THE SURFACE 20 cm OF SOIL BY THE H20 (XI) AND HCI-H2SO4 (X2) SOILTEST METHODS. . . . . . . . . . . . . . . . . . . . . . . . 116

21 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN AMOUNTS OF P EXTRACTED BY THE HCI-H2S04 METHOD FROM TWO SOIL
POSITIONS AND THREE SOIL DEPTHS . . . . . . . . . . . . . . 116

22 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE
-! SURFACE 20 cm OF SOIL AND THAT EXTRACTED FROM WITHIN
THE EFFECTIVE SOIL DEPTH (VOLUME) AT PREDICTING RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3, AND 5
YEARS' GROWTH ON 72 FIELD SITES ......... . . . . . 119

23 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES AND RELATIVE HEIGHT OF SLASH PINE 1, 3, AND 5 YEARS
AFTER P FERTILIZATION ON 72 SITES . ............ 120


ix








LIST OF TABLES (continued)


Table Page


24 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN SELECTED SOIL AND SITE PROPERTIES OF 72 FIELD SITES . ....... ..... . 121

25 REGRESSION COEFFICIENTS FOR MULTIPLE REGRESSION EQUATIONS OF RELATIVE HEIGHT ON DEPTH TO LH, HCI-H SO -EXTRACTABLE
P(O-20 cm) AND THE SQUARED TERMS OF TRESE TWO PARAMETERS. . 123

26 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN 'BEST FIT' MULTIPLE REGRESSION EQUATIONS OF RELATIVE HEIGHT ON SOIL AND
SITE PARAMETERS AND THE SQUARED TERMS OF THESE PARAMETERS . 124

27 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL DEPTHS AND P FERTILIZER REQUIRED TO ACHIEVE 90, 95, AND 100% OF MAXIMUM HEIGHT GROWTH AFTER 1, 3, AND 5 YEARS' GROWTH ON
72 FIELD SITES. .......... . . . .. . .... . 127

28 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE SURFACE 20 cm OF SOIL WITH THAT EXTRACTED FROM WITHIN THE EFFECTIVE SOIL DEPTH (VOLUME) AT PREDICTING THE FERTILIZER P REQUIRED
TO ACHIEVE 90, 95, AND 100% OF MAXIMUM HEIGHT GROWTH AFTER I,
3, AND 5 YEARS' GROWTH ON 72 FIELD SITES. . ........ . 132

29 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES AND P FERTILIZER REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT AFTER
1, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES . ....... 134

30 RELATIONSHIPS BETWEEN HCI-H2SO4-EXTRACTABLE P(O-20 cm) AND P FERTILIZER REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT ON
GROUPS OF SOILS CLASSED ACCORDING TO THEIR AMOUNT OF
NH4OAc (pH 4.8)-EXTRACTABLE Al . ............. 135

31 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN 'BEST FIT' MULTIPLE REGRESSION EQUATIONS OF P FERTILIZER REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT ON SOIL AND SITE PARAMETERS
AND THE SQUARED TERMS OF THESE PARAMETERS . ........ 137

32 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL DEPTHS AND HEIGHT OF SLASH PINE IN THE ABSENCE OF P FERTILIZER
AFTER 1, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES . .... 138

33 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE SURFACE 20 cm OF SOIL AND THAT EXTRACTED FROM WITHIN THE
EFFECTIVE SOIL DEPTH (VOLUME) AT PREDICTING HEIGHT OF SLASH
PINE AFTER 1, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES. . . 139

34 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES AND HEIGHT OF SLASH PINE AFTER 1, 3, AND 5 YEARS' GROWTH ON 72
FIELD SITES . . . . . . . . . . . . . . . . . . . . . . . . 142


x










LIST OF TABLES (continued)


Table Page


35 REGRESSION COEFFICIENTS FOR MULTIPLE REGRESSION EQUATIONS
OF RELATIVE HEIGHT ON DEPTH TO LH, H20-EXTRACTABLE
P (0-20 cm) AND THE SQUARED TERMS FOR THESE TWO
PARAMETERS . . . . . . . . . . . . . . . . . . . . . . . . 144

36 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN 'BEST
FIT' MULTIPLE REGRESSION EQUATIONS OF HEIGHT AT AGE 1, 3,
AND 5 YEARS ON SOIL AND SITE PARAMETERS AND THE SQUARED
TERMS OF THESE PARAMETERS. . .............. . 146

37 RELATIONSHIPS BETWEEN HEIGHT, RELATIVE HEIGHT, AND P
FERTILIZER REQUIREMENTS (95%) AT AGE 1, 3, AND 5 YEARS, AND P CONCENTRATIONS IN THE FOLIAGE OF 4-YEAR-OLD SLASH
PINE. ......... . .. .......... . . . . . . . 148

38 RELATIONSHIPS BETWEEN EXTRACTABLE-SOIL P AT TWO POSITIONS,
THREE DEPTHS, AND WITHIN THE EFFECTIVE SOIL DEPTH
(VOLUME), AND P CONCENTRATIONS IN THE FOLIAGE OF 4-YEAROLD SLASH PINE . . . . . . . . . . . . . . . . . . . . . .. 148

39 SOIL PROPERTIES AND THEIR CORRELATION WITH P RETENTION. . . 159

40 EXTRACTABLE Al AND Fe VALUES AND THEIR CORRELATION WITH
P RETENTION. . . . . . . . . . . . . . . . . . . . .. . 162

41 CORRELATIONS BETWEEN DIFFERENT FORMS OF Al OR Fe, AND P
RETENTION. .................. ..... . 164

42 REGRESSION EQUATIONS RELATING P RETENTION (%) TO
DIFFERENT FORMS OF SOIL Al AND Fe . ........... 166

43 ALUMINUM AND Fe EXTRACTED BY FOUR SOIL-TEST METHODS AND
THEIR CORRELATION WITH P RETENTION . ........... 168

44 FOLIAR NUTRIENT CONCENTRATIONS AND THEIR CORRELATION WITH
SOIL-P RETENTION . . . . . . . . . . . . . . . . . . . . . 170

45 REGRESSION EQUATIONS OF Al EXTRACTED BY HCI-H2SO4 (YI),
BRAY I (Y2), AND BRAY 2 (Y3) ON Al EXTRACTED 8Y NH4OAc (X) 175

46 HEIGHT, DRY WEIGHT, AND P CONCENTRATION OF SLASH PINE
SEEDLINGS AFTER I AND 2 YEARS' GROWTH ON 10 SOILS IN THE
GREENHOUSE RECEIVING FOUR P TREATMENTS . ......... 187

47 AVERAGE HEIGHTS OF SLASH PINE AS AFFECTED BY P TREATMENTS
AFTER 1, 3, AND 5 YEARS' GROWTH IN THE FIELD ON 10 SOILS 194

xi








LIST OF TABLES (continued)


Table Page


48 DRY WEIGHT, P CONCENTRATION, AND P UPTAKE IN TOPS AND
ROOTS OF SLASH PINE SEEDLINGS AFTER I YEAR OF GROWTH
ON NINE SOILS IN THE GREENHOUSE. . ...... . . . . . . 196

49 AMOUNTS OF P EXTRACTED FROM 10 SOILS BY SOIL-TEST METHODS 197

50 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS BETWEEN SOIL-TEST VALUES AND SEEDLING PARAMETERS FROM
GREENHOUSE STUDY I . . . . . . . . . . . . . . . . . . . . 202

51 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS BETWEEN SOIL-TEST VALUES AND TREE PARAMETERS FROM 10
SELECTED FIELD TRIALS. . . . .......... ...... .. . 207

52 DRY WEIGHT AND P UPTAKE OF ENTIRE SLASH PINE SEEDLINGS, AS AFFECTED BY P SOURCE, AFTER THE 8 MONTHS OF GROWTH ON
TWO SOILS IN THE GREENHOUSE. . .............. . 212

53 ANALYSIS OF VARIANCE OF SEEDLING DRY WEIGHTS AND P UPTAKES OF GREENHOUSE TRIAL 2. . ...... . ............ . . 213

54 .SOIL CLASSIFICATION, SITE PROPERTIES, AND SELECTED CHEMICAL PROPERTIES OF UNFERTILIZED SOILS (0-20 cm), FOR THE 24
FIELD SITES . .............. . . . . . . . 214

55 AMOUNTS OF P EXTRACTED FROM SOILS, COLLECTED FROM CONTROL PLOTS OF 24 FIELD TRIALS, BY FIVE SOIL-TEST METHODS, AND AMOUNTS OF Ca, Mg, K, and Al EXTRACTED BY NH40Ac(pH 4.8)
AND SOIL pH. . . . . . . . . . . . . . . . . . . . . . . . 218

56 PHYSICAL PROPERTIES OF SOILS COLLECTED FROM 24 FIELD TRIALS 231

57 HEIGHT AND RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3, AND 5 YEARS' GROWTH ON 72 SITES IN THE FIELD . ..... . . . . 241

58 PHOSPHATE FERTILIZER (CSP) REQUIRED TO ACHIEVE 90, 95, AND 100% OF MAXIMUM HEIGHT OF SLASH PINE 1, 3, AND 5 YEARS
AFTER FERTILIZATION OF 72 FIELD SITES. ...... . . . . 244

59 CONCENTRATIONS OF N, P, K, Ca, Mg, Al, AND Fe IN FOLIAGE COLLECTED FROM 4-YEAR-OLD SLASH PINE GROWING IN THE
CONTROL PLOTS OF 24 FIELD TRIALS . ....... ..... . 248

60 CLASSIFICATION, SELECTED PHYSICAL AND CHEMICAL PROPERTIES, AND P-RETENTION CHARACTERISTICS (LANGMUIR AND SATURATION
MAXIMA) OF 42 LOWER COASTAL PLAIN FOREST SOILS . ..... 251

xii









LIST OF TABLES (continued)


Table Page


61 AMOUNTS OF Al AND Fe EXTRACTED FROM 42 COASTAL PLAIN
FOREST SOILS BY SIX CHEMICAL EXTRACTANTS. .......... 253

62 AMOUNTS OF Al AND Fe EXTRACTED FROM 42 COASTAL PLAIN
FOREST SOILS BY FOUR SOIL P-TEST METHODS. . ......... 255

63 AMOUNTS OF TOTAL P IN THE SURFACE 20 cm OF SOIL COLLECTED
FROM THE CONTROL, Pi(56 Kg P/ha), AND P2(224 Kg P/ha)
PLOTS OF 10 SELECTED FIELD TRIALS 4 YEARS AFTER PFERTILIZER APPLICATION. . .................. 257







































xiii














LIST OF FIGURES


Figure Page


1 RELATIONSHIP BETWEEN R2 VALUES FOR REGRESSIONS OF RELATIVE
HEIGHT AT AGE 1, 3, AND 5 YEARS ON SOIL-TEST VALUES, AND
MEAN AMOUNTS OF P EXTRACTED BY SOIL-TEST METHODS . .... 75

2 PHOSPHORUS UPTAKE BY SLASH PINE SEEDLINGS AFTER 8 MONTHS'
GROWTH ON TWO SOILS TREATED WITH EIGHT P COMPOUNDS . . .. 96

3 .DRY MATTER OF SLASH PINE SEEDLINGS AFTER 8 MONTHS' GROWTH
ON TWO SOILS TREATED WITH-EIGHT P COMPOUNDS. . ...... . 97

4 RELATIONSHIP BETWEEN HCI-H2SO4-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 1 YEAR AFTER P FERTILIZATION. . ............ Ill

5 RELATIONSHIP BETWEEN HCI-H2SO4-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 3 YEARS AFTER FERTILIZATION . ........... . 112

6 RELATIONSHIP BETWEEN HCI-H2S04-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 5 YEARS AFTER P FERTILIZATION . ........... 113

7 RELATIONSHIPS BETWEEN HCI-H2SO4-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT OF SLASH
PINE i, 3, AND 5 YEARS (Yl, Y3, AND Y5) AFTER P
FERTILIZATION. . ..... . ... ......... . . . 114

8 RELATIONSHIP BETWEEN DEPTH TO LIMITING HORIZON (X) AND
RELATIVE HEIGHT (Y) OF SLASH PINE 5 YEARS AFTER
FERTILIZATION. . . . . . . . . . . . . . . . . . . . .. . 122

9 RELATIONSHIPS BETWEEN HC1-H2SO-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND AMOUNT OF P FERTILIZER (CSP) REQUIRED TO ACHIEVE 90% OF MAXIMUM HEIGHT GROWTH OVER
PERIODS OF 1, 3, AND 5 YEARS (YI, Y3, AND Y5)
FOLLOWING FERTILIZATION. . ................ . 128

10 RELATIONSHIPS BETWEEN HCI-H2SO4-EXTRACTABLE P(X) IN THE SURFACE 20 cm OF SOIL AND AMOUNT OF P FERTILIZER (CSP) REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT GROWTH OVER
PERIODS OF 1, 3, AND 5 YEARS (YI, Y3, AND Y5)
FOLLOWING FERTILIZATION. . ................ . 129

xiv









LIST OF FIGURES (continued)


Figure Page


11 RELATIONSHIP BETWEEN DEPTH TO LIMITING HORIZON (X) AND
THE HEIGHT OF SLASH PINE AFTER 5 YEARS' GROWTH . .... 143

12 RELATIONSHIPS BETWEEN P CONCENTRATIONS IN FOLIAGE (X)
OF 4-YEAR-OLD SLASH PINE AND RELATIVE HEIGHT OF SLASH
PINE 1, 3, AND 5 YEARS (Y1, Y3, AND Y5) AFTER P
FERTILIZATION. . . . . . . . . . . . . . . . . .. . . . 149

13 RELATIONSHIP BETWEEN HCI-H2S04-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND P CONCENTRATION IN FOLIAGE
(Y) OF 4-YEAR-OLD SLASH PINE .............. 151

14 PHOSPHORUS-ADSORPTION ISOTHERMS OF FOUR LOWER COASTAL
PLAIN SOIL TYPES REPRESENTATIVE OF FOUR SOIL ORDERS. . 156

15 RELATIONSHIP BETWEEN % P RETENTION FROM THREE P
SOLUTIONS (100, 300 AND 2,500 0g P/g SOIL) AND
NH40Ac (pH 4,8)-EXTRACTABLE Al . ............ 158

16 RELATIONSHIP BETWEEN SURFACE (0-20 cm) RETENTION OF P APPLIED AS CSP 4 YEARS PREVIOUSLY AND NH40Ac (pH 4.8)EXTRACTABLE Al AT 10 SITES GROWING SLASH PiNE. . . . . . 174

17 PHOSPHORUS RETENTION AS DETERMINED FROM ADDITION OF 2,500 pg P/g SOIL IN A AND Bh HORIZONS OF SIX
SPODOSOLS. . . . . . . . . . . . . . . . . . . . .... 177



















xv









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


SOIL TESTING AS A GUIDE TO PHOSPHORUS FERTILIZATION
OF SLASH PINE (Pinus elliottii var. elliottii Engelm.) By

Russell Ballard

August, 1974

Chairman: Dr. W.L. Pritchett
Major Department: Soil Science

This investigation was designed to develop and calibrate a soil test or tests for predicting the amount and type of phosphatic fertilizer, if any, needed at time of planting slash pine (Pinus elliottii var. elliottii Engelm.) on forest soils in the southeastern lower Coastal Plain.

In a preliminary screening of over 100 different procedures for extracting or characterizing soil P, it was found that soil-test methods which extracted relatively small amounts of P from the soil (H20, NH40Ac pH 4.8, neutral salts, very dilute acid solutions at narrow soil:solution ratios, anion exchange resin) provided the best index of first-year growth and response of slash pine to P fertilizer both in greenhouse and field fertilizer trials on the same 10 soils. However, soil-test methods capable of extracting larger quantities of soil P, irrespective of the extracting reagent used (strong acid, alkaline, or complexing agents), were more effective predictors of growth and response achieved over longer growth periods. Procedures which extracted total, or specific fractions of soil P (organic-P, Ca-P, Fe-P, AI-P) were less effective


xvi









predictors of P fertilizer needs than more conventional test methods for available P.

An examination of the solubility of P compounds (mono and dicalcium phosphates, fluorapatite, colloidal Al and Fe phosphates, K taranakite, wavellite, and strengite) in extractants of several common soil-P test methods and their utilization by slash pine seedlings, showed that soiltest methods which provided the best index of early growth and response in preliminary screenings extracted little more P from these compounds than did H20 alone. Methods which were more successful predictors over longer growth periods extracted appreciable quantities of P from the Ca phosphates, colloidal phosphates, and K taranakite, which were effective P sources for seedlings; but, they did not extract much P from wavellite and strengite, which were ineffective P sources for seedlings. Methods utilizing strong acids did, however, overestimate the availability of fluorapatite to seedlings on a soil of pH > 5.

Five soil-test methods (H20, NH40Ac pH 4.8, 0.05N HCI + 0.025N H2S04, 0.03N NH4F + 0.025N HCI, 0.03N NH4F + 0.IN HCI) selected on the basis of results from the preliminary screening, were calibrated against slash pine growth and response information obtained from 72 field fertilizer trials I, 3, and 5 years after establishment. The method involving use of 0.05N HCI + 0.025N H2S04 provided the best prediction of height response and P fertilizer requirements over the 3- and 5-year growth periods. A surface soil (0-20 cm) value of 5 ppm by this method pro-vided an effective delineation of P responsive sites. Soils testing between 5.0 and 2.5 ppm P required ca. 20-40 kg P/ha and those testing below 2.5 ppm P required ca. 40-80 kg P/ha to provide an adequate P supply to slash pine over the above growth periods.


xvii









Laboratory-determined P-retention capacity of 42 forest soils was significantly related to Al extracted by either conventional procedures (KCI leaching, pyrophosphate, oxalate, and CDB extraction) or four soilP test methods. Soil Al extracted by these four soil-P test methods was calibrated against P leaching losses over a 4-year period in the field following application of concentrated superphosphate (CSP). Soils containing less than 40, 120, 300, or 400 ppm Al extractable by NH40Ac pH 4.8, 0.05N HCI + 0.025N H2S04, 0.03N NH4F + 0.025N HCI, or 0.03N NH4F + 0.01N HCI, respectively, exhibited excess leaching losses of P from soluble CSP fertilizer; use of less soluble P fertilizers, such as rock phosphate, was suggested on such soils.


































xviii














INTRODUCTION


Fertilization is one of many cultural practices available for increasing the productivity of forest lands. Increasing demand for wood products, associated with a decreasing land base for commercial timber production, has led to an upsurge of interest in intensive forest management techniques in the last decade.

In the southeastern Coastal Plain, intensive forest cultural practices such as clear-cut harvesting, site preparation, use of improved genetic stock, and fertilization are now widely used. Climatic conditions which favor long growing seasons and high biological activity, the adoption of intensive management techniques which increase growth rates, and the inherently infertile acid sandy soils, have been the major factors contributing to the successful results obtained from the use of fertilizer in southeastern Coastal Plain forests. Results from more than 225 field experiments established throughout the Coastal Plain by the Cooperative Research in Forest Fertilization (CRIFF) program have shown that about two-thirds of the tested sites were responsive to nitrogen and/or phosphate fertilizers (CRIFF annual report, 1973)1 Trials in young plantations indicated that about two-thirds of the sites were responsive to phosphate fertilizers and three-fourths responsive to nitrogen and/or phosphate (Pritchett and Smith, 1972).

Approximately 770,000 acres of pines are planted annually in the

Southeast, of which it has been projected 500,000 acres would benefit from fertilizer additions (CRIFF annual report, 1973)1 Considerably less than





2


this projected acreage is actually fertilized. This, in part, can be attributed to the usual caution of forest managers in accepting new practices, but another major reason is the lack of a precise diagnostic technique for delineating areas where economic responses can be expected from fertilizer additions. This inadequacy was emphasized in a recent survey of 50 forestry organizations in the South soliciting opinions on forest fertilization research priorities. The concensus of most respondents was that one of the most pressing needs was "for a predictive mechanism telling where, when, how and with what to fertilize " (Moehring, 1972).

The two major diagnostic techniques used in forestry are foliar and soil analyses. Foliar analysis has generally proved to be more reliable than soil analysis for predicting increased growth of southern pines from P fertilization (Pritchett, 1968; Wells and Crutchfield, 1969). However foliar analysis cannot be used in the absence of growing stock on the site and, consequently, an effective soil test has the advantage over foliar analysis in that it can be used to predict the need for fertilizer at the time of planting.

Accurate diagnosis is essential for the success of any forest fertilization program. The diagnosis should provide information on the degree of response to be expected, the amount of fertilizer required to obtain this response, and the most effective nutrient source to use. The question of most effective P source is of particular importance to P fertilization in the Coastal Plain. Humphreys and Pritchett (1971) found that the P-retention capacity of soils in the Coastal Plain region had a strong influence on the long-term effectiveness of P sources of different solubility.

The general purpose of this study was to develop a soil test or

tests which would provide the necessary information required for a forest




3




manager to make the correct decisions regarding P fertilization practices at time of planting. Specific objectives were as follows:

1. Evaluation of the effectiveness of a wide range of soil-P testing

techniques for predicting both growth and response of P fertilizer

applied to slash pine at time of planting.

2. Examination of the effects that soil and environmental parameters

have on the predictive accuracy of soil-P tests.

3. Calibration of'the best soil-P test against the degree of response

obtained from fertilizer additions and the amount of fertilizer

required to produce this response.

4. Development of a predictive procedure for delineating sites where

phosphate sources of different solubility should be used.

5. Determination of the capacity of slash pine to utilize a range of

soil-P compounds, and relate this ability to the solubility of these

compounds in the various extractants used in soil-P tests.



NOTE


CRIFF annual report, 1973. (unpublished). Soil Science Department, University of Florida, Gainesville, Florida.















LITERATURE REVIEW


Forest Fertilization

Historical

The value of fertilizers for increasing the productivity of timber lands was first recognized in Europe. Fertilizer trials were established on peat lands in Sweden as early as 1898 (Hagner, 1967). During the period 1900-1925, forest fertilizer trials were established in many European countries including Finland (Salonen, 1967), Belgium, Germany, and Denmark (Tamm, 1968), and Britain (Leyton, 1958). Experimentation expanded in Europe between 1925 and 1960, establishing a firm scientific foundation to tree nutrition which led to the first operational applications of fertilizer in Germany (Tamm, 1968) and Britain (Leyton, 1958). In Australia, the importance of phosphatic fertilizers in achieving healthy growth of introduced pines planted on infertile sandy soils was established as early as 1930 (Kessell and Stoate, 1938).

Research on forest fertilization is relatively new in North America. Some of the earliest trials on this.continent were N experiments established on hardwood stands in the northeastern USA by Mitchell and Chandler (1939). Westveld established a phosphate fertilizer trial in the southeastern Coastal Plain in 1945 (Pritchett and Swinford, 1961), while research on forest fertilization was started around 1950 in the Pacific Northwest (Gessell, 1968).


4




5



The apparent lack of interest shown in forest fertilization by most North American foresters prior to 1960 has been attributed to their preoccupation with extensive rather than intensive management techniques (Bengtson, 1972). In addition, the results of most soil-site research, particularly that by Coile (1952), indicated that soil physical characteristics provided better indices of site productivity than did soil chemical properties.


N and P Fertilization

Several general reviews on forest fertilization (White and Leaf, 1957; -Stoeckler and Arneman, 1960; Swan, 1965; Mustanoja and Leaf, 1965) and specific reviews of progress in countries such as Sweden (Hagner, 1967; Holmen, 1967), Finland (Salonen, 1967; Paarlahti, 1967), Norway, (Jerven, 1967), Japan (Kawana, 1969), Britain (Leyton, 1958; Binns, 1969), Australia (Raupach, 1967; Gentle and Humphreys, 1968), Canada (Swan, 1969; Krause, 1973), and the USA (Gessell, 1968; White, 1968; Bengtson, 1968 and 1972; Safford, 1973), have illustrated the tremendous advances in forest fertilization research and the application of the results during the last decade. Most recorded responses have been to N and P fertilizers although isolated responses to K fertilizers have been reported in Japan (Kawana, 1969), Australia (Hall and Purnell, 1961), Finland (Salonen, 1967), midwestern Europe (Hagner, 1971), and the northeastern USA (Stone and Leaf,

1967).

The most spectacular response of coniferous species to N applications have been recorded in Scandinavia (Hagner, 1967) and the Douglasfir region of the northwestern USA and southwestern Canada (Strand and Miller, 1969). In the Scandinavian countries, operational fertilization began about 1960 (Hagner, 1967) with urea as the principal N source. In




6



recent years there has been a switch to ammonium nitrate because it appeared to give better results than N applied as urea (Hagner, 1971). Normal application rates, designed to give a response over a 5-year period, were about 120 to 180 kg N/ha. In 1970, approximately 143,000 ha of established forest stands were fertilized in Scandinavia. In Sweden alone, over 500,000 ha had received N fertilizer by the end of 1970 (Hagner, 1971). Operational fertilization with urea-N began in the Pacific Northwest in 1965 (Anderson, 1969). About 60,000 ha had been fertilized up to the end of 1970, with some 40-50,000 ha fertilized during 1970 (Hagner, 1971). The average rate applied in the Douglas-fir region was about 200 kg N/ha, which gives a response over a period of 5 to 7 years (Strand and Miller, 1969).

Pritchett and Smith (1970) reported several studies in which slash pine (Pinus elliottii var. elliottii Engelm.) responded to N applications in the USA southeastern Coastal Plain. In their studies, most responses were obtained in established stands, where drainage conditions apparently had little effect on growth response. However, in one study (Broerman, 1967), stand density had a strong modifying influence on tree response to N. In operational fertilization, urea has been the preferred N source and has been applied at rates of 70 to 100 kg N/ha (Pritchett and Hanna, 1969). In 28 (NxP factorial) fertilizer trials established at time of planting slash pine throughout the Coastal Plain, N applied alone had little effect on the growth of seedlings in any tests except those in the somewhat poorly-drained flatwood sites (Pritchett and Smith, 1972). On P-deficient sites, the application of N alone tended to suppress growth, but when applied in conjunction with P, N significantly increased growth over that obtained from P alone in 72% of all tests.





7



Spectacular growth responses following addition of phosphate fertilizers to conifers have been recorded in many countries. As a consequence, routine application of P fertilizer, particularly at time of planting, is now practiced on P-deficient sites in New Zealand (Conway, 1962; Armitage, 1969), Australia (Gentle and Humphreys, 1968), Japan (Kawana, 1969), Britain (Leyton, 1958), Finland (Salonen, 1967), Norway (Jerven, 1967), France and Germany (Baule and Fricker, 1970) and the southeastern USA (Pritchett and Hanna, 1969). In 1970, 70-80,000 ha were fertilized at time of planting in Japan, about 30,000 ha in midwest Europe and 10,000 ha in both the US Southeast and Oceania (Australia and New Zealand) (Hagner, 1971). Superphosphate was the principal fertilizer used.

Applications of P fertilizer to young stands at rates of 78 kg P/ha have maintained responses for a minimum of 15 years (Pritchett and Swinford, 1961; Gentle, Humphreys,and Lambert, 1965). However in other fertilizer trials in both the US Southeast and Australia, growth response and foliar concentrations have shown that applications of superphosphate up to 78 kg P/ha began to lose their effectiveness 7 to 15 years after application (Gentle and Humphreys, 1968; Humphreys and Pritchett, 1971; Pritchett and Smith, 1974). The loss in efficiency of soluble forms of P fertilizer was attributed to either leaching from the rooting zone in soils of low P-retention capacity, or binding in unavailable forms in soils of very high P-retention capacity. The practical and economic problems of providing an adequate P supply throughout a rotation has led to a policy in New Zealand and Australia of applying only sufficient fertilizer at time of planting to provide an adequate P supply for pinus species up





8



to canopy closure at age 5 to 10 years (Armitage, 1969; Craig, 1972). Once canopy closure is reached the stands are refertilized by aerial application.

The characteristics of forest soils in the southeastern USA lower Coastal Plain have been described by Pritchett and Smith (1970). These soils are predominantly infertile, acid sands with drainage characteristics ranging from excessively drained to poorly drained. Responses of slash pine to P fertilizers in the lower Coastal Plain have been related to the drainage characteristics of the sites. In eight fertilizer trials, Pritchett and Llewellyn (1966) found that 3 to 5 years after application; significant responses to P fertilizer had occurred in 5 out of the 8 trials, all of which were on sites classified as poorly or somewhat poorly drained. On two well-drained sites no response to P was recorded, although they did respond to N application. Responses from the same trials, recorded 7 to 11 years after fertilizer application (Humphreys and Pritchett, 1971), showed that the greatest response had occurred on the poorly drained sites, followed by the somewhat poorly drained sites, with no P response on the well-drained sites. In a more recent series of trials established at planting time on 28 sites throughout the Coastal Plain, Pritchett and Smith (1972) reported that 3 years after fertilizer application, 6 of 7 experiments located on poorly drained sites, 8 of 10 on somewhat poorly drained sites, 4 of 8 on moderately well drained sites, and none of the 3 on excessively drained sites responded significantly to P applications. In the same series it was found that N alone significantly increased growth only in 8 of the experiments on the somewhat poorly drained sites, although when N was applied with P, it increased the response over that obtained with P alone in 72% of the 28 experiments.





9



P Sources Used in Forestry

A number of P sources of different solubility have been used successfully in forestry. Ordinary (OSP) and concentrated superphosphate (CSP) are the most extensively used sources (Pritchett and Hanna, 1969; Gentle and Humphreys, 1968), but ground rock phosphate (RP) has proved to be an effective source for many forest soils (Young, 1948; Gentle et al., 1965; Pritchett and Humphreys, 1971; Pritchett and Smith, 1974). Its use in operational fertilization is recommended on certain soils in Britain (Leyton, 1958), Australia (Gentle and Humphreys, 1968), Finland (Salonen, 1967), Ireland (O'Carroll, 1967) and the southeastern USA (CRIFF annual report, 1973). The soils on which slowly soluble rock phosphates are recommended are principally acid sands and peats with low P-retention ability. Other P sources used to a limited extent in operational fertilization include diammonium phosphate (DAP) (Pritchett and Hanna, 1969) and basic slag (O'Carroll, 1967). In order to fully understand the difference in effectiveness of P sources of different solubility, a knowledge of their reaction with soil constituents is necessary. Reaction in Soils

Water-soluble monocalcium phosphate (MCP) is the principal P compound in both OSP and CSP. The reactions and transformations of MCP following addition to soil have been reviewed in several publications (Huffman, 1962 and 1968; Olsen and Flowerday, 1971). Initially MCP undergoes hydrolysis producing an extremely acid (pH 1.48) zone of phosphoric acid around the pellet. The phosphoric acid diffuses into the soil leaving a residue of dicalcium phosphate dihydrate (DCPD) (Lindsay and Stephenson, 1959 a). The acid H3P04 solution may cause dissolution of Fe, Al, Mn, Ca,





10



and other soil constituents. As the pH of the diffusing fertilizer solution rises following reaction with soil constituents and dilution by the soil solution, the solubility products of various Al and Fe phosphates are exceeded and they precipitate (Lindsay and Stephenson, 1959 b). The initial precipitates in an acid soil (Hartsells fine sandy loam, pH 4.6) were identified as colloidal products of the type (Fe, Al, X)PO4.nH20 and various crystalline acidic phosphates of the type X20.3(Al; Fe)203.6P205 .20H20. The Fe and Al may substitute for one another, and the X indicates a cation other than Fe and Al (Lindsay and Stephenson, 1959 b; Lindsay, Frazier, and Stephenson, 1962).

Many of the initial reaction products formed in the soil following addition of MCP have been reported to be good sources of plant-available P (Taylor, Gurney and Lindsay, 1960; Juo and Ellis, 1968). However, as the initial colloidal and poorly crystalline products age, they become more crystalline and less available sources of P (Juo and Ellis, 1968). Stable crystalline phosphate minerals of the variscite-strengite series are considered to be the ultimate end-product of the reactions and transformations of MCP in acid soils (Wright and Peech, 1960). Juo and Ellis (1968) reported that Fe phosphates crystallize more rapidly than Al phosphates in acid soils. They observed that amorphous Fe-P forms crystallized within 9 months at 35 C, while amorphous Al-P forms remained completely amorphous over the same period. This slower rate of crystallization of Al phosphates was cited by Juo and Ellis (1968) as the probable explanation why colloidal Al-P was more effective than colloidal Fe-P as a long term source of P for plants (Taylor et al., 1960) and explained the numerous reports of close correlations between Al-P (extracted by alkaline NH4F) and plant-P uptake (Thomas and Peaslee, 1973).









The transformations of the residual DCPD are greatly influenced by soil pH. Moreno, Brown, and Osborn (1960) found that DCPD in aqueous solutions hydrolysed to the more basic octacalcium phosphate (OCP) at a pH of 6.38 and that the rate of hydrolysis increased at higher pH values. Bell and Black (1970 a) reported that DCPD reverted to OCP at 35 C in soil at pH 6.4, but at 25 C a pH of at least 6.9 was required before OCP became detectable. Lehr and Brown (1958) found that both MCP and DCPD applied to alkaline soils (pH 8.0) reverted to OCP and hydroxyapatite. In acid soils, DCPD dissolves releasing Ca2+ and HP042- (Olsen and Flowerday, 1971). The rate of dissolution of DCPD, according to Moreno, Lindsay, and Osborn (1960) is a function of the rate of removal of Ca and P from solution by plants and/or the soil sorption and exchange sites. It has been reported that DCPD persists in acid soils for several months (Lehr and Brown, 1958). Moreno et al. (1960) found that once the dissolution of DCPD in acid soils was complete, P concentrations in soil solution decreased abruptly.

The reactions and transformations of RP following addition to soils have been less subject to research than has MCP. Apatite is the principal P form in RP, but it varies in degree of crystallinity and proportions of fluor and hydroxyapatite, depending on location of the RP source (Olsen and Flowerday, 1971). The dissolution of RP in soils is closely related to soil pH. Lindsay and Moreno (1960) developed solubility diagrams for predicting P in solution at various pH ranges in the presence of fluor and hydroxyapatite. For fluorapatite, the equation developed for predicting P in solution was:

p H2P04 = -5.13 + 2 pH





12



Huffman (1962) found that the rate of dissolution of fluor and hydroxyapatites was proportional to their surface areas and that this rate decreased with a rise in pH. Mattson et al. (1951) reported that the addition of organic matter, which complexed Ca, increased the solubility of hydroxyapatite about tenfold. Even in acid soils, the dissolution rate of RP has been found to be slow. Chu, Moschler,and Thomas (1962) reported that 18% of the RP applied 4 years previously to a soil at pH 5.7 (Nason silt loam) had dissolved, while only 7.5% had dissolved 7 years after application to a soil at pH 6.1 (Frederick silt loam). Gentle et al. (1965) found that most of the RP applied to a podsolized soil at pH 5 was still present as Ca phosphates after 15 years. Similarly, in the southeastern Coastal Plain,Humphreys and Pritchett (1971) reported that 7 to II years after the addition of RP to 7 soils ranging in pH from 4.2 to 5.4, substantial amounts of P were recovered in the Ca-P fraction. In all three of these studies, the transformation of RP produced an increase in the AI-P and Fe-P fractions.

Diammonium phosphate (DAP) is water soluble and upon dissolution in the soil produces an alkaline solution of about pH 8.0. The initial reaction of the fertilizer solution produces basic phosphates of the general type (Ca, Mg) (NH )2(HP04)2.2H20 (Lindsay et al., 1962; Bell and Black, 1970 b). In slightly acid soils, Bell and Black (1970 b) found DCPD was formed once the fertilizer solution had moved several centimeters from the fertilizer source and the pH decreased. Bell and Black (1970 b) also reported that in comparison with MCP and monoammonium phosphate, which both form acid solutions, the movement through soil columns of P from DAP was far greater, irrespective of pH and clay content of the soil.





13



Phosphate-fractionation studies following P additions and work relating P-retention capacity of soils to various soil properties support the findings discussed above that in acid soils Al and Fe are the principal soil constituents with which P reacts. The literature on both these subjects is voluminous and for the sake of clarity and brevity will be illustrated by work in the southeastern USA.

Yuan, Robertson, and Neller (1960) found in laboratory tests that water-soluble P rapidly reverted to Al-P upon addition to three acid sandy Florida soils. The addition of CSP to a Lakeland fine sand in a short-term lysimeter study was found by Hortenstine (1969) to significantly increase only water soluble and Al-P fractions. Fiskell and Spencer (1964) reported that 6 years after the addition of a heavy application of CSP to Lakeland fine sand, the increase in P fractions followed the order AI-P>Fe-P>Ca-P. A similar result was reported by Robertson, Thompson, and Hutton (1966) with a Red Bay fine sandy loam and a Norfolk loamy fine sand. However, they reported that the ratio of Al-P to Fe-P decreased with time. They attributed this decrease to uptake of Al-P by plants and a transformation of Al-P to less soluble and less available Fe-P. Phosphorus-fractionation studies on seven forest soils in the southeastern USA (Humphreys and Pritchett, 1971) revealed that 7 to 11 years after application of OSP the majority of the recoverable P was found in the Al-P and Fe-P fractions. The distribution between the two fractions varied between soils. Equal amounts of each fraction were found in a Bladen soil, while nearly all applied P was recovered in the Al-P form in a Rutlege soil. Two Kershaw soils were found to be intermediate in behavior between the Bladen and Rutlege soils. The interpretation of these fractionation studies must take into consideration the limitations of the fractionation





14



procedure. Although all studies, except that by Hortenstine (1964), used the Chang and Jackson (1957) prodecure with Fife's (1959) modification of the NH4F extraction, work by Bromfield (1967) has shown that on recently fertilized soils the prodecure can provide a misleading estimate of the relative amounts of Al-P and Fe-P. This is due to the ability of NH4F at pH 8.2 to extract appreciable DCPD which leads to an over estimation of

AI-P

In acid soils, Al and Fe contents have been shown to be closely

correlated with P-retention capacity. Coleman, Thorup,and Jackson (1960), working with subsoil samples from 60 North Carolina Piedmont soils, found that exchangeable Al extracted by N KCI was most closely correlated with P retention. Similar findings were reported by Syers et al. (1971) using 15 topsoil samples from Brazil and by Udo and Uzu (1972) using 10 acid Nigerian soils. Aluminum extracted by (NH4)2C204 (pH 3.0), a reagent which extracts amorphous forms of Fe and Al, has been reported to be closely related to P retention in a range of acid soils (Bromfield,1965; Saunders, 1965; Yuan and Breland, 1969). Yuan and Breland (1969) used the horizons of 43 virgin soils representative of the soil orders found in the southeastern Coastal Plain. Iron extracted by oxalate generally proved to be less effective than oxalate-extractable Al for predicting P retention (Bromfield, 1965; Yuan and Breland, 1969; Syers et al., 1971). However on some soils rich in iron oxides, oxalate-extractable Fe proved to be more successful than Al at predicting P retention (Ahenkorah, 1968). Extractants which remove crystalline forms of Fe and Al (citrate-dithionatebicarbonate, termed herein as CDB) were generally less effective than oxalate for extracting amounts of Al and Fe related to P retention. However CDB extractable-Al appears to be more closely related to P retention




15



than CDB extractable-Fe (Syers et al., 1971; Udo and Uzu, 1972). These results have been interpreted as indicating that the relative activity of the various soil forms of Al and Fe in P retention follows the order: exchangeable>amorphous>crystalline; while, within each form, Al is more active than Fe on a per unit weight basis (Bromfield, 1965; Syers et al., 1971; Udo and Uzu, 1972). Soil constituents that contribute to P retention in soils other than various Al and Fe forms include phyllosilicate clays and CaCO3 (Kuo and Lotse, 1972). Calcium carbonate is of little significance in virgin acid soils, but kaolinite, which is more effective at retaining P than 2:1 clays (Ramula, Pratt, and Page, 1967) can retain up to 187og P/g (Kuo and Lotse, 1972). Nevertheless, kaolinite is less reactive than other forms of Al and Fe in the soil (Huffman, 1968). P Effectiveness in Forestry

.The majority of fertilizer trials comparing the effectiveness of

different P sources in forestry have involved water-soluble superphosphate (OSP or CSP) and slowly soluble rock phosphate (RP) sources.

Early trials in Australia comparing the effectiveness of OSP and RP gave similar results: Both RP and OSP applied at equivalent rates of P were equally effective as long-term P sources (Young, 1948; Richards, 1956; Richards, 1961; Gentle et al., 1965). Although an adequate description of the soil properties was not provided in most of these reports, it was mentioned that all soils were acid. Young (1948) and Richards (1956) pointed out that RP was a more economical source than OSP, because it was cheaper and more effective per unit weight of fertilizer applied. Gentle and Humphreys (1968) reported that a trial established in Penrose State Forest, New South Wales, in which OSP and RP were applied at equal rates of P (77 kg/ha), showed OSP to have an early superiority, but after




16



15 years, current increment in the RP treatment was significantly greater than that in the OSP treatment. The decline in effectiveness of the OSP was attributed to the high P-retention capacity of the soil. Working on the sandy coastal plain of Western Australia, Hopkins (1960) found RP to be the most effective source of P for P. pinaster. This was attributed to less leaching of P from RP than from OSP on these leached, acid,sandy soils.

In summarizing the results from fertilizer trials in New South

Wales, Gentle and Humphreys (1968) suggested that on soils with a high Pretention capacity, banded applications of a semi-soluble source should be used,whilst on near neutral soils with a medium P-retention capacity, OSP or CSP would be preferred. On acid soils with a low P-retention capacity they suggested the use of RP.

Rock phosphates are the preferred source of P for fertilizing

forests on acid peats in England (Leyton, 1958), Finland (Salonen, 1967) and Ireland (Dickson, 1971). The success of RP is probably due to the acidity and low P-retention capacity of these peats, although there is no reported evidence in the literature to indicate leaching losses of soluble P sources applied to peats. In some instances, however, the selection of RP was based on economic considerations, since OSP and RP were equally effective (Dickson, 1971). Hagenzieker (1958) also reported that OSP, RP and basic slag were equally effective sources on the acid forest soils of Holland.

Bengtson (1970) examined the comparative value of CSP and RP as P sources for slash pine in greenhouse studies with a Lakeland fine sand (pH 4.8). Using different placements, he found CSP to be superior to RP in all tests. The performance of RP was improved by mixing it with





17



the soil as opposed to broadcast or banded placement. Brendemuehl (1970) reported a similar improvement of performance in field trials from mixing RP with the same soil type.

The results from a series of trials established using slash pine to compare the effectiveness of OSP and RP on soils in the southeastern lower Coastal Plain have been reported in three publications (Pritchett and Llewellyn, 1966; Pritchett, 1968; Humphreys and Pritchett, 1971). The response in 1964, 3 to 5 years after establishment, showed that at equal rates of P, OSP was superior to RP in all trials (Pritchett and Llewellyn, 1966). Increment growth during the next 3 years, 1964 to 1967, was still greater from OSP than RP in all trials except on a Leon fs. Response from the highest RP treatment, which contained eight times as much P as the highest OSP treatment, was generally greater than that from the OSP treatments (Pritchett, 1968). Humphreys and Pritchett (1971) reported growth data only for the highest OSP and RP treatments which did not have comparable P rates. However, 7 to 11 years after the fertilizer application,they found that in the three Spodosols with virtually no P-retention capacity, all the P applied as OSP had been leached from the upper horizons, while most of that applied as RP was still retained in the surface horizons. In the remaining soils, which ranged in P-retention capacity from medium to very high, the majority of P applied either as OSP or RP was retained in the surface horizons. In the most retentive soil (Bladen), a considerable portion of the OSP had been converted to Fe-P, a relatively unavailable form as reflected by the extremely low foliar P levels reported for this treatment.

Humphreys and Pritchett (1971) suggested that selection of a suitable P source for the acid Coastal Plain soils should be based on their





18



P-retention capacity. On soils with virtually no P retention (Spodosols) they recommended the use of RP or other slowly soluble sources, whilst on soils of low to intermediate retention capacity they suggested a mixture of a soluble source, to provide initial high availability, and a slowly soluble source to maintain the P supply in later years (5 to 15 years). For soils with high P-retention capacity they suggested either repeated applications of a soluble source or banded application of a suitable mixture of soluble and slowly soluble P sources.

Pritchett and Smith (1974) recently reported the growth and response data obtained 10 years after application of fertilizer in the experiment on the site with the greatest P-retention capacity (Bladen fine sandy loam). Their results showed the classical trend anticipated from a longterm comparison of a soluble and slowly soluble source on a highly Pretentive acid soil: Initial response was greatest from OSP, but after 8 years the response from an equivalent rate of RP was equally great. Current annual increment during the ninth year showed RP to have a considerable advantage over OSP. This was reflected in the foliar P levels, because only the trees in the RP treatments had foliar P levels above the critical response range of 0.085-0.090%.

Although RP has been used successfully in many field fertilizer trials, its use in operational fertilization has been restricted by its bulk, which adds significantly to transportation costs, and the difficulty in spreading these finely ground materials--particularly by air (Pritchett and Smith, 1969; Conway, 1962).

The success with RP in forest fertilizer trials is generally attributed to three factors, namely: (1) The acidity of most forest soils (Gentle and Humphreys, 1968; Bengtson, 1970), (2) the relatively long









vegetative growth period and large root system of trees (Terman, 1968), and (3) the presence of mycorrhizal rootlets on coniferous trees which have been reported to increase the trees ability to utilize less soluble forms of P (Bowen, 1973).

Although not extensively tested, DAP has proved to be an effective fertilizer in the Southeast, particularly on sites where responses to N and P are synergistic (Pritchett and Smith, 1974). This source is recommended for operational fertilization on these sites; however, its long term effectiveness on soils of high P-retention capacity is likely to be limited. Furthermore, White and Pritchett (1970) found that 5 years after surface application of DAP to a Leon soil less than 11% of the applied P remained in the surface 20 cm of this soil having a low P-retention capacity.






20



Diagnostic Methods

This section of the review is concerned principally with the use of foliar and soil analysis as diagnostic aids in predicting P deficiency and the need for fertilizer in pine plantations. Major emphasis is given to the use of soil analysis since this is the method under investigation in this study.


Foliar Analysis

The principles, physiological basis,and problems inherent in the use of.tissue analysis for diagnosing the nutritional status of crops have been extensively reviewed by Goodall and Gregory (1947) and Smith (1962). The value and specific problems associated with the use of foliar analysis in forestry have been the subject of several extensive reviews (Tamm, 1964; Qureshi and Srivastava, 1966; Raupach, 1967; Richards and Bevege, 1972 a; Armson, 1973; Leaf, 1973).

Published information on foliar P concentrations in relation to responses of slash and loblolly pines (P. elliottii and P. taeda) to P fertilizer is summarized in Table 1. Whereas most research on the use of foliar analysis in forestry has been concerned with relating foliar nutrient levels to productivity or site index (Leaf, 1973), foliar analysis research with Southern pine has been almost exclusively related to establishing 'critical levels' for both predicting fertilizer needs and following the effectiveness of fertilizer applications (Table 1).

The critical level of foliar P has been defined by Pritchett (1968) as the point at which trees with a higher concentration of needle P would not be expected to respond significantly to applications of phosphate fertilizers, but at which trees with a lower needle P






21

Table 1. Foliar P concentrations prior to fertilization in relation to
response of southern pines to P fertilizer


Tree Foliar P Response to Reference
age (Unfertilized) P fertilizer

yr. %

(A) Slash pine
(Pinus elliottii)

8 0.053 yes Young (1948)

6 0.048 yes Baur (1959) 15 0.075 yes Pritchett and
0.105 yes Swinford (1961)
9 0.08 no Walker and Youngberg (1962)

3-5 <0.10 yes Pritchett and >0.10 no Llewellyn (1967) 5-8 <0.09-0.10 yes Pritchett (1968) >0.09-0.10 no

1 0.08 no Schultz (1969) 1-30 <0.075-0.80 yes (90%)* Richards and >0.075-0.80 no (90%) Bevege (1972b)
<0.085-0.10 yes (100%) >0.085-0.10 no (100%)

(B) Loblolly pine
(Pinus taeda)
8 0.064 yes Young (1948)

6 0.057 yes Baur (1959) 1 0.089 yes Fowells and Krauss (1959)
5-9 0.090 no Zahner (1959) 5-10 0.137 no Maki (1960) 22 0.124 no Thompson (1960)





22


Table 1. Continued



Tree Foliar P Response to age (Unfertilized) P fertilizer Reference


yr. %

1 <0.11 yes Wells and >0.11 no Crutchfield (1969)

10 0.11 no Moschler, Jones, and Adams (1970)

1-40 <0.095-0.105 yes (90%)* Richards and
>0.095-0.105 no (90%) Bevege (1972 b)
<0.13-0.14 yes (100%) >0.13-0.14 no (100%)

3 <0.10 yes Wells et al. (1973) >0.10 no


*Associated with 90 or 100% of maximum yield.





23



concentration would normally respond. This definition corresponds to the 'optimum concentration' proposed by Richards and Bevege (1972 a) who defined the 'critical concentration' as the concentration associated with 90% of maximum yield.

Richards and Bevege (1972 a) pointed out that critical levels can be established with accuracy only where all nutrients apart from the one under consideration are not limiting. However, Leyton and Armson (1955) suggested that difficulties in interpretation caused by nutrient interactions could be reduced by extending investigations over a wide range of conditions to allow for a statistical analysis of the relationships. Information in Table I shows that where a number of trials and statistical analyses were used, there is good agreement between the proposed critical levels for slash pine (Pritchett, 1968; Richards and Bevege, 1972 b) and loblolly pine (Wells and Crutchfield, 1969; Richards and Bevege, 1972 b; Wells et al., 1973). These data suggest that the best current estimates of critical foliar P concentrations (Pritchett, 1968) for P. elliottii and P. taeda are in the range 0.085-0.10% and 0.095-0.105% respectively. The failure of foliar P values to provide the correct diagnosis in some of the single trials reported in Table I can probably be attributed to other factors such as other nutrients or site factors limiting production or to differences in sampling procedures (Leaf, 1973).

In studies where both foliar and soil analysis have been used to predict response to P fertilizer, foliar analysis has generally proved to be the more effective (Pritchett, 1968; Wells and Crutchfield, 1969; Wells et al., 1973). However, it has been pointed out that foliar analysis has practical disadvantages as a diagnostic tool in forestry. Its use is largely limited to areas where trees are established (Pritchett, (1968). Restricted sampling periods and the time required to collect an





24


adequate sample, even in small areas of mature stands, makes foliar analisis rather impracticable for large scale routine diagnosis (Wilde, 1958).

Richards and Bevege (1972 a) suggested that for management purposes, plant analysis must be quantitative: They stated that

... at the very least it must provide some measure of the degree
of nutrient deficiency, while ideally it should enable us to
predict the magnitude of response to a given application of fertilizer. (p. 110)

The response curves shown by Pritchett (1968) are apparently the only reported fully quantitative foliar analysis data for trees. A report by Wells et al. (1973) related foliar P to a response index but provided no information on the amount of fertilizer required to achieve a particular response. Cain (1959) pointed out that the use of foliar analysis to predict quantitative fertilizer requirements could be difficult. He concluded this because the amount of fertilizer required to produce a desired increment of concentration in plant tissue varies greatly from site to site with variation caused by climatic and soil factors relating to the movement and retention of nutrients in the soil.


Soil Analysis

Soil nutrient analysis has been used in forestry, both as a

technique to predict site productivity and as a diagnostic aid in determining the need for fertilizer (Armson, 1973). Until recently the emphasis in soil-site studies was on soil physical properties (Coile, 1952). This was commented on by Voigt (1958):

Students of forest soils commonly come away from the literature with the impression that nearly all tree growth can be explained
almost completely by the so-called physical properties of the
soil--particularly those related to its moisture regime. (p. 31) Ralston (1964) attributed the neglect of soil fertility parameters to their frequent correlation with other soil properties commonly used in





25


site index studies, and the lack of soil testing techniques of known significance in relation to tree requirements. However, since the widespread demonstration of the responsiveness of trees to fertilizers (Mustanoja and Leaf, 1965), soil nutrient analysis has received more attention as a technique for predicting site productivity (Pawluk and Arneman, 1961; Wilde et al., 1964 a; Schomaker and Rudoph, 1964; la Bastide and van Goor, 1970; Alban, 1972).

As a diagnostic technique for predicting site nutrient status and the need for fertilizer, soil analysis has received considerably less attention than foliar analysis (Leaf, 1968). According to Leaf (1968), current thinking on the value of soil and foliar analysis still reflects the ideas of Mitchell and Chandler (1939) who stated:

The distinct advantage of the method of leaf analysis is that by
chemical analysis of the leaves we can obtain a more reliable
estimate of the amount of the various nutritional elements which
have been adsorbed by, and therefore available to, plants growing in a given soil. One therefore uses a natural biological rather
than an artificial extraction method for estimating available
nutrients. (p. 75)

Several factors have been cited in the literature as reasons why soil testing has not been widely accepted as a diagnostic tool by forest managers. Most of these are related to insufficient information on methods and procedures.


Sampling procedures

The determination of tree nutrient requirements is complicated by a deep root system. Tamm (1964) suggested that the main rooting zone should be sampled thoroughly and consideration should be given to lower soil horizons when they contain notable root concentrations. In discussing this problem, Voigt (1958) pointed out that most forest trees, as opposed to agricultural crops, have two distinct rooting regions; one






26


predominantly organic and the other predominantly inorganic. He suggested that it is doubtful if a single extraction agent could yield a valid estimate of nutrient availability in view of the behavior differences of organic and inorganic colloids in nutrient retention.

The value of sampling subsurface horizons has been examined in many studies. Kessell and Stoate (1938) recognized the importance of the vertical distribution of soil P in their site classification studies. In outlining the P levels required for successful afforestation of radiata pine (P.radiata D. Don), they wrote:

A P205 content of 400 parts per million (ppm) i.s required in the.
-surface and subsurface soils. Three hundred parts per million
-may be satisfactory if this content is maintained for a depth of
two to three feet. (p. 28)

In a survey of factors influencing the growth of radiata pine in New Zealand, Jackson (1973) found that available soil P within the effective rooting zone was more closely correlated with growth than with available P in the surface 7.5 cm of soil. Similarly, White and Leaf (1964) reported that the content of HNO3 extractable K in the upper 1.7 metres of soil was more closely correlated with height and K content in the biomas of red pine (P. resinosa AIT) than was the extractable K in the surface horizon. The importance of subsurface nutrient supply was illustrated by the work of Ellerbe and Smith (1963) who found that occasional serious underestimation of site quality by the soil-site predictions of Coile (1952) in the lower Coastal Plain of South Carolina could be attributed to the presence of underlying phosphate marl at these sites. Will (1966) also reported that the disappearance of Mg deficiency in radiate pine growing on volcanic soils in New Zealand coincided with the root penetration of Mg-rich buried topsoil at age 6 to 8 years.





27

Several investigators have reported P levels in the surface

horizon to be more closely related to growth and foliar P levels than P levels in lower horizons. Humphreys (1963) outlined Australian work in which it was reported that soil P below 37 cm contributed little to the P supply of mature pines. Pawluk and Arneman (1961) found that available P in the A horizon, but not the B horizon, was correlated with site productivity of jack pine (P. banksiana Lamb). Similarly, Wells (1965) reported that variation in concentration of foliar P of 5-year-old loblolly pine was better explained by available P in the A than in the B horizons. Alban (1972) also found that where P extracted from the 25 to 100 cm depth was included, it did not improve the estimation of the site index of red pine in Minnesota over that obtained using P extracted from the 0 to 25 cm depth. The apparently greater importance of P in the surface horizon is attributed by most workers to the concentration of fine feeder roots in the upper profile (Alban, 1972).

As in agriculture, horizontal soil variability is also a problem in securing representative soil samples. The significance and extent of variability of various soil properties in forest soils has been discussed (York, 1959; Mader, 1963) and examined (McFee and Stone, 1965; Metz et al., 1966). In reviewing these publications, Leaf (1968) pointed out that the patterns and magnitudes of variability differ between soil characteristics at a particular site and determination of required sampling intensity is a statistical problem.

In his review article, Leaf (1968) mentioned a problem in soil

sampling which is fairly unique when dealing with deep rooted perennials. He stated:

...different degrees of biologically important soil heterogenity
exist on any site depending on the stage of tree development, e.g., seedling or establishiment stage vs. sapling stage vs. mature tree
stage, with the associated soil volume being tapped by the roots.
(p. 93)





28

This aspect has received little attention in the literature, but its importance is well illustrated by the work of Will (1966) discussed earlier.


Extraction methods

Tamm (1964) pointed out that most soil analysis techniques were

those developed for agricultural crops and are of unknown value for predicting tree nutrient requirements. Most techniques used in agriculture today for determining plant-available P are largely empirical (Williams, 1962). However, considerable research relating the availability to agricultural plants of various soil-P fractions and determining the solubility of these same fractions in the extractants used in soil-P tests, has provided a theoretical basis for these soil-P tests (Thomas and Peaslee, 1973). Little research of this nature has been conducted with trees. However, knowing the forms of P extracted by various soil tests, some insight into the forms of P utilized by trees can be obtained.


Total P. Early work in Australia indicated that total soil P provided a good index of site productivity and adequately delineated areas of P deficiency (Kessell and Stoate, 1938; Young, 1940 and 1948). This work helped foster the concept that trees could utilize sources of P unavailable to other plants. This ability has been attributed to the presence of mycorrhizae (Pritchett, 1968) which have been shown to solubilize otherwise insoluble sources of P (Rosendahl, 1943). However, Tamm (1964) has suggested the ability of tree roots to utilize P unavailable to agricultural crops could be the consequence not so much of higher efficiency as of longer persistence.

Stoate (1950) found that certain anomalies occurred when using total P for predicting site nutrient status. He reported that these





29

anomalies could be accounted for by differences in the level of P extracted by 1% citric acid. Baur (1959) reported that total-P values were related to site productivity only within limited localities and not over more heterogenous sites, a finding substantiated by Humphreys (1964). Similarly, Ballard (1970 a) found that total P was significantly correlated with productivity of radiata pine only on groups of genetically similar soils. When all soils were grouped together, the Olsen (O.5M NaHCO3, pH 8.5) and Bray 2 (0.03N NH4F + O.IN HCI) tests for available-P provided the best index of productivity and foliar-P concentrations. In the southeastern USA, Pritchett and Llewellyn (1966) found total P to. be of little value in delineating sites responsive to P fertilizer. Wells (1965) also found total P to be inferior to available P for predicting foliar-P concentrations of loblolly pine.


Extractable P vs site productivity. Despite the many suggestions that soil-test methods derived for agricultural purposes are unlikely to be satisfactory for use in forestry, reports of their successful use in forestry are frequent in recent literature. Pawluk and Arneman (1961) reported a significant correlation between Bray 1 extractable P (0.03N NH4F + 0.025N HCI) in the A2 horizon and the site index of 50-yearold jack pine in Minnesota. On acid soils, such as used in this study, the Bray 1 test extracts P predominantly from the Al-P and Fe-P fractions (Thomas and Peaslee, 1973). On non-phreatic sandy soils in Wisconsin ranging in pH from 4.3 to 7.0, Wilde et al. (1964 a and 1964 b) reported a significant correlation between P extracted by the Truog soil test (0.002N H2SO4 + 3g (NH4)2S04/1) and the site index of both Jack pine and red pine. The Truog test extracts P from the Ca-P and to a lesser extent the AI-P fractions (Thomas and Peaslee, 1973).





30

Working with young plantations of loblolly pine in the lower

Coastal Plain of South Carolina, Wells and Crutchfield (1969) reported that soil P extracted by O.05N HCI + 0.025N H2SO4 was not significantly correlated with the first year height growth. The authors suggested the soil-test values would prove more reliable as the trees became older. This was confirmed by Wells et al. (1973), who found soil P extracted by the above extractant to be significantly correlated with the height of unfertilized loblolly pine at age 3 years in the same region. They also reported the Bray 2 extractant to be equally effective. According to Thomas and Peaslee (1973), 0.05N HCI + 0.025N H2SO4 extracts P predominantly from the Ca-P and Al-P fractions. However, the acid soils of the southeastern lower Coastal Plain contain little Ca-P (Humphreys and Pritchett, 1971). The effect of age of trees on the value of soil-P tests as predictors of height growth was also observed by Ballard (1974). Working with radiata pine on heavy textured acid soils in New Zealand, he found that levels of soil P were most closely related to height at age 3, followed by height at age 2 and least with height at age 1 year. The Bray 2 and Olsen extractants (0.5M NaHCO3, pH 8.5) were the most successful in this study. On acid soils, the Olsen test extracts P from the Al-P and Fe-P fractions, although with somewhat less intensity than extractants,such as Bray,containing F (Thomas and Peaslee, 1973).


Extractable P vs foliar P. There are numerous reports in the literature of correlations between foliar-P concentrations in pine species and levels of soil P extracted by various soil tests; Wells (1965) found P extracted by the Truog and Bray 2 tests to satisfactorily predict foliar P of 5-year old loblolly pine growing in the South Carolina Piedmont. However, Metz et al. (1966), in an intensive study on one soil





31

(Helena sandy loam) in South Carolina, found P extracted by the Truog test accounted for only 17% of the variability of foliar P in loblolly pine. Phosphorus extracted by the Olsen and Bray 2 tests was found to be significantly correlated with foliar P of 40-year radiata pine (Ballard, 1970 a). In an intensive study involving 63 plots in Florida and 60 in Australia, water-extractable P was found to provide the single best index of foliar-P concentrations in slash pine (Humphreys and Pritchett, 1972). However, the inclusion of NaOH-extractable P in their prediction equation significantly improved the estimate of foliar P. The authors suggested that both the intensity and quantity factors of soil-P supply are of importance in determining foliar-P concentrations. Alban (1972) also reported that extractants which removed relatively small amounts of P from the soil (H20, 0.002N H2S04, 0.OIN HCI) provided a better index of foliarP concentrations of 49-to 94-year-old red pine. Although Alban also examined stronger extractants and found them less effective than the above weaker extractants, he did not use multiple regression techniques to examine the possibility that the combined use of a weak and strong extractant may have improved the prediction, as did Humphreys and Pritchett (1972). Using a greenhouse trial, Baker and Brendemuehl (1972) found P extracted by NH4OAc (pH 4.8) correlated significantly with both foliar-P concentrations and dry weights of sand pine (P. clausa Vasey) seedlings growing on a Lakeland fine sand. This weak acid extractant removes small quantities of P from the soil by forming weak complexes with polyvalent cations (Al, Fe, Ca) in soils (Thomas and Peaslee, 1973). Pritchett and Llewellyn (1966) also found NH OAc (pH 4.8) to extract P that provided a good index of the growth of 3-to 5-year-old slash pine in the absence of fertilizer. The trees were growing on acid sandy soils in the lower Coastal Plain which varied in drainage characteristics from





32

excessively to poorly drained. In their study, P extracted by NH40Ac (pH 4.8) was reported to be superior to that extracted by other extractants, including 0.03N NH4F + 0.025N HCI and 0.05N NCI + 0.025NH2SO4, at predicting response to P fertilizer. However significant correlations were obtained only when the well-drained sites were eliminated from the statistical analysis.


Forms of tree-available P. McKee (1973) concluded from a greenhouse fertilizer trial with N and P rates using slash pine grown in a Caddo silt loam, that Ca-P and Al-P were the sources of P available to the pine seedlings. This conclusion was based on the finding that the combined value of these two fractions in the soil at the end of the experiment was significantly correlated with P uptake by the seedlings. He found individual fractions of Ca-P, Fe-P, and AI-P were not related to P uptake, nor was Fe-P in combination with either Ca-P or Al-P. The relevance of results from greenhouse trials to the nutrition of field grown trees must however be questioned in view of the finding by Mead and Pritchett (1971) that the response, and hence nutrition, of greenhouse grown seedlings differed from that of field grown trees on the same

soils.

The literature reviewed in this section provides no conclusive

evidence as to what forms of soil P are utilized by trees and which soil testing procedures are likely to be most successful in forestry. The situation appears to parallel that in agriculture in which it has been found that the value of any particular soil test depends upon the objective in soil testing (predicting growth, nutrient uptake, or responsiveness to fertilizer addition), the range of soils used, and the species of plant involved (Williams, 1962; Williams and Knight, 1963). However, despite some evidence that mycorrhizae may increase the availability of





33

insoluble sources of inorganic and organic P (Bowen, 1973), the evidence reviewed above tends to indicate that extractant which remove plantavailable P are likely to be of more value than total P analyses for predicting forest site fertility.

A problem in evaluating nutrient status of forest sites which has received little attention was pointed out by Voigt (1958). He suggested that the perennial nature of trees is likely to present problems in evaluating the .nutrient status since the rate of nutrient cycling rather than the level of a particular fraction in the soil may be of importance in determining whether the trees' nutrient requirements are met. For P, this is likely to be of particular importance in closed canopy stands of trees, as Will (1964) reported that after canopy closure the P requirements of radiata pine were met almost entirely by nutrient cycling. However, he found that prior to canopy closure the tree-P requirements were met almost entirely by net withdrawl from the soil. Ballard (1970 a) suggested that the reason for the significant correlations he found between P extracted by several available-P tests and foliar-P concentrations of 40-year-old radiata pine was due to these methods reflecting the availability of organic P once it had been mineralized.


Interpreting soil test results

The successful use of soil tests depends entirely upon their prior calibration against crop performance in the field (Williams, 1962). In forestry, most studies have been concerned with calibrating soil-test values against site productivity, their objective being to establish critical levels below which growth performance of trees decline.


Prediction of productivity. Young (1940) found that a total-P

value in. the topsoil of 47 ppm delineated areas of slash pine which showed





34

symptoms of fused needle in Australia. Young (1940) showed that fused needle, a symptom of unthrifty pine trees, coul-d be corrected by the application of P fertilizer. On the basis of significant correlations between levels of P extracted by the Truog test and the productivity of both jack and red pine, Wilde et al. (1964 a, 1964 b) proposed that minimum P levels required for the establishment of these species were

6.5 and 5.4 ppm respectively in the surface of 15 cm of soil. Ballard (1970 a) proposed that values of 3.5 and 5.0 ppm P by the Olsen and Bray

2 tests were required in the surface 10 cm of soil for satisfactory growth of radiata pine.

Minimum or critical values established from relationships between soil P and productivity cannot be used with confidence for predicting ameliorative practices for sites testing below these values. Dahl, Selmer-Anderssen, and Saether (1961) pointed out that a significant correlation between nutrient levels in the soil and site productivity is not proof that this nutrient is controlling growth. For instance, these authors reanalyzed the data of Viro (1955), who recorded a significant correlation between EDTA-extractable P and the site index of Scots pine (P. sylvestris). They found that extractable P was related to site index only indirectly through its correlation with extractable soil Ca levels. Proof of a casual relationship must be obtained by either field fertilizer trials or the establishment of a positive relationship between site productivity and both the concentration of a nutrient in the foliage and its availability in the soil (Leyton and Armson, 1955).


Prediction of fertilizer response. Before soil-test results can

be used in developing fertilizer recommendations, field studies must show that the test will be useful for differentiating soils into those groups






35

on which trees will respond and those on which trees will not respond to fertilization (Wells et al., 1973). According to Mader (1973), the paucity of well-designed fertilizer trials, suitable for use in calibrating soil-test results against response over a wide range of conditions, is the major reason for the lack of suitably calibrated soil tests. However, this situation has been alleviated to some extent by the establishment of a large number of uniform fertilizer trials by forest fertilization cooperatives based in Florida, North Carolina and Washington State (Bengtson, 1972).

Information published on the response of southern pines in relation to soil-P levels is summarized in Table 2. Early work in Australia showed that total soil-P values could be used for delineating sites responsive to P fertilizers (Young, 1948; Richards, 1956; Baur, 1959). However, as discussed earlier, these relationships were found to hold true only over limited areas. This is exemplified in Table 2 by the different critical values reported by the various authors. The only attempt at using total-P values for other than delineating unresponsive sites was that reported by Baur (1959). He found that the quantity of P fertilizer required to produce adequate growth in the field corresponded closely with the quantities calculated to raise total P to the optimum of 70 ppm P (based on 125 kg/ha of superphosphate being equivalent to 6 ppm P). The studies reported by Pritchett (1968) and Pritchett and Llewellyn (1966) for slash pine, and Wells et al. (1973) for loblolly pine are the only ones in which relationships were derived statistically using a range of sites. These studies were conducted in the Coastal Plain region of the southeastern USA. Pritchett and Llewellyn (1966) and Pritchett (1968) reported that 2.0 ppm P by NH40Ac (pH 4.8) extraction separated sites responsive to P fertilizer from unresponsive sites. This test proved





36


superior to several others examined, including 0.05N HCI + 0.025N H2S04 and Bray 2. These authors also gave calibration curves indicating the amount of fertilizer required to give a specific degree of response at a particular soil-test value. Wells et al. (1973) reported that P extracted by 0.05N HCI + 0.025N H2SO4 was superior to that extracted by the Bray 2 test at delineating responsive sites. From their response curves, they proposed a value of 3.0 ppm P extracted by 0.05N HCl + 0.025N H2SO4 as' providing the best separation of responsive and unresponsive sites. Their calibration curves indicated only the degree of response expected at any particular soil-test value and not the amount of fertilizer required to achieve this response. The critical level found by Wells et al. (1973) for loblolly pine probably accounts for the lack of response by loblolly pine to P recorded by Carter and Lyle (1966) and Moschler et al. (1970) on soils testing above this critical level (Table 2).

Wells et al. (1973) reported several sites which,although ~ falling into the responsive category according to their soil-P test value, did not respond to P fertilizer. They attributed this in most cases to N being more limiting than P on these sites. This illustrates the need in calibration studies, pointed out by Williams (1962) and Richards and Bevege (1972), for utilizing field trials in which all nutrients except the one under observation are nonlimiting to growth. Wells et al. (1973) also pointed out that their critical value established for the response of 3-year-old loblolly, may not hold true for older trees due to the increasing P requirements of trees with age. As mentioned earlier, the importance of stand age or development to the nutritional demands of trees was discussed by Leaf (1968) and illustrated by the work of Will (1964).





37


Table 2. Soil-P values (surface horizon) of unfertilized soils in
relation to response of southern pines to P fertilizer


Tree Soil test . Response to Soil P Reference age type P fertilizer

yr. ppm

(A) Slash pine
(Pinus elliottii)

9 Total P <52 yes Young (1948) >52 no

21 Total P <65 yes Richards (1956) >65 no

3+ Total P <70 yes Baur (1959) >70 no

9 0.05N HCI + 29.5 no Walker and Youngberg
0.025N H2S04 (1962)

5-8 NH40Ac(pH 4.8) <2.0 yes Pritchett (1968) >2.0 no Pritchett and Llewellyn (1966)


(B) Loblolly pine
(Pinus taeda)

9 Total P <59 yes Young (1948) >59 no
21 Total P <91 yes Richards (1956) >91 no

6 0.03N NH F 7 no Merrifield and Foil
+ 0. IN HCI (1967)

6 0.05N HCI + 8 no Carter and Lyle
0.025N H2S04 (1966)

10 0.05N HCI 4.5 no Moschler et al.
+ 0.025N H2S04 (1970)

3 0.05N HCI <3.0 yes Wells et al. (1973)
+ 0.025N H2S04 >3.0 no





38


Provided soil analysis results can be suitably calibrated against tree growth and response to P fertilizer (and evidence suggests they can) soil analysis has certain advantages over foliar analysis. Pritchett (1968) outlined advantages for soil testing: (1) It can be used for predicting fertilizer needs in areas prior to planting. (2) Collecting and analyzing soil samples may be less laborious than collecting and analyzing needle samples, particularly in old stands. Ballard (1970 a) also suggested that soil analysis has an advantage in that soil samples can be collected at any time of the year, whereas foliage samples cannot.














MATERIALS AND METHODS


Introduction

A series of field, greenhouse and laboratory experiments designed to examine the effectiveness of soil-test methods for predicting the need for P fertilization of slash pine, included: a) A preliminary screening of a wide range of soil-test methods to

relate soil-P values with tree height, height response to P fertilizer and fertilizer requirements of slash pine in field and greenhouse experiments (greenhouse trial 1) on 10 soils. The most effective of these test methods were then calibrated against tree growth

and response parameters from 72 field fertilizer trials.

b) The solubility of a range of P compounds in soil-test extractants was

related to the capacity of slash pine seedlings to utilize P from

these compounds in a greenhouse test (greenhouse trial 2).

c) A P-retention study was conducted in both the laboratory and field,

using soils that varied widely in their ability to retain fertilizer

P.













39





40


Field Trials

A series of uniform fertilizer trials was established on each of 29 sites in March 1968 by the Cooperative Research in Forest Fertilization (CRIFF) program. The sites were selected to represent the principal forest soils of the Coastal Plain region. They included 22 soil types in the Spodosol, Ultisol, Entisol and Inceptisol orders. Fertilizer treatments were applied I to 2 months after planting of the sites with slash pine seedlings. Twenty-four of the trials which were still functional in 1972 were used in this study for evaluating soil testing procedures.

Prior to planting, most sites were burned and disced to remove

residual vegetation. This was followed on all sites by ridging to form beds approximately 1.22 m wide on 3.05 m centers on which the pine seedlings were planted. The experimental design consisted of three replications of 12 fertilizer treatments in randomized complete blocks. Each treatment plot was 27.5 by 30.5 m ( 0.08 ha) while the net measurement plot was 18.3 by 26.8 m. Nine of the twelve treatments formed a 3 x 3 factorial with N and P fertilizers. Application rates of both N and P were 0 (Po, No), 22.5 (PI, Ni), and 90 (P2, N2) kg/ha. Additional treatments numbered 10, 11, and 12 involved applications of K and micronutrients but these treatments were not included in the present study. Nitrogen was applied as ammonium nitrate and P as CSP. All fertilizer materials were applied in 1.22-m wide bands down tops of beds. This effectively gave nutrient concentrations within the bed of 2.5

-times those shown above.


Soil Sampling

Soil samples were collected from each replication at each site





41


prior to treatment by CRIFF personnel (Pritchett and Smith, 1972). The samples were collected from undisturbed areas between bedded tree rows from the 0-20, 20-40 and 40-60 cm depths.

Additional samples, specifically for use in this study, were collected in December-January of 1971/72, 4 years after the establishment of the trials. From the control plot (PoNo) of each replication at each site, four composite soil samples were collected using a 4-cm diameter, closed cylinder soil auger. Two composite samples were collected from the 0-20 cm depth--one from the bedded area and the other from the undisturbed interbed area. Each sample consisted of 12 cores collected randomly from within the net plot area. Four soil pits, one in each quarter of the plot, were dug in the interbed area. During excavation, two cores were taken at each pit from both the 20-40 and 40-60 cm depths. The eight cores from each depth were combined to provide a composite sample for each of these depths.

Bulk soil samples, for use in greenhouse trials (described

below), were collected from 10 trials. The trials were selected on the basis of soil and P-response information (Pritchett and Smith, 1972) to provide a wide range of soil-P levels and P-response characteristics. The bulk samples were collected from the 0-20 cm depth from a control plot in each of the 10 selected trials.. The samples consisted of subsamples collected randomly from the undisturbed interbed area.

In order to determine the extent of P leaching in the field, additional samples were collected from 10 selected trials in JanuaryFebruary, 1973. The trials were selected, on the basis of laboratory determinations of P-retention capacity, to provide sites with a range of retention capacities. Samples were collected from the 0-20 cm depth in the PINo and P2No plots of each replicate. The samples were








collected from the bedded area using the same equipment and procedure as used in the collection of the samples from control plots.

Following excavation of the four soil pits in each control plot, the following information was recorded: Depth and color of the Al horizon; and depth to mottling, spodic horizon, or fine-textured layer where they occurred within the surface 90 cm. The experimental sites were classified within one of five drainage classes by CRIFF personnel at the time of establishment (Pritchett and Smith, 1972).

All samples were air-dried after collection. Samples for soil

analysis were passed through a 2-mm sieve. Where analysis procedures required it, subsamples were ground to pass a 0.2-mm sieve. Bulk samples used in greenhouse trials were screened through a 6-mm sieve while subsamples for soil analysis were passed through a 2-mm sieve. Foliage Sampling

Foliage samples were collected in December-January of 1971/72 from the control plot of each of the three replicates at each site. Each sample consisted of a composite of needles collected from a minimum of five trees within the net plot. The trees were selected to represent the range in tree size and vigor found in the plot. Needles were taken from the previous spring flush on the uppermost whorl bearing secondary branches. The foliage samples were stored on ice while in transit to the laboratory. All samples were dried at 70C and ground to pass a I-mm sieve in a stainless steel Wiley mill. Growth and Response Parameters

Heights of all living trees were determined by CRIFF personnel at the end of the first, second, third and fifth growing seasons. The measurements taken after the second growing season were not used in this





43


study. Responses recorded after three growing seasons were reported by Pritchett and Smith (1972).

This study was concerned with evaluating the ability of soil testing procedures to predict (a) tree height growth in the absence of P fertilizer, (b) height response to P fertilizer, and (c) the amount of fertilizer required to obtain the optimum response. Since N was a limiting factor on many of the sites and N x P interactions were common (Pritchett and Smith, 1972), procedures for computing growth and Presponse parameters were designed to adjust for the N effect. This was deemed necessary as meaningful relationships between growth and/or response and a particular nutrient level can only be obtained if the nutrient in question is the only one limiting growth (Williams, 1962; Richards and Bevege, 1972 a).

The index of growth in the absence of P fertilizer was taken as the tallest mean height of the Po Ni treatments, where i = o, i, or 2. These values were obtained for each replicate at each site for growth periods of 1, 3, and 5 years.

Relative tree height was used as the index of response to P fertilizer. This was calculated from:

Relative height = Mean height of Po Ni treatment x 100.
Maximum height from P addition

Maximum height from P addition, adjusting for the effect of N, was predicted by first fitting response curves to the height data using a second degree polynomial equation,

Y = a + bX + cX2

where Y = tallest mean height in the N treatments, X = P-application rate, and a, b, and c are constants. The predicted maximum height was then obtained by first differentiating the quadratic equation to give





44


dY
- b + 2cX,
dX
dY
then solving X for dY = O, to give

b + 2cX = 0, and by rearranging

X = -b
2c

We obtained the value of Y corresponding to this value of X from the original quadratic equation by substitution

Y = a + b + c , and rearranging b2
Y = a - c , which is maximum tree height.

Relati-ve height was computed for each replicate at each site for growth periods of 1, 3, and 5 years. In some cases it was not possible to obtain a predicted maximum height in this manner, because the response was either linear or increased exponentially with increasing rates of P. In these cases, the maximum height was taken as the actual mean height in the tallest PiNi treatment. Where no increase in height occurred following P application, relative heights were recorded as 100%.

The amount of P fertilizer required to achieve maximum height was obtained by differentiating the quadratic equation, setting the derivative equal to zero and solving for X (this = -b/2c). In addition, the amounts of fertilizer required to achieve 90 and 95% of maximum height were also computed. These values were obtained by substituting 90 and 95% of maximum tree height values in the quadratic equation and solving for X. In cases where actual rather than predicted maximum tree heights were used, P fertilizer required for maximum height was taken as the actual P rate of the treatment that produced the maximum height.





45

Greenhouse Trial 1

This greenhouse trial was established in order to obtain growth and P response information for use in the preliminary screening of the effectiveness of soil-test methods. This screening was necessary as it was not practical to evaluate a large number of soil P-test using the soils from all 72 sites. Only the most successful methods in the preliminary screening were used in the final evaluation employing soils and growth information from all sites.


Establishment

The greenhouse trial with the 10 bulk soil samples was established in March 1972. Eight kilograms of air-dried, sieved soil were placed in 21-cm diameter, 7.6-liter glazed pots. Single drainage holes near the base of each pot were filled with porous fibre glass plugs. Sufficient pots of each soil were prepared to give three replicates of four P rates. One additional pot of each soil was prepared for use in determining root: shoot ratios at the first harvest, as explained later. The P treatments were applied to the surface of the potted soils as CSP to emulate field application--on a surface-area basis at 0, 56.3, 112.5, and 225 kg P/ha. The first two and last application rates were identical to the P application rates, per unit surface area, as used in the field trials. To insure P was the only nutrient limiting to growth, all pots received a basal application of KCI (225 kg K/ha), micronutrient frit (140.6 kg FTE 503/ha), and NH4NO3 (225 kg N/ha). Both the KCI and FTE 503 were mixed with the soil at potting, while the NH4NO3 was added as a dilute aqueous solution in three equal applications, 1, 5,and 10 months after planting of slash pine seeds. Phosphorus treatments were assigned at random to the 10 extra pots. All P treatments were applied I month after planting of seeds.





46


Thirty slash pine seeds, previously soaked in 0.1% citric acid for 24 hours to facilitate germination, were planted in each pot. Seedlings were thinned by stages to eight seedlings per pot 2 months after planting. Soil moisture in the pots was maintained between 50 and 100% of field capacity by watering to predetermined weight with distilled water; no leaching losses were observed. Heating and cooling equipment in the greenhouse maintained the temperature between 24C and 35C. Day length was not controlled during the first 9 months, but following the first harvest in December of 1972, day length was increased to 15 hours using lowintensity incandescent light. The position of the pots on the greenhouse benches was altered periodically to minimize micro-environment effects. Harvesting

In December of 1972, seedling heights from root collar to terminal bud were measured to the nearest millimeter. Four seedlings per pot, with an average height approximating that of all eight seedlings, were harvested removing the tops at root-collar level. The roots of the harvested seedlings were left undisturbed in the soil. Root: shoot relationships at the first harvest were established by harvesting the entire plants from the extra pot of each soil included in the trial.

Root harvests in these extra pots were made by washing the soil

and root mass on a 6-mm sieve followed by a rinsing of the recovered roots in distilled water. A final harvest of the remaining four seedlings per pot was made in October, 1973, following height measurements. Tops and roots were harvested separately. Old roots left from the first harvest were separated and discarded during sieving. All tissues samples were dried at 70C, weighed, and ground to pass a 1-mm sieve using a stainless steel Wiley mill.





47


Growth and Response Parameters

In computing the growth and response parameters for the greenhouse trial, no adjustment for the N effect was required, since all treatments received a uniform N application. Mean height in pots not receiving P was taken as index of growth in the absence of P fertilizer. Relative height and P-fertilizer requirements were computed in a similar manner to those for the field trials using a response curve fitted with a quadratic equation. Growth and response parameters were computed from the data obtained at each harvest. In order to compare the growth and response parameters from the greenhouse study with those of the field trials growing on the same soils, the field parameters for the 10 trials were computed using PoN2, PIN2,and P2N2 treatments only. The N2 treatment in the field was the same rate, per unit surface area, as that used in the greenhouse trial. In addition, the parameters were computed using all three replicates at each of the 10 sites, rather than just the replicate from which the bulk soil sample was collected. However, for 3 of the 10 sites only the replicate from which the sample was collected was used, because soil properties and response information indicated substantial variation between replicates on these three sites.

Other parameters determined were the concentration of P in the

seedling tops, and total uptake of P (mg/pot). Since only the tops were harvested at the end of the first growth season, total P uptake at this stage was determined using the relationship between quantity of P in tops and total P in tops and roots computed from the 10 extra pots.





48


Greenhouse Trial 2

This trial was established for the purpose of examining the ability of slash pine seedlings to utilize different P compounds. Uptake of P was determined for 6-month-old slash pine seedlings growing in two soils to which the P compounds were added. This P uptake was related to the solubility of the P compounds, alone or after mixing with the above two soils, in chemical extracts used as soil-P tests. P Compounds

The eight P compounds used in the trial, and their chemical composition, are shown in Table 3. The composition of all compounds was checked by determining Ca, Fe, Al and P in solution following dissolution of the compounds in hot 6N HCl. Procedures used in the chemical analysis are outlined in a later section. The compounds were also checked by Xray analysis using a General Electric XRU-7 instrument with Ni-filtered CuKa radiation. The principal peaks of the six crystalline compounds, shown in Table 3, confirmed their identification according to published standards (Lehr et al., 1967). The colloidal Al and Fe phosphates were prepared by the procedures outlined by Deming and Cate (1963) and Cate, Huffman, and Deming (1959), respectively. The X-ray analysis confirmed that these prepared compounds were amorphous.

Monocalcium phosphate is the principal P form in most phosphatic fertilizers. Dicalcium phosphate, colloidal Al and Fe phosphates, K taranakite, and strengite were used in this study because they have all been identified as soil-P fertilizer reaction products (Lindsay, Frazier, and Stephenson, 1962). Fluorapatite is a common primary phosphate mineral in some soils and may also be formed as a soil-P fertilizer reaction prcduct following reversion of less basic Ca phosphates in the






49

presence of fluoride (Olsen and Flowerday, 1971). Wavellite has not been identified as a soil-P fertilizer reaction product, but is naturally occurring mineral which was included as a representative of insoluble stable Al phosphates.


Establishment

Six-month-old slash pine seedlings were used in this greenhouse

trial. The seedlings were grown from half-sibling seeds in the greenhouse. Ten seeds were sown in each of 35 closed plastic pots containing 1,750 g of thoroughly mixed Ona fine sand (0-20 cm) collected from CRIFF site A23, known to be responsive to both N and P (Pritchett and Smith, 1972). Fertilizer was not added to these pots. Soil moisture was maintained between 50 and 100% of field capacity by watering to predetermined weight with distilled water. Day length was maintained at 14 hours using lowintensity incandescent light. Seedlings were thinned to provide five uniform seedlings in each pot at 6 months. At this stage, the seedlings were removed from the pots and the roots cleansed of the sandy soil by gentle agitation under water. The roots were then rinsed in distilled water prior to transplanting immediately into the potted soils used in the experiment. The roots of all seedlings were observed to be heavily infected with mycorrhizal fungi.

The experimental design used was a randomized block design, with two blocks each with nine treatments. The blocks consisted of two soils, selected from the 10 bulk samples to provide soils of widely different P-retention capacity and pH. Properties of the two soils, an Immokalee fine sand from CRIFF site A16 and a McLaurin fine sandy loam from CRIFF site A28, are shown in Table 4. Treatments consisted of a control, to which amendments were not added, and eight P treatments, using the P





50














Table 3. Properties of phosphorus sources used in greenhouse trial 2



Composition Principal P Ca Al Fe X-ray peaks

---------- () ------Monocalcium phosphate* 24.6 15.9 - - 11.74, 3.88, 3.69 Dicalcium phosphate* 22.8 29.5 - - 3.35, 2.96, 2.72 Fluorapatitet 18.0 39.7 - - 3.44, 2.80, 2.70 Colloidal aluminum phosphate 17.9 - 14.4 Potassium taranakitet 19.0 - 10.1 - 15.84, 7.91, 3.81 Wavellite� 10.8 - 16.0 - 8.71, 8.42, 3.22 Colloidal ferric phosphate 13.9 - - 27.1 Strengitet 17.0 - - 35.5 5.49, 4.37, 3.01


* Analytical grade reagents.
t Provided by courtesyof the Tennessee Valley Authority. � Provided by courtesy of Dr. F.N. Blanchard, Geology Dept., University
of Florida.





51

compounds shown in Table 3. Each treatment was replicated three times for the Immokalee soil but only twice for the McLaurin soil because it was in short supply.

Closed plastic pots containing 1,500 g of soil were used. The P

compounds were applied at a rate designed to raise the total P content of the soil by 100 ppm. Required quantities of each compound, all previously ground to pass a 0.0105-mm sieve, were thoroughly mixed, throughout the soil for each pot. Three carefully graded seedlings were transplanted into each pot. Moisture conditions and day length were maintained as outlined above.


Harvesting

Eight months after transplanting, the heights of all seedlings were recorded. Following height determinations, the seedlings were harvested using the procedure outlined for greenhouse trial 1. The tops and roots from each pot were dried at 70C, weighed, and ground to pass a 1-mm sieve. Phosphorus uptake per pot was determined as the combined product of P concentration in the tops and roots and their corresponding dry weights.


Extraction of P Compounds

Each of the eight P compounds (Table 3) was added at a rate of

100 ppm P to duplicate 100-g samples of each of the two soils. All samples, including control soil samples, were thoroughly mixed by passing through a

0.5-mm sieve several times. One of the duplicate sets was incubated at room temperature in the dark for 2 months, with moisture being maintained at field capacity by the addition of distilled water. Following this incubation period, the samples were air-dried and passed through a 0.5-mm sieve.





52






Table 4. Physical and chemical properties of soils used in greenhouse
trial 2


Soil type
Property Immokalee fs McLaurin fsl pH (1:2, H20) 4.2 4.9 pH (1 N KCI) 3.0 3.8 Clay, % 1.9 9.8 Silt, % 5.6 28.7 Organic matter, % 3.3 2.2 CEC, meg/lOOg 5.5 4.7 NH40Ac (pH 4.8) extractable

Ca, ppm 164 104 P, ppm 6.3 0.4

0.3 M NH4C204 (pH 3.0) extractable

Al, ppm 25 1025 Fe, ppm 60 545 P retention,* Ug P/g soil 30 750


* Equilibration with 2,500 ug P/g soil for 6 days.





53



Several soil-test methods, details of which are given later, were used to extract P from subsamples of the above incubated and nonincubated mixtures. In addition, the extractability of the P compounds in the soil-test extractants was also determined. Amounts of P compounds used for the extraction in the absence of soil were calculated to provide the same amount of P per unit of extractant as that in the corresponding soil mixture.





54


Phosphorus-Retention Study

Phosphorus-retention characteristics of a range of forest soils

were determined in the laboratory and related to various soil properties. The value of using the concentration of various elements in the foliage of slash pine to predict the P-retention capacity of forest soils was also examined. Parameters found to correlate closely with laboratory determined P retention were examined for their ability to predict leaching losses of P Fertilizers in the field. Soil and Foliage Samples

A total of 42 surface (0-20 cm) soil samples was used in the laboratory phase of this study. Twenty-four of the samples were those collected from the bedded area of the replicate 2 control plots in the NP field fertilizer trials. The 10 bulk samples were also included. The eight remaining samples were collected from other uncultivated sites in Florida and used in this study to increase the range of soil types.

Foliage samples were those collected from replicate 2 control plots and control plots from which bulk samples had been collected. Foliage samples were not available from the above eight remaining sites.


Determination of P Retention

Phosphate retention characteristics were determined by equilibrating 5-g, air-dry soil samples with 25 ml of 0.01M CaC12 containing Ca(H2P04)2.H20 at concentrations ranging from 0 to 500 ppm P (0-2500 vg P/g soil). Two drops of toluene were added to inhibit microbial activity and the samples were equilibrated for 6 days at 25C with





55



intermittent shaking. Following centrifugation, P was determined in solution. Phosphorus-adsorption isotherms were plotted for vg P adsorbed/g soil against pg P/ml in equilibrium solution. From these plots, values were interpolated to fit the linear form of the Langmuir adsorption isotherm equation (Olsen and Watanabe, 1957)

C I C
- +
x/m kb b '

where C = equilibrium P concentration, x/m = amount of P adsorbed per unit weight of soil, k = a constant related to binding energy, and b = P-adsorption maximum.

Significant deviations from the Langmuir equation have been shown to occur at equilibrium values in excess of 10-20 ppm P (Fox and Kamprath, 1970). Also the Langmuir maximum, b, depends to a certain extent on the maximum equilibration value included in the computation (Guna.ry, 1970). Therefore, interpolated values were restricted to cases where equilibration values were 10 ppm or less with 10 ppm being included as the top value for all soils. The Langmuir adsorption maximum was obtained from the reciprocal of the slope following use of linear regression analysis to obtain the Langmuir equation from the interpolated data.

In addition to the Langmuir maximum, the amount of P adsorbed

from the highest level of application (2,500 1g P/g soil) was also used as an index of the P-retention capacity of these soils.

Retention of field-applied P was determined from the 0-20 cm

samples collected from the Po, PI, and P2 plots of each replicate of the 10 field trials selected for this purpose. The samples were analyzed for total P and this value converted to kg P/ha in the surface 20 cm using the bulk density of the samples. The amount of P retained in the





56


surface 20 cms, 4 years after fertilizer application, was determined by difference in total P (kg P/ha) between control plots and plots which had received 56 or 224 kg P/ha. This P-retention value was then expressed as a percentage of the P applied.





57



Sample Analysis


Soil Characterization

Bulk density of all soil samples collected using a closed-cylinder soil auger was computed from the dry weight of the samples and their volume. The volume was calculated from the diameter of the auger, the depth of sampling and number of cores. Particle-size distribution was determined by the hydrometer method (Bouyoucus, 1951). Moisture contents at 15, 1/3, and 1/10 atmospheres were determined by use of a pressuremembrane apparatus (Richards, 1965). Loss on ignition was determined as the weight loss (%) of an air-dried sample following ignition in a furnace at 550C for 1 hour.

Soil pH was determined potentiometrically by insertion of a combination glass electrode assembly in the supernatant of a 1:2 soil-water and/or soil-N KCI suspension. Soil organic matter was determined by a modified Walkley-Black method (Allison, 1965); total N by the macroKjeldahl procedure (Bremner, 1965); and cation exchange capacity (CEC) by NH4 saturation using N NH40Ac at pH 7.0 (Chapman, 1965). Extractable cations were determined in the filtrate following extraction with

0.7N NH40Ac + 0.54N HOAc buffered at pH 4.8. Calcium and Mg in the filtrates were determined by atomic absorption (Perkin-Elmer 303) and K by flame emission (Beckman B) spectroscopy. Lanthanum chloride was added to suppress anion interference in the Ca and Mg determinations (Breland, 1966). These analyses were carried out by the Analytical Research Laboratory of the Soil Science Department, University of Florida. Soil P Analysis

Reagents used to extract soil P, and details of the extraction





58


procedures are given in Table 5. Unless mentioned to the contrary below, all procedures involved extraction of air-dry samples (<2mm) on a reciprocating shaker followed by filtration and the determination of P using an aliquot from the filtrate. The P extracted was expressed as ppm on a soil basis.

Saturation extracts were obtained by placing 10Og of air-dry soil in a Nalgene filter assembly (0.20-pi membrane). A measured quantity of distilled H20 was added to produce saturation, and the soil thoroughly stirred to remove all air bubbles. After I hour, the solution was filtered with a suction of 300 mm Hg. The volume of filtrate was recorded and an equivalent amount of distilled H20 was returned to the soil, stirred, and again filtered as before an hour later. This procedure was repeated a total of 10 times. Phosphorus was determined on each filtrate and expressed as ppm P in soil solution at 20% moisture content. Values for the first and tenth extracts were used as soil-P parameters, as was the total Pg of P extracted by all 10 extractions. In addition, a measure of the soil capacity to buffer solution P levels against depletion was obtained by dividing the difference between the first and tenth extracts into the first extract value. This parameter will be referred to as Capacity(l). Two measurements were obtained of the soil capacity to buffer against P additions; a parameter reported as significant in determining fertilizer requirements of trees (Humphreys and Pritchett, 1971) and agricultural crops (Ozanne and Shaw, 1968). These two parameters, determined from the P-adsorption isotherms, were the amount of P addition required (pg/g of soil) to increase the solution equilibrium P concentrations to 0.3 pg P/ml and 3.0 Pg P/ml. These two parameters will be referred to as Capacity(2) and Capacity(3) respectively.





59


Table 5. Phosphorus extraction methods


Soil: Extraction
Method* Extractant pH solution time


min
H20 H20 7.0 1:5 30 H20(2) H20 saturated paste ( 1st extract) H20(3) H20 saturated paste (10th extract) H20(4) H20 saturated paste (IO extracts) NaCI 0.O1M NaCI/HCIt 4.0 1:5 30 Na2SO4 0.01M Na2SO4/H2S 4 4.0 1:5 30 Na MoO 0.01M NaMo04/HCI 4.0 1:5 30 Na2 B407 0.01M Na28407/HCI 4.0 1:5 30 Na2B407(2) 0.01M Na2B407/NaOH 10.0 1:5 30

NH40Ac 0.7N NH4OAc + 0.54N HOAc 4.8 1:5 30 NH40Ac(2) IN HOAc/NH40H 3.8 1:5 30NH40Ac(3) IN HOAc/NH40H 2.8 1:5 30 HOAc 2.5% HOAc (v/v) 2.5 1:40 120 HOAc(2) " 2.5 1:40 30 HOAc(3) " 2.5 1:5 120 HOAc(4) " 2.5 1:5 30 Lactate 0.IN CH3CHOHCOONH4 + 0.4N HOAc 3.5 1:20 240 Lactate (2) " 3.5 1:20 30 Lactate(3) " 3.5 1:5 240 Lactate (4) " 3.5 1:5 30 Citrate 1% Citric acid (w/v) 2.2 1:10 24(hr) Citrate(2) " 2.2 1:5 30





60


Table 5. Continued


Soil: Extraction
Method* Extractant pH solution time min

Oxalate 0.2M (NH4)2C204 + O.1M H2C204 3.0 1:50 60

Truog 0.002N H2SO4/(NH4)2SO4 3.0 1:100. 30 Truog(2) " 3.0 1:20 30 Truog(3) " 3.0 1:5 30 Truog(4) 0.02N H2S04/(NH4)2S04 2.1 1:100 30 Truog(5) " 2.1 1:5 30 Truog(6) " 2.1 1:5 30 Truog(7) 0.2N H2SO4/(NH4)2S04 1.1 1:100 30 Truog(8) " 1.1 1:20 30 Truog(9) 1.1 1:5 30 Truog(lO0) 0.02N H2SO4/Na2MoO4 2.5 1:100 30 Truog(ll) " 2.5 1:20 30 Truog(12) " 2.5 1:5 30 H2SO4 0.002N H2S04 2.7 1:100 30 H2S04(2) " 2.7 1:20 30 H2S04(3) " 2.7 1:15 30 H2S04(4) 0.02N H2S04 1.7 1:100 30 H2S04(5) 1.7 1:20 30 H2Sq4 (6) 1" .7 1:5 30 H2SO (7) 0.2N H2SO4 0.8 1:100 30 H2S04(8) " 0.8 1:20 30 H2S0o4(9) " 0.8 1:5 30




61


Table 5. Continued


Soil: Extraction
Method* Extractant pH solution solution time

min

HCl-H2SO4 0.5N HCI + 0.025N H2S04 1.3 1:4 5 Olsen(2) 0.5M NaHCO 3/NH4 OH 8.5 1:5 30 Olsen " 8.5 1:20 30 Olsen(3) " 8.5 1:50 30 Olsen(4) " 8.5 1:5 240 Olsen(5) " 8.5 1:20 240 Olsen(6) " 8.5 1:50 240 Olsen(7) " 8.5 1:5 16(hr) Olsen(8) " 8.5 1:20 16(hr) Olsen(9) " 8.5 1:50 16(hr) Bray 2(2) 0.03N NH4F + 0.IN HCl 1.5 1.5 1 Bray 2(3) " 1.5 1.5 30 Bray 2 " 1.5 1:10 1 Bray 2(4) " 1.5 1:10 30 Bray 2(5) " 1.5 1:50 1 Bray 2(6) " 1.5 1:50 30 Bray 1(2) 0.03N NH4F + 0.025N HCI 2.5 1:5 1 Bray 1(3) " 2.5 1:5 30 Bray 1 " 2.5 1:10 1 Bray 1(4) " 2.5 1:10 30 Bray 1(5) " 2.5 1:50 1





62


Table 5. Continued


Soil: Extraction
Method* Extractant pH solution time min

Bray 1(6) 0.03N NH F + 0.025N HCI 2.5 1:50 30 Bray 3 0.03N NH F + 0.0125N HCI 4.1 1:10 1 Bray 3(2) It 4.1 1:50 1 Bray 4 0.03N NH4F + 0.005N HCI 4.6 1:10 1 Bray 4(2) " 4.6 1:50 1 Bray 5 . IN NH F + 0.1N HCI 1.8 1:10 1 Bray 5(2) " 1.8 1:50 1 Bray 6- 0.1N NH4F + 0.025N HCI 2.9 1:10 1 Bray 6(2) " 2.9 1:50 i Bray 7 0.01N NH4F + 0.1N HCI 1.3 1:10 1 Bray 7(2) 1.3 1:50 1 Bray 8 0.01N NH4F + 0.025N HCI 2.4 1:10 Bray 8(2) " 2.4 1:50 1 HC1 0.1N HCI 1.1 1:5 30 HCI(2) " 1.1 1:10 1 HCl(3) " 1.1 1:50 1 HCl(4) 0.025N HCI 1.7 1:10 1 HC1(5) " 1.7 1:50 1 HCl(6) 0.0125N HCI 2.0 1:10 1





63


Table 5. Continued


Soil: Extraction
Method* Extractant pH solution solution ti me

min

HCI(7) 0.025N HCI 2.0 1:50 1 HCI(8) 0.005N HCI 2.3 1:10 HCl(9) " 2.3 1:50 1



Resin Amberlite IRA-400 resin (2g) 2g soil + 25m H20 1 Resin(2) " 2g soil + 25ml H20 60 Resin(3) " 2g soil + 25ml H20 4(hr) Resin(4) " 2g soil + 25ml H20 16(hr) Resin(5) " 5g soil + 3ml H20 6(days)


* In the text and tables the methods are referred to by these abbreviations.

t Reagents following a slash mark were used to adjust the pH of the
reagents preceding the slash to that shown in the pH column.








Resin-extractable P was determined using a strongly basic,

quaternary ammonium, anion-exchange resin RN(CH3)3+ Cl. The soil was ground to pass a 0.25mm-sieve and all resin exceeded this sieve size. Following the soil extraction, the soil and resin were separated by washing the soil-resin mixture with distilled water over the 0.25mm sieve. The resin was quantitatively transferred to a filtering apparatus and leached successively with five 5-ml portions of 2N NaOH followed by five 5-ml portions of 2N HCI. This procedure was found in preliminary work to recover greater than 90% of the P sorbed by the resin. Phosphorus was determined in this leachate.

-Soil P was fractionated by a modified Chang and Jackson procedure developed by Fife (Ballard, 1970 b). Fractions determined included soluble-P, Al-P, Fe-P, Ca-P and organic P. Total P was extracted by Na2CO3 fusion (Jackson, 1958). Dispersed material present in alkaline extracts was removed by addition of P-free activated carbon at the filtering stage.

Phosphorus in all solutions was determined colorimetrically by the Murphy and Riley (1962) technique using ascorbic acid and molybdatesulphuric acid as modified by Watanabe and Olsen (1965). All measurements were made using a Unicam SP 600 spectrophotometer with a wavelength setting of 880 mp. Aliquots taken for developments were adjusted to pH 5 with H2S04, using para-nitro phenol indicator, prior to the addition of the developing reagent.

Extractants containing NH40Ac and HOAc required adjustment of the aliquote to pH 2.5 with H2SO4, using 2,4 dinitrophenol indicator, in order to obtain color development. Fluoride interference in the extracts containing NH4F was eliminated by addition of saturated boric acid (Kurtz, 1942). Oxalate and citrate were found to interfere with color





65

development; an aliquot was taken down to dryness on a hot plate and ignited in a furnace at 500C for 30 minutes and the residue taken up in a known amount of O.lN NCI and an aliquot taken from these solutions for P determination. Concentrations of other potentially interfering ions were all below the interference level, as indicated by John (1970) or as established in preliminary work.


Soil Al and Fe Analysis

Soil Al and Fe were extracted by several extractants. These 'included CDB using the procedure of Mehra and Jackson (1960), 0.3M (NH4)2C204 (Saunders, 1965), 0.IM Na4P207 (McKeague, 1967), and 0.05M EDTA (Viro, 1955). Exchangeable Al was also extracted by leaching with IN KCI (Yuan, 1965). Aluminum and Fe were also determined in the extracts of several of the P-extraction methods outlined in Table 3. These methods included NH40Ac, HCI-H2S04, Bray 2, Bray i, Bray 1(3), and HCI.

Aluminum and Fe in all extracts except KCI were determined by atomic adsorption at the Soil Science Analytical Research Laboratory, Universtiy of Florida (Yuan and Breland, 1969). Pyrophosphate extracts were digested in HNO3-H2S04 and oxalate extracts ignited at 500C and taken up in 0.IN HCI before atomic absorption analysis. Aluminum in the KCI extracts was determined colorimetrically by the aluminon procedure (Yuan and Fiskell, 1959).


Plant Tissue Analysis

One-gram samples of plant material were ashed at 480C for 5 hours. The ash was dissolved in 6N HCI, digested on a hot plate for 15 minutes and quantitatively transferred with distilled H20 to a 50-ml volumetric flask. Suit3-ble aliquots for elemental analysis were obtained from this





66

solution. Phosphorus was determined using the above soil P technique. Calcium, Mg, Al and Fe were determined by atomic absorption and K by flame emission as described above. Tissue N was determined by the macroKjeldahl technique (Bremner, 1965).





67


Statistical Analysis

Statistical analysis of data was performed using either a

Hewlett-Packard model 9100B desk top computer or the computer facilities at the University of Florida Computer Center. Library Stat-Pac programs were used with the Hewlett-Packard. All statistical analyses done at the Computer Center were performed using programs available in the Statistical Analysis System (SAS) package (Barr and Goodnight, 1972). Specific programs used are.given in the following sections where appropriate.















RESULTS AND DISCUSSION


Preliminary Screening of Soil-Test Methods

The classification and selected properties of the 19 soils used in preliminary screening of soil-test methods are shown in Table 6. They are typical of many forest soils in the lower Coastal Plain. These are acid, sandy, and relatively low in organic matter. Levels of extractable P vary considerably among the 10 soils.

Heights, relative heights, P-fertilizer requirements and tissue-P parameters of slash pine grown on these 10 soils in both greenhouse and field trials are summarized in Tables 7 and 8, respectively. Data in these two tables were summarized or computed from the original data of the greenhouse and field trials presented in Appendix Tables 46 and 47, respectively. Phosphorus uptake values at the first year's harvest in the greenhouse trial were computed using the regression equation Y = 1.604X - 4.951 (r = 0.999)

where Y = mg P in tops and roots per pot, and X = mg P in tops per pot. This equation was derived using plant uptake data obtained from extra pots which were subject to a complete harvest after the first year's growth (Appendix Table 48).

The amounts of P extracted from the 10 soils by all soil-test methods are given in Appendix Table 49. Also included in this table are the amounts of various P fractions and the buffering capacities of these soils.

Relationships between extractable soil-P values and plant growth

and response parameters are usually curvilinear, according to Grigg (1965).


68









Table 6. Classification and selected properties of 10 soils used in greenhouse study 1


Soil pH Silt + Organic Ext. Al Extractable P
Type Order (H20) clay CEC matter NH4OAc NH4OAc H20 Bray 1(3) % me/lOOg % ------------ppm--------------Bladen scl* Ultisol 4.8 43.5 8.67 3.27 222 0.7 0.5 1.0 Blanton fs Ultisol 5.2 5.5 2.17 1.24 40 4.3 1.7 48.9 Plummer fs Ultisol 5.0 8.5 3.94 2.23 105 13.8 3.0 185.5 Marlboro fsl Ultisol 5.0 38.5 4.70 2.57 149 0.4 0.2 1.6 McLaurin fsl Ultisol 5.3 31.0 4.70 2.20 91 0.3 0.3 1.5 Immokalee fs Spodosol 4.3 7.5 5.46 '3.27 8 6.3 5.5 6.6 Leon fs Spodosol 4.1 6.5 9.28 4.52 11 7.8 6.4 7.9 Ona fs Spodosol 4.2 10.3 4.90 3.76 55 2.1 2.4 4.7 Kershaw fs Entisol 5.2 4.0 2.09 1.31 35 2.6 0.7 25.9 Lakeland fs Entisol 5.4 14.8 4.90 2.57 66 5.9 2.2 130.0


* scl = silty clay loam; fs = fine sand; and fsl = fine sandy loam textures.
ar'








Table 7. Height, relative height, P-fertilizer requirements, and P concentration and uptake
of slash pine seedlings after 1 and 2 years' growth on 10 soils in the greenhouse


Soil Height Rel. height Fert. reqm.t P in tops P uptake� type 1* 2 1 2 1 2 1 2 1 2

-----cm----- -----%------ --kg P/ha--- ------%------- --mg P/pot--Bladen scl 10.6 17.1 77.3 45.8 56.6 92.5 0.060 0.044 10.2 19.5 Blanton fs 12.3 29.1 89.3 91.7 13.8 0.0 0.126 0.081 25.4 62.1 Plummer fs 15.3 35.8 98.3 90.3 0.0 0.0 0.132 0.089 45.7 105.6 Marlboro fsl 12.0 19.4 65.0 52.7 70.8 95.6 0.058 0.038 8.1 15.2 McLaurin fsl 12.3 23.8 71.3 64.6 81.4 93.3 0.055 0.047 8.8 22.2 Immokalee fs 14.8 29.4 100.0 98.4 0.0 0.0 0.096 0.039 51.6 35.5 Leon fs 13.0 26.2 100.0 98.1 0.0 0.0 0.108 0.068 44.3 40.7 Ona fs 15.2 34.1 87.2 84.2 4.8 35.8 0.091 0.044 37.6 48.7 Kershaw fs 10.0 26.7 79.9 93.2 48.5 0.0 0.107 0.075 10.6 38.1 Lakeland fs 15.2 32.9 94.5 87.4 0.0 15.7 0.123 0.082 53.6 79.8


* Age in years at harvesting or measurement.
t Fertilizer required to achieve 90% of maximum height growth. � First year uptake based on eight seedlings/pot and 2-year uptake based on four seedlings/pot over 2 years.








Table 8. Height, relative height and P-fertilizer requirements of slash pine after 1, 3, and 5
years' growth, and foliar P concentration after 4 years' growth in the field on 10
soils


Soil Height Rel. height Fert. regm.t Foliar P type 1* 3 5 1 3 5 1 3 5 4


--------cm------- ---------- % ---------- ------kg P/ha--------- %

Bladen scl 29.5 90 227 62.2 69.9 79.5 71.2 61.0 42.4 0.067 Blanton fs 55.4 243 451 94.2 98.1 o100. 0.0 0.0 0.0 0.115 Plummer fs 44.4 216 433 100.0 100.o I00.0 0.0 0.0 0.0 0.109 Marlboro fsl 23.5. 136 350 65.1 80.0 89.9 114.0 50.1 1.5 0.076 McLaurin fsl 31.6 174 400 69.6 66.6 81.8 89.0 75.0 42.9 0.083 Immokalee fs 50.3 209 436 100.0 82.5 86.0 0.0 47.6 25.6 0.090 Leon fs 88.1 291 498 99.4 90.1 90.0 0.0 0.0 0.0 0.092 Ona fs 42.0 156 342 68.4 76.6 81.1 94.3 119.5 115.6 0.081 Kershaw fs 26.4 135 293 89.2 100.0 100.0 0.0 0.0 0.0 0.103 Lakeland fs 27.5 129 276 96.5 100.0 100.0 0.0 0.0 0.0 0.101


* Age in years at measurement.
t Fertilizer required to achieve 90% of maximum height growth.




72



This was true for data from this study, as statistical models which allow for curvilinearity (models 2, 3, and 4 in Table 9) provided better fits between soil-test values and tree parameters than an untransformed linear model (model 1 in Table 9). For the selected data used in this comparison, most models using a logarithmic or an arctangent transformation of the independent variable (soil-test value) and a quadratic model provided better fits, as indicated by the R2 value, than the untransformed linear model. There was little difference between the transformed models and for convenience the logarithmic model was used for all computations in the preliminary screening. The square of the multiple correlation coefficient

(R2) was used as the index of success of soil-test methods as predictors of the tree parameters.


Relationships Between Soil-Test Values and Relative Height

Relative heights, the heights of unfertilized trees expressed as a percent of heights of fertilized trees, after 1 year's growth in the greenhouse (Table 7) were almost identical to those in the field after I year of growth (Table 8). This suggested a similarity in the modes of nutrition of young seedlings in both the greenhouse and field. Responses (relative heights) to added P, after 5 years of growth in the field, were similar to those obtained after 3 years, but responses obtained at both periods tended to differ from those obtained during shorter growth periods in either field or greenhouse trials. This is in agreement with the findings of Mead and Pritchett (1971) who reported a poor relationship existed between response in the greenhouse and that obtained after several years in the field.

Response trends developed with time for the various soils can be categorized fairly well in accordance with the relationship between P








Table 9. Comparison of the goodness of fit of four statistical models, as indicated by the square
of the multiple correlation coefficient (R2), relating selected tree parameters (dependent
variable) and soil-test values (independent variable)


Soil-test Greenhouse (1 year) Field (3 years) Foliar P
method Modelt Height Rel.height Fert.reqm. Height Rel. height Fert.reqm. (4 years)

---------------------------------------R2----------------------------------------H20 1 0.288 0.741** 0.601** 0.545* 0.067 0.059 0.056
2 0.411* 0.951** 0.897** 0.421* 0.217 0.087 0.228 3 0.489* 0.947** 0.951** 0.410* 0.257 0.086 0.265 4 0.676** 0.935** 0.949** 0.548* 0.285 0.061 0.338

Truog 1 0.358 0.355 0.339 0.036 0.421* 0.286 0.346
2 0.342 0.658** 0.686** 0.113 0.658** 0.413* 0.496* 3 0.203 0.702** 0.803** 0.132 0.645** 0.350 0.424* 4 0.386 0.626* 0.653** 0.096 0.616* 0.433 0.461*

HCl-H2S04 1 0.282 0.280 0.265 0.040 0.425* 0.287 0.392
2 0.262 0.579* 0.607** 0.151 0.783** 0.494* 0.706** 3 0.193 0.689** 0.764** 0.218 0.734** 0.386 0.642**
4 0.292 0.413 0.450* 0.053 0.616* 0.510* 0.615*

Bray 1 1 0.251 0.205 0.211 0.007 0.434* 0.289 0.386
2 0.094 0.241 0.279 0.022 0.785** 0.551* 0.704** 3 0.034 0.355 0.419* 0.064 0.908** 0.640** 0.712**
4 0.282 0.207 0.228 0.014 0.642** 0.446* 0.575*

* Significant at the 5% level.
** Significant at the 1% level.
Model 1, Y = bX + c; Model 2, Y = b logX + c; Model 3, Y - b arctan X + c; Model 4, Y = aX + bX2 + c.









extracted by the H20 and Bray 1(3) methods (Table 6), which can be taken as indicative of the intensity and quantity factors of soil P supply, respectively (Williams and Knight, 1963). For soils containing high extractable P by H20 and moderately low by Bray 1(3) methods--!mmokalee and Leon--no growth response occurred in the first year, but a response was evident after 3 and 5 years. Those soils with a moderate to low extractable P by H20 but high by Bray 1(3) methods--Blanton, Kershaw, and Lakeland--showed the opposite trend with some response in the early growth period which disappeared with increasing time. Response occurred in the early growth period and persisted with time on soils with low extractable P by both H20 and Bray 1(3) methods--Bladen, Marlboro, and McLaurin-while on the one soil with moderate to high extractable P by H20 and Bray 1(3) methods--Plummer--no response occurred at all during the first 5 years. Apparently the amount of water-soluble P (intensity factor) is of importance during the early establishment and growth of the pine seedling while the amount of solid-phase P in equilibrium with soil-solution P (quantity factor) becomes more important to tree growth with increasing time of persistence on the site.

The response trends discussed above are well illustrated by the effectiveness of P extracted by various soil-test methods at predicting relative heights in the greenhouse (Appendix Table 50) and the field (Appendix Table 51). The effectiveness (R2) of soil-test methods as predictors of relative height at any age appeared to be closely related to the amount of P extracted, but independent of the type of extractant used where similar quantities of P were extracted (Fig. 1). In view of this, the results will be illustrated and discussed principally using those obtained with some of the more commonly used soil-test methods




75


1 YEAR


t 0.68 *

0.4'

LL




0
Oo :'' W YEARS . < ..





0 0.2
O ' L 1 YEARS

- O.8 * * . * * .: *: .*.*.

S0 0 * '1 0.4 0 *; .*
. *..



O 10 .20SO 40 50 60
MEAN EXTRACTABLE P, ppm


Fig. 1. Relationship between R 2 values for regressions of relative
height at age 1, 3, and 5 years on soil-test values, and
mean amounts of P extracted by soil-test methods.
MEA EXRCAL PE opp

Fig . Reaiosi bewe 2vle o ersin frltv hegta g ,3 n easo olts aus n
mea amut fPetatd ysi-etmtos




76



which vary in their capacities to extract P from soils. The similar effectiveness of different extractants which remove equivalent quantities of P suggests that they extract the same form(s) of P from the soil. Aluminum-P is probably the form most commonly involved since this fraction dominates the inorganic P component of the 10 soils (Appendix Table 49) and most soil-test extractants are capable of solubilizing Al-P (Thomas and Peaslee, 1973).

Soil-test methods involving use of weak extractants, such as H201 and NH OAc which extract only small amounts of P from most soils, provided the best index of tree response to P fertilizers over short growth periods in both the greenhouse and field trials (Table 10). Waterextractable P was slightly superior to that extracted by NH4OAc at predicting response after I year's growth of seedlings in the greenhouse, while NH4O0Ac was a better predictor of response than H20 after 1 year's growth in the field. Other soil-test methods which provided a good index of response over short growth periods included those involving use of neutral salts, short term extractions with anion exchange resin, and very dilute acid solutions at a narrow soil:solution ratio (Appendix Tables 49 and 50). The effectiveness of other soil-test methods as indicators of short term response tended to be inversely proportional to the amount of P extracted (Fig. I and Table 10).

Responses after 3 and 5 years' growth in the field were most

closely related to P extracted by soil-test methods which removed larger amounts of P from the soil than H20 and NH4O0Ac. As the growth period increased from I to 3 or 5 years, R2 values decreased for the soil-test methods involving use of weak extractants. The R2 values for response after 3 and 5 years' growth increased with increasing amounts of P








Table 10. Relationships between selected soil-test values and relative
height growth of slash pine In field and greenhouse experiments on 10 soils


Soil-test Mean Relative height method or extractP form able P Greenhouse Field i 2 1 3 5 ppm -------------------------R2
H20 2.3 0.951** 0.715** 0.614** 0.217 0.055 NH 4Ac 4.4 0.863** 0.729** 0.817** 0.589** 0.348 Truog 8.9 0.658** 0.440* 0.658** 0.657** 0.481* Olsen 12.3 0.533* 0.515* 0.702** 0.787** 0.666** HCl-H2S04 15.2 0.579* 0.538* 0.722** 0.782** 0.639** H2SO4(9) 32.5 0.349 0.331 0.540* 0.765** 0.716** Bray 1(3) 41.3 0.187 0.478* 0.623** 0.845** 0.778** NH4F(pH8.5) 57.0 0.296 0.060 0.465* 0.718** 0.699** Organic-P 42.6 0.050 0.010 0.011 0.040 0.056 Total-P 136.1 0.040 0.001 0.065 0.229 0.306 t Tree age at time of measurement (years).
* Significant at the 5% level, using the model Y = b logX +c, where
Y = relative height and X = soil-test value.
** Significant at the 1% level.




78


extracted by the soil-test methods up to a plateau (Fig. I and Table 10). The plateau, where R2 values were almost independent of quantities of P extracted, was reached at lower levels of mean extractable P for response after 3 years (ca. 12 ppm) than after 5 years (ca. 20 ppm). As mean extractable P values increased above ca. 50 ppm, the R2 values began to decline indicating the soil-test methods involved were beginning to extract forms of P unavailable for tree use at this stage of growth. Soil-test methods extracted-amounts of P corresponding to the above-mentioned plateau included those involving use of strong acid and alkaline solutions (HCI, H2S04 and NaHC03) and strong complexing agents (lactate, NH4F).

Amounts of Fe-P and Ca-P in the 10 soils were very low and were not related to fertilizer response at any stage of tree growth (Appendix Tables 50 and 51). Aluminum-P, which dominated the inorganic P fractions of all 10 soils, was significantly related to tree response when they were 3 and 5 years old. However, the mean amount present in the soil, 54.3ppm P, suggests that not all this fraction was available for tree use during the first 5 years on the basis of data in Fig. I. This is perhaps not surprising since different Al-P compounds found in soils vary considerably in their availability to plants (Taylor et al., 1960). Soil-test methods which discriminate between Al-P compounds of different solubility are apparently better indicators of soil-P status than methods which do not so discriminate.

Neither total nor organic-P were significantly related to response at any stage of growth (Table 10). This concurs with the shortcomings of total analyses reported previously (Pritchett, 1968; Ballard, 1970a). However, it is of interest to note that R2 values for both methods of analysis increased with increasing growth period. This result is consistent with the premise that an increasing proportion of the total P in





79



soil becomes involved in meeting the trees'P requirements as growth periods increase. However, it is unlikely that total P would ever become an effective predictor of soil-P status, even over very long growth periods (rotation length, ca. 25 years) since certain fractions of the soil's total P are occluded and extremely insoluble.


Relationships Between Soil-Test Values and P Uptake and Tissue P

Phosphorus uptake by seedlings during the first year's growth in the greenhouse was most closely correlated with P extracted by weak extractants e.g. H20 and NH 4Ac (Table 11). However, P uptake over the 2year growth period was more closely correlated with P extracted by soiltest methods which removed larger amounts of P from the soil (strong extractants). In contrast to P uptake, P concentrations in seedling tops at the end of the first year were more closely correlated with P extracted by the stronger extractants than the weaker extractants. After two years' growth in the greenhouse the advantage of the stronger extractants over the weaker ones for predicting P concentrations in the tops was more evident.

The magnitude and ranking of the R2 values for the second year P concentrations in tops and for foliar P of 4-year-old field trees were strikingly similar. Raw data in Tables 7 and 8 show that both first and second year P concentrations in tops of greenhouse grown seedlings wereclosely correlated with foliar P of field trees (r = 0.897 and 0.875 respectively), although the P concentration in seedling tops at the end of the first year in the greenhouse was quantitatively more closely related to foliar P than was the P concentration in tops at the end of the second year. Terman and Bengtson (1973) reported close agreement between their established 'critical' P concentration in tops of 8-to 12-month-old slash








Table 11. Relationships between selected soil-test values and tissue P
parameters of greenhouse and field slash pine grown on 10 soils


Soil-test Mean Greenhouse Field method or extract- P uptake %P in tops foliar P P form able P 2 1 2 4 1 2 1 2 4 ppm ------------------------R2.----------H20 2.3 0.81.0** 0.322 0.537* 0.150 0.228 NH40Ac 4.4 0.709** 0.607** 0.847** 0.454* 0.511* Truog 8.9 0.634** 0.828** 0.821** 0.599** 0.495* Olsen 12.3 0.447* 0.875** 0.931** 0.806** 0.765** HCI-H2S04. 15.2 0.516* 0.871** 0.925** 0.753** 0.706** H2SO4(9) 32.5 0.314 0.873** 0.811** 0.846** 0.698** Bray 1(3) 41.3 0.348 0.850** 0.895** 0.827** 0.817** NH4F(pH 8.5) 57.0 0.266 0.722** 0.742** 0.746** 0.638** Organic-P 42.6 0.141 0.459* 0.070 0.164 0.019 Total-P 136.1 0.066 0.600** 0.213 0.442* 0.175


tTree age at time of sampling (years).
* Significant at the 5% level, using the model Y = b logX + c, where Y =
P parameter and X = soil-test value.
**Significant at the 1% level.





81



pine and the 'critical' foliar P values reported by other workers for this species in the field. The relationships reported above could be of value in extrapolating greenhouse trial results to field conditions.


Relationships Between Soil-Test Values and Fertilizer Requirements

The association found between mean quantities of P extracted by soil-test methods and their ability to predict response to P fertilizer was also apparent in the relationships between soil-test values and Pfertilizer requirements (Table 12). Phosphorus extracted by weak extractants was most closely correlated with fertilizer requirements over short growth.periods, while P extracted by stronger extractants was most closely correlated with fertilizer requirements over longer growth periods.

Although R2 values for the relationships between soil-test values and fertilizer requirements in the greenhouse were of a similar magnitude to those for relative height, R2 values for fertilizer requirements calculated from field tests were lower than those for relative height in field tests. This was particularly pronounced for the fifth, and to a lesser extent, the third year data. This could be due to variation inthe P-retention capacity of the soils. For instance, Lewis and Harding (1963) reported lower P-fertilizer requirements for pines growing on soils of low P-retention capacity compared to those on soils of high P-retention capacity, but with similar extractable soil-P values. One would expect differences due to P retention to become more pronounced with time due to the time dependency of the reversion of relatively soluble phosphates to less soluble forms (Juo and Ellis, 1968). If variable P retention was the major factor accounting for the low R2 values shown in Table 12, a multiple regression equation including both soil-test values and a measure








Table 12. Relationships between selected soil-test values and P-fertilizer requirements of slash pine in field and greenhouse experiments on 10 soils


Soil-test Mean P-fertilizer requirementst method or extract- Greenhouse Field P form able P l 2 3 5 ppm ---------------------------R2-----------------------H20 2.3 0.897** 0.703** 0.508* 0.088 0.000 NH40Ac 4.4 0.851** 0.863** 0.745** 0.357 0.100 Truog 8.9 0.685** 0.607** 0.613** 0.414* 0.154 Olsen 12.3 0.552* 0.678** 0.694** 0.513* 0.201 HCI-H2SO4 15.2 0.606** 0.697** 0.686** 0.494* 0.192 H2S04(9) 32.5 0.377 0.488* 0.567* 0.519* 0.230 Bray 1(3) 41.3 0.436* 0.616** 0.601** 0.506* 0.217 NH4F(pH 8.5) 57.0 0.322 0.396* 0.491* 0.507* 0.247 Organic-P 42.6 0.050 0.001 0.003 0.016 0.003 Total-P 136.1 0.043 0.015 0.077 0.193 0.123

t P fertilizer required to achieve 90% of maximum growth.
6 Tree age at time of determination (years).
* Significant at the 5% level, using the model Y - b logX + c, where Y = P fertilizer
requirement and X = soil-test value.
** Significant at the 1% level.




Full Text
271
Walker, L. C., end C. T. Youngberg. Response of slash pine to nitrogen
and phosphorus fertilization. Soil Sci. Soc. Amer. Proc, 2.6:
399-401.
Watanabe, F. S., and S. R. Olsen. 1965. Test of an ascorbic acid
method for determining phosphorus in water and NaHC03 extracts
from soil. Soil Sci. Soc. Amer. Proc. 29:677-678.
Wells, C. G. 1965. Nutrient relationships between soils and needles
of loblolly pine (Pinus taeda). Soil Sci. Soc. Amer, Proc. 29:
621-624.
Wells, C. G., and D. M. Crutchfield. 1969. Foliar analysis for pre
dicting loblolly pine response to phosphorus fertilization on wet
sites. Southeast. For. Exp. Sta., USDA For. Serv. Res. Note SE-128.
4 p.
Wells, C. G., D. M. Crutchfield, N. M. Berenyi, and C. B. Davey. 1973-
Soil and foliar guidelines for phosphorus fertilization of loblolly
pine. Southeast. For. Exp. Sta., USDA For. Serv. Res. Paper SE-110-
15 P.
White, D. P., and A. L. Leaf. 1957. Forest fertilization: A biblio
graphy, with abstracts on the use of fertilizers and soil amendments
in forestry. State Univ. Coll. For. at Syracuse Univ. Tech. F'ubl .
81 .
White, D. P. 1968. Progress and needs in tree nutrition research in
the Lake States and Northeast, p. 226-233. J_n^ Forest fertiliza
tion theory and practice. Tennessee Valley Authority, Muscle
Shoals, Ala.
White, E. H., and A. L. Leaf. 1964. Soil and tissue K levels related
to tree growth: 1. HNO3 extractable soil K. Soil Sci. 98:395"402.
White, E. H., and W. L. Pritchett. 1970. Water table control and fer
tilization of a flatwood soil for pine production. Fla, Agr, Exp.
Sta. Tech. Bull. No. 743. 41 p,
Wilde, S. A. 1958. Diagnosis of nutrient deficiency by foliar and soil
analysis in silvicultural practice, p. 138-140. _hi T. D. Stevens
and R. L. Cook (ed.) First North American forest soils conference.
Michigan State Univ., E. Lansing, Mich,
Wilde, S. A., J. G. Iyer, C. H. Tanzer, W. L. Trautman, and K, G. Wat-
terston. 1964a. Growth of jack pine (Pinus banksiana. Lamb)
plantations in relation to fertility of non phreatic sandy soils.
Soil Sci. 98:162-169-
Wilde, S. A., J. G. Iyer, C. H. Tanzer, W. L. Trautman, and K. G. Wat-
terston. 1964b. Growth of red pine plantations in relation to
fertility on non phreatic sandy soils. For. Sci. 10:463-470.


249
Table 59- Continued
Field Repl i-
s i te
cate
N
P
K
Ca
Mg
A1
Fe
ppm--
A1 7
1
1.10
0.076
0.44
0.24
0.108
520
39
2
1.07
0.080
0.45
0.24
0.116
533
37
3
1.04
0.079
0.40
0.22
0.101
488
36
A18
1
0.87
0.085
0.31
0.20
0.116
437
37
2
0.91
0.087
0.40
0.24
0.120
431
24
3
0.87
0.085
0.34
0.21
0.129
400
35
A19
1
0.91
0.090
0.45
0.18
0.120
262
50
2
0.90
0.087
0.47
0.22
0.104
263
30
3
0.98
0.099
0.48
0.21
0.121
244
35
A20
1
1.56
0.060
0.31
0.15
0.080
281
66
2
1.70
0.061
0.37
0.15
0.076
236
87
3
1.94
0.061
0.39
0.15
0.076
344
86
A21
1
1.24
0.109
0.41
0.24
0.091
625
39
2
1.11
0.093
0.31
0.25
0.108
538
22
3
1.21
0.112
0.39
0.29
0.099
656
32
A22
1
0.88
0.080
0.48
0.14
0.086
344
32
2
0.93
0.082
0.49
0.12
0.089
350
34
3
1.01
0.084
0.52
0.14
0.086
350
34
A23
1
0.94
0.081
0.36
0.26
0.113
263
34
2
1.01
0.081
0.38
0.23 .
0.114
544
25
3
1.07
0.081
0.34
0.23
0.101
475
35
A24
1
1.08
0.072
0.40
0.17
0.073
469
52
2
J .00
0.080
0.41
0.21
0.086
466
38
3
1.02
0.075
0.49
0.20
0.073
500
58
A25
1
l .24
0.106
0.45
0.18
0.094
600
52
2
1.16
0.107
0.50
0.21
0.094
525
28
3
1.08
0.091
0.50
0.21
0.093
544
40
A26
1
1.13
0.087
0.26
0.17
0.083
575
46
2
1.06
0.088
0.26
0.14
0.073
525
51
3
1.07
0.082
0.29
0.17
0.089
456
45
A27
1
1.08
O.C83
0.48
0.18
0.078
675
107
2
0.97
0.101
0.60
0.19
0.090
650
80
3
1.08
0.075
0.48
0.21
0.088
581
99
A28
1
1.11
0.083
0.53
0.20
0.081
683
55
2
1.07
0.086
0.49
0.15
0.069
694
92
3
1.08
0.079
0.49
0.17
0.085
556
75


268
Pritchett, W. L. 1968. Progress in the development of techniques and
standards for soil and foliar diagnosis of phosphorus deficiency
in slash pine. p. 81-8?. Jin Forest fertilization theory and prac
tice. Tennessee Valley Authority, Muscle Shoals, Ala.
Pritchett, W. L., and H. Hanna. 1969. Fertilization 11. Results to
date, potential gains in Southeast generate cautious optimism.
Forest Ind. 96:26-28.
;* Pritchett, W. L., and W. H. Smith. 1969. Sources of nutrients and their
reactions in forest soils. Soil and Crop Sci. Soc. Fla. Proc. 29:
149-158.
Pritchett, W. L., and W. H. Smith, 1970. Fertilizing slash pine on the
sands of the Lower Coastal Plain, p. 194l. In C. I. Youngberg
and C. B. Davey (ed.) Tree growth and forest soils. Oregon State
Univ. Press, Corvallis.
v- Pritchett, W. L., and W. H. Smith. 1972. Fertilizer response in young
pine plantations. Soil Sci. Soc. Amer. Proc. 36:660-663.
Pritchett, W. L,, and W, H, Smith. 1974. Management of wet savanna
forest soils for pine production. Univ. Fla. Agrie. Expt. Sta.
Tech. Bul 1. 762. 22 p.
Qureshi, I. M., and P. B. L. Srivastava. 1966. Foliar diagnosis and
mineral nutrition of forest trees. Indian For. 92:447"460.
Ralston, C. W. 1964. Evaluation of forest site productivity. Int. Rev.
For. Res. 1:171201.
Ramulu, U. S. S., P. F. Pratt, and A. L. Page. 1967- Phosphorus fixation
by soils in relation to extractable iron oxides and mineralogical
composition. Soil Sci, Soc. Amer. Proc. 31:193"196.
# Raupach, M. 1967. Soil and fertilizer requirements for forests of Pinus
rad ¡ata. Advan. Agron. 19:307353.
Richards, B. N. 1956. The effect of phosphate on slash and loblolly
pines in Queensland. Queensland For. Serv. Res. Note 5*
Richards, B. N. 1961. Fertilizer requirements of Pinus taeda in the
coastal lowlands of subtropical Queensland. Queensland For. Dept.
Bull. 16. 24 p.
;Richards, B. N., and D. I. Bevage. 1972a. Principles and practice of
foliar analysis as a basis for crop-logging in pine plantations.
I. Basic considerations. Plant and Soil 36:109-119.
Richards, B. N., and D. I. Bevage. 1972b. Principles and practice of
foliar analysis as a basis for crop-logging in pine plantations.
II. Determination of critical phosphorus levels. Plant and Soil
37:159-169.


The apparent lack of interest shown in forest fertilization by most
North American foresters prior to I960 has been attributed to their pre
occupation with extensive rather than intensive management techniques
(Bengtson, 1972). In addition, the results of most soil-site research,
particularly that by Coile (1952), indicated that soil physical charac
teristics provided better indices of site productivity than did soil
chemical properties.
N and P Fertilization
Several general reviews on forest fertilization (White and Leaf,.
1957; -Stoeckler and Arneman, I960; Swan, 1965; Mustanoja and Leaf, 1965)
and specific reviews of progress in countries such as Sweden (Hagner,
1967; Holmen, 1967), Finland (Salonen, 1967; Paarlahti, 1967), Norway,
(Jerven, 1967), Japan (Kawana, 1969), Britain (Leyton, 1958; Binns, 1969),
Australia (Raupach, 1967; Gentle and Humphreys, 1968), Canada (Swan, 1969;
Krause, 1973), and the USA (Gessell, 1968; White, 1968; Bengtson, 1968 and
1972; Safford, 1973), have i 1 lust rated the tremendous advances in forest
fertilization research and the application of the results during the last
decade. Most recorded responses have been to N and P fertilizers although
isolated responses to K fertilizers have been reported in Japan (Kawana,
1969), Austra1ia (Hall and Purnell, 1961), Finland (Salonen, 1967), mid-
western Europe (Hagner, 1971), and the northeastern USA (Stone and Leaf,
1967).
The most spectacular response of coniferous species to N applica
tions have been recorded in Scandinavia (Hagner, 1967) and the Douglas-
fir region of the northwestern USA and southwestern Canada (Strand and
Miller, 1969). In the Scandinavian countries, operational fertilization
began about I960 (Hagner, 1967) with urea as the principal N source. In


156
Table 48. Dry weight, P concentration, and P uptake in
tops and roots of slash pine seedlings after 1
year of growth on nine soils in the greenhouse
Soi 1
type
Treat
-ment
Seedling
component
Dry
we ight
P
cone.
p
uptake
g
%
mg/pot
Blanton fs
P2
Shoot
17-8
0.192
34.09
Root
8.3
0.184
14.23
Plummer fs
Pi
Shoot
27.h
0.150
41.10
Root
11.9
0.150
17.85
Marlboro fsl
Po
Shoot
17.9
0.063
11.19
Root
6.8
0.059
3.98
McLaurin fsl
Pi
Shoot
31.7
0.090
28.63
Root
13.0
0.076
9.85
Immokalee fs
P2
Shoot
27-9
0.320
89.28
Root
13-4
0.361
48.41
Leon fs
Po
Shoot
31 .0
0.084
26.04
Root
13.9
0.089
12.30
Ona fs
Pa
Shoot
35.9
0.298
106.80
Root
19-4
0.320
62.08
Kershaw fs
Po
Shoot
7.3
0.188
13.69
Root
5.9
0.107
6.31
Lakeland fs
P2
Shoot
33.0
0.125
41.25
Root
14.4
0.115
16.30
* Total weight is for eight seedlings per pot.


125
increased importance of P from lower soil depths with increase in tree
age. Apparently, because of its correlation with topsoil P, which in
turn is correlated with P at lower depths, depth to LH provides an inte
grated measure of available soil P within the rooting zone of the older
trees, of which topsoil P is only one component.
All the soil and site properties listed in Table 23 plus P
extracted by the H^O and HCl-H^SO^ methods and the squared terms of each
of these parameters were used as independent variables in a stepwise
regression procedure designed to maximize R^ (Barr and Goodnight, 1972)
for models with relative heights as the dependent variables. The 'best
fit' models, in which the coefficients for all the included independent
variables were significant at least at the 5% level are shown in Table 26.
Although these models could be used as prediction models, they provide no
information on causal relationships. It is interesting, however, that
the majority of the independent variables included are those which were
significantly correlated with P extracted by either the HCl-I^SQ/j or H^O
methods (Table 24).The relatively low R^ values for these 'best fit'
models suggest either unmeasured parameters are contributing to variation
in relative height, or the relative heights derived from the raw plot
data are not completely true reflections of the responsiveness of the
control-plot soils because of 'within site' variability or lack of
optimum fertilizer application.
Thus far, the data have shown that various site or soil physical
parameters provide better indications of the P responsiveness of the
sites used in this investigation than extractable-P values. However,
since these parameters are unlikely to reflect causal relationships, ex
treme caution should be used in extrapolating the results to other sites.
Where relationships are known to be causal, such as between soil-P levels


173
The relationship between retention of field-applied P and A! ex
tracted by NH^OAc is shown in Fig. 16. Each point is the mean of deter
minations from three pairs of plots taken from the 3 blocks in each trial.
In cases where A] values between blocks varied considerably, blocks were
treated as separate samples. Where A1 values between paired plots within
blocks varied considerably, the block was discarded from the analysis.
Total P and NH^OAc-extractable A1 values for soils of the PQ, Pj and P2
plots in the 10 trials selected for use in this calibration are given in
Appendix Table 63. A number of factors probably caused deviations from
the expected relationship between P retention and extractable soil A1.
These included differences in uptake of applied P between trials, initial
variability in total P between control and treated plots, and differences
in the leaching caused by variable rainfall and soil porosity between
trials. The relationship between P retention and extractable A1 (Fig.
l6) appeared to be almost independent of application rates used in these
trials. For diagnostic purposes, it was assumed that excessive leaching
losses of P occur at sites with P retention below 50%. These sites can
be identified with reasonable accuracy as those with kO ppm, or less, of
NH^OAc-extractable A] (Fig. 16).
The amounts of A1 extracted by the four soil tests were signifi
cantly related (Table **5) and using the regression equations relating
each to A] extracted by NH^OAc, values corresponding to ^0 ppm A1 ex
tracted by the NH^OAc method were as follows: 300 ppm for Bray 1, 400
ppm for Bray 2, and 120 ppm for HCl-H^SO^. On sites which test below
these critical values, the use of soluble phosphate fertilizers, such as
superphosphate, should be avoided with preference given to slowly soluble
sources such as ground rock phosphate. On soils with marginal retention


124
Table 26. Regression coefficients for variables included in 'best fit*
multiple regression
site parameters and
equations of.
the squared
relative height on soil and
terms of these parameters
Source
Coefficients
Sign!f¡canee
R2
Relative height at age 1 year
0.582**
Mean
57.426
Silt + clay
-0.869
JL .L
Depth to LH
1.141
XJU
(Depth to LH)2
-0.008
**
CEC
1.005
**
(NH^OAc-Al)2
0.001
k
Relative height at age 3 years
0.522**
Mean
66.084
Silt + clay
-1.732
**
Drainage class
18.310
**
CEC
1.225
**
(Silt + cl ay)^
0.022
JL
(Drainage class)2
-0.207
JL
Relative height at age 5 years
0.522**
Mean
54.917
Depth to LH
1.813
**
(Depth to LH)2
-0.014
JU JL
Silt + clay
-1.840
JL JL
(Silt + clay)
0.034
JL JL
(h2o-p)2
-0.231
JL
* Significant at the 5% level.
**Signif¡cant at the \% level.


Table 60. Classification, selected physical and chemical properties, and P-retention characteristics
(Langmuir and saturation maxima) of 42 lower Coastal Plain forest soils
k
Field
s i te
Soi 1
type
pH
(KC1)
C 1 ay
Silt
Loss on
ignition
f
Exchange
able Ca
P retention
Langmuir Saturation
maximum maximum
i-
meq/1OOg
p / r
yg
v' 9
Entisols
A5b
Kershaw fs
5.1
1.5
2.3
1.62
0.60
85
289
A5
Kershaw fs
5.2
2.3
3.8
1 .43
0.47
89
338
A15
Chipley fs
b.3
1.9
5-6
2.20
0.62
171
469
A25b
Lakeland fs
5-4
4.0
9-5
3.30
1 .06
204
638
A25
Lakeland fs
5.2
4.0
9-5
3.70
1 .20
256
713
I nceptisols
A7
Rutlege fs
b. 0
1.7
5.7
3-09
0.17
176
513
A10
Rutlege fs
4.1
2.9
6.6
4.29
0.33
308
738
A23
Rutlege fs
b.b
2.5
10.8
3.05
0.27
257
588
Spodosols
A1 6 b
immokalee fs
3.8
1.4
6.0
4.07
0.82
0
0
A16
Immokalee fs
3.6
2.9
4.6
4.99
0.62
0
38
A3
Leon fs
3.7
2.3
4.2
2.52
0.48
0
75
A4
Leon fs
3.7
1 .0
5-9
2.51
0.40
0
100
a6
Leon fs
3.7
0.0
4.7
4.61
0.82
0
100
A18
Leon fs
3.7
1.1
7.4
5.39
0.25
0
50
A19b
Leon fs
3.6
0.9
7.7
3-94
0.46
0
29
A19
Leon fs
3-5
0.9
7.4
3.23
0.30
Q
88
A21
Leon fs
4.1
2.9
7.5
1.42
0.21
bl
163
A22
Leon fs
3.9
1.4
8.1
2.31
0.49
0
63
A23b
Ona fs
3.9
1.6
10.3
3-78
0.44
111
225
FI
Qna fs
4,0
1 .2
4.3
2.90
0.17
165
375
F2
Myakka fs
3-1
0.2
2.2
3.02
0.87
0
75
F3
Myakka fs
4.5
0.2
1.0
2.22
2.28
0
0


102
(Table 11). Water and NH^OAc, which extract similar amounts of most com
pounds were by far the most effective tests. This is consistent with the
previous statement that for young seedlings the amounts of H^O-soluble P
are critical in determining P utilization. However, since H20-extractable
P was not related to long-term P availability (Table 10), it must be
concluded that the utilization of P compounds by seedlings grown in this
greenhouse study provided little indication of the long-term contribution
of such compounds towards meeting the P requirements of trees.
The large build up of H^O-soluble P following incubation of CA1P
and CFeP mixed with Immokalee soil (Table 16) is in agreement with the
finding of Taylor, Gurney, and Lehr (1963). These authors reported that
amorphous phosphates hydrolyzed rapidly to release P upon incubation in a
Hartsells fine sandy loam (pH 4.2). They also found that KTK hydrolyzed
at a much slower rate than amorphous phosphates, which is supported by
the data for KTK in Table 16.
General Discussion
It is apparent (Table 14) that the solubility of P compounds in
chemical extractants cannot be adequately categorized according to the
broad groupings of Ca-P, Al-P and Fe-P as has been the practice in past
studies (Thomas and Peaslee, 1973). Variation in solubility between mem
bers within such groups was as great as that between groups, and these
differences were of equal significance to plant utilization. Although
examination of the solubility of isolated P compounds, known to occur in
the soil and of known availability to plants, probably provides a better
indication of the prediction value of various extractants, the results
can still be misleading. This is because, in practice, these compounds
occur in the soil and the soil properties can appreciably alter their


123
Table 25. Regression coefficients for multiple regression equations of
relative height on depth to LH, HCl-H-SO^-extractable
P(0-20 cm) and the squared terms of these two parameters
Source
Coefficients
Significance
R2
Relative height at
age
1 year
0.402**
Intercept
40.96
"k'k
Depth to LH
1.43
**
(Depth to LH)2
-0.01
**
hci-h2so4-p
1.68
XX
(hci-h2so4-p)2
-0.03
**
Relative height at
age
3 years
0.378**
Intercept
39.99
"kk
Depth to LH
1.40
XX
(Depth to LH)2
-0.01
**
hci-h2so4-p
1.18
*
(hc'i-h2so4-p)2
-0.02
NS
Relative height at
age
5 years
0.357**
Intercept
44.99
XX
Depth to LH
1.45
XX
(Depth to LH)2
-0.01
kk
HC1-H SO -P
2 o
0.42
NS
(hci-h2so4-p)z
-0.01
NS
* Significant at the 5%
1evel.
**Significant at the \%
level.


Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy
SOIL TESTING AS A GUIDE TO PHOSPHORUS FERTILIZATION
OF SLASH PINE (Pinus elliottii var. elliottii Engelm.)
By
Russel 1 Bal lard
August, 1974
Chairman: Dr. W.L. Pritchett
Major Department: Soil Science
This investigation was designed to develop and calibrate a soil
test or tests for predicting the amount and type of phosphatic fertili
zer, if any, needed at time of planting slash pine (Pinus elliottii var.
elliottii Engelm.) on forest soils in the southeastern lower Coastal
Plain.
In a preliminary screening of over 100 different procedures for
extracting or characterizing soil P, it was found that soil-test methods
which extracted relatively small amounts of P from the soil (H2O, NH^OAc
pH 4.8, neutral salts, very dilute acid solutions at narrow soi1:solution
ratios, anion exchange resin) provided the best index of first-year
growth and response of slash pine to P fertilizer both in greenhouse and
field fertilizer trials on the same 10 soils. However, soil-test methods
capable of extracting larger quantities of soil P, irrespective of the
extracting reagent used (strong acid, alkaline, or complexing agents),
were more effective predictors of growth and response achieved over
longer growth periods. Procedures which extracted total, or specific
fractions of soil P (organic-P, Ca-P, Fe-P, Al-P) were less effective
xv i


LIST OF TABLES (continued)
Table Page
61 AMOUNTS OF A1 AND Fe EXTRACTED FROM 42 COASTAL PLAIN
FOREST SOILS BY SIX CHEMICAL EXTRACTANTS. 253
62 AMOUNTS OF A1 AND Fe EXTRACTED FROM 42 COASTAL PLAIN
FOREST SOILS BY FOUR SOIL P-TEST METHODS 255
63 AMOUNTS OF TOTAL P IN THE SURFACE 20 cm OF SOIL COLLECTED
FROM THE CONTROL, Pi(56 Kg P/ha), AND P2(224 Kg P/ha)
PLOTS OF 10 SELECTED FIELD TRIALS 4 YEARS AFTER P-
FERTILIZER APPLICATION 257
x i 1 !


250
Table 59- Continued
Field
s i te
Repli-
cate
N
P
K
Ca
Mg
A1
Fe
%
- ppm--
A29
1
1.08
0.076
0.46
0.19
0.076
500
94
2
0.98
0.071
0.40
0.14
0.068
519
70
3
1.05
0.073
0.43
0.22
0.088
500
70


Table 60. Continued
&
Field
s i te
Soil
type
pH
(KC1)
Clay
Silt
Loss on
ignition
Exchange
able Ca
P retention
Langmuir Saturation
maximum maximum
meq/1OOg
yg p/g -
Spodosols
F4
Myakka fs
3.1
0.2
1.4
2.09
"0.46
23
38
F5
Pomello fs
3.6
0.1
0.1
2.43
1.14
0
38
F6
Wabasso fs
A.6
0.4
1.2
3.32
1.94
70
272
F7
Oldsmar fs
4.7
0.1
0.5
1.66
1.04
0
200
F8
St. Johns fs
6.2
0.9
2.9
3-32
4.64
117
340
U1tisols
A1
Plummer fs
4.2
3.9
12.5
2. 1 1
0.15
174
413
A17
Plummer 1fs
3.6
15.1
24.9
6.48
0.43
128
313
A20
Plummer lfs
4.0
7.5
24.6
3.03
0.13
195
504
A21 b
Plummer fs
4.9
2.8
7.5
2.59
0.51
199
560
A2b
Bladen scl
4.4
20.4
27.0
5.80
2.05
425
1 ,050
A2
Bladen scl
4.4
16.3
26.7
6.71
3.44
511
1 ,288
A8b
Blanton fs
4.8
1.8
4.3
1.51
0.27
93
330
A8
Blanton fs
4.8
1.2
4.4
2.04
0.35
147
438
A29
Kalmia fsl
4.9
5.1
22.8
2.46
0.29
214
475
A26
Lynchburg lfs
4.2
5.2
12.8
4.00
0.39
265
713
A27
McLaurin fsl
5.4
9.0
26.8
4.34
2.29
280
800
A28b
McLaurin fsl
5.1
7.3
23-5
3.86
0.77
308
735
A28
McLaurin fsl
5.1
7.3
23.5
3,86
0.69
276
750
* CRIFF identification code (A series); samples were those collected from 0-20 cm depth within beds
of replicate 2 control plots except where identified by b, which were the bulk samples used in
greenhouse trials; samples labelled F were'those provided by Dr. J. G. A. Fiskell.
t Extracted with 1N_ NH^OAcipH 7.0).


180
methods which extracted larger amounts of P from the soil were more ef
fective predictors of growth and response of slash pine achieved over
growth periods in excess of 1 year. Phosphorus extracted by these soil-
test methods was also the most closely correlated with concentrations of
P in both the tops of greenhouse-grown seedlings and the foliage of 4-
year-old field trees. These latter soil-test methods involved use of
extractants such as strong acids, bases, and reagents such as F and lac
tate which are capable of complexing di- and trivalent ions.
Total soil P and organic soil P were ineffective predictors of
growth and response over all growth periods up to 5 years, although their
value as predictors did show an improvement as the growth period increased.
Aluminum-P, which dominated the inorganic P fraction of the 10 soils, was
the only P fraction significantly correlated with growth response to
added P at any growth stage, but was never as closely related to response
as P extracted by the best soil-test methods.
In an attempt to account for the degree of success in the prelimi
nary screening of several more commonly used soil-test methods, their
ability to solubilize several P compounds, which occur naturally or as
soil-P fertilizer reaction products in the soil, was determined and re
lated to the utilization of these compounds by slash pine seedlings
transplanted into two soils to which these compounds had been added.
The solubility of the P compounds in the soil-test extractants
could not be categorized adequately by broad groupings as Ca-P, A1 -P, and
Fe-P. The degree of crystallinity of the compounds appeared to affect
their solubility as much as their chemical composition. The presence of
soil affected the extractabi1ity of the P compounds, the direction of the
effect being a complicated function of the properties of the soil,


/^N.
17
the soil as opposed to broadcast or banded placement. Brendemuehl (1970)
reported a similar improvement of performance in field trials from mixing
RP with the same soi1 type.
The results from a series of trials established using slash pine to
compare the effectiveness of OSP and RP on soils in the southeastern lower
Coastal Plain have been reported in three publications (Pritchett and
Llewellyn, 1966; Pritchett, 1968; Humphreys and Pritchett, 1971). The
response in 1964, 3 to 5 years after establishment, showed that at equal
rates of P, OSP was superior to RP in all trials (Pritchett and Llewellyn,
1966). Increment growth during the next 3 years, 1964 to 1967, was still
greater from OSP than RP in all trials except on a Leon fs. Response
from the highest RP treatment, which contained eight times as much P as
the highest OSP treatment, was generally greater than that from the OSP
treatments (Pritchett, 1968). Humphreys and Pritchett (1971) reported
growth data only for the highest OSP and RP treatments which did not have
comparable P rates. However, 7 to 11 years after the fertilizer appl¡ca
tion,they found that in the three Spodosols with virtually no P-retention
capacity, all the P applied as OSP had been leached from the upper hori
zons, while most of that applied as RP was still retained in the surface
horizons. In the remaining soils, which ranged in P-retention capacity
from medium to very high, the majority of P applied either as OSP or RP
was retained in the surface horizons. In the most retentive soil
(Bladen), a considerable portion of the OSP had been converted to Fe-P,
a relatively unavailable form as reflected by the extremely low foliar P
levels reported for this treatment.
Humphreys and Pritchett (1971) suggested that selection of a suit
able P source for the acid Coastal Plain soils should be based on their


119
Table 22. Comparison of the effectiveness of P extracted from the sur
face 20 cm of soil and that extracted from within the effec
tive soil depth (volume) at predicting relative height of
slash pine after 1, 3 and 5 years' growth on 72 field sites
Soi1-test
method
Soi 1
sample
Relative height
1 year
3 years
5 years
r2
V
0 20 cm
0.306**
0.132**
0
0
0
*
*
Volume
0.372**
0.239**
0.191**
NH^OAc
0 20 cm
0.270**
0.137**
0.072*
Volume
0.253**
0.220**
0.142**
hci-h2so4
0 20 cm
0.264**
0.224**
0.126**
Volume
0.274**
0.322**
0.211**
Bray 1(3)
0 20 cm
0.235**
0.198**
0.095**
Volume
0.282**
0.284**
0.1-62**
Bray 2
0 20 cm
0.152**
0.178**
0.094**
Volume
0.211**
0.303**
0.202**
* Significant
at the 5% level,
using the model
Y=b log X + c
** Significant at the 1% level.


pine and the 'critical' foliar P values reported by other workers for
this species in the field. The relationships reported above could be of
value in extrapolating greenhouse trial results to field conditions.
Relationships Between Soil-Test Values and Fertilizer Requirements
The association found between mean quantities of P extracted by
soil-test methods and their ability to predict response to P fertilizer
was also apparent in the relationships between soil-test values and P-
fertilizer requirements (Table 12). Phosphorus extracted by weak extrac
tants was most closely correlated with fertilizer requirements over short
growth.periods, while P extracted by stronger extractants was most closely
correlated with fertilizer requirements over longer growth periods.
Although R values for the relationships between soil-test values
and fertilizer requirements in the greenhouse were of a similar magnitude
to those for relative height, R4 values for fertilizer requirements cal
culated from field tests were lower than those for relative height in
field tests. This was particularly pronounced for the fifth, and to a
lesser extent, the third year data. This could be due to variation in
the P-retention capacity of the soils. For instance, Lewis and Harding
(1963) reported lower P-fertilizer requirements for pines growing on soils
of low P-retention capacity compared to those on soils of high P-retention
capacity, but with similar extractable soil-P values. One would expect
differences due to P retention to become more pronounced with time due to
the time dependency of the reversion of relatively soluble phosphates to
less soluble forms (Juo and Ellis, 1968). If variable P retention was
the major factor accounting for the low R2 values shown in Table 12, a
multiple regression equation including both soil-test values and a measure


Table 56. Continued
Field
s i te
Repli-
cate
Depth
Part icl e
size d istribution
Moisture
content
Bulk
density
C1 ay
Silt
Sand
1/3 atm.
15 atm.
cm
0/
'0
g/ cc
A6
1
0-20
1.5
3.9
94.6
5-3
4.7
1.20
20-40
0.8
3.9
95-3
3-5
' 2.2
1.75
40-60
4.0
3-7
92.3
7.3
3.9
1.77
2
0-20
0.0
4.7
95.3
5.7
5-0
1.30
20-AO
0.0
3.0
97.0
2.6
1.7
1.66
40-60
2.7
3-8
93.5
5-0
2.2
1.64
3
0-20
0.5
4.1
95.4
4.2
3.8
1.07
20-40
0.0
3.2
96.8
2.4
1.8
1.68
40-60
2.7
2.5
94.6
6.2
2.7
1.06
A7
1
0-20
1 1
4.8
94.1
5-1
4.5
1.10
20-40
1.5
4.7
93.8
3.6
3-1
1.69
40-60
2.4
4.3
93.3
2.6
1.1
1.80
2
0-20
1.7
5.7
92.6
4.9
4.2
1.18
20-40
1.6
5.1
93.3
3.2
2.8
1.63
40-60
2.5
4.5
93.0
3-2
2.0
1.85
3
0-20
1.4
4.5
94.1
4.6
3-8
1.21
20-40
1.2
4.4
94.4
2.8
0.2
1.57
40-60
1.7
3.9
94.4
4.6
1.3
1.60
A8.
1
0-20
0.9
3.9
95.2
3.2
2.7
1.43
20-40
1.7
94.9
1.8
0.8
1.78
40-60
1.3
3-7
95.0
1.5
0.8
1.78
2
0-20
1.2
4.4
94.4
3.0
2.2
1.40
20-40
1.9
3.5
94.6
2.0
0.9
1.69


Table 55- Continued
Field
s i te
Repli-
cate
Depth
H2
P
NH,OAc
4
P
he i -
H2S0i.
P
Bray 1(3)
P
Bray 2
P
NH,
4
OAc
pH
(h2o)
Ca
Mg
K
A1
cm
ppm
A15
1
0-20
0.5
2.2
20.8
63.6
75.6
30
5
4
60
5-4
20-40
0.2
1.3
15-3
35.9
54.5
15
5
2
62
5.3
40-60
0.4
1.3
12.7
30.0
46.7
15
5
2
52
5-4
2
0-20B
0.8
3.5
21.2
75.8
90.0
30
5
4
52
5-3
0-20
0.6
4.0
27.9
91.7
104.2
30
21
4
47
5.3
20-40
0.4
2.2
18.8
56.8
72.4
15
14
2
47
5.4
40-60
0.3
1.8
12.8
31.4
45.9
15
14
2
50
5.4
3
0-20B
0.8
1.3
7.5
30.3
27.8
45
14
2
28
4.9
0-20
0.8
1.8
8.6
30.3
30.7
45
19
4
25
4.8
20-40
0.4
1.8
16.1
58.5
56.0
15
19
2
33
5.0
40-60
0.3
1.3
20.8
59.1
58.2
30
19
1
28
5.1
A16
1
0-20B
7.3
7.4
8.4
10.6
11.5
215
66
15
5
4.1
0-20
6.2
5.7
6.5
7.5
7.6
135
61
12
2
4.2
20-40
0.9
0.9
1.3
1.5
1.7
45
25
2
2
4.6
40-60
0.6
0.7
1.1
1.7
1.8
15
25
2
37
4.5
2
0-20B
7.8
8.3
7.8
10.7
11.6
200
110
18
6
3-9
0-20
6.5
5.7
9.0
9.1
9.0
135
82
12
6
4.0
20-40
0.8
0.9
1.0
1.5
1.5
60
38
2
5
4.5
40-60
0.6
0.7
1.7
3.8
3.2
15
30
2
60
4.4
3
0-20B
5-6
6.1
6.7
9.1
8.6
170
72
12
6
4.0
0-20
8.7
7.4
9-1
12.6
10.9
200
72
18
6
3-8
20-40
0.7
0.7
0.9
1.3
1 .2
' 30
19
2
2
4.7
40-60
0.5
0.5
0.7
0.9
1 .0
15
14
2
2
4.7


Table 46.. Continued'
T reat-
ment
Repli-
cate
1 year
2 years
Ht.+
Dry wt.5
tops
P in
tops
Ht.+
Dry wt.
Tops Roots
P
Tops
Roots
cm
9
T~
cm
-g .
c
0
Leon
fine sand
(A19)
1
13-5
13.5
0.123
24.5
39-6
22.0
0.0-1
0.076
Po
2
13.3
15-6
0.105
28.5
28.5
19.1
0.067
0.067
3
12.3
12.7
0.095
25-5
42.0
16.1
0.066
0.066
1
12.9
16.7
0.178
29-0
41.3
17.1
0.146
0.107
Pi
2
12.8
17-7
0.224
24.5
45.0
15-8
0.151
0.118
3
12.8
13.2
0.220
26.5
46.4
20.5
0.142
0.134
1
12.0
14.4
0.320
22.3
49.2
16.4
0.195
1.141
P2
2
15.4
15.3
0.3H
28.3
46.1
17-3
0.201
0.108
p r.
3
11.3
13.3
0.31 1
24.8
44.0
16.3
0.196
0.140
1
11.0
13.0
0.340
24.5
41.0
14.1
0.259
0.157
P 3
2
13-3
10.8
0.361
27-5
43.2
15.2
0.31 1
0.179
3
10.3
13-0
0.320
22.3
36.0
17.3
0.235
0.164
Ona
fine sand
(A23)
1
13.3
15.2
0.092
33.8
77.1
25-1
0.048
0.061
Po
2
17.1
13-4
0.089
36.0
79.0
24.5
0.044
0.057
3
14.8
15.0
0.093
32.5
74.3
25.0
0.040
0.057
1
16.3
21 .2
0.149
40.8
87.0
34.0
0.086
0.100
Pi
2
21.8
18.3
0.167
39-5
109-1
32.6
0.093
0.095
3
15-3
20.1
0.154
41.0
88.8
49-5
0.073
0.104


Table 56. Continued
Field
s i te
Rep 1i-
cate
Depth
Particle
size distribut ion
Moisture content
Bulk
density
C1 ay
Silt
Sand
1/3 atm.
15 atm.
g/cc
'0
A21
1
0-20
2.8
7.5
89.7
5.1
2.8
1.32
20-40
3.9
5-9
90.2
3.1
' 1.5
1.71
4o-6o
3.9
5.9
90.2
3.1
1.4
1.80
2
0-20
2.9
7-5
89.6
5.1
3.0
1.50
20-40
4.3
5.8
89.9
3.1
1.5
1 .78
40-60
4.1
6.1
89.8
2.9
1.3
1.82
3
0-20
4.0
6.9
89.I
5-7
3.0
1-35
20-40
4.0
5.8
90.2
3.6
1.8
1.77
40-60
3-9
4.3
91 .8
2.9
1.4
1.85
A22
1
0-20
1.5
7-5
91.0
5.6
2.3
1.29
20-40
1.3
6.1
92.6
2.8
0.8
1.75
40-60
5-4
8.6
86.0
9-2
3.0
1.73
2
0-20
1.4
8.1
90.5
4.9
2.2
1.25
20-40
1.5
7.0
91.5
3.0
0.9
1.78
40-60
5-7
8.7
85.6
8.1
3.1
1.76
3
0-20
1.9
7.8
90.3
4.3
2.1
1.28
20-40
1.7
7.0
90.6
3.3
0.9
1.70
40-60
5-3
8.5
86.2
8.0
3.0
1.75
A23
1
0-20
2.0
9.0
89.O
5-5
3.5
1 -32
20-40
1.2
7.4
91.4
2.4
1 .0
1.64
40-60
4.4
'8.5
87.1
7.7
3.0
1.73
2
0-20
2.5
10.8
86.7
7.9
5.0
1.39
20-40
2.7
9-8
87.5
5.1
2.0
1.63


s
55
intermittent shaking. Following centrifugation, P was determined in
solution. Phosphorus-adsorption isotherms were plotted for yg P
adsorbed/g soil against yg P/ml in equilibrium solution. From these
plots, values were interpolated to fit the linear form of the Langmuir
adsorption isotherm equation (Olisen and Watanabe, 1957)
where C = equilibrium P concentration, x/m = amount of P adsorbed per
unit weight of soil, k = a constant related to binding energy, and
b = P-adsorption maximum.
Significant deviations from the Langmuir equation have been shown
to occur at equilibrium values in excess of 10-20 ppm P (Fox and
Kamprath, 1970). Also the Langmuir maximum, b, depends to a certain
extent on the maximum equilibration value included in the computation
(Gunary, 1970). Therefore, interpolated values were restricted to cases
where equilibration values were 10 ppm or less with 10 ppm being
X
included as the top value for all soils. The Langmuir adsorption
maximum was obtained from the reciprocal of the slope following use of
i
linear regression analysis to obtain the Langmuir equation from the in
terpolated data.
In addition to the Langmuir maximum, the amount of P adsorbed
from the highest level of application (2,500 yg P/g soil) was also used
as an index of the P-retention capacity of these soils.
Retention of field-applied P was determined from the 0-20 cm
samples collected from the Po, Pi, and P2 plots of each replicate of the
10 field trials selected for this purpose. The samples were analyzed
for total P and this value converted to kg P/ha in the surface 20 cm
using the bulk density of the samples. The amount of P retained in the


103
solubility in any extractant; such soil effects were not predictable
(Table 14).
The results observed in the preliminary screening of soil-test
methods can be equated to some extent with the extractabi 1ity of P com
pounds shown in Table 14. Considering all 10 soils used in the prelimi
nary screening, the average amounts of P extracted by the soil-test
methods used were in the order NH^OAc < Truog < Olsen < HCl-h^SO^ < Bray.
This same order was observed for the extractabi 1 ity of CA1P and KTK, but
not CFeP, which was much more soluble in the Olsen than in HCI-h^SO^
extractant. This is consistent with Al-P being the dominant inorganic
form in these soils. The success of the Olsen, HCl-h^SO^, and Bray method
for predicting response over a 3~year growth period can be probably
attributed to these methods extracting forms of soil Al-P similar to CA1P
and KTK. The superiority of the Bray method at predicting response over
a 5~year growth period can probably be attributed to its extraction of
some less soluble, crystalline forms of Al-P, some forms of Fe-P, and its
ability to discriminate against insoluble, completely unavailable forms
of Al-P and Fe-P similar in nature to WA and STR.
In Florida, the NH^OAc method is used on a routine basis for pre
dicting the fertilizer requirements of agricultural crops (Page et al.,
1965). Because of the success achieved with this method by Pritchett
(1968), it has also been adopted for predicting P-fertilizer requirements
of trees. Data from this current study showed that NH^OAc is likely to
seriously underestimate the long-term supply of P to trees growing on
soils which contain P compounds similar to colloidal Al and Fe phosphates
and KTK. In the next section of this study, an attempt will be made to
calibrate soil-test methods other than NH^OAc and which theoretically
should prove superior to NH^OAc for forestry purposes.


Table 6l. Continued
l\ 4-
Field IN KC1 0.3M (NH,.) C0;, CDB 0. 1M Na, P0., 0.5MEDTA 0.1M.HC1
o|te A ] h /. h H / 1=
A] Fe A1 Fe A1 Fe A] Fe A1 Fe
Ppm
F8
2
200
160
250
148
400
140
225
120
202
27
A1
66
350
350
400
520
2,000
610
375
225
152
95
A17
113
450
280
575
372
600
290
475
315
307
83
A20
94
300
440
525
720
1 ,650
430
525
170
247
65
A21 b
53
850
360
800
530
3,675
585
888
178
397
39
A2b
200
825
1,000
825
2,380
11 ,075
3,795
725
460
303
140
A2
185
850
1,560
1050
4,168
10,150
5,220
750
800
300
220
A8b
23
250
130
375
196
950
165
350
70
122
17
A8
37
450
200
550
340
2,100
300
525
95
197
17
A29
49
600
350
725
1,405
2,500
870
500
135
235
23
A26
94
700
490
750
725
2,450
570
650
340
332
86
A27
8
950
490
1,350
4,355
4,800
2,070
550
155
345
19
A28b
44
1,025
545
1,475
3,830
5,000
2,045
825
148
310
17
A28
42
900
270
1 ,200
3,530
4,200
1 ,920
575
150
275
22
* As per Table 60.
+ Citrate-dithionate-bicarbonate extraction (Mehra and Jackson, 1960)-


261
Bremner, J. M. 1965* Total nitrogen, p. 1149-1178. _l_n^C. A. Black
(ed.) Methods of soil analysis. American Society of Agronomy, Madi
son, Wis.
Brendemuehl, R. H. 1970. The phosphorus placement problem in forest
fertilization, p. 43-50. J_r^ C T. Youngberg and C. B. Davey (ed.)
Tree growth and forest soils. Oregon State Univ. Press, Corvallis,
Oregon.
Broerman, F. S. 1967. Nitrogen-phosphorus fertilization of slash pine.
Union Camp Corp., Savannah, Ga., Woodlands Res. Note No. 18. 4 p.
Bromfield, S. M. 1965. Studies on the relative importance of iron and
aluminum in the sorption of phosphate by some Australian soils.
Austr. J. Soil Res. 3^31 -44.
Bromfield, S. M. 1967- An examination of the use of ammonium fluoride
as a selective extractant for aluminum-bound phosphate in partially
p'hosphated systems. Austr. J. Soil Res. 5:225-234.
Cain, J. C. 1959- Plant tissue analysis. Duke Llniv., School of For.
Bull. 15. 184 p.
Carter, M. C., and E. S. Lyle. 1966. Fertilization of loblolly pine
on two Alabama soils. Effects on growth and foliar mineral content
Ala. Agrie. Expt. Sta. Bull. 370. 18 p.
Cate, W. E., E. 0. Huffman, and M. E. Deming. 1959- Preparation of cry
stalline ferric phosphates. Soil Sci. 88:130-132.
Cate, R. B., and L. A. Nelson. 1965. A rapid method for correlation of
soil test analyses with plant response data. N. C. State Univ.
Agrie. Expt. Sta. Intnl. Soil Testing Series Bull. 1. pp. 24.
Chapman, H. D. 1965. Cation-exchange capacity, p. 891-901. C. A.
Black (ed.) Methods of soil analysis. American Society of Agronomy
Madison, Wis.
Chu, C. R., W. W. Moschler, and G. W. Thomas. 1962. Rock phosphate
transformations in acid soils. Soil Sci. Soc. Amer. Proc. 26:
476-478.
Coile, T. S. 1952. Soil and the growth of forests. Advan. Agron. 4:
329-393.
Coleman, N. T., J. T. Thorup, and W. A. Jackson. 1960. Phosphate-
sorption reactions that involve exchangeable A1. Soil Sci. 90:1-7.
Colwell, J. D. 1959. Phosphate sorption by iron and aluminum oxides.
Aust. J. Appl. Sci. 10:95-103-


I
*3
study. Responses recorded after three growing seasons were reported by
Pritchett and Smith (1972).
This study was concerned with evaluating the ability of soil test
ing procedures to predict (a) tree height growth in the absence of P
fertilizer, (b) height response to P fertilizer, and (c) the amount of
fertilizer required to obtain the optimum response. Since N was a limit
ing factor on many of the sites and N x P interactions were common
(Pritchett and Smith, 1972), procedures for computing growth and P-
response parameters were designed to adjust for the N effect. This was
deemed necessary as meaningful relationships between growth and/or re
sponse and a particular nutrient level can only be obtained if the nu
trient in question is the only one limiting growth (Williams, 1962;
Richards and Bevege, 1972 a).
The index of growth in the absence of P fertilizer was taken as
the tallest mean height of the Po Ni treatments, where i = o, l, or 2.
These values were obtained for each replicate at each site for growth
periods of 1, 3, and 5 years.
Relative tree height was used as the index of response to P
fertilizer. This was calculated from:
Relative height = MeaP. height of Po_Nj. treatment x 100>
Maximum height from P addition
Maximum height from P addition, adjusting for the effect of N, was pre
dicted by first fitting response curves to the height data using a second
degree polynomial equation,
Y = a + bX + cX2
where Y = tallest mean height in the N treatments, X = P-application
rate, and a, b, and c are constants. The predicted maximum height was
then obtained by first differentiating the quadratic equation to give


Table 51. Continued
Soi1-test
Mean
extract
Height
Rel
. height
Fert. reqm,
Fol1ar
P
method
able P
jl
r
3
5
1
3
5
1
3
5
4
ppm
,r2
Bray 6
44.3
0.000
0.025
0.002
0.480
0.768
0.774
0.513
0.538
0.275
0.707
Bray 6(2)
49.0
0.000
0.029
0.003
0.462
0.734
0.737
0.487
0.494
0.236
0.700
Bray 7
35.6
0.009
0.057
0.011
0.561
0.803
0.744
0.583
0.515
0.221
0.744
Bray 7(2)
44.1
0.004
0.048
0.009
0.524
0.767
0.734
0.536
0.501
0.226
0.720
Bray 8
'31.4
0.007
0.036
0.002
0.525
0.782
0.698
O.588
0.499
0.200
0.679
Bray 8(2)
41.2
0.005
0.049
0.010
0.540
0.790
0.752
0.545
0.519
0.243
0.719
HC1
15.4
0.160
0.199
0.066
0.720
0.732
0.509
0.708
0.391
0.101
0.651
HC1 (2)
6.1
0.139
0.172
0.058
0.740
0.697
0.479
0.725
0.396
0.114
0.595
HC1 (3)
9.5
0.044
0.119
0.191
0.644
0.770
0.658
0.552
0.477
0.236
0.626
HC1(A)
3.3
0.285
0.331
0.090
0.831
0.589
0.355
0.735
0.325
0.084
0.550
HC1(5)
5.1
0.169
0.204
0.225
0.777
0.661
0.434
0.738
0.375
0.104
0.568
HC1(6)
2.3
0.406
0.386
0.125
0.775
0.467
0.229
0.675
0.249
0.047
0.411
HC1(7)
3.9
0.286
0.265
0.291
0.748
0.571
0.323
0.699
0.345
0.095
0.429
HC1(8)
1.8
0.51 1
0.459
0.242
0.720
0.310
0.113
0.603
0.154
0.013
0.314
HC1(9)
2.6
0.414
0.389
0.079
0.654
0.246
O.O89
0.544
0.115
0.001
0.311
Res i n
1.4
0.399
0.371
0.209
0.738
0.417
0.196
0.650
0.213
0.025
0.404
Res!n(2)
4.5
0.317
0.340
0.199
0.754
0.470
0.264
0.626
0.265
0.054
0.444
Res n(3)
6.2
0.244
0.285
0.169
0.680
0.467
0.281
0.542
0.247
0.051
0.435
Res In(4)
9.6
0.133
0.176
0.081
0.632
0.532
0.371
0.572
0.325
0.085
0.494
Res i n (5)
4.6
0.260
0.219
0.076
0.724
0.564
0.309
0.732
0.299
0.046
0.469
P fractions
Sol -P
2.7
0.401
0.396
0.243
0.687
0.428
0.198
0.571
0.150
0.008
0.421
Al-P
54.3
0.045
0.002
0.018
0.196
0.557
0.647
0.248
0.388
0.206
0.501
Sol-P+Al-P
57.0
0.001
0.031
0.003
0.465
0.718
0.699
0.491
0.507
0.247
0.638
Fe-P
7.7
0.219
0.106
0.136
0.045
0.204
0.288
0.067
0.137
0.069
0.166
o


Fig. 5 Relationship between HC1-H^SO^-extractable P(X) in the surface 20 cm of soil and relative
height(Y) of slash pine 3 years after fertilization.


Table 56. Continued
Field
s i te
Repl¡-
cate
Depth
Particle
size distribut ion
Moisture
content
Bulk
density
Clay
Silt
Sand
1/3 atm.
15 atm.
"
g/cc
cm
A3
2
40-60
3-1
6.2
90.7
3.9
1.9
1.66
3
0-20
0.7
5-3
94.0
2.9
2.4
1.35
20-40
0.3
4.8
94.9
1.5
1 .3
1.71
40-60
3.8
7.1
89.1
5-5
2.6
1.62
A4
1
0-20
0.8
6.4
92.8
4.0
3.6
1.23
20-40
2.1
6.0
91.9
3-8
2.5
1.58
40-60
4.3
6.4
89.3
6.0
3.0
1.71
0-20
0.1
4.9
93.1
4.2
, 2.8
1.24
20-40
1.8
5-1
93.1
3.5
2.0
1 .60
40-60
4.0
6.0
90.0
6.0
5.5
1.56
3
0-20
1.4
6.0
92.6
3.8
3-2
1.27
20-40
0.7
6.3
93.0
3.1
1.5
1 .61
40-60
4.0
6.4
89.6
5-5
2.5
1 .67
A5
1
0-20
1.5
2.4
96.1
2.2
1.6
1.35
20-40
1.4
1.9
96.7
1.6
1.2
1.67
40-60
1.4
1.2
97.4
1.6
0.8
1.65
2
0-20
1.5
2.3
96.2
2.0 <
1.5
1.34
20-40
1.2
1.9
96.9
1.4
0.9
1.66
40-60
1 .4
1.9
96.7
1.7
0.8
1.68
3
0-20
1.6
2.4
36.0
2.1
1.6
1.33
20-40
1.4
0.9
97.7
1.5
1.0
1.61
40-60
1.5
0.9
97-6
1.3
0.6
1.66


266
Maki, T. E. I960. Some effects of fertilizers on loblolly pine. Trans.
VII Int. Congr. Soil Sci. 3:363375.
Martens, D. C., J. A. Lutz, and G. 0. Jones. 1969. Form and availa
bility of P in selected Virginia soils as related to available P
tests. Agron. J. 61:616-621.
Mattson, S., E. Koutler-Andersson, R. B. Miller, and K. Vahtras. 1951.
Phosphate relationships of soil and plant. VIII. Electrokinetics,
amphoteric behavior and solubility relationships of calcium phos
phates. Kgl. Lantbruks-Hogskol. Ann. 18:128-153.
McFee, W. W., and E. L. Stone. 1965. Quality, distribution and variance
of organic matter and nutrients in a forest podzol in New York.
Soil Sci. Soc. Amer. Proc. 29:432-436.
McKeague, J. A. 1967. An evaluation of 0.M pyrophosphate and citrate
dithionate in comparison with oxalate as extractants of the accumu
lation products in podzols and some other soils. Can. J. Soil Sci.
47:95-99.
McKee, V/. H. 1973* Slash pine response to nitrogen and phosphorus on
imperfectly drained soil of the West Gulf Coastal Plain. Soil Sci.
Soc. Amer. Proc. 37:784-788.
McLachlan, K. D. 1965. The nature of available phosphorus in some acid
pasture soils and a comparison of estimating procedures. Austr. J.
Expt. Agrie. An. Husb. 5:125132.
Mead, D. J., and W. L. Pritchett. 1971. A comparison of tree responses
to fertilizers in field and pot experiments. Soil Sci. Soc. Amer.
Proc. 35:346-349.
Mehra, 0. P., and M. L. Jackson. 1960. Iron oxide removal from soils
and clays by dithionate-citrate system buffered with sodium bicar
bonate. Clays Clay Miner. 7:317326.
Merrifield, R. G., and R. R. Foil. 1967. The effects of fertilization
on growth and nutrient concentration in young loblolly pine. La.
State Univ. Agrie. Expt. Sta. Bull. 622. 23 p.
Metz, L. J., C. G. Wells, and B. F. Swindel. 1966. Sampling soil and
foliage in a pine plantation. Soil Sci. Soc. Amer. Proc. 30:397-
399.
Mitchell, H. L., and R. F. Chandler, Jr. 1939. The nitrogen nutrition
and growth of certain deciduous trees of Northeastern United States.
Black Rock Forest Bull. 11. 94 p.
Moehring, D. M. (ed.). 1972. Forest fertilization research in the
South: a review and analysis. Southern Cooperative Series Bull.
158. 80 p.


LIST OF TABLES (continued)
Table Page
13 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND HEIGHT
GROWTH IN THE ABSENCE OF P FERTILIZER OF SLASH PINE IN
FIELD AND GREENHOUSE EXPERIMENTS ON 10 SOILS 84
14 SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS IN THE
PRESENCE AND ABSENCE OF TWO SOILS 89
15 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS
BETWEEN SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS
AND THE UPTAKE OF P FROM THESE COMPOUNDS BY SLASH PINE
SEEDLINGS GROWN ON TWO SOILS IN THE GREENHOUSE 100
16 .SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS FOLLOWING
. 2 MONTHS' INCUBATION IN TWO SOILS 101
17 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL
DEPTHS AND RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3,
AND 5 YEARS' GROWTH ON 72 FIELD SITES 107
18 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN AMOUNTS OF
I P EXTRACTED FROM THE SURFACE 20 cm OF SOIL (BEDDED
AREA) BY FIVE SOIL-TEST METHODS 108
19 SEPARATION OF 72 FIELD SITES INTO RESPONSE QUADRANTS
USING THE TECHNIQUE OF CATE AND NELSON (1965) AND A
CRITICAL HC1-H2S04-EXTRACTABLE P VALUE OF 5 ppm 108
20 REGRESSION EQUATIONS RELATING RELATIVE HEIGHT OF SLASH
PINE AT AGE 1, 3, AND 5 YEARS ( Y1, Y3, AND Y5) TO
' THE LOG TRANSFORMED P EXTRACTED FROM THE SURFACE 20
cm OF SOIL BY THE H20 (Xl) AND HCl-H-SO. (X2) SOIL-
TEST METHODS 116
21 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN AMOUNTS OF
P EXTRACTED BY THE HC1-H2S0/, METHOD FROM TWO SOIL
POSITIONS AND THREE SOIL DEPTHS 116
22 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE
II SURFACE 20 cm OF SOIL AND THAT EXTRACTED FROM WITHIN
THE EFFECTIVE SOIL DEPTH (VOLUME) AT PREDICTING
RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3, AND 5
YEARS' GROWTH ON 72 FIELD SITES 119
23 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES
AND RELATIVE HEIGHT OF SLASH PINE 1, 3, AND 5 YEARS
AFTER P FERTILIZATION ON 72 SITES 120
i x


2*<
adequate sample, even in small areas of mature stands, makes foliar anl
isis rather impracticable for large scale routine diagnosis (Wilde, 1958).
Richards and Bevege (1972 a) suggested that for management
purposes, plant analysis must be quantitative: They stated that
... at the very least it must provide some measure of the degree
of nutrient deficiency, while ideally it should enable us to
predict the magnitude of response to a given application of fer
tilizer. (p. HO)
The response curves shown by Pritchett (1968) are apparently the only
reported fully quantitative foliar analysis data for trees. A report by
Wells et al. (1973) related foliar P to a response index but provided no
information on the amount of fertilizer required to achieve a particular
response. Cain (1959) pointed out that the use of foliar analysis to
predict quantitative fertilizer requirements could be difficult. He
concluded this because the amount of fertilizer required to produce a
desired increment of concentration in plant tissue varies greatly from
site to site with variation caused by climatic and soil factors relating
to the movement and retention of nutrients in the soil.
Soil Analysis
Soil nutrient analysis has been used in forestry, both as a
technique to predict site productivity and as a diagnostic aid in deter
mining the need for fertilizer (Armson, 1973). Until recently the
emphasis in soil-site studies was on soil physical properties (Coile,
1952). This was commented on by Voigt (1958):
Students of forest soils commonly come away from the literature
with the impression that nearly all tree growth can be explained
almost completely by the so-called physical properties of the
soil particularly those related to its moisture regime, (p. 31)
Ralston (196*0 attributed the neglect of soil fertility parameters to
their frequent correlation with other soil properties commonly used in


DEDICATION
To
Phi 11ipa


153
The above recommendations for the HCl-f^SO^ method should be used
only in conjunction with a knowledge of the limitations imposed by both
the conditions under which the calibration was conducted and the short
comings of the soil-test method. These can be summarized: (a) The rec
ommendations are valid only for slash pine plantations over a 3" to 5
year period following planting on unfertilized, acid, sandy soils in the
Coastal Plain. For establishing the fertilizer needs of older establish
ed stands, or the need to refertilize stands 5 or more years after
fertilization at establishment, foliar analysis should probably be used,
(b) The recommendations are based on soil-test values for the surface
20 cm of soil collected from within the bed. Four years after bedding,
when the samples for this study were collected, there was only a small
difference between extractable P in surface samples collected from the
bed and interbed area. However, shortly after bedding these differences
may be substantial due to enhanced mineralization rates. Thus, in order
to apply the above recommendations, soil samples should be collected
prior to bedding, or from the undisturbed area after bedding. For sites
where lower, root-penetrable horizons have substantially higher extract-
able P levels than the surface 20 cm, use of the surface sample may
underestimate the P status. Similarly, where the effective rooting
depth is relatively large, use of surface samples may underestimate the
P status of the site. For instance, from the relationship between
foliar P (Y) and HC1-H2SO/4-extractable P(X) on the 20 sites where depth
to LH was greater than 75 cm was
Y = 0.0029X 0.00007X2 + 0.0796 (R2 = 0.588)
and the extractable P value corresponding to a foliar P concentration of
O-085- was calculated to be 2 ppm, which was considerably less than the


Table 55* Continued
Field
s i te
Rep 1 i -
cate
*
Depth
H2
P
NH,OAc
4
P
HC 1 -
P
Bray 1(3)
P
Bray 2
P
NH,
4
OAc
pH
(h2o)
Ca
Mg
K
A1
cm
ppm__.
A24
2
20-40
0.1
0.1
0.3
0. 1
1 .2
30
30
7
141
4.9
40-60
0.1
0. 1
0.2
0.1
1.2
15
42
7
148
5.1
3
0-20B
0. 1
0.2
1.5
1.6
3.8
60
42
20
104
5-1
0-20
0. 1
0.2
1.5
1.6
3.8
60
42
20
104
5.1
20-40
0.1
0.1
0.3
0.1
1.3
30
38
10
119
5-1
40-60
0.1
0.1
0.2
0.1
1.2
45
94
20
155
5-1
A25
1
0-20B
0.8
4.4
31.0
111.5
150.0
60
48
7
64
5-2
0-20
0.5
3-5
32.6
95-5
141.8
60
42
10
62
5.4
20-40
0.2
2.6
25-2
68.2
112.8
30
30
2
64
5-4
40-60
0.1
2.2
20.4
52.1
87-9
45
30
2
57
5.4
2
0-20B
1 .4
4.8
30.7
121.2
146.0
170
42
12
45
5-4
0-20
1.0
4.8
30.7
128.8
137.0
120
38
12
42
5.5
20-40
0.4
3-5
33-1
103.1
130.0
45
21
2
40
5-7
40-60
0.4
2.6
24.3
72.1
94.0
45
25
2
38
5.7
3
0-20B
0.9
4.4
19-2
83.3
92.0
135
42
15
38
5.4
0-20
0.8
3.1
20.8
77-3
88.5
90
38
. 7
35
5-5
20-40
0.2
1.3
17.4
57.6
72.6
30
30
2
33
5.6
40-60
0.2
1.3
12.9
40.9
51.5
30
25
1
25
5.7
A26
1
0-203
0.3
0.9
2.8
4.6
6.8
45
48
18
126
5.1
0-20
0.2
0.7
2.2
3.2
5.6
15
38
10
114
4.9
20-40
0.1
0.2
1 .0
1.3
3-5
30
21
2
109
4.9
40-60
0.1
0.2
0.6
0.6
3-0
30
21
2
83
4.9
N>
N>
CO


procedure. Although all studies, except that by Hortenstine (1964), used
the Chang and Jackson (1957) prodecure with Fife's (1959) modification of
the NH^F extraction, work by Bromfield (1967) has shown that on recently
fertilized soils the prodecure can provide a misleading estimate of the
relative amounts of A1 -P and Fe-P. This is due to the ability of NH^F at
pH 8.2 to extract appreciable DCPD which leads to an over estimation of
A1 P
In acid soils, A1 and Fe contents have been shown to be closely
correlated with P-retention capacity. Coleman, Thorup, and Jackson (i960),
working with subsoil samples from 60 North Carolina Piedmont soils, found
that exchangeable A1 extracted by N_ KC1 was most closely correlated with
P retention. Similar findings were reported by Syers et al. (1971) using
15 topsoil samples from Brazil and by Udo and Uzu (1972) using 10 acid
--n
Nigerian soils. Aluminum extracted by (NH/t)2C20{ (pH 3-0), a reagent
which extracts amorphous forms of Fe and Al, has been reported to be
closely related to P retention in a range of acid soils (Bromf?eId,1965;
Saunders, 1965; Yuan and Breland, 1969). Yuan and Breland (1969) used
the horizons of 43 virgin soils representative of the soil orders found
in the southeastern Coastal Plain. Iron extracted by oxalate generally
proved to be less effective than oxalate-extractable Al for predicting P
retention (Bromfield, 1965; Yuan and Breland, 1969; Syers et al., 1971).
However on some soils rich in iron oxides, oxalate-extractable Fe proved
to be more successful than Al at predicting P retention (Ahenkorah, 1968).
Extractants which remove crystalline forms of Fe and Al (citrate-dithionate
bicarbonate, termed herein as CDS) were generally less effective than
oxalate for extracting amounts of Al and Fe related to P retention. How
ever CDB extractable-A] appears to be more closely related to P retention


2
this projected acreage is actually fertilized. This, in part, can be at
tributed to the usual caution of forest managers in accepting new practices,
but another major reason is the lack of a precise diagnostic technique for
delineating areas where economic responses can be expected from fertilizer
additions. This inadequacy was emphasized in a recent survey of 50 fores
try organizations in the South soliciting opinions on forest fertilization
research priorities. The concensus of most respondents was that one of
the most pressing needs was "for a predictive mechanism telling where,
when, how and with what to fertilize (Hoehring, 1972).
The two major diagnostic techniques used in forestry are foliar and
soil analyses. Foliar analysis has generally proved to be more reliable
than soil analysis for predicting increased growth of southern pines from
P fertilization (Pritchett, 1968j Wells and Crutchfield, 1969). However
foliar analysis cannot be used in the absence of growing stock on the site
and, consequently, an effective soil test has the advantage over foliar
analysis in that it can be used to predict the need for fertilizer at the
time of planting.
Accurate diagnosis is essential for the success of any forest fer
tilization program. The diagnosis should provide information on the de
gree of response to be expected, the amount of fertilizer required to ob
tain this response, and the most effective nutrient source to use. The
question of most effective P source is of particular importance to P fer
tilization in the Coastal Plain. Humphreys and Pritchett (1971) found
that the P-retent ion capacity of soils in the Coastal Plain region had a
strong influence on the long-term effectiveness of P sources of different
solubi 1ity.
The general purpose of this study was to develop a soil test or
tests which would provide the necessary information required for a forest


137
Table 31. Regression coefficients for variables included in 'best fit1
multiple regression equations of P fertilizer required to
achieve 95% of maximum height on soil and site parameters and
the squared terms of these parameters
Source
Coefficients
Signif¡canee
R2
P-fertilizer requirements over
1 year
0.450**
Mean
45-757
h2o-p
-13.333
-
(H20-P)2
1.280
^A
NH^OAc-Al
-0.279
**
hci-h2so2i-p
-1.776
*
(hci-h2so4-p)2
0.028
*
P-fertilizer requirements over
3 years
0.249**
Mean
16.590
CEC
1.726
hci-h2so^-p
-0.629
JUJU
(h2o-p)2
0.306
aT
P-fertilizer requirements over
5 years
0.109*
Mean
51.284
Drainage class
-23.890
JL
(Drainage class)
3.510
*
* Significant at the 5% level.
**Significant at the 1% level.


83
of soil's P-retention capacity as independent variables would provide a
better prediction of fertilizer requirements. This hypothesis will be
examined in a later section.
Relationships Between Soil-Test Values and Height
In the ensuing discussion, height refers to tree height in the ab
sence of P fertilizer. Extractable soil P was much less effective at
predicting tree height (Table 13) than at predicting relative height
(Table 10). This fact was particularly pronounced under field conditions.
Only certain tests which extracted small quantities of P from the soil
were significantly correlated with the height at the end of 1 year's
growth in the greenhouse and 1 and 3 years' in the field. No single test
was significantly correlated with height at age 3 in the field (Table 10,
Appendix Table 51).
As would be anticipated from trends shown by response and P uptake
data, the effectiveness of soil-test methods involving use of stronger
extractants increased as the growth period in the greenhouse increased.
If inadequate available P was the major factor limiting height growth in
the field, one would have anticipated a similar trend in the field to
that shown in the greenhouse. Although the effectiveness of soil-test
methods which extracted small quantities of P decreased with increasing
growth period, there was no indication of any increase over time in the
effectiveness of soil-test methods involving use of stronger extractants.
This suggests that factors other than available soil P were contributing
to variation in height growth in the field, particularly over longer
growth periods. The effect of site factors other than soil P on height
growth in the field will be examined in a later section.


Table 56. Physical properties of soils collected from 2k field trials
Field
s i te
Repli-
cate
Depth
Particle
size distribution
Moisture
content
Bulk
densit>
C 1 ay
Silt
Sand
1/3 atm.
15 atm.
cm
g/cc
A1
1
0-20
9.0
8.1
82.9
4.6
3.6
1.36
20-40
4.2
13.1
82.7
4.0
1.9
1 .80
40-60
6.9
1 1.4
81 .7
5.5
2.5
1 .84
2
0-20
3.9
12.5
83.6
6.0
4.0
1.37
20-40
4.0
1 1.6
84.4
4.8
2.5
1.76
40-60
4.1
1 1.6
84.3
3.8
1.7
1.79
3
0-20
3.3
11.3
85.4
5.5
3.9
1.29
20-40
3-9
10.2
85.9
3.5
2.0
I.69
40-60
4.5
10.1
85.4
3.7
1.8
1.79
A2
1
0-20
17.2
37.0
45.8
18.4
8.4
1.54
20-40
34.0
30.8
35.2
25.2
12.7
1.56
40-60
56.6
15.6
27.8
32.5
18.4
1.56
2
0-20
16.3
26.7
57.0
17.0
7.5
1.54
20-40
28.8
22.4
48.8
23-3
10.7
1.47
40-60
38.6
19.6
41.8
28.3
15.1
1.45
3
0-20
20.4
27.0
52.6
18.4
8.3
1.58
20-40
29.6
26.4
44.0
23.O
11.6
1.64
40-60
46.0
18.2
35.8
30.8
16.6
1.60
A3
1
0-20
2.2
6.5
91.3
4.8
3.4
1.13
20-40
2.1
5-3
92.6
2.2
1.7
1.66
40-60
1.5
11 .0
87.5
3.6
3.3
1.66
2
0-20
2.3
4.2
93.5
3.4
3.1
1.24
20-40
1.1
4.6
94.3
2.2
1.7
1 .71
ro


Table 55* Continued
Field
s i te
Repli
cate
Depth
h20
p
NH.OAc
P
HC 1 -
H2S04
P
Bray 1 (3)
P
Bray 2
P
Ca
NH,OAc
4
Mg K
A1
pH
(h2)
ppm
A17
1
0-20B
1.7
0.9
3-1
7.8
6.0
60
34
12
38
3-9
0-20
1.3
0.9
2.7
5.8
5.1
60
42
15
40
4.0
20-40
0. 1
0.2
0.4
0.8
0.8
15
14
1
9
4.9
40-60
0. 1
0.5
0.6
1.0
1.1
15
14
1
14
4.7
2
0-20B
2.4
1.1
4.7
9.5
9-7
105
72
28
52
3-8
0-20
1.5
0.9
2.7
7.1
5-6
60
42
18
50
4.0
20-40
0.2
0.5
0.9
1 -9
1.6
60
42
2
14
5.1
40-60
0.2
0.5
1.2
3.4
3-0
30
19
1
16
5.1
3
0-20B
1 .9
1.1
4.5
10.4
9.4
120
61
25
69
4.0
0-20
1.4
1.1
2.6
5-9
5.6
60
30
15
50
4.1
20-40
0.1
0.5
0.4
0.8
0.9
15
10
2
23
4.9
40-60
0. 1
0.2
0.7
1.4
1.3
30
25
1
12
5-0
A18
1
0-20B
4.5
4.4
6.1
7.6
6.1
90
105
18
12
3.7
0-20
4.5
4.0
4.5
6.1
5-4
75
85
12
12
3.8
20-40
0.3
0.5
0.4
0.6
0.7
30
42
1
5
4.7
40-60
0.4
0.5
2.2
4.3
3-8
10
19
1
60
4.5
2
0-203
7.0
6.1
8.8
11.2
9.5
90
88
18
9
3.6
0-20
5.6
5.3
5.8
7.9
7.7
60
76
15
12
3.8
20-40
0.2
0.7
0.6
1.2
1.3
45
30
2
12
4.6
40-60
0.4
0.5
1.6
3-8
3.3
15
21
2
83
4.5
3
0-20B
3.0
1.3
4.7
8.5
8.1
90
40
15
25
3.8
0-20
1.8
1.3
2.9
5.3
5-3
45
33
10
35
4.2
20-40
0.4
0.5
1.1
2.7
2.5
60
34
2
35
4.7
N>
ro
Jr-


54
Phosphorus-Retention Study
Phosphorus-retention characteristics of a range of forest soils
were determined in the laboratory and related to various soil properties.
The value of using the concentration of various elements in the foliage
of slash pine to predict the P-retention capacity of forest soils was
also examined. Parameters found to correlate closely with laboratory
determined P retention were examined for their ability to predict leach
ing losses of P fertilizers in the field.
Soil and Foliage Samples
A total of 42 surface (0-20 cm) soil samples was used in the lab
oratory phase of this study. Twenty-four of the samples were those
collected from the bedded area of the replicate 2 control plots in the
NP field fertilizer trials. The 10 bulk samples were also included.
The eight remaining samples were collected from other uncultivated
sites in Florida and used in this study to increase the range of soil
types.
Foliage samples were those collected from replicate 2 control
plots and control plots from which bulk samples had been collected.
Foliage samples were not available from the above eight remaining
s i tes.
Determination of P Retention
Phosphate retention characteristics were determined by equilib
rating 5"9, air-dry soil samples with 25 ml of 0.0M CaCl2 containing
Ca(^PO/j)2.H2O at concentrations ranging from 0 to 500 ppm P
(0-2500 yg P/g soil). Two drops of toluene were added to inhibit micro
bial activity and the samples were equilibrated for 6 days at 25C with


40
Field Trials
A series of uniform fertilizer trials was established on each of
29 sites in March 1968 by the Cooperative Research in Forest Fertiliza
tion (CRIFF) program. The sites were selected to represent the princi
pal forest soils of the Coastal Plain region. They included 22 soil
types in the Spodosol, Ultisol, Entisol and Inceptisol orders. Fertil
izer treatments were applied 1 to 2 months after planting of the sites
with slash pine seedlings. Twenty-four of the trials which were still
functional in 1972 were used in this study for evaluating soil testing
procedures.
Prior to planting, most sites were burned and disced to remove
residual vegetation. This was followed on all sites by ridging to form
beds approximately 1.22 m wide on 3-05 m centers on which the pine
seedlings were planted. The experimental design consisted of three re
plications of 12 fertilizer treatments in randomized complete blocks.
Each treatment plot was 27-5 by 30.5 m ( 0.08 ha) while the net measure
ment plot was 18.3 by 26.8 m. Nine of the twelve treatments formed a
3x3 factorial with N and P fertilizers. Application rates of both N
and P were 0 (Po, No), 22.5 (P1, Ni), and 90 (P2, N2) kg/ha. Addi
tional treatments numbered 10, 11, and 12 involved applications of K and
micronutrients but these treatments were not included in the present
study. Nitrogen was applied as ammonium nitrate and P as CSP. All fer
tilizer materials were applied in 1.22-m wide bands down tops of beds.
This effectively gave nutrient concentrations within the bed of 2.5
-times those shown above.
So?1 Sampling
Soil samples were collected from each replication at each site


LIST OF FIGURES (continued)
Figure Page
11 RELATIONSHIP BETWEEN DEPTH TO LIMITING HORIZON (X) AND
THE HEIGHT OF SLASH PINE AFTER 5 YEARS' GROWTH 143
12 RELATIONSHIPS BETWEEN P CONCENTRATIONS IN FOLIAGE (X)
OF 4-YEAR-OLD SLASH PINE AND RELATIVE HEIGHT OF SLASH
PINE 1, 3, AND 5 YEARS (Y1, Y3, AND Y5) AFTER P
FERTILIZATION 149
13 RELATIONSHIP BETWEEN HC1-H2S0^-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND P CONCENTRATION IN FOLIAGE
(Y) OF 4-YEAR-OLD SLASH PINE 151
14 PHOSPHORUS-ADSORPTION ISOTHERMS OF FOUR LOWER COASTAL
PLAIN SOIL TYPES REPRESENTATIVE OF FOUR SOIL ORDERS. . 156
15 RELATIONSHIP BETWEEN % P RETENTION FROM THREE P
SOLUTIONS (100, 300 AND 2,500 yg P/g SOIL) AND
NH^OAc (pH 4,8)-EXTRACTABLE A1 158
16 RELATIONSHIP BETWEEN SURFACE (0-20 cm) RETENTION OF P
APPLIED AS CSP 4 YEARS PREVIOUSLY AND NH^OAc (pH 4.8)-
EXTRACTABLE A1 AT 10 SITES GROWING SLASH PINE 174
17 PHOSPHORUS RETENTION AS DETERMINED FROM ADDITION OF
2,500 yg P/g SOIL IN A AND Bh HORIZONS OF SIX
SPODOSOLS 177
xv


269
Richards, S. J. 1965. Soil suction measurements with tensiometers.
p. 153~ 163. J_n C. A. Black (ed.) Methods of soil analysis. Ameri
can Society of Agronomy, Madison, Wis.
Robertson, W. K., L. G. Thompson, and C. E. Hutton. 1966. Availability
and fractionation of residual phosphorus in soils high in aluminum
and iron. Soil Sci. Soc. Amer. Proc. 30:446-450.
Rosendahl, R. 0. 1942. The effect of mycorrhizal and non-mycorrhiza1
fungi on the availability of difficultly soluble potassium and phos
phorus. Soil Sci. Soc. Amer. Proc. 7:477479.
Safford, L. 0. 1973. Forest fertilization in the eastern United States,
p. 206-210. _Jjt_ Forest fertilization symposium proceedings. USDA
For. Serv. General Tech. Rep. NE-3.
Salonen, L. K. 1967. Evolution of forest fertilization in Finland.
p. 41-48. Forest fertilization. Proc. Vth Colloquium of the
International Potash Institute, Jyvaskyla, Finland.
Saunders, W. M. H. 1965. Phosphate retention by New Zealand soils and
its relationship to free sesqu¡oxides, organic matter, and other
soil properties. New Zeal. J. Agrie. Res. 8:30-57.
Schnitzer, M. 1969. Reactions between fulvic and a soil humic compound
and inorganic soil constitnents. Soil Sci. Soc. Amer. Proc. 33:
75-81.
Schomaker, C. E., and V. J. Rudolph. 1964. Nutritional relationships
affecting height growth of planted yellow poplar in southwest
Michigan. For. Sci. 10:66-76.
Schultz, R. P. 1969. Effect of seed source and fertilization on slash
pine seedling growth and development. USDA Southeast. For. Expt.
Sta. Res. Paper SE-49-
Smith, P. F. 1962. Mineral analysis of plant tissue. Ann. Rev. Plant 'S*
Physiol. 13:81-108.
Snedecor, G. W. and W. G. Cochran. 1967. Statistical Methods. Iowa
State Univ. Press, Ames, Iowa.
Stoate, T. N. 1950. Nutrition of the pine. Commw. For. Timb. Bur. Bull.
30. 61 p.
it.-.
Stoeckler, J. H., and H. F. Arneman. i960. Fertilizers in forestry.
Advan. Agron. 12:127-195.
Stone, E. L., and A. L. Leaf. 1967- Potassium deficiency and response
in young conifer forests in Eastern North America, p. 217-229. J_n
Forest fertilization. Proc. Vth Colloquium of the International
Potash Institute, Jyvaskyla, Finland.


LIST OF TABLES (continued)
Table Page
48 DRY WEIGHT, P CONCENTRATION, AND P UPTAKE IN TOPS AND
ROOTS OF SLASH PINE SEEDLINGS AFTER 1 YEAR OF GROWTH
ON NINE SOILS IN THE GREENHOUSE 196
49 AMOUNTS OF P EXTRACTED FROM 10 SOILS BY SOIL-TEST METHODS 197
50 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS
BETWEEN SOIL-TEST VALUES AND SEEDLING PARAMETERS FROM
GREENHOUSE STUDY 1 202
51 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS
BETWEEN SOIL-TEST VALUES AND TREE PARAMETERS FROM 10
SELECTED FIELD TRIALS 207
52 DRY WEIGHT AND P UPTAKE OF ENTIRE SLASH PINE SEEDLINGS, AS
AFFECTED BY P SOURCE, AFTER THE 8 MONTHS OF GROWTH ON
TWO SOILS IN THE GREENHOUSE 212
53 ANALYSIS OF VARIANCE OF SEEDLING DRY WEIGHTS AND P UPTAKES
OF GREENHOUSE TRIAL 2 213
54 .SOIL CLASSIFICATION, SITE PROPERTIES, AND SELECTED CHEMICAL
PROPERTIES OF UNFERTILIZED SOILS (0-20 cm), FOR THE 24
FIELD SITES 214
55 AMOUNTS OF P EXTRACTED FROM SOILS, COLLECTED FROM CONTROL
PLOTS OF 24 FIELD TRIALS, BY FIVE SOIL-TEST METHODS, AND
AMOUNTS OF Ca, Mg, K, and A1 EXTRACTED BY NH^OAc(pH 4.8)
AND SOIL pH 218
56 PHYSICAL PROPERTIES OF SOILS COLLECTED FROM 24 FIELD TRIALS 231
57 HEIGHT AND RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3, AND 5
YEARS' GROWTH ON ¡2 SITES IN THE FIELD 241
58 PHOSPHATE FERTILIZER (CSP) REQUIRED TO ACHIEVE 90, 95, AND
100* OF MAXIMUM HEIGHT OF SLASH PINE 1, 3, AND 5 YEARS
AFTER FERTILIZATION OF 72 FIELD SITES 244
59 CONCENTRATIONS OF N, P, K, Ca, Mg, A1, AND Fe IN FOLIAGE
COLLECTED FROM 4-YEAR-OLD SLASH PINE GROWING IN THE
CONTROL PLOTS OF 24 FIELD TRIALS 248
60 CLASSIFICATION, SELECTED PHYSICAL AND CHEMICAL PROPERTIES,
AND P-RETENTI ON CHARACTERISTICS (LANGMUIR AND SATURATION
MAXIMA) OF 42 LOWER COASTAL PLAIN FOREST SOILS 251
xi i


Table b6. Continued
Trea ti
men t
Repli
cate
1 year
2 years
+
Ht.
Dry wt.^
tops
P in
tops
Ht.
Dry wt.
Tops Roots
F
Tops
>
Roots
cm
9
'o
cm
g

./
'0
Blanton fine sand (A8)
1
13.8
13.4
0.161
29.0
58.8
20.7
0.131
0.162
P 2
2
14.8
12.1
0.171
35.8
59-9
23-7
0.131
0.166
3
13-4
10.9
0.144
30.0
64.9
25.0
0.115
0.166
1
13.3
12.6
0.220
31.0
60.4
25.5
0.175
0.231
P 3
2
14.8
13-6
0.172
32.8
64.2
20.0
0.156
0.162
3
12.1
12.4
0.218
29.0
59.3
20.8
0.158
0.245
Plummer fine sand (A21)
1
13-9
12.4
0.140
35-5
11.2
30.4
0.086
0.112
P 0
2
16.3
1 1.4
0.120
36.5
96.4
23.2
0.090
0.118
3
15-6
11.9
0.138
37-3
75.2
31 .8
0.090
0.110
1
14.6
11.6
0.145
33-5
78.2
25.2
0.090
0.134
P 1
2
16.4
16.1
0.128
32.5
86.2
30.2
0.103
0.114
3
14.0
10.9
0.136
36.0
85-3
40.7
0.106
0.127
1
14.9
14.6
0.155
38.0
15.1
22.3
0.114
0.166
P 2
2
16.0
17-9
0.134
46.5
78.0
26.7
0.120
0.137
3
14.0
13.3
0.139
33.0
84.3
23-9
0.111
0.166
1
16.1
18.0
0.168
40.0
104.0
36.0
0.106
0.164
P 3
2
15-5
18.5
, 0.154
40.8
84.3
27.1
0.126
0.161
3
15.0
16.6
0.165
38.0
88.4
24.8
0.121
0.175
Co
CO


Table 42. Regression equations relating P retention {%) to different forms of soil A1 and Fe
Prediction equations*
R2
Equation
number
V = 3.708 + 0.040
(am.
Fe) +
0.835
[2]
-<
II
O
+
O
o
o
(am.
Fe) +
0.015
/
(am. Al)
0.930
[3]
Y = 0.794 + 0.022
(am.
Fe) +
0.017
(am. Al) + 0.054 (ex. Al)
0.944
[4]
Y = 1.456 + 0.023
(am.
Al) +
0.012
(cry. Al) + 0.029 (ex. Al) + 0.190 (ex. Fe)
0.955
[5]
* All regression coefficients significant at least at the 5% level.
^ Symbols; am. = amorphous; ex. = exchangeable; cry. = crystalline, all expressed as ppm on soil basis.


which vary in their capacities to extract P from soils. The similar ef
fectiveness of different extractants which remove equivalent quantities
of P suggests that they extract the same form(s) of P from the soil.
Aluminum-P is probably the form most commonly involved since this fraction
dominates the inorganic P component of the 10 soils (Appendix Table kS)
and most soil-test extractants are capable of solubilizing A1 -P (Thomas
and Peas lee, 1973)
Soil-test methods involving use of weak extractants, such as H^O
and NH^OAc which extract only small amounts of P from most soils, pro
vided the best index of tree response to P fertilizers over short growth
periods in both the greenhouse and field trials (Table 10). Water-
extractable P was slightly superior to that extracted by NH^OAc at pre
dicting response after 1 year's growth of seedlings in the greenhouse,
while NH^OAc was a better predictor of response than H20 after 1 year's
growth in the field. Other soil-test methods which provided a good index
of response over short growth periods included those involving use of
neutral salts, short term extractions with anion exchange resin, and
very dilute acid solutions at a narrow soi1:solution ratio (Appendix
Tables *9 and 50). The effectiveness of other soil-test methods as indi
cators of short term response tended to be inversely proportional to the
amount of P extracted (Fig. 1 and Table 10).
Responses after 3 and 5 years' growth in the field were most
closely related to P extracted by soil-test methods which removed larger
amounts of P from the soil than H20 and NH^OAc. As the growth period
2
increased from 1 to 3 or 5 years, R values decreased for the soil-test
methods involving use of weak extractants. The R2 values for response
after 3 and 5 years' growth increased with increasing amounts of P


141
Y = 26.24 + 6.25X 0.05X2 (R2 = 0.323)
Barnes and Ralston (1955), in an examination of soil factors affecting
growth of slash pine in Florida, also found depth to a fine-textured
horizon and depth to mottling to be significantly related to site quality
of slash pine. The form of the relationship reported by these authors,
was the same as that shown in Fig. 11. Height increased with increasing
depth to limiting horizon up to a maximum, thereafter a further increase
in depth was associated with a decline in height. Barnes and Ralston
(1955) interpreted this relationship in terms of the effect of soil
moisture and aeration on growth. The optimum depth to a fine-textured
horizon or mottling of 20 to 30 inches (51 to 76 cm) reported by these
authors corresponds closely to that shown in Fig. 11.
In order to examine the possibility that extractable-P levels may
have contributed significantly towards variation in height within
specific levels of depth to LH, multiple regression equations of height
on depth to LH, H^O-extractable P, and the squares of these terms were
run. Results (Table 35) showed that only H^O-P made a significant
contribution towards accounting for variation in height at age 1 year.
However, as the growth period was increased from 1 to 3 to 5 years,
H^O-P became less significant and depth to LH more so. Since both H^O-P
and depth to LH were significantly related to height at all ages, and
yet these two variables were not linearly correlated (Table 24), inter
pretation of these data appeared difficult. However, a regression of
H^O-P on depth to LH and its squared term was highly significant
(R2 = 0.384). Thus, it appears probable that the significant relation
ship between H20-extractab1e P and height at age 5 years was in part due
to the relationship of H20-extractable P to depth to LH. This is con
sistent with the hypothesis formulated from the results of the prelimi-


115
quantity factor since it removes h^O-soluble P in addition to certain
solid phase forms of soil-P. Thus, the HCl-H^SO^ method can probably be
considered to provide a composite index of the quantity and intensity
factors of soil-P supply.
Effect of soil sampling position and depth
The amount of P extracted from surface (0-20 cm) samples
collected from within the bedded tree rows was more closely related to
relative height than P extracted from surface samples collected from the
undisturbed interbed area (Table 17). This was not surprising since
nearly all the rooting activity of the young trees is within the bedded
area. All soil-test methods extracted more P from soils collected within
the beds than from those collected between beds. The increase in avail
able P following bedding has been recorded previously (Haines and
Pritchett, 1965). These microsite differences in available P have impli
cations for the operational use of soil tests in forestry. It may be
more convenient to collect soil samples from areas prior to bedding, in
which case critical values somewhat lower than those established for
samples from the bedded areas should be used. For soils collected in this
study, the relationship between P extracted by the HC1-H2S0^ method from
interbed samples (Y) and bedded samples (X) was:
Y = 1.11X 0.9* (r= 0.983)
Using this relationship, the HC1-H2S0^-extractable P value for the inter
bed samples, corresponding to 5 ppm P for the bedded samples, was 4.6 ppm
P.
The quantity of P extracted from samples collected from the
20-40 cm depth was generally less closely related to relative height than
P extracted from surface samples (Table 17). However, the amount of P


Table 43- Aluminum and Fe extracted by four soil-test methods and thier correlation with P
retention
Soi1-test
method
pH Element Mean
Range
P retention
Langmuir Saturation
maximum maximum
ppm r
NH.OAc
A. 8
A1
69
8-222
0.937**
0.934**
Fe
14
0-155
0.659**
0.651**
A1 + Fe
-
-
0.932**
0.906**
HCl-H2S0ii
1.3
A1
159
23-435
0.843**
0.820**
Fe
35
6-223
0.724**
0.713**
A1 + Fe
-
-
0.904**
0.881**
Bray 1
2.5
A1
487
30-1390
0.909**
0.912**
Fe
77
13-395
0.783**
0.786**
A1 + Fe
-
-
0.926**
0.928**
Bray 2
1.5
A1
545
60-1480
0.901**
0.907**
Fe
109
20-565
0.829**
0.823**
A! + Fe
-
-
0.91.9**
0.927**
** .Significant at the 1% level.


170
Table 44.
Foliar nutrient concentrations and their correlation with
soi1-P retention
Element
Mean
Range
Correlation
with P
retentionT
£
r
N
n
1.07
o
co
i
o
0.244
P
/
0.087
o.oA 0.54 0.115
0.172
K
0.44
0.24 0.60
0.010
Ca
0.18
0.12 0.25
0.045
Ca + Mg
0.107
--ppm
Mn
216
69 -
663
0.147
A1
436
231 -
694
0.425*
Fe
47
22 -
92
0.5!4**
A1 + Fe
0.478**
Al/P
0.570**
Fe/P
0.499**
A1 + Fe/P

0.593**
* Significant at the 5% level.
** Significant at the 1% level.
^ P retention determined as % adsorption of 2,500 pg P/g soil.


Table 62. Continued
&
Field
site
NH^OAc
HCl-H2S0Zf
Bray 1
Bray 2
A1
Fe
A1
Fe
A1
Fe
A1
Fe
-ppm-
F6
32
16
80
53
165
94
160
123
F7
8
2
23
15
50
30
60
44
F8
35
3
143
18
55
49
245
75
A1
85
46
120
90
450
200
395
240
A17
88
19
190
63
420
104
430
155
A20
132
24
185
58
600
101
600
138
A21 b
105
8
300
35
1,010
100
1,110
134
A2b
222
92
212
139
925
225
1,025
323
A2
200
155
208
223
1,030
395
1,100
565
A8b
40
5
95
16
355
55
400
68
A8
70
7
180
19
660
79
670
93
A29
92
8
220
30
780
127
900
154
A26
140
36
245
75
760
155
820
220
A27
82
5
280
23
960
52
1 ,250
96
A28b
91
6
257
27
998
58
1,175
90
A28
88
6
230
27
950
73
1,250
113
As per Table 60.


53
Several soil-test methods, details of which are given later, were
used to extract P from subsampies of the above incubated and non-
incubated mixtures. In addition, the extractabi 1ity of the P compounds
in the soil-test extractants was also determined. Amounts of P
compounds used for the extraction in the absence of soil were calculated
to provide the same amount of P per unit of extractant as that in the
corresponding soil mixture.


BIOGRAPHICAL SKETCH
Russell Ballard was born on February 21, 19^ in Bromyard,
Herefordshire, England. His family moved to Kenya in 19^6 where he
attended school, graduating from high school in 1962. In 1963 he
went to New Zealand to study agriculture at Massey University, Palmer
ston North. He graduated from this institute with a Bachelor of Agri
cultural Science in 1966 and in 1968 completed a Master of Agricultural
Science with first class honors in Soil Science. While studying for his
master's degree he was the recipient of the Murray and the David Henry
Scholarships.
During the period January, 1969 to August, 1971 he was employed
as a Scientist in the Soils and Tree Nutrition section of the Forest
Research Institute, Rotorua, New Zealand. In August, 1971 he was given
leave to study for a Ph.D at the University of Florida. For this purpose
he was awarded a National Research Advisory Council Fellowship and a
Fulbright travel grant.
He is a member of the New Zealand, American, and International
Societies of Soil Science. He is also a member of Sigma X¡ and is the
author of 13 scientific publications.
Russell Ballard is married to the former Phil lipa Anne Gifford
of Hastings, New Zealand.
273


184
2. The supply of soil P to trees over growth periods longer than 1
year is determined not only by P in the solution phase, but also by
the amount of solid-phase P in equilibrium with it.
3. In unfertilized forest soils of the lower Coastal Plain, A1 -P is the
main source of solid-phase inorganic P contributing to the P require
ments of trees over the first 5 years' growth.
4. Soil-test methods which extract P from the more soluble fractions of
the Al-P component provide the best index of the P supply to trees
over the first 5 years. These include the HCI-^SO, and two Bray
methods.
5. The contribution of P derived from soil depths below 20 cm to the
tree's requirements increases with the age of trees during the first
5 years. This contribution is modified by the degree of penetration
of lower soil horizons by the tree roots.
6. Slash pine planted on soils containing less than 5 ppm of HC1-H S0^~
extractable P in the surface 20 cm should respond significantly over
the first 5 years to P-fertilizer applications. Equivalent critical
levels for the Bray 1(3) and Bray 2 methods are 7-5 and 10 ppm of P,
respect ively.
7. Soils testing between 5 and 2.5 ppm of HC1-H^SO^-extractable P wi11
need an application of ca. 20-40 kg P/ha to maintain an adequate P
supply for slash pine over the first 5 years or so of growth. Soils
testing below 2.5 ppm will need ca. 40-80 kg P/ha. Soils with a very
high P-retention capacity may require higher application rates than
those proposed above, while soils with no restriction to root pene
tration within the surface 75 cm may require less P.


LIST OF TABLES (continued)
Table Page
35 REGRESSION COEFFICIENTS FOR MULTIPLE REGRESSION EQUATIONS
OF RELATIVE HEIGHT ON DEPTH TO LH, H20-EXTRACTABLE
P (0-20 cm) AND THE SQUARED TERMS FOR THESE TWO
PARAMETERS 144
36 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN 'BEST
FIT1 MULTIPLE REGRESSION EQUATIONS OF HEIGHT AT AGE 1, 3,
AND 5 YEARS ON SOIL AND SITE PARAMETERS AND THE SQUARED
TERMS OF THESE PARAMETERS 146
37 RELATIONSHIPS BETWEEN HEIGHT, RELATIVE HEIGHT, AND P
FERTILIZER REQUIREMENTS (95%) AT AGE 1, 3, AND 5 YEARS,
- AND P CONCENTRATIONS IN THE FOLIAGE OF 4-YEAR-OLD SLASH
- PINE 148
38 RELATIONSHIPS BETWEEN EXTRACTABLE-SOIL P AT TWO POSITIONS,
THREE DEPTHS, AND WITHIN THE EFFECTIVE SOIL DEPTH
(VOLUME), AND P CONCENTRATIONS IN THE FOLIAGE OF 4-YEAR-
OLD SLASH PINE 148
39 SOIL PROPERTIES AND THEIR CORRELATION WITH P RETENTION. . 159
40 EXTRACTABLE A1 AND Fe VALUES AND THEIR CORRELATION WITH
P RETENTION 162
41 CORRELATIONS BETWEEN DIFFERENT FORMS OF A1 OR Fe, AND P
RETENTION 164
42 REGRESSION EQUATIONS RELATING P RETENTION (%) TO
DIFFERENT FORMS OF SOIL A1 AND Fe 166
43 ALUMINUM AND Fe EXTRACTED BY FOUR SOIL-TEST METHODS AND
THEIR CORRELATION WITH P RETENTION 168
44 FOLIAR NUTRIENT CONCENTRATIONS AND THEIR CORRELATION WITH
SOIL-P RETENTION 170
45 REGRESSION EQUATIONS OF A1 EXTRACTED BY HCI-H2S0 (Yl),
BRAY 1 (Y2), AND BRAY 2 (Y3) ON A1 EXTRACTED BY NH^OAc (X) 175
46 HEIGHT, DRY WEIGHT, AND P CONCENTRATION OF SLASH PINE
SEEDLINGS AFTER 1 AND 2 YEARS' GROWTH ON 10 SOILS IN THE
GREENHOUSE RECEIVING FOUR P TREATMENTS 187
47 AVERAGE HEIGHTS OF SLASH PINE AS AFFECTED BY P TREATMENTS
AFTER 1, 3, AND 5 YEARS' GROWTH IN THE FIELD ON 10 SOILS 194
x!


Table 63.
Amounts of
total P
in the surface
20 cm
of soil
collected
from the
control,
Pi(56 Kg P/ha),
and P2 (22^+
Kg P/ha)
plots of 10 selected
field
trials 4 years after
P-ferti1
izer applicat ion
Field
Repli-
Bulk
Tota 1
P/20 cm
. f
P retention
NH.OAc (pH 4.8)
-A1
s i te
cate
density
Po
Pi
P2
pi
P 2
Po
Pi
P 2
g/cc
Kg/ha--
-----
-%

-""Ppm

A2
2
1.16
292
330
437
68
65
175
175
190
3
1.16
261
400
-
62
175

190
A5
1
1.52
212
245
318
59
47
35
35
33
2
1.52
247
273
355
46
48
33
35
37
3
1.52
217
242
360
45
64
35
33
35
A8
2
1.38
266
295
-
52
-
57
50
-
3
1.38
280
316

64
-
40
52

A15
1
1.3**
389
416
515
48
56
70
77
67
2
1.34
443
492
-
87
-
67
67
*
A16
1
1.14
123
124
98
2
-11
5
5
5
2
1.14
123
133
96
19
-12
7
3
5
3
1.14
123
113
94
-19
-13
7
3
5
A17
1
1.10
184
213
245
52
27
52
40
37
A18
1
1.19
120
119
90
-2
-13
12
10
7
2
1.19
113
90
105
-41
-4
12
10
10
3
1.19
127
12 7
110
0
-3
37
12
18
A21
1
1.21
1158
1228
1375
125
99
98
127
115
3
1 .21
742
8cf9
996
120
100
100
123
112
A22
1
1.25
104
108
126
1 1
10
15
12
12
2
1.25
117
123
135
12
8
15
15
18
N>
V/l


TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ¡¡ i
LIST OF TABLES vi
LIST OF FIGURES xlv
ABSTRACT xvi
INTRODUCTION 1
LITERATURE REVIEW 4
Forest Fertilization 4
Historical 4
N and P Fertilization 5
P Sources Used in Forestry 9
Reactions in Soils 9
P Effectiveness in Forestry 15
/Diagnostic Methods 20
i Foliar Analysis 20
Soil Analysis 24
? Sampling procedures 25
Extraction methods 28
Interpreting soil test results 33
u Prediction of productivity 33
Prediction of fertilizer response 34
MATERIALS AND METHODS 39
Introduction 39
FieldTrials 40
So i 1 Sampl ing 40
Foliage Sampling 42
Growth and Response Parameters 42
Greenhouse Trial 1 45
Establishment 45
Harvesting 46
Growth and Response Parameters 47
Greenhouse Trial 2 48
P Compounds 48
Establishment 49
v


46
Thirty slash pine seeds, previously soaked in 0.1% citric acid for
24 hours to facilitate germination, were planted in each pot. Seedlings
were thinned by stages to eight seedlings per pot 2 months after planting.
Soil moisture in the pots was maintained between 50 and 100% of field ca
pacity by watering to predetermined weight with distilled water; no
leaching losses were observed. Heating and cooling equipment in the
greenhouse maintained the temperature between 24C and 35C. Day length was
not controlled during the first 9 months, but following the first harvest
in December of 1972, day length was increased to 15 hours using low-
intensity incandescent light. The position of the pots on the greenhouse
benches was altered periodically to minimize micro-environment effects.
Harvesting
In December of 1972, seedling heights from root collar to terminal
bud were measured to the nearest millimeter. Four seedlings per pot,
with an average height approximating that of all eight seedlings, were
harvested removing the tops at root-collar level. The roots of the
harvested seedlings were left undisturbed in the soil. Root: shoot rela
tionships at the first harvest were established by harvesting the entire
plants from the extra pot of each soil included in the trial.
Root harvests in these extra pots were made by washing the soil
and root mass on a 6-mm sieve followed by a rinsing of the recovered roots
in distilled water. A final harvest of the remaining four seedlings per
pot was made in October, 1973, following height measurements. Tops and
roots were harvested separately. Old roots left from the first harvest
were separated and discarded during sieving. All tissues samples were
dried at 70C, weighed, and ground to pass a 1-mm sieve using a stainless
steel Wiley mill.


36
superior to several others examined, including 0.05N_ HC1 + 0.025^ I^SO^
and Bray 2. These authors also gave calibration curves indicating the
amount of fertilizer required to give a specific degree of response at a
particular soil-test value. Wells et al. (1973) reported that P
extracted by 0.05N HC1 + 0.025N_ was superior to that extracted by
the Bray 2 test at delineating responsive sites. From their response
curves, they proposed a value of 3.0 ppm P extracted by 0.05N_ HC1
+ 0.025N_ I^SO^ as providing the best separation of responsive and unre
sponsive sites. Their calibration curves indicated only the degree of
response expected at any particular soil-test value and not the amount
of fertilizer required to achieve this response. The critical level
found by Wells et al. (1973) for loblolly pine probably accounts for the
lack of response by loblolly pine to P recorded by Carter and Lyle (1966)
and Moschler et al. (1970) on soils testing above this critical level
(Table 2).
Wells et al. (1973) reported several sites which,although -
falling into the responsive category according to their soil-P test
value, did not respond to P fertilizer. They attributed this in most
cases to N being more limiting than P on these sites. This illustrates
the need in calibration studies, pointed out by Williams (1962) and
Richards and Bevege (1972), for utilizing field trials in which all nu
trients except the one under observation are nonlimiting to growth.
Wells et al. (1973) also pointed out that their critical value estab
lished for the response of 3-year-old loblolly, may not hold true for
older trees due to the increasing P requirements of trees with age. As
mentioned earlier, the importance of stand age or development to the
nutritional demands of trees was discussed by Leaf (1968) and illus
trated by the work of Will (1964).


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ESGRPUNBX_GIS55Y INGEST_TIME 2015-03-25T20:17:16Z PACKAGE AA00029998_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


collected from the bedded area using the same equipment and procedure as
used in the collection of the samples from control plots.
Following excavation of the four soil pits in each control plot,
the following information was recorded: Depth and color of the A1
horizon; and depth to mottling, spodic horizon, or fine-textured layer
where they occurred within the surface 90 cm. The experimental sites
were classified within one of five drainage classes by CRIFF personnel at
the time of establishment (Pritchett and Smith, 1972).
All samples were air-dried after collection. Samples for soil
analysis were passed through a 2-mm sieve. Where analysis procedures re
quired it, subsamples were ground to pass a 0.2-mm sieve. Bulk samples
used in greenhouse trials were screened through a 6-mm sieve while sub
samples for soil analysis were passed through a 2-mm sieve.
Foliage Sampling
Foliage samples were collected in December-January of 1971/72 from
the control plot of each of the three replicates at each site. Each
sample consisted of a composite of needles collected from a minimum of
five trees within the net plot. The trees were selected to represent the
range in tree size and vigor found in the plot. Needles were taken from
the previous spring flush on the uppermost whorl bearing secondary
branches. The foliage samples were stored on ice while in transit to the
laboratory. All samples were dried at 70C and ground to pass a 1-mm
sieve in a stainless steel Wiley mill.
Growth and Response Parameters
Heights of all living trees were determined by CRIFF personnel at
the end of the first, second, third and fifth growing seasons. The
measurements taken after the second growing season were not used in this


Table 50. Continued
Soi1-test
Mean
extract-
Helght
Re 1 1
height
Fert.
reqm.
%? In
tops
P uptake
method
able P
*
1
2
1
2
1
2
1
2
1
2
ppm
_2
- - r\
P fractions
Sol-P+Al-P
57.0
0.145
0.436
0.296
0.237
0.322
0.396
0.742
0.818
0.266
0.874
Fe-P
7.7
0.052
0.173
0.006
0.001
0.007
0.013
0.158
0.411
0.027
0.470
So 1-P+Al-P+Fe-P
64.7
0.134
0.412
0.239
0.190
0.250
0.322
0.666
0.811
0.221
0.858
Ca-P
0.6
0.251
0.009
0.005
0.072
0.014
0.124
0.083
0.095
0.043
0.005
Sol-P+Al-P+Fe-P
+Ca-P
65.3
0.130
0.414
0.243
0.197
0.255
0.332
0.678
0.823
0.221
0.859
Organic-P
42.6
0.288
0.176
0.050
0.010
0.050
0.001
0.070
0.164
0.141
0.459
Tota 1-P
136.1
0.099
0.151
0.040
0.000
0.043
0.015
0.213
0.442
0.066
0.600
Buffering capacity
Capacity(1)
-
0.276
0.471
0.212
0.108
0.242
0.194
0.516
0.605
0.251
0.904
Capacity(2)
-
0.141
0.070
0.498
0.448
0.432
0.338
0.113
0.009
0.404
0.001
Capacity(3)
-
0.075
0.021
0.431
0.361
0.308
0.271
0.065
0.014
0.326
0.008
* Age in years at harvesting or measurement.
t
Using the model Y = b logX
cant at the 5% level, R^ >
+ c, where Y = seedling parameter and
0.586 significant at the 1% level.
X
soi1-test value.
R2 > 0.397 signifi-


167
between many of the Fe and A1 forms, noninclusion as a significant com
ponent cannot be taken to indicate nonparticipation in P retention. The
data concur with other findings (Colwell, 1959) that the order of activity
of A1 and Fe per unit weight in P retention is exchangeable > amorphous >
crystalline (equation [5]). Although various extractants of A1 usually
provide the best single index of P retention, these multiple regression
data suggest that the contribution of active forms of Fe to P retention
is probably considerable, particularly on a per unit-weight basis.
Aluminum and Fe extracted by soil-P test methods
It is apparent from the results presented above, that any of the
extraction procedures conventionally used to remove various forms of A1
and Fe from the soil could be used as an aid to identifying soils of dif
ferent P-retention capacity. However, some of these procedures, such as
the oxalate, pyrophosphate, and CDB, presented difficulties in the deter
mination of the A1 and Fe and which, if used, would mean an additional
operation in any soil-testing program. However, A1 extracted by the
NH^OAc procedure, which is also used to extract soil P, could be satis
factorily used to predict soil-P retention. But, evidence presented in
an earlier section showed that other soil tests such as the HCl-P^SO^ and
Bray methods were more suitable than NH^OAc for predicting the long-term
P status of forest soils. Thus, the utility of A1 and/or Fe extracted by
these methods as a means of predicting P retention was examined.
Aluminum and Fe extracted by the NH^OAc, HCl-H2SOi), Bray 1 and
Bray 2.methods were significantly correlated with soil-P retention (Table
A3). Amounts of A] and Fe extracted from the individual soils by the
four soil-test methods are given in Appendix Table 62. For all four


74
extracted by the H0 and Bray 1(3) methods (Table 6), which can be taken
as Indicative of the intensity and quantity factors of soil P supply, re
spectively (Williams and Knight, 1963). For soils containing high ex
tractable P by and moderately low by Bray 1(3) methods--Immokalee and
Leon--no growth response occurred in the first year, but a response was
evident after 3 and 5 years. Those soils with a moderate to low extract
able P by H2O but high by Bray 1(3) methods--Blanton, Kers-haw, and Lake-
land--showed the opposite trend with some response in the early growth
period which disappeared with increasing time. Response occurred in the
early growth period and persisted with time on soils with low extractable
P by both H2O and Bray 1(3) methods--Bladen, Marlboro, and McLaurln--
while on the one soil with moderate to high extractable P by H^O and Bray
1(3) methods--Plummeino response occurred at all during the first 5
years. Apparently the amount of water-soluble P (intensity factor) Is
of importance during the early establishment and growth of the pine seed
ling while the amount of solid-phase P in equilibrium with soil-solution
P (quantity factor) becomes more important to tree growth with increasing
time of persistence on the site.
The response trends discussed above are well illustrated by the
effectiveness of P extracted by various soil-test methods at predicting
relative heights in the greenhouse (Appendix Table 50) and the field
2
(Appendix Table 51). The effectiveness (R ) of soil-test methods as
predictors of relative height at any age appeared to be closely related
to the amount of P extracted, but independent of the type of extractant
used where similar quantities of P were extracted (Fig. 1). In view of
this, the results will be Illustrated and discussed principally using
those obtained with some of the more commonly used soil-test methods


163
Al extracted by the other reagents were all very similar despite large
difference in the amounts of A1 extracted, suggesting that they are all
extracted essentially the same form of A1 but with different degrees of
efficiency. It appears reasonable to assume, in view of the relatively
small differences in mean A1 extracted by oxalate (amorphous) and CDB
(amorphous and crystal 1ine),that the dominant form of A1 in most of these
soils is amorphous, excluding that in phyl1 os i 1icate clays.
Correlations between P retention and Fe extracted by the various
reagents differed markedly. Unlike the situation for extractable A1 where
the correlation coefficients were essentially independent of amounts ex
tracted, correlations between P retention and extractable Fe tended to
increase with the amount of Fe extracted up to an optimum for the oxalate
extraction (amorphous) and then declined for the stronger CDB and pyro
phosphate extractions. There appears to be two possible explanations for
this phenomenon. One possibility is that extractants that were least ef
ficient in removing Fe did not dissolve all the active Fe fraction in
volved in P retention while the more efficient reagents extracted other
Fe forms in addition to those active in P retention, such as crystalline
oxides and interlayer Fe (Ramulu, Pratt, and Page, 1967). Secondly, the
amounts of Fe extracted by the oxalate reagent may be more closely cor
related with all soil components involved in P retention than are amounts
extracted by the other reagents. The presence of appreciable crystalline
Fe in these soils, indicated by the relatively large CDB/oxalate ratio
(McKeague, 1967), and the highly significant correlations between amor
phous Fe and all other forms of A] and Fe in the soil (Table 4l) suggest
both above explanations are quite possible.


Fig. 8. Relationship between depth to limiting horizon(X) and relative
height(Y) of slash pine 5 years after fertilization.
ro


APPENDIX


138
Table 32. Relationships between soil-test values at three soil depths
and height of slash pine in the absence of P fertilizer
after 1, 3, and 5 years' growth on 72 field sites
Soi 1-test
So i 1
Height
method
depth
1 year
3 years
5 years
.... r2
H 0
0-20B
0.370**
0.297**
0.193**
2
0-20
0.369**
0.324**
0.209**
20-40
0.195**
0.148**
0.062*
40-60
0.139**
0.094**
0.033
NH^OAc
0-20B
0.189**
0.179**
0.086*
0-20
0.193**
0.195**
0.084*
20-40
0.040
0.045
0.014
40-60
0.033
0.058
0.017
HC1-H.S0,
0-20B
0.024
0.087*
0.072*
4. H
0-20
0.012
0.082*
0.063*
20-40
0.001
0.022
0.009
40-60
0.030
0.069
0.038
Bray 1(3)
0-20B
0.003
0.046
0.042
0-20
0.002
0.049
0.040
20-40
0.024
0.015
0.001
40-60
0.076*
0.053*
0.018
Bray 2
0-20B
0.015
0.007
0.014
0-20
0.018
0.006
0.010
20-40
0.008
0.004
0.001
40-60
0.007
0.044
0.023
* Sign i ficant at
the
5% level,
using the model
Y = b logX + c.
**Significant at
the
1 % 1 eve 1.


Table 55. Continued
Field
s i te
A18
A19
A20
Rri-
Rep 1 i- H2Q NH^OAc l^SO^ Bray 1(3) Bray 2 NH^OAc pH
Depth' P P P P P Ca M K AT (HO)
cm ppm
3
40-60
0.1
0.2
1.3
3.3
2.8
15
34
2
76
4.9
1
0-20B
9-0
5-3
10.0
10.7
10.6
155
88
15
6
3.8
0-20
9-1
6.6
7.5
11.4
8.6
105
82
18
5
3-8
20-40
1 .0
1.1
1.3
1 .8
1.5
60
38
2
2
4.5
40-60
0.5
0.7
1.1
1.8
1.9
30
25
2
3
4.4
2
0-20B
6.6
5-7
8.5
8.5
8.9
135
88
15
9
3.8
0-20
6.0
5.3
7-3
7.0
8.1
105
61
12
9
3-9
20-40
1 .0
1.1
1.5
2.3
2.3
45
42
2
9
4.4
40-60
0.3
0.7
1 -3
3-1
2.4
15
30
2
66
4.5
3
0-20B
4.5
5.3
5-5
7.6
7.1
170
110
18
6
3-8
0-20
4.2
4.4
4.9
6.1
6.5
90
72
10
5
3.9
20-40
1.1
1.1
1.2
1.8
2.5
45
42
2
5
4.3
40-60
0.8
1.3
3.8
9-5
9.2
15
34
2
64
4.3
1
0-20B
0.3
0.9
1.9
3-3
4.8
30
34
7
86
4.3
0-20
0.2
0.7
1.3
1.9
4.5
15
34
4
69
4.4
20-40
0.2
0.5
0.6
0.9
1.6
15
30
2
40
4.8
40-60
0.2
0.2
0.2
0.3
0.4
45
25
2
14
5-1
2
0-20B
0.3
1.1
2.6
4.1
5.6
15
25
7
78
4.3
0-20
0.2
0.7
2.2
3-5
4.8
15
38
7
/
69
4.6
20-40
0.2
0.2
0./
1.0
1.7
15
38
2
52
4.7
40-60
0.2
0.2
0.4
0.5
0.9
15
34
2
55
4.7
3
0-20B
0.2
0.7
1.8
2.5
3-7
15
34
7
73
4.4
ro
ro
VJT


Table 14. Solubility of P compounds in chemical extractants in the presence and absence of two soils
Soi 1
Soil-test method
H2O
NH/jOAc
T ruog
HC1-H2S0Z,
01 sen
Lactate
Bray 1
Bray 2
N H *4 F
_ 0/
'b
Monocalcium phosphate(MCP)
None
100.0
100.0
100.0
100.0
90.6
100.0
100.0
100.0
90.8
1mmoka1ee
97.6
89.6
100.0
95-3
79.1
90.9
95.7
97.3
95.6
McLaurin
*0.9
50.5
78.5
72.4
69.7
79. 1
90.8
97.1
92.2
Dicalcium phospha
te(DCP)
None
13.2
96.7
100.0
100.0
39^8
100.0
100.0
100.0
86.0
1mmoka1ee
91 .2
72.6
98.0
87-3
1.9
87.4
88.5
93.3
88.9
McLaurin
38.4
50.7
85.0
69.1
3.0
11.1
83.0
86.6
17.4
FIuorapatite(FA)
None
0.3
4.6
29.5
77.0
0.2
42.5
16.5
51.5
0.1
1mmokalee
3.2
0.6
25.5
63.0
0.2
12.8
0.8
35-9
1.9
McLaurin
0.2
0.1
8.6
49-4
0.1
7-9
0.3
17.8
0.2
Colloidal a
luminum phosphate(CAIP)
None
2.3
1 .9
6.3
35.8
12.7
67.1
93-7
93-6
87.8
1mmoka1ee
15.7
3-7
7.8
35.2
13.6
39.1
90.7
85.8
79.2
McLaurin
9.8
0.9
4.6
23.6
4.5
39-9
80.2
84.1
71 .9
Potassium taranaki te(KTK)
None
*.0
4.4
4.3
23.8
11.4
7-3
97.0
97.5
99.0
1mmoka1ee
3.7
3.4
3.0
7.8
8.2
4.2
86.7
83.8
79-2
McLaurin
0.2
0.8
2.7
4.7
2.8
2.2
84.6
87.7
80.3
Wavel1 i te(WA)
None
0. 1
0.1
0.1
0.1
0.2
0.5
8.2
7-8
1.9
1mmoka1ee
0.3
0.1
0.2
0.8
0.1
0. 1
6.9
7.6
3.5
McLaurin
0. 1
0.1
0.5
0. 1
0.2
0.1
3.1
2.3
0.1
Colloidal
ferric phosphate(CFeP)
None
2.6
4.2
2.2
17.4
29.1
49.3
90.5
97.5
8.1
1mmokalee
19.8
4.8
2.4
22.4
51.7
88.2
91 .2
85-3
57.0
McLaurin
17.9
2.2
4.2
18.5
40.4
79.7
65.8
66.7
64.1
Strengite(STR)
None
0.1
0. 1
0.1
0.1
0.3
0.1
0.1
0.1
0.2
1mmokalee
0. 1
0.1
0.1
0.5
0.4
0. 1
0.5
1.7
0.4
McLaurin
0. 1
0.1
0. 1
0.1
0.2
0.1
1 1
0.4
0.2


147
foliar P are given in Table 37- Actual P concentrations in the foliage
collected from the control plots of each replication of each trial are
presented in Appendix Table 59-
The values for the relationship between height and foliar P
increased with increasing age of trees. This can probably be attributed
to two factors. First, P concentrations in the foliage of 4-year-old
trees was shown earlier to be a function of the quantity factor of soil-P
supply, while height growth in the first year was related to the intensi
ty factor of soil-P supply, with the quantity factor becoming more
important over longer growth periods. Second, because of internal cycl
ing of-P within trees, one would anticipate a closer relationship between
foliar P and height of trees subsequent to the collection of foliage
samples. Foliar samples were collected when the trees were 4 years old.
The significant relationship between foliar P and height of 5-year-old
trees tends to support the earlier contention that soil P does contribute
to the variability in height of the older trees.
Foliar P was more closely related to relative height than soil P
extracted by any of the soil-test methods. This is in agreement with the
findings of other workers (Wells et al. 1973)* Since foliar P is a more
direct measurement of P available to trees, which integrates all factors
of soil-P supply including depth and time functions, this is perhaps not
surprising. Models for the relationships between relative height at age
1, 3, and 5 years and foliar P are shown in Fig. 12. Foliar~P values
corresponding to a relative height of 90% at ages 1, 3, and 5 years were
in the range 0.085-0.095%, which corresponds almost identically to the
critical range proposed as the best current estimate in the review of
literature. Although it is not the objective of this study to calibrate
foliar-P levels against growth and response of slash pine, the critical


P-retention capacity. On soils with virtually no P retention (Spodosols)
they recommended the use of RP or other slowly soluble sources, whilst on
soils of low to intermediate retention capacity they suggested a mixture of
a soluble source, to provide initial high availability, and a slowly
soluble source to maintain the P supply in later years (5 to 15 years).
For soils with high P-retention capacity they suggested either repeated
applications of a soluble source or banded application of a suitable mix
ture of soluble and slowly soluble P sources.
Pritchett and Smith (197*0 recently reported the growth and response
data obtained 10 years after application of fertilizer in the experiment
on the site with the greatest P-retention capacity (Bladen fine sandy
loam). Their results showed the classical trend anticipated from a long
term comparison of a soluble and slowly soluble source on a highly P-
retentive acid soil: Initial response was greatest from OSP, but after 8
years the response from an equivalent rate of RP was equally great. Cur-
i
rent annual increment during the ninth year showed RP to have a consider
able advantage over OSP. This was reflected in the foliar P levels,
because only the trees in the RP treatments had foliar P levels above the
critical response range of 0.085-0.090%.
Although RP has been used successfully in many field fertilizer
trials, its use in operational fertilization has been restricted by its
bulk, which adds significantly to transportation costs, and the difficulty
in spreading these finely ground mater¡als--parti cu lar 1y by air (Pritchett
and Smith, 1969; Conway, 1962).
The success with RP in forest fertilizer trials is generally at
tributed to three factors, namely: (l) The acidity of most forest soils
(Gentle and Humphreys, 1968; Bengtson, 1970), (2) the relatively long


SUMMARY AND CONCLUSIONS
This investigation was designed to develop and calibrate a soil-
test method or methods for predicting the amount, and type of phosphatic
fertilizer, if any, needed at time of planting slash pine on forest soils
in the southeastern lower Coastal Plain.
In a preliminary screening, over 100 different procedures for ex
tracting or characterizing soil P were evaluated as methods for pre
dicting height growth, increased height growth due to P fertilization,
and the amount of P fertilizer required to provide the optimum response
of slash pine on 10 Coastal Plain soils. Height, height response, and P
fertilizer parameters for slash pine were obtained from field and green
house fertilizer trials established on these 10 soils. Field trials
were measured 1, 3, and 5 years after planting and fertilization, while
trees in the greenhouse trial were measured 1 and 2 years after estab-
1ishment.
Soil-test methods which extracted relatively small amounts of P
from the soil provided the best index of .growth and response of slash
pine after 1 year's growth in both the field and the greenhouse. The
effectiveness of these methods as predictors declined when longer growth
periods were considered. These methods involved use of extractants such
as i^O, weak acids (NH^OAc), neutral salts, anion exchange resin, and
very dilute solutions of strong acids at narrow soi1:solution ratios
which removed little of the solid phase components of soil P. Soil-test
179


Phosphate-fractionation studies following P additions and work re
lating P-retention capacity of soils to various soil properties support
the findings discussed above that in acid soils A1 and Fe are the princi
pal soil constituents with which P reacts. The literature on both these
subjects is voluminous and for the sake of clarity and brevity will be
illustrated by work in the southeastern USA.
Yuan, Robertson, and Neller (i960) found in laboratory tests that
water-soluble P rapidly reverted to Al-P upon addition to three acid
sandy Florida soils. The addition of CSP to a Lakeland fine sand in a
short-term lysimeter study was found by Hortenstine (1969) to signifi
cantly increase only water soluble and Al-P fractions. Fiskell and
Spencer (1964) reported that 6 years after the addition of a heavy appli
cation of CSP to Lakeland fine sand, the increase in P fractions followed
the order Al-P>Fe-P>Ca-P. A similar result was reported by Robertson,
Thompson, and Hutton (1966) with a Red Bay fine sandy loam and a Norfolk
loamy fine sand. However, they reported that the ratio of Al-P to Fe-P
decreased with time. They attributed this decrease to uptake of Al-P by
plants and a transformation of Al-P to less soluble and less available
Fe-P. Phosphorus-fractionation studies on seven forest soils in the south
eastern USA (Humphreys and Pritchett, 1971) revealed that 7 to 11 years
after application of OSP the majority of the recoverable P was found in the
Al-P and Fe-P fractions. The distribution between the two fractions varied
between soils. Equal amounts of each fraction were found in a Bladen soil,
while nearly all applied P was recovered in the Al-P form in a Rutlege
soil. Two Kershaw soils were found to be intermediate in behavior between
the Bladen and Rutlege soils. The interpretation of these fractionation
studies must take into consideration the limitations of the fractionation


95
the only ones with a significantly greater P uptake than the STR treat
ment.
The significantly greater P uptake over all P treatments on the
Immokalee soil compared to the McLaurin soil can probably be attributed
to three main factors: (a) First, there was a greater uptake of P from
the untreated Immokalee soil than the untreated McLaurin soil. The Imm
okalee soil (site A16) was not P deficient for young seedlings (Table 7)
while the McLaurin soil was deficient. (b) Second, the Immokalee soil
had a negligible P-retention capacity while that of the McLaurin soil
was relatively large (Table 4). The P-retention ability of the McLaurin
soil probably maintained solution P at low levels. (c) Third, the pH of
the Immokalee soil was lower than that of the McLaurin soil (Table 6).
Since acidic conditions are necessary for the solubilization of the more
basic Ca phosphates (Olsen and Flowerday, 1971), this difference
undoubtedly accounted for the much greater uptake from fluorapatite on
the Immokalee soil than on the McLaurin soil. The complete lack of.
utilization of FA by trees on the McLaurin soil is inconsistent with the
theory that mycorrhizae have the ability to solubilize insoluble
phosphates. The roots of the seedlings at both transplanting and
harvesting were heavily infected with mycorrhizal fungi. This tends to
support the hypothesis that improved P uptake resulting from mycorrhizal
infection is a consequence of the increase in effective root-absorption
surface, which requires that the P be at least partially soluble.
The significant S x P interaction can be attributed, in the main,
to the differences in utilization of FA between the two soils. Also,
the less effective utilization of MCP and DCP compared to KTK, CA1P and
CFeP in the McLaurin soil probably contributed to a small degree to the
interact ion.


Table 58. Continued
Field
Repli-
Ferti
lizer P requirement
s i te
cate
1 year
3 years
5 years
90
95
100
90
95 100 90
95
100
kg/ha
A8
1
0
0
90
0
0
0
0
0
90
2
2
16
50
0
12
62
0
17
85
3
0
0
64
0
0
111
0
7
93
A10
1
0
0
0
0
0
168
0
0
90
2
0
0
0
16
28
56
0
16
55
3
0
0
0
0
0
45
0
8
48
A15
1
10
21
49
0
15
47
2
16
46
2
0
0
0
1
16
48
0
13
45
3
74
82
90
0
63
90
0
55
90
Al
1
6
0
90
16
29
61
8
21
55
2
0
0
90
0
77
90
0
0
90
3
0
0
90
54
73
90
59
115
254
Al 7
1
18
29
58
18
28
55
6
18
49
2
22
36
68
61
76
90
56
74
90
3
23
32
52
27
36
55
24
35
62
Al 8
1
0
0
0
0
67
90
0
20
100
2
0
8
47
19
35
73
11
27
67
3

0
90
50
72
90
0
0
0
Al 9
1
0
35
90
11
21
46
12
22
46
2
0
0
0
67
80
90
61
77
90
3
2
13
44
0
69
90
0
41
90
Ni
x-
U~1


LIST OF TABLES
Table Page
1 FOLIAR P CONCENTRATIONS PRIOR TO FERTILIZATION IN RELATION
TO RESPONSE OF SOUTHERN PINES TO P FERTILIZER 21
2 SOIL-P VALUES (SURFACE HORIZON) OF UNFERTILIZED SOILS IN
RELATION TO RESPONSE OF SOUTHERN PINES TO P FERTILIZER. . 37
3 PROPERTIES OF PHOSPHORUS SOURCES USED IN GREENHOUSE TRIAL 2 50
4 PHYSICAL AND CHEMICAL PROPERTIES OF SOILS USED IN
GREENHOUSE TRIAL 2 52
5 PHOSPHORUS EXTRACTION METHODS 59
6 CLASSIFICATION AND SELECTED PROPERTIES OF 10 SOILS USED
IN GREENHOUSE STUDY 1 69
7 HEIGHT, RELATIVE HEIGHT, P-FERTILIZER REQUIREMENTS, AND
P CONCENTRATION AND UPTAKE OF SLASH PINE SEEDLINGS AFTER
1 AND 2 YEARS' GROWTH ON 10 SOILS IN THE GREENHOUSE .... 70
8 HEIGHT, RELATIVE HEIGHT AND P-FERTILIZER REQUIREMENTS OF
SLASH PINE AFTER 1, 3, AND 5 YEARS' GROWTH, AND FOLIAR P
CONCENTRATION AFTER 4 YEARS' GROWTH IN THE FIELD ON
10 SOILS 71
9 COMPARISON OF THE GOODNESS OF FIT OF FOUR STATISTICAL
MODELS, AS INDICATED BY THE SQUARE OF THE MULTIPLE
CORRELATION COEFFICIENT (R2), RELATING SELECTED TREE
PARAMETERS (DEPENDENT VARIABLE) AND SOIL-TEST VALUES
(INDEPENDENT VARIABLE) 73
10 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND
RELATIVE HEIGHT GROWTH OF SLASH PINE IN FIELD AND
GREENHOUSE EXPERIMENTS ON 10 SOILS 77
11 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND
TISSUE P PARAMETERS OF GREENHOUSE AND FIELD SLASH
PINE GROWN ON 10 SOILS SO
12 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND P-
FERTILIZER REQUIREMENTS OF SLASH PINE IN FIELD AND
GREENHOUSE EXPERIMENTS ON 10 SOILS 82
v i! i


1*5
nary screening, that H2-extractable P Is a good index only of the short
term P supply to trees.
Results from multiple regressions (not shown) of height on depth
to LH, P extracted by HCl-h^SO^ (0-20 cm) and the squares of these terms
showed no significant contribution of HC1-H^SO^-extractab1e P towards ac
counting for variation in height at any age. However, since P extracted
by HC1-H2S0/1 was significantly related to depth to LH, these multiple
regression results cannot be taken to imply that HC1 -H2S0/1-extractable P,
which previous data have shown to provide a good index of the longer
term P supply to trees, does not play a causal role in determining tree
height over longer growth periods. It is likely that depth to LH
provides an integrated index of several factors influencing growth such
as available P within the effective rooting volume (see earlier discus
sion), and soil moisture and aeration conditions.
'Best fit' models derived using the procedure and soil and site
parameters outlined previously are given in Table 36. These models could
be used for prediction purposes, but the extra time and effort required
to measure several parameters may not justify the improved accuracy
obtained over using a single parameter model such as depth to LH. These
models accounted for a fairly high proportion of the variation in height
growth of slash pine. Unmeasured parameters such as genetic variation,
competing vegetation, insect and disease attack, amount and distribution
of rainfall, and temperature ranges and fluctuations would probably
account'for much of the remaining variation. ~ ~ "
Relationships Between Foliar-P Concentrations and Tree Parameters
Multiple correlation coefficients (R2) for the regressions of
height, relative height, and frtilizer-P requirements on 1og-transformed


178
vertically through the spodic horizon, it is unlikely that any of the P
leached from the surface horizons will move below the spodic horizon.
However, even if all the leached P is retained in the spodic horizon, it
is not likely that trees will be able to utilize it because the fluctu
ating water table of such soils confines most of the rooting activities
to above the spodic horizon (White and Pritchett, 1970). A high water
table on these soils may sometimes lead to lateral drainag above the
spodic horizon. This could lead to P movement into stream waters on sites
where recent applications of soluble P sources cannot be retained in the
surface horizons. Obviously, it is in the interests of all concerned to
use less soluble sources of P, such as finely ground rock phosphate, on
such soils whenever P fertilizers are required.


63
Table 5- Continued
Method*
Extractant
pH
Soil:
solution
Extract ion
t i me
min
HC1(7)
0.025N_ HC1
2.0
1 :50
1
HC1(8)
0.005N. HC1
2.3
1 :10
1
HC1(9)
1 1
2.3
1 :50
1
Res i n
Amber lite 1RA-^00 resin (2g)
2g soil
+ 25ml H20
1
Res in(2)
1 1
2g soil
+ 25ml H20
60
Res in(3)
II
2g soil
+ 25ml H20
Mhr)
Resin(4)
II
2g soil
+ 25ml H20
16 (h r)
Resin(5)
II
5g soil
+ 3ml H20
6(days)
* In the
text and tables the methods are
referred
to by these
abbrevia-
tions.
t Reagents following a slash mark were used to adjust the pH of the
reagents preceding the slash to that shown in the pH column.


41
prior to treatment by CRIFF personnel (Pritchett and Smith, 1972). The
samples were collected from undisturbed areas between bedded tree rows
from the 0-20, 20-40 and 40-60 cm depths.
Additional samples, specifically for use in this study, were col
lected in December-January of 1971/72, 4 years after the establishment
of the trials. From the control plot (PoNo) of each replication at each
site, four composite soil samples were collected using a 4-cm diameter,
closed cylinder soil auger. Two composite samples were collected from
the 0-20 cm depth--one from the bedded area and the other from the un
disturbed interbed area. Each sample consisted of 12 cores collected
randomly from within the net plot area. Four soil pits, one in each
quarter of the plot, were dug in the interbed area. During excavation,
two cores were taken at each pit from both the 20-40 and 40-60 cm depths.
The eight cores from each depth were combined to provide a composite
sample for each of these depths.
Bulk soil samples, for use in greenhouse trials (described
below), were collected from 10 trials. The trials were selected on the
basis of soil and P-response information (Pritchett and Smith, 1972) to
provide a wide range of soil-P levels and P-response characteristics.
The bulk samples were collected from the 0-20 cm depth from a control
plot in each of the 10 selected trials. The samples consisted of sub
samples collected randomly from the undisturbed interbed area.
In order to determine the extent of P leaching in the field, ad
ditional samples were collected from 10 selected trials in January-
February, 1973- The trials were selected, on the basis of laboratory
determinations of P-retention capacity, to provide sites with a range
of retention capacities. Samples were collected from the 0-20 cm depth
in the PiNo and P2N0 plots of each replicate. The samples were


243
Table 57* Continued
Field
Repli-
Height
and
year
Relative
height
and year
s i te
cate
1
3
5
1
3
5
A29
1
41.7
-cm
165
317
66.4
%
60.2
69.0
2
39.6
163
305
70.6
66.5
73-9
3
39.9
166
313
63.8
57.8
73-7


22
Table 1. Continued
T ree
age
Foliar P
(Unferti 1ized)
Response to
P ferti 1izer
Reference
yr.
%
1
<0. 1 1
>0.1 1
yes
no
Wells and
Crutchfield (1969)
10
0.11
no
Hoschler, Jones, and
Adams (1970)
1 -bO
<0.095-0.105
>0.095-0.105
<0.13-0.14
>0.13-0.14
yes (90%)*
no (90%)
yes (100%)
no (100%)
Richards and
Bevege (1972 b)
3
<0.10
>0.10
yes
no
Wells et al. (1973)
Associated with 90 or 100% of maximum yield.


Table 63. Continued
Field
s i te
Repli-
cate
Bu 1 k
density
Tota 1
P/20 cm"
P retention'
NHjOAc(pH 4.8)
-A 1
Po
Pi
P2
Pi
P2
Po
Pi
P2
A22
3
g/cc
1 .25
127
Kg/ha
272
65
4o
ppm
50
A23
1
1.18
112
120
144
14
14
12
10
12
2
1.18
168
' -
270
46
122
-
57
3
1.18
125
1*3
185
32
27
55
27
30
* Calculated from total P analyses (ppm), depth (20 cm), and bulk density.
t Calculated by expressing difference in total P/20 cm between control and treated plots as a per cent
of the P applied 4 years previously as CSP.


92
competition for available F between native soil cations and FA-Ca.
Colloidal aluminum phosphate (CAiP)
Water and NH^OAc extracted very small quantities of CAIP. Soil-
test methods involving use of inorganic acid extractants removed
increasing quantities of P with increasing acid strength, but even the
extractant with the lowest pH (HC1-H2S0^) removed considerably less P
than extractants containing complexing agents (lactate, NH^F, Bray 1,
Bray 2). The Olsen method, which released P from A1 and Fe by
hydrolysis, extracted relatively small quantities of P from CAIP; it
rated between the Truog and HCl-H^SO^ methods in extraction efficiency.
Adsorption of P during extraction accounts for the effect of soil on the
P extractabi1ity by all methods except for H^O and NH^F.
The increased water solubility of CAIP in the presence of soil
was not anticipated. The slight reduction in pH from 7 to the vicinity
of A to 5 would theoretically be expected to reduce the amount of P in
solution (Bache, 1963). A possible explanation is that the H^O extrac
tion mobilized organic compounds, through ionization of functional
groups and these formed complexes with A1, thereby increasing the re
lease of P (Schnitzer, 1969). The reduced P recovery by the NH^F
extraction probably resulted from the precipitation of Ca phosphates
since exchangeable soil Ca was not removed prior to extraction, as was
recommended when this method is to be used for determining Al-P (Fife,
1959).
Potassium taranakite (KTK)
The solubility of KTK was almost identical to that of CAIP in all
extractants containing F (Bray 1, Bray 2, NH^F). Likewise, the order of
extractabi 1ity in the inorganic acid extractants (Truog, HC1-H2S0^) and


The transformations of the residual DCPD are greatly influenced
by soil pH. Moreno, Brown, and Osborn (i960) found that DCPD in aqueous
solutions hydrolysed to the more basic octacalcium phosphate (OCP) at a
pH of 6.38 and that the rate of hydrolysis increased at higher pH values.
Bell and Black (1970 a) reported that DCPD reverted to OCP at 35 C in soil
at pH 6.4, but at 25 C a pH of at least 6.9 was required before OCP be
came detectable. Lehr and Brown (1958) found that both MCP and DCPD
applied to alkaline soils (pH 8.0) reverted to OCP and hydroxyapatite.
In acid soils, DCPD dissolves releasing Ca^+ and HP0^ (Olsen and
Flowerday, 1971). The rate of dissolution of DCPD, according to Moreno,
Lindsay, and Osborn (i960) is a function of the rate of removal of Ca and
P from solution by plants and/or the soil sorption and exchange sites.
It has been reported that DCPD persists in acid soils for several months
(Lehr and Brown, 1958). Moreno et al- (i960) found that once the dissolu
tion of DCPD in acid soils was complete, P concentrations in soil solution
decreased abruptly.
The reactions and transformations of RP following addition to soils
have been less subject to research than has MCP. Apatite is the principal
P form in RP, but it varies in degree of crystallinity and proportions of
flor and hydroxyapatite, depending on location of the RP source (Olsen
and Flowerday, 1971). The dissolution of RP in soils is closely related
to soil pH. Lindsay and Moreno (i960) developed solubility diagrams for
predicting P in solution at various pH ranges in the presence of flor
and hydroxyapatite. For f1uorapatite, the equation developed for pre
dicting P ¡n solution was:
p h2poa
-5-13 + 2 pH


LIST OF FIGURES
gure Page
1 RELATIONSHIP BETWEEN R2 VALUES FOR REGRESSIONS OF RELATIVE
HEIGHT AT AGE 1, 3, AND 5 YEARS ON SOIL-TEST VALUES, AND
MEAN AMOUNTS OF P EXTRACTED BY SOIL-TEST METHODS 75
2 PHOSPHORUS UPTAKE BY SLASH PINE SEEDLINGS AFTER 3 MONTHS'
GROWTH ON TWO SOILS TREATED WITH EIGHT P COMPOUNDS .... 96
3 .DRY MATTER OF SLASH PINE SEEDLINGS AFTER 8 MONTHS GROWTH
ON TWO SOILS TREATED WITH EIGHT P COMPOUNDS 97
4 RELATIONSHIP BETWEEN HC1-H-SO^-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 1 YEAR AFTER P FERTILIZATION Ill
5 RELATIONSHIP BETWEEN HC 1 -H2 SO j-EXTRACTA BLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 3 YEARS AFTER FERTILIZATION 112
6 RELATIONSHIP BETWEEN HC 1 -H^O^-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 5 YEARS AFTER P FERTILIZATION 113
7 RELATIONSHIPS BETWEEN HC1-H2SO^-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT OF SLASH
PINE 1, 3, AND 5 YEARS (Y1, Y3, AND Y5) AFTER P
FERTILIZATION 11 4
8 RELATIONSHIP BETWEEN DEPTH TO LIMITING HORIZON (X) AND
RELATIVE HEIGHT (Y) OF SLASH PINE 5 YEARS AFTER
FERTILIZATION 122
9 RELATIONSHIPS BETWEEN HC 1-H-SO.-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND AMOUNT OF P FERTILIZER (CSP)
REQUIRED TO ACHIEVE 30% OF MAXIMUM HEIGHT GROWTH OVER
PERIODS OF 1, 3, AND 5 YEARS (Y1, Y3, AND Y5)
FOLLOWING FERTILIZATION 128
10RELATIONSHIPS BETWEEN HC1-H-SO.-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND AMOUNT OF P FERTILIZER (CSP)
REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT GROWTH OVER
PERIODS OF 1, 3, AND 5 YEARS (Y 1, Y3, AND Y5)
FOLLOWING FERTILIZATION 129
xi V


93
alkaline NaHCO^ (Olsen) was the same as for CA1P, although less P was
extracted. The only major differences in extractabi 1ity between CAIP
and KTK occurred with the lactate method, and the H2O extraction in the
presence of soil. The lower solubility of KTK in lactate probably
resulted from the lower solubility of KTK in weakly acid solutions. Ap
parently at pH 3-7, insufficient dissolution occurred to enable the
complexing capacity of lactate to appreciably alter the dissolution rate
of KTK. A similar explanation, with natural organic compounds instead
of lactate acting as the complexing agents, probably accounts for the
failure of KTK-water solubility to increase in the presence of soil.
Wavellite (WA)
Extractants containing F were the only ones to remove other than
trace quantities of P from WA. If A1 -P compounds of similar stability
to WA exist in the soil, alkaline NH^F will underestimate the amount of
Al-P in the soil.
Colloidal ferric phosphate (CFeP)
This compound was less soluble than CA1P in inorganic acid extract
ants but more soluble in the Olsen extractant. Solubility in both Bray
extractants was similar to CA1P, although the recovery was less in the
presence of the McLaurin soil; probably as a result of F preferentially
complexing with the large amounts of A1 in this soil. Substantial
increases in P solubility in the presence of soil occurred with H2O,
Olsen, NH^F and lactate methods. In the case of the three methods utiliz
ing nonacidic extractants, the increased solubility probably resulted
from ionized organic functional groups complexing Fe (Schnitzer, 1969).
The possibility that soil organic matter and other soil constituents
contributed to solubilization by reducing Fe^+ to Fe2+ cannot be


25
site index studies, and the lack of soil testing techniques of known
significance in relation to tree requirements. However, since the wide
spread demonstration of the responsiveness of trees to fertilizers
(Mustanoja and Leaf, 1965) soil nutrient analysis has received more at
tention as a technique for predicting site productivity (Pawluk and
Arneman, 1961; Wilde et al., 1964 a; Schomaker and Rudoph, 1964; la
Bastide and van Goor, 1970; Alban, 1972).
As a diagnostic technique for predicting site nutrient status and
the need for fertilizer, soil analysis has received considerably less
attention than foliar analysis (Leaf, 1968). According to Leaf (1968),
current thinking on the value of soil and foliar analysis still reflects
the ideas of Mitchell and Chandler (1939) who stated:
The distinct advantage of the method of leaf analysis is that by
chemical analysis of the leaves we can obtain a more reliable
estimate of the amount of the various nutritional elements which
have been adsorbed by, and therefore available to, plants growing
in a given soil. One therefore uses a natural biological rather
than an artificial extraction method for estimating available
nutrients. (p. 75)
Several factors have been cited in the literature as reasons why
soil testing has not been widely accepted as a diagnostic tool by forest
managers. Most of these are related to insufficient information on
methods and procedures.
Sampling procedures
The determination of tree nutrient requirements is complicated by
a deep root system. Tamm (1964) suggested that the main rooting zone
should be sampled thoroughly and consideration should be given to lower
soil horizons when they contain notable root concentrations. In discus
sing this problem, Voigt (1958) pointed out that most forest trees, as
opposed to agricultural crops, have two distinct rooting regions; one


16
15 years, current increment in the RP treatment was significantly greater
than that in the OSP treatment. The decline in effectiveness of the OSP
was attributed to the high P-retention capacity of the soil. Working on
the sandy coastal plain of Western Australia, Hopkins (i960) found RP to
be the most effective source of P for P. pinaster. This was attributed
to less leaching of P from RP than from OSP on these leached, acid,sandy
soils.
In summarizing the results from fertilizer trials in New South
Wales, Gentle and Humphreys (1968) suggested that on soils with a high P-
retention capacity, banded applications of a semi-soluble source should
be used, whi1st on near neutral soils with a medium P-retention capacity,
OSP or CSP would be preferred. On acid soils with a low P-retention ca
pacity they suggested the use of RP.
Rock phosphates are the preferred source of P for fertilizing
forests on acid peats in England (Leyton, 1958), Finland (Salonen, 1967)
and Ireland (Dickson, 1971)- The success of RP is probably due to the
acidity and low P-retention capacity of these peats, although there is no
reported evidence in the literature to indicate leaching losses of soluble
P sources applied to peats. In some instances, however, the selection of
RP was based on economic considerations, since OSP and RP were equally
effective (Dickson, 1971). Hagenzieker 0958) also reported that OSP, RP
and basic slag were equally effective sources on the acid forest soils of
Hoi land.
Bengtson (1970) examined the comparative value of CSP and RP as P
sources for slash pine in greenhouse studies with a Lakeland fine sand
(pH A.8). Using different placements, he found CSP to be superior to
RP in all tests. The performance of RP was improved by mixing it with


relative height.
11
Fig. 7- Relationships between HC1-H^SO^-extractab1e P(X) in the
surface 20 cm of soil and relative height of slash pine
1, 3 and 5 years (Y1, Y3, and Y5) after P fertilization.


50
Table 3. Properties of phosphorus sources used in greenhouse trial 2
Source
Composition
Principal
X-ray peaks
P
Ca
A1
Fe
(%)
(ft
Monocalcium phosphate*
24.6
15.9
-
-
11.74,
3-88,
3.69
D¡calcium phosphate*
22.8
29.5
-
-
3.35,
2.36,
2.72
FI uorapatite +
18.0
39.7
-
-
3.44,
2.80,
2.70
Colloidal aluminum phosphate
17.9
-
14.4
-
-
Potassium taranakitef
19.0
-
10.1
-
15.84,
7.91*
3.81
Wavel1 ite §
10.8
-
16.0
-
8.71,
8.42,
3.22
Colloidal ferric phosphate
13.9
-
-
27.1
-
Strengite+
17.0
-
-
35.5
5.49,
4.37,
3.01
* Analytical grade reagents.
t Provided by coirtesyof the Tennessee Valley Authority.
§ Provided by courtesy of Dr. F.N. Blanchard, Geology Dept., University
of Florida.


Table 46. Corit i nued
T reat-
ment
Rep 1i-
cate
1 year
2 years
t
Ht.
Dry wt.?
tops
P in
tops
t
Ht.
Dry
Tops
wt.
Roots
Tops
P
Roots
cm
g
f
cm
g
~%
McLaur¡n
fine sandy
loam (A28)
1
17.0
17.3
0.097
33-8
87.1
23-5
0.076
0.096
P 2
2
17.0
15.5
0.087
36.0
80.3
27.0
0.089
0.084
3
15-0
18.1
0.117
33-5
77.1
26.5
0.082
0.084
I
18.1
22.0
0.106
34.0
78.9
28.4
0.094
0.097
P 3
2
18.9
22.8
0.115
38.3
86.5
27-9
0.088
0.104
3
14.6
23.3
0.107
36.5
88.7
23.6
0.084
0.103
Immokalee fine sand (A 16)
r>
1
14.6
18.7
0.088
28.5
64.2
21 .8
0.040
0.050
Po
2
15.8
20.5
0.094
29.8
65.0
17.2
0.042
0.057
3
14.1
16.1
0.107
30.0
65-5
20.4
0.035
0.048
1
14.6
15.4
0.192
26.8
76.2
22.2
0.122
0.159
Pi
2
14.1
19.0
0.228
28.8
65..0
24.5
0.117
0.090
3
13.9
16.6
0.217
31.8
62.1
19.1
0.111
0.142
1
13-5
17.0
0.305
27.0
64.2
21.2
0.105
0.149
P2
2
15.0
18.0
0.320
29.0
61.0
18.8
0.152
0.163
3
13-6
16.3
0.250
35.3
67.1
29-0
0.104
0.184
1
14.0
17.5
0.360
32.5
60.0
20.7
0.136
0.159
!p3
2
15.5
15.8
, 0.388
28.0
69.I
21.0
0.190
0.148
3
12.3
11.2
0.340
25.0
53.6
24.5
0.161
0.290
V>
o


108
Table 18. Simple correlation coefficients(r) between amounts of P
extracted from the surface 20 cm of soil(bedded area) by
five soil-test methods
Methods H20 NH^OAc HCl-^SO^ Bray 1(3) Bray 2
r
H20
1.000
0.684**
0.081
-0.058
-0.077
NH^OAc
1 .000
0.721**
0.606**
0.587**
HCl-H2S02j
1.000
0.972**
0.968**
Bray 1(3)
1.000
0.996**
Bray 2
1.000
Mean extractable
P, ppm
1.76
2.39
6.96
17.93
20.82
** Significant at the 1% level.
Table 19- Separation of 72 field sites into response quadrants using
the technique of Cate and Nelson (1965) and a critical HC1 -
i^SO^-extractable P value of 5 ppm
T ree
HC1-H2S04 < 5
ppmP
HC1-HjSO^ > 5
ppmP
Correct*
Age
Response No
response
Response No
response
prediction
1
34
12
9
17
%
70.8
3
34
12
6
20
75.0
5
27
19
6
20
65.3
* Correct prediction = Response(<5ppmP) + No response(>5ppmP) x ]0(J
Total number of sites


176
capacity (20-40 ppm of NH^OAc-Al), partially acidulated rock phosphates
containing a small proportion of soluble phosphate might be the most suit
able source. This source provides initially available P at a level small
enough to be retained on the low number of retention sites plus a slowly
soluble source to maintain available P over an extended period of time.
Such sources may also be of value on sites with very high P-retention
capacity where long-term utilization of soluble sources is restricted by
fixation of P in difficultly available forms (Humphreys and Pritchett,
1971). Fertilization of soils with negligible P-retention capacity
(NH^OAc-Al < 20 ppm) should be with slowly soluble sources, since any
soluble P wi11 leach in these soils. Use of soluble sources on low-
retention sites should be restricted to situations where frequent small
applications can be economically used, such as in nursery operations.
Data in Fig. 16 showed that on some soils (Spodosols) fertilized
plots contained less total P than did control plots, a phenomenon also
shown by the results of Humphreys and Pritchett (1971) for certain
Spodosols. The degree of vertical and lateral P movement may determine
both the ability of the tree to ultimately utilize the P lost from the
upper portions of the profile and also the extent of any adverse environ
mental effects. Very little, if any, of the leached P is likely to be
retained in the A2 horizons of the Spodosols (Humphreys and Pritchett,
1971), but the spodic (Bh) horizons of these soils have very high P-
retention capacities. This is shown in Fig. 17 comparing P retention
from the addition of 2,500 pig P/g soil in the A and Bh horizons of several
Spodosols used in this study. The P-retention capacity of some spodic
horizons was greater than that for any surface horizons of soils shown in
Appendix Table 59. In the event that drainage in these soils occurs


P adsorbed, /jg/g soil
Fig. 14. Phosphorus-adsorption isotherms of four Lower Coastal riain
soil types representative of four soil orders.


Table 49. Continued
Soi]-test Soi1 (CRIFF identification code)
method ~~ A2 B A21 A24 A28 A16 A19 A23 A5 ~A25
ppm-
01 sen(6)
1.7
21.5
63.0
1.1
0.9
7.9
8.5
3-9
10.6
41.3
01 sen(7)
0.5
23.5
66.5
0.5
0.5
5.9
7-9'
2.9
10.9
40.3
01 sen(8)
0.7
25.5
64.0
1.1
0.6
6.4
8.6
4.3
11.3
44.6
01 sen(9)
0.8
28.9
87.0
1.3
1.0
6.2
8.8
3.8
12.0
58.3
Bray 2(2)
2.9
46.5
178.5
1 .8
1.9
5.4
6.6
3.8
23.6
125.5
Bray 2(3)
2.2
52.3
232.0
1 .9
1.4
5.8
7.9
4.6
27-5
152.0
Bray 2
2.7
45.6
193.8
2.0
2.0
3.7
4.6
2.7
23.6
135.0
Bray 2(4)
3.1
55.5
257.0
2.8
2.6
4.7
5-5
3.2
31.2
170.4
Bray 2(5)
5-7
49.1
215-9
4.0
3.4
5.5
6.0
5.6
27.1
150.8
Bray 2(6)
7.1
55.1
303.7
4.5
4.8
5.8
6.5
5.1
33.1
183.4
Bray 1(2)
1.9
41.6
156.0
1.8
1.7
5.2
6.7
4.0
21 .4
104.0
Bray 1(3)
1.0
48.9
185-5
1 .6
1.5
6.6
7.9
4.7
25.9
130.0
Bray 1
2.7
43.6
187.8
2.1
1.5
3.8
4.6
2.6
23.3
128.6
Bray 1 (4)
1.7
50.6
218.9
1.9
2.0
4.8
6.0
3.0
29-5
152.8
Bray 1 (5)
3.8
47.6
198.3
2.6
3.0
4.5
4.8
2.9
24.6
145.7
Bray 1(6)
6.8
53.6
256.0
4.7
5.5
6.3
7.1
5-2
30.6
182.2
Bray 3
2.5
40.9
162.2
2.2
1.9
3-5
4.5
2.5
22.9
115.1
Bray 3(2)
4.0
43.8
175.7
3.0
2.6
4.1
5.0
3.3
23.5
134.7
Bray 4
1.8
34.3
121 .3
1.7
1.5
4.3
5.0
2.5
18.1
96.0
Bray 4(2)
3.5
41.5
171.3
2.9
2.8
4.8
5.5
2.8
20.6
117.8
Bray 5
3.6
52.5
234.0
2.9
3.0
4.0
4.9
2.6
28.8
162.0
Bray 5(2)
4.9
50.5
235.0
3.4
4.1
5-4
6.3
3.9
30.8
165.0
Bray 6
3.4
44.0
207-5
3.0
2.7
5-8
6.2
3.7
26.7
140.5
Bray 6(2)
4.3
53.8
230.0
3-5
3.7
6.8
7.4
5-5
25.0
150.0
Bray 7
2.5
41.6
154.4
1.9
1.8
6.1
7.5
4.6
20.6
115.0
Bray 7(2)
3.1
46.0
207.5
2.6
2.6
6.6
7.9
5.1
21.1
139-0
Bray 8
3.6
35.7
135.0
]. 6
1.2
5.7
6.6
4.5
19.7
100.0
U>


FERTILIZER REQUIREMENT(90%), kgP/ha.
128
Fig. 9- Relationships between HC1-f^SO,-extractab1e P(X) in the
surface 20 cm of soil and amount of P fertilizer (CSP)
required to achieve 90% of maximum height growth over
periods of J, 3, and 5 years (Y1, Y3, and Y5)
following fertilization.


37
Table 2. Soil-P values (surface horizon) of unfertilized soils in
relation to response of southern pines to P fertilizer
T ree
age
Soil test
type
Soi 1 P
Response to
P ferti 1izer
Reference
yr.
ppm
(A) Slash pine
(Pinus e11iottii)

9
Total P
<52
yes
Young (1948)
>52
no
21
Total P
<65
yes
Richards (1956)
>65
no
3+
Total P
<70
yes
Baur (1959)
>70
no
9
0.05N HC1 +
29.5
no
Walker and Youngberg
0.025N HjSOj,
(1962)
5-8
NH^OAc(pH 4.8)
<2.0
yes
Pritchett (1968)
>2.0
no
Pritchett and
Llewellyn (1966)
(B) Loblolly pine
(Pinus taeda)
9
Total P
<59
yes
Young (1948)
>59
no
21
Total P
<91
yes
Richards (1956)
>91
no
6
0.03N NH.F
7
no
Merrifield and Foil
+ 0.1NTHC1
(1967)
6
0.05N HC1 +
8
no
Carter and Lyle
0.025^1
(1966)
10
0.05N HC1
4.5
no
Moschler et al.
+ 0.025N H2S04
(1970)
3
0.05N HC1
<3-0
/yes
Wei 1s et a 1. (1973)
+ 0.025N H S0^
>3.0
no


Table 56. Continued
Field
s i te
Repli-
cate
Depth
Particle
size distribution
Moisture
content
Bu 1 k
density
C1 ay
Silt
Sand
1/3 atm.
15 atm.
cm
g/cc
A28
2
40-60
16.4
20.9
62.7
11 .8
5.2
1.54
3
0-20
6.5
24.4
69.1
9-1
3.8
1.20
20-40
14.0
21.5
64.5
9.3
4.6
1.63
40-60
15.3
.21.6
63.1
10.4
5.2
1.60
A29
1
0-20
10.2
32.4
57.4
16.3
5-6
0.96
20-40
1 1.0
34.0
55.0
11.8
3-7
1.53
40-60
12.3
32.3
55.4
13.2
4.3
1.69
2
0-20
5.1
22.8
72.1
8.8
2.9
1.14
20-40
6.1
22.7
71.2
5-9
1.8
1.48
40-60
6.9
22.8
70.3
7.5
2.0
1.52
3
0-20
5.8
23-9
70.3
8.6
3-8
1.02
20-40
7.0
22.3
70.7
7.8
3.2
1.53
40-60
7.3
22.5
70.2
6.7
2.1
1.67
o


99
Relationships Between Seedling Utilization and Solubility of P Compounds
Multiple correlation coefficients (R ) for relationships between
seedling utilization of P compounds and solubility of these compounds in
chemical extractants, in the absence and presence of soil, are given in
Table 15. Also included are R2 values for the relationships between
seedling utilization and the extractabi 1ity of the P compounds following
2 months of incubation in the two soils. The extractabi 1ity of the P
compounds by various soil-test solutions following 2 months' incubation
in the two soils, are presented in Table 16.
The R values for the Immokalee soil were generally higher than
those for the McLaurin soil. This was particularly pronounced for acidic
extractants (NH^OAc, Truog, HCl-h^SO^, lactate, Bray 1, Bray 2). The
solubilization of FA by these extractants, which was not reflected in its
availability in the McLaurin soil as it was in the Immokalee soil,
probably accounts for this difference. For groups of soils which contain
native or added FA (rock phosphates) which is unlikely to be available
for plant use due to a lack of soil acidity, soil-test methods using
neutral or alkaline extractants are likely to provide a better index of P
availability than acid extractants.
In general, P extracted from incubated samples was more closely
related to P uptake than was P extracted from the compounds alone or in
the presence of soil. This was expected since P uptake over 8 months in
the greenhouse was influenced by the reaction between the applied P com
pounds and the soils, and this reaction in turn influenced the extract-
ability of P compounds from the incubated soils.
The R2 values using extraction data from incubated samples of both
soils showed an almost identical ranking for soil tests to that found for
the greenhouse data in the preliminary screening of soil-test methods


This dissertation
College of Agriculture
as partial fulfillment
Phi 1osophy.
was submitted to the Graduate Faculty of the
and to the Graduate Council, and was accepted
of the requirements for the degree of Doctor of
August, 197^
Dean, Graduate School


185
8. The P supply of soils with a pH > 5 which contain either fluorapatite,
or added rock phosphate will be overestimated by the three above
soil-test methods.
9. The height growth of slash pine in the absence of P fertilizer on
lower Coastal Plain soils (with no N deficiency) is determined by a
combination of the soil-P supply and moisture availability. Depth
to limiting horizon, which provides an integrated index of these
two factors, is the best single variable for predicting height
growth.
10. Excess leaching losses of P applied as a soluble fertilizer such as
CSP can be anticipated on soils containing less than bO, 120, 300,
or ^00 ppm of A1 extracted by the NH^OAc, HC1-H2SO^, Bray 1 and Bray
2 methods, respectively.
11. Slowly-soluble P fertilizers, such as rock phosphate, should be used
if P fertilization is required on soils containing less extractable
A1 than the proposed critical levels. However, rock phosphate
should not be used on soils with a pH > 5.
It can be concluded from the above that carefully selected and
calibrated soil tests can serve as effective diagnostic aids for predict
ing the amount and type of phosphatic fertilizer, if any, needed at
planting time to achieve optimum growth of young slash pine plantations
in the lower Coastal Plain region.


Table 50. Multiple correlation coefficients (R^) for relationships between soil-test values and seedling
parameters from greenhouse study 1
Mean
So!1-test
extract-
Height
Rel.
height
Fert.
reqm.
%P in
tops
P uptake
method
able P
l"
2
1
2
1
2
1
2
1
2
ppm
R2
k2o
2.3
0.411T
0.506
0.951
0.715
0.897
0.703
0.537
0.150
0.810
0.322
H20(2)
7.1
0.160
0.144
0.497
0.546
0.465
0.426
0.135
0.002
0.428
0.003
H20(3)
0.3
0.350
0.705
0.612
0.813
0.789
0.809
0.699
0.274
0.556
0.401
H20(4)
243.8
(yg p)
0.258
0.446
0.658
0.853
0.715
0.763
0.464
0.086
0.577
0.150
NaC 1
2.2
0.279
0.572
0.750
0.908
0.818
0.889
0.657
0.213
0.642
0.303
Ma2S0^
2.3
0.301
0.587
0.769
0.905
0.837
0.887
0.667
0.217
0.665
0.320
NaoMoO/f
7.5
0.404
0.478
0.562
0.319
0.475
0.395
0.618
0.544
0.562
0.802
2.4
0.284
0.456
0.773
0.835
0.807
0.796
0.527
0.120
0.664
0.206
Na2B407(2)
15.0
0.243
0.673
0.626
0.666
0.649
0.806
0.073
0.740
0.513
0.814
MH4OAC
4.4
0.307
0.570
0.863
0.729
0.851
0.863
0.847
0.607
0.709
0.454
N'Hi(0Ac(2)
4.7
0.328
0.564
0.872
0.731
0.904
0.857
0.807
0.542
0.740
0.372
NH/jOAc (3)
4.6
0.222
0.526
0.812
0.692
0.849
0.845
0.769
0.481
0.621
0.359
HOAc
10.3
0.297
0.630
0.757
0.599
0.769
0.765
0.901
0.616
0.633
0.787
HOAc(2)
7.6
0.365
0.647
0.775
0.613
0.806
0.750
0.856
0.541
0.696
0.761
HOAc(3)
4.9
0.307
0.639
0.835
0.815
O.863
0.922
O.878
0.452
0.688
0.592
HOAc(4)
3.9
0.328
0.629
0.847
0.785
0.886
0.895
0.847
0.412
0.701
0.580
Lactate
16.0
0.263
0.638
0.634
0.536
O.689
0.712
0.930
0.712
0.545
0.850
Lactate(2)
9-7
0.332
0.639
O.689
0.577
0.716
0.733
0.917
0.649
0.625
0.839
Lactate (3)
13.3
0.299
0.679
9.684
0.679
0.760
0.833
0.956
0.624
0.604
0.776
Lactate(4)
8.0
0.351
0.693
0.804
0.706
0.833
0.841
0.906
0.532
0.688
0.741
Citrate
32.9
0.246
0.592
0.514
0.468
0.572
0.633
0.927
0.777
0.452
0.884
202


48
Greenhouse Trial 2
This trial was established for the purpose of examining the abili
ty of slash pine seedlings to utilize different P compounds. Uptake of P
was determined for 6-month-old slash pine seedlings growing in two soils
to which the P compounds were added. This P uptake was related to the
solubility of the P compounds, alone or after mixing with the above two
soils, in chemical extracts used as soil-P tests.
P Compounds
The eight P compounds used in the trial, and their chemical compo
sition, are shown in Table 3- The composition of all compounds was
checked by determining Ca, Fe, A1 and P in solution following dissolution
of the compounds in hot 6N HC1. Procedures used in the chemical analysis
are outlined in a later section. The compounds were also checked by X-
ray analysis using a General Electric XRU-7 instrument with Ni-filtered
CuKa radiation. The principal peaks of the six crystalline compounds,
shown in Table 3 confirmed their identification according to published
standards (Lehr et al., 1967). The colloidal A1 and Fe phosphates were
prepared by the procedures outlined by Deming and Cate (1963) and Cate,
Huffman, and Deming (1959), respectively. The X-ray analysis confirmed
that these prepared compounds were amorphous.
Monocalcium phosphate is the principal P form in most phosphatic
fertilizers. Dicalcium phosphate, colloidal Al and Fe phosphates, K
taranakite, and strengite were used in this study because they have all
been identified as soil-P fertilizer reaction products (Lindsay, Frazier,
and Stephenson, 1962). Fluorapatite is a common primary phosphate
mineral in some soils and may also be formed as a soil-P fertilizer
reaction product following reversion of less basic Ca phosphates in the


35
on which trees will respond and those on which trees will not respond to
fertilization (Wells et al., 1973). According to Mader (1973), the
paucity of well-designed fertilizer trials, suitable for use in calibrat
ing soil-test results against response over a wide range of conditions,
is the major reason for the lack of suitably calibrated soil tests. How
ever, this situation has been alleviated to some extent by tne establish
ment of a large number of uniform fertilizer trials by forest fertiliza
tion cooperatives based in Florida, North Carolina and Washington State
(Bengtson, 1972).
Information published on the response of southern pines in rela
tion to soil-P levels is summarized in Table 2. Early work In Australia
showed that total soil-P values could be used for delineating sites
responsive to P fertilizers (Young, 1948; Richards, 1956; Baur, 1959)-
However, as discussed earlier, these relationships were found to hold
true only over limited areas. This is exemplified in Table 2 by the dif
ferent critical values reported by the various authors. The only attempt
at using total-P values for other than delineating unresponsive sites was
that reported by Baur (1959). He found that the quantity of P fertilizer
required to produce adequate growth in the field corresponded closely
with the quantities calculated to raise total P to the optimum of 70 ppm
P (based on 125 kg/ha of superphosphate being equivalent to 6 ppm P). The
studies reported by Pritchett (1968) and Pritchett and Llewellyn (1966)
for slash pine, and Wells et al. (1973) for loblolly pine are the only
ones in which relationships were derived statistically using a range of
sites. These studies were conducted in the Coastal Plain region of the
southeastern USA. Pritchett and Llewellyn (1966) and Pritchett (1968)
reported that 2.0 ppm P by NH^OAc (pH 4.8) extraction separated sites
responsive to P fertilizer from unresponsive sites. This test proved


FOLIAR
2 4 6 8O 12
HCI-H2SO4 EXTRACTABLE P, ppm
Relationship between HC1 -HS0;^-extractab 1 e P(X) in the surface 20 cm of soil and
P concentration in foliagefY) of 4-year-old slash pine.


23
concentration would normally respond. This definition corresponds to
the 'optimum concentration1 proposed fay Richards and Bevege (1972 a)
who defined the 'critical concentration' as the concentration asso
ciated with 90% of maximum yield.
Richards and Bevege (1972 a) pointed out that critical levels can
be established with accuracy only where all nutrients apart from the one
under consideration are not limiting. However, Leyton and Armson (1955)
suggested that difficulties in interpretation caused by nutrient inter
actions could be reduced by extending investigations over a wide range of
conditions to allow for a statistical analysis of the relationships. In
formation in Table 1 shows that where a number of trials and statistical
analyses were used, there is good agreement between the proposed critical
levels for slash pine (Pritchett, 1968; Richards and Bevege, 1972 b) and
loblolly pine (Wells and Crutchfield, 1969; Richards and Bevege, 1972 b;
Wells et al., 1973)- These data suggest that the best current estimates
of critical foliar P concentrations (Pritchett, 1968) for P. el 1iottii
and P. taeda are in the range 0.085-0.10% and 0.095-0.105% respectively.
The failure of foliar P values to provide the correct diagnosis in some
of the single trials reported in Table 1 can probably be attributed to
other factors such as other nutrients or site factors limiting production
or to differences in sampling procedures (Leaf, 1973).
In studies where both foliar and soil analysis have been used to
predict response to P fertilizer, foliar analysis has generally proved to
be the more effective (Pritchett, 1968; Wells and Crutchfield, 1969;
Wells et al., 1973). However, it has been pointed out that foliar analy
sis has practical disadvantages as a diagnostic tool in forestry. Its
use is largely limited to areas where trees are established (Pritchett,
(1968). Restricted sampling periods and the time required to collect an


Table 50. Continued
Mean
Soi1-test
extract-
Height
Rel. he
i ght
Fert.
reqm.
i n
tops
P uptake
method
able P
1*
2
1
2
1
2
1
2
1
2
ppm
Bray 5
49.8
0.073
0.359
0.184
0.202
0.204
0.343
0.679
0.849
0.200
0.807
Bray 5(2)
40.9
C.087
0.382
0.198
0.205
0.218
0.345
0.681
0.841
0.175
0.799
Bray 6
44.3
0.113
0.435
0.250
0.266
0.279
0.418
0.748
0.841
0.226
0.827
Bray 6(2)
49.0
0.137
0.465
0.264
0.260
0.298
0.407
0.751
0.845
0.241
0.862
Bray 7
35-6
0.162
0.532
0.361
0.374
0.409
0.538
0.855
0.843
0.322
0.868
Bray 7(2)
44.1
0.168
0.516
0.329
0.323
0.362
0.476
0.811
0.837
0.302
0.883
Bray 8
31.4
0.139
0.490
0.369
0.338
0.422
0.528
0.841
0.816
0.315
0.846
Bray 8(2)
41.2
0.170
0.514
0.333
0.335
0.377
0.489
0.818
0.833
0.313
0.974
HC1
15.4
0.263
0.642
0.699
0.647
0.755
0.827
0.961
0.628
0.580
0.762
HC1(2)
6.1
0.289
0.623
0.729
0.614
0.768
0.795
0.918
0.594
0.624
0.768
HC1(3)
9.5
0.301
0.568
0.490
0.421
0.551
0.579
0.835
0.661
0.496
0.860
HC1(4)
3.3
0.350
0.645
0.856
0.779
0.852
0.884
0.859
0.450
0.727
0.630
HC1(5)
5.1
0.311
0.641
0.771
0.672
0.782
0.825
0.892
0.552
0.664
0.732
HC1(6)
2.3
0.337
0.564
0.904
0.784
0.889
0.863
0.751
0.324
0.751
0.486
HC1(7)
3-9
0.287
0.515
0.835
0.642
0.838
0.804
0.802
0.428
0.692
0.591
HC1(8)
1.8
0.334
0.517
0.904
0.827
0.840
0.813
0.603
0.205
0.744
0.332
HC1(9)
2.6
0.462
0.626
0.933
0.706
0.820
0.680
0.596
0.259
0.799
0.480
Res i n
1.4
0.392
0.614
0.945
0.743
0.913
0.813
0.753
0.345
0.790
0.548
Resin(2)
4.5
0.460
0.662
-.901
0.673
0.860
0.742
0.780
C.432
0.809
0.685
Res in(3)
6.2
0.517
0.696
0.836
0.574
0.820
0.651
0.767
0.460
0.791
0.781
Resin (4)
9.6
0.387
0.640
0.706
0.461
0.674
0.586
0.791
0.609
0.632
0.881
Resi n (5)
P fractions
4.6
0.273
0.584
0.835
0.720
0.862
0.851
0.844
0.465
0.693
0.576
N>
Sol P
2.7
0.322
0.624
0.788
0.903
O.858
0.899
0.709
0.249
0.678
0.372
O
vjn
A1 -P
54.3
0.046
0.268
0.071
0.060
0.100
0.160
0.491
0.746
O.C66
0.722


Table 50. Continued
Mean
Soi1-test
method
extract-
able P
Height
Rel. height
Fert.
reqm.
%P i n
tops
P uptake
d'c
1
2
1
2
1
2
1
2
1
2
ppm
01 sen(2)
9.9
0.259
0.652
0.583
0.61 1
0.616
0.756
0.955
0.749
0.506
0.848
01 sen
12.3
0.218
0.601
0.533
0.515
0.552
0.6787
0.931
0.806
0.447
0.875
01 sen (3)
14.7
0.195
0.548
0.466
0.437
0.460
0.595
0.880
0.826
0.387
0.875
01 sen(4)
12.5
0.213
0.630
0.560
0.650
0.597
0.794
0.973
0.768
0.474
0.793
01 sen(5)
13.2
0.186
0.518
0.470
0.505
0.516
0.676
0.915
0.777
0.419
0.794
Olsen(6)
16.0
0.224
0.606
0.577
0.557
0.618
0.734
0.953
0.754
0.491
0.835
01 sen(7)
15.9
0.217
0.642
0.569
0.645
0.621
0.798
0.981
0.761
0.482
0.801
01 sen(8)
16.7
0.238
0.648
0.529
0.611
0.621
0.764
0.972
0.739
0.443
0.847
01 sen(9)
20.8
0.231
0.629
0.486
0.553
0.549
0.696
0.946
0.792
0.443
0.847
Bray 2(2)
39.6
0.125
0.478
0.318
0.328
0.353
0.495
0.819
O.858
0.271
0.845
Bray 2(3)
48.7
0.147
0.522
0.344
0.373
0.400
0.543
0.858
0.848
0.305
0.858
Bray 2
41.6
0.095
0.414
0.231
0.251
0.262
0.404
0.741
O.858
0.201
0.813
Bray 2(4)
53.6
0.094
0.413
0.220
0.248
0.251
0.398
0.732
0.852
0.194
0.809
Bray 2(5)
47.3
0.107
0.411
0.207
0.197
0.250
0.344
0.694
0.813
0.192
0.829
Bray 2(6)
60.9
0.087
0.368
0.178
0.165
0.201
0.301
0.643
0.815
0.159
0.806
Bray 1(2)
34.4
0.148
0.529
0.338
0.379
0.386
0.539
0.854
O.858
0.298
0.860
Bray 1(3)
41.3
0.187
0.607
0.376
0.478
0.436
0.616
0.895
0.827
0.348
0.850
Bray 1
40.0
0.094
0.409
0.241
0.255
0.279
0.418
0.754
O.850
0.209
0.810
Bray 1 (4)
47.1
0.117
0.478
0.274
0.338
0.308
0.486
0.802
0.866
0.244
0.822
Bray 1(5)
43.8
0.086
0.380
0.204
0.210
0.224
0.353
0.690
0.839
0.178
0.795
Bray 1(6)
55.8
0.095
0.376
0.185
0.173
0.205
0.304
0.647
0.819
0.170
0.807
Bray 3
35.8
0.087
0.402
,0.221
0.249
0.252
0.402
0.738
0.861
0.191
0.801
Bray 3(2)
4o.O
0.087
0.381
0.202
0.203
0.234
0.352
0.697
0.839
0.180
0.801
Bray 4
28.6
0.124
0.467
0.297
0.333
0.332
0.491
0.810
0.855
0.264
0.825
Bray 4(2)
37-4
0.098
0.387
0.227
0.227
0.246
0.274
0.711
0.842
0.200
0.807
N>
O
Jr-


263
Gessell, S. P. 1968. Progress and needs in tree nutrition research in
the Northwest, p. 216-225. _l_n Forest fertilization theory and
practice. Tennessee Valley Authority, Muscle Shoals, Ala.
Goodall, D. W., and F. G. Gregory. 1947. Chemical composition of plants
as an index of their nutritional status. Imp. Bur. Hort. and Plant
Crops, E. Mailing, U. K. Tech. Commun. 17.
Grigg, J. L. 1965. Prediction of plant response to fertilizers by
means of soil tests. New Zeal. J. Agrie. Res. 8:893_90i4.
Gunary, D. 1970. A new adsorption isotherm for phosphate in soils. J.
Soi1 Sci. 21:72-77-
Haas, A. R. C. 1936. Phosphorus nutrition of citrus and the beneficial
effect of aluminum. Soil Sci. 42:187-201.
Hagenzieker, F. 1958. Fertilizing forest trees. World Crops. 10:
369-372.
Hagner, D. S. 1967- The evolution of forest fertilization in Sweden,
p. 41-48. J_r^ Forest fertilization. Proc. Vth Colloquiumof the
International Potash Institute, Jyvaskyla, Finland.
Hagner, D. S. 1971. Techniques in silvicultural operations. IUFRO
Cong. 15 Oiv. 3:184-207- Sveriges Skogsvardforbands Forlag, Stock
holm.
Haines, L. W., and W. L. Pritchett. 1965. The effect of site prepara
tion on the availability of soil nutrients and on slash pine growth.
Soil and Crop Sci. Soc. Fla. Proc. 25:356-364.
Hall, M. J., and H. M. Purnell. 1961. Potassium deficiency in Pinus
radiata iri Eastern Australia. Austr. For. 25:111-115.
Holmen, D. H. 1967- Forest fertilization in Sweden, p. 291-297. In
Forest fertilization. Proc. Vth Colloquiumof the International
Potash Institute, Jyvaskyla, Finland.
Hopkins, E. R. i960. The fertilizer factor in Pinus pinaster planta-
tions on sandy soils of the Swan Coastal Plain, Western Australia.
West. Austr. For. Dept. Bull. 68. 26 p.
Hortenstine, C. C. 1964. The effect of lime and phosphorus fertiliza
tion on oats and soil phosphorus in Lakeland fine sand in small
lysimeters. Soil and Crop Sci. Soc. Fla. Proc. 24:3542.
Hsu, P. H. 1965. Fixation of phosphate by aluminum and iron in acidic
soils. Soil Sci. 99:398-402.
Huffman, E. 0. 1962. Reactions of phosphate in soils: recent research
by TVA. Proc. Fert. Soc. no. 71:1-48.


33
insoluble sources of inorganic and organic P (Bowen, 1973), the evidence
reviewed above tends to indicate that extractant which remove plant-
available P are likely to be of more value than total P analyses for pre
dicting forest site fertility.
A problem in evaluating nutrient status of forest sites which has
received little attention was pointed out by Voigt (1958). He sug
gested that the perennial nature of trees is likely to present problems
in evaluating the .nutrient status since the rate of nutrient cycling
rather than the level of a particular fraction in the soil may be of im
portance in determining whether the trees' nutrient requirements are met.
For P, this is likely to be of particular importance in closed canopy
stands of trees, as Will (1964) reported that after canopy closure the P
requirements of radiata pine were met almost entirely by nutrient cycling.
However, he found that prior to canopy closure the tree-P requirements
were met almost entirely by net withdrawl from the soil. Ballard
(1970 a) suggested that the reason for the significant correlations he
found between P extracted by several available-P tests and foliar-P con
centrations of 40-year-old radiata pine was due to these methods reflect
ing the availability of organic P once it had been mineralized.
m
Interpreting soil test results
The successful use of soil tests depends entirely upon their prior
calibration against crop performance in the field (Williams, 1962). In
forestry, most studies have been concerned with calibrating soil-test
values against site productivity, their objective being to establish
critical levels below which growth performance of trees decline.
Prediction of productivity. Young (1940) found that a total-P
value in the topsoil of 47 ppm delineated areas of slash pine which showed


Table 36. Regression coefficients for variables included in 'best fit'
multiple regression equations of height at age 1, 3, and 5
years on soil and site parameters and the squared terms of
these parameters
Source
Coefficients
S i gnifi canee
R2
Height at age 1 year
0.540**
Mean
49.444
.
h2o-p
3.706
AJ.
r\
(Drainage class)z
-0.626
JUJU
Silt + clay
-0.007
JUJU
Height at age 3 years
0.705**
Mean
62.944
h2o-p
8.525
JUJL
Drainage class
99-196
JU n
(Drainage class)z
-16.709
**
Silt + clay
-0.017
Height at age 5 years
0.610**
Mean
82.1 39
(h2o-p)2
0.958
A
Drainage class
105.065
AA
(Drainage class)2
-21.181
AA
Depth to LH
6.194
AA
(Depth to LH)2
-0.042
JUJU
Significant at the 5% level.
'Significant at the 1% level.


Table 58. Continued
Jf
Field
s i te
Repli-
cate
Fertilizer P reauirement
1 year
3 years
5 years
90
95
100
90
95
100
90
95
100
K9/ na
A20
1
33
41
61
35
43
59
35
43
62
2
38
47
66
30
36
60
35
42
59
3
38
47
69
36
43
59
35
42
58
A21
1
0
0
0
0
0
0
0
0
0
2
0
0
50
0
0
0
0
0
0
3
9
18
41
0
2
41
0
0
9
A22
1
0
0
90
0
0
0
0
0
90
2
6
35
105
15
26
54
6
22
65
3
0
0
0
0
0
90
0
0
0
A23
1
20
28
47
0
0
0
0
0
0
2
66
80
90
31
49
91
23
39
75
3
28
48
86
53
73
90
55
74
90
A24
1
38
54
91
62
77
90
0
62
90
2
34
46
77
11
25
58
0
17
59
3
25
35
58
16
28
60
4
20
57
A25
1
0
0
90
0
0
90
0
83
90
2
7
26
73
0
0
0
0
0
0
3
3
38
0
0
0
0
0
0
0
A26
1
0
0
0
0
0
90
0
0
90
2
0
0
90
0
0
0
0
0
64
3
0
0
90
0
2
50
0
1
44
NJ
J-
ON


than CDB extractable-Fe (Syers et al., 1971; Udo and Uzu, 1972). These
results have been interpreted as indicating that the relative activity of
the various soil forms of Al and Fe in P retention follows the order:
exchangeable>amorphous>crystal1ine; while, within each form, Al is more
active than Fe on a per unit weight basis (Bromfield, 1965; Syers et al.,
1971; Udo and Uzu, 1972). Soil constituents that contribute to P reten
tion in soils other than various Al and Fe forms include phy1 los i 1icate
clays and CaCO^ (Kuo and Lotse, 1972). Calcium carbonate is of little
significance in virgin acid soils, but kaolinite, which is more effective
at retaining P than 2:1 clays (Ramula, Pratt, and Page, 1967) can retain
up to l87)jg P/g (Kuo and Lotse, 1972). Nevertheless, kaolinite is less
reactive than other forms of Al and Fe in the soil (Huffman, 1968).
P Effectiveness in Forestry
.The majority of fertilizer trials comparing the effectiveness of
different P sources in forestry have involved water-soluble superphosphate
(OSP or CSP) and slowly soluble rock phosphate (RP) sources.
Early trials in Australia comparing the effectiveness of OSP and
RP gave similar results: Both RP and OSP applied at equivalent rates of
P were equally effective as long-term P sources (Young, 1948; Richards,
1956; Richards, 1961; Gentle et al., 1965). Although an adequate de
scription of the soil properties was not provided in most of these reports,
it was mentioned that all soils were acid. Young (1948) and Richards
(1956) pointed out that RP was a more economical source than OSP, because
it was cheaper and more effective per unit weight of fertilizer applied.
Gentle and Humphreys (1968) reported that a trial established in Penrose
State Forest, New South Wales, in which OSP and RP were applied at equal
rates of P (77 kg/ha), showed OSP to have an early superiority, but after


Table 56. Continued
Field
s i te
Rep 1i-
cate
Depth
Particle
size distribut ion
Moisture
content
Bulk
density
C lay
Si It
Sand
1/3 atm.
15 atm.
g/cc
'O
A18
2
40-60
6.5
9-7
83.8
8.6
2.6
1 .97
3
0-20
2.3
9.6
88.1
7.2
4.3
1 -33
20-40
0.7
8.2
91.1
3.1
1.4
1.90
40-60
4.6
9-6
85-8
7.5
2.7
2.00
A19
1
0-20
0.9
7.7
91.4
5.5
5.0
1.18
20-40
0.8
6.5
92.7
1 .8
1.1
1.78
40-60
5-7
8.2
86.1
8.4
3.2
2.00
2
0-20
0.9
7.4
91.7
6.1
5.0
1 .20
20-40
1.3
6.6
92.1
2.2
0.7
1 .11
40-60
5.0
8.0
87.0
7.9
3.4
1.87
0
0-20
1.2
8.5
90.3
6.5
4.8
1.19
20-40
0.9
6.8
92.3
2.2
1.1
1.88
40-60
5.9
8.9
85.2
8.3
3-7
1.86
A20
1
0-20
3.1
26.7
70.2
8.1
5.1
1.62
20-40
4.0
22.4
73.6
5.1
1.6
1.72
40-60
4.6
21.3
74.1
5.9
1.5
1.90
2
0-20
7.5
24.6
67-9
1 1.0
5.8
1.45
20-40
4.6
26.4
69.O
8.3
2.6
1.73
40-60
5.0
26.2
68.8
7.3
1.8
1.90
3
0-20
3.0
25-2
71.8
7.3
4.5
1.40
20-40
4.4
25.2
70.4
8.2
2.5
1.71
40-60
5-5
24.4
70.1
10.2,
2.0
1.93


66
solution. Phosphorus was determined using the above soil P technique.
Calcium, Mg, A1 and Fe were determined by atomic absorption and K by
flame emission as described above. Tissue N was determined by the macro
Kjeldahl technique (Bremner, 1965).


C7)
Spectacular growth responses following addition of phosphate fer
tilizers to conifers have been recorded in many countries. As a conse
quence, routine application of P fertilizer, particularly at time of
planting, is now practiced on P-deficient sites in New Zealand (Conway,
1962; Armitage, 1969), Australia (Gentle and Humphreys, 1968), Japan
(Kawana, 1969), Britain (Leyton, 1958), Finland (Salonen, 1967), Norway
(Jerven, 1967), France and Germany (Baule and Fricker, 197-0) and the
southeastern USA (Pritchett and Hanna, 1969). In 1970, 70-80,000 ha were
fertilized at time of planting in Japan, about 30,000 ha in midwest
Europe and 10,000 ha in both the US Southeast and Oceania (Australia and
New Zealand) (Hagner, 1971). Superphosphate was the principal fertilizer
used.
Applications of P fertilizer to young stands at rates of 78 kg P/ha
have maintained responses for a minimum of 15 years (Pritchett and Swin-
ford, 1961; Gentle, Humphreys, and Lambert, 1965). However in other fer
tilizer trials in both the US Southeast and Australia, growth response
and foliar concentrations have shown that applications of superphosphate
up to 78 kg P/ha began to lose their effectiveness 7 to 15 years after
application (Gentle and Humphreys, 1968; Humphreys and Pritchett, 1971;
Pritchett and Smith, 197*0. The loss in efficiency of soluble forms of
P fertilizer was attributed to either leaching from the rooting zone in
soils of low P-retention capacity, or binding in unavailable forms in
soils of very high P-retention capacity. The practical and economic prob
lems of providing an adequate P supply throughout a rotation has led to a
policy in New Zealand and Australia of applying only sufficient fertilizer
at time of planting to provide an adequate P supply for pinus species up


139
Table 33. Comparison of the effectiveness of P extracted from the sur
face 20 cm of soil and that extracted from within the effec
tive soil depth (volume) at predicting height of slash pine
after 1, 3, and 5 years' growth on 72 field sites
Soi1-test
method
Soi 1
sample
Height
1 year
3 years
5 years
r2
H,0
0-20 cm
0.369**
0.324**
0.209*
Volume
0.238**
0.329**
0.238**
NH^OAc
0-20 cm
0.193**
0.195**
0.084*
Vo1ume
0.071*
0.122**
0.062*
HC1 -H2S0it
0-20 cm
0.012
0.082*
0.063*
Volume
0.000
0.059*
0.055*
Bray 1(3)
0-20 cm
0.002
0.049
0.040
Volume
0.006
0.056
0.041
Bray 2
0-20 cm
0.018
0.006
0.010
Volume
0.012
0.024
0.028
* Significant at the 5% level, using the model Y = b logX + c
**Signif¡cant at the 1? level.


86
between the intensity and quantity factors of P supply for the soils used
in any study. Such was the case in this study (Table 6). Although P ex
tracted by the NH^OAc method was found to be unrelated to response over a
5 year period in this study, Pritchett (1968) reported a significant cor
relation between P extracted by NH^OAc and response of slash pine on six
Coastal Plain soils 5 to 8 years after fertilizer addition. However,an
examination of the relationship between NH^OAc and Bray 2-extractable P
reported for these six soils showed them to be significantly correlated
(r = 0.902). This close relationship between the intensity and quantity
measurements of soil P on these six soils probably accounts for the suc
cess achieved with the NH^OAc procedure.
:- The trend for intensity measurements of soil P to decline in value
and quantity measurements to improve in value as predictors of tree growth
and response with increasing age of trees probably accounts for the effect
of tree age on the success of soil-test methods at predicting tree height
reported by Wells and Crutchfield (1969), Wells et al. (1973) and Ballard
(197*0. These authors found the ability of P extracted by strong extrac
tants (HCl-H^SO^, Bray 2, Olsen) to predict height growth of young trees
improved with age of trees from time of planting. It appears reasonable
to assume that had these studies included a measure of soil-? intensity
they would have found the same trends as observed in this study.
The growth of seedlings in pots is analagous to growth of seedlings
in nurseries. Thus,it can be expected that soil-test methods using ex
tractants such as H^O and NH^OAc should be good predictors of growth and
fertilizer requirements of nursery stock. However, a good transplant
seedling should not only be large and sturdy, but it should also possess
a high nutrient content. Since the nutrient contents tend to be related


Table 49. Continued
Soi1-test
Soil (CRIFF ident I
f1 cat ion
code)
method
A2
A8
A21
A24
A28
A1 6
A19
A23
A5
A25
HP1"
Bray 8(2)
2.9
40.9
187.5
2.9
2.4
6.9
8.5
5.0
21.5
133.5
HC1
0.7
21.6
70.5
0.3
0. 1
7.8
10.1
4.5
7.0
31.3
HC1 (2)
0.8
6.1
25-5
0.4
0.3
5.1
4.7
2.2
3.5
12.8
HC1(3)
1.5
8.6
41.5
2.5
1.0
5-5
5-6
3.0
5.3
20.9
HC1(4)
0.3
3.1
11.2
0.2
0.2
5-3
4.8
1.5
1.8
4.5
HC1(5)
1.0
4.5
18.0
0.6
0.5
5-7
5-5
2.6
3.9
9.2
HC1(6)
0.3
1.8
5.7
0.2
0.2
4.4
5.4
1 .5
1 .2
2.4
HC1(7)
0.9
3.0
12.6
0.5
0.3
5.1
6.9
2.0
2.4
5.2
HC1(8)
0.2
1.1
2.7
0.1
0.2
4.3
5-6
1.4
0.9
1.4
HC1(9)
0.3
1.3
6.9
0.1
0.4
4.8
6.5
2.0
0.6
3.0
Res ¡ n
0.3
1.3
3.4
0.2
0.2
2.3
2.8
1 .2
0.7
1.8
Resin (2)
1 .1
3.6
13.0
0.7
1 .0
5.5
7.5
3-2
2.2
6.9
Res In(3)
1 .8
4.8
19.9
1.3
1 .6
6.3
8.5
4.7
2.9
10.0
Res In(4)
3.1
7.3
28.3
1.6
2.2
6.7
9.4
5.2
4.9
17.1
Resin (5)
1.4
4.9
9.0
0.5
0.6
5.7
7.7
3.7
3.9
9.2
P fractions
Soi1-P
0.0
2.0
3.1
0.0
0.0
7.4
8.9
2.8
0.9
1 .6
Al-P
8.1
51.4
284.5
5.5
4.6
2.4
3-5
4.7
26.4
151 -9
Fe-P
2.6
1.6
28.1
1.3
4.0
0.7
0.7
1.5
5.4
31.7
Ca-P
0.2
4.0
0.0
0.0
0.0
0.0
0.7
0.1
0.8
0.0
Organic-P
40.8
25.8
88.8
33.1
39.8
26.1
28.1
37.1
24.5
81.8
Tota 1-P
118.0
108.0
410.0
85.0
85.5
51.5
58.0
64.5
80.5
300.0
Buffering capacity
Capad ty (1)
1 .0
3.9
73.0
1 .0
1 .0
1.0
1 .0
1 .6
1 .7
19-0
200


152
first 3 to 5 years following slash pine establishment, the HCl-F^SO^ or
either of the Bray methods were superior to the NH^OAc method currently
used in Florida. Of these three methods, the HCl-F^SO^ method is prob
ably the most suitable for several reasons: (a) Its prediction value
was at least as good as either of the Bray methods. (b) The method can
also be used to determine the amounts of available cations in the soil
(Page et al., 1365). (c) The method is currently used on a routine
basis by several southern states (Page et al.,1965), and thus its use by
Florida would offer an opportunity for improved standardization and
calibration. (d) The method is suited for large-scale routine determi
nations in soil-testing laboratories since the analytical procedure is
rapid and presents no problem from ion interference in the colorimetric
determination of P.
From the results presented above, the following interpretation
of HC1-H^SO^ test results is suggested. (a) Soils containing < 5 ppm
of HC1-H^SO^-extractable P should,in the main, respond significantly to
applications of P fertilizer. (b) Soils testing between 2.5 and 5 ppm
P should receive applications of ca. 20-*t0 kg P/ha applied in a band
1.2 metres wide down the tree rows at or near the time of planting.
Soils testing below 2.5 ppm P should receive applications of ca. *)0-80
kg P/ha. If the fertilizer is broadcast, rather than banded in the tree
row, field experience has indicated application rates approximately 50%
greater than those shown above should be used. For soils testing below
2.5 ppm which have a high P-retention capacity, the highest application
rate of 80 kg P/ha should be used. Evidence has suggested however that
even this rate may be inadequate over a 5_year period from establishment
on some of the most highly P-retentive soils (Pritchett and Smith, 197*0


51
compounds shown in Table 3- Each treatment was replicated three times
for the Immokalee soil but only twice for the McLaurin soil because it
was in short supply.
Closed plastic pots containing 1,500 g of soil were used. The P
compounds were applied at a rate designed to raise the total P content of
the soil by 100 ppm. Required quantities of each compound, all previous
ly ground to pass a 0.0105_mm sieve, were thoroughly mixed, throughout
the soil for each pot. Three carefully graded seedlings were transplant
ed into each pot. Moisture conditions and day length were maintained as
outlined above.
Harvesting
Eight months after transplanting, the heights of all seedlings
were recorded. Following height determinations, the seedlings were
harvested using the procedure outlined for greenhouse trial 1. The tops
and roots from each pot were dried at 70C, weighed, and ground to pass a
1-mm sieve. Phosphorus uptake per pot was determined as the combined
product of P concentration in the tops and roots and their corresponding
dry weights.
Extraction of P Compounds
Each of the eight P compounds (Table 3) was added at a rate of
100 ppm P to duplicate 100-g samples of each of the two soils. All samples,
including control soil samples, were thoroughly mixed by passing through a
0.5-mm sieve several times. One of the duplicate sets was incubated at
room temperature in the dark for 2 months, with moisture being maintained
at field capacity by the addition of distilled water. Following this in
cubation period, the samples were air-dried and passed through a 0.5-mm
sieve.


181
extractant, and P compound involved. Soil-test methods which were success
ful in the preliminary screening as predictors of response over a 1-year
growth period extracted little more P from the compounds than did h^O
alone. However, those methods which successfully predicted response over
longer growth periods extracted P from colloidal A1 and Fe phosphates and
K taranakite, which were effective P sources for seedlings; these ex
tractants did not remove P from wavel1 ite and strengite, which were inef
fective P sources. A strongly basic P compound (f1uorapatite) was a good
P source on an acid soil, but a poor source on a soil of pH > 5- So i 1 -
test methods using acidic extractants overestimated the availability of
this P compound to seedlings growing on less acid soils.
Five soil-test methods ^0, NH^OAc, HCl-H2SOi4, Bray 1(3), Bray 2),
selected on the basis of results from the preliminary screening, were
calibrated against height, height response, and P-fertilizer requirements
of slash pine after 1, 3, and 5 years' growth on 72 sites throughout the
Coastal Plain. Soil-test values were also calibrated against the concen
tration of P in foliage collected from A-year-old slash pine in the con
trol plots of the fertilizer trials on each of the 72 sites. Relation
ships between soil and site variables, other than soil P, and the growth
and response parameters of slash pine were examined. In addition, the
effect these variables had on the predictive ability of extractable-soi1
P was determined.
The effectiveness of the soil-test methods at predicting growth, re
sponse, and P-fertilizer requirements of slash pine on the 72 sites was sim
ilar to that found in the preliminary screening. The three soil-test methods
which extracted some of the solid phase component of soil P (HC1-H2S0j,,
Bray 1 (3), Bray 2) provided the best index of response and P-fertilizer


Table 54. Soil classification, site properties, and selected chemical properties of unfertilized
soils (0-20 cm), for the 24 field sites
:'c
Field
s i te
Repi i-
cate
Soi 1
Type
Order
Drainage
* Vc
c 1 ass
Depth
to LH+
Depth
of A1
Organic
matter
Tota 1
N
CEC
A1
1
Plummer fs
U1tisol
1
cm-
20
18
%-
1 .28
0.053
meq/1OOg
3.12
2
Plummer fs
U111 sol
1
20
18
1 .48
0.063
3.20
3
Plummer fs
U1tisol
1
23
20
1.68
0.059
3.07
A2
1
Bladen scl
U1tisol
1
20
15
3.04
0.102
13.43
2
Bladen scl
111 t i sol
1
20
15
2.57
0.085
12.16
3
Bladen scl
Ultisol
1
20
15
2.11
0.098
12.28
A3
1
Leon fs
Spodosol
2
45
15
2.01
0.055
5.08
2
Leon fs
Spodosol
2
40
13
1.38
0.041
3.20
3
Leon fs
Spodosol
2
50
13
1.51
0.035
2.72
A4
1
Leon fs
Spodosol
2
33
20
1.97
0.053
4.44
2
Leon fs
Spodosol
2
35
15
1.68
0.064
3-65
3
Leon fs
Spodosol
2
20
18
1.64
0.065
3.83
A5
1
Kershaw fs
Entisol
5
90
20
0.85
0.048
2.33
2
Kershaw fs
Entisol
5
90
20
0.85
0.039
1.98
3
Kershaw fs
Enti so 1
5
90
20
0.85
0.037
1.95
A6
1
Leon fs
Spodosol
2
63
25
2.53
0.080
6.19
2
Leon fs
Spodosol
2
43
25
2.28
0.078
6.27
3
Leon fs
Spodosol
2
40
18
1.71
0.060
4.36
A7
1
Rutlege fs
1nceptisol
I
25
25
2.30
0.101
3.58
2
Rutlege fs
1nceptisol
1
25
25
2.04
0.103
3-73
3
Rutlege fs
1nceptisol
1
25
25
1.74
0.076
2.99


Table 47. Average heights of slash pine as affected by P treat
ments after 1, 3, and 5 years' growth in the field on
10 soils
Rep 1i-
cate
Height
1 year 3 years 5 years
Po Pi P 2 Po Pi P 2 Po Pi P 2
cm'
Bladen silty clay loam (A2)
1
27.0
40.5
l3^~
"76
108
110
185
236
258
2
33.1
39-0
43.8
98
119
122
242
271
268
3
31.4
38.3
37-1
94
117
113
245
284
270
B1 an ton
fine sand (A8)
3
55.4
56.9
"50
243
247
212
451
448
422
.
P1ummer
fine sand (A21)
1
44.4
40.6
42.6
216
172
168
433
428
418
Marlboro fi
ne sandy
' loam
(A2 4)
1
24.9
32.0
T8TT
149
148
l8o"
376
358
397
2
23-9
26.6
38.0
130
168
169
336
395
399
3
19.6
27.1
33.4
121
149
157
324
361
376
McLauri
n fine sandy
loam
(A28)
1
31-5
38.5
39-5
168"
216
228
387
443
447
2
33-5
33-5
48.4
192
214
269
428
443
507
3
30.0
41.3
45.9
164
237
242
386
460
463
1mmoka1ee
fine sand (A16)
1
54.6
52.1
230
261
266
457
521
518
2
50.8
47.0
49.-
227
217
248
467
460
494
3
447
48.5
50.3
178
205
246
392
427
482


9^
ignored. This may be one of the explanations for the increased
extractabi1ity noted with the lactate method. The marked increase in P
extracted from CFeP by alkaline NH^F in the presence of soil, presumably
caused by ionized organic functional groups complexing Fe, points to an
error in the assumption that alkaline NH^F is a selective extractant of
soil A1 -P.
Strengite (STR)
This ferric phosphate was practically insoluble in all extractants
examined. Only the Bray 2 method, in the presence of Leon soil, extract
ed more than 1% of the P in this compound.
Utilization of P Compounds by Slash Pine Seedlings
The uptake of P and dry matter production of slash pine seedlings
grown on two soils treated with the eight P compounds described in a
previous section are shown in Fig. 2 and Fig. 3, respectively. The raw
data from the greenhouse trial are presented in Appendix Table 52.
An analysis of variance of P uptake (Appendix Table 52) showed
significant differences between soils (S) and P compounds (P). The
S x P interaction was also highly significant and thus differences in
uptake between P treatments were examined within soils by the least
significant difference (LSD) method (Snedecor and Cochran, 1967)- Data
in Fig. 2 showed that on the Immokalee soil significantly greater
amounts of P were taken up by the seedlings in the MCP, DCP, FA, CA1P,
KTK, and CFeP treatments than in the control (C), WA and STR treatments.
There were no significant differences in P uptake within the former and
the latter groups mentioned above. On the McLaurin soil, significantly
greater P uptake occurred in the MCP, CA1P, CFeP and KTK treatments than
in the control, FA and WA treatments. The CFeP and KTK treatments were


Physical and chemical properties
In many studies relating P retention and soil properties, linear
regression analysis is used to relate P retention, determined as the ad
sorption of an arbitrary amount of applied P, to soil properties (Yuan
and Breland, 1969; Syers et al., 1971; Udo and Uzu, 1972). Data in Fig.
15 illustrate the relationship between adsorption of P applied at three
different levels and extractable Al by NH^OAc (pH *1.8), a -property pre
viously reported as closely related to P retention (Yuan and Breland,
1969). The relationship was not linear until sufficient P had been added
to saturate retention sites. Increasing the application rate from 100 to
2,500 yg P/g soil improved the linear correlation coefficient from r =
0.8l to r = 0.93. For soils used in this study, an application rate of
2,500 yg P/g soil appeared to be more than adequate to ensure such linear
relationships.
Correlations between P retention and soil properties commonly used
to characterize soils are shown in Table 39- Also shown are the mean
values and range of these properties. Specific values of these properties
for each soil are presented in Appendix Table 60. Soil pH, clay, silt,
and loss on ignition were significantly correlated with P retention. All
of these soil properties have been previously recorded as being related
to P retention in other soils (Saunders, 1965; Syers et al., 1971; Udo
and Uzu, 1972). The significant positive correlation between pH and P
retention is of interest in that this relationship is usually reported to
be negative (Udo and Uzu, 1972). Although this would be anticipated from
the greater activity of Al and Fe at lower pH, the soils with the lowest
pH in the Coastal Plain are Spodosols, which are strongly leached sands
containing very little Al and Fe of any form in the Al horizon and,


ACKNOWLEDGMENTS
I would like to thank Dr. W. L. Pritchett, the chairman of my
supervisory committee for his valuable guidance and counsel during the
tenure of this study. The interest and assistance given by the other
members of my committee, Dr. J.G.A. Fskell, Dr. F.G. Martin, Dr. W.H.
Smith and Dr. C.A. Hollis are very much appreciated.
I would like to acknowledge the encouragement given by Dr. D.F.
Eno, chairman of the Soil Science Department and other members of the
faculty. In particular I would like to thank Dr. L.W. Zelazny and Dr.
J.G.A. Fiskell for their contribution to my academic enrichment and Dr.
D.F. Rothwell for his administrative work on my behalf.
Thanks are also due to the CRIFF laboratory personnel, in par
ticular to Mary McLeod for her help in the laboratory phase of the work,
and to Dr. H.L. Breland and the staff of the Soils Analytical laboratory
and to Dr. J. NeSmith and the staff of the Soil Testing laboratory for
their expert analysis of numerous samples.
I have enjoyed the companionship of the graduate students of the
Soil Science Department who provided necessary light relief and much
intellectual stimulation. Specifically I would like to thank my fellow
graduate students in the CRIFF program, Terry Sarigumba and Roy Voss,
who willingly helped with some of the heavy work.
My most sincere thanks to my wife Pip, who has been actively in
volved in all phases of this work from initial collection of samples
through to the final typing.
i i i
i


169
soil-test methods, extractable A1 was more closely correlated with P re
tention than extractable Fe, accounting for up to 80% of the variation in
P retention. Summed A1 and Fe values improved the correlation coefficient
obtained over that with A1 alone in three of the four tests. In the case
of extractable A1, it appears that extractants capable of complexing A1
(acetate and fluoride) provide better indices of P retention than tests
which rely solely on acidity to extract the A1, irrespective of quanti-
i .
ties extracted. However, for extractable Fe, it appears the greater the
quantities of Fe removed by the extractant the better the correlation
with P retention. In view of the highly significant relationships between
soil-P retention and amounts of A1 extracted by these same tests, they
apparently could be used to provide an index of the soil-P retention.
With the possible exception of the HC1-H2S0( method, the additional de
termination of Fe in the extract would hardly be justified by the small
degree of improvement in predictability.
Relationships Between P Retention and Foliar Nutrient Concentrations
Since forest fertilization is not limited to young plantations and
the need for fertilizer in established plantations is more usually based
on foliar analysis, the usefulness of foliar elemental concentrations to
predict soil-P retention was examined using 3b out of b2 sites from which
foliage samples were obtainable. Element concentrations in ^-year-old
slash pine foliage collected from control plots of replication 2 are pre
sented in Appendix Table 59.
Foliar Fe and A1 were the only elements significantly correlated
with soil-P retention (Table bb). On acid soils, this could have been
predicted since the activity of Fe and A1 in the soil should determine
both P retention (Hsu, 1965) and foliar levels of Fe and A1, provided


Table 49. Continued
Soi1-test Soil (CRIFF identification code)
method
A2
a8
A21
A24
A28
A16
A19
A23
A5
A25
PJJMI
T ruog
2.4
6.6
32.8
1.6
1.0
6.5
7.5
3.6
4.6
22.4
T ruog(2)
0.7
5.6
22.8
0.5
0.7
6.2
7.0
3.1
3.7
12.1
Troug(3)
0.1
4.0
6.3
0.1
0. 1
5-5
7.2
2.0
2.1
3.1
T ruog(4)
3.9
16.5
82.2
2.1
2.5
6.5
7.6
4.1
11.7
50.0
T ruog(5)
1.3
14.1
65.9
1.3
1 .2
6.1
7-6
3.8
9.0
38.9
Truog(6)
0.6
11.9
50.4
0.5
0.6
5-8
7.3
3-0
7.1
24.3
Truog(7)
3.5
26.5
134.0
2.4
3-6
6.0
8.5
4.5
18.2
115-0
Truog(8)
4.4
31.6
143.0
2.5
2.8
5-7
6.8
4.4
18.7
110.0
T ruog(9)
2.9
29.9
136.3
2.1
2.2
5-4
7.5
3.6
18.4
104.5
T ruoq(10)
11.0
19-0
65-5
8.8
10.0
6.8
10.8
5.8
7.5
36.0
Truog (1 1)
6.0
9-9
55.2
6.2
5-2
4.4
5-5
2.5
6.0
27-8
Truog(12)
3.8
8.1
41.3
3.0
3-3
3.5
4.9
1.8
5.1
21.3
H2S0a
1.4
5.8
30.2
0.6
0.4
5.9
8.0
3-3
3-6
13.8
H2So4 (2)
0.3
4.5
17.4
0.1
0.2
5-6
6.9
2.6
2.8
8.5
H2S04(3)
0.1
2.4
3.1
0.1
0.1
5-9
7-0
2.0
1.4
1.4
H2SCM4)
2.6
15-6
81.5
2.4
2.6
7.4
11.5
6.0
10.4
50.8
H2S0j (5)
1.6
14.9
68.0
1 .4
1.4
6.5
8.3
4.2
8.9
39.0
h2so.(6)
0.5
12.4
49.3
0.5
0.5
6.0
7.8
3.4
6.6
8.5
h2so4(7)
3-1
29-5
155.0
2.8
3.1
6.2
8.6
5.1
21.2
126.0
H2S04(8)
3.4
31.2
153-0
2.2
2.5
6.2
7.6
4.7
19.8
116.0
h2so4(9)
3.2
30.0
146.0
2.0
2.3
5.8
7.8
4.2
18.6
104.8
HC1-H2S0^
1.6
14.5
68.2
1.3
1 .2
7.0
9-1
3-8
9-1
36.4
Olsen(2)
0.5
11.6
48.5
0.6
0.6
4.5
5-7
2.1
5.7
19.8
01 sen
1.7
16.4
50.0
1.2
1.4
5-6
6.9
3-3
8.3
28.2
Olsen(3)
2.0
18.4
60.9
1.8
1.9
6.0
6.9
3-1
9-9
35.8
01 sen(4)
0.4
18.1
51.5
0.5
0.5
5-1 .
6.6
2.0
8.8
31.0
01 sen(5)
1 .0
17.5
53.9
1.7
0.8
5.5
7.6
2.2
9.1
33.0


RELATIVE HEIGHT,
Fig. 12. Relationships between P concentrations in foliage(x) of ^-year-old slash
pine and relative height of slash pine 1, 3, and 5 years (Yl, Y3, and Y5)
after P fertilization.


hi
Growth and Response Parameters
In computing the growth and response parameters for the greenhouse
trial, no adjustment for the N effect was required, since all treatments
received a uniform N application. Mean height in pots not receiving P
was taken as index of growth in the absence of P fertilizer. Relative
height and P-fertilizer requirements were computed in a similar manner to
those for the field trials using a response curve fitted with a quadratic
equation. Growth and response parameters were computed from the data
obtained at each harvest. In order to compare the growth and response
parameters from the greenhouse study with those of the field trials grow
ing on the same soils, the field parameters for the 10 trials were
computed using P0N2, PiN2,and P2N2 treatments only. The N2 treatment in
the field was the same rate, per unit surface area, as that used in the
greenhouse trial. In addition, the parameters were computed using all
three replicates at each of the 10 sites, rather than just the replicate
from which the bulk soil sample was collected. However, for 3 of the 10
i
sites only the replicate from which the sample was collected was used,
because soil properties and response information indicated substantial
variation between replicates on these three sites.
Other parameters determined were the concentration of P in the
seedling tops, and total uptake of P (mg/pot). Since only the tops were
harvested at the end of the first growth season, total P uptake at this
stage was determined using the relationship between quantity of P in tops
and total P in tops and roots computed from the 10 extra pots.


applied P during the early growth period which disappeared over longer
growth periods as the trees made use of the less soluble, HCl-l-^SO^-
extractable P. Third, the number of nonresponsive sites with extract-
able P < 5 ppm increased with an increase in the growth period from 3 to
5 years. The probably explanations for this were discussed earlier.
Values of P extracted by the HC1-H2S0^ method corresponding to a
relative height of 30% using the log transformed models shown in Figs. 4,
5, and 6 were above the critical value of 5 ppm P determined by the Cate
and Nelson (1965) technique. However, these models tended not to fit
the data in the initialy steep part of the response curve. Quadratic
models, excluding all sites which tested above 10 ppm of HCl-H^O^-
extractable P, did however provide P values corresponding to 90% relative
height in the vicinity of 5 ppm (Fig. 7). This critical value appeared
to be essentially independent of the age of the trees.
Humphreys and Pritchett (1972), suggested on the basis of rela
tionships between soil-test values and foliar P concentrations of slash
pine that a combination of intensity and quantity measurements of soil P
should provide the best index of the P status of soils for slash pine.
Regression equations of relative height at age 1, 3, and 5 years on the
P extracted from the surface 20 cm of soil by the H2O (intensity) and
HC1-H2S0^ (quantity) soil-test methods showed that only at age 1 year
did both of these soil tests contribute significantly to explaining
variation in relative height (Table 20). At age 3 and 5 years, the coef
ficient for H20-extractable P was nonsignificant when included in a
multiple regression equation with HC1-H2S0^- extractable P. This is con
sistent with the hypothesis concerning the importance of H^O-soiuble P
in early growth as proposed earlier. It should, however, be appreciated
that the HC!-H2S0^ method does not provide an exclusive measure of the


TABLE OF CONTENTS (Continued)
Page
Relationships Between Other Soil and Site Parameters
and P-Fertilizer Requirements 133
Relationships Between Soil-Test Values and Height . 136
Relationships Between Other Soil and Site Parameters
and Height 1^0
Relationships Between Foliar-P Concentrations and
Tree Parameters 1^5
Relationships Betv/een Foliar P and Soil-Test Values . 150
General Discussion 150
Phosphorus-Retent ion Study 155
Relationships Between P Retention and Soil Properties 155
Physical and chemical properties 157
Extractable A1 and Fe 161
Relative contribution of A1 and Fe to P retention 165
Aluminum and Fe extracted by soil-test methods . 16?
Relationships Between P Retention and Foliar Nutrient
Concentrations 169
Calibration of Extractable A1 Against Field-P Retention 172
SUMMARY AND CONCLUSIONS 179
APPENDIX 186
LITERATURE CITED 259
BIOGRAPHICAL SKETCH 273
VI i


8?
to quantity measurements of soil P, nursery fertilizer practices based
solely on intensity measurements could result in seedlings of adequate
size but possessing less than optimal nutrient reserves for maximum growth
following transplanting. A sound practice would be to base fertilizer
recommendations on the measurement of both quantity and intensity factors
of soi1-P supply.
It should be appreciated that soil-test methods which provide a
good measure of the quantity factor of soi1-P supply for acid sandy soils
which have their inorganic P fraction dominated by Al-P, such as those
used in this study, will not necessarily work as well for a different
range of soils. For instance, Alban (1972), using soils of high base
saturation which probably had insoluble basic Ca-P as a major component
of the inorganic P fraction, found a poor correlation between site index
of red pine and P extracted by solutions with a low pH (HC1-H2S0^, Bray
2). This can probably be attributed to the acidic extractants removing
insoluble basic Ca-P which was not available for tree use.
On the basis of results reported in this section, five soil test
methods--H20, NH^OAc, HCl-H2S0/t, Bray 1 (3) Bray 2--were selected for use
in the calibration study using 72 sites. These test methods extract P
with a range of intensity and they also include the tests currently used
on a routine basis by most of the southern states (Page et al., 1965).


Table 8. Height, relative height and P-fertilizer requirements of slash pine after 1, 3, and 5
years' growth, and
soils
foliar
P concentration after 4 years' growth in the
field on
10
Soi 1
type
Height
Rel. height
Fert. reqm.'f
Foliar P
4
1*
3
5
1
3
5
1
3
5
9

ts.y r / iica
X)
Bladen scl
29.5
90
227
62.2
69.9
79.5
71.2
61.0
42.4
0.067
Blanton fs
55-4
c+\
-3*
451
94.2
98.1
100.0
0.0
0.0
0.0
0.115
Plummer fs
44.4
216
433
100.0
100.0
100.0
0.0
0.0
0.0
0.109
Marlboro fsl
23.5.
136
350
65.1
80.0
89.9
114.0
50.1
1.5
0.076
McLaurin fsl
31.6
174
400
69.6
66.6
81.8
89.0
75.0
42.9
O.O83
Immokalee fs
50.3
209
436
100.0
82.5
86.0
0.0
47.6
25.6
0.090
Leon fs
88.1
291
498
99.4
90.1
90.0
0.0
0.0
0.0
0.092
Ona fs
42.0
156
342
68.4
76.6
81.1
94.3
119.5
115.6
0.081
Kershaw fs
26.4
135
293
89.2
100.0
100.0
0.0
0.0
0.0
0.103
Lakeland fs
27.5
129
276
96.5
100.0
100.0
0.0
0.0
0.0
0.101
* Age in years at measurement.
" Fertilizer required to achieve 90% of maximum height growth.


131
growth periods. This may be a valid situation resulting from a reduced
requirement for high concentrations of P in the surface soil with roots
exploiting greater volumes of soil as the trees increase in age. It may
also result from the fact that height is an insensitive indicator of
response of older trees, and/or that inadequate P-fertilizer applications
were made in some of the field trials to maintain a true response for the
full 5 years. Since the models for age 5 years shown in Figs. 9 and 10
are not significant, and because of the doubt concerning the validity of
the data from which they were derived, these age 5 prediction models are
not satisfactory for predicting operational fertilizer requirements.
Effect of soil sampling position and depth
The effect of sample position or depth on the prediction of fertil
izer requirements from extractable-P levels was similar to that observed
for the prediction of relative height (Table 27). The improved predic
tion resulting from use of soil samples from deeper profile depths
compared to the use of surface soil samples appears to have no simple ex
planation. Since P extracted from all depths and positions was signifi
cantly interrelated (Table 21) and subsoil extractable P is less likely
to be subject to short-term fluctuation associated with variable climatic
conditions which affect mineralization, P extracted from lower depths may
have provided a better indication of the average P status of the surface
soil than the level of P extracted from the surface soil at a particular
time of col lection.
Comparison of the effectiveness of P extracted from the surface
20 cm of soil with that extracted from within the effective soil depth
(Table 28), again revealed evidence of the increasing importance of soil
P in lower,, root-penetrable horizons with increasing age of trees. The


FERTILIZER REQUIREMENTS0/), kg'ha
129
60f \
10
oY(1)=-33.27logX + 47.33 (R2=0.277)
oY(3)--21.54logX + 41.10 (2=0.188)
aY(5)=-9.01 log X + 27.36 (R2=0.022)
0 2 4 6 8 10
HCI- H2SC>4 EXTRACTABLE P, ppm
Fig. 10. Relationships between HC1-H^SO^-extractable P(X) in the
surface 20 cm of soil and amount of P fertilizer (CSP)
required to achieve 95% of maximum height growth over
periods of 1, 3, and 5 years (Y 1, Y3. and Y5)
following fertilization.


260
Barnes, R. L., and C. W. Ralston. 1955* Soil factors related to growth
and yield of slash pine plantations. Fla. Agrie. Exp. Sta. Tech.
Bull. 559. 23 p.
Barr, A. J., and J. H. Goodnight. 1972. A user's guide to the statis
tical analysis system. Student Supply Stores, N. C. State Univ.,
Raleigh, N. C.
Baule, H., and C. Fricker. 1970. The fertilizer treatment of forest
trees. BLV, Mnchen, Germany.
Baur, G. N. 1959. A soil survey of a slash pine plantation, Barcoongere,
New South Wales. Aust. For. 23:78-87.
Bell, L. C., and C. A. Black. 1970a. Transformation of dibasic calcium
phosphate dihydrate and octacalcium phosphate in slightly acid and
alkaline soils. Soil Sci. Soc. Amer. Proc. 34:583-587.
Bell, L. C., and C. A. Black. 1970b. Crystalline phosphates produced
by interaction of orthophosphate fertilizers with slightly acid and
alkaline soils. Soil Sci. Soc. Amer. Proc. 34:735_740.
Bengtson, G. W. 1968. Progress and needs in forest fertilization
research in the South, p. 234-241. J_n_ Forest fertilization theory
and practice. Tennessee Valley Authority, Muscle Shoals, Ala.
Bengtson, G. W. 1970. Placement influences the effectiveness of phos
phates for pine seedlings, p. 51-63. _l_n_ C. T. Youngberg and C. B.
Davey (ed.) Tree growth and forest soils. Oregon State Univ. Press,
Corva 11is, Oregon.
Bengtson, G. W. 1972. Forest fertilization: Promises and problems.
p. 231-261 J_n_ Sound American Forestry. Proc. National Convention,
Society of American Foresters. Hot Springs, Arkansas.
Binns, W. 0. 1969. Fertilizers in forestry the future. J. Sci. Food
Agrie. 20:424-426.
Black, C. A. 1943. Phosphate fixation by kaolinite and other clays as
affected by pH, phosphorus concentration and time of contact. Soil
Sci. Soc. Amer. Proc. 7:123-133.
Bouyoucus, C. J. 1951. A recalibration of the hydrometer method for
making mechanical analysis of soils. Agron. J. 43:434-438.
Bowen, G. D. 1973- Mineral nutrition of ectomycorrhizae. p. 151-205.
j_n_ G. C. Marks and T. T. Kozlowski (ed.) Ectomycorrhizae their
ecology and physiology. Academic Press, N. Y. C.
Breland, H. L. 1966. Atomic absorption methods of analysis for agri
cultural samples. Soil and Crop Sci. Soc. Fla. Proc. 26:54-64.


105
Field Calibration of Selected Soil-Test Methods
Five soil-test methods, 1^0, NH^OAc, HCl-l^SO/j, Bray 1(3) and Bray
2, were selected for use in this calibration study. The amounts of P
extracted by each method from the four soil samples collected from the
control plot of each replicate of each trial (72 sites in total), are
given in Appendix Table 55* Tree heights in the absence of P fertilizer
and relative heights of slash pine on each of these 72 sites after 1, 3,
and 5 years' growth are shown in Appendix Table 7- The raw plot data
from which these results were obtained were too voluminous for inclusion
in this dissertation. The data are on file in the Forest Soils Section
of the Soil Science Department, University of Florida, Gainesville.
Relationships Between Soil-Test Values and Relative Height
The effectiveness (R^ values) of the five methods at predicting
relative heights, on the basis of P extracted from the surface 20 cm of
soil, showed similar, though less clearly defined trends to those found
in the preliminary screening (Table 17)* The values for water-extract
able P were most closely related to relative height of trees after 1
year's growth in the field, while P extracted by the HCl-H^SO^ and both
Bray methods tended to be most closely correlated with relative height
after 3 and 5 years' growth. The HCl-l^SO^ method was the most effective
overa 11.
As was the case for the soils used in the preliminary screening,
there was a poor relationship between l^O-extractable P and P extracted
by methods which removed larger quantities of P from the soil (Table 18).
The amounts of phosphate extracted by the HCl-H^SO^, Bray 1(3), and
Bray 2 methods were closely related, even though the HCl-f-^SO^ method
extracted appreciably less P than either of the two Bray methods. How-


33
Provided soil analysis results can be suitably calibrated against
tree growth and response to P fertilizer (and evidence suggests they can)
soil analysis has certain advantages over foliar analysis. Pritchett
(1968) outlined advantages for soil testing: (1) It can be used for pre
dicting fertilizer needs in areas prior to planting. (2) Collecting and
analyzing soil samples may be less laborious than collecting and analyz
ing needle samples, particularly in old stands. Ballard (1970 a) also
suggested that soil analysis has an advantage in that soil samples can
be collected at any time of the year, whereas foliage samples cannot.


12
Huffman (1962} found that the rate of dissolution of flor and hydroxy
apatites was proportional to their surface areas and that this rate de
creased with a rise in pH. Mattson et al. (1951) reported that the addi
tion of organic matter, which complexed Ca, increased the solubility of
hydroxyapatite about tenfold. Even in acid soils, the dissolution rate
of RP has been found to be slow. Chu, Moschler,and Thomas (1962) reported
that 18% of the RP applied k years previously to a soil at pH 5-7 (Nason
silt loam) had dissolved, while only 7*5% had dissolved 7 years after
application to a soil at pH 6.1 (Frederick silt loam). Gentle et al.
(1965) found that most of the RP applied to a podsolized soil at pH 5 was
still present as Ca phosphates after 15 years. Similarly, in the south
eastern Coastal Plain, Humphreys and Pritchett (1971) reported that 7 to 11
years after the addition of RP to 7 soils ranging in pH from k.2 to 5.^,
substantial amounts of P were recovered in the Ca-P fraction. In all
three of these studies, the transformation of RP produced an increase in
the Al-P and Fe-P fractions.
Diammonium phosphate (DAP) is water soluble and upon dissolution
in the soil produces an alkaline solution of about pH 8.0. The initial
reaction of the fertilizer solution produces basic phosphates of the gen
eral type (Ca, Mg) (NH^)^ (HPO^)2H2O (Lindsay et al., 1962; Bell and
Black, 1970 b). In slightly acid soils, Bell and Black (1970 b) found
DCPD was formed once the fertilizer solution had moved several centimeters
from the fertilizer source and the pH decreased. Bell and Black (1970 b)
also reported that in comparison with MCP and monoammonium phosphate,
which both form acid solutions, the movement through soil columns of P
from DAP was far greater, irrespective of pH and clay content of the soil.


27
Several investigators have reported P levels in the surface
horizon to be more closely related to growth and foliar P levels than P
levels in lower horizons. Humphreys (1963) outlined Australian work in
which it was reported that soil P below 37 cm contributed little to the P
supply of mature pines. Pawluk and Arneman (1961) found that available
P in the A horizon, but not the B horizon, was correlated with site
productivity of jack pine (P. banksiana Lamb). Similarly, Wells (1965)
reported that variation in concentration of foliar P of 5~year-old
loblolly pine was better explained by available P in the A than in the B
horizons. Alban (1972) also found that where P extracted from the 25 to
100 cm depth was included, it did not improve the estimation of the site
index of red pine in Minnesota over that obtained using P extracted from
the 0 to 25 cm depth. The apparently greater importance of P in the
surface horizon is attributed by most workers to the concentration of
fine feeder roots in the upper profile (Alban, 1972).
As in agriculture, horizontal soil variability is also a problem in
securing representative soil samples. The significance and extent of
variability of various soil properties in forest soils has been discussed
(York, 1959; Mader, 1963) and examined (McFee and Stone, 1965; Metz et al.,
1966). In reviewing these publications, Leaf (1968) pointed out that the
patterns and magnitudes of variability differ between soil characteris
tics at a particular site and determination of required sampling intensity
is a statistical problem.
In his review article, Leaf (1968) mentioned a problem in soil
sampling which is fairly unique when dealing with deep rooted perennials.
He stated:
...different degrees of biologically important soil heterogenity
exist on any site depending on the stage of tree development, e.g.,
seedling or establishiment stage vs. sapling stage vs. mature tree
stage, with the associated soil volume being tapped by the roots.
(p. 93)


49
presence of fluoride (Olsen and Flowerday, 1970* Wavellite has not been
identified as a soil-P fertilizer reaction product, but is naturally oc
curring mineral which was included as a representative of insoluble
stable A1 phosphates.
Establishment
Six-month-old slash pine seedlings were used in this greenhouse
trial. The seedlings were grown from half-sibling seeds in the greenhouse.
Ten seeds were sown in each of 35 closed plastic pots containing 1,750 g
of thoroughly mixed Ona fine sand (0-20 cm) collected from CRIFF site
A23, known to be responsive to both N and P (Pritchett and Smith, 1972).
Fertilizer was not added to these pots. Soil moisture was maintained be
tween 50 and 100% of field capacity by watering to predetermined weight
with distilled water. Day length was maintained at 14 hours using low-
intensity incandescent light. Seedlings were thinned to provide five
uniform seedlings in each pot at 6 months. At this stage, the seedlings
were removed from the pots and the roots cleansed of the sandy soil by
gentle agitation under water. The roots were then rinsed in distilled
water prior to transplanting immediately into the potted soils used in
the experiment. The roots of all seedlings were observed to be heavily
infected with mycorrhizal fungi.
The experimental design used was a randomized block design, with
two blocks each with nine treatments. The blocks consisted of two soils,
selected from the 10 bulk samples to provide soils of widely different
P-retention capacity and pH. Properties of the two soils, an Immokalee
fine sand from CRIFF site A16 and a McLaurin fine sandy loam from CRIFF
site A28, are shown in Table 4. Treatments consisted of a control, to
which amendments were not added, and eight P treatments, using the P


106
ever this was consistent with results reported in an earlier section
which showed these methods solubilized similar forms of P, although to
different degrees.
Unlike the resuits found in the preliminary screening, the effec
tiveness of the soil-test methods which extracted the greatest amounts
of P did not improve with increases in duration of the growing period.
In fact, values for all soil-test methods, except those for the Bray 2
method after 3 years, decreased with increasing growth periods from 1 to
5 years. Also R^ values for the relationships derived using all 72 sites
were considerably smaller than for those derived using only 10 sites in
the preliminary screening. However, the 10 sites used in the preliminary
screening were carefully selected to represent a range of P-responses and
replicates showing soil variability were excluded. This was not the case
when all sites were used so that soil variability within certain sites
probably contributed to the lower R values. Soil samples were
collected only from the control piots under the assumption that each ade
quately represented the experimental area of 12 plots.
The general decline in effectiveness of soil-test methods with in
creasing growth period can probably be attributed to several factors, or
their combination: (a) Results of analysis of foliage samples collected
from all plots by CRIFF personnel after the fifth year showed that, at
several sites, P concentrations in the foliage of trees in the highest P
treatments were at or below the accepted critical value of 0.085 % P.
Thus, the computed relative heights at age 5 years for these sites
probably underestimated the true responsiveness obtainable with adequate
P-fertilizer applications. (b) As trees grow older their root systems
exploit greater depths and volumes of soil. Consequently, P concentra
tions at depth greater than 20 cm may become important in determining


23
This aspect has received iittie attention in the literature, but its
importance is well illustrated by the work of Will (1366) discussed
earlier.
Extraction methods
Tamm (1964) pointed out that most soil analysis techniques were
those developed for agricultural crops and are of unknown value for pre
dicting tree nutrient requirements. Most techniques used in agriculture
today for determining plant-available P are largely empirical (Williams,
1962). However, considerable research relating the availability to
agricultural plants of various soil-P fractions and determining the solu
bility of these same fractions in the extractants used in soil-P tests,
has provided a theoretical basis for these soil-P tests (Thomas and
Peaslee, 1973)- Little research of this nature has been conducted with
trees. However, knowing the forms of P extracted by various soil tests,
some insight into the forms of P utilized by trees can be obtained.
Total P. Early work in Australia indicated that total soil P
provided a good index of site productivity and adequately delineated
areas of P deficiency (Kessell and Stoate, 1938; Young, 1940 and 1948).
This work helped foster the concept that trees could utilize sources of P
unavailable to other plants. This ability has been attributed to the
presence of mycorrhizae (Pritchett, 1968) which have been shown to solu
bilize otherwise insoluble sources of P (Rosendahl 1943). However, Tamm
(1964) has suggested the ability of tree roots to utilize P unavailable
to agricultural crops could be the consequence not so much of higher ef
ficiency as of longer persistence.
Stoate (1950) found that certain anomalies occurred when using
total P for predicting site nutrient status. He reported that these


79
soil becomes involved in meeting the trees' P requirements as growth peri
ods increase. However, it is unlikely that total P would ever become an
effective predictor of soi1-P status, even over very long growth periods
(rotation length, ca. 25 years) since certain fractions of the soil's
total P are occluded and extremely insoluble.
Relationships Between Soil-Test Values and P Uptake and Tissue P
Phosphorus uptake by seedlings during the first year's growth in
the greenhouse was most closely correlated with P extracted by weak ex
tractants e.g. H^O and NH^OAc (Table 11). However, P uptake over the 2-
year growth period was more closely correlated with P extracted by soil-
test methods which removed larger amounts of P from the soil (strong ex
tractants). In contrast to P uptake, P concentrations in seedling tops
at the end of the first year were more closely correlated with P extracted
by the stronger extractants than the weaker extractants. After two years
growth in the greenhouse the advantage of the stronger extractants over
the weaker ones for predicting P concentrations in the tops was more
evident.
2
The magnitude and ranking of the R values for the second year P
concentrations in tops and for foliar P of 4-year-old field trees were
strikingly similar. Raw data in Tables 7 and 8 show that both first and
second year P concentrations in tops of greenhouse grown seedlings were
closely correlated with foliar P of field trees (r = 0.897 and 0.875 re
spectively), although the P concentration in seedling tops at the end of
the first year in the greenhouse was quantitatively more closely related
to foliar P than was the P concentration in tops at the end of the second
year. Terman and Bengtson (1973) reported close agreement between their
established 'critical' P concentration in tops of 8-to 12-month-old slash


241
Table 57. Height and relative height of slash pine after
1, 3, and 5 years' growth on 72 sites in the
field
Field Repli- Height and year Relative height and year
site cate 1 3 5 1 3 5
cm %
A1
1
53.6
164
2
40.4
158
3
47.5
152
A2
1
29.7
96
2
33.0
105
3
31.5
106
A3
1
52.1
163
2
52.6
182
3
55.6
182
A 4
1
57-4
213
2
48.0
168
3
57.2
180
A5
1
30.5
147
2
27-2
131
3
34.8
158
A6
1
53-8
253
2
58.4
278
3
51.8
254
A7
1
42.7
172
2
39.1
136
3
42.7
169
A8
1
52.6
233
2
51.6
219
3
59.4
243
A10
1
68.1
180
2
73.9
173
3
61.5
173
A15
l
31.5
212
2
36.3
206
3
27.2
180
A16
1
54.6
230
2
53.1
227
3
49.0
195
263
72.7
68.1
62.9
283
46.1
58.8
72.1
253
64.5
61.3
57.1
266
63.8
78.9
97.0
269
74.5
78.7
88.3
269
75-1
72.3
76.6
302
89.8
85.5
97.0
313
90.1
96.9
100.0
311
96.0
92.0
91.5
397
94.0
97.4
98.5
330
75-3
74.6
90.7
352
93-6
67-9
73.3
314
84.5
100.0
100.0
300
88.3
94.7
98.7
342
100.0
100.0
100.0
455
99.5
94.2
93.6
483
91.5
94.2
97.2
467
83.2
85.8
92.6
356
75-8
94.2
88.6
305
61.6
51.7
57.2
354
63-3
65.1
69.5
447
98.5
100.0
98.3
421
89-3
92.0
91.8
451
91.8
95.0
94.1
419
100.0
95.5
97.9
410
100.0
80.6
90.4
392
100.0
97.3
92.7
417
83.8
89.8
88.9
412
100.0
89.3
90.1
362
83.0
90.1
90.8
474
97.2
81.8
86.7
467
93.8
91.5
94.5
430
97.5
79.4
82.8


Table 15- Multiple correlation coefficients (R^) for relationships between solubility of P compounds in
chemical extractants and the uptake of P from these compounds by slash pine seedlings grown on
two soils in the greenhouse
P compound
treatmentt
Soil-
test method
h2o
NH^OAc
T ruog
HC1-H2SO4
01 sen
Lactate
Bray 1
Bray 2
NH^F
r2
Immokalee
fine sand(n=8)
No soil
0.706**
0.837**
0.841**
0.9^8**
0.655*
0.936**
0.712**
0.736**
0.427
+ Soil
0.913**
0.793**
0.827**
0.913**
0.478
0.949**
0.498*
0.889**
0.677*
+ So i 1
(incubated)
0.973**
0.936**
0.962**
0.931**
0.954**
0.912**
0.750**
0.766**
0.841**
McLaurin fine sandy loam(n=8)
No soil
0.546*
0.2 76
0.137
0.275
O.783**
0.265
0.429
0.312
0.681*
+ Soi 1
0.374
0.383
0.147
0.220
0.737**
0.379
0.833**
0.480
0.942**
-+ Soil
(incubated)
0.838**
0.879**
0.434
0.383
0.838**
0.584*
0.787**
0.450
0.853**
Immokalee
+ McLaurin(n=l6)
No soil
0.380*
0.354*
0.307*
0.390**
0.411**
0.383*
0.355*
0.322*
0.300*
+ Soi 1
0.530**
0. **82**
0.337*
0.397*
0.383*
0.450**
0.388**
0.483**
0.550**
+ Soil
(incubated)
0.958**
0.832**
0.594**
0.482**
0.647**
0.561**
0.551**
0.421**
0.600**
^ Treatment prior to extraction.
* Significant at
of P compound.
the 5% level,
using the model Y=b logX + c, where Y=P uptake above control and X=Solubility
**Significant at the
1 % 1evel.
o
o


242
Table 57- Continued
Field
Repl i -
Height and
year
Relative
: heiqht
and year
s i te
cate
1
3
5
1
3
5
A17
1
35.8
cm--
159
391
78.8
%
78.5
87.6
2
31.0
1 19
325
11.1
72.3
82.4
3
31.5
117
307
65.3
60.8
73.2
A18
1
88.9
286
**83
100.0
90.7
92.0
2
8^4.3
269
**81
92.6
81.8
85.8
3
86.6
262
576
95-2
81.6
100.0
A19
1
90.9
300
501
92.5
82.4
81.3
2
93.7
291
504
100.0
85.8
87.7
3
92.5
28**
505
89.1
90.1
93.0
A20
1
39-1
126
268
53.1
42.7
48.3
2
35.1
103
221
45.5
40.1
43.6
3
36.6
102
187
51.7
36.4
33.8
A21
1
44.5
216
*33
100.0
100.0
100.0
2
**0.6
162
417
95.5
100.0
100.0
3
**6.7
183
452
83.9
94.3
98.7
A22
1
5**.3
206
358
97.0
100.0
97.1
2
66.8
217
357
88.8
81.0
88.1
3
79.5
21 1
395
100.0
96.2
100.0
CO
CNI
<
1
51.6
177
350
10.1
100.0
100.0
2
52.3
157
351
86.3
76.4
78.5
3
**9.0
167
356
79-8
81.8
84.5
A24
1
26.9
1**9
378
70.1
82.4
93.8
2
25-9
150
370
67-6
84.4
90.1
3
25.**
139
355
69.I
82.2
88.0
A25
1
27-9
121
263
96.2
93.0
90.9
2
21.k
13*
277
87.8
100.0
100.0
3
30.2
133
285
94.1
100.0
100.0
A26
1
*3-7
2**0
490
100.0
98.4
95.0
2
*3.9
21**
452
99.3
99-9
98.3
3
31.1
20**
450
96.5
94.5
94.9
A27
1
**0.1
205
447
82.8
83.3
96.8
2
***t.7
221
458
84.5
100.0
100.0
3
3**.0
178
404
68.0
76.7
91.9
A28
1
39-*
203
434
84.0
89.1
96.3
2
36.6
200
430
10.1
12.2
82.4
3
38.9
210
429
80.0
82.9
89.5


Table 13. Relationships between selected soil-test values and height growth in the
absence of P fertilizer of slash pine in field and greenhouse experiments
on 10 soils
Soi1-test
method or
P form
Mean
extract-
able P
Height growth
Greenhouse
Field
lf
2
1
3
5
r2
ppm
h2o
2.3
0.411*
0.506*
0.525*
0.421*
0.258
NH^OAc
4.4
0.307
0.570*
0.317
0.332
0.178
T ruog
8.9
0.342
0.562*
0.085
0.113
0.037
01 sen
12.3
0.218
0.601**
0.067
0.147
0.059
hci-h2so4
15.2
0.262
0.624**
0.078
0.150
0.062
H2S\
32.5
0.159
0.503*
0.006
0.045
0.006
Bray 1(3)
41.3
0.187
0.607**
0.018
0.093
0.034
NH4F(pH 8.5)
57.0
0.145
0.268
0.001
0.031
0.003
Organic-P
42.6
0.288
0.176
0.083
0.052
0.056
Total-P
136.1
0.099
0.515*
0.090
0.030
0.048
+ _
1 Tree age at
time of measurement (years
).
* Significant
at the 5%
level, using the
model Y -
b logX + c, where Y =
height
and X = soil-test value.
** Significant at the ]% level.


264
Huffman, E. 0. 1968. Behaviour of fertilizer phosphates. Trans. Intnl.
Cong. Soil Sci. (Australia) 2:745754.
Humphreys, F. R. 1963. Some limitations in the use of foliage and soil
analysis. Tech. Paper For. Comm. NSW, no. 1. 6 p.
Humphreys, F. R., and R. Truman. 1964. Aluminum and phosphorus require
ments of Pinus radiata. Plant and Soil, 20:131-134.
Humphreys, F. R., and W. L. Pritchett. 1971. Phosphorus adsorption and
movement in some sandy forest soils. Soil Sci. Soc. Amer. Proc.
35:495-500.
Humphreys, F. R. ,'and W. L. Pritchett. 1972. A soil testing procedure
for predicting phosphorus status of P i us e11 iott i i plantations.
Plant and Soil. 37:495-487.
Jackson, M. L. 1958. Soil chemical analysis. Prentice Hall Inc.,
Englewood Cliffs, N. J. 498 p.
Jackson, D. S. 1973. Soil factors that should influence allocations of
land for forestry and agriculture. New Zeal. J. For. 18:55-62.
Jerven, 0. 1967. A brief summary on the evolution of forest fertiliza
tion in Norway, p. 4954. J_r^ Forest fertilization. Proc. Vtn
Colloquium of the International Potash Institute, Jyvaskyla, Fin
land.
John, M. K. 1970. Colorimetric determination of phosphorus in soil and
plant materials with ascorbic acid. Soil Sci. 109:214-220.
Juo, A. S. R., and B. G. Ellis. 1968. Chemical and physical properties
of iron and aluminum phosphates and their relation to phosphorus
availability. Soil Sci. Soc. Amer. Proc. 32:216-221.
Kawana, A. 1969. Forest fertilization in Japan. J. For. 67:485-487.
Kessell, S. L., and T. N. Stoate. 1938. Pine nutrition. West Austr.
For. Dept. Bull. 50. 45 p.
Krause, H. H. 1973- Forest fertilization in Eastern Canada, with
emphasis on New Brunswick studies, p. 188-205. j_n_ Forest fertili
zation symposium proceedings. USDA For. Serv. General Tech. Rep.
NE-3.
Kuo, S., and E. G. Lotse. 1972. Kinetics of phosphate adsorption by
calcium carbonate and Ca-kaolinite. Soil Sci. Soc. Amer. Proc.
36:723-729.
Kurtz, L. T. 1942. Elimination of fluoride interference in the molyb
denum blue reaction. Indus. Engin. Chern. Analyt. Ed. 14:855.


Table>55* Cohtinued
Field
s i te
Repli-
cate
*
Depth
H2
P
NH.OAc
4
P
HC1-
H2S4
P
Bray 1 (3)
P
Bray 2
P
NH,
4
OAc
pH
(h2o)
Ca
Mg
K
A1
N. 1 1 '
ppm
A6
2
0-20B
4.6
4.4
5-1
5-8
7.6
280
76
18
12
4.0
0-20
4.2
3.1
4.4
5.2
5-3
230
72
12
12
4.2
20-40
0.8
1.8
3.4
9-6
9-3
45
25
12
12
4.4
40-60
0.4
2.6
15-6
54.6
70.8
15
21
2
138
4.7
3
0-20B
5.2
3.1
5.0
7.0
5-9
250
56
10
6
4.3
0-20
5-4
4.4
5.7
7.8
6.3
230
50
12
9
4.2
20-40
0.5
2.6
11.5
35.6
33.4
30
10
2
40
4.7
40-60
0.4
2.2
13-5
44.4
45.8
15
2
2
128
4.7
A7
1
0-20B
0.5
1.1
2.4
4.7
4.4
15
19
7
78
4.3
0-20
0.3
0.9
1.9
3.3
3.1
15
19
4
78
4.4
20-40
0.1
0.7
0.8
1.8
1.7
15
19
2
71
4.6
40-60
0.1
0.5
0.6
1.3
1.5
15
14
1
73
4.7
2
0-20B
0.3
0.9
2.2
3.9
3.8
15
14
4
81
4.3
0-20
0.3
0.7
1.7
3.2
3.3
15
14
4
76
4.5
20-40
0.1
0.7
1.1
2.2
2.1
15
14
2
73
4.6
40-60
0.1
0.5
0.8
1.3
1.9
15
5
2
76
4.7
3
0-20B
C.4
0.9
2.6
4.6
5.4
30
10
18
86
4.3
0-20
0.3
0.9
1.9
4.0
3-7
15
5
7
81
4.4
20-40
0.1
0.7
1 .0
2.0
1 -9
15
2
2
76
4.5
40-60
0.1
0.5
0.8
1.5
1.7
15
21
2
71
4.6
A8
1
0-20B
0.7
1.3
4.3
13.3
12.4
60
19
2
25
4.8
0-20
0.6
1.8
8.1
23.5
23.2
45
19
2
33
5.0
20-40
0.2
0.9
7.3
23.2
20.3
30
14
2
28
5-2
N>
NJ


2.67
Moreno, E. C., W. E. Brown, and G. Osborn. I960. Stability of dicai
cium phosphate dihydrate in aqueous solutions and solubility of
octacalcium phosphate. Soil Sci. Soc. Amer. Proc. 24:99-102.
Moreno, E. C., W. L. Lindsay, and G. Osborn. I960. Reactions o> dical
cium phosphate dihydrate in soils. Soil Sci. 90:58-68.
Moschler, W. W., G. D. Jones, and R. E. Adams. 1970. Effects Oi lob
lolly pine fertilization on a Piedmont soil: growth, foliar com
position, and soil nutrients 10 years after establishment. Soil
Sci. Soc. Amer. Proc. 34:683-685.
Murphy, J., and J. P. Riley. 1962. A modified single solution method
for the determination of phosphate in natural waters. Anal. Chim.
Acta. 27:31-36.
Mustanoja, K. J., and A. L. Leaf. 1965. Forest fertilization research
1357-1964. Bot. Rev. 31:151-246.
O'Carroll, N. 1967. Forest fertilization in the Republic of Ireland,
p. 271-274. j_n Forest fertilization. Proc. Vth Colloquium of the
International Potash Institute, Jyvaskyla, Finland.
Olsen, S. R., and A. D. Flowerday. 1971. Fertilizer phosphorus inter
actions in alkaline soils, p. 153-185. In Ferti 1izer technology
and use. Soil Sci. Soc. Amer., Madison, Wis.
Olsen, S. R., and F. S. Watanabe. 1957- A method to determine a phos
phorus adsorption maximum of soils as measured by the Langmuir
isotherm. Soil Sci. Soc. Amer. Proc. 21:144-149.
Ozanne, P. G., and T. C. Shaw. 1967. Phosphate sorption by soils as
a measure of the phosphate requirements for pasture growth. Austr.
J. Agrie. Res. 18:601-612.
Paarlahti, L. H. 1967. Forest fertilization experiments in Finland.
P* 311-312. J_n_ Forest fertilization. Proc. Vth Colloquium of the
International Potash Institute, Jyvaskyla, Finland
Page, N. R., G. W. Thomas, H. F. Perkins, and R. D. Rouse. 1965. Pro
cedures used by state soil-testing laboratories in the southern
region of the United States. Southern Cooperative Series Bull. 102.
49 P-
Pawluk, S., and H. F. Arneman. 1961. Some forest soil characteristics
and their relationship to jack pine growth. For. Sci. 7:160-172.
Pritchett, W. L., and K. R. Swinford. 1961. Response of slash pine to
colloidal phosphate fertilization. Soil Sci. Soc. Amer. Proc 25*
397-400.
Pritchett, W. L., and W. R. Llewellyn. 1966. Response of slash pine
to phosphorus in sandy soils. Soil Sci. Soc. Amer. Proc 30-
509-19. '


predictors of P fertilizer needs than more conventional test methods for
available P.
An examination of the solubility of P compounds (mono and dicalcium
phosphates, f1uorapatite, colloidal A1 and Fe phosphates, K taranakite,
wavellite, and strengite) in extractants of several common soil-P test
methods and their utilization by slash pine seedlings, showed that soil-
test methods which provided the best index of early growth and response
in preliminary screenings extracted little more P from these compounds
than did H20 alone. Methods which were more successful predictors over
longer growth periods extracted appreciable quantities of P from the Ca
phosphates, colloidal phosphates, and K taranakite, which were effec
tive P sources for seedlings; but, they did not extract much P from
wavellite and strengite, which were ineffective P sources for seedlings.
Methods utilizing strong acids did, however, overestimate the availa
bility of fluorapatite to seedlings on a soil of pH > 5-
Five soil-test methods (H20, NH^OAc pH 4.8, 0.05N_ HC1 + 0.025N_
H2S0it, 0.03 NH^F + 0.025N. HC 1 0.03^ NH^F + 0.1N HC1) selected on the
basis of results from the preliminary screening, were calibrated against
slash pine growth and response information obtained from 72 field ferti
lizer trials 1, 3, and 5 years after establishment. The method involving
use of 0.05]^ HCI + 0.025N^ H2SO^ provided the best prediction of height
response and P fertilizer requirements over the 3 and 5_year growth
periods. A surface soil (0-20 cm) value of 5 ppm by this method pro
vided an effective delineation of P responsive sites. Soils testing
between 5-0 and 2.5 ppm P required ca. 20-40 kg P/ha and those testing
below 2.5 ppm P required ca. 40-80 kg P/ha to provide an adequate P
supply to slash pine over the above growth periods.
xv 1


160
consequently, have virtually no P-retention capacity. The possible con
tribution of Ca to the above correlation in these soils, which have a mean
pH of b.3b, was probably minimal.
Relationships between P retention and such soil properties as clay,
silt, pH, and organic matter are frequently considered to be indirect
through the association of these properties with the causal agents of P
retention, such as the various forms of Fe and A1. Partial correlations
(recorded in parentheses in Table 39) > showed that correcting the signif
icant relationships for the indirect association through A1 (NH^OAc-
extractable) significantly reduced correlations in all cases, except for
pH. The relationship between P retention and soil pH was independent of
any effects soil pH may have had on NH^OAc-extractable A1, while that of
clay was in part related to its association with NH^OAc-extractable A1,
but it also had a significant independent effect. Clay, particularly 1:1
lattice clays which are prevalent in many Southeastern soils (Fiskell and
Perkins, 1970), may contribute directly to P fixation particularly St low
soil pH values (Black, 19^3)- The possibility that clay levels may be
related to active Fe, or active forms of A1 not extracted by NH^OAc may
also have contributed to the significance of the independent relationship.
The positive effect of pH on P retention within specific A] levels ap
pears to have r.o explanation in terms of current knowledge of the effects
of pH on P retention by sesquioxides and a 1umino-si 1icate clays. The
explanation probably lies in the association of low pH with podzolized
soils within which most of the sesquioxides have been elluviated from
surface horizons into the spodic horizon. At specific levels of NH^OAc-
A1 increasing pH is probably associated with the presence of higher
levels of active Fe and/or active forms of A1 not extracted by NH.OAc.


34
symptoms of fused needle in Australia. Young (1940) showed that fused
needle, a symptom of unthrifty pine trees, could be corrected by the ap
plication of P fertilizer. On the basis of significant correlations be
tween levels of P extracted by the Truog test and the productivity of
both jack and red pine, Wilde et al. (1964 a, 1964 b) proposed that
minimum P levels required for the establishment of these species were
6.5 and 5.4 ppm respectively in the surface of 15 cm of soil. Ballard
(1970 a) proposed that values of 3-5 and 5-0 ppm P by the Olsen and Bray
2 tests were required in the surface 10 cm of soil for satisfactory
growth of radiata pine.
Minimum or critical values established from relationships between
soil P and productivity cannot be used with confidence for predicting
ameliorative practices for sites testing below these values. Dahl,
Selmer-Anderssen, and Saether (1961) pointed out that a significant cor
relation between nutrient levels in the soil and site productivity is
not proof that this nutrient is controlling growth. For instance, these
authors reanalyzed the data of Viro (1955), who recorded a significant
correlation between EDTA-extractab1e P and the site index of Scots pine
(P. sy1 vestris). They found that extractable P was related to site
index only indirectly through its correlation with extractable soil Ca
levels. Proof of a casual relationship must be obtained by either field
fertilizer trials or the establishment of a positive relationship
between site productivity and both the concentration of a nutrient in
the foliage and its availability in the soil (Leyton and Armson, 1955).
Prediction of fertilizer response. Before soil-test results can
be used in developing fertilizer recommendations, field studies must show
that the test will be useful for differentiating soils into those groups


126
and response to P fertilizer, they can be used with, some confidence for
prediction purposes on sites with soil properties within the range en
countered in the calibration work.
Relationships Between Soil-Test Values and P-Fertil?zer Requirements
Multiple correlation coefficients (R^) for relationships between
the amount of P extracted from the soil by the five soil-test methods and
the amount of P fertilizer required to achieve 90, 95, and 100% of
maximum height growth after 1, 3, and 5 years' growth in the field are
presented in Table 27. Actual amounts of fertilizer required to achieve
these growth rates at each of the 72 sites are given in Appendix Table 58.
The effect of increasing growth period on the effectiveness of the
five soil-test methods as predictors of fertilizer requirements was
similar to that observed for relative height. The amounts of P extracted
by a weak extractant (NH^OAc) were most closely related to P requirements
of 1-year-old trees, while at ages 3 and 5 years, P extracted by the
stronger extractants was most closely related to fertilizer P require
ments. While the HCl-i^SO^ method was the most effective predictor of
relative height at ages 3 and 5 years, there was a tendency for the Bray
methods to be the most effective predictors of fertilizer requirements of
trees at ages 3 and 5 years.
As was the case for relative height, only a very small proportion
of the variation in fertilizer requirements at age 5 years was explained
by extractable P in the surface soil. Factors such as the contribution
of P from lower soil depths, the insensitivity of the height measurement
at age 5 years, inadequate fertilizer applications on some sites, and
the increased length of time for other site factors to influence the ef
fectiveness of fertilizer may have contributed to variation unaccountable


Table 55- Continued
Field
s i te
A20
A21
A22
HC1 -
Repli- H20 NH^OAc Bray 1(3) Bray 2 NH^OAc
Depth* P P P P P Ca Mg K 1 (HO)
cm ppm
3
0-20
0.4
0.7
2.1
2.5
5.1
15
38
10
86
4.3
20-40
0.2
0.5
0.7
0.9
1.6
15
34
2
42
4.6
40-60
0.2
0.1
0.4
0.5
0.9
30
30
2
45
4.7
1
0-20B
1 .3
10.1
56.3
174.2
212.0
45
50
10
62
5.0
0-20
1 .3
10.5
70.2
197.0
239-0
15
48
4
62
5.1
20-40
0.6
11.8
103.0
245.5
284.0
30
42
2
64
5.1
40-60
0.3
14.0
109.0
231.8
285.0
15
34
2
73
5.0
2
0-20B
1.5
2.2
3.1
6.2
7.0
75
38
4
19
4.5
0-20
1.1
1 1
2.5
5-8
6.7
60
38
4
14
4.5
20-40
0.4
12.2
3-0
8.5
8.6
90
38
2
14
4.8
40-60
0.9
10.9
99-3
209.1
251.0
15
38
2
66
5.0
3
0-20B
1.6
10.9
55-5
149.0
170.0
90
50
7
62
5.0
0-20
1.4
10. 1
53-9
141.8
161.2
60
50
7
66
5.0
20-40
0.9
11 .8
84.8
168.2
203.0
30
48
2
73
5.2
40-60
0.7
4.4
13.3
35.8
33.5
30
38
2
28
5.0
1
0-20B
3.0
3.1
4.7
5-9
7.2
120
50
10
14
4.2
0-20
2.2
4.0
3-5
5.0
5.0
105
50
15
16
4.3
20-40
1.4
5.7
26.6
75.2
67.1
15
48
L
52
4.5
40-60
0.7
4.8
74.0
206.1
171.5
15
42
2
92
4.6
2
0-20B
3.3
3.5
4.9
7.1
7.7
135
42
12
16
4.3
0-20
2.7
2.6
4.0
5.5
7-4
135
61
10
16
4.3
20-40
1 .2
2.6
6.2
21.2
16.4
45
38
2
21
4.6
40-60
0.9
6.6
54.5
187.9
149.0
15
34
4
88
4.7


TABLE OF CONTENTS (Continued)
Page
Harvesting 51
Extraction of P Compounds 51
Phosphorus-Retention Study 51*
Soil and Foliage Samples 5**
Determination of P Retention 5^
Sample Analysis 57
Soil Characterization 57
Soil P Analys is 57
Soil A1 and Fe Analysis 65
Plant Tissue Analysis 65
Statistical Analysis 67
RESULTS AND DISCUSSION 68
Preliminary Screening of Soil-Test Methods 68
Relationships Between Soil-Test Values and Relative
Height 72
Relationships Between Soil-Test Values and P Uptake
and Tissue P 79
Relationships Between Soil-Test Values and
Fertilizer Requirements 8l
Relationships Between Soil-Test Values and Height . 83
General Discussion 85
Phosphorus Compounds 88
Solubility of P Compounds in Chemical Extractants . 88
Monocalcium phosphate (MCP) 88
Dicalcium phosphate (DCP) 90
Fluorapatite (FA) 91
Colloidal aluminum phosphate (CA1P) 92
Potassium taranakite (KTK) 92
Wavel 1 i te (WA) 93
Colloidal ferric phosphate (CFeP) 93
Strengite (STR) 9^
Utilization of P Compounds by Slash Pine Seedlings 9^
Relationships Between Seedling Utilization and
Solubility of P Compounds 99
General Discussion 102
Field Calibration of Selected Soil-Test Methods 105
Relationships Between Soil-Test Values and Relative
Height 105
Effect of soil sampling position and depth ... 115
Relationships Between Other Soil and Site
Parameters and Relative Height 117
Relationships Between Soil-Test Values and P-
Fertilizer Requirements 126
Effect of soil sampling position and depth . 131
vi


DRY MATTER, gm/pot
37
PHOSPHORUS COMPOUND
Fig. 3* Dry matter of slash pine seedlings after 8
months' growth on two soils treated with
eight P compounds.


57
Sample Analysis
Soil Characterization
Bulk density of all soil samples collected using a closed-cylinder
soil auger was computed from the dry weight of the samples and their
volume. The volume was calculated from the diameter of the auger, the
depth of sampling and number of cores. Particle-size distribution was
determined by the hydrometer method (Bouyoucus, 1951). Moisture contents
at 15, 1/3, and 1/10 atmospheres were determined by use of a pressure-
membrane apparatus (Richards, 1965). Loss on ignition was determined as
the weight loss (%) of an air-dried sample following ignition in a
furnace at 550C for 1 hour.
Soil pH was determined potent iometrica1ly by insertion of a com
bination glass electrode assembly in the supernatant of a 1:2 soil-water
and/or soil-N^ KC1 suspension. Soil organic matter was determined by a
modified Walkley-Black method (Allison, 1965); total N by the macro-
Kjeldahl procedure (Bremner, 1965); and cation exchange capacity (CEC)
by NH^ saturation using N_ NH/jOAc at pH 7-0 (Chapman, 1965). Extract-
able cations were determined in the filtrate following extraction with
0.7^NH^0Ac + 0.5^N_ HOAc buffered at pH k.8. Calcium and Mg in the
filtrates were determined by atomic absorption (Perkin-Elmer 303) and K
by flame emission (Beckman B) spectroscopy. Lanthanum chloride was added
to suppress anion interference in the Ca and Mg determinations (Breland,
1966). These analyses were carried out by the Analytical Research Lab
oratory of the Soil Science Department, University of Florida.
Soil P Analysis
Reagents used to extract soil P, and details of the extraction


Table 10. Relationships between selected soil-test values and relative
height growth of slash pine in field and greenhouse experi
ments on 10 soils
Soi1-test
method or
P form
Mean
extract-
able P
Relative height
Greenhouse
Field
1 +
2
1
3
5
d2_
ppm
h2
2.3
0.951**
0.715**
0.614**
0.217
0.055
NH^OAc
4.4
0.863**
0.729**
0.817**
0.589**
0.348
T ruog
8.9
0.658**
0.440*
0.658**
0.657**
0.481*
01 sen
12.3
0.533*
0.515*
0.702**
0.787**
0.666**
HC1 -H2S04
15.2
0.579*
0.538*
0.722**
0.782**
0.639**
h2so/4(9)
32.5
0.349
0.331
0.540*
0.765**
0.716**
Bray 1 (3)
41.3
0.187
0.478*
0.623**
0.845**
0.778**
NHAF(pH8.5)
57.0
0.296
0.060
0.465*
0.718**
0.699**
Organic-P
42.6
0.050
0.010
0.011
0.040
0.056
Tota 1-P
136.1
0.040
0.001
0.065
0.229
0.306
+ Tree age at time of measurement (years).
* Significant at the 5% level, using the model Y = b logX +c, where
Y = relative height and X = soil-test value.
** Significant at the \% level.


Table 24. Simple correlation coefficients(r) between selected soil and site properties of 72 field sites
Soil or site
property
Extractable P
pH
NH^OAc
Silt +
clay
Depth
to LH
Drainage
class
y
HCl-H2S0i<
Ca
A1
pH
-0.713**
0.250*
1.000
-
-
-
-
Organic matter
0.168
-0.097
-0.394**
0.036
0.041
0.360**
-0.227
-O.35I**
N i t rogen
-0.047
-0.157
-0.293*
-0.004
0.207
0.458**
-O.289*
-O.387**
CEC
0. 101
-0.124
-0.295*
0.263*
0.202
0.494**
-0.207
-O.3I3**
NH^OAc-Ca
0.249*
-0.113
-0.033
1 .000
-
-
-
-
-Mg
0.469**
-0.075
-O.337**
0.727**
0.092
0.344**
-0.168
-0.185
-K
0.077
-0.199
0.009
0.535**
0.377**
0.639**
0.067
0.099
-A 1
-0.623**
-0.126
0.401**
0.019
1 .000
-
-
-
Silt + clay
-0.424**
-O.263*
0.271*
0.274**
0.715**
1 .000
-
-
Ava i Iable moisture
-0.329**
-O.I80
0.169
0.215
0.587**
0.907**
-0.030
-0.055
Depth of A1 horizon
0.141
0.232*
-0.007
-0.216
-0.177
-0.252*
0.122
-0.063
Depth to LH^
-0.070
0.292*
0.602**
0.021
-0.161
-0.049
1 .000
Drainage class
-0.224
0.188
0.697**
0.012
-0.017
0.077
0.852**
1 .000
* Significant at the 5% level.
**Significant at the 1% level.
' Depth to limiting horizon.
N>


133
prediction of laboratory-determined P retention were calibrated against
the amount of P, applied as CSP, retained in the surface horizon of 10
soils sampled years after P fertilizer application in the field.
Phosphorus retention in these soils ranged from negligible in many
of the A1 horizons of Spodosols to very high (l,200 ygP/g soil) in some
of the finer-textured Ultisols. However, there was considerable overlap
in P-retention capacities between soil orders. The amount of A1 and Fe
extracted by a range of methods provided the best index of the P-retention
capacity. Several other soil properties, including pH, silt plus clay,
and exchangeable Ca were also significantly related to P retention, but
the relationships were mainly indirect through association with A1 and
Fe levels. Aluminum extracted by soil-P test methods (NH^OAc, HCl-H^SO,,
Bray 1, Bray 2) was as effective for predicting P retention as that ex
tracted by methods conventionally used to characterize soil A1, such as
KC1 leaching, pyrophosphate, oxalate, and CDB extraction. Foliar levels
of A1 and Fe were only poorly related to the P-retention capacity of
soils. Retention in the surface 20 cm of soil of P applied as CSP h
years previously ranged from -1^ to +125%. Retention was essentially in
dependent of the rate of application of CSP and was closely related to A]
extracted by the four soil-P test methods. Critical levels of A1 for
each of these soil-test methods were proposed, below which excess leach
ing losses of soluble forms of P fertilizers could be expected.
The following conclusions appear to be justified from the results
of this investigation:
1. The amount of soil P available for utilization by slash pine seedlings
during their initial year of development is determined mainly by the
amount of P in the solution phase in equilibrium with solid-phase P.


Table 58. Phosphate fertilizer (CSP) required to achieve 90, 95, and 100 % of maximum height
of slash pine 1, 3, and 5 years after ferti iization of 72 field sites
Field
s i te
Repli-
cate
Ferti1
1 izer P requ
i rement
1 year
3 years
5 years
90
95
100
90
95
100
90
95
100
. 1 /U-
A1
1
21
35
55
28
38
64
53
69
1 10
2
31
38
55
27
35
53
21
30
53
3
34
45
67
46
61
94
34
44
66
A2
1
26
35
55
26
41
81
0
0
90
2
27
40
73
25
41
82
22
62
90
3
19
29
54
20
29
50
17
26
48
A3
1
71
81
90
30
63
90
0
0
90
2
0
17
56
0
0
90
0
0
0
3
0
0
34
0
9
39
0
10
40
A4
1
0
43
90
0
0
90
0
0
90
2
24
35
64
20
30
54
0
19
80
3
0
8
71
21
29
48
18
26
45
A5
1
13
29
67
0
0
0
0
0
0
2
4
20
59
0
2
69
0
0
27
3
0
0
0
C
0
0
0
0
0
A6
1
0
0
18
0
3
47
0
4
43
2
0
13
58
0
15
69
0
0
90
/. ) '
3
1 1
21
47
8
20
55
0
8
46
A7
1
19
29
55
0
0
64
3
16
47
2
31
40
64
30
37
55
27
35
53
3
30
39
62
26
35
'55
26
36
59


Table 56. Continued
Field
s i te
Repli-
cate
Depth
Particle
size distribut ion
Moisture
content
Bulk
density
C1 ay
Silt
Sand
1/3 atm.
15 atm.
y
g/cc
'0
A26
1
0-20
5.5
12.6
81.9
7.6
4.0
1.26
20-40
6.3
11.7
82.0
5-9
2.5
1.75
40-60
8.4
1 1.0
80.6
6.6
2.7
1.75
2
0-20
5.2
12.8
82.0
6.3
4.1
1.22
20-40
6.6
11.7
81.1
5.0
2.6
1.77
40-60
7.2
11.0
81.5
4.6
2.7
1 .98
3
0-20
3.7
12.0
84.3
6.1
4.1
1.31
20-40
4.5
10.9
84.6
4.2
2.2
1.67
40-60
4.8
11.6
83.6
5-0
1.8
1.71
A27
1
0-20
8.7
23.2
68.1
11.0
4.5
1.11
20-40
16.2
19-7
64.1
11.2
5.2
1.53
40-60
15.1
20.7
60.2
12.1
6.3
1.62
2
0-20
9.0
26.8
64.2
10.6
5.0
1 .30
20-40
17.1
23-4
59.5
12.0
5-4
1.59
40-60
21.0
23.9
55.1
12.8
6.6
1.60
3
0-20
10.6
27.9
61.5
13.2
5.2
1.24
20-40
17-9
27-7
54.4
12.9
5.7
1 .71
40-60
19-0
27-9
53.1
13-9
6.8
1 .76
A28
1
0-20
8.2
21.3
70.5
10.0
4.4
1 .21
20-40
12.8
20.0
67.2
9.7
4.3
1.53
40-60
15-1
£0.4
64.5
10.1
4.7
1.54
2
0-20
7.3
23-5
69.2
10.3
3.9
1.21
20-40
14.5
20.9
64.6
10.5
4.7
1 -59


Table 27.
Relationships between
soil-test values at
three soil
depths and P
fertilizer
requ1 red
to achieve
90, 95,and
100% of maximum height
growth after 1, 3
and 5 years'
growth on
72 field
s i tes
Fert i
1izer requirements
Soil
1 year
3 year
5 year
Depth
90
95
100
90
95
100
90
95
100
cm
0-20B
p2
0.271**
0.250**
0.041
0.039
h2o
0.000
0.001
0.005
0.005
0.002
0-20
0.209**
0.192**
0.035
0.035
0.000
0.000 .
0.005
0.003
0.003
20-40
0.223**
0.206**
0.041
0.070*
0.010
0.002
0.012
0.004
0.001
40-60
0.318**
0.298**
0.083**
0.100**
0.019
NH/,0Ac
0.023
0.007
0.001
0.002
0-20B
0.359**
0.364**
0.094**
0.116**
0.034
0.014
0.014
0.009
0.001
0-20
0.332**
0.334**
0.106**
0.098**
0.023
0.012
0.013
0.004
0.000
20-40
0.265**
0.266**
0.082*
0.164**
0.127**
0.090*
0.038
0.026
0.056*
40-60
0.242**
0.234**
0.082*
0.199**
0.129**
HCl-H2SO/4
0.110**
0.056*
0.031
0.053
0-20B
0.258**
0.277**
0.085*
0.189**
0.118**
0.071*
0.052
0.022
0.044
0-20
0.181**
0.199**
0.074*
0.173**
0.093**
0.070*
0.054*
0.015
0.029
20-40
0.180**
0.182**
0.060*
0.235**
0.224**
0.118**
0.090*
0.053*
0.082*
40-60
0.239**
0.240**
0.070*
0.285**
0.191**
Bray 1(3)
0.102**
0.135**
0.075*
0.074*
0-20B
0.198**
0.224**
0.076*
0.194**
0.144**
0.095**
0.056*
0.023
0.065*
0-20
0.164**
0. 180**
0.070*
0.170**
0.120**
0.091**
0.047
0.014
0.050
20-40
0.303**
0.309**
0.130**
0.164**
0.153**
0.063*
0.026
0.016
0.043
40-60
0.328**
0.335**
0.123**
0.217**
0.136**
Bray 2
0.057*
0.053*
0.041
0.039
0.069*
0-20B
0.147**
0.170**
0.048
0.192**
O.56**
0.Ill**
0.070*
0.022
0-20
0.103**
0.120**
0.051
0.164**
0.126**
0.111**
0.058*
0.012
0.052
20-40
0.124**
0.124**
0.046
0.241**
0.252**
0.126**
0.114**
0.058*
0.096**
40-60
0.185*
0.182**
0.048
O.315**
0.230**
0.109**
0.188**
0.097**
0.087*
* Significant at the 5% level. ** Significant at the level, using the model Y=b logX + c.
Ni
Vl


Table 51
Multiple correlation coefficients (R^) for relationships between soil-test values and tree para
meters from 10 selected field trials
Mean Foliar
Soi1-test
extract-
Height
Rel
. height
Fert. reqm,
P
method
able P
1*
3
5
1
3
5
1
3
5
4
ppm
,r2
h2o
2.2
0.525+
0.421
0.258
0.614
0.217
0.055
0.508
0.088
0.000
0.228
H20(2)
7.1
0.569
0.416
0.324
0.328
0.038
0.001
0.212
0.007
0.004
0.029
h20(3)
0.3
0.271
0.253
0.126
0.485
0.424
0.199
0.415
0.083
0.003
0.407
h2o(4)
243.8
(ug P)
0.433
0.441
0.306
0.554
0.255
0.081
0.415
0.070
0.000
0.250
NaC 1
2.2
0.403
0.378
0.225
0.662
0.401
0.174
0.558
0.139
0.006
0.382
Na2S0zj
2.3
0.406
0.385
0.232
0.670
0.402
0.175
0.559
0.139
0.006
0.386
Na2MoOZj
7.5
0.104
0.210
0.146
0.622
0.466
0.399
0.473
0.399
0.222
0.453
Na2B/(0y
2.4
0.505
0.430
0.275
0.614
0.287
0.083
0.491
0.101
0.003
0.258
Na2BZ,07(2)
15.0
0.122
0.212
0.098
0.771
0.767
0.600
0.748
0.456
0.146
0.785
NH^OAc
4.4
0.317
0.332
0.178
0.817
0.589
0.348
0.745
0.357
0.100
0.511
NH/jOAc (2)
4.7
0.336
0.317
0.165
0.760
0.544
0.291
0.682
0.288
0.063
0.440
NHijOAc (3)
4.6
0.273
0.205
0.070
0.642
0.479
0.038
0.684
0.222
0.018
0.395
HOAc
10.3
0.166
0.186
O.O69
0.716
0.643
0.426
0.710
0.373
0.091
0.564
H0Ac(2)
7.6
0.209
0.228
0.106
0.708
0.617
0.389
0.630
0.337
0.084
0.494
HGAc(3)
4.9
0.317
0.346
0.190
0.807
0.604
0.356
0.729
0.315
0.069
0.561
HOAc(4)
3.9
0.325
0.335
0. .83
0.771
0.571
0.320
0.68-7
0.284
0.056
0.503
Lactate
16.0
0.098
0.129
0.036
0.653
0.719
0.527
0.664
0.404
0.106
0.626
Lactate(2)
9.7
0.127
0.196
0.094
0.758
0.724
0.544
0.689
0.443
0.162
0.642
Lactate (3)
13-3
0.152
0.213
0.099
0.754
0.761
0.544
0.693
0.405
0.121
0.664
Lactate(4)
8.0
0.211
0.253
0.126
0.763
0.627
0.400
0.702
0.325
0.072
0.582
Citrate
32.9
0.069
0.153
0.066
0.655
0.785
0.676
0.621
0.491
0.199
0.754
Ci trate(2)
11.3
0.131
Q.159
0.063
0.622
0.606
0.41 1
0.586
0.321
0.074
0.520


Table 34. Relationships between
height of slash pine
field sites
selected s
after 1, 3,
oil and site
and 5 years'
properties and
growth on 72
Soil or site
Height
property
1 year
3 years
5 years
r2
pH
0.412**
0.224**
0.142**
Organic matter
0.083*
0.074
0.096*
N i trogen
0.029
0.074
0.017
CEC
0.056
0.089*
0.105*
NH^OAc (pH 4.8) extractable
-Ca
0.134**
0.151**
0. J17*
-Mg
0.113*
0.086*
0.076
-K
0.092*
0.067
0.101*
-A1
0.307**
0.296**
0.148**
Silt + clay
0.232**
0.240**
0.105*
Available moisture
0.213**
0.207**
0.059
Depth of A1 horizon
0.008
0.094*
0.154**
Depth to LH
0.231**
0.323**
0.424**
Drainage class
0.306**
0.497**
0.456**
* Significant at the 5% level, using the model Y = aX + bX2 + c.
**Signif¡cant at the 1% level.


MATERIALS AND METHODS
Introduction
A series of field, greenhouse and laboratory experiments designed
to examine the effectiveness of soil-test methods for predicting the need
for P fertilization of slash pine, included:
a) A preliminary screening of a wide range of soil-test methods to
relate soil-P values with tree height, height response to P frtil-'
izer and fertilizer requirements of slash pine in field and green
house experiments (greenhouse trial 1) on 10 soils. The most effec
tive of these test methods were then calibrated against tree growth
and response parameters from 72 field fertilizer trials.
b) The solubility of a range of P compounds in soil-test extractants was
related to the capacity of slash pine seedlings to utilize P from
these compounds in a greenhouse test (greenhouse trial 2).
c) A P-retention study was conducted in both the laboratory and field,
using soils that varied widely in their ability to retain fertilizer
P.
39


Table 46. Height, dry weight, and P concentration of slash pine seedlings after 1
and 2 years' growth on 10 soils in the greenhouse receiving four P treat
ments
T reat-
ment
Repli-
cate
1 year
2 years
Ht.+
Dry wt
tops
P in
tops
Ht.
Dry wt.
Tops Roots
Tops
P
Roots
cm
g
/
'0
cm
-g

-*
B1 aden
silty clay
loam (A2)^
1
10.3
6.5
o.oo
17.8
22.0
7.1
0.048
0.043
Po
2
11.4
10.8
0.086
20.3
60.0
18.2
0.050
0.058
3
10.1
3-1
0.033
13.3
7.2
4.7
0.035
0.039
1
12.8
12.8
0.069
31.8
72.3
24.7
0.059
0.058
Pi
2
15-5
17.3
0.093
34.0
91.3
27.8
0.067
0.077
3
10.9
6.5
0.058
26.5
50.6
16.1
0.058
0.051
1
11.5
11.7
0.088
35-5
78.6
26.0
0.081
0.081
P 2
2
14.3
13.8
0.117
37.8
79-9
28.0
0.067
0.061
3
12.1
11.3
0.100
33.0
78.7
31.4
0.071
0.084
1
13.5
19.4
0.104
35-8
84.7
30.0
0.082
0.084
P 3
2
15.5
17.8
0.089
36.3
101.8
33.2
0.086
0.095
3
12.0
13-5
0.069
34.3
90.9
28.2
0.089
0.121
Blanton fine sand (A8)
1
12.4
10.0
0.105
26.5
54.0
21.6
0.080
0.112
Po
2
13.3
7.3
0.114
29.3
55.7
17.4
0.081
0.102
3
11.1
6.0
0.159
31 .5
49.7
14.3
0.084
0.100
1
14.3
13.7
0.125
31.3
59.6
25.8
0.081
0.111
Pi
2
12.4
8.3
0.140
30.5
53.0,
14.8
0.120
0.155
OO
3
11.1
7.7
0.168
30.8
51.9
16.2
0.126
0.156


Table 49. Amonts of P extracted from 10 soils by soil-test methods
Soi1-test So i 1 (CRIFF identification code)
method ~~ A2 a8 A21 A2A A28 Al Al9 A23 A5 A25
Ppm
h2o
0.5
1.7
3.0
0.2
0.3
5.5
6.4
2.4
0.7
2.2
H20(2)
0.1
0.5
0.4
0. 1
0. 1
31 -2
34.5
3.1
0.7
0.4
H20(3)
0.0
0.4
0.4
0.0
0.0
0.3
0.6
1 .1
0.3
0.4
H20(4)
0.7
78.6
50.4
1.9
1.4
884.7
1064.8
245.1
66.1
44.6
NaCl
0.0
1.3
1.4
0.0
0.0
6.5
8.0
2.3
1.1
1.2
h'a2S04
0.0
1.4
1.7
0.0
0.0
7.0
8.1
2.4
0.9
1.3
Na2MoO/j
2.5
5-7
27.7
2.7
3.1
6.1
7.8
2.9
3.2
13.6
Na2B^0y
0.1
1 .0
1.2
0.1
0.0
7-8
10.2
2.1
0.7
0.9
Na2B/}07 (2)
1 .0
22.6
60.8
0.6
1 .0
8.0
9-3
4.1
10.2
32.2
NH/^OAc
0.7
4.3
13.8
0.4
0.3
6.3
7.8
2. 1
2.6
5.9
NH^OAc(2)
0.6
4.2
13.1
0.3
0.2
7.8
9.5
2.7
2.3
6.3
NH/,0Ac (3)
0.3
3-9
14.4
0.0
0.0
8.2
8.5
2.7
2.6
5.7
HOAc
1.7
9.1
43.7
0.5
0.6
7.2
9-9
4.2
5.5
21.0
HOAc(2)
1.4
5-4
29.2
1.0
0.7
6.1
10.3
4.5
4.3
13.0
HOAc(3)
0.2
5.4
16.3
0.1
0.1
7.1
8.7
2.3
2.6
6.2
HOAc(4)
0.3
3-6
12.9
0.2
0. 1
5.7
7.2
2.2
1.9
4.6
Lactate
2.3
16.7
64.8
0.9
0.9
6.9
10.8
6.2
9.4
41.3
Lactate(2)
0.9
8.7
44.8
0.7
0.5
6.1
6.8
2.5
4.0
21.6
Lactate (3)
0.8
14.1
58.5
0.7
0.4
8.6
10.8
4.9
8.0
26.6
Lactate (4)
0.8
7.8
33.6
0.4
0.4
8.2
8.2
3.9
4.1
13.1
Ci trate
4.1
57.0
125.0
4.0
3.0
13.4
17.1
9.3
16.0
80.0
Citrate(2)
2.7
10.1
50.3 '
1.3
1.1
7.3
9.3
6.3
5-3
20.0
Oxalate
14.6
58.4
282.5
13.1
15.0
8.8
13.9
14.9
40.1
247-5


136
retention capacity of the soil does modify the relationship between
extractable soil P and P requirements over longer periods than one year
following fertilization. However, the improvement in the significance
of the regressions within groups of soils with similar NH^OAc-extract-
able A1 over that for all soils combined does not appear sufficient to
warrant use of separate regression equations for prediction purposes.
'Best fit1 models derived using the procedure and soil and site
parameters outlined previously are given in Table 31- As mentioned
previously, such equations are of interest as prediction models and for
seeing how much variation in the independent variable can be accounted
for by the independent parameters, but not as indicators of causal
relationships. The most noteworthy feature of these models is their
low R^ values, particularly for the models predicting fertilizer
requirements for 3" and 5-year growth periods. This suggests either un
measured parameters were contributing significantly to variation in
fertilizer requirements, or, as suggested previously, the calculated
fertilizer requirements did not reflect the true fertilizer requirements
of the sites.
Relationships Between Soil-Test Values and Height
In order to convert predicted relative height response into an
absolute growth response, which is the criteria needed for economic
justification of fertilizer additions, a prediction of height (height
growth in the absence of P fertilizer) is required. Relationships be
tween soil-test values in the surface soil and height (Table 32) showed
that only H^O-extractable P provided a reasonable estimate of height.
The R^ value for this relationship declined with an increase in the
growth period, as was also observed in the preliminary screening. It was


155
Phosphorus-Retention Study
Phosphorus-adsorption isotherms of representative soils from each
soil order are shown in Fig. \h. Isotherms similar to these were plotted
for all h2 soils used in this P retention study, from which data were
obtained for fitting the Langmuir adsorption isotherm. Two indices of
the P-sorption capacity of the soils, determined from the Langmuir equa
tion and saturation with 2,500 yg P/g soil, are presented in Appendix
Table 60. These data illustrate the wide range in P-sorption capacities
of soils in the Coastal Plain. A considerable overlap in sorption capa
cities of soils between various soil orders was obtained suggesting that
these broad classification units cannot be used to separate the soils
into distinct P-sorption groups, although soils classified as Spodosols
form a fairly distinctive group in the lower P-sorption range. Thirteen
of the 19 Spodosols in this study had a zero Langmuir adsorption maximum.
The Langmuir maximum, the derivation of which is based on the as
sumption of monolayer adsorption, gave consistently lower values than the
adsorption which can be obtained by saturation of the soil with a concen-
tracted phosphate solution. This i 1lustrates the theoretical weakness of
assuming monolayer adsorption as the sole mechanism operative in P reten
tion in soil systems. However, for the hi soils, the values for satura
tion maximum (V) and Langmuir maximum (X) were significantly related:
Y = 2.33X + 63.22, (r = 0.986).
This close relationship suggests that although the Langmuir data
do not provide absolute indices of P retention, they adequately reflect
relative P-retention capacities.
Relationships Between P Retention and Soil Properties


88
Phosphorus Compounds
Although correlation studies relating soil-P fractions to plant-P
uptake and P extracted by soil-test methods have provided some theoret
ical basis for selecting soil-test methods (Thomas and Peaslee, 1973),
this information can be misleading for several reasons: (a) The selec
tivity of some P-fractionation procedures are questionable (Bromfield,
1967)- (b) Phosphorus compounds within fractionation categories, such as
Ca-P, Al-P and Fe-P, may vary considerably in their availability to
plants (Juo and Ellis, 1968). (c) Correlation techniques used in most
studies may not provide definitive evidence of direct relationships be
cause of possible unknown mutual correlations (Thomas and Peaslee, 1973)-
(d) Variable P readsorption and buffering effects of different soils
during extraction can confuse the relationships between P fractions and
extractable P (Martens, Lutz, and Jones, 1969)-
Examination of the solubility in chemical extractants of P com
pounds of known P availability to plants should provide more direct evi
dence (a) on which to base the selection of soil-test methods, and (b)
to explain the success or failure of various soil-test methods for pre
dicting plant needs for P fertilizers.
Solubility of P Compounds in Chemical Extractants
The solubilities of P compounds in the extracting solutions of
several well-known soil-test methods, in the presence and absence of two
soils are shown in Table \h. Soil-test values are corrected for the
amount of native P extracted from the untreated soils by the correspond
ing extractant.
Monocalcium phosphate (MCP)
Dissolution of MCP was complete in all extractants except the


RESULTS AND DISCUSSION
Preliminary Screening of Soil-Test Methods
The classification and selected properties of the 19 soils used in
preliminary screening of soil-test methods are shown in Table 6. They
are typical of many forest soils in the lower Coastal Plain. These are
acid, sandy, and relatively low in organic matter. Levels of extractable
P vary considerably among the 10 soils.
Heights, relative heights, P-fertilizer requirements and tissue-P
parameters of slash pine grown on these 10 soils in both greenhouse and
field trials are summarized in Tables 7 and 8, respectively. Data in
these two tables were summarized or computed from the original data of the
greenhouse and field trials presented in Appendix Tables 46 and 47, re
spectively. Phosphorus uptake values at the first year's harvest in the
greenhouse trial were computed using the regression equation
Y = 1.604X 4.951 (r = 0.999)
where Y = mg P in tops and roots per pot, and X = mg P in tops per pot.
This equation was derived using plant uptake data obtained from extra pots
which were subject to a complete harvest after the first year's growth
(Appendix Table 48).
The amounts of P extracted from the 10 soils by all soil-test meth
ods are given in Appendix Table 49- Also included in this table are the
amounts of various P fractions and the buffering capacities of these soils
Relationships between extractable soil-P values and plant growth
and response parameters are usually curvilinear, according to Grigg (1965)
68


133
superiority of P extracted from within the effective soil depth compared
to that extracted from the surface soil was more apparent for the 3 and
5~year growth periods.
Relationships Between Other Soil and Site Parameters and P-Fertilizer
Requi rements
Drainage class was the only parameter which provided a better pre
diction than extractable-soil P of the amount of P fertilizer required
to achieve 95% of maximum height over any growth period (Table 29). Most
of the other parameters recorded above as significantly related to
fertilizer requirements were also significantly related to extractable P
(Table 2b).
The hypothesis that extractable soil-P values provide a better
indication of P-fertilizer requirements, within groups of soils with
similar P-retention capacities, was examined. This was done by determin
ing relationships between extractable-soil P and P requirements within
classes of soils grouped according to the amounts of their NH^OAc-s.
extractable A1. Aluminum extracted by NH^OAc has been reported as sig
nificantly correlated with the P-retention capacity of soils in the
Coastal Plain (Yuan and Breland, 1969)- Results (Table 30) showed that
after 1 year, the significance level of the regression of P-fertilizer
requirements (95%) on P extracted by the HCl-i^SO^ method and its
squared term was less within P-retention categories than for all soils
combined. However, after 3 years, the significance level of the regres
sions for the two soil groups with higher levels of NH^OAc-extractable
A1 was greater than that for all soils combined. After 5 years, the
level of significance of the regressions for all three groups was
greater than that for all soils combined, although none reached even
the 10% level of significance. These data indicate that the P-


118
Table 23. Actual values for the parameters included in Table 23 are
given in Appendix Tables 5^, 55, and 56.
As prediction models, some of the relationships shown in Table 23
are superior to models involving use of extractable soil P. However,
many of the significant relationships can be attributed to the influence
of these variables on soil P. Silt + clay, NH^OAc-extractable Ca and A1,
available moisture, and pH were all significantly correlated with H2O-
extractable P (Table 2k). Since relative height is a measured response
to P fertilizer application in the field, it is most likely that the sig
nificant relationships shown in Table 23 exist because the parameters
are either related to available soil-P levels or in some manner influence
the response of the trees to added P fertilizer.
Depth to a limiting horizon (LH) and drainage class, which are
closely related (Table 2k) both showed increased values with increas
ing tree age, a result not shown by any other soil parameters, including
any form of extractable P. Depth to LH, although significantly
correlated with HC1-H2S0^, extractable P, is markedly superior to
HC1-H2S0^ -extractable P as a predictor of relative height, particularly
at age 5 years. A plot of depth to LH against relative height at age 5
years is shown in Fig. 8. Multiple regression equations of relative
height on depth to LH, HCl-H2S0^-extractable P (0-20 cm) and the squared
terms of these two parameters showed that whereas depth to LH and its
squared term were significantly related to relative height at ail three
measurement periods, the significance of the contribution of HC1-H2S0/J-
extractable P decreased with increasing age (Table 25). At tree age of
5 years, neither of the terms for HC1 H2S0t-P made a significant contri
bution towards accounting for the variation in relative height. An
explanation for these results probably lies in the previously observed


RELATIVE HEIGHT,
HCI-H2SO4 EXTRACTABLE P, ppm
Fig. 6. Relationship between HC1-H^SO^-extractable P(X) in the surface 20 cm of soil and relative
height(Y) of slash pine 5 years after P fertilization.


Table 55- Continued
Field
s i te
A2
A3
A4
Repl i-
cate
*
Depth
V
P
NH,OAc
4
P
HcY-
H SO.
2 4
P
Bray 1(3)
P
Bray 2
P
NH.OAc
4
pH
(h2o)
Ca
Mg
K
A1
cm
ppm--
3
0-20B
0. 1
0.7
1 .8
0.6
4.7
315
110
22
166
4.5
0-20
0. 1
0.5
1 .2
0.4
2.8
230
110
20
206
4.6
20-40
0.1
0.5
0.5
0.1
1 .7
495
182
30
403
4.9
40-60
0.1
0.2
0.2
0.1
0.9
360
262
43
593
5-0
1
0-20B
3.4
3-1
4.1
5.3
5.8
230
66
12
9
4.0
0-20
2.8
3.1
4.6
4.9
5.0
105
50
10
9
4.1
20-40
0.5
0.7
0.8
1-3
1,1
45
19
2
14
4.6
40-60
0.2
0.5
1 .1
4.8
2.3
75
19
4
177
4.4
2
0-20B
2.8
3.5
3.6
5.3
4.5
170
50
10
9
4.1
0-20
2.0
2.2
3.0
4.2
3.6
105
30
12
6
4.1
20-40
0.5
0.7
1.0
1.9
1.8
15
21
2
33
4.5
40-60
0.4
0.7
2.1
4.2
4.4
15
19
2
121
4.5
3
0-20B
1.9
2.2
3.3
4.6
4.2
135
42
4
6
4.2
0-20
1.3
2.2
1.8
2.5
2.2
90
30
7
2
4.5
20-40
0.5
0.9
0.6
0.8
1.3
105
25
2
0
4.8
40-60
0.2
0.5
1.6
3-3
4.7
30
21
4
161
4.5
1
0-20B
1.5
1.3
2.8
4.6
3.8
120
21
7
23
4.1
0-20
0.8
0.9
1.8
3-0
2.9
60
14
4
27
4.4
20-40
0. 1
0.5
1.4
2.5
5.1
15
10
2
119
4.7
40-60
0.1
0.2
0.7
1.3
2.4
15
10
1
71
5-0
2
0-20B
2. 1
2.2
3-1
4.7
4.9
120
25
7
2
4.1
0-20
1.7
2.2
2.4
3-0
2.8
. 135
30
7
2
4.3
20-40
0.2
0.7
1.1
2.3
2.1
15
19
2
22
4.6
N>
V£>


67
Statistical Analysis
Statistical analysis of data was performed using either a
Hewlett-Packard model 9100B desk top computer or the computer facilities
at the University of Florida Computer Center. Library Stat-Pac programs
were used with the Hewlett-Packard. All statistical analyses done at the
Computer Center were performed using programs available in the Statisti
cal Analysis System (SAS) package (Barr and Goodnight, 1972). Specific
programs used aregiven in the following sections where appropriate.


159
Table 39- Soil properties and their correlation with P retention
Soi 1
property
Mean
Range
P retent ion
Langmuir Saturation
maximum maximum
CEC, meq/100g
5.02
1.48-11.65
0.214
_ r
0.190
Exch. Ca, meq/100g
0.83
0.13-3.44
0.222
0-251
Clay, %
3.95
0.1-20.4
0.771**
0.755** .
(0.323*)'
Silt, %
10.68
0.1-33.8
CO
O
o
0.727**
(0.257)
Silt + Clay, %
7-31
0.2-47.4
0.785**
0.757**
pH
4.34
3.1-6.2
0.479**
0.535**
(0.551**)
Loss on
ignition, %
3.33
1.42-6.71
0.455**
0.425**
(0.085)
Significant at the 5% level.
Significant at the 11 level.
Partial correlation coefficients in parentheses are corrected for
effect of NH^OAc-Al.


120
Table 23. Relationships between selected soil and site properties and
relative height of slash pine 1, 3 and 5 years after P
fertilization on 72 sites
Soil or site^
property
Relative height
1 year
3 years
5 years
r2
pH
0.089*
0.083*
0.097*
Organic matter
0.013
0.038
0.008
Nitrogen
0.149**
0.216**
0.143**
CEC
0.020
0.030
0.035
NH^0Ac(pH 4.8) extractable
Ca
0.217**
0.175**
0.252
Mg
0.019
0.005
0.029
K
0.047
0.044
0.019
A1
0.257**
0.145**
0.144
Si 11 + clay
0.347**
0.225**
0.149**
Available moisture
0.227**
0.095
0.017
Depth of A1 horizon
0.056
0.050
0.032
Depth to limiting horizon
0.281**
0.314**
0.345**
Drainage class
0.189**
0.237**
0.274**
Except for depth functions and drainage class, all properties are for
the surface 20 cm of soil.
* Significant at the 5% level, using the model Y=aX + bX^ + c.
**Signif¡cant at the \% level.


Table 56. Continued
Field
s i te
Repli-
cate
Depth
Particle
size d istribut ion
Moisture content
Bulk
density
Clay
Silt
Sand
1/3 atm.
15 atm.
cm
g/cc
A23
2
-40-60
3.5
8.8
87.7
6.5
3.1
1.70
3
0-20
1.6
10.3
88.1
7.5
4.1
1.28
20-40
1 .4
7.8
90.8
3.9
1.4
1.61
40-60
3.4
9.5
87.1
7.3
2.6
1 .67
A24
1
0-20
11.8
30.4
57.8
14.2
5.0
1 .21
20-40
15.0
28.0
57.0
14.2
5.0
1.56
40-60
19-8
26.4
53.4
14.9
6.5
1.58
2
0-20
11.6
33.8
54.6
14.7
4.3
1.12
20-40
14.9
31.2
53.9
16.2
4.5
1.65
40-60
21.7
29.0
49.3
17.0
7.1
1.73
3
0-20
8.6
24.8
66.6
11.9
3.5
1.17
20-40
10.8
23-6
65.6
1 1 .2
3.2
1.69
40-60
15-0
23.8
61.2
13.5
4.8
1.67
A25
1
0-20
4.5
10.0
85.5
5.6
2.5
1 .36
20-40
4.8
8.2
87.0
4.2
2.0
1 .65
40-60
4.8
8.5
86.7
4.3
1 .9
1.73
2
0-20
4.0
9-5
86.5
5-6
2.2
1.28
20-40
3.8
9.2
87.0
4.2
1.8
1.65
1
40-60
4.8
8.9
87.2
4.0
1.4
1.68
3
0-20
3.7
1*3.3
83.O
4.9
2.1
1.29
20-40
2.7
11.9
85.4
3.7
1.9
1 .65
40-60
3-2
12.8
84.0
3.0
1.1
1.70


Table 46. Continued
T reat-
ment
Repli-
cate
1 year
2 years
Ht.+
Dry wt.^
tops
P in
tops
Ht.+
Dry wt.
Tops Roots
F
Tops
>
Roots
cm
9
f
cm
g
' 'o
Ona
fine sand (A23)
1
16.5
23.1
0.189
38.8
94.2
33.2
0.120
0.157
P2
2
18.9
22.5
0.236
39.8
93.2
26.3
0.128
0.126
3
14.5
14.2
0.181
38.5
89.3
30.7
0.117
0.125
1
13-9
17.0
0.242
41.5
93-2
28.8
0.156
0.226
P 3
2
21.4
22.9
0.254
42.8
103.2
33.1
0.151
0.140
3
15.5
21 .2
0.278
39.0
91 .2
27.2
0.146
0.148
Kershaw fine sand (A5)
1
10.6
5-4
0.170
20
42.9
10.4
0.094
0.124
Po
2
10.9
4.4
0.097
29.0
38.2
11.7
0.060
0.074
3
8.4
3.4
0.053
22.3
26.9
13.6
0.072
0.078
1
12.3
9.1
0.145
26.5
48.0
21 .4
0.076
0.107
Pi
2
11.9
8.3
0.189
29-8
48.7
17.8
0.094
0.141
3
11.6
8.5
0.157
27o
49.0
20.8
0.099
0.149
1
13.5
11.4
0.168
29.8
49.6
19.3
0.114
0.162
P2
2
13.1
7-7
0.195
29.5
56.0
17.1
0.134
0.213
3
11.8
9-4
0.181
26.8
49.6
20.0
0.129
0.184
1
13.1
13-5
0.231
29-5
55.0
24.2
0.152
0.241
P 3
2
11.6
9.7
0.264
26.5
51.6
15.7
0.183
0.238
3
11.8
10.4
0.292
24.8
50.2
21 .0
0.168
0.230
v^o
ho


Table 62. Amounts of A1 and Fe extracted from 42 Coastal Plain forest soils by four soil P-test
methods
Field NH^OAc HCl-H^SO^ Bray 1 Bray 2
S'te A ~ ~ A1 Fe A1 Fe A1 Fe
- --ppm
A5b
35
5
114
21
370
84
433
116
A5
38
5
115
22
405
108
395
126
A15
82
9
203
29
880
123
895
150
A25b
66
4
280
27
1,020
67
1,205
112
A25
70
5
268
27
1 ,030
80
1,250
136
A7
127
17
173
39
440
56
475
83
A10
148
24
290
58
840
112
895
171
A23
138
19
285
52
670
110
720
150
Al6b
8
0
35
9
78
15
85
21
A16
8
1
50
11
95
18
105
25
A3
15
1
38
12
80
15
80
20
A4
12
1
4o
11
85
19
85
22
A6
15
1
65
12
130
22
135
31
A18
18
1
80
20
145
31
150
48
A1 9b
11
1
38
10
83
13
83
20
A19
r: 10
1
38
1 1
75
15
85
21
A21
25
7
43
17
140
33
125
40
A22
20
2
55
13
110
21
115
28
A23b
55
6
133
28
295
49
308
70
FI
98
3
300
13
500
23
550
37
F2
15
1
68
, 7
120
13
135
20
F3
8
2
55
6
30
14
65
22
F4
15
1
85
8
85
22
90
30
F5
8
2
70
16
80
37
85
51
ro
VJ"I


52
Table 4. Physical and chemical properties of soils used in greenhouse
trial 2
Property
Soil type
Immokalee fs McLaurin fsl
pH (1:2, H20)
4.2
4.9
pH (1 N_ KC 1)
3.0
3.8
Clay, %
1.9
9.8
Silt, %
5.6
28.7
Organic matter, %
3.3
2.2
CEC, meg/100g
5.5
li 7
NH/jOAc (pH 4.8) extractable
Ca, ppm
164
104
P, ppm
6.3
0. 4
0.3 M NH^C20^ (pH 3-0) extractable
A1, ppm
25
1025
Fe, ppm
60
5^5
P retention,* yg P/g soil
30
750
* Equilibration with 2,500 yg P/g soil for 6 days.


Table 7. Height, relative height, P-fertilizer requirements, and P concentration and uptake
of slash pine
seed 1ings
after
1 and 2
years 1
growth on 10 soil
s in the
greenhouse
Soi 1
type
Height
Re 1 .
height
Fert.
reqm.f'
P in
tops
P uptake^
1*
2
1
2
1
2
1
2
1
2
cm-

%

--kg
P/ha-
%
--mg
P/pot
Bladen scl
10.6
17.1
77.3
45.8
56.6
92.5
0.060
0.044
10.2
19.5
Blanton fs
12.3
29.1
89-3
91.7
13.8
0.0
0.126
0.081
25.4
62.1
Plummer fs
15-3
35.8
98.3
90.3
0.0
0.0
0.132
0.089
45.7
105.6
Marlboro fsl
12.0
19.4
65.0
52.7
70.8
95.6
0.058
0.038
8.1
15.2
McLaurin fsl
12.3
23.8
71.3
64.6
81.4
93.3
0.055
0.047
8.8
22.2
Immokalee fs
14.8
29.4
100.0
98.4
0.0
0.0
0.096
0.039
51.6
35.5
Leon fs
13.0
26.2
100.0
98.1
0.0
0.0
0.108
0.068
44.3
40.7
Ona fs
15.2
34.1
87.2
84.2
4.8
35.8
0.091
0.044
37.6
48.7
Kershaw fs
10.0
26.7
79.9
93.2
48.5
0.0
0.107
0.075
10.6
38.1
Lakeland fs
15.2
32.9
94.5
87.4
0.0
15-7
0.123
0.082
53.6
79.8
* Age in years at harvesting or measurement.
1 Fertilizer required to achieve 30% of maximum height growth.
§ First year uptake based on eight seedlings/pot and 2-year uptake based on four seedlings/pot
over 2 years.


70
60
50
40
30
20
10
0
50
40
30
20
10
0
Fig
96
IMMOKALEE FINE SAND
(A16)
LSD
..(5/.)
MCLAURIN FINE SANDY LOAM
(A28)
LSD
(5%)
C MCP DCP FA CAIP KTK WA CFeP STR
PHOSPHORUS COMPOUND
2. Phosphorus uptake by slash pine seedlings after
8 months' growth on two soils treated with
eight P compounds.


19
vegetative growth period and large root system of trees (Terman, 1968),
and (3) the presence of mycorrhizal rootlets on coniferous trees which
have been reported to increase the trees ability to utilize less soluble
forms of P (Bowen, 1973)-
Although not extensively tested, DAP has proved to be an effective
fertilizer in the Southeast, particularly on sites where responses to N
and P are synergistic (Pritchett and Smith, 197*0- This source is recom
mended for operational fertilization on these sites; however, its long
term effectiveness on soils of high P-retention capacity is likely to be
limited. Furthermore, White and Pritchett (1970) found that 5 years af
ter surface application of DAP to a Leon soil less than 11% of the applied
P remained in the surface 20 cm of this soil having a low P-retention
capacity.


Table 12. Relationships between selected soil-test values and P-fertilizer require
ments of slash pine in field and greenhouse experiments on 10 soils
Soi1-test
method or
P form
Mean
extract
able P
P-fertilizer requirements^
Greenhouse
Field
1?
2
1
3
5
ppm
h2o
2.3
0.897**
0.703**
0.508*
0.088
0.000
NH/jOAc
4.4
0.851**
0.863**
0.745**
0.357
0.100
T ruog
8.9
0.685**
0.607**
0.613**
0.414*
0.154
01 sen
12.3
0.552*
0.678**
0.694**
0.513*
0.201
hci-h2so4
15.2
0.606**
0.697**
0.686**
0.494*
0.192
h2so4(9)
32.5
0.377
0.488*
0.567*
0.519*
0.230
Bray 1(3)
41 -3
0.436*
0.616**
0.601**
0.506*
0.217
NH4F(pH 8.5)
57.0
0.322
0.396*
0.491*
0.507*
0.247
Organic-P
42.6
0.050
0.001
0.003
0.016
0.003
Total-P
136.1
0.043
0.015
0.077
0.193
0.123
4.
' P fertilizer required
to achieve 90% of maximum growth.
s> Tree age at
time of determination (years).
* Significant
at the 5%
level, using the model Y =
b logX + c,
where Y = P
ferti 1izer
requirement
and X = soil-test value.
** Significant at the 1% level.
oo
NJ


LITERATURE REVIEW
Forest Fertilization
H i storica1
The value of fertilizers for increasing the productivity of timber
lands was first recognized in Europe. Fertilizer trials were established
on peat lands in Sweden as early as 1898 (Hagner, 1967). During the pe
riod 1900-1925, forest fertilizer trials were established in many European
countries including Finland (Salonen, 1967), Belgium, Germany, and Denmark
(Tamm, 1968), and Britain (Leyton, 1958). Experimentation expanded in
Europe between 1925 and I960, establishing a firm scientific foundation
to tree nutrition which led to the first operational applications of fer
tilizer in Germany (Tamm, 1968) and Britain (Leyton, 1958). In Austra
lia, the importance of phosphatic fertilizers in achieving healthy growth
of introduced pines planted on infertile sandy soils was established as
early as 1930 (Kessell and Stoate, 1938).
Research on forest fertilization is relatively new in North Ameri
ca. Some of the earliest trials on this-continent were N experiments es
tablished on hardwood stands in the northeastern USA by Mitchell and
Chandler (1939). Westveld established a phosphate fertilizer trial in
the southeastern Coastal Plain in 19^5 (Pritchett and Swinford, 1961),
while research on forest fertilization was started around 1950 in the
Pacific Northwest (Gessell, 1968).


10
and other soil constituents. As the pH of the diffusing fertilizer solu
tion rises following reaction with soil constituents and dilution by the
soil solution, the solubility products of various A1 and Fe phosphates
are exceeded and they precipitate (Lindsay and Stephenson, 1959 b). The
initial precipitates in an acid soil (Hartsells fine sandy loam, pH 4.6)
were identified as colloidal products of the type (Fe, A1, X)P0^.nH20 and
various crystalline acidic phosphates of the type X20.3(A1, Fe)20^.6P205
.2OH2O. The Fe and A1 may substitute for one another, and the X indicates
a cation other than Fe and A1 (Lindsay and Stephenson, 1959 b; Lindsay,
Frazier, and Stephenson, 1962).
Many of the initial reaction products formed in the soil following
addition of MCP have been reported to be good sources of plant-available
P (Taylor, Gurney and Lindsay, I960; Juo and Ellis, 1968). However, as
the initial colloidal and poorly crystalline products age, they become
more crystalline and less available sources of P (Juo and Ellis, 1968).
Stable crystalline phosphate minerals of the vari sc ite-strengite series
are considered to be the ultimate end-product of the reactions and trans
formations of MCP in acid soils (Wright and Peech, I960). Juo and Ellis
(1968) reported that Fe phosphates crystallize more rapidly than A1 phos
phates in acid soils. They observed that amorphous Fe-P forms crystal
lized within 9 months at 35 C, while amorphous A1 -P forms remained com
pletely amorphous over the same period. This slower rate of crystalliza
tion of A1 phosphates was cited by Juo and Ellis (1968) as the probable
explanation why colloidal Al-P was more effective than colloidal Fe-P as
a long term source of P for plants (Taylor et al., I960) and explained
the numerous reports of close correlations between Al-P (extracted by
alkaline NH^F) and plant-P uptake (Thomas and Peaslee, 1973).


56
surface 20 cms, *4 years after fertilizer application, was determined by
difference in total P (kg P/ha) between control plots and plots which
had received 56 or 22*4 kg P/ha. This P-retention value was then
expressed as a percentage of the P applied.


kk
dY
dX
dY
= b + 2cX,
then solving X for = 0, to give
b + 2cX = 0, and by rearranging
x = ik
2c
We obtained the value of Y corresponding to this value of X from the
original quadratic equation by substitution
Y = a + b
-b'
-b
2c
+ c
2c
, and rearranging
Y = a which is maximum tree height.
Relative height was computed for each replicate at each site for growth
periods of 1, 3, and 5 years. In some cases it was not possible to
obtain a predicted maximum height in this manner, because the response
was either linear or increased exponentially with increasing rates of P.
In these cases, the maximum height was taken as the actual mean height
in the tallest P i N i treatment. Where no increase in height occurred fol
lowing P application, relative heights were recorded as 100%.
The amount of P fertilizer required to achieve maximum height was
obtained by differentiating the quadratic equation, setting the deriva
tive equal to zero and solving for X (this = -b/2c). In addition, the
amounts of fertilizer required to achieve 90 and 95% of maximum height
were also computed. These values were obtained by substituting 90 and
95% of maximum tree height values in the quadratic equation and solving
for X. In cases where actual rather than predicted maximum tree heights
were used, P fertilizer required for maximum height was taken as the
actual P rate of the treatment that produced the maximum height.


8
to canopy closure at age 5 to 10 years (Armitage, 1969; Craig, 1972).
Once canopy closure is reached the stands are refertilized by aerial
application.
The characteristics of forest soils in the southeastern USA lower
Coastal Plain have been described by Pritchett and Smith (1970). These
soils are predominantly infertile, acid sands with drainage characteris
tics ranging from excessively drained to poorly drained. Responses of
slash pine to P fertilizers in the lower Coastal Plain have been related
to the drainage characteristics of the sites. In eight fertilizer trials,
Pritchett and Llewellyn (1966) found that 3 to 5 years after application',
significant responses to P fertilizer had occurred in 5 out of the 8
trials, all of which were on sites classified as poorly or somewhat poor
ly drained. On two well-drained sites no response to P was recorded,
although they did respond to N application. Responses from the same
trials, recorded 7 to 11 years after fertilizer application (Humphreys
and Pritchett, 1971), showed that the greatest response had occurred on
the poorly drained sites, followed by the somewhat poorly drained sites,
with no P response on the well-drained sites. In a more recent series of
trials established at planting time on 28 sites throughout the Coastal
Plain, Pritchett and Smith (1972) reported that 3 years after fertilizer
application, 6 of 7 experiments located on poorly drained sites, 8 of 10
on somewhat poorly drained sites, k of 8 on moderately well drained sites,
and none of the 3 on excessively drained sites responded significantly to
P applications. In the same series it was found that N alone significantly
increased growth only in 8 of the experiments on the somewhat poorly
drained sites, although when N was applied with P, it increased the re
sponse over that obtained with P alone in 72% of the 28 experiments.


Table 41. Correlations between different forms of A1 or Fe, and P retention
Variable
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(1)
Exchangeable A1
1.000
0.382
0.199
0.810**
0.748**
0.302
0.705**
(2)
Amorphous A1
1 .000
0.490**
0.238
0.686**
0.66!**
0.850**
(3)
Crystal line A1
1.000
0.065
0.454**
0.841**
0.498**
(4)
Exchangeable Fe
1.000
0.822**
0.314
0.651**
(5)
Amorphous Fe
1.000
0.696**
0.914**
(6)
Crystal line Fe
1.000
0.699**
(7)
P retention
1 .000
** Significant at the 1% level.


Table 55. Continued
Field
s i te
A4
A5
A6
HC1
Repli- H20 NH^OAc h2S04 BraV 1(3) Bray 2 NH^OAc pH
Depth" P P P P P Ca Mg K T (HO)
cm ppm
2
40-60
0. 1
0.5
1 -5
2.3
4.5
15
19
2
34
4.7
3
0-20B
1 .0
0.9
2.4
4.3
5-1
75
21
7
11
4.4
0-20
0.8
0.9
1 -9
3.5
3-1
60
25
4
9
4.5
20-40
0.1
0.5
1 -5
2.2
2.8
10
5
2
24
5.0
40-60
0.1
0.5
1.0
1.6
2.9
15
14
2
21
5.1
1
0-20B
0.5
2.6
9-1
36.5
33.5
135
19
7
33
5-2
0-20
0.5
2.6
9-1
36.5
33.5
135
19
7
33
5.2
20-40
0.3
1.3
12.1
39.8
40.2
75
19
2
33
5-3
40-60
0.3
1.3
13.6
40.9
41.0
15
21
2
30
5.4
2
0-20B
0.5
1.1
6.9
22.7
21.0
135
30
4
28
5.4
0-20
0.5
1.1
6.9
22.7
21.0
135
30
4
28
5-4
20-40
0.2
1.1
7.4
26.5
25.3
30
19
2
33
5-4
40-60
0.1
0.9
9.1
29.4
30.5
30
19
2
30
5.5
3
0-20B
0.7
1.3
8.1
30.3
27.6
60
19
2
28
5-3
0-20
0.7
1.3
8.1
30.3
27.6
60
19
2
28
5.3
20-40
0.2
1.1
6.7
26.5
27.6
15
19
2
38
5-3
40-60
0.2
1.1
10.1
38.3
AO.2
30
19
2
28
5.2
1
0-20B
5.2
6.6
5.7
8.6
6.8
315
88
18
2
4.2
0-20
4.2
6.6
4.4
5.3
5.7
330
94
22
6
4.1
20-40
0.7
0.9
0.8
1.2
1.2
75
21
2
6
4.6
40-60
0.5
0.9
1.3
2.1
2.1
105
30
2
14
4.9
rv>
ho


Laboratory-determined P-retention capacity of 42 forest soils was
significantly related to A1 extracted by either conventional procedures
(KC1 leaching, pyrophosphate, oxalate, and CDB extraction) or four soil-
P test methods. Soil A1 extracted by these four soil-P test methods was
calibrated against P leaching losses over a 4-year period in the field
following application of concentrated superphosphate (CSP). Soils con
taining less than 40, 120, 300, or 400 ppm A1 extractable by NH^QAc
pH 4.8, 0.05^ HC1 + 0.025N^ H2S0i>, 0.03^ NH^F + 0.025N.HC1, or 0.03N_
NH^F + O.OIN^ HC1, respectively, exhibited excess leaching losses of P
from soluble CSP fertilizer; use of less soluble P fertilizers, such as
rock phosphate, was suggested on such soils.
XVIII


171
that no major interaction with other elements occurred during thier up
take and translocation to the foliage. Although the relationship be
tween soil-P retention and foliar A] was significant, it accounted for
only 19% of the variation in P retention and is, therefore, not particu
larly useful for prediction purposes. Phosphorus retention was more
closely correlated with foliar Fe than with foliar A!, but even this re
lationship accounted for only 26% of the variation in P retention and the
predictability was not improved by using summed A1 and Fe values.
A graphical comparison of foliar A1 with soil A1 values (NH^OAc-
extractable) showed that foliar levels seriously underestimated soil A1 on
the sites where foliar P was low (P < 0.08%). An attempt was made to
correct this anomaly by using the ratio of foliar A1 and/or Fe to P as
the index of P retention. Use of such ratios improved the predictability
for A1 and A1 + Fe, but not for Fe (Table bb). However, even the best
relationship still accounted for only 35% of the variation in the P re
tention. By excluding sites where foliar P was below 0.08% (7 out of 3*0
from the analysis, the correlation between foliar A1 and soil-P retention
improved considerably to r = 0.752. This same relationship for the seven
low-P status sites was not significant (r = 0.070). On these low P sta
tus sites foliar A1 and foliar P concentrations were significantly cor
related (r = 0.691) while on the 27 sites where tissue P was > 0.03% this
relationship was not significant (r = 0.331)* Apparently on sites of low
P status foliar A1 levels are determined more by soil P than soil A1
levels. This may explain why foliar A] failed to provide a good index of
P retention over all sites. An examination of the relationships between
P retention and foliar Fe on sites of low and adequate P status did not
produce any improvement on the correlation over that obtained with Fe
us i ng al 1 si tes.


78
extracted by the soil-test methods up to a plateau (Fig. 1 and Table 10).
The plateau, where R values were almost independent of quantities of P
extracted, was reached at lower levels of mean extractable P for response
after 3 years (ca. 12 ppm) than after 5 years (ca. 20 ppm). As mean ex-
2
tractable P values increased above ca. 50 ppm, the R values began to de
cline indicating the soil-test methods involved were beginning to extract
forms of P unavailable for tree use at this stage of growth. Soil-test
methods extracted-amounts of P corresponding to the above-mentioned pla
teau included those involving use of strong acid and alkaline solutions
(HC1, H^SO/j and NaHCO^) and strong complexing agents (lactate, NH,F).
Amounts of Fe-P and Ca-P in the 10 soils were very low and were
not related to fertilizer response at any stage of tree growth (Appendix
Tables 50 and 51). Aluminum-P, which dominated the inorganic P fractions
of all 10 soils, was significantly related to tree response when they
were 3 and 5 years old. However, the mean amount present in the soil,
5**.3ppm P, suggests that not all this fraction was available for tree use
during the first 5 years on the basis of data in Fig. 1. This is perhaps
not surprising since different A1 P compounds found in soils vary consid
erably in their availability to plants (Taylor et al., I960). Soil-test
methods which discriminate between Al-P compounds of different solubility
are apparently better indicators of soil-P status than methods which do
not so discriminate.
Neither total nor organic-P were significantly related to response
at any stage of growth (Table 10). This concurs with the shortcomings of
total analyses reported previously (Pritchett, 1968; Ballard, 1970a).
However,it is of interest to note that R values for both methods of
analysis increased with increasing growth period. This result Is consis
tent with the premise that an increasing proportion of the total P in


272
Will, G. M. 1964. Dry matter production and nutrient uptake by Pinus
radiata in New Zealand. Commw. For. Rev. 43:57"70,
Will, G. M. 1966. Magnesium deficiency: the cause of spring needle-
tip chlorosis in young pine on pumice soils. New Zeal. J. For.
11:88-94.
Williams, E. G. 1962. Chemical soil tests as an aid to increased pro
ductivity. Trans. Int. Soc. Soil Sci. Comm. IV and V:82C-834.
Williams, E. G., and A. H. Knight. 1963. Evaluations of soil phosphate
status by pot experiments, conventional extraction methods, and
labile phosphate values estimated with the aid of phosphorus 32.
J. Sci. Food Agric,14:555"563
Wright, B. C., and M. Peech. i960. Characterization of phosphate reac
tion products in acid soils by application of solubility criteria.
Soil Sci. 90:32-43.
York, E. T. 1959- Agronomic view point of field experiments in tree
nutrition. Duke Univ. Sch. For. Bull. 15- 184 p.
Young, H. E. 1940. Fused needle disease and its relation to the nutri
tion of Pinus. Queensland J. Agrie. 45:45-54.
Young, H. E. 1948. The response of loblolly and slash pine to phos
phate manures. Queensland J. Agrie. Sci. 5:77"105.
Yuan, T. L., and J. G. A. Fiskell. 1959* Aluminon studies: 1. Soil
and plant analysis of aluminon by modification of the aluminon
method. J. Agrie. Food Chem. 7:115_117.
Yuan, T. L., W. K. Robertson, and J. R. Neller. i960. Forms of newly
fixed phosphorus in three acid sandy soils. Soil Sci. Soc. Amer.
Proc. 24:447-450.
Yuan, T. L. 1965- A survey of the aluminum status in Florida soils.
Soil and Crop Sci. Soc. Fla. Proc. 25:l43~152.
Yuan, T. L., and H. L. Breland. 1969- Correlation of A1 and Fe as
extracted by different reagents with phosphate retention in sev
eral soil groups. Soil and Crop Sci. Soc. Fla. Proc. 29:78-86.
Zahner, R. 1959* Fertilizer trials with loblolly pine in southern
Arkansas. J. For. 57:812-816.


107
Table 17- Relationships between soil-test values at three soil depths
and relative height of slash pine after 1, 3. and 5 years'
growth on 72 field sites
Soi1-test
Soil
Relative height
method
depth
1 year
3 years
5 years
em R^
Ho0
0-20B+
0.322**
0.149**
0.120**
z
0-20
0.306**
0.132**
0.100**
20-40
0.165**
0.132**
0.074*
40-60
0.179**
0.088*
0.024
NH.OAc
0-20B
0.281**
0.157**
0.084*
*4
0-20
0.270**
0.137**
0.072*
20-40
0.174**
0.172**
0.086*
40-60
0.194**
0.203**
0.114**
HC1H0SO.
0-20B
0.306**
0.246**
0.143**
0-20
0.264**
0.224**
0.126**
20-40
0.199**
0.233**
0.127**
40-60
0.317**
0.323**
0.208**
Bray 1(3)
0-20B
0.259**
0.221**
0.113**
0-20
0.235**
0.198**
0.095**
20-40
0.191**
0.140**
0.052
40-60
0.300**
0.228**
0.117**
Bray 2
0-20B
0.188**
0.210**
0.123**
0-20
0.152**
0.178**
0.094**
20-40
0.150**
0.229**
0.132**
40-60
0.282**
0.362**
0.251**
t
0-20B represents soil samples collected from within the bedded tree
row.
* Significant at the 5% level, using the mode Y=b logX + c.
**Significant at the 1% level.


Table 51. Continued
Mean
Foliar
Soi1-test
extract
Height
Re 1
. height
Fert
. reqm.
P
method
able P
JU
r
3
5
1
3
5
1
3
5
4
ppm
,r2
P fractions
Sol-P+Al-P+Fe-P
64.7
0.001
0.016
0.000
0.405
0.649
0.666
0.434
0.471
0.234
0.606
Ca-P
0.6
0.190
0.092
0.014
0.016
0.080
0.041
0.092
0.079
0.005
0.090
Sol-P+Al-P+Fe-P
+Ca-P
65.3
0.000
0.019
0.001
0.414
0.660
0.675
0.444
0.481
0.238
0.620
Organic-P
42.6
0.083
0.052
0.056
0.011
0.040
0.056
0.003
0.016
0.003
0.019
Total-P
136.1
0.090
0.030
0.048
0.065
0.229
0.306
0.077
0.193
0.123
0.175
Buffering capacity
Capacity(1)
0.006
0.009
0.002
0.264
0.446
0.464
0.224
0.251
0.110
0.423
Capacity(2)
-
0.61 1
0.450
0.359
0.217
0.022
0.006
0.213
0.006
0.004
0.023
Capacity(3)
-
0.605
0.452
0.373
0.332
0.013
0.007
0.229
0.027
0.009
0.01 1
* Age in years at sample collection or measurement.
2
t Using the model Y = b logX + c, where Y = tree parameter and X = soil-test value. R > 0.397 significant
at the 5% level, R^ > 0.586 significant at the 1% level.


9S
An analysis of variance of dry matter production (Appendix Table
53) revealed a significant difference between soils, but no difference
between P compounds. However, since there was a significant S x P inter
action, differences between P compounds within soils were examined by the
LSD method. Data in Fig. 3 show there were no significant differences in
seedling yields between P compounds on the Immokalee soil. On the
McLaurin soil, the MCP, DCP, CA1P, and CFeP treatments all produced sig
nificantly greater yields than the control and FA treatments. Yields as
sociated with the MCP, DCP, and CA1P treatments were significantly greater
than those of the WA treatment, while yields from the CA1P treatment was'
also significantly greater than those of the STR and KTK treatments.
Although the Immokalee soil was not P deficient for seedlings, dry
matter production on this soil was significantly lower than that on the
P-deficient McLaurin soil. This result was not anticipated. However,
the possibility of a N deficiency limiting dry matter production on the
Immokalee soil existed. Furthermore, soon after establishment, some salt
burn was observed on seedlings in some of the Immokalee pots. Seedlings
on the Immokalee soil appeared to be slower to recover from transplanting
shock than seedlings on the McLaurin soil. This may have been associated
with the lower pH of the Immokalee soil.
The effective utilization of P from KTK and the two colloidal
phosphates by pine seedlings supports the conclusion, derived from many
experiments with agricultural crops, that these forms of P are readily
available to plants (Taylor et al., I960; Taylor et al., 1963; Juo and
Ellis, 1968). The very poor utilization of insoluble crystalline STR
and WA also concurs with previously reported findings (Taylor et al.,
1963; McLachlan, 1965; Juo and Ellis, 1968).


60
Table 5. Continued
Method*
Extractant
pH
Soil: Extract
solution time
min
Oxalate
0.2M (NH4)2C204 + 0.1M H2C204
3.0
1 :50
60
T ruog
0.002N_ H2S04/(NH4)2S04
3.0
1:100.
30
Truog(2)
1 1
3.0
1 :20
30
T ruog(3)
1 1
3.0
1:5
30
T ruog(A)
0.02N^ H2S04/(NH4)2S04
2.1
1:100
30
T ruog(5)
1 1
2.1
1:5
30
Truog(6)
II
2.1
1:5
30
Truog(7)
0.2N H2S04/(NH4)2S04
1.1
1 : 100
30
T ruog(8)
II
1.1
1:20
30
Truog(9)
II
1.1
1:5
30
Truog(10)
0.02N^ H2S04/Na2Mo04
2.5
1:100
30
T ruog(11)
1 1
2.5
1:20
30
Truog(12)
II
2.5
1:5
30
h2so4
0.002h[ H2S04
2.7
1:100
30
H2S04(2)
1 1
2.7
1:20
30
h2so4(3)
II
2.7
1:15
30
h2so4(A)
0 -02N^ H2S04
1.7
1:100
30
h2so4(5)
II
1.7
1:20
30
h2so4(6)
II
1.7
1:5
30
h2so4(7)
0.2NI H2S04
0.8
1:100
30
h2so4(8)
II
0.8
1 :20
30
h2so4(9)
II
0.8
1:5
30


Table 47. Continued
Repli-
cate
Height
1 year 3 years 5 years
Po Pi P 2 Po Pi P2 P0 Pi P2
crrv
Leon fine sand (A 19)
1
91.0
92.6
86TT
300
347
305
501
586
507
2
83.9
85.0
85-4
291
282
339
504
492
575
3
85.4
86.2
83-5
280
283
315
488
497
543
Ona f i
ne sand
(A23)
3
42.0
50.8
61.4
156
170
204
342
359
422
Kershaw
fine sand (A5)
1
29.1
23.4
31.1
131
102
130
262
227
263
2
24.2
26.0
27-5
129
122
109
287
260
251
3
27.9
25-7
31.8
146
144
132
329
312
282
Lakeland fine sand (A25)
1
27.4
24.7
25.5
121
109
130
263
237
290
2
24.9
24.7
31.0
134
115
119
277
252
258
3
29.1
25.4
28.8
133
117
131
285
260
280
* Po = 0, Pi = 22.5. P2 = 90 kg P/ha. All plots received N.


2!
Table 1. Foliar P concentrations prior to fertilization in relation to
response of southern pines to P fertilizer
T ree
age
Foliar P
(Unferti 1ized)
Response to
P fertilizer
Reference
yr.
%
(A) Slash pine
(Pinus el 1iottii)
8
0.053
yes
Young (1948)
6
0.0148
yes
Baur (1959)
15
0.075
0.105
yes
yes
Pritchett and
Swinford (1961)
9
0.08
no
Walker and
Youngberg (1962)
3-5
<0.10
>0.10
yes
no
Pritchett and
Llewellyn (1967)
5-8 .
<0.09-0.10
>0.09-0.10
yes
no
Pritchett (1968)
1
0.08
no
Schultz (1969)
1-30
<0.075-0.80
>0.075-0.80
<0.085-0.10
>0.085-0.10
yes (90%)*
no (90%)
yes (100%)
no (100%)
Richards and
Bevege (1972b)
(B) Loblolly pine
(Pinus taeda)
8
0.064 .
yes
Young (1948)
6
0.057
yes
Baur (1959)
1
0.089
yes
Fowel1s and
Krauss (1959)
5-9
0.090
no
Zahner (1959)
5-10
0.137
no
Maki (I960)
22
0.124
no
Thompson (1960)


30
Working with young plantations of loblolly pine in the lower
Coastal Plain of South Carolina, Wells and Crutchfield (1969) reported
that soil P extracted by 0.05N. HCI + 0.025N. H2S0^ was not significantly
correlated with the first year height growth. The authors suggested the
soil-test values would prove more reliable as the trees became older.
This was confirmed by Wells et al. (1973), who found soil P extracted by
the above extractant to be significantly correlated with the height of
unfertilized loblolly pine at age 3 years in the same region. They also
reported the Bray 2 extractant to be equally effective. According to
Thomas and Peaslee (1973), 0.05N_ HC I + 0.025N_ extracts P predomi
nantly from the Ca-P and Al-P fractions. However, the acid soils of the
southeastern lower Coastal Plain contain little Ca-P (Humphreys and
Pritchett, 1971). The effect of age of trees on the value of soil-P tests
as predictors of height growth was also observed by Ballard (197^). Work
ing with radiata pine on heavy textured acid soils in New Zealand, he
found that levels of soil P were most closely related to height at age 3,
i
followed by height at age 2 and least with height at age 1 year. The
Bray 2 and Olsen extractants (0.5M_ NaHCO^, pH 8.5) were the most success
ful in this study. On acid soils, the Olsen test extracts P from the
Al-P and Fe-P fractions, although with somewhat less intensity than
extractants,such as Bray,containing F (Thomas and Peaslee, 1973).
Extractable P vs foliar P. There are numerous reports in the lit
erature of correlations between foliar-P concentrations in pine species
and levels of soil P extracted by various soil tests; Wells (1965) found
P extracted by the Truog and Bray 2 tests to satisfactorily predict
foliar P of 5-year old loblolly pine growing in the South Carolina
Piedmont. However, Metz et al. (1966), in an intensive study on one soil


116
Table 20. Regression equations relating relative height of slash pine at
age I, 3, and 5 years (Y1, Y3, and Y5) to the log transformed
P extracted from the surface 20 cm of soil by the h^O (XI) and
HCl-f^SO^ (X2) soil-test methods
Regress ion
equations
R2
Y1
=
13.88**X1
+ 85.00
0.322**
Y1
=
19.63**X2
+ 71.53
0.306**
Y1
=
11.86**X2
+ 9.05**X1 + 77.25
0.395**
Y3
=
9.99**X1
+ 84.28
0.149**
Y3
=
18.65**X2
+ 71.83
0.246**
Y3
=
15.49X2
+ 3.68X1 + 74.15
0.259**
Y5
=
8.25**X1
+ 88.13
0.120**
Y5
=
13.07**X2
+ 79.27
0.143**
Y5
=
9.20* X2
+ 4.50X1 + 82.11
0.166**
* Significant
at the 5%
leve 1.
**Significant
at the 1%
level.
Table 21.
Simple correlation coefficients(r)
between
amounts of P
extracted by
the HC1-H?S0^ method
from two
soi1 positions
and three soil depths
Depth
o :
20B 0-20
20 40
40 60
0 20B+
1.000 0.983**
r
0.928**
0.468**
0-20
1 .000
0.934**
0.505**
o
-cr
i
o
CM
1 .000
0.614**
40 60
1 .000
^ Surface sample collected from within the bedded tree row.
**S ignif¡cant at the level.


91
this effect may be offset by Na from the extractant causing release of
Ca from soil exchange sites. Alkaline NH^F, the extractant used to
selectively remove Al-P in most P-fractionation procedures, extracted
appreciable amounts of DCP. This has been recorded previously
(Bromfield, 1967) and attributed to CaF2 formation. The degree of ex
traction by the NH^F method was unaltered in the presence of Immokalee
soil but substantially reduced in the presence of McLaurin soil. The
explanation for this low P recovery was not clear as it could not be
explained by adsorption of P during extraction.
Fluorapatite (FA)
The solubility of FA in the extractants correlated fairly close
ly with the pH of the extractants. However, soil-test methods with
extractants capable of forming stable Ca complexes (lactate) and those
utilizing wide so i 1:solution ratios (Truog) and long extraction periods
(lactate) extracted more P than was anticipated, based on the pH of
these extractants. The reduced recovery of P from FA by the NH^OAc,
Truog, and HC1-H2S0^.methods in the presence of soil was consistent with
the P adsorption anticipated on the basis of the MCP data. However,
reduced recovery of P in the presence of soil by the lactate and both
Bray methods was greater than anticipated from P adsorption alone. In
the lactate method, this may have been caused by soil cations competing
with FA-Ca for lactate and acetate functional groups. The considerable
reduction in P recovery by the Bray 1 method may partially be explained
by the poor buffering capacity of this extractant; the final pH of the
Immokalee soil extraction solution was 2.7 and that of the McLaurin
3.0. The more acidic Bray 2 extractant was better buffered and it can
only be speculated that the reduced P extractabi 1ity arose from


LITERATURE CITED
Ahenkorah, Y. 1968. Phosphorus retention capacities of some cocoa
growing soils of Ghana and their relationship with soil properties.
Soil Sci. 105:24-30.
Alban, D. H. 1372. The relationship of red pine site index to soil
phosphorus extracted by several methods. Soil Sci. Soc. Amer. Proc.
36:664-666.
Allison, L. E. 1965. Organic carbon, p. 1346-1366. ]r C. A. Black
(ed.) Methods of soil analysis. American Society of Agronomy, Madi
son, Wis.
Anderson, H. W. 1969. Forest fertilization: Douglas-fir region aerial
fertilization makes strides in a brief span of years. Forest Ind.
96:30-32.
Armitage, I. P. 1969- The establishment of second rotation radiata
pine in Riverhead forest. New Zeal. J. For. 14:184-194.
Armson, K. A. 1973- Soil and plant analysis techniques as diagnostic
criteria for evaluating fertilizer needs and treatment response,
p. 155-166. I_n^ Forest fertilization symposium proceedings. USDA
For. Serv. General Tech. Rep. NE-3.
Bache, B. W. 1963. Aluminum and iron phosphate studies relating to
soils: 1. Solution and hydrolysis of variscite and strengite. J.
Soil Sci. 14:113-123.
Baker, J. B., and R. H. Brendemuehl. 1972. Soil phosphorus level ade
quate for growth of Ocala sand pine seedlings, a greenhouse evalu
ation. Soil Sci. Soc. Amer. Proc. 36:666-667.
Ballard, R. 1370a. The phosphate status of the soils of Riverhead
forest in relation to growth of radiata pine. New Zeal. J. For.
15:88-99.
Ballard, R. 1970b. A preliminary investigation of the phosphate status
of Riverhead forest soils. New Zeal. J. Sci. 13:312-322.
Ballard, R. 1974. Use of soil testing for predicting phosphate ferti
lizer requirements of radiata pine at time of planting. New Zeal.
. J. For. Sci. 4:27-34.
259


Table 56. Continued
Field
s i te
Repli-
cate
Depth
Particle
size distribut ion
Moisture content
Bulk
density
C 1 ay
Silt
Sand
1/3 atm.
15 atm.
g/cc
'O
A16
1
0-20
1.4
6.0
92.6
5-7
. 5.3
1.14
20-40
2.0
4.3
93.7
1.7
1.3
1 .66
40-60
1.4
3-3
95-3
1.3
1.1
1.72
2
0-20
2.9
4.6
92.5
5.4
3.9
1.14
20-40
2.0
4.0
94.0
1.6
1.5
1.67
40-60
2.0
3.3
9^.7
1.2
0.8
1.68
3
0-20
2.4
4.6
93.0
4.9
3.8
1.14
20-40
2. 1
3.4
94.5
1.8
1.5
1.70
40-60
2.0
3.7
94.3
1.5
0.8
1.74
A17
1
0-20
6.9
28.1
65.0
20.6
10.7
1.46
20-40
1.5
1.8
96.7
1.8
1 -5
1.73
40-60
1.7
1.4
96.9
1.6
1.3
1.72
2
0-20
15.1
24.9
60.0
31.0
14.4
1.44
20-40
1.9
1.6
96.5
1.3
0.7
1.62
40-60
1.5
2.7
95.8
1.1
0.9
1.72
3
0-20
11.2
22.8
66.0
23-3
12.1
1.45
20-40
1.9
0.6
97-5
1.7
1.4
1.70
40-60
1.8
2.4
95-8
2.0
0.5
1.74
A18
1
0-20
1.5
9-3
89.2
8.7
6.3
1.20
20-40
0.6
h.7
92.7
3-2
0.9
1.85
40-60
6.6
10.0
83.4
9-0
2.9
2.04
2
0-20
1.1
7.4
91.5
4.6
2.9
1.22
20-40
1.0
6.2
92.8
1.9
1.1
1.87
N>
V>J
U~1


134
Table 29. Relationships between selected soil and site properties and
P fertilizer required to achieve 95% of maximum height after
1, 3, and 5 years' growth on 72 field sites
Soil or site
property
Frtilizer
requirements
(95%)
1 year
3 years
5 years
r2
pH
0.073
0.126**
0.046
Organic matter
0.003
0.199**
0.066
Nit rogen
0.083*
0.140**
0.069
CEC
0.011
0.144**
0.061
NH^OAc (pH !*.8) extractable
-Ca
0.069
0.003
0.031
-Mg
0.000
0.096*
0.007
-K
0.179**
0.079
0.000
-A1
0.116*
0.005
0.025
Silt + clay
0.268**
0.087*
0.051
Available moisture
0.225**
0.098*
0.064
Depth of A1 horizon
0.015
0.009
0.002
Depth to LH
0.135**
0.036
0.062
Drainage class
0.038
0.067
0.109*
* Significant at the 5% level,
using the model
Y=aX + bX2 +
c.
**Significant at the 1% level.


Table 11. Relationships between selected soil-test values and tissue P
parameters of greenhouse and field slash pine grown on 10 soils
Soi1-test
method or
P form
Mean
ext ract-
able P
Greenhouse
Field
P uptake
%P in tops
foliar P
1 +
2
1
2
4
r2
ppm
h2o
2.3
0.81.0**
0.322
0.537*
0.150
0.228
NH/jOAc
4.4
0.709**
0.607**
0.847**
0.454*
0.511*
T ruog
8.9
0.634**
0.828**
0.821**
0.599**
0.495*
Olsen
12.3
0.447*
0.875**
0.931**
0.806**
0.765**
HCl-H2S0il
15.2
0.516*
0.871**
0.925**
0.753**
0.706**
h2S04(9)
32.5
0.314
0.873**
0.811**
0.846**
0.698**
Bray 1 (3)
*1 .3
0.348
0.850**
0.895**
0.827**
0.817**
NH4F(pH 8.5)
57.0
0.266
0.722**
0.742**
0.746**
0.638**
Organic-P
42.6
0.141
0.459*
0.070
0.164
0.019
Tota 1-P
136.1
0.066
0.600**
0.213
0.442*
0.175
fTree age at time of sampling (years).
* Significant at the 5% level, using the model Y = b logX + c, where Y =
P parameter and X = soil-test value.
**Signifleant at the 1% level.


132
requirements over the 3 and 5-year growth periods and foliar P concen
trations at age A years. Expressing the soil-test results as amount of
extractable P within the effective rooting volume improved the predictive
ability of these tests over that obtained with results expressed on a
concentration in the surface 20-cm basis. The soil-test methods which
extracted small quantities of P from the soil (H^O, NH^OAc) provided the
best index of height growth in the absence of P fertilizer over all
growth periods considered, although the relationships deteriorated with
increasing age of trees.
Several soil and site variables, including pH, extractable Ca,
silt plus clay, drainage class, and depth to limiting horizon were sig
nificantly correlated with growth and response parameters. Drainage
class and depth to limiting horizon were more closely related to growth
and response at age 5 years than P extracted from the surface 20 cm of
soil by any of the five soil-test methods. This was attributed to these
variables providing an integrated index of both P supply and moisture
conditions. The significance of most of the other variables was ex
plained in terms of their relationship to available soil-P levels.
The P-retention characteristics were determined in the laboratory
for surface horizons of k2 forest soils, representative of the major soil
types used for forestry in the lower Coastal Plain. Relationships be
tween a range of soil physical and chemical properties, conventionally
used to characterize soils, and the P-retention capacity of the soils
were examined. In addition, the effectiveness of A1 and Fe, extracted
from the soil by soil-P test methods and present in the foliage of slash
pine growing on these soils, at predicting the P-retention capacity of
the soil was determined. Soil properties found to provide the best


109
the responsiveness over longer growth periods. This might cause a dete
rioration in the relationship between topsoil-P values and responsiveness,
particularly where topsoil and lower horizon-P values are not closely
related. This aspect is examined in some detail below. (c) With in
creasing age the relative requirements of trees for such factors as nu
trients other than P and moisture may change. It can be speculated that
some factors may not limit early responsiveness to P fertilizer, but may
do so at a later stage of growth. (d) Height measurements tend to become
less sensitive indices of growth performance with increasing age of trees
so that by age 5. these may not adequately reflect true growth response
to P fertilizer.
The relationships between P extracted by HC1-H2S0^ from the
0-20 cm depth and relative height of slash pine 1, 3, and 5 years after P
fertilization are shown in Figs. 4, 5, and 6, respectively. These rela
tionships are strikingly similar to those reported by Wells et al. (1973)
between response to P fertilization of loblolly pine and HCl-H^SO^--.
extractable P. Applying the Cate and Nelson (1965) technique of deter
mining the critical level of soil P, it was found that a soil-test value
of 5 ppm P provided the best separation of responsive and nonresponsive
sites, assuming a relative height value of 90% or above represented no
significant response. The separation of the 72 field sites into response
quadrants using values above or below 5 ppm P and 90% relative height are
shown in Table 19. The data in Table 19 illustrate three points. First,
the use of 5 ppm of HC1-H^SO^-extractable P provides a reasonably good
separation between responsive and nonrespons ive sites. Second, as the
growth period increases, the number of responsive sites with extractable
P > 5 ppm decreased. This can probably be attributed to some of these
sites having low water-soluble P, which resulted in a tree response to


Table 46. Continued
T reat-
ment
Repl i -
cate
1 year
l years
Ht.T
Dry wt.^
tops
P In
tops
Ht.+
Dry wt.
Tops Roots
P
Tops
Roots
cm
9
/
'O
cm
c
3 .
%
Marl boro
fine sandy
loam
(A24)
1
11.5
XT
0.058
18.8
28.1
8.8
0.036
0.045
Po
2
12.3
6.4
0.055
19.8
31.0
10.0
0.042
0.052
3
12.3
7.9
0.062
19-8
25.2
9.1
0.036
0.043
1
17.1
17.8
0.076
35-8
83.8
25.2
0.054
0.053
Pi
2
16.5
12.6
0.076
31.5
76.8
20.6
0.055
0.058
3
16.3
19-6
0.081
32.3
87.0
26.0
0.057
0.058
1
17.0
26.1
O.O65
3.13
88.2
33.0
0.041
0.051
P 2
2
18.9
22.6
0.084
37.0
99.0
25.8
0.066
0.065
3
17.9
25-7
0.097
34.5
100.0
31 .2
0.049
0.052
1
17.6
26.2
0.114
35-8
106.8
40.8
0.078
0.092
P 3
2
18.5
21.2
0.117
38.5
120.7
33.2
0.084
0.086
3
15.9
21.7
0.132
39-8
103.4
28.3
0.089
0.089
McLaurin
fine sandy
loam
(A28)
1
10.8
"8.0
0.058
20.0
29.4
10.5
0.048
0.052
Po
2
14.3
7.1
0.040
27.5
35-8
14.1
0.050
0.053
3
12.0
7.9
0.068
23.8
36.3
1 1.0
0.044
0.050
1
13-9
11.6
0.078
26.0
59.5
22.8
0.051
0.069
Pi
2
16.4
16.6
> 0.081
31 .8
74.2
22.5
0.065
0.066
3
14.0
15.3
0.070
30.3
56.9
16.1
0.059
0.068
oo
U>


Table 58. Continued
Field
Rep 1i-
Ferti
1 izer P
requirement
s i te
cate
1 year
3 yea
rs
5 years
90
95
100
90
95
100
90
95
100
A27
1
77
84
90
55
74
90
0
0
90
2
84
86
90
0
0
0
0
0
0
3
40
56
92
31
49
92
0
15
71
A28
1
25
52
118
3
25
77
0
0
63
2
72
101
171
27
39
68
16
30
64
3
44
66
90
13
27
59
1
19
61
A29
1
31
43
69
26
34
52
21
30
50
2
33
47
80
28
38
63
22
32
57
3
27
36
56
29
37
56
22
32
58
* Fertilizer applied in 1.22 m bands down tree rows on 3*05 m centers; rates within bands are
2.5 times those shown above.


Table 55* Continued
Field
s i te
A8
A10
TTcHP
RepiI -
cate
A
Depth
h2o
P
NH,OAc
4
P
H2S04
P
Bray 1(3)
P
Bray 2
P
NH.
4
OAc
pH
(h2o)
Ca
Mg
K
A1
cm
ppm--.
1
40-60
0.2
0.7
6.1
19.7
17.9
90
10
2
33
5-3
2
0-20B
0.7
2.2
8.9
32.1
34.3
75
10
7
47
5.1
0-20
0.4
2.2
12.5
38.9
41.7
75
25
7
47
5.1
20-40
0.3
1.3
12.1
34.6
41.4
30
14
2
55
5.2
40-60
0. 1
1.1
8.8
24.6
26.8
30
19
2
50
5.3
3
0-20B
0.8
4.0
15-3
50.6
53.2
75
21
12
33
5.1
0-20
1.0
2.6
11.5
42.7
37.6
45
30
2
28
5.1
20-40
0.5
2.2
24.9
75-0
70.0
30
30
2
30
5-2
40-60
0.4
1.3
19.1
56.1
53.2
30
30
2
30
5-3
1
0-20B
2.5
1.8
4.0
5-8
5.5
135
61
10
16
4.3
0-20
1.9
2.2
3-3
4.4
4.5
135
56
12
19
4.3
20-40
0.2
0.5
0.7
1.7
1 .2
15
19
2
64
4.9
40-60
0.1
0.2
1.5
2.4
2.7
10
14
2
135
4.9
2
0-20B
0.4
0.9
2.7
5.3
5.3
90
38
10
95
4.5
0-20
0.4
0.7
2.5
5-3
5-0
45
25
10
104
4.6
20-40
0.1
0.2
1.3
2.0
3.4
45
14
2
119
5.0
40-60
0.1
0.2
1.6
2.3
5-3
15
10
2
90
5.2
3
0-20B
0.3
0.7
2.5
5.3
5.8
30
21
12
123
4.8
0-20
0.2
0.5
2.0
4.8
6.1
15
10
7
111
4.8
20-40
0.1
0.2
1.4
2.8
4.1
10
2
2
128
5.1
40-60
0.1
0.2
1.7
2.8
5.9
15
2
2
125
5.2
1
0-20B
0.7
3.1
18.4
57.0
69 6
30
5
7
64
5-2
A15
222


Resin-extractable P was determined using a strongly basic,
quaternary ammonium, anion-exchange resin RN(CH^)^ Cl The soil was
ground to pass a 0.25mm-sieve and all resin exceeded this sieve size.
Following the soil extraction, the soil and resin were separated by wash
ing the soil-resin mixture with distilled water over the 0.25mm sieve.
The resin was quantitatively transferred to a filtering apparatus and
leached successively with five 5_ml portions of 2N^ NaOH followed by five
5-ml portions of 2N_ HC1. This procedure was found in preliminary work
to recover greater than 90% of the P sorbed by the resin. Phosphorus was
determined in this leachate.
Soil P was fractionated by a modified Chang and Jackson procedure
developed by Fife (Ballard, 1970 b). Fractions determined included
soluble-P, Al-P, Fe-P, Ca-P and organic P. Total P was extracted by
Na2C0^ fusion (Jackson, 1958). Dispersed material present in alkaline
extracts was removed by addition of P-free activated carbon at the
filtering stage.
Phosphorus in all solutions was determined color¡metrically by the
Murphy and Riley (1962) technique using ascorbic acid and molybdate-
sulphuric acid as modified by Watanabe and Olsen (1965). All measure
ments were made using a Unicam SP 600 spectrophotometer with a wavelength
setting of 880 my. Aliquots taken for developments were adjusted to pH 5
with H2S0^, using para-nitro phenol indicator, prior to the addition of
the developing reagent.
Extractants containing NH/jOAc and HOAc required adjustment of the
aliquote to pH 2.5 with HjSO^, using 2,4 dinitrophenol indicator, in
order to obtain color development. Fluoride interference in the extracts
containing NH^F was eliminated by addition of saturated boric acid
(Kurtz, 1942). Oxalate and citrate were found to interfere with color


Table 55*. Continued
Field
s i te
Repli-
cate
*
Depth
V
P
NH,OAc
4
P
~RTT=
H2S4
P
Bray 1(3)
P
Bray 2
P
Ca
NH.
4
Mg
O
<
o
A1
pH
(h2o)
ppm--.
A28
1
40-60
0.1
0.1
0.3
0.1
1 .2
135
82
12
71
5-3
2
0-20B
0.2
0.5
1.3
1.2
3.3
200
61
28
62
5.3
0-20
0.2
0.2
1.2
1 .0
3.0
155
88
20
73
5.4
20-40
0. 1
0. 1
0.3
0.1
1 .0
105
88
15
86
5.3
40-60
0.1
0. 1
0.2
0.1
1.0
120
66
20
66
5.3
3
0-20B
0.1
0. 1
1 .0
0.9
2.6
120
66
20
66
5-3
0-20
0.2
0.2
1.2
1.3
2.0
155
66
28
62
5.3
20-40
0. 1
0.1
0.3
0.1
1.0
90
72
12
69
5-3
40-60
0. 1
0.1
0.3
0.1
1 .2
90
82
15
86
5-3
A29
1
0-20B
0.2
0.7
2.2
2.9
5-6
30
50
28
185
4.8
0-20
0.1
0.7
2.0
2.3
4.6
15
42
20
193
4.8
20-40
0.1
0.1
0.5
0.3
1.5
15
30
4
156
4.9
40-60
0.1
0.1
0.3
0. 1
1.0
45
34
4
131
4.9
2
0-20B
0. 1
0.2
1 .2
1.6
3-3
45
34
7
66
5.1
0-20
0.1
0.5
1.3
1.5
3.0
45
30
7
64
5.2
20-40
0. 1
0.1
0.4
0.3
1.3
60
30
4
45
5-2
40-60
0. 1
0. 1
0.3
0. 1
1.3
90
25
4
35
5-2
3
0-20B
0. 1
0.5
1.8
1.8
4.2
30
38
20
145
5.0
0-20
0.1
0.5
1.6
1.7
4.0
15
34
18
128
5-1
20-40
0.1
0.2
0.4
0.1
1.3
30
34
4
114
5.1
40-60
0.1
0.1
0.2
0.1
1 .0
120
56
12
119
5.0
* 0-20B = sample collected from bedded area.


26
predominantly organic and the other predominantly inorganic. He sug
gested that it is doubtful if a single extraction agent could yield a
valid estimate of nutrient availability in view of the behavior differ
ences of organic and inorganic colloids in nutrient retention.
The value of sampling subsurface horizons has been examined in
many studies. Kessell and Stoate (1938) recognized the importance of
the vertical distribution of soil P in their site classification studies.
In outlining the P levels required for successful afforestation of
radiata pine (P.radiata D. Don), they wrote:
A P20c content of *00 parts per million (ppm) is required in the-
surface and subsurface soils. Three hundred parts per million
may be satisfactory if this content is maintained for a depth of
two to three feet. (p. 28)
In a survey of factors influencing the growth of radiata pine in New
Zealand, Jackson (1973) found that available soil P within the effective
rooting zone was more closely correlated with growth than with available
P in the surface 7-5 cm of soil. Similarly, White and Leaf (1964)
reported that the content of HNO^ extractable K in the upper 1.7 metres
of soil was more closely correlated with height and K content in the
biomas of red pine (P. resinosa AIT) than was the extractable K in the
surface horizon. The importance of subsurface nutrient supply was il
lustrated by the work of Ellerbe and Smith (1963) who found that
occasional serious underestimation of site quality by the soil-site
predictions of Coile (1952) in the lower Coastal Plain of South
Carolina could be attributed to the presence of underlying phosphate
marl at these sites. Will (1966) also reported that the disappearance
of Mg deficiency in radiate pine growing on volcanic soils in New
Zealand coincided with the root penetration of Mg-rich burred topsoil
at age 6 to 8 years.


Table 6. Classification and
selected
properties of
10 soils
used in greenhouse
study
1
Soil
pH
Silt +
Organic
Ext. A1
Extractable P
Type
Order
(h2o)
clay
CEC
matter
NH/jOAc
NH^OAc
h2o
Bray 1(3)
£
me/100g
%
ppm--
Bladen scl*
U11 i sol
4.8
43-5
8.67
3.27
222
0.7
0.5
1 .0
Blanton fs
U 11isol
5.2
5.5
2.17
1.24
40
4.3
1.7
48.9
Plummer fs
U11i so 1
5.0
8.5
3-94
2.23
105
13.8
3.0
185.5
Marlboro fsl
U1tisol
5.0
38.5
4.70
2.57
149
0.4
0.2
1.6
McLaurin fsl
U1tisol
5.3
31.0
4.70
2.20
91
0.3
0.3
1.5
Immokalee fs
Spodosol
4.3
7.5
5.46
3.27
8
6.3
5.5
6.6
Leon fs
Spodosol
4.1
6.5
9.28
4.52
11
7.8
6.4
7.9
Ona fs
Spodosol
4.2
10.3
4.90
3.76
55
2.1
2.4
4.7
Kershaw fs
Entisol
5-2
4.0
2.09
1.31
35
2.6
0.7
25-9
Lakeland fs
Entisol
5.4
14.8
4.90
2.57
66
5-9
2.2
130.0
* sel = silty clay loam; fs =* fine sand; and fsl = fine sandy loam textures.


20
Diagnostic Methods
This section of the review is concerned principally with the use
of foliar and soil analysis as diagnostic aids in predicting P deficien
cy and the need for fertilizer in pine plantations. Major emphasis is
given to the use of soil analysis since this is the method under inves
tigation in this study.
Foliar Analysis
The principles, physiological basis,and problems inherent in the
use of tissue analysis for diagnosing the nutritional status of crops
have been extensively reviewed by Goodall and Gregory (19^7) and Smith
(1962). The value and specific problems associated with the use of
foliar analysis in forestry have been the subject of several extensive
reviews (Tamm, 1964; Qureshi and Srivastava, 1966; Raupach, 1967;
Richards and Bevege, 1972 a; Armson, 1973; Leaf, 1973)-
Published information on foliar P concentrations in relation to
responses of slash and loblolly pines (P. el 1iottii and P. taeda) to P
fertilizer is summarized in Table 1. Whereas most research on the use
of foliar analysis in forestry has been concerned with relating foliar
nutrient levels to productivity or site index (Leaf, 1973). foliar
analysis research with Southern pine has been almost exclusively related
to establishing 'critical levels' for both predicting fertilizer needs
and following the effectiveness of fertilizer applications (Table 1).
The critical level of foliar P has been defined by Pritchett
(1968) as the point at which trees with a higher concentration of
needle P would not be expected to respond significantly to applications
of phosphate fertilizers, but at which trees with a lower needle P


32
excessively to poorly drained. In their study, P extracted by NH^OAc
(pH 4.8) was reported to be superior to that extracted by other extract
ants, including 0.03N_ NH^F + 0.025N_ HC 1 and 0.051 NCI + 0.025N H2S04, at
predicting response to P fertilizer. However significant correlations
were obtained only when the well-drained sites were eliminated from the
statistical analysis.
Forms of tree-available P. McKee (1973) concluded from a green
house fertilizer trial with N and P rates using slash pine grown in a
Caddo silt loam, that Ca-P and Al-P were the sources of P available to
the pine seedlings. This conclusion was based on the finding that the
combined value of these two fractions in the soil at the end of the ex
periment was significantly correlated with P uptake by the seedlings. He
found individual fractions of Ca-P, Fe-P, and Al-P were not related to P
uptake, nor was Fe-P in combination with either Ca-P or Al-P. The rele
vance of results from greenhouse trials to the nutrition of field grown
trees must however be questioned in view of the finding by Mead and
Pritchett (1971) that the response, and hence nutrition, of greenhouse
grown seedlings differed from that of field grown trees on the same
soils.
The literature reviewed in this section provides no conclusive
evidence as to what forms of soil P are utilized by trees and which soil
testing procedures are likely to be most successful in forestry. The
situation appears to parallel that in agriculture in which it has been

found that the value of any particular soil test depends upon the objec
tive in soil testing (predicting growth, nutrient uptake, or responsive
ness to fertilizer addition), the range of soils used, and the species
of plant involved (Williams, 1962; Williams and Knight, 1963). However,
despite some evidence that mycorrhizae may increase the availability of


Table 55* Continued
Field
s i te
A26
A27
HCi-
Repl i -
cate
/V
Depth
V
P
NH.OAc
4
P
H2S4
P
Bray 1 (3)
P
Bray 2
P
NH.OAc
4
pH
Ca
Mg
K
A1
cm
ppm--
2
0-20B
0.6
1.8
6.8
14.5
18.5
45
34
12
95
4.6
0-20
0.3
0.7
2.6
4.8
6.7
30
34
10
97
4.6
20-40
0.1
0.5
1 .0
1.4
2.9
45
25
4
102
4.9
40-60
0.1
0.2
0.7
1 .2
3.0
45
25
2
86
4.9
3
0-20B
0.6
1.3
8.1
13.6
20.7
60
34
12
88
4.6
0-20
0.2
0.9
2.1
3.7
5.1
30
30
7
88
4.9
20-40
0. 1
0.5
1.3
1.8
3.9
15
25
2
90
5-0
40-60
0.1
0.2
1.8
2.8
6.8
45
25
2
66
5.1
1
0-20B
0.2
0.5
1.5
1.5
3.5
75
50
20
78
5.1
0-20
0. 1
0.5
1.3
1.4
3-5
75
48
25
83
5-2
20-40
0.1
0.1
0.3
0.1
1.2
75
50
10
88
5.2
40-60
0.1
0.1
0.3
0.1
1.2
60
66
10
92
5.2
2
0-20B
0.3
0.2
2.7
5.4
6.4
415
88
38
50
5-5
0-20
0.5
0. 1
2.5
3.2
5.6
380
82
35
55
5-5
20-40
0.1
0. 1
0.4
0.1
1.4
215
76
20
66
5-5
40-60
0.1
0. 1
0.3
0.1
1.9
135
76
18
111
5-3
3
0-208
0. 1
0.2
1.2
1.5
3.0
460
66 '
18
62
5-5
0-20
0. 1
0.2
1.2
1.1
2.5
215
61
18
81
5.4
20-40
0.1
0.2
0.4
0.1
1.3
135
76
10
123
5-2
40-60
0.1
0.1
0.2
0.1
1 .2
60
72
10
180
5-2
1
0-20B
0.2
0.2
1.2
0.9
3.4
105
61
25
66
5-3
0-20
0. 1
0.2
1.0
0.9
2.7
90
61
22
71
5.3
20-40
0. 1
0.1
0.3
0.1
1.2
105
76
10
73
5-3
A28


Table 52. Dry weight and P uptake of entire slash pine seedlings, as affected by P source,
after the 8 months of growth on two soils in the greenhouse
1mmoka1ee
fine :
sand
McLaurin fine
sandy
1 oam
Dry weight
*v
P uptake
Dry weight
P uptake
P source
2
3
1
2
3
i
2
1
2
-g/pot~-
~
-mg P/pot-

g/pot
-mg
P/pot-
None
15.2
11.9
11.6
15.99
17.58
14.99
15-6
16.2
10.53
12.19
Monocalcium
phosphate
9.2
11.6
13.2
40.99
51.29
52.51
19-5
20.4
19-31
29.07
Dica1cium
phosphate
10.6
13.**
14.1
50.04
52.34
48.84
21.0
17-7
18.44
24.01
FIuorapatite
11.2
10.2
12.2
39-67
40.74
42.89
17.3
14.0
13.56
8.24
Colloidal alum
inum phosphate
10.5
12.8
11.7
46.71
28.62
41.37
21.1
21.4
23.97
23-39
Potassium
taranak? te
12.6
12.6
10.1
44.4
40.60
38.26
16.9
19-2
30.34
25.53
Wavel1ite
12.3
13-5
12.3
20.19
19.56
20.54
16.6
16.0
12.77
1 1 .20
Colloidal ferric
phosphate
12.8
13-2
12.5
45-17
42.19
44.41
18.5
19-1
29.32
24.24
Strengite
12. 4
12.4
12.4
20.95
18.74
19.91
19.4
16.8
15.46
14.03
* Shoots + roots of three seedlings per pot.
t Replicates.
212


148
Table 37. Relationships between height, relative height, and P-
fertilizer requirements (95%) at age 1, 3, and 5 years,
and P concentrations in the foliage of 4-year-old slash
p i ne
Age
Height
Rel. height
Fert. reqm.
d2
yr
1
0.015
0.506**
0.103**
3
0.194**
0.544**
0.125**
5
0.278**
0.485**
0.100**
** Significant at the
1% level, using
the model Y = b
logX + c, where
X = foliar P.
Table 38.
Relationships between extractablc-soi1
three depths, and within the effective
and P concentrations in the foliage of
pine
P at two positions,
soi1 depth (volume)
4-year-old slash
Soi1-test
Soi1 sample
method
0-20B cm
0-20 cm
20-40 cm
40-60 cm
Volume
r2
H2
0.159**
0.183**
0.154**
0.140**
0.357**
NH.OAc
4
0.273**
0.271**
0.281**
0.292**
0.479**
HC1-H SO,
2 4
0.588**
0.644**
0.494**
0.473**
0.708**
Bray 1(3)
0.602**
0.618**
0.299**
0.319**
0.627**
Bray 2
0.601**
0.617**
0.462**
0.462**
0.675**
** Significant at the 1% level, using the model Y = b logX + c, where
X = soi1-test value.


90
alkaline extractants of the Olsen and NH^F methods. Because of the high
solubility of MCP, recovery values in the presence of soil provide a
good index of the amount of P adsorption which occurs during extraction.
Adsorption in the presence of the Immokalee soil, which has virtually no
P-retention capacity (Table A), was minimal for all soil-test methods.
But in the presence of the McLaurin soil, which has a relatively large
P-retention capacity, adsorption ranged from 59% in water to only 3% in
the Bray 2 extractant. Soil-test methods in which appreciable adsorption
occurs during extraction are considered to provide a good index of the
intensity factor of soil-P supply (Williams and Knight, 1963). It is
noteworthy that almost as much adsorption occurred during extraction
with NH^OAc as with H^O, although one of the main purposes of the ace
tate ion in certain extraction solutions is supposedly to prevent
adsorption of P removed by other ions (Thomas and Peaslee, 1973)-
Dicalcium phosphate (DCP)
a
Acidic extractants completely dissolved DCP, and in the presence
of soil the degree of recovery was similar to that for MCP. Dicalcium
phosphate was only slightly soluble in water, but in the presence of
soil the degree of P recovery almost matched that for MCP, suggesting
almost complete dissolution of DCP. The acidity of the soil-water
systems (pH 4.4 and 5-1 for the Immokalee and McLaurin soils, respective
ly, at the conclusion of the extraction) was apparently sufficient to
dissolve DCP. The Olsen method (alkaline NaHCO^) extracted 40% of the
DCP but, in the presence of soils recovery values fell well below those
anticipated if adsorption was the only mechanism reducing P recovery.
Although the solubility of DCP was enhanced as a result of the bicarbon
ate ion reducing the activity of Ca in solution (Olsen et al. 1954),


Table 54. Continued
4?
1
Field5'
Repl i-
;Soi 1
Drainage
s i te
cate
Type
Order
;Wc
class
A20
1
Plummer 1fs
U1tisol
1
2
Plummer 1fs
U1tisol
1
3
Plummer 1fs
U1tisol
1
A21
1
Plummer fs
U1tisol
2
2
Plummer fs
U1tisol
2
3
Plummer fs
U11i so 1
2
A22
1
Leon fs
Spodosol
2
2
Leon fs
Spodosol
2
3
Leon fs
Spodosol
2
A23
1
Leon fs
Spodosol
2
2
Rutlege fs
1nceptisol
1
3
Ona fs
Spodosol
1
Mb
1
Marlboro fsl
U1tisol
4
2
Marlboro fsl
1)11 i sol
k
3
Marlboro fsl
U11i so 1
b
A25
1
Lakeland fs
Entisol
5
2
Lakeland fs
Enti so 1
5
3
Lakeland fs
Enti so 1
5
A26
1
Lynchburg Ifs
U1tisol
3
2
Lynchburg Ifs
U1tisol
3
3
Lynchburg Ifs
Ultisol
3
Depth
to LHT
Depth
of A1
Organ i c
matter
Total
N
CEC

cm
%-
meq/1OOg
30
30
2.20
0.098
3-76
20
20
2.3 b
0.119
3.43
18
18
1 -77
0.101
2.74
63
25
1.91
0.070
4.54
58
25
2.07
0.068
4.59
55
23
2.60
0.071
5.28
33
20
1.77
0.041
4.97
43
23
1.98
0.062
4.49
38
20
1.24
0.042
3.48
b8
25
2.54
0.074
6.06
28
.28
3-72
0.096
10.10
28
25
2.85
0.077
6.90
65
18
3.24
0.084
7.59
60
18
2.89
0.060
5.58
bO
15
2.00
0.057
4.92
so
23
1.37
0.057
3.96
so
20
1.31
0.046
4.06
so
23
1.20
0.035
3.43
53
23
1 .82
0.066
4.80
50
23
2.21
0.076
5.02
50
20
2.15
0.064
5.20


61
Table 5. Continued
Method*
Extractant
pH
Soil:
solution
Extraction
t i me
min
hci-h2so4
0.5N HC1 + 0.025N_ H2S04
1.3
1 :4
5
01 sen(2)
0.5M NaHCO^NH^OH
8.5
1:5
30
01 sen
1 1
8.5
1 :20
30
01sen(3)
II
8.5
1:50
30
01sen(4)
II
8.5
1:5
240
01 sen(5)
8.5
1 :20
240
01 sen(6)
II
8.5
1:50
240
01sen(7)
II
8.5
1:5
l6(hr)
01sen(8)
II
8.5
1:20
I6(hr)
01sen(9)
II
8.5
1:50
l6(hr)
Bray 2(2)
0.03N NH^F + 0.1N_ HC1
1.5
1.5
1
Bray 2(3)
1 1
1.5
1.5
30
Bray 2
II
1.5
1:10
1
Bray 2(4)
II
1.5
1:10
30
Bray 2(5)
1 1
1.5
1 :50
1
Bray 2(6)
II
1.5
1:50
30
Bray 1(2)
0.03N. NH^F + 0.025N. HC1
2.5
1:5
1
Bray 1(3)
II
2.5
1:5
30
Bray 1
II
2.5
1: 10
1
Bray 1(4)
II
2.5
1:10
30
Bray 1(5)
II
2.5
1:50
1


59
Table 5. Phosphorus extraction methods
Soil: Extract ion
Method* Extractant pH solution time
H20
h2o
7.0
1:5
min
30
H20(2)
h2o
saturated
paste
( 1st extract)
H20(3)
h2o
saturated
paste
(10th extract)
H20(4)
h2o
saturated
paste
(E10 extracts)
NaCl
0.01M NaCl/HClt
4.0
1:5
30
Na2S04
0.01M Na2SO4/H2SOi+
4.0
1:5
30
Na2MoO^
0.01M NaoMoO^/HCl
4.0
1 :5
30
Na2B47
0.01M Na2B^07/HCl
4.0
1:5
30
Na2B407(2)
0.01M Na2B407/Na0H
10.0
1:5
30
NH^OAc
0.7N NH^OAc + 0.54N_ HOAc
4.8
1:5
30
NH40Ac(2)
IN HOAc/NH^OH
3.8
1:5
30~
NH4OAc(3)
IN HOAc/NH^OH
2.8
1:5
30
HOAc
2.5% HOAc (v/v)
2.5
1:40
120
HOAc(2)
II
2.5
1:40
30
H0Ac(3)
II
2.5
1:5
120
HOAc(4)
II
2.5
1:5
30
Lactate
0.1N CH3CHOHCOONH4 + O.kH
HOAc 3-5
1:20
240
Lactate(2)
II
3.5
1:20
30
Lactate(3)
1 1
3-5
1:5
240
Lactate(4)
II
3-5
1:5
30
Citrate
1 % Citric acid (w/v)
2.2
1:10
24(hr)
Citrate(2)
II
2.2
1:5
30


LIST OF TABLES (continued)
Table Page
24 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN SELECTED SOIL AND
SITE PROPERTIES OF 72 FIELD SITES 121
25 REGRESSION COEFFICIENTS FOR MULTIPLE REGRESSION EQUATIONS OF
RELATIVE HEIGHT ON DEPTH TO LH, HC1-H SO,-EXTRACTABLE
P(0-20 cm) AND THE SQUARED TERMS OF THESE TWO PARAMETERS. 123
26 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN '.BEST FIT1
MULTIPLE REGRESSION EQUATIONS OF RELATIVE HEIGHT ON SOIL AND
SITE PARAMETERS AND THE SQUARED TERMS OF THESE PARAMETERS 124
27 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL DEPTHS
AND P FERTILIZER REQUIRED TO ACHIEVE 90, 95, AND 100% OF
MAXIMUM HEIGHT GROWTH AFTER 1, 3, AND 5 YEARS' GROWTH ON
72 FIELD SITES 127
28 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE SURFACE
20 cm OF SOIL WITH THAT EXTRACTED FROM WITHIN THE EFFECTIVE
SOIL DEPTH (VOLUME) AT PREDICTING THE FERTILIZER F REQUIRED
TO ACHIEVE 90, 95, AND 100% OF MAXIMUM HEIGHT GROWTH AFTER 1,
3, AND 5 YEARS' GROWTH ON 72 FIELD SITES 132
29 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES AND P ,
FERTILIZER REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT AFTER
1, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES 134
30 RELATIONSHIPS BETWEEN HC1 -H2S0/j-EXTRACTABLE P(0-20 cm) AND P
FERTILIZER REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT ON
GROUPS OF SOILS CLASSED ACCORDING TO THEIR AMOUNT OF
NH^OAc (pH 4.8)-EXTRACTABLE A1 135
31 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN 'BEST FIT'
MULTIPLE REGRESSION EQUATIONS OF P FERTILIZER REQUIRED TO
ACHIEVE 95% OF MAXIMUM HEIGHT ON SOIL AND SITE PARAMETERS
AND THE SQUARED TERMS OF THESE PARAMETERS 137
32 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL DEPTHS
AND HEIGHT OF SLASH PINE IN THE ABSENCE OF P FERTILIZER
AFTER 1, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES 138
33 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE SUR
FACE 20 cm OF SOIL AND THAT EXTRACTED FROM WITHIN THE
EFFECTIVE SOIL DEPTH (VOLUME) AT PREDICTING HEIGHT OF SLASH
PINE AFTER I, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES. . 139
34 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES AND
HEIGHT OF SLASH PINE AFTER 1, 3, AND 5 YEARS' GROWTH ON 72
FIELD SITES 142
x


suggested in the preliminary screening, that the failure of P extracted
by stronger extractants, such as the HCl-H^SO^, Bray 1(3), and Bray 2
methods, to provide a reasonable estimate of height over longer growth
periods was due to factors other than inadequate available soil P re
stricting growth. The failure of extractable P within the effective soil
depth to improve the estimate of height compared to extractable P in the
surface 20 cm (Table 33), lends support to this contention.
Relationships Between Other Soil and Site Parameters and Height
Soil pH, NH^OAc-extractab1e Ca and A1, silt + clay, available
moisture, depth to LH, and drainage class were all significantly related
at the 1% level to tree height at age 1 year (Table 34). However, all
these parameters, except depth to LH and drainage class, were significant
ly correlated with H^O-extractable P (Table 24), suggesting they
influenced growth only through their association with H^O-extractable P.
As with H^O-extractable P, the value of these parameters as predictors of
height decreased with increasing growth period. The value as predictors
of growth of both drainage class and depth to LH increased with increas
ing growth period. These two site factors were closely related (Table
24). The close correlation between drainage class and depth to LH has
been observed on other soil groups (DeMent and Stone, 1968).
For prediction purposes, depth to LH probably has an advantage
over drainage class since its measurement involves a less subjective as
sessment than drainage class. The relationship between depth to LH and
l " 1 . , - .. ......
height of slash pine after 5 years' growth is shown in Fig. 11. Although
not shown, the relationship between depth to LH (X, cm) and height of
slash pine after 3 years' growth (Y, cm) was described by the quadratic
equation:


Table 55- Continued
Field
s i te
Repli
cate
Vc
Depth
H2
P
NH^OAc
P
HC1 -
H2S0A
P
Bray 1(3)
P
Bray 2
P
NH.
4
OAc
pH
(h2o)
Ca
Mg
K
A1
cm
ppm
A22
3
0-20B
2.6
2.6
4.2
5-5
7.2
135
50
10
14
4.1
0-20
1.9
1.8
3-1
4.3
4.2
90
56
7
14
4.3
20-40
1 .2
4.0
16.1
72.1
70.9
45
48
2
50
4.6
40-60
0.4
2.2
26.4
121 .2
122.8
15
21
2
86
5.1
A23
1
0-208
4.3
4.4
6.5
7.4
7.2
185
66
12
6
4.3
0-20
3-9
5-3
5.2
5-6
6.1
170
66
15
6
4.4
20-40
1.8
1.3
1.8
2.5
2.7
90
38
2
6
4.6
40-60
0.2
0.7
1.8
5.3
5-1
15
25
2
100
4.7
2
0-20B
0.3
0.7
2.3
4.1
5.0
30
25
7
78
4.7
0-20
0.3
0.7
2.9
4.6
4.9
45
30
7
81
4.8
20-40
0.1
0.2
0.8
1.7
2.2
45
30
2
78
5.1
40-60
0.1
0.2
0.7
1.5
2.2
15
19
2
66
5-1
3
0-20B
1.2
1 1
2.7
5-3
5.1
75
38
10
45
4.3
0-20
0.5
0.7
2.1
3-7
4.1
45
30
7
62
4.6
20-40
0.1
0.5
1 .2
2.3
2.9
15
21
2
88
4.9
40-60
0.1
0.2
1.4
2.7
3.7
45
21
2
76
5-1
A24
1
0-20B
0.1
0.2
0.9
1.3
2.7
135
38
15
83
5-2
0-20
0.1
0.2
0.9
1.3
2.7
135
34
4
88
5-2
20-40
0.1
0.2
0.3
0.2
1.1
30
34
4
88
5-2
40-60
0.1
0.1
0.2,
0.1
1.1
30
50
10
114
5.2
2
0-208
0. 1
0.5
1.0
0.8
3.1
30
30
18
135
5.0
0-20
0. 1
0.5
1 .0
0.8
3.1
30
30
18
135
5-0
N>
NO


265
la Bastide, J. G. A., and C. P. van Goor. 1970. Growth-site relation
ships in plantations of Pinus el 1iott?i and Auraucaria angustifolia
in Brazil. Plant and Soil 32:349"36.
Leaf, A. L. 1963. K, Mg, and S deficiencies in forest trees, p. 88-
122. jjri^ Forest fertilization theory and practice. Tennessee
Valley Authority, Muscle Shoals, Ala.
Leaf, A. L. 1973. Plant analysis as an aid in fertilizing forests,
p. 427-454. J_n L. M. Walsh and J. D. Beaton (ed.) Soil testing
and plant analysis. Soil Sci. Soc. Amer., Madison, Wis.
Lehr, J. R., and W. E. Brown. 1958. Calcium phosphate fertilizers:
11. A petrographic study of their alteration in soils. Soil Sci. Soc.
Amer. Proc. 22:24-32.
Lehr, J. R., E. H. Brown, A. W. Frazier, J. P. Smith, and R. D. Thrasher.
1967. Crystallographic properties of fertilizer compounds. Ten
nessee Valley Authority, Chem. Eng. Bull. 6. 166 p.
Lewis, N. B., and J. H. Harding. 1963- Soil factors in relation to
pine growth in South Australia. Austr. For. 27:27~34.
Leyton, L., and K. A. Armson. 1955- The mineral composition of the
foliage in relation to the growth of Scots pine. For. Sci. 1:
210-218.
Leyton, L. 1958. Forest fertilization in Britain. J. For. 56:104-106.
Lindsay, W. L., and H. F. Stephenson. 1959a. Nature of the reactions
of monocalcium phosphate monohydrate in soils: 1. The solution
that reacts with the soil. Soil Sci. Soc. Amer. Proc. 23:12-18.
Lindsay, W. L., and H. F. Stephenson. 1959b. Nature of the reactions
of monocalcium phosphate monohydrate in soils: 2. Dissolution
and precipitation reactions involving iron, aluminum, manganese,
and calcium. Soil Sci. Soc. Amer. Proc. 23:18-22.
Lindsay, W. L., and E. C. Moreno. I960. Phosphate equilibria in soils.
Soil Sci. Soc. Amer. Proc. 24:177-182.
Lindsay, W. L., A. W. Frazier, and H. F. Stephenson. 1962. Identifi
cation of reaction products from phosphate fertilizers in soils.
Soil Sci. Soc. Amer. Proc. 26:446-452.
Mader, D. L. 1963- Soil variability: a serious problem in soil-site
studies in the Northeast. Soil Sci. Soc. Amer. Proc. 27:707~709.
Mader, D. L. 1973- Fertilizer needs and treatment responses for wood
fiber production: field assessment, p. 140-154. J_n Forest ferti
lization symposium proceedings. USDA For. Serv. General Tech. Rep.
NE-3.
v/


Table 45.
Regression equations of A] extracted by
HC1-H2S04 (Yl), Bray 1 (Y2), and Bray 2
(Y3) on A1 extracted by NH^OAc (X)
Regression equation
Correlation
coefficient
Yl = 1.5X + 59-3
r
0.803**
Y2 = 5.OX + 100.2
0.819**
Y3 = 5.5X + 180.3
0.718**
** Significant at the 1
% level .


29
anomalies could be accounted for by differences in the level of P
extracted by ]% citric acid. Baur (1959) reported that total-P values
were related to site productivity only within limited localities and not
over more heterogenous sites, a finding substantiated by Humphreys (1964).
Similarly, Ballard (1970 a) found that total P was significantly corre
lated with productivity of radiata pine only on groups of genetically
similar soils. When all soils were grouped together, the Olsen
(0.5M NaHC03, pH 8.5) and Bray 2 (0.03N. NH^F + 0.1N HC1) tests for avail-
able-P provided the best index of productivity and foliar-P concentrations.
In the southeastern USA, Pritchett and Llewellyn (1966) found total P to.
be of little value in delineating sites responsive to P fertilizer. Wells
(1965) also found total P to be inferior to available P for predicting
foliar-P concentrations of loblolly pine.
Extractable P vs site productivity. Despite the many suggestions
that soil-test methods derived for agricultural purposes are unlikely to
be satisfactory for use in forestry, reports of their successful use in
forestry are frequent in recent literature. Pawluk and Arneman (1961)
reported a significant correlation between Bray 1 extractable P
(O.O 3 N_ NH/jF + 0.02 5jN HCI) in the A2 horizon and the site index of 50-year-
old jack pine in Minnesota. On acid soils, such as used in this study,
the Bray 1 test extracts P predominantly from the Al-P and Fe-P fractions
(Thomas and Peaslee, 1973). On non-phreatic sandy soils in Wisconsin
\
ranging in pH from 4.3 to 7-0, Wilde et al. (1964 a and 1964 b) reported
a significant correlation between P extracted by the Truog soil test
.(0.002N^ H2S0^ + 3g (NHt)2S0t/l) and the site index of both Jack pine and
red pine. The Truog test extracts P from the Ca-P and to a lesser extent
the Al-P fractions (Thomas and Peas lee, 1973).


144
Table 35* Regression coefficients for multiple regression equations of
relative height on depth to LH, F^O-extractab1e P (0-20 cm)
and the squared terms for these two parameters
Source
Coefficients
S ignifi canee
R2
Height at age 1 year
Intercept
42.447
**
0.460**
Depth to LH
-0.056
NS
(Depth )to LH)2
-0.001
NS
h2o-p
7 Ml
**
(h2o-p)2
-0.351
NS
Height at age 3 years
Intercept
86.100
**
0.495**
Depth to LH
2.616
*
(Depth to LH)2
-0.018
NS
h2o-p
13-209
(10£ level)
h2o-p2
-0.144
NS
Height at age 5 years
Intercept
126.410
**
0.473**
Depth to LH
8.986
**
(Depth to LH)2
-0.069
**
h2o-p
2.845
NS
(h2o-p)2
0.969
NS
* Significant at the 5% level.
**Significant at the ]% level.


Table 51* Continued
Mean Foliar
Soi1-test
method
extract-
able P
Height
Rel. height
Fert
:. reqm.
P
jt,
r
3
5
1
3
5
1
3
5
4
ppm
,.r2
01 sen (3)
14.7
0.035
0.115
0.046
0.665
0.799
0.725
0.651
0.554
0.257
0.767
01 sen(4)
12.5
0.100
0.198
0.093
0.784
0.845
0.686
0.748
0.528
0.209
0.819
Olsen(5)
13.2
0.064
0.156
0.071
0.731
0.888
0.780
O.683
0.613
0.314
0.771
01 sen(6)
16.0
0.077
0.142
0.051
0.724
0.803
0.646
0.724
0.498
0.184
0.738
01 sen(7)
15.9
0.100
0.186
0.080
0.759
0.837
0.675
0.735
0.500
0.181
0.799
01 sen(8)
16.7
0.084
0.169
0.075
0.716
0.861
0.705
0.664
0.484
O.I83
0.781
01 sen(9)
20.8
0.062
0.156
0.070
0.704
0.852
0.738
0.651
0.517
0.217
0.805
Bray 2(2)
39.6
0.004
0.042
0.005
0.529
0.791
0.753
0.574
0.535
0.242
0.735
Bray 2(3)
48.7
0.008
0.054
0.010
0.554
0.826
0.766
0.578
0.525
0.229
0.748
Bray 2
41.6
0.000
0.023
0.001
0.458
0.765
0.779
0.504
0.541
0.274
0.713
Bray 2(4)
53.6
0.001
0.022
0.001
0.455
0.767
0.788
0.495
0.541
0.281
0.714
Bray 2(5)
47-3
0.005
0.008
0.001
0.383
0.709
0.725
0.429
0.470
0.226
0.638
Bray 2(6)
60.9
0.009
0.005
0.001
0.362
0.672
0.713
0.413
0.481
0.247
0.619
Bray 1(2)
34.4
0.008
0.061
0.014
0.560
0.820
0.772
0.578
0.527
0.233
0.771
Bray 1(3)
41.3
0.018
0.093
0.034
0.623
0.845
0.778
0.601
0.506
0.217
0.817
Bray 1
40.0
0.000
0.022
0.001
0.464
0.786
0.787
0.413
0.550
0.279
0.703
Bray 1(4)
47.1
0.001
0.047
0.010
0.436
0.817
0.810
0.557
0.557
0.277
0.776
Bray 1(5)
43.8
0.003
0.014
0.000
0.425
0.719
0.755
0.477
0.530
0.280
0.682
Bray 1(6)
55.8
0.007
0.007
0.000
0.376
0.672
0.716
0.421
0.487
0.254
0.630
Bray 3
35.8
0.000
0.023
0.001
0.458
0.776
0.795
0.504
0.555
0.288
0.716
Bray 3(2)
40.0
0.003
0.01 1
0.000
0.408
0.734
0.761
0.459
0.523
0.270
0.665
Bray 4
28.6
0.002
0.047
0.008
0.547
0.816
0.798
0.572
0.566
0.283
0.759
Bray 4(2)
37.4
0.000
0.024
0.002
0.459
0.742
0.771
0.497
0.556
0.304
0.694
Bray 5
49.8
0.004
0.014
0.000
0.418
0.734
0:781
0.467
0.552
0.304
0.686
Bray 5(2)
50.9
0.004
0.012
0.000
0.416
0.715
0.751
0.465
0.524
0.275
0.670


RELATIVE HEIGHT,
Fig. A. Relationship between HC1-h^SO^-extractable P(X) in the surface 20 cm of soil and relative
height(Y) of slash pine 1 year after P fertilization.


oL
4tf
Table 59- Concentrations of N, P, K, Ca, Mg. A1, and Fe in
foliage collected from 4-year-old. si ash pine grow
ing in the control plots of 24 field trials
Field Rep 1i-
s i te
cate
N
P
K
Ca
Mg
A1
Fe
A1
1
1.02
0.055
-%
0.29
0.18
0.104
ppm
281 60
2
0.90
0.054
0.24
0.13
0.100
300
58
3
0.95
0.055
0.28
0.15
0.115
281
65
A2
1
1.10
0.066
0.49
0.23
0.096
306
86
2
1.14
0.070
0.49
0.19
0.103
350
80
3
1.07
0.065
0.46
0.18
0.086
338
70
A3
1
1.00
0.086
0.50
0.21
0.109
306
55
2
0.91
0.081
0.50
0.18
0.125
263
45
3
0.9b
0.078
0.46
0.18
0.120
169
45
A b
1
0.97
0.082
0.49
0.19
0.086
363
50
2
0.87
0.082
0.43
0.16
0.119
263
38
3
1.02
0.082
0.56
0.21
0.094
475
58
A5
1
1.33
0.102
0.54
0.15
0.085
394
48
2
1 .22
0.100
0.59
0.13
0.076
363
38
3
1.26
0.106
0.54
0.15
0.077
425
46
A6
1
1.05
0.084
0.36
0.16
0.099
250
34
2
1.0b
0.088
0.46
0.17
0.084
319
38
3
1.15
0.090
0.48
0.21
0.092
325
44
A7
1
1.08
0.071
0.33
0.18
0.096
350
38
2
1.20
0.066
0.25
0.17
0.088
263
31
3
1.16
0.063
0.29
0.12
0.077
288
38
A8
1
1.21
0.105
0.45
0.19
0.075
581
40
2
1.19
0.106
0.48
0.17
0.089
644
25
3
1.15
0.115
0.52
0.18
0.074
538
45
A10
1
1.06
0.083
0.58
0.29
0.104
388
44
2
1.01
0.076
0.49
0.19
0.089
450
25
3
1.11
0.086
0.45
0.24
0.095
425
41
A15
1
1.16
0.100
0.36
0.14
0.081
570
45
2
1.16
0.107
0.31
0.14
0.101
533
74
3
1.07
0.098
0.36
0.13
0.071
513
50
A16
1
1.08
0.090
0.48
0.17
0.099
231
38
2
0.97
0.090
0.49
0.16
0.101
281
32
3
1.0b
0.089
0.44
0.18
0.101
275
38


45
Greenhouse Trial 1
This greenhouse trial was established in order to obtain growth and
P response information for use in the preliminary screening of the effec
tiveness of soil-test methods. This screening was necessary as it was not
practical to evaluate a large number of soil P-test using the soils from
all 72 sites. Only the most successful methods in the preliminary screen
ing were used in the final evaluation employing soils and growth informa
tion from all si tes.
Establishment
The greenhouse trial with the 10 bulk soil samples was established
in March 1972. Eight kilograms of air-dried, sieved soil were placed in
21-cm diameter, 7-6-1 iter glazed pots. Single drainage holes near the
base of each pot viere filled with porous fibre glass plugs. Sufficient
pots of each soil were prepared to give three replicates of four P rates.
One additional pot of each soil was prepared for use in determining root:
shoot ratios at the first harvest, as explained later. The P treatments
were applied to the surface of the potted soils as CSP to emulate field
applicationon a surface-area basis at 0, 56.3, 112.5, and 225 kg P/ha.
The first two and last application rates were identical to the P applica
tion rates, per unit surface area, as used in the field trials. To
insure P was the only nutrient limiting to growth, all pots received a
basal application of KCI (225 kg K/ha), micronutrient frit (140.6 kg FTE
503/ha), and NH^NO^ (225 kg N/ha). Both the KCI and FTE 503 were mixed
with the soil at potting, while the NH^NO^ was added as a dilute aqueous
solution in three equal applications, 1, 5,and 10 months after planting
of slash pine seeds. Phosphorus treatments were assigned at random to
the 10 extra pots. All P treatments were applied 1 month after planting
of seeds.


Financial assistance from the CRIFF program, the National Research
Advisory Council of New Zealand, and the Fu 1bright-Hays Foundation is
gratefully acknowledged.
Finally, I would like to thank the Forest Research Institute,
Rotorua, New Zealand for granting me leave of absence to pursue this
study.


72
This was true for data from this study, as statistical models which allow
for curvi1inearity (models 2, 3, and A in Table 9) provided better fits
between soil-test values and tree parameters than an untransformed linear
model (model 1 in Table 9). for the selected data used in this compari
son, most models using a logarithmic or an arctangent transformation of
the independent variable (soil-test value) and a quadratic model provided
better fits, as indicated by the R value, than the untransformed linear
model. There was little difference between the transformed models and
for convenience the logarithmic model was used for all computations in the
preliminary screening. The square of the multiple correlation coefficient
(R^) was used as the index of success of soil-test methods as predictors
of the tree parameters.
Relationships Between Soil-Test Values and Relative Height
Relative heights, the heights of unfertilized trees expressed as a
percent of heights of fertilized trees, after 1 year's growth in the
greenhouse (Table 7) were almost identical to those in the field after 1
year of growth (Table 8). This suggested a similarity in the modes of
nutrition of young seedlings in both the greenhouse and field. Responses
(relative heights) to added P, after 5 years of growth in the field, were
similar to those obtained after 3 years, but responses obtained at both
periods tended to differ from those obtained during shorter growth periods
in either field or greenhouse trials. This is in agreement with the
findings of Head and Pritchett (1971) who reported a poor relationship
existed between response in the greenhouse and that obtained after several
years in the field.
Response trends developed with time for the various soils can be
categorized fairly well in accordance with the relationship between P


3
manager to make the correct decisions regarding P fertilization practices
at time of planting. Specific objectives were as follows:
1. Evaluation of the effectiveness of a wide range of soil-P testing
techniques for predicting both growth and response of P fertilizer
applied to slash pine at time of planting.
2. Examination of the effects that soil and environmental parameters
have on the predictive accuracy of soil-P tests.
3. Calibration of the best soil-P test against the degree of response
obtained from fertilizer additions and the amount of fertilizer
required to produce this response.
k. Development of a predictive procedure for delineating sites where
phosphate sources of different solubility should be used.
5. Determination of the capacity of slash pine to utilize a range of
soil-P compounds, and relate this ability to the solubility of these
compounds in the various extractants used in soil-P tests.
*
NOTE
CRIFF annual report, 1973- (unpublished). Soil Science Department,
University of Florida, Gainesville, Florida.


Table 53- Analysis of variance of seedling dry
weights and P uptakes of greenhouse trial
2
Mean
square
Source
d.f.
Dry weight
P uptake
Total
44
T reatments
17
26.86**
489.88**
Soi1(S)
1
386.65**
3,150.24**
P compounds(P)
8
3.12
530.25**
S x P
8
5.62*
116.96**
Error
27
1.88
32.09
* Significant at
the 5%
1evel.
** Significant at
the 1%
1 eve 1.


65
development; an aliquot was taken down to dryness on a hot plate and
ignited in a furnace at 5Q0C for 30 minutes and the residue taken up in a
known amount of 0.1N_NC1 and an aliquot taken from these solutions for P
determination. Concentrations of other potentially interfering ions were
all below the interference level, as indicated by John (1970) or as
established in preliminary work.
Soil A1 and Fe Analysis
Soil A1 and Fe were extracted by several extractants. These
included CDB using the procedure of Mehra and Jackson (I960), 0.3M_
(NH^)2C20i, (Saunders, 1965), 0.1M Na/^Oy (McKeague, 1967), and 0.05M
EDTA (Viro, 1955)- Exchangeable A1 was also extracted by leaching with
1N_ KC1 (Yuan, 1965). Aluminum and Fe were also determined in the
extracts of several of the P-extraction methods outlined in Table 3-
These methods included NH^OAc, HCl-h^SO^, Bray 2, Bray 1, Bray 1(3). and
HC1.
Aluminum and Fe in all extracts except KC1 were determined by
atomic adsorption at the Soil Science Analytical Research Laboratory,
Universtiy of Florida (Yuan and Breland, 1969). Pyrophosphate extracts
were digested in HNO^-^SO^ and oxalate extracts ignited at 500C and
taken up in 0.1N_ HC1 before atomic absorption analysis. Aluminum in the
KC1 extracts was determined color¡metrically by the aluminon procedure
(Yuan and Fiskell, 1959).
Plant Tissue Analysis
One-gram samples of plant material were ashed at 480C for 5 hours.
The ash was dissolved in 6N_ HC1, digested on a hot plate for 15 minutes
and quantitatively transferred with distilled H20 to a 50-ml volumetric
flask. Suitable aliquots for elemental analysis were obtained from this


Table 49. Continued
Soi1-test
method
Soil (CRIFF identification code)
" A2 AB "aTI Az4 A28 A16 A9 23 A3 A25
Capacity (2)
Capacity(3)
210.0 20.0 44.0 178.0 146.0 0.0 0.0 3.5 22.5 50.0
296.0 57-5 119.0 246.0 236.0 0.0 0.0 42.0 55-5 123-0


165
Relative contribution of A] and Fe to P retention
The existence of closer relationships between P retention and ex
tractable A1 than between P retention and extractable Fe cannot be taken
to indicate that on a weight basis A] is more active than Fe in the sorp
tion of P, as has been suggested (Syers et al., 1971). Exchangeable A1
(KC1-extractable) and Fe (NH^OAc-extractable), amorphous Al (oxalate minus
KC1-extractable) and Fe (oxalate minus NH^OAc-extractable), and crystal
line Al and Fe (CDB minus oxalate-extractable) were used as independent
variables in a stepwise multiple regression analysis (Barr and Goodnight,
1972) with the retention of 2,500 pg P/g soil as the dependent variable
(Y,%). The correlation matrix for these variables is shown in Table Al.
Of all forms, amorphous Fe was the most closely correlated with P reten
tion, followed by amorphous Al. However, it is noteworthy that amorphous
Fe was significantly correlated at the 1% level with all other Fe and Al
forms. Regression equations from the stepwise regression analysis, in
which all partial regression coefficients were significant at least at
the 5% level, are shown in Table 42.
Partial regression coefficients in equation [3] relating P reten
tion to amorphous Al and Fe suggest that on a unit-weight basis, amor
phous Fe was more active in P retention than was amorphous Al. However,
this may have been due to the close relationship of amorphous Fe to other
active components such as exchangeable Al and Fe not included in the
equation. Such a possibility is supported by the reduction in the
partial regression coefficient for amorphous Fe upon inclusion of ex
changeable Al in equation [4] and complete loss from the equation as a
significant component when exchangeable Fe and crystalline Al were in
cluded as shown in equation 5]. Because of the close interrelationships


135
Table 30. Relationships between HCl-H^SO^-extractable P(0-20 cm) and P
fertilizer required to achieve 95% of maximum height on groups
of soils classed according to their amount of NH^OAc (pH 4.8)-
extractable A1
NH OAc(pH 4.8)
A1
Number
of soils
P ferti
lizer requirements (95%)
1 year
3 years
5 years
ppm
-
<20
22
0.021(81.50)+
o.189(13.49)
0.138(24.30)
20 to 80
34
0.3S0( 0.07)
0.297( 0.45)
0.043(51.37)
o
CO
A
16
0.609( 0.26)
0.533( 0.72)
0.256(14.43)
Whole A1 range
72
0.205( 0.06)
0.117( 1.36)
0.016(58.60)
^ Values in parentheses represent the significance level of the regression
in percent, using the model Y=aX + bX2 + c, where X = soil test-value.


Table 56. Continued
Field
s I te
Repli
cate
Depth
Particle
size distri
' bution
Moisture
content
Bu 1 k
density
C lay
Sil t
Sand
1/3 atm.
15 atm.
cm
'i,J
g/cc
0
A8
2
40-60
1 .6
3.6
94.8
1.8
0.6
1.70
3
0-20
1.8
4.3
93-9
3.2
1.8
1.39
20-40
2.3
4.7
93.0
2.8
1.2
1.69
40-60
1.8
4.6
93.6
2.6
1.0
1.69
A10
1
0-20
1.1
4.7
94.2
4.5
3.9
1.10
20-40
0.5
4.2
95-3
1.7
1.3
1.72
40-60
4.1
5.5
90.4
4.1
3.5
1.65
2
0-20
2.9
6.6
90.5
5.6
3.4
1.00
20-40
3.5
5.9
90.6
4.3
2.1
1.57
40-60
3.9
5.7
90.4
4.7
1.9
1.55
3
0-20
3.9
6.4
89.7
5.8
4.7
1.04
20-40
4.3
6.0
89.7
6.1
2.9
1.49
40-60
4.0
6.0
90.0
5.1
2.6
1.66
A15
1
0-20
2.7
5.7
91.6
4.0
1.6
1.20
20-40
3.0
4.3
92.7
2.8
1.1
1.64
40-60
2.8
4.5
92.7
3-0
0.8
1.65
2
0-20
1.9
5.6
92.5
4.3
3.4
1.28
20-40
3.2
3-9
92.9
3.1
1.0
1.58
. i:
1
40-60
2.8
4.6
92.6
2.6
0.9
1.62
3
0-20
2.0
4.7
93.3
4.7
2.7
1.21
20-40
3.5
3.1
93.4
2.0
1.0
1,54
40-60
2.0
4.5
93-5
2.2
0.7
1.64
ho


262
Conway, M. J. 1962. Aerial application of phosphate fertilizers to
radiata pine forests in New Zealand. Commw. For. Rev. 41:234-245.
Craig, F. G. 1972. Current fertilizer practice in Victorian state
forests, p. 389-392. In R. Boardman (ed.) Australian forest-tree
nutrition conference. Forestry and Timber Bureau, Canberra.
Oahl, E., Chr, Selmer-Anderssen, and R. Saether. 1961. Soil factors
and the growth of scots pine: A statistical re-interpretation of
data presented by Viro. Soil Sci. 92:36737l.
BeKent, J. A., and E. L. Stone. 1968. Influence of soil and site fac
tors on red pine plantations in New York. II. Soil type and physi
cal properties. Cornell Univ. Agrie. Expt. Sta. Bull. 1020. 25 p.
Deming, M. E., and W. E. Cate. 1963- Preparation of variscite. Soil
Sci. 95:206-208.
Dickson, D. A. 1971. The effect of form, rate, and position of phos-
phatic fertilizers on growth and nutrient uptake of Sitka spruce
on deep peat. Forestry 44:17-26.
Ellerbe, C. M., and G. E. Smith. 1963- Apparent influence of phosphate
marl on the site index of loblolly pine in the lower coastal plain
of South Carolina. J. For. 61:284-286.
Fife, C. V. 1959. An evaluation of ammonium fluoride as a selective
extractant for aluminum-bound soil phosphate: 11. Preliminary
studies on soils. Soil Sci. 87:83"88.
Fiskell, J. G. A., and H. F. Perkins. 1970. Selected Coastal Plain
soil properties. Southern Cooperative series Bull. 148. Published
by Fla. and Ga. Agr. Expt. Stations. 141 p.
Fiskell, J. G. A., and W. F. Spencer. 1964. Forms of phosphate in
Lakeland fine sand after six years of heavy phosphate and lime
applications. Soil Sci. 97:320-327.
Fowells, H. A., and R. W. Krauss. 1959. The inorganic nutrition of
loblolly pine and Virginia pine with special reference to nitrogen
and phosphorus. For. Sci. 5:95_112.
Fox, R. L., and E. J. Kamprath. 1970. Phosphate sorption isotherms for
evaluating the phosphate requirements of soils. Soil Sci. Soc.
Amer. Proc. 34:902-907.
Gentle, S. W., F. R. Humphreys, and M. Lambert. 1965- An examination
of Pinus radiata fertilizer trial 15 years after treatment. For.
Sci. 11:315-324.
Gentle, S. W. and F. R. Humphreys. 1968. Experience with phosphatic
fertilizers in man-made forests of Pinus radiata in New South Wales.
9th Commonwealth Forestry Conference, India. 36 p.


SOIL TESTING AS A GUIDE TO PHOSPHORUS FERTILIZATION
OF SLASH PINE (Pinus elliottii var. elliottii Engelm.)
By
Russell Ballard
A Dissertation Presented to the Graduate
Council of the University of Florida
in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
UNIVERSITY OF FLORIDA
1974


(Helena sandy loam) in South Carolina, found P extracted by the Truog
test accounted for only 17% of the variability of foliar P in loblolly
pine. Phosphorus extracted by the Olsen and Bray 2 tests was found to be
significantly correlated with foliar P of 40-year radiata pine (Ballard,
1970 a). In an intensive study involving 63 plots in Florida and 60 in
Australia, water-extractable P was found to provide the single best index
of foliar-P concentrations in slash pine (Humphreys and Pritchett, 1972).
However, the inclusion of NaOH-extractable P in their prediction equation
significantly improved the estimate of foliar P. The authors suggested
that both the intensity and quantity factors of soil-P supply are of im
portance in determining foliar-P concentrations. Alban (1972) also
reported that extractants which removed relatively small amounts of P from
the soil (H2O, 0.002f[ H^SO^, 0.01N_HCI) provided a better index of foliar-
P concentrations of 49-to 94-year-old red pine. Although Alban also
examined stronger extractants and found them less effective than the above
weaker extractants, he did not use multiple regression techniques to
examine the possibility that the combined use of a weak and strong
extractant may have improved the prediction, as did Humphreys and
Pritchett (1972). Using a greenhouse trial, Baker and Brendemuehl (1972)
found P extracted by NH^OAc (pH 4.8) correlated significantly with both
foliar-P concentrations and dry weights of sand pine (P. clausa Vasey)
seedlings growing on a Lakeland fine sand. This weak acid extractant
removes small quantities of P from the soil by forming weak complexes
with polyvalent cations (Al, Fe, Ca) in soils (Thomas and Peaslee, 1973)-
Pritchett and Llewellyn (1966) also found NH^OAc (pH 4.8) to extract P
that provided a good index of the growth of 3-to 5-year-old slash pine in
the absence of fertilizer. The trees were growing on acid sandy soils in
the lower Coastal Plain which varied in drainage characteristics from


Table 40. Extractable A1 and Fe values and their correlation with P retention
,
P
retent ion
Extractant
pH
Element
Mean
Range
Langmuir
maximum
Saturation
maximum
ppm r
1JN KC1
7.0
A1
47.7
1.5-200.0
0.745**
0.705**
1 N_ NHjjOAc
4.8
A1
68.7
8-222
0.937**
0.934**
Fe
13-9
0-155
0.659**
0.651**
A1
+ Fe
-
-
0.932**
0.906**
0.3M (NHlf)2C20if
3.2
A1
450.0
25-1250
0.881**
0.890**
Fe
288.5
55-1560
0.898**
0.903**
A1
+ Fe
-

0.94]**
0.949**
CDBt
8.2
A1
526.1
50-1680
0.887**
0.881**
Fe
837.2
39-4360
0.784**
O.709**
A1
+ Fe
-
-
0.856**
0.856**
0. 1M Na^P^
10.0
Al
Fe
1926.2
703.8
100-11100
40-5220
0.881**
0.830**
0.889**
0.835**
A1
+ Fe
-
-
0.879**
0.882**
0.05M EDTA
9.0
Al
398.9
25-1000
0.901**
0.894**
Fe
144.9
40-800
0.769**
0.798**
A1
+ Fe
-
-
0.942**
0.933**
0. 1 N_ HC1
1.1
Al
197.9
25-435
0.865**
0.849**
Fe
37.8
9-220
0.641**
0.593**
A1
+ Fe
-
-
0.904**
0.887**
** Significant at the ]% level.
1 Citrate-dithionate-bicarbonate (Mehra and Jackson, I960).


130
by extractable P in the top 20 cm of soil. Some of the possible factors
are examined below.
The level of P extracted by the five soil-test methods tended to
be equally effective at predicting amounts of P fertilizer required to
achieve either 90 or 95% of maximum height growth (Table 27). There was
a pronounced decline in their effectiveness at predicting fertilizer
required to achieve 100% of maximum height growth, particularly at ages
1 and 3 years. This was probably a function of the method of determining
P requirements, rather than a true deterioration in the value of soil P
for predicting P-fertilizer requirements. As mentioned in the Materials
and Methods section, where response curves could not be fitted to data
from any site, fertilizer requirements were taken as the actual field ap
plication rate which provided the greatest height growth. At several
sites, small differences of less than 5% (relative height > 95%) were
observed between the control and highest P treatment; these were
recorded as requiring this amount of P fertilizer to achieve 100% of
maximum grov/th. Many of these small differences were probably nonsignif
icant and unrelated to soil-P levels.
The relationships between P extracted by the HCl-h^SO^ method
(0-20 cm bed samples) and the amount of P fertilizer required to achieve
90 and 95% of maximum height growth are shown in Figs. 9 and 10,
respectively. The critical value of 5 ppm extractable P, proposed from
the relationships with relative height, also provided a reasonable
distinction between sites requiring a practical P-fertilizer application
(>10-20 kg P/ha) and those which did not. The prediction curves
for ages 1 and 3 years were somewhat similar, but the fertilizer
requirements over the 5-year period were considerably lower for any soil-
test value below 5 ppm, than the requirements over the 1- and 3~year


Table 16. Solubility of P compounds In chemical extractants following 2 months' incubation in two soils
Sol 1-test method
Sol 1
h2o
NH/jOAc
T ruog
HCl-h^SO^ Olsen Lactate
Bray 1
Bray 1
NH^F
%
Monocalcium phosphate(MCP)
Immokalee
79.9
72.6
89.0
81.0
77.8
87.2
92.1
84.0
83-5
McLaurin
1.1
8.6
29.2
37.9
41.2
47.2
66.1
74.4
67.2
Dicalcium phosphate(DCP)
82.5
82.0
1mmoka1ee
72.9
71.7
84.8
81.0
77.2
78.7
82.1
McLaurin
1.6
8.4
29-5
39-5
39-0
29-9
83.O
86.6
69-2
FIuorapatite(FA)
48.3
1mmoka1ee
8.9
10.4
21.4
79-7
10.4
34.0
11.7
13.1
McLaurin
0.1
0.1
10.4
47.2
0.4
8.4
0.2
25.0
0.2
Colloidal aluminum phosphat
e(CAlP)
/
89-8
!mmokalee
60.4
54.6
69-7
78.5
6778
83.8
85-5
85.O
McLaurin
0.6
9.0
18.7
37.6
35.8
43.0
63-7
71.4
63.4
Potassium taranaki te(KTK)
89.1
80.8
86.0
1mmoka1ee
6.8
7.5
8.4
12.8
15.8
9-8
McLaur¡n
1.1
9-0
15.0
37.2
36.9
42.3
67-5
75.9
68.8
Wavel1 i te(WA)
1mmoka1ee
0. I
0.6
0.3
0.6
0.4
0.8
4.9
5.3
1.9
McLaurin
0.1
0.1
0.1
0. 1
0. 1
0.1
1.5
1.9
0.3
Colloidal
ferric phosphate(CFeP)
85.0
Immokalee
49.1
45.2
60.8
68.2
70.2
83.8
87.O
79.6
McLaurin
1.4
6.5
10.6
37.2
36.9
29.4
62.8
75.9
53.8
St rengi te(STR)
0.1
Immokalee
0.1
0.1
0.1
0.3
0.1
,0.1
0.1
0.1
McLaurin
0.1
0.1
0.1
0. 1
0.1
0.1
0.1
0.1
0. 1


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
/ W¡ 1
Professor (Forest Soils)
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
John G. A. Fiskel1
Professor (Soil Biochemist)
I certify that 1 have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Frank G. Martin
Associate Professor (Statistician)
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Assistant Professor (Forest
Physiologist)


Table 61. Amounts of A1 and Fe extracted from 42 Coastal Plain forest soils by six chemical extractants
>V " j 1 1 "
Field IN KC1 0.3M (NH^C^ CDB O.IMNa^O 0.5H EDTA 0. ItL HC1
Sl te A1 ~A Fe Al Fe 1 Fe A1 Fe A1 Fe
A5b
14
275
230
350
405
1 ,000
340
275
83
136
21
A5
18
200
220
400
395
950
310
300
90
147
20
A15
43
650
300
650
490
2,000
450
575
125
255
29
A25b
27
1,125
440
875
710
2,900
570
825
148
361
25
A25
38
1,250
460
850
735
4,000
770
800
155
330
25
A7
84
500
190
575
235
1,100
240
450
140
237
48
A10
108
750
320
850
435
1 ,400
390
800
275
385
67
A23
87
650
280
725
365
1,500.
350
625
200
347
56
A16b
14
25
60
75
45
200
45
63
43
45
13
A16
32
50
70
120
60
150
60
100
55
67
15
A3
21
50
60
125
48
200
70
75
40
50
14
A4
19
50
70
100
43
'200
70
100
50
50
15
A6
33
100
90
175
77
250
90
150
70
87
16
A18
63
100
90
175
113
200
260
175
105
115
30
AI9b
20
50
55
125
65
150
50
75
48
52
16
A19
28
100
70
100
55
100
80
75
40
57
16
A21
17
100
80
125
68
200
70
100
75
62
23
A22
21
100
60
100
57
150
50
100
50
72
16
A23b
67
275
135
400
158
500
145
313
128
207
35
FI
41
500
140
525
100
750
no
550
70
352
14
F2
16
25
60
175
39
200
40
100
45
80
10
F3
2
50
70
100
48
150
50
50
40
40
9
F4
1 1
100
60
100
45
200
50
75
50
57
10
F5
3
50
100
50
135
150
100
75
70
42
22
F6
3
150
260
200
227
300
. 280
150
175
110
70
F7
2
25
70
50
68
100
60
25
55
25
18


Another point of practical importance, illustrated by the results
from the greenhouse trial with P compounds, concerns the use of RP in
forestry. The use of RP is now recommended on certain sites in the
Coastal Plain region (CRIFF annual report, 1373) where either leaching or
excess retention of soluble forms of P fertilizers are likely to occur.
In view of the critical role of soil pH in governing RP solubilization,
and hence its availability to pine trees, this source should not be rec
ommended on soils with a pH above 5*


161
Extractable Al and Fe
Amounts of Al and Fe removed by a range of extractants and their
correlation with the two measurements of P retention are shown in Table
*40. Specific values of A1 and Fe removed by selected extractants are
presented in Appendix Table 6l. Despite some variation in the size of
the correlation coefficients, the relationships between P retention and
levels of A1 and Fe extracted by all reagents were significant at the 1%
level and were independent of the method of measuring P retention. The
results agree with the findings of many investigators (Yuan and Breland,
1969; Udo and Uzu, 1972; Syers et al., 1971) that extractable soil A1
provides the best single index of soil-P retention over a range of soils.
However, this does not imply that Al is more active than Fe in the ad
sorption of P or that extractable Fe may not be more closely correlated
with P retention within certain soil groups, as was found by Yuan and
Breland (1969).
Several investigators have reported that exchangeable Al (KC1 -
extractable) provided the best index of soil-P retention (Coleman et al.,
I960; Syers et al., 1971; Udo and Uzu, 1972) but in this study it proved
inferior to all other forms. This may be attributed to the small quanti
ties of exchangeable Al found in soils with a pH > 5-0 but which fre
quently had large amounts of amorphous Al (oxalate extractable). Of all
extractants of Al NH^OAc (pH *1.8), which on the average extracted only
slightly more Al than KC1, mainly from the less acid soils, provided the
best index of P retention. This reagent apparently removed Al from the
soil in proportion to the amount of active forms present, as indicated
by its correlations with Al extracted by other reagents ( r > 0.80 in all
cases except for KCl). Correlation coefficients between P retention and


INTRODUCTION
Fertilization is one of many cultural practices available for in
creasing the productivity of forest lands. Increasing demand for wood
products, associated with a decreasing land base for commercial timber
production, has led to an upsurge of interest in intensive forest manage
ment techniques in the last decade.
In the southeastern Coastal Plain, intensive forest cultural prac-
9
tices such as clear-cut harvesting, site preparation, use of improved ge
netic stock, and fertilization are now widely used. Climatic conditions
/
which favor long growing seasons and high biological activity, the adop-
tion of intensive management techniques which increase growth rates, and
the inherently'infertile acid sandy soils, have been the major factors
contributing to the successful results obtained from the use of fertilizer
in southeastern Coastal Plain forests. Results from more than 225 field
experiments established throughout the Coastal Plain by the Cooperative
Research in Forest Fertilization (CRIFF) program have shown that about
two-thirds of the tested sites were responsive to nitrogen and/or phos
phate fertilizers (CRIFF annual report, 1973). Trials in young planta
tions indicated that about two-thirds of the sites were responsive to
phosphate fertilizers and three-fourths responsive to nitrogen and/or
phosphate (Pritchett and Smith, 1972).
Approximately 770,000 acres of pines are planted annually in the
Southeast, of which it has been projected 500,000 acres would benefit from
fertilizer additions (CRIFF annual report, 1973)- Considerably less than
1


Table 9* Comparison of the goodness of fit of
of the multiple correlation coefficient
four statistical
(R2),
models, as indicated by the square
relating selected tree parameters (dependent
vari able)
and soil
-test values
(i nde-pendent
variable)
Soi1-test
method
Mode 1^
Greenhouse (l
year)
Field (3 years)
Fol iar P
(4 years)
Height
Rel.height
Fert.reqm.
Height
Rel. height
Fert.reqm.
H0
1
0.288
0.741**
0.601**
0.545*
O.O67
0.059
0.056
L
2
0.411*
0.951**
0.897**
0.421*
0.217
0.087
0.228
3
0.489*
0.947**
0.951**
0.410*
0.257
0.086
0.265
4
0.676**
0.935**
0.949**
0.548*
0.285
0.061
0.338
T ruog
1
0.358
0.355
0.339
0.036
0.421*
0.286
0.346
2
0.342
0.658**
0.686**
0.113
0.658**
0.413*
0.496*
3
0.203
0.702**
0.803**
0.132
0.645**
0.350
0.424*
4
0.386
0.626*
0.653**
0.096
0.616*
0.433
0.461*
HC1-H-S0,,
1
0.282
0.280
0.265
0.040
0.425*
0.287
0.392
2
0.262
0.579*
0.607**
0.151
0.783**
0.494*
0.706**
3
0.193
0.689**
0.764**
0.218
0.734**
0.386
0.642**
4
0.292
0.413
0.450*
0.053
0.616*
0.510*
0.615*
Bray 1
1
0.251
0.205
0.211
0.007
0.434*
0.289
0.386
2
0.094
0.241
0.279
0.022
0.785**
0.551*
0.704**
3
0.034
0.355
0.419*
0.064
0.908**
0.640**
0.712**
4
0.282
0.207
0.228
0.014
0.642**
0.446*
0.575*
* Significant at the 5% level.
** Significant at the 1% level.
+ Model 1, Y = bX + c; Model 2, Y = b logX + c; Model 3, Y = b arctan X + c; Model 4, Y = aX + bXz +


HEIGHT AT AGE 5 YEARS,
Fig. 11. Relationship between depth to limiting horizon(X) and the height
of slash pine after 5 years' growth.


270
Strand, R. F., and R. E. Miller. 1969- Douglas-fir can be increased
report from Pacific Northwest shows. Forest Ind. 96:28-30.
Swan, H. S. D. 1965. Reviewing the scientific use of fertilizers in
forestry. J. For. 63:501-508.
Swan, H. S. D. 1969. Fertilizers, their role in reforestation. Pulp
and Pap. Res. Inst. Can. Woodlands Pap. 9- lip.
Syers, J. K., T. D. Evans, J. D. H. Williams, and J. T. Murdock. 1971-
Phosphate sorption parameters of representative soils from Rio
Grande do Sul, Brazil. Soil Sci. 112:267-275
Tamm, C. 0. 1964. Determination of the nutrient requirements of forest
stands. Int. Rev. For. Res. 1:115-170.
Tamm, C. 0. 1968. The evolution of forest fertilization in European
silviculture, p. 242-247- J_n Forest fertilization theory and
practice. Tennessee Valley Authority, Muscle Shoals, Ala.
Taylor, A. W,, E. L. Gurney, and W. L. Lindsay. I960. An evaluation of
some iron and aluminum phosphates as sources of phosphate for
plants. Soil Sci. 90:25~31 *
Taylor, A. W., E. L. Gurney, and J. R. Lehr. 1963- Decay of phosphate
fertilizer reaction products in an acid soil. Soil Sci. Soc. Amer.
Proc. 27:145-148.
Terman, G. L. 1968. Fertilizer, soil, and plant properties affecting
crop response to P fertilizers, p. 7780. _l_n Forest fertilization
theory and practice. Tennessee Valley Authority, Muscle Shoals, Ala.
Terman, G. L., and G. W. Bengtson. 1973- Yield-nutrient concentration
relationships in slash and loblolly pine seedlings. Soil Sci. Soc.
Amer. Proc. 37:445-450.
Thomas, G. W., and D. E. Peaslee. 1973- Testing soils for phosphorus,
p. 115-132. Jjn L. M. Walsh and J. D. Beaton (ed.) Soil testing
and plant analysis. Soil Sci. Soc. Amer. Madison, Wis.
Thompson, E. F. i960. Relation of soil and foliar nutrient levels to
growth of loblolly pine on different sites. Thesis. North Carolina
State Univ. Library, Raleigh.
Udo, E. J., and F. 0. Uzu. 1972. Characteristics of phosphorus adsorp-
> tion by some Nigerian soils. Soil Sci. Soc. Amer. Proc. 36:879-883.
Viro, P. J. 1955- The use of EDTA in soil analysis. 11. Determination
of soil fertility. Soil Sci. 80:69"74.
Voigt, G. K. 1958. Plant and soil factors in chemical soil analysis,
p. 31-41. J_n T. D. Stevens and R. L. Cook (ed.) First North Ameri
can forest soils conference. Michigan State Univ., E. Lansing,
Mich.


Table 50. Continued
Mean
Soi1-test
method
extract-
able P
Height
Rel. height
Fert.
reqm.
%P in
tops
P uptake
1*
2
1
2
1
2
1
2
1
2
ppm
Ci trate(2)
11.3
0.357
0.651
0.708
0.494
0.752
0.662
0.862
0.594
0.623
0.864
Oxalate
>
69-9
0.092
0.311
0.040
0.036
0.070
0.098
0.373
0.554
0.065
0.672
T ruog
8.9
0.342
0.562
0.658
0.440
0.685
0.607
0.821
0.599
0.634
0.828
Truog(2)
6.2
0.378
0.726
0.785
0.729
0.787
0.824
0.904
0.574
0.701
0.768
T ruog(3)
3.0
0.280
0.637
0.806
0.914
0.834
0.955
0.824
0.385
0.650
0.477
T ruog(4)
I8.7
0.191
0.506
0.418
0.330
0.421
0.495
0.799
0.801
0.370
0.889
T ruog(5)
15.0
0.251
0.610
0.537
0.483
0.570
0.647
0.905
0.769
0.484
0.890
Truog(6)
11.1
0.278
0.695
0.636
0.674
0.673
0.808
0.964
0.701
0.554
0.821
Truog(7)
32.2
0.167
0.495
0.326
0.308
0.339
0.440
0.765
0.843
0.307
0.867
Truog (8)
33.0
0.137
0.451
0.296
0.260
0.322
0.415
0.753
0.832
0.268
0.857
Iruog(9)
31.3
0.145
0.479
0.332
0.323
0.358
0.480
0.804
0.854
0.298
0.856
Truog(10)
18.1
0.145
0.206
0.162
0.030
0.132
O.O89
0.372
0.593
0.158
0.715
Truog (1 1)
12.9
0.110
0.183
0.123
0.032
0.099
0.098
0.363
0.585
0.134
0.672
T ruog(12)
9-6
0.127
0.273
0.232
0.116
0.190
0.222
0.526
0.720
0.21 1
0.767
H2SO4
7.3
0.324
0.598
0.774
0.570
0.804
0.744
0.8-8
0.551
0.671
0.759
H2S04(2)
4.9
0.336
0.695
0.826
0.790
0.852
0.893
0.895
0.499
0.703
0.662
H2so((3)
2.3
0.226
0.537
0.797
0.838
0.847
0.912
0.727
0.279
0.614
0.366
H2S0/,(4)
19.1
0.299
0.635
0.515
0.451
0.542
O.58I
0.861
0.769
0.497
0-922
H2SO/4(5)
15.4
0.277
0.649
0.549
0.520
0.582
0.666
0.916
0.766
0.503
0.897
H2S0/, (6)
9*5
0.253
0.692
0.651
0.741
0.682
0.865
0.943
0.637
0.508
0.754
h2so^(7)
36.1
0.171
0.515
0.318
0.321
0.351
0.461
0.787
0.839
0.308
0.873
H2S04(8)
34.7
0.162
0.507
/0.34o
0.323
0.370
0.477
0.801
0.836
0.311
0.874
h2so*(9) '
32.5
0.159
0.503
0.349
0.331
0.377
0.488
0.811
0.846
0.314
0.873
ro
0
HC1-H2S04
15.2
0.262
0.624
0.579
0.538
0.606
0.697
0.925
0.753
0.516
0.871
V_aJ


Table 54. Continued
Field'"'
s i te
Repli-
cate
So i 1
Type
Order
Drainage
V *v
c lass
Depth
to LI!
Depth
of A!
Organic
matter
Total
N
CEC

-cm
meq/1OOg
A27
1
McLaurin fsl
U1tisol
4
90
20
2.07
0.096
5-96
2
McLaurin fsl
U1tisol
4
90
20
2.07
0.085
6.52
3
McLaurin fsl
U11i so 1
4
90
20
2.60
0.094
7.64
A28
1
McLaurin fsl
U11i so 1
4
90
25
-rr
00
0.075
5-63
2
McLaurin fsl
U111 sol
4
90
23
1 .51
0.076
4.72
3
McLaurin fsl
U11i so 1
4
90
23
1.45
0.064
4.36
A29
1
Kalmia fsl
Ultisol
4
70
20
2.92
0.103
7.87
2
Kalmia fsl
Ultisol
4
90
18
1.68
0.066
4.11
3
Kalmia fsl
Ultisol
4
90
15
1.88
0.069
4.54
* CRIFF identification code.
** Drainage class: (l) poorly; (2) somewhat poorly; (3) moderately well; (4) well; and (5) some
what excessively drained.
t Depth to limiting horizon (mottled, spodic, argillic). If no limiting horizon occurred within the
surface 90 cm, a value of 90 was assigned.


Table 54. Continued
Field Repli-
site cate
SoM
Type Order
Drainage
*wV
class
A8
1
Blanton fs
U11i so 1
3
2
Blanton fs
U11i so 1
3
3
Blanton fs
U1tisol
3
A10
.1
Leon fs
Spodosol
2
2
Rut lege fs
1nceptisol
1
3
Rutlege fs
1nceptisol
1
A15
1
Chipley fs
Entisol
3
2
Chipley fs
Entisol
3
3
Chipley fs
Entisol
3
A16
1
Immokalee fs
Spodosol
2
2
1mmoka1ee fs
Spodosol
2
3
Immokalee fs
Spodosol
2
A17
1
Plummer Ifs
U1tisol
1
2
Plummer Ifs
Ultisol
I
3
Plummer Ifs
U1tisol
1
A18
1
Leon fs
Spodosol
2
2
Leon fs
Spodosol
2
3
Leon fs
Spodosol
2
A19
1
Leon fs
Spodosol
2
2
Leon fs
Spodosol
2
3
Leon fs
Spodosol
2
Depth
to LH
90
90
90
40
23
33
90
90
90
68
53
65
45
33
4o
50
48
53
60
53
50
Depth
of A1
Organic
matter
Total
N
CEC
1
%-
meq/1OOg
20
1 .-68
0.055
6.60
20
1 .44
0.046
4.16
20
1.07
0.034
2.92
20
2.41
0.066
5-99
20
1.98
0.116
4.92
20
2.75
0.112
6.24
23
1.54
0.050
3-09
23
1.85
0.060
4.97
23
1.88
0.062
5.05
25
4.08
0.051
7.64
25
3.06
0.066
6.14
20
2.27
0.061
4.62
20
7-97
0.215
16.95
20
11.99
0.320
27-44
23
9-03
0.265
20.33
18
4.54
0.123
14.19
18
1.74
0.046
3-98
18
2.92
0.076
7.72
23
2.34
0.078
6.42
23
2.99
0.075
6.83
23
5.15
0.072
6.50


172
Foliar analysis, using current sampling procedures, appears to
have little merit as a means of predicting soil-P retention. Although
foliar A1 accounted for a significant proportion of the variation in P
retention when sites of low P status were excluded, such a relationship
can be of little practical use, as it is these low P status sites that
one is concerned with in forest fertilization. It is questionable whether
a good predictive equation could be developed using tissues other than
those currently sampled, particularly if the poor relationship on sites
of low-P status reflects either direct or indirect effects of insufficient
P uptake on the intake and translocation of A1 within the plant. Humph
reys and Truman (1964) found that in three pine species, increasing a-
mounts of A1 in the substrate increased both A1 and P in roots and shoots
provided that the substrate contained adequate P, and Haas (1936) re
ported a similar phenomenon in citrus. These reports, in conjunction with
the findings reported above that increasing foliar A1 appeared to be as
sociated with increasing foliar P on sites of low P status, suggest that
A1 and P uptake and/or translocation are linked somehow at low levels of
either element.
Calibration of Extractable A1 Against Field-P Retention
Relationships between soil properties and P retention determined
by addition of P solutions to soils in the laboratory cannot be used di
rectly to predict retention of phosphate fertilizers applied in the field.
This is because the nature of the reaction between P and soil in the re
gion of the fertilizer particles is likely to be somewhat different from
that between soil and P solutions used in the laboratory. Laboratory
reactions, however, provide clues about properties most likely to cor
relate with field retention.


85
The success of soil-P tests for predicting response to P fertili
zer, despite the apparent limiting effect on growth of site factors other
than soil-P levels, can probably be attributed to the use of relative
height as the measure of response. By expressing height response to P
fertilizer at each site as a function of height in the absence of P fer
tilizer, the influence of site factors other than soil P among sites is
effectively reduced.
General Discussion
Research results indicate that P-intensity measurements (weak ex
tractants) correlated better with yield and response of annual crops,
which depend upon rapid uptake in the early stages of growth, than do P-
quantity measurements (strong extractants). However,P-quantity measure
ments correlate better with nutrient concentrations and uptake data,
which depend upon absorption over the entire growth period (Williams and
Knight, 1983). Data for the first year's tree growth in the greenhouse
test concur with these findings. The trend for quantity measurements to
become better predictors of growth and response to added P, when growth
periods of greater than 12 months were considered, can probably be re
lated to the efficiency of internal P recycling within plants in addition
to the longer absorption periods involved. Growth and response in years
subsequent to the first year will be determined not only by current uptake
from the soil but also reserves of P accumulated in tissues from the pre
vious year(s) uptake. This latter (P uptake) is determined by the quan
tity factor of soil-P supply (Williams and Knight, 1963).
Clear distinction between the value of intensity and quantity mea
surements of soil-P supply for predicting growth, response to P fertili
zer, and nutrient uptake will only be apparent if there is no relationship


150
values (range) obtained from the close relationships shown in Table 37
and Fig. 12 can be used to confirm critical soil-test values proposed
from relationships between relative height and soil-test values.
Relationships Between Foliar P and Soil-Test Values
Multiple correlation coefficients (R^) for relationships between
foliar P and soil-test values illustrated the same trends reported in
the preliminary screening (Table 38). Soil-test methods which extracted
relatively large amounts of P from the soil (HCl-h^SO^, Bray 1(3), Bray
2) were more closely related to foliar P of 4-year-old trees than soil-
test methods which extracted smaller quantities of P from the soil
(H2O, NH^OAc). The importance of soil P at rooting depths below the
surface 20 cm. was again illustrated by the larger R^ values obtained
using P extracted from the soil effective rooting depth.
The relationship between foliar P and HC1-HjSO^-extractab1e P in
the surface 20 cm (bed) of soil is illustrated in Fig. 13. The
HCl-H^SO^-P value corresponding to the foliar-P concentration of 0.085%
was 4.5 ppm, which is in good agreement with the value of 5 ppm proposed
earlier as useful for separating P responsive and nonresponsive sites.
From the equation relating foliar P (Y) to Bray 1(3)-extractable P (X),
Y = 0.0189 logX + 0.0684
the Bray 1(3)-extractab1e P value corresponding to 0.085% P in the
foliage was calculated to be 7-5 ppm, while the corresponding value
for Bray 2 was calculated to be 9-8 ppm from the following equation:
Y = 0.0231 logX + 0.0620.
General Discussion
Data from the calibration study indicated that for the purpose of
predicting P-responsive sites and the P-fertilizer requirements for the


P reten t ion jug/gsoil
2000
160G
1200
800
400
:
EZ3
Ed
Ed
a
n
Lmi
L_i

Cl

Cl

a
a
ci
n
a


ci
r1
.P--' A,
On a
f.s.
Cl p
Cl bIl
CTTl
n
Cl
d
Cl

-
ci
CJ
Cl
Cl
Ed
CJ
Cl
Cl
td
Ed
n
__
Cl
Cl
0
Cl
Cl
r;i
Cl
Cl
Cl

Cl
Cl
11
Cl
Ed
Cl
Cl
; j
Ed
O
Cl
Cl
1
Cl
Cl
Cl
td
-1
d
1
r ~1
St.Johns Myakka Pome! !o Leon Leon
"f.S. f.s. f.s. f.S. f.S-
Fig. 17. Phosphorus retention as determined from addition of 2,500 yg P/g
soil in A and Bh horizons of six Spodosols.


P retention,
NH4OAC (pH 4.8) extractable Al, ppm
Fig. 15. Relationship between % P retention from three P solutions (100, 300 and 2,500 ug P/g soil) and
NH^OAc (pH 4.8)-extractable Al. For clarity, the points for only 15 soils are shown but the
linear correlation coefficients (r) were computed using data of all 42 soils.


SOIL TESTING AS A GUIDE TO PHOSPHORUS FERTILIZATION
OF SLASH PINE (Pinus elliottii var. elliottii Engelm.)
By
Russell Ballard
A Dissertation Presented to the Graduate
Council of the University of Florida
in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
UNIVERSITY OF FLORIDA
1974

DEDICATION
To
Phi 11ipa

ACKNOWLEDGMENTS
I would like to thank Dr. W. L. Pritchett, the chairman of my
supervisory committee for his valuable guidance and counsel during the
tenure of this study. The interest and assistance given by the other
members of my committee, Dr. J.G.A. Fskell, Dr. F.G. Martin, Dr. W.H.
Smith and Dr. C.A. Hollis are very much appreciated.
I would like to acknowledge the encouragement given by Dr. D.F.
Eno, chairman of the Soil Science Department and other members of the
faculty. In particular I would like to thank Dr. L.W. Zelazny and Dr.
J.G.A. Fiskell for their contribution to my academic enrichment and Dr.
D.F. Rothwell for his administrative work on my behalf.
Thanks are also due to the CRIFF laboratory personnel, in par
ticular to Mary McLeod for her help in the laboratory phase of the work,
and to Dr. H.L. Breland and the staff of the Soils Analytical laboratory
and to Dr. J. NeSmith and the staff of the Soil Testing laboratory for
their expert analysis of numerous samples.
I have enjoyed the companionship of the graduate students of the
Soil Science Department who provided necessary light relief and much
intellectual stimulation. Specifically I would like to thank my fellow
graduate students in the CRIFF program, Terry Sarigumba and Roy Voss,
who willingly helped with some of the heavy work.
My most sincere thanks to my wife Pip, who has been actively in
volved in all phases of this work from initial collection of samples
through to the final typing.
i i i
i

Financial assistance from the CRIFF program, the National Research
Advisory Council of New Zealand, and the Fu 1bright-Hays Foundation is
gratefully acknowledged.
Finally, I would like to thank the Forest Research Institute,
Rotorua, New Zealand for granting me leave of absence to pursue this
study.

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ¡¡ i
LIST OF TABLES vi
LIST OF FIGURES xlv
ABSTRACT xvi
INTRODUCTION 1
LITERATURE REVIEW 4
Forest Fertilization 4
Historical 4
N and P Fertilization 5
P Sources Used in Forestry 9
Reactions in Soils 9
P Effectiveness in Forestry 15
/Diagnostic Methods 20
i Foliar Analysis 20
Soil Analysis 24
? Sampling procedures 25
Extraction methods 28
Interpreting soil test results 33
u Prediction of productivity 33
Prediction of fertilizer response 34
MATERIALS AND METHODS 39
Introduction 39
FieldTrials 40
So i 1 Sampl ing 40
Foliage Sampling 42
Growth and Response Parameters 42
Greenhouse Trial 1 45
Establishment 45
Harvesting 46
Growth and Response Parameters 47
Greenhouse Trial 2 48
P Compounds 48
Establishment 49
v

TABLE OF CONTENTS (Continued)
Page
Harvesting 51
Extraction of P Compounds 51
Phosphorus-Retention Study 51*
Soil and Foliage Samples 5**
Determination of P Retention 5^
Sample Analysis 57
Soil Characterization 57
Soil P Analys is 57
Soil A1 and Fe Analysis 65
Plant Tissue Analysis 65
Statistical Analysis 67
RESULTS AND DISCUSSION 68
Preliminary Screening of Soil-Test Methods 68
Relationships Between Soil-Test Values and Relative
Height 72
Relationships Between Soil-Test Values and P Uptake
and Tissue P 79
Relationships Between Soil-Test Values and
Fertilizer Requirements 8l
Relationships Between Soil-Test Values and Height . 83
General Discussion 85
Phosphorus Compounds 88
Solubility of P Compounds in Chemical Extractants . 88
Monocalcium phosphate (MCP) 88
Dicalcium phosphate (DCP) 90
Fluorapatite (FA) 91
Colloidal aluminum phosphate (CA1P) 92
Potassium taranakite (KTK) 92
Wavel 1 i te (WA) 93
Colloidal ferric phosphate (CFeP) 93
Strengite (STR) 9^
Utilization of P Compounds by Slash Pine Seedlings 9^
Relationships Between Seedling Utilization and
Solubility of P Compounds 99
General Discussion 102
Field Calibration of Selected Soil-Test Methods 105
Relationships Between Soil-Test Values and Relative
Height 105
Effect of soil sampling position and depth ... 115
Relationships Between Other Soil and Site
Parameters and Relative Height 117
Relationships Between Soil-Test Values and P-
Fertilizer Requirements 126
Effect of soil sampling position and depth . 131
vi

TABLE OF CONTENTS (Continued)
Page
Relationships Between Other Soil and Site Parameters
and P-Fertilizer Requirements 133
Relationships Between Soil-Test Values and Height . 136
Relationships Between Other Soil and Site Parameters
and Height 1^0
Relationships Between Foliar-P Concentrations and
Tree Parameters 1^5
Relationships Betv/een Foliar P and Soil-Test Values . 150
General Discussion 150
Phosphorus-Retent ion Study 155
Relationships Between P Retention and Soil Properties 155
Physical and chemical properties 157
Extractable A1 and Fe 161
Relative contribution of A1 and Fe to P retention 165
Aluminum and Fe extracted by soil-test methods . 16?
Relationships Between P Retention and Foliar Nutrient
Concentrations 169
Calibration of Extractable A1 Against Field-P Retention 172
SUMMARY AND CONCLUSIONS 179
APPENDIX 186
LITERATURE CITED 259
BIOGRAPHICAL SKETCH 273
VI i

LIST OF TABLES
Table Page
1 FOLIAR P CONCENTRATIONS PRIOR TO FERTILIZATION IN RELATION
TO RESPONSE OF SOUTHERN PINES TO P FERTILIZER 21
2 SOIL-P VALUES (SURFACE HORIZON) OF UNFERTILIZED SOILS IN
RELATION TO RESPONSE OF SOUTHERN PINES TO P FERTILIZER. . 37
3 PROPERTIES OF PHOSPHORUS SOURCES USED IN GREENHOUSE TRIAL 2 50
4 PHYSICAL AND CHEMICAL PROPERTIES OF SOILS USED IN
GREENHOUSE TRIAL 2 52
5 PHOSPHORUS EXTRACTION METHODS 59
6 CLASSIFICATION AND SELECTED PROPERTIES OF 10 SOILS USED
IN GREENHOUSE STUDY 1 69
7 HEIGHT, RELATIVE HEIGHT, P-FERTILIZER REQUIREMENTS, AND
P CONCENTRATION AND UPTAKE OF SLASH PINE SEEDLINGS AFTER
1 AND 2 YEARS' GROWTH ON 10 SOILS IN THE GREENHOUSE .... 70
8 HEIGHT, RELATIVE HEIGHT AND P-FERTILIZER REQUIREMENTS OF
SLASH PINE AFTER 1, 3, AND 5 YEARS' GROWTH, AND FOLIAR P
CONCENTRATION AFTER 4 YEARS' GROWTH IN THE FIELD ON
10 SOILS 71
9 COMPARISON OF THE GOODNESS OF FIT OF FOUR STATISTICAL
MODELS, AS INDICATED BY THE SQUARE OF THE MULTIPLE
CORRELATION COEFFICIENT (R2), RELATING SELECTED TREE
PARAMETERS (DEPENDENT VARIABLE) AND SOIL-TEST VALUES
(INDEPENDENT VARIABLE) 73
10 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND
RELATIVE HEIGHT GROWTH OF SLASH PINE IN FIELD AND
GREENHOUSE EXPERIMENTS ON 10 SOILS 77
11 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND
TISSUE P PARAMETERS OF GREENHOUSE AND FIELD SLASH
PINE GROWN ON 10 SOILS SO
12 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND P-
FERTILIZER REQUIREMENTS OF SLASH PINE IN FIELD AND
GREENHOUSE EXPERIMENTS ON 10 SOILS 82
v i! i

LIST OF TABLES (continued)
Table Page
13 RELATIONSHIPS BETWEEN SELECTED SOIL-TEST VALUES AND HEIGHT
GROWTH IN THE ABSENCE OF P FERTILIZER OF SLASH PINE IN
FIELD AND GREENHOUSE EXPERIMENTS ON 10 SOILS 84
14 SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS IN THE
PRESENCE AND ABSENCE OF TWO SOILS 89
15 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS
BETWEEN SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS
AND THE UPTAKE OF P FROM THESE COMPOUNDS BY SLASH PINE
SEEDLINGS GROWN ON TWO SOILS IN THE GREENHOUSE 100
16 .SOLUBILITY OF P COMPOUNDS IN CHEMICAL EXTRACTANTS FOLLOWING
. 2 MONTHS' INCUBATION IN TWO SOILS 101
17 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL
DEPTHS AND RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3,
AND 5 YEARS' GROWTH ON 72 FIELD SITES 107
18 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN AMOUNTS OF
I P EXTRACTED FROM THE SURFACE 20 cm OF SOIL (BEDDED
AREA) BY FIVE SOIL-TEST METHODS 108
19 SEPARATION OF 72 FIELD SITES INTO RESPONSE QUADRANTS
USING THE TECHNIQUE OF CATE AND NELSON (1965) AND A
CRITICAL HC1-H2S04-EXTRACTABLE P VALUE OF 5 ppm 108
20 REGRESSION EQUATIONS RELATING RELATIVE HEIGHT OF SLASH
PINE AT AGE 1, 3, AND 5 YEARS ( Y1, Y3, AND Y5) TO
' THE LOG TRANSFORMED P EXTRACTED FROM THE SURFACE 20
cm OF SOIL BY THE H20 (Xl) AND HCl-H-SO. (X2) SOIL-
TEST METHODS 116
21 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN AMOUNTS OF
P EXTRACTED BY THE HC1-H2S0/, METHOD FROM TWO SOIL
POSITIONS AND THREE SOIL DEPTHS 116
22 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE
II SURFACE 20 cm OF SOIL AND THAT EXTRACTED FROM WITHIN
THE EFFECTIVE SOIL DEPTH (VOLUME) AT PREDICTING
RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3, AND 5
YEARS' GROWTH ON 72 FIELD SITES 119
23 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES
AND RELATIVE HEIGHT OF SLASH PINE 1, 3, AND 5 YEARS
AFTER P FERTILIZATION ON 72 SITES 120
i x

LIST OF TABLES (continued)
Table Page
24 SIMPLE CORRELATION COEFFICIENTS (r) BETWEEN SELECTED SOIL AND
SITE PROPERTIES OF 72 FIELD SITES 121
25 REGRESSION COEFFICIENTS FOR MULTIPLE REGRESSION EQUATIONS OF
RELATIVE HEIGHT ON DEPTH TO LH, HC1-H SO,-EXTRACTABLE
P(0-20 cm) AND THE SQUARED TERMS OF THESE TWO PARAMETERS. 123
26 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN '.BEST FIT1
MULTIPLE REGRESSION EQUATIONS OF RELATIVE HEIGHT ON SOIL AND
SITE PARAMETERS AND THE SQUARED TERMS OF THESE PARAMETERS 124
27 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL DEPTHS
AND P FERTILIZER REQUIRED TO ACHIEVE 90, 95, AND 100% OF
MAXIMUM HEIGHT GROWTH AFTER 1, 3, AND 5 YEARS' GROWTH ON
72 FIELD SITES 127
28 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE SURFACE
20 cm OF SOIL WITH THAT EXTRACTED FROM WITHIN THE EFFECTIVE
SOIL DEPTH (VOLUME) AT PREDICTING THE FERTILIZER F REQUIRED
TO ACHIEVE 90, 95, AND 100% OF MAXIMUM HEIGHT GROWTH AFTER 1,
3, AND 5 YEARS' GROWTH ON 72 FIELD SITES 132
29 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES AND P ,
FERTILIZER REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT AFTER
1, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES 134
30 RELATIONSHIPS BETWEEN HC1 -H2S0/j-EXTRACTABLE P(0-20 cm) AND P
FERTILIZER REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT ON
GROUPS OF SOILS CLASSED ACCORDING TO THEIR AMOUNT OF
NH^OAc (pH 4.8)-EXTRACTABLE A1 135
31 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN 'BEST FIT'
MULTIPLE REGRESSION EQUATIONS OF P FERTILIZER REQUIRED TO
ACHIEVE 95% OF MAXIMUM HEIGHT ON SOIL AND SITE PARAMETERS
AND THE SQUARED TERMS OF THESE PARAMETERS 137
32 RELATIONSHIPS BETWEEN SOIL-TEST VALUES AT THREE SOIL DEPTHS
AND HEIGHT OF SLASH PINE IN THE ABSENCE OF P FERTILIZER
AFTER 1, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES 138
33 COMPARISON OF THE EFFECTIVENESS OF P EXTRACTED FROM THE SUR
FACE 20 cm OF SOIL AND THAT EXTRACTED FROM WITHIN THE
EFFECTIVE SOIL DEPTH (VOLUME) AT PREDICTING HEIGHT OF SLASH
PINE AFTER I, 3, AND 5 YEARS' GROWTH ON 72 FIELD SITES. . 139
34 RELATIONSHIPS BETWEEN SELECTED SOIL AND SITE PROPERTIES AND
HEIGHT OF SLASH PINE AFTER 1, 3, AND 5 YEARS' GROWTH ON 72
FIELD SITES 142
x

LIST OF TABLES (continued)
Table Page
35 REGRESSION COEFFICIENTS FOR MULTIPLE REGRESSION EQUATIONS
OF RELATIVE HEIGHT ON DEPTH TO LH, H20-EXTRACTABLE
P (0-20 cm) AND THE SQUARED TERMS FOR THESE TWO
PARAMETERS 144
36 REGRESSION COEFFICIENTS FOR VARIABLES INCLUDED IN 'BEST
FIT1 MULTIPLE REGRESSION EQUATIONS OF HEIGHT AT AGE 1, 3,
AND 5 YEARS ON SOIL AND SITE PARAMETERS AND THE SQUARED
TERMS OF THESE PARAMETERS 146
37 RELATIONSHIPS BETWEEN HEIGHT, RELATIVE HEIGHT, AND P
FERTILIZER REQUIREMENTS (95%) AT AGE 1, 3, AND 5 YEARS,
- AND P CONCENTRATIONS IN THE FOLIAGE OF 4-YEAR-OLD SLASH
- PINE 148
38 RELATIONSHIPS BETWEEN EXTRACTABLE-SOIL P AT TWO POSITIONS,
THREE DEPTHS, AND WITHIN THE EFFECTIVE SOIL DEPTH
(VOLUME), AND P CONCENTRATIONS IN THE FOLIAGE OF 4-YEAR-
OLD SLASH PINE 148
39 SOIL PROPERTIES AND THEIR CORRELATION WITH P RETENTION. . 159
40 EXTRACTABLE A1 AND Fe VALUES AND THEIR CORRELATION WITH
P RETENTION 162
41 CORRELATIONS BETWEEN DIFFERENT FORMS OF A1 OR Fe, AND P
RETENTION 164
42 REGRESSION EQUATIONS RELATING P RETENTION (%) TO
DIFFERENT FORMS OF SOIL A1 AND Fe 166
43 ALUMINUM AND Fe EXTRACTED BY FOUR SOIL-TEST METHODS AND
THEIR CORRELATION WITH P RETENTION 168
44 FOLIAR NUTRIENT CONCENTRATIONS AND THEIR CORRELATION WITH
SOIL-P RETENTION 170
45 REGRESSION EQUATIONS OF A1 EXTRACTED BY HCI-H2S0 (Yl),
BRAY 1 (Y2), AND BRAY 2 (Y3) ON A1 EXTRACTED BY NH^OAc (X) 175
46 HEIGHT, DRY WEIGHT, AND P CONCENTRATION OF SLASH PINE
SEEDLINGS AFTER 1 AND 2 YEARS' GROWTH ON 10 SOILS IN THE
GREENHOUSE RECEIVING FOUR P TREATMENTS 187
47 AVERAGE HEIGHTS OF SLASH PINE AS AFFECTED BY P TREATMENTS
AFTER 1, 3, AND 5 YEARS' GROWTH IN THE FIELD ON 10 SOILS 194
x!

LIST OF TABLES (continued)
Table Page
48 DRY WEIGHT, P CONCENTRATION, AND P UPTAKE IN TOPS AND
ROOTS OF SLASH PINE SEEDLINGS AFTER 1 YEAR OF GROWTH
ON NINE SOILS IN THE GREENHOUSE 196
49 AMOUNTS OF P EXTRACTED FROM 10 SOILS BY SOIL-TEST METHODS 197
50 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS
BETWEEN SOIL-TEST VALUES AND SEEDLING PARAMETERS FROM
GREENHOUSE STUDY 1 202
51 MULTIPLE CORRELATION COEFFICIENTS (R2) FOR RELATIONSHIPS
BETWEEN SOIL-TEST VALUES AND TREE PARAMETERS FROM 10
SELECTED FIELD TRIALS 207
52 DRY WEIGHT AND P UPTAKE OF ENTIRE SLASH PINE SEEDLINGS, AS
AFFECTED BY P SOURCE, AFTER THE 8 MONTHS OF GROWTH ON
TWO SOILS IN THE GREENHOUSE 212
53 ANALYSIS OF VARIANCE OF SEEDLING DRY WEIGHTS AND P UPTAKES
OF GREENHOUSE TRIAL 2 213
54 .SOIL CLASSIFICATION, SITE PROPERTIES, AND SELECTED CHEMICAL
PROPERTIES OF UNFERTILIZED SOILS (0-20 cm), FOR THE 24
FIELD SITES 214
55 AMOUNTS OF P EXTRACTED FROM SOILS, COLLECTED FROM CONTROL
PLOTS OF 24 FIELD TRIALS, BY FIVE SOIL-TEST METHODS, AND
AMOUNTS OF Ca, Mg, K, and A1 EXTRACTED BY NH^OAc(pH 4.8)
AND SOIL pH 218
56 PHYSICAL PROPERTIES OF SOILS COLLECTED FROM 24 FIELD TRIALS 231
57 HEIGHT AND RELATIVE HEIGHT OF SLASH PINE AFTER 1, 3, AND 5
YEARS' GROWTH ON ¡2 SITES IN THE FIELD 241
58 PHOSPHATE FERTILIZER (CSP) REQUIRED TO ACHIEVE 90, 95, AND
100* OF MAXIMUM HEIGHT OF SLASH PINE 1, 3, AND 5 YEARS
AFTER FERTILIZATION OF 72 FIELD SITES 244
59 CONCENTRATIONS OF N, P, K, Ca, Mg, A1, AND Fe IN FOLIAGE
COLLECTED FROM 4-YEAR-OLD SLASH PINE GROWING IN THE
CONTROL PLOTS OF 24 FIELD TRIALS 248
60 CLASSIFICATION, SELECTED PHYSICAL AND CHEMICAL PROPERTIES,
AND P-RETENTI ON CHARACTERISTICS (LANGMUIR AND SATURATION
MAXIMA) OF 42 LOWER COASTAL PLAIN FOREST SOILS 251
xi i

LIST OF TABLES (continued)
Table Page
61 AMOUNTS OF A1 AND Fe EXTRACTED FROM 42 COASTAL PLAIN
FOREST SOILS BY SIX CHEMICAL EXTRACTANTS. 253
62 AMOUNTS OF A1 AND Fe EXTRACTED FROM 42 COASTAL PLAIN
FOREST SOILS BY FOUR SOIL P-TEST METHODS 255
63 AMOUNTS OF TOTAL P IN THE SURFACE 20 cm OF SOIL COLLECTED
FROM THE CONTROL, Pi(56 Kg P/ha), AND P2(224 Kg P/ha)
PLOTS OF 10 SELECTED FIELD TRIALS 4 YEARS AFTER P-
FERTILIZER APPLICATION 257
x i 1 !

LIST OF FIGURES
gure Page
1 RELATIONSHIP BETWEEN R2 VALUES FOR REGRESSIONS OF RELATIVE
HEIGHT AT AGE 1, 3, AND 5 YEARS ON SOIL-TEST VALUES, AND
MEAN AMOUNTS OF P EXTRACTED BY SOIL-TEST METHODS 75
2 PHOSPHORUS UPTAKE BY SLASH PINE SEEDLINGS AFTER 3 MONTHS'
GROWTH ON TWO SOILS TREATED WITH EIGHT P COMPOUNDS .... 96
3 .DRY MATTER OF SLASH PINE SEEDLINGS AFTER 8 MONTHS GROWTH
ON TWO SOILS TREATED WITH EIGHT P COMPOUNDS 97
4 RELATIONSHIP BETWEEN HC1-H-SO^-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 1 YEAR AFTER P FERTILIZATION Ill
5 RELATIONSHIP BETWEEN HC 1 -H2 SO j-EXTRACTA BLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 3 YEARS AFTER FERTILIZATION 112
6 RELATIONSHIP BETWEEN HC 1 -H^O^-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT (Y) OF SLASH
PINE 5 YEARS AFTER P FERTILIZATION 113
7 RELATIONSHIPS BETWEEN HC1-H2SO^-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND RELATIVE HEIGHT OF SLASH
PINE 1, 3, AND 5 YEARS (Y1, Y3, AND Y5) AFTER P
FERTILIZATION 11 4
8 RELATIONSHIP BETWEEN DEPTH TO LIMITING HORIZON (X) AND
RELATIVE HEIGHT (Y) OF SLASH PINE 5 YEARS AFTER
FERTILIZATION 122
9 RELATIONSHIPS BETWEEN HC 1-H-SO.-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND AMOUNT OF P FERTILIZER (CSP)
REQUIRED TO ACHIEVE 30% OF MAXIMUM HEIGHT GROWTH OVER
PERIODS OF 1, 3, AND 5 YEARS (Y1, Y3, AND Y5)
FOLLOWING FERTILIZATION 128
10RELATIONSHIPS BETWEEN HC1-H-SO.-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND AMOUNT OF P FERTILIZER (CSP)
REQUIRED TO ACHIEVE 95% OF MAXIMUM HEIGHT GROWTH OVER
PERIODS OF 1, 3, AND 5 YEARS (Y 1, Y3, AND Y5)
FOLLOWING FERTILIZATION 129
xi V

LIST OF FIGURES (continued)
Figure Page
11 RELATIONSHIP BETWEEN DEPTH TO LIMITING HORIZON (X) AND
THE HEIGHT OF SLASH PINE AFTER 5 YEARS' GROWTH 143
12 RELATIONSHIPS BETWEEN P CONCENTRATIONS IN FOLIAGE (X)
OF 4-YEAR-OLD SLASH PINE AND RELATIVE HEIGHT OF SLASH
PINE 1, 3, AND 5 YEARS (Y1, Y3, AND Y5) AFTER P
FERTILIZATION 149
13 RELATIONSHIP BETWEEN HC1-H2S0^-EXTRACTABLE P(X) IN THE
SURFACE 20 cm OF SOIL AND P CONCENTRATION IN FOLIAGE
(Y) OF 4-YEAR-OLD SLASH PINE 151
14 PHOSPHORUS-ADSORPTION ISOTHERMS OF FOUR LOWER COASTAL
PLAIN SOIL TYPES REPRESENTATIVE OF FOUR SOIL ORDERS. . 156
15 RELATIONSHIP BETWEEN % P RETENTION FROM THREE P
SOLUTIONS (100, 300 AND 2,500 yg P/g SOIL) AND
NH^OAc (pH 4,8)-EXTRACTABLE A1 158
16 RELATIONSHIP BETWEEN SURFACE (0-20 cm) RETENTION OF P
APPLIED AS CSP 4 YEARS PREVIOUSLY AND NH^OAc (pH 4.8)-
EXTRACTABLE A1 AT 10 SITES GROWING SLASH PINE 174
17 PHOSPHORUS RETENTION AS DETERMINED FROM ADDITION OF
2,500 yg P/g SOIL IN A AND Bh HORIZONS OF SIX
SPODOSOLS 177
xv

Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy
SOIL TESTING AS A GUIDE TO PHOSPHORUS FERTILIZATION
OF SLASH PINE (Pinus elliottii var. elliottii Engelm.)
By
Russel 1 Bal lard
August, 1974
Chairman: Dr. W.L. Pritchett
Major Department: Soil Science
This investigation was designed to develop and calibrate a soil
test or tests for predicting the amount and type of phosphatic fertili
zer, if any, needed at time of planting slash pine (Pinus elliottii var.
elliottii Engelm.) on forest soils in the southeastern lower Coastal
Plain.
In a preliminary screening of over 100 different procedures for
extracting or characterizing soil P, it was found that soil-test methods
which extracted relatively small amounts of P from the soil (H2O, NH^OAc
pH 4.8, neutral salts, very dilute acid solutions at narrow soi1:solution
ratios, anion exchange resin) provided the best index of first-year
growth and response of slash pine to P fertilizer both in greenhouse and
field fertilizer trials on the same 10 soils. However, soil-test methods
capable of extracting larger quantities of soil P, irrespective of the
extracting reagent used (strong acid, alkaline, or complexing agents),
were more effective predictors of growth and response achieved over
longer growth periods. Procedures which extracted total, or specific
fractions of soil P (organic-P, Ca-P, Fe-P, Al-P) were less effective
xv i

predictors of P fertilizer needs than more conventional test methods for
available P.
An examination of the solubility of P compounds (mono and dicalcium
phosphates, f1uorapatite, colloidal A1 and Fe phosphates, K taranakite,
wavellite, and strengite) in extractants of several common soil-P test
methods and their utilization by slash pine seedlings, showed that soil-
test methods which provided the best index of early growth and response
in preliminary screenings extracted little more P from these compounds
than did H20 alone. Methods which were more successful predictors over
longer growth periods extracted appreciable quantities of P from the Ca
phosphates, colloidal phosphates, and K taranakite, which were effec
tive P sources for seedlings; but, they did not extract much P from
wavellite and strengite, which were ineffective P sources for seedlings.
Methods utilizing strong acids did, however, overestimate the availa
bility of fluorapatite to seedlings on a soil of pH > 5-
Five soil-test methods (H20, NH^OAc pH 4.8, 0.05N_ HC1 + 0.025N_
H2S0it, 0.03 NH^F + 0.025N. HC 1 0.03^ NH^F + 0.1N HC1) selected on the
basis of results from the preliminary screening, were calibrated against
slash pine growth and response information obtained from 72 field ferti
lizer trials 1, 3, and 5 years after establishment. The method involving
use of 0.05]^ HCI + 0.025N^ H2SO^ provided the best prediction of height
response and P fertilizer requirements over the 3 and 5_year growth
periods. A surface soil (0-20 cm) value of 5 ppm by this method pro
vided an effective delineation of P responsive sites. Soils testing
between 5-0 and 2.5 ppm P required ca. 20-40 kg P/ha and those testing
below 2.5 ppm P required ca. 40-80 kg P/ha to provide an adequate P
supply to slash pine over the above growth periods.
xv 1

Laboratory-determined P-retention capacity of 42 forest soils was
significantly related to A1 extracted by either conventional procedures
(KC1 leaching, pyrophosphate, oxalate, and CDB extraction) or four soil-
P test methods. Soil A1 extracted by these four soil-P test methods was
calibrated against P leaching losses over a 4-year period in the field
following application of concentrated superphosphate (CSP). Soils con
taining less than 40, 120, 300, or 400 ppm A1 extractable by NH^QAc
pH 4.8, 0.05^ HC1 + 0.025N^ H2S0i>, 0.03^ NH^F + 0.025N.HC1, or 0.03N_
NH^F + O.OIN^ HC1, respectively, exhibited excess leaching losses of P
from soluble CSP fertilizer; use of less soluble P fertilizers, such as
rock phosphate, was suggested on such soils.
XVIII

INTRODUCTION
Fertilization is one of many cultural practices available for in
creasing the productivity of forest lands. Increasing demand for wood
products, associated with a decreasing land base for commercial timber
production, has led to an upsurge of interest in intensive forest manage
ment techniques in the last decade.
In the southeastern Coastal Plain, intensive forest cultural prac-
9
tices such as clear-cut harvesting, site preparation, use of improved ge
netic stock, and fertilization are now widely used. Climatic conditions
/
which favor long growing seasons and high biological activity, the adop-
tion of intensive management techniques which increase growth rates, and
the inherently'infertile acid sandy soils, have been the major factors
contributing to the successful results obtained from the use of fertilizer
in southeastern Coastal Plain forests. Results from more than 225 field
experiments established throughout the Coastal Plain by the Cooperative
Research in Forest Fertilization (CRIFF) program have shown that about
two-thirds of the tested sites were responsive to nitrogen and/or phos
phate fertilizers (CRIFF annual report, 1973). Trials in young planta
tions indicated that about two-thirds of the sites were responsive to
phosphate fertilizers and three-fourths responsive to nitrogen and/or
phosphate (Pritchett and Smith, 1972).
Approximately 770,000 acres of pines are planted annually in the
Southeast, of which it has been projected 500,000 acres would benefit from
fertilizer additions (CRIFF annual report, 1973)- Considerably less than
1

2
this projected acreage is actually fertilized. This, in part, can be at
tributed to the usual caution of forest managers in accepting new practices,
but another major reason is the lack of a precise diagnostic technique for
delineating areas where economic responses can be expected from fertilizer
additions. This inadequacy was emphasized in a recent survey of 50 fores
try organizations in the South soliciting opinions on forest fertilization
research priorities. The concensus of most respondents was that one of
the most pressing needs was "for a predictive mechanism telling where,
when, how and with what to fertilize (Hoehring, 1972).
The two major diagnostic techniques used in forestry are foliar and
soil analyses. Foliar analysis has generally proved to be more reliable
than soil analysis for predicting increased growth of southern pines from
P fertilization (Pritchett, 1968j Wells and Crutchfield, 1969). However
foliar analysis cannot be used in the absence of growing stock on the site
and, consequently, an effective soil test has the advantage over foliar
analysis in that it can be used to predict the need for fertilizer at the
time of planting.
Accurate diagnosis is essential for the success of any forest fer
tilization program. The diagnosis should provide information on the de
gree of response to be expected, the amount of fertilizer required to ob
tain this response, and the most effective nutrient source to use. The
question of most effective P source is of particular importance to P fer
tilization in the Coastal Plain. Humphreys and Pritchett (1971) found
that the P-retent ion capacity of soils in the Coastal Plain region had a
strong influence on the long-term effectiveness of P sources of different
solubi 1ity.
The general purpose of this study was to develop a soil test or
tests which would provide the necessary information required for a forest

3
manager to make the correct decisions regarding P fertilization practices
at time of planting. Specific objectives were as follows:
1. Evaluation of the effectiveness of a wide range of soil-P testing
techniques for predicting both growth and response of P fertilizer
applied to slash pine at time of planting.
2. Examination of the effects that soil and environmental parameters
have on the predictive accuracy of soil-P tests.
3. Calibration of the best soil-P test against the degree of response
obtained from fertilizer additions and the amount of fertilizer
required to produce this response.
k. Development of a predictive procedure for delineating sites where
phosphate sources of different solubility should be used.
5. Determination of the capacity of slash pine to utilize a range of
soil-P compounds, and relate this ability to the solubility of these
compounds in the various extractants used in soil-P tests.
*
NOTE
CRIFF annual report, 1973- (unpublished). Soil Science Department,
University of Florida, Gainesville, Florida.

LITERATURE REVIEW
Forest Fertilization
H i storica1
The value of fertilizers for increasing the productivity of timber
lands was first recognized in Europe. Fertilizer trials were established
on peat lands in Sweden as early as 1898 (Hagner, 1967). During the pe
riod 1900-1925, forest fertilizer trials were established in many European
countries including Finland (Salonen, 1967), Belgium, Germany, and Denmark
(Tamm, 1968), and Britain (Leyton, 1958). Experimentation expanded in
Europe between 1925 and I960, establishing a firm scientific foundation
to tree nutrition which led to the first operational applications of fer
tilizer in Germany (Tamm, 1968) and Britain (Leyton, 1958). In Austra
lia, the importance of phosphatic fertilizers in achieving healthy growth
of introduced pines planted on infertile sandy soils was established as
early as 1930 (Kessell and Stoate, 1938).
Research on forest fertilization is relatively new in North Ameri
ca. Some of the earliest trials on this-continent were N experiments es
tablished on hardwood stands in the northeastern USA by Mitchell and
Chandler (1939). Westveld established a phosphate fertilizer trial in
the southeastern Coastal Plain in 19^5 (Pritchett and Swinford, 1961),
while research on forest fertilization was started around 1950 in the
Pacific Northwest (Gessell, 1968).

The apparent lack of interest shown in forest fertilization by most
North American foresters prior to I960 has been attributed to their pre
occupation with extensive rather than intensive management techniques
(Bengtson, 1972). In addition, the results of most soil-site research,
particularly that by Coile (1952), indicated that soil physical charac
teristics provided better indices of site productivity than did soil
chemical properties.
N and P Fertilization
Several general reviews on forest fertilization (White and Leaf,.
1957; -Stoeckler and Arneman, I960; Swan, 1965; Mustanoja and Leaf, 1965)
and specific reviews of progress in countries such as Sweden (Hagner,
1967; Holmen, 1967), Finland (Salonen, 1967; Paarlahti, 1967), Norway,
(Jerven, 1967), Japan (Kawana, 1969), Britain (Leyton, 1958; Binns, 1969),
Australia (Raupach, 1967; Gentle and Humphreys, 1968), Canada (Swan, 1969;
Krause, 1973), and the USA (Gessell, 1968; White, 1968; Bengtson, 1968 and
1972; Safford, 1973), have i 1 lust rated the tremendous advances in forest
fertilization research and the application of the results during the last
decade. Most recorded responses have been to N and P fertilizers although
isolated responses to K fertilizers have been reported in Japan (Kawana,
1969), Austra1ia (Hall and Purnell, 1961), Finland (Salonen, 1967), mid-
western Europe (Hagner, 1971), and the northeastern USA (Stone and Leaf,
1967).
The most spectacular response of coniferous species to N applica
tions have been recorded in Scandinavia (Hagner, 1967) and the Douglas-
fir region of the northwestern USA and southwestern Canada (Strand and
Miller, 1969). In the Scandinavian countries, operational fertilization
began about I960 (Hagner, 1967) with urea as the principal N source. In

6
recent years there has been a switch to ammonium nitrate because it ap
peared to give better results than N applied as urea (Hagner, 1971)
Normal application rates, designed to give a response over a 5_year period,
were about 120 to 180 kg N/ha. In 1970, approximately 143,000 ha of es
tablished forest stands were fertilized in Scandinavia. In Sweden alone,
over 500,000 ha had received N fertilizer by the end of 1970 (Hagner,
1971). Operational fertilization with urea-N began in the Pacific North
west in 1965 (Anderson, 1969). About 60,000 ha had been fertilized up to
the end of 1970, with some 40-50,000 ha fertilized during 1970 (Hagner,
1971). The average rate applied in the Douglas-fir region was about 200
kg N/ha, which gives a response over a period of 5 to 7 years (Strand and
Hiller, 1969).
Pritchett and Smith (1970) reported several studies in which slash
pine (Pinus el iiottii var. el 1iottii Engelm.) responded to N applications
in the USA southeastern Coastal Plain. In their studies, most responses
were obtained in established stands, where drainage conditions apparently
had little effect on growth response. However, in one study (Broerman,
1967), stand density had a strong modifying influence on tree response to
N. In operational fertilization, urea has been the preferred N source
and has been applied at rates of 70 to 100 kg N/ha (Pritchett and Hanna,
1969). In 28 (NxP factorial) fertilizer trials established at time of
planting slash pine throughout the Coastal Plain, N applied alone had
little effect on the growth of seedlings in any tests except those in the
somewhat poorly-drained flatwood sites (Pritchett and Smith, 1972). On
P-deficient sites, the application of N alone tended to suppress growth,
but when applied in conjunction with P, N significantly increased growth
over that obtained from P alone in 11% of all tests.

C7)
Spectacular growth responses following addition of phosphate fer
tilizers to conifers have been recorded in many countries. As a conse
quence, routine application of P fertilizer, particularly at time of
planting, is now practiced on P-deficient sites in New Zealand (Conway,
1962; Armitage, 1969), Australia (Gentle and Humphreys, 1968), Japan
(Kawana, 1969), Britain (Leyton, 1958), Finland (Salonen, 1967), Norway
(Jerven, 1967), France and Germany (Baule and Fricker, 197-0) and the
southeastern USA (Pritchett and Hanna, 1969). In 1970, 70-80,000 ha were
fertilized at time of planting in Japan, about 30,000 ha in midwest
Europe and 10,000 ha in both the US Southeast and Oceania (Australia and
New Zealand) (Hagner, 1971). Superphosphate was the principal fertilizer
used.
Applications of P fertilizer to young stands at rates of 78 kg P/ha
have maintained responses for a minimum of 15 years (Pritchett and Swin-
ford, 1961; Gentle, Humphreys, and Lambert, 1965). However in other fer
tilizer trials in both the US Southeast and Australia, growth response
and foliar concentrations have shown that applications of superphosphate
up to 78 kg P/ha began to lose their effectiveness 7 to 15 years after
application (Gentle and Humphreys, 1968; Humphreys and Pritchett, 1971;
Pritchett and Smith, 197*0. The loss in efficiency of soluble forms of
P fertilizer was attributed to either leaching from the rooting zone in
soils of low P-retention capacity, or binding in unavailable forms in
soils of very high P-retention capacity. The practical and economic prob
lems of providing an adequate P supply throughout a rotation has led to a
policy in New Zealand and Australia of applying only sufficient fertilizer
at time of planting to provide an adequate P supply for pinus species up

8
to canopy closure at age 5 to 10 years (Armitage, 1969; Craig, 1972).
Once canopy closure is reached the stands are refertilized by aerial
application.
The characteristics of forest soils in the southeastern USA lower
Coastal Plain have been described by Pritchett and Smith (1970). These
soils are predominantly infertile, acid sands with drainage characteris
tics ranging from excessively drained to poorly drained. Responses of
slash pine to P fertilizers in the lower Coastal Plain have been related
to the drainage characteristics of the sites. In eight fertilizer trials,
Pritchett and Llewellyn (1966) found that 3 to 5 years after application',
significant responses to P fertilizer had occurred in 5 out of the 8
trials, all of which were on sites classified as poorly or somewhat poor
ly drained. On two well-drained sites no response to P was recorded,
although they did respond to N application. Responses from the same
trials, recorded 7 to 11 years after fertilizer application (Humphreys
and Pritchett, 1971), showed that the greatest response had occurred on
the poorly drained sites, followed by the somewhat poorly drained sites,
with no P response on the well-drained sites. In a more recent series of
trials established at planting time on 28 sites throughout the Coastal
Plain, Pritchett and Smith (1972) reported that 3 years after fertilizer
application, 6 of 7 experiments located on poorly drained sites, 8 of 10
on somewhat poorly drained sites, k of 8 on moderately well drained sites,
and none of the 3 on excessively drained sites responded significantly to
P applications. In the same series it was found that N alone significantly
increased growth only in 8 of the experiments on the somewhat poorly
drained sites, although when N was applied with P, it increased the re
sponse over that obtained with P alone in 72% of the 28 experiments.

9
P Sources Used in Forestry
A number of P sources of different solubility have been used suc
cessfully in forestry. Ordinary (OSP) and concentrated superphosphate
(CSP) are the most extensively used sources (Pritchett and Hanna, 1969;
Gentle and Humphreys, 1968), but ground rock phosphate (RP) has proved to
be an effective source for many forest soils (Young, 1948; Gentle et al.,
1965; Pritchett and Humphreys, 1971; Pritchett and Smith, 197*0. Its use
in operational fertilization is recommended on certain soils in Britain
(Leyton, 1958), Australia (Gentle and Humphreys, 1968), Finland (Salonen,
1967), Ireland (O'Carroll, 1967) and the southeastern USA (CRIFF annual
report, 1973). The soils on which slowly soluble rock phosphates are
recommended are principally acid sands and peats with low P-retention
ability. Other P sources used to a limited extent in operational ferti
lization include diammonium phosphate (DAP) (Pritchett and Hanna, 1969)
and basic slag (O'Carroll, 1967). In order to fully understand the dif-

ference in effectiveness of P sources of different solubility, a knowledge
of their reaction with soil constituents is necessary.
Reaction in Soils
Water-soluble monocalcium phosphate (MCP) is the principal P com
pound in both OSP and CSP. The reactions and transformations of MCP fol
lowing addition to soil have been reviewed in several publications (Huff
man, 1962 and 1968; Olsen and Flowerday, 1971). Initially MCP undergoes
hydrolysis producing an extremely acid (pH 1.48) zone of phosphoric acid
around the pellet. The phosphoric acid diffuses into the soil leaving a
residue of dicalcium phosphate dihydrate (DCPD) (Lindsay and Stephenson,
1959 a). The acid H^PO^ solution may cause dissolution of Fe, Al, Mn, Ca,

10
and other soil constituents. As the pH of the diffusing fertilizer solu
tion rises following reaction with soil constituents and dilution by the
soil solution, the solubility products of various A1 and Fe phosphates
are exceeded and they precipitate (Lindsay and Stephenson, 1959 b). The
initial precipitates in an acid soil (Hartsells fine sandy loam, pH 4.6)
were identified as colloidal products of the type (Fe, A1, X)P0^.nH20 and
various crystalline acidic phosphates of the type X20.3(A1, Fe)20^.6P205
.2OH2O. The Fe and A1 may substitute for one another, and the X indicates
a cation other than Fe and A1 (Lindsay and Stephenson, 1959 b; Lindsay,
Frazier, and Stephenson, 1962).
Many of the initial reaction products formed in the soil following
addition of MCP have been reported to be good sources of plant-available
P (Taylor, Gurney and Lindsay, I960; Juo and Ellis, 1968). However, as
the initial colloidal and poorly crystalline products age, they become
more crystalline and less available sources of P (Juo and Ellis, 1968).
Stable crystalline phosphate minerals of the vari sc ite-strengite series
are considered to be the ultimate end-product of the reactions and trans
formations of MCP in acid soils (Wright and Peech, I960). Juo and Ellis
(1968) reported that Fe phosphates crystallize more rapidly than A1 phos
phates in acid soils. They observed that amorphous Fe-P forms crystal
lized within 9 months at 35 C, while amorphous A1 -P forms remained com
pletely amorphous over the same period. This slower rate of crystalliza
tion of A1 phosphates was cited by Juo and Ellis (1968) as the probable
explanation why colloidal Al-P was more effective than colloidal Fe-P as
a long term source of P for plants (Taylor et al., I960) and explained
the numerous reports of close correlations between Al-P (extracted by
alkaline NH^F) and plant-P uptake (Thomas and Peaslee, 1973).

The transformations of the residual DCPD are greatly influenced
by soil pH. Moreno, Brown, and Osborn (i960) found that DCPD in aqueous
solutions hydrolysed to the more basic octacalcium phosphate (OCP) at a
pH of 6.38 and that the rate of hydrolysis increased at higher pH values.
Bell and Black (1970 a) reported that DCPD reverted to OCP at 35 C in soil
at pH 6.4, but at 25 C a pH of at least 6.9 was required before OCP be
came detectable. Lehr and Brown (1958) found that both MCP and DCPD
applied to alkaline soils (pH 8.0) reverted to OCP and hydroxyapatite.
In acid soils, DCPD dissolves releasing Ca^+ and HP0^ (Olsen and
Flowerday, 1971). The rate of dissolution of DCPD, according to Moreno,
Lindsay, and Osborn (i960) is a function of the rate of removal of Ca and
P from solution by plants and/or the soil sorption and exchange sites.
It has been reported that DCPD persists in acid soils for several months
(Lehr and Brown, 1958). Moreno et al- (i960) found that once the dissolu
tion of DCPD in acid soils was complete, P concentrations in soil solution
decreased abruptly.
The reactions and transformations of RP following addition to soils
have been less subject to research than has MCP. Apatite is the principal
P form in RP, but it varies in degree of crystallinity and proportions of
flor and hydroxyapatite, depending on location of the RP source (Olsen
and Flowerday, 1971). The dissolution of RP in soils is closely related
to soil pH. Lindsay and Moreno (i960) developed solubility diagrams for
predicting P in solution at various pH ranges in the presence of flor
and hydroxyapatite. For f1uorapatite, the equation developed for pre
dicting P ¡n solution was:
p h2poa
-5-13 + 2 pH

12
Huffman (1962} found that the rate of dissolution of flor and hydroxy
apatites was proportional to their surface areas and that this rate de
creased with a rise in pH. Mattson et al. (1951) reported that the addi
tion of organic matter, which complexed Ca, increased the solubility of
hydroxyapatite about tenfold. Even in acid soils, the dissolution rate
of RP has been found to be slow. Chu, Moschler,and Thomas (1962) reported
that 18% of the RP applied k years previously to a soil at pH 5-7 (Nason
silt loam) had dissolved, while only 7*5% had dissolved 7 years after
application to a soil at pH 6.1 (Frederick silt loam). Gentle et al.
(1965) found that most of the RP applied to a podsolized soil at pH 5 was
still present as Ca phosphates after 15 years. Similarly, in the south
eastern Coastal Plain, Humphreys and Pritchett (1971) reported that 7 to 11
years after the addition of RP to 7 soils ranging in pH from k.2 to 5.^,
substantial amounts of P were recovered in the Ca-P fraction. In all
three of these studies, the transformation of RP produced an increase in
the Al-P and Fe-P fractions.
Diammonium phosphate (DAP) is water soluble and upon dissolution
in the soil produces an alkaline solution of about pH 8.0. The initial
reaction of the fertilizer solution produces basic phosphates of the gen
eral type (Ca, Mg) (NH^)^ (HPO^)2H2O (Lindsay et al., 1962; Bell and
Black, 1970 b). In slightly acid soils, Bell and Black (1970 b) found
DCPD was formed once the fertilizer solution had moved several centimeters
from the fertilizer source and the pH decreased. Bell and Black (1970 b)
also reported that in comparison with MCP and monoammonium phosphate,
which both form acid solutions, the movement through soil columns of P
from DAP was far greater, irrespective of pH and clay content of the soil.

Phosphate-fractionation studies following P additions and work re
lating P-retention capacity of soils to various soil properties support
the findings discussed above that in acid soils A1 and Fe are the princi
pal soil constituents with which P reacts. The literature on both these
subjects is voluminous and for the sake of clarity and brevity will be
illustrated by work in the southeastern USA.
Yuan, Robertson, and Neller (i960) found in laboratory tests that
water-soluble P rapidly reverted to Al-P upon addition to three acid
sandy Florida soils. The addition of CSP to a Lakeland fine sand in a
short-term lysimeter study was found by Hortenstine (1969) to signifi
cantly increase only water soluble and Al-P fractions. Fiskell and
Spencer (1964) reported that 6 years after the addition of a heavy appli
cation of CSP to Lakeland fine sand, the increase in P fractions followed
the order Al-P>Fe-P>Ca-P. A similar result was reported by Robertson,
Thompson, and Hutton (1966) with a Red Bay fine sandy loam and a Norfolk
loamy fine sand. However, they reported that the ratio of Al-P to Fe-P
decreased with time. They attributed this decrease to uptake of Al-P by
plants and a transformation of Al-P to less soluble and less available
Fe-P. Phosphorus-fractionation studies on seven forest soils in the south
eastern USA (Humphreys and Pritchett, 1971) revealed that 7 to 11 years
after application of OSP the majority of the recoverable P was found in the
Al-P and Fe-P fractions. The distribution between the two fractions varied
between soils. Equal amounts of each fraction were found in a Bladen soil,
while nearly all applied P was recovered in the Al-P form in a Rutlege
soil. Two Kershaw soils were found to be intermediate in behavior between
the Bladen and Rutlege soils. The interpretation of these fractionation
studies must take into consideration the limitations of the fractionation

procedure. Although all studies, except that by Hortenstine (1964), used
the Chang and Jackson (1957) prodecure with Fife's (1959) modification of
the NH^F extraction, work by Bromfield (1967) has shown that on recently
fertilized soils the prodecure can provide a misleading estimate of the
relative amounts of A1 -P and Fe-P. This is due to the ability of NH^F at
pH 8.2 to extract appreciable DCPD which leads to an over estimation of
A1 P
In acid soils, A1 and Fe contents have been shown to be closely
correlated with P-retention capacity. Coleman, Thorup, and Jackson (i960),
working with subsoil samples from 60 North Carolina Piedmont soils, found
that exchangeable A1 extracted by N_ KC1 was most closely correlated with
P retention. Similar findings were reported by Syers et al. (1971) using
15 topsoil samples from Brazil and by Udo and Uzu (1972) using 10 acid
--n
Nigerian soils. Aluminum extracted by (NH/t)2C20{ (pH 3-0), a reagent
which extracts amorphous forms of Fe and Al, has been reported to be
closely related to P retention in a range of acid soils (Bromf?eId,1965;
Saunders, 1965; Yuan and Breland, 1969). Yuan and Breland (1969) used
the horizons of 43 virgin soils representative of the soil orders found
in the southeastern Coastal Plain. Iron extracted by oxalate generally
proved to be less effective than oxalate-extractable Al for predicting P
retention (Bromfield, 1965; Yuan and Breland, 1969; Syers et al., 1971).
However on some soils rich in iron oxides, oxalate-extractable Fe proved
to be more successful than Al at predicting P retention (Ahenkorah, 1968).
Extractants which remove crystalline forms of Fe and Al (citrate-dithionate
bicarbonate, termed herein as CDS) were generally less effective than
oxalate for extracting amounts of Al and Fe related to P retention. How
ever CDB extractable-A] appears to be more closely related to P retention

than CDB extractable-Fe (Syers et al., 1971; Udo and Uzu, 1972). These
results have been interpreted as indicating that the relative activity of
the various soil forms of Al and Fe in P retention follows the order:
exchangeable>amorphous>crystal1ine; while, within each form, Al is more
active than Fe on a per unit weight basis (Bromfield, 1965; Syers et al.,
1971; Udo and Uzu, 1972). Soil constituents that contribute to P reten
tion in soils other than various Al and Fe forms include phy1 los i 1icate
clays and CaCO^ (Kuo and Lotse, 1972). Calcium carbonate is of little
significance in virgin acid soils, but kaolinite, which is more effective
at retaining P than 2:1 clays (Ramula, Pratt, and Page, 1967) can retain
up to l87)jg P/g (Kuo and Lotse, 1972). Nevertheless, kaolinite is less
reactive than other forms of Al and Fe in the soil (Huffman, 1968).
P Effectiveness in Forestry
.The majority of fertilizer trials comparing the effectiveness of
different P sources in forestry have involved water-soluble superphosphate
(OSP or CSP) and slowly soluble rock phosphate (RP) sources.
Early trials in Australia comparing the effectiveness of OSP and
RP gave similar results: Both RP and OSP applied at equivalent rates of
P were equally effective as long-term P sources (Young, 1948; Richards,
1956; Richards, 1961; Gentle et al., 1965). Although an adequate de
scription of the soil properties was not provided in most of these reports,
it was mentioned that all soils were acid. Young (1948) and Richards
(1956) pointed out that RP was a more economical source than OSP, because
it was cheaper and more effective per unit weight of fertilizer applied.
Gentle and Humphreys (1968) reported that a trial established in Penrose
State Forest, New South Wales, in which OSP and RP were applied at equal
rates of P (77 kg/ha), showed OSP to have an early superiority, but after

16
15 years, current increment in the RP treatment was significantly greater
than that in the OSP treatment. The decline in effectiveness of the OSP
was attributed to the high P-retention capacity of the soil. Working on
the sandy coastal plain of Western Australia, Hopkins (i960) found RP to
be the most effective source of P for P. pinaster. This was attributed
to less leaching of P from RP than from OSP on these leached, acid,sandy
soils.
In summarizing the results from fertilizer trials in New South
Wales, Gentle and Humphreys (1968) suggested that on soils with a high P-
retention capacity, banded applications of a semi-soluble source should
be used, whi1st on near neutral soils with a medium P-retention capacity,
OSP or CSP would be preferred. On acid soils with a low P-retention ca
pacity they suggested the use of RP.
Rock phosphates are the preferred source of P for fertilizing
forests on acid peats in England (Leyton, 1958), Finland (Salonen, 1967)
and Ireland (Dickson, 1971)- The success of RP is probably due to the
acidity and low P-retention capacity of these peats, although there is no
reported evidence in the literature to indicate leaching losses of soluble
P sources applied to peats. In some instances, however, the selection of
RP was based on economic considerations, since OSP and RP were equally
effective (Dickson, 1971). Hagenzieker 0958) also reported that OSP, RP
and basic slag were equally effective sources on the acid forest soils of
Hoi land.
Bengtson (1970) examined the comparative value of CSP and RP as P
sources for slash pine in greenhouse studies with a Lakeland fine sand
(pH A.8). Using different placements, he found CSP to be superior to
RP in all tests. The performance of RP was improved by mixing it with

/^N.
17
the soil as opposed to broadcast or banded placement. Brendemuehl (1970)
reported a similar improvement of performance in field trials from mixing
RP with the same soi1 type.
The results from a series of trials established using slash pine to
compare the effectiveness of OSP and RP on soils in the southeastern lower
Coastal Plain have been reported in three publications (Pritchett and
Llewellyn, 1966; Pritchett, 1968; Humphreys and Pritchett, 1971). The
response in 1964, 3 to 5 years after establishment, showed that at equal
rates of P, OSP was superior to RP in all trials (Pritchett and Llewellyn,
1966). Increment growth during the next 3 years, 1964 to 1967, was still
greater from OSP than RP in all trials except on a Leon fs. Response
from the highest RP treatment, which contained eight times as much P as
the highest OSP treatment, was generally greater than that from the OSP
treatments (Pritchett, 1968). Humphreys and Pritchett (1971) reported
growth data only for the highest OSP and RP treatments which did not have
comparable P rates. However, 7 to 11 years after the fertilizer appl¡ca
tion,they found that in the three Spodosols with virtually no P-retention
capacity, all the P applied as OSP had been leached from the upper hori
zons, while most of that applied as RP was still retained in the surface
horizons. In the remaining soils, which ranged in P-retention capacity
from medium to very high, the majority of P applied either as OSP or RP
was retained in the surface horizons. In the most retentive soil
(Bladen), a considerable portion of the OSP had been converted to Fe-P,
a relatively unavailable form as reflected by the extremely low foliar P
levels reported for this treatment.
Humphreys and Pritchett (1971) suggested that selection of a suit
able P source for the acid Coastal Plain soils should be based on their

P-retention capacity. On soils with virtually no P retention (Spodosols)
they recommended the use of RP or other slowly soluble sources, whilst on
soils of low to intermediate retention capacity they suggested a mixture of
a soluble source, to provide initial high availability, and a slowly
soluble source to maintain the P supply in later years (5 to 15 years).
For soils with high P-retention capacity they suggested either repeated
applications of a soluble source or banded application of a suitable mix
ture of soluble and slowly soluble P sources.
Pritchett and Smith (197*0 recently reported the growth and response
data obtained 10 years after application of fertilizer in the experiment
on the site with the greatest P-retention capacity (Bladen fine sandy
loam). Their results showed the classical trend anticipated from a long
term comparison of a soluble and slowly soluble source on a highly P-
retentive acid soil: Initial response was greatest from OSP, but after 8
years the response from an equivalent rate of RP was equally great. Cur-
i
rent annual increment during the ninth year showed RP to have a consider
able advantage over OSP. This was reflected in the foliar P levels,
because only the trees in the RP treatments had foliar P levels above the
critical response range of 0.085-0.090%.
Although RP has been used successfully in many field fertilizer
trials, its use in operational fertilization has been restricted by its
bulk, which adds significantly to transportation costs, and the difficulty
in spreading these finely ground mater¡als--parti cu lar 1y by air (Pritchett
and Smith, 1969; Conway, 1962).
The success with RP in forest fertilizer trials is generally at
tributed to three factors, namely: (l) The acidity of most forest soils
(Gentle and Humphreys, 1968; Bengtson, 1970), (2) the relatively long

19
vegetative growth period and large root system of trees (Terman, 1968),
and (3) the presence of mycorrhizal rootlets on coniferous trees which
have been reported to increase the trees ability to utilize less soluble
forms of P (Bowen, 1973)-
Although not extensively tested, DAP has proved to be an effective
fertilizer in the Southeast, particularly on sites where responses to N
and P are synergistic (Pritchett and Smith, 197*0- This source is recom
mended for operational fertilization on these sites; however, its long
term effectiveness on soils of high P-retention capacity is likely to be
limited. Furthermore, White and Pritchett (1970) found that 5 years af
ter surface application of DAP to a Leon soil less than 11% of the applied
P remained in the surface 20 cm of this soil having a low P-retention
capacity.

20
Diagnostic Methods
This section of the review is concerned principally with the use
of foliar and soil analysis as diagnostic aids in predicting P deficien
cy and the need for fertilizer in pine plantations. Major emphasis is
given to the use of soil analysis since this is the method under inves
tigation in this study.
Foliar Analysis
The principles, physiological basis,and problems inherent in the
use of tissue analysis for diagnosing the nutritional status of crops
have been extensively reviewed by Goodall and Gregory (19^7) and Smith
(1962). The value and specific problems associated with the use of
foliar analysis in forestry have been the subject of several extensive
reviews (Tamm, 1964; Qureshi and Srivastava, 1966; Raupach, 1967;
Richards and Bevege, 1972 a; Armson, 1973; Leaf, 1973)-
Published information on foliar P concentrations in relation to
responses of slash and loblolly pines (P. el 1iottii and P. taeda) to P
fertilizer is summarized in Table 1. Whereas most research on the use
of foliar analysis in forestry has been concerned with relating foliar
nutrient levels to productivity or site index (Leaf, 1973). foliar
analysis research with Southern pine has been almost exclusively related
to establishing 'critical levels' for both predicting fertilizer needs
and following the effectiveness of fertilizer applications (Table 1).
The critical level of foliar P has been defined by Pritchett
(1968) as the point at which trees with a higher concentration of
needle P would not be expected to respond significantly to applications
of phosphate fertilizers, but at which trees with a lower needle P

2!
Table 1. Foliar P concentrations prior to fertilization in relation to
response of southern pines to P fertilizer
T ree
age
Foliar P
(Unferti 1ized)
Response to
P fertilizer
Reference
yr.
%
(A) Slash pine
(Pinus el 1iottii)
8
0.053
yes
Young (1948)
6
0.0148
yes
Baur (1959)
15
0.075
0.105
yes
yes
Pritchett and
Swinford (1961)
9
0.08
no
Walker and
Youngberg (1962)
3-5
<0.10
>0.10
yes
no
Pritchett and
Llewellyn (1967)
5-8 .
<0.09-0.10
>0.09-0.10
yes
no
Pritchett (1968)
1
0.08
no
Schultz (1969)
1-30
<0.075-0.80
>0.075-0.80
<0.085-0.10
>0.085-0.10
yes (90%)*
no (90%)
yes (100%)
no (100%)
Richards and
Bevege (1972b)
(B) Loblolly pine
(Pinus taeda)
8
0.064 .
yes
Young (1948)
6
0.057
yes
Baur (1959)
1
0.089
yes
Fowel1s and
Krauss (1959)
5-9
0.090
no
Zahner (1959)
5-10
0.137
no
Maki (I960)
22
0.124
no
Thompson (1960)

22
Table 1. Continued
T ree
age
Foliar P
(Unferti 1ized)
Response to
P ferti 1izer
Reference
yr.
%
1
<0. 1 1
>0.1 1
yes
no
Wells and
Crutchfield (1969)
10
0.11
no
Hoschler, Jones, and
Adams (1970)
1 -bO
<0.095-0.105
>0.095-0.105
<0.13-0.14
>0.13-0.14
yes (90%)*
no (90%)
yes (100%)
no (100%)
Richards and
Bevege (1972 b)
3
<0.10
>0.10
yes
no
Wells et al. (1973)
Associated with 90 or 100% of maximum yield.

23
concentration would normally respond. This definition corresponds to
the 'optimum concentration1 proposed fay Richards and Bevege (1972 a)
who defined the 'critical concentration' as the concentration asso
ciated with 90% of maximum yield.
Richards and Bevege (1972 a) pointed out that critical levels can
be established with accuracy only where all nutrients apart from the one
under consideration are not limiting. However, Leyton and Armson (1955)
suggested that difficulties in interpretation caused by nutrient inter
actions could be reduced by extending investigations over a wide range of
conditions to allow for a statistical analysis of the relationships. In
formation in Table 1 shows that where a number of trials and statistical
analyses were used, there is good agreement between the proposed critical
levels for slash pine (Pritchett, 1968; Richards and Bevege, 1972 b) and
loblolly pine (Wells and Crutchfield, 1969; Richards and Bevege, 1972 b;
Wells et al., 1973)- These data suggest that the best current estimates
of critical foliar P concentrations (Pritchett, 1968) for P. el 1iottii
and P. taeda are in the range 0.085-0.10% and 0.095-0.105% respectively.
The failure of foliar P values to provide the correct diagnosis in some
of the single trials reported in Table 1 can probably be attributed to
other factors such as other nutrients or site factors limiting production
or to differences in sampling procedures (Leaf, 1973).
In studies where both foliar and soil analysis have been used to
predict response to P fertilizer, foliar analysis has generally proved to
be the more effective (Pritchett, 1968; Wells and Crutchfield, 1969;
Wells et al., 1973). However, it has been pointed out that foliar analy
sis has practical disadvantages as a diagnostic tool in forestry. Its
use is largely limited to areas where trees are established (Pritchett,
(1968). Restricted sampling periods and the time required to collect an

2*<
adequate sample, even in small areas of mature stands, makes foliar anl
isis rather impracticable for large scale routine diagnosis (Wilde, 1958).
Richards and Bevege (1972 a) suggested that for management
purposes, plant analysis must be quantitative: They stated that
... at the very least it must provide some measure of the degree
of nutrient deficiency, while ideally it should enable us to
predict the magnitude of response to a given application of fer
tilizer. (p. HO)
The response curves shown by Pritchett (1968) are apparently the only
reported fully quantitative foliar analysis data for trees. A report by
Wells et al. (1973) related foliar P to a response index but provided no
information on the amount of fertilizer required to achieve a particular
response. Cain (1959) pointed out that the use of foliar analysis to
predict quantitative fertilizer requirements could be difficult. He
concluded this because the amount of fertilizer required to produce a
desired increment of concentration in plant tissue varies greatly from
site to site with variation caused by climatic and soil factors relating
to the movement and retention of nutrients in the soil.
Soil Analysis
Soil nutrient analysis has been used in forestry, both as a
technique to predict site productivity and as a diagnostic aid in deter
mining the need for fertilizer (Armson, 1973). Until recently the
emphasis in soil-site studies was on soil physical properties (Coile,
1952). This was commented on by Voigt (1958):
Students of forest soils commonly come away from the literature
with the impression that nearly all tree growth can be explained
almost completely by the so-called physical properties of the
soil particularly those related to its moisture regime, (p. 31)
Ralston (196*0 attributed the neglect of soil fertility parameters to
their frequent correlation with other soil properties commonly used in

25
site index studies, and the lack of soil testing techniques of known
significance in relation to tree requirements. However, since the wide
spread demonstration of the responsiveness of trees to fertilizers
(Mustanoja and Leaf, 1965) soil nutrient analysis has received more at
tention as a technique for predicting site productivity (Pawluk and
Arneman, 1961; Wilde et al., 1964 a; Schomaker and Rudoph, 1964; la
Bastide and van Goor, 1970; Alban, 1972).
As a diagnostic technique for predicting site nutrient status and
the need for fertilizer, soil analysis has received considerably less
attention than foliar analysis (Leaf, 1968). According to Leaf (1968),
current thinking on the value of soil and foliar analysis still reflects
the ideas of Mitchell and Chandler (1939) who stated:
The distinct advantage of the method of leaf analysis is that by
chemical analysis of the leaves we can obtain a more reliable
estimate of the amount of the various nutritional elements which
have been adsorbed by, and therefore available to, plants growing
in a given soil. One therefore uses a natural biological rather
than an artificial extraction method for estimating available
nutrients. (p. 75)
Several factors have been cited in the literature as reasons why
soil testing has not been widely accepted as a diagnostic tool by forest
managers. Most of these are related to insufficient information on
methods and procedures.
Sampling procedures
The determination of tree nutrient requirements is complicated by
a deep root system. Tamm (1964) suggested that the main rooting zone
should be sampled thoroughly and consideration should be given to lower
soil horizons when they contain notable root concentrations. In discus
sing this problem, Voigt (1958) pointed out that most forest trees, as
opposed to agricultural crops, have two distinct rooting regions; one

26
predominantly organic and the other predominantly inorganic. He sug
gested that it is doubtful if a single extraction agent could yield a
valid estimate of nutrient availability in view of the behavior differ
ences of organic and inorganic colloids in nutrient retention.
The value of sampling subsurface horizons has been examined in
many studies. Kessell and Stoate (1938) recognized the importance of
the vertical distribution of soil P in their site classification studies.
In outlining the P levels required for successful afforestation of
radiata pine (P.radiata D. Don), they wrote:
A P20c content of *00 parts per million (ppm) is required in the-
surface and subsurface soils. Three hundred parts per million
may be satisfactory if this content is maintained for a depth of
two to three feet. (p. 28)
In a survey of factors influencing the growth of radiata pine in New
Zealand, Jackson (1973) found that available soil P within the effective
rooting zone was more closely correlated with growth than with available
P in the surface 7-5 cm of soil. Similarly, White and Leaf (1964)
reported that the content of HNO^ extractable K in the upper 1.7 metres
of soil was more closely correlated with height and K content in the
biomas of red pine (P. resinosa AIT) than was the extractable K in the
surface horizon. The importance of subsurface nutrient supply was il
lustrated by the work of Ellerbe and Smith (1963) who found that
occasional serious underestimation of site quality by the soil-site
predictions of Coile (1952) in the lower Coastal Plain of South
Carolina could be attributed to the presence of underlying phosphate
marl at these sites. Will (1966) also reported that the disappearance
of Mg deficiency in radiate pine growing on volcanic soils in New
Zealand coincided with the root penetration of Mg-rich burred topsoil
at age 6 to 8 years.

27
Several investigators have reported P levels in the surface
horizon to be more closely related to growth and foliar P levels than P
levels in lower horizons. Humphreys (1963) outlined Australian work in
which it was reported that soil P below 37 cm contributed little to the P
supply of mature pines. Pawluk and Arneman (1961) found that available
P in the A horizon, but not the B horizon, was correlated with site
productivity of jack pine (P. banksiana Lamb). Similarly, Wells (1965)
reported that variation in concentration of foliar P of 5~year-old
loblolly pine was better explained by available P in the A than in the B
horizons. Alban (1972) also found that where P extracted from the 25 to
100 cm depth was included, it did not improve the estimation of the site
index of red pine in Minnesota over that obtained using P extracted from
the 0 to 25 cm depth. The apparently greater importance of P in the
surface horizon is attributed by most workers to the concentration of
fine feeder roots in the upper profile (Alban, 1972).
As in agriculture, horizontal soil variability is also a problem in
securing representative soil samples. The significance and extent of
variability of various soil properties in forest soils has been discussed
(York, 1959; Mader, 1963) and examined (McFee and Stone, 1965; Metz et al.,
1966). In reviewing these publications, Leaf (1968) pointed out that the
patterns and magnitudes of variability differ between soil characteris
tics at a particular site and determination of required sampling intensity
is a statistical problem.
In his review article, Leaf (1968) mentioned a problem in soil
sampling which is fairly unique when dealing with deep rooted perennials.
He stated:
...different degrees of biologically important soil heterogenity
exist on any site depending on the stage of tree development, e.g.,
seedling or establishiment stage vs. sapling stage vs. mature tree
stage, with the associated soil volume being tapped by the roots.
(p. 93)

23
This aspect has received iittie attention in the literature, but its
importance is well illustrated by the work of Will (1366) discussed
earlier.
Extraction methods
Tamm (1964) pointed out that most soil analysis techniques were
those developed for agricultural crops and are of unknown value for pre
dicting tree nutrient requirements. Most techniques used in agriculture
today for determining plant-available P are largely empirical (Williams,
1962). However, considerable research relating the availability to
agricultural plants of various soil-P fractions and determining the solu
bility of these same fractions in the extractants used in soil-P tests,
has provided a theoretical basis for these soil-P tests (Thomas and
Peaslee, 1973)- Little research of this nature has been conducted with
trees. However, knowing the forms of P extracted by various soil tests,
some insight into the forms of P utilized by trees can be obtained.
Total P. Early work in Australia indicated that total soil P
provided a good index of site productivity and adequately delineated
areas of P deficiency (Kessell and Stoate, 1938; Young, 1940 and 1948).
This work helped foster the concept that trees could utilize sources of P
unavailable to other plants. This ability has been attributed to the
presence of mycorrhizae (Pritchett, 1968) which have been shown to solu
bilize otherwise insoluble sources of P (Rosendahl 1943). However, Tamm
(1964) has suggested the ability of tree roots to utilize P unavailable
to agricultural crops could be the consequence not so much of higher ef
ficiency as of longer persistence.
Stoate (1950) found that certain anomalies occurred when using
total P for predicting site nutrient status. He reported that these

29
anomalies could be accounted for by differences in the level of P
extracted by ]% citric acid. Baur (1959) reported that total-P values
were related to site productivity only within limited localities and not
over more heterogenous sites, a finding substantiated by Humphreys (1964).
Similarly, Ballard (1970 a) found that total P was significantly corre
lated with productivity of radiata pine only on groups of genetically
similar soils. When all soils were grouped together, the Olsen
(0.5M NaHC03, pH 8.5) and Bray 2 (0.03N. NH^F + 0.1N HC1) tests for avail-
able-P provided the best index of productivity and foliar-P concentrations.
In the southeastern USA, Pritchett and Llewellyn (1966) found total P to.
be of little value in delineating sites responsive to P fertilizer. Wells
(1965) also found total P to be inferior to available P for predicting
foliar-P concentrations of loblolly pine.
Extractable P vs site productivity. Despite the many suggestions
that soil-test methods derived for agricultural purposes are unlikely to
be satisfactory for use in forestry, reports of their successful use in
forestry are frequent in recent literature. Pawluk and Arneman (1961)
reported a significant correlation between Bray 1 extractable P
(O.O 3 N_ NH/jF + 0.02 5jN HCI) in the A2 horizon and the site index of 50-year-
old jack pine in Minnesota. On acid soils, such as used in this study,
the Bray 1 test extracts P predominantly from the Al-P and Fe-P fractions
(Thomas and Peaslee, 1973). On non-phreatic sandy soils in Wisconsin
\
ranging in pH from 4.3 to 7-0, Wilde et al. (1964 a and 1964 b) reported
a significant correlation between P extracted by the Truog soil test
.(0.002N^ H2S0^ + 3g (NHt)2S0t/l) and the site index of both Jack pine and
red pine. The Truog test extracts P from the Ca-P and to a lesser extent
the Al-P fractions (Thomas and Peas lee, 1973).

30
Working with young plantations of loblolly pine in the lower
Coastal Plain of South Carolina, Wells and Crutchfield (1969) reported
that soil P extracted by 0.05N. HCI + 0.025N. H2S0^ was not significantly
correlated with the first year height growth. The authors suggested the
soil-test values would prove more reliable as the trees became older.
This was confirmed by Wells et al. (1973), who found soil P extracted by
the above extractant to be significantly correlated with the height of
unfertilized loblolly pine at age 3 years in the same region. They also
reported the Bray 2 extractant to be equally effective. According to
Thomas and Peaslee (1973), 0.05N_ HC I + 0.025N_ extracts P predomi
nantly from the Ca-P and Al-P fractions. However, the acid soils of the
southeastern lower Coastal Plain contain little Ca-P (Humphreys and
Pritchett, 1971). The effect of age of trees on the value of soil-P tests
as predictors of height growth was also observed by Ballard (197^). Work
ing with radiata pine on heavy textured acid soils in New Zealand, he
found that levels of soil P were most closely related to height at age 3,
i
followed by height at age 2 and least with height at age 1 year. The
Bray 2 and Olsen extractants (0.5M_ NaHCO^, pH 8.5) were the most success
ful in this study. On acid soils, the Olsen test extracts P from the
Al-P and Fe-P fractions, although with somewhat less intensity than
extractants,such as Bray,containing F (Thomas and Peaslee, 1973).
Extractable P vs foliar P. There are numerous reports in the lit
erature of correlations between foliar-P concentrations in pine species
and levels of soil P extracted by various soil tests; Wells (1965) found
P extracted by the Truog and Bray 2 tests to satisfactorily predict
foliar P of 5-year old loblolly pine growing in the South Carolina
Piedmont. However, Metz et al. (1966), in an intensive study on one soil

(Helena sandy loam) in South Carolina, found P extracted by the Truog
test accounted for only 17% of the variability of foliar P in loblolly
pine. Phosphorus extracted by the Olsen and Bray 2 tests was found to be
significantly correlated with foliar P of 40-year radiata pine (Ballard,
1970 a). In an intensive study involving 63 plots in Florida and 60 in
Australia, water-extractable P was found to provide the single best index
of foliar-P concentrations in slash pine (Humphreys and Pritchett, 1972).
However, the inclusion of NaOH-extractable P in their prediction equation
significantly improved the estimate of foliar P. The authors suggested
that both the intensity and quantity factors of soil-P supply are of im
portance in determining foliar-P concentrations. Alban (1972) also
reported that extractants which removed relatively small amounts of P from
the soil (H2O, 0.002f[ H^SO^, 0.01N_HCI) provided a better index of foliar-
P concentrations of 49-to 94-year-old red pine. Although Alban also
examined stronger extractants and found them less effective than the above
weaker extractants, he did not use multiple regression techniques to
examine the possibility that the combined use of a weak and strong
extractant may have improved the prediction, as did Humphreys and
Pritchett (1972). Using a greenhouse trial, Baker and Brendemuehl (1972)
found P extracted by NH^OAc (pH 4.8) correlated significantly with both
foliar-P concentrations and dry weights of sand pine (P. clausa Vasey)
seedlings growing on a Lakeland fine sand. This weak acid extractant
removes small quantities of P from the soil by forming weak complexes
with polyvalent cations (Al, Fe, Ca) in soils (Thomas and Peaslee, 1973)-
Pritchett and Llewellyn (1966) also found NH^OAc (pH 4.8) to extract P
that provided a good index of the growth of 3-to 5-year-old slash pine in
the absence of fertilizer. The trees were growing on acid sandy soils in
the lower Coastal Plain which varied in drainage characteristics from

32
excessively to poorly drained. In their study, P extracted by NH^OAc
(pH 4.8) was reported to be superior to that extracted by other extract
ants, including 0.03N_ NH^F + 0.025N_ HC 1 and 0.051 NCI + 0.025N H2S04, at
predicting response to P fertilizer. However significant correlations
were obtained only when the well-drained sites were eliminated from the
statistical analysis.
Forms of tree-available P. McKee (1973) concluded from a green
house fertilizer trial with N and P rates using slash pine grown in a
Caddo silt loam, that Ca-P and Al-P were the sources of P available to
the pine seedlings. This conclusion was based on the finding that the
combined value of these two fractions in the soil at the end of the ex
periment was significantly correlated with P uptake by the seedlings. He
found individual fractions of Ca-P, Fe-P, and Al-P were not related to P
uptake, nor was Fe-P in combination with either Ca-P or Al-P. The rele
vance of results from greenhouse trials to the nutrition of field grown
trees must however be questioned in view of the finding by Mead and
Pritchett (1971) that the response, and hence nutrition, of greenhouse
grown seedlings differed from that of field grown trees on the same
soils.
The literature reviewed in this section provides no conclusive
evidence as to what forms of soil P are utilized by trees and which soil
testing procedures are likely to be most successful in forestry. The
situation appears to parallel that in agriculture in which it has been

found that the value of any particular soil test depends upon the objec
tive in soil testing (predicting growth, nutrient uptake, or responsive
ness to fertilizer addition), the range of soils used, and the species
of plant involved (Williams, 1962; Williams and Knight, 1963). However,
despite some evidence that mycorrhizae may increase the availability of

33
insoluble sources of inorganic and organic P (Bowen, 1973), the evidence
reviewed above tends to indicate that extractant which remove plant-
available P are likely to be of more value than total P analyses for pre
dicting forest site fertility.
A problem in evaluating nutrient status of forest sites which has
received little attention was pointed out by Voigt (1958). He sug
gested that the perennial nature of trees is likely to present problems
in evaluating the .nutrient status since the rate of nutrient cycling
rather than the level of a particular fraction in the soil may be of im
portance in determining whether the trees' nutrient requirements are met.
For P, this is likely to be of particular importance in closed canopy
stands of trees, as Will (1964) reported that after canopy closure the P
requirements of radiata pine were met almost entirely by nutrient cycling.
However, he found that prior to canopy closure the tree-P requirements
were met almost entirely by net withdrawl from the soil. Ballard
(1970 a) suggested that the reason for the significant correlations he
found between P extracted by several available-P tests and foliar-P con
centrations of 40-year-old radiata pine was due to these methods reflect
ing the availability of organic P once it had been mineralized.
m
Interpreting soil test results
The successful use of soil tests depends entirely upon their prior
calibration against crop performance in the field (Williams, 1962). In
forestry, most studies have been concerned with calibrating soil-test
values against site productivity, their objective being to establish
critical levels below which growth performance of trees decline.
Prediction of productivity. Young (1940) found that a total-P
value in the topsoil of 47 ppm delineated areas of slash pine which showed

34
symptoms of fused needle in Australia. Young (1940) showed that fused
needle, a symptom of unthrifty pine trees, could be corrected by the ap
plication of P fertilizer. On the basis of significant correlations be
tween levels of P extracted by the Truog test and the productivity of
both jack and red pine, Wilde et al. (1964 a, 1964 b) proposed that
minimum P levels required for the establishment of these species were
6.5 and 5.4 ppm respectively in the surface of 15 cm of soil. Ballard
(1970 a) proposed that values of 3-5 and 5-0 ppm P by the Olsen and Bray
2 tests were required in the surface 10 cm of soil for satisfactory
growth of radiata pine.
Minimum or critical values established from relationships between
soil P and productivity cannot be used with confidence for predicting
ameliorative practices for sites testing below these values. Dahl,
Selmer-Anderssen, and Saether (1961) pointed out that a significant cor
relation between nutrient levels in the soil and site productivity is
not proof that this nutrient is controlling growth. For instance, these
authors reanalyzed the data of Viro (1955), who recorded a significant
correlation between EDTA-extractab1e P and the site index of Scots pine
(P. sy1 vestris). They found that extractable P was related to site
index only indirectly through its correlation with extractable soil Ca
levels. Proof of a casual relationship must be obtained by either field
fertilizer trials or the establishment of a positive relationship
between site productivity and both the concentration of a nutrient in
the foliage and its availability in the soil (Leyton and Armson, 1955).
Prediction of fertilizer response. Before soil-test results can
be used in developing fertilizer recommendations, field studies must show
that the test will be useful for differentiating soils into those groups

35
on which trees will respond and those on which trees will not respond to
fertilization (Wells et al., 1973). According to Mader (1973), the
paucity of well-designed fertilizer trials, suitable for use in calibrat
ing soil-test results against response over a wide range of conditions,
is the major reason for the lack of suitably calibrated soil tests. How
ever, this situation has been alleviated to some extent by tne establish
ment of a large number of uniform fertilizer trials by forest fertiliza
tion cooperatives based in Florida, North Carolina and Washington State
(Bengtson, 1972).
Information published on the response of southern pines in rela
tion to soil-P levels is summarized in Table 2. Early work In Australia
showed that total soil-P values could be used for delineating sites
responsive to P fertilizers (Young, 1948; Richards, 1956; Baur, 1959)-
However, as discussed earlier, these relationships were found to hold
true only over limited areas. This is exemplified in Table 2 by the dif
ferent critical values reported by the various authors. The only attempt
at using total-P values for other than delineating unresponsive sites was
that reported by Baur (1959). He found that the quantity of P fertilizer
required to produce adequate growth in the field corresponded closely
with the quantities calculated to raise total P to the optimum of 70 ppm
P (based on 125 kg/ha of superphosphate being equivalent to 6 ppm P). The
studies reported by Pritchett (1968) and Pritchett and Llewellyn (1966)
for slash pine, and Wells et al. (1973) for loblolly pine are the only
ones in which relationships were derived statistically using a range of
sites. These studies were conducted in the Coastal Plain region of the
southeastern USA. Pritchett and Llewellyn (1966) and Pritchett (1968)
reported that 2.0 ppm P by NH^OAc (pH 4.8) extraction separated sites
responsive to P fertilizer from unresponsive sites. This test proved

36
superior to several others examined, including 0.05N_ HC1 + 0.025^ I^SO^
and Bray 2. These authors also gave calibration curves indicating the
amount of fertilizer required to give a specific degree of response at a
particular soil-test value. Wells et al. (1973) reported that P
extracted by 0.05N HC1 + 0.025N_ was superior to that extracted by
the Bray 2 test at delineating responsive sites. From their response
curves, they proposed a value of 3.0 ppm P extracted by 0.05N_ HC1
+ 0.025N_ I^SO^ as providing the best separation of responsive and unre
sponsive sites. Their calibration curves indicated only the degree of
response expected at any particular soil-test value and not the amount
of fertilizer required to achieve this response. The critical level
found by Wells et al. (1973) for loblolly pine probably accounts for the
lack of response by loblolly pine to P recorded by Carter and Lyle (1966)
and Moschler et al. (1970) on soils testing above this critical level
(Table 2).
Wells et al. (1973) reported several sites which,although -
falling into the responsive category according to their soil-P test
value, did not respond to P fertilizer. They attributed this in most
cases to N being more limiting than P on these sites. This illustrates
the need in calibration studies, pointed out by Williams (1962) and
Richards and Bevege (1972), for utilizing field trials in which all nu
trients except the one under observation are nonlimiting to growth.
Wells et al. (1973) also pointed out that their critical value estab
lished for the response of 3-year-old loblolly, may not hold true for
older trees due to the increasing P requirements of trees with age. As
mentioned earlier, the importance of stand age or development to the
nutritional demands of trees was discussed by Leaf (1968) and illus
trated by the work of Will (1964).

37
Table 2. Soil-P values (surface horizon) of unfertilized soils in
relation to response of southern pines to P fertilizer
T ree
age
Soil test
type
Soi 1 P
Response to
P ferti 1izer
Reference
yr.
ppm
(A) Slash pine
(Pinus e11iottii)

9
Total P
<52
yes
Young (1948)
>52
no
21
Total P
<65
yes
Richards (1956)
>65
no
3+
Total P
<70
yes
Baur (1959)
>70
no
9
0.05N HC1 +
29.5
no
Walker and Youngberg
0.025N HjSOj,
(1962)
5-8
NH^OAc(pH 4.8)
<2.0
yes
Pritchett (1968)
>2.0
no
Pritchett and
Llewellyn (1966)
(B) Loblolly pine
(Pinus taeda)
9
Total P
<59
yes
Young (1948)
>59
no
21
Total P
<91
yes
Richards (1956)
>91
no
6
0.03N NH.F
7
no
Merrifield and Foil
+ 0.1NTHC1
(1967)
6
0.05N HC1 +
8
no
Carter and Lyle
0.025^1
(1966)
10
0.05N HC1
4.5
no
Moschler et al.
+ 0.025N H2S04
(1970)
3
0.05N HC1
<3-0
/yes
Wei 1s et a 1. (1973)
+ 0.025N H S0^
>3.0
no

33
Provided soil analysis results can be suitably calibrated against
tree growth and response to P fertilizer (and evidence suggests they can)
soil analysis has certain advantages over foliar analysis. Pritchett
(1968) outlined advantages for soil testing: (1) It can be used for pre
dicting fertilizer needs in areas prior to planting. (2) Collecting and
analyzing soil samples may be less laborious than collecting and analyz
ing needle samples, particularly in old stands. Ballard (1970 a) also
suggested that soil analysis has an advantage in that soil samples can
be collected at any time of the year, whereas foliage samples cannot.

MATERIALS AND METHODS
Introduction
A series of field, greenhouse and laboratory experiments designed
to examine the effectiveness of soil-test methods for predicting the need
for P fertilization of slash pine, included:
a) A preliminary screening of a wide range of soil-test methods to
relate soil-P values with tree height, height response to P frtil-'
izer and fertilizer requirements of slash pine in field and green
house experiments (greenhouse trial 1) on 10 soils. The most effec
tive of these test methods were then calibrated against tree growth
and response parameters from 72 field fertilizer trials.
b) The solubility of a range of P compounds in soil-test extractants was
related to the capacity of slash pine seedlings to utilize P from
these compounds in a greenhouse test (greenhouse trial 2).
c) A P-retention study was conducted in both the laboratory and field,
using soils that varied widely in their ability to retain fertilizer
P.
39

40
Field Trials
A series of uniform fertilizer trials was established on each of
29 sites in March 1968 by the Cooperative Research in Forest Fertiliza
tion (CRIFF) program. The sites were selected to represent the princi
pal forest soils of the Coastal Plain region. They included 22 soil
types in the Spodosol, Ultisol, Entisol and Inceptisol orders. Fertil
izer treatments were applied 1 to 2 months after planting of the sites
with slash pine seedlings. Twenty-four of the trials which were still
functional in 1972 were used in this study for evaluating soil testing
procedures.
Prior to planting, most sites were burned and disced to remove
residual vegetation. This was followed on all sites by ridging to form
beds approximately 1.22 m wide on 3-05 m centers on which the pine
seedlings were planted. The experimental design consisted of three re
plications of 12 fertilizer treatments in randomized complete blocks.
Each treatment plot was 27-5 by 30.5 m ( 0.08 ha) while the net measure
ment plot was 18.3 by 26.8 m. Nine of the twelve treatments formed a
3x3 factorial with N and P fertilizers. Application rates of both N
and P were 0 (Po, No), 22.5 (P1, Ni), and 90 (P2, N2) kg/ha. Addi
tional treatments numbered 10, 11, and 12 involved applications of K and
micronutrients but these treatments were not included in the present
study. Nitrogen was applied as ammonium nitrate and P as CSP. All fer
tilizer materials were applied in 1.22-m wide bands down tops of beds.
This effectively gave nutrient concentrations within the bed of 2.5
-times those shown above.
So?1 Sampling
Soil samples were collected from each replication at each site

41
prior to treatment by CRIFF personnel (Pritchett and Smith, 1972). The
samples were collected from undisturbed areas between bedded tree rows
from the 0-20, 20-40 and 40-60 cm depths.
Additional samples, specifically for use in this study, were col
lected in December-January of 1971/72, 4 years after the establishment
of the trials. From the control plot (PoNo) of each replication at each
site, four composite soil samples were collected using a 4-cm diameter,
closed cylinder soil auger. Two composite samples were collected from
the 0-20 cm depth--one from the bedded area and the other from the un
disturbed interbed area. Each sample consisted of 12 cores collected
randomly from within the net plot area. Four soil pits, one in each
quarter of the plot, were dug in the interbed area. During excavation,
two cores were taken at each pit from both the 20-40 and 40-60 cm depths.
The eight cores from each depth were combined to provide a composite
sample for each of these depths.
Bulk soil samples, for use in greenhouse trials (described
below), were collected from 10 trials. The trials were selected on the
basis of soil and P-response information (Pritchett and Smith, 1972) to
provide a wide range of soil-P levels and P-response characteristics.
The bulk samples were collected from the 0-20 cm depth from a control
plot in each of the 10 selected trials. The samples consisted of sub
samples collected randomly from the undisturbed interbed area.
In order to determine the extent of P leaching in the field, ad
ditional samples were collected from 10 selected trials in January-
February, 1973- The trials were selected, on the basis of laboratory
determinations of P-retention capacity, to provide sites with a range
of retention capacities. Samples were collected from the 0-20 cm depth
in the PiNo and P2N0 plots of each replicate. The samples were

collected from the bedded area using the same equipment and procedure as
used in the collection of the samples from control plots.
Following excavation of the four soil pits in each control plot,
the following information was recorded: Depth and color of the A1
horizon; and depth to mottling, spodic horizon, or fine-textured layer
where they occurred within the surface 90 cm. The experimental sites
were classified within one of five drainage classes by CRIFF personnel at
the time of establishment (Pritchett and Smith, 1972).
All samples were air-dried after collection. Samples for soil
analysis were passed through a 2-mm sieve. Where analysis procedures re
quired it, subsamples were ground to pass a 0.2-mm sieve. Bulk samples
used in greenhouse trials were screened through a 6-mm sieve while sub
samples for soil analysis were passed through a 2-mm sieve.
Foliage Sampling
Foliage samples were collected in December-January of 1971/72 from
the control plot of each of the three replicates at each site. Each
sample consisted of a composite of needles collected from a minimum of
five trees within the net plot. The trees were selected to represent the
range in tree size and vigor found in the plot. Needles were taken from
the previous spring flush on the uppermost whorl bearing secondary
branches. The foliage samples were stored on ice while in transit to the
laboratory. All samples were dried at 70C and ground to pass a 1-mm
sieve in a stainless steel Wiley mill.
Growth and Response Parameters
Heights of all living trees were determined by CRIFF personnel at
the end of the first, second, third and fifth growing seasons. The
measurements taken after the second growing season were not used in this

I
*3
study. Responses recorded after three growing seasons were reported by
Pritchett and Smith (1972).
This study was concerned with evaluating the ability of soil test
ing procedures to predict (a) tree height growth in the absence of P
fertilizer, (b) height response to P fertilizer, and (c) the amount of
fertilizer required to obtain the optimum response. Since N was a limit
ing factor on many of the sites and N x P interactions were common
(Pritchett and Smith, 1972), procedures for computing growth and P-
response parameters were designed to adjust for the N effect. This was
deemed necessary as meaningful relationships between growth and/or re
sponse and a particular nutrient level can only be obtained if the nu
trient in question is the only one limiting growth (Williams, 1962;
Richards and Bevege, 1972 a).
The index of growth in the absence of P fertilizer was taken as
the tallest mean height of the Po Ni treatments, where i = o, l, or 2.
These values were obtained for each replicate at each site for growth
periods of 1, 3, and 5 years.
Relative tree height was used as the index of response to P
fertilizer. This was calculated from:
Relative height = MeaP. height of Po_Nj. treatment x 100>
Maximum height from P addition
Maximum height from P addition, adjusting for the effect of N, was pre
dicted by first fitting response curves to the height data using a second
degree polynomial equation,
Y = a + bX + cX2
where Y = tallest mean height in the N treatments, X = P-application
rate, and a, b, and c are constants. The predicted maximum height was
then obtained by first differentiating the quadratic equation to give

kk
dY
dX
dY
= b + 2cX,
then solving X for = 0, to give
b + 2cX = 0, and by rearranging
x = ik
2c
We obtained the value of Y corresponding to this value of X from the
original quadratic equation by substitution
Y = a + b
-b'
-b
2c
+ c
2c
, and rearranging
Y = a which is maximum tree height.
Relative height was computed for each replicate at each site for growth
periods of 1, 3, and 5 years. In some cases it was not possible to
obtain a predicted maximum height in this manner, because the response
was either linear or increased exponentially with increasing rates of P.
In these cases, the maximum height was taken as the actual mean height
in the tallest P i N i treatment. Where no increase in height occurred fol
lowing P application, relative heights were recorded as 100%.
The amount of P fertilizer required to achieve maximum height was
obtained by differentiating the quadratic equation, setting the deriva
tive equal to zero and solving for X (this = -b/2c). In addition, the
amounts of fertilizer required to achieve 90 and 95% of maximum height
were also computed. These values were obtained by substituting 90 and
95% of maximum tree height values in the quadratic equation and solving
for X. In cases where actual rather than predicted maximum tree heights
were used, P fertilizer required for maximum height was taken as the
actual P rate of the treatment that produced the maximum height.

45
Greenhouse Trial 1
This greenhouse trial was established in order to obtain growth and
P response information for use in the preliminary screening of the effec
tiveness of soil-test methods. This screening was necessary as it was not
practical to evaluate a large number of soil P-test using the soils from
all 72 sites. Only the most successful methods in the preliminary screen
ing were used in the final evaluation employing soils and growth informa
tion from all si tes.
Establishment
The greenhouse trial with the 10 bulk soil samples was established
in March 1972. Eight kilograms of air-dried, sieved soil were placed in
21-cm diameter, 7-6-1 iter glazed pots. Single drainage holes near the
base of each pot viere filled with porous fibre glass plugs. Sufficient
pots of each soil were prepared to give three replicates of four P rates.
One additional pot of each soil was prepared for use in determining root:
shoot ratios at the first harvest, as explained later. The P treatments
were applied to the surface of the potted soils as CSP to emulate field
applicationon a surface-area basis at 0, 56.3, 112.5, and 225 kg P/ha.
The first two and last application rates were identical to the P applica
tion rates, per unit surface area, as used in the field trials. To
insure P was the only nutrient limiting to growth, all pots received a
basal application of KCI (225 kg K/ha), micronutrient frit (140.6 kg FTE
503/ha), and NH^NO^ (225 kg N/ha). Both the KCI and FTE 503 were mixed
with the soil at potting, while the NH^NO^ was added as a dilute aqueous
solution in three equal applications, 1, 5,and 10 months after planting
of slash pine seeds. Phosphorus treatments were assigned at random to
the 10 extra pots. All P treatments were applied 1 month after planting
of seeds.

46
Thirty slash pine seeds, previously soaked in 0.1% citric acid for
24 hours to facilitate germination, were planted in each pot. Seedlings
were thinned by stages to eight seedlings per pot 2 months after planting.
Soil moisture in the pots was maintained between 50 and 100% of field ca
pacity by watering to predetermined weight with distilled water; no
leaching losses were observed. Heating and cooling equipment in the
greenhouse maintained the temperature between 24C and 35C. Day length was
not controlled during the first 9 months, but following the first harvest
in December of 1972, day length was increased to 15 hours using low-
intensity incandescent light. The position of the pots on the greenhouse
benches was altered periodically to minimize micro-environment effects.
Harvesting
In December of 1972, seedling heights from root collar to terminal
bud were measured to the nearest millimeter. Four seedlings per pot,
with an average height approximating that of all eight seedlings, were
harvested removing the tops at root-collar level. The roots of the
harvested seedlings were left undisturbed in the soil. Root: shoot rela
tionships at the first harvest were established by harvesting the entire
plants from the extra pot of each soil included in the trial.
Root harvests in these extra pots were made by washing the soil
and root mass on a 6-mm sieve followed by a rinsing of the recovered roots
in distilled water. A final harvest of the remaining four seedlings per
pot was made in October, 1973, following height measurements. Tops and
roots were harvested separately. Old roots left from the first harvest
were separated and discarded during sieving. All tissues samples were
dried at 70C, weighed, and ground to pass a 1-mm sieve using a stainless
steel Wiley mill.

hi
Growth and Response Parameters
In computing the growth and response parameters for the greenhouse
trial, no adjustment for the N effect was required, since all treatments
received a uniform N application. Mean height in pots not receiving P
was taken as index of growth in the absence of P fertilizer. Relative
height and P-fertilizer requirements were computed in a similar manner to
those for the field trials using a response curve fitted with a quadratic
equation. Growth and response parameters were computed from the data
obtained at each harvest. In order to compare the growth and response
parameters from the greenhouse study with those of the field trials grow
ing on the same soils, the field parameters for the 10 trials were
computed using P0N2, PiN2,and P2N2 treatments only. The N2 treatment in
the field was the same rate, per unit surface area, as that used in the
greenhouse trial. In addition, the parameters were computed using all
three replicates at each of the 10 sites, rather than just the replicate
from which the bulk soil sample was collected. However, for 3 of the 10
i
sites only the replicate from which the sample was collected was used,
because soil properties and response information indicated substantial
variation between replicates on these three sites.
Other parameters determined were the concentration of P in the
seedling tops, and total uptake of P (mg/pot). Since only the tops were
harvested at the end of the first growth season, total P uptake at this
stage was determined using the relationship between quantity of P in tops
and total P in tops and roots computed from the 10 extra pots.

48
Greenhouse Trial 2
This trial was established for the purpose of examining the abili
ty of slash pine seedlings to utilize different P compounds. Uptake of P
was determined for 6-month-old slash pine seedlings growing in two soils
to which the P compounds were added. This P uptake was related to the
solubility of the P compounds, alone or after mixing with the above two
soils, in chemical extracts used as soil-P tests.
P Compounds
The eight P compounds used in the trial, and their chemical compo
sition, are shown in Table 3- The composition of all compounds was
checked by determining Ca, Fe, A1 and P in solution following dissolution
of the compounds in hot 6N HC1. Procedures used in the chemical analysis
are outlined in a later section. The compounds were also checked by X-
ray analysis using a General Electric XRU-7 instrument with Ni-filtered
CuKa radiation. The principal peaks of the six crystalline compounds,
shown in Table 3 confirmed their identification according to published
standards (Lehr et al., 1967). The colloidal A1 and Fe phosphates were
prepared by the procedures outlined by Deming and Cate (1963) and Cate,
Huffman, and Deming (1959), respectively. The X-ray analysis confirmed
that these prepared compounds were amorphous.
Monocalcium phosphate is the principal P form in most phosphatic
fertilizers. Dicalcium phosphate, colloidal Al and Fe phosphates, K
taranakite, and strengite were used in this study because they have all
been identified as soil-P fertilizer reaction products (Lindsay, Frazier,
and Stephenson, 1962). Fluorapatite is a common primary phosphate
mineral in some soils and may also be formed as a soil-P fertilizer
reaction product following reversion of less basic Ca phosphates in the

49
presence of fluoride (Olsen and Flowerday, 1970* Wavellite has not been
identified as a soil-P fertilizer reaction product, but is naturally oc
curring mineral which was included as a representative of insoluble
stable A1 phosphates.
Establishment
Six-month-old slash pine seedlings were used in this greenhouse
trial. The seedlings were grown from half-sibling seeds in the greenhouse.
Ten seeds were sown in each of 35 closed plastic pots containing 1,750 g
of thoroughly mixed Ona fine sand (0-20 cm) collected from CRIFF site
A23, known to be responsive to both N and P (Pritchett and Smith, 1972).
Fertilizer was not added to these pots. Soil moisture was maintained be
tween 50 and 100% of field capacity by watering to predetermined weight
with distilled water. Day length was maintained at 14 hours using low-
intensity incandescent light. Seedlings were thinned to provide five
uniform seedlings in each pot at 6 months. At this stage, the seedlings
were removed from the pots and the roots cleansed of the sandy soil by
gentle agitation under water. The roots were then rinsed in distilled
water prior to transplanting immediately into the potted soils used in
the experiment. The roots of all seedlings were observed to be heavily
infected with mycorrhizal fungi.
The experimental design used was a randomized block design, with
two blocks each with nine treatments. The blocks consisted of two soils,
selected from the 10 bulk samples to provide soils of widely different
P-retention capacity and pH. Properties of the two soils, an Immokalee
fine sand from CRIFF site A16 and a McLaurin fine sandy loam from CRIFF
site A28, are shown in Table 4. Treatments consisted of a control, to
which amendments were not added, and eight P treatments, using the P

50
Table 3. Properties of phosphorus sources used in greenhouse trial 2
Source
Composition
Principal
X-ray peaks
P
Ca
A1
Fe
(%)
(ft
Monocalcium phosphate*
24.6
15.9
-
-
11.74,
3-88,
3.69
D¡calcium phosphate*
22.8
29.5
-
-
3.35,
2.36,
2.72
FI uorapatite +
18.0
39.7
-
-
3.44,
2.80,
2.70
Colloidal aluminum phosphate
17.9
-
14.4
-
-
Potassium taranakitef
19.0
-
10.1
-
15.84,
7.91*
3.81
Wavel1 ite §
10.8
-
16.0
-
8.71,
8.42,
3.22
Colloidal ferric phosphate
13.9
-
-
27.1
-
Strengite+
17.0
-
-
35.5
5.49,
4.37,
3.01
* Analytical grade reagents.
t Provided by coirtesyof the Tennessee Valley Authority.
§ Provided by courtesy of Dr. F.N. Blanchard, Geology Dept., University
of Florida.

51
compounds shown in Table 3- Each treatment was replicated three times
for the Immokalee soil but only twice for the McLaurin soil because it
was in short supply.
Closed plastic pots containing 1,500 g of soil were used. The P
compounds were applied at a rate designed to raise the total P content of
the soil by 100 ppm. Required quantities of each compound, all previous
ly ground to pass a 0.0105_mm sieve, were thoroughly mixed, throughout
the soil for each pot. Three carefully graded seedlings were transplant
ed into each pot. Moisture conditions and day length were maintained as
outlined above.
Harvesting
Eight months after transplanting, the heights of all seedlings
were recorded. Following height determinations, the seedlings were
harvested using the procedure outlined for greenhouse trial 1. The tops
and roots from each pot were dried at 70C, weighed, and ground to pass a
1-mm sieve. Phosphorus uptake per pot was determined as the combined
product of P concentration in the tops and roots and their corresponding
dry weights.
Extraction of P Compounds
Each of the eight P compounds (Table 3) was added at a rate of
100 ppm P to duplicate 100-g samples of each of the two soils. All samples,
including control soil samples, were thoroughly mixed by passing through a
0.5-mm sieve several times. One of the duplicate sets was incubated at
room temperature in the dark for 2 months, with moisture being maintained
at field capacity by the addition of distilled water. Following this in
cubation period, the samples were air-dried and passed through a 0.5-mm
sieve.

52
Table 4. Physical and chemical properties of soils used in greenhouse
trial 2
Property
Soil type
Immokalee fs McLaurin fsl
pH (1:2, H20)
4.2
4.9
pH (1 N_ KC 1)
3.0
3.8
Clay, %
1.9
9.8
Silt, %
5.6
28.7
Organic matter, %
3.3
2.2
CEC, meg/100g
5.5
li 7
NH/jOAc (pH 4.8) extractable
Ca, ppm
164
104
P, ppm
6.3
0. 4
0.3 M NH^C20^ (pH 3-0) extractable
A1, ppm
25
1025
Fe, ppm
60
5^5
P retention,* yg P/g soil
30
750
* Equilibration with 2,500 yg P/g soil for 6 days.

53
Several soil-test methods, details of which are given later, were
used to extract P from subsampies of the above incubated and non-
incubated mixtures. In addition, the extractabi 1ity of the P compounds
in the soil-test extractants was also determined. Amounts of P
compounds used for the extraction in the absence of soil were calculated
to provide the same amount of P per unit of extractant as that in the
corresponding soil mixture.

54
Phosphorus-Retention Study
Phosphorus-retention characteristics of a range of forest soils
were determined in the laboratory and related to various soil properties.
The value of using the concentration of various elements in the foliage
of slash pine to predict the P-retention capacity of forest soils was
also examined. Parameters found to correlate closely with laboratory
determined P retention were examined for their ability to predict leach
ing losses of P fertilizers in the field.
Soil and Foliage Samples
A total of 42 surface (0-20 cm) soil samples was used in the lab
oratory phase of this study. Twenty-four of the samples were those
collected from the bedded area of the replicate 2 control plots in the
NP field fertilizer trials. The 10 bulk samples were also included.
The eight remaining samples were collected from other uncultivated
sites in Florida and used in this study to increase the range of soil
types.
Foliage samples were those collected from replicate 2 control
plots and control plots from which bulk samples had been collected.
Foliage samples were not available from the above eight remaining
s i tes.
Determination of P Retention
Phosphate retention characteristics were determined by equilib
rating 5"9, air-dry soil samples with 25 ml of 0.0M CaCl2 containing
Ca(^PO/j)2.H2O at concentrations ranging from 0 to 500 ppm P
(0-2500 yg P/g soil). Two drops of toluene were added to inhibit micro
bial activity and the samples were equilibrated for 6 days at 25C with

s
55
intermittent shaking. Following centrifugation, P was determined in
solution. Phosphorus-adsorption isotherms were plotted for yg P
adsorbed/g soil against yg P/ml in equilibrium solution. From these
plots, values were interpolated to fit the linear form of the Langmuir
adsorption isotherm equation (Olisen and Watanabe, 1957)
where C = equilibrium P concentration, x/m = amount of P adsorbed per
unit weight of soil, k = a constant related to binding energy, and
b = P-adsorption maximum.
Significant deviations from the Langmuir equation have been shown
to occur at equilibrium values in excess of 10-20 ppm P (Fox and
Kamprath, 1970). Also the Langmuir maximum, b, depends to a certain
extent on the maximum equilibration value included in the computation
(Gunary, 1970). Therefore, interpolated values were restricted to cases
where equilibration values were 10 ppm or less with 10 ppm being
X
included as the top value for all soils. The Langmuir adsorption
maximum was obtained from the reciprocal of the slope following use of
i
linear regression analysis to obtain the Langmuir equation from the in
terpolated data.
In addition to the Langmuir maximum, the amount of P adsorbed
from the highest level of application (2,500 yg P/g soil) was also used
as an index of the P-retention capacity of these soils.
Retention of field-applied P was determined from the 0-20 cm
samples collected from the Po, Pi, and P2 plots of each replicate of the
10 field trials selected for this purpose. The samples were analyzed
for total P and this value converted to kg P/ha in the surface 20 cm
using the bulk density of the samples. The amount of P retained in the

56
surface 20 cms, *4 years after fertilizer application, was determined by
difference in total P (kg P/ha) between control plots and plots which
had received 56 or 22*4 kg P/ha. This P-retention value was then
expressed as a percentage of the P applied.

57
Sample Analysis
Soil Characterization
Bulk density of all soil samples collected using a closed-cylinder
soil auger was computed from the dry weight of the samples and their
volume. The volume was calculated from the diameter of the auger, the
depth of sampling and number of cores. Particle-size distribution was
determined by the hydrometer method (Bouyoucus, 1951). Moisture contents
at 15, 1/3, and 1/10 atmospheres were determined by use of a pressure-
membrane apparatus (Richards, 1965). Loss on ignition was determined as
the weight loss (%) of an air-dried sample following ignition in a
furnace at 550C for 1 hour.
Soil pH was determined potent iometrica1ly by insertion of a com
bination glass electrode assembly in the supernatant of a 1:2 soil-water
and/or soil-N^ KC1 suspension. Soil organic matter was determined by a
modified Walkley-Black method (Allison, 1965); total N by the macro-
Kjeldahl procedure (Bremner, 1965); and cation exchange capacity (CEC)
by NH^ saturation using N_ NH/jOAc at pH 7-0 (Chapman, 1965). Extract-
able cations were determined in the filtrate following extraction with
0.7^NH^0Ac + 0.5^N_ HOAc buffered at pH k.8. Calcium and Mg in the
filtrates were determined by atomic absorption (Perkin-Elmer 303) and K
by flame emission (Beckman B) spectroscopy. Lanthanum chloride was added
to suppress anion interference in the Ca and Mg determinations (Breland,
1966). These analyses were carried out by the Analytical Research Lab
oratory of the Soil Science Department, University of Florida.
Soil P Analysis
Reagents used to extract soil P, and details of the extraction

58
procedures are given in Table 5. Unless mentioned to the contrary below,
all procedures involved extraction of air-dry samples (<2mm) on a recip
rocating shaker followed by filtration and the determination of P using
an aliquot from the filtrate. The P extracted was expressed as ppm on a
soil basis.
Saturation extracts were obtained by placing lOOg of air-dry soil
in a Nalgene filter assembly (0.20-y membrane). A measured quantity of
distilled was added to produce saturation, and the soil thoroughly
stirred to remove all air bubbles. After 1 hour, the solution was
filtered with a suction of 300 mm Hg. The volume of filtrate was record
ed and an equivalent amount of distilled i^O was returned to the soil,
stirred, and again filtered as before an hour later. This procedure was
repeated a total of 10 times. Phosphorus was determined on each
filtrate and expressed as ppm P in soil solution at 20% moisture content.
Values for the first and tenth extracts were used as soil-P parameters,
as was the total yg of P extracted by all 10 extractions. In addition,
a measure of the soil capacity to buffer solution P levels against de
pletion was obtained by dividing the difference between the first and
tenth extracts into the first extract value. This parameter will be
referred to as Capacity(l). Two measurements were obtained of the soil
capacity to buffer against P additions; a parameter reported as signifi
cant in determining fertilizer requirements of trees (Humphreys and
Pritchett, 1970 and agricultural crops (Ozanne and Shaw, 1968). These
two parameters, determined from the P-adsorption isotherms, were the
amount of P addition required (yg/g of soil) to increase the solution
equilibrium P concentrations to 0.3 yg P/ml and 3.0 yg P/ml. These two
parameters will be referred to as Capacity(2) and Capacity(3) respective
ly.

59
Table 5. Phosphorus extraction methods
Soil: Extract ion
Method* Extractant pH solution time
H20
h2o
7.0
1:5
min
30
H20(2)
h2o
saturated
paste
( 1st extract)
H20(3)
h2o
saturated
paste
(10th extract)
H20(4)
h2o
saturated
paste
(E10 extracts)
NaCl
0.01M NaCl/HClt
4.0
1:5
30
Na2S04
0.01M Na2SO4/H2SOi+
4.0
1:5
30
Na2MoO^
0.01M NaoMoO^/HCl
4.0
1 :5
30
Na2B47
0.01M Na2B^07/HCl
4.0
1:5
30
Na2B407(2)
0.01M Na2B407/Na0H
10.0
1:5
30
NH^OAc
0.7N NH^OAc + 0.54N_ HOAc
4.8
1:5
30
NH40Ac(2)
IN HOAc/NH^OH
3.8
1:5
30~
NH4OAc(3)
IN HOAc/NH^OH
2.8
1:5
30
HOAc
2.5% HOAc (v/v)
2.5
1:40
120
HOAc(2)
II
2.5
1:40
30
H0Ac(3)
II
2.5
1:5
120
HOAc(4)
II
2.5
1:5
30
Lactate
0.1N CH3CHOHCOONH4 + O.kH
HOAc 3-5
1:20
240
Lactate(2)
II
3.5
1:20
30
Lactate(3)
1 1
3-5
1:5
240
Lactate(4)
II
3-5
1:5
30
Citrate
1 % Citric acid (w/v)
2.2
1:10
24(hr)
Citrate(2)
II
2.2
1:5
30

60
Table 5. Continued
Method*
Extractant
pH
Soil: Extract
solution time
min
Oxalate
0.2M (NH4)2C204 + 0.1M H2C204
3.0
1 :50
60
T ruog
0.002N_ H2S04/(NH4)2S04
3.0
1:100.
30
Truog(2)
1 1
3.0
1 :20
30
T ruog(3)
1 1
3.0
1:5
30
T ruog(A)
0.02N^ H2S04/(NH4)2S04
2.1
1:100
30
T ruog(5)
1 1
2.1
1:5
30
Truog(6)
II
2.1
1:5
30
Truog(7)
0.2N H2S04/(NH4)2S04
1.1
1 : 100
30
T ruog(8)
II
1.1
1:20
30
Truog(9)
II
1.1
1:5
30
Truog(10)
0.02N^ H2S04/Na2Mo04
2.5
1:100
30
T ruog(11)
1 1
2.5
1:20
30
Truog(12)
II
2.5
1:5
30
h2so4
0.002h[ H2S04
2.7
1:100
30
H2S04(2)
1 1
2.7
1:20
30
h2so4(3)
II
2.7
1:15
30
h2so4(A)
0 -02N^ H2S04
1.7
1:100
30
h2so4(5)
II
1.7
1:20
30
h2so4(6)
II
1.7
1:5
30
h2so4(7)
0.2NI H2S04
0.8
1:100
30
h2so4(8)
II
0.8
1 :20
30
h2so4(9)
II
0.8
1:5
30

61
Table 5. Continued
Method*
Extractant
pH
Soil:
solution
Extraction
t i me
min
hci-h2so4
0.5N HC1 + 0.025N_ H2S04
1.3
1 :4
5
01 sen(2)
0.5M NaHCO^NH^OH
8.5
1:5
30
01 sen
1 1
8.5
1 :20
30
01sen(3)
II
8.5
1:50
30
01sen(4)
II
8.5
1:5
240
01 sen(5)
8.5
1 :20
240
01 sen(6)
II
8.5
1:50
240
01sen(7)
II
8.5
1:5
l6(hr)
01sen(8)
II
8.5
1:20
I6(hr)
01sen(9)
II
8.5
1:50
l6(hr)
Bray 2(2)
0.03N NH^F + 0.1N_ HC1
1.5
1.5
1
Bray 2(3)
1 1
1.5
1.5
30
Bray 2
II
1.5
1:10
1
Bray 2(4)
II
1.5
1:10
30
Bray 2(5)
1 1
1.5
1 :50
1
Bray 2(6)
II
1.5
1:50
30
Bray 1(2)
0.03N. NH^F + 0.025N. HC1
2.5
1:5
1
Bray 1(3)
II
2.5
1:5
30
Bray 1
II
2.5
1: 10
1
Bray 1(4)
II
2.5
1:10
30
Bray 1(5)
II
2.5
1:50
1

62
Table 5. Continued
Method*
Extractant
pH
Soil:
solution
Extraction
time
#
min
Bray 1(6)
0.03N NH^F + 0.025JN HCl
2.5
1:50
30
Bray 3
0.03N NH^F + 0.0125N_ HCl
4.1
1:10
1
Bray 3(2)
1 1
4.1
1:50
1
Bray 4
0.03N NH^F + 0.005N_ HCl
4.6
1:10
1
Bray 4(2)
II
4.6
1 :50
1
Bray 5
0.1 NH^F + 0.1 N_ HCl
1.8
1:10
1
Bray 5(2)
II
1.8
1:50
1
Bray 6-
0.1N_ NH^F + 0.025N. HCl
2.9
1:10
1
Bray 6(2)
1 1
2.9
1 :50
1
Bray 7
0.01 NH^F + 0. 1 N_ HCl
1.3
1 :10
1
Bray 7(2)
II
1.3
1 :50
1
Bray 8
0.01N NH^F + 0.025 HCl
2.4
1:10
1
Bray 8(2)
1 1
-3"

CM
1:50
1
HC1
0.1 N_ HCl
1.1
1:5
30
HCl(2)
1 1
1.1
1:10
1
HC1(3)
1 1
1.1
1:50
1
HCl(4)
0.025N^ HCl
1.7
1:10
1
HCl(5)
II
1.7
1:50
1
HCl(6)
0.0125N HCl
2.0
1:10
1

63
Table 5- Continued
Method*
Extractant
pH
Soil:
solution
Extract ion
t i me
min
HC1(7)
0.025N_ HC1
2.0
1 :50
1
HC1(8)
0.005N. HC1
2.3
1 :10
1
HC1(9)
1 1
2.3
1 :50
1
Res i n
Amber lite 1RA-^00 resin (2g)
2g soil
+ 25ml H20
1
Res in(2)
1 1
2g soil
+ 25ml H20
60
Res in(3)
II
2g soil
+ 25ml H20
Mhr)
Resin(4)
II
2g soil
+ 25ml H20
16 (h r)
Resin(5)
II
5g soil
+ 3ml H20
6(days)
* In the
text and tables the methods are
referred
to by these
abbrevia-
tions.
t Reagents following a slash mark were used to adjust the pH of the
reagents preceding the slash to that shown in the pH column.

Resin-extractable P was determined using a strongly basic,
quaternary ammonium, anion-exchange resin RN(CH^)^ Cl The soil was
ground to pass a 0.25mm-sieve and all resin exceeded this sieve size.
Following the soil extraction, the soil and resin were separated by wash
ing the soil-resin mixture with distilled water over the 0.25mm sieve.
The resin was quantitatively transferred to a filtering apparatus and
leached successively with five 5_ml portions of 2N^ NaOH followed by five
5-ml portions of 2N_ HC1. This procedure was found in preliminary work
to recover greater than 90% of the P sorbed by the resin. Phosphorus was
determined in this leachate.
Soil P was fractionated by a modified Chang and Jackson procedure
developed by Fife (Ballard, 1970 b). Fractions determined included
soluble-P, Al-P, Fe-P, Ca-P and organic P. Total P was extracted by
Na2C0^ fusion (Jackson, 1958). Dispersed material present in alkaline
extracts was removed by addition of P-free activated carbon at the
filtering stage.
Phosphorus in all solutions was determined color¡metrically by the
Murphy and Riley (1962) technique using ascorbic acid and molybdate-
sulphuric acid as modified by Watanabe and Olsen (1965). All measure
ments were made using a Unicam SP 600 spectrophotometer with a wavelength
setting of 880 my. Aliquots taken for developments were adjusted to pH 5
with H2S0^, using para-nitro phenol indicator, prior to the addition of
the developing reagent.
Extractants containing NH/jOAc and HOAc required adjustment of the
aliquote to pH 2.5 with HjSO^, using 2,4 dinitrophenol indicator, in
order to obtain color development. Fluoride interference in the extracts
containing NH^F was eliminated by addition of saturated boric acid
(Kurtz, 1942). Oxalate and citrate were found to interfere with color

65
development; an aliquot was taken down to dryness on a hot plate and
ignited in a furnace at 5Q0C for 30 minutes and the residue taken up in a
known amount of 0.1N_NC1 and an aliquot taken from these solutions for P
determination. Concentrations of other potentially interfering ions were
all below the interference level, as indicated by John (1970) or as
established in preliminary work.
Soil A1 and Fe Analysis
Soil A1 and Fe were extracted by several extractants. These
included CDB using the procedure of Mehra and Jackson (I960), 0.3M_
(NH^)2C20i, (Saunders, 1965), 0.1M Na/^Oy (McKeague, 1967), and 0.05M
EDTA (Viro, 1955)- Exchangeable A1 was also extracted by leaching with
1N_ KC1 (Yuan, 1965). Aluminum and Fe were also determined in the
extracts of several of the P-extraction methods outlined in Table 3-
These methods included NH^OAc, HCl-h^SO^, Bray 2, Bray 1, Bray 1(3). and
HC1.
Aluminum and Fe in all extracts except KC1 were determined by
atomic adsorption at the Soil Science Analytical Research Laboratory,
Universtiy of Florida (Yuan and Breland, 1969). Pyrophosphate extracts
were digested in HNO^-^SO^ and oxalate extracts ignited at 500C and
taken up in 0.1N_ HC1 before atomic absorption analysis. Aluminum in the
KC1 extracts was determined color¡metrically by the aluminon procedure
(Yuan and Fiskell, 1959).
Plant Tissue Analysis
One-gram samples of plant material were ashed at 480C for 5 hours.
The ash was dissolved in 6N_ HC1, digested on a hot plate for 15 minutes
and quantitatively transferred with distilled H20 to a 50-ml volumetric
flask. Suitable aliquots for elemental analysis were obtained from this

66
solution. Phosphorus was determined using the above soil P technique.
Calcium, Mg, A1 and Fe were determined by atomic absorption and K by
flame emission as described above. Tissue N was determined by the macro
Kjeldahl technique (Bremner, 1965).

67
Statistical Analysis
Statistical analysis of data was performed using either a
Hewlett-Packard model 9100B desk top computer or the computer facilities
at the University of Florida Computer Center. Library Stat-Pac programs
were used with the Hewlett-Packard. All statistical analyses done at the
Computer Center were performed using programs available in the Statisti
cal Analysis System (SAS) package (Barr and Goodnight, 1972). Specific
programs used aregiven in the following sections where appropriate.

RESULTS AND DISCUSSION
Preliminary Screening of Soil-Test Methods
The classification and selected properties of the 19 soils used in
preliminary screening of soil-test methods are shown in Table 6. They
are typical of many forest soils in the lower Coastal Plain. These are
acid, sandy, and relatively low in organic matter. Levels of extractable
P vary considerably among the 10 soils.
Heights, relative heights, P-fertilizer requirements and tissue-P
parameters of slash pine grown on these 10 soils in both greenhouse and
field trials are summarized in Tables 7 and 8, respectively. Data in
these two tables were summarized or computed from the original data of the
greenhouse and field trials presented in Appendix Tables 46 and 47, re
spectively. Phosphorus uptake values at the first year's harvest in the
greenhouse trial were computed using the regression equation
Y = 1.604X 4.951 (r = 0.999)
where Y = mg P in tops and roots per pot, and X = mg P in tops per pot.
This equation was derived using plant uptake data obtained from extra pots
which were subject to a complete harvest after the first year's growth
(Appendix Table 48).
The amounts of P extracted from the 10 soils by all soil-test meth
ods are given in Appendix Table 49- Also included in this table are the
amounts of various P fractions and the buffering capacities of these soils
Relationships between extractable soil-P values and plant growth
and response parameters are usually curvilinear, according to Grigg (1965)
68

Table 6. Classification and
selected
properties of
10 soils
used in greenhouse
study
1
Soil
pH
Silt +
Organic
Ext. A1
Extractable P
Type
Order
(h2o)
clay
CEC
matter
NH/jOAc
NH^OAc
h2o
Bray 1(3)
£
me/100g
%
ppm--
Bladen scl*
U11 i sol
4.8
43-5
8.67
3.27
222
0.7
0.5
1 .0
Blanton fs
U 11isol
5.2
5.5
2.17
1.24
40
4.3
1.7
48.9
Plummer fs
U11i so 1
5.0
8.5
3-94
2.23
105
13.8
3.0
185.5
Marlboro fsl
U1tisol
5.0
38.5
4.70
2.57
149
0.4
0.2
1.6
McLaurin fsl
U1tisol
5.3
31.0
4.70
2.20
91
0.3
0.3
1.5
Immokalee fs
Spodosol
4.3
7.5
5.46
3.27
8
6.3
5.5
6.6
Leon fs
Spodosol
4.1
6.5
9.28
4.52
11
7.8
6.4
7.9
Ona fs
Spodosol
4.2
10.3
4.90
3.76
55
2.1
2.4
4.7
Kershaw fs
Entisol
5-2
4.0
2.09
1.31
35
2.6
0.7
25-9
Lakeland fs
Entisol
5.4
14.8
4.90
2.57
66
5-9
2.2
130.0
* sel = silty clay loam; fs =* fine sand; and fsl = fine sandy loam textures.

Table 7. Height, relative height, P-fertilizer requirements, and P concentration and uptake
of slash pine
seed 1ings
after
1 and 2
years 1
growth on 10 soil
s in the
greenhouse
Soi 1
type
Height
Re 1 .
height
Fert.
reqm.f'
P in
tops
P uptake^
1*
2
1
2
1
2
1
2
1
2
cm-

%

--kg
P/ha-
%
--mg
P/pot
Bladen scl
10.6
17.1
77.3
45.8
56.6
92.5
0.060
0.044
10.2
19.5
Blanton fs
12.3
29.1
89-3
91.7
13.8
0.0
0.126
0.081
25.4
62.1
Plummer fs
15-3
35.8
98.3
90.3
0.0
0.0
0.132
0.089
45.7
105.6
Marlboro fsl
12.0
19.4
65.0
52.7
70.8
95.6
0.058
0.038
8.1
15.2
McLaurin fsl
12.3
23.8
71.3
64.6
81.4
93.3
0.055
0.047
8.8
22.2
Immokalee fs
14.8
29.4
100.0
98.4
0.0
0.0
0.096
0.039
51.6
35.5
Leon fs
13.0
26.2
100.0
98.1
0.0
0.0
0.108
0.068
44.3
40.7
Ona fs
15.2
34.1
87.2
84.2
4.8
35.8
0.091
0.044
37.6
48.7
Kershaw fs
10.0
26.7
79.9
93.2
48.5
0.0
0.107
0.075
10.6
38.1
Lakeland fs
15.2
32.9
94.5
87.4
0.0
15-7
0.123
0.082
53.6
79.8
* Age in years at harvesting or measurement.
1 Fertilizer required to achieve 30% of maximum height growth.
§ First year uptake based on eight seedlings/pot and 2-year uptake based on four seedlings/pot
over 2 years.

Table 8. Height, relative height and P-fertilizer requirements of slash pine after 1, 3, and 5
years' growth, and
soils
foliar
P concentration after 4 years' growth in the
field on
10
Soi 1
type
Height
Rel. height
Fert. reqm.'f
Foliar P
4
1*
3
5
1
3
5
1
3
5
9

ts.y r / iica
X)
Bladen scl
29.5
90
227
62.2
69.9
79.5
71.2
61.0
42.4
0.067
Blanton fs
55-4
c+\
-3*
451
94.2
98.1
100.0
0.0
0.0
0.0
0.115
Plummer fs
44.4
216
433
100.0
100.0
100.0
0.0
0.0
0.0
0.109
Marlboro fsl
23.5.
136
350
65.1
80.0
89.9
114.0
50.1
1.5
0.076
McLaurin fsl
31.6
174
400
69.6
66.6
81.8
89.0
75.0
42.9
O.O83
Immokalee fs
50.3
209
436
100.0
82.5
86.0
0.0
47.6
25.6
0.090
Leon fs
88.1
291
498
99.4
90.1
90.0
0.0
0.0
0.0
0.092
Ona fs
42.0
156
342
68.4
76.6
81.1
94.3
119.5
115.6
0.081
Kershaw fs
26.4
135
293
89.2
100.0
100.0
0.0
0.0
0.0
0.103
Lakeland fs
27.5
129
276
96.5
100.0
100.0
0.0
0.0
0.0
0.101
* Age in years at measurement.
" Fertilizer required to achieve 90% of maximum height growth.

72
This was true for data from this study, as statistical models which allow
for curvi1inearity (models 2, 3, and A in Table 9) provided better fits
between soil-test values and tree parameters than an untransformed linear
model (model 1 in Table 9). for the selected data used in this compari
son, most models using a logarithmic or an arctangent transformation of
the independent variable (soil-test value) and a quadratic model provided
better fits, as indicated by the R value, than the untransformed linear
model. There was little difference between the transformed models and
for convenience the logarithmic model was used for all computations in the
preliminary screening. The square of the multiple correlation coefficient
(R^) was used as the index of success of soil-test methods as predictors
of the tree parameters.
Relationships Between Soil-Test Values and Relative Height
Relative heights, the heights of unfertilized trees expressed as a
percent of heights of fertilized trees, after 1 year's growth in the
greenhouse (Table 7) were almost identical to those in the field after 1
year of growth (Table 8). This suggested a similarity in the modes of
nutrition of young seedlings in both the greenhouse and field. Responses
(relative heights) to added P, after 5 years of growth in the field, were
similar to those obtained after 3 years, but responses obtained at both
periods tended to differ from those obtained during shorter growth periods
in either field or greenhouse trials. This is in agreement with the
findings of Head and Pritchett (1971) who reported a poor relationship
existed between response in the greenhouse and that obtained after several
years in the field.
Response trends developed with time for the various soils can be
categorized fairly well in accordance with the relationship between P

Table 9* Comparison of the goodness of fit of
of the multiple correlation coefficient
four statistical
(R2),
models, as indicated by the square
relating selected tree parameters (dependent
vari able)
and soil
-test values
(i nde-pendent
variable)
Soi1-test
method
Mode 1^
Greenhouse (l
year)
Field (3 years)
Fol iar P
(4 years)
Height
Rel.height
Fert.reqm.
Height
Rel. height
Fert.reqm.
H0
1
0.288
0.741**
0.601**
0.545*
O.O67
0.059
0.056
L
2
0.411*
0.951**
0.897**
0.421*
0.217
0.087
0.228
3
0.489*
0.947**
0.951**
0.410*
0.257
0.086
0.265
4
0.676**
0.935**
0.949**
0.548*
0.285
0.061
0.338
T ruog
1
0.358
0.355
0.339
0.036
0.421*
0.286
0.346
2
0.342
0.658**
0.686**
0.113
0.658**
0.413*
0.496*
3
0.203
0.702**
0.803**
0.132
0.645**
0.350
0.424*
4
0.386
0.626*
0.653**
0.096
0.616*
0.433
0.461*
HC1-H-S0,,
1
0.282
0.280
0.265
0.040
0.425*
0.287
0.392
2
0.262
0.579*
0.607**
0.151
0.783**
0.494*
0.706**
3
0.193
0.689**
0.764**
0.218
0.734**
0.386
0.642**
4
0.292
0.413
0.450*
0.053
0.616*
0.510*
0.615*
Bray 1
1
0.251
0.205
0.211
0.007
0.434*
0.289
0.386
2
0.094
0.241
0.279
0.022
0.785**
0.551*
0.704**
3
0.034
0.355
0.419*
0.064
0.908**
0.640**
0.712**
4
0.282
0.207
0.228
0.014
0.642**
0.446*
0.575*
* Significant at the 5% level.
** Significant at the 1% level.
+ Model 1, Y = bX + c; Model 2, Y = b logX + c; Model 3, Y = b arctan X + c; Model 4, Y = aX + bXz +

74
extracted by the H0 and Bray 1(3) methods (Table 6), which can be taken
as Indicative of the intensity and quantity factors of soil P supply, re
spectively (Williams and Knight, 1963). For soils containing high ex
tractable P by and moderately low by Bray 1(3) methods--Immokalee and
Leon--no growth response occurred in the first year, but a response was
evident after 3 and 5 years. Those soils with a moderate to low extract
able P by H2O but high by Bray 1(3) methods--Blanton, Kers-haw, and Lake-
land--showed the opposite trend with some response in the early growth
period which disappeared with increasing time. Response occurred in the
early growth period and persisted with time on soils with low extractable
P by both H2O and Bray 1(3) methods--Bladen, Marlboro, and McLaurln--
while on the one soil with moderate to high extractable P by H^O and Bray
1(3) methods--Plummeino response occurred at all during the first 5
years. Apparently the amount of water-soluble P (intensity factor) Is
of importance during the early establishment and growth of the pine seed
ling while the amount of solid-phase P in equilibrium with soil-solution
P (quantity factor) becomes more important to tree growth with increasing
time of persistence on the site.
The response trends discussed above are well illustrated by the
effectiveness of P extracted by various soil-test methods at predicting
relative heights in the greenhouse (Appendix Table 50) and the field
2
(Appendix Table 51). The effectiveness (R ) of soil-test methods as
predictors of relative height at any age appeared to be closely related
to the amount of P extracted, but independent of the type of extractant
used where similar quantities of P were extracted (Fig. 1). In view of
this, the results will be Illustrated and discussed principally using
those obtained with some of the more commonly used soil-test methods

MULTIPLE CORRELATION COEFFICIENT,
75
MEAN EXTRACTABLE P, ppm
Fig. 1. Rejationshlp between R values for regressions of relative
height at age 1, 3, and 5 years on soil-test values, and
mean amounts of P extracted by soil-test methods.

which vary in their capacities to extract P from soils. The similar ef
fectiveness of different extractants which remove equivalent quantities
of P suggests that they extract the same form(s) of P from the soil.
Aluminum-P is probably the form most commonly involved since this fraction
dominates the inorganic P component of the 10 soils (Appendix Table kS)
and most soil-test extractants are capable of solubilizing A1 -P (Thomas
and Peas lee, 1973)
Soil-test methods involving use of weak extractants, such as H^O
and NH^OAc which extract only small amounts of P from most soils, pro
vided the best index of tree response to P fertilizers over short growth
periods in both the greenhouse and field trials (Table 10). Water-
extractable P was slightly superior to that extracted by NH^OAc at pre
dicting response after 1 year's growth of seedlings in the greenhouse,
while NH^OAc was a better predictor of response than H20 after 1 year's
growth in the field. Other soil-test methods which provided a good index
of response over short growth periods included those involving use of
neutral salts, short term extractions with anion exchange resin, and
very dilute acid solutions at a narrow soi1:solution ratio (Appendix
Tables *9 and 50). The effectiveness of other soil-test methods as indi
cators of short term response tended to be inversely proportional to the
amount of P extracted (Fig. 1 and Table 10).
Responses after 3 and 5 years' growth in the field were most
closely related to P extracted by soil-test methods which removed larger
amounts of P from the soil than H20 and NH^OAc. As the growth period
2
increased from 1 to 3 or 5 years, R values decreased for the soil-test
methods involving use of weak extractants. The R2 values for response
after 3 and 5 years' growth increased with increasing amounts of P

Table 10. Relationships between selected soil-test values and relative
height growth of slash pine in field and greenhouse experi
ments on 10 soils
Soi1-test
method or
P form
Mean
extract-
able P
Relative height
Greenhouse
Field
1 +
2
1
3
5
d2_
ppm
h2
2.3
0.951**
0.715**
0.614**
0.217
0.055
NH^OAc
4.4
0.863**
0.729**
0.817**
0.589**
0.348
T ruog
8.9
0.658**
0.440*
0.658**
0.657**
0.481*
01 sen
12.3
0.533*
0.515*
0.702**
0.787**
0.666**
HC1 -H2S04
15.2
0.579*
0.538*
0.722**
0.782**
0.639**
h2so/4(9)
32.5
0.349
0.331
0.540*
0.765**
0.716**
Bray 1 (3)
41.3
0.187
0.478*
0.623**
0.845**
0.778**
NHAF(pH8.5)
57.0
0.296
0.060
0.465*
0.718**
0.699**
Organic-P
42.6
0.050
0.010
0.011
0.040
0.056
Tota 1-P
136.1
0.040
0.001
0.065
0.229
0.306
+ Tree age at time of measurement (years).
* Significant at the 5% level, using the model Y = b logX +c, where
Y = relative height and X = soil-test value.
** Significant at the \% level.

78
extracted by the soil-test methods up to a plateau (Fig. 1 and Table 10).
The plateau, where R values were almost independent of quantities of P
extracted, was reached at lower levels of mean extractable P for response
after 3 years (ca. 12 ppm) than after 5 years (ca. 20 ppm). As mean ex-
2
tractable P values increased above ca. 50 ppm, the R values began to de
cline indicating the soil-test methods involved were beginning to extract
forms of P unavailable for tree use at this stage of growth. Soil-test
methods extracted-amounts of P corresponding to the above-mentioned pla
teau included those involving use of strong acid and alkaline solutions
(HC1, H^SO/j and NaHCO^) and strong complexing agents (lactate, NH,F).
Amounts of Fe-P and Ca-P in the 10 soils were very low and were
not related to fertilizer response at any stage of tree growth (Appendix
Tables 50 and 51). Aluminum-P, which dominated the inorganic P fractions
of all 10 soils, was significantly related to tree response when they
were 3 and 5 years old. However, the mean amount present in the soil,
5**.3ppm P, suggests that not all this fraction was available for tree use
during the first 5 years on the basis of data in Fig. 1. This is perhaps
not surprising since different A1 P compounds found in soils vary consid
erably in their availability to plants (Taylor et al., I960). Soil-test
methods which discriminate between Al-P compounds of different solubility
are apparently better indicators of soil-P status than methods which do
not so discriminate.
Neither total nor organic-P were significantly related to response
at any stage of growth (Table 10). This concurs with the shortcomings of
total analyses reported previously (Pritchett, 1968; Ballard, 1970a).
However,it is of interest to note that R values for both methods of
analysis increased with increasing growth period. This result Is consis
tent with the premise that an increasing proportion of the total P in

79
soil becomes involved in meeting the trees' P requirements as growth peri
ods increase. However, it is unlikely that total P would ever become an
effective predictor of soi1-P status, even over very long growth periods
(rotation length, ca. 25 years) since certain fractions of the soil's
total P are occluded and extremely insoluble.
Relationships Between Soil-Test Values and P Uptake and Tissue P
Phosphorus uptake by seedlings during the first year's growth in
the greenhouse was most closely correlated with P extracted by weak ex
tractants e.g. H^O and NH^OAc (Table 11). However, P uptake over the 2-
year growth period was more closely correlated with P extracted by soil-
test methods which removed larger amounts of P from the soil (strong ex
tractants). In contrast to P uptake, P concentrations in seedling tops
at the end of the first year were more closely correlated with P extracted
by the stronger extractants than the weaker extractants. After two years
growth in the greenhouse the advantage of the stronger extractants over
the weaker ones for predicting P concentrations in the tops was more
evident.
2
The magnitude and ranking of the R values for the second year P
concentrations in tops and for foliar P of 4-year-old field trees were
strikingly similar. Raw data in Tables 7 and 8 show that both first and
second year P concentrations in tops of greenhouse grown seedlings were
closely correlated with foliar P of field trees (r = 0.897 and 0.875 re
spectively), although the P concentration in seedling tops at the end of
the first year in the greenhouse was quantitatively more closely related
to foliar P than was the P concentration in tops at the end of the second
year. Terman and Bengtson (1973) reported close agreement between their
established 'critical' P concentration in tops of 8-to 12-month-old slash

Table 11. Relationships between selected soil-test values and tissue P
parameters of greenhouse and field slash pine grown on 10 soils
Soi1-test
method or
P form
Mean
ext ract-
able P
Greenhouse
Field
P uptake
%P in tops
foliar P
1 +
2
1
2
4
r2
ppm
h2o
2.3
0.81.0**
0.322
0.537*
0.150
0.228
NH/jOAc
4.4
0.709**
0.607**
0.847**
0.454*
0.511*
T ruog
8.9
0.634**
0.828**
0.821**
0.599**
0.495*
Olsen
12.3
0.447*
0.875**
0.931**
0.806**
0.765**
HCl-H2S0il
15.2
0.516*
0.871**
0.925**
0.753**
0.706**
h2S04(9)
32.5
0.314
0.873**
0.811**
0.846**
0.698**
Bray 1 (3)
*1 .3
0.348
0.850**
0.895**
0.827**
0.817**
NH4F(pH 8.5)
57.0
0.266
0.722**
0.742**
0.746**
0.638**
Organic-P
42.6
0.141
0.459*
0.070
0.164
0.019
Tota 1-P
136.1
0.066
0.600**
0.213
0.442*
0.175
fTree age at time of sampling (years).
* Significant at the 5% level, using the model Y = b logX + c, where Y =
P parameter and X = soil-test value.
**Signifleant at the 1% level.

pine and the 'critical' foliar P values reported by other workers for
this species in the field. The relationships reported above could be of
value in extrapolating greenhouse trial results to field conditions.
Relationships Between Soil-Test Values and Fertilizer Requirements
The association found between mean quantities of P extracted by
soil-test methods and their ability to predict response to P fertilizer
was also apparent in the relationships between soil-test values and P-
fertilizer requirements (Table 12). Phosphorus extracted by weak extrac
tants was most closely correlated with fertilizer requirements over short
growth.periods, while P extracted by stronger extractants was most closely
correlated with fertilizer requirements over longer growth periods.
Although R values for the relationships between soil-test values
and fertilizer requirements in the greenhouse were of a similar magnitude
to those for relative height, R4 values for fertilizer requirements cal
culated from field tests were lower than those for relative height in
field tests. This was particularly pronounced for the fifth, and to a
lesser extent, the third year data. This could be due to variation in
the P-retention capacity of the soils. For instance, Lewis and Harding
(1963) reported lower P-fertilizer requirements for pines growing on soils
of low P-retention capacity compared to those on soils of high P-retention
capacity, but with similar extractable soil-P values. One would expect
differences due to P retention to become more pronounced with time due to
the time dependency of the reversion of relatively soluble phosphates to
less soluble forms (Juo and Ellis, 1968). If variable P retention was
the major factor accounting for the low R2 values shown in Table 12, a
multiple regression equation including both soil-test values and a measure

Table 12. Relationships between selected soil-test values and P-fertilizer require
ments of slash pine in field and greenhouse experiments on 10 soils
Soi1-test
method or
P form
Mean
extract
able P
P-fertilizer requirements^
Greenhouse
Field
1?
2
1
3
5
ppm
h2o
2.3
0.897**
0.703**
0.508*
0.088
0.000
NH/jOAc
4.4
0.851**
0.863**
0.745**
0.357
0.100
T ruog
8.9
0.685**
0.607**
0.613**
0.414*
0.154
01 sen
12.3
0.552*
0.678**
0.694**
0.513*
0.201
hci-h2so4
15.2
0.606**
0.697**
0.686**
0.494*
0.192
h2so4(9)
32.5
0.377
0.488*
0.567*
0.519*
0.230
Bray 1(3)
41 -3
0.436*
0.616**
0.601**
0.506*
0.217
NH4F(pH 8.5)
57.0
0.322
0.396*
0.491*
0.507*
0.247
Organic-P
42.6
0.050
0.001
0.003
0.016
0.003
Total-P
136.1
0.043
0.015
0.077
0.193
0.123
4.
' P fertilizer required
to achieve 90% of maximum growth.
s> Tree age at
time of determination (years).
* Significant
at the 5%
level, using the model Y =
b logX + c,
where Y = P
ferti 1izer
requirement
and X = soil-test value.
** Significant at the 1% level.
oo
NJ

83
of soil's P-retention capacity as independent variables would provide a
better prediction of fertilizer requirements. This hypothesis will be
examined in a later section.
Relationships Between Soil-Test Values and Height
In the ensuing discussion, height refers to tree height in the ab
sence of P fertilizer. Extractable soil P was much less effective at
predicting tree height (Table 13) than at predicting relative height
(Table 10). This fact was particularly pronounced under field conditions.
Only certain tests which extracted small quantities of P from the soil
were significantly correlated with the height at the end of 1 year's
growth in the greenhouse and 1 and 3 years' in the field. No single test
was significantly correlated with height at age 3 in the field (Table 10,
Appendix Table 51).
As would be anticipated from trends shown by response and P uptake
data, the effectiveness of soil-test methods involving use of stronger
extractants increased as the growth period in the greenhouse increased.
If inadequate available P was the major factor limiting height growth in
the field, one would have anticipated a similar trend in the field to
that shown in the greenhouse. Although the effectiveness of soil-test
methods which extracted small quantities of P decreased with increasing
growth period, there was no indication of any increase over time in the
effectiveness of soil-test methods involving use of stronger extractants.
This suggests that factors other than available soil P were contributing
to variation in height growth in the field, particularly over longer
growth periods. The effect of site factors other than soil P on height
growth in the field will be examined in a later section.

Table 13. Relationships between selected soil-test values and height growth in the
absence of P fertilizer of slash pine in field and greenhouse experiments
on 10 soils
Soi1-test
method or
P form
Mean
extract-
able P
Height growth
Greenhouse
Field
lf
2
1
3
5
r2
ppm
h2o
2.3
0.411*
0.506*
0.525*
0.421*
0.258
NH^OAc
4.4
0.307
0.570*
0.317
0.332
0.178
T ruog
8.9
0.342
0.562*
0.085
0.113
0.037
01 sen
12.3
0.218
0.601**
0.067
0.147
0.059
hci-h2so4
15.2
0.262
0.624**
0.078
0.150
0.062
H2S\
32.5
0.159
0.503*
0.006
0.045
0.006
Bray 1(3)
41.3
0.187
0.607**
0.018
0.093
0.034
NH4F(pH 8.5)
57.0
0.145
0.268
0.001
0.031
0.003
Organic-P
42.6
0.288
0.176
0.083
0.052
0.056
Total-P
136.1
0.099
0.515*
0.090
0.030
0.048
+ _
1 Tree age at
time of measurement (years
).
* Significant
at the 5%
level, using the
model Y -
b logX + c, where Y =
height
and X = soil-test value.
** Significant at the ]% level.

85
The success of soil-P tests for predicting response to P fertili
zer, despite the apparent limiting effect on growth of site factors other
than soil-P levels, can probably be attributed to the use of relative
height as the measure of response. By expressing height response to P
fertilizer at each site as a function of height in the absence of P fer
tilizer, the influence of site factors other than soil P among sites is
effectively reduced.
General Discussion
Research results indicate that P-intensity measurements (weak ex
tractants) correlated better with yield and response of annual crops,
which depend upon rapid uptake in the early stages of growth, than do P-
quantity measurements (strong extractants). However,P-quantity measure
ments correlate better with nutrient concentrations and uptake data,
which depend upon absorption over the entire growth period (Williams and
Knight, 1983). Data for the first year's tree growth in the greenhouse
test concur with these findings. The trend for quantity measurements to
become better predictors of growth and response to added P, when growth
periods of greater than 12 months were considered, can probably be re
lated to the efficiency of internal P recycling within plants in addition
to the longer absorption periods involved. Growth and response in years
subsequent to the first year will be determined not only by current uptake
from the soil but also reserves of P accumulated in tissues from the pre
vious year(s) uptake. This latter (P uptake) is determined by the quan
tity factor of soil-P supply (Williams and Knight, 1963).
Clear distinction between the value of intensity and quantity mea
surements of soil-P supply for predicting growth, response to P fertili
zer, and nutrient uptake will only be apparent if there is no relationship

86
between the intensity and quantity factors of P supply for the soils used
in any study. Such was the case in this study (Table 6). Although P ex
tracted by the NH^OAc method was found to be unrelated to response over a
5 year period in this study, Pritchett (1968) reported a significant cor
relation between P extracted by NH^OAc and response of slash pine on six
Coastal Plain soils 5 to 8 years after fertilizer addition. However,an
examination of the relationship between NH^OAc and Bray 2-extractable P
reported for these six soils showed them to be significantly correlated
(r = 0.902). This close relationship between the intensity and quantity
measurements of soil P on these six soils probably accounts for the suc
cess achieved with the NH^OAc procedure.
:- The trend for intensity measurements of soil P to decline in value
and quantity measurements to improve in value as predictors of tree growth
and response with increasing age of trees probably accounts for the effect
of tree age on the success of soil-test methods at predicting tree height
reported by Wells and Crutchfield (1969), Wells et al. (1973) and Ballard
(197*0. These authors found the ability of P extracted by strong extrac
tants (HCl-H^SO^, Bray 2, Olsen) to predict height growth of young trees
improved with age of trees from time of planting. It appears reasonable
to assume that had these studies included a measure of soil-? intensity
they would have found the same trends as observed in this study.
The growth of seedlings in pots is analagous to growth of seedlings
in nurseries. Thus,it can be expected that soil-test methods using ex
tractants such as H^O and NH^OAc should be good predictors of growth and
fertilizer requirements of nursery stock. However, a good transplant
seedling should not only be large and sturdy, but it should also possess
a high nutrient content. Since the nutrient contents tend to be related

8?
to quantity measurements of soil P, nursery fertilizer practices based
solely on intensity measurements could result in seedlings of adequate
size but possessing less than optimal nutrient reserves for maximum growth
following transplanting. A sound practice would be to base fertilizer
recommendations on the measurement of both quantity and intensity factors
of soi1-P supply.
It should be appreciated that soil-test methods which provide a
good measure of the quantity factor of soi1-P supply for acid sandy soils
which have their inorganic P fraction dominated by Al-P, such as those
used in this study, will not necessarily work as well for a different
range of soils. For instance, Alban (1972), using soils of high base
saturation which probably had insoluble basic Ca-P as a major component
of the inorganic P fraction, found a poor correlation between site index
of red pine and P extracted by solutions with a low pH (HC1-H2S0^, Bray
2). This can probably be attributed to the acidic extractants removing
insoluble basic Ca-P which was not available for tree use.
On the basis of results reported in this section, five soil test
methods--H20, NH^OAc, HCl-H2S0/t, Bray 1 (3) Bray 2--were selected for use
in the calibration study using 72 sites. These test methods extract P
with a range of intensity and they also include the tests currently used
on a routine basis by most of the southern states (Page et al., 1965).

88
Phosphorus Compounds
Although correlation studies relating soil-P fractions to plant-P
uptake and P extracted by soil-test methods have provided some theoret
ical basis for selecting soil-test methods (Thomas and Peaslee, 1973),
this information can be misleading for several reasons: (a) The selec
tivity of some P-fractionation procedures are questionable (Bromfield,
1967)- (b) Phosphorus compounds within fractionation categories, such as
Ca-P, Al-P and Fe-P, may vary considerably in their availability to
plants (Juo and Ellis, 1968). (c) Correlation techniques used in most
studies may not provide definitive evidence of direct relationships be
cause of possible unknown mutual correlations (Thomas and Peaslee, 1973)-
(d) Variable P readsorption and buffering effects of different soils
during extraction can confuse the relationships between P fractions and
extractable P (Martens, Lutz, and Jones, 1969)-
Examination of the solubility in chemical extractants of P com
pounds of known P availability to plants should provide more direct evi
dence (a) on which to base the selection of soil-test methods, and (b)
to explain the success or failure of various soil-test methods for pre
dicting plant needs for P fertilizers.
Solubility of P Compounds in Chemical Extractants
The solubilities of P compounds in the extracting solutions of
several well-known soil-test methods, in the presence and absence of two
soils are shown in Table \h. Soil-test values are corrected for the
amount of native P extracted from the untreated soils by the correspond
ing extractant.
Monocalcium phosphate (MCP)
Dissolution of MCP was complete in all extractants except the

Table 14. Solubility of P compounds in chemical extractants in the presence and absence of two soils
Soi 1
Soil-test method
H2O
NH/jOAc
T ruog
HC1-H2S0Z,
01 sen
Lactate
Bray 1
Bray 2
N H *4 F
_ 0/
'b
Monocalcium phosphate(MCP)
None
100.0
100.0
100.0
100.0
90.6
100.0
100.0
100.0
90.8
1mmoka1ee
97.6
89.6
100.0
95-3
79.1
90.9
95.7
97.3
95.6
McLaurin
*0.9
50.5
78.5
72.4
69.7
79. 1
90.8
97.1
92.2
Dicalcium phospha
te(DCP)
None
13.2
96.7
100.0
100.0
39^8
100.0
100.0
100.0
86.0
1mmoka1ee
91 .2
72.6
98.0
87-3
1.9
87.4
88.5
93.3
88.9
McLaurin
38.4
50.7
85.0
69.1
3.0
11.1
83.0
86.6
17.4
FIuorapatite(FA)
None
0.3
4.6
29.5
77.0
0.2
42.5
16.5
51.5
0.1
1mmokalee
3.2
0.6
25.5
63.0
0.2
12.8
0.8
35-9
1.9
McLaurin
0.2
0.1
8.6
49-4
0.1
7-9
0.3
17.8
0.2
Colloidal a
luminum phosphate(CAIP)
None
2.3
1 .9
6.3
35.8
12.7
67.1
93-7
93-6
87.8
1mmoka1ee
15.7
3-7
7.8
35.2
13.6
39.1
90.7
85.8
79.2
McLaurin
9.8
0.9
4.6
23.6
4.5
39-9
80.2
84.1
71 .9
Potassium taranaki te(KTK)
None
*.0
4.4
4.3
23.8
11.4
7-3
97.0
97.5
99.0
1mmoka1ee
3.7
3.4
3.0
7.8
8.2
4.2
86.7
83.8
79-2
McLaurin
0.2
0.8
2.7
4.7
2.8
2.2
84.6
87.7
80.3
Wavel1 i te(WA)
None
0. 1
0.1
0.1
0.1
0.2
0.5
8.2
7-8
1.9
1mmoka1ee
0.3
0.1
0.2
0.8
0.1
0. 1
6.9
7.6
3.5
McLaurin
0. 1
0.1
0.5
0. 1
0.2
0.1
3.1
2.3
0.1
Colloidal
ferric phosphate(CFeP)
None
2.6
4.2
2.2
17.4
29.1
49.3
90.5
97.5
8.1
1mmokalee
19.8
4.8
2.4
22.4
51.7
88.2
91 .2
85-3
57.0
McLaurin
17.9
2.2
4.2
18.5
40.4
79.7
65.8
66.7
64.1
Strengite(STR)
None
0.1
0. 1
0.1
0.1
0.3
0.1
0.1
0.1
0.2
1mmokalee
0. 1
0.1
0.1
0.5
0.4
0. 1
0.5
1.7
0.4
McLaurin
0. 1
0.1
0. 1
0.1
0.2
0.1
1 1
0.4
0.2

90
alkaline extractants of the Olsen and NH^F methods. Because of the high
solubility of MCP, recovery values in the presence of soil provide a
good index of the amount of P adsorption which occurs during extraction.
Adsorption in the presence of the Immokalee soil, which has virtually no
P-retention capacity (Table A), was minimal for all soil-test methods.
But in the presence of the McLaurin soil, which has a relatively large
P-retention capacity, adsorption ranged from 59% in water to only 3% in
the Bray 2 extractant. Soil-test methods in which appreciable adsorption
occurs during extraction are considered to provide a good index of the
intensity factor of soil-P supply (Williams and Knight, 1963). It is
noteworthy that almost as much adsorption occurred during extraction
with NH^OAc as with H^O, although one of the main purposes of the ace
tate ion in certain extraction solutions is supposedly to prevent
adsorption of P removed by other ions (Thomas and Peaslee, 1973)-
Dicalcium phosphate (DCP)
a
Acidic extractants completely dissolved DCP, and in the presence
of soil the degree of recovery was similar to that for MCP. Dicalcium
phosphate was only slightly soluble in water, but in the presence of
soil the degree of P recovery almost matched that for MCP, suggesting
almost complete dissolution of DCP. The acidity of the soil-water
systems (pH 4.4 and 5-1 for the Immokalee and McLaurin soils, respective
ly, at the conclusion of the extraction) was apparently sufficient to
dissolve DCP. The Olsen method (alkaline NaHCO^) extracted 40% of the
DCP but, in the presence of soils recovery values fell well below those
anticipated if adsorption was the only mechanism reducing P recovery.
Although the solubility of DCP was enhanced as a result of the bicarbon
ate ion reducing the activity of Ca in solution (Olsen et al. 1954),

91
this effect may be offset by Na from the extractant causing release of
Ca from soil exchange sites. Alkaline NH^F, the extractant used to
selectively remove Al-P in most P-fractionation procedures, extracted
appreciable amounts of DCP. This has been recorded previously
(Bromfield, 1967) and attributed to CaF2 formation. The degree of ex
traction by the NH^F method was unaltered in the presence of Immokalee
soil but substantially reduced in the presence of McLaurin soil. The
explanation for this low P recovery was not clear as it could not be
explained by adsorption of P during extraction.
Fluorapatite (FA)
The solubility of FA in the extractants correlated fairly close
ly with the pH of the extractants. However, soil-test methods with
extractants capable of forming stable Ca complexes (lactate) and those
utilizing wide so i 1:solution ratios (Truog) and long extraction periods
(lactate) extracted more P than was anticipated, based on the pH of
these extractants. The reduced recovery of P from FA by the NH^OAc,
Truog, and HC1-H2S0^.methods in the presence of soil was consistent with
the P adsorption anticipated on the basis of the MCP data. However,
reduced recovery of P in the presence of soil by the lactate and both
Bray methods was greater than anticipated from P adsorption alone. In
the lactate method, this may have been caused by soil cations competing
with FA-Ca for lactate and acetate functional groups. The considerable
reduction in P recovery by the Bray 1 method may partially be explained
by the poor buffering capacity of this extractant; the final pH of the
Immokalee soil extraction solution was 2.7 and that of the McLaurin
3.0. The more acidic Bray 2 extractant was better buffered and it can
only be speculated that the reduced P extractabi 1ity arose from

92
competition for available F between native soil cations and FA-Ca.
Colloidal aluminum phosphate (CAiP)
Water and NH^OAc extracted very small quantities of CAIP. Soil-
test methods involving use of inorganic acid extractants removed
increasing quantities of P with increasing acid strength, but even the
extractant with the lowest pH (HC1-H2S0^) removed considerably less P
than extractants containing complexing agents (lactate, NH^F, Bray 1,
Bray 2). The Olsen method, which released P from A1 and Fe by
hydrolysis, extracted relatively small quantities of P from CAIP; it
rated between the Truog and HCl-H^SO^ methods in extraction efficiency.
Adsorption of P during extraction accounts for the effect of soil on the
P extractabi1ity by all methods except for H^O and NH^F.
The increased water solubility of CAIP in the presence of soil
was not anticipated. The slight reduction in pH from 7 to the vicinity
of A to 5 would theoretically be expected to reduce the amount of P in
solution (Bache, 1963). A possible explanation is that the H^O extrac
tion mobilized organic compounds, through ionization of functional
groups and these formed complexes with A1, thereby increasing the re
lease of P (Schnitzer, 1969). The reduced P recovery by the NH^F
extraction probably resulted from the precipitation of Ca phosphates
since exchangeable soil Ca was not removed prior to extraction, as was
recommended when this method is to be used for determining Al-P (Fife,
1959).
Potassium taranakite (KTK)
The solubility of KTK was almost identical to that of CAIP in all
extractants containing F (Bray 1, Bray 2, NH^F). Likewise, the order of
extractabi 1ity in the inorganic acid extractants (Truog, HC1-H2S0^) and

93
alkaline NaHCO^ (Olsen) was the same as for CA1P, although less P was
extracted. The only major differences in extractabi 1ity between CAIP
and KTK occurred with the lactate method, and the H2O extraction in the
presence of soil. The lower solubility of KTK in lactate probably
resulted from the lower solubility of KTK in weakly acid solutions. Ap
parently at pH 3-7, insufficient dissolution occurred to enable the
complexing capacity of lactate to appreciably alter the dissolution rate
of KTK. A similar explanation, with natural organic compounds instead
of lactate acting as the complexing agents, probably accounts for the
failure of KTK-water solubility to increase in the presence of soil.
Wavellite (WA)
Extractants containing F were the only ones to remove other than
trace quantities of P from WA. If A1 -P compounds of similar stability
to WA exist in the soil, alkaline NH^F will underestimate the amount of
Al-P in the soil.
Colloidal ferric phosphate (CFeP)
This compound was less soluble than CA1P in inorganic acid extract
ants but more soluble in the Olsen extractant. Solubility in both Bray
extractants was similar to CA1P, although the recovery was less in the
presence of the McLaurin soil; probably as a result of F preferentially
complexing with the large amounts of A1 in this soil. Substantial
increases in P solubility in the presence of soil occurred with H2O,
Olsen, NH^F and lactate methods. In the case of the three methods utiliz
ing nonacidic extractants, the increased solubility probably resulted
from ionized organic functional groups complexing Fe (Schnitzer, 1969).
The possibility that soil organic matter and other soil constituents
contributed to solubilization by reducing Fe^+ to Fe2+ cannot be

9^
ignored. This may be one of the explanations for the increased
extractabi1ity noted with the lactate method. The marked increase in P
extracted from CFeP by alkaline NH^F in the presence of soil, presumably
caused by ionized organic functional groups complexing Fe, points to an
error in the assumption that alkaline NH^F is a selective extractant of
soil A1 -P.
Strengite (STR)
This ferric phosphate was practically insoluble in all extractants
examined. Only the Bray 2 method, in the presence of Leon soil, extract
ed more than 1% of the P in this compound.
Utilization of P Compounds by Slash Pine Seedlings
The uptake of P and dry matter production of slash pine seedlings
grown on two soils treated with the eight P compounds described in a
previous section are shown in Fig. 2 and Fig. 3, respectively. The raw
data from the greenhouse trial are presented in Appendix Table 52.
An analysis of variance of P uptake (Appendix Table 52) showed
significant differences between soils (S) and P compounds (P). The
S x P interaction was also highly significant and thus differences in
uptake between P treatments were examined within soils by the least
significant difference (LSD) method (Snedecor and Cochran, 1967)- Data
in Fig. 2 showed that on the Immokalee soil significantly greater
amounts of P were taken up by the seedlings in the MCP, DCP, FA, CA1P,
KTK, and CFeP treatments than in the control (C), WA and STR treatments.
There were no significant differences in P uptake within the former and
the latter groups mentioned above. On the McLaurin soil, significantly
greater P uptake occurred in the MCP, CA1P, CFeP and KTK treatments than
in the control, FA and WA treatments. The CFeP and KTK treatments were

95
the only ones with a significantly greater P uptake than the STR treat
ment.
The significantly greater P uptake over all P treatments on the
Immokalee soil compared to the McLaurin soil can probably be attributed
to three main factors: (a) First, there was a greater uptake of P from
the untreated Immokalee soil than the untreated McLaurin soil. The Imm
okalee soil (site A16) was not P deficient for young seedlings (Table 7)
while the McLaurin soil was deficient. (b) Second, the Immokalee soil
had a negligible P-retention capacity while that of the McLaurin soil
was relatively large (Table 4). The P-retention ability of the McLaurin
soil probably maintained solution P at low levels. (c) Third, the pH of
the Immokalee soil was lower than that of the McLaurin soil (Table 6).
Since acidic conditions are necessary for the solubilization of the more
basic Ca phosphates (Olsen and Flowerday, 1971), this difference
undoubtedly accounted for the much greater uptake from fluorapatite on
the Immokalee soil than on the McLaurin soil. The complete lack of.
utilization of FA by trees on the McLaurin soil is inconsistent with the
theory that mycorrhizae have the ability to solubilize insoluble
phosphates. The roots of the seedlings at both transplanting and
harvesting were heavily infected with mycorrhizal fungi. This tends to
support the hypothesis that improved P uptake resulting from mycorrhizal
infection is a consequence of the increase in effective root-absorption
surface, which requires that the P be at least partially soluble.
The significant S x P interaction can be attributed, in the main,
to the differences in utilization of FA between the two soils. Also,
the less effective utilization of MCP and DCP compared to KTK, CA1P and
CFeP in the McLaurin soil probably contributed to a small degree to the
interact ion.

70
60
50
40
30
20
10
0
50
40
30
20
10
0
Fig
96
IMMOKALEE FINE SAND
(A16)
LSD
..(5/.)
MCLAURIN FINE SANDY LOAM
(A28)
LSD
(5%)
C MCP DCP FA CAIP KTK WA CFeP STR
PHOSPHORUS COMPOUND
2. Phosphorus uptake by slash pine seedlings after
8 months' growth on two soils treated with
eight P compounds.

DRY MATTER, gm/pot
37
PHOSPHORUS COMPOUND
Fig. 3* Dry matter of slash pine seedlings after 8
months' growth on two soils treated with
eight P compounds.

9S
An analysis of variance of dry matter production (Appendix Table
53) revealed a significant difference between soils, but no difference
between P compounds. However, since there was a significant S x P inter
action, differences between P compounds within soils were examined by the
LSD method. Data in Fig. 3 show there were no significant differences in
seedling yields between P compounds on the Immokalee soil. On the
McLaurin soil, the MCP, DCP, CA1P, and CFeP treatments all produced sig
nificantly greater yields than the control and FA treatments. Yields as
sociated with the MCP, DCP, and CA1P treatments were significantly greater
than those of the WA treatment, while yields from the CA1P treatment was'
also significantly greater than those of the STR and KTK treatments.
Although the Immokalee soil was not P deficient for seedlings, dry
matter production on this soil was significantly lower than that on the
P-deficient McLaurin soil. This result was not anticipated. However,
the possibility of a N deficiency limiting dry matter production on the
Immokalee soil existed. Furthermore, soon after establishment, some salt
burn was observed on seedlings in some of the Immokalee pots. Seedlings
on the Immokalee soil appeared to be slower to recover from transplanting
shock than seedlings on the McLaurin soil. This may have been associated
with the lower pH of the Immokalee soil.
The effective utilization of P from KTK and the two colloidal
phosphates by pine seedlings supports the conclusion, derived from many
experiments with agricultural crops, that these forms of P are readily
available to plants (Taylor et al., I960; Taylor et al., 1963; Juo and
Ellis, 1968). The very poor utilization of insoluble crystalline STR
and WA also concurs with previously reported findings (Taylor et al.,
1963; McLachlan, 1965; Juo and Ellis, 1968).

99
Relationships Between Seedling Utilization and Solubility of P Compounds
Multiple correlation coefficients (R ) for relationships between
seedling utilization of P compounds and solubility of these compounds in
chemical extractants, in the absence and presence of soil, are given in
Table 15. Also included are R2 values for the relationships between
seedling utilization and the extractabi 1ity of the P compounds following
2 months of incubation in the two soils. The extractabi 1ity of the P
compounds by various soil-test solutions following 2 months' incubation
in the two soils, are presented in Table 16.
The R values for the Immokalee soil were generally higher than
those for the McLaurin soil. This was particularly pronounced for acidic
extractants (NH^OAc, Truog, HCl-h^SO^, lactate, Bray 1, Bray 2). The
solubilization of FA by these extractants, which was not reflected in its
availability in the McLaurin soil as it was in the Immokalee soil,
probably accounts for this difference. For groups of soils which contain
native or added FA (rock phosphates) which is unlikely to be available
for plant use due to a lack of soil acidity, soil-test methods using
neutral or alkaline extractants are likely to provide a better index of P
availability than acid extractants.
In general, P extracted from incubated samples was more closely
related to P uptake than was P extracted from the compounds alone or in
the presence of soil. This was expected since P uptake over 8 months in
the greenhouse was influenced by the reaction between the applied P com
pounds and the soils, and this reaction in turn influenced the extract-
ability of P compounds from the incubated soils.
The R2 values using extraction data from incubated samples of both
soils showed an almost identical ranking for soil tests to that found for
the greenhouse data in the preliminary screening of soil-test methods

Table 15- Multiple correlation coefficients (R^) for relationships between solubility of P compounds in
chemical extractants and the uptake of P from these compounds by slash pine seedlings grown on
two soils in the greenhouse
P compound
treatmentt
Soil-
test method
h2o
NH^OAc
T ruog
HC1-H2SO4
01 sen
Lactate
Bray 1
Bray 2
NH^F
r2
Immokalee
fine sand(n=8)
No soil
0.706**
0.837**
0.841**
0.9^8**
0.655*
0.936**
0.712**
0.736**
0.427
+ Soil
0.913**
0.793**
0.827**
0.913**
0.478
0.949**
0.498*
0.889**
0.677*
+ So i 1
(incubated)
0.973**
0.936**
0.962**
0.931**
0.954**
0.912**
0.750**
0.766**
0.841**
McLaurin fine sandy loam(n=8)
No soil
0.546*
0.2 76
0.137
0.275
O.783**
0.265
0.429
0.312
0.681*
+ Soi 1
0.374
0.383
0.147
0.220
0.737**
0.379
0.833**
0.480
0.942**
-+ Soil
(incubated)
0.838**
0.879**
0.434
0.383
0.838**
0.584*
0.787**
0.450
0.853**
Immokalee
+ McLaurin(n=l6)
No soil
0.380*
0.354*
0.307*
0.390**
0.411**
0.383*
0.355*
0.322*
0.300*
+ Soi 1
0.530**
0. **82**
0.337*
0.397*
0.383*
0.450**
0.388**
0.483**
0.550**
+ Soil
(incubated)
0.958**
0.832**
0.594**
0.482**
0.647**
0.561**
0.551**
0.421**
0.600**
^ Treatment prior to extraction.
* Significant at
of P compound.
the 5% level,
using the model Y=b logX + c, where Y=P uptake above control and X=Solubility
**Significant at the
1 % 1evel.
o
o

Table 16. Solubility of P compounds In chemical extractants following 2 months' incubation in two soils
Sol 1-test method
Sol 1
h2o
NH/jOAc
T ruog
HCl-h^SO^ Olsen Lactate
Bray 1
Bray 1
NH^F
%
Monocalcium phosphate(MCP)
Immokalee
79.9
72.6
89.0
81.0
77.8
87.2
92.1
84.0
83-5
McLaurin
1.1
8.6
29.2
37.9
41.2
47.2
66.1
74.4
67.2
Dicalcium phosphate(DCP)
82.5
82.0
1mmoka1ee
72.9
71.7
84.8
81.0
77.2
78.7
82.1
McLaurin
1.6
8.4
29-5
39-5
39-0
29-9
83.O
86.6
69-2
FIuorapatite(FA)
48.3
1mmoka1ee
8.9
10.4
21.4
79-7
10.4
34.0
11.7
13.1
McLaurin
0.1
0.1
10.4
47.2
0.4
8.4
0.2
25.0
0.2
Colloidal aluminum phosphat
e(CAlP)
/
89-8
!mmokalee
60.4
54.6
69-7
78.5
6778
83.8
85-5
85.O
McLaurin
0.6
9.0
18.7
37.6
35.8
43.0
63-7
71.4
63.4
Potassium taranaki te(KTK)
89.1
80.8
86.0
1mmoka1ee
6.8
7.5
8.4
12.8
15.8
9-8
McLaur¡n
1.1
9-0
15.0
37.2
36.9
42.3
67-5
75.9
68.8
Wavel1 i te(WA)
1mmoka1ee
0. I
0.6
0.3
0.6
0.4
0.8
4.9
5.3
1.9
McLaurin
0.1
0.1
0.1
0. 1
0. 1
0.1
1.5
1.9
0.3
Colloidal
ferric phosphate(CFeP)
85.0
Immokalee
49.1
45.2
60.8
68.2
70.2
83.8
87.O
79.6
McLaurin
1.4
6.5
10.6
37.2
36.9
29.4
62.8
75.9
53.8
St rengi te(STR)
0.1
Immokalee
0.1
0.1
0.1
0.3
0.1
,0.1
0.1
0.1
McLaurin
0.1
0.1
0.1
0. 1
0.1
0.1
0.1
0.1
0. 1

102
(Table 11). Water and NH^OAc, which extract similar amounts of most com
pounds were by far the most effective tests. This is consistent with the
previous statement that for young seedlings the amounts of H^O-soluble P
are critical in determining P utilization. However, since H20-extractable
P was not related to long-term P availability (Table 10), it must be
concluded that the utilization of P compounds by seedlings grown in this
greenhouse study provided little indication of the long-term contribution
of such compounds towards meeting the P requirements of trees.
The large build up of H^O-soluble P following incubation of CA1P
and CFeP mixed with Immokalee soil (Table 16) is in agreement with the
finding of Taylor, Gurney, and Lehr (1963). These authors reported that
amorphous phosphates hydrolyzed rapidly to release P upon incubation in a
Hartsells fine sandy loam (pH 4.2). They also found that KTK hydrolyzed
at a much slower rate than amorphous phosphates, which is supported by
the data for KTK in Table 16.
General Discussion
It is apparent (Table 14) that the solubility of P compounds in
chemical extractants cannot be adequately categorized according to the
broad groupings of Ca-P, Al-P and Fe-P as has been the practice in past
studies (Thomas and Peaslee, 1973). Variation in solubility between mem
bers within such groups was as great as that between groups, and these
differences were of equal significance to plant utilization. Although
examination of the solubility of isolated P compounds, known to occur in
the soil and of known availability to plants, probably provides a better
indication of the prediction value of various extractants, the results
can still be misleading. This is because, in practice, these compounds
occur in the soil and the soil properties can appreciably alter their

103
solubility in any extractant; such soil effects were not predictable
(Table 14).
The results observed in the preliminary screening of soil-test
methods can be equated to some extent with the extractabi 1ity of P com
pounds shown in Table 14. Considering all 10 soils used in the prelimi
nary screening, the average amounts of P extracted by the soil-test
methods used were in the order NH^OAc < Truog < Olsen < HCl-h^SO^ < Bray.
This same order was observed for the extractabi 1 ity of CA1P and KTK, but
not CFeP, which was much more soluble in the Olsen than in HCI-h^SO^
extractant. This is consistent with Al-P being the dominant inorganic
form in these soils. The success of the Olsen, HCl-h^SO^, and Bray method
for predicting response over a 3~year growth period can be probably
attributed to these methods extracting forms of soil Al-P similar to CA1P
and KTK. The superiority of the Bray method at predicting response over
a 5~year growth period can probably be attributed to its extraction of
some less soluble, crystalline forms of Al-P, some forms of Fe-P, and its
ability to discriminate against insoluble, completely unavailable forms
of Al-P and Fe-P similar in nature to WA and STR.
In Florida, the NH^OAc method is used on a routine basis for pre
dicting the fertilizer requirements of agricultural crops (Page et al.,
1965). Because of the success achieved with this method by Pritchett
(1968), it has also been adopted for predicting P-fertilizer requirements
of trees. Data from this current study showed that NH^OAc is likely to
seriously underestimate the long-term supply of P to trees growing on
soils which contain P compounds similar to colloidal Al and Fe phosphates
and KTK. In the next section of this study, an attempt will be made to
calibrate soil-test methods other than NH^OAc and which theoretically
should prove superior to NH^OAc for forestry purposes.

Another point of practical importance, illustrated by the results
from the greenhouse trial with P compounds, concerns the use of RP in
forestry. The use of RP is now recommended on certain sites in the
Coastal Plain region (CRIFF annual report, 1373) where either leaching or
excess retention of soluble forms of P fertilizers are likely to occur.
In view of the critical role of soil pH in governing RP solubilization,
and hence its availability to pine trees, this source should not be rec
ommended on soils with a pH above 5*

105
Field Calibration of Selected Soil-Test Methods
Five soil-test methods, 1^0, NH^OAc, HCl-l^SO/j, Bray 1(3) and Bray
2, were selected for use in this calibration study. The amounts of P
extracted by each method from the four soil samples collected from the
control plot of each replicate of each trial (72 sites in total), are
given in Appendix Table 55* Tree heights in the absence of P fertilizer
and relative heights of slash pine on each of these 72 sites after 1, 3,
and 5 years' growth are shown in Appendix Table 7- The raw plot data
from which these results were obtained were too voluminous for inclusion
in this dissertation. The data are on file in the Forest Soils Section
of the Soil Science Department, University of Florida, Gainesville.
Relationships Between Soil-Test Values and Relative Height
The effectiveness (R^ values) of the five methods at predicting
relative heights, on the basis of P extracted from the surface 20 cm of
soil, showed similar, though less clearly defined trends to those found
in the preliminary screening (Table 17)* The values for water-extract
able P were most closely related to relative height of trees after 1
year's growth in the field, while P extracted by the HCl-H^SO^ and both
Bray methods tended to be most closely correlated with relative height
after 3 and 5 years' growth. The HCl-l^SO^ method was the most effective
overa 11.
As was the case for the soils used in the preliminary screening,
there was a poor relationship between l^O-extractable P and P extracted
by methods which removed larger quantities of P from the soil (Table 18).
The amounts of phosphate extracted by the HCl-H^SO^, Bray 1(3), and
Bray 2 methods were closely related, even though the HCl-f-^SO^ method
extracted appreciably less P than either of the two Bray methods. How-

106
ever this was consistent with results reported in an earlier section
which showed these methods solubilized similar forms of P, although to
different degrees.
Unlike the resuits found in the preliminary screening, the effec
tiveness of the soil-test methods which extracted the greatest amounts
of P did not improve with increases in duration of the growing period.
In fact, values for all soil-test methods, except those for the Bray 2
method after 3 years, decreased with increasing growth periods from 1 to
5 years. Also R^ values for the relationships derived using all 72 sites
were considerably smaller than for those derived using only 10 sites in
the preliminary screening. However, the 10 sites used in the preliminary
screening were carefully selected to represent a range of P-responses and
replicates showing soil variability were excluded. This was not the case
when all sites were used so that soil variability within certain sites
probably contributed to the lower R values. Soil samples were
collected only from the control piots under the assumption that each ade
quately represented the experimental area of 12 plots.
The general decline in effectiveness of soil-test methods with in
creasing growth period can probably be attributed to several factors, or
their combination: (a) Results of analysis of foliage samples collected
from all plots by CRIFF personnel after the fifth year showed that, at
several sites, P concentrations in the foliage of trees in the highest P
treatments were at or below the accepted critical value of 0.085 % P.
Thus, the computed relative heights at age 5 years for these sites
probably underestimated the true responsiveness obtainable with adequate
P-fertilizer applications. (b) As trees grow older their root systems
exploit greater depths and volumes of soil. Consequently, P concentra
tions at depth greater than 20 cm may become important in determining

107
Table 17- Relationships between soil-test values at three soil depths
and relative height of slash pine after 1, 3. and 5 years'
growth on 72 field sites
Soi1-test
Soil
Relative height
method
depth
1 year
3 years
5 years
em R^
Ho0
0-20B+
0.322**
0.149**
0.120**
z
0-20
0.306**
0.132**
0.100**
20-40
0.165**
0.132**
0.074*
40-60
0.179**
0.088*
0.024
NH.OAc
0-20B
0.281**
0.157**
0.084*
*4
0-20
0.270**
0.137**
0.072*
20-40
0.174**
0.172**
0.086*
40-60
0.194**
0.203**
0.114**
HC1H0SO.
0-20B
0.306**
0.246**
0.143**
0-20
0.264**
0.224**
0.126**
20-40
0.199**
0.233**
0.127**
40-60
0.317**
0.323**
0.208**
Bray 1(3)
0-20B
0.259**
0.221**
0.113**
0-20
0.235**
0.198**
0.095**
20-40
0.191**
0.140**
0.052
40-60
0.300**
0.228**
0.117**
Bray 2
0-20B
0.188**
0.210**
0.123**
0-20
0.152**
0.178**
0.094**
20-40
0.150**
0.229**
0.132**
40-60
0.282**
0.362**
0.251**
t
0-20B represents soil samples collected from within the bedded tree
row.
* Significant at the 5% level, using the mode Y=b logX + c.
**Significant at the 1% level.

108
Table 18. Simple correlation coefficients(r) between amounts of P
extracted from the surface 20 cm of soil(bedded area) by
five soil-test methods
Methods H20 NH^OAc HCl-^SO^ Bray 1(3) Bray 2
r
H20
1.000
0.684**
0.081
-0.058
-0.077
NH^OAc
1 .000
0.721**
0.606**
0.587**
HCl-H2S02j
1.000
0.972**
0.968**
Bray 1(3)
1.000
0.996**
Bray 2
1.000
Mean extractable
P, ppm
1.76
2.39
6.96
17.93
20.82
** Significant at the 1% level.
Table 19- Separation of 72 field sites into response quadrants using
the technique of Cate and Nelson (1965) and a critical HC1 -
i^SO^-extractable P value of 5 ppm
T ree
HC1-H2S04 < 5
ppmP
HC1-HjSO^ > 5
ppmP
Correct*
Age
Response No
response
Response No
response
prediction
1
34
12
9
17
%
70.8
3
34
12
6
20
75.0
5
27
19
6
20
65.3
* Correct prediction = Response(<5ppmP) + No response(>5ppmP) x ]0(J
Total number of sites

109
the responsiveness over longer growth periods. This might cause a dete
rioration in the relationship between topsoil-P values and responsiveness,
particularly where topsoil and lower horizon-P values are not closely
related. This aspect is examined in some detail below. (c) With in
creasing age the relative requirements of trees for such factors as nu
trients other than P and moisture may change. It can be speculated that
some factors may not limit early responsiveness to P fertilizer, but may
do so at a later stage of growth. (d) Height measurements tend to become
less sensitive indices of growth performance with increasing age of trees
so that by age 5. these may not adequately reflect true growth response
to P fertilizer.
The relationships between P extracted by HC1-H2S0^ from the
0-20 cm depth and relative height of slash pine 1, 3, and 5 years after P
fertilization are shown in Figs. 4, 5, and 6, respectively. These rela
tionships are strikingly similar to those reported by Wells et al. (1973)
between response to P fertilization of loblolly pine and HCl-H^SO^--.
extractable P. Applying the Cate and Nelson (1965) technique of deter
mining the critical level of soil P, it was found that a soil-test value
of 5 ppm P provided the best separation of responsive and nonresponsive
sites, assuming a relative height value of 90% or above represented no
significant response. The separation of the 72 field sites into response
quadrants using values above or below 5 ppm P and 90% relative height are
shown in Table 19. The data in Table 19 illustrate three points. First,
the use of 5 ppm of HC1-H^SO^-extractable P provides a reasonably good
separation between responsive and nonrespons ive sites. Second, as the
growth period increases, the number of responsive sites with extractable
P > 5 ppm decreased. This can probably be attributed to some of these
sites having low water-soluble P, which resulted in a tree response to

applied P during the early growth period which disappeared over longer
growth periods as the trees made use of the less soluble, HCl-l-^SO^-
extractable P. Third, the number of nonresponsive sites with extract-
able P < 5 ppm increased with an increase in the growth period from 3 to
5 years. The probably explanations for this were discussed earlier.
Values of P extracted by the HC1-H2S0^ method corresponding to a
relative height of 30% using the log transformed models shown in Figs. 4,
5, and 6 were above the critical value of 5 ppm P determined by the Cate
and Nelson (1965) technique. However, these models tended not to fit
the data in the initialy steep part of the response curve. Quadratic
models, excluding all sites which tested above 10 ppm of HCl-H^O^-
extractable P, did however provide P values corresponding to 90% relative
height in the vicinity of 5 ppm (Fig. 7). This critical value appeared
to be essentially independent of the age of the trees.
Humphreys and Pritchett (1972), suggested on the basis of rela
tionships between soil-test values and foliar P concentrations of slash
pine that a combination of intensity and quantity measurements of soil P
should provide the best index of the P status of soils for slash pine.
Regression equations of relative height at age 1, 3, and 5 years on the
P extracted from the surface 20 cm of soil by the H2O (intensity) and
HC1-H2S0^ (quantity) soil-test methods showed that only at age 1 year
did both of these soil tests contribute significantly to explaining
variation in relative height (Table 20). At age 3 and 5 years, the coef
ficient for H20-extractable P was nonsignificant when included in a
multiple regression equation with HC1-H2S0^- extractable P. This is con
sistent with the hypothesis concerning the importance of H^O-soiuble P
in early growth as proposed earlier. It should, however, be appreciated
that the HC!-H2S0^ method does not provide an exclusive measure of the

RELATIVE HEIGHT,
Fig. A. Relationship between HC1-h^SO^-extractable P(X) in the surface 20 cm of soil and relative
height(Y) of slash pine 1 year after P fertilization.

Fig. 5 Relationship between HC1-H^SO^-extractable P(X) in the surface 20 cm of soil and relative
height(Y) of slash pine 3 years after fertilization.

RELATIVE HEIGHT,
HCI-H2SO4 EXTRACTABLE P, ppm
Fig. 6. Relationship between HC1-H^SO^-extractable P(X) in the surface 20 cm of soil and relative
height(Y) of slash pine 5 years after P fertilization.

relative height.
11
Fig. 7- Relationships between HC1-H^SO^-extractab1e P(X) in the
surface 20 cm of soil and relative height of slash pine
1, 3 and 5 years (Y1, Y3, and Y5) after P fertilization.

115
quantity factor since it removes h^O-soluble P in addition to certain
solid phase forms of soil-P. Thus, the HCl-H^SO^ method can probably be
considered to provide a composite index of the quantity and intensity
factors of soil-P supply.
Effect of soil sampling position and depth
The amount of P extracted from surface (0-20 cm) samples
collected from within the bedded tree rows was more closely related to
relative height than P extracted from surface samples collected from the
undisturbed interbed area (Table 17). This was not surprising since
nearly all the rooting activity of the young trees is within the bedded
area. All soil-test methods extracted more P from soils collected within
the beds than from those collected between beds. The increase in avail
able P following bedding has been recorded previously (Haines and
Pritchett, 1965). These microsite differences in available P have impli
cations for the operational use of soil tests in forestry. It may be
more convenient to collect soil samples from areas prior to bedding, in
which case critical values somewhat lower than those established for
samples from the bedded areas should be used. For soils collected in this
study, the relationship between P extracted by the HC1-H2S0^ method from
interbed samples (Y) and bedded samples (X) was:
Y = 1.11X 0.9* (r= 0.983)
Using this relationship, the HC1-H2S0^-extractable P value for the inter
bed samples, corresponding to 5 ppm P for the bedded samples, was 4.6 ppm
P.
The quantity of P extracted from samples collected from the
20-40 cm depth was generally less closely related to relative height than
P extracted from surface samples (Table 17). However, the amount of P

116
Table 20. Regression equations relating relative height of slash pine at
age I, 3, and 5 years (Y1, Y3, and Y5) to the log transformed
P extracted from the surface 20 cm of soil by the h^O (XI) and
HCl-f^SO^ (X2) soil-test methods
Regress ion
equations
R2
Y1
=
13.88**X1
+ 85.00
0.322**
Y1
=
19.63**X2
+ 71.53
0.306**
Y1
=
11.86**X2
+ 9.05**X1 + 77.25
0.395**
Y3
=
9.99**X1
+ 84.28
0.149**
Y3
=
18.65**X2
+ 71.83
0.246**
Y3
=
15.49X2
+ 3.68X1 + 74.15
0.259**
Y5
=
8.25**X1
+ 88.13
0.120**
Y5
=
13.07**X2
+ 79.27
0.143**
Y5
=
9.20* X2
+ 4.50X1 + 82.11
0.166**
* Significant
at the 5%
leve 1.
**Significant
at the 1%
level.
Table 21.
Simple correlation coefficients(r)
between
amounts of P
extracted by
the HC1-H?S0^ method
from two
soi1 positions
and three soil depths
Depth
o :
20B 0-20
20 40
40 60
0 20B+
1.000 0.983**
r
0.928**
0.468**
0-20
1 .000
0.934**
0.505**
o
-cr
i
o
CM
1 .000
0.614**
40 60
1 .000
^ Surface sample collected from within the bedded tree row.
**S ignif¡cant at the level.

117
extracted from the 40-60 cm samples by the HCl-h^SO^, and both Bray
methods was more closely related to relative height than P extracted
from surface samples. This was not the case for H^O-extractab1e P, but
was for NH^OAc-extractab1e P at ages 3 and 5 years. Multiple regressions
of relative height at each age on the amount of P extracted by the
HC1-H2S0^ method from each of the three depths showed only the coeffi
cient for P extracted from the 40-60 cm depth to be significant, indicat
ing that extractable P at al 1 three depths were significantly correlated.
These regressions are not shown, but the significant interrelationships
between P extracted from the different depths were confirmed (Table 21).
Since the levels of P extracted from all three depths were interrelated,
multiple regression techniques were inadequate to indicate whether P
from lower soil depths contributed significantly towards meeting the P
requirements of the young slash pine. However, data comparing the ef
fectiveness of extractable P in the surface soil with the amount of
extractable P within the effective rooting volume (computed from thS
depth to a limiting horizon, and extractable P and bulk density values
of soil above the limiting horizon) showed the latter to be a more ef
fective predictor of relative height (Table 22). The superiority of the
1 rooting-vo1ume P tended to become more pronounced with increasing age
of trees. This strongly suggested a significant contribution to the
tree's P supply from lower depths, where roots were capable of penetrat
ing free of impedence from restrictive horizons (spodic, mottled or
argi11ic).
Relationships Between Other Soil and Site Parameters and Relative Height
Multiple correlation coefficients (R^) for relationships between
a variety of soil and site properties and relative height are shown in

118
Table 23. Actual values for the parameters included in Table 23 are
given in Appendix Tables 5^, 55, and 56.
As prediction models, some of the relationships shown in Table 23
are superior to models involving use of extractable soil P. However,
many of the significant relationships can be attributed to the influence
of these variables on soil P. Silt + clay, NH^OAc-extractable Ca and A1,
available moisture, and pH were all significantly correlated with H2O-
extractable P (Table 2k). Since relative height is a measured response
to P fertilizer application in the field, it is most likely that the sig
nificant relationships shown in Table 23 exist because the parameters
are either related to available soil-P levels or in some manner influence
the response of the trees to added P fertilizer.
Depth to a limiting horizon (LH) and drainage class, which are
closely related (Table 2k) both showed increased values with increas
ing tree age, a result not shown by any other soil parameters, including
any form of extractable P. Depth to LH, although significantly
correlated with HC1-H2S0^, extractable P, is markedly superior to
HC1-H2S0^ -extractable P as a predictor of relative height, particularly
at age 5 years. A plot of depth to LH against relative height at age 5
years is shown in Fig. 8. Multiple regression equations of relative
height on depth to LH, HCl-H2S0^-extractable P (0-20 cm) and the squared
terms of these two parameters showed that whereas depth to LH and its
squared term were significantly related to relative height at ail three
measurement periods, the significance of the contribution of HC1-H2S0/J-
extractable P decreased with increasing age (Table 25). At tree age of
5 years, neither of the terms for HC1 H2S0t-P made a significant contri
bution towards accounting for the variation in relative height. An
explanation for these results probably lies in the previously observed

119
Table 22. Comparison of the effectiveness of P extracted from the sur
face 20 cm of soil and that extracted from within the effec
tive soil depth (volume) at predicting relative height of
slash pine after 1, 3 and 5 years' growth on 72 field sites
Soi1-test
method
Soi 1
sample
Relative height
1 year
3 years
5 years
r2
V
0 20 cm
0.306**
0.132**
0
0
0
*
*
Volume
0.372**
0.239**
0.191**
NH^OAc
0 20 cm
0.270**
0.137**
0.072*
Volume
0.253**
0.220**
0.142**
hci-h2so4
0 20 cm
0.264**
0.224**
0.126**
Volume
0.274**
0.322**
0.211**
Bray 1(3)
0 20 cm
0.235**
0.198**
0.095**
Volume
0.282**
0.284**
0.1-62**
Bray 2
0 20 cm
0.152**
0.178**
0.094**
Volume
0.211**
0.303**
0.202**
* Significant
at the 5% level,
using the model
Y=b log X + c
** Significant at the 1% level.

120
Table 23. Relationships between selected soil and site properties and
relative height of slash pine 1, 3 and 5 years after P
fertilization on 72 sites
Soil or site^
property
Relative height
1 year
3 years
5 years
r2
pH
0.089*
0.083*
0.097*
Organic matter
0.013
0.038
0.008
Nitrogen
0.149**
0.216**
0.143**
CEC
0.020
0.030
0.035
NH^0Ac(pH 4.8) extractable
Ca
0.217**
0.175**
0.252
Mg
0.019
0.005
0.029
K
0.047
0.044
0.019
A1
0.257**
0.145**
0.144
Si 11 + clay
0.347**
0.225**
0.149**
Available moisture
0.227**
0.095
0.017
Depth of A1 horizon
0.056
0.050
0.032
Depth to limiting horizon
0.281**
0.314**
0.345**
Drainage class
0.189**
0.237**
0.274**
Except for depth functions and drainage class, all properties are for
the surface 20 cm of soil.
* Significant at the 5% level, using the model Y=aX + bX^ + c.
**Signif¡cant at the \% level.

Table 24. Simple correlation coefficients(r) between selected soil and site properties of 72 field sites
Soil or site
property
Extractable P
pH
NH^OAc
Silt +
clay
Depth
to LH
Drainage
class
y
HCl-H2S0i<
Ca
A1
pH
-0.713**
0.250*
1.000
-
-
-
-
Organic matter
0.168
-0.097
-0.394**
0.036
0.041
0.360**
-0.227
-O.35I**
N i t rogen
-0.047
-0.157
-0.293*
-0.004
0.207
0.458**
-O.289*
-O.387**
CEC
0. 101
-0.124
-0.295*
0.263*
0.202
0.494**
-0.207
-O.3I3**
NH^OAc-Ca
0.249*
-0.113
-0.033
1 .000
-
-
-
-
-Mg
0.469**
-0.075
-O.337**
0.727**
0.092
0.344**
-0.168
-0.185
-K
0.077
-0.199
0.009
0.535**
0.377**
0.639**
0.067
0.099
-A 1
-0.623**
-0.126
0.401**
0.019
1 .000
-
-
-
Silt + clay
-0.424**
-O.263*
0.271*
0.274**
0.715**
1 .000
-
-
Ava i Iable moisture
-0.329**
-O.I80
0.169
0.215
0.587**
0.907**
-0.030
-0.055
Depth of A1 horizon
0.141
0.232*
-0.007
-0.216
-0.177
-0.252*
0.122
-0.063
Depth to LH^
-0.070
0.292*
0.602**
0.021
-0.161
-0.049
1 .000
Drainage class
-0.224
0.188
0.697**
0.012
-0.017
0.077
0.852**
1 .000
* Significant at the 5% level.
**Significant at the 1% level.
' Depth to limiting horizon.
N>

Fig. 8. Relationship between depth to limiting horizon(X) and relative
height(Y) of slash pine 5 years after fertilization.
ro

123
Table 25. Regression coefficients for multiple regression equations of
relative height on depth to LH, HCl-H-SO^-extractable
P(0-20 cm) and the squared terms of these two parameters
Source
Coefficients
Significance
R2
Relative height at
age
1 year
0.402**
Intercept
40.96
"k'k
Depth to LH
1.43
**
(Depth to LH)2
-0.01
**
hci-h2so4-p
1.68
XX
(hci-h2so4-p)2
-0.03
**
Relative height at
age
3 years
0.378**
Intercept
39.99
"kk
Depth to LH
1.40
XX
(Depth to LH)2
-0.01
**
hci-h2so4-p
1.18
*
(hc'i-h2so4-p)2
-0.02
NS
Relative height at
age
5 years
0.357**
Intercept
44.99
XX
Depth to LH
1.45
XX
(Depth to LH)2
-0.01
kk
HC1-H SO -P
2 o
0.42
NS
(hci-h2so4-p)z
-0.01
NS
* Significant at the 5%
1evel.
**Significant at the \%
level.

124
Table 26. Regression coefficients for variables included in 'best fit*
multiple regression
site parameters and
equations of.
the squared
relative height on soil and
terms of these parameters
Source
Coefficients
Sign!f¡canee
R2
Relative height at age 1 year
0.582**
Mean
57.426
Silt + clay
-0.869
JL .L
Depth to LH
1.141
XJU
(Depth to LH)2
-0.008
**
CEC
1.005
**
(NH^OAc-Al)2
0.001
k
Relative height at age 3 years
0.522**
Mean
66.084
Silt + clay
-1.732
**
Drainage class
18.310
**
CEC
1.225
**
(Silt + cl ay)^
0.022
JL
(Drainage class)2
-0.207
JL
Relative height at age 5 years
0.522**
Mean
54.917
Depth to LH
1.813
**
(Depth to LH)2
-0.014
JU JL
Silt + clay
-1.840
JL JL
(Silt + clay)
0.034
JL JL
(h2o-p)2
-0.231
JL
* Significant at the 5% level.
**Signif¡cant at the \% level.

125
increased importance of P from lower soil depths with increase in tree
age. Apparently, because of its correlation with topsoil P, which in
turn is correlated with P at lower depths, depth to LH provides an inte
grated measure of available soil P within the rooting zone of the older
trees, of which topsoil P is only one component.
All the soil and site properties listed in Table 23 plus P
extracted by the H^O and HCl-H^SO^ methods and the squared terms of each
of these parameters were used as independent variables in a stepwise
regression procedure designed to maximize R^ (Barr and Goodnight, 1972)
for models with relative heights as the dependent variables. The 'best
fit' models, in which the coefficients for all the included independent
variables were significant at least at the 5% level are shown in Table 26.
Although these models could be used as prediction models, they provide no
information on causal relationships. It is interesting, however, that
the majority of the independent variables included are those which were
significantly correlated with P extracted by either the HCl-I^SQ/j or H^O
methods (Table 24).The relatively low R^ values for these 'best fit'
models suggest either unmeasured parameters are contributing to variation
in relative height, or the relative heights derived from the raw plot
data are not completely true reflections of the responsiveness of the
control-plot soils because of 'within site' variability or lack of
optimum fertilizer application.
Thus far, the data have shown that various site or soil physical
parameters provide better indications of the P responsiveness of the
sites used in this investigation than extractable-P values. However,
since these parameters are unlikely to reflect causal relationships, ex
treme caution should be used in extrapolating the results to other sites.
Where relationships are known to be causal, such as between soil-P levels

126
and response to P fertilizer, they can be used with, some confidence for
prediction purposes on sites with soil properties within the range en
countered in the calibration work.
Relationships Between Soil-Test Values and P-Fertil?zer Requirements
Multiple correlation coefficients (R^) for relationships between
the amount of P extracted from the soil by the five soil-test methods and
the amount of P fertilizer required to achieve 90, 95, and 100% of
maximum height growth after 1, 3, and 5 years' growth in the field are
presented in Table 27. Actual amounts of fertilizer required to achieve
these growth rates at each of the 72 sites are given in Appendix Table 58.
The effect of increasing growth period on the effectiveness of the
five soil-test methods as predictors of fertilizer requirements was
similar to that observed for relative height. The amounts of P extracted
by a weak extractant (NH^OAc) were most closely related to P requirements
of 1-year-old trees, while at ages 3 and 5 years, P extracted by the
stronger extractants was most closely related to fertilizer P require
ments. While the HCl-i^SO^ method was the most effective predictor of
relative height at ages 3 and 5 years, there was a tendency for the Bray
methods to be the most effective predictors of fertilizer requirements of
trees at ages 3 and 5 years.
As was the case for relative height, only a very small proportion
of the variation in fertilizer requirements at age 5 years was explained
by extractable P in the surface soil. Factors such as the contribution
of P from lower soil depths, the insensitivity of the height measurement
at age 5 years, inadequate fertilizer applications on some sites, and
the increased length of time for other site factors to influence the ef
fectiveness of fertilizer may have contributed to variation unaccountable

Table 27.
Relationships between
soil-test values at
three soil
depths and P
fertilizer
requ1 red
to achieve
90, 95,and
100% of maximum height
growth after 1, 3
and 5 years'
growth on
72 field
s i tes
Fert i
1izer requirements
Soil
1 year
3 year
5 year
Depth
90
95
100
90
95
100
90
95
100
cm
0-20B
p2
0.271**
0.250**
0.041
0.039
h2o
0.000
0.001
0.005
0.005
0.002
0-20
0.209**
0.192**
0.035
0.035
0.000
0.000 .
0.005
0.003
0.003
20-40
0.223**
0.206**
0.041
0.070*
0.010
0.002
0.012
0.004
0.001
40-60
0.318**
0.298**
0.083**
0.100**
0.019
NH/,0Ac
0.023
0.007
0.001
0.002
0-20B
0.359**
0.364**
0.094**
0.116**
0.034
0.014
0.014
0.009
0.001
0-20
0.332**
0.334**
0.106**
0.098**
0.023
0.012
0.013
0.004
0.000
20-40
0.265**
0.266**
0.082*
0.164**
0.127**
0.090*
0.038
0.026
0.056*
40-60
0.242**
0.234**
0.082*
0.199**
0.129**
HCl-H2SO/4
0.110**
0.056*
0.031
0.053
0-20B
0.258**
0.277**
0.085*
0.189**
0.118**
0.071*
0.052
0.022
0.044
0-20
0.181**
0.199**
0.074*
0.173**
0.093**
0.070*
0.054*
0.015
0.029
20-40
0.180**
0.182**
0.060*
0.235**
0.224**
0.118**
0.090*
0.053*
0.082*
40-60
0.239**
0.240**
0.070*
0.285**
0.191**
Bray 1(3)
0.102**
0.135**
0.075*
0.074*
0-20B
0.198**
0.224**
0.076*
0.194**
0.144**
0.095**
0.056*
0.023
0.065*
0-20
0.164**
0. 180**
0.070*
0.170**
0.120**
0.091**
0.047
0.014
0.050
20-40
0.303**
0.309**
0.130**
0.164**
0.153**
0.063*
0.026
0.016
0.043
40-60
0.328**
0.335**
0.123**
0.217**
0.136**
Bray 2
0.057*
0.053*
0.041
0.039
0.069*
0-20B
0.147**
0.170**
0.048
0.192**
O.56**
0.Ill**
0.070*
0.022
0-20
0.103**
0.120**
0.051
0.164**
0.126**
0.111**
0.058*
0.012
0.052
20-40
0.124**
0.124**
0.046
0.241**
0.252**
0.126**
0.114**
0.058*
0.096**
40-60
0.185*
0.182**
0.048
O.315**
0.230**
0.109**
0.188**
0.097**
0.087*
* Significant at the 5% level. ** Significant at the level, using the model Y=b logX + c.
Ni
Vl

FERTILIZER REQUIREMENT(90%), kgP/ha.
128
Fig. 9- Relationships between HC1-f^SO,-extractab1e P(X) in the
surface 20 cm of soil and amount of P fertilizer (CSP)
required to achieve 90% of maximum height growth over
periods of J, 3, and 5 years (Y1, Y3, and Y5)
following fertilization.

FERTILIZER REQUIREMENTS0/), kg'ha
129
60f \
10
oY(1)=-33.27logX + 47.33 (R2=0.277)
oY(3)--21.54logX + 41.10 (2=0.188)
aY(5)=-9.01 log X + 27.36 (R2=0.022)
0 2 4 6 8 10
HCI- H2SC>4 EXTRACTABLE P, ppm
Fig. 10. Relationships between HC1-H^SO^-extractable P(X) in the
surface 20 cm of soil and amount of P fertilizer (CSP)
required to achieve 95% of maximum height growth over
periods of 1, 3, and 5 years (Y 1, Y3. and Y5)
following fertilization.

130
by extractable P in the top 20 cm of soil. Some of the possible factors
are examined below.
The level of P extracted by the five soil-test methods tended to
be equally effective at predicting amounts of P fertilizer required to
achieve either 90 or 95% of maximum height growth (Table 27). There was
a pronounced decline in their effectiveness at predicting fertilizer
required to achieve 100% of maximum height growth, particularly at ages
1 and 3 years. This was probably a function of the method of determining
P requirements, rather than a true deterioration in the value of soil P
for predicting P-fertilizer requirements. As mentioned in the Materials
and Methods section, where response curves could not be fitted to data
from any site, fertilizer requirements were taken as the actual field ap
plication rate which provided the greatest height growth. At several
sites, small differences of less than 5% (relative height > 95%) were
observed between the control and highest P treatment; these were
recorded as requiring this amount of P fertilizer to achieve 100% of
maximum grov/th. Many of these small differences were probably nonsignif
icant and unrelated to soil-P levels.
The relationships between P extracted by the HCl-h^SO^ method
(0-20 cm bed samples) and the amount of P fertilizer required to achieve
90 and 95% of maximum height growth are shown in Figs. 9 and 10,
respectively. The critical value of 5 ppm extractable P, proposed from
the relationships with relative height, also provided a reasonable
distinction between sites requiring a practical P-fertilizer application
(>10-20 kg P/ha) and those which did not. The prediction curves
for ages 1 and 3 years were somewhat similar, but the fertilizer
requirements over the 5-year period were considerably lower for any soil-
test value below 5 ppm, than the requirements over the 1- and 3~year

131
growth periods. This may be a valid situation resulting from a reduced
requirement for high concentrations of P in the surface soil with roots
exploiting greater volumes of soil as the trees increase in age. It may
also result from the fact that height is an insensitive indicator of
response of older trees, and/or that inadequate P-fertilizer applications
were made in some of the field trials to maintain a true response for the
full 5 years. Since the models for age 5 years shown in Figs. 9 and 10
are not significant, and because of the doubt concerning the validity of
the data from which they were derived, these age 5 prediction models are
not satisfactory for predicting operational fertilizer requirements.
Effect of soil sampling position and depth
The effect of sample position or depth on the prediction of fertil
izer requirements from extractable-P levels was similar to that observed
for the prediction of relative height (Table 27). The improved predic
tion resulting from use of soil samples from deeper profile depths
compared to the use of surface soil samples appears to have no simple ex
planation. Since P extracted from all depths and positions was signifi
cantly interrelated (Table 21) and subsoil extractable P is less likely
to be subject to short-term fluctuation associated with variable climatic
conditions which affect mineralization, P extracted from lower depths may
have provided a better indication of the average P status of the surface
soil than the level of P extracted from the surface soil at a particular
time of col lection.
Comparison of the effectiveness of P extracted from the surface
20 cm of soil with that extracted from within the effective soil depth
(Table 28), again revealed evidence of the increasing importance of soil
P in lower,, root-penetrable horizons with increasing age of trees. The

Table 28. Comparison of the effectiveness of P extracted from the surface 20 cm of soil with that extracted
from within the effective soil depth (volume) at predicting the fertilizer P required to achieve
90, 95>and 100% of maximum height growth after 1, 3, and 5 years' growth on 72 field sites
Fertilizer P requirements
Soi 1
Depth
1 year
3 years
5 years
90
95
100
90 95
100
90
95
100
c>2
0-20 cm
Volume
0.209**
0.237**
O.192**
0.220**
0.035
0.034
H20
0.035 0.000
0.084* 0.006
0.000
0.003
0.005
0.042
0.003
0.018
0.003
0.001
0-20 cm
Volume
0.332**
0.251**
0.334**
0.239**
0.106**
0.071*
NH.OAc
4
0.098** 0.023
0.179** 0.077*
0.012
0.034
0.013
0.056*
0.004
0.015
0.000
0.032
0-20 cm
Vo 1 ume
0.181**
0.158**
0.199**
0.172**
0.074*
0.043
HC1-H2S04
0.173** 0.093**
0.256** 0.172**
0.070*
0.125**
0.054*
0.128**
0.015
0.056*
0.029
0.073*
Bray 1(3)
0-20 cm
Volume
0.164**
0.199**
0.180**
0.215**
0.070*
0.066*
0.170** 0.120**
0.245** 0.179**
0.091**
0.117**
0.047
0.098**
0.014
0.045
0.050
0.076*
0-20 cm
Volume
0.103**
0.109**
0.120**
0.120**
0.051
0.030
Bray 2
0.164** 0.126**
0.257** 0.204**
0.111**
0.146**
0.058*
0.149**
0.012
0.061*
0.052
0.092**
*Significant at the 5% level, using the model Y=b logX + c.
1
**Significant at the 1% level.

133
superiority of P extracted from within the effective soil depth compared
to that extracted from the surface soil was more apparent for the 3 and
5~year growth periods.
Relationships Between Other Soil and Site Parameters and P-Fertilizer
Requi rements
Drainage class was the only parameter which provided a better pre
diction than extractable-soil P of the amount of P fertilizer required
to achieve 95% of maximum height over any growth period (Table 29). Most
of the other parameters recorded above as significantly related to
fertilizer requirements were also significantly related to extractable P
(Table 2b).
The hypothesis that extractable soil-P values provide a better
indication of P-fertilizer requirements, within groups of soils with
similar P-retention capacities, was examined. This was done by determin
ing relationships between extractable-soil P and P requirements within
classes of soils grouped according to the amounts of their NH^OAc-s.
extractable A1. Aluminum extracted by NH^OAc has been reported as sig
nificantly correlated with the P-retention capacity of soils in the
Coastal Plain (Yuan and Breland, 1969)- Results (Table 30) showed that
after 1 year, the significance level of the regression of P-fertilizer
requirements (95%) on P extracted by the HCl-i^SO^ method and its
squared term was less within P-retention categories than for all soils
combined. However, after 3 years, the significance level of the regres
sions for the two soil groups with higher levels of NH^OAc-extractable
A1 was greater than that for all soils combined. After 5 years, the
level of significance of the regressions for all three groups was
greater than that for all soils combined, although none reached even
the 10% level of significance. These data indicate that the P-

134
Table 29. Relationships between selected soil and site properties and
P fertilizer required to achieve 95% of maximum height after
1, 3, and 5 years' growth on 72 field sites
Soil or site
property
Frtilizer
requirements
(95%)
1 year
3 years
5 years
r2
pH
0.073
0.126**
0.046
Organic matter
0.003
0.199**
0.066
Nit rogen
0.083*
0.140**
0.069
CEC
0.011
0.144**
0.061
NH^OAc (pH !*.8) extractable
-Ca
0.069
0.003
0.031
-Mg
0.000
0.096*
0.007
-K
0.179**
0.079
0.000
-A1
0.116*
0.005
0.025
Silt + clay
0.268**
0.087*
0.051
Available moisture
0.225**
0.098*
0.064
Depth of A1 horizon
0.015
0.009
0.002
Depth to LH
0.135**
0.036
0.062
Drainage class
0.038
0.067
0.109*
* Significant at the 5% level,
using the model
Y=aX + bX2 +
c.
**Significant at the 1% level.

135
Table 30. Relationships between HCl-H^SO^-extractable P(0-20 cm) and P
fertilizer required to achieve 95% of maximum height on groups
of soils classed according to their amount of NH^OAc (pH 4.8)-
extractable A1
NH OAc(pH 4.8)
A1
Number
of soils
P ferti
lizer requirements (95%)
1 year
3 years
5 years
ppm
-
<20
22
0.021(81.50)+
o.189(13.49)
0.138(24.30)
20 to 80
34
0.3S0( 0.07)
0.297( 0.45)
0.043(51.37)
o
CO
A
16
0.609( 0.26)
0.533( 0.72)
0.256(14.43)
Whole A1 range
72
0.205( 0.06)
0.117( 1.36)
0.016(58.60)
^ Values in parentheses represent the significance level of the regression
in percent, using the model Y=aX + bX2 + c, where X = soil test-value.

136
retention capacity of the soil does modify the relationship between
extractable soil P and P requirements over longer periods than one year
following fertilization. However, the improvement in the significance
of the regressions within groups of soils with similar NH^OAc-extract-
able A1 over that for all soils combined does not appear sufficient to
warrant use of separate regression equations for prediction purposes.
'Best fit1 models derived using the procedure and soil and site
parameters outlined previously are given in Table 31- As mentioned
previously, such equations are of interest as prediction models and for
seeing how much variation in the independent variable can be accounted
for by the independent parameters, but not as indicators of causal
relationships. The most noteworthy feature of these models is their
low R^ values, particularly for the models predicting fertilizer
requirements for 3" and 5-year growth periods. This suggests either un
measured parameters were contributing significantly to variation in
fertilizer requirements, or, as suggested previously, the calculated
fertilizer requirements did not reflect the true fertilizer requirements
of the sites.
Relationships Between Soil-Test Values and Height
In order to convert predicted relative height response into an
absolute growth response, which is the criteria needed for economic
justification of fertilizer additions, a prediction of height (height
growth in the absence of P fertilizer) is required. Relationships be
tween soil-test values in the surface soil and height (Table 32) showed
that only H^O-extractable P provided a reasonable estimate of height.
The R^ value for this relationship declined with an increase in the
growth period, as was also observed in the preliminary screening. It was

137
Table 31. Regression coefficients for variables included in 'best fit1
multiple regression equations of P fertilizer required to
achieve 95% of maximum height on soil and site parameters and
the squared terms of these parameters
Source
Coefficients
Signif¡canee
R2
P-fertilizer requirements over
1 year
0.450**
Mean
45-757
h2o-p
-13.333
-
(H20-P)2
1.280
^A
NH^OAc-Al
-0.279
**
hci-h2so2i-p
-1.776
*
(hci-h2so4-p)2
0.028
*
P-fertilizer requirements over
3 years
0.249**
Mean
16.590
CEC
1.726
hci-h2so^-p
-0.629
JUJU
(h2o-p)2
0.306
aT
P-fertilizer requirements over
5 years
0.109*
Mean
51.284
Drainage class
-23.890
JL
(Drainage class)
3.510
*
* Significant at the 5% level.
**Significant at the 1% level.

138
Table 32. Relationships between soil-test values at three soil depths
and height of slash pine in the absence of P fertilizer
after 1, 3, and 5 years' growth on 72 field sites
Soi 1-test
So i 1
Height
method
depth
1 year
3 years
5 years
.... r2
H 0
0-20B
0.370**
0.297**
0.193**
2
0-20
0.369**
0.324**
0.209**
20-40
0.195**
0.148**
0.062*
40-60
0.139**
0.094**
0.033
NH^OAc
0-20B
0.189**
0.179**
0.086*
0-20
0.193**
0.195**
0.084*
20-40
0.040
0.045
0.014
40-60
0.033
0.058
0.017
HC1-H.S0,
0-20B
0.024
0.087*
0.072*
4. H
0-20
0.012
0.082*
0.063*
20-40
0.001
0.022
0.009
40-60
0.030
0.069
0.038
Bray 1(3)
0-20B
0.003
0.046
0.042
0-20
0.002
0.049
0.040
20-40
0.024
0.015
0.001
40-60
0.076*
0.053*
0.018
Bray 2
0-20B
0.015
0.007
0.014
0-20
0.018
0.006
0.010
20-40
0.008
0.004
0.001
40-60
0.007
0.044
0.023
* Sign i ficant at
the
5% level,
using the model
Y = b logX + c.
**Significant at
the
1 % 1 eve 1.

139
Table 33. Comparison of the effectiveness of P extracted from the sur
face 20 cm of soil and that extracted from within the effec
tive soil depth (volume) at predicting height of slash pine
after 1, 3, and 5 years' growth on 72 field sites
Soi1-test
method
Soi 1
sample
Height
1 year
3 years
5 years
r2
H,0
0-20 cm
0.369**
0.324**
0.209*
Volume
0.238**
0.329**
0.238**
NH^OAc
0-20 cm
0.193**
0.195**
0.084*
Vo1ume
0.071*
0.122**
0.062*
HC1 -H2S0it
0-20 cm
0.012
0.082*
0.063*
Volume
0.000
0.059*
0.055*
Bray 1(3)
0-20 cm
0.002
0.049
0.040
Volume
0.006
0.056
0.041
Bray 2
0-20 cm
0.018
0.006
0.010
Volume
0.012
0.024
0.028
* Significant at the 5% level, using the model Y = b logX + c
**Signif¡cant at the 1? level.

suggested in the preliminary screening, that the failure of P extracted
by stronger extractants, such as the HCl-H^SO^, Bray 1(3), and Bray 2
methods, to provide a reasonable estimate of height over longer growth
periods was due to factors other than inadequate available soil P re
stricting growth. The failure of extractable P within the effective soil
depth to improve the estimate of height compared to extractable P in the
surface 20 cm (Table 33), lends support to this contention.
Relationships Between Other Soil and Site Parameters and Height
Soil pH, NH^OAc-extractab1e Ca and A1, silt + clay, available
moisture, depth to LH, and drainage class were all significantly related
at the 1% level to tree height at age 1 year (Table 34). However, all
these parameters, except depth to LH and drainage class, were significant
ly correlated with H^O-extractable P (Table 24), suggesting they
influenced growth only through their association with H^O-extractable P.
As with H^O-extractable P, the value of these parameters as predictors of
height decreased with increasing growth period. The value as predictors
of growth of both drainage class and depth to LH increased with increas
ing growth period. These two site factors were closely related (Table
24). The close correlation between drainage class and depth to LH has
been observed on other soil groups (DeMent and Stone, 1968).
For prediction purposes, depth to LH probably has an advantage
over drainage class since its measurement involves a less subjective as
sessment than drainage class. The relationship between depth to LH and
l " 1 . , - .. ......
height of slash pine after 5 years' growth is shown in Fig. 11. Although
not shown, the relationship between depth to LH (X, cm) and height of
slash pine after 3 years' growth (Y, cm) was described by the quadratic
equation:

141
Y = 26.24 + 6.25X 0.05X2 (R2 = 0.323)
Barnes and Ralston (1955), in an examination of soil factors affecting
growth of slash pine in Florida, also found depth to a fine-textured
horizon and depth to mottling to be significantly related to site quality
of slash pine. The form of the relationship reported by these authors,
was the same as that shown in Fig. 11. Height increased with increasing
depth to limiting horizon up to a maximum, thereafter a further increase
in depth was associated with a decline in height. Barnes and Ralston
(1955) interpreted this relationship in terms of the effect of soil
moisture and aeration on growth. The optimum depth to a fine-textured
horizon or mottling of 20 to 30 inches (51 to 76 cm) reported by these
authors corresponds closely to that shown in Fig. 11.
In order to examine the possibility that extractable-P levels may
have contributed significantly towards variation in height within
specific levels of depth to LH, multiple regression equations of height
on depth to LH, H^O-extractable P, and the squares of these terms were
run. Results (Table 35) showed that only H^O-P made a significant
contribution towards accounting for variation in height at age 1 year.
However, as the growth period was increased from 1 to 3 to 5 years,
H^O-P became less significant and depth to LH more so. Since both H^O-P
and depth to LH were significantly related to height at all ages, and
yet these two variables were not linearly correlated (Table 24), inter
pretation of these data appeared difficult. However, a regression of
H^O-P on depth to LH and its squared term was highly significant
(R2 = 0.384). Thus, it appears probable that the significant relation
ship between H20-extractab1e P and height at age 5 years was in part due
to the relationship of H20-extractable P to depth to LH. This is con
sistent with the hypothesis formulated from the results of the prelimi-

Table 34. Relationships between
height of slash pine
field sites
selected s
after 1, 3,
oil and site
and 5 years'
properties and
growth on 72
Soil or site
Height
property
1 year
3 years
5 years
r2
pH
0.412**
0.224**
0.142**
Organic matter
0.083*
0.074
0.096*
N i trogen
0.029
0.074
0.017
CEC
0.056
0.089*
0.105*
NH^OAc (pH 4.8) extractable
-Ca
0.134**
0.151**
0. J17*
-Mg
0.113*
0.086*
0.076
-K
0.092*
0.067
0.101*
-A1
0.307**
0.296**
0.148**
Silt + clay
0.232**
0.240**
0.105*
Available moisture
0.213**
0.207**
0.059
Depth of A1 horizon
0.008
0.094*
0.154**
Depth to LH
0.231**
0.323**
0.424**
Drainage class
0.306**
0.497**
0.456**
* Significant at the 5% level, using the model Y = aX + bX2 + c.
**Signif¡cant at the 1% level.

HEIGHT AT AGE 5 YEARS,
Fig. 11. Relationship between depth to limiting horizon(X) and the height
of slash pine after 5 years' growth.

144
Table 35* Regression coefficients for multiple regression equations of
relative height on depth to LH, F^O-extractab1e P (0-20 cm)
and the squared terms for these two parameters
Source
Coefficients
S ignifi canee
R2
Height at age 1 year
Intercept
42.447
**
0.460**
Depth to LH
-0.056
NS
(Depth )to LH)2
-0.001
NS
h2o-p
7 Ml
**
(h2o-p)2
-0.351
NS
Height at age 3 years
Intercept
86.100
**
0.495**
Depth to LH
2.616
*
(Depth to LH)2
-0.018
NS
h2o-p
13-209
(10£ level)
h2o-p2
-0.144
NS
Height at age 5 years
Intercept
126.410
**
0.473**
Depth to LH
8.986
**
(Depth to LH)2
-0.069
**
h2o-p
2.845
NS
(h2o-p)2
0.969
NS
* Significant at the 5% level.
**Significant at the ]% level.

1*5
nary screening, that H2-extractable P Is a good index only of the short
term P supply to trees.
Results from multiple regressions (not shown) of height on depth
to LH, P extracted by HCl-h^SO^ (0-20 cm) and the squares of these terms
showed no significant contribution of HC1-H^SO^-extractab1e P towards ac
counting for variation in height at any age. However, since P extracted
by HC1-H2S0/1 was significantly related to depth to LH, these multiple
regression results cannot be taken to imply that HC1 -H2S0/1-extractable P,
which previous data have shown to provide a good index of the longer
term P supply to trees, does not play a causal role in determining tree
height over longer growth periods. It is likely that depth to LH
provides an integrated index of several factors influencing growth such
as available P within the effective rooting volume (see earlier discus
sion), and soil moisture and aeration conditions.
'Best fit' models derived using the procedure and soil and site
parameters outlined previously are given in Table 36. These models could
be used for prediction purposes, but the extra time and effort required
to measure several parameters may not justify the improved accuracy
obtained over using a single parameter model such as depth to LH. These
models accounted for a fairly high proportion of the variation in height
growth of slash pine. Unmeasured parameters such as genetic variation,
competing vegetation, insect and disease attack, amount and distribution
of rainfall, and temperature ranges and fluctuations would probably
account'for much of the remaining variation. ~ ~ "
Relationships Between Foliar-P Concentrations and Tree Parameters
Multiple correlation coefficients (R2) for the regressions of
height, relative height, and frtilizer-P requirements on 1og-transformed

Table 36. Regression coefficients for variables included in 'best fit'
multiple regression equations of height at age 1, 3, and 5
years on soil and site parameters and the squared terms of
these parameters
Source
Coefficients
S i gnifi canee
R2
Height at age 1 year
0.540**
Mean
49.444
.
h2o-p
3.706
AJ.
r\
(Drainage class)z
-0.626
JUJU
Silt + clay
-0.007
JUJU
Height at age 3 years
0.705**
Mean
62.944
h2o-p
8.525
JUJL
Drainage class
99-196
JU n
(Drainage class)z
-16.709
**
Silt + clay
-0.017
Height at age 5 years
0.610**
Mean
82.1 39
(h2o-p)2
0.958
A
Drainage class
105.065
AA
(Drainage class)2
-21.181
AA
Depth to LH
6.194
AA
(Depth to LH)2
-0.042
JUJU
Significant at the 5% level.
'Significant at the 1% level.

147
foliar P are given in Table 37- Actual P concentrations in the foliage
collected from the control plots of each replication of each trial are
presented in Appendix Table 59-
The values for the relationship between height and foliar P
increased with increasing age of trees. This can probably be attributed
to two factors. First, P concentrations in the foliage of 4-year-old
trees was shown earlier to be a function of the quantity factor of soil-P
supply, while height growth in the first year was related to the intensi
ty factor of soil-P supply, with the quantity factor becoming more
important over longer growth periods. Second, because of internal cycl
ing of-P within trees, one would anticipate a closer relationship between
foliar P and height of trees subsequent to the collection of foliage
samples. Foliar samples were collected when the trees were 4 years old.
The significant relationship between foliar P and height of 5-year-old
trees tends to support the earlier contention that soil P does contribute
to the variability in height of the older trees.
Foliar P was more closely related to relative height than soil P
extracted by any of the soil-test methods. This is in agreement with the
findings of other workers (Wells et al. 1973)* Since foliar P is a more
direct measurement of P available to trees, which integrates all factors
of soil-P supply including depth and time functions, this is perhaps not
surprising. Models for the relationships between relative height at age
1, 3, and 5 years and foliar P are shown in Fig. 12. Foliar~P values
corresponding to a relative height of 90% at ages 1, 3, and 5 years were
in the range 0.085-0.095%, which corresponds almost identically to the
critical range proposed as the best current estimate in the review of
literature. Although it is not the objective of this study to calibrate
foliar-P levels against growth and response of slash pine, the critical

148
Table 37. Relationships between height, relative height, and P-
fertilizer requirements (95%) at age 1, 3, and 5 years,
and P concentrations in the foliage of 4-year-old slash
p i ne
Age
Height
Rel. height
Fert. reqm.
d2
yr
1
0.015
0.506**
0.103**
3
0.194**
0.544**
0.125**
5
0.278**
0.485**
0.100**
** Significant at the
1% level, using
the model Y = b
logX + c, where
X = foliar P.
Table 38.
Relationships between extractablc-soi1
three depths, and within the effective
and P concentrations in the foliage of
pine
P at two positions,
soi1 depth (volume)
4-year-old slash
Soi1-test
Soi1 sample
method
0-20B cm
0-20 cm
20-40 cm
40-60 cm
Volume
r2
H2
0.159**
0.183**
0.154**
0.140**
0.357**
NH.OAc
4
0.273**
0.271**
0.281**
0.292**
0.479**
HC1-H SO,
2 4
0.588**
0.644**
0.494**
0.473**
0.708**
Bray 1(3)
0.602**
0.618**
0.299**
0.319**
0.627**
Bray 2
0.601**
0.617**
0.462**
0.462**
0.675**
** Significant at the 1% level, using the model Y = b logX + c, where
X = soi1-test value.

RELATIVE HEIGHT,
Fig. 12. Relationships between P concentrations in foliage(x) of ^-year-old slash
pine and relative height of slash pine 1, 3, and 5 years (Yl, Y3, and Y5)
after P fertilization.

150
values (range) obtained from the close relationships shown in Table 37
and Fig. 12 can be used to confirm critical soil-test values proposed
from relationships between relative height and soil-test values.
Relationships Between Foliar P and Soil-Test Values
Multiple correlation coefficients (R^) for relationships between
foliar P and soil-test values illustrated the same trends reported in
the preliminary screening (Table 38). Soil-test methods which extracted
relatively large amounts of P from the soil (HCl-h^SO^, Bray 1(3), Bray
2) were more closely related to foliar P of 4-year-old trees than soil-
test methods which extracted smaller quantities of P from the soil
(H2O, NH^OAc). The importance of soil P at rooting depths below the
surface 20 cm. was again illustrated by the larger R^ values obtained
using P extracted from the soil effective rooting depth.
The relationship between foliar P and HC1-HjSO^-extractab1e P in
the surface 20 cm (bed) of soil is illustrated in Fig. 13. The
HCl-H^SO^-P value corresponding to the foliar-P concentration of 0.085%
was 4.5 ppm, which is in good agreement with the value of 5 ppm proposed
earlier as useful for separating P responsive and nonresponsive sites.
From the equation relating foliar P (Y) to Bray 1(3)-extractable P (X),
Y = 0.0189 logX + 0.0684
the Bray 1(3)-extractab1e P value corresponding to 0.085% P in the
foliage was calculated to be 7-5 ppm, while the corresponding value
for Bray 2 was calculated to be 9-8 ppm from the following equation:
Y = 0.0231 logX + 0.0620.
General Discussion
Data from the calibration study indicated that for the purpose of
predicting P-responsive sites and the P-fertilizer requirements for the

FOLIAR
2 4 6 8O 12
HCI-H2SO4 EXTRACTABLE P, ppm
Relationship between HC1 -HS0;^-extractab 1 e P(X) in the surface 20 cm of soil and
P concentration in foliagefY) of 4-year-old slash pine.

152
first 3 to 5 years following slash pine establishment, the HCl-F^SO^ or
either of the Bray methods were superior to the NH^OAc method currently
used in Florida. Of these three methods, the HCl-F^SO^ method is prob
ably the most suitable for several reasons: (a) Its prediction value
was at least as good as either of the Bray methods. (b) The method can
also be used to determine the amounts of available cations in the soil
(Page et al., 1365). (c) The method is currently used on a routine
basis by several southern states (Page et al.,1965), and thus its use by
Florida would offer an opportunity for improved standardization and
calibration. (d) The method is suited for large-scale routine determi
nations in soil-testing laboratories since the analytical procedure is
rapid and presents no problem from ion interference in the colorimetric
determination of P.
From the results presented above, the following interpretation
of HC1-H^SO^ test results is suggested. (a) Soils containing < 5 ppm
of HC1-H^SO^-extractable P should,in the main, respond significantly to
applications of P fertilizer. (b) Soils testing between 2.5 and 5 ppm
P should receive applications of ca. 20-*t0 kg P/ha applied in a band
1.2 metres wide down the tree rows at or near the time of planting.
Soils testing below 2.5 ppm P should receive applications of ca. *)0-80
kg P/ha. If the fertilizer is broadcast, rather than banded in the tree
row, field experience has indicated application rates approximately 50%
greater than those shown above should be used. For soils testing below
2.5 ppm which have a high P-retention capacity, the highest application
rate of 80 kg P/ha should be used. Evidence has suggested however that
even this rate may be inadequate over a 5_year period from establishment
on some of the most highly P-retentive soils (Pritchett and Smith, 197*0

153
The above recommendations for the HCl-f^SO^ method should be used
only in conjunction with a knowledge of the limitations imposed by both
the conditions under which the calibration was conducted and the short
comings of the soil-test method. These can be summarized: (a) The rec
ommendations are valid only for slash pine plantations over a 3" to 5
year period following planting on unfertilized, acid, sandy soils in the
Coastal Plain. For establishing the fertilizer needs of older establish
ed stands, or the need to refertilize stands 5 or more years after
fertilization at establishment, foliar analysis should probably be used,
(b) The recommendations are based on soil-test values for the surface
20 cm of soil collected from within the bed. Four years after bedding,
when the samples for this study were collected, there was only a small
difference between extractable P in surface samples collected from the
bed and interbed area. However, shortly after bedding these differences
may be substantial due to enhanced mineralization rates. Thus, in order
to apply the above recommendations, soil samples should be collected
prior to bedding, or from the undisturbed area after bedding. For sites
where lower, root-penetrable horizons have substantially higher extract-
able P levels than the surface 20 cm, use of the surface sample may
underestimate the P status. Similarly, where the effective rooting
depth is relatively large, use of surface samples may underestimate the
P status of the site. For instance, from the relationship between
foliar P (Y) and HC1-H2SO/4-extractable P(X) on the 20 sites where depth
to LH was greater than 75 cm was
Y = 0.0029X 0.00007X2 + 0.0796 (R2 = 0.588)
and the extractable P value corresponding to a foliar P concentration of
O-085- was calculated to be 2 ppm, which was considerably less than the

15*
value of 5 ppm calculated from the relationship for all sites. (c) In
this study, HC1 -I^SO^-extractable P values were calibrated against P
response information obtained from sites on which possible N deficiencies
had been corrected by applications of N fertilizer. Thus it is possible
that on sites where both N and P are deficient, a response to P
fertilizer may not be obtained, even though the soil test indicates a
response should be obtained, until the N deficiency has first been cor
rected. This situation will exist for concurrent deficiencies of P and
other growth-limiting factors. (d) Data presented in a previous section
indicated that where insoluble basic calcium phosphates occur in soils
with a pH > 5, the HCl-H2S0/< method will overestimate the P status of
the soil. Thus high HC1 -H2S0j-extractable P values for near neutral
soils should be treated with caution. (e) The fertilizer recommenda
tions are based on the use of CSP. As discussed in the review of liter
ature, this source may be inferior to less soluble sources such as RP
on sites of either excessively low or high P-retention capacity. The
value of using soil-test methods to identify these sites of extremely
low and high P-retention capacities will be examined in a following sec
tion.

155
Phosphorus-Retention Study
Phosphorus-adsorption isotherms of representative soils from each
soil order are shown in Fig. \h. Isotherms similar to these were plotted
for all h2 soils used in this P retention study, from which data were
obtained for fitting the Langmuir adsorption isotherm. Two indices of
the P-sorption capacity of the soils, determined from the Langmuir equa
tion and saturation with 2,500 yg P/g soil, are presented in Appendix
Table 60. These data illustrate the wide range in P-sorption capacities
of soils in the Coastal Plain. A considerable overlap in sorption capa
cities of soils between various soil orders was obtained suggesting that
these broad classification units cannot be used to separate the soils
into distinct P-sorption groups, although soils classified as Spodosols
form a fairly distinctive group in the lower P-sorption range. Thirteen
of the 19 Spodosols in this study had a zero Langmuir adsorption maximum.
The Langmuir maximum, the derivation of which is based on the as
sumption of monolayer adsorption, gave consistently lower values than the
adsorption which can be obtained by saturation of the soil with a concen-
tracted phosphate solution. This i 1lustrates the theoretical weakness of
assuming monolayer adsorption as the sole mechanism operative in P reten
tion in soil systems. However, for the hi soils, the values for satura
tion maximum (V) and Langmuir maximum (X) were significantly related:
Y = 2.33X + 63.22, (r = 0.986).
This close relationship suggests that although the Langmuir data
do not provide absolute indices of P retention, they adequately reflect
relative P-retention capacities.
Relationships Between P Retention and Soil Properties

P adsorbed, /jg/g soil
Fig. 14. Phosphorus-adsorption isotherms of four Lower Coastal riain
soil types representative of four soil orders.

Physical and chemical properties
In many studies relating P retention and soil properties, linear
regression analysis is used to relate P retention, determined as the ad
sorption of an arbitrary amount of applied P, to soil properties (Yuan
and Breland, 1969; Syers et al., 1971; Udo and Uzu, 1972). Data in Fig.
15 illustrate the relationship between adsorption of P applied at three
different levels and extractable Al by NH^OAc (pH *1.8), a -property pre
viously reported as closely related to P retention (Yuan and Breland,
1969). The relationship was not linear until sufficient P had been added
to saturate retention sites. Increasing the application rate from 100 to
2,500 yg P/g soil improved the linear correlation coefficient from r =
0.8l to r = 0.93. For soils used in this study, an application rate of
2,500 yg P/g soil appeared to be more than adequate to ensure such linear
relationships.
Correlations between P retention and soil properties commonly used
to characterize soils are shown in Table 39- Also shown are the mean
values and range of these properties. Specific values of these properties
for each soil are presented in Appendix Table 60. Soil pH, clay, silt,
and loss on ignition were significantly correlated with P retention. All
of these soil properties have been previously recorded as being related
to P retention in other soils (Saunders, 1965; Syers et al., 1971; Udo
and Uzu, 1972). The significant positive correlation between pH and P
retention is of interest in that this relationship is usually reported to
be negative (Udo and Uzu, 1972). Although this would be anticipated from
the greater activity of Al and Fe at lower pH, the soils with the lowest
pH in the Coastal Plain are Spodosols, which are strongly leached sands
containing very little Al and Fe of any form in the Al horizon and,

P retention,
NH4OAC (pH 4.8) extractable Al, ppm
Fig. 15. Relationship between % P retention from three P solutions (100, 300 and 2,500 ug P/g soil) and
NH^OAc (pH 4.8)-extractable Al. For clarity, the points for only 15 soils are shown but the
linear correlation coefficients (r) were computed using data of all 42 soils.

159
Table 39- Soil properties and their correlation with P retention
Soi 1
property
Mean
Range
P retent ion
Langmuir Saturation
maximum maximum
CEC, meq/100g
5.02
1.48-11.65
0.214
_ r
0.190
Exch. Ca, meq/100g
0.83
0.13-3.44
0.222
0-251
Clay, %
3.95
0.1-20.4
0.771**
0.755** .
(0.323*)'
Silt, %
10.68
0.1-33.8
CO
O
o
0.727**
(0.257)
Silt + Clay, %
7-31
0.2-47.4
0.785**
0.757**
pH
4.34
3.1-6.2
0.479**
0.535**
(0.551**)
Loss on
ignition, %
3.33
1.42-6.71
0.455**
0.425**
(0.085)
Significant at the 5% level.
Significant at the 11 level.
Partial correlation coefficients in parentheses are corrected for
effect of NH^OAc-Al.

160
consequently, have virtually no P-retention capacity. The possible con
tribution of Ca to the above correlation in these soils, which have a mean
pH of b.3b, was probably minimal.
Relationships between P retention and such soil properties as clay,
silt, pH, and organic matter are frequently considered to be indirect
through the association of these properties with the causal agents of P
retention, such as the various forms of Fe and A1. Partial correlations
(recorded in parentheses in Table 39) > showed that correcting the signif
icant relationships for the indirect association through A1 (NH^OAc-
extractable) significantly reduced correlations in all cases, except for
pH. The relationship between P retention and soil pH was independent of
any effects soil pH may have had on NH^OAc-extractable A1, while that of
clay was in part related to its association with NH^OAc-extractable A1,
but it also had a significant independent effect. Clay, particularly 1:1
lattice clays which are prevalent in many Southeastern soils (Fiskell and
Perkins, 1970), may contribute directly to P fixation particularly St low
soil pH values (Black, 19^3)- The possibility that clay levels may be
related to active Fe, or active forms of A1 not extracted by NH^OAc may
also have contributed to the significance of the independent relationship.
The positive effect of pH on P retention within specific A] levels ap
pears to have r.o explanation in terms of current knowledge of the effects
of pH on P retention by sesquioxides and a 1umino-si 1icate clays. The
explanation probably lies in the association of low pH with podzolized
soils within which most of the sesquioxides have been elluviated from
surface horizons into the spodic horizon. At specific levels of NH^OAc-
A1 increasing pH is probably associated with the presence of higher
levels of active Fe and/or active forms of A1 not extracted by NH.OAc.

161
Extractable Al and Fe
Amounts of Al and Fe removed by a range of extractants and their
correlation with the two measurements of P retention are shown in Table
*40. Specific values of A1 and Fe removed by selected extractants are
presented in Appendix Table 6l. Despite some variation in the size of
the correlation coefficients, the relationships between P retention and
levels of A1 and Fe extracted by all reagents were significant at the 1%
level and were independent of the method of measuring P retention. The
results agree with the findings of many investigators (Yuan and Breland,
1969; Udo and Uzu, 1972; Syers et al., 1971) that extractable soil A1
provides the best single index of soil-P retention over a range of soils.
However, this does not imply that Al is more active than Fe in the ad
sorption of P or that extractable Fe may not be more closely correlated
with P retention within certain soil groups, as was found by Yuan and
Breland (1969).
Several investigators have reported that exchangeable Al (KC1 -
extractable) provided the best index of soil-P retention (Coleman et al.,
I960; Syers et al., 1971; Udo and Uzu, 1972) but in this study it proved
inferior to all other forms. This may be attributed to the small quanti
ties of exchangeable Al found in soils with a pH > 5-0 but which fre
quently had large amounts of amorphous Al (oxalate extractable). Of all
extractants of Al NH^OAc (pH *1.8), which on the average extracted only
slightly more Al than KC1, mainly from the less acid soils, provided the
best index of P retention. This reagent apparently removed Al from the
soil in proportion to the amount of active forms present, as indicated
by its correlations with Al extracted by other reagents ( r > 0.80 in all
cases except for KCl). Correlation coefficients between P retention and

Table 40. Extractable A1 and Fe values and their correlation with P retention
,
P
retent ion
Extractant
pH
Element
Mean
Range
Langmuir
maximum
Saturation
maximum
ppm r
1JN KC1
7.0
A1
47.7
1.5-200.0
0.745**
0.705**
1 N_ NHjjOAc
4.8
A1
68.7
8-222
0.937**
0.934**
Fe
13-9
0-155
0.659**
0.651**
A1
+ Fe
-
-
0.932**
0.906**
0.3M (NHlf)2C20if
3.2
A1
450.0
25-1250
0.881**
0.890**
Fe
288.5
55-1560
0.898**
0.903**
A1
+ Fe
-

0.94]**
0.949**
CDBt
8.2
A1
526.1
50-1680
0.887**
0.881**
Fe
837.2
39-4360
0.784**
O.709**
A1
+ Fe
-
-
0.856**
0.856**
0. 1M Na^P^
10.0
Al
Fe
1926.2
703.8
100-11100
40-5220
0.881**
0.830**
0.889**
0.835**
A1
+ Fe
-
-
0.879**
0.882**
0.05M EDTA
9.0
Al
398.9
25-1000
0.901**
0.894**
Fe
144.9
40-800
0.769**
0.798**
A1
+ Fe
-
-
0.942**
0.933**
0. 1 N_ HC1
1.1
Al
197.9
25-435
0.865**
0.849**
Fe
37.8
9-220
0.641**
0.593**
A1
+ Fe
-
-
0.904**
0.887**
** Significant at the ]% level.
1 Citrate-dithionate-bicarbonate (Mehra and Jackson, I960).

163
Al extracted by the other reagents were all very similar despite large
difference in the amounts of A1 extracted, suggesting that they are all
extracted essentially the same form of A1 but with different degrees of
efficiency. It appears reasonable to assume, in view of the relatively
small differences in mean A1 extracted by oxalate (amorphous) and CDB
(amorphous and crystal 1ine),that the dominant form of A1 in most of these
soils is amorphous, excluding that in phyl1 os i 1icate clays.
Correlations between P retention and Fe extracted by the various
reagents differed markedly. Unlike the situation for extractable A1 where
the correlation coefficients were essentially independent of amounts ex
tracted, correlations between P retention and extractable Fe tended to
increase with the amount of Fe extracted up to an optimum for the oxalate
extraction (amorphous) and then declined for the stronger CDB and pyro
phosphate extractions. There appears to be two possible explanations for
this phenomenon. One possibility is that extractants that were least ef
ficient in removing Fe did not dissolve all the active Fe fraction in
volved in P retention while the more efficient reagents extracted other
Fe forms in addition to those active in P retention, such as crystalline
oxides and interlayer Fe (Ramulu, Pratt, and Page, 1967). Secondly, the
amounts of Fe extracted by the oxalate reagent may be more closely cor
related with all soil components involved in P retention than are amounts
extracted by the other reagents. The presence of appreciable crystalline
Fe in these soils, indicated by the relatively large CDB/oxalate ratio
(McKeague, 1967), and the highly significant correlations between amor
phous Fe and all other forms of A] and Fe in the soil (Table 4l) suggest
both above explanations are quite possible.

Table 41. Correlations between different forms of A1 or Fe, and P retention
Variable
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(1)
Exchangeable A1
1.000
0.382
0.199
0.810**
0.748**
0.302
0.705**
(2)
Amorphous A1
1 .000
0.490**
0.238
0.686**
0.66!**
0.850**
(3)
Crystal line A1
1.000
0.065
0.454**
0.841**
0.498**
(4)
Exchangeable Fe
1.000
0.822**
0.314
0.651**
(5)
Amorphous Fe
1.000
0.696**
0.914**
(6)
Crystal line Fe
1.000
0.699**
(7)
P retention
1 .000
** Significant at the 1% level.

165
Relative contribution of A] and Fe to P retention
The existence of closer relationships between P retention and ex
tractable A1 than between P retention and extractable Fe cannot be taken
to indicate that on a weight basis A] is more active than Fe in the sorp
tion of P, as has been suggested (Syers et al., 1971). Exchangeable A1
(KC1-extractable) and Fe (NH^OAc-extractable), amorphous Al (oxalate minus
KC1-extractable) and Fe (oxalate minus NH^OAc-extractable), and crystal
line Al and Fe (CDB minus oxalate-extractable) were used as independent
variables in a stepwise multiple regression analysis (Barr and Goodnight,
1972) with the retention of 2,500 pg P/g soil as the dependent variable
(Y,%). The correlation matrix for these variables is shown in Table Al.
Of all forms, amorphous Fe was the most closely correlated with P reten
tion, followed by amorphous Al. However, it is noteworthy that amorphous
Fe was significantly correlated at the 1% level with all other Fe and Al
forms. Regression equations from the stepwise regression analysis, in
which all partial regression coefficients were significant at least at
the 5% level, are shown in Table 42.
Partial regression coefficients in equation [3] relating P reten
tion to amorphous Al and Fe suggest that on a unit-weight basis, amor
phous Fe was more active in P retention than was amorphous Al. However,
this may have been due to the close relationship of amorphous Fe to other
active components such as exchangeable Al and Fe not included in the
equation. Such a possibility is supported by the reduction in the
partial regression coefficient for amorphous Fe upon inclusion of ex
changeable Al in equation [4] and complete loss from the equation as a
significant component when exchangeable Fe and crystalline Al were in
cluded as shown in equation 5]. Because of the close interrelationships

Table 42. Regression equations relating P retention {%) to different forms of soil A1 and Fe
Prediction equations*
R2
Equation
number
V = 3.708 + 0.040
(am.
Fe) +
0.835
[2]
-<
II
O
+
O
o
o
(am.
Fe) +
0.015
/
(am. Al)
0.930
[3]
Y = 0.794 + 0.022
(am.
Fe) +
0.017
(am. Al) + 0.054 (ex. Al)
0.944
[4]
Y = 1.456 + 0.023
(am.
Al) +
0.012
(cry. Al) + 0.029 (ex. Al) + 0.190 (ex. Fe)
0.955
[5]
* All regression coefficients significant at least at the 5% level.
^ Symbols; am. = amorphous; ex. = exchangeable; cry. = crystalline, all expressed as ppm on soil basis.

167
between many of the Fe and A1 forms, noninclusion as a significant com
ponent cannot be taken to indicate nonparticipation in P retention. The
data concur with other findings (Colwell, 1959) that the order of activity
of A1 and Fe per unit weight in P retention is exchangeable > amorphous >
crystalline (equation [5]). Although various extractants of A1 usually
provide the best single index of P retention, these multiple regression
data suggest that the contribution of active forms of Fe to P retention
is probably considerable, particularly on a per unit-weight basis.
Aluminum and Fe extracted by soil-P test methods
It is apparent from the results presented above, that any of the
extraction procedures conventionally used to remove various forms of A1
and Fe from the soil could be used as an aid to identifying soils of dif
ferent P-retention capacity. However, some of these procedures, such as
the oxalate, pyrophosphate, and CDB, presented difficulties in the deter
mination of the A1 and Fe and which, if used, would mean an additional
operation in any soil-testing program. However, A1 extracted by the
NH^OAc procedure, which is also used to extract soil P, could be satis
factorily used to predict soil-P retention. But, evidence presented in
an earlier section showed that other soil tests such as the HCl-P^SO^ and
Bray methods were more suitable than NH^OAc for predicting the long-term
P status of forest soils. Thus, the utility of A1 and/or Fe extracted by
these methods as a means of predicting P retention was examined.
Aluminum and Fe extracted by the NH^OAc, HCl-H2SOi), Bray 1 and
Bray 2.methods were significantly correlated with soil-P retention (Table
A3). Amounts of A] and Fe extracted from the individual soils by the
four soil-test methods are given in Appendix Table 62. For all four

Table 43- Aluminum and Fe extracted by four soil-test methods and thier correlation with P
retention
Soi1-test
method
pH Element Mean
Range
P retention
Langmuir Saturation
maximum maximum
ppm r
NH.OAc
A. 8
A1
69
8-222
0.937**
0.934**
Fe
14
0-155
0.659**
0.651**
A1 + Fe
-
-
0.932**
0.906**
HCl-H2S0ii
1.3
A1
159
23-435
0.843**
0.820**
Fe
35
6-223
0.724**
0.713**
A1 + Fe
-
-
0.904**
0.881**
Bray 1
2.5
A1
487
30-1390
0.909**
0.912**
Fe
77
13-395
0.783**
0.786**
A1 + Fe
-
-
0.926**
0.928**
Bray 2
1.5
A1
545
60-1480
0.901**
0.907**
Fe
109
20-565
0.829**
0.823**
A! + Fe
-
-
0.91.9**
0.927**
** .Significant at the 1% level.

169
soil-test methods, extractable A1 was more closely correlated with P re
tention than extractable Fe, accounting for up to 80% of the variation in
P retention. Summed A1 and Fe values improved the correlation coefficient
obtained over that with A1 alone in three of the four tests. In the case
of extractable A1, it appears that extractants capable of complexing A1
(acetate and fluoride) provide better indices of P retention than tests
which rely solely on acidity to extract the A1, irrespective of quanti-
i .
ties extracted. However, for extractable Fe, it appears the greater the
quantities of Fe removed by the extractant the better the correlation
with P retention. In view of the highly significant relationships between
soil-P retention and amounts of A1 extracted by these same tests, they
apparently could be used to provide an index of the soil-P retention.
With the possible exception of the HC1-H2S0( method, the additional de
termination of Fe in the extract would hardly be justified by the small
degree of improvement in predictability.
Relationships Between P Retention and Foliar Nutrient Concentrations
Since forest fertilization is not limited to young plantations and
the need for fertilizer in established plantations is more usually based
on foliar analysis, the usefulness of foliar elemental concentrations to
predict soil-P retention was examined using 3b out of b2 sites from which
foliage samples were obtainable. Element concentrations in ^-year-old
slash pine foliage collected from control plots of replication 2 are pre
sented in Appendix Table 59.
Foliar Fe and A1 were the only elements significantly correlated
with soil-P retention (Table bb). On acid soils, this could have been
predicted since the activity of Fe and A1 in the soil should determine
both P retention (Hsu, 1965) and foliar levels of Fe and A1, provided

170
Table 44.
Foliar nutrient concentrations and their correlation with
soi1-P retention
Element
Mean
Range
Correlation
with P
retentionT
£
r
N
n
1.07
o
co
i
o
0.244
P
/
0.087
o.oA 0.54 0.115
0.172
K
0.44
0.24 0.60
0.010
Ca
0.18
0.12 0.25
0.045
Ca + Mg
0.107
--ppm
Mn
216
69 -
663
0.147
A1
436
231 -
694
0.425*
Fe
47
22 -
92
0.5!4**
A1 + Fe
0.478**
Al/P
0.570**
Fe/P
0.499**
A1 + Fe/P

0.593**
* Significant at the 5% level.
** Significant at the 1% level.
^ P retention determined as % adsorption of 2,500 pg P/g soil.

171
that no major interaction with other elements occurred during thier up
take and translocation to the foliage. Although the relationship be
tween soil-P retention and foliar A] was significant, it accounted for
only 19% of the variation in P retention and is, therefore, not particu
larly useful for prediction purposes. Phosphorus retention was more
closely correlated with foliar Fe than with foliar A!, but even this re
lationship accounted for only 26% of the variation in P retention and the
predictability was not improved by using summed A1 and Fe values.
A graphical comparison of foliar A1 with soil A1 values (NH^OAc-
extractable) showed that foliar levels seriously underestimated soil A1 on
the sites where foliar P was low (P < 0.08%). An attempt was made to
correct this anomaly by using the ratio of foliar A1 and/or Fe to P as
the index of P retention. Use of such ratios improved the predictability
for A1 and A1 + Fe, but not for Fe (Table bb). However, even the best
relationship still accounted for only 35% of the variation in the P re
tention. By excluding sites where foliar P was below 0.08% (7 out of 3*0
from the analysis, the correlation between foliar A1 and soil-P retention
improved considerably to r = 0.752. This same relationship for the seven
low-P status sites was not significant (r = 0.070). On these low P sta
tus sites foliar A1 and foliar P concentrations were significantly cor
related (r = 0.691) while on the 27 sites where tissue P was > 0.03% this
relationship was not significant (r = 0.331)* Apparently on sites of low
P status foliar A1 levels are determined more by soil P than soil A1
levels. This may explain why foliar A] failed to provide a good index of
P retention over all sites. An examination of the relationships between
P retention and foliar Fe on sites of low and adequate P status did not
produce any improvement on the correlation over that obtained with Fe
us i ng al 1 si tes.

172
Foliar analysis, using current sampling procedures, appears to
have little merit as a means of predicting soil-P retention. Although
foliar A1 accounted for a significant proportion of the variation in P
retention when sites of low P status were excluded, such a relationship
can be of little practical use, as it is these low P status sites that
one is concerned with in forest fertilization. It is questionable whether
a good predictive equation could be developed using tissues other than
those currently sampled, particularly if the poor relationship on sites
of low-P status reflects either direct or indirect effects of insufficient
P uptake on the intake and translocation of A1 within the plant. Humph
reys and Truman (1964) found that in three pine species, increasing a-
mounts of A1 in the substrate increased both A1 and P in roots and shoots
provided that the substrate contained adequate P, and Haas (1936) re
ported a similar phenomenon in citrus. These reports, in conjunction with
the findings reported above that increasing foliar A1 appeared to be as
sociated with increasing foliar P on sites of low P status, suggest that
A1 and P uptake and/or translocation are linked somehow at low levels of
either element.
Calibration of Extractable A1 Against Field-P Retention
Relationships between soil properties and P retention determined
by addition of P solutions to soils in the laboratory cannot be used di
rectly to predict retention of phosphate fertilizers applied in the field.
This is because the nature of the reaction between P and soil in the re
gion of the fertilizer particles is likely to be somewhat different from
that between soil and P solutions used in the laboratory. Laboratory
reactions, however, provide clues about properties most likely to cor
relate with field retention.

173
The relationship between retention of field-applied P and A! ex
tracted by NH^OAc is shown in Fig. 16. Each point is the mean of deter
minations from three pairs of plots taken from the 3 blocks in each trial.
In cases where A] values between blocks varied considerably, blocks were
treated as separate samples. Where A1 values between paired plots within
blocks varied considerably, the block was discarded from the analysis.
Total P and NH^OAc-extractable A1 values for soils of the PQ, Pj and P2
plots in the 10 trials selected for use in this calibration are given in
Appendix Table 63. A number of factors probably caused deviations from
the expected relationship between P retention and extractable soil A1.
These included differences in uptake of applied P between trials, initial
variability in total P between control and treated plots, and differences
in the leaching caused by variable rainfall and soil porosity between
trials. The relationship between P retention and extractable A1 (Fig.
l6) appeared to be almost independent of application rates used in these
trials. For diagnostic purposes, it was assumed that excessive leaching
losses of P occur at sites with P retention below 50%. These sites can
be identified with reasonable accuracy as those with kO ppm, or less, of
NH^OAc-extractable A] (Fig. 16).
The amounts of A1 extracted by the four soil tests were signifi
cantly related (Table **5) and using the regression equations relating
each to A] extracted by NH^OAc, values corresponding to ^0 ppm A1 ex
tracted by the NH^OAc method were as follows: 300 ppm for Bray 1, 400
ppm for Bray 2, and 120 ppm for HCl-H^SO^. On sites which test below
these critical values, the use of soluble phosphate fertilizers, such as
superphosphate, should be avoided with preference given to slowly soluble
sources such as ground rock phosphate. On soils with marginal retention

r e t ention
125
100}
a=56 kg/ha
o=224 ii
I
Fig. 16.
NH4OAC Al, ppm
Relationship between surface (0-20 cm) retention of P applied as CSP b years previously
and NH^OAc (pH 4.8)-extractable Al at 10 sites growing slash pine.
'-J

Table 45.
Regression equations of A] extracted by
HC1-H2S04 (Yl), Bray 1 (Y2), and Bray 2
(Y3) on A1 extracted by NH^OAc (X)
Regression equation
Correlation
coefficient
Yl = 1.5X + 59-3
r
0.803**
Y2 = 5.OX + 100.2
0.819**
Y3 = 5.5X + 180.3
0.718**
** Significant at the 1
% level .

176
capacity (20-40 ppm of NH^OAc-Al), partially acidulated rock phosphates
containing a small proportion of soluble phosphate might be the most suit
able source. This source provides initially available P at a level small
enough to be retained on the low number of retention sites plus a slowly
soluble source to maintain available P over an extended period of time.
Such sources may also be of value on sites with very high P-retention
capacity where long-term utilization of soluble sources is restricted by
fixation of P in difficultly available forms (Humphreys and Pritchett,
1971). Fertilization of soils with negligible P-retention capacity
(NH^OAc-Al < 20 ppm) should be with slowly soluble sources, since any
soluble P wi11 leach in these soils. Use of soluble sources on low-
retention sites should be restricted to situations where frequent small
applications can be economically used, such as in nursery operations.
Data in Fig. 16 showed that on some soils (Spodosols) fertilized
plots contained less total P than did control plots, a phenomenon also
shown by the results of Humphreys and Pritchett (1971) for certain
Spodosols. The degree of vertical and lateral P movement may determine
both the ability of the tree to ultimately utilize the P lost from the
upper portions of the profile and also the extent of any adverse environ
mental effects. Very little, if any, of the leached P is likely to be
retained in the A2 horizons of the Spodosols (Humphreys and Pritchett,
1971), but the spodic (Bh) horizons of these soils have very high P-
retention capacities. This is shown in Fig. 17 comparing P retention
from the addition of 2,500 pig P/g soil in the A and Bh horizons of several
Spodosols used in this study. The P-retention capacity of some spodic
horizons was greater than that for any surface horizons of soils shown in
Appendix Table 59. In the event that drainage in these soils occurs

P reten t ion jug/gsoil
2000
160G
1200
800
400
:
EZ3
Ed
Ed
a
n
Lmi
L_i

Cl

Cl

a
a
ci
n
a


ci
r1
.P--' A,
On a
f.s.
Cl p
Cl bIl
CTTl
n
Cl
d
Cl

-
ci
CJ
Cl
Cl
Ed
CJ
Cl
Cl
td
Ed
n
__
Cl
Cl
0
Cl
Cl
r;i
Cl
Cl
Cl

Cl
Cl
11
Cl
Ed
Cl
Cl
; j
Ed
O
Cl
Cl
1
Cl
Cl
Cl
td
-1
d
1
r ~1
St.Johns Myakka Pome! !o Leon Leon
"f.S. f.s. f.s. f.S. f.S-
Fig. 17. Phosphorus retention as determined from addition of 2,500 yg P/g
soil in A and Bh horizons of six Spodosols.

178
vertically through the spodic horizon, it is unlikely that any of the P
leached from the surface horizons will move below the spodic horizon.
However, even if all the leached P is retained in the spodic horizon, it
is not likely that trees will be able to utilize it because the fluctu
ating water table of such soils confines most of the rooting activities
to above the spodic horizon (White and Pritchett, 1970). A high water
table on these soils may sometimes lead to lateral drainag above the
spodic horizon. This could lead to P movement into stream waters on sites
where recent applications of soluble P sources cannot be retained in the
surface horizons. Obviously, it is in the interests of all concerned to
use less soluble sources of P, such