Methods in Agricultural Chemical Analysis: A Practical Handbook (CABI Publishing)

METHODS ANALYSIS

AGRICULTURAL CHEMICAL

A Practical Handbook

To Alwyn, Antonia and Bethan

METHODS IN AGRICULTURAL CHEMICAL ANALYSIS A Practical Handbook

N.T. Faithfull Institute of Rural Studies University of Wales Aberystwyth UK

CABI Publishing

CABI Publishing is a division of CAB International CABI Publishing CABI Publishing CAB International 10 E 40th Street Wallingford Suite 3203 Oxon OX10 8DE New York, NY 10016 UK USA Tel: +44 (0)1491 832111 Fax: +44 (0)1491 833508 E-mail: [emailprotected] Web site: www.cabi-publishing.org

Tel: +1 212 481 7018 Fax: +1 212 686 7993 E-mail: [emailprotected]

©CAB International 2002. All 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, London, UK. Library of Congress Cataloging-in-Publication Data Faithfull, N.T. (Nigel T.) Methods in agricultural chemical analysis: a practical handbook / N.T. Faithfull. p. cm. Includes bibliographical references (p. 206). ISBN 0–85199–608–6 1. Soils--Analysis--Handbooks, manuals, etc. 2. Plants--Analysis--Hand-books, manuals, etc. 3. Chemistry, Analytic--Handbooks, manuals, etc. I. Title. S593 .F19 2002 630’.2’43--dc21 2002005768 ISBN 0 85199 608 6 Typeset by Wyvern 21 Ltd, Bristol Printed and bound in the UK by Biddles Ltd, Guildford and King’s Lynn

Contents

Preface About the Author Disclaimer Acknowledgements Abbreviations and Acronyms Chapter 1 Experimental Planning Experimental Design Plot size Equipment Considerations Autoanalysis Samplers Peristaltic pumps Chemistry module Heating bath and dialyser Colorimeter and spectrophotometer Chart-recorders Chart reader Flow injection Batch Size Sampling Protocol Soils Composts Feeds

xii xv xvi xvii xx 1 1 2 2 2 4 4 5 5 6 6 7 8 8 9 9 10 10 v

Contents

Plant components Microbiological analysis Biological substances Fertilizers

15 16 16 16

Chapter 2 Sample Preparation Pre-treatment of Samples and Sample Contamination Trace Element Analysis Sub-sampling Drying Techniques Air-drying Oven-drying Vacuum oven Freeze-drying Desiccation Milling, Grinding and hom*ogenization Freezer mill hom*ogenization Storage of milled samples

17 17 17 18 19 19 19 19 20 22 22 23 23 24

Chapter 3 Weighing and Dispensing Weighing Errors Corrections of weighings in vacuo Incorrect calibration of the balance Static charge Convection currents Absorption of moisture by the sample Absorption of moisture by the sample container Dispensing Errors Bottle top dispensers Syringe pipettes

26 26 26 27 27 27 27 28 28 29

Chapter 4 Acid-digestion, Ashing and Extraction Procedures Acid-digestion and Washing Acid-digestion of soils Total soil nitrogen Acid-digestion of plant materials Digestion systems Distillation systems Microwave acid-digestion Dry ashing Extraction Procedures – Plant-based Materials Oils, fats and waxes Fibre, lignin, cellulose, nitrogen-free extract and starch In vitro digestibility Nitrate and water-soluble carbohydrate Water content in silage

30 30 30 31 32 32 32 34 35 36 37 38 42 48 50

vii

Contents

Extraction Procedures – Soils pH extractants Phosphate extractants Potassium extractants Trace element extractants

50 51 52 54 54

Chapter 5 Analysis of Soil and Compost Soil Analytical Procedures Method 5.1. Determination of extractable boron Method 5.2. Cation exchange capacity, exchangeable bases and base saturation Method 5.3. Determination of effective cation exchange capacity (ECEC) Method 5.4. Determination of fulvic and humic acids Discussion 5.5. Determination of available nitrogen Method 5.5a. Determination of nitrate by selective ion electrode Discussion 5.5b. Determination of total mineralized nitrogen Method 5.5b.i. Determination of extractable ammonium-N Method 5.5b.ii. Determination of extractable nitrate-N Discussion 5.6. Determination of organic plus ammonium nitrogen Method 5.6a. Determination of soil nitrogen by autoanalysis Method 5.6a.i. Reduction of nitrate before digestion and colorimetric analysis Method 5.6b. Determination of organic plus ammonium-N by digestion and distillation Discussion 5.7. Determination of soil organic matter Method 5.7a. Determination of soil organic matter by loss on ignition Method 5.7b. Determination of easily oxidizable organic C by Tinsley’s wet combustion Discussion 5.8. Determination of pH and lime requirement Method 5.8a. Measurement of pH Method 5.8b. Determination of lime requirement Method 5.8c. Determination of pH in soils with soluble salts Discussion 5.9. Determination of extractable phosphorus Method 5.9a. Determination of extractable phosphorus (manual method) Method 5.9b. Determination of extractable phosphorus (automated method) Method 5.9c. Determination of resin extractable phosphorus (automated method) Method 5.10. Determination of extractable magnesium, potassium and sodium Method 5.11. Determination of extractable trace elements Discussion 5.12. Determination of extractable sulphur Method 5.12a. Determination of extractable sulphur (manual method)

57 57 57 59 66 68 71 71 72 73 74 74 75 75 76 78 78 79 81 82 82 83 84 84 86 87 89 91 93 94

viii

Contents

Method 5.12b. Determination of extractable sulphur (automated method) The Analysis of Composts Method 5.13. Determination of CEC in composts Method 5.14. Determination of Ca, K, Mg and P in composts Method 5.15. Determination of heavy metals in compost

96 101 103 104

Chapter 6 The Analysis of Fertilizers Fertilizer Analytical Procedures Discussion 6.1. Determination of total nitrogen in presence of nitrate and organic N Method 6.1a. Determination of total nitrogen in presence of nitrate and organic N, with final determination by distillation Method 6.1b. Determination of total nitrogen in presence of nitrate and organic N, with final determination by autoanalysis Discussion 6.2. Determination of phosphorus in fertilizers Method 6.2a. Determination of water-soluble phosphorus (extraction) Method 6.2a.i. Determination of water-soluble phosphorus (autoanalysis) Method 6.2a.ii. Determination of water-soluble phosphorus (manual method) Method 6.2b. Determination of 2% citric acid-soluble phosphorus – method for basic slags (Thomas phosphate) Method 6.2c. Determination of total phosphorus in the acid digest from Method 6.1b. with final determination by autoanalysis Discussion 6.3 Determination of potassium in fertilizers Method 6.3a. Determination of water-soluble potassium Method 6.3b. Determination of ammonium oxalate-soluble potassium Method 6.3c. Determination of potassium in the acid digest from Methods 6.1a. or 6.1b Liming Materials Method 6.4. Determination of the moisture and neutralizing value of liming materials Method 6.5. Determination of fineness of grinding (150 µm/100 mesh fraction)

106 107

Chapter 7 The Analysis of Animal Feed and Plant Materials Discussion 7.1. Determination of acid detergent fibre, cellulose and lignin Method 7.1a. Determination of acid detergent fibre Method 7.1b. Determination of lignin Method 7.1c. Determination of cellulose and ash Method 7.2. Determination of crude fibre Method 7.3. Determination of modified acid detergent fibre (MAD fibre)

124

107 108 109 110 114 114 116 117 118 119 120 120 121 121 122 123

125 125 126 127 128 130

Contents

Method 7.4. Determination of neutral cellulase plus gamanase digestibility (NCGD) of feeding stuffs 131 Method 7.5. Determination of neutral detergent fibre (NDF) or plant cell-wall constituents 133 Method 7.6. Determination of nitrate in plant material by autoanalysis 135 Discussion 7.7. Determination of total nitrogen (crude protein) in plant material and feeding stuffs 137 Method 7.7a. Determination of total nitrogen (crude protein) in plant material by autoanalysis 138 Discussion 7.8. Determination of oil in feeding stuffs by extraction with petroleum spirit 141 Method 7.8a. Determination of oil in feeding stuffs by extraction with petroleum spirit 142 Method 7.8b. Determination of oil in rapeseed 142 Method 7.9. Determination of pepsin–cellulase digestibility of plant material 143 Discussion 7.10. Determination of total phosphorus in plant material and feeding stuffs 144 Method 7.10a. Determination of total phosphorus in plant material by autoanalysis 145 Discussion 7.11. Determination of total potassium in plant material and feeding stuffs 146 Method 7.11a. Preparation of plant sample solution by dry combustion 147 Method 7.11b. Determination of potassium in plant material by flame photometry (dry ashing extract) 148 Method 7.11c. Determination of potassium in plant material by flame photometry (Kjeldahl acid digest) 148 Discussion 7.12. Determination of starch by acid hydrolysis 149 Method 7.12a. Determination of starch in potatoes by hydrolysis and autoanalysis 149 Discussion 7.13. Determination of trace elements in plants and feeds 150 Method 7.13a. Determination of trace elements in plants and feeds 151 Method 7.14. Determination of water soluble carbohydrate by autoanalysis 151 Chapter 8 The Analysis of Silage Method 8.1. Determination of ammonium-N in silage Method 8.2. Determination of moisture in silage Method 8.3. Determination of pH in silage Discussion 8.4. Determination of volatile fatty acids (VFAs) in silage Method 8.4. Extraction method for obtaining silage juice for analysis for VFAs

154 154 156 159 159 164

Contents

Chapter 9 Near Infrared Spectroscopy Prediction of Metabolizable Energy (ME) Applications of NIR Interpreting NIR Spectra Interferences

167 167 168 169 170

Chapter 10 Methods in Equine Nutrition Toxic Effects of Some Elements Method 10.1a. Application of ytterbium marker to feed Method 10.1b. Feeding of ytterbium marked feed and faecal collection and preparation Method 10.1c. Preparation of ytterbium marked feed for analysis Mobile Bag Technique (MBT) Method 10.2. Determination of digestibility using the mobile bag technique Hemicellulose Non-starch polysaccharides (NSP) Method 10.3. Determination of total non-starch polysaccharides

172 172 175

Chapter 11 Methods for Organic Farmers and Growers Origins Balance Albrecht Basic cation saturation ratio Other ratios N:K balance Trace elements Fertilizers Commercial Analytical Services BCSR versus SLAN Phosphate Analysis Organic phosphorus Determination of organic P Method 11.1. Determination of extractable organic and inorganic soil P The Balzer Methods pH determination Humus Humification (see also Chapter 5, Method 5.4) Phosphorus, potassium and magnesium Trace elements The UK Situation

187 187 188 189 189 190 190 190 191 191 192 193 193 194

Chapter 12 Quality Assurance and Control Replicates Standard Reference Materials Quality Systems

200 201 202 203

176 177 177 179 181 182 183

194 196 197 197 198 198 199 199

Contents

Matrix Interference Standard Additions, or ‘Spiking’

204 204

References Appendix 1: Equipment Suppliers Appendix 2: Soil Index Table Appendix 3: Lime Application Rates for Arable Land Appendix 4: Lime Application Rates for Grassland Appendix 5: Nitrate and Nitrite in Soil, Plant and Fertilizer Extracts Appendix 6: Phosphate in Soil, Plant and Fertilizer Extracts Appendix 7: Analytical Methods Used by ADAS for the Analysis of Organic Manures Appendix 8: Laboratory Safety Appendix 9: Chemical Composition Data Sources for Plants, Feeds, Blood, Urine and Soils Appendix 10: Atomic Weights, Units and Conversion Tables

206 223 225 226 227 228 233

Subject Index Commercial Index

259 265

241 247 251 253

Preface

The need for this publication has arisen in four ways. The first is that relatively few staff engaged in agricultural research in educational institutions have sufficient knowledge of chemistry to make informed decisions regarding choice of the most suitable analytical method for their purposes. For example, an unsuitable sample drying process can destroy or seriously degrade the component being estimated. Second, there has been a need for a book containing methods of soil and crop analysis suitable for use in undergraduate practical classes. Lecturers under pressure to carry out publishable research and burdened with administrative duties have little time for scouring libraries and the Web for such methods. For the benefit of those lacking much experience in laboratory experimentation, the methods are described in greater practical detail than found in many publications. Third, the useful manual The Analysis of Agricultural Materials, MAFF/ADAS Reference Book 427, HMSO, 1986, is now out of print. Lastly, the growth in organic farming, and the establishment of the Organic Farming Centre for Wales, funded by the National Assembly for Wales and based in the University of Wales, Institute of Rural Studies, Aberystwyth, has engendered a fresh interest in analytical methods more suitable for sustainable agriculture, and a chapter is included on this area of analysis. The nature of the contents will be determined by the practicability of the methods in undergraduate teaching, by their acceptability for research publications, and by their affordability by public sector institutions. The use of very expensive instruments may be referred to, but not described in detail. This background knowledge will assist the choice of whether to send samples xii

xiii

Preface

away for analysis. The methods have not been chosen for their suitability in legal proceedings, although references to such will be made, and where published on the World Wide Web, the respective websites will be given. These official methods tend to be more elaborate, longer to perform, and far more rigorous than required in our case. The use of the web is growing apace, and website addresses will often be inserted in the text to aid further research. There is no attempt to include every possible procedure, but to provide the most useful selection. It is anticipated that another author will publish a volume concentrating on chemical analysis dealing with ruminant animal nutrition. To avoid duplication, this volume will not cover that area in depth.

Soils and composts Analyses will be those in common use for soils from fields for both grass and arable crops. MAFF/ADAS (now DEFRA) methods and Index Tables are reproduced by permission of the Controller of Her Majesty’s Stationery Office (Ref. 20001327). Analyses for nitrogen mineralization are included. Special consideration will be given to composts and recycled urban waste.

Fertilizers It is quite common for the researcher to check the specified minerals as stated on the label. Some methods will cover the usual elements.

Plant materials Research methods demand large numbers of samples. The only way the throughput can be achieved is by using some form of automatic processing. Such methods using segmented flow analysers were conceived by Skeggs (1957) for use in clinical analysis, but found wide application in water, soil and plant analyses in the mid-1960s. In 1968, the author established an analytical laboratory at the University of Wales, Aberystwyth, with the remit to install this type of equipment for the analysis of plant samples from the Departments of Agriculture, Biology and Biochemistry. Further discussion of segmented flow analysis will be found in Chapter 1. These are the reasons why, in addition to the well-established manual procedures such as for fibre, automatic methods will be preferred if they exist, but references to the equivalent manual methods will usually be provided if available.

Feeds This relates closely to animal nutrition, which may in the future be published

xiv

Preface

elsewhere as indicated previously. Considerations relating to grass, hay, haylage, silage, compound feeds and grain will be given.

Plant components Research samples may contain just one part of the structure of the plant: leaf, stem, root, etc. Certain precautions may be applicable.

Biological substances This will not be covered in depth, but some types of material may occasionally be presented for analysis, so a few selected procedures and references will be given.

Equine nutrition In some ways this is a developing area, and certainly lags behind ruminant nutrition in the published material. The discussion will include any details of recent work in this field.

Microbiological analysis This is really not chemical analysis, but some references are suggested.

About the Author

Nigel Faithfull spent 4 years in the laboratories of RTZ at Avonmouth before proceeding to the University of Wales at Aberystwyth. He graduated with Honours in Chemistry (1968), and immediately took charge of the newly established Agricultural Sciences Analytical Laboratory. Research into automated methods in herbage analysis led to an MSc, and further studies involving atomic absorption spectrophotometry resulted in a PhD in 1975. He is a Chartered Chemist and a Member of the Royal Society of Chemistry. He was a member of college safety committees for about 30 years. Following a merger with the Welsh Agricultural College, the Analytical Laboratory is now located at the UWA Institute of Rural Studies, Llanbadarn Campus, Aberystwyth, where a commercial soil analysis service for farmers has been in operation for several years.

N.T. Faithfull, MSc PhD (Wales) CChem MRSC Institute of Rural Studies University of Wales Llanbadarn Campus Aberystwyth Ceredigion SY23 3AL UK

Disclaimer

Trade names in this publication are used only to provide specific information and to illustrate the type of equipment being discussed. Mention of a trade name does not imply an endorsem*nt of that product or constitute a recommendation of it in preference to any other product which is not mentioned. The purchaser of any equipment must ensure it is suitable for the purpose for which it is intended, and compatible with any items to which it is to be connected. All methods should be carried out only by competent persons and with adequate supervision when necessary. All obligations under The Control of Substances Hazardous to Health Regulations 1999 (COSHH), should be observed, and risk-assessment documentation completed. Appropriate personal protective equipment should be provided and worn whenever recommended. Persons carrying out the procedures in this manual do so entirely at their own risk, and neither the author, publishers, or anyone mentioned in, or connected with this publication can be held in any way responsible for any accidents no matter how caused.

xvi

Acknowledgements

First, I would like to express my gratitude to the teachers, lecturers and industrial scientists who have instilled a high regard for practical analytical chemistry, with the need for care, accuracy, and the development of a skilful and safe technique. There are a number of individuals who have been most helpful in my search for information. It is difficult to remember everyone I have consulted over a period of 18 months, and my sincere apologies for any omissions. They include the following, with the area of advice in parentheses: Professor W.A. Adams (soil science), University of Wales, Aberystwyth, UK F.M. Balzer (analysis for organic farming), Labor Dr F.M. Balzer, Wetter, Germany Zoltán Bodor (fertilizer analysis), Kemira, Finland Joao Coutinho (soil sulphate analysis), Universidade de Trás-os Montes e Alto Douro (UTAD), Portugal Steve Cuttle (soil analysis for organic farming), Institute for Grassland and Environmental Research, Aberystwyth, UK Sue Fowler (organic farming), Institute of Rural Studies, University of Wales, Aberystwyth, UK Professor D.I. Givens (NIR), ADAS Nutritional Sciences Research Unit, Stratford-upon-Avon, UK John Hollies (soil analysis), The Potash Development Association, Laugharne, Carmarthen, UK D. Iorwerth Jones (cellulase digestibility) Sue Lister (NIR), Institute for Grassland and Environmental Research, Aberystwyth xvii

xviii

Acknowledgements

Bob Llewelyn (organic manure analysis; ADAS analytical methods), ADAS Laboratories, Wolverhampton, UK Peter J. Loveland (ADAS and BCSR soil analysis approaches), Silsoe Campus, Cranfield University, UK Ramadan Al-Mabruk (animal nutrition) Institute of Rural Studies, University of Wales, Aberystwyth, UK Isabel McMann (references), Bodleian Library, Oxford, UK Tim Meeks (information), Tennessee Valley Authority Meriel Moore-Colyer (equine nutrition), Institute of Rural Studies, University of Wales, Aberystwyth, UK Cornelia Moser (reference), Landwirtschaftlicher Informationsdienst (LID), Bern, Switzerland Suzanne Padel (organic farming), Institute of Rural Studies, University of Wales, Aberystwyth, UK Dan Powell (organic farming), Aberystwyth, UK David Rowell (soil analysis), Department of Soil Science, University of Reading, UK Mustapha Bello Salawu (digestibility spreadsheets; silage VFA analysis), Institute of Rural Studies, University of Wales, Aberystwyth; now at Commonwork Group Ltd, Kent, UK Steve Smith (reference), Stapledon Library, Institute for Grassland and Environmental Research, Aberystwyth, UK Stuart Smith (autoanalysis methodology), Divisional Manager, BL-Analytics, Bran+Luebbe, UK Rebecca Stubbs (editorial support), CABI Publishing, Wallingford, UK Claudine Tayot (reference), OPOCE online helpdesk Paul Thomas (NSP methodology), Institute for Grassland and Environmental Research, Aberystwyth, UK Ann Vaughan (lab techniques), Institute of Rural Studies, University of Wales, Aberystwyth, UK Joanne Vessey (fertilizer analysis), Hydro Agri (UK) Limited, Immingham Dock, UK Keith Way (NCGD method), Laboratory Consultant Gabriele Weigmann-Dramé (reference), VDLUFA-Verlag, Darmstadt, Germany Lorraine Whyberd (information), Royal Society of Chemistry, London, UK David Wilman (Fig. 1.5; agronomy), Institute of Rural Studies, University of Wales, Aberystwyth, UK I appreciate the efforts of the staff of the Huw Owen and Thomas Parry Libraries, University of Wales, Aberystwyth, for retrieving large old tomes from the store in an attempt to trace the origins of some methods. I am also grateful to the university for permission to use the facilities of the Information Services subsequent to taking early ‘retirement’. I would like to thank ADAS, Wolverhampton, for permission to reproduce the information in Appendix 7; Bran+Luebbe, for permission to reproduce the methods in Appendices 5 and 6; and the Controller of Her Majesty’s

Acknowledgements

xix

Stationery Office for a licence to reproduce material from the MAFF/ADAS publications: The Analysis of Agricultural Materials Reference Book 427, and Fertiliser Recommendations Reference book 209 5th and 6th edns. Finally, many thanks to my wife Eileen, for her continuous support, patience and encouragement throughout this task.

Abbreviations and Acronyms

AA AAS ac ADAS ADF AES AOAC BCSR CEC cmolc COSHH CP CRM CSL DEFRA detn DM DMD DMSO DOMD xx

atomic absorption atomic absorption spectrophotometry alternating current Agricultural Development and Advisory Service (UK) acid detergent fibre atomic emission spectrometry Association of Official Analytical Chemists basic cation saturation ratio cation exchange capacity centimole charge Control of Substances Hazardous to Health Regulations (1999) crude protein Certified Reference Material Central Science Laboratory Department for Environment, Food and Rural Affairs (UK) determination dry matter dry matter digestibility dimethylsulphoxide digestible organic matter in dry matter

xxi

Abbreviations

DTPA ECEC ED EDTA

EPA FAPAS FEPAS FIA FP GC GLC GLP HDPE HPLC HS HSE ICP ICP-MS ICP-OES ID IGER IMS IR ISE ISO IV IVDMD LGC m MADF MAFF (now DEFRA) MBT ME mol MS NAMAS

diethylenetriaminepentaacetic acid, or diethylenetrinitrilopentaacetic acid effective cation exchange capacity effective degradability diaminoethanetetraacetic acid, ethylenediaminetetraacetic acid, or (ethylenedinitrilo)tetraacetic acid Environmental Protection Agency (US) Food Analysis Performance Assessment Scheme Food Examination Performance Assessment Scheme flow-injection analysis flame photometry gas chromatography gas–liquid chromatography Good Laboratory Practice high density polyethylene high-performance liquid chromatography humic substances The Health and Safety Executive (UK) inductively coupled plasma emission spectroscopy hyphenated technique of ICP followed immediately by MS inductively coupled plasma optical emission spectroscopy inside diameter Institute for Grassland and Environmental Research industrial methylated spirits infrared ion-selective electrode International Organization for Standardization in vitro in vitro dry matter digestibility Laboratory of the Government Chemist mass (when used in % m/m) modified acid detergent fibre Ministry of Agriculture, Fisheries and Food (UK) mobile bag technique metabolizable energy molecular weight in grams (relative molecular mass) mass spectrometry National Measurement Accreditation Service (now FAPAS)

xxii

Abbreviations

NCGD NDF NIR; NIRS NIST NMR NSP NVLAP OD OECD OEM OMD OSHA PAH PPE PTFE PVC RI rpm RPRs SGS SI SLAN SNV-D SOC TCA TEA TEB TOM t.p.i. UKAS UKASTA URL USDA UV v VAM VFA WPBS WSC

neutral cellulase plus gamanase digestibility neutral detergent fibre near infrared, or near infrared reflectance spectroscopy National Institute of Standards and Technology nuclear magnetic resonance non-starch polysaccharide National Voluntary Laboratory Accreditation Program outside diameter Organisation for Economic Co-operation and Development original equipment manufacturer organic matter digestibility Occupational Safety and Health Administration (US) polycyclic aromatic hydrocarbon(s) personal protective equipment polytetrafluoroethane polyvinyl chloride refractive index revolutions per minute reactive phosphate rocks Société Générale de Surveillance Statutory Instrument sufficiency level of available nutrient standard normal variate and detrend soil organic carbon The Composting Association triethanolamine total exchangeable bases total organic matter threads per inch United Kingdom Accreditation Service United Kingdom Agricultural Supply Trade Association Uniform Resource Locator (Internet website address) United States Department of Agriculture ultraviolet volume valid analytical measurement volatile fatty acid Welsh Plant Breeding Station (now Institute for Grassland and Environmental Research) water soluble carbohydrate

Experimental Planning

Experimental Design Experimental design is not directly related to chemical analysis, but it is important in that it determines the number of samples for processing. This could mean that there are too many tests for the laboratory to fit into its schedule, bearing in mind that there are many other customers clamouring for laboratory services. It could also mean that the cost is prohibitive for the funds available for the project. Some of the books on the design of scientific experiments appear far too theoretical for use in college field trials. However, three books in particular have proved useful in this Institute: • Statistical Procedures for Agricultural Research, 2nd edn. Gomez, K.A. and Gomez, A.A. John Wiley & Sons, 1984. • Agricultural Experimentation. Little, T.M. and Hills, F.J. John Wiley & Sons, 1978. • Statistical Methods in Agriculture and Biology, 2nd edn. Mead, R., Curnow, R.N. and Hasted, A.M. Chapman and Hall, 1993. For example, the book by Gomez and Gomez describes many possible designs such as the Latin square and the lattice designs. The former can handle simultaneously two known sources of variation among experimental units. Chapters deal with ‘Sampling in experimental plots’, and the ‘Presentation of research results’.

Chapter 1

Plot size The field plot size is chosen to give the required degree of precision for measurement of the selected characteristic. Sampling only a fraction of the plot, providing the sampling error is acceptable, may save time and expense. The sampling error is the difference between the value of the fraction and the value if the whole plot (population) had been sampled. If adequate precision is retained, it may be possible to bulk samples together at a later stage to reduce the numbers for chemical analysis.

Equipment Considerations Autoanalysis There is usually no problem of access to basic laboratory instruments and associated glassware, however, the only means of handling large numbers of tests is to apply some form of automation. An added advantage is that it improves the analytical precision and reproducibility. The most suitable technique has been based on the segmented continuous-flow principle invented by Skeggs (1957), and which was first marketed as the Technicon® AutoAnalyzer. The system consists of a number of modules powered from a stabilized 110 V supply, and a typical layout is shown in Fig. 1.1. This was improved with the next generation AutoAnalyzer II, which provided the peristaltic pump with a metering air-bar. This aided a more regular bubble pattern with further improvement in precision. The current AutoAnalyzer 3 system offers several useful features. The Compact Sampler has random access, which means that if there is an over-range sample, which may distort the succeeding two peaks, the software will automatically instruct the sampler to repeat the affected peaks. This system saves a lot of time because the operator does not have to work out the repeats after a long run and reload the cups to be repeated. The pump in the current model has the option of dilution valves that allow automatic rerun of off-scale samples at a higher dilution. The segmented stream can pass through the colorimeter flowcell without debubbling, the software switches off the detection signal when a bubble is present. The redesigned flowcell has a square-edge planar window and uses fibre optics to ensure parallel light transmission and hence a reduction of interference from variation in refractive index of the liquid stream. More information can be found on the manufacturer’s website: http://www.bran-luebbe.de/en/autoanalyzer.html The price range for a basic system with a colorimeter is about £20k to £27k depending on options (e.g. PC and flame photometer) and whether educational discount applies. Other manufacturers of segmented-flow analysers are Burkard Scientific, see: http://www.burkardscientific.co.uk/Analytical/Systems_Analysers_SF A2000.htm

Experimental Planning

(a) (a)

Sampler

Peristaltic pump

Chemistry Heating bath module

Colorimeter

Output to chartrecorder or personal computer

(b)

Fig. 1.1. (a) Modular layout of a typical segmented continuous flow system. (b) Simplified design for a 40-place tray to hold 8.5 ml industrial type autoanalyser cups (not to scale). It would be useful to number the cup positions. The 2.5 mm holes are for the staple which sets the stopping position of some models of sampler. ø, diameter.

Chapter 1

and Skalar (UK Ltd), who publish a comprehensive soil and plant analysis manual, see: http://www.skalar.com/uk/products2-1.html A micro-bore analyser is manufactured by Astoria-Pacific Inc. USA, see: http://www.astoria-pacific.com/analyzer.html and is marketed in the UK by Advanced Medical Supplies Ltd, see: http://www.ams-med.com/ It is possible to build a basic system with chart-recorder output using components from various manufacturers. Suppliers of used equipment are another possible source. Sometimes there are equipment auctions but, having learnt from bitter experience, unless one can actually go and see (and preferably test) the items listed in the catalogue, this method of purchase should be avoided. Very often parts will be missing and, being obsolete, no longer obtainable. Used equipment suppliers always include some form of guarantee, and that is worth its cost. Some used or refurbished equipment suppliers are listed in Appendix 1.

Samplers In addition to the above manufacturers, autosamplers (Series 4000) may be obtained from: Hook & Tucker zenyx, Harwood House, Clarendon Court, Carrs Road, Cheadle, Cheshire SK8 2LA, UK Tel: +44 (0) 161 428 0009 Fax: +44 (0) 161 428 0019 (Price range is about £2.8k to £3.0k).

Peristaltic pumps In addition to the above manufacturers, suitable peristaltic pumps with a minimum of 12 tube channels, such as the Ismatec IPC-16 and IPC-24 versions (cost £2000+), are obtainable from: Bennett Scientific: http://www.bennett-scientific.com/ismatec/peri.htm and from Cole Parmer Instrument Co. Ltd at: http://www.coleparmer.co.uk/ Other multi-channel pumps are manufactured by Watson-Marlow Bredel (the 200 Series) http://www.watson-marlow.com/wmb-gb/index.htm and distributed by Fisher: http://www.catalogue.fisher.co.uk/ where one can browse the catalogue without needing to complete the registration form; or from Patterson Scientific: http://www.patterson-scientific.co.uk/index.htm Another brand is the Cole Parmer Masterflex, which can accept a maximum of 12 cartridges to give 12 channels. The peristaltic pump is fitted with colour-coded PVC tubes of varying diameters. The flow rate is governed by the diameter and indicated by the colours of the collar at each end. Standard quality is usually of acceptable

Experimental Planning

tolerance, but flow-rated tubing with a higher precision is available. In addition to PVC, other tube materials are available. Silicone rubber is free of additives and plasticizers and less likely to age over time. Solvent resistant yellow PVC retains its flexibility when used to pump solvents, and vulcanized black rubber tubing is used with concentrated acids.

Chemistry module The outflow end of the pump tubing is connected directly to the components of the chemistry module. This module consists mainly of connectors and glass mixing coils. Proprietary modules are available, of course, but it is perfectly feasible to assemble the necessary components on a plastic tray fitted with four legs. Pump tubing and connectors are available from many suppliers. Apart from the OEMs, sources include: Gradko International Ltd: http://members.aol.com/gradkoin/ homepage.htm, who can also supply refurbished modules. Industrial 8.5-ml autoanalyser cups (Part No. 127-0080-01) are available from: Gradko (see above); or LIP (Equipment & Services) Ltd, 111 Dockfield Road, Shipley, West Yorkshire BD17 7SJ, UK. Tel: +44 (0) 1274 593411 Fax: +44 (0) 1274 589439. The advantage of the 8.5 ml as opposed to the 2-ml or 4-ml conical cups, is that it is easier to pour into them, several analyses are possible before they need refilling, and they are more easily washed if reuse is considered. The snag is that the 40-place 8.5-ml industrial cup sample trays are no longer made by Bran+Luebbe, but Gradko can supply them for about £112 each. In case these sample trays become unavailable, a dimensional diagram is given in Fig. 1.1b of a simplified version. The original trays were made from a glass fibre filled resin. It is suggested that suitable materials would be Nylon 66 rod, 25 mm diameter for the handle; cast Nylon 6 rod, 100 mm diameter for the underside; cast Nylon 6 available as 10 × 500 × 500 mm sheet, sufficient for four trays. These are available from RS Components Ltd, at the website: http://rswww.com Pump tubing is supplied by the above sources, also Elkay Laboratory Products UK Ltd: http://www.elkay-uk.co.uk/

Heating bath and dialyser These modules are options that can be incorporated within the chemistry module with newer systems, but stand-alone units are also possible.

Chapter 1

Colorimeter and spectrophotometer Many types are available, single or dual channel, expanded absorption ranges, digital or analogue, linear or logarithmic output, etc. If purchasing a complete new system, then the colorimeter is to be preferred. It is designed for the job with excellent long-term stability and freedom from drift; also, the sensitivity will suit the recommended chemistry, and the signal output will be compatible with the software and hardware. If building a system from variously sourced modules, the spectrophotometer will be a more useful choice. This is because it will accept standard flowcells, can be adjusted to any wavelength within its range, and usually has an output suitable for a chart-recorder. A spectrophotometer is also likely to have a scale-expansion facility allowing the measurement of absorbance values in excess of 1.0 Å, perhaps to 2.0 Å, and enabling lower values of possibly 0.1 Å to have the sensitivity increased to give a full-scale reading. This saves a lot of extra work in diluting or concentrating sample solutions. A colorimeter requires a separate filter for each wavelength. Interference filters are often required in pairs, and can be expensive. The one essential component is the flowcell (flow-through cell), which must either be the manufacturer’s own special fitting, or else a more universal design (e.g. 12.5 mm external square cross-section) as is common with most spectrophotometers. They are available with an optional built-in debubbler. These flowcells for continuous flow analysis must not be confused with flowcells with tube connections at the top and bottom of the cell, which merely allow filling and emptying by means of an external syringe mechanism. There will be two (or three with a debubbler) connections at the top of the cell. If the wavelength is to be in the UV region, a quartz or silica cell is required, otherwise an optical glass cell is adequate. The internal cell dimensions should be cylindrical, and a path length of 10 mm × 3 mm diameter giving a volume of 0.07 ml is usually suitable. This is a micro flow-through cell. A larger cell would cause too much internal mixing and interference between wash and samples, but a smaller (ultra-micro) cell volume would emphasize noise from differences in refractive index unless specified for low flow rate methods and the particular measuring instrument. Some manufacturers are: Hellma Cells: http://hellma-worldwide.com/tochter/Tochter2.htm to get the website for your area. For UK use: http://www.hellma.demon.co.uk/ Optiglass Ltd (Starna® Brand): http://www.optiglass.co.uk/

Chart-recorders A complete new system would have the benefit of system-control and data processing via a personal computer and proprietary software. An in-house system, however, would probably output readings to a chart-recorder. This can give a further opportunity of adjusting the scale-expansion to accommodate

Experimental Planning

extra-low or extra-high peaks. It is useful to be able to vary the chart speed; this will allow the peak width to be kept at an optimum width despite any variation of the sampling rate with different methodologies. Continuous-flow methodologies mean that the recorder is left running unattended for long periods. It is vital that the sprocket pins at each end of the chart paper drive are long enough to engage positively in the holes in the paper. It is annoying to find that they slip out, perhaps at one end, and an hour’s readings are wasted. It could be that the holes are fractionally out of sync with the pins, or that the paper has buckled at the end. We found with our in-house systems that a friction drive avoided these problems. The Houston Instrument OmniScribe® is of the friction type. Alternatively, a couple of large bulldogclips attached to the end of the chart that overhangs the bench may solve the problem.

Chart reader The reading of hundreds of peaks from a chart trace can be daunting. It is facilitated by means of a simple device known as a chart reader, apparently no longer available. It is a clear plastic A3 size sheet, originally having 15 sections, each consisting of ten vertical lines, which is laid over the chart. A baseline is first drawn on the chart under the peaks by linking the trace from aspirating the wash solution between tray changes. This compensates for baseline drift. The bottom of the vertical lines on the reader are next aligned with the baseline as it passes under the peaks from the standard solutions, which are included at the start, then after each tray to compensate for any change in sensitivity. The heights of the peaks are marked on the reader with a black grease-pencil (e.g. Royal Sovereign 808 Chinagraph) and labelled with the corresponding concentration. A connecting line is drawn to link the marks to give a standard curve. This is checked for each set of standards and corrected if necessary. The reader is laid on the chart, the bottom of the vertical lines aligned with the baseline, and the curve aligned with the top of the sample peak. The corresponding concentration is read off. A way to make a chart reader is given below. 1. Use a computer graphics program to draw eight sets of ten lines. This is printed in duplicate in portrait mode onto two sheets of laser transparency film. Corel Draw™ has a Graph Paper tool on the Polygon tool flyout. Select 40 columns and one row (the maximum number of columns is 50). Drag a rectangular graph to fill the left half of the page, make a copy and paste to the right as closely in line as possible. Go to Arrange and then Align and Distribute. Select left-hand image and align top to grid; repeat with right-hand image. Align right side of left image to grid, also left side of right image, and they should now be perfectly joined together. Select both and Group together. Adjust line width to 0.20 mm. Now draw a vertical line and adjust height to that of the graph, and line width to 0.60 mm. Copy, paste and drag to lie exactly over every tenth line. Save to file.

Chapter 1

2. Print duplicate copies using a laser printer on to laser transparency film. 3. Guillotine a vertical edge of each copy 5 mm from the thicker border line so that the two copies will form a single graph when placed together with edges overlapping and the thicker lines aligned. Tack together with a minimum of adhesive at the top, bottom and centre. 4. Laminate using 250 µm gloss film.

Flow injection The other type of automated wet-chemistry analysis is flow injection analysis (FIA), which was first described by Ruzicka and Hansen (1975). This is a nonsegmented continuous flow method – i.e. no air bubbles are introduced to aid mixing and to separate sample and wash segments. The small diameter of the tubing and the optimized flow rate, together with precise electronic control, enable sufficient separation of samples and wash. By allowing colorimetric reactions to go only partially to completion, high throughput rates are possible, up to 300 h–1. Although reagent consumption is low compared with the older segmented flow methods, the newer systems are even more economical than FIA. Two areas particularly suited to FIA are stopped-flow analysis as used in some immunoassays, and enzymatic analyses. FIA systems are manufactured by: • Burkard Scientific: http://www.burkardscientific.co.uk/Analytical/Systems_ Analysers_FIAflo2000.htm • ChemLab Instruments Ltd: http://home-1.worldonline.nl/~chemlab/ (The ChemLab instrument is in use at the Department of Soil Science, the University of Reading: http://www.rdg.ac.uk/soil/SoilSci/FACILITIES/analytical. html) • Foss Tecator: the FIAstar® 5012 System: http://www.foss.dk/foss.asp • Note: a useful list of scientific equipment suppliers is available at: http://chem.yonsei.ac.kr/~lsk/company.html We will only be dealing with segmented-flow methodology in this manual, but there are sure to be equivalent FIA methods available.

Batch Size The total number of samples will be determined by the experimental design (above), but the batch size should be chosen to suit the equipment used to process the samples ready for analysis. There will be a maximum load for the boiling units, heating blocks and shakers etc., so forward planning will optimize throughput. If an herbage batch size for a researcher is 80, it may be advantageous if he added half of the following batch in order to bring the batch for analysis to 120. This is because of the larger capacity of the heat-

Experimental Planning

ing block and the four available autoanalyser sample trays. Another approach would be to accumulate, say 240, samples and digest in two batches. The first batch of digested sample solutions could then be analysed while the second batch is digesting, or they could all be stored until a suitable time for autoanalysis. The main point to make is that the analytical laboratory needs to inform users well in advance of the best protocol for submitting samples. This includes other factors such as amount of sample required, recommended drying procedure, labelling, when to bring them in, and what authorized cost code is to be used in charging for the work. The question of prioritizing samples for certain users and situations in which queue jumping is allowed should also be addressed.

Sampling Protocol In this section we will consider some precautions necessary for the sampling of various materials before analysis. In general, samples should be representative of the bulk samples from which they were taken. It is shown that the variation associated with field sampling is 5 to 10 times greater than that associated with laboratory procedure. It would therefore be better to increase the number of core samples taken from the field than try to improve the accuracy of the analytical methods if the precision of the results from our field experiments is to be improved. (Allen and Whitfield, 1964).

Enough core samples should be taken throughout the field or mass of material to give a representative bulk sample. This may weigh several kilograms, so should be thoroughly mixed and sub-sampled, perhaps on site, to obtain a truly hom*ogeneous sample of a size suitable for processing. Galvanized sampling tools should not be used for trace element analysis. Usually from 20 to 25 cores are taken in a ‘W’ pattern across the whole area. An alternative approach is to traverse the whole area in a zig-zag manner, sampling at random along different sections of the area (Scott et al., 1971). The cores should be broken up and mixed well in a bucket, then about 200 g retained in a labelled polythene bag.

Soils With soil sampling from agricultural fields, it is usual to avoid any small patches of different soil (e.g. boggy or very stony); dung/urine patches, gateways and headlands should also be excluded. Large areas within the field that have had a different manuring/fertilizing history should be sampled separately. An auger, bulb-planter or trowel should be used to remove a core from an appropriate depth of 7.5 cm for grassland and 15 cm for arable. Stones and plant debris should be discarded. Sampling should be avoided after heavy rain or in time of drought. Sampling should also be avoided for

Chapter 1

P, K or Mg analysis for 8 weeks after applying fertilizer, 12 weeks after manure or slurry, or 12 months for pH determination after liming. Further details are available from the Potash Development Association (PDA, 1999c). If the soil is to be analysed for nitrate, it should be kept moist in a grip-top polythene bag and placed in ice as soon as possible before transport to the laboratory. Unless analysed immediately, which is unlikely, it should be frozen until a convenient time for analysis. This is to arrest microbial metabolism causing denitrification (conversion of nitrate by reduction to ammonium nitrogen and nitrous oxide gases). Biological activity and other problems have been discussed by Cresser (1990).

Composts Composts can be made from most biodegradable materials, and could derive from many unusual sources. If it originates from municipal solid waste, however, care should be taken that no toxic and non-degradable materials remain after the supplier’s separation processes. Small pieces of brick and concrete, glass and plastic (inerts), lead residues from old car batteries and cadmium from electroplated items are possible. A useful work on specifications and recommended chemical analyses of composts is the book by Bertoldi et al., 1987. The analyses specifying the compost include: ammoniacal nitrogen calcium carbon C:N ratio conductivity heavy metals inerts magnesium moisture

nitrate-nitrogen nitrogen organic matter (ignition 450–600°C) particle size pH phosphorus potassium total solids

Feeds Bagged feeds Instructions can be found in the AOAC Official Methods of Analysis (Padmore, 1990, p. 69). A pointed corer consisting of a single or double tube, or slotted tube and rod, is used to remove a diagonal core from end to end of the horizontal bag. Bulk feeds should have ten or more cores from different regions. The sample should be stored in such a way that deterioration and change in composition are prevented (BS 5766, 1979).

Experimental Planning

Silage, hay and haylage A suitable corer is needed to remove sample cores from within clamps and stacks. A motor driven corer is used in some research institutes, but is rare in other establishments. One of the first designed for research work was that of Alexander (1960). His design was just 183 cm in length, and not long enough for the depth of the average farm clamp today. We designed a threesection clamp in stainless steel to resist corrosion by the volatile fatty acids in silage (Faithfull, 1997). This is clipped inside a wooden box and will fit into the boot of a car. Table 1.1 compares the two corers. Construction details are shown in Fig. 1.2. Table 1.1. Comparison of Alexander pattern and modified design of silage corer. Property

Alexander pattern

Modified pattern

Reason

Assembled length (cm) Number of sections Material

182.9 2 Mild steel

281.5 3 Stainless steel

Greater depth required Ease of handling Avoid corrosion products contaminating sample

To sample the clamp, make two cuts in the membrane about 3 cm long in the form of a cross. Insert the tommy-bar into the corer, thrust the corer down vertically, and finish with a twisting action. Pull up and thrust down and twist again, repeating until the corer is full. Great care should be taken not to hit the concrete base of the clamp, as this will buckle the cutting edge. A penetration of about 38 cm was needed to fill the 15.7 cm long corer tube because of the greater compaction in the tube. The sample is removed from the tube using the tommy-bar and immediately placed in a labelled grip-top polythene bag. The middle and top bar sections are added to reach greater depths. They are secured with cross-pins held in place with insulation tape. Sampling positions Alexander (1960) commented on the distortion of the horizontal layers in the physical structure of the silage clamp (Fig. 1.3), and concluded that the most likely points to be representative of the whole pit would be the mid-points of the half-diagonals (Fig. 1.4). A vertical core through the centre of the clamp would include more of the top layer, which would have wilted longer, than the lower layers. Conversely, a core taken near the edge of the clamp would include relatively more of the lowest, moister layer. A core through the half diagonals would be more representative of each layer, although the optimum position might need to be determined by a more careful examination of the geometry of the clamp. Grass and herbage species It is vital that sufficient weight of sample is taken for the planned analyses, extra being added in case further unforeseen tests are required. Plant materials

Chapter 1

Fig. 1.2. (a) Stainless still silage corer. Units in mm (and inches (in) when appropriate).

Experimental Planning

(b)

(c)

(d)

(e) Fig. 1.2. (b) Stainless steel silage corer in wooden carrying case. From the top: file to sharpen cutting edge; tommy bar; bottom section with corer; middle section; top section. (c) Stainless steel silage corer; close-up of bottom end with corer. (d) Stainless steel silage corer; bottom and middle sections. (e) Fully assembled silage corer with metre rule.

Chapter 1

Central core biased to top layer

End core biased to bottom layer

Half -diagonal core is more representative

Fig. 1.3. Effect of layer structure on sample core bias.

•

• Front of clamp

Fig. 1.4. Sampling positions in order of sampling.

are high in moisture content, and young growth could lose 85% of its fresh weight after drying (Wilman and Wright, 1978). Contamination by soil should be carefully avoided. In animal nutrition studies, however, ingestion of some soil adhering to forage leaves and stems should be considered as normal for herbivores, and thus be taken into account when assessing the mineral and trace-element status of the forage. It is normal to allow 2 weeks between grazing and sampling to avoid contamination by trampling. Washing foliage should be kept to a minimum to reduce leaching, and large smooth leaves can be wiped with a damp cloth. Atmospheric deposition immediately before sampling should be considered, especially if within 10 miles downwind of a coastal region. An assessment of the degree of contamination can be obtained from the level of titanium in dry matter. If this exceeds 10 µg g–1, it can be considered as contaminated (Berrow, 1988). Some plant species possess a high moisture content, little structural fibre, and are very delicate. Such a species is chickweed (Stellaria media (L.) Vill.)

Experimental Planning

with about 91.3% moisture (Derrick et al., 1993). When this is thawed after being stored in a freezer, most of this moisture exudes out and so various soluble components will be lost unless poured back over the foliage before drying. Even before thawing, it forms ice crystals within the polythene sample bag, so these should be added to the sample if freeze-drying.

Plant components The chemical composition varies between roots, stem and leaf. For a wholeplant analysis, it is essential that no root fibres are left in the ground, and that no other parts snap off and are left out of the sample. As much material as possible should be collected to minimize errors from variations in heterogeneity. If sampling at the pollen shedding stage, the heads should be contained in paper or polythene bags to collect pollen and anthers (Wilman and Altimimi, 1982). It is possible to separate many types of plant components. The variation of chemical nutrients within these components and the change in them with maturity is relevant to animal nutrition. It could also influence the cutting height of crops for conservation. Some ryegrass components are shown in Fig. 1.5. Spikelets Rachis

Head-bearing internode

Blade

Sheath 2

Sheath 1 Node 1 Internode 1

Node 2 Internode 2

Sheath 3 Node 3

Sheath 4

Internode 3

Internode 4

Sheath 5

Roots

Fig. 1.5. Components of a typical ryegrass plant (Lolium perenne L.), adapted from Wilman and Altimimi (1982).

Chapter 1

Microbiological analysis Samples taken for chemical analysis may also be used for microbiological analysis. This may be the case for silage samples, when harmful clostridia could spoil the beneficial fermentation of Lactobacillus. It is therefore essential that the treatment of the sample immediately after collection should both prevent the further growth of the microbial species present and protect from the ingress of any harmful microorganisms or fungal spores. Although biased towards food samples, Microbiology for the Analytical Chemist by R.K. Dart (1996) is a helpful publication.

Biological substances Such samples include milk, blood, urine and faeces. Most samples will only need to be placed in an ice-box after sampling, this will help to prevent degradation and oxidation of sensitive compounds like vitamin E (tocopherol). The treatment may depend on the analyte to be measured, so it is essential to study the published sampling protocol before arriving to take the sample. Blood may need to be collected in a heparin tube if plasma is to be later prepared by centrifugation. The blood should be mixed with the heparin by slowly inverting several times, but never vigorously shaken. A heparin tube is not required before centrifugation for serum preparation. Samples may be kept for several months in a freezer at –20°C, but for longer than 6 months at –80°C. If sem*n is to retain its activity, it should be kept in liquid nitrogen.

Fertilizers For the sampling of fertilizers, consult Johnson (1990b), also refer to Chapter 2 ‘Sub-sampling’ and Chapter 6.

Sample Preparation

Pre-treatment of Samples and Contamination Care must be taken to avoid contamination of samples before analysis. Common causes of contamination are: • lime or fertilizer blowing on to plots from adjacent plots/fields, • use of tap water instead of deionized or distilled water when washing plants or extracting soluble components, • failure to wash earth from roots thoroughly before analysis.

Trace Element Analysis Extreme care is necessary in trace element analysis. Before use, polythene containers for storing sample and standard solutions should be washed successively with: • • • •

0.05 M EDTA (14.63 g EDTA + 4.0 g NaOH l–1) H2O, deionized 1.5 M HNO3 H2O, triply deionized or distilled (Adriano et al., 1971).

Chapter 2

Solutions and leachates for analysis must be particle free, and should therefore be centrifuged in polypropylene tubes and not filtered unless this is specified in the methodology.

Sub-sampling A bulk sample should be thoroughly mixed until hom*ogeneous, then a sub-sample taken. There are two main ways to achieve this when dealing with solid samples. First, there is the manual cone and quartering method. A spoon-shaped spatula is used to take portions randomly from the bulk sample, which are then transferred to a clean surface to form a new conical pile. Each successive portion is poured on to the apex of the cone until the entire heap has been transferred. The cone is then flattened, divided into quarters, and opposite quarters removed. These are mixed to form a smaller conical pile, and again quartered. This is repeated until a sample of suitable weight is obtained (Jeffery et al., 1989, p. 154; MAFF/ADAS, 1986, p. 2). A variation on this method is to place the sample in the centre of a square sheet of paper and thoroughly mix by alternately lifting opposite corners of the paper so as to roll the sample particles towards the centre, rather than allowing them to slide. The pile is made approximately circular and quartered as above (Triebold, 1946). Second, easily flowing granules or powder may be riffled. This is recommended for fertilizers (Johnson, 1990b). Riffle boxes (sample dividers or splitters) are available to BS812 and BS1377 from: A.J. Cope & Son Ltd, 11/12 The Oval, Hackney Road, London E2 9DU, UK Tel: +44 (0) 20 7729 2405 Fax: +44 (0) 20 7729 2657 E-mail: [emailprotected] also from Merck [VWR International at http://www.merckeurolab.ltd.uk/], and larger ones from Fritsch. A rotary cone type sample divider would probably be too sophisticated for fertilizers, plants and soils, with the simpler manually operated types being adequate. Cresser (1990) recommends a chute splitter or spinning riffler for environmental samples. Pascall Engineering Co. Ltd market the Rotary Wholestream, (see: http://www.pascalleng.co.uk/sampling/representative_ sampling.htm). This divider is intended mainly for providing samples for chemical analysis. The samples are taken from a moving stream of powder by a set of rotating stainless steel containers, the powder being fed from an adjustable hopper onto a vibrating feeder. They also make the Centrifugal, which is used mainly for seed samples and is used by the Official Seed Testing Station of England and Wales and by seed merchants and seed associations throughout the world. The Rules for Seed Testing, issued by the International Seed Testing Association (http://www.seedtest.org), give details of the unit and its use. Gross samples can be divided in seconds, and the model is suitable for all but the chaffiest of seeds.

Sample Preparation

Drying Techniques It is important to find out the correct drying method for the nature of sample and the type of analysis to be carried out. If samples are presented for the analysis of water-soluble carbohydrates having been dried for 24 h at 100 °C, it will be a waste of time as the sugars will be partially degraded. This also applies to cell-wall analysis by the neutral detergent fibre procedure. The original fresh herbage sample will have wilted, unless frozen, so the outcome could be disastrous. Often a compromise will be necessary if both above ambient temperature and time degrade the material. Thus the choice of drying technique will be between a low temperature for a longer period or a higher temperature for a shorter period. The decision may have to be in line with the conditions published in current journals for similar experiments, and these references may be cited to justify the choice.

Air-drying Air-drying is the usual method for soils. Large numbers of samples could be placed in cardboard or expanded polystyrene trays on metal shelving units in a ventilated warm room. A space-heater could be used to raise the temperature to no more than 30°C.

Oven-drying Fresh plant material is generally dried in a forced-draught oven. Samples may be placed in aluminium trays with mesh in the base to allow circulation of air. Fairly dry herbage can be placed first into labelled manilla envelopes or brown paper bags, being careful not to let them touch the interior surfaces of the oven. For in vitro digestibility, the drying time at 80°C should not exceed 6 h. To avoid losses when determining fluoride and selenium, the temperature should not exceed 50°C, and for boron 60°C (in an unlined tray). Although a short drying time of 2 h at 102°C has been given for water-soluble carbohydrates (MAFF/ADAS, 1986b), we would only recommend freezedrying (see below). Suggested drying conditions are given in Table 2.1. It should be noted that sample drying conditions are sometimes different from those used for dry-matter determinations, which are often more severe and for which a sub-sample is taken.

Vacuum oven This is one of the recommended drying methods for moisture in animal feed (Padmore, 1990, p. 69). About 2 g animal feed is dried to constant weight at 95–100°C under a pressure of 100 mmHg for about 5 h. A high molasses content requires 50 mmHg at 70°C. Vacuum ovens are ideal for drying

Chapter 2

Table 2.1. Some drying times for various feeds. Reference MAFF/ADAS, MAFF/ADAS, MAFF/ADAS, MAFF/ADAS,

Sample 1986, 1986, 1986, 1986,

p. p. p. p.

4 4 4 4

Drying Temperature time (h) (°C)

Herbage, hay (silage dry matter) 18 Brassicas 18 Root crops (carrots, swedes, etc.) 48 Potatoes, artichokes 24 plus 18 MAFF/ADAS, 1986, p. 4 Cereal grains 40 Wallinga et al., 1995 Herbage 24 Isaac, 1990, p. 41 Herbage 24 Byrne, 1979 Herbage 16

100±2 100±2 60 60 100±2 100±2 70 80 95

products that are heat labile at low temperatures. For dietary fibre analysis, drying at 100–105°C may cause Maillard reaction products which analyse as lignin. It is recommended that a vacuum oven at 60°C or freeze-drying is used (Southgate, 1995, p. 46). Makers of vacuum ovens include Gallenkamp (from Fisher), Heraeus GmbH & Co. KG, Jouan, and Townson and Mercer (see http://www. sanyogallenkamp.com; http://www.heraeus.com; http://www.jouaninc.com; http:// www.townson-mercer.co.uk).

Freeze-drying This is one of the best methods for drying sensitive materials, but has relatively little mention in the literature. It is the only way water can be almost completely removed from tissue or organic material with minimal damage to the cell structure. The fresh herbage is first deep frozen as soon after harvesting as possible. It is transferred to the freeze-dryer chamber, and the methodology used is described below. A vacuum is applied, and a controlled supply of heat may be provided. This is to allow the ice to sublimate or evaporate, but never to melt. The extracted water vapour condenses on the surface of the refrigerated chamber at about –40°C. A small amount of water vapour escapes condensation and passes out to the vacuum pump and through the oil reservoir, thence through an oil mist filter to the atmosphere. If the oil is not hot, the water vapour will condense in the oil and sink to the bottom. The vacuum is measured using a Pirani gauge, with readout on a meter; this may have a separate electrical switch. Notes: A freeze-dryer must be purchased that either has a large built-in chamber, or to which a chamber can be attached. Some smaller ones are mainly for smallscale work when multiple samples are held in glass flasks or ampoules, which are attached to a manifold equipped with isolation valves.

Sample Preparation

It is vital to maintain clean contacts on the connection fitting of the Pirani gauge, because it is very sensitive to resistivity changes due to tarnishing or dust. Freeze-drying methodology The freeze-dryer is switched on and the pump started in order to warm up the oil within the pump housing. This is to prevent condensation of moisture in the oil. The gas ballast valve is also opened for the first half of the drying process, when most of the moisture is removed, to purge out any moisture from the oil. One manufacturer (ChemLab) recommends leaving the ballast valve open for the whole drying process. Although this will keep the accumulation of water in the oil to a minimum, it will mean the ultimate vacuum possible with the pump will not be reached, and samples will have slightly more residual moisture. When the condenser temperature indicator reads less than –30°C, the frozen sample may be loaded into the chamber. Samples may be placed in trays, paper bags or microporous bread bags, but preferably not in polythene bags, which could hinder the evaporation process. Samples of a lumpy consistency should be broken up while still frozen to speed the evaporation process. The chamber may have rubber seals which need greasing. The minimum amount of silicone grease should be applied, and the seals should be wiped scrupulously clean before the application, as any particles of sample material adhering to the seals will allow ingress of air owing to poor sealing at that point. With the chamber lid or door in place, the drain valve should be closed, and the vacuum valve opened. When the Pirani gauge reads 66.5 Pa (500 millitorr or 0.5 mmHg), the heater may be switched on. The drying time will depend on the nature and water content of the samples, but 2–4 days is normal. The condensation chamber will have a certain capacity, perhaps 3 l, so the total sample water content should not exceed this, and there is a loss in efficiency if about two-thirds of this value is exceeded. At the end of the process, when the pressure has been about 13.3 Pa (100 millitorr or 0.1 mmHg) for several hours, the isolation valve is closed, the drain tap opened to allow ingress of air, and the defrost switch turned on. The lid may then be removed and the samples checked for dryness. Any larger samples should be inspected, to ensure there are no remaining areas of ice at the centre. The freeze-dried samples should be stored in a desiccator before milling, which should be done as soon as possible. Freeze-dried samples are somewhat more sticky than oven-dried ones, and the crisper they are, the better they mill. After milling, the samples should be stored in airtight sample tubes or grip-top polythene bags to prevent rehydration and fungal attack. Samples will not be as dry as oven-dry material, and will typically contain between 3% and 10% moisture depending on the sample. When results from the analysis of freeze-dried material have to be expressed in terms of oven-dry matter, a sub-sample must be taken at the time of weighing for a separate oven-dry matter determination. This will enable a correction for residual moisture to be made. Microporous bags can be obtained from Cryovac at the website of Sealed Air Corporation:

Chapter 2

http://www.sealedair.com/products/food/bakery_fs.htm Websites of freeze-dryer manufacturers/suppliers are: Virtis: http://www.virtis.com/ ChemLab: http://home-1.worldonline.nl/~chemlab/ CHRIST: http://www.phscientific.co.uk/html/ Heto: http://www.heto-holten.com/camel.htm

Desiccation Although infrequently used, an alternative method for drying samples at room temperature is in a vacuum desiccator. These used to be made of glass with a possible risk of implosion. Modern ones are made from polycarbonate or polypropylene base with a polycarbonate cover, and are cheaper than the glass equivalent. The main limitation is that they are for room temperature use only, and not for use with organic solvents or vapours. An efficient water pump should be adequate, however, a guard-tube containing desiccant should be inserted between the pump and the desiccator. Manufacturers include Kartell and Nalgene. Nalgene: http://nalgenelab.nalgenunc.com/

Milling, Grinding and hom*ogenization Animal tissue is often blended in a high-speed blender until completely hom*ogeneous. For trace element work, solid samples should not be ground in a mill constructed with materials containing the elements to be determined, such as iron, chromium and manganese. In this case, a mortar grinder (mortar and pestle mill) or ball mill would be suitable. The former may be constructed of agate, with the pestle and mortar being independently motor driven (Pascall Model 00, Agate), or may consist of a vibrating ball and mortar (Fritsch Pulverisette). The ball mill may be porcelain with fused magnesium silicate balls (Pascall Model No. 9). There are centrifugal, planetary and roller type ball mills. The physical characteristics of the sample material may determine which type is best for the purpose, and the manufacturer’s advice should be sought. Some manufacturers are listed below: Glen Creston Ltd: http://www.glencreston.co.uk/ Christy: http://www.christy-norris.co.uk/ Fritsch GmbH: http://www.fritsch_lab.de/englisch/english.htm/ IKA®: http://www.ika.net/ Merck: http://www.merckeurolab.ltd.uk/ Pascall Engineering Co. Ltd: http://www.pascalleng.co.uk/Mixing.htm Retsch GmbH & Co. KG: http://www.retsch.de/english/zerkleinern_e.html The fineness of grind is important and can influence the result, especially when the sample is being subjected to partial dissolution in detergent or enzyme containing solutions. A mesh of the appropriate size can usually be

Sample Preparation

inserted in the mill. Herbage is usually graded to 1 mm particle size. For available carbohydrates in cereal mixes, the sample should be ground to 0.5 mm. In general, for small dry samples, a micro hammer-cutter mill will tackle anything from cotton to small rocks. For more than 100 g dry herbage, a larger cutter or knife mill will be more efficient. We use a Christy-Norris 20.3 cm (8 in) cutter mill with a 1.47 kW (2 hp) 415 V (three-phase) motor. A 12.7 cm cross beater hammer mill at approximately 13,000 rpm is also suitable. If the receiving container is a cotton bag, it should be turned inside out and shaken between samples. An 18 × 36 cm grill of 2.5 cm wire mesh set into the bench in front of the mill and connected to a suction fan (415 V, 2.5 A, 1400 rpm, 1.1 kW) via ducting through an outside wall, removes the dust at source. It is essential to clean the mill between samples to prevent cross-contamination, and a paint brush and vacuum nozzle are used. However, if milling samples weighing about 500 g, and the component to be measured only differs by a maximum of 0.5% between samples, then a residue of 5 g in the milling chamber will only affect results by 0.005%. If results are given to 0.1%, the tedious cleaning process might be considered unnecessary. The lignified and cutinized tissues of cereal grains need the more vigorous disintegration of ball-milling to produce a hom*ogeneous sample. Prolonged ball-milling, however, can depolymerize cellulose, therefore wet ball-milling in an organic solvent or suitable extractant is recommended (Southgate, 1995, p. 47).

Freezer mill For extra sensitive or rubbery samples, a freezer mill is available. This uses liquid nitrogen at –195.8°C which renders most ductile or elastic substances friable, and is suitable for those with a low melting point or which are unstable at room temperature. The sample is placed in a polycarbonate tubular sample container with a stainless steel impactor and closed with two end caps. It is immersed in liquid nitrogen and a magnetic field oscillates the impactor against the end caps to powder the sample. Samples of up to 3 ml can be ground in less than 4 min, while others wait in a separate compartment in the milling bath. A full day’s operation may require 20 l of liquid nitrogen.

hom*ogenization This may find application in several areas. The first example is the hom*ogenization of animal tissues in a high-speed blender, which enables a hom*ogeneous sample to be obtained for subsequent analysis. This is used, for example, in the analysis of arsenic or copper in liver (Ross, 1990). A second area is the extraction of volatile fatty acids from silage. Typically, 10 g fresh silage is hom*ogenized for between 1 and 10 min with 100 ml water in a blender before filtration (Lessard et al., 1961). The last area is the dry

Chapter 2

hom*ogenization of linseed seeds prior to oil determination. These resist crushing and so about 4 g of seeds are hom*ogenized at 11,500–13,000 rpm for 2 min. There are several types of blender. One of the most popular is the Waring blender, operating on a similar principle to domestic jug-type food blenders. Containers are available in glass, stainless steel and polycarbonate. It should be noted that the working capacities of the containers are a maximum of 70%, and a minimum of 10% nominal capacity. Accessory containers enable volumes as low as 12 ml to be handled. Silage is hom*ogenized in a Waring type blender. There are also the dispersing shaft type hom*ogenizers, which may be hand held or stand mounted. The shaft has a tip with teeth rotating within a fixed crenated stator, which imparts impact, shock, shearing and cavitation effects. Working volumes as low as 0.03 ml (PRO20 hom*ogenizer) can be handled. The materials, however, should be free-flowing, and usually suspended in a liquid. For solid materials, like seeds, a blade rather than a dispersing tool is required. PROTM market a Safety-Seal® Chamber Assembly with a 25.4-mm blade which can handle a minimum of 10 ml (supplied by Radleys). Status hom*ogenizers can be equipped with their AX60 Analytical Mill attachment (supplied by Philip Harris Scientific). This has a cooling jacket that can be used with liquid nitrogen for temperature sensitive samples. Manufacturers’ websites: Büchi Labortechnik AG: http://www.buchi.com/ Fisherbrand: http://www.fisher.co.uk/ IKA®: http://www.ika.net/ http://www.labworld-online.com/ika/index1.html Kinematica (Polytron®): Kinematica AG at http://www.kinematica.ch/ Tel.: +41 41 2501257 Fax: +41 41 2501460 Supplied by Philip Harris (RossLab plc) http://www.phscientific.co.uk/html/ PRO Scientific Inc: http://www.proscientific.com/ Radleys (R.B. Radley & Co. Ltd): http://www.radleys.co.uk/

Storage of milled samples Once milled, the samples should be stored in air-tight containers and kept in a cool place away from direct sunlight. Powders are suitably stored in 50 × 25 mm glass specimen tubes capped with polythene push-in closures. They may be handled in aluminium or polycarbonate freezer trays. Two sizes of trays are useful – a larger size holding 120 tubes (10 × 12) is suitable for oven redrying before weighing. A smaller tray holding 25 (5 × 5) will fit most desiccators for temporary storage after redrying for subsequent weighing. The sample tubes should be numbered consecutively from 1 up. Any plot codes, identifying letters, etc., should be kept by the person submitting the samples for later interpretation. This simplifies the sample labelling and record

Sample Preparation

keeping of the analytical laboratory. (Note: Unique batch/sample numbers are required for UKAS accreditation.) When sub-sampling from a kilogram or more of milled herbage samples or sieved soils, it would be wise to store the remainder of those giving low, medium and high values for future use as reference samples. These can be included with each batch of similar samples, and thus any excessive standard deviation from the mean (obtained by repeated analyses over a period of time) will indicate that an error has arisen in the analytical procedure. A protocol should therefore be established that if one (or more) of the low, medium and high control samples included with the sample batch gives a result lying outside of, say, ± 2s (where s = standard deviation), the whole analytical procedure should be repeated after checking from where the error could have arisen. This is discussed in greater depth in Chapter 12.

Weighing and Dispensing

Weighing Errors There are various sources of error that can occur when weighing samples for analysis.

Correction of weighings to ‘in vacuo’ If, as is usual, the sample has a lower density than the stainless steel balance calibration weight, the buoyancy effect of air on the sample mass means that a litre of water would indicate a weight of 1.05 g less than expected. Where weights of sample components are expressed as percentages or ratios, this error almost reduces to zero. This correction is usually ignored and considered well within acceptable experimental error (Jeffery et al., 1989, p. 76).

Incorrect calibration of the balance Between periods of servicing of the balance, it is wise to check the accuracy with a calibration weight. Some balances incorporate a self-calibration facility.

Weighing and Dispensing

Static charge If a glass vessel is cleaned by wiping or brushing, a static charge may build up. For a vessel of 150 cm2, the force on the balance pan could amount to 100 mg. Clearly, when weighing samples of less than a gram, this is an unacceptable error of over 10%. Static may also be present on the person weighing or on the sample particles; milled herbage can jump from the spatula blade on to the walls of the weighing vessel. The apparent weight often alters as the spatula is lowered into the weighing container. The final reading should therefore only be taken after withdrawal of the spatula and when the draught shielding door has been closed. Electronic anti-static devices are available, but as they usually incorporate a fan, it is necessary to position them carefully to avoid the effect of the draught.

Convection currents It is essential that both samples and crucibles have cooled to room temperature in a desiccator before weighing. This may take 30–40 min.

Absorption of moisture by the sample Although dried herbage samples are kept in a desiccator before weighing, it is possible for samples to absorb moisture from the atmosphere during weighing (Faithfull, 1970). This arises from the repeated removal and replacement of the desiccator lid. This was investigated using three types of sample: grass, barley and faeces. Using a large desiccator holding 80 samples, the lid would be removed that number of times over a 2-h period. The absorption of moisture by the last sample to be weighed amounted to 0.95% for grass, 0.83% for faeces and 0.77% for barley. This effect can be reduced to about 0.1% by using smaller desiccators holding about 12 sample tubes. The sample will continue to absorb moisture while on the balance pan, the initial rate being about 0.01% min–1. The use of a well-balanced spatula (e.g. a wooden handled 75 mm stainless steel-bladed palette knife) and glass weighing funnel will speed the weighing process and reduce moisture absorption.

Absorption of moisture by the sample container Glass and porcelain are particularly susceptible to adsorption of atmospheric moisture on exposed surfaces. Containers to be heated in gravimetric procedures (e.g. oven-dry matter or ash content) should therefore be pre-heated to the same temperature as that procedure and cooled in a desiccator before measuring the tare weight.

Chapter 3

Dispensing Errors Dispensing errors can arise from the use of inappropriate or inaccurate equipment. Measuring cylinders are suitable for making up reagents, but are not accurate enough for the dilution of sample solutions. For the latter, a pipette and graduated (volumetric) flask are used. The accuracy of these may be checked by weighing the dispensed or contained amount of water. Flasks are calibrated on manufacture at 20°C. If a 1-l borosilicate glass measuring flask is used at 15°C, the contraction of the glass wall increases the volume by 0.05 ml, thus a correction of –0.05 ml is required. The water has itself contracted by 0.84 ml, so an additional correction of +0.84 ml should be added, making the total correction +0.79 ml, or +0.079%. Although this is an acceptable error, when combined with other sources of error, the maximum possible error can be surprisingly high. Thus each source of error should be minimized as far as is practicable. Volumetric glassware is available in Class A and Class B qualities. A Class A 10-ml bulb (one-mark) pipette has a tolerance of ±0.020 ml, and a Class B ±0.040 ml. Class B is adequate for routine agricultural chemical analysis. Bulb, or transfer pipettes, are usually made to deliver a stated volume of liquid under standard conditions of temperature and with a draining time of 15 s while the tip is in contact with the wall of the receiving vessel. Previously, the tip should be touched against the wall of the container from which the liquid has been aspirated in order to allow any adhering droplet to drain away. A pipette filler should be used to avoid the danger of liquid entering the mouth when the unsafe mouth suction technique is used. Graduated pipettes with straight sides may deliver a volume from zero at the top to any graduation line, or from a graduation line to zero at the jet tip. Some are blow-out pipettes which require the last drop to be blown out from the tip, and these are indicated by a white or etched ring near the top of the pipette. Normal bulb pipettes should never be blown out to try and save time. Bulb pipettes are graduated to BS1583 and graduated pipettes to BS700 and ISO835, and most are colour coded. The latter are divided into types as given in Table 3.1.

Bottle top dispensers Bottle top dispensers are invaluable for the repetitive measurement of a certain volume of reagent into sample containers for extraction. For example, they find extensive use in soil analysis for dispensing the extracting reagents for phosphate, potassium and magnesium. If the volume setting is adjustable, it is essential to check the amount delivered by weighing the water. Careful priming should ensure that there are no trapped air bubbles. Some manufacturers are: Bibby Sterilin Ltd: http://www.bibby-sterilin.co.uk/ Brand GmbH & Co. KG: http://www.brand.de/ Eppendorf AG: http://www.eppendorf.com/

Weighing and Dispensing

Jencons (Scientific Ltd)/Zippette: http://www.jencons.co.uk/ John Poulten Ltd/Ultra Volac: http://www.pdd.co.uk/experience/medical/volac.htm and http://www.labpages.com/std_home/page0107.html Table 3.1. Various types of graduated pipette. Type 1 Type 2 Type 3 Type 4

Calibrated for delivery (EX) from zero at the top to any graduation line down to the shoulder Calibrated for delivery (EX) from any graduation line down to zero at the jet Calibrated for delivery (EX) from zero at the top to any graduation line down to the jet Calibrated to deliver from zero at the top down to the jet with the last drop expelled by blowing

Syringe pipettes Syringe or micropipettors can be of two types: positive displacement or air displacement. In the former type the liquid comes into direct contact with the piston, which may lead to carry-over from one sample to the next, albeit usually negligible. This would not matter if the same reagent solution were being dispensed. The latter type, however, is usually used. They have either fixed or adjustable ranges and are available from 1 µl to 10 ml. Micropipettors with disposable tips are useful for dispensing the extracted and filtered soil solutions. The delivered volume should be checked as above, and care taken to use the correct technique. With the pipette in an upright position, the push-button should be slowly depressed until the first resistance is felt (first stop position). With the tip well immersed in the liquid to be dispensed, the push-button is slowly released. When aspirating the solution, no air should be admitted by exposing the tip above the liquid surface. If this happens, liquid will contaminate the piston chamber, which should be cleaned before further use. The liquid is delivered by slowly depressing the push-button until the first stop. After a couple of seconds, press the plunger to blow out the droplet to empty the tip. When checking the volume of water delivered by weighing, Table 3.2 will enable a graph to be plotted and the volume at the exact temperature of measurement determined. Table 3.2. Volume of 1 g water at temperatures between 10°C and 30°C. °C

Vol. (ml)

°C

Vol. (ml)

10.00 12.00 14.00 16.00 18.00 20.00

1.0013 1.0015 1.0017 1.0021 1.0023 1.0027

22.00 24.00 26.00 28.00 30.00

1.0033 1.0037 1.0044 1.0047 1.0053

Acid-digestion, Ashing and Extraction Procedures

The actual analytical methods will be detailed in the appropriate chapters, but here we will just comment on the techniques involved.

Acid-digestion and Washing Acid-digestion of soils There are three main reasons for digesting soils in hot acid – to determine the organic carbon content, to extract mineral elements for their total content, and to determine total nitrogen by the Kjeldahl digestion. The first is called Tinsley’s wet combustion (Tinsley, 1950), and uses a highly corrosive mixture of sodium dichromate, and concentrated perchloric and sulphuric acids. For undergraduate practical classes, the safer loss on ignition method might be considered more appropriate. The second reason for acid-digestion is the determination of the total soil elemental content of, e.g. potassium, phosphorus or trace elements. This is seldom done for potassium in normal soil samples, mainly because ‘the total K in soils is of no value as an index to the availability of K to plants, nor is it always of value in tracing the movement or accumulation of applied fertilizer K’ (Pratt, 1965). The unreactive soil phosphorus is obtained by subtracting the naturally leached reactive phosphorus from the total phosphorus, and a method for determining the latter by extraction with sulphuric acid and potassium persulphate is cited by Turner and Haygarth (2000). They analysed 30

Acid-digestion, Ashing and Extraction Procedures

the reactive phosphate by flow injection analysis using a Tecator 5020 with autosampler, and using Method Application ASN 60–03/83 (Tecator Ltd, Sweden). The safety aspect is an important reason for avoiding, if possible, total elemental determination in soils, because the reagents often involve hydrofluoric acid (48% m/m) and perchloric acid (60% m/m). The former causes horrific burns, possibly fatal if not treated immediately, but is necessary to dissolve the potassium-bearing silica, and the latter, necessary for completely dissolving organic matter, may cause explosions if evaporated to dryness with carbonaceous materials or metals. Alkali fusion is another method for total elements in soil. Acid-digestion is often used with composts derived from municipal wastes, sewage and slurry, where toxic amounts of heavy metals may cause problems on the land to which they are applied. It is probably more convenient to determine total elements in soils by a benchtop X-ray fluorescence spectroscopy (XRF) instrument. This only requires the soil to be ground, and several reference standards of a similar soil. A Reference Materials Catalogue, Issue 5, 1999, is available from LGC’s Office of Reference Materials, Queens Road, Teddington, Middlesex TW11 0LY, UK. Tel. +44 (0)20 8943 7565; Fax +44 (0)20 8943 7554. Alkali fusion, hydrofluoric acid (HF) digestion and XRF give true total values as required for geochemical purposes, but digestion in aqua regia (see Method 5.15) gives total environmentally available concentrations, which are most meaningful for agricultural and environmental purposes. Transition metals may be more effectively extracted by using a pressured microwave digestion system such as the Anton Paar Multiwave Microwave Sample Preparation System. An example of sewage sludge analysis by this system is given at: http://www.lab123.com/app_data/mswave.htm

Total soil nitrogen Soils mainly contain nitrogen in its reduced state such as ammonium compounds and organic amino complexes. The standard Kjeldahl technique is therefore suitable to estimate the organic (plus ammonium) nitrogen, which it does by oxidizing the organic matter in hot sulphuric acid containing a catalyst and converting the nitrogen to ammonium sulphate, which can be measured by distillation and titration, or by a colorimetric procedure. The distillation is carried out after first adding excess sodium hydroxide to the acid digest to liberate the ammonia gas from the ammonium sulphate. (NH4)2SO4 + 2NaOH = Na2SO4 + 2NH3↑ + 2H2O

This is distilled into a receiving flask containing boric acid indicator mixture and titrated against 0.001 M HCl. The colorimetric method using the autoanalyser is based on that used for plant materials (see below), but care should be taken that any precipitate formed does not collect in the flowcell, which must be occasionally inverted or cleared by passing a bubble of air through it. Any nitrate (and nitrite, which is usually insignificant) should be reduced

Chapter 4

to ammonium by adding salicylic acid followed by zinc dust (see Method 5.6a.i) before digestion for the autoanalysis method with a colorimetric procedure, or Devarda’s alloy (see Method 5.5b.ii) before the distillation, if it is to be included to give a total nitrogen value.

Acid-digestion of plant materials The original method for the determination of nitrogen by sulphuric aciddigestion was published by Kjeldahl in 1883 and fully described by Burns (1984). Many modifications have since been made with various catalysts and acid mixtures. The digestions can be carried out in up to 40-place multiple heating units using specialized glassware which is commercially available; some suppliers are listed below: Digestion systems: Büchi Labortechnik AG: http://www.buchi.com/ Gerhardt UK Ltd: http://www.gerhardt.de/gb/kb.htm Foss (Digestor 2000 System): http://www.foss.dk/foss.asp Distillation systems: Büchi Labortechnik AG: http://www.buchi.com/ Foss (Kjeltec® 2300 Analyzer Unit): http://www.foss.dk/foss.asp Gerhardt UK Ltd http://www.gerhardt.de/gb/vap.htm We have devised a method enabling the digestion of up to 152 samples at a time, and with the wearing of some essential personal protective equipment (PPE), it has proved successful for over 32 years (Faithfull, 1969). Acid-digestion unit The major expense is the hotplate, which has to have a sufficiently large working surface area and be able to sustain a temperature of 310°C. A suitable hotplate is the Gerhardt HC 63, nominal voltage 400 VAC, 4800 W, working area 650 × 300 mm, and a maximum temperature of 400°C ± 5°C. In the UK this is available from: C. Gerhardt UK Ltd, Unit 5, Avonbury Court, County Road, Brackley, Northants. NN13 7AX. Tel. +44 (0) 1280 706772; Fax. +44 (0) 1280 706088 Other suitable hotplates are available from S & J Juniper & Co.: http://www.sjjuniper.com/general_purpose.shtml On the centre of the work surface are positioned two aluminium blocks, 440 × 100 × 100 mm (w × d × h), with the bottom surface machined flat to ensure good thermal contact with the hotplate. These are each drilled with 17 mm diameter holes to a depth of 86 mm and arranged in four rows of 19 holes. Thus each block accommodates 76 digestion tubes. These tubes are 150 mm long and 16 mm diameter, heavy wall (BS 3218) borosilicate glass rimless type; they are supplied by Fisher as TES-674-150S. The exposed areas of the work surface may be covered with a heat-resistant insulating material.

Acid-digestion, Ashing and Extraction Procedures

The whole unit is accommodated in a fume cupboard fitted with a scrubber unit to remove the acidic fumes before emission to the atmosphere. The constructional materials of the fume cupboard should be able to withstand the heat radiated from the hotplate and heating blocks. A digestion tube containing a 350°C thermometer with the bulb embedded in a 2-cm layer of sand occupies one hole in each block. The hotplate can be connected to the power source via a time-switch, which can be set to come on approximately 1 hour before commencement of work; this saves valuable time lost waiting for it to warm up. Acid-digestion procedure The acid used for the Kjeldahl digest is analytical quality concentrated sulphuric acid which contains 4 g l–1 selenium. This is prepared by heating a 250-ml portion of acid with 4 g selenium powder (Aldrich 20,965-1, 100 mesh) in a 1-l beaker on a hotplate in a fume cupboard, carefully stirring with a glass rod until dissolved to form a green solution (protective gloves, and safety spectacles or visor must be worn at all times when handling concentrated acids). After cooling, the solution is poured via a funnel into a glass storage bottle, and the balance of 750 ml acid added. Note: a dust mask should be worn when weighing selenium as it is easily absorbed by the lungs and is a possible teratogen. The beakers should be removed from the hotplate and left to cool in the fume cupboard after placing a watch glass over the top of the beaker. After cooling, the solution is poured via a funnel into a bottle or reservoir fitted with a bottle-top dispenser adjusted to 5 ml. All components with which the solution comes into contact must be resistant to concentrated sulphuric acid. Warning: the solution is highly corrosive and even when cold rapidly dissolves cellulosic materials. Wipe up any drips immediately with a wad of tissue and soak with plenty of running water before disposal. This is to both protect personnel involved in waste disposal and to prevent spontaneous combustion. Acid on the skin should be flooded with water for 1 min and medical advice sought for any blisters or burns; contaminated clothing should be removed and washed before reuse. Exactly 0.1000 g milled plant sample is weighed into a glass weighing funnel and transferred to the numbered digestion tubes with the aid of a small paintbrush. The digestion tubes are held in stainless steel racks and either stoppered or covered with sheets of paper until ready for digestion. The tubes should have been previously marked with two scratch lines around the outside at the levels of 5 ml and 10 ml. The acid is dispensed carefully into each tube; if it is admitted too rapidly, fine sample powder as well as acid may be ejected from the tube. A few tubes at a time are loaded into the blocks. Some types of sample are prone to frothing, and if this occurs, it is easier to remove a few tubes and allow them to cool in their racks, rather than risk some frothing right over before they can be removed. The most tedious aspect of the procedure is, after about an hour, to run a thin (4 mm diameter) glass rod vertically around the inside of the digestion tube in a downward spiralling motion in order to reintroduce any sample

Chapter 4

particles back into the acid. PPE must be worn for this. The temperature must not exceed 320°C because sulphuric acid boils at 330°C, which could cause injury; the two thermometers should be checked before the cleaning operation. (Note: the normal Kjeldahl procedure uses a salt such as sodium sulphate to raise the boiling point of the acid.) The samples are allowed to digest for a total of 4.25 h, when they are removed with stainless steel tongs and allowed to cool in their racks. The acid level is then adjusted dropwise with concentrated sulphuric acid to the 5-ml mark to replace any lost as fumes. Deionized water is then slowly added from a wash bottle, directing the jet down the side of the tube, up to the 10-ml mark so as to form two layers. Note: normally the safe way is to add concentrated sulphuric acid to water, especially when contained in a beaker – this is to prevent violent boiling. This does not happen here because of the restricted surface area and the formation of separate layers. The two layers are mixed by slowly oscillating a thin glass rod with one end flattened to form an 8–10 mm disc. Mixing should start from the junction of the layers, slowly working towards the top and bottom. The solution will contract after cooling, so the level must be again adjusted to the 10-ml mark and mixed with the rod. This final adjustment to 10 ml is best done immediately before analysis otherwise the tubes will need to be stoppered to avoid absorption of atmospheric moisture. The advantages of this digestion technique are the large number of samples that can be processed at one time, the simple and cheap glassware involved, and the fact that the digest may be used for the subsequent determination of not only nitrogen, but calcium, magnesium, potassium, sodium, phosphorus and iron. Summary of the indophenol blue colorimetric determination of nitrogen To determine the nitrogen content of herbage and soils by autoanalysis, one must first carry out a Kjeldahl digest in concentrated sulphuric acid with selenium (0.4% w/v) catalyst; this converts protein nitrogen to ammonium nitrogen, as shown in Fig. 4.1. The density of the blue colour is proportional to the nitrogen content. It is measured using a spectrophotometer at a wavelength of 640 nm and the height of the peaks on a chart-recorder compared with those of known standards to obtain the nitrogen content of the original material. Protein content = %N × 6.25.

Microwave acid-digestion The digestion of a wide range of matrices, from fish to rocks, is possible in a stainless steel pressure vessel fitted with a PTFE container. It is particularly useful for demanding trace element analyses. It was first developed by Professor Tölg, the method being described by Kotz et al. (1972). Pressure vessels are expensive, but digestion times can be as little as 60 s to dissolve fish tissue in nitric acid. An article comparing closed vessel microwave digestion versus conventional digestion procedures for the determination of

Acid-digestion, Ashing and Extraction Procedures

H2SO4/Se; 310 ˚C; 4.25 hours RCH-COOH I NH2 amino acid in protein

(NH4)2SO4 Using autoanalysis, ammonium ions are reacted with sodium phenate and sodium hypochlorite to give the indophenol blue colour by the Berthelot reaction (Berthelot, 1859).

Indophenol Blue Fig. 4.1. Colorimetric reaction converting protein nitrogen to the indophenol blue colour.

mercury in fish tissue by cold vapour AAS using a basic laboratory microwave is given by D.C. Stockton and B. Schuppener at: http:/www.epa.gov/earth1r6/6lab/mercury.htm Typical vessels are described at the following website: http://www.berghofusa.com/berghof.htm Digestion systems are supplied by CEM Corporation in the USA: http://www.cem.com/applctns/AcdDgst.html The UK supplier is CEM (Microwave Technology) Ltd, Unit 2 Middle Slade, Buckingham Industrial Park, Buckingham MK18 1WA, UK Tel. +44 (0) 1280 822873; Fax. +44 (0) 1280 822342. Their HP-500 Plus vessel system can handle 14 soil or plant digestions at a time. Microwave systems are also used for accelerated Soxhlet extractions with reduced solvent consumption, and microwave muffle furnaces with airexhaust for rapid ashing.

Dry ashing Dry ashing is normally carried out in a muffle furnace. Large numbers of silica basins or crucibles take up a considerable amount of floor area within the furnace, therefore the larger the capacity the better. It may be advantageous to have two furnaces. Typical specifications would be: interior dimensions (depth × width × height) volume maximum power rating

457 × 305 × 203 mm 27 l 7 kW

Chapter 4

Suitable furnaces are available from: Stuart Scientific (Bibby): http://www.bibby-sterilin.com/cat/stuart/ furnaces.htm#muffle Carbolite Furnaces Ltd: Aston Lane, Hope, Hope Valley, Sheffield S30 2RR, UK. Tel. +44 (0) 1433 620011 Fax. +44 (0) 1433 621198 See also the following websites: http://www.catalogue.fisher.co.uk http://www.keison.co.uk/carbolite/carb39.htm Ashing in a furnace is a compromise between total oxidation of carbon and some vaporization of the element of interest. When this is for trace metals, such losses only become significant with the more volatile metals such as cadmium and lead. Even some iron may be lost if chlorides are present, as ferric chloride is appreciably volatile at 450°C. A useful review of ashing biological material for the determination of trace metals was provided by Middleton and Stuckey (1953, 1954). They recommended an ashing temperature of 500– 550°C (dull red heat) as the lowest temperature at which combustion can be completed in a reasonable time when the trace metal is volatile. Sometimes the sample is first moistened with sulphuric acid when lead is being determined, to convert it to lead sulphate, which is involatile below 550°C. When ashing for trace element determination, we prefer to err on the side of caution, and recommend ashing overnight at 450°C . Sometimes an additional treatment is required such as for manganese solubilization. Ashing converts manganese salts to manganese dioxide, which is virtually insoluble in dilute acids. The ash is therefore moistened with concentrated HCl and heated carefully on a hotplate until it has fumed dry. This converts the manganese dioxide to manganous chloride: MnO2 + 4HCl = MnCl2 + Cl2↑ + 2H2O

The residue is then dissolved in 0.1 M HCl in the normal way for the subsequent determination of elements by atomic absorption spectrophotometry. It is usually best to avoid the production of flames from the sample while ashing (deflagration) as this can result in some loss of analyte. The sample should therefore be placed in the muffle at room temperature, with the chimney vent open, allowing the combustible gases to evolve without ignition as it heats up to the final ashing temperature. The vent is then closed to prevent a downdraught blowing the light ash out of the crucibles, and ashing is continued for the stated time. Dry ashing of animal tissues is problematic, and the above authors suggest using a mixture of nitric and sulphuric acids for the purpose (Middleton and Stuckey, 1954). This would be safer than mixtures involving perchloric acid, which may be explosive.

Extraction Procedures – Plant-based Materials Other extraction procedures are used for determining: (i) oils, fats and waxes; (ii) fibre, lignin, cellulose, nitrogen-free extract and starch; (iii) in vitro

Acid-digestion, Ashing and Extraction Procedures

digestibility; (iv) nitrate and water soluble carbohydrate; (v) water content in silage.

Oils, fats and waxes All substances in feedstuffs logically belong to one of the six components or groups of a ‘proximate analysis’. This concept goes back nearly 150 years to the first state agricultural research stations in Germany, and is also known as the Weende methods (Henneberg, 1864). They give crude, but useful, measurements of the components of feedstuffs, and adaptations of the original methods are used today. The components are as follows: water protein fat/oil fibre ash nitrogen-free extract

(from dry matter determination) (from nitrogen determination) (also known as ether extract) (also called crude fibre) (mineral content) (subtract sum of above from 100%; it is mainly carbohydrate/starch)

Fat includes triglycerides, sterols, lecithins (phospholipids), essential oils, fat-soluble pigments such as chlorophyll, and similar substances. The AOAC recommends that anhydrous diethyl ether kept over freshly cut sodium pieces is used for the extractant (Padmore, 1990, p. 79), but we prefer to use petroleum spirit, also called light petroleum and petroleum ether, with a boiling range of 40–60°C, as it is a less hazardous solvent. The sample should not be oven dried before analysis as this could oxidize or degrade the oil and result in too low a value. A separate sample should be taken for a dry matter determination allowing the result to be corrected to percentage fat in dry matter. The ground sample is placed in a cellulose extraction thimble of the correct size for the Soxhlet extraction glassware. The Whatman extraction thimbles are available in two thicknesses, and it is the more robust double thickness that is preferred. The size of Soxhlet flask should match that of the heating mantle recess. The 250-ml capacity recess is most appropriate, and heating units are available with three or six recesses, with the six-recess model being more economical per recess, and more suited to handle multiple samples. The heating units are specially designed to be spark free in normal operation. The flask is pre-dried and weighed, so a flat-bottomed flask is easier to handle. After extraction, the remaining solvent is evaporated off on a boiling water bath. When there is no longer any smell of solvent, the flask is again dried in an oven (102°C), cooled and weighed. The weight of oil remaining in the flask is found by difference. For multiple samples (c. 5 g) of seeds such as oilseed rape, another approach is to crush them and enclose them in small packets of pre-dried Whatman No. 4 filter circles, which are then stapled, labelled with a pencil and weighed. From 10 to 16 of these packets may be extracted in one large

Chapter 4

600 ml Soxhlet with a 2-litre flask. In this case, the packets are reweighed, and the weight loss gives the crude oil content (Hughes, 1969). Oil in compound feeds and feeding stuffs This method published by MAFF (1993a) is not applicable to oilseeds or compound feeds containing milk powder. The sample is extracted with light petroleum and the residue then heated with 3 M HCl. This is filtered, washed, dried and re-extracted with light petroleum. There are several producers of automatically controlled Soxhlet extractors, which require their proprietary glassware. Some examples are: Büchi Labortechnik AG: http://www.buchi.com/ Foss: http://www.foss.dk/foss.asp Gerhardt: http://www.gerhardt.de/gb/soxt.htm Soxhlet heating mantles are produced by: Electrothermal Engineering Ltd: http://www.electrothermaluk.com/files/prodcore.htm

Fibre, lignin, cellulose, nitrogen-free extract and starch Fibre can mean many things. Crude fibre is an attempt to measure the roughage material in a feedstuff that is indigestible as far as the animal is concerned. It is an attempt to approximate the effect on the feedstuff of the digestive processes within the digestive tract by the use of inorganic chemicals, in this case, boiling dilute sulphuric acid, then boiling dilute sodium hydroxide, and the weight loss on ignition (which corrects for mineral ash content) of the residue is the fibre content. There are many modifications of this method, they may be to make the process more representative of the ruminant digestive system, or the desired residue may be just the plant cell walls. It is not always possible to say that one procedure is better than another, therefore the chosen procedure may be that which has been used by workers involved in animal nutrition over a number of years in a certain geographical area. The decision may be to use the usual procedure favoured by the referees for research papers in a particular journal. For several decades one of the leading authorities on the extraction of fibre from feedstuffs with particular reference to ruminant nutrition has been Professor P.J. Van Soest. His book, Nutritional Ecology of the Ruminant, has many helpful details (Van Soest, 1982, 1994). Various detergents are used to fractionate forage matter into its components. Neutral detergent is useful for separating the insoluble plant cell wall fraction, which is only partially digested by ruminal microorganisms (Van Soest and Wine, 1967). He considered that rather than the crude fibre (which entails the loss of soluble fibre components and variable amounts of the hemicelluloses and lignin), it is the proportion of plant cell wall and its degree of lignification that best determines the character and nutritive value of feeds and forage. The cellular contents determine the proportion of completely available nutrients, and consist

Acid-digestion, Ashing and Extraction Procedures

of the bulk of the protein, starch, sugars, lipids, organic acids and soluble ash. The various processes are given in Table 4.1. Table 4.1. Effect of detergents and reagents in forage analysis. Residue type

Reagent

Process

Products

Acid detergent fibre (ADF)a

Cetyl Boil for 1 h trimethylammonium bromide in 0.5 M H2SO4

Lignocellulose + insoluble mineral

Neutral detergent fibre (NDF)

Sodium lauryl sulphate, EDTA, pH 7.0

Boil for 1 h

Herbage cell wall minus pectins

Unavailable N

Acid detergent

Kjeldahl nitrogen on ADF residue

Maillard products plus lignified N

Cellulosea

None required

Ash from lignin step

By weight loss

Lignina

72% H2SO4 on ADF

3 h @ 20°C

Crude lignin

Hemicellulose

Not required

Calculate NDF-ADF Hemicellulose by difference

Silica (SiO2)

Conc. HBr treatment of ADF ash

Add dropwise to ash,1 h @ 25°C

SiO2 residue

Van Soest and Wine (1968).

In the NDF method, the sodium lauryl sulphate (sodium dodecyl sulphate) forms strong protein complexes which are soluble under the right conditions. The EDTA-disodium salt complexes with any Ca or Mg which would otherwise be included with the cell walls. The addition of sodium sulphite to cleave disulphide bridges in any added animal protein (e.g. keratin) in the feed is usually omitted unless major quantities of such substances are present. This is because the sulphite will also dissolve cell wall lignin, reducing its recovery (Moir, 1982; Van Soest et al., 1991). The reagent 2-ethoxyethanol, which aids solution of starches, is toxic and has been replaced by triethylene glycol (Cherney, 2000). Van Soest later recommended omission of anti-foaming decahydronaphthalene (decalin) because it greatly slowed the filtration step (Van Soest, 1973). The ADF method has tended to replace the crude fibre procedure, especially when further fractionating the feed into lignin and cellulose. However, for an improved correlation between acid detergent fibre and ruminant digestibility, the modified acid detergent fibre (MADF) method of Clancy and Wilson was developed in Ireland (Clancy and Wilson, 1966). Although NIRS is currently the preferred technique to predict OMD (organic matter digestibility) which is then converted to a ME (metabolizable energy) value, this expensive procedure is rarely available to smaller laboratories. Prediction equations

Chapter 4

were developed for ME from MADF values, and although somewhat less accurate, they may give a working basis for ration formulation. Some examples are given below: Fresh grass: ME (MJ kg–1 DM) = 16.20 – 0.0185[MADF] (Givens et al., 1990) Grass hays: ME = 15.86 – 0.0189[MADF] (Moss and Givens, 1990) Grass silage: ME = 15.0 – 0.0140[MADF] (Givens et al., 1989)

This subject is discussed in depth in Chapter 4, ‘Feed evaluation and diet formulation’ in the Agriculture and Food Research Council (AFRC) advisory manual Energy and Protein Requirements of Ruminants (Alderman and Cottrill, 1993), and more recently by Coleman et al. (1999). With non-ruminants, the only fibre determination required is by neutral detergent. Ruminants and other herbivores, which can partially digest fibre, will need the ADF or MADF methods. Lignin and cellulose Lignin, like fibre, is a complex substance. Lignins are phenolic polymers that occur in plant cell walls, and they impart, with cellulose, rigidity to stems. There are several molecular building blocks in lignin. When oxidized with nitrobenzene, lignin from angiosperms (grasses, herbs and flowers) yield p-hydroxybenzaldehyde, vanillaldehyde (from coniferyl alcohol component) and syringaldehyde. Lignin from gymnosperms (coniferous trees), however, lacks the syringyl group (Harborne, 1984). A typical lignin structural unit is shown in Fig. 4.2. Estimation of lignin is complicated by the presence of strongly bound proteins. Other contaminants are carbohydrates, chemically bonded cinnamic acids, cutins and tannins. A partial loss of lignin may also occur in the determination, and it is not yet possible to prepare a pure analytical lignin fraction. The relative merits of about 15 procedures are reviewed by Cherney (2000). In the procedure described by Van Soest and Wine (1968), the crucible plus residue from the ADF method is left to stand for 1.5 h in a buffered potassium permanganate solution to dissolve the lignin. The cellulose residue is reacted with demineralizing solution until white, washed successively with 80% ethanol and acetone, dried overnight at 100°C, cooled and weighed. The loss in weight is equal to the lignin content. The crucible may be ashed for 3 h at 500°C and the loss in weight is the cellulose content. Nitrogen-free extract (Nifext) This is obtained by subtracting the sum of the percentages of water, protein, fat, fibre and ash from 100. It represents the starch, gums, sugars and organic acids (all N-free), which may be extracted by water or diastase from cleaned, dried and defatted foods. As it is mainly starch, it will be high in the case of cereal grains and lower with seeds containing more oil and protein.

Acid-digestion, Ashing and Extraction Procedures

CH2OH CH CHOH

p-hydroxybenzyl section

HOCH2

Coniferyl group

OCH3

CH2OH CH CH

Syringyl group H3CO

OCH3 OH

Fig. 4.2. Typical lignin structural unit.

Starch Starch may be determined using specific enzymes such as amyloglucosidase, but the extraction and hydrolysis stage is slow, enzyme activity can vary, reagents are expensive and complete hydrolysis is difficult. Although acidhydrolysis lacks specificity, with the simple case of starch in potatoes, it becomes an ideal procedure (Faithfull, 1990). The freeze-dried, milled sample is washed with 10% v/v ethanol/water to remove sugars, dextrins and tannins which can amount to about 12%. Note: The use of 80% v/v ethanol/water is recommended for pre-extraction with enzyme methods, followed by heat treatment to gelatinize the starch; 90% v/v ethanol/water tends to make the starch resistant to enzymatic hydrolysis (Hall et al., 2001). The suspension is centrifuged, washed into McCartney bottles using 1 M HCl and heated at 106°C for 40 min. After adjusting the pH to 3.0, it is diluted to 100 ml, and a further dilution with saturated benzoic acid solution provides the solution for analysis of the products of starch hydrolysis. Starch has been given the formula C36H62O31.12H2O, with residue units of C6H10O5. The hydrolysis is to units of glucose, C6H12O6, so when using the weight of glucose to

Chapter 4

determine the initial weight of starch, a correction factor of ×0.9 is required. Although one would anticipate the hydrolysis to yield only glucose, some of the glucose is subsequently converted by the hot acid to fructose and 5-hydroxymethylfurfural which produce more colour with the anthrone reagent than glucose itself. The use of fructose solutions as standards corrects for this effect, otherwise a correction factor of ×0.8 is used with glucose standards. Partial hydrolysis of potato cell walls to chromogenic products led to a further correction factor of ×0.98, and a correction for any moisture must be made.

In vitro digestibility The estimation of animal digestibility of a feedstuff is usually achieved in one of three ways: in vivo, in sacco or in vitro. The first uses real animals in feeding trials and gives the most realistic results to which the other methods are correlated; the second method allows feed samples contained in small permeable plastic (e.g. nylon) bags to be inserted through a cannula into the rumen or another section of the digestive tract. The last method allows the digestion of feed samples to occur in the laboratory using digestive juices obtained from a fistulated animal, commercially obtainable enzymes, detergent solutions, or any combination of these, with the aim of imitating naturally occurring digestive processes. Extractions using detergent plus enzyme Neutral cellulase plus gamanase digestibility (NCGD) of feeding stuffs. This method originated at a time when compound feeds contained less starch and more digestible fibre and oil than when ME prediction equations were derived in 1985. In this method published by MAFF (1993b), using a fat-free sample, neutral detergent removes soluble cell contents, α-amylase dissolves starch, while cellulase/polysaccharase dissolves cellulose and hydrolyses any polysaccharides in the feed, and gamanase hydrolyses galactomannans which occur in palm kernel products. Neutral detergent (plus amylase) fibre (NDF) of feeding stuffs. This other MAFF (1993c) method removes cell contents from the fat-free sample by boiling with neutral detergent solution. The α-amylase converts any starch (which would enhance the fibre content) to soluble sugars. The residue is designated neutral detergent (plus amylase) fibre; the abbreviation given is NDF, but this would confuse it with the Van Soest and Wine NDF. Perhaps ND(+A)F would be clearer. Rumen liquor plus neutral detergent To obtain the in vitro true digestibility, the residue from the first buffered rumen liquor stage of the Tilley and Terry (1963) procedure is digested with neutral detergent solution. The ordinary true digestibility is found by

Acid-digestion, Ashing and Extraction Procedures

subjecting the faeces to neutral detergent digestion. The neutral detergent soluble non-cell-wall fraction of faeces equates to the endogenous and bacterial loss. Tilley and Terry (1963) procedure This eponymous method is widely used and as originally proposed or in modified form has served as a benchmark for other methods. In fact, it is often referred to simply as the in vitro digestibility. The first stage involves anaerobic incubation at 38°C in the dark with partially filtered rumen liquor which has been buffered with McDougall’s artificial saliva solution, previously saturated with CO2. After 48 h, 5 ml of M Na2CO3 is added to aid sedimentation immediately before centrifugation. Although mercuric chloride was added to inhibit bacterial activity, immediately centrifuging after 48 h rather than storing samples avoids this. It also avoids disposal of a toxic reagent. The supernatant is decanted into a fine nylon cloth filter and any particles returned to the tube. The particles on the rubber stopper and those adhering to the sides of the tube are washed down to the pellet which is then broken up before adding the acid pepsin solution. This is incubated for a further 48 h, then filtered through a porous alumina crucible (unpublished modification to the original method) before oven-drying, weighing, and possibly ashing. In vitro calculations The Tilley and Terry method (X) correlates with in vivo results (Y) as follows: Y = 0.99X –1.01

One particular correction is advisable. Standards of known in vivo and in vitro values covering the lower and higher digestibility range (about 50% and 70% respectively) should be obtained, possibly from a research station, and included with the sample batch. Rumen liquor varies in potency from week to week, therefore a proportional adjustment must be made to enable comparison of results from analyses performed at different times. This variation does not equally affect the low and high standards; one may decrease and the other increase. A graph should be drawn relating the difference of the measured standard from the stated value to the concentration. This could be a positive or negative slope. The samples’ measured digestibility should be corrected according to the corresponding adjustment read off the graph. A typical example is shown in Fig. 4.3. Alternatively, a spreadsheet program such as Microsoft Excel may be used to achieve the correction automatically. A typical example is shown in Table 4.2. Digestibility equations There are several ways of expressing the in vitro rumen liquor digestibility of a sample: the DOMD or D-value, the DMD value and the OMD value. These are defined below:

Chapter 4

1. The DOMD (D-value) is the digestible organic matter in dry matter: = OM sample – (OM residue – OM blank) × 100% DM sample

Measured DMD (%)

where OM is the organic matter in the original dried and milled sample (sample minus sample ash), OM residue is the organic material in the residue

90.0 80.0 70.0 60.0 50.0 40.0 30.0 -1.0

-0.5

0.0

0.5

1.0

Correction (%) Fig. 4.3. Typical graph for correcting measured sample dry matter digestibility (DMD) values in proportion to deviation of low and high standards from their declared values. Table 4.2. Typical spreadsheet for correcting the measured sample dry matter digestibility values in proportion to deviation of low and high standards from their declared values. Spreadsheet for correction of digestibility values between batches

High standard Low standard

Stated value (%)

Measured value (%)

Correction required

72.9 50.9

71.9 51.4

1.0 -0.5

Corrected value (%) Sample 1 Sample 2 Sample 3

51.8 62.5 73.2

52.2 62.2 72.2

Let correction graph be y = mx + c y = correction to be applied m = slope x = measured sample value mx = Q.10 c = intercept on y-axis Vc = corrected value; Vm = measured value y = mx1 + c = y = mx2 + c y = mx3 + c Vc = Vm + y

Spreadsheet formulas

Result

((B4-C4) - (B5-C5))/(C4-C5) 0.073 C8 52.200 ((B4-C4)-(B5-C5))/(C4-C5)*C8 3.820 (D4-(C4/(C4-C5))*((B4-C4)-B5-C5))) -4.261

((B4-C4)-(B5-C5))/(C4-C5)*C8+(D4-(C4/(C4-C5))*((B4-C4)-(B5-C5))) ((B4-C4)-(B5-C5))/(C4-C5)*C9+(D4-(C4/(C4-C5))*((B4-C4)-(B5-C5))) ((B4-C4)-(B5-C5))/(C4-C5)*C10+(D4-(C4/(C4-C5))*((B4-C4)-(B5-C5))) For Sample 1: B8 = C8 + y C8+E20 For Sample 2: B9 = C9 + y C9+E21 For Sample 3: B10 = C10 + y C10+E22

-0.441 0.290 1.022 51.759 62.490 73.222

Acid-digestion, Ashing and Extraction Procedures

following digestion (residue weight minus ashed residue weight), OM blank is the organic matter in the rumen liquor itself, and DM sample is a dry matter determination done on a separate sample. This equation must be translated into the actual weighings required so that a spreadsheet can be drawn up. Errors can easily occur in the calculations unless the individual steps are understood. The sample weight is 0.5000 g and the calculation formula and any spreadsheet must be designed for this and allow for the fact that the original ash is carried out on 1.0000 g. The residue of undigested sample contains four components of the calculation: • • • •

undigested sample organic matter sample residue ash blank (rumen liquor) organic matter blank ash.

We are interested in the first component, so need to subtract the other components. The residue from the rumen liquor blank contains both blank organic matter and blank ash. When this value is subtracted from the above we get the sum of sample organic matter plus sample ash. After weighing the dried residue it is subsequently ashed and weighed. This gives an ash comprising: • sample residue ash • blank ash. A separate ashing of a rumen liquor blank sample gives a figure for blank ash. When subtracted from the above residue ash, the difference gives the sample residue ash. Subtracting this value from the sum of sample organic matter plus ash, leaves us with the undigested residue sample organic matter. Finally, this is subtracted from the original sample organic matter to give the amount of digestible organic matter, which is corrected for dry matter content of the sample and expressed as a percentage or as g kg–1 digestibility. DOMD =  0.5 − original   sample residue −  sample residue ash −     −   −   ash    blank residue   blank residue ash     × 100% 0.5

2. The OMD value is the organic matter digestibility: = OM sample (OM residue OM blank) × 100% OM sample OMD =  0.5 − original   sample residue −  sample residue ash −     −   −   ash    blank residue   blank residue ash     × 100% 0.5 − original ash

Chapter 4

3. The DMD value is the dry matter digestibility: = [sample (sample residue blank residue)] × 100% sample DMD =

{0.5 − (sample residue − blank residue)} × 200% A suggested spreadsheet for the above calculations is shown in Table 4.3. The values for residues and ash are entered from results sheets printed with columns for crucible weights, etc., unless the laboratory is equipped with computerized balances, when a more sophisticated spreadsheet could be devised. When planning for Tilley and Terry digestibilities, it is common practice to ensure that the sheep or cattle have been fed for a couple of weeks on a basal diet similar to the test samples to be analysed. This is to ensure a buildup of the appropriate rumen flora resulting in a corresponding optimal activity. Whether or not this is necessary is open to question, and this and other sources of error have been discussed by Ayres (1991). It is also customary not to feed the animal on the morning planned for extracting the rumen liquor. 4. True dry matter digestibility (True DMD) (Van Soest et al.,1966) This is expressed by the equation: True DMD = {(% cell content in DM × 0.98) + (% digestible cell wall in DM)}

The % cell content in DM is (100 – % cell wall in DM), which is derived from (100 – NDF). The % digestible cell wall in DM is the (% cell wall in DM – % indigestible cell wall in DM). The % indigestible cell wall in DM is the residual DM after digestion in rumen liquor (48 h) followed by the neutral detergent procedure and expressed as % sample DM. Various aspects of in vitro methods, from its first use in 1880 to the 1980s have been discussed by the author (Faithfull, 1984). In particular, the effect of pH on tannin complexes, phosphates and sulphides have been studied. The concept of fistulated animals may seem abhorrent. It should be observed, however, that properly tended animals appear to be quite contented, and that their lifetime as an experimental animal is far longer than it would otherwise have been. Nevertheless, it is impossible to prevent the animal from knocking the cannula, and it is easy for leaks to occur causing irritation to the skin around it. It is also expensive to maintain such animals in an acceptable way, and to justify this if long periods exist between experiments. The procedure is favoured by experienced researchers as it facilitates comparison of results with earlier published work, and may give more consistent results over periods of time. However, improved within-batch precision, economy of time, money and convenience, and improved public

Name In vitro DOMD, OMD & DMD measurement Sample

Corrected

Sample Sample Sample Fractional S dry ID wt. DM sample wt. DM

Sample Fractional Sample ash %. ash OM

1 2

7.24 6.80

3 4 5 6 7 8 9 10 Standard L1 L2 H1 H2 M19 1 2

0.5000 0.5000

98.01 97.22

0.5000 0.5000 0.5000 0.5000 0.5000 0.5000 0.5000 0.5000 0.5000 0.5000 0.5000 0.5000

98.25 97.56 98.18 97.73 97.66 97.94 98.12 96.97 98.44 97.89 98.35 97.17

0.9801 Formula =D5/100 0.9825 0.9756 0.9818 0.9773 0.9766 0.9794 0.9812 0.9697 0.9844 0.9789 0.9835 0.9717

0.4901 =samplewt *E5 0.4913 0.4878 0.4909 0.4887 0.4883 0.4897 0.4906 0.4849 0.4922 0.4895 0.4918 0.4859

8.43 7.48 6.61 7.53 7.96 8.01 8.22 7.87 6.98 7.34 7.97 6.84

0.0355 =G5/ 100*F5 0.0414 0.0365 0.0324 0.0368 0.0389 0.0392 0.0403 0.0382 0.0344 0.0359 0.0392 0.0332

Residual Residual Residual OM wt. ash

0.4546 0.1891 =F5-G5 0.1773

0.0085 0.0072

0.4498 0.4513 0.4585 0.4519 0.4494 0.4505 0.4503 0.4467 0.4578 0.4535 0.4526 0.4526

0.0088 0.0079 0.0076 0.0081 0.0077 0.0083 0.0080 0.0077 0.0082 0.0090 0.0083 0.0076 0.0009 0.0009

0.1896 0.1907 0.1814 0.1882 0.1901 0.1866 0.1871 0.1888 0.1903 0.1855 0.1899 0.1977 0.0040 0.0040

0.1775 =J5-K5 -M19 0.1777 0.1797 0.1707 0.1770 0.1793 0.1752 0.1760 0.1780 0.1790 0.1734 0.1785 0.1870 0.0031 =J19-K19

OMD g/kg

DOMD g/kg DM

DMD g/kg

Average

610 =(I5-L5) /I5*1000 605 602 628 608 601 611 609 602 609 618 606 587

565 =(I5-L5) /F5*1000 554 557 586 562 553 562 559 554 567 572 557 547

614 =(F5-J5) /F5*1000 614 609 630 615 611 619 619 611 613 621 614 593

=AVERAGE (M4:M5)

603

Acid-digestion, Ashing and Extraction Procedures

Table 4.3. Typical spreadsheet in Microsoft Excel for calculating the various digestibility values.

618 606 605 613 596

0.0031

The row for Sample 2 is used to display the formulae, which are normally hidden.

Chapter 4

perception all point to alternative methods as being the way forward. One such method uses faecal liquor and has been discussed by Omed et al. (2000). Replacing the acid pepsin stage with biological washing liquid produced digestibilities very close to the known in vivo values for a variety of grasses, legumes and hays (Solangi, 1997). The two-stage pepsin–cellulase method (see below) is probably the best alternative to the Tilley and Terry procedure. Cellulase digestibility A convenient procedure for assessing the digestibility of forages is the cellulase digestibility technique. This was refined by Jones and Hayward (1973) at the Welsh Plant Breeding Station (WPBS) in Aberystwyth (since 1992, the Institute for Grassland and Environmental Research). It was later extended to a two-stage procedure with a pepsin pre-treatment (Jones and Hayward, 1975). The pepsin removes protein from the cell walls and possibly modifies the cell wall polysaccharide in such a way as to render it more susceptible to attack by the cellulase enzyme. It also allows cellulases from different sources to be used with less effect from variation in enzyme activity. The single stage cellulase technique is suggested for screening in plant breeding programmes, but in this case, the higher activity enzyme from Trichoderma viride will yield a higher correlation with in vivo and in vitro digestibility. One might expect less precision when digesting with enzymes versus rumen liquor, because enzymes lack the ability of microorganisms in adapting to a substrate. Stakelum et al. (1988), however, found a similar accuracy in predicting in vivo digestibility when using the rumen liquor–pepsin, pepsin–cellulase or neutral detergent–cellulase methods.

Nitrate and water-soluble carbohydrate The same extractant is used for both nitrate and water-soluble carbohydrate (WSC) determinations, however the ratio of sample to extractant is different. The herbage may be oven-dried for nitrate, but must be freeze-dried for WSC determination. The extractant is saturated benzoic acid solution. Benzoic acid is sparingly soluble in cold water, and the solution is made by adding an excess quantity to deionized water at ambient temperature in a blender, which is then switched on for about a minute. It is filtered through a Whatman No. 4 paper into a storage container fitted with a tap. If the ambient temperature should fall several degrees, it is possible for some crystals to separate out. These would make little quantitative difference, but might block the sample capillary probe or tubing. If this is thought likely, the containers for samples and standards should be warmed and shaken gently to redissolve the crystals. The benzoic acid acts as a preservative, allowing the storage of sample extracts at room temperature almost indefinitely, so they can be analysed at a convenient time. It has been noticed, however, that the concentration of nitrite (as opposed to nitrate, which is stable) decreases to zero after a day or

Acid-digestion, Ashing and Extraction Procedures

so. Methods estimating nitrite, therefore, must use an extractant such as water, followed by immediate analysis. Nitrate The autoanalysis method was developed at the WPBS and modified by using benzoic acid extractant solution. It is based on the method of Follett and Ratcliff (1963), which was itself based on that of Grace and Mirna (1957). It relies on the reduction of nitrate to nitrite by adding the sample solution to an ammonium chloride buffer (pH 7.5) containing EDTA disodium salt and copper sulphate and passing through a glass tube containing cadmium filings which become copper-plated. The nitrite immediately reacts with sulphanilamide to form a diazo salt which couples with 8-aminonaphthalene-2-sulphonic acid (Cleve’s acid) to form an orange acid azo dye which is measured at 470 nm on a spectrophotometer. Another method used for nitrate determination on dried and milled herbage employs the nitrate selective electrode. One of the first published methods was that of Paul and Carlson (1968). Other anions, especially chloride, can interfere. These authors removed chloride with silver resin, but Barker et al. (1971) omitted the resin because it tended to foul the electrode and cause excessive drift. Normally the Cl–:NO3– ratio is so low as not to interfere, but saline precipitation from coastal plots could affect this. The method was further modified to allow storage of extracts for up to 64 h by adding a preservative of phenyl-mercuric acetate and dioxane, both very toxic (Baker and Smith, 1969). This paper mentions the need to change the electrode’s membrane, filling solution and liquid ion exchanger every 2 months to minimize chloride interference. It is easy to overlook electrode maintenance between batches of nitrate analyses, and this can lead to errors and sluggish performance. The method was extended from plants to include soils and waters by Milham et al. (1970). They point out that nitrate reductase activity in fresh plant samples often causes a rapid decline in nitrate content, so samples collected from remote sites should be frozen in dry ice. A trace of chloroform was used to protect soil and water samples before freezing. We are now more aware of the harmful effects of chloroform inhalation and suggest immediate freezing without preservative and analysis within a few days as a safer alternative – especially with student projects. One drawback with selective ion electrodes is their slow response at low concentrations of analyte, perhaps below 2 mg l–1 NO3-N. It can take several minutes to equilibrate, and slow drifting can give a measure of uncertainty as to the equilibration point. If this cannot be remedied by reducing the dilution factor, an alternative method should be sought. They are also sensitive to changes in temperature, in excess of 1°C being significant. Mechanically driven magnetic stirrers get warm, therefore electronic ones are preferable.

Chapter 4

Water soluble carbohydrate This is basically the anthrone method of Yemm and Willis (1954) which was developed at the WPBS for use with an autoanalyser (Thomas, 1977) and modified by using benzoic acid extractant solution. The extract is reacted with anthrone in 76% sulphuric acid. Heating to 95°C develops the green colour which is measured at 620 nm. Fructose, sucrose and inulin give the colour at room temperature, but heating is necessary for glucose, maltose, fucose and rhamnose to react (Van Handel, 1967). Fructose and glucose are hydrolysed by hot sulphuric acid to 5-hydroxymethyl-furfural which reacts with anthrone to give 10-{5-(anthron-10-ylmethyl)-2-furfurylidene} anthrone, which couples with brown resin by-products to give the colour (ho*rmann, 1968).

Water content in silage Silage moisture consists of both water and volatile fatty acids (VFAs). To ovendry silage would remove both the water and the nutritional VFAs which should be included with the DM. The most widely used method to correct for this loss is the toluene distillation method of Dewar and McDonald (1961). The Karl Fischer titration is probably the most accurate, but uses anhydrous methanol. Oven drying and using correction equations makes assumptions which may not be valid in every case. NIRS involves very expensive equipment and extensive calibration. Various pros and cons have been discussed by Givens et al. (2000). The main deficiency of the toluene distillation is its inability to account for the alcohol content of the volatiles. The other drawback is the large quantities of toluene involved. It is, however, a simple method and is widely quoted. We suggest a smaller scale procedure which uses less solvent, and recovers used solvent by distillation and drying over anhydrous sodium sulphate. If the accuracy requires the alcohol content to be determined, this may be done separately by GLC. The method also enables small core samples to be analysed, which simplifies the profiling of silage clamps for nutrient analyses (Faithfull, 1998).

Extraction Procedures – Soils There are many different types of soil, and extractant formulations have been fine-tuned to suit the soil. The particular extractant may also be chosen on the basis of familiarity over the years, and because it is easier to compare results with those previously obtained, and hence make recommendations to correct deficiencies based on experience. Usually one is not interested in the total amount of a soil nutrient, rather in the amount that is in a form available to the roots of the plant. Regional advisory laboratories over a long period may have developed index tables relating to the found concentration of nutrient in local soil types and the corrective amount of fertilizer required. It would probably be wise to adopt the same methods that have been used to

Acid-digestion, Ashing and Extraction Procedures

derive these tables, unless they have been found to be inadequate. We will refer to the UK MAFF/ADAS publications in the appropriate chapter. There are, however, published procedures on the web, particularly from the USA. Two such manuals are available from Delaware Cooperative Extension (1995): Recommended Soil Testing Procedures for the Northeastern United States, 2nd Edition, and from the Missouri Agricultural Experiment Station (1998): Recommended Chemical Soil Test Procedures for the North Central Region at their respective websites: http://bluehen.ags.udel.edu/deces/prod_agric/title-95.htm http://muextension.missouri.edu/xplorpdf/miscpubs/sb1001.pdf The United States Department of Agriculture (USDA, 1996) has also published a methods manual.

pH extractants The apparently simplest of procedures faces one with a choice of about four extractants. The commonest extractant is water, and the ratio we use is 10 ml soil:25 ml water, i.e. 1:2.5 v/v. Other ratios used by the Northeastern United States are 1:1 v/v, 1:1 w/v and 1:2 v/v soil/water (Delaware Cooperative Extension, 1995, Appendix). Some soils have a significant soluble salt content, which can affect the measured pH. The concentration of these salts in the soil varies with the season, with dry season pH values being lower than wet season ones. This is because salts such as sulphates and nitrates, which lower pH, accumulate in dry periods and are leached away in rainy periods. To overcome this effect, a 1 M KCl extractant was first used. The pH values so obtained are 1.5–2.0 units less than those with water extractant, and are also affected by variations in the soil:extractant ratio. It is still used to assess the aluminium status of the soil. Values below pH 5 indicate significant amounts of Al, and if very much lower than 5, almost all the acidity is in the form of Al (USDA, 1996, p. 149). The aluminium acts by displacing hydrogen ions from the exchange sites on the surface of clay and humus particles to increase the acidity by raising the hydrogen ion (H+) concentration. It was later proposed that more suitable extractants to overcome, and also measure, the salt effects which displace hydrogen ions in a seasonal manner would be either 0.1 M KCl or 0.01 M CaCl2, with the latter being more widely used (Schofield and Taylor, 1955). The effect on mineral soils with a permanent negative charge, or on organic soils with a negative charge which varies with pH, is to displace H+ and lower the measured pH by about 0.5 units compared with water extractant. The effect on mineral soils dominated by sesquioxides, kaolinite and allophane with variable charge is that the salt causes adsorption of H+ onto reactive sites, raising the pH by about 0.5 units (Rowell, 1994, p. 161). The difference in pH between water and salt solution extracts is known as the salt effect, and given the symbol ∆ pH. Thus, ∆ pH = soil pH in salt solution – soil pH in water

Chapter 4

and ∆ pH values are positive for soils with a net positive charge, and negative for soils with a net negative charge, with magnitude proportional to charge.

Phosphate extractants Phosphorus occurs in various soil fractions: as soil minerals combined with Ca, Fe, Al, which are of low solubility; bound to particle surfaces of, e.g. sesquioxides, calcite, to Al on humus surfaces; in soil solution; in the organic matter, primarily as esters. Again, there are several choices of extractant, and the preferred one depends mainly on the type of soil under test. One of the most widely used procedures is the Olsen method (Olsen et al., 1954), which was developed in the USA to correlate crop response to fertilizer on calcareous soils. The amount of P extracted will vary with temperature (increases by 0.43 mg P kg–1 per degree rise between 20°C and 30°C) and shaking speed, so conditions should be standardized. The extractant is 0.5 M sodium bicarbonate adjusted to pH 8.5. The bicarbonate competes with phosphate on the adsorption sites extracts, and removes most, but not all of it, together with some soluble calcium phosphate. Addition of phosphate-free activated carbon before shaking is necessary if coloured soil extracts are obtained, and then they will require filtration. The northeastern United States have soils where the P chemistry is affected by aluminium phosphates. They therefore use dilute acid extractants to dissolve these minerals and extract the P. They use several procedures: (i) The Mehlich 1 Extraction (dilute double acid extractant) containing 0.0125 M H2SO4 + 0.05 M HCl (Mehlich, 1953). (ii) The Mehlich 3 Extraction using 0.2 M acetic acid + 0.25 M ammonium nitrate + 0.015 M NH4F + 0.013 M nitric acid + 0.001 M EDTA (ethylenediaminetetraacetic acid) (Mehlich, 1984). The pH should be 2.5. (iii) The Morgan Extraction using 0.72 M NaOAc (sodium acetate) + 0.52 M acetic acid at pH 4.8 (Morgan, 1941). (iv) The Modified Morgan Extraction (McIntosh, 1969) using 0.62 M NH4OH + 1.25 M acetic acid at pH 4.8. The resulting extracts are used for the appropriate colorimetric reaction and absorbances are measured on a colorimeter or spectrophotometer, possibly coupled to an autoanalyser. Note: the Olsen method is not to be confused with the Olson method (Olson et al., 1954), which uses sodium carbonate. The North Central Region, in addition to the Olsen method, uses the Bray and Kurtz P-1 test for phosphorus (Bray and Kurtz, 1945), which has proved to be well correlated with crop response to phosphate fertilizer on acid to neutral soils in the region. Each state experiment station has developed correlations and calibrations for the particular soil conditions within its own state, so field experience over a number of years or decades is necessary when deciding which methods to adopt. When bringing samples from remote sites back to the laboratory, it is therefore important to assess the nature of the soil at that site in order to choose the optimum method. If the same method has

Acid-digestion, Ashing and Extraction Procedures

to be used for reasons of comparability, then it is necessary to state that the available phosphorus content was obtained using a particular named method. Another method used by a few laboratories will be briefly mentioned; that is determination by resin extraction. The latest fertilizer recommendations by MAFF/ADAS (2000) include a classification of soils from the resin P values obtained using the method of Hislop and Cooke (1968). This was developed in the 1960s at the Levington Research Station, Ipswich. The method was intended to reflect the soil phosphate capacity, intensity and kinetic (rate of release) components. It was also designed to avoid inducing any major change in the chemical constitution of the soil as a result of the applied extraction procedure. The anion exchange resin (De Acidite FF 510, particle size >0.5 mm) was considered to be an inert phosphate sink. The method is outlined as follows: a subsample of 20 g air-dry soil, 2.0 mm, is ground in a Glen Creston Micro Hammer Mill fitted with a 0.5-mm screen. A 2-ml scoop of soil is then transferred to a 6 ounce (170 ml) bottle followed by a 5-ml scoop of washed and dried resin and 100 ml distilled water. It is shaken end-overend for 16 h at 25°C, after which it is filtered through approximately 0.5 mm terylene netting and washed; this retains the resin and allows the soil to pass through. The resin is then transferred to a leaching tube and 50 ml sodium sulphate (70 g l–1) solution is added and the leaching controlled to last 20 min. The phosphate in the leachate is determined colorimetrically, either manually or using an autoanalyser, the method being based on Fogg and Wilkinson (1958). The resin procedure was correlated with the Olsen bicarbonate method and gave a correlation coefficient of 0.877 (significant at P0.001) for non-calcareous soils, and 0.830 for calcareous soils. The amount of phosphate extracted by the resin during 16 h shaking approaches a maximum, and reflects the quantity or capacity factor which dominates under agricultural conditions. For glasshouse soils, however, full extraction is not approached, and it is rather the intensity and kinetic factors which are reflected more than capacity, and which are considered to be more relevant in this situation. Others use the resin method of Somasire and Edwards, 1992. The latter involves extracting 5 g soil 1:20 (m/v) using 100 ml of water, 2.8 ml cation exchange resin and 4.0 ml anion exchange resin with shaking for 16 h; this is followed by extraction with 1 M ammonium chloride, pH 2.0, with 30 min shaking. The above extractants for phosphate have been mainly developed for conventional agriculture. Some methods have been developed for assessment of soils managed on the organic system, which will be discussed in a separate chapter. Tip: finding soil analysis methods on the web requires a more powerful search engine. Try searching for ‘soil test procedures’ using http://www. alltheweb.com or http://www.google.co.uk

Chapter 4

Potassium extractants Potassium occurs in soil clay minerals, feldspars and micas. The unweathered illite region of the clay mineral contains non-exchangeable K+, the weathered vermiculite region has exchangeable K+, while the intermediate region has slowly exchangeable K+. There is also available potassium in the soil solution. The extractant will leach the free potassium ions and displace the exchangeable and some slowly exchangeable K+ by replacing the K+ with Na+, H+ or NH4+, depending on the extractant. Various extractants are listed in Table 4.4. Table 4.4. Some extractants for potassium in soils used by various regional laboratories. Regional methodology Extractant North Central USA

North Eastern USA

MAFF/ADAS UK

Comments

Mehlich 3 1 M ammonium acetate

Non-calcareous soils Calcareous soils if Ca and Mg also to be extracted Morgan, Modified Morgan, All are acidic and could Mehlich 1 and Mehlich 3 extract some non-exchangeable K+. Ammonium based reagents extract more K+ than H+ or Na+ based ones 1 M ammonium nitrate Also used for Mg and Na

Trace element extractants The determination of total amounts in soil is valid for finding whether there are toxic levels of certain metals (e.g. after repeated slurry applications), and comparisons can be made with published tables of maximum recommended levels. Some typical and maximum values are shown in Table 4.5 (ADAS, 1987; DOE/NWC, 1981). Dutch values differ from those developed in the UK in that the intention is to allow the return of contaminated land to any potential use, rather than tailoring the level of remediation to the intended use of the land. The most recent values include general targets and intervention values (http://www.athene.freeserve.co.uk/sanaterre/guidelines/dutch.htm). The soil sample is ground to pass a 0.5-mm sieve, and 2.5 g taken for the analysis. There are two possible extractants. Firstly 1:4 HClO4 (60% by weight perchloric acid):HNO3 (70% by weight nitric acid), of which 25 ml is added to the sample. It is allowed to stand overnight, then heated at 100°C, next 180–200°C and finally at 240°C. The residue is dissolved in 6 N HCl, boiled, cooled, made to 50 ml and filtered before analysis by atomic absorption spectrophotometry. This can be a dangerous procedure with a risk of explosion, and the full details should be carefully followed as given in the original reference (MAFF/ADAS, 1986, p. 31). An alternative acid mixture,

Acid-digestion, Ashing and Extraction Procedures

Table 4.5. Typical and maximum recommended levels for some trace elements in soil. Metal

Typical value in uncontaminated soil

Maximum recommended level

Earth/sediment (mg kg–1 dry matter)

(mg kg–1)

(kg ha–1)a

(mg kg–1)

(kg ha–1)a

Target value

Intervention value

80 20 25 0.5 50

160 40 50 1 100

300 135 75 3 250

600 270 150 6 500

140 36 35 0.8 85

720 190 210 12 530

Zinc Copper Nickel Cadmium Lead a

Assumes 2000 t ha–1 to depth of 15 cm.

aqua regia, is now suggested, and although a highly corrosive reagent, there should be no risk of an explosion. The availability of the trace metals is easily determined without any of the above risks, and the results used to assess both deficiencies and toxicities. The metals need to be removed from the sites where they are bound to the soil particles by use of an even stronger binding agent than the soil. This is achieved with two possible complexing reagents: EDTA and DTPA. They are a class of chemicals known as complexones, which form complex molecules with metals in a cage-like structure called a chelate. CO O CH 2 Zn

N CH 2 COO

CH 2

CH 2 CH 2

CH 2 COO

Fig. 4.4. A four co-ordinated Zn-EDTA complex ion.

Chapter 4

EDTA is ethylenediaminetetraacetic acid, also called (ethylenedinitrilo) tetraacetic acid, mol. wt 292.24. Although having four carboxylic groups, it behaves as a dicarboxylic acid with two strongly acidic groups. It is used as the disodium or ammonium salt, the latter being formed in situ. One mole of the EDTA salt reacts in all cases with one mole of the metal irrespective of its valency state. A four co-ordinated zinc EDTA complex is shown in Fig. 4.4, and a six co-ordinated cobalt EDTA complex in Fig. 4.5. EDTA is known as hexa- (or sexa-) dentate, having up to six active metal-complexing sites per molecule. DTPA is diethylenetriaminepentaacetic acid, also known as diethylenetrinitrilopentaacetic acid, mol. wt 393.36. It is octo-dentate, having eight active metal-complexing sites per molecule. A diagrammatic representation of the DTPA molecule is shown in Fig. 4.6. The amount of metal extracted from the soil by both EDTA and DTPA is dependent on the pH, the metal being extracted, the soil:solution ratio, the concentration of chelating agent, the shaking time, the temperature, and the sample preparation procedure. Clearly, the methodology used should be clearly described and closely followed if repeatable work is to be possible, and comparison of results is to be meaningful. CO O

CH2

CH2 N

CH2

O CO

CH2 CH2 O CO

Fig. 4.5. A six co-ordinated Co-EDTA complex ion.

HOOC-CH2

CH2 - COOH CH2CH2 N

N CH2CH2 HOOC-CH2

CH2 - COOH

CH2 COOH Fig. 4.6. A diagrammatic representation of a molecule of DTPA.

Analysis of Soil and Compost

Soil Analytical Procedures Method 5.1. Determination of extractable boron The predominant form of boron in soil solution is H3BO3, but above pH 9.2, H2BO3 may predominate. Hot-water extraction is the most widely accepted procedure for determining the amount of boron that is available to plants, and correlated best with the incidence of black spot in garden beets (Missouri Agricultural Experiment Station, 1998). The final determination is best performed using an ICP spectrometer, but this may not always be available, so a colorimetric method will be described. Methods using either curcumin or azomethine-H are possible, but the latter will be suggested here. It is not only the reagent used in the MAFF/ADAS (1986, pp. 20–22) handbook, which is the method to be described (with Crown Copyright permission), but has been adopted by the Delaware Cooperative Extension (1995) as being rapid, reliable and requiring less sample preparation and handling than the curcumin method. The American methodology, however, omits the removal by ashing of any organic matter in the filtrate, which might interfere with the determination; it also adds 0.1% m/v CaCl2.2H2O to the water extractant to promote soil flocculation. Also, the filtration step is replaced by centrifugation in a plastic centrifuge tube at 2700 g for 15 min. The following method could be modified similarly if appropriate, but once adopted, should be adhered to for future comparison of results. Boron is obviously a component of borosilicate glassware, which should therefore be avoided. Apparatus should therefore be made of PTFE, soda © 2002 CAB International. Methods in Agricultural Chemical Analysis: a Practical Handbook (N.T. Faithfull)

Chapter 5

glass or silica. A fibre digestion apparatus may be suitable. It may even be possible to extract the soil by boiling with the extracting water in a sealed plastic bag or pouch (Mahler et al., 1984). Having said that, however, silica (or quartz) apparatus is expensive, especially for educational purposes. The comment by Bingham (1982) should therefore be noted: ‘We have not found it necessary to use special low-B glassware for the analysis of water, soil, or plant samples. Pyrex glassware or plastic ware has been entirely satisfactory.’ Presumably the magnitude of the blank reading would show whether there was a contamination by extraneous boron. Apparatus. • Flasks, 250 ml, conical with ground joint – silica (quartz), or soda glass. • Condenser – either silica cold finger condensers, effective length 140 mm, or soda glass air condensers, approximately 750 mm. • Evaporating basins – 20 ml, translucent silica, shallow form, with round bottom and spout. • Polyethylene tubes – 20 ml with hinged cap. Reagents. Note: all reagents must be stored in polyethylene containers. • Azomethine-H reagent – Dissolve 0.45 g of azomethine-H in 100 ml of 1% m/v L-ascorbic acid solution. Prepare fresh weekly and store in a refrigerator. • Boron stock standard solution, 100 µg B ml–1 – Dissolve 0.572 g of boric acid (H3BO3) in water and dilute to 1 l and mix. • Boron intermediate standard solution, 20 µg B ml–1 – Pipette 20 ml of the boron stock standard solution into a 100 ml volumetric flask and make up to the mark with water and mix. • Boron working standard solutions, 0–3 µg B ml–1 – Pipette 0, 1.0, 2.0, 5.0, 10.0 and 15.0 ml into 100 ml volumetric flasks and make up to the mark with water and mix. This will provide solutions containing 0, 0.2, 0.4, 1.0, 2.0 and 3.0 µg ml–1 of boron. • Buffer masking reagent – Dissolve 250 g of ammonium acetate and 15 g of EDTA, disodium salt, in 400 ml water. Carefully add 125 ml of glacial acetic acid. • Calcium hydroxide solution, saturated. • Hydrochloric acid, approximately M – Dilute 85 ml of hydrochloric acid, approximately 36% m/m HCl, to 1 l with water. • Sucrose. Procedure. Transfer 40 ml (2 × 20 ml plastic scoopfuls, struck off level without tapping) of air-dry soil, sieved to 2 mm, into a flask. Measure 80 ml of cold water into a boron-free container and bring to the boil. Transfer the boiling water to the flask containing the soil and attach a condenser. Reheat to boiling as quickly as possible, and continue to boil for exactly 5 min. Remove the flask from the heat source, and allow to stand for exactly 5 min. Filter under reduced pressure through a 125 mm Hartley funnel fitted with a 125 mm

Analysis of Soil and Compost

Whatman No. 2 filter paper, collecting the filtrate in a boron-free tube inside the filter flask. Terminate the filtration after 5 min, and retain the filtered extract for the determination of boron. Carry out a blank determination. Pipette 5 ml of each boron working standard solution into a silica evaporating basin, and add 0.25 g of sucrose and 2 ml of satd calcium hydroxide solution. Evaporate to dryness on a boiling water bath. Place the basin in a cold muffle furnace, slowly increase the temperature to 450°C and maintain this temperature for 2 h. Allow to cool, then add exactly 5 ml of approximately M hydrochloric acid and dissolve all soluble material. Filter through a 90-mm Whatman No. 541 filter paper. Transfer 1 ml of the filtrate to a polyethylene tube. Add 2 ml of buffer masking reagent, mix and add 2 ml of azomethine-H reagent. Mix well and allow to stand for 45 min. Measure the absorbance in a 10 mm optical cell at 420 nm. Construct a graph relating absorbance to µg of boron present. The absorbances corresponding to 0 and 3.0 µg of boron are approximately 0.2 and approximately 0.7 and should differ by approximately 0.45. Pipette 5 ml of each soil extract into a silica evaporating basin, add 0.25 g sucrose and 2 ml of satd calcium hydroxide solution, and proceed as above as far as measuring the absorbance at 420 nm. Calculation. Read from the standard graph the number of µg boron equivalent to the absorbance of the sample, and the absorbance of the blank. Multiply the difference by 2 to give the mg l–1 of boron in the air-dry soil sample. To express in terms of oven-dry soil, see Method 5.2, Calculation (2).

Method 5.2. Cation exchange capacity, exchangeable bases and base saturation

Definitions Percentage base saturation: Cation exchange capacity (CEC):

= 100 × TEB/CEC7 the sum total of exchangeable cations that a soil can adsorb CEC7 or CEC-7: the CEC determined with 1 M ammonium ethanoate (ammonium acetate) buffered at pH 7.0 Effective cation exchange capacity (ECEC): the sum of the exchangeable cations (Al3+, H+, Ca2+ and Mg2+) extracted by 1 M potassium chloride Total exchangeable bases (TEB): the sum of the exchangeable ‘basic’ cations (Ca2+, Mg2+, K+, Na+ and NH4+) acted with 1 M ammonium ethanoate at pH 7.0.

Chapter 5

Discussion. The colloidal clay and humus soil fractions are negatively charged and therefore attract and adsorb positive ions (cations) on to exchange sites. These may be the so-called basic cations defined above, or the acidic cations H+ and Al3+. These cations are not soluble in water when in the adsorbed state, but can exchange with H+ which is present in the acidic vicinity of the plant root system. They are now in solution and able to be absorbed into the plant. The extent to which the exchange sites are saturated with cations, together with the ratios of the cations to each other, indicates the nutrient supplying power of the soil. The principle behind the determination of the CEC is that ammonium ions will leach the adsorbed metallic cations from the soil (soil ammonium having a ratio small enough to be ignored in this group of calculations) as a solution suitable for analysis by flame emission and atomic absorption techniques. The reagent M ammonium ethanoate is universally adopted for this purpose. The presence of any free basic cations as salts in solution would give an exaggerated TEB value, therefore some workers suggest an initial leaching with aqueous ethanol. This may be 95% ethanol, or more economically for class work, 95% or even 60% industrial methylated spirits (IMS), which is also used to remove excess ammonium ethanoate. The initial leaching is not usually necessary for temperate (UK) soils. In certain cases, ethanol may remove some adsorbed NH4+, and should be replaced with isopropanol. The amount of ammonium ion adsorbed on to all the exchange sites is a measure of the CEC. It may be determined either by leaching with acidified KCl (100 g l–1 KCl + 2.5 ml M HCl) to remove the ammonium ions, then a 25 ml aliquot of this solution is made alkaline to convert NH4+ to NH3 which is steam-distilled over and titrated, or the entire soil sample may be steam-distilled. The latter method has two disadvantages: it is difficult to transfer the entire soil sample to the distillation flask, and any non-exchangeable ammonium in the sample could be liberated to give an inflated CEC value. The TEB value may be obtained by either the sum of the individually measured cations or by evaporating and igniting a portion of the ammonium ethanoate leachate to convert the metallic cations to oxides and carbonates, followed by addition of excess acid (to convert carbonates to chlorides) and back-titration with alkali. The latter method is difficult if the soil is insufficiently base-rich to provide an adequate amount of bases for the titration. On the other hand, the calcium carbonate in calcareous soils may be partially leached by the ammonium ethanoate at pH 7.0 in addition to the exchangeable bases and thus give an exaggerated TEB value and a percentage base saturation in excess of 100%. The TEB by ignition/titration can serve as a check on the values from the summation method. If the percentage base saturation as defined above is 60%, this provides an indication of the need for estimation of exchangeable aluminium and hydrogen, in addition to calcium and magnesium, by the ECEC procedure. It must be strongly emphasized that the charge on the humus and mineral particles depends not only on the nature of the surface but on the pH, the negative charge, hence CEC, rising with increase in pH. The CEC7 can therefore be far higher than it would be in the field. It is therefore necessary to

Analysis of Soil and Compost

ascertain the experimental conditions when assessing published data. One way to compensate for this effect is to carry out the leaching with unbuffered 1 M KCl solution (some methods use unbuffered NH4Cl), which will not affect the in situ pH of the soil. Obviously this precludes subsequent analysis of potassium, but this is one of the minor cations. This analysis is termed the effective cation exchange capacity (ECEC). The exchange complex in acidic soils tends to be dominated by Al3+ rather than H+ according to the reaction: 3H+ + Al(OH)3.3H2O s Al3+ + 6H2O

The solubilized aluminium is the main toxic agent to plants in acidic soils, and acid tolerant (calcifuge) plants are usually also aluminium tolerant. The ECEC method determines the levels of Al3+, H+, Ca2+ and Mg2+ extracted by 1 M potassium chloride and is described in Method 5.2. A detailed discussion of the above topics together with a selection of class projects and test calculations is given in Chapter 7 of Rowell (1994, pp. 131–152), but note his calculation of CEC has an error in that it is based on 250 ml KCl extract, not on 100 ml as per given methodology, and thus requires correction (confirmed by personal communication, 2001). Directions for using the Foss/Perstorp Analytical Tecator Kjeltec Auto 1035/1038 Sample System for CEC determinations are given in USDA, 1996 (pp. 203–210). Determination of CEC and exchangeable cations Reagents. Note: deionized water and analytical grade chemicals are used throughout unless otherwise stated. • Ammonium ethanoate, M – dilute approximately 230 ml glacial ethanoic (acetic) acid to 1 l. Dilute approximately 220 ml ammonia solution (ammonium hydroxide) approximately 35% m/m NH3 to 1 l in a fume cupboard. Mix together in a 5-l graduated beaker and adjust the pH to 7.0 using ethanoic acid or ammonia solution added using a disposable polyethylene pasteur pipette. Stir with a glass rod between additions, but allow solution to become still before reading the pH. Dilute to 4 l and transfer to a polythene storage bottle. • Ethanol, 95% (or industrial methylated spirits, 95%) – dilute ethanol (or IMS) to give 95% v/v ethanol/water. • Potassium chloride solution – dissolve approximately 100 g KCl in water and make up to 1 l. Add 2.5 ml M HCl, and check the pH is approximately 2.5. Extraction. Transfer 5 g sieved ( 2 mm) air-dried soil to a 100-ml glass beaker, add 20 ml M ammonium ethanoate, stir and let stand overnight. Transfer the contents to a filter funnel fitted with a 125 mm Whatman No. 44 filter paper and held in a 250 ml volumetric (graduated) flask. Wash the beaker with ammonium ethanoate reagent from a wash (squeeze) bottle to remove all the sample, then add successive 25 ml volumes of reagent to leach the soil in the funnel, allowing it to drain between additions. With the collected leachate volume approaching 250 ml, remove the funnel to a rack or

Chapter 5

place in a 250 or 500 ml conical flask, and make up the volume to the 250 ml mark with reagent and retain for analysis of exchangeable bases. The soil in the funnel is washed free of excess reagent by five successive additions of 95% ethanol, allowing to drain between washings. A wash bottle containing ethanol enables the interior surface of the funnel, the outside of the stem, the exposed surface of the paper and the soil to be thoroughly washed. Any remaining ammonium ethanoate will elevate the final CEC value. The washings, which are flammable, should be collected in a waste solvents bottle for safe disposal. The funnel is now placed in a 100-ml volumetric flask and leached with successive 25-ml portions of potassium chloride solution, allowing draining between additions, until nearly 100 ml has been collected. Make up to the mark and retain for determination of CEC. Measurement of calcium and magnesium by AAS This is achieved by using an atomic absorption spectrophotometer (or less accurately with a flame photometer). Some details could be instrument specific, so refer to the manufacturer’s handbook, application data sheets, and obtain technical support if you lack experience in this area. Some general guidelines will be noted here. The use of a nitrous oxide-acetylene flame obviates the need for releasing agents to be added to samples and standards, but may be hazardous to use. It also requires addition of a reagent of an easily ionized compound, such as potassium, to be added to suppress ionization. It is suggested that an air-acetylene flame is more appropriate for routine use. Releasing agents are chemicals which protect the analyte atoms in the flame from forming compounds with other molecular or ionic species, which will depress the absorption in an erratic manner. Either strontium or lanthanum salts are used for this purpose. It is essential that all standard solutions are made up in the same reagent as the samples. This ensures that they not only have any impurities introduced by the reagent solution, but that they have the same viscosity (which can greatly affect the rate of aspiration by the nebulizer) and exert the same interference effect in the flame. A blank solution should always be included, and a control obtained from a bulk sample is good practice for any analysis, and enables one to detect if a systematic error or instrument malfunction should arise. The sample solutions will often require dilution to suit the sensitivity of the particular instrument, however, the sensitivity may be able to be reduced either electronically or by rotation of the burner, and so avoid this extra step. If the standard curve begins to level out towards the horizontal, the flame is probably becoming saturated with the analyte, and dilution is essential. Wavelengths for AAS. Calcium is measured at 422.7 nm and magnesium at 285.21 nm.

Analysis of Soil and Compost

Reagents. • Calcium stock solution, 1000 µg Ca2+ ml–1 – stock solutions of many elements for determination by AAS are available commercially. Details for inhouse preparation will also be given. Anhydrous calcium nitrate, Ca(NO3)2, is dried for 1 h at 105°C, then cooled in a desiccator. Transfer 2.05 g to a 100-ml beaker containing water and stir to dissolve. Immediately add 1 ml HCl (36% m/m) to prevent hydrolysis, add with washings to a 500 ml volumetric flask, make up to the mark with water, and mix by shaking. • Calcium standards, 50 and 0–5.0 µg Ca2+ ml–1 – pipette 25 ml stock solution into a 500-ml volumetric flask, make up to the mark with M ammonium ethanoate reagent and mix to give a solution of 50 µg Ca2+ ml–1. Pipette 0, 0.5, 1, 2.5, 5 and 10 ml of this solution into 100 ml volumetric flasks and make up to the mark with ammonium ethanoate reagent. Standard values are 0, 0.25, 0.5, 1.0, 2.5 and 5.0 µg Ca2+ ml–1. • Releasing agent – dissolve 2.68 g lanthanum chloride heptahydrate (LaCl3.7H2O) in water and make up to 100 ml. • Magnesium stock solution, 1000 µg Mg2+ ml–1 – dissolve 1.6581 g magnesium oxide (previously dried at 105°C overnight and cooled in a desiccator) in the minimum of hydrochloric acid (approximately 5 M). Dilute with water to 1 l in a volumetric flask to obtain a solution of 1000 µg Mg2+ ml–1. • Magnesium standards, 10 and 0–1 µg Mg2+ ml–1 – pipette 5 ml stock solution into a 500-ml volumetric flask and dilute to the mark with M ammonium ethanoate reagent to obtain a stock solution of 10 µg Mg2+ ml–1. Pipette 0, 2, 4, 6, 8 and 10 ml of the 10 µg Mg2+ ml–1 stock solution into 100 volumetric flasks and make up to the mark with M ammonium ethanoate reagent and mix. This will give solutions containing 0, 0.2, 0.4, 0.6, 0.8 and 1.0 µg Mg2+ ml–1. Analysis of solutions. Pipette 20 ml of sample and standard solutions into 50-ml beakers, then pipette 1 ml releasing agent solution into each beaker and mix. If readings are off-scale, pipette 5 ml extract plus 15 ml M ammonium ethanoate and the 1 ml releasing agent and retest. Whatever dilution is necessary, ensure the sample plus M ammonium ethanoate solution add up to 20 ml before addition of the 1 ml releasing agent. Measurement of potassium and sodium by flame photometry These elements are best determined using flame photometry, as their high atomic emission energy in the flame exceeds their absorption of energy, which results in a higher sensitivity than with atomic absorption spectrophotometry. Check that the appropriate filter is in place for the element being determined, ignite the air–propane flame and ensure an adequate warm-up time. Aspirate the blank solution and adjust the reading to zero. Aspirate the highest standard to allow sensitivity adjustment to give an emission of about 90% maximum reading, and then re-check the zero with the blank. Ensure the standard curve is reasonably linear, then proceed to analyse the samples. Repeat the standards

Chapter 5

at about 10-min intervals to permit correction for any changes in sensitivity. A quality control sample may be analysed at intervals of about 48 samples. Reagents. • Potassium stock solution, 1000 µg K+ ml–1 – weigh 1.293 g potassium nitrate (previously dried for 1 h at 105°C and cooled in a desiccator) into a 100-ml beaker. Dissolve in water, add 1 ml hydrochloric acid (approximately 36% m/m HCl) and 1 drop of toluene, then transfer with washings to a 500-ml volumetric flask, make up to the mark and mix well by shaking. • Potassium standard solutions, 100 and 0–10 µg K+ ml–1 – pipette 10 ml of the stock solution into a 100-ml volumetric flask and dilute with M ammonium ethanoate reagent to the mark and mix to give a solution of 100 µg K+ ml–1. Pipette 0, 2, 4, 6, 8 and 10 ml of this solution into 100-ml volumetric flasks and dilute to the mark with M ammonium ethanoate reagent and mix. These will contain 0, 2, 4, 6, 8 and 10 µg K+ ml–1. • Sodium stock solution, 1000 µg Na+ ml–1 – weigh 2.542 g sodium chloride (previously dried for 1 h at 105°C and cooled in a desiccator) into a 100-ml beaker. Dissolve in water, add 1 ml hydrochloric acid (approximately 36% m/m HCl) and 1 drop of toluene, then transfer with washings to a 1000-ml volumetric flask, make up to the mark and mix well by shaking. • Sodium standard solutions, 100 and 0–10 µg Na+ ml–1 – pipette 10 ml of the stock solution into a 100-ml volumetric flask and dilute with M ammonium ethanoate reagent to the mark and mix to give a solution of 100 µg Na+ ml–1. Pipette 0, 1, 2, 3, 4 and 5 ml of this solution into 100-ml volumetric flasks and dilute to the mark with water and mix. These will contain 0, 1, 2, 3, 4 and 5 µg Na+ ml–1. Calculation (1). Results have traditionally been expressed as milliequivalents per 100 g soil. An alternative more recent expression is centimole charge per kilogram soil (cmolc kg–1), but both expressions give the same numbers. The concentrations of cations using the above methods may be obtained by multiplying the concentration of cation (µg ml–1) in the sample extract solution (obtained by comparing sample readings with the standard curve) by the following factors (plus any dilution factors to bring readings on scale): Calcium, 0.249; magnesium, 0.412; potassium, 0.128; sodium, 0.2175 Explanation: If a reading of X µg K ml–1 is obtained for a solution of 5 g soil in 250 ml extractant (1 in 50 dilution), this amounts to X/(39.098 × 103) milliequivalents K ml–1, or 250X/(39.098 × 103) milliequivalents K in 250 ml extractant. This is derived from 5 g soil, thus 100 g soil would contain (20 × 250 × X)/(39.098 × 103) = 0.128 milliequivalents K. Thus if 2 g soil were taken instead of 5 g, an additional factor of × 5/2 should be used. If the sample solution for calcium determination was diluted 5 ml solution plus 15 ml M ammonium ethanoate reagent before addition of 1 ml releasing agent, then an additional factor of × 4 will be necessary.

Analysis of Soil and Compost

Determination of cation exchange capacity (CEC) The ammonium extracted by the potassium chloride reagent is analysed by steam distillation. This may be carried out using an automatic instrument such as the Kjeltec Auto 1035 Analyzer (USDA, 1996, pp. 203–210), or a micro (or semi-micro) steam distillation unit such as that described by Bremner and Keeney (1965), or the readily available Markham still. We will describe the manual procedure. Reagents. • Ammonium-N standard solution, 140 µg ml–1 nitrogen – weigh 0.661 g ammonium sulphate (dried at 105°C for 1 h and cooled in a desiccator) into a 100-ml beaker and dissolve in ammonia-free water (distil deionized water acidified with sulphuric acid), transfer with washings to a 1-l volumetric flask and make up to the mark with the ammonia-free water and mix. This should be stored in a refrigerator, but a quantity allowed to warm to room temperature in a stoppered container before use. • Boric acid solution, approximately 2% m/v – prepare fresh weekly. • Mixed indicator – dissolve 0.3 g methyl red and 0.2 g methylene blue in 250 ml ethanol. • Magnesium hydroxide suspension – heat magnesium oxide (heavy) for 2 h at 800°C. After cooling in a desiccator, make a suspension of 17 g in 100 ml water. • Octan-2-ol – antifoam agent: use 1 drop when flasks 7 as alkaline, and if the pH is 7, as neutral. Pure water in equilibrium with atmospheric CO2 has a pH of 5.6. If a soil pH is lower or higher than this, it is acting as an acid or base respectively. Several soil components act as buffers (hydroxy aluminium monomers or polymers, soil organic matter and undissolved carbonates), therefore lime requirement tests may also be required.

Chapter 5

Method 5.8a. Measurement of pH Reagents. • Buffer capsules/tablets – dissolve to make solutions of pH 4.0, 7.0 and 9.0. Procedure. Calibrate the pH meter according to the manufacturer’s instructions using buffers to cover the pH range of the soil samples. Transfer a 10ml scoop of sieved (2 mm) air-dry soil (struck off level without tapping) into a flat-bottom plastic vending cup, add 25 ml water and a magnetic PTFEencased stirrer bar. Place on a multi-position electronic stirrer unit (e.g. 15place Variomag) and stir for 15 min. Tilt the cup, if necessary, to ensure the pH electrode is sufficiently immersed (the soil suspension should reach the porous plug liquid junction on the side of a combination glass electrode). The electrode should not be abraded by the abrasive soil at the bottom of the cup. Swirl a couple of times and allow the drift in pH to stabilize (about 30 s) before taking the reading. Rinsing between samples is not necessary unless soils have widely differing pH values. Recalibrate the meter hourly. If required, retain the suspension for the determination of lime requirement.

Method 5.8b. Determination of lime requirement Definition. The lime requirement of a mineral soil is the number of tonnes of calcium carbonate calculated to raise the pH of a hectare of soil 200 mm deep, under field conditions, to, and maintain at, 6.5. A low pH indicates that lime is required, but not the quantity. Excess is not only wasteful, but may render certain elements (e.g. Fe, Mn, and B) unavailable to plants. There are several methods for determining the lime requirement, including adding excess calcium bicarbonate and back-titrating the excess; adding increasing amounts of calcium hydroxide and monitoring the pH; and the use of a buffer solution (MAFF/ADAS, 1986, pp. 150–151), which will be described below. Reagents. • Buffer solution, double strength – add to 4.5 l water, 400 g oven-dried calcium acetate, 80 g 4-nitrophenol and 6 g of light magnesium oxide. [Safety note: 4- (or para-) nitrophenol may cause eye irritation or irreversible eye injury, and may be harmful by absorption through skin or ingestion.] Heat the mixture to dissolve the solids and dilute to 5 l. Filter if the solution is not clear. The pH of this solution should lie between 6.9 and 7.1; adjust by the addition of hydrochloric acid or magnesium oxide as necessary. • Buffer solution, single strength – add 1 volume of double strength buffer solution to 1 volume of water and mix. The pH of this solution should lie between 6.9 and 7.1; if necessary, adjust as above. Procedure for mineral soils of pH 5.0 to 6.4 inclusive. Add 20 ml of single strength buffer solution to the soil suspension retained from the pH determi-

Analysis of Soil and Compost

nation, and stir for 5 min. Mix 25 ml water with 20 ml single strength buffer with pH adjusted as above and recalibrate the pH meter to indicate pH 7.00 using this solution. Measure the pH of the stirred sample. If the pH is < 6.0, again stir a fresh sample of 10 ml soil with 25 ml water for 15 min and proceed as for mineral soils of pH < 5.0 as described below. Calculation for mineral soils of pH 5.0–6.4. Subtract the indicated pH from 7.00 and multiply by 11.2. The result gives the lime requirement as tonnes ha–1 calcium carbonate. Procedure for mineral soils of pH less than 5.0. Add 20 ml of double strength buffer solution to the soil suspension retained from the pH determination, and stir for 5 min. Mix 25 ml water and 20 ml double strength buffer with pH adjusted to between 6.9 and 7.1, and use to recalibrate the pH meter to read 7.00. Read the pH of the stirred sample. Calculation for mineral soils of pH less than 5.0. Subtract the indicated pH from 7.00 and multiply by 22.4 to obtain the lime requirement expressed as tonnes ha–1 calcium carbonate. Notes on the calculation. A full explanation of soil buffer capacity and derivation of the above factors is given by Rowell (1994, pp. 171–172).

Method 5.8c. Determination of pH in soils with soluble salts See the discussion under ‘pH extractants’ in Chapter 4. Reagents. • Calcium chloride, 1.0 M – completely dissolve 14.7 g CaCl2.2H2O in water and make up to 100 ml. • Calcium chloride, 0.01 M – completely dissolve 1.47 g CaCl2.2H2O in water and make up to 1 l. Procedure. Proceed as in Method 5.8a., replacing water with a solution of 0.01 M CaCl2. Alternatively, add 5 drops (0.25 ml) of 1 M CaCl2 to the suspension following the pH determination in Method 5.8a. Calculation. The difference in pH between water and salt solution extracts is known as the salt effect, and given the symbol ∆ pH. Thus, ∆ pH = soil pH in CaCl2 solution – soil pH in water

and ∆ pH values are positive for soils with a net positive charge, and negative for soils with a net negative charge, with magnitude proportional to charge.

Chapter 5

Discussion 5.9. Determination of extractable phosphorus See the discussion under ‘Phosphate extractants’ in Chapter 4 to determine the most appropriate extractant.

Method 5.9a. Determination of extractable phosphorus (manual method) The following manual procedure is based on MAFF/ADAS, 1986, pp. 183–185 (with Crown Copyright Permission). Phosphorus is extracted from soil at 20 ±1°C with a solution of sodium bicarbonate at pH 8.5. The absorbance of the molybdenum blue complex produced by the reduction with ascorbic acid of the phosphom*olybdate formed when acid ammonium molybdate reacts with phosphate is measured using a spectrophotometer at 880 nm. Reagents (extraction). • Polyacrylamide solution, 0.05% m/v – dissolve 0.5 g of polyacrylamide in approximately 600 ml of water by stirring for several hours. When dissolved, dilute to 1 l. • Sodium bicarbonate reagent – dissolve 420 g sodium bicarbonate (sodium hydrogen carbonate) in water, add 50 ml of the polyacrylamide solution and dilute to 10 l. Add approximately 50% m/m sodium hydroxide solution, stirring with a glass rod, until the pH meter reading is steady at 8.50 at 20°C (a plastic Pasteur pipette is useful for dropwise addition approaching the required pH). Procedure (extraction). Transfer 5 ml (scoop filled and struck off level without tapping) of air-dry soil, sieved to 2 mm into a bottle (e.g. wide-mouth, square HDPE). Add 100 ml of sodium bicarbonate reagent, pH 8.50, cap the bottle and shake on a reciprocating shaker, at approximately 275 strokes of 25 mm length per minute, for 30 min at 20°C. Filter a portion immediately through a Whatman No. 2 filter paper, rejecting the first few millilitres of filtrate. Carry out a blank determination. Reagents (determination). • Ammonium molybdate reagent, 1.2% m/v – dissolve 24 g powdered ammonium molybdate (ammonium paramolybdate, (NH4)6Mo7O24.4H2O), and 0.6 g antimony potassium tartrate (Note: cumulative poison) in 1200 ml water. Slowly add 296 ml sulphuric acid (approximately 98% m/m H2SO4), stir slowly with a glass rod, and dilute to 2 l. (Note: sulphuric acid is highly corrosive and generates heat when diluted; standing the beaker in a sink with a few centimetres depth of cold water before adding the acid will reduce any likelihood of localized boiling. Wear PPE for this step.) Store in a dark glass bottle in a refrigerator. • Ammonium molybdate reagent, 0.15% m/v – dilute 1 vol. of 1.2% m/v

Analysis of Soil and Compost

• •

•

ammonium molybdate reagent to 8 vol. with water. Store in a dark glass bottle in a refrigerator. Ascorbic acid solution, 1.5% m/v – prepare immediately before use, allowing 5 ml per standard, blank and sample, with some spare for any repeats. Phosphorus stock standard solution, 1 mg P ml–1 – dry potassium dihydrogen orthophosphate at 102°C for 1 h and cool in a desiccator. Dissolve 0.879 g of the dried salt in water and add 1 ml of hydrochloric acid, approximately 36% m/m HCl. Dilute to 200 ml and add 1 drop of toluene to the solution. Phosphorus intermediate standard solution, 20 µg ml–1 – pipette 10 ml of the phosphorus stock standard solution, 1 mg ml–1, into a 500-ml volumetric flask, make up to the mark and mix. Add 1 drop of toluene to the solution. Phosphorus working standard solutions, 0–7 µg P ml–1 – prepare fresh daily solutions by pipetting 0, 5, 10, 15, 20, 25, 30 and 35 ml of the phosphorus intermediate standard solution, 20 µg ml–1, into 100-ml volumetric flasks, make up to the mark with sodium bicarbonate reagent, and mix. These will contain 0, 1, 2, 3, 4, 5, 6 and 7 µg P ml–1 respectively. Sulphuric acid, approximately 1.5 M – slowly with stirring, add 80 ml sulphuric acid, approximately 98% m/m H2SO4, to about 800 ml water in a 2-l beaker (Note: sulphuric acid is highly corrosive and generates heat when diluted; standing the beaker in a sink with a few centimetres of cold water before adding the acid will reduce any likelihood of localized boiling. Wear PPE for this step.) Cool and dilute to 1 l.

Procedure. Pipette 5 ml of each phosphorus working standard solution (i.e. 0, 5, 10, 15, 20, 25, 30 and 35 µg P) into a plastic vending cup (or 100-ml conical flask). Add 1 ml of approximately 1.5 M sulphuric acid and swirl the solution to assist the release of CO2. Add 20 ml of 0.15% m/v ammonium molybdate reagent, 5 ml of ascorbic acid solution, 1.5% m/v, swirl to mix and allow to stand for 30 min for colour development. Measure the absorbance in a 10 mm optical cell at 880 nm. The colour is stable for several hours. Construct a graph relating absorbance to µg P. The absorbance values should be approximately 0 to 0.8 for the 0 and 35 µg P standards respectively. Similarly pipette 5 ml of the soil extract into a plastic vending cup, followed by 5 ml sulphuric acid, 1.5 M. If the soil extract solution is highly coloured after addition of the acid, pipette a duplicate sample and add the 5 ml of sulphuric acid, 1.5 M. Add the other reagents, as detailed above, to the first sample, but only add the ammonium molybdate reagent to the duplicate. Measure the absorbance at 880 nm. For coloured extracts, subtract the absorbance of the duplicate without ascorbic acid, which will not develop the blue colour, from the absorbance of the sample extract with ascorbic acid. Calculation. Read from the standard graph the number of µg of P equivalent to the absorbances of the sample and blank determinations. Subtract the blank from the sample value, and multiply the difference by 4. The result gives the

Chapter 5

mg l–1 extractable phosphorus in the air-dry soil. This can be expressed for oven-dry soil as in Method 5.2, Calculation (2). Notes: 1. If one of the standards produces an absorbance that lies significantly away from the standard graph produced by the other standards, or if the whole graph is erratic, repeat as necessary. Detergents containing phosphates should be avoided, but ones such as Decon 90 are phosphate-free. 2. The use of phosphate-free carbon to decolorize soil extracts has been found to give erratic results. 3. A slightly less accurate determination is possible using a colorimeter with a wide bandpass filter, e.g. a simple (non-interference type) purple-red filter but an interference filter in a quality instrument gives results comparable to a spectrophotometer.

Method 5.9b. Determination of extractable phosphorus (automated method) An automated method for the Lachat QuikChem Automatic Flow Injection Ion Analyzer is given in Missouri Agricultural Experiment Station (1998), pp. 27–29, and is available free of charge from Lachat (Lachat Instruments, 1988). Sun et al. (1981) describe a method for the Tecator FIAstar® flow injection system. A method for a segmented continuous flow procedure for both phosphate and potassium was devised by Armitage (1965). The parameters for the phosphate analysis using a dilute HCl soil extractant are outlined below. The manifold diagram (Fig. 5.3) has been modified to allow for the fact that Armitage later changed the Sampler I to a Sampler II module. -1 Sampler 40 h Pump

To waste Heating bath

0.60 ml min 1.60 ml min 2.50 ml min 0.80 ml min

Debubbler Colorimeter 660 nm

To sampler wash

3.40 ml min 1.20 ml min 1.60 ml min

To chart recorder

-1 -1 -1 -1 -1 -1 -1

sample ascorbic acid molybdate air 0.3 M HCl wash waste flowcell waste

Fig. 5.3. Manifold for the automated determination of phosphorus in soil extracts.

Analysis of Soil and Compost

Reagents. • Hellige–Truog extractant, 0.3 M HCl – dilute 25.8 ml hydrochloric acid, 36% m/m, to 1 l. • Ammonium molybdate reagent – dissolve 4 g powdered ammonium molybdate in 1 l of water. Slowly with stirring add 80 ml sulphuric acid, 98% m/m H2SO4, and dilute to 2 l. • Ascorbic acid, 0.1% (m/v) – prepare fresh daily. Procedure. The above reagents, together with the segmenting stream of air, are mixed in a double mixing coil, with the sample being introduced halfway along the coil. A further single mixing coil provides a final mixing before the solution passes through a single glass coil in a thermostatically controlled heating bath at 95°C. After development of the colour, the solution passes to a debubbler, whence it proceeds to the colorimeter fitted with a 15-mm flowcell, and a pair of 660-nm interference filters. The output is to a chart recorder or personal computer. Standards and blanks should be prepared in the extractant solution according to the normal protocol. It is expected that a sampling rate of 40 h–1, with a time ratio of 2:1 sample:wash would be suitable. This may need adjusting with soils of widely differing extractable P concentrations, where large peaks may obscure very small ones. Notes: 1. A pump tube has been included for the wash waste for reservoirs that do not have a gravity overflow. 2. The pump tubes in some published manifold diagrams are labelled with the internal diameters of the tubes. We will use the more common convention of labelling with flow rates in ml min–1. 3. Pump tubes are designated by a colour code, which may also be used in the literature. Some common colour codes are listed in Table 5.1. 4. Flow rates are for standard PVC tubing. It is good practice to avoid using the smallest or largest sizes whenever possible. 5. Pump tubes are available in different materials depending on the liquid to be pumped, e.g. for solvents or concentrated acids. Flow rates for these other materials will be different than for standard PVC. Pump tubes may also be available in a specially calibrated or measured flow rate quality at extra cost. Unless specified for medical purposes or to meet regulations, the standard quality is normally adequate. See Chapter 1, ‘Peristaltic pumps’.

Method 5.9c. Determination of resin extractable phosphorus (automated method) The extraction method of Hislop and Cooke (1968), has been outlined in Chapter 4, ‘Phosphate extractants’. A blank determination without soil should be carried out. The autoanalysis manifold is shown in Fig. 5.4. Some adjustments to dilution and/or readout sensitivity may be necessary to handle both

Chapter 5

Table 5.1. Shoulder colour code for peristaltic pump tubing. Colour code Orange red Orange blue Orange green Orange yellow Orange white Black Orange White Red Grey Yellow Yellow blue Blue Green Purple Purple black Purple orange Purple white

Delivery (ml min–1) 0.03 0.05 0.10 0.16 0.23 0.32 0.42 0.60 0.80 1.00 1.20 1.40 1.60 2.00 2.50 2.90 3.40 3.90

agricultural (lower P) and glasshouse (higher P) soils. The above authors referred to P2O5, but we have converted values to P. Reagents. • Ascorbic acid, 1% (m/v) – prepare fresh daily. • Ammonium molybdate – sulphuric acid stock reagent – dissolve 10 g powdered ammonium molybdate in approximately 70 ml water and dilute to 100 ml. Carefully add 150 ml sulphuric acid, 98% m/m H2SO4, to 150 ml water in a 600/800 ml beaker while stirring with a glass rod, and allow to cool. Add the molybdate solution with careful stirring and allow to cool. • Ammonium molybdate – sulphuric acid autoanalysis reagent – dilute 100 ml of the ammonium molybdate – sulphuric acid stock reagent to 1 l with water and mix. • Phosphorus stock standard solution, 1 mg P ml–1 – Dry potassium dihydrogen orthophosphate at 102°C for 1 h and cool in a desiccator. Dissolve 0.879 g of the dried salt in water and add 1 ml of hydrochloric acid, approximately 36% m/m HCl. Dilute to 200 ml and add 1 drop of toluene to the solution. • Phosphorus intermediate standard solution, 100 µg ml–1 – pipette 50 ml of the phosphorus stock standard solution, 1 mg ml–1, into a 500-ml volumetric flask, make up to the mark with sodium sulphate extractant and mix. Add 1 drop of toluene to the solution. • Phosphorus working standard solutions, 0–35 µg P ml–1 – prepare fresh daily solutions by pipetting 0, 5, 10, 15, 20, 25, 30 and 35 ml of the phos-

Analysis of Soil and Compost

Sampler 40 h-1 To sampler wash Heating Bath

Pump

To waste

2.50 ml min-1 molybdate

95oC DMC Debubbler Colorimeter 625 nm

1.20 ml min-1 air

2.00 ml min-1 sample 0.32 ml min-1 ascorbic acid

To wash

2.50 ml min-1 sod. sulphate 2.90 ml min-1 wash waste 3.40 ml min-1 flowcell waste

To chart recorder Fig. 5.4. Manifold for the automated determination of phosphorus in soil resin extracts.

phorus intermediate standard solution, 100 µg ml–1, into 100-ml volumetric flasks, make up to the mark with sodium sulphate extractant, and mix. These will contain 0, 5, 10, 15, 20, 25, 30 and 35 µg P ml–1 respectively, and are suitable for glasshouse soils that are approximately ×8 higher in P than agricultural soils; a lower range of 0, 1, 2, 3, 4, and 5 µg P ml–1 should be prepared for the latter. • Sodium sulphate extractant/wash, 7% (m/v). Calculation. The 2-ml scoop of soil was extracted via resin into 50 ml sodium sulphate extractant; therefore the concentration must be multiplied by 25 to give the µg P ml–1 in soil by resin extraction. Hislop and Cooke (1968) classified the soils with respect to mg P l–1 air-dry soil as follows: • agricultural soils: low, 65 • glasshouse soils: low, 436. ADAS have indexed resin P values as follows: 0, 0–19; 1, 20–30; 2, 31–49; 3, 50–85; 4, 86–132, 5, >132 mg P l–1.

Method 5.10. Determination of extractable magnesium, potassium and sodium Magnesium, potassium and sodium are extracted from the soil with 1 M ammonium nitrate.

Chapter 5

Reagent (extraction). • Ammonium nitrate, 1 M – dissolve 400 g of ammonium nitrate in water and make up to 5 l. Procedure (extraction). Transfer 10 ml (scoop filled and struck off level without tapping) of air-dry soil, sieved to 2 mm, into a bottle (e.g. wide-mouth, square HDPE), and shake on a reciprocating shaker (approximately 275 strokes of 25 mm per min) for 30 min. Filter through a Whatman No. 2 filter paper, discard the first few millilitres, and retain the rest for analysis of the required elements. Carry out a blank determination. Reagents (determination). • Releasing agent – dissolve 13.4 g lanthanum chloride heptahydrate (LaCl3. 7H2O) in water and make up to 500 ml. • Magnesium stock standard solution, 1000 µg Mg2+ ml–1 – dissolve 1.6581 g magnesium oxide (previously dried at 105°C overnight and cooled in a desiccator) in the minimum of hydrochloric acid (approximately 5 M). Dilute with water to 1 l in a volumetric flask to obtain a solution of 1000 µg Mg2+ ml–1. • Magnesium standards, 10 and 0–1 µg Mg2+ ml–1– pipette 5 ml stock solution into a 500-ml volumetric flask and dilute to the mark with M ammonium nitrate reagent to obtain a stock solution of 10 µg Mg2+ ml–1. Pipette 0, 2, 4, 6, 8 and 10 ml of the 10 µg Mg2+ ml–1 stock solution into 100-ml volumetric flasks, add 5 ml releasing agent and make up to the mark with 1 M ammonium nitrate reagent and mix. This will give solutions containing 0, 0.2, 0.4, 0.6, 0.8 and 1.0 µg Mg2+ ml–1. • Potassium stock standard solution, 1 mg ml–1 of potassium – dry potassium nitrate at 102°C for 1 h and cool in a desiccator. Dissolve 1.293 g of the dried salt in water and add 1 ml of hydrochloric acid (approximately 36% m/m HCl). Dilute to 500 ml and add 1 drop of toluene. • Potassium working standard solutions, 0–50 µg ml–1 of potassium. Pipette 0, 1, 2, 3, 4 and 5 ml of the potassium stock standard solution into 100ml volumetric flasks, dilute to the mark with 1 M ammonium nitrate solution and mix. These will contain 0, 10, 20, 30, 40 and 50 µg K ml–1. • Sodium stock standard solution, 1 mg ml–1 of sodium – dry sodium chloride at 105°C for 1 h and cool in a desiccator. Dissolve 0.254 g in water, make up to 100 ml and mix. • Sodium intermediate standard solution, 20 µg ml–1 of sodium – pipette 10 ml of the sodium stock standard solution into a 500-ml volumetric flask, make up to the mark with 1 M ammonium nitrate reagent and mix. • Sodium working standard solutions, 0–2 µg ml–1 of sodium – pipette 0, 2, 4, 6, 8 and 10 ml of the sodium intermediate standard solution into a 100-ml volumetric flask, make up to the mark with 1 M ammonium nitrate solution and mix. These will contain 0, 0.4, 0.8, 1.2, 1.6 and 2.0 µg Na+ ml–1.

Analysis of Soil and Compost

Procedure (determination). Magnesium is determined by atomic absorption spectrophotometry (see Method 5.2, ‘Measurement of calcium and magnesium by AAS’). Potassium and sodium are determined by flame photometry (see Method 5.2 ‘Measurement of potassium and sodium by flame photometry’). Analyse the standards and adjust the zero and maximum standard readings in the usual way. • Magnesium: pipette 2 ml sample solution into a 100-ml volumetric flask, add 5 ml releasing agent, make up to the mark with 1 M ammonium nitrate and mix. Nebulize into the AAS and record the readings (computer, chart recorder or manually, as appropriate). • Potassium: nebulize the extract without further dilution. • Sodium: pipette 10 ml extract into a 100-ml volumetric flask, make up to the mark with 1 M ammonium nitrate reagent and mix. Calculations. 1. Magnesium. From the standard graph determine the number of µg ml–1 of magnesium in the sample, subtract the blank value and multiply the difference by 250 (initial extraction ratio of 5 multiplied by subsequent 50 dilution of the extract solution). The result is the number of mg l–1 extractable magnesium in the air-dry soil. Include any extra dilution factors, and, if required, convert to oven-dry soil using the appropriate factor, as in Method 5.2, Calculation (2). 2. Potassium. From the standard graph determine the number of µg ml–1 of potassium in the sample, subtract the blank value and multiply the difference by 5 (initial extraction ratio). The result is the number of mg l–1 extractable potassium in the air-dry soil. Include any extra dilution factors, and, if required, convert to oven-dry soil using the appropriate factor, as in Method 5.2, Calculation (2). 3. Sodium. From the standard graph determine the number of µg ml–1 of sodium in the sample, subtract the blank value and multiply the difference by 50 (initial extraction ratio of ×5 multiplied by subsequent ×10 dilution of the extract solution). The result is the number of mg l–1 extractable sodium in the air-dry soil. Include any extra dilution factors, and, if required, convert to oven-dry soil using the appropriate factor, as in Method 5.2, Calculation (2).

Method 5.11. Determination of extractable trace elements For a discussion on the nature of the extractants, see Chapter 4 ‘Trace element extractants’. The method described below will use the complexing reagent DTPA (diethylene-triaminepentaacetic acid) to extract, by chelation, copper, iron, manganese and zinc (including zinc on calcareous soils); it also shows promise for monitoring cadmium, nickel and lead in soils receiving sludge appli-

Chapter 5

cations. The amount of nutrient extracted will vary with extractant pH and concentration, shaking time and temperature, and soil:solution ratio. Keeping these parameters constant will enable valid comparisons with subsequent experiments or advisory tests. The stock standard solutions may be purchased ready-made for AAS. If made in-house, then the appropriate spectroscopically pure metals, oxides or non-hydrated salts should be used, and oven-dried at 102–105°C for 1 h before weighing. To avoid significant weighing errors, at least 0.2 g of substance should be weighed. Metals and oxides should be dissolved in spectroscopically pure grade acids. At the lower wavelengths used for some of these micronutrients (250 nm), background absorption from molecular flame species, such as CaO, arising from compounds in the soil extracts can have an interfering effect and cause an elevation in the observed absorption. Some AAS instruments have a background correction facility (e.g. by using the Zeeman effect), and this should be used. An approximate assessment of this effect can be achieved by measuring the absorption with a spectral line close to the one being used, but one not showing an absorption for a dilute solution of that particular element (Slavin, 1968; Christian and Feldman, 1970), while keeping the sensitivity of the instrument the same. Another approach is to make up the standards in a matrix of approximately the same levels of soluble salts as found in the soil extracts. Background interference can be more troublesome with electrothermal than with flame atomizers (Fuller, 1977). Reagents. • DTPA extractant – dissolve 3.933 g DTPA in a mixture of 29.844 g TEA (triethanolamine) and 22.22 ml water; stir until dissolved. Add 2.944 g calcium chloride (CaCl2.2H2O) to 1.1 l of water, and when dissolved, add to the DTPA/TEA solution and make up to about 1.9 l with water. Adjust the pH to 7.3 using hydrochloric acid (approximately 36% m/m HCl) and make up to 2 l. • Releasing agent – dissolve 2.68 g lanthanum chloride heptahydrate (LaCl3.7H2O) in water and make up to 100 ml. • Stock standard solutions, 1 mg ml–1 of the metal – purchase or make up as appropriate. • Working standard solutions – dilute 5 ml of the stock standard solutions to 500 ml with DTPA extractant to give intermediate standards of 10 µg ml–1 of the metal. Prepare a range of standards in DTPA extractant for each metal. Suggested values are: cadmium, 0, 0.1, 0.2, 0.3, 0.4 and 0.5 µg Cd ml–1; copper, lead or manganese, 0, 1.0, 2.0, 4.0, 6.0 and 8.0 µg Cu, Pb or Mn ml–1; iron, 0, 2.0, 5.0, 10.0, 15.0 and 20.0 µg Fe ml–1; nickel, 0, 0.2, 0.5, 1.0, 2.0, 3.0 µg Ni ml–1; zinc, 0, 0.5, 1.0, 2.0, 3.0 and 4.0 µg Zn ml–1. Procedure. Weigh 10 g air-dry soil, sieved to 2 mm (10 mesh) using a stainless steel sieve into a 175-ml square HDPE (e.g. Nalgene) plastic screw-cap

Analysis of Soil and Compost

bottle. Add 20 ml of the DTPA extractant and shake on a reciprocating shaker (275 oscillations of 25 mm length per minute, or similar, but keep constant for all extractions) for 2 h. Carry out a blank extraction. Filter through a Whatman No. 42 filter paper, rejecting the first couple of millilitres, into a polythene hinged-cap sample tube. Pipette 10 ml of filtrate and standards into 25-ml beakers and add 0.5 ml releasing agent to each and mix. They are analysed for the required trace elements by atomic absorption spectroscopy using a suitable range of standards made up in the DTPA extractant. Samples may be diluted with DTPA extractant to reduce excessively high readings to the normal range of the instrument. The wavelengths (nm) of the most sensitive resonance lines for AAS are as follows: Cd, 228.8; Cu, 324.8; Fe, 248.3; Pb, 217.0; Mn, 279.5; Ni, 232.0 and Zn, 213.9. Calculation. The concentration of trace element (µg ml–1) in the extract is read from the standard curve and the blank reading subtracted; the difference is multiplied by 2 to give the µg g–1 (= mg kg–1) of the trace element in the airdry soil. Include any extra dilution factors, and, if required, convert to ovendry soil using the appropriate factor, as in Method 5.2, Calculation (2).

Discussion 5.12. Determination of extractable sulphur A helpful discussion of sulphur in soils and its availability to plants is found in Combs et al. (1998) and Rowell (1994, pp. 213–215). Plants absorb sulphur mainly in the form of sulphate, which is the main form of sulphur occurring in the soil solution. The SO4-S is therefore the fraction usually measured. Over 90% of the surface soil sulphur occurs in combination with organic molecules from where it is mineralized to sulphate. The SO4-S concentration has been found to increase from approximately 5 kg ha–1 in the first 30 cm depth of a Wisconsin soil, to approximately 10 kg ha–1 in the 30–60 cm profile, and approximately 15 kg ha–1 in the 60–90 cm profile. It is therefore recommended that subsoil, as well as topsoil, cores are also taken for analysis. There are other sources of sulphur available to the plant, such as the seasonal effect of precipitation of sulphate-containing rain, especially near industrial areas and conurbations, and the sulphur in applied manure. There is also the sulphur adsorbed by clays and oxides of iron and aluminium, which will increase as the pH decreases below 6.5. The extractant may be water or 10 mM calcium chloride solution, but the latter may displace some adsorbed sulphate. In acidic soils, the available sulphur should include the adsorbed sulphate, therefore calcium phosphate [Ca(H2PO4)2] or potassium phosphate (KH2PO4), which will extract the adsorbed sulphate, are the extractants of choice. Calcium phosphate is preferred, because the calcium ion depresses the solubility of organic matter to produce a clearer filtrate. This is the method described below. For good reproducibility, it is essential to duplicate the conditions used to form the suspension. These include the temperature and the standing time before measuring the absorbance. A known quantity of sulphate ‘seed solution’ is usually added to improve the reproducibility of the sus-

Chapter 5

pension formation. This not only provides a nucleus to assist the uniform growth of barium sulphate crystals, but also ensures that the solubility product of barium sulphate has been exceeded in the final solution, and thus avoids a concave standard curve in the lower range. Acacia powder (gum acacia, gum arabic) is added to the ‘seed solution’ to stabilize the BaSO4 precipitate when larger amounts of sulphate are encountered. It may be omitted for soils low in sulphate. The presence of HCl in the seed solution prevents the co-precipitation of barium carbonate, phosphate or hydroxide, which would add to the turbidity.

Method 5.12a. Determination of extractable sulphur (manual method) Instrumentation. The determination of extracted sulphate-S may be carried out by ICP, ion chromatography or turbidimetry. The ICP procedure measures both organic and inorganic S present in the extract, but has a low methodological error. Ion chromatography may be affected by interference from phosphates and speed of sample throughput, but a suitable method is given by Combs et al. (1998). The most widely used method is by turbidimetry, where sulphate is precipitated as a white suspension of barium sulphate by the addition of barium chloride solution to the soil extract. The absorption of light is often measured using a nephelometer, spectrophotometer, or colorimeter at 480 nm, however, measurements at lower wavelengths increase sensitivity, but may incur possible curvature of the calibration graph. BaCl2 + SO42– = BaSO4↓ + 2Cl–

Reagents (extraction). • Activated charcoal, purified – place about 50 g Darco G-6 activated carbon in a wide-neck screw-cap container, add sufficient calcium phosphate extractant to completely wet it, then cap the bottle and shake for 5 min. Filter slowly with suction through a Buchner funnel, then wash three times successively with deionized water. Test the final leachate with a solution of barium chloride (approximately 1.4% m/v in 0.3 M HCl). If turbidity indicates the presence of sulphate, return the charcoal to a beaker, thoroughly mix with deionized water (boil for 15 min if necessary to get a clear test), refilter, wash and test for S as above. When satisfactory, dry overnight at 105°C and store in a tightly capped bottle. • Calcium phosphate extractant, 500 mg l–1 of phosphorus – dissolve 20.3 g calcium phosphate [Ca(H2PO4).2H2O] in water and make up to 10 l. Procedure (extraction). Weigh 10 g air-dry soil sieved to 2 mm (10 mesh) into a 50-ml conical flask. Add 25 ml of calcium phosphate extractant (50 ml for peat or compost) and shake on a reciprocating shaker (at approximately 200–275 oscillations of 25 mm per minute) for 30 min. If the presence of sol-

Analysis of Soil and Compost

uble organic matter is suspected, use a scoop to add 0.15 g purified activated charcoal (carbon), and shake for a further 3 min to enable a subsequent clear filtrate. Filter through a 125-mm Whatman No. 40 (or 42) filter paper, rejecting the first few millilitres. Carry out a blank determination. Reagents (determination). • Barium chloride crystals – sieve the powdered BaCl2.2H2O crystals, retaining the 520–860 µm (30–20 mesh) fraction. Warning: barium chloride is poisonous; wear PPE when handling. • Seed solution, 20 µg ml–1 of SO4-S – dissolve 0.1087 g of K2SO4 in 500 ml water in a 2-l beaker, and add 500 ml of hydrochloric acid (approximately 36% w/v HCl). Carefully place a Teflon-coated magnetic stirrer-bar into the beaker, place on to a magnetic stirrer and switch on. Add 2 g of acacia powder (see above discussion) slowly while stirring so as to avoid formation of any lumps. Transfer to a bottle and store in a refrigerator. Safety note: Acacia powder will cause severe irritation to the eyes, and also irritates the skin and the digestive and respiratory tracts; it is also a mutagen. Wear appropriate PPE in handling acacia powder. • Sulphate stock standard solution, 500 µg ml–1 of SO4-S – dissolve 2.717 g potassium sulphate (K2SO4), previously dried at 105°C for 1 h and cooled in a desiccator, in calcium phosphate extractant, then transfer to a 1–l volumetric flask with washings and make up to the mark with extractant. • Sulphate working standard solutions, 0–12 µg ml–1 of SO4-S – pipette 1, 2, 4, 6, 8 and 12 ml of the 500 µg ml–1 sulphate stock standard solution into 500-ml volumetric flasks, make up to the mark with calcium phosphate extractant and mix. This will give solutions containing 1, 2, 4, 6, 8 and 12 µg ml–1 of sulphate-S. Procedure (determination). If charcoal has been used in the sample extraction stage, then 25-ml aliquots of the working standards should be shaken with a 0.15 g scoop of purified charcoal for 3 min, and filtered through a 125 mm Whatman No. 40 (or 42) filter paper, rejecting the first few millilitres. Pipette 10-ml aliquots of standards, blank and sample extracts into a 50-ml conical flask containing a magnetic stirrer bar, add 1 ml of the ‘seed solution’, swirl to mix. Next add a 0.3-g scoop of BaCl2.2H2O crystals and stir magnetically for 1 min. Within the next 8 min, read the absorbance at 420 nm on a colorimeter or spectrophotometer fitted with a 40 mm optical cell. The background absorbance resulting from fine clay particles passing through the filter paper should also be measured on the soil extract plus ‘seed solution’, but without addition of BaCl2 crystals. Calculation. Read the µg ml–1 SO4-S for all the solutions from the standard graph. Add the values for the blank and background absorbance, and subtract the sum from the value of the sample solution to give a corrected value. Since 10 g soil provided 25 ml extract, multiply the corrected value of µg ml–1 SO4S in the extract by 2.5 to give the µg g–1 SO4-S (= mg kg–1 SO4-S) in the air-

Chapter 5

dry soil. Include any extra dilution factors, and, if required, convert to ovendry soil using the appropriate factor, as in Method 5.2, Calculation (2).

Method 5.12b. Determination of extractable sulphur (automated method) An automated method should improve reproducibility by maintaining constant conditions for the formation of the BaSO4 precipitate. A method suitable for the segmented-flow analysis of sulphate in soil and plant extracts using Skalar Analytical equipment has been proposed by Coutinho (1997). Soils are said to be extracted with double-distilled water according to the method in MAFF/ADAS (1986, pp. 215–216), however, that reference uses approximately 1.5 M HCl (10 g soil extracted with 70 ml water and 10 ml HCl, 36% m/m); and the absorbance is measured at 420 nm using a 50-mm cell path. It is suitable for soils up to 10 µg SO4-S ml–1 without a dilution step, and for soils up to 100 µg SO4-S ml–1 with automatic dilution of the sample. A method for soils low in sulphur (up to 1 µg SO4-S ml–1) using Technicon (Bran and Luebbe) AutoAnalyzer equipment was described by Bettany and Halstead (1972). The Turner Model 111 Fluorometer was modified to enable it to function as a nephelometer, but presumably a colorimeter could be substituted with measurement at 480 nm using a long path (40–50 mm) optical flowcell. The method, with some amendments, is summarized below. Reagents. • Barium chloride reagent – add 10.0 g of polyvinyl alcohol, 40.0 g of barium chloride (BaCl2.2H2O, turbidimetric grade) and 120 ml of 1.0 M HCl to 1800 ml of water. Warning: barium chloride is poisonous; wear PPE when handling. Heat on a stirrer-heater unit until the solution clarifies, then allow to cool before adjusting to 2 l. Allow to stand for 2–3 days, then filter through glass wool prior to use. • Calcium chloride extractant, 0.01 M – completely dissolve 1.47 g CaCl2.2H2O in water and make up to 1 l. • EDTA buffer wash – dissolve 40 g tetrasodium EDTA, and 6.75 g ammonium chloride in approximately 800 ml water, then, in a fume cupboard, carefully add 57 ml ammonia solution (approximately 35% m/m NH3). Stir and make up to 1 l with water. • Hydrochloric acid, 1.0 M. • Hydrochloric acid, 0.5 M. • Hydrochloric acid, 0.085 M. • Sodium peroxide solution, 2% m/v. (Note: sodium peroxide is highly corrosive, very irritant to skin and mucous membranes, and may ignite combustible materials; wear suitable PPE.) • Sulphate stock standard solution, 500 µg ml–1 of SO4-S – dissolve 2.717 g potassium sulphate (K2SO4), previously dried at 105°C for 1 h and cooled in a desiccator, in water, then transfer to a 1–l volumetric flask with washings and make up to the mark with extractant.

Analysis of Soil and Compost

• Sulphate intermediate standard solution, 50 µg ml–1 – pipette 10 ml of the sulphate stock standard into a 100-ml volumetric flask, make up to the mark with water and mix. • Sulphate working standards, 0.25–1.5 µg ml–1 – pipette aliquots of 1, 2, 3, 4, 5 and 6 ml of the sulphate intermediate standard solution into 200-ml volumetric flasks, make up to the mark with water and mix. This gives a series of standard solutions of 0.25, 0.50, 0.75, 1.00, 1.25 and 1.50 µg SO4-S ml–1. Procedure (extraction). Add 25 g air-dry soil, sieved to 2 mm (10 mesh), into a 125-ml conical flask. Add 50 ml calcium chloride extractant, 0.01 M, and shake on a reciprocating shaker (at approximately 200–275 oscillations of 25 mm per minute) for 30 min. Filter through a Whatman No. 30 filter paper, rejecting the first few millilitres, and pipette 25 ml of the filtrate into a 50-ml beaker. Slowly add 2 ml sodium peroxide solution using a plastic pasteur pipette. Allow to stand for 5 min, after which the gelatinous precipitate of interfering cations and organic matter is filtered off using a Whatman No. 42 filter paper. Wash the precipitate and filter paper with water, adjust the filtrate to approximately pH 3 with 0.5 M HCl, transfer with rinsing to a 50-ml volumetric flask, make up to the mark with water and mix. Carry out a blank determination. Procedure (determination). Samples, blank, and standards are poured into the sample cups. Industrial 8.5-ml autoanalyser cups (Part No. 127-0080-01) are available from Gradko or LIP (Equipment & Services) Ltd (see Chapter 1, ‘Chemistry module’). Technicon Sampler IV would require the 10-ml cups, same Part No. from Bran + Luebbe. Every 10th cup should contain the EDTA buffer wash solution followed by a cup containing water; this is to prevent build-up of barium sulphate on the walls of the mixing coil and flowcell. Pump the reagents for about 30 min at the start to flex the pump tubes. Sample the highest standard several times and adjust the sensitivity of the colorimeter/spectrophotometer and/or chart recorder to give a reading of about 90% full-scale. Adjust the zero setting and baseline reading to about 5% full-scale for aspiration of the wash solution. If the high standard reading is too low or high, alter the ratio of flow rates of the sample and barium chloride reagent as appropriate. If the readings from the soil samples are too high, take a smaller weight of sample for the extraction. The flow diagram is shown in Fig. 5.5. Calculation. The 25 g soil was shaken with 50 ml calcium chloride extractant, and 25 ml of this extract was diluted to 50 ml. There is therefore a ×4 dilution factor. Calculate the concentration of SO4-S in the blank and samples by comparison with the standard curve. Subtract the blank value from the sample values and multiply by 4 to give the µg SO4-S g–1 (= mg SO4-S kg–1) air-dry soil. Include any extra dilution factors, and, if required, convert to oven-dry soil using the appropriate factor, as in Method 5.2, Calculation (2).

Chapter 5

Sampler 20 h-1 ratio 1:3 sample:wash To waste

Pump Double e g mixing coil

2.50 ml min-1 sample 2.50 ml min-1 sample 2.50 ml min-1 BaCl2 reagent 2.00 ml min-1 airr

Debubbler Colorimeter 480 nm

To chart recorder

To sampler wash

2.50 ml min-1 0.085 M HCl 2.50 ml min-1

From sampler wash (if required)

2.50 ml min-1 2.90 ml min-1 3.40 ml min-1 flowcell waste

Fig. 5.5. Manifold for the automated determination of sulphate in soil extracts.

The Analysis of Composts The term compost has been defined by Zucconi and Bertoldi (1987) as ‘the stabilized and sanitized product of composting which is beneficial to plant growth. It has undergone an initial, rapid stage of decomposition and is in the process of humification.’ The initial thermophilic stage of decomposition is the means of self-sanitizing and removing pathogens. If the compost is insufficiently humified, it is immature, and the wide C:N ratio causes it to immobilize soil nitrogen as it continues to actively decompose in the soil. If sufficiently sanitized and humified, the compost is said to be biomature. The development of globally accepted criteria for compost specifications is still at an early stage, so some scientists have proposed biomaturity tests (Mathur et al., 1993). Methods on-line The United States Environmental Protection Agency (EPA) has a website where physical and chemical test methods for evaluating solid wastes may be downloaded as pdf files: http://www.epa.gov/sw-846/main.htm We understand here by compost a marketed product of an organic based material derived from a variety of sources. These might be treated municipal

Analysis of Soil and Compost

waste, spent mushroom compost, a bracken- or seaweed-based compost, agricultural and food processing wastes etc., which might be put to agricultural use. Composts are often very heterogeneous, which makes it difficult to prepare a sufficiently hom*ogeneous sample. The high humus content makes them similar to peat soils, where organic matter can exceed 95%, which can affect not only the analytical method, but also the interpretation of the results in making fertilizer recommendations. Typical specifications Typical parameters and nutrient levels for assessment of compost quality are shown in Table 5.2. These are combined values from a variety of sources, including Bertoldi et al. (1987), and are merely intended to help in setting up analytical procedures. Some typical and preferred heavy and trace element concentrations for soils and municipal composts are shown in Table 5.3. The levels in soils are typical for dilute aqueous extractants such as 0.05 M EDTA, 0.5 M acetic acid, hot water for boron, and, for molybdenum, Tamm’s reagent (acid ammonium oxalate; Reisenauer, 1965). Tables in the literature often give total values obtained spectrographically, by XRF, or by extraction with hot

Table 5.2. Typical parameters and nutrient levels for assessment of compost quality.

Parameter Dry matter (DM) Organic matter pH Salinity (as NaCl) (conductivity)

Minimum

Normal range

5.5

40–60% 20–40% DM 6.5–8.0

Nutrient (g kg–1 DM) Calcium (CaO) 20 Calcium (Ca) 14 Magnesium (MgO) 3 Magnesium (Mg) 1.8 Nitrogen (Kjeld.) 6 Organic-N % total 90% NH4-N NO3-N NO3-N/ NH4-N CN Phosphorus (P2O5) 5 Phosphorus (P) 2.2 Potassium (K2O) 3 Potassium (K) 2.5

1–2 dS m–1

Maximum

8.0 2.0 g l–1 2.0 dS m–1

Preferred values lower ×2 organic C 7.0–8.0 0.5 dS m–1

35–140 25–100 4–16 2.4–9.7 6–13 2–5 50–200 ×20–80 ratio ×10–20 ratio 3.5–14 1.5–6.0 4.5–18 3.7–15

High 0.4

Low High 60

22 High High High High

100

Chapter 5

Table 5.3. Some typical and preferred heavy and trace element concentrations for soils and composts; various local regulations specify different maximum levels. Heavy and trace elements Boron Cadmium Cobalt Copper Chromium Iron Lead Manganese Mercury Molybdenum Nickel Zinc

Symbol B Cd Co Cu Cr Fe Pb Mn Hg Mo Ni Zn

Typical/ Municipal preferred compost in soil (typical) (mg kg–1 DM) (mg kg–1 DM) 0.01–10/0.5–

Methods in Agricultural Chemical Analysis: A Practical Handbook (CABI Publishing) - PDF Free Download (2024)