Some Effects of Plant Ash and Heating of Soil

Minerals on Soils Affected by Bushfires

Baiq Emielda Yusiharni BSc. in Soil Science, University of Mataram, Indonesia

MSc. University of Western Australia, Australia

This thesis is presented for the degree of

Doctor of Philosophy of The University of Western Australia

School of Earth and Environment

Faculty of Natural and Agricultural Sciences

2012

i

ABSTRACT

Bushfires are very common in Australia and severely modify ecosystems including

soils. The effects of heating on soil chemical, mineralogical and morphological

properties including the growth of plants have been investigated using laboratory,

glasshouse and field observations. This thesis focuses on three soil minerals that are

affected by heating: kaolinite, gibbsite and goethite. The thesis also considers the

nature of ash created by bushfires and its reaction with soil.

A laboratory study investigated dehydroxylation of pure kaolinite, gibbsite and

goethite and their rehydroxylation under wet conditions. Rehydroxylation of heated

gibbsite was extensive at 95oC with bayerite and gibbsite forming during 14 days, the

process was much slower at 55oC. Metakaolinite formed from kaolinite and hematite

formed from goethite by heating did not recrystallise but did aquire structural water

during incubation in water. The specific surface area of all three dehydroxylated

minerals was increased substantially by wet incubation. Dehydroxylated minerals

and probably their partly rehydroxylated forms may exist in soils heated by fire and

thus affect the chemical behaviour of these soils.

Soil heated a day earlier in a bushfire at Wundowie in the Darling Range, Western

Australia was taken from under and adjacent to burnt Eucalyptus and grass tree

(Xanthorrhoea pressii) logs. Conventional and synchrotron XRD patterns of heated

and unheated soil show that the main crystalline compounds of unheated soil are

quartz, kaolinite, gibbsite and goethite. In heated soil, kaolinite had dehydroxylated

to form metakaolinite, gibbsite altered into an amorphous phase, while goethite

transformed into hematite (hydrohematite). The bushfire added calcite in plant ash to

the soil which considerably increased the pH. Increases in soil EC simply reflect the

addition of soluble salts in plant ash. Heating increased amounts of extractable Al, Fe

and Si due to crystalline minerals becoming amorphous as they were dehydroxylated

by heating. Evidently dehydroxylated minerals and possibly their rehydroxylated

forms are present in naturally heated soils and may exert a significant influence on

the chemical behavior of the soil.

ii

A glasshouse experiment was carried out on the impact of heating a lateritic podzolic

soil on phosphate availability and plant growth. Forest soil was heated at 250, 350

and 500oC which are temperatures attained by topsoil during bushfires. As in the

laboratory and field experiment, heating soil caused kaolinite, gibbsite and goethite

to dehydroxylate and to partly alter into metakaolinite, amorphous alumina and

hematite respectively. Heating increased soil pH and EC although EC then relatively

decreased for 350oC and 500oC heating. Yield of ryegrass decreased with increasing

temperature of heating for unfertilized soil and for heated soils supplied with

phosphate (P) fertilizer. The P concentration in ryegrass for each of three harvests

ranged from 0.03% to 0.30% and decreased in the same sequence as for yield (i.e

unheated soil>250oC>350oC>500oC heated soil). Clearly heating of soil by bushfires

may reduce the availability to plants of native and added phosphate.

Another study determined the amounts and forms of plant nutrient elements in the

ash of several Australian native plant species and investigated the reactions of ash

with soil. Ash may contain much calcium, magnesium, potassium, sodium or silicon

with amounts varying depending on plant species and plant part. Many minor

elements are also present including elements with no biological function. All

elements are mostly present in crystalline compounds which were identified using

XRD and SEM. Minerals present in ash include calcite (CaCO3), fairchildite

(K2Ca(CO3)2),, nesquehonite (MgCO3.H2O), sylvite (KCl), lime (CaO), scolecite

(CaAl2Si3O10.3(H2O)), quartz (SiO2), portlandite (Ca(OH)2, periclase (MgO), and an

apatite, probably resembling hydroxyl-apatite (Ca5(PO4)3(OH)) and wilkeite (Ca5((P,

S,Si)O4)3(OH,CO3)). Ash has important liming and fertilizer values and its

effectiveness in these roles is a consequence of the properties of these minerals and

their reaction with soil as was demonstrated in an ash plus soil incubation

experiment.

This study concludes that bushfires may have considerable impacts on soil chemical

and mineralogical properties. The impacts are diverse depending on the nature of the

fire, including the fire temperature, soil mineralogy and vegetation composition.

iii

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude and appreciation to my supervisor

Winthrop Prof. Robert Gilkes for his excellent support, advice, encouragement, and

great attention throughout my study.

I would also like to acknowledge the Australian Government through the Department

of Innovation, Industry, Science and Research (DIISR) for funding my International

Postgraduate Research Scholarship (IPRS) and Scholarship for International

Research Fees (SIRF). I would like to acknowledge the Graduate Research School

and School of Earth and Environment, University of Western Australia for providing

me with a completion scholarship for the last few months of my study.

I thank Michael Smirk for his technical assistance, Australian Synchrotron for beam

time and staff at the Centre for Microscopy at the University of Western Australia for

assistance with electron-optical analysis. I thank the amazing women in the Soil

Science office, Margaret Pryor, Gail Ware and Karen Newnham for their assistances

during my study. Special thanks to Ksawery Kuligowsky, Nattaporn Prakongkep,

Rick Roberts and family, Cameron Duggin and Andrijana Eded for support,

assistance and friendship.

I would like to thank all mineralogy group members and postgraduate students at the

School of Earth and Environment, UWA for your friendship and sharing of

knowledge, experiences, thoughts, and laboratory equipment.

Finally, I would like to thank my family and friends; I could not have finished my

study without your constant prayer and support.

iv

DEDICATIONS

I dedicate this thesis to special people in my life. To my husband, Husnan Ziadi for

his constant love, encouragement, support and sometimes-technical assistance

throughout my study. To my daughters, Dhiyaul Aulia Huda and Fadila Almira

Huda, thank you for being so understanding, I could not do this without your

constant smiles and love. Thank you for always letting Mum concentrate on doing

the PhD study, for giving Mum the best support while multitasking as a Mum and a

student at the same time. I also dedicate this thesis to my late grandmother, Siti

Harah and my grandfather Saharudin. I wish to thank my parents, Suharni and Lalu

Yusuf for showering me with love and support. I also wish to thank my mother in

law, Aminah Hafs for her prayer and love. I dedicate this thesis to my late father in

law, Ruba’I, who believed in me since the first time we met, for encouraging me in

pursuing my dream. I finally made it.

v

LISTS OF CONTENTS

ABSTRACT i ACKNOWLEDGEMENTS iii DEDICATIONS iv LISTS OF CONTENTS v LIST OF TABLES viii LIST OF FIGURES x LIST OF APPENDICES xiv

Chapter 1. Introduction 1

1.1 General Introduction 1

1.2 Objectives of this Study 2

1.3 Structure of the thesis 2

Chapter 2. Literature review 4

2.1 Forest fires issues in the world 4

2.2 Forest fires issues in Australia 7

2.3 Impacts of forest fire on soil properties 10

2.4 Studies on ash 12

2.5 Heating effect on soil mineralogy 13

2.6 Studies on mineral reversion 15

Chapter 3 Rehydration of heated gibbsite, kaolinite and goethite: an

assessment of properties and environmental significance

18

3.1 Introduction 18

3.2 Materials and Methods 19

3.2.1 Pure minerals and heating procedures 19

3.2.2 Chemical and morphological analysis 19

3.3 Results and Discussion 21

3.3.1 X-ray diffraction and chemical data 21

3.3.2 Thermal analysis 26

3.3.3 Specific surface area and phosphate adsorption 30

3.3.4 Infrared analysis 31

3.3.5 Electron microscopy 39

3.4 Conclusions 41

vi

Chapter 4 Short term effects of heating a lateritic podzolic soil on the

availability to plants of native and added phosphate

42

4.1 Introduction 42

4.2 Materials and methods 43

4.2.1 Soil and glasshouse experiment 43

4.2.2 Soil and plant analysis techniques 46

4.3 Results and discussions 47

4.3.1 XRD and SEM 47

4.3.2 Heating impacts on chemical properties 51

4.3.3 Forms of Fe, Al and Si in the heated soils 53

4.3.4 Plant dry matter 56

4.3.5 Plant Analysis 57

4.4 Conclusions 60

Chapter 5 Changes in the mineralogy and chemistry of a lateritic soil

due to a bushfire at Wundowie, Darling Range, Western

Australia

61

5.1 Introduction 61

5.2 Materials and methods 62

5.2.1 Soil samples burnt in forest fires 62

5.2.2 Analytical techniques 65

5.3 Results and discussions 66

5.3.1 Carbon and nitrogen 66

5.3.2 Element concentrations, pH, EC, available P

and K, and extractable Fe, Al and Si in soil

samples

67

5.3.3 Mineralogical and morphological effects of

bushfire heating on soil minerals

72

5.4 Conclusions 78

Chapter 6 Minerals in the ash of Australian native plants 81

6.1 Introduction 81

vii

6.2 Materials and methods 83

6.2.1 Ash preparation 83

6.2.2 Characterization of the ash 83

6.2.3 Incubation of ash-soil mixture 84

6.3 Results and discussions 85

6.3.1 Characteristics of the ash 85

6.3.2 Total and water extractable elements, carbon

and nitrogen concentrations

87

6.3.3 Mineralogy and morphology of ash (XRD and

SEM)

92

6.3.4 Ash and soil incubation 95

6.4 Conclusions 96

Chapter 7 General Summary, Limitations and Future Work 104

7.1 Introduction 104

7.2 General Summary 105

7.2.1 Dehydroxylation and rehydroxylation of soil

minerals

105

7.2.2 Effects of heating soil on the availability of P

to plants

106

7.2.3 The properties of soil heated in bushfire 107

7.2.4 The composition of plant ash 109

7.2 Limitations to this research and suggested future work 111

Chapter 8 Publications from this thesis 113

8.1 Conference publications 113

8.2 Journal publications 114

8.3 Other publications 114

REFERENCES 115

APPENDICES 127

viii

LIST OF TABLES

Table Page 2.1 Example of a fire intensity and an associated severity rating for

eucalypt-dominated sclerophyll vegetation communities in south eastern Australia based on Cheney (1981), Jasper (1999) and (Shakesby et al, 2003). Fire intensity and severity are broadly related.

5

2.2 Specific surface area (SSA) and water content, expressed as TGA weight loss of metakaolinite, kaolinite and rehydrated kaolinite samples (Rocha et al, 1990).

16

3.1 Properties of mineral samples used in the study. 20 3.2 Infrared absorption spectra maxima for variously

dehydroxylated kaolinite samples and 600 oC heated kaolinite wet incubated at 55/95 oC for 400 days, references values and assignments.

36

3.3 Infrared absorption spectra maxima for variously dehydroxylated goethite samples and 350 oC heated goethite wet incubated at 55/95 oC for 400 days, references values and assignments.

37

3.4 Infrared absorption spectra maxima for variously dehydroxylated gibbsite samples and 350 oC heated gibbsite wet incubated at 55/95 oC for 400 days, references values and assignments.

38

4.1 Properties of original and heated Bakers Hill (BH) soil samples (n = 1).

46

4.2 p-values for significant difference with level of P added (0, 1.66, 3.33, 6.66, and 13.33 mg P/kg) and heating temperature.

51

5.1 The nomenclature for the samples of heated soil from under burnt eucalyptus (Eucalyptus marginata) and grass tree (Xanthorrhoea preissii) logs at the Wundowie bush fire site.

64

5.2 Mean values of element concentrations (mg/kg), pH and EC for the unheated and heated soil (-2 mm), gravel, ash and charcoal after the bushfire.

68

5.3 Mean values of extractable Al, Fe and Si for soil, gravel, ash and charcoal under and adjacent to burnt grass tree (GT) and eucalyptus (EU) logs. SP = sodium pyrophosphate extractant, Oxalate = sodium oxalate extractant and DCB = dithionite-citrate-bicarbonate extractant.

72

6.1 The nomenclature for plant ash and some properties of the ash. SSA= specific surface area (m2/g), EC= electrical conductivity (mS/cm), Bic P= bicarbonate P (mg/kg).

82

6.2 Total element concentrations in plant ash (mg/kg). Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

86

6.3 (a) Concentration of water-soluble elements in plant ash. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf

87

ix

and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

(b) Proportion of element in plant ash that is water-soluble. 87 6.4 Crystalline compounds in plant ash identified by SXRD. Silver

wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

94

x

LIST OF FIGURES

Figure Page 2.1 Scheme of fire development phases and fire spread regimes. 8 2.2 Fire detection map for 15th December 2000 in Australia,

overlaid on a pan-Australian vegetation cover map. Note that Eastern Australia experiences a wet climate in this season. Source: Satellite Remote Sensing Services Department of Land Administration (DOLA).

9

2.3 The times of peak fire danger over Australia (Australian Climate Extreme Fires, 2008).

10

2.4 A summary of the thermal reactions of kaolinite (Frost et al, 2003).

14

3.1 Conventional XRD patterns for original and dehydroxylated kaolinite (A), goethite (B) and gibbsite (C) heated at the indicated temperatures.

22

3.2 XRD patterns for heated kaolinite (A), goethite (B) and gibbsite (C), wet incubated at 55/95 oC, 0-400 days, Cu Kα radiation. I=illite, A=anatase, H=hematite, Bo=boehmite, Gi=gibbsite, and Ba=bayerite.

23

3.3 Synchrotron XRD patterns for unheated and 600oC heated kaolinite (A), 350oC heated goethite (B) and 350oC heated gibbsite (C), wet incubated at 95oC, for 200 and 400 days, Cu Kα radiation.

25

3.4 Thermal analysis results for original kaolinite (A), goethite (B) and gibbsite (C) showing dehydroxylation peaks.

27

3.5 Thermal analysis results for kaolinite (A) previously heated at 600oC, goethite (B) and gibbsite previously heated at 350oC (C): then wet incubated at 95oC for 400 days.

28

3.6 Weight loss measured using TGA (110-840oC) for dehydroxylated kaolinite (600oC), goethite (350oC) and gibbsite (350oC) incubated for 0, 14, 70, 200 and 400 days at 55 and 95oC.

29

3.7 Specific surface area (SSA) of variously dehydroxylated kaolinite, goethite and gibbsite samples incubated for 0, 14, 70, 200 and 400 days at 55 and 95oC.

32

3.8 Langmuir P sorption maxima for variously dehydroxylated kaolinite, goethite and gibbsite incubated for 0, 14, 70, 200 and 400 days at 55 and 95oC.

33

3.9 P sorption maximum versus specific surface area for variously dehydroxylated kaolinite (A), goethite (B) and gibbsite (C) incubated for 0, 14, 70, 200 and 400 days at 55 and 95oC.

34

3.10 Infrared spectra of original, dehydroxylated and rehydrated kaolinite, goethite and gibbsite incubated for 400 days at 55 and 95oC.

39

3.11 Electron micrographs of original, dehydroxylated and rehydroxylated kaolinite, goethite and gibbsite incubated for 400 days at 95 oC.

40

xi

4.1 Conventional XRD patterns (Cu Kα radiation), with inset for the clay fraction (a) and synchrotron XRD (SXRD) patterns (b) of original Bakers Hills (BH) soil and soil heated to 250oC, 350oC and 500oC. Q = quartz, K = kaolinite, Gi = gibbsite and Go = goethite. The broad background scattering for the SXRD patterns is due to the glass capillary containing the sample.

45

4.2 Scanning electron micrographs (SEM) and X-ray spectra of the indicated particles for original Bakers Hills (BH) soil and soil heated to 250oC, 350oC and 500oC.

49

4.3 Mean (n = 3) soil pH (a), EC (b) and plant available soil phosphorus (Bic P) (c) after the last harvest versus rate of P applied. Means having different letters are significantly different at P ≤ 0.05.

50

4.4 Mean (n = 3) data for the effect of heating on extractable soil Al, Fe and Si for BH, BH250, BH350 and BH500 after the last harvest for zero P applied. Ox = ammonium oxalate, SP = sodium pyrophosphate, and DCB = dithionite-citrate-bicarbonate.

54

4.5 Relationships between plant yield and rate of P applied for harvests 1, 2, and 3 for ryegrass grown on variously heated Bakers Hill soil (n = 3). Representative standard error values are shown in Fig. 4.5b.

55

4.6 Relationships between plant phosphorus concentration and rate of P applied for harvests 1, 2, and 3 for ryegrass grown on variously heated Bakers Hill soil (n = 3). Representative standard error values are shown in Fig. 6b.

58

4.7 Internal efficiency curves (yield versus plant P content) for three harvests of ryegrass grown on variously heated Bakers Hill soil (n = 3).

59

5.1 View of the study site one day after the Wundowie bushfire showing scorched eucalyptus and grass trees (a) and the position of a burnt eucalyptus log showing ash and charcoal (b).

63

5.2 Flow chart of analyses performed during the study. 64 5.3 Mean values of carbon and nitrogen concentrations (%) for 0-

1cm soil, gravel, ash and charcoal under and adjacent to burnt grass tree (GT) and eucalyptus (EU) logs.

69

5.4 Mean values of bicarbonate soluble P and K for soil (a) and gravel (b) under and adjacent to burnt grass tree (GT) and eucalyptus (EU) logs.

70

5.5 Conventional XRD patterns (a= fine soil fraction and b=gravel) and Synchrotron XRD (c=whole soil) of soil under and adjacent to a burnt grass tree (GT) log (Q=quartz, K=kaolinite, Gi=gibbsite, H=hematite and C=calcite).

75

5.6 Conventional XRD patterns (a=fine soil fraction and b=gravel) and Synchrotron XRD (c=whole soil) of soil under and adjacent to a burnt eucalyptus (EU) log (Q=quartz, K=kaolinite, Gi=gibbsite, H=hematite and C=calcite).

76

5.7 Conventional (a, b) and Synchrotron XRD (c) patterns of ash and charcoal for burnt grass tree (GT) and eucalyptus (EU)

77

xii

logs (Q=quartz, A= apatite, and C=calcite). 5.8 Scanning electron micrograph (SEM) and X-ray spectra of the

indicated particles for grass tree ash, where large rhombic calcite crystal (CaCO3) (a), microcrystalline apatite (Ca10(PO4)6(OH)2) with calcite (b) and mixed potassium and calcium salts (c and d) are present.

79

5.9 Scanning electron micrograph (SEM) and X-ray spectra of the indicated particles for eucalyptus ash where rhombic calcite (CaCO3) (a), microcrystalline mixed calcium and magnesium carbonates and sulphates (b), an alumino-silicate, possibly K-feldspar (KAlSi3O8) with calcite (c) and phosphorous enriched calcite (d) (possibly calcite with apatite) are present.

80

6.1 Principal component analysis of log total element concentration for native plant ash, variables (elements) (a) and plant material (b). Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

90

6.2 Principal component analysis of log water soluble element concentrations, variables (a) and cases (b) and log proportion that is water soluble, variables (c) and plant material (d) for native plant ash. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

91

6.3 Mean (n=3) values with standard error bars of carbon (C) and nitrogen (N) concentrations (%) for native plant materials (original and ash). Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

93

6.4 (a) Synchrotron XRD patterns for SW, SL, PM, WW and WL. (b) Enlargements of part of the SXRD patterns (F=fairchildite (K2Ca(CO3)2), N=nesquehonite (MgCO3.H2O), C=calcite (CaCO3), S=sylvite (KCl), A=apatite (Ca5(PO4)3(OH), L=lime (CaO), E=scolecite (CaAl2Si3O10.3(H2O), Q=quartz (SiO2), Po=portlandite (Ca(OH)2 and P=periclase (MgO).

97

6.5 (a) Synchrotron XRD patterns for RW, RL, GT, JL, and HH. (b) Enlargements of parts of the SXRD patterns (F=fairchildite (K2Ca(CO3)2), N=nesquehonite (MgCO3.H2O), C=calcite (CaCO3), S=sylvite (KCl), A=apatite (Ca5(PO4)3(OH)), L=lime (CaO), E=scolecite (CaAl2Si3O10.3(H2O)), Q=quartz (SiO2), Po=portlandite (Ca(OH)2 and P=periclase (MgO).

98

6.6 Scanning electron micrograph (SEM) and X-ray spectra of the indicated ash particles for SW, SL, PM, WW and WL ash. Calcite crystals (CaCO3), microcrystalline hydroxyl apatite and mixed potassium, magnesium and calcium salts are present.

99

xiii

6.7 Scanning electron micrograph (SEM) and X-ray spectra of indicated ash particles for RW, RL, GT, JL, and HH ash. Calcrystals (CaCO3), microcrystalline hydroxyl apatite and mixpotassium, magnesium and calcium salts are present.

100

6.8 Principal component analyses of log SEM EDS elemental analyses results for native plant ash particles. Variables (a) and plant material (b). Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

101

6.9 Mean (n=3) values with standard error bars of pH (a), EC (b) and bicarbonate P (c) for unburnt soil incubated with native plant ash. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

102

6.10 Mean (n=3) values with standard error bars of pH (a), EC (b) and bicarbonate P (c) for burnt soil incubated with native plant ash. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

103

7.1 Flow chart for the dehydration and rehydration of kaolinite, gibbsite and goethite

106

7.2 Flow chart for the glasshouse study conducted on heated lateritic soil.

107

7.3 Flow chart for the field study of soil minerals heated in a bushfire at Wundowie, Darling Range, Western Australia.

109

7.4 Flow chart for the Australian native plants ash study. 111

xiv

LIST OF APPENDICES

Appendix Page 1 Thermal analysis results for kaolinite previously heated at 500oC:

wet incubated at 55 and 95oC for 14, 70 and 200 days 127

2 Thermal analysis results for kaolinite previously heated at 550oC: wet incubated at 55 and 95oC for 14, 70 and 200 days

128

3 Thermal analysis results for goethite previously heated at 250oC: wet incubated at 55 and 95oC for 14, 70 and 200 days

129

4 Thermal analysis results for goethite previously heated at 300oC: wet incubated at 55 and 95oC for 14, 70 and 200 days

130

5 Thermal analysis results for gibbsite previously heated at 250oC: wet incubated at 55 and 95oC for 14, 70 and 200 days

131

6 Thermal analysis results for gibbsite previously heated at 300oC: wet incubated at 55 and 95oC for 14, 70 and 200 days

132

7 Analyses of dry plant tops (leaves and shoots) of annual ryegrass (Lolium rigidum Gaud) for Harvest 1.

133

8 Analyses of dry plant tops (leaves and shoots) of annual ryegrass (Lolium rigidum Gaud) for Harvest 2.

134

9 Analyses of dry plant tops (leaves and shoots) of annual ryegrass (Lolium rigidum Gaud) for Harvest 3.

135

10 Conventional XRD patterns for (a) silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL); (b) red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH), (N=nesquehonite (MgCO3.H2O), C=calcite (CaCO3), S=sylvite (KCl), A=apatite (Ca5(PO4)3(OH)), E=scolecite (CaAl2Si3O10.3(H2O)), Q=quartz (SiO2), Po=portlandite (Ca(OH)2) and P=periclase (MgO).

136

1

Chapter 1

1.0 Introduction

1.1. General Introduction

Forest fires affect ecosystems worldwide and contribute to forest loss. Millions of

hectares of forest burn annually, which consumes several billion tones of dry

biomass (FAO, 2001). About 60 million hectares of forest worldwide were burned

per year during the period 2003-2007 (FRA, 2010).

Fire is often used as a tool to manage natural ecosystems (Neary et al., 1999). Fire is

the oldest method used to clear land for farming and other uses, and it is still widely

used in many countries. However, fire is a major disturbance factor that has

beneficial and detrimental effects on the forest ecosystem (FRA, 2010). Some

ecosystems are adapted to fire, however fires often get out of control and destroy

forest vegetation and biomass, providing an added threat to biodiversity and may

have a considerable impact on land degradation (FRA, 2010). Moreover, forest fire

has also has caused other types of catastrophic impact including the loss of human

lives and assets. For example, bushfires in Victoria, Australia in 2009 caused 173

people to die (Teague et al., 2009), while wildfires in Greece in 2007 have claimed

the lives of 80 people (FRA, 2010).

Heating during fire, which burns litter and both fallen and standing timber, may

impose significant impacts on soil physical, chemical and biological properties

(Raison, 1979; Neary et al., 1999; Certini, 2005). Earlier studies for several biomass

types were conducted under both laboratory and field conditions and at different

temperatures (Etiegni and Campbell, 1991; Misra et al., 1993; Bodi et al., 2011).

However, relatively few in depth studies have addressed the impacts of fire on ash

properties, nutrient solubility and the mineralogical and morphological properties of

soil. Ash characterization is required to determine the amounts and forms of

nutrients available for plants and for ecosystem re-establishment (Pereira et al.,

2011).

Increases in soil fertility and elevated plant nutrient concentrations in regrowth

vegetation occurr after a single fire (Christensen, 1994). Both surface and lower soil

2

horizons may be affected in high intensity fires. The effects of burning on soil

characteristics varies with the duration and intensity of fire, fuel and soil types,

including the amount of soil organic matter present at the time of fire (Raison et al.,

1985). The effects of fire on plant nutrient cycles have been widely studied (Raison,

1979; Certini, 2005). In contrast, the effect of heating on soil mineralogy and the

growth of plants after heating of soil have not been extensively studied although

modest (200-600oC) temperatures cause mineral transformations and may influence

the availability of several nutrients (Ketterings et al., 2000).

Kaolinite alters into metakaolinite at temperatures between 500oC and 700oC

(Richardson, 1972; Ulery and Graham, 1993) losing lattice water. Gibbsite

commonly alters to an amorphous phase on heating at 200oC (Rooksby, 1972), and

goethite is transformed to hematite at ≈300 oC (Cornell and Schwertmann, 1996).

The majority of research on the stability of minerals under heating conditions has

been conducted in the laboratory with both synthetic and purified pure minerals and

not on soils. The persistence of the dehydroxylated compounds in heated soils in the

field, which may affect the availability of nutrients for plants, is unknown and

deserves further investigation.

1.2 Objectives of this study

The specific aim of the study was to investigate the potential impacts of forest fire on

a highly weathered Western Australian soil containing heat sensitive hydroxylated

minerals. The study investigated soil and ash properties, including chemistry,

mineralogy and morphology that may affect the growth of plants after fire.

Experiments on how heating impacts on soil minerals including subsequent mineral

reversion by rehydroxylation were also conducted. The research involved studies on

ideal mineral samples as well as soil samples. The research was carried out under

laboratory, glasshouse and field conditions.

1.3. Structure of the thesis

This thesis consists of 7 chapters in which each chapter specifies and discusses

different aspects of the research and also discusses the related literature. Justification

and objectives of the research are introduced in Chapter 1. A general literature

review is given in Chapter 2. Chapter 3, 4, 5 and 6 have been written in journal

3

format and the manuscripts are currently under review, in press or published in high

impact journals.

Rehydration of heated gibbsite, kaolinite and goethite: an assessment of properties

and environmental significance is Chapter 3. Short term effects of heating a lateritic

podzolic soil on the availability to plants of native and added phosphate is in Chapter

4. Changes in the mineralogy and chemistry of a lateritic soil due to a bushfire at

Wundowie, Darling Range, Western Australia is Chapter 5. Minerals in the ash of

Australian native plants is Chapter 6. General discussion and conclusions, limitations

of this work and suggestions for further work are presented in Chapter 7. Tables and

figures are placed within the text and all the references cited are listed at the end of

the thesis followed by the appendices.

4

Chapter 2

2.0 Literature Review

2.1 Forest fires issues in the world

Fire is one of major disturbances on forest management where millions of hectares

of forest are affected by fire each year (FAO, 2001). An average of 60 million

hectares of forest were burned per year during the period of 2003-2007 worldwide.

This data only reflected 63% of global forest area as the information on forest fires

continues to be poorly reported. During the same period, an average of 156 000

forest fires occurred per year. The largest number of fires were in United States of

America, the Russian Federation, India, Poland and China where in total about

10000 fires occurred per year (FRA, 2010).

The major causes of fire are lightning, volcanoes and human action. Human activity

is now believed to be the main reason of forest fire (Shakesby et al., 2003; Turekian

et al., 1998). Forest fire affects the ecosystem and biodiversity patterns along with

the landscape (Myers et al., 2004). It affects the existence of individual plants and

animal species within the area of fire. Some other impacts of forest fires are

economic losses, destruction of the biological environment of the forest and hence

man’s natural environment (Karlikowski, 1982). Fires affect the ecosystem in several

ways (FRA, 2000): regulating plant succession, regulating fuel accumulation,

controlling age, structure and species composition of vegetation, affecting insect and

disease populations, influencing nutrient cycles and energy flows, regulating biotic

productivity, diversity and stability, and determining habitats for wildlife. Moreover,

fires have reduced the amount of some soil animals in the topsoil (Malmstrom,

2008). The top layer soil animals are killed immediately during fires, while animals

that live in deeper layers may survive or be killed (Wikars and Schimmel, 2001).

5

Table 2.1. Example of a fire intensity and an associated severity rating for eucalypt-

dominated sclerophyll vegetation communities in south eastern Australia based on

Cheney (1981), Jasper (1999) and (Shakesby et al, 2003). Fire intensity and severity

are broadly related.

Fire intensity (a)

(kW/m)

Max flame height

(m)

Severity

rating

Post-fire vegetation

characteristics

≤ 500 1.5 Low Only ground fuel and

shrubs <2m high burnt

501-3000 5.0 Moderate All ground fuel and shrub

vegetation <4m high

consumed

3001-7000 10.0 High All ground and shrub

vegetation consumed and

lower tree canopy <10m

high scorched

7001-70,000 10-30 Very high All green vegetation

including tree canopy up to

30m and woody vegetation

<5mm diameter consumed

70,001-100,000+ 20-40 Extreme All green and woody

vegetation <10mm

diameter consumed a The fire intensity index as defined by Byram (1959).

While fire has been the primary agent of forest degradation, some forest ecosystems

depend on fire for their regeneration and to retain their vigour and reproductive

capacity (FRA, 2010). Several tree species may take advantage of fire and periodic

controlled burns can contribute to overall forest health. Fire normally moves through

forests burning lower branches and clearing dead wood from the forest floor, which

improves restoration by providing ideal growing conditions for many plant species.

It may also provide a better forest floor habitat for some species that favor relatively

open spaces. Ketterings et al. (1999) stated in their recent survey that fire is

6

commonly used for forest clearing to provide an easy and economical means of

increasing access. Fire is also commonly used to burn logging slash and facilitate

seedbed formation for agricultural practises (Rab, 1996).

Shakesby et al. (2006) mentioned that the impacts of forest fire are influenced by the

frequency and severity of the fire itself. Frequencies of fire are determined by the

types of vegetation and climate. On the other hand, fire severity relies on the

interactions of burning intensity and duration, as well as the characteristics of fuels

(biomass), soil and local climate. Myers et al. (2004) stated that a landscape fire

regime is established by several factors, including:

- The human, physical, and biological properties of the landscape; these along

with weather and fuel characteristics, which influence the chance of ignition

and the speed and extent of spread

- The sequence of individual fires, including the characteristics and timing of

each fire

- The time elapsed between fires, which influences the recovery of the

landscape and its species composition

- The spectrum of potential different fire regimes as determined by the number

and size of fires and also the weather.

There are two main type of forest fires: controlled (prescribed) and wildfires

(Certini, 2005). Prescribed fires are controlled application of fire under specified

environmental conditions to reduce fuel levels and to avoid the severity of wildfires.

This type of fire is normally at low intensity and applied when soil is moderately

moist. Wildfires in general, are uncontrolled fire with the presence of massive fuel

loads and have high severity.

There are several different schemes use to classify the fire regimes. According to

Shakesby et al. (2006), forest fire is classified into three different types: ground fires

that influence the organic layer such as leaves and other parts of the plants, surface

fires which affect plants and bushes and burn the bases and crowns of trees, and

canopy fires that flame the higher leaves and branches. Shakesby et al. (2006)

summarised the fire regimes based on the intensity and severity (Table 2.1). The

7

scheme of fire development is described by Viegas (1998) in Figure 2.1. Brown and

Smith (2000) on the other hand, classify fire regimes as follows:

- Understorey fire (occur on forests and woodlands). Fires are generally non-

lethal to the dominant vegetation and do not substantially change the

structure of the dominant vegetation.

- Stand replacement fire (occur on forests, woodlands, shrublands, and

grasslands). Fires destroy the dominant vegetation and alters the structure of

aboveground vegetation. Approximately 80% or more of the top layer

dominant vegetation is either consumed or destroyed as a result of fires.

- Mixed severity fire (occur on forests and woodlands). Severity of fire either

causes selective mortality in dominant vegetation, depending on different tree

species susceptibility to fire, or varies between understorey and stand

replacement.

- Non-fire regime. There is a little possibility that it will experience natural

fire.

Clearly with increasing severity of fire, there is a greater probability of the soil being

heated and altered.

2.2 Forest fires issues in Australia

Fire has been part of the Australian natural ecosystem for millions of years. Fire has

been used as the most powerful land-use tool by the Indigenous Australians to

manage grasslands, forest and fauna (Gill and Moore, 1990). Australia is one of the

most fire prone continents with a huge variety of vegetation and fire regimes (FRA,

2000). Eucalyptus, acacias and grasses are well-adapted to fire regimes. Many

Australian plants and animals have evolved to survive fire events and most

Australian ecosystems have developed very strong relationships with fire. However,

fire also has caused fatalities in Australia where it has claimed over 800 lives since

1851 (Haynes et al., 2008). The recent devastating “Black Saturday” bushfires in

Victoria in 2009 caused 173 people to lose their lives

Based on Australian fire reports, about 115,000-230,000 fires were observed per year

by remote sensing during fire periods between 1998-2000. A map of fires detection

for one day in late 2000 is presented in Figure 2.2. Earlier reports stated that wildfire

8

has caused about 1million hectares of the forestland to be burned during the 1956-

1971 period.

Fire Growth Fire Decay

Secondary growth Decay

Initial growth Flame

extinction

Flaming

Glowing Glow

extinction

Ignition Extinction

Pyrolysis

Pre-heating Cooling

Figure 2.1 Scheme of fire development phases and fire spread regimes (Viegas,

1998).

Fire regimes in Australia are driven by climate and land use (Myers et al., 2004).

Figure 2.3. shows distribution of fire danger over seasons in Australia. Myers et al.

(2004) summarised that there are three major fire regions:

- The wet-dry tropical savanna region. Landscape-scale fires occur annually.

Fuels, such as grasses and herbs, accumulate during the wet season. During

the dry season, the fuels cure and there are spells of moderate to extreme fire

weather. Fires can also be ignited by people and by lightning. Fires tend to

occur in the dry season months of May to December.

Crown Fire

Surface Fire Surface Fire

Ground Fire Ground Fire

Initial state Final state

9

- The semi arid and arid interior. Landscape-scale fires occur episodically,

typically at intervals of up to a decade. Extensive fires only happen after

periods of exceptional growth when fuel is increased by the growth of

annuals between Spinifex hummocks, due to above average rainfall. The hot,

dry climate promotes curing of fuels every year. Fires usually occur in

spring-summer, from September to January.

- The southern temperate zone. Landscape-scale fires occur episodically, and

at intervals of decades. Major fires occur in those rare years when there is

both drought and severe fire weather. These fires are usually associated with

forests where fuels are woody rather than grassy. Fires typically occur from

October to March. (The present research relates to vegetation and fire in this

zone).

Figure 2.2. Fire detection map for 15th December 2000 in Australia, overlaid

on a pan-Australian vegetation cover map. Note that Eastern Australia

experiences a wet climate in this season. Source: Satellite Remote Sensing

Services Department of Land Administration (DOLA).

10

Figure 2.3. The times of peak fire danger in Australia (Australian Climate Extreme

Fires, 2008).

2.3 Impacts of forest fire on soil properties

Soil heating may eliminate the beneficial effects of surface organic layer on soil

properties. The effects of burning on soil characteristic may vary within the duration

and intensity of fire, fuel and soil types, and also the amount of soil organic matter

present at the time of fire (Flinn et al., 1984). Certini (2005) in his review on the

effects of forest fire on soil properties points out that fire alteration of soil properties

depends on several factors. The major factor is severity that is controlled by the

amount, nature and moisture of live and dead fuel, air temperature and humidity,

wind speed and also the topography of the area. Fire severity reflects both the

intensity and duration of fire (Certini, 2005; Keely, 2009).

Substantial amounts of organic matter are lost during fire at temperatures above

300oC thus affecting soil properties (Terefe et al., 2008). Fires increase soil

temperature, which can considerably disturb ecosystem dynamics by changing

nutrient quantity and cycling by affecting soil chemical, biological and physical

properties. The nutrients particularly N, S and P may experience oxidation to

gaseous form, organic matter volatilisation, ash particles convection, and water

transport both by leaching and residue transport (Binkley et al., 1992).

11

Heat transferred to soil depends on surface temperature and exposure duration

(Steward et al., 1990). Overall, soil heating during a fire occurs in topsoil to depths

of 10-15 cm (Ketterings et al., 2000). Ghuman and Lal (1989) conducted research on

soil temperature during a fire. They measured the soil temperature at 1cm depth in a

tropical rainforest during burning at 218oC. The temperature reached 150, 104 and

70oC at 5, 10 and 20cm depths respectively.

The effects of fire on plant nutrient cycles have been widely studied in native and

managed forests as well as for slash and burn agriculture in tropical forest

(Kauffman et al., 1993; Ketterings et al., 1999). However, the impacts of burning of

vegetation for several different temperatures have not been widely investigated.

Some studies reported increase in soil fertility and plant nutrient concentrations

shortly after a single fire (Christensen, 1994). Soils heated during a fire may

experience significant impacts on several soil properties (Raison, 1979). The effects

occur in both surface and lower horizons for both low and high fire intensities.

Commonly, burning raises the pH of surface soil (approximately 0-3cm),

exchangeable Ca2+, extractable P and other nutrients. These increases take place

quickly and the elevated concentrations may continue for up to 1 year or more

(Tomkins et al., 1991). Nitrogen and phosphorus are the nutrient elements that are

most affected by fire (Ferran et al., 2005). The fire interval needed to enable

recovery for these elements has been calculated to be in the order of 10-12 years for

N and 20 for P (Raison et al., 1985). Fire may be the dominant factor affecting C and

N losses from some forests (Cadwell et al., 2002).

Fire may cause P loss by volatilisation when temperatures exceed 360oC and during

burning P is also removed as ash in smoke (Cotton and Wilkinson, 1988). Raison et

al. (1985) found that P exported to the atmosphere may constitute 50% of the total P

in the combusted fuels. In contrast, many researchers believe that there is a large

increase of P in the soil surface shortly after fire as much plant P is retained on site

in ash (Wilbur and Christensen, 1985). Ferran et al (1991) reported an increase of

total P in the topsoil (0-5cm) after a wildfire.

12

Bauhus et al. (1993) analysed the effect of fire on carbon and nitrogen mineralisation

and nitrification in an Australian forest soil. In an incubation experiment they

determined the occurrence of nitrification in ash-beds and unburnt soils in native

eucalypt forests. In contrast to such detailed information on the forms and fate of

nitrogen in burnt soil, there are few observations on the chemistry, mineralogy and

morphology of other nutrient elements in soil.

2.4. Studies on ash

Burning creates ash-bed as fire removes some or all of the vegetation and litter

cover. The presence of ash may influence hydrological behaviour and soil erosion

processes (Woods and Balfour, 2008).

Khanna and Raison (1986) have conducted experiment on the function of ash as a

nutrient source and the process involved in the alteration of forms of elements

present near the soil surface. They found that the availability of plant nutrients

increased during the first year after fire.

Burning is used to remove plant residues in forestry and agricultural areas and also

for natural vegetation management (Khanna and Raison, 1986). Burning plant

materials may have beneficial and harmful effects on soil properties and the growth

of plants (Raison, 1979). Beneficial effects may include the increased availability of

plant nutrients, while harmful effects on the other hand could be associated with

organic matter losses and nutrient transport in smoke during the burning (Khanna

and Raison, 1994).

Nutrient losses due to fire depend on fuel utilization, fire behaviour, microclimate,

plant composition and structure, fire severity, fuel moisture content and fuel

compactness (Raison, 1979; Kauffman et al., 1993). Vegetation burning has resulted

in nutrient losses to the atmosphere (Kauffman et al., 1993). Both the black carbon

remaining after fire and nutrients are deposited on the soil surface in the form of ash,

which may be lost by wind or water erosion or leached through the soil (Raison et

al., 1985; Kauffman et al., 1993).

13

Several researchers have investigated the variability of ash layer thickness (Cerdá

and Doerr, 2008; Woods and Balfour, 2008), which normally ranges from less than 1

to 10 cm (Goforth et al., 2005). Gabet and Sternberg (2008) identified up to 20cm

thick ash layers in heavy fuel combustion areas. The amount of ash and its chemical

composition varied with the combustion temperatures (Etiegni and Campbell, 1991).

The major elements present in ash are calcium, potassium, magnesium, silicon,

manganese, aluminium, iron, phosphorus, sodium and zinc (Etiegni and Campbell,

1991; Misra et al., 1993; Liodakis et al., 2005). Misra et al., (1993) studied the

chemical and mineralogical composition of various types of wood ash and found that

for 600oC combustion, ash mainly consists of calcite (CaCO3) and fairchildite

(K2Ca(CO3)2) and for 1300oC lime (CaO) and periclase (MgO) dominate. Liodakis et

al., (2005) observed that the main compounds present in the ash of several forest

species were oxides, carbonates and sulfates of calcium, magnesium and potassium.

Combustion above 600oC resulted in decomposition of dolomite ((CaMg)CO3),

fairchildite (K2Ca(CO3)2), sylvite (KCl), arcanite (K2SO4) and potash (K2CO3). Lime

(CaO) and periclase (MgO) were present in wood ash combusted at 600oC.

Published information on ash properties is mainly for wood ash; little information is

available for the ash of leaves, bark and other parts of plants. However, Ulery and

Graham (1993) stated that ash composition depends on several factors, including

plant species, part of plant (wood, leaves, bark), plant age, soil type, climate and

condition of combustion.

2.5 Heating Effect on Soil Mineralogy

We will consider some soil minerals with structural (OH) that are affected by heating

(dehydroxylation) at quite low temperatures. Kaolinite is a dioctahedral 1:1 layer

silicate with the crystal chemical formula Si2Al2O5(OH)4. The structure contains

hydroxyl groups (Rocha, 1999). Goethite (α-FeOOH) is widespread soil mineral.

Gibbsite is one of the forms of aluminum hydroxide (Al(OH3) that occur in soils.

Little is known on the impact of soil heating on mineralogical properties of the soil

although modest (250-500oC) fires and extreme fires (>500oC) cause various mineral

transformations. Ketterings et al. (2000) investigated the effect of heat intensity on

14

the mineralogy of Oxisols in the Sepunggur area, Jambi Province, Sumatra,

Indonesia where slash and burn agriculture is commonly used. They evaluated the

effect of fire on mineralogy and soil texture in field and laboratory experiments.

They found that changes in soil properties with burning mostly affect the top layer

(0-5cm). The soil texture became coarser after burning, and heating soil reduced

gibbsite and kaolinite concentrations converting goethite into ultra fine maghemite.

The majority of the research on the stability of minerals under heating condition has

been conducted in the laboratory with both synthetic and purified natural minerals.

However, Ulery et al. (1996) observed effects of heating in the field environment.

Fire caused the collapse of some 2:1 phllyosilicates and destroyed kaolinite. The

occurrence of maghemite in soils has often been related to the dehydroxylation of

goethite or lepidocrocite by heating in fires in the presence of organic matter (Anand

and Gilkes, 1987).

Heating kaolinite alters it to metakaolinite, the Si-O arrangement remains intact and

the Al-O network reorganises itself. Frost et al. (2003) summarised thermal reactions

of kaolinite as shown in Fig. 2.4.

predehydroxylation state

Thermal Reactions of Kaolinite

Al2(OH)4Si2O5

450-550oC

Metakaolinite Spinel

Mullite

950-980oC

1000-1100oC

Figure 2.4. A summary of the thermal reactions of kaolinite (Frost et al, 2003).

Many studies have been conducted on the transformation of goethite into hematite

(Gonzalez et al., 2000) by heating and mechanochemical (dry) grinding. Those

15

studies were related based on the wide interest of the technological applications of

iron oxides. The transformation reactions can be either: (a) a direct change from

goethite into hematite, 2 α-FeOOH α-Fe2O3+ H2O, or (b) a transformation with

the formation of an intermediate superstructure phase before final formation into

hematite; α-FeOOH superstructure (FeOOH) α-Fe2O3 (Watari et al., 1979).

Fan et al. (2006) showed that goethite may transform into protohematite and then

hydrohematite and finally into hematite on heating.

Heating will transform gibbsite into an amorphous material (MacKenzie et al.,

1999). During thermal treatment, gibbsite may transform into chi alumina, then

progressively into gamma, theta and alpha alumina (corundum) (Bokhimi et al.,

2002). Wang et al. (2006) summarized dehydration of gibbsite as a complex process

depending on particle size and heating rate as follows:

Gibbsite amorphous phase

Gibbsite boehmite amorphous phase

Clearly there are several possible transformations that reflect the nature of the

gibbsite and heating regime.

2.6 Studies on Minerals Reversion

As discussed above, there have been many studies on the transformations of

kaolinite, gibbsite and goethite on heating. However, very few studies have

considered the reversion of heated kaolinite and gibbsite and there are no published

data on the reversion of heated goethite.

Rocha et al. (1990) studied rehydroxylation of metakaolinite to kaolinite using Solid

State NMR and cognate techniques. They measured specific surface area and weight

loss of the mineral based on reaction times for several hydrothermal treatments

(Table 2.2). Metakaolinite was transformed back into kaolinite by suitable

experimental conditions. Rocha et al., (1990) achieved rehydroxylation of

metakaolinite to kaolinite by heating in an autoclave at 155oC for 1, 2, 7 and 14 days,

at 200oC for 1 and 2 days, and at 250oC for 3 and 6 days. Infrared spectra and XRD

results showed that at 155oC for 14 days, 250oC for 3 days or more the

characteristics of kaolinite reappeared. The electron microscopy and specific surface

area (SSA) measurements support their finding. Hydrothermally treated

16

metakaolinite reverted kaolinite to had a much larger SSA and smaller particle size

than the parent kaolinite. Rocha and Klinowski (1991) considered that the alteration

of kaolinite to metakaolinite can be reversed completely. They considered that there

are three possible mechanisms for the rehydration process: (i) dissolution of

metakaolinite particles followed by subsequent crystallization; (ii) local dissolution

of micro-regions of the metakaolinite particles followed by crystallization of small

kaolinitic nuclei which then increase in size; (iii) a purely solid state process

whereby chemical bonds are rearranged.

Table 2.2. Specific surface area and water content, expressed as TGA weight loss for

metakaolinite, kaolinite and rehydrated kaolinite samples. (Rocha et al, 1990).

Dehydration treatment day (temperature oC) Surface area (m2g-1a) Weight Loss (%c)

0 (metakaolinite)

2 (155)

7 (155)

14 (155)

3 (250)

6 (250)

kaolinite

17

31

46

87

77

74

10b

1.1

3.7

5.6

12.8

12.9

13.6

13.8

a ± 4 m2g-1

b ± 1 m2g-1

c ± 0.1%

Frost et al. (2003) conducted experiments on the transformation of mechanically

dehydroxylated kaolinite into kaolinite by using a formamide-intercalated

intermediate. They showed that ageing of mechanochemically activated kaolinite

allowed the reformation of kaolinite through the disruption of clay layers.

Intercalation of reorganised kaolinite with formamide resulted in an enlarged phase

with a d-spacing of 10.2 Å. They considered that dehydroxylation arises through a

consistency procedure involving transfer of protons. Mechanochemically activated

kaolinite with intercalation of formamide aged in the presence of water results in de-

intercalation of formamide and the de-intercalated kaolinite returns to its original d-

spacing.

17

Miśta and Wrzyszcz (1999) studied the rehydration of transition aluminas prepared

by flash calcination of gibbsite. The results show the relation of the phase transitions

associated with the recrystallization of amorphous alumina that are produced through

the contact of water with the alumina surface. The final rehydration product was

crystalline Al(OH)3, mainly bayerite, which was observed by XRD and TEM after

100 hours of rehydration at 25oC. XRD results illustrated the alteration of an

amorphous phase that transform into pseudoboehmite then into Al(OH)3, mainly

bayerite. After rehydration at 216 hours at 50oC, small amounts of gibbsite were

detected.

In this review of the literature, it has been established that soil is heated by forest fire

and that:

(a) hydroxylated soil minerals (particularly kaolinite, gibbsite and goethite) may

become dehydroxylated creating diverse, sometimes disordered minerals.

These minerals may exhibit increased chemical reactivity and so affect

adsorption of plant nutrient and other ions. The dehydroxylated minerals may

rehydroxylate under ambient conditions.

(b) ash deposited during bushfires may contains diverse compounds that will

variously affect soil properties. The nature of ash compounds is poorly

understood and more detailed information will help with the prediction of the

extent and rate of dissolution of plant nutrient ions in ash.

This thesis is focussed on these two important issues.

18

Chapter 3

3.0 Rehydration of heated gibbsite, kaolinite and goethite: an assessment of

properties and environmental significance

3.1. Introduction

Several common micrometric minerals in soils contain structural hydroxyl ions that

are lost on heating. Kaolinite (Si2O5(OH)4Al2), goethite (α-FeOOH) and gibbsite (Al

(OH)3) are major constituents of highly weathered soils (Schwertmann and Taylor,

1989). These minerals will be affected by heating during managed and natural forest

fires. Forest fires may heat topsoils to temperatures in excess of 500oC (Sertsu and

Sanchez, 1978; Chandler et al., 1983). The effects of fire on soil mineralogy are

poorly known, although these heating temperatures will dehydroxylate several

minerals (Ketterings et al., 2000).

There have been numerous laboratory studies of the dehydroxylation of kaolinite,

goethite and gibbsite (Rocha et al., 1990; Ruan et al., 2002; De Faria and Lopes,

2007; Landers and Gilkes, 2007). Kaolinite dehydroxylates to form metakaolinite at

temperatures between 450oC to 600oC (Grim, 1968; Richardson, 1972; Babuskhin et

al., 1985). Gibbsite is dehydroxylated at about 200oC to produce a mixture of

boehmite and amorphous alumina (Rooksby, 1972) and goethite transforms to

hematite at about 300oC (Cornell and Schwertmann, 1996). The question arises as to

whether rehydroxylation of these three dehydroxylated minerals is possible. Heated

gibbsite does rehydroxylate readily in the laboratory (Miśta and Wrzyszcz, 1999)

and a small amount of kaolinite may be regenerated from metakaolinite although the

process is slow (Grim and Bradley, 1948; Rocha et al., 1990). There are no

corresponding studies of the rehydroxylation of heated goethite. Dehydroxylation of

these minerals may affect their specific surface area and surface reactions, which

will be of significance to chemical reactions in soils that involve adsorption of ions

such as phosphate (Ketterings et al., 2002).

Soils mostly provide a humid environment so that minerals that have been

dehydroxylated in a bush fire may tend to rehydroxylate in the soil. The apparent

absence of large amounts of dehydroxylated minerals in frequently burnt soils

19

supports this proposition but the topic has not been adequately investigated. This

chapter investigates the rehydration of dehydroxylated kaolinite, goethite and

gibbsite in the laboratory to identify if rehydroxylation is likely to occur under soil-

like conditions.

3.2. Material and methods

We investigated how heating impacts three soil minerals (kaolinite, goethite and

gibbsite) and their subsequent rehydroxylation.

3.2.1. Pure minerals and heating procedures

A pure synthetic gibbsite sample was supplied by Alcoa, Western Australia,

kaolinite came from the McNamee Pit Bath, South Carolina, United States and

goethite from a lateritic soil at Koniambo, New Caledonia. The three minerals were

heated for one hour at temperatures above and below their DTA dehydroxylation

maxima. Kaolinite (K) was heated at 500oC (K500), 550oC (K550) and 600oC

(K600); gibbsite (Gi) was heated at 250oC (Gi250), 300oC (Gi300) and 350oC

(Gi350); and goethite (Go) at 250oC (Go250), 300oC (Go300) and 350oC (Go350).

For the rehydration experiment, 6.5 g of the heated mineral were mixed with 26 cm3

of water before heating in a sealed container in an oven for 0, 14, 70, 200 and 400

days at two temperatures (55 and 95oC).

3.2.2. Chemical and morphological analysis

Mineral properties were investigated using several techniques. Conventional XRD

analysis was conducted with a Philips PW3020 diffractometer with a graphite

diffracted beam monochromator (CuKα, 50kV, 20 mA) and scans from 4 to 70o 2θ.

Synchrotron XRD (SXRD) analysis was performed at the Australian Synchrotron,

where powder samples were mounted into glass capillaries and scanned from 4-60o

2θ. The wavelength for SXRD was set at ~1.0 Å to provide a high peak/background

in order to identify minor constituents. Thermal analysis (TGA, DTGA, DTA) was

done on a STA 6000 instrument (Perkin-Elmer, Norwalk, CT, USA), and

transmission electron microscopy on a JEOL 3000 FEG electron microscope

equipped with an Oxford Instruments INCA 200 Energy Dispersive Spectrometer

(EDS). Specific surface area (SSA) was measured using a Micrometrics Gemini

2375 instrument with VacPrep 061 using a five point B. E. T. method with N2 as the

20

absorbate. Phosphate (P) adsorption (x) was measured following the Ozanne and

Shaw (1967) method with the P concentration in the filtrate (c) being determined by

the molybdate blue method (Murphy and Riley 1962). The P adsorption data were

fitted to the linear form of the Langmuir equation as follows, where xm is the P

adsorption maximum:

c/x = (bxm)-1 +(c/xm) (Barrow 1978)

Fourier transform infrared (FTIR) spectra were obtained with a Perkin Elmer

Spectrum One spectrometer with samples dried at 105oC for 24 hours prior to

analysis. Samples were prepared in KBr disks with sample to KBr ratio of 1: 300.

Elemental composition of the minerals (Table 3.1) was determined using an

inductively coupled plasma optical emission spectrometer (ICP-OES) (Perkin-Elmer,

Norwalk, CT, USA) after perchloric acid digestion where minerals were completely

dissolved. This acid digestion procedure was validated using standard minerals and

rocks.

Table 3.1. Properties of mineral samples used in the study.

Properties Kaolinite Goethite Gibbsite

pH (1:5 H2O) 4.54 4.81 7.80

EC (1:5 H2O) μS/cm 79.3 48.1 53.4

Total P (mg/kg) 249 nd nd

Total Ca (mg/kg) 948 37 nd

Total Mg (mg/kg) 1180 1271 nd

Total K (mg/kg) 34 nd 45

Total Na (mg/kg) 204 nd 100

Total S (mg/kg) 62 2132 8

Total Cd (mg/kg) 0.2 nd nd

Total Ni (mg/kg) 19 10270 nd

Total As (mg/kg) 3 nd 2

Total Pb (mg/kg) 27 nd nd

Total Cu (mg/kg) 15 12 nd

Total Mn (mg/kg) 42 4131 nd

Total Zn (mg/kg) 13 255 nd

Total Fe (%) 0.39 54.7 0.04

Total Al (%) 12 2 36

Note: nd-Not detected.

21

3.3. Results and Discussion

3.3.1. X-ray diffraction and chemical data

Examples of conventional XRD random powder patterns of original and variously

dehydroxylated kaolinite, goethite and gibbsite are shown in Figure 3.1. Original

kaolinite contains minor amounts of illite and anatase. The quite high Mg content

represents exchangeable Mg (Table 3.1). The high Ni and quite high Mg and Mn

content of the goethite are a consequence of its origin in a weathered ultramafic rock.

Soil goethite commonly contains substantial amounts of Al (Cornell and

Schwertmann, 1996).

On heating at 500-600oC, kaolinite dehydroxylated to form metakaolinite (Figure

3.1A), a compound where the two dimensional Si-O arrangement remains but the

Al-O network is disorganised (Yokozeki et al., 2004). Heating at 300 and 350oC

transformed goethite into poorly ordered hematite (Figure 3.1C). Goethite only

partly dehydroxylates at these temperatures so that the hematite is more

appropriately described as hydrohematite (Pomies et al., 1998). Fan et al. (2006)

considered that goethite progressively transforms into protohematite, then into

hydrohematite and finally into hematite on heating.

Gibbsite heated at 350oC altered to an amorphous phase and minor boehmite (Figure

3.1E). Wang et al. (2006) showed that the structural transformations during

dehydration of gibbsite is a complex process that is affected by crystal size and

heating conditions as follows:

Gibbsite amorphous phase and simultaneously

Gibbsite boehmite amorphous phase

Synchrotron XRD patterns of original and heated kaolinite goethite and gibbsite in a

glass capillary were also obtained as they offer better resolution of weak and

adjacent reflections than conventional XRD (Williams et al., 2003), but the glass of

the capillaries contributed to the broad scattering characteristic of amorphous

compounds so that conventional XRD was also used (Figure 3.1).

22

A. Kaolinite, Conventional XRD.

5 15 25 35 45 55 65

2θ Cu Kα

Kaolinite

K500

Metakaolinite K600

K550 I

A

B. Goethite, Conventional XRD.

5 15 25 35 45 55 65

2θ Cu Kα

Go350 Hematite

Goethite

Go300

Go250

C. Gibbsite Conventional XRD.

5 15 25 35 45 55 65

2θ Cu Kα

Gi350

Gibbsite

Boehmite and amorphous material

Gi250

Gi300

Boehmite

Figure 3.1. Conventional XRD patterns for original and dehydroxylated kaolinite

(A), goethite (B) and gibbsite (C) heated at the indicated temperatures.

23

A. Kaolinite heated at 600oC.

5 15 25 35 45 55 65

2θ Cu Kα

Kaolinite

K600

K600-0

K600-55-14

K600-95-14

K600-95-70

K600-55-70

K600-55-200

K600-95-200

Metakaolinite

K600-55-400

K600-95-400 Metakaolinite I I A

B. Goethite heated at 350oC.

5 15 25 35 45 55 65

2θ Cu Kα

Goethite

Go350

Go350-0

Go350-55-14

Go350-95-14

Go350-95-70

Go350-55-70

Go350-55-200

Go350-95-200

Hydrohematite

Go350-55-400

Go350-95-400 H

H H

C. Gibbsite heated at 350oC.

5 15 25 35 45 55 65

2θ Cu Kα

Gi350-55-14 Gi350-95-14

Gi350 Gibbsite

Gi350-0

Gi350-95-70 Gi350-55-70

Gi350-55-200 Gi350-95-200

Bo Gi

Ba

Ba

Ba

Gi

Gi Ba Ba Gi Bo Bo

Gi Ba

Boehmite and amorphous material

Bo

Gi350-55-400

Gi350-95-400

Figure 3.2. XRD patterns for heated kaolinite (A), goethite (B) and gibbsite (C), wet

incubated at 55/95 oC, 0-400 days, Cu Kα radiation. I=illite, A=anatase, H=hematite,

Bo=boehmite, Gi=gibbsite, and Ba=bayerite.

24

Figure 3.2 and 3.3 show XRD and SXRD patterns respectively for heated kaolinite,

goethite and gibbsite after wet incubation for various times and temperatures. SXRD

showed a much better peak to background discrimination than conventional XRD

but the broad scattering maximum due to glass was dominant for SXRD patterns

(Figure 3.3). Both techniques showed that metakaolinite formed from kaolinite at

600oC (Fig. 3.2 A and 3.3 A) persisted during wet incubation, and rehydration of this

material to reform kaolinite did not occur for up to 400 days of incubation, at 55 or

95oC. This results contrasts with findings of Rocha and Klinowski (1991) and Frost

et al. (2003) who considered that dehydration of kaolinite to form metakaolinite is

completely reversible. They state that there are three possible mechanisms for this

rehydration process: (i) dissolution of metakaolinite particles followed by

crystallization; (ii) local dissolution of micro-regions of the metakaolinite particles

followed by crystallization of small kaolinitic nuclei, which then increase in size;

(iii) a purely solid state process whereby chemical bonds are rearranged. However

under the condition of the present experiment, which more closely resembles

ambient conditions than those used by the cited authors, none of these changes were

observed.

Hydrohematite formed from goethite at 350oC (Figure 3.2 B and 3.3 B) also showed

no reversion during wet incubation so that after 400 days of incubation at 55 and

95oC, the broad hematite reflections persisted. There are no published studies on the

rehydroxylation of hematite into goethite. Heated gibbsite (initially boehmite and

amorphous material) rapidly recrystallised during wet incubation and after 14 days at

95oC, boehmite (Bo), bayerite (Ba) and gibbsite (Gi) had formed (Figure 3.2C and

3.3C). The rehydroxylation process was much slower at 55oC but was extensive after

400 days.

25

5 15 25 35 45 55 65

2θ angle

Kaolinite

K600-95-400

K600

K600-95-200

Metakaolinite

5 15 25 35 45 55 65

2θ angle

Go350

Go350-95-200

Go350-95-400

Hydrohematite

Goethite

5 15 25 35 45 55 65

2θ angle

Gibbsite

Gi350-95-400

Gi350

Gi350-95-200

Boehmite and amorphous material

Figure 3.3. Synchrotron XRD patterns for unheated and 600oC heated kaolinite (A),

350oC heated goethite (B) and 350oC heated gibbsite (C), wet incubated at 95oC, for

200 and 400 days, Cu Kα radiation.

A

B

C

26

3.3.2. Thermal analysis

TGA and DTA plots for kaolinite are shown in Figure 3.4A. The endotherm at

520oC corresponds to dehydroxylation of kaolinite. The sample weight was reduced

by 15.3 wt.% during the dehydroxylation process, which is close to the theoretical

value of 14 wt.% (Ptáček et al., 2010). The additional 1.3 wt.% presumably

corresponds to strongly adsorbed water that was lost between 110 and 400oC.

Goethite dehydroxylation is shown in Figure 3.5B with an endotherm and associated

water loss peaking at 280oC (Prasad et al., 2006). TGA and DTA data for gibbsite

(Figure 3.4C) show a strong endotherm at 300oC (Pereira et al., 2009). This

endothermic peak is due to gibbsite dehydroxylation to boehmite and amorphous

material. A small endotherm at 530oC is associated with the dehydroxylation of

boehmite (MacKenzie, 1957).

The TGA and DTA results for the three minerals after the rehydration treatment are

shown in Figure 3.5. The complete thermal analysis results for kaolinite, goethite

and gibbsite, wet incubated at 55 and 95oC are presented in Appendix 1, 2, 3, 4, 5

and 6. All three minerals had acquired structural water (we define this as water lost

at T > 110oC) during the rehydration treatment. Neither rehydrated heated kaolinite

nor goethite provided the sharp DTGA/DTA dehydroxylation peaks that would

indicate that crystalline minerals had formed. This result is consistent with XRD data

that indicated that the metakaolinite formed from kaolinite heated at 600oC and

hematite formed from goethite at 350oC had not developed an ordered structure

during the wet incubation. However thermal analysis data show that rehydroxylation

and recrystallisation of heated gibbsite was extensive at 95oC and both DTA/TGA

and XRD results indicate that boehmite (Bo), bayerite (Ba) and gibbsite (Gi) had

formed (Figure 3.2C).

27

A

B

C

Figure 3.4. Thermal analysis results for original kaolinite (A), goethite (B) and

gibbsite (C) showing dehydroxylation peaks.

DTGA

DTA

TGA

520oC

DTGA

DTA

TGA

520oC

TGA

DTA

DTGA

280oC

TGA

DTA

DTGA

280oC

DTGA

DTA

TGA

300oC

510oC

DTGA

DTA

TGA

300oC

510oC

28

A

B

C

Figure 3.5. Thermal analysis results for kaolinite (A) previously heated at 600oC,

goethite (B) and gibbsite previously heated at 350oC (C): then wet incubated at 95oC

for 400 days.

DTGA

DTA

TGA

DTGA

DTA

TGA

DTGA

DTA

TGA

DTGA

DTA

TGA

DTGA

DTA

TGA

DTGA

DTA

TGA

300oC

510oC

DTGA

DTA

TGA

300oC

510oC

29

Figure 3.6. Weight loss measured using TGA (110-840oC) for dehydroxylated

kaolinite (600oC), goethite (350oC) and gibbsite (350oC) incubated for 0, 14, 70, 200

and 400 days at 55 and 95oC.

A histogram showing weight loss (105-840oC) from the incubated heated minerals

after various incubation times (Figure 3.6) indicates that water had been incorporated

into the structure of the rehydroxylated minerals rather than being simply adsorbed

0

2

4

6

8

10

12

14

16

18

Kaolinite Kaolinite600

Kaolinite600-0

Kaolinite600-55-

14

Kaolinite600-95-

14

Kaolinite600-55-

70

Kaolinite600-95-

70

Kaolinite600-55-

200

Kaolinite600-95-

200

Kaolinite600-55-

400

Kaolinite600-95-

400

TGA Weight Loss(%)

Kaolinite

0

5

10

15

20

25

30

35

40

Gibbsite Gibbsite350

Gibbsite350-0

Gibbsite350-55-

14

Gibbsite350-95-

14

Gibbsite350-55-

70

Gibbsite350-95-

70

Gibbsite350-55-

200

Gibbsite350-95-

200

Gibbsite350-55-

400

Gibbsite350-95-

400

TGA Weight Loss(%)

Gibbsite

0

2

4

6

8

10

12

14

16

18

20

Goethite Goethite350

Goethite350-0

Goethite350-55-

14

Goethite350-95-

14

Goethite350-55-

70

Goethite350-95-

70

Goethite350-55-

200

Goethite350-95-

200

Goethite350-55-

400

Goethite350-95-

400

TGA Weight Loss(%)

Goethite

30

(i.e. adsorbed water is assumed lost at <105oC). Rehydrated kaolinite that had been

first heated at 600oC and then incubated for 400 days at 95oC contained nearly 14%

water, which is 11% more than the 3% water that remained in the heated kaolinite.

We consider that the water incorporated into metakaolinite during incubation may

represent an initial step towards the eventual recrystallisation of kaolinite, although

this water is not lost during a distinct dehydroxylation endotherm at 520oC as occurs

for kaolinite. Hematite (hydrohematite) formed at 350oC contained about 6%

residual water and only gained a small amount (up to 2%) of additional water during

incubation. This water was lost over a wide range of temperatures rather than at

280oC as occurred for dehydroxylation of goethite. Gibbsite heated at 350oC retained

about 21% H2O in boehmite and amorphous alumina and acquired up to 10% water

during incubation, some of which was incorporated into the structures of boehmite,

bayerite and gibbsite that recrystallised during incubation. The distinct

dehydroxylation endotherms at 300 and 510oC correspond to these minerals.

3.3.3. Specific surface area and phosphate adsorption

Dehydroxylation of goethite at 300 and 350oC caused substantial increases in

specific surface area (SSA) with no corresponding systematic effect for kaolinite and

gibbsite (Figure 3.7). Landers et al. (2009) observed a 1.5-2.6-fold increase in

specific surface area due to dehydroxylation of goethite that reflected the

development of micropores in the newly formed OH-hematite. The SSA of

dehydroxylated goethite (0, 250, 300, 350oC) was not systematically affected by

rehydration (0-400 days) at 55 and 95oC. The SSA of dehydroxylated kaolinite (500,

550, 600oC) increased substantially with rehydration period (0-400 days) at 55 and

95oC. This result is consistent with observations of Rocha et al. (1990) who found

that the SSA of metakaolinite treated at 155oC for 2-14 days in an autoclave

increased from 10 to 87 m2/g with increasing reaction time. The SSA of heated (250,

300, 350oC) gibbsite initially increased with incubation times of 14 or 70 days for

both incubation temperatures (55, 95oC) then decreased substantially due to the

growth of crystals of gibbsite, boehmite and bayerite. These changes were

particularly large for 350oC heated gibbsite.

The increase of specific surface area and reduced crystalline order of heated and

rehydroxylated minerals is likely to affect some soil chemical reactions. Sanchez

31

(1976) states that minerals have a higher P sorption capacity in the order: 2:1

minerals < kaolinite < gibbsite = goethite < amorphous oxides. Ketterings et al.

(2002) considered that the loss of sorption sites due to the dehydroxylation of

kaolinite, gibbsite and goethite is offset by the formation of amorphous phases with a

higher specific surface area.

Phosphate adsorption maxima (Xm) values for kaolinite (Figure 3.8) were increased

to a minor extent by dehydroxylation at 500-600 oC. There were mostly systematic

increases in P adsorption maxima with incubation treatment for heated kaolinite. The

largest increase was by 697 μg P/g for 600oC heated kaolinite incubated at 95oC for

70 days. Dehydroxylation of goethite at 250-350oC did not affect P adsorption

maximum but the P adsorption maximum (μg/g) of dehydroxylated goethite (0, 250,

300, 350oC) increased substantially (up to 596 μg P/g) with rehydroxylation at 55oC

for 400 days (Figure 3.8). The P adsorption maximum of gibbsite was increased

slightly (≈ 166 μg P/g) by dehydroxylation at 250 to 350oC (Figure 3.8) and mostly

did not change systematically with rehydration time (0-400 days) at 55 and 95oC.

The relationships between P adsorption maximum and specific surface area of the

rehydroxylated minerals are shown in Figure 3.9. There is no clear systematic

relationship between specific surface area and P adsorption maximum for any

mineral or rehydroxylation treatment. Presumably the nature of the exposed surfaces

has a greater effect on P adsorption than does the SSA.

3.3.4. Infrared analysis

Infrared spectra for kaolinite, goethite, gibbsite and their dehydroxylated and

rehydrated products are shown in Figure 3.10. The complexity of the infrared spectra

with many of the peaks being due to ordered arrays of ions was reduced by heating

as three dimensional ordering was decreased or destroyed. The loss of peaks and

development of broad bands for all three minerals coincided with the loss of sharp

XRD reflections. Heating causes those bands due to lattice modes to generally shift

to lower wavenumbers (Freund, 1974). Tables 3.2, 3.3 and 3.4 show infrared

absorption peaks for kaolinite, goethite, gibbsite and their various dehydroxylated

and rehydrated products compared with published assignments for the absorption

peaks.

32

0

10

20

30

40

50

60

70

80

K

K-5

5-14

K

-95-

14

K-5

5-70

K

-95-

70

K-5

5-20

0 K

-95-

200

K-5

5-40

0 K

-95-

400

K50

0 K

500-

0 K

500-

55-1

4 K

500-

95-1

4 K

500-

55-7

0 K

500-

95-7

0 K

500-

55-2

00

K50

0-95

-200

K

500-

55-4

00

K50

0-95

-400

K

550

K55

0-0

K55

0-55

-14

K55

0-95

-14

K55

0-55

-70

K55

0-95

-70

K55

0-55

-200

K

550-

95-2

00

K55

0-55

-400

K

550-

95-4

00

K60

0 K

600-

0 K

600-

55-1

4 K

600-

95-1

4 K

600-

55-7

0 K

600-

95-7

0 K

600-

55-2

00

K60

0-95

-200

K

600-

55-4

00

K60

0-95

-400

SS

A (

sq m

/g)

500 oC heated

550 oC heated

600 oC heated

Kaolinite

Original

0

20

40

60

80

100

120

140

160

Go

Go-

55-1

4 G

o-95

-14

Go-

55-7

0 G

o-95

-70

Go-

55-2

00

Go-

95-2

00

Go-

55-4

00

Go-

95-4

00

Go2

50

Go2

50-0

G

o250

-55-

14

Go2

50-9

5-14

G

o250

-55-

70

Go2

50-9

5-70

G

o250

-55-

200

Go2

50-9

5-20

0 G

o250

-55-

400

Go2

50-9

5-40

0 G

o300

G

o300

-0

Go3

00-5

5-14

G

o300

-95-

14

Go3

00-5

5-70

G

o300

-95-

70

Go3

00-5

5-20

0 G

o300

-95-

200

Go3

00-5

5-40

0 G

o300

-95-

400

Go3

50

Go3

50-0

G

o350

-55-

14

Go3

50-9

5-14

G

o350

-55-

70

Go3

50-9

5-70

G

o350

-55-

200

Go3

50-9

5-20

0 G

o350

-55-

400

Go3

50-9

5-40

0

SS

A (s

q m

/g)

Goethite

Original

250 oC heated 300 oC heated 350 oC heated

0

10

20

30

40

50

60

70

80

Gi

Gi-5

5-14

G

i-95-

14

Gi-5

5-70

G

i-95-

70

Gi-5

5-20

0 G

i-95-

200

Gi-5

5-40

0 G

i-95-

400

Gi2

50

Gi2

50-0

G

i250

-55-

14

Gi2

50-9

5-14

G

i250

-55-

70

Gi2

50-9

5-70

G

i250

-55-

200

Gi2

50-9

5-20

0 G

i250

-55-

400

Gi2

50-9

5-40

0 G

i300

G

i300

-0

Gi3

00-5

5-14

G

i300

-95-

14

Gi3

00-5

5-70

G

i300

-95-

70

Gi3

00-5

5-20

0 G

i300

-95-

200

Gi3

00-5

5-40

0 G

i300

-95-

400

Gi3

50

Gi3

50-0

G

i350

-55-

14

Gi3

50-9

5-14

G

i350

-55-

70

Gi3

50-9

5-70

G

i350

-55-

200

Gi3

50-9

5-20

0 G

i350

-55-

400

Gi3

50-9

5-40

0

SS

A (s

q m

/g)

Gibbsite

Original 250 oC heated

300 oC heated

350 oC heated

Figure 3.7. Specific surface area (SSA) of variously dehydroxylated kaolinite,

goethite and gibbsite samples incubated for 0, 14, 70, 200 and 400 days at 55 and

95oC.

Original 250oC heated

Original

33

0

200

400

600

800

1000

1200

1400

1600

K

K-55

-14

K-95

-14

K-55

-70

K-95

-70

K-55

-200

K-95

-200

K-55

-400

K-95

-400

K500

K500

-0

K500

-55-

14

K500

-95-

14

K500

-55-

70

K500

-95-

70

K500

-55-

200

K500

-95-

200

K500

-55-

400

K500

-95-

400

K550

K550

-0

K550

-55-

14

K550

-95-

14

K550

-55-

70

K550

-95-

70

K550

-55-

200

K550

-95-

200

K550

-55-

400

K550

-95-

400

K600

K600

-0

K600

-55-

14

K600

-95-

14

K600

-55-

70

K600

-95-

70

K600

-55-

200

K600

-95-

200

K600

-55-

400

K600

-95-

400

P So

rptio

n M

axim

um (μ

g/g)

Kaolinite

Original 500 oC heated 550 oC heated

600 oC heated

0

200

400

600

800

1000

1200

1400

1600

Go

Go-

55-1

4

Go-

95-1

4

Go-

55-7

0

Go-

95-7

0

Go-

55-2

00

Go-

95-2

00

Go-

55-4

00

Go-

95-4

00

Go2

50

Go2

50-0

Go2

50-5

5-14

Go2

50-9

5-14

Go2

50-5

5-70

Go2

50-9

5-70

Go2

50-5

5-20

0

Go2

50-9

5-20

0

Go2

50-5

5-40

0

Go2

50-9

5-40

0

Go3

00

Go3

00-0

Go3

00-5

5-14

Go3

00-9

5-14

Go3

00-5

5-70

Go3

00-9

5-70

Go3

00-5

5-20

0

Go3

00-9

5-20

0

Go3

00-5

5-40

0

Go3

00-9

5-40

0

Go3

50

Go3

50-0

Go3

50-5

5-14

Go3

50-9

5-14

Go3

50-5

5-70

Go3

50-9

5-70

Go3

50-5

5-20

0

Go3

50-9

5-20

0

Go3

50-5

5-40

0

Go3

50-9

5-40

0

P So

rptio

n M

axim

um (μ

g/g)

Goethite

Original

250 oC heated 300 oC heated 350 oC heated

0

200

400

600

800

1000

1200

1400

1600

Gi

Gi-5

5-14

Gi-9

5-14

Gi-5

5-70

Gi-9

5-70

Gi-5

5-20

0

Gi-9

5-20

0

Gi-5

5-40

0

Gi-9

5-40

0

Gi2

50

Gi2

50-0

Gi2

50-5

5-14

Gi2

50-9

5-14

Gi2

50-5

5-70

Gi2

50-9

5-70

Gi2

50-5

5-20

0

Gi2

50-9

5-20

0

Gi2

50-5

5-40

0

Gi2

50-9

5-40

0

Gi3

00

Gi3

00-0

Gi3

00-5

5-14

Gi3

00-9

5-14

Gi3

00-5

5-70

Gi3

00-9

5-70

Gi3

00-5

5-20

0

Gi3

00-9

5-20

0

Gi3

00-5

5-40

0

Gi3

00-9

5-40

0

Gi3

50

Gi3

50-0

Gi3

50-5

5-14

Gi3

50-9

5-14

Gi3

50-5

5-70

Gi3

50-9

5-70

Gi3

50-5

5-20

0

Gi3

50-9

5-20

0

Gi3

50-5

5-40

0

Gi3

50-9

5-40

0

P So

rptio

n M

axim

um (μ

g/g)

Gibbsite

Original 250 oC heated 300 oC heated 350 oC heated

Figure 3.8. Langmuir P sorption maxima for variously dehydroxylated kaolinite,

goethite and gibbsite incubated for 0, 14, 70, 200 and 400 days at 55 and 95oC.

250oC heated

34

A.

0 10 20 30 40 50 60 70 80

0 200 400 600 800 1000 1200

SSA

(sq m

/g)

P Sorption Maximum (μg/g)

Kaolinite

K

K500

K550

K600

B.

0 20 40 60 80

100 120 140 160

0 200 400 600 800 1000 1200 1400 1600

SSA

(sq

m/g

)

P Sorption Maximum (μg/g)

Goethite

Go

Go250

Go300

Go350

C.

0 10 20 30 40 50 60 70 80

0 200 400 600 800 1000

SSA

(sq m

/g)

P Sorption Maximum (μg/g)

Gibbsite

Gi

Gi250

Gi300

Gi350

Figure 3.9. P sorption maximum versus specific surface area for variously

dehydroxylated kaolinite (A), goethite (B) and gibbsite (C) incubated for 0, 14, 70,

200 and 400 days at 55 and 95oC.

35

Kaolinite mostly altered into metakaolinite at 550oC and fully altered at 600oC. The

several sharp OH stretching vibrations (3800-3500 cm-1) and the Si-O stretching and

Si-O-Al combination bands (1300-40 cm-1) were replaced by broad bands centered at

about 3450 cm-1 and 1050 cm-1 respectively (Figure 3.10). The infrared spectrum of

heated kaolinite after incubation for 400 days at 55 and 95oC was unchanged

showing that metakaolinite persisted with no recrystallisation of kaolinite, which is

consistent with XRD and TGA results.

Broad bands at 3399-3411 and 3188 cm-1 for goethite may be assigned to H-O-H and

O-H stretching vibrations (Figure 3.9 and Table 3.3) (Prasad et al., 2006). Ruan et al.

(2001) observed bands between 3450-3445 and 3212-3194 cm-1 in the spectra of

hematite formed by dehydroxylation of goethite at 200oC whereas hematite formed

at 300oC and 350oC had one broad band centered at 3400 cm-1. Hydroxyl

deformation bands at 898 and 795 cm-1 did persist in hematite. Bands with

wavenumbers lower than 700 cm-1 are due to lattice FeO6 vibrations and persisted in

hematite (Ruan et al., 2001). There was no significant change in the infrared

spectrum for heated goethite after rehydration treatment for 400 days at 55 and 95oC.

The OH stretching and bending vibrations of dehydroxylated and rehydroxylated

gibbsite are shown in Figure 3.10 and Table 3.4. Gibbsite has four stretching

vibrations in the range of 3621-3371 cm-1, OH bending vibrations are at about 914,

968 and 1022 cm-1. Other vibrations are in the range 666-483 cm-1 (Kloprogge et al.,

2001). Gibbsite dehydroxylated to form a mixture of boehmite and an amorphous

phase. According to Ryskin (1974), boehmite has an OH stretching band with two

strong bands centered at 3297 and 3090 cm-1. In our study gibbsite heated at 350oC

showed strong OH stretching bands at 3292 cm-1 and, 3098 cm-1, which are assigned

to boehmite (Table 3.4). The boehmite OH bending region of the spectrum is

characterized by absorption at 1160 and 1080 cm-1 with an additional band at 755

cm-1 (Kloprogge et al., 2001). Heated gibbsite wet incubated at 55 and 95oC for 400

days had OH stretching modes for bayerite at 3658, 3546, 3419 and 3496 cm-1

(Table 3.4) as has been observed by Lee and Condrate (1995).

36

Table 3.2. Infrared absorption spectra maxima for variously dehydroxylated kaolinite

samples and 600oC heated kaolinite wet incubated at 55/95oC for 400 days,

references values and assignments.

Wavenumber of major absorption peaks (cm-1) Reference Spectra Assignment

Kaolinite K500 K550 K600 K600-55-400 K600-95-400 Kaolinitea,b Metakaoliniteb

470

539

470

539

469

539

467 467 467 472

540

560

Si-O

stretching

and Si-O-Al

combination

bands

696

754

791

696

754

791

688

799

662

811

662

811

662

811

700

754

790

665

810

912

936

912

936

912 915

940

1007 1007 1012

1031

1114

1031

1114

1059

1066

1066

1066

1038

1108

1070

3435

3620

3651

3435

3620

3651

3435

3621

3656

3435 3435 3435

3620

3650

3434

OH

Stretching

3669

3695

3669

3695

3695

3670

3695

aWhite & Roth (1986), bRocha et al. (1990)

37

Table 3.3. Infrared absorption spectra maxima for variously dehydroxylated goethite

samples and 350oC heated goethite wet incubated at 55/95oC for 400 days,

references values and assignments.

Wavenumber of major absorption peaks (cm-1) Reference Spectra Assignment Band

Component Goethite Go250 Go300 Go350 Go350-55-400 Go350-95-400 Goethite a,b Hematite a

460 455 454

538

454

538

454

538

454

538

461 454

536

Fe-O

Fe-O

Low-

frequency

region

618 618 619 619 Fe-O-H

795 797 799 800 γ(O-H) Hydroxyl

deformation

898 898 930 930 929 929 893 884 δ(O-H)

1630 1630

1637

1686

1640

1674

γ'(O-H)

δ'(O-H)

Water

bending

region

3188 3188 3206 3235 ν(O-H) Hydroxyl

stretching

3411 3399 3411 3411 3429 3429 3450 3446 ν(H-O-H) region

aRuan et al. (2001), bRussell and Fraser (1994).

38

Table 3.4. Infrared absorption spectra maxima for variously dehydroxylated gibbsite

samples and 350oC heated gibbsite wet incubated at 55/95oC for 400 days, references

values and assignments.

Wavenumber of major absorption peaks (cm-1) Reference Spectra Assignment Band

Component Gibbsite Gi250 Gi300 Gi350 Gi350-55-400 Gi350-95-400 Gibbsitea Boehmitea Bayeriteb 731 731 735 735 758 747 743 755 γ(OH) Hydroxyl

deformation 800 800 798 802 914 914 914 δ(OH) Water

bending 968 968 967 975 974 958 region 1022 1022 1022

1078 1019 1071

1019 1073

1020 1080 1160

3392 3435

3392 3435

3392 3447

3098 3435

3098 3419

3093 3292 3419

3380 3428

3090 3297

3440 3465

ν(OH) Hydroxyl stretching region

3479 3496 3477 3526 3526 3526

3548 3546

3520 3546 3550

3621 3621 3620 3617 3658

3616 3658

3617 3656

aKloprogge et al. 2001, bLee and Condrate (1995).

39

K500

K550

K600

Go250

450950145019502450295034503950

Wavenumber

Ab

so

rban

ce

Goethite

Goe-250

Goe-250-55-400

Goe-250-95-400

Go300

450950145019502450295034503950

Wavenumber

Ab

so

rban

ce

Goethite

Goe-300

Goe-300-55-400

Goe-300-95-400

Go350

450950145019502450295034503950

Wavenumber

Ab

so

rban

ce

Goethite

Goe-350

Goe-350-55-400

Goe-350-95-400

Gi250

Gi300

Gi350

Figure 3.10. Infrared spectra of original, dehydroxylated and rehydrated kaolinite,

goethite and gibbsite incubated for 400 days at 55 and 95oC.

3.3.5. Electron microscopy

Figure 3.11 shows TEM micrographs for original, dehydroxylated and rehydrated

kaolinite, goethite and gibbsite with particles ranging from <10 to >200 nm.

Metakaolinite obtained from kaolinite heated at 600oC had pseudohexagonal

platelets and remained unchanged after the incubation for 400 days at 95oC. TEM

micrographs of heated goethite indicate that the materials consist of particles with

diverse shapes and sizes with many particles being lathes. Gibbsite heated to 350oC

had a microporous, granular appearance due to rapid loss of structural water forming

voids. After incubation for 14 days at 95oC a complex intergrowth of more finely

divided hexagonal crystals had formed, which is consistent with XRD results that

450950145019502450295034503950

Wavenumber

Ab

so

rban

ce

Kaolinite

K500

K500-55-400

K500-95-400

OH Stretching

Si-O strectching and Si-O-Al combination bands

450950145019502450295034503950

Wavenumber

Ab

so

rban

ce

Kaolinite

K550

K550-55-400

K550-95-400

Si-O strectching and Si-O-Al combination bandsOH Stretching

450950145019502450295034503950

Wavenumber

Ab

sorb

ance

Kaolinite

K600

K600-55-400

K600-95-400

Si-O strectching and Si-O-Al combination bandsOH Stretching

450950145019502450295034503950

Wavenumber

Ab

so

rban

ce

Gib250

Gibbsite

Gib250-55-400

Gib250-95-400

γ(OH) ð(OH)

ν(OH)

Boehmite

450950145019502450295034503950

Wavenumber

Ab

so

rban

ce

Gib300

Gibbsite

Gib300-55-400

Gib300-95-400

ð(OH)

γ(OH) ν(OH)

Boehmite

450950145019502450295034503950

Wavenumber

Ab

so

rban

ce

Gib350

Gibbsite

Gib350-55-400

Gib350-95-400

ν(OH)

ð(OH)

γ(OH)

Boehmite

40

show that a mixture of boehmite, bayerite and gibbsite had replaced the amorphous

alumina.

Figure 3.11. Electron micrographs of original, dehydroxylated and rehydroxylated

kaolinite, goethite and gibbsite incubated for 400 days at 95oC.

240 nm K240 nm240 nm K 240 nm K600oC240 nm240 nm K600oC 240 nm K600-95-400240 nm240 nm K600-95-400

240 nm Go240 nm Go 240 nm Go350240 nm Go350 240 nm Go350-95-400240 nm Go350-95-400

240 nm Gi240 nm Gi 240 nm Gi350240 nm Gi350 240 nm Gi350-95-400240 nm Gi350-95-400

41

3.4. Conclusions

Kaolinite, goethite and gibbsite dehydroxylated to form metakaolinite, hematite and

boehmite and amorphous material respectively. This process may also occur in

natural environments such as in soils heated by bushfires. Dehydroxylation of these

three minerals caused structural changes and slight to moderate increases in both

specific surface area and phosphate adsorption maxima. The greater P adsorption

could be associated with the greater surface reactivity. Dehydroxylated kaolinite and

goethite showed no change in conventional and synchrotron XRD patterns due to

rehydration treatment. However, rehydration of 350oC heated gibbsite was extensive

at 95oC and after 14 days boehmite, bayerite and gibbsite had formed, the process

was slower at 55oC.

Infrared spectra of metakaolinite and hematite subjected to rehydration did not

change but thermal analysis indicated that the structural water content (>105oC)

increased substantially for metakaolinite. We propose that dehydroxylated minerals

such as kaolinite, goethite and gibbsite and their fully or partially rehydrated forms

may be present in naturally heated soils and may exert significant effects on the

chemical behaviour of soils. The effects may include changes in the availability of

phosphate and other nutrients.

42

Chapter 4

4.0 Short term effects of heating a lateritic podzolic soil on the availability to

plants of native and added phosphate

4.1. Introduction

Wildfire severely impacts ecosystems (FRA 2010) due to the destruction of

vegetation together with degradation of air quality, surface and ground water and soil

(Turrión et al., 2010). Fire is also used to remove plant residues in forest and

agricultural areas and also for management of natural vegetation (Khanna and

Raison, 1986). Forest fires may have significant impacts on physical, chemical,

mineralogical and biological properties of soils, thus affecting growth of plants

(Raison, 1979; Wan et al., 2001; Certini, 2005). These effects vary depending on

several environmental factors that control the combustion process, such as amount

and type of fuel, air temperature and humidity, wind speed and topography of the

area (Certini, 2005). Fire severity, intensity and soil type also affect soil properties

(Mataix-Solera et al., 2011).

Phosphorus (P) is one of the plant nutrients that is most affected by fire cycles

(Ferran et al., 2005). There may be a large increase in P in the soil surface shortly

after fire due to accumulation of ash (Wilbur and Christensen, 1985; Debano and

Klopatek, 1988), but the nett long-term loss of P due to burning reduces forest

productivity (Romanyá et al., 1994). Fire may lead to the transformation of the

chemical bonding of soil nutrients including P. This also includes the alteration of

organic forms into inorganic forms with associated changes in availability (Galang et

al., 2010). P losses in soluble and particulate forms can increase due to organic

matter mineralization and fire-induced soil erosion (Saa et al., 1993). Elemental P

may be volatilised (boiling point 280oC) at temperatures reached in fires and P is

also removed in smoke during fire (Cotton and Wilkinson, 1988). Raison et al.

(1985) found that P exported to the atmosphere may attain 50% of the total P in the

combusted fuels.

The effects of fire on other components of the P cycle are generally poorly known

(Debano and Klopatek, 1988; Saa et al., 1993) and in particular little is known of the

43

effects of heating on the transformation of dominant soil minerals and consequent

effects on P availability to plants.

Fire affects the nature of soil minerals and presumably their interactions with plant

nutrients. Soil temperatures in excess of 500oC can be reached during fires, which

will alter many hydroxylated soil minerals thereby changing their nature and nutrient

retention properties. This process may be of particular importance for

kaolinite/sesquioxide dominated soils as these minerals are particularly sensitive to

dehydroxylation (Ketterings et al., 2000).

Dehydroxylation of kaolinite, gibbsite and goethite by heating produces highly

reactive minerals. Dehydroxylation between 500oC and 700oC will destroy kaolinite

and form metakaolinite (Babuskhin et al., 1985). Gibbsite alters to complex

amorphous and crystalline mineral assemblages on heating at about 200oC (Rooksby,

1972) and goethite transforms to partly ordered hematite at about 300oC (Cornell and

Schwertmann, 1996). The existence and persistence of these reactive dehydroxylated

minerals in heated soils is unknown but they have markedly different properties from

the precursor minerals. They commonly exhibit greatly reduced structural order and

increased surface reactivity while retaining their original particle morphology

(Rooksby, 1972; Babuskhin et al., 1985; Cornell and Schwertmann, 1996). The

interactions of plant nutrients with the reactive minerals present in heated soils

clearly requires investigation and has not been evaluated by other workers. Most

previous work on the effects of fire on soils has focussed on plant nutrients in ash

and not on the presence and roles of heated soil minerals. The objective of this

glasshouse study was to evaluate the effects of heating soil containing kaolinite,

goethite and gibbsite on the availability to plants of native and added phosphate

which is known to be strongly sorbed by these minerals. Wild and managed fires

affect large areas of soil so that knowledge of the effects of heating on plant nutrient

availability is clearly of value.

4.2. Material and methods

4.2.1 Soil and Glasshouse experiment

The top soil (0-10 cm) of the Yalanbee, lateritic podzolic soil which is an Alfisol

(USDA, 2010) was collected from a virgin eucalypt forest site at Bakers Hill, 73 km

44

east of Perth, Western Australia at 31°46'34.21"S latitude and 116°28'31.81"E

longitude. These soils have developed from colluvium derived from lateritised

granite. The landscape consists of laterite-capped uplands and colluvium mantled

moderate slopes, the samples were taken from slope sites (McArthur, 1991). The

average annual rainfall is 552 mm with an average of 64 rain days each year for

Bakers Hill with most rain in the May–September period and highest monthly

rainfall during June and July. (Bureau of Meteorology, 2011). The same sandy,

highly P deficient soil had been used in a study of the P fertilizer value of chicken

litter ash by Yusiharni et al. (2007) so that its P-response characteristics and basal

nutrient requirements are known. The sampled soils are very gravely sandy earths

derived from laterite colluvium. The soil was passed through a <2 mm sieve for use

in the glasshouse experiment and for chemical analysis. Soil properties are listed in

Table 4.1. To prepare heated soils under conditions that resemble heating of soils in

a wildfire, subsamples of soil were heated under oxidising conditions in shallow

trays for one hour at 250 (BH250), 350 (BH350) and 500oC (BH500) which are

temperatures reached by top soils in bush fires (Raison, 1979). These particular

temperatures are above and below dehydroxylation maxima of the major

hydroxylated minerals in the soil (Grim, 1968; Babuskhin et al., 1985; Cornell and

Schwertmann, 1996) and the dehydroxylation reactions are largely completed within

1 h.

Phosphate (P) as monocalcium phosphate (MCP) (zero, 1.66, 3.33, 6.66 and 13.33

mg of P/kg) was mixed with 200 g of <2 mm original and the 3 preheated soil

samples in a plastic bag and placed in a non draining plastic pot. Basal fertilizer

excluding P was applied to all pots, the soil was then mixed thoroughly and

incubated at field capacity for 4 days before seeding with ryegrass. Twenty

pregerminated seeds of annual ryegrass (Lolium rigidum Gaud) were placed in the

pots at 1 cm depth and thinned to 10 plants per pot at the two-leaf-stage of growth.

This is a commonly used bioassay species for P fertilizer studies as the nutrient

requirement and glasshouse culture procedures for ryegrass are well established

(Snars et al., 2004). The treatments were replicated three times, the location of pots

on the glasshouse table was randomised every week and pots were maintained at

constant weight with deionised water. Plants were harvested 3 times at 4-week

45

intervals by cutting the tops at about 1 cm above the soil surface. Plant material was

dried at 60oC until constant weight occurred then ground and analysed.

a)

b)

5 10 15 20 25 30 35

2θ angle

BH

BH250

BH350

BH500

Figure 4.1. Conventional XRD patterns Cu Kα radiation, with inset for the clay

fraction (a) and synchrotron XRD (SXRD) patterns (b) of original Bakers Hills

(BH) soil and soil heated to 250oC, 350oC and 500oC. Q = quartz, K = kaolinite, Gi =

gibbsite and Go = goethite. The broad background scattering for the SXRD patterns

is due to the glass capillary containing the sample.

5 10 15 20 25 30 35 40 45

2θ Cu Kα

BH

BH250

BH350

BH500 K

Q

Q

Q, Gi, GoQ, K, Go Q Q

Gi Go

F

10 15 20 25

2θ Cu Kα

BH250

BH350

BH500

BHK

Gi Go K

5 10 15 20 25 30 35 40 45

2θ Cu Kα

BH

BH250

BH350

BH500 K

Q

Q

Q, Gi, GoQ, K, Go Q Q

Gi Go

F

10 15 20 25

2θ Cu Kα

BH250

BH350

BH500

BHK

Gi Go K

46

Table 4.1. Properties of original and heated Bakers Hill (BH) soil samples (n= 1).

4.2.2 Soil and plant analysis techniques

XRD analysis of unheated and heated soil samples was carried out using two

techniques. Conventional XRD analysis was conducted with a Philips PW3020

diffractometer and synchrotron XRD (SXRD) analysis was performed at the

Australian Synchrotron. Samples of soil for scanning electron microscopy and

energy dispersive X-ray spectrometry (EDS) using a JEOL 6400 instrument were

placed on metal stubs. Total carbon and nitrogen were determined on an Elementar

CNS (Vario Macro) analyzer.

Soil samples were analysed for pH and electrical conductivity (EC) (1:5 H2O) and

available P (Bic P) using 0.01M sodium bicarbonate extractant at pH 8.5 followed by

colorimetric determination of dissolved P (Colwell, 1963). To determine poorly

ordered minerals in heated soils, extractable forms of (iron) Fe, aluminium (Al) and

Properties Unheated BH Soil

BH 250oC

BH 350oC

BH 500oC

pH (1:5 H2O) 4.63 4.67 5.40 5.45 EC (1:5 H2O) μS/cm 39.9 119.1 91.8 77.1 Total P (mg/kg) 54 45 46 51 Bic P (mg/kg) 1.55 2.17 3.27 1.93 Total N (%) 0.080 0.078 0.072 0.023 Total C (%) 1.64 1.56 0.75 0.015 Total Ca (mg/kg) 471 Total Mg (mg/kg) 437 Total K (mg/kg) 498 Total Na (mg/kg) 53 Total S (mg/kg) 56 Total Fe (mg/kg) 9115 Total Ni (mg/kg) 7 Total Pb (mg/kg) 20 Total Cu (mg/kg) 2 Total Mn (mg/kg) 55 Total Zn (mg/kg) 7 Total Al (mg/kg) 39,840 Total Ba (mg/kg) 40 Clay Percentage (%) 4 Silt Percentage (%) 6 Sand Percentage (%) 90

47

silicon (Si) in the soil were dissolved in dithionite-citrate-bicarbonate (DCB),

ammonium oxalate (Ox) and sodium pyrophosphate (SP) solutions (Rayment and

Higginson 1992) and elements were determined by inductively coupled plasma

optical emission spectroscopy (ICP-OES) (Perkin-Elmer, Norwalk, CT, USA). Soils

were analysed before planting and after the last harvest.

The harvested plant tops were digested in perchloric acid and analysed for P,

calcium (Ca), magnesium (Mg), potassium (K), sodium (Na), sulphur (S) and trace

elements by inductively coupled plasma optical emission spectrometry (ICP-OES)

(Perkin-Elmer, Norwalk, CT, USA) (Rayment and Higginson, 1992).

Replicate (3) plant data were averaged and statistically analysed including fitting of

internal efficiency curves. An analysis of variance followed by Tukey’s Honestly

Significant Differences test were performed to identify the effects of different levels

of P added and heating treatments using SAS (SAS Institute, 1999). Differences for

levels of P added in values of pH, EC and available P after the last harvest were

assessed using one way-ANOVA. All results are presented as significant differences

at p < 0.05.

4.3. Results and discussions

4.3.1 XRD and SEM

Conventional and synchrotron XRD patterns of heated and unheated Yalanbee soil

show the effect of heating on soil mineralogical properties (Fig 4.1). The SXRD

patterns had much better sensitivity and peak to background discrimination than

conventional XRD. Both techniques showed that the main crystalline compounds of

unheated soils are quartz and kaolinite together with minor amounts of gibbsite and

goethite, which were more clearly revealed by conventional XRD of the clay

fraction. After heating at 250oC gibbsite reflections had disappeared (Rooksby,

1972; Kloprogge et al., 2002). Wang et al. (2006) observed similar results and

considered that dehydroxylation caused gibbsite to alter into an amorphous phase

and boehmite. After heating at 350oC, gibbsite and goethite reflections had

substantially disappeared as amorphous alumina (no reflections) and hematite

respectively had formed (Zanelli et al., 2006). Nornberg et al. (2009) found that

48

when soil temperature reaches about 320oC, goethite transforms to hematite as

observed in this work. After heating at 500oC the kaolinite peak intensity was greatly

reduced due to the formation of metakaolinite (Grim, 1968). Quartz was unaltered

for all the heating temperatures as would be anticipated for a high temperature

igneous mineral (Ketterings et al., 2000).

Phyllosilicate and iron oxides minerals are affected by the high temperatures

generated at the soil surface during heating (Sertsu and Sanchez, 1978; Ulery et al.,

1996). Fires altered topsoil in 1 to 2% of the land area studied by Ulery et al. (1996)

and destroyed kaolinite. Ketterings et al. (2000) have reported that mineralogical

analyses of clay fractions from field-burned soil in Indonesia showed that the

amount of kaolinite was significantly reduced, gibbsite was partially decomposed

(<300oC) while goethite was transformed to ultra-fine maghemite as a result of soil

heating. The formation of highly reactive compounds (amorphous alumina, hematite

and metakaolinite) by heating might be expected to increase sorption of P from soil

solution (Kwari and Batey, 1991) and this would reduce the agronomic effectiveness

of both native and added P. Ketterings et al. (2002) described an increase of

maximum P-sorption capacity of the top soil layer (0-5 cm) exposed to fire (450+ oC)

from 2200 mg P kg-1 to 2800 mg P kg-1 but this may not occur for soil types with a

different mineralogy. Furthermore, increased P retention may be a reflection of the

presence of calcite in ash within soil and the precipitation/sorption of P due to calcite

and high pH.

SEM micrographs and EDS spectra of unheated Bakers Hills soil and soil heated at

250, 350 and 500oC show the diverse particle sizes, shapes and compositions present

in this material (Fig. 4.2). Apart from quartz crystals, most grains seen in the

micrographs are kaolin-rich aggregates and consequently consist mostly of

aluminium and silicon, and contain only a little P, Fe and other elements. For

example, the grain shown for BH250 (indicated in Fig. 2) contains much Si and Al

and is probably kaolinite. The Al and Si rich grains remain after heating at 350oC

and 500oC but are now metakaolinite rather than kaolinite. The indicated grain for

BH350 contains P and S in addition to Al and Si, which may indicate a mixture of

kaolinite and plant ash.

49

BH BH250

BH analysed grain

1 3 5 7 9 Energy (keV)

Al Si

K Ti Fe

BH250 analysed grain

1 3 5 7 9 Energy (keV)

Al Si

K Ti Fe

BH350 BH500

BH350 analysed grain

1 3 5 7 9 Energy (keV)

Al

Si

K Fe

P

BH500 analysed grain

1 3 5 7 9 Energy (keV)

Al Si

K Fe Ti

Figure 4.2. Scanning electron micrographs (SEM) and X-ray spectra of the indicated

particles for original Bakers Hills (BH) soil and soil heated to 250oC, 350oC and

500oC.

50

0

1

2

3

4

5

6

7

0 1.67 3.33 6.67 13.33

Soi

l pH

Rate of P Applied (mg/kg)

Soil pH (a)

BH Unburnt

BH 250

BH 350

BH 500

b b ab ab a a b bc bc c

a ab bc bc c a b b b b

0

100

200

300

400

500

600

700

800

900

0 1.67 3.33 6.67 13.33

Soil

EC (μ

S/cm

)

Rate of P Applied (mg/kg)

Soil EC (b)

BH Unburnt

BH 250

BH 350

BH 500

a a

ab

bc

c

a

ab

ab

ab b

a a a

b b

a ab

bc

c

c

0

2

4

6

8

10

12

14

0.00 1.67 3.33 6.67 13.33

Soi

l Bic

P (m

g/kg

)

Rate of P Applied (mg/kg)

Soil Bic P (c)

BH Unburnt

BH 250

BH 350

BH 500

a

c c c

b

a

ab

ab b

b

a

b

bc c c

a

b

c c c

Figure 4.3. Mean (n = 3) soil pH (a), EC (b) and plant available soil phosphorus (Bic

P) (c) after the last harvest versus rate of P applied. Means having different letters

are significantly different at P ≤ 0.05.

51

4.3.2 Heating impacts on chemical properties

Other characteristics of the unheated and heated soil used in the study are provided

in Table 4.1. The very low concentrations of plant nutrient elements are evident and

are typical of unfertilised lateritic soils in this region (Robson and Gilkes, 1981).

Total soil carbon and N decreased with the increasing heating temperature due to

combustion of soil organic matter and loss of gaseous oxides of C and N at 250oC

and higher temperatures. These results agree with findings by Terefe et al. (2008)

where little change in total carbon occurred when soil was heated to 100oC but

amounts significantly decreased for soil heated at 200oC, 300oC and eventually all

carbon was lost from soil heated at 500oC. Soil nitrogen decreased simultaneously

with carbon. Gonzàlez-Pérez et al. (2004) stated that C/N ratio in soil after burning

is usually lower than in the original soil.

Table 4.2. p-values for significant difference with level of P added (0, 1.66, 3.33,

6.66, and 13.33 mg P/kg) and heating temperature.

The initial pH of unheated soil was 4.63 and pH increased to 4.67, 5.40 and 5.45 for

BH250, BH350 and BH500 respectively (Table 4.1), which may indicate that

oxidised organic matter released alkali cations (Terefe et al., 2008) and heating also

caused the loss of organic acids (Certini, 2005; Fernandez et al., 1997). Electrical

conductivity (EC) of the heated soil increased substantially for BH250 then

decreased for BH350 and BH500. Combustion of organic compounds in soils and

litter produces soluble salts (e.g. CaCO3, CaSO4, KCl) and at 350oC and 500oC these

may have reacted with soil constituents to produce insoluble compounds (e.g.

calcium silicates) thereby reducing EC (Terefe et al., 2008). Ùbeda et al. (2009) also

observed decreasing EC of soil solution for soil heated above 450oC due to the high

amount of calcite (CaCO3) that formed. Earlier studies also documented the decrease

of EC in soil solution after heating above 400oC to 500oC (Iglesias et al., 1997;

Heating Temperature pH EC Bic-P Unburnt 0.0019 0.0010 0.0001

250 0.0389 0.0012 0.0001 350 0.0001 0.0338 0.0201 500 0.0062 0.0002 0.0002

52

Badía and Martí, 2003). It is unlikely that the quite small changes in pH and EC

observed in the present research would have a major effect on P retention by the

heated soils.

Total P concentration of the various soil samples was almost constant (49± 4.2 mg/g

P) whereas plant available P (Bic P) increased substantially on heating the soil to

350oC and then decreased on heating at 500oC. This increase reflects inter alia the

conversion of organic P to soluble inorganic forms on combustion as indicated in Fig

2. Changes in soil pH and additionally the dehydroxylation and recrystallisation of

minerals may liberate adsorbed P (Ketterings et al., 2002). At 500oC chemical

reactions with soil constituents may cause some P to become unavailable (insoluble).

The impacts of P additions to heated soils on soil pH, EC and Bic-P measured at the

end of the plant growth experiment are shown in Fig. 4.3 and a list of p-values for

associated correlations is shown in Table 4.2. There was no systematic effect of level

of P added on the pH of heated and unheated soil. Soil heated at 250, 350 and 500oC

showed a systematic increase in pH. The values of soil pH with increasing heating

temperature are statistically different p≤0.05 (Table 4.2). This increment in pH was

not affected by P fertilisation. Earlier studies also observed the increase of soil pH

for heated soil (Kutiel and Naveh, 1987; Iglesias et al., 1997). These increases could

be associated with the incorporation of ash into soil from the combustion of soil

organic constituents. According to Ulery et al. (1993) the higher pH values of burned

soil are possibly produced by the formation of oxides, hydroxides and potassium and

sodium carbonates. The EC of unfertilised soil at the end of pot experiment was not

affected by heating (Fig 4.3) in contrast to the substantial differences in EC existing

at the beginning of pot experiment. EC of soil for the highest P level differed

significantly for unheated, 250, 350 and 500oC heated soil (p values of 0.0010,

0.0012, 0.00338, and 0.0002 respectively). The EC values decreased with increasing

rate of P application for all heating temperatures. The decrease is due to increasing

uptake of salts by the progressively larger plants produced by higher P applications

(Fig. 4.5). Yusiharni et al. (2007) also observed a decrease in EC with increasing

MCP application rate and plant weight in a similar glasshouse experiment on the

same soil. Available P after the last harvest increased with level of P applied but did

53

not differ systematically between unheated and heated soils. Most added P was

recovered by the bicarbonate extractant. Ketterings et al. (2002) found that for low

intensity fires, phosphorus is more available because of mineralisation of litter P and

because soil temperatures are generally too low to affect soil minerals. However in

the present work, soil P sorption increased for medium to high fire intensity and this

resulted in P being less available for the growth of plants. The present results do not

show a systematic effect of heating on the availability of soil P as measured by

bicarbonate extraction.

Analyses of the plant tops (Appendix 7,8 and 9) indicate that the concentrations of

plant nutrient elements (Na, Mg, K, Ca, Si, P, S, Cl, Mn, Cu, and Zn) were within

the normal range for ryegrass (Reuter and Robinson 1997) so that it is likely that no

element toxicity or deficiency other than P occurred during the experiment.

4.3.3 Forms of Fe, Al and Si in the heated soils

Crystalline, amorphous and organic compounds of Fe, Al and Si in soil were

measured by selective extraction methods. Dithionite-citrate-bicarbonate (DCB)

extractant removes clay size oxides of Fe and associated Al/Si including crystalline

and poorly crystalline, amorphous and organic forms. Ammonium oxalate (Ox)

dissolves poorly crystalline and organically bound Fe, Al and Si. sodium

pyrophosphate (SP) extracts organically bound Fe, Al and possibly Si (McKeague et

al., 1971).

54

0

500

1000

1500

2000

2500

3000

3500

BH BH 250 BH 350 BH 500

Ext

ract

able

Al (

mg/

kg)

Al (Ox) Al (SP) Al (DCB)

a a

b b

b

a

c c a

b a a

(a)

0

500

1000

1500

2000

2500

3000

3500

4000

BH BH 250 BH 350 BH 500

Ext

ract

able

Fe

(mg/

kg)

Fe (Ox) Fe (SP) Fe (DCB)

b b b

a

c c b a a a b a

(b)

-100

0

100

200

300

400

500

600

700

800

BH BH 250 BH 350 BH 500

Ext

ract

able

Si (

mg/

kg)

Si (Ox) Si (SP) Si (DCB)

b b b

a c c b

a

a

b b b

(c)

Figure 4. 4. Mean (n = 3) data for the effect of heating on extractable soil Al, Fe and

Si for BH, BH250, BH350 and BH500 after the last harvest for zero P applied. Ox =

ammonium oxalate, SP = sodium pyrophosphate, and DCB = dithionite-citrate-

bicarbonate.

55

0

100

200

300

400

500

600

0 5 10 15

Yield

(mg/k

g)

Rate of P Applied (mg/kg)

Harvest 1 (a)

BH Unburnt BH 250 BH 350 BH 500

0

100

200

300

400

500

600

0 5 10 15

Yield

(mg/

kg)

Rate of P Applied (mg/kg)

Harvest 2 (b)

BH Unburnt BH 250 BH 350 BH 500

0

100

200

300

400

500

600

0 5 10 15

Yield

(mg/k

g)

Rate of P Applied (mg/kg)

Harvest 3 (c)

BH Unburnt BH 250 BH 350 BH 500

Figure 4.5. Relationships between plant yield and rate of P applied for harvests 1, 2,

and 3 for ryegrass grown on variously heated Bakers Hill soil (n = 3). Representative

standard error values are shown in Fig. 4.5b.

56

The mean values of the three forms of extractable Al, Fe and Si for unheated and

heated soil after the last harvest are shown in Fig 4.4. For each heating temperature,

there was no systematic difference in the amounts of Al, Fe and Si extracted for the

several levels of P added so only data for zero P added are shown in Fig 4.4.

Extractable Al values for the three extractants mostly increased with heating

temperature. Oxalate Al was the most abundant form of Al and may be present in

amorphous aluminium compounds and in particular amorphous alumina and

metakaolinite. Thus heating at 500oC created substantial additional amounts (2.8

g/kg) of oxalate soluble Al that presumably mostly originates in amorphous

metakaolinite and alumina that form at this temperature (Ulery et al., 1996). The

amount of iron extracted with dithionite-citrate-bicarbonate (DCB) is considerably

higher than for the other extractants and this come mostly from crystalline iron

oxides. The small increase in DCB-Fe for BH350 presumably reflects the increased

solubility of disordered hematite created from goethite at this temperature. For all

three extractants relatively little Si dissolved but amounts increased substantially for

soil heated at 500oC, presumably reflecting the greater solubility of Si in

metakaolinite. Quartz, the dominant Si mineral in these soils is not soluble in these

solutions and is not affected by heating at these temperatures (Ketterings et al.,

2000).

4.3.4 Plant dry matter

For all three harvests dry matter yield generally decreased for plants grown on

unfertilised heated soil reflecting changes in the forms of native organic and

inorganic P discussed above. For harvest I, there was a large negative response of

plant dry matter yield to soil heating for each rate of fertiliser P addition (Fig 4.5).

The effectiveness of fertilizer (yield increment/added P) decreased with heating

temperature (i.e response curves diverge with increasing rate of P application). For

the second and third harvests, the decrease in yield due to heating the soil persisted

but there was no systematic change in fertilizer effectiveness for the three lowest

heating temperatures (i.e response curves are approximately coincident). The

response curve for BH500 soil is systematically below those for lower heating

temperatures and has a lower initial slope value for all three harvests. This result

may indicate the major effect of metakaolinite in increasing P-retention as this

57

compound only formed at 500oC. Other workers have suggested that P deficient soils

with a large sorption capacity as a result of heating above 420oC can retain large

quantities of added P (Ketterings et al., 2002), which may be unavailable for the

growth of plants.

4.3.5 Plant analysis

Heating the soil reduced P concentration in plants for the non P fertilized heated

soils indicating that heating soil reduced the availability of native P. The P

concentration in plants increased with increasing rate of P application and heating

soil decreased the P concentration in plants for all three harvests, with the response

curve for BH500 being substantially lowered relative to the other heated soils for all

three harvests (Fig. 4.6). Kwari and Batey (1991) found that heating soil

significantly reduced the P concentration in maize for fertilized soils but there was

no significant difference in plant P concentration for unfertilized soils. The

concentration of P in ryegrass for all the harvests ranged from 0.03% to 0.26%. The

lower concentrations are indicative of P deficiency based on criteria of Reuter and

Robinson (1997), however, the cultivar of ryegrass (Lolium rigidum Gaud) used in

this experiment has a particularly low demand for P and produces substantial plant

growth at nominally deficient concentrations of P (Snars et al., 2004; Yusiharni et

al., 2007).

Plots of internal efficiency of P utilisation by plants (i.e plant dry matter yield v. the

P content of plants) for each harvest (Fig. 4.7) showed a single internal efficiency

curve for all heated soils. This result indicates that for each harvest the differences in

yield were primarily due to differences in the P content of the plants and not to the

supply of other nutrients or differences in other soil conditions (Palmer and Gilkes,

1983). These plant data (Fig. 4.6 and Fig. 4.7) clearly indicate that heating the soil

greatly decreased the amount of plant available P for both unfertilized and fertilized

soil.

58

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

0 5 10 15

P C

once

ntra

tion

(%)

P Applied (mg/kg)

Harvest 1 (a)

BH Unburnt BH 250 BH 350 BH 500

-0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

0 5 10 15

P C

once

ntra

tion

(%)

P Applied (mg/kg)

Harvest 2 (b)

BH Unburnt BH 250 BH 350 BH 500

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

0 5 10 15

P C

once

ntra

tion

(%)

P Applied (mg/kg)

Harvest 3 (c)

BH Unburnt BH 250 BH 350 BH 500

Figure 4.6. Relationships between plant phosphorus concentration and rate of P

applied for harvests 1, 2, and 3 for ryegrass grown on variously heated Bakers Hill

soil (n = 3). Representative standard error values are shown in Fig. 6b.

59

y = 0.078Ln(x) + 0.31 R2 = 0.98

0

0.1

0.2

0.3

0.4

0.5

0.6

0.0 0.2 0.4 0.6 0.8 1.0

Yiel

d (g

/pot

)

Plant P Content (mg/pot)

Harvest 1 (a)

BH Unburnt

BH 250

BH 350

BH 500

y = 0.16Ln(x) + 0.58 R2 = 0.97

0

0.1

0.2

0.3

0.4

0.5

0.6

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Yiel

d (g

/pot

)

Plant P Content (mg/pot)

Harvest 2 (b)

BH Unburnt

BH 250

BH 350

BH 500

y = 0.12Ln(x) + 0.50 R2 = 0.98

0

0.1

0.2

0.3

0.4

0.5

0.6

0.00 0.20 0.40 0.60 0.80 1.00

Yiel

d (g

/pot

)

Plant P Content (mg/pot)

Harvest 3 (c)

BH Unburnt

BH 250

BH 350

BH 500

Figure 4.7. Internal efficiency curves (yield versus plant P content) for three harvests

of ryegrass grown on variously heated Bakers Hill soil (n = 3).

60

4.4. Conclusions

Heating the soil increased Bic-P and had little effect on the retention of added P as

indicated by bicarbonate extraction. However heating soil decreased the agronomic

effectiveness of both native and applied P with smaller reductions occurring for

lower heating temperatures. We ascribe this change to the presence of reactive

dehydroxylated minerals in the heated soils that adsorb P. Heating at 500oC greatly

increased reactive Si and Al due to the formation of reactive metakaolinite and

amorphous alumina, which are highly reactive compounds and are presumably

responsible for the reduced effectiveness of the added P fertilizers.

This research has indicated that soil heating during a bushfire may have significant

impacts on soil mineralogical and chemical properties. The availability to plants of

native and added P may be substantially affected. These effects of fire on soil

characteristics will vary with the duration and intensity of fire, fuel and soil type and

these topics deserve further investigation. The lateritic podzolic soil used in this

research has a fine fraction dominated by three readily dehydroxylated minerals

(kaolinite, gibbsite and goethite) and consequently is particularly sensitive to

heating. Phosphate sorption by soils that are predominantly composed of more

thermally stable minerals (e.g quartz, illite , hematite) is unlikely to be so strongly

affected by heating. Similarly for soils where much P resides in organic matter and

litter, the impact of fire on P availability may be more strongly affected by the forms

of P in ash than the transformations of minerals.

61

Chapter 5

5.0 Changes in the mineralogy and chemistry of a lateritic soil due to a bushfire

at Wundowie, Darling Range, Western Australia

5.1. Introduction

Australia has a history of severe bush fires. Fire was used as a powerful land-use tool

by indigenous Australians to manage grasslands, forest and fauna (Gill and Moore,

1990) and this practice persists under contemporary management regimes. The

nature of the Australian environment, which is generally hot, dry and prone to

drought and with a dominance of volatile natural vegetation, makes Australian

landscapes particularly vulnerable to wildfires.

Heating of soil by burning plant materials during bushfires may have significant

impacts on soil properties and consequently the growth of plants (Raison, 1979; Wan

et al., 2001). The effects of burning on soil vary with the duration and intensity of

fire and soil type. The amount of fuel present at the time of fire affects the

temperature reached by heated soil, which influences the degree of alteration of soil

properties (Raison, 1979). Durán et al. (2008) found that P availability to plants

declined after fire although it tended to recover over time. Some studies have

reported an increase in soil fertility and nutrient concentrations in plants grown

shortly after a single fire (Christensen, 1994). Ferran et al. (2005) observed an

increase in total P in the topsoil (0-5 cm) after a wildfire and the available P in

surface horizons can rise immediately after fire (Debano and Klopatek, 1988). These

effects are at least partly due to the accelerated recycling of nutrients present in the

ash of combusted plant materials and soil organic matter (Shakesby and Doerr, 2006)

but the contribution of heated soil minerals is not known.

Some soil minerals are affected by heating in natural and managed fires.

Temperatures during forest fire typically ranging from 100 to 300oC and may be in

excess of 1500oC (Neary et al., 1999). Higher temperatures occur during high

intensity forest fire, especially under concentrated fuel sources such as logs, slash

piles or in a tree stump and these temperatures could decompose soil minerals (Ulery

et al., 1996). Heating of soil by both modest (250-500oC) fires and extreme fires

(>500oC) will cause some hydroxylated soil minerals to dehydroxylate (Ketterings et

62

al., 2000). Laboratory studies have shown that kaolinite decomposes at temperatures

between 450oC and 700oC losing lattice hydroxyl and forming metakaolinite

(Babuskhin et al., 1985; Frost, 2003). Ulery et al. (1996) found that kaolinite in a

topsoil was destroyed by fire. Gibbsite commonly alters to a mixture of an

amorphous phase and boehmite on heating at about 200oC (Wang et al., 2006) and

goethite is transformed to a disordered mineral known as hydro-hematite at about

300oC (Watari et al., 1979). Eggleton and Taylor (2008) studied the impact of fire on

the Weipa Bauxite, northern Australia and they found that gibbsite dehydrated to

boehmite or alumina and Fe-oxyhydroxides were converted to maghemite. Several

other studies also reported that soil heating may also alter goethite into maghemite

(Ketterings et al., 2000; Terefe et al., 2008; Nørnberg et al., 2009). The possibility

that these various dehydroxylated minerals may form and persist in soils heated in

bushfires in Southwest Australia is the subject of this investigation.

5.2. Material and Methods

5.2.1 Soil samples burnt in forest fires

Samples were obtained from soils that had been heated under burnt logs at

Wundowie in the Eastern Darling Range, Western Australia. Samples were removed

as a 1cm thick layer from under burnt eucalyptus (Eucalyptus marginata) and grass

tree (Xanthorrhoea pressii) logs from replicate sites on a soil formed from lateritic

colluvium (Fig. 5.1a). The soil is known locally as the Yalanbee gravel and is

classified as a lateritic podzolic soil, which is an Alfisol (USDA, 2010). Sampling

took place shortly after the fire in early March 2009 (Fig. 5.1b). Samples were taken

from a 1 cm deep and 10 cm wide strip below a burnt log and from strips 10-20 cm

and 20-30 cm away from the log (Table 5.1). Soil samples were also taken from 10-

20 cm depth to represent unheated soil. The soils are very gravelly and were sieved

to obtain the <2 mm fraction. Ash and charcoal derived from the burnt eucalyptus

and grass tree trunks were also collected. Eckmeier et al. (2010) defined

macrocharcoals as carbonized wood fragments that are visible by naked eye with >2

mm in diameter. To remove macroscopic charcoal prior to analysis of the ash, it was

sieved through <2 mm mesh. A flow chart of analysis performed during the study is

given in Fig.5.2.

63

a)

b)

Figure 5.1. View of the study site one day after the Wundowie bushfire showing

scorched eucalyptus and grass trees (a) and the position of a burnt eucalyptus log

showing ash and charcoal (b).

64

Table 5.1. The nomenclature for the samples of heated soil from under burnt

eucalyptus (Eucalyptus marginata) and grass tree (Xanthorrhoea preissii) logs at the

Wundowie bush fire site.

Figure 5.2. Flow chart of analyses performed during the study.

Key Explanation GT 0-10 cm Heated 0-1cm soil, 0-10cm from burnt grass tree log GT 10-20 cm Heated 0-1cm soil, 10-20cm from burnt grass tree log GT 20-30 cm Heated 0-1cm soil, 20-30cm from burnt grass tree log EU Unburnt Unburnt soil 0-10cm depth EU 0-10 cm Heated 0-1cm soil, 0-10cm from burnt eucalyptus log EU 10-20 cm Heated 0-1cm soil, 10-20cm from burnt eucalyptus log EU 20-30 cm Heated 0-1cm soil, 20-30cm from burnt eucalyptus log

Samples collected from field Under burnt logs

Soil <2 mm

Gravel >2 mm

Charcoal >2 mm

Ash <2 mm

Chemical Analysis: pH EC Bic P and K C and N Extractable Fe, Al and Si

Mineralogy Analysis: XRD and SXRD

SEM Analysis

Grind

65

5.2.2 Analytical techniques

The soil samples were analyzed for bicarbonate extractable P (Bic P) (Colwell 1963)

and for pH and electrical conductivity (EC) in a 1:5 deionised water extract using

methods from Rayment and Higginson (1992). The amounts of carbon and nitrogen

were determined on an Elementar CNS (Vario Macro) analyzer.

Whole soil samples, the fine earth fraction (<2 mm) and ground gravels were

analyzed as random powders by Conventional XRD and Synchrotron XRD (SXRD).

Conventional XRD was conducted on a Philips PW3020 diffractometer with a

diffracted beam monochromator (CuKα, 50kV, 20mA). Powder samples were

scanned from 4 to 70o 2θ, using a step size of 0.02o 2θ and a scan speed of 0.04o 2θ

sec-1. SXRD analysis was performed at the Australian Synchrotron where powder

samples were mounted into glass capillaries with a 1.0Å wavelength set for this

analysis in order to provide a high peak/background and adequate resolution for

identifying minor constituents.

Major and minor elements were analyzed using a PE ELAN 600 inductively coupled

plasma with optical emission spectroscopy (ICP-OES) instrument (Perkin-Elmer,

Norwalk, CT, USA) after concentrated perchloric acid (HClO4) digestion. 0.2 g

samples were pretreated drop wise with concentrated nitric acid until the reaction

became less vigorous because of the high amounts of carbonate that were present in

some samples.

Extractable forms of iron (Fe), aluminum (Al) and silica (Si) in <2 mm fraction of

the soils and ground gravels were dissolved in solutions of dithionite-citrate-

bicarbonate (DCB) (Mehra and Jackson 1960), ammonium oxalate and sodium

pyrophosphate (Rayment and Higginson 1992) and concentrations were determined

by inductively coupled plasma optical emission spectroscopy (ICP-OES) (Perkin-

Elmer, Norwalk, CT, USA). For DCB extraction, one gram of sample was placed

into a 50 mL centrifuge tube to which 45 mL of solution (0.3M Na-citrate + 0.1 M

Na bicarbonate) was added. The centrifuge tube then placed in a water bath at 70oC

and one gram of sodium-dithionite was added. The mixture was then stirred

constantly for one minute and occasionally during the next 15 minutes. The tubes

were then centrifuged at 2000 rpm for 15 minutes. The extraction was repeated twice

66

then the combined volume was made up to 250 mL with DI water, and analyzed by

ICP-OES. Ammonium oxalate extraction used one gram of sample which was placed

into a 25 mL centrifuge tube, 10 mL of 0.2 M ammonium oxalate solution at pH 3.0

was added and the tube was shaken for 4 hours in a dark room. Five drops of 0.4%

Superfloc were added to the mixture, which was then centrifuged. Clear supernatant

was then analyzed by ICP-OES. For pyrophosphate extraction, one gram of sample

was weighed into a 50 mL centrifuge tube, 30 mL of 0.1 M Na-pyrophosphate

solution was added and the tube was shaken overnight. The mixture was centrifuged

and the clear supernatant was analyzed by ICP-OES.

The composition and morphology of the plant ash was examined by scanning

electron microscopy (SEM) and energy dispersive X-ray spectrometry (EDS) using a

JEOL 6400 instrument. Samples for SEM analysis were placed on metal stubs and

carbon coated before analysis.

5.3. Results and Discussion

5.3.1 Carbon and nitrogen

Mean concentrations of carbon and nitrogen in soil, gravel, ash and charcoal for

unburnt and burnt soil under grass tree (GT) and eucalyptus (EU) logs are presented

in Fig. 5.3. Soil contains much higher amounts of carbon and nitrogen than gravel

and there was no significant difference in values for these elements between the two

types of logs. Burnt soil 0-10 cm and one centimeter deep from (i.e under) burnt logs

contains much less C and N than soil 10-20 cm, 20-30 cm from the burnt logs. The

same trend occurred for C and N in gravel. These trends may be due to the lower

temperatures and incomplete combustion of soil organic matter experienced by soil

at >10 cm from the burning log. The large concentrations of C and N in these 0-1 cm

deep soil samples is partly due to incorporation of charcoal into the soil surface.

Atanassova et al. (2009) also found increased topsoil organic carbon after fire in

Lyulin Mountain, Bulgaria. This increase was associated with incorporation of burnt

plant residues into soil (Johnson and Curtis, 2001). González-Pérez et al. (2004)

considered that the amount of soil organic matter retained in soil after heating

depends on the type of fire, slope and the intensity of the fire. Low intensity fires

had no effect on soil organic carbon (Terefe et al., 2008), however organic C was

67

totally eliminated when soil was heated at 400oC (Fernández, 1997). The present

results are consistent with these observations with less C and N in the soil directly

under the burnt logs. During a simulated fire at 250oC, Kutiel and Shaviv (1992)

reported 15 to 30% loss of total soil nitrogen and similar results have been reported

by Kutiel and Naveh (1987) and Kutiel and Shaviv (1989). Quintana et al. (2007)

demonstrated that total nitrogen was reduced by volatilization when soil was heated

to 400oC and these losses reached around 64% when temperature reached 500oC.

Carbon concentrations for charcoal are similar for grass tree and eucalyptus samples.

The ash from both species contains a similar concentration of C. Charcoal from both

species contains appreciable amounts of N but ash contain little N. Plant ash

generally contains little nitrogen (Khanna et al., 1994) especially when the

combustion of the fuel is nearly complete (Raison et al., 1985).

5.3.2 Element concentrations, pH, EC, available P and K, and extractable Fe, Al

and Si in soil samples

Table 5.2 shows the total element concentrations together with pH and EC values of

the soil samples. The element composition of soil post fire depends on plant species,

age of plant and also on which part of the plant was burnt (foliage, bark, wood, etc)

(Iglesias, 1997). During the study we found that concentrations of Ca, K, Mg, Mn, P

and soluble salts (EC) increased in heated soil. Values of pH also increased

substantially. Similar results were observed by Quintana et al. (2007) for the heated

soil organic horizon from a Spanish juniper (Juniperus thurifera L.) woodland.

These increases are at least partly due to addition of ash to the soil. Earlier studies

also pointed out that the increase was possibly due to the incorporation of ash and

organic remains from burnt plant materials (Kutiel and Naveh, 1987; Ulery et al.,

1996). The increased pH of the soil could also be associated with the solubility of

calcite in ash, which may persist for more than three years after the fire (Ulery and

Graham, 1993). Furthermore, Ulery et al. (1993) noticed that the elevated pH at the

soil surface immediately after the fire was the result of the formation of various

oxides, hydroxides and potassium and sodium carbonates.

68

Analyses of GT and EU ash shown in Table 5.2 demonstrate that several elements

are abundant in ash. There were much smaller increases in the concentrations of Ca,

K, Mg, Mn and P for gravel, presumably because little ash is attached to gravel

particles. The pH and EC of both soil and gravel was increased by addition of ash.

These results may also indicate greater oxidation of organic matter present in soil

under burnt logs and this has caused the release of additional cations (Terefe et al.,

2008). Calcium, magnesium, potassium and silicon are dominant cations in plant ash

(Etiegni and Campbel, 1991; Ulery et al., 1993; Khanna et al., 1994) and the P and

Mn in wood are also concentrated in ash (Badĭa and Marti, 2003). The concentration

of Fe in ash is relatively much lower than in unburnt soil so there is no effect of ash

addition to soil on the Fe concentration in the burnt soil.

Table 5.2. Mean values of element concentrations (mg/kg), pH and EC for the

unheated and heated soil (-2 mm), gravel, ash and charcoal after the bushfire.

Sample Key Element Concentrations (mg/kg) pH EC (μS/cm)

Ca Fe K Mg Mn P (1:5) (1:5) Soil Unburnt 1564 10908 553 243 111 177 7.0 35

GT 0-10cm 48400 9017 2571 6299 556 721 9.6 525 GT 10-20cm 23669 17595 120 3447 732 696 8.5 851 GT 20-30cm 29541 12190 318 4538 670 797 8.3 915 EU 0-10cm 18472 9097 770 2392 576 449 9.0 600 EU 10-20cm 31747 12107 1108 3678 829 776 8.5 700 EU 20-30cm 32982 14094 1867 3806 839 648 7.9 591

Gravel Unburnt 159 37035 320 101 21 21 6.0 44 GT 0-10cm 477 65550 279 83 31 59 8.1 100 GT 10-20cm 507 96287 349 130 116 60 7.5 132 GT 20-30cm 393 79825 348 93 24 29 7.3 94 EU 0-10cm 643 55682 414 116 31 30 7.9 140 EU 10-20cm 668 47678 531 104 34 52 7.5 108 EU 20-30cm 769 72159 521 155 41 76 7.3 86

Ash and charcoal GT Ash 95038 447 2662 11139 195 671 10.6 1529 GT Charcoal 5346 21 162 1154 6 38 8.3 358 EU Ash 67750 1015 4763 16915 1494 508 12.2 3865 EU Charcoal 10112 968 901 2160 260 135 8.4 243

69

a)

0

5

10

15

20

25

0-10cm 10-20cm 20-30cm

C %

C Soil, GT C Soil, Eu

b)

0

2

0-10cm 10-20cm 20-30cm

C %

C Gravel, GT C Gravel, Eu

c)

0

1

0-10cm 10-20cm 20-30cm

N %

N Soil, GT N Soil, Eu

d)

0.0

0.1

0-10cm 10-20cm 20-30cm

N %

N Gravel, GT N Gravel, Eu

e)

0

20

40

60

80

100

120

GT Ash GT Charcoal EU Ash EU Charcoal

C %

f)

0

1

GT Ash GT Charcoal EU Ash EU Charcoal

N %

Figure 5.3. Mean values of carbon and nitrogen concentrations (%) for 0-1cm soil,

gravel, ash and charcoal under and adjacent to burnt grass tree (GT) and eucalyptus

(EU) logs.

70

a)

0

5

10

15

20

25

30

Unburnt GT 0-10cm GT 10-20cm GT 20-30cm EU 0-10cm EU 10-20cm EU 20-30cm

mg

/kg

Soil Available P and K (mg/kg)

Bic P

Bic K

b)

0

2

4

6

8

10

12

Unburnt GT 0-10cm GT 10-20cm GT 20-30cm EU 0-10cm EU 10-20cm EU 20-30cm

mg

/kg

Gravel Available P and K (mg/kg)

Bic P

Bic K

Figure 5.4. Mean values of bicarbonate soluble P and K for soil (a) and gravel (b)

under and adjacent to burnt grass tree (GT) and eucalyptus (EU) logs.

The concentrations of available (bicarbonate soluble) P and K in soil and gravel after

the bushfire are shown in Fig. 5.4. Available P and K were increased substantially by

burning (Giardina et al., 2000) for both soil and gravel. These increases are a result

of the combustion of organic P in the soil (Galang et al., 2010) and especially the

addition of substantial amount of P and K to the soil in ash (Kutiel and Shaviv,

71

1989), with a possible contribution from the transformation of soil minerals (Table

5.2).

The amounts of Fe, Al and Si in soils removed by the three extractants (Table 5.3)

are commonly taken to indicate particular forms of these elements in soil (Zanelli et

al., 2007). Sodium pyrophosphate (p) is assumed to extract organically bound

elements and also hydroxyl-like Al in the interlayers of clay minerals (Kleber et al.,

2004). Heated soil contained more Al-p than unburnt soil. Amounts of Fe-p and Si-p

showed no systematic effect of heating the soil. There were no systematic

differences due to heating on levels of Al, Fe, and Si-p in gravels. The considerable

amounts of Al-p and Si-p in GT and EU ash may be from heated kaolinite

incorporated into the ash and are also possibly due to oxidation of organic-mineral

complexes during combustion.

Sodium oxalate (o) is assumed to dissolve poorly ordered and amorphous minerals in

soils. Amount of Al-o and Si-o were greatly increased for heated soil relative to

unburnt soil with some increase in Fe-o. These changes probably reflect the

dissolution of poorly ordered and amorphous compounds formed by dehydroxylation

of kaolinite (an aluminosilicate), gibbsite (aluminum hydroxide) and goethite (iron

oxyhydroxide). There were no comparable trends for Al-o, Fe-o, Si-o in gravel and

the substantial value of Al-o and Si-o for ash may be due to heated soil minerals

being incorporated into the ash during collection. Plants generally contain little Al so

that the high Al-o values for ash are indicative of contamination with heated

kaolinite and gibbsite from the soil.

The DCB (d) extractant dissolves free iron oxides and some proportions of free

aluminum oxides and Si minerals depending on their composition and crystal size

(McKeague et al., 1971). DCB extractant dissolves several forms of organic and

inorganic Al and Fe, some Al in hydroxy interlayers and allophane (Wada, 1977;

Eckmeier et al., 2010). Heating increased both Al-d and Si-d but there was no

systematic effect of heating on Fe-d. There was no systematic effect of heating on

Al-d, Fe-d and Si-d in gravel.

72

Eckmeier et al. (2010) investigated the amount of Al and Fe extracted with

dithionite, pyrophosphate and oxalate extractants for heated black soils in the

Southern Switzerland. They found that the high amount of charred organic matter

was presumably responsible for the darker soil color and is closely related to Fe-p

and Al-p concentrations.

Table 5.3. Mean values of extractable Al, Fe and Si for soil, gravel, ash and charcoal

under and adjacent to burnt grass tree (GT) and eucalyptus (EU) logs. SP = sodium

pyrophosphate extractant, Oxalate = sodium oxalate extractant and DCB =

dithionite-citrate-bicarbonate extractant.

Sample Key SP Oxalate DCB

Al Fe Si Al Fe Si Al Fe Si mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

Soil Unburnt 2498 792 378 4711 1380 773 4007 10908 413

GT 0-10cm 6494 477 420 12274 3265 3806 11329 9017 1517

GT 10-20cm 6604 1275 406 10103 3074 1945 10479 17595 1123

GT 20-30cm 5043 938 414 7456 2293 1737 6492 12190 858

EU 0-10cm 4055 743 327 10516 3680 3538 5393 9097 861

EU 10-20cm 3109 721 354 6034 2218 1741 5347 12107 966

EU 20-30cm 2804 676 398 4531 1801 1172 5121 14094 835

Gravel Unburnt 2764 480 337 7401 4598 3763 8029 5704 1455

GT 0-10cm 2898 894 344 9940 6756 3119 7910 6555 1713

GT 10-20cm 2596 409 370 7979 4809 2378 9636 9629 1687

GT 20-30cm 3245 442 385 8706 4988 2446 9167 7982 1552

EU 0-10cm 2570 929 347 7979 5039 2029 7083 5568 1106

EU 10-20cm 1984 405 234 5111 5836 1281 6157 4768 836

EU 20-30cm 3423 534 294 6386 6647 2937 6134 7216 707

Ash and charcoal GT Ash 919 8 809 1221 341 1199 545 447 433

GT Charcoal 34 20 79 51 28 39 25 21 29

EU Ash 1981 8 2036 3190 775 2603 1011 1015 703

EU Charcoal 368 85 168 407 161 118 370 968 140

5.3.3. Mineralogical and morphological effects of bushfire heating on soil

minerals

There have been numerous laboratory studies of the dehydroxylation mechanisms of

kaolinite, goethite and gibbsite (De Faria and Lopes, 2007; Landers and Gilkes,

2007; Yusiharni and Gilkes, 2011). These are the major hydroxylated minerals in the

73

Wundowie soil. Kaolinite dehydroxylates to form metakaolinite at temperatures

above 500oC, gibbsite altered into boehmite and amorphous alumina and goethite

transforms to hematite (hydrohematite) at about 300oC.

Conventional and SXRD patterns of heated and unheated soil materials from the

Wundowie bushfire site show the effect of fire on the mineralogy of the fine fraction

of soil and gravel (Fig. 5.5 and 5.6). The main crystalline compounds of soil and

gravel are quartz, gibbsite, kaolinite, and a trace amount of goethite. There is

abundant hematite in gravel but little in <2 mm soil.

The conventional and SXRD patterns for both the fine fraction of soil and gravel

heated under burnt GT and EU logs show that at the largest distance (20-30 cm)

from the logs all three hydroxylated soil minerals persisted. Close to the burnt logs

gibbsite and kaolinite intensities had decreased substantially indicating that these

minerals had been partly dehydroxylated. Iglesias et al. (1997) found that vermiculite

and kaolinite experienced alteration in the surface soil (0-5 cm). Eggleton and Taylor

(2009) considered that heat from burning bark of Eucalyptus tetrodonta (Darwin

stringybark) might be sufficient to convert gibbsite into boehmite. They presented

results where soil in an unburned area had 26% gibbsite and 20% boehmite whereas

after burning, the amount of gibbsite decreased to 11% and boehmite increased to

29%. Nørnberg et al. (2009) set up an experimental forest fire where they measured

soil temperature, they found that goethite transformed to hematite at 320oC.

Similarly Ketterings et al. (2000) found that burning the topsoil destroyed gibbsite at

a quite low temperature and kaolinite at 600oC.

Calcite is most abundant in the soil under the burnt logs (0-10 cm) as this mineral is

a common constituent of plant ash (Harper et al., 1982; Ulery et al., 1993). The

presence of calcite in soil under burnt vegetation is also confirmed by Iglesias et al.

(1997) where the XRD peaks appeared for soils under burnt Juniperus oxycedrus.

Quintana et al. (2007) observed the occurrence of calcium oxalate or known as

whewellite (Ca(C2O4).(H2O)) in unheated Spanish juniper (Juniperus thurifera L.), a

mineral that turns into calcite at heating around 400oC. Wattez and Courty (1987)

pointed out that calcite crystals formed from the transformation of calcium oxalate

during the combustion of plant tissue, especially wood from different tree species.

74

Ulery et al. (1996) estimated about 1 to 2% of the land area she investigated had

concentrated fuel such as logs which burnt at higher temperatures resulting in soil

mineral alteration. Kaolinite was completely destroyed within 1 to 8 cm of topsoil.

We estimate that at Wundowie, about 0.1% of the forest floor is occupied by fallen

trees with each fallen tree trunk occupying about 4 m2. This estimate includes trees

that fall during fires. At Wundowie fires occur every 10 years on average so that for

each hectare of forest floor, 10 m2 of topsoil will be strongly heated under a burning

log. The landscape is extremely ancient and stable (McArthur, 1991) with soil ages

in excess of 104 years so that on average 104 m2/ha (ie 100%) of soil will have been

strongly heated. Bioturbation, minor colluviation, etc will have caused some mixing

of heated soil materials into the subsoil but clearly heating under a burning log is

likely to have significantly affected topsoil in this forest.

X-ray diffraction patterns of grass tree and eucalyptus ash and charcoal are shown in

Fig 5.7. The major crystalline compounds present for both ashes are calcite, apatite

and quartz. SXRD patterns show strong and sharp apatite reflections. Charcoal

contains abundant amorphous material (mostly carbon) as indicated by the broad

background scatter centered at 15o and 44o 2θ together with calcite in grass tree

charcoal and quartz impurity in eucalyptus charcoal. Harper et al. (1982) also

reported that calcite, apatite and quartz may are present in the ash of plant materials

with quartz being an impurity. Ulery et al. (1993) identified calcite together with

minor unidentified peaks in XRD patterns of wood ash. Calcite, lime, portlandite and

calcium silicate have been identified in wood ash from several tree species (Wattez

and Courty, 1987; Etiegni and Campbell, 1991).

75

a) Fine soil fraction

b) Gravel

c) Whole soil

Figure 5.5. Conventional XRD patterns (a= fine soil fraction and b=gravel) and

Synchrotron XRD (c=whole soil) of soil under and adjacent to a burnt grass tree

(GT) log (Q=quartz, K=kaolinite, Gi=gibbsite, H=hematite and C=calcite).

GT Soil 20-30cm

GT Soil 0-10cm

GT Soil 10-20cm

K K

Q

Q

H

GoGi C

10 20 30

2θ (degrees)

K KQ K

Q

GT Gravel 0-10cm

GT Gravel 10-20cm

GT Gravel 20-30cm

H

Gi CGo

10 20 30

2θ (degrees)

K K

QK

Q

GT Soil 0-10cm

GT Soil 10-20cm

GT Soil 20-30cm

K

C

Gi HGo

76

a) Fine soil fraction

b) Gravel

c) Whole soil

Figure 5.6. Conventional XRD patterns (a=fine soil fraction and b=gravel) and

Synchrotron XRD (c=whole soil) of soil under and adjacent to a burnt eucalyptus

(EU) log (Q=quartz, K=kaolinite, Gi=gibbsite, H=hematite and C=calcite).

EU Soil 20-30cm

EU Soil 0-10cm

EU Soil 10-20cm

K

Q

Gi K Go

Q

C

H

10 20 30

2θ (degrees)

K Q H

Q

EU Gravel 0-10cm

EU Gravel 10-20cm

EU Gravel 20-30cm

CH

Gi K KGo

10 20 30

2θ (degrees)

K GoQ K

Q

EU Soil 0-10cm

EU Soil 10-20cm

EU Soil 20-30cmCGiH

H

77

a) Ash-CXRD

15 20 25 30 35 40 45 50 55 60 65

EU Ash

GT Ash

Q

C

C

C C C C

C

A A A A A

b) Charcoal-CXRD

15 20 25 30 35 40 45 50 55 60 65

EU Charcoal

GT Charcoal

C

Carbon

Q Carbon

Carbon

Carbon

c) Ash-SXRD

EU ASH Synchrotron

GT ASH Synchrotron Q

Q

A

C

A A

C

Q

C C C

Q

C C

A Q A

Figure 5.7. Conventional (a, b) and Synchrotron XRD (c) patterns of ash and

charcoal for burnt grass tree (GT) and eucalyptus (EU) logs (Q=quartz, A= apatite,

and C=calcite).

78

Scanning electron microscopy of grass tree ash (Fig. 5.8) indicates that the ash

contains particles of inorganic compunds with sizes ranging from <1 μm to 100 μm

with both euhedral and irregular shapes. Many calcite (CaCO3) (5.8a) particles range

from <10 μm to 100 μm and have a rhombic shape, Ca and P-rich particles are

apatite (Ca10(PO4)6(Cl,OH)2) (5.8b) with sizes <10 μm and an irregular shape.

Potassium rich particles (5.8c and 5.8d) are <10 μm and the potassium salt(s) may

include sulfates, which are mixed with calcite and possible calcium sulfate. SEM

micrographs and associated EDS X-ray spectra of eucalyptus ash (Fig. 5.9) shows

rhombic and microcrystalline calcite (5.9a), mixed calcium-magnesium carbonate

(5.9b), silt-size grains of the soil minerals, quartz and K-feldspar (KAlSi3O8) (5.8c,

5.8d) mixed or coated with calcite. Humphreys et al., (1987) and Yusiharni et al.

(2007) also found that plant ash mainly consisted of calcite however Etiegni and

Campbell (1991) observed that quicklime (CaO) and calcium silicate (Ca2SiO4)

could also be present. The present results indicate the complex nature of ash, with a

high diversity of composition, shape and size of particles.

5.4. Conclusion

The Wundowie bushfire added calcite and other salts in plant ash to the soil, which

considerably increased soil pH and EC values. Plant ash also increased

concentrations of Ca, K, Mg, Mn, and P in the heated soil. The amount of C and N in

burnt soil and gravel were less closer to center of burnt logs. Soil heated under burnt

logs experienced alteration of some soil minerals (Ulery et al., 1996) and amorphous

compounds may have formed. These compounds are generally chemically reactive

with a high capacity to adsorb ions and thus, may affect soil fertility (Ketterings et

al., 2000). Gibbsite, goethite and kaolinite in the topsoil were partly dehydroxylated

by heating in the bushfire. Dehydroxylation of kaolinite and gibbsite created

amorphous compounds, which increased oxalate extractable Al and Si. Clearly

dehydroxylated minerals and possibly their rehydroxylated forms must be present in

naturally heated soils and these complex minerals may have a significant effect on

the chemical behaviour of the soil.

79

0

5000

10000

15000

20000

25000

1 2 3 4 5 6 7 8 9 10

Energy (keV)

Ca

Ca

a)

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

1 2 3 4 5 6 7 8 9 10

Energy (keV)

Ca

Ca

P

b)

0 500

1000 1500 2000 2500 3000 3500 4000 4500 5000

1 2 3 4 5 6 7 8 9 10

Energy (keV)

Ca

Ca

K

S

c)

0

2000

4000

6000

8000

10000

12000

1 2 3 4 5 6 7 8 9 10 Energy (keV)

Ca

Ca

K

S d)

Figure 5.8. Scanning electron micrograph (SEM) and X-ray spectra of the indicated

particles for grass tree ash, where large rhombic calcite crystal (CaCO3) (a),

microcrystalline apatite (Ca10(PO4)6(OH)2) with calcite (b) and mixed potassium and

calcium salts (c and d) are present.

A

B

CD

A

B

CD

80

0 2000 4000 6000 8000

10000 12000 14000 16000 18000 20000

1 2 3 4 5 6 7 8 9 10 Energy (keV)

Ca

Ca a)

0

2000

4000

6000

8000

10000

12000

14000

1 2 3 4 5 6 7 8 9 10 Energy (keV)

Ca

Ca

Mg

S K Si P

b)

0 1000 2000 3000 4000 5000 6000 7000 8000 9000

10000

1 2 3 4 5 6 7 8 9 10 Energy (keV)

Ca

Ca

K

Si

Al

Mg

c)

0 2000 4000 6000 8000

10000 12000 14000 16000 18000 20000

1 2 3 4 5 6 7 8 9 10

Energy (keV)

Ca

Ca K S P

Al

d)

Figure 5.9. Scanning electron micrograph (SEM) and X-ray spectra of the indicated

particles for eucalyptus ash where rhombic calcite (CaCO3) (a), microcrystalline

mixed calcium and magnesium carbonates and sulfates (b), an alumino-silicate,

possibly K-feldspar (KAlSi3O8) with calcite (c) and phosphorus enriched calcite (d)

(possibly calcite with apatite) are present.

A

B

CD

A

B

CD

81

Chapter 6

6.0 Minerals in the ash of Australian native plants

6.1. Introduction

Wild and managed fires burn many thousands of hectares of forest annually and

consume some or all of the vegetation and litter (Ulery et al., 1996). The intensity of

forest burning is related to weather conditions, amount of fuel available and the

condition of the fuel. Fires create ash consisting of organic and inorganic residues

from the combustion process (Ùbeda et al., 2009). Ash may be defined as a complex

mixture of charcoal, char and minerals (i.e inorganic compounds) (Scott, 2010). The

characteristics and the amount of forest plant ash produced under natural or

controlled conditions depend on several factors, including the species and the part of

plant that was combusted (leaves, fruit, bark or wood), plant age and combustion

degree (Clapham and Zibilske, 1992; Ulery et al., 1993; Vance and Mitchell 2000;

Demeyer et al., 2001).

Most elements in plants are retained in the ash with most N and some S being lost

(Perkiomaki, 2004). Consequently the N content in ash is usually small, particularly

where fuels are completely combusted (Khanna et al., 1994). The carbon (C) content

of ash varies with combustion degree, with C remaining as unburnt plant material,

charcoal and carbonate minerals (Vance and Mitchell, 2000). The color of ash varies

from black through grey, brown to white with the type of plant and combustion

process influencing ash color. Dark colored (black) ash contains higher amounts of

charred organic material compared to the light colored ash that is mostly composed

of crystalline or amorphous inorganic compounds (Knicker, 2007). Recent studies

have proposed that ash color could be use as an indicator of fire severity (Keeley,

2009; Pereira et al., 2011; Bodí et al., 2011) but the nature and amount of inorganic

compounds need to be considered and differ with vegetation type. Several laboratory

studies have examined the relationship between ash color and fire severity for

various biomass types and combustion temperatures (Misra et al., 1993; Liodakis et

al., 2005; Liodakis et al., 2009). For combustion at 600oC, CaCO3 and K2Ca(CO3)2

were present while for 1300oC combustion, CaO and MgO were the dominant

compounds (Misra et al., 1993). However, in general the nature of minerals in plant

ash is poorly understood and is the subject of this study.

82

Ash is generally an alkaline material with pH ranging from 9.0 to 13.5 (Khanna et

al., 1994; Liodakis et al., 2005) and ash has been used as a combined liming agent

and nutrient source (Etiegni and Campbell, 1991; Yusiharni et al., 2007). For

example, application of ash to an acid soil increased soil pH, and levels of

extractable Ca, Mg and SO4 (Voundi Nkana et al., 2002). Ash generally contains

significant amounts of most plant nutrient elements and thus can be regarded as a

multi nutrient fertilizer (Vance and Mitchell, 2000). Furthermore addition of ash to

soil may increase the availability of some soil nutrients to plants because of

exchange with ions dissolved from ash (Voundi Nkana et al., 2002).

Several studies have investigated the physical and chemical properties of ash from

diverse plant species (Etiegni and Campbell, 1991; Khanna et al., 1994; Ùbeda et al.,

2009; Gabet and Bookter, 2011). However, the forms of plant nutrients in ash

including the forms of water-soluble elements and their interactions with soils have

received little attention. The research described in this paper determined the amounts

and mineral forms of elements in ash derived from several Australian native plant

species and evaluated the reactions of plant ash with heated and unheated soil.

Table 6.1. The nomenclature for plant ash and some properties of the ash. SSA= specific surface area (m2/g), EC= electrical conductivity (mS/cm), Bic P= bicarbonate P (mg/kg). Key Species and plant material Ash % Ash Color SSA

(m2/g)

pH

(1:5)

EC (1:5)

(mS/cm)

Bic P

(mg/kg)

SW

SL

PM

WW

WL

RW

RL

GT

JL

HH

Silver wattle wood

Silver wattle leaf

Prickly moses leaf and twig

Wandoo wood

Wandoo leaf

Red gum wood

Red gum leaf

Grass tree leaf

Jarrah leaf

Harsh hakea leaf and twig

1.72

3.39

2.81

3.07

3.02

4.71

4.07

3.40

3.27

4.01

G1 8/N

G1 5/N

G1 4/N

2.5Y 7/2

G1 5/N

G1 7/N

G1 6/N

G1 3/N

2.5Y 6/2

G1 4/N

5.8

4.2

8.3

9.2

2.4

2.4

2.5

2.7

2.9

7.7

13.8

12.8

13.1

13.6

13.3

13.8

13.4

13.5

13.2

12.3

18.9

13.9

8.5

15.1

14.3

23.2

18.4

22.9

18.9

16.3

145

198

199

176

32

183

30

145

196

164

83

6.2. Material and Methods

6.2.1. Ash preparation

The Australian native plants used for the study are major constituents of forest

vegetation at Bakers Hills, Darling Range, Western Australia: silver wattle (Acacia

retinodes), prickly moses (Acacia pulchella), wandoo/white gum (Eucalyptus

wandoo), red gum or marri (Corymbia calophylla), grass tree (Xanthorrhoea

preissii), jarrah (Eucalyptus marginata) and harsh hakea (Hakea prostrata). The

leaves and wood samples were dried in an oven at 60oC until the weight was

constant (up to 7 days) and then cut to 1-2 cm to make the sample homogenous. A

subsample of approximately 50 g then placed in an aluminum oven tray and burned

in open air for about 15 min with agitation to maximize burning, to simulate

combustion in a forest fire in the field. Temperature was not recorded during the

combustion process. Combustion was almost complete for each material with only

small amounts of charcoal remaining in the ash. A key to the materials investigated

and the abbreviations used to identify these materials are given in Table 6.1.

6.2.2. Characterization of the ash

Ash color was classified using a Munsell Color Chart (Munsell Color Co., 1998).

Specific surface area (SSA) was measured using a Micrometrics Gemini 2375

instrument with VacPrep 061 using a five point BET method with N2 as the

absorbate. The pH of the ash was determined in a 1:5 deionized water extract. The

samples were analyzed for bicarbonate extractable P (Bic P) (Colwell, 1963). Total

carbon and nitrogen were determined on an Elementar CNS (Vario Macro) analyzer.

Water-soluble elements in ash were determined using Association of Official

Analytical Chemistry (AOAC) standard methods (AOAC, 1975). 0.5 g of ash was

mixed with 100 ml of DI water and placed on a mechanical shaker for 24 h. Extracts

were then filtered through a 0.22 μm Millipore filter. Water-soluble elements were

then measured using an inductively coupled plasma optical emission spectrometer

(ICP-OES) (Perkin-Elmer, Norwalk, CT, USA). Ash content of plant material was

determined by a separate dry combustion in a muffle furnace at 550oC for 2 hours

(Rayment and Higginson, 1992).

Total elements were determined in duplicate with a model PE ELAN 600 inductively

coupled plasma – atomic emission spectroscopy (ICP-OES) instrument (Perkin-

84

Elmer, Norwalk, CT, USA) after concentrated perchloric acid digestion of the ash

(Rayment and Higginson, 1992). Conventional XRD of the ash was conducted on a

Philips PW3020 diffractometer with a diffracted beam monochromator (CuKα,

50kV, 20mA). Powder samples were scanned from 4 to 70o 2θ, using a step size of

0.02o 2θ and a scan speed of 0.04o 2θ sec-1. Synchrotron XRD (SXRD) analysis was

performed at the Australian Synchrotron where powder samples were mounted into

glass capillaries with a 1.0Å wavelength set for this analysis in order to provide a

high peak/background and adequate resolution for identifying minor constituents.

The composition and morphology of the plant ash was examined by scanning

electron microscopy (SEM) and energy dispersive X-ray spectrometry (EDS) using a

JEOL 6400 instrument. Samples for SEM analysis were placed on metal stubs and

carbon coated before analysis. Principal component analysis was carried out on the

bulk chemical data and on element concentrations of ash particles derived from SEM

EDS spectra using STATISTICA (STATISTICA, 2011).

6.2.3. Incubation of ash-soil mixture

Topsoil (0-10 cm) from Bakers Hills, (Yalanbee soil, a lateritic podzolic soil, which

is an Alfisol (USDA, 2010)) was used in the incubation study. The same acid sandy,

highly P-deficient soil had been used in a study of the P fertilizer value of chicken

litter ash by Yusiharni et al. (2007). Soil properties (<2 mm fraction) are as follow;

pH (1:5 H2O) 4.5, Electrical conductivity (EC) (1:5 H2O) 0.04 mS/cm, total P 0.25

%, bicarbonate extractable P 0.49 ppm (Colwell 1963), total carbon 1.64%, total

nitrogen 0.08%, clay, silt and sand were 4%, 6%, and 90% respectively (Yusiharni,

2007). Subsamples of unheated soil were air dried and passed through a 2 mm sieve

for use in the incubation experiment and for chemical analysis, while other soil

subsamples were first heated at 200oC in a muffle furnace for one hour to simulate

heating in a bushfire. 100 g of unheated and heated soil were thoroughly mixed with

each ash (SW, SL, PM, WW, WL, RW, RL, GT, JL and HH) in vials at rates of 0

and 1 g/100g soil. Each mixture was replicated two times, wetted to about 20% water

content with addition of toluene as a microbial inhibitor. The tops were firmly

screwed onto the vials, which were placed in the dark at 20oC for 0, 1 and 5 weeks.

After each incubation period a subsample was removed for analysis for Bic-P, EC

and pH as described above.

85

6.3. Results and Discussion

6.3.1. Characteristics of the ashes

The ash content of the plant materials as the result of furnace burning ranged from

1.7 to 4.7%. Properties of plant ash produced by open-air burning are shown in Table

6.1. Ash was white (G1 8/N, G1 7/N, G1 6/N, G1 5/N, or G1 4/N) and grey (G1 3/N,

2.5Y 7/2 or 2.5Y 6/2). Grey ash contains more carbon (charred organic material)

than white ash possibly reflecting differences in the intensity of combustion (Khanna

et al., 1994; Neary et al., 1999; Lentile et al., 2006; Roy et al., 2010). White ash

generally contains abundant calcite (CaCO3) (Ulery and Graham, 1993) as will be

discussed subsequently.

All the ashes were alkaline with pH ranging from 12.3 to 13.8. These high values are

associated with the presence of carbonates, oxides and hydroxides of base cations in

ash (Ulery et al., 1993). The electrical conductivity of the ash extracts (1:5) was

high, indicating that the ash contains considerable amounts of soluble salts, with red

gum wood ash containing the highest amount (23.2 mS/cm) and prickly moses

leaf/twig ash the least (8.5 mS/cm). Other workers have made similar observations

(Kutiel and Naveh, 1987; Iglesias et al., 1997; Badía and Martií, 2003). The specific

surface area of the ashes ranges from 2 to 9 m2/g indicating that ash particles

(crystals) are very small (micron size) and are thus likely to be quite reactive.

Amounts of available phosphorus (P) in the ash ranged from 29 to 198 mg/kg

reflecting the combustion of organic P compounds and the presence of partly soluble

P minerals in ash (Kutiel and Shaviv, 1989; Galang et al., 2010).

The ash produced by open-air combustion in this study may not be identical to ash

produced during wildfires, prescribed fires or ash derived from laboratory ignition at

controlled temperatures. Raison (1979) mentioned that ash generated in a muffle

furnace would be different from ash produced during a wildland fire. Bodí et al.

(2011) also found in their recent study that ash derived at several temperatures in a

laboratory study may not have effectively replicated ash created by a bushfire.

86

Table 6.2. Total element concentrations in plant ash (mg/kg). Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

Element (mg/kg)

Ash Type SW SL PM WW WL RW RL GT JL HH

Al 1014 1802 3499 2028 1375 840 2540 1085 3689 6091 As 3 5 4 3 2 0 0 1 3 5 B 242 282 164 197 1215 153 657 135 936 248 Ba 494 372 1092 913 842 169 149 561 55 60 Be 0.2 0.4 0.3 0.3 0.5 0.1 0.1 0.1 0.3 0.2 Ca 289300 206800 134200 248300 177600 219400 147900 162801 149800 158600 Ce 6 27 51 43 23 4 6 11 9 15 Co 3 2 3 1 1 2 6 2 3 2 Cr 42 24 44 18 41 11 30 131 62 45 Cu 122 453 92 71 104 97 97 88 131 99 Fe 2688 4098 13792 4067 2920 1518 4473 4177 14850 22870 K 32370 41660 37400 29380 42110 32660 35810 45790 33000 34080 La 6 30 58 31 15 2 4 10 3 8 Mg 58670 76801 21570 49760 60940 43420 66990 69610 111300 33920 Mn 387 1005 522 2998 3150 694 822 890 3494 3955 Mo 1 1 3 1 1 1 1 4 3 4 Na 1792 10080 6319 26520 92980 24340 60220 9406 83140 47990 Nb 5 5 8 5 5 3 5 4 8 8 Ni 7 9 15 8 20 10 16 30 35 14 P 4986 11190 6097 2649 7715 9940 13470 10750 8867 7074 Pb 3 3 23 14 5 0 13 21 35 8 Rb 148 210 179 187 170 205 196 229 92 107 S 1842 6046 5086 2437 5490 2121 5489 8738 6454 23820 Si 5610 10010 15870 11450 9140 9090 18300 10600 17220 7010 Ta 26 23 26 28 23 22 22 23 32 35 Ti 168 291 764 242 226 119 375 206 887 706 U 16 11 8 13 9 13 9 10 9 9 Y 1 8 17 6 6 1 1 3 2 7 Zn 61 322 273 92 1039 292 335 149 131 317 Zr 4 5 12 3 4 3 5 4 3 7

87

Table 6.3.

(a) Concentration of water-soluble elements in plant ash. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

Ash Type Water Soluble Element Concentration (mg/kg)

Na Mg K Ca P S B Cl SW 1100 3500 7217 25990 5 54 22 3600 SL 1629 6332 11330 22700 13 150 20 11290 PM 892 5500 6195 38650 17 132 16 6607 WW 5103 4137 4680 23320 63 64 32 200 WL 15020 6581 8666 16510 52 148 163 15370 RW 4655 6429 11810 25980 185 52 17 717 RL 10080 4650 16180 31830 221 143 85 4670 GT 1733 2545 20131 26250 129 234 10 14410 JL 12600 5800 12730 37550 166 164 74 2997 HH 8658 6650 9410 23430 57 303 18 9226

(b) Proportion of element in plant ash that is water-soluble.

Ash Type Proportion of Element that is Water Soluble

Na Mg K Ca P S B SW 0.61 0.06 0.22 0.09 0.00 0.03 0.09 SL 0.16 0.08 0.27 0.11 0.00 0.02 0.07 PM 0.14 0.25 0.17 0.29 0.00 0.03 0.10 WW 0.19 0.08 0.16 0.09 0.02 0.03 0.16 WL 0.16 0.11 0.21 0.09 0.01 0.03 0.13 RW 0.19 0.15 0.36 0.12 0.02 0.02 0.11 RL 0.17 0.07 0.45 0.22 0.02 0.03 0.13 GT 0.18 0.04 0.44 0.16 0.01 0.03 0.07 JL 0.15 0.05 0.39 0.25 0.02 0.03 0.08 HH 0.18 0.20 0.28 0.15 0.01 0.01 0.07

* The proportion of water soluble Cl could not be calculated as no total Cl values are available due to use of perchloric acid dissolution.

6.3.2. Total and water extractable elements, carbon and nitrogen concentrations

Table 6.2 lists the mean element concentrations for the ashes. Plant ash contains

most plant nutrient elements and many other elements (Khanna et al., 1994). The

most abundant elements in these ash samples were Ca, K, Mg, P, S, Fe, Mn, Na, and

Si, which is in agreement with other work (Etiegni and Campbell, 1991; Misra et al.,

1993; Gabet and Bookter, 2011).

88

Total element concentrations in ash differed substantially between plant materials as

has been reported by Khanna et al. (1994) and Liodakis et al. (2009). The chemical

composition of ash is highly dependent upon the fuel (e.g. wood and leaf

proportions) and the combustion degree (Khanna et al., 1994). Silver wattle wood

ash (SW) had the highest concentration of calcium (289,300 mg/kg), jarrah leaf ash

(JL) had the highest concentration of magnesium (111,300 mg/kg) and grass tree ash

(GT) had the highest concentration of potassium (45,700 mg/kg). The calcium

concentration is generally higher for wood and twig ash compared to leaf ash,

concentrations of magnesium and potassium were higher in leaf ash than in wood

ash (Table 6.2 and Fig. 6.1). Other elements are present at relatively small

concentrations in ash, with some elements being most likely present in minerals

derived from soil/dust contaminating the plant materials (i.e. Ti, Al, Zr, Pb) (Fig.

6.1) (Ludwig et al., 2005). The plant samples were not washed prior to burning as

this would remove soluble elements and in addition soil and dust minerals and their

thermal reaction products would be present in natural ash.

The concentrations of water-soluble elements in ash are presented in Table 6.3a. The

solubility in water of elements in ash is determined by the type of minerals formed

during the fire and for some elements solubility is also a consequence of the ash pH

values, as these elements (e.g Fe, Zn, Cu) are less soluble in the pH range above 7

(Holden, 2005). The ashes studied contain significant amounts of water-soluble P, S,

Cl and B, which indicates that plant ash will be a significant source to plants of these

nutrient elements (Khanna et al., 1994). Water-soluble K ranged from 6190 to 20100

mg/kg, Ca from 33000 to 77300 mg/kg with no systematic differences between

wood and leaf ash, Mg ranged from 2540 to 6600 mg/kg and Na from 892 to 15020

mg/kg. Pereira et al. (2011) also demonstrated that amounts of water-soluble

elements in ash vary substantially according to species and that ash contains

considerable amounts of water-soluble Ca, Mg, K and Na (Etiegni and Campbell,

1991; Khanna et al., 1994).

The proportion of the total element concentration that is soluble in water was

calculated and is presented in Table 6.3b. For P and S, only very minor proportions

of these elements dissolved. For alkali elements, higher proportions dissolved but

generally most of these elements and also B remained insoluble. We conclude that

89

this low solubility of elements in ash is probably due to the presence of sparingly

soluble minerals in ash as is discussed later and also to the low solubility of some

minerals at high pH.

Principal component analysis provides a convenient procedure for assessing a large

body of data for total element concentrations in ash (Fig. 6.1a and 6.1b). This

analysis provides a general overview of the differences in element concentration and

the grouping of elements for all materials derived from several plant species and

plant parts. It seems that the elements are mostly not closely grouped reflecting the

diverse nature of these plant materials. However, Ca is distinctly separated and

negatively related to the majority of elements and this distribution is partly due to the

high amounts of Ca in the wood ash (Fig. 6.1b). Some quite closely associated

elements (Ti, Al, Zr) may have originated from dust/soil contamination and are more

abundant in leaf/twig ash than in wood ash, which is to be expected, as the large

exposed area of leaf and twig will be more readily contaminated with dust than

would wood (Ulery et al., 1993). The original plant materials were not washed to

remove dust as we wished to simulate natural conditions during forest fires. We

considered that some of the crystalline compounds in ash might be due to plant

constituents chemically reacting with dust/soil constituents during combustion.

Principal component analysis for the concentration of elements soluble in water

including their solubility as a proportion of total element concentrations is illustrated

in Figures 6.2a, b, c and d. Water-soluble calcium is well separated from other

elements and it is more strongly associated with wood ash (RW, WW and SW) and

also PM and SL ash. The association of water-soluble P, Mg, Na and B is mostly

related to leaf ash materials (WL, JL and RL). Water soluble Cl, K and S are

associated and are relatively abundant in grass tree ash. Quite distinct groupings (e.g.

Ca, K, P) occur when elements are expressed as log proportion of total element

soluble in water (Fig. 6.2c and d). Explanation for these groupings will be provided

by the XRD and SEM/EDS results discussed below. Several studies have observed

that water-soluble element concentrations varied according to plant species and

combustion temperature (Gray and Dighton, 2006; Pereira et al., 2009). A recent

study by Pereira et al. (2011) of ash from cork oark (Quercus suber) observed

substantial water soluble Ca, Mg, Na, Si and S in the ash.

90

a)

Al

As

B

Ba

Be

C

Ca

Ce

Co Cr

Cu

Fe

K

La

Mg

Mn Mo

N

Na

Nb

Ni P

Pb

Rb

S

Si

Ta

Ti

U

Y

Zn

Zr

-1.0 -0.5 0.0 0.5 1.0

Factor 1 : 32.52%

-1.0

-0.5

0.0

0.5

1.0Fa

ctor

2 :

19.6

2%

Al

As

B

Ba

Be

C

Ca

Ce

Co Cr

Cu

Fe

K

La

Mg

Mn Mo

N

Na

Nb

Ni P

Pb

Rb

S

Si

Ta

Ti

U

Y

Zn

Zr

Contaminants in dust

b)

SW

SL

PM

WW

WL

RW

RL

GT

JL

HH

-8 -6 -4 -2 0 2 4 6 8 10

Factor 1: 32.52%

-8

-6

-4

-2

0

2

4

6

8

Fact

or 2

: 19.

62%

SW

SL

PM

WW

WL

RW

RL

GT

JL

HH

Ca-richwood ash

Figure 6.1. Principal component analysis of log total element concentration for native plant ash, variables (elements) (a) and plant material (b). Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

91

(a)

Na Mg

K

Ca

P

S

B

Cl

-1.0 -0.5 0.0 0.5 1.0

Factor 1 : 32.33%

-1.0

-0.5

0.0

0.5

1.0Fa

ctor

2 :

25.5

0%

Na Mg

K

Ca

P

S

B

Cl

(b)

SW

SL PM

WW

WL RW

RL

GT

JL

HH

-4 -3 -2 -1 0 1 2 3 4 5

Factor 1: 32.33%

-5

-4

-3

-2

-1

0

1

2

3

4

Fact

or

2: 25.5

0% SW

SL PM

WW

WL RW

RL

GT

JL

HH

(c)

Na

Mg

K

Ca

P

S

B

-1.0 -0.5 0.0 0.5 1.0

Factor 1 : 29.87%

-1.0

-0.5

0.0

0.5

1.0

Fact

or 2

: 25

.33%

Na

Mg

K

Ca

P

S

B

(d)

SW

SL

PM

WW WL

RW

RL

GT

JL

HH

-5 -4 -3 -2 -1 0 1 2 3 4

Factor 1: 29.87%

-4

-3

-2

-1

0

1

2

3

4F

act

or

2: 25.3

3%

SW

SL

PM

WW WL

RW

RL

GT

JL

HH

Figure 6.2. Principal component analysis of log water soluble element concentrations, variables (a) and cases (b) and log proportion that is water soluble, variables (c) and plant material (d) for native plant ash. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

Mean values of total carbon and nitrogen concentrations in raw plant materials and

ash are presented in Fig. 6.3. Total carbon was above 40% for all the raw plant

materials and decreased to below 10% for their ash. Total carbon concentrations are

similar for all the plant samples. The reduced concentrations of total carbon in plant

ash are due to the almost complete removal of organic material from the ash with

92

remaining carbon being mostly present as carbonates (Misra et al., 1993; Gabet and

Bookter, 2011). Wandoo leaf ash (WL) and harsh hakea ash (HH) had the highest

total carbon and their ash color was considered to be white (G1 5/N and G1 4/N

respectively) rather than grey. Burning plant materials greatly reduced the nitrogen

concentration in the ashes. Almost complete loss of N occurs where combustion of

the fuel is nearly complete (Khanna et al., 1994; Raison et al., 1985). The N content

of ash produced at temperatures above 600 oC is normally low, which is presumably

due to the conversion of most plant nitrogen to NH3, NOx and N2 gases during the

combustion process (Misra et al., 1993).

6.3.3. Mineralogy and Morphology of ash (XRD and SEM)

Conventional XRD patterns of SW, SL, PM, WW, WL, RW, RL, GT, JL and HH

showed that the major compounds present in the ashes are calcite, apatite and quartz

(Appendix 10). Harper et al. (1982) also found that calcite, apatite and quartz were

present in the ash of plant materials where quartz was identified as being an impurity

from dust or soil. Ulery et al. (1993) observed calcite and some minor unidentified

peaks in plant ash. The sensitivity of conventional XRD is insufficient to identify

trace amounts of minerals, therefore synchrotron XRD (SXRD) patterns of ashes

(Fig. 6.4a, b and 6.5a, b) in a glass capillary were also obtained. These patterns offer

better detection and resolution of weak and adjacent reflections than is provided by

conventional XRD (Williams et al., 2003). However the glass of the capillary

contributes to the broad background scattering seen in Fig. 6.4 and 6.5 so that

amorphous silica (e.g. phytoliths) can not be detected.

The main compounds identified by SXRD in the plant ashes were various oxides,

carbonates and hydroxides of Ca, Mg and K as has also been reported by Misra et al.

(1993). The minerals present were fairchildite (K2Ca(CO3)2), nesquehonite

(MgCO3.H2O), calcite (CaCO3), sylvite (KCl), lime (CaO), scolecite

(CaAl2Si3O10.3(H2O),), portlandite (Ca(OH)2, periclase (MgO), apatite group

minerals probably resembling hydroxyl-apatite (Ca5(PO4)3(OH)) and wilkeite

(Ca5((P, S, Si)O4)3(OH,CO3)) and quartz (SiO2 (Table 6.4). Fairchildite,

nesquehonite, scolecite, portlandite and periclase were not identified using

conventional XRD. As discussed above, the proportion of water-soluble elements in

93

ash (Table 6.3.b) is low, which may be attributed to the presence of these mostly

poorly soluble minerals. Hydroxide minerals form when the ash is exposed to air as

oxides of Ca, Mg and K react with atmospheric water (Misra et al. 1993). Liodakis et

al. (2005) detected periclase (MgO), lime (CaO) and portlandite (Ca(OH)2 in wood

ash prepared at 600oC. Reaction of these oxides with atmospheric carbon dioxide

results in the formation of calcite and other carbonates. However, carbonates also

form during combustion. Periclase and lime probably formed directly via the

combustion of magnesium- and calcium-containing organic matter rather than via

decomposition of MgCO3 and CaCO3 (Liodakis et al., 2005).

0

10

20

30

40

50

60

SW SL PM WW WL RW RL GT JL HH

C %

% C in Original Plant % C in Ash

0

1

2

3

SW SL PM WW WL RW RL GT JL HH

N %

% N in Original Plant % N in Ash

Figure 6.3. Mean (n=3) values with standard error bars of carbon (C) and nitrogen (N) concentrations (%) for native plant materials (original and ash). Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

94

SXRD could not detect the present of halite (sodium chloride) in the ash, even

though ICPO-ES data showed abundant sodium (up to 93000 mg/kg of total Na) and

chloride (up to 15370 mg/kg water soluble Cl) in the ash so these ions may not be

present as halite. This interpretation is consistent with the low proportion (0.15-0.29)

of total Na that was soluble in water (Table 6.3b).

Scanning electron microscopy of particles and associated EDS X-ray spectra for all

the ashes show that diverse particle sizes, shapes and compositions occur (Fig. 6.6

and 6.7). Ash contains particles with sizes ranging from <1 μm to 100 μm. Apart

from distinct mostly prismatic, large calcite crystals, most grains seen in the

micrographs are K- and Mg-rich aggregates of very small particles often containing

Cl and they commonly contain little P, Fe, S and Si. The sensitivity of EDS for Na is

extremely poor, so that Na peak is not visible in EDS spectra. Liodakis et al. (2009)

found that SEM spectra of particles of plant ash contained lines due to Ca, Mg, K, P

and Si as observed in this work. The present results indicate the complex nature of

ash, with a high diversity of composition, shape and size of particles.

Table 6.4. Crystalline compounds in plant ash identified by SXRD. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

Compound Ash Type SW SL PM WW WL RW RL GT JLA HH

Calcite *** *** *** *** *** *** *** *** *** *** Sylvite ** *** *** * *** ** *** *** *** *** Fairchildite ** *** ** - - - - ** - - Nesquehonite - - - * *** * - - - - Scolecite - ** ** - - - *** - *** ** Periclase * ** * *** ** * ** ** ** * Apatite ** *** ** ** *** ** ** ** ** *** Lime ** * * ** ** * * * * * Portlandite - - - *** - ** *** - ** ** Quartz - - *** * * - * - * ***

The EDS spectra and XRD results indicate that grains are mostly complex mixture of

extremely small crystals including calcite, fairchildite, nesquehonite, sylvite, lime,

scolecite, quartz, portlandite, periclase, apatite (hydroxyl-apatite and wilkeite).

95

Principal component analysis of log element concentrations derived from SEM-EDS

spectra of ash particles (Figs. 6.8a, b) is helpful in determining the underlying

associations of elements in ash particles. Calcium was quite well separated from

other elements presumably representing particles of calcite, lime or portlandite,

while P, Mg, K, Cl and Na are grouped representing the various halide, phosphate

and mixed cation carbonates that are present. The grouping of Al, Fe, Si and Ti

probably represent impurity particles derived from dust and these include the soil

minerals quartz (SiO2), anatase (TiO2), kaolin (Al2Si2O5(OH)4), goethite (FeOOH)

and hematite (Fe2O3). Leaf and twig ash contains more of these contaminating

minerals than wood ash as a consequence of the greater surface area of these plant

materials. Some Si also occurs as amorphous silica within plant materials, as it is a

structural constituent of many grasses, prickly shrubs, etc (Lanning et al., 1958). The

plant ash materials are not tighltly grouped in Fig. 6.8b due to the diversity of

minerals present and the various comcentrations of the impurity minerals.

6.3.4. Ash and soil incubation

The effects on pH, EC and Bic P of the addition of ash to unburnt and burnt soil are

illustrated in Figures 6.9 and 6.10. The soil had an original pH of 4.5, which was not

affected by heating but pH increased substantially when ash was added to both the

unburnt and burnt soil. The addition of all ashes increased soil pH by about 3 units,

with wandoo leaf ash (WL) increasing the pH of unburnt and burnt soil by about 4

units. The EC of burnt soil was more than for unburnt soil and was greatly increased

by the addition of ash. EC decreased slightly after 5 weeks of incubation. The

increases of pH and EC for soil plus ash are due to the dissolution of salts include

carbonates in the ash. Differences in EC values for different ashes reflect the

different amounts of soluble halides, carbonates, hydroxides and oxides present in

the various ash materials (Goforth et al., 2005; Terefe et al., 2008). The decrease of

EC for the 5 week incubation for burnt and unburnt soil plus ash possibly reflects the

crystallization of less soluble minerals during incubation (Kutiel and Shaviv, 1992).

The amount of plant-available P (bicarbonate soluble P) in burnt soil increased

relative to unburnt soil and increased substantially with addition of ash for 0 and 1

weeks of incubation. Values had decreased after 5 weeks incubation time

96

presumably due to P-fixation. The increase in Bic P for burnt soil reflects the

combustion of organic P compounds in the soil (Galang et al., 2010), with a possible

contribution from the transformation of soil minerals. The addition of substantial

amount of P to the soil in ash, which contains apatite and wilkeite clearly increased

the available P in the soil (Kutiel and Shaviv, 1992).

The considerable amounts of plant nutrient elements in plant ash can be used to

improve soil fertility. The understanding of mineralogical and morphological

properties of ash provided by this research helps explain and predict some effects of

ash on soil fertility. A major finding is that some compounds in ash are sparingly

soluble in water so that plant nutrient elements (e.g. P in apatite) are not readily

released to soil solution

Ash deposited during wildfire and prescribed fires contains diverse compounds that

will variously affect soil properties, especially for acid soils. Addition of ash to

heated and unheated acid soil during the incubation study increased the pH and EC

values due to dissolution of the salts in the ash. The different pH and EC values for

each ash are explained by the different amounts and occurrences of carbonates and

other salts in ash. The increased amount of plant available phosphate in soil to which

ash has been added is at least partly due to dissolution of the apatite in ash.

6.4. Conclusion

The ash of these native plant species has diverse chemical, mineralogical and

morphological properties and differs in solubility, depending on plant species and

plant part. Compounds in the ash include fairchildite (K2Ca(CO3)2), nesquehonite

(MgCO3.H2O), calcite (CaCO3), sylvite (KCl), lime (CaO), scolecite

(CaAl2Si3O10.3(H2O)), quartz (SiO2), portlandite (Ca(OH)2, periclase (MgO). Much

of the P in ashes is present as the mineral apatite. These compounds are more soluble

when applied to acid soil where they could provide substantial amount of P for

plants whereas P in apatite may be much less available if the ash is deposited on

naturally alkaline soil or on a soil that has been raised to a high pH value due to the

liming action of ash. These effects will occur in the field for prescribed or wild fires

and their magnitude will presumably vary with the duration and intensity of fire, the

type and abundance of fuel and soil type. These topics deserve further investigation.

97

(a)

5 15 25 35 2 Theta (deg)

N

C

N

SW

F

N

S

A

L N

F

C

L

N

S

P L

C

C Q

Po

C

C N

C

A N A S

A

C

C

SL

PM

WW

WL

A

Q A

A A

C P

C C

S C

C C S

P A

E

F

E

(b)

14 24 2 Theta (deg)

N

C

N

SW

F

N

S

L

F

P L

Q

Po

C

P

N

A

S

N

Q

C

SL

PM

WW

WL

A

A A

C

C

C

E F E

Figure 6.4. (a) Synchrotron XRD patterns for SW, SL, PM, WW and WL. (b)

Enlargements of part of the SXRD patterns (F=fairchildite (K2Ca(CO3)2),

N=nesquehonite (MgCO3.H2O), C=calcite (CaCO3), S=sylvite (KCl), A=apatite

(Ca5(PO4)3(OH), L=lime (CaO), E=scolecite (CaAl2Si3O10.3(H2O), Q=quartz (SiO2),

Po=portlandite (Ca(OH)2 and P=periclase (MgO).

98

a)

5 15 25 35 2 Theta (deg)

E

C

E

RW

A

S

A

A

C

N

C

E

S

Po

C Q

Po

C

C

P

E

C

Q A

S S

C

P

RL

GT

JL

HH

Q

E

A P

Po

P

C

C C C C

A

F

A

N

L E

A Q Q

C S C

C

C C S

Q

F

b)

14 24 2 Theta (deg)

C

RW

A

S C

E Po

Q

L

C

P E

C

W

S C

RL

GT

JL

HH

L

P

A

A

A

N

E

C

Po

Figure 6.5. (a) Synchrotron XRD patterns for RW, RL, GT, JL, and HH. (b)

Enlargements of parts of the SXRD patterns (F=fairchildite (K2Ca(CO3)2),

N=nesquehonite (MgCO3.H2O), C=calcite (CaCO3), S=sylvite (KCl), A=apatite

(Ca5(PO4)3(OH)), L=lime (CaO), E=scolecite (CaAl2Si3O10.3(H2O)), Q=quartz

(SiO2), Po=portlandite (Ca(OH)2 and P=periclase (MgO).

99

SW

SW enlarged grain

SW analysed grain

1 3 5 7 Energy (keV)

Ca

K Mg

P

Mg = 7.5% P = 0.96% K = 4.7% Ca = 19.18%

SL

SL enlarged grain

SL analysed grain

1 3 5 7 Energy (keV)

Cl Ca

K

Mg P

Mg = 3.45% P = 2.48% Cl = 7.58% K = 24.28% Ca = 18.87%

PM

PM enlarged grain

PM analysed grain

1 3 5 7 Energy (keV)

Ca K

Cl

P

Mg Mg = 17.61% Al = 0.18% Si = 0.25% P = 2.76% Cl = 7.37% K = 11.9% Ca = 10.87%

Al Si

WW

WW enlarged grain

WW analysed grain

1 3 5 7 Energy (keV)

Ca K

Cl Si

Mg

P

Mg = 12.02% Si = 0.59% P = 0.83% Cl = 0.22% K = 0.64% Ca = 4.19%

WL

WL enlarged grain

WL analysed grain

1 3 5 7 Energy (keV)

Ca K

Cl S P

Si

Mg

Mg = 12.95% Si = 0.94% P = 2.02% Cl = 4.24% K = 6.55% Ca = 9.61%

Figure 6.6. Scanning electron micrograph (SEM) and X-ray spectra of the indicated ash

particles for SW, SL, PM, WW and WL ash. Calcite crystals (CaCO3), microcrystalline

hydroxyl apatite and mixed potassium, magnesium and calcium salts are present.

100

RW

RW enlarged grain

RW analysed grain

1 3 5 7 Energy (keV)

Ca K

Cl P

Mg

Mg = 5.74% P = 1.07% Cl = 0.51% K = 8.02% Ca = 7.57%

RL

RL enlarged grain

RL analysed grain

1 3 5 7 Energy (keV)

Ca

K

Cl P Mg

Mg = 6.22% P = 3.22% Cl = 2.8% K = 17.3% Ca = 12.9%

GT

GT enlarged grain

GT analysed grain

1 3 5 7 Energy (keV)

Ca

K

Cl P Si

Mg

Mg = 5.72% Si = 1.76% P = 2.95% Cl = 4.28% K = 17.62% Ca = 12.86%

JL

JL enlarged grain

JL analysed grain

1 3 5 7 Energy (keV)

Ca

K

Cl P

Si Al

Mg

Mg = 4.45% Al = 6.34% Si = 3.24% P = 1.36% Cl = 0.67% K = 5.6% Ca = 14.28%

HH

HH enlarged grain

HH analysed grain

1 3 5 7 Energy (keV)

Fe

Ca K

Cl P

Si

Al Al = 10.67% Si = 3.23% P = 0.6% Cl = 2.79% K = 6.41% Ca = 8.67% Fe = 4.27%

Figure 6.7. Scanning electron micrograph (SEM) and X-ray spectra of the indicated ash

particles for RW, RL, GT, JL, and HH ash. Calcite crystals (CaCO3), microcrystalline hydroxyl

apatite and mixed potassium, magnesium and calcium salts are present.

101

a)

Na

Mg

Al

Si

P

Cl

K

Ca

Ti Fe

-1.0 -0.5 0.0 0.5 1.0

Factor 1 : 33.99%

-1.0

-0.5

0.0

0.5

1.0

Facto

r 2 :

21.9

1%

Na

Mg

Al

Si

P

Cl

K

Ca

Ti Fe

Calcite

Possible Soil Contamination

Other Salts(Ca also present but minor)

b)

SW SL PM GT RL JL HH WL RW WW-8 -6 -4 -2 0 2 4 6 8

Factor 1: 33.99%

-8

-6

-4

-2

0

2

4

6

8

Fact

or 2

: 21.

91%

Figure 6.8. Principal component analyses of log SEM EDS elemental analyses

results for native plant ash particles. Variables (a) and plant material (b). Silver

wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM),

wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL),

grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

102

0

1

2

3

4

5

6

7

8

9

10

0 1 5

pH

Incubation Time (week)

pH Unburnt (a)

Control SW SL PM WW WL RW RL GT JL HH

0

20

40

60

80

100

120

140

0 1 5

EC (μ

S/cm

)

Incubation Time (week)

EC Unburnt (b)

Control SW SL PM WW WL RW RL GT JL HH

-10

0

10

20

30

40

50

0 1 5

Bic P

(mg/k

g)

Incubation Time (week)

Bic P Unburnt (c)

Control SW SL PM WW WL RW RL GT JL HH

Figure 6.9. Mean (n=3) values with standard error bars of pH (a), EC (b) and bicarbonate P (c) for unburnt soil incubated with native plant ash. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

103

0

1

2

3

4

5

6

7

8

9

10

0 1 5

pH

Incubation Time (week)

pH Burnt (a)

Control SW SL PM WW WL RW RL GT JLA HH

0

20

40

60

80

100

120

140

160

0 1 5

EC (μ

S/cm)

Incubation Time (week)

EC Burnt (b)

Control SW SL PM WW WL RW RL GT JL HH

-10

0

10

20

30

40

50

60

0 1 5

Bic P

(mg/k

g)

Incubation Time (week)

Bic P Burnt (c)

Control SW SL PM WW WL RW RL GT JL HH

Figure 6.10. Mean (n=3) values with standard error bars of pH (a), EC (b) and bicarbonate P (c) for burnt soil incubated with native plant ash. Silver wattle wood (SW), silver wattle leaf (SL), prickly moses leaf and twig (PM), wandoo wood (WW), wandoo leaf (WL), red gum wood (RW), red gum leaf (RL), grass tree leaf (GT), jarrah leaf (JL), and harsh hakea leaf and twig (HH).

104

Chapter 7

7.0 General Summary, Limitations and Future Work

7.1 Introduction

A main objective of the study as stated in the general introduction was to

investigated changes in soil and ash properties due to heating: these properties

include chemistry, mineralogy and morphology and include changes that may affect

plant growth in forests after a fire has occured. Several studies were conducted

including:

• dehydration (dehydroxylation) and subsequent rehydration of the three soil

minerals (gibbsite, kaolinite and goethite) that are most commonly and

severely affected by heating during natural fires. Significant changes in

mineral properties were discovered and the environmental significance of

these changes has been considered.

• a glasshouse study assessing the impacts of soil heating on the availability of

native and added phosphate for a lateritic soil.

• a field study of soil mineralogy and chemistry immediately after a bushfire at

Wundowie, Darling Range, Western Australia

• a thorough chemical, mineralogical and morphological investigation of ash

derived from several Australian native plant species. The reactions of ash

with unheated and heated soil were also considered.

Many studies have been conducted on the effect of fire on soil properties and have

focussed variously on nutrient availability, pH and EC, organic matter content,

texture, structural stability, cation exchange capacity, water repellence and soil

erosion immediately after fire (Raison et al., 1979; Khanna and Raison, 1986;

Etiegni and Campbell, 1991; Ulery et al., 1993; Iglesias et al., 1997; Neary et al.,

1999; Goforth et al., 2005; Cerdá and Doerr, 2008; Galang et al., 2010; Pereira et al.,

2011). There have been few prior studies on the chemistry, mineralogy and

morphology of heated soil and plant ash which are addressed in this thesis. Past

studies highlight the complexity of the effects of fire on soil properties but have not

focussed on the detail of the processes that change morphological, mineralogical and

chemical properties of soil and consequently the growth of plants, land stability, etc.

105

The present research has partly adressed this deficiency and provides a deeper

understanding on the complex relationships between ash and soil and how fire

affects the availability of several nutrients to plants.

7.2 General Summary 7.2.1. Dehydroxylation and rehydroxylation of soil minerals A comprehensive study on three soil minerals (kaolinite, goethite and gibbsite)

commonly affected by heating was presented in Chapter 3. It evaluates

dehydroxylation and rehydration after dehydroxylation of these soil minerals. A flow

chart of the study is given in Figure 7.1. The minerals experienced alteration in

structure and composition when they heated to their dehydroxylation temperature.

Kaolinite transformed into metakaolinite when heated between 500- 600oC, goethite

altered into poorly ordered hematite when heated at 300-350oC. Heating at 350oC

transformed gibbsite into boehmite and an amorphous phase.

Several published studies have investigated the reversion of metakaolinite to

kaolinite or the rehydroxylation of goethite and gibbsite under laboratory conditions

and only a few of them have been conducted under conditions resembling natural

soil conditions. Hydrothermal treatment of dehydroxylated minerals at 55 and 95oC

for 0, 14, 70, 200 and 400 days was aimed at inducing crystal regrowth in

dehydroxylated minerals. The results of the rehydroxylation treatments for the three

dehydroxylated minerals examined in this research indicate that this process does

occur to various extents and thus may occur in natural environments.

Dehydroxylation of kaolinite, goethite and gibbsite caused slight to moderate

increases in SSA and P sorption maximum. Rehydroxylation of 350oC heated

gibbsite was extensive during hydrothermal treatment at 95oC and after 14 days

boehmite, bayerite and gibbsite had formed, the process was slower at 55oC. Heated

goethite rehydroxylated with a considerable proportion of this water being lost at

265oC (same temperature as goethite) but no goethite was observed by XRD.

Similarly, metakaolinite showed no change in either XRD or SXRD patterns but

bound-water content increased and SSA and P sorption capacity were variously

affected by rehydroxylation.

106

• 

• 

• 

Figure 7.1 Flow chart for the dehydration and rehydration of kaolinite, gibbsite and

goethite.

7.2.2. Effects of heating soil on the availability of P to plants

The soil used in the glasshouse experiment is a lateritic podzolic which has a fine

fraction dominated by three readily dehydroxylated minerals (kaolinite, gibbsite and

goethite) and consequently is particularly sensitive to temperatures that commonly

occur in the surface of soils heated by bushfires. A focus of the study was on the

impact of heating soil minerals on phosphate (P) availability to plants. Soil from the

topsoil of the lateritic podzolic from a virgin forest site was heated at 250, 350 and

500oC which are temperatures that may be experienced by topsoil during bushfires

(Figure 7.2). The P-response of annual ryegrass (Lolium rigidum Gaud) to the

heating of the soil and to the application of several levels of P as monocalcium

phosphate (MCP) was determined for three succesive harvests.

Heating caused kaolinite, gibbsite and goethite to dehydroxylate and to partly alter

into metakaolinite, amorphous alumina and hematite respectively. Heating increased

soil pH and EC due to combustion of soil organic matter to produce salts, although

EC then decreased for 350oC and 500oC heating as soluble salts chemically reacted

with soil constituents at these temperatures. Yield of ryegrass decreased with

107

increasing temperature of heating for unfertilized soil and for heated soils supplied

with P fertilizer. The P concentration in ryegrass for each of three harvests ranged

from 0.03% to 0.3% and decreased in the same sequence as for yield (i.e unheated

soil>250oC>350oC>500oC heated soil). Heating the soil increased Bic-P and had

little effect on the retention of added P as indicated by bicarbonate extraction.

Heating the soil at 500oC greatly increased amounts of reactive Si and Al in the soil

due to the formation of metakaolinite and amorphous alumina, which are highly

reactive compounds, which are presumably responsible for the reduced effectiveness

of the added P fertilizers. Clearly heating of soil by bushfires can reduce the

availability to plants of native and added soil phosphate and forest managers should

be aware of this process.

• 

• • 

Figure 7.2. Flow chart for the glasshouse study conducted on heated lateritic soil.

7.2.3. The properties of soil heated in bushfire

Heating of soil during a fire, which burns abundant surface litter and fallen timber,

can raise surface soil temperature to 500oC or more and may significantly impact on

soil properties. Several soil minerals are likely to be affected by heating in natural

and managed fires. The effects of soil heating on the mineralogy of soil have not

been extensively studied although quite modest (250oC and above) temperatures

108

affect minerals. Kaolinite decomposes at temperatures between 450oC and 600oC

losing lattice water and forms metakaolinite. Gibbsite alters to an amorphous phase

and boehmite on heating at ≈200oC, and goethite is transformed to a disordered

mineral known as hydro-hematite at ≈300oC. The possibility that these

dehydroxylated compounds persist in soils heated in bush fires has received little

attention and this investigation was aimed at providing additional understanding of

the behaviour of hydroxylated soil minerals heated in a bushfire.

Samples were collected the day after a bushfire at Wundowie in the Darling Range,

Western Australia in early March 2009. Samples were removed as a 1cm thick soil

layer from under burnt Eucalyptus and grass tree (Xanthorrhoea pressii) logs from

up-slope and down slope sites in a lateritic colluvium catena. The soils are very

gravelly and were sieved to obtain the <2mm fraction for analysis (Figure 7.3).

Conventional and synchrotron XRD patterns of heated and unheated soil samples

from the Wundowie bushfire site show the effect of fire on soil minerals. The main

crystalline compounds of unheated soil are quartz, kaolinite, gibbsite and goethite.

The XRD patterns of heated soil show that kaolinite had dehydroxylated to form

metakaolinite, gibbsite had altered into an amorphous phase, while goethite had

transformed into hematite (hydrohematite) but quartz was unaltered. The bushfire

added calcite in plant ash to the soil. The addition of ash during the fire considerably

increased the pH of all soil samples. Soil EC values for the heated soil also being

considerably higher than for unburnt soil. The increases in EC simply reflect the

addition to the soil of soluble salts in plant ash. Heating had increased amounts of

extractable Al, Fe and Si due to crystalline minerals becoming amorphous or poorly

ordered. Clearly dehydroxylated minerals and possibly their rehydroxylated forms

must be present in naturally heated soils and may exert significant effects on the

chemical behaviour of the soil.

109

• 

• • 

• 

• 

Figure 7.3. Flow chart for the field study of soil minerals heated in a bushfire at

Wundowie, Darling Range, Western Australia.

7.2.4. The composition of plant ash

Wildfires and prescribed fires burn many thousand of hectares of forest annually and

generate abundant amounts of ash. This material is generally alkaline with pH

ranging from 9.0 to 13.5 and may consist of diverse compounds containing

substantial amounts of plant nutrients.

Ash was created by burning various plant materials from several Australian native

plant species growing on lateritic soil at Bakers Hill (near Wundowie, WA): silver

wattle (Acacia retinodes), prickly moses (Acacia pulchella), wandoo/white gum

(Eucalyptus wandoo), red gum or marri (Corymbia calophylla), grass tree

(Xanthorrhoea pressie) jarrah (Eucalyptus marginata), and harsh hakea (Hakea

prostrata). A flow chart for the study is in Figure 7.4

Ash percentage ranges from 1.7 to 4.7% for the complete combustion of the plant

materials (leaf, twig, wood, etc). Specific surface area of the ashes ranges from 2 to 9

m2/g indicating that the particles of ash (mostly crystals) are very small (micron size)

and consequently they are likely to be quite reactive in soil. The ashes are alkaline,

pH ranging from 12.3 to 13.8. Available phosphorus (Colwell–P) in the ashes ranged

110

from 30 to 199 mg/kg and EC values were high for all the ashes, indicating that the

ashes contain substantial amounts of soluble salts. Minerals present in the ashes

include calcite (CaCO3), fairchildite (K2Ca(CO3)2),, nesquehonite (MgCO3.H2O),

sylvite (KCl), lime (CaO), scolecite (CaAl2Si3O10.3(H2O)), quartz (SiO2) (derived

from dust), portlandite (Ca(OH)2), periclase (MgO), and apatite, probably

resembling hydroxyl-apatite (Ca5(PO4)3(OH)) and wilkeite (Ca5((P, S,

Si)O4)3(OH,CO3)). The amount of carbon was above 40% for all the raw plant

materials and significantly reduced to below 15% for their ashes where carbon was

mostly present as carbonates. Burning plants resulted in loss of most nitrogen with

low nitrogen concentrations in ash.

Scanning electron microscopy and associated X-ray spectra of particles in the ashes

show the diverse particle sizes, shapes and compositions present in these materials.

Apart from distinct calcite crystals, most grains seen in the micrographs are K and

Mg-rich aggregates often of sub-micron-size particles (the above mentioned

minerals) containing alkali elements and chlorine and commonly also containing a

little P, Fe and S. These results indicate the complex nature of ash, with a high

diversity in composition, shape and size of particles.

Addition of ash to heated and unheated soil increased the pH and EC values due to

dissolution of the salts in the ash. The different pH and EC values for each ash can

be explained by the diverse amounts and occurences of carbonates and other salts in

ash. The increased amount of plant available phosphate in soil to which ash has been

added is at least partly due to dissolution of the apatite and possibly wilkeite present

in ash.

111

• 

• 

Figure 7.4. Flow chart for the Australian native plants ash study.

7.3 Limitations to this research and suggested future work

This research has contributed to understanding relationships between the chemical,

mineralogical and morphological properties of heated soil, the nature of ash and P

availability to plants for soils affected by bushfire. The research has shown how a

knowledge of soil and ash materials may allow us to explain and predict some

effects of fire on soils. However, there are several limitations to this work and

further research is needed, as follows:

1) The study of soil materials focused on laboratory, glasshouse and field work

shortly after a bushfire or laboratory heating; there is a need to extend the

study to several years after the bushfire or heating to determine the

persistence of the various effects on soils and minerals described in this

thesis.

2) Kaolinite, goethite and gibbsite in natural soil have been shown to be affected

by heating during a bushfire. Additional soil minerals may also be sensitive

to heating such as the hydroxylated and hydrated minerals: allophane,

halloysite, lepidicrocite and gypsum. The various soil carbonate minerals

(calcite, dolomite, magnesite) may lose carbon dioxide if heated to a

sufficient temperatures (>800oC)

112

3) Minerals in ash were studied using chemical analysis, XRD, SXRD and

SEM. Because particles of ash are often very small (<1μm) and they occur in

aggregates, further work is required to characterise single crystals in ash.

This can be achieved using a combination of TEM EDS, electron

diffraction/lattice imaging and EELS although specimen preparation of these

materials for TEM will be difficult.

4) Ryegrass was used in the glasshouse experiment to assess the impacts of

addition of ash on phosphate uptake by plants from soil. In the field, the fate

of plant nutrients in ash is likely to be more complex than simulated in this

experiment and will reflect several interacting properties and processes in the

ash-soil-plant continuum. For example, the availability of element (e.g. Ca)

in ash to plants will depend (inter alia) on:

• the chemical form of Ca (which minerals) in the ash

• the particle size of these minerals

• the extent to which these minerals are protected from dissolution by

the presence of other ash constituents through occlusion, elevated pH

due to carbonates, etc

• the reaction of dissolved Ca (possibly as a complex with CO3 or other

anion) with reactive heated minerals (e.g. adsorption and possibly

specific sorption (chemisorption)

• the location of Ca in ash at the soil surface where direct contact with

subsurface roots and their exudates will be limited.

113

Chapter 8 8.0 Publications from this thesis

8.1. Conferences Publications

Yusiharni, E., Gilkes, R.J., 2009. Rehydroxylation of heated gibbsite, kaolinite and

goethite: an assesment of properties and environmental significance. Proceedings,

XIV International Clay Conference, June 14-20, 2009 Castellaneta Marina, Italy.

pp:545-545

Yusiharni, E., Gilkes, R.J. 2009. Investigating the effects of bushfire on soil minerals

using thermal analysis. Proceedings, The Asian Conference on Thermal Analysis and

Applications, December 17-18, 2009 Bangkok, Thailand. pp: 115-119

Yusiharni, E., Gilkes, R.J., 2010. Do soil minerals recover after they are damaged by

bushfires? Proceedings, The 19th World Congress of Soil Science (WCSS) August

1-6, 2010, Brisbane, Queensland, Australia. pp:104-107

Yusiharni, E., Gilkes, R.J., 2010. Do dehydroxylated gibbsite, kaolinite and goethite

rehydroxylate? Proceedings, Australian Clay Minerals Society’s 21st Conference

August 7-8, 2010, Brisbane, Queensland, Australia. pp: 131-134

Gilkes, R.J., Yusiharni, E., 2011. The effects of heating a lateritic podzolic soil on

soil phosphate availability, a glasshouse study? Proceedings, III International

Meeting of Fire Effects on Soil Properties, March 15-19, 2011, University of

Minho, Guimaraes, Portugal. pp: 44-44

Yusiharni, E., Gilkes, R.J., 2011. Mineralogical and chemical changes in a lateritic

soil due to a bushfire in the Darling Range, Western Australia. Proceeding, III

International Meeting of Fire Effects on Soil Properties (15-19 March 2011),

University of Minho, Guimaraes, Portugal. pp: 49-49

114

8.2. Journal Publications

Yusiharni, E., Gilkes, R.J., 2012. Rehydration of heated gibbsite, kaolinite and

goethite: an assessment of properties and environmental significance. Applied Clay

Science (In press, Corrected Proof, Available online 25 January 2012 ).

http://dx.doi.org/10.1016/j.clay.2011.12.005

Yusiharni, E., Gilkes, R.J., 2012. Short term effects of heating a lateritic podzolic

soil on the availability to plants of native and added phosphate. Geoderma (In press,

Corrected Proof, Available online 1 February 2012).

http://dx.doi.org/10.1016/j.geoderma.2012.01.002

Yusiharni, E., Gilkes, R.J., 2012. Changes in the mineralogy of a lateritic soil due to

a bushfire at Wundowie, Darling Range, Western Australia. Geoderma (In press,

Corrected Proof, Available online 2 March 2012).

http://dx.doi.org/10.1016/j.geoderma.2012.01.030

Yusiharni, E., Gilkes, R.J., 2012. Minerals in the ash of Australian native plants.

Geoderma (In Press).

8.3. Other Publications

Yusiharni, B. E., H. Ziadi., and Gilkes, R. J., 2007. A laboratory and glasshouse

evaluation of the byproducts; chicken litter ash, wood ash and iron smelting slag as

liming agents and P fertilizers. Aust. J. of Soil. Res. 45(5) 374–389.

Kuligowski, K., Gilkes, R.J., Poulsen, T.G., Yusiharni, B.E., 2010. The composition

and dissolution in citric extractants of ash from the thermal gasification of pig

manure. Chemical Engineering Journal 163, 1–9.

Kuligowski, K., Gilkes, R.J., Poulsen, T.G., Yusiharni, B.E., 2012. Ash from the

thermal gasification of pig manure - effects on ryegrass yield, element uptake and

soil properties. Soil Res. (In Press).

115

References

Anand, R.R., Gilkes, R.J., 1987. The association of maghemite and corundum in Darling Range laterites, Western Australia. Aust. J. Soil Res. 35, 303-311. Association of Official Agricultural Chemists, official methods of analysis, 15th ed., AOAC, Virginia, 1990. Atanassova, I., Teoharov, M., Ivanov, P., 2009. Characterisation of a fire affected catena sequence from Lyulin Mountain Bulgaria. Bulg. J. Agric. Sci. 15, 333-340. Australian Climate Extreme Fires, 2008. Fire. Accessed online at http://www.bom.gov.au/climate/c20thc/fire.shtml. 4 May 2008. Babushkin, V.I., Matveyev, G.W., Mchedlov-Petrossyan, O.P., 1985. Thermodynamics of Silicates. Springer, Berlin, pp. 165-167. Badĭa, D., Marti, C., 2003. Plant ash and heat intensity effects on chemical and physical properties of two contrasting soils. Arid Land Res. Manage. 17, 23-41. Barrow, N.J., 1978. The description of phosphate adsorption curves. J. Soil Sci. 29, 447-462. Bauhus, J., Khanna, P.K., Raison, R.J., 1993. The effect of fire on carbon and nitrogen mineralization and nitrification in an Australian forest soil. Aust. J. Soil Res. 31, 621–639. Binkley, D., D, R., David, M. B., Caldwell, B., 1992. Soil chemistry in a loblolly/longleaf pine forest with interval burning. Ecol. Applic. 2, 157-164. Bodí, M.B., Mataix-Solera, J., Doerr, S.H., Cerdá, A., 2011. The wettability of ash from burned vegetation and its relationship to Mediterranean plant species type, burn severity and total organic carbon content. Geoderma 160, 599-607. Bokhimi, X., Sanchez-Valente, J., Pedraza, F., 2002. Crystallization of sel-gel boehmite via hydrothermal annealing. J. Solid State Chem. 166: 182-190. Brown, J. K., Smith J.K., 2000. Wildland fire in ecosystems: effects of fire on flora. Gen. Tech. Rep. RMRS_GTR-42. Ogden, Utah, US. Department of Agriculture, Forest service, Rocky Mountain Station: 257pp. Bureau of Meteorology, 2011. http://www.bom.gov.au/wa. Burrows, N.D., 2008. Linking fire ecology and fire management in south-west Australian forest landscapes. Forest Fire Manage. 255, 2394-2406 Byram, G.M., 1959. Combustion of forest fuels. In: Forest fire: Control and use. K. P. Davis (Ed). New York, McGraw-Hill: 61-89.

116

Cadwell, T. G, Johnson, D. W., Miller, W. W., Qualls, R.G., 2002. Forest floor carbon and nitrogen losses due to prescription fire. Soil Sci. Soc. Am. J. 66, 262-267. Cerdá, A., Doerr, S.H., 2008. The effect of ash and needle cover on surface runoff and erosion in the immediate post-fire period. Catena 74, 256-263. Certini, G., 2005. Effects of fire on properties of forest soils: a review. Oecologia 143, 1-10. Chandler, C., Cheney, P., Thomas, L., Trabaud, L., Willimas, D., 1983. Fire in forestry. Vol. 1. Forest fire behaviour and effects. John Wiley & Sons, New York. Cheney, N.P., 1981. Fire behaviour. In: Fire and the Australian biota. A. M. Gill, R. H. Groves and I. R. Noble (Eds). Canberra, Australian Academy of Science. Choromanska, U., DeLuca, T.H., 2001. Prescribed fire alters the impact of wildfire on soil biochemical properties in a Ponderosa Pine Forest. Soil Sci. Soc. Am. J. 65, 232-238. Christensen, N.L., 1994. The effects of fire on physical and chemical properties of soils in Mediterranean climate shrublands. In. The role of fire in Mediterranean type ecosystems. J. M. Moreno and W. C. Oechel (Eds). New York, USA, Springer-Verlag: 79-95. Clapham, W. M., and Zibilske, L. M., 1992, Wood ash as a liming amendment, Commun., Soil Sci. Plant Anal. 23, 1209-1227. Colwell, J.D., 1963. The estimation of the phosphorus fertilizer requirements of wheat in southern New South Wales by soil analysis. Aust. J. Exp. Agric. Animal Husbandry. 3, 190-197. Cornell, R.M., Schwertmann, U., 1996. The iron oxides: Structure, properties, reactions, occurence and uses. Berlin, VCH Weinheim. Cotton, F.A., Wilkinson, G., 1988. Advanced inorganic chemistry. 5th edn. Wiley, New York. DeBano, L.F., Klopatek, J.M., 1988. Phosphorus dynamics of Pinyon-Juniper soils following simulated burning. Soil Sci. Soc. Am. J. 52, 271-277. De Faria, D.L.A., Lopes, F.N., 2007. Heated goethite and natural hematite: Can Raman spectroscopy be used to differentiate them? Vibrational Spectroscopy. 45, 117-121. Demeyer, A., Voundi Nkana, J.C., Verloo, M.G., 2001. Characteristics of wood ash and influence on soil properties and nutrient uptake: an overview. Bioresour. Technol. 77, 287-295.

117

Durán, J., Rodrĭguez, A., Fernández-Palacios, J.M., Gallardo, A. Changes in soil N and P availability in a Pinus canariensis fire chronosequence. For. Ecol. Manag. 256, 384-387. Eckmeier, E., Egli, M., Schmidt, M.W.I., Schlumpf, N., Nötzli, M., Minikus-Stary, N., Hagedorn, F., 2010. Preservation of fire-derived carbon compounds and sorptive stabilisation promote the accumulation of organic matter in black soils of the Southern Alps. Geoderma, 159, 147-155. Eggleton, R. A., Taylor, G. G., 2008. Impact of fire on the Weipa Bauxite, northern Australia. Australian Journal of Earth Sciences, 55, S83-S86. Etiegni, L., Campbell, A.G., 1991. Physical and chemical characteristics of wood ash. Bioresour. Technol. 37, 173-178. Etiegni, L., Campbell, A.G., Mahler, R.L., 1991. Evaluation of wood ash disposal on agricultural land. I. Potential as a soil additive and liming agent, Commun. Soil Sci. Plant Anal, 22, 243-256. Fan, H., Song, B., Li, Q., 2006. Thermal behavior of goethite during transformation to hematite. Materials Chemistry and Physics 98, 148-153. FAO, 2001. Global forest fire assesment 1999-2000. Rome, Food and Agriculture Organization of the United Nations, Forestry Department, Forest Resources Assessment Programme 55. Fernandez, I., Cabaneiro, A., Carballas, T., 1997. Organic matter changes immediately after a wildfire in an Atlantic forest soil: comparison with laboratory soil heating. Soil Biol. Biochem. 29, 1-11. Ferran, A., Delitti, W., Vallejo, V.R., 2005. Effects of fire reccurence in Quercus coccifera L. shrublands of the Valencia Region (Spain): II. plant and soil nutrients. Plant Ecology 177, 71-83. Ferran, A., Serrasolsas, I., and Valejo, V. R., 1991. Soil evolution after fire in Quercus ilex and Pinus halepensis forests. In A. Teller (Ed). Responses of forest ecosystems to environmental changes. Elsevier Applied Science, London. Flinn, D. W., Farrell, P.W., Stewart, H.T.L., Leitch, C.J., Hopmans, P., 1984. The effects of fire in eucalyptus forests on soils, nutrient cycling, tree growth, and catchment hydrology: A review with particular reference to fuel reduction burning. Res. Bran. Rep. 44pp. FRA, 2000. Global forest fire assessment. Forest Resources Assessment. Rome, Forestry Department, Food and Agriculture Organization of the United nations. FRA, 2010. Global forest fire assessment. Forest Resources Assessment. Main Report, Forestry Department, Food and Agriculture Organization of the United Nations paper 163, 340 pp.

118

Freund, F., 1974. Ceramics and thermal transformations of minerals. In: Farmer, V.C. (Ed), The Infrared Spectra of Minerals. Vol. Monograph 4. Mineralogical Society London, pp. 465-482. Frost, R.L., Horvarth, E., Mako, E., Kristof, J., Redey, A., 2003. Slow transformation of mechanically dehydroxylated kaolinite to kaolinite- an aged mechanochemically activated formamide-intercalated kaolinite study. Thermoch. Acta 408, 103-113. Gabet, E.J., Bookter, A., 2011. Physical, chemical and hydrological properties of Ponderosa pine ash. Int. J. Wildland Fire 20, 443-452. Gabet, E.J., Sternberg, P., 2008. The effects of vegetative ash on infiltration capacity, sediment transport, and the generation of progressively bulked debris flows. Geomorphology 101, 666-673. Galang, M.A., Markewitz, D., Morris, L.A., 2010. Soil phosphorous transformations under forest burning and laboratory heat treatments. Geoderma. 155, 1959-1970. Ghuman, B.S., Lal, R., 1989. Soil temperature effects of biomass burning in windrows after clearing a tropical rainforest. Field Crops Res. 22, 1-10. Giardina, C.P., Sandford Jr., R.L., Dockersmith, I.C., 2000. Changes in soil phosphorus and nitrogen during slash-and-burn clearing of a dry tropical forest. Soil.Sci.Amer.J. 64, 399-405. Gill, A.M., Moore, P.H.R., 1990. Fire intensities in eucalypt forests of south eastern Australia. International Conference on Forest Fire Research, Coimbra, Portugal. Goforth, B.R., Graham, R.C., Hubbert, K.R., Zanner, C.W., Minnich, R.A., 2005. Spatial distribution and properties of ash and thermally altered soils after high severity forest fire, southern California. Int. J. Wildland Fire 14, 343-354. González-Pérez, J.A., González-Vila, F.J., Almendros, G., Knicker H., 2004. The effect of fire on soil organic matter-a review. Environ. Int. 30, 855-870. Gonzalez, G., Sagarzazu, A., and Vilalba, R. (2000). Study of mechano-chemical transformation of goethite to hematite by TEM and XRD. Materials Research Bulletin, 35, 2295-2308 Grim, R., 1968.Clay mineralogy. New York, McGraw-Hill. Grim, R.E., Bradley, W.F., 1948. Rehydration and dehydration of the clay minerals: American Mineralogist 33, 50-59. Harper, R.J., Gilkes, R.J., Robson, A.D., 1982. Biocrystallization of quartz and calcium phosphates in plants- a re-examination of the evidence. Aust. J. Agric. Res 33, 565-571.

119

Haynes, K., Tibbits, A., Coates, L, Ganewatta, G., Handmer, J., McAneney, J., 2008. 100 years of Australian Civilian bushfire fatalities: exploring the trends in relation to the ’stay or go’ policy. Report prepared for the Bushfire Co-operative Research Centre, Risk Frontiers, Sydney, Australia. Holden, J., 2005. An introduction to physical geography and the environment. Pearson Education Limited. Essex. Humphreys, G.S., Hunt, P.A., Buchanan, R., 1987. Wood ash stone near Sydney, N. S. W.: A carbonate pedological feature in an acidic soil. Aust J. Soil Research 25, 115-124. Iglesias, T. Cala, V., González, J. 1997. Mineralogical and chemical modifications in soils affected by a forest fire in the Mediterranean area. Sci. of the Tot. Environ. 204, 89-96. Jasper, R., 1999. The changing direction of land managers in reducing the threat from major bushfires on the urban interface of Sydney. Australian Bushfire Conference. Albury, NSW, Bushfire 99: 175-185. Johnson, D.W., Curtis, P.S., 2001. Effects of forest management on soil C and N storage: meta analysis. For. Ecol. Manag. 140, 227-238. Karlikowski, T., 1982. Forest fire detection system. Proceedings of an International Seminar organized by the Timber Committee of the Unite Nations Economic Commission for Europe, warsaw, Poland, Martinus Nijhoff/Dr W. Junk Publisher. Kauffman, J.B., Sanford, R.L., Cummings, D.L., Salcedo, I.H., Sampaio, E.V.S.B., 1993. Biomass and nutrient dynamics associated with slash fires in Neotropical dry forests. Ecology 74, 140-151. Keeley, J.E., 2009. Fire intensity, fire severity and burn severity: a brief review and suggested usage. Int. J. Wildland Fire 18, 116-126. Ketterings, Q. M., T. T. Wibowo, Noordwijk, M.V., Penot, E., 1999. Farmers perspectives on slash and burn as a land clearing method for small scale rubber producers in Sepunggur, jambi Province, Sumatra, Indonesia. For. Ecol. Manage. 120, 158-169. Ketterings, Q.M., Bigham, J.M., Valerie, L., 2000. Changes in soil mineralogy and texture caused by slash and burn fires in Sumatra Indonesia. Soil Sci. Soc. Amer. J. 64, 1108-1117. Ketterings, Q.M., Noordwijk, M.V., Bigham, J.M., 2002. Soil phosphorous availability after slash-and burn fires of different intensities in rubber agroforests in Sumatra, Indonesia. Agric. Eco. Environ. 92, 37-48. Khanna, P.K., Raison, R.J., 1986. Effect of fire intensity on solution chemistry of surface soil under a Eucalyptus pauciflora forest. Aust. J. Soi Res. 24, 423-434.

120

Khanna, P.K., Raison, R.J., Falkiner, R.A., 1994. Chemical properties of ash derived from Eucalyptus litter and its effects on forest soils. For. Ecol. Manage. 66, 107-125. Kleber, M., Mikutta, C., Jahn, R., 2004. Andosols in Germany- pedogenesis and properties. Catena. 56, 67-83. Kloprogge, J.T., Ruan, H.D., Frost, R.L., 2001. Thermal decomposition of bauxite minerals: infrared emission spectroscopy of gibbsite, boehmite and diaspore. J. Mat. Sci. 37, 1121-1129. Knicker, H., 2007. How does fire affect the nature and stability of soil organic nitrogen and carbon? A review. Biogeochemistry 85, 91-118. Kutiel, P., Naveh, Z., 1987. The effect of fire on nutrients in a pine forest soil. Plant Soil. 104, 269-274. Kutiel, P., Shaviv, A., 1989. Changes of soil N-P status in a laboratory simulated forest fire. Plant Soil. 120, 57-63. Kutiel, P., Shaviv, A., 1992. Effects of soil type, plant composition and leaching on soil nutrients following a simulated forest fire. For. Ecol. Manag. 53, 329-343. Kwari, J.D., Batey, T., 1991. Effect of heating on phosphate sorption and availability in some northeast Nigerian soils. Euro. J. Soil Sci. 52, 97-102. Landers, M., Gilkes, R.J., 2007. Dehydroxylation and dissolution of nickeliferous goethite in New Caledonian lateritic Ni ore. App. Clay Sci. 35, 162-172. Landers, M., Gilkes, R.J., Wells, M., 2009. Dissolution kinetics of dehydroxylated nickeliferous goethite from limonitic lateritic nickel ore. Applied Clay Science 42, 615-624. Lanning, F.C., Ponnaiya, B.W.X., Crumpton, C.F., 1958. The chemical nature of silica in plants. Plant Phys. 33, 339-343. Lee, D.H., Condrate, R.A. Sr., 1995. An FTIR spectral investigation of the structural species found on alumina surfaces. Materials Letters 23, 241-246. Lentile, L.B., Holden, A.Z., Smith, A.M.S., Flakiwski, M.J., Hudak, A.T., Morgan, P., Lewis, S.A., Gessler, P.E., Benson, N.C., 2006. Remote sensing techniques to asses active fire characteristics and post-fire effects. Int. J. Wildland Fire 15, 319-354. Liodakis, S., Katsigiannis, G., Kakali, G., 2005. Ash properties of some dominant Greek forest species. Thermochimic. Acta 437, 158-176. Liodakis, S., Tsoukala, M., Katsigiannis, G., 2009. Laboratory study of leaching properties of Mediterranean forest species ashes. Water Air Soil Pollut. 203, 99-107.

121

Ludwig, B., Khanna, P., Prenzel, J., Beese, F., 2005. Heavy metal release from different ashes during serial batch tests using water and acid. Waste Manage. 25, 1055-1066. MacKenzie, K.J.D., Temujin, J., and Okada, K., 1999. Thermal decomposition of mechanically activated gibbsite. Thermo. Acta. 327:103-108. MacKenzie, R.C., 1957. The Differential Thermal Investigation of Clays. Mineralogical Society, Clay Minerals Group. London. Malmstrom, A., 2008. Temperature tolerance in soil microarthropods: Simulation of forest fire heating in the laboratory. Pedobiologia. 51, 419-426. Mataix-Solera, J., Cerda, A., Arcenegui, V., Jordán, A., Zavala, L.M., 2011. Fire effects on soil aggregation: A review. Earth-Science Rev. 109, 44-60. McArthur, W.M., 1991. Reference soils of south-western Australia. Australian Society of Soil Science, WA Branch Inc. Perth. Australia. McKeague, J.A., Brydon, J.E., Miles, N.M., 1971. Differentiation of forms of extractable iron and aluminium in soils. Soil Sci. Soc. Amer. J. 35, 8-33. Mehra, O.P., Jackson, M.L., 1960. Iron oxide removal from soils and clays by a dithionite-citrate system buffered with sodium bicarbonate. Clays Clay Miner. 7, 317-327. Misra, M.K., Ragland, K.W., Baker, A.J., 1993. Wood ash composition as a function of furnace temperature. Biomass Bioenergy 4, 103-116. Miśta, W., Wrzyszcz, J., 1999. Rehydration of transition aluminas obtained by flash calcination of gibbsite. Thermochimica Acta 331, 67-72. Munsell Colour Co., 1998. Munsell and Soil Colour Charts. Munsell Colour Co., Maryland, USA. Murphy, J., Riley, J.P., 1962. A modified single solution method for the determination of phosphate in natural waters. Analytical Chemistry Acta 27, 31-36. Myers, B., G., A., Bradstock, R., Dias, L., Duff, G., Jacklyn, P., Landsberg, J., Morrison, J., Russel-Smith, J., Williams, R., 2004. Managing for biodiversity in the rangelands In Fire Management. Australian Government Department of the Environment and Heritage, Australia, pp. 6-9. Neary, D.G., Klopatek, C.C., DeBano, L.F., Ffolliot, P.F., 1999. Fire effects on belowground sustainability: a review and synthesis. For. Ecol. Manage. 122, 51-71. Nornberg, P., Vendelboe, A.L., Gunnlaugsson, H.P., Merrison, J.P., Finster, K., Jensen, S.K., 2009. Comparison of the mineralogical effects of an experimental forest fire on a goethite/ferryhydrite soil with a topsoil that contains hematite, maghemite and goethite. Clay Min. 44, 239-247.

122

Ozanne, P.G., Shaw, T.C., 1967. Phosphate sorption by soils as a measure of the phosphate requirement for pasture growth. Aust. J. Agric. Res. 19, 601-612. Palmer, B., Gilkes, R.J., 1983. The influence of application rate on the relative effectiveness of calcined Christmas Island C-grade rock phosphate and super phosphate when applied as mixtures. Fert. Res. 4, 45-50. Pereira, J.A.M., Schwaab, M., Dell’Oro, E., Carlos Pinto, J., Monteiro, J.L.F., Henriques, C.A., 2009. The kinetics of gibbsite dissolution in NaOH. Hydrometallurgy 96, 6-13. Pereira, P., Ũbeda, X., Martin, D., Mataix-Solera, J., Guerrero, C., 2011. Effects of low severity prescribed fire on water-soluble elements is ash from a cork oak (Quercus suber) forest located in the northeast of the Iberian Peninsula. Environ. Res. 111, 237-247. Perkiömakï, J., 2004. Wood ash use in coniferous forests, a soil microbiological study into the potential risk of cadmium release, Metsantutkimuslaitoksen Tiedonantoja 917, Finish Forest Research Institute, Research Papers 917. Pomies, M.P., Morin, G., Vignaud, C., 1998. XRD study of the goethite-hematite transformation: application to the identification of heated prehistoric pigments. Euro. J. Solid State Inorg. Chem. 35, 9-25. Prasad, P.S.R., Sviva Parasad, K., Krishna Chaitanya, V., Babu, E.V.S.S.K., Sreedhar, B., Ramana Murthy, S., 2006. In situ FTIR study on the dehydration of natural goethite. J. Asian Earth Sci. 27, 503-511 Ptáček, P., Kubátova, D., Havlica, J., Brandštetr, J., Šoukal, F., Opravil, T., 2010. Isothermal kinetic analysis of the thermal decomposition of kaolinite: The thermogravimetric study. Thermochimica Acta 501, 24-29. Quintana, J.R., Cala, V., Moreno, A.M. & Parra, J.G., 2007. Effect of heating on mineral components of the soil organic horizon from a Spanish juniper, (Juniperus thurifera L.) woodland. J. Arid Environ. 71, 45-56. Rab, M. A., 1996. Soil physical and hydrological properties following logging and slash burning in the Eucalyptus regnans forest of southeastern Australia. Forest Ecol. Manage. 84, 159-176. Raison, R.J., 1979. Modification of the soil environment by vegetation fires, with particular reference to nitrogen transformations: a review. Plant Soil. 51, 73-108. Raison, R.J., Khanna, P.K., Woods, P.V., 1985. Transfer of elements to the atmosphere during low intensity prescribed fires in three Australian subalpine eucalypt forest. Can. J. For. Res. 15, 657-664. Rayment, G.E., Higginson, F.R., 1992. Australian Laboratory Handbook of Soil and Water Chemical Methods. Melbourne, Inkata Press.

123

Reuter, D.J., Robinson, J.B., 1997. Plant analysis: an interpretation manual. 2nd edn. Melbourne, CSIRO Publishing. Richardson, H.M., 1972. Phase changes which occur on heating kaolin clays. In: Brown, G. (Ed). The x-ray identification and crystal structures of clay minerals. Mineralogical Society London, pp. 132-142. Robson, A.D., Gilkes, R.J., 1981. Fertilizer responses (N, P, K, S, micronutrients) on lateritic soils in southwestern Australia, a review. In: Balkema AA, (Ed) Laterisation Processes. Rotterdam. Rocha, J., 1999. Single and triple quantum 27Al MAS NMR study of the thermal transformation of kaolinite. J. Phys. Chem. 103, 9801-9804. Rocha, J., Adams, J.M., Klinowski, J., 1990. The rehydration of metakaolinite to kaolinite: Evidence from Solid State NMR and Cognate techniques. J. Solid State Chem. 89, 260-274. Rocha, J., Klinowski, J., 1991. The rehydration of metakaolinite to kaolinite. J. Chem. Soc., Chem. Commun. 582-584. Romanyà, J., Khanna, P. K., Raison, R. J., 1994. Effects of slash burning on soil phosphorus fractions and sorption and desorption of phosphorus. For. Ecol. Manag. 65, 89-103. Rooksby, H.P., 1972. Oxides and hydroxides of aluminium and iron. In: Brown, G. (Ed). The x-ray identification and crystal structures of clay minerals. Mineralogical Society London. Roy, D.P., Boschetti, L., Maier, S.W., Smith, A.M.S., 2010. Field estimation of ash and char colour-lightness using a standard grey scale. Int. J. Wildland Fire 19, 698-704. Ruan, H.D., Frost, R.L., Kloprogge, J.T., 2001. The behaviour of hydroxyl units of synthetic goethite and its dehydroxylated product hematite. Spectrochimica Acta Part A 57, 2575-2586. Ruan, H.D., Frost, R.L., Kloprogge, J.T., Duong, L., 2002. Far-infrared spectroscopy of alumina phases. Spectrochimica Acta Part A 58, 265-272. Russel, J.D., Fraser, A.R., 1994. Infrared methods. In: Wilson, M.J. (Ed), Clay Mineralogy: Spectroscopic and Chemical determinative Methods. Chapman & Hall, London, pp. 49-52. Ryskin, Y.I., 1974. The vibrations of protons in minerals: hydroxyl, water and ammonium. In: Farmer, V.C. (Ed), The Infrared Spectra of Minerals. Vol. Monograph 4. Mineralogical Society London, pp. 137-181.

124

Saa, A., Trasar-Cepeda, M.C., Gil-Sotres, F., Carballas, T., 1993. Changes in soil phosphorus and acid phosphatase activity immediately following forest fires. Soil Biol. Biochem. 25 (9), 1223-1230. Sanchez, P.A., 1976. Properties and management of soils in the tropics. Wiley/Interscience, London. SAS Institute, 1999. SAS/STAT User’s guide. V 8. SAS Inst, Car, NC. Schwertmann, U., Taylor, R.M., 1989. Iron Oxides. In: Dixon, J.B., Weed, S.B., (Eds.), Minerals in Soil Environments, 2nd Edition pp. 379-438. Scott, A.C., 2010. Charcoal recognition, taphonomy and uses in palaeoenvironmental analysis. Palaeogeogr. Palaeclimatol. Palaeocol. 291, 11-39. Sertsu, S.M., Sanchez, P.A., 1978. Effects of heating on some changes in soil properties in relation to an Ethiopian land management practice. Soil Sci. Am. J. 42, 940-944. Shakesby, R.A., Chafer, C.J., Doerr, S.H., Blake, W.H., Humphreys, G.S., Wallbrink, P., Harrington, B.H., 2003. Fire severity, water repellency characteristics and hydrogeomorphological changes following the Christmas 2001 forest fires. Aust. Geog. 34, 147-175. Shakesby, R.A., Doerr, S.H., 2006. Wildfire as a hydrological and geomorphological agent. Earth Sci. 74, 269-307. Snars, K., Hughes, J.C., Gilkes, R.J., 2004. The effects of addition of bauxite red mud to soil on P uptake by plants. Aust. J. Agric. Res. 55:25–31. STATISTICA 10, 2011. Electronic Statistics Textbook. StatSoft, Inc., Tulsa, OK, USA. WEB: http://www.statsoft.com/textbook/stathome.html. Steward, F.R., Peter, S., Richan J.B., 1990. A method for predicting the depth of lethal heat penetration into mineral soils exposed to fires of various intensities. Can. J. For. Res. 20, 919-926. Teague, B., McClead, R., Pascoe, S., 2009. 2009 Victorian Bushfire Royal Commission Interim Report. Australia: Government Printer for the State of Victoria. Terefe, T., Mariscal-Sancho, I., Peregrina, F., Espejo R., 2008. Influence of heating on various properties of six Mediterranean soils. A laboratory study. Geoderma. 143, 273-280. Tomkins, I.B., Kellas, J.D., Tolhurst, K.G., Oswin, D.A. 1991. Effects of fire intensity on soil chemistry in a Eucalypt forest. Aust. J. Soil. Res. 29, 25-47. Turekian, V.C., Macko, S., Ballentine, D., Swap, R.J., Garstang, M., 1998. Causes of bulk carbon and nitrogen isotopic fractionations in the products of vegetation burns: laboratory studies. Chem. Geol. 152, 181-192.

125

Turrión, M.B., Lafuente, F., Aroca, M.J., López, O., Mulas, R., Ruipérez, C., 2010. Characterization of soil phosphorus in a fire-affected forest cambisol by chemical extractions and 31P-NMR spectroscopy analysis. Sci. Tot. Environ. 408, 3342-3348. Ùbeda, X., Pereira, P., Outeiro, L., Martin, D.A., 2009. Effects of fire temperature on the physical and chemical characteristics of the ash from two plots of cork oak (Quercus suber). Land Degrad. and Dev. 20, 589-608. Ulery, A.L., Graham, R.C., 1993. Forest fire effects on soil color and texture. Soil Sci. Soc. Amer. J. 57, 135-140. Ulery, A.L., Graham, R.C., Amrhein, C., 1993. Wood-ash composition and soil pH following intense burning. Soil Sci. 156, 358-364. Ulery, A.L., Graham, R.C., Bowen, L.H., 1996. Forest fire effects on soil phyllosilicates in California. Soil Sci. Soc. Amer. J. 60, 309-315. Underwood, R.J., Christensen, P.E.S., 1981. Forest fire management in Western Australia. Perth, Western Australia. USDA, 2010. Keys to Soil Taxonomy, Eleventh Edition. United States Department of Agriculture, Natural Resources Conservation Service. Vance, E. D., and Mitchell, C. C., 2000. Beneficial use of wood ash as an agricultural soil amendment: case studies from the United States forests products industry. In Land Application of Agricultural, Industrial, and Municipal By-Products, (Eds. J. F. Power and W. A. Dick), Soil Science Society of America, Book series no 6, pp 567-582. Viegas, D.X., 1998. Weather, fuel status and fire occurrence: predicting large fires. In Large Forest Fires. J. M. Moreno (Ed). Leiden, The Netherlands, Backhuys publisher: 31-48. Voundi Nkana, J.C., Demeyer, A., Verloo, M.G., 2002. Effect of wood ash application on soil solution chemistry of tropical acid soils: incubation study. Bioresour. Technol. 85, 323-325. Wada, K., 1977. Allophane and imogolite. In: Dixon, J.B., Weed, S.B. (Eds), Minerals in Soil Environment. Soil Science Society of America, Madison, Wis., USA, pp. 603-638. Wan, S., Hui, D., Luo, Y., 2001. Fire effects on nitrogen pools and dynamics in terrestrial ecosystems : a meta analysis. Ecol. Appl. 11, 1349-1365. Wang, H., Xu, B., Smith, P., Davies, M., Desilva, L., Wingate, C., 2006. Kinetic modeling of gibbsite dehydration/amorphization in the temperature range 823-923K. J. Phys. Chem. Solid. 67, 2567-2582.

126

Watari, F., Landuyt, J.V., Delavignette, P., and Amelinckx, S., 1979. Electron microscopy study of dehydration transformations 1. Twin formation and Mosaic structure hematite derived goethite. J. Solid State Chemis. 29, 137-150. Wattez, J., Courty, M.A., 1987. Morphology of ash of some plant materials. Soil micromorphology. In: Fedoroff, N., Bresson, L.M., Courty, M.A. (Eds.). Proceedings of the Seventh International Working Meeting on Soil Micromorpholgy, Paris, pp. 677-683. White, J.L., Roth, C.B., 1986. Infared Spectrometry. In: Klute, A. (Ed) Methods of soil analysis. Part 1. Physical and Mineralogical Methods Second Edition. Soil Science Society of America, Inc. Madison, Wisconsin USA, pp. 291-330. Wikars, L.O., Scmimmel, J., 2001. Immediate effects of fire severity on soil invertebrates in cut and uncut pine forests. Forest Ecol. Manage. 141, 189-200. Wilbur, R.B., Christensen, N.L., 1985. Effects of fire on nutrient availability in a North Carolina coastal plain pocosin. Am. Midl. Nat. 110, 54-61. Williams, J.P., Biernacki, J.J., Bai, J., Rawn, C.J., 2003. Assesment of a synchrotron X-ray method for quantitative analysis of calcium hydroxide. Cement and Concrete Research 33, 1553-1559. Woods, S.W., Balfour, V., 2008. The effect of as on runoff and erosion after a forest wildfire, Montana, U.S.A. Int. J. Wildland Fire 17, 535-548. Yokozeki, K., Watanabe, K., Sakata, N., Otsuki, N., 2004. Modeling of leaching from cementitious materials used in underground environment. Applied Clay Sci. 26, 293-308. Yusiharni, B.E., Gilkes, R.J., 2011. Rehydration of heated gibbsite, kaolinite and goethite: an assessment of properties and environmental significance. Appl. Clay Sci. In Press.

Yusiharni, B.E., Ziadi, H., Gilkes, R.J., 2007. A laboratory and glasshouse evaluation of chicken litter ash, wood ash and iron smelting slag as liming agents and P fertilisers. Aust. J. Soil Res. 45, 374-389. Zanelli, R., Egli, M., Mirabella, A., Plőtze, M., Abdelmoula, M., 2006. ‘Black’ soils in the Southern Alps: clay mineral formation and transformation, amorphous phases and Fe forms. Clay and Clay Min. 54, 705-722. Zanelli, R., Egli, M., Mirabella, A., Giaccai, D., Abdelmoula, M., 2007. Vegetation effects on pedogenetic forms of Fe, Al and Si and on clay minerals in soils in southern Switzerland and northern Italy. Geoderma. 141, 119-129.

127

Appendix 1: Thermal analysis results for kaolinite previously heated at 500oC: wet

incubated at 55 and 95oC for 14, 70 and 200 days.

500oC 55oC 14 days

TGA

DTA

DTGA

500oC 95oC 14 days

DTA

DTGA

TGA

500oC 55oC 70 days

DTGA

DTA

TGA

500oC 95oC 70 days

DTGA

DTA

TGA

500oC 55oC 200 days

DTA

DTGA

TGA

500oC 95oC 200 days

DTGA

DTA

TGA

128

Appendix 2: Thermal analysis results for kaolinite previously heated at 550oC: wet incubated at 55 and 95oC for 14, 70 and 200 days.

550oC 55oC 14 days

DTA DTGA

TGA

550oC 95oC 14 days

DTGA DTA

TGA

550oC 55oC 70 days

DTGA DTA

TGA

550oC 95oC 70 days

DTA

TGA

DTGA

550oC 55oC 200 days

DTGA DTA

TGA

550oC 95oC 200 days

DTGA

DTA

TGA

129

Appendix 3: Thermal analysis results for goethite previously heated at 250oC: wet

incubated at 55 and 95oC for 14, 70 and 200 days.

250oC 55oC 14 days

DTGA

DTA

TGA

250oC 95oC 14 days

DTGA

DTA

TGA

250oC 55oC 70 days

TGA

DTGA

DTA

250oC 95oC 70 days

TGA

DTGA

DTA

250oC 55oC 200 days

DTGA

DTA

TGA

250oC 95oC 200 days

TGA

DTA

DTGA

130

Appendix 4: Thermal analysis results for goethite previously heated at 300oC: wet incubated at 55 and 95oC for 14, 70 and 200 days.

300oC 55oC 14 days

DTGA

TGA

DTA

300oC 95oC 14 days

DTGA

DTA

TGA

300oC 55oC 70 days

TGA

DTA

DTGA

300oC 95oC 70 days

TGA

DTA

DTGA

300oC 55oC 200 days

TGA

DTGA

DTA

300oC 95oC 200 days

DTGA

TGA

DTA

131

Appendix 5: Thermal analysis results for gibbsite previously heated at 250oC: wet

incubated at 55 and 95oC for 14, 70 and 200 days.

250oC 55oC 14 days

DTA

DTGA

TGA

250oC 95oC 14 days

DTA

DTGA

TGA

250oC 55oC 70 days

TGA

DTGA

DTA

250oC 95oC 70 days

DTA

DTGA

TGA

250oC 55oC 200 days

TGA

DTGA

DTA

250oC 95oC 200 days

TGA

DTA

DTGA

132

Appendix 6: Thermal analysis results for gibbsite previously heated at 300oC: wet incubated at 55 and 95oC for 14, 70 and 200 days.

300oC 55oC 14 days

DTGA

DTA

TGA

300oC 95oC 14 days

DTGA

DTA

TGA

300oC 55oC 70 days TGA

DTA

DTGA

300oC 95oC 70 days

TGA DTGA

DTA

300oC 55oC 200 days

TGA

DTA

DTGA

300oC 95oC 200 days

TGA

DTA

DTGA

133

Appendix 7: Analyses of dry plant tops (leaves and shoots) of annual ryegrass (Lolium rigidum Gaud) for Harvest 1.

Rate of P Na Mg Al S K Ca Mn Fe Cu Zn Mo Pb applied (mg/kg) % % mg/kg % % % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg BH Unburnt 0 0.11 0.31 377 0.41 2.68 1.34 172 84 6.8 53.9 2.8 0.8 BH Unburnt 1.67 0.11 0.26 343 0.37 3.02 1.51 140 82 6.8 49.4 1.2 0.7 BH Unburnt 3.33 0.07 0.25 148 0.30 3.04 1.52 113 59 3.9 28.9 0.9 0.5 BH Unburnt 6.67 0.08 0.31 68 0.23 2.97 1.49 110 47 3.4 24.8 0.9 0.6 BH Unburnt 13.33 0.06 0.27 84 0.20 3.28 1.64 92 46 3.2 19.9 0.5 0.4 BH 250 0 0.13 0.29 192 1.01 3.23 1.62 190 73 9.3 53.2 1.4 1.1 BH 250 1.67 0.20 0.31 349 1.09 4.28 2.14 209 105 11.4 62.9 1.4 0.8 BH 250 3.33 0.14 0.24 60 0.82 3.20 1.60 107 52 6.3 33.9 0.5 0.5 BH 250 6.67 0.14 0.24 102 0.85 4.53 2.26 137 91 6.4 42.5 0.6 0.4 BH 250 13.33 0.18 0.23 42 0.56 3.87 1.94 116 55 4.6 27.1 0.4 0.7 BH 350 0 0.16 0.32 320 0.69 2.56 1.28 219 155 3.8 34.6 4.9 0.4 BH 350 1.67 0.15 0.35 187 1.04 4.19 2.10 252 196 4.1 44.5 3.0 0.4 BH 350 3.33 0.17 0.36 93 1.03 4.53 2.26 273 324 3.9 57.5 3.3 0.6 BH 350 6.67 0.20 0.32 269 0.77 4.58 2.29 265 377 3.6 35.0 2.9 0.5 BH 350 13.33 0.11 0.29 83 0.62 4.31 2.16 208 196 2.4 36.2 1.9 0.5 BH 500 0 0.03 0.28 226 0.26 2.66 1.33 197 111 3.7 29.0 5.9 0.5 BH 500 1.67 0.02 0.22 123 0.19 2.30 1.15 173 104 2.8 33.9 5.9 0.4 BH 500 3.33 0.02 0.24 201 0.22 2.82 1.41 195 121 4.4 33.1 8.2 0.3 BH 500 6.67 0.01 0.23 138 0.18 2.75 1.38 233 2610 2.0 46.0 4.9 0.2 BH 500 13.33 0.01 0.26 139 0.18 2.42 0.81 253 84 2.9 22.9 5.0 0.6

134

Appendix 8: Analyses of dry plant tops (leaves and shoots) of annual ryegrass (Lolium rigidum Gaud) for Harvest 2.

Rate of P Na Mg Al S K Ca Mn Fe Cu Zn Mo Pb applied (mg/kg) % % mg/kg % % % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg BH Unburnt 0 0.64 1.12 626 0.53 2.17 1.76 1166 172 8.3 191.8 2.1 1.0 BH Unburnt 1.67 0.56 0.80 672 0.61 2.60 1.38 894 211 7.3 171.3 1.0 0.9 BH Unburnt 3.33 0.15 0.37 218 0.34 3.22 0.53 285 85 4.9 65.8 1.6 0.6 BH Unburnt 6.67 0.10 0.47 318 0.35 2.80 0.88 241 103 5.0 44.3 1.4 0.6 BH Unburnt 13.33 0.14 0.42 192 0.34 2.69 0.64 250 53 3.8 31.7 1.0 0.3 BH 250 0 0.91 0.72 240 0.87 3.14 0.94 594 45 8.7 91.1 2.3 0.7 BH 250 1.67 0.99 0.67 287 0.95 3.32 1.14 804 83 3.1 103.1 0.6 0.4 BH 250 3.33 0.72 0.61 67 1.03 3.35 0.98 422 30 7.4 91.8 1.7 0.3 BH 250 6.67 0.41 0.48 240 0.59 2.49 0.69 306 91 4.7 45.0 1.1 0.4 BH 250 13.33 0.33 0.48 288 0.49 2.43 0.82 251 133 4.8 56.7 1.4 0.4 BH 350 0 0.47 0.52 666 0.41 1.98 0.92 580 287 2.3 46.9 3.9 1.6 BH 350 1.67 0.31 0.49 228 0.55 3.18 0.80 499 142 3.0 42.3 3.5 0.2 BH 350 3.33 0.29 0.43 271 0.46 3.04 0.70 465 160 2.8 48.0 3.0 0.4 BH 350 6.67 0.27 0.43 398 0.36 2.43 0.69 499 156 1.8 23.3 1.4 0.3 BH 350 13.33 0.13 0.36 83 0.33 1.90 0.51 391 67 1.2 17.2 0.9 0.2 BH 500 0 0.18 0.57 804 0.30 1.88 1.79 749 326 4.1 64.3 11.6 0.4 BH 500 1.67 0.15 0.59 1158 0.22 1.94 1.76 981 363 3.7 79.6 12.3 0.6 BH 500 3.33 0.07 0.42 1036 0.22 2.03 1.14 690 331 3.4 63.0 14.7 0.5 BH 500 6.67 0.04 0.33 340 0.27 2.26 0.95 592 142 2.7 50.6 13.4 0.1 BH 500 13.33 0.03 0.27 129 0.22 2.49 0.81 426 78 2.4 29.7 9.6 0.1

135

Appendix 9: Analyses of dry plant tops (leaves and shoots) of annual ryegrass (Lolium rigidum Gaud) for Harvest 3.

Rate of P Na Mg Al S K Ca Mn Fe Cu Zn Mo Pb applied (mg/kg) % % mg/kg % % % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg BH Unburnt 0 1.37 1.18 1940 1.14 2.11 3.23 1074 340 9.3 191.6 6.0 1.9 BH Unburnt 1.67 1.10 1.24 3130 1.01 2.16 3.15 1170 694 9.3 223.3 0.9 3.8 BH Unburnt 3.33 0.52 0.49 659 0.44 2.51 0.83 477 116 6.4 105.1 1.0 1.1 BH Unburnt 6.67 0.42 0.50 269 0.38 2.40 0.62 400 75 5.9 67.7 1.2 0.7 BH Unburnt 13.33 0.40 0.42 493 0.31 2.04 0.51 290 99 5.5 43.1 1.0 0.5 BH 250 0 1.50 1.39 3829 2.07 2.14 4.05 992 559 9.9 105.6 1.1 4.0 BH 250 1.67 1.51 1.09 2651 1.57 2.10 3.32 1163 1240 8.2 84.8 0.8 4.6 BH 250 3.33 1.28 0.90 397 0.99 2.41 2.18 637 100 8.0 123.5 1.3 0.8 BH 250 6.67 0.81 0.57 243 0.62 2.32 0.94 476 70 6.8 73.3 1.2 0.5 BH 250 13.33 0.49 0.52 553 0.49 1.99 0.71 347 105 5.8 46.6 1.2 0.6 BH 350 0 1.59 1.15 3260 1.50 1.66 3.95 1058 496 4.4 42.9 2.8 2.5 BH 350 1.67 1.30 0.74 440 0.82 1.98 2.05 925 125 3.7 53.5 2.0 0.4 BH 350 3.33 0.99 0.58 771 0.63 1.87 1.34 793 174 3.9 59.4 2.1 0.8 BH 350 6.67 0.56 0.45 265 0.42 2.19 0.82 736 92 2.3 33.7 1.2 0.6 BH 350 13.33 0.39 0.41 157 0.36 1.73 0.61 552 80 2.1 20.7 1.0 0.4 BH 500 0 0.61 0.94 1039 0.94 1.38 3.55 824 232 6.6 83.8 14.0 0.5 BH 500 1.67 0.62 1.05 542 0.39 1.68 3.23 1333 205 5.5 115.8 8.6 1.4 BH 500 3.33 0.47 0.92 736 0.54 1.84 3.24 988 134 3.1 89.8 7.8 0.9 BH 500 6.67 0.25 0.41 479 0.40 1.85 1.37 768 105 3.2 65.7 10.7 0.3 BH 500 13.33 0.12 0.31 45 0.22 1.89 0.72 707 46 2.1 34.7 5.2 0.1

136

Appendix 10: Conventional XRD patterns for (a)

; (b) (N=nesquehonite

(MgCO3.H2O), C=calcite (CaCO3), S=sylvite (KCl), A=apatite (Ca5(PO4)3(OH)), E=scolecite (CaAl2Si3O10.3(H2O)), Q=quartz (SiO2), Po=portlandite (Ca(OH)2) and P=periclase (MgO).

5 15 25 35 45 55 65

2θ (degrees)

SW

SL

PM

WW

WL C

C

S C C C C

C C C C C

Q

Q N N A A

A

5 15 25 35 45 55 65

2θ (degrees)

RW

RL

GT

JL

HH

C

C

E E

S Q

Q R

C

A E

A

C C C P C C C C

Po E P S Po

H