developing a tarodigilib.library.usp.ac.fj/gsdl/collect/usplibr1/index/... · 2016. 4. 25. ·...

DEVELOPING A TARO (Colocasia esculenta)

PRODUCTION SYSTEM BASED ON GENOTYPE AND

FALLOW SYSTEM FOR ECONOMIC AND

ENVIRONMENTAL SUSTAINABILITY UNDER LOCAL

CONDITIONS IN SAMOA

by

Sanjay Anand

A thesis submitted in fulfilment of the

requirements for the Degree of

Doctor of Philosophy

Copyright © 2016 by Sanjay Anand

School of Agriculture and Food Technology

Faculty of Business and Econmoics

The Univesrity of the South Pacific

March 2016

iii

ACKNOWLEDGEMENTS

I hereby would like to extend my sincere gratitude and appreciation to a great number of

people who were very helpful and generous during the course of my research. Without

the assistance of these people, things would not have gone nearly as well. I am greatly

indebted to all of them.

I wish to express my profound appreciation and sincere thanks to Dr. Danilo F. Guinto,

Senior Lecturer and Head of Department of Soil Science and my mentor, for his

valuable guidance and inestimable help throughout the period of this investigation and in

the preparation of the manuscript.

I would like to express my sincere and profound gratitude to Associate Professor

Mohammed Umar, the Head of School of Agriculture and Food technology, for his

magnificient guidance, administrative, financial and technical support towards the

completion of this first ever Doctor of Philosophy dissertation since the inception of the

school.

I also wish to extend my heartfelt thanks to the Australian Centre for International

Agricultural Research (ACIAR) for coordinating the Soil Health Project in collaboration

with the Secretariat of the Pacific Community (SPC), and the University of the South

Pacific (USP) in Samoa, facilitating the financial support for the research as well as

providing technical expertise through its personnel, especially Dr. Tony Pattison and Dr.

Mike Smith, but for which this study would not have been possible.

My sincere appreciation to Mr. Tolo Iosefa for assisting me in identifying the research

sites as well as towards procurement of taro planting materials. I am indeed grateful to

the following taro farmers who kindly permitted me to use their land for the entire

duration the soil health research: Mr. Tuala of Safaatoa; Mr. Unasa of Siufaga, Faga;

iv

Mr. Falelima of Aopo; and Reverend Maselusi Misa of the Falealili Assemblies of God

Church Farm.

My unfathomable appreciation is hereby extended to the entire crew of the Soil Science

Department at the Alafua Campus: Mr. Daya Perera, Samuelu Saulia, Dean Seuoti and

Phillip Reti for accompanying and helping me tirelessly during the execution of the field

and laboratory tasks. I would also like to thank all my student friends and colleagues:

Ami, Ashika, Binesh, Bimlesh, Toloi, Edmund, Shonal, Rohit, Amit and Dinesh, for

helping me with executing field tasks, upkeep of the field sites as well as during the data

collection phase.

My very special thanks to Mr. David Hunter from the Scientific Research Organisation

of Samoa (SROS) and his team for collaborative research efforts towards the Samoa Soil

Health Project.

I would also like to thank the University of the South Pacific’s Scholarship Committee

for awarding me the opportunity to pursue my studies, having confidence in my

capabilities to carry out the undertakings of the Samoa Soil Health Project.

I express my profound debt of gratitude to my family members for their consistent

encouragement and unfailing help in many facets of this work.

And finally to the glory of the almighty, my saviour to whom I owe everything. You are

the pillar of my strength.

v

ABSTRACT

Developing agricultural production systems able to conserve soil quality is essential to

guarantee the current and future capacity of soil to provide ecosystem services. This

study investigated the efficacy of six month fallow effects of four contrasting cover

crops namely, grass, mucuna, erythrina and biochar over four agro-ecological taro

growing zones in Samoa. Understanding mineralisation patterns of green manure

residues is crucial in the synchronisation of nutrient release from plant residue and

uptake by plants. This research work focused on tailoring the release of nutrients from

green manure that has been applied as mulch, with the aim of improving the efficiency

with which green manure nutrients are used in taro cropping systems. All the organic

soil amendments were evaluated in the screen house and multi location field conditions

in an attempt to estimate the rates of decomposition of green manure residues and their

effects on the dynamics of soil labile C, fluorescein diacetate hydrolysis activity (broad

spectrum soil biological activity) and the potentially mineralisable N pools. The mineral

N fluxes (NH4+-N and NO3

--N) from the embedded covered core in-situ aerobic

incubations as well as the net mineralisation potentials from subsequent mineralisation

of green manure mulch residues were also studied. The influence of the green manure

supply on the yield and nutrient content of two taro (Colocasia esculenta) cultivars was

also determined. The effects of the decomposition of cover crop mulches on nematode

population were also evaluated. In addition to the biochemical indicators mentioned

above, soil phosphatase and urease assays were investigated during 90 days incubation

period of the mulch residues at different rates in pots under the screen house conditions.

The nutrient uptake and nutrient use efficiency of the two taro cultivars: Samoa 1 and

Samoa 2; used for this research was determined in a separate pot experiment.

Results from this study indicated that all the fallow treatments significantly improved

the soil active carbon stocks upon decomposition; which however, were largely

dependent on the biomass production. Mucuna fallow contributed to the largest

additions of biomass across all the agro-ecological sites and as such proved to be the

superior cover crop with regards to improving soil active carbon, soil biological activity

vi

as well as the potentially mineralisable N pools. Mucuna fallows also resulted in

significantly greater inputs of mineral N to the soil system and showed significant net

mineralisation potentials over two of the sites. It significantly contributed to the

suppression of plant parasitic nematodes while enhancing the activity of free living

genera. Nutrient uptake and the corresponding yields of taro were comparatively higher

under the mucuna fallow, both with and without supplementation with complete mineral

fertilisers. Comparable yields under biochar treated plots were observed, owing to

appreciable amounts of K uptake.

The comparative economic analysis of the mucuna fallow technology against the

traditional grass fallow revealed 98% and 48% higher gross margin for Salani and

Safaatoa sites, respectively; while 21 fold increases was observed for the Siufaga site.

The potential practical benefits of the mucuna legume technology for South Pacific taro

farmer’s looks promising in terms of increased yields reduced labour requirements,

reduced fertiliser inputs, and suppression of weeds and plant parasitic nematode

population.

Significant positive associations were found to exist between the yield of cultivar Samoa

2 and mean levels of FDA, PMN, NH4+-N and NO3

- -N. However, yield of cultivar

Samoa 1 showed no significant correlation with these soil parameters. Mean levels of

labile C did not correlate with the yields of any of the cultivars.

The rate of decomposition and subsequent release of mineral N was also favourable for

the decomposing mucuna litter. Although erythrina had higher N content than mucuna,

most of it was released so rapidly during the initial stages of decomposition that it could

not be resourcefully made use of by the taro crop. Conversely, the rate of decomposition

and the N release pattern from the mucuna residues was more gradual and synchronised

well with the vegetative growth phase of the taro crop.

vii

The dry matter accumulation and nutrient uptake of the two taro cultivars revealed that

the cultivars exhibited significant differences for the various nutrients with regards to

their efficiency of utilisation towards production of a unit of edible dry matter. Cultivar

Samoa 1 had a higher nutrient use efficiency for N, P, K, Mg, Mn and Cu over cultivar

Samoa 2. However, for Ca, Fe and Zn, cultivar Samoa 2 had a higher nutrient use

efficiency over cultivar Samoa.

viii

TABLE OF CONTENTS

Acknowledgements iii

Abstract v

List of Tables xiii

List of Figures xviii

List of Plates xxi

List of Appendices xxii

CHAPTER 1 INTRODUCTION 1

1.1 Research objectives 5

CHAPTER 2 LITERATURE REVIEW 7

2.1 Background on Samoa 7

2.2 Soil ecosystems 10

2.3 Soil degradation and soil health 11

2.4 Properties and indicators of healthy soils 12

2.4.1 Properties of healthy soils 12

2.4.2 Measurement of soil quality 13

2.4.3 Soil health indicators 14

2.4.3.1 Biological indicators 14

2.4.3.2 Chemical indicators 15

2.4.3.3 Physical indicators 15

2.4.3.4 Minimum data set (MDS) 15

2.5 Role of nematodes in soil nutrient cycling 16

2.6 Goals in developing nematode management soil ecosystem 18

2.7 Importance of maintaining soil functional diversity 20

2.8 Indicative value of nematode trophic group abundance and food web indices

22

2.9 Agricultural practices compatible with soil ecosystem management

23

2.10 Cover crops 25

2.11 The importance of organic matter in soil fertility and crop 29

ix

health

2.12 Nitrogen-fixing plants to improve soil fertility and health 39

2.12.1 Mucuna pruriens 43

2.12.2 Erythrina 44

2.13 The use of biochar in agriculture 46

2.13.1 Biochar for sustainable agriculture 48

2.14 Microbial and biochemical indicators of soil health 49

2.14.1 Enzyme activity tests as soil quality indicators 50

2.14.2 Soil microbial activity 52

2.14.2.1 Fluorescein diacetate hydrolysis (FDA) activity

53

2.14.2.2 Soil urease activity 53

2.14.2.3 Soil phosphatase activity 57

2.14.3 Nitrogen mineralisation 58

2.14.4 Potentially mineralisable nitrogen (PMN) 60

2.14.4.1 Factors affecting PMN 61

2.14.4.2 Relationship of PMN to soil functions 62

2.14.4.3 PMN problems associated with poor activity 62

CHAPTER 3 MATERIALS AND METHODS 63

3.1 Experiment 1 The soil health fallow trial 63

3.1.1 Research sites 63

3.1.2 Site characterisation and history 66

3.1.2.1 Salani, Falealili, - high rainfall zone, Upolu 66

3.1.2.2 Safaatoa, Lefaga, - low rainfall zone, Upolu 70

3.1.2.3 Siufaga, Faga, - high rainfall zone, Savaii 75

3.1.2.4 Aopo, Auala, - low rainfall zone, Savaii 79

3.1.3 The fallow treatments 83

3.1.4 Plant culture 85

3.1.5 Experimental design and size 87

3.1.6 Data collection 87

3.1.6.1 Meteorological data collection 87

x

3.1.6.2 Soil parameters 87

3.1.6.2.1 Biochemical indicators 88

i. Soil labile carbon 88

ii. Fluorescein diacetate hydrolysis activity

89

iii. Potentially mineralisable nitrogen 89

iv. Mineral nitrogen fluxes 90

3.1.6.2.2 Biological indicators – nematode study

91

3.1.6.3 Plant parameters 94

3.1.6.3.1 Dry matter yields and nutrient uptake of cover crops

94

3.1.6.3.2 Harvesting and yield of taro crop 94

3.1.6.3.3 Dry matter yields and nutrient uptake of taro corms

95

3.1.7 Statistical analysis 95

3.2 Experiment 2 The soil incubation experiment 95

3.2.1 Background 95

3.2.2 The trial description 96

3.2.3 Application of organic amendments 96

3.2.4 Treatment, factors and levels 97

3.2.5 Experimental design 98


3.2.6.1 Analysis of biochemical soil health indicators 99

3.2.6.2 Assay of soil urease activity 99

3.2.6.3 Assay of soil phosphatase activity 100


3.3 Experiment 3 The taro nutrient budgeting experiment 101

3.3.1 Background 101

3.3.2 Description of the trial 101

3.3.3 Nutrient supplementation and incubation 102

3.3.4 Experimental design, layout and size 102


xi


CHAPTER 4 RESULTS AND DISCUSSION 104

4.1 Experiment 1 The soil health fallow trial 104

4.1.1 Meteorological variables 104

4.1.1.1 Rainfall 104

4.1.1.2 Temperature 106

4.1.2 Soil biochemical indicators 108

4.1.2.1 Labile carbon 108

4.1.2.2 Fluorescein diacetate hydrolysis activity (FDA) 113

4.1.2.3 Potentially mineralisable nitrogen (PMN) 119

4.1.2.4 Mineral N fluxes from embedded in-situ covered core aerobic incubation

124

4.1.2.4.1 Ammonium nitrogen 124

4.1.2.4.2 Nitrate nitrogen 129

4.1.2.4.3 Cumulative net N mineralisation 134

4.1.2.5 Associations between the evaluated biochemical soil parameters

138

4.1.3 Nematode community analysis 139

4.1.3.1 Salani site 139

4.1.3.2 Safaatoa site 145

4.1.4 Cover crop dry matter yields, nutrient concentrations and nutrient uptake over the four sites

149

4.1.5 Taro yields 153

1 4.1.5.1 Fresh corm yields 153

4.1.5.2 Associations between fresh corm yield and mean levels of the evaluated biochemical parameters

158

4.1.6 Corm nutrient uptake by the two cultivars produced under the different fallow systems over the three sites

159

4.1.6.1 Salani site 159

4.1.6.2 Safaatoa site 159

xii

4.1.6.3 Siufaga site 160

4.1.6.4 Associations between corm dry matter and mean levels of macronutrient uptake

164

4.1.7 Marginal economic analysis of taro grown under the mucuna fallow versus the traditional grass fallow

165

4.2 Experiment 2 The soil incubation pot trial 167

4.2.1 Labile carbon measurement 167

4.2.2 Fluorescein diacetate hydrolysis (microbial activity) measurements (FDA)

181

4.2.3 Potentially mineralisable nitrogen (PMN) 175

4.2.4 Ammonium nitrogen (NH4+ - N) 179

4.2.5 Nitrate nitrogen (NO3- - N) 183

4.2.6 Assay of soil phosphatase activity 186

4.2.7 Assay of soil urease activity 190

4.3 Experiment 3 The taro nutrient uptake pot trial 194

4.3.1 Dry matter accumulation by various plant organs 194

4.3.2 Nutrient uptake of the two taro cultivars as influenced by plant age

197

4.3.3 Nutrient concentration of the two taro cultivars 201

CHAPTER 5 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 204

5.1 Summary 204

5.2 Conclusions 208

5.3 Recommendations for future researchers and farmers 210

References 211

Appendices 245

xiii

LIST OF TABLES

Table 3.1 Field experimental sites for fallow trial 63

Table 3.2 (a) Characterisation of Salani site

(b) Soil profile description of Salani site

(c) Selected soil physical, chemical and fertility indicators (0-15cm)

of Salani site

67

68

70

Table 3.3 (a) Characterisation of Safaatoa site

(b) Soil profile description of Safaatoa site


of Safaatoa site

72

73

74

Table 3.4 (a) Characterisation of Siufaga site

(b) Soil profile description of Siufaga site


of Siufaga site

76

77

78

Table 3.5 (a) Characterisation of Aopo site

(b) Soil profile description of Aopo site


of Aopo site

80

81

82

Table 3.6 Fallow treatments 83

Table 3.7 Characterisation of biochar 84

Table 3.8 Actual dates of fallow establishment, killing of cover crops and

planting and harvesting of the taro crop for the four sites

85

Table 3.9 Characterisation of the taro cultivars 86

Table 3.10 Treatments, factors and levels of the soil incubation pot experiment 97

Table 3.11 Biochemical soil health indices and their significance 98

Table 4.1 (a) Table of predicted fallow means from repeated measures analysis

for labile carbon under different fallow systems across all time

points for the four sites

(b) Table of predicted estimates for fallow x time interaction from

111

111

xiv

repeated measures split-plot in time analysis for labile carbon

across all fallow systems for all sites

(c) Within time point comparison among fallow types for labile C

for each site

(d) Between-site comparison for labile C

112

112


for FDA under different fallow systems across all time points for

the four sites


repeated measures split-plot in time analysis for FDA across all

fallow systems for all sites

(c) Within time point comparison among fallow types for FDA for

each site

(d) Between-site comparison for FDA

116

117

117

118


for PMN under different fallow systems across all time points for

the four sites


repeated measures split-plot in time analysis for PMN across all

fallow systems for all sites

(c) Within time point comparison among fallow types for PMN for

each site

(d) Between-site comparison for PMN

121

122

122

123


for ammonium-N under different fallow systems across all time



repeated measures split-plot in time analysis for ammonium-N

across all fallow systems for all sites

(c) Within time point comparison among fallow types for

126

127

xv

ammonium-N for each site

(d) Between-site comparison for ammonium-N

127

128


for nitrate-N under different fallow systems across all time



repeated measures split-plot in time analysis for nitrate-N across

all fallow systems for all sites.

(c) Within time point comparison among fallow types for nitrate-N

for each site

(d) Between-site comparison for nitrate-N

131

132

132

133


for cumulative net mineral N under different fallow systems

across all time points for the four sites


repeated measures split-plot in time analysis for cumulative net

mineral N across all fallow systems for all sites

(c) Within time point comparison among fallow types for

cumulative net mineral N for each site

(d) Between-site comparison for cumulative net mineral N

136

137

137

138

Table 4.7 Pearson’s product-moment correlation analyses between the soil

biochemical indicators

139

Table 4.8 (a) Nematode enumeration, classification and analysis of principal

components and indices at different timea of the fallow

experiment for Salani site

(b) Differences in the nematode count, distribution and indices

indicating shifts in activity across the various trophic guilds

between pre and post decomposition of fallow cover crop

residues for Salani site

143

144

Table 4.9 (a) Nematode enumeration, classification and analysis of principal 147

xvi

components and indices at different timea of the fallow

experiment for Safaatoa site

(b) Differences in the nematode count, distribution and indices

indicating shifts in activity across the various trophic guilds

between pre and post decomposition of fallow cover crop

residues for Safaatoa site

148

Table 4.10 (a) Dry matter yields and nutrient concentrations of the fallow crops

grown over the four sites

(b) Dry matter yields and nutrient uptake by fallow crops over the

four sites

151

152

Table 4.11 (a) Predicted mean taro yields for individual sites

(b) Predicted mean taro yields for the two cultivars

(c) Predicted mean corm yield of taro under different fallows within

sites

153

154

154

Table 4.12 Pearson’s product-moment correlation analyses between the

evaluated biochemical indicators and the fresh taro corm yields of the

two cultivars

158

Table 4.13 (a) Dry matter yields and macronutrient uptake of the two cultivars

of taro corms produced under the different fallow practices at

Salani site

(b) Dry matter yields and macronutrient uptake of the two cultivars


Safaatoa site

(c) Dry matter yields and macronutrient uptake of the two cultivars


Siufaga site

61

162

163

Table 4.14 Pearson’s product-moment correlation analyses between corm dry

matter and macronutrient uptake

164

Table 4.15 Marginal economic analysis of the mucuna fallow technology versus

the current farmer’s practice, without the use of chemical fertilisers

166

xvii

Table 4.16 Repeated measures analysis of parameters for soil labile C under the

incubation trial

168

Table 4.17 Repeated measures analysis of parameters for soil FDA under the

incubation trial

171

Table 4.18 Repeated measures analysis of parameters for soil PMN under the

incubation trial

175

Table 4.19 Repeated measures analysis of parameters for ammonium-N under

the incubation trial

179

Table 4.20 Repeated measures analysis of parameters for nitrate-N under the

incubation trial

183

Table 4.21 Repeated measures analysis of parameters for phosphate

mineralisation under the incubation trial

187

Table 4.22 Repeated measures analysis of parameters for urease activity under

the incubation trial

191

Table 4.23 Analysis of variance for effects of cultivar and days after planting on

total dry weight and plant uptake of various nutrients

196

Table 4.24 Maximum levels of nutrient uptake by the two cultivars 197

Table 4.25 Percent nutrient concentration in the laminar of the third uppermost

leaf of the two taro cultivars at various stages of growth

200

xviii

LIST OF FIGURES

Figure 2.1 Complex components of soil health/quality indicators 13

Figure 2.2 Roles of nematodes in organic matter decomposition 21

Figure 2.3 Nematode community succession in relation to C:N ratios of soil

amendments 22

Figure 2.4 Functional guilds of nematodes characterised by trophic groups and

life history 23

Figure 2.5 Organic compounds in soil before cultivation 30

Figure 2.6 Effects of cultivation on native organic matter levels 38

Figure 2.7 Contributions of nitrogen fixing trees to the soil ecosystem 41

Figure 4.1 Rainfall pattern for the two year research period for the four

experimental sites 105

Figure 4.2 Mean day temperature over the two year research period for the four

experimental sites

107

Figure 4.3 Labile carbon trends for the four fallow sites under various fallow

systems

108

Figure 4.4 Overall labile C trend 110

Figure 4.5 Microbial activity trends for the four fallow sites under various

fallow systems

113

Figure 4.6 Overall microbial activity trend 115

Figure 4.7 Potentially mineralisable nitrogen (PMN) trends for the four fallow

sites under various fallow systems

119

Figure 4.8 Potentially mineralisable N trend 120

Figure 4.9 Ammonium nitrogen fluxes for the four fallow sites under various

fallow systems

124

Figure 4.10 Ammonium N trend 125

xix

Figure 4.11 Nitrate nitrogen fluxes for the four fallow sites under various fallow

systems

129

Figure 4.12 Nitrate N trend 130

Figure 4.13 Net cumulative N mineralisation trends for the four sites under

various fallow systems

134

Figure 4.14 Cumulative net N mineralisation trend 135

Figure 4.15 Actual corm yields of the two cultivars of taro from the three sites 137

Figure 4.16 (a) Labile C dynamics for Salani soil incubated with different organic

mulches at different rates in pots without plants under screen

house conditions

(b) Labile C dynamics for Safaatoa soil incubated with different

organic mulches at different rates in pots without plants under

screen house conditions

169

173

Figure 4.17 (a) FDA dynamics for Salani soil incubated with different organic


house conditions

(b) FDA dynamics for Safaatoa soil incubated with different organic


house conditions

173

174

Figure 4.18 (a) PMN dynamics for Salani soil incubated with different organic


house conditions

(b) PMN dynamics for Safaatoa soil incubated with different organic


house conditions

177

178

Figure 4.19 (a) Ammonium-N fluxes for Salani soil incubated with different



(b) Ammonium-N fluxes for Safaatoa soil incubated with different


181

182

xx


Figure 4.20 (a) Nitrate-N fluxes for Salani soil incubated with different organic


house conditions

(b) Nitrate-N fluxes for Safaatoa soil incubated with different organic


house conditions

184

185

Figure 4.21 (a) Phosphatase activity fluxes for Salani soil incubated with different



(b) Phosphatase activity fluxes for Safaatoa soil incubated with

different organic mulches at different rates in pots without plants

under screen house conditions

188

189

Figure 4.22 (a) Urease activity fluxes for Salani soil incubated with different



(b) Urease activity fluxes for Safaatoa soil incubated with different



192

193

Figure 4.23 Dry weights of plant organs of the two taro cultivars as influenced by

age 195

Figure 4.24 Macronutrient contents of the two taro cultivars as influenced by

plant age 199

Figure 4.25 Micronutrient contents of the two taro cultivars as influenced by plant

age 200

Figure 4.26 Relationship between corm dry matter yield and macronutrient

content of the two cultivars

202

Figure 4.27 Relationship between corm dry matter yield and micronutrient

content of the two cultivars

203

xxi

LIST OF PLATES

Plate 2.1 Location of Samoa 9

Plate 3.1 Location of the research sites 65

xxii

LIST OF APPENDICES

Appendix 1 Trial design, layout and randomisation for the soil health fallow

experiment 248

Appendix 2 Layout and randomisation of the soil incubation experiment 249

Appendix 3 Typical rates of nutrient supplementation for the nutrient budgeting

experiment 250

Appendix 4 Experimental layout and randomisation for the taro nutrient

budgeting experiment 251

Appendix 5 Life cycle of a taro plant 252

Appendix 6 Parts of a taro plant 253

Appendix 7 Index leaf of a taro plant 254

Appendix 8 Labile carbon determination 255

Appendix 9 Fluorescein diacetate hydrolysis analysis 256

Appendix 10 Nematode extraction procedure 258

Appendix 11 Repeated measures analysis for soil labile C for field trial 259

Appendix 12 Repeated measures analysis for soil microbial activity (FDA) for field

trial 271

Appendix 13 Repeated measures analysis for potentially mineralisable N (PMN)

for field trial 279

Appendix 14 Repeated measures analysis for ammonium N (NH4+ - N) for field

trial 286

Appendix 15 Repeated measures analysis for nitrate N (NO3- - N) for field trial 294

Appendix 16 Repeated measures analysis for net cumulative N mineralisation 301

xxiii

Appendix 17 Nested classification analysis of variance for between site

comparisons of biochemical parameters

309

Appendix 18 Correlation analysis for test of association between soil biochemical

properties 318

Appendix 19 Nested classification analysis of variance for yield 321

Appendix 20 Repeated measures analysis for the soil incubation experiment 323

1

CHAPTER 1

INTRODUCTION

Intensification of agriculture has a major impact on the soil environment (OECD, 1999).

Adverse impacts of agriculture include loss of biodiversity, nitrogen discharges into

surface water, eutrophication of surface water, contamination of groundwater from

pesticides and nitrate, and ammonia volatilisation due to over-fertilisation with manure

(OECD, 1999). These impacts are exacerbated by infrastructure development, increasing

urbanisation, waste disposal and forestry practices.

As contemporary agriculture struggles to find a balance between feeding the world and

managing problems such as salinity, soil acidification, declining bio-diversity, pesticide

resistance and human and animal health concerns, a renaissance in integrative thinking is

permeating agricultural policy and research. Researchers are beginning to investigate

organic farming systems in the hope that they may provide some solutions to improving

agricultural sustainability (Neeson, 2001).

Declining soil fertility is thought to present a major threat to sustainable agricultural

development in the South Pacific Island Countries, as smallholders respond to economic

incentives to supply growing urban and export markets, while lacking the technologies

and knowledge to underpin the sustainability of these newly intensified production

systems.

The conventional approaches to soils research have tended to focus on the physical and

chemical properties of soils and, perhaps, the function of soils in plant nutrition and

water relations. A ‘soil health’ approach focuses explicitly on the functions of soil -

including plant nutrition and water relations, but giving attention also to biological

processes and to the 'ecosystem services provided by soils, for instance in the biological

suppression of soil-borne pests and diseases. Although these underlying processes may

be complex, hard to measure directly and difficult to understand completely,

2

considerable practical progress can be made in solving soil-related production problems

through adaptive research, especially to restore the biological functions of degraded soils

(Karlen et al., 2001).

A definition of soil health based on this concept would encompass only a small fraction

of the many roles soil play (Singer et al., 2000). Soil health is the net result of on-going

conservation and degradation processes, depending highly on the biological component

of the soil ecosystem, and influences plant health, environmental health, food safety and

quality (Halvorson et al., 1997; Parr et al., 1992).

Several definitions of soil health have been proposed during the last decades.

Historically, the term soil quality described the status of soil as related to agricultural

productivity or fertility (Singer et al., 2000). In the 1990s, it was proposed that soil

quality was not limited to soil productivity but instead expanded to encompass

interactions with the surrounding environment, including the implications for human and

animal health. In this regard, several examples of definitions of soil quality have been

suggested (Doran et al., 1994). In the mid-1990s, the term soil health was introduced.

Soil health is defined as the continued capacity of soil to function as a vital living

system, by recognising that it contains biological elements that are key to the ecosystem

function within land-use boundaries (Doran and Ziess, 2000; Karlen et al., 2001). These

functions are able to sustain biological productivity of the soil, maintain or enhance the

quality of surrounding water and air, as well as promote plant, animal and human health

(Doran et al., 1996a). Healthy soil functions optimally through balanced interactions

amongst its biological, physico-chemical and mineral components. The mineral

component consists of sand, silt and clay particles; the physico-chemical component

consists of soil aggregates, pore space, reactive surfaces, and organic and inorganic

compounds; and the biological component consists of roots, insects, invertebrates and

microorganisms. Healthy soils function to:

• sustain biological productivity;

• store and cycle water and nutrients ;

3

• decompose organic matter;

• inactivate toxic compounds;

• suppress pathogens; and,

• protect water quality and enhance catchment health.

Soil is part of the terrestrial environment and supports all forms of terrestrial life. Soil

health is the result of continuous conservation and degradation processes and represents

the continued capacity of soil to function as a vital living ecosystem. A unique balance

of chemical, physical and biological (including microbial) components contribute to

maintaining soil health. Evaluation of soil health, therefore, requires indicators of all

these components (Doran and Ziess, 2000; Karlen et al., 2001).

Healthy soils are essential for the integrity of terrestrial ecosystems to remain intact or to

recover from disturbances, such as drought, climate change, pest infestation, pollution,

and human exploitation including agriculture. Protection of soil is, therefore, of high

priority, and a thorough understanding of ecosystem processes is a critical factor in

assuring that soils remain sustainably productive for generations to come (Ellert et al.,

1997).

To manage and maintain soil in a sustainable fashion, the definition of soil health must

be broad enough to encompass the many functions of soil, e.g. environmental filter,

plant growth and water regulation (Doran et al., 1997). Definitions of air and water

quality standards have existed for a long time, while a similar definition does not exist

for soil. There is, however, little if any parallels between air or water quality and soil

health (Sojka et al., 1999). Air and water quality standards are usually based on

maximum allowable concentration of materials hazardous to human health.

Soil quality is defined as the capacity of a reference soil to function, within natural and

managed ecosystem boundaries, to sustain plant and animal productivity, maintain or

enhance water and air quality, and support human health and habitation (Karlen et al.,

1997).

4

Subsequently the two terms, ‘soil health’ and ‘soil quality’, are used interchangeably

(Karlen et al., 2001) although it is important to distinguish that, soil quality is related to

soil functions (Karlen et al., 2003; Letey et al., 2003), whereas soil health presents the

soil as a finite, non-renewable and dynamic living resource (Doran and Ziess, 2000).

Hierarchy and emergence are properties of all systems including soils. These properties

imply there are higher level components and functions of the system that depend on, and

emerge from, lower level components and functions. They enable the whole to be more

than the sum of the parts. The biological and organic component and functions of soils

depend on, and emerge from, the physiochemical and mineral components. Hence the

abundance, diversity and functioning of these organisms is a key indicator of soil health.

Protection of soil quality under intensive land use and fast economic development is a

major challenge for sustainable resource in the developing world (Doran et al., 1996b).

The basic assessment of soil health and quality is necessary to evaluate the degradation

status and changing trends following natural disasters, such as tsunami or different land-

use management interventions. Adverse effects of soil health and quality arise from

nutrient imbalance in soil, excessive fertilisation, soil pollution and soil loss process

(Zhang et al., 1996; Hedlund et al., 2003).

The soil health concept is well-grounded scientifically on accepted principles and

practical experiences in developed countries of the world - and in these countries, is

gaining mainstream recognition (Doran and Parkin, 1994). In the South Pacific Island

Countries, however, the practical demonstrations of the effectiveness of soil health

approaches have been quite limited and poorly documented. Thus, although there is

some support for soil health approaches among Research & Development leaders and

among specific sectors of the horticultural industry in the Pacific (especially those

involved in the Pacific Organic movement), the dominant paradigm in the research and

extension community in the Pacific remains that of conventional soil science (and

market incentives to boost production). In view of the rapid erosion of the natural

5

resource base that is occurring, as Pacific islands seek to intensify agricultural

production to meet economic aspirations, without in general being able to offer growers

new and more sustainable technologies to underpin this intensification, changes in

production practices and land management are urgently needed (Maathuis and van Meer,

2003).

This research, based on soil health concepts, focuses on selecting those soil

improvement tactics, chosen from a range of options that have proven effective

elsewhere, that are best adapted to local social and environmental conditions. Emphasis

had been placed on cost-effectiveness, so that experiences gained in pilot sites can be

used as a model for developing and advocating solutions for comparable problems

elsewhere. Emphasis is also placed on the development of 'indicators' that can be used

by researchers, extension workers and farmers (at least the more progressive among

them), to monitor their progress towards restoring the biological health of soils. Because

soil processes can be cryptic and hard to measure directly, these indicators need to be

rapid, inexpensive tests that can provide evidence of the current status of soils (Carter et

al., 1997).

This project focuses on testing, with farmers, the best-bet strategies for increasing soil

organic matter content, supported by developing research-based indicators that growers

and extension officers can use to assess soil health status (including key chemical,

physical and biological variables).

1.1 Research Objectives

The objectives of this research are to:

1.1.1 Form a minimum data set (MDS), including selected soil physical, chemical and

biological characteristics as indicators of soil health and quality to be used for

assessment of the four trial sites in Samoa;

1.1.2 Compare the dry matter yield and nutrient uptake of the fallow crop treatments

and biochar as influenced by site differences;

6

1.1.3 Compare the yield response of two taro varieties as influenced by different

fallow crop treatments and biochar;

1.1.4 Investigate nutrient use efficiency for two improved taro varieties grown under

semi-controlled screen house environment, through regular destructive sampling;

1.1.5 Quantify and compare the changes in selected soil biochemical parameters,

namely, labile carbon, fluorescein diacetate hydrolysis activity (soil biological

activity), and potentially mineralisable nitrogen, as influenced by the different

fallow treatments, between the wet and the dry zones of the islands of Savaii and

Upolu, over the fallow duration, and the subsequent decomposition (and taro

growing) phase;

1.1.6 Study fluxes of soil mineral nitrogen using the covered core in-situ incubation

method, as influenced by the different fallow treatments, between the wet and the

dry zones of the islands of Savaii and Upolu, over the decomposition and

subsequent mineralisation of the fallow cover crops and biochar;

1.1.7 Investigate the existence of any relationships between the evaluated biochemical

indicators;

1.1.8 Examine the effects of site variation (for Upolu sites) on the decomposition phase

of different rates of cover crops under semi-controlled screen house conditions;

1.1.9 Compare the response of nematode numbers to mulch application and resulting

decomposition of different cover crops, for Upolu sites; and,

1.1.10 Investigate the associations between the evaluated biochemical soil parameters.

7

CHAPTER 2

LITERATURE REVIEW

2.1 Background on Samoa Located in the South Pacific between 15-17o S and 171 and 173o W, Samoa consists of

four volcanic islands and a series of islets (Plate 2.1). Upolu and Savaii are the two main

islands. The climate of Samoa is humid tropical and receives an annual rainfall varying

from 2,000 to 7,000 mm with a strong seasonality of distribution. The average annual air

temperature is 24–32oC (Maathuis and van Meer, 2003). The climate is suitable for

growing a wide range of tropical crops, and many that thrive here include coconuts

(Cocos nucifera), banana (Musa spp.), mango (Mangifera indica) and a wide range of

other fruits, vegetables, root crops and flowers. There is a marked wet season

(November - April) and a relatively dry season (May - October). Heavy rains are mainly

experienced during the rainy season that runs from November until May. Much of the

islands’ terrain is steepland that reaches up to 500 metres in elevation (Morrison, 1991).

The hills dissect the area, causing steep slopes that are completely covered with tropical

rainforest or shrub, in which up to now agriculture is practiced traditionally. The islands

are composed of a mass of successive olivine basalt flows, varying in age from the early

Pleistocene to the present century (Kear and Woods, 1959). Due to this, the Samoan

soils are mainly latosolic soils derived from basalt and basic andesite. Under the USDA

classification, many Samoan soils are classified as ‘Inceptisols’.

In Samoa, like the other Pacific island countries, the increasing population pressure and

emerging trends of socio-economic marginalisation of rural population are putting heavy

strain on the delicate ecosystems. Since there is greater use of minimum tillage practices

on small fragmented farm holdings, the traditional agricultural systems, such as shifting

cultivation and mixed agroforestry systems, are considered to be most appropriate and

sustainable. Traditional agriculture is completely interwoven with the forest areas, as an

integral part of the food security system of the villages and provides protection against

natural disasters. Although Samoa was the first South Pacific island nation to gain

8

independence and the first to join the United Nations, it still remains among the poorest

in the world, with most of its people living at the subsistence level (Maathuis and van

Meer, 2003). In order to find ways to economically develop the country, Samoa is ever

more searching for ways to replace this traditional system of cash crop farming. Besides

the need for the cash crop, the increasing population causes a rapid increase in food

demand. Therefore, farmers could abandon traditional farming systems to adopt high

input (commercial) production methods to both satisfy the need for domestic food

supply and exportation of cash crops.

Shifting cultivation is the traditional agricultural practice of Samoan farmers and is a

deeply rooted tradition; many decades of contact with other farming techniques have

scarcely altered the basic practices. It is still by far the easiest way of raising crops and

requires only a very small outlay on agricultural inputs (Wright, 1963).

It is fairly certain that, in the foreseeable future, shifting cultivation will continue to be

an integral part of the Samoan way of life, and any plans for accelerated agricultural

development in the territory must allow for a substantial volume of produce grown in the

old Samoan way (Maathuis and van Meer, 2003). It is, however, most desirable that

shifting cultivation is restricted to the raising of subsistence crops rather than export

crops. It is folly to encourage the growing of export crops using a technique that is

wasteful of the soil resources of the country (Wright, 1963).

9

Pl

ate

2.1

Lo

catio

n of

Sam

oa

10

2.2 Soil ecosystems

Soil ecosystems are highly complex, containing a tremendous amount of species.

Indigenous microbial populations in soil are of fundamental importance for ecosystem

functioning, through determining nutrient cycling, organic matter decomposition and

energy flow (Doran and Zeiss, 2000). Despite all attempts to measure fluxes and gross

microbial pools, the soil and its microbiota still remain a ‘black box’. Most soil

microorganisms are still unknown, while very few have been isolated, cultured and

identified, and directly related to their function in agroecosystems. Although new culture

media have been recently developed to maximise the recovery of diverse microbial

groups from soils (Balestra and Misaghi, 1997; Mitsui et. al., 1997) the comparison

between direct microscopic counts and plate counts indicates that less than 0.1% of

agricultural soil microorganisms are culturable (Atlas and Bartha, 1998).

In recent years, the approaches for studying soil microbiota have moved from

biochemical and microbiological determinations such as enzyme activities, microbial

biomass and respiration coefficients towards the investigation of bacterial diversity and

microbial community structure (Hill et al., 2000).

Little information is available about the contribution of soil microbial diversity in soil

ecosystem functioning. An important benefit of such diversity may be to provide greater

resistance to environmental stresses and external disturbance. Several studies (Baath,

1989; Engelem et al., 1998; Jonsen et al., 2001; Smit et al., 1997) have found decreased

levels of microbial biomass and diversity, and altered community structure due to

disturbance mostly caused by pesticides, heavy metals and sludge amendments.

However, it is not precisely known whether a decreased diversity of soil organisms will

cause declines in resistance to external stresses and how it affects soil capacity to

function as a vital living system (Degens et al., 2001), largely because soil microbes

have greater functional redundancy than higher organisms (Othonen et al., 1997). This

would mean that, even though anthropogenic activities affect the genetic composition of

soil microbial communities, gross microbial processes and their potential role in

11

maintaining soil quality might remain unaffected. On the other hand, natural or human-

induced perturbations may influence the level of soil microbial activities without causing

compositional shifts in soil community structure (Giller et al., 1997).

2.3 Soil degradation and soil health

Many agricultural practices increase the soil’s vulnerability to degradation processes

such as erosion, acidification, salinisation, soil structure decline and contamination.

These degradation processes reduce the functional capacity of soils and, at a catchment

level, can reduce the quality of water draining to streams and rivers. Hence soil and

water quality degradation can be thought of as symptoms of poor soil health (Zhang et

al., 1996). The challenge for management of agricultural soils is to develop production

systems that not only prevent soil degradation but also enhance soil health. The

biological component of the soil system has a high dependence on the chemical and

physical soil components and hence tends to be a sensitive indicator to disturbance or

degradation processes (Doran et al., 1996b).

Ecosystem functions can be characterised in terms of their resistance to change by an

imposed disturbance and their resilience, or potential to recover following

disturbance/degradation (Pimm, 1984). These concepts are equally valid for assessing

the sustainability of agricultural production systems (Herrick, 2000). Useful indicators to

evaluate the sustainability of different management practices may be the amount and

rate of change in soil biological functions, and the amount and rate of recovery. The

most sustainable practices will be those which cause little or no negative change in

functional capacity and/or which enable rapid recovery. Some soil properties and

functions undergo changes when disturbed that are effectively irreversible within

management time scales.

12

2.4 Properties and indicators of healthy soils

2.4.1 Properties of healthy soils

Soil health integrates all components of the soil system and is assessed by indicators that

describe or quantify biological, chemical and physical properties (Karlen et al., 2001).

Attributes of a healthy soil include:

i. protected soil surface and low erosion rates;

ii. high soil organic matter;

iii. high biological activity and biological diversity;

iv. high available moisture storage capacity;

v. favourable soil pH;

vi. deep root zone;

vii. balanced stores of available nutrients;

viii. resilient and stable soil structure;

ix. adequate internal drainage;

x. favourable soil strength and aeration;

xi. favourable soil temperature;

xii. low levels of soil borne pathogens; and,

xiii. low levels of toxic substances (Karlen et al., 2001).

2.4.2 Measurement of soil quality

Soil quality is simply defined as “the capacity of a specific kind of soil to function.” The

concept of soil health and soil quality has consistently evolved with an increase in the

understanding of the soil and soil quality attributes (Karlen and Stott, 1994). Soil quality

cannot be measured directly, but soil properties that are sensitive to changes in

management can be used as indicators (Andrews and Cambardella, 2004). Soil health

indicators are needed that help smallholder farmers to understand the chain of cause and

effect that links farm decisions to ultimate productivity and health of plant and animals.

The soil health approach is better applied when specific goals are defined for a desired

13

outcome from a set of decisions. Therefore, the soil health and quality evaluation

process consists of a series of actions:

� Selection of soil health indicators

� Determination of minimum data set (MDS)

� Development of an interpretation scheme of indices

� On farm assessment and validation

2.4.3 Soil health indicators

The quality of soil is rather dynamic and can affect the sustainability and productivity of

land use. It is the end product of soil degradative and conserving processes and is

controlled by chemical, physical and biological components of the soil and their

interactions as shown in Figure 2.1 (Papendick and Parr, 1992). That is, soil quality

integrates the physical, chemical, and biological components of soil and their

interactions. Therefore, when measuring soil quality, it is important to evaluate the

physical, chemical, and biological properties of the soil so as to capture the holistic

nature of soil quality or health. Indicators, however, will vary according to the location

and the level of sophistication at which measurements are likely to be made (Riley,

2001).

Figure 2.1 Complex components of soil health/quality indicators

Indicators selected based on management goals

Physical Indicators

Chemical Indicators

Biological Indicators

Scoring Functions

SOIL HEALTH / QUALITY INDEX

14

Measurement procedures and guideline optimum ranges for soil chemical, physical and

biological properties are available (USDA, 1999). However, agricultural productivity is

determined by a large number of direct and indirect interactions between plant and

animal characteristics, climatic conditions, soil properties, pest conditions and

management practices. Satisfactory crop production may still occur if soil properties are

outside guideline ranges because of plant tolerance, compensatory climatic conditions,

or compensatory management practices. Hence, it is usually not possible to predict

animal or crop production from soil properties alone. This limits the value of generalised

soil quality guidelines.

2.4.3.1 Biological indicators

Identification of biological indicators of soil quality is reported as critically important by

several authors (Doran and Parkin, 1994; Abawi and Widmer, 2000) because soil quality

is strongly influenced by microbiologically mediated processes (nutrient recycling,

nutrient retention, aggregation). Biological indicators of soil quality that are commonly

measured include soil organic matter, respiration, microbial biomass and mineralisable

nitrogen. Soil organic matter plays a key important role in soil function, determining soil

quality, water holding capacity and susceptibility of soil to degradation (Giller and

Cadisch, 1997; Feller et al., 2001). In addition, soil organic matter may serve as a source

or sink to atmospheric CO2 (Lal, 1997) and an increase in the soil C content is indicated

by a higher microbial biomass and elevated respiration (Sparling et al., 2003). It is also

the principal reserve of nutrients such as N in the soil and some tropical soils may

contain large quantities of mineral N in the top 2 m depth (Havlin et al., 2005).

Nematodes can be used as bio-indicators of soil health because they are ubiquitous and

have diverse feeding behaviours and life strategies, ranging from colonisers to persisters.

Some nematodes can survive harsh, polluted, or disturbed environments better than

others, and some have short life cycles and respond to environmental changes rapidly

(Bongers and Bongers, 1998; Neher, 2001).

15

2.4.3.2 Chemical indicators

In order to achieve high crop yields smallholder farmers have to provide soil nutrients in

large quantities (Sanchez and Swaminathan, 2005). Therefore, it is possible to alter the

pool of available nutrients by adding inorganic fertilisers, incorporating cover crops, and

using other organic materials in the form of manures and composts (Stocking, 2003).

Results of chemical tests are soil quality indicators which provide information on the

capacity of the soil to supply mineral nutrients, which is dependent on the soil pH. Soil

pH is an estimate of the activity of hydrogen ions in the soil solution. It is also an

indicator of plant available nutrients. High activity is not desirable and soil may require

liming with base cations Ca or Mg in order to bring the soil back to neutral.

2.4.3.3 Physical indicators

Soil physical properties are estimated from the soil’s texture, bulk density (a measure of

compaction), porosity and water holding capacity (Hillel, 1982). The presence of hard

pans usually presents barriers to rooting depth. These properties are all improved

through addition of organic matter. Therefore, the suitability of the soil for sustaining

plant growth and biological activity is a function of its physical properties (porosity,

water holding capacity, structure, and tilth).

2.4.3.4 Minimum data set (MDS)

Soil quality assessment or interpretation should be considered a process through which

soil resources are evaluated on the basis of soil function (what the soil does) and change

in soil function in response to a specific natural or introduced stress, or management

practice.

A minimum data set is used to measure soil quality and its changes due to management

practices through selection of key indicators such as soil texture, organic matter, pH,

16

nutrient status, bulk density, electrical conductivity and rooting depth (Larson and

Pierce, 1994). Collecting a minimum data set helps to identify locally relevant soil

indicators and to evaluate the link between selected indicators and significant soil and

plant properties (Arshad and Martin, 2002). It is a minimum set of indicators required to

obtain a comprehensive understanding of the soil attributes evaluated. More importantly

they serve as a useful tool for screening the condition, quality and health of soil (Doran

et al., 1996; Larson and Pierce, 1994; and Doran and Parkin, 1994).

It is also important to emphasize that soil quality evaluations must consider biological,

chemical, and physical properties and processes. For interpretation, the measurements

must be evaluated with respect to their long-term trends or signs of sustainability. A

general sequence of how to evaluate soil quality is to (1) define the soil functions of

concern, (2) identify specific soil processes associated with those functions, and (3)

identify soil properties and indicators that are sensitive and considered to be an

indication of the level of functioning. However, indicator data is not meaningful unless a

baseline or some reference condition is available for comparison or unless relative

comparisons between management systems are made (Doran et al., 1996; Doran and

Parkin, 1994).

2.5 Roles of nematodes in soil nutrient cycling

Detritus and organic residues must decompose to release nutrients for plant uptake.

Decomposition of organic matter in a soil food web can be divided into two energy

channels, a faster bacterial channel and a slower fungal-based channel. Soil ecosystem

types and nutrient forms (e.g. C:N ratios) determine the predominant decomposition

channels (Ferris and Matute, 2003; Ingham et al., 1995). Although bacteria and fungi are

the primary decomposers in the soil food web (Figure 2.2), these microbes also can

immobilise inorganic nutrients in the soil (Ingham et al., 1995). As an extension of these

decomposition channels, when the bacterivorous and fungivorous nematodes graze on

these microbes, they give off CO2 and NH4+ and other nitrogenous compounds, affecting

C and N mineralisation directly (Ingham et al., 1995). Indirectly, nematodes can

17

disseminate microbial propagules throughout the soil (Freckman, 1988.), which

advances the colonisation of substrates and mineralisation of nutrients. Nematode

metabolites may also stimulate specific bacterial growth by releasing growth-limiting

nutrients (such as N and vitamins). However, overgrazing of bacterial or fungal

populations by nematodes can result in a reduction of the overall activity of these

decomposers. Fortunately, in the hierarchy of the soil food web, generalist predators

prey on these bacterivorous and fungivorous nematodes, improving nutrient cycling and

allowing more nutrients to be released (Yeates and Wardle, 1996).

Figure 2.2 Roles of nematodes in organic matter decomposition (Source: Ingham et al., 1985)

Therefore, nematodes play important roles in soil nutrient cycling. Nematode excretion

may contribute up to 19% of soluble N in soil (Neher, 2001). This is due to the fact that

nematodes (C:N ratio of 8-12) have a lower N content than the bacteria (C:N ratio of 3-

4) they consume (Wasilewska and Bienkowski, 1985). In addition, the growth

efficiency of nematodes (< 25%) is smaller than those of the bacteria (> 30%) (Hunt et

Detrital - N, P, K

Fungal – N, P, K

Inorganic – N, P, K

Plant– N, P, K

Bacterial – N, P, K

Fungal – feeding nematode Bacterial – feeding nematode

Omnivorous and predatory nematode

18

al., 1987). Therefore, nematodes excrete a majority of both the assimilated C and N that

they consumed from the bacteria. Bacteria, on the other hand, usually respire most of the

assimilated C, but immobilise most of the assimilated N. Therefore the contribution

made by nematodes to N mineralisation is relatively high compared to bacteria in soil

ecosystems.

Besides contributing to N mineralisation, the abundance of many free-living nematodes,

especially bacterivorous, omnivorous, and predatory nematodes, also correlates with

concentrations of many other soil nutrients in a fallow field (Wang et al., 2004),

suggesting the possibility of nematodes mineralising many other soil nutrients. More

clear-cut relationships between nematodes and soil nutrients has been observed in a field

that had been in fallow for 1.5 years compared to a recently cultivated field (Wang et al.,

2004).

2.6 Goals in developing nematode management soil ecosystem

Maintaining soil nutrient availability and plant-parasitic nematode suppression are two

of the most important issues in nematode management for soil health. Plant-parasitic

nematodes cause damage to plant roots, resulting in root systems which are less able to

take up nutrients and water. Enhancing soil nutrient availability not only supplies

nutrients for plant uptake, but also provides plants with materials needed to grow

functional roots, thus increasing the plant tolerance to nematode damage. On the other

hand, a great resource in most soil ecosystems for suppressing plant-parasitic nematodes

is the pool of natural enemies of nematodes in the soil.

Therefore, the overall goal is to develop soil ecosystem nematode management systems

that encompass: (i) the enhancement of free-living nematodes that are significantly

involved in soil nutrient cycling; (ii) the suppression of multiple nematode pests; (iii) the

enhancement of natural enemies of plant-parasitic nematodes, and (iv) he improvement

19

of plant health. Several studies of nematode communities provide useful examples of

how some of these goals may be achieved.

Enhancing free-living nematodes that are significantly involved in soil nutrient cycling.

Due to their efficiency in nutrient mineralisation, particularly N, the abundance of

bacterivorous nematodes may be a useful indicator of enhanced soil nutrient cycling

(Freckman, 1988). Ferris and Matute, (2003), specified that to provide a healthy soil

ecosystem with adequate soil fertility, enrichment-opportunists need to be maintained at

a high level.

Suppressing multiple nematode pests. Conventional, non-chemical management

practices for plant-parasitic nematodes often target the suppression of a single nematode

pest. For instance, some crop rotations or inundative biological control methods may

target only a single pest. In soil ecosystem nematode management, ideally one should

aim for suppressing multiple plant-parasitic nematode species. This might be difficult to

achieve with a single procedure. However, the use of a cover crop that is a poor host to

multiple species of plant-parasitic nematodes might be closer to this goal. For instance,

sun hemp (Crotalaria juncea), used as a cover crop, was suppressive to root-knot

nematodes as well as reniform nematodes (Wang, et al., 2002).

Enhance natural enemies of plant-parasitic nematodes. Some scientists believe that

using soil amendments can increase the activity of free-living nematodes and enhance

soil suppressiveness to plant-parasitic nematodes. Van den Boogert et al. (1994)

supported this hypothesis and concluded that organic matter stimulated bacteria

numbers, provided a food base for free-living nematodes, which in turn became a food

source for nematode-trapping fungi. Some nematode antagonistic fungi occupy the

rhizosphere in preference to the general soil mass. Therefore, planting cover crops

would be preferable to fallowing between crop seasons to enhance the populations and

activity of these antagonists.

20

Parasitism of some nematode antagonistic fungi is dependent on the population densities

of host nematodes, a phenomenon known as host density-dependent parasitism.

However, dependency on nematodes varies among nematophagous fungi; those that

produce adhesive networks appear to be less dependent on nematodes than are those that

produce constricting rings and adhesive knobs (Cooke, 1963). Jaffee et al., (1993)

demonstrated that the greatest dependency on host density was observed in Hirsutella

rhossiliensis (an endoparasitic fungus), intermediate for Monacrosporium ellipsosporum

(a trapping fungus forming adhesive knobs) and A. dactyloides (a trapping fungus

forming constricting rings), and least for A. oligospora (a trapping fungus forming

adhesive three-dimensional networks) and M. cionopagum (a trapping fungus forming

adhesive branches and two-dimensional adhesive networks). Therefore, it is not

surprising to find higher population densities of host density-dependent nematode-

antagonistic fungi in soils heavily infested with plant-parasitic nematodes.

Improve plant health. The ultimate goal of soil ecosystem management is to improve

plant health. Amending soil with organic matter is a basic practice for soil ecosystem

management because even though this practice does not always suppress soil pests, it

still can increase crop yields, which is the main concern of a grower. In a field trial of

yellow squash, amendment with yard-waste compost (very high C:N ratio, minimal

nutrient release) at 269 metric t/ha, either incorporated into the soil or as mulch, did not

suppress M. incognita population densities at the end of the experiment. However,

incorporating compost resulted in an increase of squash yield compared to the control

treatment, indicating an increase in plant tolerance against plant-parasitic nematode

infection, attributed to improved water-holding capacity in mulched plots (McSorley,

and Gallaher, 1995).

2.7 Importance of maintaining soil functional diversity

Stability of soil ecosystems. Microbial diversity in soil is usually assessed as species or taxa

diversity. However, structural and functional diversity are more important to soil health.

Functional redundancy, which refers to a reserve pool of quiescent organisms or a community

21

with vast interspecific overlaps and trait plasticity, allows an ecosystem to maintain long-term

stability of soil function. Considerable functional redundancy exists and merits protection because

this may be why decomposition processes are maintained in highly disturbed soils despite the

extinction of many species (Ettema, 1998).

Succession on nematode communities. During the decomposition of organic matter with a mixture

of C:N ratios, populations of enrichment-opportunist bacterivorous nematodes increased rapidly

in response to low C:N materials, and to a lesser extent, to more complex materials. The general-

opportunist bacterivores increased at a slower rate. Fungivorous nematodes increased gradually as

higher C:N ratio residues became more abundant, but increased most rapidly in soil amended with

higher C:N and more complex materials (Ferris and Matute, 2003). These in turn were replaced

by persisters which encompass most of the omnivores and predators (Figure 2.3).

Green manure Recalcitrant plant

(low C:N) materials (high C:N)

= Increase in the abundance of the nematode specified over time

= Mineralisation

Figure 2.3 Nematode community succession in relation to C:N ratios of soil amendments.

Succession of nematodes is not limited to the trophic group level, but also occurs among taxa

(genera or species) within a feeding group. In general, nematode succession follows the pattern as

demonstrated in a typical succession of nematodes after adding organic material to the soil (Wang

et al., 2013).

Bactivores, Ba1

Omnivores / Predators

Bacterial

decomposition

pathway

Bactivores, Ba1

Fungal

decomposition

pathway

Fungivores,

22

2.8 Indicative value of nematode trophic group abundance and food web indices

Yeates et al. (1993) categorised nematodes into five generally recognised trophic groups:

bacterivores, fungivores, predators, omnivores and plant parasites. Bongers (1990) classified

nematodes along a coloniser-persister (c-p) continuum of 1-5. Nematodes with c-p value equal to

one are short lived, have high fecundity, feed on enriched media whereas those of c-p value five

are have large body size, longer life span, low fecundity, susceptible to disturbance and are

predominantly omnivores and predators (Bongers, 1990). C-p classifications of nematodes lead to

the formation of the maturity index (MI), which is a weighted mean frequency of c-p scaling

across the entire nematode community and provides the information of the likely condition of the

soil environment (Bongers, 1990). The development of MI represented a significant advancement

in interpreting the relationships between the ecology of nematode communities and functions of

the soil (Neher et. al., 2005). However, Ferris (1993) argued that use of trophic groups could often

lead to ambiguous results since the trophic groups encompass an enormous diversity of life

history and physiological characteristics. Calculation of the MI index also assumes a progression

of soil conditions from stressed or polluted to pristine exactly congruent with the continuum of

nematode life history characteristics in the c-p classification as suggested by Bongers (1990).

Ferris et al. (2001) observed that the most abundant nematode taxa under stressed conditions are

those in c-p 2, while the enrichment opportunists (c-p 1) respond positively to disturbances that

result in enrichment of the food web. Therefore, in an attempt to improve the indicator capabilities

of nematodes, Ferris et al. (2001) assigned weights to indicator nematode guilds representing

basal, enriched and structured conditions of the food web. This concept leads to the development

of food web indices including enrichment index (EI) and structure index (SI). EI is based on the

expected responsiveness of the opportunistic guilds (bacterivore nematodes with c-p value equals

one) to organic resources enrichment. Therefore, EI describes whether the soil environment is

nutrient enriched (high EI) or depleted (low EI). SI represents an aggregation of functional guilds

with c-p values ranging from 3-5 and describes whether the soil ecosystem is structured with

greater trophic links (high SI) or degraded (low SI) with fewer trophic links (Ferris et al., 2001).

Plotting of EI and SI provide a model framework of nematode faunal analysis as an indicator of

the likely conditions of the soil food web (Figure 2.4).

23

Figure 2.4 Indicator guilds of soil food web condition (basal, structured, enriched) are designated and

weightings of the guilds along the structure and enrichment trajectories are provided, for

determination of the enrichment index (EI) and structure index (SI) of the food web (Ferris

et al., 2001). The functional guilds of soil nematodes are characterised by their feeding

habits or carbon flow channels and by life history expressed along a coloniser-persister (c-

p) scale (cp scale proposed by Bongers and Bongers, (1998). Bax

(bacterivores), Fux

(fungivores), Cax

(carnivores), Omx

(omnivores) (where value of x = 1-5 on the c-p scale)

represents various functional guilds.

Further, Ferris et al. (2001) also proposed the channel index (CI), which is a percentage of

fungivores among the total fungivores and c-p one opportunists bacterivores. CI indicates

predominant decomposition channels in the soil food web, a high CI (> 50 %) indicates fungal

decomposition channels whereas low CI (< 50 %) suggests bacterial decomposition channels

(Ferris et al., 2001). In later studies use of these indices provided critical information about below

ground processes in distinct agroecosystems (Bulluck et al., 2002b; Ferris and Matute, 2003;

Neher et al., 2005).

2.9 Agricultural practices compatible with soil ecosystem management

Two major characteristics of farming systems that are compatible with soil ecosystem

management are the addition of organic amendments and avoiding the application of synthetic

24

pesticides, similar to organic farming practices. A major concern in organic systems is the

maintenance of adequate soil fertility at key crop-growth periods. Availability of nutrients from

an organic matter source varies substantially among different kinds of organic matter ((Ferris and

Matute, 2003), mainly due to differences in C:N ratios. With low C:N ratio organic substrates,

bacterial growth is C-limited, and N-immobilisation by microbes will be minimal; but at high C:N

ratios, bacterial growth will be N-limited, and there may be rapid immobilisation of newly

mineralised N. Therefore, it is critical to select an organic amendment with a low C:N ratio if

rapid nutrient availability for the crop plant is needed. If high C:N ratio amendments are used, soil

food webs will be selected for fungal dominated decomposition pathways, thus a slower

mineralisation rate but a longer lasting supply of organic materials will be available (Wang et. al.,

2004).

Practices that enhance soil health. Although an important practice that enhances soil health is the

use of organic matter, the decomposition rate and products of organic materials in the soil depend

on their nature and C:N ratios and the time-course of decomposition. Nitrogen may become

immobilised in microbial tissue when organic material has a C:N ratio greater than 20:1 but

mineralised in the form of NH4+ or NO3

- when the C:N ratio is less than 20:1 (Ferris and Matute,

2003).

Ecosystem monitoring. Monitoring the soil ecosystem prior to the cropping season will help to

determine if the practices selected are compatible with the achievement of goals of soil ecosystem

management. For example, if natural enemies of major pests in a particular site are not present,

bio-control agents might be added to the soil. Since the history of agricultural practices can affect

soil ecosystems, knowing the history of an agricultural site can also be helpful in making

decisions (Ellert et al., 1997). For instance, rates of cover crop decomposition are more consistent

in organically managed soils than in conventionally farmed soils. In addition, the C:N ratio of the

cover crop (higher for small grains, lower for legumes) will affect the rate of its decomposition.

Besides monitoring biotic factors, monitoring abiotic factors in soil also will be important because

nematode community structure is influenced by a combination of previous land use, and soil

factors such as soil type and texture.

25

Timing. The timing of application of organic matter prior to crop planting differs among the

specific fertilisers used. Therefore, growers could time planting and fertiliser application so that

the crop will be ready to take up the nutrients within 2 weeks after the residues are applied. On the

other hand, when using a manure with the relatively high C:N ratio as a source of organic matter,

it was recommended that the manure be incorporated at the end of the previous crop to increase

the abundance, biomass, and activity of bacterivorous nematodes for the next crop (Ferris et al.,

1996).

2.10 Cover crops

What is a cover crop?

A cover crop is "any crop whose main purpose is to benefit the soil and/or other crops in one or

more ways, but is not intended to be harvested for feed or sale" (Roos, 2006). Cover crops have

been part of agriculture for at least a few thousand years, but have recently received renewed

attention as the result of environmental and economic concerns.

Desirable Characteristics of Leguminous Cover Crops

The selection of leguminous cover crops is essentially based on the combination of the following

characteristics:

� Very vigorous growth

� Easy establishment and low seed rate

� Non-palatability to animals

� High drought tolerance

� Shade tolerance

� Presence of allelopathic chemicals to enhance competitive ability against weed growth.

� High biomass production

� Tolerance to pest and diseases

� Low labour and chemical requirements for its establishment

� Good control against soil erosion.

Benefits of Cover Crops

Conventional agricultural practices can result in environmental problems such as soil erosion,

26

surface and groundwater pollution, and overdependence on fossil fuels and other inputs. There is

increasing interest in developing sustainable agricultural systems that decrease reliance on

chemical and fossil fuel inputs by enhancing biological processes. Cover crops are an important

component of a sustainable system (Craig and Wilkinson, 1998).

Improve soil fertility

There is increasing interest in the use of cover crops to improve soil fertility. Much of this interest

stems from a heightened awareness of the negative environmental impact of synthetic fertilisers,

including ground- and surface water contamination, long-term soil productivity, and the energy-

intensiveness of fertiliser production. Cover crops are capable of trapping residual nitrogen in the

soil and, in the case of legumes, fixing atmospheric nitrogen.

A successfully established leguminous cover crop can replace some or all of the nitrogen fertiliser

needed to produce crops. Both legumes and non-legumes can help recycle and increase the

availability of phosphorous, potassium and micronutrients (Roos, 2006). The quantity and

availability of nitrogen provided by the cover crop depends on many factors:

� the current level of nitrogen in the soil - legume nitrogen fixation is reduced by 6.2 kg per

hectare for every one kilogram of available soil nitrogen; to facilitate nitrogen fixation,

precede the cover crop with a crop that will uptake high levels of nitrogen

� the cover crop species - different legume species contribute varying quantities of nitrogen;

also, the nitrogen content of the same legume species can vary according to environmental

conditions and management strategies.

� the growth stage of the cover crop when killed - some studies have shown that the highest

nitrogen levels are achieved when the cover crop is killed at the full bloom or pod stage;

however, yield loss could occur if the cover crop delays planting; for best results, time cover

crop planting and incorporation so that they don't interfere with optimal cash crop planting

schedules.

� landscape position - one study revealed that nitrogen fixation by peas was higher on

bottomland than on slope and ridge sites

� method of cover crop suppression and incorporation - tillage operations can affect nitrogen

27

availability; a leguminous mulch may be vulnerable to nitrogen loss from volatility in areas

of high moisture and temperature (Roos,2006).

Improve soil structure and reduce soil erosion

Many farmers are now planting cover crops on land that they used to leave bare in fallow periods.

Cover crops can reduce the impact of raindrops on the soil surface and minimise runoff. Cover

crops increase the organic matter content of soils and improve infiltration. Cover crop biomass

production depends on many factors, including soil type, environment, fertilisation, planting date,

and time of kill. Deep-rooted cover crops can break up plough pans and improve soil tilth and

water-holding capacity (Schouten et al., 2000).

Suppress weeds

Because so few herbicides are registered for vegetables, many farmers are interested in the effect

of cover crops on weed populations. Increasing public concern about herbicides has also

contributed to the interest in alternative weed management strategies. Many cover crops and their

residues can suppress weed growth by altering light and temperature. Cover crops also present a

barrier to emerging weed seedlings (Schouten et al., 2000).

Another way that cover crops suppress weed emergence is through allelopathy, the release of

toxic compounds by one plant to a neighboring plant. Many cover crops have exhibited

allelopathic effects. Rye residues produce allelopathic effects that can suppress weeds for 30-75

days after the cover crop has been killed (Roos, 2006). The use of rye cover crops in tomato

production can eliminate the need for soil-applied herbicides.

Rapeseed, another brassica, provided similar weed control as a herbicide in one study, reducing

weed density and biomass 73-85% and 50-96%, respectively (Roos, 2006). However, excessive

rapeseed residues can also have a negative effect on cash crop growth and development. An

integrated weed management approach that supplements the use of cover crops with timely

cultivation and/or herbicide application is recommended for optimal results.

28

Reduce insect problems

Cover crops can be an important part of an integrated pest management program. Cover crops can

attract beneficial insects. However, caution is needed with this approach, since they can also

attract pests. Pest management is a complex issue that warrants further research.

Cover crop management affects their capacity to attract beneficial insects. Usually, cover crops

are incorporated soon after flowering but this may have a negative effect on beneficial insect that

inhabit the cover crop. Insects that can reproduce and mature quickly are less likely to be harmed

(Siciliano and Roy, 1999). Method of incorporation can have an impact on insect populations.

Mowing is the most damaging while no-till causes the least disturbance.

Reduce disease and nematode problems

Some disease problems can be reduced with an appropriate cover crop rotation. Incorporated

alfafa and white sweet clover residues can reduce the fungus Sclerotium rolfsii, while rye has

demonstrated a capacity to reduce the incidence of Pythium. Cover cropping can also suppress

plant-parasitic nematode populations. Sorghum, hairy indigo, cowpea, and jointvetch have all

reduced nematode populations (Schouten et al., 2000).

Reduce groundwater contamination

Groundwater contamination caused by leaching nitrates (NO3--N) from residual nitrogen

fertilisers and the mineralisation of soil organic matter is becoming a serious problem in crop

production in some areas, especially during the winter fallow (Roos, 2006). Non-leguminous

cover crops can help ameliorate the problem because they are capable of immobilising as much as

70% of the available NO3--N in the upper soil profile. To effectively reduce nitrate leaching, a

cover crop must grow rapidly and produce an extensive root system conditions without

supplemental inputs (Siciliano and Roy, 1999). Cover crops such as annual ryegrass, grain rye,

and brassicas have all reduced nitrate leaching. Generally, non-legumes are about three times

more efficient at reducing nitrate leaching than legumes.

29

Save money

Cover crops can potentially reduce production costs by decreasing fertiliser and pesticide bills.

They may also cut fuel and irrigation costs since deep-rooted cover crops can break up plough

pans and improve soil tilth and water-holding capacity, reducing fuel and irrigation costs.

Disadvantages of Using Cover Crops

Cover crops have some potentially negative effects:

� depletion of soil moisture

� reduced spring soil temperatures

� disruption of field operations

� allelopathy

� habitat for pests and disease

2.11 The importance of organic matter in soil fertility and crop health

What is organic matter?

Organic matter is the vast array of carbon compounds in soil. Originally created by

plants, microbes, animals and other organisms, these compounds play a variety of roles

in nutrient, water, and biological cycles. For simplicity, organic matter can be divided

into two major categories: stabilised organic matter which is highly decomposed and

stable, and the active fraction which is being actively used and transformed by living

plants, animals, and microbes. Two other categories of organic compounds are living

organisms and fresh organic residue (Figure 2.5). These may or may not be included in

some definitions of soil organic matter (Grundon, 2009).

Stabilised organic matter

Many soil organisms decompose plant and animal tissues, and transform the organic

matter into new compounds. After years or decades of these transformations, what

remains are large, complex compounds that few microbes can degrade. Other

30

compounds become bound inside soil aggregates where microbes cannot reach. These

hard-to-decompose, or stabilised, substances make up a third to a half of soil organic

matter (Grundon, 2009). Scientists often divide stabilised organic matter into three

chemical groups: humic acids, fulvic acids, and humins. Fungi and actinomycetes create

many of the humic acids that combine in soil to become stable compounds. Much of the

stabilised matter in soils originated from plants that grew one or more centuries ago.

Some of these old compounds are bound to clay, and are important in gluing together

tiny aggregates of soil particles (Goh, 1980).

Stabilised organic matter acts like a sponge and can absorb six times its weight in water.

In sandy soils, water held by organic matter will make the difference between crop

failure or success during a dry year.

Figure 2.5 Organic compounds in soil before cultivation (Goh, 1980).

Living organism (5%)

Fresh organic matter(10%)

Stabilised organic matter(33% to 50%)

Active fraction(decomposing organicmatter) (33% to 50%)

31

Organic compounds in soil

Figure 5 represents organic matter in soil before cultivation. After land has been

cultivated for one or two decades, much of the active fraction is lost and stabilised

organic matter makes up more than half of the soil organic matter (Goh, 1980).

Both organic and clay particles can hold on to nutrients electrochemically - like a

magnet holds on to iron filings (Feller et al., 2001). The amount of nutrients that the

organic compounds and clay could carry and make available to plants is called the soil’s

cation exchange capacity (CEC). Although the amount of clay in a soil cannot be

changed easily, the amount of organic matter in soil can easily be decreased or (with

more difficulty) be increased.

In addition to nutrients, stabilised organic matter holds on to pesticides and

contaminants. This prevents pesticides from moving into water supplies and improves

the decomposition of the compounds, but it also makes pesticides less effective by

preventing their contact with the target organism.

The active fraction

Up to 15% of soil organic matter is fresh organic material and living organisms (Giller

and Cadisch, 1997). Another third to one half is partially and slowly decomposing

material that may last decades. This decomposing material is the active fraction of soil

organic matter.

The active organic matter, and the microbes that feed on it, are central to nutrient cycles.

Many of the nutrients used by plants are held in organic matter until soil organisms

decompose the material and release ammonium and other plant-available nutrients.

Organic matter is especially important in providing nitrogen, phosphorus, sulfur, and

iron. Depending on the rate of decomposition, these nutrients may become available to

plants in a year, but it is difficult to predict the decomposition rate (Goh, 1980).

32

Why focus on soil organic matter?

Many soil properties impact soil quality, but organic matter deserves special attention. It

affects several critical soil functions, can be manipulated by land management practices,

and is important in most agricultural settings across the tropics. Because organic matter

enhances water and nutrient holding capacity and improves soil structure, managing soil

organic matter can enhance productivity and environmental quality, and can reduce the

severity and costs of natural phenomena, such as drought, flood, and disease. In

addition, increasing soil organic matter levels can reduce atmospheric CO2 levels that

contribute to climate change (Greenland, 1988).

One of the greatest challenges producers face when beginning organic management is

providing adequate fertility to meet crop needs. Synthetic fertilisers provide nutrients in

an inorganic form, and are therefore immediately available for uptake by the crop. In the

absence of these fertilisers, organic nutrient sources are needed to supply fertility. These

sources require processing by the soil microbial community before plants can utilise

them (Pusparajah, 1997). Soils that have been under conventional management often do

not support the levels of organic matter to supply plant nutrients, or an active microbial

community to efficiently process those nutrients and make them available to the crop

(Sanchez et al., 1989). However, careful planning of fertility programs can alleviate

nutrient deficiencies that may occur in the transition years, as well as help to build

healthy, disease and pest resistant soils and crops.

Soil organic matter is the most fundamental source of fertility in organic agriculture and

it is important for producers to understand the basics of organic matter cycling in the soil

(Goh, 1980). Soil organic matter is that portion of the soil that consists of biological

residues, from plant to animal to microorganism, in various stages of decay. These

residues are decomposed by soil fauna, including relatively large organisms such as

earthworms (macrofauna), nematodes and springtails (mesofauna) and microorganisms

(fungi and bacteria). Depending on the carbon to nitrogen (C:N) ratio of the residues, the

fate of the decomposition products are different. High carbon residues such as corn

33

stalks decompose slowly, because a lack of nitrogen limits the ability of microorganisms

to break the material down. The majority of the nitrogen that is extracted from these

residues is immediately incorporated into the bodies of the soil fauna, therefore making

it unavailable to the plant, at least for a short time (Grundon, 2009). High nitrogen

residues such as legumes, on the other hand, will decompose quickly and, due to an

excess of nitrogen in relation to the needs of the microorganisms, release nitrogen that is

readily available for plant uptake (Craig and Wilkinson, 1998).

Regular additions of organic matter are important as food for microorganisms, insects,

worms, and other organisms, and as habitat for some larger organisms. Soil organisms

degrade potential pollutants, help control disease, and bind soil particles into larger

aggregates. Well-aggregated, crumbly soil allows good root penetration, improves water

infiltration, makes tillage easier, and reduces erosion (Grundon, 2009).

Very fresh organic matter can cause problems to crops in two ways - nitrogen tie-up and

allelopathy. A temporary nitrogen deficiency for crops occurs if the organic matter is

low in nitrogen. "Allelopathic" chemicals are formed when some residues decay, and

can inhibit plant growth.

The various components of plant and animal residues also have various fates. Certain

parts may break down easily, liberating nutrients, while other portions will continue to

be worked on and altered by microorganisms until they can no longer be broken down.

At this point these materials are called humus. Humus is extremely important in

increasing and maintaining soil fertility. It possesses an overall negative charge, which

translates into a very high cation exchange capacity. This means it is able to attract and

effectively store positively charged ions, or cations. As the majority of plant macro and

micronutrients in the soil (with the notable exception of phosphorus) are cations, humus

can be thought of as a bank which holds nutrients and releases them in response to plant

or microorganism secretions. Additionally, nutrients in the soil are subject to a complex

array of chemical reactions and these also affect their absorption and release (Goh,

1980).

34

From the above it can be seen that soil organic matter is the heart of balanced soil

fertility. Organic matter supplies not only readily available nutrient sources but also the

building blocks of humus. Including a broad selection of crops in a rotation ensures

diverse sources of organic matter, and is an important strategy for increasing the overall

organic matter content of the soil (Goh, 1980).

The inclusion of combinations of materials with low and high C:N ratios is very

important. Low C:N ratio materials contain relatively large amounts of nitrogen and

decompose quickly, contributing very little to the building of humus. High C:N ratio

residues, on the other hand, break down more slowly in the soil due to the presence of

more stubborn compounds. These residues increase soil organic matter and humus

contents but contribute relatively fewer readily available nutrients. Therefore a diversity

of crop residues ensures sufficient organic C and N for humus formation and ultimately

produces a pool of potentially available nutrients that can become mobilised according

to crop demand (Greenland, 1998).

Organic matter is more than fertiliser

Organic matter is not just N, P, K, and carbon. Two sources of organic matter with the

same nutrient content or total organic matter content might not have equal effects on

crops and soils.

Plant residues also differ greatly as a source of organic matter. Above-ground growth

has a different action in soil than roots, even when it is tilled into the soil. All roots do

not act the same. For example, tap-rooted plants such as alfalfa create vertical pores in

the soil, whereas the finely branched roots of grasses enhance soil aggregation (Sanchez

et al.,1989).

Organic matter also affects nutrient cycles by chelating (chemically holding on to)

nutrients, and preventing them from becoming insoluble and therefore unavailable to

35

plants. For example, humic substances help make iron available to plants, even in

medium-to-high pH soils.

Soil microfauna and microorganisms mediate the release of these nutrients, and diverse

residue sources sustain a microbial community that is more efficient and has more

functional diversity. For instance, bacterial biomass is associated with readily available

organic matter fractions, while the abundance of fungi increases in relation to the

materials with higher C:N ratios. In general, increasing soil C is linked to greater soil

microbial biomass, which is an important sink and source of nutrients (Greenland,

1988). Although the incorporation of nutrients into living microbial biomass can, at least

initially, reduce availability for plant uptake, over time the cycling of nutrients through

microbial biomass should reach short term qusi-equilibrium, at which point nutrients are

readily available for crop needs. This process is hastened by the presence of predators,

such as bacteria-feeding nematodes, which have been shown to double the rate of N

release (Kennedy and Roughley, 2002).

Furthermore, it is estimated that 20 to 70% of the soil cation exchange capacity is due to

humus, which highlights the importance of organic matter for nutrient storage (Goh,

1980). High organic matter contents also have a positive effect on soil physical

properties. For example, soils with high organic matter contents contain a greater

abundance of water-stable aggregates and have a greater exchange capacity, which

translates into better structure and water-holding and nutrient absorption capacities.

Larger aggregates also slow organic matter degradation, producing a slowly mineralising

pool of nutrients (Sanchez et al.,1989).

Organic matter can also reduce crop attractiveness to insect pests. In fact, plants growing

in soils receiving diverse organic matter inputs have been shown to be less attractive to

some insect pests, as a result of a more nutritionally-balanced growth medium (Grundon,

2009). The effect of fertilisation based predominantly on one nutrient out of balance

with other essential nutrients often leads to an environment attractive to insect pests. In

36

fact, inorganic fertilisers, most commonly nitrogen, are known to increase plant

predation. This is due to the response of the plant to nutrient imbalances. A plant grown

in mineral-balanced soil will initially produce simple metabolic compounds, such as

amino acids and sugars, which are subsequently assimilated into secondary metabolic

compounds that selectively promote (1) vegetative/reproductive growth and (2)

enhanced insect and disease resistance (Greenland, 1988). On some level a trade-off

must be made between these two pathways; however, in mineral-balanced soil these

pathways are interrelated and positively correlated. In environments with an excess of

readily available N, on the other hand, the plant will accumulate a large amount of

simple compounds, effectively unable to metabolise these compounds further due to the

nutrient imbalance. So not only do the absence of secondary metabolites reduce pest

resistance, but the simple compounds are metabolically more accessible to insect

herbivores. These simple compounds act as feeding and egg-laying stimuli for many

herbivorous insects, and therefore it is no surprise that the development and fitness of

these insects is linked to their abundance (Hall, 1988).

In many cases, healthy soils can also promote the suppression of common soil-borne

crop diseases (Grundon, 2009). Two types of suppression, general and specific, work to

inhibit the activity and fitness of disease-causing agents. Many plant pathogens are poor

competitors in the soil and therefore general suppression of these pathogens results from

competition for resources by other non-pathogenic microorganisms. Interspecies

relations such as amensalism (a relationship between two species where species A

negatively influences the fitness of species B without gaining any benefit) and non-

selective predation also help to define general suppression. No single species is

responsible for general suppression, rather the community as a whole acts as an overall

regulator of the individual populations. Therefore, this type of suppression is a result of

a diverse microbial community, and can effectively lead to biostatis, or conditions which

disfavour the inordinate growth of any specific species (Ferris et al., 1996). Under these

conditions it is likely that populations of pathogenic microorganisms will be held at

levels below those necessary for a disease outbreak to occur. Ultimately, a soil system

37

that is nutrient deficient, or lacks the proper mineral balance, will often lack an active

microbial community, creating inefficiencies and imbalances in the community which

pathogens can exploit (Greenland, 1988).

While general suppression stems from the dynamics of the entire microbial community,

specific suppression describes the direct antagonism of a pathogen by a non-pathogen at

some point in the life cycle of either species, usually through predation (Grundon, 2009).

There is much study still to be done on mechanisms of specific suppression, which can

at times be difficult to separate from general suppression.

Soil fertility can also be related to weed abundance. Increasing organic matter content

has been found to be related to decreased weed abundance due to a higher abundance of

bacteria toxic to germinating seeds (Kothandram et al., 1999). In fact, direct microbial

predation has been determined to be a significant fate of weed seeds in the soil. Weed

seed predation by microarthropods and invertebrates such as crickets and beetles is also

extremely significant and is enhanced by increasing ground cover. The distribution of

weeds in a field also has some links with varying soil properties. Weeds are

evolutionarily endowed with the ability to adapt to and survive in a vast array of soil

conditions, which is much of the reason that simple relationships between weeds and

soil properties can be difficult to observe consistently. However, it is generally accepted

that any plant evolves in such a way that optimal conditions produce optimal fitness;

therefore, it seems extremely likely that differing soil conditions may favour certain

species of weeds, at the expense of either the growing crop or other weeds. In fact, the

competitive advantage of a weed is probably less likely to be related to a singular soil

property but instead to the ratio of one nutrient concentration to any number of other

nutrient concentrations, or the interaction of various soil physical properties

(Kothandram et al., 1999).

The benefits of healthy soils to crops are many, and management is the key to ensure

that a soil is functioning correctly (Sanchez et al., 1989). Practices that can help to build

38

healthy soils include crop rotation, organic matter additions, or using high-residue tillage

implements. The inclusion of green manures and cover crops in a rotation is an excellent

way to sponsor fertility, suppress weeds and provide a break in pest cycles.

Incorporating several different species of crops in a rotation, along with manures and/or

compost, ensures a diversity of organic matter sources (Greenland, 1988). This diversity

leads to a more minerally-balanced soil and a pool of nutrients which become available

slowly over time, reducing leaching, waste and toxicity that can result from

immediately-available inorganic fertiliser additions. Ultimately, managing for good soil

fertility is extremely important because the soil and water environment and the

surrounding air environment are in reality virtually inseparable, and the establishment of

a functional and stable system in one environment can have far-reaching impacts in the

other (Doran and Zeiss, 2000).

Most organic matter losses in soil occurred in the first decade or two (Figure 2.6) after

land is cultivated (Tuivavalagi et al., 2002). Native levels of organic matter may not be

possible under agriculture, but many farmers can increase the amount of active organic

matter by reducing tillage and increasing organic inputs.

Tillage Begins

25 years after tillage began

Time

% o

f nat

ive

orga

nic

mat

ter l

evel

s

50%

100%

Figure 2.6 Effect of cultivation on native organic matter levels (Sanchez et al., 1989)

39

2.12 Nitrogen fixing plants to improve soil fertility and health

Nitrogen fixing plants

Nitrogen fixation is a pattern of nutrient cycling which has successfully been used in

perennial agriculture for millennia. Specifically, tree legumes (nitrogen fixing trees,

hereafter called NFTs) are especially valuable in subtropical and tropical agroforestry.

They can be integrated into an agroforestery system to restore nutrient cycling and

fertility self-reliance (Craig and Wilkinson, 1998).

On unvegetated sites, "pioneer" plants (plants which grow and thrive in harsh, low-

fertility conditions) begin the cycling of nutrients by mining and accumulating available

nutrients. As more nutrients enter the biological system and vegetative cover is

established, conditions for other non-pioneering species become favourable. Pioneers

like nitrogen fixing trees tend to benefit other forms of life by boosting fertility and

moderating harsh conditions.

NFTs are often deep rooted, which allows them to gain access to nutrients in subsoil

layers. Their constant leaf drop nourishes soil life, which in turn can support more plant

life. The extensive root system stabilises soil, while constantly growing and atrophying,

adding organic matter to the soil while creating channels for aeration (Figure 2.7). There

are many species of NFTs that can also provide numerous useful products and functions,

including food, wind protection, shade, animal fodder, fuel wood, living fence, and

timber,) in addition to providing nitrogen to the system (MacDicken, 1994).

How to use NFTs in a system?

Nitrogen is often referred to as a primary limiting nutrient in plant growth. Simply put,

when nitrogen is not available plants stop growing. Although lack of nitrogen is often

viewed as a problem, nature has an immense reserve of nitrogen everywhere plants

grow-in the air. Air consists of approximately 80% nitrogen gas (N2), representing about

6400 kg of N above every hectare of land. However, N2 is a stable gas, normally

40

unavailable to plants. Nitrogen fixation, a process by which certain plants "fix" or gather

atmospheric N2 and make it biologically available, is an underlying pattern in nature

(NAS, 1979).

In the tropics, most of the available nutrients (over 75%) are not in the soil but in the

organic matter. In subtropical and tropical forests, nutrients are constantly cycling

through the ecosystem. Aside from enhancing overall fertility by accumulating nitrogen

and other nutrients, NFTs establish readily, grow rapidly, and regrow easily from

pruning. They are perfectly suited to jump-start organic matter production on a site,

creating an abundant source of nutrient-rich mulch for other plants. Many fast-growing

NFTs can be cut back regularly over several years for mulch production (NFTA, 1988-

1994).

The NFTs may be integrated into a system in many different ways including clump

plantings, alley cropping, contour hedgerows, shelter belts, or single distribution

plantings. As part of a productive system, they can serve many functions: microclimate

for shade-loving crops like coffee or citrus (cut back seasonally to encourage fruiting);

trellis for vine crops like vanilla, pepper, and yam; mulch banks for home gardens; and

living fence and fodder sources around animal fields.

41

Figure 2.7 Contributions of a nitrogen fixing tree to the soil ecosystem (MacDicken,

1994).

How biological nitrogen fixation works in legumes?

Nitrogen is essential in plant growth and nitrogen-fixing plants are hugely beneficial in

any ecosystem. Plants can’t absorb nitrogen as gas from the air because they require

fixed (combined) forms of nitrogen known as nitrates. Some plants, known as ‘nitrogen

fixers’ interact with certain type of microbes in order to transform nitrogen gas in

nitrates. The huge botanical family Fabacae (previously known as Leguminosae) forms

a relationship with rhizobium and bradyrhizobium bacteria. These are the plants we

call legumes and they are commonly used in organic agriculture and also in main-stream

agriculture in the form of ‘green manures’. Leguminous plants range from small to large

weeds, to crops, to shrubs, to trees.

Working with a group of bacteria called rhizobia, legumes are able to pull nitrogen out

of the air and accumulate it biologically. The bacteria, which are normally free-living in

the soil in the native range of a particular legume, infect (inoculate) the root hairs of the

42

plant and are housed in small root structures called nodules. Energy is provided by the

plant to feed the bacteria and fuel the nitrogen fixation process. In return, the plant

receives nitrogen for growth (FAO, 1984).

A number of complex biochemical reactions defining the mechanism for biological

nitrogen fixation can be summarised by the following chemical equation, in which two

moles of ammonia are produced from one mole of nitrogen gas, at the expense of 16

moles of ATP and a supply of electrons and protons (hydrogen ions):

N2 + 8H+ + 8e- + 16 ATP 2NH3 + H2 + 16 ADP + 16 Pi

There are thousands of strains of rhizobia. Certain of these will infect many hosts;

certain hosts will accept many different strains of rhizobia. Certain hosts may be

nodulated by several strains of rhizobia, but growth may be enhanced only by particular

strains. Therefore, when introducing hosts to a new area it is extremely important to also

introduce a known effective symbiotic rhizobia strain. Such effective strains have been

identified for thousands of the important nitrogen fixing legumes, and can be purchased

at low cost for the value returned. The best method for ensuring effective nitrogen

fixation is to introduce a known effective strain of Rhizobium to the potting medium at

the time of sowing. Large, healthy nodules may also be used to inoculate seeds. To

determine if the nodule is effective, it may be cut open. Effective nodules will have a

pink to dark red pigment inside (FAO, 1984).

In conventional cropping systems it is estimated that 50-800 kg of nitrogen per hectare

per year are accumulated by nitrogen fixing plants, depending on species, soil and

climate, Rhizobium efficiency, and management. Equivalent quantities of manufactured

nitrogen are produced using an energy intensive process, and the end product is high-

priced nitrogen in a form which can be detrimental to soil ecology (Verma, et al., 1993).

43

2.12.1 Mucuna pruriens

Mucuna pruriens is a tropical legume known as velvet bean or cow itch and by other

common names. The plant is infamous for its extreme itchiness produced on contact,

particularly with the young foliage and the seed pods. It has a high value in agricultural

and horticultural use and has a range of medicinal properties (Agharkar, 1991).

The plant is an annual, twinning, climbing shrub with long vines that can reach over 15

m in length. When the plant is young, it is almost completely covered with fuzzy hairs,

but when older, it is almost completely free of hairs. The leaves are trifoliate (tripinnate),

ovate, reverse ovate, rhombus shaped or widely ovate. The sides of the leaves are often

heavily grooved and the tips are pointy, gray-silky beneath; petioles are long and silky,

6.3–11.3 cm. Leaflets are membranous, terminal leaflets are smaller, lateral very

unequal sided. Dark purple flowers (6 to 30) occur in drooping racemes. Fruits are

curved, 4–6 seeded. The longitudinally ribbed pod is densely covered with persistent

pale-brown or grey trichomes that cause irritating blisters. Seeds are black ovoid and 12

mm long (Sastry and Kavathekar 1990; Agharkar 1991; Verma et al. 1993).

The vines grow very fast by branching from each node. The leaves are thermonastic –

when temperature rises or falls, the leaves close up. The thickness of green vegetation of

mucuna on the ground ranges up to 1 m. Such luxuriant growth of cover plants is of

much value in suppressing noxious weeds and reducing soil temperature. The probable

presence of allelo-chemicals in the tips of the young vein inhibits the growth of engulfed

weeds. The mulch of dried leaves increases the microbial activity and enriches the

nutrient status of the soil. Roots developed from the nodes of vines touching the ground

are fibrous. Nodules formed on such roots are small and round. The very presence of

nodules indicates the penetration of rhizobium (Dutta, 1970). The main root grows to a

depth of 2-3 m. As found by Wycherley (1963) deep rooted plants may increase the

fertility of the surface soil by extracting nutrients from the deeper layer of soil and

depositing them on the surface in the form of organic matter.

44

Mucuna litter was observed to decompose slowly. The slower decomposition rate and

gradual nutrient mineralisation of cover plant litter are valuable attributes, as the mulch

would have a more lasting effect and nutrients would be available over a longer period

for uptake (Tan et al., 1996).

The plant exhibits strong tolerance to drought and shade. Only older leaves shed during

the dry months and new shoots sprout in about 30 days. The plant has no serious pest or

disease problems as high level of phenolic compounds deter the insects (Kothandram et

al., 1989).

Mucuna as an improved fallow legume

The first use of mucuna as a green manure was reported in Bali, Java and Sumatra in the

17th century (Burkill, 1966). Buckles et al. (1998) reported additions of about 155-200

kg/ha of N from a sole crop of mucuna beans. The same study confirmed that mucuna

beans accumulated large quantities of calcium (140 kg/ha on average, 70% of it in the

litter), potassium (100 kg/ha, 82% in the live sub-fractions) and phosphorus (15-20

kg/ha, 45% in the litter). In the south-eastern United States mucuna has been reported to

have a positive effect on the suppression of plant-parasitic nematodes and other soil

borne pathogens (Hartemink, 2003). Boateng (2005) reported on improvements in soil’s

physical, chemical and biological properties under the mucuna fallow systems.

In a study in Ghana, Fosu et al. (2004) reported the dry matter yields for mucuna to be in

the range of 5-15 t/ha depending on the amount of rainfall and fallow duration.

2.12.2 Erythrina

The genus Erythrina is an attractive alternative for agro-forestry systems. Studies have

shown promising results when alley cropped with annual crops, shade, live fencing, and

for animal feed.

45

Erythrina is native to the tropics and subtropics from Bolivia to Panama, and naturalised

to Central America, Africa, Malaysia, and the Pacific. Erythrina grows rapidly, produces

abundant biomass, is easily propagated from cuttings, coppices well, and fixes nitrogen.

The wood, however, is generally of poor quality. Traditionally, in the Pacific, farmers

know Erythrina for its potential to improve soil quality.

As a source of green manure, Erythrina yielded almost 12 tons/ha/yr when cut back to

the crown twice yearly. Biomass Production in cocoa plantations, pruned biomass

averaged over 9 t/ha (Ramirez, et al., 1990).

As a protein supplement in animal fodder, Erythrina was more palatable to ruminants

than G. sepium. When compared with concentrates for increasing weight gain and

productivity, Erythrina was not as effective, but is a more affordable alternative.

However, the presence of alkaloids in the seeds may limit potential use as animal feed.

The presence of such alkaloids in the foliage of selected clones is being evaluated

(Ramirez, et al., 1990).

Erythrina has also being evaluated as mulch for cassava, maize and beans. Applied at a

rate of 40 tons/ha (8 t/ha DM) a mulch of E. poeppiginia was successful in maintaining

maize and bean yields at 3 t/ha and 1 t/ha for more than eight years, respectively,

without any application of mineral fertiliser (Russo, 1990).

Alley cropping trials were conducted to ascertain the effect of E. peoppiginia on

associated maize and beans. Trees were planted at 6 x 3 m spacing and pruned twice

yearly. Bean yields increased significantly with the application of prunings; maize yields

began showing increases only in the 7th and 8th years of the trial. Soil tests showed

increases in K but decreases of available P in the alley-cropping system: accumulation of

P in the tree stump may be significant where P is limiting (Ramirez, et al., 1990).

46

There is little quantification of the N fixed by Erythrina; long-term studies are needed

for quantification purposes. In alley cropping systems, a response to mineral N was

frequently observed with beans but not with maize.

Erythrina is widely used as a shade tree for coffee and cacao because of its rapid growth

rate. A 2.5 m stake of E. poeppiginia planted can grow up to 4.5 m in 6 months, with a

diameter of 8-10cm (Russo, 1990).

As a coffee shade, E. poeppiginia can produce the same yield increase as 132 kg/ha of

mineral N. In Central America, Erythrina is primarily used for live fencing. Biomass

production, but not fodder quality, was found to decrease when trees were pruned less

than every six months; however it is unlikely that a fence can be a significant source of

animal feed (Russo, 1990).

2.13 The use of biochar in agriculture

Biochar is the carbon-rich product obtained by heating biomass in a closed system under

limited supply of oxygen. Currently, there are several thermochemical technologies such

as pyrolysis, gasification, and hydrothermal conversion to produce biochar.

Biochar can be used directly as a replacement for pulverised coal as a fuel. But one of

the major distinctions between biochar and charcoal (or char) is that the former is

produced with the intent to be added to a soil as a means of sequestering carbon and

enhancing soil quality. When used as a soil amendment, biochar has been reported to

boost soil fertility and improve soil quality by raising soil pH, increasing moisture

holding capacity, attracting more beneficial fungi and microbes, improving cation

exchange capacity (CEC), and retaining nutrients in soil (Lehmann et al., 2006;

Lehmann, 2007). Another major benefit associated with the use of biochar as a soil

amendment is its ability to sequester carbon from the atmosphere-biosphere pool and

transfer it to soil (Winsley, 2007; Guant and Lehmann, 2008; Laird, 2008). Biochar may

persist in soil for millennia because it is very resistant to microbial decomposition and

47

mineralisation. This particular characteristic of biochar depends strongly on its

properties, which is affected in turn by the pyrolysis conditions and the type of feedstock

used in its production. Previous studies indicate that a bioenergy strategy that includes

the use of biochar in soil not only leads to a net sequestration of CO2 (Woolf et al.,

2010), but also may decrease emissions of other more potent greenhouse gases such as

N2O and CH4 (Spokas et al., 2009).

Similar to activated carbon, biochar can serve as a sorbent in some respects. Biochar

usually has greater sorption ability than natural soil organic matter due to its greater

surface area, negative surface charge, and charge density (Liang et al., 2006). Biochar

can not only efficiently remove many cationic chemicals including a variety of metal

ions, but also sorb anionic nutrients such as phosphate ions, though the removal

mechanism for this process is not fully understood (Lehmann, 2007). Thus, the addition

of biochar to soil offers a potential environmental benefit by preventing the loss of

nutrients and thereby protecting water resources. Furthermore, soils containing biochar

have a strong affinity for organic contaminants (Yang and Sheng, 2003a; 2003b; Yu et

al., 2009). For example, one study revealed that unmodified biochar pyrolysed from

waste biomass could effectively sorb two triazine pesticides, effectively retarding their

transport through the soil (Zheng et al., 2010). Additionally, some modified biochars

(i.e., biochar modified by some specific physical and chemical activation treatments)

have demonstrated the potential to effectively remove a variety of organic contaminants

from water as a sorbent (Chen et al., 2008; Cao et al., 2009). The use of biochar as a

cost-effective sorbent is an emerging research topic.

Sustainable biochar is produced from sustainably procured waste biomass such as crop

residues, manures, timber and forestry residues, and green waste using modern pyrolysis

technology (Woolf et al., 2010). Therefore, sustainable biochar production and its use as

a soil amendment have been suggested as a means of abating climate change by

sequestering carbon, while simultaneously reducing waste, improving soil quality, and

protecting natural resource (Winsley, 2007; Laird, 2008; Guant and Lehmann, 2008;

Zheng et al., 2010).

48

2.13.1 Biochar for Sustainable Agriculture

Sustainable agriculture is a way of raising food that is healthy for consumers and

animals without causing damage to ecosystem health. Low nutrient content and

accelerated mineralisation of soil organic matter (SOM) are the two major constraints

currently encountered in sustainable agriculture (Renner, 2007). Nutrients are retained in

soil and remain available to crops mainly by adsorption to minerals and soil organic

matter. Usually, the addition of organic matter such as compost and manure into soil can

help retain nutrients. Biochar is considered much more effective than other organic

matter in retaining and making nutrients available to plants. Its surface area and complex

pore structure are hospitable to bacteria and fungi that plants need to absorb nutrients

from the soil. Moreover, biochar is a more stable nutrient source than compost and

manure (Chan et al., 2007).

The modality of biochar in its ability to act as an effective soil amendment is similar to

the traditional “slash-and-burn” fertilisation method, where farmers remove the

vegetation and release a pulse of nutrients to fertilise the soil. But the “slash-and-burn”

practice has an unfavourable environmental reputation because it is associated with

deforestation and air pollution. In contrast, biochar production under a controlled system

may provide a higher yield and have fewer detrimental effects on the environment.

These characteristics make biochar an exceptional soil amendment for use in sustainable

agriculture (Lehmann and Joseph, 2008; Verhejien et al., 2010).

Several greenhouse and field studies have been conducted to examine the effect of

biochar on crop yields (Glaser et al., 2002; Yamato et al., 2006; Chan et al., 2007 and

2008). Most studies showed that biochar addition increased crop yields. For example, a

plot trial where soil was amended with a greenwaste-derived biochar, showed benefits

that included increased crop yield and improved soil quality (Chan et al., 2007). Field

experiments have also reported substantial crop yield increase in response to soil biochar

application (Glaser et al., 2002; Yamato, et al., 2006). Most of these experiments,

however, were conducted in the tropics using biochar produced in local earthen kilns

49

and applied to soils with low organic matter content (Laird, 2008). In a few cases, either

no difference or negative results have been found on use of biochar as a soil amendment

(Laird, 2008). The varying effects on crop yield appear to depend on such factors as

biochar quality, biochar quantities added, soil type, and crop tested.

2.14 Microbial and biochemical indicators of soil health

Soils perform a wide range of ecosystem functions that are crucial for the majority of

terrestrial life. Soils provide microbial habitat spaces of diverse size and architecture, as

well as reservoirs of chemicals such as nutrients, and ecosystem services such as water

filtration and storage. In addition, soils support biological activities such as

decomposition and recycling of dead organic matter, and play a major role in mitigating

climate change through the sequestration of carbon.

Biological and biochemical indicators are proposed as sensitive parameters to slightest

modifications that the soil can undergo under the action of any applied agent (Klein et

al. 1985). Since soil microorganisms, due to their quick metabolism, can respond to

stress/disturbance factors more rapidly, they should preferentially be considered when

monitoring soil status.

Reliable soil microbiological and biochemical indicators to determine soil health should

be simple to measure, work equally well in all environments and reveal which problems

exist wherever. It is unlikely that a sole indicator can be defined with a single

measurement because of the multitude of microbiological components and biochemical

pathways. Microbial indicators of soil health cover a diverse set of microbial

measurements due to the multifunctional properties of microbial communities in the soil

system (Bloem et al. 2006).

50

2.14.1 Enzyme activity tests as soil quality indicators

Soil enzymes serve as catalysts for all the biochemical processes in soils. They are found

in living organisms as well as free in the environment. Most enzymes released into the

environment, either as truly extracellular enzymes or as a result of cell death, are rapidly

degraded. However, some are stabilised on soil colloids (clay and organic matter) where

they are protected from degradation and can retain their activity (Skujins, 1976;

Nannipieri et al., 1996). This suggests that some soil enzyme activities may be useful as

integrative measures of soil quality. Soil enzyme activities can be interpreted from the

perspective of total and/or specific activities. Total enzyme activity provides an estimate

of the rate at which the product of activity is made available in the soil (Landi et al.

2000). Specific enzyme activities provide insight of how suitable the organic matter is to

degradation by each specific enzyme, thus providing a measure of organic matter quality

(Sinsabaugh et al. 2008).

Soil enzymes are involved in many vital functions including decomposition of organic

inputs, cycling of vital plant nutrients (e.g. N, P, K, and S) as well as the global C cycle,

and detoxification of xenobiotics. Soil quality has been defined as "the ability of the soil

to perform functions that are required for the biological components of an ecosystem

within the constraints of local environmental factors" (Dick, 1997). Consequently, the

integrative nature of soil enzyme assays makes them attractive as indicators of biological

function. In addition, soil colloids have the ability to stabilise enzymes outside of living

cells, allowing them to maintain their catalytic activity (Skujins, 1976; Hope and Bums,

1987; Nannipieri et al., 1996). Thus, the enzyme activities measured in the lab are the

result of the combined activities of enzymes associated with living cells and those

stabilised in the soil matrix. Skujins (1976) proposed the name "abiontic" for these

enzymes of biotic origin no longer associated with living cells. For enzymes with a large

abiotic component, the soil itself can be viewed as an organism, capable of performing

biochemical transformations.

51

Soil enzyme activity has been shown to be sensitive to a variety of ecosystem stresses.

Kuperman and Carreiro (1997) studied a grassland ecosystem contaminated with a wide

range of heavy metal concentrations. Activities of a wide variety of enzymes (N-

acetylglucosaminidase, -glucosidase, endocellulase, and acid and alkaline phosphatases)

were shown to decrease with increasing soil heavy metal concentrations. Madejon et al.

(2001) performed a similar experiment, but amended soils with heavy metal-

contaminated organic materials and incubated them in the lab for 40 weeks. They

observed a flush of microbial activity and a concomitant increase in enzyme activity

early in the incubation, but it was short-lived and enzyme activities leveled off near or

below beginning values. They attribute the flush of activity to the addition of organic

matter. The contrast between this work and that of Kuperman and Carreiro (1997)

suggests that the reduction of enzyme activity by heavy metals requires a longer period

of time than the 40 week incubation period to become evident. Soil enzymes have also

been shown to be sensitive to cultivation. Gupta and Germida (1988) studied the effect

of 69 years of tillage on a Saskatchewan soil. A comparison was made between a field

cultivated for 69 years and an adjacent native prairie. Both arylsulfatase and acid

phosphatase activities were found to be significantly lower in the cultivated field than in

the native prairie. Farrell et al. (1994) reported similar results in three different systems

where they compared native soil to a variety of lengths and intensities of cultivation.

They found that arylsulfatase activity was sensitive to tillage treatment in all cases. In

addition, they reported that increasing the length of time and intensity of cultivation

further reduced activity.

Many people have reported sensitivity of soil enzyme activities to the use of organic soil

amendments and green manures. Verstraete and Voets (1976) compared the effect of

four different organic fertiliser regimes with that of a control field. They found that

phosphatase, β-glucosidase, saccharase, and urease activities were sensitive to these

management techniques. Mendes et al. (1999) found that both 3-glucosidase activity and

FDA hydrolysis were able to differentiate between a winter fallow treatment and two

different cover crop treatments. Dick et al. (1988b) studied the effect of a variety of

52

residue management activities on soil enzyme properties. Straw was incorporated into

the soil in all treatments. Treatments included fall or spring burn, 0, 45, or 90 kg N/ha,

pea vine (2.2 Mg/ha/y), or manure (22.2 Mg/ha/y). They found that acid and alkaline

phosphatase, arylsulfatase, β-glucosidase, urease, and amidase activities were sensitive

to residue management. Bandick and Dick (1999) studied the effect of cover crops and

manure amendments on soil enzyme activities at two sites in Oregon. They found that a-

and 3-glucosidase, a- and 3-galactosidase, amidase, arylsulfatase, deaminase, FDA,

invertase, cellulase, and urease were all sensitive to these treatments at both sites.

2.14.2 Soil Microbial Activity

Soil microbial activities are of critical importance for biogeochemical cycles. Microbial

activity is regulated by many factors including oxygen and water availability,

temperature and soil pH. Soil microbial activity can be measured under either field or

laboratory conditions. In the field, variations in meteorological conditions during an

experiment are inevitable, i.e. soil aeration, moisture and temperature will change and

may strongly influence the results (Madsen, 1996). Furthermore, field measurements are

difficult to interpret. For example, soil respiration determined in the field comprises of

activity of microorganisms and other organisms such as macro fauna and plants, which

vary significantly in different systems and throughout the season (Dilly et al. 2000).

Laboratory procedures are usually carried out on sieved and stabilised soil samples at

standardised temperature and moisture content. Such measurements generally include

assays of enzyme activities, C and N mineralisation. These, and eventually other

microbial activity measurements, may be helpful to evaluate the effects of soil

management, landuse and specific environmental conditions on microbial activity

(Burns, 1978). Laboratory methods allow the standardisation of environmental factors,

and thus, the comparison of results from various spatial and temporal measurements.

53

2.14.2.1 Fluorescein Diacetate Hydrolysis (FDA) Activity

More than 90% of the energy in a soil system passes through microbial decomposers.

Consequently total microbial activity of microbial decomposers is a good general

measure of organic matter turnover and a good indicator of soil health.

Fluorescein diacetate is composed of a fluorescein conjugated to two acetate radicals.

FDA provides a broad-spectrum indicator of total microbial activity in soil. Fluorescein

diacetate (3', 6' diacetylfluorescein) can be hydrolysed by many free and membrane

bound enzymes such as proteases, lipases, and esterases, which turns the colourless

compound yellow. The esters of fluorescein diacetate are non-polar and are easily

transported through the membranes of active cells in comparison to fluorescein

molecules which are polar and remain inside the cells. The activity necessary to

hydrolyse it has been found among many groups of soil bacteria, protozoa, algae and

fungi (Dick et al., 1996). Because of its ubiquitous nature, FDA hydrolysis has been

used as a broad spectrum measure of microbial activity. It has been shown to be highly

correlated with some of the most sensitive measures of microbial activity such as ATP

content and cell density studies (Stubberfield and Shaw, 1990) as well as radio labeled

thymidine incorporation into microbial DNA (Federle et al., 1990).

FDA hydrolysis also has been shown to be sensitive to management. Mendes et al.

(1999) showed that FDA hydrolysis can be used to differentiate between a winter fallow

treatment and a cereal or legume cover crop. Bandick and Dick (1999) also showed that

FDA is sensitive to cover cropping and N fertilisation.

2.14.2.2 Soil Urease Activity

The requirement of nitrogenous fertilisers for agricultural production is well known.

Among the commercially available nitrogen fertilisers, urea is the most widely used

source of N (Gautney et.al., 1986). The use of urea is steadily increasing worldwide and

this trend is likely to continue (Sahrawat, 1980).

54

In most arable soils, urea is decomposed enzymatically to CO2 and NH3. When applied

to soil, urea is hydrolysed by the enzyme urease to NH4+. Urease (urea amidohydrolase)

hydrolyses non peptide C-N bonds in linear amides, leading to plant available NH4+-N.

Depending on soil pH, the NH4+ may form NH3, which can be volatilised at the soil

surface, as represented in the following reactions:

CO(NH2)2 + H+ + 2H2O 2NH4+ + HCO3

-

NH4+ NH3 + H+

Urease hydrolysis, as in any enzymatic reaction, may only be needed to reduce

activation energy for the formation of intermediate products. Even in the absence of

enzymes, urea can be hydrolysed physio-chemically. However, chemical hydrolysis is

very slow compared to biochemical enzymatic hydrolysis (Chin and Kroontje, 1963).

Therefore, it can be concluded that urea hydrolysis in soils is mainly brought about by

the action of the enzyme urease.

Origin of soil urease

Urease activity in soil may originate from plant residues, animal waste or soil microbes

containing urease (Lai and Tabatabai, 1992). Plants are rich sources of ureases

(Frankenberger and Tabatabai, 1982). However, there is no direct evidence for the

production of urease by plant roots (Esterman et al., 1961). Urease has been reported to

be present in animal intestines and excreta. Therefore, the addition of plant materials and

animal wastes may supply urease to the soil. Skunjins (1976) reported that soil urease is

of microbial origin. Sumner (1951) also identified some species of bacteria, yeast and

fungi which contained urease. Most of the Nitrosomonas and Nitrosospira isolated from

soils in Scotland were capable of hydrolysing urea (Allison and Prosser, 1991).

55

Factors affecting urease activity in soil

Temperature – numerous studies have shown that urease activity in soils increased with

increase in temperature from 10o to 40oC (Lai and Tabatabai, 1992). In some soils,

urease activity increased very markedly with the increase in temperature from 40o to

70oC; but decreased rapidly above this range. Urease activity in Indian Alfisols and

Vertisols at 100oC was close to zero. When soils were heated at 105oC for 24 hours,

urease activity was inactivated completely (Sahrawat, 1984). Zantua and Bremner

(1977) suggested that soil urease is protected from inactivation at higher temperatures

and immobilisation of enzymes enhances their thermal stability. Bremner and Zantua

(1975) found that urease activity could be detected in soils at -10o or -20oC, but not in

soils at -30oC.

Soil pH – generally, free urease is most active at neutral pH and soil urease is most

active at slightly alkaline pH levels. But different pH optima values, ranging between

5.87 and 9.0 have been reported. This divergence may be related to the differences in the

buffers and urea concentrations adopted in these investigations, in addition to the

variability in soil types. However, there are some reports which indicate that urease

activity is unrelated to pH (Zantua et al., 1977) or negatively related to pH (Dash et al.,

1981).

Moisture content – urea hydrolysis increases with increasing soil water up to near field

capacity, followed by a decreasing trend thereafter (Velk and Carter, 1983; Savant et al.,

1985; Antil et al., 1993). Sahrawat (1984) observed a constant urease activity when the

moisture content was increased further beyond field capacity. Savant et al. (1985)

observed a higher rate of urea hydrolysis in soils at field capacity than in water logged

soils after 24 hours of incubation. Urea hydrolysis rates decrease below the permanent

wilting point (Velk and Carter, 1983) and in dry soils (Sahrawat, 1984).

56

Urea concentration – studies have shown that the rate of hydrolysis of urea by soil

urease increases with an increase in substrate (urea) concentration until the quantity of

urea added is saturating and its activity becomes constant (Bremner and Mulvaney,

1978; Tabatabai and Bremner, 1972). Zantua et al., (1977) reported urease activity to be

positively correlated to the total N in the soil.

Oxygen - Bremner and Mulvaney, (1978), concluded that since the urea added to soil is

hydrolysed largely, if not entirely, by native soil urease, there is no apparent reason why

the activity of urease should be affected by O2. Savant et al. (1985) found that O2

becomes a limiting factor after 12 hours of submergence. Velk and Carter (1983)

observed substantial reduction of urease activity under flooded conditions of some soils.

Delaune and Patrick (1970) observed no difference in the rate of urea hydrolysis in soils

under waterlogged and 0.33 bar moisture conditions.

Organic matter – many workers have found that urease activity in soils is positively

correlated with organic C and total N (Zantua et al., 1977; Dash et al., 1981; Reynolds et

al., 1985), which are indices of organic matter content. Zantua et al. (1977) suggested

that organic matter content of a soil accounted for most of the variations in urease

activity. Several workers have observed an increase in soil enzyme activities after

incorporation of organic matter in the soil (Balasubramanium et al., 1972; Zantua and

Bremner, 1976). The increased level of urease activity in the organic amended soil has

generally been attributed to the increased microbial biomass although additional

evidence has shown that plant materials and sludges may directly contribute enzymes to

the soil. Microorganisms associated with the organic materials may also contribute to the

urease in the soil enzyme pool. The urease activity in soil varies depending upon the

type and amount of organic matter added. On addition of decomposed organic matter

and farmyard manure, urease activity increases significantly. However, incorporation of

undecomposed dried grass had no effect on urease activity (Kumar and Wagenet, 1984).

57

2.14.2.3 Soil Phosphatase Activity

The element phosphorus is essential for plant growth and metabolism. It is generally

accepted that plant roots take up phosphorus as soluble inorganic phosphate (Bieleski

and Ferguson, 1983). Since a large proportion of the P in soil is organically bound (a

very important amount of P in soils especially of arid climates is bound inorganically), the

mineralisation of this organic fraction is of major agricultural importance (Speir and

Ross, 1978). Several enzymes are involved in the decomposition of organic phosphorus

compounds. Those enzymes that catalyse the hydrolysis of both P and anhydrides of

H3PO4 esters are commonly called phosphatases (Alexander, 1977). Phosphatases

(orthophosphoric monoester phosphohydrolases) are important in soils because phosphatase

catalyses the hydrolysis of organic esters and anhydrides of H3PO4 to orthophosphate;

thus, they form an important link between plant-unavailable and soluble P fractions in

soil (Amador et al., 1997).

Phosphatase activities in soil can be associated with active cells (animal, plant,

microbial), entire dead cells and cell debris as well as being complexed with clay

minerals and humic colloids (Pascual et al., 2002). In addition, the sorption of

phosphatases on clay, oxides or humic substances can change enzyme conformation and

reduce activity (Dick and Tabatabai, 1987; Nannipieri et al., 1988). Phosphatases are

distinguishable not only by the chemical nature of the substrates hydrolysed but also by

pH ranges for their optimal activity. Among them are: acid phosphatase, optimal pH 4-6;

neutral phosphatase, optimal pH 7; and alkaline phosphatase, optimum pH 8-10 (Speir

and Ross, 1978).

Phosphatase activity is affected by soil physico-chemical (clay content, soil moisture,

soil depth, temperature, organic matter, pH and nutrients) and biological (microbial

population and their activities) properties (Speir and Ross, 1978) and these properties

play a key role among them. As far as physico-chemical soil properties are concerned,

numerous studies have focused on the carbon content and its positive impact

phosphatase activity (e.g., Herbien and Neal, 1990; Pagliai and De-Nobili, 1993;

58

Marinari et al., 2000), relationships between organic matter and the other elements in the

organic bounds (e.g., N and P) and pH. Relationships between phosphatase activities and

total P, organic P, available P have been described by Gavrilova et al. (1973), Speir and

Ross (1978), Haynes and Swift (1988) and Nahas et al. (1994). On the contrary, little

information on the relationships between phosphatase activities and inorganic P

fractions in soils is available. In moderately well drained soils with a relatively low level

of inorganic P, Amador et al. (1997) have observed a positive correlation between

inorganic P and phosphatase activity. The high concentration of inorganic P in soils has

been shown to reduce phosphatase activity. For example, orthophosphate inhibited

phosphatase activity in soils (Juma and Tabatabai, 1978), as well as the synthesis and

catalytic action of phosphatases of microorganisms in soil (Woolhouse, 1969). Chen

(2003) reported positive correlations between phosphatase activities (acid and neutral)

and inorganic P fractions (iron and aluminium phosphates) in acidic Chinese forest soils.

There is currently great interest in the use of extracellular enzymes as biological

indicators of soil health, because they are relatively simple to determine, have microbial

ecological significance, are sensitive to environmental stress and respond rapidly to

changes in land management (Dick, 1997; Yakovchenko et al., 1996). Phosphatase

activity may be a particularly useful enzyme for soil health monitoring because of its

central role in soil organic matter cycling, which is generally regarded as an important

component of soil quality. Research has shown that phosphatase is the most abundant

and easily detected of the enzymes involved in organic P compounds decomposition in

soil and is rarely substrate limited, thus making it ideal to examine the importance of soil

P status. Indeed, it provides an early indication of changes in organic P and organic

matter status and turnover (Gavrilova et al., 1973; Speir and Ross, 1978).

2.14.3 Nitrogen mineralisation

Almost all soil nitrogen (N) is present in the form of organic compounds that cannot be

used directly by plants and is not susceptible to loss through leaching. The soil’s

capacity to transform organic N into inorganic N, i.e. its N mineralisation potential, is

59

often used as an index of the N available to plants (Robertson et al. 1999; Nannipieri and

Paul, 2009). It is perhaps the most common and best means available to assess N fertility

(Keeney, 1980; Binkley and Hart, 1989) as it is related to both the size of the labile soil

organic N pool and the activity of organisms responsible for the mineralisation

processes. Mineralisation potentials, i.e. the net production of inorganic N released from

the mineralisable organic fraction in soil under constant moisture and temperature

conditions, are better than inorganic soil N concentrations (pool size) as an indicator of

site fertility, because the supply rate of a limiting nutrient affects more its availability

than its instantaneous concentration. Most mineralisation assays are designed to exclude

plant uptake and leaching but include microbial immobilisation and denitrification, thus

providing net mineralisation potentials.

Nitrogen mineralisation assays usually refer to the net increase in both ammonium

(NH4+) and nitrate (NO3

-) in soil, since any nitrate formed must be derived from

ammonium. The term “net” refers to the difference between the gross N mineralisation

and gross N immobilisation (Davidson et al., 1991). While other forms of inorganic N

are also produced during mineralisation assays (e.g. NO2-, N2O and NOx), in most soils

their appearance is highly transient and relative pools are quickly converted to another

form or their fluxes are in consequential relative to increases in the NH4+ and NO3

- pools

(Robertson and Tiedje, 1985). While N mineralisation assays have their limitations as

measures of N availability, they can nevertheless provide substantial insight into soil

fertility and ecosystem functioning at many sites, and they are widely used as indicators

of soil health and quality (Sparling, 1997).

Large differences between sites or experimental treatments, for example, imply large

differences in plant-available N, as well as large differences in the potential loss of N

from the ecosystem. Nitrate, for example, is more readily lost from most ecosystems

than ammonium, so large potential nitrification rates at a site can indicate a higher

likelihood of nitrogen loss, all else being equal.

60

Greater mineralisation potentials have been found in a variety of cropping systems under

organic than conventional management (Doran et al. 1987; Drinkwater et al. 1995;

Gunpala and Scow 1998) reflecting the increased role of decomposers in determining N

availability. The ratio of N mineralised to total organic N can be a sensitive indicator of

differences in soil organic matter. The percent of total soil N mineralised in short term

anaerobic incubations was more than two fold greater in soils with organic compared to

conventional management, indicating qualitative differences in soil organic matter

(Drinkwater et al. 1995). N mineralisation and total N pools can serve as indicators on

the status of N dynamics in the soil. High mineralisation potentials in conjunction with

high concentration of mineral N, especially during times of reduced crop uptake, could

indicate susceptibility of N losses through leaching. Organically managed soils are

sometimes characterised by high microbial activity and potentially mineralisable N,

together with small concentration of mineral N when compared with soils receiving

conventional mineral fertilisers (Drinkwater et al. 1995). The combination of low

mineral N concentration and enhanced microbial activity in the organic soils are

indicative of a more tightly coupled N cycle (Sprent, 1987; Jackson et al., 1989;

Jenkinson and Parry 1989), with higher turnover rates of mineral N pools than in

conventional soils. 2.14.4 Potentially mineralisable nitrogen (PMN)

PMN can be defined as the fraction of organic nitrogen converted to plant available (or

mineral) forms under specific conditions of temperature, moisture, aeration and time.

Determining levels of PMN can provide an estimate of available N in the soil

(Drinkwater et al. 1996).

PMN originates mainly from microbial biomass and plant and animal tissues – the main

source of the organic nitrogen pool. It represents the fraction of nitrogen easily

decomposable by soil microorganisms and is considered to be an indirect measure of

nitrogen availability during the period of measurement (Doran, 1987).

61

Potentially mineralisable N (PMN) as determined by anaerobic incubation (Waring and

Bremner, 1964; Keeney and Bremner, 1966) is often considered to reflect the organic

matter pools being mineralised (Stockdale and Rees, 1994). It is a simple biochemical

assay that is affected less by leaching events (Cookson and Murphy 2004) and climatic

fluctuations (Murphy et al. 1998) than actual gross N mineralisation rates. Previous

studies have thus used PMN as an index to assess within-site variation of soil N supply

on a scale applicable to farm management practices (Baxter et al. 2003). In that

particular study, Baxter et al. (2003) illustrated that for a 6-ha paddock cropped to

winter barley, PMN exhibited sufficient spatial structure to create a map with the

intended purpose of zoning for a variable N management strategy. Such an approach

implies that PMN can be interpreted with direct relevance to the supply of crop N

demand. While this was not reported by Baxter et al. (2003), other studies have found

similar indices of soil N supply to be significantly related to crop N uptake (e.g.

McTaggart and Smith 1993), although the percentage of yield variability explained is

often low (<40%) (Walley et al., 2002).

2.14.4.1 Factors affecting PMN

Inherent – levels of PMN may be greatest in humid climates and lower in drier climates

because humid climates usually enhance biomass production. Clay soils have the

capacity to physically protect organic matter and organic nitrogen and, thus, associated

PMN from degradation by microorganisms. During the soil incubation to measure PMN,

soil clay particles can attract and temporarily retain ammonium on cation exchange

complexes. The depth to bedrock affects soil hydrologic properties (e.g. fluctuation of

water table and subsequent soil hydromorphy, causing excess or low moisture amounts),

which in turn determine the chemical end products of N mineralisation, namely

ammonium or nitrate. Low areas (topographic depressions) of a field tend to accumulate

more organic matter and total N, and probably available N and PMN, than elevated sites

(Stanford and Smith, 1972).

Dynamic – soil properties and soil management practices that affect organic matter and

organic N dynamics will ultimately affect available N and PMN levels. Continuous

62

cultivation of lands without the replenishment of external organic matter depletes the

land’s organic matter and nitrogen reserve and the related available PMN. Conversely,

repetitive additions of farm manure or crop residues (as under no-till or cover crops

usually increase the levels of available N and probably PMN), no-till significantly

increases PMN levels compared to plough and fertiliser additions. Soils with stable

aggregates protect soil organic matter and associated available N from microbial

degradation compared to soils with unstable aggregates. Small aggregates in soil

reportedly contains a larger proportion of readily mineralisable organic N, therefore a

greater amount of PMN may be obtained in soils with small aggregates than in those

with larger aggregates. Accumulation and mineralisation of N also depend upon the C:N

ratio of the amendment material added to the soil (Doran, 1987).

2.14.4.2 Relationship of PMN to soil functions

As a readily available fraction of total N, PMN is an important source of N for crop

growth and yield, especially in synthetic N fertiliser free agricultural operations (e.g.

organic farming). PMN can be source of available N for microorganisms and indirectly

enhance microbial growth and activities, including C and N cycling. In well drained

soils, PMN is made available to the plants and microorganisms, mostly in the form of

nitrate, through aerobic mineralisation. In poorly drained soils (such as rice fields), PMN

is made available, in the form of ammonium, through anaerobic mineralisation (Islam et

al., 1998).

2.14.4.3 PMN problems associated with poor activity

Soils naturally low in organic matter or depleted by poor management will have low

PMN content. In the absence of live vegetation, a high amount of available N delivered

by the PMN pool can build up and become a potential source of nitrate contamination

for ground water. An excess of nitrate from the PMN pool can be lost to the atmosphere

as gaseous nitrogen products during subsequent very wet seasons or under heavy

irrigation (many of those products, like nitrous oxide, are greenhouse gases) (Islam et

al., 1998).

63

CHAPTER 3

MATERIALS AND METHODS

3.1 Experiment 1 The soil health fallow trial

3.1.1 Research sites

This research involved a multi-location trial in an attempt to assess and improve the soil

biological health of continuously cropped taro fields through investigating the efficacy

of selected organic soil amendments. The experiment was conducted on four sites on the

two larger islands of Samoa (Plate 3.1): Upolu and Savaii. There were two research sites

on each of the islands, with one on each island being situated in the wet zone and the

other in the dry zone (Plate 3.1). The four trial sites identified for this research are

shown in Table 3.1.

Table 3.1 Field experimental sites for fallow trial

Island Village Rainfall zone Annual rainfall range (mm)

Upolu Salani High 4,000 – 5,000

Safaatoa Low 2,000 – 4,500

Savaii Siufaga High 4,500 – 6,000

Aopo Low 1,500 – 2,000

64

Plat

e 3.

1 Lo

catio

n of

the

rese

arch

site

s

Aop

o -

Low

ra

infa

ll

Siuf

aga –

high

ra

infa

ll si

te

Siuf

aga

- L

ow

rain

fall

Sala

ni –

hi

gh

rain

fall

65

3.1.2 Site characterisation and history

All the four sites were characterised with regards to their relative locations on the two

islands of Samoa. The dryland taro cropping systems and the farmer’s package of

management practices were described together with the historical details of each site.

The soil form and geological evolution of the sites were also detailed starting from their

early volcanic formations through subsequent stages of weathering thereafter. The sites

were also characterised for their soil profile (up to a depth of 100 cm), with details of the

horizons outlined.

Preliminary data for selected soil physical, chemical and fertility indicators for the top 0-

15 cm of the soil was also ascertained. The physical indicators included particle size

determination analysis using the hydrometer method and bulk density using 100 mL

core samples. The chemical indicators comprised of measurements of EC and pH H2O

(1:5 w/v) using EC 300 and EUTech pH meter, respectively; and determination of CEC

by 1 M NH4OAc percolation (pH 7.0) (Blakemore et al. 1987).

Determination of fertility indices were carried out through measurement of organic C

using the wet digestion method (Walkley and Black, 1934); Total N using semi-micro

Kjeldahl method (Blakemore et al. 1987); available P by Olsen et al. (1954) described

by (Blakemore et al. 1987); exchangeable cations by 1 M NH4OAc (pH 7.0) shaking

extraction method described by Daly et al. (1984); and, DTPA-extractable micro-

nutrient elements by method of Lindsay and Norvell (1978).

3.1.2.1 Salani, Falealili, High Rainfall Zone, Upolu

Crop management

The trial area was initially under dense grass ground cover. The farmer's practice usually

involves slashing and allowing for re-growth, before spraying with a systemic herbicide

followed by planting. Taro is the main crop with banana plants in the plot. The varieties

66

planted by the farmer are mainly the new hybrids from USP/SPC taro improvement

breeding program. There are no external inputs of chemical fertilisers and there were no

nitrogen fixing species planted in the plot. The farmer practices shifting cultivation and

entirely depends on the accumulation of nutrients through fallowing. The field has been

under continual cropping of taro for the past five years with short fallow durations of up

to four months.

Soil Form and Geology

Salani volcanics – a little over 30 cm thickness of surface weathering and consist mainly

of grey to black porphyritic basalt, vesicular pahoehoe basalt and rubbly aa basalt lava

most, with few, vesicle surfaces coated with a zeolite film; not eroded far enough to

expose either dykes or compacted rock. In those few places where rivers have cut into

Salani rocks, they have exposed a high proportion of rubbly rock. Scoria cones are

usually low, rounded, and breached. In places a layer of reddish highly weathered

material separates the flow rocks of the Salani volcanics from the rocks of older

formation but it is not possible to say whether this represents an old soil. In places the

Salani lavas are slightly andesitic in character. Selected characteristics features of the

site are outlined in Table 3.2 (a), (b) and (c) (Vasuidreketi, 2015).

Table 3.2(a) Characterisation of Salani Site Address Salani Village

Location South Eastern Upolu

Relative location 67.5 m to the right of the Falealili Assembly of God church

Elevation (above

mean sea level)

About 30 m

Soil Series Falealili clays, boulder and stony

Soil Name Falealili clay

Soil Classification

(Soil Taxonomy)

Oxic Humitropept, clayey-skeletal, oxidic, isohyperthermic

[As mapped by Wright (1963) and subsequent surveys by

67

Rijkse and MacLeod (1989)]

Land use Taro (Colocasia esculenta), bananas (Musa spp.) and t-grasses

(Paspalum conjugatum)

Date described 21/11/2013

Terrain Mainly easy rolling

Slope: 0-1o

Annual rainfall (mm) Average annual rainfall of 4,000 mm to 5,500 mm with a very

weak dry season (0 - 1 month with less than 100 mm rainfall)

Parent material Basalt (vesicular)

Erosion None to slight

Drainage Good

Topsoil depth (cm) 31 cm

Total rooting depth

(cm)

>80 cm

Limiting Horizon None

Table 3.2(b) Soil Profile Description of Salani Site

Horizon Depth (cm) Description

Ap

0-31

7.5YR 4/4 brown (moist) clay; fragments of reddish

weathered rock; fine to medium, moderately strong sub-

angular blocky structure; friable consistence; few fine

pores; few fine to medium roots, smooth diffuse boundary,

earthworm present (1.5 inch long).

Bw1 31-50

7.5YR 4/3 brown (moist) clay; fragments of reddish

weathered rocks (Fe oxides); fine to medium and weak to

moderate subangular blocky structure; friable consistence;

few fine pores; few fine roots, clear, wavy boundary.

50-64 7.5YR 4/3 brown (moist) clay (more than 60%), fragments

68

Bw2

of reddish weathered rock; fine to medium and weak to

moderate subangular blocky structure; friable consistence;

few fine pores; few fine roots; abrupt, wavy clear

boundary.

Bw3

64-86

7.5YR 4/2 (moist) brown gravelly clay; fragments of

reddish weathered rock; fine to medium and weak to

moderate subangular blocky structure; friable to firm

consistence; very few fine pores; very few fine roots;

abrupt and wavy boundary.

Cr >86 Weathered vesicular basalt, firm.

69

Sele

cted

soil

indi

cato

rs

Sele

cted

soil

indi

cato

rs o

f the

Sal

ani s

ite a

re p

rese

nted

in T

able

3.2

(c) b

elow

.

Tabl

e 3.

2(c)

.

S

elec

ted

soil

phys

ical

, che

mic

al a

nd fe

rtilit

y in

dica

tors

(0-1

5 cm

) of S

alan

i site

Phys

ical

Indi

cato

rs

Che

mic

al In

dica

tors

N

utrit

iona

l Ind

icat

ors

Parti

cle

size

ana

lysi

s (%

) B

ulk

Den

sity

(M

g/m

3

)

EC

(dS/

m)

pH

H2O

(1

:5)

CEC

pH

7

(cm

ol(+

)/kg)

Mac

ro-n

utrie

nts

DTP

A E

xtra

ctab

le M

icro

-nu

trien

ts

(mg/

kg)

Sand

Si

lt C

lay

OC

(%

)

Tota

l N

(%

)

Ols

en

P (m

g/kg

)

Exch

ange

able

B

ases

(cm

ol(+

)/kg)

K

Ca

Mg

Fe

Mn

Cu

Zn

43

24

33

0.77

0.

12

5.5

20.8

3 2.

99

0.28

3.

20

0.28

9.

01

0.68

47

.59

21.0

7 3.

76

0.85

Cla

y lo

am

Low

,

poro

us,

good

aera

tion

Ver

y lo

w1

Mod

erat

e

ly a

cidi

c1 M

ediu

m1

Low

1 Lo

w1

Ver

y lo

w1

Ver

y

low

1 M

ediu

m1

Low

1 V

ery

Hig

h2

Ver

y

Hig

h2

Ver

y

Hig

h2

Med

ium

2

1 Rat

ing

of B

lake

mor

e et

al.

(198

7)

2 Rat

ing

of B

uchh

olz

(198

3)

70

3.1.2.2 Safaatoa, Lefaga, Low Rainfall Zone, Upolu

Crop management

The land has been under continuous cultivation of taro for the past 5 years. The total

land area has been subdivided into subplots, and the harvested plots are fallowed for at

least 6 months before replanting. The farmer never uses any kind of fertiliser and there

are a few Erythrina trees (NFT species) around the area but not in the cultivated area.

The taro varieties of farmer's choice are mainly the new hybrid lines obtained from the

USP/SPC taro breeding program, grown at a spacing of 1.0 m x 1.0 m. Crop

maintenance practices include two weedings before the crop is 4 months old after which

the dense canopy formation shades out the weeds. The crop is harvested between 6 and 7

months of age.


Lefaga volcanics – are relatively unweathered, and without cavity filling. They are

assumed to have the lowest bulk density of any volcanic formation in Samoa, mainly

because the ratio of lava: rubble is low as possibly 1:10. The Lefaga volcanics consist

mainly of grey-black vitreous, porphyritic, and non-porphyritic basalts, more or less

vesicular and interbedded with aa. Andesitic lavas are seldom found. Lefaga volcanics

are thought to be younger and contain a higher proportion of scoria in thick irregular

beds with many volcanic bombs and lapilli. Selected characteristics features of the site

are outlined in Table 3.3 (a), (b) and (c) (Vasuidreketi, 2015).

71

Table 3.3(a) Characterisation of Safaatoa Site

Address Safaatoa

Location South Western Upolu

Relative location Safaatoa hills overlooking the Lefaga Bay

Elevation (above mean sea

level)

About 100 m

Soil series Lefaga clays, boulder and stony

Soil Name Lefaga clay

Soil Classification (Soil

Taxonomy)

Typic Fulvudand, medial-skeletal, amorphic,

isohyperthermic [As mapped by Wright (1963) and

subsequent surveys by Rijkse and MacLeod (1989)]

Land use/vegetation Taro (Colocasia esculenta), bananas (Musa spp.), t-

grasses (Paspalum comjugatum), goatweed

(Ageratum conyzoides), African tulip (Spathodea

campanulata), little bell (Ipomea triloba)


Terrain Easy to strongly rolling with very stony surface;

stones are loose fragments of highly vesicular aa lava

Slope: Pit site = 1°; Surrounding area = < 5°

Annual rainfall (mm) Average annual rainfall of 2,000 mm to 3,500 mm

with a weak dry season (1 - 2 months with less than

30 mm rainfall)

Parent material Basalt

Erosion Pit site = Slight; Surrounding area (taro field):

moderate

Drainage Well drained

Topsoil depth (cm) 0-24 cm

Total rooting depth (cm) > 100 cm

Limiting Horizon None

72

Table 3.3(b) Soil Profile Description of Safaatoa Site

Horizon Depth

(cm) Description

Ap

0-24

7.5YR 4/6 strong brown (moist) clay (higher clay content

than Salani site); fine to medium sub-angular blocky

structure; sticky and friable consistence; few fine to

medium pores, few fine roots; diffuse, smooth boundary.

[Note: topsoil from experimental site has a brown colour

(7.5YR 4/4) (Suspect that this is part of the B horizon since

moderate erosion is observed)]

Bw1

24-52

7.5YR 4/4 brown (moist) clay; fine to coarse, sub-angular

blocky structure; strong and slightly friable consistence;

few fine pores, few fine roots; diffuse, smooth boundary;

bouldery basalt rock embedded in horizon.

Bw2

52-70

7.5YR 5/6 strong brown (moist) clay; fine to coarse sub-

angular blocky structure; strong and slightly friable

consistence; few fine roots, few fine pores; diffuse, smooth

boundary; bouldery basalt rock embedded in horizon.

Bw3

70-100

7.5YR 4/6 strong brown (moist) clay; coarse to moderate,

subangular blocky and blocky structure; strong and slightly

friable consistence; few fine roots; few fine pores; diffuse,

smooth boundary.

Cr >100 Weathered basalt

73

Sele

cted

soil

indi

cato

rs

Sele

cted

soil

indi

cato

rs o

f the

Saf

aato

a si

te a

re p

rese

nted

in T

able

3.3

(c) b

elow

.

Tabl

e 3.

3(c)

.

S

elec

ted

soil

phys

ical

, che

mic

al a

nd fe

rtilit

y in

dica

tors

(0-1

5 cm

) of S

afaa

toa

site

Phys

ical

Indi

cato

rs

Che

mic

al In

dica

tors

N

utrit

iona

l Ind

icat

ors

Parti

cle

size

an

alys

is (%

) B

ulk

Den

sity

(M

g/m

3 )

EC

(dS/

m)

pH

H2O

(1

:5)

CEC

pH

7

(cm

ol(+

)/kg)

M

acro

-nut

rient

s D

TPA

Ext

ract

able

Mic

ro-

nutri

ents

(mg/

kg)

Sand

Si

lt C

lay

OC

(%

) To

tal

N

(%)

Ols

en

P (m

g/kg

)

Exch

ange

able

Bas

es

(cm

ol(+

)/kg)

K

Ca

Mg

Fe

Mn

Cu

Zn

40.1

23

.5

36.4

0.

84

0.22

5.

62

19.2

3 4.

63

0.43

2.

85

0.32

2.

74

1.07

38

.97

49.4

6 3.

45

2.86

Cla

y lo

am

Low

, por

ous,

good

aera

tion

Ver

y lo

w1

Mod

erat

ely

acid

ic1

Med

ium

1 M

ediu

m1

Med

ium

1 V

ery

low

1 Lo

w 1

Low

1 M

ediu

m1

Ver

y

Hig

h2

Ver

y

Hig

h2

Ver

y

Hig

h2

Ver

y

Hig

h2

1 Rat

ing

of B

lake

mor

e et

al.

(198

7)

2 Rat

ing

of B

uchh

olz

(198

3)

74

3.1.2.3 Siufaga, Faga, High Rainfall Zone, Savaii

Crop management

The research site has most coconut trees felled to make way for taro and other food

crops. Taro varieties planted are mostly a mixture of new improved taros from USP/SPC

taro breeding program. There are no external inputs of chemical fertilisers and there

were no nitrogen fixing species planted in the plot. The farmer practices shifting

cultivation and entirely depends on the accumulation of nutrients through fallowing. The

field has been under continual cropping of taro for the past five years with short fallow

durations of up to four months.


Lefaga volcanics – are relatively unweathered, and without cavity filling. They are

assumed to have the lowest bulk density of any volcanic formation in Samoa, mainly

because the ratio of lava:rubble is low as possibly 1:10. The Lefaga volcanics consist

mainly of grey-black vitreous, porphyritic, and non porphyritic basalts, more or less

vesicular and interbedded with predominantly aa flows. Andesitic lavas are seldom

found. Lefaga volcanics are thought to be younger and contain a higher proportion of

scoria in thick irregular beds with many volcanic bombs and lapilli. This is similar to the

Safaatoa site as the soil form and geology for both the sites originate from the similar

types of volcanic rocks. Selected characteristics features of the site are outlined in Table

3.4 (a), (b) and (c).

75

Table 3.4(a) Characterisation of Siufaga Site Address Siufaga

Location Eastern Savaii

Relative location 2 km inland from Siufaga Beach Resort in the

coconut zone


level)

About 15 m

Soil series Olomauga sandy clays, boulder and stony

Soil Name Olomauga gritty clay


Taxonomy)

Typic-Eutropept, fine oxidic, isohyperthermic [As

mapped by Wright (1963) and subsequent surveys by

Rijkse and MacLeod (1989)]

Land use/vegetation Taro (Colocasia esculenta), bananas (Musa spp.), t-

grasses (Paspalum comjugatum), goatweed

(Ageratum conyzoides), coconuts (Cocos nucifera),

Honolulu rose (Clerodendron fragrans), crowsfoot

grass (Eleusine indica).


Terrain Flat


with a weak dry season (1 - 2 months with less than

60 mm rainfall)

Parent material Basalt

Erosion None (trial site field)



Total rooting depth (cm) > 61 cm

Limiting Horizon Cr –Strong gravelly weathering scoria

76

Table 3.4(b) Soil Profile Description of Siufaga Site

Horizon Depth

(cm) Description

Ap 0-16

5YR 3/2 dark reddish brown gritty (silty) clay; fine to

coarse; sub-angular blocky structure; sticky and friable

consistence; few fine to medium pores, few fine roots;

diffuse, smooth boundary.

Bw1

16-47

5YR 4/3 reddish brown (moist) clay; medium-coarse to

slightly gravelly, sub-angular blocky structure; strong and

slightly friable consistence; medium pores, few fine roots;

diffuse, smooth boundary; bouldery basaltic rock

embedded in horizon.

Bw2

47-60

7.5YR 4/6 strong dark brown (moist) clay; coarse to

gravelly, sub-angular blocky structure; strong and slightly

friable consistence; few fine roots, medium macro pores;

diffuse, smooth boundary; bouldery basalt rock embedded

in horizon.

Cr > 61 Weathering scoria

77

Sele

cted

soil

indi

cato

rs

Sele

cted

soil

indi

cato

rs o

f the

Siu

faga

site

are

pre

sent

ed in

Tab

le 3

.4(c

) bel

ow.

Tabl

e 3.

4(c)

.

S

elec

ted

soil

phys

ical

, che

mic

al a

nd fe

rtilit

y in

dica

tors

(0-1

5 cm

) of S

iufa

ga si

te

Phys

ical

Indi

cato

rs

Che

mic

al In

dica

tors

N

utrit

iona

l Ind

icat

ors

Parti

cle

size

an

alys

is (%

) B

ulk

Den

sity

(M

g/m

3 )

EC

(dS/

m)

pH

H2O

(1

:5)

CEC

pH

7

(cm

ol(+

)/kg)

M

acro

-nut

rient

s D

TPA

Ext

ract

able

Mic

ro-

nutri

ents

(mg/

kg)

Sand

Si

lt C

lay

OC

(%

) To

tal

N

(%)

Ols

en

P (m

g/kg

)

Exch

ange

able

Bas

es

(cm

ol(+

)/kg)

K

Ca

Mg

Fe

Mn

Cu

Zn

57.2

21

.6

21.2

0.

75

0.45

5.

98

26.1

9 6.

33

0.60

5.

38

0.27

6.

99

5.02

80

.67

52.3

0 8.

17

6.04

Sand

y cl

ay lo

am

Low

, por

ous,

good

aera

tion

Med

ium

1 Sl

ight

ly

acid

ic1

Hig

h1 M

ediu

m1

Med

ium

1 V

ery

low

1 V

ery

Low

1

Med

ium

1 H

igh1

Ver

y

Hig

h2

Ver

y

Hig

h2

Ver

y

Hig

h2

Ver

y

Hig

h2

1 Rat

ing

of B

lake

mor

e et

al.

(198

7)

2 Rat

ing

of B

uchh

olz

(198

3)

78

3.1.2.4 Aopo, Low Rainfall Zone, Savaii

Crop management

An area of approximately 1.6 ha has been cropped while the rest of the land is still under

bush fallow. Since the area has been recently opened up for planting, the organic matter

levels are high giving good yields. There are no external inputs of chemical fertilisers

and there were no nitrogen fixing species planted in the plot. The farmer practices

shifting cultivation and entirely depends on the accumulation of nutrients through

fallowing. The field has been under continual cropping of taro for the past five years

with short fallow durations of up to four months.


Aopo volcanics – are relatively unweathered, and consist mainly of ropy pahoehoe and

rubbly aa vesicular and sometimes porphyritic basalts (olivine and feldspar phenocryst),

from two historic eruptions. The scoria cones associated with these historic eruptions are

small and insignificant from the point of view of soil-forming parent material. The soils

are mainly scrub-covered or lightly forested rocky land of undulating to rolling relief.

The surface has partly smooth sheets of pahoehoe lava and partly cindery aa. Soils are

mainly confined to rock fissures in pahoehoe areas, but a shallow, very gravelly or stony

soil forms a more or less continuous sheet in the aa areas. Narrow ‘pressure ridges’ of

upthrust lava blocks also occur; the margins of individual flows are likely rugged and

blocky. Selected characteristics features of the site are outlined in Table 3.5 (a), (b) and

(c).

79

Table 3.5(a) Characterisation of Aopo Site Address Aopo

Location North Western Savaii

Relative location 2 km Inland from foothill of Aopo village


level)

About 20 m

Soil series Aopo loamy clay , boulder and stony

Soil Name Aopo loamy sand


Taxonomy)

Lithic, Hapludoll, fragmental, mixed,

isohyperthermic [As mapped by Wright (1963) and

subsequent surveys by Rijkse and MacLeod (1989)]

Land use/vegetation Coconuts (Cocus nucifera), taro (Colocasia

esculenta), xanthosoma (Xanthosoma saggitifolium),

mile-a-minute (Mikania mikrantha).


Terrain Flat with fairy rugged surface


with a strong distinct dry season (7-8 months with

less than 30 mm rainfall)

Parent material Olivine Basalt

Erosion None (trial site field)



Total rooting depth (cm) > 35cm

Limiting Horizon Cr – > 36 cm almost continuous sheet of underlying

parent rock

80

Table 3.5(b) Soil Profile Description

Horizon Depth

(cm) Description

Ap 0-9

5YR 4/2 dark reddish grey (peaty) loamy sand; medium to

coarse; moderately developed fine nutty structure breaking

to crumb and single grain structure; non-sticky and non-

plastic; friable consistence; few fine to medium pores, few

fine roots; diffuse, sharp boundary.

Bw1

9-35

7.5YR 3/1 dark brown (moist) fine sandy loam; medium-

coarse very stony loamy sand, massive breaking to strong

crumb structure; non-sticky and non-plastic; very friable

consistence; medium pores, few fine roots; diffuse, sharp

boundary; bouldery basaltic rock embedded in horizon.

Cr > 36 Underlying more or less continuous sheet of cindery aa.

81

Sele

cted

soil

indi

cato

rs

Sele

cted

soil

indi

cato

rs o

f the

Aop

o si

te a

re p

rese

nted

in T

able

3.5

(c) b

elow

.

Tabl

e 3.

5(c)

.

S

elec

ted

soil

phys

ical

, che

mic

al a

nd fe

rtilit

y in

dica

tors

(0-1

5 cm

) of A

opo

site

Phys

ical

Indi

cato

rs

Che

mic

al In

dica

tors

N

utrit

iona

l Ind

icat

ors

Parti

cle

size

an

alys

is (%

) B

ulk

Den

sity

(M

g/m

3 )

EC

(dS/

m)

pH

H2O

(1

:5)

CEC

pH

7

(cm

ol(+

)/kg)

M

acro

-nut

rient

s D

TPA

Ext

ract

able

Mic

ro-

nutri

ents

(mg/

kg)

Sand

Si

lt C

lay

OC

(%

) To

tal

N

(%)

Ols

en

P (m

g/k

g)

Exch

ange

able

Bas

es

(cm

ol(+

)/kg)

K

Ca

Mg

Fe

Mn

Cu

Zn

54.0

13

.6

32.4

0.

57

0.35

6.

27

38.8

7 10

.18

0.89

7.

11

0.63

15

.40

5.50

26

.12

3.87

1.

98

3.48

Sand

y cl

ay lo

am

Low

, por

ous,

good

aera

tion

Low

1 M

ediu

m

Slig

htly

acid

ic1

Hig

h1 H

igh1

Hig

h1 V

ery

low

1 M

ediu

m1

Hig

h1 H

igh1

Ver

y

Hig

h2

Ver

y

Hig

h2

Ver

y

Hig

h2

Ver

y

Hig

h2

1 Rat

ing

of B

lake

mor

e et

al.

(198

7)

2 Rat

ing

of B

uchh

olz

(198

3)

82

3.1.3 The fallow treatments

The organic soil amendment (and inorganic combinations) of the fallow treatments

selected for use in this research in taro farming systems are given in Table 3.6. The

characterisation of biochar was done as per Table 3.7.

Table 3.6 Fallow treatments

Fallow Treatment Treatment Description

T1 - Farmer's practice

Management of organic residues from fallow and weed

control by slashing and burning or by use of herbicide

allowing subsequent decomposition of the surface mulch

(Control).

T2 - Mucuna

A six-month green manure cover crop of Mucuna pruriens

with the entire biomass produced being decomposed as

mulch

T3 - Erythrina Biomass and residues produced from Erythrina subumbrans

grown on- site for six months and decomposed as mulch

T4 - Mucuna + 200

kg/ha NPK

A six-month green manure cover crop with the entire

biomass produced being decomposed as mulch plus NPK

(12-5-20) chemical fertiliser at half the research

recommended rate of 400 kg/ha (i.e. 200 kg/ha).

T5 - Farmer’s practice

+

400 kg/ha NPK

Farmer’s practice together with the application of NPK

chemical fertiliser at the research recommended rate of 400

kg/ ha.

T6 - Biochar*

Biochar produced from coconut shells applied at a rate of 15

t/ha. Grass that grew in biochar treated plots was controlled

by the use of herbicide allowing subsequent decomposition

of the surface mulch.

* Biochar treatment was included only on the two Upolu sites.

83

Table 3.7 Characterisation of biochar

Property Units Methods Typical

values

Sample homogenisation mm Crush and sieve <2

Moisture content g/g Oven drying 0.14

Specific surface m2/g Carter et al. (1986) as modified by

Cerato and Lutenegger (2002).

290

pH (20:1 solution:solid

ratio)

pH Calibrated pH electrode 10.2

Cation exchange

capacity

µmol/g Boehm (1994), Goertzen et al.

(2010) and Oickle et al. (2010)

234

Maximum water

holding capacity

g/g Briggs and McLane, 1907; Briggs

and Shantz, 1912

0.312

Wettability (surface

tension)

mN/m Roy and McGill (2002) 68.3

Total C content g/g Walkley and Black (1934) 80

Total N % Blakemore et al. (1987) 0.31

P % Blakemore et al. (1987) 0.19

K % Blakemore et al. (1987) 1.16

Ca % Blakemore et al. (1987) 0.41

Mg % Blakemore et al. (1987) 0.29

Fe mg/kg Blakemore et al. (1987) 4537

Mn mg/kg Blakemore et al. (1987) 6

Cu mg/kg Blakemore et al. (1987) 33

Zn mg/kg Blakemore et al. (1987) 61

84

Application of soil amendments (treatments)

All the soil amendments were applied to take effect over a six month fallow period after

the existing taro crop was harvested. Only the two research sites on the island of Upolu

received biochar treatments while the two on the island of Savaii received only five

fallow treatments. The actual dates for fallow establishment, killing of fallow covers and

planting of taro, and harvesting of the taro crop for the four sites is given in Table 3.8

below. However, harvesting was only done for the three sites, as the produce was stolen

from the Aopo site.

Table 3.8 Actual dates of fallow establishment, killing of cover crops and planting

and harvesting of the taro crop for the four sites

Site Planting of fallow

cover crops

Spraying of the

fallow covers and

planting of the taro

crop

Harvesting of the taro

crop

Salani Jan 9, 2013 July 10-12, 2013 March 19, 2014

Safaatoa Jan 11, 2013 July 16-18, 2013 March 23, 2014

Siufaga Feb 20, 2013 Aug 28-29, 2013 April 24, 2014

Aopo Feb 21, 2013 Aug 30-31, 2013 No harvesting done *

* Theft of the taro crop at the Aopo site precluded harvesting

3.1.4 Plant culture

After the six month fallow duration, the following two taro cultivars from the Taro

Improvement Program, bred for taro leaf blight resistance, were planted at a spacing of

1.0 m x 1.0 m under dry land conditions:

i. C1 - Samoa I

ii. C2 - Samoa II

85

Characterisation of the taro cultivars

The taro cultivars were characterised using the standardised morphological descriptors

used for characterising aroid germplasm (Table 3.9).

Table 3.9 Characterisation of the taro cultivars

Descriptor Samoa 1 Samoa 2

Germplasm type Cultivated Cultivated

Progeny Cross between Pacific and

Indonesian lines

Cross between Pacific and

Malaysian lines

Growing conditions Lowland and upland Lowland and upland

Altitude Lowlands to high altitudes Lowlands to high altitudes

Botanical variety Dasheen Dasheen

Growth habit Erect Erect

Stolon formation Partly present Absent

Plant height (peak growth) Tall (100-150 cm) Tall (100-150 cm)

Shape of lamina Drooping lobes Drooping lobes

Orientation of lamina Tip pointing downwards Tip pointing downwards

Leaf lamina margin Undulated (broad waves) Entire

Lamina colour Dark green Normal green

Variegation of lamina Absent Absent

Sinus Narrow pointed (<45o) Wide pointed (>45o)

Vein junction Light green Light purple

Colour of leaf petiole Brown-green Dark purple

Variation on petiole Upper part darker Upper part darker

Flowering Often flowering Often flowering

Resistance against leaf

blight

Resistant Resistant

Maturity Intermediate (6-8 months) Intermediate (6-8 months)

Corm shape Conical Elliptical

Corm weight Medium (0.5-2.0 kg) Medium (0.5-2.0 kg)

86

Corm flesh colour Pink White

Eating quality/texture Good - fibrous texture and

slight hardness and dryness

Excellent - smooth texture

and slight sweetness

Yield potential (under

research package)

10-12 t/ha 12-15 t/ha

Sucker production 10-15 6-10

Storage qualities (without

weight loss at room

temperature)

7 days 10 days

3.1.5 Experimental design and size

The experimental arrangement that was employed for this field trial research was a split-

plot arrangement with the soil amendment applications arranged as main plot treatments,

which was then split to accommodate the two cultivars. For each of the four research

sites, the experiment was laid out using a randomised complete block design with four

replications (Appendix 1). The gross main plot size was 6 m x 6 m, which was split into

two sub-plots of 6 m x 3 m, to accommodate 18 plants of each of the two taro cultivars.

Of the 18 plants in each gross split plot, 8 plants from each of the two net split plots

were used for data collection.

3.1.6 Data collection

3.1.6.1 Meteorological data collection

The data for the mean monthly rainfall and temperature were collected from the

meteorological stations situated nearest to the four experimental sites for the entire

duration of the research.

3.1.6.2 Soil Parameters

To demonstrate the holistic interactions that occur due to the application of the different

organic amendments, measurements of biological/biochemical indicators were carried

87

out at all the four sites to assess the changes in soil properties over the fallow duration,

as well as over the growing phase of the taro crop, on a regular monthly basis.

Each test is considered to be an indication of the level of functioning. However,

indicator data are not meaningful unless a baseline or some reference condition is

available for comparison or unless relative comparisons between management systems

are made. Therefore, initial measurements of these parameters were evaluated prior to

the application of the soil amendment treatments.

The following data were collected for this trial:

3.1.6.2.1 Biochemical Indicators

The measurements for the following bio-chemical indicators were made on a monthly

basis during the fallow period as well as during the cover crop decomposition period and

the entire growth cycle of the taro crop.

i. Soil labile carbon

Soil labile carbon was analysed by the potassium permanganate (KMnO4) oxidation

method of Weil et al. (2003). Five grams of ≤ 2 mm sieved air dried soil, 25 mL of

33mM KMnO4 and 1 mL of 0.1M CaCl2 were mixed in 50 mL centrifuge tube. A

mechanical rocker was used to shake the samples for 2 minutes. After 2 minutes of

shaking, 5 minutes of settling time was allowed for the suspension. A clear supernatant

was obtained as 0.1M CaCl2 stimulated the flocculation of the soil particles and

therefore hastened the settling out of the suspension. One mL of the solution was

pipetted and added to 49 mL of distilled water into a separate clean tube and mixed

thoroughly. The absorbance of this solution was measured using a spectrophotometer set

at 550 nm. The quantity of labile soil C (oxidised by KMnO4) was distinguished from

the recalcitrant soil C (not oxidised by KMnO4), using a standard calibration curve. All

the samples were run in duplicates and results were expressed as mg/kg of labile soil C.

88

ii. Soil biological activity – Fluorescein diacetate hydrolysis activity (FDA)

Soil FDA hydrolysis activity was assayed by the method of Bonanomi et al. (2010). For

each sample, 4 x 5g ≤ 2 mm sieved of air dried were weighed in 50 mL acetone resistant

centrifuge tubes. Two mL of distilled water was added to the samples which were then

incubated at room temperature (~30oC) for 7 days. After 7 days of incubation, 20 mL of

potassium phosphate buffer (pH 7.6) and 200 µL of 2000 µg FDA/L (dissolved in

acetone) solution was added to the reaction tubes. The samples were mechanically

shaken (using a mechanical rocker) for 30 minutes to undergo the hydrolysis reaction.

Exactly after 30 minutes from the addition of the FDA solution, 20 mL of acetone

(equivalent volume of the phosphate buffer) was added to the tubes to terminate the

reaction. The tubes were centrifuged for 10 minutes at 2000 rpm. Additions of 20 mL of

potassium phosphate buffer (pH 7.6), followed by 20 mL of acetone was carried out for

the blank samples. The blank tubes were shaken by hand and centrifuged for 10 minutes

at 2000 rpm. The absorbance of the supernatant from the reaction and blank tubes were

measured at 490 nm using a spectrophotometer. The absorbance values of the blanks

were subtracted from the reaction tubes and the corresponding concentrations of

fluorescein released were calculated using a standard calibration curve. All the samples

were run in duplicates and the results were converted and expressed as mg of hydrolysed

FDA/kg of soil/hour.

iii. Potentially mineralisable nitrogen (PMN) – (anaerobic incubation)

The potentially mineralisable nitrogen (PMN) was analysed using the biological method

of Waring and Bremner (1964), which involves estimation of ammonium nitrogen

(NH4+-N) produced under waterlogged (anaerobic) conditions. The method employed

involved incubation of 8 g of ≤ 2 mm sieved fresh soil samples under anaerobic

conditions created by adding 10 mL of distilled water in a tube with as little head space

as possible. The samples were incubated at 40oC for 7 days. Biological activity during

the incubation period ensured development and maintenance of anaerobic conditions,

eliminating any possible nitrification-denitrification reactions at the soil-water interface.

89

After 7 days of anaerobic incubation, 30 mL of 2.67M KCl was added to the tubes,

creating 2.0 M solution. The samples were mechanically shaken for 1 hour and then

centrifuged for 10 minutes at 2000 rpm. Then 20 mL of supernatant together in the

presence of 0.2 g nitrogen-flushed heavy MgO (heated in an electric muffle furnace at

650oC for 2 hours), was analysed for ammonium concentration using the steam

distillation apparatus.

The amount of NH4+-N in the soil before incubation was determined by the same

procedure using 40 mL 2 M KCl and 0.2 g nitrogen-flushed MgO. All the samples were

run in duplicates and the mineralisable N was calculated from the difference in the

results between before and after incubation analysis. Results were expressed as mg/kg of

mineralisable N.

iv. Mineral nitrogen fluxes using the covered core in-situ incubation method

The covered core in situ aerobic incubation method of Adams and Attiwill (1986) was

employed to estimate the N mineralisation potential and mineral N fluxes of the soils

treated with different fallow cover crops on a monthly basis over their decomposition of

8 months. Since the cover crop residues were allowed to decompose as surface mulches,

5 cm deep cores were used; as 10 cm deep cores reflected large dilution effects within a

sample. These 5 cm deep PVC cores were perforated and covered with plastic sheet,

which allowed soil moisture and temperature to equilibrate with the surrounding soil

environment. The cores were placed confined adjacent to each other, so as to minimise

the effects of spatial heterogeneity. Duplicate soil samples at each monthly sampling,

were removed from the core with minimal disturbance, placed in sampling bags and

transported to the laboratory under frozen conditions.

The determination of the amount of mineral-N in solution by the distillation method of

Bremner and Keeney (1965) was adopted for laboratory analyses. In the laboratory, 10 g

of ≤ 2 mm-sieved fresh soil was extracted with 100 mL of KCl (10:1 extractant: soil) by

shaking for 1 hour. The shaken extracts were filtered, (and stored at 4-6oC where

90

distillation was not carried out on the same day), and 20 mL of the filtrate was used for

NH4+-N and NO3

--N determination using steam distillation apparatus.

For an individual sample, ammonia was separately distilled off first in the presence of

0.2 g nitrogen-flushed MgO (heated in an electric muffle furnace at 650oC for 2 hours),

for 3 minutes. Then 0.2 g of Devarda’s alloy was added to convert the remaining N to

ammonia. Nitrate-N was reduced to ammonia by Devarda’s alloy and the ammonia was

distilled over for 5 minutes, in the presence of 0.2 g nitrogen-flushed MgO. All the

samples were run in duplicates and the results were reported as mg/kg of NH4+-N and

NO3--N.

3.1.6.2.2 Biological indicators – study of nematode community analyses

Nematode extraction, enumeration and identification.

Nematode extraction, enumeration and identification was carried out prior to the

establishment of the fallow cover crops, after the fallow cover crops were sprayed at the

end of the six month growing period and at the end of fallow trial (at the time of taro

harvest).

Nematode extraction

Quantitative extraction of mobile nematodes was carried out by spreading out 200 g of

freshly sampled soil on paper tissue in 23 x 33 cm aluminium trays of 8 mesh/cm

phosphor-bronze gauze, just resting on shallow water (250 mL). The samples were

soaked for about 48 hours at room temperature (~30oC), with the trays being covered by

plastic sheets so as to avoid any evaporation. The suspension obtained after 48 hours,

was concentrated to 20 -30 mL without loss of sedimentation by carefully decanting in a

vial (Whitehead and Hemming, 2008) (Appendix 10).

Enumeration of total nematode counts

An estimation of the total number of nematodes extracted from the sample was made as

soon as possible, by measuring the volume of water in the vial and counting the number

91

of nematodes in 1 ml of the suspension, under a compound microscope. Where

nematode counts were not performed immediately, samples were kept in a refrigerator at

4-6oC. Nematode counts were expressed as number of nematodes/100g soil.

Identification of nematodes

Once the total number of nematodes had been estimated, a cover slip was placed on the

slide and the nematodes were immobilised for identification by gently heating the glass

slide over a low flame of a methanol burner. The following recognition features in key

areas of the nematode were used to get to most taxa levels:

� Mouth cavity – stoma, spear, stylet;

� Obvious features such as head ornaments;

� Stylet/spear shape and length;

� Median bulb;

� Basal bulb; and

� Tail shape.

Based on the above discretions and pictorial guides (Bongers, 1990), nematodes were

identified and categorised nematodes into five generally recognised trophic groups:

bacterivores, fungivores, predators, omnivores and plant parasites.

Calculation of principal components analysis of nematode indices

The following nematode indices were calculated:

i. Enrichment Index = 100[e/e(e+b)]

Enrichment (e) : [B1 and F2, where B = bacterivores, F = fungivores] and numbers

represent the coloniser-persister (c-p) value 1-5 (Bongers, 1990), structure

Basal(b): (B2 and F2) nematode communities (Ferris et al., 2001).

92

ii. Structure Index = 100[s/(s+b)]

Structure (s): (B3-B5, F3-F5, O3-O5, P2-P5, where O = omnivores, P = predators)

Basal(b): (B2 and F2) nematode communities (Ferris et al., 2001).

iii. Channel Index = 100 [0.8F2/(3.2B1 + 0.8 F2]

The proportion of C entering the soil food web through the decomposition of detritus (d)

was estimated by the sum of microbivorous nematodes belonging to fungal or bacterial

guilds with their respective weightings. Similarly the proportion of C being recycled in

the soil ecosystem through predation (p) was estimated by determining the sum of

nematodes belonging to predatory or omnivorous guilds. The proportion of C entering

the soil ecosystem through the activity of plant-parasitic nematodes was estimated by

their potential to cause damage. Plant-parasitic nematodes with the greater potential to

do damage were given a greater ranking than those that are commonly associated with

roots. For example, the proportion of C estimated to enter the soil through the root

channel (r) was determined through the sum of plant-parasitic nematodes such that r =

(0.8 R2 + 1.8 R3), where R2 = plant-parasitic nematodes with low damage potential,

which included Rotylenchulus reniformis, Heticotylenchus dihystera, Meloidogyne spp.

and other herbivorous stylet-bearing nematodes and R3 = plant-parasitic nematodes with

high damage potential to roots which included Radopholus similis, Helicotylenchus

multicinctus, Pratylenchus spp. and Hoplolaimus spp. The proportion of carbon moving

through nematodes representing the three channels were calculated as detritus (D),

predatory (P) and roots (R):

Detritus Index = 100 [d/(d + p + r)]

Predator Index = 100 [p/(d + p + r)]

Root Index = 100 [r/(d + p + r)]

93

Where: d = (3.2 B1 + 0.8B2 + 1.8 B3 + 0.8 F2), p = (1.8 P3 + 3.2 P4 + 3.2 O4), r = (0.8

R2 + 1.8 R3) and B = bacterivores, F = fungivores, O = omnivores, P = predators and R

= plant-parasites and numbers represent the coloniser-persister (c-p) value 1-5 (Bongers,

1990).

3.1.6.3 Plant Parameters

3.1.6.3.1 Dry matter yield and nutrient uptake of cover crops

The dry matter yields of grass fallow (farmer's practice), Erythrina subumbrans,

Mucuna pruriens and vegetation under coconut shel biochar treated plots were recorded.

Samples of the various cover crops and fallow vegetation were collected at six months

of age (before herbicide spraying) and oven dried to a constant weight at 65oC for dry

matter yield, nutrient concentration and nutrient uptake determination. Nitrogen,

phosphorus and potassium were measured after Kjeldahl digestion method for plant

samples as described by Blakemore et al. (1987) and Daly et al. (1984). Determination

of N was done by steam distillation, P by molybdovanadophosphoric acid (IBSNAT,

1987), and K, Ca, Mg, Zn, Fe, Mn, and Cu by atomic absorption spectrophotometry

(Chapman and Pratt, 1961; Prasad and Spiers, 1978). Nutrient contents were calculated

as the product of dry matter content and tissue nutrient concentration. Nutrient uptake

was calculated by multiplying the percentage nutrient content by the dry matter yields,

expressed as kg/ha.

3.1.6.3.2 Harvesting and yield of the taro crop

The taro was harvested at 8 months of age, the corms washed and dried before weighing.

The yields of the two cultivars of taro grown under the different fallow practices over

the 3 sites were ascertained and expressed as t/ha of fresh corm weight.

94

3.1.6.3.3 Dry matter yield and nutrient uptake of taro corms

Longitudinal sections of corm samples of the two taro cultivars grown under the various

fallow treatments over the three sites were collected and oven dried to a constant weight

at 65oC for dry matter yield and nutrient uptake determination. The procedure for

analyses of plant samples, as described above for the fallow cover crops, were employed

to determine the accumulation of N, P, K, Ca and Mg by the corms of the two cultivars.

Nutrient contents were calculated as the product of dry matter content and tissue nutrient

concentration. Nutrient uptake was calculated by multiplying the percentage nutrient

content by the dry matter yields, expressed as kg/ha.

3.1.7 Statistical analysis

All the data for the biochemical soil indices were subjected to repeated measures

analysis of variance to compare the significance of the fallow and time effects, together

with their interactions. Yield and nematode data collected were subjected to analysis of

variance for split- plot experiments. Paired sample t-test was employed to ascertain the

mean difference in the nematode activity before and after the soil was treated with the

organic amendments. Mean comparisons were carried out using least significant

differences where significant differences were found. Linear associations between the

evaluated biochemical soil parameters were determined using product-moment

correlation coefficients between variates. All the data analyses were carried out using the

Discovery Edition of the Genstat statistical software package (VSN International Ltd.,

2011).

3.2 Experiment 2 The soil incubation trial

3.2.1 Backgound

Soil biological activity varies on soils in different climatic zones as prevailing moisture

and temperature regimes as well as fallow systems and crop cover biomass production,

95

all influence the overall activity of soil organisms and the processes involved in the

release of plant nutrients contained in the organic residues. Therefore, this trial aims to

investigate the effects of site variation, on the decomposition phase of different rates of

cover crops under semi-controlled screen house conditions. Evaluation of selected

biochemical indicators were used as a measure of the overall performance and activity of

the soil as well the cover crops.

3.2.2 The trial description

The experiment was laid out in pots to study the effect of rate of decomposition of

selected organic material on selected soil bio-chemical indicators. This experiment

involved studying the decomposition phase of the organic materials applied at three

different rates on soil from the two Upolu sites over a period of three months under

semi-controlled (screen house conditions) environment. The pots were not rotated,

however, blocking was done within the screen house.

3.2.3 Application of the organic amendments

The soils from the two sites were incubated with three different rates (15, 30 and 45 t/ha

dry matter equivalent) of the four organic amendments used for the fallow trial: grass

(farmers practice - control), Mucuna pruriens, Erythrina subumbrans and biochar from

coconut shells, over a duration of four months under semi controlled conditions of a

screen house. The amendments were applied as shredded surface mulches so as to

simulate the field conditions of the fallow experiment. All the applications were made

on dry matter equivalent basis. The pots were kept at field capacity moisture content to

ensure optimal micro-biological activity.

96

3.2.4 Treatments, factors and levels

The treatments, factors and levels of the soil incubation experiment are given in Table

3.10 below.

Table 3.10 Treatments, factors and levels of the soil incubation experiment

Soil Fallow Crop Rates of Fallow Crop Treatment Combinations

Salani - SAL

Grass - G

15 t/ha T1 - SAL-G-15

30 t/ha T2 - SAL-G-30

45 t/ha T3 - SAL-G-45

Erythrina - E

15 t/ha T4 - SAL-E-15

30 t/ha T5 - SAL-E-30

45 t/ha T6 - SAL-E-45

Mucuna - M

15 t/ha T7 - SAL-M-15

30 t/ha T8 - SAL-M-30

45 t/ha T9 - SAL-M-45

Biochar - B

15 t/ha T10 - SAL-B-15

30 t/ha T11 - SAL-B-30

45 t/ha T12 - SAL-B-45

Safaatoa - SAF

Grass - G

15 t/ha T13 - SAF-G-15

30 t/ha T14 - SAF-G-30

45 t/ha T15 - SAF-G-45

Erythrina - E

15 t/ha T16 - SAF-E-15

30 t/ha T17 - SAF-E-30

45 t/ha T18 - SAF-E-45

Mucuna - M

15 t/ha T19 - SAF-M-15

30 t/ha T20 - SAF-M-30

45 t/ha T21 - SAF-M-45

Biochar - B

15 t/ha T22 - SAF-B-15

30 t/ha T23 - SAF-B-30

45 t/ha T24 - SAF-B-45

97

3.2.5 Experimental design

The treatment structure employed for this research was a three-factor factorial

arrangement with 2 soil types, 4 fallow crop/soil amendment types and 3 rates of fallow

crop/soil amendment applied to pots for a total of 24 treatment combinations. The

experiment was laid out using randomised complete block design with 3 replications

(Appendix 2).


The data for selected biochemical indices were collected on a monthly basis for three

month decomposition period. The indices and their significance in nutrient recycling are

given in Table 3.11 below.

Table 3.11 Biochemical soil health indices and their significance

Soil health

index

Organic

matter

substances

involved

End product Significance Predictor of soil

function

Labile

carbon organic matter carbon

energy for

microorganisms

organic matter

levels

Fluorescein

diacetate

hydrolysis

activity

(FDA)

organic matter

carbon and

various

nutrients

energy and nutrients

for microorganisms,

measure

microbial biomass

Biological

activity, organic

matter

decomposition

nutrient cycling

98

Potentially

mineralisable

nitrogen

(PMN)

nitrogen

compounds

ammonium

(NH4+-N)

Potentially plant

available NH4+- N

Potential available

N pool

Mineral

nitrogen organic matter

ammonium

(NH4+-N)

nitrate (NO3--

N)

Plant available N,

leaching potential

mineralisation

potential

Urease assay nitrogen (urea)

ammonia

(NH3) and

carbon dioxide

(CO2)

plant available

NH4+- N

N mineralisation

potential

Phosphatase

assay Phosphorus

phosphate

(PO42-)

plant available P P mineralisation

potential

3.2.6.1 Analysis of biochemical soil health indicators

The laboratory analysis for the soil labile carbon, fluorescein diacetate hydrolysis

activity (FDA), potentially mineralisable nitrogen (PMN) and mineral nitrogen for this

pot experiment were carried out using the same methods as described under the fallow

trial. The methods for soil urease and phosphatase assays are described in the succeeding

sections.

3.2.6.2 Assay of soil urease activity

Soil urease activity was assayed using the unbuffered method of Kandeler and Gerber

(1988). For each sample, 2.5 g of ≤ 2 mm sieved air dried soil was weighed into two

separate centrifuge tubes (reaction and blank). The sample in the reaction tube was

wetted with 1.25 mL of 0.08 M aqueous urea solution while the blank was wetted with

1.25 mL of distilled water. The samples were incubated at 37oC for 2 hours. At the end

99

of the incubation period and immediately before KCl addition, 1.25 mL of urea solution

was added to the blanks and 1.25 mL of distilled water to the reaction tubes. After the 2

hours of incubation, 25 mL of 2 M KCl was added to the samples and the mixture was

mechanically shaken for 30 minutes. The resulting suspension was centrifuged for 10

minutes at 2000 rpm. One milliliter of filtrate was diluted to 10 mL with distilled water

in a separate tube, and successively, 5 mL sodium salicylate and 2 mL of 0.1% sodium

dichloriocyanurate were added. The sodium salicylate solution was prepared for use by

mixing 100 mL of 0.12% sodium nitroprusside, 100 mL of 17% sodium salicylate and

100 mL of distilled water.

Optical density was determined at 690 nm after 30 minute incubation at room

temperature. The concentrations of NH4+-N of the samples were colorimetrically

calculated by reference to a calibration graph plotted from the results obtained with

diluted standards containing 0, 0.5, 1.0, 1.5 and 2.0 µg/mL NH4+-N. All the samples

were run in duplicates with the urease activity expressed as µg N hydrolysed /g dry soil

per 2 hours at 37oC. Where, colorimetric analyses were not carried out soon after

preparation, extracts were stored up to 24 hours in a refrigerator at 4-6oC.

3.2.6.3 Assay of soil phosphatase activity

The acid soil phosphatase activity was assayed by the modified method of Tabatabai and

Bremner (1969). For each sample, 1 g of ≤ 2 mm sieved air dried soil was weighed into

2 separate tubes (reaction and blanks). Additions of 200 µL of Tween 80 (5.25 mL of

Tween 80 dissolved with 94.75 mL distilled water) and 4 mL of McIlvaine (citrate-Na

phosphate) buffer (pH 6.5) were made to all the tubes (reaction and blanks). One

milliliter of 0.5M 4-nitrophenyl disodium orthophosphate hexahydrate (dissolved in

McIlvaine buffer) was added to the reaction tubes and an additional 1 mL of McIlvaine

buffer to the blanks. The samples were incubated for 37oC for 1 hour. After an hour of

incubation, the reaction was terminated by additions of 1 mL of 0.5M CaCl2 and 4 mL

of 0.5M NaOH. The samples were centrifuged for 10 minutes at 2000 rpm and the

optical density was measured at 405 nm. The concentration of p-nitrophenol released

100

was calculated with reference to a calibration curve of standards. All the samples were

run in duplicates and the phosphatase activity was expressed as µmol of p-nitrophenol

released/g soil/hour at 37oC.


All the data collected were subjected to repeated measures analysis of variance for 3

factor factorial treatment structure laid out in a randomised complete block design

structure. All the data analyses were carried out using the Discovery Edition of the

Genstat statistical software package.

3.3 Experiment 3 The taro nutrient uptake and partitioning experiment

3.3.1 Background

There is a scarcity of basic information regarding dry matter accumulation and nutrient

uptake and partitioning for the taro crop, particularly under intensive cropping systems

which are aimed at satisfying the crop demand of a growing population and supplying

corms for export markets. These data are essential for the development of technological

packages, especially involving nutrient inputs, growth simulation models, and decision

support system. This information is also critical for the establishment of taro breeding

programs aimed at raising the yield potential of taro.

3.3.2 Description of the trial

This experiment was conducted to investigate the nutrient uptake of the two improved

(blight resistant) taro cultivars grown for the fallow experiment, Samoa 1 and Samoa 2.

The experiment was executed under the semi controlled environment of a screen house,

with the taro being grown in pots. The soil used was a well drained Inceptisol (Oxic

Humitropept, clayey-skeletal oxidic isohyperthermic) with pH = 6.0; organic carbon =

3.2%; and exchangeable bases = 10.3 cmol(+)/kg of soil. The soil was air dried and

sieved through a 1 cm mesh. The potting bags were filled with 10 kg of soil each.

101

3.3.3 Nutrient supplementation and incubation

The entire package of macro and macronutrient elements, based on the soil pH, was

included for nutrient supplementation to each pot, carried out at recommended levels by

Usher and Grundon (2004) (Appendix 3). An incubation time of two weeks was allowed

before the planting of the two taro varieties.

3.3.4 Experimental design, layout and size

Suckers of two improved taro cultivars, Samoa 1 and Samoa 2, were planted in pots and

laid out in a split-plot arrangement, using randomised complete block design with five

replications. Each replication consisted of two main plots as the cultivars which were

split to accommodate eight monthly biomass harvests (Appendix 4), sampled for dry

matter accumulation and nutrient uptake and partitioning at different stages of plant

growth (Appendix 5). There were six data plants of each variety from each block (each

sub-plot) for each of the eight harvests totaling to 240 plants for each cultivar (480

plants for the whole experiment). The cultivars and harvests were completely

randomised within a block.


Six taro plants of each cultivar from a block were harvested at 30, 60, 90, 120, 150, 180,

210, and 240 days after planting (DAP), to ascertain the dry matter measurements and

total chemical analysis of individual plant parts. Plants in the sub-plots were harvested,

washed and separated into petioles, corms, roots and sucker components (Appendix 6).

Samples of the various plant parts were oven dried to a constant weight at 65oC for dry

matter determination. The dried samples were ground to pass through a 1.0-mesh screen

and analysed for N, P, K, Ca, Mg, Fe, Mn, Cu and Zn. The third most upper leaf laminar

(Appendix 7) was also analysed for these elements at 30, 60, 90, 120, 150, 180, 210, and

240 days after planting (DAP). Nitrogen was determined by the micro-Kjeldahl

procedure (IBSNAT, 1987), P by molybdovanadophosphoric acid (IBSNAT, 1987), and

K, Ca, Mg, Zn, Fe, Mn, and Cu by atomic absorption spectrophtometry (Chapman and

102

Prat, 1961; Prasad and Speirs, 1978). Nutrient content were calculated as the product of

dry matter content and tissue nutrient concentration.


All the data collected were subjected to analysis of variance using ANOVA for split plot

treatment arrangement laid out in a randomised complete block design structure. Best-fit

curves were determined using polynomial regression procedures of the Genstat program.

Only coefficients significant at P < 0.05 were retained in the model.

103

CHAPTER 4 RESULTS AND DISCUSSION

4.1 Experiment 1 The soil health fallow trial

4.1.1 Meteorological variables

4.1.1.1 Rainfall

The intra-annual distribution of rainfall data for the four experimental sites over the two

year research period, expressed as magnitude of monthly totals, is given in Figure 4.1a-d

below. The annual rainfall for the Salani site for year 2013 and 2014 was 4,379 and

5539 mm, respectively while for the Safaatoa site corresponding annual totals were

3,400 and 3,436 mm. For the Savaii sites, the annual totals were comparatively lower

with the Siufaga site receiving 3,032 and 2,945 mm respectively for year 2013 and 2014

while the Aopo site receiving 2,521 and 2,485 mm for the corresponding periods.

The decomposition of the annual rainfall into magnitude of its monthly components

revealed relatively drier months with the seasonality more noticeably pronounced in the

low rainfall sites. The month of October was found out to be receiving comparatively

lower rainfall, even in the high rainfall zones. The Aopo site showed an extended period

of reasonably lower rainfall (Fig. 4.1d).

104

0

100

200

300

400

500

600

700

800

JanFebMarApr

MayJunJul

AugSepOct

NovDecJanFebMarApr

MayJunJul

AugSepOct

NovDec

2013

2014

YEA

R

Monthly rainfall (mm)

0

100

200

300

400

500

600

700

800

JanFebMarApr

MayJunJul

AugSepOct

NovDecJanFebMarApr

MayJunJul

AugSepOct

NovDec

2013

2014

YEA

R


0

100

200

300

400

500

JanFebMarApr

MayJunJul

AugSepOct

NovDecJanFebMarApr

MayJunJul

AugSepOct

NovDec

2013

2014

YEA

R

Monthly rainfall (mm) 0

100

200

300

400

500

JanFebMarApr

MayJunJul

AugSepOct

NovDecJanFebMarApr

MayJunJul

AugSepOct

NovDec

2013

2014

YEA

R


(

a) S

alan

(b) S

afaa

toa

(c)

Siu

faga

(

d) A

opo

Figu

re 4

.1

Rai

nfal

l pat

tern

for t

he tw

o ye

ar re

sear

ch p

erio

d fo

r the

four

exp

erim

enta

l site

s.

105

4.1.1.2 Temperature

The monthly mean day temperature for the four experimental sites over the two year

research period, is given in Figure 4.2a-d below. The mean daily air temperature

recorded for the Salani site over the experimental period was 26.5oC with maximum

daily range of 23-34oC and minimum daily range of 20-26.5oC. The parallel data for the

Safaatoa site revealed a mean daily air temperature of 26.9oC with maximum daily range

of 25.8-35.5oC and minimum daily range of 18.2-27.2oC.

For the Savaii sites, the mean daily air temperature recorded for the Siufaga site over the

experimental period was 27.8oC with maximum daily range of 25.7-35.1oC and

minimum daily range of 17.8-32.8oC. Corresponding data for the Aopo site revealed a

mean daily air temperature of 27.4oC with maximum daily range of 26.4-35.5oC and

minimum daily range of 15.2-28.4oC.

106

2223242526272829

JanFebMarApr

MayJunJul

AugSepOct

NovDecJanFebMarApr

MayJunJul

AugSepOct

NovDec

2013

2014

YEA

R

Monthly mean day temperature (oC)

242526272829

JanFebMarApr

MayJunJul

AugSepOct

NovDecJanFebMarApr

MayJunJul

AugSepOct

NovDec

2013

2014

YEA

R


2526272829

JanFebMarApr

MayJunJul

AugSepOct

NovDecJanFebMarApr

MayJunJul

AugSepOct

NovDec

2013

2014

YEA

R


2526272829

JanFebMarApr

MayJunJul

AugSepOct

NovDecJanFebMarApr

MayJunJul

AugSepOct

NovDec

2013

2014

YEA

R

Mean monthly day temperature (oC)

(a) S

alan

i

(b)

Saf

aato

a

(c) S

iufa

ga

(d)

Aop

o

Figu

re 4

.2

Mea

n da

y te

mpe

ratu

re o

ver t

he tw

o ye

ar re

sear

ch p

erio

d fo

r the

four

exp

erim

enta

l site

s.

107

4.1.2 Soil biochemical indicators

4.1.2.1 Labile carbon

The labile carbon trends under the different cover crop fallow systems for the four

experimental sites over the entire duration of the research are presented in Figure 4.3 (a-

d).

Figure 4.3 Labile carbon trends for the four fallow sites under various fallow systems

108

The general labile C fluxes across all the fallow treatments over the two Upolu sites

revealed a very similar trend during the fallow phase as well the decomposition phase

(Fig 4.3a-d) However, the research sites from the island of Savaii revealed a very

different trend. The Upolu sites showed more fluctuations and compared to the Savaii

sites. This can be partially explained by the comparative land use intensites as well as

the age of the soils. Since the sites were sprayed with a systemic herbicide prior to the

establishment of the fallow crops, the organic matter from the remnant litter contributed

to the significant initial increase (P<0.001) in the labile C. However, the labile C levels

later declined as the cover crops got established primarily due to the soil organic carbon

being fixed into the biological structures of the growing cover crops. The suppression of

native organic carbon mineralisation was highest under biochar; supporting the findings

of Singh and Cowie (2014). During the decomposition phase, the labile C significantly

increased to a unimodal peak and then fluctuated thereafter as a net result of

decomposition and carbon assimilation by the taro crop before levelling off towards the

end of the decomposition period. The overall labile C trend across all the sites and

fallow practices over the entire duration of the research showed a highly significant

positive quadratic response (P<0.001; R2 = 0.26) (Fig 4.4). The findings are consistent

with the work of Liu et al. (2013) on soil organic carbon pools in dry land farming in

northwest China. However, Cowie et al. (2013) found no significant increases in their

work on carbon farming practices on soil carbon stocks in NSW, Australia.

109

Figure 4.4 Overall labile C trends

Repeated measures analysis for labile C under the different fallow systems over the four

sites revealed significant differences among the fallows with mucuna treated plots

excelling all the other fallow types except at the Aopo site (Table 4.1a). This can

partially be attributed to significantly higher biomass production by the mucuna cover

crop over all the sites. Analogous results were reported by Wang et al. (2013) in their

study on mineral soils to forest conversion in the subtropics. The relatively higher levels

of active carbon at Aopo site can be related to the practice of seasonal burning and forest

fires the site was historically subjected to. Zhao et al. (2012) reported similar temporary

increase in soil organic carbon pools following fires in North-eastern China.

Y = -0.0081x2 + 3.9574x + 851.42 R² = 0.2632

500

700

900

1100

1300

1500

1700

0 30 60 90 120 150 180 210 240 270 300 330

Labi

le C

(mg/

kg)

Time (days)

110

Table 4.1 (a) Table of predicted fallow means from repeated measures analysis for

labile carbon (mg/kg) under different fallow systems across all time


Fallow Site

Salani Safaatoa Siufaga Aopo

Grass 1076 b 1124 bc 1183 b 1499 a

Mucuna 1176 a 1206 a 1264 a 1502 a

Erythina 1104 b 1158 b 1206 b 1487 b

Biochar 1096 b 1090 c - -

F.Pr <0.001 <0.001 <0.001 <0.001

s.e.d 36 21 15 5.7

Repeated measures split plot in time analyses showed that the changes in labile carbon

levels were highly significant (P<0.001) over time across all the fallow systems for the

four sites. However, fallow-time interaction was only significant for the Safaatoa site

(P<0.013) (Table 4.1b). Nested classification repeated measures analyses for fallow

within a time point showed significant effects (P<0.05) of fallows on the soil labile

carbon as the decomposition progressed (Table 4.1c).

Table 4.1 (b) Table of predicted estimates for fallow x time interaction from repeated

measures split plot in time analysis for labile carbon across all fallow systems for the

four sites

Fallow x

Time

Site


F.Pr 0.272 0.013 0.771 0.056

s.e.d 81 49 46 19

111

Table 4.1 (c) Within time point comparison among fallow types for labile C for each

site

Time (fallow)

Site Salani Safaatoa Siufaga Aopo

F.Pr 0.036 <0.001 <0.001 <0.001 s.e.d 81 49 46 19

Nested classification analysis of variance for between sites revealed significant spatial

variations (P<0.001) between the sites with regards to mean levels of labile C, with the

sites on the island of Upolu having comparatively lower labile C than the Savaii sites

(Table 4.1d). This result agrees with Guinto et al. (2015) in their survey of the soil

health status of 40 taro exporting farms in Samoa where mean labile C carbon for Upolu

was 1229 mg/kg while that for Savaii was 1391 mg/kg. This may have resulted from the

Upolu sites being subjected to more intensive cultivation than the Savaii soils. The Aopo

site on the island of Savaii in the dry zone had the highest mean labile C as a result of

cumulative effects of seasonal forest fires that the site has been vulnerable to. Fallows

within site comparisons revealed significant fallow effects (P<0.048) on mean levels of

labile carbon with mucuna treated plots being comparatively higher than all the other

fallow types (Table 4.1a).

Table 4.1 (d) Between-site comparison for labile C (mg/kg).

Site Predicted mean for labile C

Salani 1113 c

Safaatoa 1145 c

Siufaga 1218 b

Aopo 1496 a

LSD (5%) 42.75

112

4.1.2.2 Fluorescein diacetate hydrolysis activity (FDA)

The broad spectrum microbial activity trends, measured as hydrolysis of fluorescein

diacetate, under the different cover crop fallow systems for the four experimental sites

over the entire duration of the research are illustrated in Figure 4.5 (a-d).

Figure 4.5 Microbial activity trends for the four fallow sites under various fallow systems

113

The general microbial activity fluxes across all the fallow treatments over the four sites

revealed highly variable trends among the four sites during the fallow phase as well as

the decomposition phase (Fig 4.5a-d). The Upolu sites were quite similar; however, the

research sites from the island of Savaii revealed a very different trend. Since the sites

were sprayed with a systemic herbicide prior to the establishment of the fallow crops,

the organic matter from the remnant litter contributed to the significant initial increase

(P<0.001) in microbial activity except for the Siufaga site where the activity

significantly decreased (P<0.001) due to preceding extended bare fallowing with

minimal organic matter inputs. During the decomposition phase, the microbial activity

significantly increased (P<0.001) to a unimodal peak at around 60-90 days and then

declined significantly thereafter. The rapid increase in microbial activity during the

onset of the decomposition phase can be ascribed to the higher levels of soil labile C

additions from the subsequent organic matter mineralisation resulting into an increased

microbial population. This finding is consistent with that of Tiemann and Grandy

(2014), reporting on prompt increase in extracellular enzyme activity following biomass

inputs in Uganda. In addition, greater rooting activity and associated microbial activity

during the establishment phase of the taro crop can also be linked to the high biological

activity during this time period. Balota and Chaves (2011) reported similar findings from

their investigation on microbial activity in soil cultivated with different summer legumes

in coffee crop in Brazil. The marked decline in microbial activity after peaking could be

attributed to the decline in microbial population as a result of a decline in labile C. The

continual decline thereafter can be partially explained by the suppressiveness of organic

matter levels with no significant additions, following mineralisation as well as the

relative decline in the plant parasitic nematode population as revealed by nematode

enumeration data at the end of the experiment. Potter et al. (1988) and McBride et al.

(2000) reported highly significant nematicidal effect of incorporated residues from

Brassica species and cereal rye, respectively, hypothesising that low molecular organic

acids could be responsible for nematode suppressing qualities.

114

Y= 1E-05x3 - 0.0078x2 + 1.2035x + 49.896 R² = 0.2778

0

20

40

60

80

100

120

140

160

0 30 60 90 120 150 180 210 240 270 300 330

mg

FDA

hyd

roly

sed/

kg s

oil/h

r

Time (days)

The overall microbial activity trend across all the sites and fallow practices over the

entire duration of the research showed a highly significant cubic response (P<0.001; R2

= 0.28) (Fig. 4.6). The soil biological activity increased significantly across all fallow

systems and sites and then decreased to all time low as the decomposition progressed

before showing signs of an increase towards the end of the decomposition period. This

can be attributed to a lot of sucker production by both the cultivars of taro as well as

organic additions from weeds and the mother plant approaching senescence.

Figure 4.6 Overall microbial activity trends

Repeated measures analysis for microbial activity under the different fallow systems

over the four sites revealed significant differences among the fallows with mucuna

treated plots outdoing all the other fallow types (Table 4.2a). This can partially be

attributed to significantly higher biomass production of the mucuna cover crop over all

the sites, subsequently, resulting into higher microbial population. Furthermore, the

115

nitrogen fixing phenomenon of the legumes could have also played a significant role in

contributing to the high levels of biological activity under mucuna treated plots as also

reported by Balota and Chaves (2011). Comparable levels of biological activity under

biochar-treated plots can be ascribed to the higher moisture retention properties of the

coconut husk biochar material.

Table 4.2 (a) Table of predicted fallow means from repeated measures analysis for FDA

(mg/kg) under different fallow systems across all time points for the four

sites

Fallow Site


Grass 82 b 90 b 73 b 88 c

Mucuna 89 a 100 a 87 a 103 a

Erythina 80 b 94 b 77 b 96 b

Biochar 84 b 89 b - -

F.Pr 0.029 0.034 <0.001 <0.001

s.e.d 3.1 3.7 2.3 2.8

Repeated measures split plot in time analyses showed that the changes in microbial

activity levels were highly significant (P<0.001) over time points across all the fallow

systems for the four sites. Fallow-time interactions were also highly significant for all

the sites except Siufaga (Table 4.2b). This can be accredited to relatively higher biomass

production by all the cover crops at the Siufaga site which falls in the high rainfall zone

of Savaii. Nested classification repeated measures analyses for fallow within a time

point showed highly significant effects (P<0.001) of fallows on the soil microbial

activity as the decomposition of the fallow litter progressed (Table 4.2c).

116

Table 4.2b Table of predicted estimates for fallow x time interaction from repeated

measures split plot in time analysis for FDA (mg/kg) across all fallow

systems for the four sites

Fallow x Time Site


F.Pr 0.001 0.020 0.273 0.023

s.e.d 8.0 7.4 8.1 7.5

Table 4.2c Within time point comparison among fallow types for FDA (mg/kg) for

each site

Time (fallow) Site


F.Pr <0.001 0.020 <0.001 <0.001

s.e.d 8.0 7.4 8.1 7.5


variations (P<0.001) between the sites with regards to mean levels of soil microbial

activity, with the sites on the drier sides of both the islands being comparatively more

biologically active soils (Table 4.2a and Table 4.2d). This can be ascribed to the higher

weed infestation in the Safaatoa site whereby chemical weed control ensured regular

additions of organic matter to the soil environment. For the Aopo site, the comparatively

higher biologically active soil can rationally be linked to the higher levels of organic

carbon. Reddy et al. (2013) reported parallel findings using the same FDA hydrolysis

method. Fallows within site comparisons revealed significant fallow effects (P<0.003)

on mean levels of biological functioning with mucuna treated plots being comparatively

higher than all the other fallow types (Table 4.2a). Mendes et al. (1999) showed that

FDA hydrolysis can be used to differentiate between a winter fallow treatment and a

cereal or legume cover crop. Bandick and Dick (1999) also showed that FDA is sensitive

to cover cropping and N fertilisation.

117

Table 4.2 (d) Between site comparison for soil biological activity (mg FDA

hydrolised/kg soil/hr.).

Site Predicted mean for soil biological activity

Salani 84 b

Safaatoa 94 a

Siufaga 79 b

Aopo 95 a

LSD (5%) 5.0

118

4.1.2.3 Potentially mineralisable nitrogen (PMN) (Anaerobic incubation)

The potentially mineralisable nitrogen trends under the different cover crop fallow

systems for the four experimental sites over the entire duration of the research are shown

in Figure 4.7 (a-d).

Figure 4.7 Potentially mineralisable nitrogen trends for the four fallow sites under various fallow systems

119

Y = -2E-05x3 + 0.0124x2 - 1.2825x + 56.282 R² = 0.751

0

40

80

120

160

200

240

280

0 30 60 90 120 150 180 210 240 270 300 330Pote

ntia

lly m

iner

alis

able

N (m

g/kg

)

Time (days)

The potentially mineralisable N dynamics across all the fallow treatments over the four

sites revealed a very similar trend during the fallow phase as well the decomposition

phase (Fig 4.7a-d). The PMN levels declined slightly as the cover crops established

indicating N immobilisation by microbes as well as cover crop vegetation after initial

mineralisation of the remnant organic litter. The significant increase in PMN from the

onset of the decomposition phase to a unimodal maximum at around 120 days illustrates

net mineralisation of N. This is actually the time point at which the taro crop achieved its

maximum vegetative growth. The fluctuations thereafter show the net result of N

mineralisation synchronised together with the corm development phase of the taro crop.

Wagger (1989) reported similar findings during his exponential modelling of soil

nitrogen mineralisation. The overall labile N (biologically active soil nitrogen) trend

across all the sites and fallow practices over the entire duration of the research showed a

highly significant cubic response (P<0.001; R2 = 0.75) (Fig 4.8).

Figure 4.8 Potentially mineralisable N trend

120

Repeated measures analysis for potentially mineralisable nitrogen under the different

fallow systems over the four sites revealed significant differences (P<0.05) among the

fallows with mucuna treated plots out-mineralising all the other fallow types (Table

4.3a). This can be ascribed to significantly higher biomass production of the mucuna

cover crop over all the sites, subsequently, resulting into higher additions of N to the

mineralisable pool. Furthermore, the biological nitrogen fixing phenomenon of the

legumes could have also played a significant role in contributing to the high levels of

mineralisable N pool under mucuna treated plots. Goh and Chin (2007) and Ngome et

al. (2011) reported that 70% of N uptake by mucuna fallow was through biological

fixation while Chikowo et al. (2004) and Sanginga et al. (2001) concluded that mucuna

fallow crop biologically fixed up to 96% and 91% of the accumulated N, respectively.

Decker et al. (1994) reported similar contributions of legumes towards soil PMN levels.

Table 4.3(a) Table of predicted fallow means from repeated measures analysis for

PMN (mg/kg) under different fallow systems across all time points for

the four sites

Fallow Site


Grass 67 b 67 c 75 b 86 b

Mucuna 84 a 88 a 108 a 121 a

Erythina 64 b 76 b 84 b 89 b

Biochar 70 b 64 c - -

F.Pr 0.001 <0.001 0.029 <0.001

s.e.d 4.2 4.5 5.9 5.8

Repeated measures split plot in time analyses showed that the changes in mineralisable

N pools were highly significant (P<0.001) over time points across all the fallow systems

for the four sites. Fallow-time interactions were also highly significant for the all the

sites (Table 4.3b). This can be attributed to the progression of the organic matter

121

mineralisation process which ensured continuous additions to the soil pool of

biologically active nitrogen. Nested classification repeated measures analyses for fallow

within a time point showed highly significant effects (P<0.001) of fallows on the

mineralisable pool of N, as the decomposition progressed (Table 4.3c).

Table 4.3b Table of predicted estimates for fallow x time interaction from repeated

measures split plot in time analysis for PMN (mg/kg) across all fallow

systems for the four sites

Fallow x Time Site

Salani Safaatoa Siufaga Aopo F.Pr <0.001 <0.001 <0.001 <0.001 s.e.d 10.1 10.4 10.3 12.0

Table 4.3c Within time point comparison among fallow types for PMN (mg/kg) for each site

Time (fallow) Site

Salani Safaatoa Siufaga Aopo F.Pr <0.001 <0.001 <0.001 <0.001 s.e.d 10.1 10.4 10.3 12.0


variations (P<0.001) between the sites with regards to mean levels of biologically active

N pools, with the sites on the island of Savaii supporting comparatively larger active N

pools than the sites on the island of Upolu (Table 4.3a and Table 4.3d). This can be

ascribed to the Upolu sites being subjected to more intensive cultivation than the Savaii

soils and therefore, N recovery being lower during fallow periods. The findings of

Sharifi et al. (2008) confirm these dynamics of nitrogen availability comparisons

between farming systems. In addition, the Siufaga site had the highest biomass

production comparatively. For the Aopo site, the comparative higher labile N levels soil

can reasonably be linked to the higher levels of organic carbon. Fallows within site

122

comparisons revealed significant fallow effects (P=0.001) on mean levels of biologically

active N with mucuna treated plots being comparatively higher than all the other fallow

types (Table 4.3a).

Table 4.3 (d) Between site comparison for potentially mineralisable N (mg/kg).

Site Predicted mean for potentially mineralisable N

Salani 71 b

Safaatoa 74 b

Siufaga 89 a

Aopo 99 a

LSD (5%) 11.6

123

0

10

20

30

40

50

60

70

80

0 30 60 90 120 150 180 210 240

NH

4+ - N

(mg/

kg)

Time (Days) Grass Mucuna

0

10

20

30

40

50

60

70

80

90

0 30 60 90 120 150 180 210 240N

H4+ -

N (m

g/kg

)

Time (Days) Grass Mucuna Erythrina Biochar

0

10

20

30

40

50

60

70

80

90

100

0 30 60 90 120 150 180 210 240

NH

4+ - N

(mg/

kg)


0

20

40

60

80

100

120

0 30 60 90 120 150 180 210 240

NH

4+ - N

(mg/

kg)


4.1.2.4 Mineral nitrogen fluxes from in-situ covered core aerobic incubation

4.1.2.4.1 Ammonium nitrogen

The mean ammonium nitrogen mineralisation trends from in-situ covered core aerobic

incubation, under the different cover crop fallow systems, for the four experimental sites

over the cover crop decomposition and taro growing phase of the research are presented


(a) Salani (b) Safaatoa

(c) Siufaga (d) Aopo

Figure 4.9 Ammonium nitrogen fluxes for the four fallow sites under various fallow

systems

124

Y = 1E-05x3 - 0.0048x2 + 0.7899x + 10.667 R² = 0.168

0

20

40

60

80

100

120

0 30 60 90 120 150 180 210 240

NH

4+ - N

(mg/

kg)

Time (days)

The ammonium nitrogen dynamics, from the embedded covered core in-situ aerobic

incubation procedure, across all the fallow treatments over the four sites revealed highly

variable trends during the decomposition phase (Fig 4.9a-d; Table 4.4a). The Salani site

had an initial increase with the onset of the decomposition across all the fallow

treatments while the Siufaga and Aopo sites had initial declines. The Safaatoa site

responded differently to different fallow practices. However, all the sites attained a

unimodal peak at around 120 days except for the Aopo site which peaked at around 90

days of incubation. The sites that were in the high rainfall zones of both the islands and

subsequently had higher biomass production had comparatively longer decomposition

period before attaining short term qusi-equilibrium at around 180 days from the onset of

decomposition. The Safaatoa site achieved some short term qusi-equilibrium at around

150 days while the Aopo site showed a non-equilibrium pattern particularly owing to a

higher natural level of organic carbon.

The overall ammonium N mineralisation trend across all the sites and fallow practices

over the entire duration of the research showed a highly significant cubic response

(P<0.001; R2 = 0.17) (Fig 4.10).

Figure 4.10 Ammonium N trend

125

Repeated measures analysis for mineralised ammonium nitrogen under the different

fallow systems over the four sites revealed significant differences (P<0.05) among the

fallows with mucuna treated plots exhibiting a dominant effect over all the other fallow

types (Table 4.4a). This can be attributed to significantly higher biomass production of

the mucuna cover crop over all the sites, subsequently, resulting in higher additions of N

to the mineralised ammonium pool. Furthermore, the biological nitrogen fixing

phenomenon of the legumes could have also played a significant role in contributing to

the high levels of mineralised ammonium N to the soil pool under mucuna treated plots.

. Goh and Chin (2007) and Ngome et al. (2011) reported that 70% of N uptake by

mucuna fallow was through biological fixation while Chikowo et al. (2004) and

Sanginga et al. (2001) concluded that mucuna fallow crop biologically fixed up to 96%

and 91% of the accumulated N, respectively. Briggs et al. (2005) and Handayanto et al.

(1997) reported comparable nitrogen accumulations in organic farming.


ammonium - N (mg/kg) under different fallow systems across all time


Fallow Site


Grass 34 c 32 b 45 b 50 b

Mucuna 57 a 53 a 70 a 74 a

Erythina 38 bc 36 b 52 b 57 b

Biochar 42 b 35 b - -

F.Pr <0.001 <0.001 0.010 <0.001

s.e.d 2.4 2.8 4.6 4.5

CV (%) 53.1 52.5 41.61 94.3

126

Repeated measures split plot in time analyses showed that the changes in mineralised

ammonium N pools were highly significant (P<0.001) over time points across all the

fallow systems for the four sites. Fallow-time interactions were also highly significant

for all the sites (Table 4.4b). This can be attributed to the progression of the organic

matter mineralisation process which ensured continuous additions to the soil

ammoniacal nitrogen pool. Nested classification repeated measures analyses for fallow

within a time point showed highly significant effects (P<0.001) of fallows on the soil

ammoniacal nitrogen pool as the decomposition progressed (Table 4.4c). This can be

credited to the differences in the C:N ratios of the decomposing cover crop residues.

Legumes (mucuna and erythrina) having lower C:N ratios decomposed more rapidly

than grass and the highly recalcitrant biochar.

Table 4.4 (b) Table of predicted estimates for fallow x time interaction from repeated

measures split plot in time analysis for ammonium-N (mg/kg) across all

fallow systems for the four sites

Fallow x Time Site


F.Pr 0.022 0.905 0.060 0.606

s.e.d 7.0 7.2 7.2 9.4

Table 4.4 (c) Within time point comparison among fallow types for ammonium-N

(mg/kg) for each site

Time (fallow) Site

Salani Safaatoa Siufaga Aopo F.Pr <0.001 <0.001 0.014 0.006 s.e.d 7.0 7.2 7.2 9.4

127


variations (P<0.001) between the sites with regards to mean levels of mineralised

ammonium N pools, with the sites on the island of Savaii supporting comparatively

larger ammonium N pools than the sites on the island of Upolu (Table 4.4a and Table

4.4d). This can be ascribed to the Upolu sites being subjected to more intensive

cultivation than the Savaii soils resulting into lower levels of native N and C. In

addition, the Siufaga site had the highest biomass production comparatively. For the

Aopo site, the comparative higher ammonium N levels soil can reasonably be linked to

the higher levels of organic carbon. Fallows within site comparisons revealed significant

fallow effects (P=0.001) on mean levels of biologically active ammonium N with

mucuna treated plots being comparatively higher than all the other fallow types (Table

4.4a).

Table 4.4 (d) Between site comparison for ammonium N (mg/kg).

Site Predicted mean for ammonium N

Salani 43 c

Safaatoa 39 c

Siufaga 56 b

Aopo 61 a

LSD (5%) 4.7

128

0

200

400

600

800

1000

1200

1400

1600

0 30 60 90 120 150 180 210 240

NO

3- -N

(mg/

kg)


0

200

400

600

800

1000

1200

0 30 60 90 120 150 180 210 240N

O3- -

N (m

g/kg

)


0

200

400

600

800

1000

1200

1400

1600

0 30 60 90 120 150 180 210 240

NO

3- - N

(mg/

kg)

Time (Days)

Grass Mucuna Erythrina

0

200

400

600

800

1000

1200

1400

1600

1800

2000

0 30 60 90 120 150 180 210 240

NO

3- - N

(mg/

kg)

Time (Days)

Grass Mucuna Erythrina

4.1.2.4.2 Nitrate nitrogen

The mean nitrate nitrogen mineralisation trends from the in-situ covered core aerobic

incubation, under the different cover crop fallow systems, for the four experimental sites

over the cover crop decomposition and taro growing phase of the research are presented



(c) Siufaga (d) Aopo Figure 4.11 Nitrate nitrogen fluxes for the four fallow sites under various fallow

systems

129

Y = -0.0243x2 + 9.4894x + 58.176 R² = 0.2925

0

400

800

1200

1600

2000

0 30 60 90 120 150 180 210 240

NO

3- - N

(mg/

kg)

Time (days)

The nitrate nitrogen dynamics, from the embedded covered core in-situ aerobic

incubation procedure across all the fallow treatments over the four sites revealed a very

similar trend during the decomposition phase (Fig 4.11a-d). All the sites attained a

unimodal peak at around 90 - 120 days except for the Safaatoa site which peaked at

around 150 days of incubation. This can be partially explained by slower rate of

decomposition process in this low rainfall site. All the sites achieved some short term

qusi-equilibrium thereafter. Comparatively, nitrate nitrogen levels were much higher

than ammonium nitrogen levels, indicating rapid nitrification under the given climatic

and soil environments. This also denotes a higher level of N leaching as nitrate remains

in the soil solution, where it is subjected to leaching losses and not adsorbed to the soil

colloids. Comparable trends were observed by Schmidt et al. (1999) and Watson et al.

(1997) with the use of legume breaks in organic farming rotations.

The overall nitrate N mineralisation trend across all the sites and fallow practices over

the entire duration of the research showed a highly significant quadratic response

(P<0.001; R2 = 0.29) (Fig 4.12).

Figure 4.12 Nitrate N trend

130

Repeated measures analysis for mineralised nitrate nitrogen under the different fallow

systems over the four sites revealed significant differences (P<0.05) among the fallows

with mucuna treated plots exhibiting an overriding effect over all the other fallow types

(Table 4.5a). This can be attributed to significantly higher biomass production of the

mucuna cover crop over all the sites, subsequently, resulting into higher additions of N

to the mineralised nitrate pool. Furthermore, the biological nitrogen fixing phenomenon

of the legumes could have also played a significant role in contributing to the high levels

of mineralised nitrate N to the soil pool under mucuna treated plots. Chikowo et al.

(2004) and Sanginga et al. (2001) concluded that mucuna fallow crop biologically fixed

up to 96% and 91% of the accumulated N, respectively.


nitrate - N (mg/kg) under different fallow systems across all time points

for the four sites

Fallow Site


Grass 596 bc 413 b 720 b 968 b

Mucuna 925 a 653 a 1160 a 1341 a

Erythina 531 c 485 b 889 b 1100 b

Biochar 681 b 475 b - -

F.Pr <0.001 <0.001 <0.001 <0.001

s.e.d 70 67 109 94

CV (%) 65.7 66.8 74.5 88.4

Repeated measures split plot in time analyses showed that the changes in mineralised

nitrate N pools were highly significant (P<0.001) over time points across all the fallow

systems for the four sites. Fallow-time interactions was only significant for the Aopo site

(P=0.032) (Table 4.5b). The interaction at other sites being not significant can be

attributed to a large degree of variability in the data over time and between fallow

131

systems. Nested classification repeated measures analyses for fallow within a time point

showed highly significant effects (P<0.001) of fallows on the soil nitrate nitrogen pool

as the decomposition progressed, except for the Safaatoa site (Table 4.5c). This can

credited to the differences in the C:N ratios of the decomposing cover crop residues.

Legumes (mucuna and erythrina) having lower C:N ratios, decomposed more rapidly

than grass and the highly recalcitrant biochar.

Table 4.5(b) Table of predicted estimates for fallow x time interaction from repeated

measures split plot in time analysis for nitrate - N (mg/kg) across all

fallow systems for the four sites

Fallow x Time Site

Salani Safaatoa Siufaga Aopo F.Pr 0.112 0.948 0.882 0.032 s.e.d 166 135 189 162

Table 4.5(c) Within time point comparison among fallow types for nitrate - N (mg/kg)

for each site

Time (fallow) Site

Salani Safaatoa Siufaga Aopo F.Pr 0.004 0.479 0.007 0.012 s.e.d 166 135 189 162

Nested classification analysis of variance for between sites revealed highly significant

spatial variations (P<0.001) between the sites with regards to mean levels of mineralised

nitrate N pools, with the sites on the island of Savaii having comparatively larger nitrate

N pools than the sites on the island of Upolu (Table 4.5a and Table 4.5d). This can be

ascribed to the Upolu sites being subjected to more intensive cultivation than the Savaii

soils resulting into lower levels of native N and C. This is consistent with the findings of

Philipps et al. (1995). In addition, the Siufaga site had the highest biomass production

comparatively. For the Aopo site, the comparative higher nitrate N levels soil can

132

reasonably be linked to the higher levels of organic carbon. Fallows within site

comparisons revealed significant fallow effects (P<0.001) on mean levels of biologically

active nitrate N with mucuna treated plots being comparatively higher than all the other

fallow types (Table 4.5a).

Table 4.5 (d) Between site comparison for nitrate N (mg/kg).

Site Predicted mean for nitrate N

Salani 683 c

Safaatoa 507 d

Siufaga 913 b

Aopo 1136 a

LSD (5%) 99.6

133

0

500

1000

1500

2000

2500

0 30 60 90 120 150 180 210

Min

eral

N (m

g/kg

)

Time (Days) Grass Erythrina Mucuna Biochar

0

500

1000

1500

2000

0 30 60 90 120 150 180 210

Min

eral

N (m

g/kg

)

Time (Days) Grass Erythrina Mucuna Biochar

0

500

1000

1500

0 30 60 90 120 150 180 210

Min

eral

N (m

g/kg

)

Time (Days) Grass Erythrina

0

500

1000

1500

2000

2500

0 30 60 90 120 150 180 210

Min

eral

N (m

g/kg

)

Time (Days) Grass Erythrina

4.1.2.4.3 Cumulative net nitrogen mineralisation

The sum of ammonium-N and nitrate-N at each time point for each of the fallow types

was calculated. The difference between two successive measurements was used to

calculate the net N mineralisation. The cumulative net nitrogen mineralisation trends

from the embedded in-situ covered core aerobic incubation, under the different cover

crop fallow systems, for the four experimental sites over the cover crop decomposition

and taro growing phase of the research are shown in Figure 4.13 (a-d).


(c) Siufaga (d) Aopo

Figure 4.13 Net cumulative nitrogen mineralisation trends for the four sites under

various fallow systems

134

Y = -0.0435x2 + 22.598x - 394.06 R² = 0.5324

0

1000

2000

3000

4000

5000

6000

0 30 60 90 120 150 180 210

Cum

ulat

ive

net m

iner

al N

(m

g/kg

)

Time (days)

The cumulative mineral nitrogen dynamics, from the embedded covered core in-situ

aerobic incubation procedure across all the fallow treatments over the four sites was

monitored to reveal very similar trends during the decomposition phase (Fig 4.13a-d).

All the sites attained short term qusi-equilibrium at around 120 days of incubation

except for the Safaatoa site which peaked at around 150 days of incubation. This can be

partially explained by slower rate of decomposition process in this low rainfall site.

The dynamics of the availability of mineral nitrogen from various organic matter

amended soils showed that maximum availability synchronised well with the peak

vegetative growth and high levels were maintained during the corm development phase

of the taro crop.

The overall cumulative net N mineralisation trend across all the sites and fallow

practices over the entire duration of the research showed a highly significant quadratic

response (P<0.001; R2 = 0.53) (Fig 4.14). Carpenter-Boggs et al. (2000) reported

parallel trends from their study on soil nitrogen mineralisation by crop rotation and N

fertilisation.

Figure 4.14 Cumulative net N mineralisation trend

135

Repeated measures analysis for cumulative net mineral N under the different fallow

systems over the four sites revealed significant differences (P<0.05) among the fallows

for the Salani and the Aopo sites with mucuna treated plots exhibiting a superseding

effect over all the other fallow types. This can be due to the high biomass production and

a low C:N ratio of 11:1 of mucuna. However, no statistical significance was found

between the fallow systems for the Siufaga and the Safaatoa sites (Table 4.6a). This may

be linked to higher immobilisation rates and slower decomposition due to the higher clay

content of the soils from these sites.


cumulative net mineral - N (mg/kg) under different fallow systems across

all time points for the four sites

Fallow Site


Grass 1416 b 1246 a 1249 a 1505 c

Mucuna 2390 a 1131 a 1460 a 2183 a

Erythina 1458 b 1156 a 1805 a 1888 b

Biochar 1494 b 1096 a - -

F.Pr 0.009 0.225 0.141 0.012

s.e.d 164 150 165 158

Repeated measures split plot in time analyses showed that the accumulation in plant

available N pools, from the embedded covered core in-situ aerobic incubation procedure,

were highly significant (P<0.001) over time points across all the fallow systems for the

four sites. Larger inputs to the mineralised pool occurred prior to the qusi-equilibrium

phase after which only minor additions followed. Fallow-time interactions were also

only significant for the Salani (P=0.005) and Aopo (P<0.001) sites (Table 4.6b). Nested

classification repeated measures analyses for fallow within a time point also showed

highly significant effects (P<0.001) of fallows on the soil mineral nitrogen pool for the

136

Salani (P=0.001) and Aopo (P<0.001) sites (Table 4.6c). The fallow-time and fallow

within a time point interactions at Safaatoa and Siufaga sites being not significant can be

attributed to a large degree of variability in the data over time and between fallow

systems as reported by Kolberg et al. (1997;1999).

Table 4.6(b) Table of predicted estimates for fallow x time interaction from repeated

measures split plot in time analysis for cumulative net mineral - N

(mg/kg) across all fallow systems for the four sites

Fallow x Time Site

Salani Safaatoa Siufaga Aopo F.Pr 0.005 0.998 0.234 <0.001 s.e.d 280 197 237 210

Table 4.6(c) Within time point comparison among fallow types for cumulative net

mineral - N (mg/kg) for each site

Time (fallow) Site

Salani Safaatoa Siufaga Aopo F.Pr 0.001 0.981 0.218 <0.001 s.e.d 280 197 237 210

Nested classification analysis of variance for between sites revealed highly significant

spatial variations (P<0.001) between the sites with regards to mean levels of cumulative

net mineral N pools (Table 4.6a and Table 4.6d). The Siufaga site had the highest

biomass production comparatively. For the Aopo site, the comparative higher mineral N

levels soil can reasonably be linked to the higher levels of organic carbon. Fallows

within site comparisons revealed significant fallow effects (P<0.001) on mean levels of

biologically active N pools with mucuna treated plots being comparatively higher than

all the other fallow types for Salani and Aopo sites (Table 4.6a). Azam et al. (1995)

reported analogous findings on mineralisation of N from leguminous residues.

137

The system is so nitrate-saturated that adopting the practice of mucuna fallow

technology could pose environmental problems. For instance, too much nitrate-N could

lead to excessive leaching potentially polluting the groundwater leading to stream

eutrophication.

Table 4.6(d) Between site comparisons for cumulative net mineral N (mg/kg).

Site Predicted mean for cumulative net mineral N

Salani 1690 ab

Safaatoa 1157 c

Siufaga 1504 b

Aopo 1859 a

LSD (5%) 257

4.1.2.5 Associations between the evaluated biochemical soil parameters

Correlation analyses were carried out to determine significant associations between the

soil biochemical parameters evaluated. Linear associations were determined using

Pearson’s product-moment correlation coefficients between variates. Table 4.7 outlines the

details of associations between the various indicators evaluated during the study.

Highly significant associations (P<0.001) were observed between the various evaluated

indicators. Positive associations were observed between the magnitude of soil labile C

pool and the biological activity that it supported. Statistical significance was also found

to exist with regards to positive associations between soil labile C and the nitrogen

availability indices (PMN, NH4+ - N and NO3

- - N)

(Table 4.7). Significant negative associations were found between biological activity and

mineral nitrogen indices. This can be credited to the decline in plant parasitic nematode

population and activities upon additions and subsequent mineralisation of the organic

matter from the fallow cover crops. Potter et al. (1988) and McBride et al. (2000)

138

reported similar nematicidal effects of organic matter incorporations, hypothesising that

low molecular organic acids were responsible for the effect. Another probable

mechanism of suppression of plant parasitic nematodes is the accumulation of toxic

nitrogenous compounds, particularly ammonia (Oka, 2010).

Table 4.7 Pearson’s product-moment correlation analyses between the biochemical

indicators.

Variable X1 Variable X2 N r - value p - value

Labile C FDA 616 0.2376 <0.001

Labile C PMN 616 0.4606 <0.001

FDA PMN 616 -0.3103 <0.001

Labile C NH4+ - N 448 0.4692 <0.001

Labile C NO3- - N 448 0.4400 <0.001

FDA NH4+ - N 448 -0.2035 <0.001

FDA NO3- - N 448 -0.2767 <0.001

PMN NH4+ - N 448 0.5613 <0.001

PMN NO3- - N 448 0.5505 <0.001

NH4+ - N NO3

- - N 448 0.5800 <0.001

4.1.3 Nematode community analysis

4.1.3.1 Salani site

There was a mean overall decline in the total nematode counts for the Salani site,

following the decomposition of all the cover crop biomass residues, however, under

erythrina-treated plots a slight increase was observed. The breakdown distribution of this

decline across the various trophic guilds revealed significant reductions (P<0.05) in the

population of plant parasitic nematodes under grass and mucuna fallows, while

reduction in fungal feeding nematode population was only significant under erythina

fallow. Bacterial feeding nematode population significantly declined under all the fallow

139

strategies except erythrina. Decomposition of all the fallows resulted in an increase in

the population of predatory and omnivores nematodes, however, no statistically

significant differences were found (Table 4.8 a and b). Possible mechanisms involved in

nematode suppression as reported by Oka (2010), are: (1) release of pre-existing

nematicidal compounds in soil amendments, (2) generation of nematicidal compounds,

such as ammonia and fatty acids during degradation, (3) enhancement and/or

introduction antagonistic microorganisms, (4) increase in plant tolerance and resistance,

and (5) changes in soil physiology that are unsuitable for nematode behaviour.

Combinations of these mechanisms, rather than a single one, appeared to produce

nematode suppression in organically amended soils. Barros et al., (2014) reported on the

toxicological effects of plant organic volatile compounds on plant parasitic soil

nematode mobility, pathogenicity and reproduction.

Food web evaluation

The effect of management strategies on nematode faunal indices revealed a significant

increase (P<0.05) in the enrichment index (E.I.) following decomposition of the cover

crops across all the fallow treatments, however, it was not significant under mucuna.

This showed a significant overall nutrient enrichment across all the fallow treatments

except mucuna, where the nutrient enriched pool supported a relatively larger microbial

population since the biomass production was the highest. Such increases in E.I. have

been reported by Neher et al. (2005). In addition, the nutrient enriched pool under

mucuna fallow supported a higher biomass and comparatively higher yields of the taro

crop.

Initially, prior to the establishment of the fallow cover crops, the food webs were highly

enriched (Table 4.8a). The six month duration of cover cropping saw a marked decline

in soil enrichment, particularly due to nutrient immobilisation into cover crop

vegetation. However, following decomposition of the residues, the E.I. increased to near

initial levels, though no statistical differences were found between fallows. External

organic matter inputs in the form of cover crops increased energy availability to the soil

140

microbes, thereby enhancing microbial activity. However, microbivorous organisms

such as nematodes feed on the microbes and mineralisation of nutrients occurs, which

may otherwise be immobilised in the body of the microbes. E.I. provides an indication

of the response of primary decomposers or enrichment opportunists towards the labile

sources of organic materials.

The analysis of food web indices reveal significant increases (P<0.05) in the structure

index (S.I.) across all the fallow treatments following decomposition (Table 4.8a & b).

This represents an aggregation of functional guilds and denotes a soil ecosystem with

greater trophic links. Since the cover crop residues were allowed to decompose as

mulches without any tillage or physical disturbance, the decomposition process can

reasonably be attributed to these increases of trophic links in the soil food web. S.I.

value provides information about the levels of trophic links indicated by the abundance

of high coloniser-persister value nematodes mainly omnivores and predatory nematodes.

The increase indicates a better structured food web with more number of highly active

trophic guilds. Comparable results were reported by Briar (2007) in a comparison

between cropland and forest soil web analysis as well as with composted and non-

composted additions of organic matter. Lower S.I. values were observed when

composted organic residue was added as opposed to fresh organic materials.

The decline in the channel index (CI) following decomposition exhibited a shift in the

decomposition channel from a predominantly fungal pathway towards a bacterial

decomposition pathway, however, this was not significant. This showed that both the

channels of decomposition were important for the Salani site. The only exception was

for the mucuna treated plots, where the channel index significantly increased (P<0.05),

indicating a stronger, if not absolute, fungal decomposition channel. This can again be

attributed to relatively higher organic matter production with high moisture retention.

Ferris et al. (2001) reported parallel findings on their work on nematode faunal analysis

and diagnostic concepts.

141

The analysis of carbon flow proportions revealed a significant decline (P<0.05) in the

detritus index across all the fallow systems, denoting a decline in the population and

subsequent activity fungal and bacterial feeding nematodes. This scenario mostly occurs

towards the end of the decomposition period.

A significant increase (P<0.05) in the predation index across all the cover crop fallows

indicates an increase in the population and activity of predatory and omnivorous

nematodes. A decline in the root index, though not significant, indicates a decline in the

population and activity of plant parasitic nematodes. Tsifouli et al. (1997) from the

analysis of trophic structure, reported that there was a gradual reduction of plant feeders

from conventional to the older organic cultivation, while the exact opposite trend was

revealed in the case of nematodes that feed on decomposers, i.e. bacteria and fungi. Van

Diepeningen et al. (2006) hypothesised that the differences between the organic and

inorganic management types are more gradual than black and white, and are in

agreement with the findings of several authors (e.g. Neher & Olson, 1999; García-

Álvarez et al., 2004), who reported an increase of decomposer feeders and especially

bacterivores under organic cultivation.

The general objective of suppressing plant parasitic nematodes and enhancing the build

up of free living nematode population through organic management strategies have been

met to varying degree with selected organic mulches. Bulluck et al. (2002) reported

analogous results from their study on the influences of organic and synthetic soil fertility

amendments on nematode trophic groups and community dynamics under tomatoes.

142

Tabl

e 4.

8(a)

N

emat

ode

enum

erat

ion,

cla

ssifi

catio

n an

d an

alys

is o

f prin

cipa

l com

pone

nts a

nd in

dice

s at d

iffer

ent t

imes

of

the

fallo

w e

xper

imen

t for

Sal

ani s

ite.

Tim

e of

m

easu

rem

ent

Fallo

w

type

Nem

atod

e co

unt/1

00g

soil

Food

web

indi

ces

Car

bon

flow

pro

porti

ons

Tota

l Pl

ant

para

sitic

Free

livi

ng

Enric

hmen

t in

dex

Stru

ctur

e in

dex

Cha

nnel

in

dex

Det

ritus

in

dex

Pred

atio

n in

dex

Roo

t in

dex

Fung

al

feed

ing

Bac

teria

l fe

edin

g Pr

edat

ory

Om

nivo

res

Initi

al

- 11

62

237

264

365

23

273

59.9

67

.9

40.7

44

.2

45.6

10

.2

Bef

ore

plan

ting

(Afte

r 6 m

onth

fa

llow

per

iod)

Gra

ss

1088

49

6 21

0 26

3 12

10

7 39

.1 a

b 54

.7

82.5

40

.8

27.6

31

.6

Muc

una

1488

82

8 16

6 27

8 15

20

1 42

.9 a

68

.0

64.4

30

.9

40.7

28

.4

Eryt

hrin

a 92

3 11

8 16

5 53

0 0

110

30.7

b

55.5

77

.9

54.1

33

.7

12.2

Bio

char

10

41

263

192

442

4 14

0 33

.8 a

b 58

.8

78.6

48

.0

35.7

16

.3

LSD

(5%

) 92

8n.s.

861n.

s. 11

7n.s.

395n.

s. 22

n.s.

94n.

s. 11

.26*

23

.28

n.s.

21.6

4 n.

s. 27

.6 n.

s. 23

.20

n.s.

23.0

4 n.

s.

F.Pr

(5%

) 0.

567

0.32

5 0.

791

0.39

7 0.

428

0.15

3 0.

034

0.57

3 0.

313

0.32

0 0.

656

0.23

9

CV

(%)

51.1

12

6.2

40.0

65

.4

179.

7 42

.0

19.2

24

.6

17.8

39

.7

42.1

65

.1

Afte

r tar

o ha

rves

t

Gra

ss

462

116

126

a 64

12

14

4 55

.1

78.7

b

66.2

ab

22.1

55

.2

22.7

Muc

una

537

74

134

a 33

16

27

9 46

.9

81.3

ab

86.3

a

16.4

70

.2

13.4

Eryt

hrin

a 54

8 13

6 54

b

100

27

231

62.2

90

.4 a

39

.6 b

17

.9

67.8

14

.2

Bio

char

52

5 12

8 11

1 a

77

15

194

55.4

82

.0 a

b 63

.2 a

b 21

.0

61.9

17

.1

LSD

(5%

) 17

7 n.

s. 12

6 n.

s. 39

.9*

68 n

.s.

30 n

.s.

143 n

.s.

19.1

3 n.s.

9.

97*

37.7

3*

13.0

6 n.s.

18

.26 n

.s.

21.5

4 n.s.

F.Pr

(5%

) 0.

705

0.70

1 0.

006

0.23

5 0.

683

0.24

8 0.

404

0.01

1 0.

014

0.75

0 0.

313

0.76

2

CV

(%)

21.4

69

.2

23.5

62

.2

105.

2 42

.2

21.8

7.

5 37

.0

42.2

17

.9

79.9

* M

ean

com

paris

ons w

ithin

a c

olum

n ar

e si

gnifi

cant

at 5

% le

vel.

n.s. M

ean

com

paris

ons w

ithin

a c

olum

ns a

re n

ot si

gnifi

cant

143

Tabl

e 4.

8(b)

D

iffer

ence

s in

the

nem

atod

e co

unt,

dist

ribut

ion

and

indi

ces i

ndic

atin

g sh

ift in

act

ivity

acr

oss t

he v

ario

us

troph

ic g

uild

s bet

wee

n pr

e an

d po

st d

ecom

posi

tion

of fa

llow

cov

er c

rop

resi

dues

for S

alan

i site

.

Tim

e of

m

easu

rem

ent

Fallo

w

type

Nem

atod

e co

unt/1

00g

soil

Food

web

indi

ces

Car

bon

flow

pro

porti

ons

Tota

l Pl

ant

para

sitic

Free

livi

ng

Enric

hmen

t in

dex

Stru

ctur

e in

dex

Cha

nnel

in

dex

Det

ritus

in

dex

Pred

atio

n in

dex

Roo

t in

dex

Fung

al

feed

ing

Bac

teria

l fe

edin

g Pr

edat

ory

Om

nivo

res

Diff

eren

ces i

n m

eans

(B

efor

e –

Afte

r)

Gra

ss

-626

* -3

80*

-84 n

.s.

-199

* 0 n

.s.

+37 n

.s.

+16.

0*

+24.

0*

-16.

3 n.s.

-1

8.7*

+2

7.6*

-8

.9 n

.s.

Muc

una

- 951

* -7

54*

-32 n

.s.

-245

* +1

n.s.

+7

8 n.s.

+4

.0 n

.s.

+13.

3*

+21.

9*

-14.

4*

+29.

5*

-15.

0 n.s.

Eryt

hrin

a -3

75 n

.s.

+18 n

.s.

-111

* -4

00 n

.s.

+27*

+1

21*

+31.

5*

+34.

9*

-38.

3 n.s.

-3

6.2*

+3

4.1*

+2

.0 n

.s.

Bio

char

-5

16*

-135

n.s.

-8

1 n.s.

-3

65*

+11 n

.s.

+54 n

.s.

+21.

6*

+23.

2*

-15.

4 n.s.

-2

7.0*

+3

4.1*

-0

.8 n

.s.

LSD

(5%

) for

co

mpa

rison

be

twee

n be

fore

and

af

ter

deco

mpo

sitio

n

Gra

ss

446

376

88

79

0 95

. 8.

0 11

.1

18.0

15

.3

21.1

24

.9

Muc

una

785

707

84

95

21

109

14.9

9.

6 21

.1

8.3

15.6

15

.3

Eryt

hrin

a 45

6 91

10

7 42

8 20

95

19

.2

23.8

44

.2

28.1

20

.5

14.0

Bio

char

39

8 20

5 13

7 23

3 12

64

11

.9

16.5

28

.5

16.7

15

.1

14.9

-/+ N

egat

ive

valu

es in

dica

te a

dec

reas

e fo

llow

ing

deco

mpo

sitio

n w

here

as p

ositi

ve v

alue

s in

dica

te a

n in

crea

se f

ollo

win

g th

e

deco

mpo

sitio

n of

the

cove

r cro

ps

* S

igni

fican

t diff

eren

ces i

n m

eans

at 5

% le

vel.

n.

s. Not

sign

ifica

nt

144

4.1.3.2 Safaatoa site

For the Safaatoa site, highly contrasting results were obtained with regards to total

nematode counts with increases observed under the cover crop fallows, however, not

significant. The only exception was under the mucuna fallow which recorded a

significant decline in the total nematode counts (Table 4.9a). The distribution of changes

in the nematode population across the various trophic guilds revealed a significant

decline in the plant parasitic nematode population under mucuna treated plots;

significant increases in fungal feeding nematodes across all the fallow treatments except

erythrina; a significant decline in bacterial feeding nematodes under the mucuna fallow;

and significant increases in omnivorous nematode populations under mucuna and

erythrina fallows. The decline in the population of plant parasitic nematodes and

subsequently total nematode population counts under mucuna fallow can be attributed to

comparatively higher biomass production by the leguminous fallows leading to a greater

accumulation of nitrogenous compounds, particularly ammonia. Biochar fallow only

contributed to a significant increase in the population of fungal feeding nematodes with

no statistical significance for changes under other trophic guilds.

Food web evaluation

Effects of management strategies on nematode faunal indices revealed a significant

increase in the enrichment index only under erythrina fallow and a non-significant

decline under mucuna fallow. This can be attributed to the significant increase in the

fungal feeding nematodes under both the fallow systems. However, larger increase in the

population of omnivorous nematodes under erythrina fallow ensured greater

mineralisation and release of much of the immobilised nutrients by the soil microbes,

leading to a significant nutrient enrichment of the soil.

There was no statistical significance observed with regards to changes in the structure

index under all the fallow systems following decomposition of the fallow residues. A

145

moderately structured food web was the resultant effect; however, better structured than

the pre-experiment (initial) levels.

The channel index significantly increased under the grass and the mucuna fallow

practices indicating a significant shift towards a predominant fungal decomposition

channel of the fallow litter. For the mucuna fallow, high biomass production and

moisture retention could be the reason for this shift while for the grass fallow, the

relatively high C:N ratio of the decomposing material could be responsible for this shift

(Table 4.9 a and b).

Analysis of carbon flow proportions revealed a significant decline in the detritus index

under mucuna and erythrina fallows denoting a decline in the combined population and

subsequent activity of the fungal and bacterial feeding nematodes. A significant increase

in the predation index was also observed under mucuna and grass fallows, indicating an

increase in the population and activity of predatory and omnivorous nematodes. A

significant decline in the root index was only observed under the mucuna fallow

suggesting a decline in parasitism. Tsiafouli et al. (2007) reported comparable findings

with regards to changing management regime from conventional to organic cultivation

causing profound modifications in generic structure and composition of the nematode

community.

The interactive responses observed with regards to changes in the food web indices

support the idea that mucuna and erythrina fallowing practices leads to suppression of

plant parasitic nematodes while enhancing the population of free living nematodes.

146

Tabl

e 4.

9 (a

) N

emat

ode

enum

erat

ion,

cla

ssifi

catio

n an

d an

alys

is o

f prin

cipa

l com

pone

nts a

nd in

dice

s at d

iffer

ent t

imes

of

the

fallo

w e

xper

imen

t for

Saf

aato

a si

te.

Tim

e of

m

easu

rem

ent

Fallo

w

type

Nem

atod

e co

unt/1

00g

soil

Food

web

indi

ces

Car

bon

flow

pro

porti

ons

Tota

l Pl

ant

para

sitic

Free

livi

ng

Enric

hmen

t in

dex

Stru

ctur

e in

dex

Cha

nnel

in

dex

Det

ritus

in

dex

Pred

atio

n in

dex

Roo

t in

dex

Fung

al

feed

ing

Bac

teria

l fe

edin

g Pr

edat

ory

Om

nivo

res

Initi

al

- 84

8 45

2 12

5 24

7 10

14

52

.4

39.7

58

.6

48.1

9.

8 42

.1

Bef

ore

plan

ting

(Afte

r 6 m

onth

fa

llow

per

iod)

Gra

ss

935

301

133

350

11

140

46.2

65

.4

43.9

b

45.6

35

.7

18.7

Muc

una

957

414

109

311

26

97

52.7

68

.9

42.2

b

44.5

29

.3

26.2

Eryt

hrin

a 78

8 36

4 92

25

6 0

76

34.2

61

.1

64.2

a

44.7

26

.3

29.0

Bio

char

82

6 27

7 11

8 28

7 7

137

40.6

68

.1

58.2

ab

43.3

36

.5

20.2

LSD

(5%

) 33

3

n.s.

302 n

.s.

63 n

.s.

114 n

.s.

29 n

.s.

72 n

.s.

15.2

5 n.s.

20

.81 n

.s.

16.6

4*

14.1

8 n.s.

14

.63 n

.s.

25.0

6 n.s.

F.Pr

(5%

) 0.

619

0.73

8 0.

539

0.35

1 0.

283

0.19

0 0.

106

0.83

1 0.

039

0.98

6 0.

372

0.76

5

CV

(%)

23.8

55

.7

34.7

23

.6

166.

5 39

.8

22.0

19

.7

20.0

19

.9

28.6

66

.5

Afte

r tar

o ha

rves

t

Gra

ss

1162

26

7 31

9 32

6 16

23

4 49

.0

66.0

70

.6

34.1

45

.9

20.0

Muc

una

613

84

217

143

34

135

46.8

66

.9

80.1

36

.1

54.4

9.

5

Eryt

hrin

a 11

19

316

226

235

3 33

9 51

.1

77.1

55

.2

24.3

47

.8

27.9

Bio

char

12

05

231

317

345

2 31

0 50

.0

70.1

60

.8

34.4

49

.1

16.5

LSD

(5%

) 63

3 n.s.

35

3 n.s.

18

3 n.s.

24

5 n.s.

48

n.s.

19

8 n.s.

8.

85 n

.s.

21.5

7 n.s.

22

.66 n

.s.

18.3

7 n.s.

18

.53 n

.s.

19.4

7 n.s.

F.Pr

(5%

) 0.

192

0.51

5 0.

458

0.28

8 0.

322

0.16

1 0.

727

0.65

9 0.

136

0.49

5 0.

761

0.26

4

CV

(%)

38.6

98

.1

42.4

58

.5

221.

1 48

.5

11.2

19

.3

21.3

35

.6

23.5

65

.9

* M

ean

com

paris

ons w

ithin

a c

olum

n ar

e si

gnifi

cant

at 5

% l

evel

. n.s.

Mea

n co

mpa

rison

s with

in a

col

umns

are

not

sign

ifica

nt

147

Tabl

e 4.

9(b)

D

iffer

ence

s in

the

nem

atod

e co

unt,

dist

ribut

ion

and

indi

ces i

ndic

atin

g sh

ift in

act

ivity

acr

oss t

he v

ario

us

troph

ic g

uild

s bet

wee

n pr

e an

d po

st d

ecom

posi

tion

of fa

llow

cov

er c

rop

resi

dues

for S

afaa

toa

site

Tim

e of

m

easu

rem

ent

Fallo

w

type

Nem

atod

e co

unt/1

00g

soil

Food

web

indi

ces

Car

bon

flow

pro

porti

ons

Tota

l Pl

ant

para

sitic

Free

livi

ng

Enric

hmen

t in

dex

Stru

ctur

e in

dex

Cha

nnel

in

dex

Det

ritus

in

dex

Pred

atio

n in

dex

Roo

t in

dex

Fung

al

feed

ing

Bac

teria

l fe

edin

g Pr

edat

ory

Om

nivo

res

Diff

eren

ce

(Bef

ore

–Afte

r)

Gra

ss

+227

n.s .

-3

4 n.s.

+1

86*

-24 n

.s.

+5 n

.s.

+94 n

.s.

+2.8

n.s.

+0

.6 n

.s.

+26.

7*

-11.

5 n.s.

+1

0.2 n

.s.

+1.3

n.s.

Muc

una

-344

* -3

30*

+108

* -1

68*

+8 n.

s. +3

8*

-5.9

n.s.

-2.0

n.s.

+37.

9*

-8.4

* +2

5.1*

-1

6.7*

Eryt

hrin

a +3

31 n.

s. -4

8 n.

s. +1

34n.

s -2

1 n.

s. +3

n.s.

+263

* +1

6.9*

+1

6.0

n.s

-9.0

n.s

-20.

4*

+21.

5*

-1.1

n.s

Bio

char

+3

79 n

.s -4

6 n

.s +1

99*

+58

n.s

-5 n

.s +1

73 n

.s +9

.4 n

.s +2

.0 n

.s +2

.6 n

.s -8

.9 n

.s +1

2.6

n.s

-3.7

n.s

LSD

(5%

) for

co

mpa

rison

be

twee

n be

fore

an

d af

ter

deco

mpo

sitio

n

Gra

ss

686

251

184

237

21

132

11.2

18

.3

12.3

11

.9

15.4

11

.5

Muc

una

147

127

71

80

58

23

8.2

15

19.6

16

.8

9.9

9.1

Eryt

hrin

a 45

0 38

6 17

3 10

0 4.

5 19

6 14

.5

18.0

29

.6

14.4

18

.9

29.7

Bio

char

71

7 17

9 19

0 30

5 8

197

7.7

9.4

23.4

11

.0

14.6

13

.8

-/+ N

egat

ive

valu

es in

dica

te a

dec

reas

e fo

llow

ing

deco

mpo

sitio

n w

here

as p

ositi

ve v

alue

s in

dica

te a

n in

crea

se f

ollo

win

g th

e

deco

mpo

sitio

n of

the

cove

r cro

ps

* Si

gnifi

cant

diff

eren

ces i

n m

eans

at 5

% le

vel.

n.s. N

ot si

gnifi

cant

148

4.1.4 Cover crop dry matter yields, nutrient concentrations and nutrient uptake over

the four sites

The dry matter yields (t/ha) of all the fallow cover crops together with their nutrient

concentration and uptake is given in Table 4.10 (a) and (b). The dry matter production of

the mucuna cover crop significantly out yielded all the other fallow cover crops across

all the four sites. This can be ascribed to the inherent growth characteristic of the

mucuna cover as well as its ability to fix atmospheric nitrogen biologically. Lal (2013)

also reported similar findings from a trial in Taveuni, Fiji. Goh and Chin (2007) and

Ngome et al. (2011) concluded that mucuna fallow crop fixed 70% of atmospheric N

through biological symbiosis while the remaining was thought to have been taken up

from the soil. However, Chikowo et al. (2004) stated that mucuna fallow fixed 96% of

the accumulated nitrogen. Sanginga et al. (2001) found that 91% of the total N was

fixed by mucuna cover crop.

The nutrient concentration analyses of the cover crop revealed significant differences

between fallow crops across all the four sites. Plant tissue analyses revealed that the

nitrogen and the phosphorus content of erythrina was significantly higher than the

mucuna cover across all the four sites. However, since the biomass production of

erythrina was much lower, total uptake of these two nutrients was significantly much

lower than the mucuna cover (Table 4.10b). Since, the nutrient uptake of mucuna across

all the sites were very high (196-700 kg N/ha), it is rational to assume that at taro

harvesting time (eight months after spraying the fallow crop), almost all the N contained

in the mucuna biomass has been mineralised. This can be attributed to the low C:N ratio

of 11:1 for mucuna. Parallel findings were reported by Ibewiro et al. (2000) showing

that mucuna decomposition can be quite fast, losing 60% of its biomass within the first

28 days of decomposition while releasing up to 174 kg N/ha during that time period. The

C:N ratios for erythrina, grass and biochar were 20:1, 25:1 and 70:1, respectively. These

comparatively higher C:N ratios coupled with significantly lower biomass production

than mucuna can be linked to the much lower N inputs under these fallow systems.

149

Prominently, biochar supported vegetation resulted in significantly higher concentrations

of K than all the fallow practices at the two biochar treated sites on the island of Upolu.

Significant notable concentrations of Mg and all the micronutrients (Fe, Mn, Cu and Zn)

were also observed for the biochar supported vegetation in the high rainfall zone only.

For the Savaii sites, erythrina had the higher K content.

Generally, nutrient uptake was significantly higher under mucuna fallow systems across

all the sites, owing to the higher biomass production, comparatively.

150

Tabl

e 4.

10(a

) D

ry m

atte

r yie

lds a

nd n

utrie

nt c

once

ntra

tions

of f

allo

w tr

eatm

ents

gro

wn

over

the

four

site

s

Site

Fa

llow

trea

tmen

t D

ry m

atte

r yi

eld

(t/ha

)

Nut

rient

con

tent

M

acro

nutri

ent (

%)

Mic

ronu

trien

t (m

g/kg

) N

P

K

Ca

Mg

Fe

Mn

Cu

Zn

Sala

ni

Gra

ss

8.21

b

0.96

d

0.15

b

1.28

d

0.31

c

0.57

b

7777

b

345

b 19

c

48 b

M

ucun

a 22

.11

a 2.

20 b

0.

16 b

1.

38 c

0.

77 b

0.

34 c

63

54 c

23

4 c

26 b

33

d

Eryt

hrin

a 8.

13 b

2.

42 a

0.

32 a

1.

87 b

0.

82 a

0.

57 b

41

7 d

125

d 13

d

42 c

B

ioch

ar

6.50

b

1.15

c

0.17

b

2.19

a

0.26

d

0.80

a

1486

2 a

479

a 30

a

92 a

LS

D (5

%)

6.98

0.

038

0.02

6 0.

04

0.01

5 0.

032

119.

92

5.61

1.

53

2.07

Safa

atoa

Gra

ss

7.62

b

1.18

c

0.19

b

1.09

c

0.66

c

0.45

b

5024

b

207

b 20

c

74 a

M

ucun

a 17

.85

a 2.

29 b

0.

17 b

1.

21 b

1.

23 b

0.

40 c

71

24 a

24

3 a

29 a

36

c

Eryt

hrin

a 8.

13 b

3.

16 a

0.

30 a

1.

21 b

1.

40 a

0.

61 a

17

0 d

164

c 22

b

40 b

B

ioch

ar

10.8

7 b

1.18

c

0.16

b

2.12

a

0.39

d

0.26

d

2838

c

163

c 21

bc

37 c

LS

D (5

%)

5.84

0.

068

0.03

3 0.

042

0.05

0.

04

103.

65

5.84

1.

85

2.39

Siuf

aga

Gra

ss

7.03

b

1.02

c

0.15

c

0.16

c

0.56

c

0.57

a

5757

a

194

a 16

b

42 a

M

ucun

a 35

.92

a 1.

95 b

0.

30 b

1.

41 b

1.

03 b

0.

39 b

16

82 b

14

1 b

26 a

33

c

Eryt

hrin

a 4.

49 c

3.

47 a

0.

34 a

2.

51 a

1.

99 a

0.

33 c

35

2 c

69 c

10

c

39 b

LS

D (5

%)

11.5

5 0.

16

0.03

0.

04

0.01

0.

04

238.

9 5.

00

3.9

1.6

Aop

o

Gra

ss

4.36

b

1.24

c

0.17

b

2.11

b

0.90

b

0.66

a

489

a 13

b

15 b

50

a

Muc

una

9.62

a

2.04

b

0.18

b

1.77

c

0.93

b

0.27

c

195

b 19

a

18 a

30

c

Eryt

hrin

a 3.

57 b

2.

51 a

0.

36 a

2.

53 a

1.

08 a

0.

38 b

16

0 c

21 a

19

a

37 b

LS

D (5

%)

4.49

0.

18

0.04

0.

12

0.10

0.

07

14.4

3 3.

5 2.

34

3.86

LSD

(5%

) for

be

twee

n si

te

com

paris

on

Gra

ss

3.32

0.

07

0.02

0.

07

0.04

0.

05

191

5.89

1.

59

4.00

M

ucun

a 11

.03

0.12

0.

03

0.07

0.

05

0.03

12

9.5

4.23

3.

44

1.89

Er

ythr

ina

2.04

0.

15

0.03

0.

05

0.06

0.

04

38.2

2 5.

16

1.80

2.

82

Bio

char

7.

35

0.04

0.

04

0.04

0.

03

0.02

15

6.6

0.27

1.

33

2.84

151

Tabl

e 4.

10(b

) D

ry m

atte

r yie

lds a

nd n

utrie

nt u

ptak

e by

the

fallo

w tr

eatm

ents

ove

r the

four

site

s

Site

Fa

llow

trea

tmen

t D

ry m

atte

r yi

eld

(t/ha

)

Nut

rient

upt

ake

Mac

ronu

trien

t (kg

/ha)

M

icro

nutri

ent (

kg/h

a)

N

P K

C

a M

g Fe

M

n C

u Zn

Sala

ni

Gra

ss

8.21

b

78.7

7 c

12.3

1 b

105.

02 b

25

.44

bc

46.7

7 b

63.8

1 b

2.83

b

0.15

b

0.40

b

Muc

una

22.1

1 a

486.

31 a

35

.37

a 30

5.05

a

170.

21 a

75

.16

a 14

0.45

a

5.18

a

0.57

a

0.73

a

Eryt

hrin

a 8.

13 b

19

6.75

b

26.0

2 a

152.

03 b

66

.67

b 46

.34

b 3.

39 c

1.

02 c

0.

10 b

0.

34 b

B

ioch

ar

6.50

b

74.7

5 c

11.0

5 b

14

2.35

b

16.9

0 c

52.0

0 b

96.6

0 b

3.11

b

0.19

b

0.60

ab

LSD

(5%

) 6.

98

117.

4 13

.47

84.9

43

.60

22.0

7 41

.54

1.43

0.

19

0.29

Safa

atoa

Gra

ss

7.62

b

89.9

5 c

14.4

8 b

83.0

9 b

50.3

1 bc

34

.30

b 38

.30

b 1.

58 b

0.

15 b

0.

56 a

M

ucun

a 17

.85

a 40

8.65

a

30.3

4 a

215.

92 a

21

9.49

a

71.3

8 a

127.

13 a

4.

34 a

0.

51 a

0.

65 a

Er

ythr

ina

8.13

b

256.

91 b

24

.39

ab

98.3

7 b

113.

82 b

49

.59

ab

1.38

b

1.33

b

0.18

b

0.33

b

Bio

char

10

.87

b 33

.21

c 17

.38

ab

230.

34 a

42

.37

c 28

.25

b 30

.84

b 1.

77 b

0.

22 b

0.

40 a

LS

D (5

%)

5.84

13

0.5

13.1

6 72

.8

64.3

26

.06

38.1

3 1.

40

0.19

0.

27

Siuf

aga

Gra

ss

7.03

b

71.7

1 b

10.5

5 b

81.5

5 b

39.3

7 b

40.0

7 b

40.4

7 b

1.36

b

0.11

b

0.30

b

Muc

una

35.9

2 a

700.

49 a

12

2.14

a

506.

51 a

37

0.00

a

140.

10 a

60

.44

a 5.

07 a

0.

94 a

1.

19 a

Er

ythr

ina

4.49

c

155.

80 b

15

.27

b 11

2.70

b

89.3

5 b

14.8

2 c

1.58

c

0.31

b

0.05

b

0.18

b

LSD

(5%

) 11

.55

241.

0 41

.10

169.

9 10

6.4

54.8

2 26

.80

1.79

0.

22

0.4

Aop

o

Gra

ss

4.36

b

54.1

0 b

7.42

a

92.0

5 a

39.2

6 b

28.7

9 a

2.13

a

0.06

b

0.07

b

0.22

ab

Muc

una

9.62

a

196.

15 a

17

.31

a 17

0.19

a

89.4

2 a

25.9

6 a

1.87

a

0.18

a

0.18

a

0.29

a

Eryt

hrin

a 3.

57 b

89

.54

b 12

.84

a 90

.26

a 38

.53

b 13

.56

a 0.

57 b

0.

07 a

b 0.

07 b

0.

13 b

LS

D (5

%)

4.49

10

0.5

10.5

8 91

.1

39.5

17

.08

1.06

0.

118

0.1

0.15

LSD

(5%

) for

be

twee

n si

te

com

paris

on

Gra

ss

3.32

32

.70

5.89

45

.09

17.2

5 19

.16

21.2

9 0.

85

0.06

0.

17

Muc

una

11.0

3 24

5.4

34.1

6 15

9.1

107.

5 46

.14

50.1

3 2.

08

0.27

0.

36

Eryt

hrin

a 2.

04

66.8

6.

57

46.1

4 34

.66

8.50

0.

62

0.22

0.

03

0.08

B

ioch

ar

7.35

25

.77

16.6

9 86

.9

31.9

0 20

.79

34.2

6 0.

05

0.24

0.

46

152

4.1.5 Taro yields

4.1.5.1 Fresh corm yields

The taro crop was grown over all the four sites following a six month fallow duration.

However, harvest data could only be ascertained from three of the sites due to theft of

the crop from the Aopo site. The taro corm yield data from the three remaining sites are

shown in Figure 4.15.

The yield data from all the three sites was subjected to a nested classification analysis of

variance, where blocks were nested into locations; the six fallow practices were nested

into blocks and the two cultivars were nested into fallows. The differences in the

predicted mean taro yields between sites was found to be highly significant (P<0.001),

with the Salani site out yielding the other two sites (Table 4.11a). This can be attributed

to the relatively higher amount of annual rainfall received by the Salani site (4,959 mm)

as opposed to the Safaatoa (3,418 mm) and Siufaga (2,989 mm) sites.

Table 4.11 (a) Predicted mean taro yields (t/ha) for individual sites

Site Predicted mean taro yield (t/ha)

Salani 10.76 a

Safaatoa 8.53 b

Siufaga 7.77 c

LSD (5%) 0.75

Significant difference was also found (P<0.001) between the predicted mean yields of

two cultivars within the same fallow treatment, with cultivar Samoa 2 out yielding

cultivar Samoa 1 (Table 4.11b).

153

Table 4.11 (b) Predicted mean taro yields (t/ha) for the two cultivars

Cultivar Predicted mean taro yield (t/ha)

Samoa 1 8.14 b

Samoa 2 10.00 a

LSD (5%) 1.5

Fallows within a site were also highly significant (P<0.001) with regards to the mean

corm yield of taro produced over the three sites (Table 4.11c).

Table 4.11(c) Predicted mean corm yield of taro (t/ha) under different fallows within

sites.

Fallow Site

Salani Safaatoa Siufaga

Grass 8.6 d 6.9 c 5.6 c

Mucuna 11.7 b 7.6 c 9.1 ab

Erythina 9.7 a 8.8 abc 6.5 c

Mucuna + 200kg/ha NPK 13.6 a 10.2 a 10.1 a

Grass + 400 kg/ha NPK 10.5 bc 9.8 ab 7.6 bc

Biochar 10.5 bc 7.5 c -

(LSD 5%) 2.5 1.8 1.5

The six month cover crop fallow practice with mucuna together with modest application

(200 kg/ha) of complete fertiliser (NPK 12-5-20) to the taro crop, that is a corresponding

supplementation of 24 kg N/ha, 10 kg P/ha and 40 kg/K/ha, resulted in significantly

higher (P<0.001) mean yields at the Salani site, out yielding all the fallow practices.

Taro grown under the mucuna with no fertiliser supplementation, grass fallow with the

recommended 400kg/ha complete fertiliser supplementation, (that is, 48 kg N/ha, 20 kg

P/ha and 80 kg/K/ha) and the biochar fallow treatments did not differ significantly from

154

each other; however, they significantly out yielded the traditional grass fallow at the

Salani site. This indicated that reasonable taro yields can be obtained under mucuna

fallows with no supplementation as opposed to grass fallows with 400 kg/ha of complete

fertiliser supplementation and biochar additions during fallow periods increases taro

yields of the succeeding crop for the Salani site. The yield differences can be attributed

to the possible P effect, as the native Olsen P values of the soils are well below the

critical range, making it the most limiting nutrient.

The fallow practice of mucuna together with modest application (200 kg/ha) of complete

fertiliser (NPK 12-5-20) and grass with 400 kg/ha complete fertiliser supplementation

showed no statistical significance with regards to the mean yields of taro for the

Safaatoa site. Only mucuna, erythrina and biochar treatments were not significant from

the traditional grass fallow for the Safaatoa site. This denotes that maximum taro yields

at this particular site are largely dictated by fertiliser inputs.

For the Siufaga site, mucuna with 200 kg/ha of complete fertiliser supplementation did

not significantly increase the taro yields as compared to mucuna with no

supplementation. Mucuna with no supplementation did not significantly differ from

grass fallow with 400 kg/ha complete fertiliser supplementation. This signifies that

optimum taro yields are possible by switching from the traditional grass fallowing to

mucuna fallows.

The increased mean taro yields obtained under mucuna fallowing systems can be

attributed to the greater biomass production and subsequently greater nutrient uptake by

the vegetation cover. In addition, the phenomenon of biological fixation of atmospheric

nitrogen by the legume cover crop can also be credited to the comparatively higher mean

yields obtained under these fallows. These results are in agreement with the work of

other authors involved with cover cropping trials including leguminous and non-

155

leguminous cover crops (Clarke et al., 1994; Vaughn and Evanylo, 1988; Kuo and

Jellum, 2002).

The general yield trend under different fallow practices across all the sites indicates that

mucuna with modest supplementation of the taro crop with complete fertilisers can help

maintain optimum yields. However, it appears that the yield responses of the taro crop to

fallow treatments are site-specific. In Safaatoa, the yield increase relative to control was

only 10%. In Salani, it was 36% and in Siufaga it was 62%. Moreover, more yields can

be obtained if a positive change to the traditional grass fallow is made by opting for the

economically best site specific improved fallow alternative. Mucuna fallow systems

appear to increase yields of Samoan taro soils that generally have medium levels of

organic C and total N; however, are low in plant available P and exchangeable K. Sakala

et al. (2003) reported significant increase in maize yields under short term mucuna

fallow compared to natural fallow. Carsky et al. (1998) attributed the improved yield of

the succeeding crop under mucuna fallow to increased soil moisture retention and

improved fertility.

156

7.2

9.9

10.1

13.2

9.6

9.9

12.4

14.8

10.0

11.0

9.0

11.9

5.7

8.0

6.1

9.1

7.5

10.1

9.1

11.4

9.2

10.5

7.4

7.6

5.5

5.7

7.4

10.7

6.3

6.8

7.9

12.3

7.9

7.3

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Samoa 1

Samoa 2

Gra

ssM

ucun

aEr

ythr

ina

Muc

una

+½

NPK

Gra

ss +

NPK

Bio

char

Gra

ssM

ucun

aEr

ythr

ina

Muc

una

+½

NPK

Gra

ss +

NPK

Bio

char

Gra

ssM

ucun

aEr

ythr

ina

Muc

una

+½

NPK

Gra

ss +

NPK

Sala

niSa

faat

oaSi

ufag

a

Taro corm yield (t/ha)

Site

LSD

(5%

): 2.

6

Cul

tiva

r Fallo

w

Figu

re 4

.15A

ctua

l cor

m y

ield

s of t

he tw

o cu

ltiva

rs o

f tar

o fr

om th

e th

ree

site

s

157

4.1.5.2 Associations between fresh corm yield and mean levels of the evaluated

biochemical parameters

Correlation analyses were carried out to determine significant associations between the

various soil biochemical parameters and the fresh taro corm yields of the two cultivars.

Linear associations were determined using Pearson’s product-moment correlation

coefficients between variates. Table 4.12 outlines the details of associations.

There was no statistical significance with regards to associations between the fresh corm yields

of the two cultivars of taro and the mean levels of soil labile C across the experimental time

duration. The mean levels of all the other biochemical indicators significantly correlated with the

fresh corm yield of cultivar Samoa 2 but not with cultivar Samoa 1 (Table 4.12). This can be

attributed to the genotypic differences with cultivar Samoa 2 exhibiting very vigorous vegetative

growth, comparatively.

Table 4.12 Pearson’s product-moment correlation analyses between the evaluated

biochemical indicators and the fresh taro corm yields of the two cultivars

Variable X1 Variable X2 Cultivar N r-value p-value

Yield Labile C Samoa 1 68 -0.1405 0.2532

Samoa 2 68 0.0617 0.6174

Yield FDA Samoa 1 68 0.1130 0.3590

Samoa 2 68 0.2840 0.0189*

Yield PMN Samoa 1 68 0.1082 0.3800

Samoa 2 68 0.2646 0.0292*

Yield NH4+ - N Samoa 1 68 0.1109 0.3678

Samoa 2 68 0.2554 0.0356*

Yield NO3- - N Samoa 1 68 0.0980 0.4265

Samoa 2 68 0.2580 0.0337*

Yield Cumulative

mineral N

Samoa 1 68 0.1620 0.1869

Samoa 2 68 0.2516 0.0385*

*Significant association at P=0.05.

158

4.1.6 Corm nutrient uptake by the two cultivars produced under different fallow

systems over the three sites

The mean corm dry matter yields (t/ha) and the macronutrient uptake (kg/ha) of the two

cultivars produced under the different fallow systems over the three sites are presented

in Table 4.13 (a-c).

4.1.6.1 Salani site

The mean corm dry matter yields produced under the various fallow practices showed

that mucuna with 200 kg/ha of complete fertiliser out yielded all the other fallow

treatments. Corm nitrogen uptake was significantly higher under mucuna fallows, either

with or without fertiliser supplementation. Corm P levels significantly increased under

all the improved fallow practices compared to the traditional grass fallow while mucuna

with NPK supplementation resulted into significantly higher K uptake. Ca uptake was

significantly higher under mucuna fallows, either with or without fertiliser

supplementation, while Mg uptake was highest under mucuna with the taro crop being

supplemented with NPK.

Significant differences were found between the two taro cultivars for N, P and K uptake

at the Salani site. Cultivar Samoa 2, showed higher mean nutrient uptake and mean corm

dry matter yields than Samoa 1. However, Ca and Mg uptake did not differ significantly

between cultivars. Jacobs and Clarke (1993), from an earlier experiment in Samoa,

suggested that N availability in the soil is one of the determinants of biomass production

in taro and the management of N nutrition for corm production may need to take account

of changes in partitioning of dry matter and N that occur under varying supplies of N.

4.1.6.2 Safaatoa site

For the Safaatoa site, significantly higher mean corm dry matter yields were observed

under the fallows practices where the taro crop was supplemented by complete fertiliser.

Mean corm nutrient uptake was highest under mucuna with fertiliser supplementation

159

treatment, however, not significantly from the grass fallow with taro supplemented by

complete fertiliser. Erythrina showed comparable levels of uptake of N and P, but not

for the other macronutrients.

Significant differences were found between the two cultivars for all the macro nutrients

except Mg. Cultivar Samoa 2, showed higher mean nutrient uptake and mean corm dry

matter yields than Samoa 1.

4.1.6.3 Siufaga site

For the Siufaga site, comparatively higher mean corm dry matter yields were recorded

for taro produced under mucuna fallows. Mean corm nutrient uptake were also higher

under both the mucuna fallows, however, K was significantly lower when the taro crop

was not supplemented with complete fertiliser.

Cultivars differed significantly for the mean uptake of all the macronutrients except Mg.

Cultivar Samoa 2, with higher nutrient uptake, was considered to have better nutrient

utilisation efficiency than Samoa 1.

160

Tabl

e 4.

13 (a

) D

ry m

atte

r yie

lds a

nd m

acro

nutri

ent u

ptak

e of

the

two

culti

vars

of t

aro

corm

s pro

duce

d un

der t

he

diff

eren

t fal

low

pra

ctic

es a

t Sal

ani s

ite.

Fallo

w

Cul

tivar

Fallo

w x

Cul

tivar

mea

ns

Fallo

w m

eans

Cor

m d

ry

mat

ter y

ield

(t/

ha)

Cor

m n

utrie

nt u

ptak

e (k

g/ha

) C

orm

dry

m

atte

r yi

eld

(t/ha

)

Cor

m n

utrie

nt u

ptak

e (k

g/ha

)

N

P K

C

a M

g N

P

K

Ca

Mg

Gra

ss

Sam

oa I

2.97

29

.1

5.0

15.6

10

.8

4.2

3.53

c

38.0

c

6.0

b 18

.3 c

12

.1 b

4.

9 b

Sam

oa II

4.

08

47.0

7.

1 20

.9

13.4

5.

6

Muc

una

Sam

oa I

4.02

48

.8

7.6

28.4

16

.5

6.2

4.55

b

61.2

a

9.2

a 32

.8 b

17

.7 a

6.

3 b

Sam

oa II

5.

07

73.7

10

.8

37.1

18

.8

6.5

Eryt

hrin

a Sa

moa

I 4.

06

39.0

7.

8 25

.3

14.2

6.

3 4.

09 b

c 41

.6 b

c 7.

7 ab

26

.2 b

c 13

.9 b

5.

8 b

Sam

oa II

4.

11

44.1

7.

7 27

.2

13.7

5.

3 M

ucun

a +

200k

g/ha

N

PK

Sam

oa I

5.20

54

.2

7.6

29.4

18

.9

8.1

5.81

a

65.1

a

9.3

a 41

.9 a

19

.8 a

8.

2 a

Sam

oa II

6.

43

76.0

11

.1

54.4

20

.8

8.4

Gra

ss +

400

kg/h

a N

PK

Sam

oa I

4.09

38

.5

6.4

26.2

14

.2

5.5

4.30

bc

45.1

bc

7.2

ab

29.8

b

14.5

b

5.6

b Sa

moa

II

4.51

51

.8

8.1

33.4

14

.8

5.6

Bio

char

Sa

moa

I 3.

74

34.5

7.

0 20

.6

13.6

5.

7 4.

28 b

c 45

.8 b

c 7.

9 ab

26

.4 b

c 14

.6 b

6.

1 b

Sam

oa II

4.

82

57.1

8.

7 32

.1

15.6

6.

4

Cul

tivar

mea

ns

Sam

oa I

4.01

b

40.7

b

6.9

b 24

.2 b

14

.7 a

6.

0 a

Sa

moa

II

4.84

a

58.3

a

8.9

a 34

.2 a

16

.2 a

6.

3 a

LSD

(5%

) C

ultiv

ar

0.37

4.

90

0.93

4.

40

1.67

0.

60

Fa

llow

0.

86

13.2

5 2.

85

8.42

3.

83

1.54

LSD

(5%

) C

x F

Sa

me

leve

ls

0.91

12

.01

2.28

10

.77

4.09

1.

46

D

iff. l

evel

s 1.

06

15.2

3 3.

18

10.9

2 4.

63

1.79

161

Tabl

e 4.

13 (b

) D

ry m

atte

r yie

lds a

nd m

acro

nutri

ent u

ptak

e of

the

two

culti

vars

of t

aro

corm

s pro

duce

d un

der t

he

diff

eren

t fal

low

pra

ctic

es a

t Saf

aato

a si

te.

Fallo

w

Cul

tivar

Fallo

w x

Cul

tivar

mea

ns

Fallo

w m

eans

Cor

m d

ry

mat

ter

yiel

d (t/

ha)

Cor

m n

utrie

nt u

ptak

e (k

g/ha

) C

orm

dry

m

atte

r yi

eld

(t/ha

)

Cor

m n

utrie

nt u

ptak

e (k

g/ha

)

N

P K

C

a M

g N

P

K

Ca

Mg

Gra

ss

Sam

oa I

2.24

25

.0

2.6

10.5

8.

5 3.

1 2.

99 c

35

.0 c

4.

6 c

14.5

c

13.0

b

4.1

c Sa

moa

II

3.74

45

.0

6.6

18.4

17

.5

5.2

Muc

una

Sam

oa I

2.55

30

.3

4.3

14.7

10

.3

3.6

3.21

bc

41.7

bc

5.5

bc

18.1

bc

15.9

b

4.5

bc

Sam

oa II

3.

88

53.2

6.

7 21

.5

21.4

5.

4

Eryt

hrin

a Sa

moa

I 3.

43

33.4

4.

5 18

.8

13.6

5.

0 3.

80 b

44

.2 a

b 6.

1 ab

21

.4 b

15

.9 b

5.

3 b

Sam

oa II

4.

18

55.1

7.

6 24

.0

18.1

5.

6 M

ucun

a +

200k

g/ha

NPK

Sa

moa

I 3.

97

40.6

4.

7 22

.6

16.5

5.

7 4.

68 a

51

.2 a

7.

1 a

29.9

a

24.2

a

6.4

a Sa

moa

II

5.38

61

.7

9.4

37.3

31

.8

7.1

Gra

ss +

40

0kg/

ha N

PK

Sam

oa I

4.49

41

.3

5.7

24.3

17

.5

6.6

4.49

ab

45.8

ab

6.6

ab

27.0

ab

22.8

a

6.2

ab

Sam

oa II

4.

48

50.2

7.

5 29

.7

28.0

5.

8

Bio

char

Sa

moa

I 3.

19

28.8

4.

8 17

.8

14.6

4.

6 3.

31 b

c 33

.8 c

5.

5 bc

20

.0 b

c 16

.2 b

4.

7 bc

Sa

moa

II

3.43

38

.7

6.3

22.3

17

.7

4.7

Cul

tivar

mea

ns

Sam

oa I

3.31

b

33.2

0 b

4.4

b 18

.1 b

13

.5 b

4.

8 a

Sa

moa

II

4.18

a

50.7

0 a

7.4

a 25

.5 a

22

.4 a

5.

6 a

LSD

(5%

) C

ultiv

ar

0.53

3 6.

52

0.79

3.

94

2.62

0.

84

Fa

llow

0.

71

8.80

1.

26

6.00

4.

51

0.98

LSD

(5%

) C

x

F

Sam

e le

vels

1.

31

15.9

6 1.

94

9.65

6.

41

2.06

Diff

. le

vels

1.

12

13.8

0 1.

79

8.75

6.

15

1.70

162

Tabl

e 4.

13 (c

) D

ry m

atte

r yie

lds a

nd m

acro

nutri

ent u

ptak

e of

the

two

culti

vars

of t

aro

corm

s pro

duce

d un

der t

he

diff

eren

t fal

low

pra

ctic

es a

t Siu

faga

site

.

Fallo

w

Cul

tivar

Fallo

w x

Cul

tivar

mea

ns

Fallo

w m

eans

Cor

m d

ry

mat

ter

yiel

d (t/

ha)

Cor

m n

utrie

nt u

ptak

e (k

g/ha

) C

orm

dry

m

atte

r yi

eld

(t/ha

)

Cor

m n

utrie

nt u

ptak

e (k

g/ha

)

N

P K

C

a M

g N

P

K

Ca

Mg

Gra

ss

Sam

oa I

1.83

24

.8

3.7

10.9

10

.0

2.9

2.04

d

29.0

b

3.8

b 12

.1 c

10

.2 b

2.

9 b

Sam

oa II

2.

25

33.2

4.

0 13

.2

10.5

2.

9

Muc

una

Sam

oa I

2.64

36

.0

4.6

18.8

11

.7

4.0

3.43

ab

51.8

a

6.7

a 24

.2 b

14

.2 a

4.

6 ab

Sa

moa

II

4.22

67

.7

8.8

29.7

16

.8

5.2

Eryt

hrin

a Sa

moa

I 2.

42

27.4

4.

2 14

.9

12.0

3.

6 2.

48 c

d 29

.2 b

4.

3 b

15.9

c

11.0

b

3.3

b Sa

moa

II

2.54

31

.0

4.5

16.8

10

.0

3.0

Muc

una

+ 20

0kg/

ha N

PK

Sam

oa I

2.74

38

.4

4.9

25.3

12

.7

4.2

3.63

a

54.8

a

7.3

a 35

.5 a

15

.4 a

4.

9 a

Sam

oa II

4.

52

71.1

9.

7 45

.6

18.2

5.

6 G

rass

+ 4

00kg

/ha

NPK

Sa

moa

I 2.

85

33.3

4.

9 23

.6

13.3

4.

1 2.

87 b

c 35

.2 b

5.

0 b

25.2

b

12.7

ab

3.8

b Sa

moa

II

2.89

37

.2

5.2

26.7

12

.4

3.5

Cul

tivar

mea

ns

Sam

oa I

2.50

b

32.0

b

4.5

b 18

.7 b

12

.0 b

3.

8 a

Sa

moa

II

3.28

a

48.0

a

6.4

a 26

.4 a

13

.5 a

4.

0 a

LSD

(5%

) C

ultiv

ar

0.28

5.

18

0.83

2.

74

1.49

0.

43

Fa

llow

0.

58

9.54

1.

28

6.29

3.

35

0.99

LSD

(5%

) C

x V

Sa

me

leve

ls

0.63

11

.59

1.85

6.

13

3.33

0.

96

D

iff. l

evel

s 0.

70

11.9

8 1.

74

7.32

3.

92

1.15

163

4.1.6.4 Associations between corm dry matter (CDM) and mean levels of macro

nutrient uptake

Correlation analyses were carried out to determine significant associations between

corm dry matter and mean levels of various macro nutrient uptakes. Strengths of

associations between the different nutrients for their uptake were determined. Linear

associations were determined using Pearson’s product-moment correlation coefficients

between variates. Highly significant associations (P<0.001) were observed between all

the tests (Table 4.14).

Table 4.14 Pearson’s product-moment correlation analyses between CDM and

macronutrients.

Variable X1 Variable X2 N r-value p-value

Corm dry matter N 34 0.8585 <0.001 Corm dry matter P 34 0.8999 <0.001

Corm dry matter K 34 0.8579 <0.001

Corm dry matter Ca 34 0.7429 <0.001

Corm dry matter Mg 34 0.9580 <0.001

N P 34 0.9294 <0.001

N K 34 0.8816 <0.001

N Ca 34 0.6779 <0.001

N Mg 34 0.7447 <0.001

P K 34 0.8860 <0.001

P Ca 34 0.6640 <0.001

P Mg 34 0.8101 <0.001

K Ca 34 0.6239 <0.001

K Mg 34 0.7557 <0.001

Ca Mg 34 0.7001 <0.001

164

4.1.7 Marginal economic analysis of taro grown under the mucuna fallow versus the

traditional grass fallow

The comparative gross margin analysis of taro grown under the improved mucuna

fallow and the traditional grass fallow is given in Table 4.15. Succeeding taro crop

grown after the six month fallow period showed 98% higher gross margin under the

mucuna fallow as opposed to the traditional grass fallow for the Salani site. For the

Safaatoa site, the corresponding increase was 48%. The largest increase was noted for

the Siufaga site where the gross margin of taro grown under mucuna fallow systems was

21 folds higher than that produced under the traditional grass fallowing system.

The gross margin differences observed between the sites can be credited to the yield

differences of taro which is a function of the interactive effects of the following factors:

� The mean rainfall received over the taro life cycle

� The amount of biomass production by the fallow covers

� Nutrient uptake of the cover crops

� The limiting nutrient for individual sites

� The inherent soil properties of the sites

Eilitta et al., (2004) reported extremely low profitability without the use of mucuna in a

comparative cost-benefit study on maize in Mexico.

165

Tabl

e 4.

15

Mar

gina

l eco

nom

ic a

naly

sis o

f the

muc

una

fallo

w te

chno

logy

ver

sus t

he c

urre

nt fa

rmer

’s fa

llow

pra

ctic

e w

ithou

t

the

use

of c

hem

ical

ferti

liser

s

Parti

cula

rs

Uni

t Pr

ice

$SA

T

Sala

ni si

te

Safa

atoa

Site

Si

ufag

a Si

te

Muc

una

Fallo

w

Gra

ss fa

llow

M

ucun

a Fa

llow

G

rass

fallo

w

Muc

una

Fallo

w

Gra

ss fa

llow

Qua

nti

ty

Cos

t $S

AT

Qua

ntity

C

ost

$SA

T Q

uant

ity

Cos

t $S

AT

Qua

nti

ty

Cos

t $S

AT

Qua

ntity

C

ost

$SA

T Q

uant

ity

C

ost

$SA

T

Inco

me

Taro

yie

ld (t

/ha)

11.7

8.6

7.

6

6.9

9.

1

5.6

G

ross

inco

me

($)

$2,1

00/

mt

$2

4,57

0

$18,

06 0

$15,

960

$1

4,49

0

$19,

110

$1

1,76

0

Expe

nditu

re

Fallo

w e

stab

lishm

ent:

-Her

bici

de &

labo

ur

272

1 27

2

1

272

1 27

2

Taro

pla

ntin

g:

- Spr

ayin

g (f

allo

w)

- Pla

ntin

g m

ater

ial

(suc

kers

) - L

abou

r (m

an-d

ay)

15

5 1 20

1 10

,000

12

15

5 10

,000

24

0

1 10

,000

12

15

5 10

,000

24

0

1 10

,000

12

15

5 10

,000

24

0

1 10

,000

12

15

5 10

,000

24

0

1 10

,000

12

15

5 10

,000

24

0

1 10

,000

12

15

5 10

,000

24

0

Mai

nten

ance

: - H

erbi

cide

/Lab

our

275

1

275

2

550

1

275

2

550

1

275

2

550

Har

vest

ing

/labo

ur/c

arta

ge

500

1 50

0 1

500

1 50

0 1

500

1 50

0 1

500

Tota

l cos

ts ($

/ha)

11,4

42

11

,445

11,4

42

11

,445

11,4

42

11

,445

Gro

ss m

argi

n ($

SAT/

ha)

13

,128

6,61

5

4,51

8

3,04

5

7,66

8

315

166

4.2 Experiment 2 The soil incubation pot trial

4.2.1 Labile carbon measurement

The general mean labile C trend for both the soils incubated with the different organic

materials at various rates in pots without plants under screen house conditions showed

similar trends (Figure 4.16a and b). Significant increases from initial low levels were

observed for both the soils after 30 days of incubation. The overall mean effects of soil

and cover crop mulches were highly significant (P<0.001), however, the application rate

was not (P=0.979) (Table 4.16). By the end of the 90-day incubation period, the mean

levels of active C under all the organic mulches was still on an increasing trend.

Repeated measures analysis of variance test for second order ante-dependence

(correlation model structure employed to test for changes in treatment effects between

particular time points and the treatment effects combined over all the time points)

revealed highly significant interactive effect (P<0.001) of soil and fallow for active C

with the mucuna mulch adding higher mean levels to the labile pool on Salani soil

compared to all the other mulches on both the soils. The response of the mean soil labile

C pool to the interactive effects of the fallow mulches and their rates of application on

was also found to be highly significant (P<0.001), with mucuna and erythrina fallow

mulches having higher mean C mineralisation than the grass and the biochar fallows.

This can be attributed to the legume mulches of mucuna and erythrina having a

comparatively lower C:N ratio and consequently having a faster rate of mineralisation

than grass and the highly recalcitrant biochar fallows. Mucuna was only significant from

erythrina at the highest rate of application (45t/ha).

No statistical significance was found (P=0.983) for the interaction between the different

soils and the application rates of the mulches on the changes to the mean soil labile C

pool. This denotes that both the soils responded identically to the different rates of

various organic applications during the pot incubations. The interaction between the

167

soils, fallow cover crop mulches and the rates of application also showed no statistical

significance (P=0.106) (Table 4.16).

Table 4.16 Repeated measures analysis of parameters for soil labile C

Fixed term F-statistic Fpr. s.e.d

Soil 21.92 <0.001 5.213

Fallow 61.43 <0.001 7.372

Rate 0.02 0.979 6.384

Soil.fallow 10.25 <0.001 10.43

Soil.rate 0.02 0.983 9.029

Fallow.rate 10.4 <0.001 12.77

Soil.fallow.rate 1.81 0.106 18.06

168

0

200

400

600

800

1000

1200

1400

1600

day 030 days60 days90 days












15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Sala

ni

Laabile C (mg/kg)

Figu

re 4

.16

(a)

Labi

le c

arbo

n dy

nam

ics

for

Sala

ni s

oil i

ncub

ated

with

diff

eren

t org

anic

mul

ches

at d

iffer

ent r

ates

in

pots

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

169

0

200

400

600

800

1000

1200

1400

1600













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Safa

atoa

Labile C (mg/kg)

Figu

re 4

.16

(b)

Labi

le c

arbo

n dy

nam

ics

for S

afaa

toa

soil

incu

bate

d w

ith d

iffer

ent o

rgan

ic m

ulch

es a

t diff

eren

t rat

es in

pots

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

170

4.2.2 Fluorescein diacetate (FDA) hydrolysis activity measurement

The general mean trend for FDA hydrolysis activity (biological activity) for both the

soils incubated with the different organic materials at various rates in pots without plants

under screen house conditions showed similar trends (Figure 4.17a and b). Significant

increases from initial low levels were observed for both the soils after 30 days of

incubation. The individual mean effects of soils, cover crop mulches as well as their

application rates were all highly significant (P<0.001) with regards to the mean levels of

biological activity during pot incubations.

Repeated measures analysis of variance test for second order ante dependence revealed

highly significant interactions (P<0.001) between levels of all the three factors with

regards to the mean levels of soil biological functioning (Table 4.17).

Table 4.17 Repeated measures analysis parameters for soil biological activity (FDA

hydrolysis activity)


Soil 793170.27 <0.001 0.01897

Fallow 121606.92 <0.001 0.02683

Rate 54595.56 <0.001 0.02324

Soil.fallow 212012.38 <0.001 0.03795

Soil.rate 130720.93 <0.001 0.03287

Fallow.rate 167242.75 <0.001 0.04648

Soil.fallow.rate 110817.17 <0.001 0.06573

171

Grass and erythrina cover crops showed highest levels of mean biological activity at

their highest rate of application (45t/ha) on both the soils. Contrastingly, mucuna and

biochar covers showed highest levels of mean biological activity at the lowest rate of

application (15t/ha). This can be attributed to the difference in the rate of decomposition

of the fallow cover crops. Erythrina decomposed much faster over the 90-day incubation

period than the mucuna, thereby increasing the overall soil biological activity via

quicker nutrient release. Mucuna, conversely, followed a gradual pattern of

decomposition and consequently contributed to a slower release of nutrients. At higher

rates of application decomposition rate of mucuna further reduced, possibly owing to its

higher lignin content, comparatively (Odhiambo, 2010; Palm and Sanchez, 1991).

The significant decreasing trend in mean biological activity with increasing rates of

biochar applications on both the soils can be attributed to the increase in surface

compaction by the fine biochar material, imposing severe limitations on adequate

aeration to the underlying soil in the pots. This scenario may have been exacerbated by

watering to keep the soil’s moisture content to near field capacity.

On average, Safaatoa soil was more biologically active than Salani soils across all the

fallow covers except mucuna during the pot incubations under screen house conditions.

172

020406080100

120













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Sala

ni

mg FDA hydrolysed/kg soil/hr

Figu

re 4

.17

(a)

Fluo

resc

ein

diac

etat

e hy

drol

ysis

act

ivity

dyn

amic

s fo

r Sa

lani

soi

l in

cuba

ted

with

diff

eren

t or

gani

c

mul

ches

at d

iffer

ent r

ates

in p

ots w

ithou

t pla

nts u

nder

scre

en h

ouse

con

ditio

ns

173

020406080100

120













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Safa

atoa

mg FDA hydrolysed/kg soil/hr

Figu

re 4

.17

(b)

Fluo

resc

ein

diac

etat

e hy

drol

ysis

act

ivity

dyn

amic

s fo

r Sa

faat

oa s

oil

incu

bate

d w

ith d

iffer

ent

orga

nic

mul

ches

at d

iffer

ent r

ates

in p

ots w

ithou

t pla

nts u

nder

scre

en h

ouse

con

ditio

ns

174

4.2.3 Potentially mineralisable nitrogen (PMN)

The general mean trend for potentially mineralisable nitrogen (PMN) for both the soils

incubated with the different organic materials at various rates in pots without plants



incubation. The individual mean effects of soils, cover crop mulches as well as their

application rates were all highly significant (P<0.001) with regards to the mean levels of

PMN during pot incubations. Analogous scenarios have been elaborated on by several

other authors (Odhiambo, 2010; Neely et al., 1991; Hunt, 1977).


highly significant interactions (P<0.001) between levels of all the three factors with

regards to the mean levels of mineralisable pool of soil N (Table 4.18).

Table 4.18 Repeated measures analysis parameters for potentially mineralisable N


Soil 57722.53 <0.001 0.01355

Fallow 14205.37 <0.001 0.01916

Rate 10627.18 <0.001 0.01659

Soil.fallow 14276.45 <0.001 0.02709

Soil.rate 21018.51 <0.001 0.02346

Fallow.rate 31152.29 <0.001 0.03318

Soil.fallow.rate 32214.66 <0.001 0.04692

175

Erythrina fallow made the highest mean contribution to the mineralisable N pool for

both the soils over the 90 day incubation period. The application rate of 30 t/ha of dry

matter equivalent of the fallow crops made the mean highest additions to the

mineralisable N pools for both the soils across all the fallow covers. Higher mean PMN

levels were recorded for the Salani soil, owing to its significantly higher levels of active

C.

All the fallow mulches contributed significantly higher mean mineralisable N levels to

the soil pools for the Salani soil comparatively, except for the grass fallow which had a

higher significant mean effect for the Safaatoa soil. Grass and mucuna fallows elevated

the mean mineralisable N levels at the lowest rate of application (15 t/ha) while mucuna

contributed more N at 30 t/ha of application. Biochar treated soils significantly increased

the mineralisable N inputs with increasing rates, owing to its predominant negative

charges affinity for ammonium ions. All the other fallows showed a significant decline

in mineralisable N at higher rates. This can be ascribed to the net effect of an increase in

the microbial population and N immobilisation. Increases in PMN pools as a percentage

of initial, over the 90 day incubation period for Salani soil treated with mucuna cover,

were 31%, 87% and 188% for application rates of 15 t/ha, 30 t/ha and 45 t/ha,

respectively. Corresponding increases for Safaatoa soil were 48%, 81% and 134%.

176

050100

150

200

250













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Sala

ni

Potentially mineralisable N (mg/kg)

Figu

re 4

.18

(a)

Pote

ntia

lly m

iner

alis

able

N d

ynam

ics

for

Sala

ni s

oil

incu

bate

d w

ith d

iffer

ent

orga

nic

mul

ches

at

diff

eren

t rat

es in

pot

s with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

177

0102030405060708090100













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Safa

atoa

Potentially Mineralisable N (mg/kg)

Figu

re 4

.18

(b)

Pote

ntia

lly m

iner

alis

able

N d

ynam

ics

for

Safa

atoa

soi

l in

cuba

ted

with

diff

eren

t or

gani

c m

ulch

es a

t

diff

eren

t rat

es in

pot

s with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

178

4.2.4 Ammonium nitrogen (NH4+ - N)

The general mean fluxes for ammonium nitrogen (NH4+ - N) for both the soils

incubated with the different organic materials at various rates in pots without plants



incubation except for the grass cover. The individual mean effects of cover crop

mulches as well as their application rates were all highly significant (P<0.001) with

regards to the mean levels of NH4+ - N during pot incubations. Javier and Tabien (2003)

reported parallel findings of N dynamics in soils amended with different organic

fertilisers in pot incubations. However, the difference between the two soils was only

marginally significant.


highly significant interactions (P<0.001) between soil and fallow; and between fallow

and rate with regards to the mean levels of NH4+ - N (Table 4.19).

Table 4.19 Repeated measures analysis parameters for NH4+ - N


Soil 3.99 0.051 1.087

Fallow 82.58 <0.001 1.537

Rate 48.79 <0.001 1.331

Soil.fallow 6.36 <0.001 2.173

Soil.rate 1.49 0.235 1.882

Fallow.rate 9.28 <0.001 2.662


179

The overall trend of release of NH4+ - N by the cover crop mulches illustrates a very

high early release by the erythrina mulch peaking just after 30 days of incubation for the

soils; declining sharply thereafter during subsequent measurement. On the contrary, the

pattern of release of NH4+ - N by the mucuna mulch exhibited an increasing trend over

the successive measurement during the 90-day incubation period. At the end of the 90-

day incubation period, the mean levels of NH4+ - N under all the rates of mucuna

mulches was still on an increasing trend. This shows that the peak release of NH4+ - N

by the mucuna cover is expected to occur at a later stage of decomposition. The early

release of NH4+ - N by the erythina green manure is therefore, considered to be too rapid

for a crop of taro to fully utilise it over its growth cycle. Conversely, the comparative

much slower release of NH4+ - N by the mucuna cover would be much desirable for a

crop of taro which has a mean life cycle of 210 days. The overall mean NH4+ - N

mineralisation was significantly higher under the mucuna treated soils than all the other

fallow mulches. Grass and biochar treated plots had the lowest mean NH4+ - N

mineralisation after the 90 days incubation across both the soils.

The interactions between fallows and soils revealed that while erythrina and mucuna

both had significantly higher mean levels of NH4+ - N mineralisation under Salani soil,

only mucuna was higher under the Safaatoa soil. Fallow and application rate interactions

showed significant mean increase in NH4+-N mineralisation with increasing rates of

mucuna and erythrina; however, no such statistical significance was found for the grass

and biochar treated soils. This can be attributed to the high N contents of the leguminous

cover crop mulches.

Increases in NH4+-N release as a percentage of initial, over the 90 day incubation period

for Salani soil treated with mucuna cover, were 46%, 118% and 213% for application

rates of 15 t/ha, 30 t/ha and 45 t/ha, respectively. Corresponding increases for Safaatoa

soil were 130%, 336% and 485%.

180

050100

150

200

250













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Sala

ni

NH4+ - N (mg/kg)

Figu

re 4

.19

(a)

Am

mon

ium

N fl

uxes

for S

alan

i soi

l inc

ubat

ed w

ith d

iffer

ent o

rgan

ic m

ulch

es a

t diff

eren

t rat

es in

pot

s

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

181

020406080100

120

140













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Safa

atoa

NH4+ - N (mg/kg)

Figu

re 4

.19

(b)

Am

mon

ium

N f

luxe

s fo

r Sa

faat

oa s

oil i

ncub

ated

with

diff

eren

t or

gani

c m

ulch

es a

t di

ffer

ent

rate

s in

pots

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

182

4.2.5 Nitrate nitrogen (NO3- - N)

The general mean fluxes for nitrate nitrogen (NO3- - N) for both the soils incubated with

the different organic materials at various rates in pots without plants under screen house

conditions showed similar trends (Figure 4.20a and b). Significant increases from initial

low levels were observed for both the soils over the 90-day incubation period except for

the biochar cover which contributed to an increase during the first 30 days after

incubation and subsequently declined thereafter. The individual mean effects of cover

crop mulches as well as their application rates were significant (P<0.05) with regards to

the mean levels of NO3- - N during pot incubations. On average, mucuna had the highest

rate of NO3- - N mineralisation. Significant increases in rate of NO3

- - N mineralisation

were observed with increasing rate of application of all the fallow covers on both the

soils. Analogous scenarios have been elaborated on by Odhiambo (2010).

Increases in NO3--N release as a percentage of initial, over the 90 day incubation period

for Salani soil treated with mucuna cover, were 253%, 321% and 584% for application

rates of 15 t/ha, 30 t/ha and 45 t/ha, respectively. Corresponding increases for Safaatoa

soil were 111%, 248% and 468%.


no significant interactions between various levels of the three factors towards the mean

levels of NO3--N mineralisation (Table 4.20).

Table 4.20 Repeated measures analysis parameters for NO3

- - N

Fixed term F-statistic Fpr. s.e.d Soil 0.10 0.757 46.28

Fallow 11.89 <0.001 65.46

Rate 4.20 0.021 56.69

Soil.fallow 0.02 0.996 92.57

Soil.rate 0.12 0.891 80.17

Fallow.rate 1.33 0.261 113.4


183

0

500

1000

1500

2000

2500













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Sala

ni

NO3- - N (mg/kg)

Figu

re 4

.20

(a)

Nitr

ate

N f

luxe

s fo

r Sa

lani

soi

l in

cuba

ted

with

diff

eren

t or

gani

c m

ulch

es a

t di

ffer

ent

rate

s in

pot

s

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

184

0

200

400

600

800

1000

1200

1400

1600

1800

2000













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Safa

atoa

NO3- - N (mg/kg)

Figu

re 4

.20

(b)

Nitr

ate

N f

luxe

s fo

r Sa

faat

oa s

oil i

ncub

ated

with

diff

eren

t or

gani

c m

ulch

es a

t di

ffer

ent

rate

s in

pot

s

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

185

4.2.6 Assay of soil phosphatase activity

The mean dynamics for soil phosphatase activity for both the soils incubated with the

different organic materials at various rates in pots without plants under screen house

conditions showed similar trends (Figure 4.21a and b). Significant increases from initial

low to maximum peak levels were observed across all the fallows for both the soils

during the first 30 days of incubation followed by subsequent declines thereafter. Long-

term experiments have shown that repeated applications of manure increase both acid

and alkaline phosphomonoesterase activities, particularly immediately after manure

addition to the soil (Dick et al., 1988; Colvan et al., 2001), due to the stimulation of

microbial growth. When the monitoring period is prolonged, stimulation of microbial

synthesis of enzymes by easily degradable organic substrates decreases (Garcia et al.,

1993; Nannipieri, 1994).

The individual mean effects of cover crop mulches and the two soils were significant

(P<0.05) with regards to the mean levels of phosphate mineralisation during pot

incubations. However, the application rates of cover crop were not. On average, mucuna

had the highest significant rate of phosphate mineralisation while the Salani soil

expressed greater significant phosphate activity than Safaatoa soil.


significant interactions between the different fallows and their application rates towards

the mean levels of phosphate mineralisation (Table 4.21).

186

Table 4.21 Repeated measures analysis parameters for phosphate mineralisation


Soil 38.62 <0.001 0.5998

Fallow 15.64 <0.001 0.8482

Rate 1.08 0.345 0.7346

Soil.fallow 2.50 0.064 1.200

Soil.rate 0.70 0.500 1.039

Fallow.rate 2.35 0.037 1.469


Significant interactions between the fallow covers and their application rates revealed

that mean levels of phosphate mineralisation and release of plant available P increases

significantly with increasing rates of mucuna application. Conversely, increasing rates of

biochar application resulted in a decrease in phosphate mineralisation. This can be

partially explained by the limited mineralisation of the highly recalcitrant biochar

material. No statistical significance was found in phosphate mineralisation with

increasing rates of grass and erythrina mulches.

187

051015202530354045day 0

30 days60 days90 days












15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Sala

ni

µmol p-nitrophenol released/g soil/hr

Figu

re 4

.21

(a)

Phos

phat

ase

activ

ity fl

uxes

for S

alan

i soi

l inc

ubat

ed w

ith d

iffer

ent o

rgan

ic m

ulch

es a

t diff

eren

t rat

es in

pots

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

188

05101520253035day 0

30 days60 days90 days












15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Safa

atoa

µmol p-nitrophenol released/g soil/hr

Figu

re 4

.21

(b)

Phos

phat

ase

activ

ity fl

uxes

for S

afaa

toa

soil

incu

bate

d w

ith d

iffer

ent o

rgan

ic m

ulch

es a

t diff

eren

t rat

es

in p

ots w

ithou

t pla

nts u

nder

scre

en h

ouse

con

ditio

ns

189

4.2.7 Assay of soil urease activity

The mean dynamics for soil urease activity for both the soils incubated with the different

organic materials at various rates in pots without plants under screen house conditions

showed similar trends (Figure 4.22a and b). Significant increases from initial low to

maximum peak levels were observed across all the fallows for both the soils during the

90-day incubation period. Incorporation of organic materials into soil promotes

microbial activity and also soil urease activity (Nannipieri et al., 1983; Boltan et al.,

1985; Balasubramanium et al., 1972; Zantua and Bremner, 1976; Al-Rashidi and Al-

Jabri, 1990). Many researchers have reported that urease activity in soils is positively

correlated with organic C and total N (Zantua et al. 1977; Spier et al., 1980; Dash et al.,

1981; Reynold et al., 1985; Frankenberger and Dick, 1983), which are indices of organic

matter content. Zantua et al. (1977) suggested that organic matter accounted for most of

the variations in soil urease activity. Further, the constituents of the organic matter also

determine the activity of urease in soils. The individual mean effects of cover crop

mulches, their application rates and the two soils were all highly significant (P<0.001)

with regards to the mean levels of soil urease activity during pot incubations. On

average, mucuna had the highest significant rate of hydrolysis of urea, while the

Safaatoa soil expressed greater significant urease activity than Salani soil across all the

mulch treatments. This can reasonably be linked to the significantly higher microbial

activity (FDA) of the Safaatoa soil. There was an increasing trend of the hydrolysis of

urea with increasing application rates of cover crop mulches. Zantua et al. (1977)

reported urease activity to be positively correlated to the total N in the soil. The

increased levels of urease activity in the organic amended soil has generally been

attributed to the increased microbial biomass although additional evidence suggests that

plant materials may directly contribute enzyme to the soil. The urease activity in the soil

varies depending upon the type and amount of organic matter added (Kumar and

Wagenet 1984; Zantua and Bremner, 1976; Frankenberger, 1983).

190


significant interactions between the different fallows and their effect of the two soils

towards the mean levels of urease activity (Table 4.22). Erythrina mulch showed

significantly lower urease activity than the grass fallow for the Salani soil, however,

significantly exceeded that for the Safaatoa soil.

Mucuna, supporting the highest levels of the hydrolysis reaction of urea, suggests that

plant available ammonium N would be made readily available for uptake.

Table 4.22 Repeated measures analysis parameters for urease activity


Soil 456.70 <0.001 10.71

Fallow 21.08 <0.001 15.15

Rate 82.05 <0.001 13.12

Soil.fallow 4.92 0.004 21.43

Soil.rate 2.60 0.084 18.56

Fallow.rate 1.68 0.143 26.25


191

0

100

200

300

400

500

600

700

800













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Sala

ni

µg NH4+ - N/g soil/2hrs

Figu

re 4

.22

(a)

Ure

ase

activ

ity fl

uxes

for S

alan

i soi

l inc

ubat

ed w

ith d

iffer

ent o

rgan

ic m

ulch

es a

t diff

eren

t rat

es in

pot

s

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

192

0

100

200

300

400

500

600

700

800

900

1000













15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

15 t/

ha30

t/ha

45 t/

ha15

t/ha

30 t/

ha45

t/ha

Gra

ssEr

ythr

ina

Muc

una

Bio

char

Safa

atoa

µg NH4+ - N/g soil/2hrs

Figu

re 4

.22

(b)

Ure

ase

activ

ity f

luxe

s fo

r Sa

faat

oa s

oil i

ncub

ated

with

diff

eren

t org

anic

mul

ches

at d

iffer

ent r

ates

in

pots

with

out p

lant

s und

er sc

reen

hou

se c

ondi

tions

193

4.3 Experiment 3 The taro nutrient uptake and partitioning experiment

4.3.1 Dry matter accumulation by various plant organs

The accumulation of dry matter by various plant organs of the two cultivars is illustrated

in Figure 4.23a-f. Total dry matter did not differ significantly between cultivars

throughout the experimental period (Table 4.23). The first 90 days after planting (DAP)

were characterised by low rates of total dry matter production by both the cultivars

(Figure 4.23a). During this period, leaves and petioles accounted for 58% of the total dry

matter produced in each cultivar (Figure 4.23a-c). Following 210 DAP, the dry matter

content in the leaves and petioles declined to less than 25% of the total dry matter, but it

increased significantly in corms and suckers (Figure 4.23e and f). During the first 90

DAP, roots of cultivars Samoa 1 and Samoa 2 represented about 13% and 18% of the

total dry matter content, respectively. Following 180 DAP, the dry matter content in the

roots was never higher than 8% for Samoa 1 and 12% for Samoa 2. Cultivar Samoa 2

accumulated significantly higher root dry matter than Samoa 1 throughout the

experimental period. It is noteworthy that, between 150 and 240 DAP, the suckers were

a significant sink of dry matter in the taro plant. During this period, these organs

accounted for 22% of the total plant dry matter in Samoa 1 and 13% in Samoa 2. These

results are of particular importance because, when taro is grown under upland

conditions, cormels of suckers seldom reach a marketable size; and they may compete

for assimilates with the marketable main corm. Maximum significant dry matter

accumulation in the corms of both the cultivars was recorded between 210 and 240

DAP, accounting for about 46% of the total plant dry matter.

194

Y =

-0.0

005x

3 +

0.17

54x2

- 11

.962

x +

419.

89

R² =

0.9

39

Y =

-0.0

002x

3 +

0.08

98x2

- 3.

549x

+ 2

14.2

1 R

² = 0

.983

9

020

040

060

080

010

0012

0014

0016

00

030

6090

120

150

180

210

240

Dry matter (kg/ha)

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y

= -7

E-05

x3 + 0

.016

x2 +

0.15

24x

+ 18

.879

R

² = 0

.964

6

Y =

-4E-

05x3 +

0.0

069x

2 + 0

.940

8x +

9.4

5 R

² = 0

.921

1

050100

150

200

250

030

6090

120

150

180

210

240

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y =

-0.0

002x

3 + 0

.059

1x2 -

3.3

942x

+ 1

20.5

4 R

² = 0

.973

9

Y =

-0.0

001x

3 + 0

.039

9x2 -

2.1

891x

+ 9

4.95

R

² = 0

.944

6

050100

150

200

250

300

350

400

030

6090

120

150

180

210

240

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y =

-3E-

05x3 +

0.0

085x

2 - 0

.134

8x +

20.

879

R² =

0.9

171

Y =

-3E-

05x3 +

0.0

082x

2 + 0

.319

5x +

14.

629

R² =

0.9

85

020406080100

120

140

160

030

6090

120

150

180

210

240

Dry matter (kg/ha)

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y =

4E-

06x3 +

0.0

083x

2 + 0

.339

3x +

27.

079

R² =

0.9

909

Y =

-0.0

001x

3 + 0

.053

2x2 -

3.8

726x

+ 1

18.1

6 R

² = 0

.995

6

010

020

030

040

050

060

070

080

0

030

6090

120

150

180

210

240

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y =

-0.0

083x

2 + 5

.571

3x -

530.

54

R² =

0.9

888

Y=

-0.0

106x

2 + 6

.352

6x -

589.

35

R² =

0.9

918

050100

150

200

250

300

350

030

6090

120

150

180

210

240

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

(a) T

otal

(b) L

eave

s

(c

) Pet

iole

(d) R

oots

(e)

Cor

ms

(

f) Su

cker

s Fi

gure

4.2

3 D

ry w

eigh

ts o

f pla

nt o

rgan

s of t

he tw

o ta

ro c

ultiv

ars a

s inf

luen

ced

by a

ge

195

Tabl

e 4.

23

Ana

lysi

s of

var

ianc

e fo

r ef

fect

s of

cul

tivar

and

day

s af

ter

plan

ting

on t

otal

dry

wei

ght

and

plan

t up

take

of

vario

us n

utrie

nts

So

urce

df

M

ean

Squa

res

Tota

l dry

mat

ter

N

P K

C

a M

g Fe

M

n C

u Zn

Blo

ck

4 13

024

634.

9 29

.54

283

1274

.6

13.6

2 8.

70

0.02

3 0.

0014

790

0.00

6684

Cul

tivar

(CV

) 1

4236

81

7756

.7

507.

49*

3320

6 13

42.4

29

7.99

* 0.

68

0.19

0 0.

0011

110

0.14

8525

**

Erro

r (a)

4

6985

7 18

23.3

44

.33

4351

48

2.7

38.8

6 11

.31

0.04

6 0.

0005

035

0.00

7752

Day

s afte

r

plan

ting

(DA

P)

7 18

1628

0***

20

029.

5***

14

15.5

4***

68

008*

**

3111

6.7*

**

618.

67**

* 18

6.70

***

1.00

0***

0.

0079

679*

**

0.21

0027

***

CV

x

DA

P 7

1227

85

1634

.0

183.

46**

* 58

60

1017

.3

105.

35**

* 5.

43

0.06

1 0.

0001

913

0.02

5468

***

Erro

r (b)

56

55

551

816.

5 21

.13

2868

62

8.1

12.9

2 11

.29

0.03

7 0.

0005

920

0.00

4510

*, *

*, *

** S

igni

fican

t at 0

.05,

0.0

1 an

d 0.

001

prob

abili

ty le

vels

, res

pect

ivel

y.

196

4.3.2 Nutrient uptake of the two taro cultivars as influenced by plant age

Except for P, Mg and Zn, there was no statistical significance between the cultivars for

the quantity of nutrients taken up by plants (Table 4.23). In general, the nutrient uptake

was very similar between cultivars during the first 150 DAP; thereafter, the quantity of

all the nutrients taken up by plants of cultivar Samoa 1 was lower than that of cultivar

Samoa 2, however, only significant for P, Mg and Zn. The only exception was for Fe

uptake where uptake by cultivar Samoa 1 was higher than culticar Samoa 2, however,

this was not significant (Figure 4.24a-e and Figure 4.25 a-d). Maximum uptake values

for the two cultivars are given in Table 4.24.

Table 4.24 Maximum levels of nutrient uptake by the two cultivars (kg/ha)

Macronutrient Samoa I Samoa II

N 146 176

P 35 41

K 259 321

Ca 165 183

Mg 20 28

Micronutrient

Fe 21 10

Mn 0.9 1.1

Cu 0.07 0.08

Zn 0.39 0.54

It is noteworthy that cultivar Samoa 1 plants absorbed 20% less K and 17% less N than

those of cultivar Samoa 2 with the uptake uniformly distributed over the entire life cycle

of the crop. These results also confirms that, as with most root crops, taro has a high

requirement for K relative to N (Goenaga and Chardon, 1995; Norman et al., 1994).

Mergedus et al. (2014) also reported analogous findings with the corm being

characterised by high concentrations of K.

197

The third uppermost leaf laminar is often used to determine the nutritional status of aroid

crops including taro (Goenaga and Chardon, 1995). In general, the concentrations of all

the nutrients except Fe in the leaf laminar of cultivar Samoa 1 plants had greater

concentrations than cultivar Samoa 2 plants.

Since there were no significant differences in the total and corm dry matter productions

between the cultivars (Table 4.23), it can rationally be said that cultivar Samoa 1 had a

higher nutrient use efficiency, (kg of edible dry matter produced per kg of nutrient taken

up), for N, P, K, Mg, Mn and Cu over cultivar Samoa 2. However, for Ca, Fe and Zn, it

is logical to consider that cultivar Samoa 2 had a higher nutrient use efficiency over

cultivar Samoa 1 (Figure 4.26 and Figure 4.27). Mergedus et al. (2014) concluded that

the effect of the taro genotype was significant for more than half of the analysed

minerals (i.e., Mg, Ca, Zn, Fe, Mn).

198

Y =

-6E-

05x3

+ 0

.019

6x2

- 1.1

347x

+ 4

4.55

R

² = 0

.918

1

Y =

-0.0

028x

2 +

1.54

41x

- 34.

194

R² =

0.9

227

050100

150

200

030

6090

120

150

180

210

240

Nitrogen uptake (kg/ha)

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y =

-1E-

05x3 +

0.0

056x

2 - 0

.453

4x +

12.

261

R² =

0.9

279

Y =

0.2

239x

- 9.

418

R² =

0.9

605

01020304050

030

6090

120

150

180

210

240

Phosphorus uptake (kg/ha)

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y =

-0.0

001x

3 + 0

.050

9x2 -

3.8

547x

+ 1

12.4

6 R

² = 0

.935

9

Y =

-8E-

05x3 +

0.0

282x

2 - 1

.043

1x +

36.

282

R² =

0.9

103

050100

150

200

250

300

350

030

6090

120

150

180

210

240

Potassium uptake (kg/ha)

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y =

0.0

006x

3 - 0

.093

5x2 +

5.3

173x

- 89

.222

R

² = 0

.865

1

Y =

0.0

006x

3 - 0

.087

4x2 +

4.8

894x

- 80

.471

R

² = 0

.875

1

04080120

160

200

030

6090

120

150

180

210

240

Calcium uptake (kg/ha)

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

Y =

-1E-

05x3 +

0.0

033x

2 - 0

.178

2x +

4.0

309

R² =

0.9

685

Y =

-6E-

06x3 +

0.0

022x

2 - 0

.063

7x +

1.4

56

R² =

0.9

675

051015202530

030

6090

120

150

180

210

240

Magnesium uptake (kg/ha)

Day

s af

ter p

lant

ing

(DA

P)

Sam

oa I

Sam

oa II

(a)

Nitr

ogen

(b) P

hosp

horu

s

(c) P

otas

sium

(d) C

alci

um

(e)

Mag

nesi

um

Figu

re 4

.24

Mac

ronu

trien

t con

tent

s of t

he tw

o ta

ro c

ultiv

ars a

s inf

luen

ced

by p

lant

age

199

Y = -1E-07x4 + 5E-05x3 - 0.0052x2 + 0.1878x R² = 0.8585

Y = -8E-08x4 + 3E-05x3 - 0.0031x2 + 0.1508x - 2.077 R² = 0.935

0

5

10

15

20

25

0 30 60 90 120 150 180 210 240

Iron

upt

ake

(kg/

ha)

Days after planting (DAP)

Samoa I Samoa II

Y = -5E-07x3 + 0.0002x2 - 0.0137x + 0.3784 R² = 0.9513

Y = -4E-07x3 + 0.0002x2 - 0.0128x + 0.3461 R² = 0.9808

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 30 60 90 120 150 180 210 240

Man

gane

se u

ptak

e (k

g/ha

)


Samoa I Samoa II

Y = -3E-06x2 + 0.0011x - 0.0329 R² = 0.7917

Y = -3E-06x2 + 0.0011x - 0.0309 R² = 0.8634

0.00

0.02

0.04

0.06

0.08

0.10

0 30 60 90 120 150 180 210 240

Cop

per u

ptak

e (k

g/ha

)


Samoa I Samoa II

Y = -8E-06x2 + 0.0038x - 0.127 R² = 0.8203

Y= 0.0026x - 0.0561 R² = 0.9753

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 30 60 90 120 150 180 210 240

Zinc

Upt

ake

(kg/

ha)


Samoa I Samoa II

(a) Iron (b) Manganese

(c) Copper (d) Zinc

Figure 4.25 Micronutrient contents of the two taro cultivars as influenced by plant age

200

4.3.3 Nutrient concentration of the two taro cultivars

Table 4.25 Percent nutrient concentration in the lamina of the third uppermost leaf of the two taro cultivars at various stages of growth.

Days after

planting

Cultivar Nutrient Content (%)

N P K Ca Mg Fe Mn Cu Zn

30 Samoa I 4.73 0.33 3.85 0.97 0.18 0.104 0.068 0.003 0.014

Samoa II 4.87 0.34 2.87 0.78 0.30 0.120 0.060 0.004 0.019

60 Samoa I 4.21 0.44 3.33 1.82 0.16 0.425 0.059 0.005 0.017

Samoa II 4.39 0.46 3.11 1.54 0.18 0.489 0.075 0.005 0.028

90 Samoa I 3.89 0.41 2.93 2.30 0.37 0.625 0.049 0.006 0.028

Samoa II 4.01 0.44 2.89 2.21 0.45 0.513 0.050 0.007 0.036

120 Samoa I 4.37 0.41 3.75 2.23 0.37 0.675 0.053 0.010 0.024

Samoa II 4.53 0.43 3.71 1.76 0.37 0.733 0.045 0.011 0.040

150 Samoa I 3.94 0.41 4.01 1.42 0.35 1.052 0.069 0.007 0.031

Samoa II 4.29 0.39 4.00 1.17 0.29 0.940 0.067 0.007 0.036

180 Samoa I 2.94 0.39 3.78 12.11 0.31 1.103 0.085 0.006 0.033

Samoa II 3.43 0.38 3.63 10.82 0.31 1.364 0.090 0.006 0.036

210 Samoa I 3.14 0.47 3.35 2.40 0.27 1.649 0.071 0.006 0.031

Samoa II 3.18 0.40 3.35 1.97 0.29 0.645 0.074 0.005 0.036

240 Samoa I 4.18 0.56 3.34 1.95 0.36 0.575 0.066 0.006 0.029

Samoa II 3.41 0.48 2.91 2.41 0.34 0.050 0.064 0.005 0.038

201

Y =

4.6

613x

- 11

5.63

R

² = 0

.891

2

Y =

3.9

048x

- 99

.433

R

² = 0

.848

9

0

100

200

300

400

500

600

700

800

040

8012

016

020

0

Corm dry matter (kg/ha)

Plan

t N (k

g/ha

) Sa

moa

1Sa

moa

2

Y =

18.

84x

- 11.

35

R² =

0.9

733

Y =

15.

562x

+ 7

.506

3 R

² = 0

.991

4

0

100

200

300

400

500

600

700

800

010

2030

4050


Plan

t P (k

g/ha

)

Sam

oa 1

Sam

oa 2

Y =

2.5

586x

- 62

.466

R

² = 0

.965

2

Y =

2.1

956x

- 69

.9

R² =

0.8

727

010

020

030

040

050

060

070

080

0

050

100

150

200

250

300

350


Plan

t K (k

g/ha

) Sa

moa

1Sa

moa

2

Y =

3.6

179x

+ 8

3.84

R

² = 0

.930

3

Y =

3.9

214x

+ 7

5.44

1 R

² = 0

.938

6

0

100

200

300

400

500

600

700

800

040

8012

016

020

0


Plan

t Ca

(kg/

ha)

Sam

oa 1

Sam

oa 2

Y =

28.

63x

- 45.

401

R² =

0.9

Y =

26.

131x

- 62

.108

R

² = 0

.908

3

0

100

200

300

400

500

600

700

800

05

1015

2025

30


Plan

t Mg

(kg/

ha)

Sam

oa 1

Sam

oa 2

Fi

gure

4.2

6 R

elat

ions

hip

betw

een

corm

dry

mat

ter y

ield

and

mac

ronu

trien

t con

tent

s of t

he tw

o cu

ltiva

rs

202

Y = 30.67x + 76.94 R² = 0.9303

Y = 46.495x + 75.888 R² = 0.9275

0

100

200

300

400

500

600

700

800

0 5 10 15 20 25

Cor

m d

ry m

atte

r (kg

/ha)

Plant Fe (kg/ha)

Samoa 1 Samoa 2

Y = 701.32x - 39.503 R² = 0.9714

Y= 635.66x - 31.192 R² = 0.9683

0

100

200

300

400

500

600

700

800

0 0.2 0.4 0.6 0.8 1 1.2

Cor

m d

ry m

atte

r (kg

/ha)

Plant Mn (kg/ha)

Samoa 1 Samoa 2

Y = 1614.9x - 30.827 R² = 0.9686

Y = 1308.7x - 61.955 R² = 0.9609

0

100

200

300

400

500

600

700

800

0 0.1 0.2 0.3 0.4 0.5 0.6

Cor

m d

ry m

atte

r (kg

/ha)

Plant Zn (kg/ha) Samoa 1 Samoa 2

Y = 7782.7x - 51.864 R² = 0.8888

Y = 8187.3x - 87.092 R² = 0.8135

0

100

200

300

400

500

600

700

800

0 0.02 0.04 0.06 0.08 0.1

Cor

m d

ry m

atte

r (kg

/ha)

Plant Cu (kg/ha) Samoa 1 Samoa 2

Figure 4.27 Relationship between corm dry matter yield and micronutrient contents of

the two cultivars

203

CHAPTER 5

SUMMARY, CONCLUSION AND RECOMMENDATIONS

5.1 Summary

Society has long known of its dependence on soils, but has only recently appreciated its

biologically active component. By any measure, soils worldwide are in trouble from

being over exploited into continuous submission, and there is a critical and immediate

need to apply knowledge about the role of below ground functioning in sustaining soils.

A major feature of global change in the tropical regions is the land use change associated

with agricultural intensification. Intensification is necessary to ensure global food

supplies, but it can have many negative consequences in terms of diminished delivery of

ecosystem services, including loss of primary productivity through changes in soil

fertility and/or increases in soil-borne diseases, loss of cleansing potential for wastes and

pollutants, disruptions of global elemental cycles, feedbacks of greenhouse gas fluxes

and erosion. This, in turn, frequently results in reduction of soil biodiversity and

subsequent functioning and presents a challenge to increased agricultural productivity in

regions that are degraded. As intensification proceeds, above-ground biodiversity is

reduced with consequences for the below-ground diversity and thence the biological

regulation of soil based ecosystem services. These regulatory functions are often

described as being ‘substituted’ by inputs such as the use of mechanical tillage, chemical

fertilisers and pesticides. These are also assumed to reduce the ability of agricultural

systems to withstand unexpected periods of stress, bringing about undesirable effects.

For any agricultural system, regardless of whether it is classified as a low input or high

input system, inputs must not be less than outputs - if they are, then mining of existing

stocks occur and the system must deteriorate in time. Where there is constant removal of

nutrients in crops harvested for human consumption which are not returned to the

system, then no such system can be sustainable without some further external inputs. No

204

group of technologies, whether based on organic or inorganic sources of nutrients, can

contravene this basic law of mass conservation.

Large numbers of farmers in the tropics have limited access to inputs but in an attempt

to intensify production, the complexity of their agro-ecosystems has been drastically

reduced nonetheless. The maintenance of diversity of crops and other plants in cropping

systems is a widely accepted management practice that buffers farmers against short

term risks. Shifting cultivators have traditionally alternated periods of crop production

with periods of fallow to restore soil fertility and suppress weeds. In most cases the

cropping period is comparatively shorter than the fallow period during which time the

land is usually unproductive in terms of generating a livelihood. In recent years, the

principal foci of researchers have been on the ways to shorten this period. Thus an

accelerated fallow, which involves the cultivation of specific fast growing leguminous or

non-leguminous trees, shrubs, vines, legumes and other plants, are used to improve the

soil fertility faster then would occur otherwise. This has been the focus of this research

considering the economic climate under which most Pacific Island farmers operate.

The interaction between the processes of soil disturbances (and recovery) and the

resultant effects on ecosystem C and N fluxes are poorly understood in the South Pacific

region. An understanding of the dynamics of soil organic C and N as affected by

farming practices is imperative for maintaining soil productivity and mitigating risks

associated with climate change. Soil organic matter ultimately governs the ability of the

soil to provide long term sustainable agro-ecosystems and the microbial activity that

governs them. This research investigated the efficacy of selected green cover crop fallow

practices with and without judicious use of a complete chemical fertiliser and biochar

towards improving the yield and biological functioning of taro soil over different agro-

ecological zones of Samoa. Selected soil biochemical indicators were evaluated to

reflect nutrient recycling dynamics in the soil environment together with the nutrient

uptake of cover crops, nematode population and activity differentials. The dry matter

accumulation and nutrient use efficiency of two cultivars of taro were also evaluated.

205

Mucuna fallow-treated soils showed a significant mean increase in active carbon over all

the other fallow treatments across all the sites. In general, fallowing increased the soil

labile C across all the treatments. Comparatively, the Savaii sites recorded higher mean

levels of active C than Upolu sites. The mean levels of soil biological activity (FDA

hydrolysis activity) and the mineralisable N pools were also significantly higher under

mucuna fallows comparatively, owing to its relatively higher biomass production over

all the sites.

Mineral N fluxes showed similar trends for NH4+-N and NO3

- -N with both reaching a

maximum 120-150 days after decomposition and then levelling off. The NO3- -N levels

under all the fallow systems across all the sites was much higher than NH4+-N,

signifying a high potential for N leaching. Mucuna fallow mulch significantly

mineralised more plant available N than all the other fallows, while the Savaii sites

recorded significantly higher levels than the Upolu sites. Cumulative net mineralisation

potentials did not show any statistical significance between the fallow covers for the

Safaatoa and Siufaga sites; however, mucuna comparatively exhibited significant

increases in the mineral N pools for Salani and Aopo sites.

Nematode community analyses revealed significant declines in total and plant parasitic

population counts under mucuna fallow while the population of predatory and

omnivorous nematodes improved. Analyses of food web indices showed significant

nutritionally enriched soils under all the fallows while no statistical significance was

detected between any of the fallows at the Safaatoa site. Significantly better structured

food web with greater number of active trophic levels resulted from all the fallows at

Salani site; however, structure index between fallows was not significant for the

Safaatoa site. Analyses of channel index showed that the predominant channel of

decomposition of all the fallows covers at both the sites was the fungal decomposition

pathway. Significant declines in bacterial and fungal feeding nematodes and significant

increases in predatory and omnivorous nematode population and activities were

observed through carbon flow proportions under all the fallows at Salani site but not at

206

the Safaatoa site. A significant decrease in plant parasitism was only recorded under

mucuna fallow practice at the Safaatoa site.

Analyses of all the cover crops for their nutrient concentrations and uptake showed that

while generally the nutrient content of erythrina was significantly higher than that of

Mucuna, the latter had higher nutrient uptake over the six month fallow duration, owing

to its comparatively higher biomass production over all the four sites. This was well

reflected on the taro yields for Salani and Siufaga sites (high rainfall zones), where

biomass production was comparatively higher. The yield of taro under mucuna with no

supplementation of any fertiliser was not significant from taro grown under traditional

grass with the crop being supplemented by the recommended rate of 400 kg/ha of

complete fertiliser.

The analyses of taro corm dry matter yields and nutrient uptake also revealed that

significantly higher mean corm dry matter production and nutrient uptake occurred

under mucuna fallows across all the sites. Comparable yields under biochar fallow can

be attributed to the biochar fallow to enhance appreciable quantities of K uptake.

The controlled rate of decomposition under screen house conditions for the fallow

covers on soils over a 90 days incubation period revealed significant differences and

interactive responses of fallow, soil and rate on the biochemical soil indicators of

nutrient recycling and mineralisation. Mean soil labile C and biological activities

increased following organic matter mineralisation across all the fallows. Nitrogen

mineralisation potentials illustrated that the rate of release of plant available N was too

rapid (within the first 30 days after incubation) under the erythrina cover to be fully

utilised by a crop of taro. On the contrary, the rate of release of plant available N was

much slower under the mucuna fallow cover and had not peaked during the 90 days

incubation period, providing a more desirable option for N availability for the taro crop.

The mean phosphate mineralisation was again significantly higher under the mucuna

treated mulch with comparative higher levels for the Salani soil. Most of the phosphate

207

mineralisation occurred during the first 30 days after incubation, desirable for early root

development. Assay of soil urease activity suggested that the highest levels of hydrolysis

of urea occurred under the mucuna mulch, while Salani soil supported higher hydrolysis

than Safaatoa soil.

The nutrient budget pot experiment revealed that while the concentrations of all the plant

nutrients (except Fe) uptake was higher for cultivar Samoa 2 than cultivar Samoa 1, the

nutrient use efficiency (kg of edible dry matter produced per kg of nutrient taken up) for

cultivar Samoa 1 was higher for N, P, K, Mg, Mn and Cu over cultivar Samoa 2.

However, for Ca, Fe and Zn cultivar Samoa 2 had a higher nutrient use efficiency over

cultivar Samoa 1.

5.2 Conclusions

The central tenet is that the functional significance of biological systems is highly

dynamic across scales in space and time. Broadly, this means that what is measured, and

where and when it is measured has to take account of the features and management that

are likely to affect overall ecosystem services in different types of comparisons.

Nutrient recycling by the vast array of soil biota (litter transformers and ecosystem

engineers) is essential for all forms of agriculture. Additions and subsequent

decomposition of organic matter is largely mediated by the enzymatic activity of

bacteria and fungi and affects the entire biochemical soil environment.

The soil health assessment of Samoan taro soils was best achieved with the help of a

minimum data set consisting of selected physical, chemical, biological and biochemical

indicators. The physical (bulk density and particle size) and the chemical indicators (pH,

EC, CEC together with the routine nutritional and fertility indicators) used for site

characterisation revealed remarkable spartial variations between the climatic zones as

well as the islands and provided a sound platform for resoning out the treatment

208

differences. The biological (nematode community analysis) and biochemical indicators

(labile carbon, FDA activity, PMN, and N mineralisation measurements) were highly

sensitive to management and provided an important insight into short term temporal

changes with regards to biochemical transformation and mineralisation fluxes. These

indicators were particularly sensitive to the fallow treatments. Consequently, the

combination of the indicators selected for the study enabled a reasonably comprehensive

assessment of the Samoan taro sites under study.

This study confirmed the integrative benefits of organic matter management through

green manure cover crop fallows through a series of field and supporting pot

experiments. From this investigation, it was perceptible that mucuna fallow offers multi

benefits to the soil ecosystem with regards to the biochemical processes that it supports

and the overall biological functioning of Samoan taro soils over the traditional farmers

practice of grass fallow. It has also resulted in significantly higher taro yields across

different agro ecological zones of Samoa, which connotes that a wider adoption of the

fallow practices by the taro farmers of Samoa could be promising.

The incorporation of mucuna into fallow programmes can significantly reduce reliance

on chemical fertilisers, reduce soil erosion, comparatively shorten the fallow durations,

help suppress plant parasitic nematode, improve soil fertility through organic matter

recovery and N additions and enhance nutrient uptake and generally improve taro yields.

It can thereby improve the farmer’s ability to mitigate short term risks associated with

declining soil fertility. All these benefits are only achievable if the decomposition phase

of the fallow cover is temporally well synchronised with the life cycle of the taro crop

and the cover crops are left to decompose as surface mulches, under the prevailing

tropical environment of Samoa.

Finally, the practice of fallowing with nitrogen fixing leguminous cover crop, such as

mucuna, can be seen as the way forward for Samoan taro farmers towards sustaining

yields and maintaining healthy soils.

209

5.3 Recommendations for future researchers and farmers

Based on the findings of this investigation the following recommendations can be made:

1. Given the multiple benefits that it offers, the inclusion of mucuna as an improved

leguminous fallow system for the Samoan taro soils is highly recommended.

2. Future researches need to investigate the comparative effects of other proven short

term and practical best-bet leguminous systems for fallow purposes for Samoa soils.

3. Future researches needs to focus on the effect of longer fallow durations to

determine the long term effects of fallow crops on soil organic C and N dynamics.

4. It is highly recommended that the decomposition of the fallow litter be well

synchronised temporally with the critical periods of nutrient requirements during

life cycle of the crop.

5. This research is based on mostly evaluating and comparing the few biological and

biochemical indices of the soil ecosystem. However, future research needs to

include a detailed investigation of physicochemical and biophysical components of

the soil systems to reflect a complete picture of the soil ecosystems in assessing soil

health status of treated soils.

6. Quantifying biological fixation of N and its subsequent release is also an important

research area which needs to be looked into.

7. Other problematic soils (marginal, degraded, saline and atoll soils) that are found in

the Pacific region should be subjected to research work with green manure cover

cropping.

8. The use of biochar as a carbon sequester should be further investigated to fully

understand the benefits that it has to offer.

9. The effects of green manure cover cropping should also be applied to other crop and

site specific situations.

10. The comparative study between the methods of cover crop residue management

systems, namely mulching and ploughed incorporations also needs to be

determined.

11. Comparative economic analysis of fallow cropping systems is highly

recommended.

210

REFERENCES Abawi, G.S., and Widmer, T.L. 2000. Impact of soil health management practices on

soil borne pathogens, nematodes and root diseases of vegetable crops. Applied Soil

Ecology 15, 37-47.

Adams, M.A. and P.M. Attiwill. 1986. Nutrient cycling and nitrogen mineralisation in

eucalyptus forests of south-eastern Australia. II. Indices of nitrogen mineralisation. Plant

and Soil. 92: 341-362.

Agharkar, S.P. 1991. Medicinal plants of Bombay presidency. P. 1–2. Scientific

Publication Jodhpur. India.

Alexander, M. 1977. Introduction to Soil Microbiology. 2nd Edition. John Wiley and

Sons Inc., New York, ISBN-13: 9780894645129, Pages: 467.

Al-Rashidi, K., and M.M. Al-Jabri. 1990. Nitrification and urea hydrolysis in amended

calcareous saline soils. Arid Soil Research and Rehabilitation. 4(4): 543-549.

Amador, J.A., A.M. Glucksman, J.B. Lyons and J.H. Gores. 1997. Spatial distribution of

soil phosphatase activity within a riparian forest. Soil Science Journal., 162: 808-825.

Allison, S.M..J. and J.I. Proser. 1991. Urease activity in neutrophilic autotropic

ammonia oxidising bacteria isolated from acid soils. Soil Biology and Biochemistry.

23(1):45-51.

Andrews, S.S., D.L. Karlen, and C.A. Cambardella. 2004. The soil management

assessment framework: A quantitative soil quality evaluation method. Soil Science

Society of America Journal 68, 1945-1962.

Anonymous, 1934. Green manures. Straits Settlements and Federal Malaya States,

Department of Agriculture. Leaflet No. 7.

211

Antil, R.S., A.P. Gupta and R.P. Narwal. 1993. Effects of substrate concentration, soil

moisture and organic amendments on urease activity of soil containing variable nickel

amounts. Acid Soil Research and Rehabilitation. 7(4): 381-387.

Arshad, M.A., and S. Martin. 2002. Identifying critical limits for the soil quality

indicators in agro-ecosystems. Agriculture, Ecosystems and Environment 88:153-160.

Atlas, R.M., and R. Bartha. 1998. Microbial Ecology: Fundamentals and Applications.

Benjamin Cummings, Dordrecht, 640pp.

Azam, F., R.L. Mulvaney, and F.W. Simmons. 1995. Effects of ammonium and nitrate

on mineralisation of nitrogen from leguminous residues. Biol. Fertil. Soils 20:49–52.

Baath, E. 1989. Effects of heavy metals in soil on microbial processes and populations (a

review). Water, Air and Soil Pollution. 47: 325–375.

Balasubramanium, A., R. Siddaramappa and G. Rangaswami. 1972. Effects of organic

manuring on the activities of the enzymes hydrolysing sucrose and urea on soil

aggregation. Plant and Soil. 37(2):319-328.

Balestra, G.M., and I.J. Misaghi. 1997. Increasing the efficiency of the plate counting

method for estimating bacterial diversity. Journal of Microbiological Method 30, 111–

117.

Balota, E.L. and J.C.D. Chaves. 2011. Microbial activity in soil cultivated with different

summer legumes in coffee crop. Brazillian Archives of Biology and Tecnology. Vol. 54

No: 1.

Bandick, A.K. and R.P. Dick 1999. Field management effects on soil enzyme activities.

Soil Biology and Biochemistry 31: 147 1-1479.

Barros, A.F., V.P. Campos, J.C.P. da Silva, M.P. Pedroso, F.H.V. Medeiros, E.A. Pozza

and A.L. Reale. 2014. Nematicidal activity of volatile organic compounds emitted by

212

Brassica juncea, Azadirachta indica, Canavalia ensiformis, Mucuna pruriens and

Cajanas cajan against Meloidogyne incognita. Applied Soil Ecology. 80: 34-43.

www.elsevier.com/locate/apsoil.

Baxter S.J., M.A. Oliver, and J. Gaunt. 2003. A geostatistical analysis of the spatial

variation of soil mineral nitrogen and potentially available nitrogen within an arable

field. Precision Agriculture 4, 213-226. doi: 10.1023/A:1024565507688.

Bieleski, R.L. and I.B. Ferguson, 1983. Physiology and Metabolism of Phosphate and its

Compounds. In: Inorganic Plant Nutrition, Lauchli, A. and R.L. Bieleski (Eds.).

Springer-Verlag, Berlin, pp: 415-449.

Binkley, D. and S.C. Hart. 1989. The components of nitrogen availability assessments in

forest soils. Advances in Soil Science, 10: 55-112.

Blakemore, L.C., P.L. Searle, and B.K. Daly. 1987. Methods of Chemical Analysis of

Soils. New Zealand Soil Bureau Scientific Report 80. Pp: 103.

Bloem, J., A.J. Schouten, M. Rutgeers, A van der Werf, and A.M. Breure. 2006.

Monitoring and evaluating soil quality. In: Bloem, J et al. (eds.): Microbial Methods for

Assessing Soil Quality. CAB International, p. 23-49.

Boateng, S.A. 2005. Mucuna pruriens and its effects on some physical, chemical and

biological properties of a forest Acrisol. West African Journal of Applied Ecology. 8(1).

http://ajol.info/index/wajne/articles/view/45789.

Boehm, H.P. 1994. Some aspects of the surface chemistry of carbon blacks and other

carbons. Carbon 32:759-769.

Bolton, H. jr., L.T. Elliot, R.I. Papendick and D.E. Bezdicek. 1985. Soil microbial

biomass and selected soil enzyme activites: effect of fertilization and cropping practices.

Soil Biology and Biochemistry. 17(3): 297-302.

213

Bonanomi G, V. Antignani, M. Capodilupo, F. Scala. 2010. Identifying the

characteristics of organic soil amendments that suppress soilborne plant diseases. Soil

Biology and Biochemistry 42, 136-144.

Bongers, T., 1990. The maturity index: an ecological measure of environmental

disturbance based on nematode species composition. Oecologica 83, 14-19.

Bongers, T., and Bongers, M. 1998. Functional diversity of nematodes. Applied Soil

Ecology 10:239-251.

Bremner, J.M. and D.R. Keeney. 1965. Steam distillation methods for determination of

ammonium, nitrate and nitrite. Analytica Chimica Acta. 32: 485-495.

Bremner, J.M. and R.L. Mulvaney. 1978. Urea activity in soils. In: Soil Enzymes (Ed.

R.G. Burns). Pp. 149-196. Academic Press, New York.

Bremner, J.M. and M.I. Zantua. 1975. Enzymes activity in soils at sub zero

temperatures. Soil Biology and Biochemistry. 7(6):383-387.

Briar, S.S. 2007. Nematode as bioindicators of soil food web health in agroecosystems:

A critical analysis. PhD Dissertation. Ohio State University.

Briggs, L.J. and J.W. McLane. 1907. The moisture equivalents of soils. U.S. Department

of Agriculture, Bureau of Soils, Bulletin 45. Government Printing Office, Washington

DC. 23 pp.

Briggs, L.J. and H.L. Shantz. 1912. The wilting coefficient for different plants and its

indirect determination. U.S. Department of Agriculture, Bureau of Plant Industry,

Bulletin 230. Government Printing Office, Washington DC. 83 pp.

Briggs, S.R., S. Cuttle, G. Goodlap, D. Hatch, J. King, S Roderick and M. Shepherd.

2005. Soil nitrogen building crops in organic farming. Defra Research Project OF0316

publications.

214

Buchholz, D.D. 1983. Soil test interpretations and recommendations handbook.

University of Missouri.

Buckles, D., T. Benard and Gustaros. 1998. Cover crops in hillside agriculture – Farmer

innovation in Mucuna. http://www.idrc.ca/openbooks/269-4/.

Budelman, A. 1988. The decomposition of the leaf mulches of Leucaena leucocephala,

Gliricidia sepium and Flemingia macrophylla under humid tropical conditions.

Agroforestry Systems 7, 3345.

Bulluck III, L.R., K.R. Barker and J.B. Ristaino. 2002. Influences of organic and

synthetic soil fertility amendments on nematode trophic groups and community

dynamics under tomatoes. Applied Soil Ecology 21, 233-250.

Burkill, I.H. 1966. A dictionary of the economic products of the Malay Peninsula.

Governments of Malaysia and Singapore, Kuala Lumpur, Malaysia.

Burns, R.G. 1978. Enzyme activity in Soil: Some Theoretical and Practical

Considerations. In: Soil Enzymes, Burns, R.G. (ed.). Academic Press, New York, USA.

Pp.295-340.

Cao, X., L. Ma, B. Gao, and W. Harris. 2009. Dairy-manure derived biochar effectively

sorbs lead and atrazine. Environmental Science and Technology. 43:3285-3291.

Carpenter-Boggs, C., J.L. Pikhul Jr., M.F. Vigil and W.E. Riedell. 2000. Soil nitrogen

mineralisation influenced by crop rotation and nitrogen fertilization. Soil Science Society

of America Journal. 64:2038-2045.

Carsky, J.R., S.A. Tarawali, M. Becker, D. Chikoye, G. Tian, and N. Sanginga. 1998.

Mucuna: a herbaceous cover legume with potential for multiple uses. International

Institute of Tropical Agriculture, Ibadan, Nigeria. RCMD Research Monograph No. 25.

215

Carter, D.L., M.M. Mortland, and W.D. Kemper. 1986. Specific surface. p. 413-423. In:

A. Klute (ed.) Methods of Soil Analysis, Part 1, Physical and Mineralogical Methods

2nd ed.. American Society of Agronomy, Madison, WI.

Carter, M.R., E.G. Gregorich, D.W. Anderson, J.W. Doran, H.H. Janzen, and F.J. Pierce.

1997. Concepts of soil quality and their significance. pp. 1-20. In: E.G. Gregorich and

M.R. Carter. (eds.) Soil quality for crop production and ecosystem health. Elsevier

Science, Amsterdam.

Cerato, A.B. and A.J. Lutenegger. 2002. Determination of surface area of fine-grained

soils by the ethylene glycolmonoethyl ether (EGME) method. Geotechnical Testing

Journal 25:315-321.

Chan, K.Y., L.V. Zwieten, I. Meszaros, A. Downie, and S. Joseph. 2007. Agronomic

values of greenwaste biochar as a soil amendment. Australian Journal of Soil Research.

45, 629-634.

Chan, K.Y., L.V. Zwieten, I. Meszaros, A. Downie, and S. Joseph. 2008. Using poultry

litter biochar as soil amendments. Australian Journal of Soil Research.46, 437-444.

Chapman, H.D. and P.F. Pratt. 1961. Methods for Analysis of Soils, Plants and Water.

Division of Agricultural Sciences, University of California, Riverside, California.

Chen, B., D. Zhou, and L. Zhu. 2008. Transitional adsorption and partition of nonpolar

and polar aromatic contaminants by biochars of pine needles with different pyrolytic

temperature. Journal Environmental Science and Technology. 42. 5137-5143.

Chen, H.B. 2003. Phosphatase activity and P fractions in soils of an 18-year-old Chinese

fir (Cunninghamia lanceolata) plantation. Forest Ecology Management. 178: 301-310.

Chikowo, R., P. Mapfumo, P. Nyamugafata and K.E. Giller. 2004. Mineral N dynamics,

leaching and nitrous oxide losses under maize following two year improved fallow on a

sandy loam soil in Zimbabwe. Plant Soil. 259:315-330.

216

Chin, W. and W. Kroontje. 1963. Urea hydrolysis and subsequent loss of ammonia. Soil

Science Society of America Proceedings. 27(3): 316-319.

Clark, A.J., A.M. Decker and J.J. Meisinger. 1994. Seeding rate and kill date effects on

hairy vetch-cereal rye cover crop mixtures for corn production. Agronomy Journal.

86:1065-1070.

Cloonan, M. J. 1993. Black nodules Dolichos sp. Australian Journal of Science 26:121

Colvan, S.R., J.K. Syers and A.G. O’Donnel. 2001. Effect of long-term fertiliser use on

acid and alkaline phosphomonoesterase and phosphodiesterase activities in managed

grassland. Biol Fertil Soils 34:258–263

Cookson W.R., and D.V. Murphy. 2004. Quantifying the contribution of dissolved

organic matter to soil nitrogen cycling using [sup.15] N isotopic pool dilution. Soil

Biology & Biochemistry 36, 2097-2100. doi: 10.1016/j.soilbio. 2004.05.013.

Cooke, R. C. 1963. Ecological characteristics of nematode-trapping fungi

Hyphomycetes. Ann. Rev. Applied Biology 52:431-437.

Cowie, A.L., V.E. Lonergan, S.M Fazle Rabbi, F. Fornasier, C. Mcdonald, S. Harden, A.

Kawasaki and B.K. Singh. 2013. Impact of carbon farming practices on soil carbon in

Northern New South Wales. Soil Research 51(8). 707-718. CSIRO Publishing.

Craig, E. and K. Wilkinson. 1998. Nitrogen fixing trees: multipurpose pioneers.

Permaculture International Journal. No. 55. Daly, B.K., V.T. Manu and S.M. Halavatau. 1984. Soil and plant analysis methods for use at

Agricultural Research Station, Vaini, Tonga. New Zealand Soil Bureau Laboratory Report AN2.

Dash, M.C., P.C. Mishra, R.K. Mohanty, and N. Bhatt. 1981. Effects of specific conductance

and temperature on urease activity in some Indian soils. Soil Biology and Biochemistry 13(1):

73-74.

217

Datta, A.C. 1990. Botany for degree students, Bombay, Calcutta and Madras. Oxford

University Press.

Davidson, E.A. S.C. Hart and M.K. Firestone. 1991. Internal Cycling of Nitrates in soils

of a Mature Coniferous Forest. Ecology 73(4). Pp. 1148-1156. Ecology Society of

America.

Decker, A.M., A.J. Clark, J.J. Meisinger, F.R. Mulford and M.S. McIntosh. 1994.

Legume cover crop contributions to no-tillage corn production. Agron J. 86(1):126–135

Degens, B.P., L.A. Schipper, G.P. Sparling, and L.C. Dunca. 2001. Is the microbial

community in a soil with reduced catabolic diversity less resistant to stress or

disturbance? Soil Biology and Biochemistry 33, 1143–1153.

Delaune, R.D. and W. H.jr. Patrick. 1970. Urea conversion to ammonia in waterlogged

soils. Soil Science Society of America Proceedings. 34(2): 603-607.

Dick, R.P. 1997. Soil enzyme activities as integrative indicators of soil health. In:

Biological indicators of soil health. C. E. Pankhurst, B. M. Doube and V.V.S.R. Gupta.

New York, CAB International: 121-156.

Dick, R.P. and M.A. Tabatabai, 1987. Factors affecting hydrolysis of polyphosphates in

soils. Soil Science, 143: 97-104.

Dick, R.P., D.P. Breakwell and R. F. Turco. 1996. Soil enzyme activities and

biodiversity measurements as integrative microbiological indicators. Methods for

assessing soil quality. 6775. Segoe Rd. Madison, WI 53771, USA. Soil Science Society

of America. 49: 9-17.

Dick, R.P., P.E. Rasmussen and E.A. Kerle 1988. Influence of long-term residue

management on soil enzyme activities in relation to soil chemical properties of a wheat-

fallow system. Biology and Fertility of Soils 6: 159-164.

218

Dilly, O., H.J Bach, F. Buscot, C. Eschenbach, W.L. Kutsch, U. Middlehoff, K. Pritsch,

and J.C. Munch. 2000. Characteristics and energetic strategies of the rhizosphere in

ecosystems of the Bornhoved Lake District. Applied Soil Ecology. 15: 201-210.

Doran, J.W. 1987. Microbial biomass and mineralisable nitrogen distributions in no

tillage and ploughed soils. Plant Biology and Fertility of Soils. 5: 68-75.

Doran, J.W., D.G. Fraser, M.N. Culik, and W.C. Liebhardt. 1987. Influence of

alternative and conventional agriculture management on soil microbial processes and

nitrogen availability. Americal Journal of Alternative Agriculture. 2: 99-106.

Doran, J. W., Sarrantonio, M., and Liebig, M. A., eds. 1996. Soil health and

sustainability, Advance Agronomy 56:1-54.

Doran, J.W., and Parkin, T.B. 1994. Defining and assessing soil quality. pp 3-21. In:

Doran, J.W., D.C. Coleman, D.F. Bezdicek, and B.A. Stewart. (Eds.), Defining Soil

Quality for a Sustainable Environment. Soil Science Society of America Special

Publication, vol. 35. Soil Science Society of America, Madison, Wisconsin.

Doran, J. W. and M. Safley. 1997. Defining and assessing soil health and sustainable

productivity. pp. 1-28. In: Biological Indicators of Soil Health. Pankhurst, E., B.M.

Doube, and V.V.S.R. Gupta, (eds.). AB International.

Doran, J.W. and M.R. Zeiss. 2000. Soil Health and Sustainability: Managing the Biotic

Component of Soil Quality. Applied Soil Ecology 15, 3-11.

Doran, J.W. and A.J. Jones. (Eds.). 1996. Methods for Assessing Soil Quality. Soil

Science Society of America Special Publication, vol. 49. Soil Science Society of

America, Madison, Wisconsin.

Doran, J.W. and T.B. Parkin. 1994. Defining and Assessing Soil Quality. pp. 3-20. In:

Doran, J.W., Coleman, D.C., Bezdicek, D.F., Stewart, B.A. (Eds.), Defining Soil Quality

219

for a Sustainable Environment. Soil Science Society of America Special Publication, vol.

35. Soil Science Society of America, Madison, Wisconsin.

Doran, J.W., M. Sarrantonio, M. Liebig, 1996. Soil Health and Sustainability. pp. 1-54.

In: Sparks, D.L. (Ed.), Advances in Agronomy, Vol. 56. Academic Press, San Diego.

Drinkwater, L.E., C.A. Cambardella, J. Reeder, and C. Rice. 1996. Potentially

mineralisable nitrogen as an indicator of biologically active soil nitrogen. In: Doran J.W.

and A.J. Jones (eds.) Methods for Assessing Soil Quality. Soil Science Society of

America. Special Publication 49:217-229.

Drinkwater, L.E., D.K. Letourneau, F. Workneh, A.C.H. van Bruggen, and C. Shenan.

1995. Fundamental differences between conventional and organic tomato

agroecosystems in California. Applied Ecology. 5:1098-1112.

Eilitta, M., F.L. Arteaga, M. Diaz, C, Guerrero, B. Herrera, G. Narvaez, L. Pare, G.C.A.

Robles and B. Triomphe. 2004. Cultivating maize with mucuna in Los Tuxtlas region of

south eastern Veracruz, Mexico. In: E. Maratta, J. Mureithi and R. Derpsch. 2004. Green

manure/cover crop systems of smallholder farmers – experiences from tropical and

subtropical regions. Kluwer Academic Publishers, New York.

http://www.ebooks.kluweronline.com

Ellert, B. H., Clapperton, M. J., and Anderson, D. W. 1997. An ecosystem perspective of

soil quality. pp. 115-141. In: Soil Quality for Crop Production and Ecosystem Health.

Gregorich, E. G. and Carter, M. R. (eds.). Elsevier, Amsterdam.

Engelem, B., Meiken, K., Wintzingerode, F., Heuer, H., Malkomes, H.-P., Backhaus, H.,

1998. Monitoring impact of a pesticide treatment on bacterial soil communities by

metabolic and genetic fingerprinting in addition to conventional testing procedures.

Applied and Environmental Microbiology 64, 2814–2821.

Esterman, E.F. and A.D. McLaren. 1961. Contribution of rhizosplane organisms to total

capacity of plants to utilise organic nutrients. Plant and Soil. 15(3): 243-260.

220

Ettema, C. H. 1998. Soil nematode diversity, species coexistence and ecosystem

function. Journal of Nematology. 30:159-169.

FAO. 1984. Legume Inoculants and Their Use, FAO of the United Nations, Rome. Excellent

practical handbook for inoculation.

FAO. 1991. Production Yearbook. Volume 45, pp. 67-184. Rome, Italy.

Farrell, R.E., V.V.S.R. Gupta and J.J. Germida. 1994. Effects of cultivation on the

activity and kinetics of arylsulfatase in Saskatchewan soils. Soil Biology and

Biochemistry 26(8): 1033-1040.

Federle, T.W., R.M. Ventullo and D.C. White 1990. Spatial distribution of microbial

biomass, activity, community structure and biodegradation of linear alkyl benzene

sulfonate (las) and linear alcohol ethoxylate (lae) in the subsurface. Microbial Ecology

20: 297-3 13.

Feller, C., J. Balesdent, B. Nicolardot and C. Cerri. 2001. Approaching soil organic

matter pools through particle size fractionation - Examples of tropical soils. pp. 53-67

CRC. In: Lal, R., K.M. Kimble, and R.F. Follet,. (eds.) Assessment methods for soil

carbon. Advances in soil science.

Ferris, H., 1993. New frontiers in nematode ecology. Journal of Nematology 25: 374-382.

Ferris, H., and M.M. Matute. 2003. Structural and functional succession in the nematode

fauna of a soil food web. Applied Soil Ecology 23:93-110.

Ferris, H., R.C. Venette, and S.S. Lau. 1996. Dynamics of nematode communities in

tomatoes grown in conventional and organic farming systems, and their impact on soil

fertility. Applied Soil Ecology 3:161-175.

Ferris, H., T. Bongers, and R.G.M de Geode. 2001. A framework for soil food web

diagnostics: extension of the nematode faunal analysis concept. Applied Soil Ecology. 18:

13-29.

221

Fosu, M., R.F. Kuhne and P.L.G. Vlek. 2004. Improving maize yield in the Guinea

savannah zone of Ghana with leguminous cover crops and P-K fertilisation. Agronomy

Journal 3(2): 115-121.

Frankenberger, W.T. jr. and M.A. Tabatabai. 1982. Amidase and urease activity in

plants. Plant and Soil. 64(2):153-166.

Frankenberger, W.T. jr. and W.A. Dick. 1983. Relationship between enzyme activityes

and microbial growth and indices in soil. Soil Science Society of America Journal. 47(5):

945-951.

Frankenberger, W.T. jr., J.B. Johanson and C.O. Nelson. 1983. Urease activity in

sewage sludge amended soils. Soil Biology and Biochemistry. 15(5): 543-549.

Freckman, D.W. 1988. Bacterivorous nematodes and organic-matter decomposition.

Agricultural Ecosystems and Environments 24:195-217.

Gautney, J., A.R. Barnard, D.B. Penney and Y.K. Kim. 1986. Solid state decomposition

kinetics of phenyl phosphorodiamidate. Soil Science Society of America Journal. 50 (3):

792-797.

Gaunt, J. and J. Lehmann. 2008. Energy balance and emissions associated with biochar

sequestration and pyrolysis bioenergy production. Environmental Science and

Technology. 42, 4152-4158.

García-Álvarez, A., M. Arias, M.A. Díez-rojo, and A. Bello. 2004. Effect of agricultural

management on soil nematode trophic structure in a Mediterranean cereal system. Appl.

Soil Ecol., 27: 197-210.

Garcia C, T. Hernandez, C. Costa, B. Ceccanti, G. Masciandaro, and C. Ciardi. 1993. A

study of biochemical parameters of composted and fresh municipal wastes. Bioresour

Technol 44: 17–23.

222

Gavrilova, A.N., N.A. Shimko and V.F. Savcenko. 1973. Dynamics of organic

phosphorus compounds and phosphatase activity in pale yellow sod-podzolic soil. Soviet

Soil Science Journal., 3: 320-328.

Giller, K.E., and G. Cadisch. 1997. Driven by nature: a sense of arrival or departure. pp

393-399. In: Cadisch, G. and K.E. Giller, (eds), Driven by nature: Plant litter quality and

decomposition. CAB International, Wallingford, UK.

Giller, K.E., M.H. Beare, P. Lavelle, A.M. Izac, and M.J. Swift. 1997. Agricultural

intensification, soil biodiversity and agroecosystem function. Applied Soil Ecology 4, 3–

16.

Glaser, B., J. Lehmann and W. Zech. 2002. Ameliorating physical and chemical

properties of highly weathered soils in the tropics with charcoal-a review. Biological

Fertility of Soils 35, 219-230.

Goenaga, R., U. Singh, F.H. Beinroth, and H. Prasad. 1991. SUBSTOR – Aroids: A

model in the making. Agrotechnology Transfer 14: 1-4.

Goenaga, R., and U. Chardon. 1993. Nutrient, uptake, growth and yield performance of

three tanier (Xanthosoma spp.) cultivars grown under intensive management. J. Agri.

Univ. P.R. 78:78-98.

Goertzen, S.L., K.D. Theriault, A.M. Oickle, A.C. Tarasuk, and H.A. Andreas. 2010.

Standardization of the Boehm titration. Part I. CO2 expulsion and endpoint

determination. Carbon 48:1252-1261.

Goh, K.M. 1980. Dynamics of stability of organic matter. In: B.K.G. Theng (ed). Soils

with variable charges. Soil Bureau, Department of Scientific and Industrial Research,

Lower Hutt, New Zealand.

223

Goh, K.J. and S.B. Chin. 2007. Mucuna bracteata: A cover crop and living green

manure [Online]. Available at:

http://books.google.com.fj/books?id=NtbZxS2hxnAC&dq=N+FIXED+BY+MUCUNA

&sitesec=reviews.

Greenland, D.J. 1998. Soil organic matter in relation to crop nutrition and management.

Pp. 85-90. In: K.M. Pretty and S.F. Sowdle (eds.). Proceedings of International

Conference on Management and Fertilization of Upland Soils in the Tropics and Sub-

tropics. 1986, September 7-11. Nanjing, People’s Republic of China.

Grundon, J. 1998. The importance of organic matter in soil fertility and crop health.

Organic Broadcaster; Midwest Organic and Sustainable Education Service.

Guinto, D.F.; Lauga, S.; Dauara, L.; Walasi, E.; Autufuga, D.; Perera, H.; Seuoti, D.;

Sae, A. and Tauati, S. 2015. Soil health assessment of taro (Colocasia esculenta) farms

in Samoa. In: Moving farm systems to improved nutrient attenuation. (Eds. L. D. Currie

and L. L. Burkitt http://flrc.massey.ac.nz/publications.html. Occasional Report No. 28.

Fertilizer and Lime Research Centre, Massey University,Palmerston North, New

Zealand. 7 pages.

Gunapala, N. and K.M. Scow. 1998. Dynamics of soil microbial mass and activity in

conventional and organic farming systems. Soil Biology and Biochemistry. 30:805-816.

Gupta, V.V.S.R. and J.J. Germida. 1988. Distribution of microbial biomass and its

activity in different soil aggregate size classes as affected by cultivation. Soil Biology

and Biochemistry 20(6): 777-786.

Hagerman, S. M., M. D. Jones, G. E. Bradfield, M. Gillespie and Hall, I.R. 1988.

Potential for exploiting vesicular – arbuscular mycorrhizas in agriculture. pp. 141-174.

In: A. Mizrahi (ed.). Advances in Biotechnological Processes: Biotechnology in

Agriculture. Alan R. Liss., Inc. New York.

224

Halvorson, J.J., J.L. Smith, and R.I. Papendick. 1997. Issues of scale for evaluating soil

quality. Journal of Soil and Water Conservation. January-February: 26-30.

Handayanto, E., G. Cadish, and K.E. Giller. 1997. Regulating N mineralisation from

plant residues by manipulation of quality. In: Driven By Nature: Plant Litter Quality and

Decomposition, eds G Cadish & KE Giller, CAB International Wallingford UK pp175-

185. 22

Hartley, C.W.S. 1987. The oil palm. West African Institute for oil palm research.

London and New York: Longman.

Hartemink, A.E. 2003. Soil fertility decline in the tropics - with case studies on

plantations. Netherlands.56-213.

Havlin, J.L., J.D Beaton, S.L. Tisdale, and W.L. Nelson. 2005. Soil fertility and nutrient

management. An introduction to nutrient management. 7th edition. Pearce/Prentice Hall.

Upper Saddle River, NJ.

Haynes, R.J. and R.S. Swift. 1988. Effect of lime and phosphate additions on changes in

enzyme activities, microbial biomass and levels of extractable nitrogen, sulphur and

phosphorus in acid soil. Biological Fertility of Soils. 6: 153-158.

Hedlund, A., E. Witter, B.X. An. 2003. Assessment of N, P and K management by

nutrient balances and flows on peri-urban smallholder farms in Southern Vietnam.

European Journal of Agronomy 20, 71-87.

Herbien, S.A. and J.L. Neal. 1990. Soil pH and phosphatase activity. Communications in

Soil Science and Plant Analysis. 21: 439-456.

Herrick, J. 2000. Soil quality: an indicator of sustainable land management. Applied Soil

Ecology 15:75-83.

225

Hill, G.T., N.A. Mitkowski, L. Aldrich-Wolfe, L.R. Emele, D.D. Jurkonie, A. Ficke,

Maldonado-Ramirez, S. Lynch and S.T. Nelson, E.B., 2000. Methods for assessing the

composition and diversity of soil microbial communities. Applied Soil Ecology 15, 25–

36.

Hillel, D. 1982. Introduction to soil physics. Academic Press, San Diego, CA.

http://soilhealth.cals.cornell.edu/ http://soils.usda.gov/.USDA, Natural Resource

Conservation Service.

Hope, C.F.A. and R.G. Burns 1987. Activity, origins and locations of cellulases in a silt

loam soil. Biology and Fertility of Soils 5: 164-170.

Hunt HW (1977). A simulation model for decomposition in grasslands. Ecol. 58: 469-484. Hunt, H.W., D.C. Coleman, E.R. Ingham, E.T. Elliott, J.C. Moore, S.L. Rose, C.P.P.

Reid, and C.R. Morley. 1987. The detrital foodweb in a shortgrass prairie. Biological

Fertility of Soils 3:57-68.

Ibewiro, B., N. Sanginga, B. Vanlauwe and R. Merckx. 2000. Nitrogen contributions

from decomposing cover crop residues to maize in a tropical savanna. Nutrient Cycling

in Agroecosystems. 57: 131-140.

IBSNAT. 1987. Field and laboratory methods for IBSNAT. Technical Report 2 pp. 37-

38. Department of Agronomy and Soil Science, College of Tropical Agriculture and

Human Resources, University of Hawaii, Honolulu, Hawaii.

Ingham, R.E., J.A. Trofymow, E.R. Ingham, and D.C. Coleman. 1985. Interactions of

bacteria, fungi, and their nematode grazers: Effects on nutrient cycling and plant growth.

Ecology Monograph. 55:119-140.

Islam, M.M., F. Iyamuremye, and R.P. Dick. 1998. Effects of organic residue

amendment on mineralisation of nitrogen in flooded rice soils under laboratory

conditions. Communications in Soil Science and Plant Analysis. 29: 971-981.

226

Jackson, L.E., J.P. Schimel, and M.K. Firestone. 1989. Short term partitioning of

ammonium and nitrate between plants and microbes in an annual grassland. Soil Biology

and Biochemistry. 21: 409-415.

Jacobs, B.C. and J. Clarke. 1993. Accumulation and partitioning of dry matter and

nitrogen in traditional and improved cultivars of taro (Colocasia esculenta (L.) Schott)

under varying nitrogen supply. Field Crops Research. 31:317-328.

Jaffee, B.A., E.C. Tedford, and A.E. Muldoon. 1993. Tests for density-dependent

parasitism of nematodes by nematode-trapping and endoparasitic fungi. Biological

Control 3:329- 336.

Javier, E.F and R.E. Tabien. 2003. Nitrogen dynamics in soils amended with different

fertilisers. Philippine Journal of Crop Science. 28(3): 49-60.

Jenkinson, D.S. and L.C. Parry. 1989. The nitrogen cycle in the Broadbalk Wheat

Experiment: A model for the turnover of nitrogen through the soil microbial mass. Soil

Biology and Biochemistry. 21: 535-541.

Jonsen, K., C.S. Jacobsen, and V. Torsvik. 2001. Pesticide effects on bacterial diversity

in agricultural soils - a review. Biology and Fertility of Soils 33, 443–453.

Juma, N.G. and M.A. Tabatabai, 1978. Distribution of phosphomonoesterases in soils.

Soil Science. 126: 101-108.

Kadono, A. S. Funakawa and T. Kosaki. 2010. Factors controlling potentially

mineralisable and recalcitrant soil organic carbon in humid Asia. Pedologist, 135-142.

Kandeler, E. and H. Gerber. 1988. Short-term assay of soil urease activity using

colorimetric determination of ammonium. Biology and Fertility of Soils. 6: 68-72.

Karlen, D.L., M.J. Mausbach, J.W. Doran, R.G. Cline, R.F. Harry, and G.E. Schuman,

1997. Soil Quality: a concept, definition, and framework for evaluation. Soil Science

Society of America Journal. 61: 4-10.

227

Karlen, D.L., S.S. Andrews and J.W. Doran. 2001. Soil Quality: Current concepts and

applications. Advances in Agronomy 74: 1-40.

Karlen, D.L., J.W. Doran, B.J.Weinhold and S.S.Andrews. 2003. Soil Quality:

Humankind’s foundation for survival. Journal of Soil and Water Conservation 58.

Karlen, D.L. and D.E. Stott. 1994. Framework for evaluating physical and chemical

indicators and soil quality. pp. 53-72 In: J.W. Doran, D.C. Leman, D.F. Bezdicek and

B. A. Stewarts, (eds.) Defining soil quality for a sustainable environment; Proceeding of

a symposium, Minneapolis, MN, USA, 4-5 November 1992. Soil Science Society of

America, Inc., Madison, WI.

Kaspar, T.C. and J.W. Swinger. 2011. The use of cover crops to manage soil. American

Society of Agronomy. In: Hatfield, J.L. and T.J. Sauer (eds) Soil Management: Building

a stable base for agriculture. Madison, Wisconsin.

Kear D. and B.L. Wood. 1959. The geology and hydrology of Western Samoa. New

Zealand Geological service, Wellington, New Zealand.

Keeney, D. 1980. Prediction of soil nitrogen availability in forest ecosystems: A

literature review. Forest Science 26: 159-171.

Keeney D.R., and J.M. Bremner. 1966. Comparison and evaluation of laboratory method

of obtaining an index of soil nitrogen availability. Agronomy Journal 58, 498-503.

Kennedy, I.R. and R.J. Roughley. 2002. The inoculants biofertiliser and its potential to

increase yield and reduce costs of production: The need for quality control. pp. 4-9. In:

I.R. Kennedy and A.T.M.A. Choudhury (eds.). Biofertilisers in Action. Rural Industries

Research and Development Cooperatives (RIRDC) Publication No.2/086, Sydney

Australia.

228

Klein, D.A., D.L. Sorensen, and E.F. Redente. 1995. Soil enzymes: a predictor of

reclamation and progress. In: Tate R.L., and D.A. Klein (eds.). Soil reclamation process.

Microbiological analysis and applications. Marcel Dekker, New York.

Kolberg, R.L., B. Rouppet, D.G. Westfall, and G.A. Peterson. 1997. Evaluation of an in

situ net soil nitrogen mineralisation method in dryland agroecosystems. Soil Sci. Soc.

Am. J. 61:504–508.

Kolberg, R.L., D.G. Westfall, and G.A. Peterson. 1999. Influence of cropping intensity

and nitrogen fertiliser rates on in situ nitrogen mineralisation. Soil Sci. Soc. Am. J.

63:129–134.

Kothandram, R., J. Mathew; A.R. Krishnakumar; J. Kochuresiamma; K. Jayarathnam

and M.R. Sethuraj. 1999. Comparative efficiency of M. bracteata D.C. and Pueraria

phaseoloides Benth, on soil nutrient enrichment, microbial population and growth of

Hevea. Short communications. Rubber Research Institute of India. Rubber Board, India.

Kumar, V. and R.J. Wagenet. 1984. Urease and kinetics of urea transformations in soils.

Soil Science. 137(4): 263-269.

Kuo, S. and E.J. Jellum. 2002. Influence of winter cover crop and residue management

on soil nitrogen availability and corn yield. Agronomy Journal. 61:145-152.

Kuperman, R.G. and M.M. Carreiro. 1997. Soil heavy metal concentrations, microbial

biomass and enzyme activities in a contaminated grassland ecosystem. Soil Biology and

Biochemistry 9 (2): 179-190.

Lai, C.M. and M.A. Tabatabatai. 1992. Kinetic parameters of immobilized urease. Soil

Biology and Biochemistry. 24(3): 225-228.

Laird, D.A. 2008. The charcoal vision: A win-win-win scenario for simultaneously

producing bioenergy, permanently sequestering carbon, while improving soil and water

quality. Agronomy Journal. 100, 178-181.

229

Lal, R. (ed.) 1997. Methods for assessment of soil degradation.

Lal, R. 1998. Basic concepts and global issues: soil quality and agricultural

sustainability. pp 3-12. In: Lal, R (Ed.), Soil Quality and Agricultural Sustainability.

Ann Arbor Science, Chelsea, MI, USA.

Landi, L., G. Renella, J.L. Moreno, L. Falchini and P. Nannipieri. 2000. Influence of

cadmium on the metabolic quotient, L-:D-glutamic acid respiration ratio and enzyme

activity: Microbial biomass ratio under laboratory conditions. Biological Fertility of

Soils. 32: 8-16.

Larson, W.E. and F.J. Pierce, 1994. The dynamics of soil quality as a measure of

sustainable management. pp. 37-51. In: J.W. Doran, D.C. Leman, D.F. Bezdicek and

B.A. Stewarts, (eds.) Defining soil quality for a sustainable environment; Proceeding of

a symposium, Minneapolis, MN, USA, 4-5 November 1992. Soil Science Society of

America, Inc., Madison, WI.

Lehmann, J. 2007. Bio-energy in the black. Frontier Ecology and Environment. 5, 381-387.

Lehmann, J., J. Gaunt, and, M. Rondon. 2006. Bio-char sequestration in terrestrial

ecosystems-A review. Mitigating and Adapting Strategies for Global Change. 11, 403-427.

Lehmann, J. and S. Joseph (Eds). 2008. Biochar for Environmental Management

Science and Technology, Earthscan: Sterling, VA.

Letey, J., R.E. Sojka, D.R. Upchurch, D.K. Cassel, K.R. Olson, W.A. Payne, S.E.

Petrie, G.H. Price, R.J. Reginato, H.D. Scott, P.J. Smethurst and G.B. Triplett. 2003.

Deficiencies in the soil quality concept and its application. Journal of Soil and Water

Conservation 58: 180-187.

Letey, J. 1985. Relationship between soil physical properties and crop production.

Advances in Soil Science 1:277-301.

230

Liang, B., J. Lehmann, D. Solomon, J. Kinyangi, J. Grossman, B. O’Neill, J.O.

Skjemstad, J. Thies, F.J. Luizão, J. Petersen, and E.G. Neves. 2006. Black carbon

increases cation exchange capacity in soils. Soil Science Society of America Journal. 70:

1719-1730.

Lindsay, W.L, and W.A. Norvell. 1978. Development of a DTPA soil test for zinc, iron,

manganese and copper. Soil Science Society of America Journal 42: 421-428.

Liu, E., Y Changrong, X. Mei, Y. Zhang and T. Fan. 2013. Long term effects of manure

and fertiliser on soil organic carbon pools in dry land farming in Northwest China. PloS

One. 8(2) e56536. US National Library of Medicine. National Institute of Health.

Maathuis, K. and H. van Meer. 2003. Spatial evaluation of saturated hydraulic

conductivity and soil erosion in Alafua catchment Samoa. MSc. thesis submitted to the

Wageningen University.

MacDicken, K.G. 1994. Selection and Management of Nitrogen-Fixing Trees. Winrock

International Institute for Agricultural Development, Morrilton, Arkansas, USA.

Madejon, E., P. Burgos, R. Lopez and F. Cabrera. 2001. Soil enzymatic response to

addition of heavy metals with organic residues. Biology and Fertility of Soils 34: 144-

150.

Madsen, E.L., 1996. A critical analysis of methods for determining the composition of

biogeochemical activities of soil microbial communities in situ. In: Stotzky, G. and J.M.

Bollag (Eds.), Soil Biochemistry, Vol. 9. Marcel Dekker, New York, pp. 287-370.

Marinari, S., G. Masciandaro, B. Ceccanti and S. Grero. 2000. Influence of organic and

mineral fertilisers on soil biological and physical properties. Bioresour. Technology., 72:

9-17.

Mathews, C. 2006. The introduction and establishment of a new leguminous cover crop,

Mucuna bracteata under oil palm in Malaysia.

231

McBride, R.G., R.L. Mikkelsen and K.R. Barker. 2000. The role of low molecular

weight organic acids from decomposing rye in inhibiting root-knot nematode

populations in soil. Appl. Soil Ecol. 15:243–251.

McTaggart, I.P. and K.A. Smith. 1993. Estimation of potentially mineralisable nitrogen

in soil by KCl extraction. II. Comparison with soil N uptake in the field. Plant and Soil

157, 175 184. doi: 10.1007/BF00011046

McSorley, R., and Gallaher, R.N. 1995. Cultural practices improve crop tolerance to

nematodes. Nematropica 25:53-60.

Mendes, I.C., A.K. Bandick, R.P. Dick and P.J. Bottomley. 1999. Microbial biomass and

activities in soil aggregates affected by winter cover crops. Soil Science Society of

America Journal 63: 873-881.

Mergedus, A., J. Kristl, A. Ivancic, A. Soba. V.Sustar, T. Krizan and V. Lebot. 2014.

Variation of mineral composition in different parts of taro (Colocasia esculenta) corms.

Food Chemistry. 170: 37-46.

Mitsui, H.K., K. Gorlach, H.J. Lee, R. Hattori, and T. Hattori. 1997. Incubation time and

media requirements of culturable bacteria from different phylogenetic groups. Journal of

Microbiological Methods 30, 103–110.

Morrison, R.J. 1991. Analysis and prediction of the properties of Western Samoa soil.

Ph.D thesis. University of Hawaii, Honolulu, USA.

Murphy D.V., I.R.P. Fillery, and G.P. Spading. 1998. Seasonal fluctuations in gross N

mineralisation, ammonium consumption and microbial biomass in a Western Australian

soil under different land use. Australian Journal of Agricultural Research 49, 523 535.

doi: 10.1071/A97096.

232

Nahas, E., J.F. Centurion and L.C. Assis, 1994. Efeito das caractericas quimicas dos

solos sobre os microorganismos solubilisatores de fosfato e produtores de fosfatases.

Revista Brasileira de Ciencia de Solo., 18: 49-53.

Nannipieri, P. 1994. The potential use of soil enzymes as indicators of productivity,

sustainability

and pollution. In: Pankhurst CE, Doube BM, Gupta VVSR, Grace PR (eds) soil biota:

management in sustainable farming systems. CSIRO, East Melbourne, pp 238–244.

Nannipieri, P., B. Ceccanti and D. Bianchi, 1988. Characterisation of humus-

phosphatase complexes extracted from soil. Soil Biology and Biochemistry. 20: 683-691.

Nannipieri, P., L. Muccini and C. Ciardi. 1983. Microbial biomass and enzyme

activities: production and perisitence. Soil Biology and Biochemistry. 15(6): 679-685.

Nannipieri, P. and E.A. Paul. 2009. The chemical and functional characterisation of soil

N and its biotic components. Soil Biology and Biochemistry. 41: 2357-2369.

Nannipieri, P., P. Sequi and P. Fusi. 1996. Humus and enzyme activity. Humic

substances in terrestrial ecosystems. A. Piccolo. New York, Elsevier: 293-328.

National Academy of Sciences. 1979. Tropical Legumes: Resources for the Future,

National Academy Press, Washington, D.C.

Neely, C.L., M.H. Beare, W.L. Hargrove and D.C. Coleman. 1991. Relationship

between fungal and bacterial substrate-induced respiration (SIR), biomass and plant

residue decomposition. Soil Biol. Biochem. 16:459-464.

Neeson, R. 2001. Organic agriculture soil health strategies. Wollongbar. 21-22.

Neher, D. 2001. Role of nematodes in soil health and their use as indicators. Journal of

Nematology 33:161-168.

233

Neher, D.A., 1999. Nematode communities in organically and conventionally managed

agricultural soils. Journal of Nematology 31, 142-154.

Neher, D.A., J. Wub, M.E. Barbercheck and O. Anas. 2005. Ecosystem type affects

interpretation of soil nematode community measures. Applied Soil Ecology 30: 47-64.

Neher D.A. and R.K. Olson. 1999. Nematode communities in soil of four farm cropping

management systems. Pedobiologia, 43: 430-438.

Ngome, A.F.E., M. Becker and K.M. Mtei. 2011. Leguminous cover crops differentially

affect maize yields in three contrasting soil types of Kakamega, Western Kenya. Journal

of Agriculture and Rural Development in the Tropics and Subtropics. 112(1): 1-10.

Nitrogen Fixing Tree Association (Currently the FACT Net). 1989-1994. NFT

Highlights. Nitrogen Fixing Tree Association, Morrilton, Arkansas, USA.

Norman, M.J.T., C.J. Pearson and P.G.E. Searle. 1984. Sweet potato (Ipomea batatas),

pp. 245-257. In: The Ecology of Tropical Food Crops. Cambridge University Press,

Cambridge, Great Britain.

Odiambo, J.J.O. 2010. Decomposition and nitrogen release by green manure residues in

different soil types. African Journal of Agricultural Research. Vol. 5(1). Pp. 090-096.

OECD. 1999. Environmental performance reviews, Denmark. OECD (Organisation for

Economic Co-operation and Development) Publications, Paris.

Oickle, A.M., S.L. Goertzen, K.R. Hopper, Y.O. Abdalla, and H.A. Andreas. 2010.

Standardization of the Boehm titration: Part II. Method of agitation, effect of filtering

and dilute titrant. Carbon 48:3313-3322.

Oka, Y. 2010. Mechanisms of nematode suppression by organic amendments – A

Review. Applied Soil Ecology. 44: 101-115. www.elsevier.com/locate/apsoil

234

Olsen, S.R., C.V. Cole, F.S. Watanabe, and L.A. Dean. 1954. Estimation of available

phosphorus in soils by extraction with sodium bicarbonate. USDA Department Circular

939.

Othonen, R., S. Aikio, and H. Va¨re. 1997. Ecological theories in soil biology. Soil

Biology and Biochemistry 29, 1613–1619. Pagliai, M. and M. De-Nobili. 1993. Relationships between soil porosity, root

development and soil enzyme activity in cultivated soils. Geoderma, 56: 243-256.

Palm, C.A., and P.A. Sanchez. 1991. Nitrogen release from the leaves of some tropical

legumes as affected by their lignin and polyphenolic contents. Soil Biol. Biochem.

23:83-88.

Papendick, R.I., and J. Parr. 1992. Soil quality – the key to sustainable agriculture.

American Journal of Alternative Agriculture 7:2-3.

Parr, J.F., R.I. Papendick, S.B. Hornick and R.E. Meyer. 1992. Soil quality: Attributes

and relationship to alternative and sustainable agriculture. American Journal of

Alternative Agriculture 7:5-11.

Pascual, J.A., J.L. Moreno, T. Hernandez and C. Garcia. 2002. Persistence of

immobilised and total urease and phosphatase activities in a soil amended with organic

wastes. Bioresour. Technology, 82: 73-78.

Philipps, L., C.E. Stopes and L. Woodward. 1995. The impact of cultivation practice on

nitrate leaching from organic farming systems. In: H.F. Cook and H.C. Lee (eds) Soil

management in sustainable agriculture. pp 488-496. Wye College press.

Pimm S.L. 1984. The complexity and stability of ecosystems. Nature 307, 321-326.

Potter, M.J., K. Davies and A.J. Rathjen. 1998. Suppressive impact of glucosinolates in

brassica vegetative tissues on root lesion nematode Pratylenchus neglectus. J. Chem.

Ecol. 24:67–80.

235

Prasad M. and M. Spiers. 1978. Comparative study of ashing techniques for the

digestion of horticultural plant samples. Agricultural and Food Chemistry. 26: 824-827.

Pusparajah, E. 1997. Nutrition and fertiliser use in Hevea and associated covers in

Peninsula Malaysia – A Review. Journal of Rubber Research Institute. Sri lanka, 54,

270-283.

Ramirez C., and D. Kass. 1990. Advances in Erythrina Research at CATIE, p. 96-105.

In: Werner, D. and P. Mueller, (eds.). Fast Growing Trees and Nitrogen Fixing Trees.

International Conference. Marburg, October 8-12, 1989. Fischer Verlag, Stuttgart.

Reddy, K.N., R.M. Zablotowicz, M.A. Locke and C.H. Koger. 2003. Cover crop, tillage,

and herbicide effects on weeds, soil properties, microbial populations, and soybean

yield. Weed Sci. 51:987–994.

Renner, R. 2007. Rethinking biochar. Environ. Sci. Technol. 41, 5932-5933.

Reynolds, C.M., D.C. Wolf, and J.A. Armbuster. 1985. Factors related to urea

hydrolysis in soil. Soil Science Society of America Journal. 49(1): 104-108.

Richards B.N. 1974. Introduction to the soil ecosystem. Longmans NY.

Riley, J. 2001. The indicator explosion: local needs and international challenges.

Agriculture, Ecosystems, and Environment 87: 119-120.

Robertson, G.P., D. Wedin, P.M. Groftman, J.M. Blair, E.A. Holland, K.J. Nadelhoffer,

and D. Harris. 1999. Soil carbon and nitrogen availability: Nitrogen mineralisation,

nitrification, and soil respiration potentials. Pp. 258-271. In: G.P. Robertson et al. (eds.).

Standard soil methods for long term ecological research. Oxford University Press. New

York.

Robertson, G.P. and J.M Tiedje. 1985. An automated technique for sampling the

contents of stoppered gas collection vials. Plant and Soil. 83:453-457.

236

Roos, D. 2006. Cover crops: benefits and challenges. North Carolina Cooperative

Extension, Chatham County Center, Pittsboro, North, Carolina, USA.

Roy, J.L. and W.B. McGill. 2002. Assessing soil water repellency using the molarity of

ethanol droplet (MED) test. Soil Science 167:83-97.

Russo, R.O. 1990. Erythrina (Leguminosae: Papilionoideae): A Versatile Genus for

Agroforestry Systems in the Tropics. Journal of Sustainable Agriculture, Vol. 1(2):1991.

Sahrawat, K.L. 1980. Control of urea hydrolysis and nitrification in soils by chemicals –

prospects and problems. Plant and Soil. 57(2-3): 335-352.

Sahrawat, K.L. 1984. Effects of temperature and moisture on urease activity in semi-

acid tropical soils. Plant and Soils 78(2): 401-408.

Sakala, W.D., A.R. Saka and D. Kayira. 2003. The effect of mucuna-maize rotations and

short term improved fallows on soil fertility and maize yields under smallholder farms in

Malawi. Malawi Journal of Agriculture Science. 2(1):34-40.

Sanchez, P.A. and M.S. Swaminathan. 2005. Hunger in Africa: the link between

unhealthy people and unhealthy soils. The Lancet 365: 442-444.

Sanchez, P.A., C.A. Palm, L.T. Szott. E Cuevas, and R. Lal. 1989. Organic input

management in tropical agroecosystems. pp. 125-145. In: D.C. Coleman, J.M. Oades

and G. Uhera (eds.). Dynamics of soil organic matter in tropical ecosystems. NifTAL

Project, University of Hawaii, USA.

Sanginga, N., P. Houngnandan, N. Okogun, B. Vanlauwe, R. Merckx and O. van

Cleemput. 2011. Assessment of soil factors limiting growth and establishment of velvet

bean in farmer’s fields in the derived savanna of the Benin Republic. Biology and

Fertility of Soils. 33:416-422.

237

Savant, N.K., A.F. James, and G.H. McLellen. 1985. Effects of soil submergence on

urea hydrolysis. Soil Science. 140(2): 81-88.

Schmidt, H., L. Philipps, J.P. Welsh and P. von Fragstein. 1999. Legume breaks in

stockless organic farming rotations: nitrogen accumulation and influence on the

following crops. Biological Agriculture and Horticulture. 17: 159-170.

Schouten, A.J., J. Bloem, W.A.M. Didden, M. Rutgers, H. Siepel, L. Posthuma, and A.

M. Breure. 2000. Development of a biological indicator for soil quality. SETAC Globe

July-August: 30-32.

Sharifi, M., B.J. Zebarth, D.L. Burton, C.A. Grant, S. Bittman, C.F. Drury, B.G.

McConkey and N. Ziadi. 2008. Response of potentially mineralisable soil nitrogen and

indices of nitrogen availability to tillage system. Soil Sci Soc Am J 72(4):1124–1131.

doi:10.2136/sssaj2007.0243

Shashtry, R.K. 1995. Bhartiya Jabi-butiyon tatha vrikcho ke chamatkar. Dehati Pustak

Bhandar, Delhi.

Siciliano, S.D. and R.D. Roy. 1999. The role of soil microbial tests in ecological risk

assessment: differentiating between exposure and effects. Human and Ecological Risk

Assessment 5:671-682.

Singer A.C., E.S. Gilbert, E. Luepromchai, D.E. Crowley. 2000. Bioremediation of

polychlorinated biphenyl-contaminated soil using carvone and surfactant-grown

bacteria. Applied Microbiology and Biotechnology 54:838-843.

Singh, B.P. and Cowie, A.L. 2014. Long term influence of biochar on native organic

mineralisation in low carbon clayee soil. Scientific Reports. 4(3687).

Sinsabaugh, R.L., C.L. Lauber and M.N. Weintraub. 2008. Stoichiometry of soil

enzymes activity at global scale. Ecology Letter 11: 1252-1264.

238

Shearer S.A., J.A. Thomasson, T.G. Meuller, J.P. Fulton, S.F. Higgins, and S. Sampson.

1999. Yield prediction using a neural network classifier obtained using soil landscape

features and soil fertility data. ASAE-CSAE-SCGR Annual International Meeting,

Toronto, Ontario, Canada. ASAE Paper No. 993041.

Skujins, J.J. 1976. Extracellular enzymes in soil. CRC Critical Reviews in Microbiology

4: 383-421.

Smit, E., P. Leeflang, K. Wernars. 1997. Detection of shifts in microbial community

structure and diversity in soil caused by copper contamination using amplified ribosomal

DNA restriction analysis. FEMS Microbiology Ecology 23, 249–261. 1882 C. Crecchio

et al./Soil Biology & Biochemistry 36 (2004) 1873–1883.

Smith, J.L. and J.W. Doran. 1996. Measurement and use of pH and electrical

conductivity for soil analysis, in: Doran, J.W., Jones, A.J. (Eds.), Methods for Assessing

Soil Quality, SSSA, Special Publication No. 49, Madison, WI, USA.

Solaiman, M.Z. and L.K. Abbott. 2003. Phosphorus uptake by a community of

arbuscular mycorrhizal fungi in jarrah forest. Plant and Soil. 284: 313-320.

Sojka, R.E. and D.R. Upchurch. 1999. Reservations regarding the soil quality concept.

Soil Science Society of America Journal 63:1039-1054.

Sparling, G.P. 1997. Soil microbial biomass, activity and nutrient cycling as indicators

of soil health. In: C.E. Pankurst, B.N. Doube and V.S.R. Gupta (eds.) Biological

Indicators of Soil Health. CAB: Wallingford 4: 97-119.

Sparling, G., R.L. Parfitt, A.E. Hewitt, L.A. Schipper. 2003. Three approaches to define

desired soil organic matter contents. Journal of Environmental Quality 32: 760-766.

Speir, T.W. and D.J. Ross. 1978. Soil Phosphatase and Sulphatase. In: Soil Enzymes,

Burns, R.G. (Ed.). Academic Press, London, pp: 197-250.

239

Spier, T.W., R. Lee, E.A. Pansier and A. Carnis. 1980. A comparison of sulfatase,

urease and protease activities in planted and fallow soils. Soil Biology and Biochemistry.

12(3): 281-291.

Spokas, K.A., W.C. Koskinen, J.M. Baker, and D.C. Reicosky. 2009. Impacts of

woodchip biochar additions on greenhouse gas production and sorption/degradation of

two herbicides in a Minnesota soil. Chemosphere 77, 574-581.

Sprent, J.I. 1987. The ecology of the nitrogen cycle. Cambridge Press. Cambridge.

Stanford, G., and S.J. Smith. 1972. Nitrogen mineralisation potentials of soils. Soil

Science Society of America Proceedings. 36: 465-472.

Stocking, M.A. 2003. Tropical soils and food security: The next 50 years. Science 302,

1356-1359.

Stockdale E.A, and R.M. Rees. 1994. Relationships between biomass nitrogen and

nitrogen extracted by other nitrogen availability methods. Soil Biology & Biochemistry

26, 1213-1220. doi: 10.1016/0038-0717(94)90146-5.

Stubberfield, L.C.F. and P.J.A. Shaw. 1990. A comparison of tetrazolium reduction and

FDA hydrolysis with other measurements of microbial activity. Journal of

Microbiological Methods 12: 15 1-162.

Sumner, J.B. 1951. Urease. In: The Enzymes Vol. 1. Ed. J.B. Sumner. Pp . 873-892.

Academic Press. New York.

Tabatabai, M.A., and J.M. Bremner. 1969. Use of p-nitrophenylphosphate for assay of

soil phosphatase activity. Soil Biology and Biochemistry. 1:301-307.

Tabatabai, M.A. and J.M. Bremner. 1972. Assay of urease activity in soil. Soil Biology

and Biochemistry. 4:479-487.

240

Tan, K.H., E. Pusparajah, R. Sheperd and T.C. Hai. 1996. Proceedings of planters

conference, 1996:45-67.

Tiemann, L.K. and A.S. Grandy. 2014. Mechanisms of soil carbon accruals and storage

in bioenergy cropping systems. Soil Biology and Ecology. Michigan State University.

Tsiafouli, M.A., M.D. Argyropoulou, G.P. Stamouand S.P. Sgordelis. 2007. Is duration

of organic management reflected on nematode communities of cultivated soils? Belgian

Journal of Zoology. 137(2) 165-175.

Tuivavalagi, N.S., D.J. Hunter and F. Amosa. 2002. Tackling land degradation and

unsustainable land use in Samoa – with emphasis on agriculture sector. Proceedings of

the National Environmental Forum held in Apia, Samoa in 2001. pp. 32 – 36.

USDA, 1999. Soil quality test kit guide. United States Department of Agriculture,

Agricultural Research Service and Natural Resource Conservation Centre.

http://soils.usda.gov/sqi/kit2.hmtl.

van den Boogert, P.H.J.F., Velvis, H., Ettema, C.H., and Bouwman, L.A. 1994. The role

of organic matter in the population dynamics of the endoparasitic nematophagous

fungus, Drechmaria coniospora in microcrosms. Nematological Science 7:139-145.

van Diepeningen A.D., OJ. de Vos, G.W. Korthals & A.H.C van Bruggen. 2006. Effects

of organic versus conventional management on chemical and biological parameters in

agricultural soils. Appl. Soil Ecol., 31: 120-135.

Vasuidreketi, T. 2015. Influence of leguminous fallow crops on the suppression of

nematodes on taro soils in Samoa. MSc Dissertation. The University of the South

Pacific.

Vaughn, J.D. and G.K. Evanylo. 1998. Corm response to cover crop species, spring

desiccation time and residue management. Agronomy Journal. 90:536-544.

Velk, P.L.G. and M.F. Carter. 1983. The effects of soil environment and fertiliser

modifications on the rate of urea hydrolysis. Soil Science. 136(1): 56-63.

241

Verhejien, F., S. Jeddery, A. Bastos, C.M. van der Velde, and I. Diafas. 2010. Biochar

Application to Soils, JRC Scientific and Technical Reports: Institute for Environment and

Sustainability, European Communities.

Verma, D.M., N.P. Balakrishnan, and R.D. Dixit. 1993. Flora of Madhya Pradesh.

Botanical Survey of India. pp. 190–191. Lucknow, India.

Verstraete, W. and J.P. Voets. 1976. Soil microbial and biochemical characteristics in

relation to soil management and fertility. Soil Biology and Biochemistry 9: 253-258.

VSN International Ltd. 2011. Genstat Discovery Edition. Rothamsted Experimental

Station. http://discovery.genstat.co.uk.

Wagger, M.G. 1989. Cover crop management and nitrogen rate in relation to growth and

yield of no-till corn. Agron J. 81(3):533–538.

Wahab, M. 1997. RRIM. Short courses on rubber planting and nursery technique. Kuala

Lumpur: Rubber Research Institute of Malaysia.

Walkley, A and I.A. Black. 1934. An examination of the Degtareff method for

determining organic matter, and a proposed modification of the chromic and titration

method. Soil Science 37: 29-38.

Walley, F. T. Yates, J.W. Van Groenigen, and C Van Kessel. 2002. Relationships

between soil nitrogen availability indices, yield, and nitrogen accumulation of wheat.

Soil Science Society of America Journal 66, 1549-1561.

Wang, K.-H., R. McSorley, and R. N. Gallaher. 2004. Effects of Crotalaria juncea

amendment on squash infected with Meloidogyne incognita. Journal of Nematoogy.

36:290-296.

242

Wang, K.H., R. McSorley, and R.N. Gallaher. 2004. Relationships of nematode

communities and soil nutrients in cultivated soils. Soil Crop Science Society. Florida

63:105-113.

Wang, K.H., R. McSorley, A.J. Marshall, and R.N. Gallaher. 2004. Nematode

community changes associated with decomposition of Crotalaria juncea amendment in

litterbags. Applied Soil Ecology. 27:31-45.

Wang, Q., F. Xiao, T. He and S. Wang. 2013. Responses of labile soil carbon organic

matter and enzyme activity in mineral soils to forest conversion in the subtropics. Annals

of Forest Science. INRA and Springer-Verlay, France.

Wang W.J., C.J. Smith and D. Chen. 2004. Predicting soil nitrogen mineralisation

dynamics with a modified double exponential model. Soil Sci Soc Am J 68(4):1256–

1265.

Waring, S.A., and J.M. Bremner. 1964. Ammonium production in soil under

waterlogged conditions as an index of nitrogen availability. Nature 201: 951-952.

Wasilewska, L., and P. Bienkowski. 1985. Experimental study on the occurrence and

activity of soil nematodes in decomposition of plant material. Pedobiologia 28:41-57.

Watson C.A., C. Stopes and L. Philipps. 1997. Nitrogen cycling in organic managed

crop rotations: Importance of rotation design. From Resource use in organic farming –

proceedings of the third ENOF workshop: Ancona 5-6 June 1997

Weil, R.R., K.R. Islam. K, M.A. Stine, J.B. Gruver and S.E.S-Liebig. 2003. Estimating

active carbon for soil quality assessment: A simplified method for laboratory and field

use. American Journal of Alternative Agriculture. Volume 18, Number 1.

Whitehead A. G. and J. R. Hemming. 2008. A comparison of some quantitative methods

of extracting small vermiform nematodes from soil. Annals of Applied Biology. Vol.

55:1.

243

Wienhold, B.J., S.S. Andrews, and D.L. Karlen. 2004. Soil Quality: A review of the

science and experiences in the USA. Environmental Geochemistry Health. 26, 89-95.

Wilson, G.F., B.T. Kang, and K. Mulongoy. 1986. Alley cropping: trees as sources of

green manure and mulch in the tropics. Biological Agriculture and Horticulture 3, 251-

267.

Winsley, P. 2007. Biochar and bioenergy production for climate change mitigation. New

Zealand Science Review 64, 5-10.

Woolf, D., J.E. Amonette, F.A. Street-Perrott, J. Lehmann, and S. Joseph. 2010.

Sustainable biochar to mitigate global climate change. Nature Comm. 1, 56-64.

Woolhouse, H.W. 1969. Differences in the Properties of the Acid Phosphatase of Plant

Roots and Their Significance in the Evolution of Edaphic Ecosystem. In: I.H. Rovison.

(Ed.). Ecological Aspects of the Mineral Nutrition of Plants, Blackwell Scientific

Publications, Oxford England, pp: 357-380.

Wright, A.C.S. 1963. Soils and land use of Western Samoa. New Zealand Soils Bureau

Bulletin. 22.

Wycherley, P.R. 2003. The range of cover crops. Planters’ Bulletin. 68-117.

Yakovchenko, V., L.J. Sikora and D.D. Kaufma., 1996. A biologically based indicator

of soil quality. Biological Fertility of Soils. 21: 245-251.

Yamato, M., Y. Okimori, I.F. Wibowo, S. Anshori and M. Ogawa. 2006. Effect of the

application of charred bark of Acacia mangium on the yield of maize, cowpea and peanut,

and soil chemical properties in South Sumatra, Indonesia. Soil Science and Plant Nutrition

52, 489-495.

Yang, Y. and G. Sheng. 2003a. Enhanced pesticide sorption by soils containing particulate

matter from crop residue burns. Environmental Science and Technology. 37, 3635-3639.

244

Yang, Y. and G. Sheng. 2003b. Pesticide adsorptivity of aged particulate matter arusing

from crop residue burns. Journal of Agriculture and Food Chemistry. 5, 5047-5051.

Yeates, G.W., and D.A. Wardle. 1996. Nematodes as predators and prey: relationships

to biological control and soil processes. Pedobiologia 40:43-50.

Yeates, G.W., T. Bongers, R.G.M. De Goede, D.W. Freckman, and S.S. Georgieva. 1993.

Feeding habits in soil nematode families and genera–an outline for soil ecologists. Journal

of Nematology. 25: 315-331.

Yu, X.-Y., G.-G. Ying and, R.S. Kookana. 2009. Reduced plant uptake of pesticides with

biochar additions to soil. Chemosphere 76, 665-671.

Zantua, M.I. and J.M. Bremner. 1977. Stability of urease in soils. Soil Biology and

Biochemistry. 9(2): 135-140.

Zantua, M.I. and J.M. Bremner. 1976. Production and persistence of urease activity in

soils. Soil Biology and Biochemistry. 8(5): 369-374.

Zantua, M.I., L.C. Dumenil and J.M. Bremner. 1977. Relationship between soil urease

activity and other soil properties. Soil Science Society of America Journal. 41(2): 350-

352.

Zhang, W.L., Z.X. Tia, N. Zhang and, Q.X Li. 1996. Nitrate pollution of groundwater in

Northern China. Agriculture, Ecosystem Environment 59, 223-231.

Zhao, H., D.Q. Tong, Q. Lin, X. Liu and G. Wang. 2012. Effects of fires on soil organic

carbon pool and mineralisationin Northeastern China.

www.elsevier.com/locate/geoderma. Paper 424.

Zheng, W., M. Guo, T. Chow, D.N. Bennett, and N. Rajagopalan. 2010. Sorption

properties of greenwaste biochar from two trizaine pesticides. Journal of Hazardous

Materials. 181, 121-126.

245

APP

EN

DIC

ES

App

endi

x 1

Tri

al d

esig

n, la

yout

and

ran

dom

isat

ion

for

the

soil

heal

th fa

llow

exp

erim

ent

Mai

n Pl

ot (F

allo

w T

reat

men

ts)

C2

C

1

C2

C

1

C2

C

1

C1

C

2

C1

C

2

C1

C2

C1

C

2

C2

C

1

C1

C

2

C2

C

1

C1

C

2

C2

C1

C2

C

1

C2

C

1

C2

C

1

C1

C

2

C1

C

2

C1

C2

C1

C

2

C1

C

2

C1

C

2

C2

C

1

C2

C

1

C2

C1

6 m

1

m

BLO

CK

I

6

m

Split

plo

ts

(Cul

tivar

sT

5 T

1 T

2 T

3 T

6

BLO

CK

II

T

3 T

6 T

2 T

1 T

4

1 m

BLO

CK

II

I T

4 T

6 T

5 T

3 T

1

BLO

CK

IV

T

1 T

4 T

3 T

2 T

5

T 4

T 5 T 2

T 6

FALL

OW

TR

EATM

ENTS

T1 –

Con

trol (

Farm

er’s

Pra

ctic

e)

T2 –

Muc

una

T3 –

NFT

: Er

ythr

ina

T4 –

Muc

una

+ 0.

5 N

.P.K

T5 –

Far

mer

’s P

ract

ice

+ N

.P.K

T6 –

Bio

-cha

rcoa

l

TAR

O C

ULT

IVA

RS

C1

– Sa

moa

I

C2

– Sa

moa

II

Th

e ab

ove

layo

ut is

typ

ical

ran

dom

isat

ion

for

one

rese

arch

site

. For

eac

h of

the

four

res

earc

h si

tes

sepa

rate

ran

dom

isat

ion

sche

mes

wer

e fo

llow

ed. T

he ra

ndom

isat

ion

sche

mes

wer

e ge

nera

ted

usin

g G

enst

at st

atis

tical

softw

are

pack

age.

246

Appendix 2 Layout and randomisation of the soil incubation experiment

Block

I

T17 T23 T1 T16 T8 T21 T2 T15 T9 T4 T24 T13

T14 T3 T19 T5 T7 T12 T18 T11 T20 T10 T6 T22

Block

II

T24 T13 T2 T17 T6 T21 T19 T4 T15 T3 T8 T20

T10 T22 T18 T23 T5 T14 T11 T9 T16 T1 T12 T7

Block

III

T11 T23 T3 T1 T20 T2 T17 T18 T4 T16 T9 T15

T10 T22 T8 T13 T5 T14 T24 T19 T6 T21 T12 T7

247

Appendix 3 Typical rates of nutrient supplementation for the nutrient budgeting experiment

Solution No.

Element Rates of application of element

(kg/ha)

Compound Molecular weight

Weight conversion

factor –element to

salt

Rates of application

of saltkg/ha

Concentration of stock

solution (g salt/L)

1 N 100 NH4NO3 80.04 2.86 286.00 104.02

2 P 60 NaH2PO4.2H2O 178.00 5.75 346.00 125.60

3 K 80 KCl 78.56 2.01 161.00 58.60

4 Ca 35 CaCl2 112.00 2.79 98.00 35.70

5 Mg 30 MgCl2.6H2O 203.30 8.35 250.00 91.00

6 S 25 Na2SO4 142.00 4.42 111.00 40.40

7 Fe 5 Sequestrene 138 - 16.70 100.00 36.40

8 B 2 H3BO3 61.84 5.72 11.40 4.14

9 Zn 4 ZnCl2 136.30 2.08 8.34 3.02

10 Mn 5 MnCl2.4H2O 179.90 3.27 16.35 5.96

11 Cu 3 CuCl2.2H2O 170.50 2.68 8.04 2.92

12 Mo 0.4 [NH4]6Mo7O24.H2O 1236.00 12.88 5.15 1.87

13 Co 0.1 CoCl2.6H2O 237.95 4.04 0.404 0.15

14 Ni 0.1 NiCl2.6H2O 237.72 4.05 0.405 0.15

Source: Usher and Grundon, 2004

248

Appendix 4 Experimental layout and randomisation for the taro nutrient budget

(uptake and partitioning) experiment.

H1 H2 H4 H8 H6 H7 H5 H3

H3 H5 H7 H4 H2 H1 H8 H6

H8 H5 H2 H1 H3 H7 H4 H6

H6 H7 H8 H1 H5 H4 H3 H2

H6 H2 H1 H8 H3 H4 H7 H5

H5 H3 H6 H2 H8 H7 H4 H1

H3 H2 H4 H6 H8 H7 H5 H1

H1 H7 H2 H5 H3 H6 H4 H8

H7 H6 H3 H1 H4 H2 H8 H5

H3 H5 H1 H6 H7 H4 H8 H2

Taro cultivars: C1 – Samoa I; and C2 – Samoa II

Biomass harvests: H1 – 30 days after planting; H2 – 60 days after planting; H3 –

90 days after planting; H4 – 120 days after planting; H5 – 150 days after planting; H6 –

180 days after planting; H7 – 210 days after planting; H8 – 240 days after planting;

Split plots – Eight Randomised Biomass Harvests

1 m

1 m

BLOCK I Main Plots

(Taro Cultivars)

C1

C2

C2

C2 C1

BLOCK II

C2

BLOCK III

C2 C1

C1

BLOCK IV

C2 C2

C2

BLOCK V

C2 C1

249

Appendix 5 Life cycle of a taro plant

250

Appendix 6 Parts of a taro plant

251

Appendix 7 Index leaf of a taro plant

0 - Unopened leaf whorl

1 - Young, fully open leaf

2 - The indx leaf - (sampled for nutritional studies)

3 - Older leaf

252

Appendix 8 Labile carbon determination

The following equation was used to calculate the amount of Labile C:

C (g/kg) = [(M0-M1) x 26 x 9]

5 x 1000

M0= Initial concentration of KMnO4 (33 mM)

M1 = Concentration of KMnO4 (mM) after oxidation

(calculated from the calibration curve)

26 = Volume in each tube

9 = Conversion factor for 1 mMol of concentration of KMnO4 is reduced for every 9 mg of C

5 = Weight of the soil

1000 = Conversion to kg of soil

253

Appendix 9 Fluorescein diacetate hydrolysis activity

254

Calculation of the Rate of Soil Biological Activity

The concentration of fluorescein in the solution is converted to mg of FDA

hydrolysed/kg soil/hour using the following equation:

FDA hydrolysed/kg soil/hour = Fluorescein conc. (µg/ml) x 40.2 x 2 x 200

1000

Where 40.2 = total volume of solutions in tubes

2 = conversion to 1 hour from 30 minutes

200 = conversion to 1kg of soil from 5g

1,000 = conversion to mg from µg of fluorescein.

Molecular structure of fluorescein and fluorescein diacetate

Fluorescein C20H12O5 Fluorescein diacetate C24H16O7

MW: 332.31 MW: 416.38

Colour Change

255

Appendix 10 Nematode community analysis procedure

256

APPENDIX 11

REPEATED MEASURES ANALYSIS FOR SOIL LABILE C FOR FIELD

TRIAL

REML variance components analysis for labile C for Salani site

Response variate: Labile_C Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 176

Plot.Time used as residual term with covariance structure as below

Covariance structures defined within terms:

Term Factor Model Order No. rows Plot.Time Plot Identity 0 16 Time Uniform (het) 1 11

Tests for fixed effects

Sequentially adding terms to fixed model

Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 1188.66 10 105.45 43.6 <0.001 Fallow 26.39 3 8.80 18.0 <0.001 Time.Fallow 41.04 30 1.20 58.5 0.272

Dropping individual terms from full fixed model

Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 41.04 30 1.20 58.5 0.272 Table of effects for Constant

621.2 Standard error: 42.26

Table of effects for Time

Time 1 2 3 4 5 6 7 8 0.0 527.1 257.3 279.8 721.3 297.6 593.6 513.4

Time 9 10 11 631.4 631.6 551.9

Standard errors of differences

Average: 79.39 Maximum: 178.0 Minimum: 39.50

Average variance of differences: 8097.

257

Table of effects for Fallow

Fallow 1 2 3 4 0.00 80.50 -32.25 13.75

Standard error of differences: 59.76

Table of effects for Time.Fallow

Fallow 1 2 3 4 Time 1 0.00 0.00 0.00 0.00 2 0.00 -18.00 65.12 28.87 3 0.00 -66.63 -26.50 -65.75 4 0.00 373.87 331.00 253.00 5 0.00 42.00 53.50 26.75 6 0.00 -189.88 75.12 -19.13 7 0.00 115.50 163.87 72.62 8 0.00 58.62 116.87 -2.63 9 0.00 -29.00 -68.38 -90.13 10 0.00 -53.50 -29.13 -119.38 11 0.00 -26.63 -17.88 -17.88




Table of predicted means for Constant

1113. Standard error: 12.6 Table of predicted means for Time

Time 1 2 3 4 5 6 7 8 637 1183 854 1156 1389 901 1318 1193

Time 9 10 11 1221 1218 1173




Table of predicted means for Fallow

Fallow 1 2 3 4 1076 1176 1104 1096

258


Table of predicted means for Time.Fallow

Fallow 1 2 3 4 Time 1 621 702 589 635 2 1148 1211 1181 1191 3 879 892 820 827 4 901 1355 1200 1168 5 1343 1465 1364 1383 6 919 810 962 914 7 1215 1411 1347 1301 8 1135 1274 1219 1146 9 1253 1304 1152 1176 10 1253 1280 1192 1147 11 1173 1227 1123 1169




Standard error of differences for same level of factor:

Time Fallow

Average: 75.91 79.39

Maximum: 195.3 178.0

Minimum: 41.10 39.50

Average variance of differences: 8630. 8097.

259

REML variance components analysis for labile C for Safaatoa site



Sparse algorithm with AI optimisation

Covariance structures defined for random model










Time 1 2 3 4 5 6 7 8 0.0 419.6 25.5 389.4 595.5 158.2 474.0 407.6

Time 9 10 11 457.2 464.5 365.6





Fallow 1 2 3 4 0.00 -26.75 47.00 -47.00

260



Fallow 1 2 3 4 Time 1 0.00 0.00 0.00 0.00 2 0.00 21.00 -84.12 47.88 3 0.00 115.50 20.50 3.25 4 0.00 161.75 -15.37 -8.12 5 0.00 142.00 -13.50 -12.00 6 0.00 188.00 24.25 -6.75 7 0.00 182.13 -2.75 77.75 8 0.00 125.38 -46.12 0.38 9 0.00 154.13 41.50 28.00 10 0.00 -23.62 -37.25 -78.00 11 0.00 130.25 -32.37 89.13




Table of predicted means for Constant 1145. Standard error: 7.4 Table of predicted means for Time

Time 1 2 3 4 5 6 7 8 776 1192 836 1200 1401 986 1314 1204

Time 9 10 11 1289 1206 1188



Average variance of differences: 592.6

261

Table of predicted means for Fallow Fallow 1 2 3 4 1124 1206 1158 1090 Standard error of differences: 20.98 Table of predicted means for Time.Fallow Fallow 1 2 3 4 Time 1 783 756 830 736 2 1202 1197 1165 1203 3 808 897 876 764 4 1172 1307 1204 1117 5 1378 1494 1412 1319 6 941 1102 1012 887 7 1257 1412 1301 1288 8 1190 1289 1191 1144 9 1240 1367 1328 1221 10 1247 1197 1257 1122 11 1148 1252 1163 1191 Standard errors of differences Average: 49.08 Maximum: 74.01 Minimum: 25.39 Average variance of differences: 2543. Standard error of differences for same level of factor: Time Fallow Average: 48.21 47.49 Maximum: 74.01 68.19 Minimum: 25.39 26.10 Average variance of differences: 2595. 2371.

262

REML variance components analysis for Labile C for Siufaga site











Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 17.48 20 0.73 36.1 0.771

Table of effects for Constant



Time 1 2 3 4 5 6 7 8 0.0 9.5 -39.3 87.1 362.0 143.5 552.9 382.9

Time 9 10 11 425.4 358.7 358.2




263

Table of effects for Fallow Fallow 1 2 3 0.00 83.25 44.37



Fallow 1 2 3 Time 1 0.00 0.00 0.00 2 0.00 -58.00 -38.25 3 0.00 31.50 -29.50 4 0.00 22.75 14.63 5 0.00 15.63 -12.75 6 0.00 82.50 -49.75 7 0.00 -58.00 -32.87 8 0.00 5.75 -11.62 9 0.00 8.00 -22.37 10 0.00 -58.12 -36.75 11 0.00 -16.37 -17.75





1218. Standard error: 6.1

Table of predicted means for Time

Time 1 2 3 4 5 6 7 8 986 963 947 1085 1349 1140 1508 1367

Time 9 10 11 1406 1313 1333




264


Fallow 1 2 3 1183 1264 1206



Fallow 1 2 3 Time 1 943 1026 988 2 953 978 959 3 904 1019 919 4 1030 1136 1089 5 1305 1404 1337 6 1087 1252 1081 7 1496 1521 1508 8 1326 1415 1359 9 1368 1460 1390 10 1302 1327 1310 11 1301 1368 1328





Time Fallow

Average: 44.00 45.56

Maximum: 73.83 72.85

Minimum: 16.17 21.10


265

REML variance components analysis for Labile C for Aopo site




Covariance structures defined for random model Covariance structures defined within terms:

Term Factor Model Order No. rows Plot.Time Plot Identity 0 12 Time Uniform (het) 1 11 Tests for fixed effects





Table of effects for Constant

1425. Standard error: 28.6 Table of effects for Time

Time 1 2 3 4 5 6 7 8 0.00 57.50 -4.88 28.75 84.87 115.62 91.62 87.50

Time 9 10 11 129.00 115.25 108.00




266


Fallow 1 2 3 0.00 -27.88 -37.00



Fallow 1 2 3 Time 1 0.00 0.00 0.00 2 0.00 -9.00 -15.75 3 0.00 89.25 13.38 4 0.00 -4.50 -7.87 5 0.00 42.63 30.88 6 0.00 45.38 40.38 7 0.00 41.63 49.38 8 0.00 44.88 55.50 9 0.00 34.88 42.25 10 0.00 24.00 32.50 11 0.00 33.13 30.50







Time 1 2 3 4 5 6 7 8 1404 1453 1433 1428 1513 1548 1526 1525

Time 9 10 11 1558 1538 1533 Standard errors of differences




Fallow 1 2 3 1499 1502 1487


267


Fallow 1 2 3 Time 1 1425 1397 1388 2 1483 1446 1430 3 1420 1482 1397 4 1454 1422 1409 5 1510 1525 1504 6 1541 1558 1544 7 1517 1531 1529 8 1513 1530 1531 9 1554 1561 1559 10 1540 1537 1536 11 1533 1538 1527





Time Fallow

Average: 16.60 19.30

Maximum: 40.45 40.38

Minimum: 2.284 2.414

Average variance of differences: 476.8 488.4

268

APPENDIX 12

REPEATED MEASURES ANALYSIS FOR SOIL MICROBIAL (FDA)

ACTIVITY

REML variance components analysis for microbial activity for Salani site

Response variate: FDA Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 176


Sparse algorithm with AI optimisation Covariance structures defined for random model Covariance structures defined within terms:



Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 407.25 10 36.07 43.1 <0.001 Fallow 11.47 3 3.82 17.4 0.029 Time.Fallow 86.77 30 2.53 58.0 0.001





Time 1 2 3 4 5 6 7 8 0.00 17.53 15.34 68.79 28.39 -4.74 11.19 -2.74

Time 9 10 11 2.50 6.65 22.96




269


Fallow 1 2 3 4 0.000 4.275 -19.375 7.525



Fallow 1 2 3 4 Time 1 0.000 0.000 0.000 0.000 2 0.000 -3.863 20.625 3.625 3 0.000 10.687 19.037 -0.638 4 0.000 12.837 11.837 -21.763 5 0.000 23.762 30.837 1.037 6 0.000 14.562 30.837 -5.438 7 0.000 9.412 20.737 -4.513 8 0.000 9.250 17.162 -0.338 9 0.000 -8.788 17.850 -4.275 10 0.000 -25.525 15.100 -21.775 11 0.000 -10.938 10.287 -6.213





83.74 Standard error: 1.088 Table of predicted means for Time

Time 1 2 3 4 5 6 7 8 64.91 87.53 87.52 134.42 107.20 70.16 82.50 68.69

Time 9 10 11 68.60 63.51 86.15





Fallow 1 2 3 4 81.88 89.01 80.17 83.92

270



Fallow 1 2 3 4 Time 1 66.80 71.08 47.43 74.33 2 84.32 84.74 85.57 95.47 3 82.14 97.10 81.80 89.02 4 135.59 152.70 128.05 121.35 5 95.19 123.22 106.65 103.75 6 62.06 80.90 73.52 64.15 7 77.99 91.67 79.35 81.00 8 64.06 77.59 61.85 71.25 9 69.30 64.79 67.77 72.55 10 73.45 52.20 69.17 59.20 11 89.76 83.10 80.67 91.07 Standard errors of differences




Time Fallow

Average: 7.708 7.925

Maximum: 14.03 12.65

Minimum: 3.953 3.913


REML variance components analysis for soil microbial activity for Safaatoa site Response variate: FDA Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 176

Plot.Time used as residual term with covariance structure as below Sparse algorithm with AI optimisation Covariance structures defined for random model Covariance structures defined within terms: Term Factor Model Order No. rows Plot.Time Plot Identity 0 16 Time Uniform (het) 1 11

271

Tests for fixed effects Sequentially adding terms to fixed model

Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 1120.28 10 99.78 43.8 <0.001 Fallow 9.58 3 3.19 11.6 0.064 Time.Fallow 63.82 30 1.87 57.6 0.020 Dropping individual terms from full fixed model




Time 1 2 3 4 5 6 7 8 83.88 116.59 53.50 129.60 126.77 73.47 84.67 86.03

Time 9 10 11 79.43 82.82 112.08 Standard errors of differences Average: 3.470 Maximum: 4.349 Minimum: 2.425



Fallow 1 2 3 4 90.04 100.22 94.33 89.53



Fallow 1 2 3 4 Time 1 83.04 85.76 86.32 80.37 2 114.29 122.87 116.33 112.88 3 43.00 69.09 50.93 50.98 4 122.71 149.86 135.62 110.20 5 110.68 134.69 131.30 130.43 6 66.88 83.79 77.12 66.07 7 80.35 90.69 85.70 81.95 8 82.83 92.80 79.83 88.65 9 79.59 76.62 83.65 77.85 10 90.33 80.96 81.20 78.80 11 116.74 115.27 109.68 106.65

272

Standard errors of differences Average: 7.410 Maximum: 9.796 Minimum: 4.850



Time Fallow

Average: 7.476 6.941

Maximum: 9.796 8.698

Minimum: 5.046 4.850

Average variance of differences: 58.14 49.00 REML variance components analysis for soil microbial activity for Siufaga site

Response variate: FDA Fixed model: Constant + Time + Time.Fallow Random model: Plot.Time Number of units: 132


Sparse algorithm with AI optimisation Covariance structures defined for random model Covariance structures defined within terms:



Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 245.10 10 20.81 30.4 <0.001 Time.Fallow 92.05 22 3.35 34.0 <0.001


Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 92.05 22 3.35 34.0 <0.001 Table of predicted means for Constant


273


Time 1 2 3 4 5 6 7 8 83.26 63.12 78.19 95.19 102.98 104.24 80.08 77.55

Time 9 10 11 61.78 63.22 60.38





Fallow 1 2 3 Time 1 76.51 87.68 85.60 2 55.35 74.58 59.45 3 74.49 82.76 77.32 4 79.49 109.16 96.92 5 86.11 116.66 106.18 6 100.25 113.68 98.80 7 74.71 91.46 74.08 8 73.01 82.65 77.00 9 57.83 64.34 63.18 10 67.79 67.68 54.20 11 58.53 63.64 58.97 Standard errors of differences




Time Fallow

Average: 7.775 8.137

Maximum: 11.61 11.54

Minimum: 3.001 3.704


274

REML variance components analysis for soil microbial activity for Aopo site

Response variate: FDA Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 132







Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 50.74 20 2.13 37.0 0.023 Table of predicted means for Constant



Time 1 2 3 4 5 6 7 8 114.78 113.71 112.80 114.42 107.58 106.15 85.05 78.49

Time 9 10 11 69.62 60.39 87.14





Fallow 1 2 3 87.53 103.09 95.78 Standard error of differences: 4.764

275

Table of predicted means for Time. Fallow Fallow 1 2 3 Time 1 116.75 111.97 115.63 2 101.16 125.31 114.65 3 97.99 128.23 112.18 4 101.91 123.98 117.38 5 93.79 120.14 108.83 6 101.69 111.79 104.98 7 79.11 95.96 80.08 8 75.84 84.19 75.45 9 65.76 73.36 69.73 10 60.59 63.18 57.40 11 68.29 95.86 97.28 Standard errors of differences




Time Fallow

Average: 7.608 6.658

Maximum: 13.45 10.74

Minimum: 3.625 3.872

Average variance of differences:

66.05 47.69

276

APPENDIX 13

REPEATED MEASURES ANALYSIS FOR POTENTIALLY MINERALISABLE

N (PMN)

REML variance components analysis for PMN for Salani site

Response variate: PMN Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 176






Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 4202.72 10 371.24 42.6 <0.001 Fallow 24.48 3 8.16 19.0 0.001 Time.Fallow 91.72 30 2.67 57.6 <0.001


Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 91.72 30 2.67 57.6 <0.001



Time 1 2 3 4 5 6 7 8 47.81 19.06 13.23 29.74 52.05 108.50 100.41 102.53

Time 9 10 11 113.09 100.69 93.31




277


Fallow 1 2 3 4 66.71 83.93 63.55 69.60



Fallow 1 2 3 4 Time 1 43.75 57.50 40.00 50.00 2 21.25 20.00 15.00 20.00 3 17.50 12.69 8.75 14.00 4 26.84 35.44 28.70 27.98 5 50.75 60.81 51.13 45.50 6 102.38 126.00 101.50 104.12 7 92.75 135.62 91.00 82.25 8 100.25 119.62 88.50 101.75 9 106.75 126.87 94.50 124.25 10 90.63 109.87 100.25 102.00 11 81.00 118.75 79.75 93.75





Time Fallow

Average: 9.390 9.903

Maximum: 20.03 17.85

Minimum: 2.895 2.936


278

REML variance components analysis for PMN for Safaatoa site





Covariance structures defined within terms: Term Factor Model Order No. rows Plot.Time Plot Identity 0 16 Time Uniform (het) 1 11



Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 2527.06 10 222.29 41.7 <0.001 Fallow 25.88 3 8.63 19.7 <0.001 Time.Fallow 90.74 30 2.63 56.7 <0.001






Time 1 2 3 4 5 6 7 8 21.09 17.19 17.20 33.57 50.86 84.98 121.84 100.00

Time 9 10 11 133.66 109.16 122.06





Fallow 1 2 3 4 67.20 88.13 76.13 63.68


279


Fallow 1 2 3 4 Time 1 20.63 22.50 22.50 18.75 2 15.00 16.25 22.50 15.00 3 15.75 19.80 15.75 17.50 4 24.09 42.42 29.40 38.37 5 44.19 65.62 49.87 43.75 6 77.00 107.19 83.12 72.62 7 104.13 152.25 122.50 108.50 8 92.75 119.00 108.50 79.75 9 123.38 150.50 141.75 119.00 10 108.50 128.62 119.00 80.50 11 113.75 145.25 122.50 106.75





Time Fallow

Average: 9.831 10.15

Maximum: 19.28 17.04

Minimum: 2.324 3.054

Average variance of differences: 125.5 115.7 REML variance components analysis for PMN for Siufaga site







280









Time 1 2 3 4 5 6 7 8 24.58 15.77 29.76 43.75 86.64 74.37 159.78 131.75

Time 9 10 11 136.25 159.22 117.28





Fallow 1 2 3 74.82 108.32 83.91



Fallow 1 2 3 Time 1 25.00 16.25 32.50 2 16.64 17.16 13.53 3 28.22 32.95 28.10 4 38.46 47.92 44.85 5 76.54 101.27 82.10 6 56.62 101.00 65.48 7 131.89 197.87 149.58 8 94.77 173.14 127.35 9 112.66 161.35 134.75 10 146.24 189.22 142.20 11 95.92 153.34 102.58

281





Time Fallow

Average: 10.02 9.491

Maximum: 16.71 13.98

Minimum: 1.683 3.081

Average variance of differences: 123.5 97.07 REML variance components analysis for PMN for Aopo site









Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 2054.53 10 172.59 29.6 <0.001 Fallow 25.45 2 12.73 22.5 <0.001 Time.Fallow 101.04 20 4.19 36.3 <0.001


Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 101.04 20 4.19 36.3 <0.001 Table of predicted means for Constant


282


Time 1 2 3 4 5 6 7 8 24.17 20.30 33.51 38.38 56.63 76.40 157.31 120.10

Time 9 10 11 204.04 186.84 167.80





Fallow 1 2 3 85.62 121.06 89.35



Fallow 1 2 3 Time 1 26.25 28.75 17.50 2 18.67 20.09 22.13 3 27.02 37.51 36.00 4 36.21 44.21 34.70 5 50.06 65.76 54.08 6 55.84 102.78 70.58 7 142.56 188.20 141.18 8 91.61 162.21 106.48 9 171.36 255.06 185.70 10 168.02 216.66 175.83 11 154.24 210.48 138.70 Standard errors of differences




Time Fallow

Average: 11.01 11.63

Maximum: 26.64 21.23

Minimum: 2.390 2.360


181.4 163.1

283

APPENDIX 14

REPEATED MEASURES ANALYSIS FOR AMMONIUM - N (NH4+ - N)

REML variance components analysis for ammonium N for Salani Response variate: Ammonium Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 128 Plot.Time used as residual term with covariance structure as below Covariance structures defined within terms: Term Factor Model Order No. rows Plot.Time Plot Identity 0 16 Time Uniform (het) 1 8 Tests for fixed effects Sequentially adding terms to fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 198.72 7 25.60 33.5 <0.001 Fallow 92.07 3 30.69 12.9 <0.001 Time.Fallow 48.90 21 2.07 42.9 0.022 Dropping individual terms from full fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 48.90 21 2.07 42.9 0.022 Table of predicted means for Constant

42.61 Standard error: 0.847 Table of predicted means for Time Time 1 2 3 4 5 6 7 8 27.02 50.31 33.69 46.70 36.97 51.24 46.70 48.23 Standard errors of differences Average: 3.552 Maximum: 5.031 Minimum: 1.687 Average variance of differences: 13.27

284

Table of predicted means for Fallow Fallow 1 2 3 4 33.82 56.74 38.09 41.78 Standard error of differences: 2.396 Table of predicted means for Time.Fallow Fallow 1 2 3 4 Time 1 25.16 29.75 28.66 24.50 2 42.00 68.25 41.13 49.88 3 27.78 47.03 30.63 29.31 4 26.25 76.56 35.88 48.13 5 22.31 57.31 30.63 37.63 6 46.59 58.62 48.56 51.19 7 42.88 54.69 42.88 46.38 8 37.62 61.69 46.38 47.25 Standard errors of differences Average: 7.032 Maximum: 10.06 Minimum: 3.270 Average variance of differences: 52.39 Standard error of differences for same level of factor: Time Fallow Average: 6.782 7.104 Maximum: 10.01 10.06 Minimum: 3.270 3.374 Average variance of differences: 52.18 53.08

285

REML variance components analysis for ammonium N for Safaatoa site Response variate: Ammonium Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 128 Plot.Time used as residual term with covariance structure as below Sparse algorithm with AI optimisation Covariance structures defined for random model Covariance structures defined within terms: Term Factor Model Order No. rows Plot.Time Plot Identity 0 16 Time Uniform (het) 1 8 Tests for fixed effects Sequentially adding terms to fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 124.84 7 16.09 33.1 <0.001 Fallow 79.59 3 26.53 10.6 <0.001 Time.Fallow 13.83 21 0.59 41.8 0.905 Dropping individual terms from full fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 13.83 21 0.59 41.8 0.905 Table of predicted means for Constant 38.78 Standard error: 1.005 Table of predicted means for Time

Time 1 2 3 4 5 6 7 8 26.14 28.11 27.40 68.58 36.53 38.72 42.44 42.33 Standard errors of differences

Average: 3.552 Maximum: 4.939 Minimum: 2.453 Average variance of differences: 13.11

286

Table of predicted means for Fallow Fallow 1 2 3 4 31.96 52.66 35.77 34.73 Standard error of differences: 2.843 Table of predicted means for Time.Fallow Fallow 1 2 3 4 Time 1 22.75 36.31 20.12 25.37 2 19.69 43.75 28.00 21.00 3 22.97 39.81 23.63 23.19 4 55.13 87.06 64.75 67.38 5 26.69 55.56 29.75 34.13 6 38.50 50.75 37.62 28.00 7 34.13 55.12 41.12 39.38 8 35.88 52.94 41.12 39.38 Standard errors of differences Average: 7.170 Maximum: 11.69 Minimum: 4.623 Average variance of differences: 53.64 Standard error of differences for same level of factor: Time Fallow Average: 7.068 7.103 Maximum: 11.69 9.879 Minimum: 4.623 4.906 Average variance of differences: 53.98 52.45

287

REML variance components analysis for ammonium – N for Siufaga site Response variate: Ammonium Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 96 Plot.Time used as residual term with covariance structure as below Sparse algorithm with AI optimisation Covariance structures defined for random model Covariance structures defined within terms: Term Factor Model Order No. rows Plot.Time Plot Identity 0 12 Time Uniform (het) 1 8 Tests for fixed effects Sequentially adding terms to fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 614.47 7 77.41 24.5 <0.001 Fallow 22.02 2 11.01 6.1 0.010 Time.Fallow 31.85 14 1.98 27.9 0.060 Dropping individual terms from full fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 31.85 14 1.98 27.9 0.060 Table of predicted means for Constant 55.64 Standard error: 1.871 Table of predicted means for Time Time 1 2 3 4 5 6 7 8 15.02 35.44 59.94 78.02 41.42 73.35 70.73 71.17 Standard errors of differences Average: 3.709 Maximum: 4.315 Minimum: 3.016 Average variance of differences: 13.87

288

Table of predicted means for Fallow Fallow 1 2 3 45.12 69.56 52.23 Standard error of differences: 4.583 Table of predicted means for Time.Fallow Fallow 1 2 3 Time 1 14.87 17.50 12.69 2 27.56 49.00 29.75 3 50.31 78.75 50.75 4 62.12 98.44 73.50 5 33.69 54.69 35.88 6 64.75 86.19 69.13 7 54.69 84.88 72.63 8 52.94 87.06 73.50 Standard errors of differences Average: 7.200 Maximum: 8.901 Minimum: 5.225 Average variance of differences: 52.61 Standard error of differences for same level of factor: Time Fallow Average: 7.501 6.424 Maximum: 8.901 7.473 Minimum: 5.765 5.225 Average variance of differences: 57.42 41.61

289

REML variance components analysis for ammonium – N for Aopo site Response variate: Ammonium Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 96 Plot.Time used as residual term with covariance structure as below Sparse algorithm with AI optimisation Covariance structures defined for random model Covariance structures defined within terms: Term Factor Model Order No. rows Plot.Time Plot Identity 0 12 Time Uniform (het) 1 8 Tests for fixed effects Sequentially adding terms to fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time 256.47 7 31.78 23.9 <0.001 Fallow 47.58 2 23.79 11.7 <0.001 Time.Fallow 14.07 14 0.86 28.2 0.606 Dropping individual terms from full fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 14.07 14 0.86 28.2 0.606 Table of predicted means for Constant 60.59 Standard error: 1.830 Table of predicted means for Time Time 1 2 3 4 5 6 7 8 36.60 38.06 88.37 59.35 76.27 54.83 65.48 65.77 Standard errors of differences Average: 5.303 Maximum: 8.566 Minimum: 2.775 Average variance of differences: 31.45

290

Table of predicted means for Fallow Fallow 1 2 3 49.93 74.43 57.42 Standard error of differences: 4.481 Table of predicted means for Time.Fallow Fallow 1 2 3 Time 1 33.69 41.13 35.00 2 29.75 47.69 36.75 3 70.87 111.13 83.13 4 46.81 70.00 61.25 5 65.62 93.19 70.00 6 45.06 66.94 52.50 7 55.12 85.31 56.00 8 52.50 80.06 64.75 Standard errors of differences Average: 9.426 Maximum: 18.88 Minimum: 4.622 Average variance of differences: 100.1

Standard error of differences for same level of factor: Time Fallow Average: 9.098 9.185

Maximum: 18.88 14.84

Minimum: 4.622 4.806


291

APPENDIX 15

REPEATED MEASURES ANALYSIS FOR NITRATE - N (NO3- - N)

REML variance components analysis for nitrate – N for Salani site

Response variate: Nitrate Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 128














Time 1 2 3 4 5 6 7 8 233.5 460.9 218.8 1027.8 839.0 921.8 843.5 919.8 Standard errors of differences



292


Fallow 1 2 3 4 595.5 925.1 531.1 680.9 Standard error of differences: 69.80


Fallow 1 2 3 4 Time 1 197.1 279.0 198.2 259.9 2 395.5 685.1 379.8 383.2 3 206.5 323.8 164.9 179.8 4 901.2 1325.2 818.1 1066.6 5 752.1 1143.6 622.1 838.3 6 789.2 1265.2 709.6 923.1 7 720.1 1156.8 626.5 870.6 8 801.9 1221.9 729.8 925.8




Standard error of differences for same level of factor: Time Fallow

Average: 153.9 165.1

Maximum: 338.6 292.3

Minimum: 36.70 41.34

Average variance of differences: 32417. 31480. REML variance components analysis for nitrate – N for Safaatoa site





293










Time 1 2 3 4 5 6 7 8 268.8 485.3 286.6 302.6 910.1 578.9 603.7 616.4




Table of predicted means for Fallow Fallow 1 2 3 4 412.9 653.3 485.2 474.9


Table of predicted means for Time.Fallow Fallow 1 2 3 4 Time 1 165.8 394.2 239.8 275.6 2 458.1 548.6 537.3 397.3 3 171.1 405.8 340.4 229.3 4 204.8 416.9 266.0 322.9 5 842.2 1026.4 798.0 973.9

294

6 470.3 782.3 502.3 560.9 7 473.4 832.1 577.5 532.0 8 518.0 819.9 620.4 507.5 Standard errors of differences




Time Fallow

Average: 134.0 129.3

Maximum: 182.6 166.1

Minimum: 75.72 76.48 Average variance of differences: 19585. 17302. REML variance components analysis for nitrate – N for Siufaga site












295



Time 1 2 3 4 5 6 7 8 523.6 362.5 1059.5 965.4 1032.6 1120.6 1150.0 1171.0





Fallow 1 2 3 719.8 1160.3 889.4



Fallow 1 2 3 Time 1 443.0 644.0 483.9 2 282.6 546.9 258.1 3 857.1 1281.0 1040.4 4 511.4 1326.9 1057.9 5 853.1 1222.8 1022.0 6 920.9 1334.8 1106.0 7 934.5 1452.5 1063.1 8 955.5 1473.5 1084.1





Time Fallow

Average: 183.0 189.2

Maximum: 287.3 257.6

Minimum: 84.56 107.6


296

REML variance components analysis for nitrate –N for Aopo site












Time 1 2 3 4 5 6 7 8 577 373 1537 1563 1244 1235 1277 1285




297


Fallow 1 2 3 968 1341 1100



Fallow 1 2 3 Time 1 529 620 584 2 248 448 422 3 1217 1760 1635 4 1324 1898 1467 5 1236 1349 1148 6 1070 1540 1096 7 1048 1555 1227 8 1075 1562 1218





Time Fallow

Average: 157.3 151.5

Maximum: 269.9 220.2

Minimum: 64.74 57.85


298

APPENDIX 16

REPEATED MEASURES ANALYSIS FOR NET CUMULATIVE N

MINERALISATION

REML variance components analysis for net cumulative N mineralisation for Salani site

Response variate: Cumulative_Mineral_N Fixed model: Constant + Time + Fallow + Time.Fallow Random model: Plot.Time Number of units: 112













Time 1 2 3 4 5 6 7 251 510 1376 1984 2335 2572 2800

299





Fallow 1 2 3 4 1416 2390 1458 1494



Fallow 1 2 3 4 Time 1 215 445 194 149 2 418 827 422 373 3 1232 1858 1137 1278 4 1576 2790 1699 1869 5 1960 3270 2013 2096 6 2206 3621 2207 2253 7 2307 3919 2536 2439





Time Fallow

Average: 279.9 256.3

Maximum: 334.2 293.2

Minimum: 125.9 147.6


300

REML variance components analysis for net cumulative N mineralisation for Safaatoa site















Time 1 2 3 4 5 6 7 315 516 668 1244 1605 1784 1969




301


Fallow 1 2 3 4 1246 1131 1156 1096



Fallow 1 2 3 4 Time 1 337 339 377 207 2 621 491 578 372 3 739 656 752 526 4 1348 1234 1249 1144 5 1708 1554 1537 1620 6 1871 1736 1726 1803 7 2097 1908 1873 1998





Time Fallow

Average: 204.7 158.9

Maximum: 287.3 201.3

Minimum: 133.2 111.1


44900. 26034. REML variance components analysis for net cumulative N mineralisation for Siufaga site




302



Term Factor Model Order No. rows Plot.Time Plot Identity 0 12 Time Uniform (het) 1 7 Residual variance model

Term Factor Model(order) Parameter Estimate s.e. Plot.Time Sigma2 1.000 fixed Plot Identity - - - Time Uniform het theta1 0.3464 0.1332 Scale row 1 38078. 20799. Scale row 2 161014. 78719. Scale row 3 83433. 32663. Scale row 4 152778. 54517. Scale row 5 155864. 55034. Scale row 6 151273. 54314. Scale row 7 153053. 54821. Tests for fixed effects




Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Time.Fallow 18.07 12 1.36 32.7 0.234 Table of predicted means for Constant 1505. Standard error: 67.3 Table of predicted means for Time

Time 1 2 3 4 5 6 7 235 945 1363 1735 1912 2160 2182



303



Fallow 1 2 3 1249 1460 1805



Fallow 1 2 3 Time 1 135 299 271 2 698 1063 1074 3 1032 1380 1678 4 1529 1609 2068 5 1653 1838 2245 6 1838 2004 2639 7 1857 2027 2661





Time Fallow

Average: 247.6 206.0

Maximum: 283.7 227.6

Minimum: 138.0 143.6

Average variance of differences: 63964. 42880. REML variance components analysis for net cumulative N mineralisation for Aopo site




304












Time 1 2 3 4 5 6 7 215 1430 1685 2033 2300 2562 2784 Standard errors of differences




Fallow 1 2 3 1505 2183 1888



Fallow 1 2 3 Time 1 284 165 196 2 1293 1540 1456 3 1537 1855 1664 4 1644 2381 2074 5 1831 2703 2367 6 1934 3173 2580 7 2011 3464 2877

305




Standard error of differences for same level of factor: Time Fallow

Average: 216.9 176.9

Maximum: 293.3 216.2

Minimum: 70.77 111.3


306

APPENDIX 17

NESTED CLASSIFICATION ANALYSIS OF VARIANCE FOR BETWEEN

SITE COMPARISONS OF BIOCHEMICAL PARAMETERS

Analysis of an unbalanced design using GenStat regression for labile C Variate: Labile_C Accumulated analysis of variance Change d.f. s.s. m.s. v.r. F pr. + Block 3 53710. 17903. 0.46 0.713 + Site 3 13050358. 4350119. 110.70 <.001 + Site.Fallow 10 731088. 73109. 1.86 0.048 Residual 599 23537865. 39295. Total 615 37373021. 60769. Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations. Response variate: Labile_C Prediction se Site 1 1115 15.39 2 1153 15.39 3 * * 4 * * Standard error of differences between predicted means 21.77 Least significant difference (at 5.0%) for predicted means 42.75 Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations.

307

Response variate: Labile_C Fallow 1 2 Prediction se Prediction se Site 1 1076 29.88 1176 29.88 2 1124 29.88 1206 29.88 3 1183 29.88 1264 29.88 4 1499 29.88 1502 29.88 Fallow 3 4 Prediction se Prediction se Site 1 1104 29.88 1096 29.88 2 1158 29.88 1090 29.88 3 1206 29.88 * * 4 1487 29.88 * * Standard error of differences between predicted means 42.26 Least significant difference (at 5.0%) for predicted means 83.00 Analysis of an unbalanced design using GenStat regression for soil microbial activity (FDA) Variate: FDA Accumulated analysis of variance Change d.f. s.s. m.s. v.r. F pr. + Block 3 1185.4 395.1 0.73 0.537 + Site 3 26400.9 8800.3 16.16 <.001 + Site.Fallow 10 14788.3 1478.8 2.72 0.003 Residual 599 326130.4 544.5 Total 615 368505.1 599.2 Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations.

308

Response variate: FDA Prediction se Site 1 83.72 1.812 2 94.10 1.812 3 * * 4 * * Standard error of differences between predicted means 2.562 Least significant difference (at 5.0%) for predicted means 5.032 Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations. Response variate: FDA Fallow 1 2 Prediction se Prediction se Site 1 81.88 3.518 89.01 3.518 2 90.04 3.518 100.22 3.518 3 73.10 3.518 86.75 3.518 4 87.53 3.518 103.09 3.518 Fallow 3 4 Prediction se Prediction se Site 1 80.17 3.518 83.92 3.518 2 94.33 3.518 89.53 3.518 3 77.43 3.518 * * 4 95.78 3.518 * * Standard error of differences between predicted means 4.975 Least significant difference (at 5.0%) for predicted means 9.770

309

Analysis of an unbalanced design using GenStat regression for potentially mineralisable N Variate: PMN Accumulated analysis of variance Change d.f. s.s. m.s. v.r. F pr. + Block 3 2283. 761. 0.26 0.852 + Site 3 76486. 25495. 8.82 <.001 + Site.Fallow 10 86172. 8617. 2.98 0.001 Residual 599 1730590. 2889. Total 615 1895530. 3082. Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations. Response variate: PMN Prediction se Site 1 71.14 4.174 2 75.23 4.174 3 * * 4 * * Standard error of differences between predicted means 5.903 Least significant difference (at 5.0%) for predicted means 11.59 Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations.

310

Response variate: PMN Fallow 1 2 Prediction se Prediction se Site 1 66.71 8.103 83.93 8.103 2 67.20 8.103 88.13 8.103 3 74.82 8.103 108.32 8.103 4 85.62 8.103 121.06 8.103 Fallow 3 4 Prediction se Prediction se Site 1 63.55 8.103 69.60 8.103 2 76.13 8.103 63.68 8.103 3 83.91 8.103 * * 4 89.35 8.103 * * Standard error of differences between predicted means 11.46 Least significant difference (at 5.0%) for predicted means 22.51 Analysis of an unbalanced design using GenStat regression for ammonium – N Variate: Ammonium_N Accumulated analysis of variance Change d.f. s.s. m.s. v.r. F pr. + Block 3 2150.4 716.8 2.06 0.105 + Site 3 35412.4 11804.1 33.95 <.001 + Site.Fallow 10 38210.6 3821.1 10.99 <.001 Residual 431 149856.2 347.7 Total 447 225629.5 504.8 Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations.

311

Response variate: Ammonium_N Prediction se Site 1 42.73 1.698 2 39.36 1.698 3 * * 4 * * Standard error of differences between predicted means 2.401 Least significant difference (at 5.0%) for predicted means 4.719 Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations. Response variate: Ammonium_N Fallow 1 2 Prediction se Prediction se Site 1 33.82 3.296 56.74 3.296 2 31.96 3.296 52.66 3.296 3 45.12 3.296 69.56 3.296 4 49.93 3.296 74.43 3.296 Fallow 3 4 Prediction se Prediction se Site 1 38.09 3.296 41.78 3.296 2 35.77 3.296 34.73 3.296 3 52.23 3.296 * * 4 57.42 3.296 * * Standard error of differences between predicted means 4.662 Least significant difference (at 5.0%) for predicted means 9.162

312

Analysis of an unbalanced design using GenStat regression for nitrate - N Variate: Nitrate_N Accumulated analysis of variance Change d.f. s.s. m.s. v.r. F pr. + Block 3 2150810. 716937. 4.63 0.003 + Site 3 24934644. 8311548. 53.63 <.001 + Site.Fallow 10 9326416. 932642. 6.02 <.001 Residual 431 66797084. 154982. Total 447 103208953. 230893. Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations. Response variate: Nitrate_N Prediction se Site 1 683.5 35.85 2 511.1 35.85 3 * * 4 * * Standard error of differences between predicted means 50.69 Least significant difference (at 5.0%) for predicted means 99.64 Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations. Response variate: Nitrate_N Fallow 1 2 Prediction se Prediction se Site 1 595.5 69.59 925.1 69.59 2 412.9 69.59 653.3 69.59 3 719.8 69.59 1160.3 69.59 4 968.4 69.59 1341.4 69.59

313

Fallow 3 4 Prediction se Prediction se Site 1 531.1 69.59 680.9 69.59 2 485.2 69.59 474.9 69.59 3 889.4 69.59 * * 4 1099.5 69.59 * * Standard error of differences between predicted means 98.42 Least significant difference (at 5.0%) for predicted means 193.4 Analysis of an unbalanced design using GenStat regression for cumulative net N mineralisation Variate: Cumulative_N Accumulated analysis of variance Change d.f. s.s. m.s. v.r. F pr. + Block 3 10711200. 3570400. 3.97 0.008 + Site 3 27531117. 9177039. 10.21 <.001 + Site.Fallow 10 29631726. 2963173. 3.30 <.001 Residual 375 336984452. 898625. Total 391 404858495. 1035444. Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations. Response variate: Cumulative_N Prediction se Site 1 1718 92.27 2 1166 92.27 3 * * 4 * * Standard error of differences between predicted means 130.5 Least significant difference (at 5.0%) for predicted means 256.6

314

Predictions from regression model The standard errors are appropriate for interpretation of the predictions as summaries of the data rather than as forecasts of new observations. Response variate: Cumulative_N Fallow 1 2 Prediction se Prediction se Site 1 1416 179.1 2390 179.1 2 1246 179.1 1131 179.1 3 1249 179.1 1460 179.1 4 1505 179.1 2183 179.1 Fallow 3 4 Prediction se Prediction se Site 1 1458 179.1 1494 179.1 2 1156 179.1 1096 179.1 3 1805 179.1 * * 4 1888 179.1 * * Standard error of differences between predicted means 253.4 Least significant difference (at 5.0%) for predicted means 498.2

315

APPENDIX 18

CORRELATION ANALYSIS FOR TEST OF ASSOCIATION BETWEEN SOIL

BIOCHEMICAL PROPERTIES

Correlations FDA Labile_C 0.2376 FDA Labile_C Number of observations: 616 Two-sided test of correlations different from zero probabilities FDA Labile_C 0.0000 FDA Labile_C Correlations PMN FDA -0.3103 PMN FDA Number of observations: 616 Two-sided test of correlations different from zero probabilities PMN FDA 0.0000 PMN FDA Correlations

Labile_C PMN 0.4606 Labile_C PMN

Number of observations: 616

Two-sided test of correlations different from zero probabilities

Labile_C PMN 0.0000 Labile_C PMN

316

Correlations Amm_N Labile_C 0.4692 Amm_N Labile_C

Number of observations: 448 Two-sided test of correlations different from zero probabilities

Amm_N Labile_C 0.0000 Amm_N Labile_C Correlations Labile_C Nitr_N 0.4400 Labile_C Nitr_N Number of observations: 448

Two-sided test of correlations different from zero probabilities Labile_C Nitr_N 0.0000 Labile_C Nitr_N

Correlations

FDA Amm_N -0.2035 FDA Amm_N


Two-sided test of correlations different from zero probabilities

FDA Amm_N 0.0000 FDA Amm_N

317

Correlations

FDA Nitr_N -0.2767 FDA Nitr_N


Two-sided test of correlations different from zero probabilities FDA Nitr_N 0.0000 FDA Nitr_N Correlations

PMN Amm_N 0.5613 PMN Amm_N Number of observations: 448 Two-sided test of correlations different from zero probabilities PMN Amm_N 0.0000 PMN Amm_N

Correlations

PMN Nitr_N 0.5505 PMN Nitr_N


Two-sided test of correlations different from zero pobabilities PMN Nitr_N 0.0000 PMN Nitr_N Correlations Amm_N Nitr_N 0.5800 Amm_N Nitr_N


Two-sided test of correlations different from zero probabilities Amm_N Nitr_N 0.0000 Amm_N Nitr_N

318

APPENDIX 19

NESTED CLASSIFICATION ANALYSIS OF VARIANCE FOR TARO YIELD

Analysis of an unbalanced design using GenStat regression Variate: Yield Accumulated analysis of variance Change d.f. s.s. m.s. v.r. F pr. + Block 3 4.604 1.535 0.46 0.712 + Site 2 219.938 109.969 32.80 <.001 + Site.Treat 14 300.886 21.492 6.41 <.001 + Site.Treat.Cultivar 17 180.565 10.621 3.17 <.001 Residual 99 331.881 3.352 Total 135 1037.874 7.688 Predictions from regression model Response variate: Yield Prediction Site 1 10.764 2 8.532 3 * Least significant difference (at 5.0%) for predicted means 0.7480 Predictions from regression model Response variate: Yield Prediction Treat 1 2 3 4 5 Site 1 8.553 11.652 9.725 13.573 10.520 2 6.871 7.616 8.820 10.235 9.821 3 5.583 9.052 6.514 10.100 7.587 Treat 6 Site 1 10.456 2 7.473 3 *

319

Least significant difference (at 5.0%) for predicted means 1.816 Predictions from regression model Response variate: Yield Prediction Cultivar 1 2 Site Treat 1 1 7.174 9.933 2 10.147 13.157 3 9.577 9.874 4 12.385 14.762 5 10.048 10.992 6 8.985 11.926 2 1 5.704 8.039 2 6.140 9.091 3 7.545 10.095 4 9.070 11.399 5 9.188 10.454 6 7.369 7.578 3 1 5.509 5.656 2 7.421 10.684 3 6.253 6.774 4 7.906 12.294 5 7.900 7.274 6 * * Least significant difference (at 5.0%) for predicted means: 2.569

320

APPENDIX 20

REPEATED MEASURES ANALYSIS FOR THE SOIL INCUBATION EXPERIMENT

REML variance components analysis for labile C

Response variate: Labile_C Fixed model: Constant + Soil + Fallow + Rate + Soil.Fallow + Soil.Rate + Fallow.Rate + Soil.Fallow.Rate Random model: Pot.Time Number of units: 288

Pot.Time used as residual term with covariance structure as below




Term Factor Model Order No. rows Pot.Time Pot Identity 0 72 Time Antedependence 2 4



Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil 21.92 1 21.92 93.2 <0.001 Fallow 184.29 3 61.43 93.2 <0.001 Rate 0.04 2 0.02 93.2 0.979 Soil.Fallow 30.74 3 10.25 93.2 <0.001 Soil.Rate 0.03 2 0.02 93.2 0.983 Fallow.Rate 60.22 6 10.04 93.2 <0.001 Soil.Fallow.Rate 10.86 6 1.81 93.2 0.106


Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil.Fallow.Rate 10.86 6 1.81 93.2 0.106 Table of predicted means for Constant


Table of predicted means for Soil

321

Soil 1 2 1318 1294



Fallow 1 2 3 4 1303 1334 1338 1249


Table of predicted means for Rate

Rate 1 2 3 1307 1305 1305


Table of predicted means for Soil.Fallow

Fallow 1 2 3 4 Soil 1 1307 1329 1353 1284 2 1299 1339 1323 1214


Table of predicted means for Soil.Rate

Rate 1 2 3 Soil 1 1320 1317 1318 2 1295 1293 1293


Table of predicted means for Fallow.Rate

Rate 1 2 3 Fallow 1 1297 1300 1312 2 1318 1347 1335 3 1316 1343 1355 4 1298 1229 1220


322

Table of predicted means for Soil.Fallow.Rate

Rate 1 2 3 Soil Fallow 1 1 1317 1295 1308 2 1318 1341 1326 3 1327 1364 1369 4 1316 1266 1269 2 1 1276 1305 1316 2 1319 1354 1345 3 1305 1323 1340 4 1279 1192 1170 Standard error of differences: 18.06 REML variance components analysis for soil biological (FDA) activity

Response variate: FDA Fixed model: Constant + Soil + Fallow + Rate + Soil.Fallow + Soil.Rate + Fallow.Rate + Soil.Fallow.Rate Random model: Pot.Time Number of units: 288





Term Factor Model Order No. rows Pot.Time Pot Identity 0 72 Time Antedependence 2 4 Tests for fixed effects


Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil 793170.27 1 793170.27 14.8 <0.001 Fallow 364820.76 3 121606.92 14.8 <0.001 Rate 109191.11 2 54595.56 14.8 <0.001 Soil.Fallow 636037.13 3 212012.38 14.8 <0.001 Soil.Rate 261441.86 2 130720.93 14.8 <0.001 Fallow.Rate 1003456.51 6 167242.75 14.8 <0.001 Soil.Fallow.Rate 664902.99 6 110817.17 14.8 <0.001


323

Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil.Fallow.Rate 664902.99 6 110817.17 14.8 <0.001 Table of predicted means for Constant



Soil 1 2 74.62 81.70



Fallow 1 2 3 4 78.63 81.59 74.99 77.43



Rate 1 2 3 78.70 76.35 79.43



Fallow 1 2 3 4 Soil 1 72.97 79.43 75.89 70.19 2 84.30 83.74 74.09 84.67



Rate 1 2 3 Soil 1 72.52 75.11 76.23 2 84.87 77.59 82.63



Rate 1 2 3 Fallow 1 78.16 76.34 81.40 2 73.92 81.79 89.05 3 80.25 71.40 73.33 4 82.45 75.88 73.95

324

Standard error of differences: 0.04648 Table of predicted means for Soil.Fallow.Rate

Rate 1 2 3 Soil Fallow 1 1 64.11 74.12 80.67 2 66.68 84.66 86.97 3 81.45 73.99 72.23 4 77.83 67.69 65.05 2 1 92.21 78.56 82.13 2 81.16 78.93 91.13 3 79.05 68.82 74.42 4 87.08 84.07 82.85 Standard error of differences: 0.06573 REML variance components analysis for potentially mineralisable N

Response variate: PMN Fixed model: Constant + Soil + Fallow + Rate + Soil.Fallow + Soil.Rate + Fallow.Rate + Soil.Fallow.Rate Random model: Pot.Time Number of units: 288








Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil 57722.53 1 57722.53 30.0 <0.001 Fallow 42616.11 3 14205.37 30.0 <0.001 Rate 21254.36 2 10627.18 30.0 <0.001 Soil.Fallow 42829.34 3 14276.45 30.0 <0.001 Soil.Rate 42037.02 2 21018.51 30.0 <0.001 Fallow.Rate 186913.74 6 31152.29 30.0 <0.001 Soil.Fallow.Rate 193287.97 6 32214.66 30.0 <0.001 Dropping individual terms from full fixed model

325

Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil.Fallow.Rate 193287.97 6 32214.66 30.0 <0.001




Soil 1 2 44.72 42.33



Fallow 1 2 3 4 43.77 45.12 42.64 42.55



Rate 1 2 3 43.76 44.26 42.56



Fallow 1 2 3 4 Soil 1 43.19 46.97 44.40 44.32 2 44.36 43.27 40.89 40.79



Rate 1 2 3 Soil 1 46.38 44.73 43.05 2 41.13 43.78 42.06

326

Standard error of differences: 0.02346 Table of predicted means for Fallow.Rate

Rate 1 2 3 Fallow 1 45.49 42.04 43.79 2 42.04 49.20 44.13 3 45.53 43.81 38.60 4 41.97 41.97 43.71 Standard error of differences: 0.03318

Table of predicted means for Soil.Fallow.Rate Rate 1 2 3 Soil Fallow 1 1 45.48 38.56 45.53 2 42.06 52.79 46.07 3 49.02 45.55 38.62 4 48.95 42.00 41.99 2 1 45.50 45.53 42.05 2 42.02 45.60 42.19 3 42.03 42.06 38.57 4 34.98 41.94 45.43 Standard error of differences: 0.04692

327

REML variance components analysis for ammonium N

Response variate: NH4_N Fixed model: Constant + Soil + Fallow + Rate + Soil.Fallow + Soil.Rate + Fallow.Rate + Soil.Fallow.Rate Random model: Pot.Time Number of units: 288








Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil 3.99 1 3.99 51.4 0.051 Fallow 247.74 3 82.58 51.4 <0.001 Rate 97.57 2 48.79 51.4 <0.001 Soil.Fallow 19.09 3 6.36 51.4 <0.001 Soil.Rate 2.98 2 1.49 51.4 0.235 Fallow.Rate 55.70 6 9.28 51.4 <0.001 Soil.Fallow.Rate 1.82 6 0.30 51.4 0.933


Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil.Fallow.Rate 1.82 6 0.30 51.4 0.933 Table of predicted means for Constant



Soil 1 2 46.17 44.00



Fallow 1 2 3 4 37.27 49.66 56.83 36.56


328


Rate 1 2 3 38.41 45.29 51.55



Fallow 1 2 3 4 Soil 1 37.92 54.68 56.80 35.27 2 36.62 44.64 56.86 37.85



Rate 1 2 3 Soil 1 38.44 46.19 53.87 2 38.37 44.38 49.24



Rate 1 2 3 Fallow 1 35.36 37.36 39.10 2 40.44 50.65 57.90 3 44.37 54.85 71.27 4 33.45 38.30 37.93



Rate 1 2 3 Soil Fallow 1 1 35.04 37.57 41.16 2 43.48 54.98 65.58 3 43.25 55.63 71.53 4 32.01 36.60 37.19 2 1 35.69 37.15 37.04 2 37.39 46.31 50.23 3 45.50 54.07 71.01 4 34.90 39.99 38.68


329

REML variance components analysis for nitrate - N

Response variate: NO3_N Fixed model: Constant + Soil + Fallow + Rate + Soil.Fallow + Soil.Rate + Fallow.Rate + Soil.Fallow.Rate Random model: Pot.Time Number of units: 288




Covariance structures defined within terms: Term Factor Model Order No. rows Pot.Time Pot Identity 0 72 Time Antedependence 2 4



Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil 0.10 1 0.10 46.9 0.757 Fallow 35.68 3 11.89 46.9 <0.001 Rate 8.39 2 4.20 46.9 0.021 Soil.Fallow 0.06 3 0.02 46.9 0.996 Soil.Rate 0.23 2 0.12 46.9 0.891 Fallow.Rate 8.00 6 1.33 46.9 0.261 Soil.Fallow.Rate 1.38 6 0.23 46.9 0.965


Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil.Fallow.Rate 1.38 6 0.23 46.9 0.965



330


Soil 1 2 536.6 551.0 Standard error of differences: 46.28


Fallow 1 2 3 4 427.1 447.8 774.6 525.5



Rate 1 2 3 463.6 540.1 627.6



Fallow 1 2 3 4 Soil 1 417.7 450.0 760.8 517.8 2 436.5 445.7 788.4 533.2 Standard error of differences: 92.57

Table of predicted means for Soil.Rate Rate 1 2 3 Soil 1 465.8 539.1 604.8 2 461.3 541.0 650.5 Standard error of differences: 80.17


Rate 1 2 3 Fallow 1 370.3 458.0 453.0 2 402.8 453.1 487.6 3 602.7 703.7 1017.4 4 478.5 545.5 552.5


331


Rate 1 2 3 Soil Fallow 1 1 348.1 506.0 399.0 2 418.7 455.0 476.1 3 574.1 692.8 1015.4 4 522.2 502.6 528.7 2 1 392.5 409.9 507.1 2 386.8 451.2 499.1 3 631.2 714.7 1019.4 4 434.8 588.4 576.4 Standard error of differences: 160.3 REML variance components analysis for soil phosphatise activity

Response variate: Phosphatase Fixed model: Constant + Soil + Fallow + Rate + Soil.Fallow + Soil.Rate + Fallow.Rate + Soil.Fallow.Rate Random model: Pot.Time Number of units: 288








Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil 38.62 1 38.62 92.1 <0.001 Fallow 46.91 3 15.64 92.1 <0.001 Rate 2.15 2 1.08 92.1 0.345 Soil.Fallow 7.50 3 2.50 92.1 0.064 Soil.Rate 1.40 2 0.70 92.1 0.500 Fallow.Rate 14.09 6 2.35 92.1 0.037 Soil.Fallow.Rate 6.17 6 1.03 92.1 0.412

332

Dropping individual terms from full fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil.Fallow.Rate 6.17 6 1.03 92.1 0.412

Table of predicted means for Constant 15.54 Standard error: 0.300 Table of predicted means for Soil

Soil 1 2 17.41 13.68



Fallow 1 2 3 4 14.60 14.31 19.09 14.17



Rate 1 2 3 15.12 15.35 16.15



Fallow 1 2 3 4 Soil 1 15.68 15.36 21.48 17.10 2 13.52 13.27 16.69 11.24



Rate 1 2 3 Soil 1 17.34 16.73 18.14 2 12.90 13.97 14.16 Standard error of differences: 1.039

333


Rate 1 2 3 Fallow 1 15.36 13.52 14.91 2 13.94 14.06 14.94 3 17.13 18.33 21.81 4 14.06 15.50 12.94 Standard error of differences: 1.469 Table of predicted means for Soil.Fallow.Rate

Rate 1 2 3 Soil Fallow 1 1 15.58 14.71 16.75 2 17.00 14.02 15.07 3 19.74 20.31 24.41 4 17.05 17.89 16.35 2 1 15.14 12.34 13.07 2 10.89 14.10 14.81 3 14.53 16.34 19.21 4 11.06 13.12 9.54

Standard error of differences: 2.078 REML variance components analysis for soil urease activity

Response variate: Urease Fixed model: Constant + Soil + Fallow + Rate + Soil.Fallow + Soil.Rate + Fallow.Rate + Soil.Fallow.Rate Random model: Pot.Time Number of units: 288




334





Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil 456.70 1 456.70 54.3 <0.001 Fallow 63.24 3 21.08 54.3 <0.001 Rate 164.09 2 82.05 54.3 <0.001 Soil.Fallow 14.77 3 4.92 54.3 0.004 Soil.Rate 5.20 2 2.60 54.3 0.084 Fallow.Rate 10.11 6 1.68 54.3 0.143 Soil.Fallow.Rate 15.44 6 2.57 54.3 0.029 Dropping individual terms from full fixed model Fixed term Wald statistic n.d.f. F statistic d.d.f. F pr Soil.Fallow.Rate 15.44 6 2.57 54.3 0.029




Soil 1 2 393.1 622.1



Fallow 1 2 3 4 476.5 459.2 567.3 527.4



Rate 1 2 3 430.6 495.0 597.2

335



Fallow 1 2 3 4 Soil 1 390.7 316.6 459.0 406.2 2 562.4 601.8 675.5 648.7



Rate 1 2 3 Soil 1 333.3 370.6 475.5 2 527.8 619.5 719.0 Standard error of differences: 18.56 Table of predicted means for Fallow.Rate

Rate 1 2 3 Fallow 1 403.7 479.5 546.3 2 388.2 431.2 558.3 3 481.6 532.4 687.9 4 448.7 537.1 596.4



Rate 1 2 3 Soil Fallow 1 1 313.9 373.5 484.6 2 306.0 271.6 372.2 3 376.0 429.6 571.4 4 337.2 407.7 473.6 2 1 493.6 585.6 608.0 2 470.3 590.8 744.4 3 587.1 635.1 804.4 4 560.3 666.5 719.2


developing a tarodigilib.library.usp.ac.fj/gsdl/collect/usplibr1/index/... · 2016. 4. 25. ·...

Documents