Soil Sampling and Methods of Analysis, Second Edition

Second Edition Soil Sampling and Methods of Analysis Soil Sampling and Methods of Analysis Second Edition Edited by ...

Author: E.G. Gregorich | M.R. Carter

279 downloads 2625 Views 9MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Second Edition

Soil Sampling and Methods of Analysis

Soil Sampling and Methods of Analysis Second Edition

Edited by

M.R. Carter E.G. Gregorich

Carter • Gregorich

Canadian Society of Soil Science

3586_Cover_FinalMechanical.indd 1

6/8/07 10:09:41 AM

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page i 27.6.2007 1:53pm Compositor Name: JGanesan


E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page ii 27.6.2007 1:53pm Compositor Name: JGanesan

In physical science the first essential step in the direction of learning any subject is to find principles of numerical reckoning and practicable methods for measuring some quality connected with it. I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely in your thoughts advanced to the state of science, whatever the matter may be. Lord Kelvin, Popular Lectures and Addresses (1891–1894), vol. 1, Electrical Units of Measurement

Felix qui potuit rerum cognoscere causas. Happy the man who has been able to learn the causes of things. Virgil: Georgics (II, 490)

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page iii 27.6.2007 1:53pm Compositor Name: JGanesan


Edited by

M.R. Carter E.G. Gregorich

Canadian Society of Soil Science

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page iv

27.6.2007 1:53pm Compositor Name: JGanesan

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2008 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-13: 978-0-8493-3586-0 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http:// www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Soil sampling and methods of analysis / edited by M.R. Carter and E.G. Gregorich. -- 2nd ed. p. cm. Includes bibliographical references and index. ISBN-13: 978-0-8493-3586-0 (alk. paper) ISBN-10: 0-8493-3586-8 (alk. paper) 1. Soils--Analysis. 2. Soils--Sampling. I. Carter, Martin R. II. Gregorich, E. G. III. Title. S593.S7425 2007 631.4’1--dc22 Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

2006102606

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page v 27.6.2007 1:53pm Compositor Name: JGanesan

PREFACE This volume is an update of the book, Soil Sampling and Methods of Analysis, first published in 1993. The aims of this second edition remain the same as those of the earlier edition—to provide a compilation of soil analytical and sampling methods that are commonly used, straightforward, and relatively easy to use. The materials and procedures for these methods are presented with sufficient detail and information, along with key references, to characterize the potential and limitation of each method. As methods develop, so do their degree of sophistication. Taking these developments into account, the second edition includes several chapters that serve as ‘‘primers,’’ the purpose of which is to describe the overall principles and concepts behind a particular type or types of measurement, rather than just methods alone. All of the chapters retained from the earlier edition have been modified and updated. The second edition also introduces new chapters, particularly in the areas of biological and physical analyses, and soil sampling and handling. For example, the ‘‘Soil Biological Analyses’’ section contains new chapters to reflect the growing number and assortment of new microbiological techniques and the burgeoning interest in soil ecology. New chapters are offered describing tools that characterize the dynamics and chemistry of soil organic matter. A new section devoted to soil water presents up-to-date field- and laboratory-based methods that characterize saturated and unsaturated soil hydraulic properties. This second edition of Soil Sampling and Methods of Analysis comprises 7 sections and a total of 85 chapters and 2 appendices written by 140 authors and co-authors. Each section is assembled by two section editors and each chapter reviewed by at least two external reviewers. We are grateful to these people for their diligent work in polishing and refining the text and helping to bring this new volume to fruition. We particularly thank Elaine Nobbs for her support in working with the many authors involved in writing this book. We offer this new edition of Soil Sampling and Methods of Analysis in the belief that it will continue as a useful tool for researchers and practitioners working with soil. M.R. Carter and E.G. Gregorich Editors

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page vi


CANADIAN SOCIETY OF SOIL SCIENCE The Canadian Society of Soil Science is a nongovernmental, nonprofit organization for scientists, engineers, technologists, administrators, students, and others interested in soil science. Its three main objectives are .

To promote the wise use of soil for the benefit of society

.

To facilitate the exchange of information and technology among people and organizations involved in soil science

.

To promote research and practical application of findings in soil science

The society produces the international scientific publication, the Canadian Journal of Soil Science, and each year hosts an international soil science conference. It sponsored the first edition of Soil Sampling and Methods of Analysis (Lewis Publishers, CRC Press, 1993) and also promoted the publication of the popular reference book Soil and Environmental Science Dictionary (CRC Press, 2001). The society publishes a newsletter to share information and ideas, and maintains active liaison and partnerships with other soil science societies. For more information about the Canadian Society of Soil Science, please visit www.csss.ca.

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page vii 27.6.2007 1:53pm Compositor Name: JGanesan

EDITORS M.R. Carter holds degrees in agriculture and soil science from the University of Alberta and obtained a PhD in soil science from the University of Saskatchewan in 1983. Since 1977, he has held agricultural research positions with Agriculture and Agri-Food Canada (AAFC) and is currently a research scientist at the AAFC Research Center, Charlottetown, Prince Edward Island. Dr. Carter is a fellow and past-president of the Canadian Society of Soil Science, and past editor of the Canadian Journal of Soil Science. He edited the first edition of Soil Sampling and Methods of Analysis, (CRC Press, 1993) and also edited Conservation Tillage in Temperate Agroecosystems (CRC Press, 1994) and Structure and Organic Matter Storage in Agricultural Soils (CRC Press, 1996). In collaboration with Dr. Gregorich, he edited Soil Quality for Crop Production and Ecosystem Health (Elsevier, 1997) and Soil & Environmental Science Dictionary (CRC Press, 2001). Dr. Carter presently serves as editorin-chief for the international scientific journal Agriculture Ecosystems & Environment. E.G. Gregorich is a research scientist with Agriculture and Agri-Food Canada at the Central Experimental Farm in Ottawa, Canada. His work focuses on soil biochemistry, particularly carbon and nitrogen cycling in soil. He is a fellow and past-president of the Canadian Society of Soil Science, and has served the Soil Science Society of America as chair of the soil biology and biochemistry division. Dr. Gregorich has been a member of the International Panel on Climate Change, has conducted field studies in Scotland, New Zealand, and Antarctica, and directs a Canadian international development project in Vietnam. He has served as associate editor for the Journal of Environmental Quality; Agriculture, Ecosystems & Environment; European Journal of Soil Science; and the Canadian Journal of Soil Science. This is the third book on which he and Dr. Carter have collaborated as editors.

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page viii


E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page ix


CONTRIBUTORS D. Acosta-Mercado Department of Biology University of Puerto Rico Mayaguez, Puerto Rico J.A. Addison School of Sustainability and Environment Royal Roads University Victoria, British Columbia, Canada

B.C. Ball Scottish Agricultural College Edinburgh, Scotland, United Kingdom M.H. Beare New Zealand Institute for Crop and Food Research Christchurch, New Zealand

S.M. Adl Department of Biology Dalhousie University Halifax, Nova Scotia, Canada

E.G. Beauchamp Department of Land Resource Science University of Guelph Guelph, Ontario, Canada

D.W. Anderson Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada

V.M. Behan-Pelletier Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

Denis A. Angers Agriculture and Agri-Food Canada Quebec, Quebec, Canada

N. Be´langer Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada

H. Antoun Department of Soils and Agrifood Engineering Laval University Quebec, Quebec, Canada

Normand Bertrand Agriculture and Agri-Food Canada Quebec, Quebec, Canada

J.M. Arocena College of Science and Management University of Northern British Columbia Prince George, British Columbia, Canada V.L. Bailey Biological Sciences Division Pacific Northwest National Laboratory Richland, Washington, United States G.H. Baker Entomology Commonwealth Scientific and Industrial Research Organization Glen Osmond, South Australia, Australia J.A. Baldock Land and Water Commonwealth Scientific and Industrial Research Organization Glen Osmond, South Australia, Australia

R.P. Beyaert Agriculture and Agri-Food Canada London, Ontario, Canada H. Bolton, Jr. Biological Sciences Division Pacific Northwest National Laboratory Richland, Washington, United States Jeff Braidek Saskatchewan Agriculture and Food Saskatoon, Saskatchewan, Canada E. Bremer Symbio Ag Consulting Lethbridge, Alberta, Canada J.A. Brierley Agriculture and Agri-Food Canada Edmonton, Alberta, Canada

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page x 27.6.2007 1:53pm Compositor Name: JGanesan

P.C. Brookes Agriculture and Environment Division Rothamsted Research Harpenden, Hertfordshire, United Kingdom M.S. Bullock Holly Hybrids Sheridan, Wyoming, United States B.J. Cade-Menun Department of Geological and Environmental Sciences Stanford University Stanford, California, United States C.A. Campbell Agriculture and Agri-Food Canada Ottawa, Ontario, Canada J. Caron Department of Soils and Agrifood Engineering Laval University Quebec, Quebec, Canada

H.P. Cresswell Land and Water Commonwealth Scientific and Industrial Research Organization Canberra, Australian Capital Territory Australia J.A. Crumbaugh Canadian Forest Service Natural Resources Canada Edmonton, Alberta, Canada J.L.B. Culley Agriculture and Agri-Food Canada Ottawa, Ontario, Canada M.P. Curran British Columbia Ministry of Forests Nelson, British Columbia, Canada Denis Curtin New Zealand Institute for Crop and Food Research Christchurch, New Zealand

M.R. Carter Agriculture and Agri-Food Canada Charlottetown, Prince Edward Island Canada

Y. Dalpe´ Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

Martin H. Chantigny Agriculture and Agri-Food Canada Quebec, Quebec, Canada

Pauline De´fossez French National Institute for Agricultural Research Laon, France

M.J. Clapperton Agriculture and Agri-Food Canada Lethbridge, Alberta, Canada F.J. Cook Land and Water Commonwealth Scientific and Industrial Research Organization Indooroopilly, Queensland, Australia F. Courchesne Department of Geography University of Montreal Montreal, Quebec, Canada

J.R. de Freitas Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada C.F. Drury Agriculture and Agri-Food Canada Harrow, Ontario, Canada K.E. Dunfield Department of Land Resource Science University of Guelph Guelph, Ontario, Canada

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xi

M. Duquette SNC-Lavalin Montreal, Quebec, Canada B.H. Ellert Agriculture and Agri-Food Canada Lethbridge, Alberta, Canada J.A. Elliott Environment Canada Saskatoon, Saskatchewan, Canada D.E. Elrick Department of Land Resource Science University of Guelph Guelph, Ontario, Canada R.E. Farrell Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada Ty P.A. Ferre´ Department of Hydrology and Water Resources University of Arizona Tucson, Arizona, United States


C.D. Grant School of Earth and Environmental Sciences University of Adelaide Glen Osmond, South Australia, Australia E.G. Gregorich Agriculture and Agri-Food Canada Ottawa, Ontario, Canada M. Grimmett Agriculture and Agri-Food Canada Charlottetown, Prince Edward Island Canada P.H. Groenevelt Department of Land Resource Science University of Guelph Guelph, Ontario, Canada Umesh C. Gupta Agriculture and Agri-Food Canada Charlottetown, Prince Edward Island Canada C. Hamel Agriculture and Agri-Food Canada Swift Current, Saskatchewan, Canada

C.T. Figueiredo Department of Renewable Resources University of Alberta Edmonton, Alberta, Canada

X. Hao Agriculture and Agri-Food Canada Lethbridge, Alberta, Canada

T.A. Forge Agriculture and Agri-Food Canada Agassiz, British Columbia, Canada

S.C. Hart School of Forestry and Merriam-Powell Center for Environmental Research Northern Arizona University Flagstaff, Arizona, United States

C.A. Fox Department of Renewable Resources Agriculture and Agri-Food Canada Harrow, Ontario, Canada

A. Hartmann National Institute of Agronomic Research Dijon, France

J.J. Germida Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada

W.H. Hendershot Department of Renewable Resources McGill University Sainte Anne de Bellevue, Quebec, Canada

Tee Boon Goh Department of Soil Science University of Manitoba Winnipeg, Manitoba, Canada

Ganga M. Hettiarachchi School of Earth and Environmental Sciences University of Adelaide Glen Osmond, South Australia, Australia

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xii 27.6.2007 1:53pm Compositor Name: JGanesan

D.W. Hopkins Scottish Crop Research Institute Dundee, Scotland, United Kingdom

C.G. Kowalenko Agriculture and Agri-Food Canada Agassiz, British Columbia, Canada

H.H. Janzen Agriculture and Agri-Food Canada Lethbridge, Alberta, Canada

D. Kroetsch Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

R.G. Kachanoski Department of Renewable Resources University of Alberta Edmonton, Alberta, Canada

H. Lalande Department of Renewable Resources McGill University Sainte Anne de Bellevue, Quebec, Canada

Klaus Kaiser Soil Sciences Martin Luther University Halle-Wittenberg, Halle, Germany

David R. Lapen Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

Karsten Kalbitz Soil Ecology University of Bayreuth Bayreuth, Germany Y.P. Kalra Canadian Forest Service Natural Resources Canada Edmonton, Alberta, Canada A. Karam Department of Soils and Agrifood Engineering Laval University Quebec, Quebec, Canada Thomas Keller Department of Soil Sciences Swedish University of Agricultural Sciences Uppsala, Sweden J. Kimpinski Agriculture and Agri-Food Canada Charlottetown, Prince Edward Island Canada Peter J.A. Kleinman Pasture Systems and Watershed Management Research Center U.S. Department of Agriculture University Park, Pennsylvania United States

F.J. Larney Agriculture and Agri-Food Canada Lethbridge, Alberta, Canada R. Lessard Environmental Division Bodycote Testing Group Edmonton, Alberta, Canada B.C. Liang Environment Canada Gatineau, Quebec, Canada N.J. Livingston Department of Biology University of Victoria Victoria, British Columbia, Canada D.H. Lynn Department of Integrative Biology University of Guelph Guelph, Ontario, Canada J.D. MacDonald Agriculture and Agri-Food Canada Quebec, Quebec, Canada D.G. Maynard Pacific Forestry Centre Natural Resources Canada Victoria, British Columbia, Canada

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xiii

R.A. McBride Department of Land Resource Science University of Guelph Guelph, Ontario, Canada


D.C. Olk U.S. Department of Agriculture Agriculture Research Service National Soil Tilth Laboratory Ames, Iowa, United States

W.B. McGill College of Science and Management University of Northern British Columbia Prince George, British Columbia Canada

D. Pare´ Natural Resources Canada Canadian Forest Service Quebec, Quebec, Canada

G.R. Mehuys Department of Renewable Resources McGill University Sainte Anne de Bellevue, Quebec, Canada

L.E. Parent Department of Soils and Agrifood Engineering Laval University Quebec, Quebec, Canada

A.R. Mermut Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada

G.W. Parkin Department of Land Resource Science University of Guelph Guelph, Ontario, Canada

J.C. Michel INH–INRA–University of Angers Angers, France

G.T. Patterson Agriculture and Agri-Food Canada Truro, Nova Scotia, Canada

Jim J. Miller Agriculture and Agri-Food Canada Lethbridge, Alberta, Canada

Dan Pennock Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada

J.O. Moir Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada D.D. Myrold Department of Crop and Soil Science Oregon State University Corvallis, Oregon, United States R. Naasz Department of Soils and Agrifood Engineering Laval University Quebec, Quebec, Canada I.P. O’Halloran University of Guelph Ridgetown, Ontario, Canada

Caroline Preston Pacific Forestry Centre Natural Resources Canada Victoria, British Columbia, Canada D. Pre´vost Agriculture and Agri-Food Canada Quebec, Quebec, Canada P. Qian Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada D. Reyes Department of Renewable Resources McGill University Sainte Anne de Bellevue, Quebec, Canada

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xiv 27.6.2007 1:53pm Compositor Name: JGanesan

W.D. Reynolds Agriculture and Agri-Food Canada Harrow, Ontario, Canada Guy Richard French National Institute for Agricultural Research Olivet, France Philippe Rochette Agriculture and Agri-Food Canada Quebec, Quebec, Canada L. Rock Agriculture and Agri-Food Canada Lethbridge, Alberta, Canada P.M. Rutherford College of Science and Management University of Northern British Columbia Prince George, British Columbia, Canada S. Sauve´ Department of Chemistry University of Montreal Montreal, Quebec, Canada J.J. Schoenau Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada Andrew N. Sharpley Crop, Soil and Environmental Sciences University of Arkansas Fayetteville, Arkansas, United States S.C. Sheppard ECOMatters Inc. W.B. Lewis Business Centre Pinawa, Manitoba, Canada B.C. Si Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada Myrna J. Simpson Department of Physical and Environmental Sciences University of Toronto Toronto, Ontario, Canada

J.O. Skjemstad Land and Water Commonwealth Scientific and Industrial Research Organization Glen Osmond, South Australia, Australia J.L. Smith U.S. Department of Agriculture Agriculture Research Service Washington State University Pullman, Washington, United States Y.K. Soon Agriculture and Agri-Food Canada Beaverlodge, Alberta, Canada P. St-Georges Agriculture and Agri-Food Canada Ottawa, Ontario, Canada C. Swyngedouw Environmental Division Bodycote Testing Group Calgary, Alberta, Canada M. Tenuta Department of Soil Science University of Manitoba Winnipeg, Manitoba, Canada Y.-C. Tien Agriculture and Agri-Food Canada London, Ontario, Canada H. Tiessen Inter-American Institute for Global Change Research Sao Jose dos Campos Sao Paulo, Brazil E. Topp Agriculture and Agri-Food Canada London, Ontario, Canada G. Clarke Topp Agriculture and Agri-Food Canada Ottawa, Ontario, Canada T. Sen Tran Institute of Research and Development in Agroenvironment Quebec, Quebec, Canada

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xv 27.6.2007 1:53pm Compositor Name: JGanesan

M.-C. Turmel Department of Geography University of Montreal Montreal, Quebec, Canada A.J. VandenBygaart Agriculture and Agri-Food Canada Ottawa, Ontario, Canada Ken C.J. Van Rees Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada R.P. Voroney Department of Land Resource Science University of Guelph Guelph, Ontario, Canada C. Wang Agriculture and Agri-Food Canada Ottawa, Ontario, Canada Jennifer L. Weld Department of Crop and Soil Sciences The Pennsylvania State University University Park, Pennsylvania, United States G. Wen Lemington, Ontario, Canada

O.O.B. Wendroth Department of Plant and Soil Sciences University of Kentucky Lexington, Kentucky, United States J.P. Winter Nova Scotia Agricultural College Truro, Nova Scotia, Canada N. Wypler Leibniz-Centre for Agricultural Landscape Research Institute for Soil Landscape Research Mu¨ncheberg, Germany X.M. Yang Agriculture and Agri-Food Canada Harrow, Ontario, Canada Thomas Yates Department of Soil Science University of Saskatchewan Saskatoon, Saskatchewan, Canada N. Ziadi Agriculture and Agri-Food Canada Quebec, Quebec, Canada

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xvi 27.6.2007 1:53pm Compositor Name: JGanesan

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xvii 27.6.2007 1:53pm Compositor Name: JGanesan

TABLE OF CONTENTS

I.

SOIL SAMPLING AND HANDLING Section Editors: G.T. Patterson and M.R. Carter

1. 2. 3. 4. 5.

II.

Soil Sampling Designs Dan Pennock, Thomas Yates, and Jeff Braidek Sampling Forest Soils N. Be´langer and Ken C. J. Van Rees Measuring Change in Soil Organic Carbon Storage B.H. Ellert, H.H. Janzen, A.J. VandenBygaart, and E. Bremer Soil Sample Handling and Storage S.C. Sheppard and J.A. Addison Quality Control in Soil Chemical Analysis C. Swyngedouw and R. Lessard

1 15 25 39 51

DIAGNOSTIC METHODS FOR SOIL AND ENVIRONMENTAL MANAGEMENT Section Editors: J.J. Schoenau and I.P. O’Halloran

6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

III.

Nitrate and Exchangeable Ammonium Nitrogen D.G. Maynard, Y.P. Kalra, and J.A. Crumbaugh Mehlich 3-Extractable Elements N. Ziadi and T. Sen Tran Sodium Bicarbonate-Extractable Phosphorus J. J. Schoenau and I.P. O’Halloran Boron, Molybdenum, and Selenium Ganga M. Hettiarachchi and Umesh C. Gupta Trace Element Assessment W.H. Hendershot, H. Lalande, D. Reyes, and J.D. MacDonald Readily Soluble Aluminum and Manganese in Acid Soils Y.K. Soon, N. Be´langer, and W.H. Hendershot Lime Requirement N. Ziadi and T. Sen Tran Ion Supply Rates Using Ion-Exchange Resins P. Qian, J.J. Schoenau, and N. Ziadi Environmental Soil Phosphorus Indices Andrew N. Sharpley, Peter J.A. Kleinman, and Jennifer L. Weld Electrical Conductivity and Soluble Ions Jim J. Miller and Denis Curtin

71 81 89 95 109 121 129 135 141 161

SOIL CHEMICAL ANALYSES Section Editors: Y.K. Soon and W.H. Hendershot

16.

Soil Reaction and Exchangeable Acidity W.H. Hendershot, H. Lalande, and M. Duquette

173

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xviii 27.6.2007 1:53pm Compositor Name: JGanesan

17.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

IV.

Collection and Characterization of Soil Solutions J.D. MacDonald, N. Be´langer, S. Sauve´, F. Courchesne, and W.H. Hendershot Ion Exchange and Exchangeable Cations W.H. Hendershot, H. Lalande, and M. Duquette Nonexchangeable Ammonium Y.K. Soon and B.C. Liang Carbonates Tee Boon Goh and A.R. Mermut Total and Organic Carbon J.O. Skjemstad and J.A. Baldock Total Nitrogen P.M. Rutherford, W.B. McGill, J.M. Arocena, and C.T. Figueiredo Chemical Characterization of Soil Sulfur C.G. Kowalenko and M. Grimmett Total and Organic Phosphorus I.P. O’Halloran and B.J. Cade-Menum Characterization of Available P by Sequential Extraction H. Tiessen and J.O. Moir Extractable Al, Fe, Mn, and Si F. Courchesne and M.-C. Turmel Determining Nutrient Availability in Forest Soils N. Be´langer, D. Pare´, and W.H. Hendershot Chemical Properties of Organic Soils A. Karam

179

197 207 215 225 239 251 265 293 307 317 331

SOIL BIOLOGICAL ANALYSES Section Editors: E. Topp and C.A. Fox

29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

Cultural Methods for Soil and Root-Associated Microorganisms J. J. Germida and J.R. de Freitas Arbuscular Mycorrhizae Y. Dalpe´ and C. Hamel Root Nodule Bacteria and Symbiotic Nitrogen Fixation D. Pre´vost and H. Antoun Microarthropods J.P. Winter and V.M. Behan-Pelletier Nematodes T.A. Forge and J. Kimpinski Earthworms M.J. Clapperton, G.H. Baker, and C.A. Fox Enchytraeids S.M. Adl Protozoa S.M. Adl, D. Acosta-Mercado, and D.H. Lynn Denitrification Techniques for Soils C.F. Drury, D.D. Myrold, E.G. Beauchamp, and W.D. Reynolds Nitrification Techniques for Soils C.F. Drury, S.C. Hart, and X.M. Yang

341 355 379 399 415 427 445 455 471 495

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xix 27.6.2007 1:53pm Compositor Name: JGanesan

39.

40. 41. 42. 43.

44.

V.

Substrate-Induced Respiration and Selective Inhibition as Measures of Microbial Biomass in Soils V.L. Bailey, H. Bolton, Jr., and J.L. Smith Assessment of Soil Biological Activity R.P. Beyaert and C.A. Fox Soil ATP R.P. Voroney, G. Wen, and R.P. Beyaert Lipid-Based Community Analysis K.E. Dunfield Bacterial Community Analyses by Denaturing Gradient Gel Electrophoresis E. Topp, Y.-C. Tien, and A. Hartmann Indicators of Soil Food Web Properties T. A. Forge and M. Tenuta

515 527 547 557

567 577

SOIL ORGANIC MATTER ANALYSES Section Editors: E.G. Gregorich and M.H. Beare

45. 46. 47. 48. 49. 50. 51. 52. 53.

54.

VI.

Carbon Mineralization D.W. Hopkins Mineralizable Nitrogen Denis Curtin and C.A. Campbell Physically Uncomplexed Organic Matter E.G. Gregorich and M.H. Beare Extraction and Characterization of Dissolved Organic Matter Martin H. Chantigny, Denis A. Angers, Klaus Kaiser, and Karsten Kalbitz Soil Microbial Biomass C, N, P, and S R.P. Voroney, P.C. Brookes, and R.P. Beyaert Carbohydrates Martin H. Chantigny and Denis A. Angers Organic Forms of Nitrogen D.C. Olk Soil Humus Fractions D.W. Anderson and J.J. Schoenau Soil Organic Matter Analysis by Solid-State 13C Nuclear Magnetic Resonance Spectroscopy Myrna J. Simpson and Caroline Preston Stable Isotopes in Soil and Environmental Research B.H. Ellert and L. Rock

589 599 607 617 637 653 667 675

681 693

SOIL PHYSICAL ANALYSES Section Editors: Denis A. Angers and F.J. Larney

55. 56.

Particle Size Distribution D. Kroetsch and C. Wang Soil Shrinkage C.D. Grant

713 727

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xx 27.6.2007 1:53pm Compositor Name: JGanesan

57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67.

68.

VII.

Soil Density and Porosity X. Hao, B.C. Ball, J.L.B. Culley, M.R. Carter, and G.W. Parkin Soil Consistency: Upper and Lower Plastic Limits R.A. McBride Compaction and Compressibility Pauline De´fossez, Thomas Keller, and Guy Richard Field Soil Strength G. Clarke Topp and David R. Lapen Air Permeability C.D. Grant and P.H. Groenevelt Aggregate Stability to Water Denis A. Angers, M.S. Bullock, and G.R. Mehuys Dry-Aggregate Size Distribution F.J. Larney Soil Air R.E. Farrell and J.A. Elliott Soil-Surface Gas Emissions Philippe Rochette and Normand Bertrand Bulk Density Measurement in Forest Soils D.G. Maynard and M.P. Curran Physical Properties of Organic Soils and Growing Media: Particle Size and Degree of Decomposition L.E. Parent and J. Caron Physical Properties of Organic Soils and Growing Media: Water and Air Storage and Flow Dynamics J. Caron, D.E. Elrick, J.C. Michel, and R. Naasz

743 761 771 783 803 811 821 833 851 863

871

885

SOIL WATER ANALYSES Section Editors: W.D. Reynolds and G. Clarke Topp

69. 70. 71. 72.

73. 74. 75. 76.

Soil Water Analyses: Principles and Parameters W.D. Reynolds and G. Clarke Topp Soil Water Content G. Clarke Topp, G.W. Parkin, and Ty P.A. Ferre´ Soil Water Potential N.J. Livingston and G. Clarke Topp Soil Water Desorption and Imbibition: Tension and Pressure Techniques W.D. Reynolds and G. Clarke Topp Soil Water Desorption and Imbibition: Long Column W.D. Reynolds and G. Clarke Topp Soil Water Desorption and Imbibition: Psychrometry W.D. Reynolds and G. Clarke Topp Saturated Hydraulic Properties: Laboratory Methods W.D. Reynolds Saturated Hydraulic Properties: Well Permeameter W.D. Reynolds

913 939 963

981 999 1007 1013 1025

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xxi 27.6.2007 1:53pm Compositor Name: JGanesan

77. 78. 79. 80. 81. 82. 83. 84. 85.

Saturated Hydraulic Properties: Ring Infiltrometer W.D. Reynolds Saturated Hydraulic Properties: Auger Hole G. Clarke Topp Saturated Hydraulic Properties: Piezometer G. Clarke Topp Unsaturated Hydraulic Conductivity: Laboratory Tension Infiltrometer F.J. Cook Unsaturated Hydraulic Properties: Laboratory Evaporation O.O.B. Wendroth and N. Wypler Unsaturated Hydraulic Properties: Field Tension Infiltrometer W.D. Reynolds Unsaturated Hydraulic Properties: Instantaneous Profile W.D. Reynolds Estimation of Soil Hydraulic Properties F.J. Cook and H.P. Cresswell Analysis of Soil Variability B.C. Si, R.G. Kachanoski, and W. D. Reynolds

1043 1057 1065 1075 1089 1107 1129 1139 1163

APPENDIX A. B.

Site Description G.T. Patterson and J.A. Brierley General Safe Laboratory Operation Procedures P. St-Georges

INDEX

1193 1197

1205

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C000 Final Proof page xxii 27.6.2007 1:53pm Compositor Name: JGanesan

I. SOIL SAMPLING AND HANDLING Section Editors: G.T. Patterson and M.R. Carter

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C001 Final Proof page 1 10.6.2007 5:52pm Compositor Name: BMani

Chapter 1 Soil Sampling Designs Dan Pennock and Thomas Yates University of Saskatchewan Saskatoon, Saskatchewan, Canada

Jeff Braidek Saskatchewan Agriculture and Food Saskatoon, Saskatchewan, Canada

1.1 INTRODUCTION Sampling involves the selection from the total population of a subset of individuals upon which measurements will be made; the measurements made on this subset (or sample) will then be used to estimate the properties (or parameters) of the total population. Sampling is inherent to any field research program in soil science because the measurement of the total population is impossible for any realistic study. For example, even a single 10 ha field contains about 100,000 1 m2 soil pits or 1107 10 cm2 cores, and sampling of the entire population would be more of an unnatural obsession than a scientific objective. Sampling design involves the selection of the most efficient method for choosing the samples that will be used to estimate the properties of the population. The definition of the population to be sampled is central to the initial formulation of the research study (Eberhardt and Thomas 1991; Pennock 2004). The sampling design defines how specific elements will be selected from the population, and these sampled elements form the sample population. There are many highly detailed guides to specific sampling designs and the statistical approaches appropriate for each design. The goal of this chapter is to present the issues that should be considered when selecting an appropriate sampling design. In the final section, specific design issues associated with particular research designs are covered. Suggested readings are given in each section for more in-depth study on each topic.

1



2

1.2 APPROACHES TO SAMPLING 1.2.1 HAPHAZARD, JUDGMENT,

AND

PROBABILITY SAMPLING

Sample locations can be chosen using (a) haphazard sampling, (b) judgment sampling, or (c) probability sampling. Haphazard, accessibility, or convenience sampling involves a series of nonreproducible, idiosyncratic decisions by the sampler and no systematic attempt is made to ensure that samples taken are representative of the population being sampled. This type of sampling is antithetical to scientific sampling designs. Judgment sampling (also termed purposive sampling [e.g., de Gruijter 2002]) involves the selection of sampling points based on knowledge held by the researcher. Judgment sampling can result in accurate estimates of population parameters such as means and totals but cannot provide a measure of the accuracy of these estimates (Gilbert 1987). Moreover the reliability of the estimate is only as good as the judgment of the researcher. Probability sampling selects sampling points at random locations using a range of specific sample layouts, and the probability of sample point selection can be calculated for each design. This allows an estimate to be made of the accuracy of the parameter estimates, unlike judgment sampling. This allows a range of statistical analyses based on the estimates of variability about the mean to be used, and is by far the most common type of sampling in soil science.

1.2.2 RESEARCH DESIGNS USING JUDGMENT SAMPLING Pedogenetic and soil geomorphic studies focus on determining the processes that formed the soil properties or landscapes under study and the environments that controlled the rates of these processes. Pedon-scale studies are closely associated with the development of soil taxonomic systems, and focus on vertical, intrapedon processes. Soil geomorphic studies are the interface between quaternary geology and soil science, and soil geomorphologists focus on lateral transfer processes and the historical landscape evolution. Both types of studies involve the identification of soil and=or sediment exposures that are highly resolved records of the sequence of processes that have formed the soil landscape. The researcher locates these exposures by using his judgment as to the landscape positions where optimum preservation of the soil–sediment columns is most likely. The development of the chronological sequence can be done with a detailed analysis of a single exposure; no replication of exposures is required. Surveys are designed to define the extent of spatial units. Soil surveyors map the distribution of soil taxonomic units and provide descriptive summaries of the main properties of the soils. In soil survey the association between soil classes and landscape units is established in the field by judicious selection of sampling points (termed the free survey approach). This type of judgment sampling can be an extremely efficient way of completing the inventory. Contaminant surveys are most typically undertaken by private-sector environmental consultants, and the specific objective may range from an initial evaluation of the extent of contamination to the final stage of remediation of the problem. Laslett (1997) states that consultants who undertake these surveys almost always employ judgment sampling and place their samples where their experience and prior knowledge of site history suggest the contamination might be located. In many jurisdictions the sampling design may also be constrained by the appropriate regulatory framework.


Soil Sampling Designs

3

1.2.3 RESEARCH DESIGNS USING PROBABILITY SAMPLING Inventory studies share the common goal of measuring the amount of a property or properties under study and the uncertainty surrounding our estimate of the amount. For example, in agronomic sampling we may wish to estimate the amount of plant-available nutrients in a given field; in contaminant sampling the goal may be to estimate the amount of a contaminant present at a site. In comparative mensurative experiments, comparisons are drawn among classes that the researcher defines but cannot control— for example, sampling points grouped by different soil textures, landform positions, soil taxonomic classes, and drainage class. Their location cannot be randomized by the researcher, unlike imposed treatments such as tillage type or fertilizer rates where randomization is essential. In manipulative experiments the treatments can be directly imposed by the researcher—ideally as fixed amounts that are applied precisely. Many studies are hybrid mensurative–manipulative designs—for example, the measurement of yield response to different fertilizer rates (imposed treatment) in different landform positions (characteristic or inherent treatment). The role of sampling in inventory, mensurative, and manipulative designs is very similar—to allow statistical estimation of the distribution of the parent population or populations. In inventory studies the statistical estimates may be the end point of the study. Pattern studies are undertaken to assess and explain the spatial or temporal pattern of properties. Two main types of pattern studies exist: (a) the quantification of the spatial and temporal variability in properties and (b) hypothesis generation and testing using point patterns. In pattern studies the initial goal may be a visual assessment of the pattern of observations in time or space, and statistical estimation of the populations may be a secondary goal. Geostatistical and other spatial statistical studies are undertaken to model the spatial pattern of soil properties, to use these models in the interpolation of values at unsampled locations, to assess the suitability of different spatial process models, or to assist in the design of efficient sampling programs.

1.3 STATISTICAL CONCEPTS FOR SAMPLING DESIGN 1.3.1 MEASURES OF CENTRAL TENDENCY AND DISPERSION The key characteristics of the distribution of attributes are measures of its central tendency and the dispersion of values around the measure of central tendency. In the initial stage of study formulation the researcher defines the population, which is composed of the sampling units and one or more attributes measured on these sampling units. Each attribute has a distribution of values associated with it, which can be characterized by parameters such as the population mean (m) and variance (s2 ). A sample of the sampling units is drawn from the population, and statistics such as the sample mean ( x) and variance (s2 ) are calculated, which serve as estimates of the population parameters. Calculations of these statistics are readily available and will not be repeated here. The number of samples taken is denoted as n. For sample populations that are more or less normally distributed the arithmetic mean ( x) is an appropriate measure of central tendency. The variance (s2 ) is a common measure of the deviation of individual values from the mean and its square root; the standard deviation (s)



4

reports values in the same units as the mean. The coefficient of variation (CV) is a normalized measure of the amount of dispersion around the mean, and is calculated by CV ¼ (s= x)100

(1:1)

Sample populations in the soil science commonly show a long tail of values to the right of the distribution (i.e., they are right-skewed). In this case a log normal or other right-skewed distribution should be used. The mathematical properties of the normal distribution are well understood and the probability that the true population mean lies within a certain distance of the sample mean can be readily calculated. For sample populations the estimated standard error of the sample mean is pffiffiffi s( x) ¼ s= n

(1:2)

For a sample population that has a large sample size or where the standard error is known and that approximates a normal distribution, the true mean will be within +1.96 standard errors of the sample mean 95 times out of 100 (i.e., where the probability P ¼ 0:05). The range defined these limits are the 95% confidence interval for the mean and these limits are the 95% confidence limits. The value 1.96 is derived from the t distribution, and values of t can be derived for any confidence limit. For sample populations based on a small sample size or where the standard error is not known the value of 1.96 must be replaced by a larger t-value with the appropriate degrees of freedom. A probability of exceeding a given standard error (a) may be selected for any sample distribution that approximates the normal distribution and the appropriate confidence limits calculated for that distribution.

1.3.2 INDEPENDENCE, RANDOMIZATION,

AND

REPLICATION

The goal of sampling is to produce a sample that is representative of the target population. If the choice of samples is not probability based then a strong likelihood exists that the sample will not be representative of the population. For example, selection of sampling locations convenient to a farmyard (instead of distributed throughout the field) may lead to overestimates of soil nutrients due to overapplication of farmyard manure near the source of the manure through time. The use of probability-based sampling designs (i.e., the designs discussed in Section 1.4) confers a design-specific independence on the sample selection process, which satisfies the need for independence of samples required by classical statistical analysis (a theme developed in great detail by Brus and de Gruijter 1997). Replication is an important consideration in mensurative and manipulative experiments. In a manipulative study, replication is the repeated imposition of a set of treatments (e.g., fertilizer or pesticide rates). In a pattern or mensurative study, replication is the repeated, unbiased selection and sampling of population elements in a particular class—for example, the selection of multiple 5 5 m slope elements in a field that have markedly convex downslope curvatures. Replication provides an estimate of the experimental error, and increasing replication improves precision by reducing the standard error of treatment or class means (Steel and Torrie 1980). Correct identification and sampling of replicates is critical for estimating the parameters of the class the sample is drawn from and is required for statistically correct procedures. Pseudoreplication (Hurlbert 1984) occurs when a researcher assumes a very general effect from a limited sampling and often occurs because the target population has not been clearly defined at the outset of the research.



5

Randomization is a consideration in manipulative designs. Steel and Torrie (1980, p. 135) summarizes the need for randomization: ‘‘ . . . it is necessary to have some way of ensuring that a particular treatment will not be consistently favored or handicapped in successive replications by some extraneous sources of variation, known or unknown. In other words, every treatment should have an equal chance of being assigned to any experimental unit, be it unfavorable or favorable.’’ Randomization is implemented by ensuring the random placement of treatment plots within a field design; the repeated imposition of the same sequence of treatments in a block of treatments may cause an erroneous estimate of the experimental error. The random order of treatment placement is achieved using random number tables or computer-generated randomizations.

1.4 SAMPLE LAYOUT AND SPACING Although many types of sampling designs exist (reviewed in Gilbert 1987; Mulla and McBratney 2000; de Gruijter 2002) only two main types (random and systematic) are commonly used in the soil and earth sciences. Inventory studies can be completed using any of the designs discussed in the following two sections. Pattern and geostatistical studies typically use transect or grid designs, as is discussed in more detail in Section 1.5.

1.4.1 SIMPLE RANDOM AND STRATIFIED RANDOM SAMPLING In simple random sampling all samples of the specified size are equally likely to be the one chosen for sampling. In stratified random sampling, points are assigned to predefined groups or strata and a simple random sample chosen from each stratum. The probability of being selected can be weighted proportionally to the stratum size or the fraction of points sampled can vary from class to class in disproportionate sampling. Disproportionate sampling would be used if the degree of variability is believed to vary greatly between classes, in which case a higher number of samples should be drawn from the highly variable classes to ensure the same degree of accuracy in the statistical estimates. Stratified sampling (correctly applied) is likely to give a better result than simple random sampling, but four main requirements should be met before it is chosen (Williams 1984): 1

Population must be stratified in advance of the sampling.

2

Classes must be exhaustive and mutually exclusive (i.e., all elements of the population must fall into exactly one class).

3

Classes must differ in the attribute or property under study; otherwise there is no gain in precision over simple random sampling.

4

Selection of items to represent each class (i.e., the sample drawn from each class) must be random.

The selection of random points in a study area has been greatly facilitated by the widespread use of Global Positioning System (GPS) receivers in field research. The points to be sampled can be randomly selected before going to the field, downloaded into the GPS unit, and then the researcher can use the GPS to guide them to that location in the field.



6

TABLE 1.1 Sample Sizes Required for Estimating the True Mean m Using a Prespecified Relative Error and the Coefficient of Variation Confidence level

Relative error, dr 10

20

40

50

100

150

0.80

0.10 0.25 0.50 1.0

2

7

27 6

42 7 2

165 27 7 2

370 60 15 4

0.90

0.10 0.25 0.50 1.0

2

12

45 9

70 12 2

271 45 13 2

609 92 26 8

0.95

0.10 0.25 0.50 1.0

4

17

63 12

97 17 4

385 62 16 9

865 139 35 16

Coefficient of variation (CV), %

Source: Adapted from Gilbert, R.O., in Statistical Methods for Environmental Pollution Monitoring, Van Nostrand, Reinhold, New York, 1987, 320 pp.

Determination of Sample Numbers in Inventory Studies A necessary and important step in the planning stages of a project is to determine the number of samples required to achieve some prespecified accuracy for the estimated mean. One approach is to use prior knowledge about the CV of the property under study to estimate sample numbers required to achieve a certain prespecified relative error. The relative error (dr ) is defined as dr ¼ jsample mean population meanj=population mean

(1:3)

The sample numbers required to achieve a specified relative error at a selected confidence level can be estimated from Table 1.1. For example, at a confidence level of 0.95 and a relative error of 0.25, 16 samples are required if the CV is 50% and 139 samples are required if the CV is 150%. Estimates of CV for different soil properties are widely available, and are summarized in Table 1.2.

1.4.2 SYSTEMATIC SAMPLING The most commonly used sampling design for many field studies is systematic sampling using either transects or grids. Systematic sampling designs are often criticized by statisticians but the ease with which they can be used and the efficiency with which they gather information makes them popular in the field of earth sciences. Ideally the initial point of the transect or grid and=or its orientation should be randomly selected. The major caution in the use of systematic sampling with a constant spacing is that the objects to be sampled must not be arranged in an orderly manner which might correspond to the spacing along the transect or the grid. The choice of a transect or a grid depends on several factors. Certain types of research designs require particular types of systematic designs—as discussed below, wavelet analysis requires long transects whereas geostatistical designs more typically use grid designs. Grids are often used for spatial pattern studies because of the ease with which pattern maps can be derived from the grids. The complexity of landforms at the site is also a consideration.



7

TABLE 1.2 Variability of Soil Properties Coefficient of variation Moderate (CV 15%–35%)

Low (CV 0:90, except for Ca and Al concentrations in LF, and Mn and C in LF and H horizons). Similarly, Bruckner et al. (2000) investigated the impact of bulking soil samples on microarthropod abundance on a Norway spruce plantation in Austria. It was assumed that the grinding action of soil particles during mixing would injure or kill part of the population and thus underestimate the population relative to a mean weighted from samples of the population analyzed individually. However, using a special mixing procedure of the extracts, Bruckner et al. (2000) came to the conclusion that no microarthropod was lost or damaged because a large number of samples were bulked in a systematic manner and mixed in equal amounts.

REFERENCES Arp, P.A. and Krause, H.H. 1984. The forest floor: lateral variability as revealed by systematic sampling. Canada. Can. J. Soil Sci. 64: 423–437. Aust, W.M., Tippett, M.D., Burger, J.A., and McKee, W.H. Jr. 1995. Compaction and rutting during harvesting affect better drained soils more than poorly drained soils on wet pine flats. South. J. Appl. Forest 19: 72–77.

A case study with Norway spruce. Can. J. Forest Res. 34: 560–572. Bock, M.D. and Van Rees, K.C.J. 2002. Forest harvesting impacts on soil properties and vegetation communities in the Northwest Territories. Can. J. Forest Res. 32: 713–724.

Beatty, S.W. and Stone, E.L. 1986. The variety of soil microsites created by tree falls. Can. J. Forest Res. 16: 539–548.

Bruckner, A., Barth, G., and Scheibengraf, M. 2000. Composite sampling enhances the confidence of soil microarthropod abundance and species richness estimates. Pedobiologia 44: 63–74.

Be´langer, N., Pare´, D., Bouchard, M., and Daoust, G. 2004. Is the use of trees showing superior growth a threat to soil nutrient availability?

Cade-Menun, B.J., Berch, S.M., Preston, C.M., and Lavkulich, L.M. 2000. Phosphorus forms and related soil chemistry of Podzolic soils on

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C002 Final Proof page 23 9.6.2007 10:51am

Sampling Forest Soils northern Vancouver Island. I. A comparison of two forest types. Can. J. Forest Res. 30: 1714–1725. Carter, R.E. and Lowe, L.E. 1986. Lateral variability of forest floor properties under secondgrowth Douglas-fir stands and the usefulness of composite sampling techniques. Can. J. Forest Res. 16: 1128–1132. Clinton, B.D. and Baker, C.R. 2000. Catastrophic windthrow in the southern Appalachians: characteristics of pits and mounds and initial vegetation responses. Forest Ecol. Manag. 126: 51–60. Coˆte´, B., Hendershot, W.H., Fyles, J.W., Roy, A.G., Bradley, R., Biron, P.M., and Courchesne, F. 1998. The phenology of fine root growth in a maple-dominated ecosystem: relationships with some soil properties. Plant Soil 201: 59–69. Dijkstra, F.A. and Smits, M.M. 2002. Tree species effects on calcium cycling: the role of calcium uptake in deep soils. Ecosystems 5: 385–398.

Compositor Name: BMani

23 Hamilton, W.N. and Krause, H.H. 1985. Relationship between jack pine growth and site variables in New Brunswick plantations. Can. J. Forest Res. 15: 922–926. Ike, A.F. and Clutter, J.L. 1968. The variability of forest soils of the Georgia Blue Ridge Mountains. Soil Sci. Soc. Am. Proc. 32: 284–288. Klinka, K., Green, R.N., and Trowbridge, R.L. 1981. Taxonomic classification of humus forms in ecosystems of British Columbia. First approximation. British Columbia Ministry of Forest, 53 pp. Kulmatiski, A., Vogt, D.J., Siccama, T.G., and Beard, K.H. 2003. Detecting nutrient pool changes in rocky forest soils. Soil Sci. Soc. Am. J. 67: 1282– 1286. Lichter, J.M. and Costello, L.R. 1994. An estimation of volume excavation and core sampling techniques for measuring soil bulk density. J. Arboric. 20: 160–164.

Eriksson, H.M. and Rosen, K. 1994. Nutrient distribution in a Swedish tree species experiment. Plant Soil 164: 51–59.

McFee, W.W. and Stone, E.L., 1965. Quantity, distribution and variability of organic matter and nutrients in a forest podzol in New York. Soil Sci. Soc. Am. Proc. 29: 432–436.

Expert Committee on Soil Survey. 1987. The Canadian System of Soil Classification, 2nd edn. Agriculture Canada Publ. 1646, Supplies and Services, Ottawa, Canada, 164 pp.

Olsson, B.A., Bengtsson, J., and Lundkvist, H. 1996. Effects of different forest harvest intensities on the pools of exchangeable cations in coniferous forest soils. Forest Ecol. Manag. 84: 135–147.

Finzi, A.C., van Breemen, N., and Canham, C.D. 1998. Canopy tree–soil interactions within temperate forests: species effects on soil carbon and nitrogen. Ecol. Appl. 8: 440–446.

Page-Dumroese, D.S., Jurgensen, M.F., Brown, R.E., and Mroz, G.D. 1999. Comparison of methods for determining bulk densities of rocky forest soils. Soil Sci. Soc. Am. J. 63: 379–383.

Green, R.N., Trowbridge, R.L., and Klinka, K. 1993. Towards a taxonomic classification of humus forms. Forest Science Monograph 29. Society of American Foresters, Bethesda, MD, 50 pp.

Palmer, C.J., Smith, W.D., and Conkling, B.L. 2002. Development of a protocol for monitoring status and trends in forest soil carbon at a national level. Environ. Pollut. 116: S209–S219.

Grier, C.C. and McColl, J.G. 1971. Forest floor characteristics within a small plot in Douglas-fir in Western Washington. Soil Sci. Soc. Am. Proc. 35: 988–991.

Parfitt, R.L., Percival, H.J., Dahlgren, R.A., and Hill, L.F. 1997. Soil and solution chemistry under pasture and radiata pine in New Zealand. Plant Soil 191: 279–290.

Hamel, B., Be´langer, N., and Pare´, N. 2004. Productivity of black spruce and jack pine stands in Quebec as related to climate, site biological features and soil properties. Forest Ecol. Manag. 191: 239–251.

Pennock, D.J. 2004. Designing field studies in soil science. Can. J. Soil Sci. 84: 1–10. Powers, R.F. 1991. Are we maintaining productivity of forest lands? Establishing guidelines

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C002 Final Proof page 24 9.6.2007 10:51am

24 through a network of long-term studies. In A.E. Harvey and L.F. Neuenschwander, compilers. Proceedings—Management and Productivity of Western Montane Forest Soils. General Technical Report. INT-GTR-280. April 10–12, 1990, USDA Forest Service, Washington, DC, pp. 70–89. Quesnel, H.J. and Lavkulich, L.M. 1980. Nutrient availability of forest floors near Port Hardy, British Columbia, Canada. Can. J. Soil Sci. 60: 565–573.

Compositor Name: BMani

Soil Sampling and Methods of Analysis Snedecor, G.W. and Cochran, W.G. 1980. Statistical Methods, 7th edn. Iowa State University Press, Ames, AI, 507 pp. Van Wesemael, B. and Veer, M.A.C. 1992. Soil organic matter accumulation, litter decomposition and humus forms under Mediterranean-type forests in southern Tuscany, Italy. J. Soil Sci. 43: 133–144.

E.G. Gregorich/Soil Sampling and Methods of Analysis, Second Edition 3586_C003 Final Proof page 25

9.6.2007 11:53am Compositor Name: BMani

Chapter 3 Measuring Change in Soil Organic Carbon Storage B.H. Ellert and H.H. Janzen Agriculture and Agri-Food Canada Lethbridge, Alberta, Canada

A.J. VandenBygaart Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

E. Bremer Symbio Ag Consulting Lethbridge, Alberta, Canada

3.1 INTRODUCTION Organic carbon (C) must be among the most commonly analyzed soil constituents, starting with the earliest soil investigations. Already in the nineteenth century, chemists were routinely analyzing soil C (e.g., Lawes and Gilbert 1885). Initially, these analyses were done to investigate pedogenesis and to assess soil productivity, both of which are closely linked to organic C (Gregorich et al. 1997). But more recently, scientists have been analyzing soil organic C (SOC) for another reason: to measure the net exchange of C between soil and atmosphere (Janzen 2005). Indeed, building reserves of SOC has been proposed as a way of slowing the rising atmospheric CO2 concentrations caused by burning fossil fuel (Lal 2004a,b). Measuring SOC to quantify soil C ‘‘sinks’’ requires more stringent sampling and analyses than measuring SOC to evaluate productivity. Where once it was sufficient to measure relative differences in concentration over time or among treatments, now we need to know the change in amount of C stored in Mg C per ha. Reviews of SOC measurement typically focus on the chemical methods of determining the SOC concentrations after samples have been brought to the laboratory. Here we emphasize soil sampling procedures and calculation approaches to estimate temporal changes in SOC stocks. Uncertainties along the entire chain of procedures, from designing the soil sampling plan, to sampling in the field, to processing and storing the samples, through to chemical analysis and calculating soil C stocks need to be considered (Theocharopoulos et al. 2004).

25




26

SOC is dynamic: newly photosynthesized C is added regularly in the form of plant litter, and existing SOC is gradually decomposed back to CO2 by soil biota. Management or environmental conditions that change the relative rates of inputs and decomposition will effect a change in the amount of SOC stored. Rates of change in SOC (typically less than 1 0:5 Mg C ha1 year ) are quite small, however, compared to the large amounts of SOC often present (as high as 100 Mg C ha1 , or more, in the top 30 to 60 cm soil layer). Thus changes in SOC can only be reliably measured over a period of years or even decades (Post et al. 2001). Since the distribution of SOC in space is inherently variable, temporal changes (e.g., attributable to management practices, environmental shifts, successional change) must be distinguished from spatial ones (e.g., attributable to landform, long-term geomorphic processes, nonuniform management). Temporal changes in SOC can be defined in two ways (Figure 3.1): as an absolute change in stored C (SOC at t ¼ x minus SOC at t ¼ 0), or as a net change in storage among treatments (SOC in treatment A minus SOC in treatment B, after x years). The former provides an estimate of the actual C exchange between soil and atmosphere; the latter provides an estimate of the C exchange between soil and atmosphere, attributable to treatment A, relative to a control (treatment B). Both expressions of temporal change may be available from manipulative experiments with appropriate samples collected at establishment (assesses spatial variability) and at various intervals (say 5 to 10 years) thereafter.

Soil organic C (Mg C ha−1)

This chapter provides selected methods for measuring the change in C storage, either absolute or net, typically for periods of 5 years or more. To be effective, the method needs to: measure organic (not total) C, provide estimates of C stock change (expressed in units of C mass per unit area of land to a specified soil depth and mass), be representative of the land area or management treatment under investigation, and provide an indication of confidence in the measurements. These methods are applicable, with minor modification, to a range of scales and settings, including benchmarks sites and replicated research experiments.

A B

Net change in treatment A Absolute change in treatment A

x

0 Time (years)

FIGURE 3.1. Illustration of hypothetical changes in soil organic C in two treatments, A and B. For treatment A, the absolute change is the difference in SOC at time ¼ x, compared to that at time ¼ 0. The net change is the difference between SOC in treatment A and that in treatment B, at time ¼ x, assuming that SOC was the same in both treatments at time ¼ 0. The latter approach is often used to measure the effect on SOC of a proposed treatment (e.g., no-till) compared to a standard ‘‘control’’ (e.g., conventional tillage).



Measuring Change in Soil Organic Carbon Storage

27

3.2 SELECTING THE SAMPLING LOCATIONS AND PATTERN Determining the optimum number and spatial arrangement of sampling points to estimate SOC storage remains as much an art as a science. Nevertheless, careful study of the site, along with clearly articulated objectives can improve the cost-effectiveness and precision of the estimates (VandenBygaart 2006).

3.2.1 MATERIALS 1

Descriptions of soil properties, landscape characteristics, and agronomic history at the study site, from sources such as: soil maps and reports, aerial photos, scientific publications, cropping records, and yield maps.

3.2.2 PROCEDURE Two general approaches can be used in sampling a study area (e.g., a plot, field, watershed): a Nonstratified sampling, where the entire study area is considered to be one unit, and sampled in a systematic or random manner. b Stratified sampling, where the study area is first subdivided into relatively homogeneous units, based on factors such as topography (e.g., slope position), and each unit is sampled separately.

3.2.3 NONSTRATIFIED SAMPLING 1

Obtain an estimate of the likely sample variance and required accuracy for SOC at the study site, based on previously compiled information.

2

Using as much information as available, calculate the number of samples required using Equation 3.1. The required number of samples will increase as variability and the required accuracy increase (Figure 3.2) (Garten and Wullschleger 1999; Wilding et al. 2001). Required accuracy is expressed as in the same units used for the sample mean, and often is less than 10% of that value because even small changes in the mean imply appreciable pedosphere–atmosphere C exchange over large tracts of land.

3

Select an appropriate grid or linear sampling pattern, suited to the study site and sampling equipment.

3.2.4 STRATIFIED SAMPLING 1

Subdivide the study site into areas likely to have similar SOC stocks, based on factors such as topography or management history.

2

Select the number of sampling sites within each subarea, using Equation 3.1, or Figure 3.2 as a guide, or by fixed allotment. In the latter case, for example, one or several sampling sites may be designated for each of three slope positions within a large research plot.




28 30 Minimum detectable difference (MDD), Mg C ha–1

σ 2 = 4, cv = 5% σ 2 = 17, cv = 10%

25

σ 2 = 34, cv = 15% σ 2 = 67, cv = 20%

20

σ 2 = 100, cv = 25% 15

10

5

0 1

10

100

1000

Number of soil samples, n

FIGURE 3.2. Decrease in the minimum detectable difference (MDD) between mean soil C at two sampling times for contrasting levels of variance as the number of samples collected at each time doubles (4, 8, 16, . . .). The MDD was calculated for a ¼ 0.05 significance and (1b) ¼ 0.90 statistical power (i.e. probability of rejecting the null hypothesis when it really is false and should be rejected). The lines correspond to increasing variance (s2) selected for a hypothetical soil layer containing a mean of 40 Mg C ha1 with the coefficient of variation (cv) increasing from 5% to 25%. (Adapted from Garten, C.T. and Wullschleger, S.D., J. Environ. Qual., 28, 1359, 1999. With permission.)

3.2.5 CALCULATIONS nreq ¼

t 2 s2 (d mean)2

(3:1)

where nreq is the required number of samples, t is the Student’s t-value, at the desired confidence level (typically 1a ¼ 0:90 or 0.95), s2 is the sample variance, d is the required accuracy or maximum acceptable deviation from the mean (e.g. d ¼ 0.10), and mean is the arithmetic sample mean.

3.2.6 COMMENTS Sampling patterns and intensities will vary widely, depending on site characteristics and on other factors, notably economic considerations. Often, the number of samples required to achieve the desired sensitivity is exceedingly expensive, and the number of sampling points is somewhat arbitrarily reduced. As well, sampling intensity may have to be reduced in small plots, such as long-term experiments, where excessive soil removal may disturb the site to the extent that future research is jeopardized. But such compromises, if carried too far, may reduce the chance of measuring any differences with reasonable reliability. Studies with insufficient sampling points typically lack statistical power to assess treatment effects. Consequently, the ‘‘cost’’ of erroneous conclusions drawn from such data (when the data really are inconclusive) may greatly exceed the ‘‘savings’’ provided by reduced sample numbers.



Measuring Change in Soil Organic Carbon Storage

29

Precisely measuring temporal changes in SOC first depends on identifying or minimizing spatial changes. Spatial changes can be minimized by pairing sampling locations in space (Ellert et al. 2001, 2002; VandenBygaart 2006). This approach allows for effective measurement of SOC changes in time at comparatively few sampling points, but measured C stock change values at these points are not necessarily representative of the entire study site. Conant and Paustian (2002) and Conant et al. (2003) have evaluated similar sampling strategies.

3.3 EXTRACTING AND PROCESSING SOIL CORES The following procedure is intended for the extraction of soil cores, from agricultural plots or landscapes, for subsequent organic C analysis. It is provided as an illustration, recognizing that individual studies may require modification to satisfy specific objectives and local conditions.

3.3.1 MATERIALS 1

Truck-mounted hydraulic soil coring device.

2

Soil coring tube, with slots 1 cm wide by 30 cm long, and a cutting bit with inside diameter of about 7 cm. The bit usually has slightly smaller diameter (by 1 to 4 mm) than the tube; this difference should be small enough to avoid soil mixing, but large enough to prevent sticking. In dry, coarse-textured soils with weak consolidation this difference should be reduced so there is enough friction to hold the core when the tube is pulled from the soil. The diameter of the coring bit should be measured accurately and recorded for future calculations of soil core density.

3

Piston to push the soil core out of tube. A simple piston can be constructed by attaching a rubber stopper to the end of a wooden dowel.

4

Knife, steel ruler, scissors, wire brush.

5

Aluminum foil trays ( 24 30 6 cm, used in steam tables for serving food), coolers for transporting trays from field, and heavy polyethylene bags ( 30 50 cm) to contain trays of field-moist soil.

6

Analytical balance (3000 g capacity, resolution to 0.01g), moisture tins (8 cm diameter 6 cm tall), drying oven (1058C).

7

Paper ‘‘coffee’’ bags with plastic lining and attached wire ties (e.g., Zenith Specialty Bag Co., 11 6 cm base 23 cm height).

8

‘‘Rukuhia’’ perforated drum grinder, with 2 mm perforations (Waters and Sweetman 1955); or another coarse soil grinder and a 2 mm soil sieve.

9

Equipment to measure soil sampling locations. This may be a simple surveyor’s tape to measure locations relative to permanent marker stakes in long-term field experiments, or a Global Positioning System (GPS) receiver. For precise pairing (in space) of samples collected at sequential time intervals of several years, a twostage measuring approach may be useful: the general location is measured relative to permanent reference points or is recorded using a simple GPS receiver,




30

and the position of the initial cores is marked by burying an electromagnetic marker originally developed to identify underground utilities (Whitlam 1998). Alternatively, high-resolution GPS is available in many regions.

3.3.2 PROCEDURE 1

Before sampling, label paper bags with name, sampling date, location, and soil depth. These bags, eventually to be used for storing the air-dried soils, also serve as labels throughout the sampling process. Weigh the aluminum trays, one for each sample, and record the weight on the tray.

2

In the field, for each sampling point, lightly brush away surface residue and extract a core to a depth of at least 60 cm. Move the core from the vertical to a horizontal position (e.g., in a sectioning trough made of 10 to 15 cm diameter pipe cut lengthwise), and measure the depths of any visible discontinuities (e.g., depth of Ap horizon). Be prepared to discard cores that are unrepresentative (e.g., excessively compacted during sampling, evidence of atypical rodent activity, gouged by a stone pushed along the length of the core during sampling). It may prove useful to push the core (from the deepest end) out in increments, using the top end of the tube as a guide to make perpendicular cuts. Cut the core into carefully measured segments (for example: 0 to 10, 10 to 20, 20 to 30, 30 to 45, and 45 to 60 cm), and place segments into aluminum trays, avoiding any loss of soil. Repeat the procedure for a second core, about 20 cm apart, and composite with the first core segments. Place aluminum trays inside a polyethylene bag, along with the labeled paper bag, fold over polyethylene bag, and store in cooler before subsequent processing indoors.

3

In the laboratory, remove aluminum trays from the polyethylene bags and air-dry at room temperature. Except for very sandy soils, it will be much easier to grind the soils if the field-moist soil cores are broken apart by hand before air drying and subsequent grinding. Great care is required to avoid sample losses during processing and contamination by dust, plant material, paper, or other C-rich contaminants during drying. Wear rubber gloves when handling soil to avoid contamination.

4

Once samples are air-dry, record weight of sample þ aluminum tray. Remove a small, representative subsample (e.g., 50 to 80 g, excluding stones and large pieces of plant residue), and determine air-dry moisture content by oven-drying for 48 h at 1058C. Alternatively, the weights of field-moist cores plus trays may be recorded immediately after removal from the polyethylene bag and before they are broken apart and air-dried. In this case, accurate field moisture contents are crucial to estimate the densities of core segments, but spillage when cores are broken apart and mixed may be less consequential than the case when cores are dried before weighing. Thoroughly mix soils before subsampling to determine field moisture content and possibly to retain a field-moist subsample for biological analyses.

5

Crush or grind entire samples to pass a 2 mm sieve, and screen out gravel >2 mm in diameter. All organic material in the sample should be included; if necessary, separately grind roots and other large organic debris to 2 mm in diameter. 6

Place coarsely ground samples in labeled ‘‘coffee’’ bags for storage under cool, dry conditions, before analysis. For permanent storage (longer than 1 year), soil samples should be placed in sealed glass or plastic jars, and kept under cool, dry, and dark conditions. If finely ground soil is required (e.g., for elemental microanalysis), the coarsely ground (358C until used (Vaden et al. 1987).

6

ATP standard stock solution is prepared by reconstituting a vial containing ~1 mg (2 mM) of ATP disodium hydrate (formula weight 551.1) (Sigma-Aldrich) in 10 mL of cold water, giving a solution that is ~0.2 mM ATP (see label on vial for actual amount). This standard stock solution can be divided into 0.5 mL portions (contained in 4 mL plastic vials with lids) and, if not immediately used, kept frozen (158C) for later use.

7

The ATP standard calibration curve relating ATP concentrations to light intensity expressed in relative light units (RLU) is prepared in ATP assay mix dilution buffer. Typically, it should span an ATP assay concentration range from 0 to 0.60 pM ATP assay1 when added as a component to the reaction mixture. Prepare an ATP working standard for establishing the standard curve by taking a 0.1 mL aliquot of the ATP standard stock solution and bring up to 10 mL with cold water. Take a 0.1 mL aliquot of this solution and bring to 10 mL with ATP assay mix dilution buffer. This ATP working standard, containing 20 pM ATP mL1 , is thoroughly mixed, and kept on ice. Prepare 1 mL of each of the ATP standards using the ATP working standard solution and the ATP assay mix dilution buffer as shown in Table 41.1. These standards are substituted for the pure ATP assay mix dilution buffer (100 mL) addition in the reaction mixture. A plot of the logarithm of RLU generated by the luciferin–luciferase reaction vs. the logarithm of the ATP standard concentration gives a linear relationship.

8

The ATP spiking solution for measurement of the recovery efficiency is prepared by taking a 0.3 mL aliquot of the ATP standard stock solution and adding, while mixing, to 0.3 mL with HEPES buffer. This ATP spiking solution contains 0:10 mM ATP mL1 HEPES and is kept on ice.



550

TABLE 41.1 Construction of the ATP Standard Calibration Curve

9

ATP standard concentration (pM ATP assay1 )

ATP working standard (mL)

ATP assay mix dilution buffer (mL)

0 0.1 0.2 0.3 0.4 0.5 0.6

0 50 100 150 200 250 300

1000 950 900 850 800 750 700

An autoclaved soil extract for preparing the ATP standard calibration curve is prepared by weighing 5 g soil (oven-dry weight) into a 100 mL glass centrifuge tube and immediately autoclaving it at 1218C for 20 min. After cooling, the tube is capped and kept frozen until used. After thawing, the soil is extracted as described below for fresh soil.

41.2.3 EXTRACTION PROCEDURE 1

Triplicate 5 g (dry weight) portions of fresh soil are weighed into 100 mL glass centrifuge tubes (equipped with screw caps) to which is added 50 mL of extractant, while gently shaking, and placed in an ice bath.

2

ATP recovery efficiency is measured by adding 50 mL of the ATP spiking solution (containing 5 nM ATP) to a tube containing 5 g soil þ 50 mL extractant (also gently shaking and kept in an ice bath). A soil control is prepared by adding 50 mL HEPES buffer alone to a tube containing 5 g soil þ 50 mL extractant.

3

The soil plus extractant is homogenized and sonicated for 1 min at a setting of 7 using a Brinkmann Polytron Homogenizer equipped with a stainless steel tip (PTA 20S, Kinematica), or for 2 min at full power using a 20 kHz 140 W Branson Sonifier (Model 200) equipped with a 12.5 mm diameter probe. During sonication, the centrifuge tube is kept cooled in an ice bath and covered with parafilm to prevent losses by splashing. (Between soil samples, the tip is carefully washed with water, with the machine turned on for a brief period, and dried.) The centrifuge tubes are capped and shaken on a wrist-action shaker (180 strokes min1 ) for 30 min, after which the tubes are centrifuged for 20 min at 31,000 g and 48C.

4

Duplicate 0.2 mL aliquots of the supernatant from the soil sample extract are transferred to 4 mL glass sample cups and neutralized by addition of 3.2 mL Tris buffer.

5

The same procedure is repeated for neutralization of the autoclaved soil extract (or the extractant alone) for use in analysis of ATP standards.

6

The neutralized extracts should be either kept at 48C and analyzed immediately or immediately frozen and stored at 158C until ATP determinations can be


Soil ATP

551

carried out. Frozen extracts are thawed immediately before measurement and kept at 48C.

41.2.4 ASSAY PROCEDURE 1

The final reaction mixture for the assay (325 mL in total) is prepared by adding to a cuvette the following, in order: (i)

100 mL of water;

(ii) 100 mL of pure ATP assay mix dilution buffer; (iii) 25 mL of neutralized soil extract; (iv) 100 mL of luciferin–luciferase solution; the cuvette is immediately placed in the instrument for counting. 2

For construction of the ATP standard calibration curve, the final reaction mixture for the assay (325 mL in total) is prepared by adding to a cuvette the following, in order: (i)

100 mL of water;

(ii) 100 mL ATP assay mix dilution buffer containing the ATP standards; (iii) 25 mL of neutralized extract of autoclaved soil (preferable) or neutralized extractant; (iv) 100 mL of luciferin–luciferase solution; the cuvette is immediately placed in the instrument for counting. 3

The most reliable method of determining ATP concentrations is with light detecting instruments, such as a commercially available photometer or a liquid scintillation counter (operated in noncoincident mode) set to an integral counting mode. If a luminometer (Lumac Model 1070, Lumac Systems Inc. P.O. Box 2805, Titusville, FL 32780, USA) is used, it should be turned on for 30 min prior to measurements to allow the electronics to stabilize. The delay time is set to 5 s to avoid an error reading of the immediate light flash, produced during mixing of luciferin– luciferase with ATP. The counting time integration setting is 10 s. Determinations of ATP in each extract are carried out in duplicate. Periodic standardization should be performed during the day to check for instrument stability. If the grid electrical supply is highly variable, it is advisable to install a UPS power filter to ensure greater instrument stability.

41.3 CALCULATION OF SOIL ATP CONTENT 41.3.1 ASSAY ATP CONCENTRATION ATP content in each assay is obtained using the equation of the standard calibration curve relating the logarithm of the assay RLU vs. the logarithm of the ATP standard concentrations (expressed in pM ATP assay1 ).



552

41.3.2 SOIL ATP RECOVERY EFFICIENCY A measure of the soil ATP recovery efficiency is made from the measurement of the spike ATP recovered in the assay using the following two equations: Recovery efficiency (RE) ¼

Spike ATP measured in assay ATP added in spiked sample assay

(41:1)

where Spike ATP measured in assay ¼ (ATP in soil þ spike assay) ATP in soil assay (41:2) ATP added in spike assay ¼ 0.1469 pM ATP.

41.3.3 SOIL ATP CONTENT IN

THE

ASSAY

The soil ATP content in the assay is determined from the measured assay ATP content and the recovery efficiency using the following equation: Soil ATP in the assay (pM assay1 ) ¼

Measured assay ATP content RE

(41:3)

41.3.4 SOIL ATP CONTENT Soil ATP content is calculated from the soil ATP in the assay, the volume of neutralized extract (NE) assayed (25 mL), the quantity of neutralized extract (3.4 mL), the quantity of soil extract (SE) neutralized (0.2 mL), the quantity of soil extract (50 mL extractant þ soil water content), and the quantity of soil extracted (5 g) using the following equation: Soil ATP (pM g1 soil) ¼

Soil ATP (pM) 3:4 mL NE (50 þ SW) mL SE 0:025 mL NE assay 0:2 mL SE 5 g soil

(41:4)

where SW is the water in the soil sample calculated using the following equation: Soil water (SW) (mL) ¼ 5 g (oven-dry soil) gravimetric soil water content (%) (41:5) Results are typically expressed as nM g1 soil: Soil ATP (nM g1 soil) ¼

Soil ATP (pM g1 soil) 1000

(41:6)

41.4 COMMENTS 1

Most of the reagents used, except where noted, are readily obtainable and are certified analytical grade. Deionized water (Type 1) should be used throughout for preparation of the reagent solutions.

2

Luciferase activity and the wavelength of light emitted are pH sensitive, therefore, neutralization of the soil extract and adjustment of the reaction


Soil ATP

553

mixture to an optimal pH are critical for maximizing assay sensitivity (Wen et al. 2005). 3

An autoclaved soil extract should be used for preparing the ATP standard calibration curve to ensure a similar chemistry in the ATP standards as that of the soil extract.

4

The composition of the reaction mixture gives precise control of the reaction mixture pH (pH should be ~7.75) while diluting components in the extract that may affect enzyme activity.

5

It is critical that all glassware be very clean and that new gloves be worn to avoid contamination of the samples with foreign ATP. Undiluted, these reagents are able to detect concentrations as low as 0.002 pM ATP mL1 .

6

Soil ATP measurements using this method are relatively precise; three replicate determinations on the same soil sample should be able to detect differences in ATP content of 5%–10% at a 0.05% level of probability.

41.5 ATP AND MICROBIAL BIOMASS The soil ATP method is an extremely sensitive method for studying soil microbial biomass and its activity. Because the method can be carried out quickly, it has potential for use in studying soils under rapidly changing environmental conditions (e.g., wetting and drying, freezing, and thawing). It can also be used for samples derived from select soil microhabitats such as aggregate surfaces and intra-aggregate spaces within the soil matrix or from the rhizosphere (Nannipieri et al. 1990). Further information about the average physiological state of the soil microbial biomass can be obtained by measurements of the adenylate energy charge (AEC) (Brookes et al. 1987; Vaden et al. 1987; Raubuch et al. 2002; Joergensen and Raubuch 2003; Raubuch et al. 2006). The AEC is a relation between the concentrations of the adenine nucleotides, according to the following equation: Adenylate energy charge (AEC) ¼

[ATP] þ 0:5 [ADP] [ATP] þ [ADP] þ [AMP]

(41:7)

where ATP, ADP, and AMP are the soil concentrations (in mM g1 soil) of adenosine triphosphate, adenosine diphosphate, and adenosine monophosphate, respectively. AEC values of 0.7–0.95 have been reported for fresh soils and 0.4–0.5 for air-dried soils. Further efficiencies in analysis can be gained by measuring the content of all three adenylates in one step using ion-paired reverse-phase high-performance liquid chromatography (HPLC) techniques (Dyckmans and Raubuch 1997).

REFERENCES Brookes, P.C., Newcombe, A.D., and Jenkinson, D.S. 1987. Adenylate energy charge measurements in soil. Soil Biol. Biochem. 19: 211–217.

Chander, K.C., Dyckmans, J., Joergensen, R.G., Meyer, B., and Raubuch, M. 2001. Different sources of heavy metals and their long-term


554


effects on soil microbial properties. Biol. Fert. Soils 34: 241–247.

soils differing in the intensity of disturbance. Soil Biol. Biochem. 35: 1161–1164.

Ciardi, C. and Nannipieri, P. 1990. A comparison of methods for measuring ATP in soil. Soil Biol. Biochem. 22: 725–727.

Karl, D.M. and LaRock, P.A. 1975. Adenosine triphosphate measurements in soil and marine sediments. J. Fish. Res. Board Can. 5: 599–607.

Contin, M., Jenkinson, D.S., and Brookes, P.C. 2002. Measurement of ATP in soil: correcting for incomplete recovery. Soil Biol. Biochem. 34: 1381–1383.

Maire, N. 1984. Extraction de l’adenosine triphosphate dans les sols: une nouvelle methode de calcul des pertes en ATP. Soil Biol. Biochem. 16: 361–366.

De Nobili, M., Diaz-Ravinã, M., Brookes, P.C., and Jenkinson, D.S. 1996. Adensoine 50 -triphosphate measurements in soils containing recently added glucose. Soil Biol. Biochem. 28: 1099–1104.

Martens, R. 2001. Estimation of ATP in soil: extraction methods and calculation of extraction efficiency. Soil Biol. Biochem. 33: 973–982.

Dilly, O. and Nannipieri, P. 2001. Response of ATP content, respiration rate, and enzyme activities in an arable and a forest soil to nutrient additions. Biol. Fert. Soils 34: 64–72. Dyckmans, J., Chander, K., Joergensen, R.G., Priess, J., Raubuch, M., and Sehy, U. 2003. Adenylates as an estimate of microbial biomass C in different soil groups. Soil Biol. Biochem. 35: 1485–1491. Dyckmans, J. and Raubuch, M. 1997. A modification of a method to determine adenosine nucleotides in forest organic layers and mineral soils by ion-paired reverse-phase high performance liquid chromatography. J. Microbiol. Methods 30: 13–20. Eiland, F. 1983. A simple method for quantitative determination of ATP in soil. Soil Biol. Biochem. 15: 665–670. Jenkinson, D.S. 1988. Determination of microbial biomass carbon and nitrogen in soil. In J.R. Wilson, Ed. Advances in Nitrogen Cycling in Agricultural Ecosystems. CAB International, Wallingford, UK, pp. 368–386. Jenkinson, D.S., Davidson, S.A., and Powlson, D.S. 1979. Adenosine triphosphate and microbial biomass in soil. Soil Biol. Biochem. 11: 521–527. Joergensen, R.G. and Raubuch, M. 2002. Adenylate energy charge of a glucose-treated soil without adding a nitrogen source. Soil Biol. Biochem. 34: 1317–1324. Joergensen, R.G. and Raubuch, M. 2003. Adenylate energy charge and ATP-to-biomass C ratio in

Nannipieri, P., Grego, S., and Ceccanti, B. 1990. Ecological significance of the biological activity in soil. In J.M. Bollag and G. Stotzky, Eds. Soil Biochemistry, Vol. 6. Marcel Dekker, New York, NY, pp. 293–355. Raubuch, M., Campos, A., and Joergensen, G.R. 2006. Impact of cycloheximide addition on adenylates in soil. Soil Biol. Biochem. 38: 222–228. Raubuch, M., Dyckmans, J., Joergensen, R.G., and Kreutzfeldt, M. 2002. Relation between respiration, ATP content, and adenylate energy charge (AEC) after incubation at different temperatures and after drying and rewetting. J. Plant Nutr. Soil Sci. 165: 435–440. Renella, G., Reyes Ortigoza, A.L., Landi, L., and Nannipieri, P. 2003. Additive effects of copper and zinc on cadmium toxicity to phosphatase activities and ATP content of soil as estimated by the ecological dose (ED50 ). Soil Biol. Biochem. 35: 1203–1210. Shannon, D., Sen, A.M., and Johnson, D.B. 2002. A comparative study of the microbiology of soils managed under organic and conventional regimes. Soil Use Manage. 18: 274–283. Tate, K.R. and Jenkinson, D.S. 1982. Adenosine triphosphate (ATP) and microbial biomass in soil: effects of storage at different temperature and different moisture levels. Commun. Soil Sci. Plant Anal. 13: 899–908. Vaden, V.R., Webster, J.J., Hampton, G.J., Hall, M.S., and Leach, F.R. 1987. Comparison of methods for extraction of ATP from soil. J. Microbiol. Methods 7: 211–217.


Soil ATP

555

Verstraete,W., Van de Werf, H., Kucnerowicz, F., Ilaiwi, M., Verstraeten, L.M.J., and Vlassak, K. 1983. Specific measurement of soil microbial ATP. Soil Biol. Biochem. 15: 391–396.

Wen, G., Voroney, R.P., Curtin, D., Schoenau, J.J., Qian, P.Y., and Inanaga, S. 2005. Modification and application of a soil ATP determination method. Soil Biol. Biochem. 37: 1999–2006.

Webster, J.J., Hampton, G.J., and Leach, F.R. 1984. ATP in soil: a new extractant and extraction procedure. Soil Biol. Biochem. 16: 335–342.

Wen, G., Voroney, R.P., Schoenau, J.J., Yamamoto, T., and Chikushi, J. 2001. Assessment of ionic quenching on soil ATP bioluminescence reaction. Soil Biol. Biochem. 33: 1–7.


E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C042 Final Proof page 557 9.6.2007 11:58am Compositor Name: BMani

Chapter 42 Lipid-Based Community Analysis K.E. Dunfield University of Guelph Guelph, Ontario, Canada

42.1 INTRODUCTION Soil microbial communities are incredibly complex, with estimates of more than 4000 bacterial genomes in a single soil sample (Torsvik et al. 1990; Amann et al. 1995). However, evidence has shown that less than 1% of soil bacteria are cultivable using common laboratory media under standard conditions (Torsvik et al. 1990; Ovreas and Torsvik 1998). Therefore, it is important to choose a method of community analysis that does not rely on isolation and cultivation techniques. Methods that extract cellular components that are representative of most bacterial species directly from soils have become popular. These methods are based on the characterization of bacterial cell constituents such as lipids and nucleic acids that can be directly extracted from a soil sample without the need for isolating bacterial cells (Drenovsky et al. 2004). This chapter will focus on lipid-based community analysis, while nucleic acid based methods are described in Chapter 43. Bacterial lipids are key energy storage and cellular membrane compounds that include free fatty acids, hydrocarbons, fatty alcohols, and membrane bound fatty acids such as phospholipids and glycolipids (Kennedy 1994). Bacterial taxa often have unique fatty acid profiles that can be identified and used as indicators of microbial community structure (White et al. 1979; Sasser 1990). Lipid analysis of soil microbial communities consists of recovering lipids by extraction in organic solvents followed by analysis with high-resolution fused-silica capillary gas chromatography (Kennedy 1994). Fatty acids are identified with the aid of computer programs such as the Sherlock Microbial Identification System (MIDI, Inc., Newark, DE), by comparison of their retention times in a column against the retention times of bacterial fatty acid standards. Two protocols for extracting lipids from soil are commonly used. Fatty acid methyl ester (FAME) analysis targets all soil fatty acids, and phospholipid fatty acid (PLFA) analysis specifically targets the phospholipid fatty acids found in viable microbial cells. Fatty acid methyl ester analysis uses a one day, 4-step chemical extraction protocol to saponify and methylate all lipids in a soil sample, extract the FAMEs and analyse them

557


558


through gas chromatography, resulting in a fatty acid profile of the soil. Soil FAME profiles can be compared in order to monitor shifts in the overall community structure and are highly reproducible when obtained from communities under similar environmental conditions (Haack et al. 1994). MIDI-FAME analysis has been used successfully to compare the microbial communities in two or more soils, or in a soil under different management practices or cropping regimes (Klug and Tiedje 1993; Cavigelli et al. 1995; Dunfield and Germida 2001, 2003). This method extracts all lipids in living or dead cells, including animal and plant biomass in various stages of decomposition (Ibekwe and Kennedy 1999). A variety of approaches can be taken during the statistical analysis to minimize interference from plant and animal sources, such as ignoring known plant fatty acids, or fatty acids with chain lengths exceeding 20 carbons, which are more characteristic of eukaryotes than prokaryotes (Buyer and Drinkwater 1997; Fang et al. 2001). However, due to the existence of common fatty acids in plants, animals, and microbes, this method cannot reliably be used for taxonomic characterization of the microbial community by lipid biomarker analysis. Phospholipid fatty acid analysis is a more intensive six day protocol that extracts all fatty acids, isolates phospholipids from other soil lipids using solid-phase extraction, converts them into FAMEs and analyzes them through gas chromatography (Bligh and Dyer 1959; Bobbie and White 1980; Bossio and Scow 1998). This method has the advantage that phospholipids are found only in live bacterial cell membranes, and are degraded upon cell death; therefore, a PLFA profile is indicative of the live microbial community in the soil (White et al. 1979; Zelles et al. 1992). Several studies have used PLFA to monitor shifts in the overall soil microbial community structure (Ba˚a˚th et al. 1992; Bossio and Scow 1998; Feng et al. 2003). In addition, given that PLFA profiles are derived from viable microbial biomass, specific biomarker fatty acids can be used as indicators of a particular group of bacteria, providing a taxonomic representation of the soil microbial community (Pankhurst et al. 2001). For example, hydroxyl fatty acids are derived primarily from Gram-negative bacteria, especially Pseudomonas spp., whereas, cyclopropane fatty acids are indicators of other groups of Gram-negative bacteria, such as Chromatium, Legionella, Rhodospirillum, and Campylobacter (Harwood and Russell 1984; Wollenweber and Rietschel 1990; Cavigelli et al. 1995). Branched fatty acids, such as a 15:0, are commonly thought to be markers for Gram-positive bacteria such as Clostridium and Bacillus (Ratledge and Wilkinson 1988; Ibekwe and Kennedy 1999), and 18:3 v6, 9, 12c, is a fatty acid primarily found in lower fungi (Harwood and Russell 1984). A detailed survey of biomarker fatty acids used as taxonomic indicators has been published in a review by Zelles (1999). The MIDI-FAME (42.2) and PLFA (42.3) protocols are presented in this chapter. Detailed comparisons of the two methods can be found in recent literature (Pankhurst et al. 2001; Petersen et al. 2002; Drenovsky et al. 2004). In general, PLFA represents a functionally more well-defined fraction of soil lipids than MIDI-FAME, specifically analyzes microbial community composition, and provides more consistent fatty acid profiles among sample replicates. However, the MIDI-FAME extraction is not as time consuming, and requires a smaller sample mass to recover a reliable community fingerprint (Petersen et al. 2002; Drenovsky et al. 2004). Researchers should decide which method to use by taking into consideration the above comparisons of the methods, specifically whether taxonomic analysis of the community through biomarker analysis is necessary or whether time and sample size are limiting.


Lipid-Based Community Analysis

559

42.2 FATTY ACID METHYL ESTER ANALYSIS (SASSER 1990; MODIFIED FOR SOIL BY CAVIGELLI ET AL. 1995) 42.2.1 MATERIALS AND REAGENTS 1

25 mL test tubes with Teflon lined caps.

2

Reagent 1 (saponification reagent). 4 M NaOH in 50% methanol. 45 g sodium hydroxide, 150 mL methanol, and 150 mL distilled water.

3

Vortex.

4

Water bath (1008C).

5

Reagent 2 (methylation reagent). 6.0 M HCl in 50% methanol. 325 mL certified 6.0 M HCl, and 275 mL methanol.

6

Water bath (80 + 18C).

7

Cold water bath.

8

Reagent 3 (extraction reagent). 1:1 (v=v) hexane:methyl tert-butyl ether (MTBE). 200 mL hexane, and 200 mL MTBE.

9

Rotary shaker.

10

Centrifuge.

11

Pasteur pipettes, tips flamed to remove contaminants.

12

15 mL test tubes with Teflon lined caps.

13

Reagent 4 (wash reagent). 0.3 M NaOH. 10.8 g NaOH dissolved in 900 mL distilled water.

14

Gas chromatograph (GC) vials.

15

GC equipped with a 25 m (5% phenyl)-methylpolysiloxane column, programable 1708C to 2608C at 28C min1 equipped with flame ionization detector and integrator.

16

Hydrogen gas (99.999% pure).

17

Nitrogen gas (99.999% pure).

18

Air, industrial grade, dry.

19

Computer containing MIDI-Sherlock peak identification software (Microbial ID Inc., Newark, DE).



560

42.2.2 PROCEDURE 1

Place 5 g dry weight of soil into a 25 mL test tube with Teflon lined cap.

2

Saponify fatty acids by adding 5 mL of Reagent 1. Vortex for 10 s. Incubate at 1008C for 5 min. Vortex for 10 s. Incubate at 1008C for 25 min. Allow to cool to room temperature.

3

Add 10 mL Reagent 2. This drops the pH of the solution below 1.5 and causes methylation of the fatty acid. Vortex for 10 s. Incubate at 808C for 10 min. Cool rapidly in cold water bath.

4

Add 1.5 mL Reagent 3. This extracts the FAMEs into the organic phase for use with the GC. Shake on rotary shaker for 10 min. Centrifuge at 121 g for 5 min. Transfer top phase to a 15 mL test tube.

5

Add 3.0 mL Reagent 4. To wash samples and reduce contamination of the injection port liner, the column, and the detector. Shake on a rotary shaker for 5 min. Centrifuge at 129 g for 3 min. Transfer top phase to a GC vial.

6

Separate FAMEs by gas chromatography. GC conditions are controlled by the MIDI Sherlock program (MIDI, Inc., Newark, DE). Basically, a 2 mL injection of the samples is analyzed with a GC at an initial temperature of 1708C, ramped to 2608C at 28C min1 using hydrogen as the carrier gas, N2 as the make up gas, and air to support the flame. Peaks are identified using bacterial fatty acid standards and MIDI-Sherlock peak identification software.

42.2.3 COMMENTS 1

All reagents should be of high-performance liquid chromatography grade.

2

Caution should be taken in Step 2. Tube contents may boil up and leak out the test tube, which could cause fatty acids to volatilize and escape (Schutter and Dick 2000).

3

All glassware should be washed and then fired in a muffle furnace at 4508C for a minimum of 4 h to remove traces of lipids.

42.3 PHOSPHOLIPID FATTY ACID ANALYSIS (BLIGH AND DYER 1959; MODIFIED BY BOSSIO AND SCOW 1998; SMITHWICK ET AL. 2005) 42.3.1 FATTY ACID EXTRACTION Materials and Reagents 1

Teflon centrifuge tubes.

2

One-phase extraction mixture, 1:2:0.8 (v=v=v) chloroform:methanol:phosphate buffer.



561

3

Phosphate buffer (1 M). 39 mL 1 M K2 HPO4 , 61 mL KH2 PO4 , fill to 1 L, adjust to pH 7.0.

4

Shaker.

5

Centrifuge.

6

Separatory funnel.

7

Chloroform (CHCl3 ).

8

Glass tubes.

9

Fume hood.

10

Vacuum aspirator.

11

Compressed nitrogen gas cylinder.

Procedure 1

Place 8 g dry weight of freeze-dried soil into a Teflon centrifuge tube.

2

Add 23 mL of one-phase extraction mixture. Shake for 2 h in dark. Centrifuge for 10 min at 756 g. Decant supernatant to a separatory funnel.

3

Re-extract soil by adding 23 mL of one-phase extraction mixture and shaking for 30 min. Centrifuge for 10 min at 756 g. Decant supernatant and combine with supernatant from Step 2.

4

Add 12 mL phosphate buffer and 12 mL CHCl3 to combined supernatants. Vortex for 1 min, vent periodically. Let stand in dark overnight to allow phases to separate.

5

In a fume hood, remove upper aqueous phase with a vacuum aspirator. Decant the bottom CHCl3 layer containing the lipids into a clean glass tube, and dry under N2 gas at 328C.

6

Add 0.5 mL CHCl3 to wash lipids. Swirl to dissolve the fatty acid residue, and transfer to a clean glass tube. Repeat four times (2 mL total). Dry down samples.

Comments 1

Glassware should be cleaned, then fired in a muffle furnace at 4508C for a minimum of 4 h to remove traces of lipids. Rinse items that will not tolerate heat with hexane.

2

Cloudiness of sample in Step 6 indicates the presence of water in the sample. This is problematic because the water will attack the double bonds in the fatty acids. Therefore, add methanol with a dropper until the solution clears and continue with Step 6 (Balser 2005).



562

42.3.2 SOLID-PHASE EXTRACTION OF PHOSPHOLIPIDS (ZELLES AND BAI 1993) Materials and Reagents 1

Solid-phase extraction columns packed with 0.5 g silica.

2

Chloroform (CHCl3 ).

3

Acetone.

4

Methanol.

5

Glass tubes.

6

Compressed N2 gas cylinder.

Procedure 1

Set up solid-phase extraction columns packed with 0.50 g Silica. Condition column with 3 mL CHCl3 .

2

Add 250 mL CHCl3 to glass tube containing dried lipids. Transfer to the column. Repeat four times (1 mL total).

3

Add 5 mL CHCl3 to the column, and allow draining by gravity. This fraction contains the neutral lipids, discard or save for further analysis.

4

Add 5 mL acetone to the column, and allow draining by gravity. Repeat one time (10 mL total). This fraction contains the glycolipids, discard or save for further analysis.

5

Add 5 mL methanol to the column, and allow draining by gravity. This fraction contains phospholipids, save and dry under N2 at 328C.

Comments Researchers are sometimes interested in analyzing other lipids. The neutral lipid fraction in Step 3 can be saved if interested in analyzing sterols for estimates of fungal biomass, and the glycolipid fraction can be saved if interested in analyzing polyhydroxyalkanoates (PHAs) (White and Ringelberg 1998).

42.3.3 CONVERSION TO FATTY ACID METHYL ESTERS METHANOLYSIS

BY

MILD ALKALINE

Materials and Reagents 1

1:1 (v=v) Methanol:toluene. 100 mL methanol and 100 mL toluene to hexane rinsed bottle.

2

0.2 M methanolic KOH, freshly prepared. Dissolve 0.28 g KOH in 25 mL methanol.



563

3

Vortex.

4

Ultrapure water.

5

1 M acetic acid. Add 58 mL glacial acetic acid to water for a total volume of 1 L.

6

Hexane.

7

Centrifuge.

8

Amber GC vial.

9

Compressed N2 gas cylinder.

10

19:0 methyl ester in hexane (25 ng mL1 ).

11

GC vial with glass insert.

Procedure 1

Redissolve dried phospholipids in 1 mL of 1:1 methanol:toluene and 1 mL 0.2 M methanolic KOH. Vortex briefly. Incubate at 378C for 15 min. Allow sample to cool to room temperature.

2

Extract FAMEs by adding 2 mL water, 0.3 mL 1 M acetic acid, and 2 mL hexane. Shake tube then vortex for 30 s. Separate phases by centrifugation for 5 min at 484 g. Remove hexane (upper) layer, and transfer to an amber GC vial.

3

Wash aqueous phase from Step 2, and remove remaining FAMEs by adding 2 mL hexane to the bottom layer. Shake tube then vortex for 30 s. Separate phases by centrifugation for 5 min at 480 g. Remove hexane (upper) layer, and transfer to GC vial containing FAMEs from Step 2. Dry sample under N2 at room temperature. Store sample at 208C in the dark.

4

Suspend samples in 150 mL hexane containing 19:0 methyl ester (25 ng mL1 ) as an internal standard, and transfer to a GC vial containing a glass insert.

42.3.4 GAS CHROMATOGRAPHIC ANALYSIS Phospholipid fatty acid samples can be analysed by gas chromatography as described in the procedure for MIDI-FAME analysis (Step 6 of Section 42.2.2).

42.4 CHARACTERIZING FATTY ACID DATA 42.4.1 FATTY ACID IDENTIFICATION After GC analysis FAMEs are identified by their equivalent chain length (ECL) values using programs such as the Sherlock Microbial Identification System (MIDI Inc., Newark, DE). Straight-chain saturated fatty acids are assigned an ECL value corresponding to the number of carbons in the FAME chain (e.g., 11:0 ¼ ECL 11.000). Because ECLs are a constant



564

property of a specific FAME, published ECLs in a library of FAMEs can be used for identification (White and Ringelberg 1998).

42.4.2 NOMENCLATURE Terminology to describe fatty acids is described by A:B vC, where A indicates the total number of carbon atoms, B the number of double bonds, and vC indicates the position of the double bond from the methyl end of the molecule. The prefixes i and a refer to iso and anteiso methyl branching. The suffixes c for cis and t for trans refer to geometric isomers. Hydroxy groups are indicated by OH. Cyclopropyl groups are denoted by ‘cy.’ 10 ME refers to a methyl group on the tenth carbon from the carboxylic end of the fatty acid (White and Ringelberg 1998; Smithwick et al. 2005).

42.4.3 STATISTICAL ANALYSES Multivariate analysis such as principal components analysis (PCA) is often used to compare fatty acid profiles of soil microbial communities. PCA is used to summarize data in which multiple variables have been measured for each sample (Cavigelli et al. 1995). This is particularly useful for lipid-based community analysis where more than 40 fatty acids are commonly found in the profile of a single soil. An analysis of variance comparing the average peak area or percent of total of each fatty acid for each treatment is possible, but time-consuming, and ecologically meaningless. Principal components analysis linearly transforms an original set of variables (fatty acids) into a substantially smaller set of uncorrelated variables (principal components) that represent most of the information in the original data set (Dunteman 1989). Principal component (PC) are ordered with respect to their variation so that the first few account for most of the variation present in the data. A visual representation of the variation in the data can be presented in a PC plot, where the score of the first two or three PC are plotted in a two- or three-dimensional graph. PC plots are commonly presented in literature to compare the diversity of a soil microbial community based on FAME or PLFA data (Fang et al. 2001; Dunfield and Germida 2001, 2003; Feng et al. 2003). Further information is obtained by examining the eigenvalue loadings associated with each PC. This value reflects the contribution of each original variable to the variation in the PC, and can be used to identify the fatty acids that contribute the most to variation in the communities. These fatty acids can then be selected to undergo further analysis.

REFERENCES Amann, R., Ludwig, W., and Schleifer, K.H. 1995. Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol. Rev. 59: 143–169. ˚ ., and Fritze, H. 1992. Ba˚a˚th, E., Frostega˚rd, A Soil bacterial biomass, activity, phospholipids fatty acid pattern, and pH tolerance in an area polluted with alkaline dust deposition. Appl. Environ. Microbiol. 58: 4026–4031.

Balser, T. 2005. Phospholipid-fatty acid analysis. (Online). Available at http:==www.cnr.berkeley. edu=soilmicro=methods=BalserPLFA.pdf. Verified March 2006. Bligh, E.G. and Dyer, W.J. 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37: 911–917. Bobbie, R.J. and White, D.C. 1980. Characterzation of benthic microbial community structure by


Lipid-Based Community Analysis high-resolution gas chromatography of fatty acid methyl esters. Appl. Environ. Microbiol. 39: 1212–1222. Bossio, D.A. and Scow, K.M. 1998. Impacts of carbon and flooding on soil microbial communities: phospholipids fatty acid profiles and substrate utilization patterns. Microb. Ecol. 35: 265–278.

565 Haack, S.K., Garchow, H., Odelson, D.A., Forney, L.J., and Klug, M.J. 1994. Accuracy, reproducibility, and interpretation of fatty acid profiles of model bacterial communities. Appl. Environ. Microbiol. 60: 2483–2493. Harwood, J.L. and Russell, N.J. 1984. Lipids in Plants and Microbes. George Allen and Unwin Ltd., Herts, U.K.

Buyer, J.S. and Drinkwater, L.E. 1997. Comparison of substrate utilization assay and fatty acid analysis of soil microbial communities. J. Microbiol. Methods 30: 3–11.

Ibekwe, A.M. and Kennedy, A.C. 1999. Fatty acid methyl ester (FAME) profiles as a tool to investigate community structure of two agricultural soils. Plant Soil 206: 151–161.

Cavigelli, M.A., Roberson, G.P., and Klug, M.J. 1995. Fatty acid methyl ester (FAME) profiles as measures of soil microbial community structure. Plant Soil 170: 99–113.

Kennedy, A.C. 1994. Carbon utilization and fatty acid profiles for characterization of bacteria. In R.W. Weaver et al., Eds. Methods of Soil Analysis, Part 2—Microbiological and Biochemical Properties. Soil Science Society of America, Madison, WI, USA, pp. 543–556.

Drenovsky, R.E., Elliott, G.N., Graham, K.J., and Scow, K.M. 2004. Comparison of phospholipids fatty acid (PLFA) and total soil fatty acid methyl esters (TSFAME) for characterizing soil microbial communities. Appl. Environ. Microbiol. 36: 1793–1800. Dunfield, K.E. and Germida, J.J. 2001. Diversity of bacterial communities in the rhizosphere and root-interior of field-grown genetically modified Brassica napus. FEMS Microbiol. Ecol. 38: 1–9. Dunfield, K.E. and Germida, J.J. 2003. Seasonal changes in the rhizosphere of microbial communities associated with field-grown genetically modified canola (Brassica napus). Appl. Environ. Microbiol. 69: 7310–7318. Dunteman, G.H. 1989. Principal Components Analysis No. 69. In Lewis-Beck, M.S., ed. Quantitative Applications in the Social Sciences. Sage Publications, Newbury Park. Fang, C., Radosevich, M., and Fuhrmann, J.J. 2001. Characterization of rhizosphere microbial community structure in five similar grass species using FAME and BIOLOG analyses. Soil Biol. Biochem. 33: 679–682. Feng, Y., Motta, A.C., Reeves, D.W., Burmester, C.H., van Stanten, E., and Osborne, J.A. 2003. Soil microbial communities under conventionaltill and no-till continuous cotton systems. Soil Biol. Biochem. 35: 1693–1703.

Klug, M.J. and Tiedje, J.M. 1993. Response of microbial communities to changing environmental conditions: chemical and physiological approaches. In Guerrero, R. and Pedros-Alio, C., Eds. Trends in Microbial Ecology. Spanish Society for Microbiology, Barcelona, Spain, pp. 371–374. Ovreas, L. and Torsvik, V. 1998. Microbial diversity and community structure in two different agricultural soil communities. Microb. Ecol. 36: 303–315. Pankhurst, C.E., Yu, S., Hawke, B.G., and Harch, B.D. 2001. Capacity of fatty acid profiles and substrate utilization patterns to describe differences in soil microbial communities associated with increased salinity or alkalinity at three locations in South Australia. Biol. Fert. Soils 33: 204–217. Petersen, S.O., Frohne, P.S., and Kennedy, A.C. 2002. Dynamics of a soil microbial community under spring wheat. Soil Sci. Soc. Am. J. 66: 826–833. Ratledge, C. and Wilkinson, S.G. 1988. Microbial Lipids. Academic Press, Inc., New York, NY. Sasser, M. 1990. Identification of bacteria by gas chromatography of cellular fatty acids. Tech. Note #101. Microbial ID, Newark, DE. Schutter, M.E. and Dick, R.P. 2000. Comparison of fatty acid methyl ester (FAME) methods for


566


characterizing microbial communities. Soil Sci. Soc. Am. J. 64: 1659–1668.

Techniques in Microbial Ecology. Oxford University Press, New York, NY, pp. 255–272.

Smithwick, E.A.H., Turner, M.G., Metzger, K.L., and Balser, T.C. 2005. Variation in NH4 þ mineralization and microbial communities with stand age in lodgepole pine (Pinus contorta) forests, Yellowstone National Park (USA). Soil Biol. Biochem. 37: 1546–1559.

Wollenweber, H.W. and Rietschel, E.T. 1990. Analysis of lipopolysaccharide (lipid A) fatty acids. J. Microbiol. Methods 11: 195–211.

Torsvik, V., Goksoyr, J., and Daae, F.L. 1990. High diversity in DNA of soil bacteria. Appl. Environ. Microbiol. 56: 782–787. White, D.C., Davis, W.M., Nickels, J.S., King, J.D., and Bobbie, R.J. 1979. Determination of the sedimentary microbial biomass by extractable lipid phosphate. Oecologia 40: 51–62. White, D.C. and Ringelberg, D.B. 1998. Signature lipid biomarker analysis. In Burlage, R.S., Atlas, R., Stahl, D., Geesey, G., and Sayler, G., Eds.

Zelles, L. 1999. Fatty acid patterns of phospholipids and lipopolysaccharides in the characterization of microbial communities in soil: a review. Biol. Fert. Soils 29: 111–129. Zelles, L. and Bai, Q.Y. 1993. Fractionation of fatty acids derived from soil lipids by solid phase extraction and their quantitative analysis by GC–MS. Soil Biol. Biochem. 25: 495–507. Zelles, L., Bai, Q.Y., Beck, T., and Beese, F. 1992. Signature fatty acids in phospholipids and lipopolysaccharides as indicators of microbial biomass and community structure in agricultural soils. Soil Biol. Biochem. 24: 317–323.


Chapter 43 Bacterial Community Analyses by Denaturing Gradient Gel Electrophoresis E. Topp and Y.-C. Tien Agriculture and Agri-Food Canada London, Ontario, Canada

A. Hartmann National Institute of Agronomic Research Dijon, France

43.1 INTRODUCTION In ‘‘traditional’’ microbiology, soil bacteria are quantified by methods that detect viable cells, by plate counting or most probable number (MPN) enumeration, for example. The major weakness of this approach is that only those organisms that are viable and able to grow in the chosen media at the specified incubation conditions (e.g., temperature) will be detected. Microbiologists have long suspected that the bacteria that are amenable to culturing using conventional methods represent only a tiny fraction of those in soil. Most likely, because these have fastidious requirements that have foiled the development of suitable cultivation techniques, or because they are in obligate association with other organisms such as protozoa. The reannealing behavior of DNA isolated from soil suggests that a single gram of soil may contain up to 10,000 bacterial types (Torsvik et al. 1996). A gram of agricultural soil typically contains a billion or more bacteria. There are estimated to be about 51030 prokaryotic individuals on Earth, of which 491027 are in the top meter of cultivated land on the Earth’s surface (Whitman et al. 1998). The exploration of this hitherto unseen microbial world is now feasible using methods that can elucidate the abundance, identity, and activity of bacteria without relying on culturing. Methods that exploit the sequence of bacterial nucleic acids extracted directly from soil are particularly powerful in this regard. There are three types of nucleic acids that are informative in soil microbial ecology: DNA, ribosomal RNA (rRNA), and messenger RNA (mRNA). A fundamental dogma in biology is that each organism carries its own genetic ‘‘blueprint,’’ or genome, composed of DNA (some viruses being the exception to the rule). Through the processes of transcription and

567



568

translation, the genes encoded in the DNA control the synthesis and specify the amino acid composition of each and every protein that can be made by the organism. Each protein is encoded by a unique DNA sequence, which is transcribed to an mRNA molecule that directs the order of amino acid assembly by the protein synthesis machinery. The critical component of the protein synthesis machinery is the ribosome, which is built of protein and three different types of rRNA molecules. One of the ribosomal RNA genes, which encodes the 16S rRNA molecule, is a tool of choice in bacterial taxonomy. Portions of the gene sequence are highly conserved among different bacterial groups, and can be used to identify bacteria, and establish their evolutionary relatedness. Overall, the presence of specific rDNA sequences is informative with respect to specific types of bacteria, and the quantity of these molecules informative with respect to the abundance in soil of their bacterial owners. Denaturing gradient gel electrophoresis–polymerase chain reaction (DGGE–PCR) analysis is a relatively tractable and powerful culture-independent approach that can be used to characterize bacterial community composition (Muyzer and Smalla 1998; Muyzer 1999). The method yields a community ‘‘fingerprint’’ that can be responsive to soil treatments. For example, 16S rDNA from soil extract is PCR amplified using primers that are universal for bacteria that will hybridize to conserved sequences and amplify a fragment of this gene from most bacteria in the soil. The composition of the PCR product mixture, with respect to the number of products and their DNA sequences, will vary according to the diversity and identity of the bacteria in the soil. The base sequence of the DNA bands resolved in the gel can be elucidated and compared to known sequences to gain insights into the identity of the bacteria. This is accomplished by excising individual gel bands, eluting the DNA into a buffer, cloning the fragment into a suitable vector, sequencing the cloned DNA, and comparing the sequence with archived sequences (e.g., using the basic local alignment search tool [BLAST] program to match with sequences in the National Center for Biotechnology Information [NCBI] database [McGinnis and Madden 2004]).

43.2 OBTAINING DNA FROM SOILS (MARTIN-LAURENT ET AL. 2001) 43.2.1 CONSIDERATIONS AND PRINCIPLES The objective is to extract and purify DNA from the soil microbial community such that it is suitable for amplification by PCR. The PCR will not work if humic materials and other inhibitory substances are not removed from the extract. Two types of methods for obtaining DNA are generally used. The first method consists of extracting DNA from soil microorganisms isolated directly from the soil matrix. This method is tedious since density gradient centrifugation is used to isolate microorganisms, and a strong bias in the community structure can be introduced since the efficiency of the recovery varies from 5% to 20% depending on the soil. The second and more generally used method consists of extracting DNA directly from bulk soil. The operational challenge with this approach is to extensively purify the DNA, removing humic materials that will otherwise interfere with the PCR. Here, we describe a method based on combined mechanical and chemical lysis of microbes from bulk soil and subsequent purification of DNA. This method has proven to be efficient on a wide range of soils and is rather simple to set up and operate.


Bacterial Community Analyses by Denaturing Gradient Gel Electrophoresis

569

There are a number of commercial kits now available for extracting DNA from soil, and different options for optimizing yield and quality of the extracted DNA. The method described here is based on the method described by Martin-Laurent et al. (2001) with some modifications. Genomic DNA is extracted from bulk soil by a mechanical lysis of microbial cells, which is achieved by bead beating and by a chemical lysis achieved by sodium dodecyl sulfate (SDS), anionic detergent. Soil is then eliminated by centrifugation and DNA precipitated by isopropanol in the presence of potassium acetate. DNA is first purified on a poly(vinylpolypyrrolidone) (PVPP) column and then on a glass milk cartridge (Geneclean Turbo Kit, Bio 101 Systems, Qbiogene).

43.2.2 MATERIALS AND REAGENTS 1

Ice bucket, ice

2

Screw cap tubes (2 mL, sterile) and snap cap microtubes (2 mL, 1.5 mL, sterile)

3

Glass beads 0.1 and 2 mm in diameter (sterile)

4

Bead beater (Mikrodismembrator S, B. Braun Biotech International)

5

Water bath or dry heating block for 2 mL tubes adjustable to 708C

6

Freezer (208C)

7

Microbiospin chromatography columns (Biorad, #732-6204)

8

Refrigerated benchtop centrifuge fitted with a 1.5 mL microtube rotor, capable of 14,000 g

9

Horizontal agarose gel electrophoresis equipment and power supply (capable of at least 300 V)

10

Gel image capture equipment (e.g., Alphalmager, Alpha Innotech Corporation)

11

Micropipettors and sterile tips

12

Lysis buffer: Tris–HCl pH 8, 100 mM; Na2 EDTA pH 8, 100 mM; NaCl, 100 mM; SDS, 2% (w=v)

13

Potassium acetate (CH3 COOK) pH 5.5, 3 M

14

Poly(vinylpolypyrrolidone) (PVPP; Sigma–Aldrich Chemical Co.)

15

Isopropanol and ethanol

16

Calf thymus DNA

17

Geneclean Turbo Kit (Bio 101 Systems, Qbiogene)



570 18

Agarose and TBE buffer: per liter of 5 stock, 54 g Tris base, 27.5 g boric acid, 20 mL 0.5 M EDTA, pH 8.0

19

Ethidium bromide staining solution (0.4 mg L1 )

43.2.3 PROCEDURE 1

DNA extraction from soil should be performed on freshly collected moist (not dried) soil. Alternatively fresh soil samples could be stored in plastic microtubes at 808C after rapid freezing in liquid nitrogen. Air drying the soil will change its microbial composition and is to be avoided.

2

Moist soil (equivalent to 250 mg dry weight) is dispensed into a 2 mL screw cap tube. In each tube, 0.5 g of 0.1 mm diameter glass beads and two 2 mm diameter glass beads are added. One mL of lysis buffer is added. The tubes are shaken in a bead beater for 30 s at 1600 rpm. Then tubes are incubated at 708C for 20 min in the dry heating block or water bath. Tubes are centrifuged at 14,000 g for 1 min and the supernatant transferred into a new sterile 2 mL tube. The supernatant volume is measured and one-tenth of the volume of 3 M potassium acetate pH 5.5 is added. The tube is incubated on ice for 10 min and centrifuged at 14,000 g for 5 min. The supernatant volume is transferred into a new 2 mL tube. To precipitate the DNA, one volume of cold (208C) isopropanol is added and tubes are incubated for 30 min at 208C, then centrifuged at 14,000 g for 30 min and the supernatant carefully discarded. The DNA pellet is washed in 200 mL of cold 70% ethanol and air dried for 30 min at room temperature. DNA is finally resuspended in 100 mL of sterile water.

3

DNA is cleaned up by passage through PVPP columns, prepared just prior to use by the following procedure: Microbiospin columns are placed in 2 mL microtubes and filled with 92 mg of PVPP powder; 400 mL of sterile water is added, and then the assembly is centrifuged at 1000 g for 2 min (108C). Add a second portion of 400 mL of sterile water and centrifuge again. Place the column in a new 2 mL tube and carefully load the DNA extracted in the previous step onto the top of the column. Columns are incubated for 5 min on ice and then centrifuged for 4 min at 1000 g (108C). The volume of eluted DNA solution should be 8090 mL.

4

DNA eluted from PVPP columns is further purified using the Geneclean Turbo Kit according to the manufacturer recommendations (protocol 5.1 rapid isolation of DNA from PCR reactions and other enzymatic solutions). DNA is finally recovered in a volume of 30–50 mL of water.

5

Purified DNA is quantified by visualizing the band following agarose gel electrophoresis (1% agarose gel in TBE buffer). Known amounts of calf thymus DNA (e.g., 10, 50, 100, and 200 ng per well) are loaded besides the purified soil DNA samples.

6

After electrophoresis, gels are stained with ethidium bromide and photographed under UV light illumination. After image analysis, soil DNA quantities are computed from the regression curve obtained from the calf thymus DNA standards.



571

43.3 AMPLIFYING DNA FROM SOIL BY PCR (DIEFFENBACH AND DVEKSLER 2003) 43.3.1 CONSIDERATIONS AND PRINCIPLES The PCR selectively and exponentially replicates specific DNA sequences, corresponding to the oligonucleotide primers used in the reaction. For studies of soil microbial ecology, the PCR is extremely useful in a number of ways. It can be used to generate large amounts of DNA needed for establishing community composition and identity using various electrophoretic fingerprinting or hybridization approaches. The PCR can be primed using specific oligonucleotides that amplify genes of interest only, a positive reaction indicating the presence of those sequences. Finally, target sequences in the soil DNA can be quantified if a thermocycler capable of doing real-time quantitative PCR is available (e.g., a LightCycler PCR System, Roche Applied Science). Primers amplifying specific gene sequences of interest can be obtained from the literature, or readily derived using software packages designed for this purpose (e.g., PrimerSelect, DNAStar Inc.). For any given primers used, the various temperature and time steps used to program the thermocycler, repeatedly denaturing the DNA template, annealing the primers, and extending the new strand of DNA, can be obtained from the literature or optimized in preliminary experiments.


Ice bucket and ice

2

Micropipettors (dedicated to PCR only) suitable for delivering various volumes (1–1000 mL capacity), and sterile aero seal tips

3

Plastic PCR tubes suitable for running 25 or 100 mL reactions

4

Thermocycler, preferably with heated lid

5

PCR preparation hood with UV lamp for destroying ambient DNA contaminants (recommended)

6

Reagents for PCR master mix; Taq polymerase (5 units mL1 ) and 10 reaction buffer (provided by the Taq supplier), dNTPs, and 25 mM MgCl2

7

Oligonucleotide primers. We use universal Bacterial 16S rDNA primers one of which is GC clamped according to Santegoeds et al. (1998)

8

Purified soil DNA template for the PCR

43.3.3 PROCEDURE 1

A series of autoclaved 0.5 mL PCR tubes are labeled and set on ice. A PCR master mix containing all components except the soil DNA template is prepared and dispensed into the PCR tubes. The total volume of the master mix is calculated and adjusted on the basis of the number of reactions to be undertaken. The final mixture volume for each reaction is 100 mL with the following composition: 10 mL of 10 PCR reaction buffer; 0:5 mM of each of the primers; 0.2 mM of



572

each dNTP; three units of Taq polymerase, sufficient autoclaved Milli-Q water to bring the volume to 96 mL. 2

While still on ice, template DNA (generally 4 mL of 1:10 dilution of soil DNA, containing about 40 ng of DNA) is added, the solution is mixed by repeatedly pipetting up and down, and placed into the thermocycler. The DNA is amplified by ‘‘touchdown’’ PCR, which has been optimized for the universal bacterial primers used in the experiments. The thermocycler is programed as follows: 5 min at 948C, three cycles of 1 min at 948C, 1 min at 658C, 2 min at 728C; three cycles of 1 min at 948C, 1 min at 638C, 2 min at 728C; three cycles of 1 min at 948C, 1 min at 618C, 2 min at 728C; three cycles of 1 min at 948C, 1 min at 598C, 2 min at 728C; three cycles of 1 min at 948C, 1 min at 588C, 2 min at 728C; three cycles of 1 min at 948C, 1 min at 578C, 2 min at 728C; three cycles of 1 min at 948C, 1 min at 568C, 2 min at 728C; 14 cycles of 1 min at 948C, 1 min at 558C, 2 min at 728C; and final extension of 5 min at 728C. PCR tubes are removed and stored at 208C.

3

Yield and PCR fragment size are evaluated by electrophoresis through 1% agarose as specified in Section 43.2.3.

43.4 REVEALING COMMUNITY COMPOSITION BY DGGE (MUYZER AND SMALLA 1998) 43.4.1 CONSIDERATIONS AND PRINCIPLES The PCR products produced from a mixed soil DNA template will be of near identical size, reflecting the generally uniform size of DNA between the conserved PCR priming sites. The mixture cannot, therefore, be resolved using standard agarose gel electrophoresis on the basis of size separation. Using DGGE, PCR products are separated on the basis of their melting behavior, determined by the DNA composition, in a gel that contains a vertical gradient of denaturant consisting of increasing concentrations of urea and formamide. The PCR is done with one of the two oligonucleotide DNA primers having a so-called GC-clamp added. This is a GC-rich segment of DNA that does not readily denature or melt. The double-stranded GC-clamped PCR products migrate through the DGGE gel to the point where the lowest temperature melting domain is sufficiently unstable that the double-stranded DNA unravels into the single-stranded forms. The partially denatured molecule stops migrating in the gel when held together by the still double-stranded GC-clamp. Thus, a PCR mixture that contains many molecules varying in their melting behavior will yield a mixture of bands that have migrated to different locations in a DGGE gel. The distribution of bands represents a community fingerprint whose characteristics will vary with the number and identity of the bacteria. DGGE is quite flexible in its application. Targets chosen for analysis may be 16S rDNA or functional genes encoding enzymes of interest. The primers can be chosen to reveal very broad or more distinct groups of bacteria. The gel composition (i.e., concentration of denaturants) and running conditions can be varied to optimize band resolution according to the melting properties of the PCR products.


DGGE apparatus: These can most easily be purchased commercially. We use the BioRad DCode Universal Mutation Detection System. It includes a temperature



573

control module and a ‘‘sandwich’’ core upon which two 16 cm gels can be electrophoresed simultaneously. It also comes with a kit for gel casting consisting of two sets of plates, two sets of clamps and 1 mm spacers, two single-well 1 mm prep combs, and comb gaskets. 2

Gradient gel-casting apparatus: We use a simple acrylic two-chamber linear gradient maker. The solutions are delivered by gravity into the gel-casting sandwich by means of a narrow 25 cm long plastic tube connected to a 20-gauge syringe needle via a luer-lock fitting.

3

Electrophoresis power supply: A direct current (DC) voltage power supply, which has a maximum voltage limit of 500 V DC and a maximum power limit of 50 W.

4

Pipettors for loading samples: A capacity of 100 mL fitted with long pipet tips used for loading polyacrylamide DNA sequencing gels.

5

Trays for staining and destaining gels, a UV transilluminator for revealing stained bands, and a conventional or digital camera for capturing gel images.

6

Software for digitizing, archiving, and analyzing images is very useful.

7

Stock solutions for casting polyacrylamide gel: The concentration of denaturant to be used is adjusted according to the desired gradient range, in the example here from 35% to 65% denaturant, where 100% is defined as 7 M urea (H2 NCONH2 ; 420:4 g L1 water) and 40% v=v formamide (CH3 NO). Caution: Acrylamide monomer is teratogenic (can potentially cause cancer, birth defects) and is a neurotoxin. It must be handled with extreme care and in a fume hood only. Once the gel polymerization has taken place, it is safe to handle with gloves.

8

50 Tris-acetate-EDTA (TAE) buffer: Add the following to 900 mL distilled water. 242 g Tris base, 57.1 mL glacial acetic acid, and 18.6 g EDTA. Adjust volume to 1 L with additional distilled water. Caution: Glacial acetic acid is extremely volatile and corrosive and must be manipulated in a fume hood. In a complete TAE buffer (pH of 8.3), the acetate is no longer volatile. Stock solutions for the low and the high denaturant concentrations have the following composition:

40% Acrylamide=Bis (37.5:1) 50 TAE Formamide (deionized) Urea Total volume

Low (35%)

High (65%)

25 mL 2 mL 14 mL 14.7 gm 100 mL

25 mL 2 mL 26 mL 27.3 gm 100 mL



574 9

Reagents for catalyzing acrylamide polymerization: A freshly prepared solution of 10% (w=v) ammonium persulfate ((NH4 )2 S2 O8 ) in water and N,N,N0 ,N0 -tetramethylethylenediamine (TEMED). Caution: (NH4 )2 S2 O8 is an extremely powerful oxidizing agent, and TEMED is extremely volatile, flammable, corrosive, and fatal if inhaled. These materials must be manipulated in a fume hood.

10

Gel loading dye: The 6 loading buffer contains 0.25% bromophenol blue, 0.25% xylene cyanol FF, and 70% glycerol in distilled water.

11

Staining solution: 1 TAE buffer containing 1 SYBR Green I.

12

1 TAE running buffer for gel electrophoresis is made by diluting 50 TAE (see 8 above) concentrated stock solution 50-fold.

43.4.3 PROCEDURE 1

At least 1 mg of GC-clamped DNA from each sample is obtained by PCR. The quantity and quality of the DNA is evaluated and adjusted. It is important to optimize the PCR reaction to minimize unwanted products that may interfere with gel analysis. The PCR products should be evaluated for purity by agarose gel electrophoresis before being loaded onto the DGGE gel. A clear, bright PCR fragment with limited primer dimer signal is expected for DGGE gel running. In order to obtain sufficient DNA for the DGGE analysis, it may be necessary to combine the products of several PCR reactions, and concentrate them by ethanol precipitation to an ideal concentration of about 150 ng mL1 . DNA can be quantified accurately in a fluorometer in a standard 2 mL assay. Adjust the volume of DNA in each sample to 20 mL with water; add 5 mL of loading dye and mix by vortexing briefly.

2

Fifteen mL of the high and 15 mL of the low concentration denaturant are added into the gradient maker. The high concentration is always added into the chamber that will empty first, the low concentration solution to the chamber that will empty last. Acrylamide polymerization is initiated immediately prior to gel casting by adding to each chamber 150 mL of 10% (NH4 )2 S2 O8 and 15 mL of TEMED. The solutions are allowed to drain into the gel-casting sandwich by gravity, and the comb is set into the still-liquid solution. The gel is allowed to polymerize for at least 1 h at room temperature (about 208C).

3

Buffer reservoir is filled with 1 TAE, and the gel sandwich placed into the reservoir. The temperature setting is adjusted to 608C, and the apparatus is allowed to warm to the set point. The 25 mL DNA samples are carefully added into the bottom of each well using a sequencing gel pipet tip. The gel apparatus is then run for 16 h at a constant voltage of 100 V.

4

Carefully disassemble the gel sandwich and remove the gel from the glass plates. Place the gel into a tray containing 150 mL of 1 TAE buffer and 15 mL of 10,000 SYBR Green I. Stain for 40 min, then carefully transfer the gel into a tray containing 250 mL of 1 TAE buffer, and destain for 5–20 min. Place the gel on a UV transilluminator and photograph for DNA capture.



575

43.4.4 ANALYSIS Information can be extracted both from the DGGE fingerprint patterns and the DNA sequences of specific bands. Gel images can be visually compared to readily detect significant differences or consistent features in the community fingerprint pattern, and how these vary according to soil treatment or conditions. Examples of studies that have employed this approach include evaluating the impact of cropping with transgenic potatoes on rhizospheric communities, the effect of temperature on the community structure of ammonia-oxidizing bacteria in soil, the impact of soil management and plant species on nitrogen-transforming bacteria, and the effect of long-term contamination with organic and heavy metal pollutants on global bacterial community structure (Heuer et al. 2002; Avrahami and Conrad 2003; Becker et al. 2006; Patra et al. 2006). Various statistical analyses are available to establish treatment effects (Fromin et al. 2002; Kropf et al. 2004). The identity of the bacteria from which specific bands in the profile originate can be established by excision, elution, cloning, DNA sequencing, and comparison with published databases. Examples of studies that have employed this approach include establishing the identity of bacteria associated with the decomposition of rice straw in anoxic soils, the elucidation of soil bacteria associated with cysts of the soybean cyst nematode (Heterodera glycines), and the identification of antibiotic-producing rhizospheric pseudomonads that suppress soilborne plant pathogens (Weber et al. 2001; Nour et al. 2003; Bergstra-Vlami et al. 2005). Overall, these studies provide examples of various analytical strategies that can be chosen by the investigator based on the specific research question, the detail of answer required, and the resources available.

REFERENCES Avrahami, S. and Conrad, R. 2003. Patterns of community change among ammonia oxidizers in meadow soils upon long-term incubation at different temperatures. Appl. Environ. Microbiol. 69: 6152–6164. Becker, J.M., Parkin, T., Nakatsu, C.H., Wilbur, J.D., and Konopka, A. 2006. Bacterial activity, community structure, and centimeter-scale spatial heterogeneity in contaminated soil. Microb. Ecol. 51: 220–231. Bergsma-Vlami, M., Prins, M.E., Staats, M., and Raaijmakers, J.M. 2005. Assessment of genotypic diversity of antibiotic-producing pseudomonas species in the rhizosphere by denaturing gradient gel electrophoresis. Appl. Environ. Microbiol. 71: 993–1003.

Cuvelle, S., Gillet, F., Aragno, M., and Rossi, P. 2002. Statistical analysis of denaturing gel electrophoresis (DGE) fingerprinting patterns. Environ. Microbiol. 4: 634–643. Heuer, H., Kroppenstedt, R.M., Lottmann, J., Berg, G., and Smalla, K. 2002. Effects of T4 lysozyme release from transgenic potato roots on bacterial rhizosphere communities are negligible relative to natural factors. Appl. Environ. Microbiol. 68: 1325–1335. Kropf, S., Heuer, H., Gruning, M., and Smalla, K. 2004. Significance test for comparing complex microbial community fingerprints using pairwise similarity measures. J. Microbiol. Methods 57: 187–195.

Dieffenbach, C.W. and Dveksler G.S., Eds. 2003. PCR Primer: A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press. Plainview, NY, USA.

Martin-Laurent, F., Philippot, L., Hallet, S., Chaussod, R., Germon, J.-C., Soulas, G., and Catroux, G. 2001. DNA extraction from soils: Old bias for new microbial diversity analysis methods. Appl. Environ. Microbiol. 67: 2354–2359.

Fromin, N., Hamelin, J., Tarnawski, S., Roesti, D., Jourdain-Miserez, K., Forestier, N., Teyssier-

McGinnis, S. and Madden, T.L. 2004. BLAST: At the core of a powerful and diverse set of


576 sequence analysis tools. Nucleic Acids Res. 32: W20–W25. Muyzer, G. 1999. DGGE=TGGE: A method for identifying genes from natural ecosystems. Curr. Opin. Microbiol. 2: 317–322. Muyzer, G. and Smalla, K. 1998. Application of denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology. Ant. van Leeuw. 73: 127–141. Nour, S.M., Lawrence, J.R., Zhu, H., Swerhome, G.D.W., Welsh, M., Welacky, T.W., and Topp, E. 2003. Bacteria associated with cysts of the soybean cyst nematode (Heterodera glycines). Appl. Environ. Microbiol. 36: 607–615. Patra, A.K., Abbadie, L., Clays-Josserand, A., Degrange, V., Grayston, S.J. Guillaumaud, N., Loiseau, P., Louault, F., Mahmood, S., Nazaret, S., Philipot, L., Poly, F., Prosser, J.I., and Le Roux, I. 2006. Effects of management regime and plant

Soil Sampling and Methods of Analysis species on the enzyme activity and genetic structure of N-fixing, denitrifying, and nitrifying bacterial communities in grassland soils. Environ. Microbiol. 8: 1005–1016. Santegoeds, C.M., Ferdelman, T.G., Muyzer, G., and de Beer, D. 1998. Structural and functional dynamics of sulfate-reducing populations in bacterial biofilms. Appl. Environ. Microbiol. 64: 3731–3739. Torsvik, V., Sorheim, R., and Goksoyr, J. 1996. Total bacterial diversity in soil and sediment communities—a review. J. Ind. Microbiol. Biotechnol. 17: 170–178. Weber, S., Stubner, S., and Conrad, R. 2001. Bacterial populations colonizing and degrading rice straw in anoxic paddy soil. Appl. Environ. Microbiol. 67: 1318–1327. Whitman, W.B., Coleman, D.C., and Wiebe, W.J. 1998. Prokaryotes: The unseen majority. Proc. Nat. Acad. Sci. USA 95: 6578–6583.


Chapter 44 Indicators of Soil Food Web Properties T.A. Forge Agriculture and Agri-Food Canada Agassiz, British Columbia, Canada

M. Tenuta University of Manitoba Winnipeg, Manitoba, Canada

44.1 INTRODUCTION The soil food web is a conceptual simplification of the soil biota, in which the microflora and microfauna are aggregated into general trophic groups that describe major flows of energy (organic C) and nutrients (primarily N and P) (Figure 44.1). Analyses of soil food web structure can provide insight into how soil management practices influence microbial immobilization, and turnover of energy and nutrients (Wardle 2002). Roots, plant residues, agrochemicals, and animal manures are the primary inputs of energy and nutrients to cropped soil. Bacteria and some fungi are the initial decomposers of such organic inputs, and soil food webs can be compartmentalized into bacterial and fungal channels representing very different storage efficiencies and turnover rates (Edwards 2000; Wardle 2002). Thus, separation of bacterial and fungal biomass is a fundamental component of soil food web analyses. Soil protozoa, nematodes, and microarthropods are the principal consumers of the microbial biomass (Figure 44.1). Through their grazing, they regulate microbial community structure and enhance mineralization of nutrients. The protozoa are primarily bacterivorous. The soil nematode community includes bacterivores, fungivores, omnivores, predators (which consume other microfauna), and root-grazers, in addition to the true plant parasites, which are covered in Chapter 33 (Edwards 2000; Wardle 2002). The soil microarthropod community is dominated by Collembola and Acarina (mites), and includes fungivore, bacterivore, omnivore, and predator trophic groups (Edwards 2000; Wardle 2002). Methods for extraction of microarthropods, nematodes, protozoa, and total microbial biomass are described in Chapters 32, 33, 36, and 49, respectively. It is important to note that resource preferences of many taxonomic groups of soil fauna are not directly known, rather, they are inferred from morphological similarity with species of

577



578 Phytophagous nematodes Roots

Predacious mites Collembolans

Mycorrhizal hyphae

Nematode feeding mites

Cryptostigmatid mites Noncryptostigmatid mites

Fungi

Detritus

Fungivorous nematodes

Predacious nematodes

Omnivorous nematodes

Flagellates Amoebae Bacteria

Bacterivorous nematodes

FIGURE 44.1. A soil food web. (Adapted from Hunt, H.W., Coleman, D.C., Ingham, E.R., Ingham, R.E., Elliot, E.T., Moore, J.C., Rose, S.L., Reid, C.P.P., and Morley, C.R., Biol. Fert. Soils, 3, 57, 1987.)

known feeding habits. Consequently, the categorization of soil fauna into broad trophic groups, such as in Figure 44.1, is a gross simplification of the complex interactions that occur in real soil food webs. Microarthropod feeding habits have been assessed on the basis of gut enzymes in addition to mouthpart morphology (Behan-Pelletier 1999). New approaches involving stable isotope and fatty acid analyses are making it possible to confirm food resources for various groups of soil fauna, and measure actual fluxes of C through components of the food web (Fitter et al. 2005; Ruess et al. 2005). In the future, these approaches may become powerful tools for assessing the relative strengths of different pathways within the soil food web without the need for detailed microscopic observation and identification. One approach for describing soil food webs involves obtaining data on biomass of the major trophic groups, and then using the food web model originally described by Hunt et al. (1987) (see Figure 44.1) to estimate flows of C, N, and P through the trophic groups. This model has been tested against actual measurements, of soil C and N mineralization (e.g., De Ruiter et al. 1993; Hassink et al. 1994; Berg et al. 2001), and fluxes of 13 C through some components of the food web (Leake et al. 2006). A truncated version of the model, utilizing only data on bacteria, fungi, protozoa, and nematodes, captures the trophic interactions that make the greatest direct contributions to N mineralization (Hassink et al. 1994; Forge et al. 2005). Introductory information on the conversion of bacterial, fungal, nematode, and protozoan abundance data to biomass, for use in soil food web model analyses, is included in this chapter. Actual construction of soil food web models is beyond the scope of this chapter, and readers are referred to Irvine et al. (2006), De Ruiter et al. (1993), and Hunt et al. (1987) for more details. Another approach for describing soil food webs involves focusing on one or a few organism groups, and using information on changes in community structure within those groups as indicators of changes in properties of the broader soil food web.


Indicators of Soil Food Web Properties

579

For example, the ratio of fungal biomass to total microbial biomass provides information on the relative extent to which C, N, and P are channeled through fungus–fungivore pathways of the soil food web, which is of functional importance because it influences nutrient turnover rates and storage efficiencies. Similarly, some specific descriptors of nematode community structure are becoming popular as indicators of changes in soil food web properties, and will be described in more detail in this chapter. Microarthropod community data can be used in a similar way (e.g., Behan-Pelletier 1999; Parisi et al. 2005), but specific indices of microarthropod community structure have not been adopted as widely as nematode indices for routine assessment of soil food web properties. This chapter will describe methods for distinguishing bacterial and fungal biomass, obtaining data on nematode community structure, and calculating nematode community indices that provide information on the properties of soil food webs.

44.2 DIFFERENTIATING FUNGAL AND BACTERIAL BIOMASS Bacterial and fungal biomass can be differentiated via biochemical, physiological, and microscopic approaches. Guggenberger et al. (1999) used glucosamine and muramic acid as markers of fungal and bacterial contributions to microbial-derived soil organic matter, respectively. Other researchers have inferred bacterial biomass from the difference between total biomass, determined via chloroform fumigation–extraction (Chapter 49), and fungal biomass, determined via measurement of ergosterol (Montgomery et al. 2000). Phospholipid fatty acid (PLFA) profiles of bacteria and fungi differ, and specific PLFA compounds have been used as biomarkers of bacterial and fungal biomass in soil (Chapter 49; Frostegard and Ba˚a˚th 1996; Bossio et al. 1998). The substrate-induced respiration (SIR) method for measuring total biomass has been adapted for separate measurement of fungal and bacterial biomass (Chapter 39; Beare et al. 1990; Lin and Brookes 1999). SIR of soil samples amended with cycloheximide (fungal inhibitor) and streptomycin (bacterial inhibitor) gives estimates of respiration activity of bacterial and fungal biomass, respectively. This procedure assesses only metabolically active components of the respective groups, and is therefore an indicator of biomass, not a direct measure of total biovolume or biomass per se. Fierer et al. (2005) designed general bacterial and fungal primers, allowing for the estimation of bacterial and=or fungal DNA in soil by quantitative polymerase chain reaction; such DNA-based techniques may provide a rapid and reproducible indication of bacterial and fungal biomass in soil. Measurement of microbial biovolumes via microscopy is the most direct method for measuring the biomass of bacteria and fungi, but it is also the most labor-intensive. Recent advances in digital image analysis and confocal laser scanning microscopy have great potential to drastically reduce the labor associated with direct microscopic assessment of microbial biovolumes (Bloem et al. 1995; Bo¨lter et al. 2002). Since the agar film method was first described by Jones and Mollison (1948), numerous variations have been developed (Bottomley 1994). Membrane filter techniques have become popular for fungal biovolume measurement, but they may not necessarily be more effective than the agar film technique (Ba˚a˚th and So¨derstro¨m 1980). Modifications of agar film and membrane filter techniques involving the use of fluorescein diacetate (FDA) or tetrazolium chloride make it possible to discriminate empty or dead hyphae and bacteria, respectively, from metabolically active hyphae and bacterial cells (Bottomley 1994). The following procedure is an adaptation of the agar film technique. It is particularly convenient because



580

it allows for estimation of total (but not metabolically active) fungal and bacterial biomass on the same agar film slides.


Waring blender

2

Fluorescent immunology slides with 15 mm diameter rings (e.g., VWR cat. No. 48349-057); rinsed in 70% ethanol and air-dried before use

3

Nonfluorescent immersion oil (Cargille type A) and #1 cover slips

4

Staining jars

5

Epifluorescence microscope with ocular micrometer and=or grid micrometer

6

Formalin solution (37% – 40% (v=v) formaldehyde)

7

Purified agar solution (0.15%): 1.5 g purified agar in 1000 mL distilled water, bring to boil, and then keep molten in 508C water bath

8

Acridine orange stock solution: 1 g in 500 mL distilled water

44.2.2 PROCEDURE 1

Place 5.0 g fresh soil into a Waring blender with 500 mL distilled water. Blend on highest setting for 1 min. Allow sand to settle for 10 s.

2

Transfer 1 mL to a test tube with 3.5 mL 0.15% agar solution (508C) and 0.5 mL formalin (final conc. 3.7% formaldehyde), resulting in a 1:500 soil suspension in 0.1% agar.

3

Vortex and immediately deliver 0.1 mL onto each of the two 15 mm diameter circular areas on an immunology slide. Using a needle, ensure that the suspension spreads to the edge of each circular area. Place in warm, dust-free area to air-dry (about 4 h).

4

Immerse slides in acridine orange staining solution (5 mL stock solution þ 95 mL water) for 30 min. Remove and gently rinse by dunking in a beaker of distilled water, and allow to air-dry.

5

Place a drop of nonfluorescent immersion oil (e.g., Cargille type A) or glycerol on each smear, and cover with a #1 cover slip.

6

Observe 20 randomly chosen fields-of-view from two perpendicular transects through the diameter of each smear (10 fields-of-view=transect) at 400 to 600 under phase contrast. Using an ocular grid, estimate the length of hyphae in each field-of-view via the gridline intersect method (Newman 1966). Alternatively, use an ocular micrometer to estimate the length of each hyphal fragment encountered in each transect.


Indicators of Soil Food Web Properties 7

581

Observe under oil immersion at 1000 using epi-illumination and a filter set appropriate for acridine orange (excitation 490 nm, emission 520 nm). Bacterial cells will appear green to yellow against a dark background. Count the cells in each of 20 randomly chosen fields-of-view within each agar film, and categorize into the following five size=shape categories (values in mm): (i ) 2:0, and (v) >1:0 >1:0 as described by Lundgren (1984). Nominal volumes for these size classes are 0.03, 0.22, 0.34, 0.75, and 1:77 mm3 , respectively (Lundgren 1984). Calculate the average number of cells of each size class per field-of-view.

44.2.3 CALCULATIONS Using a stage micrometer, determine the area covered by each field-of-view for your microscope using the 40 and 100 objectives, Av40 and Av100 , respectively. Calculate the total area of each smear, As (177 mm2 for 15 mm diameter smears); then calculate biomass for each size class: Bacterial biomass C=g moist soil (Bb ) ¼ [Nave (As=Av100 )=0:1 mL D Vb Cv ] (44:1) where Nave is the average number of cells (for the size class) per field-of-view, As is the total area of the smear, Av is the area of the field-of-view, D is the dilution factor (500 for the above example), Vb is average biovolume for the size class (mm3 ), and C is the specific carbon content of bacteria (fg C=mm3 ). Bloem et al. (1995) reported C to be 196 fg=mm3 : Fungal biomass C=g moist soil (Bf ) ¼ (pr 2 Lave ) (As=Av40 )=0:1 mL D (b=v) Cm

(44:2)

where Lave is the average hyphal length (mm) per field-of-view and (b=v) is the ratio of biomass=biovolume for fungi. Values ranging from 200 to 330 fg=mm3 have been used (Bottomley 1994). Cm is the C content as a fraction of total mass; a value of 0.5 has been used (Van Veen and Paul 1979).

44.2.4 COMMENTS 1

Above example starts with a 1:500 soil suspension. Depending on the soil, it may be necessary to use more or less dilute suspensions. Ideally, bacterial counts should average 15 to 30 cells per field-of-view. Larger cell densities tend to lead to counting fatigue and error. The occurrence of cell clusters is problematic for counting, as they can result in underestimation and large variability. Many researchers have used dispersants such as 0.1% sodium pyrophosphate or Calgon, but Bloem et al. (1995) performed a systematic comparison and found that dispersants did not significantly improve bacterial counts.

2

Sodium dithionate solution (3.5 g in 100 mL distilled water) apparently helps reduce photobleaching during epifluorescence viewing, and can be used as a mounting medium.

3

In the absence of an epifluorescence microscope, phenolic aniline blue, acetic aniline blue, or tryptophan aniline blue can be used. Bacterial cells will appear


582

Soil Sampling and Methods of Analysis dark blue against a bright background. Fungal hyphae take up aniline blue stains to varying degree, but phase contrast is generally the most reliable method for observing fungal hyphae.

44.3 NEMATODE FEEDING GROUPS AND INDICATORS OF FOOD WEB STRUCTURE Several specific indices of nematode community structure are particularly useful for assessing changes in structure of the soil food web as a whole. In this chapter we use the term ‘‘structure’’ to refer to the presence and relative dominance of major trophic groups (Figure 44.1) as well as the degree to which the trophic groups are composed of a diversity of functional groups or guilds. Bongers (1990; 1999) ranked nematode families on a scale of 1 to 5 on the basis of their positions in the colonizer-persister continuum (c-p ranking). The maturity index (MI), a weighted mean representing the c-p ranking of the community, is a general indicator of community response to disturbance, environmental stress, and addition of easily decomposed, high-N organic materials. Ferris et al. (2001) expanded on the MI concept and differentiated changes in nematode communities into enrichment and structure trajectories, and applied weightings to reflect the differential importance of certain nematode families to either food web enrichment or development of structure (Table 44.1). Organic matter inputs, tillage, and other changes that result in increased microbial production=turnover also result in increased abundance of enrichment opportunist nematodes, represented primarily by bacterivores in the families Rhabditidae, Diplogasteridae, and Panagrolaimidae, and fungivores in the families Aphelenchidae and Aphelenchoididae (Table 44.1). Ferris et al. (2001) described an enrichment index (EI) that measures the increased abundance of bacterivore and fungivore enrichment opportunists. The channel index (CI) is a measure of the extent to which the biomass of decomposers is dominated by bacteria or fungi; high values of the CI reflect fungal-dominated decomposition pathways. The EI and overall abundance of bacterivorous nematodes have both been positively correlated with N mineralization (Hassink et al. 1993; Forge and Simard 2001; Ferris and Matute 2003; Parfitt et al. 2005). A few families of bacterivores and fungivores, and all omnivores and carnivores, indicate more structured food webs (i.e., being higher trophic feeders and indicative of greater functional diversity within broad trophic groups); these taxa have been assigned greater enrichment-structure (E-S) weightings (Table 44.1). The structure index (SI) is a measure of the extent to which the nematode community is dominated by these taxa. While the SI is not a measure of taxonomic diversity per se, taxonomic diversity increases with the addition of taxa with high c-p and structure weightings (Ferris et al. 2001). The significance of this relationship is that the SI, based on family-level nematode identification rather than specieslevel identification, may be used for assessment of changes in faunal biodiversity, and perhaps overall soil biodiversity. This section will describe how to assess nematode community structure, assuming the ability to identify nematodes at the level of family (preferably genus). Instruction on basic nematode identification is outside the scope of this chapter. Freckman and Baldwin (1990) is a good primer on soil nematode identification. Some nematology laboratories provide intensive workshops on nematode identification.



583

TABLE 44.1 Feeding Groups, Colonizer-Persister Rankings (c-p) and Enrichment-Structure (E-S) Weightings for Common Families of Terrestrial Nematodes Family

Feeding group

c-p ranking

Tylenchidae Aphelenchidae Aphelenchoididae Neotylenchidae Anguinidae Iotonchiidae Rhabditidae Bunonematidae Diplogasteridae Tylopharingidae Panagrolaimidae Cephalobidae Teratocephalidae Monhysteridae Plectidae Achromadoridae Desmodoridae Odontolaimidae Basianiidae Prismatolaimidae Ironidae Tripylidae Alaimidae Mononchidae Anatonchidae Nygolaimidae Dorylaimidae Chrysonematidae Thornenematidae Nordiidae Qudsianematidae Aporcelaimidae Belondiridae Actinolaimidae Discolaimidae Leptonchidae Diphtherophoridae

F-Rh F F F F F B B B B B B B B B B B B B B B B B C C C O O O O O O–C O C C F F

3 2 2 2 2 2 1 1 1 1 1 2 3 2 2 3 3 3 3 3 4 3 4 4 4 5 4 5 5 4 4 5 5 5 5 4 3

E-S weighting 1.8 0.8 0.8 0.8 0.8 0.8 3.2 3.2 3.2 3.2 3.2 0.8 1.8 0.8 0.8 1.8 1.8 1.8 1.8 1.8 3.2 1.8 3.2 3.2 3.2 5.0 3.2 5.0 5.0 3.2 3.2 5.0 5.0 5.0 5.0 3.2 1.8

(s) (b,e) (b,e) (b,e) (b,e) (b,e) (e) (e) (e) (e) (e) (b) (s) (b) (b) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s) (s)

Source: Data extracted from Bongers, T., Plant Soil, 212, 13, 1999 and Ferris, H., Bongers, T., and de Goede, R.G.M., Appl. Soil Ecol., 18, 13, 2001, with the exception of the Tylenchidae, which were classified as root feeders in the original scheme of Bongers (1990, 1999). F-Rh, fungivore-root-hair feeders; F, fungivore; B, bacterivore; O, omnivore; C, carnivore. Values in parentheses designate the summation groups in which the family is included: b, basal; e, enrichment; s, structure.


Inverted microscope with mechanical stage or compound, and stereomicroscope

2

Gridded counting dish capable of holding >4 mL suspension to fit inverted microscope or stereomicroscope



584 3

Conical-bottom 15 mL centrifuge tubes

4

Microscope slides, #1 cover slips, fingernail polish

5

Pasteur pipettes

6

Extracted nematodes (Chapter 33)

44.3.2 COUNTING 1

Pour nematode sample onto counting dish, observe with dissecting stereomicroscope, and determine total nematodes as described in Chapter 33.

2

Rinse sample into a 15 mL conical-bottom centrifuge tube, and allow nematodes to settle for >2 h or, alternatively, centrifuge at 420 g for 5 min. Using a Pasteur pipette or pipette attached to vacuum, remove all but the bottom 0.5 mL of suspension. Immerse tube in 608C water bath for 1 min to heat-kill nematodes.

3

Suspend nematodes by shaking and remove 0.25 mL using Pasteur pipette, and transfer one drop to each of the two positions on a microscope slide. Cover each drop with a #1 coverslip and seal with fingernail polish. These temporary mounts will generally last through a single workday but often dry out over longer periods. It is best to keep slides cool by placing in a refrigerator if observation is to occur hours after slide preparation.

4

Observe with compound microscope at 400. Make regularly spaced transects through each coverslip, identifying each nematode encountered according to genus, family, or trophic group until 100 nematodes have been identified.

5

Alternatively, a high-quality inverted microscope with 4 and 40 objectives can be used, which allows nematodes to be counted (40), and then identified (400) in one counting dish. When utilizing this approach, it is helpful to first use the compound microscope (400 to 1000) to observe a large number of nematodes (i.e., 500) from the experimental site, identify the dominant taxa present, and learn to recognize the major genera or families during routine counting or categorizing with the inverted microscope at 40 to 400.

44.3.3 CALCULATIONS Calculate the relative abundance ( pi ) of each family listed in Table 44.1 from the identification of 100 nematodes: pi ¼ ni=100. Multiply total nematode abundance by relative abundance of each family to get the absolute abundance (Pi ) of each family. Apply C-P rankings to the families as tabulated by Bongers (1990, 1999; Table 44.1), and E-S weightings according to the scheme of Ferris et al. (2001; Table 44.1). Relevant weighted abundances are calculated as follows, with summation across relevant families: Weighted abundance of basal taxa (b) ¼

X

(vb nb )

(44:3)



585

where nb is the abundance of each family representing basal characteristics of the food web and vb is the weighting associated with each of those families (0.8): Weighted abundance of enrichment opportunists (e) ¼

X

(ve ne )

(44:4)

where ne is the abundance of each family representing enrichment characteristics of the food web and ve is the weighting associated with each of those families (1.8 or 3.2): Weighted abundance of fungivore enrichment opportunists ( fe ) ¼

X

(vfe nfe ) (44:5)

where nfe is the abundance of each fungivorous family representing enrichment characteristics of the food web and vfe is the weighting associated with each of those families (0.8): Weighted abundance of structure taxa (s) ¼

X

(vs ns )

(44:6)

where ns is the abundance of each family representing structural characteristics of the food web and vs is the weighting associated with each of those families (0.8 to 5): Enrichment index ¼ e=(b þ e) 100

(44:7)

Structure index ¼ s=(b þ s) 100

(44:8)

Channel index ¼ fe =e 100

(44:9)

44.3.4 COMMENTS Obtaining total counts is best accomplished with live nematodes, as movement aids detection in the counting dish at 40. If the identification step cannot be accomplished promptly, the samples can be reduced to 0.9 mL, heat-killed as described before, and then 0.1 mL of 16% formaldehyde (40% formalin solution) added as a preservative.

44.4 CALCULATION OF MICROFAUNAL BIOMASS FOR FOOD WEB MODELING If the nematode data will be used in a food web model for calculating nutrient fluxes or simulating food web dynamics, it is necessary to convert numbers of organisms in each trophic group to biomass. Nematode lengths and widths can be determined during the identification step with a compound microscope, and converted to biovolume according to Andrassy (1956). Alternatively, biomass can be inferred from published measurements of each taxon (Forge et al. 2005). After extraction and quantification of protozoa (Chapter 36), protozoan biovolume can be estimated from measurements of mean cell diameter, assuming a spherical shape. Forge et al. (2003) obtained an average diameter of 10 mm for a flagellate-dominated protozoan assemblage. Protozoan biovolume can then be converted to biomass-C using a conversion factor of 0:212 pg C mm3 (Griffiths and Ritz 1988).



586

REFERENCES Andrassy, I. 1956. The determination of volume and weight of nematodes. Acta Zool. (Hungarian Acad. Sci.) 2: 1–15. Ba˚a˚th, E. and So¨derstro¨m, B. 1980. Comparisons of the agar-film and membrane-filter methods for the estimation of hyphal lengths in soil, with particular reference to the effect of magnification. Soil. Biol. Biochem. 12: 385–387. Beare, M.H., Neely, C.L., Coleman, D.C., and Hargrove, W.L. 1990. A substrate-induced respiration (SIR) method for measurement of fungal and bacterial biomass on plant residues. Soil Biol. Biochem. 22: 585–594. Behan-Pelletier, V.M. 1999. Oribatid mite biodiversity in agroecosystems: Role for bioindication. Agric. Ecosyst. Environ. 74: 411–423. Berg, M., DeRuiter, P., Didden, W., Jansen, M., Schouten, T., and Verhoef, H. 2001. Community food web, decomposition and nitrogen mineralization in a stratified Scots pine forest soil. Oikos 94: 130–142. Bloem, J., Veninga, M., and Shepherd, J. 1995. Fully automatic determination of soil bacterium numbers, cell volumes, and frequencies of dividing cells by confocal laser scanning microscopy and image analysis. Appl. Environ. Microbiol. 61: 926–936. Bo¨lter, M., Bloem, J., Meiners, K., and Moller, R. 2002. Enumeration and biovolume determination of microbial cells—A methodological review and recommendations for applications in ecological research. Biol. Fert. Soils 36: 249–259. Bongers, T. 1990. The maturity index: An ecological measure of environmental disturbance based on nematode species composition. Oecologia 83: 14–19. Bongers, T. 1999. The Maturity Index, the evolution of nematode life history traits, adaptive radiation and cp-scaling. Plant Soil 212: 13–22. Bossio, D.A., Scow, K.M., Gunapala, N., and Graham, K.J. 1998. Determinants of soil microbial communities: effects of agricultural management, season, and soil type on phospholipid fatty acid profiles. Microb. Ecol. 36: 1–12.

Bottomley, P. 1994. Light microscopic methods for studying soil microorganisms. In R.W. Weaver et al., Eds. Methods of Soil Analysis, Part 2—Microbiological and Biochemical Properties. Soil Science Society of America, Madison, WI, pp. 81–106. De Ruiter, P.C., Van Veen, J.A., Moore, J.C., Brussaard, L., and Hunt, H.W. 1993. Calculation of nitrogen mineralization in soil food webs. Plant Soil 157: 263–273. Edwards, C.A. 2000. Soil invertebrate controls and microbial interactions in nutrient and organic matter dynamics in natural and agroecosystems. In D.C. Coleman and P.F. Hendrix, Eds. Invertebrates as Webmasters in Ecosystems. CABI Publishing, Wallingford, UK, pp. 141–159. Ferris, H., Bongers, T., and de Goede, R.G.M. 2001. A framework for soil food web diagnostics: Extension of the nematode faunal analysis concept. Appl. Soil Ecol. 18: 13–29. Ferris, H. and Matute, M.M. 2003. Structural and functional succession in the nematode fauna of a soil food web. Appl. Soil Ecol. 23: 93–110. Fierer, N., Jackson, J.A., Vilgalys, R., and Jackson, R.B. 2005. Assessment of soil microbial community structure by use of taxon-specific quantitative PCR assays. Appl. Environ. Microbiol. 71: 4117–4120. Fitter, A.H., Gilligan, C.A., Hollingworth, K., Kleczkowski, A., Twyman, R.M., Pitchford, J.W., and Members of the NERC Soil Biodiversity Programme. 2005. Biodiversity and ecosystem function in soil. Funct. Ecol. 19: 369–377. Forge, T.A., Bittman, S., and Kowalenko, C.G. 2005. Responses of grassland soil nematodes and protozoa to multi-year and single-year applications of dairy manure slurry and fertilizer. Soil Biol. Biochem. 37: 1751–1762. Forge, T.A., Hogue, E., Neilsen, G., and Neilsen, D. 2003. Effects of organic mulches on soil microfauna in the root zone of apple: implications for nutrient fluxes and functional diversity of the soil food web. Appl. Soil Ecol. 22: 39–54.


Indicators of Soil Food Web Properties Forge, T.A. and Simard, S.W. 2001. Structure of nematode communities in forest soils of southern British Columbia: Relationships to nitrogen mineralization and effects of clearcut harvesting and fertilization. Biol. Fert. Soils 34: 170–178. Freckman, D.W. and Baldwin, J.G. 1990. Nematoda. In D.L. Dindal, Ed. Soil Biology Guide. John Wiley and Sons, New York, NY, pp. 155–200. Frostegard, A. and Ba˚a˚th, E. 1996. The use of phospholipid fatty acid analysis to estimate bacterial and fungal biomass in soil. Biol. Fert. Soils 22: 59–65. Griffiths, B.S. and Ritz, K. 1988. A technique to extract, enumerate and measure protozoa from mineral soils. Soil Biol. Biochem. 20: 163–173. Guggenberger, G., Frey, S.D., Six, J., Paustian, K., and Elliott, E.T. 1999. Bacterial and fungal cellwall residues in conventional and no-tillage agroecosystems. Soil Sci. Soc. Am. J. 63: 1188–1198. Hassink, J., Bouwman, L.A., Zwart, K.B., and Brussaard, L. 1993. Relationships between habitable pore space, soil biota and mineralization rates in grassland soils. Soil Biol. Biochem. 25: 47–55. Hassink, J., Neutel, A.M., and De Ruiter, P.C. 1994. C and N mineralization in sandy and loamy grassland soils: The role of microbes and microfauna. Soil Biol. Biochem. 11: 1565–1571. Hunt, H.W., Coleman, D.C., Ingham, E.R., Ingham, R.E., Elliott, E.T., Moore, J.C., Rose, S.L., Reid, C.P.P., and Morley, C.R. 1987. The detrital food web in a shortgrass prairie. Biol. Fert. Soils 3: 57–68. Irvine, L., Kleczkowski, A., Lane, A.M.J., Pitchford, J.W., Caffrey, D., and Chamberlain, P.M. 2006. An integrated data resource for modelling the soil ecosystem. Appl. Soil Ecol. 33: 208–219. Jones, P.C.T. and Mollison, J.E. 1948. A technique for the quantitative estimation of soil microorganisms. J. Gen. Microbiol. 2: 54–69. Leake, J.R., Ostle, N.J., Rangel-Castro, J.I., and Johnson, D. 2006. Carbon fluxes from plants

587 through soil organisms determined by field 13 CO2 pulse-labelling in an upland grassland. Appl. Soil Ecol. 33: 152–175. Lin, Q. and Brookes, P.C. 1999. Comparison of substrate induced respiration, selective inhibition and biovolume measurements of microbial biomass and its community structure in unamended, ryegrass-amended, fumigated and pesticidetreated soils. Soil Biol. Biochem. 31: 1999–2014. Lundgren, B. 1984. Size classification of soil bacteria: Effects on microscopically estimated biovolumes. Soil Biol. Biochem. 16: 283–284. Montgomery, H.J., Monreal, C.M., Young, J.C., and Seifert, K.A. 2000. Determination of soil fungal biomass from soil ergosterol analyses. Soil Biol. Biochem. 32: 1207–1217. Newman, E.I. 1966. A method for estimating the total length of root in a sample. J. Appl. Ecol. 3: 139–145. Parfitt, R.L., Yeates, G.W., Ross, D.J., Mackay, A.D., and Budding, P.J. 2005. Relationships between soil biota, nitrogen and phosphorus availability, and pasture growth under organic and conventional management. App. Soil Ecol. 28: 1–13. Parisi, V., Menta, C., Gardi, C., Jacomini, C., and Mozzanica, E. 2005. Microarthropod communities as a tool to assess soil quality and biodiversity: A new approach in Italy. Agric. Ecosyst. Environ. 105: 323–333. Ruess, L., Schu¨tz, K., Haubert, D., Ha¨ggblom, M.M., Kandeler, E., and Scheu, S. 2005. Application of lipid analysis to understand trophic interactions in soil. Ecology 86: 2075–2082. Van Veen, J.A. and Paul, E.A. 1979. Conversion of biovolume measurements of soil organisms, grown under various moisture tensions, to biomass and their nutrient content. Appl. Environ. Microbiol. 37: 686–692. Wardle, D.A. 2002. Communities and Ecosystems: Linking the Aboveground and Belowground Components. Princeton University Press, Princeton, NJ, USA.


V. SOIL ORGANIC MATTER ANALYSES Section Editors: E.G. Gregorich and M.H. Beare

E.G. Gregorich/Soil Sampling and Methods of Analysis 3586_C045 Final Proof page 589 11.6.2007 12:10pm Compositor Name: TSuresh

Chapter 45 Carbon Mineralization D.W. Hopkins Scottish Crop Research Institute Dundee, Scotland, United Kingdom

45.1 INTRODUCTION Organic matter in soils is the complex mixture of organic compounds derived from the dead and decaying remains of plants, animals, and microorganisms, and their corpses and metabolic wastes at different stages of decomposition. Mineralization of organic carbon (C) is the conversion from the organic form to inorganic compounds as a result of decomposition reactions carried out by decomposer organisms, the vast majority of which are microorganisms (bacteria and fungi) (Gregorich et al. 2001). In the process of utilizing soil organic matter, heterotrophic soil organisms release CO2 during respiration. The release of CO2 as a metabolic by-product of organic matter decomposition is referred to as C mineralization. Because soil organic matter is a complex mixture of organic compounds of different biological origins and at different stages of decay, C mineralization is the result of a complex set of biochemical processes conducted by a wide range of organisms. Despite the fact that it is a simplification of the actual process, C mineralization measurements are commonly used in investigations of soils and the data have a wide range of applications in agriculture, forestry, ecology, and the environmental sciences. One reason for this is the relative ease with which CO2 can be measured in the laboratory. There are a wide range of methods for measuring CO2 production in the field and at the landscape scale, but this chapter is concerned with measuring C mineralization under controlled laboratory conditions and only limited reference is made to field methods to illustrate some principles. Data on mineralization of soil C may be used in two ways. The rate of C mineralization measured over periods from a few days to a few weeks is commonly used as an indicator of general biological activity because it is an integrated measure of the combined respiration rate of all the organisms active in the soil under specific conditions. However, with time and without inputs of fresh organic matter, the rate of C mineralization declines as the most readily available soil organic matter is depleted. The total CO2 -C released when the rate of production subsides is an index of the readily mineralizable fraction of organic C in soil. Given enough time, however, all, or virtually all, soil organic matter will be mineralized and therefore the total mineralizable C fraction is equivalent, or close, to the total organic C content of the soil. It is important to distinguish between the total amount of C that will be mineralized eventually and the fraction readily mineralized during the initial period of rapid

589


590


decomposition when the most easily utilized and accessible components are decomposed. This chapter focuses on the readily mineralizable fraction of the soil organic matter, which is believed to be a biologically meaningful, albeit operationally defined, fraction of the soil organic matter. However, defining biologically meaningful fractions is fraught with difficulties (Hopkins and Gregorich 2005). Because the readily mineralizable C is one such operationally defined fraction, the conditions under which it is measured need to be carefully specified. It should also be recognized that there is no inherent linkage between the size of the readily mineralizable C fraction and the rate of C mineralization measured over the shortterm. Two soils may contain the same amount of readily mineralizable C, but because of more favorable conditions for decomposition, one may have a much faster initial rate of C mineralization than the other.

45.2 SOIL PREPARATION AND INCUBATION CONDITIONS Before the start of the mineralization assay, some degree of sample preparation is inevitable, but in general, this should be kept to a minimum consistent with being able to prepare a representative and suitable sample. Soil is usually sieved (5–10 days) unless the headspace is flushed Not suitable for soils with pH above neutrality because some CO2 is absorbed in the soil solution Usually only suitable for short-term incubations

Closed chamber incubation with CO2 trapping

Can be inexpensive Can have simple equipment requirements Usually easily replicated Usually suitable for both shortand long-term incubations

Composition of the atmosphere changes because of O2 and CO2 depletion, therefore may unsuitable for long-term incubations if there is a large O2 demand Automated, multichannel respirometers are expensive Manual titration of alkali traps can be time consuming and produce toxic waste products that require disposal

Open chamber incubation with continuous flushing and CO2 trapping

Suitable for both long- and short-term incubations

More expensive More complex equipment Less easily replicated

as it is produced (usually in alkali solution) and then determined. In the third, the soil is incubated in a flow-through system in which the headspace is replaced by a stream of CO2 free air and the CO2 released from the soil is trapped or measured continuously as the air flows out of the chamber. The particular choice of approach will depend on the equipment and other resources (e.g., financial) available to the investigator and a consideration of the advantages and disadvantages of the different methods (Table 45.1). The method of CO2 analysis is determined by a combination of the incubation approach adopted and the instrumentation available. Four methods commonly used to determine CO2 produced from soil are outlined below.

45.3.1 ACID–BASE TITRATIONS Carbon dioxide can be trapped in alkali (typically KOH or NaOH) and then determined by backtitration of the excess alkali with a dilute acid (Hopkins et al. 1988; Schinner et al. 1996). In its simplest form, this can be done by a manual titration using a burette with a pH indicator. Automatic titrators that measure pH with an electrode and deliver acid from a mechanized burette can increase the precision, although rarely the sample throughput.



592

45.3.2 INFRARED GAS ANALYSIS Carbon dioxide absorbs radiation in the infrared region and detection of this absorbance is at the heart of infrared gas analyzers (IRGAs) used to determine CO2 in both closed and open chamber incubation systems (e.g., Bekku et al. 1995; Schinner et al. 1996; Rochette et al. 1997; King and Harrison 2002). There are a range of IRGAs commercially available, and many of those used for measuring CO2 from soil are modifications of systems used for photosynthesis measurements.

45.3.3 CONDUCTIOMETRY Carbon dioxide trapped in alkali can be determined conductiometrically on the principle that the impedance of the alkali solution declines as CO2 is absorbed. Although stand-alone conductiometric systems can be assembled (Chapman 1971; Anderson and Ineson 1982), this method of CO2 detection is usually an integral part of multichannel respirometers (Nordgren 1988) which are expensive, but permit a high degree of replication and near-continuous measurements.

45.3.4 GAS CHROMATOGRAPHY Gas chromatography (GC) provides very precise analysis, but is suitable only for incubation approaches in which CO2 accumulates. There is wide variety of GCs available for CO2 determination and a review of the different types is beyond the scope of this chapter. However, the commonest GC methods involve separation on packed columns and detection using either a thermal conductivity (i.e., hot-wire) detector (e.g., Hopkins and Shiel 1996; Schinner et al. 1996). One advantage of GC is that the instruments are very versatile and can be modified for use in many types of analyses other than CO2 determination by reconfiguring the injector, column, and detector.

45.4 CLOSED CHAMBER INCUBATION WITH ALKALI CO2 TRAPS 45.4.1 MATERIALS AND REAGENTS 1

Incubation jars with gastight lids (Mason or Kilner types; Figure 45.1)

2

Glass vials (20–50 mL) for the alkali solution and water

3

M NaOH solution

4

0.5 M HCl solution

5

Phenolphthalein solution

6

1 M BaCl2

7

Pipettes

8

Burette or automatic titrator

9

Magnetic stirrer (optional)

10

Incubator or controlled environment room (optional)


Carbon Mineralization

593

H2O

Alkali

Soil

FIGURE 45.1. Closed incubation vessel with NaOH traps for CO2 .

45.4.2 PROCEDURE Weigh 100–150 g (dry weight equivalent) into jars and record the weight of each jar plus soil without its lid. Place one vial containing 10 mL of 1 M NaOH and one vial containing water into each jar and seal them with the lids (Figure 45.1). Incubate the jars in the dark and at the desired temperature. The CO2 can be assayed at intervals of 3–10 days typically. For each mole of CO2 trapped in the NaOH, 2 moles of NaOH will be converted to Na2 CO3 (Equation 45.1). Therefore, the total CO2 produced is twice the depletion of NaOH in the trap. Remove the vials of water and NaOH and then backtitrate the excess NaOH with HCl (Equation 45.2) using phenolphthalein as an indicator after having removed dissolved CO2 and carbonates by precipitation with the addition of 2 mL of BaCl2 . 2NaOH þ CO2 ! Na2 CO3 þ H2 O

(45:1)

NaOH þ HCl ! NaCl þ H2 O

(45:2)

For example, if 5 mL of 0.5 M HCl was required to backtitrate the excess NaOH in an alkali trap that originally contained 10 mL of 1.0 M NaOH after precipitating the carbonates with BaCl2 , then the CO2 content of the traps would be calculated as CO2 in trap ¼ 0:5 (((VNaOH CNaOH )=1000) ((VHCl CHCl )=1000))

(45:3)

where VNaOH is the initial volume of NaOH (mL), CNaOH is the initial molar concentration of NaOH, VHCl is the volume of HCl used in the titration (mL), and CHCl is the molar concentration of HCl used in the titration. So, CO2 in the trap ¼ 0:5 [((10 1:0)=1000) ((5 0:5)=1000)] ¼ 0:00375 mol C



594

Where the incubation involved 100 g of dry weight equivalent soil and an incubation time of 48 h, the C mineralization rate would be calculated as C mineralization rate ¼ CO2 in the trap=(soil mass in g incubation time in h) ¼ 0:00375=(100 48) ¼ 0:00000078 mol C g1 soil h1 or 0:78 mmol C g1 soil h1 (45:4) If the incubation is to be continued, wipe any condensation from the inside of the jar and the lid, weigh the jars, and correct for any weight loss by addition of water. Then put fresh NaOH and water vials in the jars, reseal them, and continue the incubation.

45.4.3 COMMENTS The method given here is very general and may be adapted to address a wide range of specific research questions. Among other factors, the amount of soil, the temperature and moisture conditions, the concentration and amount of NaOH, and the incubation time can all be adjusted to suit particular applications. It is, however, important to be sure that the headspace in the jars is large enough to avoid the risk of anaerobiosis during long-term incubations. Typically, 100– 150 g soil in a 1000 mL vessel is suitable for 3–4 days incubation intervals. It is also important to ensure that the amount of NaOH is adequate to trap all the CO2 produced. If the amount of CO2 produced is small, reducing the NaOH concentration will increase the sensitivity of the assay. Carbonic anhydrase can be added to the analyte to catalyze the dissolution of CO2 in water and allow titration between two pH endpoints, 8.3 to 3.7 (Underwood 1961). An automatic titrator and a magnetic stirrer can be used to help improve the precision of the titration. However, these are not essential as the assays can be carried out satisfactorily using manual equipment provided the operator is careful and skilful. Commonly used protocols that employ closed chamber incubations to measure soil biological activity and to quantify the amount of readily mineralizable C in soil are given below. Closed chamber techniques involving alkali traps for measuring CO2 production in the field have also been described by Anderson (1982) and Zibilske (1994).

45.5 CLOSED CHAMBER INCUBATION WITH CO2 ACCUMULATION 45.5.1 MATERIALS AND REAGENTS 1

Miniaturized incubation vessels (Figure 45.2a and Figure 45.2b)

2

1% CO2 gas standard mixture

3

Gas chromatograph

4

Incubator (optional)

45.5.2 PROCEDURE This procedure is based on that of Heilmann and Beese (1992) as modified by Hopkins and Shiel (1996). Weigh 10–15 g (dry weight equivalent) soil into glass vials, put them into the



595 Hypodermic needle Septum

1 mL syringe

Three-way valve

Glass vial Soil 60 mL syringe H2O

Soil

(a)

(b)

FIGURE 45.2. Two different (a and b) closed incubation vessels in which CO2 can accumulate.

incubation chambers, and set the volume of the incubation chamber by adjusting the plunger before closing the three-way tap (Figure 45.2a). After 2–3 days, remove a sample of the headspace gas using the smaller sampling syringe, flushing it several times to ensure mixing. Analyze the gas sample by GC. Many GC configurations can be used. In the method of Hopkins and Shiel (1996), a GC fitted with a 1.32 m long 3 mm internal diameter stainless steel column packed with 80=100 mesh Poropak Q and a thermal conductivity detector was used. After sampling the gas from the headspace, the air in the incubation chambers should be replenished before they are resealed and the incubation continued. The incubation chamber shown in Figure 45.2b is an adaptation of the chamber used in Figure 45.1, which can be used for CO2 accumulation.

45.5.3 COMMENTS Soils may contain CO2 sinks, such as alkaline soil solution in which bicarbonate may accumulate (Martens 1987) and chemoautotrophic bacteria which reduce CO2 (Zibilske 1994). The importance of these sinks is often overlooked, but in alkaline soils, where the capacity for CO2 dissolution is large or where the respiratory CO2 flux is small they may lead to underestimates of C mineralization, methods in which CO2 is trapped may be preferable. The incubation chambers can be assembled from easily available materials; however, because some grades of plastic are permeable to CO2 and the joints between components may leak, it is advisable either to check plastic materials before starting or to use glass equipment. If plastic syringes are used, care should be taken to ensure that the insides of the syringe barrels and the plungers do not get scored by soil particles as this will cause them to leak. Because the headspace volume is relatively small, prolonged incubation without replenishing the headspace is not advisable as this will increase the chance of anaerobiosis and will also increase the risk of leakage.



596

To pump Airflow Airflow

Soil NaOH

H2O

Enlarged view NaOH

NaOH

Multiple NaOH traps

FIGURE 45.3. Open incubation vessel in which CO2 released into a CO2 -free air stream is trapped in NaOH traps. (Adapted from Zibilske, L.M., in R.W. Weaver, S. Angle, P. Bottomly, D. Bezdieck, S. Smith, A. Tabatabai, and A. Wollum (Eds), Methods of Soil Analysis, Part 2—Microbiological and Biochemical Processes, Soil Science Society of America, Madison, Wisconsin, 1994.)

45.6 OPEN CHAMBER INCUBATION The system outlined in Figure 45.3 is suitable for collecting CO2 in an open chamber incubation in which the airflow is maintained either by a suction pump or vacuum line to draw air through the apparatus, or by air pumps (such as a diaphragm aquarium pump) or compressed gas cylinders to force air through the apparatus. Depending on the source of the air, it is necessary to consider the purity of the gas and if necessary use supplementary concentrated H2 SO4 scrubbers to remove organic contaminants from the compressed gas cylinder or carried over from the pumps. The CO2 bubble traps on the upstream side of the soil can be replaced with soda lime traps. After the incubation, the contents of the NaOH traps are quantitatively transferred to a beaker and the CO2 produced is determined by titration as described in Section 45.4.2. The main advantages of this approach are that there is no risk of anaerobiosis or leakage of accumulated CO2 , and soil drying is reduced by the air flowing through the water bottle immediately upstream of the incubation chamber. The equipment can be assembled from easily available laboratory glassware. However, for replicated measurements multiple systems will be required and this will increase the amount of laboratory space required.

45.7 CONDUCTIOMETRIC RESPIROMETERS There is a range of dedicated multichannel respirometers, which can be used to measure CO2 production (and in some cases other gases) in soils, sediments, composts, animals, and cell cultures (including microorganisms). A systematic account of the operation of these instruments is beyond the scope of this chapter. The instrument which appears to be most widely used in soil research is the Respicond instrument (Nordgren 1988). This instrument comprises up to 96 chambers (Figure 45.4) in which CO2 is trapped in KOH. Absorbed CO2 leads to a fall in the conductance of the trap as the KOH concentration falls. This change in



597 Pt electrodes Stopper

Alkali (KOH)

Soil

FIGURE 45.4. Closed incubation vessel using within the Respicond respirometer. (Adapted from Nordgren, A., Soil Biol. Biochem., 20, 955, 1988.)

conductance can be measured as frequently as every 30–45 min and the instrument can run for many months with only minimal interruptions. The conductance measurement is very sensitive to temperature fluctuations and variations in the electrical supply to the instrument. Although the instrument has integral temperature control, best results are obtained when it is located in a temperature-controlled room with an isolated electricity supply.

REFERENCES Anderson, J.M. and Ineson, P. 1982. A soil microcosm system and its application to measurements of respiration and nutrient leaching. Soil Biol. Biochem. 14: 415–416.

Chow, A.T., Tanji, K.K., Gao, S.D., and Dahlgren, R.A. 2006. Temperature, water content and wet–dry cycle effects on DOC and carbon mineralization in agricultural peat soils. Soil Biol. Biochem. 38: 477–488.

Anderson, J.P.E. 1982. Soil respiration. In A.L. Page et al., Eds. Methods of Soil Analysis, Part 2—Chemical and Biological Properties, 2nd ed. Agronomy Society of America, Madison, WI, pp. 831–871.

Fang, C.M., Smith, P., Moncrieff, J.B., and Smith, J.U. 2005. Similar responses of labile and resistant soil organic matter pools to changes in temperature. Nature 433: 57–59.

Bekku, Y., Koizumi, H., and Iwaki, H. 1995. Measurement of soil respiration using closedchamber method—an IRGA technique. Ecol. Res. 10: 369–373.

Fierer, N., Craine, J.M., McLauchlan, K., and Schimel, J.P. 2005. Litter quality and the temperature sensitivity of decomposition. Ecology 82: 320–326.

Bol, R., Bolger, T., Cully, R., and Little, D. 2003. Recalcitrant soil organic materials mineralize more efficiently at higher temperatures. Z. Pflanzen. Boden. 166: 300–307.

Gregorich, E.G., Turchenek, L.W., Carter, M.R., and Angers, D.A. 2001. Soil and Environmental Sciences Dictionary. CRC Press, Boca Raton, FL.

Chapman, S.B. 1971. Simple conductiometric soil respirometer for field use. Oikos 22: 348.

Heilmann, B. and Beese, F. 1992. Miniaturized method to measure carbon dioxide production and


598 biomass of soil microorganisms. Soil Sci. Soc. Am. J. 56: 596–598. Hopkins, D.W. and Gregorich, E.G. 2005. Carbon as a substrate for soil organisms. In R.D. Bardgett, M.B. Usher, and D.W. Hopkins, Eds. Biodiversity and Function in Soils, British Ecological Society Ecological Reviews. Cambridge University Press, Cambridge, pp. 57–79. Hopkins, D.W. and Shiel, R.S. 1996. Size and activity of soil microbial communities in longterm experimental grassland plots treated with manure and inorganic fertilizers. Biol. Fert. Soils 22: 66–70. Hopkins, D.W., Shiel, R.S., and O’Donnell, A.G. 1988. The influence of sward species composition on the rate of organic matter decomposition in grassland soil. J. Soil Sci. 39: 385–392. Hopkins, D.W., Sparrow, A.D., Elberling, B., Gregorich, E.G., Novis, P., Greenfield, L.G., and Tilston, E.L. 2006. Carbon, nitrogen and temperature controls on microbial activity in soils from an Antarctic dry valley. Soil Biol. Biochem 38: 3130–3140. King, J.A. and Harrison, R. 2002. Measuring soil respiration in the field: an automated closed chamber system compared with portable IRGA and alkali adsorption methods. Commun. Soil Sci. Plant Anal. 33: 403–423. Martens, R. 1987. Estimation of microbial biomass in soil by the respiration methods: importance of soil pH and flushing methods for respired CO2 . Soil Biol. Biochem. 19: 77–81. Nordgren, A. 1988. Apparatus for the continuous, long-term monitoring of soil respiration rate in large numbers of samples. Soil Biol. Biochem. 20: 955–957.

Soil Sampling and Methods of Analysis Powlson, D.S. 1980. The effects of grinding on microbial and non-microbial organic matter in soil. J. Soil Sci. 31: 77–85. Rey, A., Petsikos, C., Jarvis, P.G., and Grace, J. 2005. Effect of temperature and moisture on rates of carbon mineralization in a Mediterranean oak forest soil under controlled and field conditions. Eur. J. Soil Sci. 56: 589–599. Rochette, P., Ellert, B., Gregorich, E.G., Desjardins, R.L., Pattey, E.L., Lessard, R., and Johnson, B.G. 1997. Description of a dynamic closed chamber for measuring soil respiration and its comparison with other techniques. Can. J. Soil Sci. 77: 195–203. ¨ hlinger, R., Kandeler, E., and Schinner, F., O Margesin, R. 1996. Methods in Soil Biology. Springer-Verlag, Berlin, Germany. Sˇimek, M., Elhottova´, D., Klimesˇ, F., and Hopkins, D.W. 2004. Emissions of N2 O and CO2 , denitrification measurements and soil properties in red clover and ryegrass stands. Soil Biol. Biochem. 36: 9–21. Underwood, A.L. 1961. Carbonic anhydrase in the titration of carbon dioxide solutions. Anal. Chem. 33: 955–956. Wu, J. and Brookes, P.C. 2005. The proportional mineralization of microbial biomass and organic matter caused by air-drying and rewetting of a grassland soil. Soil Biol. Biochem. 37: 507–515. Zibilske, L.M. 1994. Carbon mineralization. In R.W. Weaver, S. Angle, P. Bottomly, D. Bezdieck, S. Smith, A. Tabatabai, and A. Wollum, Eds. Methods of Soil Analysis, Part 2—Microbiological and Biochemical Processes, SSSA Book Series No. 5. Soil Science Society of America, Madison, WI, pp. 835–863.


Chapter 46 Mineralizable Nitrogen Denis Curtin New Zealand Institute for Crop and Food Research Christchurch, New Zealand

C.A. Campbell Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

46.1 INTRODUCTION Nitrogen (N) is generally the most common growth-limiting nutrient in agricultural production systems. The N taken up by crops is derived from a number of sources, particularly from fertilizer, biological N fixation and mineralization of N from soil organic matter, crop residues, and manures (Keeney 1982). The contribution of mineralization to crop N supply may range from 200 kg N ha1 (Goh 1983; Cabrera et al. 1994) depending on the quantity of mineralizable organic N in the soil and environmental conditions (soil temperature and moisture) that control the rate of mineralization. Large amounts of mineralizable N can accumulate under grassland with the result that crops grown immediately after cultivation of long-term grass may derive much of their N from mineralization. In contrast, soils that have been intensively cropped often mineralize little N, leaving crops heavily dependent on fertilizer N. Potentially mineralizable N is a measure of the active fraction of soil organic N, which is chiefly responsible for the release of mineral N through microbial action. Mineralizable N is composed of a heterogeneous array of organic substrates including microbial biomass, residues of recent crops, and humus. Despite a continuing research effort (Jalil et al. 1996; Picone et al. 2002), chemical tests that are selective for the mineralizable portion of soil N are not available and incubation assays remain the preferred way of estimating mineralizable N. Stanford and Smith (1972) proposed a method to estimate potentially mineralizable N based on the mineral N released during a 30 week aerobic incubation of a soil:sand mixture under optimum temperature and moisture conditions. Although this procedure is regarded as the standard reference method, its main application is as a research tool because it is too time-consuming for routine use. Shortened versions of the aerobic incubation method have been found useful in evaluating soil N supplying power (Paul et al. 2002; Curtin and

599



600

McCallum 2004), but these assays still take several weeks to complete and require considerable technical expertise. An anaerobic incubation method for estimating mineralizable N was proposed by Keeney and Bremner (1966). This anaerobic (i.e., waterlogged soil) technique has significant practical and operational advantages over aerobic techniques in that the incubation period is relatively short (7 days) and the need for careful adjustment of soil water content is avoided. This assay is occasionally used for routine soil fertility testing by commercial laboratories. Although Keeney and Bremner (1966) reported good correlations between anaerobically mineralizable N (AMN) and plant N uptake under greenhouse conditions, subsequent work with field-grown crops has given mixed results (Thicke et al. 1993; Christensen et al. 1999).

46.2 POTENTIALLY MINERALIZABLE N 46.2.1 THEORY In theory, potentially mineralizable N is the amount of N that will mineralize in infinite time at optimum temperature and moisture. It is estimated by incubating soil under optimal conditions and measuring N mineralized as a function of time by periodically leaching mineral N from the soil. Potentially mineralizable N is calculated using a first-order kinetic model: Nmin ¼ N0 (1ekt )

(46:1)

where Nmin is cumulative N mineralized in time t, N0 is potentially mineralizable N, and k is the mineralization rate constant. This equation has two unknowns (N0 and k), which are usually estimated by least-squares iteration using appropriate statistics software.

46.2.2 MATERIALS 1

Incubator capable of maintaining temperatures of up to 408C (and humidity near 100% so that soils do not dry out during incubation).

2

Vacuum pump to extract leachate at 80 kPa.

3

Leaching units to hold incubating soils. These can be purpose-made leaching tubes (Campbell et al. 1993), commercially available filter units (e.g., 150 mL membrane filter units; MacKay and Carefoot 1981), or Buchner funnels (Ellert and Bettany 1988; Benedetti and Sebastiani 1996).

4

Glass wool to make a pad ~6 mm thick at the bottom and 3 mm on top of the incubating sample.

5

Acid-washed, 20 mesh quartz sand.

6

0:01 M CaCl2 leaching solution (made from a stock solution of CaCl2 ).

7

N-free nutrient solution containing 0:002 M CaSO4 , 0:002 M MgSO4 , 0:005 M Ca(H2 PO4 )2 , and 0:0025 M K2 SO4 to replace nutrients removed from the soil during leaching.


Mineralizable Nitrogen

601

46.2.3 PROCEDURE 1

Soils are usually air-dried and sieved before incubation, but field-moist soil may also be used. The N mineralization rate can be quite sensitive to sample pretreatment, particularly in the early phase of incubation (see Section 46.2.5).

2

Mix 15–50 g of soil with sand at a soil:sand ratio of 1:1 for medium-textured soils and 1:2 for fine-textured soils. It may be helpful to apply a light mist of water to prevent particle=aggregate size segregation during transfer to the leaching tubes.

3

Sand–soil mixture is supported in the leaching tube on a glass wool pad or by a sandwich of glass wool=Whatman glass microfiber filter=glass wool. A thin pad of glass wool is placed on top of the soil–sand mixture to prevent aggregate disruption when leaching solution is applied.

4

Native mineral N is leached using 100 mL of 0:01 M CaCl2 , applied in small increments (~10 mL) followed by 25 mL of N-free nutrient solution. The soil–sand mixture is initially allowed to drain naturally, then a vacuum (80 kPa) is applied to remove excess water. Discard the first leachate.

5

Tubes are stoppered at both ends and placed in an incubator at 358C. A hypodermic needle (38 mm, 16–18 gauge) is inserted in the bottom to facilitate aeration. Twice per week the top stopper is briefly removed to facilitate aeration.

6

Step 4 (leaching) is repeated every 2 weeks for the first 8–10 weeks of incubation and every 4 weeks thereafter. The collected leachate is filtered through a prewashed Whatman No. 42 filter paper and analyzed for NO3 - and NH4 -N.

7

Incubation can be terminated when cumulative N mineralized approaches a plateau. This usually occurs after about 20 weeks (see Section 46.2.5).

46.2.4 CALCULATIONS Nonlinear least-square regression is the preferred statistical technique to estimate N0 and k in the first-order kinetic model (Campbell et al. 1993; Benedetti and Sebastiani 1996). Rough estimates of N0 and k are needed to initiate the calculation. We suggest an initial estimate of k 0:10 per week (values normally between 0.05 and 0.20 per week) and N0 can be assumed to be about 50% greater than cumulative mineralized N at the end of the incubation period (Campbell et al. 1993).

46.2.5 COMMENTS 1

The most appropriate way of handling samples before incubation has not been established. Both air-dry soil and field-moist samples have been used. Where moist samples are to be used, they should be refrigerated (about 48C) in the period between sampling and incubation. Campbell et al. (1993) recommend air-drying after collection, which may be appropriate in regions where soils become air-dry in the field. Air-drying can kill off part of the microbial biomass and rapid mineralization of this microbial-N will occur upon rewetting.


602

Soil Sampling and Methods of Analysis The single-exponential model (Equation 46.1) may not adequately describe the initial flush of mineralization that occurs after rewetting (Cabrera 1993) and data for the first 2 weeks have sometimes been excluded when estimating N0 (Stanford and Smith 1972). The degree of sample disturbance (e.g., fineness of sieving) may also influence the results. However, Stenger et al. (2002) found little difference in N mineralization (6 month incubation) between intact and sieved (10 kPa) because of the absence of capillary rise for quasisaturated materials, but this is of minor importance because materials at water potentials >100 kPa are already highly hydrophilic (Michel 1998).

68.3 BULK DENSITY Bulk density (BD) is an easily measured property correlated with many other peat properties and is used for trade and characterization purposes. In organic soils, it is simply determined in the same way as for mineral soils using a cylinder of known volume forced into the soil, then air dried and weighed. Special sampler shapes have been proposed (Sheppard et al. 1993). A cylinder with a sharpened edge can also be used. However, the bottom surface contact area of the core should be very large, as small cores (5 cm diameter and less) can compact the organic soil easily on a ratio of 5, 6, or even 7 to 1 (against about 1.2 to 1 for mineral soil with the same instrument). For peat-based substrates (as well as other organic growing media), specific methods have been developed to prevent any disturbance, because of their sensitivity to settling and loosening. Therefore, a reliable methodology to prepare samples and measure bulk densities is of great importance. Also, because of this sensitivity, sample preparation should yield BD values close to those obtained under cultivation. Worldwide, two groups of standardized methods exist. The first group uses large-volume samples (20 L) prepared without compaction and have been designed and are used mainly for trade purposes (e.g., CEN method, 12580; Morel et al. 1999). A second group of methods uses smaller volumes (usually less than 1 L), which are naturally drained after saturation or onto which an external pressure is applied in addition to overburden pressure (Hidding 1999). These methods attempt to mimic natural settling conditions in a potted substrate under cultivation and are mainly used for characterization purposes. Additional details on these two groups can be found in Caron and Rivie`re (2003).

68.3.1 BULK DENSITY WITH CORES (FOR ORGANIC SOILS OR LOOSE SUBSTRATES) This method is commonly used with organic soils or on cores filled with loose substrates. Materials and Reagents 1

Core samplers or McCauley sampler and sharpened knife

2

Forced-air oven

Procedure 1

For organic soils, trim a core of undisturbed peat with a sharpened knife to fit roughly into the core sampler. A core sample can also be extracted with a McCauley sampler.

2

Alternatively, for peat substrates, they should be loosely packed into cores (as described below for the water desorption curves, using the CEN=TC 223 method). Samples should be prepared as for the water desorption procedure with the above part of the cylinder removed (see water desorption), and the


Organic Soils: Water and Air Storage and Flow Dynamics

891

sample wetted from underneath and then saturated again. Then the substrates should be drained on a tension table (see below) at a tension of 1 kPa. 3

Dry samples at 105 C or 70 C, then compute the BD from BD ¼

SDW V

(68:5)

where BD is bulk density (mass per unit volume), SDW is the sample dry weight, and V is the core volume. Comments Care should be taken in order to minimize peat compression by the cylinder when sampling. The BD of peat materials is affected by water content. Preequilibration of the substrate at 5 kPa before core filling is therefore now a standard practice in Europe (Verdonck and Gabriels 1992). The correlation between BD and ash content is high (Grigal et al. 1989). BD is also highly correlated to the degree of decomposition (Silc and Stanek 1977). The above method, as well as other methods using the oven-dry weight at 105 C (see below), might slightly overestimate the water content of peat, since peat drying at a temperature exceeding 85 C can result in some loss of organic matter (Macfarlane 1969). Hence, drying at 70 C may be preferred. Prior checks should be made with the type of material in order to set a reference drying temperature.

68.3.2 BULK DENSITY IN SITU Techniques have also been developed to measure BD directly in the potted substrates without any substrate manipulation (Paquet et al. 1993). After the compaction process, either natural or artificial, BD is determined using time domain reflectometry (TDR) (Topp et al. 1980) and is calculated from the total porosity (water content at saturation) determination measured on soil cores or cylinders using TDR. The same technique can be used at a saturated depth in the profile (see Section 68.4.2).

68.4 WATER AND AIR STORAGE 68.4.1 TOTAL POROSITY Total porosity is the first point of the water desorption curve since it gives the total volume available for water and air storage. It is one of the important parameters in assessing the quality of commercial growing media (Hidding 1999). It can be measured directly in situ for potted growing media or directly in the field for organic soils. Alternatively, it can be measured on soil or substrate cores. For loosely filled substrate cores, samples are prepared the same way as for the water desorption curve (see below).

68.4.2 TOTAL POROSITY IN SITU The approach is based on the volumetric water content determination on cores or potted substrates slowly rewetted from underneath, and assumes no air entrapment. The same technique can be used at a saturated depth in the profile.



892 Materials 1

Samples in a pot or in cores or alternatively a saturated depth in the profile.

2

A water bath for slow rewetting from underneath (for pots or cores).

3

Distilled or deionized water.

4

TDR probes of the length equivalent to the desired volume to be sampled.

5

A time domain reflectometer. Alternatively, CS-615 or any other time domain, frequency domain, or capacitive probe could be used, but they tend to be less accurate.

6

An appropriate calibration curve for the probe, if necessary.

Procedure 1

Rewet a sample in a pot or core from underneath by gradual elevation of the water level and immersion for 24 h, with the water level about 1 cm below the top of the substrate. This achieves full saturation of the substrate, without having significant changes in total porosity as a result of the release of the overburden pressure.

2

Insert the TDR probe into the sample. As the TDR technique represents the average water content along the sampling probe, the probe should be made to sample the total pot height (for a vertical sample), desired soil depth (for sampling in peat bogs), or the pot diameter (for horizontal sampling). The probe should be fully inserted into the substrate, or its full length and should sample the whole depth. Additional measurements could be taken, if necessary, by reinserting the probes at other locations within the sample. The TDR apparatus determines the apparent dielectric constant of the medium, from which volumetric water content is derived. The apparatus must be calibrated a priori against volumetric water content determined from weighing, as equations relating the apparent dielectric constant to volumetric water content derived for mineral soils do not apply to organic soils (Paquet et al. 1993). Paquet et al. (1993), Anisko et al. (1994), and Da Sylva et al. (1998) have published different calibration equations for various organic–mineral soil mixtures.

Calculations Calculate volumetric water content from the dielectric constant measurement. Total porosity is then assumed to be equal to the volumetric water content. BD can be deduced from total porosity estimates, obtained by measuring water content after complete rewetting of the substrate and from particle density (PD) estimates (see next section). Comments It is critical with this approach to resaturate slowly (resaturation by slowly raising the water level in the bath to reach full saturation of the sample). This procedure is frequently used for



893

well-moist blonde peat (Caron et al. 2002), but may lead to serious errors with air-dried white or black peat materials (Wever 1995), and therefore the period of rewetting should be prolonged if necessary. One way to check is to monitor, over a long period of time, the readings obtained with TDR to make sure they do not increase and to deaerate the sample if needed by putting the core in a chamber or the pot within a cell (Nemati et al. 2002). Potted plants, recently cut, with many roots, have been suspected of generating lot of gases, and care should be taken to apply a vacuum when resaturating the sample to avoid any gas entrapment. Gas entrapment (methane) may also occur in field samples (Buttler et al. 1991).

68.4.3 TOTAL POROSITY BY CALCULATION Physical parameters, such as total pore (TP) space, can be calculated from knowing the BD and ash content, after determining the BD obtained on soil cores (see above). Calculations Total porosity can be calculated from BD and PD. Particle density is estimated from ash content (Paquet et al. 1993), assuming a PD of 1.55 for the organic fraction (OM) and 2.65 for the mineral fraction (Verdonck et al. 1978): %OM ¼ 100% %ash PD ¼

F 1:55

1 þ 1F 2:65

(68:6)

(68:7)

where F¼

%OM %solids

(68:8)

BD PD

(68:9)

TP ¼ 1

Alternatively, PD can be measured on samples using kerosene or ethanol and a pycnometer using the classical mineral soil approach (Blake and Hartge 1986). Comments Macfarlane (1969) reported that estimates of PD calculated from ash content can deviate up to 18% from the actual value.

68.4.4 WATER DESORPTION CURVE ON CORES This method is widely used in Europe and has been adopted by the International Society of Horticultural Science to characterize growing media used in nurseries and greenhouses (Verdonck and Gabriels 1992; Hidding 1999). It provides estimates of the volumetric water content (sometimes referred to as water-holding capacity) at different potentials.



894

When characterizing peat material, past literature and common use have referred to the water-holding capacity of peat as one characteristic, often obtained by the soak and drain method (Parent and Caron 1993), where water in saturated peat is extracted by gravity. However, since the water potential at which the water-holding capacity is measured is equal to half the height of the peat sample, the value of water-holding capacity is inaccurate as the method does not provide information on the height of the cylinder at which the water content is measured. Hence, much more complete information is provided by the water desorption curve, and this characterization is recommended for obtaining information on the water-holding capacity of materials. Materials 1

Double rings (5 cm height and 10 cm i.d.) as well as fixing collars made of temperature-resistant polypropylene should be used. Fix the collar at the periphery of the bottom ring of known volume VR (2 cm high, glued at a height of 1.5 cm on the lower ring, and hence overlapping 0.5 cm on the upper ring).

2

Nylon gauze fitting the bottom of the rings.

3

A sand box or a tension table (Topp and Zebchuk 1979).

4

Large plastic containers, perforated at the bottom.

5

Steel frame and nylon cloth about 30 60 cm.

6

A large plastic water bath (about 5 L or more).

Procedures 1

Transfer around 10 L of material in several containers.

2

Cover the pot with a nylon cloth.

3

Place the pot on a steel frame into the water bath.

4

Fill the bath slowly up to 1 cm under the top of the substrate.

5

Stand overnight.

6

Remove the pot and leave 48 h on the sandbox, applying a potential of 5 kPa pressure head measured from the bottom of the plastic pot.

7

Then fill the assembled rings (with a cheese cloth or very coarse nylon cloth secured at the bottom of the lower ring) with the material using a large spoon in increments of about 100 mL without causing compaction (filling the hollow spaces though) of the removable rings.

8

Cover the upper ring with a nylon cloth to prevent substrate from floating.

9

Transfer the filled double rings to the bath.



895

10

Flood slowly to 1 cm from the top of the upper ring.

11

Saturate for 24 h.

12

Transfer to the tension table.

13

Cover the table and the sample with a cloth.

14

Apply a pressure head of 1 kPa (10 cm of suction), calculated from the middle of the lower ring.

15

Equilibrate for 48 h.

16

Remove the sample from the table and then the upper ring slowly, exposing the uppermost part to the material.

17

Strike off the excess material keeping a flat surface using a sharp knife without causing compaction. This is a delicate operation that should be performed as precisely as possible. Very fibrous material should be cut with large scissors.

18

Determine the weight of the sample present in the lower ring (WR ).

19

Return the sample for other potential determinations (2, 5, 10 kPa), leaving them at least 24 h between each measurement. Check for equilibrium (constant weight) at a given potential before applying a new potential.

20

Dry the sample in the oven at 105 C to estimate dry BD and volumetric water content at each corresponding potential. For shrinking material, measurement of height of the substrate within the ring may be useful to correct the volumetric water content by adjusting VR accordingly. Adjust for cheese cloth weight.

Calculations 1

Volumetric water content at 1 kPa, u1 , is determined as u1 ¼

W1 WR VR

(68:10)

where WR is the dry weight of the lower ring and VR the ring volume. The volumetric water content for other potentials is then determined accordingly by replacing W1 with the weight corresponding to the potential applied. 2

Air volume (AFP for air-filled porosity) is determined from the difference between total porosity (TP) and u1 : AFP ¼

TP u1 VR

(68:11)



896 3

Easily available water (EAW) is the water volume between 1 and 5 kPa and is calculated from EAW ¼

4

(68:12)

Available water (AW) is the water volume between 1 and 10 kPa, and is calculated from AW ¼

5

u1 u5 VR

u1 u10 VR

(68:13)

Buffer capacity (BC) is the water volume between 5 and 10 kPa and is calculated as BC ¼

u5 u10 VR

(68:14)

Comments Because of pronounced hysteretic effects, the measurements should be performed at different potentials upon rewetting. This is particularly relevant when the substrates are used with subirrigation devices (ebb and flow, gullies, capillary mats) to be more representative of growing conditions.

68.4.5 WATER DESORPTION OF POTTED SUBSTRATES For diagnostic purposes, in research facilities, nurseries, or greenhouses, attempts have been made to infer properties existing in pots from the water desorption curves measured on independent samples. Paquet et al. (1993) have shown that substrate disturbance alters these properties and Fonteno (1989) exemplified the effects that container size and geometry may have on existing water and air contents in containers. It is also our experience that once potted, substrate physical properties will evolve significantly, as a result of settling and compaction, decomposition, particle reorganization, and root activity (Allaire-Leung et al. 1999). Hence, measurements taken directly in the pot before and during plant growth are advisable for accurate diagnosis and proper characterization.

Materials 1

Potted substrates with the containers open at the bottom

2

TDR apparatus and probe or a weighing scale

3

Polyethylene sheets, a screen or nylon gauze

4

Tension table apparatus (Topp and Zebchuk 1979)



897

Procedure 1

When using potted substrates with growing plants, first cut off the top part of the plant.

2

Resaturate the substrate from below by slowly increasing the water content to saturate the sample, measuring water content with a vertically inserted probe.

3

Bring the water level up overnight, and finish raising it up to 1 cm below the top of the substrate.

4

Measure the water content during the day to make sure it reaches a constant value. Drift may suggest a high level of entrapped air. Flush trapped air with carbon dioxide if necessary, and then resaturate with deaerated water or put the sample into a vacuum chamber.

5

Determine water content at saturation (it then equals TP).

6

Let the sample drain for about 1.5 to 2 h, and then measure the TDR values vertically at regular intervals during drainage to make sure they have reached a constant value, i.e., the volumetric water content at container capacity, which may differ from that at 1 kPa, since the equivalent applied potential, due to the weight of water, varies with container height (Fonteno 1989). Alternatively, the container can be weighed (in absence of TDR measurements).

7

AFP can then be calculated from the difference between volumetric water content at saturation and that after saturation and drainage (uc ) using AFP ¼ TP uc

(68:15)

8

Cover the substrate top surface with a plastic sheet to restrict evaporation.

9

Take a second measurement with a horizontally inserted probe and note the height of the probe to estimate the corresponding potential.

10

Put the substrate in contact with the tension table (Figure 68.1). Make sure a good contact is established between the sample and the tension table by making a slurry if necessary, or by making additional holes at the bottom of the pot and fixing a screen or a gauze, if necessary, to retain the substrate.

11

Apply a series of water potentials, the most common being 1, 2, 5, and 10 kPa, by lowering the opening of the drainage tube of the tension table to a fixed distance (10, 20 cm, etc.) from the probe height. Equilibrate for 24 h between each potential point, measuring the volumetric water content or weigh the pot between each point. For each measurement, record the water content and the corresponding height. Using TDR instead of a weighing balance has the advantage of accuracy at a known height and prevents removal of the potted substrate from the tension table.

12

Measurements should be performed on rewetting and on drainage, as the properties clearly differ because of hysteresis, which affects air content and AW.



898 Partly buried TDR or capacitive probe

Tensiometer

Tension table

Outflow tubing

FIGURE 68.1. The system to measure the soil water desorption curve on potted substrates or in cylinders showing horizontally or vertically installed TDR probes. The plant should be cut before putting the pot onto the tension table.

13

Draw the water desorption curves and calculate the parameter as mentioned above (Equation 68.11 through Equation 68.15). In the absence of a TDR, TP should be calculated from the bulk and particle densities, which requires determination of the whole container weight at 105 C for use in the calculation of AFP in Equation 68.15.

Comments The entrapment of air can be a problem with the tension table method, and care should be taken to avoid breaking the water column. The above procedure has been described in more detail elsewhere for pots (Paquet et al. 1993).

68.4.6 POINT OF AIR ENTRY Measurement of point of air entry is critical if an indirect assessment of the gas diffusivity is desired (Caron and Nkongolo 2004). It may also be used for modeling purposes and in the design of growing systems for identifying the height of the water-saturated zone at the bottom of containers, which then may result in oxygen-deficient zones. Caron et al. (2002) and Nemati et al. (2002) carried out a comparison of different approaches to identify the



899

air-entry value, and the one based on the water desorption curve is presented here (Nemati et al. 2002) for determination of air entry at the bottom of the pot. Nemati et al. (2002) have also proposed and validated another approach based on pressure transducer measurements, which can be automated. Other approaches have been proposed and tested but have yielded inaccurate estimates for a large number of growing media (Caron et al. 2002). Materials 1

Substrates potted in PVC cylinders or in greenhouse or nursery containers, loosely filled or taken directly from the production area should be used. The walls of cylinders or pots should be perforated with two tapped holes: one for the TDR probe to monitor water content and the other for the tensiometer to monitor matric potential.

2

A TDR probe consisting of three 145-mm long stainless steel rods, 2 mm diameter and spaced 15 mm apart, forming a plane.

3

Minitensiometers approximately 80 mm long 8 mm outside diameter.

4

A fast-response tension table, an Erlenmeyer flask, and a metal stand (see Figure 1 in Nemati et al. 2002).

Procedure 1

Before starting the measurements, evaluate the zone of influence of the TDR probe. It can be determined by immersing horizontal probes in water and measuring at what distance from the horizontal position of the probes any changes in water content can be accurately detected.

2

Prepare the substrates in cylinders as mentioned above (see Section 68.4.5) or obtain the containers from the production area.

3

Saturate the sample from the bottom, slowly raising the water level overnight. Bring the water level to about 1 cm from the top of the substrate.

4

Slowly drain the substrate and put the cylinder in contact with the tension table.

5

Insert the TDR probe (30 mm from the bottom of the pot) and the tensiometers horizontally at the precise height corresponding to the top zone of influence of the probe (Nemati et al. 2002), and apply a potential of þ0:2 kPa to the table (i.e., with a 20 mm height of water on the surface of the tension table).

6

Cover the cylinders or pots to restrict evaporation, but leave small holes to allow air to enter at the surface.

7

Monitor water content and matric potential of the substrates daily after each decrease of water level at the rate of 10 mm per day. This step can be fast or slow depending on the hydrodynamics of the system. An assessment of the dynamics of equilibrium should be performed first by checking (after potential changes) the pace at which the water content changes, by monitoring the water content with the TDR probes at different times.


900


8

Adjust the matric potential data for the height of the water column in the tensiometer as well as for the distance between the tensiometer and the TDR probe.

9

Plot water content as a function of water potential u(c). It will result in a curve with at least two distinct zones: an initial zone showing nearly constant water content with decreasing water potential and a second zone showing sharp decrease in water content with water potential (in some cases followed by a third zone of lesser decrease). The air-entry value is estimated from the intersection of the fitted lines for the first two zones as illustrated by Nemati et al. (2002).

68.5 WATER AND GAS MOVEMENT 68.5.1 SATURATED HYDRAULIC CONDUCTIVITY ON CORES OR

IN THE

FIELD

This method is preferably used with organic soils and substrate-filled cores. Hydraulic conductivity of organic soils is determined in the same manner as mineral soils (Klute and Dirksen, 1986). One should be aware that organic soils are composed of rather loose materials and that, consequently, the walls of auger holes may be unstable while measuring hydraulic conductivity or the infiltration rate of water in the field. However, measurements in situ are obviously more preferable than measurements performed on disturbed cores because of the high sensitivity of the structure. Measurements of saturated hydraulic conductivity (Ks ) by the constant-head or the falling-head method may be invalidated by the presence of large particles in the peat sample. Also, mathematical models developed especially to estimate saturated hydraulic conductivity from field measurements might not apply to organic soils (Hemond and Goldman 1985). Verification of the applicability of Darcy’s law in peat substrates has however led to the conclusion that low flow is laminar and Darcy’s approach applies (Allaire et al. 1994). If so, then measurements made directly in the field using the Guelph permeameter are possible above the water table but may require an independent estimation of a (a parameter used to characterize the unsaturated hydraulic conductivity). Organic soils in our studies have yielded estimates of a of about 9 m1 on internal drainage tests (J. Caron, unpublished data). These values are comparable to that of some peat substrates (about 7---9 m1 ) (Caron et al. 1998; Jobin et al. 2004), but may be higher. Saturated hydraulic conductivity values sometimes are comparable to that of sandy loam or loam soils, which is surprising, but may be linked to their low pore effectiveness (Caron and Nkongolo 2004). Readers are referred to Ks measurement procedures for mineral soils (see Section ‘‘Soil Water Analyses’’) for further details on Ks determinations, either in the field or in the laboratory (if the decision is made to use cores).

68.5.2 SATURATED HYDRAULIC CONDUCTIVITY OF POTTED GROWING MEDIA The particularly sensitive structure of growing media makes it highly advisable to measure the saturated hydraulic conductivity on undisturbed substrates. The measurement is sometimes performed with plants present, which incorporate changes in the structure of decomposition, reorganization, settling, and root enmeshment, all of which are known to affect Ks determinations (Allaire-Leung et al. 1999). The procedure uses constant-head permeameter devices to establish steady-state conditions and obtains flow measurements in a container (Figure 68.2). Since these flow measurements are affected by the container geometry and hole distribution (a 3-D process), the flow is corrected to provide an estimate of the equivalent saturated hydraulic conductivity from Darcy’s law as if it was a 1-D process measured in a cylinder. Performing these measurements in a cylinder may be problematic



901

Graduated water filled cylinder

Water flow

Air entry Water height

Open ring

Substrate surface

FIGURE 68.2. Schematic representation of the special head of the Coˆte´’s infiltrometer to measure saturated hydraulic conductivity.

because of the very sensitive structure, and important deformations affecting the Ks determination have been observed. Allaire et al. (1994) have instead proposed correcting the flow from the measurement performed directly in the pot and have shown this approach to be appropriate. Obviously, the container geometry, hole distribution, and hydraulic gradient affect the correction factor. If no correction factor is available, Allaire et al. (1994) describe an experimental procedure on how to derive it from solving Laplace’s equation in 3-D. Materials 1

A constant-head permeameter (Coˆte´’s infiltrometer, Figure 68.2), with a special head that can be moved directly onto a pot surface (Banton et al. 1991)

2

An interpretation chart for different container heights and pot-hole distributions (Allaire et al. 1994)

3

A ruler

4

A substrate field container, with or without the actively growing plant

5

A water pail

Procedure 1

Immerse containers in a distilled-water or tap-water filled bath for 24 h by slowly rewetting the sample from underneath. A prewetting period may be necessary if hydrophobicity is suspected.

2

Take the container slowly off the bath and put it onto a metal stand having a perforated surface.



902 3

If there is an empty space between the container side and the substrate, fill the gap with bentonite to avoid preferential flow.

4

Cover the substrate surface with a screen or a porous pad (scouring pad) to prevent particles from floating and plugging the opening of the infiltrometer or to reduce particle displacement.

5

Establish steady state when maintaining a known height of water above the substrate surface using the infiltrometer.

6

Measure the water drop in the permeameter as a function of time.

7

Calculate flux after reaching steady-state conditions (Q).

8

Determine the known height of water above the substrate surface (h).

9

Determine the final height after running the experiment (L).

10

Calculate the substrate surface area (A).

11

Find the flux reduction ratio obtained from Figure 5 in Allaire et al. (1994). For 1 L pots with multiple holes (type Ultra) commonly found in greenhouses, we determined that the Rf value is equal to 1.66 with a substrate height of 7 and 3 cm of water head.

12

Calculate the saturated hydraulic conductivity Ks from Ks ¼

QLRf A(h þ L)

(68:16)

68.5.3 UNSATURATED HYDRAULIC CONDUCTIVITY ON SOIL CORES AND POTTED SUBSTRATES Instantaneous profile methods (Watson 1966; Wind 1969; Hillel et al. 1972; Vachaud and Dane 2002) have successfully been used in the field at different depths for characterizing the whole-unsaturated hydraulic conductivity curve on both drainage and rewetting. The approach has been used for potted substrates and on soil cores (Naasz et al. 2005) and is presented here. The method is highly variable, close to saturation (1 to 0 kPa) (Caron and Elrick 2005; Naasz et al. 2005), but provides useful and rapid estimates in the range of water availability (1 to 20 kPa) in this kind of substrate. Unsaturated hydraulic conductivity (using constant-head methods) is not accurate enough for an adequate characterization at potentials lower than 2 kPa, and therefore is not presented herein. Materials 1

Two large PVC cylinders (14 cm diameter and 14 cm height).

2

Small PVC cylinders (10 cm diameter and 12 cm height, V ¼ 942 cm3 ).



903

3

A TDR system: two TDR miniprobes (three wires, 80 mm long, with 4-mm uncoated stainless steel rods and a spacing of 10 mm). The probes are connected to a Tektronix 1502C system (Tektronix, Beaverton, OR) via a multiplexer run by WINTDR software (Time Domain Reflectometry Soil Sample Analysis Program V. 6.1, Utah State University, Logan, Utah).

4

A tensiometer system: two minitensiometers (2.2 mm diameter and 20 mm length, ceramic cell, SDEC 220 [Socie´te´ De´veloppement et Commercialization, Reignac, France]) are connected to pressure transducers (differential pressure sensors, precision: 0:03%, response time: 103 s) monitored by a real-time multitasking computer to control the measurement and to collect data.

5

Small ventilators to impose top-surface sample evaporation.

6

A Mariotte bottle to maintain a constant-head bottom infiltration.

Procedure 1

Since the physical properties of organic substrates are largely influenced by preparation and, more precisely, by the packing of materials, the procedure of substrates preparation is standardized: manually fill with substrate (but without packing) two large PVC cylinders (14 cm diameter and 14 cm height). Slowly wet cylinders (30 min) from the bottom, saturate with distilled water for 24 h, and allow samples to equilibrate for 48 h to a water potential of 5 kPa (on a tension table). Empty cylinders, homogenize substrate, and fill small PVC cylinders (10 cm diameter and 12 cm height) with substrate without packing. Finally, slowly saturate material from the bottom for 24 h.

2

Horizontally insert (at an angle of 90 ) the two pairs of sensors (TDR probes and minitensiometers to determine volumetric water content u and the water potential c, respectively) at two levels, h1 and h2 , from the bottom of the small PVC cylinder (h1 ¼ 9 cm and h2 ¼ 3 cm, Figure 68.3).

3

Seal the bottom of the column to prevent water loss and then slowly saturate the substrate from the bottom with the Mariotte bottle.

4

When readings (TDR and tensiometers) indicate that the core is in hydrostatic equilibrium (after about 30 min), the evaporation experiment could begin.

5

Impose controlled top-surface sample evaporation with the small ventilators.

6

The evaporation experiment is terminated when the uppermost tensiometer in the substrate core reaches a suction level of approximately 30 kPa.

7

The sample is then subjected to infiltration. During the infiltration experiment, impose three pressure levels (stepwise increases): 15, 5, 0 kPa.

8

The drying–wetting cycle lasts for approximately 1 month and represents 8000 sets of water content=water potential data (measurements were taken every 5 min).



904 Fan

Mariotte reservoir bottle

Substrate column H1 TDR moisture miniprobes

Minitensiometers H2

∆H

FIGURE 68.3. Schematic representation of the experimental design for measuring the unsaturated hydraulic conductivity in cores. (Redrawn from Naasz, R., Michel, J.-C., and Charpentier, S., Soil Sci. Soc. Am. J., 69, 13, 2005.)

Calculations Calculate the unsaturated hydraulic conductivity by direct measurement from water content and water pressure data obtained during evaporation and infiltration experiments with the Darcy equation. First, calculate the water flow through the column q from the temporal changes (Dt) in water storage at two depths (h1 and h2 ) as follows: q¼

Duh1 Duh2 Dt

(68:17)

where Dt ¼ t2 t1 is the time interval; t1 is time 1 and t2 is time 2. The unsaturated hydraulic conductivity k(c) is then obtained by dividing the flux density calculated above with the matric head differences (dc) at the same positions (dz ¼ h1 h2 ) and times as follows: k(c) ¼

q dc 1 dz

(68:18)

Comments The introduction of TDR miniprobes and minitensiometers in the substrate column could possibly lead to a small change in the structure of materials and, as a result, in the hydraulic properties. The volume occupied by all sensors in the column has been calculated and only represents cFC drains away too quickly to be captured by plant roots, whereas water at cm < cPWP is held too strongly by the soil to be extracted by the roots (compare PWP discussion). The proposed minimum PAWC for optimum plant growth and minimum susceptibility to droughtiness is 0.20–0.30 m3 m3 (Verdonck et al. 1983; Cockroft and Olsson 1997; Bilderback et al. 2005). Recent research (Olness et al. 1998; Reynolds et al. 2002) suggests that the optimal balance between root-zone soil water and soil air is achieved in rain-fed crops when FC=Porosity ¼ 0:66

(69:7)

AC=Porosity ¼ 0:34

(69:8)

or alternatively, when

These criteria are based on the finding that maximum production of crop-available nitrogen by aerobic microbial mineralization of organic matter occurs when about 66% of the soil pore space in the root zone is water-filled, or alternatively, when 34% of the pore space is airfilled (Skopp et al. 1990). The rationale for applying Equation 69.7 and Equation 69.8 to rain-fed crops is that root-zone soils with these ratios are likely to have desirable water and air contents (for good microbial production of nitrogen) more frequently and for longer periods of time (especially during the critical early growing season) than root-zone soils that have larger or smaller ratios.

69.4.3 DETERMINATION OF DESORPTION AND IMBIBITION CURVES The generally accepted ‘‘ideal’’ for obtaining soil water desorption and imbibition curves is to collect simultaneous field-based measurements of volumetric water content, uv , and


Soil Water Analyses: Principles and Parameters

921

matric head, cm , in an undisturbed vertical profile under conditions of steady drainage (desorption) or steady wetting (imbibition). Several approaches are available for achieving this (e.g., Bruce and Luxmoore 1986), with the most popular approach being the ‘‘instantaneous profile’’ method (see Chapter 83). Several factors inhibit or complicate the field-based methods, however, including complex and poorly controlled boundary conditions (e.g., varying water table depth, strong and varying temperature gradients); limited instrumentation for determining cm (e.g., tensiometers have a narrow operating range and often fail after a period of time); difficulty in maintaining continuous wetting or drainage throughout the soil profile (e.g., periodic rainfalls can induce hysteretic effects); complicated and labor-intensive experimental setups (e.g., installation of many pairs of uv and cm sensors over a substantial depth range with minimum soil disturbance, equipment for applying large volumes of water to saturate the soil profile, complex electronics and data logging equipment for simultaneous and long-term monitoring of uv and cm , limited ability for spatial replication); and potentially very long measurement times (it can take several weeks to months to obtain adequate desorption or imbibition curve over the required soil depth because of slow wetting and drainage rates). As a result, experimentally determined desorption and imbibition curves are usually obtained in the laboratory on relatively small soil cores or columns where uv and cm sensors are more easily installed and maintained, and where initial and boundary conditions can be precisely defined and controlled. Desorption and imbibition curves can also be estimated from basic soil data via pedotransfer functions (see Chapter 84); from flow experiments, such as the evaporation method (see Chapter 81) and the instantaneous profile method (see Chapter 83); or from inverse modeling procedures (Hopmans et al. 2002). Laboratory determination of desorption and imbibition curves that are representative of field conditions requires (i) the collection of soil cores or columns that are large enough to adequately sample the antecedent soil structure and (ii) use of collection, handling, and analysis procedures that maintain the soil structure intact. Bouma (1983, 1985) suggests that the volume encompassed by the core=column should include at least 20 soil structural units (e.g., peds, worm holes, abandoned root channels, etc.), which is especially important for the cm > 3:3 m range and if saturated hydraulic conductivity (see Chapter 75) is to be determined on the same sample. For relatively structureless sandy soils, the minimum recommended core=column inside diameter and length is on the order of 7.6 cm, whereas structured loamy and clayey soils should use a core length and diameter of at least 10 cm (McIntyre 1974). The samples should be collected when the soil is near its field capacity water content, uFC , which generally makes the soil strong enough to resist compaction and structural collapse during core=column insertion, but still plastic enough to prevent shattering and breakage of peds. Recommended procedures for the collection of minimally disturbed soil samples are given in McIntyre (1974) and Chapter 80. Excavated soil cores should be trimmed flush with the ends of the sampling cylinder, capped to prevent damage of the core ends, wrapped in plastic to prevent evaporation, and transported to the laboratory in cushioned coolers to minimize vibration-induced damage and large temperature-changes. Sample storage before analysis should be in darkened facilities maintained at 0 C4 C, which is cold enough to inhibit faunal–bacterial–fungal–algal activity, but not so cold as to cause freezing and ice lens formation. Soil water desorption–imbibition methods are described in Chapter 72 through Chapter 74 and include the tension table, tension plate, and pressure extractor methods (Chapter 72), the long column method (Chapter 73), and the dew point psychrometer method (Chapter 74). The approximate matric head ranges of these methods are compared in Figure 69.4.



922

Long column Soil water content, q (m3 m−3)

qs Low tension table/plate High tension table/plate Pressure plate extractor Dew point psychrometer

qr −10000

−1000

−100

−10

Matric head,

−1

− + 0

(m)

FIGURE 69.4. Approximate matric head ranges of the long column, tension table, tension plate, pressure extractor, and dew point psychrometer methods for measuring desorption and imbibition curves. us is the saturated water content and ur is the residual water content. These methods are described in Chapter 72 through Chapter 74.

69.5 SATURATED HYDRAULIC PROPERTIES The saturated hydraulic properties are used to describe and predict water movement in permeable porous material (e.g., soil, building fill, sand, rock, etc.) when the pore water pressure (or matric) head in the material is greater than or equal to the water-entry value or air-entry value (see Section 69.4 for explanation of water-entry and air-entry values). The saturated soil hydraulic properties of greatest relevance include saturated hydraulic conductivity, field-saturated hydraulic conductivity, and the so-called capillarity parameters such as matric flux potential, sorptivity, sorptive number, Green–Ampt wetting front pressure head, and FWM pore size and pore number. Saturated hydraulic conductivity, Ks [LT1 ], and field-saturated hydraulic conductivity, Kfs [LT1 ], are measures of the ‘‘ease’’ or ‘‘ability’’ of a permeable porous medium to transmit water. The Ks parameter applies when the water-conducting pores in the porous medium are completely water-filled (saturated), and the Kfs parameter applies when the water-conducting pores contain entrapped or encapsulated bubbles of air or gas (field-saturated). The capillarity parameters measure various aspects of the suction or ‘‘capillary pull’’ that unsaturated soil exerts on infiltrating water; and measurement or estimation of the soil’s capillarity is usually required when Ks or Kfs are measured in initially unsaturated soil (e.g., soil above the water table). The Ks and Kfs parameters are discussed below and the capillarity parameters are discussed in Section 69.6 (unsaturated hydraulic properties). The Ks and Kfs parameters are defined by Darcy’s law, which may be written in the form q ¼ Ksat i

(69:9)

where q is the water flux density through the porous medium (volume of water flowing through a unit cross-sectional area of porous medium per unit time), i is the hydraulic head gradient in the porous medium (dimensionless), and Ksat ¼ Ks or Kfs , depending on whether



923

the porous medium is completely saturated or field-saturated, respectively. As implied by Equation 69.9, the dimensions of Ksat are the same as those for q (i.e., volume of water per unit cross-sectional area of flow per unit time); however, these dimensions are usually simplified to length per unit time so that Ksat may be expressed in the more convenient (but physically incorrect) units of velocity (i.e., cm s1 , cm s1 , cm h1 , m days1 , etc.). The Ksat value is a constant when the porous medium is rigid, homogenous, isotropic, and stable; when in-situ biological activity such as earthworm burrowing and algal=fungal growth are negligible; and when the flowing water maintains constant physical and chemical properties (e.g., temperature, viscosity, dissolved air content, dissolved salt content, etc.) and does not chemically or physically interact with the porous medium. The primary factors determining the magnitude of Ksat include the physical characteristics of the porous medium and the physical and chemical characteristics of the flowing water (discussed further below). The physical characteristics of the porous medium affecting Ksat include the size distribution, roughness, tortuosity, shape, and degree of interconnectedness of the water-conducting pores. For soils, Ksat increases greatly with coarser texture (larger grain sizes), increasing numbers of biopores (e.g., worm holes, root channels), and increasing structure (e.g., aggregates, interpedal spaces, shrinkage cracks), as these factors increase the number of water-conducting pores that are relatively large, straight (i.e., low tortuosity), smooth, rounded, and interconnected. Soils and other porous media that are coarse-textured, structured, and bioporous consequently tend to have larger Ksat values than those that are fine-textured, structureless, and devoid of biopores. In addition, texture, structure, and biopores can interact in such a way that it is not uncommon for a fine-textured material with structure or biopores (e.g., a clay soil with shrinkage cracks or worm holes) to have a substantially larger Ksat than a coarse-textured material that is devoid of structure and biopores (e.g., single-grain sandy soil). An important implication of this texture–structure– biopore interaction is that the physical condition of the porous medium must be preserved by the measuring technique in order for the measured Ksat value to be representative of the porous medium in its ‘‘natural’’ or in-situ condition. Hydraulic conductivity is inversely related to water viscosity, which is inversely related to temperature (Bouwer 1978, p. 43). Consequently, the measured value of Ksat will increase with the temperature of the water used; and an increase in water temperature from 10 C to 25 C will result in a 45% increase in Ksat , all other factors remaining equal. Temperature effects can be important if the water used in a field measurement differs greatly in temperature from that of the resident soil water or groundwater, or if laboratory measurements of field samples (e.g., intact cores) are conducted at temperatures that differ greatly from the field temperature. Precise measurements and comparisons of Ksat values should therefore always be referenced to a specific water temperature, which is usually 20 C (Bouwer 1978, p. 43), as it yields a water viscosity of nearly 1 cP. Note in passing that the temperature of ‘‘deep’’ soil water and shallow groundwater is fairly constant and close to the local mean annual air temperature, for example, about 10 C at 408N–458N latitude (Bouwer 1978, p. 378). The concentration and speciation of dissolved salts in the water can affect Ksat through swelling, flocculation, or dispersion of silt and clay within the porous medium, and through the creation or dissolution of precipitates. The Ksat value will usually increase if silt and clay particles are flocculated, or if precipitates are dissolved, as this tends to increase the size and interconnectedness of water-conducting pores. Alternatively, formation of precipitates and swelling=dispersion of silt and clay particles will usually decrease Ksat through narrowing and plugging of pores. Reduction in Ksat most commonly occurs in silt- and clay-rich soils


924


when the cationic speciation is changed or the concentration of the resident soil water is diluted by incoming rainfall, irrigation water, or groundwater. The relative concentrations of sodium, calcium, and magnesium in solution and sorbed onto the porous medium exchange sites are particularly important in this respect (Bouwer 1978, p. 44). In extreme cases, such as when water low in dissolved salts (e.g., rainwater) is introduced into saline soil, the resulting silt and clay dispersion can reduce Ksat to virtually zero. The water used for measuring the Ksat of a natural porous medium should therefore be either ‘‘native’’ water extracted from the porous medium, or a laboratory ‘‘approximation,’’ which has about the same major ion composition and concentrations as the native water. Local municipal tap water is often an adequate approximation to native soil water, although this should always be checked as some municipal water treatment facilities can change major ion chemistry radically. Distilled or deionized water should never be used for measuring the Ksat of a natural porous medium, as it will almost always induce clay swelling or dispersion of silt and clay particles. Entrapped bubbles tend to constrict or block the water-conducting pores in a porous medium. As a result, Kfs (i.e., field-saturated Ksat ) is usually less than Ks (i.e., completely saturated Ksat ) with the degree of reduction largely dependent on the mechanism responsible for bubble formation. Bubbles can become encapsulated in pores through physical entrapment of resident air during wetting of an initially unsaturated porous medium (Bouwer 1966); by accumulation of biogases (e.g., methane) as a result of microbial activity (Reynolds et al. 1992); and by ‘‘exsolution’’ of dissolved air as a result of changes in the temperature or chemistry of the pore water (Bouwer 1978, p. 45). Air encapsulation as a result of rapid wetting (e.g., ponded infiltration) often causes Kfs to be on the order of 0:5 Ks (Bouwer 1966; Stephens et al. 1987; Constantz et al. 1988), while gradual accumulation of biogases and exsolved air can cause much greater reductions (Bouwer 1978; Reynolds et al. 1992). Further information concerning the theoretical basis and other aspects of Ks , Kfs , and their associated capillarity parameters can be obtained from Bouwer (1978), Koorevaar et al. (1983), Smith (2002), Reynolds and Elrick (2005), and references contained therein. Saturated hydraulic property methods are described in Chapter 75 through Chapter 79 and Chapter 84; and they include the constant and falling head core methods (Chapter 75), selected constant and falling head well permeameter methods (Chapter 76), selected constant and falling head ring infiltrometer methods (Chapter 77), the auger hole method (Chapter 78), the piezometer method (Chapter 79), and selected estimation methods (Chapter 84).

69.6 UNSATURATED HYDRAULIC PROPERTIES Unsaturated hydraulic properties are used to describe and predict water movement in permeable porous material (e.g., soil, building fill, sand, rock, etc.) that is only partially saturated and has a pore water matric head that is less than the material’s air-entry value or water-entry value (see Section 69.4 for explanation of air-entry and water-entry values). The unsaturated hydraulic properties of greatest relevance include unsaturated hydraulic conductivity, K(c) or K(u) [LT1 ], sorptivity, S(c) [LT1=2 ], sorptive number, a*(c) [L1 ], flux potential, f(c) [L2 T1 ], FWM pore diameter, PD(c) [L], and the number of FWM pores per unit area, NP(c) [L2 ]. The K(c) or K(u) parameter quantifies the ability of an unsaturated porous material to transmit water as a result of a hydraulic head gradient, while S(c) measures the ability of the material to imbibe water as a result of capillarity forces (Philip 1957). The a*(c) parameter, on the other hand, indicates the relative magnitudes of gravity and capillarity forces during unsaturated flow (Raats 1976),



925

while the f(c) parameter relates to the ‘‘potential’’ for water flow (Gardner 1958). The PD(c) parameter represents the effective equivalent mean pore size conducting water during constant head infiltration, and NP(c) indicates the number of PD(c) pores that are active (Philip 1987). These parameters and their interrelationships are discussed briefly below. Vertical water flow in rigid, homogeneous, variably saturated porous material (e.g., soil) can be described by (Richards 1931) @u @ @H @ @H ¼ K(c) ¼ K(u) ; H ¼cþz @t @z @z @z @z

(69:10)

where u [L3 L3 ] is volumetric water content, t[T] is time, K(c) [LT1 ] is the hydraulic conductivity (K ) versus pore water matric head (c) relationship, K(u) [LT1 ] is the hydraulic conductivity (K ) versus volumetric water content (u) relationship, H [L] is hydraulic head, and z [L] is elevation or gravitational head above an arbitrary datum (positive upward). (Note that the ‘‘v’’ and ‘‘m’’ subscripts on u and c, respectively, have been dropped to simplify the nomenclature.) Equation 69.10 indicates that the rate of water flow through the porous medium is determined by the magnitude of the hydraulic head gradient, @H=@z, and by the hydraulic conductivity function, K(c) or K(u). The K(c) or K(u) term is the porous material’s water transmission relationship, and it gives the permeability of the porous material to water as a function of either pore water matric head, c [L], or volumetric water content, u [L3 L3 ]. The K(c) and K(u) relationships depend strongly on the magnitude and shape of the pore water desorption–imbibition relationship, u(c) [L3 L3 ], which itself describes the change in volumetric water content with changing pore water matric head (Section 69.4). As a result, the K(c) and K(u) relationships decrease from the Ksat maximum (Section 69.5) as c and u decrease from their respective maximum values at porous medium saturation (i.e., c ¼ 0 and u ¼ us ). Through their connection with the u(c) relationship, K(c) and K(u) depend on the number and size distribution of the porous medium pores, which in turn depend on porosity, structure, texture, organic matter content, and clay mineralogy. Unlike u(c), however, K(c) and K(u) also depend on pore morphology parameters such as tortuosity, roughness, connectivity, and continuity. These various dependencies cause K(c) and K(u) to change by many orders of magnitude over the range in c applicable to plant growth (i.e., 150 m c 0). Due to the extreme sensitivity of unsaturated hydraulic conductivity to pore size and pore morphology, the magnitude and shape of the K(c) and K(u) relationships change substantially with the texture and structure of the porous medium. Figure 69.5 gives schematic examples of K(c) and K(u) relationships for a representative ‘‘sandy’’ soil, and for a representative ‘‘loamy’’ soil with and without structure, where structure refers to the presence of aggregates, peds, cracks, root channels, worm holes, etc. For convenience, the structured loam was assumed to have the same u(c) relationship as the unstructured loam. Note in these figures that for a rigid (nonswelling) porous material, K(c) and K(u) are maximum and constant when the material is saturated, i.e., K(c) ¼ K(u) ¼ constant ¼ Ksat ; c ce , u ¼ usat

(69:11)

where Ksat [LT1 ] is the saturated or field-saturated hydraulic conductivity, ce [L] is the airentry or water-entry matric head, and usat [L3 L3 ] is saturated or field-saturated volumetric water content (see Section 69.4 and Section 69.5). Note also that the near-saturated hydraulic



926

Hydraulic conducitvity, K( ) (cm s–1)

1e+0 1e− 1

Sand Loam Structured loam

1e− 2 1e− 3 1e− 4 1e− 5 1e− 6 1e− 7 1e− 8 −300

− 250

− 200

− 150

−100

Pore water pressure head,

(a)

−50

0

50

0.6

0.7

(cm)

Hydraulic conducitvity, K(q ) (cm s−1)

1e + 0 Sand Loam Structured loam

1e − 1 1e − 2 1e − 3 1e − 4 1e − 5 1e − 6 1e − 7 1e − 8 0.0

(b)

0.1

0.2

0.3

0.4

0.5

Volumetric water content, q (cm3 cm−3)

FIGURE 69.5. (a) Hydraulic conductivity, K (c), versus pore water matric (or pressure) head, c and (b) hydraulic conductivity, K (u), versus volumetric water content, u, for a representative sandy soil (Sand), and a representative loamy soil with structure (structured loam) and without structure (loam).

conductivity relationship in a structured porous medium can change very rapidly (by orders of magnitude) with only small changes in c or u, and that the hydraulic conductivity of a fine-textured material with structure can be either greater than or less than the hydraulic conductivity in a coarse-textured material, depending on the value of c or u. Texture and structure effects are also illustrated in the Ksat values, where it is seen that the Ksat of the sandy soil is two orders of magnitude greater than the Ksat of the unstructured loam (texture effect), but two orders of magnitude less than the Ksat of the structured loam (structure effect).



927

The sorptivity parameter, S(c) [LT1=2 ], is related to K(c) and f(c) by (Philip 1957; White and Sully 1987) " S(c0 ) ¼ g ½uðc0 Þ uðci Þ

#1=2

ð c0 K(c)dc

¼ ½g½uðc0 Þ uðci Þfðc0 Þ1=2 ;

ci

u(ci ) u(c0 ) us ,

c i c0 0

(69:12)

where it is seen that the matric flux potential, f(c0 ) [L2 T1 ], is defined by (Gardner 1958) f(c0 ) ¼

ð c0 K(c)dc;

1 < ci c0 0

(69:13)

ci

In Equation 69.12 and Equation 69.13, c0 [L] is the pore water matric head at the infiltration (sorption) surface, ci [L] is the background or antecedent pore water matric head in the porous medium at the time of the infiltration measurement, u(c0 ) [L3 L3 ] is the porous medium volumetric water content at c ¼ c0 , u(ci ) [L3 L3 ] is the porous medium volumetric water content at c ¼ ci , and g ¼ 1:818 is a dimensionless empirical constant (White and Sully 1987) related to the shape of the wetting (or drainage) front (g ¼ 1:818 for wetting, but may be smaller for drainage). The shape and magnitude of the S(c0 ) and f(c0 ) relationships is thus controlled by the shape and magnitude of the K(c) relationship, as well as the magnitude of ci . Figure 69.6 gives the S(c0 ) and f(c0 ) relationships corresponding to the K(c) (and u(c)) relationships for our three representative soils, and it is seen that S(c0 ) and f(c0 ) are essentially ‘‘subdued replicas’’ of K(c). Note from Equation 69.12 and Equation 69.13, however, that S(c0 ) ¼ f(c0 ) ¼ 0 when u(c0 ) ¼ u(ci ) or when c0 ¼ ci ; and that S(c0 ) and f(c0 ) do not exist for positive pore water pressure heads (i.e., cp > 0). If the K(c) relationship is represented by the Gardner (1958) exponential function K(c) ¼ Ksat exp (ac);

c0

(69:14)

then Equation 69.13 becomes f(c0 ) ¼

K(c0 ) K(ci ) ; ci < c0 , K(ci ) < K(c0 ) a(c0 )

(69:15)

where the ‘‘alpha parameter,’’ a(c0 ) [L1 ] gives the slope of ln K versus c. For most natural porous materials at field capacity or dryer, K(ci ) K(c0 ), and Equation 69.15 can consequently be simplified to f(c0 )

K(c0 ) ; a*(c0 )

K(ci ) K(c0 )

(69:16)

which defines the ‘‘sorptive number,’’ a*(c0 ) [L1 ]. The a*(c0 ) parameter is generally used rather than a(c0 ) because it avoids having to determine K(ci ) in Equation 69.15, which can be extremely difficult or impossible. Large a(c0 ) and a*(c0 ) values indicate dominance of the gravitational force (gravity) over the porous medium adsorption forces (capillarity) during infiltration, whereas small a(c0 ) and a*(c0 ) values indicate the reverse (Raats 1976). The a(c0 ) relationships corresponding to our three representative soils are given in



928 1

Sorptivity, S( ) (cm s−1/2)

Sand Loam Structured loam 0.1

0.01

0.001

0.0001 − 140

− 120

− 100

−80

−60

− 40


(a)

−20

0

20

0

20

(cm)

Matric flux potential, f ( ) (cm2 s−1)

1e+0 1e− 1 1e− 2 1e− 3 1e− 4 1e− 5 1e− 6 −140 (b)

Sand Loam Structured loam

− 120

− 100

− 80

− 60

− 40


− 20 (cm)

FIGURE 69.6. (a) Sorptivity, S(c), versus pore water matric (or pressure) head, c, and (b) matric flux potential, f(c), versus pore water matric head, c, for a representative sandy soil (sand), and a representative loamy soil with structure (structured loam) and without structure (loam).

Figure 69.7a; and generally speaking, a(c0 ) increases as c0 increases, indicating an increase in the importance of the gravity component of infiltration relative to the capillarity component as the soil gets wetter. Note, however, that the a(c0 ) relationships have complex slopes, and the sand and unstructured loam produce curves with local maxima and minima. This occurs because a(c0 ) is based on the exponential K(c) function (i.e., Equation 69.14), whereas the actual K(c) relationships were not exponential, especially those for the sand and unstructured loam (see Figure 69.5a). Generally speaking, the closer the K(c) relationship is to a monotonic exponential function (i.e., Equation 69.14), the closer the a(c0 ) relationship is to a single constant value. Figure 69.7b compares a*(c0 ) to a(c0 ) for the



929

Alpha parameter, a ( ) (cm−1)

1 Sand Loam Structured loam 0.1

0.01

0.001

0.0001 −300

−250

Alpha, a (y ), or sorptive number, a ∗( ) (cm−1)

(a)

(b)

− 200

− 150

− 100


− 50

0

(cm)

1

0.1

0.01 a( ) a ∗( ) 0.001 − 300

−250

− 200

− 150

− 100


− 50

0

(cm)

FIGURE 69.7. (a) Alpha parameter, a(c), versus pore water matric (or pressure) head, c, for a representative sandy soil (sand) and a representative loamy soil with structure (structured loam) and without structure (loam) and (b) alpha parameter, a(c), and sorptive number, a*(c), versus pore water pressure head, c, for the structured loamy soil.

structured loam, where it is seen that a*(c0 ) diverges progressively for c0 < 50 cm. This occurred because K(ci ) ¼ K(260 cm) in this scenario, and the assumption K(ci ) K(c0 ) became progressively more incorrect as c0 decreased, resulting in increasing error in a*(c0 ) with smaller (more negative) c0 values. The a*(c0 ) parameter (and relationships based on the a*(c0 ) parameter) must consequently be used with caution when K(ci ) is not substantially less than K(c0 ), such as might occur in very wet porous materials, or in fine-textured materials where K(c) does not decrease rapidly with decreasing c.



930

Substituting Equation 69.16 into Equation 69.12 produces

K(c0 ) 1=2 S(c0 ) ¼ g[u(c0 ) u(ci )] a*(c0 )

(69:17)

which shows that the ability of a porous medium to imbibe water (i.e., its sorptivity as indicated by the magnitude of S(c0 )) depends on the available water-storage capacity (u(c0 ) u(ci )), the K(c) relationship, and the a*(c0 ) relationship. Hence, a porous material’s sorptivity decreases with increasing antecedent water content (i.e., decreasing available water-storage capacity), decreasing hydraulic conductivity, and increasing sorptive number. Note also that the accuracy of Equation 69.17 will depend strongly on the accuracy of the a*(c0 ) relationship, as discussed above. The FWM pore diameter, PD(c0 ) [L], is defined as (Philip 1987) PD(c0 ) ¼

2sK(c0 ) 2sa*(c0 ) ¼ rgf(c0 ) rg

(69:18)

where s [MT2 ] is the air–pore water interfacial surface tension, r [ML3 ] is the pore water density, and g [LT2 ] is the acceleration due to gravity. The PD(c0 ) parameter is often referred to as the effective ‘‘equivalent mean’’ pore diameter conducting water when infiltration occurs at c0 (White and Sully 1987). It may be more accurate, however, to view PD(c0 ) as an index parameter that represents the mean ‘‘water-conductiveness’’ of the hydraulically active pores, rather than an actual pore size. This is because the PD(c0 ) parameter is derived from a flow measurement (associated with the measurement of K(c0 ); Equation 69.18), and must consequently reflect in some way the combined sizes, tortuosities, roughnesses, and connectivities of all water-conducting pores at c ¼ c0 (Reynolds et al. 1997). Associated with PD(c0 ) is the ‘‘concentration’’ of pore sizes, NP(c0 ) (number of pores L2 ), which may be derived from Poiseuille’s law for flow in smooth, cylindrical capillary tubes (Philip 1987): NP(c0 ) ¼

128mK(c0 ) rg[PD(c0 )]4

(69:19)

where m [ML1 T1 ] is the dynamic viscosity of water and the other parameters are as defined above. The NP(c0 ) parameter is an indicator of the number of hydraulically active pores per unit area of infiltration surface, which have FWM diameter, PD(c0 ). The relationships among PD(c0 ), NP(c0 ), and K(c0 ) for the structured loam soil are illustrated in Figure 69.8, where it is seen that a two-order of magnitude increase in flow-weighted mean pore diameter, PD(c0 ), corresponded to about a six-order of magnitude increase in K(c0 ), and about a four-order of magnitude decrease in NP(c0 ). Equation 69.14 through Equation 69.19 also apply when measuring saturated flow parameters in unsaturated porous materials (see Section 69.5). In this case, c0 is at its maximum value in the equations (i.e., c0 ¼ 0), and consequently the K(c), f(c0 ), a*(c0 ), u(c0 ), S(c0 ), PD(c0 ), and NP(c0 ) relationships become maximum-valued constants, which are indicated by Ksat (i.e., Ks or Kfs ), fm , a*, usat (i.e., us or ufs ), S, PD, and NP, respectively. As mentioned in Section 69.5, the matric flux potential (fm ), sorptive number (a*), and sorptivity (S) are measures of the capillary suction=pull or ‘‘capillarity’’ that unsaturated hydrophilic porous materials exert on infiltrating water. Mathematically, fm is the area under the K(c) curve between c ¼ c0 ¼ 0 and c ¼ ci (Equation 69.13); and as a result, the magnitude of a


K( ) (cm s−1) or NP( ) (pores m−2)

Soil Water Analyses: Principles and Parameters 1e+ 8 1e+ 7 1e+ 6 1e+ 5 1e+ 4 1e+ 3 1e+ 2 1e+ 1 1e+ 0 1e− 1 1e− 2 1e− 3 1e− 4 1e− 5 1e− 6 1e− 7 1e− 8 0.001

931

K( ) NP( )

0.01

0.1

1

10

Flow-weighted mean pore diameter, PD( ) (mm)

FIGURE 69.8. Hydraulic conductivity, K (c), and number of FWM pores per unit area, NP(c), versus FWM pore diameter, PD(c), for the structured loamy soil.

material’s capillarity depends on the shape and magnitude of the K(c) curve, and on the antecedent pore water matric head, ci . Porous media that are coarse-textured, structured, bioporous, or wet consequently tend to have lower capillarity (i.e., smaller area under the K(c) curve) than porous media that are fine-textured, structureless, dry, or devoid of biopores. Furthermore, all porous media (regardless of texture or structure) have zero capillarity (i.e., fm ¼ 0) when they are saturated or field-saturated because under that condition, c0 ¼ ci ¼ 0 in Equation 69.13. If the K(c) function is represented by Equation 69.14, it can be shown that for porous materials at field capacity or drier (Mein and Farrell 1974; Scotter et al. 1982; Reynolds et al. 1985; see also Section 69.4): a a* (Ksat =fm ) c1 f ; cf < 0 < a*

(69:20)

where a* [L1 ] is the maximum sorptive number (for the material in question) and cf [L] is the Green–Ampt wetting front matric head (negative quantity). Near-zero cf (large a*) occurs primarily in porous materials that are coarse-textured and=or highly structured and=or highly bioporous, while large negative cf (small a*) occurs primarily in materials that are fine-textured or structureless or devoid of biopores. When c0 ¼ 0, the S, fm , Ksat , a*, and cf parameters are related by S ¼ [g(ufs ui )fm ]

1=2

Ksat ¼ g(ufs ui ) a*

1=2

¼ [g(ui ufs )Ksat cf ]1=2

(69:21)

where ufs [L3 L3 ] is the field-saturated volumetric water content (Section 69.4), ui [L3 L3 ] is the initial or antecedent volumetric water content, and the other parameters are as previously defined. Note that in Equation 69.21, S decreases to zero as ui increases to ufs , indicating (as expected) that field-saturated porous material has no ability to absorb or store



932

additional water. The PD and NP parameters (Equation 69.18 and Equation 69.19, respectively) are often used to quantify temporal and management-induced changes in porous medium structure as they relate to water flow (e.g., White et al. 1992; Reynolds et al. 1995). In structured porous materials, it is often important to distinguish between ‘‘matrix’’ flow parameters and ‘‘macropore’’ flow parameters, given that macropores (e.g., large cracks, worm holes, abandoned root channels, large interaggregate spaces, etc.) can have a substantial effect on near-saturated water flow and solute transport. Matrix pores are defined as all pores that are small enough to remain water-filled at a specified pore water matric head, cmat [L], whereas macropores are pores that are too large to remain water-filled at cmat . The value of cmat is not yet agreed upon (i.e., various values have been proposed such as 3, 5, 10 cm); however, growing experimental evidence suggests that cmat ¼ 10 cm is appropriate (Jarvis et al. 2002), which corresponds to an equivalent pore diameter of 0.3 mm according to classical capillary rise theory (Or and Wraith 2002). Using this criterion, all pores with equivalent diameters 0:3 mm (c cmat ¼ 10 cm) are matrix pores, whereas those with equivalent diameters >0:3 mm (c > cmat ¼ 10 cm) are macropores. The various ‘‘total porous medium’’ flow parameters described above (i.e., Equation 69.11 through Equation 69.21 that apply to all pore sizes) can be recast as matrix flow parameters by simply restricting c0 to the range, ci < c0 < cmat . Macropore flow parameters can be similarly defined by restricting c0 to the range, cmat < c0 < 0; however, the hydraulic conductivity relationships must be rewritten as Kp (c) ¼ K(c) K(cmat );

cmat c 0

Kp (u) ¼ K(u) K[u(cmat )]; u(cmat ) u us

(69:22) (69:23)

where the subscript ‘‘p’’ denotes the macropore flow domain, and K(c) and K(u) refer to the total porous medium (i.e., both matrix pores and macropores). As a result of these definitions, the flow parameters in the matrix domain are at their maximum values when c0 ¼ cmat ; whereas the flow parameters in the macropore domain are either zero (Kp (c) ¼ Kp (u) ¼ f(c0 ) ¼ S(c0 ) ¼ 0) or undefined (PD(c0 ) and NP(c0 )) when c0 ¼ cmat . Figure 69.9 and Figure 69.10 illustrate selected flow parameter relationships for the matrix, macropore, and total porous medium flow domains in our representative structured loam soil. Note in these figures that the matrix and total porous medium flow parameters are coincident when c0 cmat because the macropores are empty, and thus only the matrix pores are waterconducting. Note also that the macropore relationships produce complex patterns and may have values that are greater than, equal to, or less than the corresponding matrix and total porous medium values, depending on the value of c0 . The primary physical and chemical factors affecting the above unsaturated flow parameters include porous medium texture and structure, pore water viscosity, the concentration and speciation of dissolved salts in the pore water, and porous medium hydrophobicity. All of the unsaturated flow parameters are highly sensitive to porous medium texture and structure (compare Figure 69.5 through Figure 69.7), and hence measuring techniques must preserve the porous medium in its natural=in-situ=antecedent condition to as great an extent as possible. The effects of pore water viscosity and dissolved salts on the unsaturated flow parameters are similar to those described for saturated and field-saturated hydraulic conductivity (see Section 69.5). A hydrophobic soil is nonwetting (i.e., it partially or completely repels water rather than attracts water), and this in turn impedes infiltration because of reduced (or even negative) capillarity. Soil hydrophobicity can be caused by accumulation of certain naturally water-repelling organic constituents (such as pine tree



933

Hydraulic conductivity, K ( ) (cm s−1)

1e+ 0 Kp ( ) Km ( ) Kt ( )

1e− 1

1e− 2

1e− 3

1e− 4

1e− 5

−40

−30

− 20

− 10

0

10

Pore water pressure head, (cm)

(a)

Sorptivity, S( ) (cm s−1/2)

1

0.1

0.01

0.001

0.0001 − 40 (b)

St ( ) Sm ( ) Sp ( )

−30

−20

− 10

0

10

Pore water pressure head, (cm)

FIGURE 69.9. For the structured loamy soil: (a) hydraulic conductivity, K (c), versus pore water matric (or pressure) head, c, in the total soil (Kt (c)), matrix flow domain, (Km (c)), and macropore flow domain (Kp (c)) and (b) sorptivity, S(c), versus pore water pressure head, c, for the total soil (St (c)), matrix flow domain (Sm (c)), and macropore flow domain (Sp (c)).

needles), or by extreme or prolonged drying (such as after a long drought or after a forest fire), which causes certain organic materials and mineral oxides lining the soil pores to become partly or completely water-repellent. Hydrophobicity reduces the capillarity parameters (i.e., f(c0 ), a(c0 ), a*(c0 ), S(c0 ), PD(c0 ), NP(c0 )) relative to a hydrophilic (water-wetting) situation, all other factors remaining equal. Although soil hydrophobicity can be initially strong enough to prevent infiltration of even shallow-ponded water, it usually breaks down over time, allowing normal soil capillarity to eventually return. Further information on soil hydrophobicity and its impacts on soil hydraulic processes and properties



934

FWM pore diameter, PD (mm)

1.2

1.0

PDp( ) PDm( ) PDt( )

0.8

0.6

0.4

0.2

0.0 − 40

− 30

(a)

− 20

− 10


0

10

(cm)

Number of FWM pores/m2 (NP)

40000 NPp( ) NPm( ) NPt( )

30000

20000

10000

0 − 40 (b)

− 30

− 20

− 10


0

10

(cm)

FIGURE 69.10. For the structured loamy soil: (a) FWM pore diameter (PD), versus pore water matric (or pressure) head, c, in the total soil (PDt (c)), matrix flow domain (PDm (c)), and macropore flow domain (PDp (c)) and (b) number of FWM pores per unit area, NP, versus pore water pressure head, c, in the total soil (NPt (c)), matrix flow domain (NPm (c)), and macropore flow domain (NPp (c)).

can be found in Bauters et al. (1998, 2000), Nieber et al. (2000), and references contained therein. Unsaturated hydraulic property methods are described in Chapter 80 through Chapter 84 and include the laboratory tension infiltrometer (Chapter 80), the evaporation method (Chapter 81), the field tension infiltrometer (Chapter 82), the instantaneous profile method (Chapter 83), and selected estimation methods (Chapter 84).



935

REFERENCES Bauters, T.W.J., DiCarlo, D.A., Steenhuis, T.S., and Parlange, J.-Y. 1998. Preferential flow in water repellent sands. Soil Sci. Soc. Am. J. 62: 1185–1190. Bauters, T.W.J., Steenhuis, T.S., DiCarlo, D.A., Nieber, J.L., Dekker, L.W., Ritsema, C.J., Parlange, J.-Y., and Haverkamp, R. 2000. Physics of water repellent soils. J. Hydrol. 231–232: 233–243. Bilderback, T.E., Warren, S.L., Owen, J.S., and Albano, J.P. 2005. Healthy substrates need physicals too! HortTechnology 15: 747–751. Bouma, J. 1983. Use of soil survey data to select measurement techniques for hydraulic conductivity. Agric. Water Manage. 6: 177–190. Bouma, J. 1985. Soil variability and soil survey. In J. Bouma and D.R. Nielsen, Eds. Proceedings of Soil Spatial Variability Workshop. PUDOC, Wageningen, The Netherlands, pp. 130–149. Bouwer, H. 1966. Rapid field measurement of air-entry value and hydraulic conductivity of soil as significant parameters in flow system analysis. Water Resour. Res. 2: 729–738. Bouwer, H. 1978. Groundwater Hydrology. McGraw-Hill, Toronto, ON, Canada. Bruce, R.R. and Luxmoore, R.J. 1986. Water retention: Field methods. In A. Klute, Ed., Methods of Soil Analysis, Part I—Physical and Mineralogical Methods. 2nd ed. American Society of Agronomy, Madison, WI, pp. 663–683.

management of structure for roots in duplex soils. In E.G Gregorich and M.R. Carter, Eds., Soil Quality for Crop Production and Ecosystem Health. Developments in Soil Science, Vol. 25. Elsevier, New York, pp. 339–350. Constantz, J., Herkelrath, W.N., and Murphy, F. 1988. Air encapsulation during infiltration. Soil Sci. Soc. Am. J. 52: 10–16. Gardner, W.R. 1958. Some steady-state solutions of the unsaturated moisture flow equation with application to evaporation from a water table. Soil Sci. 85: 228–232. Grable, A.R. and Siemer, E.G. 1968. Effects of bulk density, aggregate size, and soil water suction on oxygen diffusion, redox potentials, and elongation of corn roots. Soil Sci. Soc. Am. Proc. 32: 180–186. Gupta, S.C. and Hanks, R.J. 1972. Influence of water content of electrical conductivity of the soil. Soil Sci. Soc. Am. Proc. 36: 855–857. Hillel, D. 1980. Applications of Soil Physics. Academic Press, Toronto, ON, Canada. Hopmans, J.W., Simunek, J., Romano, N., and Durner, W. 2002. Simultaneous determination of water transmission and retention properties: inverse methods. In J.H. Dane and G.C. Topp, Eds., Methods of Soil Analysis, Part 4— Physical Methods, Soil Science Society of America, Madison, WI, pp. 963–1004.

Campbell, G.S. 1987. Soil water potential measurement. In R.J. Hanks and R.W. Brown, Eds., Proceedings of International Conference on Measurement of Soil and Plant Water Status, Vol. 1. Logan, UT, July 1987, pp. 115–119.

Jarvis, N.J., Zavattaro, L., Rajkai, K., Reynolds, W.D., Olsen, P.-A., McGechan, M., Mecke, M., Mohanty, B., Leeds-Harrison, P.B., and Jacques, D. 2002. Indirect estimation of near-saturated hydraulic conductivity from readily available soil information. Geoderma 108: 1–17.

Cassel, D.K. and Nielsen, D.R. 1986. Field capacity and available water capacity. In A. Klute, Ed., Methods of Soil Analysis, Part 1—Physical and Mineralogical Methods. 2nd ed. American Society of Agronomy, Madison, WI, pp. 901–926.

Koorevaar, P., Menelik, G., and Dirksen, C. 1983. Elements of Soil Physics. Elsevier, New York, 228 pp.

Cockroft, B. and Olsson, K.A. 1997. Case study of soil quality in south-eastern Austrialia:

McIntyre, D.S. 1974. Soil sampling techniques for physical measurements. In J. Loveday, Ed., Methods for Analysis of Irrigated Soils. Tech.


936 Commun. No. 54, Commonwealth Agricultural Bureau, Australia, pp. 12–20. Mein, R.G. and Farrell, D.A. 1974. Determination of wetting front suction in the Green–Ampt equation. Proc. Soil Sci. Soc. Am. 38: 872–876. Nieber, J.L., Bauters, T.W.J., Steenhuis, T.S., and Parlange, J.-Y. 2000. Numerical simulation of experimental gravity-driven unstable flow in water repellent sand. J. Hydrol. 231–232: 295–307. Olness, A., Clapp, C.E., Liu, R., and Palazzo, A.J. 1998. Biosolids and their effects on soil properties. In A. Wallace and R.E. Terry, Eds., Handbook of Soil Conditioners. Marcel Dekker, New York, pp. 141–165. Or, D. and Wraith, J.M. 2002. Soil water content and water potential relationships. In A.W. Warrick, Ed., Soil Physics Companion. CRC Press, Boca Raton, FL, pp. 49–84. Passioura, J.B. 1980. The transport of water from soil to shoot in wheat seedlings. J. Exp. Bot. 3: 1161–1169.

Soil Sampling and Methods of Analysis Angers, and G.C. Topp, Eds., Proceedings of 3rd Eastern Canada Soil Structure Workshop. Universite´ Laval, Sainte-Foy, Quebec, Canada, pp. 235–248. Reynolds, W.D. and Elrick, D.E. 2005. Measurement and characterization of soil hydraulic properties. In J. Alvarez-Benedi and R. MunozCarpena, Eds., Soil–Water–Solute Process Characterization: An Integrated Approach. CRC Press, Boca Raton, FL, pp. 197–252. Reynolds, W.D., Elrick, D.E., and Clothier, B.E. 1985. The constant head well permeameter: effect of unsaturated flow. Soil Sci. 139: 172–180. Reynolds, W.D., Gregorich, E.G., and Curnoe, W.E. 1995. Characterization of water transmission properties in tilled and untilled soils using tension infiltrometers. Soil Till. Res. 33: 117–131. Richards, L.A. 1931. Capillary conduction of liquids in porous mediums. Physics 1: 318–333.

Philip, J.R. 1957. The theory of infiltration. 4: Sorptivity and algebraic infiltration equations. Soil Sci. 84: 257–264.

Robinson, D.A., Jones, S.B., Blonquist, J.M. Jr., and Friedman, S.P. 2005. A physically derived water content=permittivity calibration model for coarse-textured, layered soils. Soil Sci. Soc. Am. J. 69: 1372–1378.

Philip, J.R. 1987. The quasilinear analysis, the scattering analog, and other aspects of infiltration and seepage. In Y.S. Fok, Ed., Infiltration, Development and Application. Water Resources Research Centre, Honolulu, HI, pp. 1–27.

Romano, N. and Santini, A. 2002. 3.3 Water Retention and Storage, 3.3.3 Field. In J.H. Dane and G.C. Topp, Eds., Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 721–738.

Raats, P.A.C. 1976. Analytical solutions of a simplified flow equation. Trans. ASAE 19: 683–689.

Scotter, D.R., Clothier, B.E., and Harper, E.R. 1982. Measuring saturated hydraulic conductivity and sorptivity using twin rings. Aust. J. Soil Res. 20: 295–304.

Reynolds, W.D., Bowman, B.T., Drury, C.F., Tan, C.S., and Lu, X. 2002. Indicators of good soil physical quality: density and storage parameters. Geoderma 110: 131–146. Reynolds, W.D., Brown, D.A., Mathur, S.P., and Overend, R.P. 1992. Effect of in-situ gas accumulation on the hydraulic conductivity of peat. Soil Sci. 153: 397–408. Reynolds, W.D., Bowman, B.T., and Tomlin, A.D. 1997. Comparison of selected water and air properties in soil under forest, no-tillage, and conventional tillage. In J. Caron, D.A.

Skopp, J., Jawson, M.D., and Doran, J.W. 1990. Steady-state aerobic microbial activity as a function of soil water content. Soil Sci. Soc. Am. J. 54: 1619–1625. Smith, R.E. 2002. Infiltration Theory for Hydrologic Applications. Water Resources Monograph 15, American Geophysical Union, Washington, DC, 212 pp. Soil Science Society of America. 1997. Glossary of Soil Science Terms. Soil Science Society of America, Madison, WI.


Soil Water Analyses: Principles and Parameters Stephens, D.B., Lambert, K., and Watson, D. 1987. Regression models for hydraulic conductivity and field test of the borehole permeameter. Water Resour. Res. 23: 2207–2214. Topp, G.C., Davis, J.L., and Annan, A.P. 1980. Electromagnetic determination of soil–water content: Measurement in coaxial transmission lines. Water Resour. Res. 16: 574–582. Topp, G.C. and Ferre´, Ty. P.A. 2002. 3.1 Water content. In J.H. Dane and G.C. Topp, Eds., Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 417–545. Topp, G.C. and Reynolds, W.D. 1998. Time domain reflectometry: a seminal technique for measuring mass and energy in soil. Soil Till. Res. 47: 125–132. Townend, J., Reeve, M.J., and Carter, A. 2001. Water release characteristic. In K.A. Smith and

937

C.E. Mullins, Eds., Soil and Environmental Analysis: Physical Methods, 2nd ed. Marcel Dekker, New York, pp. 95–140. Verdonck, O., Penninck, R., and De Boodt, M. 1983. Physical properties of different horticultural substrates. Acta Hortic. 150: 155–160. White, I. and Sully, M.J. 1987. Macroscopic and microscopic capillary length and time scales from field infiltration. Water Resour. Res. 23: 1514–1522. White, I., Sully, M.J., and Perroux, K.M. 1992. Measurement of surface-soil hydraulic properties: disk permeameters, tension infiltrometers, and other techniques. In G.C. Topp, W.D. Reynolds, and R.E. Green, Eds., Advances in Measurement of Soil Physical Properties: Bringing Theory into Practice. SSSA Special Publication No. 30, Soil Science Society of America, Madison, WI, pp. 69–103.



Chapter 70 Soil Water Content G. Clarke Topp Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

G.W. Parkin University of Guelph Guelph, Ontario, Canada

Ty P.A. Ferre´ University of Arizona Tucson, Arizona, United States

70.1 INTRODUCTION Water in soil is a vital link in the hydrological cycle that controls exchange with the atmosphere above and with the groundwater below. Water in soil acts both as a lubricant and as a binding agent among the soil particulate materials, thereby influencing the structural stability and strength of soil and geologic materials. The high heat capacity of water causes a moderation of diurnal and seasonal temperature cycles at the soil surface. Chemically, water serves as the transport agent for the dissolved inorganic chemicals and suspended biological components that are involved in the processes of soil development and degradation. Biological production from soil, either as forest products or agricultural crops, is influenced primarily by water availability. The measurement of soil water content then is important directly for quantifying water balance, for estimates of plant water status, and for characterizing most soil physical, chemical, and biological processes. The measurement of soil water content has undergone revolutionary advancements in the last 20 years. From having gravimetric sampling and neutron moderation as the primary field methods in the early 1980s, we now have numerous options, such as time-domain reflectometry (TDR), capacitance (and impedance) devices, ground penetrating radar (GPR), airborne=satellite active radar, and passive microwave methods (Gardner et al. 2001; Topp and Ferre´ 2002). These five newer methods are all based on electromagnetic (EM) measurements. Information on EM properties of soil and their use in soil water content measurements can be found in Topp et al. (1980), Ferre´ and Topp (2002), and Topp and Reynolds (1998). All of the EM methods make use of the high relative permittivity (dielectric constant) of the

939


940


water (80) in soil compared with the permittivities of the other soil components, which range from one for air to 3–5 for typical soil solids. Due to this contrast, methods that measure the bulk dielectric permittivity of soil are effective for the measurement of volumetric water content. A selection of EM methods is the focus of this chapter, as these offer a variety of sample geometries and spatial coverage, are minimally site disruptive, collect data digitally allowing real- or near real-time information, and measure on volumetric basis directly. In Section 69.2 the basic water content parameters and expressions for water content are defined, such as volumetric, gravimetric, and degree of saturation. In addition, the principles behind the use of EM methods, including how dielectric permittivity relates to water content appear in Section 69.2.

70.2 GRAVIMETRIC WITH OVEN DRYING The thermogravimetric method is conceptually simple. Initially, a moist soil sample is weighed. The sample is then oven dried at 105 C and reweighed. The gravimetric water content is defined as the ratio of the mass lost, attributed to water initially present in the sample, to total mass of the fully dried soil. The method is apparently straightforward and is commonly thought to yield absolute results. In fact, this is not so for several reasons. Water is retained by the components of the soil at a wide range of energy levels and there is no absolute time at which the soil reaches a ‘‘dry’’ state when maintained at 105 C. Soil samples continue to decrease in mass slowly at 105 C for many days (Gardner 1986). In addition, many soil samples contain organic materials, some of which are volatile at 105 C, so some of the decrease in mass may be due to volatilization of components other than water. Finally, there is the problem of temperature control. Although the drying ovens in common use in most soil laboratories can maintain temperatures in the range of 100 C–110 C with careful adjustment, temperatures within the oven vary depending on the location in the oven chamber. Given that the actual temperature of the soil sample is not measured, this variability can lead to differential heating among soils placed in the same oven for the same amount of time. In spite of these imperfections, however, the oven-drying method is a commonly used and convenient method to obtain a good estimate of soil water content. The use of microwave ovens is not as rigorously standardized, as in the case of incandescent heating ovens. A more complete discussion of the procedures and limitations of the gravimetric method are given in many standard texts (e.g., Topp and Ferre´ 2002).

70.3 TIME-DOMAIN REFLECTOMETRY It was just 30 years ago that TDR was first applied to measurement in soil and earth materials (Davis and Chudobiak 1975). Since those first measurements, TDR has been used to measure water content at many scales and under a broad range of conditions (Topp and Reynolds 1998; Robinson et al. 2003a), and has become a standard method of water content measurement. The popularity of the method for soil=environmental monitoring and research arises from a combination of its accuracy in a wide range of soils and its relative ease of use compared with many other available techniques. TDR provides real-time, in-situ soil water content measurements. Measurement systems can be multiplexed and data-logged, allowing for remote automated monitoring. For most soils, the accuracy of measurements of volumetric water content change is within 0:02 m3 m3 without the need for soil-specific calibration, and better absolute water contents can be achieved with calibration. There is considerable flexibility in the design and placement of TDR probes, allowing users to modify water content measurement networks to conform to the requirements of any specific study. Finally, because TDR determines the volumetric water content, the data are directly applicable


Soil Water Content

941

to hydrologic water balance analyses with no need for the measurement of supporting soil parameters such as bulk density.

70.3.1 MATERIAL

AND INSTRUMENTS

TDR instrumentation consists of four basic components: a timing circuit, a pulse generator, a sampling receiver, and a display or recording device. It is the pulse generator, which launches a pulse or wave whose travel is analyzed. Most commercial instruments have all components in a single unit, which also performs analyses of the TDR traces, displaying and recording the interpreted water contents. Any of these instruments can be used to measure water content as long as they provide stable low noise readings with high time base accuracy, typically with a pulse transition time of 0:2 ns (Hook and Livingston 1995). For custom analyses involving highly precise applications, the capability to record the entire waveform is necessary. An additional useful feature is the ability to provide automated water content analysis. The capability of displaying the actual TDR trace, and manually interpreting it, is very useful for assuring that the instrument is operating properly and that the automated interpretation is reasonable. Connection of the TDR instrument to a multiplexer for sequential measurement at a number of locations increases greatly the efficiency of data collection possibilities. The initial and still widely used TDR instrument is the portable cable tester (Model 1502 B or C, Tektronix). In this instrument, the trace is displayed and analyses may be performed, and recorded manually, or data may be recorded and analyzed digitally on a PC (Or et al. 2003). The cable tester and a PC were incorporated into a number of custom systems designed to achieve automated TDR trace analysis, and multiplexing (Ferre´ and Topp 2002). Most commercial instruments now offer automated analysis as a part of the basic instrument with the multiplexing capability as an option. Some of the features of commercially available instruments are listed in Table 70.1. The basic elements of a TDR probe are conductive components, often parallel metallic rods, which act as wave-guides, and the soil material in which the wave or signal propagates (Figure 70.1). Currently, the most common soil probes are of the balanced pair transmission line, consisting of two parallel rods, with rods that vary in length, depending on the measurement requirement, from 0.1 to 1.0 m and with probe separations from 0.01 to 0.1 m (Topp and Davis 1985). The minimum practical probe length for standard equipment is 0.1 m. The upper limit on length of probe is largely determined by electrical conductivity, clay content, and maximum water content expected. Although no firm guide can be offered, Dalton (1992) showed that probe lengths will have to be reduced to 0.2 m in clayey soil of EC > 0:1 S m1 . Coated probes, discussed later, overcome this limitation to some extent. Zegelin et al. (1989) introduced multipronged probes where one prong or wire is centrally located and variable numbers of prongs are located circumferentially around the central wire. These configurations, even with only two outer prongs, act electrically to emulate a coaxial transmission line and result in a marginally improved TDR reflection. The extra rods, however, make for greater installation difficulty and associated soil disturbance than from a parallel pair. The configuration of the wave-guide or probe determines the extent and shape of the measured soil sample. Earlier experimental and theoretical analyses have demonstrated that the distribution along the length of probes has an effect, which is represented by a linear-weighted average (Hook and Livingston 1995). Specific refinements may be required for layered soils (Robinson et al. 2003b).

MP917 1502B=C TDR100 TRASE Uses Tek 1502 B=C TRIME TDR

Tektronix Campbell Scientific, Inc. Soilmoisture Equipment Corp. Dynamax, Inc. MESA Systems Co.

Model

Environmental Sensors, Inc.

Supplier

Data logging

No SDMX50 TRASE 6003 TR-200 TRIME-MUX6

Yes

Multiplexing

MP917 compatible, with shorting diodes Custom Custom TRASE compatible Custom TRIME compatible

Soil probes

Yes Yes Yes Yes No

No

Waveform

Yes Yes Yes No No

No

EC

942

No CR10X or CR23X Internal PC-based PC-based

Internal option

TABLE 70.1 TDR Instruments, Listing Options and Capabilities




Soil Water Content

943

FIGURE 70.1. A collection of TDR from a limited number of suppliers. The numbers indicate the sources of those shown as: 1 and 2 are from Soilmoisture Equipment Corp.; 3, 4, and 5 are from Environmental Sensors Inc. (ESI); 6 are custom design developed in our laboratory.

For the lateral distribution, the situation is more complex. Knight (1992) and Knight et al. (1994) examined theoretically the spatial weighting function for parallel pair and multiwire acoaxial probes inserted in a medium of nearly uniform permittivity. The analytical expressions and approaches from Knight (1992) form the basis for probe design specifications and for evaluation of probe performance. For example, Knight (1992) showed that the ratio of the wire or prong spacing to the wire diameter in a soil probe is an important geometric descriptor of all TDR probes and should be considered for design and installation purposes. Knight (1992) proposed that the ratio of wire spacing to wire diameter should not exceed 10. It is reasonable that the wire diameter should be at least 10 times the representative pore size or particle diameter to provide sensible averages. One important finding of this analytical investigation is that the sample area of TDR is independent of the water content of the medium. We have found that 6 mm diameter rods spaced at 50 mm have worked well in a variety of studies in tilled and untilled agricultural soil. Many other probe configurations have come into current use and these can be used successfully with due consideration of the limitation applying to each probe type.

70.3.2 PROCEDURE The TDR method is straightforward but varies for different types of applications such as laboratory or field; surficial or at depth; point specific or spatially referenced; and so on. Insert Soil Probe or Transmission Line into the Soil Sample The installation of TDR probes is also important for high-quality measurements. Air gaps around the probes can cause erroneously low water content measurements. However, Knight et al. (1997) and Ferre´ et al. (1998) applied a numerical analysis to show


944


that partial air gaps, surrounding only a fraction of the probe perimeter, will not adversely affect the measured relative permittivity. Rods of the probe should be installed in parallel; however, minor deviations from parallel alignment will not lead to significant errors unless the rods come into contact with each other. Installation by insertion of nonparallel or poorly aligned rods or probes may lead to air gaps along the rods; this should be avoided at all times. Care should be taken to minimize disturbance of the soil when inserting the rods, especially in compressible media. In laboratory and near-surface field measurements, it is important that the cross-section and length of the probe be chosen so that the EM field associated with the TDR signal is contained within the soil sample (Knight 1992). In addition, the maximum probe length is limited by excessive conductive loss in the soil. Connect Probe to TDR Instrument Using Coaxial Cable and Initiate Signal Transmission and Recovery of the TDR Waveform The length of cable connecting probe, and instrument is best limited to 25 m to achieve acceptable signal-to-noise ratio and prevent excessive signal attenuation. Some instruments have introduced compensation for signal loss due to cable length allowing the use of greater cable lengths. Choice of acceptable cable length should be based on signal quality from the measurements in the wettest, most conductive conditions. The use of multiplexers introduces additional signal deterioration and may restrict additionally the separation between probe and instrument. Analyze the Waveform to Determine the Time of Travel of the Signal in the Soil, Which Serves to Determine the Relative Permittivity Of interest for water content determination is the two-way travel time of the TDR signal in the soil in and surrounding the probe. Two times are measured; the time of arrival of signal reflected from the probe-to-soil interface (t1 in Figure 70.2) and the time of arrival of the signal reflected from the end of the probe (t2 in Figure 70.2). The TDR waveform in Figure 70.2 shows the recommended way of estimating the two times. The intersection of tangential lines on either side of the identifying signal reflection is the most precise indication of the desired times. The time difference (t2 t1 ) is a measure of the two-way travel time for the pulse or wave along the length of the rods. For some probes, the choice of where to pick t1 may be difficult under some conditions. Robinson et al. (2003b) present a method for probe calibration using only water and air, claiming this to be highly accurate. Periodic Measurements in a Reference Liquid to Detect Instrument Drift and Malfunction Reference liquids of known dielectric permittivity are useful to check measurement repeatability and instrument drift. We have used repeated measurements with the TDR probe immersed in isopropyl alcohol or water and recorded at hourly intervals during field measurement. It is important to ensure that the container is sufficiently large to contain the signal entirely within the reference fluid.

70.3.3 CALCULATIONS For many instruments calculations of volumetric water content are made within the instrument. The simple calculation sequence given here applies to those instruments where travel time measurement is made explicit.


945

1.2

1.2

1.1

1.1

1.0

Voltage (V0)

Voltage (V0)

Soil Water Content

0.9 0.8 0.7 t1

0.6 Schematic TDR probe

0.5 25

30

40

1.4

0.8 0.6 0.4

Probe specific dip

0.2

0

20

40

60 80 Time (ns)

Vf = 1.03 V0 100

120

Voltage (V0)

Voltage (V0)

t2

0.8

0.6 40

45

45

50 Time (ns)

55

60

1.2

1.0

(a)

t1

1.4

1.2

0.0

0.9

0.7

t2 35 Time (ns)

1.0

1.0 0.8 0.6 0.4

Probe specific dip

0.2 0.0

0

20

40

60 80 Time (ns)

Vf = 0.78 V0

100

120

(b)

FIGURE 70.2. Two TDR curves from 20 cm probes in silty clay loam soil. The soils are at similar water content but the soil solution is more conductive in (b), giving a smaller return reflection and resulting lower V f . In (a) uv ¼ 0:304 m3 m3 and s0 ¼ 57 mS m1 and in (b) uv ¼ 0:271 m3 m3 and s0 ¼ 95 mS m1 . (From Topp, G.C. and Ferre´, Ty P.A., in D. Hillel et al. (Eds.), Encyclopedia of Soils in the Environment, Vol. 4, Elsevier, Oxford, UK, 2004, 174–181. With permission.)

Convert Travel Time to Relative Permittivity The travel time (t2 t1 ) from second section, p. 944, is converted to propagation velocity and then to apparent relative permittivity as follows: pffiffiffiffiffiffi «ra 2L ¼ (70:1) (t2 t1 ) ¼ v c where L is the length of the probe, v is the velocity of propagation, «ra is the apparent relative permittivity, and c is the velocity of light or other EM waves in vacuum (3 108 m s1 ). pffiffiffiffiffiffi Convert «ra to Volumetric Water Content Using a Selected Calibration Relationship Although calibration relationships should be validated for each soil, experience has shown that the empirical relationship given by Topp et al. (1980) is widely applicable. The simpler-to-use linear relationship is recommended for soil where a calibration has not been developed:



946

pffiffiffiffiffiffi uv ¼ 0:115 «ra 0:176

(70:2)

Soils where it may become advisable to develop a specific calibration include those high in clay and=or salt and having organic matter above 0.05 kg kg1 . The effect of high clay cannot be made specific as its effect depends on grain size and mineralogy of the clay. The high clay, salt, and organic matter alter the slope (0.115) of Equation 70.2 and may introduce curvature as well (Topp et al. 2000). Dense soils may influence the intercept (0:176) in Equation 70.2, as the magnitude of that term is dependent on the soil solids composition.

70.3.4 COMMENTS Measuring Water Content Profiles in the Field For profiles near to (approximately 1.5 m depth) and extending to the surface, three general approaches have been used (Ferre´ and Topp 2002). Each offers certain advantages along with some limitations. With a series of differing length probes, vertically installed from the surface, it is possible to segregate the water into layers in the profile. Water in each layer is assumed evenly distributed over the appropriate length interval. Spatial variability laterally contributes to the uncertainty or error associated with this type of profile determination, which can be as large as 0:03 m3 m3 (Topp 1987). The longest probes are useful for water balance calculation, where a single measurement gives the total water quantity over the depth spanned and is not dependent on the depth distribution of the water. Vertically installed rods tend to generate cracks in the soil between them at the surface and=or gaps around the individual rods. These soil openings each affect the infiltration of rainfall or irrigation and also affect the TDR reading. Vertical rods will tend to be moved vertically out of the soil during winter by the processes of frost-heave. A second method involves installation of a number of horizontally oriented probes, one at each measured depth. These provide a more precise profile of water content that is not influenced strongly by lateral spatial variations, but these cannot compensate for major discontinuities in the vertical water content distribution. The total profile storage for water balance estimates involves sums of values measured at each depth, being less precise than from a single vertical probe. Horizontally installed probes generally require opening a pit or hole into the soil, creating the possibility of disturbance to the region to be measured. Additionally the cable and probe connection must be hermetically sealed. An optimized profiling option uses parallel rods installed from the soil surface but 45 off the vertical. These can be placed so that the resulting water content profile is a single vertical profile, and affected less by lateral variability and each depth increment provides for equal magnitude lateral and vertical integration. Schwartz and Evett (2003) give an evaluation of 30 installations for wetting front evaluation in a soil column. The two disadvantages of angled installations are the greater difficulty of making installations at an angle with the required precision to know the actual depth at the end of the installed rods. The increased probe length to achieve an angled installation decreases the total vertical depth that can be measured in clayey soils.


Soil Water Content

947

Hook et al. (1992) describe the use of diode shorting to segment probes, which are constructed as profiling probes for use with model MP917 TDR instrument from Environmental Sensors, Inc. These probes allow for determination of water content profiles having the same accuracy for each segment of 0:01 m3 m3 . The EM field for TDR measurement propagates both in the resin comprising the probe, and in the surrounding soil, with more of the field in soil in wet than in dry soil, meaning that the sampling volume changes with water content. Installation of this type of probe is of critical importance as any disturbance of the soil adjacent to the probe is in the most sensitive of the measured region. These edge effect factors have not been adequately evaluated to allow specific quantification. Recently, a number of attempts have been made to determine the water content profile using waveform analysis (Todoroff and Luk 2001; Heimovaara et al. 2004). With additional research and development, this approach may become the method of choice to overcome limitations cited above.

Coated-Rod Probes Applying electrically resistive coatings to the rods can minimize the signal attenuation and loss of signal in conductive soils. Ferre´ et al. (1996) extended the work of Annan (1977) to show that coated rods do not measure the arithmetic average of the dielectric permittivities of the coatings and the surrounding medium. Because common coating materials have low dielectric permittivities, coated rods are more sensitive to lower water contents than to higher water contents. One result of this variable sensitivity is that, unlike uncoated rods, coated rod probes do not measure the correct length-weighted average water content along their length if the water content varies along their length. Therefore, probes that measure the water content through coatings should be installed in a manner that minimizes water content differences throughout their sample volume. In addition, the reduced sensitivity of coated rods to conductive losses reduces the usefulness of these probes for electrical conductivity measurement.

70.4 GROUND PENETRATING RADAR GPR has been widely applied in the geosciences (Neal 2004) and methods have recently been developed for measuring soil volumetric water content (Davis and Annan 2002; Huisman et al. 2003). GPR methods offer the advantage of providing data from larger spatial regions than for TDR. Another significant advantage of surface and airborne GPR methods over TDR is that both methods are nonintrusive. The borehole GPR method is intrusive requiring installation of GPR transmitter and receiver in horizontal or vertical boreholes (Parkin et al. 2000; Rucker and Ferre´ 2003). Surface and airborne GPR methods are most appropriate for root zone investigations, whereas borehole methods are more appropriate for deeper vadose zone applications. This discussion is limited to above ground methods, which offer greater spatial coverage than downhole methods. However, many of the concepts presented are equally applicable to borehole GPR. The physics of the GPR method is identical to TDR (Weiler et al. 1998). Both methods rely on measuring the travel time or amplitude of EM wave fields. Energy emitted from the GPR transmitter travels through air and soil to the receiver. Depending on the method used, the



948

(b)

0

Amplitude Ar

Tx

Rx

Time

(a)

Amplitude

Tx

Rx

0 t0

t gw

Time

FIGURE 70.3. A schematic diagram showing two different GPR antennae configurations for measuring soil water content. (a) Surface GPR, direct ground wave, (b) Airlaunched surface reflectivity. Tx and Rx are GPR transmitter and receiver antennae, respectively.

travel time or amplitude of energy from reflected or direct pathways is measured and converted to the soil relative permittivity. Equation 70.2 (or a soil-specific calibration) is then used to convert the measured relative permittivity to volumetric water content. More details on the principles of using GPR to measure soil water content are found in Davis and Annan (2002) and Huisman et al. (2003). Figure 70.3 shows two of the GPR antenna configurations that have been used to measure soil water content of which the methods are described herein. The air-launched surface reflectivity method has the advantage over the surface method in that the antenna can be suspended above the land surface, with the energy directed downwards. The chosen antenna frequency and height above ground will depend on the desired size of the energy footprint (area sampled) on the ground. For example, Huisman et al. (2003) show that when using 1 GHz and 225 MHz antenna systems elevated 1 m above the ground, approximate footprint areas are 0.79 m by 0.79 m and 1.76 m by 1.76 m, respectively. The following equation (Davis and Annan 2002; Redman et al. 2002) calculates the soil relative permittivity from the amplitude of the reflected energy, Ar , relative to a maximum amplitude, Am , from a perfect reflector such as a metal sheet placed on the ground, which has a reflection coefficient of 1: «r ¼

1 þ AAmr 1 AAmr

2 2

(70:3)


Soil Water Content

949

Active microwave remote sensing (synthetic aperture radar or SAR) operates on the same principle as air-launched surface reflectivity GPR. As SAR operates at a higher frequency (above 1 GHz) there are additional complications caused by soil surface roughness and growing plants. One of the primary motivations behind SAR satellites is to be able to estimate surface soil conditions, including water content. Mapping soil water content with SAR has been extensively researched and several projects have demonstrated the feasibility of deriving water content using SAR. As the methods continue to be under development microwave remote sensing is not given detailed coverage in this manual. McNairn et al. (2002) prepared a state of the art summary of the SAR approach as a method for the measurement of soil water content. The surface GPR method differs from the air-launched in that the antennae are placed in direct contact with the ground surface and the relative permittivity is determined by measuring the GPR wave velocity. The velocity of direct or reflected waves between the GPR transmitter and receiver is converted to relative permittivity and water content as shown above for TDR. For either method, fixed or multiple antennae separation distances (offsets) can be used to collect velocity data. Multiple offset methods include wide-angle reflection and refraction (WARR) and common midpoint (CMP). A WARR survey involves keeping the receiver antenna at a fixed location and moving the transmitter antenna away from the receiver a set increment, measuring the ground wave velocity at each offset. A CMP begins with the transmitter and receiver placed very close together, and then incrementally moving them in opposite directions, again measuring the ground wave velocity at each offset. The ground wave velocity is more straightforward to measure when using the two multiple offset methods as opposed to the fixed offset (FO) method. The FO method requires picking the arrival time of the direct ground wave, whereas the WARR and CMP methods can use ground wave peak amplitude arrival times at different offsets to measure ground wave velocity.

70.4.1 MATERIAL

AND INSTRUMENTS

There are no commercially available GPR instruments designed specifically for measuring soil water content. Davis and Annan (2002) give GPR manufacturers whose equipment can be adapted for measuring soil water content. The procedures described forthwith assume that users are familiar with the basic operational methods, licensing requirements, and potential health and safety issues of their GPR equipment and will therefore not be repeated.

70.4.2 PROCEDURE Air-Launched Surface Reflectivity 1

Position the GPR transmitter and receiver about 1 m above the ground surface using a cart or vehicle (Davis and Annan 2002; Huisman et al. 2003). Use at least a 10 MHz system so that the electrical conductivity of the ground does not substantially influence the electrical current flow (Davis and Annan 2002).

2

Place a metal sheet, larger than the energy footprint on the ground, under the GPR antennae. Measure Am , the maximum amplitude of the energy reflected from the metal sheet.


950 3

Soil Sampling and Methods of Analysis Measure Ar over the soil of interest and then use Equation 70.3 to calculate «r . Finally, Equation 70.2 or a soil-specific calibration can be used to convert «r to soil volumetric water content.

Surface GPR (Direct Ground Wave, Fixed Offset) The following procedure is adapted after the studies by Grote et al. (2003), and Galagedara et al. (2003, 2005), and the review by Huisman et al. (2003): 1

Perform a WARR survey to determine the GPR system airwave velocity calibration (time zero, t0 ), to identify clearly the ground wave on the GPR output of energy versus time, and to select the best antennae offset distance for the FO survey. Time zero is defined as the start of the transmitter pulse, which may vary due to thermal drift and flexing of the fiber optic cables of the GPR system. It is critical to determine, using Equation (70.4), an accurate t0 calibration using the airwave velocity measurement from the WARR survey as all direct ground wave arrival times are measured relative to it: tab ¼ tgw t0

(70:4)

where tab is the absolute ground wave arrival time, and tgw is the measured ground wave arrival time (leading edge of the ground wave). For more information on time zero, and other issues related to this method see Galagedara et al. (2003). As a general recommendation, Galagedara et al. (2005) suggest using seven offsets (from 0.5 to 2.0 m) for the WARR survey. 2

After selecting the best offset (one for which the ground wave is clearly separated from reflected waves), perform the FO survey by keeping the GPR antennae at the selected offset (Galagedara et al. 2005 recommend 1.5–2.0 m) and moving along the survey line (Huisman et al. 2003). Measurements can be taken at a very small time increment (few seconds), depending on the speed at which the antennae are moving, and the desired measurement resolution. Synchronizing the GPR measurements with a GPS system facilitates analysis of spatial variability of soil water content.

3

Measure the travel time of the direct ground wave by picking its leading edge arrival time with a 5% threshold, for example.

4

Convert the measured travel time to velocity using the fixed antenna offset distance. Then determine the apparent relative permittivity from the velocity using Equation 70.1; then use Equation 70.2 or a soil-specific calibration to determine the soil water content.

Surface GPR (Direct Ground Wave, Multiple Offsets) Either the WARR or CMP methods can be used to gather data on the velocity of the direct ground wave. Huisman et al. (2001) found that soil water contents measured with the WARR method were more accurate than those measured with the FO method. The WARR and CMP methods do not rely on an accurate measurement of t0 or picking of the leading edge of the ground wave; they only depend on the slope of the peak arrival time versus antenna offset relationship:


Soil Water Content

951

1

Perform WARR or CMP measurements as discussed above.

2

Select the arrival times of the peak amplitude of the ground wave for each antenna offset.

3

Measure the velocity of the ground wave using the slope of the peak arrival time versus antennae offset distance data.

4

Repeat step 4 given under the first section on p. 950.

Surface GPR (Reflected Ground Wave, Single and Multiple Offsets) The single offset method relies on knowing the depth to a subsurface reflector, whereas the multiple offset method only requires the presence of reflecting horizons at depth in the soil profile, such as the water table, lithologic boundaries, and buried objects (Davis and Annan 2002). For the multioffset method, the average soil water content between the ground surface and the reflector, and the depth to the reflector can be estimated since multiple travel pathways occur. See Greaves et al. (1996) for an example using this method: 1

Perform single or multiple offset GPR surveys.

2

Convert measured travel times of reflected energy to velocity using the known distance to the reflector (single offset method) or measured depth to the reflector (multiple offset method).

3

Determine the relative permittivity from the velocity using Equation 70.1; then use Equation 70.2 or a soil-specific calibration to determine the soil water content.

70.4.3 COMMENTS The GPR-based methods described above seem straightforward; however, data processing is not routine and would be very difficult to automate (Davis and Annan 2002). The major questions=limitations surrounding the GPR-based methods are: 1

Potential users of any of the GPR-based methods are advised to contact their local regulatory agency for any limitations on the use of EM energy emitting devices.

2

The surface methods work best in soils of low signal loss (low electrical conductivity, EC). For instance, the direct ground wave method works best when soil EC is less than 20 mS m1 (Davis and Annan 2002).

3

Depth of penetration of the direct ground wave is a function of many variables including the GPR frequency, antennae offset distance, soil EC and water content, and soil heterogeneities.

4

Little is known about the effects of surface roughness, and soil water heterogeneities on the value of the reflection coefficient, and the depth of penetration in the air-launched method (Huisman et al. 2003).


952


5

Methods that rely on subsurface reflectors may be limited by nature of reflectors and conductive loss.

6

The larger sample volume with GPR over TDR may be an advantage or disadvantage depending on the purpose of the investigation.

Notwithstanding the aforementioned questions=limitations surrounding the GPR methods, they do hold great promise as a means of rapidly determining soil water content variability at the field scale. Although GPR-based methods remain largely at the research and development stage, it may one day be the standard method of measuring soil water content at the field scale, if GPR technology advances occur as for TDR over the past 25 years. Active microwave remote sensing radar (SAR) operates on the same principle as airlaunched GPR but at 1.4 and 6 GHz frequencies. Although relative permittivity (water content) is the main factor influencing radar backscatter, the soil and surface factors also have major influence on radar backscatter, especially, the surface geometry of the soil (random roughness related to tillage; soil aggregation; tillage row direction; and microtopography), and the vegetation characteristics (size and geometry of the leaves; stalks and fruits; and the vegetation water content). Other soil properties, such as texture, and bulk density have a minor effect. The relationships between radar backscatter and soil water content are usually based on regression analyses as surface and plant factors cannot be characterized adequately. Getting reliable soil data as ‘‘ground truth’’ for the regression analyses is complicated by the shallow depth of penetration of the radar wave ( 0:03 S m1 , hence a soil-based calibration procedure follows. Calibration procedures will vary with the electrode configuration of the capacitance sensor. Laboratory and field calibrations are the same and only a laboratory procedure is given. Although the gravimetric method is the usual standard, a more convenient standard for these calibrations is TDR. TDR measures on a volumetric basis as do capacitance and impedance instruments. When using the gravimetric method, it is necessary to make additional measurements of bulk density to convert gravimetric mass basis values to volumetric: 1

Choose a container size and shape to accommodate the sensor’s primary zone of influence. Minimum container volume can be estimated using the general



956

procedure described by Starr and Paltineanu (2002). For ring-type sensors, the container diameter is 25 cm. In addition, there should be 5 cm of uniform soil above the uppermost sensor and below the lowermost sensor to assure adequate sample size. 2

Screen the required mass of soil through a 5 mm sieve.

3

Thoroughly mix the soil after air-drying.

4

Weigh the desired mass of soil for a 2 cm soil depth to be packed to the chosen soil density.

5

Pack the soil carefully to the desired bulk density (i.e., to give the desired depth increment).

6

Repeat steps 4 and 5 until the container is filled.

7

Install the probe carefully in the soil so that there are no air gaps between the probe and soil and so that soil density is not altered by the probe installation.

8

Record capacitance probe reading in the container.

9

Record a matching TDR value in the container, if TDR is the reference method. If gravimetric method is the reference method, a series of additional steps are required as follows: a. Record weight of container plus soil, to be used for bulk density measurement. b. Take a minimum of three subsamples for wet and dry weights for mass basis values. c. If soil packing was done with adequate precision, subsampling for bulk density is not required. If assurance of density uniformity is required, obtain carefully a minimum of three subsamples using bulk density cylinders near the location of the probe.

10

Prepare the soil for the next calibration point. This step includes selecting the wetting increments so that four or five discrete soil water contents will result; and adding water and mixing the soil to distribute the water uniformly. Spread soil uniformly on a plastic sheet or large tray (a thickness not greater than 4 cm is ideal). Mist-spray one measured volume of water on the soil in several stages, mixing the soil after each stage. The mixing may be accomplished at each stage by lifting and turning with a flat lifter or by allowing the soil to cascade over itself, caused by lifting the plastic sheet from one corner toward its diagonal opposite and so on from other corners.

11

Repeat steps 4 to 10 for each calibration point.

12

Record the data for use in calculating a calibration equation.


Soil Water Content

957

70.5.4 MEASUREMENT Good results require that great care be given to probe installation to ensure a tight fit of the electrodes or access pipe to the soil (i.e., no air gaps along the electrodes or access pipe), and with minimal soil disturbance (i.e., change of soil density or structure, in or near the sensing volume). Rod-type probes are pushed carefully into the soil surface, or into the wall of a soil pit for probes that are to be buried, without creating air gaps along the rods or compressing the soil with the electrode housing: 1(a) Connect probe to impedance instrument to measure impedance, following the operations manual. 1(b) Connect probe to capacitance instrument to measure resonant frequency as directed by the operations manual. 2

For monitoring involving data storage, initiate the data logging capability to retrieve the measured data.

3

Transmit or retrieve stored data for calculation and analysis.

70.5.5 CALCULATIONS The first calculation step is to convert measured impedance or capacitance (frequency) values to either apparent relative permittivity or water content depending on adopted calibration procedure. As most instruments have a voltage output as indicative of impedance or capacitance, these calculations will adopt that assumption. Normalization Here it is convenient to represent voltage output from the impedance measurements as VZ and that from the capacitance measurements as VF : 1

pffiffiffiffiffiffi Plot and=or perform a linear regression of 1=VZ or 1=VF against «ra for air and water, along with any other liquids that were used for the normalization measurements.

2

If the above does not automatically give a zero intercept, then the normalization involves subtracting the intercept to get equations having similar form of Equation 70.5 and Equation 70.8. This normalizes the data to air as the lowest reading.

3

Calibration can be achieved by substituting this simple regression relationship pffiffiffiffiffiffi into Equation 70.2 or an equivalent relationship between «ra and water content.

For soils needing additional soil-based calibration another step is required. Calibration Using Soil 1

Plot and=or perform a regression of 1=VZ or 1=VF against uv . For soils of low electrical conductivity, this relationship is expected to be linear. Increasing electrical conductivity will contribute curvature to the relationship.



958 2

The resulting calibration relationship can be used to convert all measurements to water content on a volume basis.

70.5.6 COMMENTS The zone of influence of the rod-type sensors is similar to that of equivalent TDR probes. The cylindrical ring sensors, however, have a limited zone of influence in the soil. The zone of influence has both axial (vertically along the sensor) and radial (perpendicular to the sensor) components. The electric field giving rise to the capacitance measure is most heavily weighted to the region between the rings and drops off very rapidly, both axially and radially, from that point. Research aimed at quantifying the electric field strength and radial weighting of the capacitance is ongoing. Commercial suppliers have claimed radial reach upto 10 cm, and axial reach at 5 cm, which are seldom accompanied by quantitative verification. The electric field diminishes radically in the access pipe, and adjacent soil. Therefore, it is imperative that installation is tight-fitting and soil disturbance around the access pipe is minimized. Electrical conductivity has a major influence on capacitance measurements (Robinson et al. 1998; Kelleners et al. 2004b). Although Robinson et al. (1998) and Kelleners et al. (2004b) show that corrections can be made to the affected capacitance measurement, one needs additional circuit and probe information and an estimate of bulk electrical conductivity. As electrical conductivity varies with water content, the error from unknown conductivity will be greater at higher water contents. This is not likely to be a discernable problem at conductivity below 0:03 S m1 .

REFERENCES Annan, A.P. 1977. Time domain reflectometry— air gap problem for parallel wire transmission lines. Report of Activities, Part B. Geological Survey of Canada. Paper 77-1B: 59–62. Dalton, F.N. 1992. Development of time-domain reflectometry for measuring soil water content and bulk soil electrical conductivity. In G.C. Topp et al., Eds. Advances in Measurement of Soil Physical Properties: Bringing Theory into Practice. Soil Science Society of America, Madison, WI, pp. 143–167. Davis, J.L. and Annan, A.P. 2002. Ground penetrating radar to measure soil water content. In J.H. Dane and G.C. Topp Eds. Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society of America, Madison, WI, pp. 446–463. Davis, J.L. and Chudobiak, W.J. 1975. In situ meter for measuring relative permittivity of soils. Geol. Surv. Can. Pap. 75–1A: 75–79.

Dean, T.J., Bell, J.P., and Bary, A.J.B. 1987. Soil moisture measurement by an improved capacitance technique. Part 1. Sensor design and performance. J. Hydrol. 93: 67–78. Ferre´, P.A., Rudolph, D.L., and Kachanoski, R.G. 1996. Spatial averaging of water content by time domain reflectometry: Implications for twin rod probes with and without dielectric coatings. Water Resour. Res. 32: 271–279. Ferre´, P.A., Rudolph, D.L., and Kachanoski, R.G. 1998. The water content response of a profiling time domain reflectometry probe. Soil Sci. Soc. Am. J. 62: 865–873. Ferre´, P.A. (Ty) and Topp, G.C. 2002. Time domain reflectometry. In J.H. Dane and G.C. Topp Eds. Methods of Soil Analysis, Part 4— Physical Methods, Soil Science Society of America, Madison, WI, pp. 434–446. Galagedara, L.W., Parkin, G.W., and Redman, J.D. 2003. An analysis of the GPR direct ground wave


Soil Water Content method for soil water content measurement. Hydrol. Process. 17: 3615–3628. Galagedara, L.W., Parkin, G.W., Redman, J.D., von Bertoldi, P., and Endres, A.L. 2005. Field studies of the GPR ground wave method for estimating soil water content during irrigation and drainage. J. Hydrol. 301: 182–197. Gardner, C.M.K., Robinson, D., Blyth, K., and Cooper, D. 2001. Soil water content. In K.A. Smith and C.E. Mullins Eds. Soil and Environmental Analysis: Physical Methods, 2nd ed., Marcel Dekker, New York, NY, pp. 1–64. Gardner, W.H. 1986. Water content. In A. Klute Ed. Methods of Soil Analysis, Part 1—Physical and Mineralogical Methods, 2nd ed. American Society of Agronomy, Soil Science Society of America, Madison, WI, pp. 493–544. Gaskin, G.J. and Miller, J.D. 1996. Measurement of soil water content using a simplified impedance measuring technique. J. Agric. Eng. Res. 63: 153–160. Greaves, R.J., Lesmes, D.P., Lee, J.M., and Toksoz, M.N. 1996. Velocity variations and water content estimated from multioffset, groundpenetrating radar. Geophysics 61: 683–695. Grote, K., Hubbard, S., and Rubin, Y. 2003. Field-scale estimation of volumetric water content using ground-penetrating radar ground wave techniques. Water Resour. Res. 39(11): 1321–1335. Heimovaara, T.J., Huisman, J.A., Vrugt, J.A., and Bouten, W. 2004. Obtaining the spatial distribution of water content along a TDR probe using the SCEM-UA Bayesian inverse modeling scheme. Vadose Zone J. 3: 1128–1145. Hook, W.R. and Livingston, N.J. 1995. Propagation velocity errors in time domain reflectometry measurements. Soil Sci. Soc. Am. J. 59: 92–96. Hook, W.R., Livingston, N.J., Sun, Z.J., and Hook, P.B. 1992. Remote diode shorting improves measurement of soil water by time domain reflectometry. Soil Sci. Soc. Am. J. 56: 1384–1391.

959 Huisman, J.A., Hubbard, S.S., Redmond, J.D., and Annan, A.P. 2003. Measuring soil water content with ground penetrating radar: A review. Vadose Zone J. 2: 476–491. Huisman, J.A., Sperl, C., Bouten, W., and Verstraten, J.M. 2001. Soil water content measurements at different scales: Accuracy of time domain reflectometry and ground penetrating radar. J. Hydrol. 245: 48–58. Kelleners, T.J., Soppe, R.W.O., Ayars, J.E., and Skaggs, T.H. 2004a. Calibration of capacitance probe sensors in a saline silty clay soil. Soil Sci. Soc. Am. J. 68: 770–778. Kelleners, T.J., Soppe, R.W.O., Robinson, D.A., Schaap, M.G., Ayars, J.E., and Skaggs, T.H. 2004b. Calibration of capacitance probe sensors using electric circuit theory. Soil Sci. Soc. Am. J. 68: 430–439. Knight, J.H. 1992. The sensitivity of time domain reflectometry measurements to lateral variations in soil water content. Water Resour. Res. 28: 2345–2352. Knight, J.H., Ferre´, P.A., Rudolph, D.L., and Kachanoski, R.G. 1997. A numerical analysis of the effects of coatings and gaps upon relative dielectric permittivity measurement with time domain reflectometry. Water Resour. Res. 33: 1455–1460. Knight, J.H., White, I., and Zegelin, S.J. 1994. Sampling volume of TDR probes used for water content monitoring. In K.M. O’Connor et al., Eds. Time Domain Reflectometry in Environmental Infrastructure and Mining Applications. U.S. Bureau of Mines, Minneapolis and Northwestern University, Evanston, Spec. Publ. SP 19–94: 93–104. Lapen, D.R., Topp, G.C., Bugden, J.L., and Pattey, E. 2004. A new TDR probe for evaluating airborne SAR data for soil water content estimates. In Baoji Wang, Quanzhong Huang, Qing Li, Jianhan Lin, Yu Chen, Feng Mei, Qing Wei, and Jieqiang Zhuo Eds. Proceedings of the 2004 CIGR International Conference (Agricultural Engineering), Bejing, China. Oct. 11–14. (Published on CD-ROM).


960 McNairn, H., Pultz, T.J., and Boisvert, J.B. 2002. Active microwave remote sensing methods. In J.H. Dane and G.C. Topp Eds. Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 475–488. Neal, A. 2004. Ground-penetrating radar and its use in sedimentology: Principles, problems, and progress. Earth Sci. Rev. 66 (3–4): 261. Or, D., Jones, S.B., VanShaar, J.R., and Wraith, J.M. 2003. WinTDR 6.0 Users Guide (Windows-based TDR program for soil water content and electrical conductivity measurement). Available at: http:==129.123.13.101= soilphysics=wintdr=documentation.htm (posted Fall 2003; verified 6 June 2004). Utah Agric. Exp. Stn. Res., Logan, UT. Parkin, G., Redman, D., von Berotldi, P., and Zhang, Z. 2000. Measurement of soil water content below a waste water trench using ground penetrating radar. Water Resour. Res. 36: 2147–2154. Redman, J.D., Davis, J.L., Galagedara, L.W., and Parkin, G.W. 2002. Field studies of GPR air launched surface reflectivity measurements of soil water contents. Proceedings of Ninth Conference on Ground Penetrating Radar. SPIE 4758: 156–161. Robinson, D.A., Gardner, C.M.K., Evans, J., Cooper, J.D., Hodnett, M.G., and Bell, J.P. 1998. The dielectric calibration of capacitance probes for soil hydrology using an oscillation frequency response mode. Hydrol. Earth Sys. Sci. 2: 83–92. Robinson, D.A., Jones, S.B., Wraith, J.M., Or, D., and Friedman, S.P. 2003a. A review of advances in dielectric and electrical conductivity measurements in soils using time domain reflectometry. Vadose Zone J. 2: 444–475. Robinson, D.A., Schaap, M., Jones, S.B., Friedman, S.P., and Gardner, C.M.K. 2003b. Considerations for improving the accuracy of permittivity measurement using TDR: Air=water calibration, effects of cable length. Soil Sci. Soc. Am. J. 67: 62–70. Rucker D.F. and Ferre´, Ty.P.A. 2003. Nearsurface water content estimation with borehole

Soil Sampling and Methods of Analysis ground penetrating radar using critically refracted waves. Vadose Zone J. 2: 247–252. Schwartz, R.C. and Evett, S.R. 2003. Conjunctive use of tension infiltrometry and time-domain reflectometry for inverse estimation of soil hydraulic parameters. Vadose Zone J. 2: 530–538. Starr, J.L. and Paltineanu, I.C. 2002. Capacitance devices. In J.H. Dane and G.C. Topp Eds. Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society of America, Madison, WI, pp. 463–474. Todoroff, P. and Luk, J. 2001. Calculation of in situ soil water content profiles from TDR signal traces. Measure. Sci. Technol. 12: 27–36. Topp, G.C. 1987. The application of time-domain reflectometry (TDR) to soil water content measurement. In Proceedings of International Conference on Measurement of Soil and Plant Water Status. Logan, Utah, pp 1: 85–93. Topp, G.C. and Davis, J.L. 1985. Measurement of soil water content using TDR: A field evaluation. Soil Sci. Soc. Am. J. 49: 19–24. Topp, G.C., Davis, J.L., and Annan, A.P. 1980. Electromagnetic determination of soil-water content: Measurement in coaxial transmission lines. Water Resour. Res. 16: 574–582. Topp, G.C. and Ferre´, Ty.P.A. 2002. Water content. In J.H. Dane and G.C. Topp Eds. Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society of America, Madison, WI, pp. 417–545. Topp, G.C. and Ferre´, Ty.P.A. 2004. Timedomain reflectometry. In D. Hillel et al., Eds. Encyclopedia of Soils in the Environment, Vol. 4, Elsevier, Oxford, UK, pp. 174–181. Topp, G.C. and Reynolds, W.D. 1998. Time domain reflectometry: A seminal technique for measuring mass and energy in soil. Soil Till. Res. 47: 125–132. Topp, G.C., Zegelin, S.J., and White, I. 2000. Impacts of the real and imaginary components of relative permittivity on TDR measurements in soils. Soil Sci. Soc. Am. J. 64: 1244–1252.


Soil Water Content Weiler, K.W., Steenhuis, T.S., Boll, J., and Kung, K.-J.S. 1998. Comparison of ground penetrating radar and time-domain reflectometry as soil water sensors. Soil Sci. Soc. Am. J. 62: 1237–1239.

961 Zegelin, S.J., White, I., and Jenkins, D.R. 1989. Improved field probes for soil water content and electrical conductivity measurement using time domain reflectometry. Water Resour. Res. 25: 2367–2376.



Chapter 71 Soil Water Potential N.J. Livingston University of Victoria Victoria, British Columbia, Canada

G. Clarke Topp Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

71.1 INTRODUCTION Soil water potential may be considered as the ‘‘energy status’’ of the water in the soil pores, relative to some standard reference condition or datum. Soil water potential is defined along with the units of measurement now in common practice (see Section 69.3). Soil water potential is used primarily for determining the direction and rate of water flow between locations with differing potentials (i.e., flow due to a potential gradient or hydraulic head gradient). The potential of soil pore water varies over several orders of magnitude, ranging from positive values in saturated soil to extremely negative values in dry soil. There are numerous instruments and techniques for direct and indirect measurement of soil water potential, but no single approach applies for the entire water potential range commonly found in soils or other natural porous materials. Direct measurement of soil water potential involves determining water pressure or water surface elevation relative to a datum (e.g., pressure transducer, standpipe water level, etc.), while indirect measurement involves measuring some surrogate property that correlates with water potential (e.g., electrical resistance or conductivity, water vapor pressure, water content, plant xylem potential, etc.). For example, a direct determination of positive pore water potential is obtained from the elevation of the water surface in a piezometer pipe, while an indirect determination of negative water potential can be obtained from electrical resistance or relative humidity. This chapter focuses on selected direct and indirect methods for measuring soil water potential, which are well established and commonly used, namely the piezometer method for saturated soils, and the tensiometer, resistance block, and psychrometer methods for unsaturated soils. The emphasis is on basic principles and practical application. More detailed descriptions of these methods and other methods for measuring soil pore water potential can be found in Richards (1965), Brown and van Haveren (1972), Hanks and Brown (1987), Young (2002), Young and Sisson (2002), Andraski and Scanlon (2002),

963



964

Scanlon et al. (2002), and Strangeways (2003). A more general overview of measuring water potential and flow is given in Kramer and Boyer (1995).

71.2 PIEZOMETERS Piezometers are generally of small diameter ( 0:010:1 m), nonpumping wells, which are used primarily to determine water potential (hydraulic head) and the direction of saturated water flow (Young 2002), but can also be used as water quality sampling wells and as a technique for in situ measurement of saturated hydraulic conductivity. We focus here on the measurement of water potential. Piezometers consist essentially of a subsurface intake connected to the bottom of a standpipe or riser pipe which extends above the surface (Figure 71.1) (see also Figure 69.1 for principles of operation of piezometers). The intake in the soil may consist of a prefabricated well screen, a slotted section of the standpipe, or simply the open end of the standpipe. A piezometer operates by allowing water to move freely in or out through the intake in response to imposed changes in standpipe water level, and=or natural changes in piezometric surface elevation, water table elevation, or barometric pressure. The top of the piezometer is normally fitted with a vented cap to maintain atmospheric pressure within the pipe and to prevent unintended entry of water and foreign materials. Piezometers measure the total water potential (ct ) and pressure potential (cp ) in saturated porous materials. If units of energy per unit weight are used, then ct is equivalent to the total Protective surface casing

Ground surface

Protective surface casing

Ground surface

Grout Piezometer riser pipe

Cement grout to ground surface Piezometer riser pipe Sieved native material or engineering sand

Piezometer screen

Blank pipe length

(a)

Engineering sand material Bentonite plug, thickness 1.0–1.5 m

Piezometer screen

Blank pipe length

Sieved native material or engineering sand, extending 60 cm above top of screen

(b)

FIGURE 71.1. A piezometer that is not hydrologically isolated (a) and a piezometer that is hydrologically isolated (b), with isolation being achieved through installation of an impermeable bentonite ‘‘plug’’. (From Young, M.H. and Sisson, J.B., in J.H. Dane and G.C. Topp (Eds.), Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society of America, Madison, WI, 2002. With permission.)


Soil Water Potential

965

elevation of the water surface (piezometric surface) in the piezometer standpipe and the cp is equivalent to the depth of water in the piezometer standpipe (see Figure 69.1). Piezometers do not measure matric potential, which does not exist in saturated materials (see Section 69.3), and the gravitational potential (cg ) is determined by measuring the elevation difference between the midpoint of the piezometer intake and an arbitrary datum, such as mean sea level (see Figure 69.1). Note that a ‘‘water table well’’ may be viewed as a special case piezometer where the intake (e.g., slotted section) extends from the piezometer base to near the porous medium surface. A water table well is used primarily to measure the elevation of the water table in unconfined aquifers.

71.2.1 MATERIAL

AND

SUPPLIES

Pipe Materials Piezometer pipes are constructed from iron, aluminum, stainless steel, acrylonitrile butadiene styrene (ABS) plastic, and other materials. ABS plastic is most commonly used, as it is inexpensive and easily handled. If measurements of pore water chemistry or electrical conductivity are planned, care should be taken to ensure that the selected pipe materials will not affect the results through leaching, adsorption, or electrical interferences. The diameter of the pipe must be greater than the diameter of the devices, selected for measuring cp (often referred to as the ‘‘piezometric head’’ or ‘‘pressure head’’) and for collecting water samples if required. Piezometer Intake The piezometer intake is usually constructed by attaching a length of prefabricated well screen, or by cutting a series of saw slots along the pipe at the desired position (which is usually at the base of the pipe) and covering the slots with a filter cloth (e.g., tile sock). In stable, highly permeable soils (usually coarse-textured sandy materials), a simple wire mesh and=or filter cloth covering the open base of the pipe provides a sufficient intake. Drilling Capability Drilling requirements depend largely on the depth of installation, diameter of the hole, and soil conditions (e.g., texture, density, stoniness, etc.). Options include drill rigs, which are used primarily for stony soils, deep boreholes, or large-diameter boreholes, impact hammers (e.g., Cobra rock drill), which are used for vibrating piezometer pipes into place (work best in saturated medium-coarse sandy materials), and various motorized or manual augers, which are designed for specific soil types and conditions. Surveying Capability To obtain measures of total water potential, the elevation (measured relative to mean sea level or relative to some arbitrarily selected datum) of the piezometer intake and=or ground surface adjacent to the piezometer must be determined to account for variations in surface topography and depth of piezometer installation. Piezometric water levels (i.e., pressure potentials or heads) are usually collected as depth of water from the ground surface and then converted to elevation using the known depth of the piezometer intake.


966


Measurement and Monitoring Water levels can be monitored using either manual or automated devices. Manual collection of water levels from piezometers is most often accomplished using electronic water level indicators, electric tapes, or simple measuring tapes fitted with a ‘‘popper.’’ A popper is simply a small cube or cylinder of closed pore foam (e.g., Styrofoam), split lengthwise, and screwed or cemented around the end of the measuring tape, so that the zero end of the tape floats at the water level. Automated sensors are usually electronic and combined with a data logger. A variety of automated sensors are available, most being pressure-sensing devices based on piezoresistive, strain gauge, or vibrating wire diaphragms, or nonpressuresensing devices based on sonic, radar, or time-domain reflectometry (TDR) technologies. Young (2002) gives a detailed discussion of the setup and use of automated water level sensors. The main advantages of manual water level devices are that they are simple, inexpensive, rugged, and easily portable so that many piezometers can be measured using a single device. Important advantages of automated water level devices include greatly improved temporal resolution and acquisition of data in a format that can be directly downloaded into a computer spreadsheet. Water Extraction Equipment Manual bailer, hand pump, or motorized pump. Sealing Material Bentonite pellets, grout, or cement for preventing ‘‘short circuit’’ flow (see Piezometer Installation, p. 969). Backfill Material Sieved soil from the installation sites may be used. Sandy material is often chosen because of its ‘‘fluidity’’ for pouring around the pipe. Selected Hand Tools and Consumables Pipe wrenches, saw, rope, shovels, glue, couplings, etc. are needed for construction and installation of the piezometers.

71.2.2 PROCEDURE Piezometer Selection and Construction As indicated above, a wide variety of materials are used for piezometers along with a variety of intake designs. The intake is usually tailored to soil conditions to ensure unimpeded water flow, for example, high-hydraulic conductivity soils require less intake area than lowconductivity soils. As a result, sandy soils often use piezometers that are open only at the bottom. Large artificial gravel packs are not normally placed around piezometer intakes (unlike water supply wells), as this can decrease the accuracy of water potential readings, especially when hydraulic gradients exist.



967

Borehole Construction Construct the borehole using an appropriate technology (e.g., drill rig, hand=motorized auger, impact hammer, etc.). Young (2002) gives a summary of the approaches commonly used for piezometer installation, and the shallow boreholes typically required for agrienvironmental applications are amenable to small, mobile drill rigs, impact hammers, and various hand=motorized augers designed for specific soil types and conditions. In loose, sandy materials, installation of shallow piezometers is sometimes possible without digging a borehole, i.e., a conical tip is fitted to the bottom of the piezometer pipe (below the intake), and the pipe is then simply pushed (using a drill rig) or vibrated (using an impact hammer) to the required depth. Piezometer Installation There are two installations under piezometer installations namely the single installations and the nested installations. Single installations are those where one piezometer is installed in the borehole, and nested installations are those where several piezometers are installed at different depths in the borehole. Single installations may be hydrologically ‘‘isolated,’’ where a bentonite or grout seal is placed just above the intake, or ‘‘nonisolated,’’ where no seal is used. Nested installations almost always require hydrologic isolation of each piezometer intake by placement of bentonite, grout or cement seals, above and below ea!ch intake (Figure 71.1). The objective of hydrologic isolation is to prevent short-circuit water flow (i.e., direct water flow between nested intakes, water flow along the standpipe wall and through the backfill material, leakage from the surface), which can invalidate measurements. 1

Connect the piezometer intake (e.g., well screen, slotted pipe section) to the bottom of the first section of standpipe. Lower the intake and standpipe into the borehole, add additional standpipe sections as required to reach the bottom of the hole.

2

Place backfill material into the borehole using either a shovel or funnel, or a tremie tube to fill the hole around the intake to 0.3–0.6 m above the intake (ASTM 1995). A tremie tube is a small-diameter pipe lowered to the bottom of the borehole, into which air-dried and granulated backfill material is poured to provide more uniform and more accurate placement of backfill material. If the piezometer is nonisolated, backfilling is continued to the surface. Backfill should be tamped as much as possible during placement to prevent settlement, and also mounded at the surface to prevent collection of surface water around the standpipe.

3

For isolated piezometers, a seal of bentonite, grout, or cement (see ASTM 1995 for cement compositions) is applied just above each intake in an annular ring around the piezometer standpipe (1–1.5 m thickness) to prevent short-circuit water flow (see above). Tamped backfill is placed between each seal, and once the seal is in place, backfilling is continued to the surface as for nonisolated piezometers.

4

Place a vented cap on the piezometer to maintain ambient atmospheric pressure inside the pipe and prevent entry of rain water or foreign materials.



968

71.2.3 PIEZOMETER RESPONSE TIME AND DEVELOPMENT Piezometer response time is the time required for the water level in the piezometer standpipe to re-equilibrate after an imposed change in pipe water level (as a result of bailing or slug testing), or after a change in the surrounding piezometric surface or water table elevation (as a result of precipitation, drainage, or groundwater extraction). Piezometer response time should reflect the permeability of the porous medium, not the permeability of the intake or the adjacent borehole wall. The installation process often causes partial plugging of the intake with fine particles, and also smearing and compaction of the borehole wall. To ensure that the piezometer response time reflects the permeability of the porous medium, newly installed piezometers are often ‘‘developed,’’ which involves ‘‘surging’’ and rapid extraction of the piezometer water. A pump or bailer can be used to extract water, and surging can be conveniently accomplished by ‘‘bouncing’’ the bailer in and out of the water before removal. Surging seems to be particularly effective at washing off smeared or compacted borehole surfaces and dislodging fines from the intake. Surging and water extraction are continued until the extracted water is free of suspended silt and clay.

71.2.4 MONITORING AND DATA ACQUISITION 1

Establish for each piezometer a ‘‘reference height’’ against which the piezometric head will be measured over time, and determine (by surveying or other means) the elevation of each reference height. The top of the piezometer standpipe is often chosen as the reference height.

2

Check the calibration and operation of the chosen manual or automatic water level monitoring device, and operate as directed in the operator’s manual.

3

Initiate the data collection as required by the study.

71.2.5 CALCULATION CONSIDERATIONS The use of piezometers to calculate total water potential (i.e., ct ¼ cp þ cg ) and total potential gradient (i.e., Dct =Dcg ) involves summations, subtractions, and divisions of values that are often of dissimilar magnitude (e.g., cg ¼ 300 m, cp ¼ 2 m). This can greatly magnify measurement errors, and it is therefore critically important that substantial effort be expended in ensuring that the cp and cg measurements are as precise and accurate as possible.

71.3 TENSIOMETERS Tensiometers are instruments for in situ measurement of total pore water potential in unsaturated porous materials, where the total potential, ct , is the summation of the negative matric potential, cm , and the gravitational potential, cg (i.e., ct ¼ cm þ cg ) (refer to Figure 69.1 for the similarities and the differences between piezometer and tensiometer operation). Tensiometers are used widely in the determination of water potential, water potential gradients, aeration, and water availability to plants. They are relatively inexpensive, simple, and easy to install—particularly well suited to studies where large numbers of measurements



969

may be required. Tensiometers are used primarily in unsaturated porous materials, and thus complement piezometers, which operate only in saturated porous materials. Tensiometers typically consist of a water-filled plastic tube, which has a saturated porous cup (usually made from fired ceramic or porous metals) sealed to the bottom, a removable cap, rubber stopper, or rubber septum sealed to the top, and some device for measuring matric potential (e.g., manometer, pressure gauge, pressure transducer) (Figure 71.2). When installed in unsaturated soil, the saturated porous cup prevents air entry into the tensiometer, and provides hydraulic connection between the tensiometer water and the soil water. Hence, the matric potential of the tensiometer water equilibrates to the negative matric potential of the soil water, which is recorded by the measuring device. As solutes pass freely through the porous cup, tensiometers are insensitive to osmotic potential, cp . As the matric potential decreases (becomes more negative), the water in the tensiometer tends to gradually degas and vaporize. Degassing can reduce or prevent tensiometer response by forming obstructive bubbles, whereas progressive vaporization will cause the tensiometer to empty gradually, which can cause erroneous matric potential readings and eventual failure

FIGURE 71.2. Some selected tensiometers. From left to right: tensimeter1 fitted with septum cap and tensimeter readout device; customized tensimeter modified to accept a gauge or pressure transducer readout; tensiometer showing two optional screw caps and a manual gauge readout; pressure transducer for use as an alternative to a manual gauge; tensiometer fitted with a flexible ‘‘spaghetti’’ tubing and a mini porous cup; scaled-down tensiometer with midsized cup. Note that all the porous cups are ceramics, which is the most popular and least expensive of the available porous materials.


970


of the tensiometer. As a result, the effective measurement range of tensiometers is about 0:08 MPa cm 0. Methods for dealing with air bubbles are discussed by Miller and Salehzadeh (1993).

71.3.1 MATERIAL AND SUPPLIES Tensiometer Cups, Tubes, and Caps Fully assembled tensiometers are available from suppliers, such as Soil Measurement Systems, Soilmoisture Equipment Corp., and Irrometer Co. (Figure 71.2). It is also possible to purchase component parts and assemble tensiometers for specific applications. A detailed list of the materials required and the procedures necessary to construct tensiometers are available from Soil Measurement Systems. Pressure-Indicating or Recording Devices The companies listed above also provide a variety of pressure-sensing devices. Soil Measurement Systems offers a tensimeter, which uses a hypodermic needle and a battery-powered pressure transducer to measure the partial vacuum above the water in the tensiometer tube (upper left in Figure 71.2). The other two companies use either mechanical pressure gauges or electronic pressure transducers, depending on the requirement of the application. U-tube manometers containing heavy liquids (e.g., mercury) were used frequently in the past, but are now avoided because of inconvenience, health and environmental concerns. Pressure Readout Devices Mechanical gauges and manometers must be read and recorded manually. Pressure transducers are usually battery-powered and some (e.g., the tensimeter) have a digital readout, while others require data logging. Selected Hand Tools Hand-operated vacuum pump, wrenches, shovels, and augers, etc.

71.3.2 PROCEDURE Filling the Tensiometer Saturate the tensiometer cup in de-aired, temperature equilibrated water, a process that can be quickened by applying a slight vacuum (suction) to the tensiometer tube. Once the ceramic tip is saturated (water appears in the tensiometer tube), fill the tensiometer to the prescribed level by removing the cap or septum and directly pouring de-aired temperature equilibrated water into the top of the tube. Attach the pressure-indicating device and assure that its tubing and connection are water-filled. Note that the tensimeter requires a small airspace below the rubber septum, while other pressure-indicating devices respond more effectively in the absence of air. Field Installation For relatively shallow installations (say, 1 m) in relatively stone-free soils, bore a hole to the desired depth using an auger with the same outside diameter as the tensiometer (sliding fit without gap), and then carefully insert the tensiometer, ensuring good hydraulic



971

connection between the tensiometer cup and the soil at the bottom of the hole. Two approaches are recommended for establishing good hydraulic connection between the soil and the cup: (i) place a small amount of water-slurried soil (all stones and grit are removed) in the bottom of the hole before tensiometer insertion and (ii) fully insert the tensiometer, then wet the soil immediately surrounding the cup by removing the cap=septum to allow a small amount of water outflow. After the tensiometer is installed, refill the tensiometer and replace the cap (if required), and then wait for the tensiometer to equilibrate to the matric potential of the soil. Note that the time required for initial equilibration increases with the amount of water-slurried soil used, or the amount of water outflow allowed, when establishing the hydraulic connection between soil and cup. For a full description of field installed tensiometers equipped with pressure gauges, see Marthaler et al. (1983). For relatively deep installations (say, >1 m) and unstable soils, an access tube (metal or hard plastic) can be installed, and then the tensiometer inserted through the access tube until the porous cup protrudes below the access tube and about 0.05–0.1 m into the soil. In stoney soils, considerable care is required to avoid damaging the porous cup during tensiometer installation and placement of stone-free, water-slurried soil in the bottom of the hole is usually necessary to establish good hydraulic connection between the soil and the cup. Above ground, tubing should be kept to a minimum to reduce bubble formation, and it should also be shielded to avoid damage and solar heating effects. Recording the Pressures After the initial equilibration, the tensiometer pressure is a continuous indication of soil matric potential. Data logged transducers are easily programmed to sample at the desired frequency and time of day, whereas manual readings taken early in the morning in order to minimize temperature and solar heating effects are considered to be best. Tensiometer Maintenance Dissolved air diffuses slowly through the porous cup and can exsolve inside the tensiometer tube, forming undesirable bubbles, which impede water movement and thereby impair (increase) response time. To maintain optimum performance, it is necessary to remove exsolved air periodically, which accumulates more rapidly as the matric potential approaches the low end of the tensiometer operating range (i.e., 0:08 MPa). If the matric potential decreases below 0:1 MPa, then the tensiometer rapidly fills with exsolved air and becomes operative again only after it is recharged with water, which can be done in place when the soil water matric potential is again greater than 0:08 to 0:1 MPa.

71.3.3 CALCULATION CONSIDERATIONS Tensiometer readout devices usually give matric potential directly, and thus no additional calculations are necessary. As tensiometers are often used to determine gradients in potential, they are subjected to the same accuracy and precision concerns as piezometers, which in turn influences the choice of readout device and the required frequency of recalibration.

71.3.4 COMMENTS 1

For a vertically installed tensiometer, the absolute pressure head measured by the tensiometer gauge, P (positive quantity), is given by


972

Soil Sampling and Methods of Analysis P ¼ A þ cm h

(71:1)

where cm is the negative matric potential (head) at the tensiometer cup (negative quantity), h is the vertical height of the gauge above the tensiometer cup (positive quantity), and A is the ambient atmospheric pressure head (positive quantity). The minimum measurable cm is therefore given by cm ¼ A þ h

(71:2)

as this corresponds to P ¼ 0 (i.e., complete vacuum) at the tensiometer gauge; and the measurable range of cm is consequently (A þ h) cm 0

(71:3)

Given that the average atmospheric pressure is A ¼ 10 m, then h < 10 m is required to provide a usable range of cm and for most field applications, h ¼ 4 m is the practical maximum, as this yields a measurable matric potential range of 6 m cm 0 according to Equation 71.3. As a result, matric potential measurements at depths greater than about 3.5 m are best achieved using short, buriable transducer-based tensiometers, which are fitted with wire leads that extend to the surface, to allow monitoring. 2

It is advisable to incorporate a short section of a clear plastic tubing at the upper end of the tensimeter tube if the tensiometer readout system is to be used (Figure 71.2). The clear plastic allows the headspace (air gap) at the top of the tensiometer to be seen so that one can ensure that the tensimeter needle is inserted into air and not into water.

3

Tensiometers can provide accurate and reliable measurements of soil water matric potential in moist soils (i.e., 0:08 MPa cm 0), with a precision of about 0:0001 MPa if the pressure-sensing devices are well maintained and calibrated.

4

If a large number of tensiometers are required, costs can be reduced by using a single-pressure transducer to read several tensiometers via an automated switching (scanning) valve and a data logging system.

71.4 RESISTANCE BLOCK Resistance blocks are a relatively inexpensive means of providing a continuous estimate of the negative matric potential, cm , of relatively dry soil (i.e., cm 0:05 MPa). They are typically composed of an engineered matrix of hydrophilic porous materials, such as gypsum, fiberglass, or nylon, within which two electrodes are embedded. The blocks are buried at the desired depth in the soil, where they absorb or desorb water until the energy status of the block water equals that of the soil water. Soil matric potential is then inferred by measuring the electrical conductivity (or resistance) across the two embedded electrodes, and then applying a conductivity (or resistance) versus matric potential calibration curve. The electrical conductivity of a dry block is effectively zero, and it normally increases with block water content in a nonsaline environment. Note, however, that resistance blocks do not work well in saline environments (e.g., saline soils; saline irrigation water) as electrical conductivity later becomes insensitive due to changes in water content.



973

Resistance blocks react relatively slowly to changes in soil water potential (because they equilibrate by absorbing or desorbing water), and should therefore not be used to track the movement of wetting fronts. In addition, very large measurement errors can arise if the blocks are not in complete equilibrium with the soil, or if they are subjected to large temperature variations (Carlson and El Salam 1987).

71.4.1 MATERIAL

AND

SUPPLIES

Resistance Blocks and Readout Devices Blocks of various materials are commercially available from a wide variety of suppliers. Each supplier offers compatible voltage supply and data logger or readout alternatives. General data loggers and power supplies having a.c. voltage output may also be used with most resistance blocks. Hand Tools Selected hand tools used primarily for block installation (e.g., shovels, augers, etc.).

71.4.2 PROCEDURE Calibration The logarithmic calibration curve that relates soil water matric potential to electrical conductivity or resistance is usually hysteretic and specific to each block and soil type. Hence, each block should be individually calibrated for wetting up and drying down using the same soil from which the field or laboratory measurements will be collected. It is further recommended that the pressure plate apparatus can be used to conduct the calibration, as it operates well within the measurement range of resistance blocks, and it allows electrical conductance, water content, and matric potential to be measured simultaneously. Note also that resistance blocks tend to degrade over time (e.g., due to slow gypsum dissolution; formation of small water-conducting cracks), and therefore need recalibration at approximately 3 month intervals. Installation Before installation, presoak the blocks for at least 24 h using water slurry made from the appropriate soil. Considerable care should be taken during installation to ensure minimum soil disturbance and good hydraulic connection between the block and soil. Note that soil disturbance (e.g., compaction) or damage to surface vegetation can lead to dramatic changes in the original soil water balance.

71.4.3 CALCULATION CONSIDERATIONS The resistance block calibration curves can often be programmed into data logger and readout devices, which consequently allows direct readout of soil water potential (and sometimes also water content) without further manipulation. The precision of resistance block cm data is rather low (typically ranging from 0:1 to 0:5 MPa), which precludes their use for determining gradients in water potential. They are also sensitive to temperature



974

(i.e., matric potential changes by approximately 1.5% for each Kelvin change in temperature), although partial temperature correction is possible if block temperature is monitored using thermistors or thermocouples.

71.4.4 COMMENTS An important and unique advantage of resistance blocks is that they can provide a continuous and automated measure of both matric potential and water content for dry soil. On the other hand, equally important disadvantages include the requirement for regular laboratory recalibration (which usually means laborious and careful removal and reinstallation), slow equilibration, and generally low precision (which precludes their use for estimating potential gradients).

71.5 THERMOCOUPLE PSYCHROMETERS Thermocouple psychrometers provide an excellent complement to tensiometers in that they operate in the ct 0:1 MPa water potential range, whereas tensiometers operate in the 0:08 MPa ct 0 range. There are both field-based psychrometer systems that measure water potential in situ (e.g., Wescor, Inc., Logan, Utah) and laboratory-based systems that measure water potential in intact or disturbed samples (e.g., Decagon Devices, Inc., Pullman, Washington). Thermocouple psychrometers operate by relating the total potential of the liquid water in the sample to the equilibrium water vapor pressure in the air above the sample. The relation between water potential and relative water vapor pressure at thermodynamic equilibrium is given by RT e ln ct ¼ Vw es

(71:4)

where R is the universal gas constant (J mol1 K1 ), T is the absolute temperature (K), e is the water vapor pressure of the air (Pa), and es is the saturation vapor pressure (Pa) at the air temperature. The dimensionless ratio, e=es , is the relative water vapor pressure or ‘‘relative humidity.’’

71.5.1 IN SITU SOIL PSYCHROMETERS Rawlins and Dalton (1967), Lang (1968), and Weibe et al. (1971) describe soil psychrometer devices that can measure soil water potential in situ. A soil psychrometer usually consists of a small porous cup (about 1 cm in diameter and 1 cm in length) that contains a single thermocouple (50100 mm in diameter). The cup is usually made of porous ceramic, brass, or stainless steel (330 mm pore size), which allows water vapor to diffuse between the soil and the inside of the cup until vapor pressure equilibrium is established. The sensing junction of the thermocouple is constructed of very fine, welded chromel and constantan wires, while the reference junction is connected to much larger (>0:40 mm diameter) copper wires. The open end of the porous cup is sealed with a Teflon plug, and in some psychrometers (Szietz 1975), another thermocouple is embedded in this region to provide a measure of the psychrometer temperature.



975

Modes of Operation Psychrometric Mode of Operation When an appropriate current is applied to the sensing junction in a thermally equilibrated psychrometer system, the junction cools by the Peltier effect. With continued cooling, the junction temperature falls below the dew point of the air in the porous cup so that a drop of water condenses on the junction. The maximum cooling is about 5 C below the ambient cup temperature. When the soil water potential is less than 0 MPa, the relative humidity inside an equilibrated porous cup will be less than 100% and the water drop will consequently re-evaporate and cool the junction to a temperature that can be related to the relative humidity inside the cup. This decline in temperature can be measured with a voltmeter that has microvolt or nanovolt sensitivity, and the output voltage is typically about 5 mV MPa1 . The temperature of the psychrometer is also measured so that the soil water potential can be calculated from Equation 71.4. The relationship between psychrometer output voltage and water potential can be obtained by immersing the psychrometer in a range of constant temperature salt solutions with varying water potentials (Lang 1967). Very small temperature depressions are generated at the sensing junction during measurements so that any temperature gradient between the sensing junction and the reference junction will lead to large errors. For example, a temperature difference of 0:001 C corresponds to an error of 0.01 MPa (0.1 bar). It is therefore imperative that there are no temperature gradients across the sensor or leads, and as a result, soil psychrometers usually cannot be used where large temperature gradients exist (e.g., within 0.15–0.30 m of the soil surface). There have been numerous attempts to design thermocouple psychrometers that can accurately measure soil water potential in the presence of temperature gradients. For example, Campbell (1979) designed a psychrometer containing both high thermal conductivity materials to minimize temperature gradients and symmetrically arranged ceramic ‘‘windows’’ to improve vapor exchange with the soil and thereby reduce internal condensation. These improvements reduced measurement errors due to temperature gradients to about one-third of those for earlier designs. Dew Point Mode of Operation Many psychrometers can be used in a ‘‘continuous feedback’’ dew point mode (Neumann and Thurtell 1972), which is often referred to as the ‘‘dew point hygrometer’’ system. Here, Peltier cooling is again used to condense a water droplet on the sensing junction; however, the cooling current is continuously adjusted so that there is no net gain or loss of water vapor. In unsaturated soil conditions, the dew point temperature will be below ambient and this temperature difference will be measured as a differential voltage between the reference junction and the sensing junction when no current is flowing. Although dew point hygrometers are still sensitive to internal temperature gradients, they are much less sensitive than psychrometers to changes in ambient temperature, and therefore require no temperature correction. In addition, they provide a relatively large output signal with a sensitivity of approximately 0:75 mV MPa1 , which is about 50% greater than that provided by psychrometers. Further, since there is no net movement of water from the thermocouple junction to the chamber at the dew point, the signal is stable for long periods and the vapor equilibrium in the chamber is not disturbed.



976 Apparatus and Procedures 1

Psychrometers have two main components: the porous cup with its contained sensing and reference junctions and the instrument for generating the electrical current and measuring the psychrometer output. Wescor Inc. offers soil psychrometers, combined psychrometer–hygrometer systems, and various data logging systems for continuous monitoring.

2

To minimize thermal gradients, psychrometers should be installed with the axis of the sensor parallel to the soil surface, and in situ psychrometers should not be installed at depths shallower than 0.15–0.30 m. To reduce heat conduction along lead wires, at least two loops of wire (about 0.04 m long) should be wrapped up behind the psychrometer sensing head and buried at the same depth. Rundel and Jarrell (1989) recommend that if continuous data records are required, extra psychrometers should be installed to replace those that fail or are removed for calibration checks. Use of psychrometers to measure the water potential in greenhouse or outdoor pots is often not successful because of large temperature gradients that are difficult to control.

Psychrometer Maintenance For valid and accurate results, it is critically important that thermocouple psychrometer components remain free of contamination and corrosion, as contaminated and corroded surfaces dramatically delay vapor equilibration, and also invalidate the calibration relationship by causing a nonuniform internal vapor concentration and nonconstant evaporation from the sensing junction. Unfortunately, the porous nature of the psychrometer cup allows contamination of its internal surfaces and structures by allowing the entry of dissolved soil salts that form precipitates and promote corrosion. Hence, all internal surfaces in the psychrometer cup must be cleaned periodically. Simply running water over the cups for several hours can sometimes effect adequate cleaning, although most commercial units are now easily disassembled for more thorough cleaning. It is often recommended that thermocouples and their mounts be cleaned by immersion in steam or solvents (such as reagent grade acetone or 10% ammonium hydroxide solution), then thoroughly rinsed in distilled or deionized water (especially if solvent cleaners were used). Wescor Inc. strongly recommends that the water used for rinsing should be pure enough to have an electrical resistance of at least 1 megohm (106 V) cm3 . Psychrometers should be dried by blowing with clean (filtered) air. Savage et al. (1987) further recommend that the screen cage covers attached to commercially available soil psychrometers be removed and soaked in a 10:1 mixture of water and hydrochloric acid to remove any traces of rust. This is particularly important if the devices were calibrated in salt solutions. The screens should then be soaked in acetone to reduce the possibility of fungal growth. Psychrometers fabricated from stainless steel are highly resistant to internal corrosion, but are generally much more expensive than those made from brass. Corrosion resistance of brass psychrometers can be quite improved considerably by chrome or nickel plating. In saline soils, psychrometers should be checked frequently for corrosion of the fine thermocouple wires.

71.5.2 LABORATORY DEW POINT PSYCHROMETERS The dew point psychrometer, which was first developed for the food industry, has been adapted to soils applications (Scanlon et al. 2002) as it is much less sensitive to temperature



977

effects than the thermocouple psychrometer. It operates by measuring the dew point temperature inside a chamber, which is in thermal and vapor pressure equilibrium with a soil sample. Laboratory dew point psychrometers are commercially available, and this section will focus on the system offered by Decagon Devices Inc., known as the WP4 ‘‘Dewpoint PotentiaMeter.’’ The WP4 Dewpoint PotentiaMeter uses a sealed chamber to equilibrate the liquid-phase water of the sample with the vapor-phase water in the headspace above the sample. A mirror situated in the headspace above the sample is Peltier-cooled until the dew point is reached, which is detected as a sudden decrease in mirror reflectance because of condensation. The dew point and sample temperatures are then recorded and used to calculate the headspace water vapor pressure (e) and the saturated water vapor pressure (es ), respectively. Water potential is then calculated using Equation 71.4.

Apparatus and Procedures 1

Load each sample (7 mL volume) into a separate psychrometer sample holder. Slightly compress using a rubber stopper or square-ended metal rod to produce a flat surface and a uniform thickness of 0:5 cm (surface leveling and slight compression tends to produce more reliable results—Gee et al. 1992). The loaded sample holders should be sealed in vapor-tight containers to prevent loss or gain of water before analysis.

2

Insert one loaded sample holder into the psychrometer and start the measuring process (samples are inserted and measured one at a time). The WP4 Dew point PotentiaMeter requires 5 min per measurement and displays both soil water matric potential and sample temperature after internal calculations.

Comments 1

The operating range of the WP4 is about 40 MPa ct 0:1 MPa. A precision of 0:1 MPa can be achieved if the difference between the sample temperature and the dew point temperature is known within 0:005 C, and if the sample temperature is within 0:5 C of the chamber temperature. Hence, the temperature sensors of the instrument must be extremely accurate, and it is advisable to house the prepared samples and psychrometer instrument together in a constant temperature room (e.g., 20 C 1 C).

2

With a precision limit of 0:1 MPa, the WP4 instrument is most effectively applied to very dry soils, i.e., ct 0:4 MPa. When working with disturbed field samples at such low potentials, the main source of measurement error arises from changes in sample water content (and thereby water potential) during sample collection, transport, and storage. This is because the soil water characteristic curve is very flat in this water potential range, and thus a very small change in water content can produce a very large change in water potential. Hence, great care must be taken to prevent water evaporation or condensation when the samples are collected, transported, and stored.



978

71.6 CONCLUDING REMARKS 1

Selection of the most appropriate devices for measuring water potential depends strongly on the anticipated water potential range (i.e., positive, slightly negative, strongly negative), the intended use of the data, and the limitations of the devices (e.g., operating range, accuracy, etc.). Furthermore, all devices must be carefully installed and maintained, else incorrect and misleading data will likely result.

2

Note that a tensiometer is capable of operating as a piezometer (i.e., record positive water potentials in saturated porous materials), but its response time may be impractically or unacceptably slow because of flow impedance by the porous cup.

3

Piezometers remain the simplest and most reliable method for measuring saturated porous materials with positive water potentials (i.e., cp > 0), while tensiometers are recommended for wet but unsaturated materials (i.e., 0.08 MPa cm 0). Measurements in dry porous materials (i.e., cm 0:1 MPa) are best achieved using thermocouple psychrometers, thermocouple hygrometers, and dew point psychrometers.

Note, however, that great care should be taken when using thermocouple-based devices since very large errors can occur if they are incorrectly calibrated, poorly maintained, or subjected to strong temperature gradients. Moisture blocks should only be used to obtain a general indication of water potential, even when they are carefully installed and maintained.

REFERENCES Andraski, B.J. and Scanlon, B.R. 2002. Thermocouple psychrometry. In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4— Physical Methods, Soil Science Society of America, Madison, WI, pp. 609–642. ASTM (American Society for Testing and Materials). 1995. Standard practice for design and installation of ground water monitoring wells in aquifers—Method D5092-90(1995)e1. ASTM, West Conshohocken, PA. Brown, R.W. and van Haveren, B.P. 1972. Psychrometry in water relations research. Proceedings of the Symposium on Thermocouple Psychrometers. Utah Agricultural Experimental Station, Utah State University, Logan, UT, pp. 1–27. Campbell, G.S. 1979. Improved thermocouple psychrometers for measurement of soil water potential in a temperature gradient. J. Phys. E. Sci. Instrum. 12: 739–743.

Carlson, T.N. and El Salam, J. 1987. Measurement of soil moisture using gypsum blocks. In R.J. Hanks and R.W. Brown, Eds. Proceedings of International Conference on Measurement of Soil and Plant Water Status. Vol. 1. Logan, Utah State University, UT, July 1987, pp. 193–200. Gee, G.W., Campbell, M.D., Campbell, G.S., and Campbell, J.H. 1992. Rapid measurement of low soil water potentials using a water activity meter. Soil Sci. Soc. Am. J. 56: 1068–1070. Hanks, R.J. and Brown, R.W. 1987. Eds. Proceedings of International Conference on Measurement of Soil and Plant Water Status, Vol 1, Logan, UT, July 1987, pp. 115–119. Kramer, P.J. and Boyer, J.S. 1995. Water Relations of Plants and Soil. Academic Press, San Diego, CA. Lang, A.R.G. 1967. Osmotic coefficients and water potentials of sodium chloride solutions from 0 to 40 C. Aust. J. Chem. 20: 2017–2023.


Soil Water Potential Lang, A.R.G. 1968. Psychrometric measurement of soil water potential in situ under cotton plants. Soil Sci. 106: 460–464. Marthaler, H.P., Vogelsanger, W., Richard, F., and Wierenga, P.J. 1983. A pressure transducer for field tensiometers. Soil Sci. Soc. Am. J. 47: 624–627. Miller, E.E. and Salehzadeh, A. 1993. Stripper for bubble-free tensiometry. J. Soil Sci. Soc. Am. 57: 1470–1473.

979 potential. In R.J. Hanks and R.W. Brown, Eds. Proceedings of International Conference on Measurement of Soil and Plant Water Status. Vol. 1. Logan, Utah State University, UT, July 1987, pp. 119–124. Scanlon, B.R., Andraski, B.J., and Bilskie, J. 2002. Miscellaneous methods for measuring matric or water potential. In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society of America, Madison, WI, pp. 643–670.

Neumann, H.H. and Thurtell, G.W. 1972. A Peltier cooled thermocouple dewpoint hygrometer for in situ measurement of water potential. In R.W. Brown and B.P. van Haveren, Eds. Psychrometry in Water Relations Research. Utah Agricultural Experimental Station, Logan, Utah State University, UT, pp. 103–112.

Strangeways, I. 2003. Measuring the Natural Environment. Cambridge University Press, Cambridge, UK.

Rawlins, S.L. and Dalton, F.N. 1967. Psychrometric measurements of soil water potential without precise temperature control. Soil Sci. Soc. Am. Proc. 31: 297–300.

Weibe, H.H., Campbell, G.S., Gardner, W.H., Rawlins, S., Cary, J.W., and Brown, R.W. 1971. Measurement of plant and water status. Utah Agricultural Experimental Bulletin 484. Utah State University, Logan, UT.

Richards, L.A. 1965. Physical conditions of water in soil. In C.A. Black et al., Eds. Methods of Soil Analysis. American Society of Agronomy, Madison, WI, pp. 128–152. Rundel, P.W. and Jarrell, W.R. 1989. Water in the environment. In R.W. Pearcy, J. Ehleringer, H.A. Mooney, and P.W. Rundel, Eds. Plant Physiological Ecology. Chapman and Hall, New York. Savage, M.J., Ritchie, J.T., and Khuvutlu, I.N. 1987. Soil hygrometers for obtaining water

Szietz, G. 1975. Instruments and their exposure. In J.L. Monteith, Ed. Vegetation and the Atmosphere. Vol. 1. Academic Press, London.

Young, M.H. 2002. Piezometry. In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4—Physical Methods, SSSA Book Series No. 5, Soil Science Society of America, Madison, WI, pp. 547–573. Young, M.H. and Sisson, J.B. 2002. Tensiometry. In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society of America, Madison, WI, pp. 575–608.



Chapter 72 Soil Water Desorption and Imbibition: Tension and Pressure Techniques W.D. Reynolds Agriculture and Agri-Food Canada Harrow, Ontario, Canada

G. Clarke Topp Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

72.1 INTRODUCTION Soil water desorption refers to the decrease in soil volumetric water content with decreasing pore water matric head (drainage), while imbibition refers to the increase in volumetric water content with increasing matric head (wetting). A discussion of the principles and parameters associated with the determination of desorption and imbibition curves is given in Chapter 69. This chapter describes the tension table, tension plate, and pressure extractor methods for measuring soil water desorption and imbibition curves. Alternative methods include the long column (Chapter 73), dewpoint psychrometer (Chapter 74), soil core evaporation (Chapter 81), instantaneous profile (Chapter 83), and estimation techniques (Chapter 84).

72.2 TENSION TABLE AND TENSION PLATE The tension table and tension plate methods are used primarily for soil cores that are less than 20 cm diameter by 20 cm long, although larger samples can be used. These methods involve establishing a continuous hydraulic connection between the sample and the tension medium or plate, and then sequentially equilibrating the sample to a series of preselected matric heads set on the table or plate. The water content of the sample after equilibration represents one point on the desorption or imbibition curve. If the soil sample is initially saturated and the preselected matric heads are set in a descending sequence (i.e., successively more negative), a desorption curve is obtained. If the soil sample is initially dry and the preselected matric heads are set in an ascending sequence (successively less negative), an imbibition curve is obtained. Scanning curves are obtained by reversing the sequence of matric heads

981



982

(i.e., from ascending to descending or vice versa) at some intermediate point along the desorption or imbibition curves (see Chapter 69 for details).

72.2.1 MATERIALS AND SUPPLIES 1

Tension table: A tension table (Figure 72.1a) consists of a circular or rectangular tank containing a saturated layer of tension medium, a port for allowing water inflow and outflow, and an apparatus for changing and controlling the matric head, cm , of the pore water in the tension medium (Stakman et al. 1969). The tension medium has an air-entry value, ca (see Chapter 69), that is lower (more negative) than the minimum matric head set on the medium, plus a high-saturated hydraulic conductivity to minimize sample equilibration time (Topp and Zebchuk 1979). For convenience and improved equilibration times, ‘‘low-tension’’ and ‘‘high-tension’’ tables are often set up, with low-tension tables operating in the range of 1 m cm 0 m, and high-tension tables operating in the 5 m cm 1 m range (Topp and Zebchuk 1979). The tension medium in low-tension tables is usually natural fine sand (< 50 mm particle diameter), fine glass beads (42 mm mean particle diameter), or silica flour (1050 mm particle diameter), while the high-tension tables generally use glass bead powder (25 mm mean particle diameter) or aluminum oxide powder (9 mm mean particle diameter) (see also Table 72.1). Beneath the tension medium is a water inflow– outflow port that is connected to the apparatus for setting and maintaining matric head, a drainage system to facilitate water movement into or out of the tension medium, and a fine mesh retaining screen to prevent loss of tension medium through the inflow–outflow port. The top of the tension medium may be protected by a cloth cover (nylon mesh) to prevent the tension medium from adhering to the soil core samples. Two main tension table designs are currently in use: (i) a rectangular perspex (acrylic) tank with a drainage system comprised of a glass microfiber retaining screen overlying a network of channels cut into the tank bottom (Ball and Hunter 1988) and (ii) a cylindrical polyethylene or PVC tank (Figure 72.1a) with a drainage system comprised of a retaining screen made of finegrade nylon mesh (620 mm openings depending on fineness of tension medium) overlying a coarse mesh woven stainless steel screen ( 3 mm openings) placed on the tank bottom (Topp and Zebchuk 1979). Both designs should have loosefitting, opaque lids for the following reasons: (i) to allow easy air exchange, and thereby prevent possible buildup of vacuum or pressure when the matric head is changed; (ii) to prevent evaporative water loss from the sample surfaces during the course of the measurements; and (iii) to omit light to inhibit fungal=algal=microbial growth on and in the samples during the course of the measurements. The tank needs to be stiff enough to resist flexing over the applied matric head range, as this can result in air entry due to cracking of the tension medium or breaking of the seal between the tension medium and tank wall (for cylindrical polyethylene or PVC tanks, a wall and base thickness of at least 0.5 cm is recommended for the low-tension system, and at least 1.5 cm for the high-tension system). A 60 cm diameter tension tank can accommodate up to 30 cores with a 7.6 cm diameter, and up to 12–15 cores with a 10 cm diameter. Although tension tables are relatively easy and inexpensive to construct, they can require frequent maintenance as a result of algal growth and periodic plugging of the retaining screen or tension medium with silt and clay; and they can also be unreliable at


Soil Water Desorption and Imbibition: Tension and Pressure Techniques

Constant head burette

10 9 8 7

1 6 5 4

5

Tape measure

10

2

3

(a)

983

1. Tray 2. Drainage mesh 3. Cementing seal 4. Retaining mesh 5. Tension medium

6. Cloth cover 7. Rubber band 8. Nylon cloth 9. Core cylinder 10. Cover disc

200

Manometer

Tray outlet

Vacuum regulator

Trap

To vacuum source Overflow reservoir

(b)

Gauge Drain outlet

Regulator

10 9 8 7 12 11

Pressure input Pressure vessel

(c)

11. Hydraulic contact slurry 12. Ceramic plate

FIGURE 72.1. Tension table=plate and pressure extractor systems: (a) a low-tension table with matric head control by constant head burette; (b) a controlled vacuum system for use with high-tension tables or plates; and (c) a pressure extractor system. Note that the burette reservoir and trap (storage) flask are the water sources for the imbibition curve.



984

TABLE 72.1 Approximate Equilibration Times (Days) for Desorption from Saturation. The Tension Table and Tension Plate Times Apply to 7.6 cm Long Intact (Undisturbed) Soil Cores, or 10 cm Long Intact Cores (Bracketed Values). The Pressure Extractor Times Refer to 7.6 cm Long Intact Cores (First Column) or 1.0 cm Thick Samples that Have Been Granulated to 2 mm (Second Column) Tension table or tension plate with corresponding tension medium or contact material

Matric head, c m (m)

Glass beads (42 mm mean particle diameter)

Glass beads (25 mm mean particle diameter)

Aluminum oxide (9 mm mean particle diameter)

Silica flour (1050 mm particle diameter)

0 0:05 0:1 0:2 0:3 0:4 0:5 0:6 0:75 0:8 1:0 1:5 2:0 3:3 5:0 10 40 150

0.5 (1) 1 (2) 1 (2) 1 (2) 2 (3) 3 (4) 3 (4) 3 (4) 4 (5) 4 (5) 6 (7) — — — — — — —

— — — — — — — — — — — 8 (9) 10 (11) 12 (13) — — — —

— — — — — — — — — — — — — 12 (13) 14 (15) — — —

— — 4 (5) 5 (6) 6 (7) 7 (8) 8 (9) 9 (10) 9 (10) 9 (10) 10 (11) 11 (12) 12 (13) 14 (15) — — — —

a

Pressure Pressure extractor extractor (intact (granulated core) sample) — — — — — — — — — — — — — 10a 11a 12a 15a 20a

— — — — — — — — — — — — — — — 3a 7a 10a

Equilibration times for pressure extractors are very sensitive to sample and plate hydraulic characteristics, and to the degree of hydraulic connection between sample and plate. Hence, determination of equilibration is best achieved by monitoring plate outflow (e.g., Figure 72.3).

matric heads < 1 m due to air entry and air accumulation problems (Topp and Zebchuk 1979; Townend et al. 2001). 2

Tension plate: An alternative to the tension table is the so-called ‘‘tension plate’’ (Figure 72.2). The essential difference between a tension plate and a tension table is that the tension plate uses a large-diameter ceramic disc or ‘‘plate’’ (e.g., 50 cm diameter) instead of tension medium. The ceramic plate is usually designed to have a relatively high saturated hydraulic conductivity and minimum operating matric heads of 5 to 10 m. Although tension plates are more expensive than tension tables and must usually be purchased from a commercial supplier (e.g., Soilmoisture Equipment Corp., California), they are very reliable and much easier to operate and maintain. Good hydraulic connection between the plate and the core samples is established and maintained using a thin layer ( 0:5 cm) of saturated ‘‘contact material’’ made of fine-grade glass beads (42 mm



985

FIGURE 72.2. Example of tension plate apparatus (high-tension system) loaded with 10 cm inside diameter by 11 cm long intact soil cores. The cores are resting on a nylon cloth (15 mm pore size), which overlies a 0.5 cm thick layer of glass bead tension medium (25 mm mean particle diameter), which in turn overlies a 53 cm diameter ceramic disk (10 m bubbling pressure) sealed within a PVC backing plate. To minimize biological growth and evaporative water loss, the cores are loosely capped with opaque plastic lids and the entire tension plate is covered with an acrylic lid. The trap (small flask) and overflow reservoir (large flask) for the vacuum system (Figure 72.1b) can be seen below the plates. Tygon tubing (1=4-in. i.d. by 1=8-in. wall) connects the trap to the inflow–outflow port of the tension plate via a hole drilled through the bench top.

mean particle diameter for the ‘‘low-tension’’ system; 25 mm mean particle diameter for the ‘‘high-tension’’ system) topped with fine-mesh nylon cloth (15 mm openings) to prevent the contact material from adhering to the samples. To minimize biological growth and evaporative water loss, the samples are loosely capped with opaque plastic lids and the entire tension plate is covered with an acrylic lid (Figure 72.2). A 50 cm diameter tension plate can accommodate up to 20–22 cores of 7.6 cm diameter, and up to 10–11 cores of 10 cm diameter. 3

Matric head control: Control of the matric head (cm ) applied to the water saturating the tension medium=plate and the coincident drainage=uptake of water by the samples is achieved using a hanging water column–constant head burette system (Figure 72.1a), or a regulated vacuum—manometer=transducer system (Figure 72.1b). The constant head burette system is recommended for low-tension tables=plates (1 m cm 0), while the vacuum system is recommended for high-tension tables=plates (5 m cm 1 m). For desorption



986

curve measurements, water removed from the samples drains out the burette outflow or into the overflow flask; and for imbibition curve measurements, water taken up by the samples is extracted from the burette reservoir or trap (storage) flask. Both systems are fitted with shutoff valves so that water flow can be stopped when needed (see Section 72.3.1). The matric head datum can be set at the top, bottom, or mid height position on the soil cores; however, the mid height position is most logical as it locates the average matric head in the core once equilibrium is established. For the constant head burette system, the mid-core datum position is most easily established by simply setting the zero point of the measuring scale (i.e., the top of the tape measure in Figure 72.1a) at the height, L=2, above the tension medium or contact sand surface, where L is the length of the soil core. 4

Appropriately collected and prepared soil cores: Collect soil cores (as recommended in Chapter 69), then trim the soil flush with the bottom end of the sampling cylinder using a sharp, thin-bladed knife or hacksaw blade to prevent smearing and loss of material. If there are gaps or holes at the bottom of the soil sample, fill them with fine-grade sand or glass beads (e.g., 2542 mm mean particle diameter) so that hydraulic contact between the sample and the tension medium or plate is maximized. Place a piece of nylon cloth (53 mm openings) over the bottom end of the core and hold in place with a stout, elastic band. The upper end of the core is covered with a loose-fitting disc to prevent losses or gains of soil and water.

5

Rake (tension table) or putty knife (tension plate): Use a small handheld rake or similar implement suitable for raking and leveling the tension medium surface. Use a wide plastic putty knife or similar implement for scraping contact material off the tension plate. Do not use metal implements to scrape the tension plate, as they may damage the ceramic surface.

6

Balance: A weigh balance with the appropriate range (usually 0–3 kg) and sensitivity (usually 0.1–0.01 g).

7

Controlled temperature: Temperature controlled room (20 C 1 C) for housing the tension tables or plates, sample preparation, and sample weighing.

8

Drying oven: Forced air or convection oven for drying soil cores at 105 C 5 C.

9

Cooling box: Box with water vapor-tight seal suitable for cooling soil cores and other samples from oven temperature to room temperature in the presence of a desiccant.

72.2.2 PROCEDURE Desorption Curve (i)

Saturate the soil cores at room pressure and temperature (20 C 1 C) using deaired, temperature equilibrated water. Convenient procedures for deairing water include the following: (a) boiling or autoclaving in large-vacuum flasks (4 L), then applying airtight seals to the flasks and cooling, (b) filling a vacuum desiccator with water and applying a 65 kPa vacuum for 45 min, and (c) direct application of vacuum (via vacuum pump or tap aspirator) to water in



987

large-vacuum flasks. An advantage of autoclaving is that the water is also sterilized, which helps to reduce biological growth in the samples and in the tension medium or plate. Water used for saturation should have similar major ion speciation and concentrations as the native soil water to prevent aggregate slaking and dispersion=aggregation of silt and clay. Local tap water is often adequate, but this should always be checked. Saturate the cores by placing them in an empty ‘‘wetting tank’’ (Chapter 75), and submerge one-third and subsequent thirds of the core length each 24 h period, so that the ponded water is at the top of the core by the third day. This promotes more complete saturation of the sample (e.g., reduces air entrapment within the sample), and allows time for fine-textured soils to swell completely. Leave the cores in the wetting tank until free water appears on the core surface. Weigh the saturated core by (a) weighing it under water (using a cradle) or (b) quickly removing the core from the wetting tank and placing it in a tared weigh boat so that the weight of rapidly drained macropore water is included in the total core weight. Record the saturated core weight, Mc (c1 ), where c1 ¼ 0 is the first (and largest) cm value on the desorption curve, which yields the saturated water content value, us . (ii)

Place the saturated cores on the presaturated tension table or plate with the constant head burette set to yield cm ¼ 0 m at the tension medium surface (tension table) or contact material surface (tension plate). Close the burette inflow–outflow valve to prevent water flow. Establish good hydraulic connection between the bottom of the cores and the tension medium or contact material by pushing and twisting the cores slightly to deform the material to the shape of the core base. Open the burette valve and allow the cores to equilibrate, then remove and weigh to obtain core weight, Mc (c2 ), where c2 ¼ L=2 is the second cm value on the desorption curve and L is the core length. The time required for equilibration (equilibration time) depends on the matric head (equilibration time increases with decreasing head due to decreasing soil core hydraulic conductivity); the height of the soil core (equilibration time increases roughly as the square of the core length because of increased flow path); the quality of the hydraulic connection between soil core and tension medium or contact material (equilibration time increases as the contact area between core and medium decreases); and the saturated hydraulic conductivity of the tension medium=contact material (equilibration time may be somewhat greater for hightension media relative to low-tension media). Approximate equilibration times for 7.6 cm long soil cores are given in Table 72.1, and 1–2 days are generally added to these times for 10 cm long cores. Note that the times are approximate only, and individual equilibration time tests are recommended for accurate work with any particular soil.

Imbibition Curve (i)

Weigh the unsaturated soil cores to obtain, Mc (ci ), where ci is the initial cm value of the samples. This weight will be used later to determine the initial soil water content of the cores.

(ii)

Presaturate the tension table (or tension plate and contact material) using deaired, temperature-equilibrated water (see desorption curve procedures for details), and



988

set the constant head burette or vacuum to produce the desired minimum (most negative) matric head relative to the chosen datum (e.g., mid-core height), and close the water inflow–outflow valve. Establish contact between the soil cores and the tension table or plate by wetting the bottom of the cores, then firmly placing the cores on the tension medium or contact material with a slight push and twist to establish a good hydraulic connection. Open the inflow– outflow valve and allow the cores to equilibrate (imbibe water), then remove and weigh to obtain core weight, Mc (c1 ), where c1 is the first (and most negative) cm value on the imbibition curve. Equilibration times for imbibition are generally longer than those for desorption because rewet soil hydraulic conductivity is usually less than drainage hydraulic conductivity due to hysteretic and air entrapment effects. Unfortunately, equilibration time guidelines for imbibition curves are not yet established, hence preliminary equilibration tests are required. Desorption and Imbibition Curves (i)

Next matric head (cm ) 1 m (low-tension system): After Mc (c1 ) is measured, close the water inflow–outflow valve. To reestablish hydraulic connection between the cores and the tension table or plate, dampen the tension medium or contact material surface using a spray bottle (desorption curve), or wet the bottom of the cores (imbibition curve), and then return the cores to the same locations used for the previous head using a slight push and twist to ensure good core contact. Set the next desired head (i.e., c2 ) by adjusting the height of the constant head burette, and then open the water inflow– outflow valve. Allow the soil cores to equilibrate, then remove and weigh to obtain, Mc (c2 ). Low-tension tables and plates usually do not accumulate significant exsolved air (within and under the tension medium or plate) for matric heads 1 m. It is always advisable, however, to flush low-tension systems periodically to prevent possible buildup of exsolved air over time. This is most conveniently accomplished by closing the inflow–outflow valve, setting the matric head at 1 m, ponding 3–6 L of deaired temperature-equilibrated water on the surface, opening the inflow–outflow valve to allow the water to drain through, and then resetting the matric head to zero at the surface. At the end of a sequence of low-tension measurements on a batch of soil cores (e.g., cores successively equilibrated to cm ¼ 0, 0:05, 0:1, 0:3, 0:5, 0:75, 1 m), it is advisable to ‘‘purge’’ the low-tension table or plate using the procedures given below for the high-tension system. It is also advisable to replace the contact material at the end of each sequence of low-tension measurements to prevent potential plugging of the ceramic plate by silt and clay; and to replace the tension medium when the tank drainage rate starts to decline, which usually signals incipient plugging of the medium by silt and clay. Idle tanks and plates should always be left with a small amount of ponded water on the surface, the matric head set to near-zero, and the inflow–outflow valve closed.

(ii)

Next matric head (cm ) 20 cm

Water table

Close-up of spiked roller

FIGURE 76.2. Roller-type desmearing–decompaction apparatus for the constant-head well permeameter method. (From Reynolds, W.D. and Elrick, D.E., in J.H. Dane and G.C. Topp (Eds.), Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society of America, Madison, Wisconsin, 2002. With permission.)


1028

Soil Sampling and Methods of Analysis (Figure 76.2) up and down the measurement zone (Figure 76.1) several times to break up and pluck off the smeared=compacted surface (Reynolds and Elrick 1986). Other implements that have been used for removing smeared and compacted surfaces include a stiff cylindrical brush (available from many commercial auger and well=borehole permeameter suppliers), a pick-like ‘‘plucking’’ instrument (Campbell and Fritton 1994; Bagarello 1997), and soil peels made from quick-setting resin (Koppi and Geering 1986) (see Comment 1). If the removal of smeared or compacted surfaces results in an appreciable increase in well radius, this new radius should be measured and used in the calculations (Section 76.2.2).

2

Stand the empty permeameter in the well and attach it to some kind of stabilizing apparatus (here we refer specifically to ‘‘vertical’’ permeameters based on in-hole Mariotte bottle systems as illustrated in Figure 76.1—other systems may not require stabilizing). The stabilizing apparatus should hold the permeameter upright, give the permeameter good stability against wind, and carry the weight of the permeameter (when full of water) so that the water outlet tip (Figure 76.1) does not sink into the base of the well during the measurements. A simple tripod (similar to a surveyor’s transit or camera tripod) that clamps solidly to the permeameter reservoir works very well for this purpose (Figure 76.1). To prevent possible collapse of the well when measuring unstable porous materials, it is advisable to install a well screen, or to backfill around the permeameter to the top of the measurement zone using pea gravel or coarse sand (Figure 76.1). Backfill material (which must have a much greater permeability than the material tested to avoid flow impedance effects) also helps to reduce siltation (see Comment 1), as well as produce faster and more uniform bubbling of the Mariotte bottle when measuring low-permeability materials (i.e., porous materials with low hydraulic conductivity).

3

Close the water outlet of the Mariotte bottle by pushing the air tube down into the outlet tip and then fill the bottle with water (Figure 76.1). Use water at ambient temperature to minimize the accumulation within the reservoir of bubbles of degassed air, which can obscure the reservoir scale. Do not use distilled or deionized water, as this may encourage clay=silt dispersion and subsequent siltation of the well surface during the measurement (see also Chapter 69). In many cases, local tap water can be used, as its major ion speciation and concentrations are often sufficient to prevent clay and silt dispersion. In porous media that are particularly susceptible to siltation (primarily materials with high silt content), it may be necessary to use native water (i.e., water extracted from the porous medium), or water with flocculent added. Fill the permeameter reservoir to the top, leaving no air space. This minimizes overfilling of the well when flow is started (see Comment 2).

4

Lift the air tube out of the outlet tip to establish and maintain the desired depth (head) (H) of water in the well (Figure 76.1). The air tube should be raised slowly to prevent a sudden rush of water against the well surface. A sudden rush of water can erode the well (especially if backfill material has not been used), promote well siltation by stirring silt and clay into suspension, and cause excessive air entrapment within the porous medium. Perforated well liners and screens have also been used to protect the well surface and prevent collapse


Saturated Hydraulic Properties: Well Permeameter

1029

(Bagarello 1997). The desired head (H) is obtained by setting the base of the air tube at the appropriate level, which is usually accomplished using a calibrated height marker and scale (Figure 76.1). The Mariotte bottle is operating properly when air bubbles rise regularly up through the outflow tube and into the reservoir (Figure 76.1). 5

The rate of water flow or discharge (Q) out of the Mariotte bottle and into the porous medium is measured by monitoring the rate of fall, R, of the water level in the reservoir. This can be accomplished using a scale attached to the reservoir (Figure 76.1) and a stopwatch, or an automated pressure transducer–data logger system similar to that described by Ankeny (1992). The rate of fall, R, decreases with increasing time and approaches a constant value (Rs ) as the flow rate becomes quasisteady (Qs ). Quasisteady flow is usually assumed when effectively the same R value (Rs ) is obtained over four or five consecutive R measurements (Section 76.2.2).

6

If the single-head analysis is to be used, proceed to Section 76.2.2. If the twohead analysis is desired, raise the air tube again to obtain steady flow for a second head (H2 ); and if the multiple-head analysis is required, raise the air tube two or more times to obtain steady flows at H2 , H3 , . . ., and so on. For both the two-head and multiple-head analyses, H1 must be ponded first with H1 < H2 < H3 < , and the water level in the well must not be allowed to fall when switching from one head to the next higher head.

76.2.2 ANALYSIS AND EXAMPLE CALCULATIONS The original constant-head well permeameter analysis is based on the approximate Glover relationship (Zangar 1953) Kfs ¼ CG Qs =(2pH 2 )

(76:1)

where Qs [L3 T1 ] is the quasisteady flow rate out of the permeameter and into the porous medium, H [L] is the steady depth (head) of water in the well (set by the height of the air tube), and CG is a dimensionless shape factor given by CG ¼ sinh

1

1=2 H a 2 a þ1 þ a H H

(76:2)

where a [L] is the radius of the well. In Mariotte-type permeameters (such as the one illustrated in Figure 76.1), Qs is conveniently obtained by measuring the quasisteady rate of fall of the water level in the reservoir, Rs [LT1 ], and then multiplying by the reservoir cross-sectional area, A [L2 ] (i.e., Qs ¼ ARs ). Although the Glover analysis (i.e., Equation 76.1 and Equation 76.2) is simple and easy to use, it is seriously limited because only Kfs is calculated; and of the three components of flow out of the well (i.e., pressure, gravity, and capillarity), only the pressure component is taken into account (Reynolds et al. 1985). As a consequence, a*, fm , cf , S, PD, and NP cannot be determined with this analysis, and the Kfs value can be overestimated by an order of magnitude or more in dry, finetextured materials (Philip 1985; Reynolds et al. 1985). The accuracy of the Glover analysis tends to improve with increasing H=a ratio, however; Amoozegar and Wilson (1999) suggest that the systematic overestimate of Kfs by Equation 76.1 and Equation 76.2



1030

can usually be reduced to reasonable levels by maintaining H=a 5, although the original work of Zangar (1953) indicates that H=a 10 is required. Extended and improved well permeameter analyses have been developed, which include single-head, two-head, and multiple-head procedures (Reynolds et al. 1985; Reynolds and Elrick 1986; Elrick et al. 1989). These updated analyses account for all three components of flow out of the well (i.e., pressure, gravity, and capillarity); and as a result, they yield more accurate Kfs values, plus simultaneous estimates of the capillarity parameters. The extended single-head analysis (which is directly comparable to the original single-head Glover analysis) determines Kfs using Kfs ¼

Cw Qs [2pH2 þ Cw pa2 þ (2pH=a*)]

(76:3)

where the a* parameter is visually estimated using the texture–structure categories in Table 76.1 (see Comment 6), and the dimensionless shape parameter, Cw , is determined by using (Zhang et al. 1998) Cw ¼

H=a 2:074 þ 0:093(H=a)

Cw ¼

H=a 1:992 þ 0:091(H=a)

Cw ¼

H=a 2:081 þ 0:121(H=a)

0:754

0:683

0:672

for a* 0:09 cm1

(76:4a)

for a* ¼ 0:04 cm1

(76:4b)

for a* ¼ 0:01 cm1

(76:4c)

Once a* is estimated and Kfs calculated, estimates of fm , S, PD, NP, and cf can be obtained using Equation 69.16 through Equation 69.21, recognizing that for this case (field-saturated flow) K(c0 ) ¼ Kfs , a*(c0 ) ¼ a*, u(c0 ) ¼ ufs , u(ci ) ¼ ui , PD(c0 ) ¼ PD, and NP(c0 ) ¼ NP. TABLE 76.1 Texture–Structure Categories for Visual Estimation of a* Texture–structure category

a* (cm21)

Compacted, structureless, clayey, or silty materials such as landfill caps and liners, lacustrine, or marine sediments, etc. Porous materials that are both fine textured and massive; include unstructured clayey and silty soils, as well as fine structureless sandy materials Most structured and medium textured materials; include structured clayey and loamy soils, as well as unstructured medium sands. This category is generally the most appropriate for agricultural soils Coarse and gravelly sands; may also include some highly structured soils with large and numerous cracks and biopores

0.01 0.04

0.12

0.36

Source: Adapted from Elrick, D.E., Reynolds, W.D., and Tan, K.A., Ground Water Monit. Rev., 9, 184, 1989.



1031

The two-head analysis (two heads ponded successively in the well) and the multiple-head analysis (three or more heads ponded successively in the well) allow simultaneous calculation of both Kfs and a*, i.e., the a* parameter does not have to be estimated. The two-head and multiple-head approaches make use of (Reynolds and Elrick 1986) Cwi Qsi ¼ P1 Hi 2 þ P2 Hi þ P3 ; i ¼ 1, 2, 3, . . . , n; n 2

(76:5a)

which is least squares fitted to Cw Qs versus H data (Cw Qs on Y-axis; H on X-axis), where Qsi is the steady flow rate corresponding to steady ponding head, Hi , the Cwi parameter is the Cw value corresponding to Hi =a (Equation 76.4), and P1 ¼ 2pKfs ;

P2 ¼ 2p

Kfs ; a*

P3 ¼ Y-axis intercept

(76:5b)

For the two-head approach, n ¼ 2 in Equation 76.5a (i.e., H1 , H2 ; Q1 , Q2 ; Cw1 , Cw2 ), and for the multiple-head approach, n 3 in Equation 76.5a (i.e., H1 , H2 , H3 , . . . ; Q1 , Q2 , Q3 , . . . ; Cw1 , Cw2 , Cw3 , . . .). Equation 76.5 can be solved for Kfs and a* using a custom-built computer program (see Reynolds and Elrick 1986, for working equations), or using the regression function of a computer spreadsheet (quadratic curve). The two-head approach can also be solved using simultaneous equations presented in Reynolds et al. (1985) and Reynolds and Elrick (1986). Once Kfs and a* are determined, the fm , S, PD, NP, and cf parameters are calculated as above for the extended single-head analysis. Further detail on the extended single-head, two-head, and multiple-head techniques can be found in Reynolds et al. (1985), Reynolds and Elrick (1986), and Reynolds and Elrick (2002). All three of the extended analyses prevent the systematic overestimation of Kfs that occurs with the Glover analysis (illustrated in Table 76.2). Example data sheets and calculations for the single-head, two-head, and multiple-head analyses are given in Table 76.2 and Table 76.3.

76.2.3 COMMENTS 1

Because water flows out of the well and into the porous medium in the constanthead well permeameter method, any significant smearing, compaction, or siltation of the porous medium in the measurement zone (Figure 76.1) can result in unrepresentative (inaccurate) Kfs , fm , S, a*, cf , PD, or NP values. In wet, structured silty clay soils, for example, the Kfs and fm values can be reduced by more than an order of magnitude if ‘‘normal’’ augering techniques are used, rather than the ‘‘two-finger=two-turn’’ rule. Proper and careful augering of the well is therefore essential in materials susceptible to smearing, compaction, and siltation. Removal of any smeared or compacted areas in the measurement zone is also strongly recommended, although it appears that none of the currently available methods are completely effective in susceptible materials (i.e., silty and clayey soils).

2

If the Mariotte bottle does not respond after the air tube has been raised to the desired H-level (i.e., no bubbling), the well may be overfilled with water (i.e., the water level in the well is above the base of the air tube). To remedy this, withdraw water from the well until bubbling starts, which indicates that the set H-level has been reached. One way to accomplish this is to tape a small-diameter ‘‘extraction



1032

TABLE 76.2 Example Data Sheet and K fs Calculations for the Constant-Head Well Permeameter Method, Single-Head Approach (Equation 76.1 through Equation 76.4) Cumulative time, t (min)

Reservoir scale reading, L (cm)

Rate, R (DL=Dt) (cm=2 min)

0 2 4 6 8 10 12 14 16 18 20 22 24

5.8 8.6 10.6 12.4 14.1 15.7 17.2 18.7 20.1 21.5 22.9 24.3 25.7

— 2.8 2.0 1.8 1.7 1.6 1.5 1.5 1.4 1.4 1.4 1.4 1.4

Well radius, a ¼ 5 cm; well depth, d ¼ 50 cm; depth of ponding, H ¼ 10 cm. Mariotte reservoir cross-sectional area, A ¼ 12:57 cm2 . Porous material type: structureless clay loam soil (estimated a* ¼ 0:04 cm1 from Table 76.1). ufs ¼ 0:65; ui ¼ 0:40; g ¼ 1:818. The R value is the rate of fall of water level in the Mariotte reservoir. Rs ¼ DL=Dt ¼ 1:4 cm=2 min ¼ 0:7 cm= min ¼ 1:1667 102 cm s1 : Qs ¼ ARs ¼ (12:57 cm2 )(1:1667 102 cm s1 ) ¼ 1:4665 101 cm3 s1 : Glover analysis

Extended analysis

CG ¼ 0:8256 (Equation 76.2) Kfs ¼ 1:93 104 cm s1 (Equation 76.1)

Cw ¼ 0:9446 (Equation 76.4b) Kfs ¼ 6:09 105 cm s1 (Equation 76.3)

Note: 1. The Kfs from the Glover analysis overestimates the Kfs from the extended analysis by a factor of 3.2. 2. Any convenient timing interval can be used, and it need not be constant.

tube’’ to the outflow tube (Figure 76.1) before the Mariotte bottle is inserted into the well, and then connect a syringe to the top of the extraction tube to withdraw the excess water from the well. Alternatively, one can simply wait for the water level to fall and the Mariotte bottle will start itself when the H-level is eventually reached. A slowly bubbling Mariotte bottle means that the porous medium has low hydraulic conductivity or the well is smeared=compacted=silted up. In low hydraulic conductivity materials where the Mariotte bottle bubbles slowly, it is advisable to shade the reservoir from direct, hot sun in order to minimize solar heating of the headspace (Figure 76.1) above the water surface. Thermal expansion of the air in the headspace can prevent bubbling. In addition, extreme solar heating of the water in the reservoir will cause a significant reduction in water viscosity, which will introduce calculation errors if not corrected (see Chapter 69 for details). A Mariotte-based permeameter that will not bubble but still registers flow (i.e., dropping water level in the reservoir) may have an air (vacuum) leak in the reservoir.

3.0 4.1 5.0 5.8 6.5 7.3 8.0 8.8 9.5 10.3 11.0 11.8 12.5

0 2 4 6 8 10 12 14 16 18 20 22 24

17.8 19.6 21.3 22.9 24.4 25.9 27.3 28.8 30.2 31.6 33.0 34.4 —


H2 5 9 cm

— 1.8 1.7 1.6 1.5 1.5 1.4 1.5 1.4 1.4 1.4 1.4 —


R2 ¼ 1:4 cm=2 min ¼ 0:01167 cm s1 Qs2 ¼ AR2 ¼ 0:42 cm3 s1 Cw2 ¼ 0:8476 (Equation 76.4a)

26 28 30 32 34 36 38 40 42 44 46 48 —

Cumulative time, t (min)

Well radius, a ¼ 5 cm; well depth, d ¼ 50 cm. Mariotte reservoir cross-sectional area, A ¼ 36:0 cm2 . Porous material type: structured clay loam soil. ufs ¼ 0:65; ui ¼ 0:40; Du ¼ 0:25; g ¼ 1:818. The R value is the rate of fall of water level in the Mariotte reservoir.

— 1.1 0.9 0.8 0.7 0.8 0.7 0.8 0.7 0.8 0.7 0.8 0.7





H1 5 3 cm

41.0 43.9 46.8 49.6 52.2 54.6 56.9 59.1 61.3 63.5 65.7 — —


(continued)

— 2.9 2.9 2.8 2.6 2.4 2.3 2.2 2.2 2.2 2.2 — —

Rate, R (DL=Dtb) (cm=2 min)


50 52 54 56 58 60 62 64 66 68 70 — —


H3 5 15 cm

TABLE 76.3 Example Data Sheet and Calculations for the Constant-Head Well Permeameter Method, Two-Head and Multiple-Head Analyses (Equation 76.5)


Saturated Hydraulic Properties: Well Permeameter 1033

Kfs ¼ 3:06 104 cm s1 Kfs ¼ 3:35 104 cm s1 Kfs ¼ 3:63 104 cm s1

a* ¼ 0:1288 cm1 0:13 cm1 (Chapter 69, Equation 69.16) (Chapter 69, Equation 69.20) (Chapter 69, Equation 69.17) (Chapter 69, Equation 69.18) (Chapter 69, Equation 69.19)

a* ¼ 0:0959 cm1 0:10 cm1 a* ¼ 0:1098 cm1 0:11 cm1 a* ¼ 0:1397 cm1 0:14 cm1

Note: 1. The two-head results can also be used to calculate fm , cf , S, PD, and NP. 2. For the PD and NP calculations: s ¼ 72:75 g s2 ; r ¼ 0:9982 g cm3 (20 C); g ¼ 980:621 cm s2 ; m ¼ 1:002 g cm1 s1 (20 C).

Kfs ¼ 3:54 104 cm s1 fm ¼ 2:75 103 cm2 s1 cf ¼ 7:76 cm S ¼ 3:53 102 cm s1=2 PD ¼ 0:0191 cm NP ¼ 1:11 106 pores m2

Multiple-Head Analysis (H1 , H2 , H3 ) (Equation 76.5)

H1 , H2 H1 , H3 H2 , H3

Two-Head Analysis (Equation 76.5)

TABLE 76.3 (continued) Example Data Sheet and Calculations for the Constant-Head Well Permeameter Method, Two-Head and Multiple-Head Analyses (Equation 76.5)


1034 Soil Sampling and Methods of Analysis



1035

3

The time required for a well permeameter to reach quasisteady flow (equilibration time) is determined primarily by the hydraulic conductivity of the material tested, but also by the antecedent water content of the material, the radius of the well, and the depth of water in the well. Generally speaking, equilibration time increases with decreasing hydraulic conductivity, decreasing antecedent water content, increasing well radius, and increasing depth of water ponding (Reynolds and Elrick 1986). Equilibration times range from about 5 to 60 min in moderate to highly permeable materials (Kfs 104 cm s1 ) to as much as two or more hours in low permeability materials (Kfs 105 cm s1 ). The range of Kfs that can be measured practically with a Mariotte bottle permeameter is on the order of 102106 cm s1 , although this range can be extended somewhat with careful use and adjustments in the size of the reservoir, outflow tube, and air tube.

4

The Cw value (shape factor) relationships given in Equation 76.4 are calibrated for approximately 1 cm a 5 cm, 0:5 cm H 20 cm, and 0.25 cm H=a 20 cm. They are based on discrete data points obtained from numerical solution of Richards’ equation for steady, three-dimensional saturated–unsaturated flow around the well (Reynolds and Elrick 1987). If a, H, or H=a values substantially outside these ranges are required, it is recommended that new Cw values be calculated using the procedures in Reynolds and Elrick (1987). Note that Equation 76.4a applies for all a* 0:09 cm1 (because of the decreasing influence of capillarity with increasing a*), and is thus the appropriate Cw versus H=a relationship for the a* ¼ 0:12 cm1 category and the a* ¼ 0:36 cm1 category in Table 76.1.

5

The primary advantage of the two-head and multiple-head approaches is that simultaneous measurements of the Kfs , fm , S, a*, cf , PD, and NP parameters can be obtained. An important limitation, however, is that heterogeneity in the form of layering, horizonation, cracks, worm holes, root channels, etc. can result in unrealistic and invalid (e.g., negative) parameter values (Elrick et al. 1989). This occurs because both the infiltration surface and the wetted bulb around the well increase with increasing H, which increases the likelihood of encountering heterogeneities. In addition, the coefficient matrices in the two-head and multiple-head analyses are ill-conditioned, which further increases sensitivity to heterogeneity (Philip 1985). When the two-head or multiple-head analysis produces a negative Kfs and fm value, or when the calculated a* value falls substantially outside the physically realistic range of 0:01 cm1 a* 1 cm1 , then the extended single-head analysis (Equation 76.3 and Equation 76.4) should be applied to each head (H value) and the resulting Kfs and capillarity averaged (Elrick and Reynolds 1992). Further discussion on analyzing constanthead well permeameter data can be found in Amoozegar (1993) and Elrick and Reynolds (1993).

6

Relative to the two-head and multiple-head approaches (Equation 76.5), the primary advantages of the updated single-head analysis (Equation 76.3) include time saving and avoidance of negative Kfs values, as a result of using only one ponded head and site estimation of a*, as illustrated in Table 76.1. Disadvantages include lack of simultaneous calculation of fm , S, a*, cf , PD, and NP, and potentially reduced parameter accuracy through inappropriate selection of a* from Table 76.1. Fortunately, the categories are broad enough in Table 76.1


1036

Soil Sampling and Methods of Analysis that one should not be in error by more than one a* category when visually estimating the texture and structure at the measurement site. This in turn introduces an ‘‘error’’ into the Kfs and fm calculations, which is generally less than a factor of 2 and often less than 25% (Reynolds et al. 1992). This is sufficient accuracy for many practical applications, given the inherent variability of these parameters. In addition, the sensitivity of Kfs and fm to the choice of a* can be reduced further by adjusting the H-level. The sensitivity of Kfs to the choice of a* decreases as H increases, while the sensitivity of fm to a* decreases as H decreases (Reynolds et al. 1992). Consequently, if one is interested primarily in Kfs , then the H-level used for the single-head approach should be as large as possible. On the other hand, if interest is primarily in fm or the other capillarity parameters, then the H-level should be as small as possible. It should also be kept in mind, however, that fm , S, a*, cf , PD, and NP can be of low accuracy when obtained using ponded infiltration techniques (regardless of analysis procedure) because of the usual dominance of the pressure and gravity components of flow over the capillarity component of flow in a ponded environment.

7

Alternative constant-head well permeameter designs, which are based on various Mariotte bottle or float valve arrangements, can be found in Amoozegar and Warrick (1986), Koppi and Geering (1986), Stephens et al. (1987), Jenssen (1989), Bell and Schofield (1990), and Amoozegar (1992). Alternative procedures for collecting and analyzing constant-head well permeameter data may be found in Philip (1985), Amoozegar and Warrick (1986), Stephens et al. (1987), Amoozegar (1992), Elrick and Reynolds (1992), Reynolds et al. (1992), and Xiang (1994). Further discussion of the apparatus and analyses presented here may be found in Reynolds and Elrick (1986), Elrick et al. (1989), and Elrick and Reynolds (1992).

76.3 FALLING-HEAD WELL PERMEAMETER The falling-head well permeameter method involves ponding a known head (depth) of water in a tightly cased well, and monitoring the decline in head with time as water flows out through the base of the casing and into the unsaturated porous material (Figure 76.3). The diameter and shape of the well are usually similar to those of the constant-head well permeameter method (i.e., 4–10 cm diameter, flat bottom), although there are no theoretical restrictions on well diameter. As flow is entirely through the bottom of the well, the calculated water transmission parameters (Kfs , fm , S, a*, cf , PD, NP) are primarily relevant to vertical flow, as occurs during infiltration or near-surface drainage. Overall strengths of the falling-head method relative to steady flow methods include greatly reduced measurement times in low-permeability materials, and the ability to measure lower Kfs values than what is practical with the steady flow (e.g., constant head) approaches. Overall weaknesses include the necessity for a watertight seal between the well liner (casing) and the well wall (which can be difficult or impossible in some porous materials), and limited comparison to more established methods.



1037

Soil surface

Solid casing

H0

Ht 2a

FIGURE 76.3. Schematic of the falling-head well permeameter method, cased well. (From Reynolds, W.D. and Elrick, D.E., in J. Alvarez-Benedi and R. MunozCarpena (Eds.), Soil–Water–Solute Process Characterization: An Integrated Approach, CRC Press, Boca Raton, Florida, 2005. With permission.)

76.3.1 APPARATUS AND PROCEDURES 1

Construct a cylindrical, flat-bottomed well using the procedures given for the constant-head well permeameter method (Section 76.2). The bottom of the well should be sufficiently above the water table or capillary fringe to avoid ‘‘groundwater mounding’’ up into the well, which is not accounted for in the analysis. Use the procedures in Section 76.2 and Chapter 78 (auger hole method) to remove smearing or compaction from the well base, as these conditions can result in unrepresentatively low values for Kfs and the capillarity parameters.

2

Slide a solid casing (i.e., solid-walled pipe open only at its ends) down to the bottom of the well (Figure 76.3). The casing must be tight-fitting in the well to prevent water ‘‘leakage’’ between the casing and the well wall (required by theory) (see Comment 1). The casing should extend at least to the surface, but may extend above the surface to accommodate the desired initial water depth (H0 ) at the start of the measurement (Figure 76.3).

3

Quickly add a calibrated volume of water into the casing to produce the desired water level, H ¼ H0 , at time t ¼ 0 (Figure 76.3). This may be accomplished by simply pouring the water into the casing, or alternatively, by adding the water through a tube that extends to the well base. It may be necessary in some porous materials (e.g., silty and clayey soils) to place a screen (2 mm openings) or layer of pea gravel (5 cm thickness) in the bottom of the casing to prevent both



1038

erosion of the well base when water is first added, and subsequent siltation of the well base by the eroded material as water flows out of the casing and into the porous medium (Figure 76.3). The hydraulic conductivity of the screen or pea gravel must be substantially greater than that of the porous medium to prevent flow impedance, which may cause unrepresentative results. After the water is added, measure the decline in water level (H) with time (t) and calculate the various hydraulic parameters as indicated below.

76.3.2 ANALYSIS AND EXAMPLE CALCULATIONS The decline in water level with time is given by (Philip 1993) 3 2 1 A 1 3 A1 ln 3 ln 1þ p2 a 6 A r3t 2A A rt 6 pffiffi2A ffi t¼ 4 A þ 2r Aþ2 3 8Kfs þ arctan pffiffiffi t arctan pffiffiffi A 3A 3A

3 7 7 5

(76:6a)

where 1 p2 a þ 3 H0 þ a* 8 A3 ¼ þ1 a(Du) rt3 ¼

3(H0 Ht ) þ1 a(Du)

(76:6b) (76:6c)

where t [T] is the time since initiation of flow out of the well, a [L] is the inside radius of the casing (Figure 76.3), Du ¼ (ufs ui ) [L3 L3 ] is the difference between the fieldsaturated volumetric water content (ufs ) and the antecedent volumetric water content (ui ) of the porous medium, H0 [L] is the initial water depth in the well (at t ¼ 0), Ht [L] is the depth of water in the well at time, t, and the arctan functions are in radian measure. Equation 76.6 thus describes the time-dependent decline in water level (Ht versus t) as water flows out through the bottom of the cased well and into the porous medium. For simultaneous determination of Kfs and a* using Equation 76.6, a minimum of two H versus t data points are required, which Philip (1993) and Munoz-Carpena et al. (2002) chose as t at H ¼ H0 =2 and t at H ¼ 0. An alternative (and perhaps more robust) approach is to numerically curve-fit Equation 76.6 to a sequence of Ht versus t data points. The ufs , ui , and H0 parameters must be measured independently. Once Kfs and a* are determined, fm , S, PD, NP, and cf are calculated using Equation 69.16 through Equation 69.21, recognizing that for this case (field-saturated flow) K(c0 ) ¼ Kfs , a*(c0 ) ¼ a*, u(c0 ) ¼ ufs , u(ci ) ¼ ui , PD(c0 ) ¼ PD, and NP(c0 ) ¼ NP. If only Kfs is of interest, Equation 76.6a and Equation 76.6c can be simplified to (Elrick and Reynolds 2002) 3 3 1 A 1 3 A1 1þ ln 3 ln 7 p2 a 6 A rb 3 2A A rb 6 pffiffi2A ffi 7 Kfs ¼ 3 A þ 2rb Aþ2 5 8tb 4 arctan pffiffiffi arctan pffiffiffi þ A 3A 3A 2

(76:7a)


Saturated Hydraulic Properties: Well Permeameter rb 3 ¼

3(H0 Hb ) þ1 a(ufs ui )

1039 (76:7b)

where A is defined by Equation 76.6b and Hb [L] is the measured water level in the casing at time, tb [T]. In Equation 76.7a and Equation 76.7b, the initial water level at zero time (H0 ) is preset by adding a calibrated volume of water to the well, and the a* parameter is site-estimated using the texture–structure categories in Table 76.1. The simplified Kfs calculation thus requires measurements of tb , Hb , ufs , and ui , remembering that H0 (i.e., H at t ¼ 0) is already known because a calibrated volume of water was added to the well. If the well empties within a reasonable amount of time (due to favorable well dimensions or adequate permeability of the porous medium), it is feasible to set tb as the time required for the well to empty, in which case Hb ¼ 0 in Equation 76.7b and only tb , ufs , and ui need to be measured. The simplifications also allow Kfs to be obtained directly from Equation 76.7a without resorting to simultaneous equation or numerical curve-fitting methods. Elrick and Reynolds (1986) developed a falling-head analysis for an uncased well, which is derived from the constant-head well permeameter method (Equation 76.3). Although simultaneous determination of Kfs and a* (and thereby fm , S, cf , PD, and NP as well) is possible with this method, numerical curve fitting to a sequence of H versus t data points is required (i.e., no simplified analytical expressions are available), and steady flow at H ¼ H0 must be attained before the falling-head phase is started. This approach is also susceptible to the sensitivity and heterogeneity problems discussed under the two-head and multiple-head approaches for the constant-head method (Section 76.2). Further details on the falling-head well permeameter methods can be found in Elrick and Reynolds (1986, 2002), Munoz-Carpena et al. (2002), and Reynolds and Elrick (2005). Example calculations based on Equation 76.6 and Equation 76.7 are given in Figure 76.4 and Table 76.4, respectively.

76.3.3 COMMENTS 1

It is critically important to obtain a watertight seal between the casing and the well wall, as ‘‘short circuit’’ flow in this area can cause substantial overestimation of Kfs and the capillarity parameters. One way of accomplishing this is to use an auger that fits snugly (12 mm clearance) inside a sharpened, thinwalled metal casing, and advance the slightly oversized casing as the well is being dug (see Chapter 79 and Amoozegar 2002 for details). Other possible approaches include applying grease to the outside of the casing, or attaching an inflatable ‘‘packer’’ to the base of the casing that can then be expanded once the casing is in place. The measurement should be abandoned if leakage of free water occurs between the casing and well wall.

2

As indicated in Equation 76.7b, Hb can be any value less than H0 , including zero (i.e., empty well), although setting Hb ¼ 0 may result in impractically long tb times. Generally speaking, the most practical Hb values are H0 =2. Note also that one can measure the decline in water level using either specified tb and measured Hb or specified Hb and measured tb .



1040 100

Fitted Equation 76.6a Data

Ponded head, H (cm)

90

80

70

60

50 0

2

4

6

8

10

12

14

16

Time, t (h)

FIGURE 76.4. Curve-fit of Equation 76.6a to H versus t for the falling-head well permeameter method. Input parameters: a ¼ 2 cm; H0 ¼ 100 cm; Du ¼ 0:25. Fitted parameters: Kfs ¼ 2:29 105 cm s1 ; a* ¼ 0:12 cm1 . Calculated parameters: fm ¼ 1:91 104 cm2 s1 (Chapter 69, Equation 69.16); cf ¼ 8:3 cm (Chapter 69, Equation 69.20); S ¼ 9:31 103 cm s1=2 (Chapter 69, Equation 69.17); PD ¼ 0:0178 cm (Chapter 69, Equation 69.18); NP ¼ 9:51 104 pores per m2 (Equation 69.19).

3

Equation 76.6 and Equation 76.7 are relatively insensitive to Du, e.g., a change of a factor of 2 in Du often results in less than 20% change in Kfs . It may consequently be feasible in some situations (e.g., when only ‘‘ball park’’ Kfs values are required) to further simplify the method by using a single ‘‘field average’’ Du rather than individual values. The field average Du could be obtained from a series of TDR measurements (Chapter 70), or perhaps even estimated from average texture and antecedent wetness. TABLE 76.4 Example Data Sheet and Kfs Calculation for the Falling-Head Well Permeameter Method, Simplified Analysis (Equation 76.6 and Equation 76.7) Cumulative time, t (h) 0 3 A3 ¼ 665:8044 rb3 ¼ 91:0 Kfs ¼ 2:27 105 cm s1

(t0 ) (tb ) (Equation 76.6b) (Equation 76.7b) (Equation 76.7a)

Water height above well base, H (cm) 100 85

(H0 ) (Hb )

Well radius, a ¼ 2 cm; well depth, d ¼ 50 cm. Initial ponded head, H0 ¼ 100 cm. Water volume required to produce H0 ¼ 1:257 L. Porous material type: structured silt loam soil (estimated a* ¼ 0:12 cm1 from Table 76.1). ufs ¼ 0.65; ui ¼ 0.40; Du ¼ 0.25


Saturated Hydraulic Properties: Well Permeameter 4

1041

The falling-head cased permeameter method often overestimates Kfs and a* relative to other methods (Munoz-Carpena et al. 2002). The reasons for this are unclear at this time, but may be related to simplifying assumptions in the falling-head theory (Philip 1993), to differing flow geometries and sampling volumes among methods, or to experimental problems such as leakage between the casing and well wall.

REFERENCES Amoozegar, A. 1992. Compact constant head permeameter: a convenient device for measuring hydraulic conductivity. In G.C. Topp et al., Eds. Advances in Measurement of Soil Physical Properties: Bringing Theory into Practice. SSSA Special Publication, 30. Soil Science Society of America, Madison, WI, pp. 31–42. Amoozegar, A. 1993. Comments on ‘‘Methods for analyzing constant-head well permeameter data.’’ Soil Sci. Soc. Am. J. 57: 559–560. Amoozegar, A. 2002. Piezometer method (saturated zone). In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 859–878. Amoozegar, A. and Warrick, A.W. 1986. Hydraulic conductivity of saturated soils: field methods. In A. Klute, Ed. Methods of Soil Analysis, Part 1— Physical and Mineralogical Methods. Agronomy 9, Soil Science Society of America, Madison, WI, pp. 758–763. Amoozegar, A. and Wilson, G.V. 1999. Methods for measuring hydraulic conductivity and drainable porosity. In R.W. Skaggs and J. van Schilfgaarde, Eds. Agricultural Drainage. Agron. Monogr. 38. American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, Madison, WI, pp. 1149–1205. Ankeny, M.D. 1992. Methods and theory for unconfined infiltration measurements. In G.C. Topp et al., Eds. Advances in Measurement of Soil Physical Properties: Bringing Theory into Practice. Soil Science Society of America. Special Publication, 30. Soil Science Society of America, Madison, WI, pp. 123–141. Bagarello, V. 1997. Influence of well preparation on field-saturated hydraulic conductivity measured with the Guelph permeameter. Geoderma 80: 169–180.

Bell, R.W. and Schofield, N.J. 1990. Design and application of a constant head well permeameter for shallow high saturated hydraulic conductivity soils. Hydrol. Process. 4: 327–342. Campbell, C.M. and Fritton, D.D. 1994. Factors affecting field-saturated hydraulic conductivity measured by the borehole permeameter technique. Soil Sci. Soc. Am. J. 58: 1354–1357. Elrick, D.E. and Reynolds, W.D. 1986. An analysis of the percolation test based on threedimensional, saturated–unsaturated flow from a cylindrical test hole. Soil Sci. 142: 308–321. Elrick, D.E. and Reynolds, W.D. 1992. Methods for analyzing constant-head well permeameter data. Soil Sci. Soc. Am. J. 56: 320–323. Elrick, D.E. and Reynolds, W.D. 1993. Reply to ‘‘Comments on ‘Methods for analyzing constanthead well permeameter data’ ’’. Soil Sci. Soc. Am. J. 57: 560–563. Elrick, D.E. and Reynolds, W.D. 2002. Measuring water transmission parameters in vadose zone using ponded infiltration techniques. Agri. Sci. 7: 17–22. Elrick, D.E., Reynolds, W.D., and Tan, K.A. 1989. Hydraulic conductivity measurements in the unsaturated zone using improved well analyses. Ground Water Monit. Rev. 9: 184–193. Jenssen, P.D. 1989. The constant head percolation test—improved equipment and possibilities of assessing the saturated hydraulic conductivity. In H.J. Morel-Seytoux, Ed. Unsaturated Flow in Hydrologic Modeling: Theory and Practice. NATO ASI Series C: Mathematical and Physical Sciences, Vol. 275, Kluwer Academic Publishers, the Netherlands, pp. 481–488. Koppi, A.J. and Geering, H.R. 1986. The preparation of unsmeared soil surfaces and an improved


1042


apparatus for infiltration measurements. J. Soil Sci. 37: 177–181.

Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 844–858.

Munoz-Carpena, R., Regalado, C.M., AlvarezBenedi, J., and Bartoli, F. 2002. Field evaluation of the new Philip-Dunne Permeameter for measuring saturated hydraulic conductivity. Soil Sci. 167: 9–24.

Reynolds, W.D. and Elrick, D.E. 2005. Measurement and characterization of soil hydraulic properties. In J. Alvarez-Benedi and R. Munoz-Carpena, Eds. Soil–Water–Solute Process Characterization: An Integrated Approach. CRC Press, Boca Raton, FL, pp. 197–252.

Philip, J.R. 1985. Approximate analysis of the borehole permeameter in unsaturated soil. Water Resour. Res. 21: 1025–1033. Philip, J.R. 1993. Approximate analysis of falling-head lined borehole permeameter. Water Resour. Res. 29: 3763–3768. Reynolds, W.D. 1993. Saturated hydraulic conductivity: field measurement. In M.R. Carter, Ed. Soil Sampling and Methods of Analysis. Canadian Society of Soil Science, Lewis Publishers, Boca Raton, FL, pp. 599–613. Reynolds, W.D. and Elrick, D.E. 1986. A method for simultaneous in situ measurement in the vadose zone of field-saturated hydraulic conductivity, sorptivity and the conductivity–pressure head relationship. Ground Water Monit. Rev. 6: 84–95. Reynolds, W.D. and Elrick, D.E. 1987. A laboratory and numerical assessment of the Guelph permeameter method. Soil Sci. 144: 282–299. Reynolds, W.D. and Elrick, D.E. 2002. Constant head well permeameters (vadose zone). In J.H. Dane and G.C. Topp, Eds. Methods of Soil

Reynolds, W.D., Elrick, D.E., and Clothier, B.E. 1985. The constant head well permeameter: effect of unsaturated flow. Soil Sci. 139: 172–180. Reynolds, W.D., Vieira, S.R., and Topp, G.C. 1992. An assessment of the single-head analysis for the constant head well permeameter. Can. J. Soil Sci. 72: 489–501. Stephens, D.B., Lambert, K., and Watson, D. 1987. Regression models for hydraulic conductivity and field test of the borehole permeameter. Water Resour. Res. 23: 2207–2214. Xiang, J. 1994. Improvements in evaluating constant-head permeameter test data. J. Hydrol. 162: 77–97. Zangar, C.N. 1953. Theory and problems of water percolation. U.S. Department of the Interior, Bureau of Reclamation, Eng. Monogr. No. 8, Denver, CO. Zhang, Z.F., Groenevelt, P.H., and Parkin, G.W. 1998. The well shape-factor for the measurement of soil hydraulic properties using the Guelph permeameter. Soil Till. Res. 49: 219–221.


Chapter 77 Saturated Hydraulic Properties: Ring Infiltrometer W.D. Reynolds Agriculture and Agri-Food Canada Harrow, Ontario, Canada

77.1 INTRODUCTION The ring infiltrometer methods are used primarily for in situ or field measurement of fieldsaturated hydraulic conductivity, Kfs [LT1 ], in unsaturated porous materials (e.g., soil). They can also be used, however, for in situ determination of the capillarity parameters: matric flux potential, fm [L2 T1 ], sorptive number, a* [L1 ], the effective Green–Ampt wetting front pressure head, cf [L], sorptivity, S [LT1=2 ], flow-weighted mean pore diameter, PD [L], and the number of PD pores per unit area, NP [L2 ]. A discussion of the principles associated with determination of Kfs and the capillarity parameters appears in Chapter 69. Ring infiltrometers are thin-walled, open-ended metal or plastic cylinders with the bottomend sharpened to ease insertion into the porous medium. Most ring infiltrometers are 5– 20 cm long by 10–50 cm in diameter, although much smaller and much larger ring diameters have been used for special-purpose applications (e.g., Youngs et al. 1996; Leeds-Harrison and Youngs 1997). Ring infiltrometers are operated by inserting one or more rings into the soil (usually to a depth of 3–10 cm), ponding one or more known heads of water inside the rings, and measuring the rate of water flow out of the rings and into the unsaturated porous medium. Both constant head and falling head analyses are available. Other field methods for measuring Kfs and the associated capillarity parameters include the well permeameter (Chapter 76), auger hole (Chapter 78), and piezometer (Chapter 79). Laboratory methods for measuring saturated hydraulic conductivity, Ks , are described in Chapter 75. Selected methods for estimating Ks from surrogate porous medium properties are given in Chapter 84. The overall strengths of ring infiltrometer methods include: (i) accurate measures of vertical Kfs ; (ii) simple and robust equipment and procedures; (iii) relatively easy and rapid spatial=temporal replication of measurements; (iv) ability to measure water transmission parameters at the porous medium surface; and (v) widespread acceptance by the science and engineering communities because of long-term usage in a vast range of porous materials. The general weaknesses of ring methods include: (i) difficult use in stony porous media

1043



1044

(rings difficult to insert); (ii) potential disturbance=alteration of the measured porous medium volume during the ring insertion process; (iii) inconvenience for subsurface measurements (relatively large access pits have to be dug); and (iv) capillarity parameter determinations (i.e., fm , S, a*, cf , PD, NP) of potentially reduced accuracy because ponded infiltration maximizes the hydrostatic pressure and gravity components of flow at the expense of the capillarity component of flow.

77.2 SINGLE RING INFILTROMETER The single ring infiltrometer involves inserting a solitary cylinder into unsaturated porous medium, ponding one or more heads of water in the cylinder, and measuring the rate of water flow out of the cylinder and into the porous medium. Data analysis options include the single constant head approach, the multiple constant head approach, and the falling head approach. The single constant head approach involves ponding one constant head of water in the cylinder, and measuring the quasisteady rate of water flow out of the cylinder (Figure 77.1a).

Mariotte reservoir Measuring cylinder H d

a

Water diffuser

Wetting front Single ring infiltrometer (cross section) (a)

Buffer cylinder

Measuring cylinder

H a

d

Wetting front Double or concentric ring infiltrometer (cross section) (b)

FIGURE 77.1. A single ring constant head infiltrometer (a) and double=concentric ring constant head infiltrometer (b). (From Reynolds, W.D., Elrick, D.E., and Youngs, E.G., in J.H. Dane and G.C. Topp (Eds.), Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society of America, Madison, Wisconsin, 2002. With permission.)


Saturated Hydraulic Properties: Ring Infiltrometer

1045

∆ Hi

∆ H2

∆ H1

Soil surface q fs

K fs

∆ ∆

∆ Field-saturated bulb

Wetting front

i = 1, 2, 3, …, n ; n ≥ 2

FIGURE 77.2. The single ring, multiple constant head infiltrometer. (From Reynolds, W.D. and Elrick, D.E., in J. Alvarez-Benedi and R. Munoz-Carpena (Eds.), Soil–Water–Solute Process Characterization: An Integrated Approach, CRC Press, Boca Raton, Florida, 2005. With permission.)

The multiple constant head approach extends the single head analysis in that it ponds two or more constant heads in succession, and measures the corresponding quasisteady flow rates for each head (Figure 77.2). The falling head approach, on the other hand, monitors the fall of water level with time from an initial ponded head (Figure 77.3), (see also Comment 4 in Section 77.2.3).

Standpipe reservoir H0

2rs

Ht

Soil surface qfs

Wetting front Kfs

Ring qi

Ki 2rr

FIGURE 77.3. The single ring, falling head infiltrometer with attached standpipe reservoir. (From Reynolds, W.D. and Elrick, D.E., in J. Alvarez-Benedi and R. MunozCarpena (Eds.), Soil–Water–Solute Process Characterization: An Integrated Approach, CRC Press, Boca Raton, Florida, 2005. With permission.)



1046

77.2.1 APPARATUS AND PROCEDURES 1

The single ring infiltrometer is typically 10–20 cm long by 10–50 cm in diameter, although diameters as large as 100 cm are used occasionally (see Comment 1 in Section 77.2.3). The cylinder should be sturdy (e.g., made of metal or high-density plastic), but thin-walled (e.g., 1–5 mm wall thickness) with a sharp outside beveled cutting edge at the base to minimize resistance and porous medium compaction=shattering during cylinder insertion. Using an appropriate insertion technique (e.g., drop-hammer apparatus, hydraulic ram, etc.), insert the cylinder into the porous medium to a depth of 3–10 cm. The cylinder should be held as straight (vertical) as possible during the insertion process to ensure one-dimensional vertical flow through the porous medium. To allow the desired water ponding heights, H, the top of the inserted cylinder should either extend to at least the maximum H-level above the porous medium surface (Figure 77.1 and Figure 77.2), or the cylinder should be coupled to a standpipe attachment (Figure 77.3) or a Mariotte bottle attachment (Figure 77.4). Scraping, leveling, or similar disturbance

Air tube Water supply tube

Tripod

Reservoir

Standpipe ≈60 cm

H

Wetting front

Wing bolt (4)

Metal support strut (3)

H

Water outlet port

O-ring

Transparent polycarbonate end cap O-ring

Splash plate 30° Angle

Infiltration surface

Soil surface

Stainless-steel ring

FIGURE 77.4. The single ring ‘‘pressure infiltrometer’’ apparatus, which uses a combined standpipe—Mariotte bottle system to allow single constant head, multiple constant head or falling head analyses. (From Reynolds, W.D., in M.R. Carter (Ed.), Soil Sampling and Methods of Analysis, Canadian Society of Soil Science, Lewis Publishers, Boca Raton, Florida, 1993. With permission.)



1047

of the infiltration surface is not recommended, as this may alter the porous medium’s hydraulic properties and thereby produce unrepresentative results. 2

Prevent short circuit flow or leakage around the cylinder wall by lightly tamping the contact between the porous medium and the inside surface of the cylinder. Larger gaps between the porous medium and the cylinder wall should be backfilled with powdered bentonite or fine clay.

3

For a single constant head approach, pond a constant head (depth) of water inside the cylinder (H), and monitor the rate at which water infiltrates the porous medium until the infiltration rate becomes steady (Figure 77.1a). The depth of water ponding is usually in the order of 5–20 cm. For the multiple constant head approach, repeat the single head process for each head (H1 , H2 , H3 , . . .), starting with the smallest head (i.e., H1 < H2 < H3 . . .) (Figure 77.2). A Mariotte reservoir provides a simple and convenient method for simultaneously maintaining a constant head and measuring the infiltration rate (Figure 77.1a through Figure 77.4); i.e., the height of the Mariotte bubble tube sets the depth of ponding, and the rate of fall of the water level in the Mariotte reservoir can be used to calculate the steady infiltration rate, qs [LT1 ], or discharge rate, Qs [L3 T1 ]. Alternative constant head approaches include the use of a float valve arrangement connected via flexible tubing to a gravity-feed reservoir (often useful for high infiltration rates), and simple manual addition of water (often useful for low infiltration rates). In the manual approach, some kind of pointer or ‘‘hook gauge’’ is positioned above the infiltration surface, and when the water surface in the cylinder drops to the pointer=hook gauge level, water is manually added to bring the water surface back up to a preset mark on the cylinder wall. Average infiltration rate in the manual approach is determined using the volume of water added, and the time interval between additions. The depth of water ponding is estimated as the mid-way elevation between the mark on the cylinder wall and the height of the pointer or hook gauge. In the falling head approach, water is quickly added (within a few seconds) to the cylinder to attain the desired initial head, H ¼ H0 at t ¼ 0, and then the fall of water level with time (H vs. t) is measured (Figure 77.3).

4

Constant head infiltration through a cylinder into unsaturated porous material normally decreases through an early-time transient and becomes quasisteady within finite time. The time required to reach quasisteady flow (equilibration time), generally increases with finer porous medium texture, decreasing porous medium structure, increasing depth of water ponding (H), increasing depth of cylinder insertion (d ), and increasing cylinder radius (a). For the single constant head approach (or the initial head of the multiple constant head approach), equilibration times can be as short as 10–60 min for relatively small cylinders (e.g., 5–10 cm diameter) and=or materials that are coarse textured or well structured (Scotter et al. 1982), to as long as several hours or days for large cylinders (e.g., 30–60 cm diameter and larger) and=or materials, which are moderate to fine textured and unstructured (Scotter et al. 1982; Daniel 1989). Generally speaking, the equilibration times for the succeeding constant heads in the multiple head approach (i.e., H2 , H3 , etc.) are substantially less than that for the initial head (H1 ). The falling head approach obviously does not have an equilibration time, as infiltration rate is entirely transient due to the falling head.



1048

77.2.2 ANALYSIS AND EXAMPLE CALCULATIONS Single Constant Head Calculate Kfs using (Reynolds and Elrick 1990) Kfs ¼

qs [H=(C1 d þ C2 a)] þ f1=[a*(C1 d þ C2 a)]g þ 1

(77:1)

where qs ¼ Qs =pa2 [LT1 ] is the quasisteady infiltration rate out of the cylinder, Qs [L3 T1 ] is the corresponding quasisteady flow rate, a [L] is the inside radius of the cylinder, H [L] is the steady depth (head) of ponded water in the cylinder, d [L] is the depth of cylinder insertion into the porous medium, a* [L1 ] is the sorptive number of the porous medium (estimated from Chapter 76, Table 76.1 or measured independently—see Comment 2 in Section 77.2.3), and C1 ¼ 0:316p and C2 ¼ 0:184p are dimensionless quasiempirical constants that apply for d 3 cm and H 5 cm (Reynolds and Elrick 1990; Youngs et al. 1993). Once a* is estimated and Kfs calculated, estimates of fm , S, PD, NP, and cf can be obtained using Equation 69.16 through Equation 69.21, recognizing that for this case (fieldsaturated flow) K(c0 ) ¼ Kfs , a*(c0 ) ¼ a*, u(c0 ) ¼ ufs , u(ci ) ¼ ui , PD(c0 ) ¼ PD, and NP(c0 ) ¼ NP. Note in Equation 77.1 that the magnitude of Kfs depends not only on flow rate, but also on depth of ponding (H), cylinder radius (a), depth of cylinder insertion (d), and porous medium capillarity (a*). As a result, the traditional constant head ring infiltrometer analysis, Kfs ¼ qs

(77:2)

overestimates Kfs to varying degrees, depending on the magnitudes of H, a, d, and a* (Reynolds et al. 2002). The accuracy of Equation 77.2 improves as H decreases, and as a, d, and a* increase, although d and a must generally be impractically large before Equation 77.2 is sufficiently accurate (Table 77.1). An example data sheet and calculations based on Equation 77.1 and Equation 77.2 are given in Table 77.1.

TABLE 77.1 Example Data Sheet and Kfs Calculation for the Single Ring Infiltrometer Method, Single Constant Head (Equation 77.1 and Equation 77.2) Cylinder radius, a ¼ 15 cm Depth of cylinder insertion, d ¼ 5 cm Depth of ponding, H ¼ 10 cm Porous material type: Unstructured clay loam soil (estimated a* ¼ 0:04 cm1 —Table 76.1, Chapter 76) Quasiempirical constant, C1 ¼ 0:316p ¼ 0:9927 Quasiempirical constant, C2 ¼ 0:184p ¼ 0:5781 Measured quasisteady infiltration rate, qs ¼ Qs =pa2 ¼ 8:56 105 cm s1 Extended analysis (Equation 77.1) Kfs ¼ 2:40 105 cm s1

Traditional analysis (Equation 77.2) Kfs ¼ 8:56 105 cm s1

Note: The traditional analysis overestimates the extended analysis by a factor of 3.6. The degree of overestimation by the traditional analysis is systematic and depends on the values of a*, H, d, and a.



1049

Multiple Constant Head Ponding two or more constant heads allows simultaneous calculation of both Kfs and a*, i.e., the a* parameter does not have to be estimated as in the single constant head analysis (see analysis above). If two heads are ponded, Kfs and a* can be determined using the two simultaneous equations (Reynolds and Elrick 2005) T(q2 q1 ) (H2 H1 )

(77:3a)

(q2 q1 ) [q1 (H2 þ T) q2 (H1 þ T)]

(77:3b)

Kfs ¼ a* ¼ where

T ¼ C1 d þ C2 a

(77:3c)

and q1 is qs at H1 ; q2 is qs at H2 ; q1 < q2 ; H1 < H2 ; and the other parameters are as defined in Equation 77.1. For two or more ponded heads, qs is linearly related to H via the relationship (Reynolds and Elrick 2005) qi ¼

Kfs 1 Hi þ Kfs þ1 ; T a*T

i ¼ 2, 3, 4, . . . , n; n 2

(77:4a)

where T is given by Equation 77.3c. Least squares fitting procedures can consequently be used to obtain Kfs from the regression slope, Kfs ¼ T slope

(77:4b)

a* ¼ Kfs =[T(intercept Kfs )]

(77:4c)

and a* from the regression intercept,

of the qi vs. Hi data points (qi on Y-axis; Hi on X-axis). Once Kfs and a* are determined, the capillarity parameters, fm , S, PD, NP, and cf are calculated using Equation 69.16 through Equation 69.21, recognizing that for this case (field-saturated flow) K(c0 ) ¼ Kfs , a*(c0 ) ¼ a*, u(c0 ) ¼ ufs , u(ci ) ¼ ui , PD(c0 ) ¼ PD, and NP(c0 ) ¼ NP. Note that Equation 77.3 and Equation 77.4 yield identical results for two ponded heads because linear regression is equivalent to simultaneous equations when only two qs vs. H data points are used. Example data sheets and calculations based on Equation 77.3 and Equation 77.4 are given in Table 77.2. Falling Head Falling head infiltration through a solitary cylinder into unsaturated porous material can be described by (Elrick et al. 2002)



1050

TABLE 77.2 Example Data Sheet and Calculation of Kfs and the Capillarity Parameters for the Single Ring Infiltrometer Method, Multiple Constant Heads (Equation 77.3 and Equation 77.4) Cylinder radius, a ¼ 15 cm Depth of cylinder insertion, d ¼ 5 cm Steady depth of ponding, H: H1 ¼ 5 cm, H2 ¼ 10 cm, H3 ¼ 20 cm Porous material type: Structured loam soil Quasiempirical constant, C1 ¼ 0:316p ¼ 0:9927 Quasiempirical constant, C2 ¼ 0:184p ¼ 0:5781 T ¼ C1 d þ C2 a ¼ 13:635 ufs ¼ 0:65, ui ¼ 0:40 Measured quasisteady infiltration rates, qs : q1 ¼ 6:53 104 cm s1 , q2 ¼ 7:74 104 cm s1 , q3 ¼ 10:16 104 cm s1 Two-head analysis (Equation 77.3)

Regression analysis (Equation 77.4)

Using (H1 , q1 ) and (H3 , q3 )

Regress q1 , q2 , q3 (Y-axis) against H1 , H2 , H3 (X-axis) Regression slope ¼ 2:42 105 s1 Regression intercept ¼ 5:32 104 cm s1 Kfs ¼ 3:3 104 cm s1 (Equation 77.4b) a* ¼ 0:12 cm1 (Equation 77.4c) (Chapter 69, Equation 69.16) (Chapter 69, Equation 69.20) (Chapter 69, Equation 69.17) (Chapter 69, Equation 69.18) (Chapter 69, Equation 69.19)

Kfs ¼ 3:3 104 cm s1 (Equation 77.3a) a* ¼ 0:12 cm1 (Equation 77.3b)

fm ¼ Kfs =a* ¼ 2:75 103 cm2 s1 cf ¼ 1=a* ¼ 8:3 cm S ¼ [g(ufs ui )fm ]1=2 ¼ 3:54 102 cm s1=2 PD ¼ 0:0178 cm NP ¼ 1:37 106 pores m2

For the S calculation, g ¼ 1:818 was assumed. For the PD and NP calculations, s ¼ 72:75 g s2 ; r ¼ 0:9982 g cm3 (20 C); g ¼ 980:621 cm s2 ; m ¼ 1:002 g cm1 s1 (20 C). Note: In this example, the two-head analysis produces the same result regardless of which combination of two (H, q) data pairs is used. The regression analysis can also be used when only two (H, q) data pairs are available.

2

1 H0 þ Du 6 R(H H ) a* 0 t 6 t¼ CKfs 4 Du C

0

13

B 7 CR(H0 Ht ) C 7; C 6¼ 0 C lnB @1 þ A 5 1 Du H0 þ a*

(77:5)

where t [T] is the time since initiation of ponded infiltration, Du ¼ (ufs ui ) [L3 L3 ] is the difference between the field-saturated volumetric water content (ufs ) and the antecedent volumetric water content (ui ) of the porous medium, C ¼ 1 (Du=R) ¼ constant, H0 is the depth of ponding at t ¼ 0, Ht [L] is the depth of ponding at time, t, and R ¼ As =Ac is the cross-sectional area of the standpipe reservoir (As ¼ prs2 ) divided by the crosssectional area of the cylinder (Ac ¼ prc2 ) (Figure 77.3). Note that Equation 77.5 is both implicit and nonlinear in Ht , and, as a result, simultaneous determination of Kfs and a* is best achieved by curve-fitting the equation to a sequence Ht vs. t data points using numerical optimization procedures (e.g., Figure 77.5). The Du, R, and H0 parameters in Equation 77.5 must be measured independently, although R ¼ 1 if water is ponded directly in the cylinder without using a standpipe reservoir. Once Kfs and a* are obtained, fm , S, PD, NP, and cf are determined using Equation 69.16 through Equation 69.21 (see Figure 77.5).



1051

0.52 Fitted Equation 77.5 Data

Water level in standpipe, H (m)

0.50 0.48 0.46 0.44 0.42 0.40 0.38 0.0

0.5

1.0

1.5

2.0

2.5

Time, t (h)

FIGURE 77.5. Curve-fit of Equation 77.5 to H vs. t data for the falling head ring infiltrometer method. Input parameters Du ¼ 0:50; R ¼ 0:0278; C ¼ 16:9856; H0 ¼ 0:50 m. Fitted parameters: Kfs ¼ 1:41 107 cm s1 ; a* ¼ 0:04 cm1 . Calculated parameters: fm ¼ 3:53 106 cm2 s1 (Chapter 69, Equation 69.16); cf ¼ 25 cm (Chapter 69, Equation 69.20); S ¼ 1:79 103 cm s1=2 (Chapter 69, Equation 69.17); PD ¼ 0:0059 cm (Chapter 69, Equation 69.18); NP ¼ 4:85 104 pores m2 (Chapter 69, Equation 69.19).

If Kfs is the primary interest, Equation 77.5 can be rewritten in the form (Bagarello et al. 2003) 0 13 1 H0 þ B 7 Du 6 CR(H0 Ha ) C a* 6R(H0 Ha ) 7; C lnB 1þ Kfs ¼ 4 @ A 5 1 Cta Du C Du H0 þ a* 2

C 6¼ 0

(77:6)

where Ha [L] is the measured standpipe water level at time, ta [T]. The initial standpipe water level at zero time (i.e., H0 ) is preset by adding a calibrated volume of water to the reservoir, and the a* parameter is estimated using the texture–structure categories in Table 76.1. These simplifications reduce the number of required measurements to only three (i.e., Ha , ta , and Du), and also allow Kfs to be obtained directly from Equation 77.6 without resorting to numerical optimization procedures. Even further simplification can be achieved when it is feasible to allow the standpipe water level to fall all the way to the porous medium surface to produce Ha ¼ 0 at t ¼ ta . Equation 77.6 can then be solved for Kfs using only two measurements (ta and Du) because Ha is now 0. An example data sheet and calculation based on Equation 77.6 is given in Table 77.3.

77.2.3 COMMENTS 1

Bouma (1985) suggests that the volume encompassed by a ring infiltrometer (i.e., infiltration surface area multiplied by depth of ring insertion) must include at least 20 structural units (e.g., soil peds delineated by a polygonal cracking pattern,



1052

TABLE 77.3 Example Data Sheet and K fs Calculation for the Single Ring Infiltrometer Method, Simplified Falling Head Analysis (Equation 77.6) Cylinder radius, rc ¼ 15 cm Depth of cylinder insertion, d ¼ 10 cm Standpipe (reservoir) radius, rs ¼ 2:5 cm Initial ponded head, H0 ¼ 50 cm Porous material type: Unstructured clay loam soil (estimated a* ¼ 0:04 cm1 —Table 76.1, Chapter 76) u ¼ 0:70; ui ¼ 0:20 Cumulative time, t (s) 0 8000 R ¼ As =Ac ¼ prs2 =prc2 ¼ 0:0278 Du ¼ ufs ui ¼ 0:50 C ¼ 1 (Du=R) ¼ 16:9856 Kfs ¼ 1:41 107 cm s1 (Equation 77.6)

Water level in standpipe, H (cm) (t0 ) (ta )

50 40

(H0 ) (Ha )

large aggregates, worm holes, abandoned root channels, etc.) before representative measures of water transmission parameters are obtained. Limited field data in Youngs (1987), Lauren et al. (1988), Richards (1987), and Bouma (1985) suggest that cylinder diameters need to be 5---10 cm for single-grain sandy material and uniform structureless materials (e.g., compacted landfill liners); 30 cm for stony=heterogeneous sands, structured sandy loams, and structured silty loams; and 50 cm for structured clays and clay loams. 2

Estimation (rather than measurement) of a* using the texture–structure categories in Table 76.1 (Chapter 76) may introduce some error into the single constant head and simplified falling head calculations (Equation 77.1 and Equation 77.6), given that a* varies continuously with porous medium capillarity. The error (or potential for error) can be decreased, however, by reducing the importance of the a* terms in Equation 77.1 and Equation 77.6 relative to the other terms. For the single constant head approach (Equation 77.1), the importance of the a* term is reduced by making the cylinder radius (a), and the water ponding depth (H) as large as practicable. For the simplified falling head approach (Equation 77.6), the importance of the a* terms is reduced by making both H0 and Ha as large as practicable. The maximum potential error introduced in Kfs by estimating a* via Table 76.1 can be determined by recalculating Equation 77.1 or Equation 77.6 using the a* category immediately above and immediately below the selected category. The categories in Table 76.1 are broad enough that visual estimation of the appropriate category will not likely be in error by more than 1 category. If the estimated maximum potential error in Kfs (and corresponding capillarity parameters) is unacceptably large, then the importance of the a* terms should be further reduced, or an alternative analysis used, such as the multiple head approach (Equation 77.3 or Equation 77.4), or the complete falling head approach (Equation 77.5). Incorrect estimation of a* by 1 texture–structure category in Table 76.1 generally introduces an error of 50% and capillarity parameter calculations, which is usually of no great concern given the high natural variability of these parameters.

3

The falling head analysis (Equation 77.5 and Equation 77.6) is based on the Green–Ampt model for transient ponded infiltration, which assumes



1053

one-dimensional vertical flow with a step-function wetting front that is horizontal, stable, and downward migrating (Guyonnet et al. 2000; Reynolds and Elrick 2005). The falling head analysis consequently requires that the wetting front remain distinct, horizontal, and within the cylinder during the falling head measurements (i.e., no divergent flow out through the bottom of the cylinder). As a result, the falling head method is often not practical in coarse textured and highly structured materials (e.g., coarse and medium single-grain sands; loams and clays with many shrinkage cracks and biopores), as the wetting front either reaches the base of the cylinder too quickly to allow sufficient H vs. t measurements, or the wetting front is unstable and discontinuous (due to preferential or finger flow). An approximation of the maximum time available for H vs. t measurements can be obtained from 2

Du 4 d td ¼ CKfs

H0 þ a1* C

0

Cd

13

A5; ln@1 þ H0 þ a1*

H0 >

Du d R

(77:7)

where td [T] is the time required for the Green-Ampt wetting front to migrate from the infiltration surface to the base of the cylinder, d [L] is the depth of cylinder insertion, and the other parameters are as previously defined. The constraint on H0 in Equation (77.7) ensures that the standpipe does not run dry before d is reached. Equation (77.7) is applied by specifying R, H0 and d, and then calculating td for likely values (or likely ranges) of Du, C, Kfs and a*. 4

The so-called ‘‘pressure infiltrometer’’ is a convenient and commercially available apparatus (Soilmoisture Equipment Corp., Goleta, CA) that is capable of applying the single constant head, multiple constant head, and falling head approaches to single ring infiltrometer measurements. The apparatus consists essentially of a Mariotte-based dual reservoir system (one small inner reservoir for slow flows located inside a larger outer reservoir for rapid flows), and an end cap that clamps onto the infiltrometer cylinder (Figure 77.4). During operation, water flows out of the reservoir, through the end cap, and into the cylinder. The Mariotte system can be engaged (Mariotte bottle mode) to apply the single constant head or multiple constant head approaches, or it can be disengaged (standpipe mode) to apply the falling head approach. Infiltration rates are determined by measuring the rate of fall of the water level in the inner or outer reservoir. Detailed discussions of the apparatus and its application appear in Reynolds (1993) and Reynolds et al. (2002), where the equations provided are different forms of those given here.

5

Strengths of the steady (constant head) analyses include reasonably accurate and robust determination of vertical Kfs , extensive field testing, relatively simple measurements (qs , H, d, a, a*), and relatively large sample volume (often comparable to ring volume). Weaknesses of the steady analyses include potentially long equilibration times and extensive water consumption for large rings or highly structured=permeable porous materials. Exacerbating these weaknesses is the fact that ring diameters may need to be 10 cm in single-grain sands and uniform structureless materials, 30 cm in heterogeneous sands and structured loams, and 50 cm in structured clays and clay loams to obtain truly representative measures of Kfs and the capillarity parameters. The main strengths of the transient (falling


1054

Soil Sampling and Methods of Analysis head) analyses (relative to steady analyses) include reduced measurement time (especially in low-permeability soils) and potentially simpler equipment. Weaknesses of the transient analyses include potentially small sample volume (sampled volume depends on the rate of wetting front migration and measurement duration), the need to measure Du, which requires separate collection of soil samples or use of expensive in situ techniques (e.g., see TDR; Chapter 70), and questionable usability in coarse textured and=or highly structured porous materials (due to rapid flow and potentially unstable wetting fronts).

77.3 OTHER RING INFILTROMETER METHODS Other important ring infiltrometer methods include the ‘‘double’’ or ‘‘concentric’’ ring infiltrometer, the ‘‘twin’’ or ‘‘dual’’ ring infiltrometer, the ‘‘multiple’’ ring infiltrometer, and the ‘‘air-entry’’ permeameter. The double=concentric ring infiltrometer consists of an inner ‘‘measuring’’ cylinder placed concentrically inside an outer ‘‘guard’’ or ‘‘buffer’’ cylinder (Figure 77.1b). It is an established ASTM standard method, where the measuring cylinder and buffer cylinder have set diameters of 30 and 60 cm, respectively (Lukens 1981). Inclusion of the outer buffer cylinder is an attempt to improve the accuracy of Equation 77.2 by reducing flow divergence under the measuring cylinder, which results from ponding (H ), finite cylinder radius (a), and porous medium capillarity (a*). However, laboratory sand tank studies (Swartzendruber and Olsen 1961) and numerical simulation studies (Wu et al. 1997; Smettem and Smith 2002) have shown that the double ring system still systematically overestimates Kfs when Equation 77.2 is applied, although the degree of overestimation tends to be decreased somewhat relative to the single ring infiltrometer. The overestimate continues to occur because the physical barrier provided by the outer buffer cylinder is not effective for eliminating flow divergence. There is consequently no real advantage gained by using the double ring infiltrometer over the single ring infiltrometer, and Equation 77.2 should be avoided regardless of cylinder arrangement. The twin=dual ring and multiple ring infiltrometers employ adjacent cylinders with a single constant head (H) but different cylinder diameters. Two adjacent cylinders are used in the twin ring system, and three or more adjacent cylinders are used in the multiple ring system. The cylinders are typically 5–50 cm in diameter by 5–20 cm long, and are installed individually (not concentrically as in the double ring infiltrometer) with just enough separation to prevent the wetting fronts from merging before steady infiltration is achieved. Both infiltrometer systems can determine Kfs , fm , S, a*, cf , PD, and NP (via constant head steady flow analyses); however, they are not widely used because they are more laborintensive than the single ring approaches (e.g., two or more cylinders vs. one cylinder) and they are highly sensitive to lateral heterogeneity. Further detail on the twin ring and multiple ring infiltrometer methods can be found in Scotter et al. (1982), Reynolds et al. (2002), and Reynolds and Elrick (2005). The air-entry permeameter includes a single cylinder connected to a constant head reservoir. The method estimates Kfs and a* from measurements of the constant head infiltration rate (qs ), the time (tf ) required for the wetting front to reach a specified depth within the cylinder (wetting front depth, zf , must be less than or equal to the cylinder insertion depth, d), and an estimate of the Green–Ampt wetting front pressure head (cf ), which is based on a measurement of the porous medium’s air-entry pressure head (ca ). This method is not extensively used because of



1055

awkward and delicate equipment, somewhat complicated procedures, and frequent difficulty in estimating zf and cf accurately. Further information on the air-entry permeameter method can be found in Topp and Binns (1976) and Reynolds and Elrick (2005).

REFERENCES Bagarello, V., Iovino, M., and Elrick, D.E. 2003. A simplified falling-head technique for rapid determination of field-saturated hydraulic conductivity. Soil Sci. Soc. Am. J. 68: 66–73. Bouma, J. 1985. Soil variability and soil survey. In J. Bouma and D.R. Nielsen, Eds. Proc. Soil Spatial Variability Workshop. PUDOC, Wageningen, The Netherlands, pp. 130–149. Daniel, D.E. 1989. In situ hydraulic conductivity tests for compacted clay. J. Geotech. Eng. 115: 1205–1226. Elrick, D.E., Angulo-Jaramillo, R., Fallow, D.J., Reynolds, W.D., and Parkin, G.W. 2002. Infiltration under constant head and falling head conditions. In P.A.C. Raats, D. Smiles, and A.W. Warrick, Eds. Environmental Mechanics: Water, Mass and Energy Transfer in the Biosphere. Geophysical Monograph 129, American Geophysical Union, Washington, DC, pp. 47–53. Guyonnet, D., Amraoui, N., and Kara, R. 2000. Analysis of transient data from infiltrometer tests in fine-grained soils. Ground Water 38: 396–402. Lauren, J.G., Wagenet, R.J., Bouma, J., and Wo¨sten, J.H.M. 1988. Variability of saturated hydraulic conductivity in a Glassaquic Hapludalf with macropores. Soil Sci. 145: 20–28. Leeds-Harrison, P.B. and Youngs, E.G. 1997. Estimating the hydraulic conductivity of soil aggregates conditioned by different tillage treatments from sorption measurements. Soil Till. Res. 41: 141–147. Lukens, R.P. Ed. 1981. Annual Book of ASTM Standards, Part 19: Soil and Rock. American Standards of Materials and Testing. Washington, DC, pp. 509–514. Reynolds, W.D. 1993. Saturated hydraulic conductivity: field measurement. In M.R. Carter, Ed. Soil Sampling and Methods of Analysis. Canadian

Society of Soil Science. Lewis Publishers, Boca Raton, FL, pp. 605–611. Reynolds, W.D. and Elrick, D.E. 1990. Ponded infiltration from a single ring: I. Analysis of Steady Flow. Soil Sci. Soc. Am. J. 54: 1233–1241. Reynolds, W.D. and Elrick, D.E. 2005. Measurement and characterization of soil hydraulic properties. In J. Alvarez-Benedi and R. MunozCarpena, Eds. Soil–Water–Solute Process Characterization: An Integrated Approach. CRC Press, Boca Raton, FL, pp. 197–252. Reynolds, W.D., Elrick, D.E., and Youngs, E.G. 2002. Ring or cylinder infiltrometers (vadose zone). In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 818–843. Richards, N.E. 1987. A comparison of three methods for measuring hydraulic properties of field soils. MSc thesis. University of Guelph, Guelph, ON, Canada. Scotter, D.R., Clothier, B.E., and Harper, E.R. 1982. Measuring saturated hydraulic conductivity and sorptivity using twin rings. Aust. J. Soil Res. 20: 295–304. Smettem, K.R.J. and Smith, R.E. 2002. Field measurement of infiltration parameters. In R.E. Smith, Ed. Infiltration Theory for Hydrologic Applications. Water Resources Monograph 15, American Geophysical Union, Washington, DC, pp. 135–157. Swartzendruber, D. and Olsen, T.C. 1961. Model study of the double ring infiltrometer as affected by depth of wetting and particle size. Soil Sci. 92: 219–225. Topp, G.C. and Binns, M.R. 1976. Field measurements of hydraulic conductivity with a modified air-entry permeameter. Can. J. Soil Sci. 56: 139–147.


1056


Wu, L., Pan, L., Robertson, M.L., and Shouse, P.J. 1997. Numerical evaluation of ring-infiltrometers under various soil conditions. Soil Sci. 162: 771–777.

Youngs, E.G., Elrick, D.E., and Reynolds, W.D. 1993. Comparison of steady flows from infiltration rings in ‘‘Green and Ampt’’ soils and ‘‘Gardner’’ soils. Water Resour. Res. 29: 1647–1650.

Youngs, E.G. 1987. Estimating hydraulic conductivity values from ring infiltrometer measurements. J. Soil Sci. 38: 623–632.

Youngs, E.G., Spoor, G., and Goodall, G.R. 1996. Infiltration from surface ponds into soils overlying a very permeable substratum. J. Hydrol. 186: 327–334.


Chapter 78 Saturated Hydraulic Properties: Auger Hole G. Clarke Topp Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

78.1 INTRODUCTION The auger-hole method is a field technique for measuring the in situ saturated hydraulic conductivity, Ks [LT1 ], of porous materials within the saturated zone (i.e., below the water table); and it is perhaps the most reliable and trusted means of obtaining Ks values for the design of subsurface tile drainage systems. An alternative in situ method for Ks measurement in the saturated zone is given in Chapter 79 (piezometer), while several laboratory and estimation techniques for Ks determination are given in Chapter 75 and Chapter 84, respectively. A discussion of the principles and parameters associated with the determination of Ks appears in Chapter 69. The auger-hole method is based on an application of Darcy’s law (Chapter 69) where the initial equilibrium or ‘‘static’’ water level in the auger hole (the static water level is usually equivalent to the water table level) is rapidly raised or lowered, and then the recovery to the static level is monitored through time as water flows between the auger hole and the surrounding porous material. The method and most of the equipment are similar to the piezometer method (Chapter 79). The Ks value is determined using (Boast and Kirkham 1971; Youngs 2001) Ks ¼ CAH

Dy Dt

(78:1)

where Dy (cm) is the change in water level in the auger hole (relative to the initial static water level) during time interval, Dt (s), and CAH is a dimensionless shape factor. The CAH parameter is related to the radius of the auger hole, rc , the depth of the auger hole below the static water level, H, and the depth from the base of the auger hole to an impermeable or highly permeable porous medium layer, F (Figure 78.1). The CAH values can be obtained

1057


Location:

Date:

Site:

Operator:

Soil surface

Auger-hole data sheet

Reference level

Auger-hole method yb

Piezometer method

r

ya

E

W

F

Notes:

W yb

d D

ya

H hc

T = 19°C

hc = 27 cm

rc = 5 cm

F = ∞

F = (F − D − E ) = ∞

r = 3.65 cm

2rc

E = 41 cm

D = 50 cm

H = (E + D − W) = 42 cm

yi = (yi − E )

E = cm

D = cm

H = (E + D − W) = cm

∆y/∆t

t (s)

1

0

85.5

2

15

84.8

0.0467

2

3

30

84.1

0.0467

3

4

45

83.4

0.0467

4

i

5

60

82.7

0.0467

5

75

82.1

0.04

6

(ya − yb) = 85.5 − 82.1

tb − t a =

0.0453

Average y /H = 0.829

H/rc = 8.4 −3

CAH = 25.1 10

H/rc = −1

Ks = CAH ( ya − yb)/(tb − ta) cm s

Ks = 1.14 10−3 cm s−1

t (s)

yi = (yi − E ) ∆y/∆t

y (cm)

1

6 tb − ta = 75

W = (W − E) = cm

W = cm

i

y (cm)

F

2rc

F/H = ∞

W = (W − E) = 8 cm

W = 49 cm

Impermeable or highly permeable layer


1058

CAH = Ks =

(ya − yb) = Average y /H = Ks = CAH ( ya − yb)/(tb − ta) cm s−1 cm s−1

FIGURE 78.1. Data sheet for the auger-hole method, including schematic diagrams of the equipment for the auger-hole method and the piezometer method (Chapter 79).

from Table 78.1 (Boast and Kirkham 1971; Youngs 2001), or from graphs (van Beers 1970; Bouwer and Jackson 1974), or empirical equations (Ernst 1950; Amoozegar 2002). The auger hole must be large enough to sample a sufficient volume of porous medium to yield representative Ks values. Unfortunately, the complexity of both the flow regime and auger hole geometry precludes a precise determination of the volume of porous medium sampled, although it is clear that the porous material adjacent to the wall and base of the auger hole exerts by far the greatest influence on the results. Generally speaking, the larger the auger hole diameter, the more representative the Ks measurements; and auger hole diameters of 10 cm or more are typical. Most analyses assume flat-bottomed auger holes (i.e., right cylindrical geometry), and it is generally recommended that the bottom of the hole be at least 30 cm below the static water level (water table). The basic procedure for making an auger-hole measurement is as follows: (i) auger a hole that extends to at least 30 cm below the static water level; (ii) allow the water in the auger

1 0.75 0.5 1 0.75 0.5 1 0.75 0.5 1 0.75 0.5 1 0.75 0.5 1 0.75 0.5 1 0.75 0.5

1

518 544 643 215 227 271 60.2 63.6 76.7 21.0 22.2 27.0 6.86 7.27 8.90 1.45 1.54 1.90 0.43 0.46 0.57

0

490 522 623 204 216 261 56.3 60.3 73.5 19.6 21.0 25.9 6.41 6.89 8.51 1.37 1.47 1.82 0.41 0.44 0.54

0.05 468 503 605 193 208 252 53.6 57.9 71.1 18.7 20.2 24.9 6.15 6.65 8.26 1.32 1.42 1.79 0.39 0.42 0.52

0.1 435 473 576 178 195 240 49.6 54.3 67.4 17.5 19.1 23.9 5.87 6.38 7.98 1.29 1.39 1.74 0.39 0.42 0.52

0.2 375 418 521 155 172 218 44.9 49.5 62.5 16.4 18.0 22.6 5.58 6.09 7.66 1.24 1.35 1.69 0.38 0.41 0.51

0.5 331 376 477 143 160 203 42.8 47.6 60.2 15.8 17.4 22.0 5.45 5.97 7.52 1.22 1.32 1.67 0.37 0.41 0.51

1 306 351 448 137 154 196 41.9 46.6 59.1 15.5 17.2 21.8 5.40 5.90 7.44

2 296 339 441 135 152 194

5 295 338 440 133 152 194 41.5 46.4 58.8 15.5 17.2 21.7 5.38 5.89 7.44 1.21 1.31 1.66 0.37 0.41 0.51

F=H ¥ 292 335 437 133 151 193

5 280 322 416 131 148 190 41.2 45.9 58.3 15.4 17.1 21.6 5.36 5.88 7.41

2

247 287 376 123 140 181 40.1 44.8 57.1 15.2 16.8 21.3 5.31 5.82 7.35 1.19 1.30 1.65 0.32 0.39 0.50

1

193 230 306 106 123 161 37.6 42.1 54.1 14.6 16.2 20.6 5.17 5.67 7.16 1.18 1.28 1.61 0.36 0.39 0.50

0.5

F=H for underlying impermeable or highly permeable layer

a

rc ¼ Radius of the auger hole, H ¼ depth of water in the hole at static level, y ¼ average distance from the water level in the hole to the static level for two consecutive measurements, and F ¼ distance from the bottom of the borehole to an impermeable or highly permeable layer.

Source: From Boast, C.W. and Kirkham, D., Soil Sci. Soc. Am. Proc. 35, 365, 1971, as modified by Youngs, E.G., in K.A. Smith and C.E. Mullins (Eds.), Soil and Environmental Analysis: Physical Methods, 2nd ed., Marcel Dekker, New York, New York, 2001, 157–160.

100

50

20

10

5

2

y=H

H=rc

F=H for underlying impermeable layer

TABLE 78.1 Values of CAH 3 103 for the Auger-Hole Method (Equation 78.1) with an Underlying Impermeable or Highly Permeable Layera


Saturated Hydraulic Properties: Auger Hole 1059


1060


hole to equilibrate to the static level; (iii) add or remove water from the auger hole to initiate water flow into or out of the hole; and (iv) monitor the early-time change in water level in the auger hole as it reequilibrates to the static level.

78.2 APPARATUS AND SUPPLIES 1

Soil auger(s) of 10 cm diameter or larger. A variety of designs are available for specific porous medium conditions; for example, ‘‘dutch=mud’’ augers are often best in clayey soils, while ‘‘bucket=sand’’ or ‘‘screw’’ augers are usually best for sandy and stony soils. Once a hole is augered to the desired depth, a ‘‘planer’’ auger can be used to produce the flat bottom assumed in the auger hole analysis (see e.g., Figure 29.11 in Amoozegar and Warrick 1986).

2

Equipment for producing a rapid change in auger hole water level. This usually involves rapid addition or removal of water. Rapid removal of water from the auger hole is easily achieved using a bailer or water pump (e.g., Figure 3.2.1.6 in Young 2002). If a pump is used, it should have a pumping rate > 0:5 L s1 so that water level decline occurs quickly. Rapid addition of water can be accomplished by simply pouring water into the hole (see Comment 3 in Section 78.5). If addition or removal of water is problematic, a rapid initial increase in water level can be produced by quickly submerging a ‘‘slug’’ (see Procedure 5; Comment 3, Section 78.5).

3

Timer. A stopwatch or equivalent timer (graduated in seconds) for timing the rise or fall of the water level in the auger hole.

4

Water level measuring device. Several possibilities exist depending either on the rate of rise or fall of the water level. If water level change is relatively slow (e.g., 1 cm per interval, but short enough that several measurements can be made while y=H 0:5. Intervals of 10–20 s are usually optimal.

5

Produce a rapid change in water level. Once the water level returns to its static level after the auger hole conditioning process, quickly lower or raise the water level in the hole to initiate water level change. Lowering the water level is achieved using a ‘‘bailer’’ or pump (e.g., Figure 3.2.1.6 in Young 2002) to remove water; raising the water level by a known distance can be accomplished by rapidly submerging a ‘‘slug’’ (solid cylinder of known volume suspended on a rope), or by simply adding water to the borehole. Water removal or use of a slug is usually preferred over water addition (see Comment 3 in Section 78.5). Dispose removed water well away from the auger hole (or place it in a container) so that it cannot perturb the measurements by re-entering the auger hole or disturbing the static water level.

6

Record the rate of water level rise or fall. Quickly place the water level recorder in the auger hole and initiate monitoring the change in water level with time (some water level recording systems, e.g., pressure transducers, can be placed in the auger hole before step 5). Record ti and yi0 as illustrated in Figure 78.1, and attempt to obtain several values before y=H ¼ 0:5. At least five times of water level readings are recommended. Step 5 and step 6 can be repeated to obtain additional or confirming measurements.


1062


78.4 CALCULATIONS 1

Calculate Dy=Dt following the example on the left in Figure 78.1. These values are expected to decrease monotonically with time for water level rise. If Dy=Dt is constant or decreasing consistently, then either a single timing interval or multiple timing intervals may be chosen for the next step.

2

Choose a time interval (tb ta ) and corresponding (ya 0 yb 0 ), then calculate Dy=Dt, the average y=H, and H=rc (see Figure 78.1). Note that the average water level depth relative to the static level for two consecutive measurements is y ¼ [(ya 0 þ yb 0 )=2] W 0 .

3

Estimate CAH from Table 78.1 using graphical interpolation for maximum accuracy.

4

Calculate Ks using Equation 78.1; calculate CAH and Dy=Dt, as illustrated in Figure 78.1.

78.5 COMMENTS 1

The auger-hole method offers a reliable and relatively rapid procedure for determining the Ks of a relatively large volume of porous material in the saturated zone (Amoozegar 2002). For large values of H=rc , the auger-hole method approaches a measure of the horizontal Ks . Amoozegar (2002) provides a detailed method for using the auger-hole method in layered soils.

2

The example calculation given in Figure 78.1 uses Table 78.1 to find CAH via linear interpolation between adjacent table entries. The CAH coefficient is not linearly dependent on H=rc and y=H, however, and greater precision in CAH is usually obtained via nonlinear graphical interpolation based on simulation modeling results and quasiempirical regression relationships. For the example in Figure 78.1, applying nonlinear graphical interpolation resulted in a 20% reduction in Ks relative to linear interpretation, which may or may not be important depending on the intended Ks application. Nonlinear graphical interpolation of the CAH values in Table 78.1 is recommended for maximum Ks accuracy (see van Beers 1970; Bouwer and Jackson 1974).

3

In principle, the auger-hole method works equally well regardless of whether the rise or fall of water is recorded. Adding water to the hole to achieve a decline in water level is not recommended, however, as rapid addition of water may introduce silt and clay into suspension (e.g., through erosion of the borehole wall); and this entrained material can cause progressive siltation (partial plugging) of the auger hole and thereby result in unrepresentative low Ks values.

4

The terms ‘‘highly permeable layer’’ and ‘‘impermeable layer’’ in Figure 78.1 do not indicate absolute Ks values or ranges, but rather large differences in Ks relative to the Ks of the zone being measured. For example, Amoozegar (2002) specifies that an ‘‘impermeable layer’’ has Ks , that is 20% of the Ks in the zone being measured.


Saturated Hydraulic Properties: Auger Hole

1063

REFERENCES Amoozegar, A. 2002. Auger-hole method (saturated zone). In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 859–869. Amoozegar, A. and Warrick, A.W. 1986. Hydraulic conductivity of saturated soils: field methods. In A. Klute, Ed. Methods of Soil Analysis, Part 1: Physical and Mineralogical Methods, 2nd ed. Soil Science Society of America, Madison, WI, pp. 735–770. Boast, C.W. and Kirkham, D. 1971. Auger hole seepage theory. Soil Sci. Soc. Am. Proc. 35: 365–373. Bouwer, H. and Jackson, R.D. 1974. Determining soil properties. In J. van Schilfgaarde, Ed. Drainage for Agriculture. American Society of Agronomy, Madison, WI, pp. 611–672. Ernst, L.F. 1950. Een nieuwe formule voor de berekening van de doorlaatfactor met

deboorgatenmethode. Groningen, The Netherlands. Rap. Landbouwproefsta. en Bodemkundig Inst. T.N.O. van Beers, W.F.J. 1970. The auger-hole method: a field measurement of the hydraulic conductivity of soil below the water table. International Institute for Land Reclamation and Improvement (ILRI), Bull No. 1. H. Veenman and Zonen, Wageningen, the Netherlands. Young, M.H. 2002. Water Potential. In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 547–573. Youngs, E.G. 2001. Hydraulic conductivity of saturated soils. In K.A. Smith and C.E. Mullins, Eds. Soil and Environmental Analysis: Physical Methods, 2nd ed. Marcel Dekker, New York, NY, pp. 157–160.



Chapter 79 Saturated Hydraulic Properties: Piezometer G. Clarke Topp Agriculture and Agri-Food Canada Ottawa, Ontario, Canada

79.1 INTRODUCTION The piezometer method is a well-established field technique for measuring the in situ saturated hydraulic conductivity, Ks [LT1 ], of porous materials within the saturated zone (i.e., below the water table). An alternative in situ method for Ks measurement in the saturated zone is given in Chapter 78 (Auger Hole), while several laboratory and estimation techniques for Ks determination are given in Chapter 75 and Chapter 84, respectively. A discussion of the principles and parameters associated with the determination of Ks appears in Chapter 69. The piezometer method measures Ks through an open-ended pipe or casing inserted into a borehole that extends into the saturated zone of a confined or an unconfined aquifer (Figure 79.1). The pipe may extend to the bottom of the borehole, or it may terminate above the bottom leaving a cylindrical ‘‘piezometer cavity’’ (Figure 79.1). Often the borehole is flat-bottomed, although other shapes can be used (Youngs 1968). It is important that the pipe is sealed against the borehole wall so that leakage or short-circuit flow along the outside wall of the pipe is prevented (Figure 79.1). The principle of piezometer operation is the same as that of the auger-hole method (Chapter 78), which is explained as follows: first, it consists of allowing the water level in the piezometer pipe to equilibrate to the static or equilibrium water level, then quickly changing the water level in the pipe (usually by adding or removing water), and then monitoring the return of the water level back to the static level. The static water level is the water table elevation in unconfined aquifers, and the phreatic surface in confined aquifers. For water level rise in the piezometer, the saturated hydraulic conductivity is given by (Figure 79.1) Ks ¼

pr 2 ln (ya =yb ) CP (tb ta )

(79:1)

1065



1066

Piezometer data sheet

Site:

Operator:

Auger-hole method Soil surface

Date:

Notes:

Reference level

yb ya W

r

yb

d D

rc = 5 cm

hc = 27 cm

F = ∞

F = (F − D − E) = ∞

W = 51.5 cm

E = 44 cm

hc 2rc

F/H = ∞

d = (H − hc ) = (E + D − W − hc) = 57 cm

yi = (yi−E )

W=

yi (cm)

yi /yi +1

I

95

51

1.06

1

60

92.3

48.3

1.06

2

3

120

89.7

45.7

1.05

3

4

180

87.6

43.6

1.06

4

5

240

85

41

1.06

5

6

300

82.8

38.8

t (s)

1

0

2

cm

E=

2rc

cm

t (s)

y (cm)

F

W = (W − E ) =

d = (H − hc) = (E + D − W − hc) =

y (cm)

I

ya

H

r = 3.65 cm

W = (W − E) = 7.5 cm

E F

W

T = 19C

Piezometer method Impermeable or highly permeable layer

Location:

yi (cm)

cm

cm yi = (yi−E ) yi /yi +1

6

tb − ta = 300

yb /ya = 51/38.8 = 1.31

ln(yb /ya) = 0.273 tb − ta =

hc /rc = 7.4

d /rc = 15.6

C P /rc = 27

Ks = πr ln(yb / ya)/CP (tb − ta) (cm

C P = 7.40

Ks = 9.65 10−3 cm s−1

F/rc = ∞

2

s−1)

ln(yb /ya ) =

yb /ya =

hc /rc =

d /rc =

C P / rc =

Ks = πr 2ln (yb /ya)/CP(tb−ta) (cm s−1)

F /rc =

CP =

Ks =

cm s−1

FIGURE 79.1. Data sheet for the piezometer method, including schematic diagrams of the equipment for the piezometer method and the auger-hole method (Chapter 78).

where ya and yb are the water depths below the static level at times ta and tb , respectively; r is the inside radius of the pipe, and CP is a shape factor (Table 79.1) that depends on the height of the static water level above the end of the pipe (d ), on the length (hc ), on the radius (rc ) of the piezometer cavity, and on the distance below the bottom of the borehole to a layer of greatly different hydraulic conductivity (F) (Youngs 1968, 2001; Amoozegar 2002).

79.2 APPARATUS AND SUPPLIES 1

Soil augers of 10 cm diameter or larger. A variety of designs are available for specific porous medium conditions; for example, ‘‘dutch=mud’’ augers are often best for clayey soils, while ‘‘bucket=sand’’ or ‘‘screw’’ augers are usually best for

d =rc

20 16 12 8 4

20 16 12 8 4

20 16 12 8 4

20 16 12 8 4

20 16 12 8 4

hc =rc

0

0.5

1.0

2.0

4.0

18.6 19.0 19.4 19.8 21.0

13.8 13.9 14.0 14.3 15.0

10.6 10.7 10.8 11.0 11.5

8.7 8.8 8.9 9.0 9.5

5.6 5.6 5.6 5.7 5.8

¥

18.0 18.4 18.8 19.4 20.5

13.5 13.6 13.7 14.1 14.9

10.4 10.5 10.6 10.9 11.4

8.6 8.7 8.8 9.0 9.4

5.5 5.5 5.5 5.6 5.7

8.0

17.3 17.6 18.0 18.7 20.0

12.8 13.0 13.2 13.6 14.5

10.0 10.1 10.2 10.5 11.2

8.3 8.4 8.5 8.7 9.0

5.3 5.3 5.4 5.5 5.6

4.0

16.3 16.6 17.1 17.6 19.1

11.9 12.1 12.3 12.7 13.7

9.3 9.4 9.5 9.8 10.5

7.7 7.8 8.0 8.2 8.6

5.0 5.0 5.1 5.2 5.4

2.0

15.3 15.6 16.0 16.4 17.8

10.9 11.0 11.2 11.5 12.6

8.4 8.5 8.6 8.9 9.7

7.0 7.0 7.1 7.2 7.5

4.4 4.4 4.5 4.6 4.8

1.0

F=rc for impermeable layer

14.6 14.8 15.1 15.5 17.0

10.1 10.2 10.4 10.7 11.7

7.6 7.7 7.8 8.0 8.8

6.2 6.2 6.3 6.4 6.5

3.6 3.6 3.7 3.8 3.9

0.5

13.6 13.8 14.1 14.5 15.8

9.1 9.2 9.4 9.6 10.5

6.3 6.4 6.5 6.7 7.3

4.8 4.8 4.8 4.9 5.0

0 0 0 0 0

0

18.6 19.0 19.4 19.8 21.0

13.8 13.9 14.0 14.3 15.0

10.6 10.7 10.8 11.0 11.5

8.7 8.8 8.9 9.0 9.5

5.6 5.6 5.6 5.7 5.8

¥

19.8 20.0 20.3 20.6 21.5

14.1 14.3 14.4 14.8 15.4

11.0 11.0 11.1 11.2 11.6

8.9 9.0 9.1 9.3 9.6

5.6 5.6 5.7 5.7 5.8

8.0

20.8 20.9 21.2 21.4 22.2

15.0 15.1 15.2 15.5 16.0

11.6 11.6 11.7 11.8 12.1

9.4 9.4 9.5 9.6 9.8

5.8 5.8 5.9 5.9 6.0

4.0

22.7 22.8 23.0 23.3 24.1

16.5 16.6 16.7 17.0 17.6

12.8 12.8 12.8 12.9 13.1

10.3 10.3 10.4 10.5 10.6

6.3 6.4 6.5 6.6 6.7

2.0

25.5 25.6 25.8 26.0 26.8

19.0 19.1 19.2 19.4 20.1

14.9 14.9 14.9 14.9 15.0

12.2 12.2 12.2 12.3 12.4

7.4 7.5 7.6 7.7 7.9

1.0

F=rc for highly permeable layer

TABLE 79.1 Values of CP =rc for the Piezometer Method (Equation 79.1) with an Impermeable or Highly Permeable Layer below the Boreholea

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0

Saturated Hydraulic Properties: Piezometer (continued)

29.9 29.9 30.0 30.2 31.5

23.0 23.1 23.2 23.3 23.8

19.0 19.0 19.0 19.0 19.0

15.2 15.2 15.3 15.3 15.4

10.2 10.3 10.4 10.5 10.7

0.5


1067

20 16 12 8 4

8.0

26.9 27.4 28.3 29.1 30.8

¥ 26.3 26.6 27.2 28.2 30.2

8.0 25.5 25.8 26.4 27.4 29.6

4.0 24.0 24.4 25.1 26.1 28.0

2.0

a

23.0 23.4 24.1 25.1 26.9

1.0 22.2 22.7 23.4 24.4 25.7

0.5 21.4 21.9 22.6 23.4 24.5

0 26.9 27.4 28.3 29.1 30.8

¥ 29.6 29.8 30.0 30.3 31.5

8.0 30.6 30.8 31.0 31.2 32.8

4.0 32.9 33.1 33.3 33.8 35.0

2.0 36.1 36.2 36.4 36.9 38.4

1.0

F=rc for highly permeable layer

40.6 40.7 40.8 41.0 43.0

0.5

1 1 1 1 1 1

0

d ¼ Length of piezometer pipe below the static water level, rc ¼ radius of the piezometer cavity, hc ¼ length of the piezometer cavity, and F ¼ depth from the base of the piezometer cavity to an impermeable or highly permeable layer (see Figure 79.1).

Source: From Youngs, E.G., Soil Sci., 106, 235, 1968.

d =rc

hc =rc

F=rc for impermeable layer

TABLE 79.1 (continued) Values of CP =rc for the Piezometer Method (Equation 79.1) with an Impermeable or Highly Permeable Layer below the Boreholea


1068 Soil Sampling and Methods of Analysis


Saturated Hydraulic Properties: Piezometer

1069

sandy and stony soils. Once a hole is augered to the desired depth, a ‘‘planer’’ auger can be used to produce the flat bottom assumed in the piezometer analysis (see e.g., Figure 29.11 in Amoozegar and Warrick 1986). 2

Equipment for producing a rapid change in piezometer water level. This usually involves rapid addition or removal of water. Rapid water removal is easily achieved using a bailer or water pump (see e.g., Figure 3.2.1.6 in Young 2002). If a pump is used, it should have a pumping rate > 0:5 L s1 so that water level decline occurs quickly. Water addition involves simply pouring water into the piezometer pipe, although this approach is not recommended (see Comment 3 in Section 79.5). One can also change the piezometer water level by quickly lowering a ‘‘slug’’ (solid cylinder of known volume) into the piezometer to displace the water level a known distance upward, or quickly removing a slug (after the static level is reestablished) to displace the water level a known distance downward.

3

Timer. A stopwatch or equivalent timer (graduated in seconds) for timing the rise or fall of water in the pipe.

4

Water level measuring device. Several possibilities exist depending somewhat on the rate of rise or fall of the water level. If water level change is relatively slow (e.g., 0:5 cm cm1


Unsaturated Hydraulic Properties: Laboratory Evaporation

1105

in the 5T setup, these thresholds being determined by experience and by the accuracy and precision of the pressure transducers. Other methods, such as the tension infiltrometer (Chapter 80 and Chapter 82), should be used to obtain K (c) in the 10 cm c 0 range (Wendroth and Simunek 1999). 2

Method assumptions: Critical assumptions of the method are that changes in water content and pore water matric head between adjacent tensiometers are linear, which further implies that water flow is steady and the soil sample is homogeneous and isothermal. Wendroth et al. (1993) validated these assumptions for the 2T setup over the range, 650 cm c 0, and also avoided potential sample boundary effects (Becher 1975) by locating the tensiometers away from the sample ends. Some attempts have been made to extend the method into the ‘‘dry soil’’ range (i.e., c < 700 cm) by installing gypsum blocks or other devices (e.g., Becher 1971a,b). This is not advisable for the 2T setup, however, as the assumption of linear water flux profile would likely be violated, and the hydraulic head gradient causing flux may not be adequately represented by c measurements at only two depths. Moreover, allowing evaporation to continue would eventually produce a downwardmigrating ‘‘drying front,’’ which would cause a substantial decrease in evaporation rate as well as invalidate the application of Darcy’s law for liquid water flow (Idso et al. 1974). A relatively steady evaporation rate is generally a good indicator that the quasisteady liquid water flow regime required by theory has been achieved (Willis 1960).

3

Additional information: Additional information on experimental procedures using two or more tensiometers is given in Schindler (1980) and Tamari et al. (1993). The optimum sample diameter or length and number of tensiometers depend on the soil condition, the objective of the investigation, and operator experience. On the one hand, it is not advisable to install many tensiometers in short samples, as the optimum tensiometer spacing for accurate determination of hydraulic head gradient is 2–3 cm. On the other hand, increasing sample length to accommodate more tensiometers often invalidates the assumption of sample homogeneity, which may complicate the results (e.g., Figure 81.4b). Sample diameters of 8–10 cm, sample lengths of 5–10 cm, and installation of two to five tensiometers with 2–3 cm spacing are used most commonly and most successfully. The top and bottom tensiometers should be installed at least 1–2 cm from the ends of the sample to avoid possible end effects. Criteria for minimum representative sample size in structured and unstructured soils are given in Chapter 69.

REFERENCES Becher, H.H. 1971a. Ein Verfahren zur Messung der ungesa¨ttigten Wasserleitfa¨higkeit. Z. Pflanzenern. Bodenkd. 128: 1–12. Becher, H.H. 1971b. Ergebnisse von Wasserleitfa¨higkeitsmessungen im wasserungesa¨ttigten Zustand. Z. Pflanzenern. Bodenkd. 128: 227–234.

Becher, H.H. 1975. Bemerkungen zur Ermittlung der ungesa¨ttigten Wasserleitfa¨higkeit unter nichtstationa¨ren Bedingungen. Z. Pflanzenern. Bodenkd. 132: 1–12. Campbell, G.S. 1974. A simple method for determining conductivity from moisture retention data. Soil Sci. 117: 311–314.


1106 Groenevelt, P.H. and Grant, C.D. 2004. A new method for the soil-water retention curve that solves the problem of residual water contents. Eur. J. Soil Sci. 55: 479–485. Idso, S.B., Reginato, R.J., Jackson, R.D., Kimball, B.A., and Nakayama, F.S. 1974. The three stages of drying of a field soil. Soil Sci. Soc. Am. Proc. 38: 831–837. Press, W.H., Flannery, B.P., Teukolsky, S.A., and Vetterling, W.T. 1990. Numerical Recipes. Cambridge University Press. New York, NY, 702 pp. Schindler, U. 1980. Ein Schnellverfahren zur Messung der Wasserleitfa¨higkeit im teilgesa¨ttigten Boden an Stechzylinderproben. Arch. Acker-Pflanzenbau Bodenkd. 24: 1–7. Tamari, S., Bruckler, L., Halbertsma, J., and Chadoeuf, J. 1993. A simple method for determining soil hydraulic properties in the laboratory. Soil Sci. Soc. Am. J. 57: 642–651. van Genuchten, M.Th. 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Am. J. 36: 380–383. van Genuchten, M.Th., Leij, F.J., and Yates, S.R. 1991. The RETC code for quantifying the hydraulic functions of unsaturated soils, Version

Soil Sampling and Methods of Analysis 1.0. EPA Report 600=2-91=065, U.S. Salinity Laboratory, USDA, ARS, Riverside, CA. Watson, K.K. 1966. An instantaneous profile method for determining the hydraulic conductivity of unsaturated porous materials. Water Resour. Res. 2: 709–715. Wendroth, O., Ehlers, W., Hopmans, J.W., Kage, H., Halbertsma, J., and Wo¨sten, J.H.M. 1993. Reevaluation of the evaporation method for determining hydraulic functions in unsaturated soils. Soil Sci. Soc. Am. J. 57: 1436–1443. Wendroth, O. and Simunek, J. 1999. Soil hydraulic properties determined from evaporation and tension infiltration experiments and their use for modeling field moisture status. In Van Genuchten, M.Th. and F.J. Leij, Eds. Proceedings of the International Workshop on ‘‘Characterization and Measurement of the Hydraulic Properties of Unsaturated Porous Media,’’ Riverside, CA, pp. 737–748. Willis, W.O. 1960. Evaporation from layered soils in the presence of a water table. Soil Sci. Soc. Am. Proc. 24: 239–242. Wind, G.P. 1968. Capillary conductivity data estimated by a simple method. In P.E. Rijtema and H. Wassink, Eds. Water in the Unsaturated Zone. Proceedings of the Wageningen Symposium. June 1966. Vol. 1, IASAH, Gentbrugge, Belgium, pp. 181–191.


Chapter 82 Unsaturated Hydraulic Properties: Field Tension Infiltrometer W.D. Reynolds Agriculture and Agri-Food Canada Harrow, Ontario, Canada

82.1 INTRODUCTION The tension or disk infiltrometer is used primarily for field (in situ) measurement of nearsaturated hydraulic conductivity, K(c) [LT1 ], and sorptivity, S(c) [LT1=2 ]. It can also be used, however, to determine near-saturated sorptive number, a*(c) [L1 ], flow-weighted mean pore diameter, PD(c) [L], and number of flow-weighted mean pores per unit area, NP(c) [L2 ]. By ‘‘near-saturated’’ we mean measurements at pore water matric heads (c) within the range, 20 cm c þ2 cm, although some infiltrometers can operate at matric heads as low as 40 cm. An alternative field method for K(c) determination in the ‘‘wet-end’’ (i.e., 200 cm c 0) is given in Chapter 83. Laboratory methods for K(c) or K(u) determination are given in Chapter 80 and Chapter 81. Selected methods for estimating K(c) from surrogate porous medium properties are given in Chapter 84. A discussion of the principles and parameters associated with the determination of K(c) and the capillarity relationships [i.e., S(c), a*(c), PD(c), NP(c)] appears in Chapter 69. Given that flow from field-based tension infiltrometers is three dimensional, the measured K(c) and capillarity relationships are most relevant to three-dimensional flow. Tension infiltrometers consist essentially of a 10–20 cm diameter ‘‘infiltrometer plate’’ containing a hydrophilic porous disk (e.g., ceramic, porous plastic, porous metal) or membrane (e.g., nylon mesh, sieve screen) connected to a water reservoir and a Mariotte-type bubble tower. The reservoir supplies water to the disk=membrane and the bubble tower determines the water matric head, c0 , on the disk=membrane (20 c0 þ2 cm). When the infiltrometer is placed on an unsaturated porous medium, the capillarity of the porous medium ‘‘sucks’’ the water out of the infiltrometer such that water infiltrates the porous medium under matric head, c0 . A layer of contact sand is frequently placed under the infiltrometer to ensure good hydraulic connection between the disk=membrane and the porous medium. Tension infiltrometer measurements can usually be obtained within a few minutes to several hours, depending on the type of analysis used and number of matric heads (c0 ) set on the disk=membrane. Several steady-flow and transient-flow analyses are

1107



1108

available, which involve single, twin, or multiple infiltrometers and single or multiple matric heads (e.g., Elrick and Reynolds 1992; White et al. 1992; Clothier 2000; Smettem and Smith 2002). We will focus here, however, on one steady-flow analysis and one transient-flow analysis—both are physically based, practical, well established, and applicable to a single disk and one or more matric heads (see Section 82.4.6). We will also focus on the ‘‘overhead reservoir’’ infiltrometer design (e.g., the ‘‘CSIRO’’ or ‘‘Guelph’’ tension infiltrometers), notwithstanding that other valid designs exist (see Section 82.4.6).

82.2 APPARATUS AND PROCEDURES 1

For a measurement at the porous medium surface, choose a level site (or cut a level ‘‘bench’’ if on a continuous hillslope), which is at least as large as the retaining ring for the contact material (Figure 82.1) (see Section 82.4.5). Clip vegetation flush with the surface and remove all debris that are both loose and ‘‘large’’ (e.g., plant residues, stones, etc., which are more than about 0.4 cm high), as this material may cause poor contact between the porous medium and the contact sand. Avoid measurement sites containing large ‘‘attached’’ debris, such as partially exposed roots, partially buried plant residues and stones, large clods, etc., which are greater than about 0.4 cm high. The disruption of the porous medium caused by removal of attached debris or extraction of plant roots may change the hydraulic properties of the infiltration surface. For a subsurface measurement (or measurement on a bench cut into a hillslope), the site should

Disk for leveling contact sand

11−21 cm 5−10 cm

Retaining ring

Contact sand Soil surface

1.0 cm

~0.5 cm

270 Mesh “Guard” cloth Soil macropore

FIGURE 82.1. Schematic of the retaining ring, leveling disk, contact sand, and guard cloth for use in the tension infiltrometer method. (Adapted from Reynolds, W.D., in M.R. Carter (Ed.), Soil Sampling and Methods of Analysis, Canadian Society of Soil Science, Lewis Publishers, Boca Raton, Florida, 1993. With permission.)


Unsaturated Hydraulic Properties: Field Tension Infiltrometer

1109

be excavated using procedures that minimize smearing and compaction, which can also alter hydraulic properties. 2

Gently press a sharpened ‘‘retaining ring’’ into the infiltration surface to a depth of about 0.5 cm (Figure 82.1) (see Section 82.4.1). The retaining ring, which is conveniently constructed from PVC sewer pipe, should be just large enough to allow the chosen size of infiltrometer (10–20 cm diameter) to fit inside (Figure 82.2). Reservoir air tube (blocked) Reservoir

Tripod ~60 cm

Soil surface Air entry

Bubble tower air tube (≥0.25 cm i.d.) Cap Water supply tube

O-ring

Second air tube (≥0.25 cm i.d.) Bubble tower Sealed air tube base

Transparent polycarbonate infiltrometer plate (10–20 cm diameter)

Water outlet port

Retaining ring Contact material

Porous disk/ membrane

Soil surface

270 Mesh nylon bolting cloth

FIGURE 82.2. Schematic of the tension infiltrometer apparatus—‘‘Guelph’’ version with concentric overhead reservoirs. (Adapted from Reynolds, W.D., in M.R. Carter (Ed.), Soil Sampling and Methods of Analysis, Canadian Society of Soil Science, Lewis Publishers, Boca Raton, Florida, 1993. With permission.)


1110


3

Lay a circle of flexible 270 mesh ‘‘guard’’ cloth (e.g., 53 mm pore size, ‘‘Nitex’’ nylon bolting cloth, precut to the same inside diameter as the retaining ring) on the infiltration surface inside the ring (see Section 82.4.1). Pour air-dry contact sand (discussed further below) on top of the cloth to an average depth of 1 cm (a 1 cm reference line on the inside of the retaining ring helps achieve this). Level, smooth, and lightly tamp the contact sand using a wooden disk or similar device (Figure 82.1).

4

Attach the presaturated infiltrometer plate and supply tube assembly (see Section 82.4.2 for details concerning the construction and saturation of the infiltrometer plate) to the water reservoir (e.g., see Chapter 76) and stand upright with the infiltrometer disk submerged in a flat-bottomed pail below a few centimeters of water. Support the infiltrometer using a large tripod (Figure 82.2). Close the water outlet of the infiltrometer plate by firmly pushing the base of the reservoir air tube down into the outlet port, and then fill the reservoir to the top with water. The base of the reservoir air tube should be sealed (clear silicon caulking or a small rubber bung works well for this), so that the bubble tower air tube provides the only source of air (Figure 82.2). Using a syringe (illustrated in Figure 82.6), adjust the water level in the bubble tower to establish the minimum (most negative) desired matric head. Open the water outlet of the infiltrometer plate by lifting the base of the reservoir air tube out of the water outlet port (as indicated in Figure 82.2). The water used in the infiltrometer should meet the same specifications as given for the well permeameter method (see Chapter 76).

5

Lift the infiltrometer out of the bucket and tilt slightly to remove water from the upper surface of the infiltrometer plate. Carefully lower the infiltrometer plate onto the contact sand, using a slight twist to ensure contact. Use the tripod to hold the infiltrometer in place and to carry the weight of the reservoir and water (Figure 82.2, inset). When initial contact with the contact sand is made, air bubbles should rise rapidly in the bubble tower and up into the reservoir, indicating rapid saturation of the contact sand. Air bubbles should not appear from any location other than the bubble tower air tube and the connection between the bubble tower and the water supply tube (Figure 82.2; see also Section 82.4.2). The infiltrometer is operating properly when air bubbles rise regularly up into the reservoir from the connection between the bubble tower and the supply tube. Measurements should be limited to porous materials, which are at field capacity or drier (i.e., antecedent pore water matric head, ci 50 to 100 cm), otherwise there may be insufficient capillarity in the porous medium to draw water out of the infiltrometer against the tension applied by the bubble tower. Also, tension infiltrometer theory (Chapter 69) requires that the porous medium be sufficiently dry (i.e., ci 50 to 100 cm) to ensure that the antecedent K (c) value is negligibly small relative to the minimum K (c) being measured (steady-flow analysis), and that the early-time transient-flow phase is long enough to provide relevant measurements (transient-flow analysis). Working with ‘‘wet’’ porous materials may consequently require preliminary measurement of ci (e.g., using a portable=handheld tensiometer—Chapter 71) to ensure that the material is ‘‘dry’’ enough for tension infiltrometer application.

6

The rate of water flow or discharge (Q) out of the infiltrometer and into the porous medium is measured by monitoring the rate of fall, R [LT1 ], of the water level in



1111

the infiltrometer reservoir. This can be accomplished using a handheld stopwatch and a scale attached to the reservoir, or via an automated pressure transducer– data logger system similar to that described by Ankeny (1992). For the transientflow analysis, start measurements as soon as the infiltrometer plate touches the contact sand, and collect frequent early-time measurements (e.g., after 0, 5, 10, 20, 40, . . . s). This ensures that sufficient flow data are collected to both identify contact sand effects and delineate the time period over which the transient analysis is applicable. For the steady-flow analysis, measurements do not need to be started immediately, as only the steady-flow rate is required. 7

For the transient-flow analysis, collect a small disturbed sample (only a few millimeters deep) from the infiltration surface immediately after infiltration is stopped (by quickly removing the infiltrometer and contact sand), and collect a second sample adjacent to the infiltration surface but just outside the wetted zone. Seal these samples in water vapor–tight containers and transport to the laboratory for water content determination (Chapter 70). The difference in water content between the wetted and unwetted (background) soil, Du ¼ u(c0 ) u(ci ), is required to determine K (c0 ) from the transient-flow analysis, and S(c0 ) from the steady-flow analysis (see Section 82.3). Small in situ TDR probes (Chapter 70) have also been used to obtain Du (e.g., Vogeler et al. 1996; Wang et al. 1998).

8

Calculate the near-saturated K (c) and capillarity parameters as indicated below.

82.3 ANALYSIS AND EXAMPLE CALCULATIONS 82.3.1 STEADY FLOW Under a constant matric head, c0 , the rate of fall of the water level in the infiltrometer reservoir, R [LT1 ], normally decreases with increasing time, and approaches a constant value, Rs , as the flow rate becomes quasisteady (Qs ). Quasisteady flow is usually assumed when effectively the same R-value (Rs ) is obtained over four or five consecutive R measurements (i.e., R constant within about 1%–5% with no trend evident—see Table 76.3, Chapter 76, for example Rs and Qs calculations; see also Section 82.4.3). Once steady flow is attained at the current matric head (e.g., c1 ), the water level in the bubble tower is adjusted to obtain the next desired matric head (e.g., c2 ), and flow is again monitored until the next steady-flow rate is attained. The method requires that at least two matric heads (c1 , c2 ) are set sequentially on the infiltrometer plate (by adjusting the water level in the bubble tower), and measurement of the corresponding steady-flow rates [Qs (c1 ), Qs (c2 )]. The matric heads should be set in ascending order of magnitude (i.e., c1 < c2 < c3 , . . .), with the first and most negative matric head (c1 ) no less than about 20 cm, and the last and largest matric head (cn ) at or near zero. It is recommended that 3–5 matric heads be used to provide adequate definition of the various near-saturated flow parameters (see also Section 82.4.4). Once Qs is obtained for the desired number of matric heads, the K(c0 ), a*(c0 ), S(c0 ), PD(c0 ), and NP(c0 ) relationships can be calculated as described below. Steady constant-head infiltration from the tension infiltrometer can be described using the Wooding (1968) ‘‘shallow pond’’ relationship written in the form Qs (c0 ) ¼ pa2 K(c0 ) þ

a f(c0 ) G

(82:1)



1112

where Qs (c0 ) [L3 T1 ] is the steady-flow rate out of the infiltrometer and into the porous medium when c ¼ c0 [L] on the infiltration surface, a [L] is the inside radius of the retaining ring, K(c0 ) [LT1 ] is the hydraulic conductivity at the infiltration surface, G ¼ 0:237 is a shape factor constant (Reynolds and Elrick 1991), and f(c0 ) [L2 T1 ] is the matric flux potential at the infiltration surface (see Equation 69.13, Chapter 69). Substituting the Gardner (1958) exponential K(c) relationship (Equation 69.14, Chapter 69) into Equation 82.1 and assuming K(ci ) K(c0 ) produces Qs (c0 ) ¼

a Kfs exp [a(c0 )c0 ] pa þ a*(c0 ) 2

(82:2)

where a*(c0 ) [L1 ] is the sorptive number (Equation 69.16, Chapter 69) and Kfs [LT1 ] is the field-saturated hydraulic conductivity (Equation 69.9, Chapter 69) of the porous medium. If it is assumed that a*(c0 ) is constant between any two adjacent matric heads set on the infiltrometer plate (c0 ), Equation 82.2 can be written in the form (Reynolds and Elrick 1991): " ln Qs (c0 ) ¼

a0x,xþ1 c0 þ ln

a

2

pa þ

!

a0x,xþ1 G

#

0 Kx,xþ1

x ¼ 1, 2, 3, . . . , n 1;

;

(82:3)

n2

where n is the total number of matric heads set on the infiltrometer plate, cx and cxþ1 are matric heads set in succession on the plate (first cx , then cxþ1 , with cx more negative than 0 cxþ1 ), and a0x,xþ1 and Kx,xþ1 are parameters related to a*(c0 ) and K(c0 ), respectively. Equation 82.3 describes a piece-wise linear plot of ln Qs (c0 ) versus c0 , from which a0x,xþ1 is determined from the piece-wise slope: a0x,xþ1 ¼

ln [Qs (cx )=Qs (cxþ1 )] (cx cxþ1 )

(82:4)

0 and Kx,xþ1 is determined from the piece-wise intercept:

0 Kx,xþ1 ¼

Ga0x,xþ1 Qs (cx ) a(1 þ Ga0x,xþ1 pa)[Qs (cx )=Qs (cxþ1 )]P

(82:5)

where P ¼ cx =(cx cxþ1 ). For the intermediate c0 values (c2 , c3 , . . . , cn1 ) the K(c0 ) values are calculated using K(cx ) ¼

0 0 [Kx1,x exp (a0x1,x cx ) þ Kx,xþ1 exp (a0x,xþ1 cx )] 2

(82:6)

while K(c) for the first c0 value (i.e., c1 ) is given by 0 K(c1 ) ¼ K1,2 exp (a01,2 c1 )

(82:7)

and K(c) for the last c0 value (i.e., cn ) is given by 0 exp (a0n1,n cn ) K(cn ) ¼ Kn1,n

(82:8)



1113

Once the K(c0 ) values are calculated, the corresponding a*(c0 ) values are determined by solving Equation 82.1 for f(c0 ), then substituting in Equation 69.16 (Chapter 69) to produce a*(c0 ) ¼

aK(c0 ) G[Q(c0 ) pa2 K(c0 )]

(82:9)

The S(c0 ), PD(c0 ), and NP(c0 ) relationships can then be calculated using Equation 69.17 through Equation 69.19, respectively, in Chapter 69. Note that the S(c0 ) calculation requires independent determination of Du ¼ u(c0 ) u(ci ), which for multiple head analyses, would require installation of an in situ water content measuring device (e.g., TDR probe—Chapter 70) under the infiltrometer plate to avoid interrupting flow and destructive sampling of the infiltration surface.

82.3.2 TRANSIENT FLOW Under a constant matric head, early-time transient flow out of the infiltrometer can be described using a two-term infiltration equation similar in form to that developed by Philip (1957) for one-dimensional infiltration: I(c0 ) ¼ E1 t1=2 þ E2 t

(82:10)

where I(c0 ) [L3 L2 ] is cumulative infiltration from the infiltrometer at c ¼ c0 , t [T] is time, and E1 and E2 are constants related to the porous medium hydraulic properties. Equation 82.10 is applied by differentiating with respect to t1=2 to produce (Vandervaere et al. 2000a) dI(c0 ) ¼ E1 þ 2E2 t1=2 dt1=2

(82:11)

which implies that a plot of dI(c0 )=dt1=2 versus t1=2 should be linear with a Y-axis intercept of E1 and a slope of 2E2 . Haverkamp et al. (1994) showed that for three-dimensional infiltration at short-to-medium times (not approaching steady state) E1 ¼ S(c0 )

(82:12)

and E2 ¼

2b vS(c0 )2 K(c0 ) þ a[u(c0 ) u(ci )] 3

(82:13)

where b 0:6 (Haverkamp et al. 1994), v ¼ 0:75 (Smettem et al. 1994), S(c0 ) [LT1=2 ] is porous medium sorptivity at c ¼ c0 , u(c0 ) is porous medium volumetric water content at c ¼ c0 , and u(ci ) is the antecedent volumetric water content at the antecedent pore water matric head, ci . Next, Equation 82.11 is discretized to produce (Vandervaere et al. 2000a) (Ijþ1 Ij ) dI(c0 ) DI(c0 ) ¼ E1 þ 2E2 *tj 1=2 ¼ 1=2 1=2 dt Dt (tjþ1 1=2 tj 1=2 )

(82:14)

where j ¼ 1, 2, 3, . . . (n 1), n is the number of DI(c0 )=Dt1=2 versus t1=2q data points, and ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi *tj 1=2 is calculated as the geometric mean of tj 1=2 and tjþ1 1=2 i:e:, *tj 1=2 ¼

tjþ1 1=2 tj 1=2 .

Then, simple linear regression is used to calculate both the magnitudes and standard errors of



1114

E1 and E2 , noting that calculation of K(c0 ) from E2 (Equation 82.13) requires independent measurement of Du ¼ u(c0 ) u(ci ). Vandervaere et al. (2000b) showed that determination of K(c0 ) from Equation 82.13 is most accurate when vE1 2 E2 < a[u(c0 ) u(ci )] 2

(82:15)

and determination of S(c0 ) from Equation 82.12 is most accurate when the porous medium sorptivity is close to a so-called optimal value given by S(c0 )opt

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi a[u(c0 ) u(ci )](2 b)K(c0 ) ¼ 3v

(82:16)

Both of these criteria minimize the interference (masking effect) of source geometry (finite disk radius) in the K(c0 ) and S(c0 ) calculations, and are best achieved by using large infiltrometer plates. Selecting relatively wet antecedent moisture conditions further assists in achieving the K(c0 ) criterion (i.e., Equation 82.15), whereas selecting relatively dry antecedent moisture conditions further assists in achieving the S(c0 ) criterion (i.e., Equation 82.16). Once K(c0 ) and S(c0 ) are calculated, a*(c0 ), PD(c0 ), and NP(c0 ) can be determined via Equation 69.17 through Equation 69.19, respectively, in Chapter 69. An important and elegant feature of Equation 82.14 is the ease with which the transient analysis can be validated; that is, a plot of DI(c0 )=Dt1=2 versus t1=2 must be linear with both E1 (intercept) and 2E2 (slope) positive. If these criteria are not met, the basic assumptions of the method (i.e., constant-head infiltration into rigid, homogeneous, isotropic porous medium with uniform antecedent water content) are seriously violated and the calculated E1 and E2 values are not likely to have physical meaning. Vandervaere et al. (2000a) also showed that Equation 82.14 provides a convenient and sensitive means for detecting loss of hydraulic contact between the infiltrometer disk=membrane and the porous medium, and for eliminating flow data perturbed by contact sand (see Section 82.3.3). The transient-flow approach is limited to varying degrees in its ability to apply a succession of c0 values to a single infiltration surface. This is because the transient analysis is valid only for early-time flow not approaching steady state, and the transition toward steady state can start very soon after the initiation of flow, depending on the c0 value, the infiltrometer radius, and the texture, structure, and antecedent water content of the porous medium. Applying a succession of c0 values to a single infiltration surface would also require installation of an in situ water content measuring device (e.g., TDR probe—Chapter 71) under the infiltrometer plate to obtain Du without interrupting flow or destructive sampling of the infiltration surface. Routine determination of K(c0 ), S(c0 ), a*(c0 ), PD(c0 ), and NP(c0 ) for a range of c0 values may consequently require the use of several adjacent infiltration surfaces (rather than a single infiltration surface), which may in turn increase the uncertainty of the results because of small-scale spatial variability.

82.3.3 ACCOUNTING

FOR

CONTACT SAND

A layer of ‘‘contact sand’’ (usually natural sand, uniform glass beads, or some other fine particulate material) should be placed under the tension infiltrometer to establish and maintain good hydraulic connection or linkage between the disk=membrane and the porous medium. This should be done regardless of whether the porous medium surface has been



1115

smoothed, leveled, or left undisturbed (e.g., Perroux and White 1988; Bagarello et al. 2001; Vandervaere 2002; Smettem and Smith 2002), and regardless of whether steady-state or transient analyses are used (Vandervaere 2002). The contact sand layer can introduce artifacts, however, which must be accounted for in tension infiltrometer analyses (Reynolds and Zebchuk 1996). For both steady state and transient flow, the saturated hydraulic conductivity of the contact sand, Kcs [LT1 ], must be greater than the maximum measured K(c0 ) of the porous medium. Also, the water-entry matric head of the contact sand, cw [L] (i.e., the matric head at which the contact sand spontaneously saturates from a dry state), must be smaller (more negative) than the minimum matric head set on the infiltrometer plate, c0 . If these criteria are not met, the hydraulic conductivity of the contact sand may fall below that of the porous medium at one or more of the set c0 values, and the sand layer may consequently restrict the flow and cause the infiltrometer measurements to be unrepresentative of the porous medium. Reynolds and Zebchuk (1996) recommend a fine glass bead material with cw ¼ 30 cm and Kcs ¼ 102 cm s1 , which should be adequate for c0 20 cm and for use on most agricultural soils. For steady flow, elevation and hydraulic head-loss effects induced by the contact sand layer can produce an important difference (or offset) between the matric head set on the infiltrometer plate (c0 ) and the matric head actually applied to the soil surface (cs ). If the contact sand layer is contained within a retaining ring (Figure 82.2), flow through the sand layer is steady, saturated, and rectilinear, which allows the cs value to be estimated accurately using Darcy’s law in the form (Reynolds and Zebchuk 1996): Q(c ) cs ¼ c0 þ 1 2 0 Tcs pa Kcs

(82:17)

where Q(c0 ) [L3 T1 ] is the flow rate out of the infiltrometer at c ¼ c0 , Kcs [LT1 ] is the saturated hydraulic conductivity of the contact sand layer, Tcs [L] is the mean thickness of the contact sand layer (Figure 82.1), and a [L] is the inside radius of the retaining ring. The value of cs thus depends on c0 , Q(c0 ), Kcs , Tcs , and a. Using numerical simulations and controlled laboratory experiments, Reynolds and Zebchuk (1996) determined that cs during steady flow can range between cs c0 and cs (c0 þ Tcs ) when Kcs Kfs . Given that Tcs ¼ 1:0 cm is often considered a practical minimum for field studies due to the roughness of undisturbed soil surfaces (Thony et al. 1991), the cs value can differ from c0 by 1.0 cm or more. Tension infiltrometer analyses based on steady flow should consequently use cs (calculated using Equation 82.17) rather than c0 to maintain accuracy when contact sand is used. The contact sand layer can also introduce substantial artifact effects into transient infiltration from the tension infiltrometer, although this may be difficult to detect from plots of cumulative infiltration versus time (e.g., Figure 82.3), or infiltration flux versus time (Vandervaere et al. 2000a). Plotting infiltration data according to Equation 82.14, however, reveals a distinct early-time negative slope region followed by a later-time positive slope region (Figure 82.4). The nonlinear negative slope region indicates contact sand effects, whereas the linear positive slope region represents infiltration into the porous medium (Vandervaere et al. 2000a; Vandervaere 2002). Thus, only the data in the linear positive slope region are used for the determination of K(c0 ) and S(c0 ). Inclusion of the contact sand–affected data, even when the sand layer is only a few millimeters thick, can result in large errors and=or physically meaningless (e.g., negative) K(c0 ) and S(c0 ) values (Vandervaere et al. 2000a). Equation 82.14 consequently provides an easy and convenient means for identifying contact sand



1116

Cumulative infiltration, I (cm)

4

3

2

1

0 0

200

400

600

800

1000

Time, t (s)

FIGURE 82.3. Example of cumulative tension infiltration, I, versus time, t, into a layer of air-dry contact sand placed over an unsaturated loamy sand soil. The matric head set on the infiltrometer membrane was c0 ¼ 3 cm. Contact sand thickness and saturated hydraulic conductivity were Tcs ¼ 1 cm and Kcs ¼ 1:0 102 cm s1 , respectively. The cumulative infiltration versus time data also appear in Table 82.2.

0.40 0.35

∆I /∆t 1/2 (cm s−1/2)

0.30 0.25 0.20 0.15 0.10 0.05 0.00 0

5

10

15

Geometric mean time,

20

*t 1/2

25

30

(s1/2)

FIGURE 82.4. The previous figure plotted as infiltration rate, DI=Dt 1=2 , versus geometric mean square root time, *t 1=2 (see Equation 82.14). The initial negative slope portion (first six data points) represents wetting and saturation of the contact sand, and the linear positive slope portion represents infiltration into the unsaturated loamy sand soil. The solid straight line is a least squares regression through the linear data (final eight data points), from which near-saturated soil sorptivity, S(c0 ), and near-saturated soil hydraulic conductivity, K (c0 ), can be calculated using Equation 82.12 and Equation 82.13, respectively. In this example, S(c0 ) ¼ 4:03 102 cm s1=2 and K (c0 ) ¼ 1:08 103 cm s1 , with c0 ¼ 3 cm. The infiltration rate versus geometric mean square root time data also appear in Table 82.2.



1117

effects, and for eliminating these data from the transient analysis. As with the steady-state analyses, cs should be used in place of c0 in Equation 82.12 and Equation 82.13 to account for elevation and head-loss effects caused by the contact sand layer, especially when the layer must be thick due to surface roughness. This is somewhat less straightforward for transient analyses, however, because Q(c0 ) changes with time, and this may in turn cause cs to change with time. For example, Figure 82.5 illustrates the predicted change in cs with Kcs and time (using Equation 82.17) for the same transient infiltration data used to generate Figure 82.3 and Figure 82.4, but after the contact sand layer was fully wetted (i.e., for times greater than 150 s). Note that when Kcs is an order of magnitude greater than the K(c) of the porous medium, cs is effectively constant (as required by the transient analysis), but offset from c0 by þ0:8 cm; hence, use of c0 instead of cs in the flow parameter calculations would introduce substantial error. On the other hand, when Kcs is equal to the K(c) of the porous medium, cs not only is offset substantially from c0 (by about 1:6 to 0:7 cm), but also changes with time (by about 20%), which violates the constant-head requirement of the transient analysis. Successful application of the transient analysis with contact sand consequently requires both accounting for the offset between cs and c0 (via Equation 82.17) and ensuring that Kcs is sufficiently greater than the K(c) of the porous medium to maintain cs constant. Another important consequence of using contact sand is that cs can be slightly positive when c0 is close to zero. Given that the tension infiltrometer relationships apply only for cs 0, then either c0 must be adjusted to prevent cs > 0 (e.g., by accounting for the thickness of the contact sand layer in the bubble tower calibration and=or setting the maximum c0 < 0),

−2.0

Pressure head,

(cm)

−2.5 −3.0

−3.5 −4.0 −4.5

−5.0 100

200

300

400 500 600 Geometric mean time, t (s)

700

800

900

FIGURE 82.5. Impact of transient, constant-head tension infiltration, and contact sand on the matric head at the infiltration surface, cs . The cs values were calculated using Equation 82.17 and the infiltration rate data in Figure 82.4 (and Table 82.2) for a loamy sand soil. The heavy solid line represents the matric head set on the infiltrometer membrane, c0 ¼ 3 cm; the circles give cs when Kcs ¼ 1:0 102 cm s1 ; and the triangles give cs when Kcs ¼ 1:0 103 cm s1 . The K (c0 ) of the soil is 1:08 103 cm s1 , and the contact sand thickness is Tcs ¼ 1 cm.

0.90 1.10 1.55 2.00 2.95 3.95 5.40

16 11 6 4 2 1 0

15.1667 10.2037 5.2871 3.3704 1.5463 0.7315 0.00007

cs c (cm) 0.04043 0.06975 0.13299 0.21307 0.35826 0.42748

a0 d (cm1 ) 9.2562E04 1.7973E03 3.5281E03 5.5542E03 8.0451E03 8.7914E03

K 0e (cm s1 )

j

i

h

g

f

e

d

c

b

a

Pressure head set on infiltrometer disk or membrane. Calculated from steady rate of fall of water level, Rs (as in Table 76.3, Chapter 76). Calculated using Equation 82.17. Calculated using Equation 82.4. Calculated using Equation 82.5. Calculated using Equation 82.6, Equation 82.7, or Equation 82.8 depending on c0 . Calculated using Equation 82.9. Calculated using Equation 69.17, Chapter 69. Calculated using Equation 69.18, Chapter 69. Calculated using Equation 69.19, Chapter 69.

5.01E04 7.47E04 1.49E03 2.48E03 4.31E03 6.31E03 8.79E03

K (c s )f (cm s1 ) 0.040 0.054 0.097 0.167 0.272 0.390 0.427

a*(cs )g (cm1 ) 0.22 0.24 0.26 0.28 0.31 0.33 0.35

u(cs ) (cm3 cm3 ) 5.2008E02 5.9479E02 6.7095E02 6.9660E02 7.7733E02 8.2204E02 9.6679E02

S(cs )h (cm s1=2 )

0.060 0.080 0.14 0.25 0.40 0.58 0.64

PD(c s )i (mm)

1:60E þ 08 7:64E þ 07 1:47E þ 07 2:70E þ 06 6:70E þ 05 2:32E þ 05 2:25E þ 05

NP (cs )j (p m2 )

1118

Note: Air-pore water interfacial surface tension, s ¼ 72:75 g s2 . Assumed pore water density, r ¼ 0:9982 g cm3 . Assumed pore water viscosity, m ¼ 1:002 g cm1 s1 . Acceleration due to gravity, g ¼ 980:621 cm s2 .

Qs b (cm3 s1 )

c 0a (cm)

Retaining ring inside radius, a ¼ 12:5 cm Contact sand: thickness, T cs ¼ 1 cm; saturated hydraulic conductivity, Kcs ¼ 102 cm s1 Porous material type: loamy sand soil b ¼ 0:6; v ¼ 0:75; g ¼ 1:818; u(ci ) ¼ 0:10

TABLE 82.1 Example K (c), S(c), a*(c), PD(c), and NP (c) Calculations for the Tension Infiltrometer Method, Steady-Flow Analysis, Overhead Reservoir System (Figure 82.2). The Tension Infiltrometer Data Collection Sheet Is Similar in Format to Table 76.3, Chapter 76





1119

or alternative relationships must be used. Reynolds and Zebchuk (1996) provide approximate steady-flow expressions for determining K(cs ) from tension infiltrometer data when cs > 0. Example data sheets and calculations for the steady-state and transient-flow methods are given in Table 82.1 and Table 82.2, respectively. TABLE 82.2 Example Data Sheet and K (c0 ), S(c0 ), a*(c0 ), PD(c 0 ), and NP (c 0 ) Calculations for the Tension Infiltrometer Method, Transient-Flow Analysis, Overhead Reservoir System (see Figure 82.2) Reservoir inside radius, r ¼ 5 cm; retaining ring or disk radius, a ¼ 10 cm Initial height of water in reservoir, h0 ¼ 30 cm Pressure head on infiltrometer disk or membrane, c0 ¼ 3 cm Contact sand: thickness, Tcs ¼ 1 cm; saturated hydraulic conductivity, Kcs ¼ 102 cm s1 Porous material type: loamy sand; u(c0 ) ¼ 0:35; u(ci ) ¼ 0:10 b ¼ 0:6; v ¼ 0:75; g ¼ 1:818 Reservoir scale Time, reading, tj (s) Lj (cm)

Flow rate, Qj a (cm3 s1 )

Cumulative infiltration, lj b 3 (cm cm2 )

Square root time 1=2 tj (s1=2 )

0f 5 10 20 40 70 120 200 300 400 500 600 700 800 900

50.2655 16.6096 8.8085 4.4809 1.8850 1.0681 0.8050 0.7219 0.6452 0.6237 0.5807 0.5450 0.5435 0.5341

0 0.800 1.064 1.345 1.630 1.810 1.980 2.185 2.415 2.620 2.819 3.004 3.177 3.350 3.520

0 2.24 3.16 4.47 6.32 8.37 10.95 14.14 17.32 20.00 22.36 24.49 26.46 28.28 30.00

0 3.20 4.26 5.38 6.52 7.24 7.92 8.74 9.66 10.48 11.27 12.01 12.71 13.40 14.08

Geometric mean time, tjGM c (s)

Geometric mean square root time, 1=2 *tj d (s1=2 )

Dlj =Dtj e (cm s1=2 )

0 7.07 14.14 28.28 52.92 91.65 154.92 244.95 346.41 447.21 547.72 648.07 748.33 848.53

0 2.66 3.76 5.32 7.27 9.57 12.45 15.65 18.61 21.15 23.40 25.46 27.36 29.13

0.3578 0.2854 0.2141 0.1540 0.0881 0.0657 0.0643 0.0723 0.0766 0.0841 0.0866 0.0884 0.0947 0.0991

1=2

Note: Values for s, r, m, and g are given in Table 82.1. The accuracy criteria for hydraulic conductivity (Equation 82.15) and sorptivity (Equation 82.16) were both met in this example. a b c d e f

Qj ¼ [(Ljþ1 Lj )=(tjþ1 tj )](pr 2 ); j ¼ 1, 2, 3, . . . , n 1; n ¼ 15 ¼ number of measurements. Ij ¼ Lj (r 2 =a2 ). tj GM ¼ [tjþ1 tj ]1=2 . *tj 1=2 ¼ [tjþ1 1=2 tj 1=2 ]1=2 . DIj =Dtj 1=2 ¼ (IJþ1 Ij )=(tjþ1 1=2 tj 1=2 ). Regular font refers to the initial wetting and saturation of the contact sand layer; bold font refers to wetting of the unsaturated porous medium under the contact sand.

Plot DIj =Dtj 1=2 versus *tj 1=2 and determine the Y-axis intercept and slope for the data in bold (Figure 82.4): Intercept ¼ 4:0293 102 cm s1=2 ; Slope ¼ 1:9849 103 cm s1 . Sorptivity, S(c0 ) ¼ Intercept ¼ 4:03 102 cm s1=2 (Equation 82.12). (Equation 82.13). Hydraulic conductivity, K(c0 ) ¼ 1:08 103 cm s1 Sorptive number, a*(c0 ) ¼ 0:303 cm1 (Equation 69.17, Chapter 69). Flow-weighted mean pore diameter, PD(c0 ) ¼ 0:451 mm (Equation 69.18, Chapter 69). Number of flow-weighted mean pores, NP(c0 ) ¼ 1.09 105 pores m2 (Equation 69.19, Chapter 69).



1120

82.4 COMMENTS 82.4.1 PURPOSE OF RETAINING RING

AND

GUARD CLOTH

The purpose of the retaining ring is to (i) ensure that the flow cross-section is circular and that flow within the contact sand is vertical and rectilinear (required by theory); (ii) prevent development of ‘‘horizontal wicks’’ caused by slumping or spillage of contact sand beyond the edge of the infiltrometer plate; and (iii) prevent possible horizontal leakage of free water from under the infiltrometer plate when zero or slightly positive matric heads occur at the infiltration surface. Note that the small depth of ring insertion ( 0:5 cm) causes negligible disturbance and has no appreciable effect on flow within the porous medium (Reynolds and Elrick 1991; Reynolds and Zebchuk 1996). The primary purpose of the guard cloth is to prevent the contact sand from infilling cracks, worm holes, and other macropores present in the infiltration surface (Figure 82.1). Significant infilling of macropores can change the hydraulic properties of the porous medium, as well as increase greatly the amount of contact sand required. The guard cloth can be omitted if macropores are not present, although the cloth provides a very convenient and effective means for reclaiming the contact sand from the infiltration surface for subsequent reuse at other sites (after air-drying and sieving to remove entrained debris). In loose, single-grain materials (e.g., loose sandy soils), the contact sand, guard cloth, and retaining ring may not be necessary (i.e., the infiltrometer plate can be placed directly on the infiltration surface), although some work suggests that contact sand should always be used (Bagarello et al. 2001).

82.4.2 DESIGN, SATURATION, LEAK TESTING,

AND

CALIBRATION

The tension infiltrometer consists essentially of a porous disk or membrane covering a conical or hemispherical cavity in the underside of a transparent plastic plate (Figure 82.2 and Figure 82.6). Water is supplied to the infiltrometer via the water supply tube. Tension is set on the porous disk=membrane via the bubble tower. The seal between the porous disk or membrane and the infiltrometer plate (often made using epoxy, resin glue, or a tight-fitting O-ring) must be airtight or it will leak air when tension is applied. The porous disk or membrane must be hydrophilic (water wettable) and have an airentry matric head that is more negative than the minimum matric head set by the bubble tower. If the air-entry matric head is exceeded, air will leak through the disk=membrane and appear as a stream of bubbles entering the cavity above the disk=membrane. Porous disks may be constructed of porous metal (stainless steel, brass), sintered glass, high-flow ceramic, or porous plastic that has been treated to make it hydrophilic. Membranes are usually made of fine-mesh (270 mesh or smaller) nylon cloth or metal sieve screen. The wettability of the disk or membrane can often be enhanced by first cleaning it with a solvent (e.g., ethanol, dithionite–citrate solution, dilute bleach solution) to remove hydrophobic oils, oxides, organic matter, etc., and then treating it with a wetting agent (e.g., a solution of isopropyl alcohol, distilled water, and commercial surfactant). Porous disks and membranes should have an air-entry matric head 20 cm and a saturated hydraulic conductivity of 101 to 103 cm s1 . The porous disk or membrane is saturated by submerging the entire infiltrometer plate in a water bath containing several centimeters of deaired, temperature-equilibrated water, and then drawing the water up through the disk=membrane (by sucking through a oneway valve inserted into the top of the water supply tube) until the cavity and supply tube are full (Figure 82.6). The bubble tower should also be filled to the maximum level with water (discussed below). Leave the infiltrometer standing in the water bath for a couple of days



1121

Vacuum One-way valve

Air entry

Bubble tower air tube (≥0.25 cm i.d.) Cap

Water supply tube

O-ring WL 1

WL 2

P

Syringe

Infiltrometer plate

Z Contact sand layer 1=

Tension table

0

P −Z 1

Measuring scale

Hanging water column

FIGURE 82.6. Calibration of a tension infiltrometer using a vacuum source and a tension table– hanging water column system. (Adapted from Reynolds, W.D., in M.R. Carter (Ed.), Soil Sampling and Methods of Analysis, Canadian Society of Soil Science, Lewis Publishers, Boca Raton, Florida 1993. With permission.)

periodically jarring the plate against the wall of the water bath to remove air that may be entrapped in the disk=membrane. Test the infiltrometer by removing it from the water bath and setting it on several layers of dry paper towel. Air bubbles should rise rapidly from the air tube and up into the supply tube as water flows out of the infiltrometer and into the paper towel. If air bubbles rise from the disk=membrane itself, then the infiltrometer is not completely saturated or contains a hole=large pore or there is a break in the seal between the disk=membrane and the plate. Repeating the above procedures should stop the leak if incomplete saturation is the problem. The leak will persist if it is due to a hole in the disk=membrane, or to a break in the seal. Such holes and breaks can usually be plugged using a couple of drops of epoxy=resin glue,


1122


although membranes with several holes may need to be replaced. To prevent damage and maintain saturation during transport, place the infiltrometer on padding (e.g., sponge, waterresistant carpet underlay, etc.) in a flat-bottomed pail with the plate submerged below several centimeters of water. For long-term storage, the infiltrometer should be air-dried to prevent algal growth, corrosion, etc. The bubble tower should be calibrated to ensure that the matric heads applied to the infiltration surface are correct and accurate. This can be accomplished using a tension table–hanging water column arrangement (Figure 82.6). Adjust the water level in the bubble tower (using a syringe) so that it is about 2 cm below the top of the second air tube (the ‘‘second tube’’ is identified in Figure 82.2). Lower the bubble tower air tube until its bottom end is about P ¼ (c1 þ Z) cm below the water surface, where c1 is the minimum (most negative) matric head to be set on the disk=membrane and Z is the height of the bubble point above the disk=membrane (Figure 82.6). Place the saturated infiltrometer on a prewetted tension table= Buchner funnel, which is set at zero matric head (Figure 82.6). Attach a one-way air valve to the supply tube and gently draw air through the bubble tower air tube and into the water supply tube by applying a vacuum (Figure 82.6). Using the syringe, ‘‘fine-tune’’ the water level in the bubble tower (by adding or removing water) until the minimum (most negative) matric head to be set on the disk=membrane (c1 ) is established, as indicated by the water elevation in the hanging water column (i.e., c1 ¼ P Z in Figure 82.6). Mark and label the water level on the side of the bubble tower (WL 1 in Figure 82.6), then remove water to find and mark the next desired matric head (WL 2 in Figure 82.6). Repeat this procedure until all desired matric heads have been marked. Note that if the bubble tower and air tubes are highly uniform in diameter, the change in matric head is equal to the change in water level in the bubble tower, which allows the succeeding water levels (i.e., WL 2, WL 3, etc.) to be obtained by simply using a ruler to measure downward from WL 1. A matric head sequence of c1 ¼ 15 cm, c2 ¼ 10 cm, c3 ¼ 5 cm, c4 ¼ 3 cm, c5 ¼ 1 cm, c6 ¼ 0 cm is often used.

82.4.3 EQUILIBRATION TIMES, RESERVOIR REFILLING,

AND

EVAPORATIVE LOSSES

Deciding when steady flow (Rs ) is attained is somewhat arbitrary (as seen in the procedures section), and it will depend to some extent on the experience of the operator. Nevertheless, steady flow is usually reached within 15–120 min for the first (and most negative) matric head set on the infiltrometer; and the succeeding (less negative) matric heads usually require less time. At lower matric heads (say 20 to 5 cm), the steady-flow rates are often slow enough to require use of the smaller of the two concentric reservoirs in order to obtain adequate measurement accuracy. When the small reservoir is used, it will require periodic refilling. This is best accomplished at the end of a measurement for a given matric head. Adjust the reservoir valve to allow slow flow of water from the large reservoir into the small reservoir. At the same time, slowly lower the water level in the bubble tower (by withdrawing water via the syringe) to establish the next matric head. This technique accomplishes refilling, while at the same time ensuring that the change from one matric head to the next is monotonic, thereby preventing possible hysteresis effects. Details on switching between the two concentric reservoirs are given in the Soilmoisture Equipment Corp. procedure manual for the Guelph permeameter. An infiltrometer that is bubbling slowly is susceptible to solar heating effects, and the procedures for minimizing these adverse effects (as well as for general ‘‘troubleshooting’’) are similar to those given in Section 76.2. The base of the infiltrometer should be covered with a plastic sheet or transparent lid to prevent evaporative water loss from the annulus



1123

between the retaining ring and infiltrometer plate (Figure 82.2), as such losses can be sufficient to perturb the measurements when flow is slow and atmospheric evaporative demand is high.

82.4.4 FACTORS AFFECTING ACCURACY AND K(c) RANGE As discussed in Section 82.3, the steady-flow approach determines K(c0 ) values by fitting an exponential curve segment between successive (c0 , Qs ) data pairs. Since K(c) relations are generally exponential over only small ranges, then the accuracy of the K(c0 ), a*(c0 ), S(c0 ), PD(c0 ), and NP(c0 ) values tends to increase with the number of matric heads (and thereby the number of exponential curve segments) used within a particular matric head range; and as a consequence, the steeper the K(c) curve of the porous medium, the greater the number of matric heads required to maintain flow parameter accuracy (see Reynolds and Elrick 1991). Note as well that Reynolds and Zebchuk (1996) provide steady-flow expressions for the determination of K(c0 ) from a single c0 value when a priori knowledge of a*(c0 ) is available. The range of K(c0 ) values that can be measured conveniently with 10–20 cm diameter tension infiltrometers appears to be on the order of 102 to 106 cm s1 . Excessive restriction of flow by the contact sand can occur at K(c0 ) values greater than 0:008---0:01 cm s1 (Reynolds et al. 2000). For the steady-flow analysis, impractically slow flows (and correspondingly long equilibration times) can occur when K(c0 ) < 106 cm s1 . Both the steady-flow and transient-flow analyses will occasionally yield unrealistic or invalid (e.g., negative) parameter values. This is usually caused by porous medium heterogeneities or strong water content gradients. The steady-flow method can also produce unrealistic results if the matric head set on the infiltrometer is changed to the next head before steady flow is achieved. Sometimes, the tension infiltrometer measurement can be salvaged by eliminating obviously aberrant data points, such as a steady-flow rate measurement that decreased with increasing matric head, or a transient infiltration rate measurement that deviates from linearity with t1=2 . Other factors affecting the accuracy of tension infiltrometer measurements include macrostructure collapse under the infiltrometer during the infiltration measurement, and inadequate or changing hydraulic connection between the infiltrometer and the infiltration surface. Macrostructure collapse is caused by the weight of the infiltrometer combined with a decline in porous medium strength as the porous medium wets up. Inadequate or changing hydraulic connection is usually caused by wind-induced vibration of the infiltrometer, and by the decreasing weight of the infiltrometer as the water empties out of the reservoir. Both of these problems can be reduced by supporting the infiltrometer with a large tripod (as recommended above), which clamps solidly to the reservoir.

82.4.5 SENSITIVITY TO CALIBRATION

AND

SLOPE

Near-saturated soil hydraulic properties are often extremely sensitive to small changes in pore water matric head, e.g., K(c) often changes by 2–3 orders of magnitude over the range, 15 cm c 0 (Reynolds and Elrick 1991; Thony et al. 1991; Reynolds et al. 1995). It is consequently important that the infiltrometer is accurately calibrated (using a precise calibration method such as in Section 82.4.2); and that the infiltrometer plate is level when making a measurement (gravity effects cause the ‘‘high’’ and ‘‘low’’ sides of a sloping plate



1124

to have different c0 values than the value set by the bubble tower). Placement of contact sand is the best approach for leveling (and smoothing) rough or undulating surfaces (Figure 82.1), as it also ensures good hydraulic contact over the entire infiltration surface, and it does not alter surface hydraulic properties. For continuously sloping surfaces (e.g., hillslopes), it is generally recommended that the infiltrometer is placed on a level bench cut into the slope. It has recently been found, however, that infiltrometers as large as 20 cm diameter can still yield valid results when placed without leveling on slopes as great as 20% (Bodhinayake et al. 2004). Evidently, the difference in matric head (and thereby infiltration rate) between the upslope and downslope sides of the infiltrometer tends to be compensating as long as the infiltration surface has reasonably uniform hydraulic properties.

82.4.6 ALTERNATIVE DESIGNS AND ANALYSES Additional information concerning the tension infiltrometer method, including several alternative infiltrometer designs and analyses, may be found in Smettem and Clothier (1989), Ankeny (1992), Elrick and Reynolds (1992), White et al. (1992), Vandervaere et al. (2000b), Clothier and Scotter (2002), and Smettem and Smith (2002). Although many of these designs and analyses have specific advantages, the apparatus presented here (Figure 82.2) tends to be the most versatile, and single disk analyses usually yield the most stable, accurate, and repeatable results (e.g., Hussen and Warrick 1993). Commercial manufacturers of tension infiltrometers include Soil Measurement Systems, Tucson, Arizona; and Soilmoisture Equipment Corp., Goleta, California.

82.5 STRENGTHS AND WEAKNESSES OF TENSION INFILTROMETER METHODS An important strength of the tension infiltrometer method is that the apparatus is simple, inexpensive, portable, easily applied in both field and laboratory studies, and requires only small volumes of water. Hence, laboratory or greenhouse studies, detailed investigations of spatial variability, field studies in areas with difficult access, and large-scale surveys are more feasible with this method than with many other methods. The apparatus also does not require soil disturbance such as augering a well or deep insertion of a ring, and can thus provide highly plausible estimates of the hydraulic properties of fragile aggregates and soil macropores. Perhaps the main theoretical strength of the method is its ability to determine a number of important water transmission parameters (e.g., K(c0 ), S(c0 ), a*(c0 ), PD(c0 ), NP(c0 )) in the near-saturated range, 0:20 m c0 0, where both parameter values and water=solute movement can change dramatically with even small changes in c0 . The tension infiltrometer method consequently has the ability to relate ‘‘macropore’’ and ‘‘matrix pore’’ flow parameters to changes in soil condition or soil management. For example, changes in PD(c0 ) have been used to quantify the effects of macrostructure collapse and macropore infilling (White et al. 1992), soil cracking (Thony et al. 1991), tillage practices (Sauer et al. 1990; White et al. 1992; Reynolds et al. 1995), root growth (White et al. 1992), and sediment erosion-deposition (White et al. 1992). Measurements of near-saturated K(c0 ) and S(c0 ) have been used to quantify changes in water transmission as a result of different tillage procedures (Sauer et al. 1990; Reynolds et al. 1995), faunal activity (Clothier et al. 1985), soil structural changes during the growing season (Messing and Jarvis 1993), soil textural changes (Jarvis and Messing 1995), wheel trafficking (Ankeny et al. 1991), and development



1125

of soil hydrophobicity (Clothier et al. 1996, 2000; Hallett et al. 2004). Tension infiltrometers have also been used to characterize near-saturated mobile–immobile soil water contents (Clothier et al. 1992; Angulo-Jaramillo et al. 1997), solute transport characteristics (Jaynes et al. 1995; Clothier et al. 1996; Vogeler et al. 1996), solute sorption isotherms (Clothier et al. 1995), and the hydraulic properties of surface crusts (Vandervaere et al. 1997). Perhaps the primary weaknesses of the tension infiltrometer method are the potential difficulties associated with working on sloped surfaces, and the need to use contact sand to establish and maintain good hydraulic connection between the infiltrometer and the porous medium surface. Tension infiltrometer measurements on hillslopes may require cutting a level bench so that the applied matric head is the same at all points on the infiltration surface, although some recent work suggests that benches may not be necessary for slopes 20% (see Section 82.4.5). The requirement for cutting a bench would obviously preclude measuring intact (undisturbed) hillslope surfaces. The contact sand must meet specific and somewhat restrictive performance criteria, and the presence of contact sand must be accounted for in the data analysis procedures. Specifically, the saturated hydraulic conductivity, Kcs , and water-entry matric head, cw , of the contact sand must be such that the sand never restricts flow into the porous medium. Recommended values are Kcs 102 cm s1 , which is greater than the Kfs of most unstructured agricultural soils, and cw 20 cm, which is less than the minimum matric head set on most infiltrometer plates. The Kcs and cw values should also be stable with a narrow standard deviation to minimize contact sand–induced variations. The contact material should be strongly hydrophilic and single grain with a narrow particle-size distribution so that it levels easily and readily establishes good hydraulic connection. The material should also be easily obtained, inexpensive, and reusable. Most natural soil materials cannot meet all of the above performance criteria; however, the glass bead material proposed by Reynolds and Zebchuk (1996) appears to be a good choice according to field tests conducted by Bagarello et al. (2000, 2001). The steady-state and transient tension infiltrometer analyses also have strengths and weaknesses. Important strengths of the steady-flow, multiple-head approach (Equation 82.4 through Equation 82.9) include well-established and tested theory, robustness, provision of measurements at several matric heads on a single infiltration surface, relatively large (and thereby more representative) sample volumes, and the ability to accurately account for the effects of a contact sand layer. Weaknesses of the steady-flow analysis include potentially long equilibration times and potentially greater susceptibility to error associated with porous medium heterogeneities (e.g., layering) and nonuniform water contents (due to large sample volume). Some recent simulation studies suggest, however, that the steady-flow approach may be far less sensitive to near-surface layering than previously supposed (Smettem and Smith 2002). The main strengths of the transient analysis (Equation 82.12 through Equation 82.16) include shorter measurement-times (because steady flow is not required), simple but effective procedures for determining valid results and for eliminating flow perturbations caused by contact sand, and straightforward use of linear regression to obtain the hydraulic conductivity and sorptivity values. On the other hand, important weaknesses of the transient analysis include imprecise knowledge of the b parameter and the matric head at the porous medium–contact sand interface (cs ), and limited ability to obtain measurements on a single infiltration surface for a sequence of c0 values. Generally speaking, steady-flow analyses are more effective for determining K(c0 ), whereas transient analyses are more effective for determining S(c0 ).



1126

REFERENCES Angulo-Jaramillo, R., Moreno, F., Clothier, B.E., Thony, J.L., Vachaud, G., Fernandez-Boy, E., and Cayuela, J.A. 1997. Seasonal variation of hydraulic properties of soils measured using a tension disk infiltrometer. Soil Sci. Soc. Am. J. 61: 27–32. Ankeny, M.D. 1992. Methods and theory of unconfined infiltration measurements. In G.C. Topp, W.D. Reynolds, and R.E. Green, Eds., Advances in Measurement of Soil Physical Properties: Bringing Theory into Practice, SSSA Special Publication No. 30, Soil Science Society of America, Madison, WI, pp. 123–141. Ankeny, M.D., Ahmed, M., Kaspar, T.C., and Horton, R. 1991. Simple field method for determining unsaturated hydraulic conductivity. Soil Sci. Soc. Am. J. 55: 467–470.

Clothier, B.E. and Scotter, D.R. 2002. Unsaturated water transmission parameters obtained from infiltration. In J.H. Dane and G.C. Topp, Eds., Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 879–898. Clothier, B.E., Scotter, D.R., and Harper, E. 1985. Three-dimensional infiltration and trickle irrigation. Trans. Am. Soc. Agric. Eng. 28: 497–501. Clothier, B.E., Vogeler, I., and Magesan, G.N. 2000. The breakdown of water repellency and solute transport through a hydrophobic soil. J. Hydrol. 231: 255–264.

Bagarello, V., Iovino, M., and Tusa, G. 2000. Factors affecting measurement of the near saturated soil hydraulic conductivity. Soil Sci. Soc. Am. J. 64: 1203–1210.

Elrick, D.E. and Reynolds, W.D. 1992. Infiltration from constant-head well permeameters and infiltrometers. In G.C. Topp, W.D. Reynolds, and R.E. Green, Eds., Advances in Measurement of Soil Physical Properties: Bringing Theory into Practice, SSSA Special Publication No. 30, Soil Science Society of America, Inc., Madison, WI, pp. 1–24.

Bagarello, V., Iovino, M., and Tusa, G. 2001. Effect of contact material on tension infiltrometer measurements. Trans. Am. Soc. Agric. Eng. 44: 911–916.

Gardner, W.R. 1958. Some steady-state solutions of the unsaturated moisture flow equation with application to evaporation from a water table. Soil Sci. 85: 228–232.

Bodhinayake, W., Si, B.C., and Noborio, K. 2004. Determination of hydraulic properties in sloping landscapes from tension and double-ring infiltrometers. Vadose Zone J. 3: 964–970.

Hallett, P.D., Nunan, N., Douglas, J.T., and Young, I.M. 2004. Millimeter-scale spatial variability in soil water sorptivity. Soil Sci. Soc. Am. J. 68: 352–358.

Clothier, B.E. 2000. Infiltration. In K.A. Smith and C.E. Mullins, Eds., Soil and Environmental Analysis: Physical Methods. Marcel Dekker, Inc., New York, pp. 239–280.

Haverkamp, R., Ross, P.J., Smettem, K.R.P., and Parlange, J.-Y. 1994. Three-dimensional analysis of infiltration from the disc infiltrometer: 2. Physically based infiltration equation. Water Resour. Res. 30: 2931–2935.

Clothier, B.E., Green, S.R., and Katou, H. 1995. Multidimensional infiltration: points, furrows, basins, wells, and disks. Soil Sci. Soc. Am. J. 59: 286–292.

Hussen, A.A. and Warrick, A.W. 1993. Alternative analyses of hydraulic data from disc tension infiltrometers. Wat. Resour. Res. 29: 4103–4108.

Clothier, B.E., Kirkham, M.B., and McLean, J.E. 1992. In situ measurement of the effective transport volume for solute moving through soil. Soil Sci. Soc. Am. J. 56: 733–736.

Jarvis, N.J. and Messing, I. 1995. Near-saturated hydraulic conductivity in soils of contrasting texture measured by tension infiltrometers. Soil Sci. Soc. Am. J. 59: 27–34.

Clothier, B.E., Magesan, G.N., Heng, L., and Vogeler, I. 1996. In situ measurement of the solute adsorption isotherm using a disc permeameter. Water Resour. Res. 32: 771–778.

Jaynes, D.B., Logsdon, S.D., and Horton, R. 1995. Field method for measuring mobile=immobile water content and solute transfer rate coefficient. Soil Sci. Soc. Am. J. 59: 352–356.


Unsaturated Hydraulic Properties: Field Tension Infiltrometer Messing, I. and Jarvis, N.J. 1993. Temporal variation in the hydraulic conductivity of a tilled clay soil as measured by tension infiltrometers. J. Soil Sci. 44: 11–24. Perroux, K.M. and White, I. 1988. Designs for disc permeameters. Soil Sci. Soc. Am. J. 52: 1205–1215. Philip, J.R. 1957. The theory of infiltration. 4. Sorptivity and algebraic infiltration equations. Soil Sci. 84: 257–264. Reynolds, W.D. 1993. Saturated hydraulic conductivity: Field measurement. In M.R. Carter, Ed., Soil Sampling and Methods of Analysis. Canadian Society of Soil Science. Lewis Publishers, Boca Raton, FL, pp. 599–613. Reynolds, W.D., Bowman, B.T., Brunke, R.R., Drury, C.F., and Tan, C.S. 2000. Comparison of tension infiltrometer, pressure infiltrometer, and soil core estimates of saturated hydraulic conductivity. Soil Sci. Soc. Am. J. 64: 478–484. Reynolds, W.D. and Elrick, D.E. 1991. Determination of hydraulic conductivity using a tension infiltrometer. Soil Sci. Soc. Am. J. 55: 633–639. Reynolds, W.D., Gregorich, E.G., and Curnoe, W.E. 1995. Characterization of water transmission properties in tilled and untilled soils using tension infiltrometers. Soil Till. Res. 33: 117–131. Reynolds, W.D. and Zebchuk, W.D. 1996. Use of contact material in tension infiltrometer measurements. Soil Tech. 9: 141–159. Sauer, T.J., Clothier, B.E., and Daniel, T.C. 1990. Surface measurements of the hydraulic properties of a tilled and untilled soil. Soil Till. Res. 15: 359–369. Smettem, K.R.J. and Clothier, B.E. 1989. Measuring unsaturated sorptivity and hydraulic conductivity using multiple disc permeameters. J. Soil Sci. 40: 563–568.

1127

Applications. Water Resources Monograph 15, American Geophysical Union, Washington, DC, pp. 135–157. Thony, J.-L., Vachaud, G., Clothier, B.E., and Angulo-Jaramillo, R. 1991. Field measurement of the hydraulic properties of soil. Soil Tech. 4: 111–123. Vandervaere, J.-P. 2002. Early-time observations. In J.H. Dane and G.C. Topp, Eds., Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 889–894. Vandervaere, J.-P., Peugeot, C., Vauclin, M., Angulo-Jaramillo, R., and Lebel, T. 1997. Estimating hydraulic conductivity of crusted soils using disc infiltrometers and minitensiometers. J. Hydrol. 188: 203–223. Vandervaere, J.-P., Vauclin, M., and Elrick, D.E. 2000a. Transient flow from tension infiltrometers: I. The two-parameter equation. Soil Sci. Soc. Am. J. 64: 1263–1272. Vandervaere, J.-P., Vauclin, M., and Elrick, D.E. 2000b. Transient flow from tension infiltrometers: II. Four methods to determine sorptivity and conductivity. Soil Sci. Soc. Am. J. 64: 1272–1284. Vogeler, I., Clothier, B.E., Green, S.R., Scotter, D.R., and Tillman, R.W. 1996. Characterizing water and solute movement by time domain reflectometry and disk permeametry. Soil Sci. Soc. Am. J. 60: 5–12. Wang, D., Yates, S.R., and Ernst, F.F. 1998. Determining soil hydraulic properties using tension infiltrometers, time domain reflectometry, and tensiometers. Soil Sci. Soc. Am. J. 62: 318–325.

Smettem, K.R.J., Parlange, J.-Y., Ross, P.J., and Haverkamp, R. 1994. Three-dimensional analysis of infiltration from the disc infiltrometer: 1. A capillary-based theory. Water Resour. Res. 30: 2925–2929.

White, I., Sully, M.J., and Perroux, K.M. 1992. Measurement of surface-soil hydraulic properties: disk permeameters, tension infiltrometers, and other techniques. In G.C. Topp, W.D. Reynolds, and R.E. Green, Eds., Advances in Measurement of Soil Physical Properties: Bringing Theory into Practice. SSSA Special Publication No. 30, Soil Science Society of America, Madison, WI, pp. 69–103.

Smettem, K.R.J. and Smith, R.E. 2002. Field measurement of infiltration parameters. In R.E. Smith, Ed., Infiltration Theory for Hydrologic

Wooding, R. 1968. Steady infiltration from a shallow circular pond. Water Resour. Res. 4: 1259–1273.



Chapter 83 Unsaturated Hydraulic Properties: Instantaneous Profile W.D. Reynolds Agriculture and Agri-Food Canada Harrow, Ontario, Canada

83.1 INTRODUCTION The instantaneous profile method (also known as the internal drainage method) is used primarily for direct (in situ) field measurement of the hydraulic conductivity (K ) versus matric head (c) relationship, K(c) [LT1 ], and=or the hydraulic conductivity versus volumetric water content (u) relationship, K(u) [LT1 ]. It is also occasionally used for in situ determination of u(c) desorption curves; and it may potentially be useful for in situ estimation of the capillarity relationships (Chapter 69), although this has not yet been attempted (see Comment 7 in Section 83.4). An alternative field method for determining near-saturated K(c) and capillarity relationships is given in Chapter 82. Laboratory methods for K(c) or K(u) determination are given in Chapter 80 and Chapter 81. Selected methods for estimating K(c), K(u), and u(c) from surrogate porous medium properties are given in Chapter 84. A discussion of the principles and parameters associated with the determination of K(c), K(u), u(c), and the capillarity relationships appears in Chapter 69. The instantaneous profile method involves installing probes for in situ measurement of volumetric water content, u [L3 L3 ] (e.g., time-domain reflectometer (TDR) probes, Chapter 70), and pore water matric head, c [L] (e.g., tensiometers, Chapter 71), at selected depths below the porous medium surface. The porous medium is then wetted to field saturation, and the K(c) and K(u) relationships are derived from periodic measurements of u and c during drainage, which coincidentally provides an in situ u(c) desorption curve (Chapter 69). The resulting K(c), K(u), and u(c) relationships are most relevant for one-dimensional vertical flow and are usually limited to the ‘‘wet end’’ matric head range of about 200 c 0 cm (Dirksen 2001). Although the method may be somewhat involved and laborious, it is often considered the ‘‘benchmark’’ of precision and relevance against which other field-based K(c), K(u), and u(c) methods are evaluated (Vachaud and Dane 2002).

1129


1130


83.2 APPARATUS AND PROCEDURES 1

Enclose a level area 12 m2 in the field using a lined earthen berm, wooden planking, or other materials that will allow ponding of water. The area should be large enough that the initial infiltration and subsequent drainage are effectively vertical in the center of the enclosure (see also Comment 1 in Section 83.4). Clear the enclosed area of vegetation by clipping level with the soil surface.

2

In the center of the enclosed area, install the water content (u) and matric head (c) probes at the desired depths below the surface (see Chapter 70 and Chapter 71 for probe designs; see also Comment 2 in Section 83.4). The u and c probes should be fast-acting (e.g., response time of 12 min) to accurately portray the potentially rapid changes in porous medium water content and matric head at the start of measurements. Depth increments of 15–30 cm are often recommended, with at least one pair of uc probes per porous medium layer or soil horizon. Installation procedures should be used that prevent ‘‘short-circuit’’ flow down the probe access holes, e.g., use of tight-fitting access holes, or backfilling oversized access holes with powdered bentonite and=or clay. In some cases, it may also be feasible to use inclined access holes so that any leakage down the hole is diverted into the porous medium matrix before contact with the probe sensors. Even small amounts of short-circuit flow=leakage to the probe sensors may cause erroneous results. The antecedent water content of the porous medium or soil profile should be low enough that a wide range of u and c can be measured between antecedent and field-saturated conditions. The porous medium should not be so dry, however, that the pore water matric head probes (e.g., tensiometers) lose hydraulic contact with the porous material (which usually occurs for porous ceramic cup tensiometers when the antecedent c falls below about 800 to 900 cm).

3

The enclosed area is flooded with water until the porous material is either fieldsaturated (i.e., u ¼ ufs , c 0—see Chapter 69), or u and c are constant at all instrumented depths. The area is then insulated and covered (e.g., straw=bark mulch laid down and a plastic sheet placed on top) to impose gravity drainage under a surface boundary condition of zero flux and constant temperature. Gravity drainage causes u and c to decrease with time throughout the initially wetted volume of porous material.

4

Collect u versus time and c versus time data at all depths from the time when flooding stops (t ¼ 0) until u and c decreases to effectively constant values, a time period that can extend from hours to months, depending on the porous medium. Data collection often needs to be frequent (or even continuous) for the first hour or so after irrigation is stopped, as initial drainage rates are often quite rapid, which in turn causes rapid decreases in u and c. Data collection can be much less frequent as drainage slows, however; and it is often recommended that the later-time collection intervals follow geometric time increments (e.g., at 3, 6, 12, 24 h after irrigation is stopped) to better define the often exponentiallike decline of u versus time. If the field setup is maintained, replicate measurements can be made over several wetting=drainage cycles and=or at different times of the year.


Unsaturated Hydraulic Properties: Instantaneous Profile

1131

83.3 ANALYSIS AND EXAMPLE CALCULATIONS The amount of water, W [L], stored in the porous medium profile at time, t [T], is given by W(z,t) ¼

Zz

u(z,t)dz; 0 z L

(83:1)

0

where z [L] is depth below the porous medium surface (z ¼ 0 at surface), and L [L] is the depth to the deepest pair of u and c probes. Given that a no-flow boundary is imposed on the porous medium surface, the flux density, q [LT1 ], for drainage at any depth and time in the measurement zone is given by Rz q(z,t)z ¼

u(z,t)dz

0

@t

@W(z,t) ¼ @t z

(83:2)

which states that the time rate of decrease in water storage between the porous medium surface and depth, z, equals the drainage flux density at depth, z. Darcy’s law for water flow at any depth and time in the porous medium profile can be written as @H(z,t) (83:3) q(z,t)z ¼ K(u)z @z z where H(z,t) ¼ c(z,t) z is hydraulic head (z positive downward), and K(u)z [LT1 ] is the hydraulic conductivity–volumetric water content relationship at depth, z. Substituting Equation 83.2 in Equation 83.3 and solving for K(u)z produces @W(z,t) @H(z,t) K(u)z ¼ @t z @z z

(83:4)

which indicates that K(u)z at depth, z, can be determined from the time rate of decrease in water storage between the porous medium surface and depth, z, divided by the hydraulic head gradient at depth, z. The u(c)z and K(c)z relationships can also be determined because both u(z,t) and c(z,t) are measured at each depth, i.e., replace u with the corresponding c in K(u)z to produce K(c)z . The accuracy of Equation 83.4 is obviously dependent on the accuracy with which @H(z,t)=@z and @W(z,t)=@t can be determined. The accuracy of @H(z,t)=@z is improved by careful calibration of the pore water matric head probes (tensiometers) and by minimizing the depth increments between the probes. The accuracy of @W(z,t)=@t, on the other hand, is improved by curve-fitting empirical, time-differentiable expressions to W(z,t) versus t data in order to better describe both the very rapid decrease in W(z,t) at early time and the very slow decrease in W(z,t) at late time. It has been found, for example, that the empirical relationships (Vachaud and Dane 2002) W(t) ¼ a ln t þ b

(83:5)

W(t) ¼ ctd

(83:6)

or



1132

can often provide good fits to a wide range of porous medium data, where a, b, c, and d are curve-fitting constants. These expressions are then differentiated with respect to time and substituted into Equation 83.4 to produce @H(z,t) (83:7) K(u)z ¼ [a=t]z @z z or K(u)z ¼ [cdt

(d1)

@H(z,t) ]z @z z

(83:8)

where Equation 83.7 incorporates Equation 83.5, and Equation 83.8 incorporates Equation 83.6. The determination of K(u) can be simplified substantially if the porous medium profile is homogeneous over the depth range of interest. In this special case, drainage produces @H(z,t)=@z 1 and thus tensiometer measurements of c(z,t) are not required. In addition, @W(z,t)=@t can be represented by @W(z,t) @u(t)z @u(t)z ¼z ¼ zm @t @t @t z

z

(83:9) z

where u(t)z is the average water content from the porous medium surface to depth, z (see Comment 3 in Section 83.4), which can be represented by the empirical relationship (Libardi et al. 1980) u(t)z ¼ mu(t)z þ n

(83:10)

where u(t)z is the measured water content at depth, z, and m and n are curve-fitting constants (m is generally close to 1). It has also been established that the empirical relationship (Vachaud and Dane 2002) u(t)z ¼ t tm

(83:11)

applies for drainage in homogeneous profiles, where t and m are coefficients obtained from a regression analysis of u versus t data at each depth, z. Substituting Equation 83.9 through Equation 83.11 into Equation 83.4 and remembering that @H(z,t)=@z 1 produces a simplified K(u) analysis K(u)z ¼ zmt (1=m) mu[(m1)=m] z

(83:12)

which provides a simple analytic relationship between K(u)z and uz at any depth, z, in a homogeneous porous medium profile, and avoids direct estimation of uncertain time and space derivatives associated with W(z,t) and H(z,t), respectively. Further details on the justification and derivation of Equation 83.9 through Equation 83.12 can be found in Libardi et al. (1980) and Vachaud and Dane (2002). Although the simplified analysis (Equation 83.12) is technically restricted to homogeneous porous media, some applications suggest that it can still provide useful results in profiles that



1133

are moderately heterogeneous with depth (Hillel et al. 1972; Vachaud and Dane 2002). In addition, the simplified analysis avoids serious error propagation problems associated with inaccurate tensiometer readings (e.g., insufficient equilibration when c is changing rapidly; lack of sensitivity when c is close to zero; partial or complete loss of hydraulic contact when c is low), and with inaccurate estimation of hydraulic head gradients (Flu¨hler et al. 1976). It may consequently be advisable to apply the simplified analysis even when tensiometer data are available. A comparison of the complete analysis (Equation 83.7 and Equation 83.8) and the simplified analysis (Equation 83.12) is given in Vachaud and Dane (2002). An example data sheet and calculation for the simplified analysis is given in Table 83.1 and Figure 83.1 through Figure 83.3. TABLE 83.1 Example Data Sheet and Calculation of K(u) for the Instantaneous Profile Method, Simplified Analysis (Equation 83.12). Analysis Based on Volumetric Water Content Data Obtained from a Probe Located at Depth, z ¼ 50 cm Table (a): u versus t at depth, z ¼ 50 cm (Figure 83.1) Time, t (min)

ln (t)

u (cm3 cm3 )

ln (u)

1 3 10 20 40 80 160 320 640 1280 2560 5120 8100

0 1.0986 2.3026 2.9957 3.6889 4.3820 5.0752 5.7683 6.4615 7.1546 7.8478 8.5409 8.9996

0.50 0.47 0.44 0.42 0.40 0.39 0.37 0.36 0.34 0.33 0.32 0.31 0.30

0:6932 0:7550 0:8210 0:8675 0:9163 0:9416 0:9943 1:0217 1:0788 1:1087 1:1394 1:1712 1:2040

Regression fit of Equation 83.11 to data in Table (a): Intercept ¼ ln t ¼ 0:6964. Therefore, t ¼ 0:4984. Slope ¼ m ¼ 0:0569. R 2 ¼ 0:9979. Table (b): u(t )z versus u(t )z at depth, z ¼ 50 cm (Figure 83.2) u(t )z (cm3 cm23)

u(t )z (cm3 cm23)

0.50 0.48 0.44 0.40 0.38 0.34 0.30

0.47 0.46 0.42 0.38 0.36 0.34 0.30

Regression fit of Equation 83.10 to data in Table (b). The u(t)z values were measured at depth, z, and the u(t)z values were calculated to depth, z, using Equation 83.13: Intercept ¼ n ¼ 0:0401. Slope ¼ m ¼ 0:8624. R 2 ¼ 0:9929. (continued)



1134

TABLE 83.1 (continued) Example Data Sheet and Calculation of K(u) for the Instantaneous Profile Method, Simplified Analysis (Equation 83.12). Analysis Based on Volumetric Water Content Data Obtained from a Probe Located at Depth, z ¼ 50 cm Table (c): K(u)z versus uz at depth, z ¼ 50 cm (Figure 83.3) K(u)z (cm min1 )

uz (cm3 cm3 )

K(u)z (cm s1 ) 2:16 102 6:85 103 2:01 103 8:48 104 3:43 104 2:14 104 8:06 105 4:84 105 1:67 105 9:62 106 5:43 106 3:01 106 1:64 106

0

1:30 10 4:11 101 1:21 101 5:09 102 2:06 102 1:29 102 4:83 103 2:91 103 1:00 103 5:77 104 3:26 104 1:81 104 9:83 105

0.50 0.47 0.44 0.42 0.40 0.39 0.37 0.36 0.34 0.33 0.32 0.31 0.30

K (u)z values in Table (c) are calculated using Equation 83.12: z ¼ 50 cm: m ¼ 0:8624: t ¼ 0:4984: m ¼ 0:0569: The above process is repeated for each measurement depth=probe.

−0.6 Data ln(q ) = −0.0569 ln(t ) −0.6964 R 2 = 0.9979

−0.7

In(q ) (cm3 cm−3)

−0.8 −0.9 −1.0 −1.1 −1.2 −1.3 0

2

4

6

8

10

In(t ) (min)

FIGURE 83.1. Example measurements of decrease in volumetric water content, u, with time, t, at depth, z, for the instantaneous profile method. The straight line is a regression through the data to obtain the coefficients in Equation 83.11. The data appear in Table 83.1a.



1135

0.50 Data qav(t )z = 0.8624 q(t )z + 0.0401 qav(t )z (cm3 cm−3)

0.45

R 2 = 0.9929

0.40

0.35

0.30 0.30

0.35

0.40

0.45

0.50

q(t )z (cm3 cm−3)

FIGURE 83.2. Example of average volumetric water content to depth z, uav (t)z , versus measured volumetric water content at depth z, u(t)z , for the instantaneous profile method. The straight line is a regression through the data to obtain the coefficients in Equation 83.10. The data appear in Table 83.1b.

Hydraulic conductivity, K(q )z (cm s−1)

10−1

10−2

10−3

10−4

10−5

10−6 0.30

0.35

0.40

Volumetric water content, qz

0.45

0.50

(cm3 cm−3)

FIGURE 83.3. Example calculation of hydraulic conductivity, K (u)z , versus volumetric water content, uz , at depth z using the instantaneous profile method (Equation 83.12). The data appear in Table 83.1c.



1136

83.4 COMMENTS 1

A critical requirement of the method is that flow be entirely vertical and due only to gravity drainage. Hence, situations that induce lateral flow (e.g., presence of subsurface flow-impeding layers; presence of a shallow water table), or upward flow (e.g., surface evaporation=transpiration) must be avoided.

2

Use of single tube, multidepth matric head probes (e.g., multilevel tensiometers) and water content probes (e.g., TDR or neutron probes that slide down an access tube) may be advisable in the instantaneous profile method because they can reduce the number of access holes to as few as two (i.e., one for matric head and the other for water content), which in turn minimizes profile disturbance and potential pathways for short-circuit flow to the sensors.

3

If the vertical spacing between the water content probes is highly variable (as might be required to accommodate subsurface layers, horizons, etc.), it may be advisable to calculate u(t)z as a weighted arithmetic mean, with the weights determined by the depths of the water content probes; i.e., u(t)z ¼

n 1X u(t)i Ti ; i ¼ 1, 2, 3, . . . , n; z i¼1

n>1

(83:13)

where z [L] is the depth (probe) under consideration, n is the number of water content probes between the surface and depth z, u(t)i [L3 L3 ] is the water content versus time values measured at probe i, and Ti is the weighing factor at probe i which is given by Ti ¼

(di þ 1 di1 ) ; i ¼ 2, 3, 4, . . . , (n 1) 2

(83:14a)

(d1 þ d2 ) ; i¼1 2

(83:14b)

(dn dn 1 ) ; i¼n 2

(83:14c)

T1 ¼ Tn ¼

where di [L] is the depth from the surface to water content probe i, d1 is the depth to the shallowest water content probe (i ¼ 1), and dn (i ¼ n) corresponds to the depth (z) of the water content probe under consideration. When the probe under consideration is the shallowest probe (i ¼ 1), or in the special case when only one water content probe was installed (n ¼ 1), then T1 ¼ d1 ¼ z, and u(t)z ¼ u(t)z . 4

The complete analysis (Equation 83.7 and Equation 83.8) is generally incapable of determining the K uc relationship reliably within the top 20–30 cm of the porous medium surface because the near-surface hydraulic head gradient is very low, and actually goes to zero at the surface because q(z,t) ¼ 0 is imposed at the surface (see Equation 83.2). The simplified analysis (Equation 83.12) applies at all depths, however; and ‘‘at surface’’ estimates of K (u) are also possible via timed collection of core samples from the porous medium surface and subsequent determination of u by oven-drying.



1137

5

The instantaneous profile method may not yield realistic representations of K (u), K (c), and u(c) in highly structured or low-permeability materials. In highly structured materials, flow through preferential flow zones (e.g., worm holes, cracks, finger-flow zones, etc.) may be missed partially or completely by the relatively small u and c probes. In low-permeability materials, the time required for profile saturation and drainage can be impractically long.

6

The empirical relationships between W(t) and t (Equation 83.5 and Equation 83.6), and between u(t)z and t (Equation 83.11) are not fixed. Alternative empirical relationships that provide better fits to particular data sets are permissible; however, W(t) versus t must be differentiable by t, and u(t)z versus t must be both differentiable by t and explicitly solvable in terms of t.

7

Although not yet attempted, it should be possible to estimate the capillarity relationships, a*(c), S(c), PD(c), and NP(c) (see Chapter 69), from instantaneous profile determinations of K (c) and u(c). This would likely involve the following: (i) fitting an appropriate function to the K (c) data; (ii) integrating under the fitted K (c) function to obtain f(c) via Equation 69.13; (iii) calculation of a*(c) via Equation 69.16 (i.e., the a*(c) values are ‘‘integrally correct’’); and (iv) using a*(c), K (c), and u(c) in Equation 69.17 through Equation 69.19 to obtain S(c), PD(c), and NP(c), respectively. The success of this would probably depend on how well the K (c) function fits the K (c) data. The capillarity relationships obviously cannot be determined if the simplified analysis is used (i.e., Equation 83.12), as no c measurements are available in that case.

83.5 STRENGTHS AND WEAKNESSES OF THE INSTANTANEOUS PROFILE METHOD The main strengths of the instantaneous profile method are that it can yield simultaneous in situ estimates of K(u), K(c), and u(c) for a number of depths and horizons during active drainage, and it does not require assuming any particular functional form for K(u), K(c), and u(c). No other method is capable of doing this. On the other hand, important weaknesses of the method include the need for complex and delicate equipment (e.g., TDR probes, tensiometers) installed at various depths below the surface; measurement times that can run from hours to months; the potential need for large volumes of water to effect profile saturation; extensive effort required for replication (because of extensive equipment, time, and water requirements); and determination of only the wet end (200 c 0 cm) of the K(u), K(c), and u(c) relationships. Particular strengths of the simplified approach for homogeneous porous materials (i.e., Equation 83.9 through Equation 83.12) include avoidance of ‘‘finicky’’ pore water matric head probes (e.g., tensiometers); no requirement to evaluate uncertain water content and hydraulic head gradients; and provision of a straightforward analytical relationship between K(u)z and uz (Equation 83.12) that applies at all depths. Disadvantages of the simplified approach include loss of useful u(c) and capillarity parameter data (because c is not measured); potentially inadequate representations of the data by the available empirical relationships; and limited applicability to nonuniform (e.g., layered) porous medium profiles.



1138

REFERENCES Dirksen, C. 2001. Unsaturated hydraulic conductivity. In K.A. Smith and C.E. Mullins, Eds. Soil and Environmental Analysis: Physical Methods. Marcel Dekker, Inc., New York, NY, pp. 218–219. Flu¨hler, H., Ardakani, M.S., and Stolzy, L.H. 1976. Error propagation in determining hydraulic conductivities from successive water content and pressure head profiles. Soil Sci. Soc. Am. J. 40: 830–836. Hillel, D., Krentos, V.D., and Stylianou, Y. 1972. Procedure and test of an internal drainage method

for measuring soil hydraulic characteristics in situ. Soil Sci. 114: 395–400. Libardi, P.L., Reichardt, K., Nielsen, D.R., and Biggar, J.W. 1980. Simple field methods for estimating soil hydraulic conductivity. Soil Sci. Soc. Am. J. 44: 3–7. Vachaud, G. and Dane, J.H. 2002. Instantaneous profile. In J.H. Dane and G.C. Topp, Eds. Methods of Soil Analysis, Part 4—Physical Methods. Soil Science Society of America, Madison, WI, pp. 937–945.


Chapter 84 Estimation of Soil Hydraulic Properties F.J. Cook Commonwealth Scientific and Industrial Research Organization Indooroopilly, Queensland, Australia

H.P. Cresswell Commonwealth Scientific and Industrial Research Organization Canberra, Australian Capital Territory, Australia

84.1 INTRODUCTION Soil hydraulic properties should be directly measured whenever possible, as they are critically important to the transport and storage of water and solutes (see other chapters in the Section Soil Water Analysis), and they are highly variable in space and time. However, direct measurement is not always feasible, due to restrictive budgets, insufficient time, and substantial difficulty associated with certain measurements, such as the dry end of the unsaturated hydraulic conductivity relationship. In these situations, the only practical option may be to estimate soil hydraulic properties from more easily measured parameters, such as texture, bulk density, porosity, and soil water desorption–imbibition relationships. Although the prudence of estimating soil hydraulic properties has been questioned (particularly when predicting or modeling water–solute movement; Philip 1991; Addiscott 1993; Passioura 1996), it can nonetheless provide insights that would otherwise be difficult to achieve; and it also allows extrapolation beyond the specific soil conditions under which measurements must be conducted. A plethora of methods have been developed over the years for estimating the saturated or field-saturated hydraulic conductivity and water content, the soil water desorption– imbibition relationships, and the unsaturated hydraulic conductivity relationships (e.g., Mackenzie and Cresswell 2002). We will restrict ourselves here, however, to the wellestablished and versatile methods of Cresswell and Paydar (1996), Minasny et al. (1999), Saxton et al. (1986), and van Genuchten et al. (1991).

84.2 SOIL WATER DESORPTION–IMBIBITION RELATIONSHIPS The soil water desorption relationship describes the release of water from soil, while the imbibition relationship describes the uptake of water. These so called u(c) relationships are

1139



1140

usually represented by empirical or quasiempirical functions or ‘‘models’’ describing the change in soil volumetric water content, u [L3 L3 ], with soil pore water pressure or matric head, c [L].

84.2.1 BROOKS–COREY (1964) u(c) MODEL One of the most popular quasiempirical u(c) models is the modified Brooks and Corey (1964) function (Campbell 1974; Hutson and Cass 1987): ui 2b ; ¼ usat 1 þ 2b u ¼ usat

1=b jcj ; a

a > 0;

b>0

u ui ; c c i < 0

ui c2 1 u usat ¼1 2b ; usat ui a2 usat

ci < c 0

(84:1a)

(84:1b)

(84:1c)

where usat [L3 L3 ] is the saturated or field-saturated soil volumetric water content, a [L] is a constant analogous to the air-entry pressure head of the desorption curve or the water-entry pressure head of the imbibition curve, b is a dimensionless empirical constant, and (ci , ui ) is the point of inflexion on the desorption–imbibition relationship (represented by Equation 84.1a) where the power curve component of the relationship (represented by Equation 84.1b) joins the parabolic curve component (represented by Equation 84.1c). The advantage of this form of the Brooks and Corey (1964) equation is that it avoids a mathematical discontinuity when jcj ¼ a. The parameters in Equation 84.1a through Equation 84.1c are obtained using three main approaches: (i) least-squares fitting of Equation 84.1b when water content and matric head data are available over the appropriate range of u(c); (ii) the ‘‘two-point’’ method where only two widely spaced data points on the u(c) curve are required; and (iii) the ‘‘surrogate data’’ method where the parameters are derived from other soil data such as texture, bulk density, average particle density, etc. These approaches are described further below. Regression Method for Brooks–Corey u(c) Rewriting Equation 84.1b into the form log10 u ¼

1 log10 jcj þ C1 ; b

u ui ; c c i

(84:2a)

where C1 ¼ log10 usat þ

1 log10 a b

(84:2b)

describes a straight line with slope 1=b, and a y-axis intercept of C1 . The b and C1 values are consequently found by least-squares regression of log10 u against log10 jcj. Once b is known,


Estimation of Soil Hydraulic Properties

1141

ui is readily determined from Equation 84.1a. Given that Equation 84.1b and Equation 84.2a apply for ui and ci , then ci can be obtained by rearranging Equation 84.2a into log10 jci j ¼ b( log10 ui C1 )

(84:2c)

which in turn allows a to be obtained by rewriting Equation 84.1b in the form

ui a ¼ jci j usat

b (84:2d)

Example calculations are given in Section 84.4. Two-Point Method for Brooks–Corey u(c) (Cresswell and Paydar 1996) Estimate usat using usat

rb ¼ C2 1 rs

(84:3a)

where rb [Mg m3 ] is the dry bulk density of the soil sample, rs ¼ 2:65 Mg m3 is the average soil particle density, and C2 ¼ 0:93 is a coefficient accounting for entrapped air, which is usually present in saturated (or field-saturated) soil. Measure two widely spaced points on the sample’s desorption or imbibition curve within the range of validity of Equation 84.1b, i.e., (c1 , u1 ) and (c2 , u2 ) where c1 and c2 are ci , and u1 and u2 are ui . This assumes that log10 u vs. log10 c is roughly linear in the range of validity of Equation 84.1b. Calculate b using b¼

Dx (log10 jc1 j log10 jc2 j) ¼ Dy log10 u1 log10 u2

(84:3b)

Note that b is the negative inverse slope of the straight line between the two points (Figure 84.1). Back substitute usat and b into Equation 84.1a to obtain ui . Calculate ci by rearranging Equation 84.3b into log10 jci j ¼ b(log10 ui log10 u1 ) þ log10 jc1 j

(84:3c)

where the data point, (c1 , u1 ), falls within the range of validity of Equation 84.1b. Example calculations are given in Section 84.4. Surrogate Data Method for Brooks–Corey u(c) (Saxton et al. 1986) The surrogate data method of Saxton et al. (1986) employs empirical regression relationships based on soil texture. The usat , b, and a parameters in Equation 84.1a through Equation 84.1c are given by



1142 −0.2

All data points Regressed data points Linear regression

−0.3

Points for two-point method

Log q

−0.4

−0.5

−0.6

−0.7

−3

−2

−1

0

1

2

3

Log

FIGURE 84.1. Soil water content (u) vs. pore water matric head (c) relationship for the data in Table 84.3. The data points used in the Brooks–Corey ‘‘regression’’ and ‘‘twopoint’’ methods are indicated.

usat ¼ A þ B(Sa) þ D ln (Cl)

(84:4a)

b ¼ E þ F(Cl)2 þ G(Sa)2 (Cl)

(84:4b)

. a ¼ 100 exp H þ I(Cl) þ J(Sa)2 þ L(Sa)2 (Cl) ubsat [kPa]

(84:4c)

where A, B, D, E, F, G, H, I, J, L are dimensionless empirical coefficients (Table 84.1), Cl is the % clay by weight (

Soil Sampling and Methods of Analysis, Second Edition

Sampling theory and methods

Sampling theory and methods

Sampling theory and methods

Physical and Chemical Methods in Soil Analysis

Sampling: Design and Analysis

Principles of Soil Dynamics , Second Edition

Fundamentals of Soil Ecology, Second Edition

Sampling Methods: Exercises and Solutions

Methods of geometry, Second Edition

Sampling Methods: Exercises and Solutions

Mosquito Ecology: Field Sampling Methods (Third Edition)

Fundamentals of environmental sampling and analysis

Fundamentals of Environmental Sampling and Analysis

Handbook of soil analysis: mineralogical, organic and inorganic methods

Handbook of Soil Analysis: Mineralogical, Organic and Inorganic Methods

Numerical Analysis (Second Edition)

Mathematical Analysis, Second Edition

Steroid Analysis, Second Edition

Mathematical Analysis, Second Edition

Real Analysis, Second Edition

Mathematical Analysis, Second Edition

Real Analysis, Second Edition

Sampling and Analysis of Indoor Microorganisms

Analytical methods for drinking water advances in sampling and analysis

Mathematical Analysis, Second Edition

Regression Analysis, Second Edition

Analytical Methods for Drinking Water: Advances in Sampling and Analysis

Bayes and Empirical Bayes Methods for Data Analysis, Second Edition

Multivariate Analysis, Design Experiments, and Survey Sampling

Stochastic Optimization Methods, Second Edition

Soil Sampling and Methods of Analysis, Second Edition

Sampling theory and methods

Sampling theory and methods

Sampling theory and methods

Physical and Chemical Methods in Soil Analysis

Sampling: Design and Analysis

Principles of Soil Dynamics , Second Edition

Fundamentals of Soil Ecology, Second Edition

Sampling Methods: Exercises and Solutions

Methods of geometry, Second Edition

Sampling Methods: Exercises and Solutions

Mosquito Ecology: Field Sampling Methods (Third Edition)

Fundamentals of environmental sampling and analysis

Fundamentals of Environmental Sampling and Analysis

Handbook of soil analysis: mineralogical, organic and inorganic methods

Handbook of Soil Analysis: Mineralogical, Organic and Inorganic Methods

Numerical Analysis (Second Edition)

Mathematical Analysis, Second Edition

Steroid Analysis, Second Edition

Mathematical Analysis, Second Edition

Real Analysis, Second Edition

Mathematical Analysis, Second Edition

Real Analysis, Second Edition

Sampling and Analysis of Indoor Microorganisms

Analytical methods for drinking water advances in sampling and analysis

Mathematical Analysis, Second Edition

Regression Analysis, Second Edition

Analytical Methods for Drinking Water: Advances in Sampling and Analysis

Bayes and Empirical Bayes Methods for Data Analysis, Second Edition

Multivariate Analysis, Design Experiments, and Survey Sampling

Stochastic Optimization Methods, Second Edition

Recommend Documents