See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/232141777

Tropical Soil Biology and Fertility: A Handbook of Methods

Article  in  Soil Science · April 1994

DOI: 10.2307/2261129

CITATIONS READS

3,177 17,553

2 authors, including:

John S. I. Ingram
University of Oxford
102 PUBLICATIONS   10,748 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Food systems and natural resources - linking food consumption, food production and actors in the food system View project

Food from the oceans: how can more food and biomass be obtained from the oceans in a way that does not deprive future generations of their benefits? View project

All content following this page was uploaded by John S. I. Ingram on 08 February 2017.

The user has requested enhancement of the downloaded file.

 ISSS

TROPICAL SOIL BIOLOGY AND FERTILITY

A Handbook of Methods
Second Edition

Edited by

J.M. Anderson
Rothamsted International
Rothamsted EJ.perimental Station
Hmpenden. AL5 2JQ
UK
and

J.S.!. Ingram
IGBP-Global Change and Terrestrial Ecosystems
Associate Project Office
Department of Plant Sciences
University of Oxford
Oxford OX 1 3RB UK

CA· B Infcr;nafional
C·A·B International Tel: WaHingford (049\) 32111
Wallingford Telex: 847964 (COMAGG G)
Oxon OX 10 8DE Telecom Gold/Dialcom: 84: CAUOOI
UK Fax: (0491) 33508

©C·A·B International 1993. Ali rights reserved. No part of this

publication may be reproduced in any form or by any means,
electronically. mechanically, by photocopying, recording or
otherwise. without the prior permission of the copyright owners.

A catalogue record for this book is available from the British Library

ISBN 0 85198 821 0

Printed and bound in the UK

 Table of Contents

Preface to the Second Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Acknow1edgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv
Principal Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. xv

Chapter 1 THE TROPICAL SOIL BIOLOGY AND FERTILITY PROGRAMME,

TSBF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . '. . . . .. 1

1.1 THE TSBF RATIONALE, PHILOSOPHY AND BACKGROUND ....... 1

1.2 TSBF RESEARCH CONCEPTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.3 TSBF EX AMPLE "SYNCHRONY" EXPERIMENT . . . . . . . . . . . . . . . . 4

1.3. 1 General hypothesis . . . . . . . . . . . . . . . . . . . . . . " . . . . . . . . . 5
l.3.2 Working hypotheses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
l.3.3 Experimental treatments . . . . . . . . . . . . . . . . . . . . ". . . . . . .. 5
1. 3 .4 Treatment rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5
1.3.5 Experimental design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
l. 3.6 Experimental installation . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7
1.3.7 Mea.surements.................................. 7
l.3.8 Analysis and interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . 8
l.3.9 Transition to on-farm experimentation . . . . . . . . . . . . . . . . . . .. 8

l.4 SITE CHARACTERISATION . . . . . . . . . . . . . . . . . . . . . . . • . . . . .. 8

Chapter 2 SITE DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.1 CLIMATE . . . . . . . . . . . . . . . . . . . . . . . . . • . • . • . . . . . . . . . . .. 13
2.1.1 Climatic characterisatiori of site locality . . . . . . . . . . . . . . . . . .. 13
2.1.2 Meteoro1ogical observations . . . . . . . . . . . . . . . . . . . . . . . . .. 13

2.2 SOIL CHARACTERISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 15

2.3 LAND MANAGEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 16

2.3.1 I...and use . . . . . . . . . . . . . . . . . . . . . . . . . "............ 16
2.3.2 Fire........................................ 19
2.3.3 Socioeconomic components . . . . . . . . . . . . . . . . . . . . . . . . .. 19

iii
TSBF: Ji Handbook of Merhods

Chapter 3 FIELD PROCEDURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.1 ABOVE-GROUND VEGETATION . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.1.1 Woody vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 22
3.1.2 Herbaceous plants and short duration crops . . . . . . . . . . . . . . . . 27

3.2 ABOVE-GROUND INPUTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3.2.1 Tree and shrub litter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.2 Herbaceous litter and above-ground crop residues . . . . . . . . . . . .. 30
3.2.3 Other inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

3.3 ROOTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.3.1 Root preparation on profùe walls . . . . . . . . . . . . . . . . . . . . . . 32
3.3.2 Root mapping on profile walls . . . . . . . . . . . . . . . . . . . . . . . . 33
3.3.3 Pinboard monolith sampling . . . . . . . . . . . . . . . . . . . . . . . . .. 34

3.4 LITTER DECOMPOSITION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.4.1 Introduction................................... 36
3.4.2 Litter bags and decomposition rate constants. . . . . . . . . . . . . . .. 36
3.4.3 Decomposition of unconfined litter and turnover rates . . . . . . . . .. 37
3.4.4 Litter bag methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.4.5 Decomposition standards . . . . . . . . . . . . . . . '. . . . . . . . . . . .. 40

3.5 SOIL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , 41
3.5.1 Soil CO2 evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.5.2 Soil fauna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 44

Chapter 4 SAMPLING FOR LABORATORY ANALYSIS AND SAMPLE

PREPARATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

4.1 SOIL AND PLANT SAMPLING FOR LABORATORY ANALYSIS . . . . . . 47

4.1.1 Soil sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.1.2 Plant sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 48

4.2 SOIL, PLANT AND LITTER MATERIAL PREPARATION . . . . . . . . . .. 48

4.2.1 Soil material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 48
4.2.2 Plant material and liUer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.2.3 Sample storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Chapter 5 LABORATORY PRACTICE AND QUALITY CONTROL . . . . . .. 50

5.1 LABORATORY LAY-OUT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

5.2 SAFETY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

IV
 Tab//! of COIII/!IIIS

5.3 SERVICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3.1 Water supp1y and quality . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3.2 E1ectricity suppl Yand quality . . . . . . . . . . . . . . . . . . . . . . . .. 52
5.3.3 Work surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 52

5.4 GLASS AND PLASTICWARE WASHING . . . . . . . . . . . . . . . . . . . . . 52

5.5 CONfAMINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

5.6 LEVELS OF ACCURACY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

5.7 SAMPLE BATCHES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

5.8 DATA RECORDING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 54

5.9 GENERAL POINfS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 54

5.10 QUALITY CONTROL AND STANDARDISATION PROCEDURES . . . . . . 54

5.10.1 Blanks....................................... 55
5.10.2 Repeats within a batch . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 55
5.10.3 Internai references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.10.4 Inter-Iaboratory standardisation . . . . . . . . . . . . . . . . . . . . . . .. 56

Chapter 6 CHEMICAL ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

6.1 WATER CONTENT DETERMINATION FOR DATA CORRECTION . . . .. 57

6.2 pH AND ELECTROCONDUCTIVITY SCREENING . . . . . . . . . . . . . . . 57

6.3 SATURATED PASTE EXTRACT CONDUCTIVITY . . . . . . . . . . . . . . . 58

6.4 EXCHANGEABLE CATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

6.4.1 Exchangeable bases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 59
6.4.2 Exchangeable acidity: AI3+ (+ H+) . . . . . . . . . . . . . . . . . . . .. 60
6.4.3 Cation exchange capacity (CEC) . . . . . . . . . . . . . . . . . . . . . . 61
6.4.4 Effective cation exchange capacity (ECEC) and aluminium saturation. 62

6.5 SOIL ORGANIC MATTER AND ORGANIC CARBON . . . . . . . . . . . . . 62

6.5.1 Definitions.................................... 62
6.5.2 Total organic carbon in soils and soil extracts: background . . . . . .. 63
6.5.3 Totalorganic carbon in soils: colorimetrie method . . . . . . . . . . . . 63
6.5.4 Total organic carbon in soil extracts: titration method . . . . . . . . . . 65
6.5.5 Soil titter separation . . . . . . . . . . . . . . . -. . . . . . . . . . . . . .. 66
6.5.6 Microbiai biomass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

6.6 NITROGEN......................................... 70
6.6.1 Digestion for total nitrogen (and phosphorus) . . . . . . . . . . . . . . . 70

Il ------------------------------------------------------------------------
 78BF: .A. Handbool 01 M~rlwtb

6.6.2 Detennination of nitrogen: distillation and titration . . . . . . . . . . .. 71

6.6.3 Colorimetrie detennination of ammonium . . . . . . . . . . . . . . . . . 73
6.6.4 Colorimetrie determination of nitrate . . . . . . . . . . . . . . . . . . . . 74

6.7 NlTROGEN MINERALISATION . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

6.7.1 Introouction..... III • • • • III III • • III • III • • • • • • • III III III III • III III III • •• 76
6.7.2 Nitrogen mineralisation (field incubation methoo) . . . . . . . . . . . .. 76
6.7.3 Aerobic incubation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
6.7.4 Anaerobie N mineralisation index . . . . . . . . _ _ . _ . . . _ _ . . . .. 79

6.8 DENITRIFICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6.8.1 Introduction................................... 80
6.8.2 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 80

6.9 PHOSPHORUS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.9. 1 Total phosphorus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.9.2 Bicarbonate ex tractable phosphate . . . . . . . . . . . . . . . . . . . . . . 83
6.9.3 Resin extraetable phosphate . . . . . . . . . . . . . . . . . . . . . . . . .. 83
6.9.4 Organic phosphorus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 85
6.9.5 Colorimetrie determination of phosphorus . . . . . . . . . . . . . . . . . 87

6.10 POLYPHENOLICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

6.11 LIGNIN AND CELLULOSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

6. 11. 1 Introouction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
6.11.2 Lignin and cellulose via acid detergent fibre (ADF) . . . . . . . . . . . 90

Chapter 7 SOIL PHYSICAL ANALYSES . . . . . . . . . . . . . . . . . . . . . . . 93

7.1 MECHANlCAL ANALYSIS (TEXTURE) . . . . . . . . . . . . . . . . . . . . . . 93

7.2 BULK DENSITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

7.2.1 Method for non-stony soils . . . . . . . . . . . . . . . . . . . . . . . . . . 95
7.2.2 Metluxls for stony soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
7.2.3 Method for shrink-swell soils . . . . . . . . . . . . . . . . . __ - .... , 96

7.3 FIELD CAPACITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 97

7.4 LOWER LIMIT OF PLANT AVAILABLE WATER . . . . . . . . . . . . . . ., 98

7.4.1 The -1.5 MPa water content using a pressure chamber apparatus ... 98
7.4.2 Soil water potential by the filter paper methoo . . . . . . . . . . . . . . 99

7.5 PLANT AVAILABLE WATER CAPACITY (PAWC) . . . . . . . . . . . . . . . 100

7.6 INFILTRATION...................................... 100

7.6.1 Double ring method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1
7.6.2 Single ring methoo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1

LI~ _____________________________________V_l_______________________________ ~
 Table of ConlmlS

7.7 POROSITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

7.7.1 Total porosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
7.7.2 Pore size distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Appendix A EXAMPLE OF RAPID RURAL APPRAISAL (RRA) INTERVIEW . 105

A.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

A.2 PHASE 1: PREPARATORY WORK . . . . . . . . . . . . . . . . . . . . . . . . . . 105

A.3 PHASE II: FIELDWORK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

A.4 PHASE III: COMPLETING THE RRA . . . . . . . . . . . . . . . . . . . . . . . . 107

Appendix B THE RHIZOSPHERE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

B.1 INTRODUCTION..................................... 109

B.2 THE RHIZOSPHERE ENVIRONMENT . . . . . . . . . . . . . . . . . . . . . . . 109

B.2.1 The movement of organic substrates into soil . . . . . . . . . . . . . . . 111
B.2.2 The inorganic nature of the root surface and rhizosphere . . . . . . . . 112

B.3 THE MICROBIAL POPULATIONS ON ROOTS . . . . . . . . . . . . . . . . . . 113

B.3.1 Culture of micro-organisms . . . . . . . . . . . . . . . . . . . . . . . . . . 113
B.3.2 Direct observation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
B.3.3 Rhizosphere dynamics, interaction and experimental approaches . . . . 114
B.3.4 Interactions.................................... 115
B.3.5 The effects of the rhizosphere microflora . . . . . . . . . . . . . . . . . . 115

Appendix C MYCORRHIZAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

C.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

C.2 THE VESICULAR-ARBUSCULAR MYCORRHIZAL SYMBIOSIS . . . . . . 122

C.2.1 Characterisation and analysis of VA mycorrhizal infection . . . . . . . 122
C.2.2 Quantification of infection . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
C.2.3 Isolation and identification of VA fungi . . . . . . . . . . . . . . . . . . . 124
C.2.4 Inoculation of test plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
C.2.5 Analysis of host response . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

C.3 THE ECTOMYCORRHIZAL SYMBIOSIS . . . . . . . . . . . . . . . . . . . . . . 127

C.3.1 Characterisation and analysis of ectomycorrhizal infection . . . . . . . . 127
C.3.2 Quantification of occurrence . . . . . . . . . . . . . . . . . . . . . . . . . 127
C.3.3 Isolation and identification of ectomycorrhizal fungi . . . . . . . . . . . 128
C.3.4 Inoculation of test plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

vu

lit
TSBF: A Handbook of Methods

Appendix D ROOTS: LENGTH, BIOMASS, PRODUCTION AND MORTALITY 132

D.1 INTRODUCTION..................................... 132

D.2 QUANTIFICATION OF ROOT BIOMASS AND LENGTH . . . . . . . . . . . 132

D.2.1 Corer design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
D.2.2 Sample depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
D.2.3 Sampling intensity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
D.2.4 Root extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
D.2.5 Classifying the roots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
D.2.6 Assessment of root mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
D.2.7 Root 1ength measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
D.2.8 Mycorrhizas................................... 137
D.2.9 Sampling frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

D.3 ESTIMATION OF TOTAL ROOT PRODUCTION . . . . . . . . . . . . . . . . . 138

D.3.1 Root decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
D.3.2 In-growth cores .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
D .3.3 Minirhizotrons.................................. 140

D.4 ESTIMATE OF TOTAL CARBON INPUT TO THE SOIL . . . . . . . . . . . . 141

Appendix E SOIL SOLUTION SAMPLING AND LYSIMETRY . . . . . . . . . . . 145

E.1 ISOLATED SOIL MASSES ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

E.2 TENSION/VACUUM/SUCTION SAMPLERS . . . . . . . . . . . . . . . . . . . . 146

E.2.1 Introduction................................... 146
E.2.2 Cup and ring-based sampling systems . . . . . . . . . . . . . . . . . . . . 146
E.2.3 Porous plate samplers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
E.2.4 Fritted glass tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
E.2.5 Hollow fibres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
E.2.6 Application of tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
E.2.7 Samp1e recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
E.2.8 Solution-sampler interactions . . . . . . . . . . . . . . . . . . . . . . . . . 149
E.2.9 Sampled soil volume and calculation of element fluxes . . . . . . . . . 150
E.2.10 Overview..................................... 150

E.3 TENSIONLESS COLLECTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

E.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
E.3.2 Trough, box and funnel-based collectors . . . . . . . . . . . . . . . . . . 152
E.3.3 Sheet or tray-based collectors . . . . . . . . . . . . . . . . . . . . . . . . . 152
E.3.4 Tensionless collectors on sloping sites . . . . . . . . . . . . . . . . . . . . 153
E.3.5 Sample contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
E.3.6· Calculation of fluxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

viii
 Tabl~ 0/ COIIImlS

Appendix F NITROGEN AVAILABILITY . . . . . . . . . . . . . . . . . . . . . . . . 158

F.l INTRODUCTION..................................... 158

F.2 MINERALISATION . • • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

F.3 DENITRIFICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Appendix G TECHNIQUES FOR QUANTIFYING NITROGEN FIXATION ... 164

G.l INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

G.2 NITROGEN DIFFERENCE METHOD . . . . . . . . . . . . . . . . . . . . . . . . 164

G.3 15N-ISOTOPIC TECHNIQUES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

G.3.1 lsN-enrichment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
G.3.2 Natural lSN abundance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

G.4 ACETYLENE REDUCTION ASSAY (ARA) . . . . . . . . . . . . . . . . . . . . 165

G.5 UREIDE METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

G.6 XYLEM SOLUTE TECHNIQUE (UREIDE) FOR NITROGEN FIXATION

BY LEGUM.BS .... ......................................................................... 166
G.6.1 Sampling of N-solutes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
G.6.2 Sap storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
G.6.3 Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

Appendix H MOST PROBABLE NUMBER COUNTS OF RHIZOBIA IN SOILS . 172

H.l BACKGROUND...................................... 172

H.2 EXPERIMENTATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

H.2.1 Experimental design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
H.2.2 Materials..................................... 173
H.2.3 Plant preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
H.2.4 Soil sampling, preparation and storage prior to dilution .. . . . . . . . 174
H.2.5 Preparation of the dilution series, root inoculation and
observations of nodulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

H.3 ASSIGNING TABULAR POPULATION ESTIMATES TO RESULTS . . . . . 175

H.3.1 Constructing confidence limits . . . . . . . . . . . . . . . . . . . . . . . . 177

IX
Appendix 1 THE STUDY OF PHOSPHORUS CYCLES IN ECOSYSTEMS ... 179

1.1 INTRODUCTION..................................... 179

1.2 SOIL TESTS FOR INORGANIC P AVAILABILITY . . . . . . . . . . . . . . . . 179

1.3 PHOSPHORUS "FIXATION" . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180

I.4 ORGANIC PHOSPHORUS TRANSFORMATIONS . . . . . . . . . . . . . . . . 181

1.5 SEQUENTIAL EXTRACTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

1.6 SPATIAL VARIABILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

1.7 CURRENT APPROACHES AND DIRECTIONS . . . . . . . . . . . . . . . . . . 183

AppendÎX J ISOTOPE STUDIES IN TROPICAL SOIL BIOLOGY . . . . . . . . . 189

J.l INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

J.2 DOUBLE (14C, ISN) OR TRIPLE LABELLED (14C, ISN, 32p) RESIDUES .• 189

J.3 LABELLING THE PLANT MATERIAL . . . . . . . . . . . . . . . . . . . . . . . 190

].4 DECOMPOSITION EXPERIMENTS . . . . . . . . . . . . . . . . . . . . . . . . . 190

J.~ !;Jc METHODOLOGY IN ESTIMATING TURNOVER RATES OF SOIL

ORGANIC MATIER FRACTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . 193

J.6 14C METHODS TO QUANTIFY ROOT PRODUCTION AND TURNOVER. 194

Appendix K USINO OROWTH ANALYSIS TO ESTIMATE PLANT

NUTRIENT UPT AKE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

K.l INTRODUCTION..................................... 196

K.2 THE CLASSICAL APPROACH . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

K.3 APPLICATION TO NUTRIENT UPTAKE . . . . . . . . . . . . . . . . . . . . . . 197

Appendix L USE OF THE CENTURY PLANT-SOIL ENVIRONMENTAL

SIMULATION MODEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199

L.l COMPUTER SIMULATION OF DECOMPOSITION AND SOIL

OROANIC MATIER TRANSFORMATIONS . . . . . . . . . . . . . . . . . . . . 199

x
 Table of CommIs

L.2 MODEL STRUCTURE AND DESCRIPTION . . . . . . . . . . . . . . . . . . . . 200

L.3 HARDWARE REQUIREMENTS AND INSTALLATION . . . . . . . . . . . . . 201

LA RUNNING CENTURY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

L.5 PRINCIPAL MODEL ROUTINES . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

L.6 CENTURY OUTPUTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202

Ap~ndix M QUANTIFICATION OF NEAR-SURFACE MICROCLIMATE .... 204

M.I INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204

M.2 MICROCLIMATIC FACTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204

M.2.1 Parameter selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
M.2.2 Meteorologica1 observations . . . . . . . . . . . . . . . . . . . . . . . . . . 204
M.2.3 Micrometeorologica1 measurements . . . . . . . . . . . . . . . . . . . . . 205
M.2.4 Mulching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
M.2.5 Shading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
M.2.6 Wind protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206

M.3 PROCESSES AFFECTING MICROCLIMATE . . . . . . . . . . . . . . . . . . . 206

M.3.1 Radiation balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
M.3.2 Soil temperature and heat flow . . . . . . . . . . . . . . . . . . . . . . . . 206
M.3.3 Soil aeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
M.3.4 Crop evapotranspiration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

M.4 LINKED TRANSPORT MODELS . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

M.5 AUTOMATIC RECORDING OF MICROCLIMATIC FACTORS . . . . . . . . 208

Appendix N SUGGESTED EQUIPMENT AND REAGENT LISTS FOR

ROUTINE TSBF ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . 211

N.I EQUIPMENT AND NON-REA GENT CONSUMABLES . . . . . . . . . . . . . 211

N.2 SUGGESTED EQUIPMENT SUPPLIERS . . . . . . . . . . . . . . . . . . . . . . 218

N.3 REAGENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

N.4 SUGGESTED REAGENT SUPPLIER . . . . . . . . . . . . . . . . . . . . . . . . . 220

Appendix 0 TSBF DATA SHARING POLICY . . . . . . . . . . . . . . . . . . . . . . 221

Xl
 Preface to the Second Edition

During the time since publication of the first edition of this Handbook in 1989 the methods
have been tested in a wide range of field and laboratory environments. This both intensive
and extensive testing has lead to considerable debate and proposaIs for revision; we hope this
second edition adequately reflects the substantial collective thought, experience and input of
many scientists.

Overall the Handbook now comprises seven chapters and 15 appendices.

A completely revised Chapter 1 now complements the TSBF philosophy with a discussion of
an example TSBF-type experiment. This sets the recommended site characterisation and
methodology in better context. The frrst edition main text has been substantially re-structured
to offer a more logical and sequential approach to both field and laboratory procedures. A
condensed Chapter 2 discusses site characterisation, while Chapters 3 - 7 cover field
measurements, field sampling, laboratory practice and quality control, chemical analyses and
soil physical analyses, respectively. Additional material on sample preparation and the new
chapter on laboratory practice and quality control provide a more rounded laboratory
compendium. Subsection introductions have been developed and, where appropriate, cited
and "suggested further reading" references updated. Several new diagrams have been
inc1uded and a standardised, and c1earer format used throughout.

Many users found the appendices of great value. Most appendices from the first edition have
therefore been revised, and several (socioeconomic considerations, soil phosphorous, near-
surface microclimate and suggested equipment and reagents) have been completely re-written.
Several new appendices have been added (mycorrhizas, roots, most probable number analysis
for rhizobia, isotope studies and a description of the CENTURY model).

We believe this second edition now forms a substantially c1earer and more complete reference
work for ecosystem studies, being equally applicable to both tropical and temperate systems.

Jonathan Anderson and John Ingram

October 1992

Xll1

1'1
 Acknowledgements

The Tropical Soil Biology and Fertility Programme

. was established as one of the four programmes of the
International Union of Biological Sciences / UNESCO
Man and the Biosphere Programme

"The Decade of the Tropics".

Financial assistance from the Rockefeller Foundation and the International Soil Science
Society for the preparation of this Handbook is gratefully acknowledged.

The Editors gratefully acknowledge the editorial assistance of Dr Mary Scholes and Dr
Robert Scholes, and the assistance of Alice N'dungu in manuscript preparation.

The methods given in this Handbook are endorsed by

The International Society of Soil Science.

XlV

l
 Principal Contributors

1 Alexander, University of Aberdeen, UK

J M Anderson, Exeter University & Rothamsted International, UK
K F Baker, Natural Resources Institute, Chatham, UK
E Barrios, Instituto Venezolano de Investigaciones Cientijicas, Caracas, Venezuela
L Bohren, Colorado State University, Fort Collins, USA
GD Bowen, CSIRO Division of SoUs, Glen Osmond, Australia
S Brown, University of Illioois, Champaign, USA
E Cuevas, Instituto Venezolano de Investigaciones Cientificas, Caracas, Venezuela
J M Dangerfield, University of Botswana, Gaberone, Botswana
T Darnhofer, United Nations Environment Programme, Nairobi, Kenya
y Dommergue, ORSTOM, Nojent sur Marne, France
ET Elliott, Colorado State University, Fort Collins, USA
M Homung, Institute of Terrestrial Ecology, Merlewood, UK
J S 1 Ingram, IGBP-Global Change and Terrestrial Ecosystems, Oxford, UK
J Ladd, CSIRO Division of SoUs, Glen Osmond, Australia
P Lavelle, ORSTOM, Bondy, France
R Merckx, Katholieke University. Leuven, Belgium
R J Myers, International Board for Soil Research and Management, Bangkok, 1hailand
L Nelson, North Carolina State University, USA
M van Noordwijk, Institute for Soil Fertility, Haren, the Netherlands
C A PaIm, North Carolina State University, USA
M B Peoples, CSIRO Division of Plant Industry, Canberra, Australia
R J Raison, CSIRO Division of Forestry, Canberra, Australia
D J Read, Sheffield University, UK
P Robertson, WK Kellogg Biological Station, Hickory Corners. USA
A P Rowland, Institute of Terrestrial Ecology, Merlewood, UK
P A Sanchez, International Center for Research in Agroforestry, Nairobi, Kenya
M C Scholes, Witwatersrand University, Johannesburg. Republic of South Africa
R J Scholes, CSIR Forest Science and Techoology. Pretoria, Republic of South Africa
PD Seward, Tropical Soil Biology and Fertility Programme, Nairobi, Kenya
A R A Stapleton, Exeter University, UK
J L Stewart, Oxford Forestry Institute, UK
M J Swift, International Institute of Tropical Agriculture, Ibadan, Nigeria & Tropical Soil
Biology and Fertility Programme, Nairobi, Kenya
H Tiessen, University of Saskatchewan, S(lSkatoon, Canada
P M Vitousek, Stanford University, USA
B H WaIker, CSIRO Division of Wildlife and Ecology, Canberra, Australia
PL Woomer, Tropical Soil Biology and Fertility Programme, Nairobi, Kenya
A Young, University of East Anglia, Norwich, UK

xv
Chapter 1 THE TROPICAL SOIL BIOLOGY AND FERTILITY
PROGRAMME, TSBF

1.1 THE TSBF RATIONALE, PHILOSOPHY AND BACKGROUND

Recent years have seen a dramatic increase in the per capita food production in much of the
tropics. This improvement is largely based on the introduction of new crop varieties into
farming programmes on fertile soils with good supplies of water, fertiliser and pesticide. In
large parts of Africa, however, and in many other less fertile parts of the tropics, the
production trend is the opposite and per capita production of food has actua11y been declining
for twenty years.

The further spread of high-input farming systems is undoubtedly one answer to this situation,
and one which is being pursued vigorously by the development of new and appropriate
technologies. It is apparent however that the economic oost of such agriculture is often
unacceptably high. These farming systems are also oommonly of low efficiency in terms of
resource use and may be accompanied by rapid environmental degradation. At the same time
the more environmentally oonservative traditional forms of agriculture practised in much of
the tropics are no longer sustainable because of increased population densities and pressures
on land. The last two decades have therefore seen great interest in the development of
farming systems characterised by a relatively inexpensive level of input, a high efficiency of
internal resource use and hence more sustainable production in both economic and ecological
terms.

Sustainable use of the soil resource is a primary goal of all such farming systems and it is to
this target that the Tropical Soil Biology and Fertility Programme (TSBF) is directed. The
Programme' s stated general objective is "to determine the management options for improving
tropical soil fertility through biological processes" (Swift, 1984). TSBF differs from, and
complements, other research programmes in concentrating on research into the biological
resources and processes of soil; it draws on expertise from both agricultural and ecological
research.

In natural ecosystems productivity is sustained by the tight integration of the vegetative system
with the biological system of soil in relation to key processes such as nutrient cyc1ing and the
formation and breakdown of soil organic matter. These cruciall y important biological
processes of soil are still poorly understood by ecologists, even in natural ecosystems, and
are rarely investigated by agriculturalists. One of the reasons for this is ~e success of high-
input farming, which effectively bypasses soil biological processes through its use of
fertilisers, pesticides and the mechanised preparation of soi!.

This success leaves little apparent reason why soil biological processes should he taken
seriously. The focus on sustainable low-input agriculture described above does however
provide such a rationale. Furthermore, it is noticeable that the adoption of minimum tillage

1
TSBF: A Handhook of Methods

systems in temperate regions is refocusing attention on soil ecology. Such systems can
provide economically viable options which may be particularly applicable to the sensitive soils
of certain parts of the tropies. Moreover, much of our current knowledge of these soil
biological processes has been aehieved through pure researeh just as the understanding of the
consequences of perturbing nutrient cycles has largely come from manipulative studies of
natural ecosystems. It is therefore necessary to retain and develop an interface between
naturaI and agroecosystems as a context for integrating and building on the knowledge of
ecologists and agricultural scientists. It is evident that there are fundamental as well as
practical scientific problems to he solved; these problems are not unique to the tropics but it
is the partieular problems of the tropics that demand their solution.

TSBF was initiated in 1984 by the International Union of Biologica1 Sciences and the
UNESCO Man and Biosphere Programme. The underlying aim is to stimulate research in
the poorly-understood topie of the role of biologica1 processes in the maintenance of soU
fertility. It is one component of the collaborative research programme entitled "The Decade
of the Tropics", the objective of which is to "increase our knowledge and understanding of
the biology of the tropics from the point of view of the various biological subdisciplines"
(Solbrig and Golley, 1983).

The TSBF strategy inc1udes focus on the processes of carbon, nitrogen and water transfers
through the soH-plant system, and the key factors which regulate them. The practical targets
are to synchronise plant demand with nutrient and wateT availability, and to manipulate and
conserve soH organic matter pools. The tools with which this can be achieved are the
management of the quantity, quality and timing of organic inputs, and the manipulation of wil
organisms. To understand better the regulatory processes which underpin natura! systems,
and how they differ for managed systems, a key taerle has been the comparison of managed
systems with the natural ecosystems from which they were derived.

Further details on TSBF's background can be found in Swift (1984, 1987), Ingram and Swift
(1989) and Woomer and Ingram (1990); a comprehensive synthesis of TSBF and associated
research approach and findings is presented in Woomer and Swift (1992).

References

Ingram, J.S.1. and Swift, M.J. (OOs.) (1989) Tropical Soil Biology and Fertility
Programme: report of the Fourth International Workshop, Harare. Biology
Imernational Special Issue 20, 1-64.
Solbrig, a.T. and Golley, F.B. (1983) A decade of the tropics. Biology International
Special Issue 2, 1-15.
Swift, M.J. (ed.) (1984) Soil biological processes and tropical soil fertility (TSBF): planning
for re~eh Biology Imernational Special Issue 9, 1-24.
Swift, M.J. (ed.) (1987) Tropical Soil Biology and Fertility: Inter-regional research
planning workshop, Yurimaguas. Biology Imernational Special Issue 13, 1-68.
Woomer, P.L. and Ingram, J.S.I. (eds.) (1990) The Biology and Fertility of Tropical Soi/s.
TSBF Report: 1990. TSBF, Nairobi.
Woomer, P.L. and Swift, M.J. (eds.) (1992) The Biological Management o/Tropical Soil
Fertility (in press).

2
 C1top~r 1: The Tropical Soil Biology and Ferltlil] Programnu!, TSBF

1.2 TSBF RESEARCH CONCEPTS

Over the past few years, research has been conducted within TSBF to address. two major
Principles that consider the relationships between soil biological processes and soU fertility.

The first is termed "Synchrony", and recognises that the availability of nutrients to plants is
controlled by a complex set of interactions. Nevertheless, the release of nutrients from
decomposing organic residues can be synchronised with plant growth demand. A dynamic
equilibrium exists between nutrient supply from mineralisation (or addition of inorganic
fertiliser) and Its control by immobilisation on organic materials and losses through
volatilisation, denitrification and leaching. Biological processes regulate this equilibrium,
and, as they are open to potential manipulation, the additional competing process of plant
uptake can be optimised. It is hypothesised that the potential for such manipulation will be
greatest when there is a diversity of resources available, and when management options are
relatively flexible. The main way to promote synchrony is with the management of the
timing, placement and quality of organic residues.

The second major Principle concerns soU organic matter. Soil organic matter plays key roles
in crop sustainability, primarily through its interactions with soil chemica1 and physica1
properties in relation to nutrient release, cation retention and soil structure. The value of soil
organic matter (as distinct from the value of organic inputs described in the "Synchrony"
Principle) is weil recognised, but little is known about the processes that contribute to its
three key roles. This is in sharp contrast with the weIl understood processes underlying the
use of chemica1 fertilisers. As low-input systems regain importance in the tropics it is
essential to improve understanding of the functioning of soil organic matter.

Soil organic matter can differ in both quantity and quality (composition) resulting in different
patterns of nutrient release and availability. The quality of organic inputs may affect the
composition, and the long- and short-term nutrient release rates from soil organic matter.
The benefits of soil organic matter on soil properties a8S0Ciated with hea1thy plant growth are
weIl demonstrated; the processes which lead to the accumulation of organic matter in soils
are however extremely complex, being associated with overall environ mental conditions. To
assist in rationalising the complex processes inherent in the maintenance of soil fe rtility ,
simulation models are often used. The TSBF Programme bas adopted the CENTURY model
to simulate plant production, nutrient cycling and soil organic matter dynamics for TSBF
sites. A central feature of CENTURY is the division of soil organic carbon into functional
pools based on the residence time in soHs. A description of the CENTURY model, and its
application in TSBF studies. is given in Appendix L ("Use of the CENTURY plant-soil
environ mental simulation model").

Both these major Principles incorporate consideration of the themes of soil water and soil
fauna. AU soil biologica1 and chemica1 activities are dependent on an adequate level of soil
water. Fluctuating moi sture conditions are a feature of most tropical areas, and both an
excess or a shortage of water in the sail can limit decornposition, nutrient release, nutrlent
uptake by plants and plant growth. The duration for which water is available within the
tolerance range of a particular process therefore has an overall controlling influence on the
degree to which that process can operate.

3
TSBF: A Hmulbook of Melhods

Soil fauna, particularly earthworms and termites, are important as regulators of

decomposition, nutrient cycling and soïl organic matter dynamics; they directly and indirectly
effect soil structure and hence aeration, water infiltration and water holding capacity. The
response of soil fauna to changes in land use and cultivation practices cao therefore impart
profound effects on soil fertility through their feeding and burrowing activities. To help
understand their roles in nutrient regulation and soil organic matter turnover, soil fauna can
be classified into ecological groups based on their feeding habits and habitats.

The TSBF experimental approach has evolved from the recognition that the ability of natuml
plant communities to conserve nutrients and moisture, through the effects of canopy, litter and
soil organic matter, serve as an effective model for the development of more sustainable
agroecosystems. Soil fertility is therefore seen as the integration of plant nutrient demand,
decomposition processes, soïl fauna activities and their interaction with soil chemical and
physical properties. These processes are inextricably linked, and cultivation practices such
as green manuring, mulching or crop residue incorporation have a multiplicity of indirect
effects on soil biological processes. The suite of biologica1, physica1 and chemical variables
used in TSBF studies, provide a basis for interpreting the complex direct and indirect effects
of treatments on soil processes and conditions.

1.3 TSBF EXAMPLE "SYNCHRONY" EXPERIMENT

As discussed above, the potential to synchronise nutrient availability with plant demand
through the judicious use of crop residues offers a means of improving crop productivity
while reducing nutrient losses. In many agroecosystems, crop residues are however used in
a number of ways, each of comparative value, as e.g. animal feed, fuel or organic additions
to subsequent crops. Furthermore, many farming systems have access to only limited
quantities of nutrient-rich (high quality) plant materials or animal wastes. Synchrony
experimentation therefore seeks to maximise nu trient use-efficiency through the combinations
of inputs of contrasting chemical composition. Although this example experiment considers
only two materials of contrasting quality, severa! of the central features of synchrony-type
field experimentation apply:

• relatively few treatments are examined, but each in great detail, in order to identify
the pools and fluxes of the nutrient that limits plant procluctivity.

• equal amounts of a potentially limiting plant nutrient are applied in different forms and
combinations to all treatments (except the complete control); consequently, differences
observed between treatments result from the use-efficiency of the various resources
rather than the amounts applied.

• frequent plant sampling is required in order to estimate plant nu trient uptake and
demande

• measurement of the effects of residue addition on severa! soil properties in addition

to mineralisation of the nutrient limiting plant productivity; i.e. soil faunaI dynamics,
organic matter fractionation, soH water content and water holding capacity.

4
 Chaprer 1: 'lM Tropical Soil Biology and Fertility Programme. TSBF

1.3.1 General hypothesis

The release of nutrients from above-ground inputs and decomposing roots can be
synchronised with plant growth demands.

1.3.2 Working bypotheses

• mixtures of applied organic residues of contrasting chemical properties result in

improved plant productivity through controlling the pattern of nutrient release.

• residue application results in short-term immobilisation of plant nutrients within the

soil microbial biomass that becomes available to the crop later in the growing season
when nutrient demands are greater.

1.3.3 Experimental treatments

This example experiment uses locally available organic materials. The amount of nutrients
in the litter should he sufficient to produce a response in the plants. This can he determined
by appropriate calculations based on previous knowledge of crop demand, and nutrient
content oflitter. Pilot studies should determine litter decay rates, nutrient concentrations, and
potential toxicity effects of the various litters. Treatments are:

1 Complete control (no amendments)

2 Low quality residue resource applied at 40 kg N/ha
3 High quality residue resource applied at 40 kg N/ha
4 Low and high quality resources each applied at 20 kg N/ha

High and low quality litter have narrow and wide C:N ratios, respectively. Low quality
resources include such agricultural residues and waste products as cereal stover, sugar cane
bagasse, sawdust and wood chips. These materials characteristically contain larger amounts
of lignin and cellulose. Included among high quality resources are animal manures and other
animal products, green manures, composts, agroforestry leaf litters and prunings and
inorganic fertilisers.

1.3.4 Treatment rationale

The complete control (treatment 1) is compared to all treatments as a test of improved plant
performance due to inputs. Treatment 1 also serves as the control treatment in all soil
biologica1 process measurements. Researchers must consider that this treatment may contain
root and stubble residues.

The low quality resource treatment (treatment 2) is a measure of the potential immobilisation
of the system and serves as a control when compared to the mixed residue quality treatment.
~esearchers should be aware that the quantities of this resource required to match the nutrient
mput levels of the high quality resource (treatment 3) are often excessive (> 10 tlha) and

5
TSBF: A HanJbook 01 McthodJ

may result in changes of other soil properties (e.g. soil water holding capacity, microclimate,
etc.).

The high quaIity residue treatment (treatment 3) acts as a control of the nutrient use efficiency
in the mixed treatment (treatment 4). Researchers must be aware that any root residues
repre!lent a low quality input common across all treatments and that the high quality residue
input contains nutrients other than that which is limiting the growth of the complete control
(treatment 1).

The low + high mixed liner quality treatment (treatment 4) is compared to treatments 2 and
3 as a test of the comparative nutrient use efficiency of the high quality input due to potential
short-term immobilisation of plant nutrients resulting from the addition of the low quality
input.

1.3.5 Experimental dHÎgn

The experiment is installed as a randomised complete block design with 4 replicates (Figure
1.1). The dimensions of the individual plots are 10 m x 6 m. The distance between rows
may be either 60 or 75 cm. This allows for either 10 or 8 crop rows per plot, the outside
pair of which are not sampled. Because the residual effect of the treatments over time may
be substantial, the study will be continued on the same plots for up 10 3 years to assess these
potential cumulative effects on plant growth and soil properties .

, ..
. . . . .. . . . . . . . ..
..................... '

...........
'" •••• 1 •••••••••• ." . . . . . . . . . . . .,
. ............ '. ..... :~.::,
............ . .... ' ................. :.:.

.:,:,:1I:J::::
............
::::::#2'::::: ..:·/#.'
........ .. ............
,
::3':':' :':':#4,:';
. ................. .
.................... .

Figure 1.1 Experimentallayout is a Randomised Complete Block design, with four replicates
and four treatments. Treatments 1, 2, 3 and 4 are complete control, low quality input, high
quality input and mixed inputs, respectively.

6
 ChDpler 1: The Tropical Soi/ Bi%gy and Fertillry Programme. TSBF

1.3.6 Experimental installation

The site should be mapped according to elevation and relevant soil characteristics. The entire
field is tilled to a depth of 15 cm and the four blocks delineated based on the results of the
initial site characterisation and terrain. Thus blocks may be irregularly shaped and non-
contiguous if necessary to ensure uniformity of plots within a block. Then treatment plots
(10 m x 6 m) are randomly placed within each block. The required additions to soils are
evenly distributed across the soil surface and incorporated to 15 cm. Within the individual
plots four areas are identified that will he sequentially sampled for above-ground biomass,
root in-growth, litter decay, nutrient mineralisation and other soils properties. These sample
areas are 1.5 m x 2.25 m. A final harvest area (4 m x 8.5 m) is also marked. A locally
adopted cerea1 grain is planted at the recommended spacing and time of year.

As an exampleof this experimental approach, the rates of dry matter application of a low
quality litter (e.g. sugar cane bagasse, 0.27% N) and a high quality liUer (e.g. prunings from
a nitrogen fixing tree, 2.7% N) to supply 40 kg of N/ha are presented in Table 1.1.

Table 1.1 The quantities of litter inputs required to provide 40 kg/ha.

Litter quality --------------- Litter inputs (kg/ha)" ---------------

Sugar cane bagasse Tree prunings

High 1,482

Low 14,815

High + Low 7,407 741

Complete control

.. P, K, Ca, S, Mg, and micronutrients should he applied to ensure that only N limits
crop growth.

1.3.7 Measurements

These should be repeated over time on the same plot.

1 Daily rainfall
2 Water and total N content of organic residue and soil
3 Soil physica1 characteristics
4 SOM fractions (total organic C, microbial biomass C, soillitter C)
5 Decomposition (litter bag technique)
6 N mineralisation (in situ and/or anaerobic)
7 Plant growth, e.g. total above-ground biomass for annual crops

III
TSBF: A Hantlbook of Mdhods

8 Below-ground biomass
9 Crop yields
10 Nutrient content of above- and below-ground biomass
11 Population density of soil fauna, by functional groups, at the beginning and end of the
study

1.3.8 Analysis and interpretation

The experimental design requires that this experiment be analysed through analysis of
variance as a randomised complete block within sampling times and using regression
techniques when different treatments are being compared over severa! time points. The
comprehensive site characterisation and data collection also allows for the experimental results
to be compared to CENTURY model simulations of several treatments.

1.3.9 Transition to on-farm experimentation

Representatives of non-governmental organisations, national agricultural laboratories and

extension services should be invited to view the experiment towards its conclusion and discuss
the experimental results. The most promising experimental treatments can then be installed
into farmers fields in cooperation with interested individuals or organisations. It is advised
that such follow-up experiments be researcher-installed and farmer managed, and that only
crop yields and selected (few) soil parameters be measured at the conclusion of the
experiment.

1.4 SITE CHARACTERISATION

The TSBF Programme is in essence a methodology designed to offer a novel approach to soil
fertility research. A central feature of the TSBF approach however is the idea that in order
to understand soil fertility dynamics in a given system, a minimum set of characterisation data
are essential. A further key feature is to compare such data between research sites (this is
greatly facilitated by sharing a common methodology) as it has been realised for sorne time
that this would lcad to a more rapid advancement of the understanding of soil biological
processes, and improved management. Consequently, from the earliest stages of the
Programme, the development of this Handbook has been seen as a priority. The idea of a
minimum data set has evolved over the life of the TSBF programme, and continues to do so
as techniques and understanding improve. The balance and scope of this "minimum package"
have greatly benefited from input from a wide range of scientific disciplines, researchers
working in sites with very diverse soils, climates and vegetation, and from the rigour of
working in close collaboration with ecosystem modellers. The choice of parameters and
methods chosen is guided by their suitability for laboratories in the tropics, robustness, cost
and the amount of additional insight gained into processes operating within the ecosystem.
The variables requiring determination for TSBP Research Sites are listed in Table 1.2.
Variables were chosen so as to offer the minimum set of information required to make and
interpret inter-site comparisons of soil biologica1 processes at a gross level. Detailed studies
may weIl require additional measurements not included in the table.

8
 Chapter 1: The Tropical Soil Bi%gy and Fertiltry Programme, TSBF

Table 1.2 Site characterisation measurements for TSBF research sites.

1 Site Description

Present and past land use, topography and position on slope

n Site Variable Measurement

Variable Comment Units

CLIMATE

Mean monthly precipitation mm

Mean monthly maximum air temperature oC

Mean monthly minimum air temperature oc

sons
USDA Soil Taxonomy Family level

Depth To weathering parent material m

then for 0 - 20 cm:

pH 1:2.5 Water

Organic carbon Wet oxidation %

Total N and P H2S0iH202/Se digestion mg/kg

Exchangeable A13 + (+ H+) (if pH < 6.0) 1 M KCI Im'1~

Exchangeable K+, Ca2+ and Mg2+ 1 M NH40Ac pH 7.0 Im'1~

ECEC (if pH < 6.0) By summation of cations Im'1~

CEC (if pH ~ 6.0) 1 M NH40Ac pH 7.0/KCI Im'1~

Bicarbonate extractable P pH 8.5 mg/kg

Resin extractable P HC03-/CI- resin mg/kg

Organic P Difference on ignition mg/kg

9
l'SBF: Â Hondbook of M~lhods

Table 1.2 Continued

Micronutrients If suspected to be limiting or toxic mg/kg

Nitrogen mineralisation index Anaerobie lab mineralisation mglgld

Mierobîa1 biomus Chloroform fumigation/extraction mg/g

Soillitter Flotation in water mg/g

Mechanical analysis Hydrometer: sand, silt, clay %

Stone content (if > 5 % by mass) %

Field capacity Gravimetrie %

Bulk density Any suitable method g/cm3

VEGETATION

Description and local name for system e.g. Open Miombo woodland
Woody plants by dominant species:

Density no.lha

Canopy mean maximum height m

Basal area m2/ha

Herbaceous plams/crops:

Peak standing crop Harvesting @ 2 cm (state date) t'ha

Species contribution to peak standing crop Dry-weight-rank or sorted harvest %

Rools

Root profile By oounting m/m 3

Root mass Soil cores kg/ha

10
 Chapler 1: The Tropical SoU Biology and Fertility Programme, TSBF

Table 1.2 Continued

SOILFAUNA

Density By functional group

Biomass (fresh weight) By functional group

ORGANIC MATTER INPUTS

Woody plant litter Litter traps; reported monthly kg/ha

Herbaceous litter/crop residue Quadrats; reported monthly kg/ha

Roots [Specialist study] kg/ha

ORGANIC MATIER INPUT QUALITY

(For all inputs contributing > 10% of total input)

Total N H2SOiH20 2/Se digestion %

polyphenolics (for leaves) Folin-Denis %

Lignin via acid-detergent-fibre %

DECOMPOSmON

Decomposition standard lso d

Tree leaf litter Standard litter bags, tso d

Herbaceous Htter Standard litter bags, tso d

Litter turnover By calculation

SOCIOECONOMICS

Socioeconomic context Rapid Rural Assessment

11
 7SBF: A Handbook 0/ Mdhotb

Table 1.2 Continued

if applicable to the system:

HERBIVORY

Animal type

Animal mean live mass kg

Stocking level Per month kg/ha

FIRE

Date Month of year

Completeness % Of standing crop bumed %

Intensity From frre duration and fuel load kI/mIs

12

L~_
Chapter 2 SITE DESCRIPTION

2.1 CLIMATE

2.1.1 Climatic charaderisation of site locality

To characterise the climate of a site, long-term data (> 10 years and preferably, > 20 years)
are required, for at least monthly mean values of precipitation and temperature (maximum,
minimum). Wherever possible information on soil temperatures, relative humidity
(maximum, minimum), sunshine, wind (speed and direction) and evaporation should
complement the "Minimum Dataset". These climatic data should be obtained from the nearest
(or climatically Most similar) weather station with an indication of the duration of record for
each parameter and along with the particulars of the station such as name, latitude, longitude,
altitude. Further information on the distance of the study site from that meteorological
station, the relative altitudes, the aspects and topographical positions must be provided to
enable an assessment of eventual climatic differences.

2.1.2 Meteorological observations

The site of a meteorological station should be representative of the natural conditions in the
study area. It should:

1 be fairly level and free from obstructions;

2 have a short grass coyer or natura! coyer common to the area (not be concrete, asphalt
or rock);
3 not be c10ser to any obstruction (trees, buildings) than eight times that obstruction 1 s
height;
4 not be near areas with cold drainage, flooding or frequent sprinkling;
5 be easily accessible;
6 be fenced to exc1ude animaIs.

The daily observation times should be the same as used in the respective national
meteorological service. If only one daily observation can be carried out it should be the
moming observation. Minimum requirements for TSBF studies include:

Precipitation

Rainfall measurements should be made with standard gauges (WMO, 1983, Section 7) and
the readings made to the nearest 0.2 mm. The orifice of the gauge must be horizontal and
high enough to avoid splash from the ground, but low enough to prevent major wind
influence. To assure comparability, standards as used by the national meteorological service

13
TSBF: A Handbook of Melhods

should he followed. (Preference should be given to gauge types with at least 200 cm 2
receptive area.)

Âir temperature

To obtain reliable air temperature data thermometers (dry bulb, maximum, minimum) or
other probes used in connection with recording devices must be screened from direct radiation
but yet he exposed to free air movement. Standard wooden or plastic screens, available for
this purpose, are usually insta1led at a height of 1.2 - 2.0 m above the ground; the actual
height chosen at a particular project site should coincide with the one used by the national
meteorologica1 service. The thermometers have to be calibrated and should have a graduation
which a1lows areading accuracy of O.2°C.

Soil rempermure

Standard 50il thennomeœrs should be positioned to mea.sute the soil tcmpcrature at 0.10 m
depth. The thermometers have to he calibrated and should have a graduation which allows
a reading accuracy of O.2°C.

Âir humidity

Relative air humidity, defined as the ratio in percent of the actual water vapour pressure in
the armosphere to the one which cao be held at a given temperature (saturation vapour
pressure), can be directly measured with hygrometers or hygrographs. Absolute values of
the vapour pressure (measured in hectopascals) are detennined with psychrometers which
consist of two thermometen; one is measuring the air œmperature (dry bulb) while the othet
is covered by a wick kept moist with distilled, deionised water to depress the (wet) bulb
temperature. The wet bulb wick must he kept clean and checked frequently for rot. The
temperature difference hetween the dry and wet bulbs allows the calculation of the vapour
pressure. As ventilation affects results considerably, models with forced ventilation (Assmann
type) are strongly recommended. For recording purposes the Assmann psychrometer can be
supplemented by a hygrograph.

Minimum air temperlllure near che soil

In high altitude areas which are liable to frost it is recommended to measure the minimum
temperature ncar the ground. This is measured with a regular minimum thermometer e~posed
on a smal1 shaded support 5 cm above the ground.

Sunshine

Solar radiation is the basic meteorologica1 parameter for all heat and energy balance
assessments. If no direct measurements of the solar radiation are made, the duration of bright
sunshine allows for estimates of the energy available for physica1 and biologica1 processes.
The standard instrument to measure the duration of sunshine is the Campbell-Stokes recorder,
where the focused radiation from the sun bums a trace in a chart. Sunshine duration is
expressed in hours per day. As the sunshine patterns usually do not change over short
distances, data from nearby meteorologica1 stations may be used.

14
 Chapler 2: Sile Description

Mnd

Wind data are essential for estimations of evapo(trans)piration. The basic instrument for
measuring the wind speed is the cup anemometer; when equipped with a counting device it
allows measurement of distances of wind run over time, and thus the mean wind speed over
given time intervals can he derived. Data on wind speed should he complemented by
information on the wind direction which can he observed with a wind vane.

In agricultural meteorology the standard height for wind measurements is 2 m above the
ground but as topography and the surface coyer influence the wind field to a large extent,
observation sites have to he chosen very carefully.

Evaporation

To measure potential (maximum) evaporation directly, pans or open water tanks are exposed
to the environ mental conditions. The most commonly used evaporation pan is the Class A
pan which bas a diameter of 120.6 cm and a height of 25.4 cm. It is installed on a wooden
grid approximately 13 cm above ground level, with a wire mesh protecting it from animaIs.
The water level is controlled with a graduated hook gauge, set on a still well, or with a fixed
point to which the pan is refilled with a calibrated container each time a measurement is
made. The amount of water loss, analogous to rainfall, is expressed as the equivalent depth
(in mm) of wa~r evaporated.

Theoretica1 approaches, based on the surface energy budget, use measured data of sunshine
duration, temperature, air humidity and wind run to estimate the maximum water 10ss from
an open water surface, or from a surface covered with green grass well supplied with water
(reference crop evaporation).

Recent research has shown that correlated data from shaded Piche evaporimeters provide good
substitution for the aerodynamic term in formulas used for evaporation estimates. They
would replace humidity and wind estimates.

Reference

WMO No. 8 (1983) Guide to Meteorological Instruments and Methods of Observation. Fifth
Edition. World Meteorologica1 Organisation.

2.2 SOIL CHARACTERISTICS

Data are needed at two levels:

1 Information necessary to ensure that the study site is representative of the region

This should be based on available regional soil maps coupled with reconnaissance survey if
necessary.

15
TSDF: A Handbook. 01 Methods

2 Detailed information

Detailed information at a sca1e of 1:5000 or larger, is required for the actual site. The soils
should be classified at the Family level of USDA Soil Taxonomy (Soil Survey Staff. 1987),
and at least two (usually more depending upon the area) representative soil profile
descriptions should be given (Breimer et al., 1986) plus standard analyses (as listed in Table
1.2) of samples from 0-20 cm.

The soil parameters requiring measurement for TSBF can alsa be used to apply the Fertility
Capability Sail Classification System (Sanchez el al., 1982). This will help in the
interpretation of sail constraints in terms of sail biological processes. Soil analytical methods
are given in Chapters 6 and 7.

References

Breimer, R.F., van Kekem, A.J. and van Reuler, H. (1986) Guidelines for Soil Survey and
Land Evaluation in Erological Research. MAR Technica/ Notes 17, UNESCO, Paris.
Sanchez, P.A., Couto, W. and Buol, S.W. (1982) The Fertility Capability Soil
Classification System· Interpretation, Applicabilityand modification. Geoderma 27,
283·309.
Sail Survey Staff (1987) Keys 10 Soil TaxoMmy revised edition. Soil Conservation Service,
Washington OC.

2.3 LAND MANAGEMENT

2.3.1 Land use

It is relatively rare to find sites on which the natural vegetation and soils have not been
disturbed, to a greater or lesser degree, by land use. A description of the nature and intensity
of such use is therefore an essential basis for the study of soil/vegetation dynamics and should
be based on a classification plus a verbal description. The classification enables sites of
similar nature ta be related. The description enables account ta be taken of the wide diversity
of possible uses.

In the context of TSBF studies, it is particularly important to record the inputs and outputs
ta and from the ecosystem by the agency of man and domestic animaIs.

Area. site. present and pasl use

Land use is the use of land by man (including animaIs under control of man) for productive
andlor service purposes. In the present context "use" should be interpreted widely, ta include
bath land improvements (e.g. irrigation, drainage, terracing) and land degradation (e.g. forest
clearance, salinisation). The spectrum of land use extends from the considerable disturbance
of conditions through arable use at one end, to "undisturbed natural vegetation" as a limiting
case at the other. Two distinctions are made: that between land use of the area and of the
site; and that between present and past use.

16
 C1uJpler 2: Sile Description

The land use of the area refers to the surroundings within which the experimental site lies.
Land within a circle of 5 km radius around the site may he given as a guideline. It will
usually be necessary to describe the use in proportional terms, giving approximate
percentages, e.g. "30% arable (mainly cerea1s and legumes), 20% pasture and 50% natura!
woodland, used for grazing and collection of fuel wood and domestic timber". The land use
of the area does not directly concem the dynamics of the site, but provides a setting, enabling
it to be recognised as broadly similar to that in other countries. For example, "former
shifting cultivation, now changed by population pressure into semi-permanent arable use"
indicates a situation of widespread applicability.

The site refers to use on the actual site of the experimental work. Land use will normally
be the same over the whole site. Present land use is the use during the immediately preceding
year. It will frequently include differences in use as between wet and dry seasons, e.g.
"maize for 4-5 months commencing in the wet season, natural fallow with limited cattle
grazing for the rest of the year". Land clearance, cultivation (hand hoeing, animal or
tractor-powered ploughing, etc.), sowing/planting and harvesting methods should herecorded
with date. The full description of farming systems is a substantial task and the subject of a
wide range of publications (e.g. Simmonds, 1985).

Past land use refers to the trends of use in recent years. For the area as a whole, this
provides a further element of contextual setting. For the site, the intention is to coyer
whatever period may have appreciably affected the present condition of the soils and
vegetation. A 10-20 Year span should he included as a minimum, with brief reference to
earlier land use history where appropriate, e.g. "Forest clearance in this area is thought to
have taken place about 50-60 years ago; for the past 10 years, the site h~~ been under
semi-natural pasture, intensively grazed by cattle, sheep and goats". "

Classification (after Young, 1985):

1 Annual crops
Arable use for annual, or quasi-annual, crops. Include cassava, hill rice, vegetables
if on a field sca1e; exclude swamp rice, gardens.
2 Swamp rice
If irrigated, list also as irrigation.
3 Tree and shrub crops
Perennial crops, exc1uding field perennials.
4 Field perennials
Sugar cane, sisal, pineapple, bananas.
5 Gardens
Intensive production, of vegetables and/or fruit, on small plots.
6 Irrigation
Include rice if water brought to fields, but not if retention of rainfall only; include
irrigated pasture, forest.
7 Natura! pasture
Livestock production from natura! pasture, including nomadic grazing, ranching.
8 Improved pasture
Livestock production from substantially improved or sown pasture.

17
7SBF: A Handbook of Methodl

, Forestry: natura! forest

For timber and/or other products.
10 Forest plantations
For timber and/or other products.
Il Agroforestry
Trees interacting with crops and/or pasturesllivestock.
12 Wildlife' conservation
Flora and fauna, intentional use for this purpose.
13Water catchments
Intentional use for this purpose.
14 Engineering use
Any form of construction.
IS Unused
Natura! vegetation, with no significant use.

Formm

The following format for land-use classification is suggested:

Area Present use:

Classes (with approximate percentages):
Description:
Past use:
Description:

Site Present use:

Class:
Description:

Management; (c.g. methods of vegetation clearance, sowing/planting, weeding,

pruning, harvesting)
Inputs: (types, quantities, composition, timing and manner of application)
Organic material;
Inorganic material:
Water (irrigation):
Outputs: (barvest etc.; types, quantities, composition)
Past use:
Description:
Inputs and outputs: (general account, approximate quantities where possible)

References

Simmonds, N.W. (1985) Fanning systems research: A review. World Bank Technica1 Paper
No. 43. World Bank, Washlngton OC.
Young, A. (1985) An environmental dotabase for agroforestry. ICRAF Worldng Paper 5
(revised edition), ICRAF, Nairobi, Kenya.
 Chapter 2: Site Description

2.3.2 Fire

The required data are the date, completeness and intensity of the bum. Completeness is a
visual estimate, or the difference hetween a before and after tire sample, of the percentage
of the litter, herbaceous and woody standing crops consumed by the tire, and thus estimates
of carbon loss. Intensity is a measure of the rate of energy release by the tire related to
nitrogen and phosphorus losses. It cao he calculated from the fuelload (F, kg/m 2) , the
energy content of the fuel (E, kJ/g: for most biomass this cao he assumed to he 20 kJ/g), and
the rate of spread (V):

V (mIs) = A 1 (w x t)
Intensity (kI/mis) = F x E x V

where

A = area of plot (m 2)
w = width of flame front (m)
t = time to bum plot (s).

Altematively, intensity cao he estimated from the following rough guide (Trollope, 1984):

Category Intensity Description

(kJ/mls)

Wann >500 Flames < 1 m, easilyapproachable

Hot 500-1 ()()() Plames 1-1.5 m, approachable

Very hot 1000-2000 Flames 1.5-3 m, difticult to approach

Extremely hot >2000 Flames > 3 m, uncontrollable

In keeping with the peak-trough philosophy of vegetation monitoring, herbaceous biomass

(fuelload) should be assessed before buming.

Reference

Trollope, W.S.W. (1984) Fire Behaviour. In: Booyessen, P. de v. and Tainton, N.M.
(eds.) Ecological Effects of Fire in South African Ecosystems. Ecological Studies 48
Springer-Verlag, Berlin.

2.3.3 Socioeconomic components

The TSBF programme recognises the need to focus local research programmes on local land
use practices. The overa11 objective of TSBF is to determine management options for

19
TSBF: A HanJbooJ: 01 Ml!chods

improving tropical soil fertility through the manipulation of biologica1 processes. Within the
range of more specifie TSBF objectives, there is a need to ehoose those that have local
applicability .

There are different approaehes and levels of investigation required in a program of research
on soil fertility. Basie researeh, or strategie researeh, on soil biologica1 fertility provides the
information needed for target researeh and extension programs assisting farmers in coping
with fertility constraints sueh as low productivity due to soil erosion. The application of the
principles leamed from strategie researeh to the local soil fertility problems as defined by the
target population requires an understanding of the socioeconomic context of the farming
systems in which they operate.

There are many techniques used to gather socioeconomie data. One of these is participant
observation. Participant observation allows the researeher to gather an in-depth knowledge
of the indigenous farmers' practices and beliefs. It can be quite time consuming and is better
suited to smaller populations. A second technique is the use of questionnaires which are
standardized, struetured and suited to quantifying socioeconomic data of generally large
populations (for example, see Moran in TSBF Handbook of Methods, first edition). A
standardised questionnaire cannot respond to unanticipated site-specifie problems (Beebe,
1985) and the analysis cao be quite time consuming. A third technique is in-depth interviews
whieh are used to gather data using clearly specified criteria genera11y aimed at a better
understanding of specifie problems or processes. This technique uses a flexible data
collection/processing procedure, generally focuses on smaller populations (L. Llambi,
Caracas, Venezuela, 1988, personal communication), and can be adjusted to time constraints
and site-specifie problems. TSBF researehers used a structured questionnaire, proposed in
the 1986 TSBF Yurimaguas Workshop by Moran (1986). They concluded that struetured
questionnaires were not expedient in performing multidisciplinary land use surveys. It was
proposed that in-depth interview techniques such as Rapid Rural Appraisals whieh pay
attention to context be used as effective procedures for recognizing constraints and
opportunities for land users.

Rapid Rural Appraisals were designed to use observation and in-depth interviews to gather
data on agricultural systems in developing countries. Rapid Rural Appraisal (RRA) is an
interdisciplinary, iterative approach use<! to gain insight into land use practices. Small teams
of researchers representing different disciplines such as social seience and soil science work
together using techniques such as in-depth interviews, based on key indicators, and direct
observation to gain insight into management practices. It is a highly iterative approach which
allows the reformulation of hypotheses based on new information (Grandstaff et al., 1987).
The real value of techniques such as RRA are the use of indigenous technical knowledge
(ITK) such as the farmers' knowledge of plants, soils, and the local ecosystem to reformulate
these hypotheses. The relevance and validity of these rapid techniques are evaluated in the
short term through repeated iterations of the RRA process and in the long term by the
effectiveness of aetions based on knowledge and perceptions ~enerated by the RRA.

Organisations sueh as ICRAF (Raintree, 1987) and IBSRAM (1988) are successfully using
techniques such as RRAs to assess soil management techniques and other agroeçosystem
conœms in developing countries.

20
 Chapter 2: Site Descriprion

An example of an RRA is given in Appendix A. The decision to choose between participant

observation, standardised questionnaires, and in-depth interviews is dependent on the timing
of the project (Murphree, 1989) and the need for flexibility vs. standardised quantifiable data.
If time is not an issue, participant observation or standardised questionnaires are possible.
In-depth interviews cao also be used as a starting point with the standardised questionnaire
being used as the follow up. When time is an issue, in-depth interviews combined with
observation and repeated iterations cao be modified to suit the particular needs of a project
such as TSBF. U1timately the intended use of the data and the time allotted for data
collection should guide the choice of data collection techniques.

References

Beebe, J. (1985) Rapid Rural Appraisal: The Critical First Step in Farming Systems
Approach to Research. Farming Systems Support project, Networking Paper No. 5,
USAID, Washington, OC.
Grandstaff, S.W., Grandstaff, T.B. and Lovelace, G.W. (1987) Draft Summary Report. In:
Proceedings of the 1985 International Coriference on Rapid Rural Appraisal. Khon
Kaen, Thailand: Rural Systems Research and Farming Systems Research Projects.
IBSRAM (1988) Methodological guidelines for IBSRAM Soil Management Networks.
Second draft, April. IBSRAM, Bangkok, Thailand.
Moran, E.F. (1986) TSBF Inter-regional research planning workshop. In: Swift, M.J. (ed.)
Biology International Special Issue 13, IUBS, Paris, 53-64.
Murphree, M. (1989) Socio-economic research components in tropical soil biology and
fertility studies. Biology International Special Issue 20, 31-40.
Raintree, J.B. (1987) D and D User's Manual. An introduction 10 Agroforestry diagnosis
and design. ICRAF, Nairobi, Kenya.

21
Chapter 3 FIELD PROCEDURES

Note: For in situ N-mineralisation, see Section 6.7.2.

For in situ soH physical determinations, see Chapter 7.

3.1 ABOVE-GROUND VEGETATION

The methods described below are recommended for use by researchers engaged in TSBF
studies for site description. Intensive sampling should be confined to the actuaJ study site,
but the surrounding vegetation should be described in general terms (extent, dominant species,
overa11 structure, ecological affinities etc., and local name, vegetation type) maldng particular
note of past management history (e.g. age of stand, fire history, grazing etc.). A general text
covering vegetation description is Mueller-Dombois and EIlenberg (1974).

Reference

Mueller-Dombois, D. and Ellenberg, H. (1974) Aims and Methods of Vegetation Ec%gy.

Wiley, New York.

3.1.1 Woody vegetation

The desired information is the above-ground woody biomass, but since it is seldom practica1
to measure this directly, an indirect approach using the stem basal area (m2) and height (m)
of a sample of woody plants is used. In order to extrapolate the sample to a per hectare
basis, the total basal area per species per hectare (m2/ha) and the plant density (plants/ha) are
also required.

Species

Species should be identified by their full scientific name (Genus species Author).
Unidentified species should be given a code or local name, and a reference to a specimen in
a herbarium.

In natura! communities with a large number of species, it is not necessary to record the height
and basal area of minor species individua1ly. AlI the species contributing less than 10% of
the total biomass cao be combined into one category.

In extremely diverse communities with no c1ear biomass dominance, where no species

contributes more than 10% to the total biomass, it may be necessary to define the "species"
on a functional basis; for instance emergents, canopy trees, sub-canopy trees and understorey
shrubs, recording which species were placed in which category.

22
 Chapler 3: Field Procedures

Stem density: (i) Plot methods

If the trees are sparse, then all the trees within the study site can be individually counted and
measured. Usually this is not the case, and one or more subsamples inc1uding about 100
trees are taken.

A square quadrat of known area can be laid out with the aid of a tape measure, a compass
and string, but often a circular quadrat (ail the trees within a given distance of a centre peg)
is easier to work with.

A long thin quadrat (belt transect) is often an efficient method in dense vegetation. Lay out
a tape measure (50 m) in a straight line, and record all the trees within a given distance (1-5
m, measured with a hand-held rod) on either side of the tape. Both the length and width of
the transect can be varied to give a total sample of about 100 plants; different widths can be
used for common shrubs and sparse trees within the same sample. Individuals (or
multistemmed clumps) more than half within the plot boundary should be counted in. To
permit biomass estimates based on individual trees at a later stage, data should be recorded
as in the following example:

Plots are 5 m x 50 m

Species Height (m) Diameter (cm)

Burlœa ofricana 3.2 15 176

Ochna pulchra 1.5 5 19
Sclerocarya birrea 6 30 707

Stem density: (i) Plotless methods

Laying out plots is time consuming and can lead to error due to subjective placement of the
plots. A variety of plotless methods are available, such as the Point-Centre-Quarter method
(PCQ; Cottaro and Curtis, 1956). These methods use the distance between random points and
the nearest plant to calculate an average inter-plant distance (d x). The number of plants per
hectare (N) is then given by

N = 10000
, 1 d x2

where d x is in metres.

About 100 distance measurements are required, and at the same time the species, stem
diameters and heights of this random sample of plants are recorded.

The PCQ method is not applicable to vegetation which is regularly spaced (such as
plantations) or highly c1umped; various modifications, such as the two-distance methods of
Diggle (1975) or Cox (1975) must be used.

23
TSBF: .A. Handbook of Methods

Height

Height is the vertical distance between the ground and the highest living part. For trees taller
than 5 m an estimate 10 the nearest m is sufficient; smaller trees should be recorded 10 the
nearest 0.1 m. A measuring rod is the easiest method in low vegetation. In taU vegetation,
where the top and the bottom of the tree can he simultaneously observed, a clinometer can
be used; otherwise any practical method can be applied.

The Mean maximum height for each vegetation layer (stratum) should he estimated from the
data. For example, for a primary tropicallowland forest the emergents May reach 60 m, the
canopy trees 40 m, the subcanopy trees 20 m and the understorey shrubs 6 m.

Stem basal area

The stem basal area (A) is the cross-sectional area of the trunk (m2) measured at the lowest
point not influenced by basal swellings or huttresses. In forests it is traditionally measured
at 1.3 m above ground leve1 ("diameter at breast height" t DBH)t but in tropical forests the
trees May still be buttressed at this height, and in savannas, branching frequently occurs
below this height. The area can be calculated either from a diameter (d) measurement using
callipers, or from a circumference (c) measure using a tape.

A = 0.7854 d2 or
A = 0.0796 c2

The result will be in m2 if the diameter and circumference are measured in m; if measured
in cm, divide the answer by 10,000 to convert cm2 10 m2 • If the stems are very markedly
non-cylindrical (for instance if they are elliptical or convoluted) an empirical correction factor
should be calculated.

Where a single plant has more than one stem, the area of each stem should be measured. For
shrubs with very Many small stems an estimate of the total combined area is adequate.

The total stem basal area for the plot is calculated by multiplying the stem basal area by stem
density for each spccic;s (according to the criteria fot species as defined. above).

Trtt bioflf(lS$ tstimll1ion: 1U1lural vegetll1ion

Tree biomass is Most precisely estimated by cutting and weighing all portions of a tree. As
this is very time consuming, tree biomass is more commonly estimated by statistical means,
e.g. by sub-sampling, or on estimates based on tree attributes. When sub-sampling is not
possible, biomass is Most commonly estimated from a regression equation.

Since mass is the product of volume and density. tree biomass regressions often use diameter
at breast height (D, cm), total height (H, m) and some measure of specifie wood density (S,
tJm3). A locally developed equation is generally Most reliable, and should be used whenever
possible. When no local equation is available, one of the regression equations given below
cao be used. These have been derived from several data sets and rigorously tested for
reliability (Brown et al., 1989). Equations using D, H and S are the Most reliable, followed
by those using D and H, or just D. When H is not available one May apply double sampling

24
 Chaprer 3: Field Procedures

methods, and estimate H as a function of D, then biomass as a function of D and estimated

H. This is only better than approaches based on D alone when the H-D relationship
accurately reflects the population of interest.

The regression relationships vary slightly with climate, so there are three equations reflecting
three rainfall regimes: dry < 1500 mm, moist 1500-4000 mm and wet > 4000 mm. Not all
equations are available for all climate types, due to lack of reliable data.

Equations may be applied to individual trees, mean trees or stand tables (numbers of stems
per sire classes) to estimate total above ground biomass for a given tree, for a mean tree or
per unit area respective1y (cf. Wharton and Cunia, 1985, for further details of methods).

The following equations estimate biomass (B kg) and total height (H m) for trees in tropical
forests. Values for the samp1e sire, n, and goodness of fit, R 2adj (an adjusted R2) are
inc1uded. The non-linear models inc1ude the adjustment term (MSFJ2) together with the
intercept, when appropriate.

Forest type n R2adj

Dry B = 34.4703 - 8.0671 D + 0.6589 D2 32 0.67

H = exp{0.6711 + 0.6301 ln D} 1231 0.54

Moist B = 38.4908 - 11.7883 D + 1.1926 D2 1680 0.78

B = exp{- 3.1141 + 0.9719 ln (D2H)} 168 0.97
B = exp{-2.4090 + 0.9522 ln (D2HS)} 94 0.99
H = exp{1.0710 + 0.5677 ln D} 3824 0.61

Wet B = 13.2579 - 4.8945 D + 0.6713 D2 69 0.90

B = exp{-3.3012 + 0.9439 ln (D2H)} 69 0.90
H = exp{1.2017 + 0.5627 ln D} 69 0.74

Note: exp {... } means "e raised to the power of {... }"

It should be stressed that although these equations have been derived from large (and
sometimes very large) samp1e sires, they should he viewed with sorne caution until validated
for a given system.

Tree biomass estimation: even-aged plantations

Woody biomass estimation in even-aged plantations is simpler than in natural forest, and
requires destructive sampling of many fewer trees. The same principle, of estimating biomass
from predictor variables such as stem cross-sectional area and height, applies, but the model
relating these variables to biomass cannot be assumed to be the same for different species and
trees of different sires.

25
TSBF: A Handbook of Methods

For many species, for a given stand structure, the following linear function gives an accurate
prediction of biomass (Stewart et al., 1992):

w = ~j + bij Ld2
where
w = biomass
d = stem diameter (Dt2 being the sum of the squared diameters of all the stems of a multi-
stemmed tree)
a and b are coefficients for species i on site j.

These coefficients have to be determined separately for each species and each site.

Procedure

1 Feil a minimum of 12 trees.

2 Partition the trees into components (wood of different sizes, leaves, and pods/fruit).

3 Weighed the components immediately in the field.

4 Oven dry subsamples (> 100 g) of each to determine water content.

5 Calculate total dry weight (biomass):

The biomass and diameter(s) of the sample trees are used to determine a and b by regression
analysis. If R2 is low « 0.8) other models should also be tested. The biomass of the other
trees in the plantation is then estimated from their diameters. For a small stand this may be
done for every tree individually; for larger populations the total biomass may be estimated
using the mean diameter and the total number of stems in the stand.

Twelve is the absolute minimum number of trees per species that must be destructively
sampled, but if a larger sample is used it will give a more reliable model with closer
confidence limits. Whatever the sample size chosen, it is essential that the sampled trees
coyer the whole size range of the population. The usual procedure is to use a stratified
sample in which the trees are divided into size classes on the basis of diameter (or cross-
sectional area in the case of multi-stemmed trees), and sample trees are se1ected randomly
within each size class.

For studies on a single site, the inclusion of tree height h in the regression function is often
found to con tribu te little to the goodness of fit of the model. However, a linear model of the
form

is more likely to applicable across sites than one using diameters only, because height, unlike
diameter, cao be assumed to be independent of stocking level and silvicultural treatment
(Philip, 1983).

26
 Chapter J: Field Proced#res

References

Baskerville, G. L. (1972) Use of logarithmic regression in the estimation of plant biomass.

Canadian Journal Forest Science 2, 49-53.
Brown, S., Gillespie, A.J.R. and Lugo, A.E. (1989) Biomass estimation methods for
tropical forests with applications to forest inventory data. Forest Science 35, 881-
902.
Cottam, G. and Curtis, T. (1956) The use of distance measures in phytosociological
sampling. Ecology 37, 451-460.
Cox, T.F. (1975) The robust estimation of the density of a forest using a new conditioned
distance method. Biometrika 63, 493-499.
Diggle, P.J. (1975) Robust density estimation using distance methods. Biometrika 62,
39-47.
Philip, M.S. (1983) Measuring trees and forests. Tanzania/Dar-es-Salaam. Division of
Forestry, University of Dar-es-Salaam.
Stewart, J.L., Dunsdon, A.J., Hellin, J.J. and Hughes, C.E. (1992) Wood Biomass
Estimation in Central American Dry Zone Species. Tropical Forestry Paper 26,
Oxford Forestry Institute, Oxford. 83pp.
Wharton, H.E. and Cunia, T. (1985) Estimating tree biomass regressions and their errors.
USDA Forest Service: Northeastem Forest Experiment Station, Broomall, PA.
Report NE-GTR-117.

3.1.2 Herbaceous plants and short duration crops

The required measurements are total biomass (tJha = 0.001 x kg/ha = 0.01 x g/m 2) and the
contribution (%) of each species to that total. Total biomass is measured by harvesting,
drying and weighing a number of small subsamples, or quadrats. Quadrat size depends on
plant spacing: 0.5 m x 0.5 m is a convenient size for most grasslands; 1 m x 1 m may be
appropriate for crops such as maize. Sample number (n) should be sufficient to reduce the
standard error to about 10% of the mean. 20 - 30 samples per treatment, distributed amongst
replicates, is usually sufficient. Sample location should best be random, but systematic with
a random start is acceptable.

Procedure

1 Cut all herbaceous vegetation within the quadrat at 2 cm above the ground (to avoid
soit contamination), and sort into live (green) biomass and standing dead if possible.
2 Collect the liUer from the ground for an estimate of litter standing crop.
3 Dry all samples as soon as possible to prevent decomposition.

Species composition in mixed communities is estimated by the dry-weight-ranking technique.

The technique is based on a multiple regression for the dry weight of a mixed sample of
herbage on the weights of the three heaviest species in the mixture. Experience indicates that
it is easier to assess visually the rank order of the species in a quadrat than to estimate
accurately their biomass. Tests in a large number of different communities have shown that
the regression coefficients are fairly consistent between communities, and therefore do not
need to be recalculated each time. The technique does not work well in communities

27

III
~F: A. Handbook of Methods

completely dominated by one species, and tends to ignore rare species. The modified form
given here, where the total biomass within the quadrat is also given a visual score, gives
better results in communities where the total biomass is patchily distributed. Quadrat size
should be small enough that species ranking is simple, but large enough that Most quadrats
have at least three species in them. Quadrats 0.5 m x 0.5 m square are usually adequate in
grasslands. About 50 quadrats should be assessed per treatment.

Procedure

1 Walk around the plot to obtain a clear visual impression of what the minimum (1) and
maximum (5) quadrat biomass looks like.

2 Locate the quadrats randomly or systematically after a random start.

3 In each quadrat (i) give the total biomass a score (w j) between 1 and 5 according to
whether it is near the minimum or maximum for the plot.

4 In each quadrat, give the species (j) which contributes Most to the total quadrat
biomass a rank score (rij) of l, the second heaviest species a rank of 2, and the third
heaviest species a score of 3. If a single species contributes more than about 70% of
the biomass, give it ranks 1 and 3, or 1 and 2 (or even 1, 2 and 3 if it is the only
species in the quadrat). Similarly the second species could get a 2 and 3 if necessary.

5 When all the quadrats have been scored and ranked, calculate the score for each
species:

Calculation

where
l:w j is the sum of the quadrat scores for the quadrats where species j obtained rank r.

Add up the species scores to give a total score.

Determine the % contribution by species j to the total biomass:

Species j contribution to biomass (%) = (species score / total score) x 100.

Further reading

Gillen, R.L. and Smith, E.L. (1986) Evaluation of the dry-weight-rank method for
determining species composition in a tal1grass prairie. Journal of Range Management
39, 283-285.
Jones, R.M. and Hargreaves, J.N.G. (1979) Improvements to the dry-weight-rank method
for measuring botanical composition. Grass and Forage Science 34, 181-184.
Sandland, R.L., Alexander, J.C. and Haydock, K.P. (1982) A statistica1 assessment of the
dry-weight-rank method of pasture sampling. Grass and Forage Science 37,263-272.

28
 Chaprer 3: Field Procedltres

3.2 ABOVE-GROUND INPUTS

3.2.1 Tree and shrub litter

In a comprehensive review of tropicallitter fall data, Proctor (1983) observed that the results
of many published studies were not comparable. This resulted from inadequate siting and
replication of traps in relation to site heterogeneity, sampling for periods of less than a year
and lack of standardisation of smalllitter fractions Oeaves, twigs, reproductive structures and
"trash").

Litter trap construction

Litter traps are bags or boxes supported just c1ear of the ground with an aperture of 0.25 -
1 m2 • A circular construction is best as it minimises edge effects. Woven plastic bags are
light weight for use in remote sites and cao be tensioned into shape using lines attached to
D-rings sewn around the mouth of the bag. The traps must allow free drainage of rain water
but have a mesh size of approximately 1 mm or less to retain fine litter fractions. Trays on
the ground surface cao be used to measure litter-fall from dwarf shrubs etc., but animal
activity, drainage and wind cao present problems.

Similar considerations apply to collections of litter from quadrats on the ground, which may
be necessary for estimating falls of palm fronds and larger woody litter. Trash fractions,
however, which often have low mass but high nutrient content, will be lost by this method.

Procedure

1 Randomly locate litter traps (for material other than branches) within moderately
homogeneous plots, or in a stratified random pattern (with 10 traps per subplot) in
sites where it is necessary to inc1ude major variation in topography, soils and
vegetation structure.

Note: To achieve a 5% standard error about the mean, Newbould (1967) recommends the
use of at least 20 trapslplot. In very heterogeneous sites, however, higher numbers
of traps may be required.

2 Collect litter every 2 weeks and air dry il. More frequent collections may be
necessary for litters which decompose rapidly, e.g. sorne tree legumes, while less
frequent collections may be made under dry conditions (though the possibility of the
litter becoming contaminated with du st and/or animal faeces should be recognised).

3 Sort the dried material into:

a) leaves (inc1uding petioles and foliar rachises);

b) small woody litter (twigs < 2 cm in diameter and bark);
c) reproductive structures (flowers and fruits could be differentiated);
. d) trash (sieve fraction < 5 mm).

29
 1
1

TSBF: A. Handhook of Method.J

(For palm fronds, the leaflets, the rachis below 2 cm, and the remaining parts of the
rachis should be weighed and recorded separately.)

4 Oven-dry subsamples of litter to obtain correction factors for moi sture content (see
Section 6.1)

5 Express al1 fractions defined above on an oven dry basis in g/m2/year or tlhalyear
with 95 % confidence limits.

6 Estimate branch fall from large (e.g. 100 m2) ground quadrats. Break twigs at the 2
cm diameter point, weigh the > 2 cm diameter material, subsample for oven-dry
mass and other determinations as required.

References l
J
Newbould, P. J. (1967) Methods of Estimating the Primary Production of Forests.
Blackwell Scientific Publications, Oxford. 1
Proctor, J. (1983) Tropical Forest Litterfall 1. Problems of data comparison. In: Sutton, i
S.L., Whitmore, T.C. and Chardwick, A.C. (eds.), Tropical Rain Forest: Ecology l
'j
and Managemem. Blackwell Scientific Publications, Oxford, pp.267-273
11

3.2.2 Herbaceous litter and above-ground crop residues

The minimum level of sampling is at maximum and minimum biomass associated with major
seasonal changes or perturbations; i.e. sampling four times a year under climatic regimes with
1
a strongly bimodal pattern of rainfall. This will underestimate litter inputs as a consequence j
of material turning over between sampling dates and sampling at regular intervals every few
weeks is recommended. 1,
Procedure

1 Determine herbaceous litter (including grasses and forest ground flora) by harvesting
quadrats in conjunction with biomass estimates (Section 3.1.2).

2 Separate litter, where possible, by plant species for the MOst frequent 80% of species
and bulked for the remaining 20%. (This May be impractical in very species-rich
communitics.)

3 Determine crop residues after harvest and at the time of ploughing or other
manipulation.

4 Oven-dry liuer subsamples to obtain correction factors for moisture content (see
Section 6.1). Litter heavily contaminated by soil May need to be corrected for 'ash'
content as weIl (see Appendix D 2.6).

5 Express all fractions defined above on an oven dry basis in g/m2/year or tlha/year
with 95 % confidence limits.

30
 Chapt~r 3: Fidd Proc~dur~s

3.2.3 Other inputs

Mulches, manures and fertilisers

Information on the type of input applied, the amount applied, and the date of application is
necessary. For example, information required for mulches would include the type of mulch
(e.g. rice straw), the amount applied on a dry weight basis (t/ha), and the application date(s).
A representative subsamples cao be taken for nutrient and other analyses. Details on methods
of incorporation (e.g. surface application, ploughed to a certain depth, banded) of each of the
inputs should also be noted. Similar information is necessary for manure application.

For fertilisers, specify the material applied (e.g. urea, single superphosphate), quantity applied
and application date(s).

Herbivory

Record the dates of commencement and termination of feeding, the type of herbivore (species,
browser, grazer, etc.), the number of animals in the plot and estimates of their individual
masses. These data will permit the stocking level (kg/ha) to be calculated, given the plot
area, and the rates of plant consumption, defecation and urination (kg/ha/day) to be estimated.

3.3 ROOTS

Note: This section gives procedures for the study of the overall root pattern. A review of
Methodologies for estimating root biomass, production and estimating total root carbon
input to the soil May be found in Appendix D ("Roots: length, mass, biomass,
productivity and mortality").

Root research requires destructive sampling of the soil, often causing considerabledisturbance
to the plots. Space should be allowed for this when designing field experiments. Specific
information of root systems (both vertical and horizontal) is important when trying to
understand plant/input interactions and the correlative effects on soil fauna and
microorganisms. Information on the lateral spread of root systems is essential in deciding on
"guard" areas or borders in field experiments. Errors in interpreting results are easily made
when no mot information is available, as lateraI spread of over 5 m (cassava) or up to 20 m
(certain trees) is often more than expected. Information on rooting depth and distribution is
also essential when placement effects are to be considered regarding plant nutrient uptake.

Apart from the two classical descriptions of root methods by Schuurman and Goedewaagen
(1971) and Bôhm (1979), a number of recent reviews is available on methods to quantify root
development and functioning in the field: Caldwell and Virginia (1991), Mackie-Dawson and
Atkinson (1991), Taylor et al. (1991), van Noordwijk (1987), van Noordwijk et al. (1992).

Three methods outlined below are based on the study of soil profiles from a soil pit. In soils
without stones or woody roots a monolith sampler as described by Floris and van Noordwijk
(1984) might have the advantage of less site disturbance, but generally soil pits are needed
to have access to the root environ ment.

31
TSBF: A HandbooJ: of Melhods

3.3.1 Root preparation on pronle walls

A soil pit is dug close to the plant selected for study and by carefully removing soil close to
the stem sorne main roots are identified; their course is followed by gradually removing the
surrounding soil, using a pin. When all (major) fOOts within the tirst, say, 10 cm from the
original profilé wall have been exposed, a drawing is made e.g. on a 1:5 scale on graph paper
(Figure 3.1), using pins to mark grid points in the soil. Unfortunately most fine fOOts will
break off during this procedure, and so only a qualitative picture is obtained. Still, the
method allows width and depth of the fOOt system and branching pattern to he recorded. The
response of the root system to heterogeneities of the soil, to transitions between soil layers,
to cracks (clay soils) and channels in the soil (made by soil fauna and/or previous roots)
deserves special attention.

OepUl
cm

10

20

30

40

50 Sovabean

10

20

30

40
Maize

50

Figure 3.1 Example of fOOt observations on profile wall, showing the response of rnaize
roots to an acid subsoil (below 15 cm) and to the presence of old tree root channels (van
Noordwijk et al., 1991a).

In mixed cropping situations roots of the various components can usually be recognized after
sorne training. By spending 1 to 2 days per major species the overall patterns of shallow and
deep rooted species and lateral spread can he estimated: both are important when designing
fertilizer or crop residue experiments or when designing mixed cropping systems. Results
of this method are easily understood and can be used in discussions with e.g. farmers.

For shallow rooted crops lateral roots can also be observed frQm the stem base by digging
a small trench, following the fOOt. The presence of "sinker" roots, venica11y oriente<! branch
roots from a horizontal branch fOOt, is of special interest here. By excavating a small area
around the stem a classification of roots by diameter and orientation can be made (van
Noordwijk et al., 1991b). This way replicated observations can be made and a "typica1"
specimen can be selected for further observation.

32
 CluJpler 3: Field Procedures

3.3.2 Root mapping on proftle walls

Based on results of root pattern observations (described above) an estimate of total root
density in the profile can he obtained by root mapping on polythene (pYC) sheets. ProfIle
waIls are cleaned and straightened with a sharp Imife, after having removed stones and thick
roots which would hinder the process. Depending on soil texture, various steps can he taken
to facilitate root observations:

• on sàndy sons a water spray (knapsack sprayer) can he used to remove about 2 mm
of soit to expose roots,

• on clay soils gently brushing the son May help. Good results have aIso been obtained
covering the profile wall ovemight with a cotton cloth soaked in a 5 % sodium
hexametaphosphate solution to disperse the clay prior to spraying. In dry soils
compressed air can aIso he used to blow away crumbs of soi1; roots are remarkably
resistant to such treatment.

After preparation the profile is covered with a transparent PYC sheet and carefully searched
for roots; a predrawn 10 cm x 10 cm grid on the PYC helps to work systematically. AlI
roots are mark:ed with dots on the sheet; differently coloured pens can he used for different
size classes or plant species. Roots are only recorded where they intercept the plane of
observations; branch roots outside that plane, exposed by preparation of the profile waIl, are
neglected. Major features in soil structure and horizon boundaries should aIso be noted
(Figure 3.2 a).

(As condensation build-up hehind the PYC sheet May present a problem in some instances,
a wire (or fine string) grid May be used: the grid is made on a light-wooden frame (1 m x
0.5 ml, with the wires stretched across at 10 cm spacings. (A suitable wire mesh could he
used.) The coordinates are noted rather than marking the PVC sheet. A comparative
advantage of this method is that it aIlows a closer examination of the root condition and soil
features; a detailed spatial patterning is, however, harder to record.)

Root intensity (numher of interceptions, N, per unit map area) can now he estimated for each
soil horizon, as a function of distance to the plant. A more detailed analysis of root pattern
(regular, random or clustered) within each horizon is possible by computer analysis of the
root maps. Statistical tests of spatial correlation hetween roots and other map features (e.g.
cracks, tree roots, termite channels) can be performed on the buis of root intensity in zones
with increasing distance to the features of interest (van Noordwijk et al., 1992). The relation
between root distribution and the pattern of cracks and macropores can he studied by
t infilttating a Methylene blue solution (0.05 g/l) into the soil hefore digging an observation pit.
~.
t Counts of N per horizon indicate the relative decrease of root intensity with depth. Root
length density, 4v (cm root length per cm3 of soU) can be estimated for each horizon:

L rv =2xFxN

where the calibration factor F is ~ua1 to 1.0 for roots with random orientation, if aIl roots
are seen. The factor F depends on root orientation (between 0.5 and infinity) and improved

33
TSBF: A. Handbook of Methods

a. Root map b. Block sample

o
root A Be B
• .... c. Wash on
- termi te nest Une sieve

• d. Calibration line
o
o
• a

Lrv
cm/cm3 • •
o.
• •
• •
•
•
•• •

N. dots/cm 2
FIgUre 3.2 Root mapping on pve shoots; roots A and B might refer to different diameter
classes or species.

estimates can he obtained if mot intensities in horizontally oriented planes are recorded as
weIl (van Noordwijk, 1987). Experimental calibration factors may be considerably (e.g.
three-fold) larger than theoretical ones.

Problems with the method are: (i) roots of different plants are hard to distinguish. (H)
distinction of live and dead roots is not easy and (üi) a considerable fraction of fine roots may
be overlooked. To obtain (semi)quantitative results it is necessary to calibrate the maps by
taking small blocks of soil (e.g. 20 x 10 x 1 cm volume) from various layers on the root map
(Figures 3.2 b, c, d), wash them over a fine sieve (0.3 mm mesh) and determine root length
(see below).

3.3.3 Pinboard monolith sampling

Monolith samples can be obtained with pinboards ("fakir beds"), made by inserting U-shaped
pins (made from stainless steel) in plywood (Figure 3.3; further details are given by
Schuurman and Goedewaagen, 1971, and Bohm, 1979). The size of the pinboard is
determined by the crop (based on previous observations, such as footing depth and
distribution and practica1 considerations (samples of 100 x 60 x 10 cm of soil will weigh
about 100 kg). By washing the soil away the roots become exposed and cao be observed.
If a coarse mesh screen (e.g. material used for TSBF liner bags) is put on the pins before the
board is pushelinto the soil (perpendicular to the crop row), this screen cao help to keep the
roots in their original location while washing the sample. Washing the sample can he
facilitated by soaking ovemight in water, deep freezing (for clay soils), soaking in oxalic acid

34 ;

j
 Chopler 3: Field Procedwes

Stainless steel wire ~

10 em

Car jack

mm mesh sereen

Figure 3.3 Pinboard construction and sampling in the field; after pushing the pinboard into
a plane prome wall, the board is supported on a car jack and soil below and to the sides of
the board is cut away; extending pieces of the mesh screen can be used to tie the soil block
to the board; finally the back side of the sample is cut by "sawing" movements with a
stainless steel wire.

(for soils with free calcium carbonate) or soaking in hexametaphosphate, preferably under
vacuum. Whatever the pretreatment used, gentle washing must follow.

After washing away the soil, the root system cao he lifted on the mesh screen, photographed
(on a black cloth as background) and/or cut according to soillayers (indicated e.g. by a string
between the pins while washing the sample), depth zones and/or distance to the plant to obtain
root biomass and/or root length (see helow for root length). To estimate total biomass per
plant root weight density per zone and 4epth bas to he integrated over the relevant volume.

Although the pinboard method is more time consuming than methods 3.3.1 and 3.3.2 above,
it gives good information per unit effort spent. Major weaknesses are that roots may break
or be displaced during washing. The distinction between live and dead roots is easier via
pinboard sampling then in methods where the root system is not sampled in its entirety.

References

Bôhm, W. (1979) Methods of Studying Root Systems. Springer-Verlag, Berlin.

35
TSBF: A Handbook of Methods

Caldwell, M.M., and Virginia, R.A. (1991) Root systems. In: Pearcy, R.W., Ehleringer,
J., Mooney, H.A. and Rundel, P.W. (eds.), Plant Physiological Ecology: Field
Methods and Instrumentalion. Chapman and Hall, London, pp.367-398.
Floris, J. and van Noordwijk, M. (1984) Improved methods for the extraction of soil
samples for root research. Plant and Soil77, 369-372.
Mackie-Dawson, L.A. and Atkinson, D. (1991) Methodology fOf the study of fOOts in field
experiments and the interpretation of results. In: Atkinson, D. (ed.), Plant Root
Growth, an Ecological Perspective. Blackwell Scientific Publications, Oxford, pp.25-
47.
Schuurman, J.J. and Goedewaagen, M.A.J. (1971) Methods for the Examination of Root
Systems and Roots. Second edition, Pudoc, Wageningen, The Netherlands, 86pp.
Taylor, H.M., Upchurch, D,R., Brown, J.M. and Rogers, H.H. (1991) Sorne methods of
root investigations. In: Persson, H. and McMichael, B.L. (eds.), Plant Roots and
their Environment. Elsevier Science Publishers, Amsterdam, pp.553-564.
van Noordwijk, M. (1987) Methods for quantification of root distribution pattern and root
dynamics in the field. Proceedings of the 20th Colloquium of the International Potash
Institute, Baden bie Wien, pp. 243-262.
van Noordwijk, M., Widianto, Heinen, M. and Hairiah, K. (1991) Old tree root channels
in acid soils in the humid tropics: important for crop foot penetration, water
infiltration and nitrogen management. Plant and Soil 134: 37-44.
van Noordwijk, M. and de Willigen, P. (1991) Root functions in agricultural systems. In:
Pers son H. and McMichael, B.L. (eds.), Plant Roots and their Environment. Elsevier
Science Publishers, Amsterdam, pp.381-395.
van Noordwijk, M., Brouwer, G. and Harmanny, K. (1992) Concepts and methods for
studying interactions of roots and soil structure. Geoderma (in press).

3.4 LITTER DECOMPOSITION

3.4.1 Introduction

Decomposition is a complex process regulated by the interactions between organisms (fauna

and micro-organisms), physica1 environmental factors (particularly temperatureand moi sture)
and resource quality (defined here by lignin, nitrogen and condensed and soluble polyphenol
concentrations) (Swift et al., 1979). As decomposition progresses, soluble and particulate
materials from the litter, organism tissues and products of microbial metabolism are separated
from the original resource by leaching, physica1 fragmentation and animal feeding activities.
These products are then transported by wind, water and gravity to soil microhabitats which
have a different set of conditions regulating decomposition to those of the parent material.
An almost intractable problem in quantifying litter decomposition is the need to impose
methods which enable the experimental material to he identified without affecting the
variables which regulate the component processes.

3.4.2 Litter bags and decomposition rate constants

Enclosing litter in a mesh bag makes it possible to recover the residual experimental material
and defines the conditions under which the organisms operate. However, the mesh bag and

36
 Chapter 3: Field Procedures

compaction of the litter cao create different microclimates to conditions in unconfined litter
and the incorporation of materiaI into soil is aIso affected. A modified method is described
in Section 3.4.4 which overcomes sorne of these limitations (while imposing others). None
the less the standard litter bag remains one of the most convenient methods for comparative
measurements of decomposition between sites and treatments.

The minimum requirement for litter decomposition rates in TSBF studies is the time for 50%
loss of the initial mass (lso) but most studies in the literature report the decomposition rate
constant "kW from the negative exponential function:

where Wt is the amount of the initiaI mass (WcJ remaining at time "t".

This is more conveniently expressed as the regression function:

The use of the single exponential model assumes that decomposition rates are constant over
time. In fact, litter mass lasses often show a fast initial phase due to leaching of water
soluble materiaIs (which is largely an experimental artifact of drying and re-wetting the litter) ,
a slower second phase dominated by the decomposition of cell waIl constituents such as
hemicellulose and cellulose, and a slow phase regulated by lignin and microbial products.
This series of curves cao be described by fitting a two- or three-phase exponentiaI model with
a rate-constant for each of the fractions (Wieder and Lang, 1982).

3.4.3 Decomposition of uncommed Iitter and turnover rates

Measurements of mass losses from unconfined litter under natura! conditions are only possible
in the initiaI phases of decomposition before the material starts to disintegrate. The method
is most satisfactory for wood and other discrete materiaIs which cao be tagged or attached to
a tether (Lang and Knight, 1979). For branches and tree boles changes in relative density
(weight per unit volume) provides a useful measure of mass loss without sacrificing the whole
unit. Extensive decomposition of wood often occurs before branches and tree boles faIl to
the ground. This cao be determined by comparing the relative densities of standing dead and
forest floor material with live wood but the age of the material is difficult to calibrate (Swift
et al., 1976).

Litter decomposition rates on the soil surface cao be estimated from changes in litter standing
crops over intervaIs of time where (i) there is a pulsed input of materiaI (such as crop
residues at harvest), (ii) the material has decomposed before the next input, or (iii) age classes
of materiaIs cao be distinguished. In most natura! systems it is often difficult to distinguish
HUer cohorts; especiaIly where litter faIl is more or less continuous or where litter is
non-deciduous (e.g. grasses). Under these conditions determinations of litter turnover rates
provides a useful basis for comparing litter dynamics in different habitats in relation to
climate and site factors. The turnover coefficient "k" is calculated as the quotient of litter
input (1) divided by the standing crop (S):

37

liai
TSBF: Â Handbook of M~thods

k = Ils

Even for smalilitter, the same fractions must he represented in the inputs and standing crops
when ca.lculating the quotient and this is usually ooly possible for lea.f litter and small twigs
(Anderson and Swift, 1983). Sites subject to fires (Woods and Raison, 1982) and heavier
grazing present obvious diffieulties in estimating litter turnover by input/standing erop
detenninations.

It is unfortunate that, for historical reasons, the tumover-coefficient and decomposition-rate

constant, which have no formai relationship, are both designated by "k" and the use of this
term must therefore be specified.

3.4.4 LiUer bag methods

Modified method

For studies in agricultural systems, where inputs of crop residues, green man ures or dung
have largely decomposed by the start of the following cropping season, an aIternative
procedure has been developed using bags as mesh cylinders to retain the experimental
material in contact with the ground. The method is most appropriate for surface litter
treatments; sampling and processing buried litter requires modification of the protocol
depending on the materiaI and soil texture.

The advantages of the method are that it reduces microclimatic artifacts, aIlows natural access
by soil fauna, enables measurements to be made of litter incorporation into soil and sampling
cao aIso he combined with determinations of other soil properties and processes. The
disadvantages are that the method involves disturbing the soil (hence is most appropriate for
arable systems) and samples take longer to process. The method is also less suitable where
litter fall occurs during the experimental period though this cao be excluded by fine mesh
covers (e.g. mosquito-net). The method is unsuitable where there is a residual standing crop
of litter.

Procedure

1 Open the bags to form mesh tubes (ca. 20 cm diameter by 30 cm deep) and bury them
to about haIf of the depth in the soil. For stony or shallow soils the bag cao be eut
in half and only 5 cm inserted below ground. Placement is made easier by using a
metal sleeve which fits inside the bag to hold the cylinder rigid during handling. In
tilled soils, large stones may be removed from within the cylinder to facilitate
subsequent sampling.

2 Locate the mesh cylinders on a random basis in the field and add a known mass of
experimental material in parallel with plot treatments.

3 Sample 3 to 5 bags on a random basis at regular intervals. Collect as much of the

litter from the soil surface as possible (this may include smaIl stones and aggregates)
and put into a bag.

38
 Chapter 3: Field Procetblres

4 Suiface-litter treatments: soil cores (ca. 5 cm diameter) can he taken after the removal
of litter and divided into depth increments to determine moisture content, organic
carbon distribution, soi! organic matter fractions and N-mineralization rates in the
laboratory. The remaining material can he dug out and hand sorted for roots and soil
macro-fauna.
Buried-litter treatments: take cores for associated measurements and then remove all
the soil in the bag to a suitable location for processing.)

5 Fill in the hole with soU from the margin of the plot and mark the location to avoid
resampling.

6 Suiface-litter treatments: in the laboratory sort the bagged litter material by hand, or
sift in a coarse sieve, to separate the large litter fragments, stones and soU aggregates
such as termitelearthworm worked material.
Place the residua1litter and soi! material in a bucket of water, briefly swirl to suspend
litter fragments, and carefullydecant through a 2 mm mesh sieve. Experimental
material retained on the sieve is c1assified as litter. Use a 0.25 mm sieve to further
separate material in the washings if soillitter is to he determined (see Section 6.5.5).
Buried litter treatments: hand-sort the material for litter (fauna and roots) as far as
practica1. Subsample the soil and separate by water flotation as described above. In
soils with high clay content, subsamples may he deflocculated with a water softener
to release occluded material.

7 Drain washed litter thoroughly before air drying. Take subsamples for oven dry
weight determinations.

Note: Little leaching occurs if washing is brief but this cao be readily checked. It may be
impractica1 to use distilled water for this separation but the water chemistry should be
borne in mind in subsequent analyses. Litter which is heavily contaminated by clay
material may he separated by more prolonged washing or sonication but effects on
nutrient leaching must be assessed. Water softeners used in defloculation may have
high concentrations of phosphorus, so samples treated that way may not he sui table
for nutrient analyses. In sorne cases, such as where termites have infilled woody
materials with soU, it may be necessary to ash the experimental material and express
results on an ash-free basis.

Daia analysis

Plot the data for % initial mass, and logo % initial mass, vs. time (days) to inspect for trends.

If the untransformed data show linear trends, estimates of lso are made by calculating the
regression and solving the equation for t = 50%.

The convention is to fit a single negative exponential function to derive the decomposition
rate-constant "k"; tso cao then be estimated from the reciprocal of k/day.

39

Il. : 1
TSBF: A Handbook of Methods

The rate-constant "kil is the slope of the regression of logn % remaining vs time (Equation 2,
Section 3.4.2). If time is in days then the slope is -logn %/day. To obtain k1year multiply
by 365.

For example, for Shorea affinis (sampled over 294 days in a rainforest in Sri Lanka), k =
-O.0036/day; k1year = -0.0036 x 365 = 1.22/year (values for k are conventionally expressed
as positive values). The estimate of lso = 1/0.0036 = 278 days.

3.4.5 Decomposition standards

Birch (Betula spp.) 10llipop sticks are used as a decomposition standard for inter-site and
regional comparisons using natura! vegetation (where possible) and a derived site in each
studyarea.

The sticks (100 x 10 x 1 mm) are secured in bund1es of three with a wire or plastic tie at
each end of the bundle. Two sets of bundles are used to investigate the effects of litter and
soil conditions on decomposition rates.

Procedure

1 Place one set horizontall y on the soil surface and bury the other set upright in the soil
with the tips at the soil surface. Pairs of bundles, one of each set, are placed and
samples on a random basis.

2 Collect sampies of 10 bundles of each set at intervals of 3 months.

3 Remove the middle stick from the bundle, dry at 85°C and weigh.

A parallel series of sticks enclosed in fme stainless steel mesh (0.5 mm) is also employed at
TSBF sites to assess termite effects on mass loss.

References

Anderson, J.M. and Swift, M.J. (1983) Decomposition in tropical forests. In; Sutton, S.L.,
Whitmore, T.C. and Chadwick, A.C. (eds.), Tropical Rain Forest: Ecology and
Management. Blackwell Scientific Publications, Oxford, pp.287-309.
Lang, G.E. and Knight, D.H. (1979) Decay site for boles of tropical trees in Panama.
Biotropica 11, 316-317.
Swift, M.J., Hea1ey, LM., Hibberd, J.K., Sykes, J.M., Bampoe, V. and Mesbitt, M.E.
(1976) The decomposition of branch-wood in the canopy and floor of a mixed
deciduous woodland. Oecologia (Berl.) 26, 139-149.
Swift, M.J., Real, O.W. and Anderson, J.M. (1979) Decomposition in Terrestrial
Ecosystems. Blackwell Scientific Publications, Oxford.
Wieder, R.K. and Lang, G.E. (1982) A critique of the analytica1 methods used in examining
. decomposition data obtained from litter bags. Ecology 63, 1636-1642.
Woods, P.U. and Raison, R.J. (1982) An appraisal of techniques for the study of litter
decomposition in euca1ypt forests. Australian Journal of Ecology 7, 215-225.

40
 Chaprer 3: field Procetblres

3.5 SOIL

3.5.1 SoU COl evolution

The efflux of CO2 from soil ("soil respiration") theoretically represents an integrated measure
of root respiration, soil fauna respiration and the carbon mineralisation from al1 the different
carbon pools in soil and litter. Over the year, the total efflux should correspond to above-
and below-grown inputs from plants under steady-state conditions. Latitudinal gradients of
soil respiration in natural systems brœd1y follow trends in above- and below-ground litter
production (Raich and Nadelhoffer, 1989), but the additional component of CO2 production
derived from root respiration cao constitute up to 50 % of the flux (Singh and Gupta, 1977;
Coleman and Sasoon, 1980; Raich and Nadelhoffer, 1989). Precise allocation of CO2
production to plant or soil processes is therefore technically difficult and generally involves
the use of 14C tracer techniques. Measurements of CO2 evolution do, however, provide a
sensitive indication of the response of microbial activity to diurnal variations in temperature
and moisture, the effects of wetting and drying following rainfall events, and the differences
in plot treatments (mulching, tillage, irrigation, etc.).

The CO2 produced at the soil surface is collected in airtight chambers placed over the soil
(inc1uding the litter layer if present) and trapped in an absorbent. The specifications of the
chambers are not precise, and their size cao be varied as appropriate to the plot design (the
use of the modified plastic bucket is described below). It is however important for short-term
measurements, or when rates of CO2 afflux are low, to keep the enc10sed volume of air as
small as possible and determine the CO2 content of the head-space. Measurements of CO2
production from large areas over long periods of time avoid these problems but obviously
integrate across short-term variations in microbial activity which may be of interest. The
presence of the chambers cao also modify soil temperature and moisture conditions, and
hence rates of microbial processes. It is therefore essential to check on similarities of soil
microclimates inside and outside the chambers.

Procedure

1 Cut hard plastic buckets (of about '1.7 cm diameter at the base) around the
circumference about half way down its height (Figure 3.4 a). The bucket bottoms
will make a set of chambers about 20 cm high, of approximately 560 cm2 area at the
open end, while the bucket tops form a set of cylinders. When inverted (open end
down), the bucket bottoms will fit within the cylinders to make an airtight seal. Drill
a 2 cm hole in the bottom of the bucket and fit it with a rubber stopper.

2 One week prior to measurement, insert the plastic cylinders (bucket tops, cut edge
down) no more than 3 cm into the soil surface at random locations within the study
plot (Figure 3.4 b). Cut and remove allliving vegetation within the cylinders at the
soil surface taking care to minimise soil and litter disturbance.

3 One week later place an open tin cao containing soda lime (see below) in the centre
of each cylinder and immediately fit the chamber tightly into the cylinder, having first
removed the stopper. Replace the stopper once the chamber is firmly jammed in the
cylinder.

41
TSBF: A HlJ1Idbook of Methods

4 After 24 hours remove the chambers, close the tin cans and remove them, and remove
the cylinders from the soil.

Note: Never place chambers in the same location twice, or leave the cylinders in place from
one measurement period to another.

El

a
b

Figure 3.4 Using a modified plastic bucket as a collector for soil-respired CO2 •

Making and using the soda lime traps

Use about 40 g of reagent-grade soda lime (6-12 mesh) in a 5.2 cm tall, 8.1 cm diameter soil
tin with a tightly fitting lido Such tins will coyer about 9% of the surface area of the floor
enclose<! by the chamber for the bucket sile above. The 40 g of soda lime follows the
recommendation of Edwards (1982) of using a minimum amount of 0.6 g/cm2 of floor; the
minimum amount of soda lime for a chamber with an area of 560 cm2 would be 34 g.

Procedure

1 Weigh an empty tin can and lid {WJ. Add about 40 g of soda lime (or the amount
required according to chamber size), to each tin and dry at 105°C for 24 hours.
Remove the tin CM from the oyen, put the lid tightly on and allow to cool. When
cool, weigh tQ the nearest 0.01 g (Wb). (Use blank tins to account for CO2 absorption
during handling and oyen drying; these are similarly treated tin cans filled with soda
lime except that they remain closed during the measurement period.)

2 Take the tin cans to the field, uncover, place within the cylinder making sure that the
lid is under the tin can and place the chamber tightly. Leave in the field for 24 hours.

42
 Chapler 3: Field ProcedlU'es

3 After 24 hours remove the tin can from the ehamber, put the lid tightly on, and take
to the laboratory for drying at 105°C for 24 hours. Remove the tin can from the
oyen, put the lid on and allow to cool. When cool weigh to the nearest 0.01 g (W,J.
Make sure that the can is c10sed tightly.

4 Caleulate the amount of CO2 evolved from the soil from the mass gain of the soda
lime sample minus the average blank gain. (The factor 1.4 compensates for ehemica1
water lost during the drying process.)

Calculations

CO2 evolved/ehamber (g) = [(SLa - SL.,) - B] x 1.4

CO2 evolved/m2/hr (mg) = 1000 x teSLa - S'-1,) - B} x 1.4 x {10000 / (A x t)}

where
SLa (g) = weight soda lime after field placement = Wa - Wt
S~ (g) = weight soda lime before field placement = Wb - W t
B (g) = mean blank soda lime mass gain
A (em2) = ehamber area
t (hr) = exposure time

Note: Use soda lime no more than twice if CO2 uptake is less that 5% of initial weight of
soda lime. If weight gain is greater than 2 g do not reuse the soda lime.

Note: For short-term measurements a volumetrie procedure can be more sensitive than the
gravimetrie procedure described above. This uses aliquots of 0.1 M KOH exposed
in plastic eups for 1 - 2 hr followed by titration with standardised 0.1 M HCl after the
addition of saturated BaCl2 solution (2 ml per 25 ml KOH). BaCl2 is poisonous - do
not pipette by mouth. Larger volumes of absorbent or molar solutions can be used
if 0.1 M KOH becomes saturated during the exposure period. Absorbed CO2 is
caleulated on the basis that 1 ml 0.1 M HCl is equivalent to 2.2 mg CO2 ; or 1 m11.0
M HCI is equivalent to 22 mg CO2 •

References

Anderson, J.P.E. (1982) Soil Respiration. In: Page, A.L., Miller, R.H. and Keeney, D.R.
(eds.), Methods of Soil Analysis, Part 2. Chemical and Microbiological Properties.
Agronomy Monograph No. 9. (2nd edition), ASA, Madison, pp.831-871.
Coleman, D.C. and Sasoon, A. (1980) Decomposer sub-system. In: Breymayer, A.I. and
van Dyne, G.M. (eds.), Grasslands, System analysis and Management. Cambridge
University Press, Cambridge, pp.609-655.
Edwards, N.T. (1982) The use of soda-lime for measuring respiration rates in terrestrial
systems. Pedobiologia 23, 321-330.
Raieh, J. W. and Nadelhoffer, K.J. (1989) Below-ground carbon allocation in forest
ecosystems: global trends. &ology 70, 1346-1354.
Singh, J.S. and Gupta, S.R. (1977) Plant decomposition and soil respiration in terrestrial
ecosystems. Botanical Review 43, 449-475.

43
 .- r~

TSBF: .A. Handboolc of Methods

3.5.2 SoU fauna

Functio1llll classification

Soil invertebrates can he classified according to their feeding habits and distribution in the soil
profile as follows:

1 Epigeic species which live and feed on the soil surface. These invertebrates effeet
litter comminution and nutrient release but do not actively redistribute plant materials
(though the comminuted material may he more easily transported by wind or water
than the material from which it was derived). These species are mainly arthropods
(e.g. Myriapods, Isopods or fairly small and entirely pigmented earthworms).

2 Anecic species which remove litter from the soil surface through their feeding
activities (e.g. earthworms dragging litter into their burrows or termites removing
litter to mounds). Considerable quantities of soil, minerai elements and organic matter
may he redistributed through these activities accompanied by physica1 effeets on soil
structure and hydraulic characteristics.

3 Endogeic species which live in the soil and feed on organic matter and dead roots.
The two main groups are earthworms and humivorous termites which can tum over
a significant proportion of top soil and soil organic matter per year in sorne sites.

The quantification of these effeets on soil processes requires a detailed study but a simple
characterisation of the macrofauna is required ta assess their role in different systems and the
impact of management practices on population densities and community structure. See Table
3.1 for a list of generaIised taxonomic units and functional classification.

Sampling

The recommendation is for a minimum of 5 (but preferably 10) 25 cm x 25 cm by 30 cm

deep soil monoliths sampled towards the end of the rainy season and hand-sorted for
macro-invertebrates (body length < 2 mm). It must he stressed this technique is designed
for undertaking a soil fauna survey; manipulative experiments may need other sampling
strategies. Severa! references related ta sampling macrofauna are given at the end of this
section.

Procedure

Locate the sampling points 5 m apart along a transect with a random origin.

1 Remove litter from within a 25 cm quadrat and retain for sorting.

2 lsolate the monolith by cutting down with a spade a few centimetres outside the
quadrat and then digging a 20 cm wide by 30 cm deep trench around it. (This
facilitates cutting of the sample into horizontal strata and collecting animais esèaping
from the black.)

44
 Chapler 3: Field Procetbues

Table 3.1 Taxonomic units for soil fauna site characterisation and their functional
classification (ecologica1 category).

Taxonomie units Ecologieal eategory

Ants Epigeic/ Anecic

Arachnids (spiders, etc) Epigeic
Beetle adults Epigeic/Endogeic
Beetle larvae Epigeic
Blattoidea (cockroaches) Epigeic
Chilopoda (centipedes) Epigeic
Cicadidae Endogeic
Diplopoda (millipedes) Epigeic
Ea.rthworms, pigmented Epigeic/ Anecic
Ea.rthworms, un-pigmented Endogeic
Gastropods (slugs and snails) Epigeic
Gryllidae (crickets) Epigeic
lsopoda (woodlice) Epigeic
Termites Anecic/Endogeic
Other groups Various

3 Collect ail invertebrates longer than 10 cm excavated from the trench; these will
mainly be large millipedes and earthworms with very low population densities but
representing an important biomasse (Their abundance and biomass per m2 of these
groups is calculated on the basis of 0.42 m2 samples, i.e. the width of the block plus
two trench widths, squared.)

4 Divide the delimited block into three layers, 0 - 10 cm, 10 - 20 cm, and 20 - 30 cm,
which are then hand-sorted.

Note: In sites where soil fauna have low abundance and/or the above method produces
unacceptable disturbance:

Remove the 0 - 10 cm soillayer intact and then dig out the remaining sample to a 30
cm depth with a spade. Cut the sides of the sample with the spade as rapidly as
possible (in advance of digging out) to isolate the block and limit the loss of termites
from the sample (which is counted as a single unit).

~
5 Rand-sort soil and litter material in trays about 50 cm x 30 cm x 5 cm deep: deposit
a handful of soil on the left side of the tray and progressively move it towards the
~
right side, over the whole. surface. After removal of the animais from the tray, empty
t it, and process another sample. This method achieves a high sorting efficiency, and
,:.
"
is faster than having large quantities of material in the tray to process at a time.

l'
\:

45
 1
j
TSBF: .A. HantlbooJ: of M~thod.s J
6 When possible, separate further separate subsamples by wet sieving to assess the
accuracy of the method.

7 Preserve the invertebrates in 4 % formaldehyde, and keep earthworms separate from

other groups.

8 Record numbers and fresh (preserved) weight of invertebrates for the litter and other
strata according to the taxonomie units listed in Table 3.1. Finer taxonomie
subdivisions are desirable if the expertise is available. Termite fungus combs should
also be recorded.

Note: Large aggregations of ants or termites in woody litter, or soil chambers exposed on
the sides of the sample, are best collected as bulk material and sorted in the
laboratory .

In sites where termite mounds or ant nests are a conspicuous feature enumerate the mounds
in a suitable area and calculate densities on a hectare basis.

Further reading

Bouche, M.B and Gardener, R. (1984) Earthworm functions VII. Population estimation
techniques. Revue d'Ecologie et de Biologie du Sol 21, 37-63.
Darlington, J.P.E.C. (1984) A method for sampling the populations of large termite nests.
Annals of App/ied Biology 104, 427-436.
Ernsting, G. (1988) A method to manipulate population densities ofarthropods in woodland
litter layers. Pedobiologia 32, 1-6.
Grace, J.K. (1990) Mark-recapture studies with Reticu/itennes flavipes (Isoptera:
Rhinotermitidae). Sociobiology 16, 297-303.
Haverty, M.I., Nutting, W.L. and Lefage, J.P. (1976) A comparison of two techniques for
determining abundance of subterranean termites in an Arizona desert grassland.
Insectes Sociaux 23, 175-178.
Kretzshmar, A. (1978) Ecological quantification of burrow systems of earthworms.
Pedobiologia 18, 31-38.
Lavelle, P. (1988) Assessing the abundance and role of soil macroinvertebrates communities 1

in tropical soils: Aims and methods. In: Ghabbour, S.I. & Davis, R.C. (OOs.),l
Proceedings of the Seminar on Resources of Soil Fauna in Egypt and Africa, Cairo, l
16-17 April 1986. African Journal of SoU Zoology 102, 275-283.
Lavelle, P. and Kohlmann, B. (1984) Étude quantitative de la macrofaune du sol dans une
forêt tropicale humide de Mexique (Bonampak, Chiapas), Pedobiologia 27, 377-393.
MacCauley, B.J. (1975) Biodegradation of leaf litter in Eucalyptus paucijlora communities
1. Techniques for comparing the effects of fungi and insects. Soi! Biology and
Biochemistry 7, 341-344.
Springett, J.A. (1981) A new method for extracting earthworms from soil cores, with a
comparison of four commonly used methods for estimating earthworm populations.
Pedobiologia 21, 217-222.

46
Chapter 4 SAMPLING FOR LABORATORY ANALYSIS AND
SAMPLE PREPARATION

4.1 SOn. AND PLANT SAMPLING FOR LABORATORY ANALYSIS

Sampling and sample preparation is arguably the most important stage of any study; excellent
laboratory analysis does not compensate for poorly collected, poorly prepared or
unrepresentative sarnples. Whenever possible, label sample containers prior to going to the
field. Field-fresh soil samples are often moist; these should he collected in plastic bags or
tubs rather than paper bags. Plant materials, on the other band, should he collected in paper
bags to reduce decomposition accelerated by a moist environment prior to preparation for
analysis. Upon retum to the laboratory, examine plant samples for contamination, e.g. dust
or mud; a brief rinse with water may he required.

UnIess the analysis requires a field moist sarnple, air-dry all samples as saon as possible to
hait biologica1 transfonnations.

4.1.1 Soil sampling

Augering procedure

1 Remove surface plant and litter material from a minimum of 5 randomly selected
auger sites.

2 At each site auger to a known depth (e.g. 30 cm) and put the auger contents in a clean
plastic bucket so that allS augerings can he thoroughly mixed prior to subsampling.
Take a subsample at about 500 g (one eupfull). Discard the remaining soil.

J Repeat step 2 for the next depth required for eaeh auger hole.

Undisturbed cores procedure

1 Clean a fiat soil surface (either a vertical face in a profile pit or a horizontal surface
at any known depth).

2 Firmly drive a soil coring eylinder (of-known volume and mass) fully into the face
causing minimum disturbance and compaetion to the core.

3 eut away surrounding soil so as to he able to eut neaUy across the inserted end of the
eylinder.

47
TSBF: A Handbook of Methods

4 Carefully lift the full core away, trim the ends flush with the cylinder ends, and cap
tightly.

4.1.2 Plant sampling

The part of the plant that needs to be sampled and the sampling strategy depends on (i) the
purpose to which the samples will be put (i.e. the types of analyses), and (ii) the type of plant
to be sampled. No standard method is thus applicable. The genera! concept as discussed for
soUs applies, vis. taking severa! (minimum 5) subsampless, mixing them to make a composite
sample and finally taking a subsamples for analysis. Sampling for litter and crop residues is
described in Section 3.2 above.

4.2 SOIL, PLANT AND LITrER MATERIAL PREPARATION

4.2.1 SoU material

Note: As many of the variables of interest in TSBF studies are biotic, or biologically
mediated, severa! of the analyses must be conducted on field-fresh soil samples; the
following guidelines relate to soUs that do not need to be fie1d-fresh.

Procedure

1 Air-dry the soil by spreading it out in a shallow tray in a well ventilated place
protected from rain and contamination. Altematively soUs can be dried in a forced
air oven at a maximum temperature of 60°C. Break up any clay clods. When the
soil is dusty it is dry enough.

2 Crush the soillumps gently so that the gravel and roots etc. are separate from the
minerai soU.

3 Sieve the soil thmugh a 2 mm sieve leaving the gravel and roots etc. in the sieve.

4 Pick out the roots and save if required.

S Retain the gravel for weighing if required. This should be done if it appears to be
about 5 % or more by mass of the original sample.

6 Retain a representative sample (of the sieved soil approximately 250 g), by e.g.
coning and quartering, for analysis.

Note: Coning and quartering is a simple and effective method of subsampling loose, < 2
mm material: mix, then pile the material in a fiat cone or mound. Divide into four
quarters by cutting across the midd1e in an "X" shape with a blade. Discard two
opposite segments, and mix and re-cone the other two. Repeat the process until
sufficient material remains.

48
 Chapl~r 4: Sampli1l8 for Laboralory Analysis lJ1Id Sampl~ Preparation

4.2.2 Plant material and litter

Procedure

1 Dry green material and litter at a maximum of 60°C; dung should be dried at a
maximum of 40°C to prevent volatilisation.

2 Grind all the material to pass a 0.15 mm mesh.

3 Retain a representative sample of approximately 10 g, bye.g. coning and quartering

(see Section 4.2.1), for analysis.

4.2.3 Sample storage

Both soil and plant material will satisfactorily store for long periods (Le. > 1 year) if they
have been correctly prepared. Samples are stored in clearly labelled, air-tight,·termite-proof
containers; the storage room should be well ventilated.

49

, 1
Chapter 5 LABORATORY PRACTICE AND QUALITY CONTROL

5.1 LABORATORY LAY-OUT

The main aim in designing the lay-out of the laboratory is to create a work environ ment
which is ergonomica1ly efficient. If this is achieved, many of the issues discussed below,
especially safety, throughput and reproducibility, are maximised. In general, samples should
"flow" smoothly through the lab (Le. without being repeatedly carried backwards and
forwards), moving away from the reception and preparation areas to cleaner areas for
analysis. The designated work areas (see Table 5.1) should be chosen to afford maximum
ease of operation, and a level of cleanliness required for the task in hand. As few labs are
idea1 in design and location, compromise inevitably has to be sought.

5.2 SAFETY

Many laboratory procedures involve handling hazardous substances and eye protection and
laboratory coats must be used. Supplies of other protective equipment (e.g. thick rubber
gloves for cleaning acid spills, tongs etc.) should be provided. An eye wash station should
be readily accessible near the area of maximum hazard, probably the fume cupboard, and ail
lab personnel must be familiar with its use. Regularly inspected CO2 or foam fire
extinguishers should be easily accessible.

Only those reagents and items of equipment needed for a given procedure should be in the
work area in question; back-up supplies should be kept in a store. AlI reagent bottles must
be clearly marked, including any particular hazard information, e.g. poison or strong acid.
Stores and cabinets for hazardous chemica1s should be secure and lockable. Flammable
chemica1s should be kept in separate, flameproof cabinets. Procedures involving strong acids
and/or high temperatures must be performed in an efficient fume cupboard.

Any spills of chemica1s, solutions or water should be promptly cleared up, and ail waste bins
regularlyemptied. Work areas should be kept "clutter-free", and a good supply of cleaning
equipment, cloths and tissues should be provided.

Liquid waste should he poured carefully down a sink, with plenty of water to dilute and flush
it away. A dilution tank should be installed between the lab's sinks and the main disposai
site. Soil and plant analysis labs do produce toxic and acidic waste, but of a highly diluted
form.

Soil extracts should be disposed of into a container, as disposaI down a sink leads to severe
plumbing maintenance problems. Eating, drinking and smoking must he prohibited in the lab
at all times. Ali laboratory lab personnel must wash their hands with soap and water after
lab work, and again before eating. .

50
 Chapur 5: Loboralory Pracrice and Quallly Control

Table 5.1 Recommended work areas and tasks for standard soil and plant analysis.
Location Task Material when task
completed
Field Sample collection and labelling Field condition soil and
plant
Sample Reception Sample bag sorting and ordering; Field condition soil and
label checks; recording entries in plant
sample accession book
Soil 10 Sample Air drying; crushing; 2 mm sieving; 2 mm air-dry sOil
Preparation Area subsampling
Plant 1 0 Sample Oven drying; subsampling Oven-dry plant
Preparation Area
Soil 2 0 Sample Subsampling; finegrinding Finely-ground, air-dry
Preparation Area soil
Plant 2 0 Sample Subsampling; fine grinding Finely-ground, oven-
Preparation Area dry plant
Sample Store SoU: 2 mm, air-dry SoU: 2 mm air-dry
Plant: ground, oven-dry Plant: finely ground,
Storing samples systematical1y with oven-dry
easy access and retrieval
Wet Chemistry Lab Extraction; [digestion]; filtration; Solutions
dilution; colour development; titration
Instrument Lab Determination (pH, EC, flame Instrument readings
photometry, AAS, colorimetry);
calculation
Office Calculation; report writing Report, etc.

Note: Several soil analyses (e.g. mineral nitrogen extraction) will require field condition
soils.

5.3 SERVICES

5.3.1 Water supply and quality

Soil and plant analysis labs require large and reliable supplies of water. Mains water is
generally of adequate quality; the higher the quality, the better, as saline or dirty water
rapidly shortens the lives of deionisers and water stills. If the mains suppl Y is unreliable, a
header tank should be installed, dedicated to the lab; it should not serve toilets and other
general areas in the building.

.1 51
TSBF: A Handbook of Meth0d8

Single distilled or deionised water should he used for all analytica1 work and reagent
preparation, and the final stages of washing labware. Deionised water should have a
conductivity value < 0.2 j.1S/cm.

5.3.2 Electricity supply and quaUty

Almost ail analyses require electricity. The reliability of the supply is therefore of paramount
importance. Electricity surges, lows and blackouts are an ever present threat in all labs,
although the degree of risk may vary! AlI instruments should he surge-protected at least, to
protect the circuits, and a stabilised electricity supply is essential for most. As most
instruments are of low wattage, an uninterruptible power supply (UPS - as now widely used
for computers) provides a very stable supply, and allows the completion of a set of sample
readings in the event of a power eut; this last point is especially useful, as the instrument
would have to he re-ca1ibrated upon start up. In sorne locations (Le. very humid or dusty),
the most sensitive instruments will need to he in air-conditioned rooms with controlIed
humidity.

5.3.3 VVorksurfaces
Although Formica or plastic work surfaces are desirable, many labs have wooden benches.
These are perfectly adequate providing they are kept c1ean and vamished. Disposable,
plasticised sheeting can be obtained as an added precaution against contamination. The fume
cupboard floor should he of specialised corrosive and heat resistant material, or glazed tiles.

5.4 GLASS AND PLASTICWARE WASHING

AlI glassware used in procedures must he scrupulously c1ean. Writing should he wiped off
with a little acetone on a tissue; never use SCQurinl: powder, as it contains large amounts of
phosphorus and chlorine, and scratches the article heing "cleaned". Brushing in mains water,
followed by at least two distilled or deionised water rinses will suffice in most cases.
Detergent (phosphorus-free) need only he used If oily or greasy materials have been involved.
More powerful c1eaning agents such as chromic acid may occasionally he required.

Dirty plasticware is more difficult to clean because the internaI and externaI surfaces are
easily scratched. Writing may be removed by a mild solvent such as aqueous alcohol, as
sorne plasticware is attacked by acetone. Wiping with a moist tissue, and soaking in
phosphorus-free detergent solution or dilute acid, prior to thorough rinsing, will help to
remove residues. When they become permanently marked or scratched the items should he
discarded.

s.s CONTAMINATION

Sources of contamination must he identified and eliminated. Sorne of the more common
examples inc1ude:

52
 Chopf~r 5: Laboralory PrQCtic~ and QJUJlity COIIIrol

• externaI dust blown in from the surroundlng environment - try to minimise by c10sing
windows in windy conditions;
• internaI dust - this arises from c1eaning operations, rusty fittings, plaster and
decorating materials;
• inter-sample - only have one sample container open at once;
• reagents - store volatile reagents (especia1ly ammonia) weB away from samples;
• washing materiaIs, particularly soap and scouring powders, and cosmetics.

As far as possible, keep a set of glassware exc1usively for each type of anaIysis; for example,
sets of bottles and test-tubes marked "P" should be used only for extractions and
determinations of phosphorus.

5.6 LEVELS OF AC CURAC Y

A major saving in time cao be achieved by deciding on the level of accuracy needed for a
given operation. For instance, when making standards, maximum accuracy is needed in
weighing out the reference salt, e.g. to make a 1000 ppm ~ solution, you must weigh out
exactly 4.714 g of anaIytica1 grade ammonium sulphate, and make up to 1000 ml in
volumetrie glassware. On the other hand, when weighing out soil for a digest, maximum
accuracy is needed to the required number of decimal places, aIthough the actual amount is
not critica1, e.g: 1.089 g would satisfy the requirement of "about 1 g accurately", providing
that the exact weight of sample is used in calculating the results.

The level of accuracy required for a given measurement cao be inferred from the way it is
stated in the instructions. Materials should be measured to the number of significaot figures
quoted; i.e. if a liquid addition is given as 4.4 ml, it should be greater than 4.35 ml and up
to or inc1uding 4.44 ml.

Sorne reagents, generally those needed in excess, cao be made to a low level of accuracy, and
are often stated as "%" solutions. A 5 % potassium dichromate solution need not be made
volumetrica11y; 50 ± 5 g of GPR grade reagent dissolved (and mixed with a magnetic stirrer)
in 1 litre from a measuring cylinder is sufficient.

5.7 SAMPLE BATCHES

When large numbers of samples have to be anaIysed, it is best to group them in batches. The
batch size will usually be set by sorne aspect of equipment capacity - it may he the number
of samples that cao he simultaneously digeste<!; or the number that cao be completed, or
brought to an appropriate pause in the procedure, within a working day.

In principle, the bigger the batch sire, the better 1 providing the time to work from 1 to n is
not so long as to detrimentally modify sorne aspect of 1 (e.g. colour fading); and providing
that handIing the large batch is convenient and the delay imposed at each stage of the
procedure does not affect the results.

53

Il •
TSlJF: A H~ of Mt!dwds

The greater the batch sim, however, the grearer the advantages of colorimetrie over
titrimetrie determinations, as once calibrated, the colorimeter cao be used more quickly (and
often more accurately) than titrating each unknown.

Batches must include blanks, repeats and a reference sample (see Quality Control below).
If a digestion system cao accommodate 48 tubes, the batch would comprise new 40 samples,
4 repeats, 2 blanks, 1 reference and 1 aliquot of digest mixture for standard compensation.

5.8 DATA RECORDING

Data must be recorded in dedicated lab record books, and should include a11 data pertaining
to a given analysis (e.g. date, name of analyst, sample identity, weight, titre or absorbance
reading, etc.). The Itraw datait books should never leave the labo Data must be recorded
directly into the book, and not on bits of paper, backs of hands, cigarette packets, etc!
Calibration curve data should be recorded with the sample readings to which they pertain.
Data should be recorded to the appropriate numher of signiticaot tigures.

5.9 GENERAL POINTS

After calibrating the instrument with standards, sample readings must he done without further
altering the instrument settings. If the power fails the instrument must he re-calibrated (see
above under Electrica1 supplies).

Standards must contain equivaIent levels of background oomponents (e.g. acids or catalysts)
as samples.

Standards and samples must be analysed as one batch with the same set of reagents.

Low level standards cao he unstable and should he freshly prepared each day.

Reagents which are prone to oxidation must he made on the day of use - aim to make
sufficient, but not excessive, quantities.

Sample solutions outside the analytical range quoted should be diluted bearing in mind they
must he restored to the original reagent concentrations - never dilule the developed colour.

Dispensers and diluters must he thoroughly washed through with distilled or de-ionized water
immediately after use, or the plungers may seize. To avoid contamination from metal parts
or plastic seals, dispensers and diluters should he operated about 5 times with reagent,
discarding the dispensings, before use.

~.10 QUALITY CONTROL AND STANDARDISATION PROCEDURES

Quality control is a very important aspect of lab practice. This is especially 80 in routine
anaIysis labs where graduaI drift can occur as a result of many factors, e.g. contamination,
or changed reagents and environ ment, operator error etc. The aim is to aehieve maximum

54
 Chapler 5: Laboralory Practice and QNality COnlrol

reproducibility while maintaining adequate accuiacy. A sample re-analysed after an interval

should give the same result, within the accepted variation, as when initially analysed (except
where for instance field condition soils are analysed).

Errors can be either consistent or random, depending on the causative factor. If observed,
consistent errors are generally easier to rectify than random. There are several ways of
monitoring quality.

5.10.1 Blanks

Blanks are reaction vessels that are subjected to identica1 procedures as the samples in a given
batch, but have no sample added. They allow for corrections for any background levels
introduced from reagents, filter papers, etc. Provided the blank values are the same, the
mean value can he subtracted from the sample result. In most procedures at least two blank
determinations should he inc1uded in a given batch. A very high blank value suggests
contamination in the reagents, filter papers, etc. If observed, check, then repeat batch
analysis.

5.10.2 Repeats within a batch

About 1 in 10 samples, selected at random from the batch, should he analysed in duplicate.
(The choice of 1 in 10 is a compromise hetween the idea1 of analysing all samples in
duplicate, and time, effort and expense.) Obviously the answers for given pairs of duplicates
should he the same (i.e. within ± 2 - 5 % of the mean depending on the analysis in question),
and any discrepancy must he investigated. Once rectified, the batch should he re-analysed.

5.10.3 Internai references

Internal reference materials are required for each type of material under investigation.
Prepare a sufficient quantity of homogenous material for use in the initial testing of methods,
training of technicians and the assessment of bias in subsequent analyses. Before commencing
routine analyses obtain estimates of precision for use in quality control.

A sample taken from the well homogenised, bulk quantity should be included in each batch
as an internal reference. If there was no analytica1 variation between batches, the values
obtained for this sample would he the same in each batch. The variation from the mean value
for this sample, calculated over previous batches, indicates the "batch error".

Plot each value of the internal reference on a quality control chart (e.g. Figure 5.1) to
monitor the performance of the analysis. The "y" axis is the variable value, and the "x" axis
is successive analyses of the internal reference.

Take action should a value exceed the ± 3 standard deviation (sd) limits, or if two successive
values exceed the ± 2 sd limits. As more data are accumulated reassess the limits.

55
 'lSBF: A Handbook of Mdhod&

Analysls result 1 * +3 sd
for the reference
sample Action IImlt
x
----------------------------------------- +2 sd
x Warnlng limit
x x x x
--------x-------------------------------- MEAN
x

-2 sd
x x Warning limlt
* -3 sd
Action limit

3
Successive batches

x = satlsfactory * = action required

Figure 5.1 Quality Control Chan

5.10.4 Inter-Iaboratory standardisation

To compare data sets validly, all dam should be obtainoo by using identica1 methods. With
soil and plant analysis, the important aspect to standardise is the preparation of an analate
(e.g. the extract or digest) rather than the quantitative determination. The former is where
strict observance of the method protocol must be observed, whereas for the latter, any method
that makes the quantitative determination accurately is satisfactory.

Standardisation between collaborating labs cao be checked and improved by swapping

reference materials. and then comparing results. Such materials are referred to as "extcmal
referenccs Il •

The International Soil-Analytical Exchange programme offers an world-wide external

reference service. For more information about this, contact ISE, cIo Dr V Houba, Dept of
Soil Science and Plant Nutrition, Wageningen Agricultural University, PO Box 8005, 6700
EC Wageningen, Netherlands, tel +31 837082344, tb 45015 intas nI, fax + 31 837083766.

It is, however, strongly recommended that extemal references are also exchanged between
"local" labs wherever possible.

56
 Chapter 6 CHEMICAL ANALYSES

Note: Where "water" is specified in a procedure, it refers to either distilled or deionised

water; where "tap" water is adequate, it is explicitly stated.

Note: Solution concentrations for colorimetrie procedures are derived graphica11y in the
following sections; alternative methods using regression are nevertheless equa1ly
appropriate.

6.1 WATER CONTENT DETERMINATION FOR DATA CORRECTION

Procedure

1 Weigh about 1 g of material to 0.001 g accuraey into a dry container ofknown weight
(W1, also to 0.001 g accuraey). Record the total weight of the material plus container
(W2).

Note: For sorne materials, e.g. field fresh soils or plant material, a mueh larger sample size
should be taken. Weighings should be made to ± 0.1 % accuraey.
}
j 2 Dry at 105°C for 2 hr, or until the weight stabilizes for larger sample sizes.

3 Allow to cool in a desiccator and reweigh the container plus dry material (W3).

Calculation

Water content (%) = (W2 - W3)/(W3 - Wl) x 100

Dry mass (g) = {100/(100 + % water)} x (W2 - Wl)

6.2 pH AND ELECTROCONDUCTIVITY SCREENING

Procedure

1 Add 50 ml water to 20 ± 0.1 g soil.

2 Stir the mixture for 10 min, leave to stand for 30 min, stir again for 2 min.

3 Measure the pH of the supernatant liquid.

For samples with pH < 6, analyse for exehangeable aeidity, pH KC1 and ECEC. For samples
with pH ~6, analyse for EC and CEC.

57
TSBF: Â Handbook of Mdhods

4 Allow to seUle for 1 hr then measure the electroconductivity (Ee) of the supematant
liquide

For samples with an EC > 1.0 mS/cm consider satumted paste extmct conductivity analysis.

6.3 SATURATED PASTE EXTRACT CONDUCTIVITY

The electroconductivity screening will have identified any soils which are potentially saline.
The electroconductivity of the saturated paste extract is ffieasured to determine the level of
salinity.

Reagents

Potassium chloride.

Standards

1 Di~solve 0.7456 g KCI in 1000 nù water: 1.412 mS/cm at 25°C.

2 Dissolve 7.456 g KCl in 1000 ml water: 12.900 mS/cm at 25°C.

Procedure

1 Weigh about 300 ± 25 g soil into a plastic container.

l Add warer to the soil with stirring until it is nearly satumted.

3 Allow the mixture to ~tand covered for severa! houts to permit the soil to imbibe the
warer, and then add more water to achieve a unifonnly saturated soil-water paste. At
this point the soil paste glislens as it reflecrs light, flows slightly when the container
is tipped, slides freely and cleanly off a spatula. and consolidates easily by tapping or
jarring the container after a trench is formed in the paste with the side of a spatula.

4 After mixing, a1low the sample to stand (preferably ovemight, but at least 4 hr), and
then recheck the criteria for saturation. Pree water should not collect on the soil
surface, nor should the paSle stiffen markedly or lose its g1i~ten. If the paste îs too
wet. add additional dry soil to the paste mixture.

S Transfer to a Buchner fllter funnel titted with Whatman No. 42 filter paper. Apply
vacuum, and collect the filtra.te. If the initial tiltrate is turbid. refilter.

Ci Measure the conductivity filtrate against that of the standards.

Reference

Rhoades. J.D. (1982) Soluble Salts. In: Page, A.L., Miller R.H. and Keeney D.R. (Bds.)
Melhods 01 Soil Analysis, Pan 2. Second Edition. American Society of Agronomy,
Inc., Madison. '

58
j
 Chapter 6: Chemical Analyses

6.4 EXCHANGEABLE CATIONS

The predominant exehangeable cations are either "basic" (K+, Ca2+ and Mg2+) or "acidie"
(AI3 + and H+). The "basic" cation exehange is not pH dependent and they are extraeted with
ammonium acetate at pH 7. The exehange of the "acidie" cations is pH dependent and they
are extraeted with an un-buffered potassium ehloride solution.

Note: SI units are emol( + )/kg. This is numerica11y equal to me/l00 g.

6.4.1 Exchangeable bases

Reagents

Ammonium acetate, 1 M, pH 7.0: dissolve 77.1 g ammonium acetate in 950 ml water.

Adjust the pH to 7.0 with acetie acid or aqueous ammonia. Make up to 1000 ml with
water. Mix weil.
Calcium carbonate
Hydrochlorie acid, 2 M
Hydrochlorie acid, cone.
Lanthanum ehloride
Magnesium sulphate
Potassium ehloride
Sulphurie aeid, cone.

Standards
1 Stock standard solution (100 Jlg/ml K+): dissolve 1.9067 g dry KCI in water and
make up to 1000 ml. Dilute tenfold with water.
2 Stock standard solution (1000 Jlg/ml Ca2+): dissolve 2.4973 g dry CaC03 in about
200 ml water eontaining 5 ml cone. HCl, boil to drive off CO2 , cool and dilute to
1000 ml in a volumetrie flask with water.
3 Stock standard solution (100 ~g/ml Mg2+): dissolve 1.0136 g MgS04 .7H20 in water
containing about 1 ml conc. H2S04 • Dilute to 1000 ml with water.
4 Lanthanum chloride solution (2000 J1g/ml LaH ): dissolve 10.6939 LaCI3 .7H20 in
water with aid of 1.6 m12 M Hel and dilute to 2000 ml. (Altematively, take 40 ml
10% La solution and dilute to 2000 mL)
For working standards
5 Pipette 0, 1, 2 and 3 ml of eaeh stock standard solution into 100 ml volumetrie flasks.
6 Add 20 ml ammonium aeetate solution and 20 mllanthanum ehloride to eaeh flask and
dilute to volume with water. Mix weil .

.Procedure

1 Plug the bottom of a 60 mm diameter funnel (capacity 25 ml) with cotton wooL Add
5 ± 0.1 g or' soil. Leach with 10 successive 20 ml aliquots of 1 M ammonium
acetate over a period of not less than 2 hours, collecting the leachate in a 250 ml
volumetrie flask.
,,

~jl
59
TSBF: A Handboolc 01 MdIrotb

Note: If the determination of CEC (for soils with pH ~ 6) is required, retain the ammonium
saturated soil in the leaehing column for its CEe estimation.

2 Make the volumetrie flask up to the 250 ml mark with ammonium acetate solution and
mix well.

3 Pipette 20 ml of leachate into a 100 ml volumetrie flask, add 20 ml lanthanum

ehloride solution and make up to the mark with water. Determine the K+ content by
flame emission spectroscopy, and ea2+ and Mg 2 + by atomic absorption spectroscopy.

Calcularion

Determine graphica11y. For the 5.0 g soi1:250 mllea.chate ratio, diluted 5 times, the most
concentrated working standards prepared above correspond to 1.2 K+, 37.5 Ca2 + and 6.2
Mg2 + me/lOO g soil.

6.4.2 Exchangeable acidity: A13 + (+ H"1

The exehangeable aeidity in non-organie soils with a pH > 3 but < 5.5 will he almost
completely comprised of exehangeable aluminium. For these soils it is therefore assumed that
this extraction will also yield a value for exehangeable aluminium.

Reagems

Potassium chloride, 1 M: dissolve 74.55 g potassium ehloride in 1000 ml water

Phenolphthalein solution: dissolve 1 g phenolphthalein in 100 ml ethanol
Sodium hydroxide, 0.1 M: dissolve 4.0 g sodium hydroxide in 1000 ml water

Procedure

1 Weigh 10 ± 0.1 g soil into a 100 ml bealœr.

2 Add 25 ml 1 M potassium chloride.

3 Stir then leave for 30 min.

4 Filter through a Buchner funnel and leaeh with 5 successive 25 ml aliquots of 1 M
potassium ehloride.

5 Add 5 drops phenolphthalein solution and titrate with 0.1 M sodium hydroxide 10 the
first permanent pink endpoint. Correct for a blank of sodium hydroxide titre on 150
ml potassium chloride solution.

Calcularion

Exehangeable acidity (me/loog) = (ml NaOH sample - ml NaOH blank) x 10

60
 Chapler 6: ChemlcaJ Analyses

6.4.3 Cation exchange capacity (CEe)

Reagents

Ammonium acetate, lM, pH 7.0: dissolve 77.1 g ammonium acetate in 950 ml water.
Adjust the pH to 7.0 with acetie aeid or aqueous ammonia. Make up to 1000 ml with
water. Mix weIl.
Ethanol, 95%
Potassium ehloride, 1 M: dissolve 74.55 g potassium ehloride in 1000 ml water.

Procedure

1 Weigh 2.5 ± 0.01 g soil into a 50 ml centrifuge tube.

2 Add 33 ml 1 M potassium ehloride, stopper the tube and shake for 5 min.

3 Un-stopper the tube and centrifuge until the supematant is c1ea.r.

4 Discard the supernatant solution.

5 Repeat steps 2 - 4 four times.

6 Add 20 ml 95 % ethanol, stopper the tube and shake for 5 min.

7 Un-stopper the tube and centrifuge until the supematant is elea.r.

8 Discard the supernatant solution.

9 Repeat steps 6 - 8 two more times.

10 Add 33 ml 1 M ammonium acetate stopper the tube and shake for 5 min.

11 Un-stopper the tube and centrifuge until the supernatant is elea.r.

12 Pour off the supernatant solution into a 100 ml volumetrie flask.

13 Repeat steps 10 - 12 twice, adding the c1ea.r supernatant to the volumetrie flask.

14 Make the flask up to 100 ml with ammonium acetate and mix weIl.

15 Detennine the potassium concentration in the volumetrie tlask as in Section 6.4.1, but
do not dilute 5 times, i.e. omit step 3.

Note: If a centrifuge is not available, CEC can be determined by leaehing; use t (see 6.4.1.
above). Use the same quantities ànd reagents, but mix 5 g acid-washed sand with the
2.5 g soil to aid the ethanol elution steps. .

61
 T~'-

TSBF: Â Handbook of Melhods

Calculation

Determine graphically. For the 2.5 g soil: 100 ml ratio and no dilution, the top standard will
correspond to a CEC of 30 mel lOOg soil.

6.4.4 Effective cation exchange capacity (ECEC) and aluminium saturation

ECEC can only be use<! for soils which are not base saturated; i.e. pH < 7. ECEe gives
the cation exchange capacity of the soil near its natura! pH. (Methods such as ammonium
acetate or barium ch10ride extractions determine the CEe of the soil at "artificial" pH values,
and can seriously overestimate the ECEC of acid soils, particularly those dominated by
variable charge clays). BeEe is calculated by summing the exchangeable cations detennined
from the two extractions given in Sections 6.4.1 and 6.4.2, i.e. exchangeable bases +
exchangeable acidity. Aluminium saturation is the percentage of the BeEe occupied by
exchangeable AlH .

ECEC (mellOOg) = exch K+ + exch Ca1 + + exch Mg2+ + exch acidity

Al3 + saturation (%) -- (exch AIH- 1 BeEe) x 100

6.5 SOIL ORGANIC MA TI'ER AND ORGANIC CARBON

6.5.1 Det1nitioDS

Organic matter in soil is not a well-defined entity. It consists of a wide range of compounds
forming a biochemi~ continuum from cellular fractions of higher plant, microbial and animal
origin, through low and medium molecular weight organic substances of known structure, to
high molecular weight humus oompounds whose structure has yet to be characterised.

For the purposes of TSBF modelling and hypothesis testing the following fractions are
defined:

Soil Drganie Matter (SOM) is all the organic material in soil which has passed through a 2
mm sieve. It is measured by determining the organic carbon (OÇ) content of sieved soil.
SOM is 40-60% carbon, depending on its composition and age. SOM is often assumed to he
!i8% OC (Le. SOM = 1.724 x OÇ).

Microbial Biomass is the organic material in living bacteria, ascomycetes and fungi. It is
determined by fumigation and comprises about 5 % of the SOM. It is thought to be the most
rapidly tuming-over fraction of SOM, and is a major part of the "Active Pool" of the
CENTURY model (Appendix L).

SoU Litter consists of organic particles smaller than 2 mm, but larger than 0.25 mm, which
are only slightly transformed by decomposition. Consequently, their bulk density is relatively
low (around 1.0 g/cm3). The material which floats in water is the least altered. The material
which does not float, but is caught on a 0.25 mm sieve, is thought to consist of fragments of
resistant organic mate rials such as lignin, and may comprise most of the intermediate turnover

62
 Chapter 6: Chemical Analyses

rate pool of soil carbon (the "Slow Pool" in CENTURY). It is sometimes called "Light
Fraction", since its bulk density is low in relation to the following pool. It contributes 45-65
% of the SOM.

Heavy Fraction consists of high molecular weight organic polymers, formed by microbial
action on organic litters, in close physical association with clay and silt-sized soil particles.
This association imparts a bulk density greater than 1.6 g/cm3 • It cao be measured as the
material which sinks in a 1.6 g/cm3 fluid, but is more easily approximated as the OC content
of dispersed soil passed through a 0.25 mm sieve. This is thought to represent the slowest
turnover SOM component (the "Passive Pool" in CENTURY), and usually contributes 30-50
% of the SOM.

6.5.2 Total organic carbon in soils and soH extracts: background

The "wet" oxidation by acidified dichromate of organic carbon follows the reaction

To ensure a complete oxidation of aIl organic C in the sample, it is necessary to heat the
reaction at 150°C for 30 min; merely allowing the heat of dilution of the acid in the aqueous
dichromate solution bas been found to oxidize about 74% of organic C, averaged across many
soil types. Methods involving a partial oxidation of organic carbon, and hence a required
factor in calculation, while less suitable, are acceptable where a hea.ting block is not available;
values for factors used should be quoted with the result. (In the calculation given in Section
6.5.3 a mean oxidation factor of 0.74 is used.)

The equation above shows that, given excess dichromate, the amount of organic C in the
sample cao be determined by measuring either the amount of unreacted dichromate (given a
know initial amount) , or by the amount of chromic (C~+) produced. The former is
determined by titration (i.e. as in Section 6.5.4), and the latter by colorimetry (i.e. as in
Section 6.5.3). The titrimetric method is preferable for low concentrations of organic C,
although the colorimetric method is very reproducible for sampi es with higher levels of
organic C, and is very economical on technician time.

6.5.3 Total organic carbon in soils: colorimetrie method

Suitablefor ail soUs except with those where organic carbon <0.2%.

Reagents

Barium chloride, 0.4%: dissolve 4 g barium chloride in 1000 ml water.

Potassium dichromate, 5%: dissolve 50 g in 1000 ml water.
Sucrose
Sulphuric acid, concentrated (H2S04 , about 36 N).

Standards
1 Dry about 15 g sucrose at 105°C for 2 hr. Cool in a desiccator.

63

,1
TSBF: A Handbook of Methodl

2 Dissolve 11.886 g dry suerose in water and make up to 100 ml in a volumetrie flask.
This is a 50 mg/ml C solution.
3 Using a pipette transfer 0, 5, 10, 15, 20, 25 ml of the 50 mg/ml C stock solution into
labelled 100 ml volumetrie flasks and make up to the mark with water. Mix weIl.
These are the working standards, and contain 0, 2.5,5.0, 7.5, 10.0, 12.5 mg/ml C.
4 Pipette 2 ml of each working standard into labelled 100 ml conical flasks, and dry
at 105°C, or into labelled digestion tubes. These now contain 0, 5, 10, 15, 20, 25
mg C.

Procedure without utemal heating

1 Weigh about 1 ± 0.00 1 g ground soil « 0.15 mm) into a labelled 100 ml conical
flask. (If the soil is dark, or is suspected to be high in organie matter, use about 0.5
± 0.001 g). Record the weight of soil, W.
2 Add 10 ml5% potassium diehromate solution and allow it to completely wet the soil
or dissolve the standards.

3 CAUTION: Add 20 ml H2S04 from a fast burette and gently swirl the mixture.

4 Allow to cool, then add 50 ml 0.4% barium ehloride, swirl to mix thoroughly, and
allow to stand overnight, so as to leave a clear supernatant solution.

S Tramifer an aliquot of the supematant solution into a colorimeter cuvette, and

Menure and record each standard and wnple abwrbance at 600 nm.

Procedure with utemal hellling

1 Weigh about 1 ± 0.001 g ground soil « 0.15 mm) into a labelled 100 ml digestion
tube. (If the soil is dark, or is suspected to be high in organic matter, use about 0.5
± 0.001 g.) Record the weight of soil, W.
2 Add 2 ml water.

;,) Add 10 m15% potassium dichromate solution and allow it to completely wet the soil
or dissolve the standards.

4 CAUTION: Slowly add 5 ml H1S04 from a slow burette and gently swirl the mixture.

S Digest at 150 a C for 30 min.

6 Allow to cool, then add 50 ml 0.4% barium chloride, swirl ta mix thoroughly, and
allow ta stand ovemight, so as ID leave a clear supematant solution.

7 Transfer an aliquot of the supematant solution into a colorimeter cuvette, and

mwure and record each standard and sample absorbance at 600 nm.

64
 Chapter 6: ChemicaJ Analyses

Calculation

Plot a graph of absorbance against standard concentration. Determine solution concentrations

for eaeh unknown and the blanks. Subtract the mean blank value from the unknowns; this
gives a value for corrected concentration, K. Where W = weight of soil:

For without external hearing

% organie carbon = (K x 0.1) 1 (W x 0.74)

For with external hearing

% organie carbon = (K x 0.1) 1 W

Reference

Baker, K.F. (1976) The determination of organie carbon in soil using a probe-colorimeter.
Laborarory Practice 25, 82-83.

6.5.4 Total organic carbon in soil extracts: titration method

Suitable for determinations of microbial biomass C, soU litrer C and warer soluble organic
C, andfor digests of soUs very low in organic matter (>0.5%).

Reagents

Ferrous ammonium sulphate hexahydrate.

Potassium diehromate, 0.0667 M: dissolve 19.622 g dry potassium diehromate in 800 ml
water, and dilute to 1000 ml.
Sulphurie aeid, cone.
Acidified ferrous ammonium sulphate, 0.033 M: dissolve 12.940 g ferrous ammonium
sulphate hexahydrate in 900 ml water, add 50 ml cone. sulphuric acid, a110w to cool
and make up to 1000 ml with water; mix weIl.
Indicator solution: dissolve 1.485 g o-phenantholine monohydrate (CAUTION: this is a
poison) and 0.695 g ferrous ammonium sulphate hexahydrate in 100 ml of water.

Procedure

1 Transfer 4.00 ml of sample extraet into a digestion tube.

2 Add 1.00 ml 0.0667 M potassium dichromate.

3 CAUTION: Add 5 ml concentrated sulphuric acid, mixing all the time.

4 Prepare 2 blank tubes (i.e. with reagents but without extracts).

5 Place the sample tubes and 1 blank in a preheated block at 150°C for 30 min, remove
and allow to cool. Leave the other blank unheated.

65

1 1
 ---1

TSBF: A Handhook of Mt:thods

6 Quantitatively transfer the tube contents to labelled 100 ml conical flasks. and add 0.3
ml (3 - 4 drops; not by mou th) indicator solution.

7 Using a magnetic stirrer to ensure good mixing, titrate ail samples and blanks with
acidified ferrous ammonium sulphate solution; the endpoint is a colour change from
green/violet to red. Record the titres for each sample (mlsamp,J, the heated blank
(ml~ and the unheated blank (mlUS>.

Note: The acidified ferrous ammonium sulphate solution needs to be standardised daily
because of oxidation. This will establish its molarity, M:

1 Pipette 1 ml 0.0667 M potassium dichromate into a conical flask.

2 Add 0.3 ml (3 - 4 drops) indicator solution. (Do not use a mouth pipette.)

3 Titrate with acidified ferrous ammonium sulphate solution, and record the millilitres
used (T).

Calculations

M = 0.4/ T
Organic carbon (%) = {(A x M x 0.(03) 1 g} x (E / S) x 100

where
T = standardisation titre
M = molarity of ferrous ammonium sulphate (:;;; 0.033 M)
A = (ml HB - mlaamp10l x [(m1ua - miRV 1 mlUlll + (mlHB - mllillIllp,J
g = dry soil mass (g)
E -- extraction volume (ml) (~ 50 for above procedure)
S = digest sample volume (ml) (~ 4 for above procedure)

Reference

Nelson D.W. and Som mers L.B. (1982) Total carbon, organic carbon and organie matter.
In: Page, A.L., Miller. R.H. and Keeney, D.R. (oos.) MelhotMofSoilAnalysis. Part
2. American Society of Agronomy, Madison. pp. 539-579.

6.5.5 Soit litter sepamtion

Soil organie matter forms a continuum of particle sile and density. AlI limits set for
fractionation are arbitrary. For convenienœ it is recommended that techniques based on
flotation in water are employed, rather than higher density fluids.

Higher carbon recoveries can be achieved by dispersion, followoo by flotation in liquids of

specifie densities in the range 1.6 to 2.2 (Ford et al., 1969; Turehenek and Oades, 1979), or
by using density gradient centrifugation techniques or sonification (with or without sodiu~

66
 Chopler 6: Chemical Analyses

hexametaphosphate). Concentrations of organic C of soillitter separated in liquids of higher

specific gravity may be more than 10 times those of the respective unfractionated soils, since
they are more highly humified, but will contain increasing proportions of soil mineral
components as weIl.

For TSBF studies the soil litter is defined as that material which passes a 2 mm, but not a
0.25 mm, sieve. This may be further subdivided ioto that which floats in water (less
humified) and that which sinks (more humified). Separation of the soiltitter cao be achieved
by using a commercially available root washing apparatus, for example as marketed by
Mordue Bros., Newcastle, England or described by Smucker et al. (1982).

Procedure

1 Sieve 1 kg of soil through a 2 mm mesh sieve. Place the < 2 mm fraction in the root
washer. It is a brass container fitted internally with baffles, and with water jets set
horizontally at an angle of 5° with the container wall.

2 Allow mains pressure tap water ta disperse the soil, so that soillitter and suspended
fine organic-mineral components overflow through the opening at the apex of a brass
cone centrally located within the apparatus. (In high clay content soils overnight
soaking in 10% sodium hexametaphosphate may be required to aid dispersion prior
to putting the soil in the brass container.)

3 Collect the soillitter on the removable sieve (0.25 mm) fitted ta the bottom of the
apparatus.

Note: In the absence of specialised apparatus the soillitter cao be separated by agitating 250
g of soil « 2 mm) in 5 - 10 litres tap water and carefully decanting the organic
matter in suspension onto a 0.25 mm sieve. It is clearly difficult to standardise such
a separation method and although the use of sorne design of fluidising column is
preferable, this method gives a fairly good estimate; its comparative advantages are
cheapness, simplicity and therefore wide applicability.

4 Wash the soillitter material collected on the 0.25 mm sieve with water (distilled) and
fùter using a Buchner funnel to rem ove excess water.

5 Dry the material at 85°C to constant weight.

6 Retain a subsample (of known m~s) for possible N analysis if the soillitter is found
to constitute > 20 % of the total organic C.

7 Combust the soillitter at 550°C for 6 hr in a muffle fumace. Results are expressed
as ash-free organic matter.

Calculation

Ash-free mass (g) = dry mass - ash mass

67
If there is a large quantity of soillitter, only a subsample need be ashed. In this case,

Ash (%) = (ash mass/subsample mass) x 100 and

Ash-free mass (g) = dry mass x {(100 - ash %)I100}

Reference
Smucker, A.J.M., McBumey, S.L. and Srivastava, A.K. (1982) Quantitative separation of
roots from compacted soil profiles by the hydropneumatic elutriation system.
Agronomy Jou17Ul1 74, 500-504.

6.5.6 Microbial biomass

Biomass C denotes the amount of C in the total soil biota; it does not necessarily reflect their
"activities" .. Its measurement does not require soil fractionation to obtain the soil biota as a
separate entity (except perhaps the soil macrofauna), but rather the application of any of
severa! indirect techniques (Jenkinson and l.add, 1981) al1 of which give approximate
measures only.

The chlorofonn fumigation-incubation techru.que (Jenkinson and Powlson, 1976; Anderson

and Domsch, 1978) has been used extensively but there are problems using it for acid soils
Of soils which have recently had organic amendments. An alternative procedure in which the
soil is extracted immediatelyaiter fumigation is recommended; it is quicker and simpler, and
the results are highly correlated with those from the fumigation-incubation technique.

Reagenls

Chloroform (alcohol-free): wash commercial chlorofonn with about 5% by volume conc.

H 2SO4 by shaking in a separating funnel. Separate off the acid, and wash the
chloroform with 10 rinses of distilled water. (Store in the dark to prevent
photochemical build-up of explosive by-products.)
Potassium sulphate, 0.5 M: dissolve 87.13 g potassium sulphate in 1000 ml water.

Procedure

Note: Up to 20 samples of field-fresh soils, taken a few days after soaking rain at the
warmest time of the growing season, should be thoroughly mixed; subsamples of this
composite field-fresh soil are required for the assay.

1 Sieve the soil to remove stones, coarse roots and a1l visible litter.

2 Weigh 2 subsamples of 10 ± 0.01 g of soil into 50 ml beakers and a third subsample

of 10 ± 0.01 g sail into a 125 ml water-tight boule.

3 Extract the sample in the boille <ta sample) (see 6-8 below) whi1e the first sample is
beîng fumigated.

68
 Chopler 6: Chemical Analyses

4 Determine % water content in one of the samples in the beaker (see Section 6.1) so
as to be able to express the results on a dry-soil basis.

Fumigation

5 Place the beaker in a vacuum desiccator containing 30 ml alcohol-free chloroform in

a shaIlow dish. Close the lid and apply the vacuum until the chloroform c1early
evaporates. Close the tap on the desiccator and store in the dark for 5 days at 25°C.
After 5 days, transfer the soil to a watertight 125 ml extraction bottle.

Exrracrion

6 Add 50 ml 0.5 M K2S04 to the bottle, stopper tightly and shake for 30 min.

7 Filter the extract through a No. 42 Whatman fllter paper, and retain the flltrate for
analysis.

Analyse the extracts for dissolved organic C (as in Section 6.5.4; the external heating is not
required as aIl the C is readily oxidisable):

Microbial biomass C = (Extracted c;l - Extracted c.o) x 2.64

(Vance et al., 1988)

In addition, or altematively, microbial N can be determined by analysing for total N in the

extract after digestion (Section 6.6.1). Note that the N content in the digest is very low.

Microbial biomass N = (Extracted Ntl - Extracted Nta> x 1.46

(Brookes et al., 1985)

Microbial biomass P can be estimated using a procedure whereby inorganic P is extracted by

0.5 M sodium bicarbonate at pH 8.5. The extracted P is then determined by the ammonium
molybdate-ascorbic acid method (see Section 6.10.5). It is desirable to correct for chloroform
released P that is absorbed by soil during fumigation and extraction; an approximate
allowance is made by incorporating a known quantity of P during extraction and then
correcting for its recovery. After correcting for P extracted from unfumigated soil,

Microbial biomass P = (Extracted Ptt - Extracted PtO) x 2.5

Note: It should be stressed that these methods are very sensitive to the analytica1 conditions,
particularly with respect to the quality of the chloroform (it must he alcohol free for
carbon determinations) and the temperature and duration of fumigation. The constants
assumed in the calculations are approximate.

References

Amato; M. and Ladd, J.N. (1988) Assay for microbial biomass based on ninhydrin-reactive
nitrogen in ex tracts of fumigated soils. Soil Bi%gy and Biochemistry 20, 107-114.

69
TSBF: A HandIJoot of Merhods

Anderson, J.P.E. and Domsch, K.H. (1978) Mineralisation of bacteria and fungi in
ch10roform-fumigated soils. Soil Biology and Biochemistry 10, 215-221.
Brookes, P.C., Powlson, O.S. and Jenkinson, O.S. (1982) Measurement of microbial
biomass phosphorus in soil. Soil Biology and Biochemistry 14, 319-329.
Brookes, P.C., Landman, A., Pruden, G. and Jenkinson, O.S. (1985) Ch10roform fumigation
and the release of soil nitrogen: a rapid direct extraction method to measure microbial
biomass nitrogen in soil.. Soil Biology and Biochemistry 17, 837-842.
Heanes, D.L. (1984) Determination of total organic-C in soils by an improved chromic acid
digestion and spectophotometric procedure. Communications in Soil Science arul
Plant Analysis 15, 1191-1213.
Hughes, S. and Reynolds, B. (1991) Effects of c1earfelling on microbial biomass phosphorus
in the Oh horizon of an afforested podzol in Mid-Wales. Soit Use and Management
4, 183-188.
Jenkinson, O.S. and Powlson, O.S. (1976) The effects ofbiocidal treatments on metabolism
in soi1, V. A method for measuring soil biomass. SoU Biology and Biochemistry 8.
209-213.
Jenkin son , O.S. and Ladd,J.N. (1981) Microbial biomass in soil: measurement and
turnover. In: Paul, E.A. and Ladd, LN. (eds.) SoU Biochemistry 5, 415-471.
Dekker, New York.
Vance, B.D., Brookes, P.C. and Jenkinson, O.S. (1987) An extraction method for
measuring soil microbial biomass C. Soi! Biology and Biochemislry 19, 703-707.

6.6 NlTROGEN

6.6.1 Digestion for total nitrogen (and phosphorus)

Analysis of total nutrients requires the complete breakdown or oxidation of organic matter.
Wet oxidation is based on a Kjeldahl oxidation. Hydrogen peroxide is added as an additional
oxidising agent, selenium takes the place of the tradition al Mercury catalyst and lithium
sulphate is used to raise the boiling point. The main advantages of this method are that on1 y
the one digestion is required (for either soil or plant material) to bring nearly al1 of the
nutrients into solution. no volatilisation of metals, nitrogen or phosphorus takes place and the
method is simple and rapid.

Reagents

Hydrogen peroxide, 30%

Lithium sulphate
Selenium powder
Sulphuric acid, concentrated (H2S04 ~ 36N)
Digestion mixture: add 0.42 g selenium powder and 14 g lithium sulphate to 350 ml 30%
hydrogen peroxide and mix weIl. Slowly add·with care 420 ml eone. H 2S04 while
cooling in an ice bath. This mixture is ~table for 4 weeks if stored at 2°C.

Note: Selenium is carcinogenic, 50 this reagent must be handled in a fume cupboard.

70
 Chopler 6: Chemical Analyses

Procedure

1 Weigh about 0.2 ± 0.001 g ground soil (>0.15 mm) or plant materiaI into a
numbered digestion tube (min 75 ml size). Record the weight, W (g).

2 Add 4.4 ml digestion mixture to eaeh tube. (Note: aIso digest 40 ml of blank
digestion mixture for standard compensation; see Section 6.10.5)

3 Digest at 360°C for 2 hr. The solution should now be colourless and any remaining
solids white. If colour can still be seen, heat for a further 1 hr.

Note: Even if the soil is low in organie carbon, and a shorter digestion time may therefore
appear sufficient, boil for the full 2 hr to ensure that aIl the H 202 is OOHed off.)

4 A110w to cool.

5 CAUTION: Add about 50 ml water and mix well until no more sediment dissolves.
A110w to cool.

6 Volumetrica11y make up to 100 ml with water and mix well.

7 Allow to settle so that a c1ear solution can be taken for analysis.

8 Determine the nitrogen and phosphorus in the digests as indicated in Section 6.6.2 (or
6.6.3) and phosphorus according to Section 6.10.5. Make up the working standards
with the addition of 2.5 ml digested digestion blank.

6.6.2 Determination of nitrogen: distillation and titration

Free ammonia is liberated from solution by steam distillation in the presence of exeess alkali.
The distillate is collected in a reœiver wntaining excess borie aeid with an indicator (pH
4.5), and deterrnined by titration.

Reagems

Ammonium sulphate
Borie aeid
H ydrochlorie acid
Sodium hydroxide
Sodium thiosulphate
pH 4.5 indicator solution
Standard ammonium sulphate (100 ""g/ml ~4 +-N): dissolve 0.4714 g dry ammonium
sulphate in water and make up to lodo ml in a volumetrie flask.
Hydrochlorie aeid, M1140: prepare 0.1 M HCl and dilute to give MI140 (Le. 0.00714 M).
Standardise by distillation of 5.00 ml of the 100 J1g/ml ammonium standard. For the
titration, 1 ml of MI140 HCl will be equivaIent to 0.1 mg NH4 +-N.

71

l'
 J

l'SBF: Â HandIJook 0/ Melhods

Alkali mixture: dissolve 500 g NaOH and 25 g sodium thiosulphate in water with care. Cool
and dilute to 1000 ml.
Borie acid indicator solution: dissolve 20 g borie acid in water, add 15 ml pH 4.5 indicator
solution and dilute to 1000 ml.

Procedure
1 Set up a steam distillation apparatus. Use NH3-free water if possible.

l Pass steam through the apparatus for 30 min. Check the steam blank by collecting
50 ml distillate and titrating with Ml 140 HCI as given below. The steam blank should
require no more than 0.2 ml acid.

3 Transfer an aliquot (with a volume of A ml) of sample solution to the reaetion

ehamber and add 12 ml of alka1i mixture. For digests use 25 ml of alkali mixture.

4 Commence distillation immc<liately and collect 25 ml diltUlate in a suitable receiver

containing 5 ml of borie acid-indicator solution.

5 Titrate the distillate with Ml 140 hydrochloric acid to a grey end-point using a
mieroburette. Record the volume of hydrochlorie acid used (ca11ed the titre).

Occasionally check that the distillation recovery is satisfaetory by taking an aliquot of the
standard ammonium sulphate solution in place of the sample.

Calcula/ion

Subtmct the blank value from the sample titrations to give the oorrectcd titre. T.

As 1 ml MIl40 Hel 0.1 mg NH4 +-N then:

Total N (%) = (T x 0.1 x 0.001'') x (S/A) 1 W x 100/1

cr
= x S x 0.01) 1 (A x W)
where
Ile conversion factor from mg to g
!!!!!!

T correctcd titre (ml)

S final digest solution volume (ml)
A aliquot volume (ml)
W = sam pie weight (g)

Notes Deionised water is preferable to distilled water since the latter can give a high blank
value. Ammonium ions can be removed from water by shaking with. a strong cation
exehange resin. Mains tap warer may in faet have a lower NH4 +-N content than
distilled warer in sorne areas.

For soil extracts or aqueous solutions take an aliquot of 50 ml.

72
 Chapter 6: Chemical Analyses

The addition of the strong alkali liberates all inorganie N as NH4 +. To detennine
NH 4 +-N only, use 0.2 g MgO instead of the alkali solution. The N03 + cao then be
released in a second distillation after adding 0.4 g Devarda 's alloy. In the case of acid
digests, all the inorganie N will be in the NH4 + fonn, unless the N03- has been
protected (Bremner, 1982).

Reference

Bremner, J.M. (1982) Inorganie nitrogen. In: Page, A.L., Miller, R.H. and Keeney, D.R.
(eds.) Methods of soil analysis. Pan 2. Second Edition. American Society of
Agronomy, Madison.

6.6.3 Colorimetrie determination of ammonium

Reagents

Sodium citrate
Sodium hydroxide
Sodium hypochlorite solution, 5 % available Cl-
Sodium nitroprusside (CAUTION: poison)
Sodium salicylate
Sodium tartrate
Reagent NI: dissolve 34 g sodium salicylate, 25 g sodium citrate and 25 g sodium tartrate
together in about 750 ml water. Add 0.12 g sodium nitroprusside and when
dissolved make up to 1000 ml with water. Mix welle
Reagent N2: dissolve 30 g sodium hydroxide in about 750 ml water. Allow to cool, add 10
ml sodium hypochlorite solution and make up to 1000 ml with water. Mix weIl.

Note: Reagents NI and N2 should be made at least 24 hours before use and stored in the
dark.

Standards

1 Dry about 7 g ammonium sulphate at 105°C for 2 hr. Cool in a desiccator.

2 Dissolve 4.714 g dry ammonium sulphate in water and make up to 1000 ml in a
volumetrie flask. This is a 1()()() J.Lg/ml NH 4+-N stock solution.
3 Pipette 50 ml of the 1000 J.Lg/ml NH4 +-N solution into a 500 ml volumetrie flask and
make up to the mark with water. This is a 100 J.Lg/ml N solution.
4 Pipette 0, 5, 10, 15, 20 and 25 ml of the 100 J.Lg/ml NH4 +-N solution into labelled
100 ml volumetrie flasks. The standards must be made up in exaetly the same solution
as the final samples, exc1uding the sample. For example, when detennining the N
content of digests, eaeh_ standard must have 4.4 ml of digested digestion mixture
(6.6.1) added before making up to the mark with water and mixing weIl.

These are the working standards and contain 0,5, 10,15,20 and 2S J.Lg/ml NH4 +-N.

73
TSBF: Ji Handbool of MetJwds

Procedure

1 U5Ïng a micropipette, transfer 0.100 ml of each standard and sample into suitably
marked test tubes.

2 Add 5.00 ml of reagent NI to each test tube, mix well and leave for 15 minutes.

3 Add 5.00 ml of reagent N2 to each test tube, mix well and leave for 1 hr for full
colour d~velopment. The colour is stable for the day only.

4 Read each standard and sample absorbance at 655 nm.

Calculation

Plot a graph of absorbance against standard concentration. Determine solution concentrations

for the sample and the blanks. Subtract the mean blank value from the sample; this gives a
value for corrected concentration, C.

Nitrogen (%) = {(C x V) 1 W} x 0.0001

where
C = corrected concentration (JJ.g/ml)
V = final digest or extract volume (ml)
W = weight of sample (g)

For the digestion procedure listed in Section 6.6.1:

Nitrogen (%) = (C 1 W) x 0.01

6.6.4 Colorimetrie determinatioD of nitrate

The determination of nitrate in soil usua1ly follows an extraction in 0.5 M K'2S04' The
procedure is to shake 10.0 g of fresh soil in 20 ml of extractant for 30 minutes at 60 rpm.
Fitter (use nitrate-free paper such as Whatman 42) or centrifuge the sarnple. and dcterminc
the nitrate in the clear solution. Do not extract in 1 M KCl, since the CI- ion can cause
inference with the colorimetrie reaction. Freshly sampled soil must be used, since store<! soil
may have accumulated nitrate as a consequence of mineralisation. Correct the final value for
the soil water content.

Reagents

Sodium hydroxide, 4 M: dissolve 160 g sodium hydroxide in 1000 ml warer.

Salicylîc acîd. S%: dissolve 5 g salicylic Mid in 95 ml COne. sulphuric acid. (Make the day
before; stable for 7 days if stored in a dark, cool place),

74
 Chapur 6: ChnnicaJ Ana/yus

Standards

1 Dry about 10 g potassium nitrate at 105°C for 2 hr. Cool in a desiccator.

2 Dissolve 7.223 g dry potassium nitrate in water and make up to 1000 ml in a
volumetrie flask. This is a 1000 ~g/ml N03--N stock solution.
3 Pipette 25 ml of the 1000 ~g/ml N03--N solution into a 500 ml volumetrie flask and
make up to the mark with water. This is a 50 ~g/ml N03--N solution.
4 Pipette 0, 2, 4, 6, 8 and 10 ml of the 50 ~g/ml N03-·N solution into labelled 50 ml
volumetrie flasks. Make up to the mark with extraetant so that the standards and
samples are in identica1 solutions, and mix weIl. These are the working standards and
contain 0, 2, 4, 6, 8 and 10 ~g/ml N03--N.

Procedure

1 Mieropipette 0.5 ml of eaeh standard and sample into suitably marked test tubes.

2 Add 1.0 ml of salieylie acid solution to eaeh test tube, mix well immediately the acid
is added (use a vortex mixer, carefully) and leave for 30 minutes.

3 Add 10.0 ml of sodium hydroxide solution to eaeh test tube, mix well and leave for
1 hr for full colour development. The colour is stable for 12 hr only.

4 Read eaeh standard and sample absorbance at 410 nm.

Calculation

Plot a graph of absorbance against standard concentration.

Determine the solution concentrations for eaeh unknown and the blanks. (If the extraet is
coloured, run a blank using extraet, but use 1 ml sulphurie acid instead of the salieylie aeid
solution.) Subtraet the mean blank value from the unknowns; this gives a value for corrected
concentration, C.

N03-·N (~g/g soil)= (C x V) / W

where
C = corrected concentration (~g/ml)
V = extraet volume (ml)
W = weight of sample (g)

Reference

Cataldo, D.A., Haroon, M., Sehrader, L.E. and Youngs, V.L. (1975). Rapid colorimetrie
determination of nitrate in plant tissue by nitration of salieylie acid. Corrununications
in Soif Science and Plant Analysis 6, 71-80.

75
TSBF: Â Handbook of Melhods

6.7 NITROGEN MINERALISATION

Note: The method for determining microbial biomass nitrogen is given in Section 6.5.6
(Microbial bioIlWs).

6.7.1 Introdudion

Net nitrogen mineralisation is often equated with nitrogen availability. No single method for
estimating nitrogen availability bas gained uoiversal acceptanee, and îndeed it is unlikely that
any single method will prove applicable to aIl sites and conditions (Bremner, 1965; Keeny,
1980).

ln situ field methods employing the incubation of undisturbed cores theoretica1ly offer the best
estimate of nitrogen mineralisation. The method is however prone to problems of compaction
and water-Iogging (both leading 10 denitriflcation). especially in clayey soUs. The TSBF
Programme recommends the field incubation method (Section 6.7.2) for procçn studies
wherever possible. When field incubation is impossible. use the nerobic labomtory
incubation as the next best alternative method (see Section 6.7.4). The anaerobic
mineralisation index determined in the laboratory is the site characterisation Methode

Laboratory methods involve an incubation under either anaerobic or anaerobic conditions.

Aerobic methods are relatively simple, so many replicates can he performed. The method
does however tend to over estimate N Mineralisation rates, due to the massive disturbance
(henee aeration) of the sample. It is however useful for studying the controls on N
mineralisation between treatments. or when conditions preclude field estimations. The
anaerobic method is preferable for site characterisation. and hence for comparisons between
silcs; it yields the anaerobic N Mineralisation potential or "N mineralisation index n. The
anaerobic conditions prevent the oxidatîon of NH4 + to N03', resulting in a buîld up of the
NH4 + which is 'then analysed.

6.7.2 Nitrogen mineralisation (field incubation method)

The modification of the in situ incubation method of Raison et al. (1987) outlined here,
provides comparative estimates of N-mineralisation in sites with moderate to low annual
rainfall. TSBF does not recommend using this mcthod for estimating immobilisation,
leaching and plant uptake as describOO in the original method of Raison et al. (1987).

The original method involved sampling and analysis of paired cores to accommodate the high
spatial variability in N-mineralization. This proved very demanding in terms of sample
handling and analysis. An alternative procedure is therefore recommended which involves
random sampling in sub-Iots and bulking each set of cores in each sub-plot to provide a single
sample for analysis. This method provides a means of comparing sites and treatments.

Iron or plastic tubes (approximately 30 cm long and 50 mm internal diameter. or larger if

compaction becomes a real problem) are used to take initial soil samples and to isolate soils
during incubation. Galvanised piping should he weathered prior to use to avoid possible
effects of zinc toxicity.

76
 Chopler 6: Chemical Analyses

Procedure

1 Randomly insert 6 tubes into the soU in each sub-plot. The tubes should be driven or
pushed in 25 cm if possible (to prevent root in-growth), leaving 5 cm of tube
projecting above the soil surface.

2 Remove three of the tubes immediately, bulk the soils for each sub-plot (but see Note
below) , and determine initial mineral-N (NH4 + and N03') concentrations (timeo)
(sections 6.6.3 and 6.6.4). Coyer the remaining tubes with polyethylene (or a plastic
cup) to protect the core from leaching effects by rain.

3 Sample the remaining three cores and determine mineral-N (NH4 + and N03-)
concentrations after a period of time which will have to be determined empirica11y for
each site (1 to 2 weeks is a guideline).

Note: The cores cao be stored at 4 oC for a few days until processed, and cao he sectioned
by depth (0-10 and 10-25 cm) if a more detailed analysis is desired; this is essential
if assaying treatments which affect the surface. The soils should he processed and
analysed for mineral-N as described in Section 6.6.3 and 6.6.4. Net mineralisation
is calculated as the difference in mineral-N between the two time points (time 1 - time
0) and is reported as ~g N/g dry soilltime.

References

Bremner, J.M. (1965) Nitrogen availability indexes In: Black, C.A. (ed.), Methods ofSoil
Analys;s Pan 2. Chemical and Microbiological Properties. Americao Society of
Agronomy Inc Monograph 10. Madison, Wisconsin, pp. 1324-1345.
Keeny, D.R. (1980) Prediction of soil nitrogen availability in forest ecosystems: a literature
review. Forest Science 26, 159-171.
Matson, P.A. and Vitousek, P.M. (1981) Nitrogen mineralisation and nitrification potentials
following clearcutting in the Hoosier National Forest, Indiana. Forest Science 27,
781-791.
Raison, R.J., Connell, M.J. and Khanna, P.K. (1981) Methodology for studying fluxes of
soil mineral-N in situ. Soil Biology and Biochemistry 19, 521-530.
Smethurst, P.J. and Nambiar, E.K.S. (1989) An appraisal of the in situ core technique for
measuring nitrogen uptake by a young Pinus radiata plantation. Soil Biology and
Biochemistry 21, 939-942.
Waring, S.A. and Bremner, J.M. (1964) Ammonium production in soU under waterlogged
conditions as an index of nitrogen availability. Nature 201, 951-952.

6.7.3 Aerobic incubation

Two series of aerobic incubations are conducted, one at ambient field moisture, the other at
the best approximation of field capacity, Le. non-limiting water availability. (In practice
disturbed soils will have a lower water holding capacity than undisturbed cores.) The first
series provides an estimate of net nitrogen mineralisation at the time of sampling and reflects
the effect of moisture Status, both from seasonality and treatment related effects. The second

77
r.mF: Ji. HaniIIhMk (JJ MIIIWM

series provides an estimate of mineralisation under optimum conditions and reflects the effect
of substrate quality on mineralisation (Matson and Vitousek, 1981).

Soils used for the incubations should be composite, or bulked, samples obtained from 10-20
~i1 cores mndomly &ample<! within a specifie Mea or tnmsect. A minimum of 10 composite
samples per hectare is recommended. Sampling depth should be consistent with that of other
soil measurements; however the surface soil (0-10, 0-15 or 0-20 cm) should he incubated
separately from lower depths to enable detection of differences hetween treatments.

Reagents:

Potassium sulphate, 1 M: dissolve 174.25 g potassium sulphate in 1000 ml water.

Procedure:

1 Bulk. soil cores and take subsarnples for immediate gravimetric detennination of
moisture content (see Section 6.1). The remaining material is sorted to remove stones
and as m.uch l'Mt material as possible without wsrupting gtoU aggregate structure and
allowing the soil to dry.

2 For each composite soil sample weigh 50 g into each of two flasks or plastic bags.

3 Add water to one sample to adjust the gravimetric water content to the best
approximation of field capacity (as determined in Section 7.3), and drain e;tOcess water;
no water is added to a second sample.

4 Plug the flasks (or bags) with cotton wool (or if plastic bags are used they are tied
off), weigh and incubate at 26 ± 2°C for 14 days. The weight of the flasks should
he checked periodica1ly and corrections made for water loss. (If the temperature Îs
not within the specified range, the actual temperature mean and range should be
reponed.)

S At the beginning of the incubation (timeg) e;tOU'act a further subsample of 5 g in 1 M

potassium sulphate for 30 minuteS. A soil to solution ratio of 1; 10 is often
recommended but it may be necessary to use less extractant if nitrate and ammonium
levels arc low.

6 Filter the extract. Note: all filter papers are contaminated with some ammonium so
it is best to pre-rinse filters by passing through 50 ml of distilled water.

7 Determine mineraI NH 4 +- N and NO]--N in the filtered e;tOtract as described in Sections

6.6.3 and 6.6.4.

8 At the end of the incubation period (time l ) the two incubated soils are extracted and
analysed in the same manner as for timeo.

Net mineralisation is calculated as the difference in mineral-N between the two time periods
(time 1 - time 0). Results are expressed as ~g N/g dry weight soill14 days.

78
 Chupcer 6: ChemicaJ Analyses

6.7.4 Anaerobie N mineralisation index

This determination should he conducted on "field-fresh" soils, preferably within 1 hr of

sampling; about 100 g of stone-, root- and large litter-free sample should he retumed to the
laboratory. (If it is not possible to analyse with 1 br, transport the soils in a cold box.) The
standard 40°C/7 day incubation is followed:

ReagelUs

Potassium chloride, 1 M: dissolve 74.55 g KCI in 1000 mI water.

Potassium chloride, 2 M: dissolve 149.1 g KCI in 1000 mI water.

Procedure

1 Weigh one 100 mI airtight bottle (bottle A: WI) and weigh one 15 ml airtight
bottlelpolytop (bottle B: W3).

2 Add a 10 mI scoop of sample to bottle A and weigh (W2) and a 5 mI scoop of sample
to bottle B and weigh (W4).

3 Detennine the water correction factor on a third subsample as described in Section

6.1.

4 Add 50 ml 1 M KCI to bottle A and shake for 20 min at about 60 Hz.

5 Meanwhile add 12.5 ml water to bottle B, swirl to remove air bubbles, stopper, and
place in "an incubator (oven) set at 40°C.

6 After the 20 min extraction, determine timeo NH4 + -N in bottle A as described in

Sections 6.6.2 or 6.6.3. (An ammonia specifie electrode may he used: add 2 mIS M
NaOH to the sample after the shaking, immediately in sert the electrode and read
exactly 2 min after electrode exposure.)

7 After 7 days, remove boille B from the incubator.

8 Quantitatively transfer with a known volume of 2 M KCI to a c1ean container and

shake for 20 min at about 60 Hz.

9 After the 20 min extraction, determine timel NH4 + -N in bottle B as described in

Sections 6.6.2 or 6.6.3. (An ammonia-specific electrode may again be used as in step
6 above).

Calculation

soil mass timeo (g) = W3 - W 1

soil mass timel (g) = W4 - W2 (Express the results on a dry soil basis.)

Anaerobie N mineralisation rate (J1gN/g soil/day) = {(time. NH4 +-N) - (timeo NH 4 + -N)} /7

, 79
T.mF: A Handbook of M~rJwds

6.8 DENITRIFICATION

6.8.1 Introduction

The term denitrification mainly cavers the reduction of nitrate and nitrite to gaseous products
such as nitrous oxide and molecular nitrogen. The proccss is carricd out by a diverse group
of bacteria using nitrogen oxides as terminal electron acceptors in lieu of oxygen under
anaerobic conditions. The reductive pathway is generally accepted to be:

N03- --------> N02- --------> NO --------> N 20 --------> N2

Other pathways include the direct oxidation of ammonium to nitrous oxide during autotrophic
nitrification (Yoshida and Alexander, 1970) and by nitrate-respiring bacteria which reduce
nitrite to ammonium (Focht and Verstraete, 1977; Ingraham, 1981). The nitrate-respiring
bacteria are of interest because the disappearance of nitrate might otherwise be ascribed to
denitrification and because they effective1y reduce the loss of nitrate which accumulates in
excess of plant demande

A discussion of the methods for the quantification of nitrification can be found in Appendix
F ("Nitrogen availability") or in Payne (1991). The acetylene inhibition method represents
the most direct way of measuring denitrification in field and laboratory studies. Laboratory
studies have shown that 0.1 - 10% (v/v) acetylene effectively blocks N20, and is easily
measured against its background concentration in the atmosphere. This avoids the restrictions
of having to use mass spectrometry and uN tracer techniques to determine dinitrogen
emissions from soils.

6.8.2 Procedure

Reagenls

Calcium carbide

Procedure

1 Talee soil cores in either a plastic tube housed inside a steel tube that is driven into the
soil or in a steel tube with the core carefully transferred to a plastic tube of larger
diameter immediately after sampUng. Care must be taken to maintain the existing
core structure as completely as possible.

2 Within a few hours of collection, stopper each end of the plastic tube with a septum
and flush the tube with an equal volume of air from a large syringe to dispel
accumulated NzO.

3 Add sufficient acetylene (generated from calcium carbide plus water in a stoppered,
evacuated flask, as opPOsed to coming from a commercial cylinder) to make up a 10%
v/v (10 kPa) atmosphere. To ensure thorough acetylene distribution through the soil,

80
 Chopler 6: ChemicaJ Analyses

pump the tube atmosphere with a large syringe to altemately reduce and increase
pressure in the soil pore space. This should also be done prior to each gas sampling.

4 Incubate the cores either in the laboratory at field temperature or in the field. Take
gas samples from every core at least twice over a 12 hr incubation period, e.g. at 2
hr and 12 br following the addition of acetylene. A more conservative procedure is
to take gas sampI es at 2 hr intervals over this period. Incubation periods longer than
12 hr face potential problems with decreased nitrate and 02 concentrations; decreased
nitrate levels will tend to depress denitrification rates later in the incubation period,
while decreases in 02 concentrations will tend to accelerate rates.

5 Store samples in glass syringes or preferably in 3 ml pre-evacuated Venoject (TM)

vials (Terumo Scientific, New Jersey, USA) until analysis; to avoid sample
contamination it is best to overpressure sample vials by adding 5 ml of sample. If
stored for longer than a few hours, vials should be checked for leakage (lack of
positive pressure) before analysis.

6 Analyse N20 (and if desired CO~ by gas chromatography using a 63Ni electron
capture detector.

7 After the final sampling estimate the volume of the head +porespace in each tube and
oven-dry an aliquot of the soil to determine water content. Estimate head + pore
volume from the bulk density of the soil.

(A more accurate head + pore space volume cao be estimated by injecting a known
quantity of air into the sealed tube and measuring the pressure difference with a
pressure transducer; differences in pressure among tubes will be proportional to the
pore+air space present. A tube filled with different quantities of water cao serve to
calibrate the transducer to different head+pore space volumes.)

Calcula/ion (example)

The rate of denitrification is usually calculated on an area basis applying corrections for the
solubility of N20 in soil water using the coefficients in Table 6.1. This correction is based
on the assumption that most N20 production is initiated during incubation with acetylene and
the water phase is then in equilibrium with higher N20 concentrations than when collected
from the field.

For example, soil cores containing 20 ml water incubated at 25°C with a head+pore space
of 200 ml have a water/atmosphere quotient of 20/200 (= 0.1), and hence a correction factor
of 0.059. If the rate of N20 production was 500 ngltubell2 hr then total N20 production
would be 500 + (500 x 0.059) = 44.1 ng/tubelhr.

Express the final result in ng/g oven-dry soil/hr.

81
TSBF: A HatUlbook of Melhodl

Table 6.1 Correction factors for the solubility of NzO in soit water in c10sed incubation
flasks (Moraghan and Buresh, 1977).
..1
,1

Temperature Water volume 1 Atmosphere volume

oC 0.1 0.25 0.5 0.75 1.0 1.5

10 0.091 0.228 0.455 0.683 0.910 1.37

.)
'1

15 0.078 0.194 0.389 0.584 0.778 1.17

20 0.068 0.169 0.338 0.507 0.676 1.01

25 0.059 0.149 0.297 0.446 0.594 0.89

30 0.053 0.133 0.265 0.398 0.530 0.80

References

Focht, D,D and Verstraete, W. (1977) Biochemica1 ecology of nitrification and

denitrification. In: Alexander, M. (00.), Advances in Microhial Ecology l, 135-214.
Plenum Press, New York.
Ingraham, J.L. (1981) Microbio1ogy and genetics of denitrifiers. In: De1wiche, C.C. (00.),
Denitrification, Nitrification and Atmospheric Nitrous Oxide. John Wiley, New York,
pp,45-65.
Moraghan, J.T.· and Buresh, R. (1977) Correction for dissolved nitrous oxide in nitrogen
studies. SoU Science Society of America Proceedings 41, 1201-1202.
Payne, W.J. (1991) A review of methods for field measurements of denitrification. Forest
Ecology and Management 44. 5-14.
Yoshida, T. and Alexander, M. (1970) Nitrous oxide formation by Nitr()$()monas eur()paea
and heterotrophil;; mi~roorganisms. Soil Science Society of America Proceedings 34.
880-882.

6.9 PHOSPHORUS

Note: The method for determining microbial biomass phosphorus is given in Section 6.5.6.
1
j
1

6.9.1 Total phosphorus .~

The digestion procedure for soil and plant material is described in Section 6.6.1. For plant
~
material, 100% ret:Overy of Pis achieved, as all the material is solubilised; for soils, the % .'t:l
recovery varies. but is very high if the solid material remaining after digestion is white.

Determine P in the digests as in Section 6.9.5.

82
 Calculation

P in digests (%) = (C / W) x 0.1

where
W = weight of sample.

6.9.2 Bicarbonate exlractable phosphate

Many extraction techniques for "plant-available" phosphate have been developed; TSBF
recommends the bicarbonate extraction because of its suitability over a wide range of soil
types and pH values.

Reagenrs

Sodium bicarbonate, 0.5 M, pH 8.5: dissolve 84 g sodium bicarbonate in about 1000 ml

water. Make up to nearly 2000 ml with water, adjust the pH to 8.5 with 10% NaOH,
mix and make up to 2000 ml.

Procedure

1 Weigh 2.5 ± 0.01 g soil (W) ioto a polyethylene bottle.

2 Add 50.0 ml extracting solution.

3 Shake the bottle for 30 min then [tIter through Whatman No. 42 [tIter paper.

4 Determine the bicarbonate extracted phosphate in the filtrate as in Section 6.9.5, but
make the working standards in bicarbonate extracting solution; Le. use extracting
solution Înstead of water in steps 3 and 4.

Calculation

Bicarbonate extractable phosphate (~g/g) = (C x 20)/W

Reference

Watanabe, F.S. and Oisen, S.R. (1965) Test of an ascorbic acid method for determining
phosphorus in water and NaHCO~ extracts from soil. Soil Science Society ofAmerica
Proceedings 29, 6n-678.

6.9.3 Resin extractable phosphate

A suitable measure of labile P is resin ex tractable phosphate. This is closely related to

isotopically exchangeable P and therefore with the P pool that is in equilibrium with the soil
solution.

83
1

___ ~J ___________ - - - ---- --- ------------------- ---------------- ------- -------------~-- ----- -----------
Procedure for making min bags

1 Prepare small resin bags (about 4 cm x 3 cm) by folding polyester screening (e.g.
Estal Mono PE 4OOJJ.m) and sealing the edges of the bag in a cool flame. If you are
sewing the bags make sure you use polyester thread; other threads may rot.

2 Sieve a quantity of any gel-type strongly-basic Type 1 anion exchange resin through
a 3~0 t'm sieve.

3 Place sufficient of the > 350 l1m fraction in each bag to provide a total anion
exchange capacity of about 10 mmol. This is about 3.5 g if Dowex 1-X 8 resin is
used.

4 Heat seal or sew the open end of the bags closed, perpendicular to the c10sure at the
opposite end, sa as to form a tetrahedral shape which does not compress the resin
(see Figure 6.1).

Figure 6.1 Tetrahedmn resin bag; finished shape.

The resin in the bags is regenerated after each use and cao he used many times. Resin bags
must always be'stored in slightly acidified water between use.

Resin bag preparation

The best P extraction is obtained if the rain is in a mixed chloride-bicarbonate form;

Reagems

Sodium bicarbonate, 0.5 M: dissolve 42.0 g NaHCO, in 1000 ml water.

Procedure

1 Place the bags in a large beaker containing 100 ml 0.5 M NaHC03 per bag for 30
min, stirring occasionally. Repeat with frah 0.5 M NaHC03 • Wash the bags twice il 1
for 30 min each time in distilled water, and store them in the final· wash water until
use. They should be used within 24 br.

84
Extraction

Note: It is extremely important when trying to analyse small quantities of P that the
glassware is clean. Aeid-wash glassware in dilute Hel solution if you suspect
contamination; avoid using plastic containers if possible.

Reagems

Hydrochlorie aeid, 0.5 M.

Procedure

1 Weigh up to 4.0 ± 0.01 g (depending on P content) of sieved soil « 2 mm) into a

50 ml bottle with a tightly fitting lido

2 Add 40 ml of water and one resin bag, put on the lid and gently shake for about 14
br, i.e. ovemight.

3 Remove the resin bag, wash free of soil with distilled water and shake for 30 min
with 20 ml of 0.5 M Hel. (Allow the bubbles to escape before putting the lids baek
on and shaking.)

4 Detennine the phosphate content of the aeid by the colorimetrie procedure in Section
6.9.5 but make the working standards in 0.5 M Hel; i.e. use 0.5 M Hel instead of
water in steps 3 and 4.

Wash the bags a few times in distilled water to remove the acid and then regenerate the resin
(when required) as described above.

Calculation

Resin extractable phosphate (J1gl g) = (e x 20) 1 W

where
W = weight of sample.

Reference
Sibbesen, E. (1978) An investigation of the anion exehange resin method for soil phosphate
extraction. Plant and Soil 50, 305-321.

6.9.4 Organic phosphorus

Soil organic matter is destroyed by igniting the soil sample at 550°C in a muffle fumace.
This renders the organie-P aeid ex tractable and the difference in the aeid ex tractable P of an
ignited and an unignited sample of the same soil Ï8 a measure of the total organic-P in the
soil.

85
TSBF: A HtmdbooIc 01 MetItods

Reagems
Sulphuric acid, .1 M

Procedure

1 Weigh 1 ± 0.001 g of ground soil into a porcelain crucible, and place the crucible
in a cool muffle furnace.

2 Slowly raise the temperature of the furnace to 550°C over a period of 1 to 2 br.

3 Ash at 550°C for 1 h, allow the crucible to cool, and transfer the ignited soil to a 100
ml polypropylene centrifuge tube.

4 Weigh 1 ± 0.001 g of ground unignited soil into a separate 100 où polypropylene

centrifuge tube.

5 Add 50 ml of 1 M ~S04 to both samples, and shake the tubes ovemight.

6 Centrifuge or ftlter (Whatman No. 42, washed with lM H2S04) the samples to clear
the solution.

7 Determine P in the extracts as in Section 6.9.5, but make the working standards in 1
M H1S04; Le. use 1 M H1S04 instead of water in steps 3 and 4.

Calculation

P ignitcd (%) = (C x 0.2) 1 W

Punisnited (%) = (C x 0.2) 1 W
Organic phOSphOfUS (%) =- Pignlted - PunlgJIi~

whete
W .. wcight of sample.

References

Bowman, R.A. (1989) A sequential extraction procedure with concentrated sulphuric acid
and dilute base for soil organie phosphorus. SQi! Science Society ofAmerica Journal,
53 362-366.
Olsen, S.R. and Som mers , L.E. (1982). Phosphorus. In: Page, A.L., Miller, R.H. and
KeeneYt D.R. (OOs.). Merlwtls of SoU AMlysiS. Part 2. Second Edition. American
Society of Agronomy, Ine., Madison.

86
 Chapler 6; Chemical Analyses

6.9.S Colorimetrie determination of phosphorus

Reagents

Ammonium molybdate
Antimony sodium tartrate
Sulphurie acid, cone.
Ascorbie aeid, 1 %: dissolve 1 g ascorbie acid in 100 ml water; make a fresh solution every
day.
Molybdate reagent: dissolve 4.3 g ammonium molybdate in 400 ml water in a 1000 ml
measuririg eylinder. Dissolve 0.4 g antimony sodium tartrate in 400 ml water, then
add to the measuring eylinder. CAUTION: add carefully, with stirring, 54 ml
H 2S04 • Allow to cool and make up ta 1000 ml with water. Mix weIl. Stable for 4
weeks at 2°C.

Standards

1 Dry about 7 g KH2P04 at 105°C for 2 hr. Cool in a desiccator.

2 Dissolve 4.394 g dry KH2P04 in water and make up ta 1000 ml in a volumetrie flask.
This is a 1000 J1g1ml P stock solution.
3 Pipette 10 ml of the 1000 J1g1ml P solution inta a 500 ml volumetrie flask and make
up ta the mark with water (or with the appropriate solution as required for a given
determination). This is a 20 J1g1ml P solution.
4 Pipette 0, 5, 10, 15, 20 and 25 ml of the 20 J1g/ml P solution inta labelled 100 ml
volumetrie flasks. Make up to the mark with water (or with the appropriate solution
as required for a given determination) and mix weIl. These are the working standards
and contain 0, 1, 2, 3, 4 and 5 J1g/ml P in suitable background solution.

Procedure

1 Pipette 1 ml standard or sa.mple inta a test tube.

2 Add 4.0 ml ascorbic acid solution.

3 Add 3.0 ml molybdate reagent and mix weIl.

4 Leave for 1 hr for the colour ta develop fully.

S Read the standard and sample absorbances at 880 nm.

Calcularions

Plot a graph of absorbance against standard concentration.

Determine solution concentrations for eaeh unknown and the blanks.

C (P J1g/ml) = Psarnple - P mean blanlc:

87
1SBF: If Hantlbook of Methods

Calculations for the various P extracts are given in the appropriate sections above, where the
derived value for C is used.

6.10 POLYPHENOLICS

The following method is adapted from King and Heath (1967) and Allen et al. (1974) to
analyse for total extractable polyphenolics. Total soluble polyphenolics are analysed by the
Folin-Denis method, and inc1ude hydrolysable tannins and condensed tannins, as well as
non-tannin polyphenolics. Swain (1979) reports extraction from 30 to 95% of total
polyphenolics; this method therefore gives a very rough estimate of polyphenolics and it
inc1udes polyphenolics with varying reactivities and functions.

Reagents

Methanol, 50%
Orthophosphoric acid
Phosphomolybdic acid
Sodium carbonate, 17%: dissolve 17 g sodium carbonate in 100 ml water.
Sodium tungstate
Tannic acid
Folin-Denis reagent: add 50 g sodium tungstate, 10 g phosphomolybdic acid and 25 ml
orthophosphoric acid to 375 ml water. Reflux for 2 hr (with glass beads in the flask
to prevent superheating). Cool and dilute to 500 ml with water.

Standards

1 Tannic acid, 0.1 mgiml: dissolve 0.050 g tannic acid in 500 ml of water in a
volumetric flask, and make up to volume.
2 Using 0, 1, 2 and 4 ml of the tannic acid standard instead of the 1 ml of unknown,
follow the procedure below from step 4.

Procedure

1 Weigh about 0.75 ± 0.001 g (W) of materia! into a 50 ml beaker.

2 Add 20 ml 50% methanol. coyer with parafilm and place in a water bath at 77-80°C
for 1 hr,

3 Quantitatively filter (Whatman No 1) the extraet into a 50 ml volumetrie flask using

50% aqueous methanol to rinse, and make up with water. Mix well.

4 Pipette 1 ml of the unknown or standard into a 50 ml volumetrie flask, add 20 ml

water, 2.5 ml Folin-Denis reagent and 10 ml sodium 17% carbonate.

5 Make to the mark with water, mix weIl and stand for 20 min.

6 Read the standard and unknown absorbances at 760 nm.

88
 Chopler 6: ChemicaJ Analyses

Calculalion

Plot a graph of absorbance against standard concentration.

Determine solution concentrations for each unknown and the blanks. Subtract the mean blank
value from the unknowns; this gives a value for corrected concentration, C.

Total extractable polyphenolics (%) = (C x 5) / W

References

Allen, S.E., Max Grimshaw, A.H., Parkinson, J.A. and Quarmy, C. (1974) Chemical
Analysis of Ecological Materials. Wiley, New York.
King, H.G.C. and Heath, G.W. (1967) The chemical analysis of small samples of leaf
material and the relationship between the disappearance and composition of leaves.
Pedobiologia 7, 192-197.
Swain, T. (1979) Tannins and Ugnins. In: Rosenthal, G.A. and Janzen, D.H. (eds.),
Herbivores: ]heir InJeractions with Secondary PIani Metabolites. Academie Press,
New York.

6.11 LIGNIN AND CELLUWSE

6.11.1 Introduction

Lignin has been selected as the variable to assess the quality of the organic matter in the
litter. Lignin is defined as the residual organic fraction after chemica1 extraction which is
resistant to microbial degradation. Chemica1 methods for fractionating lignins are in general
time consuming, labour intensive and subject to interferences. Bach type of method is briefly
considered and recommendations for procedures and sample handling are given.

Klasson methods (e.g. Browning, 1967) have been widely adopted and are based on the
removal of the cellulose fraction by hydrolysis with 72 % sulphuric acid. Removal of
fats/lipids, how~ver, using alcohol and benzene or ether, soluble carbohydrates with water,
starch with dilute acid and protein with enzymes may be required prior to hydrolysis with ash
and protein correction following analysis. This method, which has usua1ly been regarded as
the standard procedure, is un sui table for large numbers of samples, but the reference bas been
inc1uded (Allen, 1974), should it be required for comparison and calibration purposes.

Other schemes (e.g. Southgate, 1967) which remove and quantify each organic fraction in
tum are also unsuitable for large scale use. An extraction method using acetyl bromide and
UV measurement (Morrison, 1972) is rapid and has been used for forage crops. This bas not
found wide applications, is subject to interferences and therefore cannot be recommended for
this study.

The chosen TSBF method is based on the acid detergent fibre method (ADF) (Van Soest and
Wine, 1968) which has bren widely adopted by agricultural analysts to provide a crude
measure of resistant organic material. The residue may be further separated into lignin,
cellulose and ash. As the method does however involve multiple steps, with concomitant

89
TSBF: Ji Ht:I1Idboot of Metltods

problems of reprodueibility, formaI TSBF collaborators are encouraged to use a common

extemallaboratory for lignin analysis.

For non-formai TSBF collaborators, and if resources are available, the full method for lignin
and cellulose analysîs on ADF should be followed. Altemative1y, the ADF fraction could be
determined for all samples, and a limited number of bulked samples examined for lignin and
cellulose.

6.11.2 Lignin and cellulose via acid detergent fibre (ADF)

Aeid detergent fibre is prepared from plant rnaterial by boiling with a sulphurie aeid solution
of cetyltrimethyl ammonium bromide (CTAB) under controlled conditions. The CTAB
dissolves nearlyall the nitroaenou5 constituents and the aeid hydrolyses the starch to leave
il residue containing Hgnin, œllu10se and a!h (Clancy and Wilson, 1966). Li~nin is removed
by oxidation with a buffered permanganate solution and then cellulose is determined by
weight 1088 upon ashing (Van Soest and Wine, 1968).

Reagents for ADF

Acetone (CAUTION: highly flammable)

Anti-foam solution, e.g. octan-2-o1
Sulphurie acidlCTAB solution: dissolve 100 g cetyltrimethyl ammonium bmmide in S titres
0.5 M sulphurie aeid. Filter if doudy.

Procedure for ADF

1 Weigh about 1 ± 0.001 g plant material into a 250 ml round bottom flask with a
ground glass joint to fit a reflux condenser (Wl).

l Add 100 ml sulphuric acid/CTAB solution and il few drops of anti-foam agent.

3 Connect a condenser and reflux. tOt 1 hr.

4 Filter hot through a vitreosil crucible (No. 1) of known weight (W2), under gentle
8uction.

5 Wash the residue with 3 x 50 ml aliquots of boiling water.

6 Wash with acetone until no more colour is removed, and suck the fibre dry.

7 Dry in an oven at 105°C for 2 hr, cool in a desiccator and re-weigh (W3).

Calculation

Ash-containing ADF (%) ;;;;; {(W3 - W2) 1 Wl} x 100

90
 Chapler 6: Chemical Analyses

Reagents for lignin and cellulose

Acetic acid, glacial

Ethanol
Ferric nitrate
Hydrochloric acid
~ethylpropan-2-01
Oxalic acid, dihydrate
Potassium permanganate
Potassium acetate
Silver nitrate
Silver sulphate
Saturated potassium permanganate: dissolve 50 g KMn04 and 0.05 g Ag 2S04 in water and
dilute to 1000 ml with wateT. Store in a dark bottle.
Lignin buffer: dissolve 6 g Fe(N03)3.9H20 and 0.15 g AgN03 in water. Add 500 ml glacial
acetic acid, 5 g potassium acetate, 400 ml methylpropan-2-01 and dilute to 1000 ml
with water. Store in a dark bottle.
Combined permanganate buffer: mix the saturated permanganate and lignin buffer solutions
in the ratio of 2: 1 by volume (stable for 1 week in the dark at 4 OC). The solution is
satisfactory for use if both purple and precipitate-free.
Demineralising solution: dissolve 100 g oxalic acid dihydrate in 1400 ml 95% ethanol. Add
100 ml conc. HCl and dilute to 2000 ml with water.
Ethanol, 80%: dilute 1690 ml 95% ethanol to 2000 ml with water.

Procedure for lignin and cellulose

1 Place the vitreosil crucible containing the dried acid detergent fibre in a shallow
enamel pan containing cold water (l cm depth). Do not wet the fibre at this stage.

2 Add 25 ml combined permanganate/buffer to the crucible.

3 Adjust the level of water (2-3 cm) in the pan to reduce the flow of solution out of the
crucibles.

4 Break up the lumps with a glass rod and draw the solution onto the sides of the
crucibles to wet all particles.

5 Allow to stand for 90 min at 20-25°C and add more combined solutions if necessary
to maintain the purple colour.

6 Filter under suction.

7 Place crucible in a clean pan and fill half full with demineralising solution.

8 Allow to stand for 15 min and then filter under suction.

9 Wash the fibre with demineralising solution until white.

91

l' 1 1
1___
TSBF: A Handbook of Methods

10 Filter and thoroughly wash with 80% ethanol. Filter under suction and repeat twice.

Il Wash twice in a similar manner with acetone.

12 Dry the crucible for 2 hr at 10Soe, cool in a desiccator and weigh (W4).

13 Ash the contents of the crucible at 550°C for 1 hr. Allow to cool in a desiccator and
weigh (W5).

Calcula/ions

Lignin (%) = {(W3 - W4) 1 W1} x 100

Cellulose (%) = {(W4 - WS) / Wl} x 100

References

Allen. S.E. (ed.) (1974) Chemical AfIlllysis 01 Ecologic111 Mll1l!rlals. Blackwell Scientific
Publications. Oxford. pp.252
Browning, B.L. (1967) M~tJwds of Wood Chemistry Vol. 2. John Wiley and Sons, New
York.
Clancey, M.J. and Wilson, R.K. (1966) Development and application of a new chemical
method for predicting the digestibility and intake of herbage samples. Proceedings
LOth lnlemaJional Grassland Congress, pp.445-453.
Morrison, LM. (1972) A semi-micro method for the determination of lignin and its use in
predicting the digestibility of forage crops. Journal of the Science of Food and
Agriculture 23, 455-465.
Southgate, D.A.T. (1967) Determination of carbohydrates in food. II. Unavailable
carbohydrates. Journal o/the Science of Food and Agriculture 20, 331-335.
Van Soest, P.J. and Wine, R.H. (1968) Determination of lignin and cellulose in
Acid-Detergent Fibre with permanganate. Journal of the Association of Official
Agricultural Chemists 51, 780-785.

92
 Chapter 7 SOIL PHYSICAL ANALYSES

7.1 MECHANICAL ANALYSIS (TEXTURE)

Reagenls

Amyl alcohol
Hydrogen peroxide, 30 % (100 volumes)
Sodium hexametaphosphate

Procedure
~
;~ 1 Weigh 50 ± 0.5 g soil into a 500 ml heat resistant (105°C) screw !id bottIe calibrated
at 250 ml.

2 Add 125 ml water and swirl the mixture to wet the soil thoroughly.

Note: Steps 3, 4 and 5 cao be omitted if the soil is very low in organic carbon, i.e. < 1%.

3 Add 20 ml 30% hydrogen peroxide and genUy swirl the bottIe. Add 1 or more (as
necessary) drops of amyl alcohol and gently swirl to minimise foaming.

4 When the reaction has subsided, heat the bottIe in a boiling water bath to complete the
reaction. Again add amy1 alcoho1 drop-wise ta contain the bubbles.

5 When the reaction has subsided, remove the bottIe from the water bath and allow ta
cool.

6 Add 2.0 g sodium hexametaphosphate and make up to the 250 ml mark with water.

7 Shake end over end for about 18 hr, Le. ovemight.

8 Transfer the contents ta a 1000 ml sedimentation cylinder, add the water washings,
and make up to the 1000 ml mark with water.

If the fluid temperature is likely ta fluctuate by ± 3 oC during the day, place the
cylinder in a water bath so as to maintain a constant temperature.

10 Make up a blank cylinder of 2.0 g sodium hexametaphosphate dissolved in water,

made up to the mark with water, and place this in the same tank. Allow the cylinders
to equilibrate for 30 min.

93
TSBF: A HandlHHJk of M~thDdI

11 Mix the sample cylinder vigorously with a plunger, and start the stop-watch the
moment the plunger is removed.

12 Record the Bouyoucos hydrometer readings at 40 sec and 5 hr, the 5 hr blank cylinder
hydrometer readin~ and the tank temperature. (If no hydrometer is available, pipette
25 ml !rom a depth of 10 cm al the S br tinte înto a prewweighed dish, evaporate to
dryness and weigh; this will give clay.)

Note: For fine resolution determination of silt and clay (e.g. for investigating faunal effects
on soil texture), use the pipette method as decribed in Gee and Bauder (1986).

Calcula/ions

40 sec (COlT) = 2(40 sec reading - 40 sec blank + T)

5 br (corr) = 2(5 br reading - 5 br blank + 1')

where
T = temperature corrections; For every 0ç ab<we 20 0Ç (d), T = 0.3 x d; for every oC
below 20°C (d) T = -0.3 X d.

% sand = 100 - 40 sec (corr)

% silt = 40 sec (corr) - ~ hr (oorr) % clay - ~ hr (oorr) where=

for= sand = 2 mm - 0.06 mm; silt = 0.06 mm - 0.002 mm; clay = <0.002 mm

TexturaI classification according to USDA should be followed (sec Figure 7.1).

Redrawn trom the standard

trlangular dlagram ~USDA, 1951)
tQ cQnv9ntiQnal 90 axes, Clay
and sUt contents have been
alloOl'lted to tM two axes as they
are BuMlvlded Into fewer partiole
slze rangea man sana.

(f)

c
seL

q ~ H ft _R " ~ " " m

Clay l':MtMt (% tly weloht)

C - clay S - sand, sandy r1

Z - SUt, sllty L - Loam, loamy
Figure 7.1 Two dimensional chart for assigning liSDA soi! texture classification (from
Booker Tropical Soil Manual, with permission).

94
 Chapeer 7: Soli PhysicaJ Analyses

Reference

Gee, G.W. and Bauder, J.W. (1986) Partic1e-size analysis. In: Klute, A. (ed) Methods of
Soil Analysis, Part 1. Second Edition. American Society of Agronomy, Inc.,
Madison.

7.2 BULK DENSITY

The idea1 soil water content for measuring bulk density is field capacity, although minor
deviations from field capacity will not result in significant errors. Determinations, however,
shou1d not be carried out when the soil is very dry.

7.2.1 Method for non-stony soils

Procedure

1 Remove 1-2 cm of surface soil from the spot where samples will be taken, and 1eve1
the spot.

2 Drive a 5 cm diameter thin-sheet metal tube of known weight (W1) and volume (V)
5 cm into the soil surface.

3 Excavate the soil from around the tube and eut the soil beneath the tube bottom.

4 Trim excess soil from the tube ends.

5 Dry at 105°C for 2 days, and weigh (W2).

Calculation

Bulle density (g/cm3) ... (W2 - Wl) / V

Reference

Blake, G.R. and Hartge, K.H. (1982) Bulk Density. In: A. Klute (Ed) Methods of Soil
Analysis, Part 1. Second Edition. American Society of Agronomy, Inc., Madison.
Lutz, J.F. (1947) Apparatus for collecting undisturbed soil samples. Soil Science 64,
399-401.
Uhland, R.E. (1950) Physica1 properties of soil modified by crops and Management. Soil
Science Society of America Proceedings 14, 361-366.

7.2.2 Methods for stony soUs

1

i 1
1 ! Procedure

( 1 Excavate an intact c10d of soil (about fist size).

\

95
'f'SBF: Â HanJbool DI M~rIwds

2 Air-dry the clod, tie a thin thre.ad around it and weigh it (WJ.

3 Dip il briefly in me1ted paraffin wax (60 0 e) ta waterproof il.

4 Weigh the coated clod (W"'> and calculate the weight of the paraffin coating
(Wp = W. - WJ.
Suspend the clad from the ba1ance arm and submerge ît completely in a beaker of
water. Record the weight (Wapw)' If it leaks air, discard il.

6 Break open the clod, take a subsample of the soH, weigh it, oven-dry il, and re-weigh.

Calculation

Correct W. to its oven-dry mus, Wdry:

W dry ;;;; W. x (weight of subsample after drying 1 weight of subsample before drying)

Bu1k density - Wdry 1 [{(W.., - W..,w)/pw} - (W,IP~]

where
Pw = density of water at temperature of determination (1.0)
Pp = density of paraffin wax (approximately 0.9)

If soil conditions are such that the above method is impractica1, an înfill method may be used;

Procedure

1 Dig a hole approximatcly 10 cm x 10 cm x 10 cm. Dry aU the soil removcd at

IO~oC for 24 hr and weigh (W).

2 Fill the hole with dry coarse sand from a kIlown volume of sand. Ensure the sand
surface i$ level with the surrounding soil surface. Record the volume of sand
remaining and hence calculate the volume used to till the hole (V).

Alternatively t press ultra-thin plastic food wcap around the sides of the hole, and fill
with water from a known volume of water. Record the volume of warer remaining
and henœ calculate the volume used to fill the hole (V),

CalCUllllio"

Bulk density (g/cm3 ) -- W1V

7.2.3 Method for sbrink-swell 50ils

ln shrink-swell soils the wet bulk density and the dry bulk density differ. It is difficult to
m~ure when the soil is either very wet or very dry. The minimum bulk density (which

,
96

1
 Chapler 7: Soli Physical Analyses

occurs when the soi! is fully wet) can be estimated by assumin that the particle density is
f
2.65 g/cm and air content when field saturated is 0.03 cm 3/cm (Shaw and Yule, 1978):

Bulk densitymu = 1 1 (Wmu + 0.4046)

where W max is the maximum gravimetric water content of the horizon in question (the field
saturated water content).

Reference

Shaw, R.J. and Yule, D.P. (1978) The Assessment of Soils for Irrigation, Emerald,
Queensland. Technical Report 13, Agricultural Chemistry Branch, Queensland Dept.
of Primary Industries, Australia.

7.3 FIELD CAPACITY

"Field capacity" is an arbitrary point on a curve. It is defined as the maximum amount of

water the soil can hold after it bas freely drained. In practice, as this defined point is hard
to ascertain, field capacity is estimated after the soil has drained for 2 - 3 days following
saturation, but before evapotranspiration bas depleted the water store.

Procedure

1 Build an earth wall around a 1 m x 1 m vegetation-free area, and fill with tap water.

2 Refill as necessary so that approximaœly 50 cm (500 litre) of water has soaked into
the soit.

3 Coyer the area with a plastic sheet to prevent evaporation and leave for 3 days.

4 Bulk 5 replicated 0-10 cm soil samples from near the centre of the area.

5 Put about 250 g of the wet soU in a container of lcnown weight (Wl). weigh (W2),
then dry at 105°C for 48 hr.

6 Weigh the dry soil again in the container rN3).

Calculation

Gravimetrie soil water content at field capacity (%) = {(W2 - W3) / (W3 - WI)} x 100

Volumetrie soil water content at field capacity (%) "'"

Gravimetrie water content at field capacity x bulk density
TSBF: Â Handhool 01 M.lhods

7.4 LOWER LIMIT OF PLANT AVAILABLE WATER

This value is sometimes known 8.8 wilting point and is oftert equated tQ the soil waœr content
at -1.5 MPa (-15 bar) water potential. A more biologically valid way of measuring it is in
the field under vegetation of the type being 5tudicd. A fully developed stand of plants is
deprived of water (usually by waiting for a dry period, if a rain-exclusion shelter is not
available) and the water content of the soil is determined at its minimum value by taking
gravimetric samples at severa! depths down 10 the bottom of the effective rooting depth or by
using a calibrated neutron probe.

In permanently wet climates this is not an acceptable practical method, and the -1.5 MPa
water content has to he measured.

7.4.1 The -1.5 MPa water content usÎDI a pressure dJamber apparatus

Procedure

1 Distribute rubher sample rings or Metal rings with cheesecloth fastened to one end
with a rubber band around a pre-soaked -1.5 MPa ceramic plate.

2 Fill each ring with soil. Do not compress or pack the soil into the ring. Prepare
triplicate samples.

3 Place the plate in a large tray and slowly add tap water until the water is abOut half
way to the top of the sample rings. Soak the samples ovemight.

4 SeaI the outflow nlhe on the œm.mic plate with a clamp. Carefully drain excess fluid
out of the tray. A syringe or siphon works weB.

s Place the plate (with samples) in the pressure chamber. Connect the outflow tube of
the œramic plate to the fitting on the inside of the cham ber . Connect another tube to
the fitting on the outside of the chamber, and place the free end of the tube in a
beaker of water. Unclamp the outflow tube so that water may flow frcely from the
ccramic plate to the beaker on the outside of the cha.mber_

6 Place a damp cloth over the samples in the chamber to maintain high humidity while
the samples are equilibraûng. Close the chamber, tighten the wing nuts, and slowly
app1y air pressure to the chamber until 1.5 MPa is reached.

7 Allow the samples to equilibrate for 2 to 4 days. The longer time is for soils with
high clay contents.

8 Before releasing air pressure. clamp the outflow tube sa that water cannot re-enter the
ceramic plate.

Release pressure slowly. Open the chamber and rernove the sarnples. Determine the
gravimetric water content as in Secûon 6.1.

98
 Chapter 7: Soil Physical Analyses

Calculation

Lower limit of plant available water (%) = {(W2 - W3) / (W3 - Wl)} x 100

where
Wl = weight of container (g)
W2 = weight of container + wet soil (g)
W3 = weight of container + oven-dry soil (g)

Reference

Klute, A. (1986) Water Retention: Laboratory Methods. In: A. Klute (ed). Methods ofSoil
Analysis, Pan 1. Second Edition. American Society of Agronomy, Inc., Madison.

7.4.2 Soil water potential by the r..:tter paper metbod

The water potential of relatively dry soil can he estimated by allowing it to equilibrate with
a piece of Whatman 42 tilter paper, and then determining the water content of the tilter
paper. There is a consistent relationship between tilter paper water content and equilibration
potential, over the range -66 to -4 500 kPa. To determine the -1.5 MPa water content of soil
using this cheap, low technology method, allow severa! soil samples of about 200 g each to
dry for varying lengths of time, and determine their water contents (6.1) and water potential
(see below). Plot the water content against the logarithm of the absolute water potential, and
interpolate the water content at -1.5 MPa.

Procedure

1 SeaI a sample of about 200 g into a air-tight container with two 70 mm diameter disks
of Whatman 42 tilter on top of the soil.

2 Leave the container in a dark, even-temperatured place (an insulated picnic box is
good) for 72 hours.

3 Open the container, and quickly weigh the upper tilter paper disk to 0.0001 g (Wl).
Dry the filter paper at lO5°C for 2 hr, and reweigh it to the same accuracy (W2).

Calcula/ion

The tilter paper water content

%WC = (WI-W2)/W2 x 100

The equilibrium water potential (WP, kPa)

WP = -exp(-0.128 x %WC + 11.424) if %WC < 56.2

WP = --exp(-O.0l9 x %WC + 5.282) if %WC > 56.2

99
TSBF: A Handbook of M~1hods

Reference
Savage, M.J., Khuvutlu, LN. and Bohne, H. (1992) Estimating water potential of porous
media using filter paper. South African Jou11Ull of Science 88, 269-274.

7.S PLANT AVAILABLE WATER CAPA CITY (PAWC)

The amount of water which a given soit horizon cao store for plant use is estimated from the
difference between the field capacity and lower limit of plant available water for that horizon.
It is expressed as an equivalent depth of water (mm), and is calculated using the values for
field capacity (Section 7.3) and the lower limit of plant available water (Section 7.4):

PAWC = {(FC - LL) /lOO} x DIOU x Z

where
FC = gravimetric field capacity (%)
LL = gravimetric lower limit of plant available water (%)
3
DIOU = bulk density of the horizon (gIcm )
Z = thickness of the horizon (mm)

The total PAWC is the sum of the PAWC of aU the horizons down to the effective rooting
depth.

Reference

Shaw, R.I. and Yule, D.F. (1978) The Â.Ssessmeni of Soils for Irrigation. Emerald.
Queensland. Technical Report 13, Agricultural Chemistry Branch, Quœnsland Dept.
of Primary Industries, Australia.

7.6 INFILTRATION

Infiltration refer~ to the vertical intake of warer ta the soH surface. The Most commonly used
method to determine infiltration is to flood an area contained within a bund, and to record
the water level over time (ponded method). Infiltration rates measured under a sprinkler or
under natural rainfall May not be the same.

The measured infiltration rate is markedly affected by cylinder diameter, the rate measured
being lower for larger diameters because of the reduced effect of lateral flow. Bener
estimates are therefore obtained if two concentric rings are used, the water lev el being
maintained in both, while measurements are made in the inner; this compensates in pan for
lateraI flow. An exœllent generaI referenœ can bc found in The Booker Tropical Soil
Manual (Landon, 1991). The procedure described in Section 7.6.1 is primarily used for sile
characterisation, notably for irrigation surveys.

A single ring of diamet.er 25 - 30 cm (Section 7.6.2) may be used to monitor treatment effects
in field experiments thereby minimizing site disturbance, and using much less water;
replication is therefore easier.

100
 Chapler 7: Soil PhysicaJ Analyses

7.6.1 Double ring method

Procedure

1 Remove surface litter from an area 1.5 m x 1.5 m.

2 Pre-wet the area by soaking for a few hours using an earth bund to contain the water.

3 Vertically drive the metal rings (inner 30 cm diameter, outer 60 cm diameter, heights
50 cm) about 15 cm into the wet soil so that the sma1ler ring is centred in the larger.
(This may be facilitated by cutting the soil with a knife and sealing the cylinder in the
soil with bentonite.) Lay sacking on the soil surface within the rings to minimize
disturbance from adding water.

4 Fill both cylinders to about 15 cm, record the time and the distance from the water
leve1 in the inner cylinder to the inner cylinder top.

5 Measure the water level at 1, 5, 10, 20, 30, 45, 60, 90 and 120 min, or more
frequently if infiltration is rapid.

6 Refill the cylinder when the level has dropped to about 10 cm, noting the water level
before and after refilling on each occasion. Try to maintain equivalent water levels
in the inner and outer rings.

7 Continue the measurements until a steady state has been reached, e.g. 1 - 4 hr.

8 Construct a table of results to show time intervals (min), cumulative time (min), intake
(cm), cumulative intake (cm) and infiltration rate (cm/hr).

9 Plot infiltration rate (cm/hr) against cumulative time (min).

The basic infiltration rate (cm/min) is that when steady state (Le. a straight line on the graph)
is attained.

Reference

Anon (1991) Soil Physics In: Landon, J.R. (ed.), Booker Tropical SoU Manual. Longman,
Harlow, UK. pp. 59-71, and pp. 213-217.

7.6.2 Single ring method

Procedure

1 Remove surface litter from a 50 cm x 50 cm area of bare soil.

2 Vertica1ly plastic or metal rings (diameter 25 - 30 cm, length 15 cm) 2 - 3 cm into the
soil. Press soil around the base to minimise water leakage, and coyer the soil surface
in the ring with sacking.

101
TSBF: A. Handbook of M~rJwds

3 Fill the cylinder to about 10 cm deep, record the time, and measure the water depth.

4 Continue as described in Section 7.6.1, from step 4.

Reference

Berryman C. and Trafford, B.D. (1974) Infiltration rates and hydraulic conductivity. Field
Drainage Experimental Unit Technical Bulletin, MAFF, London. 16pp.

7.7 POROSITY

Soil pores are classified by sire. Macro-pores (diameter > 0.1 mm) are chiefly responsible
for aeration and gravity flow; mesa-pores (diameter 30 - 100 J'm) conduct water by mpid
capillary flow; micro-pores (diameter < 30 J.Lm) are responsible for water retention and slow
capillary flow.

7.7.1 Total porosity

Total porosity (the sum of all pore size volumes) is calculated from the dry bulk density (cf.
Section 7.2) and the particle density (assumed to he 2.65 g/cm3 for most mineraI soils):

Total porosity (%) = {1 - (bu1k densiry 1 particle density)} x 100

7.7.2 Pore size distribution

Pore siu distribution is r;stimatcd on an undisturbed soil core. The water content of the soil
after all pores greater than a given size have been drained is determined volumetrically and
subtracted from the total porosity. This provides an estimate of the volume of pores above
a specified minimum pore diameter. Macro-pores and the larger meso-pores ( > 50 JJ,m) can
be drained by suction applied to the core using a tension table. To calculate the water tension
to be applied:

Tension (cm water) = 2908/minimum pore diameter (J.Lm)

e.g. all pores> 300 J'm will be drained at 2908/300 = 9.7 cm water tension. The tension
table has a range to measure water release of pores down to about 50 J.Lm.

Procedure for maldng a suitable tension table

1 Drill a hole in the centre of a plastic tray or sheet (0.5-1 m diameter) (table) and
attach a fitting for the connection of flexible PVC or polythene tubing.

2 Place a plastic sereen on the table which is about 10 cm smaller than the diameter of
the table.

102
 Chopter 7: Sotl Physlcal Analyses

3 Place blotting paper on top of the screeri. The paper should be the same size as the
table.

4 Attach tubing to the fitting on the table, and place the free end of the tubing in a flask
with an overflow.
Soli core
Plastic screen
~~~~~~~~~~~~m~~~Blotter paper
Plastic sheet or tray
Tublng

Distance : water tension

Water level -

Figure 7.2 Tension table apparatus.

Procedure

Note: Care in sampling is extremely important, because disturbances to the sail core may
greatly alter porosity measurements.

1 Sample when the soil is fairly moist (about field capacity). Use a core sampler of
known weight (W 1) with a bevelled cutting edge to minimise soil disturbance. Cores
should be about 2.5 cm high and 5 cm in diameter.

2 Obtain four to six cores. (If the top soil is very loose, sample a few cm below the
surface.)

3 Trim the excess soil from the cores and fasten muslin ta one end of each core with
a rubber band.

4 Place the cores in a pan and slowly add water until the level is just below the top of
the cores. Leave to soak overnight.

To prepare the tension table (see Figure 7.2):

1 Fill the overflow flask with water and put the PVC tube into il. Raise the flask above
the table and t'ill the tube with water.

2 The region between the blotter paper and table must be filled with water. and there
should be as few bubbles beneath the bloUer paper and in the tubing as possible. This
will take sorne practice to achieve.

103
TSBF: Â Handbool 0/ M~rJwds

3 U sing a rolling pin or your fingers, press the blotter paper against the table so that a
seal is formed when the flask is lowered beneath the table. Air should not leak. into
the area· below the blotter paper or the tubing at any time. If air does enter the
system, the measurements will be inaccurate.

4 Lower the flask so that the distance from the overflow tube to the centre of the sample
is equal to the amount of tension needed to drain pores above the desired minimum,
i.e. for 300 Ilm this distance is 9.7 cm.

5 Carefully drain the tray with the samples in it using a siphon.

6 Using care, move the samples onto the tension table (muslin side down). Coyer the
table with plastic or a damp c10th to maintain high humidity.

7 Leave the samples on the tension table for two to three days. Make sure that no air
has entered the tension table system, and that the water level in the flask is always at
the level of the overflow.

8 Remove the cores, measure their height and diameter carefully and calculate the
sample volume (V) in cm3 (1t r2 h). This measurement is very important for
caIculating porosity, and should he made using vernier or dial callipers if they are
available.

9 Weigh the samples, heing careful to not lose any water or soil while handling (W2).
Dry the samples in a 105°C oyen for 48 hr and re-weigh them (W3).

Calcularions

Volumetrie water content (%) = {(W2 - W3) / V} x 100

Macro porosity (%) = total porosity - volumetrie water content

Note! A volume for any given sire range cao be measured by equilibrating cores at tensions
corresponding to the upper and lower pore sizes of interest and using the difference
in water content at each tension to calculate the volume of pores between those pore
sizes.

References

Bouma, J. (1981) Comment on "Micro-, Meso-, and Macroporosity of Soil". SoU Science
Society of America Journal 45, 1244-1245.
Danielson, R.E. and Sutherland, P.L. (1986) Porosity. In: Klute, A. (ed.), Melhnds ofSoil
Analysis, Pan 1. Second Edition. American Society of Agronomy, Inc., Madison.
Luxmoore, R.J. (1981) Micro-, Meso-, and Macroporosity of SoiL letter to the Editor. Soil
Science Society of America Jou17Ul1 45, 671-672.

104
 Appendix A EXAMPLE OF RAPID RURAL APPRAISAL (RRA)
INTERVIEW
LelWra Bohren

A.l INTRODUCTION

According to Grandstaff et al. (1987), Rapid Rural Appraisal (RRA) is a new research
paradigm which views the world as a highly interactive set of variables which are rapidly
changing and subject to a high degree of uncertainty. The paradigm is based on the concept
of adaptive change involving feedback systems utiUzing the perceptions of differing
populations. RRA is used to:

1 explore, identify, and diagnose rural situations, problems, or issues;

2 design, implement, monitor, and evaluate programs, projects, and development;
3 help develop, extend, and transfer technology;
4 formulate poUcy and assist decision making;
5 respond to emergencies and disasters; and
6 improve, supplement, or complement other research.

A.2 PHASE 1: PREPARATORY WORK

A Selection of an interdisciplinary team according to the research topic (Grandstaff el

al., 1987). Teams can be paired in terms of their discipline as in the Sondeo
Approach (Hildebrand, 1981) and/or paired with local researchers as suggested by
Chambers (1991). Smaller teams are often preferred (Beebe, 1985).

B Team examination of relevant background information gathered from published and

unpublished reports such as archives, annual reports, census reports, maps, aerial
photographs, environ mental and social data, etc. on:

1. Natural circumstances such as rainfall, amount and reliability; seasonal

temperatures; sail characteristics, topography, 8uch as slope and depth. and
hydrology; and weeds. pest and disease incidence as a source of unœrtainty of crop
output (Grandstaff el al.• 1987; Beebe, 1985; CIMMYT, 1985; Doorman, 1991).

2. Institutional circumstances such as types of marketing and supply channe1s; types

and reliability of food distribution channels; extension and credit programs, types of
programs and numbers and types of participants; land tenure arrangements; farmer
i groups, voluntary, organized, official or unofficial, and planned actual functions; and
;. 1

1 other infrastructure considerations (Beebe. 1985: CIMMYT, 1985; Grandstaff et al.,

• • 1987), .
~.:.I
lOS

1
TSBF: Â Handbook of Methods

3. Socio-economic circumstanees such as population numbers, density and seulement

patterns; available area and production figures; marketed products, volume and trends
for outputs sold and inputs purchased; food purchased; relative volume and trends and
priees over the years and between seasons, priees, and marketing margins; and
available information (most of which will come from the in-depth interviews and
observation) on goals, beliefs, attitudes, and social obligations (Beebe, 1985;
CIMMYT, 1985; Grandstaff et al., 1987).

C Team discussion of research topie and background information aimed at developing

preliminary hypotheses (if appropriate), interview topies, methods, tools, and
techniques related to the research topic (Grandstaff et al., 1987).

D Team consultation with persons knowledgeable of the topie or geographic area if

possible (Grand staff et al., 1987).

E Team preliminary field visit to familiarize te.am with ara, to discuss the project with
local officiais, to pretest the interview checklist and identify key indicators, and to
select. if possible, local researchers (Boobe, 1985; Grandsmff it al., 1987; Chambers,
1991). Examples of checklists are in Beebe (1985), Raintree (1987), Rocheleau et al.
(1988) and IBSRAM (1988).

F Team discussion and planning of researeh sites, logisties and protocols for the
fieldwork (Grand staff et al., 1987; Raintree, 1987).

A.3 PHASE D: FIEWWORK

A Primary data collection by research team from interviews and observation. The RRA
depends on walking, s~ing, and asking.

1. Interviews are conducted using a ehecklist as a flexible guide and should be

carefully timed so as not to Interfere with work schedules. The use of semi-struetured
interviews (ehecklists) are essential for obtaining information sueh as indigenous
technica.l knowledge (lTK). Follow up interviews are conducted as needed. The
number and length of the interviews depend on the homogeneity of the population
(Beebe, 1985). The persons interviewed are:

a) Individuals representative of a cross section of the expected target population. A

cross section of farmers would include farmer leaders, women farmers, major
cropping farmers, poor farmers, rich farmers, traditional farmers, and innovative
farmers. Rhodes (1982) stresses the importance of including all farm members who
are involved in agricultural decision making, especially those who have migrated off
the farm or are working in off farm occupations.

b) Key informants cao be farmers and non-farmers. They should inc1ude those who
have experienee in the system beyond their own participation as farmers sueh as
bankers, landlords, ministry officials, merchants, middlemen, extension agents, buyers
of agricultural products, and suppliers of inputs (Becbe, 1985).

106
 Appendix A: Exampfe of Ropid Rural Appraisal (RRA) Interview

c) Groups representing the variability within the community. Chambers (1991)

suggests a chain of interviews from experts to novices in a process such as fanning.
Group interviews gather information such as natural history, local histories and
sensitive information such as land holdings. Group interviews cao be self correcting,
participants often correct each other' s information. However, in group interviews,
it is important to observe hehaviour to help distinguish hetween real and ideal
practices since the presence of others may influence answers. Sorne issues may need
to he confirmed by individual interviews. Suggested group interview techniques are
the "informal Delphi" technique or focus group techniques (Honadle, 1979; Beebe,
1985; Bryant and Bailey, 1991).

2. Observation inc1udes direct observation of what people are doing, the use of
photographs, actual participation, observing several farmers, etc. It is important to
have multiple checks on information and take good, detailed field notes. Check lists
are often used. It is important to look for patterns such as crop production, land use,
and farmer hehaviour. This cao he done by agroecological transects and field plotting
by air or foot, taking pictures, and direct observation of key indicators. Key
:1
indicators cao be soil colour as an indicator of particle size and distribution, fertility,
and drainage properties; birth weight of children as an indicator of health and
nutrition; housing as an indicator of poverty or prosperity; soap inventories as an
indicator of purchasing power; the appearance of new items in adjacent areas as an
indicator of the triclde down of benefits; transfers and turnovers as an indicator of
organizational capability (Chambers, 1980); and/or others suggested by the local
participants.

B Single or multiple visits to a particular village or visits to a numher of villages (Beebe,

1985). Also visits outside study area are important in understanding the dynamics of
the study area (Grand staff et al., 1987).

C Continuous team meetings and discussions (Grandstaff et al., 1987).

D Field research should last anywhere from four days to three weeks depending on
funding and time restraints (Grandstaff et al., 1987).

E One or several field study periods depending on the nature of the problem. Follow-up
visits are recommended (Beebe, 1985: Grandstaff et al., 1987).

A.4 PHASE m: COMPLETING THE RRA

A Team discussion and analysis immediately following fieldwork.

B Results formalized and recorded. The report is jointIy written by the team or by one
individual in consult with the rest of the team. The report should inc1ude the results,
methodology and recommendations. Data èollection checldists should be attached to
the report. Audiovisual aids should be used where appropriate (80000, 1985; Rhodes,
1982; Grandstaff el al., 1987).

107
TSBF: A Handbook of Metlwds

C Copies of the report sent to aIl interested agencies, offices, and individuals.
Publication should be done where appropriate (Grand staff et al., 1987).

D Leaming should not stop when report is written. RRAs can often result in future
research projects and need to be factored into decision making (Grand staff et al.,
1987; Beche, 1985).

E Follow-up and evaluation of how well the target population is adapting to the new
innovation is crucial. This can be done in the iteraûve process or by a follow up
study if funds are sufficient.

References

Beebe. 1. (1985) Rapid Rural Appraisal: The CriticaJ Firsl Siep in Famzing Syscems
Approach 10 Research. Farming Systems Suppon Project, Networking Paper No, 5,
USAID, Washington, D.C.
Bryant, C.A. and Bailey, D.F.C. (1991) The Use of Focus Group Research in Program
Development. In: van Willigen, 1. and Finan, T.L. (eds.), Soundings: Rapid and
Reliable Methodsfor Practising Anthropologists, NAPA Bulletin 10, 24-39, American
Anthropologica1 Association, Washington, D.C.
Chambers, R. (1980) Shon Cul Methods in lriformation Gatheringfor Rural Development
Projecls. Paper for World Bank Agricultural Sector Symposia, lanuary.
Chambers. R. (1991) In: Cernea, M.M. (00.), Putring People Firsl; Sociological Variables
in Rural Deve/opme"', 2nd Edition. Oxford University Press. New York, pp.515-
534.
CIMMYT (1985) Teaching NOies on lhe Diagnostic Phase of OFR/FSR Concepts.
Principles and Procedures. An occasional series of papers and notes on
methodologies and procedures useful in farming systems researeh in the economic
interpretation of agricultural experiments, No. 14, November. CIMMYT East
African Economies Programme, Nairobi, Kenya.
Doorman, F. (1991) A framework for the rapid appraisal of factors that influence adopûon
and impact of new agriculmra1 technology. HUflUiUJ OrganizaIion 50, 235-244.
Grandstaff, S., Grandstaff, T. and Ulve1ace, G. (1987) Proceedings of che 1985
InremMÎQnaI Cotiference on Ropid Rural Appraislli. Khon Kaçn University, Khon
Kaen, Thailand: Rural Systems Rcsear.;;h and Farming Systems Research Project.
Hildebrand, P.E. (1981) Combining Disciplines in Rapid Appraisa1: The Sondeo Approach.
Agricultural Adminiscration 8, 423-432.
Honadel, G. (1979) Organization and Administration of Integrated Rural Development.
Working Paper No. 1. Rapid Reconnaissance Approaehes to Organizational Analysis
for Development Administration. Development Alternatives, Ine. Washington, D.C.
IBSRAM (1988) Methodological Guide1ines for IBSRAM Soil Management Networks.
Second draft, ApriL IBSRAM, Bangkok, Thailand.
Raintree, J .B. (1987) D and D User's Manual. An lnt.r~tion to Agroforeslry DiagfllJsis
and Design. ICRAF, Nairobi, Kenya.
Rhodes, R. (1982) The Art of the InformaI Agriculcural Survey. Training DOCument 1982-
2, Social Science Department, International Potato Centre. Lima, Peru.
Roche1eau, D., Weber, F. and Field-Juma, A. (1988) In DrylandAfrica. ICRAF, Nairobi,
Kenya.

108
 Appendix B THE RHIZOSPHERE
Glynn Bowen

B.I INTRODUCTION

This appendix gives a general overview of the root environ ment (the rhizosphere), its
physica1, chemica1 and biologica1 nature, and the processes occurring there. A fairly detailed
entry into the literature is given by Bowen and Rovira (1976), Foster and Bowen (1982),
Rovira el al. (1983), Marschner (1986) and Kesster and Cregan (1991). As well as the
non-infecting organisms which grow in the rhizosphere, most organisms infecting plant roots,
be they symbionts or pathogens have a rhizosphere phase before infection occurs and/or
during sprea.d of the organism to other infection sites. The growth/survival of symbionts in
the rhizosphere cao be critica1 to symbioses. Nevertheless, these are not singled out for
special discussion in this appendix. Mycorrhizas are discussed in detail in Appendix C.

The rhizosphere refers to the zone of soil adjacent to roots. This bas been further subdivided
by sorne to the "outer rhizosphere", the "inner rhizosphere" and the "rhizoplane" i.e. the root
surface itself. The extent of the rhizosphere is quite ill-defined and variable depending on
the soil conditions affecting diffusion of solutes or volatiles from the root (and their biologica1
activity); - it cao be from 1 mm to approximately 10 mm (e.g. with volatiles). The inner
rhizosphere, the zone of most intensive change, is genera11y considered to extend 15-20 JJ.m
from the root.

B.2 mE RHIZOSPHERE ENVIRONMENT

Losses of organic substrates from roots to the soil cao be substantial - usually between 10%
and 20% of plant production in non-sterile soil (see Rovira el al., 1983). These large losses,
which started emerging from studies in the mid 1970s, are 10-100 times those indicated by
the earliest studies of root exudation. There are severa! reasons for this discrepancy: losses
fmm solution culture (early studîes) are considerably less than into a solid medium such as
soH (the presence of soil micm-organisms approximately doubles these IOsses) and studies in
sterile solution culture often use only very young plants (losses increase as roots age because
of an increase in root senescence).

Much of the early literature refers to losses of organics from roots as "root exudates", but
it is now' clear that "losses" cao be from severa! sources (Rovira el al., 1979). Loss of
soluble compounds fmm young, intact cens have been distinguished as:

1 "exudates" (leaking passively from cells). These low molecular weight substances are
readily available to micro-organisms for energy and inc1ude 8ugars, organic acids,
amino acids and several other compounds. The ratio of each varies with

109

l '
TSBF: A Handhook of M~lhods

environ mental factors which affect the composition of the cytoplasm (see Bowen and
Theodorou, 1973, Graham et al., 1982); often sugars and organic acids far exceed
amino acids; and

2 "secreti()nsn (compounds of low and high molecular weight lost by active movement
out of the œIl or "secretion"). Roots are usually surrounded by mucilages (high
molecular weight polysaccharides) of both plant and bacterial origin (collectively
refem:d ta as "mucigel"). The plant derived polysaccharides are probably produced
by Golgi bodies and secreted by coot cap cells and epidermal cells near the apex.
There is little doubt that this is secretion in its true sense. Similarly H+ extrusion
pumps occur in root ceUs. However, with low molecular weight compounds there are
almost no good data on whether they are leaked along a concentration gradient
(passive 1055, "exudation n) or actively expelled or "secreted". Because the
composition of the solutes lost from roots is often similar ta that of the roots
themselves, passive loss (exudation) May he the major mechanism of loss of low
molecular weight compounds from intact cells.

Cells which are senescing, however, may be a mcgor source of organic loss (perhaps ~
major source once a ceIl matures). Such losses have been identified as n1ysates n. They
increase as the root matures and ages, e.g. from the cortex and sloughed root cells. There
is sorne evidence that reports of extensive death of cortical cells of young roots (e.g. Henry
and Deacon, 1981) may be due to staining artifacts (Wenzel and McCully, 1991). With roots
growing in non-sterilised soils, a number of the cells only a centirnetre or two from the apex,
are empty and heavily colonised by bacteria; it May be that sorne of these ceUs were killed
prematurely by sub-clinical pathogens. The process of ageing and death of ceUs increases
with age of the root and is probably accelerated in the presence of soil organisms (including
sub-cliniœl pathogens). For example Martin and Kemp (1985) found when wheat wu grown
in fumigated soi1 and given a pulse of 14C01 at the carly tillering stage, sorne 14.6% of the
label appeared as COl (from root respiration and respiration of micro- organisms growing on
root substrate) in 80il over 14 days but 26.7% was lost as COz from unfumigated soil.

Sterile roots senesce with rime, releasing nlysates" and the important fact is that
micro-organisms may change the time course of this dramatically, which in tum could have
follow-<>n consequences for soil microbial processes. Even in the absence of
micro-organisms, however, there is a large throughput of organic matter from roots into soil
even quite early in the plant' S life.

How important quantitatively is carbon loss from intact cells, compared with carbon losses
via respiration and via "Iysates"? The most appropriate methods to find out involve 14COZ
labelling in closed systems and recovering 14C from shoots, roots. soil and periodic trapping
of 14C02 flushed frorn the soil (a measure of root respiration and microbial respiration from
root derived substrates) (Barber and Manin, 1976). The soil 14 Ç can be separated into water
soluble fractions (probably uncharged solutes and some cations) and water insoluble fractions
(sorne high molecular weight compounds t bacr.erial colonies, polysaccharides, sloughed root
fragments) .

The data in Table B.l from Barber and Martin (1976) indicate the relative order of various
fractions during the first three weeks of a wheat plant. Note that in sterile plants only sorne
0.9% of 14C assimilated was lost as low molecular weight compounds and may he largely

110
 Appendix B: The RhizosphLre

exudate, Le. slightly less than root respiration. In non-sterile plants the situation was less
clear, but a considerable portion of the respired 14C02 (almost 8 % of the total assimilate of
the plant compared with 1 % in sterile soil) would have arisen from increased leakage from
the root and sorne breakdown of sloughed plant parts, partly induced by the soil microflora.
There appears to he little profitability in breeding plants with less leakage of materials from
intact cells, if the 1% indicated (from sterile plants) in Table B.1 is typica1. The value of
manipulation of the "insoluble" fraction (sorne 5-6%) is debatable due to lack of knowledge
of its composition (polysaccharide or root cells) and the function and "need" for the
polysaccharide production by the root, if any.

The high molecular weight polysaccharides produced by the plant have been partly
characterised by electron-microscope studies in which various hydrolysis agents are used and
the residual polysaccharide stained (see Foster and Martin, 1981; Foster and Bowen, 1982).
Their functions have been speculated on, two of the most attractive being that they mayact
as a lubricaot for the growing root through soil, but more particularly that they may serve as
an effective continuous contact hetween soil and the root passage of nutrients and water.

Although it is genera11y thought that mot "exudation" and therefore substrate for
micro-organisms is greatest near the root tip, quite active colonisation of parts of the roots
which are 10 days old cao occur (Bowen and Rovira, 1973). Such studies of microbial
growth on roots inoculated at various points of known temporal and physiologica1 age will
indicate the available substrate for microbial growth at any one time or place; however this
will not identify the sources of the substrate - exudation or lysis.

Table B.I Distribution and loss of photosynthetica11y flXed 14C in wheat plants grown in
sterile and non-sterile soil for 3 weeks in an atmosphere containing 0.03% 14C02 and with
a 16 hr photoperiod (results are expressed as mean of five replicates ± standard errors).

Distribution (% total)

Soil condition Roots Shoots Water- Soil in- Respired

soluble soluble

Sterilised 24.45 68.18 0.88 S.4S 1.02

1.76 ±2.32 ±0.O5 ±O.56 ±O.O6

Non-sterile 32.13 51.85 0.62 7.71 7.73

±2.65 ±2.37 ±0.15 ±O.41 ±O.68

B.2.I The movement of organic substrates into soil

If low molecular substrates (solutes and volatiles) are not used by microbial growth at the root
surface, they cao diffuse from the mot. Thus a concentration gradient away from the root
occurs, the steepness of which depends on the diffusion coefficient of the substrate, and
modifications caused by interacting soil factors such as tortuosity ofpaths through soil, charge

111
TSBF: A Handbook 01 MelJuNh

on clay surfaces, soil moisture etc. The distribution of exudates and microbial growth with
respect 1'0 time and distance from the root has been modelled by Newman and Watson (1977).
Calculations of microbial growth was 234 and 23.5 J.l.g/cm3 soi! in wet soil and 7355 and 4.4
J.l.g/cm3 in dry soU at the mot surface and at 0.3 mm from the root surface respectively. Such
modelling bas many important }X)tential uses but it must be remembered that it is a simplified
version of the situation, often with many untested assumptions, and that the pre4ictions by
any mode1 must he tested.

In general, carbon dioxide and other volatiles will diffuse through soil for greater distance
than solutes. Distant from the root (e.g. 4 to 5 mm), the volatiles may not be important as
substrates, but they may he important in triggering biological activity, after which growth
may follow the concentration gradient 1'0 the root (reviewed by Stotzky and Schenck, 1976).
Roots may aIso produce exudates with specifie stimulative effects on partieular micro-
organisms, e.g. flavonoids and isoflavonoids (Bowen, 1991; Phillips et al., 1991).

Such considerations as those above are important not only for defining general microbial
growth around mots but also for studying the ecology and relations between soil populations
and root infection of specified economically important organisms (ftepidemiology"), e.g.
mycorrhizal fungi or plant pathogens. With these, and other organisms which cao be
recognised at the mot surface, questions such as the radius of influence of the mot into soil
cao be approaehed by placing propagules at med distances from a root and subsequently
examining the mot surface for the presence of the organisms (see Bowen, 1979, 1980) (see
also Rhizosphere Dynamics, etc., below).

B.2.2 The lnorganiC! nahlJ:'e of the l'VUt surface and rhÎ%osphere

The root surface cao differ from bulk soil in pH and in concentration of anions and cations
(see Nyc and Tinker. 1977). Ions which are rapidly absorbed by the root, may he at very
low concentration at the root surface. On the other band, sorne ions cao be excluded from
the mot and local high concentrations cao accumulate and even precipitate around the root
e.g. CaH (in sorne cases) and iron oxides in sorne flooded soils. As cations are absorbed they
cao be replaced by H+, and in this way there can be a reduction of rhimsphere pH from 5
to 4 with the absorption ofNH.. + from ammonium sulphate. Conversely, rhizosphere pH cao
rise following absorption of anions such as nitrate. Chang~ in rhizosphere pH and redox
J)Otentials are discussed by Marschner (1986). The pH of the rhi;r;osphere of legumes fixing
atmospheric nitrogen con also dec1ine (Robson, 1982). These pH changes and aS80Ciated
effecu on SOlubility of other ions cao have very marked selection eff~ts on the rhizosphere
microflora (Smiley, 1979; Foster and Bowen, 1982).

Present method~ for analysing the inorganic composition of rhizosphere soil include:

1 autoradiography (Lewis and Quirk, 1967; Barber and Ozanne, 1970) where ions have
appropriate radioactive tracers (a semi-quantitative methcxt)~ and

2 more generally, more careful collection of soi! adhering ta the foot on removal from
soi] (Riley and Barber, 1969; Salœr et al., 1979).

112
 Appmdix B: 'lM RlIil.ospheu

B.3 mE MICROBIAL POPULATIONS ON ROOTS

B.3.1 Culture of micro-organisms

The populations of bacteria in soil c10sely attached to the root (the rhizosphere soil) which
are isolated on general media e.g. potato dextrose agar or Czapek-Dox agar (fungi), tryptic
soy agar (bacteria) (Martin, 1975) using dilution counts or most probable number methods
range from 5 - 30 x 108/g rhizosphere soil. A comparison of these general counts (or of
specific organisms) with non-rhizosphere soi1leads to the c1assic R:S ratio, Le. a measure of
the stimulation of micro-organisms by the rhizosphere - it cao range up to 100 depending on
the plantlmicro-organismlgrowth conditions (see Rovira and Davey, 1974). Counts of
bacteria on the rhizoplane on a surface basis are not corn mon but are sorne 0.5-4 x 104/mm,
but these may be spread over only sorne 10% of the surface (see below). However the
micro-organisms growing on laboratory media may only be a small percentage of the total;
Louw and Webley (1959) found that the ratio of bacteria in total (direct microscopic) counts

,
and a laboratory media varied from 1.4 to 5.1 for rhizosphere soil. A very wide range of
i bacteria colonise the rhizosphere (see Rovira, 1965) and there is a selective stimulation of
Gram-negative organisms requiring amino acids. One major genus is Pseudomonas;

1
f~ nevertheless, it usually represents only between 5 and 10% of the total counts.

Many fungi are also isolated from the root but dilution counts on laboratory media usua1ly
refleet the relative number of propagules of each, not the biomass, unless special techniques
are used (e.g. washing and fragmentation of the roots) (see Sharley and Waid, 1955);
non-sporing fungi, and slow growing fungi are missed.

In short, therefore, plate counts of rhizosphere bacteria give a distorted and partial view both
of the range of organisms present and the microbial biomass on the root. Such counts are
useful for general comparative purposes and in particular for studying the ecology of selected
organisms. Apart from issues related to overall microbial biomass (e.g. tie up of nutrients),
the study of particular organisms (usua1ly economica1ly important ones) is the direction which
will be most rewarding for understanding rhizosphere ecology and interactions. For studying
particular organisms various selective media have been developed: e.g. Simon and Ridge
(1974) - fluorescent Pseudomonas; Brisbane and Kerr (1983) - Agrobacterium; and Tsao
(1970) - various fungi. A further development of this is the selection of an antibiotic resistant
mutant (usually a "double" mutant) of a particular species so that it cao easily be isolated
from a mixture of the same genus e.g. strains of Ps. fluorescens resistant to rifampin and
naladixic acid have been used (Weller and Cook, 1983).

B.3.2 Direct observation

Direct observations of the rhizosphere have given good infonnation on the distribution of
micro-organisms; with histochemica1 techniques, especially at the transmission electron
microscope level, insight into the physica1 and chemica1 nature of the root soil interface bas
been obtained. With the light microscope, staining of the rhizoplane microflora (e.g. Jones
and Mollison, 1948), followed by appropriaœ analysis derived from higher plant ecology. e.g.
quadrat counts, have shown that often only sorne 10% of the root surface is occupied hy

113
, ct i

III
TSBF: A Handboolc of Melhods

micro-organisms and that their distribution is non-random (see Bowen and Rovira, 1976;
Rovira el al., 1983). The sources of this non-randomness are twofold:

a) despite the high numbers of bacteria in soil, there is considerable spatial heterogeneity
ofbacterial distribution (Uley are usually associated with organic matter), thus the root
gmwing through soU is not unifonnly in contact with bacteria; and

b) although aIl parts of a root surface lose exudates, the longitudinal junctions of adjacent
cells are major sites of exudation (see Bowen, 1979); these are often (not always) the
preferred sites of growth and migration.

Fluorescence microscopy to distinguish dead from living organisms (Trolldenier, 1965), and
the use of immunofluorescence and associate techniques such as the use of monoclonal
antibody techniques to distinguish particular organisms (e.g. Malajczuk el al., 1975; Schmidt,
1991), have aIso been powerful techniques.

Scanning and transmission electron microscopy have made important contributions to our
understanding of spatial distribution of rhizosphere micro-organisms and their environ ment
(for details, see Foster and Bowen, 1982; Foster et al., 1983; Rovira et al., 1983).
Physiologica1ly, the mucigels of mot and microbial origin and localisation of enzymes have
been a major interest. Electron probe microanalysis will add a further dimension 10
rhizosphere studies. Electron microscopy bas aIso shown (i) a great diversity of bacterial
forms (often not secn in rich laboratory media), many of which may be restricted 10 a
specialised habit in grooves between epidermaI cells; (H) a ten-fold decline in microbial
numbers between the root surface and 15-20 J.Lm away (the outer rhizosphere) (Foster and
Rovira, 1978); (iii) that rhizosphere bacteria are often filled with storage materiaIs such as
polyhydroxybutyrate, polysaccharide granules and polyphosphate granules; and (iv) that the
outer cortex of apparently healthy roots may be inhabited by high numbers of microorganisms
(the "endorhizosphere") which hasten the collapse and decay of the cortex.

It is likely that DNA probes atone, or coupled with polymerase chain reaction (peR)
amplification (Holben and Tregje, 1988; Asim et al., 1990), will emerge as powerful tools
in following the dynamics of rhizosphere micro-organisms without of the need to culture
them. This will be panicu1arly useful for organisms difficult to grow on laboratory media.

B.3.3 Rhizosphere dyoamics, interaction and experimental approaches

The rhizosphere is adynamie ecosystem with the same broad considerations which occur with
macro-ecosystems. The same questions of growth rates, pioneers, succession and interactions
(synergism, competition and antagonism) occur. A knowledge of rhizosphere ecology is
important in understanding the establishment problems of inoculated beneficial
micro~rganisms (such as Rhizobium, Frankia, mycorrhizal fungi or bio-control organisms)
and in reducing the impact of disease organisms. An appreciation of the dynamics of
colonisation of and interactions in the rhimsphere CM be achieved ~ by experimental
approaches. Howen (1979, 1980) pointed out the obvious, that once the various components
of a system are identified, simple experimental approaches carl be devised to test the
importance of these components, e.g. spore germination, .movement to the fOot, growth at the
root, the influence of soil moisture or growth of organisms along the root and so on.

114
 Appendix B: The Rhir.osphere

Inductive (rnodelling) apprœches to sorne of these features such as growth through soil have
aIso been developed - see Gilligan (1983) (plant pathogens), and WaIker and Smith (1984)
(mycorrhizas).

Growth rates along roots, a corner stone of population dynamics have been studied for fungi
under various conditions by inoculation of the root at known positions and measuring growth
with time (e.g. Bowen and Theodorou, 1979). Growth rates for bacteria in the rhizoplane
were measured for various groups of organisrns by planting sterile roots 10 cm long in
non-sterile soH, harvesting particular segments each at 0, 1, 2, 3 and 4 days, counting on
selective media and constructing growth curves (Bowen and Rovira, 1973; Bowen, 1979).
These studies indicated an initiaIlogarithmic growth rate over 3-4 days (the slope of which
was interpreted as an indication of the richness of the substrate coming from the root),
followed by a slower phase (interpreted as a "statiooary" level where substrate balances
maintenance energy, and upon which migration and growth in previously uncolonised parts
is superimposed). This apprœch allows a calculation of the basic parameter, "generation"
time or "doubling" time. Comparison of generation time on roots and soil is a more correct
way to define the rhizosphere effect than the R:S ratio. The generation times of
Pseudomonas and Bacillus on the root were 5-6 and 39 hours respectively, and in soil they
were 77 and > 100 hours, respectively. The use of generation time in rhizosphere population
dynamics is increasing, albeit slowly. The use of single gene mutants of bacteria is an
emerging, powerful apprœch to questions on growth dynamics in the rhizosphere (Bowen,
1991; Lam et al., 1991).

B.3.4 Interactions

In general, extrapolation from growth and interaction of organisms in laboratory media to

what happens around roots is perilous (Bowen, 1980) and the studies must be performed in
soil. Synergism, competition and antagonism occur in microbial growth in the rhizosphere
(e.g. Bowen and Theodorou, 1979). Perception of synergism is not difficult but distinction
between competition and antagonism is more difficult (but see Bowen and Theodorou, 1979).

No statement of interactions, however brief, would be complete without reference to the very
neglected field of soil fauna-microflora interaction. The importance of grazing of bacteria
and fungi by soil micro-fauna is becoming increasingly recognised. For example, using
experimental apprœches to root colonisation, Chakraborty et al. (1985) demonstrated that
sorne amoebae caused 50-70% reduction of root colonisation by ectomycorrhizal fungi;
Wamock et al. (1982) showed that grazing ofvesicular arbuscular mycorrhiza hyphae by the
collembola Foisomia candida significantly reduced the mycorrhizal plant growth response;
Coleman et al. (1978) showed major interactions between hacteria, amoebae and nematodes
affecting biomass and nutrient dynamics in soil microcosms simulating the rhizosphere.

B.3.S The effects of the rhizosphere micronora

Non-infecting rhizosphere organisms may have several effects on plant growth. Sorne of
these are direct and sorne indirect, such as the control of plant pathogens (e.g. see Rovira and

115
'r.iBF: A HantlboNI I1f M4tMtl.J

Wildermuth, 1981; Wong, 1981). As indicated above, biological control of symbionts may
also occur.

Un1ess special precautions for axenic culture are observed for control treatments, it is often j'
extremely difficult to interpret the effects of the general rhizosphere microflora on plant
growth. Various solution culture and soU culture methods have bren devised for this (Barber,
1967; Rovira and Bowen, 1969). One of the greatest problems is in the use of soil, in which .~
autoclaving sometimes produces toxins (especially in high organic matter soils). The least
severe sterilising method is the use of wamma-irradiation but even this can produce sorne
chemica1 changes (McLaren, 1969). Severa! good soil fumigation methods exist, but with
these complete sterility may not occur; e.g. the use of aerated steam (Baker and OIson, 1960),
methyl bromide or dazomet (for small quantities of soU) (van Berkum and Hoestra, 1979).
Partial sterilisation is valuable in eliminatinw certain components of the soil microflom and
e~ where particular groups of organisms are affected, e.g. with nematicides.

The availability of nutrients such as manganese and molybdenum cao be affected by the
rhizosphere microflora (see Rovira el al., 1983). Although phosphorus solubilisation by
organisms in the rhizosphere has becn proposed for some time, the ecological imponance of
this is still open to question, as is nutrient immobilisation. There is evidenee that the
rhizosphere May sometimes enhanee denitrification but this is still an area for clarification
(Firestone, 1982). It is also claimed that ~ of sorne plants cao inhibit nitrification in soil
(Riee, 1974); this needs detailed study. Similarly, the extent of fixation of nitrogen in the
rhizosphere by associative nitrogen fixers such as Azospirillum, and Azobacler paspali is still
debatable due partly to inherent difficulties in field rneasurement. Claims for quanûûes fixed
range from a few Wha/year to posgibly 200 kglhalyear. Difficu1ties arise from the
inaccwacies of N balance methods over long period~ and from extrapolaûon trom ~mall
sample, short term incubation studies (u~ing Acetylene reducûon methodS). Nevertheless, the
evidenee indicates important accessions by associative N fixation, e.g. sometimes 30-50 kg
N/ha/year in undisturbed ecosystems (e.g. grass SWaMs) OVet a lonw period. The subject is
discussed by Dobereiner (1983), Jordan (1981) and Legg and Meisinger (1982).

There are a number of indications that non-infecting rhizosphere micro-organisms produce

biologîcally active substances affecting plants, e.g. effects on root growth and root haïr
production, the stimulation of intense tateraI branching of sorne plants (ca11ed "proteoid" or
"cluster" roots), the stimulation of phosphate uptake and transportaûon, and the induction of
early flowering. The lut of these can he caused by a number of organisms, of which one,
Azob~ttr, Wa:i previously thought to stimulate plant growth by fixing nitrogen in the
rhiwsphere. SimilarlYt increa.~J in plant Wrowth sometimeJ c.aused by inoculation with
Azospirillum May be due to hormonal effects (Beshan and Levanony, 1990). The production
of plant growth ptomoting substances is dîKussed by Rovira et al. (1983). Such micro-
organisms are referred to as Plant Growth Promoting Rhizobacteria (pGPR); responses ro
inoculation are often erratic. Depending on environ mental factors and plant species, certain
strains of rhizosphere Pseudomonas spp., and sorne other metabolites such as HCN, may
inhibit or enhance plant establishment or inhibit development of plant disease (Schippen et
al., 1991).

116
 Appmdlx B: 'I1u! Rhizosphere

References

Asim, K.B., Steffen R.J., di Cesare, J., Haff L. and Atlas R.M. (1990) Detection of
coliform bacteria in water and polymerase chain reaction and gene probes. Applied
and Environmental Microbiology 56, 307-314.
Baker, K.F. and Olsen, C.M. (1960) Aerated stearn for soil treatment. Phytopathology SO,
82.
Barber, D. A. (1967) The effects of microorganisms on the absorption of inorganic nutrients
by intact plants. 1. Apparatus and culture techniques. Journal of Experimental Botany
18, 163-169.
Barber, D.A. and Martin, J.K. (1976) The release of organic substances by cerea1 roots into
soH. New Phytologist 76, 69-80.
Bashan, Y. and Levanony, H. (1990) Current status of Azospirillum inoculation technology:
Azospirillum as a challenge for agriculture. Canadian Journal of Microbiology 36,
591-608.
van Berlrum, J.A. and Hoestra, H. (1979) Practical aspects of the chemical control of
nematodes in soil. In: Mulder, D. (ed.), Soil Disirifestation, Elsevier Scientific
Publishing Company, Amsterdam, pp.53-134.
Bowen, G.D. (1979) Integrated and experimental approaches to the study of growth of
organisms around roots. In: Schippers, B. and Gams, W. (eds.), Soil-Borne Plant
Pathogens, Academic Press, London, pp. 209-227 .
Bowen, G.D. (1980) Misconceptions, concepts and approaches in rhizosphere biology. In:
El1wood, D.C., Hedger, J.N., Latham, M.J., Lynch, J.M. and Slater, J.H. (eds.),
Contemporary Microbial Ecology, Academic Press, London, pp. 283-304.
Bowen, G.D. (1991) Microbial dynamics in the rhizosphere: Possible stategies in managing
rhizosphere populations. In: Keister, D.L. and Cregan, P.B. (eds.) Rhizosphere and
Plant Growth: Symposium Proceedings, Kluwer Academic Publishers, Dordrecht.
pp.25-32.
Bowen, G.D. and Rovira, A.D. (1973) Are modelling approaches useful in rhizosphere
biology? Bulletin ofEcological Research Communication (Stockholm) 17, 443-450.
Bowen, G.D. and Rovira, A.D. (1976) Microbial colonisation of plant roots. Annual
Review of Phytopathology 14, 121-144.
Bowen, G.D. and Theodorou, C. (1973) Growth of ectomycorrhizal fungi around seeds and
roots. In: Marks, G.C. and Kozlowski, T.T. (eds.), Ectomycorhizae - The Ecology
and Physiology, Academic Press, New York, pp.l07-150.
Bowen, G.D. and Theodorou, C. (1979) Interactions between bacteria and ectomycorrhizal
fungi. Soil Biology and Biochemistry Il, 119-126.
Brisbane, P.O. and Kerr, A. (1983) Selective media for three biovars of Agrobacterium.
Journal of Applied Bacteriology 54, 425-431.
Chakraborty, S., Theodorou, C. and Bowen, 0.0. (1985) The reduction ofrootcolonization
by mycorrhizal fungi by mycophagous amoebae. Canadian Journal of Microbiology
31, 295-297.
Coleman, D.C., Anderson, R.V., Cole, C.V., Elliot, E.T., Woods, L. and Campion, M.K.
(1978) Trophic interactions in soUs as they affect energy and nutrient dynamics. IV.
Flows of metabolic and biomass carbon. Microbial Ecology 4, 373-380.
Dobereiner, J. (1983) DiniU'Ogen fixation in rhizosphere and phyllosphere associations. In:
Lauchli. A. and Bielesld. R.L. (ed8.), lnorganic Plant Nutrition. Encyclopedia of
plant physÎology: new set. v; lj. Springer-verlag, BerlÎn, pp.331-350.

117
Firestone, M.K. (1982) Biologica1 denitrification. In: Stevenson, P.J. (ed.), Nitrogen in
Agricultural Soifs. American Society of Agronomy, Madison, Wisconsin, pp.289-326.
Foster, R.C. and Bowen, a.D. (1982) Plant surfaee and bacterial growth: the rhizosphere
and rhizoplane. In: Mount, M.S. and Laeey, G.H. (eds.) , Phytopathngenic
PMkaryotes 1. Academie Press, N~w York, pp. 159-185
Foster, R.C. and Martin, J.K. (1981) In situ analysis of soi! components of biologica1
origin. In: Paul, E.A. and Ladd, J.N. (eds.), Soil Biochemistry 5. Marcel Dekker,
New York, pp.75-111
Foster, R.C. and Rovira, A.D. (1978) The ultrastructure of the rhizosphere of Trifolium
subterraneaum L. In: Loutit, M. W. and Miles, J.A.R. (eds.), Microbial Ecology.
Springer-Verlag, Berlin, pp.278-290.
Poster, R.C., Rovira, A.D. and Cock, T.W. (1983) Ultrastructure of the Root-Soil
Inteiface. American PhytopatJwlogy Society, St. Paul, Minnesota.
Gilligan, C.A. (1983) Modelling of soi! borne pathogens. Annual Review of Phytopathnlogy
21,45-64.
Graham, J.H., Leonard, R.T. and Menge, I.A. (1982) Interaction of light intensity and soil
temperature with phosphorus inhibition ofversicular-arbuscular mycorrhiza fonnation.
New Phytologist91, 683-690.
Harley, I.L. and Waid, J.S. (1955) A method of studying active mycelia on living fOOts and
other surfaces in the soil. Transactions of1he British Mycology Society 38, 104-118.
Henry, C.M. and Deacon, J.W. (1981) Natural (non-pathogenic) death of the cortex of
wheat and barley seminal mots, as ~videnced by nuclear staining with acridine
orange. Plant and Soi/60, 255-274.
HOlben, W.F. and Tiedje, J.M. (1988) Applications of nucl~îc acid hybridization in
microbiaI eçology. Ecology 1i9, S61-~68.
Holden, J. (197S) Use of nuclear staining to assess rates of cell death in cortiees of cereal
footS. Soil Biology and Bi<>ehemislry 7, 333-334.
Jordan, D.C. (1981) Nitrogen fixation by selected free-living and associative
microorganisms. In: Gibson, A.H. and Newton, W.E. (eds.), Currenr Perspectives
in Nitrogen Fixation. Australian Acad~my of Science, Canberra, pp.317- 320.
Keister, D.L. and Cregan, P.B. (1991) The Rhizosphere and Plant Growth. Kluwer
Acad~mic Publishers, The Netherlands, 386pp.
Lam, S.F., Ellis. D.M. and Ligon, J.M. (1991) Oenetic approaches for studying
rhizosphere colonisation. Plant and Soil 129, 11-18.
Legg, 1.0. and Mersinger, J.J. (1982) Soil nitrogen budgets. In: Stevenson, F.J. (00.),
Nitrog~n in Agricullural Soi/s. American Society of Agronomy, Madison, Wisconsin,
pp.503-566
LeWiS, 0.0. and Quirk, J.P. (1967) Phosphate diffusion in soit and uptake by plants III. 3lP
movement and uptake by plants as indicated by 32P auto radiography. Plant and Soil
26, 445-'453.
Louw, H.A. and Webley, D.M. (1959) The bacteriology of the root region of the oat plant
grown under controlled pot culture conditions. Journal of Applied Bac/eriv/ogy ll,
21«$-226.
McLaren, A.D. (1969) Radiation as a technique in soil biology and biochemistry. Svil
Biology and Biochemislry l, 63-73.
Mal~czuk, N., McComb, A.l. and Parker, C.A. (1975) An immunofluorescence technique
for detecting Phytophthora cinnamomi Rands. Australian Journal of Bolany 23,
2 8 9 - 3 0 9 · î
J •. j

118
 Appendix B: The Rhizosphere

Marschner, H. (1986) Mineral Nutrition of Higher Plants. Academie Press, London.

674pp.
Martin, J.K. and Kemp, J.R. (1986) The measurement of C transfers within the rhizosphere
of wheat grown in field plots. Soil Biology and Biochemistry 18, 103-107.
Nye, P.H. and Tinker, P.B. (1977) Solute Movement in the Soil-Root System. Blackwell
Scientific Publications, Oxford.
Phillips, D.A., Maxwell, C.A., Hartwig, V.A., Joseph, C.M. and Wery, J. (1991)
Rhizosphere flavonoids released by aIfaIfa. In: Keister, D.L. and Cregan, P.B.
(eds.). Rhizosphere and Plant Growth: Symposium Proceedings. Kluwer Academie
Publishers, Dordrecht, pp. 149-154.
Rice, E.L. (1974) Allelopathy. Academie Press, London.
Riley, D. and Barber, S.A. (1969) Bicarbonate accumulation and pH changes at soyabean
(Glycine max. (L.) Merr.) root-soU interface. Soil Science Society of America
Proceedings 33, 905-908.
Robson, A.D. (1982) Mineral nutrition. In: Broughton, W.J. (ed.), Nitrogen Fixation 3.
Oxford University Press, Oxford, pp.36-52.
Rovira, A.D. (1965) Interactions between plant roots and soil microorganisms. Annual
Review of Microbiology 19, 241-266.
Rovira, A.D. and Bowen, G.D. (1969) Phosphate incorporation by sterile and non-sterile
plant roots. Australian Journal of Biological Science 19, 1167-1169.
Rovira, A.D. and Davey, C.B. (1974) Bio10gy of the rhizosphere. In: Carson, E. W. (ed.),
The Plant ROOl and Ils Environment. University Press of Virginia, Charlottesville,
pp. 153-204
Rovira, A.D. and Wildermuth, C.B. (1981) The nature and mechanisms of suppression. In:
Asher, M.J.C. and Shipton, P.J. (eds.), Biology and Control of Take-all. Academie
Press, London, pp.385-415.
Rovira, A.D., Foster, R.C. and Martin, J.K. (1979) Origin, nature and nomenclature of the
organic materiaIs in the rhizosphere. In: Harley, J.L. and Scott Russell, R. (eds.), The
Root-Soil Interface. Academie Press, London, pp. 1-4.
Rovira, A.D., Bowen, G.D. and Poster, R.C. (1983) The significance of rhizosphere
microtlora and mycorrhizas in plant nutrition. In: Lauchli, A. and Bielesld, R.L.
(eds.), Inorganic Plant Nu/rition (Encyclopedia of plant physîology: new sere v. 15).
Springer-Verlag, Berlin, pp.61-93.
Sarkar, A.N., Jenkins, D.A. and Wyn Jones, R.G. (1979) Modifications to mechanical and
mineralogical composition of soil within the rhizosphere In: Harley, J.L. and Scott
Russell, R. (eds.), The Root-Soil Interface. Academie Press, London, pp. 125-136.
Schippers, B., Bakker, A.W., Bakker, P.H.A.M. and van Peer, R. (1990) BeneficiaI and
deleterios effects of HCN-producing pseudomonads on rhiroshpere interactions. Plant
and SoU, 129, 75-83.
Schmidt, E.L. (1991) MetluJcMfor microbial au/ecology in the soil rhizosphere. In: Keister,
D.L. and Cregan, P.B. (eds.), Rhizosphere and Piani Growlh: Symposium
Proceedings, Kluwer Academie Publishers, Dordrecht. pp.81-89.
Simon, A. and Ridge, E.H. (1974) The use of ampicillin in a simplified selective medium
for the isolation of fluorescent pseudomonads. Jou17Ul1 of Applied Bacleriology 37,
459-460.
Smiley, R. W. (1979) Wheat-rhizosphere pseudomonads as antagonists of Gaeumannomyces
gaminis. Soil Biology and Biochemislry 11, 371-376.

119
 TSlJF: A Han4boolc of M~rJwds

Stotzky, G. and Schenk, S. (1976) Volatile organic compounds and microorganisms.

Critical Review 01 Microbiology 4, 333-382.
Trolldenier, G. (1965) Fluoreucnmikroskipische Untersuchung von Mikroorganism-
enkulturen in der Rhizosphare. Z. Bakrerlol. Parœitenl«l Infekt. Hyg. III 119,
2~6-259.
Tsao, P.H. (1970) Selective media for isolation of pathogenic fungi. Annual Review of
PhytoPalhology 8, 157-186.
Walker, N.A. and Smith, S.E. (1984) The quantitative study of mycorrhizal infection. II.
The relation of rate of infection and spee« of fungal growth to propagule density, the
Mean length of the infection unit and the limiting value of the fraction of the root
infected. New Phytologist 96, 55-69.
Warnok, A.J., Fitter, A.H. and USher, M.B. (1982) The influence of a springtail Folsomia
candida (Insecta, collembola) on the mycorrhizal association of the leek Allium
porrum and the vesicular-arbuscular mycorrhizal endophyte Glomuslmiculatus. New
Phyrologisl !IO, 285-292.
Weiler, D. and Cook, R.I. (1983) Suppression of take-all of wheat by seed treatment with
fluorescent psçudomonads. Phytopm/fQlogy 73. 463-469.
Wenzel, C.L. and McCully, M.F. (1991) Barly senescence of cortical cells in the mots of
cereals: How good is the evidence? American Journal 01 Botany 78, 1528-1541.
Wong, P.T.W. (1981) Biologica1 control by cross-protection. In; Asher, M.J.C. and
Shipton, P.J. (OOs.), Biology and Coll/roi 01 Toke-all. Academie Press, London,
pp.417-431.

1
.;

120
Appendix C MYCORRHIZAS
David Read

C.I INTRODUCTION

AImost all of the major plant species of natural and agroecosystems of the tropics are
mycorrhizal. Literature surveys (Trappe, 1987) suggest that 85 % of tropical plant species
are mycorrhizal and that infection of the vesicular-arbuscular (VA) type is the most widely
distributed, occurring in ca. 70% of aIl species. These inc1ude the major food crops of
tropical regions, many of which are highly responsive to infection, at least under experimental
conditions (Table C.l).

Infections of the ectomycorrhizal or sheathing type, though found on fewer species, are
important because they occur on sorne of the most valuable timber trees of tropical forests.
Orchidaceous and ericoid mycorrhizas are also found in all tropical regions, the latter mostly
being restricted to high altitude situations. Approximately 13 % of tropical plant species
examined so far are non-mycorrhizal, these being largely confined to families all members
of which are characteristically not susceptible to infection by mycorrhizal fungi such as
Caryophyllaceae, Chenopodiaceae, Commelinaceae, Cruciferae, Cyperaceae, Juncaceae.
Phytolaccaceae, Polygonaceae, Ponulacaceae. Proteaceae. Restionaceae and Sapotaceae.

In the overwhelming majority of plant species that are mycorrhizal, their fungal associates
occupy those distal parts of the roots which are involved in absorption of nutrients from soil.
Any consideration of plant growth, crop production or ecosystem function in the tropics must
therefore inc1ude an analysis of mycorrhizal biology if it is to be complete. Understanding
of the processes of nutrient capture, transport and assimilation cannot be achieved without
knowledge of the role of the fungal symbionts with which most plants evolved and upon
which many are dependent in nature.

Assessment of the status and function of a mycorrhizal infection in any plant community
normally involves a series of observations aIl or sorne of which may be required depending
upon the objectives of the study. A typical series is:

1 Characterisation of infection type and analysis of occurrence of infection.

2 Quantification of occurrence.
3 Isolation and identification of fungi.
4 Inoculation of test plants and analysis of infection.
5 Analysis of host response, first in pot cultures, then in nature.

Basic procedures for such seriai analysis are described below for the major mycorrhizal types.

121
C.2 THE VESICULAR·ARDUSCULAR MYCORRHIZAL SYMBIOSIS

This ubiquitous symbiosis is formed by zygQmycetou$ fungi of the Order Glomales (Morton
and Benny 1990). Six genera Qf glomaIean fungi are recognized vil. Glomus. Acaulospora,
Gigaspora, Enterophospora, Sclerocystis and Scutellospora. There are more than UO speçies
worldwide, most of which show low level$ of hœt speçificity. As a consequence of this it
il probable that individual plants in speçies-rieh oommunities such as those of savannah and
humid tropical forests, most of which are heavily mycorrhizal (Newman el al., 1986; Janos
1980, 1987), win he interconnected by an extensive network of VA myœlium (Read el al.,
1985). This mycelium, which mpidly infects the fOOts of seedlings. serves primarily to
inc~ the volume of soi! exploited by the plant and hence ID improve its access to nutrients.
The principle advantage to the plant is that ions, in particular phosphates. which diffuse
s]owly in soil and are readily immobilised by sorption with oxide minerals can he captured
by the fungu$ and so rendered available (Harley and Smith, 1983).

By similar mechanisms VA fungi enhnnce the ability of plants to obtain phosphorus (P)
rclcased from slowly solubilizing fertilisers such as rock phosphate (Daft, 1991). Converse1y,
large inputs of soluble P, associated for example with application of superphosphate, cao
decrease or eliminate mycorrhizal advantages by inhibiting the growth and activity of the
vegetadve tnycelium (Abbott and Rob80n, 1984).

C.2.1 Characterisation and analysis of VA mytorrhizal infection

The hyphae of VA myoorrhizaJ funai penetra.te the outer and inner cortical tissues Qf roots
of host plants. sa that 'clearing' of a sample b norma11y required before infection cao he
observc:d. lbis can he achieve<t by heating the sample in 10% KOH at 90 0 Ç for severa!
minutes (phillips and Hay man , 1970). This treatment may cause exce~$ive disruption to
delicate roots in which case an alternative method involving incubation for up to 24 hours in
KOH at fnom temperature can be employed (Read et ul., 1976). In highly pigmented or
5uberise4 roots it may be necessary. following KOH treatment, to bleach tissues by
inlmersing thern in H1 02 (10% by volume) until they appear clear to the naked eye. Alter
'clearing', roots are rinsed in water, acidified with 2% Hel and stained.

A variety of stains have becn employe4 but, as pointed O"t by Graçe and Stribley (1991)
many of th~ are haurdous, iOme lilœ toluidine olue: being potential carcinogens. The
recommended procedure involves the use of 0.05% aniline blue or methyl blue, both names
are synonyms for cotton blue in acidified 70 % glyœrol. The cleared mot sample is immcrsed
in the stain for 30 minutes before exceu ,tain is removed and the sample mounted in 10%
glyœrine.

The diagnostic structures are the vesicles and arbuscules formed within the root cortical çells
the prc5ence of which enables the type of infection to he confirmed. The spores are fonned
outside the root. and enable identification of the fungi involved. Of the fungal structures, tllat
of greatest functional significance is the arbuscule, through which nument exchange takes
place between fungus and hosto The term "arbuscular" mycorrhiza is preferable to
vesicular-arbuscular because in the a.bsence of arbuscules the mycorrhiza is unlikely to be
effeçtive.

122
 C.2.2 Quantification of infection

The most convenient and widely used method for determining the extent of V A infection is
based upon the grid intersect method (Tennant, 1975; Ambler and Young, 1977). Stained
roots are laid out over a grid drawn or photocopied onto a transparent acetate sheet which can
be placed on a glass slide for observation under the microscope. The numbers of fungus-grid
line contacts per unit root length are determined and the result is most conveniently expressed
as a proportion (usua1ly in percentage terms) of the root length occupied by the fungus.

Table C.I Responsiveness to VA infection of sorne important plants of tropical agro and
forest ecosystems.

Very responsive (i.e. Obligately mycorrhizal in most situations).

Avocado Persea gratissima

Black pepper Piper nigrum
Cassava Manihot escu/enta
Citrus Citrus spp.
Coffee Coffea arabica
Cocoa 1heobroma cacao
Cowpea Vigna sinensis
Guava Psidium guajava
Mango Mangifera indica

Responsive under sorne conditions (i.e. Facultatively rnycorrhizal)

Acacia Acacia spp.

Bamboo Bambusa spp.
Cotton Gossypîum spp.
Gmund nut Arachis hypogea
Litchi Litchi sinensis
Maize Zea mais
Mahogany Khaya grandifolia
Mung bean Vigna radiata
Oil palm Elaia guianensis
Papaya (Paw Paw) Carica papaya
Pearl millett Pennisetum glaucum
Rice (dry land) Oryza sativa
Rubber Hevea brasiliensis
Sorghum Sorghum bicolor
Sugar cane Saccharum officinarum
Tea Camelia sinensis
Terminalia Terminalia spp.

123

l '
'r.rBF: A. Htmdbook 0.1 MerJwdJI

C.2.3 Isolation and identification or VA ruugi

The true status of a fungal isolate can oo1y he assessed by back-inoculation into the host plant
to confirm that structural and functional attributes of the association are reproduced (Koch,
1882). Isolation and back-inoculation is particularly important in mycorrhizal studies because
fungal isolates even within a single category of mycorrhiza differ in their effects upon the
nutritiôn and physiology of the ho!u. Wherever possible the fungal species involved should
he identified.

VA endophytes are normally isolated as spores or sporocarps which are produced in soil,
especially in response to drying. The spores can he separated from the soil by wet seiving
(Gerdemann and Nicolson, 1963; Pacioni, 1992) or by centifugation in 2 M sucrose (Allen
et al., 1979) and can he used as sources of inoculum and for purposes of identification.

In order to ob tain reliable identification of spores iSôlated in this manner it is necessary to

produce pot ~ultures from the single spore isolates. The spôre is pl~ced on a foot of a
potcntiaJ host plant e.g. maitt, ~d the two organisms are grown in a pot of prevÎôusly
sterili7.ed medium which is low in phosphate for a period sufficient for infection development
and spore production (3-6 mônths). Identification of spores can then he attempted using keys
such as that of Schenck and Perez (1990). Confirmation of identity of isolates obtained in
this way may he provided by INV AM: Dr Joseph Morton, Div. of Plant and Soil Sciences,
401 Brooks Hall, West Virginia University, Morganstown. WV 26506-6057, USA.

C.2.4 Inoculation of test plants

Infected roots themselves, if freshly collected from the ecosystems under investigation, cao
be use<! as sources of VA inôeulum. These have the advantaae that they contain the complete
mierobial community of the mot indudinA the naturally occurring YA cndophytcs. The
disadvantage is that the taxonomie identity of endophytes growing in the inoculum is not
known. If comparative experiments involving the growth of mycorrhizal and non-mycorrhizal
plants are to be undertaken using root fragments as inoculum it is normal to add an equivalent
quantity of sterilized mot fragments to non-mycorrhizal treatments ta balance the initial
nutrient inputs. Filtrates of the root fragments containing mieroorganisms. but no VA fungi,
may aIso he added ta ensure that comparable non-mycorrhizal microbial communitics can
develop in the two trea.tments.

C.2.~ Analysis of hœt ~poose

In aIl but the most fertile conditions, where depressions of host yield can occasionaIly be
found following infection by VA fungi, a positive growth response to inoculation of sterile
soil is to be expected. The range of such responses is determined not only by the fertility of
the substrate but also by the extent to which the plant in question is responsive to or
'dependent' upon infection. Bfoadly speaking planb with fibrous root systems and weB
developed root hairs are less responsive than those with a coarse mot system made up of
tuberous foots with few root hain. The largely non-mycorrhizal families liste<! above are
characterised by the former type of root system. Many important plant families, for example
the Gramineae, whiehalso have fibrous foots, may be infected by VA fungi but be not very

124
 AppendiJc C: Myco"hizas

responsive to their presence. These infections are frequently referred to as heing 'facultative' .
In those plants, amongst which are numhered many tropical species of great agricultural
importance, with coarse root system there may he little or no growth in nature in the absence
of infection. The term 'obligate' cao he applied to these infections and the host plants
involved and are often said to he 'dependent' upon infection (Ianos, 1987). Dependence as
a concept is acceptable only if it is borne in mind that almost all autotrophs cao grow
independently of fungal symbionts providing their root systems are provided with an adequate
supply of nutrients. The fact is that only rarely in natural ecosystems or in low external input
agriculture and forestry practices are adequate supplies maintained. There are also occasions
when responses to inoculation with VA fungi are greater than those obtained by application
of superphosphate (Alexander et al., 1992).

Various attempts have been made to quantify the extent of response of a given plant or crop
to inoculation with VA fungi. Gerdemann (1975) defmed responsiveness, or relative
mycorrhizal dependency (RMD), as the extent to which a plant requires mycorrhizal infection
to produce its maximum growth and yield at a given level of soil fertility. This concept was
modified by Plenchette et al., (1983) for use with crop plants under field conditions as
follows. Relative field mycorrhizal dependency (RFMD) is

where
Wm = Dry weight of mycorrhizal plant
Wn = Dry weight of non mycorrhizal plant

A formula for use where mycorrhizal inoculum is introduced to field soils to supplement an
indigenous population (see helow) has been proposed by Bagyaraj et al., 1988. The
mycorrhizal inoculation effect (MIE):

where
W j = Dry weight of inoculated plant
Wu = Dry weight of uninoculated plant

This expression is useful in determining the extent to which introduced fungi compete with
indigenous species to induce a growth response.

If inoculum is to he added to soil in the field, large scale production of fungal propagules is
needed. This cao he achieved in stages. In the first instance starter cultures are produced
in pots. These cao he bulked up by using progressively larger pots. The spores selected for
the original inoculum may he obtained from the local site, from a culture collection or from
a commercial source (e.g. Native Plants Incorp (NPI) 417 Wakara Way, Salt Lake City, UT
84108, USA or IIRA-Agroindustrias, Apartado Âereo 427, Fulua (Valle) Colombia).
Concentrated inoculum already containing, for example, 1000-2000 spores of Glomus
manihotis per ml soil cao also he obtained commercially with a view to application to specific
crops, in this case cassava. Such inoculum cao he applied directly to sterilised seed beds on
the farm to produce further bulking up. Detailed procedures for large scale inoculum

125
TSBF: A. Hantlboot of M.:fhod.r

prOduction are described by Sieverding (1991), Feldmann and Idcuk (1992), and Bagyaraj
(1992). If pure cultures are required it is neœssary to fumigate the beds prior to introduction
of inoculum and seed sowing. Dammet, applicd as Basomid Olt 50 g/m2 is a recommendcd
soil fumigant. Using this, Sieverding (1991) 8Chieved 18,314 spores of Glomus manihotis
pee gmm dry soïl after 6 months.

In addition ta the soU based inoculum, a growth substrate consisting of porous expanded clay
particles bas bren shown by Delme and Baclchaus (1986) to be very effective for producing
vegetative mycelium and spores of VA fungi. A commeecia11y available product Lecadan
(Obtainable from LECA - Germany, D 2083 Halslenbech) is recommended. Its use for
inoculum production in tropical nurseries is described by Feldmann and Idczak (1992). Some
clays are unsuitable because of toxic properties.

Whatever the carrier of inocu1um considerable volumes are requinKl if it Î$ to be applie<t on

a field scale. Sieverding (1991) used 5000 litres of inoculum, eonsisting of roots, myœlium
and spores in soil, per ha of cassava. Even with this lëvel of application, using a highly
responsive plant, and a fungus selected for its effectiveness, benefits May not persist into the
second rotation. Loss of benefit can he attributable to failure of the introduced inoculum w
compete with indigenous endophytes or simply to the fact that the latter are better adapted and
Sc) persist longer. With the exception of severely eraded or disturbed sites, and those that
have been 115ft as open fallow for a number of years most soUs will contain a native inoculum
with attributes that are likely to he well suited to the site (800 DOOd et al., 1983). This being
the case the tirst quesrlon to be aslœd in any situation involving the possibility of inoculation
is not so much which inoculum to select but whether inoculation is îndeed the appropriate
option (see Abbott et al., 1992). The complexity of the inoculum production process and the
unœrtainties of sucuu are such that it may he preferable to consider manipulating the native
înoculum 10 maximise the benefits obtained from it.

ln crop rotation systems, and in intercropping, the use of non-mycorrhizal crop species Juch
as chenopods or crucifers (Harinikumar and Bagyaraj, 1988) will lead to a reduction in
numbers of infective progagules in soil. In contrast, use of a crop sucb as cassava which is
strongly mycorrhiul will in~rease their number, as wiU the use of a grass rather than a
non-mycorrbiul weed in any fallow system. Comparisons of row intercropping with sole
cropping of maiu and bean have shown higher rates of infeçtion and yields of bom plants in
the inrercropping system (Sieverding, 1991).

Pre-sowing cultural practices can also be modified to improve inoculum vigour. Addition of
organic manure (Harinikumar and Bagyaraj, 1989) or straw (Daft, 1992) can increase the size
and effect of VA inoculum in cropped systems, whereas heavy <loses of super-phosphate an<l
tillage will both reduce it. By fragmenting the myœlial network established under the
previous crop, tillage cau be expected 10 reduce its inoculum potential. Evidence from a
tempemt.e agricultuml cropping system involving maize now points çlearly (0 $uch effecrs.
Mycorrhizal infeçtion, phosphorus inflow and growth of the crop were all shown to be
significantly greater in a no-till than in a plou&hed system (McGonigle et al., 1990: Miller
and MçGonigle, 1992).

Within slash and hum agriculture there appears to be little adverse effect on propagule density
of VA fungi (Sieverdîng, 1991).

126
C.3 THE ECTOMYCORRHIZAL SYMBIOSIS

Most ectomycorrhizal fungi are basidiomycetes but there are sorne ascomycetous associations.
This symbiosis is found in the majority of woody species of the temperate and boreal forests
and in sorne of the most important tropical tree species. Members of the Dipterocarpaceae,
which make up ca. 80% of timber exported from South east Asia and sorne 30% of the total
tropical hardwood trade (Smits 1992) appear to be almost entirely ectomycorrhizal, as are the
Fagaceae, the sub-family Leptospermoideae of the Myrtaceae, which inc1udes Eucalyptus, and
sorne genera of the tribes Amhersteae and Detariae in the legume sub-family Caesalpinoideae
(Alexander, 1989a). Non-nodulated ectomycorrhizal trees conspicuously dominate in many
moist savannas inc1uding the genera Brachystegia, lsoberlinia, Julbemardia, Marquesia,
Monoles and Uapaca in Africa (Alexander and Hogberg, 1986; Hogberg, 1989) Selection
favouring ectomycorrhizal species is thought to be driven by nutritional factors which
ultimately determine community composition. Both in Amazonian (Singer and Araujo, 1979;
1986) and West African (Newbery et al., 1988) forests, communities dominated by
ectomycorrhizal species are characteristically restricted to extremely nutrient-poor soils with
surface accumulation of litter. There is increasing evidence that this symbiosis is involved
in the mobilisation of nitrogen as weIl as phosphorus from these substrates (Read, 1991).

In contrast to the V A fungi, the ectomycorrhizal fungi in tropical forests, show a greater level
of host-specificity (fhoen and Ba, 1987; Alexander, 1989b; Smit, 1992). This bas
implications for diversity of their host species. Alexander (1989b) points out that in the more
host specific circumstance the transfer of infection from established mycorrhizal mycelium
to germinating seedling while still being effective, would lead to a diametrically opposed
effect to that seen in V A forests of low host specificity. By increasing the survivorship of
a smalI number of compatible species, diversity is actually reduced. There is sorne evidence
from both the Amazonian and African forests that species diversity is lower in ecto than in
adjacent VA forests.

C.3.1 Characterisation and analysis of ectomycorrhizal infection

This mycorrhizal type is characterised by inter-cellular as distinct from intra-cellular

penetration by the myœlium to form a 'Hartig-net' through which nutrient exchange between
fungus and host occurs. There is also a more or less well developed 'sheath' or 'mantle' of
fungal mycelium around the root which may have a pseudoparenchymatous or
plectenchymatous structure.

Ectomycorrhizal lateral roots are often short and swollen and so cao he identified with the
naked eye. Confirmation of infection requires that a longitudinal or transverse section be
taken to revea1 the Hartig-net and the ensheathing layer or mantle of mycelium, which can
he of variable thickness around the whole mot.

C.3.2 Quantification of occurrence

Methods for sampling ectomycorrhizal fine roots and of quantifying their occurrence are
described by Fairley and Alexander (1985).

127
TSBF: A Handbook 01 M~tIwds

C.3.3 kolatioD and identification of e<:tomycorrhiDI fungi

Sporocarps provide identifiable sources of ectomycorrhizal fungi, and many species are
readily grown from vegetative parts of yOWlg sporocatps. Spores provide less reliable
sources of inoculum, often t'Cquiriog roots or root exudates to stimulate germination. Isolates
can aIso he obtained by surface sterilisation of root tips using 30% H202 for 2 minutes
(faylor and Alexander, 1991), but it is difficult to ascribe such isolates ta particular species.
Detailed ana1y~ enabling tentative identification of the causal fungus require the srudy of
mot squames or scrapings of the fungal mande which can reveal diagnostic features of
mycelial structure (Agerer, 1987; 1990; 1991; Ingleby et al., 1990).

C.3.4 Inoculation of test plants

Detailed methods for synthesis of ectomycorrhiza under laboratory conditions are provided
by Peterson and Chakravarty (1991).

Field inoculation bas often been attempted and panieularly striking responses have been
a.chieved in the rrapies when alien spooics such as pine and euca1yprus have been introduced
ioto areas which lack appropriate ectomycorrhizal fungi. One fungus, above aIl others,
Pisolithus tinctorius, has proved to be useful as an inoculant for use on seedlings to he
transplanted from nurseries ioto tropical soUs. The success of this fungus appears to he
associated with a combined ability to toJerate high temperature and to produce a very large
and vigomus external mycelium. A country by country account of the successes achieved by
inoculation with this fungus together with details of produeûon and inoculation methods is
provided by Marx (1991). Inoculum of P. nnclorius and other ectomycorrhizal fungi is
commercia11y available from Myoorr Tech. [ne. University of Pittsburgh Applied Re5eafch
Centre, Pitt~burgh, Pennsylvania 1~238, USA. P. nnctorius shows relativel}' linie host
specîficity.

References

Abbott, L.K., Robson, A,D. and DeBoer. G. (1984) The eff~t of phosphoru~ on the
formation of hyphae in soil by the vesicular-arbuscular mycorrhizal. fungus Glomus
jasciculatum. New Phytologist 97.437-446.
Abbott, L.K., Robwn, A.D. and Gazey, C. (1992) Selection of inoculant
vesicular-amuscular mycorrhizal fungî. ln: Norris, J. R., Read. D.l. and Varma.
A.K. (eds.), MetJwds in Microbiology 24. Academie Press, London. pp. 1-21.
Agerer. R. (cd.). (1987-1990) Colour atlas of 8ctomyco"hir.M ht-4th Edition.
Einhom-Verlag, Sehwabisch Gmund, Germany.
Agerer, R. (1991) Characterisation of ectomycorrhiza. In: Noms, J.R., Read, D.J. and
Vanna, A.K. (OOs.), Methotb in Microbiology 23. Academie Press, London,
pp.25-73.
Alexander, LJ. (1989a) Systematics and ecology of ectomycorrhizallegumes. In: Stirton,
C.H. and Zarucchi. J.L. (ros.), Advanc8S i1l1eg~ biology. Monogmph of MiSsouri
Botanica1 Gardens, pp. 607-624.

128
 Appendix C: Mycorrhizas

Alexander, I.J (1989b) Mycorrhizas in tropical forest. In: Proctor, J. (ed.), Mineral
Nutrients in Tropical Forest and Sa vanna Ecosystems. Blackwell Scientific
Publishers, Oxford, pp.169-188.
Alexander, I.J. and Hogberg, P. (1986) Ectomycorrhizas in tropical angiosperm trees. New
Phytologist 102, 541-549.
Alexander, I.J., Ahmad, N. and See, L.S. (1992) The role of mycorrhizas in the
regeneration of sorne Malaysian forest trees. Philosophical Transactions of the Royal
Society 335 (in press).
Allen, M.F. (1991) The ecology ofmyco"hizae. Cambridge University Press, London.
Allen, M.F., Moore, T.S. and Christensen, M. (1979) Growth of vesicular-arbuscular
mycorrhizal and non mycorrhizal Bouteloua gracilis in a defined medium. Mycologia
71, 666-669.
Ambler, J.R and Young, J.L. (1977) Techniques for determining root length infected by
vesicular-arbuscular mycorrhizae. SoU Science Society of America Journal 41, 551.
Bagyaraj, D.J. (1992) Vesicular-arbuscular mycorrhiza : Application in Agriculture. In:
Norris, J.R., Read, D.J. and Vanna, A.K. (eds.), Methods in Microbiology 24.
Academie Press, London, pp. 359-373.
Daft, M.J. (1991) Influences of genotypes, rock phosphate and plant densities on
mycorrhizal development and the growth of five different crops. Agriculture,
Ecosystems and Environment 35, 151-169.
Daft, M.J. (1992) Use of VA mycorrhizas in agriculture: Problems and Prospects. In:
Read, D.J., Lewis, D.H., Fitter, A.H. and Alexander, I.A. (eds.), Myco"hizas and
Ecosystems. CAB International, Wallingford, UK. pp.198-201.
Dehne, H-Wand Backhaus, G.F. (1986) The use of vesicular-arbuscu1ar mycorrhizal fungi
in plant production. 1. Inoculum production. Journal of Plant Diseases and
Protection 93, 415-424.
Dodd, J. C., Krikun, 1. and Hass, 1. (1983) Relative effectiveness ofindigenous populations
of vesicular-arbuscular mycorrhizal fungi from four sites in the Negev, Israel.
Journal of BOlany 32, 10-21
Fairley, R.I. and Alexander, I.J. (1985) Methods of calculating fine root production in
forests. In: Fitter, A.H., Atkinson, D., Read, D.J. and Usher, M.B. (eds.),
Ecological Interactions in SoU, Plants, Microbes and Animais. British Ecological
Society Special Publication 4. Blackwell, Oxford, pp.37-42.
Feldmann, F. and Idczak, E. (1992) Inoculum production of vesicular-arbuscular
mycorrhizal fungi for use in tropical nurseries. In: Norris, J.R., Read, D.J. and
Varma, A.K. (eds.), Methods in Microbiology 24. Academie Press, London,
pp. 339-357.
Gerdemann, J.W. (1975) Vesicular-arbuscular mycorrhizae. In: Torrey. J.G. and Clarkson,
D.T. (eds.), Tht Developmenl a,u[ Function of Roots. Academic Press, New York,
pp.575-591.
Gerdemann, J.W. and Nicolson, T.H. (1963) Spores of mycorrhizal Endogone species
extracted from soil by wet sieving and decanting. Transactions of the British
Mycological Society 46, 235-244.
Grace, C. and Stribley, D.P. (1991) A safer procedure for routine staining of
vesicular-arbuscular mycorrhizal fungi. Mycological Research 95, 1160-1162.
Harinikumar, K.M. and BagYaIaj, D.J. (1988) The effect of crop rotation on native
vesicular-arbuscular mycorrhizal propagules in soil. Plant and Soi/HO, 77-80.

129
TSBF.. Ii HanJbooA: 0/ McrhodJ

Harinikumar, K.M. and Bagyaraj, D.J. (1989) Effects of cropping sequence, fertilisers and
farmyard manure on vesicular-arbuscular myeorrhizal fungi in different crops over
three consecutive seasons. Biology and Fenility of Soirs 7, 173-175.
Harley, J.L. and Smith~ S.E. (1983) Mycorrhiml symbiosis. Academie press, London.
Hogberg, P. (1989) Root symbioses of trees in savannas. In: Proctor, J. (ed.), M;~ral
Nutrients in Tropical Forest and Savonna Ecosystems. Blackwell Seientifie
Publications, Oxford, pp.121-137.
Ingleby, K., Masan, P.A., Last, F.T. and Fleming, L.V. (1990) IdentifiCl1IÏon of
EclomycorrhiZQS. H.M.S.O. London.
Janos, D.P. (1980) Vesicular-arbuscular mycorrhizas affect lowland tropical rain forest plant
growth. Ecology 61, 151-162.
Janos, D.P. (1987) VA Mycorrhizas in humid tropical ecosystems. In; Safrr, O.S. (ed.),
Ecophysiology 01 VA Myco"hizal Plants. CRe Press, Boca Raton, LA, USA,
pp. 107-135.
Koch, R. (1882) Uber die Mllzbrandimpfung, eine Entgegnung auf den von Pasteur in Genf
gehaltenen Vortrag. Fischer, Berlin.
Marx, D.H., Ruehle, J.L. and Cordell, C.E. (1991) Methods for studying nursery and field
response of trees to specifie ectomycorrhiza. In: Norris, J.R., Read, D.J. and Vanna,
A.K. (eds.), Methods in Microbiology 23. Academic Press, London, pp.383-411.
MeGonigle, T.P., Evans, 0.0. and Miller, M.H. (1990) Effect of degree of soU disturbance
on mycorrhizal colonisation and phosphorus absorption by maize in growth chamber
and field experiments. New Phytologist 116, 629-636.
Miller, M.H. and McGonigle, T.P. (1992) Soil disturbance and the effectiveness of
arbuscular myeorrhiza in an agrioultural ecosystem, In: Read, D.J., Lewis, D.H.,
Fitter, A.H. and Alexander, I.J. (eds.), Myco"hiws in EcDsysrems. CAB
International, Wallingford, UK. pp. 156-163.
Morton, J.B. and Benny, G.L. (1990) Revised classification ofarbuscular mycorrhizal fungi
(Zygomycetes): a new order, Glomales, two new suborders t Glomineae and
Oigasporineae and two new families Ac.aulosporaceae and Gigasporaœae with an
emendation of Glomaceae. MycOloxon 37, 471-491.
Newbery, D.M., Alexander I.J., Thomas, D.W, and Gartlan, J.S. (1988) Ectomycorrhizal
rain-forest legumes and soil phosphorus in Korup National park, Cameroon. New
Phytolog;st 109, 433-350.
Newman, E.I., Child, R.D. and Patrick, C.M. (1986) My~rrhiul inf~tion in grasses of
Kenyan savanna. Journal of Ecology 74, 1179-1183.
Pacioni, G. (1992) Wet sieving and decanting techniques for the extraction of spores of
vesicular-arbuscular fungi. ln; Norris, J.R., Read~ D.J. and Varma, A.K. (oos.) ,
Melhods in Microhiology 24. Academie Press, London, pp.317-323.
Peterson, R.L. and Chakravany, P. (1991) Techniques in synthesizing ectomycorrhizas. In:
Norris, J.R., Read, D.J. and Varma, A.K. (e<1s.), MethodJ in Mir:robiology 13.
Academie Press, London, pp.75-106.
Phillips, J.M. and Hayman, D.S. (1970) Improved procedures for cleaning mots and
staining parasitic and vesicular-arbu5Cular mycorrhizal fungi. Transactions vI the
British Myr:ological Society 55, 158-160.
Plenehette, C., Fortin, J.A. and Furlan~ V. (1983) Growth response of severnl plant species
to mycorrhiza in a sail of moderate P- fertility. I. Mycorrhizal dependency under
field conditions. Plant and Soil 70, 199-209.
Read, D.J. (1991) Myeorrhizas in ecosystems. Experientia 47, 379-391.

130
 Appendix C: Mycorrhizas

Read, D.l., Koucheki, H.K. and Hodgson, 1. (1976) Vesicular-arbuscular mycorrhiza in

natural vegetation systems. 1. The occurrence of infection. New Phytologist 77,
641-653.
Read, D.l., Franics, R. and Finlay R.D. (1985) Mycorrhizal mycelia and nutrient cycling
in plant communities. In: Fitter, A.H., Atkinson, D., Read, D.l. and Usher, M.B.
(eds.), Ecological Interactions in Soil: Plants, Microbes and Animais. British
Ecologica1 Society, Special Publication 4. Blackwell, Oxford, pp.193-217.
Schenck, N. C. and Perez, V. (1990) Maruud for the Identification of VA Myco"hizal Fungi.
Synergistic Publications, Gainesville, Florida, USA.
Sieverding, E.. (1991) Vesicular-arbuscular mycorrhiza management in tropical
agroecosystems. Hartmit Bremer Verlag. 3403 Friedland 5, Germany, 371 pp.
Singer, Rand Araujo, 1. (1979) Litter decomposition and ectomycorrhiza in an Amazonian
forest. 1. A comparison of litter decomposing and ectomycorrhizal basidiomycetes
in latosol-terra-firme forest and white podzol campinarana. Acta Amazonica 9, 25-41.
Singer, R. and Araujo, 1. (1986) Litter decomposition and ectomycorrhizal basidiomycetes
in an igapo forest. Plant Systematics and Evolution 153, 107-117.
Smits, W. (1992) Mycorrhizal studies in Dipterocarp forests in Indonesia. In: Read, D.l.,
Lewis, D.H., Fitter, A.H. and Alexander, 1.1. (eds.) , Myco"hizas in Ecosystems.
CAB International, Wallingford, UK. pp.283-292.
Taylor, A.F.S. and Alexander, 1.1. (1991) Ectomycorrhizal synthesis with Tylospora
fibrillosa a member of the Corticiaceae. Mycological Research 95, 381-384.
Tennant, D. (1975) A test of a modified line intersect method of measuring mot length.
Journal of Ecology 63, 995-1001.
Thoen, D. and Ba, A.M. (1987) Observations on the fungi and the ectomycorrhiza of Afzelia
qfricana and Uapaca guineensis in Southem Senegal. In: Sylvia, D.M., Hung, L.L.,
Graham, 1.H. (eds.), Proceedings of the 7th Nonh American Coriference on
Myco"hiza. University of F10rida Press, Gainesville, Florida, USA, p.132.
Trappe, 1.M. (1987) Phylogenetic and ecological aspects of mycotrophy in the angiosperms
from an evolutionary standpoint. In: Safir, G.R. (ed.), Ecophysiology of VA
Myco"hizal plants. CRC Press, Boca Raton, LA, USA, pp.5-25.

131
Appendix D ROOTS: LENGllI, BIOMASS t PRODUCTION AND
MORTALITY
Meine van Noordwijk

D.t INTRODUCTION

ln ooth crops and natural vegetation a substantial proportion of net primary production oceues
be10w ground in the root system. This proportion varies with vegetation type, developmental
stage, soil conditions and cultural practices. Shoot:root ratios on a dry matter basis are
typically between 5: 1 and 10: 1 for annual crops at maximum standing biomass. For perennial
crops and natural vegetation. values vary between 1:2 and 5: 1. Uptalce of water and nutrients
are re1ated to mot length or mot surface area rather than mot weight. Distribution and
periodicity of root length is important 10 evaluate whether or not crop demand for nutrients
and water coincides in space and time with the available supply (synchrony and
synlocalization). Root death and mot exudaœs are a major input of organic matter 10 the soil
system and the extent, timing and location of mot death are therefore important. By leaving
a weB distributed set of continuous channels of mostly easily deoomposable organic matter,
roots have a relevance for soU biota, including roots of subRquent crops, far heyond their
often limiœd quantity of oraanic matter.

For these rcasons root investigations are an important part of the TSBF programme.
Unfonunately they prcRnt particular difficulties; extraction of roots from soi! Ï's time
consuming, labour intensive and still it is often incomplete.

D.2 QUANTIFICATION OF ROûT BIOMASS AND LENGTH

The methods describOO in Section 3.3 of this Handbook May, wîth adeq~te ~ibration, he
used for estimates of root weight and length. More accurate estimates can he obtained from
weIl repliçatcd wre (auger) samples, washed on a fine meshed sicve. To obtain reliab1e
estimates of root biomass severa! points should he noted:

• samples should be taken from representative volumes of soil; in row crops special
sampling schemes (Van Noordwijk el al., 1985) may be needed;
• samples should be taken around the expected maximum standing coot biomass; a late
season sampling may result in a high proportion of dead roots;
• the methods for sampling. storing samplcs and washing will unavoidably lead ta wme
loss of dry weight and nutrients; relevant correction factors may be obtained by
sîmulating all procedures on roots grown in nutrient solution. Losses cao be in the
range 20 - 50% for dry wcight, and thus correction factors of 1.25 - 2.0 should he
applied to the final data (Van Noordwijk, 1987j Grzebisz et al., 1989). Nutrient
contents are genera11y only reliable if sample handling is completed within one day.

132
 Appendix D: Rools: Length, Biomass, Production and Monallty

D.2.t Corer design

The corer removes a known volume of soil from a known depth in the profile, without the
need for digging a soil pit and destroying part of an experimental field. A core of 50 - 80
mm diameter is satisfactory, and the corer can he inserted either manually or mechanically.
Manual coring is difficult at depths greater than 50 cm and in clay or stony soil. In dry
sandy soil a smaller core diameter May he needed to reduce losses of soil when extracting the
core. A suitable hand corer consists of a 15 cm steel tube with a serrated cutting edge
mounted on a 1 m pipe, with a plunger to remove the core (Figure D.1; Bohm, 1979).
Marks on the pipe indicate 10 cm depth increments. Soil is extracted with the corer in
successive 10 cm increments.

plunger ,
+ plunger

40 20
marker

Figure D.I Design of a hand corer

Altemativelya sharpened steel tube of appropriate length can he driven into the soil. In this
case it is important to have a tripod and chain hoist to extract the tube, a file to resharpen the
cutting edge, and longitudinally split liners of thin metal within the tube so that the intact soil
core can he pu lIed out. A powered motor breaker can be used to drive the corer into the soil
(Welbank and Williams, 1968), but compaction of the core will occur and should be a1lowed
for when cutting subsamples from the core.

In very stony soil, or where there are many woody tree roots, coring may not be possible.
Regular, known volumes of soil (monoliths) can be taken from the face of a pit and treated
in the same way as cores.

D.2.2 Sample depth

Ideally the profile should he sampled to the limits of rooting depth. At depth, however,
rooting intensity is low and spatial variability high. Based on initial profile wall observations
a meaningful lower limit can be set. In sorne cases a linear relationship of the log of root
mass versus depth (a negative exponential root distribution) may help to extrapolate root
densities in the soil beyond sampling depth. AIl soils must be sampled to a minimum depth
of 30 cm.

Where horizon development occurs, subdivide the cores at horizon boundaries of known
depth in the first instance and within horizons in 10 cm increments. In cultivated soil or

133
where there are no clear horizons subdivide the mineraI soil in 10 cm increments. Surface
organic horizons should be treated separately.

D.2.3 Sampllng ioteDSity

Even in the most homogeneous soits a considerable spatial variability of root density will
oceur. For auger samples of about 385 cm' (10 cm height, 7 cm diameter) a coefficient of
variation in root weight of at least 40% may be expected (Van Noordwijk et al., 1985a); on
heterogeneous soils the coefficient of variation may be much higher. This variability implies
that large number of replicate samples are needed if precise estimates of mot weight are
needed.

It is advisable to obtain reliable information at one or two, weil chosen situations, rather than
non-reliable data on many. If 25 replicates would be analysed the standard error of the mean
would be five times smaller than the standard deviation of individual samples, and thus the
95% confidence interval of the mean would he plus or minus 20%, even in the most
homogeneous soil. Within each rep1îcate plot of a tIQtment Lake no fewer than 3 cores.
Within each replicate the samples at each depth increment cao he JXlOled for funher tratment.
In natural vegetation where no sample stratification strategy is obvious, take the cores on
random coordinates. Where patterns are likely to oceur (e.g. row crops, alley cropping) the
tirst stratification should be on an area buis (within rows vs. between rows), then sample
mndomly within the strata. Beware of high rrot mass directly undemeath the plants. Root
data are seldom normally distributed and an appropriate transformation «n+ l)o.~ or log n)
is often required before assessing the significance of differences between treatments or
sampling intervals.

D.2.4 Root extraction

The best approach is to wash roots from the cores immcdiately upon retum from the field.
Core samples can be stored in sealed polythylene bags in a refrigerator for a few days or deep
frceze until processed. If deep free!C faciUties are not availablc. samples can be stored air-
dried and re-wetted before wB.!lhing. l..ôsses of dry weight due to the methods used for
storage should be checked.

Soil texture. srructure, degree of compaction and organic matter content gready influence the
precision and time required to extract roots from cores. The simplest ltlr;thod involve.s ~
washing a presoaked sample over a large diameœr 0.3 - 0.5 mm mesh sieve. The work cao
be simplified by washing over a combination of sieves: one with 1.1 and one with 0.3 mm
mesh. The tint sieve will contain mostly roots. the second mostly debris. The material
removed trom the sieve(s) cao then he mixed in water and the suspended material decanted
(live roots have a specifie density of about 1.0 g/cm3). This residue is then hand sorted in
shallow dishes under wateT to remove fragments of organic .matter and dead roots; normally
it is better to pick live roots from the sample and leave debris behind in the dish. Good light
conditions, a calliper with 0.1 mm accuracy and a pair of (watchmaker) forceps are
necessary; a stereo dissecting microscope (from ~ 4 ta ~ 20) may be helpfuL .Operator
fatigue is a problem and it may take about 6 hours to S()rt one core (10 1 m depth) from maize
on a sandy loam, a similar core from forest with a high organic matter content or just one
sample of the top layer of permanent grassland.

134
 Appendix D: Roots: Length. Biomass. Production and Mortailly

A number of root washing machines have been designed (Bohm, 1979). The most successful
employ the process of elutriation, i.e. washing mots and organic debris free of soil and
separating them by flotation, onto a 0.5 mm mesh sieve leaving behind the heavier mineral
particles. The apparatus of Cahoon and Morton (1981) requires adequate water pressure and
uses a lot of water; that of Smucker et al. (1982), now commercially available, uses less
water and accepts smaller samples but requires both water pressure and compressed air.
After elutriation roots must still be sorted by band and this may take several hours. Neither
apparatus bandles organic soils. With both band and machine washing, IOS8 of fine roots
occurs and a periodic check of washing water and residues should he made to quantify such
losses; such losses should be kept less than 5% of root weight (which may imply that they
are still over 10% of root length); this loss is distinct from that due to respiration and loss
of cell contents from remaining root tissue.

Presoaking ovemight in 5 % sodium hexametaphosphate expedites the process of washing

roots from clay soils, but the chemical discolours the roots, particularly in soils with high
organic matter content and may disrupt the tissue, making subsequent identification of live
roots more difficult. Such pretreatment will also interfere with chemical analyses and should
therefore be avoided if possible. Any lengthy washing procedure may alter the element
content of root tissue and a subsample band sorted with a minimum of water and processed
on the day of sampling must be used for analysis.

D.2.S Classifying the roots

Fine roots are the most important part of the root system for water and nutrient uptake, as
they form the largest part of total root length or root surface area. For woody perennial
vegetation there is a fairly obvious distinction between the more or less pennanent,
secondarily thickened mots and the ephemeral, unthickened roots. This functional distinction
usually falls somewhere between 1 and 3 mm root diameter. Roots above 10 mm diameter
are not adequately sampled by coring. For herbaceous perennial and short lived vegetation
no such clear distinction exists. For TSBF studies these roots should he separated into < 2
mm and > 2 mm classes. In mixed vegetation, separation of roots of different spccies may
be difficult and is not necessary for TSBF research.

It is desirable to separate root samples into living and dead categories. This îs particularly
important in crop situations where old, dead roots from the previous crops may still be
present. Living roots cao be distinguished by their lightet cotour, turgid appearance and
flexible rather than friable nature when manipulated. Sorne preIirninary anatornical
investigation may help in establishing criteria for making decisions. Incubation of excised
fOOts in soil in modified titter bags can he used to establish visual clues to the mot decay
process. Cross checldng between operators and working block by block instead of treatment
by treatment help to reduce experimental error.

Congo red staining has been used to differentiate wheat roots with intact epidermis from ones
without (Ward et al., 1978). Stain in 1% aqueous solution of Congo Red for 3 min, rinse,
blot dry, then saturate for 3 min in 98 % ethanol before a final rinse. Living roots stain dark
pink to bright red. The criteria for making the living/dead distinction must be clearly stated.
Where adequate criteria cannot he developed, assess total root rnass only.

135

. .1
TSBF: A Handbook of Melhods

D.2.6 Assessm:ent of root mass

Washed root samples can be stored in sealed polyethylene bags for a short time in a
refrigerator. but deep-freeze storage is preferable. Thymol can he added as a bactericide, but
should he hand1ed carefully; classica1storage media (ethanoI, etc.) tend 10 make roots brittle.

Carry out biomass estimation on each size class of aIl samples. Oven-dry the roots and
weigh. Next the dried samples should be combusted for 5 br in a muffle fumace at 550°C
and the residue weighed. Results shou1d be expressed as ash-free oven-dry mass per unit
volume of soil.

D.2.7 Root length measurements

Root length is a relevant parameter for water and nutrient uptake. The specific root length
Oength per unit dry weight) of roots depends on diameter, variability of diameters, dry matter
content (per unit frcsh weight) and air-tilled porosity (per writ volume) (van Noordwijk,
1987). Within a species or situation variability of the specifie root length may be rather small
(e.g. with a coefficient of variation of 10-15% while root weight per unit soi! volume bas a
coefficient of variation of about 40%), 50 it is reasonable to measure the specifie root length
only for sorne subsamples. Normal values are from about 10 m/g for fine tree roots, 50-200
m/g for fine mots of dicotyledonous crops and 200-600 m/g for cereals and grasses (Van
Noordwijk and Brouwer, 1991).

Root Iength cao he estimated by counting the number of intersections between roots and
samp1e lines. This method is based on Buffon's nec:dle problem, described in 1777. where
the chance that il needle randomly thrown on a tiled floor would inrersect one or more of the
e<lges of the tiles was formulated as a function of the length of the needle and the size of the
tile. Application of the method for measuring root length is based on Newman (1966) and
Tennant (1975). Roots are spread out with random orientation in a thin layer of water on a
glass plate (about 25 x 25 cm), water is removed and a grid (photocopied on an acetate folio
for overhead projection) is put over or undemeath the sample. Line by Une all intersections
of mots and grid lines (taking the upper or left boundary of the line as criterion in case of
doubt) are added (Figure D.2). Results for horizontal (H) and vertical (V) grid lines are
added to the number N. If the grid si" is D (mm). root 1ength L (mm) is derived as:

L='l'CND/4

If Dis set equal to 40/lt, i.e. 12.7 mm, L = 10 N (mm). By adding results for H and V
lines the method is in sensitive to preferential oreintation of roots on the plate, but spreading
the rootS must be done without regard of the po5ition of the grid lines. To împrove contrast
roots can be staîned beforehand, e.g. with saffranin red (1 g/litre).

To optimise working efficiency at an acceptable random error level the grid size should be
chosen to obtain about 400 intercepts (200 H + 200 V) per sample. This method CM be
applied manually (using a magnifying lens for fine roots) or can be automated in various ways
(Rowse and Phillips, 1974; Richards et al.• 1979; Wilhelm el al., 1982). Each variant of the
method should be CaIibrated by cutting a known length of cloth to small pieces. Commercial

136
 Appendix D: Roots: Length, Biomass, Production and Morta/lty

equipment is available (Comair Root Length Scanner, Commonwealth Aircraft Corporation

Ltd, 304 Lorimer St, Port Melbourne, Victoria 32307, Australia). Recently computer image
analysis methods are used as weIl, based on a video camera or line scanner. Only small
fields of view can be analysed with sufficient resolution and root samples should be spread
out more carefully than when human eyes are used as image analyser, so the gain may be less
than expected. For TSBF purposes manual versions of the method are recommended.

H•
h 2
V / / +
/ "- 3
"\,--
--"" ~ +

[--..:(
li / 4
+
~ /
2
+
V b 4
/ 1'--
f ~ ) +
5
/
r \ V-<"VI
.......,
+
4
'\ ~ +
~ .,- V1 4
L ~
) " 1'\ l"-
r-r- +
2
v
L
1/
"'- +
3
1
1'\ \ ..1
+
3
1 r---'

36
V· 2 + 6 + 5 + 7 + 5 + 3 + 5 + 4 • 37

Figure D.2 Line intercept method for determining root length by counting the number of
interceptions between roots and horizontal (H) plus vertical (V) lines of a grid.

Measurements of the frequency distribution of root diameters can be made on the sample
spread out on a grid. by measuring on every xth interception, using a binocular microscope
wîth an <>cular micrometer. At least 20 readings per sample are requirnd (assuming three
samples per treatment per layer).

D.2.S Mycorrhizas

It is desirable to have sorne measure of the type and intensity of mycorrhizal infection in each
treatment. At peak root biomass take approximately 60 cm of fresh clean root at random
from each depth and stain to visualize fungal hyphae. The method of Phillips and Hayman
(1970) can he followed with small adaptations: short (ca. 2 cm) lengths of roots (stored in
alcohol-acetic acid mixture after washing) are heated for about 1 hr at 90°C in 10% KOH (or
left for 2-3 days at ambient temperature), washed in water, if necessary (dark tissue) bleached
in alkaline peroxide (3% NH 40H in 3% H20 2 , prepare daily) , washed in water, acidified for
3 min in 1% HCI, stained for 5 min at ambient tempearture in 0.05% trypan blue (Merck Art

137
T'SBF: A Handbool q/ IJ.dwdJ

No. 11132) in lactic acid and destained in clear lactic acid. In the original description
lactophenol (1:1:1:1 mixture of phenol, lactic acid, glycerol and water) was used instead of
lactic acid. As phenol is toxic it should not be used, unless staining results are unsatisfactory.
A modified procedure wu described by Konnanik and McCraw (1982). For roots with a
high lignin content the KOH treatment may have to he intensified. After smining the samples
are inspectc:d under a (dissecting) microscope. Vesicular arbuscular myrorrhiza (VAM)
infection is characterized by the formation of unseptated hyphae outside the mot and inter-
and/or intracelluw hyphae in the cortical cell layers of the mot (Sieverding, 1991). The
percentage of root sections which bas mycorrhizal structures is assessed. Ectomycorrhizas
are characterized by a fungal sheath and/or Hartig net; the percentage of mot tips with such
structures (which can often be recognized macroscopica.1ly and without staining) is assessed.

See a1so Appendix C "Mycorrhizas".

D.2.9 Sampling frequency

The amount of living and dead mot in the ecosystem fluctuates as a balance of new mot
growth and root death and deçay. Given the large number of replicate samples required to
obtain a reasonable Oet alone accurate) estimate of mot density at a given moment, a regular
sampling programme to monitor changes during the growing season easily beromes
unmanageable. For TSBF site characterization a minimum of two samples per year is
required.

For perennial vegetation sorne prior knowledge of the phenology of the root system will a110w
sampling to coincide with the likely peaks and tmughs in root biomass ; mot growth often
a1temates with periods in which reproductive growth is a major sink for carbohydrates.
Where no prior knowledge e}\ists samples shOlild coincide in herbaceous vegetation with
maximum and minimum above-ground biomass, and in forest with the rnonths of minimum
and maximum minfall, For annual crops maximum root development often coïncides with
flowering (transition frorn vegetative to generative stage).

Samplîng at harvest lime may lead to a considerable amount of dead roots. Sampling before
pb.nting a new crop b relevant for estimating the rate of decay of dead roots from previous
creps.

D.3 ESTIMA TION OF TOTAL ROOT PRODUCTION

To estimate total mot production the biomass estimates obtained (corrected for sampling
schemes and dry matler losses) should be corrected for root turnover between sampling dates.
Due to the large variability of results from destructive sampling it is not possible. in practice,
to obtain turnover estimates from frequent sampling schemes.

The frequent sampling method derives from the following scheme with two pools (live and
dead coots) and three rates (production, mortality and disappeau.nce);

Production (1\) --:> Live mot mass (LJ --:> Mortality (Rm) -- > Dead root mass (DJ -- >
Disappearance (Rd) .

138
 Appendix D: ROOlS: Length, Biomoss. ProdMction and Mortallry

The following equations hold:

~+l = ~ + f ~ - J~

If live and dead roots cannot be separated:

Thus:

f ~ ~ (L+D)t+l - (L+D)t

Sampling errors in determining ~ and Dt play a dominant role, however. In part of the
literature estimates of root production are based on statistically significant increments of root
masSe Statistica1 significance, however, not only depends on the size of the difference but
also on the sampling intensity. Although this method has been used in past decades, and the
literature contains many estimates based on this method, a methodological study by Singh et
al. (1984) showed that considerable ~ and ~ estimates can occur. For a full discussion
of the biologica1 and statistical aspects of the topic, including frequency of sampling in a
range of systems, consult McClaugherty et al. (1982), Fairley and Alexander (1985), Goltz
et al. (1984), Lauenroth et al. (1986), Hansson and Andren (1986), Singh et al. (1984) and
Vogt et al. (1986). More reliable methods for estimating root turnover are based on separate
study of root growth and root decay processes.

D.3.1 Root decay

Root decay (both on a root length and a root weight basis) cao be studied in a modified
litterbag method, incubating known amounts of excised roots (e.g. collected by sieving soil
at harvest time) in a ceramic pot filled with sieved, root-free soil and placing the pots in the
field (possibly at two depths), A sereen coyer on top of the pot may prevent soil fauna from
removing root tissue. At regular intervals pots are retrieved and washed on a fine mesh
sieve. Intact roots and root debris are collected separately.

D.3.2 In-growtb cores

Seasonal patterns of root growth can be studied by the in-growth core method (Steen, 1984;
Fabiao el al., 1985). Soil fmm the depth at which the in-growth core will be placed is
collecterl from the site, air-dried and roots removed by sieving. In-growth bags may he made
from plastic saclring of a minimum of 4 mm mesh, or from more rigid polypropylene
cylinders (4 ta 10 cm diameter, sea1ed at the bottom and top with mesh). For each bag an
amount of soil is weighed according to the required bulk density of the sail. Auger holes are
prepared, and the sacks or cylinders are filled (in situ) with soil, compacting cm by cm,
loosening the surface before new soil is added. The right procedure to obtain the required
bulk density cao be found by trial and error. At regular intervals (say 1 month) in-growth

139

-------------------------------------------'-------------
---- ------ -~-------
bags (minimum·3) are retrieved and washed over a fine mesh sieve. Data aIlow periods with
rapid mot growth 10 be identified.

Because of the altered physica1, chemica1 and biological conditions in the sieved and repaclcod
sail; growth of the mots into the bag may not accurately represent growth in the bulk soil,
panicularly on compaçted or clay-rich soi1s. Period! of active root growth in different laye.Q,
however, should he aœumtely reflCèted..

The technique can aIso he use<! to n:cord mot response to localise<! fertiliser application or
othee heterogeneities in the soil (Cuevas and Medina, 1988; Hamah et al., 1991). Fertiliser
application should be made by mixing the soU with a fertiliser solution prior to filling the
bags.

D.3.3 MiDirhbotrons

Simultaneous processes of root groWth and root death or dee.ay CM only be quantified if the
tà.te of individual roots can he followed. The simplest method uses a glass meet against a sail
profile wall, tmcing mots on transpanmt polyethylene sheets. By using different colours of
marking pens both new mot growth and disappearance of mots can be quantified. An
intermediate (but tedious) technique uses glass tubes inserted mto the soU and regularly
inspected (on grid lines in the glass) for mots using a miITOr and a torch. More sophisticated
versions use a fibreoptic system with a camera (Figure 0.3) and flash ligh{ or a video system
(Taylor, 1987).

F"1IUft D.l Minirhizorron system to observe dynamics of root growth and deçay under field
conditions; at each time of observation a series of images with increasing depth is taken; these
images cao be analysed as shown in Figure D.4.

If a series of images has becn obtained analysis CM procced as in Figure D.4: for ea.ch depth
in the: soit images are compared step by st.ep (Tl with ni 1'2 with 1'3 etc.) and the number
of new fOot intersections wîth a grid Îs scored plus the number of intersections which has
disappeared mnre the last observation. By adding up aIl new root intersections the actual mot
intensity on the observation plane cao he expressed as a fraction of the total annual

140
 ApperuJù D: Rootl: Length, BiomaSl, ProâMction and Monality

production. Curve fits (e.g. logistic) of mot growth and root decay per layer cao be
obtained. If at one point in time, by destructive sampling, a reasonable estimate of standing
root mass was obtained, these relative figures cao be used to estimate annual mot production.
The assumption need not be made that root length on the observation surface bas a known
or constant relation to mot length density in the soil. Calibration lines do in fact differ
between soil horizons, crops, soU types etc. (Taylor, 1987). The main assumption needed
for estimates of total mot production are that the relative pattern of mot growth and decay
on the observation surface represents that in the soil. When glass or perspex (rigid) walls are
used for the observation structure (mini-rhizotron), gaps between this surface and the soU may
he unavoidable and mots grow and die in a "gap" environment. Gijsman et al. (1991)
described an inflatable minirhizotron system which reduces the problem of gap formation.

Tl T2 T3

~ -- I~
'" D ~I
,
lIt-

( D (
.\
"'-
1--
1

""\' 1
"-
\
1

'"~ 1
t--
~
"\
--

Current 9 18
New 12 3
Dlsappeared 3 6

Figure D.4 Analysis of sequential images to derive root turnover.

D.4 FSTIMATE OF TOTAL CARBON INPUT TO mE SOIL

son include sloughed root cap cells.

In addition to structural mot tissue, carbon inputs to the
mucilage produced at the mot tip, decayed root haïrs and other cellular material, soluble
carbohydrates, amîno acîds and other exudates, and C02. from fOot respiration. These root-
rhizosphere transitions form a continuum, the study of which usually is based on 14C labelling
and sophisticated laboratory equipments.

References

Bohm, W. (1979) Methods of Studying Root Systems. Springer-Verlag, Berlin.

Bohm, W. and Kopke, U. (1977) Comparative root investigations with two profile wall
methods. Uitschrift fur Acker-Pjlanzenbau 144, 297-303.

141
TSBF: Â Handbook of Merhods

Cahoon, G.A. and Morton, E.S. (1981) An apparahls for the quantitative sepamtion of
plant rooUi from soil. p,.oceedingJ of litt A.meric;,m Sociery fO,. Hon;cullural Science
78, 593-596.
Caldwell, M.M., and Virginia. R.A. (1991) Root systems. In: Pean:y. R.W, Ehleringer, J.,
Mooney, H.A. and Rundel, R. W. (cds.), Plun/ Physiological Ecology: Field Methods
and Instrumentation. Chapman and Hall, London, pp.367-398.
Cuevas, E. and Medina, E. (1988) Nutrient dynamics within Amazonian forests: Part II.
Fine root growth, nutrient availability and litter decomposition. Oecologia 76,
222-235.
Fabiao, A., Persson, H.A., and Steen, E. (1985) Orowth dynamics of superficial root~ in
Portuguese plantations of EuÇQlyptUof glohulus l.a.bilL studied with a mesh bag
technique. Plunt and SoUS3, 2~~-242.
Fiilirley, R.I. and Alexander, I.J. (1985) Methods of calculating fine root production in
forests. In: Pitter, A.H. (00.), EcologicalllZle,.actioru in Soil. BlackweU Scientitic,
Oxford, pp.37-42.
Floris, J. and Noordwijk, M. van (1984) Improved methods for the extraction of soil
samples for mot researeh. PIani and Soil7', 369-372.
Gholz, H.L., Hendry, L.C. and Cropper, W.P. (1986) Organic matter dynamics of fine
roots in plantations of slash pine (Pi1Ul.f elliottil) in north Florida. Canmlian Journal
of Forestry Research 16, 529-538.
Gijsman, A.J., Floris, J. Noordwijk, M. van and Brouwer, G. (1991) An inflatable
minirhizotron system for root observations with improved sail/tube contact. Plant and
Soi/I34, 261-269.
Grubi$z W., Floris, J. and Noordwijk, M. van (1989) Loss of dry matter and œU contents
from fibrous roots of sugar beet due to sampling, storage and washing. Plant and Soil
113,53-57
Hairiah K., Noordwijk, M. van and Setijono, S. (1991) Tolerance to acid soil conditions
of the velvet beans Mucuna pruriens var. utilis and M. deeringiana. 1. Root
development. Plant and Soil134, 95-105.
Hansson, A.C. and Andren, O. (1986) Below-ground plant production in a perennial grau
ley (Festuca p,.arensis Huds) asscssed with different melhods. Journal of Applied
Ecology 23, 657-666.
Hansson, A.C. and Steen, E. (1984) Melhods of calculating root production and nitrogen
uptake in an annual crop. Swedish Jou11IlJ1 of Agriculture Rtstarch 14, 191-200.
Kormanik, P.P. and McGraw, A.C. (1982) Quantification of vesicular-arbuscular
mycorrhizae in plant mots. In: Schenck, N.C. (ed.), MeûlOds and Princip/es of
Mycorrhiml Research. American Phytopathological Society, St Paul, Minnesota,
pp.37-45.
lauenroth, W.K., Hunt, H.W., Swift, D.M. and Singh, J.S. (1986) Reply to Vogt et al.
Ec%gy 67, 580-582.
Maclde-Dawson, L.A. and Atkinson, D. (1991) Methodology for the study of roots in field
experiments and the interpreration of tesults. In: Atkinson, D. (ed.), Plcmt Root
Gl'Owth, ,an Ecological Perspective. BlackweU Scientific Publications, London, p.25-
47.-
MeClaugherty, C.A., Aber, J.O. and MeHUo, J.M. (1982) The role of fine mots in the
organic matter and nitrogen budgets of two forested ecosystems. Ecology 63,
1481-1491.

142
 Appendix D: ROOfS: Length, BionltUs, Prodst.ctiOll and Mortalit.y

Newman, E.I. (1966) A method for estimating the totaIlength of root in a sample. Joul7Ul1
of Applied Ecology 3, 139-145.
Perrson, H. (1978) Root dynamics in a young Scots Pine stand in central Sweden. Oikos 30.
508-519.
Phillips. J.M. and Hayman, D.S. (1970) Improved procedures for clearing and staining
parasitic and vesicular arbuscular mycorrhizal fungi for rapid assessment of infection.
Transactions of the British Mycological Society 55, 158-161.
Richards, D., Goubran, F.H., Garwoli, W.N. and Daly, M.W. (1979) A machine for
determining root length. Plant and Soil 32, 69-76.
Rowse, H.R. and Phillips, D.A. (1974) An instrument for estimating the totaIlength of root
in a sample. Jounud of Applied Ecology 11, 309-314.
Schuurman, J.J. and Goedewaagen, M.A.J. (1971) Methodsfor the Examination of Root
Systems and Roots. Second edition, Pudoc, Wageningen, the NetherJands, 86 pp.
Sieverding, E; (1991) Vesicular-arbuscular myco"hiza management. GTZ publ. No. 224.
Eschbom. ISBN 3-88085-462-9.
Singh, J.S., Lauenroth, W.K., Hunt, H.W. and Swift, D.M. (1984) Bias and random errors
in estimators of net root production: a simulation approach. Ecology 65, 1760-1764.
Srnucker, A.J.M., McBumey, S.L. and Srivastava, A.K. (1982) Quantitative separation of
roots from cornpacted soil profiles by the hydropneurnatic elutriation system.
Agronomy Jounud 74, 500-503.
Steen, E. (1984) Variation of root growth in a grass ley studied with a rnesh bag technique.
Swedish Joul7Ull of Agricultural Research 14, 93-97.
Taylor, H.M. (ed.). (1987) Minirhizotron Observation Tubes: Methods andApplicationsfor
Measuring Rhiwsphere Dynomics. American Sooiety of Agronomy Special
Publication 50, ASA, Madison, USA, 143 pp.
Taylor, H.M., Upchurch, D.R., Brown, J.M. and Rogers, H.H. (1991) Sorne methods of
root investigations. In: McMichael, B.L. and Persson, H. (eds.), Plant Roots and
their Environment. Elsevier Science Publishers, Amsterdam, p.553-564.
Tennant, D. (1975) A test of a modified line intersect method of estimating root length.
Journal of Ecology 63, 995-1001.
van Noordwijk, M. (1987) Methods for quantification of root distribution pattern and root
dynamics in the field. Proceedings ofthe 20th Colloquium ofthe International Potash
Insntute, 243-262, Baden bie Wien.
van Noordwijk" M. and Brouwer, G. (1991) Review of quantitative root length data in
aariculture. In: Penson, H. and McMichael, B.L. (OOs.) , Plant Rools and lheir
Enllironment. Elsevier Science Publishers, Amsterdam, pp.515-525.
van Noordwijk, M. and Willigen, P de. (1991) Root functions in agricultural systems. In:
Persson, H. and McMichael, B.L. (eds.), Plant Roots and thelr Environment.
Elsevier Science Publishers, Amsterdam, pp.381-395.
van Noordwijk, M., Floris, J. and de Jager, A. (1985a) Sampling schemes for estimating
root density in c::,opped fields. Netherlands Journal of Agricultural Science 33,
241-262.
van Noordwijk, M., de Jager. A. and Floris, J. (1985b) A new dimension to observations
in minirhizotrons: a stereoscopie view on root photographs. Plant and Soil 86,
447-453.
van Noordwijk, M., Widianto, Heinen, M. and Hairiah, K. (1991a) Old tree root channe1s
in acid soils in the humid tropics: important for crop root penetration, water
infiltration and nitrogen management. Plant and Soill34, 37-44.

143
TSBF: Â Handhook 01 M~tIwds

van Noordwijk, M., Kumiatun Hairiah, Syekhfani, M.S. and Flach, B. (1991) Peltophorum
pterocarpa a tree with a root distribution suitable for alley eropping. In: Persson, H.
and MeMiehael, B.L. (eds.), Plant Roots and their Environment. Elsevier Science
Publishers, Amsterdam, pp.526-532.
van Noordwijk, M., Brouwer, G. and Harmanny, K. (1992) Concepts and methods for
studying interactions of mots and sail structure. Geoderma (in press).
Vogt, K.A., Grier, G.e., Gower, S.T., Sprugel, 0.0. and Vogt. D.J. (1986)
Overestimation of net mot production: a rea1 or imaginary problem? Ecology 67,
577-579.
Ward, K.J., Klepper, B., Piekman, R.W. and Allmaras, R.R. (1978) Quantitative estimation
of living wheat root lengths in soil cores. Agronomy Joumal70. 675-677.
Welbank, P.J. and Williams, E.D. (1968) Root gmwth of a barley crop estimated by
sampling with portable powered soil-coring equipment. Jou17Illl of Applied Ecology
5, 477-481.
Wilhelm, W.W., Norman, J.M. and Newell, R.L. (1982) Sen1iautomated ,.-y-plotter-based
method of mea.suring mot length. Agronomy Journal 74, 149-152.

144
Appendix E SOIL SOLUTION SAMPLING AND LYSIMETRY
Mike Homung

Sampling systems are discussed under three main headings:

ISOLATED SOIL MASSES

TENSION/V ACUUMISUCTION SAMPLERS
TENSIONLESS COLLECTORS

Although the development of each method is outlined and the range of designs and
construction materials described, specific recommendations are not made for TSBF projects
since the choice of design often depends on site-specific constraints.

E.l ISOLATED SOIL MASSES

Sampling systems of this type collect drainage water from a mass of .soil enclosed in a
container either in the laboratory or in the field. For field studies the drainage water is
collected in suitable reservoirs situated in an adjacent trench. The systems can be divided into
those in which the container is tilled with disturbed, excavated soi! and those which use
undisturbed soH masses. The fillOO, or repackOO, containers range from rectangular brick or
concrete boxes, or tanks with surface areas of several squares metres down to repacked,
circular cross-section metal or plastic tubes with diameters of a few centimetres. The
containers should be constructed from inert materials to avoid solution--œntainer interactions.
The large in situ isolated block approach can only be used where there is a naturally
impermeable subsoil.

Undisturbed masses range from a 10 m x 10 m in situ area isolated by vertical plastic sheets
(Calder, 1976), to circular cross-section cores ca. 1 m in diameter in fibreglass (Belford,
1979) or metal (Roose and des Tureaux, 1970) containers, to cores in plastic pipes a few
centimetres in diameter. The main constraint on the technique is the expense and/or technica1
difficulties of sampling undisturbed soil masses. The use of disturbed soil allows better
replication of units than undisturbed soil-but has the disadvantage that it is difficult to recreate
the natural soil packing and structure. The larger types cao accommodate a number of plants,
which cao be cropped, to provide a nutrient and water sink. Alternatively, Anderson et al.
(1987) have used 0.5 m x 0.5 m x 0.3 m deep plastic trays filled with forest floor material
with lengths of attached tree roots introduced to the lysimeters through ports in the sides of
the trays.

The main advantages of using isolated soil masses are that a known volume and weight of soil
is used, all drainage cao he easily collected and, thus accurate element fluxes and budgets cao
be calculated. Disadvantages include the fact that only vertical drainage is possible (except

145
TSBF: Â H~/( Df Mt:11tbtM

with the large in 5itu masses), even when latera! drainage is an important factor on the
undisturbed site and that roots from the soil are also excluded from the container; this is
particularly imponant with the smaller columns.

E.l TENSION/VACUUMISUCTION SAMPLERS

E.2.1 Introduction

The extraction of soil solutions by applying a suction 10 a porous collector p1aced in the soi!
was first proposed by Briggs and McCall (1904). They used a Pasteur-Chambertin filter tube
connected to an evacuated two litre bottle. Since the cxploratory study a widc variety of
pomus materiab have been used in a number of designs (Table E.l) to produce a spectrum
of sampling systems from the simple to the complex.

Table KI The main types of pomus $uçtion samplers

Ceramic cups (Wagner, 1962)

Ceramic tubes (Krone el al., 1952; Czeratski, 1971)
Ceramic plates (Czeratski, 1959)
Alundum plates (Cole, 1958)
Acrylic plates (Bourgeois and Luvkulich, 1972)
Plastic cups (Quin and Forsythe, 1976)
Teflon rings (Morrioon, 1982)
Teflon cups (Zirnmerman el ol.. 1978)
Sintered nickel cups (Hadrick et al., 1977)
Cellulose-acetate hollow fibres (1ackson et al.• 1976)
Fritted glass tubes (Long, 1978)
Non~llulosic hollow fibre tubing (Levin and 1ackson, 1977)
Fritted glass plates (Chow, 1977)
Ceramic candies (Duke and Baise, 1973)

AlI the various systems, however, opemte on the basiç principles explored by Briggs and
MeCall (1904) - the pomus sampler is installed in the ground, a vacuum is developM within
the samp1er and pore water is drawn into the sampler through the porous section when soil
water suction is less than the applied vacuum.

E.2.2 Cup and rlng-based sampling systems

The most widely used ~mction satnplers have been ceramîc cups. They are availab1e in a
variety of shapes and sizes and with a range of air-enrry values. The most commonly ùsed
are ca. 6.0 cm in length, ca. 5.0 cm outside diameter and with an air-entry value of 1 bar.
The precise composition of the ceramic material, and the pore sizes, vary from one
manufacturer to another. The 1 bar air-entry value cups manufactured by Soilmoisture
Equipment Corp. have approximately 2.9 JJ.m pores. Small cups, or tubes, e.g. 6 mm x 65

146
 AppendU E: Soil Solution Sampling and Lysimetry

mm, have been used to extract soil solutions in pot or glasshouse studies (e.g. Harris and
Hansen, 1975).

Most ceramic cup sampler systems are based on Wagner's (1962) design. The cup is
cemented to one end of a plastic tube, which usually acts as the sample reservoir. A bung
seals the other end of the tube and a narrow bore tube through the bung serves to evacuate
the sampler and to recover the water sample. A number of adaptions to this simple design
have been reported. Two tubes are now usually incorporated through the bung; a short tube
to evacuate the sampler and a longer one to recover the sample (e.g. Reeve and Doering,
1965). Other modifications are designed for sampling at considerable depth, (Parizek and
Lane, 1970; Ward, 1973), to control sample volume (Chow, 1977), and to prevent flow of
collected water during soil drying (Knighton and Stebflow, 1981). Other systems connect the
sampler to additional samplers and/or vacuum reservoirs to increase the capacity and sampling
period. A separate sample reservoir is particularly useful when the collector is positioned
close to the soil surface and a long collector/sample tube assembly would be unstable.

Alundum, plastic, sintered nickel and teflon cup-based sampling systems are broad1y similar
in design to the above. Thus the cup is usually cemented or serewed to a tube which
genera11y acts as the sample reservoir. More detailed accounts are contained in the references
quoted in Table B.l. The teflon ring sampler introduced by Morrison (1982) provides an
interesting variation on the existing sampler designs. A ring of porous teflon forms part of
the sampler tube. The sample drains into the lower, sump-like section of the tube, thus
preventing drainage back through the porous section when loss of vacuum or soil drying
oceurs.

Cup and ring-based samplers are usua11y insta11ed vertica11y into a hole augered from the
surface to the desired depth. It is recommended that the cup is pressed into puddled soil from
the horizon being sampled, although it may be surrounded by silica sand. The cavity around
and above the sampler is then backfilled, usually by reconstructing the profile. In order to
prevent drainage outside the tube sorne workers have inserted a plug of clay around the tube
but above the level of the cup or ring; this introduction of extraneous material may itself
cause modifications to the soil water chemistry. Linden (1977) provides guidelines for both
construction and installation of ceramic cup samplers. The installation procedures are
applicable to other types of cup-based samplers.

Cup, and ring, type samplers can also be inserted horizontally into holes bored into the side
wall of a pit. This will eliminate the problem of drainage down the outside of the sampler
tube. Knighton and Steblow (1981) insta11 their cup sampler into a cavity cut into the side
wall of a pit. The cup is inserted into the upper face of the cavity which is then backfilled.
As the authors suggest, this sampler deSign is useful near to the surface and when surface
disturbance is likely.

E.2.3 Porous plate samplers

, 1
The first porous plate lysimeter reported was developed by Cole (1958) from 28 cm diameter
alundum filter dises. In the original design, a rubber sheet was fastened to the edges of the
dise; a connector valve was cemented into the rubber sheet and provided the outlet to the

147
wnple reservoir. A later design (Cole, 1968) had the alundum dise œmented to Jo plwgWs
base, which incotpOtate<1 an ouUet pipe. Ceramic plate collectors are similar in design with
the cemmic dise sealed to, or on, a base.

An mtetesting variation of the pl.iJ.te-baHd collector bas becn described by Bourgeois and
Luvkulich (1972). The base of the oollootor is made from acrylic material, Jo nm around the
upper surfaœ of the ~ prwuces a shallow well which is infi11td with a layer of siliwn
carbide powder. The authors claim that this material ptOduces '- better wntact with the
Wleven soil face than the rigid alundum or cewnic pl.iJ.tel.

Porous plate collectors are generally installed in small cavities eut into the side wall of a pit
or trench. The plate is pressed against the upper face of the cavity whieh is then bacldilled.
The sample reservoir is sepamte from the acrual colllXtor and linked to it by a feeder tube.

The reservoir May be sitêd in the pit or trench exc:avated during installation, or on the aroW1<1
sumec. Cole (1968) describes a very ~hi!ticated al\llldum plaw-ba.sed system which links
seveml plates ta flow ceUs which monitor conductivity, acidity and rate of flow with output
to a data 10iier.

E.2.4 Fritted glass tubes

Samplers of thb type use fritted glass tubular filters. These are available in a number of siu!
and with different porosities, e.g. Long (1978) used a 10 cm long filter with a 15.9 mm
outside diameter. The filter is lin1œd to a tube of some type which is then wlUlected to a
sample reservoir.

The samplers can he Înstalled vettically or horimntally in a similar way 10 the cup and
ring..based SYltmll5 discusseù carlier. Long (1978) suggeJu preuÎng the ~1Us nlfeI into a
slurry made from the soü at the bottom of the hole, stressing the care needoo because of the
fragility of the fritred glass.

E.2.S HollQW fibres

Stlkworth and GriSaI (1981) introduced their fibre bundle inta the soU horizontally inside a
polyvinylchloride (PVC) tube which had had "numerous" large diameter holes drilltd mto it.
Levin and Jackson (1977) pushoo a hollow stool tube horizontally intQ thç soil; the bundle of
fibres b then threa<led through the tube and the tube then withdrawn to allow soU to collap5e
around the fibres.

E.l.6 Application of tension

As noted earlier, Briggs and McCall (1904) exerted a tension by evacuating a two litre botUe
ID whieh their sampler was connecte<!. Evacuation of the sampling system is still the most
commonly used method of applying the suction or ten$ion. Hand or powered vacuum pumps
may be used depending on the sim and location of the system.

148

J
 Appendix: E: Soil Solution Sampling and Lysimerry

The volume of the sample and vacuum reservoir will determine the maximum sample volume
which can be collected and the maximum duration of the vacuum. AdditionaI "vacuum
reservoirs" can be incorporated into the system to extend the sampling period (e.g. Cole,
1968). If power is available on site, a vacuum pump can he permanently connected to the
system to provide a continuous suction.

A suction can aIso be applied by incorporating a "hanging column" or siphon arrangement

into the system (Krone et al., 1952; Cole, 1958). This will provide a continuous suction
whenever the siphon column is intact.

E.2.7 Sample recovery

The solution sample may be recovered from the sample reservoir directly, by disconnecting
the reservoir from the rest of the system, or by applying a suction or pressure to the sampler.
Direct recovery can only he used when the sample reservoir is separate from the porous
collector and accessible from the ground surface. When the sample reservoir is rigidly
connected to the porous section and/or buried suction or pressure may be applied. Suction,
or vacuum, recovery is used when the reservoir is relatively close to the surface whereas
pressure is used for deeply buried collectors (Parizek and Lane, 1970; Morrison, 1982).
Vacuum recovery is aIso frequently used to empty surface sited reservoirs so as to minimise
disturbance to the system.

E.2.8 Solution-sampler interactions

The chemistry of the solution samples can be modified as a result of interactions with the
sampler, mainly the porous material. A number of processes are probably involved,
inc1uding leaching of elements from the porous material, sorption of solutes onlo the sampler
surface, screening of ions by the sampler and diffusion through the porous material (Hansen
and Harris, 1975). Comparative studies involving two or more types of samplers
(Zimmerman et al., 1978; Silkworth and GrigaI, 1981; Bottcher et al., 1984) suggest that
teflon samplers produce little modification of solution chemistry. Zimmerman et al. (1978)
also examined teflon samplers and found no attenuation of a wide range of elements. Long
(1978) aIso states that he found no contamination of samples collected with fritted glass tubes.
Silkworth and Grigal, however report leaching of sodium from glass collectors.

Ceramic and alundum samplers. however. have been shown to release calcium, sodium,
magnesÎum, silica and sulphur inlo solution while phosphate and nitrate concentrations are
influenced by adsorption and ion screening (Hansen and Harris, 1975; Levin and Jackson,
1977). Trace metals such as Cu, Fe, Mn, Ni, Pb and Zn are aIsa sarbed (Wolff, 1967;
Steams, 1980). The precise reactions will vary between cups made from different original
clay mixes. Suitable preconditioning of ceramic and aIundum sampl~rs cao eliminate or
greatly reduce contamination by leaching impurities. Methods of pre-treatment are given by
Grover and Lambom (1970), Hetsch et al. (1979) and Neary and Tomassini (1985). These
involve leaching with dilute acid. usually 0.1 M hydroch1oric acid. prior to installation;
Hetsch et al. (1979) follow this by leaching with large amounts of "equilibrium solution".
After adequate preconditioning Hetsch el al. (1979) suggest that the ceramic cups they used

149
could be used for studies of H, Na, K. NH4 , Ca, Mg, Mn, Al, S, Cl and N03 • The
importance, howevet of a prolQnged equilibration period, afrer installation, must be stressed.
Solution-sampler internetions continue to influence phosphate concentrations even after
pre-treatment, indeed, Bottcher et al. (1984) found that ~id lc:aching increased the amount
of phospha.te adsorption of alundum and ~ramic materials. Na.gpa.l (1982) has also shown
an effeet on potassium rooovery (rom Sl)lurlons and suggests sampling methods ID minimise
this. Sorption of trace metals aIso persists and may he enhanced. Cellulose-acetate hollow
fibres al50 re1eue a number of elements due to leaching (Silkworth and GrigaI, 1981) e.g.
Na. The adsorption chamcreristics of the~ materials do not appear to have heM srudied in
detail.

E.l.9 Sampled soU volume and calculatioD of element fluxes

The volume of Sl)Î1 sampled by tension lysimeters cannot be defined preci~ly. It will vary
with the design and type of the col1ector, the tension applied and the size and distribution of
pores, channels and cavities in the soil. The very installation Qf the HD1pler resu1ts in soU
disturbance which May wel1 modify flow lines within the soU mass. Van der Ploeg and Beese
(1977) modelled the flQW of moi sture towards suction units and concluded that even weak
v~uums could produce large changes in seepage rate and that the radius of influence of a
suction could he "several foot". Warrick and Amoozegar-Ford (1977) have produced
expressions to enable one tu assess the likely sampling volume and the region of influence.
Further work of this type would seem to be needed, especially for different shapes of suction
samplers. It should be noted, however, that Tabma el al. (1979) ooncluded that flow
distribution around a porous cup was much clorer to mat of un-exrract.ed soi! than is predicted
from soil water flow theory. They suggest that this is due to flow impeilance near the cup
wall which reduces cup uptake.

Much work with alundum and ceramic plate collectors has used relatively low tensions, ca.
0.1 bar; this could only "extract" warer from the larger porcs. In wntrast, ceramic cups have
frequently))(en used at tensions of 1 bar in a delibetate attempt to sample "capillary" waret.
Studies using plate-type samplers commonly assumed that the water was derived from a
co1umn of soH, stretching above the plate, with a diameter equal to that of Ule plate. Fluxes
were ca1culated on the basis of Ulis assumptionj providoo that a contînUOu$ suction was
applied and sample collectors were Adequate for Ule volumes involvoo. This approach seems
questionable but may provide acceptable values with samplers sited near the surface. Cup,
ring and tube shaped samplers should always be regarded as ptoviding qualitative data.
Separate data on soil-water contents and movement is required before element fluxes can be
caIculated. The suction samplets are, therefore, increasingly used in parallel with tensiometer
or neutron probe systems. In sampling for flux. calculations, it is preferable if a continuous
suction is applioo su that the chemistry is integrated through time.

E.l.lO Overview

The choice of tension/suction sampler system for use in any given study will depend on many
factors including the aims of the study, soil characteristics, terrain. facilities available on site
and the level of funding avaîlable. The following .summarises sorne of the sampler

150
 Appendlx E: Soli So/llIton Samp/ing and Lystmetry

characteristics which should he considered and gives relative costs of the samplers - whole
system costs will depend on other considerations.

Ceramic cups: readily available, relatively cheap, fairly robust but brittle, vertical installation
causes minimal disturbance, must he preconditioned to reduce contamination problems, after
preconditioning suitable for major cations (although there are reservations about K), Cl, N03 ,
504 but not sui table for P04 or trace metals, qualitative only.

Alundum cups: similar to the above but not so readily available and hence more expensive.

Ceramic plates: more expensive than ceramic cups, thin plates rather fragile, installation more
complex than with cup-based systems, similar contamination/adsorption problems to ceramic
cups, sorne advantages in the plate design as it will only collect from a zone above the plate,
semi-quantitative near surface.

Alundum plates: more expensive than ceramic cups but similar to ceramic dises, more
disturbance of surrounding area of installation than with cup or ring-based systems,
contamination/adsorption problems as above, plate design advantage as above, robust.

Acrylic plates: have to he constructed in house, costs mainly in labour - uncertain advantages
of the plate design, improved contact with soil due to silicon carbide powder, solution
interactions with the silicon carbide uncertain.

Plastic cups: relatively cheap, constructed in-house from available plastic sections, minimal
installation disturbance of any systems, little contamination but there may be adsorption - no
details available; only available with relatively large pores which may admit colloids,
microbes etc. and lead to within sampler reactions.

Sintered nickel cups: moderately expensive, minimal disturbance of all cup systems, unstable
in acid soils, interactions with solutions limits use to a few elements.

Teflon cups: relatively expensîve, have to be made in-house, little disturbance at installation,
no known contamination problems, robust.

Teflon rings: commercia1ly available, relativelyexpensive, little disturbance at installation,

may be problems with blocking of pores, no known contamination problems, robust.

E.3 TENSIONLESS COLLECTORS

E.3.1 Introduction

Tensionless collectors are designed to intercept and collect water draining freely, ve_rucally
or obliquely, through soil. Probably the earliest use of this type of device was by Ebermeyer
(1876) in his c1assic study of nutrient transfers in German forests; he used a trough-like
collector to sample water draining from the forest floor. Since this pioneering work a wide
variety of shapes and sizes of collector have been constructed from a range of materials.
Thus copper, stainless steel, galvanised steel, perspex, polyvinylchloride and polypropylene

151
have all ~n uJed. The colle.ctors are genem11y made in-house M.d May bë consu-ucted from
sheets of the selected material or rnay adapt and incorporate prefonned units, e.g. boxes,
bowls, trays. funnels or rainwater guttering. The deviœs are usually installed into the
sidewall of a pit or trench. The collootor may he oonne.ctM directly to the wnple reservoir
or more comman1y, linked to it by a feeder tube.

E.3.2 Troup, box and funnel-based coUecton

Such collectors are generally installed in a cavity excavated in the sidewall of a pit or trench.
The deviœ may he filled with soU excavatc<t from the cavity before it is inserted and brought
into contact with the upper face of the cavity. Altematively the collector is pu shed upwards,
into the upper face of the cavity. The approach chosen will depend on factors such as soil
texture and stoniness, and density and size of rooU. As in aIl tensionless collectors, a flUer
system should be incorporated lO prevent clogging of tubes leading to the wnple coll~tor.
In filled bowls, troughs etc .• a layer of acid wash sand or plastic beads can he placed in the
base before filling. Alter insenion of the collector, the cavity should be bacldilled to keep
the deviœ in place and in ~ntacl with me overlying soiL

Jordan (1968) has described a trough-shaped device which has proved useful in a wide range
of site types. The trough is constructed from stainless steel and rests upon adjustable wooden
supports. Once inserted into the sidewall cavity the supports can he adjusted in length in
order to press the collcctor against the upper face of the cavity. A fibreglass scroon lined
with glass wool is suspended inside the trough. Steel rads run along the underside of the
fibreglass sereen 10 the ouUet tube. The glass wool, fibreglass and steel rods together help
to overcorne surface tension within the soil and ensure môvement of waret from the soH into
the collector. As Jordan notes, without sorne such device water tends (0 bang on the soil face
rather than drip into the col1~tot.

Before the collector is installed into the pre-dug cavity soil is tamped on (OP of the glass
wool-fibreglass sereen until it is level with the top of the trough. The design can be copied
using prefonned plastic guttering, nylon bolting cloth for the screen and plastic rods below
the sereen.

Boomer (1982) has described a simple trough shaped tension-free lysimeter for use in sandy
soils. The collector is a V-shaped, made from 6 mm thick plexiglass, with sides 50 cm x 7.5
cm. lt was installed in10 tunnels, cut into the end walls of soil pilS, al an angle of ca. 20°
ta the horizontal. After inserting the trough the tunnel was repacked with the excavated soiL
A "sump plate" inserted vertically into the e~posed end of the trough prevents soi! slumping
into the exposed end of the trough. A "dust cover" prevents e~traneous material entering the
exposed section. The sample bottle is housed on the floor of the soïl pit and linked to the
lysirneter with flexible plastic tube.

E.3.3 Sheet or tray-based roU~ton

A frequenUy use<! type of collector comprises a shoot or tray-lilœ device inserted horizontally,
or at a slight angle to ensure drainage, into the side wall of a pit. A popular design has a fiat

152
 Appnulix E: Soil Sollllion Sompling and Lyslmetry

base with a small front edge, side walls which reduce in height from front ta back but no
vertical rear edge.

Parizek and Lane (1970) refer ta this type of collectar as a pan lysimeter. They constructed
30 cm x 42 cm pan lysimeters from 16 gauge galvanised metal. A copper tube was soldered
inta the front, raised edge of the pan to allow the water ta drain away ta a sample reservoir.
The pans were installed inta the side wall of a trench. A sheet metal blade was first driven
inta the trench face ta provide an access slit into which the pans were pushed. Wooden sides
were put into the trench with a small gap between them and the pit face, and then this gap
was backfilled with soil. The sample boUles, linked ta the pan lysimeters, were placed on
the trench floor.

Nys (personal .communication), has constructed a sampler of this type from 1 cm

polypropylene sheet; the sections are joined by heat welding. The resulting collectar is very
robust and, with the leading edge chamferred, will cut through roots up ta ca. 1.0 cm
diameter during installation. The collectars were forced iota the pit face using hydraulic
jacks. A small quantity of acid washed sand was placed above the outlet pipe ta prevent
clogging. After forcing the plate inta the trench face a narrow sheet of plastic is inserted
above the exposed face of the plate ta prevent drainage down the soil face inta the collector.
To maintain soil temperature regimes etc., Nys places a wooded wall a short distance out
from the trench face and backfills the gap by reconstructing the profile.

Reynolds (persona! communication) bas constructed a similar, but less robust device from 3
mm pve sheet. The cut sections are cemented tagether using pve cement. This collectar
is not strong enough to he forced iota the pit face; a eut is therefore made with a knife or
trowel before insertion.

Large plastic sheets can be inserted into a pit face, at a slight angle to ensure drainage, with
a small edge left protruding from the face. The edge overhangs a trough or gutter which runs
along the face and is connected to a sample bottle. This type of collector is referred ta again
below. Trough-like containers can also be inserted into a pit face rather than into a cavity
cut into that face. The trough should have no vertical rear end piece. This type of collectar
can he constructed from plastic gutters. As the plastic material is not very robust, a cut may
have to be made into the soil face hefore insertion. Altematively, a metal former can be
driven in first ta ease insertion of the plastic trough.

The direct insertion of these pan and trough types of collector into a pit face can result in a
smearing of the sail face in contact with the plate or trough. This may effectively seal the
soil, greatly reducing the yield of water and undermine any calculation of fluxes. The risk
of smearing is reduced if the collectar is inserted during dry weather but the texture of the
soil is the main determining factor. Direct insertion is also difficult in stony soils or where
there are abundant large roots.

E.3.4 Tensionless coUectors on sloping sites

Most tensionless lysimeters have been used on level or gently sloping sites and have bren
assumed to collect water draining vertica11y through the soil masse On slopîng sites,

153
11i8F: Ji Hawlbool 0/ JI.dwtJa

hgwever, a large pan ôf the ~ter mgvement is lat.emI or obliQue. ~sek and Lane (1970)
used pan lysimetcrl to interœpt this laterally moviog water. An alternative approach uses a
colleçtign system tirst used in hydrological studies on slopes (Whipk,y, 1965) and probably
fust used in solute flux studies by Roose (1968). The approach usesa trench dug at right
anales to the slope. Shcets of plastic or metal ue pushed mto the ~p$lope face of the pit,
commonly at the base of each main sail horizon, so tha.t little of theSieetstill protrudes from
the trench face. Gutters are installed along the face of the pit und each sheet overhang.
The shoots Întereêpt the latemlly moving water which moves aCMSl; eir upper surface and
drips down ioto the guuer. 1

At one end Of each gutter is a connecto! and a feeder pipe whi(:h links it ID a sampte bOttlc
on the floor of the trench. The gutters are usually mounted onto a wadden fwne and wooden
walls ma)' he oonstnIcted to prevent collapse of the pit faces. Detai1t.td descriptions, and/or
diagrams, of this type of installation are provided b)' Roose (196~), Knapp (1973) and
Williams el al. (1984). 1

E.3.5 Sample eomamination

Sample contamlnation due to solution-sampler interactions cao be v~y eliminated by

careful selection of construction matcrials. Metals, other tha.n stainlessst:eel, mould probably
he avoided; they are gradually attacked, especially in acid soils. 1 Plastic materials are
particularly suitable but tests mould bc carried out on any new pJ.a.stiC to assess the possible
I
adsorption of phosphate or trace metals and the retease of flUers. Some plastics also "age"
relatively rapidly and their adsorption characteristici may then change- ln general, "high
denJitylt plastics have much lower adsorption than low dcnsity ones. :The use of fibreglass
filters, or screens, and of silica sand ta avoid clogging of drainage Pirœl ma)' lead ta silica
enhançements in solutions. The fibreglass May aIso have u desirablc adsorption
chamcteristic$. ,
'1

E.3.6 Calculatlon of fluxes

Wncn tcnsionless lysimeters are used in solute flux studiœ on level sit~, il is assumed that
they collcct water dmining froru a column of soit, with a cross-sectional area equal to that of
the lysimeter, stretching venica11y trom the lysimeter 10 the ground suh"ace. This may he a
vaUd assumption on sorne sites but on others routing of water do."n rnacro-pores may
concentrate. or· disperse, flow. As a result, closely adjacent collectprs may consistentJy
collect very different volumes (e.g. Russel and Ewel, 1985). Flux studles should, therefore,
use as large a number of Iysimeters as pg~ble. 1

On !doping sites, it is even more difficult to define the source area of the water collccted.
One approach il to isolate an area stretching upslope from the trenc;h housing the collecton.
This cao be done by driving sheets of material vertiœlly into the soil or excavating a
boundary rrench, inserting a vertical, waterproof barrier and backtilling the trench on eimer
side of the barrier.

In any study of fluxes, it iJ essentiaI that the sample reservoirs are large enoug11 to
accommodate aU the water draining from the l)'simeter du ring the sampling interval, or that

154
 Appmdix E: Soil Sollllion Sampling D1Ul LysIIMtry

a sample splitter is incorporated. The excavation of the pit or trench prior to installation of
the tensionless Iysimeter will cause alterations to drainage lines within the soil. If a pit, or
trench, is left open it will have a sump effect with drainage lines converging on the pit,
especiallyon sloping ground. Backfilling the pit will reduce this effect but not eliminate it.
The magnitude of the effect will vary with soil texture and other site and soil characteristics.

References

Anderson, J.M., Leonard, M.A. and Ineson, P. (1987) Lysimeters with and without tree
roots for investigating the role of macrofauna in forest soils. In: Harrison, A.F.,
Ineson, P., Hea1, a.w. (eds.), Field Methods in Terrestrial Ecosystem NU/rient
Cycling. Elsevier, Amsterdam.
Belford, R.K. (1979) Collection and evaluation of large soil monoliths for soil and crop
studies. Journal of Soil Science 30, 363-373.
Boemer, RE.J. (1982) An inexpensive, tension-free Iysimeter for use in sandy soils.
Bulletin of the Torrey Botanical Club 109, 80-83.
Bottcher, A,B., Miller, L.W. and Cambell, K.L. (1984) Phosphorus adsorption in various
soil-water extraction cup materials: Effect of acid wash. Soil Science 137, 239-245.
Bourgeois, W.W. and Lavkulich, L.M. (1972) A study of forest soils and leachates on
sloping topography using a tension lysimeter. Canadian Journal of Soil Science 52,
375-391.
Briggs, L.J. and McCall, A.F. (1904) An artificial root for inducing capillary movement of
soil moisture. Science 20, 566-569.
Calder, I.R (1976) The measurement of water los ses from forested area using a "natural"
Iysimeter. Journal of Hydrology 31, 311-325.
Chow, T.L. (1977) A porous cup soil-water sampler with volume control. Soil Science 124,
173-176.
Cole, D.W. (1958) Alundum tension Iysimeter. Soil Science 85, 293-296.
Cole, D.W. (1968) A system for measuring conductivity, acidity and rate of water flow in
a forest soil. Water Resources Board 4, 1127-1136.
Czeratzki, W. (1959) Untersuchung der Wasserbewegung im Boden mit Hilfe von
Unterdrucklysimeter. Z. Pflamenem. DUngung. Bodenkunde 87, 223-229.
Czeratzki 1 W. (1971) Saugvorrichtung für kapillar gebundene Bodenwasser. Landbauforschg.
VlJlkenrode 21, 13-14.
Duke, H.R. and Haise, H.R. (1973) Vacuum extractors to assess deep percolation losses and
chemica1 constituents of soil water. Soil Science Society ofAmerica Proceedings 37,
963-964.
Ebermayer, E. (1878) Die gesammte Lehre der Waldstreu mit Rücksicht auf die chemische
Statik des Waldbaues. Springer, Berlin.
Grover, B.L. and Lambom, R.E. (1970) Preparation of porous ceramic cups to be used for
extraction of soil water having low solute concentrations. Soil Science Society of
America Proceedings 34, 706-708.
Hansen, E.A. and Harris, A.R. (1975) Validity of soil-water samples collected with porous
ceramic cups. Soil Science Society of America Proceedings 39, 528-536.
Harris, A.R and Hansen, E.A. (1975) A new ceramic cup soil-water sampler. Soil Science
Society of America Proceedings 39, 157-158.

ISS
'ISlJF: A Handhoo! 0/ Mr:r1wcb
HelBCh. W., Beese, P. and Ulrich. B. (1979) Die Beeinflussung der Bodenlosung durch
Sauglœrzen aus Ni-Sintermemll und Keranùck. Z. Planzenemaehr. Boden/amtk 142.
29-38.
Iacmn, D.R., Brinkley. R.S. and Bondietti t B.A. (1976) Extraction of soil water mring
cellulo8e-oootate hollow nbreJ. St>iI Scie~e Sor;iery 01 America ~udin8S 40.
327-329.
1ordan. C.F. (1968) A 8imple, tenlÏon-free lysimeteI. Soil Science 105, 81-86.
Knapp, B.J. (1913) A system for the field measuremem of soil water movemellI. British
Geornorphologica1 Research Group Technica1 Bulletin Number 9.
Knighton. M.D. and Streblow, D.W. (1981) A more versatile soil water sampler. Soil
Science Society of America Jouma! 45. 158-159.
Krone, R.B .• Ludwig, H.F. and Thomas, J.F. (1952) Porous tube device for sampling soil
solutions during water-spreading operations. Soil Science 73. 211-219.
Levin, M.I. and lacleJon, D.R. (l977) A comparison of in situ ~xtraçton for sampling soil
waœI. Soil Science Sociery of ~rica Joumal41. 535-536.
Linden, D.R. (1977) Design. installalion and use ofporous ceramie samplersfor monitoring
soil-wœer qUlllity. USDA Technical Bulletin Number 1562, II pp.
Long F .1. (l978) A glass tiller soil solution sanlpler. Soil Science Society of Americ"
Jounwl42, 834-83~.
Morrison, R.D.C. (1982) A modificd vacuum-pressure lysimeter for soïl water sampling.
Soil Science 134, 206-210.
Nagpal, N.K. (1982) Comparison among and evaluation of ceramic porous cup soit water
HIIlples for nutrient mm~pon Jtudies. CUTIUClitmJoumal o/Soil Sde~e 62, 68~-694.
Nearyt A.J. and Tomanini. F. (985) Preparation of !luJ1dumlçewnic plate tension
lysimeters for soil water col1ection. Jt>unml of Soil Science 65, 169-179.
Parizek. R.R. and Lane, B.E. (1970) Soil-water sampling using pan and deep
pressure-vacuum lysimeter. Journal o/Hydrology 11, 1-21.
Quin. B.F. and Forsythe, L.J. (1976) AIl-plastic suction lysimeter~ for the rapid sampling
of percolating soil water. New Zelllll1fl1 Jou11M1 of Sr:ience 19, 143-148.
Recwe, R.C. and Doering. E.l. (1965) Sarnpling the soU SOlution for salinity apprabal. Soi!
Science 99. 339-344.
Roo8e, RI. (1968) Un dispositif de mesure du le~!iivage oblique dans les sols en pillee.
Cab. ORSTOM. ser. Pl!dnlogi~ VI. 23~-249.
Roose, E.l. and Des Tufeaux, M. (1970) Deux methodes de nature du drainage venica1 dans
les sols en place. Agronomique Tropicale 25, 1079-1087.
Russel, A.E. and Ewel, 1.1. (198~) Leaching fmm a tropical andept during big stOIIDS, a
comparison of three merbods. Soil Science 139, 181-189.
SUkworth, D.R. and Grigal, D.P. (198l) Field comparison of soit solution samplers. Soil
Science Sociecy of America Journal 45, 440-442.
Talsma, T., Hallarn, P.M. and Mansell, R.S. (1979) Evaluation of porous cup soil-water
extractors, physical factors. AlIStrolian Journal Qf Soit Resources 17, 417-422.
van der Ploeg, R.R. and Beese, F. (1917) Model ca1culations for the extraçtion of soiI water
by ceramic cups and plates. Sail Sciençe Society of America Journa/41, 466-470.
Wagner, G.H. (1962) Use of pornus eemmic cups to sample soil water within the profile.
Soil Science 94, 379-386.
Warriçk, A. W. and Amoozegar-Fard, A. (1971) Soil water regimes near porous çup water
samplers. Water Resources Research 13, 203-207.

156
 Appendlx E: Soli Sollllion &»npllng and Lysimetry

Whipkey, R.Z. (1965) Measuring subsurface storm flow from simuloted rainstorms - a plot
technique. U.S. Forest Serviee Central States Forest Experimental Station Research
Note CS-29.
Williams, A.F., Ternan, J.L. and Kent, M. (1984) Hydrochemical characteristics of a
Dartmoorhillslope. In: Burt, T.P. and WaDing, D.B. (eds.), Catclunent Experiments
in Fluvial Geomorphology. Geobooks, Norwich, U .K.
Wolff, R. (1967) Weathering of woodstock granite near Baltimore, Maryland. American
Journal of Science 265, 106-117.
Zimmermann, C.F., Priee, M.T. and Montgomery, J.R. (1978) A comparison of eeramic
and teflon in situ samplers for nutrient pore water determinations. Estuarine Coastal
Marine Science 6, 1-5.

157
Appendix F NITROGEN AVAILABlllTY
Cheryl PaIm, Phil Robertson and Peter Vitousek

F.1 INTRODUCTION

Nitrogen is undoubtedly the most intensively studied plant nutrient in both natural and
manaaed ecosystems. Despite its imponan~ to plant growth and ecosystem productivity,
there is still no reliable soil test for nitrogen nvallability (Stallford. 1982). The rcasons for
thi! are numerous (Re Stanford. 1982 and Stevenson, 1986 rôr revîews).

son nitrogen far the Most part is in the organie fom (Stevenson, 1986) and unavailable te
plantl. Organic nitrogen is mineralisod to ammonium, a plant-available form, by microbial
proeesses. The rate of mineralisation depends on many factors including temperature and
rainfall. the quality ôf the soit organic nitragen t and the quality of organic inputs to the
ayaœm. Once in the mineral form, nitrogen, li1ce other plant nutrienUi, cao he taken up by
plants or microbes (immobilisation) or lost fram the system by leaching. In addition. nitrogen
can undergo various transformations, many of them biologically mcdiated. Many of these
transrormations such as nitrification, volatilisation, and denitrifieation can ultimately lcad tu
losse~ from the system.

In natural ec.o5ystems nitrogen losses arc generally assumoo to be smalt because of efficient
nutrient cycling, mineralisation occurring in synchrony with plant uptake. In managed
systems however, this synchrony may he disrupte.d leading 10 a build up of mineral nitrogen
in the sail.

This nitrogen is susceptible to losses via the various pathwnys discussed. Leaching losses cao
be large, espedally in the humid tropics where mnrall b high and a majority of the soifs
have tittle retennon eapacity far ç.dions or anions. LosBes of nitrogen thmugh denitrification
may he as large, or Inrger, than through lcaching in disturbed or managoo systems. Little
is knawn, howevcr. about deniui.fication in tropi~ ecosy5tems, natura! ar managcd. The
potential for denitrification should be high in tropical systems where nitrification rates are
high and high rainfall cao lead to anaerobic conditions in the soil. There are indications that
denitrifieatian il higher in undisturbed tropical forest ecosystems than those in tempcrate
regions (Keller et al., 1986; Mooney et al. 1987) and that thele rates incrcase with
1

disturbance of the system (Keller et al. t 1986; Marson et al., 1987). In arid tropical
ecosystems, most nitrogen loss oceurs as a re~ml1 of fires.

One of the major objectives of the TSBF Programme Îs to find ways of minimising losses of
nitrogen (as weil as other nutrients) in managed systems by synchronising mineralisation with
plant uptake. This will requiœ an understanding of the controls of the various
transformations and fluxes of nitrogen and how these are affected by management. Part of
the problem in understanding nitrogen cycling in ecosystems has becn the lack of

158
 Appendix F: Nitrogen A vailability

methodologies for measuring the different processes. In this section methodologies for
estimating net mineralisation and denitrification and the problems associated with them are
presented. The collection of leachate is discussed in Appendix E, "Soil solution sampling and
lysimetry" .

F.2 MINERALISATION

The concentration of mineraI nitrogen in soil reflects a balance between inputs through
mineralisation of organic N and outputs through microbial immobilisation, plant uptake,
leaching, and denitrification. Most methods for estimating nitrogen availability assume it to
be equivalent to net nitrogen mineralisation (gross mineralisation minus immobilisation); they
are designed to prevent (or quantify) mot uptake, leaching, and denitrification by incubating
soil in the absence of active roots. Any method however, which involves either severing
roots or removing them from soil alters the soil environ ment to a greater or lesser extent.
Hence, any such effort to measure soil nitrogen availability interferes with the process, and
a brœd-scale measurement programme must seek a comparable and realistic index of nitrogen
availability rather than a direct measurement of the process.

No single method for estimating nitrogen availability has gained universal acceptance, and
indeed it is unlikely that any single method will prove applicable to aIl sites and conditions
(Keeney, 1980). Moreover, relatively few of the extant methods have been tested in tropical
soUs. The TSBF Programme initially adopted a field incubation method (below) as the
standard, but workers in high-rainfall areas have reported problems with flooding of soil
cores. Other studies have reponed substantial losses of tield- incubated cores to soil fauna
(Vitousek and Denslow, 1986). It was therefore decided to use aerobic incubations in the
laboratory as the simplest standard method, and to suggest optional use of more realistic field
incubations where practical (see (b) below).

Laboratory methods for estimating N-mineralisation were reviewed by Brown (1982), who
showed that a wide range of incubation times, remperatures and conditions, and extraction
procedures have been employed with littIe attempt at standardisation. Many authors have
incubated soil &lmples over long time periods assuming that mineralisation is linear over time.
Myers (1975) however. reported for tropical soils that ammonification rates in the tirst seven
days of incubation were about six times higher than between 14 and 28 days.

It is also difficult to control soil moi sture while maintaining aerobic conditions over long
periods of incubation. Quartz sand may be used to disperse soil, but ambient field moi sture
conditions cannot be maintained using this method. As a compromise on the various
limitations of laboratory incubations, the simplest aerobic incubation procedure will he used
as a baseline for TSBF studies (see Section 6.7.3 for details). Details are also given (Brown,
1982) for an anaerobic (waterlogged) laboratory incubation method of Waring and Bremner
(1964)._ This method is rapid, and it may have potential as an index of the labile pool of
plant-available N; it does not approximate field minera1isation processes in any way.

Methods used to estimate patterns of N mineralisation under field conditions include:

159
a) Incubation of disturbed (sievoo or mixed) soU in plastic bags buried in the field (e.g.
WestermaM and Crowthers, 1980).

b) Incubation of relatively undisturbed soil columns (cores) enclosed in plastic bags or

columns onder field conditions (Nadelhoffer et al•• 1985; Matson el al.• 1986; Raison
et al., 1987). Rapp el al. (1979) performed in situ incubations aRer isolating the soU
withln thin metal caus pushed into the soil.

c) Measurement of mineral-N collecœd by ion exchAll&c Ie5ms placed in the field fur
extended pcriods (e.g. Hart and BinJdey, 1985).

d) Detemùning the effects of varying temperature and moi sture on N mineralisation in

disturbed soi1s in the Iaboratory, and then assuming that these relationships can he
linlœd to variations in these paramerers in the field - i.e. a modelllng apprœch (e.g.
Cameron and Kowalenko. 1916; Marion el al., 1981; Macduff and White, 1985).

Inaddition tg ghanges fnduced by severing or mmoving rOOl$ (Hœdrickson and Robinson,

1984) the folloWÎ.fta potcntial difflculdes mayapply;

i) soil distUIbance mar1œdly affects mineralisation (e.g. Runge, 1914; Nordmeyer and
Richter, 1985). This effect will be signiticant for methods (a) and (d); it may be quite
small for method (b);

ü) rapid fluctuations in soil water content markedly affect mineralisation in many

environments. With methods (a) and (b) the soU mOÎ!lwre content at the time of
sampling is maintaitled tluoughout me incubation wriod;

iü) method (c) estimates the amount of nitrogen in percolating water (Binkley. 1984). not
net nitragen minemlbarion. It is insensitive 10 any N either absorbed by roots or
exchanged on sail surfaces; and

iv) for modellîng approaches to be useful, temperature and moisture response surfaces
must be determined for undisrurbe<1 soUs. and account nee4s to be taken of seasonal
variations in the pools ofmmemlisable otganic Nin soils (popovic. 1971; Ellis, 1974;
Theodomu and Dowen. 1983; Nordmeyer and Richter, 198.5).

Of the ahove methods, (b) appears Most sui table for a wide range of Si~i.

F.3 DENITRIFICATION

Denitrification is the biological rcducôon of soit nitrate to nitrous oxide (N20) and dinitrogen
(N:J gues. The process is carried out by a diverse group of bacteria using nitrogen oxides
as terminal elcctron acceptol'$ in lieu of oxygen under anaerobic condition!l. In well-aeratcd
soils these ~nditions most olten exist within soU aggregates or where high concenlrndon~ of
organic matter ereate strong, localised O2 !inles.

The reductive pathway for denitrification is genera11y accepted to be~

160
 Apptndix F: Nilrogtn A vailabllity

NO]- -------- > NOi ------- > NO ------- > N20 -------- > N2

with the ratio of N20:N2 liberated by the denitrifying community highly variable, underthe
control of a wide range of environ mental factors (Firestone et al., 1980). Though
denitrification is in most soils the largest single source of nitrogen gas, N 20 may additionally
arise by a variety of other pathways. These inc1ude the chemical decomposition of
hydroxylamine and nitrite reduction during nitrification (Blackmer et al., 1980; cf. Poth and
Focht, 1985), dissimilatory nitrite and nitrate reduction to ammonium (Bleakley and Tiedje,
1982), microbiaI nitrate assimilation (Satoh et al., 1981), and other pathways as yet poorly
described (Robertson and Tiedje, 1987).

Experimental methods for quantifying denitrification are reviewed by Focht (1982), Tiedje
(1982), and Tiedje et al. (1989), inc1uding lSN tracer techniques. The simplest way to
measure denitrification in field studies is the acetylene inhibition technique (Duxbury, 1986;
Tiedje et al., 1989). Acetylene at relative1y 10w concentrations « 10 kPa or < 10% v/vat
1 atmosphere) blocks the expression of N 20 reductase in most soils, so that the N 20 which
might otherwise have been reduced to N 2 remains as N20, easily measured against
background atmospheric concentrations.

Ryden and Dawson (1982) report a field method where a constant stream of acetylene gas
was injected 1 m below a pair of boxes covering a pasture sward while a second pair of
boxes was untreated. Gas products from the two series of chambers were collected over 3-4
br and scrubbed for N 20 using a molecular sieve; N 20 was then recovered and determined
by gas-liquid chromatography. The estimates of N losses by denitrification using this method
at intervals over the growing season were in close agreement with estimates obtained from
N budgets for the pasture system.

The main limitations on the acetylene inhibition technique are ensuring (i) that acetylene
diffuses to the soil microsites where denitrification is taking place, and (ü) since acetylene
aIso inhibits nitrification, that denitrification rates are quantified before nitrate becomes
limiting. Both cbamber and soil core techniques have been used to quantify denitrification
using the acetylene inhibition method; both techniques yield similar results (Ryden and
Skinner, 1987; Tiedje et al., 1989) but the static core technique has the advantages of being
relative1y inexpensive and amenable to large sample sizes, important considerations given the
high degree of spatial heterogeneity typical for this process in a variety of environments. The
static core' s principal disadvantage is that great care must he taken during sampling and
subsequent analysis to avoid disrupting anaerobic microsites in soil by disrupting soil
structure. The recommended TSBF procedure is given in Section 6.8.2.

References

Binkley, D. (1984) Ion-exchange resin bags for assessing soil N availability: the importance
of ion concentration, water regime, and microbial competition. Soil Science Society
of America Joumal48, 1181-1184.
Blackmer, A.M., Bremner, J.M. and Schmidt E.L. (1980) Production of N 20 by
ammonia-oxidising chemoautotrophic microorganisms in' soil. Applied and
Environmental Microbiology 40, 1060-1066.

161
TSBF.' A Ii~ 0/ M~t1wtb

Blealdey, B.H. and Tiedje. J.M. (1982) Nitrous oxide production by organisms other than
nitrifiers or denitrifiers. Applied and Environmental Microbiology 44, 1342-1348.
Brown, C.M. (1982) Nitrogen mineralisation in soils and sediments. In: Burns. R.G. and
Slayfer, J.H. (eds.), Experimental MicrobiaJ Ecology. Blackwell Scientific
Publications, Oxford, pp.154-163.
Cameron, D.R. and Kowalenko, C.G. (1976) Modelling nitrogen processes in soil:
mathematica1 development and relationslùps. Canmlian JOl4muJ of Soil Scil!1U:l! 56.
11·78.
Duxbury, J.M. (1986) Advanl8&e8 of the acetylene method for me.asuring denitrification. In:
Hauck, R.D. and Weaver. R.W. (oos.), FitM Measuremem of Dinitrogen Fixarion
and Denitrification. Soil Science Society of America, Madison, Wisconsin, pp.73-91.
Ellis, R.C. (1974) The sea.sonal pattern of nitrogen and carbon mineralisation in forest and
pasture soils in southem Ontario. Canadian Journal of Soil Science 54, 15-28.
Firestone, M.K., Firestone, R.B. and TIedje, J.M. (1980) Nitrous oxide from soi!
denitrification: factors controlling its biological reduction. Science 208, 749-751.
Pacht, D.D. (1982) Denitrification. In: Burns, R. and Slayter, J.H. (OOs.), ExperimemaJ
Microbial Ecology. Blackwell Scientific. Oxford, pp. 194-211.
Hart, s.c. and Binkley, D. (1985) Correlations among indices of forest soU nutrient
availability in fertilised and unfeni1iifd loblolly pine plantations. PIani and Soll8.5,
11-21.
Hendrickson, O.Q. and Robinson, I.B. (1984) Effects of roots and litter on mineralisation
proœsses in forest soil. Plant and SoU 80. 119-140.
Keeney. D.R. (1980) Prediction of soil nitrogen availability in forest ecosystems: a litemture
review. Forest Science 26. 159-171.
Keller, M., Kaplan, W.A. and Wofsy, s.e. (1986) Emission of N;zO. CH", and COl from
trovi~ forest soils. JQumul of Geophysictll Research 91, 11791-11802.
Macduff, J.H. and White, R.E. (1985) Net minernlisation and nitrification rates in a clay soil
measured and predicted in permanent grassland from soit temperarure and moisrure
wntent. Pillm and soU 86, 151-172.
Marion, G.M., Kummerow, H. and Miller, P.C. (1981) Predicting nitrogen mineralisation
in chaparral soïls. Soil Science Society of America Journal 45, 956-61.
Matson, P.A., Vitousek, P.M., Ewel, J.J., Mazzarlno, M.H. and Robertson, G.P. (1986)
Nitrogen transformations following tropical forest falling and buming on a volcanic
soU. Ecology 68, 491·502.
Mooney, H.A., Vitousek. P.M. and Matson. P.A. (1987) Exchange of materiab between
terrestrial ec.osystems and the almosphere. Science 238, 926-932.
Myers, R.I.K. (1975) Tempemture effecls on ammonificatiOn and nitrificanon in a tropical
soïl. SoU Biology and Biochtmislty 7, 83-86.
Nadelhoffer, K.1 .• Aber, J.O. and Melîllo. J.M. (1985) Fine toots, net primary production,
and soi! nitrogen availability: a new hypothesis. Ecology 66, 1377-90.
Nordmeyer, H. and Richter, J. (1985) Incubation experiments on nitrogen Mineralisation in
loess and sandy soils. Plant and Soil 83, 433-4~.
Popovic, B. (1971) Effecl of sampling data on nitrogen mobilisation during incubation
e;t!;periments. PlaJJ/ and Soi134, 381-92.
POID, M. and Focht, D.D. (1985) ISN kinetic analysis of N20 production by Nitrosomonas
europaea; an exarnination of nitrifier denitrification. Applied and Environmenlal
Microbiology 49. 1134-1141.

162
 Appendix F: Nitrogen A lIailability

Raison, R.J., Connell, M.H. and Khanna, P.K. (1987) Methodology for studying fluxes of
soil mineral-N in situ. Soil Biology and Biochemistry 19, 521-530.
Rapp, M., Leclerc, M. Cl. and Lossaint, P. (1979) The nitrogen economy of a Pinus pinea
L. stand. Forest Ecology and Managemem2, 221-231.
Robertson, G.P. and Tiedje, J.M. (1987) Nitrous oxide sources inaerobic soils: nitrification,
denitrification and other biological processes. Soil Biology and Biochemistry 19,
187-193.
Ryden, J.C. and Dawson, K.P. (1982) Evaluation of the acetylene-inhibition technique for
the measurement of denitrification in grassland soils. Journal of the Science of Food
and Agriculture 33, 1197-1206.
Ryden, J.C. and Skinner, J.H. (1987) Soil core incubation system for the field measurement
of denitrification using acetylene-inhibition. Soil Biology and Biochemistry 19,
753-757.
Satoh, T.H., Hom, S.S.M. and Shanmugam, K.T. (1981) Production ofnitrous oxide as a
product of nitrite metabolism by enteric bacteria. In: Lyons, J .M., Valentine, R.C.,
Phillips, D.A., Rains, D.W. and Huffaker, R.C. (eds.), Genetic Engineering of
Symbiotic Nitrogen Fixation and Conservation of Fixed Nitrogen. Plenum Press, New
York, pp.481-497.
Standford, G. (1982) AssesmentofSoil Nitrogen Availability pp 651-720 In: Stevenson, F.J.
(ed.), Nitrogen in Agricultural Soils. American Society of Agronomy, Madison, WI,
pp.651-720.
Stevenson, F.J. (1986) Cycles ofSoil Carbon, Nitrogen, Phosphorus, Sulfur, Micronutrlents.
John Wiley and Sons, New York.
Theodorou, C. and Bowen, G.D. (1983) Nitrogen transformations in first- and
second-rotation Pinus radiata forest soil. Australian Forest Research 13, 103-12.
Tiedje, J.M. (1982). Denitrification. In: Page, A.L., Miller, R.H. and Keeney, D.R.
(OOs.), Met1wds of SoUs Analysis, Pan 2. Chemical and Microbiological Properties.
American society of Agronomy, Madison, Wisconsin, pp. 1011-1026.
Tiedje, J.M., Simkins, S. and Groffman; P.M. (1989) Perspectives of mtasuremtnt of
denitrification in the field including recommended protocols for acetylene based
met~. (In Press).
Vitousek, P.M." and Denslow, J.S. (1986) Nitrogen and phosphorus availability in treefall
gaps of a lowland tropical rainforest. Journal of Ec%gy 74, 1167-1178.
Waring, S.A. and Bremner, J.M. (1964) Ammonium production in soil under waterlogged
conditions as an index of nitrogen availability. Nature 201, 951-952.
Wesrermann, D.T. and Crowthers, S.W. (1980) Measuring soil nitrogen mineralîsation
undet field conditions. AgroMmy Jounml 72, 1009-1012.

163
Appendix G TECHNIQUES FOR QUANTIFYING NITROGEN
FIXATION
Mark Peoples

G.l INTRODUCTION

There is no single "correct" way of measuring biological Nz fixation. Ir is unrealistic to

expect one teçhniQue [0 provide accurate field measures of N 2 fixation for ail of Lhe diverse
organisms which have the potential 10 fix Nz in symbiotic or associative relationships, or in
the free-living state. Each methodology has its own unique advantages and limitations which
make it more or le5s Imitable for particular species and envimnrnents (Ledgard and Pooplcs,
1988).

Pn;x;cdures mat have ~ommonly beeII used for meuuring symhiotic N1 fixation by legumes
will be discussed briefly in the following sections. Application of techniques to non-legume
systems have bœn reviewed by Rennie and Rennie (1983) and Boddey (1987).

G.2 NITROGEN DIFFERENCE METHOD

A measure of N1 fixation by a Iegume crop CM. be obtained by growing a ~mpanion

non-N1-fixing reference crop in the same soil and under identical conditions (usually in an
adjacent plot) as the legume. The difference in nitrogen yie1d between the legume and
reference crop is regaràed as the contribution of symbiotic fixation to the legume. It is
assumed that the legume and reference CfOps assîmilate the sanIe amount of soïl nitrogen (and
fertiliser). However, identi(;a} usage of soU nitrogen dues not a1ways OCCut in the field and
results may olten he ÎtlaéCurate (e.g. Boddey ~t al•• 1984). For sorne field leaumes, non-
nodulating isolines are available, reducing the difficulties of comparing different rooting
behaviours and nitmgen utilisation. Nonetheless the nitrQgen- difference rnethod il a
reladvely simple proçedure and is ofren used \\lhen only total nitmgen analySÎs il available.

Estimates of N2 fixation appear tu he most reliable in soils low in available nitro,en.

Examples May he found in LaRue and Patterson (1981) and Chalk (1985).

G.3 l'N-ISOTOPIC TECHNIQUES

The stable isotope of nitrogen, \SN, oceues in atmospheric Nl at a constant abundance of

0.3663 atoms %, If the ISN abundance is different in plant-avaUable soil nitrogen from that
of atmospheric N2 , measurementl of the proponions of legume nitrogen derived from each
source cau he ca1culated from measurements of the isotopie abundances in the nodulated
legume and in non-N2 -fixing reference plants which are totally dependent on soU nitroaen.

164
 Appendü: G: Techniques for QlUIIIlifyillg Mtrogen FiJuuiofl

G.3.1 I~-enrichment

The use of methods involving artificial adjustment of soil lSN enrichments to measure N 2
fixation bave been reviewed by a number of authors (e.g. Chalk, 1985; Hauck and Weaver,
1986; Danso, 1988; Ledgard and Peoples, 1988). A major assumption of the technique is
that the legume and non-N2-flXing reference plant absorb the same relative amounts of
nitrogen from added lSN and soil nitrogen. Its main advantage is that the method provides
a "time-averaged" estimate of the proportion of legume nitrogen derived from atmospheric
N2 • This represents the integraI of any changes in N2-fixing activity that may occur during
the measurement period. The choice of reference RIant is the most important single factor
affecting the accuracy in estimating N2 fixation in 1 N-enrichment studies (e.g. Witty, 1983;
see also Chalk 1985).

G.3.2 Naturall~ abundance

The small difference in the naturaI lSN enrichment of soil nitrogen compared with
atmospheric N 2 - has been utilised to estimate N2 fixation by isotope-dilution (reviewed by
Shearer and Kohl, 1986). Although the principles of the technique are the same as those of
lSN-enrichment studies, the main limitations are quite different (Ledgard and Peoples, 1988).
A mass spectrometer capable of precisely measuring differences of 0.000037 atom % lSN is
1 necessary and sample preparation requires great care in order to avoid isotopic discrimination.
, The procedure is unreliable where the lSN abundance of soil mineraI nitrogen is low and
, variable (e.g. in undisturbed naturaI ecosystems, Hansen and Pate, 1987); however, this is
not often a problem in agricultural soils and similar estimates of N;z fixation have been
obtained with either 15N-enriched or natural abundance techniques (Evans et al., 1987). For
further examples see Bergersen et al. (1985) and Rerkasem et al. (1988).

G.4 ACETYLENE REDUCTION ASSAY (ARA)

The ARA is a useful diagnostic tool for detecting nitrogenase activity and has been widely
use<! in all areas of N2 fixation research; however, it is now considered to be of limited use
for even comparative studies with legumes. Because ARA provides only an instantaneous
measure, its accuracy bas always been restricted by the requirement for many repeated
determinations to adjust for marked diurnal and seasonal changes in N2-fixing activity. But
further errors in field estimations with legumes may arise from the use of inappropriate
calibration factors to relate ethylene production to N2 fixation, incomplete recovery,
disturbance and damage of nodules, or from an acetylene-induced decline in nitrogenase
activity during assay (see WittY and Minchin, 1988).

G.S UREIDE MEmOD

The xylem sap leaving the roots of a field-grown legume carries nitrogen-compounds to the
shoot originating from (i) nodules as assimilation-products ofN2 fixation, and (ii) soil mineraI
nitrogen taken up by the roots. In many tropicallegumes (see Table G.l) such as soybean
and cowpea the xylem solutes exported from nodules (the ureides, allantoin and allantoic acid)

165
'ISHF: A Handhook of Mel1WlII

ale very different from thosç transported t'Mm tOQt5 (amino acids and nitrate) (sec Ledgard
and People!, 1988). The abundance of UIeldes relative to the other nitrogen components in
a xylem sample will be indicative of plant dependenoo upon N 1 fixation for growth (e.a. Pate
et al.• 1980; Herridge, 1984). The ureide Merbod lwi many advantages over ather more
wmplex procedures. Xylem samples CM bç ea&ily mcovered from field-grown legume$ t,ly
app1ying a mild vacuum ra the stem. the three nitrogen component8 of xylem sap (i.e.
uœides, amino aeids and nitrate) can he measured separarely by Jimple colorimetrie a.nays,
and the ureille ÇQmposition is not affectOO by time of srunpling during !:he day (Herridge el
al., 1988). The ma.jor diwlvantage of the procedure is that ir provides only shon-terM
estimat.es of a legume's relianc.e upon N~ fixation. If seasonal determinarions are required
rather than comparative mcasures of rrea.rment effects, repea.t.ed xylem collections must be
taken. Wheœ estimates of seasonal N 2 fixation have been made. reasonable agreement
between the ureide method and other techniques have becn found (e.g. Herridge et al., 1984;
Rerkasem et al. 1988). See aIso Betts and Herridge (1987) for further examples of usc.

G.ri XYLEM SOLUTE TECHNIQUE (UREIOE) FOR NlTKOGEN FIXATION DY

LEGUMES

This pmœdure il suitable for usetiÏng the N2-fixing status of sorne tropical leaumel (see
Table 0.1 and the discussion in Section 0.5). It bas been used in breeding programs to
identify Unes with enhanced N 2 fixation in high nitrate soils and ta quantify inputs of fixed
Nin cropping systems (Peoples and Herridge. 1990). The technique requires the collection
and colorimetrie analysis of }l;ylem sap for specifie N-transpon solutes dçrived from the
process of N l fixation in the nodule (thç uœides t allanrain and allantoic acid), and the uptake
and usimilation of ~H minçraJ. N (nil:nlte and amino acids). Plant reliançc upon Nz tixarlon
for growrh (PNJ is determined by relating xylem sap oomposition to elasshou~ prepou-ed
ealibration curves.

G.6.1 Sampling of N-solutes

It is possiblç ta colleet xylem exuda.te as in bleeds spontançously fmm intact root stumps of
crop legumes following decapitation of the shoot from pIanu growing in the field in the
humid tropics. In other harsher envimnments the }l;ylem contents an be reoovered from
detached stems Qr branch segments from woody legumes by mild vacuum. There appean to
be only slight differenççs between the N-solute compositions of root-bleWing exudates and
vacuu m-e}l;tracted saps.

Usually 4 replicates of 4 - 10 plants will be u5ed for sap collections. The plant material can
be retained to provide a measure of crop N. Only 1 or 2 sap collections may be required
(e.g. during flowering or early seed-filling) to compare the effect of treatments (e.g.
inOCUlation or N-feniliser) on N;z fixation, but repeated samplings (4 - 8 times) during plant
gmwth will be required to determine seasonal inputs of fi}l;w N.

Procedure for root-bleeding exudale

1 Cut the shoot close to ground level with a sharp blade or pair of secateurs.

166
 Appendix G: TechniqwesJor QIImIliJying Nirrogen FüuJtIon

2 Place a silicon or latex rubber tubing sleeVe with an internal diameter slighUy smaller
than the stem over the exposed root stump.

3 Collect the sap exuded into the tubing after 10-20 min using a Pasteur pipette or
syringe.

Procedure for vacuum-extracted exudate

1 Detach a stem or branch segment (> 3 mm diameter).

2 Immediately insert the detached stem into a close fitting rubber sleeve connected to
a syringe needle (19 or 20 gauge).

3 Push the needle through the rubber stopper of a 5 ml Vacutainer (commonly used for
collecting blood samples; &cton Dickinson, Rusherford, USA) which bas been linked
to a vacuum pump via another syringe needle connection and a flexible plastic-tubing
line. The source of vacuum may be a hand-held vacuum pump (Nalgene, Sybron
Corp, USA), or consist of a laboratory vacuum pump powered by a portable petrol-
run generator or car battery if an electricity supply is not available.

4 Apply p3.rtial vacuum and successively cut short segments (3 - 4 cm) of stem from the
top to the bottom of the shoot. Xylem sap will be collected within the Vacutainer.

Note: Sap collections should be restricted to between 9 am and 4 pm. Vacuum extraction
must œcur within 5 - 10 min of stem detachment (Herridge et al., 1988, Peoples et
al., 1989b).

G.6.1 Sap storaae

The N-solute composition of xylem sap is stable for only a few hours at temperatures above
25°C. Sap samples should be placed in an ice-box immediately after collection and remain
chilled until placed in a freezer for long-term storage. If this is not possible, samples may
he stabilised by adding an equal volume of absolute ethanol (Herridge et al., 1988).

G.6.3 AnalySfS

Apart from the specific chemica1s and acids requircd for the three assays, the following items
will be required before commencing N-solute analyses:

• A weighing balance, preferably accurate to 0.1 mg

• Test-tubes (e.g. 150 x 18 mm), racks, and a boiling-water bath
• Micropipettes and/or dispensers to cover the range 0.01 - 0.2 ml and 0.5 - 5 ml
• Vortex mixer and cold-water ice bath
• Spectrophotometer or colorimeter.

167
TSBF: Â HœuIlHH1Ic 4f Mtthod.t

Table 0.1 (JcQurrenœ of ureides in xylem sap of nodulated legumes.

Speeies in whicb Species in which ureides Species in whieh ureides

ureides Ate màJor have been dete«e<1 as a have Dot been detected
componenrs of solute N' minor CQrnponentb
AlbÎ2Îll lopbantba IIlbizja falcalllria &(I(;;a alUla
CajQIUIS coJll1t lJoss(Jjcw uquifoUum au1'Îculllf"n-tlis
Calnpogmùum .:aeruleum ErythrilllJ, val'iegdtt) aUt1J"
Cl1ilàriOf:ulyx gyroirks Flemingill O'lnguta insauvis
Cyamopsls tl!trllg()noloba GliricÎdia septum pu/cbella
Desmodium discolor Pisum arveMe Aroebis tJypogaea
re1fSonii Sesbll1tJ.a IWtrata Bauhinia spp.
unclrumtm ,'fesbml CQe"5aJpinla
Glycine mox sry/()scmthes barnmn cmolhyrsus
HllnU1fbergia spp. Vicia ervtlia Calliandra spp.
Lublub purpU1Y!UJ Sâliva Cicer llrittinum
Macroplilium UlropurpulYl#n VlmllUlI'ill funçea Çliforia spp.
Mucrorylomo unijlOrUffl Derris elliptica
~raria javanica Juncea spp.
phnseoloides Lllthyrus cicera
Psopbocarpus letragonolobus sativus
Tedbegi spp. LeUClltlU'l spp.
Vigna angularls Lens culi1lllris
mung() MIUS WmiCu111lU9
radiUla Lupi1lUJ albus
tri/ohu angusdfolîur
&mguicuIdtQ cosentinii
V~ia sublerr~â mwabllis
Medicago minima
saliva
Mimosa pigra
Sesbania grmtdt/lora
StWferruneum
repens
Vicia mlJ1Illntha
fa/}a
Zomia spp.
a 40% or more of total N of xylem sap estimated t() be in ureides
b 10-25% of total N of xylem sap collected from glasshouse-grown or field plants
estimated to be in ureides.

A volume of > 0.2 ml of sap will he required for the three analyses (0.02 - 0.5 ml of sap
is used for each assay). Usually 0.5 - 1 ml might he expectelf to be collected from mature
field-grown soybean or pigeonpea, but it is often more difficult to collect sap from mher

168
 Appendix G: Techniques for QIUJ1IliJying Mrrogen raanon

species such as cowpea and mung bean and larger numhers of plants may he required to
obtain sufficient sap for analysis.

(a) Total ureides are measured by the method of Young and Conway (1942)
(b) Amino acids are determined with ninhydrin (Yemm and Cocking, 1955) using a
50:50, asparagine:glutamine standard
(c) Nitrate is measured with salieylie acid (Cataldo et al., 1975).

Full details of the assays and the use of the ureide method are described in Peoples et al.
(1989a), copies of which may be obtained by writing to ACIAR (GPO Box 1571, Canberra,
ACT 2601, Australia).

Calculations

The ureide content of a sample is expressed in terms of a proportional value of ratio whereby
the amount of N present as ureide (4 atoms per molecu1e) is related to total sap N.
Caleulation of a relative ureide index (RU!) from the individual solute concentrations ([ ]) is
determined as:

RUI (%) = {100 (4 x [ureideD} / {(4 x [ureide]) + [amino acid] + [nitrate]}

Experiments calibrating relative ureide index against lSN-derived estimates ofN2 fixation have
been undertaken for severa11egume species. However, it appears that many ureide-exporting
species have similar relationships to soybean (Peoples and Herridge, 1990), therefore only
calibration funetions prepared for soybean (Herridge and Peoples, 1990) will be presented
below.

In soybean (and pigeonpea; Peoples et al., 1989b) there were discernible effects of plant age
on N-solute compositions resulting in two sets of calibrations - one for vegetative and
flowering plants, and another during pod-fill. For each sap collection method:

Root-bleeding exudale

The funetions describing the relationship between PNl (the proportion of plant N derived
from N 2 fixation) and the relative abundance of ureides (x) are:

PN2 (%) = 1.2 (x - 5) for vegetative and flowering stages

PN2 (%) = 1.5 (x - 21) for plants during pod-fill
Vacuum-extracted exudate

PNl (%) = 1.6 (x - 8) for vegetative and flowering stages

PN2 ('J6) = 1.6 (x - 16) for plants during pod-fill

Measurements of N 2 fixation by field-grown legumes using such glasshouse-derived

calibrations have been shown to he very similar to estimates calculated from sophisticated lSN
techniques (e.g. Herridge et al., 1990).

169
TSBJ?: A HlI1Idbwk '1 jflt.'flhll

References

Bergersen, F.J., Tumer, G.L. Gault, R.R., Chase. D.L. and Bergersen, F.I. (1985) The
naturaI abundance of 1.5N in an irrigated soybean crop and its use for the eaIculation
of nitrogen tùation. Ausl1'iÙitm J<Junuù of Agrlcultural Rtsearr;h 36, 411-423.
Betb, I.H. and Herridge, D.H. (987) Isolation of soybean linescapable of nO<lulation and
nitro,en fixation under high levels of nitrate supply. Crop Science 27, 1156-1161.
&ddey, R.M. (1987) Methods for quantification of nitrogentixation associated with
Gramineae. CRe Critical Reviews in Plant Sciences Ci, 209-266.
Boddey, R.M., Chalk, P.M., Victoria, R.L. and Matsui, E. (1984) Nitro&oo fixation by
nodulaœd soybean undâ' wpiçal field conditions Cltimated by the nN isotope dilution
tecltnique. Soil Biology tutd Bi()Chemùtry 16, 583-588.
Cataldo, D.A., Haroon. M., Schrader, L.B. and Youngs, V.L. (1975) Rapid colorimetrie
detenninations of nitrate in plant tissue by nitration of salicylic acid. Conununicatio1l8
01 Soli Science tmd Plant Analysi.r 6. 71-80.
Chalk, P.M. (1985) Review. Estimation ofN2 fixation by isotopedilution: An tlppraisal of
tecluliqUQ involving "N enrichmcot and their application. Soil Biol<Jty and
Bioclu!mîstry 17, 389-410.
Danso. S.K.A. (1988) The UR of I.5N enriched fertilizeri for estimating nitrogen fixation in
grain Md pasture leeumes. In~ Bççk. D.P. and MaterOn, L.A. (cds.), Nitrogen
Fixation by Legumes in Mediterranean Agriculture. Martinus Nijhoff, Dordrecht,
Netherlands, pp.345-458.
Evans, J., Turner. G.L., O'Connor, G.E. md Bergerse.n, F.I. (1987) Nitrogen fixation md
accretion of soil nitroeen by field-grown lupins (Lupins lUIguslifolill.'i). Field Crops
Research 16, 309-322.
Hansen, A. and Pate, J.S. (1987) Evalua.tion of the IJN naturalabundance methoo andxylem
sap analY!Î$ for tlssessing N1 fixation of understorey legumes in jarmh (Euelllyptus
I1UUgi1l/Jla Donn ex Sm) forest in SW AU$ttalia. Journal 01 Experimental Borany 38,
1446-1458.
Hauck, R,D. and weaver, R.W. (1986) Field measurernenl of dinitrogen fixation and
denitrification. SSSA Special Publicati<JN No. 18. Soil Scienee Society of America,
Madison, Wisconnn.
Herridge, D.F. (1984) Effoots of nitrate md plant developmenton the abundance of
nitrogenous solutes in root-bleedine and vacuum-extracted exudates in soybean. Crop
Sdtnct 25, 173-179.
Herridge. D.F. and Peoples, M.B. (19'Xl) Urcide UAy for rneasuring nitrO&én fixation by
nodulated soybeAn ca1ibrated by I.5N methods. Plant Physiology 93, 495-503.
Herridge, D.F •• Roughley, R.J. and Brockwell, J. (1984) Effects ofrhizobia and soH nitra.te
on the establishment and functioning of the soybean symbiOSis in the field. Ausrralian
Journal of AgricuituraJ Research 35, 149-161.
Herridge, D.F., O'ConnelI. P. md Donne11y, K. (1988) The xylem ureideassay ofnitrogen
fixation; Sampling procedures and sources of error. Journal of Etperi17U!ntaJ Botany
39, 12-22.
Herridge, D.F., Dergersen. F.J. -and PeopleJ, M.B. (1990) Measurement of nitrogen
fixation by soybean in the field using the ureide and natural UN abundance methods.
Plant PhYliology 93, 708-716.
LaRue, T.A. and Patterson, T.G. (1981) How much nitrogen do legumesfix? Advancts in
Agr01wmy 34, 15-38.

170
l
Appendù G: Techniques for QuanliJying Nilrogen Fixalion

Ledgard, S.F. and Peoples, M.B. (1988) Measurement of nitrogen fixation in the field. In:
Wilson, J.R. (ed.), Advances in Nitrogen Cycling in Agricultural Ecosystems. CAB
International, Wallingford, UK, pp.351-367.
Peoples, M.B. and Herridge, D.F. (1990) Nitrogen fixation by legumes in tropical and
subtropical agriculture. Advances in Agronomy 44, 155-223.
Peoples, M.B., Faizah, A.W., Rerkasem, B. and Herridge, D.F. (1989a) Methods for
Evaluating Nitrogen Fixation by Nodulated Legumes in the Field. ACIAR Monograph
No 11, ACIAR, Canberra. 76pp.
Peoples, M.B., Hebb, D.M., Gibson, A.H. and Herridge, D.F. (1989b) Development of
the xylem ureide assay for the measurement of nitrogen fixation by pigeonpea
(Cajanus cajan (L.) Millsp.). Journal of Experimental Botany 40, 535-542.
Rennie, R.J. and Rennie, D.A. (1983) Techniques for quantifyingN2 fixation in association
with non-Iegumes under field and greenhouse conditions. Canadian Journal of
Microbiology 29, 1022-1035.
Rerkasem, B., Rerkasem, K., Peoples, M.B., Herridge, D.F. and Bergersen, F.J. (1988)
Measurement of N2 fixation in maize (Zea mays L.) - ricebean (Vigna umbellata
[Thumb] Ohwi Ohashi) intercrops. Plant and Soi/lOS, 125-135.
Shearer, G. and Kohl, D.H. (1986) Review. N2-fixation in field settings: Estimations based
on natural ISN abundance. Australian Journal of Plant Physiology 13, 699-756.
Witty, J.F. (1983) Estimating N2-fixation in the field using ISN-labelled fertiliser: sorne
problems and solutions. Soil Biology and Biochemistry 15, 631-639.
Witty, J.F. and Minchin, F.R. (1988) Measurement of nitrogen fixation by the acetylene
reduction assay; myths and mysteries. In: Beek, D.P. and Materon, L.A. (eds.),
Nitrogen Fixation by Legumes in Medite"anean Agriculture. Martinus Nijhoff,
Dordrecht, Netherlands, pp.331-344.
Yemm, E.W. and Cocking, B.F. (1955) The determination of amino acids with ninhydrin.
Analys180, 209-213.
Young, E.G. and Conway, C.F. (1942) On the estimation of allantoin by the rimini-
schryver reaction. Journal of Biological Chemisl1"y 142, 839-853.

171
Appendix fi MOST PROBABLE NUMBER COUNTS OF RHIZOBIA
IN SOILS
Paul Woomer

B.l BACKGROUND

An estimation of the total viable rhimbia in soits and other test substrates C&n he obtained
through application of the mosr-probablc-nurnber rechnique (MPN). This technique is used
to estirnate rnicmhiAI population sizes when di.m;t quantinuive MJeSStnent of individual celIs
is not pouible.

A sui table lei ume hast is fuosr eultuted on rhimbia-free mMia, Ulen a test substrate is serially
diluted and applied to the mot systems of the legume hosts. Alter 21 days the pattern of
presence and absence of root nodules is recorded, and these data used to derive a population
cstimare of the original test sub$tl'lte.
The MPN tOOhnique ia ba8ed on the mathematical approaches of Halvorson and Zcigler
(1933). Cochmn (1950) later identified estimation of crror and Mean aeparation proce<lurea.
Computer pqrarns are now available which combine thcse operations resulting in grearer
flexibility or experimental designs Md seÎcntific procedures (Woomer etaI- 1990). Tests
J

or experimental technique that allDw œsearchers ro either acœpt or ~~t experimental results
are also available (Halvorson and Moeglein, 1940; Woodward, 1957; Stevens, 1958; deMan,
1975; Scott and Porter, 1986; Woomer el al., 198&).

The necd for the application of rhizobial inoculants and the magnitude of the response 10
applied rhiwbia are determined by dIe spedes diversity and POPulation sires of indigenous
soit rhizobia and the avaUability of mineraI nitrogen within the soH system (Theis et al,.
1991). The legume/rhimbia .symbiosis is characterised by specificity œtwecn the host and
microsymbiont; thesc are identifiec:t as cross inoculation groups (FAO, 1984). The selection
of a. range of legumes characteristic of individua.l cross-inoculation groups as hasts for MPN
assays allows for the species composition and population sizes to he characterisc:d (Woomer
et al., 1988b). The lack ofrhizobia within a soit a110ws for the compan:son of inoculated and
uninoculated treatments as a direct measure of biological nitrogen fixation using the nitrogen
difference method (sec Appendix 0).

H.l EXPERIMENTATION

B.2.1 Exper1menhd dai&D

The design of a MPN determination involves the selection of the base dilution ratio (A); the
selection of the number of units at ea.ch dilution level (N); the determination of the need of

172
 Appendb: H: Mosl Probable NlI1IIher COUlUS of Rhizobia in Soils

an initial dilution and its value (1) and the selection of the inoculation volume (V). In this
procedure, a five-fold dilution series is inoculated onto the root system of 4 replicate plants
per dilution level with 1 ml per experimental test unit.

H.l.l Materials

The following materials are required to perform the MPN determination illustrated in Figure
H.1.

1. Stock solution: 0.125 g ~HP04 and 0.05 g MgS04.7H20 in 1000 ml of distilled

water in a sterile 1 litre wide neck Erlenmeyer flask.
2. 100 g (dw) of fresh soil and 400 ml stock solution in a sterile 1 litre wide neck
Erlenmeyer flask to initiate the 1:5 dilution series.
3. 5 sterile 100 ml screw top Erlenmeyer flasks and stoppers each containing 20 ml of
sterile mineraI nu trient diluent.
4. 1 sterile 5 ml wide-mouth pipette.
5. 5 sterile 5 ml pipettes.
6. 1 sterile 1 ml pipette.
7. 1 growth pouch rack (Somesegaran and Hoben, 1985).
8. 28 growth pouches containing uniform nodulating legume seedlings.
9. 56 surface sterilised, pre-germinated legume seeds.

H.l.3 Plant preparation

ln MPN plant-infection procedures, nodulation of the legume host serves as a selective media
in determining the presence or absence of the rhizobia at a given dilution level. Therefore,
the choice of a specifie legume host greatly influences the experimental results. More
information on the specificity of legume hosts and the selection of these hosts as an indicator
of rhizobial species may he obtained from FAO (1984) and Woomer et al. (1988a). In this
procedure, the host plants are cultured in "growth pouches" (Weaver and Frederick, 1972;
Somesegaran and Hoben. 1985) available from Vaughan's Seed Company, 5300 Katrine
Ave .• Downers Grove, IL, USA (approximate oost: $301100). A technique for the culture
of small-seeded legumes in glass tubes is available in Brockwell (1963) and Brockwell et al.
(1975).

1 Surface sterilise seeds by emersion in a 1.5% solution of sodium hypochlorite (or an

equivelent concentration of household bleach) for 4 minutes followed by 5 rinses with
sterile distilled water. For" hard seeded" legumes, emersion in concentrated sulphuric
acid for 4-25 minutes followed by 8 rinses may be substituted as a means of
simultaneously surface sterilising and scarifying the seeds.

2 Transfer the seeds to a sterile germination vesse!.

3 U pon emergence of the radicàI. transfer 2 germinated seeds to the trough of growth
pouch containing 75-100 ml of sterile nitrogen-free nutrient solution (Woomer et al. ,
1988) or rhiwbia-free water, taking care that the radical is in partial contact with the

173
r.;BF: A HrDJtIbopk (Jf UIJM4I

papcr wick of the growth pouch. Do not allow the growth pouch wick to dry.
additional irrigation is required every 4-10 days. Culture host plants for 1 week prior
10 inoculation ln a clean gIushou~ or growth mom.

H.2.4 SoU samplingt preparation aud Horage prior to dilution

When collecting soit samples, care must be talen tu colleet soils with soil :sampling equipment
that is not contatninated. It is not necesS8l'Y to e1ean equipment between subumples but all
sampling equipment should be sprayed with alcohol and flame sterilise<! between different
sites or treatments. Soi1s should not be air dried prior 10 dilution, and are best stored under
refrigeration. Dilute and inoculate the soi! samples within 48 hours of collection. Prior to
dilution, mmove stones. gravel. large plant MOrs, MOt I10dules and non-dooompo~ orgatlic
reJidue$ by hand. Crush larger IW)U aggreaatQ but it lS not necessary ta sieve the soi! sinee
the soit disperses during the first dilution step.

H.2.! Prepamtion of th~ dilution uriœ, root ÏDMulation and obHcvatioDS of nodulation

Procedure

1 Plaœ 100 g (dw) soU into a large mouthed jar or flask containing 400 ml of sterile
diluent. Low concentrations of mineral salts are use<! as diluents instead of distilled
water in order to avoid differenœs in osmotic potential across UJ.e dilution series.

2 Mi~ the soit and diluent using a wrist action shaker set at about 100 Hz for 20-25
min. This is UJ.e finit dilution srep.

Tratlifer 5 ml of llie first dilution srcp te) a 100 ml sterile screw top Erlenmeyer flask:
containina the 20 ml of diluent using a sterile. wide mouthed pipette. The soi! must
not sertie to the bottom of the vesse1 before the aliquot is withdrawn. Close the test
tube and agitate using a vorte~ mixer, or by wrist action. This is the second dilution
step.
4 Continue the dilution series to 5.6 (Re Figure H.1) u~ing a sterile. narrow mouthed
pipette for each additional dilution step. Again. care must be taken to avoid
sedimentation prior to the withdrawal of the tmnsferred aliquots. These are me thir«
through sixth dilution steps.

Apply 1.0 ml of each dilution step directly to the legume root sy~tems within the
growth l'Ouches using a sterile 1.0 ml pipette. A single pipette cao be used for the
eotire dilution series of each test substrate if test units arc inoculated in order from the
highest (5-6) to the lowest (5"') dilution, and the pipette rinsed twice prior to the.
transfer of the 1.0 ml inoculant to the culture units by drawing and releasing the
dilution back into the vessel into and from the pipette before inoculation at each new
dilution level.

174
 Appendù: H: Most Probable Number COIUIIS of Rhizobia in Soils

6 Place growth pouches in a clean growth chamber or glasshouse. Check for the
appearance of root nodules daily. In general, plant infection counts yield results
between 14 and 21 days.

7 Record data by regularly examining the root systems for the presence of root nodules.
Culture units that yield negative results should not be immediately discarded, but
should continue to be checked daily until 28 days after inoculation to insure that no
delayed nodule initiation takes place.

100 9 soil (dw) 400 ml diluent

dilution step

Figure H.1 Diagrammatic procedure of an most-probable-number plant-infection count with

6 fivefold dilution steps and 4 units per dilution level.

H.3 ASSIGNING TABULAR POPYLATlON ESTIMATES TO RESULTS

The population estimates are assigned by locating the experimental results on Table H.1.
When the correct code is located, the researcher obtains a population estimate of the original
test substrate directly from the table. In order to develop a population estimate for a fivefold
seriai dilution with 4 replicates and experimental results of 4-4-3-2-0-0, the operator searches

175
llJBF. ,{ lIansIbool qf Mcdwth

Table H.l The most-probable-number md confidence interva1s Cor a six stq), fivefold
dilution series with 4 test units per dilution leve1 each inoculated with 1.0 ml.

PoliûYc At MPN Conndeuce roliûYc At MPN Conn<lence

dilution eaÎllWe ÎIltorYlI dilution ad_ ÎIlttl"YlI
12~4"'i (COI18IV (p - 0.03) 12'4$' (ccl18lV (p .. 0.05)

010000 1.0 0.4-2.9 441110 138.6 4a.l.lt9.1

100000 LI O,4-3.l 442200 141.'7 49,1-491.'
101000 l.3 0.8-6.5 44~000 Ill.1 42,0-349.1
110000 7.3 O.U.' 443010 159.8 33.4-460.8
200000 l.6 0.9-7.4 44~100 1~.7 '7.1-474.9
lO1000 3.9 1.4-11.2 443110 211.4 '7~.3-M9.$
12.0000 3.' L'l-l0.2 44'100 lll.4 75.8-6'19.7
210000 4.0 1.4-11.' 443210 1'7~.0 94.7-187.2
211000 S.4 L9-U.6 443300 283.4 911.~.II'7.t
220000 3.3 1.9-16.0 444000 202.6 70.J-~84.0
300000 4.6 1.6-13.2 444010 '27'2.4 94.3-78$.4
~01000 fi.J '1.'1.18.1 444100 ~7.1 99.6-'187.6
310000 lB 1.'-19.0 444101 37S.0 llO.UMLO
311000 •. 4 l.P-;M,3 444110 ~'79.9 l:U.J-'~.1
~2.0000 1.7 3.0-25.1 444100 4W.l 141.9-1119.5
~lIOOO 10.9 ~.a.~I.~ 444101 534.9 185.3.U4l.9
330000 11.l 3.9-32.6 444110 49$.1 I7l.0-1429.3
400000 8.0 1.11-13.' 444110 S44.6 ISS.9.U'70.1
401000 10.8 3.8·31.2 444211 698.4 242.3-2013.S
410000 IL4 4.0-33.0 444'120 '711.0 249.0-Z069.8
4 1 1000 IS.l 3.2-4'.$ 444300 610.S 11l.1-1 '7dO. 1
42.0000 16.2 5.6-46.8 444'01 110.7 281.2-2337.'1
412000 19,~ 6.&-'fi,~ 444310 830.2 1074.4-2J9J.~
411000 2l.S '7.3-61.1 444311 1074.4 3'72.. '7.lO9'7.1
421100 1'7.7 9.6-79.11 444370 IIB.4 ~I(j.l-3;w9.&
411000 211.3 9.1J,.8U 444ill !J&ti.1l 481.1-3998.0
430000 24.2 A.4-tSt.1 444330 1445.4 ~1.4-4I&U
430100 R~ 11.1-9;&.0 444400 10~3.:1 )59.1-l914.6
4:11000 3l.1 Il.4.94.6 444401 ln6.8 48 Ll-3998.0
431100 42.2 14.6-111.' 444410 1464.9 301.1.411:1.'
4~1000 43.fi 15.1.125.6 444411 Im.l 691.1·5743.4
43'1100 '4.4 11I.9-1~6.8 444410 210S1.4 73 I.HiOBL 1
433000 56.3 19.6-163.0 44441'1 2.5'711.1 894.3·7432.6
440000 40.4 14.0-116.$ 444421 2812.6 9'7'.6-8108.4
440100 54.2 11I.1.136.l 4444ll 3750.1 1300.8.101110.9
441000 57.1 19.8-164. '7 444430 32SL3 1138.2.9459. '7
441010 74.6 :M.9-2 15.0 444 4 ~ 1 4H).O "17.6-12612.6
441100 75.5 2.6.2..11'7.6 444432 S'T]7.6 'Z039.6- t '711 '7.3
442000 Al.) ~.~2J4.3 444433 TSM.I lt!QU-2.162.1.7
44101Q 106.0 36.IJ,.303.3 444440 'VoH.ti 2059.6-17117.3
4411.00 911.1 '4,O-l'~.1 444441 8750.0 ~3.U.-1U13.:l
441100 107.9 37.4·311.0 444441 16~O.1 ~J6.B-46846.7
444443 lMOO.O 6937.,-$7637.3

the first column for those experimental results and then obrains the population estimate from
the corresponding adjacent column (218 cellslg soil).

The population eSlÎmatc$ in Table H.I a58Ume that the plant roots of each experimental unit
were inoculaœd with 1.0 ml of the dUuted tell substrate and that there was no pre-dilution
of the test substrate prior to the seriaI dilution. If an inÇlÇulant volume other than 1.0 ml was
applied to the culture uuits. the population estimate must be divided by the inoculation

176
 Apperulix H: Most Probable Number COIllllS of RltizobilJ iII Solls

volume. For example, if 2.5 ml was applied, the tabular population estimate is divided by
2.5. If an pre-dilution was prepared the tabular population estimate is multiplied by that
dilution value. For example, if a test substrate is diluted 100-fold prior to the fivefold
dilution series and the pre-dilution was used as the source of the lowest dilution, the tabular
value obtained is multiplied by 100.

H.3.l Constructing confidence limits

The confidence limits are constructed by dividing and multiplying the population estimate by
a confidence factor (Cochran, 1950). The confidence limits (p = 0.05) of the MPN estimates
are given in Table H.1. For the fivefold dilution series with 4 replicates, the confidence
factor is 2.88 (p=0.05). The experimental results 4-4-3-2-0-0 yield a tabular MPN estimate
of218, the lower confidence limit (p=0.05) equals 218 / 2.88, or 76. The upper limit is 218
x 2.88 or 627. These results May be expressed as 218 (76-627, p=0.05).

References

Brockwell, J. (1963) Accuracy of a plant-infection technique for counting populations of

Rhizobium tri/olii. Applied Microbiology 11, 377-383.
Brockwell, J., Diatloff, A. Garcia, A.L. and Robinson, A.C. (1975) Use of wild soybean
(Glycine ussuriensis Regel and Maack) as a test plant in dilution-nodulation frequency
tests for counting Rhizobium japonicum. Soil Biology and Biochemistry 7, 305-311.
Cochran, W.G. (1950) Estimation of bacterial densities by means of the "most probable
number". Biometries 6, 105-116.
deMan, J.C. (1975) The probability of Most probable numbers. European Journal of
Applied Microbiology 1, 67-78.
FAO (Food and Agriculture Organisation of the United Nations) (1984) Legume lnoculants
and Their Use. FAO, Rome.
Halvorson H.O and Moeglein, A. (1940) Application of statistics to problems in
bacteriology. V. The probability of occurrence of various experimental results.
Growth 4 1 157-168.
Halvorson, H.O. and Ziegler, N.R. (1933) Applications of statistics to problems in
bacteriology. I. A means of determining bacterial population by the dilution method.
Journal of Bacteriology 25, 101-121.
Somesegaran P. and Hoben, H.J. (1985) Methods in Legume-Rhizobium Technology.
University of Hawaii Ni ITAL Project and MIRCEN. Paia, Hawaii, USA.
Theis, J.E., Singleton, P.W. and Bohl~l, B.B. (1991) Influence of the size of indigenous
rhizobial populations on the establishment and symbiotic performance of introduced
rhizobia on field-grown legumes. Applied and Environmental Microbiology 57, 19-
28.
Weaver, R.W. and Frederick, L.R. (1972) A new technique for most-pI:obable-number
counts of rhizobia. Plant and Soil 36, 219-222.
Woomer, P., Singleton, P.W. and Bohlool, B.B. (1988a) Reliability of the most-probable-
number technique for enumerating rhizobia in tropical soils. Applied and
Environmental Microbiology 54, 1494-1497

177
Woomer, P, Singleton, P.W. and Bohlool. B.B. (1988b) Ecologica1 indicators of native
rhi.wbia in tropical soUs. Applied and EnVironmental Microbiology ~4, 1112-1116.
Woomer, P, Bennett. J. and Yost. R. (1990) Overooming the inflexibility of Most-Probab1e-
Number procedure!. Agrolk>ffly Journal 82, 349-353.

178
Appendix 1 THE STUDY OF PHOSPHORUS CYCLES IN
ECOSYSTEMS
Holm Tiessen

1.1 INTRODUCTION

Phosphorus (P) is one of the main limiting elements for crop production. In many natura!
ecosystems P availability limits overall ecosystem productivity through its effect on plant
production and nitrogen fixation. Plants take up P from the soil solution. While solution P
constitutes only a minute fraction of total soil P, it is constantly replenished by the hydrolysis
of labile inorganic P (Pi) or by mineralisation of organic P (po). These fractions, in tum,
are in exchange with more stable P compounds through slow reactions. Therefore, the P
supply to vegetation depends on the quantities of labile Pi, the rates of transformations
between labile and more stable Pi, and the size and rates of transformation of the
mineralisable Po pool in the soil (fiessen et al., 1984).

1.2 SOIL TESTS FOR INORGANIC P AVAILABILITY

Soil tests for plant available Pi need to include solution Pi and varying amounts of Pi
associated with the soils solid phase, and they must he simple enough for routine application.
This is achieved with single extractants designed to moderately lower or raise pH of the soil
solution, introduce anions that solubilise P by competing with P sorption sites (H ZC03 , S04)'
or by chelating (EDTA) cations or lowering the solubility of cations (P') that bind P in the
soil. Examples are the acid ammonium fluoride (Bray and Kunz, 1945), bicarbonate (Olsen
et al., 1954; Dabin, 1967), lactate (Egner et al., 1960), or chelating (Onken et al., 1980)
extracts. An alternative extractant, ion exchange resins, simply constitute a sink for solution
Pi and thereby drive soluble or "exchangeable" Pinto solution (Amer et al., 1955). To sorne
extent, Ute extraction of P by resin is affected by the saturating anion on the resin, and by the
presence of different cations in Ute soil solution (Cunin et al., 1987), Bicarbonate resin is
frequently used in temperate !iOils but a neutrnl anion such as cr is to he preferred in acid
soils with low buffering capacîty to avoid uncontrolled pH rises.

Although !iOme· extractants have been chosen to simulate the action of plant roots which
produce bicarbonate and various organic acids, in reality, the processes occurring in the
vicinity of plant roots are likely to be much more complex, involving microbes, mycorrhiza,
specific root exudates and pH control (L6pez-Hernandez and Flores-Aguilar, 1979). The
biennial pigeonpea (Cajanus cajan) may have special abilities to sequester Fe-P, that is
unavailable to other plants, by excretion of specialised chelating substances (Ae et al., 1990).

Chemica1 extractants clearly cannot mimic such complex processes. The sucœss of simple
P tests in giving 50-60% accurat:e predictions for fertiliser requirements in Many moderately

179
ISBF: JI Handbool oJ M,I"QI/I
wealhered soi1s depends on the extraedon of a quantity of soil P thar is somehow related 10
that portion of soil P which is plant available. The regression between these two quantities
is established over years of agronomie e~perimentation and r:esting of fertiliser responses for
specifie crops on denned soil types with appropriately matched extractants (Sharpley et 'II.,
198~). On weakly wearbered Ca-rich mils, acid extraetants (Bray and Kurtz, 1945: Nelson
et aL, 1953) are mismatched because they cao dissolve P t.ha.t is not plant available, and
would not he extraç~ by more alkaline ~lutions (Olsen el al.• 1954). In açid soUS.
correlations bçtween mild aoid and a.lkAline ~tIactants may be CIOJef (Sharpl~y el QI•• 1985),
but in acid tropical soi1s that show Iapid P transformations and substaDtial P sorption,
extraction methods for P fertiliry assessment have often becn unsuccessful, and fertiliser
responses have becn ermtic (Ayodele and Agboola, 1983).

Under tropical conditions, rapid mineralisation of organic marter from slashcd vegetation or
crop residues CM Hberate sufficient P for plant gtowth to obscure the role of inorganic P
extraeted by P tests (Adepetu and Corey, 1976; Agboola and Oko, 1976). In addition ta the
faulty assessment of available p, fertiliser responses may be aitered by strong sorption of P.
ln order to conect for such P N fixation n, the fnctional recovery of added fertilise!" by a P rest
(A~boola and Ayode1e, 1983), the awOWlt of P required to taise solution P (Van der Z1l.a&
el al.• 1979; Juo and Fox. 1977) or extractable P (Agboola and Ayodele, 1983) to sorne
chosen value. direct measures of P sorption (Roche, 1983; FOJl; illd Kamprath. 1970), Of
determination of P-sorbing Fe (Anyaduba and Adepetu, 1983; LeMare, 1981) or Al (Rœhe
et al. t 1983) CM be combined with P tests.

1.3 PHOSPHORUS "FIXATION"

Phosphorus "fixation" is, as any sorption process, reversible, and in many weathered soils,
Fe- and Al-phosphates are also a potential source of plant available P (Hughes, 1982; Hughes
and LeMare t 1982). LeMare (1982) indicated that particylarly organic matter - a.ssoçiated
Fe and Al May be involved in the reversible sorptiml of P. One of the problems in assessing
the reversibility of sorption is that sorbed P undergoes funher transformations with time
(Mattingly, 1975: Parfitt el al., 1989). Suçh proœsses may involve recrystallisation (Barrow,
1983). solid state diffusion (Willeu et al., 1988), or multiple P pools whieh are not in
immediate exchange with the solution (Fardeau and Jappe. 1980) or whiçh have differing
affinities to P (Kanabo el QI., 1978). A measurement of availab1e inorganie P therefore needs
to consider both the amounts and the rates of re1ease of P from the solid phase. Among the
approaches taken to evaiuate release rates are repcated water extracts, and desorptiml
isotherms (Bache and Williams, 1971; Fox and Ka.mprath, 1970). possibly at elevated
temperatures to substitute for impractica11y long reaetion times (Barrow and Shaw t 1975).

Isotopie dilution merbods cao be used to measure the sÎze of the P pool from whkh solution
P can be replenished OVe! a few houcs, which has becn used to define an exchangeable or
available P pool. A continuing disappearance of tracer from the soil so1ution alter the initial
rapîd isotopic dilution indicates the existence of a "sink." for 32p tha[ consists of sorbing
pha.ses whiçh are not in direct exchang~ with the soil solution (Fardeau and lappe, 1980;
Saloedo et al., 1991). This sink represents an a.ctivity of less soluble or slower pools of soH
P whiçh may replenish ava.ilable P at rates varying from days tQ ycars. Parfit[ el al. (1989)

180
 Appendix 1: Melhods for lM Slfu:ly of Phosporus Cycles in Ecosyslems

confirmed that the slow sorption reaction can reduce plant available P, but the reversibility
of this process has not been tested in relation to the P acquisition rates of plants.

1.4 ORGANIC PHOSPHORUS TRANSFORMATIONS

In addition to inorganic processes, the turnover of organic matter constantly releases Pinto
the solution from Po mineralisation. Through the mineralisation of Po, the pool size of "total
available" P is strongly time dependent. In natural ecosystems where Po recycling is an
integral part of plant P supply, plant available P is a functional concept rather than a
measurable quantity, and any fertiliser responses (for purposes such as pasture improvement)
would he erratic. Therefore available P must he defined with respect to the type of ecosystem
in which it is measured, taking into consideration the type of P sink, i.e. a plant, plant
community or crop, the period over which P supply is required, such as a cropping season,
year, or growth cycle, and the rates of recycling that occur during that period. The element
of time in the definition of plant available P has not received much attention in the past, since
efforts were concentrated on annuaI crops under arable agriculture. Agroforestry, alley
cropping, shifting cultivation, managed rotations and fallows, pastures, and the concept of
"sustainability" of agriculture, all require that understanding and analytical capabilities
develop heyond the "immediately available" nutrient pool.

I.S SEQUENTIAL EXTRACTIONS

A potentially available P pool, analogous to the mineralisable N or S pools measured with

incubation and leaching techniques (Ellert and Bettany, 1988) is not feasible, because of the
reactivity of solution P with the soil' s mineraI phase. This lead to the design of sequential
fractionations which first remove labile P, and then the more stable forms. The sequential
extraction of Chang and Jackson (1957) attempted to separate chemically identifiable forms
of Pi, but problems of re-precipitation of extracted P, and the similarity in the solubilities of
Fe and Al associated P have limited its success (Williams el al., 1967; Williams and Walker,
1969). Significant amounts of organic Pare extracted with the Chang and Jackson procedure,
but have reœived little attention, although they have been shown to be important in plant
nutrition (Kelley et al. 1983).
J

A sequence of alkaline followed by acid extracts, which forms part of the Chang and Jackson
scheme gives a reliable distinction between AI+Fe and Ca-associated Pi (Kurmies, 1972).

In most unweathered soils P occurs predominantly in the form of apatites, calcium phosphates
containing carbonate, fluoride, sulphate, hydroxide, and a number of different cations, or in
primary mineraIs where P substitutes isomorphously for silicate in crystal structures.
Weathering and leaching reduce total P contents, and cause the formation of secondary Fe and
Al phosphates with low solubilities. At the same time, the accumulation of organic matter
in the soil as a result of the establishment of vegetation is accompanied by the formation of
organic P (Po) pools. This distinction is useful because the balance between primary Ca-
associated and secondary Al + Fe Pi forms reflects the weathering stage of the soil and
indicates the presence of Ca rock phosphate fertiliser in weathered soils that otherwise contain
little Ca-bound P.

181

Il
'lNBF.. A HanJOool 0/ Mcrhocù

Using these concepts, Hedley el al. (l982a) attempted 10 quantify labile, immediately plant
available Pi, Ca-Pi, Fe+ Al-Pi, and labile and more stable forms of Po using the following
extracts:

Resin and bicarbonate extractable Pi are the labile fonns of Pi which are thought to consist
of Pi adsorbed on surfaces of more crystalline P compounds, sesquioxides or carbonates
(Mattingly, 1975). Hydroxide extmctable Pi hM lower (or Jlow~r) plant ilvaib.bility (Marks,
1977) and Îs thought ta he anoeiated with atnQrphQUS M.d some crystalline Al and Fe
phosphates. A more preçise characœrisation of these Pi fonns is usua11y not possible or
desirable since mixed compounds containing Ca, Al, Fe, P and other ions predominate in
soi1s (Sawhney, 1973). Labile bicarbonate extractable Po is easily mineralisable and
contributes 10 plant available P over one growing season (Bowman and Cole, 1978). More
stable forms of Po involved in the long-œrm transformations of Pin soils are extracted with
hydroxide (Batsula and Krivonosova, 1973). By comparing the P compositions of different
field soils, each of th~ extucts wu empirically tlssigned a role in natural P transformations
associated with microbial uptake (Hedley el al., 1982a.), plant MOH (Hedky ~t al., 1982b).
cultivation (Ties!;en et al., 1983), or soi! development (Tiessen et al.• 1984; Roberts el al .•
1985). The residual. un-extractable P in a number of West African ~ib, amounting ta 20 Md
60% of total Po' was positively corre1ated to both Fe and Al whieh are responsible for the
long-tenn stabilisation of P (fiessen et al.• 1991).

Sharpley el al. (1985) related the fractions obtained by this method to a number of different
P tests on 86 different soUs with representatives from most soil orders (Tiessen el al., 19&4).
Resin ex tractable labile P was relared ta Bmy 1 and Olsen P tests in çaIÇMCQus soils. but the
correlation decreased with increa.sing weathering stage of me soils. Conversely, the Mehlich
1 P test was a good predictor of la.bile P in hi&hly weathered soils but did Dot perform weIl
on younger soils (Sharpley el al., 1985).

Although the sequential fractionation extraets severa! Po pools, their nature is even less well
defined than that of the inorganic extracts (Stewart and Tiessen, 1987). The residue of the
original fraetionation of Hedley et al. (1982a) often contained significant amounts of organie
P that sometimes participated in relatively shon-tenn transfonnations despite its recalcitrance
with respect to ehemical extraetants. Particulate organic debris which may be rapid1y
mineralised by soi1 fauna and flora usually contains P that is not extractable. On relatively
young, Ca-dominated SQils this rçsidual organic and inorganic P can be extracted by a second
NaOH treatment following the acid extraction (Condron el al.. 1990). leaving negligible
amouDts of P in the final residue. On more weathered soils, hot HCI (Mètha el al .• 1954)
extracts most of the organie and inorganic residual P (Condron et al., 1990).

Despite the suœess of fractionating most of the soils P, it has been difficult to assign
chemica1ly extracted Po fractions specifie roles in soil P transformations, because Po turnover
frequently depends on the mineralisation of organic matter during which P is released as a
side product. Most field obsety'ations of P mineralisation are not e}l;plained by the
biochemica1 release of P, but by the biological turnover of organic matter with concomitant
P release (cf. MeGill and Cole, 1981). Correlations have been uS«l successfully to show the
dependencc of Pi turnover and stabilisation on soil mineraI components such as non-crysta1line
Al and Fe, but this appmach fails when relating labile P fractions ta organic P or organic

182
 Appendix 1: Merhods for rhe Srudy of Phosporus Cycles in Ecosysrems

matter (e.g. Ayodele and Agboola, 1983), since"P availability itself limits the accumulation
of organic matter in many soils (Tiessen and Stewart, 1983).

Microbes may play an important role in the short-term cycling and availability of P by
mechanisms such as uptake-storage-release, mineralisation and solubilisation (Lee et al.,
1990; Singh et al., 1989; Parfitt et al., 1989; McLaughlin et al., 1988; Asea et al., 1988),
and microbial P bas been determined as part of the Hedley et al. (1982a) fractionation. In
sorne ecosystems such as tropical rainforests on very infertile soils, closed plant to plant or
litter to plant recycling represent a major portion of the active P cycle (Medina and Cuevas,
1989), while any chemica1 soil fractionation entirely ignores the complexity of such biotic P
transformations. Few complete ecosystem analyses of P cycles incorporating complex
biologica1 compartments have been done (Chapin, 1978).

1.6 SPATIAL VARIABILITY

In addition to the problems associated with the description of soil P in any one sample, and
its transformations with time, the forms and transformations of P are highly variable in
landscapes, profIles or even within horizons. This variability cannot be ignored when soil
fertility, plant performance or nutrient cycling are to be evaluated. While field variability of
total soil P may be low (CV < 20 %), resin extractable P and other labile fractions are very
variable (CV > 50%) since they depend to a much greater extent on micro relief, soil
moi sture , clay content, plant uptake etc. (Tiessen and Santos, 1989). Mineralogical
differences between horizons will affect P sorption, with low sorption in organic matter rich
top soils and high sorption in sesquioxide-rich accumulation horizons (Stewart et al., 1990).
Similarly, the presence of highly sorbing lateritic materials in a soil profIle will greatly
modify P cycling (fiessen et al., 1991). Cultivation and erosion will translate such
differences into lat.era! variability at field scales.

1.7 CURRENT APPROACHES AND DIRECTIONS

Short-tenn fertility assessments of P status on cultivated soils are successfully performed in

many places with mild acid (Bray) or alkaline (OIson, Dabin) extractants. Extractions with
anion exchange resins are easier to interpret since they do not modify the chemical
composition of the soil solution. Results of such extractions cao therefore more easily be
compared across different soil types and ecosystems. Ion exchange resins are available on
polymer membranes whieh ~ be eut to sui table siz.e and used repeatedly for the extraction
of soil P. Results obtained with these membranes are close but not absolutely comparable to
results obtained with the more traditional resin beads. The use of resin membranes, though,
is mueh ea8ier because soil cao he washed off easily, membranes are easily piclœd up,
transferred from Hasks, re-extracted, and re-used so that now resin extraction has become
simple enough for routine applications. Experimentation is under way for the in situ use of
resin membranes in soils but this technology has not matured enough for wide application.
Simple extraction data cao be supplemented with an evaluation of P fixation by equilibrating
soil sarriples with different levels of P (Fox and Kamprath, 1970) to arrive at a reasonably
close prediction of P requirements under conditions of annuaI amble agriculture.

183
The st\KJy of P tranlfonnations in IDixed or perennial aaricultuIal systems or in native
ecosysrems poses much greatet probiewi. The questions addressed in such lludies include;
Is P limiting a.ççumulation or transfonnarions of C and m What wnstmints does P
availability impose on current Md futuR cwp production in continuous croppÎtlg Ot rolational
agriculture? Wbat are the moohanisms of regeneration of fertility under bush fallow? In each
case the entire complex of Pi and Po transformations including mineraI interactions and the
turnover of organîc matter needs ta be studied. P sorption, Fe and Al activities. and P
fractionationl followi.na the ProcWures of KurmieJ Of Hedley and co-workers have proved
to he successful in e1ucidating Pi chemistry and transformations. These methods can follow
the tale of Pi or fertiliscr P with liming, incubations, ewultive crapping or similar
controlled manipulations. Conelative studies with other soit propertîes, .Iong environ mental
gradients, with differenoos in cultivation Of soil deve10pment have helped the underuandmg
of the peyele and promisc further successes in new e«Jsysteml.

In ecosysrems where biotic P cycling through close recycling of plant litter is important, such
as in nutrient-limited tropical min foresrs, chemical P tranuonnations in the soU are of course
of limited relevanœ. In such ecosystems detailed analyses of e«Jsystem components such as
1eaf P, litter P, plant Pt decomposition rates, root growth and mot distribution and the role
of bacreria and futlai must be evaluated in order 10 undentand the p çyclç. Such studios are
romplicated by the unavailability of a 10llg-lived iootopic tracer for P. Only the use of
repeated e1çmental budlJets in different ee.osystem comPônœts aloflB seasons has provided
inslght into the P cycle in luch erosystems (Medina and CuevH, 1989).

Studios of the availability, cycling and chemical composition of Po suffer from even greater
methooological limitations. Po fmctionations are inc1uded in Hedley's approach but the
biological funcri,ons of the chemical1y extlactable fractions are not easUy defined. While these
fractions have changed predicrably under manipulations such as incubations, fenilisations,
manuring. exhaustive cropping, etc. an understanding of the role of Po in the environmental
P cycle must be based on an understandinlJ of organic matter transformations. Separation of
reo:nt litter,· flotable organic matter, clay aSlOCiated or minerai stabilised organic mattet. and
analySis of the as.sociated Pare methods that have shOwn sorne SU~S8. Short-rerm tumover
of Po Î!l large1y mictObially mediated and microbial P rneasureruents using one of the
chlorofonn methods are useful in this œntext. In highly P sorbing soUs. the P relcased by
ch1orofonning a microbial population is rapidly removed from extractable pools and the
chloroform methods become re1atively înaceurate and requiR large fudge factors.
Neverthelcss, microbiaI P measurcments cao he uood in smIDes comparing simi1ar wils and
ecosystems.

P cycles in soHs anô œtiR ewsystems depend on the chemi$try of Pi and the transformations
of oIJanic P and henœ of organic matter. Since P often limits organic matter transformations
quite comp1ex interrelationships exÎSt and eonceptual (Tiascn el al., 1984) or mathematlcaI
(Cole et al., 1977) models may help in the analyses of soil and ecosystem data and in the
planning of research approaches.

References

Adepetu, J.A. and Corcy. R.B. (1976) Organic phosphorous as a prcdictor of plant-available
phosphorus in soUs of southem Nigeria. Soil Science 122, 159-164.

184
 Appendix 1: Methods for rhe Srudy of Plwspor/Ui Cycle6 in EC06Y6rmu

Ae, N., Arihara, J., Okada, K., Yoshihara, T. and Johansen, C. (1990) Phosphorus uptake
by pigeonpea and its role in cropping systems of the Indian subcontinent. Science
248, 477-480.
Agboola, A.A. and Oko, B. (1976) An attempt to evaluate plant available P in western
Nigerian soils under shifting cultivation. Agronomy Journal 68, 798-801.
Agboola, A.A. and Ayodele, O.J. (1983) An attempt to evaluate plant available P in
western Nigeria savannah soUs under traditional fallow systems. IMPHOS 3rd
International Congress on Phosphorus Compounds, Brussels, pp 261-287.
Amer, F., Bouldin, D.R., Black, C.A. and Duke, F.R. (1955) Characterisation of soil
phosphorus by anion exchange resin adsorption and 32p equilibration. Plant and Soil
6, 391-408.
Anyaduba, E.T. and Adepetu, I.A. (1983) Predicting the phosphorus fertilisation need of
tropical soils: significance of the relationship between critical soil solution P,
requirement of cowpea, P sorption potential, and free iron content of the soil.
Beitrllge zur tropischen Landwirtschaft und Veterinllrmedizin 21, 21-30.
Asea, P. E. A. , Kucey, R. M. N. and Stewart, 1. W. B. (1988) Inorganic phosphate
solubilization by two Penicillium species in solution culture and soil. Soil Biology and
Biochemistry 20, 459-464.
Ayodele, 0.1. and Agboola, A.A. (1983) Evaluation of phosphorus in savannah soils of
western Nigeria under bush fallow systems. Journal ofAgricultural Science 101, 283-
289, Cambridge.
Bache, B.W. and Williams, E.G. (1971) A phosphate sorption index for soUs. Journal of
Soil Science 22, 289-301.
Barrow, N.J. and Shaw, T.C. (1975) The slow reactions between soil and anions. 2. Effect
of time and temperature on the decrease in phosphate concentration in the soil
solution. Soil Science 119, 167-177.
Barrow, N.J. (1983) On the reversibility of phosphate sorption by soils. Journal of SoU
Science 34, 751-758.
Batsula, A.A. and Krivonosova. G.M. (1973) Phosphorus in the humîc and fulvic acids of
sorne UWnian soils. Soviet Soil Science S, 347-350.
Bowman, R.A. and Cole, C.V. (1978) Transformations of organic phosphorus substances
in soils as evaluated by NaHCO) extraction. Soil Science 125, 49-54.
Bray, R.H. and Kurtz, L.T. (1945) Determination of total, organic and available fonus of
phosphorus in soils. Soil Science 59, 39-45.
Chang, s.e. and Jackson, M.L. (1957) Fractionation of soil phosphorus. Soil Science 84,
133-144.
Chapin, F.S., Barsdate, R.J. and Barèl, D. (1978) Phosphorus cyc1ing in Alaskan coastal
tundra: a hypothesis for the regulation in nutrient cycling. Di/cos 31, 189-199.
Condron, L.M, Moir, J.O., Tiessen, and Stewart, J.W.B. (1990) Critical evaluation of
rnethods for determining organic phosphorus in tropical soils. Soil Science Society of
America Journal 54, 1261-1266.
Curtin, D., Syers, J.K. and Smillie, G.W. (1987) The importance of exchangeable cations
and resin-sink characteristics in the release of soil phosphorus. Journal 01 Soil
Science 38, 711-716.
Dabin, B. (1967) Methode Olsen modifié. Cahiers ORSTOM-Pedologie 5-3.
Egnér, H., Riehrn, H. and Domingo, W.R. (1960) Untersuchungen aber die chemische
Bodenanalyse ais Grundlage ftJr die Beurteilung des Nllhrstoffiustandes der BiJden.

185
 2. Ch.emilChe EurQ/aio~thcclen ;[JIr PhasphQr- und Kaliwn/)eslimmung. Kungi.
Lan<lbrulŒhOgsJm1ans Annaler, Uppula. Swedcn, 16, l%1-21S.
J3tlert, B.H. and Bmany, J.R. (1~88) Comparlsons of kinetic modeLs for describing net
sulrur and nitrogen minerali.zation. Soil Science Society 01America Journal S2, 1692-
1700.
Fardeau, J.C. and Jappe, J. (1980) Choix de la fertiJ.iWjon phosphorique des sob
tropicaux: emploi du phosphore 32. A.kro1UJf1Ûe Tropicale 35, 225-231.
Fox, R.L. and Kamptath. B.J. (1970) Phosphare J«ption isotherm, for evaluat:i.ng the
phosphate; IeQ.wremenu of lOils. SOU SCienc:e Society (JI America Proceedings M,
9fJl-907.
Hedley, M.J., Stewart, J.W.B. and Chauhan. B.S. (1~82a) Changes in worgame and
organic sail PhQ5Phorus metions iDduced hy cultivation practïçg and by laboratory
incubations. Soi! Scitlk:e Socle" (JI A.JneriCIJ JoU17JilJ 46, 970-976.
Hedlcy, M.J., White, R.E. and Nye, P.H. (1982b) Plant-inducOO changes in the
rhiwsphere of tape seedlings. m. Changm in L-valuc, soU phQlphate fractions and
phoSPhta5e activity. New PhytolDgist '1.45-56.
Hughes. J.C. (1982) High gmdieftt milinetic ~patation of some soïl clays from Niaeriat
Bruil ~d Colomhia. 1. The j.nterrcl.tionships of Îfôn and aluminium ws.tracœd hy
acid ammonium oxalaœ and carbon. Journal of Sail Science JJ, 509-519.
Hughes, J.C. and Le Mare. P.H. (1982) High gmdient magnetic separation of some soil
clays hotn Nigeria, Brazlll.lld Colombil. n. Phosphate adlOrption chara.çteristics t the
it\ter-relationshîps with iron aluminiwn and carbon. and compariwn with whole lOi!
data. J~urnal 0/ Soit Science 33, 521-~JJ.
Juo, A.S.R. and Fox. R.L. (1977) Phosphare Sôrption characterïstics of some bench-mark
$Oils of west Africa. Soli Sctt"'e 124, 370-376.
Kanabo, LA.K •• Halm, A.T. and Obeng, H.B. (1978) pho$}.)horus adsorption by suriaee
HIllples of five ilVnpan soUs of Ohana. ~iJfk1fflQ 10, 299-306.
Kelley, J., LAmbert, M,J. and Turner, J. (1963) Available pnolphorus fonus in forest sails
and their poSSible eoologiea1 $ipificance. O>l1II1Wnlcllli(J1fS in Soil Seience tmd Plant
Analysis 14, 1217-1234.
Kunnies, B. (1972) Zur Fraktionierung der Bodenphosphate. Die PhQspoorsaure 2" 118-
1~1.
Le Mare, P.H. (1981) &changeable phosphorus, estimates of it from amorphou$ iron
oxides, and soil solution phosphorus, in relation ta pholphorus ra.ken up by rnaU:e.
Jou1'1Ill1 01 SoU Science li, 285-299.
Le Mare. P.H. (1982) Sorption of isotopically exchangeable il1d non-exchangeable
phosphate by SOlUé lOils of Colombia and Braztl, and compari$Qns wirh soUs of
southem Nigeria. lDUI'fUÙ of Soli ScttllCe 33, 691-707.
Lee, D., Han, X.O. and Jordan, C.F. (1990) Soil phosphorus fractions, aluminum, and
water retention as affected by microbial activity in an Ultisol. Plant and Soil1l1,
125-136.
Lôpez-Hernândez, D. and Flores-Aguilar, D. (1979) La De80rcion de fosfuros en suelos.
Implicaciones fisioecologiws en el proccso. Acta Ciem. Veneto/ana 30, 23-3~.
Marks, G. (1977) Beitrag zur praezisierten Chamkterisierung von Pflan~n verfuegbaren
Phosphat in Ackerboed.en. Arch. Acker Pflanzenbdu IJodenlai. 21, 447-456.
Mattingly, G.E.G. (1975) Labile phosphare in lOils. Soif Science 119, 369-J7~.
MeOill, W.B. and Cole, C.V. (1981) Comparative aspects of C, N, S and P cycling
through sail organic matter during pOOogenesis. Geoderma 16, 267-286.

186
 Appendix [: Melhodr for ~ Sltuly of Plwsporllll Cycles ln ECOSYSlems

McLaughlin, M.I., Alston, A.M. and Martin, J.K. (1988) Phosphorus cycling in wheat-
pasture rotations. II. The role of the microbial biomass in phosphorus cycling.
Austra/ian Journal of Soil Research 26, 333-342.
. Medina, E. and Cuevas, E. (1989) Patterns of nutrient accumulation and release in
Amazonian forests of the upper Rio Negro Basin. In: Proctor, J. (ed.), Mineral
Nutrients in Tropicale Forest and Savanna Ecosystems, Blackwell Scientific, Oxford,
UK.
Metha, N.C., Legg, J.O., Goring, C.A.I. and Black, C.A. (1954) Determination of organic
phosphorus in soils: I. Extraction method. Soil Science Society of America
Proceedings 18, 443-449.
Nelson, W .L., Mehlich, A. and Winters, E. (1953) The development, evaluation, and use
. of soil tests for phosphorus availability.· In: Pierre, W.H. and Norman, A.G. (eds.),
Soil Fertilizer Phosphorus. Agronomy 4, 153-188. American Society of Agronomy
Madison, WI, USA.
Olsen, S.R., Cole, C.V., Watanabe, F.S. and Dean, L.A. (1954) Estimation of available
phosphorus in soUs by extraction with sodium bicarbonate. U.S. Department of
Agriculture Circular 939.
Onken, A.B., Matheson, R. and Williams, E.J. (1980) Evaluation of EDTA-extractable P
as a soil test procedure. Soil Science Society of America Journal 44, 783-786.
Parfitt, R.L., Hume, L.I. and Sparling, G.P. (1989) Loss of availability of phosphate in
New Zealand soils. Journal of Soil Science 40, 371-382.
Roberts, T.L., Stewart, I.W.B. and Bettany, I.R. (1985) The influence of topography on
the distribution of organic and inorganic soil phosphorus across a narrow
environmental gradient. Canadian Journal of Soil Science 65, 651-665.
Roche, P. (1983) Les méthodes d'appréciation du statut phosphorique des sols. Leur
application à l'estimation des besoins en engrais phosphatés. IMPHOS 3rd
International Congress on Phosphorus Compounds, Brussels, pp. 165-194.
Roche, P •• Calba, H. and Fallavier, P. (1983) Estimation du niveau de fertilité
phosphorique des sols de Cote d'Ivoire et évaluation des besoins en engrais
phosphatés. IMPHOS 3rd International Congress on Phosphorus Compounds,
Brussels, pp.289-297.
Salcedo, I.H., Bettino, F. and Sampaio, E.V.S.B. (1991) Reactivity of fertilizer P in
tropical soUs by isotopic exchange kinetics of 32p. Soil Science Society of America
Journal55, 140-145.
Sawhey, B.L. (1973) Electron microprobe analysis of phosphates in soils and sediments.
Soil Science Society of America Proceedings 37, 658-660.
i1
1
1 j
Sharpley, A.N., Jones, C.A., Gray, C., Cole, C.V., Tiessen, H. and Holzhey, C.S. (1985)
A Detaîled Phosphorus Characterisation of Seventy-eight Soils. U.S. Department of
Agriculture, Agricullura/ Resear'ch Service Publication 31, 32 pp.
Singh, J.S., Raghubanshi, A.S., Singh, R.S. and Srivastava, S.C. (1989) Microbial biomass
acts as a source of plant nutrients in dry tropical forest and savanna. Nature 338,
499-500.
Stewart, J.W.B. and Tiessen, H. _ (1987) Dynamics of soil organic phosphorus.
Biogeochemistry 4, 41-60.
Stewart, J. W.B., Tiessen, H. and Frossard, E. (1990) Phosphorus requirements of semi-
arid lands. In: Unger, P.W., Sneed, T.V., Jordan, W.R. and Jensen, R.W. (eds.),
Challenges ln Dryland Agriculture - A. Global Perspective. Proceedings of the

187
 International Conference on Dryland Fanningt AmarillO/BUShland. Texat, August 15-
19, 1988, Texas Aancultural Experiment Station, 8uahland, pp.433-435.
Tiessen, H. and Cameiro dos Santos, M. (1989) Variability of C, N allG P (;(lntent of a
tropical .semiarid 80ilas affectcd by wi! Banens, emlÎot\ and land clearing. PIani und
Soi/Il', 337-14.1.
Tie88ell, H. and Stewart, 1. W.lI. (1983) The biogeodlCmistry gf soU pho~phorus. lm
Caldwell, D.E., Brierley, J.A, and Brietley. C.L. (ed~.), PlaMlury Ecology, VaQ
Nontrand Reinhold Company, New York, pp.463-472.
Tiessen, H., Stewart, 1. W.B. and Moir, 1.0. (1983) Changes in organic and inotganic
phosphorus composition of Lwo gnwland SOm and their particle size fuwtions dunna
6()..70 years of cultivadon. JolU'IIDi of Soli Science 34, 815-823.
Tiessm, H., Stewart, 1. w.n. and Cole. C. V. (1984) Pathways of ph08Phorus
tmnsformations in soils of diffeting pedogenesis. $<>il Science Society <>f America
Journnl 48, 853-858.
Tiessen, H., Frossard, H" Mermut, A.R. and Nyamekye, A.L. (1991) PhosphoruJ
sorption, And properties of ferru8ÙlQus nodules t'mm Jemiarid mils from Ghana Md
Btuil. Geodel'l'M ~J ;'7;3-389.
Van der Zaagt T'., Fox, R.L., De La T'cna, R" LaUlJhlin, W.M' Ryslwnp, A.t Villagarcia.
I

S. and Westennann, D. T. (1979) The utility of phosphate SOrption curves for

transfetl'inlJ soil management information. Tropical Agricullure (Irtnidml).56, 153-
160.
Willett, I.R., ChMtres. C.J. and Nauyen, T.T. (1!)$8) Misratlon of Pholphate into
aggregated patticles of ferrihydrite. JolU'IIDi of S"II SCÎence 39, 275-282.
WilIiams, I.D.II. and Walker, T. W. (1969) Fractionation of PhQlphate in a maturity
JeqUeIlCl: of New Ze.atand baHluc soil profile!. Svil Science 107, 22.. 30.
Williams. J.D.H., Syen, J,K. and Wallœr, T.W, (1967) P'mctionation of soi! inorgank
phospha~ by a modification Of ChanS and Jacmn t Sprocedure, Soil SCiellce Society
of AmeriCil Proçeedings 31, 736-739.

188
Appendix J ISOTOPE STUDIES IN TROPICAL SOn.. BIOLOGY
Roel Merckx

J.l INTRODUCTION

The general interest in using (radio)isotopes to study several processes of relevance to nutrient
cycling is mainly due to the relative ease with which a certain element cao he foUowed
through its different transformations. Detection of radioactivity, or increased abundance in
the case of stable isotopes, in a given compartment of the ecosystem studied is an unequivocal
proof of its origine The most straightforward example is the amendment of radioisotope-
labelled fertilisers to a given agroecosystem to assess such things as fertiliser efficiency,
losses to various environmental compartments and turnover in soil.

Within the Tropical Soil Biology and Fertility Programme focus is on nu trient cycling in
systems characterised by a high degree of organic residue recycling. Effective management
of these residues will henefit from a detailed knowledge on the fate of the different nutrients
they contain. Traditionnally, carbon, nitrogen and phosphorus are considered as key
elements. Therefore, in this review, isotope studies related to the cycling of those three
elements are emphasised.

Ignoring the very short-lived radioisotopes of nitrogen (e.g. 13N with lfÛ[ of 10 min) which
are entirely useless in this type of ecologica1 research, we cao divide the isotope studies into
two groups according to their detection: radioisotope studies relying on (1 and/or p-
s~troscopy and stable isotope studies using mass spectrometry. With respect to the latter,
1 N is the most frequently used isotope for nitrogen. (fechniques available to quantify
nitrogen fixation using 15N methodology are reviewed in Appendix G, "Techniques for
quantifing nitrogen fixation".)

Focus here will he on the methodology associated with the use of labelled residues to study
their decomposition and cycling of their elements within a given system and on methods
estimating root material production and turnover.

Note: A general methods overview cao he found in Hardarson, G. (Ed.) (1990) Use of
Nuclear Techniques in Studies of Soil-Plant Relationships. Training Course Series No
2, International Atomic Energy Agency (IAEA), Vienna, Austria.

J.2 DOUBLE (14C, 1'N) OR TRIPLE LABELLED eC, 1~,

4
37p) RESIDUES

Decomposition experiments with uniformly labelled residues have been performed at various
places and with various plant species (Jenkinson, 1981; Paul and van Veen, 1978; Ladd and
Martin, 1984; Singer, 1992; Scharpenseel and Neue, 1984). Nevertheless the number of

189
'ISBF: .A. Handhook 0/ Mc:lhodJ

s~ie~ involved is mther limited and certainly fails to adequately çovet the hug~ range of
materiaIs of rel~vanœ to tropical fanning systems in generaI and tu agroforcstry-related
systems in particular.

The general idœ of decomposition experiments with 1abelled reaidues beio& the d~tection of
re!idue-dcrivcd clements in vat'ÏOU! compartmentJ of the system and how this evolves with
time, the fint question mould be for how long these measurements will have 10 be continued.
Thi! i! critical as it predetermin~s the level of radioaetivity ta he added to the system. A
general principle in radioisotope methodology is the so called "ALARA" principle
rooommending that radioactivity amounts should be As Low As Reasonably .â,ehievable
(lCRP, 1977). AS JUCh, ~very ~eul8tion starts with the detection limits of the counting
device and then calculating back with severaI Wisumptions on transfer processes of the hotope
between the different compartments involved. An example of Juch œ1culation is given
below. FurthennOfe, the l~ngth of the experimenral period aiS{) determines what kind of
isotopes should he use({. The use of np and np is restricted to rather short-term experiments
in view of their short halHives (14.3 days and 25.3 daY5 re,spe.ctively). Nevenheless, J~p and
JJp labellîng of reJidues MS bren used to derermine how much of the phosphorus presçnt in
mooic residues was
recovered by a wheat crop in South AuSttalia (MCl.Aughtin et ni., 1988).

J.3 LABELLING THE PLANT MATERIAL

This is undoubtedly the bottleneck of the experimenUi with labelled residues. Although
labelling with I.5N, l1p can he achieved relatively easy usmg nutrient solution techniques
(Ladd et ul., 1981; Mcuugh1irt ~t dl., 1988), labelling with 14C requires a relatively
sophistieated growth chamber, preferably with control of the specifie aetivity of th~
atmosphere.

For 811 thre.e 18bels it is mandatory to ensure homogeneous labelling, i.e. that the specifie
activity of the element is constant amongst th~ different fmctions of the materlal. Provision
of 14C at constant specifie activity requires a c10sed canopy under which plants cao grow and
where the radioactivity of the atmosphere CM be monitored and kept at a preset level. The
most economica1 way of achieving this is by supplying radioactive Na2 1"C03 into an acid bath
at rates dictated by a. preset radioactivity level in the growth chamber as monitored by a G.M.
tube (Gurr and AdeYt 1978). Homogeneous labelling with 15N or 32pP3p can be aeheived
by (a) growing the plants in a mooium contributing minimally to the N or P demand of the
plant and (b) amending the nutrients dbtributed 8CCOrding to the plant needs over the growing
season. For nirrogen, UN labelled (NH"hSO" cao he used, whereas for phosphOl'US 32p
labelled Ca(H1PO")1 or 33p labelled Ca(HzPO"h cao he u5ed. By a proper selection of the
nutrient additions according to earlier established growth rate$, homogeneous labelling is
encouraged. With respect to the use of phosphorus isotopes, the short half-lifes of both 2p,
14.3 days; 33p, 25.3 days) restrict their use to short growing periods.
e
J.4 DECOMPOSITION EXPERIMENTS

No matter how the numher of growing facilities of labellcd plant materials will increase in
the future, the relatively high cost of the labelled materiaIs willlimit their use to small-scale,
more detailed experiments. In general, the material CM he used in the field if confined to

190
 Appendù J: Is%pe SlIldies in Tropical SOU Bi%gy

small surfaces, or in small litter bags. The material cao be used as a surface litter or
incorporated in· the topsoil depending on the experimental objectives but again the limited
amount available will require that sorne fragmentation of the residues has to be done to
achieve sufficient homogeneity within one cylinder. Application rates and modes of
application cao vary according to the research needs.

Procedure

A typical experiment may offer two main options: (i) closed cores restricting root in-growth;
or (ii) mesh type cores (fSBF, 1991) allowing free horizontal passage of roots and fauna.
Effectively, the core on1y serves the purpose of confining the litter to a specified surface. A
wide-mesh coyer prevents the wind from removing residues from the surface. The procedure
described by Ladd et al. (1983) is generally applicable:

Edged, open-ended plastic cylinders (17.5 cm long by 10.3 cm diameter) are driven to 15 cm
depth in soils at predetermined field sites. At each site cores, in numbers determined by the
experimental objectives, are randomized within replicated (2 or 3) b1ocks. Soils, to say 7.5
cm depth, are removed from each cylinder, weighed, sieved (2 mm), mixed and bulked for
each block at the sites.

Soil subsamp1es, of a weight equal to that of the average weight removed from each cylinder,
are mixed with preweighed 14C-labelled plant materials, of a radioactivity set by the duration
of the experiment and type of analyses (see Example 1 below). The soils are retumed to the
cylinders, and the surfaces tamped gently so that the bulk density of the amended soil is the
same as that of soil prior to its removal.

Later, at appropriate times and especially in the tirst year of decomposition, replicated
cylinders are removed from the field, and the soils from each are dried and sieved (2 mm)
in toto. Subsamples (20 g) are then more tinely ground in a Tema mill before combusting
aliquots (usually 0.5 g) by the procedure of Amato (1983): the soil is mixed with combustion
e
liquid (6 ml) in Tecator tubes (75 ml), stoppered. then heated in a block at 130 0 for 1 hour.
The evolved 14C02 is trappe<! in ethanolamine (1 ml) contained in a glass vial and supported
in the combustion tube (but away from the heated areas) by a bent glass rode Routinely,
absorption of 14C02 is allowed to take place ovemight following cooling.

The vials containing ethanolamine with absorbed 14C02 are transferred to scintillation vials,
shaken with scintillation fluid and the radioactivity determined. Corrections are made for
background, and combustion and counting efficiencies, by using combusted standards.

Example 1. Calculation 10 determine fhe anwunr of radioaclivity which needs 10 be added

althe beginning of a 2-year experiment

To study the dynamics of the soil microbial biomass carbon in soil cores (cylinders), one
wants to follow the 14C-content of the microbial biomass during the 2 years following an
amendment of 14C-Iabelled residues; one must therefore add sufficient 14C to each cylinder
at the begining of the experiment to have sufficent 14C in the microbial biomass for analysis
after 2 years.

191
TSBF: A. Handbook 01 MeÛfodl

MSumpcions for chis example.·

1 Organic 14C n:sidue dc:composc:s at a rate similar 10 that obscrvc:d in Nigeria for
ryegrass (Jenkinson and Ayanaba. 1917). This means that after 2 yeats 14% of the
initial input remmns.

2 Alter comparable decomposition intensities the proportion of this residue present in

soil microbial biomru;s b .± 5 % (Ladd et al., 1985).

3 Microbial biomru;s 14C b measured by a fumigation extraction method (Ladd and

Amato, 1988) where 50 g soil samples are extracted in 125 ml M Kel before and
after 10 days fumigation with CHCI, at 25°C. The 14C activity in these extracts are
determined with liquid scintillation counting in a 1 ml sample. Micmbial biomass 14C
is estimated from the gain in extractable 10lC alter fumigation, multiplied by 3.25.

4 To obtain a gOôd Mmprombe berw~n Munting ume Md a~uracy at leMt 100 MUn"
pet ntin\lte arc nççdçd. or 1.67 cps. With a wwtWg çffiçicnçy of 63% this mcans
that 2.0 Bq oC 14C in 1 ml extract is requiroo.

Ca/cu/arions

Amount of 14C needed in the soU mierobial biomass 2 yeatS after amendment per 50 g
soU: 125 ml extractant = 2 Bq/ml x 125 ml x 3.25 = 0.81 kBq

Arnount of wil in the top 7.5 çm of eaçh çylinder = 625 çm~ = 700 g
(assuming a bulk density of ± 1.1 g/cm))

Then:fore the 14C nc:edc:d in the 700 g soii sample after 2 years = 0.81 kBq x (700/50) =
11.40 kBq He

Given that only S% of the initial He is in the microbial biomass (ASSumption 2). the
requirement per cylinder (700 g soU) ~ 11.40 kBq x (l00/S) ~ 0.228 MBq He

But, given this b on1y 14% (Assumption 1) of the amount added as a residue, the amount
nec:ded ta he added 10 each cylinder (700 g soU) 2 years earlier

~ 0.228 MBq x (100/14) = 1.625 MBq 14C

If a residue is added at a mte of 500 mg/eylinder (~ 200 mg C/eylinder; assume organie

materials contain 40% C) we nood to apply materials with a specifie aetivity of

1.625/200 == 8.13 kBq/mg C

The example scrves as an illustration of how to determine the amounts of mdioaetivity needed
to meet the expcrimental obj~tives; nevertheless, a1ways keep radioisotope levels at a
minimum (ALARA).

192
 Appmdbc J: Isotop~ Sludi~s ln Tropical Soli Biology

J.S IlC METHOOOLOGY IN ESTIMATING TURNOVER RATES OF SOIL

ORGANIC MATTER FRACTIONS

The method basica11y exploits the phenomenon that ~ plants differ from C4 plants in their
natural abundance of the 13C isotope due to their different photosynthetic pathways. Soil
organic matter derived from plant residues therefore reflects its origin since this 13e ratio is
preserved during decomposition. Soil organic matter derived from a ~ vegetation bas typical
mean 6 13C values of -27 %0 whereas C4 residues have 6 13C values of -12 %0 (Smith and
Epstein, 1971).·

In situations where a sudden shift from ~ to C4 vegetation took place as for instance after
forest clearing to sugar cane plantations, the graduaI change in 6 13C of the different soil
organic matter fractions allows one to calculate their respective decomposition or turnover
rates (Balesdent et al., 1990).

The method however bas two main limitations:

1 its use in its idea1 form is restricted to pure ~ - C4 transitions to maximize the
difference in 6 13C values; and

2 access to a very accurate mass spectrometer is essential.

Isotope ratios are usually expressed as 6 13C values:

6 13C ( %0 ) = {(R sample 1 R reference) - 1} x 1000

where R is the l'lC: 12C ratio for sample and reference respectively.

To. calculate the proportion (t) of organic carbon originating from source A in a mixture A
+ B whose 6 13C is equal to an experimentally determined 6, the following approximate
formula is used:

where l- A and t»B are the a l'lC values of the two sources A and B, in our case C3 and C4
derived residues respectively.

UsuaIly, we know the total carbon content (C) of the soil (or fraction) so that:

where
CA = f xC
Cs = (1 - f) C

Therefore we can write:

193
T'SBF.. A. HanJbook 01 Mt:rJroW

or

with CA, the amount of carWn derived from the source A (i.e. the C3 vegetation) M the ot1ly
Wlknown.

J.6 l''e METHOnS TO QUANTIFY aOûT PRODUCTION AND TURNOVER

Despite the substantial improvements in au10mated mot observation techniques (see ASA,
special publication no. 30) considemble difficulty persists in the tan5lation of the data
obtained inro nutrient pools and turnover rates. Labelling of roots and the associared products
of rhiwdeposition has been achieved by expo5Îng plant tops either continuously or at short
intervals 10 an atmosphere containing 14C01 (Martin, 1977; Keith et al., 1986). Methodology
is adequately desçribed in the relevant papers and it appears that for field measuremcnts
pulse-Iabelling techniques provide the most convenient wmpromise. Whereas continuous
labelling requires large and sophisttated growth chambcn, whieh are not easily
accommoclate<1 in the field, pulse labelling can he performed with relatively simple equipment
(Keith et al., 1986).
However. the bulk of titerature refers to annual crops like wheat, maize, say bean, with
relatively short growing periods. Larger plants (lnce ttees in agroforestry systems) with
longer growing periods have not been subject to these measurements for obvious reason5.

References

Amata, M. (1983) The c;lewrmination of carbon IlC and 14e in plant and soi1. S()il Bi()/()B)I
and Biochemistry 15, 611-612.
ASA Special Publication no. 50. (1987) Minirhizotron observation rubes: merhods and
applications for measuring rhizosphere dynamics. Taylor, H.M. (ed.), American
Society of Agronomy, Crop Science Society of America, Soil Science of America.
Balesdent, J., Mariotti, A. and Boisgontier, D. (1990) Effeet of tillage on soil organic
carbon mineralization estimated from 13C abundance in maire fields. ]{)U17Jll1 of Soil
Science 41, 587-596.
GUff, C.O. and Adey, R.F.J, (1978) A devi~ to wntrol radio-carbon levels in the
atmosphere of a plant growth cabinet. Laborutory Practice 27, 197-199.
lCRP (1977) Rooommendations of the ICRP, IÇRP no. 26, Pergamon Press, Oxford.
lenkinson, O.S. (1981) The fate of the plant and animal residue5 in soil. In; Greenland,
0.1. and Hayes, M.H.B. (OOs.), The Chemistry of Soil Processes. 10hn Wiley and
Sons, New York, pp.505-56I.
Jenkinson, D.S. and Ayanaba, A. (1977) Decomposition ofcarbon-14labelled plant material
under tropical conditions. Soi! ScielfCe Society of America Joumal41, 912-915.
Keith, H., Oades, J.M. and Martin, J.K. (1986) Input of carbon to soil from wheat plants.
Soil Biology and Biochemistry 18, 445--449.
J-.add, 1.N. and Martin, I.K. (1984) Soil organic matter studies. ln: L'Annunziata, M.F.
and Legg, J.O. (ws.), Isoropes and Radiation in Agriculfural Sciences 1. Academie
press, New York, pp.67-98.

194
 Appnulb: J: 1solope Siudies in Tropical Soil BWlogy

Ladd, J.N. and Amato, M. (1988) Relationships between biomass 14C and soluble organic
14C of a range of fumigated soils. Soil Biology and Biochemistry 20, 115-116.
Ladd, J.M., Oades, J.M. and Amato, M. (1981) Microbial biomass formed from 14C, 15N_
labelled plant material decomposing in soils in the field. Soil Biology and
Biochemistry 13, 119-126.
Ladd, J.N., Jackson, R.B., Amato, M. and Butler, J.H.A. (1983) Decomposition of plant
material in Australian Soils. 1. The effect of quantity added on decomposition and on
residual microbial biomasse Australian Journal of Soil Research 21, 563-570.
Ladd, J.N., Amato, M. and Oades, J.M. (1985) Decomposition of plant material in
Australian Soils. III. Residual organic and microbial biomass C and N from isotope-
labelled legume material and soil organic matter, decomposing under field conditions.
Australian Journal of Soil Research 23, 603-611.
Martin, J.K. (1977) Factors influencing the loss of organic carbon from wheat roots. Soil
Biology and Biochemistry 9, 1-7.
McLaughlin, M.J., Alston, A.M. and Martin, J.K. (1988) Phosphorus cycling in wheat-
pasture rotations. 1. The source of phosphorus taJœn up by wheat. Australian Journal
of Soil Research 26, 323-331.
Paul, B.A. and van Veen, J.A. (1978) The use of tracers to determine thedynamic nature
of organic matter. Proceedings of the llth International Congress ofSoii Science 3,
1-43.
Scharpenseel, H.W. and Neue, H. V. (1984) The use of isotopes in studying the dynamics
of organic matter in soil. In: Organic Malter and Rice. IRRI, Los Banos, Philippines,
pp.273-309.
Smith, RN. and Epstein, S. (1971) Two categories of 13C 1 l2C ratios for higher plants.
PIani Physiology 47, 380-384.
TSBF (1991) Core experiments, draft protocols version 1.0. TSBF, Nairobi.

195

LJ
Appendix K USING GROWTH ANALYSIS TO ESTIMATE PLANT
NUTRlENT UPTAKE
Robert Selw/es

K.l INTRODUCTION

Growth analysis is a procedure for the analysis of data obtained by the sequential sampling
of plant populations. individual plants or plant components. It reduces the data to a set of
standardised quantities for comparison within and between experiments. Although originally
develope.d fi) qUMtify carbon ililiimilation, it is equally applicable fi) nutrient uptake problems.
Within the TSBF programme it representJ one apprœch to estimating the nme and magniblde
of plant nutrient demand. The minimum data requimd 10 apply gmwth analysis is the dry
mass of plant material harvested at two rime~, which are a known period apart. For nutrient
applications the nutrient concentration in the tissues at œch rime must also be known.
Typically, samples are taken about five timeli during the crop growth period, yieldina four
sets of gmwth analysis statistics.

The free availabiHty of computen in recent years has had a oon~îderable impact on the way
in which the ~hnique is applied. In its "clauical" fonu. which has remained essenria1ly
unchanged for fifty yeatS, the growth of the plant is assumed ta he exponential; in other
words, by a constant proponion of the already ~sting tissue at each rime srep. This
assumption allows the use of rclatively simple equarions for estimating the instantaneous
gmwth rate, but may be erroneous during certain phases of plant growth, or when arowth
is under strong external limimtion. Precbion of estimation relies on taking a large number
of replicate liamples at each of a few time intervals during the growth of the plant.

The more modem "functional" approach makes no restrictive a priori assumprions about the
form of the growth curve. Fewer plant samples are taken, but at more times, and a smooth
empirical function is fitted to the data. The growth rate ili wmputed from the differential of
thb functîon. The functional approach is recommended ta at1yone with access to a desktop
computer.

A simple introduction 10 growth analysis is given by Hunt (1978), and a more detailed
treatment of the functionaI approach by Hunt (1982). The standard reference to the classica1
approach is Evans (1972). The approach as it relates to nutrient uptake problems Îs
developed by Nye and Tinker (1977) and Barber (1984).

K.l THE CLASSICAL APPROACH

The absolute rate of growth of a plant (g/day) chanaes continuously during the development
of the plant. Growth analysis seeks to do three things:

196
 AppendU K: Uring GroWlh Anolyris 10 Eslimale PianI NUlrienr Uptake

(i) estimate the mean instantaneous growth rate over a period by sequential measurements
of the absolute mass. The instantaneous growth rate is the slope of the plant mass
versus time curve at any one moment (dW/dt);

(ii) express the growth rate relative to the amount of tissue present at the time of the
measurement, so that comparisons cao be made between different measurement
periods and different plants. The basis of most growth analysis functions is an
equation of the form (1IW) (dW/dt); and

(iii) apportion the observed growth into the contributions due to changes in size of the
growing organs and changes in efficiency of existing organs, to assist in the
interpretation of the growth process.

If 1W is the dry mass of tissue harvested at 1t, and 2W is the dry mass at time 2t, then if the
growth of the plant is exponential, the mean Relative Growth Rate (RGR) over the period is
approximated by

(For an explanation of notation used see Note 1 below.)

The efficiency of the photosynthetic process cao be estimated by expressing the growth rate
relative to the area of leaf available for photosynthesis, (LA)' This quantity is know as the
Unit Leaf Rate, ULR, and its mean value over a period is

The mean size of the photosynthetic organs over a period, known as the Leaf Area Ratio,
LAR, is given by

1_2LAR (m2/day) = {(lLA/1W) + hLA/IW)} 12

It cm be shown that

When these concepts are developed for a crop rather than for a single plant, and are related
to the area of ground on which the crop is growing, an analogous set of quantities is defined,
including the Crop Orowth Rate, COR, and the weIl known Leaf Area Index. The above
calculations are usually (but not necessarily) performed on above-ground plant parts; other
relationships, such as the RootShoot ratio relate above-ground growth to total growth and
growth partitioning.

K.3 APPLICATION TO NUTRIENT UPT AKE

If the nutrient content of the plant is use<! in place of the plant dry weight, and the plant root
system is treated as the assimilatory organ instead of the leaf area, then a set of equations

197

Il
TSBF: .A HancIbook 01 M~rJwds

d~ribîng belôw gmund gmwth and nutrient uptalœ can he developOO. These are analogous
to those presented for above-ground carbon assimilation above.

If LR is the length of the mot system (m). (sc:e Note 2 below) and U is the nutrient content
of the plant (U .. C x W, where C is the nutrient mncentmnon in the plant tissues), then the
mean Root Extension Rate, RER, is

and the mean Relative Uptake Rate. RUR. is

The mean net inflow of nutrients per unit ôf rôO( length, l, which could he considered ta he
a root efficiency term analogous to the ULR., or altematively as a measure of the availabilily
of nutrients 10 the plant, is given by

Beçause the partitioning of nutrients between roots and shoots il nQt aIual. and considerable
retranslocanon of nutrients occurs within the plant, it is necessary ta estimate the mass and
nutrient concentration of ail the plant parts. not just the above-ground portions.

Notes

1 The notation traditionally uKd with growth analylil is rathe! peculiat. but is followed
here to avoid confusion should referen~ be made to other sources. The preceding
subscript tefers 10 the time of measurement, and the following subscript ta the
component which was measured. For example, 3W 1caf would he the leaf dry mass at
the time of the third sample.

2 Since the fine roots which are responsible for the nutrient uptake are cylinders with
a dialneter much smaller than theîr length, the 1ength of the root system is a hetter
measure ôf its assimilatory capacity than either the root surface area or root masSa
It is aIso easier to moosure that fOot surface afea.

References

Barber, S.A. (1984) Soil NU/rient Bioavailability: A Mechanistic Approach. Wiley, New
York.
Evans, G.C. (1972) The QUllntitative Analysis of Plant Grawlh. Blackwell Scientific
Publications. OJl.ford.
Hunt, R. (1978) PIani Growlh Analysis. Studies in Biology 96. Edward Arnold, London.
Hunt, R. (1982) Plant (Jrowlh Curves: the Functional Appraach ta plant Growth Analysis.
Edward Amôld, London.
Nye, P.H. and Tinker, P.B. (1977) SalUle Mavemeru in the Sail ROOf System. Blackwell
Scientitic Publications, Oxford.

198
Appendix L USE OF THE CENTURY PLANT-SOIL
ENVIRONMENTAL SIMULATION MODEL
Paul Woomer

The Tropical Soil Biology and Fertility Programme is currently evaluating the use of
environ mental simulation models as a means of testing our current perceptions of mechanisms
operative within soils and their }X>ssible interactions. At the same time, we are actively
calibrating and validating these models in order that longer-term consequences of land
management strategies may be predicted. The model most frequently used to test soil organic
matter dynamics and the effects of land use is the CENTURY model (parton et al., 1987,
1989). CENTURY is a general model of plant-soil ecosystems that may be used to represent
many different land uses inc1uding natural and managed grasslands, natural and plantation
forests and field monocrops. CENTURY simulates the dynamics of carbon, nitrogen,
phosphorus and sulphur within plants, residues and soils. This model is available in PC
format from Dr W J Parton of the Natural Resource Ecology Laboratory, Colorado State
University, Fort Collins, CO 80523, USA for US$150.00.

L.1 COMPUTER SIMULATION OF DECOMPOSITION AND SOIL ORGANIC

MATTER TRANSFORMATIONS

Detailed knowledge of decomposition rates of various fractions of applied material and soit
organic matter 1 and the transformations between those fractions over time are best utilized
through the development of computer models. Computers allow for iterative calculation of
organic matter dynamics at a fixed time step. We are no longer dea1ing with "batch "
decomposition, but detailed transformation between and loss from organic matter functional
pools. Early models of this type were developed by Jenkinson and Rayner (1977) and Van
Veen and Paul (1981).

Environmental simulation offers great opportunities in the anticipation of the effects ofvarious
land management strategies. While model output will never replace on-site experimentation,
a large number of candidate land-use strategies can be rapidly assessed allowing for a pre-
selection of experimental treatments and anticipation of experimental outcomes. Prior to this
reliance upon modelling for these purposes, however, model output must be calibrated with,
and validated by measured parameters. The Tropical Soil Biology and Fertility Programme
(TSBF) is currently investigating the usefulness of the CENTURY mode1 (Parton et al., 1987,
1989) across a range of tropical ecosystems and land uses. Idea11y, a validated model could
be used in predictive mapping of the tropics using georeferenced databases and geographic
information systems as has been done in temperate regions (Burke et al., 1990).

The CENTURY model simulates plant productivity and soil dynamics for many land uses and
management regimes of terrestrial ecosystems (panon et tl/., 1987). Early versions of

199
TSBF.- A HanJbool 0/ Mcdrcxb

CENTUR y accurately simulated productivity and soil development Wld nutrient accumulation
in sorne tropical ecosystems (panon et al., 1989).

L.2 MODEL STRUCTURE AND DESCRIPTION

Parton el al. (1987) reported a detailod description of ÇENTURY; interelited leaders are
referred to that publication. CENTURy was first developed 10 simulate grassland ecosystems
of the Great Plains region of the US, and i$ ç~rrently being evaluated in mWly other areas of
the wotld. Briefly, CENTURY simuIates soit dynamic based on the dynamics of funclional
pools. These pools are diffete1ltiatcd on residence lime in soils and the C:mineral mtios of
these pools. Plant productivity il simulate<l via uset-defined (or default) algorithms
appropriate to a partieular ooosystem and climate.

The overa11 structure of CENTURY ili presenœd in Figure L.l. CENTIIRY consists of 3
major wmponents; INITPAR, CENTURYand VIEW. INITPAR is used b) devclop site flles
that initilÛi2'.e the limulation, INlTPAR may a.1w he ulied 10 view model input and output
definitions. The CENTURY component runs the environmental simulation; CENTURY must
a1ways he nm with a data and fix file (i,e. site.VAT and site.FIX, respectively). The data
flle (.DAT) defines specifie soil and elimate attributes, the .FIX file is based on more general
ecosystem attributes. CENTURY may he fWl for a grassland t empland or forest, with
various management options avaiJable to the UJe!'. The mode! output is then diIectc:d to the
VIEW component. VIEW is the output module of TlME-ZERO: an integrated modelling
environment. VIEW allowli for the output 10 he plotted, printed, or saved. VIEW also
a110ws the user 10 change input parameters and to re-run the liimulation without re-enœring
INITPAR. .

CENTURV (dlreetory)

......•... , .. "." .. " ..... "." .... ",,, ...','!.•.','.......',,),','.'.'.

.....:.:.:.:.:-:.:':.:':.:':.:.:.:,.,.(,.,:,.•.,.,:.:•.':':':':':':':':':0:':':0:':':':0:':':':0:':0:':':':':":'

':'):':'):':':':':':':':':':':0:':':':':':':':':':':':':':':O:':':·:'):':':':':·X(':".(('.'I.'!):':':':':'
:-:':':-:':':':':':':':':0:':':':':':':':':':':0:':':0:(':0:':':':':':':':':':«((':':':0:':(':0:,',',":':':':0:':

FIgure L.t General structure of CENTURY. Files are constructed in INlTPAR, run in
CENTURY and visualise<! in VIEW.

200
 AppDldix L: Use of lM Cmlllry Planr-Soil Envir01lllfC1lal SinudtUion Madel

L.3 HARDWARE REQUIREMENTS AND INSTALLATION

The PC version of CENTURY is written in Fortran and requires an IBM PC or compatible

with at least 512K RAM. While CENTURY can run on older processors such as an 80086,
1 the simulations are quite time consuming. It is recommended that users run CENTURY on
'1
either a 80286 or 80386. For high resolution sereen output, a VGA monitor is recommended
although VIEW is supported by CGA and EGA monitors as well.

Together CENTURY and VIEW occupy 614K of bard disk drive. VIEW must be loaded as
a subdirectory of the directory where CENTURY model files are stored. CENTURY site
files (.DAT) may be stored directly in the main CENTURY directory, furthermore, site files
cannot be read from an external drive.

L.4 RUNNING CENTURY

A session with the CENTURY model is a1ways initiated by entering either INITPAR, if the
user' s intent is to create a new site.DAT file or CENTURY site.DAT if the user plans to run
a simulation with existent site files. INITPAR is a user frlendly routine whereby site files
are constructed or modified. To enter INITPAR, the user must always refer to an already
existent site.DAT file, modifying this file until representative of a new environment or land
use. INITPAR is separated into minimum and complete initialising variables. After entering
new variables through initpar, the user must save the file under a new name as CENTURY
does not overwrite existing files. The initialising variables, user management options and
suggeste.d calibration and validation data required for the adoption of the CENTURY model
into the Tropical Soi1 Biology and Fertility Programme are presented in Table L. 1.

General information on ccological attributes is contained within the site.FIX flles. For
example, this file sets limits on the relationship between precipitation and plant productivity
or on the rates of plant nutrient complexation with soil mineraIs. A HUMID.FIX file has
been developed at TSBF headquarters that is universa11y used for sites of the lowland humid
tropics with predominantly oxide soil mineralogies. These site. FIX files are more difficult
to create than site.DAT files, requiring the use of a line editor.

A site data file must a1ways accompany the CENTURY command (e.g. CENTURY
SITE.DAT). Users are than asked to select a land use, pattern of events, a site.FIX file and
management inputs. Several options of climate inputs are available, although the climate
records contained within the site.DAT files are the most conveniently used. A CENTURY
simulation is then generated for those conditions and the results of that simulation are
automatically accessed through VIEW.

L.S PRINCIPAL MODEL ROUTINES

CENTURY simulates the carbon, nitrogen, phosphorus and sulphur dynamics for grasslands,
forests and cropping systems. These different plant production systems are linked to a
common organic matter submodel that consists of 3 functional pools based on the residence

201
TSlJF: A HIR'IIJbook "1 MelM4.s

Table L.1 Minimum data. requirements for a site file to initialise CENTURY, management
options available ta the user and examples ofuseful data for model calibration and validation.

Mi,"","", file requiremenls;

clay content, numbcr of elements simulated (Ct Nt P, S), rime sreps per month, soil pH,
monthly precipitation, sand content, sUt content, length of simulation (years), monthly
minimum tempemture, monthly maximum temperature, grass growth season (months of
growth and senescence), crop growth season (months of growth, harvest and senescence),
forest growth season (months of gmwth and litter fall).

Land use 11Ul1Ulgement options:

glUS removal pattern (undislurbed, fire, graud or both), grassland fire event (years, month),
fertilisation timing and quantity (N, P, S~ amount and when npplied), addition of
stTaw/residues (quantities, frequency, eN, ç;p, C:S illd lignin content). crop/fa1low parœm
(years of vAtious land U~), tillage pattern (months, depth), timing of forest reMovallretum
events, biomass allocation of forest removal events.

Useful calibration/validation data:

total soil organic (C, N, Pl, extractable nitrate and ammonium, soil microbial biomass (C,
N. P), illnual above-ground plant producdvity, plant coMponents of yield, above-ground
standing biomas5, standing dead biomass surface liner biomass, mot biomass.

time of organic materials in soils. The plant nument ~ubroutine$ (N. P and S) are
incorporated into plant productivity and organic matter dynamics through user-defined
C:nument ratios that float within cerWn limits. The dynamics of these plant nutrients is
centred about a labile form thal is available ta plant uptake. subject to complexation with soit
mineraJ.s or loss from leaching or ero~ion.

An elOample of the overall organic matter C and N dynami,s for a I;mp-soil system is
presented in Figure L.2. Plant productivily is controlle.d by moisture availability. temperature
and the maximum a.n.(I minimum C;nutrient ratios of new planl growth. U!;er-defined
partitioning coefficienls determine the relative I>r<-PQrtion5 of grain. shoots and roots. Upon
senesœnœ, plant materials enter the litter pools that are divided between easily metabolised
and more reQ}l;itrant materials. These liuer {oms enter into the different K'il organil; matter
fractions nt rates determined by their I;hemical compositions, soil textures and soH
microclimate. System losses primarily occur as respimtion of COl' and the nutrients
associated with the respired C are assumed to become mineralised and enter thcir respective
labile pools.

L.tJ CENTURY OUTPUTS

The results of CENTURY simulation May he output as either line graphs or tables.
CENTUR.Y line graphs may contain up to silO variables on the y axis. Users may select from

202
 Apperulix L: Ule of lM COll"" Ploru-Soil Environmerual Simulation Model

a number of line styles and colours, and cao create titles and legends within view, then output
these directly to a printer or capture the image using an external grab feature. Graphs may
also be stored as files within VIEW. Tables are created using the PRINT routines. The
. column headings may be renamed (e.g. different from their respective FORTRAN codes) or
be products of combined output variables. These tables may be either printed or saved as
data files importable into a number of commercially available spreadsheet and graphies
programmes. The complete results of a simulation may also be stored within CENTURY for
future use but these files occupy nearly as much disk space as the total programme.

mass flow - - regulatlon ............. .

PET
S
~ reclp

C
nutrlent flow - - - - key coefficient ©

..
L.~ , __ -@L L-~-:s--s-~-':-1
In~;rt;~----------- ___ 1
Figure L.2 An' example of the linkage between crop productivity and soil organic matter C
and N subroutines in CENTURY (after Woomer, 1992).

References

Burke, I.C., Schimel, D.S., Yonker, C.M., Parton, W.J., Joyce, L.A. and Lauenroth, W.K.
(1990) Regional modelling of grass1and biogeœhemistry using GIS. Landscape
Ecology 4, 45-54.
Jenkinson, O.S. and Rayner, J.H. (1977) The turnover of soil organic matter in sorne of the
RothamstOO classica1 experiments. Soil Science 123, 298-305.
Parton, W.J., Schimel, O.S., Cole, C.V. and Ojima, D.S. (1987). Analysis of factors
controlling soil organic matter levels in Great Plains grasslands. Soil Science SocieLy
of America Journal 51, 1173-1179.
Parton, W.J., Sanford, R.L., Sanchez, P.A. and Stewart, J.W.B. (1989) Modelling Soil
Organic Matter Dynamics in Tropical Soils. In: Coleman, D.C., Oades, J.M. and
Uehara, G. (OOs) Dynamics of Soif Organic Matter in Tropical &osystems. NifTAL
Project, Paia, Hawaii. pp153-171.
Van Veen, J.A. and Paul, E.A. (1981) Organic Carbon Dynamics in grassland soils. 1.
Background information and computer simulation. Canadian Journal of Soil Science
61, 185-201. .
Woomer, P.L. (1992) Modelling organic matter dynamics in tropical ecosystems: model
adoption, uses and limitations. In: Molongoy, K. and Merckx, R., (eds.) Organic
Matter Dynamics in Tropical Ecosystems, prodeedings of a symposium held at
Katholieke University, Leuven, Belgium, 6-9 November, 1992. (In press).
Woomer, P.L. and Ingram, J.S.1. (1990) The Biology and Fenility of Soifs: TSBF Report
1990. TSBF, Nairobi, Kenya.

203

1 1
Appendb: M QUANTIFICATION OF NEAR-SURFACE
MICRO C LIMATE
Alan SlaplelOn

M.I INTRODUCTION

The purpose of this appendix is to provide auidelinci for the measurement of microclimare
bath abOve and bclow grouml. These measurements are require.d fôt ~tudies of management
pœ;tices employcd by farmers to preserve ôr en.lu.nCÇ: thç benefiçial effc:cts of variations in
microclimare and thôse employed ru mitiiate "Bain st adverse microclimatic factôts. The ratei
of biologi~ processes in the soit are Mntrolled by microclimate rather than macroclimate.

Microclimate is influençed by local ropography, vegetation and soil type. A distinct

microclimate exists close to the soil surface, and aIDonsst plants where vegetation cover
influences light, temperature, humidity and wind speed compared to an open site. Below
ground the soil al.so has a very distinct climate characterized by soil tempcmture, soit
humidity ôr water content ami aeration.

Microcümare influences farming praçtiœs and these influences ment investigation in the
TSBF Programme as an understanding of them may provide options for improving cmp
productivity. Stigter el al. (1986) distinguish between microclimate modifiçation and
manipulatiôn. Microclimate modifiçations are inadvenem changes in surface exposure
resulting, for example, fmm weedin& or burning. Mi~roclimate manipulations are deliberate
actioni 10 alter severe aspects of the microclimate by for cXilIIlple mulching or shading.

M.2 MICROCLIMATIC FACTORS

M.2.1 Parameter sel~tion

In microclimare investigation those façlors expected to be agronomica11y decÏ$ive should be

selected for quantification. As the importanœ may weil he indirect, agronomie signiflcance
may have to he demonstrated through field experiments wnducted over seveml seasons. In
on-faon investigations comparlson of different management practiccs requires random
sampling of different microclimatic conditions. Sampling strategies for micrometeorological
experiments are described by Stigter el al. (1976).

M.l.l Meteoroloaical QbeoatioD5

Microclimates arc largely a consequencc of the interaction of soi! or plant surfaœs with large
sca1e meteorologiçal çonditions. The meteorologica1 observations usçd for TSBF site

204
 Appendix M: QlUIIIliftcalion of Near-Slliface Microcllmale

characterisation are ful1y described in Section 2.2 in the main text. Direct observations
consist of precipitation (rainfall), air temperature (measured 1.2 - 2.0 m above ground),
relative air humidity, incidence of ground frost, solar radiation or sunshine, wind speed
(measured 2 m above ground) and pan evaporation. From these observations potential
evapotranspiration for grass as a reference crop or open water evaporation are calculated
using Penman type combination equations based on energy balance considerations.
Microclimatic observations are made using similar equipment. Further details of instruments
particularly suitable for agrometeorologica1 field stations is given by Doorenbos (1976).

M.2.3 Micrometeorological measurements

Over the last twenty years, Stigter, working in a range of tropical countries has made
extensive reviews of the literature concerning microclimate modification and management.
This work provides information about measurement methods appropriate for microclimate
measurement in the study of the major management practices that affect microclimate: these
are mulching, shading and wind protection.

M.2.4 Mulching

The main effects of mulching on microclimate are Mediation of temperature in both the mulch
and in the soil surface, and a reduction water 10ss by evaporation (Stigter, 1984b and 1988a).
Temperature is measured using thermister probes or thermometers. If exposed at or above
the ground surface, thermometers need a royer to proteet them from solar radiation but must
allow ait circùlation. Diurnal temperature fluctuation has been found to be a more sensitive
indicator of mulch effccls than average soil temperature (Othieno el al., 1985). Soil water
content can he measured gravimetricaly. or after appropriate calibration by neutron probe
(Long and French, 1967; Greacen, 1981), nylon resistance units or gypsum blacks
(Bouyoucos, 1954). Alternatively the water content of relatively moist soils cao he
determined indirectly from soil water potential measured using tensiometers (Richards, 1942).
Continuous monitoring is only possible for gypsum blocks and pressure transducer
tensiometers.

M.2.5 Sbadiq

Shading has the following effects which have been reviewed by Stigter (1984a and 1988a):
Proteetion from the sun, reduction of longwave radiation loss to the atmosphere, weed
suppression, reduction of light and radiation in mixed crops and alley cropping. Shading May
also have detrimental effects and pruning May be required to ensure that shaded crops receive
adequate light.

The main factors that are altered under shade are temperature and radiation. Radiation May
he measured using point sensors, moving sensors or sensors that integrate over an area.
Solarimeters or pyranometers measure global solar radiation (W1m2) in the 300-1100 nm
waveband and are suitab1e for continuous monitoring. Stigler and Musabilha (1982) used a
point sen sor to record the near infra-red solar radiation and determined photosynthetically

205
active radiation (PAR) by difference. Sensors are now available that measure PAR directly
as radiation in the 400-700 nm waveband and quantum sensors are available that are sensitive
10 photons. The bigher efficiency of red photons is taken into account thereby giving a more
accumte estimate of the photosynthetic aetivity of radiation rooching the crop. Moving
!IelIsors have been u5ed to measure radiation over a brger area (Stigrer et al. 1977) but theR
J

RqUÏll: an elootricity suppl y for moving the senSOFS. Integrating seniOrs have boon used by
Newman (l5lBS) ta measure photo~ynthetiçally-active nu:1iadon.

QUBDtimrive meulJIÇrnent of net longwave miliation at night has not hem !lu~uful yet.
unprotected minimum thennometers should give a more meaningful comparison in 5uch cases.

M.2.6 Wlnd protedlOD

The effeçtl of shelter introduœd for protection against meehanial damage tu Uees and cmps,
and 10 reduœ water ,treR and wind erosion are reviewcd by Stigler (1984c) and Stigter fI
al. (l988b). Although the spatial pattern of wind pmœctîon Î$ well known in front of and
behind meiterbelb. the effeçts of trees in fOreJt edges. and the pattern of pm~tion givàt by
linaIe or scattered t.œes bu not been ltudied. Protection t'rom wiOd il Quantified by
measurement of wiodspeed (mis). Specially adapted Cup anemometers need to he used as the
type used for meteorological observations are not sufficiently accurate for comparing
treatment effects. Shaded Piche evaporimeters (Stigrer and Uîso. 1981) have been usen as
alternative simple air movement meters. Measurements were made about 20 cm above the
protected surface.

M.3 PROCESSES AFFECTING MICRQCLIMATE

M.3.i Radültion bahwat

The energy which heatl crop and soil gurfaces and evaporates water from them cornes from
solar mdiation and longwave radiation from the atmosphere. Longwave radiation is also
emit.ted from the surface. Campbell (1985a) provide5 wmputational procedures fôr deriving
equation!i for heat and vapour exehange at erop and soil surfaQeS using daily observatiOns of
solu- radiation and air tempetature. If 801ar radiation mea$utemcnts are not available
estimates cao be made from observations of sunshine hour.i using stanclatd constants given
by Doorenbos and Pruitt or specifie conltants for different climadc regions (Frete and PoPOV.
1979).

M.3.2 SoU temperature and beut now

Ali of the physical, chemi~ and biologica1 proceJses that go on in the soil are influenœd by
soi! temperature and in SOrne cases growrh of above-ground plant parts is more closely
correlated with soU temperature than air temperature. Sail tempcratures may he caleulated
as a funetiM of depth and lime by numerical solution of the heat flow equation. Campbell
(l985b) provides values for ca1~ulating soU thermal conductivity and specifie heat from
proportions of minerai, water, air and organic matter in soils and a procedure for defining
the boundary condition at the soi! surface.

206
M.3.3 SoU aeraüon

Oxygen, carbon dioxide and water vapour diffuse into and out of the soil. In weil aerated
soil the carbon dioxide content is around 0.25% and the oxygen content is 20.73%. Oxygen
contents below 10-15% cao inhibit plant growth. Campbell (1985c) provides acomputational
scheme based on Fick t s Law for estimating oxygen concentrations and fluxes in the soil as
a function of aeration porosity and depth below the surface. Practices leading to soil
compaction reduce aeration porosity and hence diffusion. A decrease in oxygen concentration
is also the natural consequence of water-filled pores. Oxygen concentrations decrease towards
the anaerobic centres of microbial and soil material when pores within them are water-filled.

M.3.4 Crop evapotranspiration

Methods for estimating evapotranspiration from meteorological observations have been

developed by irrigation engineers and agronomists for determining crop water requirements.
The guidelines published by the FAO (Doorenbos and Pruitt, 1977) have now been in use in
many countries for some time. Where daily observations of mean air temperature, relative
humidity (or vapour pressure), sunshine hours (or solar radiation) and windspeed are available
the Penman combination equation is used to calculate potential evapotranspiration for grass
as a reference crop. Crop factors which vary throughout the growing season are applied to
estimate potential crop evapotranspiration. Calculation cao be on a dailyor weekly basis.
Wherever possible loca11y derived crop factors should be used, but these are rarely available
and factors provided in the FAO Guidelines are used for determining crop water
requirements.

More detailed investigations of crop water use especially under water-limiting conditions
require separation of evaporation from plant transpiration. Even when soil is covered by
vegetation evaporation is probably at least 10% of evapotranspiration. A complete analysis
of evaporation (Campbell, 1985d) takes into account both liquid and vapour flow and
incorporates the effects of temperature gmdienrs on water movement. This analysis is morc
complicated than required in many applications. If a water balance ralher than water content
profiles are required then vapour flow except at the surface might he neg1ected.

M.4 LINKE.I).TRANSPORT MODElS

In a wide variety of field and laboratory investigations concerned with the interdepence of
physical, biological and chemical processes simulation models have helped to guide our
thinking and experimental design. Pioneering work on modeling evapotranspiration, soil
water and solute flux and plant uptake was conducted at Utah State University (Hanks and
Bowers. 1962: Nîmah and Hanks, 1973). More reœntly work at the Swedish University of
Agricultural Sciences has led to the development of SOIL, a linked transport model that
simulates the flow of both heat and water in soils (Jansson, 1991). This model is particularly
appropriate for using in the investigation of management factors that influence microclimate.
The driving variables needed to simulate both soil heat and water flows in a natural vegetated
soil are precipitation, air temperature, relative humidity, wind speed and solar radiation. A
number of physical soil parameters are required; a good soil profile description combined

207
TSBF: A HœtdbooA: of M~thods

with the soit physîcal characterization &ivçO in Chapter 7 will provide the informatiun
requÎ1'td. Measurements of leaf area index (or plant cover, surface resistance and interception
are required as these are important controls of water 10ss to the atmosphere and depth of litter
and humus are required as these are important controls of soi! heat balance.

M.! AUTOMATIC RECORDING OF MlCKOCLIMATIC FACTORS

Instrumentation for continuous measurement of mi~romeroorologiçal fa~tors is continually

under development. Two basic types of recording instruments are DOW available. These are
data loggers to which a variety of sensors CM be connected and intelligent sensors which have
storage facility built in and capacity for connecting additional sensors.

Loggers are available with up to 32 analogue channels and 3 digital channels. For
micrometeorological investigation sensors such as temperature probes, rombined remperature
and relative humidity probes, tippîng bucket raingauges, cup anenometers, pyranometers, soil
tempetature probes, soil moisture blocks and ten$Îometers might he employcd. To calculate
the stomge capacity required of a loager, multiply the number of sensors by the teadings
required per day by the number of days the logger will he left unattended. A 32000 record
storage capacity will he adequate for half hourly readings of up ro 30 senSOrs stored over two
wecks. Data are analysed on an office based computer using software provided with the
logger. Inexpensive portable computers or persona! organisees can be u~ to traIlsfer data
from the logger' to the office based computer using a cable corutected to the RS232 port on
the computer.

Intelligent seD50rs commercia1ly avaîlable include tippin8 bucket mingauges, combined

temperature and relative humidity meters and pyranometers. Bach intelligent unir CM support
up to three adaitional sen sors such as œmperature probes, œnsiometers Or additional light
sensors. Altematively a four channel intelligent unir cao support up ta four ternperature
probes or tensiomereri. An advantage of these SC11wrs is the portabllity of the unit which can
be connected to the RS232 port of a computer in the office. Altematively a portable
computer coula he taken to the field to collect the stored data.

For any experimental situation the Most suitable recording system will be influe1tcW by the
availabîlity of mereorologicaJ. observations nearby and the computer facilities available. For
microclimatic studies under a forest canopy near The University of Exeter we have selected
a recording system thar consists of an, intelligent senwr to measure rainfall and air
temperature ro augment local meœorological observations and also a data logger for reoording
air and soil temperatul'Cs and soil warer content under the canopy.

References

Bouyoucos, G.J. (1954) Electrical Resistance Methods as fma1ly perfected for making
continuous measurement of soi! moisture COntent under field conditions. Michigan
Suue College QUllnerly Bulletin 37, 132-149.
Campbell, G.S. (l985a) Atmospheric Boundary Conditions. In: Soil Physics with BASIC.
Transport Models for Soil-Plant Systems. Elsevier. pp. 134-145.
Campbell, G.S., (1985b) Soil Temperature and Hear Flow. ln: Soil Physics with BASIC.
Transport Modeb for Soil-Plant Systems. Elsevier. pp.26-J9. .
\
208

1
 Appendb; M: QlUl1Il(ficazton of Nem-Slllface Microcllmale

Campbell, O.S. (1985c) Oas Diffusion in Soil. In: Soil Physics with BASIC. Transport
Models for Soil-Plant Systems. Elsevier. pp. 12-25.
Campbell, O.S. (1985d) Evaporation. In: Soil Physics with BASIC. Transport Models for
Soil-Plant Systems. Elsevier. pp.98-107.
Doorenbos, I. (1976) Agro-meteorologicalfield stations. FAO Irrigation and Drainage Paper
27, FAO, Rome.
Doorenbos, I. and Pruitt, W.O. (1977 revised) Guidelines for Predictoing Crop Water
Requirements. FAO Irrigation and Drainage Paper 24. Food and Agriculture
Organisation, Rome.
Frere, M. and Popov, O. (1979) Agrometeorological Crop Monitoring and Forecasting.
FAO Plant Production and Protection Paper No 17. Food and Agriculture
Organisation, Rome.
Oreacen, E.L. (ed.) (1981) Soil Water Assessment by the Neutron Method. CSIRO,
Australia, 149pp.
Hanks, R.I. and Bowers, S.A (1962) Numerical solution of the moi sture flow equation for
infiltration into layered soils. Soil Science Society of America Proceedings 26,
530-534.
Iansson, P-E. (1991) Simulation Modellor Soil Water and Heat Conditions - Description 01
the SOIL Model. Swedish University of Agricultural Sciences. Uppsala.
Long, I.F. and French, B.K. (1967) Measurement of soil moi sture in the field by neutron
Moderation. Journal 01 Soil Science 18, 149-166.
Newman, S.M. (1985) Low cost sensor integrators for measuring transmissivity of complex
canopies to photosynthetically active radiation. Agriculture, Forestry and Meteorology
35, 243-254.
Nimah. M.N. and Hanks, R.I. (1973) Model for estimating soil water, plant and
atmospheric interrelations: 1 ~ription and seDsidvity. SoU SCience Society of
America Proceedings 37, 522-527.
Othieno, C.O, Stigter, C.J. and Mwampcija, A.R. (1985) On the use of Stigter's ratio in
expressing thennal efficiency of grass mulches. ExperimenJal Agriculture 21, 169-174.
Richards, L.A. (1942) Soil moisture tensiometers: Materials and construction. Soi! SciellCe
53, 241-248.
Stigter, C.I. (1984a) Traditional Use of Shade: A method of microclimate modification.
Archives of Meteorology, Geophysics and Bioclimatology 34203-210.
Stigter, C.I. (1984b) Mulching as a traditional method of microclimate management.
Archives of Meteorology, Geophysics and Bioclimatology 35 147-154.
Stigter, C.J. (1984c) Wind protection in traditional microclimate management and
manipulation: examples fmm East Africa. In: Grace, I. (ed.), The Effects of Shtlltr
on lhe Physiology of Plants and Animais. Progress in Biomeleorology 2, 145-154.
Swetz and Zeitlinger, Lisse, Ne~erlands.
Stigter, C.I. (1988) Microclimate management and manipulation in agroforestry: Viewpoinrs
on Agrolorestry. In: Wiersum, K.F. (ed.). Agricultural University, Wageningen,
Netherlands.
Stigter, C.I. and Musabilha, V.M.M. (1982) The conservative ratio of photosynthetically
active to total radiation in the tropics. JourrwJ of Applied Ecology 19, 853-858.
Stig ter, C.I. and Uiso, C.B.S. (1981) Understanding the Piche evaporimeter as a simple
integrating mass transfer meter. Applied Scientific Research 37, 213-224.

209

, ,11111, ,l'
TSBF: A HandIHx1k of Methodl

Stigter, C.I. Lengkeek, I.G., and Kooijman, J. (1976) A simple worst case analysis for
estimation of corre<;t scanning rate in a micrometeorological experiment. NttMrlands
Jou17Ull of Agrieulmral Science 24, 3-16.
Stigter, C.J., Goudriaan, J. Bottemanne, F.A., Bimie, J., Lenglœ.ek, J.G. and Sibma, L.
(1977) Experimental evaluadon of a crop climale simulation mode1 for Indian corn
(ha mays L.). Agriculrural Meleorology 18, 163-166.
Stigter. C.I •• Darnhofer. T •• Karing. P.H •• Lawwn. T.D .• Popov. G.F. and von Hoy
nincen-Huenet J. (1986) MicrocU1IIllle mmmgemelU and 171llIIipulalio1& i1& Trodilional
Farming. Repon ôf the wôrIdng GMup ôf micmelimate management and
manipulation in tmlitional farming. Ninth Session of Commilsion for Agric.
MeteoMlogy, Madrid. WMO, Geneva. 74pp.
Stigler, C.J., Darnhofer,T. and Herrera, S.H. (1988) Crop protection from very strong
winds: recommendations in a Costa Rican agroforestry case study. In; Reifsnyder,
W.E. and Darnhofer t T. (oos). Applicadon of Meleorology in Agroforestry Systems
Planning and Management. Proceedings of an ICRAFIWMO/UNEP Worklhop.
Nairobi. Kenya.

210
Appendix N SUGGESTED EQUIPMENT AND REAGENT LISTS
FOR ROUTINE TSBF ANALYSES
John Ingram

N.I EQUIPMENT AND NON-REAGENT CONSUMABLES

This list is intended to serve as a basis for those wishing to establish a laboratory (or upgrade
ob5Olete equipment) 50 as to undertake measurements for TSBF studies. Prices (see helow)
are given to assist with budgeting. Although lengthy, the list is not exhaustive.

The quantity (q) given for each item is a guide, but may he considered to he the minimum
for a lab analysing perhaps 500 - 1000 samples per year, in batches of 24 (including 2 repeats
and 2 blanks), for most of the analyses in this Handbook. The unit priees (up; UK!) are
correct for April 1992 for the quoted suppliers (s; see Section N.2) but are subject to change.
The indicated items are by no means the only product that cao be used for a particular task,
but they have been found to he very satisfactory; similar (or identical) items cao be obtained
from various other suppliers both in the UK and elsewhere, but of course the priees may
vary. It is strongly recommended that an updated/recent quotation is obtained prior ta placing
orders.

Note: When ordering electrical equipment remember to specify both the voltage (V) and
A.C. current frequency (Hz) required.

i Item
s Supplier
c Supplierls code
q Quantity
up Unit priee (UKi)
tp Total line price (UIŒ)

Note: The price (in UK f) does not include paclang, freight or insurance.

i s c q up tp

1. AAS Igniter U 10171 1 12.00 12.00

2. AAS Air filter U BS303-0229 85.00 85.00

3. AAS Air compressor U BS303-0313 1 462.00 462.00

4. AAS Air tilter (replacement) U BS303-2471 1 7.00 7.00

211
TSBF: A HanJbooI.: 0/ M~dwch

5. AAS Acetylene regulator U BS303-0106 1 100.00 100.00

6. AAS &haulit liylitem U BS303-1406 1 212.00 212.00

7. AAS (with CalMg lamp) U Buck 200A 1 6545.00 6.545.00

(remember to specify gas fitting si7.es)

8. Aspirator, polythene, 9 litre F ASP-510-030W 5 8.58 42.90

9. Aspirator, polythene, 23 litre F ASP-560-01 OQ 5 10.36 51.80

10. Balance 200 x O.Olg F BCR-32o-A 2 525.30 1,050.60

11. Balance 600 x 0.18 F BCR-330-Q 1 364.62 364.62

12. Balance 310 x O.OOlg F BFH-650-01OT 1 1310.78 1,310.78

13. Beaker, Pyrex, looml (10 pack) F BNB-300-090M 5 17.55 87.75

14. Bealœr t Pyrex, 600ml (10 pack) F BNB-320-190Y 1 28.55 28.55

15. Bea1œr. Pyrex. lOOOnù (5 pack) F BNB-3OQ-230W 2 23.93 47.86

16. Beaker, plastic, 5 litre F BNH-700-170F 5 14.47 72.35

17. Bottle, screw cap (30ml) F BTF-600-03OG 5 10.63 53.15

(for ground soil stomge) (40 pack)

18. Bottle. wide mouth. 500ml F BTF-600-13OC 3 12.32 36.96

(for texture analysis) (20 pack)

19. Brush, test tube, nylon, 30mm F BUR-580-030L 5 2.73 13.65

(10 pack)

20. Brush. fiat form. brisUe F BUR-75O-C 5 1.28 6.40

21. Burette, 5ml x 0.05ml F BWF-605-05OG 5 10.73 S3.6S

22. Burette, 25m1 x 0.1 ml F BWF-610-110X 5 8.65 43.25

23. Burette, 50ml x 0.1 ml F BWF-610-15OC 5 9.06 45.30

24. Centrifuge buclret, (6 pack) B 306/0071/36 1 87.00 87.00

25. Centrifuge (Denly MkIV) B 306/0071/00 1 1226.00 1,226.00

26. Centrifuge angle head (6x50ml) B 306/0071/11 1 186.00 186.00

27. Centrifuge tubes (25 pack) 50 ml B 402/0325/06 4 45.00 180.00

212
 Appnu/ù N: SlIggeSled Eqllipmenl and Reagenl ibIS Jor RolIIine TSBF Analylel

28. Crucible lids (10 pack) F CWB-720-531 D 3 12.74 38.22

29. Crucible (10 pack) F CWB-710-0301 3 11.84 35.52
30. Cylinder, mea.suring, 10ml F CYL-302-0300 2 4.12 8.24
(3 pack)

31. Cylinder, measuring, l00ml F CYL-302-()9()L 5 4.53 22.65

(2 pack)

32. Cylinder, mea.suring, l000ml F CYL-302-15OT 5 16.48 82.40

(2 pack)

33. Deioniser F DCF-500-01OW 1 356.25 356.25

34. Deioniser, resin refill (2 pack) F DCF-505-OlOP 2 96.90 193.80
35. Deioniser, spare batteries F BMT-540-0501 2 0.91 1.82
36. Desiccator, glass F DES-250-050Y 2 68.55 137.10
37. Diluter, Brand F DHL-25O-X 2 238.78 477.56
38. Dispenser, 2-10ml, Brand F DHT-350-050D 2 121.58 243.16
39. Dispenser, 5-25ml, Brand F DHT -350-070U 2 168.99 337.98
40. Dispenser, 10-50ml, Brand F DHT-350-090X 2 235.10 470.20
41. Dispenser bottle, l000ml F BTF-680-130P 10 4.66 46.60
42. Dispenser bottle, 2000ml F BTF -680-1501 5 10.79 53.95
43. Distilled water plant F WGS-522-010Y 1 448.57 448.57
44. Distilled w. p.. spare hoUer F W400/B 1 130.78 130.78
45. Distilled w. p., spare condenser F WC48/M2 1 155.19 155.19
46. Distilled w .p., F A6/6 1 132.10 132.10
spare heater + thennostat

47. Drying cabinet, 120 0 C F OVL-235-O lOY 1 1311.00 1,311.00

48. Drying cabinet. extra shelf F OVL-240-071L 1 46.00 46.00
49. Drying tray 200 x 300 x 50cm L 100 1 100.00
(galvanise<! metal or wood)

213

L __________.,_"_.,,_.,,_________________
 TJ1J/I': A HandbooJ: Di U.dwth

50. Ee meter F CRT-500-0lOF 1 266.46 266.46

~1. EC meter. temperature probe F CRT-505-500E 1 ~4.90 ~4.90

52. EC meter, battery F DM!"-68()..09QH 2 2.81 5.62

53. Flame photometer F FGA-351-F 1 2277.33 2.277.33

~4. Flame photometer compressor F PXW-790-W 1 179.64 179.64
~~. Fl~k. Erlcnmeyer. Pyre}!;.. lOOml F FHB-375-090L 6 17.92 107.~2
(10 pack)

56. Flask. Erlenmeyer, Pyrex, 250ml F FHB-375-13OC 3 19.41 58.23

(10 pack)

~7. Flask. Erlenmeyer. Pyrex. 500ml F PHB- 375-1~OT 1 26.82 26.62

(10 pack)

56. Fluk. Buchner. Pyrex, 500ml F PHD- 355-070L 3 20.46 61.38

59. Flask, volumetric, l00ml F FHM-365-150Y 6 29.17 175.02
(5 pack)

60. Flask, volumetrie, 2~Oml F FHM -365-19OT 15 16.19 242.85

(2 pack)

61. Flask, volumetIic, l000ml P FHM-36~-230K 4 28.72 114.88

(2 pack)

62. Fritch PIS Plant grinder M Pulverisette15 1 3250.00 3,250.00

(with collector and 0.2Smm sieve)

63. Frilch stand for grinder M Stand 1 353.00 353.00

64. Funnel (3"), plastic, (10 pack) F FPH-450-030J ~ 2.89 14.45

65. Funnel, Buchner t 90mm F FPL-390-090A 3 17.83 53.49

66. Funnel, separating, 2S0ml F FPM-620-090P 2 20.12 40.24

67. Heater block lOO-450°C T D84/26 2 2320.00 4,640.00

(with insert for 12 tubes, 12 tubes and rack)

68. Hearer block, spare tubes (6 pack) T F6570 4 139.00 556.00

69. Hydrometer, jar (60 }!;. 440 mm) F HYD-800-130R 25 15.38 384.50
70. Hydrometer, soit F SLR-250-G 2 31.43 62.86

214
 Appendix N: Sllggesred Eqllipmntl and Reagml Usrs for Rollline TSBF Analyses

71. Magnetic stirrer/heater F SWf-515-QlOU 2 239.20 478.40

72. Magnetic follower 45mm (3 pack) F SWX-260-130M 2 2.79 5.58
73. Magnetic follower retriever F SWX-510-N 2 5.30 10.60
74. Mains p1ug 13 A F ECA-352-J 10 0.78 7.80
75. Micropipette, Brand (O.lml) F PMP-721-11OT 2 97.41 194.82
76. Micropipette, tips (1000 pack) F PMP-724-502H 2 11.30 22.60
77. Micropipette, 1-5m1 F PMP-850-050L 1 136.99 136.99
78. Micropipette, tips (10 pack) F PMP-772-506W 5 10.66 53.30
79. Microscope, dissecting/light F MIB-ABA-ü20M 1 178.10 178.10
80. Mixing plunger, sedimentation L 1 3 3.00
81. Muffle furnace 13A, 2.5 litre F FSE-285-01OA 1 1088.00 1,088.00
82. Oven, drying/fan, 97 litre F OVB-306-11OT 1 847.36 847.36
83. Pestle and mortar F MWA-300-071S 3 11.03 33.09
84. pH meter, bench F PHL-203-010W 1 683.20 683.20
85. pH meter, portable F PHK-080-01OK 2 44.65 89.30
86. pH meler, portable, battery (4 pk) F BMT-620-120U 1 3.59 3.59
87. pH electrode F PHQ-300-030D 3 72.46 217.38
88. Photocopier (e.g. Canon NP 112) L 1 1000.00 1,000.00
89. Pipette, 1ml (2 pack) F PMC-285-020J 5 12.56 62.80
90. Pipette, 2m1 (2 pack) F PMC-285-030G 5 12.56 62.80
91. Pipette, 5ml (2 pack) F PMC-285-070R 5 13.55 67.75
92. Pipette, 10mi (2 pack) F PMC-285-080X 5 14.91 74.55
93. Pipette, 2Sml (2 pack) F PMC-285-11QY 5 18.34 91.70
94. Pipette tiller, rubber F PMR-330-090E 3 5.93 17.79

215
188F: A HanJbook. 0/ McthocLt

95. Plastic buckets L 5 3 15.00

96. Voltage surge protectors F ECA-78o-X 5 16.24 81.20

97. Polythene bottle, 60ml F BTK-390-050D 10 3.51 35.10

(10 pack)

98. Polythene bottle 150ml J F BTK-390-070U 20 4.40 88.00

no pack)
99. Refrigerator. 215 litre F RFP-300-0ION 1 367.71 367.71

100. Safety lab coats L 10 5 50.00

101. Safety eye wash station F SAP-740-W 1 16.83 16.83

102. Safety eye wuh hottle F SAP-714-T 2 5.23 10.46

103. Safety BeF fire extinguisher F SAP-870-Q 1 64.78 64.78

104. Safety eye specs. F SAP-290-E 10 1.80 18.00

105. Safety glove&t PVC F SAR-89~70H 2 5.40 10.80

106. Safety gloves latex (50 pack)

J F SAR-710-051G 1 28.71 28.71

107. Shaker SM25

J F SGM-540-025T 1 1952.23 1.952.23

108. Sieve. brass. O.25mm F SIH4oo-220G 2 36.92 73.84

109. Sieve, brass, lid F SIH-470-504M 2 16.73 33.46

110. Sieve. S.Steel. 2mm F SIH-460-1ooU 2 44.41 88.82

Ill. Sieve. ~iver F SIH-470-506Q 2 19.47 38.94

112. Specrrophotomerer, CoU C CE 2010 1 2640.00 2.640.00

113. Spectrophotometer C 20202100 1 1410.00 1,410.00

Micro Sipette control

114. Spectrophotometer C 1010 3200 1 165.00 165.00

Holder for Micro Sipette cclI

115. Spectrophotometer C 2030703 1 390.00 390.00

Micro Sipette Flow œIl (Silica 20mm)

116. Spt:(;trophotometer t 10mm œIl C 3030726 3 14.00 42.00

(for use without Micro Sipette system)

216
 Appendbc N: SlIggelled Eqllipmou and Reagml UIII for RolllÙU! TSBF AIIoIylel

117. Spectrophotometer, spare fuses C 20202700 1 23.00 23.00

118. Spectrophotometer, spare Iamp C 23030140 1 75.00 75.00
(2 pack)

119. Spin mixer F SGP-202-01OJ 2 173.04 346.08

120. Stoppers (50 pack) F BTF-376-090'1' 1 5.69 5.69
121. Stopwatch F TKP-386-G 1 11.25 11.25
122. Test tube 150 x 16mm F TES-200-191R 5 9.81 49.05
(100 pack)

123. Test tube rack, 16mm test tube F STK-640-070L 10 4.85 48.50
(24 place)

124. Test tube holder (10 pack) F TEW-710-J 1 7.79 7.79

125. Thennometer -10 - 50°C (10 pk) F THL-210-031A 1 15.57 15.57
(for sedimentation)

126. Thennometer -10 - 400°C (10 pk) F THL-210-131T 1 15.14 15.14

(for block heaters)

127. Tongs F TNS-200-010R 5 2.78 13.90

128. Tray, polystyrene, (3 pack) F TRC-200-090R 10 11.59 115.90
129. Trolley F TSF-3S1-A 1 134.31 134.31
130. Tubing, PVC, IOm x 8 x 2mm F TWR-670-17OJ 3 2.84 8.52
131. Tubing connectars, T, 9 mm F ADF-715-070W 1 8.65 8.65
132. UPS Emerson 500VA E Mode130 1 321.00 321.00
133. Vacuum cleaner, cylinder L 1 100 100.00
134. Vacuum pump, oil free F PYB-530-OlOA 1 1001.73 1,001.73
135. Vacuum pump, spare valve seaI F PYB-536-020H 2 19.12 38.24
136. Vacuum pump, spare diaphragm F PYB-536-0IOK 2 23.90 47.80
137. Vacuum pump, water, fil ter F PXY-311-P 2 7.62 15.24
138. Vacuum pump, water, connectors F PXY-31S-500R 2 1.31 2.62

217
 '1'.mF: A If~ ~J MllItNI

139. WuhboWe, 250ml F WBS-400-02OC 10 0.72 7.20

140. Washbottle, 500ml F WBS400-04OT 10 0.86 8.60

141. Water tank L 1 10 10.00
(450 x 1400 x 400 cm deep)

142. Waret bath. 22 litre F BJL-455-21OF 1 450.85 450.85

143. Cotton wool, absorbent, 0.5kg F CTC-270-Q 4 2.64 10.56
144. FiUer papert Whatman 1 9.Ocm F FlID-24o-11OF 10 2.14 21.40
143. Filter paper, Whatman 1 15.Ocm F FHD-240-170K 10 3.71 37.10
146. Filter paper, Whatman 42 9.Ocm F PHD-460-110V 10 6.59 65.90
147. FiUer paper, Whatman 42 15.Ocm F FHD-460-170D 10 15.26 152.60

148. Pen, marker, red, (12 pack) F LAC-71O-020V 1 12.96 12.96

149. Pen, marker, blue, (12 pack) F LAC-710-030S 1 12.96 12.96

150. Pen, marlœr, black, (12 pack) F LAC-710-050M 1 12.96 12.96

N.2 SUGGESTED EQUIPMENT SUPPLIERS

B BDH 1aboratory Supplies Apparatus Division, Merck Bouse, Poole, BH15 9EL, UKt
TEL +44 202 669700, TLX 264772 bdh app, FAX +44 202666542

C Cecil Instruments Limited, Milton T~hnical Centre, Cambridge R.aad, CB4 6AZ. UK,
TEL +44 223 420821, FAX +44 223 420475

E Computer Battery Servi~s Ltd, Unit 8, Riverside Park, Famham, Surrey GUS 7UG,
UK, Tel +44252 714100, Fax +44 252 733910

F Pisons Ltd, BiShop Meadow Road, lnughborough, Leies LEll ORO, UK, TEL +44
509 231166, FAX +44 509 231893

L Local

M Christison Scientific Equipment Ltd, Albany Itoad, Gareshead, NES 3AT, UK, TEL
+44 91 4774261, FAX +44 91 490 0549

U Buck Scientific, Unit 6, Upper Wingbury Courtyard. Upper Wingbury Fann,

Wingrave, Aylsbury, Bucks, UK, TEL +44 402 349136, FAX +44 296 681293

218

~-----~--------------------------------------------~
f Appmdix N: S"gg~.sl~d Eqllipmenl and R~Qgenz U.sI.s for ROUlÎM TSBF AnaIy.s~.s

N.3 REA GENTS

This list gives the minimum quantities (or the smallest commercially available quantity,
.whichever is less) of the major reagents required to analyse 100 samples for each method in
this Handbook. The unit quoted priees (up; UK!) are correct for April 1992, from the
quoted supplier (Merck UK) but are subject to change; alternative reputable suppliers are
however equally good, and priees may vary. It is strongly recommended that a recent
quotation is obtained prior to placing orders.

s Supplier
c Supplier's code
q Quantity
up Unit price (UIŒ)
tp Total line price (UIŒ)

s c q up tp
Aceticacid,Analar 500ml l00013L 1 6.10 6.10
Acetone l000ml 270235S 2 5.40 10.80
Ammonium acetate 500g 271424C 10 6.80 68.00
Ammonium chloride, Analar 500g 1001730 1 6.60 6.60
Ammonium molybdate lOOg 271872W 1 8.70 8.70
Ammonium sulphate, Analar 500g l00333B 1 6.60 6.60
Amyl alcohol 500ml 272124U 1 4.70 4.70
Ascorbic acid lOOg 103033E 1 7.60 7.00
Barium chloride 500g 272904R 1 6.60 6.00
Borie acid 500g 274104B 1 4.30 4.30
Buffer tablets (pH 4) 50 pack 331542Q 1 10.90 10.90
Buffer tablets (pH 7) 50 pack 331552S 1 10.50 10.50
Buffer tablets (pH 9.2) 50 pack 331562U 1 8.50 8.50
Calcium carbonate, Analar 250g l00683U 1 7.80 7.80
Cetyltrimethylammonium bromide 500g 276654L 1 23.90 23.90
Chloroform l000ml 277105X 1 9.70 9.70
Devarda' s alloy 500g 280094F 1 25.10 25.10
Ethanol- 2500ml 283047K 1 16.00 16.00
Ethanol 500ml 283047K 1 16.00 16.00
Ferric nitrate 500g 283844Y 1 13.10 13.10
Ferrous ammonium sulphate 500g 271664Q 1 5.90 5.90
Hydrochloric acid, Analar* 2500ml 103076P 1 8.50 8.50
Hydrochloric acid, Analar lOOOml 101255Y 3 5.90 17.70
Hydrogen peroxide (30%)' 1000ml 286945B 1 7.20 7.20
Lanthanum chloride lOOg 103433Q 1 19.80 19.80
Lithium sulphate, Analar 500g 101474R 1 18.30 18.30
Magnesium sulphate, Analar 500g 101514Y 1 7.00 7.00
Methanol * 2500ml 2919260 1 8.00 8.60
Methanol 500ml 291924E 5 3.50 17.50
Methoxyethanol 500ml 103624V 1 5.50 5.50
Methylpropan-2-o1 500ml 103584H 1 6.80 6.80

219
TSBF: Il Handbool 01 M~r1wtb

Octan-2-o1 2500ml 294106N 1 16.20 16.20

Orthophosphoric acid, Analar 500ml 101734S 1 6.90 6.90
Oxalic acid 500g 294234U 1 10.10 10.10
pH 4.5 indicator solution 500ml 210414M 1 8.10 8.10
Phenolphthalein 100g 200892K 1 7.00 7.00
Phosphomolybdic &cid 100g 292713W 1 20.20 20.20
Potassium acetate 500g 29S814P 1 5.10 5.10
Potassium chloride 1000g 295945C 5 5.10 25.50
Potassium chloride, Analar 250g 10 1983K 1 4.40 4.40
Potassium dichromate 500g 296054D 1 14.70 14.70
Potassium dihydrogen phosphate 500g 102034B 1 8.40 8.40
Potassium nitrate, Analar 250g 102143F 1 4.80 4.80
Potassium permanganate 500g 296444N 1 6.10 6.10
Potassium sulphate 2500g 296585C 2 11.20 22.40
Salicylic acid 500g 300384B 1 7.30 7.30
Selenium powder 100g 300454V 1 20.20 20.20
Silver nitrate 25g 300872M 1 24.30 24.30
Silver sulphate 2!ig 102343L 1 30.80 30.80
Sodium bicarbonate 500g 102474.V 1 4.50 4.50
Sodium carbonate, Analar 500g 102394W 1 5.20 5.20
Sodium chloride, Analar 500g 102414] 1 4.50 4.50
Sodium citrate 500g 301284C 1 5.20 5.20
Sodium hexametaphosphate 500g 307344V 1 6.20 6.20
Sodium hydroxide 500g 30 1674M 1 4.50 4.50
Sodium hypochlorite solution 250ml 230393L 5 10.50 52.50
Sodium nitroprusside loog 301902F 1 10.40 10.40
Sodium salicylate 500g 302104K 1 8.20 8.20
Sodium tartrate 500g 3022740 1 10.60 10.60
Sodium tungstate 500g 302374H 1 27.50 27.50
Sodium thiosulphate l000g 302355E 1 4.60 4.60
Sucrose, Analar 500g 102744B 1 5.40 5.40
Sulphuric acid* 2500ml 303253D 4 7.91 31.64
Sulphuric acid 1000ml 303254E 10 5.44 54.40
Sulphuric acid, Analar l000nù 102760B 2 7.40 14.80
Tannic acid ~OOg 303374L 1 19.80 19.80

Note: The priee (UKf) doe3 not inc1ude pacldng, freight or insuraoCè.
• The 2500ml size is only allowed on a cargo aircraft; if this item is required to a
destination where only a passenger service operates, select the sma1ler sile given as the
, following item on the lista
This product is not allowed on an aircraft; try 10 obtain locally.

N.4 SUGGESTED REA GENT SUPPLIER

BDH Laboratory Supplies, Merck Ltd., Poole, Dorset, BH15 ITD, UK; Tel +44 202
660444; Fax +44 202 666856; Tlx 41186 or 418123 tetra g.

220
 Appendix 0 TSBF DATA SHARING POUCY

It is intended that participation in the TSBF Programme will require collaboration not only
in methodology. but al50 in the exehange of data. A clear statement about the ownership of
data was deemed necessary.

1. Ali raw material and derived local data remain the property of the researeh worlœr
who obtained it. No attempt ta use sueh data will be made in TSBF publications,
other than with the permission of the researeh worker concerned.

2. Eaeh worker is encouraged ta publish their work in the nonnaI way, and ta inelude
reference ta TSBF if they 50 wish. Reprints of, or references ta, sueh works would
he gratefully received by the TSBF Office, as part of the coordination exercise. TSBF
wou Id al50 aim ta use sueh references ta support applications ta funding ageneies; the
more field studies that can be eited, the greater the chance of being awarded funds.

3. One of the aims of the TSBF Programme is to compare experimental variables

measured in different ecosystems, and then ta conduet syntheses of the data from eaeh
research site. Once synthesised, data from specifie sites willlose their individuality,
and become part of the TSBF regional, or ultimately global pieture. Those condueting
a synthesis exercise on behalf of TSBF may feel they have derived sufficient data to
merlt publication in its own nght, and they should be encouraged to publish in an
appropriate journal. Any sueh paper would include suitable acknowledgement of
contributing research worlœrs, and state that it constitutes pan of the TSBF
Programme. Should published data be used in the synthesis, references must be made
in the normal style of a review. The authors will be encouraged to circulate a draft
to those workers whose data were used in the synthesis, before submission ta journaIs
is made.

221

1 !
View publication stats