Advances in Agronomy, Volume 47

DVANCES I N

igronomy

V O L U M E 47

Advisory Board Martin Alexander

Eugene J. Kamprath

Cornell University

North Carolina State University

Kenneth J. Frey

Larry P.Wilding

Iowa State University

Texas A&M University

Prepared in cooperation with the American Society of Agronomy Monographs Committee

S. H. Anderson L. P. Bush R. N. Carrow

M. A. Tabatabai, Chairman G. L. Horst R. J. Lwmoore R. H. Miller

G. A. Peterson

c.w. Stuber S. R. Yates

D V A N C E S I N

onomy VOLUME 47 Edited by

Donald L. Sparks Department of Plant and Soil Sciences University of Delaware Newark, Delaware

ACADEMIC PRESS, INC. Harcourt Brace Jovanovich, Publishers San Diego New York Boston London Sydney Tokyo Toronto

This book is printed on acid-free paper. @

Copyright 0 1992 by ACADEMIC PRESS, INC. All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Academic Press, Inc. 1250 Sixth Avenue, San Diego, California 92101

United Kingdom Edition published !y

Academic Press Limited 24-28 Oval Road, London NWI 7DX Library of Congress Catalog Number: 50-5598 International Standard Book Number: 0-12-000747-9 PRINTED IN THE UNITED STATES OF AMERICA 9 2 9 3 9 4 9 5 9 6 9 7 BC 9 8 7 6 5 4 3 2

1

Contents CONTRIBUTORS ............................................... PREFACE......................................................

ix

xi

THE EFFECTS OF ACIDIC DEPOSITION ON FORESTED SOILS Wayne P. Robarge and Dale W.Johnson I . Introduction ............................................. 1 I1. Soil Acidification ......................................... I11. Forest Soils .............................................. IV. Case Studies of Soil Change ................................ V. FutureResearch .......................................... VI. Conclusions ............................................. References ..............................................

3 6 42 63 65 66

FINGERPRINTING CROP VARIETIES J . S. C. Smith and 0. S. Smith I . Introduction ............................................. I1. Characters Used to Fingerprint Cultivated Varieties . . . . . . . . . . . . I11. Discriminational Ability of Fingerprinting Techniques . . . . . . . . . . Iv. Usage of Fingerprints ..................................... V. NewTechniques ......................................... References ..............................................

85 86 108 119 125 127

TRANSPORT OF CHEMICALS THROUGH SOIL: MECHANISMS. MODELS. AND FIELD APPLICATIONS William A.Jury and Hannes Fluhler I . Introduction ............................................. I1. Transport and Transformation Processes ...................... I11. Analysis of Process Assumptions ............................. Iv. Field Studies of Solute Transport ............................ V. Concluding Remarks ...................................... References .............................................. V

142 143 169 175 191 195

vi

CONTENTS

EVOLUTION OF CORN Walton C. Galinat I. Introduction: Teosinte Is the Wild Corn ...................... I1. Transformation by Domestication and Isolation . . . . . . . . . . . . . . . .

I11. Time Required for Transformation under Domestication . . . . . . . . IV Multiple Domestications in the Origin of Corn . . . . . . . . . . . . . . . . V. Interpathway Heterosis .................................... VI. Husk Enclosure of the Ear ................................. VII . Current Direction of Corn Evolution and Where It Is Going . . . . . VIII . Summary and Conclusions ................................. References ..............................................

203 205 212 214 221 222 224 227 229

USEOF SURFACE COMPLEXATION MODELS IN SOIL CHEMICAL SYSTEMS Sabine Goldberg I. Introduction ............................................. I1. Description of Models .....................................

I11. Application of Models to Protonation-Dissociation Reactions on Oxides. Clay Minerals. and Soils .......................... Tv: Application of Models to Metal Ion Adsorption Reactions on Oxides. Clay Minerals. and Soils ............................ V. Application of Models to Inorganic Anion Adsorption Reactions on Oxides. Clay Minerals. and Soils .......................... VI. Application of Models to Organic Ligand Adsorption Reactions onoxides ............................................... VII . Application of Models to Competitive Adsorption Reactions on Oxides ............................................... VIII . Incorporation of Surface Complexation Models into Computer Codes ......................................... IX. Summary ............................................... References ..............................................

234 235 251 274 289 308 312 319 320 32 1

CONTENTS

vii

MODELING THE TRANSPORT AND RETENTION OF INORGANICS IN SOILS H . M . Selim Introduction ............................................. Transport Equations ...................................... Equilibrium Retention Models .............................. Kinetic Retention Models .................................. Multiple-Reaction Models ................................. Transport and Ion Exchange ................................ Transport in Layered Soil .................................. References ..............................................

331 333 337 342 350 367 377 381

INDEX........................................................

385

I. I1. I11.

N. V.

VI. VII .

This Page Intentionally Left Blank

Contributors Numbers in parentheses indicate the pages on which the authors’ contributions begin.

HANNES FLUHLER (141), Institute of Terrestrial Ecology, Swiss Federal Institute of Technology (ETH), Zurich, Switzerland WALTON C. GALINAT (203), Eastern Agricultural Center, University of Massachusetts-Amherst, Waltham, Massachusetts 02 1 54 SABINE GOLDBERG (233), USDA-ARS, U.S. Salinity Laboratory, Riverside, California 92501 DALE W. JOHNSON (l), Desert Research Institute, Reno, Nevada 89J12 WILLIAM A. JURY (141), Department of Soil and Environmental Sciences, University of California, Riverside, Riverside, California 92J21 WAYNE P. ROBARGE (l), Department of Soil Science, North Carolina State University,Raleigh, North Carolina 2 769s H. M. SELIM (3 3 I), Agronomy Department, Louisiana State University, Baton Rouge, Louisiana 70803 J. S. C. SMITH (85), Plant Breeding Division, Pioneer Hi-Bred International, Inc., Johnston, Iowa 50131 0.S. SMITH (85), Plant Breeding Division, Pioneer Hi-Bred International, Inc., Johnston, Iowa 50131

ix

This Page Intentionally Left Blank

Preface Environmental quality and biotechnology are currently two major research areas in the crop and soil sciences. Thus it is appropriate that a substantial portion of this volume of Advances in Agronomy is concerned with these topics. Three chapters deal with aspects of environmental quality. Chapter 1 discusses the effects of acidic deposition on forest soils, emphasizing physical and chemical processes in forest ecosystems that interact with, modify, and respond to acidic inputs; presenting case histories of soil changes in regions where acidic deposition may be causal and in regions where it is not; and giving recommendations concerning future research that is needed to more accurately characterize and develop models for the effects of acidic deposition on forest soils. Chapters 3 and 6 examine modeling of organic and inorganic chemical transport in soils. Chapter 3 provides a detailed discussion on the mechanisms, models, and field applications of chemical transport in soils, with particular emphasis placed on the importance of approaches that accurately describe chemical transport in heterogeneous field soils. Chapter 6 describes the features of models that govern retention of inorganic solutes in soils. Single, multiple, and multicomponent or competitive models that are based on equilibrium or kinetics are presented. Many equilibrium-based models have been promulgated in the literature to describe reactions at the solid-liquid interface. One group of these models is microscopically based and is referred to as surface complexation models. These models are the subject of Chapter 5. This chapter critically reviews five common surface complexation models of the mineral-solution interface and their use in describing soil chemical systems. Common characteristics and adjustable parameters are covered, the application of models to ion adsorption on soil constituents and soils is presented, and the incorporation of surface complexation models into computer codes is discussed. Chapters 2 and 4 deal with the crop sciences. Chapter 2 is concerned with fingerprinting crop varieties. Topics that are discussed include evolution in the scope and technology of fingerprinting, characters used in fingerprinting cultivated varieties, discriminational ability of fingerprinting techniques, use of fingerprints, and new techniques. Chapter 4 deals with the evolution of corn. It critically evaluates previous research in this area and assesses current thinking on this very important crop. I deeply appreciate the fine contributions from these authors.

DONALD L. SPARKS xi

This Page Intentionally Left Blank

THE EFFECTS OF ACIDIC DEPOSITION ON FORESTED SOILS Wayne P. Robarge' and Dale W. Johnson2 'Department of Soil Science, North Carolina State University, Raleigh, North Carolina 27695 ?Desert Research Institute, Reno, Nevada 895 12

I. Introduction 11. Soil Acidification 111. Forest Soils A. Physical Factors B. Chemical Factors IV.Case Studies of Soil Change A. Categories of Soil Change and Methods for Measurement B. Intensity-Type Changes C. Capacity Changes Due to Leaching D. Capacity Changes Due to Uptake E. Seasonal Variations in Exchangeable Cations F. Interactions of Uptake and Leaching V. Future Research VI. Conclusions References

I. INTRODUCTION Acidic deposition is composed primarily of N and S acid-forming compounds that undergo gas-phase oxidation and aqueous-phase reactions in the atmosphere to form nitric and sulfuric acids (Tanner, 1989). Partial neutralization of these acids by NH3 in the atmosphere results in H + , NH;, NO;, and SO:- being the dominant ions in acidic rainwater and

1 Advances m Apmny, Volunr 47 Copyighr Q 1992 by Academic Press, Inc. All rights of reproduction in my form reserved.

2

WAYNE P. ROBARGE AND DALE W. JOHNSON

cloudwater (Weathers et al., 1988; Reisinger and Imhoff, 1989; Saxena et al., 1989; Sigmon et al., 1989; Aneja et al., 1991). Acidification of soils near point sources of N and S acid-forming substances is well known, thus one of the earliest concerns over acidic deposition occurring over a regional basis was the potential for soil acidification (Ulrich, 1980). Since then, a number of reviews have been written and models proposed to explain how acidic deposition can increase soil acidity (Ulrich et al., 1980; Reuss, 1983; van Breemen et al., 1983; Kohlmaier et al. , 1983/1984; Bloom and Grigal, 1985; Cosby et al., 1985; Rechcigl and Sparks, 1985; Reuss and Johnson, 1985; 1986; Tabatabai, 1985; Krause et al., 1986; Ulrich, 1987; Binkley et al., 1989a; De Vries et al., 1989; Reuss and Walthall, 1989; Reuss et al., 1990; Walker et al., 1990). Such models have been used to develop sensitivity criteria to determine which soils on a regional basis will be most sensitive to acidification by acidic deposition (McFee, 1983; Binkley et al., 1989a). These model reactions, based on soil chemistry theory and combined with field observations of changes in soil chemical parameters and surface water chemistry of streams, have led to the general conclusion that acidic inputs into such ecosystems are capable of increasing soil acidity (Reuss et al., 1987). Objections to this general conclusion have largely been based on the argument that the effects of acidic deposition on forest soil systems can only be evaluated from the standpoint of how acidic inputs interact with the natural processes of soil acidification (Rosenquist, 1978; Krug and Frink, 1983a,b; Tabatabai, 1985). Central to this point of view is the fact that soil formation in humid temperature climates is an acidifying process, and that a correlation between areas of high acidic deposition and the presence of acidic soils and stream waters is not sufficient cause to conclude that acidic inputs have increased soil acidity in these ecosystems. Many of the temperate forests of northern Europe and eastern North America that currently receive acidic inputs have undergone substantial changes in land use policy during the past 200 years (see, e.g. ,Brand et al., 1986). As many of these forests are now aggrading, the natural soil acidification that accompanies such regrowth cannot be attributed to acidic deposition (Krug and Frink, 1983a). The discussion concerning the pros and cons of both hypotheses on the effects of acidic deposition on soil acidity still continues (Havas et al., 1984; N. M. Johnson et al., 1984; Krug and Isaacson, 1984; Norton et al., 1989), but it has served to illustrate the limitations in our current understanding about the processes of soil acidification in forest soils. Without such knowledge it will not be possible to understand fully the effects of acidic deposition on these soils, either now, at current levels of input, or in the future, as regulatory measures begin to reduce acidic inputs. It is becoming

ACID DEPOSITION ON FORESTED SOILS

3

clear that the rate at which forest soils are changing will require a rethinking of our approach to studying soil changes in these ecosystems (D. W. Johnson et al., 1991a). This review on the effect of acidic deposition on forest soils assumes that it is no longer necessary to argue whether acidic inputs are having an impact on soil systems. The levels of input are now well characterized and it is highly unlikely that any natural ecosystem would fail to respond in some way (at the very least by increased leaching) to such sustained inputs of potential energy. Emphasis instead is on what effects may be possible given the natural physical and chemical processes acting in forest ecosystems. Discussions of possible mechanisms will differ from earlier reviews in that more emphasis will be placed on the limitations of such mechanisms, both in terms of theory and the currently available body of knowledge. Reference to agricultural and other intensively managed soil systems is excluded because it is generally accepted that the influence from acidic deposition on such soils will be minimal (McFee, 1983; Tabatabai, 1985). Our specific approach is to first attempt to define the often used but poorly understood term “soil acidification,” followed by a brief synopsis of the dominant physical and chemical processes in forest ecosystems that interact with, modify, and respond to acidic inputs. This information is then used as background to discuss published case histories of soil change in regions where acid deposition may be causal, and in regions where it is not. Last, a set of recommendations is put forward concerning areas for future research that are needed to properly characterize and develop models for the effects of acidic deposition on forest soils.

11. SOIL ACIDIFICATION Soil acidification refers to a complex set of processes that result in the formation of an acid soil (pH < 7.0). Soil acidification, therefore, in the broadest sense, can be considered as the summation of natural and anthropogenic processes that lower measured soil pH (Krug and Frink, 1983a). In forest ecosystems, natural acidifying processes include base cation uptake (by plants or microbes); natural leaching by carbonic, organic, or nitric acid; and humus formation (Ulrich, 1980). Anthropogenic acidifying processes include biomass harvesting (which simulates increased uptake) (Binkley et al., 1989a), land use conversion (Berdbn et af., 1987; Billet el af., 1988, 1990b; D. W. Johnson et af., 1988b), fertilization (van Breeman el d., 1982), as well as atmospheric inputs of acidifying compounds (Reuss and Johnson, 1986). Barring inputs of lime from anthropogenic

4

WAYNE P. ROBARGE AND DALE W. JOHNSON

sources, or marine influences, forest ecosystems developed on noncalcareous-bearing parent materials in humid environments will have acid soils. Attempts to measure soil acidification as a result of acidic inputs often center on attempting to detect changes in soil pH (see, e.g., Tamm and Hallbacken, 1986). Though this approach seems intuitively obvious, practical considerations ranging from suitable analytical methodology, spatial variability, and determination of soil horizonation to changes in land use patterns often severely limit the usefulness of this rather simple approach. Soil acidification cannot be quantitatively described by a single index parameter, even though it is often assumed that soil pH is such a parameter (Matzner, 1989). Other changes in soils that may occur during soil acidification include loss of nutrients due to leaching; loss or reduction in the availability of certain plant nutrients (such as phosphorus and molybdenum, which are more strongly retained in acid soils); an increase in the solubility of toxic metals (primarily aluminum and manganese), which may influence root growth and nutrient and water uptake; and a change in microbial populations and activities (Binkley et al., 1989a). Such changes will often be accompanied by changes in overall soil pH, but the degree of change will be dependent on a combination of properties within a given soil system. A more quantitative measure of soil acidification can be obtained by defining it as a decrease in the acid-neutralizing capacity (ANC) of a soil (Berdtn et al., 1987; De Vries and Breeuwsma, 1987). This approach is similar to that for aqueous systems (Stumm and Morgan, 198l), wherein the ANC of a soil solution can be defined as the aqueous base equivalence minus the strong acid equivalence as determined by strong acid titration to a reference pH (typically pH 4.5) (van Breemen et al., 1984). The ANC of the inorganic fraction of a mineral soil can be defined as the sum of basic components minus the strongly acidic components (van Breemen et al., 1984) ANC = 6[A1203]+ 6[Fe203]+ 2[Fe0] + 4[Mn02]

+ 2[Mn0] + 2[Ca0] + 2[Mg0] + 2[Na20] + 2[K20] - 2[SO3] - 2[P205] - [HCl]

(1) where the brackets denote molar concentrations. Note that the metal oxides include the metal cations in the soil solids, as well as those on the exchange complex and in the soil solution. A decrease in the cationic components (such as CaO) or an increase in the acidic components (such as SO3) will result in an increase in soil acidification (a decrease in ANC)

ACID DEPOSITION ON FORESTED SOILS

5

(van Breemen et al., 1984). A decrease in the cationic components could occur through biomass uptake or leaching, whereas an increase in the acidic components could result from inputs of SOT2.This approach emphasizes the mass of acidic input, as equivalents of H+ or NO, and SO:-, rather than the intensity of input as measured by pH (Binkley and Richter, 1987). Thus, a detailed H+ budget for a forest ecosystem, which attempts to quantify all of the proton-producing and -consuming processes within a given ecosystem, offers a means of separating out soil acidification due to acidic deposition from natural acidification processes (van Breeman el al., 1983). Soil acidification could then be defined as the result of an irreversible flux of protons to the soil ecosystem. The limitations of this approach are that such budgets are difficult to construct, and the budget estimates for the various processes are often associated with a large degree of uncertainty both spatially and temporally (Binkley and Richter, 1987). The components of the soil that comprise the ANC can be divided into processes that are relatively fast (approach equilibrium rapidly), slow (processes that are rate limited but for which the kinetics are known), or very slow (may be essentially ignored as having an impact on the system) (Furrer ef al., 1990). Those processes considered to be fast have an immediate impact on the composition of the soil solution, and are also referred to as intensity factors (Reuss and Johnson, 1986). The soil components involved are predominately the soil solution and those soil surfaces that react rapidly to changes in the soil solution (e.g., cation and anion exchange capacity). Slow and very slow processes are referred to as capacity factors and essentially reflect an integration of changes in a soil system over time. These processes include cation and anion plant uptake, mineralization, oxidation and reduction, and primary and secondary mineral weathering (Furrer et al., 1990). Over the long term, the capacity tactors of a soil will control the range in intensity tactors that are ooserved. The ion pools that comprise the capacity factors greatly exceed those of the intensity factors and the inputs from acidic deposition (Keuss and Johnson, 1986; Reuss and Walthall, 1989). These relatively large pools of ions in already acid soils are the basis for the assumption that the impact from acidic deposition will be small compared to natural acidification processes (Krug and Frink, 1983a), and that substantial periods of time will be required before detectable changes in these bulk soil chemical properties will occur, if at all (Tabatabai, 1985). There is a growing consensus, however, that the primary effect of acidic deposition on forest soils is via the intensity factors and that substantial changes in the capacity factors by acidic deposition are not necessary to influence the composition of the soil solution (Reuss et al., 1987;

6

WAYNE P. ROBARGE AND DALE W. JOHNSON

D. W. Johnson et al., 1991a). This approach centers on the fact that the dependence of intensity factors on capacity factors in a soil is often nonlinear, and that small changes may result in relatively large changes in soil solution composition (Reuss and Johnson, 1986; Reuss and Walthall, 1989). It is also becoming apparent that the natural acidification processes within a given ecosystem predispose that system to the way it will respond to acidic inputs. It is argued that the change in the anion composition of the soil solution caused by the introduction of the NO; and SO:- can account for the observed changes in soil solution and surface water acidity without the necessity for involving further soil acidification. Such a scenario would mean that the effects of acidic deposition on forest ecosystems would occur fairly rapidly over a variety of soil types in a relatively short period of time once a critical loading of NO; and, in particular, SO:- is exceeded. It also follows that reduction in inputs below the critical input would have an immediate positive effect. Such responses have been observed both in the field (Wright et af., 1988a) and in the laboratory (Dahlgren et al., 1990). Soil acidification has long been cited as one of the effects of acidic deposition on forest ecosystems. However, it has often not been made clear that acidification of soils from acidic inputs must be viewed from the standpoint of being superimposed upon natural acidification processes. A quantitative measure of soil acidification can be obtained by using the concept of the ANC of a soil, but the term itself does not necessarily imply that there is a specific parameter (such as soil pH) or set of parameters that can be used to measure soil acidification. It is becoming apparent that there has probably been too much emphasis on change in soil pH and cation depletion as a necessary and expected effect of acidic deposition on soil systems (D. W. Johnson et al., 1991a). The influence of acidic deposition on forest soils might be better understood by focusing on the reactions of the mineral acid anions NO; and SO:- in soils-in particular, how these anions interact with soil acidity already present from natural acidification processes.

111. FOREST SOILS Model reactions of acidic deposition with soils often only emphasize chemical reactions within a theoretical soil horizon. Actual forest ecosystems are infinitely more complicated and offer a number of physical as well as chemical factors that must be considered in order to determine the effect of acidic deposition on forest soils and surrounding surface waters.

ACID DEPOSITION ON FORESTED SOILS

7

A. PHYSICAL FACTORS 1. Canopy Interactions

Acidic deposition reaches the forest ecosystem in the form of rainwater (Schaefer and Reiners, 1989), as cloud and fog droplet impact on the forest canopy (Lovett et al., 1982; Bruck et al., 1989; Reisinger and Irnhoff, 1989); Saxena et al., 1989; Sigmon et al., 1989; Saxena and Lin, 1990), and as dry deposition (Lindberg et al., 1986; Johnson and Lindberg, 1989; Murphy and Sigmon, 1989). Dry deposition is the accumulation of particulates and gases (such as HN03 and SO2) on the forest canopy in the absence of precipitation (Davidson and Wu, 1989). Interaction of these three forms of acidic deposition with the forest canopy changes their initial chemistry before they finally reach the forest floor and the underlying mineral soil as throughfall and stemflow. Throughfall is that fraction of wet deposition that comes in contact with the canopy before reaching the forest floor, whereas stemflow is that portion that drains down the branches and trunk (Parker, 1990). In most forest ecosystems with an intact canopy, it is throughfall and stemflow that are the major inputs of acidity and other ions directly into the soil system. The degree of interaction between acidic deposition and the forest canopy is illustrated by the data in Table I, which compares the relative chemical composition of throughfall to that of cloudwater and rainwater in a high-evaluation spruce-fir ecosystem in the Black Mountains of North Carolina (Bruck er al., 1989). The relative percentage of NO, and SO:- between cloudwater, rainwater, and throughfall are almost constant, with perhaps a slight decrease in NO, and a slight increase in SOT2 in the throughfall. The largest change observed is a shift in dominant cations, with H+ and NH,f replaced by K + , Ca2+, and Mg2+. These data are representative of throughfall measurements in other forest ecosystems that receive acidic deposition (Richter et al., 1983; Lindberg et al., 1986; Bredemeier, 1988, J o s h et al., 1988; Percy, 1989; Sigmon et al., 1989; Parker, 1990) and in studies using simulated acid rain treatments (Scherbatskoy and Klein, 1983; Kelly and Strickland, 1986; Kaupenjohann er al., 1988). More detailed information on throughfall chemistry under a variety of forest canopies can be found in Parker (1983, 1990) and Bredemeier (1988). A review of the processes that control throughfall chemistry can be found in Schaefer and Reiners (1989). The release of base cations from the canopy is largely in response to the fact that SO:- and, to a lesser extent, NO, are not adsorbed by the canopy along with H+ and NHT. It is now known, through the use of 35S(Garten

WAYNE P. ROBARGE AND DALE W. JOHNSON

8

Table I Total Ion Percent (pEq liter-') per Event for Cloudwater, Rainwater, and Throughfall Samples Collected in 1986"

Throughfall (%) Red spruce Cloudwaterb Ion

Fraser fir

Rainwater'

(%I

Site 1'

Site 2d

Site 1

Anions c1NO; s0:-

47.9 1.5 11.6 34.8

45.6 2.5 8.8 34.3

47.9 3.2 8.2 36.6

42.3 5.5 9.7 21.0

48.3 3.0 7.7 37.6

Cations

52.1 27.1 13.2 3.0 1.a 5.2 3.1 12.3

54.4 30.4 12.7 2.0 3.4 4.9 1.0 11.3

52.1 19.3 3.5 1.8 9.6 13.4 4.6 29.4

51.6 15.3 3.9 3.6 10.7 18.5 5.7 38.5

51.7 16.6 6.1 1.8 10.3 12.2 4.6 28.9

H+

Nb+ Na+

K+

Ca2+ Mg'+ Sum

"After Bruck et al. (1989); reprinted by permission of Kluwer Academic Publishers. bCollected at site 1. 'Mt. Gibbes (2006 rn); from June 29, 1986 to Sept. 21, 1986. d E a ~ face t of Commissary Ridge (1760 m); from June 29, 1986 to August 15, 1986.

et af., 1988; Garten, 1990), that SO:- in particular is conserved within the canopy, and that an increase in sulfate loading, either as H2S04 or NH,HSOa in cloudwater or rainwater ( J o s h et af., 1988) or as SO2 in dry deposition, will increase the concentration of base cations in throughfall and stemflow (Parker, 1990). As an example, Johnson and Lindberg (1989) estimates that between 40 and 60% of the base cations in throughfall collected at the Walker Branch Watershed in East Tennessee during 19811983 was due to canopy exchange with deposited airborne acids. This increase in base cation loss must come at the expense of the nutrient pool within the canopy, which in turn may mean an increase in base cation uptake. For the time period cited, this extra base cation uptake due to "neutralization" of acidic input via the canopy equals a total H + input of between 0.9 and 1.1 kmol( +) ha-' yr-' of internal acidification potential within the rooting zone (Johnson and Lindberg, 1989). Calculations for the Solling Forest in West Germany (Matzner, 1989) and for forests in the Netherlands (van Breeman et af., 1986) yield similar results. Neutralization of acidic inputs via the canopy, therefore, represents an indirect means of

ACID DEPOSITION ON FORESTED SOILS

9

increasing the acid load on a forest soil (Matzner, 1989; Ulrich, 1989). In certain ecosystems, almost half of the H+ loading from acidic deposition can be transferred to the soil system before the water transporting the acidic anions enters the soil (Johnson and Lindberg, 1989; Matzner, 1989). The acidity not neutralized by the canopy enters the soil via throughfall and stemflow essentially as a salt solution dominated by Ca2+ and K+ salts of SO:-, with a relatively minor contribution from the remaining mineral acids (Johnson and Lindberg, 1989). The composition of this mixture is not constant but varies considerably depending on a variety of factors, such as season of the year (Parker, 1990), seasonal changes in deposition loading (Johnson and Lindberg, 1989), the overall nutrient status of the ecosystem (Leininger and Winner, 1988; Reynolds et al., 1989; Huettl et al., 1990; Klumpp and Guderian, 1990), and stand age (Stevens, 1987). Throughfall and stemflow composition will also vary depending on the relative health of a stand (Alenas and Skarby, 1988). On a shorter time scale, throughfall and stemflow composition will vary due to length of time between rainfall events [i.e., amount of dry deposition loading varies (Velthorst and van Breemen, 198911, and even during storm events. The overall ionic strength of throughfall and stemflow generally decreases significantly during the course of an individual event (Kelly and Strickland, 1986; Lovett et al., 1989) due to washoff of particulates from the leaf surfaces (Schaefer and Reiners, 1989) and loss in ability of the canopy to buffer the reactions with rainwater during the course of a storm (Parker, 1990). Spatial variability in throughfall composition usually exceeds 25% when expressed as the coefficient of variation due to the flow of water through the canopy (Duijsings et al., 1986). The forest canopy is both a modifier and conduit of acidic deposition into the forest ecosystem. As illustrated above, the interaction between acidic inputs and the canopy can have profound influences on both the pathways and the chemistry of acidic substances that actually enter the forest floor and the underlying mineral soil. These changes, together with varying residence times within the different soil horizons, will influence the nature of soil reactions that are likely to occur. 2. Soil Horizons Acidic inputs entering a forest soil as throughfall and stemflow encounter a gradation in organic matter content extending from the forest floor, which is essentially 100% organic matter, to the underlying parent material of the mineral soil, which usually contains no innate source of organic carbon. Depending on the interactions of the soil-forming factors (Buol et al., 1980), the gradation in organic matter content may occur as distinct

10

WAYNE P. ROBARGE AND DALE W. JOHNSON

boundaries between soil horizons (e.g., Fig. 1) or as a gradual decrease in organic matter content with depth. The mixing of organic matter and mineral soil gives a range of reactive surfaces that will respond differently to changes in the percolating soil solution, depending on the initial composition and rate of input of throughfall and stemflow, and on how the chemical composition of this initial solution is changed as it passes through each succeeding soil horizon. The nature of the chemical reactions that may occur within each soil horizon will be discussed elsewhere in this review, but Fig. 1 does serve to illustrate the point that generalizations about a “soil’s” response to acidic inputs need to be properly defined in terms of the scale of observation (Fernandez, 1989). System level studies dealing with nutrient cycling avoid the issue of differences among soil horizons by integrating observations across the entire soil pedon (Adriano and Havas, 1989). In such studies, actual mechanisms within the plantsoil system are of secondary importance, especially in terms of element cycling between the canopy and soil and between soil horizons within the soil pedon due to natural processes. Studies addressing the specific mechanisms of the effects of acidic deposition on tree growth cannot ignore differences between soil horizons, as the rooting zones of trees are seldom confined to a given soil horizon (Fernandez and Struchtemeyer, 1985; Coutts, 1989; Fernandez, 1989). The gradation of soil organic matter throughout the forest soil pedon also means that attempts at measuring differences in soil properties over time, even within the same morphological horizon, must be approached

Figure 1. Exchangeable cations (1 M NH4CI extractable) and pH (0.01 M CaCI,) from a sampling (n = 23) of undisturbed well-drained and moderately well-drained pedons under forest cover in Maine. After Fernandez (1989).

ACID DEPOSITION ON FORESTED SOILS

11

with due caution (Fernandez, 1989). A soil sample from a particular depth within a given soil pedon can be expressed as the summation of its organic matter component and its mineral soil component: soil = organic matter

+ mineral soil

(2) Because the ability of soil organic matter to retain metal ions greatly exceeds that of the mineral soil fraction on a mass basis (Sposito, 1989), relatively small changes in soil organic matter content can dominate the overall physical and chemical properties of a given soil sample. Comparison of different soil samples from the same horizon and location over time, therefore, requires attention not only to location on the landscape, but also to the proportion of soil organic matter and mineral soil within the sample itself. This is especially true if changes due to acidic inputs are restricted to very narrow spatial scales within a soil (Fernandez, 1987,1989; Haun et af., 1988). The extent of spatial variability in soil physical and chemical properties in forest ecosystems is well characterized and typically exceeds 25% coefficient of variation (Mader, 1963; McFee and Stone, 1965; Ike and Cutter, 1967; Ball and Williams, 1968; Troedsson and Tamm, 1969; Beckett and Webster, 1971; Quesnel and Lavkulich, 1980; Federer, 1982; Neilsen and Hoyt, 1982; Arp, 1984; Arp and Krause, 1984; Riha et al., 1986a,b; Wolfe et al., 1987; Bringmark, 1989; Pallant and Riha, 1990). Attempts to measure differences in soil properties over time need to account for the inherent variability in most soil systems (McBratney and Webster, 1983; Webster and Burgess, 1984; Kratochvil et a f . , C. E. Johnson et af., 1990).

3. Forest Hydrology The forest canopy together with the various soil horizons, parent material, and underlying bedrock make up the forest watershed. As discussed in the previous two sections, the forest watershed can be considered as a series of chemical reservoirs that interact with atmospheric inputs that are transported with the drainage water (Schecher and Driscoll, 1989). The degree of chemical interaction and the residence time of the drainage water in each reaction zone determine the flux of ions through the watershed and eventually into the stream-lake environment (Chen ef al., 1984). Residence time within a given soil horizon will depend on its position on the landscape (Veneman and Bodine, 1982; Veneman et al., 1984; Roberge and Plamondon, 1987) and the number and direction of water flow paths present in a given volume of soil (Whipkey and Kirkby, 1978; Beven and Germann, 1982). It should not be assumed that in most forest ecosystems water movement will be confined to the vertical direction (Schecher and

WAYNE P. ROBARGE AND DALE W. JOHNSON

12

Driscoll, 1989). Lateral movement is common in forest ecosystems on hillslopes (Jones, 1987) and can account for a substantial portion of flow, especially on an event basis (Fig. 2) (Roberge and Plamondon, 1987; Gaskin et al., 1989; Hopper et al., 1990). Rapid movement through a given soil horizon will favor control of the soil solution by intensity factors and a decrease in the ANC of the drainage water (Chen et af., 1984). Longer retention within a soil horizon will increase the ANC of the soil solution, primarily through soil mineral weathering (Peters and Driscoll, 1987). Prolonged contact between inputs in drainage water and the soil along preferred flow paths will result in changes in the nearby soil that are not evident from analysis of the bulk soil

a] AVERAGE ANION FLUX

3 Y

P

TF

SF

FF

BA,

BAl

Btv

Btl

BCv

BCI

SAMPLING LEVEL

Figure 2. Average nutrient flux for all storms at each sampling level; P, precipitation; TF, throughfall; SF, stemflow; FF, forest floor leachate; BA,, BA soil solution, vertical component; BA, , BA horizon soil solution, lateral component; Bt, , horizon soil solution, vertical component; Bt,, Bt horizon soil solution, lateral component; BC,, BC horizon soil solution, vertical component; BC, , BC horizon soil solution, lateral component. After Gaskin et al. (1989).

ACID DEPOSITION O N FORESTED SOILS

13

(Hildebrand, 1986a,b, 1987, 1990; Jardine et al., 1990). This chemical disequilibria because of soil structure (Hildebrand, 1987) can result in an underestimation of acidification and cation leaching due to acidic inputs, and an overestimation of sulfate retention capacity (Caspary, 1990). Preferred flow paths will also influence the chemical composition of soil solutions as collected by zero-tension and tension lysimeters (David and Driscoll, 1984; Turner et al., 1985; Cozzarelli et al., 1987). Soil water collected by zero-tension lysimeters travels by preferred flow paths and will have a chemical composition more closely associated with the overlying surface horizons. Soil water collected by tension lysimeters is usually different in chemical composition and more representative of the surrounding bulk soil (Cozzarelli et al., 1987). The number and type of water flow paths in a soil pedon depend on a variety of soil properties, including particle size distribution (Kirkby, 1988), changes in saturated hydraulic conductivity with depth (Gaskin et al., 1989), depth to bedrock or other impermeable layers (Driscoll et al., 1985), the presence of root channels and animal burrows (Whipkey and Kirkby, 1978), and soil moisture conditions prior to an individual storm event or snow melt (Gaskin et al., 1989; Schecher and Driscoll, 1989; Swistock et al., 1989; Hopper et al., 1990; Vogt et al., 1990). All of these factors interact within a given watershed such that the contribution of quickflow (water that moves through the upper soil horizons) (Schecher and Driscoll, 1989) and of groundwater to the stream-lake environment will vary between seasons of the year and even between storms. During dry periods, the water table height decreases in most forest watersheds. Substantial inputs of rainfall, therefore, will first be retained in the bulk soil and will increase the height of the water table. Groundwater inputs to the stream-lake environment will dominate the surface water chemistry during this period of time (DeWalle et al., 1988; Gaskin et al., 1989; Swistock et al., 1989; Hopper et al., 1990). As the height of the water table increases, the zone of saturated soil will move upslope. Additional inputs of water into the soil pedon will encounter this zone of saturated soil, which effectively acts as an impermeable layer. The presence of this saturated zone, together with a change in saturated hydraulic conductivity with depth (Gaskin et al., 1989), results in more lateral flow, especially at the interface of the forest floor and underlying mineral soils (Gaskin et al., 1989; Swistock et al., 1989; Bishop et al., 1990; Hopper et al., 1990; Rosenqvist, 1990; Vogt et al., 1990). The particular upper soil horizon that contributes to quickflow depends on the antecedent soil moisture conditions (Gaskin ef al., 1989). As the chemical composition of soil water differs markedly between soil horizons (Fig. 2), the effect of quickflow on stream water chemistry will vary between storm events. The chemical

14

WAYNE P. ROBARGE AND DALE W. JOHNSON

composition of quickflow will also change as the effective drainage area of the watershed continues to increase with the increase in the height of the water table (Jones, 1987; Gaskin et al., 1989; Hopper ef al., 1990). In the northeastern watersheds of the United States, it is now postulated that, during these periods of high water tables within a forest watershed, a significant contribution of quickflow to stream water chemistry occurs in the headwaters of the watershed (where the soils are typically shallow, coarse textured, and highly acidic, with <15% base saturation) (N. M. Johnson et al. , 1981; Cronan et al., 1990). Water entering the soil increases rapidly in ionic strength, particularly due to the presence of the mineral acid anions NO, and SO:-. This solution, in turn, releases A1 from a lower soil horizon, which is the predominant source of quickflow during periods of high antecedent moisture content. The presence of NO, and SO:- allows the rapid transport of A1 through the soil horizon to the stream environment (Driscoll et al., 1988; Cronan et al., 1990). It is further postulated that failure to see similar A1 transport in southern watersheds of the United States, despite similarities in acidic atmospheric inputs and acidic soils with low base saturations, is primarily due to the retention of NO; and SO;- in these soils. Lack of these mobile anions decreases the amount of A1 that can be transported out of the southern watersheds (Cronan et al., 1990). The transport of acid soil waters to the stream-lake environment continues to be extensively studied and a number of models have been proposed to reproduce the hydrography and chemographs of reactive species observed in stream waters draining forest watersheds with substantial inputs of acidic deposition. These models include the Birkenes model (Christophersen and Wright, 1981; Christophersen et al., 1982); the enhanced trickle-down model (Nikolaidis et d . , 1988); end-member mixing analysis (Hopper et al., 1990); the ILWAS model (Goldstein et al., 1984; Gherini et al., 1985; Goldstein et al. , 1985); and the PULSE model (Bergstrom et al., 1985). A discussion of these models is outside the scope of this review (see Schecher and Driscoll, 1989), but the role of forest hydrology in each of these models serves to emphasize its importance when considering the effects of acidic deposition on forest soils and surrounding streamlake environments. It is also equally important to note that the increased acidity of quickflow leaving the surface soil horizons in forest watersheds does not necessarily require a change in the acidity of the bulk soil. The high porosity of the surface soil horizons (Gaskin et al., 1989; Rosenqvist, 1990; Cronan et al., 1990), combined with the dominance of intensity factors in controlling solution composition (Rosenqvist, 1990; Vogt et al., 1990), and the presence of the mineral acid anions NO; and SO:- (DeWalle et al., 1988; Driscoll er al., 1988; Lynch and Corbett, 1989), are

ACID DEPOSITION ON FORESTED SOILS

1s

sufficient to release A1 from naturally acid soils. This is especially true for soils with low base saturation (<15%) and minimal capacity to retain strong acidic anions (Cronan and Goldstein, 1989). In such soils there is a strong positive relationship between increasing ionic strength and concentration of A13+ in soil solution and quickflow (Cronan and Goldstein, 1989). In these forest ecosystems, which actively biocycle A1 (David and Driscoil, 1984; Turner et al., 1985; Rustad and Cronan, 1989; Cronan et al., 1990), it is the presence of the mineral acid anions NO, and SOf- that enhances A1 transport.

B. CHEMICAL FACTORS 1. Soil pH

Soil pH is a measure of the activity of H + ions in the soil solution (Reuss and Walthall, 1989). Although a great deal of emphasis is often placed on the absolute magnitude of a measured pH value, the actual worth is as an indication of soil reaction. Soil reaction is the degree of acidity or alkalinity of a soil, usually expressed as a pH value (Soil Science Society of America, 1987). Table I1 lists several of the descriptive terms often used to describe soil reaction, along with the respective ranges in soil pH. Also included in Table I1 is a summary of the reactions that can occur in soil that could possibly buffer against changes in soil pH (Schwertmann et al., 1987;

Table I1 Descriptive Terms and Proposed Buffering Mechanisms for Various Soil pH Ranges

Descriptive termsa

PH range

Buffering mechanismb

Extremely acid Very strongly acid Strong acid Moderately acid Slightly acid to neutral Slightly alkaline

<4.5 4.5-5.0 5.1-5.5 5.6-6.0 6.1-7.3

Iron range (pH 2.4-3.8) Aluminum/iron range (pH 3.0-4.8) Aluminum range (pH 3.0-4.8) Cation exchange (pH 4.2-5.0) Silicate buffers (all pH values typically > 5 ) Carbonate (pH 6.5-8.3)

7.4-7.8

"Source: Soil Science Society of America (1987). bSource: Schwertmann et al. (1987); Ulrich (1987); Reuss and Walthall (1989).

16

WAYNE P. ROBMGE AND DALE W. JOHNSON

Ulrich, 1987; Reuss and Walthall, 1989). These buffering mechanisms are different from the ANC of a soil in that they define specific chemical reactions that will resist changes due to H+ inputs, as well as control the activity of H+ in the soil solution. Identifying such buffering mechanisms provides one means for knowing which soils may be sensitive to rapid acidification. Soils primarily in the silicate buffer range may respond rapidly because of the relatively slow rate of release of cations from the weathering of primary silicates (Bringmark, 1989). Soils with low cation exchange capacity and circumneutral pH are also thought to be at risk, and will likely suffer a reduction in soil pH with acidic inputs (McFee, 1983). Most forest soils in humid temperate regions, however, are already acid due to natural acidification processes and fall in the aluminum or cation exchange buffering ranges (Fernandez, 1989). Defining exact buffering mechanisms for soil pH in these systems becomes difficult due to the overlap of the different chemical processes, and the presence of substantial amounts of soil organic matter (Bringmark, 1989). Defining exact numerical limits for buffer ranges also has limited value due to the uncertainty regarding the acid strengths of the different functional groups present in these acid soils (Schwertmann et al., 1987), and to the fact that soil pH values are actually operationally defined parameters that depend on moisture content, solid-solution ratios, and ionic strength of the added salt solution (Soil Science Society of America, 1987; Reuss and Walthall, 1989). For the majority of acid forest soils in humid temperate regions, the chemical processes that actually control soil solution pH are probably the acid-base reactions of soil organic matter and Al-hydroxy polymers, carbonic acid dissociation, and mineral-weathering reactions (Sposito, 1989). The acid-base reaction of soil organic matter and Al-hydroxy polymers will dominate in the surface soil horizons, whereas mineral-weathering reactions will dominate at deeper soil depths. The importance of carbonic acid dissociation will depend on biological activity, soil moisture content, and hydrology. The relationship between soil solution pH and soil organic matter can be expressed as

if we assume that the functional groups on the soil organic matter behave like a weak acid (Bolt et al., 1976). The terms [HA] and [A-] refer to the concentration of functional groups present per gram of soil or per gram of soil organic matter. The p K term is the negative logarithm of a combined dissociation constant (Sposito, 1989). Titration of forest soil organic hori-

ACID DEPOSITION ON FORESTED SOILS

17

zons results in a linear relationship between pH and the amount of acid or base added in the pH 4.0-7.0 range (Federer and Hornbeck, 1985; Magdoff and Bartlett, 1985). A similar relationship is observed for other sources of soil organic matter, such as sewage sludge (Sposito et al., 1982; Sposito, 1989). The linear relationship arises because the bulk of H + present is preferentially adsorbed to soil organic matter (Bolt et al., 1976). Negligible amounts are present in the soil solution or as nonselectively adsorbed or exchangeable ions. Thus, soil solution pH in soil horizons with substantial amounts of soil organic matter is primarily a function of the [HA] present per unit volume of soil (Federer and Hornbeck, 1985; Sposito, 1989), and cannot be described by simple ion exchange models (Krug and Isaacson, 1984; James and Riha, 1986). This explains in part why soil solution pH in forest soil organic horizons is not related to percentage base saturation (Magdoff and Bartlett, 1985; C. G. Wells and W. P. Robarge, unpublished data). Changes in ionic strength, such as from the presence of NO, and SO:from acidic deposition in throughfall and stemflow, influence H+ release or adsorption by soil organic matter through reduction of electric potentials in the counterion layer, and between functional groups on the solid phase (Bolt et al., 1976). Altering the interaction of functional groups effectively changes the value of the combined dissociation constant, resulting in either a release or adsorption of H+ ions. Increasing ionic strength generally favors release of H + ions, i.e., a decrease in soil solution pH. Changes in the total amount of [HA] can result from either the addition of acid, base, or soil organic matter. Because soil organic matter is derived in part from litter that accumulates in the forest floor, a change in the acidity or basicity of the litter will influence soil solution pH. Changes in litter composition can result from either a change in soil nutrient availability or in plant species (Ineson and Wookey, 1988; Choon and Roffael, 1990; Howard and Howard, 1990). Preferential adsorption of H+ by soil organic matter should result in an equivalent release of other metal cations to maintain electrical neutrality in the soil solution. The dominant metal cation released with H + addition for forest soil organic horizons from the northeastern United States is Ca, with lesser amounts of Mg, K, Mn, Fe, Zn, A1 and Cu (James and Riha, 1986, 1987). The reaction is relatively rapid, which is consistent with field observations described in Section III,A,3 (forest hydrology). Only Ca, however, shows a linear response with H + addition. Release of A1 is constant and independent of solid :solution ratios. Furthermore, between 40 and 80% of the added H+ is retained by the soil horizon without an equivalent release of metal cations (James and Riha, 1986). The lack of an equivalent release of metal cations is thought due to the protonation of

18

WAYNE P. ROBARGE AND DALE W. JOHNSON

amino groups contained within the soil organic matter (James and Riha, 1986). Electroneutrality is maintained by adsorption of the accompanying anion or by dissolution of soil organic matter (Krug and Isaacson, 1984; James and Riha, 1986). These observations reinforce the general conclusion that forest soil organic horizons do not behave as simple ion exchanges in response to acidic inputs. Carbonic acid influences soil solution pH by being both a proton donor and a proton acceptor. In the absence of bulk soil, increasing the partial pressure of COz will lower solution pH according to the following relationship: pH = 7.8 + log(HCO7) - log pCO2

(4)

where pCOz is the partial pressure of COz, (HCO,) is the activity of the bicarbonate ion in solution, and the number 7.8 is a numerical constant (Reuss and Walthall, 1989; Sposito, 1989). Because the partial pressure of C 0 2 in forest soils (0.3-0.8% pC02) often exceeds that in the open atmosphere (0.03% pCOz) by an order of magnitude or more (Novozamsky and Beek, 1976; Fernandez and Kosian, 1987; Castelle and Galloway, 1990), carbonic acid dissociation has the potential for exerting a substantial influence on soil solution pH. Limiting this potential influence of carbonic acid on soil solution pH is the fact that carbonic acid dissociation is only significant above pH 4.5, depending on the pCOz (Reuss and Walthall, 1989). In addition, the buffer capacity of the bulk soil will most likely exceed the buffer capacity of the carbonic acid. Other than by contact with quickflow through preferred flow paths, it is unlikely that carbonic acid will be the primary agent controlling soil solution pH in most forest soils of the humid temperate regions (Reuss and Walthall, 1989). Increasing the partial pressure of C 0 2 does increase the concentration of HCO; in soil solution, however, which in turn increases the ANC of the soil solution and enhances the potential for cation leaching. This occurs even in acid soil horizons (pH 4 . 0 ) (David and Vance, 1989), which is only possible if there is dissociation of carbonic acid (Stumm and Morgan, 1981). Laboratory studies suggest that soil solution pH remains essentially constant with increasing partial pressure of COz (David and Vance, 1989). The subsequent increase in ANC must, therefore, result from the preferred adsorption of H + ions donated by carbonic acid. When the soil solution is no longer in contact with an elevated partial pressure of COz in the soil atmosphere, either by movement to the stream-lake environment or to another soil horizon (Fernandez and Kosian, 1987), the solution will degas. This process will consume H+ ions and result in a decrease in ANC and an increase in solution pH (Reuss and Johnson, 1985; David and Vance, 1989). The total anion concentration of the solution will also

ACID DEPOSITION ON FORESTED SOILS

19

decrease, which means that the cations accompanying the HCO; ions will be removed from solution as well. The influence of Al-hydroxy polymers and mineral weathering reactions on soil solution pH will be discussed in later sections as part of the overall chemical reactions involving these solid phases in the bulk mineral soil. The importance of these solid phases in regard to soil solution pH is confined to the mineral soil horizons because of the large difference in buffer capacity in the presence [lo-40 cmol (H+) kg-' pH-l] as compared to the absence [0.3 cmol (H+) kg-' pH-l] of soil organic matter (Federer and Hornbeck, 1985; James and Riha, 1986). The large buffer capacity of forest soil organic horizons means that the pH of the soil solution entering the underlying mineral horizons, in most cases, will not be directly influenced by H' inputs from acidic deposition. As referred to earlier, the composition of the soil solution passing through the organic horizons in forest soils will largely be determined by the combination of the mineral acid anions NO, and SO!- and the acidity and basicity of the litter, which is the primary source of the soil organic matter. 2. Cation Exchange and Aluminum Solubility Titration of forest soil mineral horizons results in an S-shaped relationship between pH and the amount of acid or base added in the pH 4-7 range (Federer and Hornbeck, 1985; Magdoff and Bartlett, 1985). The S-shaped curves indicate that these mineral soil horizons are well buffered at about pH 7 and below pH 4, but are poorly buffered between pH 5 and 5.6 (Magdoff and Bartlett, 1985). These observations agree with a broader survey of the pH values of 927 soils that reflect three soil buffer ranges (carbonate, pH 5.5 and 7.0; pH-dependent charge, pH 4.6-5.6; A1 release, pH 3.5-4.5) (Schwertmann and Fischer, 1982). The pH-dependent charge, or cation exchange buffering range, and the Al-release buffering range are actually interrelated, as they both depend on the nature of the exchange surfaces present in the mineral soil horizons. These exchange surfaces, together with the ionic composition of the soil solution, interact in a series of reactions to control the concentration of A1 and other base cations in the soil solution (Reuss, 1983; Reuss and Johnson, 1986; Reuss and Walthall, 1989; Reuss et al., 1990). How these reactions interact to control solution pH as a function of base saturation (BS) of the mineral soil exchange surface is shown in Fig. 3 (Robarge and Corey, 1979). With increasing BS, there is a relatively rapid increase in pH until approximately 20% BS. Between 20 and 65% BS the pH is buffered, although the actual pH observed for the midpoint of the pH buffer range is a function of ionic strength and neutral salt cation. Beyond

WAYNE P. RORARGE AND DALE W. JOHNSON

20

81

_...'.""' .

71

.-.*..."'

*...ow-...-0'''

_..

I n --t

Cap=0.0003

Na p0.003

0

20

40

60

80

100

Percent Neutralization Figure 3. Effect of ionic strength ( p = 0.0003/0.003)and neutral salt cation (Ca versus Na) on solution pH for an Al-saturated sulfonic cation exchange resin at various percent neutralizations. Adapted from Robarge and Corey (1979).

65% BS, pH buffering is minimal until above pH 7. The pH buffering at about pH 7, and the lack of buffering between pH 5 and 5.6, agree with the observations of Federer and Hornbeck (1985) and Magdoff and Bartlett (1985). Figure 3 is based on a model system using only an Al-saturated cation ion exchange resin, with base added as bicarbonate salts. Figure 3, therefore, supports the general conclusion that, below pH 7, the actual pH observed in soil solution for forest soil mineral horizons is primarily a function of the cation exchange and Al-release buffering ranges (Federer and Hornbeck, 1985; Magdoff and Bartlett, 1985; Hartikainen, 1986; Schwertmann et al., 1987; Ulrich, 1987; Reuss and Walthall, 1989). Figure 3 also demonstrates that assigning absolute pH values to these buffer ranges is confounded by the ionic strength of the soil solution. Lack of a pH buffering range below pH 4 is due to the absence of a separate A1 solid phase in the resin model system. These observations on natural and model systems illustrate that chemical models that are designed to predict the effects of acidic deposition on natiirally acid forest soil mineral horizons must take into account both cation exchange reactions and Ai solubility reactions. In most naturally acid forest soil mirleral horizons, Ca, Mg, and A1 are the dominant ions on the exchange surface. The ion exchange reactions between these three ions can be described by cation exchange reactions with the appropriate conditional or thermodynamic equilibrium constants (Sposito, 1981, 1989; Bolt, 1982). For simplicity only A13+ and Ca2+ will be included here, but

ACID DEPOSITION ON FORESTED SOILS

21

similar exchange reactions can be written for Mg, K, and Na. Following the approach of Reuss and his co-workers (1990), the exchange reaction between Ca2+ and A13+ ions can be written as (Ca2+)3[A1X]2 Kgt = (A13')2[CaX]3

where [AIX] and [CaX] denote the exchange fractions on the exchange surface, (Ca2+) and (A13+)refer to the activities in the soil solution, and Kgt is the equilibrium constant or selectivity coefficient. Rearrangement of Eq. ( 5 ) to

illustrates how the activity of the A13+ ion in solution is a function of the nature of the exchange surface (&), the BS [CaX] or acid saturation [AlX] of the exchange surface, and ionic strength of the soil solution (Ca2+). The nonlinear dependence on the activity of A13+ in solution is particularly important in regard to the nature of the base cation. Soil solution A13+ concentration increases to the 3/2 power of soil solution Ca2+ or Mg2+ concentration, and to the third power for soil solution K + , Na+, or H+ (Reuss, 1983). The exponents in Eqs. ( 5 ) and (6) are a function of the formulation of the exchange reaction and vary depending on the representation chosen for the exchange fraction. Equation ( 5 ) is based on the exchange relationship proposed by Gaines and Thomas (1953) and appears to be more consistent with the observed Ca2+-A13+ relationship in forest soil mineral horizons (Reuss et al., 1990). Other possible formulations for the exchange reaction are reviewed by Bolt (1982) and Sposito (1981, 1989). The relationship between H + and A13+ ions in most forest soil horizons can be described as follows: (AP')

= KA(H+)"

(7)

where (A13+) and (H+) refer to the activities in the soild solution and K , and a are constants (David et al., 1988). Taking the log transformation of Eq. (7) yields the more commonly cited expression between pAl and pH (Reuss and Johnson, 1986; Sposito, 1989), lOg(KA) = apH - PA1

(8)

where KA is now the equilibrium constant and a is the degree of hydrolysis of the A13+ ion. If an Al-trihydroxide or an aluminosilicate mineral solid

22

WAYNE P. ROBARGE AND DALE W. JOHNSON

phase is controlling the activity of A13+ in soil solution, then the relationship between H+ and A13+ ions can be expressed as log Ksp= 3pH - PA1

(9)

where Ksp is the solubility product (Reuss and Johnson, 1986). For gibbsite, an Al-trihydroxide species, the value of log K,, will vary between 8.0 and 9.7, depending on crystallinity (Lindsay, 1979), and the ApAl/ ApH should equal 3.0 (Reuss and Johnson, 1986). Gibbsite is often assumed to be the final control of the activity of A13+ in soil solutions because of its physical presence in forest soil mineral horizons (e.g., Cozzarelli ef af., 1987) and because of the agreement between measured and predicted activities of A13+ in soil solution (David and Driscoll, 1984; Cozzarelli et af., 1987; Matzner, 1989) and the stream-lake environment (N. M. Johnson et al., 1981; Turner ef al., 1985) when plotted on activity ratio diagrams as a function of pH. Chemical models designed to predict changes in the activity of A13+ in soil solution due to acidic inputs often assume that gibbsite is the controlling phase for the activity of A13+ (Reuss and Johnson, 1986; Schecher and Driscoll, 1989). Amorphous Altrihydroxides have a higher solubility than crystalline gibbsite and are known to be present in forest soil mineral horizons (Cozzarelli et al., 1987), but these solid phases should still have ApAllApH values approaching 3.0. Laboratory mitasurements with soil (Bache, 1974; Bloom et af., 1979) and model (Robarge and Corey, 1979) systems, however, find that ApAl/ApH is <2.5 in the pH 3.7-5.2 range. Furthermore, plots of log(A13+)versus pH for extracted soil solutions yield values of 2.29 to 2.60 for log Ksp, and 1.60 to 1.66 for the coefficient of pH in Eq. (9) (Bloom and Grigal, 1985; Reuss et al., 1990; Walker et al., 1990). These results suggest that a separate solid phase is not controlling the activity of A13+ in soil solutions, at least in the upper surface horizons (Bloom and Grigal, 1985). Additional support for this conclusion comes from the fact that dissolution of Al-trihydroxide species may be a relatively slow process, complicated by the presence of adsorbed inorganic and organic anions (Furrer and Stumm, 1983) and structural defects in natural gibbsite (Bloom and Weaver, 1982). Laboratory experiments with gibbsite suspensions do not attain equilibrium in 384 hr when maintained at pH 4 (May ef al., 1979), and only reach equilibrium when the substantial increase in activity of H+ ions at lower pH values increases the rate of dissolution (Bloom, 1983). The premise that gibbsite or other solid phases control the activity of A13+ in soil solutions and stream-lake environments based on the agreement between measured and calculated activities in activity ratio diagrams is also under question. It is argued that typical activity ratio diagrams have dependent axes and that relationships between measured and

ACID DEPOSITION ON FORESTED SOILS

23

calculated activities when plotted against such axes are autocorrelative (Neal et al., 1987; Neal, 1988). When experimentally measured activities of A13+ and H+ ions are plotted in activity ratio diagrams with independent axes, the presumed relationships with gibbsite or other A1 solid phases are no longer observed (Neal, 1988). Equation (7) does not require the presence of a separate A1 solid phase to be valid and the relationship between A13+ and H+ ions in solution can be maintained by hydrolytic reactions involving mononuclear A1 and the formation of Al-hydroxy polymers. The existence of Al-hydroxy polymers is now accepted (Bertsch, 1990), and they have been found to be stable over a relatively wide pH range. They also may be more toxic to plants than the hexaqua A13+ cation (Parker et al., 1988, 1989). These polynuclear cations have a high charge density and in the presence of a negative exchange surface are rapidly adsorbed from solution (Bloom el al., 1979; Robarge and Corey, 1979;Hodges and Zelazny, 1983; Jardine et al., 1985). The removal of these polynuclear cations from solution to an exchange surface will effectively control the activity of H + and A13+ ions in solution (Robarge and Corey, 1979; W. P. Robarge, unpublished data). The formation of an A1 polynuclear cation in soil solution can be represented by

where K is the formation constant and the remaining terms are the activities of the respective species in solution. Polynuclear A1 complexes are not removed from ion exchange surfaces by leaching with normal monovalent salt solutions (Hsu and Rich, 1960; Hsu and Bates, 1964; Turner, 1967). Thus, by adsorption onto the ion exchange surface, the polynuclear A1 cation effectively becomes part of the solid phase. Setting its activity to a value of one and rearranging Eq. (10) after substituting H + and K , (the ionization constant for water) for OH- yields the desired relationship between the A13+ and H' ions: log Ki2Kp1l6= 2pH - pAl

(11)

Both pH and pAl are controlled in solution because of the ion activity product of the polynuclear cation, the presence of the ion exchange surface, and the increase in the stability of the positively charged polynuclear A1 cation once sorbed in the presence of a negative field. The formulation for the polynuclear A1 cation in Eq. (10) is for illustration purposes only, and the coefficient for the pH term in Eq. (11) can vary between 1.0 and 2.7, depending on the composition of the polynuclear A1 cation selected (Bertsch, 1990). The formation of the polynuclear

24

WAYNE P. ROBARGE AND DALE W. JOHNSON

A1 cation in forest soil mineral horizons requires inputs of base to neutralize the H+ ions released during hydrolysis. This base is supplied in the form of HCO; ions, dissolved organic matter, and, to a certain extent, weathering of primary minerals, depending on the composition of those present in the bulk mineral soil. Hydrolysis of A13+ ions is enhanced by the ion exchange surface (Jardine et al., 1985) and is an often-cited problem in the preparation of A13+-saturated clays (e.g., see Veith, 1978). Simple passage of dilute soil solutions through the soil mineral horizons will enhance the hydrolysis of A13+ ions on exchange surfaces. The formation of the polynuclear A1 cation will increase the ANC of the bulk soil and provide a surface that will react rapidly to changes in H + ion concentration in solution because of its.position on the exchange complex. The significance of Eq. (11) in terms of formulating chemical models lies in the fact that the formation and subsequent removal from solution of the polynuclear A1 cation provides the necessary relationship between H+ and A13+ ions, without the need for invoking the presence of a separate A1 solid phase, or separate exchange reactions between H', A13+,and Ca2+ (Reuss and Johnson, 1986; Reuss and Walthall, 1989), at least in soil mineral horizons that are relatively low in organic matter content. In forest soil organic horizons, the ion exchange relationship between H+ and A13+ cannot be ignored because of the relationship between soil organic matter and H+ ions, as discussed in the previous section, and the known affinity between divalent and trivalent cations and soil organic matter (Bloom et af., 1979; Bloom, 1981; Bruggenwert and Kamphorst, 1982; Sposito, 1989). Organic A1 species are the dominant A1 species in soil solutions from soil organic horizons (Ugolini et al., 1977; Nilsson and Bergkvist, 1983) and solid-phase soil organic matter is known to adsorb both monomeric and polymeric A1 species (Bloom et al., 1979; Bloom, 1981; Jardine et al., 1985; Young and Bache, 1985; Thomas, 1988). Like the Al-hydroxy complexes, the exchange of A13+ and H + with soil organic matter provides the means for satisfying the requirements of Eq. (7) without the need for a separate A1 solid phase, especially in acidic surface organic horizons that have a low amount of permanent cation charge relative to the amount of soil organic matter present (Bloom et al., 1979). Using the approach of Cronan and his co-workers (1986), the exchange between H+ and A13+ in forest soil organic horizons can be expressed by an empirical model similar to Eq. (8), pAl = b(pH) + c

(12)

where both b and c are constants obtained from the regression lines for the plots pAl versus pH. The numerical value of both b and c are linearly related to the bound A1 ratio (moles of bound Al/total moles of carboxyl

ACID DEPOSITION ON FORESTED SOILS

2s

groups) in the soil organic matter. As indicated by Eq. (12), increasing the bound A1 ratio will result in an increase in the value of b , which will, in turn, mean an increase in the activity of A13+ in solution at any given pH value (Fig. 4). The relationship predicted by Eq. (12) is generally consistent across a variety of soil organic horizons from different geographic regions (Walker et al., 1990), and demonstrates that the equilibrium solubility of A1 in such surface horizons is dependent on soil solution pH and the degree to which the soil organic matter is saturated with Al. This relationship explains the “undersaturation” of soil solution A13+ with respect to gibbsite often observed in forest soil organic horizons (Fig. 4) (Cronan et al., 1986). The limit of applicability for this empirical relationship is proposed to be

6.0

5.0

e

._ ._ c

4“ +

m -

% 4.0

3.0

2.0

3.0

D

4.0

5.0

6.0

PH

Figure 4. Stability field for apparent A1 solubility in the presence of soil humic material in the pH range 3.0-5.0. The points along each apparent solubility line represent equilibrium pA13+ values of Oi and Oa horizons from an Adirondack Spodosol. These horizons were pretreated to yield the following four different ranges of bound aluminum ratio: A = 0-0.1; B = 0.2-0.3; C = 0.6-0.7; D = 0.9-1.0. The solubility lines for interlayer and amorphous AI(OH)3 at 25°C are shown for reference, After Cronan et af. (1986). Reprinted by permission from Nature, Vol. 324, p. 142. Copyright 01986 Macmillan Magazines Ltd.

26

WAYNE P. ROBARGE AND DALE W. JOHNSON

pH 5.2. Above this p H it is felt that amorphous Al-trihydroxide will become the Al-controlling phase (Walker et al., 1990). The limit of pH 5.2 is consistent with the cation exchange and aluminum-release buffering ranges in Table 11. Regardless of the exact mechanism controlling A1 solubility in forest soil horizons, Eqs. ( 5 ) and (7) can be combined to generate the curvilinear relationship between BS and pH shown in Fig. 3. Using the convention of Reuss and his co-workers (1990), the combined equation

can be obtained by substituting the right-hand side of Eq. (7) for (A13+)in Eq. (5). Taking the log transformation and solving for pH yields the desired relationship between pH and BS, [Cax]’ pH = (1/2a) log(K,,Ki) - (3/2a) log(Ca2+) + (1/2a) log [A1Xl2) (14)

(

if it is assumed that Ca2+ is the dominant base cation on the exchange surface (Reuss et al., 1990). The exact shape of the curvilinear relationship predicted by Eq. (14) is a function of the constant a , whereas the position of the curve on the pH and BS axes is a function of K,,, KA, and ionic strength [(3/2a) log(Ca2+)]. The inverse relationship between pH and (Ca2’) agrees with Fig. 3, in that an increase in ionic strength of the solution at a fixed BS will decrease pH and, as predicted by Eq. (7), result in an increase in the activity of A13+ in solution. The change in pH with change in neutral salt cation at equal ionic strengths is due to a change in the value of K g t . At constant ionic strength, the terms - (3/2a) log(Ca2+) are constant, and observed pH is (1/2a) log(K,,K:) essentially a function of base saturation (Bloom and Grigal, 1985). Equation (14) expresses in numerical terms the fact that soil solution composition in forest soil horizons is a function of a number of interacting factors, including BS, pH, ionic strength, metal solubility, and cation and anion retention (Cronan and Schofield, 1990). Furthermore, Eq. (14) demonstrates that relatively large changes in a given soil parameter, such as BS, are not necessary to have a significant influence on the composition of the soil solution. It is reasonable to expect a change in BS in soil systems that have received substantial inputs of acid-forming N and S compounds over an extended period of time (Matzner, 1989), but in many regions receiving acidic deposition, a decrease in BS does not appear to be a widespread induced biological stress (Richter et al., 1988). A more direct influence in soil solution composition, as reflected in Fig. 2 and as dis-

ACID DEPOSITION ON FORESTED SOILS

27

cussed in the section on forest hydrology, is the increase in ionic strength of the soil solution due to the presence of the mineral acid anions NO; and SO:- (Van Miegroet and Cole, 1985). Relatively small changes in ionic strength (0 to 0.6 mM,/liter of added salt solution) have been shown to have the largest effects in decreasing soil solution pH in forest soil horizons (Richter et a l . , 1988). As predicted by Eq. (14) and other model systems (Seip, 1980; Reuss, 1983), this change in ionic strength will influence the composition of the soil solution (Van Miegroet and Cole, 1985). Aluminum ions released due to the increase in ionic strength will be transported deeper into the soil profile or into the stream-lake environment due to the presence of the mobile NO; and SO:- anions and absence of HCO; ions (Driscoll and Schecher, 1990). Recent field observations in support of the importance of cation exchange reactions in controlling processes that lead to short-term acidification of runoff can be found in the initial results of the Reversing Acidification In Norway (RAIN) project (Wright ef a)., 1988a). This project, which has been conducting whole-catchment manipulation acidloading experiments, is designed to test the premise that a reduction in emissions of SO2 and NO, will eventually result in the restoration of acidified waters. Acid loadings at one pristine catchment (Sogndal site) at a rate comparable to that found in southern Scandinavia [l kmol (H+) ha-’ yr-’] has resulted in rapid changes in soil solution chemistry after only 4 years of treatment. During the first year, each acid rainfall event produced short-term episodes of acid, Al-rich runoff, after which there was rapid recovery to preevent concentrations. Continued additions for the past 3 years, however, has seen a steady deterioration in the ability of the soil system to recover, with a 50% decrease in the alkalinity of the soil solution. Sulfate concentrations in the soil solution have nearly doubled and are already at 50% of the calculated value at which it is expected that the SO:- flux in the runoff will equal SO:- deposition. The increase in SO:- concentration in the soil solution has seen a corresponding increase in the concentration of base cations (mainly Ca2+ and Mg2+), which is consistent with soil cation exchange processes observed for “seasalt effects” at a headwater catchment adjacent to the Sogndal site (Wright e t a ) . , 1988b). Exclusion of ambient precipitation from an acidified catchment in southernmost Norway (Risdalsheia site) [typical loading rate 1 kmol (H+) ha-’ yr-’1 has resulted in lower concentrations of NO; and SO:- in runoff (Wright qt al., 1988a). Decreases in NO; have been the most dramatic, with concentrations decreasing by 60% within 2 weeks of excluding ambient precipitation from the treatment area. The decrease in SO:- has been more gradual, and has been accompanied by a 55% decrease in the

28

WAYNE P. ROBARGE AND DALE W. JOHNSON

concentration of base cations in soil solution and a 45% increase in alkalinity. There has also been a decrease in A1 concentration in the soil solution, and an increase in the degree of dissociation of the dissolved organic acids, reflecting the reappearance of HCO; and dissolved organic matter as the dominant anions in the soil solution. This relatively rapid recovery in a previously acidified watershed by effectively lowering the ionic strength of the soil solution supports the hypothesis that cation exchange processes control the concentration of ions in soil solution at the two RAIN project sites (Wright, 1988). The relatively rapid response to acidic inputs in sensitive soils observed in the RAIN project suggests that similar rapid changes may also occur in plant nutrient availability and uptake. Such changes may appear as anomalies in the chemical composition of plant parts, such as the bole wood of trees, that serve as passive monitors of soil nutrient status over time (Momoshima and Bondietti, 1990). In long-lived, relatively undisturbed ecosystems, the cycling of base cations, especially Ca2+,between the forest floor and the overlying canopy is relatively constant (Momoshima and Bondietti, 1990). Inputs of acidic deposition represent a disturbance in this steady state. As observed in the RAIN project, the initial disturbance could be an enhancement in the mobilization of base cations in the soil solution, which would be followed later by a decrease in soil solution pH and an increase in the concentration of Al. Such an increase in the availability of base cations should be reflected in the chemical composition of bole wood from forest ecosystems impacted by acidic deposition. Analysis of bole wood chemistry of several species of trees from Maine to Tennessee in the eastern United States suggests that an increase in the availability of base cations did occur in forest ecosystems that are now known to receive substantial inputs of acidic deposition (Bondietti et al., 1990). The data suggest that the changes in sap chemistry necessary to cause an increase in the base saturation of the bole may have occurred as early as the mid-1940s. These inferred changes in sap chemistry would be the result of an increase in the availability of base cations in the soil solution such as observed in the RAIN project. By the mid-l960s, there is a decrease in base saturation of the bole wood, which would suggest a decrease in base cation availability. A decrease in base cation availability does not require a decrease in the concentration of base cations in the soil solution. An increase in the A1 concentration in the soil solution will reduce the uptake of base cations, especially Ca2+, through competition for uptake sites in fine tree roots (Shortle and Smith, 1988). Such an increase in A1 concentration would be expected as the inputs of acidic deposition increased the ionic strength of the soil solution in already naturally acid soils. The increased competition between A1 and base cat-

ACID DEPOSITION ON FORESTED SOILS

29

ions for uptake sites in the fine roots would result in essentially a Ca deficiency syndrome (Shortle and Smith, 1988). A deficiency of Ca would in turn lead to a decrease in the amount of functional sapwood, and lower the resistance of large trees to common diseases and insect pests (Shortle and Smith, 1988). The decrease in the base saturation of bole wood observed by Bondietti and his co-workers (1990) in the mid-1960s is consistent with the regional decrease in radial increment growth of declining red spruce in eastern North America. In at least one location currently receiving substantial inputs of acidic deposition (the Great Smoky Mountains National Park, Tennessee), analysis of bole wood chemistry from red spruce indicates an increase in A1 availability has occurred over this same period (Bondietti et al., 1989). The approach of Bondietti and his co-workers (1989,1990) provides one means of estimating changes in historical patterns of cation supply in forested soils that cannot be obtained from current measurements of tree foliage or soil solutions. Although the interpretation of the data obtained by this approach is not necessarily straightforward (Momoshima and Bondietti, 1990), the data strongly suggest that inputs of acidic deposition to naturally acid soils result in relatively rapid changes in soil solution chemistry that influence both the plant root and surrounding stream-lake environments. Such changes are consistent with those predicted by Eq. (14) and illustrated (in part) in Fig. 3. 3. Retention of Nitrate and Sulfate

Retention of NO, and SO;- in forest soil horizons occurs primarily by a series of inorganic and biologically mediated chemical reactions (Swank, 1986). These chemical reactions are part of the acid-base relationships of the N and S cycles in forest ecosystems (Reuss and Johnson, 1986). The discussion in this section is limited to those reactions that influence the concentration of NO, and SO$- in soil solution, and that may restrict or enhance the leaching of base cations and the mobilization of toxic cations. More detailed reviews of the N and S cycles in forest ecosystems can be found elsewhere (van Breemen et al., 1983; Reuss and Johnson, 1986; Swank, 1986; Binkley and Richter, 1987; De Vries and Breeuwsma, 1987; Gundersen and Rassmussen, 1990). a. Nitrates The dominant NO, retention mechanism in forest soil horizons is biological uptake, especially in the rooting zone (Keeney, 1980; Johnson et al., 1982; Swank, 1986; Agren and Bosatta, 1988). Nitrogen retention is high because most forest ecosystems have N as the element limiting

30

WAYNE P. ROBARGE AND DALE W. JOHNSON

growth (Binkley and Richter, 1987; Johnson and Lindberg, 1989). Nitrate removal from soil solution, therefore, will occur during the growing season when the demand for N by the biological community exceeds N inputs by atmospheric deposition and release of N from mineralization of soil organic matter. Nitrogen saturation of an intact forest ecosystem occurs when the external and internal sources of N exceed demand (Agren and Bosatta, 1988). In these systems, NO; will be one of the dominant anions in the soil solution and mobilization and leaching of cations will increase (Johnson et al., 1983; Berdkn et al., 1987; Gundersen and Rasmussen, 1990). Nitrogen saturation can be achieved by saturation of the N uptake capacity of the plant community within an ecosystem, or by saturation of the internal cycling of N within the forest soil (Agren and Bosatta, 1988). In most forest ecosystems, the uptake capacity of the plant community greatly exceeds that of the soil microbial population (Skeffington and Wilson, 1988). Mineralization of soil organic matter, followed by the biological oxidation of NHZ to NO; (nitrification), is the primary release mechanism of NO; to the soil solution in forest soils (Gundersen and Rasmussen, 1990). The conversion of NHZ to NO; can be expressed as 0

NH: 1 ,NO; H20

+ 2H+

(15)

with H+ ion generated to maintain electrical neutrality (Johnson et al., 1983). Mineralization, therefore, is a natural acidification process that has the potential to decrease soil solution pH. The resulting increase of H t ions in soil solution is the driving force for the increase in mobility and leaching of cations that accompanies mineralization and nitrification (Reuss and Johnson, 1986). The rate of mineralization is a function of several soil variables: C/N ratio of the soil organic matter (Gundersen and Rasmussen, 1990), soil temperature, pH and antecedent moisture conditions (Ulrich, 1985; Hogberg et al., 1986; Berdtn et a f . , 1987; Gundersen and Rasmussen, 1990; Stams and Marnette, 1990), and soil disturbance (Swank, 1986; Lawrence et af., 1987; Fuller el al., 1988; Lawrence and Driscoll, 1988). In forest ecosystems that are known to fix N [e.g., red alder (Alrnus rubra Bong)], there can be substantial releases of NO; with a significant decrease in soil solution pH (Van Miegroet and Cole, 1984). Increases in soil temperature, soil pH, and/or soil moisture content will also result in an increase in mineralization and a release of NO: and H' to the soil solution (Swank, 1986). Such seasonal acidification pushes (Ulrich, 1983) are known to occur in the spring and in the fall, especially after prolonged dry periods during the summer (Ulrich, 1985; Khanna et al., 1987; Matzner, 1989; van

ACID DEPOSITION ON FORESTED SOILS

31

Breemen et al., 1989; Johnson et al., 1991b). In acid soils that lie within the cation exchange and A1 buffering range, the seasonal acidification pushes can result in a marked increase in soil solution A1 and other base cations. Nitrate is often the dominant anion in soil solution during these periods (Johnson et al., 1991b). Sustained pushes of acidity are often observed after major forest ecosystem disturbances, such as whole-tree harvesting (Swank, 1986; Lawrence et al., 1987; Fuller etal., 1988; Lawrence and Driscoll, 1988). The decrease in vegetative N uptake combined with the increase in mineralization and nitrification results in a significant elevation in the concentration of NO; and cations in the soil solution. Observations at the Hubbard Brook Experimental Forest in New Hampshire have confirmed that this release of N to the soil solution can have a direct influence on the stream-lake environment draining a disturbed watershed. Stream water composition following whole-tree harvesting changes rapidly with a large increase in NO, and base cation concentrations. This is followed by a decrease in stream pH accompanied by an increase in A1 concentration. The A1 released under these conditions appears to be entirely in the inorganic form, resulting in potentially toxic concentrations in the stream-lake environment (Lawrence er al., 1987; Lawrence and Driscoll, 1988). Many forest soils have a potential for high mineralization and nitrification rates (Swank, 1986; Strader et al., 1989; Montagnini et al., 1989b; Becquer er al., 1990; Donaldson and Henderson, 1990; Gundersen and Rasmussen, 1990; Johnson et al., 1991b), but internal processes act to hold this potential in balance (Ulrich, 1983; Swank, 1986; Montagnini et al., 1989a). These internal processes also limit the overall acidification effect of nitrification, such that in nitrogen-stable ecosystems there is little net change in the acidity of the bulk soil or soil solution (Ulrich, 1983). Continued inputs of atmospheric N deposition, however, will increase the total N pool in an ecosystem and will eventually overwhelm the resistance of the ecosystem to N cycle disruptions (Gundersen and Rasmussen, 1990). The net result will be a decrease in soil solution pH and enhanced leaching of base cations that is not due to the immediate presence of NO; ions in throughfall and stemflow, but is due to the increase in the total pool of N available for mineralization: i.e., the internal cycling of N becomes saturated (Agren and Bosatta, 1988; Foster et al., 1989). Nitrogen saturation within nitrogen-unstable ecosystems, therefore, is not associated with static soil parameters [such as C : N ratio (Billett et al., 1990a)], but with the dynamics of N cycling within the ecosystem (Agren and Bosatta, 1988; Gundersen and Rasmussen, 1990). Within nitrogen-unstable ecosystems, the acidification pushes that accompany mineralization and nitrification may damage root systems and further reduce N uptake by the canopy

32

WAYNE P. ROBARGE AND DALE W. JOHNSON

(Matzner, 1989). Reduction of N uptake translates into enhanced leaching and potential transport of NO, and toxic cations to the stream-lake environment. Additional disruptions in nitrogen-unstable ecosystems, such as prolonged periods of drought or the loss of overhead canopy due to the forest decline, will only aggravate the loss of NO; from the ecosystem (Gundersen and Rasmussen, 1990). Nitrate export from such nitrogenunstable ecosystems should increase dramatically (Henriksen and Brakke, 1988a; Foster et al., 1989; Hauhs, 1989) due to the uncoupling of the deacidification processes of the bulk soil from the acidification processes of nitrification (Ulrich, 1986). Nitrogen saturation of the internal N cycling processes of a forest soil will not occur if inputs of N into a forest ecosystem, including those from acidic deposition, are approximately equal to or less than N outputs (Agren and Bosatta, 1988). If the N output for a given forest ecosystem is known, then it should be possible to calculate a critical load of N deposition (Nilsson, 1986, as cited in Skeffington and Wilson, 1988) that will not result in any long-term adverse effects. Though defining critical loads offers one approach toward regulation of N emissions, actually calculating such numbers is difficult due to the complexity of the N cycle in a forest soil, the diversity of various forest ecosystems, and the lack of agreement for what is meant by adverse effects (Skeffington and Wilson, 1988). Critical loads proposed for Scandinavia vary from 5 to 20 kg (N) ha-' yr-', depending on the productivity of the forest operation and the method of calculation (Anderson, 1989). De Vries (1988) proposed a critical load of below 4 kg (N) ha-' yr-' for Dutch forest ecosystems. Actual deposition of N compounds is approaching 30-40 kg ha-' yr-' in many areas of Central Europe and exceeds 20 kg ha-' yr-' in the southern parts of Scandinavia (Grennfelt and Hultberg, 1986). In certain areas, the actual N deposition may be twice as high. In the eastern United States, deposition of N varies from 10 to 20 kg (N) ha-' yr-', especially in the high-elevation spruce-fir forests of the Appalachian mountains (Bruck et al., 1989; Aneja et al., 1991; Johnson et al., 1991b). These figures suggest that with current levels of N deposition in Europe and the eastern United States the gradual N loading of forest soils will continue, with the eventual saturation of internal N cycling processes leading to an increase in soil acidification. b. Sulfates Sulfate is retained much more strongly than NO, in forest soils through a series of inorganic reactions, including adsorption onto mineral surfaces, precipitation of aluminosulfate minerals, and anion exchange (Bolt et al., 1976; Tabatabai; 1987; Reuss and Walthall, 1989). Sulfate is adsorbed by various A1 and Fe oxides and hydrous oxides, and it has been assumed that this is the primary retention mechanism for SO:- from acidic inputs in

ACID DEPOSITION ON FORESTED SOILS

33

forest soils (Johnson and Todd, 1983; Singh, 1984; Harrison et af., 1989). The majority of research on the SO$- adsorption capacity of forest soils has focused on the subsoil, because most of the A1 and Fe oxides and hydrous oxides in these soils are concentrated in the lower soil horizons (Johnson and Henderson, 1979; Johnson and Todd, 1983; Singh, 1984; Fernandez and Kosian, 1986; Johnson ef al., 1986b; Khanna et al., 1986; Nodvin et al., 1986; Neary et al., 1987; Dethier et af., 1988; Courchesne and Hendershot, 1989; Harrison et af., 1989; MacDonald and Hart, 1990). Attempts to classify soils in terms of their susceptibility to acidification by acidic inputs have included estimates of their SO:- adsorption capacity (Harrison et al., 1989). Factors that limit SO:- adsorption include competition from other anions, especially dissolved organic matter (Bolt et al., 1976; Sposito, 1989). Increasing the concentration of dissolved organic matter in the soil solution will enhance SO:- mobility (Tabatabai, 1987; Evans and Anderson, 1990; Evans and Zelazny, 1990). Removal of SO:- from the soil solution can occur by precipitation when the ion activity of a given alumino hydroxy sulfate is exceeded (Sposito, 1989), or by the formation of an impure solid in association with A1 oxides or hydrous oxides (Khanna et af., 1987). Precipitation from solution will consume A13', and will decrease exchangeable A13+ in mineral soil horizons (Evans and Zelazny, 1987). Formation of an impure solid in association with an A1 solid phase is thought to occur at lower soil pH levels as AI(OH),

+ SO:- + 2H+ + AIOHSO, + 2H20

(16) when the soil solution becomes saturated with respect to an alumino hydroxy sulfate solid phase (Khanna et al., 1987). Equation (16) as written is a H+-consuming reaction, and the formation of an alumino hydroxy sulfate in association with other A1 solid phases has been postulated as a primary buffering mechanism that removes both H+ and SO:- until the amount of Al(OH)3 that is readily available to react with the soil solution is exhausted (Khanna et al., 1987; Matzner, 1989). Once the supply of readily available Al(OH), is exhausted, H+ consumption by the soil continues by reaction with the alumino hydroxy sulfate phase, 2AIOHS04 + 2H+ + 2A13++ 2SO:-

+ 2H20

(17) with the release of the previously sorbed SO:- (Matzner, 1989). Equations (16) and (17) form the basis for the explanation put forward by Khanna and his co-workers (1987) for the leaching losses of SO$- that increased significantly after 1976 following an acidification phase in the Solling Forest of West Germany. Other studies have also concluded that under certain conditions an alumino hydroxy sulfate phase controls the activity of A13+ and SO:- in soil solutions (Adams and Rawajfih, 1977; Nordstrom, 1982; Weaver etaf.,

34

WAYNE P. ROBARGE AND DALE W. JOHNSON

1985; Gundersen and Beier, 1988; Courchesne and Hendershot, 1990), even though the physical evidence for the presence of this solid phase has yet to be presented (Reuss and Walthall, 1989). Conclusions regarding the presence of an alumino hydroxy sulfate solid phase are generally based on activity diagrams with pH plotted on both axes. As noted in the previous section, such plots are autocorrelative and interpretation of these plots in regard to the mineral phase controlling the activities of A13+ and SO$in soil solution is open to question (Neal et al., 1987; Neal, 1988; Reuss and Walthall, 1989). As of yet, it is still unclear as to the importance of precipitation reactions in removing or releasing SO,’- to the soil solution in forest ecosystems that receive acidic deposition (Reuss and Walthall, 1989). Acidification pushes following disturbances of N cycling in forest soils can enhance SO$- retention if the release of H + to the soil solution increases the net positive charge on the A1 and Fe oxides and hydrous oxides (Bolt et al., 1976; Sposito, 1989). At the Hubbard Brook Experimental Forest, the concentration of SO$- decreased in the soil solution and in the stream-lake environment draining the watershed following whole-tree harvesting, even though the concentration of H + and NO, increased (Nodvin et al., 1988). Retention via anion exchange due to an increase in the net positive charge on the A1 and Fe oxides could account for the removal of SO,’- from the soil solution with decreasing pH (Fuller et al., 1987; Nodvin et a l . , 1988). Later, when the pH of the soil increased due to internal deacidification processes, thereby reducing the net positive charge on the A1 and Fe oxides, the concentration of SO$- in the soil solution increased due to the release of the sorbed SO$-. Sulfate removal from soil solutions can also occur by incorporation into the soil organic matter (Strickland and Fitzgerald, 1984; Schindler et af., 1986; Swank, 1986; Schindler and Mitchell, 1987; Strickland et af., 1987; Watwood and Fitzgerald, 1988; Autry et af.,1990). Estimates using labeled 35S0,’- indicate that as much as 80% of applied SO$- may be rapidly incorporated into the organic fraction (David and Mitchell, 1987; Watwood and Fitzgerald, 1988; Dhamala et al., 1990; Stanko and Fitzgerald, 1990). Such results challenge the assumption that SO,’- adsorption by A1 and Fe oxides and hydrous oxides is the dominant SO,‘- retention mechanism in forest soils (Swank et al., 1984; Swank, 1986). Organic-bound S is the principal form of S throughout the soil profile (Maynard et al., 1987; Mitchell et al., 1989; Autry et al., 1990), with adsorbed SO$- representing a relatively small fraction of the total S present (Mitchell et af., 1989; Autry et al., 1990). The rapid incorporation of applied SO$- into soil organic matter suggests that large quantities of S may cycle through the organic fractions during the growing season. Stanko and Fitzgerald (1990) noted that

ACID DEPOSITION ON FORESTED SOILS

35

33-44% of applied SO',- immobilized by incorporation into soil organic matter was mineralized within 24 hr. Such rapid transformation of SO$would act to buffer the concentration of SO:- in the soil solution, and possibly delay the transfer of associated acidity to the stream-lake environment (Fuller et al., 1986). Such a process offers an alternative explanation of why short-term temporal variations in soil solution and accompanying stream-lake chemistries are more highly correlated with changes in NO, than SO:- (Fuller ef al., 1986; Johnson ef al., 1991b). The emphasis on SO:- retention by subsoils may have overall importance to the quality of groundwater leaving a watershed, but much less so for quickflow from the surface soil horizons or for the soil solution in the rooting zone. Unlike N, current deposition rates of S to forest ecosystems in Central Europe, southern Scandinavia, and the eastern United States exceed the amount required for plant growth. This excess S will either be retained in the soil or will be removed in the drainage water. Many watersheds in the northeastern United States are already approaching S saturation (percent S retention equals zero) (Rochelle and Church, 1987; Rochelle et al., 1987). For these watersheds, it has been proposed that critical loads for future S deposition should be set at the highest deposition rate that will not lead to harmful effects to the stream-lake environment (Henriksen and Brakke, 1988b). It is proposed that the critical load of S not lower pH levels below 5.3 in the most sensitive lakes within the watersheds. Above this pH, strong acids are not present and aluminum concentrations are below toxic levels to fish. However, translating this criterion into actual S deposition estimates is difficult, as the critical load will be dependent on the ratio of wet to dry deposition and the amounts of precipitation and runoff in a given watershed (Henriksen and Brakke, 1988b). Anderson (1989) has predicted that a reduction of 80-90% in current S deposition to forests in Scandinavia would prevent any future changes in soil base saturation due to acidic inputs. The resulting critical load with such a reduction in S deposition would be approximately 2-4 kg (S) ha-lyr-'. De Vries has proposed a critical load of below 10 kg (S) ha-' yr-' for Dutch forest ecosystems. Sulfur deposition from bulk precipitation alone is approximately 8 kg ha-' yr-' (Gorham et al., 1984) in the eastern United States, suggesting that a substantial reduction in S emissions is required if the continued S loading of forest ecosystems in the remainder of the eastern United States is to be stopped or significantly reduced. 4. Mineral Weathering

The chemical weathering or primary silicate minerals derived from the parent material of a forest soil, together with the contribution from aeolian

36

WAYNE P. ROBARGE AND DALE W. JOHNSON

dust, represents the dominant natural process releasing nutrients (except N) for plant growth in a forest ecosystem (N. M. Johnson, 1985; Wright, 1988; Lundstrom and Ohman, 1990; White et al., 1990). Furthermore, the weathering of these minerals neutralizes at least part of the Hf load entering the soil from external and internal sources and represents a substantial portion of the ANC of a forest soil (van Breemen et al., 1983; Matzner, 1989). The rate of silicate weathering can have a significant influence on the chemical state of a forest ecosystem (Matzner, 1989), and will also affect the long-term resistance of an ecosystem to acidic inputs. An increase in the rate of weathering in response to an increase in the H+ concentration of the soil solution would act to buffer against changes in soil acidity, but the potential for loss of base cations would also increase. Such losses from the primary source of nutrient cations within an ecosystem would have a direct influence on the system's ability to recover after a cessation or reduction in acidic inputs. The congruent dissolution of a primary silicate mineral can be represented as MAISi04 + 4H+ + M+ + A13+ + H4Si04 (18) where M is Ca, Mg, K, Na, or Fe or some combination thereof, and H+ represents a source of protons (N. M. Johnson, 1985). The reaction as written consumes H+, but is not a neutralization reaction because of the release of A13+. Most weathering of silicate minerals in soils, however, proceeds by incongruent dissolution (Afifi et al., 1985; N. M. Johnson, 1985; Giovanoli et al., 1988; Nesbitt and Muir, 1988; Wright, 1988; Reuss and Walthall, 1989; White et al., 1990) and can be represented by a combination of dissolution-precipitation reactions Ca(AI2Si2O8)+ 2H+ + 6HOH + Ca2++ 2AI(OH), + 2H,SiO, (anorthite)

(19)

(gibbsite)

Anorthite is only used as an example in Eq. (19), and similar reactions can be written for other primary silicate minerals and for different secondary mineral end products such as kaolinite. The overall net result is the neutralization of H + , the release of nutrients, and the formation of a new secondary mineral solid phase (Reuss and Walthall, 1989). Equation (19) demonstrates that an increase in the H+ concentration of the soil solution should result in an increase in the rate of dissolution of the silicate minerals in a forest soil. Rates of silicate weathering derived from input-output budgets for forested catchments range from 0.2 to 4.6 kmol ha-' yr-' when expressed as the sum of released base cations (Ca + Mg + K + Na) (Folster, 1985; Cronan, 1985; B e r d h et al., 1987; White et al., 1990). Rates vary between

ACID DEPOSITION ON FORESTED SOILS

37

watersheds because of differences in mineral composition of the bedrock (Colman and Dethier, 1986), precipitation amounts (White et al., 1990), and age of the catchment (N. M. Johnson, 1985; Wright, 1988). Excluding areas of known high acidic deposition, such as the Solling Forest in West Germany (Matzner, 1989), the amount of H+ loading to forest ecosystems in Central Europe, southern Scandinavia, and the eastern United States ranges from 0.2 to 2.0 kmol ha-' yr-', which suggests that at least in some forested catchments rates of silicate weathering are sufficient to neutralize present levels of H+ loading (Velbel, 1985). Estimates of the rates of silicate weathering within forested catchments, however, are difficult to obtain primarily because the actual flux of nutrients released through weathering fall within the error limits for the analytical methods used to obtain them (White et al., 1990). Many studies also suffer from the lack of necessary measurements in order to partition the fluxes from silicate weathering from those of atmospheric input, changes in biomass during the monitoring period, and changes in the soil pool of base cations stored on the cation exchange complex (Wright, 1988). Therefore, relatively few studies are available that can be used to determine whether an increase in H+ loading increases the rate of silicate weathering. Input-output budgets at the Solling Forest included measurements for the necessary nutrient fluxes and found that base cation release from weathering has apparently remained constant (0.4-1 .O kmol ha-' yr-') for the past approximately 20 years, even though the current H+ load ranges between 2.0 and 5.6 kmol ha-' yr-' (Folster, 1985). More recent estimates for a spruce site and a beech site in the Solling Forest place the rates of silicate weathering at 0.44 and 0.29 kmol ha-' yr-', respectively (Matzner, 1989). These rates are similar to those estimated for the Hubbard Brook Watershed for the past 24 years, suggesting that there has been no trend in weathering rates at this forested catchment that also receives substantial inputs of acidic deposition (Frogner, 1990). Lack of an increase in weathering rates is also the tentative conclusion of the RAIN project in Norway after 4 years of manipulation of four catchments (Frogner, 1990). The thin and patchy soils in these catchments make them sensitive to increased acidic deposition, as is evident in the increase in A13+, SO$-, Ca2+, and Mg2+ in the treated catchments, but only small increases in the chemical weathering rate have been observed with addition of H+ (Frogner, 1990). One of the objectives of the RAIN project is to determine the effect of increased Hf loading on the rate of silicate weathering, and the failure to observe a change would suggest that the lack of H + is not the rate-limiting step in the weathering mechanism for primary silicate minerals (Wright,

38

WAYNE P. ROBARGE AND DALE W. JOHNSON

1988). Such a conclusion, however, would be contrary to the results of controlled laboratory experiments that have demonstrated an increase in dissolution rates for both primary (Afifi et nl., 1985; Wollast and Chou, 1985; Holdren and Speyer, 1986; White et al., 1990) and secondary (e.g., see Carroll and Walther, 1990) silicate minerals with a decrease in solution pH. Direct chemical weathering as a result of the interaction of exposed bedrock with acid precipitation has also been observed (Giovanoli et al., 1988). A more plausible explanation for the lack of an increase in silicate weathering reported by Frogner is that silicate weathering in the RAIN catchments is a distinctly slow process and that cation exchange reactions in the bulk soil are currently controlling the runoff chemistry (Wright et al., 1988a). A similar explanation has been put forward for the Solling Forest, except that as the store of exchangeable base cations has been depleted, neutralization reactions with other more reactive A1 solid phases have acted to control soil solution pH and limit any effect on increasing the rate of silicate weathering (Matzner, 1989). Weathering of silicate minerals is a fundamental process in the formation of soils and continues to be extensively studied (Berner, 1978; Aagaard and Helgeson, 1982; Drever, 1985; Colman and Dethier, 1986; White et al., 1990). Recent numerical calculations for large-scale forested and agricultural catchments in Central Europe suggest that anthropogenic processes have increased the rate of weathering by a factor of three to five (PaEes, 1983, 1986). However, specific information on the effects of acidic deposition on mineral weathering rates in forested catchments is lacking, and more information is required before any definite conclusions can be reached in this regard (White et al., 1990). 5 . H'Budgets

Hydrogen ions participate in almost every biogeochemical reaction in a forested ecosystem (Driscoll and Likens, 1982), and the transfer of H + between the vegetation, soil solution, and primary and secondary minerals will determine the extent of acidification of the forest soil (van Breemen et al., 1983). A quantitative estimate of the net transfer of H+ ions that occurs within a given ecosystem can be obtained by measuring the inputs and outputs of all substances involved in the H+ transfer processes (van Breemen et al., 1983). Such a compilation is termed a H + budget and consists of essentially four principal components: inputs via atmospheric sources (acidic deposition, carbonic acid); internal production of H + within the ecosystem (mineralization, cation uptake, carbonic acid and organic acid formation, nitrification, oxidation of reduced compounds); internal neutralization of H+ (anion uptake, mineral weathering, reduction of oxidized compounds); and exports of H+ (either as a free ion in the

ACID DEPOSITION ON FORESTED SOILS

39

drainage waters or bound to an inorganic anion or dissolved organic matter) (Driscoll and Likens, 1982). An irreversible flux of H+ occurs when there is a spatial and/or temporal decoupling of the Ht-producing and -consuming processes (Ulrich, 1983; van Breemen et al., 1983). Inputs of strong acids from acidic deposition and loss of bicarbonate ions with associated base cations are two examples of irreversible fluxes of H + ions within an ecosystem (Binkley and Richter, 1987). An irreversible flux of H+ results in soil acidification because of the decrease in the ANC of the soil (van Breemen et al., 1984). The magnitude of the decrease in ANC is an estimate of the extent of weathering in a soil from natural acidification processes, as well as from external sources of H + . Comparing calculated changes in the ANC of a soil to the externalinternal proton ratio (EIPR), (the ratio of atmospheric inputs of H+ to internal H + loading) provides a measure of the capability of an ecosystem to withstand external sources of H + (van Breemen et al., 1984). H + budgets offer the means to calculate changes in the ANC of a soil and the surrounding watershed. In a study comparing 20 forested watersheds, van Breemen et al. (1984) noted that the highest rates of soil acidification, indicated by large negative ANC values, were not associated with the highest values of EIPR, or necessarily with the highest atmospheric loadings of H + . In fact, the ecosystems with the highest rates of soil acidification had EIPR values either equal to or much less than one, demonstrating the importance of natural processes in soil acidification (Johnson et al., 1982, 1983; Nilsson et al., 1982; Krug and Frink, 1983a). Values of EIPR significantly greater than one occurred in ecosystems with low to intermediate rates of soil acidification, i.e., low base cation content and relatively slow mineral dissolution kinetics. These same ecosystems were characterized as having acid drainage waters (pH 4 . 0 ) containing substantial concentrations of soluble Al. Van Breemen and his co-workers (1984) concluded that the amount of external loading of H+ alone was not sufficient to properly assess an ecosystem’s sensitivity to acid inputs. The EIPR values together with an estimate of the rate of soil acidification (or weathering) form a more accurate measure of how an ecosystem will respond. The forested catchments in Central Europe and the northeastern United States are particularly sensitive to acidic inputs because of their relatively low acidification rates (internal neutralization of H f occurs primarily in the A1 buffering range) and EIPR values >1. At the Hubbard Brook Experimental Forest, atmospheric inputs of H + contribute approximately 52% of the total H + sources in the ecosystem (Driscoll and Likens, 1982). External loading of H+ in northwestern Germany can exceed 70% of the total H + load (Bredemeier, 1987; Bredemeier et al., 1990). Such relatively large external

40

WAYNE P. ROBARGE AND DALE W. JOHNSON

inputs of H+ compared to the internal H+ sources, coupled with the fact that silicate mineral weathering rates have apparently remained essentially constant, have forced these ecosystems from an acid neutralizer to an acid exchanger phase. This has resulted in the loss of the capability of the soil to neutralize internal pulses of acidity, such as from nitrification, and a transfer of acidity to the deeper soil horizons (Bredemeier er al., 1990). Van Breemen and his associates (1984) projected that ecosystems with EIPR values >0.5, relatively low acidification rates, and in which precipitation inputs exceed evapotranspiration, will eventually cross into the A1 buffering range with the resulting export of H+ and A13+in the drainage waters. Comparison of EIPR values with rates of acidification within a forested catchment offers a means of estimating a critical load for external H + sources in order to limit future soil acidification due to these inputs. Such estimates, however, will require accurate H+ budgets for various ecosystems. Deriving accurate H+ budgets for an ecosystem is difficult due to the measurement uncertainties in many of the H+ transfer processes (Binkley and Richter, 1987). Of particular concern at this time are the estimates for dry deposition and for silicate mineral weathering rates. As noted by White et al., (1990), errors in measurement of nutrient fluxes from weathering often fall within the error limits for the analytical methods used to obtain them. H+ budgets are also limited both spatially and temporally (Driscoll and Likens, 1982; Binkley and Richter, 1987). Whereas atmospheric deposition generally occurs uniformly over a watershed, H+ budgets within the watershed itself may differ because of differences in vegetation, soil, and parent material. Thus, as the drainage area for a catchment is increased, the value of EIPR and rates of acidification may change because of changes in the magnitude of the H+-consuming and -generating processes. The temporal limitations of H+ budgets result from the fact that they only represent the net transformations of H+ that occur within a prescribed area during the period of measurement. Such measurements will differ for an aggrading versus a degrading ecosystem even though they have similar vegetation, soil, and parent material (Driscoll and Likens, 1982). Seasonal variations in the H+ budgets are also known to occur and their possible influence on calculated EIPR and rates of acidification should not be disregarded (Driscoll and Likens, 1982). 6. Soil Organisms

A comprehensive review of the potential effects of acidic deposition on soil organisms is given by Myrold (1990). Here we present a brief summary

ACID DEPOSITION ON FORESTED SOILS

41

of his review for the sake of completeness with the main theme of our review. Related extensive reviews in this subject area include those by Alexander (1980), Coleman (1983), Coleman el al. (1983), Binkley and Richter (1987), and Bengtsson and Tranvik (1989). Most studies dealing with the potential effects of acidic deposition on soil organisms are controlled exposures using a range of H+ inputs over relatively short periods of time. Significant effects are generally only found to occur when the H+ load is >10 kmol ha-' yr-', but effects at the lowest loading rates (<1 kmol ha-' yr-') have also been noted. These observations hold for studies dealing with effects on soil organisms, as well as on C and N cycling. The situation is further complicated by the fact that there currently seems to be an equal number of published studies for similar experimental systems reporting either positive or negative results. It is apparent that replication of a treatment effect in such systems is difficult, especially at H+ loadings typical of current inputs to most forest ecosystems (<2 kmol ha-' yr-') (Berdhn et al., 1987). There is also the underlying question of how representative most exposure studies are in regard to the actual mechanisms of acidic inputs into forest ecosystems. Many of the cited studies in Myrold (1990) used direct applications of H2S04 or a mixture of H2S04 and H N 0 3 to an isolated subsample of a forest soil horizon. In certain instances the total H+ loading for an experiment is added in one application. Extrapolation of results from such studies to the forest ecosystem is difficult at best, given the interactions acidic inputs undergo as they move through a forest soil. As summarized by Myrold (1990), acidic deposition at high loading rates definitely has the potential to affect soil organisms and biological processes, but these effects may be positive or negative, depending on the organism or processes involved. The majority of exposure studies completed thus far suggest that biologically mediated mineralization of soil organic matter is relatively unaffected by H+ loadings <1 kmol ha-' yr-'. An increase in the amount of organic matter and N in the humus layer of the forest floor in the Solling Forest has been attributed to a retardation in the decomposition of litter (Ulrich, 1985), but the deposition rate in the Solling Forest is generally >2 kmol ha-' yr-' (Matzner, 1989). The field observation at the Solling Forest supports the general conclusion that high inputs of acidity will decrease populations of soil organisms and decrease biological activity (Myrold, 1990), but the effects at more typical loading rates are apt to be much more subtle, involving gradual changes in types and/or numbers of soil organisms, or in pathways of decomposition (Coleman et al., 1983). These effects will be further compounded by the presence of other pollutants, such as ozone (0,) and/or changes in canopy structure or plant species composition following forest decline (Perkins et al., 1987; Myrold, 1990).

42

WAYNE P. ROBARGE AND DALE W. JOHNSON

A comprehensive review on potential linkages between atmospheric processes and the rhizosphere is given by Smith (1990). This review focuses on rhizosphere biology, chemistry, and possible hypotheses for linkages between atmospheric deposition and rhizosphere dynamics. Six hypotheses are proposed as the basis for future research on potential atmosphererhizosphere interactions. These hypotheses include changes in aboveground physiology (e.g., due to O3 exposure), which alters C allocation to roots and root physiology; changes in soil chemistry by atmospheric deposition, which alters rhizosphere regulation of nutrient uptake, heavy metal uptake, aluminum uptake, or microbes essential for nutrient absorption; and changes in the forest ecosystem due to atmospheric deposition that alters rhizosphere regulation of root pathogens and soil saprophyte ecology. Published studies dealing with these potential impacts on the rhizosphere of forest trees are lacking, and, in general, the influence of atmospheric deposition on the belowground ecosystem is poorly characterized (Smith, 1990).

IV. CASE STUDIES OF SOIL CHANGE A. CATEGORIES OF SOILCHANGE AND METHODS FOR MEASUREMENT As noted previously, there are two basic types of changes in the soil that bear upon the acidification issue: (1) a change in soil solution composition (intensity-type change) and (2) a change in the solid phase of the soil itself (capacity-type change) (van Breemen et al., 1983). Changes of the intensity type are quite easily induced (and reversed) with the introduction (or removal) of mineral acid anions to soil solution, and need not be accompanied by any change of the capacity type. There has never been any doubt that intensity-type changes would occur in response to either acid deposition or any other process that adds mobile, mineral acid anions to soil solution. Nor have the preferential increases in A1 in response to mineral acid anions been any surprise to soil chemists familiar with the behavior of exchange processes. However, the significance of this preferential A1 mobilization, and the fact that soil acidification was not necessary for it to occur, were not truly appreciated by the scientificcommunity studying acid deposition until Reuss (1983) pointed out these important implications of the long-known cation exchange process. There are basically two types of studies of capacity-type soil change. The first is space-for-time substitution, in which the effects of forest growth or air pollution on soil properties are deduced by comparing adjacent sites with different vegetative covers (e.g., see Alban, 1982) or pollutant input

ACID DEPOSITION ON FORESTED SOILS

43

histories (e.g., see Wolt and Lietzke, 1982). It embraces the inherent assumption that all sites were similar at one stage, introducing the potential for misinterpreting initial site differences as a result of either vegetation occupancy or pollutant inputs. The second approach involves measuring changes within individual stands or sites, determined by soil remeasurements 10-60 years apart. When this sampling is done in conjunction with measurements of nutrient budgets, it offers by far the best tool for measuring soil change and the causes of it. However, if soil sampling is conducted in isolation from either vegetation growth or nutrient budget measurements, sorting out the causes of soil change becomes nearly impossible. Possibilities of capacity-type soil change were never in doubt, given enough acid input over a significant amount of time. Well-known cases of soil acidification near smelters clearly documented the validity of that logical conclusion (e.g., Wolt and Lietzke, 1982). However, the amount of time needed for such changes under typical field conditions was assumed to be on the order of many decades to centuries, given the pool sizes of base cations and acidity on exchange sites and the available information on fluxes into and out of these pools (McFee, 1980; Johnson et al., 1982, 1985). However, several recent studies and literature reviews have conclusively demonstrated that forest soils in many parts of the world, both polluted and unpolluted, are changing much more rapidly than previously suspected (e.g., Hallbacken and Tamm, 1986; Berden et al., 1987; Binkley et af., 1989b; Johnson et al., 1988a, 1991a). In some cases, these changes are attributable primarily to nutrient uptake by vegetation (including uptake of Al; Turner and Kelly, 1977; Alban, 1982; Binkley et al., 1989a; Johnson et al., 1988a), and in other cases, the changes are attributed to both nutrient leaching and uptake (Van Miegroet and Cole, 1984; Hallbacken and Tamm, 1986; Johnson and Todd, 1990). In cases where soil change is documented and leaching is known to be strongly impacted by acid deposition, it is clear that acid deposition has played a role in causing such changes (e.g., Tamm and Hallbacken, 1986; Billett et al., 1990b; Johnson and Todd, 1990). In contrast to a decade ago, the case studies of soil change are meanwhile too numerous to be reviewed in this article. The reader is referred to Berd&n et al. (1987) and Johnson et al. (1991a) for a detailed review of recent case studies of soil change. Here, we will only refer to a seiected number of the these studies for the purposes of illustration. Specifically, we will review case studies of intensity-type changes in which SO,’- and NO, have caused marked increases in soil solution A1 concentration, and capacity-type changes caused by uptake, leaching, and the combination of the two.

44

WAYNE P. ROBARGE AND DALE W. JOHNSON

Figure 5. (a) Changes in soil exchangeable A13"at the Solling site, Germany, from 1966 to 1979. (b) Changes in soil exchangeable base cations (Ca2+, MgZ+,K + , and Na+) at the Solling site, Germany, from 1966 to 1979. After Ulrich er al. (1980).

B. INTENSITY-TYPE CHANGES One of the earliest studies of soil change in an acid deposition-impacted forest ecosystem (a beech forest at the Solling site in Germany) was the work of Ulrich and his co-workers (1980). The Solling site is one of the most intensively studied sites in the world with regard to element and Hf budgets, and these element budgets are strongly impacted by acid deposi-

ACID DEPOSITION ON FORESTED SOILS

45

tion (Ulrich et al., 1980). However, inasmuch that soils at the site were well within the aluminum buffering range before soil sampling took place, it is not surprising that changes in exchangeable base cations over the 13-year sampling period were minimal, at least in comparison to those noted by Cole and Lamon (1981; as cited in Johnson et al. (1981), Johnson et al. (1988a), and Binkley et al. (1989b)) (Fig. 5 ) . The authors did note increases in A13+ in “equilibrium soil solution,” which is a water extract of fieldmoist soil, and, to a much lesser degree, increases in exchangeable A13+. The authors interpreted the increase in exchangeable A13+ as a result of dissolution of interlayer A1 polynuclear hydroxo complexes by H + . It is obvious from the discussion of cation exchange above that increases or decreases in A1 in the equilibrium soil solution could occur over very short intervals, without any change in exchangeable A13+ or base cations, as mineral acid anions increase and decrease. In this connection, Khanna et al. (1987) noted marked decreases in soil solution pH and marked increases in soil solution A1 associated with increases in soil solution SO:at the Solling site over the period 1974-1978 (Fig. 6 ) . They interpreted these results as indicative of the dissolution of jurbanite (A1OHSO4)rather than an exchange process. The preferential displacement of A13+ from the exchange complex in acid soils in response to much shorter-term increases in mineral acid anion concentrations was illustrated by Reuss (1989) for forest soils subject to NO; pulses in a nitrogen-fixing red alder stand. It is also illustrated in Fig. 7, which shows the response of soil solution A1 to pulses of both NO; and SO$- in a red spruce stand (Johnson et al., 1991b). Aluminum concentrations in soil solution at this site ranged from 30 to over 200 pmol/ liter (600 pEq/liter in Fig. 7). These peak A1 values are well within the range noted to inhibit Ca and Mg uptake in seedling solution culture studies in the laboratory (100 pmol/liter) (Raynal et al., 1990) and are at the threshold for effects on root growth in solution culture studies (200 pmol/liter) (Thornton et al., 1987; J o s h and Wolfe, 1988). Throughout the collection period in this study, total A1 concentrations (which were 80-90% monomeric Al) were positively correlated with solution NO, and SO:- . Fluctuations in solution SO:- and NO; concentrations individually could explain over 60% of the variation in A1 concentrations, and NO; + SO:- combined could account for over 70% of the variation in measured A1 concentrations from this site (Johnson et al., 1991b). The presence of acid soils is a necessary but not sufficient condition for the mobilization of A1 via exchange processes. The introduction of mobile, mineral acid anions to an acid soil will cause increases in the concentration of A1 in soil solution, but extremely acid soils in the absence of mineral acid anions will not produce solutions that are high in Al. This is illustrated by

46

-8

WAYNE P. ROBARGE AND DALE W. JOHNSON

1

1

BEECH 0.8

.

SPRUCE

'.

'

.

0.3 '

YEAR

YWI

Figure 6. Changes in soil solution SO:- and A1 at the Solling site, Germany. After Khanna ef al. (1987).

the results of the Integrated Forest Study for the relatively unpolluted Findley Lake site as contrasted to the more poiluted Whiteface Mountain and Great Smoky Mountains National Park (GSMNP) sites (Johnson et al., 1989; Johnson and Taylor, 1989). Figure 8 shows the soil solution anion and cation concentrations in B horizons from the pristine site at Findley Lake, Washington, and the more polluted sites at Whiteface Mountain, New York, and Clingman's Dome, GSMNP, North Carolina. Though the Findley Lake site soil is as acid as those from the more polluted eastern sites, concentrations of A1 in soil solutions at Findley Lake are much lower than those at the latter sites. The differences in soil solution A1 concentrations are almost totally a function of differences in mineral acid anion (SO$- and NOT) concentration, both of which increase from Findley to Whiteface to Clingman's Dome due to a combination of

ACID DEPOSITION ON FORESTED SOILS

47

700 600 500

f .-c

400

s

300

-. ~

200 100 0 '

1000

800

-.-f

. c

600

P

W

3.

400

200

0 .

-

.

.

.

,

~ .. .

- . . .,~. .

400

Sulfate

-.-z .

300

-

200

-

100

-

c

w

3

0'

A

J O J A J O J A J O J A J O J 1985 1986 1987 1988 Figure 7. Temporal variations in A horizon soil solution Al, NO;, and SO:+ in a red spruce forest in the Great Smoky Mountains National Park, North Carolina. After Johnson etal. (1991b).

48

WAYNE P. ROBARGE AND DALE W. JOHNSON

Figure 8. Soil solution composition in the Findley Lake, Washington; Whiteface Mountain, New York; and Clingman's Dome, Great Smoky Mountains National Park (GSMNP), North Carolina. After Johnson and Taylor (1989).

increases in atmospheric deposition and internal nitrification rate (Johnson et al., 1991b).

C. CAPACITY CHANGES DUETO LEACHING In one of the most convincing case studies of soil change using the space-for-time approach, Van Miegroet and Cole (1984, 1985) documented the effects of 55 years of excessive N fixation and NO; leaching on soil properties in adjacent stands of red alder and Douglas fir [Pseudotsuga menziesii (Mirb Franco)] at the Thompson Research Center, in the foothills of the Cascade mountains of Washington. The Douglas fir stand was planted in 1931 after a series of wildfires following logging of the original old-growth forest between 1910 and 1920. The adjacent red alder forest established naturally a few years later where conifer planting ceased in the burnt area. Nitrogen fixation in the red alder stand resulted in the accumulation of approximately 5000 kg ha-' of soil N versus about 2000 kg ha-' in the adjacent Douglas fir stand. Nitrate concentrations in the soil solution measured in 1981-1983 were on the order of 200300 pEq/liter in the red alder stand versus 0.5-2 pEq/liter in the Douglas fir stand (Van Miegroet and Cole, 1988), indicating high rates of nitrification and considerable internal generation of nitric acid within the red alder soil. It was estimated that nitrification resulted in an annual H+ input of approximately 3.2 kmol ha-' yr-', a value not dissimilar from that attributable to wet and dry deposition reported for the Solling site (Matzner,

ACID DEPOSITION ON FORESTED SOILS

49

1989). This high rate of nitrate leaching decreased the surface soil pH beneath the alder compared to the Douglas fir by over one-half pH unit (pH 4.3 versus 5.0) and the base saturation by 50% (Fig. 9). The base saturation in deeper soil horizons appeared to be higher, perhaps due to displacement of base cations from the upper soil or perhaps due to increased weathering in these horizons. Such changes in soil acidity due to occupancy by Ainus species have also been observed by Franklin et al. (1968) and Binkley and Sollins (1990). In another space-for-time study, Tamm and Hallbacken (1988), through a process of elimination, attributed pH decreases in C horizons of soils in southern Sweden over the period from 1926 to 1985 to acidic deposition. In reaching this conclusion, they compared pH changes in all horizons over this 60-year period at two Norway spruce (Picea abies) sites: one in northern Sweden, where acidic deposition inputs were minimal, and one in southern Sweden, where acidic deposition inputs were elevated. At both sites, there was a clear trend toward decreasing pH in surface soils with stand age, presumably due to accumulations of acidic humus. At the northern site, no pH decreases were observed in the C horizon, whereas substantial C horizon pH decreases were noted in the southern site (Fig. 10). The possibilities that differences in forest history, climate, and soil mineralogy could account for these differences in pH were discounted, leaving leaching as the most likely cause. The authors make the tacit assumption that natural leaching rates are similar at the two sites, and therefore conclude that acidic deposition is the major cause of the C horizon pH decreases at the southern site. There are no data to support the

Figure 9. Soil exchangeable cations beneath red alder (RA) and Douglas fir (DF) stands at the Thompson Research Center, Washington. After Van Miegroet and Cole (1984).

50

WAYNE P. ROBARGE AND DALE W. JOHNSON

Figure 10. Changes in soil pH beneath spruce forests in northern (Kulbacksliden site, left) and southern (right, Tonnersjoheden) Sweden. After Tamm and Hallbacken (1988).

ACID DEPOSITION ON FORESTED SOILS

51

assumption that natural leaching rates are similar at the two sites, and the possible effect of previous occupancy by much deeper-rooted beech (Fagus sylvatica) at the southern site (Tamm and Hallbacken, 1988) raises questions about the role of uptake in causing the C horizon pH decreases. However, there is little doubt that acid deposition could either cause or substantially contribute to increases in soil acidity over such a time period. There have been several other studies of soil change in Sweden that are of note. Falkengren-Grerup (1987) reported fairly consistent increases in acidity and decreases in exchangeable bases in nine sites (both forested and nonforested) in southern Sweden from 1949 to 1985. The increases in acidity were much more substantial in soils that were initially less acid, as predicted by Reuss (1983) and as also noted by Anderson (1988) (Fig. 11). The authors hypothesize that acid deposition was a major factor in causing the increases in acidity, but have no data from element budgets to compare the relative effects of acid deposition, plant uptake, and natural leaching on these systems. Falkengren-Grerup and Eriksson (1990) noted decreases in pH and exchangeable cations between 1947 and 1988 in the C horizons of F. sylvatica and Quercus robur sites in southernmost Sweden. An unexpectedly large increase in the growth of the beech stands noted during

0 00

ce

I

0

0 5,

I

.

8 O 5. 9 6

00

.9

Figure 11. Changes in soil pH as a function of original pH in A horizon samples from a variety of sites in Sweden between 1949 and 1970 and 1984 and 1985. After FalkengrenGrerup et al. (1987).

52

WAYNE P. ROBARGE AND DALE W. JOHNSON

this period was attributed to increases in atmospheric N deposition. Once again, no element budget data were available to calculate the relative importance of acidic deposition, uptake, and natural leaching, any one of which might have caused a substantial amount of soil acidification over the four decades involved in this comparison. Subsequent to the initial studies at Solling (Ulrich et af., 1980), decreases in exchangeable Ca2' and Mg2+ have been found in many sites in Germany (Hildebrand, 1986a; Grimm and Rehfuess, 1986; Hauhs, 1989; Ulrich, 1989; see also review by Johnson et af., 1991a). The study by Hauhs (1989) at Lange Bramke was accompanied by a partial budget indicating that leaching was of major importance in causing substantial reductions in soil exchangeable base cations over a 10-year period (19741984). Biomass uptake and increment values are not presented, but the high rates of S (46 kg ha-' yr-') and N (19.4 kg ha-' yr-') deposition, the high soil solution SO$- concentrations (230-500 pEq/liter), and the trend of decreasing soil solution Ca2+ concentrations from 1977 to 1985 collectively present a very convincing case for acid deposition-caused soil acidification. The importance of leaching in causing soil change can vary considerably among individual cations, as shown by Johnson et at. (1988a) and Johnson and Todd (1990) in their studies of soil change and the reasons for it on Walker Branch Watershed, Tennessee. Johnson et al. (1988a) noted marked reductions in exchangeable Ca2' and Mg2+ in subsurface horizons over the period 1971-1982 (Fig. 12). The decreases in exchangeable Ca2+, which occurred in about half of the eight sites sampled, were expected due to large accumulations of Ca in forest biomass. The reductions in exchangeable Mg2+, which occurred in nearly every site sampled, could not be accounted for by Mg accumulation in forest biomass, however, leaving leaching as the most likely cause. Previous studies on Walker Branch had shown that ieaching was dominated by SO:- from atmospheric deposition (Richter et af., 1983; Johnson et af., 1985), and thus it was hypothesized that the reductions in exchangeable Mg2+ were due primarily to acidic deposition. In a follow-up study on four of the eight sites, Johnson and Todd (1990) constructed element budgets (using tension lysimeters to measure leaching output) to test the hypotheses posed in the initial study. They found support for the hypothesis that the Ca2+ decreases were due primarily to tree Ca accumulation, which greatly exceeded leaching in those sites where exchangeable Ca2+ decreased (Table 111).However, the authors could neither support nor reject the hypothesis that leaching was the primary cause of the Mg2+ decreases across all the sites from the data available. Leaching was clearly the dominant mechanism of Mg export in

53

ACID DEPOSITION ON FORESTED SOILS

a 0.5

w

3 0.4 \

m 5 0.3

-

0

5

OS2 0.1 0.0

26

98

281

PINE POPLAR

42

179

91

107

237

CH. OAK

OAK-HICKORY

42

91

b 1.0

w

1

0.8

x

\

Q

0

0.6

-

0

Eu

OA 0.2

0.0 26

98

281

PINE POPLAR

179

107

237

CH. OAK OAK-HICKORY

Figure 12. (a) Changes in Bt horizon exchangeable Mg2+ at selected sites in the Walker Branch Watershed, Tennessee, from 1971 to 1982. Numbers on the x axis represent individual plots. Asterisks indicate differences detected at the 5% level of significance. (b) Changes in Bt horizon exchangeable CaZ+at selected sites in the Walker Branch Watershed, Tennessee, from 1971 to 1982. Numbers on the x axis represent individual plots. Asterisks indicate differences detected at the 5% level of significance. After Johnson et al. (1988a).

54

WAYNE P. ROBARGE AND DALE W. JOHNSON Table 111 Annual Fluxes of Ca and Mg in Four Sites on Walker Branch

Watershed, Tennessee" kg ha-' yr-'

Flux Deposition Through fall Litterfall Soil solution Wood accumulation

Deposition Throughfall Litterfall Soil solution Wood accumulation

Pine

Yellow poplar

Oakhickory

Chestnut oak

Calcium 6 1954 61 +- 17 15 2 2 16.3

25 -1- 4 5127 721 21

1825 34211 1 2 0.6 45

14-1-1 31 2 16 12 3 27

*

Magnesium 1 4 2 0.6 8-1-2 621 2.8

4*1 6-1- 1 9 2 0.4 1.2

422 522 120.4 2.9

3 2 0.6 422 721 1.0

"Standard deviation for soil solutions represent variation due to concentration only. After Johnson and Todd (1990).

three of the four sites monitored, but leaching was anomalously low' at the fourth site. Leaching was dominated by atmospherically derived SO$- at three of the sites, and thus it could be concluded that the decreases in Mg2+ were due to acidic deposition. Soil solution SO$- and, consequently, total leaching were very low in the fourth site. Tree Mg accumulation was the major export mechanism in this case. Soil SO$- adsorption was known to be a major cause of ecosystem S retention on other sites on the watershed, but was clearly not responsible for the anomalously low soil solution SO:at the fourth site. The cause of the low soil solution SO:- in this case was not discovered, but SO$- reduction was suspected. Despite the anomaly of the fourth site, the hypothesis that atmospherically induced leaching caused reductions in exchangeable Mg2+ was supported in three of four cases, and remains a viable hypothesis for the other sites where exchangeable Mg+ decreases were noted. Soil acidity can be directly affected by atmospheric deposition even if it does not cause increased leaching of base cations. Specifically, soil acidity

'Soil solution sulfate and total leaching were anomalously low in comparison to the other three sites in that particular study as well as in comparison to several other studies on Walker Branch (D. W. Johnson er al., 1981; 1985; Richter et al., 1983).

55

ACID DEPOSITION ON FORESTED SOILS

can be increased by adsorption of SO’,- onto soils (Coleman and Thomas, 1967; Mehlich, 1964). This “acidity due to anions’’ was noted by Mehlich (1964) in liming experiments on soils pretreated with SO:-, OH-, and H+. Soils containing adsorbed SO$- had higher lime requirements and lower pH values than those containing adsorbed OH-. This was a logical result (at least in retrospect), and one that has implications for soils that adsorb atmospherically deposited SO$-. Numerous studies have indicated that highly weathered soils of the southeastern United States are accumulating atmospherically deposited SO$- (D. W. Johnson et al., 1980, 1986a; Rochelle et a / . , 1987), and are therefore likely experiencing increased acidity due to anion adsorption. The actual quantity of acidity accumulated due to SO:- adsorption over an 80- to 90-year period can be estimated from a two-case study by D. W. Johnson et al. (1981) near Oak Ridge, Tennessee. B horizon soil samples were taken from an Ultisol beneath and adjacent to a house that was erected on flat ground (i.e., without excavation) around the turn of the century. Soils from beneath the house had on the order of 0.30.8 cmol kg-’ less adsorbed SO$- than soils adjacent to the house (Table IV). If one assumes that these differences are due primarily to atmospheric S deposition, the contribution of SO:- adsorption to soil acidity has been substantial. This is especially so when the amount of adsorbed SO,’- is compared to cation exchange capacities and exchangeable cation values from other nearby sites (see later, Fig. 15) [D. W. Johnson et al., (1981) did not report values for exchangeable cations on these particular soils.]. Sulfate adsorption and precipitation reactions cause a reduction in acidification due to base cation leaching, but soil “acidity due to anions” is increased. The way in which this anion acidity will ultimately manifest itself

Table IV Soluble and Adsorbed S in B Horizon Samples Taken from Beneath and Adjacent to a House Erected ca. 1890“ Adsorbed‘ SO:(cmol,/kg)

Solubleb SO:(cmol,/kg)

Depth from top of B horizon (cm)

Beneath house

Adjacent to house

Beneath house

Adjacent to house

0-15 15-30

0.13 2 0.03 0.08 f 0.006

0.11 f 0.02 0.09 0.006

0.42 f 0.14 0.67 f 0.16

1.22 2 0.26 0.08 -+ 0.31

*

“After Johnson ef al. (1981). 1 :20 soil :water extraction. ‘Sulfate extracted with 0.016 M NaH2P04, 1:5 soil: water ratio, minus soluble SO:-

56

WAYNE P. ROBARGE AND DALE W. JOHNSON

depends on a number of factors. If input concentrations of SO:- are reduced and adsorbed SO$- desorbs and leaches, the anion acidity will ultimately have the same (but delayed) effect on soil exchangeable acidity as if SO:- had leached freely through the soil. In effect, the desorption of SO:- can be thought of as a net production of anions in the soil, analogous to internal HCO, or NO; formation. On the other hand, if adsorbed SO:- does not desorb when input SO:- concentrations are lowered, anion acidity remains elevated and will be manifested only if the soil is limed.

D. CAPACITY CHANGES DUETO UPTAKE Most case studies in which soil change has been attributed to uptake are of the space-for-time type. A thorough review of these studies is given by B e r d h et al. (1987) and Johnson et al. (1991a). Here, we will present a few examples for the purpose of illustration. One of the longest-documented records of soil change is at the Rothamsted site in the United Kingdom. Jenkinson (1970) reported a pH decline from 7.1 to 4.5 over a period of 82 years of natural reforestation of abandoned agricultural land. Associated with this pH decline was an increase in soil S content, which was attributed to accumulation of atmospherically deposited S by SO:- adsorption. The capacity of the soil to retain SO$- was, in turn, enhanced by increasing soil acidity. There was no implication that the pH decline was due to atmospheric deposition, only that it enhanced the adsorption of atmospherically deposited S. The pH decline itself was simply attributed to natural acidification processes due to reforestation. In a later study at the same site, Johnson et a f . , (1986a) concluded that wet deposition of H+ made a negligible contribution to soil acidification at Rothamsted, whereas dry deposition was more important. They estimated that the total contribution of atmospheric deposition (wet and dry) to the observed soil acidification was approximately 30%. Jenkinson (1970) also documented net annual soil N accumulations of more than 50 kg ha-' yr-' during an 81-year period (from 1883 to 1964) after a former agricultural site (Broadbalk) was allowed to revert to forest. This high rate of soil N accumulation was greater than was believed possible from atmospheric deposition alone and may have largely been due to the action of free-living nitrogen-fixing bacteria in the soil. Liming may have played some role in stimulating these high accumulation rates; a nearby site (Geescroft) that had not been limed showed N accumulations of only 23 kg ha-' yr-' during the same period. Alban (1982) presented very convincing evidence of tree Ca uptake resulting in changes in soil exchangeable Ca2+ by comparing nutrient distributions and soil properties of adjacent stands of quaking aspen (Populus tremuloides Michx.), white spruce (Picea glauca Moench Voss),

ACID DEPOSITION ON FORESTED SOILS

57

red pine (Pinus resinosa Ait.), and jack pine (Pinus banksiana Lamb.) in Minnesota. He found that aspen and white spruce accumulated Ca in their biomass and depleted Ca2+ from the soil exchange complex as compared to the pines. He also found that cycling of Ca at a high rate in the aspen and white spruce stands caused higher base saturation of surface horizons and lower base saturation in subsurface horizons as compared to the pine stands. Cole and Lamon sampled soils in 1962 and 1976 at the Rothamsted site referred to earlier in order to determine if the potential depletion of exchangeable K predicted by nutrient budget data would be manifested in an actual soil change (cited in D. W. Johnson et al., 1981). They found significant reductions in exchangeable K + over the sampling period of only 14 years. They also noted a trend toward lower exchangeable Ca2+ and Mg2+. The effects of acidic deposition were later studied at this site and were found to be negligible relative to the effects of natural leaching (primarily by carbonic acid) and tree uptake (Cole and Johnson, 1977). These results were significant in two respects: (1) they indicated that nutrient budget analyses were providing reasonable forecasts of potential soil change and (2) weathering was not keeping pace with base cation export by uptake and natural leaching processes. Binkley et al. (1989b) examined changes in soil chemistry across 20 years in a loblolly pine stand planted in an abandoned agricultural field in South Carolina. Soil pH declined by 0.3 (B horizon) to 0.8 units (A horizon), and pools of extractable base cations declined by 30% (B horizon) to 80% (A horizon) (Fig. 13). The loss of base cations from the soil was similar to +

Figure 13. Changes in soil exchangeable cations in a loblolly pine soil at Calhoun Forest, North Carolina. After Binkley er al. (1989b).

58

WAYNE P. ROBARGE AND DALE W. JOHNSON

expected rates of accumulation in biomass, but a possible role of natural leaching and acidic deposition could not be excluded. In a similar study, Johnson and Todd (1987) compared nutrient distributions, soil properties, and nutrient fluxes in oak-hickory (Quercus-Curyu) and loblolly pine (Pinus taedu L.) stands near Oak Ridge, Tennessee. The oak-hickory forests had accumulated considerably more Ca in their biomass than the loblolly pines, and as was the case in Alban's study, the soil under the Ca-accumulating forest had considerably lower exchangeable Ca2+ (Fig. 14). A horizon-by-horizon comparison indicated that exchangeable Ca2+ was lower throughout the profile in the mixed-oak soil, but only in the deepest horizon was exchangeable A1 greater (Fig. 15). The mixed-oak forest had greater amounts of exchangeable K + and lesser amounts of exchangeable Mg2+ than the loblolly pine soil, but these differences could not be attributed to vegetation uptake. Anderson (1988) resampled 48 sites in the Adirondack Mountains of New York that had been measured for pH and extractable Ca (with 0.2 M HCl) 50 years earlier. She found that the pH and exchangeable Ca2+ of organic horizons that were initially below 4.0 did not change over 50 years, whereas the pH of less acidic organic horizons declined by 0.3-0.5 units. These results are entirely consistent with theory, which predicts that soils already in the A1 buffering range will not acidify further (Ulrich, 1980; Reuss and Johnson, 1986). Declines in soil Ca over the period (about 19 kmol, ha-') were less than could be attributed to the accumulation of Ca in tree biomass (26 kEq ha-'). No impact of acidic

Figure 14. Distribution of Ca, K, and Mg in vegetation, forest floor, and mineral soil (exchangeable fraction) in loblolly pine (LP) and mixed-oak (MO) sites near Oak Ridge, Tennessee. After Johnson and Todd (1987).

ACID DEPOSITION ON FORESTED SOILS

59

Figure 15. Distribution of exchangeable cations in mineral soil horizons in loblolly pine (LP) and mixed-oak (MO) sites near Oak Ridge, Tennessee. After Johnson and Todd (1987).

deposition was apparent, but acidic deposition and natural leaching must have contributed to the changes to some unknown extent. One aspect of plant uptake and soil change that has not been adequately considered is the cycling of A1 (see review by Johnson et ul., 1991a). Aluminum accumulation in the leaves of coachwood (Cerufopetulurn upefulurn) in Australia has been found to have a major impact on the distribution and cycling of base cation (Turner and Kelly, 1981). The presence of coachwood as a secondary tree layer beneath brush box (Lophosternon conferfus) was found to lead to increased soil exchangeable A1 and decreased soil exchangeable Ca2+ (Turner and Kelly, 1981) (Fig. 16). The constant addition of aluminum-rich litterfall has obviously increased A1 in the surface soil and replaced CaZ+on soil exchange sites.

E. SEASONAL VARIATIONS IN EXCHANGEABLE CATIONS Several studies have documented seasonal variations in soil chemical properties. Haines and Cleveland (1981) noted large seasonal variations in many soil properties, including pH, exchangeable Ca2+, K + , and Mg2+, organic matter, and cation exchange capacity in forest soils under various treatments and vegetation cover (chopped, old field, agri-forest, pinehardwood, and young slash pine) in Georgia. The authors did not speculate on the causes of these variations, which were surprisingly large and previously unsuspected for soil parameters such as organic matter and bulk density. Similarly, Peterson and Rolfe (1982) also noted large seasonal

60

WAYNE P. ROBARGE AND DALE W. JOHNSON

Figure 16. Exchangeable cations beneath three forest species in Australia. After Turner and Kelly (1981).

Figure 17. Seasonal variations in exchangeable CaZ+in a site on Walker Branch Watershed, Tennessee. Error bars represent one standard deviation. After Johnson er al. (1988a).

variations in pH and exchangeable K + and Ca2+ in both upland and periodically flooded forest soils in central Illinois. Surface soils on Walker Branch Watershed showed large seasonal variations in exchangeable Ca2+ and K+,which diminished with soil depth. These variations were attributable to the annual cycle of uptake and return by throughfall (most important for K+), litterfall, and decomposition (most important for Ca2+).

ACID DEPOSITION ON FORESTED SOILS

61

Seasonal variations in exchangeable Mg2+ were not as marked (Fig. 17), reflecting the lower rate of annual cycling of this nutrient as compared to Ca2+ and K + . In another study on a different soil at the Walker Branch Watershed, Johnson and Todd (1984) found large seasonal variations in exchangeable A13+.

F. INTERACTIONS OF UPTAKE AND LEACHING When considering the potential causes of change in organic horizons as noted by Anderson (1988) and Falkengren-Grerup and Tyler (1987), the role of both uptake and litterfall return must be considered. Unlike the mineral soil, which can usually be considered to consist of the same basic material over a period of decades, the organic horizons are constantly being depleted by decomposition and renewed by litter input. Organic horizons are, in essence, “moving targets.” Given the fact that litterfall is normally the most important input mechanism of Ca and Mg into the forest floor (Cole and Rapp, 19Sl), it is advisable to consider the indirect effects of uptake via litterfall return on organic horizon change rather than treating organic horizons conceptually the same as mineral soil, with a relatively constant parent material composition and exchange capacity. Changes in litter quality can be brought about by many factors, including species composition, stand age, and nutritional status. The assertion that organic horizon acidification is likely most strongly affected by litter quality does not mean that acidic deposition can have little or no effect on organic horizon change. If acidic deposition causes a reduction in exchangeable base cation reserves in the soil, or if A1 mobilization by acidic deposition inhibits base cation uptake, litterfall base cation return may well be reduced, thereby causing a reduction of the base cation status of organic horizons. The uptake and cycling of A1 is likely to have major effects on soilleaching processes. Turner and Kelly (1981) did not study leaching, but presumably the displacement of exchangeable base cations by A1 in the coachwood forests caused increased base cation leaching. It would be interesting indeed to discover what effect the presence of aluminum-cycling species such as coachwood has upon soil leaching, both in terms of anion production and mobility and cation exchange processes. The uptake and sequestering of one or more base cations in biomass will indirectly affect leaching if such uptake results in the depletion of one or more base cations from exchange sites. Cation exchange equations dictate that the leaching of the depleted cation(s) will decrease relative to other exchangeable cations, given constant inputs of anions. One would assume

62

WAYNE P. ROBARGE AND DALE W. JOHNSON

this to be the case in Alban’s (1982) studies, and it definitely occurred in the studies of Johnson and Todd (1987). These differences in exchangeable Ca2+ caused by differences in biomass Ca sequestering manifested themselves clearly in leaching rates: solutions from the mixed-forest soil had lower soil solution pH, bicarbonate, Ca2+, and total cation concentrations than the loblolly pine stand (Fig. 18), and Ca2+leaching was much lower in the mixed-oak stand than in the loblolly pine stand. Interestingly, the loblolly pine stand had higher Mg2+ leaching rates, a consequence of higher Mg2+:Ca2+ ratios on the exchange sites rather than greater abso-

Figure 18. (a) Mean anion concentrations in soil solutions collected in loblolly pine and mixed-oak sites near Oak Ridge, Tennessee. (b) Atmospheric deposition and leaching loss of Ca, K , and Mg in loblolly pine (LP) and mixed-oak (MO) sites near Oak Ridge, Tennessee. After Johnson and Todd (1987).

ACID DEPOSITION ON FORESTED SOILS

63

lute quantities of exchangeable Mg2+. In the extreme case in which uptake has depleted all base cations from the exchange sites, A1 will be the dominant cation in soil solution. The effects of uptake and recycling on soils and the interactions of uptake and leaching can be even more significant than the effects of cation sequestering in biomass. The potentially important effects of recycling are amply demonstrated by the marked seasonal changes in exchangeable base cations described above. A consideration restricted to cation increment in woody biomass neglects the possibility of interhorizonal changes, wherein uptake of base cations from deeper mineral soils and recycling via litterfall and foliar leaching to surface soils could cause major changes in the distribution of base cations and acidity without causing major changes in total soil profile content (e.g., Alban, 1982). Thus, uptake can act to conserve those cation nutrients in greatest demand by vegetation. For the same reason, it can be argued that harvesting takes a greater toll of the pools of potentially limiting cations than leaching does (Johnson and Todd, 1987).

V. FUTURERESEARCH Soil acidification of present-day forest ecosystems is caused by three factors, which include stress caused by variability of climate and aggrading forests, biomass removal due to harvesting, and the deposition of acidity from air pollution (Ulrich, 1987). The rate at which many forest soils appear to be changing requires a reevaluation of our techniques for predicting and evaluating change in these ecosystems (Johnson et al., 1991a). Acidic deposition has contributed substantially to the acidification of soils where it occurs in relatively high amounts [e.g, the Solling Forest in Germany (Matzner, 1989)l. However, its role in the acidification of forest ecosystems in much of the northern humid temperate regions is less clear because of the variety of physical and chemical interactions possible in forest soils, and because of a lack of a thorough understanding of natural acidification processes. Two areas routinely ignored in most studies on natural acidification processes are A1 uptake and interhorizon changes in cation content. The gradation in organic matter content of forest soils represents a chemical redistribution of cations that are removed from the deeper mineral soil horizons and recycled via litter and foliar leaching to the surface soil. Active cycling of A1 (N. M. Johnson et al., 1981; David and Driscoll, 1984; Schecher and Driscoll, 1989) means that A1 is continually being added to

64

WAYNE P. ROBARGE AND DALE W. JOHNSON

the forest floor via litterfall. Little is known of the chemical forms of A1 in fresh litter or what transformation it undergoes during the conversion of litter to soil humus. Biocycle A1 represents a source of A1 that can interact with acidic inputs, a source that often has not been considered. The interactions of interhorizon changes with either natural or anthropogenic acidic inputs also needs to be considered in terms of the effective rooting volume of the canopy. H+ budgets usually estimate changes integrated across the entire soil profile (Binkley and Richter, 1987), but the effective rooting volume may be limited to the upper surface horizons. Significant changes in cation budgets within the rooting zone may not be detected in a whole-soil H+ budget. Interhorizon processes that could result in significant changes in cation budgets include enhanced leaching because of rapid cycling of SO;- between organic matter and the soil solution, enhanced acidification of the rooting zone due to adsorption of NHZ in the canopy, and acidification and enhanced leaching because of an increase in N mineralization and nitrification. Lack of knowledge about interhorizon changes also extends to the coupling of the various acidification and neutralization processes within the soil profile. Of particular importance is the extent of preferred flow paths that allow rapid communication between soil horizons and the stream-lake environment. The presence and number of such flow paths will be a function of soil texture and structure. In coarse-textured soils, such as those on glacial till in the northeastern United States?such preferred flow paths may not be of concern. In the more finely textured soils of the southeastern United States, such flow paths significantly influence the water movement through the soil profile. Relatively few studies have been done that incorporate the hydraulic properties of a soil with its projected capacity to retain SOf- anions and neutralize acidic inputs. The presence of preferred flow paths could also influence weathering rates of silicate minerals. The lack of agreement between studies cited for the effects of acidic deposition on soil biology illustrates the need for critical work in this area. However, future research must first solve the question of how to conduct such experiments so that meaningful results can be obtained and extrapolated to the field environment. This is particularly true for studies on the rhizosphere. Numerous greenhouse and laboratory solution growth studies have established A1 toxicity levels for various tree species (Schier, 1985; van Praag and Weissen, 1985; van Praag and Weissen, 1985; Hutchinson et al., 1986; Schier, 1987; Thornton et al., 1987; Arp et al., 1988; Stienen and Bauch, 1988; Keltjens and van Loenen, 1989; Ohno, 1989; Thornton et al., 1989; Wolfe and Joslin, 1989; DeWald et al., 1990; McQuattie and Schier, 1990), but adequate characterization of the extent of mycorrhizal association with the root systems is often lacking. Mycorrhizal associations

ACID DEPOSITION ON FORESTED SOILS

65

with the root systems is often lacking. Mycorrhizal associations with rooting systems are common in forest ecosystems and they have a pronounced effect on nutrient uptake and resistance of roots to toxic conditions. Future studies dealing with root uptake should properly characterize the extent of root-mycorrhizal association, especially in regard to the number of infected root tips and species of mycorrhiza present. An area not discussed in this review but of significance is the long-term influence of soil acidification on heavy metal mobility in forest ecosystems. Atmospheric deposition is a significant source of heavy metals in certain ecosystems, and the affinity with soil organic matter favors their accumulation in the surface horizons (Nuorteva, 1990). The solubility of heavy metals such as Pb, Cd, and Ni increases with decreasing soil pH (Bergkvist, 1986; Tyler et al., 1987), and soil acidification may enhance the biocycling of these elements (Backhaus and Backhaus, 1987; Bergkvist et al., 1989). The role of natural versus anthropogenic acidification in metal mobility in forest soils should not be ignored in future studies dealing with the effects of metal toxicity on soil microorganisms and root growth. Finally, future research on the effects of acidic deposition on forest soils should take into consideration the various models of soil acidification that are currently being developed (Van Grinsven et al., 1989; Schecher and Driscoll, 1989; De Vries et al., -1989; Furrer et al., 1990). These mathematical models will assist in our understanding of relevant soil processes and mechanisms, and in predicting soil acidification rates under different acidic loadings (Reuss et a/., 1987). ExperimentaI manipuiations designed to study specific chemical mechanisms within forest soils should provide relevant data that can be used for verification of the existing soil acidification models (e.g., see Gobran and Agren, 1989; Wright et al., 1990).

VI. CONCLUSIONS There has never been any doubt that high loading rates of acid-forming N and S compounds would eventually alter the soil chemistry of forest soils in northern humid temperate regions. What has been at issue is whether current inputs of acidic deposition are sufficient to cause the changes in soil parameters that have been reported, especially in terms of changes in soil pH, base cation content, and composition of drainage water entering the stream-lake environment. The initial emphasis on changes in soil pH and base cation content as a measure of and necessary condition for soil acidification from acidic inputs probably stems from the concerns raised by Ulrich and his co-workers (1980) at the Solling Forest, where substantial H+ loading has significantly

66

WAYNE P. ROBARGE AND DALE W. JOHNSON

altered the soil chemistry. As noted in this review, there are a number of factors that can alter soil pH and base cation content over time, and the objection raised by Krug and Frink (1983a,b) and others that a mere correlation between areas of acidic deposition and the presence of acid soils and stream waters is not sufficient cause to conclude that acidic inputs have increased soil acidity is correct. However, it was not until the work of Reuss (1983) that the true implication of this statement was realized, especially regarding how natural acidification processes can predispose a forest ecosystem to the way it will respond to acidic inputs. The observation that the dependence of intensity factors on capacity factors in a soil is often nonlinear provided the explanation for how the relatively small inputs of acidity, compared to the total exchangeable acidity in the soil, could have such a significant influence on the composition of the soil solution. The presence of the mineral acid anions NO; and SO:- only enhance the degree of change observed, displacing HCO; and dissolved organic matter as the dominant anions in solution. The observations of Wright and his co-workers (1988a, 1990) at the RAIN project in Norway, and the implications of the bole wood analysis put forward by Bondietti and his colleagus (1989, 1990), together with the variety of reports on changes in stream water chemistry (Schecher and Driscoll, 1989), serve to reinforce the conclusion that changes in the capacity factors of forest soils are not a necessary condition for soil acidification due to acidic inputs. Furthermore, the superposition of acidic deposition upon the natural acidifying processes in forest soils means that soil acidification due to acidic inputs will not be confined to only those soils in the cation exchange buffering range with low base cation content, but that all forest soils in the northern humid temperate region are at risk.

ACKNOWLEDGEMENTS W. P. Robarge’s contribution was funded (in part) by the North Carolina Agricultural Research Service. Support from the Northeastern Forest Experiment Station’s Spruce-Fir Research Cooperative within the joint U.S. Environmental Protection Agency-USDA Forest Service Program (Grant Number A8FS-20,147) is also gratefully acknowledged. D. W. Johnson’s contribution was funded by the Electric Power Research Institute (RP-2621) and the U.S. Forest Service Spruce-Fir Research Cooperative.

REFERENCES Aagaard, P., and Helgeson, H. C. (1982). Thermodynamic and kinetic constraints on reaction rates among minerals and aqueous solutions. I. Theoretical considerations. Am. J . Sci. 282. 237-285.

ACID DEPOSITION ON FORESTED SOILS

67

Adams, F., and Rawajfih, A. (1977). Basaluminite and alunite: A possible cause of sulfate retention in acid soils. Soil Sci. SOC.A m . J . 41, 686-692. Adriano, D. C., and Havas, M., eds. (1989). “Acidic Precipitation,” Vol. 1, pp. 1-311. Springer-Verlag, Berlin. Afifi, A . A., Bricker, 0. P., and Chemerys, J. C. (1985). Experimental chemical weathering of various bedrock types at different pH-values. 1. Sandstone and granite. Chem. Geol. 49, 87-113. ¥, G . I . , and Bosatta, E. (1988). Nitrogen saturation of terrestrial ecosystems. Environ. Pollut. 54, 87-113. Alban, D. H. (1982). Effects of nutrient accumulation by aspen, spruce, and pine on soil properties. Soil Sci. SOC.A m . J . 46, 853-861. Alenas, I., and Skarby L. (1988). Throughfall of plant nutrients in relation to crown thinning in a Swedish coniferous forest. Water, Air, Soil Pollut. 38, 223-237. Alexander, M. (1980). Effects of acidity on microorganisms and microbial processes in soil. In “Effects of Acid Precipitation on Terrestrial Ecosystems” (T. C. Hutchinson and M. Havas, eds.), pp. 363-374. Plenum, New York. Anderson, S. (1988). Long-term changes (1930-1932 to 1984) in the acid-base status of forest soils in the Adirondacks of New York. Ph.D. Thesis, University of Pennsylvania, Philadelphia. Anderson, F. 0. (1989). Air pollution impact of Swedish forests-Present evidence and future development. Environ. Monit. Assess. 12, 29-38. Aneja, V. P., Robarge, W. P., Claiborn, C. S., Murthy, A., Soo-Kim, D., Li, Z . , and Cowling, E. B. (1991). Chemical climatology of high elevation spruce-fir forests in the southern Appalachian mountains. Environ. Pollut. (in press). Arp, P. A. (1984). Forest floor variability and factor analysis: A case study. Can. J . Soil Sci. 64,457-461. Arp, P. A., and Krause, H. H. (1984). The forest floor: Lateral variability as revealed by systematic sampling. Can. J . Soil Sci. 64, 423-437. Arp, H., Bengtsson, B., and Jensen, P. (1988). Growth and cation uptake in spruce (Picea abies Karst.) grown in sand culture with various aluminum contents. Plant Soil 111, 127- 133. Autry, A. R., Fitzgerald, J. W., and Caldwell, P. R. (1990). Sulfur fractions and retention mechanisms in forest soils. Can. J . For. Res. 20, 337-342. Bache, B. W . (1974). Soluble aluminum and calcium-aluminum exchange in relation to the pH of dilute calcium chloride suspensions of acid soils. J. Soil Sci. 25, 320-332. Backhaus, B., and Backhaus, R. (1987). Distribution of long range transported lead and cadmium in spruce stands affected by forest decline. Sci. Total Environ. 59, 283-290. Ball, D. F., and Williams, W. M. (1968). Variability of soil chemical properties in two uncultivated brown earths. J . Soil Sci. 19, 379-391. Beckett, P. H . T., and Webster, R . (1971). Soil variability: A review. Soils Fert. 34, 1-15. Becquer, T., Merlett, D., Boudot, J. P., Rouiller, J., and Gras, F. (1990). Nitrification and nitrate uptake: Leaching balance in a declined forest ecosystem in eastern France. Plant Soil 125, 95-107. Bengtsson, G., and Tranvik, L. (1989). Critical metal concentrations for forest soil invertebrates. Water, Air, Soil Pollut. 47, 381-417. Berdtn, M.. Nilsson, S. I., Rosen, K., and Tyler, G. (1987). “Soil Acidification Extent, Causes and Consequences,” Rep. No. 3292. National Swedish Environmental Protection Board, Solna, Sweden. Bergkvist, B. (1986). Leaching of metals from a spruce forest soil as influenced by experimental acidification. Water, Air, Soil Pollut. 31, 901-916. Bergkvist, B., Folkeson, L., and Berggren, D. (1989). Fluxes of Cu, Zn, Pb, Cd, Cr and Ni in temperate forest ecosystems. A literature review. Water, Air, Soil Pollut. 47, 217-286.

68

WAYNE P. ROBARGE AND DALE W. JOHNSON

Bergstrom, S., Carlsson, B., Sandberg, G., and Maxe, L. (1985). Integrated modelling of runoff, alkalinity and pH on a daily basis. Nord. Hydrol. 16, 89-104. Berner, R. A. (1978). Rate control of mineral dissolution under earth surface conditions. Am. J . Sci. 278, 1235-1252. Bertsch, P. M. (1990). The hydrolytic products of aluminum and their biological significance. Environ. Geochem. Health 12, 7-14. Beven, K., and Germann, P. (1982). Macropores and water flow in soils. Water Resour. Res. 18, 1311-1325. Billett, M. F., Fitzpatrick, E. A., and Cresser, M. S. (1988). Long-term changes in the acidity of forest soils in North-East Scotland. Soil Use Manage. 4, 102-107. Billett, M. F., Fitzpatrick, E. A., and Cresser, M. S. (1990a). Changes in the carbon and nitrogen status of forest soil organic horizons between 1949/50 and 1987. Environ. Pollut. 66, 67-79. Billett, M. F., Parker-Jewis, F., Fitzpatrick, E. A., and Cresser, M. S. (1990b). Forest soil chemical changes between 1949/50 and 1987.1. Soil Sci. 41, 133-145. Binkley, D., and Richter, D. (1987). Nutrient cycles of H + budgets of forest ecosystems. Adv. Ecol. Res. 16, 2-51. Binkley, D., and Sollins, P. (1990). Factors determining differences in soil pH in adjacent conifer and alder-conifer stands. Soil Sci. SOC. Am. J . 54, 1427-1433. Binkley, D., Driscoll, C. T., Allen, H. L., Schoeneberger, P., and McAvoy, D. (1989a). “Acidic Deposition and Forest Soils: Context and Case Studies of the Southeastern United States,” Eco. Stud., Vol. 72. Springer-Verlag, Berlin. Binkley, D., Valentine, D., Wells, C., and Valentine, U. (1989b). An empirical model of the factors contributing to 20-yr decrease in soil pH in an old-field plantation of loblolly pine. Biogeochemistry 8, 39-54. Bishop, K. H., Grip, H., and O’Neill, A. (1990). The origins of acid runoff in a hillslope during storm events. J . Hydro. 116, 35-61. Bloom, P. R. (1981). Metal-organic matter interactions in soil. In “Chemistry in the Soil Environment,” pp. 129-150. Am. SOC.Agron. Madison, Wisconsin. Bloom, P. R. (1983). The kinetics of gibbsite dissolution in nitric acid. Soil Sci. SOC.A m . J . 47, 164-168. Bloom, P. R., and Grigal, D. F. (1985). Modeling soil response to acidic deposition in nonsulfate adsorbing soils. J . Environ. Qual. 14, 489-495. Bloom, P. R., and Weaver, R. M. (1982). Effect of the removal of reactive surface material on the solubility of synthetic gibbsites. Clays Clay Miner. 30, 281-286. Bloom, P. R., McBride, M. B., and Weaver, R. M. (1979). Aluminum organic matter in acid soils: Buffering and solution aluminum activity. Soil Sci. SOC.A m . J . 43, 448-493. Bolt, G. H., ed. (1982). “Soil Chemistry. Part B. Physico-Chemical Models”. Elsevier, Amsterdam. Bolt, G. H., Bruggenwert, M. G. M., and Kamphoret, A. (1976). Adsorption of cations by soil. In “Soil Chemistry. Part A. Basic Elements” ( G . H. Bolt and M. G. M. Bruggenwert, eds.), pp. 54-95. Elsevier, Amsterdam. Bondietti, E. A., Baes, C. F., 111, and McLaughlin, S. B. (1989). Radial trends in cation ratios in tree rings as indicators of the impact of atmospheric deposition on forests. Can. J . For. Res. 19, 586-594. Bondietti, E. A., Momoshima, N . , Shortle, W. C., and Smith, K. T. (1990). A historical perspective on divalent cation trends in red spruce stemwood and the hypothetical relationship to acidic deposition. Can. J . For. Res. 20, 1850-1858. Brand, D. G., Kehoe, P., and Connors, M. (1986). Coniferous afforestation leads to soil acidification in central Ontario. Can. J . For. Res. 16, 1389-1391. Bredemeier, M. (1987). Quantification of ecosystem-internal proton production from the ion balance of the soil. Plant Soil 101, 273-280.

ACID DEPOSITION ON FORESTED SOILS

69

Bredemeier, M. (1988). Forest canopy transformation of atmospheric deposition. Water, Air, Soil Pollut. 40, 121-138. Bredemeier, M., Matzner, E., and Ulrich, B. (1990). Internal and external proton load to forest soils in Northern Germany. J . Environ. Qual. 19, 469-477. Bringmark, E. (1989). Spatial variation in soil pH of beech forests in relation to buffering properties and soil depths. Oikos 54, 165-177. Bruck, R. I., Robarge, W. P., and McDaniel, A. (1989). Forest decline in the boreal montane ecosystems of the southern Appalachian mountains. Water, Air, Soil Pollut. 48, 161-180. Bruggenwert, M. G. M., and Kamphorst, A. (1982). Survey of experimental information on cation exchange in soil systems. In “Soil Chemistry. Part B. Physico-Chemical Models” (G. H. Bolt, ed.), pp. 141-193. Elsevier, Amsterdam. Buol, S. W., Hole, F. D., and McCracken, R. J. (1980). “Soil Genesis and Classification,” 2nd/Ed. Iowa State Uni. Press, Ames. Carroll, S. A., and Walther, J. V. (1990). Kaolinite dissolution at 25”, 60”, and 80°C. Am. J . Sci. 290, 797-810. Caspary, H. J. (1990). An ecohydrological framework for water yield changes of forested catchments due to forest decline and soil acidification. Water Resour. Res. 26,1121-1131. Castelle, A. J., and Gafloway, J. N. (1990). Carbon dioxide dynamics in acid forest soils in Shenandoah National Park, Virginia. Soil Sci. SOC.Am. J . 54, 252-257. Chen, C. W., Gherini, S. A., Dean, J . D., Hudson, R. J. M., and Goldstein, R. A. (1984). Development and calibration of the Integrated Lake-Watershed Acidification Study model. In “Modelling of Total Acid Precipitation Impacts” (J. L. Schnoor, ed.), pp. 175-203. Butterworth, Boston, Massachusetts. Choon, K. K., and Roffael, E. (1990). The acidity of five hardwood species. Holzforschung 44, 53-58. Christophersen, N., and Wright, R. F. (1981). Sulfate budget and a model for sulfate concentrations in streamwater at Birkenes, a small forested catchment in southernmost Norway. Water Resour. Res. 17, 377-389. Christophersen, N., Seip, H. M., and Wright, R. F. (1982). A model for streamwater chemistry at Birkenes, Norway. Water Resour. Res. 18, 977-996. Cole, D. W., and Johnson, D. W. (1977). Atmospheric sulfate additions and cation leaching in a Douglas-fir ecosystem. Water Resour. Res. 13, 313-317. Cole, D. W., and Rapp, M. (1981). Elemental cycling in forest ecosystems. In “Dynamic Properties of Forest Ecosystems” (D. Reichle, ed.), pp. 341-409. Cambridge University Press, London. Coleman, D. C. (1983). The impacts of acid deposition on soil biota and C cycling. Environ. EXP.Bot. 23, 225-233. Coleman, D. C., Reid, C. P. P., and Cole, C. V. (1983). Biological strategies of nutrient cycling in soil systems. Adv. Ecol. Res. 13, 1-55. Coleman, N. T., and Thomas, G. W. (1967). The basic chemistry of soil acidity. In “Soil Acidity and Liming” (R. W. Pearson and F. Adams, eds.), pp. 1-41. Am. SOC.Agron. Madison, Wisconsin. Colman, S. M., and Dethier, D. P., eds. (1986). “Rates of Chemical Weathering of Rocks and Minerals.” Academic Press, Orlando, Florida. Cosby, B. J., Hornberger, G. M., Galloway, J. N., and Wright, R. F. (1985). Modeling the effects of acid deposition: Assessment of a lumped parameter model of soil water and streamwater chemistry. Water Resour. Res. 21, 51-63. Courchesne, F., and Hendershot, W. H. (1989). Sulfate retention in some podzolic soils of the southern Laurentians, Quebec. Can 1. Soil Sci. 69, 337-350. Coutts, M. P. (1989). Factors affecting the direction of growth of tree roots. Am. Sci. For. 46, SUPPI.,277s-287s.

70

WAYNE P. ROBARGE AND DALE W. JOHNSON

Cozzarelli, I. M., Herman, J. S., and Parnell, R. A., Jr. (1987). The mobilization of aluminum in a natural soil system: Effects of hydrologic pathways. Water Resour. Res. 23, 859-874. Cronan, C. S. (1985). Chemical weathering and solution chemistry in acid forest soils: Differential influence of soil type, biotic processes, and H+ deposition. In “The Chemistry of Weathering” (J. I. Drever, ed.), pp. 175-195. Reidel, Dordrecht, The Netherlands. Cronan, C. S., and Goldstein, R. A. (1989). ALBIOS: A comparison of aluminum biogeochemistry in forested watersheds exposed to acidic deposition. In “Acidic Precipitation’’ (D. C. Adriano and M. Havas, eds.), Vol. 1, pp. 113-135. Springer-Verlag, Berlin. Cronan, C. S., and Schofield, C. L. (1990). Relationships between aqueous aluminum and acidic deposition in forested watersheds of North America and Northern Europe. Environ. Sci. Technol. 24, 1100-11105. Cronan, C. S., Walker, W. J., and Bloom, P. R. (1986). Predicting aqueous aluminum concentrations in natural waters. Nature (London) 324, 140-143. Cronan, C. S., Driscoll, C. T., Newton, R. M., Kelly, J. M., Schofield, C. L., Bartlett, R. J., and April, R. (1990). A comparative analysis of aluminum biogeochernistry in a northeastern and a southeastern forested watershed. Water Resour. Res. 26, 1413-1430. Dahlgren, R. A., McAvoy, D. C., and Driscoll, C. T. (1990). Acidification and recovery of a spodosol Bs horizon from acidic deposition. Environ. Sci. Technol. 24, 531-537. David, M. B., and Driscoll, C. T. (1984). Aluminum specification and equilibria in soil solutions of Haplorthods in the Adirondack mountains (New York, U.S.A.). Geoderma 33, 297-318. David, M. B., and Mitchell, M. J. (1987). Transformations of organic and inorganic sulfur: Importance to sulfate flu in an Adirondack Forest Soil. J . Air Pollut. Control Assoc. 37,39-44. David, M. B., and Vance G. F. (1989). Generation of soil solution acid-neutralizingcapacity by addition of dissolved inorganic carbon. Environ. Sci. Technol. 23, 1021-1024. David, M. B., Reuss, J. O., and Walthall, P. M. (1988). Use of a chemical equilibrium model to understand soil chemical processes that influence soil solution and surface water alkalinity. Water, Air, Soil Pollut. 38, 71-83. Davidson, C. I., and Wu, Y. L. (1989). Dry deposition of particles and vapors. In “Acidic Precipitation” (S. E. Lindberg, A. L. Page, and S. A. Norton, eds.), Vol. 3, pp. 103-216. Springer-Verlag, Berlin. Dethier, D. P., Jones, S. B., Feist, T. P., and Ricken, J. E. (1988). Relations among sulfate, aluminum, iron, dissolved organic carbon, and pH in upland forest soils of northwestern Massachusetts. Soil Sci. SOC. Am. J . 52, 506-512. De Vries, W. (1988). Critical deposition levels for nitrogen and sulphur on Dutch forest ecosystems. Water, Air, Soil Pollut. 42, 221-239. De Vries, W., and Breeuwsma, A. (1987). The relation between soil acidification and element cycling. Water, Air, Soil Pollut. 35, 293-310. De Vries, W., Posch, M., and Kamari, J. (1989). Simulation of the long-term soil response to acid deposition in various buffer ranges. Water, Air, Soil Pollut. 48, 349-390. DeWald, L. E., Sucoff, E. I., Ohno, T., and Buschena, c. A. (1990). Response of northern red oak (Quercus rubra) seedlings to soil solution aluminum. Can. J . For. Res. 20, 331-336. DeWalle, D. R., Sharpe, W. E., and Edwards, P. J. (1988). Biogeochemistry of two Appalachian deciduous forest sites in relation to episodic stream acidification. Water, Air, Soil Pollut. 40, 143-156. Dhamala, B. R., Mitchell, M. J., and Stam, A. C. (1990). Sulfur dynamics in mineral

ACID DEPOSITION ON FORESTED SOILS

71

horizons of two northern hardwood soils. A column study with 35S.Biogeochemistry 10, 143-160. Donaldson, J. M., and Henderson, G. S . (1990). Nitrification potential of secondarysuccession upland oak forests. I. Mineralization and nitrification during laboratory incubations. Soil Sci. SOC.A m . J . 54, 892-897. Drever, J. I., ed. (1985). “The Chemistry of Weathering.” Reidel, Dordrecht, The Netherlands. Driscoll, C. T., and Likens, G. E . (1982). Hydrogen ion budget of an aggrading forest ecosystem. Tellus 34, 283-292. Driscoll, C. T., and Schecher, W. D. (1990). The chemistry of aluminum in the environment. Environ. Ceochem. Health 12, 28-49. Driscoll, C. T., van Breeman, N., and Mulder, J. (1985). Aluminum chemistry in a forested spodosol. Soil Sci. Soc. A m . J . 49, 437-444. Driscoll, C. T., Johnson, N. M., Likens, G. E., and Feller, M. C. (1988). Effects of acidic deposition on the chemistry of headwater streams: A comparison between Hubbard Brook, New Hampshire, and Jamieson Creek, British Columbia. Water Resour. Res. 24, 195-200. Duijsings, J. J. H. M., Verstraten, J. M., and Bouten, W. (1986). Spatial variability in nutrient deposition under an oak/beach canopy. Z. Ppanzenernaehr. Bodenkd. 149, 718-727. Evans, A., Jr., and Anderson, T. J. (1990). Aliphatic acids: Influence on sulfate mobility in a forested Cecil soil. Soil Sci. SOC.Am. J . 54, 1136-1139. Evans, A . , Jr., and Zelazny, L. W. (1987). Effects of sulfate additions on the status of exchangeable aluminum in a Cecil soil. Soil Sci. 143, 410-417. Evans, A., Jr., and Zelazny, L. W. (1990). Kinetics of aluminum and sulfate release from forest soil by mono- and diprotic aliphatic acids. Soil Sci. 149, 324-330. Falkengren-Grerup, U. (1987). Long term changes in pH of forest soils in southern Sweden. Environ. Pollur. 43, 79-90. Falkengren-Grerup, U., and Eriksson, H. (1990). Changes in soil, vegetation, and forest yield between 1947 and 1988 in beech and oak sites of southern Sweden. For. Ecol. Manage. (in press). Falkengren-Grerup, U., and Tyler, G. (1987). Changes of cation pools of the humus layer in south Swedish beech forests between 1979 and 1989. Scand. J . For. Res. (in press). Falkengren-Grerup, U., Linnermark, N., and Tyler, G. (1987). Changes in acidity and cation pools of south Swedish soils. Chemosphere 16, 10-12. Federer, C. A. (1982). Subjectivity in the separation of organic horizons of the forest floor. Soil Sci. SOC. A m . J . 46, 1090-1093. Federer, C. A,, and Hornbeck, J. W. (1985). The buffer capacity of forest soils in New England. Water, Air, Soil Pollut. 26, 163-173. Fernandez, I. J. (1987). Vertical trends in the chemistry of forest soil microcosms following experimental acidification. Maine Agric. Exp. Stan., Tech. Bull. 126. Fernandez, I. J. (1989). Effects of acidic precipitation on soil productivity. In “Acidic Precipitation” (D. C. Adriano and A. H. Johnson, eds.), Vol 2, pp. 61-83. SpringerVerlag, Berlin. Fernandez, I. J . , and Kosian, P. A. (1986). Chemical response of soil leachate to alternative approaches to experimental acidification. Commun. Soil. Sci. Plant Anal. 17, 953-973. Fernandez, I. J., and Kosian, P. A. (1987). Soil air carbon dioxide concentrations in a New England spruce-fir forest. Soil Sci. Soc. A m . J . 51, 261-263. Fernandez, I. J., and Struehtemeyer, R. A. (1985). Chemical characteristics of soils under spruce-fir forests in eastern Maine. Can. J . Soil Sci. 65, 61-69. Folster, H. (1985). Proton consumption rates in Holocene and present-day weathering of acid

72

WAYNJZ P. ROBARGE AND DALE W. JOHNSON

forest soils. In “The Chemistry of Weathering” (J. I. Drever, ed.), pp. 197-209. Reidel, Dordrecht, The Netherlands. Foster, N. W., Hazlett, P. W., Nicolson, J. A., and Morrison, I. K. (1989). Ion leaching from a sugar maple forest in response to acidic deposition and nitrification. Water, Air, Soil Pollut. 48, 251-261. Franklin, J. F., Dyrness, C. T. Moore, D. G., and Tarrant, R. F. (1968). Chemical soil properties under coastal Oregon stands of alder and conifers. In “Biology of Alder” (J. M. Trappe, ed.), pp. 157-172. USDA PNW For. Range Exp. Sta., Portland, Oregon. Frogner, T. (1990). The effect of acid deposition on cation fluxes in artificially acidified catchments in western Norway. Geochim. Cosmochim. Acta 54, 769-780. Fuller, R. D., Driscoll, C. T., Schindler, S. C., and Mitchell, M. J. (1986). A simulation model of sulfur transformations in forested Spodosols. Biogeochembtry 2, 313-328. Fuller, R. D., Driscoll, C. T., Lawrence, G. B., and Nodvin, S. C. (1987). Processes regulating sulphate flux after whole-tree harvesting. Nature (London) 325, 707-710. Fuller, R. D., Simone, D. M., and Driscoll, C. T. (1988). Forest clearcutting effects on trace metal concentrations: Spatial patterns in soil solutions and streams. Water, Air, Soil Pollut. 40,185-195. Furrer, G., and Stumm, W. (1983). The role of surface coordination in the dissolution of S-A1203in dilute acids. Chimica 37, 338-341. Furrer, G., Sollins, P., and Westall, J. C. (1990). The study of soil chemistry through quasi-steady-state models. 11. Acidity of soil solution. Geochim. Cosmochim. Acta 54, 2363-2374. Gaines, G. L., and Thomas, H. C. (1953). Adsorption studies on clay minerals. 11. A formulation of the thermodynamics of exchange adsorption. J. Chem. Phys. 21, 714-718. Garten, C. T., Jr. (1990). Foliar leaching, translocation, and biogenic emission of 3sS in radiolabeled loblolly pines. Ecology 71, 239-251. Garten, C. T., Jr., Bondietti, E. A,, and Lomax, R. D. (1988). Contribution of foliar leaching and dry deposition to sulfate in net throughfall below deciduous trees. A a o s . Environ. 22, 1425-1432. Gaskin, J. W., Dowd, J. F. Nutter, W. L., and Swank, W. T. (1989). Vertical and lateral components of soil nutrient flux in a hillslope. J. Environ. Qual. 18, 403-410. Gherini, S . , Mok, L., Hudson, R. J. M., Davis, G. F., Chen C., and Goldstein, R. (1985). The ILWAS model: Formulation and application. Water, Air, Soil Pollur. 26, 95-113. Giovanoli, R., Schnoor, J. L., Sigg, L., Stumm, W., and Zobrist, J. (1988). Chemical weathering of crystalline rocks in the catchment area of acidic Ticino Lakes, Switzerland. Clays Clay Miner. 36, 521-529. Gobran, G. R., and Agren, G. I. (1989). Significance of changes in K, values for Ca-A1 exchange and its effects on soil and water acidification predictions. Ecol. Modell. 44, 165-175. Goldstein, R. A., Chen, C. W., and Gherini, S . A. (1985). Integrated lake-watershed acidification study: Summary. Water, Air, Soil Pollut. 26,327-337. Gorham, E., Martin, F. B., and Litzau, J. T. (1984). Acid rain: Ionic correlations in the eastern United States, 1980-1981. Science 225, 407-409. Grennfelt, P., and Hultberg, H. (1986). Effects of nitrogen deposition on the acidification of terrestrial and aquatic ecosystems. Water, Air, Soil Pollut. 30, 945-963. Grimm, R., and Rehfuess, K. E. (1986). Kurzfristige Veranderungen von Bodenreaktion und Kationenaustauscheigenschaftenin einem Meliorationsversuch zu Kiefer auf PodsolPseudogley in der Oberpfalz. A&. Forst- Jagdrtg. 157, 205-213. Gundersen, P., and Beier, C. (1988). Aluminum sulphate solubility in acid forest soils in Denmark. Water, Air, Soil Pollut. 39, 247-261.

ACID DEPOSITION ON FORESTED SOILS

73

Gundersen, P., and Rasmussen, L. (1990). “Nitrification in Forest Soils: Effects from Nitrogen Deposition on Soil Acidification and Aluminum Release ,”Rev. Environ. Contam. Toxicol., Vol. 113. Springer-Verlag, Berlin. Haines, S. G., and Cleveland, G. (1981). Seasonal variation in properties of five forest soils in southwest Georgia. Soil Sci. SOC.Am. J. 45,139-143. Hallbacken, L., and Tamm, C. 0. (1986). Changes in soil acidity from 1927 to 1982-1984 in a forest area of southwest Sweden. Scand. J. For. Res. 1, 219-232. Harrison, R. B., Johnson, D. W., and Todd, D . E. (1989). Sulfate adsorption and desorption reversibility in a variety of forest soils. J . Environ. Qual. 18, 419-426. Hartikainen, H. (1986). Acid and base titration behaviour of Finnish mineral soils. Z. @anzenernaehr. Bodenkd. 149, 522-432. Hauhs, M. (1989). Lange Bramke: An ecosystem study of a forested catchment. In “Acidic Precipitation” (D. C. Adriano and M. Havas, eds.), Vol. 1, pp. 273-305. SpringerVerlag, Berlin. H a m, G. W.,Wolt, J., and Reynolds, J. H. (1988). Effects of simulated acid rain on soil solution composition and orchardgrass seedling viability. Soil Sci. SOC. Am. J . 52, 1037-1043. Havas, M., Hutchinson, T. C., and Likens, G. E. (1984). Red herrings in acid rain research. Environ. Sci. Technol. 18, 176A-186A. Henriksen, A., and Brakke, D. R. (1988a). Increasing contributions of nitrogen to the acidity of surface waters in Norway. Water, Air, Soil Pollut. 42, 183-201. Henriksen, A., and Brakke, D. R. (1988b). Sulfate deposition to surface waters. Environ. Sci. Technol. 22, 8-14. Hildebrand, E. E. (1986a). Ein Verfahren zur Gewinnung der GleichgewichtsBodenporenlosung. Z. Pfanzenernaehr. Bodenkd, 149, 340-346. Hildebrand, E. E. (1986b). Zustand und Entwicklung der Austauscheigenschaften von Mineralboden aus Standorten mit erkrankten Waldbestanden. Forswirs. Centralbl. 105, 60-76. Hildebrand, E. E. (1987). “Zustand und Entwicklung chemischer Eigenschaften von Mineralbiiden aus Standorten mit erkrankten Waldbestanden,” PEFRep. No. 28. Projekt Europaisches Forschungszentrum fur MaBnahmen zur Luftreinhaltung, Karlsruhe, Germany. Hildebrand, E. E. (1990). Die Bedeutung der Bodenstruktur fur die Waldernahrung, dargestellt am Beispiel des Kaliums. Forsrwiss. Centralbf. 109, 2-12. Hodges, S . C., and Zelazny, L. W. (1983). Interactions of dilute, hydrolzed aluminum solutions with clays, peat, and resin. Soil Sci. SOC. Am. J. 47, 206-212. Hogberg, P., Granstrom, A , , Johansson, T., Lundmark-Thelin, A., and Nasholm, T. (1986). Plant nitrate reductase activity as an indicator of availability of nitrate in forest soils. Can. J. For. Res. 16, 1165-1169. Holdren, G. R., Jr., and Speyer, P. M. (1986). Stoichiometry of alkali feldspar dissolution at room temperature and various pH values. In “Rates of Chemical Weathering of Rocks and Minerals” (S. M. Colman and D. P. Dethier, eds.), pp, 61-81. Academic Press, Orlando, Florida. Hopper, R. P., Christophersen, N., and Peters, N. E. (1990). Modeling streamwater chemistry as a mixture of soilwater endmembers-An application to Panola mountain catchment, Georgia, U.S.A. J. Hydrol. 116, 321-343. Howard, P. J. A., and Howard, D. M. (1990). Titratable acids and bases in tree and shrub leaf litters. Forestry 63, 177-196. Hsu, P. H., and Bates, T. F. (1964). Fixation of hydroxy-aluminumpolymers by vermiculite. SoiI Sci. Sac. Am. Proc. 28, 763-768. Hsu, P. H., and Rich, C. I. (1960). Aluminum fixation in a synthetic cation exchanger. Soil Sci. SOC.Am. Proc. 24, 21-25.

74

WAYNE P. ROBARGE AND DALE W. JOHNSON

Huettl, R. F., Fink, S . , Lutz, H.-J., Poth, M., and Wisniewski, J. (1990). Forest decline, nutrient supply and diagnositic fertilization in southwestern Germany and in southern California. For. Ecol. Manage. 30, 341-350. Hutchinson, T. C.,Bozic, L., and Munoz-Vega, G. (1986). Responses of five species of conifer seedlings to aluminum stress. Water, Air, Soil Pollut. 31, 283-294. Ike, A. F., and Cutter, J. L. (1967). The variability of forest soils of the Georgia Blue Ridge Mountains. Soil Sci. SOC. A m . Proc. 31, 284-288. Ineson, P., and Wookey, P. A. (1988). Effects of sulphur dioxide on forest letter decomposition and nutrient release. I n “Air Pollution and Ecosystems” (P. Mathy, ed.), pp. 254-260. Reidel, Dordrecht, The Netherlands. James, B. R., and Riha, S. J. (1986). pH buffering in forest soil organic horizons: Relevance to acid precipitation. J. Environ. Qual. 15, 229-234. James, B. R., and Riha, S. J. (1987). Forest soil organic horizon acidification: Effects of temperature, time, and solution/soil ratio. Soil Sci. SOC. Am. J. 51, 458-462. Jardine, P. M., Zelazny, L. W., and Parker, J. C. (1985). Mechanisms of aluminum adsorption on clay minerals and peat. Soil Sci. SOC.Am. J. 49, 862-867. Jardine, P. M., Wilson, G . V., and Luxmoore, R. J. (1990). Unsaturated solute transport through a forest soil during rain storm events. Geoderma 46, 103-118. Jenkinson, D. S. (1970). The accumulation of organic matter in soil left uncultivated. Comm. Bur. Soil. Sci. Rothamstad Exp. Sta. Rep. for 1970. Johnson, C. E., Johnson, A. H., and Huntington, T. G. (1990) Sample size requirements for the determination of changes in soil nutrient pools. Soil Sci. 150, 637-644. Johnson, D. W., and Henderson, G. S. (1979). Sulfate adsorption and sulfur fractions in a highly weathered soil under a mixed deciduous forest. Soil Sci. 128, 34-40. Johnson, D. W., and Lindberg, S. E. (1989). Acidic deposition on Walker Branch watershed. I n “Acidic Precipitation” (D. C. Adriano and M. Havas, eds.), Vol. 1, pp. 1-38. Springer-Verlag, Berlin. Johnson, D. W., and Taylor, G. E. (1989). Role of air pollution in forest decline in eastern North America. Water, Air, Soil Pollut. 48, 21-43. Johnson, D. W., and Todd, D. E . (1983). Relationships among iron, aluminum, carbon, and sulfate in a variety of forest soils. Soil Sci. SOC.Am. J. 47, 792-800. Johnson, D. W., and Todd, D. E. (1984). Effects of acid irrigation on carbon dioxide evolution, extractable nitrogen, phosphorus, and aluminum in a deciduous forest soil. Soil Sci. SOC. Am. J. 48, 664-666. Johnson, D. W., and Todd, D . E. (1987). Nutrient export by leaching and whole-tree harvesting in a loblolly pine and mixed oak forest. Plant Soil 102, 99-109. Johnson, D. W., and Todd, D. E. (1990). Nutrient cycling in forests of Walker Branch Watershed: Roles of uptake and leaching in causing soil change. J . Environ. Qual. 19, 97-104. Johnson, D. W., Hornbeck, J. W., Kelly, J. M., Swank, W. T., and Todd, D. E. (1980). Regional patterns of soil sulfate accumulation: Relevance to ecosystem sulfur budgets. In “Atmospheric Sulfur Deposition” (D. S. Shriner, C. R. Richmond, and S. E. Lindberg, eds.), pp. 507-520. Ann Arbor Sci., Ann Arbor, Michigan. Johnson, D. W., Henderson, G. S., andTodd, D. E. (1981). Evidence of modern accumulation of sulfate in an east Tennessee forested Ultisol. Soil Sci. 132, 422-426. Johnson, D. W., Turner, J., and Kelly, J. M. (1982). The effects of acid rain on forest nutrient status. Water Resour. Res. 18, 449-461. Johnson, D. W., Richter, D. D., Van Miegroet, H., and Cole, D. W. (1983). Contributions of acid deposition and natural processes to cation leaching from forest soils: A review. J. Air Pollut. Control Assoc. 33, 1036-1041. Johnson, D. W., Richter, D. D., Lovett, G. M., and Lindberg, S. E. (1985). The effects of

ACID DEPOSITION ON FORESTED SOILS

75

atmospheric deposition on potassium, calcium, and magnesium cycling in two deciduous forests. Can. J . For. Res. 15, 773-782. Johnson, D. W., Richter, D. D., Van Miegroet, H. Cole, D. W., and Kelly, J. M. (1986a). Sulfur cycling in five forest ecosystems. Water, Air, Soil Pollut. 30, 965-979. Johnson, D. W., Cole D. W., Van Miegroet, H., and Horg, F. W. (1986b). Factors affecting anion movement and retention in four forest soils. Soil Sci. SOC. A m . J . 50, 776-783. Johnson, D. W., Henderson, G. S., and Todd, D. E. (1988a). Changes in nutrient distribution in forests and soils of Walker Branch Watershed, Tennessee, over an eleven-year period. Biogeochemistry 5, 275-293. Johnson, D. W., Kelly, J. M., Swank, W. T., Cole, D. W., Van Miegroet, H., Hornbeck, J. W., Pierce, R. S., and Van Lear, D. (1988b). The effects of leaching and whole-tree harvesting on cation budgets of several forests. J. Environ. Qual. 17, 418-424. Johnson, D. W., Friedland, A. J., Van Miegroet, H., Harrison, R. B., Miller, E., Lindberg, S. E., Cole, D. W., Schaefer, D. A., and Todd, D . E . (1989). Nutrient status of some contrasting high-elevation forests in the eastern and western United States. Proceedings of the U.S.-FRG Symposium: Effects of Atmospheric Pollutanis on the Spruce-”r Forests of the Eastern United States and the Federal Republic of Germany. Burlington, VT, Oct. 18-23, 1987. Northeastern Forest Experiment Station Gen. Tech. Rep. NE-120, Broomall, PA, pp. 453-460. Johnson, D. W., Cresser, M. S., Nilsson, S . I., Turner, J., Ulrich, B., Binkley, D., and Cole, D. W. (1991a). Soil changes in forest ecosystems: Evidence for and probable causes. Proc. R. SOC. Edinburgh Sec. B. Biol. Sci. 97, 81-116. Johnson, D. W., Van Miegroet, H., Lindberg, S . E., Todd, D. E., and Harrison, R. B. (1991b). Nutrient cycling in red spruce forests of the Great Smoky Mountains. Can. J . For. Res. 21, 769-787. Johnson, N. M. (1985). Acid rain neutralization by geologic materials. In “Geological P. Bricker, ed.), pp. 37-53. Butterworth, Boston, Aspects of Acid Deposition” (0. Massachusetts. Johnson, N. M., Driscoll, C. T., Eaton, J.S., Likens, G . E., and McDowell, W. H. (1981). ‘Acid Rain,’ dissolved aluminum and chemical weathering at the Hubbard Brook Experimental Forest, New Hampshire. Geochim. Cosmochim. Acta 45, 1421-1437. Johnson, N. M., Likens, G. E., Feller, M. C., Driscoll, C. T., Seip, H. M., Dillon, P. J., Wright, R. F., Krug, E . C., and Frink, C . R. (1984). Acid rain and soil chemistry. Science 225, 1424-1429. Jones. J. A . A. (1987). The effects of soil piping on contributing areas and erosion patterns. Earth Surf. Processes Landforms 12, 229-248. Joslin, J. D., and Wolfe, M. H. (1988). Responses of red spruce seedlings to change in soil aluminum in six amended forest soil horizons. Can. J . For. Res. 18, 1614-1623. Joslin, J. D., McDuffie, C., and Brewer, P. F. (1988). Acidic cloud water and cation loss from red spruce foliage. Water, Air, Soil Pollut. 39, 355-363. Kaupenjohann, M., Schneider, B. U., Hantschel, R., Zech, W., and Horn, R. (1988). Sulfuric acid rain treatment of Piceu abies (Karst. L): Effects on nutrient solution, throughfall chemistry, and tree nutrition. Z . Pfanzenernaehr. Bodenkd. 151, 123126. Keeney, D. R. (1980). Prediction of soil nitrogen availability in forest ecosystems: Literature review. For. Sci. 26, 159-171. Kelly, J. M., and Strickland, R. C. (1986). Throughfall and plant nutrient concentration response to simulated acid rain treatment. Water, Air, Soil Polht. 29, 219-231. Keltjens, W. G., and Van Loenen, E. (1989). Effects of aluminum and mineral nutrition of growth and chemical composition of hydroponically grown seedlings of five different forest tree species. Plant Soil 119, 39-50.

76

WAYNE P. ROBARGE AND DALE W. JOHNSON

Khanna, P. K., Weaver, G. T., and Beese, F. (1986). Effect of sulfate on ionic transport and balance in a slightly acidic forest soil. Soil Sci. SOC. A m . J. 50, 770-775. Khanna, P. K., Prenzel, J., Meiwes, K. J., Ulrich, B., and Matzner, B. (1987). Dynamics of sulfate retention by acid forest soils in an acidic deposition environment. Soil Sci. SOC. A m . J . 51,446-452. Kirkby, M. (1988). Hillslope runoff processes and models. J. Hydrol. 100,315-339. Klumpp, V. A., and Guderian, R. (1990). Leaching von Magnesium, Kalzium und Kalium aus immissionsbelasteten Nadeln junger Fichten (Picas abies [L.] Karst.). Forsrwiss. Centralbl. 109, 13-39. Kohlmaier, G. H., Sire, E. O., Brohl, H., Kilian, W . , Fischback, U., Plochl, M., Miiller, T., and Yunsheng, J. (1983/1984). Dramatic development in the dying of German spruce-fir forests: In search of possible cause-effect relationships. Ecol. Modell. 22, 45-65. Kratochvil, B., Goewie, C. E. and Taylor, J. K. (1986). Sampling theory for environmental analysis. Trends Anal. Chem. 5, 253-256. Krause, G. H. M., Arndt, U . , Brandt, C. J., Buchner, J., Kenk, G., and Matzner, E. (1986). Forest decline in Europe: Development and possible causes. Water, Air, Soil Pollut. 31, 647-668. Krug, E. C., and Frink, C. R. (1983a). Acid rain on acid soil: A new perspective. Science 221, 520-525. Krug, E. C., and Frink, C. R. (1983b). “Effects of Acid Rain on Soil and Water,” Bull. 811. Conn. Agric. Exp. St., New Haven. Krug, E. C., and Isaacson, P. J . (1984). Comparison of water and dilute acid treatment on organic and inorganic chemistry of leachate from organic-rich horizons of an acid forest soil. Soil Sci. 137, 370-378. Lawrence, G. B., and Driscoll, C. T. (1988). Aluminum chemistry downstream of a wholetree-harvested watershed. Environ. Sci. Technol. 22, 1293-1299. Lawrence, G. B., Fuller, R. D., and Driscoll, C. T. (1987). Release of aluminum following whole-tree harvesting at the Hubbard Brook Experimental Forest, New Hampshire. J. Environ. Qual. 16, 383-390. Leininger, T. D., and Winner, W. E. (1988). Throughfall chemistry beneath Quercus rubru: Atmospheric, foliar, and soil chemistry considerations. Can. J. For. Res. 18, 478-482. Lindberg, S. E., Lovett, G. M., Richter, D. D., and Johnson, D. W. (1986). Atmospheric deposition and canopy interactions of major ions in a forest. Science 231, 141-145. Lindsay, W. L. (1979). “Chemical Equilibria In Soils.” Wiley, New York. Lovett, G. M., Reiners, W. A., and Olson, R. K. (1982). Cloud droplet deposition in subalpine Balsam fir forests: Hydrological and chemical inputs. Science 218, 13031304. Lovett, G. M., Reiners, W. A., and Olson, R. K. (1989). Factors controlling throughfa!! chemistry in a balsam fir canopy: A modeling approach. Biogeochemisrry 8, 239-264. Lundstrom, U., and Ohman, L.-0. (1990). Dissolution of feldspars in the presence of natural, organic solutes. J . Soil Sci. 41, 359-369. Lynch, J. A., and Corbett,E. S. (1989). Hydrologic control of sulfate mobility in a forested watershed. Water Resour. Res. 25, 1695-1703. MacDonald, N. W., and Hart, J. B., Jr. (1990). Relating sulfate adsorption to soil properties in Michigan forest soils. Soil Sci. Soc. A m . J . 54, 238-245. Mader, D. L. (1963). Soil variability-A serious problem in soil-site studies in the northeast. Soil Sci. SOC. Am. Proc. 27, 707-709. Magdoff, F. R., and Bartlett, R. J. (1985). Soil pH buffering revisited. Soil Sci. SOC.A m . J. 49, 145-148. Matzner, E. (1989). Acidic precipitation: Case study Solling. In “Acidic Precipitation” (D. C . Adriano and M. Havas, eds.), Vol. 1, pp. 39-83. Springer-Verlag, Berlin.

ACID DEPOSITION ON FORESTED SOILS

77

May, H . M., Helmke, P. A., and Jackson, M. L. (1979). Gibbsite solubility and thermodynamic properties of hydroxy-aluminum ions in aqueous solution at 25°C. Geochim. Cosmochim. Acfa 43, 861-868. Maynard, D. G., Kalra, Y. P., and Radford, F. G . (1987). Extraction and determination of sulfur in organic horizons of forest soils. Soil Sci. SOC. A m . J . 51, 801-805. McBratney, A. B., and Webster, R. (1983). How many observations are needed for regional estimation of soil properties? Soil Sci 135, 177-183. McFee, W. W. (1980). Sensitivity of soil regions to long-term acid precipitation. In “Atmospheric Sulfur Deposition” (D. S . Shriner, C. R. Richmond, and S. E. Lindberg, eds.), pp. 495-505. Ann Arbor Sci., Ann Arbor, Michigan. McFee, W. W. (1983). Sensitivity ratings of soils to acid deposition: A review. Environ. Exp. Bot. 23, 203-210. 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. A m . Proc. 29, 432-436. McQuattie, C . J., and Schier, G. A. (1990). Response of red spruce seedlings to aluminum toxicity in nutrient solution: Alterations in root anatomy. Can. J . For Res. 20, 10011011. Mehlich, A. (1964). Influence of sorbed hydroxyl and sulfate on liming efficiency, pH, and conductivity. Soil Sci. SOC. A m . Proc. 27, 496-499. Mitchell, M. J . , Driscoll, C. T., Fuller, R. D., David, M. B., and Likens, G. E. (1989). Effects of whole-tree harvesting on the sulfur dynamics of a forest soil. Soil Sci. SOC.A m . J . 53, 933-940. Momoshima, N., and Bondietti, E. A. (1990). Cation binding in wood: Applications to understanding historical changes in divalent cation availability to red spruce. Can. J . For. Res. 20, 1840-1849. Montagnini, F., Haines, B., and Swank, W. T. (1989a). Factors controlling nitrification in soils of early successional and oak/hickory forests in the southern Appalachians. For. Ecol. Manage. 26, 77-94. Montagnini, F., Haines, B . , Swank, W. T., and Waide, J. B. (1989b). Nitrification in undisturbed mixed hardwoods and manipulated forests in the southern Appalachian mountains of North Carolina, U.S.A. Can. J . For. Res. 19, 1226-1234. Murphy, C. E., Jr., and Sigmon, J. T. (1989). Dry deposition of sulfur and nitrogen oxide gases to forest vegetation. In “Acidic Precipitation” (S. E. Lindberg, A. L. Page, and S. A. Norton, eds.), Vol. 3, pp. 217-240. Springer-Verlag, Berlin. Myrold, D. D. (1990). Effects of acidic deposition on soil organisms. In “Mechanisms of Forest Response to Acidic Deposition” (A. A. Lucier and S. G. Haines, eds.) pp. 163187. Springer-Verlag, Berlin. Neal, C. (1988). Aluminum solubility relationships in acid waters-A practical example of the need for a radical reappraisal. J . Hydrol. 104, 141-159. Neal, C., Skeffington, R. A., Williams, R., and Roberts, D. J . (1987). Aluminum solubility controls in acid waters: The need for a reappraisal. Earfh Planet. Sci. Left. 86, 113-119. Neary, A. J., Mistry, E., and Vanderstar, L. (1987). Sulphate relationships in some central Ontario forest soils. Can. J . Soil Sci. 67, 341-352. Neilsen, G. H., and Hoyt, P. B. (1982). Soil pH variation in British Columbia apple orchards. Can. J. Soil Sci. 62, 695-698. Nesbitt, H. W., and Muir, I. J. (1988). SIMS depth profiles of weathered plagioclase and processes affecting dissolved Al and Si in some acidic soil solutions. Nature (London) 334,336-338. Nikolaidis, N. P., Rajaram, H., Schnoor, J. L., and Georgakakos, K. L. (1988). A generalized soft water acidification model. WaferResour. Res. 24, 1983-1996.

78

WAYNE P. ROBARGE AND DALE W. JOHNSON

Nilsson, S . I., and Bergkvist, B. (1983). Aluminum chemistry and acidification processes in a shallow podzol on the Swedish west coast. Water, Air, Soil Pollut. 20, 311-329. Nilsson, S. I., Miller, H. G . , and Miller, J. D. (1982). Forest growth as a possible cause of soil and water acidification: An examination of the concepts. Oikos 39, 40-49. Nodvin, S. C., Driscoll, C. T., and Likens, G . E. (1986). Simple partitioning of anions and dissolved organic carbon in a forest soil. Soil Sci. 142, 27-35. Nodvin, S. C., Driscoll, C. T., and Likens, G. E. (1988). Soil processes and sulfate loss at the Hubbard Brook Experimental Forest. Biogeochemistry 5, 185-199. Nordstrom, D. K. (1982). The effect of sulfate on aluminum concentrations in natural waters: Some stability relations in the system A1203-S03-H20 at 2983. Geochirn. Cosmochim. Acta 46, 681-692. Norton, S. A., Brakke, D. F., and Henriksen, A. (1989). ‘Red Herring’ lakes and streams in the acid-rain literature. Sci. Total Environ. 83, 113-125. Novozamsky, I., and Beek, J. (1976). Common solubility equilibria in soils. In “Soil Chemistry. Part A. Basic Elements” (G. H. Bolt and M. G. M. Bruggenwert, eds.), pp. 96125. Elsevier, Amsterdam. Nuorteva, P. (1990). “Metal Distribution Patterns and Forest Decline. Seeking Achilles’ Heels for Metals in Finnish Forest Biocoenoses,” Publ. Dep. Environ. Conserv. Helsinki Univ. No. 11. University of Helsinki, Helsinki, Finland. Ohno, T. (1989). Rhizosphere pH and aluminum chemistry of red oak and honeylocust seedlings. Soil Biol. Biochern. 21, 657-660. Pates, T. (1983). Rate constants of dissolution derived from the measurements of mass balance in hydrological catchments. Geochim. Cosmochim. Acta 47, 1855-1863. Pates, T. (1986). Weathering rates of gneiss and depletion of exchangeable cations in soils under environmental acidification. J . Geol. Soc., London 143, 673-677. Pallant, E., and Riha, S. J. (1990). Surface soil acidification under red pine and Norway spruce. Soil Sci. SOC. Am. J . 54, 1124-1130. Parker, D. R., Kinraide, T. B., and Zelazny, L. W. (1988). Aluminum speciation and phytotoxicity in dilute hydroxy-aluminum solutions. Soil Sci. SOC.Am. J . 52, 438-444. Parker, D. R., Kinraide, T. B., and Zelazny, L.W.(1989). On the phytotoxicity of polynuclear hydroxy-aluminum complexes. Soil Sci. SOC. A m . J . 53, 789-796. Parker, G. G . (1983). Throughfall and stem flow in the forest nutrient cycle. Adv. Ecol. Res. 13, 58-134. Parker, G. G . (1990). Evaluation of dry deposition, pollutant damage, and forest health with throughfall studies. In “Mechanisms of Forest Response to Acidic Deposition” (A. A. Lucier and S . G. Haines, eds.), pp. 10-61. Springer-Verlag, Berlin. Percy, K. E. (1989). Vegetation, soils, and ion transfer through the forest canopy in two Nova Scotia lake basins. Water, Air, Soil Pollut. 46, 73-86. Perkins, T. D., Vogelmann, H. W., and Klein, R. M. (1987). Changes in light intensity and soil temperature as a result of forest decline on Camels Hump, Vermont. Can. J . For. Res. 17, 565-568. Peters, N. E., and Driscoll, C. T. (1987). Hydrogeologic controls of surface-water chemistry in the Adirondack region of New York State. Biogeochemistry 3, 163-180. Peterson, D. L., and Rolfe, G . L. (1982). Seasonal variation in nutrients of floodplain and upland forest soils of central Illinois. Soil Sci. SOC. A m . J . 46, 1310-1315. Quesnel, H. J., and Lavkulich, L. M. (1980). Nutrient variability of forest floor near Port Hardy, British Columbia, Canada. Can. J . Soil Sci. 60, 565-573. Raynal, D. J., Joslin, J. D., Thornton, F. C., Schadedel, M., and Henderson, G. S. (1990). Sensitivity of tree seedlings to aluminum: 111. Red spruce and loblolly pine. J . Environ. Qual. 19, 180-187. Rechcigl, J. E., and Sparks, D. L. (1985). Effect of acid rain on the soil environment: A review. Commun. Soil Sci. Plant Anal. 16, 653-680.

ACID DEPOSITION ON FORESTED SOILS

79

Reisinger, L. M., and Imhoff, R. E. (1989). Analysis of summertime cloud water measurements made in a southern Appalachian spruce forest. Water, Air, Soil Pollut. 45, 1-15. Reuss, J. 0. (1983). Implications of the calcium-aluminum exchange system for the effect of acid precipitation on soils. J . Environ. Qual. 12, 591-595. Reuss, J. 0. (1989). Soil-solution equilibria in lysimeter leachates under red alder. In “Effects of Air Pollution on Western Forests” (R. K. Olson and A. S. LeFohn, eds.), pp. 547-559. Air Waste Manage. Assoc., Anaheim, California. Reuss, J. O., and Johnson, D. W. (1985). Effect of soil processes on the acidification of water by acid deposition. J . Environ. Qual. 14, 26-31. Reuss, J . O., and Johnson, D. W. (1986). “Acid Deposition and the Acidification of Soils and Waters,” Ecol. Stud. Ser. No. 59. Springer-Verlag, Berlin. Reuss, J . 0.. and,Walthall, P. M. (1989). Soil reaction and acidic deposition. In “Acidic Precipitation” (S. A. Norton, S. E. Lindberg, and A. L. Page, eds.), Vol. 4, pp. 1-33. Springer-Verlag, Berlin. Reuss, J . 0..Cosby, B. J., and Wright, R. F. (1987). Chemical processes governing soil and water acidification. Nature (London) 329, 27-32. Reuss, J . O., Walthall, P. M., Roswall, E. C., and Hopper, R. W. E. (1990). Aluminum solubility, calcium-aluminum exchange, and pH in acid forest soils. Soil Sci. SOC.A m . J . 54, 374-380. Reynolds, B., Cape, J. N . , and Paterson, I. S. (1989). A comparison of element fluxes in throughfall beneath Larch and Sitka spruce at two contrasting sites in the United Kingdom. Forestry 62, 29-39. Ritcher, D. D., Johnson, D. W., and Todd, D. E. (1983). Atmospheric sulfur deposition, neutralization, and ion leaching in two deciduous forest ecosystems. J . Environ. Qual. 12, 263-270. Richter, D. D., Comer, P. J., King, K. S . , Swain, H. S . , and Wright, D. S. (1988). Effects of low ionic strength solutions on pH of acid forest soils. Soil Sci. SOC.A m . J . 52,261-264. Riha, S. J., James, B. R., Senesac, G . P., and Pallant, E. (1986a). Spatial variability of soil pH and organic matter in forest plantations. Soil Sci. SOC. A m . J . 50, 1347-1352. Riha, S. J . , Senesac, G. P., and Pallant, E. (1986b). Effects of forest vegetation on spatial variability of surface mineral soil pH, soluble aluminum and carbon. Water, Air, Soil Pollut. 31, 929-940. Robarge, W. P., and Corey, R. B. (1979). Adsorption of phosphate by hydroxy-aluminum species on a cation exchange resin. Soil Sci. SOC. A m . 1. 43, 481-487. Roberge, J . , and Plamondon, A. P. (1987). Snowmelt runoff pathways in a boreal forest hillslope, the role of pipe throughfall J . Hydrol. 95, 39-54. Rochelle, B . P., and Church, M. R . (1987). Regional patterns of sulfur retention in watersheds of the Eastern U.S. Water, Air, Soil Pollut. 36, 61-73. Rochelle, B. P., Chruch, M. R., and David, M. B. (1987). Sulfur retention at intensively studied sites in the U.S. and Canada. Water, Air, Soil Pollut. 33, 73-83. Rosenqvist, I. Th. (1978). Alternative sources for acidification of river water in Norway. Sci. Total Environ. 10, 39-49. Rosenqvist, I . Th. (1990). From rain to lake: Water pathways and chemical changes. J . Hydrol. 116, 3-10. Rustad, L. E., and Cronan, C. S. (1989). Cycling of aluminum and nutrients in litterfall of a red spruce (Pica rubens Sarg.) stand in Maine. Can. J . For. Res. 19, 18-23. Saxena, V. K . , and Lin, H . H. (1990). Cloud chemistry measurements and estimates of acidic deposition on an above cloudbase coniferous forest. A t o m . Environ. 24A, 329-352. Saxena, V. K., Stogner, R. E., Mendler, A. M., DeFelice, T. P., Yeh, R. J., and Lin, N. M. (1989). Monitoring the chemical climate of Mt. Mitchell State Park to evaluate its impact on forest decline. Tellus 41B, 92-109. Schaefer, D. A , , and Reiners, W. A. (1989). Throughfall chemistry and canopy processing

80

WAYNE P. ROBARGE AND DALE W. JOHNSON

mechanisms. In “Acidic Precipitation” (S. E. Lindberg, A. L. Page, and S. A. Norton, eds.), Vol. 3, pp. 241-284. Springer-Verlag, Berlin. Schecher, W. D., and Driscoll, C. T. (1989). Principles and applications of surface water acidification models. In “Hazard Assessment of Chemicals” (J. Saxena, ed.), Vol. 6, pp. 187-223. Hemisphere, New York. Scherbatskoy, T., and Klein, R. M. (1983). Reponse of spruce and birch foliage to leaching by acidic mists. J . Environ. Qua(. 12, 189-195. Schier, G . A. (1985). Response of red spruce and balsam fir seedlings to aluminum toxicity in nutrient solutions. Can. J . For. Res. 15, 29-33. Schier, G. A. (1987). Germination and early growth of four pine species on soil treated with simulated acid rain. Can. J . For. Res. 17, 1190-1196. Schindler, S . C., and Mitchell, M. J. (1987). Dynamics of 35Sin horizons and leachates from a hardwood forest Spodosol. Soil Biol. Biochem. 19, 531-538. Schindler, S. C., Mitchell, M. J . , Scott, T. J., Fuller, R. D., and Driscoll, C. T. (1986). Incorporation of 3sS-3ulfate into inorganic and organic constituents of two forest soils. Soil Sci. SOC.Am. J. 50, 457-462. Schwertmann, U . , and Fischer, W. R. (1982). pH-Verteilung und Pufferung von Boden. Z . PJ7anzenernaehr. Bodenkd. 145, 221-223. Schwertmann, U . , Siisser, P., and Natscher, L. (1987). Protonenpuffer-Substanzen in Boden. 2. Ppanzenernaehr. Bodenkd. 150, 174-178. Seip, H. M. (1980). Acidification of freshwaters: Sources and mechanisms. In “Ecological Impacts of Acid Precipitation” (D. Drablos and A. Tollan, eds.), pp. 358-366. Johs. Grefslie Trykkeri AS, Mysen, Norway. Shortle, W. C., and Smith, K. T. (1988). Aluminum-induced calcium deficiency syndrome in declining red spruce. Science 240, 1017-1018. Sigmon, J. T., Gillian, F. S., and Partin, M. E. (1989). Precipitation and throughfall chemistry for a montane hardwood forest ecosystem: Potential contributions from cloudwater. Can. J . For. Res. 19, 1240-1247. Singh, B. R. (1984). Sulfate sorption by acid forest soils. 2. Sulfate adsorption isotherms with and without organic matter and oxides of aluminum and iron. Soil Sci. 138, 294-297. Skeffington, R. A , , and Wilson, E. J . (1988). Excess nitrogen deposition: Issues for consideration. Environ. Pollut. 54, 159-184. Smith, W. H. (1990). The atmosphere and the rhizosphere: Linkages with potential significance for forest tree health. In “Mechanisms of Forest Response to Acidic Deposition” (A. A. Lucier and S. G. Haines, eds.), pp. 188-241. Springer-Verlag, Berlin. Soil Science Society of America (1987). “Glossary of Soil Science Terms.” Am. SOC.Agron., Madison, Wisconsin. Sposito, G. (1981). “The Thermodynamics of Soil Solutions.” Oxford Univ. Press. (Clarendon), London. Sposito, G . (1989). “The Chemistry of Soils.” Oxford Univ. Press, New York. Sposito, G., Holtzolaw, K. M., LeVesque, C. S . , and Johnston, C. T. (1982). Trace metal chemistry in arid-zone field soils amended with sewage sludge. 11. Comparative study of the fulvic acid fraction. Soil Sci. SOC.A m . J . 46, 265-270. Stams, A. J. M., and Marnette, E. C. L. (1990). Investigation of nitrification in forest soils with soil percolation columns. Plant Soil 125, 135-141. Stanko, K . M., and Fitzgerald, J . W. (1990). Sulfur transformations in forest soils collected along an elevational gradient. Soil Biol. Biochem. 22, 213-216. Stevens, P. A. (1987). Throughfall chemistry beneath Sitka spruce of four ages in Beddgelert Forest, North Wales, U.K. Plant Soil 101, 291-294. Stienen, H., and Bauch, J . (1988). Element content in tissue of spruce seedlings from hydroponic cultures simulating acidification and deacidification. Plant Soil 106,231-238. Strader, R. H., Binkley, D., and Wells, C. S. (1989). Nitrogen mineralization in high

ACID DEPOSITION ON FORESTED SOILS

81

elevation forests of the Appalachians. I. Regional patterns in southern spruce-fir forests. Biogeochemistry 7, 13 1- 145. Strickland, T. C., and Fitzgerald, J . W. (1984). Formation and mineralization of organic sulfur in forest soils. Biogeochernistry 1, 79-95. Strickland, T. C . , Fitzgerald, J. W., Ash, J. T., and Swank, W. T. (1987). Organic sulfur transformations and sulfur pool sizes in soil and litter from a southern Appalachian hardwood forest. Soil Sci. 143, 453-458. Stumm, W., and Morgan, J . J. (1981). “Aquatic Chemistry: An Introduction Emphasizing Chemical Equilibria in Natural Waters,” 2nd Ed. Wiley (Interscience), New York. Swank. W. T. (1986). Biological control of solute losses from forest ecosystems. In “Solute Processes” (J. T. Trudpill, ed.), pp. 87-139. New York. Swank, W. T., Fitzgerald, J. W., and Ash, J. T. (1984). Microbial transformation of sulfate in forest soils. Science 223, 182-184. Swistock, B. R . , DeWalle, D. R., and Sharpe, W. E. (1989). Sources of acidic storm flow in an Appalachian headwater stream. Water Resour. Res. 25, 2139-2147. Tabatabai, M. A . (1985). Effect of acid rain on soils. CRC Crit. Rev. Environ. Control 15. Tabatabai, M. A. (1987). Physicochemical fate of sulfate in soils. J . Air Pollut. Control ASSOC. 37, 34-38. Tamm, C. 0..and Hallbacken, L. (1986). Changes in soil pH over a 50-year period under different forest canopies in SW Sweden. Water, Air, Soil Pollut. 31, 337-341. Tamm, C. O . , and Hallbacken, L. (1988). Changes in soil acidity from the 1920s to the 1980s in two forest areas with different acid deposition. Ambio 17, 56-61. Tanner, R. L. (1989). Sources of acids, bases, and their precursors in the atmosphere. In “Acidic Precipitation” (S. E. Lindberg, A. L. Page, and S. A. Norton, eds.), Vol. 3, pp. 1-19. Springer-Verlag, Berlin. Thomas, G . W. (1988). Beyond exchangeable aluminum: Another ride on the merry-goround. Commun. Soil Sci. Plant Anal. 19, 833-856. Thornton, F. C., Schaedle, M., and Raynal, D. J. (1987). Effects of aluminum on red spruce seedlings in solution culture. Environ. Exp. Bot. 27, 489-498. Thornton, F. C., Schaedle, M., and Raynal, D. J. (1989). Tolerance of red oak and American and European beech seedlings to aluminum. J . Environ. Qual. 18, 541-545. Troedsson, T., and Tamm, C. 0. (1969). Small-scale spatial variation in forest soil properties and its implications for sampling procedures. Stud. For. Suec 74, 1-30. Turner, J . , and Kelly, J. (1977). Soil chemical properties under naturally regenerated Eucalyptus spp. and planted Douglas fir. Aust. For. Res. 7 , 163-172. Turner, J., and Kelly, J. (1981). Relationships between soil, nutrients and vegetation in a north coast forest, New South Wales. A u t . For. Res. 11, 201-208. Turner, R. C. (1967). Aluminum removed from solution by montmorillonite. Can. J . Soil Sci. 47, 217-222. Turner, R. S., Johnson, A. H., and Wang, D. (1985). Biogeochemistry of aluminum in McDonalds Branch watershed, New Jersey Pine Barrens. J . Environ. Qual. 14,314-323. Tyler, G., Berggren, D . , Bergkvist, B., Falkengren-Grerup, U., Folkeson, L., and Ruhling, A. (1987). Soil acidification and metal solubility in forests of south Sweden. In “NATO AS1 Series, Effects of Atmospheric Pollutants on Forest, Wetlands and Agricultural Ecosystems, Toronto, 12-17 May 1985” (T. C. Hutchinson and K. M. Meema, eds.), pp. 347-359. Springer, Berlin. Ugolini, F. C.. Minden, R., Dawson, H., and Zachara, J. (1977). An example of soil processes in the Abies amabilis zone of Central Cascades, Washington. Soil Sci. 124, 291-302. Ulrich, B. (1980). Production and consumption of hydrogen ions in the ecosphere. In “Effects of Acid Precipitation on Terrestrial Ecosystems” (T. C. Hutchinson and M. Havas, eds.), pp. 255-282. Plenum, New York.

82

WAYNE P. ROBARGE AND DALE W. JOHNSON

Ulrich, B. (1983). A concept of forest ecosystem stability and of acid deposition as driving force for destabilization. In “Effects of Accumulation of Air Pollutants in Forest Ecosystems” (B. Ulrich and J. Pankrath, eds.), pp. 1-29. Rei.CI, Dordrecht, The Netherlands. Ulrich, B. (1985). Interaction of indirect and direct effects of air pollutants in forests. In “Air Pollution and Plants” (C. Trozanowsky, ed.), pp. 151-181. VCH Verlagsges., Weinheim. Ulrich, B. (1986). Natural and anthropogenic components of soil acidification. Z . w a n zenernaehr. Bodenkd. 149, 702-717. Ulrich, B. (1987). Stability, elasticity, and resilience of terrestrial ecosystems with respect to matter balance. In “Potentials and Limitations of Ecosystem Analysis” (E. -D. Schulze and H. Zwolfer, eds.), Vol. 61, pp. 11-49. Springer-Verlag, Berlin. Ulrich, B. (1989). Effects of acidic precipitation on forest ecosystems in Europe. In “Acidic Precipitation” (D. C . Adriano and A. H. Johnson, eds), Vol. 2, pp. 189-272. SpringerVerlag, Berlin. Ulrich, B., Mayer, R., and Khanna, P. K. (1980). Chemical changes due to acid precipitation in a loess-derived soil in central Europe. Soil Sci. 130, 193-199. van Breemen, N., Burrough, P A., Velthorst, E. J., van Dobben, H. F., de Wit, T., Ridder, T. B., and Reijnders, H. F. R. (1982). Soil acidification from atmospheric ammonium sulphate in forest canopy throughfall. Nature (London) 299, 548-550. van Breemen, N., Mulder, J., and Driscoll, C. T. (1983). Acidification and alkalinization of soils. Plant Soil 75, 283-308. van Breemen, N., Driscoll, C. T., and Mulder, J. (1984). Acidic deposition and internal proton sources in acidification of soils and waters. Nature (London) 307, 599-604. van Breemen, N., De Visser, P. H. B., and van Grinsven, J. J. M. (1986). Nutrient and proton budgets in four soil-vegetation systems underlain by Pleistocene alluvial deposits. J . Geol. SOC.,London 143, 659-666. van Breemen, N., Boderie, P. M. A., and Booltink, H. W. G. (1989). Influence of airborne ammonium sulfate on soils of an oak woodland ecosystem in the Netherlands: Seasonal dynamics of solute fluxes. In “Acidic Precipitation” (D. C. Adriano and M. Havas, eds.), Vol. 1, pp. 209-236. Springer-Verlag, Berlin. van Grinsven, J. J. M . , Kros, J., van Breemen, N., van Riemsdijk, W. H., and van Eek, E . (1989). Simulated response of an acid forest soil to acid deposition and mitigation measures. Neth. J . Agric. Sci. 37, 279-299. Van Miegroet, H., and Cole, D. W. (1984). The impact of nitrification on soil acidification and cation leaching in a red alder forest. J . Environ. Qual. 13, 586-590. Van Miegroet, H . , and Cole, D . W. (1985). Acidification sources in red alder and Douglas fir soils-Importance of nitrification. Soil Sci. SOC.Am. J . 49, 1274-1279. Van Miegroet, H., and Cole, D. W. (1988). Influence of nitrogen-fixing alder on acidification and cation leaching in a forest soil. In “Forest Site Evaluation and Long-Term Productivity” (D. W. Cole and S. P. Gessel, eds.). University of Washington Press, Seattle. van Praag, H. J., and Weissen, F. (1985). Aluminum effects on spruce and beech seedlings. 1. Preliminary observations on plant and soil. Plant Soil 83, 331-338. Veith, J. A. (1978). Selectivity and adsorption capacity of smectite and vermiculite for aluminum of varying basicity. Clays Clay Miner. 26, 45-50. Velbel, M. A. (1985). Geochemical mass balances and weathering rates in forested watersheds of the southern Blue Ridge. Am. J . Sci. 285, 904-930. Velthorst, E. J., and van Breemen, N. (1989). Changes in the composition of rainwater upon passage through the canopies of trees and of ground vegetation in a Dutch oak-birch forest. Plant Soil 119, 81-85.

ACID DEPOSITION ON FORESTED SOILS

83

Veneman, P. L. M., and Bodine, S. M. (1982). Chemical and morphological soil characteristics in a New England drainage toposequence. Soil Sci. SOC. Am. J . 46, 359-363. Veneman, P. L. M., Jacke, P. V., and Bodine, S. M. (1984). Soil formation as affected by pit and mound microrelief in Massachusetts, USA. Geoderma 33, 89-99. Vogt, R., Seip, H. M., Christophersen, N., and Andersen, S. (1990). Assessment of aluminum mobilization and pathways in the Birkenes catchment, southern Norway. Sci. Total En~iron.96, 139-158. Walker, W. J., Cronan, C. S., and Bloom, P. R. (1990). Aluminum solubility in organic soil horizons from northern and southern forested watersheds. Soil Sci. SOC. Am. J . 54, 369-374. Watwood, M. E., and Fitzgerald, J. W. (1988). Sulfur transformations in forest litter and soil: Results of laboratory and field incubations. Soil Sci. SOC. Am. J . 52, 1478-1483. Weathers, K. C., Likens, G. E., Bormann, F. H., Bicknell, S . H., Bormann, B. T., Daube, B. C., Jr., Eaton, J. S., Galloway, J. N., Keene, W. C., Kimball, K. D., McDowell, W. H., Siccama, T. G., Smiley, D., and Tarrant, R. A. (1988). Cloudwater chemistry from 10 sites in North America. Environ. Sci. Technol. 22, 1018-1026. Weaver, G. T., Khanna, P. K., and Beese, F. (1985). Retention and transport of sulfate in a slightly acid forest soil. Soil Sci. SOC.Am. J . 49, 746-750. Webster, R., and Burgess, T. M. (1984). Sampling and bulking strategies for estimating soil properties in small regions. J . Soil Sci. 35, 127-140. Whipkey, R. Z . , and Kirkby, M. J. (1978). Flow within the soil. In “Hillslope Hydrology” (M. J. Kirkby, ed.), pp. 121-144. Wiley, London. White, G . N., Feldman, S . B., and Zelazny, L. W. (1990). Rates of nutrient release by mineral weathering. In “Mechanisms of Forest Response to Acidic Deposition” (A. A. Lucier and S . G. Haines, eds.), pp. 108-162. Springer-Verlag, Berlin. Wolfe, M. H., and J o s h , J. D. (1989). Honeylocust (Gleditsia triacanfhos L.) root response to aluminum and calcium. Plant Soil 119, 181-185. Wolfe, M. H., Kelly, J. M., and Wolt, J. D. (1987). Soil pH and extractable sulfate-sulfur distribution as influenced by tree species and distance from the stem. Soil Sci. SOC.Am. J . 51, 1042-1046. Wollast, R., and Chou, L. (1985). Kinetic study of the dissolution of albite with a continuous Bow-through fluidized bed reactor. In “The Chemistry of Weathering” (J. I. Drever, ed.), pp. 75-96. Reidel, Dordrecht, The Netherlands. Wok, J. D., and Lietzke, D. E. (1982). The influence of anthropogenic sulfur inputs upon soil properties in the Copper Basin Region of Tennessee. Soil Sci. SOC. Am. J . 46, 651-656. Wright, R. F. (1988). Influence of acid rain on weathering rates. In “Physical and Chemical Weathering in Geochemical Cycles” (A. Lerman and M. Meybeck, eds.), pp. 181-196. Kluwer, New York. Wright, R. F., Lotse, E., and Semb, A. (1988a). Reversibility of acidification shown by whole-catchment experiments. Nature (London) 334, 670-675. Wright, R. F., Norton, S . A , , Brakke, D. F., and Frogner, T. (1988b). Experimental verification of episodic acidification of freshwaters by sea salts. Nature (London) 334, 422-424. Wright, R . F., Cosby, B. J . , Flaten, M. B., and Reuss, J. 0. (1990). Evaluation of an acidification model with data from manipulated catchments in Norway. N a m e (London) 343, 1185-1187. Young, S . D., and Backe, B. W. (1985). Aluminum-organic complexation: Formation constants and a speciation model for the soil solution. J . Soil Sci. 36, 261-269.

This Page Intentionally Left Blank

FINGERPRINTING CROP VARIETIES J. S. C. Smith and 0.S. Smith Plant Breeding Division, Pioneer Hi-Bred International, Inc., Johnston, Iowa SO131

I. Introduction A. Evolution in the Scope and Technology of Fingerprinting B. The Needs to Classify and Identify Cultivated Varieties: An Introductory Overview

II. Characters Used to Fingerprint Cultivated Varieties A. Morphological Data B. Protein Data C. DNAData

ID.DiscriminationalAbility of FingerprintingTechniques A. Factors Affecting DiscriminationalAbility B. DiscriminationalAbility of Protein and DNA Descriptors for Selected Major Cultivated Species IV. Usage of Fingerprints A. The Nature of Plant Breeding and the Need for Long-Term Research Investment B. Plant Variety Protection, Registration, Certification, and Patents C. Minimum Distance D. Measurement of Genetic Diversity E. Assurance of Genetic Purity F. The Improvement of Germplasm in Terms of Agronomic Performance G. Evidence and Prevention of Misappropriation V. New Techniques References

I. INTRODUCTION A. EVOLUTION IN THE SCOPE AND TECHNOLOGY OF FINGERPRINTING During the intervening century since the coining of the term “fingerprinting,” the definition, techniques, and applications of the procedure have broadened considerably. Advances in biochemistry, genetics, and

85 Advancer in Agronomy, Volume 47 Copyright 0 1992 by Academic Press, Inc. All righu of reproduction in any form reserved

86

J. S. C. SMITH AND 0. S. SMITH

molecular biology have provided descriptors based on proteins (including isoenzymes) and deoxyribonucleic acid (DNA). These molecules, which are universal in the animal and plant kingdoms, can now be used as evidence of similarity in numerous taxa, including domesticated plant species. Previously, morphological records of fingerprint impressions, with their complex and undeciphered genetic control (Holt,1968), were restricted to individual identification. In contrast, protein- and DNA-based data are available from genetically based descriptors for which the inheritance often is known. Consequently, there are applications of these data beyond identification alone. They can also be used in tests of parentage, in genetic mapping of loci conditioning economic traits, in measurement of genetic diversity, and in discerning patterns of genetic diversity. Usage of protein and DNA descriptors has propelled the science of both human and cultivated plant identification into the mainstream of peer-reviewed and widely accepted science.

B. THE NEEDS TO CLASSIFY AND IDENTIFY CULTIVATED VARIETIES: AN INTRODUCTORY OVERVIEW Descriptive names given to landraces or varieties illustrate the practical necessity for farmers to identify sources of seed of proved performance. In many cases, processors and end-use consumers also want to know varietal identity. Plant breeders need to describe cultivated varieties because they represent the end product of the investment of time, money, and effort. There would be no incentive for further investment, in the relatively long-term process of plant improvement through breeding, without the hope of a reasonable return on current investment. Finally, descriptions based upon traits that reflect genetic variation can be used to measure genetic diversity and can, therefore, be used to monitor and promote efficient conservation and utilization of genetic diversity.

11. CHARACTERS USED TO FINGERPRINT CULTIVATED VARIETIES

A. MORPHOLOGICAL DATA 1. The Rationale for Using Morphology

Continued usage of morphological data (Pauksens, 1975; National Institute of Agricultural Botany, 1983; Troyer, 1986) to describe cultivars indicates that these data retain popularity as descriptors. However, it is

FINGERPRINTING CROP VARIETIES

87

imperative to reach a consensus on the practicality and validity of using morphological data for cultivar identification and associated uses. Such consideration is especially important because new demands that are made on cultivar description often go beyond that of identification alone. These include a need to calculate the genetic distance between cultivars, information that cannot be provided by morphological data. 2. Constraints upon Morphology

a. A Narrowing of Morphological Diversity Morphological descriptions can provide unique identification of cultivated varieties (Molina-Can0 and Elena Rossello, 1978). However, their ability to provide reliably discriminating identification is at best cumbersome (Patterson and Weatherup, 1984). Increased numbers of genetically related releases by plant breeders have made unique identification more difficult to achieve. This problem is especially acute in crops with limited levels of elite genetic diversity or when convergent selection toward similar morphologies is practiced (Wagner and McDonald, 1981). This situation has been further exacerbated in wheat with the use of nontraditional breeding crosses that have disrupted the previous correlation between grain morphology and quality (Lookhart and Bietz, 1990). b. The Effect of Environment on Morphology Morphologies reflect not only the genetic constitution of the cultivar, but also the interaction of the genotype with the environment (G x E) within which it is expressed (Gottlieb, 1977; Brown, 1978; Lin and Binns, 1984; Patterson and Weatherup, 1984). For example, “Welsh Bearded Red Rough Chaff’ rivet wheat (Triticum aestivum) has an ear color that “depends on the season and varies from dark red to pale yellowish red” (Zeven, 1990). Due to G X E effects, it is clearly inappropriate to compare morphological data for varieties that have been collected across different years and/or locations. Attempts to allow comparisons to be made across more varieties grown in different years and/or locations, or sets of locations, are made by including a list of standard check varieties at each location site in each year. However, G x E interaction effects have been found to cause aberrant adjusted means for traits such that morphological data collected in field plots can provide, at best, only an initial screen of varietal identity or distinctiveness (Higgins ef al., 1988). Subsequent repeated tests of individuality would thus still need to be made, requiring further replicated test plots of varieties. Recent investigations using digital image analysis techniques have shown that objective analysis of the morphologies of roots, leaves, and particularly of cereal grains could provide some degree of varietal identification

88

J. S. C. SMITH AND 0. S. SMITH

(Keefe and Draper, 1986,1988; Neuman et al., 1987; Symons and Fulcher, 1988). However, when image analysis was tested for its ability to correctly identify varieties on an individual grain basis for each of 14 cultivars, the percentage of correct positive identifications ranged from a low of 15% to a maximum of 96%. It was concluded that the ability to provide correct identification of cereal varieties using grain morphology was confounded by close genetic relatedness among cultivars and by environmental effects on grain shape.

c. The Inability of Morphological Data to be Used for the Estimation of Genetic Distance Descriptions based on morphological data are fundamentally flawed in their ability to provide reliable information for the calculation of genetic distance or the validation of pedigrees. First, distances that reflect the genetic constitution of cultivated varieties can only be obtained using measurements from traits that adequately sample the genome. Second, the efficiency of the sampling sites is greatly increased if there are numerous variants that can be unambiguously and repeatedly scored. Third, there needs to be a direct and unambiguous correspondence between the genotype and the phenotype. Morphological data fail to meet these criteria. First, much morphological variation cannot be consistently measured because of the effects of G x E interaction and the small effects of numerous quantitative genes. Second, much obvious morphological variation has been eliminated with the consequence that most varieties outwardly appear similar. Third, for most morphological traits, the genetic control is unknown, although it is known that multiple genotypes can give phenotypes of similar outward appearance. Therefore, it is impossible to determine how completely the genome is sampled by morphological descriptions or the extent to which similar phenotypes reflect similar genotypes. We have concluded that morphological differences cannot be interpreted to provide accurate estimates of genetic differences (J. S. C. Smith and Smith, 1989; J. S. C. Smith et al., 1991a).

B. PROTEIN DATA 1. Characters That Are Selected and Methods Used to Resolve Profiles Proteins and DNA can be used to provide varietal profiles. They are in popular usage because the variation for these markers is ubiquitous and this variation can be understood in genetic terms. These characters are

FINGERPRINTING CROP VARIETIES

89

in routine usage and are widely accepted as a source of reliable data in evolution, taxonomy, and genetics (Tanksley and Orton, 1983; Doebley, 1989; Crawford, 1990; Weir, 1990). Proteins are molecules with net electrical charges that are affected by pH. Proteins can be separated by electrophoresis on the basis of their net electrical charge, molecular weight [see Cooke (1984) for a brief listing of methods], isoelectric point, or combinations of these criteria utilizing multidimensional separations (Figs. 1-4). They can also be separated by chromatography, wherein the surface chemistry of the protein molecule determines relative retention times on an absorption matrix (Bietz, 1986) (Fig. 5). Proteins can be visualized in several ways. Variation in enzymatic proteins (isoenzymes) can be revealed by specific histochemical stains. General proteins can be revealed by various staining procedures, the most sensitive of which employ silver nitrate. Chemical stains are eclipsed in their detection power by the incorporation of radiolabeled amino acids into proteins and their subsequent exposure on film. Proteins can also be used as antigens to generate antibodies. Similarities between varieties can then be estimated by measuring the degree of response between antigen and antibodies raised to other antigens (Skerritt et al., 1984; Esen et al., 1989). Serological methods have not yet revealed unique profiles among numerous varieties. However, some differences have been revealed (Zawistowski and Howes, 1990). If greater specificity can be achieved, then serological techniques could provide very rapid and cost-effective means of identification.

2. Sampling of Tissue, Extraction, and Running Conditions Although proteins are products of the primary transcripts of DNA, environmental factors can affect qualitative and quantitative levels of protein in seed (Higgins, 1984). Proteins also interact with other compounds, especially those that are found in nonseed organs such as roots, leaves, and tubers. These interactions can detract from the reproducibility of protein profiles. Initially, a plant organ is selected as a source of protein. Then, extraction protocols are chosen to minimize the interaction of protein with other constituents that could be undesirable sources of variation in the protein profile. Also, a time period or stage of development must be identified during which the protein gives profiles that are stable for any given cultivar. Seeds provide a convenient source of tissue that is at a defined and stable stage of development (Ladizinsky and Hymowitz, 1979) and that is free from many compounds that interfere with proteins. Seedlings provide a rich source of enzymatically active proteins and are often the

90

J. S. C. SMITH AND 0. S. SMITH

Figure 1. Gliadin electrophoretic profiles from individual caryopsis extracts of two wheat varieties. Needles point to differences in banding profiles in the two lanes where the varieties are adjacent.

Figure 2. Alcohol-soluble glutenin electrophoretic profiles from individual grains of a single wheat variety. Variation in banding profiles from lane to lane shows evidence of residual variability, or “biotypes,” within the seed lot.

Figure 3. Isozyme profiles of the enzyme malate dehydrogenase separated by starch gel electrophoresis for five individuals each of maize inbred lines and hybrids.

Figure 4. Protein profile of an inbred line of maize following a two-step, two-dimensional electrophoretic separation of embryo proteins labeled with [35S]methioine. The proteins range from approximately 20 to 100 kDa (bottom to top) and isoelectric points are from 5.0 to 7.0 (left to right). Proteins were separated first according to their isoelectric points. They were then transferred to a second gel and separated according to the molecular weights of their subunits in a dimension at right angles to the first dimension.

93

J. S. C . SMITH AND 0. S. SMITH

94

0.58

-

0.5

-

0.4

-

0.3

-

0.2

-

0.1

5

10

15

20

25

30

35

40

45

50

55

MINUTES

Figure 5. Chromatogram of a wheat variety showing its gliadin profile following separation by reversed-phase high-performance liquid chromatography.

preferred choice of tissue for isoenzyme analysis or for the incorporation of radiolabeled compounds for detection purposes. Preliminary studies sampling tissue in order to establish a time or development “window” during which the protein profile remains stable and thus is potentially reflective of genotype must be undertaken. Some examples can be found in isoenzyme profiling for seedlings of maize (Zea mays) (Goodman and Stuber, 1980), seedlings of barley (Hordeurn vufgare)and ryegrass (Lofiurn spp.) (Nielsen, 1985), and leaf tissue of cocoa (Theobroma cacao) (Atkinson et a f . , 1986), apple (Mafus spp.) (Weeden and Lamb, 1985), and cassava (Manihot esculenta) (Ramirez et a f . , 1987). Critical analyses of the stability of protein profiles in relation to tissue sampling and extraction are especially vital when the data are analyzed in both quantitative and qualitative (presence or absence) terms. Preliminary quantitative investigations to test the validity of two-dimensional protein profiles are necessary. For example, in barley (Gorg et a f . ,1988) and maize

60

FINGERPRINTING CROP VARIETIES

95

inbred lines (Higginbotham and Smith, 1989; Higginbotham ef al., 1989a,b), studies were carried out to measure the genetic components of variation. Quantitative variation for protein in wheat lead Burbidge et al. (1986) to conclude that it would be preferential to consider only the qualitative aspects of wheat protein profiles. Similarly, Cross and Adams (1983a) concluded that quantitative variation for a protein in globulin profiles of maize could not be relied upon because this protein varied in its resultant expression among cultivars and was also known to decline in amount during seed maturation (Cross and Adams, 1983b).

3. Stability of Biochemical Profiles in the Face of Environmental Effects Studies that demonstrate the stability of protein profiles can be assigned to two categories: first, a test of profile stability in the face of environmental change, and second, the demonstration of genetic control of profiles. a. Variation in Protein Profiles of Varieties Grown in Different Locations and/or Years Huebner and Bietz (1988) studied the stability of protein profiles for two hard red spring wheat varieties grown in seven locations in the United States and three Australian cultivars grown in soils of three different sulfur levels. While there were baking-quality differences due to location affects, no electrophoretic differences for gliadin profiles were found among different lots of flour for each variety. Likewise, Lookhart and Finney (1984) found that frosting affected baking quality of wheat but it did not change the gliadin profiles. Similarly, Clements (1987) found no environmental effects on the gliadin electrophoretic profile of soft wheat. The electrophoretic similarities of profiles from immature and mature seed provided additional evidence of the intransigence of seed protein profiles to factors other than genetic change (Clements, 1987). However, Huebner et al. (1990) found small quantitative differences for gliadins and glutelins according to maturity and spike position in respect to chromatographic profiles; qualitatively, profiles were unaffected by these factors. Lookhart and Pomeranz (1985a) detected slightly different gliadin electrophoretic profiles for two wheat cultivars grown in soils that were severely deficient in sulfur; no differences could be detected due to various levels of soil nitrogen. Huebner and Bietz (1988) were unable to find differences in gliadin electrophoretic profiles for seedlots grown in soils with different sulfur levels. Quantitative variation was detected by reversed-phase high-performance liquid chromatography (RP-HPLC) but qualitatively the chromatographic profiles were unchanged. A small degree

96

J. S. C . SMITH AND 0. S. SMITH

of quantitative variation was also observed for varieties grown in soils with

no known fertilizer deficiencies at different locations and in different years (Huebner and Bietz, 1988). Subtle quantitative differences in peak areas were only revealed by a computed difference plot and did not preclude cultivar identification by RP-HPLC. Similar low levels of quantitative, but not qualitative, variation for chromatographic profiles have been found for seedlots produced in different environments (Marchylo and Kruger, 1984; J. S. C. Smith and Smith, 1986, 1988b; Marchylo et al., 1990). In maize, for example, variation due to environmental factors did not qualitatively affect any peaks that contributed more than 2% to the total varietal profile; 93% of variation for peak area was between genotypes (J. .S. C. Smith and Smith, 1986). For oats (Avena sativa), no qualitative differences due to environmental effects could be detected for either electrophoretic or chromatographic profiles of avenin proteins (Lookhart, 1985b; Hansen et al., 1988). In rice (Oryza sativa), Sarkar and Bose (1984) found environmental effects to be insignificant for electrophoretic profiles of globulin proteins for eight varieties grown at two locations during 2 yr. A litany of other reports of the invariance of protein profiles with respect to environmental or storage conditions has been made (Lee and Ronalds, 1967; Wrigley, 1970; Zillman and Bushuk, 1979a; Cooke, 1984). The most discriminatory technique available for the separation of proteins is two-dimensional electrophoresis. This technique provides both quantitative and qualitative data. It allows the most exhaustive test of the stability of protein profiles for seedlots grown in different environments. Gorg et al. (1988) showed sufficient repeatability of profiles for 13 varieties of barley from two cultivations to provide discrimination. Similarly, Higginbotham et al. (1989a,b, 1991a,b) identified proteins from maize embryos that were indicative of genetic differences. b. Profile Stability Shown through the Demonstration of Genetic Control of the Profile The second means of establishing clear independence of protein profiles from environmental effects is to demonstrate genetic control of proteins. It has been possible to characterize genetically isoenzyme variants for many enzyme systems in numerous species of cultivated plants (Tanksley and Orton, 1983; Nielsen, 1985). Only in esterases has the variability in banding profiles been found to be so dependent upon developmental control that, for some loci, and for some crop species, the genetic control cannot readily be established (Nielsen, 1985; Rebordinos and Perez de la Vega, 1990). The genetic control of isoenzymes, in particular, has been thoroughly researched in many species. In maize, for example, all of the current practical applications are dependent upon basic genetic studies

FINGERPRINTING CROP VARIETIES

97

carried out by Drs. Goodman and Stuber and their colleagues in work that spanned a period of approximately 15 years. During that time, thousands of crosses involving F1 and F2 crosses and back-crosses were made and tens of thousands of individual progeny were assayed in the laboratory in order to determine the details of the genetics of these markers [see Stuber et al. (1988) and numerous references cited therein]. Even in crops that are not so amenable to genetic analysis, for example, apple, it is possible to compare isoenzymatic profiles of progeny trees with those of their parents and to partially verify genetic control (Weeden and Lamb, 1985). The usage of isozymes to routinely test purity in tens of thousands of seedlots representing more than 100 different maize hybrids, (Smith and Weissinger, 1984; Smith and Wych, 1986; Brink et al., 1989) provides additional evidence of the freedom of isozyme expressions from environmental effects. Seedlots have been produced in the United States, Canada, South America, and Europe and in no instance known to us has environment been found to qualitatively affect the isozyme profiles. General protein detection either by stains or spectroscopic absorbence methods often reveal numerous bands or peaks per individual plant or variety-bands that may exhibit close or overlapping gel migration or column retention times. This level of complexity in banding patterns has prevented the establishment of genetic control for albumin and globulin proteins (Cross and Adams, 1983a), although these profiles appear to be strongly variety dependent (Koranyi, 1989a,b). Profiles of alcohol-soluble seed proteins are frequently utilized for varietal description and the genetic control of these proteins is well established, at least for some species of cultivated plants (Shewry et al., 1983; Soave and Salamini, 1983; Wilson, 1986; Wilson et al., 1989; Graybosch and Morris, 1990; Gupta and Shepherd, 1990). There is abundant evidence to show that protein profiles can be obtained for all crop species of major importance and that these profiles are independent of environmental or storage conditions. They are thus reflective of genotype. In some cases, quantitative variation can occur due to environmental effects, but this contribution to variation is small and can be taken into account when considering intervarietal comparisons. It is imperative that data having a quantitative component, for example, those derived from RP-HPLC or two-dimensional electrophoresis, be carefully examined in order to be able to evaluate their degree of significance in discriminating between qualitatively similar profiles. Environmental effects have only been found to impinge upon varietal identification in conditions of extreme sulfur deficiency; such circumstances are extremely rare and can be tested for. Protein profiles of cultivated varieties, therefore, satisfy one prerequisite of providing a fingerprint; they are stable descriptors of genotype.

98

J. S. C. SMITH AND 0. S. SMITH

4. Data Recording and Profile Extraction Optimal scoring accuracy of complex profiles can be provided by automated gel scanning that can give both qualitative and quantitative data. Large databases can be built only when variances in protein migrations or column retention times are known (Zillman and Bushuk, 1979b; Scanlon et al., 1989a,b. Rigorous use of controls is essential to provide such ability (Sapirstein and Bushuk, 1985a,b,c, 1986; Scanlon et al., 1989a,b). Critical controls are also required for meaningful analysis of two-dimensional protein data (Anderson et al., 1985; Autran and Abbal, 1988; Higginbotham et al., 1991a). Isozymic data can be more easily recorded with an assured certainty of correct allele identification because the banding profiles for individual enzymes are relatively simple and there are few possible variants to be scored for each locus. Once data are recorded, several procedures can be used to identify profile matches or to generate distance measures between varieties (Zillman and Bushuk, 1979b; Lookhart et al., 1983; Autran and Abbal, 1988; Scanlon et al., 1989a,b; Sapirstein and Bushuk, 1985a,b,c, 1986). Construction of a two-dimensional gel database can be especially challenging despite the assistance of extremely sophisticated software (Higginbotham et al., 1991a). Associations among varieties can also be shown by multivariate analyses.

C. DNADATA 1. Introduction Additional variation can be revealed through analysis of DNA. First, most proteins that have been used in isozymic or other electrophoretic or chromatographic profiling have been coded by genes that are unequally dispersed throughout the genome. This prevents them from providing thorough genomic sampling. Second, mutations occur that are not translated into conformational, net electrical charge, size, or hydrophobicity changes in the protein. Variation revealed by electrophoresis, chromatography, or immunology is dependent upon one or more of these factors. However, as much as 75% of nucleotide substitutions may have no effect on these protein characteristics and will thus remain undetected by protein assay techniques (Marshall and Brown, 1975). DNA data as revealed by restriction fragment length polymorphisms (RFLPs) are becoming routinely used and widely accepted as providing taxonomic, genetic, and phylogenetic information (Doebley and Wendel, 1989; Kirby, 1990; Weir, 1990). The ability to cleave DNA at specific nucleotide base recognition sequences with restriction endonucleases, coupled with methods to sep-

FINGERPRINTING CROP VARIETIES

99

arate, label, hybridize complementary DNA sequences, and reveal the relative position or molecular weight of DNA fragments following electrophoresis have together made possible the direct use of variation in DNA sequence as a descriptor. Polymorphisms for lengths of restricted DNA fragments (Fig. 6 ) are revealed by taking advantage of the sequence

Figure 6 . RFLP profiles of 41 inbred lines of maize arrayed in two ranks following hybridization with a single radiolabeled cloned fragment of maize DNA. In each of the two ranks of lanes there are three lanes that have been loaded with marker DNA comprising eight fragments of known molecular weight. In addition, each lane containing maize genomic DNA has also been spiked with three DNA fragments of known molecular weight. The standard “spiked” fragments can be seen to correspond in their migration to three bands in the marker lanes.

100

J. S. C. SMITH AND 0. S. SMITH

affinity in complementary base pairing to a segment of “probe” DNA. The procedure is, therefore, appropriately named restriction fragment length polymorphism technology. Fundamentally, the same technology is applied in humans, other animals, microorganisms, and in many plant species (Gill et al., 1985; Jeffreys and Morton, 1987; Loftus et al., 1988; Rogstad et al., 1988; Ryskov et al., 1988; Cooksey and Graham, 1989; Hillel et al., 1989; Walmsley et al., 1989). Basic techniques involve DNA extraction, restriction of DNA by endonucleases, separation of restricted DNA fragments by electrophoresis, transfer of separated DNA to a membrane, and subsequent repeated cycles of probing the membrane with radioactively or otherwise labeled DNA, exposure of films, and washing of the membrane (see Maniatis et al., 1982; Beckmann and Soller, 1983; Helentjaris et al., 1985).

2. Stages in the Development of Protocols for DNA Fingerprinting of Cultivated Varieties a. DNA Extraction In general, protocols are developed first for the extraction of high yields (300-500 ng/pl) of high-quality (restrictable high molecular weight) DNA. Tissues can be extracted fresh in liquid nitrogen or dried using appropriate buffers to remove and protect DNA from nucleases and proteins that could deleteriously affect DNA. General protocols for all other procedures in DNA profiling are given by Maniatis et al. (1982). b. Development of a Set of Clones as Probes Second, it is necessary to have a set of DNA clones as probes of genetic variability. Clones can be made from random pieces of genomic DNA, from DNA that is a complement of cytoplasmic RNA (cDNA), or, more rarely, from isolated cloned genes. It is also possible to synthesize oligonucleotides and use these as primers to detect length variability of fragments generated through DNA amplification in thermal cycling equipment using the polymerase chain reaction (PCR) process. If probes are constructed of highly repetitive nucleotide sequences, then profiles are more likely to be complex than would be the case for patterns produced through hybridization with probes constituted of nonrepeated DNA. The library of DNA clones can be made from cDNA that represent generally unrepeated gene coding regions. Alternately, clones can be prepared from genomic DNA cut with the methylation-sensitive restriction endonuclease PstI, which avoids fragmenting DNA in the heavily methylated and repeated regions. A subsequent screen of colonies from the library can be

FINGERPRINTING CROP VARIETIES

101

made to identify, and thus avoid selecting, highly repeated sequences as potential probes (Figdore et al., 1988; Helentjaris et al., 1986; Landry and Michelmore, 1985; Apuya et al. , 1988). In many cases, probes made from DNA of one species cross-hybridize sufficiently well with genomic DNA from another species to allow their use as probes in both species under usual conditions of hybridization stringency. Thus, for example, 80% of maize probes can be used to profile sorghum (Sorghum vulgare) inbred lines (Hulbert et al., 1990; Lee el al., 1990a) and probes made from tomato (Lycopersicon esculenturn) DNA can be used to profile potato (Solanurn tuberosurn) (Bonierbale et al., 1988). Indeed, probes made from human and bacterial DNA can be used to profile genotypes from many species of animals, birds, and plants (Dallas, 1988; Ryskov et al. , 1988). c. Screening the Probes for their Ability to Provide Fingerprints A third step is a preliminary assay of polymorphism. Usually, it is prudent to make an initial screening of clones with electrophoresed DNA restricted by several (three to six) individual restriction endonucleases extracted from relatively (5-10) few genotypes. Genotypes should encompass a range of genetic diversity sufficient to allow a reasonable chance that polymorphism can be revealed. Previous analyses of genetic diversity, knowledge of the breeding system, and an awareness of the breadth of genetic diversity encompassed by cultivated varieties can indicate the genotypic components of the initial screen. For example, screens of clones in maize need utilize only diverse elite inbred lines. In contrast, for sorghum, soybean (Glycine rnax), and tomato, genotypes of elite lines, exotic cultivars, and different species should be included. It is useful to compare banding profiles among genotypes for each probe directed against different (from three to five) individually digested genomic DNAs. This is because not all probe/restriction digest combinations perform equally well in terms of resolving clear polymorphisms. Criteria for probe selection include the ability to reveal band profiles without lane background and for which most variants can be determined unambiguously because they have quite different molecular weights. d. Number of Probes and Selection of Restriction Digests The repertoire of probes that will constitute the eventual battery of fingerprinting probes can be determined from consideration of their individual ability to reveal variation, and in relation to the purpose(s) for which the genotypic profile data are to be collected. Results obtained from the initial screen of materials will provide an indication of each probe’s ability to reveal variation among the cultivated varieties for the species of interest. For example, our screening of some 450 probes against three

102

J. S. C . SMITH AND 0. S. SMITH

individual restriction digests using BamHI, EcoRI, and Hind111 for 13 elite Corn Belt lines of maize provided a list of approximately 200 probes and some 400 probe/restriction enzyme combinations that gave clear and easily readable banding profiles, revealing polymorphism among all 13 elite lines. Of the 200 selected probes, 100 were then used to profile these same lines using an additional three restriction digests (BglII, EcoRV, and KpnI). From these data, the list of probelenzyme combinations was further refined, selecting probe/enzyme combinations that (1) gave consistently clear profiles with obvious pattern differences between genotypes; (2) that revealed the highest discriminatory power; and (3) that included probes mapped evenly to each chromosome arm in order to provide complete coverage and sampling of the genome. In addition, these data showed that most genetic diversity was revealed by a single combination of probe/restriction enzyme. Furthermore, for the probe/restriction digests that provided clear and polymorphic profiles, it was otherwise immaterial which restriction enzyme was used. It was decided to use a single restriction enzyme digest of genomic DNA for each probe. The combination was selected on the basis of giving the most polymorphism and providing clearly readable profiles. The effort that would have been expended in obtaining a profile for a second restriction enzyme for a given probe was found to be much more gainfully spent in using an additional probe (J. S. C . Smith et al., 1991e). Additional data for 37 maize inbreds profiled by 157 probes provided an opportunity to investigate the number of probes that were necessary for precise calculations of genetic distance between lines (0.S . Smith et al., 1991). Confidence intervals around genetic distance, computed using jackknifing techniques, narrowed appreciably up to the use of 60 probes. Beyond 100 probes, there was a much diminished return in the narrowing of the confidence interval. Thus, from 80 to 100 single-copy probes seems a reasonable number to provide accurate estimates of genetic distance in maize. In contrast, fewer probes can provide unique identification of maize lines. For 150 United States public lines, ~ 4 probes 6 can provide 100% unique identification (J. S. C. Smith etal., 1991~).With a database of band frequencies, it should then be possible to select from 5 to 10 probes that collectively give very low probabilities that a specific line profile could be independently derived. This potential is feasible because of the large number of variants revealed by RFLPs in maize. For other crop species, the absolute number of probes requisite for each crop will depend upon the ability of RFLPs to reveal variation in that crop. The numbers of probes could be reduced by the use of clones that simultaneously sample multiple portions of the genome and/or that reveal bands that are of a low frequency of occurrence.

FINGERPRINTING CROP VARIETIES

103

e. The Necessity of Defining Robust Methods of Acquiring RFLP Profiles So That Profile Comparisons Can Be Made from within a Large and Expanding Database Experience gained in the compilation and pratical usage of the maize isozyme database vividly demonstrates the absolute necessity of being able to repeatedly distinguish and record profiles of individual varieties. This is made possible by including on each gel entries of known banding profiles and having a standardized scoring system. In comparison to isozyme data, the challenges faced in properly and repeatedly ascribing RFLP scores are vastly greater due to the larger number of individual probe profiles that are created and the large number of variants revealed by each probe. Therefore, it is essential to have on each autoradiogram profiles of known molecular weights so that variants can be scored in terms of an inherent property-molecular weight. Migration distance of the variant is much less reliably informative because it is affected by additional factors, such as agarose gel concentration, buffer strength, power applied, and running time. Ideally, each lane should also be spiked with bands of known molecular weights so that lane to lane variation can be accounted for (J. S . C. Smith et al., 1991g). f. The Necessity of Obtaining a Genetic Map of Probes Map data provide information on the degree of completeness of genomic coverage that is afforded by the genetic markers. If major portions of the genome were not represented, then it would be necessary to continue to add clones in order to sample those regions. Also, it is necessary to know the amount of genetic linkage between the markers. If two or more markers are closely linked, then it would be inappropriate to consider them each to have individually equal weight in providing discrimination or distance measurement. However, most calculations of probabilities of identity have assumed independence of the markers (Cohen, 1990).

3. Analysis of RFLP FingerprintData a. Reading and Scoring Autoradiograms Traditionally, RFLP fingerprinting autoradiograms (Fig. 6) have been visually scored for presence or absence of bands in each lane (Lee et al., 1989; Melchinger et al., 1990; 0. S . Smith et al., 1990; 0. S. Smith and Smith, 1991). Because the source of DNA in these experiments has been from a bulked sample of several plants, only the presence or absence of bands rather than band frequencies could be scored. Autoradiograms can either be read by an audible wand signaling to stereoscopically positioned microphones or more directly by image analysis via digital camera, laser,

104

J. S. C . SMITH AND 0.S. SMITH

or other gel reader (Gill and Werrett, 1990). The molecular weight range within which differences cannot be shown to be different must be determined through experimentation. These data provide “windows” of molecular weights within which variants can be declared or “binned” for scoring and databasing. Genotypic profiles can be compared by multivariate techniques or by using software similar to that reported for protein data (Lookhart et al., 1983; Sapirstein and Bushuk, 1985b, 1986; Autran and Abbal, 1988; Scanlon et al., 1989a,b.

b. Types of Probes used to Generate Data in Plants DNA clones used as probes in plants have generally been single or very low copy versus the multilocus probes (Jeffreys, 1987). Probes have been mapped to chromosome locations using segregants of F2 populations or recombinant inbred lines. However, all of the bands found for a probe enzyme combination in the fingerprinting studies have not been mapped in an F2 or recombinant inbred. The number of probe enzyme combinations used to develop fingerprints in maize has ranged from 40 to over 100, with the total number of unique bands ranging from about 250 to greater than 1000. Some of the earlier studies restricted their analysis to using a single restriction enzyme digest per probe. When multiple restriction enzymes are used, generally the most polymorphic restriction enzyme probe combination was chosen to provide profile data. Our analyses have been based on all bands that were revealed. Polymorphisms are a result of sequence differences and each restriction enzyme can reveal variation at a different restriction site at or around the region to which the probe hybridizes. c. The Choice and Use of Distance Measures Based upon RFLP Data There have been two major objectives of the fingerprinting studies: first, to uniquely identify genotypes based on their banding patterns, and second, to develop measures of similarity between genotypes based on their RFLP profiles. Included in the second objective is often a desire to determine if measures of RFLP similarity are correlated with other measures of diversity, i.e., pedigree, F1, and heterosis. Developing useful measures of similarity among genotypes has been more of a challenge than has been the ability to achieve the unique identification of different genotypes. Potentially, this aspect will be one of the most important uses of DNA fingerprints. Knowledge of genetic relationships among individuals has several potential uses, for questions dealing with evolutionary biology (Nei and Li, 1979; Kaplan, 1983; Grafen, 1985; Lynch, 1988; Nei and Miller, 1990) and in the choice of potential mates and preferred segregants in controlled breeding programs (Lee

FINGERPRINTING CROP VARIETIES

105

et al., 1980; 0. S. Smith et al., 1990). To date, several methods of measuring genetic distance have been proposed and used [see Jacquard (1974) for a summary of these]. The several available measures of similarity imply that the definition of a “good” measure of similarity is somewhat arbitrary (Chatfield and Collins, 1980; Dillon and Goldstein, 1984). Modified Roger’s distance (MRD) (Rogers, 1972) has been used to calculate genetic distances based on allozymic data. The algorithm of Nei and Li (1979) provides a distance measure that does not depend upon knowledge of the genetic control of each variant. This algorithm is, therefore, appropriate for RFLP data where many variants are frequently detected in routine profiling and knowledge of their allelic relationships may be incomplete. The interpretation of these measures of genetic similarity in terms of relatedness, however, is not straightforward, as has been shown by Lynch (1988). Genetic relatedness is defined as the fraction of autosomal alleles that are identical by descent (IBD) and refers to those situations in which two alleles are copies of an allele inherited through common pedigree. “Identity in state” (11s) refers to those alleles that are identical in function; i.e., the phenotype and, in the case of DNA fingerprints, the DNA fragment sizes are identical at the level of discrimination of these techniques. IIS, however, implies nothing about coancestry. Therefore, similarities based on the numbers of bands that are in common between genotypes will not always accurately reflect similarities due to relationship. With these caveats in mind, coefficients of similarity based on DNA fingerprints have been analyzed to determine if these measures of similarity are correlated with (1) distance based on known pedigrees, (2) hybrid grain yield, or (3) the amount of observed heterosis. Again, all of these measures of diversity are subject to errors and are not a perfect measure of genetic similarity or genetic relatedness. They do, however, provide further multigenically based estimates of relatedness against which evidence from laboratoryderived data can be compared.

d. The Revelation of Associations among Genotypes Using Multivariate Analysis Techniques Once similarity coefficients have been calculated, multivariate techniques can be used to reduce what is a multidimensional space, i.e., a hyperspace where the number of dimensions is equal to n - 1 and n is the number of genotypes of interest, to a much smaller dimensional space, which then allows for easier interpretation (Figs. 7 , 8 , and 9). The object is to allocate the genotypes into groups of similar genotypes. In studies with cultivated varieties, the objective is not to determine an ancestral relationship, as would be the case with studies of evolution, but to ascertain if

J. S. C . SMITH AND 0. S. SMITH

106

P 3184

P 3358

P 3183

r

MFA6708

FG4522

B73iM017

B73iLH51

PM8201

G HH2500

DKXL72AA

I

1.6

I

I

I

I

1.4

1.2

1

0.8

I 0.6

I 0.4

I 0.2

I

0

DISTANCE

Figure 7. Associations among 12 maize hybrids as revealed by cluster analysis of isozymic data. Two hybrids (B73/LH51 and PM8201) had indistinguishable isozyme profiles.

these groupings are reflective of similarity based on known pedigree relationships and/or if they provide information that is useful in applied breeding programs. Two multivariate data reduction techniques in common use are principal coordinate analysis and cluster analysis. Both techniques have advantages and disadvantages. The principal coordinate analysis developed by Gower (1972) is a method for reconstruction of the point coordinates assuming that the dissimilarities are Euclidean distances. The coordinates for the genotypes are then plotted in two- or three-dimensional space from which associations among genotypes can be discerned. Cluster analysis is another

G4522

PRlN3 G967

7

1.92

6708

9

3358 0

P

E731LH51

0.75

-

B731M 017

I

------f&. XL72AA

Figure 8. Associations among 12 maize hybrids on the basis of their relative occupancies of the first three principal coordinates that individually encompass the most variation of the multidimensional space. Physical closeness among hybrids in this figure denotes relative similarity in isozymic profiles.

so 26 20 1s 10

5

PC20

-s

-

I

-3s

-33

-31

-29

-27

-2s

-23 - 2 1

-19

-17

I -1s

I -1s

I -11

PC1 Figure 9. Associations among inbred lines of maize according to their occupancy of two-dimensional space following principal coordinate analysis and according to cluster analysis. Groupings of lines encompassed by solid and dotted lines show levels at which inbred lines were associated into their various clusters. The cluster analysis of these lines is not shown separately.

3.1 6

108

J. S. C. SMITH AND 0. S. SMITH

technique that has been widely used to group objects. There are, however, many clustering algorithms [see Dillion and Goldstein (1984) for a discussion of clustering techniques]. The algorithms that have been used with DNA fingerprint data are Ward’s error sum of squares method (Ward, 1963) and the single and average linkage agglomerative methods. The utility of molecular marker data to provide “fingerprint” data that could be meaningful and useful to breeders and others is tested by comparing associations or distances among genotypes as revealed by marker data with that shown by pedigree or field performance data (Lee et al., 1989; 0. S. Smith et al., 1990, 1991; 0. S. Smith and Smith, 1991).

III. DISCRIMINATIONAL ABILITY OF FINGERPRINTING TECHNIQUES A. FACTORS AFFECTING DISCRIMINATIONAL ABILITY Factors that are related to the breadth of genetic diversity impinge upon the ability of any technique to provide unique identification. It is important to recall that plants, in contrast to most animals, make it possible, by virtue of their breeding systems, to create lines, varieties, or hybrids that can be closely related, very similar morphologically, and even genetically identical.

1. The Effect of the Breadth of the Germplasm Base Used in Breeding New Varieties upon Unique Cultivar Identification If the breeding base is narrow, then there will be, on average, fewer genetic differences segregating within breeding populations and, therefore, a reduced genetic distance between the resultant progeny and between those progeny and their parents. In the absence of new genetic diversity, genetic differences between newly bred cultivars can become fewer and individual cultivar identification, by any technique, then becomes more difficult. The choice of breeding methods also can profoundly affect diversity. Multiple recurrent back-crossing is a classic methodology to produce a line that differs for one or a few specific traits, with no difference for the majority of the genome between progeny and recurrent parent. Transformation technology using microprojectile bombardment or electroporation techniques presents faster means to make isogenic lines. However,

FINGERPRINTING CROP VARIETIES

109

even with classical breeding techniques, a few back-cross generations, on average, shift the genotype rapidly closer to the recurrent parent.

2. The Effect of the Breadth of the Germplasm Base Used in Seed Production and Agriculture The degree to which identical or very genetically similar parents are used in commercial seed production for F1 hybrid crops or varieties also determines the degree of genetic difference between cultivars. For example, in the United States and Europe, numerous breeding organizations have produced commercial F1seed using at least one common publicly available inbred line per hybrid. In the United States, the public lines B73, B37, B14, A632, Oh43, and Mo17 have been widely used. In France the line F2 is very extant in agriculture. In the United States, each foundation seed company supplies customers from common sources of inbred lines. These lines can be used as parents in many single-cross maize hybrids. There is the possibility that numerous maize hybrids sold by different organizations could be identical genetically and, therefore, be impossible to differentiate by even the most discriminative fingerprinting techniques.

3. The Effect of Descriptor Class and Separation Method on the Ability to Reveal Genetic Diversity Not all protein fractions have an equal prevalence of diverse proteins. Each cultivated species must be individually investigated to find the class or classes of protein that reveal the most genetic variation. Isozymes generally show variability for several enzyme systems, and when a number of systems are used, the level of discrimination between cultivars can be appreciable. Alcohol-soluble seed storage proteins generally also reveal significant diversity. Water- and salt-soluble proteins can also reveal variability among cultivars, but generally these profiles are qualitatively and quantitatively very complex, probably affected by the stage of physiological development, and usually are not well understood genetically. For these reasons, one-dimensional separations of such complex protein mixtures, therefore, are not usually relied upon as descriptors. However, with the advent of improved two-dimensional separation methods, coupled with an enhanced ability to identify and quantify individual spots and complex spot patterns through sophisticated software packages, it is possible to database multiple images. Comparisons can then be made using profiles with hundreds of spots per cultivar as descriptors (Dunbar et a!, , 1985; Gorg et al. ,

110

J. S. C. SMITH AND 0. S. SMITH

1988; Higginbotham et al., 1989a,b, 1991a,b. Generally, a specific combination of protein class and electrophoretic or chromatographic technique must be sought by screening several combinations of extraction and separation protocols in order to find methods that will reveal the maximum degree of variability for a set of cultivars. Oftentimes, different electrophoretic and chromatographic procedures provide complimentary identification. At the DNA level, some cultivated species (such as maize) that are immensely rich in genetic variation, as revealed by biochemical means, are also highly polymorphic when surveyed for RFLPs; some species, at least with current methodologies, for example, wheat, do not show much variation for RFLPs. More complex two-dimensional electrophoretic separations of protein may then be more appropriate to consider (Anderson et al., 1985; Dunbar et af., 1985).

B. DISCRIMINATIONAL ABILITYOF PROTEIN AND DNA DESCRIPTORS FOR SELECTED OR CULTIVATED SPECIES A high degree of unique identification among cultivated varieties and inbred lines of many species is possible by means of one or more profiling methodologies. Development of fingerprinting technology is at various stages for each crop, depending upon several factors. These include the amount of genetic variation that is present within the cultivated germ plasm, the state of technology as regards its ability to reveal variation, and the level of need expressed by breeders, seed producers, seed certification and registration agencies, and other interested parties to have in place precise and unique profiles of cultivated varieties. There is a synergistic relationship between the ability and the desire to provide fingerprints. Even when fingerprinting ability has been achieved, for example, as in maize, the work is incomplete. There remain improvements in the efficiency of profiling to be sought, yet further detailed profiles to be obtained for the benefit of plant breeders and geneticists, and standards of profile derivation and usage to be evolved and accepted in the scientific, legal, and administrational communities. The current status of profiling is, therefore, always somewhat transient and rests upon a large body of published research that is continuously built upon. In this review, we are able only to present a framework of appropriate references for each major cultivated species (Tables I and 11). For additional details, the reader is encouraged to pursue the citations given and to keep abreast of new information as it appears.

Table I Summary of Laboratory-Based Cdtivar Identification Ability for Cultivated Species of Major Economic Importance ~

~

Source material

Cultivated species

Range of genotypes

Wheat (Trificurnsupp.)

88 most widely grown U.S. cultivars in 1979

Gliadins

PAGE

88 varieties licensed in Canada

Gliadins

PAGE

68 Canadian and 2 U.S. spring wheats 51 varieties on U.K. national list 27 spring wheats

Gliadins and glutenins

Technique"

Profiling ability

Reference Jones et al. (1982)

Gradient SDS-PAGE

85% unique; cultivars with similar profiles closely related by pedigree 93% unique; cultivars with similar profiles related by pedigree 84% unique profiles

Gliadins

SGE

100% unique

Gliadins

RP-HPLC

93% unique

10 most important U.K. cultivars 15 European cultivars

Gliadins

100% unique

Gliadins

Rapid RP-HPLC RP-HPLC

Clydesdale and Draper (1982) Marchylo et al. (1988) Bietz and Cobb (1985) Burnouf et al. (1983)

Many U.S. varieties

Gliadins

RP-HPLC

14 U.K. varieties

Gliadins and glutenins

Two-dimensional electrophoresis

All but three closely related varieties unique Most unique; only some with similar pedigrees indistinguishable 100% unique

Zillman and Bushuk (1979a; b)

Marchylo et al. (1989)

Bietz et al. (1984)

Anderson et al. (1985) (continues)

Table I (Continued) Cultivated species Maize (inbreds) (Zea mays)

Range of genotypes 406 publicly available inbred lines, U.S., Canada, Europe 30 widely used U.S. lines in 1975 113 Canadian lines

62 widely used U.S. public lines, 1960-1985 14 Lancaster Sure Crop lines

Source material

Technique"

Protiling ability

Reference

Coleoptile isozymes, 23 loci

SGE

73% unique

Stuber and Goodman (1983)

Coleoptile isozymes

SGE

93% unique

Goodman and Stuber (1980)

Coleoptile isozymes

SGE

80% unique

Coleoptile isozymes, 34 loci Coleoptile isozymes and zeins Coleoptile isozymes and zeins Zeins

SGE

94% unique

Cardy and Kannenberg (1982) Smith ef al. (1987)

SGE and RP-HPLC

Unique unless related by back-crossing Unique unless related by back-crossing 92% unique

Wilson (1985)

100% unique

Wall et al. (1984)

Only split into 2 groups 100% unique

Esen el al. (1989)

J. S. C. Smith et al.

100% unique electrophoresis

(1991~) Higginbottham ef al. (1991a; b)

17 Iowa Stiff Stalk Synthetic lines 25 U.S. public inbreds 9 public lines

Zeins

15 public lines

Zeins

150 public lines

DNA

37 inbred lines

Embryo protein

SGE and RP-HPLC Isoelectric focusing Two-dimensional electrophoresis Immunology RFLPs, 46 probes Two-dimensional electrophoresis

Smith and Smith (1987) Smith and Smith (1988a)

Maize (hybrids) (Zea mays)

SGE

94% unique

SGE

95% contrasting profiles, some very similar 60% unique

47 French hybrids

Coleoptile isozymes, 22 loci Coleoptile isozymes, 21 loci Coleoptile isozymes and zeins Coleoptile isozymes and zeins DNA

106 U.S. hybrids

DNA

RFLPs, 46 probes

174 cultivars

Coleoptile isozymes, 11 enzymes DNA

SGE

155 Canadian hybrids 111 U.S. hybrids

128 U.S. hybrids

61 French hybrids

+ e W

Soybean (Glycine max)

58 lines

Barley (Hordeum vulgare)

60 lines 66 cultivars

55 U.S. cultivars 88 U.K. varieties

DNA Coleoptile isozymes and hordeins Hordeins Hordeins

SGE and RP-HPLC SGE and RP-HPLC RFLPs, 80 probes

100% different but 14 widely used hybrids similar 94% unique, 3 hybrids very similar Most unique, but some with very similar profiles 77% unique

Cardy and Kannenberg (1982) Smith (1984)

Smith (1988)

Smith (1989)

J. S. C. Smith and Smith (1991) J. S. C. Smith et 01. (1991b)

Cardy and Beavorsdorf (1984) Keim et al. (1989)

RFLPs, 17 probes RFLPs Electro phoresis

88% unique

Most unique 94% unique

Cregan et al. (1990) Nielsen and Johansen (1986)

SDS-PAGE SDS-PAGE and urea acid page

44% unique Fell into 32 groups

Heisel et al. (1986) Shewry et al. (1978)

(continues)

Table I (Continued) ~~

~~

Cultivated species

Range of genotypes

~

~

Source material

Technique"

Profiling ability

12 Canadian cultivars 9 cultivars 13 cultivars

Hordeins

RP-HPLC

100% unique

Hordeins Leaf proteins

100% unique 100% unique

40 cultivars

Hordeins

DNA probes Two-dimensional electrophoresis PAGE

48 cultivars

DNA

23 RFLPs

Fell into 17 groups Some variation

8 cultivars

Acid PAGE

100% unique

25 cultivars

Albumins and globulins Prolamins

29 U.S. cultivars

Prolamins

Isoelectric focusing RP-HPLC

27 U.S. cultivars

Glutelins

RP-HPLC

9 cultivars

DNA

70 varieties

DNA

2 human minisatellite probes 10 RFLPS

Fell into 4 groups All unique if grain type also considered Most unique, small differences for closely related varieties 100% unique

Reference Marchylo and Kruger (1984) Bunce el al. (1988) Gorg et al. (1988) Gebre et al. (1986) Graner er al. (1990)

c e

A

Rice (Oryza sativa)

83% unique

Sarkar and Bose (1984) Guo et al. (1986) Lookhart et al. (1987) Huebner et al. (1990) Dallas (1988)

Wang and Tanksley (1989)

Oats (Avena sativa)

Sorghum (Sorghum vulgare)

23 U.S. varieties

Prolamins

PAGE and RP-HPLC

100% unique

U.S. varieties

DNA

RFLPS

Variation found

7 lines

Kafirins

RP-HPLC

120 breeding lines

Coleoptile isozymes Coleoptile isozymes Coleoptile isozymes DNA

SGE SGE

Unrelated lines had different profiles 31 genotypes, 12% lines unique 38% unique

SGE

33% unique

Schertz et al. (1990)

100 probes

100% unique

Lee et al. (1990a)

Isoelectric focusing and RP-HPLC SDS-PAGE

All unique

Sastry et al. (1986)

100% unique

RFLPs, 2 probes RFLPS, 2 probes

100% unique

Loeschcke and Stegemann (1978) Gebhardt et al. (1989) Gebhardt et al. (1989)

37 converted lines 70 hybrids 25 elite U.S. lines 4 lines, 2 hybrids 8 varieties Potato (Solanurn tuberosum)

Kafirins and glutelins

600 cultivars

Tuber proteins

38 lines

DNA

20 varieties

DNA

100% unique

Lookhart (1985a); Lookhart and Pomeranz (1985b) Hoffman et al. (1990) J. S . C. Smith and Smith (1988b) Schertz et al. (1990) Schertz et al. (1990)

“Abbreviations: PAGE, polyacrylamide gel electrophoresis; SGE, starch gel electrophoresis; RP-HPLC, reverse-phase high-performance liquid chromatography; RFLPs, restriction fragment length polymorphisms; SDS, sodium dodecyl sulfate.

Table II Additional Examples of Laboratory-Based Cultivar Identification

Genus

Species

Common name

AIlium

SPP.

Onion

Arachb Beta Brassica

SPP. vulgatis SPP.

Peanut Beet Brassicas

napus oleracea

Rape Cabbage, etc.

Pepper Pecan Coffee Orange, grapefruit, etc.

Capsicum Carya

Coflea Citrus

Cucurbira Festuca

sativus SPP.

Cucumber Fescue

Source material Protein Protein DNA DNA Protein Protein DNA Protein Protein Protein Protein Protein DNA Protein Protein DNA DNA Protein and DNA DNA Protein Protein

Reference Nakamura and Tahara (1977) Hadacova et al. (1981) Havey (1990) Halward er al. (1990) Nagamine et al. (1989) Coulthart and Denford (1982) Nakamura (1977) Slocum er al. (1990) Gupta and Robbelen (1986) Arus er al. (1982) Woods and Thurman (1976) Wills el al. (1979) Wills and Wiseman (1980) Livneh ef al. (1990) Eugene and Wolfe (1982) Bade and Stegemann (1982) Durham et al. (1990) Jarrell and Roose (1990) Liou et al. (1990) Kennard et al. (1990) Hicks et al. (1982) Villamil et al. (1982) Abernethy et al. (1989)

Fragaria

manassa

Gossypium

Strawbeny

Protein

Cotton

Protein Protein DNA DNA Protein DNA DNA Protein Protein Protein Protein Protein Protein Protein DNA DNA Protein Protein DNA Protein Protein

Helianthus Humulus Juglans Lactuca

lupulus SPP. sativa

Sunflower HOP Walnut Lettuce

Lens Lolium

culinaris SPP.

Lentil Ryegrass

Malus

SPP.

Apple

Medicago

SPP.

Alfalfa, lucerne

Olea

europaea

Olive

annuus

Bnnghurst et al. (1981) Drawert et al. (1974) Kapse and Nerkar (1985) Rao et al. (1990) Khaler and Lay (1985) Kenny et al. (1990) Fjellstrom and Parfitt (1990) Kesseli and Michelmore (1986) Landry et al. (1987) Harvey and Muehlbauer (1989) Hayward and McAdam (1977) Nakamura (1979) Arcioni et al. (1980) Payne et al. (1980) Ostergaard and Nielsen (1981) Gilliland et al. (1982) Quaite and C a d i n (1985) Nybom and Schaal (1990) Nybom et ul. (1990) Bingham and Yeh (1971) Quiros (1980, 1981) Kidwell and Osborn (1990) Gardiner and Forde (1988) Pontikis et al. (1980) (continues)

Table II (Continued) Genus

Species

Persea Phaseolus

americana vulgaris

Avocado Field bean

Pisum Poa

sativum pratensus

Pea Bluegrass

Prunus

Common name

Pyrm

Peach, almond, Cherry Pear

Rosa Rubus

Rose Blackberry, raspberry

Saccharum Secale

SPP. cereale

Sugarcane Rye

Vicia Vigna Vitis

faba

Broad (field) bean Cowpea Grape

SPP. vinifera

Source material Protein Protein DNA Protein Protein Protein Protein Protein DNA Protein Protein DNA DNA DNA DNA Protein Protein Protein Protein Protein Protein Protein DNA

Reference Torres and Bergh (1980) Hussain ef al. (1986) Nodari et al. (1990) Cooke (1984) Wilkinson and Beard (1972) Wehner ef al. (1976) Wu et al. (1984) Carter and Brock (1980) Chapparo et al. (1990) Kuhns and Fretz (1978a,b) Santamour and Demuth (1980) Rajapaske et al. (1990) Nybom et al. (1989) Moore (1990) Nybom and Schaal (1990) Glaszmann er al. (1989) Steiner et al. (1984) Ramirez and Pisabarro (1985) Cooke (1984) Vaillancourt and Weeden (1990) Drawert and Gorg (1974) Wolfe (1977) Gogorcena et al. (1990)

FINGERPRINTING CROP VARIETIES

119

IV. USAGE OF FINGERPRINTS A. THE NATURE OF PLANT BREEDING AND THE NEED FOR

LONG-TERM RESEARCH INVESTMENT Plant variety protection and patents are means by which intellectual property protection can be conferred upon the end product of research investment in plant breeding. These forms of protection are means by which an environment suitable for the encouragement of further investment can be provided. Plant breeding is particularly dependent upon long-term support because elite breeding programs alone can require 10 years to produce an improved variety. Investment in new technologies can take an equal or greater time span before their practical implementation can be effected. Investment in genetic resource conservation and utilization must be made over even greater time frames.

B. PLANTVARIETY PROTECTION, REGISTRATION, CERTIFICATION, AND PATENTS In order for cultivars to be nationally or internationally registered and for breeders to be granted plant breeders’ rights of protection, varieties must successfully pass inspection for the criteria of distinctness, uniformity, and stability (DUS) (Bailey, 1983). Certification is a test of the trueness to declared varietal type. Application of these tests can help promote the breeding of novel genotypes. DUS standards help ensure that the breeding process is pursued to completion and that a seed source representing the variety is available and can be multiplied. Other tests of agronomic performance will determine whether the new variety has a performance advantage over existing varieties. Traditionally, morphological data have been used to define the parameters of certification and DUS tests. However, morphological characters whose expression is affected by environment and which exhibit continuous distribution are notoriously poor taxonomic descriptors. Therefore, there is growing interest in the use of biochemical and DNA-based methods to provide more sharply defined and repeatable genotypic descriptions. A detailed discussion of the utility of biochemical data in providing descriptors for the granting of plant breeders’ rights has been given by Bailey (1983). In the United Kingdom, electrophoretic profiles of gliadins have been used to compare wheat varieties for the past 10 years (Parnell, 1983). Electrophoregrams of commonly grown wheat varieties are published

120

J. S. C . SMITH AND 0.S. SMITH

annually with the detailed morphological descriptions (Parnell, 1983). When identical profiles have been found, additional data have been examined to test for individuality. No wheat varieties with identical gliadin profiles have been nominated to the United Kingdom’s recommended national list (Parnell, 1983; R. J. Cooke, personal communication). However, the occurrence of identical gliadin profiles as a factor in its own right has not yet been determined to be prohibitory of such listing. As of 1983, electrophoretic characters were reported to have been used to show distinctness in at least three cases involving the granting of plant breeders’ rights (Bailey, 1983). In France, isozymic data have been utilized in maize registration and, since 1990, these data have been required as part of the application. Isozymic data alone are not sufficient to allow a declaration of novelty; linkage relationships between markers are also taken into account (M. Cambolive, personal communication). Plans are also currently underway in France to RFLP profile all maize lines submitted for registration. These explorative data will be used to assess the potential of using RFLPs for cultivar identification and thence to derive standardized procedures for their use (M. Cambolive, personal communication). This approach is both creative and cautious. Such a course of action that is based on rigorous scientific testing in the practical environment is the most opportune way to explore and incorporate useful new approaches. In Germany, an index of European potato cultivars has been produced using electrophoretic data (Stegemann and Loeschcke, 1976; Stegemann, 1979). The index contains profiles of hundreds of varieties grown throughout Europe (excluding the Soviet Union) and gives standardized methods for running and scoring gels. These data have been used to check varietal identity in court and also at border crossings using a mobile laboratory (H. Stegemann, personal communication). Isozyme data are submitted as evidence in plant variety protection and patent applications for maize in the United States (Troyer, 1986; W. Kuhn, personal communication). Historically, there has been reluctance to harness the superior taxonomic descriptive power of biochemical and DNA data for several reasons. First, laboratory techniques were novel to most people and organizations involved in description work. Second, sufficient demonstration of an ability to provide a high degree of unique discrimination among cultivars was necessary. Not only had laboratory techniques to be developed but also a representative range of cultivars had to be profiled. Third, a minimum degree of standardization of methods, including tissue extraction, gel running, and profile scoring, across laboratories in different countries was necessary before a technique could be agreed upon to be useful. Fourth, the difficulties associated with the ability of biochemical data to reveal variation within cultivars had to be resolved in the context of DUS (Bailey, 1983). The International Seed Testing Association (ISTA) has, through profiling carried out in several independent laboratories, established stan-

FINGERPRINTING CROP VARIETIES

12 1

dardized procedures for electrophoretic profiling (Draper, 1987). The Union pour la Protection des Obtenions VegCtales (UPOV), an international organization that sets standards for variety protection, and the International Seed Testing Association have disseminated information on the use of laboratory-based methods in cultivar description. French authorities at the research (INRA) and cultivar identification (GEVES) institutes continue to collect and evaluate laboratory data for their utility in cultivar identification. Without such data, it will not be possible to derive standards and procedures that can be routinely applied. Fifth, there has been justified, appropriate, and critical concern that any differences, however small, that can be detected in the laboratory by subtle biochemical or DNA profiling could be the basis for the ruling of novelty. If this were to be the case, then any degree of difference, even to the extent of a single nucleotide base change, that would be trivially cosmetic in terms of the appearance or performance of the total genotype, could be sufficient to show uniqueness. Without some agreed standards as to what degree of difference would be necessary to achieve the status of novelty, the system could be open to gross abuse (Duesing and Raeber, 1989; J. S. C. Smith et al., 1991a).

C. MINIMSJMDISTANCE 1. The Concept of Minimum Distance

Minor cosmetic changes can be found, induced, or created by backcrossing, somaclonal variation, or through transformation technology. If minor differences are given weight equal to more substantial differences brought about through genetic segregation and recombination during crossing and selection of genotypes, then there could be undue reward to those seeking to plagiarize existing varieties relative to those practicing new variety development. Such concerns are heightened when the use of biochemically or DNA-based descriptions are considered, for these methods allow even minor differences between genotypes to be detected. However, many morphological differences can also be detected that are founded upon one or a few genes. Thus, the problem of cosmetic differences vis-a-vis the existence of novelty has always existed.

2. Definition of the Boundary of Minimum Distance The level of genotypic similarity at which the boundary of novelty could be adjudged to reside is spoken of as the “minimum distance.” The

122

J. S. C . SMITH AND 0.S. SMITH

decision as to where the level of minimum distance exists cannot be made easily, for it involves input from different disciplines and perspectives. We have attempted to bring a degree of objectivity to bear on the subject, at least as regards maize. By collecting pedigree, morphological, F1 yield, yield heterosis, isozymic, RFLP, and two-dimensional gel data on a common set of materials, we can combine information from several key areas. These key areas are (1) laboratory-based cultivar identification and intercultivar distance measures; (2) field-based intercultivar distance measures based on agronomically important multigenic traits such as F1 yield and yield heterosis, isozymic, RFLP, and two-dimensional gel data on a comexpected levels of intercultivar similarity based on recurrent parent dependency (back-cross breeding). These data show that RFLPs can provide intercultivar distance measures in maize that reflect genotypic similarity (J. S. C. Smith et al., 1991a). Interfacing these four sets of data allows the issue of minimum distance to be viewed from different contexts: first, from the context of similarity as revealed by RFLPs and as expected by pedigree, and second, from the context of the breeder with respect to dependency upon a previously released inbred line. Thus, measures of RFLP similarity between maize genotypes, validated by multigenic data, can be equated with degree of dependence and individual breeding effort.

There is justifiable concern about the extent of genetic diversity that is used by breeders and farmers (Duvick, 1977; 1984; National Academy of Sciences, 1972; Troyer et al., 1988). Without a continued source of variability, the ability to create new plateaus of agronomic performance that are based on complex genetic combinations could decline. Of more immediate concern is that sufficient genetic diversity should be made available to farmers. Diversity can allow the opportunity to hedge genetically based risks with respect to annually unforeseen difficulties that accrue from management, environmental, pest, and disease pressures. Unfortunately, only hindsight allows determination of what was an appropriate type or sufficient level of genetic diversity. Previously, information on diversity was gathered, with varying degrees of reliability, from field observations, pedigree surveys, results from variety trials, and rumor. However, in maize there is an opportunity to gather data that are highly informative on genetic diversity. RFLP diversity among hybrids has been shown to be highly correlated with distance based upon multigenic traits such as pedigree, F1 yield, and yield heterosis (0.S. Smith and Smith, 1991). Unique fingerprints exist among maize inbred lines (Lee et al., 1989;

FINGERPRINTING CROP VARIETIES

123

Marsan et al., 1991; 0. S. Smith et al., 1990;J. S . C. Smith et al., 1991c,d). Therefore, any lack of unique identification of maize F1 hybrids is much more a reflection of similar and even identical materials rather than being an indication' of the failure to provide discrimination.

E.

ASSURANCE

OF

GENETIC Furu-ry

A high level of genetic purity provides levels of performance that meet predicted expectations. Without a high standard of purity for varieties and inbred lines, genotypes will have the opportunity to drift. Different sources of the same named materials then would have different genetics with unpredictable performance. RFLP studies in maize have shown that residual segregation within finished lines is low to nonexistent (J. S. C. Smith et al., unpublished data; Boppenmaier et al., 1991), at least in recently finished lines. For hybrid crops such as maize, there is the additional necessity of maintaining purity of F, seed so that high performance is carried through to the commercial product. Biochemically based methods to test genetic purity as part of the overall program of quality assurance have been widely used by numerous organizations during the last 10 years for brassicas, maize, wheat, cotton, and sunflower (Motto et al., 1979). Currently, RFLP techniques are prohibitively expensive for such practical application, but technological advances could render them less expensive and thus extend their application.

F. THE IMPROVEMENTOF GERMPLASM IN TERMS OF AGRONOMIC PERFORMANCE This topic is so large and important as to warrant one or more reviews in its own right such as those given by Soller and Beckmann (1983), Edwards et al. (1987), and Walton and Helentjaris (1987). However, brief mention is appropriate here for it is the fingerprint profiles of individual segregants or varieties that should be useful as a source of information to empower plant breeders. Electrophoretically and chromatographically derived protein data are indicative of pasta and baking quality in wheat (Burnouf and Bietz, 1987) and for malting and brewing quality in barley (D. B. Smith and Simpson, 1983). Laboratory-derived data, which can be obtained more rapidly than field data, therefore, can be used to select preferentially parents and progenies that contain the desired genotypes for some key traits. Two-dimensional protein data could be useful in the identification of genotypes with special performance attributes because gene products are

124

J. S. C. SMITH AND 0. S . SMITH

viewed quantitatively, showing effects of gene regulation and/or different copy numbers (Damerval et al., 1987). The more complete genomic coverage provided in many crops by RFLPs has allowed the identification of genomic regions known as quantitative trait loci (QTL) that affect quantitatively inherited traits of agronomic importance. These data then could be used to select more rapidly progeny that carry the chromosomal segments that encompass regions with advantageous effects to agriculture. RFLP profiles can also be used to show associations among lines and identify their allocation into heterotic groups (Lee et al., 1989; 0 . S . Smith et al., 1990; J. S . C. Smith et al., 1991~).However, RFLP data cannot yet predict general and specific combining ability with the precision that would allow them to be used to identify parents of commercial F1 hybrids (Lee et al., 1989; Godshalk et al., 1990). RFLP data will allow, for the first time, calculations of relatedness among inbreds to be made on the basis of genetic data. Breeders will then be provided with coefficients of similarity data that are responsive to previous phylogeny and selection. RFLP data can allow the genetic evolutionary histories of inbred lines to be understood. RFLP data, in contrast to pedigrees, ignore neither selection nor phylogenies of lines prior to pedigree breeding. Breeders can perceive their raw material, germplasm, in its own terms and thus become more informed as to the availability and constitution of genetic resources in relation to agronomic performance.

G. EVIDENCE AND PREVENTION OF MISAPPROPRIATION Initial indications that an act of misappropriation may have been committed come from observations of commercial hybrids in farmers’ fields or varietal test plots planted for public viewing. Alternatively, routine screens of varietal profiles using protein or DNA data can identify potential instances of misappropriation. Procedures should be available for profiling that have been peer reviewed and that have gained acceptability in the scientific community; testing and data interpretation should be done in ignorance of sample identity. Suggested protocols have been made available for review and comment (Gill and Werrett, 1990; J. S . C. Smith el al., 1991f,g). Currently, RFLP data have not yet been used to address a disputation over genotypic identity relative to misappropriation. However, gliadin profiles were utilized as supplementary evidence in a case of misappropriation in wheat (Anonymous, 1986; G. L. Lookhart, personal communication). Isozymic data were ruled as unacceptable in a case involving celery [Bud Antle, Inc., v. Scattini Seed Co., D. C. N. Cal. (1985)l. This was

FINGERPRINTING CROP VARIETIES

125

largely because insufficient isozymic variability could be revealed in celery. In wheat, electrophoretic and chromatographic data have been used to discredit the pedigree of varieties (Wrigley and Shepherd, 1977; Jones ef al., 1982; Bietz, 1985). In maize, RFLP data have been used to determine the pedigree background of a line for which pedigree data did not exist (Lee et al., 1990b). Isozymic and chromatographic data have been used in the United States Federal Court to invalidate the pedigree of an inbred line of maize. In this case (U.S. District Court, 1987), it was ruled that a proprietary line had been taken and used by another organization. The precedence of utilizing laboratory-derived data in plants and the increasing usage of DNA-based fingerprinting techniques in civil and criminal cases involving humans means that any future disputes involving plants will surely utilize the most discriminative descriptors for the crop species in question. There have been justifiable concerns over the gel running and profile interpretation standards in several criminal human cases (Lander, 1989). However, in plants, all of the problems encountered in human forensics that emanate from reliance upon small quantities of degraded and possibly contaminated blood samples can be obviated by the availability of large quantities of good quality DNA and the practice of careful and thorough science (J. S . C. Smith et al., 1991f).

V. NEWTECHNIQUES Most advances in varietal descriptions will likely involve the use of DNA as a descriptor. Two-dimensional protein profiles are complex and need not be considered for cultivar identification if other highly discriminative techniques such as RFLPs or one-dimensional protein separations can suffice for a particular cultivated species. However, two-dimensional electrophoretic techniques are improving in their ability to provide useful repeatable data that simultaneously allow the quantitative assay of hundreds of gene products. Currently, RFLPs cannot assay large numbers of genetic markers as quickly. Technological advances in two areas are likely to occur: first, the development of means to identify uniquely varieties for crop species that are either difficult to profile uniquely because current RFLP procedures reveal little or no variation (e.g., wheat) or only moderate amounts of variation (e.g. soybean), and second, the development of methods to profile more rapidly and discriminate among varieties of crop species such as brassicas and maize, for which abundant variability has already been shown. Developments in both areas are obviously nonexclusive and will be in close

126

J. S. C. SMITH AND 0. S. SMITH

concert. Some technical developments coming from other genome mapping and sequencing projects will also likely come to bear on cultivar fingerprinting in the future. Intense impetus to these efforts comes from the desirability to use genetic marker data in applied plant breeding programs. Additional discrimination in several cultivated species could be afforded through the use of different probes, e.g., variable-number tandem repeat (VNTR) or M13 probes, which have already been shown to be useful (Dallas, 1988; Rogstad et al., 1988; Schafer et al., 1988; Nybom and Schaal, 1990; Nybom et al., 1990). Another approach could be to utilize four-base restriction enzyme cutters (Gebhardt et al., 1989) or, in contrast, restriction endonucleases with infrequent recognition sites followed by visualization of total DNA (Murray et al., 1990). Prehybridization of probe with restriction fragment and separation in a denaturing electrophoresis gel have been shown to be useful in discriminating among maize lines and also might be amenable to providing identification for cultivated species that currently appear to encompass less genetic variability (Losekoot et al., 1990; Riedel et al., 1990). DNA amplification could obviate the need for numerous subsequently repeated cycles of radio-labeling, membrane hybridization, washing, and film exposure. Polymorphisms revealed by primers would still need to be mapped. Use of conserved sequences that are spread throughout the genome as a result of transposable element activity could also provide sites for primer-directed DNA amplification. Alternately, DNA could be sequenced to provide primers for amplification. Portions of currently utilized probed regions or dispersed regions with variable lengths of repetitive DNA (Jeffreysetal., 1988;DeCorte etal., 1990)could be targeted. Another approach would be to insert specificsequences as cultivar identification labels via transformation procedures (Beckmann and Soller , 1986; Beckmann and Bar-Joseph, 1986). Varietal identity could then be rapidly detected via PCR. Presumedly, identifier sequences would need to be placed around the genome such that at least all chromosome arms are tagged. Currently, even though transformation has been achieved in several crop species, such a practical utilization of transformation remains speculative. It could have particular use for those species that remain intransigent to unique identification through the use of existing means to detect their inherent genetic variability. Minimum distance standards would need to be carefully applied in all cases. For identification of germplasm, it would be very important to identify any instances in which yet-to-be-developed techniques to alter specific coding regions had been specifically used to change the cultivar profile identity alone. For those several crop species for which sufficient inherent variability already exists, genetic profiling for many breeding applications, measure-

FINGERPRINTING CROP VARIETIES

127

ment of genetic diversity, mininum distance assessment, and testing for misappropriation is already not only an economic and practical feasibility, but also a necessary component of routine research. The costs of performing such tests will decline in real terms as technology develops. The technology that could alter profiles for many specific marker sites and thus render fingerprints no longer tenable for many of their useful applications will be very expensive and time consuming to achieve. Thus, for the foreseeable future, biochemical descriptors, and in particular DNA-based descriptors, will be valuable tests for providing information on genetic identity and similarity to the plant breeding and agricultural community at large. Many new techniques are in their infancy of development. It is expected that future advances in this technology will allow these techniques to become even more discriminatory with more widespread application, and they will become both easier and less expensive to use.

ACKNOWLEDGMENT We are deeply indebted to the support provided by Nancy Wiig who undertook the daunting task of typing our several evolving manuscripts.

REFERENCES Abernethy, R. H., Steiner, J. J., Wofford, D. S . , and Thil, D. S. (1989). Classification and pedigree verification of tall fescue cultivars utilizing the prolamin seed fraction. Crop Sci. 29, 791-797. Anderson, N. G., Tollaksen, S. L., Pascoe, F. H., and Anderson, L. (1985). Twodimensional electrophoretic analysis of wheat seed proteins. Crop Sci. 25, 667-674. Anonymous (1986). Plant fingerprints? Bioeng. News 7 , No. 8. Apuya, N. R . , Frazier, B. L., Keim, P., Jill Roth, E., and Lark, K. G. (1988). Restriction fragment length polymorphisms and genetic markers in soybean, Glycine max (L.) Merrill. Theor. Appl. Genet. 75, 889-901. Arcioni, S., Mariotti, D., and Ceccarelli, S. (1980). Phosphoglucoisomerase (PGI) polymorphism and cultivars distinctiveness in Lolium perenne L. Genet. Agric. 34, 101-112. Arus, P., Tanksley, S. D., Orton, T. J., and Jones, R. A. (1982). Electrophoretic variation as a tool for determining seed purity and for breeding hybrid varieties of Brassica oleracea. Euphytica 31, 417-428. Atkinson, M. D . , Withers, L. A., and Simpson, M. J. A . (1986). Characterization of Cacao and a standardization of the procedure. Euphytica 35, 741-750. Autran, J.-C., and Abbal, P. (1988). Wheat cultivar identification by a totally automatic soft-laser scanning densitometry and computer-aided analysis of protein electrophoregrams. Electrophoresis 9, 205-213. Bade, H., and Stegemann, H. (1982). Protein patterns of coffee beans. Characterization by one- and two-dimensional electrophoresis. J . Agron. Crop Sci. 151, 89-98. Bailey, D. C. (1983). Isozymic variation and plant breeders’ rights. I n “Isozymes in Plant

128

J. S. C . SMITH AND 0. S. SMITH

Genetics and Breeding” (S. D. Tanksley and T. J. Orton, eds.), pp. 425-440. Elsevier, Amsterdam. Beckmann, J. S., and Bar-Joseph, M. (1986). The use of synthetic DNA probes in breeders’ rights protection: A proposal to superimpose an alphanumerical code on the DNA. Trends Biotechnol. 4, 230-232. Beckmann, J. S., and Soller, M. (1983). Restriction fragment length polymorphisms in genetic improvement: Methodologies, mapping, and costs. Theor. Appl. Genet. 67, 35-43. Beckmann, J. S., and Soller, M. (1986). Restriction fragment length polymorphisms and genetic improvement of agricultural species. Euphyrica 35, 111-124. Bietz, J. A. (1985). High performance liquid chromatography: How proteins look in cereals. Cereal Chem. 62, 201-212. Bietz, J. A. (1986). High-performance liquid chromatography of cereal proteins. Adv. Cereal Sci. Technol. 8, 105-170. Bietz, J. A., and Cobb, L. A. (1985). Improved procedures for rapid wheat varietal identification by reversed-phase high-performance liquid chromatography of gliadins. Cereal Chem. 62, 332-339. Bietz, J. A., Burnouf, T., Cobb, L. A., and Wall, J. S . (1984). Wheat varietal identification and genetic analysis by reversed-phase high-performance liquid chromatography. Cereal Chem. 61, 129-135. Bingham, E. T., and Yeh, K. J. (1971). Electrophoretic patterns among alfalfa seed proteins from selected varieties, experimental stocks, and species accessions. Crop Sci. 11,58-61. Bonierbale, M. W., Plaisted, R. L., and Tanksley, S. D. (1988). RFLP maps based on a common set of clones reveal modes of chromosomal evolution in potato and tomato. Genetics 120, 1095-1103. Boppenmaier, J., Melchinger, A. E., Brunklaus-Jung, E., and Herrmann, R. G. (1991). Comparisons among strains of inbreds for RFLPS. Maize Genet. N e d 65, 90. Bringhurst, R. S . , Arulsekar, A., Hancock, J. F., and Voth, V. (1981). Electrophoretic characterization of strawberry cultivars. J . Am. SOC. Hortic. Sci. 106(5), 684-687. Brink, D. E., Price, S. C., Nguyen, H., Fuorst, G., and Martinez, C. (1989). Genetic purity assessment of commercial single cross maize hybrids: Isoelectric focusing of zeins. Seed Sci. Technol. 17, 91-98. Brown, A. H. D. (1978). Isozymes, plant population structure and genetic conservation. Theor. Appl. Genet. 52, 145-157. Bunce, N. A. C., Forde, B. G., Kreis, M., and Shewry, P. R. (1986). DNA restriction fragment length polymorphism at hordein loci: Application to identifying and fingerprinting barley cultivars. Seed Sci. Technol. 14, 419-429. Burbidge, M. J., Batey, I. L., Campbell, W. P., Skerritt, J. H., and Wrigley, C. W. (1986). Distinction between barley varieties by grain characteristics, electrophoresis, chromatography, and antibody reaction. Seed Sci. Technol. 14, 619-629. Burnouf, T., and Bietz, J. A. (1987). Identification of wheat cultivars and prediction of quality by reversed-phase high-performance liquid chromatographic analysis of endosperm storage proteins. Seed Sci. Technol. 15, 79-99. Burnouf, T., Bietz, J. A., Cobb, L. A., and Wall, S. J. (1983). Reversed-phase high-performance liquid chromatography of gliadins from hexaploid wheat (Triticurn aesrivum L.) varieties cultivated in France and possibility of varietal identification. C. R. Seances Acad. Sci. Ser. 3 297, 377-382. Cardy, B. J., and Beaversdorf, W. D. (1984). Identification of soybean cultivars using isoenzyme electrophoresis. Seed. Sci. Technol. 12, 943-954. Cardy, B. J., and Kannenberg, L. W. (1982). Allozymic variability among maize inbred lines and hybrids: Applications for cultivar identification. Crop Sci. 22, 1016-1020.

FINGERPRINTING CROP VARIETIES

129

Carter, G. E., and Brock, M. M. (1980). Identification of peach cultivars through protein analysis. HortScience 15, 292-293. Chaparro, J. X . , Sederoff, R. R., and Werner, D. J. (1990). Phylogenetic relationships of Prunus species in the Amygdalus subgenus. HortScience 25, 1159. Chatfield, C., and Collins, A. J. (1980). “Introduction to Multivariate Analysis.” Chapman & Hall, London. Clements, R. L. (1987). A study of gliadins of soft wheats from the Eastern United States using a modified polyacrylamide gel electrophoresis procedure. Cereal Chem. 64, 442-448. Clydesdale, A,, and Draper, S. R. (1982). A revised classification of wheat gliadin electrophoresis. J. Nutl. Inst. Agric. Bot. 16, 53-60. Cohen, J. E. (1990). DNA fingerprinting: What (really) are the odds? Chance 3, 26-32. Cooke, R. J. (1984). The characterization and identification of crop cultivars by electrophoresis. Electrophoresis 5 , 59-72. Cooksey, D. A., and Graham, J. H. (1989). Genomic fingerprinting of two pathovars of phytopathogenic bacteria by rare-cutting restriction enzymes and field inversion gel electrophoresis. Phytoputhology 79, 745-750. Coulthart, M., and Denford, K. E. (1982). Isozyme studies in Brassicu. I. Electrophoretic techniques for leaf enzymes and comparison of B. napus, B. campestris and B. oleracea using phosphoglucomutase. Can. J. Plant Sci. 62, 621-630. Crawford, D. J. (1990). “Plant Molecular Systematics, Macromolecular Approaches.” Wiley, New York. Cregan, P. B., Matthews, B. F., Tully, R. E., and Weisemann, J. M. (1990). DNA fingerprinting as a tool to determine genetic distance among soybean genotypes. A m . SOC. Agron. Abstr. p. 84. Cross, J. S., and Adams, W. R. (1983a). Differences in the embryo-specificglobulins among maize inbred lines and their hybrids. Crop Sci. 23, 1160-1162. Cross, J . W., and Adams, W. R. (1983b). Embryo-specific globulins from Zea mays L. and their subunit composition. J . Agric. Food Chem. 31, 534-538. Dallas, J. F. (1988). Detection of DNA “fingerprints” of cultivated rice by hybridization with a human minisatellite DNA probe. Proc. Nutl. Acud. Sci. U.S.A. 85, 6831-6835. Damerval, C., Hebert, Y., and deVienne, D. (1987). Is the polymorphism of protein amounts related to phenotypic variability? A comparison of two-dimensional electrophoresis data with morphological traits in maize. Theor. Appl. Genet. 74, 194-202. DeCorte, R., Cuppens, H., Marymen, P., and Cassiman, J. J. (1990). Rapid detection of hypervanable regions by the polymerase chain reaction technique. DNA Cell Biol. 9, 461-469. Dillon, W. R., and Goldstein, M. (1984). “Multivariate Analysis Methods and Applications.” Wiley, New York. Doebley, J. F. (1989). Isozymic evidence and the evolution of crop plants. In “Isozymes in Plant Biology” (D. E. Soltis and P. S. Soltis, eds.), pp. 165-191. Dioscorides Press, Portland, Oregon. Doebley, J. F., and Wendel, J. F. (1989). Applications of RFLPs to plant systematics. In “Current Communications in Molecular Biology: Development and Application of Molecular Markers to Problems in Plant Genetics” (T. Helentjaris and B. Burr, eds.), pp. 57-67. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Draper, S. R. (1987). ISTA variety committee report of the working group of biochemical tests for cultivar identification. Seed Sci. Technol. 15, 431-434. Drawert, F., and Gorg, A. (1974). On the electrophoretical differentiation and classification of proteins. 111. discelectrophoretical and iso-electrical focussing of proteins and enzymes of different grape varieties. Z. Lebensm.-Unters. -Forsch. 154, 328-338.

130

J. S. C . SMITH AND 0. S. SMITH

Drawert, F., Gorg, A., and Staudt, G. (1974). On the electrophoretical differentiation and classification of proteins. IV. Disc electrophoresis and isoelectrical focusing in polyacrylamide gels of proteins and enzymes from various species, varieties and hybrid species of strawberries. Z . Lebemm.-Unters. -Forsch. 156, 129-138. Duesing, J. H., and Raeber, J. G. (1989). Requirements of industrial firms for intellectual property protection in plant breeding and biotechnology. I n Symposium on Proprietary Rights for New Plant Material. XI1 EUCARPIA Congress, 27 February-4 March, 1989, Gottingen, Germany, “Mimeo. Ciba-Giegy Ltd., Bade, Switzerland. Dunbar, B. D., Bundman, D. S., and Dunbar, B. S. (1985). Identification of cultivar-specific proteins of winter wheat (T. aestivum L.) by high resolution two-dimensional polyacrylamide gel electrophoresis and color-based silver stain. Electrophoresis 6, 39-43. Durham, R., Moore, G., and Guy, C. (1990). Linkage of restriction fragment length polymorphisms and isozymes in a backcross of citrus and poncirus. HortScience 25, 1154. Duvick, D. N. (1977). Major United States crops in 1976. Ann. N. Y.Acud. Sci. 287,86-96. Duvick, D. N. (1984). Genetic diversity in major farm crops on the farm and in reserve. Econ. Bot. 38, 157-174. Edwards, M. D . , Stuber, C. W., and Wendel, J. F. (1987). Molecular-marker-facilitated investigations of quantitative-trait loci in maize. Genetics 116, 113-125. Esen, A., Mohammad, K., Schurig, G. G., and Aycock, H. S. (1989). Monoclonal antibodies to zein discriminate certain maize inbreds and genotypes. J. Hered. 80, 17-23. Eugene, A. M., and Wolfe, W. H. (1982). Identification of pecan cultivars with pollen isoenzymes. HortScience 17, 382-383. Figdore, S . S . , Kennard, W. C., Song, K., Slocum, M. K., and Osborn, T. C. (1988). Assessment of the degree of restriction fragment length polymorphisms in Brassicu. Theor. Appl. Genet. 75, 833-840. Fjellstrom, R., and Parfitt, D. E. (1990). Genetic diversity in walnuts using RFLP markers. HortScience 25, 1081. Gardiner, S. E., and Forde, M. B. (1988). Identification of cultivars and species of pasture legumes by sodium dodecylsulphate polyacrylamide gel electrophoresis of seed proteins. Plant Varieties See& 1, 13-26. Gebhardt, C., Blomendahl, C., Schachtschabel, U., Debener, T., Salamini, F., and Ritter, E. (1989). Identification of 2n breeding lines and 4n varieties of potato (Solanum tuberosum ssp. tuberosurn) with RFLP-fingerprints. Theor. Appl. Genet. 78, 16-22. Gebre, H., Khan, K., and Foster, A. E. (1986). Barley cultivar identification by polyacrylamide gel electrophoresis of hordein proteins: Catalogue of cultivars. Crop Sci. 26, 454-460. Gill, P., and Werrett, D. J. (1990). Interpretation of DNA profiles using a computerized database. Elecrrophoresis 11, 444-448. Gill, P., Jeffreys, A. J . , and Werrett, D. J. (1985). Forensic application of DNA ‘fingerprints.’ Nature (London) 318, 577-579. Gilliland, T. J., Camlin, M. S . , and Wright, C. E. (1982). Evaluation of phosphoglucoisomerase allozymeelectrophoresis for the identification and registration of cultivars of perennial ryegrass (Lolium perenne). Seed Sci. Technol. 10, 415-430. Glaszmann, J. L., Fautret, A., Noyer, J. L., Feldmann, P., and Lanaud, C. (1989). Biochemical genetic markers in sugarcane. Theor. Appl. Genet. 78,537-543. Godshalk, E . B., Lee, M., and Lamkey, K. R. (1990). Relationship of restriction fragment length polymorphism to single-cross hybrid performance of maize. Theor. Appl. Genet. 80,273-280. Gogorcena, Y, Arulsekar, S . , Dandekar, A., andParfitt, D. E. (1990). RFLPs for identification of grape cultivars. HortScience 25, 1081. Goodman, M. M., and Stuber, C. W. (1980). Genetic identificationof lines and crosses using isoenzyme electrophoresis. Proc. 35th Annu. Corn Sorghum Res. Conf., pp. 10-31.

FINGERPIUNTING CROP VARIETIES

131

Gorg, A., Postel, W., Domscheit, A., and Gunther, S. (1988). Two-dimensional electrophoresis with immobilized pH gradients of leaf proteins from barley (Hordeum vutgare): Method, reproducibility, and genetic aspects. Electrophoresis 9, 681-692. Gottlieb, L. D. (1977). Electrophoretic evidence and plant systematics. Ann. Mo. Bot. Card. 64, 161-180. Gower, J. C. (1972). Measures of taxonomic distance and their analysis. In “Assessment of Population Affinities in Man” (J. S. Weiner and J. Huizinga, eds.), pp. 371-380. Oxford Univ. Press, London. Grafen, A. (1985). A geometric view of relatedness. Oxford Surv. Evol. Biol. 2, 28-89. Graner, A., Siedler, H., Jahoor, A,, Herrmann, R. G . , and Wenzel, G. (1990). Assessment of the degree and the type of restriction fragment length polymorphism in barley (Hordeurn vulgare). Theor. Appl. Genet. 80, 826-832. Graybosch, R. A , , and Morris, R. (1990). An improved SDS-PAGE method for the analysis of wheat endosperm storage proteins. J. Cereal Sci. 11, 202-212. Guo, Y. J . , Bishop, R., Ferhnstrom, H . , Yu, G . Z . , Lian, Y . N., and Huang, S. D. (1986). Classification of Chinese rice varieties by electrofocusing. Cereal Chem. 63, 1-3. Gupta, S. K., and Robbelen, G. (1986). Identification of rapeseed (Brassica napus) cultivars by electrophoresis. 2.pflanzenzuecht. 96, 363-370. Gupta, S. K., and Shepherd, K. W. (1990). Two-step one-dimensional SDS-PAGE analysis of LMW subunits of glutelin. Theor. Appl. Genet. 80, 65-74. Hadacova, V., Klozova, E., Hadac, E., Turkova V . , and Pitterova, K. (1981). Comparison of esterase isoenzyme patterns in seeds of some Allium species and in cultivars of Allium cepa L. Biol. Plant. 23(3), 174-181. Halward, T. M., Kochert, G . D., Larue, E. A., and Stalker, H. T. (1990). RFLP variability among diverse germplasm resources in peanut. A m . Soc. Agron. Abstr. p. 197. Hansen, A . E., Nassuth, A., and Altosaar, I. (1988). Rapid electrophoresis of oat (Avena sativa L.) prolamins from single seeds for cultivar identification. Cereal Chem. 65, 153-154. Havey, M. J . (1990). Variability for restriction fragment length polymorphism in the chloroplast and nuclear genome of Allium species. HortScience 25, 1159. Harvey, M. J . , and Muehlbauer, F. J . (1989). Linkages between restriction fragment length, isozyme, and morphological markers in lentil. Theor. Appl. Genet. 77, 395-401. Hayward, M. D., and McAdam, N. J . (1977). Isozyme polymorphism as a measure of distinctiveness and stability in cultivars of Lolium perenne. Ppanzenzuecht. 79, 59-68. Heisel, S. E., Peterson, D. M., and Jones, B. L. (1986). Identification of United States barley cultivars by sodium dodecyl sulphate polyacrylamide gel electrophoresis hordeins. Cereal Chem. 63, 500-505. Helentjaris, T., King, G., Slocum, M . , Siedenstrang, C., and Wegman, S. (1985). Restriction fragment polymorphisms as probes for plant diversity and their development as tools for applied plant breeding. Plant Mol. Biol. 5 , 109-118. Helentjaris, T., Slocum, M., Wright, S . , Schaefer, A., and Nienhuis, J. (1986). Construction of genetic linkage maps in maize and tomato using restriction fragment length polymorphisms. Theor. Appl. Genet. 72, 761-769. Hicks, P. H., Sleper, D. A . , Randall, D. D., and Crane, C. F. (1982). Peroxides isozyme differences in tall fescue cultivars and allopolyploid accessions. Euphytica 31, 175-181. Higginbotham, J. W., and Smith, J. S . C. (1989). Reliability of the protein extraction procedure used to prepare seedling samples for two-dimensional electrophoresis. Maize Gener. Coop. Newsl. 63, 83-84. Higginbotham, J. W., Smith, J. S. C., and Smith, 0. S. (1989a). The effect of different seed lots versus genotypes on 2-D protein profiles of seedlings. Maize Genet. Coop. Newsl. 63, 84- 85. Higginbotham. J. W., Smith, J. S. C., and Smith, 0. S. (1989b). The effect of slightly

132

J. S. C. SMITH AND 0. S. SMITH

varying imbibition periods versus genotypes on 2-D protein profiles of seedlings. Maize Genet. Coop. Newsl. 63, 85. Higginbotham, J. W., Smith, J. S. C., and Smith, 0. S. (1991a). Quantitative analysis of two-dimensional protein profiles of inbred lines of maize (Zea mays L.). Electrophoresis 12,425-431. Higginbotham, J. W., Smith, J. S. C., and Smith, 0. S. (1991b). Computerized twodimensional electrophoretic protein profiles of 37 inbred lines and one hybrid. Maize Genet. Coop. Newsl. 65, 65. Higgins, J., Evans, J. L., and Law, J. R. (1988). A revised classification and description of Faba bean cultivars (Vicia faba L.). Plant Varieties Seeds 1, 27-35. Higgins, T. J. V. (1984). Synthesis and regulation of major proteins in seeds. Annu. Rev. Plant Physiol. 35, 191-221. Hillel, J., Plotsy, Y., Haberfeld, A., Lavi, U., Cahaner, A., and Jeffreys, A. J. (1989). DNA fingerprints of poultry. Anim. Genet. 20, 145-155. Hoffman, D. L., Murphy, J. P., and Guttieri, M. J. (1990). The development of a molecular linkage map in oats. A m . SOC. Agron. Abstr. p. 197. Holt, S. B. (1968). “The Genetics of Dermal Ridges.” Thomas, Springfield, Illinois. Huebner, F. R., and Bietz, J. A. (1988). Quantitative variation among gliadins of wheats grown in different environments. Cereal Chem. 65, 362-366. Huebner, F. R., Kaczowski, J., and Bietz, J. A. (1990). Quantitative variation of wheat proteins from grain at different stages of maturity and from different spike locations. Cereal Chem. 67, 464-470. Hulbert, S. H., Richter, T. E., Axtell, J. D., and Bennetzen, J. L. (1990). Genetic mapping and characterization of sorghum and related crops by means of maize DNA probes. Proc. Natl. Acad. Sci. U.S.A. 87, 4251-4255. Hussain, A,, Ramirez, H., Bushuk, W., and Roca, W. (1986). Field bean (Phaseolus vulgaris L.) cultivar identification by electrophoregrams of cotyledon storage proteins. Euphytica 35, 729-732. Jacquard, A. (1974). “The Genetic Structure of Populations.” Springer-Verlag, New York. Jarrell, D. C., and Roose, M. L. (1990). A genetic map of citrus based on isozymes and RFLPs. HortScience 25, 1154. Jeffreys, A. J. (1987). Highly variable minisatellites and DNA fragments. Biochem. SOC. Trans. 15, 309-317. Jeffreys, A. J., and Morton, D. B. (1987). DNA fingerprints of dogs and cats. Ani. Genet. 18, 1-15. Jeffreys, A. J., Wilson, V., Neumann, R., and Keyte, J. (1988). Amplification of human minisatellites by the polymerase chain reaction: Towards DNA fingerprinting of single cells. Nucleic Acids Res. 10, 10953-10970. Jones, B. L., Lookhart, G. L., Hall, S. B., and Finney, K. F. (1982). Identification of wheat cultivars by gliadin electrophoresis: Electrophoregrams of the 88 wheat cultivars most commonly grown in the United States in 1979. Cereal Chem. 59, 181-188. Kahler, A. L., and Lay, C. L. (1985). Genetics of electrophoretic variants in the annual sunflower. J. Hered. 76, 335-340. Kaplan, N. (1983). Statistical analysis of restriction enzyme map data and nucleotide sequence data. In “Statistical Analysis of DNA Sequence Data” (B. S. Weir, ed.), pp. 75-106. Dekker, New York. Kapse, S. S., and Nerkar, Y. S. (1985). Polyacrylamide gel electrophoresis of soluble seed proteins in relation to cultivar identification in cotton. Seed Sci. Technol. 13, 847-852. Keefe, P. D., and Draper, S. R. (1986). The measurement of new characters for cultivar identification in wheat using machine vision. Seed Sci. Technol. 14, 715-724. Keefe, P. D., and Draper, S. R. (1988). An automated machine vision system for the

FINGERPRINTING CROP VARIETIES

133

morphometry of new cultivars and plant genebank accessions. Plant Varieties Seeds 1, 1-11. Keim, P., Shoemaker, R. C., and Palmer, R. G. (1989). Restriction fragment length diversity in soybean. Theor. Appl. Genet. 77, 786-792. Kennard, W., Dijhuizen, A., Havey, M., and Staub, J. (1990). Progress toward development of an RFLP map for cucumber. HortScience 25, 1159. Kenny, S. T., Furter, R., Whelen, S., and Hall, B. D. (1990). Restriction fragment length polymorphisms (RFLPs) in hop. A m . Soc. Agron. Abstr. p. 199. Kesseli, R. V., and Michelmore, R. W. (1986). Genetic variation and phylogenies detected from isozyme markers in species of Lactura. 1. Hered. 77, 324-331. Kidwell, K. K., and Osborn, T. C. (1990). Polymorphism for restriction fragment lengths within diploid Medicago lines. Am. SOC. Agron. Abstr. p. 199. Kirby, L. T. (1990). “DNA Fingerprinting: An Introduction.” Stockton Press, New York. Koranyi, P. (1989a). Characterization of maize (Zea mays L.) seed samples by the electrophoretic patterns of their protein monomers. Seed Sci. Technol. 17, 153-159. Koranyi, P. (1989b). Simple purity checking of maize (Zea mays L.) lines and hybrids by protein monomer analysis. Seed Sci. Technol. 17, 161-168. Kuhns, L. J., and Fretz, T. A. (1978a). Distinguishing rose cultivars by polyacrylamide gel electrophoresis. I. Extraction and storage of protein and active enzymes from rose leaves. J . A m . SOC.Hortic. Sci. 103(4), 503-508. Kuhns, L. J., and Fretz, T. A. (1978b). Distinguishing rose cultivars by polyacrylamide gel electrophoresis. 11. Isoenzyme variation among cultivars. J. Am. SOC.Hortic. Sci. 103(4), 509-516. Laduinsky, G . , and Hymowitz, T. (1979). Seed protein electrophoresis in taxonomic and evolutionary studies. Theor. Appl. Genet. 54, 145-151. Lander, E. S. (1989). DNA fingerprinting on trial. Nature (London) 339, 501-505. Landry, B. S., and Michelmore, R. W. (1985). Selection of probes for restriction fragment length analysis from plant genomic clones. Plant Mol. Biol. Rep. 3, 174-179. Landry, B. S . , Kesseli, R., Leung, H., and Michelmore, R. W. (1987). Comparison of restriction endonucleases and sources of probes for their efficiency in detecting restriction fragment length polymorphisms in lettuce (Lactuea satium L.) Theor. Appl. Genet. 74, 646-653. Lee, J. W., Ronalds, J. A. (1967). Effect of environment on wheat gliadin. Nature (London) 213, 844-846. Lee, M., Godschalk, E. B., Lamkey, K. R., and Woodman, W. L. (1989). Association of restriction fragment length polymorphisms among maize inbreds with agronomic performance of their crosses. Crop Sci. 29, 1067-1071. Lee, M., Whitkus, R., Doebley, J., Woodman, L. and Rayapati, P. J. (1990a). Restriction fragment length polymorphisms in sorghum detected by maize and sorghum DNA clones. A m . Soc. Agron. Abstr. p. 199. Lee, M., Russell, W. A., Melchinger, A. E., and Woodman, W. L. (1990b). On the origin of the inbred line B52. Maize Genet. Coop. Newsl. 64, 21. Lin, C. S., and Binns, M. R. (1984). The precision of cultivar trials within eastern cooperative tests. Can. J . Plant Sci. 64, 587-591. Liou, P., Gmitter, F. G., and Moore, G. A. (1990). Molecular characterization and linkage mapping of the citrus genome using isozyme and RFLP markers. HortScience 25, 1154. Livneh, O., Nayler, Y., Tal., Y., Harush, S. B., Gafni, Y., Beckmann, J. S., and Sela, I. (1990). RFLP analysis of a hybrid cultivar of pepper (Capsicum annuum) and its use in distinguishing between parental lines and in hybrid identification. Seed Sci. Technol. 18, 209-2 14. Loeschcke. V., and Stegemann, H . (1978). Identification of potato cultivars by electrophoresis. Abstr., Int. Cong. Plant Pathol., 3rd, Munich, 1978, 99, 302.

134

J. S. C. SMITH AND 0. S. SMITH

Loftus, M. E., Moore, D., and Elliott, T. J. (1988). DNA polymorphisms in commercial and wild strains of the cultivated mushroom, Agaricus bisoporus. Theor. Appl. Genet. 76, 712-718. Lookhart, G. L. (1985a). Identification of oat cultivars by combining polyacrylamide gel electrophoresis and reversed-phase high-performance liquid chromatography. Cereal Chem. 62, 345-350. Lookhart, G. L. (1985b). Gliadin electrophoretic variations in foundation Arkan wheat grown at 16 Kansas Agricultural Experiment Stations in 1983. Cereal Chem. 62,355-360. Lookhart, G. L., and Bietz, J.. A. (1990). Practical wheat identification in the United States. Cereal Foods World 35, 404-407. Lookhart, G. L., and Finney, K. F. (1984). Polyacrylamide gel electrophoresis of wheat gliadins: The effect of environment and germination. Cereal Chem. 61, 496-499. Lookhart, G. L., and Pomeranz, Y. (1985a). Gliadin high-performance liquid chromatography and polyacrylamide gel electrophoresis patterns of wheats grown with fertilizer treatments in the United States and Australia on sulfur-deficient soils. Cereal Chem. 62, 227-229. Lookhart, G. L., and Pomeranz, Y. (1985b). Characterization of oat species by polyacrylamide gel electrophoresis and high performance liquid chromatography of their prolamin proteins. Cereal Chem. 62, 162-166. Lookhart, G. L., Jones, B. L., Walker, D. E., Hall, S . B., and Cooper, D. B. (1983). Computer-assisted method for identifying wheat cultivars from their gliadin electrophoregrams. Cereal Chem. 60, 111-115. Lookhart, G. L., Albers, L., Pomeranz, Y., and Webb, B. D. (1987). Identification of U.S. rice cultivars by high-performance liquid chromatography. Cereal Chem. 64, 199-206. Losekoot, M., Fodde, R., Harteudd, C. L., van Heeren, H., Giordano, P. C., and Bernini, L. F. (1990). Denaturing gradient gel electrophoresis and direct sequencing of PCR amplified genomic DNA: A rapid and reliable diagnostic approach to beta thalassaemia. Br. J . Haematol. 76, 269-274. Lynch, M. (1988). Estimation of relatedness by DNA fingerprinting. Mol. Biol. Evol. 5 , 584-599. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982). “Molecular Cloning: A Laboratory Manual.” Cold Spring Harbor Lab., Cold Spring Harbor, New York. Marchylo, B. A., and Kruger, J. E. (1984). Identification of Canadian barley cultivars by reversed-phase high-performance liquid chromatography. Cereal Chem. 61, 295-301. Marchylo, B. A,, Hatcher, D. W., and Kruger, J. E. (1988). Identification of wheat cultivars by reversed-phase high-performance liquid chromatography of storage proteins. Cereal Chem. 65,28-40. Marchylo, B. A., Handel, K. A., and Mellish, V. J. (1989). Fast horizontal sodium dolecyl sulfate gradient polyacrylamide gel electrophoresis for rapid wheat cultivar identification and analysis of high molecular weight glutenin subunits. Cereal Chem. 66, 186-192. Marchylo, B. A., Kruger, J. E., and Hatcher, D. W. (1990). Effect of environment on wheat storage proteins as determined by quantitative reversed-phase high-performance liquid chromatography. Cereal Chem. 67, 372-376. Marsan, P. A., Livini, C., Melchinger, A. E., Messmer, M. N., and Motto, M. (1991). Relatedness of inbreds based on RFLP data. Maize Genet. Coop. Newsl. 65, 26-27. Marshall, D. R., and Brown, A. H. D. (1975). The charge state model of protein polymorphism in natural populations. J. Mol. Evol. 6, 149-163. Melchinger, A. E., Lee, M., Lamkey, K. R., and Woodman, W. L. (1990). Genetic diversity for restriction fragment length polymorphisms: Relation to estimated genetic effects in maize inbreds. Crop Sci. 30, 1033-1040. Molina-Cano, J. L., and Elena Rossello, J. M. E. (1978). A further contribution to the

FINGERPRINTING CROP VARIETIES

135

classification of barley cultivars: Use of numerical taxonomy and biochemical methods. Seed Sci. Technol. 6, 593-615. Moore, P. P. (19x1). Chloroplast restriction fragment variability in raspberry. HortScience 25, 1159. Motto, M., Salamini, F., Reggiani, R., and Soave, C. (1979). Evaluation of genetic purity in hybrid corn (Zea mays L.) seed production through zein isoelectrophoretic patterns. Maydica 24, 223-233. Murray, B.E., Singh, K. V., Heath, J. D., Sharma, B. R., and Weinstock, G. M. (1990). Comparison of genomic DNAs of different enterococcal isolates using restriction endonucleases with infrequent reco nition sites. J. Clin. Microbiol. 28, 2059-2063. Nagamine, T., Catty, J. P., and For]-Lloyd, B. V. (1989). Phenotypic polymorphism and allele differentiation of isozymes in fodder beet, multigerm sugar beet and monogerm sugar beet. Theor. Appl. Genet. 77, 711-720. Nakamura, S. (1977). Studies on the determination of species and cultivars on the basis of electrophoretic patterns of seed protein and seed enzymes. I. Brassica spp., radish and Cucurbita spp. J . Jpn. SOC.Hortic. Sci. 46(1), 32-47. Nakamura, S. (1979). Separation of ryegrass species using electrophoretic patterns of seed proteins and enzymes. Seed Sci. Technol. 7 , 161-168. Nakamura, S., and Tahara, M. (1977). Studies on the determination of species and cultivars on the basis of electrophoretic patterns of seed protein and seed enzymes. 11. Allium spp., cucumber and melon, pea, and garden bean. J. Jpn. SOC. Hortic. Sci. 46(2), 233-244. National Academy of Sciences (1972). “Genetic Vulnerability of Major Crops.” Committee on Genetic Vulnerability of Major Crops, Agricultural Board, National Research Council, National Academy of Sciences, Washington, D.C. National Institute of Agricultural Botany (1983). “Detailed Descriptions of Varieties of Wheat, Barley, Oats, and Rye.” National Institute of Agricultural Botany, Cambridge, England. Nei, M., and Li, W. H. (1979). Mathematical model for studying genetic variation in terms of restriction endonucleases. Proc. Natl. Acad. Sci. U.S.A. 76, 5269-5273. Nei, M., and Miller, J. C. (1990). A simple method for estimating average number of nucleotide substitutions within and between populations from restriction data. Genetics 125, 873-879. Neuman, M., Sapirstein, H. D., Schwedyk, E., and Bushuk, W. (1987). Discrimination of wheat classes and variety by digital image analysis of whole grain samples. J . Cereal Sci. 6, 125-132. Nielsen, G. (1985). The use of isozymes as probes to identify and label plant varieties and cultivars. In “Isozymes: Current Topics in Biological and Medical Research” (M. C. Rattazzi, J. G. Scandalios, and G. S. Whitt, eds.), Vol. 12, pp. 1-32. Liss, New York. Nielsen, G., and Johansen, H. B. (1986). Proposal for the identification of barley varieties based on the genotypes for 2 hordein and 39 isoenzyme loci of 47 reference varieties. Euphytica 35, 717-728. Nodari, R., Singh, S . P., and Gepts, P. (1990). RFLPs and linkage mapping in common bean. Am. SOC. Agron. Abstr. p. 201. Nybom, H., and Schaal, B. A. (1990). DNA “fingerprints” applied to paternity analysis in apples (Maius X domestica). Theor. Appl. Genet. 79, 763-768. Nybom, H., Schaal, B. A,, and Rogstad, S. H. (1989). DNA “fingerprints” can distinguish cultivars of blackberries and raspberries. Acra Hortic. 262, 305-310. Nybom, H., Rogstad, S. H., and Schaal, B. A. (1990). Genetic variation detected by use of the M13 ‘DNA fingerprint’ probe in Malus, Prunus and Robus (Rosaceae). Theor. Appl. Genet. 79, 153-156.

J. S. C. SMITH AND 0. S. SMITH

136

Ostergaard, H, and Nielsen, G. (1981). Cultivar identification by means of isoenzymes. I. Genotypic survey of the Pgi-2 locus in tetraploid ryegrass. Z . Ppanenzuecht. 87,121-132. Parnell, A. (1983). The identification of new wheat varieties using a standard electrophoresis method. J . Natl. Inst. Agric. Bot. 16, 183-188. Patterson, H. D., and Weatherup, S . T. C. (1984). Statistical criteria for distinctness between varieties of herbage crops. J . Agric. Sci. 102, 59-68. Pauksens, J. (1975). Methods for determination of cultivar trueness and purity in maize (Zea mays L.). Seed Sci. Technol. 3, 176-181. Payne, R. C., Scott, J. A., and Koszykowski, T. J. (1980). An esterase isoenzyme difference in seed extracts of annual and perennial ryegrass. J . Seed. Technol. 5 , 14-20. Pontikis, C. A., Loukas, M., and Kousounis, G. (1980). The use of biochemical markers to distinguish olive cultivars. J . Hortic. Sci. 55(4), 333-343. Quaite, E.,Camlin, M. (1985). Electrophoretic labelling of perennial ryegrass (Lolium perenne L.) cultivars for the examination of seed mixtures. Seed Sci. Technol. 14, 553-566. Quiros, C . F. (1980). Identification of alfalfa plants by enzyme electrophoresis. Crop Sci. 20, 262-264. Quios, C . F. (1981). Starch gel electrophoresis technique used with alfalfa and other Medicago species. Can. J . Plant Sci. 61, 745-749. Rajapakse, S., Hubbard, M., Abbott, A., Kelly, J., and Ballard, R. (1990). DNA fingerprinting of rose cultivars by restriction fragment length polymorphisms. HortScience 25, 1159. Ramirez, H., Hussain, A., Roca, W., and Bushuk, W. (1987). Isozyme electrophoregrams of sixteen enzymes in five tissues of Cassava (Manihot esculenta Crantz) varieties. Euphytica 36, 39-48. Ramirez, L., Pisabarro, G. (1985). Isozyme electrophoretic patterns as a tool to characterize and classify rye (Secale cereale L.) seed samples. Euphytica 34, 793-799. Rao, T. N., Nerkar, Y. S . , and Patil, V. D. (1990). Identification of cultivars of cotton by sodium dodecylsulphate polyacrylamide gel electrophoresis (SDS-PAGE) of soluble seed proteins. Plant Varieties S e e h 3, 7-13. Rebordinos, L., and Perez de la Vega, M. (1990). Genetic variability of leaf esterases in Triticum aestivum L. 2n = 6x = 42. Theor. Appl. Genet. 79, 673-678. Riedel, G. E., Swanberg, S . L., Kuranda, K. D., Marquette, K., LaPan, P., Bledsoe, P., Kennedy, A., and Lin, B.-Y (1990). Denaturing gradient gel electrophoresis identifies genomic DNA polymorphism with high frequency in maize. Theor. Appl. Genet. 80, 1-10.

Rogers, J. S. (1972). Measures of genetic similarity and genetic distance. Univ. Tex. Publ. 7213. Rogstad, S. H., Patton, J. C., 11, and Schaal, B. A. (1988). M13 repeat probe detects DNA minisatellite-like sequences in gymnosperms and angiosperms. Proc. Natl. Acad. Sci. U.S.A. 85, 976-9178. Ryskov, A. P., Jincharadze, A. G., Prosnyat, M. I . , Ivanov, P. L., and Limborska, S. A. (1988). MI3 phage DNA as a universal marker for DNA fingerprinting of animals, plants, and microorganisms. FEES Lett. 223, 388-392. Santamour, F. S., and Demuth, P. (1980). Identification of callery pear cultivars by peroxidase isozyme patterns. J . Hered. 71, 447-449. Sapirstein, H. D., and Bushuk, W. (1985a). Computer-aided analysis of gliadin electrophoregrams. I. Improvement of precision of relative mobility determination by using a three reference band standardization. Cereal Chem. 62, 372-377. Sapirstein, H. D., and Bushuk, W. (1985b). Computer-aided analysis of gliadin electrophoregrams. 11. Wheat cultivar identification and class comparisons. Cereal Chem. 62, 377392.

FINGERPRINTING CROP VARIETIES

137

Sapirstein, H. D., and Bushuk, W. (198%). Computer-aided analysis of gliadin electrophoregrams. 111. Characterization of the heterogeneity in gliadin composition for a population of 98 common wheats. Cereal Chem. 62, 392-398. Sapirstein, H. D., and Bushuk, W. (1986). Computer-aided wheat cultivar identification and analysis of densitometric scanning profiles of gliadin electrophoregrams. Seed Sci. Technol. 14, 489-517. Sarkar, R., and Bose, S. (1984). Electrophoretic characterization of rice varieties using single seed (salt soluble) proteins. Theor. Appl. Genet. 68, 415-419. Sastry, L. V. S., Paulis, J . W., Bietz, J. A,, and Wall, J. S. (1986). Genetic variation of storage proteins in sorghum grain: Studies by isoelectric focusing and high-performance liquid chromatography. Cereal Chem. 63, 420-427. Scanlon, M. G . , Sapirstein, H. D., and Bushuk, W. (1989a). Evaluation of the precision of high-performance liquid chromatography for wheat cultivar identification. Cereal Chem. 66, 112-116. Scanlon, M. G., Sapirstein, H. D., and Bushuk, W. (1989b). Computerized wheat varietal identification by high-performance liquid chromatography. Cereal Chem. 66, 439-443. Schafer, R., Zischler, H., Birsner, U., Becker, A., and Epplen, J. T. (1988). Optimized oligonucleotide probes for DNA fingerprinting. Electrophoresis 9, 369-374. Schertz, K. F., Stec, A., and Doebley, J. F. (1990). Isozyme diversity of U.S. Sorghum lines and hybrids. A m . SOC. Agron. Abstr. p. 108. Shewry, P. R., Pratt, H.M., and Miflin, B. J. (1978). Varietal identification of single seeds of barley by analysis of hordein polypeptides. J . Sci. Food Agric. 29, 587-596. Shewry, P. R., Finch, R. A., Parmar, S., Franklin, J., and Miflin, B. J. t1983). Chromosomal location of H o d , a new locus governing storage proteins in barley. Heredity 50, 179-189. Skerritt, J. H., Smith, R. A , , Wrigley, C. W., and Underwood, P. A. (1984). Monoclonal antibodies to gliadin proteins used to examine cereal grain protein homologies. J . Cereal Sci. 2, 215-224. Slocum, M. K., Figdore, S. S., Kennard, W. C., Suzuki, J. Y . , and Osborn, T. C. (1990). Linkage arrangement of restriction fragment length polymorphism loci in Brussica oleracea. Theor. Appl. Genet. 80, 57-64. Smith, D. B., and Simpson, P. A. (1983). Relationships of barley proteins soluble in sodium dodecyl sulphate to malting quality and varietal identification. J . Cereal Sci. 1, 185-197. Smith, J. S. C. (1984). Genetic variability within U.S. hybrid maize: Multivarieties analysis of allozyme data. Crop Sci. 24, 1041-1046. Smith, J. S. C. (1988). The diversity of U. S. hybrid maize germplasm; isozymic and chromatographic evidence. Crop Sci. 28, 63-70. Smith, J . S. C. (1989). The characterization and asseessment of genetic diversity among maize (Zea mays L.) hybrids that are widely grown in France: Chromatographic data and isozyme data. Euphytica 43, 73-85. Smith, J. S. C., and Smith, 0. S. (1986). Environmental effects on zein chromatograms of maize inbred lines revealed by reversed-phase high-performance liquid chromatography. Theor. Appl. Genet. 71, 607-612. Smith, J. S. C., and Smith, 0. S. (1987). Associations among inbred lines of maize using electrophoretic, chromatographic, and pedigree data. I. Multivariate and cluster analysis of data from Lancaster Sure Crop derived lines. Theor. Appl. Genet. 73, 654-664. Smith, J. S. C., and Smith, 0. S. (1988a). Associations among inbred lines of maize using electrophoretic, chromatographic, and pedigree data. I. Multivariate and cluster analysis of data from Lancaster Sure Crop derived lines. Theor. Appl. Genet. 73, 654-664. Smith, J. S. C., and Smith, 0. S. (1988b). Variation in kafirin and alcohol-soluble glutelin

138

J. S. C. SMITH AND 0. S. SMITH

chromatograms of sorghum inbred lines revealed by reversed-phase high-performance liquid chromatography. Theor. Appl. Genet. 76, 97-107. Smith, J. S. C., and Smith, 0. S. (1989). The description and assessment of distances between lines of maize. 11. The utility of morphological, biochemical, and genetic descriptors and a scheme for the testing of distinctiveness between inbred lines. Maydica 34,151-161. Smith, J. S . C., and Smith, 0. S. (1991). Restriction fragment length polymorphisms can differentiate among U.S.maize hybrids. Crop Sci. 31, 893-899. Smith, J. S. C., and Weissinger, H. H. (1984). Rapid monitoring of purity in seed lots of hybrid maize: Modifications of current technologies. Maize Genet. Coop. Newsl. 58, 103-105. Smith, J. S. C., and Wych, R. D. (1986). The identification of female selfs in hybrid maize: A comparison using electrophoresis and morphology. Seed Sci. Technol. 14, 1-8. Smith, J. S. C., Paszkiewicz, S., Smith, 0. S., and Schaffer, J. (1987). Electrophoretic, chromatographic, and genetic techniques for identifying associations and measuring genetic diversity among corn hybrids. Proc. 42nd Annu. Corn Sorghum Ind. Res. Cone 187-203. Smith, J . S. C., Smith, 0. S., Bowen, S. L., Tenborg, R. A., and Walls, S . J . (1991a). The description and assessment of distances between inbred lines of maize. 111. A revised scheme for the testing of distinctiveness between inbred lines utilizing DNA RFLPs. Maydica 36, 213-226. Smith, J. S. C., Smith, 0. S., Wright, S., Wall, S:, and Walton, M. (1991b). Diversity of United States hybrid maize germplasm as revealed by RFLP. Crop Sci. (in press). Smith, J. S. C., Wright, S., Walton, M., Smith, 0. S . , and Wall, S . (1991~).Associations among 150 publicly available inbred lines as revealed by cluster analysis of RFLP data, Maize Genet. Coop. Newsl. 65, 67. Smith, J. S. C., Smith, 0. S., Bowen, S. L., Tenborg, R. A., Wall, S., and Duvick, D. N. (1991d). Ability of RFLPs to uniquely identify and to show associations among 160 elite lines used within a proprietary breeding program during 1930-1990. Maize Genet Coop. Newsl. 65, 67-68. Smith, J. S . C., Smith, 0. S., Grant, D., Bowen, S . L., and Tenborg, R. A. (1991e). Choice of restriction enzyme digests of genomic DNA in profiling inbred lines. Maize Genet. Coop. Newsl. 65, 68. Smith, J. S. C., Bowen, B., Grant, D., Orman, B. A., Fincher, R., and Smith, 0. S. (1991f). Standards for declaring and computing matching probabilities between genotypes of maize in cases where identities, pedigrees, ownership of germplasm are in dispute. Maize Genet. Coop. Newsl. 65, 68-69. Smith, J. S. C., Tenborg, R. A., Roth, B., Grant, D., Smith, 0. S., Brumback, R., and Beghtol, T. (1991g). Methods to precisely estimate molecular weights of fragments constituting RFLP “fingerprints” and a means to standardize measurements within and across laboratories. Maize Genet. Coop. Newsl. 65, 69. Smith, 0. S., and Smith, J. S. C. (1991). Measurement of genetic diversity among maize hybrids; a comparison of isozymic, RFLP, pedigree, and heterosis data. Maydica (in press). Smith, 0. S., Smith, J. S . C., Bowen, S. L., Tenborg, R. A., and Wall, S. F. (1990). Similarities among a group of elite maize inbreds as measured by pedigree, F, grain yield, grain yield heterosis, and RFLPs. Theor. Appl. Genet. 80, 833-840. Smith, 0. S., Smith, J. S. C., Bowen, S. L., and Tenborg, R. A. (1991). Numbers of RFLP probes necessary to show associations between lines. Maize Genet. Coop. Newsl. 65, 66. Soave, C., and Salamini, F. (1983). Genetic organization and regulation of maize storage

FINGERPRINTING CROP VARIETIES

139

proteins. In “Seed Proteins” (J. Daussant, J. Mosse, J. Vaughan, eds.), Chapter 9, pp. 205-218. Academic Press, London. Soller, M., and Beckmann, J. S. (1983). Genetic polymorphism in varietal identification and genetic improvement. Theor. Appl. Genet. 67,25-33. Stegemann, H. (1979). Characterization of proteins from potatoes, and the “Index of European Varieties.” In “The Biology and Taxonomy of the Solanaceae” (J. G. Hawkes, R. N. Lester, and A. D. Skelding, eds.), pp. 279-284. Academic Press, London. Stegemann, H., and Loeschcke, V. (1976). Index of European potato varieties. Identification by electrophoretic spectra. National registers, appraisal of characteristics, genetic data. Mitt. Biol. Bundesanst. Land Forstwirtsch., Berlin-Dahlem 168, 214. Steiner, A. M., Blender, M., and Geiger, H. H. (1984). Electrophoretic differentiation of rye inbred lines (Secale cereale L.). Z . Pfanzenuechtg. 92, 80-83. Stuber, C. W., and Goodman, M. M. (1983). Allozyme Genotypes for Popular and Historically Important Inbred Lines of Corn. Zea mays L,” USDA-ARS, ARR-S-16. U.S. Govt. Printing Office, Washington, D.C. Stuber, C. W., Wendel, J. F., Goodman, M. M., and Smith, J. 5. C. (1988). “Techniques and Scoring Procedures for Starch Gel Electrophoresis of Enzymes from Maize (Zea mays L.),” Tech. Bull. 286. North Carolina Agric. Res. Serv., Raleigh. Symons, S. J., and Fulcher, R. G. (1988). Determination of wheat kernel morphological variation by digital image analysis. 11. Variation in cultivars of soft white winter wheats. J . Cereal Sci. 8, 219-229. Tanksley, S. D., and Orton, T. J. (1983). “Isozymes in Plant Genetics and Breeding,” Parts A and B. Elsevier, Amsterdam. Torres, A. M., and Bergh, B. 0. (1980). Fruit and leaf isozymes as genetic markers in avocado. J . A m . SOC.Hortic. Sci. 105, 614-619. Troyer, A. F. (1986). United States Patent; Inbred corn line. Patent No. 4,594,810. U.S. Patent Office, Washington, D.C. Troyer, A. F., Openshaw, S. J., and Knittle, K. H. (1988). Measurement of genetic diversity among popular commercial corn hybrids. Crop Sci. 28, 481-485. U.S. District Court (1987). Pioneer Hi-Bred International, Inc., an Iowa Corporation, Plaintiff versus Holden Foundation Seed, Inc. et al., Defendants. U.S. District Court, Southern District of Iowa. Case No. 81-60-E. Vaillancourt, R. E., and Weeden, N. F. (1990). Genetic diversity in the cowpea and its wild relatives. A m . SOC. Agron. Abstr. p. 114. Villamil, C. B., Duell, R. W., Fairbrothers, D. E., and Sadowski, J. (1982). Isoelectric focusing of esterases for fine fescue identification. Crop. Sci. 22, 786-793. Wagner, C. K., and McDonald, M. B. (1981). Identification of soybean [Glycine max (L.) Merrill] cultivars using rapid laboratory techniques. Ohio Agric. Res. Dev. Cent. Res. Bull. 1133. Wall, J. S . , Fey, D. A., Paulis, J. W . , and Landry, J . (1984). Improved two-dimensional electrophoretic separation of zein proteins: Application to study of zein inheritance in corn genotypes. Cereal Chem. 61, 141 -146. Walmsley, R. M., Wikinson, B. M., and Kong, T. H. (1989). Genetic fingerprinting for yeasts. Bio Technology 7, 1168-1 170. Walton, M., and Helentjaris, T. (1987). Application of restriction fragment length polymorphism (RFLP) technology to maize breeding. Proc. 42nd Annu. Corn Sorghum Res. Con& pp. 48-75. Wang, Z . Y . , and Tanksley, S. D. (1989). Restriction fragment length polymorphisms in Oryza sativa L. Genome 32, 1113-1118. Ward, J. (1963). Hierarchical grouping to optimize an objective function. J . A m . Stat. Assoc. 58, 236-244.

140

J. S. C.SMITH AND 0. S. SMITH

Weeden, N. F., and Lamb, R. C. (1985). Identification of apple cultivars by isozyme phenotypes. J. A m . Hortic. Sci. 110, 509-515. Wehner, D. J., Duich, J. M., and Watschke, T. L. (1976). Separation of kentucky bluegrass cultivars using peroxidase isoenzyme banding patterns. Crop Sci. 16,475-480. Weir, B. S . (1990). “Genetic Data Analysis. Methods for Discrete Population Genetic Data.” Sinauer Assoc., Sunderland, Massachusetts. Wilkinson, J. F., and Beard, J. B. (1972). Electrophoretic identification of Agrostis palustris and Poa pratensis cultivars. Crop Sci. 12, 883-834. Wills, A. B., and Wiseman, E. V. (1980). Acid phosphatase isoenzymes of Brassica oleracea seedlings and their application to sib testing in F1hybrids. Ann. Appl. Biol. 94,137-142. Wills, A. B., Fyfe, S. K., and Wiseman, E. M. (1979). Testing F, hybrids of Brassicu oleracea for sibs by seed isoenzyme analysis. Ann. Appl. Biol. 91, 263-270. Wilson, C. M. (1985). Mapping of zein polypeptides after isoelectric focusing on agarose gels. Biochem. Genet. 23, 115-124. Wilson, C. M. (1986). Serial analysis of zein by isoelectric focusing and sodium dodecyl sulfate gel electrophoresis. Plant Physiol. 82, 196-202. Wilson, C. M., Sprague, G. F., and Nelsen, T. C. (1989). Linkages among zein genes determined by isoelectric focusing. Theor. Appl. Genet. 77,217-226. Wolfe, W. H. (1977). Application of isozyme bandings to the identification of cultivars and species of grapevine (Vifis).Ph.D. Dissertation, University of California, Davis (Diss. Abstr. No. DJ77-27420). Woods, S . , and Thurman, D. A. (1976). The use of seed acid phosphatases in the determination of the purity of F, hybrid brussels sprout seed. Euphyfica 25, 707-712. Wrigley, C. W. (1970). Protein mapping by combined gel electrofocusing and electrophoresis: Application to the study of genotypic variations in wheat gliadins. Biochem. Genet. 4, 509-516. Wrigley, C. W., and Shepherd, K. W. (1977). Pedigree investigation using biochemical markers: The wheat cultivar Gabo. Ausf. J. Exp. Agric. Anim. H u b . 17, 1028-1031. Wu, L., Harivondi, A. H., Harding, J. A., and Davis, W. B. (1984). Identification of Kentucky Blue Grass cultivars with esterase and phosphoglucomutase markers. Crop Sci. 24, 763-768. Zawistowski, J., and Howes, N. K. (1990). Production of monoclonal antibodies against high-M, avenin polypeptides. J. Cereal Sci. 12, 235-244. Zeven, A. C. (1990). Classification of landraces and improved cultivars of rivet wheat (Triticum turgidurn) and bread wheat ( T . aestivum) from Great Britain and described in 1934. Euphytica 47, 249-258. Zillman, R. R., and Bushuk, W. (1979a). Wheat cultivar identification by gliadin electrophoregrams. 11. Effects of environmental and experimental factors on the gliadin electrophoregram. Can. J. Plant Sci. 59, 281-286. Zillman, R. R., and Bushuk, W. (1979b). Wheat cultivar identification by gliadin electrophoregrams. 111. Catalogue of electrophoregram formulas of Canadian wheat cultivars. Can. J. Plant Sci. 59, 287-298.

TRANSPORT OF CHEMICALS THROUGH SOIL: MECHANISMS, MODELS, AND FIELDAPPLICATIONS William A. Jury1and Hannes FliihlerZ 'Department of Soil and Environmental Sciences, University of California, Riverside, Riverside, California 92521 %stitute of Terrestrial Ecology, Swiss Federal Institute of Technology (ETH), Zurich, Switzerland

I. Introduction 11. Transport and Transformation Processes A. Mass Conservation Law B. Chemical Phase Concentrations C. Chemical Mass Flux Laws D. Convection-Dispersion Representation E. Transfer Function Models F. Flux and Resident Concentrations G. Interphase Mass Transfer Laws H. Reactions of Chemicals in Soil 111. Analysis of Process Assumptions A. Dispersion Transport Processes B. Solute Adsorption IV. Field Studies of Solute Transport A. Mean Solute Velocity B. Solute Dispersion C. Solute Adsorption D. Preferential Flow E. Scale of Heterogeneity F. Indices of Solute Dispersion V. Concluding Remarks A. Preferential Flow and the Field Regime B. Rate Processes and the Field Regime C. Dispersion Processes and the Field Regime D. Lessons from History References

141 Advnnrer in Agronary, Volume 47 Copyright 0 1992 by Academic Press,Inc. All rights of reproduction in any form reserved.

142

WILLIAM A. JURY AND HANNES FLij'HLER

I. INTRODUCTION Interest in the transport of solutes through soils and aquifers has accelerated in recent years. Although this subject has always been of importance to agronomists concerned with topics such as nutrient movement or soil and groundwater salinization, the mounting worldwide concern with groundwater contamination by toxic chemicals has created a whole new audience for information on chemical movement and reactions in soil. The contamination evidence is rather alarming. Nitrates originating from fertilizers, feedlots, or septic tanks have raised groundwater concentration levels of NO3 above the public health level standard of 45 mg liter-' in countless aquifers around the world (Novotny and Chesters, 1981; Follett and Walker, 1989). Industrial solvents such as trichloroethylene (TCE) were indiscriminately disposed of by pouring waste products into the ground for decades, and now have shown up in groundwater in many locations around the country (Pye and Kelly, 1984). Chemical waste has also entered groundwater through accidental spills or leaks. For example, it has been estimated that as many as 300,000 gasoline storage tanks in the United States may be leaking at the present time (Devitt et al., 1987). A substantial amount of research has been conducted on the transport and transformations of chemicals in soil. However, the overwhelming majority of these studies have been performed in the laboratory, usually in repacked soil columns (Biggar and Nielsen, 1967; Van Genuchten and Cleary, 1979). The laboratory research has produced a large body of information on the transport and reaction processes in homogeneous soil. It is generally agreed that our understanding of the mechanisms governing chemical transport and reaction processes is relatively well advanced at this scale. In contrast to the abundance of data from laboratory studies of chemical movement and reactions, there is little quantitative experimental information available to help in deciphering the dominant transport mechanisms at the field scale. Based on the data from the few studies available, it may be concluded that there are fundamental differences between the transport characteristics of the laboratory and field environments. Under field conditions, soil structure or texture variations cause the transport and retention properties of water and solutes to vary significantly, both in the lateral and vertical directions (Warrick and Nielsen, 1977; Jury, 1985). Moreover, structural voids or other geometric anomalies can create preferential flow pathways that channel water and solutes rapidly through a small portion of the solute transport volume (Beven and Germann, 1982; White, 1985). Consequently, the time scales for mass

TRANSPORT OF CHEMICALS THROUGH SOIL

143

transfer between phases or between regions in which water flows at different rates become dramatically different at the field scale. It follows from the preceding discussion that transformation from the laboratory to the field scales cannot be achieved without considering differences in the degree of heterogeneity and in the characteristic mixing times between the two regimes. A soil column has a lateral dimension that is typically of the order of 10 cm and is 1 m or less in length. Conversely, the root zone of an agricultural field has nearly the same length scale along the direction of flow as a soil column, but its lateral dimension is more of the order of 100 m to 1 km. With such an uneven distortion of the longitudinal and lateral transport scales, it is not surprising that the transport models used in the laboratory should not apply to the field. However, all too often in the past, laboratory models have been used for just this purpose. This review will first provide a description of the chemical transport and reaction processes for compounds moving through soil. Then, the discussion will shift to the presentation of field experimental information and its significance relative to the process descriptions. Finally, the current approaches used for modeling chemical transport in natural media will be presented and critiqued.

II. TRANSPORT AND TRANSFORMATION PROCESSES A chemical transport and transformation model for a particular chemical species embodies the following elements:

1. A mass conservation law for the chemical species. 2. A division of mass into appropriate phases requiring separate description. 3. A flux law for each mobile phase, describing the rate of transport of chemical mass per unit area in that phase. 4. A series of intxphase mass transfer laws describing movement between the phases. 5 . A reaction term describing the rate of appearance or disappearance of mass per unit volume from the system. These elements are characterized with model expressions based on specific process assumptions, which are combined into a chemical transport equation or equations that can be used to calculate phase concentrations and fluxes as a function of space and time. The discussion below will review the

144

WILLIAM A. JURY AND HANNES FLmLER

principal process model descriptions for each element of the complete balance model, and will discuss the physical significance and realm of application of each process representation.

A. MASSCONSERVATION LAW The starting point for a model describing the transport and transformations of a chemical species in soil is the principle of mass conservation. This law is applied to a chemical mass in an arbitrary volume V of soil, and consists of a mathematical statement of all of the places where the chemical can move in the volume, and all of the fates that it can suffer. The law of mass conservation in words is given in Eq. (1). Rate of flow of mass into V - Rate of flow of mass out of V - Rate of increase of mass stored in V - Rate of loss of mass in V by reactions = 0

(1)

The differential equation corresponding to Eq. (1) is developed by applying the mass conservation principle for a short time interval of duration At to a small volume element of size V = Ax Ay Az and then letting the size of the time interval and volume shrink to zero (Jury et al., 1991). In one dimension the differential equation corresponding to Eq. (1) is written as

+ d J , / d z + r, = 0

(2)

where J, is the total chemical mass flux (flow of chemical mass per area per time), r, is the rate of loss of chemical mass per volume (which will be negative for source reactions), and Ciis the total resident chemical concentration (mass of chemical/soil volume). The three-dimensional form of Eq. (2) is similar, except that the chemical flux has x and y components as well. However, the one-dimensional form given in Eq. (2) is also valid for a multidimensional flow process, provided that the lateral boundaries normal to the mean flow direction do not have mass entering or leaving through them. In this case, the quantities J,, r S , and C: in Eq. (2) are averaged over the direction normal to flow. Thus, this equation could be used to describe an area-averaged mass balance for an agricultural field receiving water and chemical inputs over the inflow surface, provided that there was no net lateral inflow or outflow from the transport volume.

B. CHEMICAL PHASE CONCENTRATIONS The total resident chemical concentration C : in Eq. (2) may be divided up into a series of terms representing the chemical mass storage per volume

TRANSPORT OF CHEMICALS THROUGH SOIL

145

in various compartments or chemical phases. The division of chemical mass into phases is arbitrary, but for practical purposes should obey the following constraints:

1. The mass in each phase should behave differently than the mass in other phases. 2. Each phase should have transport and reaction laws that are known or that can be deduced by experiment. 3. The mass in a given phase should if possible be experimentally distinguishable from the mass in other phases. The most common phases into which a soil chemical is divided are gaseous, dissolved, adsorbed, and nonaqueous phase liquid (NAPL). The latter phase refers to a concentrated mass of pure or formulated liquid that does not dissolve completely into the water phase. The principal phase divisions of the total resident concentration are given in Eq. (3)

c:= ac;+ ec;+ p c ; + qc;

(3)

where a is volumetric air content (volume of air per volume of soil), 6 is volumetric water content (volume of water per volume of soil), p is soil dry bulk density (mass of dry soil per volume of soil), r] is the nonaqueous phase liquid volume fraction (volume of NAPL per volume of soil), Ci is the dissolved chemical or solute concentration (mass of solute per volume of water), Ck is the adsorbed chemical concentration (mass of adsorbed chemical per mass of soil), CL is the gaseous chemical concentration (mass of chemical vapor per volume of air), and CL is the mass density of the NAPL liquid (mass of NAPL per volume of NAPL).

C. CHEMICAL MASSFLUXLAWS When the chemical mass flux J, in Eq. (2) is expressed as a function of the phase concentrations of interest in the problem, the expression is called a flux law. A separate flux expression is required for each chemical phase that is mobile in soil. The usual form in which flux laws are expressed is through differential equations that are functions of the phase concentrations. The dissolved phase flux may also be modeled using a different formulation that will be discussed after the differential flux laws are presented. 1. Vapor Flux

The mass flux of vapor in soil is assumed to obey Fick’s law of diffusion, modified to account for the presence of the solid and liquid phases that act

146

WILLIAM A. JURY AND HANNES F L m L E R

as barriers to vapor movement. This modified expression has the general form

acy,

Jg =

- t g ( a ) D ;az

(4)

where Jg is the gas flux, D: is the binary diffusion coefficient of the vapor in air, and (,(a) is a tortuosity factor to account for the reduced crosssectional area and increased path length of a vapor molecule in soil. A variety of gas tortuosity models have been proposed for use in soil (see reviews by Sallam et al., 1984; Collin and Rasmuson, 1988). The model of Millington and Quirk (1961) given in Eq. ( 5 ) t g ( a )= a10’3/6,* (5) where 6, is the porosity, has been shown to represent vapor flow well in soil over a wide range of water contents (Sallam et al., 1984; Farmer et al., 1980). 2. NAPL Flux

The flux of nonaqueous phase liquid in unsaturated soil can only be described by using a theory that simultaneously represents the water and air phases, because the potential energy of the fluid is regulated by interfacial curvature, involving the local geometric configurations reached by the three phases at their interfaces within the soil pores (Van Dam, 1967). Moreover, the flow depends on the relative permeability of the medium to NAPL mass, which is regulated by the presence of the water phase as well as the solid matrix. When these effects are all taken into account, one writes flux equations of the form shown in Eq. (6) (assuming one-dimensional flow for simplicity)

where J, is the flux of phase p (water, NAPL, or air), k, is the relative permeability of the medium to phase p , Ksp is the saturated hydraulic conductivity of phase p in the medium, p, is the density of phase p , p* is the density of the phase used as the reference, and h, is the fluid pressure head of phasep in the soil (Abriola and Pinder, 1985; Parker et al., 1987). The fluid pressure in each phase depends on the configurations that the fluid forms at the interfaces within the soil pores. The three-phase equilibrium pressures can be characterized as a function of the relative saturation of each phase in controlled laboratory experiments. However, the relative permeability k , must be developed from an

TRANSPORT OF CHEMICALS THROUGH SOIL

147

idealized geometric model of the porous medium (Parker et al., 1987), or else related empirically to the two-phase relative permeability of water and NAPL (Stone, 1973). The flux equations for the three phases in soil are combined with mass conservation equations for each phase to produce a transport model (Abriola and Pinder, 1985). The theory of NAPL flow in soil is still in its infancy, and only a few experiments have been performed to test its hypotheses. A significant barrier to progress in modeling NAPL flow in soil is the development of fluid instabilities under a wide range of experimental conditions (Schwille, 1988).

3. Flux of Dissolved Chemical in Soil Unless a compound is very volatile, the dominant mechanism by which it is transported through soil is by bulk movement as a dissolved constituent of the water phase. At the scale of the soil pore, there are two transport mechanisms that can move solutes through the medium: convection and diffusion. Convection refers to the transport of a dissolved chemical by virtue of bulk movement of the host water phase. Thus, the vector of the local convective solute flux v, may be written as vsc = V W C I

(7)

where vw is the local water flux vector and C, is the corresponding local fluid concentration. The local diffusion flux vector Vsd within the water phase is described by Fick's law of diffusion (Bird et al., 1960)

(2

:; 2 )

vsd=-D;hl - i + - j + - k

where 0;" is the binary diffusion coefficient of the solute in water and i, j, and k are unit vectors in the x , y , and z directions, respectively. The total local flux of dissolved chemical vI = v,, + vd describing the movement of solute within the water phase is exact but impractical, because the local water flux describes three-dimensional flow around the solid and gaseous portions of the medium and is not measurable. Instead, the local quantities are volume averaged to produce a larger scale representation of the system properties. The averaging volume must be large enough so that the statistical distribution of geometric obstacles is the same from place to place. If the porous medium contains the same material and density throughout, then the mean value produced by this averaging is macroscopically homogeneous over the new transport volume containing the averaged elements.

148

WILLIAM A. JURY AND HANNES FLUHLER

For example, a quantity such as the porosity c$ is highly irregular at the pore scale, with a value of 0 or 1, depending on whether the point of characterization is within or between the solid particles forming the soil matrix. However, when many pores lie within the averaging volume, the value of the porosity represents the fraction of the volume that is not filled with solid material. This measurement will yield the same value for the porosity when the averaging volume is centered at different locations within the porous medium, provided that the material has been packed in a uniform manner using the same distribution of solid particle sizes and shapes. The minimum size of the averaging volume that yields the same value for the porosity is called its representative elementary volume (Hubbert, 1956; Bear, 1972). If the soil matrix does not have the same properties at different locations, then a representative elementary volume may not exist for that quantity, because the value for its average will then change continually as the size of the volume is altered (Baveye and Sposito, 1985). If we assume that the soil is macroscopically homogeneous, the local dissolved chemical flux can be volume averaged. The new macroscopic convective flux may be written as (assuming that the averaged flow is one dimensional for simplicity).

J,, = -J,c; (9) where now the volume-averaged water flux J , is regarded as a continuum quantity that can be described by the Buckingham-Darcy flux law (Bear, 1972). The volume-averaged diffusion flux is modeled with a tortuosity coefficient similar to the one used in Eq. (4) for the vapor flux

where the dissolved-phase tortuosity factor &(O) may be described with the Millington-Quirk model, Eq. (9,applied to the water phase

t,(e) = e10’3/c$2

(11) However, unlike the situation at the scale of the soil pore, the volumeaveraged solute flux is not the sum of the convection and diffusion fluxessome of the solute motion has been lost in the averaging process and must be reinserted into the flux equation as a new term. When the local water flux is averaged to produce a new continuum quantity, the local threedimensional fluctuations about the average motion due to movement around the air and solid phases are lost because they do not contribute to the mean water flow at the new scale. However, dissolved chemicals convected along these tortuous flow paths do contribute to solute transport

TRANSPORT OF CHEMICALS THROUGH SOIL

149

and cannot be neglected. The motion of solute due to small-scale convective fluctuations about the mean motion is called hydrodynamic dispersion (Bear, 1972). Development of models to describe the hydrodynamic dispersion flux is one of the most active research areas in soil physics and hydrology today (Gelhar, 1986). An appreciation for the physical significance of these models can only be gained after a detailed discussion of the local mixing and transport processes that contribute to the bulk flow of solute by hydrodynamic dispersion. We will illustrate the interplay between lateral mixing and longtitudinal convection using the classic analysis of Taylor (1953) on solute transport through a capillary tube. When a hydrostatic water pressure gradient is placed across a cylindrical capillary tube of radius R that is saturated with water, the water velocity distribution that forms under laminar flow is parabolic

v,(r) = vmax(l- r 2 / R 2 ) where v,,, is the maximum water velocity at the center of the tube (Jury et al., 1991). The area-averaged water velocity (v,) obtained by calculating the water volume per unit area of the entire wetted cylinder is given by (vw>=

p

v,(r)rdrd@

=

2

where CP is the angular coordinate of the circular plane in the cylindrical coordinate system (Arfken, 1985). Thus, the average water flux does not change with position along the axis of the cylinder, even though it varies with radial position within the cross-section. The average water velocity is therefore one dimensional. When a pulse of solute is added to the inlet end of the capillary, the solute at the center of the capillary initially moves ahead of the front. However, random diffusion normal to the direction of flow causes the solute to migrate into regions of slower velocity, and eventually the solute molecule samples all of the different flow paths in the cross-section. The time required for a solute molecule to sample all of these different flow paths is called the transverse mixing time fm of the capillary. The mean time required for the solute that enters the capillary tube at z = 0 to reach a given distance z along the axis of the cylinder is called the convection time tc of the solute. For this problem the convection time t, is equal to z/(vw), and the total mean convection time required to reach the end of the tube is the breakthrough time f b = L/(v,). The area- or flow-averaged solute concentration at a given distance along the tube looks very different depending on whether the mixing time is longer or shorter than the mean convection time required to reach the

lS0

WILLIAM A. JURY AND HANNES F L m L E R

point of observation. In the top part of Fig. 1, the solute pulse near the point of entry has a very short convection time compared to the mixing time, and the mixing process has just started. Therefore, some of the solute at the center of the tube has not diffused into slower channels and is still advancing at nearly twice the average velocity. In the lower part of the figure the pulse has migrated farther along the tube, the mean convection time has increased, and solute mixing has allowed each molecule to explore all of the different flow pathways. As a result, the pulse is spread out along the direction of flow (because individual molecules have spent different amounts of time in a given flow region), and the main pulse is now migrating at the average water velocity (v,) = v,,,/2. Hence, the relative size of the mixing time compared to the convection time determines the radial distribution of the solute as it arrives at a given distance from the point of entry. Although the capillary tube example is an idealized situation, it contains several elements that are important to an understanding of solute dispersion in soil. First, the mixing time of the capillary is regulated by the geometry of the tube. It will be longer for larger capillaries than for smaller ones. Second, the convection time is defined relative to the distance traveled or to the time since the pulse was injected. It does not depend on Casel: t, << ,t

z=L

z=o

Case2: t c >> ,t

z=o

z= L

Figure 1. Schematic illustration of a solute pulse advancing along the axis of a capilIary tube that has a parabolic water velocity distribution in the radial direction.

TRANSPORT OF CHEMICALS THROUGH SOIL

151

the lateral geometry, but rather on the mean convective motion and on the distance traveled. In a porous medium, the local water velocity is not distributed in an ordered parabolic pattern. Rather, it is chaotic because of the complex geometric configurations formed by the solid and air spaces that act as barriers to the water flow. However, at any scale of observation, local velocities are correlated over some distance normal to the flow direction. Velocities immediately adjacent to each other are more likely to be similar than those farther apart. At any scale of transport, there is a characteristic distance of separation, called the correlation length, within which the velocities are similar (Lumley and Panofsky, 1964). When solute has time to explore this zone of similarity, the mixing time is completed, provided that adjacent zones of this size have the same mean flow characteristics. However, if portions of the soil volume have distinctly different transport characteristics, such as extremely rapid water flow, then the area-averaged description of flow will have properties that reflect both the fast and slow regions of the medium.

4. Asymptotic Dispersion Models From this discussion it is clear that there are some representative types of solute flow that apply at different times during the transport event. These so-called asymptotic extremes of solute movement are useful both for classifying the kinds of solute behavior and for developing simple models that are only valid at certain times. For a soil with a unimodal velocity distribution (i.e., a soil having no local pathways that are extremely fast compared to the majority of the water velocities), the two extreme flows of interest are the zero-time and the infinite-time models. a. Zero-Time Model of Solute Dispersion At zero time, just as the solute enters the medium, solute molecules are caught in isolated water flow pathways (called stream tubes) moving at different velocities through the wetted pore space. Because mixing has not yet begun, diffusion or small-scale transverse convection is neglected, and the solute convection along the mean direction of flow is described as parallel flow of solute in the different stream tubes. This is called a stochastic-convective model, because all of the motion is convective but encompases a range of velocities along the direction of motion (Simmons, 1982). It applies approximately whenever the convection time is very much smaller than the mixing time. An average solute flow velocity can be defined in such a system by averaging over the motion in the different

152

WILLIAM A. JURYAND HANNES FLmLER

tubes, and the dispersion or spreading along the flow direction that occurs in the averaged medium reflects those parts of the fluid body that are convected at slower or faster than average velocities. b. Infinite-Time Model of Solute Dispersion The infinite time model of solute dispersion applies when the solute molecules have traveled by diffusion into the zones of different water velocity. Therefore, each molecule has the same mean convective flow rate through the medium. The spreading of solute molecules about this average convective motion is due to the varying amounts of time a given molecule spends in each stream tube; the average motion of the molecules is random or diffusionlike when reviewed in a frame of reference moving at the mean convective velocity. The overall motion is described as the sum of the mean drift motion and the random longitudinal spreading. This transport process representation is called a convection-dispersion model. It applies approximately whenever the mixing time is very much less than the convection time (Taylor, 1953). In a recent field-scale test of solute transport in groundwater, the longitudinal dispersion coefficient reached a constant asymptotic value after about 26 m of travel distance (Garabedian et al., 1991). In another aquifer study, however, the asymptotic limit had not been reached after 90 m of travel (Freyberg, 1986). c. Preferential Flow of Solute A field soil is full of local pathways such as structural voids or biological channels that can carry water at velocities much greater than those of the surrounding matrix, even when the entire field surface is watered uniformly. Moreover, local obstacles in a heterogeneous soil can cause water to funnel into narrow plumes under certain circumstances, even when moving within the soil matrix (Kung, 1990a,b). Also, local variations in the rate of water and solute input at the soil surface can create preferentially wetted pathways even in homogeneous soils. In any of these circumstances, the mixing and convection times of that portion of the field that is experiencing much greater than average flow are very different mixing and convection times than those within the rest of the medium. Consequently, the asymptotic limits apply to one part of the medium may not apply to other parts. In this case, a third type of limiting model is invoked that describes the two parts of the medium with different process representations. Figure 2 shows schematic illustrations of solute particles moving through a medium that could be described with these prototype dispersion models.

TRANSPORT OF CHEMICALS THROUGH SOIL

153

Convective-Dispersive Flow INLET SURFACE

Stochastic-Convective Flow INLET SURFACE

Preferential Flow

.

INLET SURFACE

0

Figure 2. Schematic illustration of three types of solute transport.

5. Stochastic Continuum Model

Modeling solute dispersion with a single continuum model from the time of solute application onward is possible only if one is able to describe both the local water flows and the mechanism of lateral mixing in terms of measurable parameters. At the pore scale, this description is not possible because it depends on the unobservable local pore geometry and the soil structural features. At larger scales a local three-dimensional model is used that has already been volume averaged, so that the local water flux can be calculated with a transport equation that contains measurable parameters (such as the saturated hydraulic conductivity if the water flow is modeled in saturated soil). The volume-averaged local solute transport model is assumed to be convective-dispersive, which amounts to assuming that the solute has sufficient time to mix through the regions of different velocity

154

WILLIAM A. JURY AND HANNES F L m L E R

within the local averaging volume. At the field scale, the local concentration and local transport parameters are assumed to be stationary random functions oscillating about their mean values. A field-scale description of solute transport is then derived by statistically averaging the stochastic local transport model (Gelhar and Axness, 1983), producing a macroscopic model that depends on the statistical properties of the random transport properties. A preliminary model of this type has already been developed for saturated soil (Dagan, 1984, 1987). The local water flux that is reaveraged to create a large-scale model is described by Darcy’s law for saturated flow in two or three dimensions, and the mixing time is characterized in terms of the correlation length scale of the saturated hydraulic conductivity, which is measurable if sufficient numbers of samples can be taken (Sudicky, 1986). This model has shown promise in characterizing the migration of solute through aquifers in a transport study for which sufficient numbers of measurements were taken to validate the theory (Freyberg, 1986; Sposito and Barry, 1987). The water flux is much more difficult to measure in unsaturated soil than it is in saturated soil, and the unsaturated hydraulic conductivity is a function of the saturation state of the soil. Also, flow in groundwater is usually parallel to the direction of stratification, and therefore the mean flow properties along the direction of flow may not change appreciably during the transport event of a practical distance. However, in unsaturated soil, the normal direction of stratification is perpendicular to the direction of flow and the mean soil properties can change significantly over short vertical distances. For these reasons, the stochastic continuum model of chemical transport is more difficult to formulate in unsaturated soil and has not been developed or tested at this time. There have been several stochastic continuum models published for water flow in unsaturated soil (Yeh et al., 1985; Mantaglou and Gelhar, 1987). Having completed this brief introduction to the principles of solute dispersion, we will now return to the discussion of specific models for the solute flux in soil.

D. CONVECTION-DISPERSION REPRESENTATION 1. Convective-Dispersive Flux The convective-dispersive model of the chemical flux in the liquid or dissolved phase J, in one dimension is given by Eq. (14) (Lapidus and

TRANSPORT OF CHEMICALS THROUGH SOIL

155

Amundson, 1952; Nielsen and Biggar, 1962)

where D, is the effective dispersion coefficient, combining the influence of diffusion in the dissolved phase [Eq. (lo)] with the long-time dispersion representation of the effect of small-scale convection on the mixing of solute about the center of motion. As stated above, the dispersion contribution to the solute motion is random or diffusionlike only after sufficient time has elapsed for the solute to mix over the range of solute velocities that are correlated. This mixing time is scale dependent, so that the model Eq. (14) is more likely to be correct after a given time in applications bounded by small transverse distance scales (such as soil columns) than by large ones (such as the field regime). 2. Convection-Dispersion Transport Model

When the only chemical phase of importance in transport and storage is the dissolved phase, Eq. (14) can be combined with the chemical mass balance equation [Eq. (2)] along with the following substitutions:

c:=ec; J, = 51

to produce (assuming r, = 0)

Equation (15) is called the convection-dispersion equation (Biggar and Nielsen, 1967). When the soil water regime is changing over time, Eq. (15) must be solved simultaneously with the water flow equation (Jury ef al., 1991) -+(h)($+ d t dz

l)]

where h is matric potential and K(h) is the unsaturated hydraulic conductivity. Equation (16) is frequently referred to as the Richards equation (Jury et al., 1991). When water flow is steady, then, if there are no sources or sinks of water in the soil, Eq. (15) reduces to the familiar expression

ac; ac; -+v-dt d‘?

a*c;

D--g=O

I56

WILLIAM A. JURY AND HANNES F L m L E R

where V = J W / Ois the pore water velocity and D = D J 6 is the effective diffusion-dispersion coefficient. Equation (17) can be solved analytically for many initial and boundary conditions (Van Genuchten and Alves, 1982). The procedure outlined above wherein differential equations are developed by combining differential flux laws with the mass conservation equation is largely restricted to convective-dispersive processes. The representation of the zero-time asymptotic extreme of stochastic-convective flux is more easily represented within the macroscopic framework of a transfer function model.

E. TRANSFER FUNCTION MODELS A transfer function model is nothing more than an application of the principles of linear superposition and mass balance to a finite soil volume bounded by an entry surface through which solute enters and an exit surface through which it leaves (Jury et al., 1986a). Although a transfer function model can be defined for an arbitrary linear solute process, it is most useful to begin by describing the transport of dissolved chemical that does not enter or leave the interior of the volume by reactions. For this case, the solute transport mechanisms within the soil volume are characterized through a preliminary experiment in which a narrow pulse of solute tracer mass M is added to the inlet end of the volume at time f = 0, and the mass leaving the volume through the exit surface is recorded as a function of time. The outflow record is then given a probabilistic interpretation as follows. If during the time period from ti to ti+l of length At a fraction mj/M of the input mass M crosses the outflow surface, then m.

P(ti+l)-P(tj)=Prob{tj
j=O,.

.., 30

(18)

where P(t) is the cumulative travel time probability distribution function (cdf), describing the probability that a solute added to the surface at t = 0 will cross the outflow surface in a time less than or equal to t (Jury and Roth, 1990). If the processes governing the travel time through the volume are time dependent, as, for example, when the water flow is transient, then Eq. (18) will be a function of the time when the solute is added and will be of, little use except as a formal definition. However, if the water flux through the system is steady, then

-

mi = JIAti Jwcf(ti) Atj

(19)

TRANSPORT OF CHEMICALS THROUGH SOIL

157

where C‘ = J,/J,,, is the flux concentration (Kreft and Zuber, 1978), and the overbar in Eq. (19) refers to the average of Cf between ti and The flux concentration may be thought of as the ratio of the solute mass crossing a unit area to the water volume crossing the area during the same time interval. The effluent concentration from a soil column is therefore a flux concentration. Equations (18) and (19) establish a relationship between the observable flux concentration Cf and the travel time distribution. If we define the travel time probability density function (pdf) f(t) as f ( t ) = lim Af -0

P ( t + At) - P(t) At

then, combining Eqs. (18)-(20), we may write

Thus, the normalized outflow flux concentration of an experiment in which a narrow pulse of solute is added to the inlet end of a transport volume under steady water flow is equal to the travel time pdf of that volume. Using the principle of superposition (valid for any solute transport process that is linear in the concentration), and the solute mass balance principle (assuming that the solute is dissolved and does not gain or lose mass by reactions in the volume), we may write the outflow flux concentration for an arbitrary input flux concentration as Eq. (22) (Jury and Roth, 1990) Cf(L,t) =

id

Cf(0,t - t ’ ) f ( L ,t’) dt’

(22)

where the indices 0 and L have been added to indicate the positions of the entry and exit surfaces of the transport volume, and the travel time pdf is now indexed to indicate the distance over which the transport measurement occurs. Equation (22) may be interpreted as follows. The chemical flux arriving at the outflow at time t contains solute that entered the volume at different times. If solute enters at time t - t’, it has a probability f ( t ’ ) dt’ of having a travel time t’ (i.e., of leaving at time t). The integral sums all of the possible contributions from travel times t‘ between 0 and t. The transfer function framework has some advantages and some disadvantages over differential equation models representing specific process hypotheses. The travel time pdf f(t) records the transfer characteristics of the medium completely, and therefore is valid for all linear solute transport

158

WILLIAM A. JURY AND HANNES F L m L E R

models; it does not require a process model assumption. However, it merely predicts outflow concentrations for a medium that is at a steady flow. It cannot be used to predict the outflow concentrations when the water flow rate is changed, nor can it predict the concentrations at any other distance within or beyond the transport volume without incorporating some hypotheses about the characteristics of the transport process. Jury (1982) extended the model so that it would apply for arbitrary water flow rates by replacing the independent time variable t by the cumulative net applied water I rt

I=

J,dt’ JO

and assuming that the new net applied water pdf f ( I ) would be the same regardless of the water flow rate. In this form, the transfer function model becomes

If this approximation is reasonably valid, then the pdf f ( Z ) becomes a characteristic macroscopic solute transport property of the transport volume, in much the same way that the hydraulic conductivity is a macroscopic water transport property of the volume over which it is measured. Dyson and White (1986) discuss the conditions under which a solute transfer function representation might be independent of the water application rate. Extension of the transfer function model to other distances from the outflow surface requires a model for the transport process. For example, the convective-dispersive flux model, Eq. (14), can be inserted into the mass balance expression [Eq. (2), with r, = 0 and C : = OCf] and solved for an input condition corresponding to a narrow pulse entry of solute. The normalized flux concentration for this model at an arbitrary depth z is then the travel time pdf to depth z. For the convective-dispersive model this (Jury and Roth, 1990) pdf is given by

f(z, t) =

exp[-(z - vt)*/4~t] 2J71.ot

Equation (25) is sometimes called the Fickian pdf (Simmons, 1982). Insertion of Eq. (25) into Eq. (22) [or into Eq. (24) using Eq. (23)] produces a prediction of Cf at any z.

TRANSPORT OF CHEMICALS THROUGH SOIL

159

1. Stochastic-Convective Transfer Function The stochastic-convective, or zero-time, dispersion model is easily created using the travel time formulation. Because solutes are trapped in separate stream tubes that move them downward in isolation from other tubes, then the probability that a solute molecule will reach depth z in a time less than or equal to t must be the same as the probability that it will reach depth L in a time less than or equal to tL/z. Thus P(2, t) = P(L, tL/z)

(26) Moreover, it is clear from Eq. (20) that f ( t ) = dP(t)/dt, and therefore we may write the stochastic-convective travel time pdf (Jury, 1982) as

-"[ (

f ( z ,t) = dP(z't) P L , -tf)] = - f dt dt

(

L , -t f )

(27)

Equation (27) establishes a relation between the pdf at a reference depth L and the pdf at other 2. Therefore, from the single flux concentration outflow measurement that determines f (L, t ) , the transfer function model framework can now incorporate the stochastic-convective hypothesis [Eq. (27)], allowing Eq. (22) or Eq. (24) to predict the flux concentration at different values of 2. 2. Convective-Lognormal Transfer Function (CLT) The relation given for the travel time pdf in Eq. (27) is valid for any outflow shape function that describes the travel time observations. Jury (1982) used the lognormal frequency distribution (Aitcheson and Brown, 1976) to represent f (L, t )

where pL uL are the constant parameters representing the mean and variance of the population of ln(t) . When the stochastic-convective hypothesis, Eq. (27), is applied to Eq. (28), the stochastic-convective lognormal travel time pdf may be written as

160

WILLIAM A. JURY AND HANNES F L m L E R

(Jury and Roth, 1990). Thus, this pdf is a function of two parameters (pLand cL),just as the convective-dispersive Fickian pdf [Eq. (25)] is a function of two parameters (V and D ) . 3. Travel-Time pdfs of Other Process Models

The travel-time pdfs that present other process model hypotheses can be generated in one of two ways. If the process model is described by a differential equation, then the travel time pdf is generated by solving for the flux concentration in response to a narrow pulse input of solute. If the transport process is described by a transfer function, then specific forms for the travel time pdf can be hypothesized for different applications. For example, Utermann et al. (1990) modeled the solute transport process through a structured unsaturated zone of thickness L that had substantial preferential flow by using a bimodal pdf

f(L 0 = ( y f , ( L t ) + (1 - . ) f 2 ( L t ) (30) where fl is the pdf of the fast zone, f2 is the pdf of the slow zone, and a is the fraction of the flow that enters zone 1. This model assumed that there was negligible mass transfer between the two zones during the time of transport.

F. FLUXAND RESIDENTCONCENTRATIONS Because the flux concentration of a dissolved chemical in one dimension is equal to J,/J,, then the mass balance condition for a chemical moving only in the dissolved phase that does not react in the transport volume is given by Eq. (31) (Jury and Roth, 1990)

ac:

-+J,--= at

dCf

at

0

Thus, once the flux concentration has been calculated for a given experimental boundary equation and a given process assumption, the corresponding total resident concentration may be calculated by [assuming that c : ( z , 0) = 01

Note that Eq. (32) requires a process model assumption, because the t dependence of the flux concentration must be known. Equation (31) estab-

TRANSPORT OF CHEMICALS THROUGH SOIL

161

lishes a relation between the total resident concentration and the flux concentration, but does not specify behavior in different phases. However, this equation allows either a convective-dispersive or a stochasticconvective flux pdf to be calibrated from soil core concentration observations if the measured resident concentration is transformed into a flux concentration using the solute mass balance equation in the form given in Eq. (33) (Jury and Roth, 1990)

A separate set of assumptions expressing the extent of mass transfer between the phases is required to describe transport when the chemical is present in more than one distinct form in the soil. This will be illustrated in the next section.

G . INTERPHASE MASS TRANSFER LAWS Thus far, the process model assumptions discussed for predicting the behavior of the flux concentration as a function of depth have only described processes that affect the movement of a tracer passively following the water flow pathways. When the chemical is divided into phases as in Eq. (3), the observed flux concentration at the outflow boundary of the transport volume cannot be extended to other depths without incorporating additional assumptions about the exchange of mass between the portions of the transport volume where water is flowing and other portions where mass may be stored. This additional set of assumptions may be called phase concentration modeling, and will be illustrated in some detail for the solute adsorption process. 1. Solute Adsorption in Soil

When a chemical is present in both the dissolved and adsorbed phases in soil, the total resident solute concentration [Eq. (3)] may be written as

The relationship between the adsorbed and dissolved phase chemical concentrations at equilibrium is called an adsorption isotherm. The three most common adsorption isotherm models are given in Eqs. (35a)-(35c).

162

WILLIAM A. JURY AND HANNES F L W L E R

c;=-c m a x c; Cf+b

Langmuir isotherm

a. Linear Isotherm The simplest representation of adsorption is the linear model [Eq. (35a)], which is characterized by the distribution coefficient Kd . When this relation is inserted into Eq. (34), the result may be written

is called the retardation factor. As seen from where R = 1 + &/6 Eq. (36), R is the ratio between the total mass density and the mass per soil volume OC; in the dissolved phase. Because this relation applies instantaneously at equilibrium, it may be interpreted in probabilistic terms as describing that the adsorbing solute spends 1/R as much time in the dissolved phase as an identical solute that does not adsorb; hence, it will take R times as long to travel a given distance as the mobile one. Therefore, if a sorbing and nonsorbing tracer are added together at the surface, their cumulative travel time probability distribution functions P (Jury and Roth, 1990) are related by Pa(z, t ) = Pm(z, t / R )

(37)

where the subscripts a and m refer to the mobile and adsorbing solutes, respectively. Thus, using Eq. (20), we may develop a relation between the pdfs of the sorbing and mobile tracers by fa(z, t ) = (l/R)fm(z, t / R )

(38)

Equation (38) was derived based on the assumption that the solute adsorbs linearly to equilibrium, and that the sorbing solute travels through the same water flow pathways as the mobile solute. No assumptions about the nature of the transport process were required to derive Eq. (38), and therefore it applies to all linear solute transport models in which the sorbing chemical experiences linear, equilibrium adsorption. It also illustrates that a two-tracer experiment (mobile and sorbing) can be performed to test the linear, equilibrium hypothesis without having to invoke a process model for the transport. The mobile tracer records the flux pdf fm(z,f)and the adsorbed tracer signal is predicted using Eq. (38). The comparison between this prediction and the observed adsorbed tracer signal constitutes a test of the linear equilibrium adsorption process with-

TRANSPORT OF CHEMICALS THROUGH SOIL

163

out having to invoke a process model for the movement of the mobile chemical (Jury and Roth, 1990). The distribution coefficient Kd in the linear model is a function of the soil as well as the compound. To provide a means of characterizing the adsorption of nonionic organic compounds that is independent of site-specific soil conditions, the distribution coefficient is written as the product of the organic carbon partition coefficient KO,and the organic carbon fraction fOc (39) This formulation has been shown to reduce the coefficient of variation of the organic carbon adsorption coefficient KO,significantly between soils for a given compound compared to the variation of the distribution coefficient Kd (Hamaker and Thompson, 1972). Kd

=foc

b. Freundlich Isotherm The adsorption isotherms of many organic compounds measured in the laboratory over a range of concentrations show deviations from linearity. The Freundlich adsorption isotherm in Eq. (35b) describes the adsorption of many nonionic organic compounds over a large range of concentrations. The power coefficient 1/N is a constant that is generally less than unity. Table I, adapted from Hamaker and Thompson (1972), gives the average value of the Freundlich power coefficient 1/N of the isotherms of some common pesticides characterized in different soils. Because the partitioning between the sorbed and dissolved phases is concentration dependent in a Freundlich isotherm, the retardation formula, Eq. (38), is not applicable. However, because the & coefficient in the formula for the retardation factor is equal to the slope of the adsorption Table I Nonlinear Freundlich Adsorption Power Law Coefficients"

1l N

Pesticide ~

Simazine Atrazine Prometon Prometryn Monuron Parathion

~

0.82 0.71 0.81 0.86 0.77 1.03

CV(%)

Soils

~~

4

11 2 1 7 2

6 19 4 4

LO 4

'l/N for various pesticides, averaged over different soils. Adapted from Hamaker and Thompson (1972).

164

WILLIAM A. JURY AND HANNES FLmLER

isotherm, we can interpret a Freundlich adsorption process as one in which compounds are more strongly retarded at low concentrations than at high ones. Thus, the main pulse is slowed down to an extent corresponding to the retardation from the average slope of the isotherm over the concentration range spanned in the transport event (Rao and Davidson, 1979). c. Langmuir Isotherm The Langmuir isotherm in Eq. (3%) exactly describes the monolayer adsorption of a pure substance to a homogeneous sorbing surface with a finite number of sorption sites (Sposito, 1981). Therefore, it more appropriately describes the sorption of positively charged solutes to soil mineral exchange sites than the bonding of nonelectrolytes to organic matter. For that reason, it is rarely used as model for pesticides or other toxic organics in soil, except for the few compounds, such as paraquat, that are positively charged in the normal pH range. d. Hysteretic Adsorption-Desorption The adsorption isotherms discussed above all assume reversible adsorption and desorption reactions. Experimental measurements of adsorption in the laboratory have shown significant hysteresis in the adsorptiondesorption isotherms of organic compounds (Van Genuchten ef al., 1974). For a hysteretic adsorption-desorption process, the partitioning between the phases is a function of the history of their interactions as well as the current concentrations. Thus, the equilibrium value between dissolved and adsorbed concentrations is different for a desorption process than for an adsorption process, and the compound is more difficult to remove from an adsorption site than it was to deposit it on the site. Models of adsorption-desorption hysteresis tend to be derived from curve fitting to the experimental data. Van Genuchten er al. (1974) used a Freundlich model, Eq. (35b), to describe the adsorption-desorption process, with one value of the power law parameter 1/N for the adsorption loop and another for the desorption loop. The desorption value of the Freundlich coefficient KF was a function of the concentration of adsorbed species. e. Rate-Limited Sorption To this point, the discussion has been confined to transport events in which the solute partitions instantaneously to equilibrium with the sorbed phase at all points in the medium. This might be a reasonable model for describing transport and sorption in homogeneously packed soiIs at low water flow rates, but may not describe the processes adequately in field soils, wherein many of the sorption sites are contained within parts of the soil that are excluded from direct contact with the solute flowing in the

TRANSPORT OF CHEMICALS THROUGH SOIL

165

mobile fluid phase (Valocchi, 1985, 1986; Brusseau and Rao, 1990). In such cases, the local exchange of mass is likely to be described better by a kinetic mass transfer model of the form p ( d C f i / d t ) = aI[F(c;)- Cfi]

(40) where CL = F(Cf) formally represents the equilibrium adsorption isotherm relation [e.g., any of Eqs. (35a)-(35c)] and cq is a rate coefficient. A summary of various types of rate-limited adsorption model is given in Van Genuchten and Cleary (1979). 2. NAPL Dissolution

When a nonaqueous phase liquid is present simultaneously with water in a porous medium, dissolution of the NAPL into the liquid phase will occur. The rate of dissolution is driven by the concentration difference, and the rate coefficient is a function of the geometry of the NAPL-water interface (Miller et al., 1990). To date, the overall dissolution process from a large NAPL body into the surrounding water phase has been represented only through simple engineering transfer coefficient models (Pfannkuch, 1984; Mackay, 1987), or mechanistically for uniform initial concentrations and soil conditions in the laboratory (Miller et af., 1990). Modeling of NAPL transport and transformations in field soil is poorly developed at present, chiefly because of inadequate experimental information. 3. Vapor-Dissolved Phase Partitioning

The partitioning between dissolved and vapor phases for a moderately soluble compound is generally described through the equilibrium relation known as Henry’s law, which in dimensionless form is given by

c; = K H c;

(41) where KH is the dimensionless Henry’s constant (Jury el al., 1983). In models describing transport of organic compounds having substantial mass storage in both the vapor and dissolved phases, Henry’s law is generally assumed to apply at all times, i.e., the phases are assumed to be in equilibrium, and the relationship is assumed to apply over the entire concentration range (Spencer and Cliath, 1970). Thus Henry’s constant may be calculated from

Cg = KHCT

(42) where the asterisk refers to the saturated state of the phase. Because both vapor density and solubility are temperature dependent, KH is also a

166

WILLWM A. JURY AND HANNES FL-LER

function of temperature. However, for an isothermal simulation, solution concentrations in the two phases are assumed to be proportional and one of them may be eliminated from the transport equations using Eq. (42).

4. Stagnant Water Phase Models In soil with natural structure, not all of the water phase is active in the transport of solute. As a result, representation of the dissolved phase with a single concentration Cf averaged over the entire volumetric water content 8 does not take into account the diffusion of solute into regions containing stagnant water, which does not move when water is flowing through other parts of the soil. To account for this effect, models have been proposed that divide the dissolved phase into a mobile and an immobile region. For nonvolatile, nonadsorbing solutes, this partitioning can be written as

c:= e,ck + eimc;,

(43)

(Coats and Smith, 1956; Van Genuchten and Wierenga, 1976, 1977), where 8, is the mobile water content, Oi, = 0 - 0, is the immobile water content, Ck is the average solute concentration in the mobile dissolved phase, and C:,,,is the average solute concentration in the immobile dissolved phase. The water-phase division model Eq. (43) is combined with a rate-limited mass transfer model to describe the transport of solute mass by diffusion from one water zone to the other 8,,(ac;m/at) = .,[C:,

- C,fm]

J , = - e,o,(acyaz)

+ J,c:,

(44) where a , is the rate coefficient describing the mass transfer process between the mobile and immobile regions. In the studies cited above, the mobile-immobile water model described by Eqs. (43) and (44) was combined with a convection-dispersion flux model for the mobile region

(45)

where D, is the effective diffusion and dispersion coefficient for the mobile zone. The flux Eq. (45)is inserted into the solute mass balance Eq. (2) (with r , = O ) to produce a transport equation for a dissolved chemical moving through a medium with mobile and stagnant water phases (Coats and Smith, 1956; Van Genuchten and Wierenga, 1976). A similar model was developed by Rao et al. (1980a,b) specifically for the case where the stagnant region consists of the water within spherical aggregates accessible to solute only by radial diffusion out of the mobile fluid. Addiscott (1977) has also developed a simple field model in which the immobile water

TRWSPORT OF CHEMICALS THROUGH SOIL

167

fraction is determined as a single field-averaged quantity. A similar approach was taken by White (1987), who measured the effective solute transport volume for a mole-drained soil and used this parameter in a transfer function to predict nitrate leaching to the drain during a season of rain fall. 5. Dual Velocity Models

The widespread occurrence of preferential flow in natural field soils has created problems for simple transport descriptions that assume only a single water flow velocity in the soil matrix. Although the mobile-immobile water model discussed above offers a reasonable conceptual framework for transport in aggregated media, it is unsatisfactory for representing media wherein transport occurs both in the preferential flow region and in the bulk matrix. This situation has been represented in the past with a two-region convection-dispersion model, in which mass is transported from one region to the other in response to differences in the resident fluid concentration. Approximate solutions to this problem have been obtained by several authors (Skopp et al., 1981; Hill, 1981; Van Duin and van der Zee. 1986), and the general analytic solution has been derived recently (WalkerJ987). However, the solution is cumbersome and contains a number of parameters that cannot be evaluated independently. An alternative formulation of the same problem, based on particle tracking by computer, has been developed by Roth et af. (1990). Although these models are not suitable for actual simulation because of the large number of parameters, they are useful for analyzing process assumptions made about specific field regimes (Roth et af., 1990). When the mass transfer between the two regions can be neglected during the time of transport through the faster region, the transport problem reduces to parallel flow and has a much simpler form and fewer parameters to evaluate (Utermann et al.. 1990). The general transport description can also be simplified by allowing solute to decay from the slow zone to the fast one (Roth et af., 1991).

Chemicals in solution in the soil can undergo a variety of reactions. Inorganic compounds can be subject to precipitation reactions that remove them from solution if their solubility product is exceeded in the soil. Prediction of chemical precipitation reactions requires calculation of the relevant ion activities, which in turn are a function of the composition of

168

WILLIAM A. JURY AND H A " E S F L m L E R

the solution. Chemical equilibrium concentrations of aqueous solutions in the presence of defined solid phases can be computed by models such as GEOCHEM (Sposito and Mattigold, 1980). Estimation of organic-inorganic reactions is more difficult, requiring a model for the dissolved organic constituents in the soil solution. Organic chemicals are subject to transformation by both chemical and biological interactions in the soil. These reactions are specific to the structure of the organic molecule and do not affect all species equally.

1. Nonbiological Transformations The major nonbiological transformations of organic chemicals are hydrolysis, oxidation-reduction, and photodecomposition (Valentine, 1986). Photodecomposition is insignificant below the top few centimeters of soil and primarily affects the breakdown of compounds during their application or after they are volatilized from the soil. Hydrolysis can be a significant factor in the breakdown of certain organic compounds, and may be the dominant one below the surface zone if biological activity is limited to the shallow soil zone. A hydrolysis reaction usually is represented as a first-order rate process whose rate coefficient depends strongly on pH and temperature (Valentine, 1986). Although certain compounds are quite susceptible to hydrolysis, many experience negligible breakdown by this reaction compared to microbiological degradation (Hamaker, 1972). The relative importance of redox reactions for organic compounds in soil is not understood quantitatively.

2. Biological Transformations Biological decomposition of organic compounds is affected by many soil factors, such as temperature, water content, and organic carbon, and by various biological factors, such as microbial species and population density. When microbial biomass is not limiting, it is common to express the biological breakdown of a mass M of the compared as a first-order reaction d M / d t = -pM

where p is a rate coefficient that is related to the half-life T by T = ln(2)/p

(47)

In laboratory experiments the disappearance of a compound is often found to occur at a rate that differs from the one predicted by a first-order rate law. This has led to the use of different process submodels for the description of microbial transformations. An example of this is the socalled second-order model used when the microbial population is limiting

TRANSPORT OF CHEMICALS THROUGH SOIL

169

to the breakdown process d M / d t = (E/Eo)pM (Jury et al., 1987), where E / E o is the fractional microbial population density relative to the optimum Eo required for a macroscopic first-order rate process to occur. Field measurements of declining microbial population with depth below the soil surface have been reported by Focht and Joseph (1973).

m. ANALYSIS OF PROCESS ASSUMPTIONS It is obvious from the above discussion that there are significant differences between various model hypotheses for the transport and transformations of chemicals in soil. Because of the large number of parameters in complex simulation models, validation of a specific process hypothesis is not always straightforward (Wierenga and Batchelet, 1988). Experimental observations are infrequent, and often consist only of effluent concentration-time curves from laboratory column transport experiments or concentration-depth curves from field cores. Usually, complex models with many parameters can achieve good agreement with outflow or depth concentration observations if parameters are adjusted to the data. However, this agreement does not constitute a “test” of the model hypotheses, but merely demonstrates the versatility of the model to describe the shapes of concentration-time or concentration-depth curves. This point was illustrated convincingly by Nkedi-Kizza et al. (1984), who showed that two distinct models of the adsorption and transport processes based on different process assumptions had the same mathematical form. Thus, their optimum fit to any single concentration-time or concentration-depth record will be identical, even though the transport and sorption processes in the two models are distinctly different. In this section the major process assumptions will be examined and the critical features that distinguish various models for describing them will be discussed. A discussion of the types of observation that must be made to evaluate a process hypothesis will also be made where appropriate.

A. DISPERSION TRANSPORT PROCESSES Figure 3, adapted from Jury and Roth (1990), shows the travel time pdfs for the CDE Eq. (25), and CLT Eq. (28) models, with their parameters optimized for maximum agreement at z = 50 cm. Note that the model

WILLIAM A. JURYAND HANNES FLUHLER

170

z 0

k

2=50 crn

I-

$z 8

$i3 0

50

100

150

200

250

300

200

250

300

TIME (days)

-

0

50

100

150

TIME (days) Figure 3. Normalized solute outflow concentrations predicted at 150-cm depths by the CDE and CLT models after common calibration at 2 = 50 cm. Reproduced with permission, from Jury and Roth (1990).

shapes are virtually indistinguishable when the two functions are fitted to each other, or to a single outflow record, even though they model the dispersion process in completely different ways. This demonstrates that a soil column experiment monitored only at the effluent end cannot be used to test the validity of a dispersion model. However, the travel time pdfs of these two models calibrated at 50cm have very different shapes at z = 150 cm, even though the mean breakthrough times of each model concentration pulse are still the same. Therefore, the appropriate type of experiment to evaluate a dispersion hypothesis is one in which the spreading rate is studied as a function of the distance from the inlet end.

TRANSPORT OF CHEMICALS THROUGH SOIL

171

This procedure was followed in the study of Khan and Jury (1990), who measured solute outflow concentrations in repacked and undisturbed soil columns 90, 45, and 22.5 cm in length (created by cutting off part of the column at the end of a series of experiments) at several column widths and flow rates. Their study showed that the CDE description of dispersion [Eq. 251 was generally valid in the repacked soil, but only at the lowest flow rate in the undisturbed columns. At higher flow rates in the undisturbed soil, the apparent dispersion coefficient increased with distance from the source, as it would for a stochastic-convective process.

B. SOLUTE ADSORPTION The effects that solute adsorption can have on solute concentrations in soil can be somewhat subtle, depending on the details of the process hypotheses. Some of the principal findings of researchers studying adsorptses will be summarized below for the main adsorption representations introduced in Section II,G,l. 1. Nonlinear Adsorption Models

a. Freundlich Adsorption The Freundlich isotherm shape, Eq. (35b), is a common form for the equilibrium partitioning of many organic chemicals. The major effects that this isotherm shape can have on transport compared to the simple linear model, Eq. (38), are to alter the symmetry of the spreading process and to make the travel time of a pulse or front concentration dependent. In a simulation study, Van Genuchten and Cleary (1979) demonstrated the effect of nonlinearity on concentration pulse shape by using Freundlich adsorption isotherms, Eq. (35b), with different powers of 1 / N . Using a numerical model that assumed equilibrium adsorption between dissolved and adsorbed phases, they predicted the concentration-depth curves of a pulse of solute whose adsorption was described by Freundlich isotherms that agreed at the initial pulse concentration, but that had Freundlich power coefficients 1/N ranging from 0.4 to 1.0. Their results showed that within the same concentration range, solute undergoing nonlinear adsorption caused the solute pulse to be more asymmetric than the one representing a solute undergoing linear adsorption. For example, with 1/N = 0.4, the partitioning to the sorbed phase is higher at low concentrations, which restricts the compound from penetrating in the forward direction in the soil and retains it more in the surface layers than when N = 1, in which case the pesticide partitions equally at all concentrations.

172

WILLIAM A. JURY AND HANNES F L m L E R

Nonlinear adsorption can have significant effects on mean travel time, as was demonstrated by Rao and Davidson (1979). These authors measured concentration outflow curves for the pesticide 2,4-D added at 50 and 5000 g ml-'. If the pesticide was linearly adsorbed to equilibrium, then the two outflow curves would have been identical. In contrast 2,4-D, which had a nonlinear Freundlich isotherm in the soil in which the experiments were performed (1/N = 0.7), had much greater mobility per unit of water applied at the higher concentration than at the lower one, arriving at the outflow end after about 2 and 6 pore volumes, respectively. This result illustrates that one should be careful to use linearized forms of nonlinear isotherms that represent the actual concentration range to be encountered in the soil, rather than to use standard linear forms of the isotherm from the literature, which might have been measured at significantly different concentrations. As shown in Table I, nonlinear isotherms are common for pesticides, which indicates that this procedure should be generally followed when simulating pesticide movement. Even when correctly linearized isotherms are used, however, the resulting model will not reproduce the assymmetry of the breakthrough curve (Jury and Ghodrati, 1989). b. Hysteretic Adsorption Models Because of the extreme difficulty in simulating hysteresis in adsorptiondesorption reactions, very little information has been obtained about the importance of this process on transport of solutes. Van Genuchten et al. (1974) used a Freundiich adsorption-desorption model with different parameters for the two cases to describe hysteresis in simulating the transport of picloram in a soil column. By comparing the predicted transport to a model using a single Freundlich isotherm, they were able to show that the main feature of hysteresis is to retard desorption, producing a longer effluent tail on a column breakthrough curve. Unfortunately, this effect is also characteristic of rate-limited nonequilibrium adsorption, so that the importance of hysteresis is difficult to establish experimentally (Van Genuchten et al., 1977). 2. Rate-Limited Adsorption

Figure 4, taken from Jury and Roth (1990), shows the calculated outflow curve for a chemical undergoing rate-limited linear adsorption (with R = 3) according to the mass transfer model, Eq. (40), as a function of the value for the rate constant a. For extremely small values of a, the compound moves as if it is completely mobile, except for a small amount of tailing (that persists to very long times) due to the limited sorption and subsequent desorption that occurred during the residence time of the molecules

TRANSPORT OF CHEMICALS THROUGH SOIL z

0 F

1.

173

= aL/V =0.005

a

[r

I-

0.

0

0.

s3 0. 0.

0

0

1

2

3

4

5

6

7

8

DIMENSIONLESSTIME T =W/L

Figure 4. Outflow concentrations for a solute undergoing rate-limited adsorption as a function of the value for the rate constant a. Reproduced with permission, from Jury and Roth (1990).

in the transport volume. When the rate coefficient a is large, the compound nearly adsorbs to equilibzium, and the breakthrough is delayed nearly proportionally to the retardation factor R = 3. At an intermediate value of the rate coefficient, the solute pulse shows both early breakthrough and pronounced tailing.

3. Stagnant Water Phase Models The division of the liquid pore volume into a mobile region and an immobile region is at best an approximate representation of the true state of nature. This framework is likely to be most successful in describing transport through a medium containing distinct aggregates in a narrow size range that hold water within small pores. For such a system, the interaggregate transport may be approximated by a single flow velocity if the aggregates are packed into a uniform configuration, and diffusive transport into the aggregates can be modeled either with an explicit, radial diffusion sink term representing flow into the aggregates (Rao et al., 1980a,b), or by representing the transfer into the immobile region with a mass transfer coefficient (Coats and Smith, 1956; Van Genuchten and Wierenga, 1976). These two approaches are not equivalent (Rao et af.,1980a), but the mass transfer coefficient can be adjusted to produce nearly the same mass exchange properties by analyzing the geometry of the region between the two water phases (Van Genuchten, 1985; Villermaux, 1987). When the aggregates have a distribution of pore sizes, the exchange process becomes more smeared out than for the case of a single aggregate geometry. This effect can be modeled if the distribution function is

174

WILLIAM A. JURY AND HANNES FLUHLER

known (Moharir et al., 1980; Rasmuson, 1985). An excellent review of mass transfer models and their characteristics is presented in the article of Brusseau and Rao (1990). The mobile-immobile water models described above have been used successfully to model solute transport in aggregated media prepared with a specified geometric configuration (Rao et al., 1980ab), and have been used to interpret rate-limited solute adsorption in a groundwater experiment in an aquifer containing relatively homogeneous sediments (Goltz and Roberts, 1986). However, other explanations based on completely different assumptions can produce similar effects when applied to the same problem. Nkedi-Kizza et al. (1984) showed that a model with an immobile and mobile water fraction assuming instantaneous sorption produced a mathematical equation with the same form as that of model wtih uniformly mobile water carrying solute that adsorbed on two types of sites, one of which was limited by diffusive mass transfer. Kabala and Sposito (1991) reanalyzed the aquifer problem studied by Goltz and Roberts (1986) with a stochastic continuum model, and demonstrated that the effective retardation factor was time dependent when equilibrium adsorption was assumed to occur on adsorption sites that were spatially distributed and correlated with the spatial variations of the water flow velocity. Thus, there are alternate mechanisms capable of reproducing a solute transport event that may not be resolvable with the given experimental information. An additional limitation to applying stagnant water-phase models to the field regime is the possibility that water partitions into a number of distinct phases having different relative motion under the influence of an external driving force. In such a circumstances, there will be many rate coefficients operating locally, which when integrated up to the field scale by volume averaging will have transport properties that differ from those of a model with a single rate coefficient between a mobile and stagnant phase. An extreme case of such behavior is encountered when the solute partitions into a rapid or preferential flow region and a slower but still mobile matrix flow region, each of which may embody a smaller but still significant degree of water flow variability (Roth et al., 1991). There are also significant complications introduced for model identification when more than one rate process is limiting solute movement and reactions in soil (Brusseau et al., 1989).

4. Preferential Flow of Solute The mechanisms that can create preferential flow are understood for the most part, but cannot be predicted from an a priori analysis of the field characteristics. Preferential flow may arise from true fluid instabilities

TRANSPORT OF CHEMICALS THROUGH SOIL

175

created by density or viscosity differences between the resident and invading fluids (Krupp and Elrick, 1969; see review by Hillel and Baker, 1988). It can be induced by structural voids that have extremely high permeability when filled with water (Kissel et af., 1974; Beven and Germann, 1982). It can develop at the interface between two soil layers of different permeability (Hill and Parlange, 1972). It can also occur in sandy soils having high matrix permeability, and which also contain discrete coarse or fine lenses of porous material. When water moves through such soils, these lenses can act as barriers to downward flow and cause focusing or funneling of water (Kung, 1990a,b). White (1985) reviewed a number of studies of water and solute transport through soil containing a pronounced macrostructure. Although each of these situations results in extremely rapid transport through a portion of the wetted cross-section, they have very different attributes and cannot be described by a single process hypothesis. Structural voids can cause preferential flow if they transmit water, even if the soil water flux is spatially uniform. The physical properties of the preferential flow region in this case are very different from those of the surrounding matrix, creating extremely high velocities when the flow channels are saturated. Moreover, sorbing chemicals moving in solution in a water-filled structural void may largely bypass the reduced number of adsorption sites. In fine-textured soils containing structural voids, the permeability of the surrounding matrix may be so low that virtually all of the convective transport occurs within the void region (Beven and Germann, 1982). In this case, the matrix acts as a stagnant water zone through which solute may move only by diffusion. The situation is quite different when preferential flow is caused by lateral movement and funneling of water flow into smaller cross-sections, or when fluid instabilities produce fingering along the direction of flow. Here the physical properties of the preferential flow region may be quite similar to those of the surrounding matrix; only the flow velocity is different. Moreover, substantial water flow may be occurring in the surrounding matrix as well as in the preferential flow region; consequently, the transport problem cannot be characterized in terms of a mobile and an immobile water region, but must include a description of transport through each zone as well as mass transfer between them.

lV.FIELD STUDIES OF SOLUTE TRANSPORT Despite the abundance of agricultural and environmental applications for solute transport modeling, very little field-scale research on transport has occurred until recently. In the few comprehensive experiments that

176

WILLIAM A. JURY AND HANNES FLUHLER

have been performed over a large surface area, only the most basic of the process assumptions can be evaluated. Several of these are discussed below.

A. MEAN SOLUTE VELOCITY When a mobile anionic tracer such as bromide or chloride is added as a pulse to the surface of a field and is subsequently leached by irrigation or rainfall, the mean velocity V , of the area-averaged pulse can be estimated by monitoring the downward movement of the chemical with solution sampling or soil coring. If the field is receiving water at a net flow rate q , then the observed field-mean velocity should be comparable to the socalled piston flow velocity V, = q / 8 , where 8 is the mean volumetric water content, provided that all of the water is participating in the transport. However, field studies that have made this comparison have come to somewhat contradictory conclusions, as shown in Table 11. The study by Biggar and Nielsen (1976) took place on 20 continuously ponded plots located within a 150-ha field area. The chloride tracer was added to each plot as a pulse and was monitored by solution samplers to 1.8 m. The measured solute velocity was calculated by fitting the breakthrough curve from a given sampler to the solution of the convection dispersion model. The 20 calculated solute velocities averaged over all of the samplers at a given site compared favorably with the piston flow velocity (slope = 1.09; r 2 = 0.84) estimated from the steady infiltration rate and the saturated water content of each site. The studies by Butters et ai. (1989) and Ellsworth et aZ(l991) took place on the same field at different times. Butters’ study involved applying a bromide pulse to the entire surface and leaching it downward by spatially uniform bidaily sprinkler irrigation, while monitoring the pulse with solution samplers in a 4 x 4 grid at different depths between 0.3 and 4.5 m. He calculated the solute velocity as the ratio of the depth of observation to the mean solute pulse arrival time, and found it to be substantially less than the net applied water flux divided by the volumetric water content in the top 1.8 m, below which the estimates agreed. When the arrival time of the peak concentration rather than the mean arrival time was used to estimate the solute velocity, the agreement with piston flow was much closer, but still was about 30% slower near the surface. The study by Ellsworth et aZ(1991) involved leaching massive plumes of solute under bidaily irrigation, after the plumes had been injected in the soil slowly in an approximate cubic configuration 1.5 or 2.0 m on a side. The plumes were sampled by soil coring at various times after application

Table II Published Comparisons between the Measured Average Solute Velocity V, and That Predicted from Piston Flow V, Water application Reference

Soil

Method

Rate (cm day-')

Sampling method

Study area (m2)

Biggar and Nielsen (1976)

Clay loam

Butters er al. (1989)

Loamy sand

Ellsworth et al. (1991)

Loamy sand

Jaynes et al. (1988)

Sandy loam

Steady ponding Bidaily sprinkler Trickler , sprinkler Ponding

Rice er al. (1988)

Sandy loam

Ponding

Starr ef al. (1978)

Layered sandy loam Loam

Ponding

47

Solution

Sprinkler

0.25-2.5

Soil cores

1.5 X 1.5 (14 plots) 1.8 X 1.8 (4 plots) 24.4 x 18.3 (14 plots) 4.6 X 6.1 (4 plots) 3x3

Layered sandy loam

Steady tricklers

Solution

8x8

Wild and Babiker (1976) Van der Pol et al. (1977)

14.6

Solution

1.1

Solution

1.1

Soil cores

35 1.4

2.0

Solution Soil cores

6.1 x 6.1 (20 plots) 80 X 80

Depth V,lV,

(4

1.0

0.3-1.8

1.6- 1.9 1.0 1.o

0.3-1.8 3.0-4.5

0.6-0.7 1.o 0.2

0-0.6 1.0-3.0 0-2.0

0-6

2.1

0-0.6

1.o

0-0.8

1.o

0-1.5

178

WILLIAM A. JURY AND HANNES FL-LER

in an irregular grid encompassing the plot boundaries. The solute velocity was estimated by the location of the center of mass of the plume, and agreed almost perfectly with the local piston flow estimate calculated from the local water application rate and the local volumetric water content profile. The studies of Rice et al. (1988) and Jaynes et al. (1988) were performed on the same field, but different relationships were observed between V, and Vp. In a bromide leaching study under intermittent ponding, Rice et al. (1988) measured a solute velocity that was almost five times greater than the one predicted by piston flow. In contrast, Jaynes et al. (1988) observed a solute velocity that was only moderately higher than the piston flow prediction in the surface 0.6 m, and found them to be similar below that depth. They postulated that soil cracking was responsible for the difference that they observed between the two studies. The study of Starr et al. (1978) was conducted under ponding in a layered soil. They calculated the solute velocity from the arrival time of the solute peak at a solution sampler location, finding that this measurement was only about half as large as the piston flow prediction. The authors noted that very early arrival was detected at deep samplers, indicating that solute was moving also by preferential flow. In contrast, Van der Pol et al. (1977) found good agreement between piston flow and solute velocity calculated from fitting solution sampler observations to the convection dispersion model in a layered clay soil irrigated by tricklers. Wild and Babiker (1976) also found good agreement between the two estimates when they calculated solute velocity from the mean location of the solute pulse in a soil core. When they used the location of the peak to estimate V,, however, it was less than half of the piston flow prediction. The research studies summarized above illustrate how the complexity of the field regime can obscure the identification of even the most basic of solute transport processes, the mean convection rate. In those cases in which piston flow underestimates the solute velocity, the most likely reason appears to be that part of the wetted pore space is not active in transport. This would not affect the mean arrival of a solute pulse viewed by soil samplers, because the early arrival would be offfset by the late tailing of an outflow curve, provided that all of the solute was observed as it passed the depth of observation (Jury and Sposito, 1985). However, it would affect the mean position of a resident concentration pulse viewed by soil coring (Jury and Roth, 1990). Moreover, the solute peak velocity of the flux concentration viewed by solution samplers is much faster than the mean velocity of the pulse (as found by Butters et al., 1989), whereas the peak velocity of the resident concentration pulse can be much slower than the center of mass velocity (as found by Wild and Babiker, 1976). Therefore, at least part of the confusion between the solute velocity and the piston

TRANSPORT OF CHEMICALS THROUGH SOIL

179

flow velocitv relates to the frame of viewing the pulse (flux or resident), and part derives from the way in which solute velocity is defined. When the area-averaged solute pulse arrives later than predicted by piston flow, one explanation offered for the disagreement between the two velocities is that part of the water flux is moving through preferential flow channels that are not observed by shallow solution samplers, as suggested by Starr et al. (1978) and Roth et al. (1991). In this case the piston flow velocity calculated by assuming that all of the water enters the matrix flow region would lead to an overestimate. This effect could account for the difference in behavior observed on the same field by Butters et al. (1989) and Ellsworth et al. (1991), because this field is known to experience substantial preferential flow in the surface zone (Ghodrati and Jury, 1990). It is possible that the shallow preferential flow was not detected by the solution samplers used by Butters et al. (1989), but was detected by the soil coring method used by Ellsworth et al. (1991).

B. SOLUTE DISPERSION With the exception of the study of Ellsworth and Jury (1991a), the transverse dispersion of a solute plume has not been monitored at the field scale in the unsaturated zone. Therefore, the discussion below will focus on the vertical or longitudinal spreading of an area-averaged solute pulse after application to the surface of a field. As noted above, the dispersion mechanism can only be revealed by monitoring solute spreading at progressive times by soil coring, or at progressive distances by solution sampling. A large-scale study of solute dispersion was conducted by Butters and Jury (1989) on a 0.64-ha alluvial loamy sand soil. A pulse of solution containing KBr was added by sprinkler irrigation to the field surface and was leached downward thereafter by irrigating every other day. The field had been placed in an initial state of quasi-steady water flux by several weeks of bidaily irrigation at the same net flux rate ( = l cm day-') as the flux during the experiment. Solution samplers were located in a 4 x 4 grid at six depths from 0.3 to 3.0 m, and six additional samplers were placed at 4.5 m. The samplers (16 or 6) at a given depth were averaged to produce a one-dimensional concentration representing the entire field area as a function of depth and time. The dispersion hypothesis was tested by fitting the model parameters of the CDE [Eq. (25)]and CLT [Eq. (28)] solutions to the area-averaged solute concentration at the shallowest (0.3 m) depth, and then comparing the predictions of the models (without further parameter adjustment) with the observed concentrations at greater depths. Figure 5 shows the comparison at 1.2, 1.8, and 3.0 m, illustrating

0

20

40

60

80

100

Net Applied Water (cm)

120

140

0

20

40 60 80 100 Net Applied Water (cm)

120

140

Figure 5. Comparison between area-averaged solute concentrations from solution samplers with CDE [Eq. (25)] and CLT [Eq. (28)] model predictions after calibration at 30 cm. Adapted from Butters (1987).

TRANSPORT OF CHEMICALS THROUGH SOIL

181

the close degree of agreement between the CLT model and the data, and the extremely poor agreement between the CDE model and the data. Recalling the foundational basis of the two dispersion models discussed in Section II,C,4, we conclude that on this field the lateral mixing time for solute is much longer than the travel time of the solute to the 3-m depth. As a result, the CLT model, which neglects mixing and treats the soil as consisting of a group of parallel independent columns with different local water velocities, provides a good description of the pulse spreading with time. Therefore, in this region the apparent dispersion coefficient is growing proportionally with distance below the point of entry. Indirect confirmation for this interpretation was offered by the transverse dispersion measurements of solute plumes by Ellsworth and Jury (1991ab), who found that the lateral spreading of solute was insignificant in the top 3 m during downward movement of the plume. Butters and Jury (1989) found that the initial stage of increasing dispersion persisted for considerable time and distance below the surface. The apparent dispersion coefficient continued to increase until the final soil sampling, when the pulse was centered at about 13 m, except for a decrease of 40% between 3 and 4.5 m. They attributed this temporary decrease to the influence of a thin layer of soil high in silt content located between the two regions. Figure 6 shows the apparent dispersivity A = D / V measured at each depth by refitting the CDE to the area-averaged data. A contrasting result to the one observed by Butters and Jury (1989) was found in the rainfall-driven field experiment reported by Roth etal. (1991). In this study, C1- movement was monitored with 110 solution samplers located in a grid arrangement extending 10 m horizontally and at depths between 0.4 and 2.3 m along the side wall of a tunnel under the field. The soil was extremely heterogeneous in the vertical direction, having eight distinct horizons in the top 3 m that varied substantially in texture and structure. The applied chloride pulse split into a main part that moved slowly downward through the soil matrix, and a series of fast pulses that moved by preferential flow, bypassing the solution samplers until they terminated at a subsurface layer of fine texture at about 2.2m. These pulses were sporadic and only present during the first part of the experiment. However, they transported approximately 58% of the total applied solute mass. The remaining 42% that moved slowly through the matrix was detected by the solution sampler network, and could be evaluated by dispersion models after the mass balance was corrected. In this case, the longitudinal dispersion of the area-averaged pulse moving within the soil matrix was described more accurately by the CDE than the CLT. This finding implies that the transverse mixing time in the soil matrix is much shorter on this field than on the one in which the study of Butters and Jury

WILLIAM A. JURY AND HANNES F L m L E R

182

0

0 Solution Samplers

0 Soil Cores

d

0

.

.

2

4

6

8

10

12

14

16

Depth (m) Figure 6. Apparent dispersivityobserved as a function of distance below the soil surface. After Butters and Jury (1989).

(1989) was conducted. This difference is undoubtedly a consequence of the soil layering in the field of Roth et al. (1991), which can greatly enhance lateral mixing of solute.

c. SOLUTE hSORFTION There have been very few field-scale studies evaluating solute adsorption. Those that have been conducted usually involved pesticides, and often were confined to shallow monitoring for persistence of residues. One of the earliest studies that monitored transport was the plot scale experiment of Rao et al. (1974). In this study, picloram was added in solution to the surface of two plots (2.74 X 2.74 m2) on a silty clay soil and was leached by separate ponded irrigations thereafter while the pulse was monitored by solution samplers between 0.03 and 1.43 m. They noted two unusual features of the leaching process. First, they observed traces of picloram at all depths after the first 24cm of irrigation. The maximum depth of appearance (1.43 m) was substantially deeper than that at which they expected to find the compound. Second, they found that the herbicide

TRANSPORT OF CHEMICALS THROUGH SOIL

183

peak migrated more slowly than they expected it to based on its modest retardation ( R = 1.6) in this soil. They attributed both phenomena to the presence of macropores, which caused part of the herbicide mass to move rapidly, and to significant depths, but which also caused subsequent irrigations to -be less efficient in leaching herbicide that had diffused into the microstructure. The first major field study of pesticide movement was reported by Jury ef al. (1986b), in which a pulse containing C1- and a relatively strongly adsorbed herbicide (napropamide) was added to the surface of a 0.64-ha loamy sand field and was subsequently leached with daily sprinkler irrigation for 2 weeks until 23 cm of water had been added. Then, soil cores were taken at 36 locations on the field to a depth of 3 m and were analyzed at 0.1-m intervals for C1- concentration. Nineteen of these cores were also analyzed for napropamide concentration. The laboratory measurements of equilibrium sorption Kd averaged 1.97 cm3 g-' on 36 samples taken near the coring locations, with a coefficient of variation of about 31% (El Abd et al., 1986). This produced a mean retardation factor R of about 9.6. Based on the variability of the water flow as revealed by the chloride profile, the authors predicted that the maximum depth the pesticide could reach if adsorbing to equilibrium was about 30cm. The pesticide profile contained a large peak at the surface depth of 10 cm, and approximately 77% of all mass averaged over the field area was found in the top 30 cm. However, the remaining 23% was located between 30 and 180cm, with a distinct secondary peak at about 90 cm. The deep movement was found in all but two of the 19 cores analyzed for pesticide. A second study was performed on field plots at the same site by Clendening (1985), who added four chemicals ranging in relative retardation from R = 1 (bromide) to R = 18 (prometryn), simultaneously in a spray to individual plots, and sampled to 3 m after 10 or 18 cm of water had been added. The study again showed both shallow and deep leaching of the strongly adsorbed pesticides, regardless of whether the plots were leached immediately after pesticide application, or whether a 72-hr equilibration period was allowed after application. A final study on the same field was performed by Ghodrati (1989), this time using three pesticides and chloride as the tracers, while varying the water application method (continuous or intermittent ponding or sprinkling), the pesticide formulation method (dissolved in water, emulsified concentrate, or wettable powder), and the condition of the soil surface (undisturbed or rototilled to 30 cm and repacked). Sixty four 1-m2 plots received 12 cm of water after chemical application and were subsequently sampled to 1.5 m with three cores that were consolidated in 10-cm depth

184

WILLZAM A. JURYAND HANNES FLmLER

intervals prior to analysis. There were no significant differences in the leaching patterns caused by the various water application methods, and all chemicals showed both shallow and deep leaching patterns. Breaking up and repacking the surface 30 cm of soil did not extinguish deep leaching of the pesticides, except when wettable powder was used. Ghodrati (1989) attributed this to physical filtering of the powder out of the solution by the smaller pores within the packed soil matrix. Other studies of pesticide transport under field conditions have reported higher than expected mobility for compounds that adsorb to the soil. Gish et al. (1986) reported that the laboratory-measured adsorption coefficient for atrazine greatly underestimated its mobility relative to bromide in a field study on a cropped silt loam soil. Hornsby et al. (1990) monitored bromide together with aldicarb and its metabolites for 200 days under a citrus grove in a sandy soil in Florida. They observed that bromide and aldicarb had comparable mobility, even though the latter was predicted to have a retardation factor of between 1.1 and 2.0 based on laboratory measurements. Because of the large amount of rainfall in this area, both the bromide and aldicarb reached groundwater at 7.2 m during the 200-day monitoring period. Bowman and Rice (1986) conducted a large study on a sandy loam field in which a water tracer (PBFA) and a mildly adsorbed herbicide (bromacil) were applied and leached under periodic ponding. Nearly all of the individual soil cores showed the bromacil to be retarded with respect to the water tracer, but both were moving faster than predicted by piston flow. The apparent retardation of bromacil relative to PBFA varied between 0.83 and 4.9 among 85 cores based on model fitting, the former number indicating that the bromacil moved faster than the companion nonsorbing tracer. Tile drains have been used as a means of monitoring nutrient and pesticide leaching losses below cropland for many years (Johnson et al., 1965; Pillsbury and Johnson, 1965; Devitt et at., 1976). Recently, they have been used to detect extremely high mobility of dissolved chemicals. Richard and Steenhuis (1988) reported chloride outflow in the 80-cm-deep drain effluent of a sandy loam, with the first outflow following application of the chemical to the surface. Subsequent rainfalls were also accompanied by additional outflow pulses of chloride. A similar phenomenon was reported by Kladivko etal. (1991) on a tile-drained silt loam soil. Following a single application to the surface, traces of the pesticides appeared in the drain tile after only 2 cm of net drainage. These pulses tapered off before the water quit flowing out of the tile, but returned with subsequent rainfall events through the season. Everts et al. (1989) conducted an analysis of four ions (including the very strongly sorbed organic dye rhodamine WT) moving to the tile drain of a loam soil, and found that traces of all four

TRANSPORT OF CHEMICALS THROUGH SOIL

185

appeared at the effluent during the first irrigation. These concentrations were as high as 30% of the input solution concentration for the bromide tracer and 6% for the rhodamine WT, indicating that substantial preferential flow with little mixing was occurring. The consensus among all of these studies is that transport of sorbing compounds through field soils cannot be completely described by a linear, equilibrium adsorption model, even when the spatial variability of the water flow regime is taken into account. Instead, a portion of the adsorbing solute displays much higher mobility than could be accounted for with a laboratory-measured adsorption coefficient. For the more strongly adsorbing compounds, extremely high mobility is often regarded as an indirect observation of preferential flow.

D. P~FERENTIALFLOW Field studies of preferential flow have consisted of two different types: direct monitoring through dye trace observations, and indirect monitoring by sorbing or nonsorbing chemical tracers. In the latter case, the concept of preferential flow is invoked to explain unusual mobility relative to the extent of transport expected to occur if the compound was undergoing normal convection and adsorption. Although preferential flow of water and chemicals has been recognized as a contributor to transport under field conditions for well over 100 years (Schumacher, 1864; Lawes et al., 1882), its significance was largely overlooked during the years of experimentation and modeling under laboratory conditions. When researchers returned to the study of transport through soil with natural structure, awareness of this phenomenon returned. With the development of the monolith lysimeter, Kissel et al. (1974) were able to monitor chemical leaching in a saturated swelling clay soil, observing 5% recovery of a chloride pulse in the effluent at 132 cm after only 5 cm of drainage. Starr et al. (1978) observed the movement of dyes through a layered (sandy loam over coarse sand) field site leached by ponding, and concluded that the majority of the water moved through narrow channels 5-20 cm in diameter that were produced through flow instabilities at the layer interface. A similar finding was observed by Quisenberry and Phillips (1976), who observed a marked increase in the coefficient of variation of solute concentration below a plow pan. They concluded that the solutes were moving in isolated channels below the interface. Aside from unstable flow arising from transport across layered soil interfaces, preferential flow has been shown to arise because of geometric voids in the soil. Ritchie et al. (1972) used a visible dye to demonstrate that

186

WILLIAM A. JURYAND HANNES FLWLER

much of the water moving through a swelling clay soil was migrating through vertical cracks. Omoti and Wild (1979) used florescent dyes in a weakly structured loamy sand to determine that earthworm channels, fissures with apertures between 0.05 and 0.10 mm, and loosely packed soil were all acting as conduits for the rapid transport of the adsorbing dyes they used in their field study. Johnston et al. (1983) used rhodamine WT to mark the pathways of water recharge to ground water in a lateritic weathering profile in Australia. They found that the water moved relatively uniformly through the 1-2 m of permeable sandy gravel comprising the surface zone. However, in the underlying clay-rich saprolites, the transport occurred mainly within cylindrical vertical channels up to 30 mm in diameter filled with coarse-textured material. The origin of the channels was believed to be decayed tree roots. Scotter and Kanchanasut (1981) reported movement of chloride to a mole drain at 0.4 m under continuous ponding within 5 min after introduction of the chemical into the infiltrating water. Dye tracing revealed that root and worm channels and occasional fracture planes were carrying the flow. The above studies for the most part describe preferential flow through structural voids. However, even structureless sandy soils have been found to exhibit preferential flow, arising from a variety of causes. Saffigna et al. (1976) used rhodamine WT dye to reveal a highly nonuniform water infiltration pattern under a potato crop on Plainfield sand in Wisconsin. They attributed the nonuniformity to stemflow along the plant and runoff from the sidewall of the hills on which the potatoes were planted. A later comprehensive study was conducted on Plainfield sand by Kung (1990a,b), who excavated a field plot after dye application. By detailed observation of the dye flow patterns, he was able to show that water funneled into increasingly narrow zones at greater depths, eventually carrying the bulk of the flow in a small fraction of the cross-sectional area. Kung (1990b) postulated that the funneling mechanism was discrete coarse sand lenses that acted as barriers to downward movement, causing the water to flow around them and focus at the edges. The high permeability of the sand allowed very high local flows in the focused regions. A similar phenomenon was reported by Ghodrati and Jury (1990) on the Etiwanda loamy sand field where deep leaching of pesticides had been reported by Jury et al. (1986b,c). Using a soluble anionic dye (amine red), Ghodrati and Jury (1990) observed that preferential flow regions comprising a small fraction of the cross-section were moving more than twice as deep as the main front under both ponded and sprinkler irrigation in undisturbed field plots. When the surface was tilled and repacked, distinct fingers formed at the plow layer. Certain porous median contain two or more distinct flow domains that

TRANSPORT OF CHEMICALS THROUGH SOIL

187

allow water to be transported through different regions at different flow rates. Jardine et al. (1990) reported a study of bromide transport through an intact soil 12-m3pedon from a silt loam that contained a fast flow region composed of large pores, and a slower flow region in the finer pores of the matrix material. They isolated the flow regions by solution samplers operating in different suction ranges, and showed that the larger pores were carrying flow rapidly to great depths and were receiving solute by mass transfer from the surrounding slower flow domain. Roth et a1 (1991) studied solute transport through a layered and structured soil in Switzerland, finding that macropore movement accounted for about 56% of the total flow. The fast movement carried substantial solute mass below 2 m during the course of the study, whereas the matrix flow was centered at about 85 cm at the end of the experiment. Preferential flow has serious implications for pesticide movement, because it is desirable to keep these compounds from migrating below the surface zone. Table I11 summarizes the extent of preferential flow observed in various studies of adsorbed chemical movement. In this table, preferential flow is operationally defined as transport more than twice as deep into the soil as would be predicted by piston flow and linear, equilibrium adsorption. All but the last two soil coring studies were conducted on the same field, although at different times and under different conditions. It is clear from these results that the leading edge or early arrival of adsorbing compounds is not consistent with a model description involving random dispersion superimposed on mean convection and equilibrium adsorption. For example, the average volumetric water content of the Etiwanda loamy sand field site where the napropamide studies were conducted was about 0.18 at the flow rates induced by the sprinkler system (Butters et al., 1989). With a retardation factor of 10, the mean predicted depth of pesticide migration after 23 cm of applied water is about 13 cm, whereas traces of the compound were found as deep as 180 cm. On the same site, triallate, with an apparent retardation factor of 100, reached 100 cm with the first 5-cm water application, whereas the mean predicted depth of leaching was only 3 mm (Clendening and Jury, 1990). Extreme mobility of strongly adsorbed compounds has been observed in other studies as well, as, for example, in the early arrival of rhodamine WT after 4.5 cm of water applied in the drain effluent of the field monitored by Richard and Steenhuis (1988), and in the appearance of pesticides pulses in the effluent of the tile lines responding to individual storms in the study of Kladivko et al. (1991). Isensee et al. (1990) observed peak atrazine concentrations in shallow groundwater during the first rainfall after application to their silt loam field. High concentrations of two other applied pesticides (carbonfuran and cyanazine) were also detected at the same time.

Table UI Preferential Flow of Adsorbing Compounds Observed in Field Studies Chemical

Soil type

Water applied (cm)

Napropamide

Loamy sand

Atrazine

Loamy sand

Prometryn

Loamy sand

Prometon Triallate Atrazine Picloram

Loamy sand Loamy sand Silt loam Clay loam

23 10 18 12 5 12 10 18 12 5 5 5 24

Rhodamine WT Carbofuran Alachlor

Loam Silt loam Silt loam

4.5 2' 2'

Retardation factor Soil coring studies 10 10 10 10 6 6 20 20 20 6 100 6 1.6

Tile drain studies 67 =2

-2

Preferential flow"

("/.I

Max depth (cm)

27 25 40 12 9 18 25 36 6 8 8 NA NA

180 110 150 110 100 110 100 120 100 100 100 60-150 143

0.15 0.05-0.94' 0.04-0.19'

110 75 75

Reference El Abd (1984) Clendening (1985) Clendening (1985) Ghodrati (1989) Clendening (1988) Ghodrati (1989) Clendening (1985) Clendening (1985) Ghodrati (1989) Clendening (1988) Clendening (1988) Isensee et al. (1990) Rao et al. (1974)

Everts et al. (1989) Kladivko et al. (1991) Kladivko et af. (1991)

"Percentage of applied mass recovered in soil cores more than twice as deep as predicted by equilibrium adsorption, or percentage leached to tile line; NA, not available. 'Tile outflow when compounds first appeared. 'Seasonal totals in different blocks.

TRANSPORT OF CHEMICALS THROUGH SOIL

189

E. SCALEOF HETEROGENEITY One of the problems encountered in drawing conclusions from field studies is that they are frequently performed on a plot scale that is too small to encompass the full heterogeneity of the field. Van Wesenbeek and Katchanoski (1991) conducted a study of solute transport monitored by solution samplers arrayed in a 0.2-m spacing along a transect to determine the lateral extent of solute monitoring that would be required to reach a scale where solute concentration averages would represent the field-scale value. By plotting the travel time variance calculated by averaging different numbers of adjacent samplers as a function of the averaging distance, they were able to show that an averaging distance of approximately 2.8 and 3.8 m was needed on two sandy field sites before the local scale and the field scale agreed. For distances less than this value, the variance was less than the field-scale variance, indicating that the full range of solute velocity was not encountered within the averaging domain. As described in Section IV,B, the nature of the transport process will shift from stochastic-convective to convective-dispersive after transverse mixing has allowed solute molecules to explore all regions over which water velocity is correlated. Based on the study above, this zone may be roughly regarded as 3 m in soils similar to the ones examined by Van Wesenbeek and Katchanoski (1991). Ellsworth and Jury (1991b) monitored the movement of massive solute plumes 1.5 or 2 m on a side and observed less than 0.5 m of lateral solute migration during more than 4 m of vertical migration when the soil was initially wet prior to solute application. However, when the field was initially dry, they observed several meters of lateral migration, caused by permeability variations in the vertical direction from small textural changes. Thus, the extent of lateral mixing is difficult to forecast, even in a relatively homogeneous soil, and may depend on the state of saturation of the soil.

F. INDICES OF SOLUTE DISPERSION Because solute dispersion is likely to be dominated by lateral differences in vertical velocity near the surface, an alternative to measuring the dispersion of a field-scale pulse or front is to evaluate the variance of the local solute velocity. Moreover, the travel time and solute velocity generally are found to be skew distributed in the field, so that the standard deviation (+ of the logarithm of either variable represents a convenient index for characterizing the extent of solute spreading in the longitudinal direction (Jury, 1985). A similar approach has long been used in groundwater

Table IV Log Variance of Travel Time or Solute Velocity Observed in Field Studies Variable

Soil type

Area

Application

Measurements

N

U

cv

Reference Starr et al. (1978) Van der Pol et al. (1977) Biggar and Nielsen (1976) Jury ef al. (1982) Butters and Jury (1989) Butters (1987) El Abd (1984) Gish et al. (1986) Jaynes et al. (1988) Schulin er al. (1987) Schulin et al. (1987) Van Wesenbeek and Katchanoski (1991) Roth (1989)

Loamy sand Clay Clay loam Loamy sand Loamy sand Loamy sand Loamy sand Silt loam Clay loam Stony Stony Sand

112 m2 64 m2 150 ha 0.64 ha 0.64 ha 0.64 ha 0.64 ha NA' 13 m2 94 m2 94 mz 9.6 md

Ponding Tricklers Ponding Rainfall Sprinkler Sprinkler Sprinkler Rainfall Ponding Rainfall Ramfall Trickers

Solution samples to 2.4 m Solution samples to 1.5 m Solution samples to 1.8 m Solution samples to 1.8 m Solution samples to 1.8 m Solution samples to 1.8 m Soil cores to 3.0 m Soil cores to 0.45 m Solution samples to 3.0 m Soil cores to 3.1 m Soil cores to 3.1 m Solution samples at 0.4 m

44 24 120 70 80 80 36 180 28 57 61 48

0.69 0.57 1.25 0.57 0.46 0.40 0.51 1.21 0.46-0.65 0.21 0.26 0.37

78 61 194 61 48 42 55 184 48-68 21 26 38

Loam, layered

10 m

Rainfall/ sprinkler

Solution samplers 0.4-2.5 m

110

0.83

100

"0 is the apparant solute transport volume Z/Z, where Z = depth of sampling. blis the net applied water required to reach depth of sampler.

'NA, Not available. Transect.

TRANSPORT OF CHEMICALS THROUGH SOIL

191

hydrology, wherein the degree of heterogeneity of an aquifer is characterized by the standard deviation of the logarithm of the saturated hydraulic conductivity (Freeze and Cherry, 1979). Table IV summarizes the published information on solute travel time or velocity obtained from field studies of mobile tracers. The studies in which ponding was used as the water application method measure velocities whose variations contain significant contributions from lateral variations in water entry rate, whereas the sprinkler and trickler studies have highly uniform entry rates. Therefore, the intrinsic variability caused by soil properties is smaller than the values given in the ponding studies. The relationship between the travel time arid solute velocity variability can be illustrated by comparing the fifth and sixth entries in the table, which are based on the same data set reported in Butters et al. (1989). The solute velocity was obtained from the method of moments analysis of individual breakthrough curves (Jury and Sposito, 1985), whereas the travel time variance was calculated from the field-scale breakthrough curves at 0.3-.8 m obtained from area averaging the local concentrations. The former is somewhat smaller than the latter, because it does not include the contribution to field-scale dispersion from local dispersion. The high value of 0.65 for the travel time variance was at 0.9 m, where there was a texture change. The 16 average solute velocities in this region are also = 0.53) (Butters, 1987). more variable (qnv Although the entries in this table span a wide range of values, they all show coefficients of variation (CV) that are significantly in excess of the CV reported for unsaturated water content in the field (Warrick and Nielsen, 1977; Jury, 1985). This finding illustrates that field-scale solute transport cannot be simulated by area averaging local one-dimensional water and solute transport model simulations, because the solute velocity in the latter is inversely proportional to the volumetric water content 8. It is clear from Table IV that the solute velocity must be three dimensional, and perhaps includes only part of the wetted pore space. Neither of these effects is included in current unsaturated zone flow and transport models that can be calibrated on field data.

V. CONCLUDING REMARKS The review has presented and evaluated current experimental information and theoretical approaches used to represent chemical transport and transformations in unsaturated soil. The majority of the review has been devoted to the field regime, because of the large number of applications for

192

WILLIAM A. JURY AND HANNES F L m L E R

chemical transport to agricultural and environmental problems in natural settings. In addition, it seems appropriate to draw a distinction between field and laboratory research, because they are conducted at such different scales. It is unfortunate that the theory of solute transport through soil was developed exclusively from stuudies conducted in the laboratory, because that theory has been transferred to the field regime without justification in many cases. The result of this action has been that models have entered the consulting, regulatory, and policy arenas, where they are used routinely in applications for which their assumptions do not apply.

A. PREFERENTIALFLOW AND THE FIELD REGIME The field evidence regarding preferential flow is quite consistent in one respect. Preferential flow can occur under a variety of circumstances, and is not restricted to clay-rich soils with significant structural voids. Rather, based on the evidence of the dye trace studies of Kung (1990a,b) and Ghodrati and Jury (1990), it seems likely that structureless sandy soils may be just as prone to developing preferential flow as are finer textured structured ones, although clearly for different reasons. A defining characteristic of preferential flow in structureless soils is high matrix permeability. This ensures that any plume resulting from funneling of flows by lateral migration around obstacles will be able to move downward at the new higher flow rate. Such flows at the present time appear to be unpredictable, in the sense that they are triggered by small-scale features that cannot be characterized by conventional sampling methods in the unsaturated zone. The conditions under which fluid instabilities caused by density or viscosity gradients will occur are predictable; however, the spatial and temporal distributions of those conditions in the field regime are not. The same may be said for instabilities caused by entrapped air or transport from a fine- to a coarse-textured soil layer. However, the controlling factors are distributed in an unknown manner throughout the soil. Finally, even fluid transport through well-defined structural voids is not predictable unless the distributions of voids, aperture sizes and shapes, depths of penetration, and interconnectivity are known. Progress is being made slowly in characterizing transport through rock fractures, but there the geometry is much more stable in time than it is in the soil regime. Moreover, subsurface flow channels can either be barriers or flow conduits depending on the local fluid pressure, and the latter is subject to many factors that are difficult to predict in nature.

TRANSPORT OF CHEMICALS THROUGH SOIL

193

Laboratory studies have demonstrated clearly that soil structure is almost certain to introduce mass transfer limitations to equilibrium between the dissolved and sorbed phases in soil. In addition, even relatively structureless soils appear to have stagnant regions where a solute can only migrate by diffusion within the dissolved phase. As a result, correct description of the solute transport process will necessarily involve either rate coefficients and mass transfer models similar to the expression in Eq. (40), or will need to use specific diffusion flux models such as the one used by Rao et al. (1980a) to represent the migration of a solute into stagnant regions. Such processes can be identified and resolved experimentally in the laboratory, where soil may be prepared in such a way as to mirror the assumptions made in the mass transfer model. However, in the field there is likely to be a wide range of rate-limiting processes operating within the averaging volume, and the volume-averaged process may not be describable with a single rate coefficient. Although several researchers have offered guidance as to the conditions under which lumped parameter coefficients may suffice (Villermaux, 1987; Brusseau et al., 1989; Parker and Valocchi, 1986), some knowledge of the local geometry is required to make this decision. When a preferential flow process is occurring, the extent of mass transfer between the preferential flow channel and the surrounding soil matrix will depend on the geometry of the interface, on the concentration difference between the regions, and on the speed of the fluid as it moves through the preferential flow channel. Moreover, the mass transfer process may not be regulated by liquid diffusion if the flow velocities are high enough to induce turbulence, or if a range of velocities are encountered within the channel.

C. DISPERSION PROCESSESAND THEFIELDREGIME The nature of the longitudinal dispersion of an area-averaged pulse or front added to the surface of a field soil is determined by the relation of the lateral mixing time to the convective residence time. As a result, the type of dispersion process may be very different from one soil to the next, even if each soil is exposed to the same boundary influences. Even the same soil will react differently to the same boundary influence as its initial condition is varied (Stephens et al., 1988; Ellsworth and Jury, 1991b). Transfer functions can be derived from observations made at a field site to develop field-scale dispersion representations, but only for conditions

194

WILLIAM A. JURY AND HANNES FLUHLER

similar to those under which the observations were made. Thus, for example, the solute transport in the field of Butters and Jury (1989) was described as stochastic-convective in the top 3 m when the soil surface was initially wet, but might behave quite differently if the solute pulse is added as the leading edge of an infiltrating water front into dry soil. A considerable amount of additional experimental information will be required to develop general guidelines for extrapolating transfer function calibrations from one application to the next on the same field. At the present time, only a single study of this type has been performed (Jury et al., 1990). On the other hand, the mechanistic description of simultaneous water and solute transport in three dimensions at the field scale is fundamentally limited by huge data requirements and incompatible scales of measurement between different devices used to characterize the transport and retention properties in siru. For example, the preferential flow of water and solute observed in the study of Kung (1990a,b) apparently was initiated and maintained by small discrete lenses of coarse sand interspersed within the finer sand matrix. These small lenses would not contribute substantially to the large volume measurements of hydraulic conductivity such as those used by Libardi et al. (1980), but they can have a dominant effect on transport under certain conditions. When spatial variability is incorporated into the description of a fieldscale transport model, the usual approach is to simulate the variability using local one-dimensional models in parallel with local values of the model parameters, and average the results to produce the field value of the concentrations. Although this produces a stochastic-convective model automatically when parallel columns are averaged (Jury and Roth, 1990), the local hydrologic model generally produces a local water velocity that is far less variable than the real one, which is three dimensional and moves through only part of the wetted pore space. Only by using a direct field scale measurement of the travel time pdf can the three-dimensional nature of the solute velocity be represented ifi the dispersion representation.

D. LESSONS FROM HISTORY Students of agricultural history can discover in early readings of field observations that much of the complexity of the field regime was perceived long ago. As pointed out by Thomas er al. (1978), concepts such as bypassing, preferential flow, and channeling of water through unsaturated soil have appeared in the literature for more than 100 years (cf. Schumacher, 1864; Lawes et al., 1882). The nature of solute dispersion and its relationship to transverse mixing and convective residence times was

TRANSPORT OF CHEMICALS THROUGH SOIL

195

laid out elegantly by Taylor in a classic paper in 1953. Yet, somehow the convection-dispersion equation has been transported to the field regime without justification of any kind for its assumptions, and it continues to be encoded into new software programs that will be widely used in the regulatory, consulting, and even academic/research arenas. The outcome of this practice will be unfortunate, both for the users of such information, and also for students in current and future generations. RERERENCES Abriola, L., and Pinder, G. F. (1985). A multiphase approach to the modeling of porous media contamination by organic compounds. Water Resour. Res. 21, 11-26. Addiscott, T. M. (1977). A simple computer model for leaching in structured soils. J . SoiZSci. 28, 554-563. Aitcheson, J., and Brown, J. A. C. (1976). “The Lognormal Distribution.” Cambridge Univ. Press, Cambridge, England. Arfken, G. (1985). “Mathematical Methods for Physicists,” 3rd Academic Press, New York. Baveye, P., and Sposito, G . (1985). Macroscopic balance equations in soils and aquifers: The case of space- and time-dependent instrumental response. Water Resour. Res. 21, 116-120. Bear, J. (1972). “Dynamics of Fluids in Porous Media.” Elsevier, New York. Beven, K., and Germann, P. (1982). Macropores and water flow in soils. Water Resour. Res. 18, 1311-1325. Biggar, J. W., and Nielsen, D. R. (1967). Miscible displacement and leaching phenomenon. Agron. Monog. 11,254-274. Biggar, J. W., and Nielsen, D. R. (1976). Spatial variability of the leaching characteristics of a field soil. Water Resour. Res. 12(1), 78-84. Bird, R. B., Stewart, W. E., and Lightfoot, E. N. (1960). “Transport Phenomena.” Wiley, New York. Bowman, R. S., and Rice, R. C. (1986). Accelerated herbicide leaching resulting from preferential flow phenomena and its implications for groundwater contamination. In “Proceedings of the Conference on Southwestern Groundwater Issues, Phoenix, Arizona.” National Well Assoc., Dublin, Ohio. Brusseau, M. L., and Rao, P. S. C. (1990). Modeling solute transport in structured soils: A review. Geoderma 46, 169-192. Brusseau, M. L., Jessup, R. E., and Rao, P. S. C. (1989). Modeling the transport of solutes influenced by multiprocess equilibrium. Water Resour. Res. 25, 1971-1988. Butters, G . L . , (1987). Field scale transport of bromide through unsaturated soil. Ph.D. Dissertation, University of California Riverside. Butters, G. L., and Jury, W. A. (1989). Field scale transport of bromide in an unsaturated soil. 2. Dispersion modeling. Water Resour Res. 25, 1582-1588. Butters, G. L., Jury, W. A., and Ernst, F. F. (1989). Field scale transport of bromide in an unsaturated soil. 1. Experimental methodology and results. Water Resour. Res. 25, 1575-1581. Clendening, L. D. (1985). A field test of pesticide mobility screening parameters. MS Thesis, University of California, Riverside. Clendening, L. D. (1988). A pesticide field mass balance study. Ph.D. Thesis, University of California, Riverside.

196

WILLIAM A. JURY AND HANNES F L ~ L E R

Clendening, L. D., and Jury, W. A. (1990). A pesticide field mass balance study. In “Long Range Transport of Pesticides” (D. Kurtz ed.). Lewis Pub., Chelseea, M1. Coats, K. H., and Smith, B. D. (1956). Dead end pore volume and dspersion in porous media. SOC.Pet. Eng. J . 4, 73-84. Collin, M., and Rasmuson, A. (1988). Comparison of gas diffusivity models for unsaturated porous media. Soil Sci. SOC.Ar. J. 52, 1559-1565. Dagan, G. (1984). Solute transport in heterogeneous porous formations. J . Fluid Mech. 145, 151-177. Dagan, G. (1987). Theory of solute transport by groundwater. Annu. Rev. Fluid Mech. 19, 183-215. Devitt, D. D., Letey, J., Lund, L. J., and Blair, J. W. (1976). Nitrate nitrogen movement through soil as affected by soil profile characteristics. J . Environ. Qual. 5, 283-287. Devitt, D. D., Evans, R. B., Starks, T. J., and Jury W. A . (1987). “Soil Gas Sensing for Detection and Mapping of Volatile Organics.” National Water Well Assoc., Dublin, Ohio. Dyson, J . S. and White, R. E. (1986). The effect of irrigation rate on solute transport in soil during steady water flow. J . Hydrol. 107, 19-29. El Abd, H. (1984). Spatial variability of the pesticide distribution coefficient. Ph. D. Thesis, University of California, Riverside. El Abd, H., Jury, W. A., and Cliath, M. M. (1986). Spatial variability of pesticide adsorption parameters. Environ. Sci. Technol. 20(3), 256-260. Ellsworth, T. E., and Jury, W. A. (1991a). A three dimensional field study of solute leaching through unsaturated soil. I. Characterization of vertical dispersion. Water Resour. Res. 27, 967-982. Ellsworth, T. E., and Jury, W. A. (1991b). “Three Dimensional Solute Plume Transport Through and Unsaturated Soil,” EPRI Top, Rep. EN-7283. Elect. Power Res. Inst., Palo Alto, California. Ellsworth, T. E., Jury, W. A., Ernst, F. E., and Shouse, P. J. (1991). A three dimensional field study of solute leaching through unsaturated soil. I. Methodology, mass balance, and mean transport. Water Resour. Res. 27, 951-966. Everts C. J., Kanwar, R. S., Alexander, E. C., Jr., and Alexander, S. C. (1989). Comparies under laboratory and field conditions. J , Environ. Qual. 18, 49 1-498. Farmer, W. J., Yang, M. S., Letey, J., and Spencer, W. F. (1980). Hexachlorobenzene: Its vapor pressure and vapor phase diffusion in soil. Soil Sci. SOC.A m . Proc. 44,676-680. Focht, D. D., and Joseph, H. (1973). An improved method for the enumeration of denitrifying bacteria. Soil Sci. SOC.A m , Proc. 37, 698-699. Follett, R. F., and Walker, D . J. (1989). Ground water quality concerns about nitrogen, Dev. Agric. Managed-For. Ecol. 21, 1-22. Freeze, R. A., and Cherry, J. A. (1979). “Groundwater.” Prentice-Hall, Englewood Cliffs, New Jersey. Freyberg, D. (1986). A natural gradient experiment on solute transport in a sand aquifer. 2. Spatial moments and advection and dispersion of nonreactive tracers. Water Resour. Res. 22, 2031-2046. Garabedian, S. P., LeBlanc, D . R., Gelhar, L. W., and Celia, M. A. (1991). Large-scale, natural-gradient tracer test in sand and gravel, Cape Cod, Masachusetts. 2. Analysis of spatial moments for a nonreactive tracer. Water Resour. Res. 27, 91 1-924. Gelhar, L. W. (1986). Stochastic subsurface hydrology from theory to application. Water Resour. Res. 22, 135s-146s. Gelhar, L. W., and Axness, C. (1983). Three dimensional analysis of macrodispersion in aquifers. Water Resour. Res. 19, 161-180.

TRANSPORT OF CHEMICALS THROUGH SOIL

197

Ghodrati, M. (1989). The influence of water application method, pesticide formulation and surface preparation method on pesticide leaching. Ph.D. Thesis, University of California, Riverside, Ghodrati, M., and Jury, W. A. (1990). A field study using dyes to characterize preferential flow of water. Soil Sci. SOC.A m . J . 54, 1558-1563. Gish, T. J., Helling, C. S., and Kearney, P. C. (1986). Chemical transport under no-till field conditions. Geoderma 38, 251-259. Goltz, M. N., and Roberts, P. V. (1986). Diffusion of sorbing solutes: Impact on contaminant transport in ground water. J . Conram. Hydrol. 1, 77-94. Hamaker, J. W. (1972). Decomposition: Quantitative aspects. In“0rganic Chemicals in the Soil Environment” (C. A. I. Goring and J . W. Hamaker, eds.), pp. 253-340. Dekker, New York. Hamaker, J . W . , and Thompson, J. M. (1972). Adsorption. In “Organic Chemicals in the Soil Environment” (C. A. I. Goring and J. W. Hamaker, eds.), pp. 49-144. Dekker, New York. Hill, D. E., and Parlange, J.-Y. (1972). Wetting front instability in layered soils. SoilSci. SOC. Am. Proc. 36 (5), 697-702. Hill, J . M. (1981). On the solution of reaction-diffusion equations. IMA J . Appl. Math. 27, 177-194. Hillel, D . , and Baker, R. S. (1988). A descriptive theory of fingering during infiltration into layered soils. Soil Sci. 146, 51-56. Hornsby, A. G., Rao, P. S . C., and Jones, R. L. (1990). Fate of Aldicarb in the unsaturated zone beneath a citrus grove. WaferResour. Res. 26(10), 2287-2302. Hubbert, M. K. (1956). Darcy law and the field equations of the Bow of underground fluids. Trans. A m . Inst. Min. Metal. Per. Eng. 207, 222-239. Isensee, A. R., Nash, R. G., and Helling, C. S . (1990). Effect of conventional vs. no-tillage on pesticide leaching to shallow ground water. J . Environ. Qual. 19, 434-440. Jardine, P. M., Wilson, G . V., and Luxmoore, R. J . (1990). Unsaturated solute transport through a forest soil during rain storm events Geoderma 46, 103-118. Jaynes, D. B., Bowman, R. S . , and Rice, R. C. (1988). Tranport of a conservative tracer in the field under continuous flood irrigation. Soil Sci. SOC.A m . J. 52, 618-624. Johnson, W. R., Ittihadich, F., Daum, R. M., and Pillsbury, A. F. (1965). Nitrogen and phosphorus in tile drainage effluent. Soil Sci. SOC. A m . Proc. 29, 287-289. Johnston, C. D., Hurle, D. H., Hudson, D. R., and Height, M. J. (1983). “Water Movement Through Preferred Paths in Lateritic Profiles of the Darling Plateau, Western Australia,” Ground Water Res. Tech. Pap. No. 1 , pp. 1-34. Commonw. Sci. Ind. Res. Organ. (CSIRO), Melbourne, Australia. Jury, W. A. (1982). Simulation of solute transport using a transfer function model. Water Resour. Res. 18, 363-368. Jury, W. A. (1985). “Spatial Variability of Soil Physical Parameters in Solute Migration: A Critical Literature Review, EPRI Top. Rep. EA 4228. Electr. Power Res. Inst., Palo Alto, California. Jury, W. A., and Ghodrati, M. (1988). Overview of organic chemical environmental fate and transport modeling approaches. In “Reactions and Movement of Organic Chemicals in Soil.” Sol1 Sci. SOC.Am. Spec. Pub. 22, pp. 271-304. Jury, W. A,, and Roth, K. (1990). “Transfer Functions and Solute Movement Through Soil: Theory and Applications.” Birkhauser, Basel and Boston. Jury, W. A., and Sposito, G. (1985). Field calibration and validation of solute transport models for the unsaturated zone. Soil Sci. SOC. A m . J . 49, 1331-1441. Jury, W. A., Stolzy, L. H., and Shouse, P. (1982). A field test of the transfer function model for predicting solute movement. Water Resour. Res. 18, 368-374.

198

WILLIAM A. JURY AND HANNES F L m L E R

Jury, W. A., Spencer, W. F . , and Farmer, W. J. (1983). Model for assessing behavior of pesticides and other trace organics using benchmark properties. I. Description of model, J . Environ. Qual. 12, 558-564. Jury, W. A., Sposito, G., and White, R. E. (1986a). A transfer function model of solute -movement through soil. I. Fundamental concepts. WaferResour. Res. 22, 243-247. Jury, W. A., El Abd, H., and Resketo, (1986b). Field study of napropamide movement through unsaturated soil. Water Resour. Ref. 22, 749-755. Jury, W. A,, El Abd, H., Clendening, L. D., and Resketo, M. (1986~).Evaluation of pesticide transport models under field conditions. ACS Symp. Ser. 315, 384-395. Jury, W. A., Focht, D. D., and Farmer, W. J. (1987). Evaluation of pesticide ground water pollution potential from standard indices of soil-chemical adsorption and biodegradation. J . Environ. Qual. 16, 422-428. Jury, W. A., Dyson, J. S . , and Butters, G. L. (1990). A transfer function model of field scale solute transport under transient water flow. Soil Sci. SOC.A m . J . 54, 327-332. Jury, W. A , , Gardner, W. R., and Gardner, W. H. (1991). “Soil Physics.” Wiley, New York. Kabala, Z. J., and Sposito, G. (1991). A stochastic model of reactive solute transport with time-varying velocity in a heterogeneous aquifer. Water Resour. Rex 27, 341-350. Khan, A. U. H., and Jury, W. A. (1990). A laboratory test of the dispersion scale effect. J . Contam. Hydrol. 5, 119-132. Kissel, D. E., Ritchie, J. T., and Burnett, E. (1974). Nitrate and chloride leaching in a swelling clay soil. J. Environ. Qua[. 3(4), 401-404. Kladivko, E. J., Van Scoyoc, G. E., Monke, E. J., Oates, K. M., and Pask, W. (1991). Pesticide and nutrient movement into subsurface tile drains on a silt loam soil in Indiana. .I Environ. . Qual. 20, 264-270. Kreft, A., and Zuber, A. (1978). On the physical meaning of the dispersion equation and its solution for different initial boundary conditions. Chem. Eng. Sci. 33, 1471-1480. Krupp, H. K., and Elrick, D. E. (1969). Density effects in miscible displacement experiments. Soil Sci. 107, 372-380. Kung, K-J. S., (1990a). Preferential flow in a sandy vadose zone. 1. Field observation. Geoderma 46, 51-58. Kung, K-J. S . , (1990b). Preferential flow in a sandy vadose zone. 2. Mechanism and implications. Geoderm 46, 59-71. Lapidus, L., and Amundson, N. R. (1952). Mathematics of adsorption in beds. VI. The effect of longitudinal diffusion in ion exchange and chromatographic columns. J. Phys. Chem. 56, 48-49. Lawes, J. B., Gilbert, J. H., and Warington, R. (1882). On the amount and composition of the rain and drainage waters collected at Rothamsted. 111. The drainage water from land cropped and manured. J . R . Agric. SOC.Engl. 18, 1-17. Libardi, P. L., Reichardt, K., Nielsen, D. R., and Biggar, J. W. (1980). Some simple field methods for estimating soil hydraulic conductivity. Soil Sci. SOC.A m . J . 44,3-6. Lumley, J. L., and Panofsky, A. (1964). “The Structure of Atmospheric Turbulence.” J. Wiley, New York. Mackay, D. (1987). Chemical and physical behavior of hydrocarbons in freshwater. In “Oil In Freshwater: Chemisrry, Biology and Countermeasure Technology” (J. Vandermeulen and S. Hrudney, eds.). Pergamon Press, New York. Mantaglou, A., and Gelhar, L. W. (1987). Stochastic modeling of large scale transient unsaturated flow systems. Water Resour. Res. 23, 57-67. Miller, C. T., Pokier-McNeill, M. M., and Meyer, A. S . (1990). Dissolution of trapped nonaqueous phase liquids: Mass transfer characteristics. Wafer Resour. Res. 26, 27832796.

TRANSPORT OF CHEMICALS THROUGH SOIL

199

Millington, R . J., and Quirk, J. M. (1961). Permeability of porous solids. Trans. Faraday SOC. 57, 1200-1207. Moharir A. S., Kunzru, D., and Saraf, D. N. (1980). Effect of adsorbent particle size distribution on breakthrough curves for molecular sieve columns. Chem. Eng. Sci. 35, 1795. Nielsen, D. R., and Biggar, J . W. (1962). Miscible displacement in soils. 3. Theoretical considerations. Soil Sci. SOC.Am. Proc. 26, 216-221. Nkedi-Kizz, P., Biggar, J. W., Selim, H. M., Van Genuchten, T. Th., Wierenga, P. J., Davidson, J. M., and Nielsen, D. R. (1984). On the equivalence of two conceptual models for describing ion exchange during transport through an aggregated Oxisol. Water Resour. Res. 20, 1123-1130. Novotny, V., and Chesters, G. (1981). “Handbook of Nonpoint Pollution.” Van NostrandReinhold, New York. Omoti, U., and Wild, A. (1979). Use of fluorescent dyes to mark the pathways of solute movement through soils under leaching conditions. 2. Field experiments. Soil Sci. 128, 93- 104. Parker, J. C . , and Valocchi, A. (1986). Constraints on the validity of equilibrium and local first order kinetic transport models in structured soils. WaferResour. Res. 20,1123-1130. Parker, J. C., Lenhard, R. J . , and Kuppsamy, T. (1987). A parametric model for constitutive properties governing multiphase flow in porous media. WaferResour. Res. 23, 618-624. Pfannkuch, H. 0. (1984). Mass exchange processes at the petroleum-water interface. In “Groundwater Contamination by Crude Oil at the Bemidji, Minnesota Research Site.” (M. F. Huh, ed.), U. S. Geol. Surv. St. Paul, Minnesota. Pillsbury, A. F., and Johnson, W. R. (1965). Tile drainage in the San Joaquin Valley of California. Publ. - Univ. Calif., WaferResour. Cent. 91. Pye, V., and Kelly, J. (1984). The extent of groundwater contamination in the United States. In “Groundwater Contamination,” pp. 23-44. National Academy Press, Washington, D.C. Quisenberry, V. L., and Phillips, R. E. (1976). Percolation of surface-applied water in the field. Soil Sci. SOC. Am. J . 40, 484-489. Rao, P. S. C . , and Davidson, J. M. (1979). Adsorption and movement of selected herbicides at high concentration in soils. Water Res. 13, 375-380. Rao, P. S. C., Green, R. E., Balasubramaniam, and Kanchiro, Y. (1974). Field study of solute movement in a highly aggregated Oxisol with intermittent flodding. 11. Picloram. J . Environ. Qual. 3, 197-202. Rao, P. S . C., Rolston, D. E., Jessup, R. E., Davidson, J . M., and Kilcrease, D. P. (1980a). Experimental and mathematical description of nonadsorbed solute transfer by diffusion in spherical aggregates. Soil Sci. SOC.Am. J . 44, 684-688. Rao, P. S. C., Rolston, D. E., Jessup, R. E., and Davidson, J. M. (1980b). Solute transport in aggregated porous media. Theoretical and experimental evaluation. Soil. Sci. SOCAm. J. 44, 1139-1146. Rasmuson, A. (1985). The influence of particles of variable size, shape, and properties on the dynamics of fixed beds. Chem. Eng. Sci. 40, 621. Rice, R. C . , Jaynes, D. B., and Bowman, R. S. (1988). Preferential flow of solutes and herbicides under irrigated fields. ASA E pap. 88-2634. Richard, T. L., and Steenhuis, T. S. (1988). Tile drain sampling of preferential flow on a field. J . Confam. Hydrol. 3, 307-325. Ritchie, J. T., Kissel, D. E., and Burnett, E. (1972). Water movement in undurbed swelling clay soil. Soil Sci. Soc. A m . Proc. 36, 874-879. Roth, K. (1989). Stofftransport im wasserungesattigten Untergrund natiirlicher, heterogener Boden unter Feldbedingungen. Dissertation, ETH Zurich No. 8907.

200

WILLIAM A. JURY AND HANNES F L m L E R

Roth, K., Fliihler, H., and Attinger, W. (1990). Transport of a conservative tracer under field conditions: Qualitative modeling with random walk in a double porous medium. In “Field Scale Water and Solute Flux in Soils” (K. Roth et al., eds.) pp. 239-249. Birkhauser, Basel. Roth, K., Jury, W. A., Fliihler, H., and Attinger, W. (1991). Field scale transport of chloride through an unsaturated field soil. Water Resour. Res. 27, 2533-2541. Saffigna, P. G., Tanner, C. B., and Keeney, D. R. (1976). Nonuniform infiltration under potato canopy caused by interception, stemflow, and hilling. Agron. J . 69, 251-257. Sallam, A., Jury, W. A., and Letey, J., Jr. (1984). Measurement of gas diffusion coefficient under relatively low air-filled porosity. Soil Sci. SOC.A m . J. 48, 3-6. Schulin, R., Van Genuchten, M. Th., Fliihler, J., and Ferlin, H. (1987). An experimental study of solute transport in a stony field soil. Water Resour. Res. 23, 1785-1794. Schumacher, W. (1864). “Die Physik des Bodens.” Berlin. Schwille, F. (1988). “Dense Chlorinated Solvents in Porous and Fractured Media: Model Experiments,” Engl. Lang. ed. Lewis, Chelsea, Michigan. Scotter, D. R., and Kanchanasut, P. (1981). Anion movement in a soil under Pasture. Aust. J . Soil Res. 19, 299-307. Simmons, C. S. (1982). A stochastic-convective transport representation of dispersion in one dimensional porous media systems. Water Resour. Res. 18, 1193-1214. Skopp, J., Gardner, W. R., and Tyler, E. J. (1981). Solute movement in soils: Two region model with small interaction. Soil Sci. SOC. Am. J . 45, 837-842. Spencer, W. F., and Cliath, M. M. (1970). Desorption of lindane from soil as related to vapor density. Soil Sci. SOC.A m . Proc. 34, 574-578. Sposito, G. (1981). “The Thermodynamics of Soil Solutions.” Oxford Univ. Press (Clarendon), London and New York. Sposito, G., and Barry D. A. (1987). On the Dagan model of solute transport through groundwater: Foundational aspects. Water Resour. Res. 23, 1867-1875. Sposito, G., and Mattigod, S. V. (1980). “GEOCHEM: A Computer Program for the Calculation of Chemical Equilibria in Soil Systems and Other Natural Water Systems.” Kearney Foundation of Soil Science, University of California, Riverside. Starr, J. L., DeRoo, H. C., Frink, C. R., and Parlange, J.-Y. (1978). Leaching characteristics of a layered field soil. Soil Sci. SOC.Am. J . 42, 386-391. Stephens, D. B . , Parsons, A. M., Mattson, E. D., Black, K., Flanigan, K., Bowman, R. S . , and Cox, W. B. (1988). A field experiment of three dimensional flow and transport in a stratified soil. In “Validation of Flow and Transport Models for the Unsaturated Zone” P. J. Wierenga, and D. Batchelet, (eds.), Ruidoso, N. M. Res. Rep. 88-ss-04. New Mexico State University, Las Cruces. Stone, H. L. (1973). Estimation of three phase relative permeability and residual oil saturation. J . Can. Pet. Technol. 12, 53-61. Sudicky, E. (1986). A natural gradient experiment on solute transport in a sand aquifer: Spatial variability of hydraulic conductivity and its role in the dispersion process. Water Resour. Res. 22, 2069-2082. Taylor, G. I. (1953). The dispersion of soluble matter flowing through a capillary tube. Proc. London Math. SOC.2, 196-212. Thomas, G. W., Philips, R. E., and Quisenberry, V. L. (1978). Characterization of water flow in soils by simple chromatographic theory. J . Soil Sci. 29, 32-37. Utermann, J., Kladivko, E. J., and Jury, W. A. (1990). Evaluation of pesticide migration in tile-drained soils with a transfer function model. J . Environ. Quai. 19, 707-714. Valentine, R. L. (1986). Nonbiological degradation in soil. In “Vadose Zone Modeling of Organic Pollutants” (S. M. Hern and S. Meloncon, eds.), pp. 223-243. Lewis, Chelsea, Michigan.

TRANSPORT OF CHEMICALS THROUGH SOIL

201

Valocchi, A. J . (1985). Validity of the local equilibrium assumption for modeling sorbing solute transport through homogeneous soils. Water Resour. Res. 21, 808-820. Valocchi, A. J. (1986). Effect of radial flow on deviations from local equilibrium during sorbing solute transport through homogeneous soils. Water Resour. Res. 22, 1693-1701. Van Dam, J. (1967). The migration of hydrocarbons in a water-bearing stratum. In “The Joint Problems of the Oil and Water Industries” (P. Hepple, ed.), Inst. Pet., London. Van der Pol, R. M. P., Wierenga, P. J . , and Nielsen, D. R. (1977). Solute movement in a field soil. Soil Sci. SOC.A m . J . 41, 10-13. Van Duin, C. J., and van der Zee, S. E. A. T. M. (1986). Solute transport parallell to an interface separating two different porous materials. Wafer Resour. Res. 22, 1779-1789. Van Genuchten, M. Th. (1985). A general approach for modeling solute transport in structured soils. M e m - h i . Assoc. Hydrogeol. 17, 513. Van Genuchten M. Th., and Alves, W. J. (1982). Analytical solutions of the one dimensional convection dispersion solute transport equation. U . S. Dep. Agric., Tech. Bull. 1661, 1-151.

Van Genuchten, M. Th., and Cleary, R. W. (1979). Movement of solutes in soil. In “Soil Chemistry. Part B . Physico-Chemical Models” ( G . H. Bolt, ed.), pp. 349-386. Elsevier, New York. Van Genuchten, M. Th., and Wierenga, P. J. (1976). Mass transfer studies in sorbing porous media. I. Analytical solutions. Soil Sci. Sac. A m . J . 40,473-480. Van Genuchten, M. Th., and Wierenga, P. J. (1977). Mass transfer studies in sorbing porous media. 111. Experimental evaluation with tritium. Soil Sci. SOC.A m . J . 41, 272-278. Van Genuchten, M. Th., Davidson, J. M., and Wierenga, P. J. (1974). An evaluation of kinetic and equilibrium equations for the prediction of pesticide movement through porous media. Soil Sci. SOC. A m . Proc. 38, 29-34. Van Genuchten, M. Th., Wierenga, P. J., and O’Connor, G . A. (1977). Mass transfer studies in sorbing porous media. 11. Experimental evaluation with 2,4,5-T. Soil Sci. SOC. A m . J . 41, 278-285. Van Wesenbeek, I. J., and Katchanoski, R. G . (1991). Spatial scale dependence of in situ solute transport. Soil Sci. SOC. A m . J . 55, 3-7. Villermaux, J. (1987). Chemical engineering approach to dynamic modeling of linear chromatography. J . Chromatogr. 406, 11-26. Walker, G. R. (1987). Solution to a class of coupled linear partial differential equations. IMA J . Appl. Math. 38, 35-48. Warrick A. W., and Nielsen, D. R. (1977). Spatial variability ofsoil physical properties in the field. In “Applications of Soil Physics” (D. I. Hillel, ed.), pp. 319-344. Academic Press, New York. White, R . E. (1985). The influence of macropores on the transport of dissolved and suspended matter through soil. Adv. Soif Sci. 3, 95-120. White, R. E. (1987). A transfer function rnodel for the prediction of nitrate leaching under field conditions. J . Hydrol. 92, 207-222. Wierenga, P. J . , and Batchelet, D., eds. (1988). “Validation of Flow and Transport Models for the Unsaturated Zone,” Ruidoso, N. M. Res. Rep. 88-ss-04. New Mexico State University, Las Cruces. Wild, A , , and Babiker, L. A. (1976). The asymmetric leaching pattern of nitrate and chloride in a loamy sand under field conditions. J . Soil Sci. 27, 460-466. Yeh, T.-C. J., Gelhar, L. W., and Gutjahr, A . (1985). Stochastic analysis of unsaturated flow in heterogeneous aquifers. Wafer Resour. Res. 21, 447-472.

This Page Intentionally Left Blank

EVOLUTION OF CORN Walton C . Galinat Eastern Agricultural Center, Department of Plant and Soil Sciences, University of Massachusetts- Amhent, Waltham, Massachusetts 02 154

I. Introduction: Teosinte Is the Wild Corn 11. The Transformation by Domestication and Isolation A. The Separation by Key Traits B. Isolating Mechanisms C. The Temporary Role of the Tunicate Locus 111. The Time Required for Transformation Under Domestication A. The Power of Domestication B. A Considerable Period of Time, an Unnecessary Assumption C. Cataclysmic Changes of an Unknown Nature, a False Tenet IV. Multiple Domestications in the Origin of Corn A. The Evidence from Cob Morphology B. The Evidence from Plant Habit V. Interpathway Heterosis VI. The Husk Enclosure of the Ear A. The Product of Protogyny B. The Hormonelike Control VII. Current Direction of Corn Evolution and Where It Is Going A. The Role of Increased Femaleness B. Partitioning of Photosynthate between rhe Sexes VIII. Summary and Conclusions References

I, INTRODUCTION TEOSINTE IS THE WILD CORN Corn [Indian corn, or maize (Zea mays L. ssp. mays)] results from the most important and spectacular plant breeding achievement in human history. It is important as one of the three main staffs of life (with rice and

203 Advancer in Agronomy, Volume 47 Copyright 8 1992 by Academic Press, Inc. All rights of reproduction in any form reserved

204

WALTON C . GALINAT

Figure 1. This figure illustrates teosinte and two of its derivatives. (a) Teosinte is corn’s wild ancestor. The spike (ear) of teosinte has two rows of single female spikelets that alternate between opposite sides of the rachis (cob). Each kernel is enclosed by a protective fruitcase that is composed of the outer glume of its spikelet and a cupule within its rachis segment. A single husk usually encloses the entire spike. (b) The first corn, as reconstructed from its oldest remains from Tehuacan, Mexico. It has four ranks of paired spikelets resulting in an eight-rowed ear. The grain is released from its fruitcase by rachilla elongation, although the entire ear is enclosed by two or more husks. Like teosinte and primitive corn, the ear may have a terminal region that is male. (c) The modern ear of corn has many ranks of paired kernels. It is exclusively female and permanently enclosed in many husks. Because man must intervene to release and sow its seed, corn is incapable of survival as a wild plant. Adapted from Galinat (1988a), with permission.

wheat) and in terms of its coevolution with human culture. It is spectacular in terms of its wide divergence from the wild types regarding ear structure. The wide divergence was made possible by the monoecious condition (unisexual flowers with both sexes on the same plant) of both corn and its apparent wild progenitor, teosinte [Zea mays (L.) ssp. mexicana Iltis and ssp. pawiglurnis Iltis and Doebley] (Fig. 1). Unisexual flowers allow

EVOLUTION OF CORN

205

anthropocentric specialization in the female secondary sex traits when under domestication. In changing its adaptive design from one for survival in the wild to that of a partnership for survival with humankind as the agent for harvesting and planting seed, teosinte’s simple female ear, bearing only two ranks of single spikelets, evolved into corn’s unique nonshattering cob bearing many ranks of paired spikelets, with the entire structure trapped within the protective confinement of husk leaves. In contrast, hermaphroditic (bisexual) expression in mature inflorescences of the small grains and most other grasses greatly limits any such specializations in their secondary sex traits that would serve only one sex or the other. Domestication of these hermaphroditic grasses would not permit the evolution of a cornlike cob with its husk enclosure because the male parts (anthers) could not function in pollination when under tight confinement. This is the case with the nonfunctional anthers carried in hermaphroditic flowers on the ears of certain mutants in corn. That teosinte is the progenitor of corn is now generally accepted (Beadle, 1939, 1972; Doebley, 1983, 1990; Galinat, 1971, 1977, 1983, 1985a,b,c, 1988a, 1991a; Harlan and deWet, 1972; Iltis, 1972, 1983; Kato, 1984; Vavilov, 1931). Doebley (1990) considers the molecular evidence for at least one origin of corn by domestication of teosinte (ssp. parviglumis, race Balsas) as definitive, and when viewed with all of the other supportive evidence, I can only add the conclusion is beyond question. The sound assumption of a teosinte origin for corn will be made here in order not to unnecessarily dwell on the arguments, but there was a period of 35 years when this view was not popular (Mangelsdorf and Reeves, 1939; Mangelsdorf, 1974). The pros and cons of the arguments have been reviewed extensively (Goodman, 1988; Beadle and Galinat, 1991) and they are primarily of historical interest. In particular, I want to call attention to a previous (Galinat, 1988a) work on the origin of corn, which parallels this article.

11. TRANSFORMATION BY DOMESTICATION AND ISOLATION The transformation of teosinte into corn started more than 8000 years ago, when the first native Americans discovered that they could copy and accelerate the forces of natural evolution through the physical separation of selected plant types; geographic isolation was accomplished by practices such as irrigated garden plots in arid locations or isolated plots from slash and burn farming where the original wild population did not exist. Changes that might take thousands of years in nature, if they could happen

206

WALTON C.GALINAT

at all, would occur in only a few hundred years when directed by the mind, eye, and hand of humans acting to ensure their own survival. Under imposed isolation where they were selectively planted, and harvested based on purposeful nurtured, human intervention, these crops would avoid genetic “swamping” by the wild populations and, thereby, would not lose their domestic advantages. It is not as if the transformation of teosinte into corn involved a series of improbable mutations, as would be the case in the slow process of natural evolution, but rather involved a situation wherein an observant, intelligent people selected the most useful variants out of large populations of teosinte in the wild and then concentrated them into isolated, evolutionary pools. All that they had to observe and act upon was that offspring tended to resemble their parents. Man transported a primitive type of corn not much different from the oldest Tehuacan corn to South America some 4000 or more years ago, where it freely diverged in complete isolation from teosinte. The interspace trait due to interspace (is) gene character of the first Tehuacan corn survived as a relic trait in “Coroico” and related races under the hand of the Guarany Indians on the eastern slopes of the Andes. The slender (string) cob due to the Sg I and Sg 2 genes with their short rachillae and reduced cupules found in some of the early Tehuacan corn survived in “Confite Morocho” from Peru. There was a delay of 2000 or more years before corn could spread into most of the area now comprising the United States because of its lack of adaptation to longer daylengths and shorter growing seasons, and eventually to a lack of human technical means to plow the dense prairie sod in the area now occupied by the United States corn belt (Galinat, 1985a). Isolation resulting from such geographic diversity has always played an important role in both natural evolution and evolution under domestication. Because of the widespread distribution of corn with the spread of farming, corn had differentiated into over 300 races by the time of Columbus. The differences between these races are mostly quantitative ear traits involving gradual changes in productivity and adaptation, in contrast with the few qualitative ear traits that are simply inherited and may rapidly transform teosinte into corn under domestication. The taxonomy of the races of corn in Latin America has been described in a series of publications over a 25-year period starting geographically with Mexico (Wellhausen et at., 1952); this has been reviewed by Goodman and Brown (1988). These races gradually became adapted to numerous sites far from corn’s original home, apparently southern Mexico, as far south as southern Argentina, and as far north as the Gasp6 Peninsula of Canada. Isolation into these diverse habitats together with the selective preferences imposed by man resulted in the present incredible array of races, each of which

EVOLUTION OF CORN

207

acquired certain adaptive, chromosomal, and genetic characteristics. This enormous diversity, now extant in germplasm banks, is the rich raw material that is available to the modern corn breeder. Now the genetic isolation is brought about through the use of paper bags by the breeder to control the pollinations, and the selection of improved designs is directed by the power of human intelligence and imagination. In considering teosinte as a food plant, variants that would double and then quadruple its productivity per ear could hardly escape the eye, mind, and hand of America’s first native peoples. The key traits of corn are not only nonadaptive in the wild, but their total combination is lethal in nature. Corn’s key traits must have been selected out of natural variation within teosinte and transported into isolation, for example, in irrigated garden plots, in the Tehuacan Valley. The first introduction to the Tehuacan Valley may have been in trade of salt for grain, and the grain was then grown by a desperate people who would invent irrigation.

A. THE SEPARATION BY KEYTRAITS To the traditional three key traits separating corn from teosinte, which I and others have described before (e.g., Galinat, 1983, 1988a) and which will be reviewed only briefly here, I shall now add a fourth trait involving a complex block of genes partially controlling the fruitcase enclosure versus naked exposure of the kernel. In an effort to keep the list of keys to a minimum, background factors such as interspace (is) and string cob (Sg I and Sg 2) are not included. The new four-part listing of keys is as follows:

1. Solitary female spikelets in teosinte in comparison with paired ones in corn. 2. Two-ranked central spike in teosinte in comparison with a manyranked one in corn. 3. Shattering rachis (cob) in teosinte in comparison with a nonshattering one in corn. 4. Fruitcase-enclosed kernel in teosinte in comparison with an exposed kernel in corn. Kernel exposure may occur by two different systems, as described later. 1. Inheritance of Solitary Versus Paired Female Spikelets

The reactivation of the suppressed second female spikelet in teosinte immediately doubled the productivity of grain per spike and, thereby, made it more adaptable as a human (and bird) food plant and less adapted

208

WALTON C . GALINAT

to natural survival. The pairing of spikelets opens up the fruitcase, giving easier access to the grain. Protection replacement evolved later from a husk enclosure system of the ear as a whole that requires human intervention to release the seed. In some segregations from corn-teosinte hybrids, the single female spikelet is expressed as a single recessive gene ( p d ) , as is known from the early studies of Collins and Kempton (1920), and its most commonly agreed-upon location is on chromosome 3 (Langham, 1940). Other reported locations (chromosomes 4, 7, and 8) apparently result from the interactions of other genes (Mangelsdorf, 1947; Rogers, 1950). In many key trait derivates, stability to p d expression is associated with the interspace and string cob traits (Galinat, 1991b). It has even been suggested that the dominant allele ( P d ) of p d had no role in the origin of the corn ear because, according to this hypothesis, the paired female spikelets came from transmutation of the teosinte tassel, which already has a paired condition of its male spikelets, into the corn ear (Iltis, 1983). Because both tassel and ear diverge developmentally from the same primordial inflorescence, not one from the other, the suggestion of Iltis (1983) is not valid. The solitary spikelets of the teosinte ear are not removed by removing the teosinte ears, but by changing the gene that produces them.

2. Inheritance of Two-Ranked Versus Many-Ranked SpikeIets The proliferation into four or more ranks of spikelets was another system for at least doubling the productivity per spike. When this system was combined with that of pairing, the productivity per spike was at least quadrupled. At this point, humanity obviously had a new and important food plant. But it was not a one-time event. Domestication would be in many stages simultaneously in the hands of many farmers of various abilities. Two-ranking versus four-ranking (decussate) is controlled by a single recessive gene (tr) when expressed in the central spike of modern eightrowed corn. But two-ranking also normally occurs in the lateral branches of a tassel with a many-ranked central spike and in the secondary ears on a plant with a many-ranked main ear, or just in stunted corn. It is those who favor a wild corn (i.e., Mangelsdorf, 1986; Wilkes, 1986) this time who would also remove a portion or a whole inflorescence (the central spike of the tassel or the main ear) in order to derive, and in my mind ridiculously so, teosinte out of corn. The tr gene is located on the short arm of teosinte chromosome 2, as shown in the data of Rogers (1950), and confirmed with interchanges be-

EVOLUTION OF CORN

209

tween corn chromosome 2 and its partial homologue Tripsacurn chromosome 9, which carries the short loci of Zea chromosome 2, including rr (Galinat, 1973). In segregations for molecular markers, the data of Doebley et al. (1990) agree in placing a major gene controlling “RANK” on the short arm of chromosome 2. A minor position for tr on chromosome 1 was also reported by Langham (1940) and Rogers (1950), but in my material this factor just involves reduced condensation, a factor known to be associated with changes in kernel row number (Anderson and Brown, 1948). 3. Inheritance of Shattering Rachis (Cob) Versus a Nonshattering One

Selective pressure against abscission layer development, or shattering, is unconscious and automatic when harvesting is at the whole-ear level in corn or any other cereal, such as the small grains. Even when harvesting is performed by beating a plant against an animal skin, intact ears are favored, as the whole ear flies into the skin. With shattered ear types, there are losses before, during, and after the operation that tend to reduce their frequency in harvested grain. Nonshattering is somewhat complex as a key trait separating corn from teosinte. It involves at least two and sometimes three components, each under control of a different incompletely dominant gene. Rind abscission extends through the rind at the node between the apex of the cupule and the divergence of the glume cushion above. When an ear with just rind abscission is snapped in cross-section, the abscission layer looks like a donut with white pith at the center. When pith abscission is added, the whole in the donut is lost. Pith abscission (Ph) appears to be linked to rind abscission (Ri) far out near the end of the short arm of chromosome 4 (Galinat, 1975b). There is a distinct layer across the pith in cobs of the Durango teosinte chromosome 1derivative of A158. This gene, designated ab, is separate from Ph and stronger in expression. Another factor that prevents shattering is really a condensation factor. It holds the rachis together by a fusion of the apex of the cupule to the glume cushion above. As heterozygotes in this back-cross to a su gl tester, the expression of Ph and Ri was poor and more frequent toward the tips of the ear, where condensation is slightly relaxed. The nonshattering rachis becomes semilethal in teosinte because it inhibits seed dispersal. But the reciprocal condition, partial abscission layers in the corn cob, may be tolerated because the comparatively high level of condensation in corn prevents complete rind abscission through a fusion of the apex of the cupule to the glume cushion above. Pith abscission is

2 10

WALTON C . GALINAT

ineffectual in the absence of complete rind abscission, so modern corn can cope with some gene flow from teosinte for either one alone of these two abscission factors because of their individual incomplete effect on the corn cob shattering.

4. The Chromosome 4 Difference That the often-described block inheritance on the short arm of chromosome 4 (4s) controlling certain difference between the ears of corn and teosinte might have originally started with a monogenic change has been suggested by Kermicle (Doebley ef al., 1990). Apparently they were considering a single mutation with pleiotropic effects rather than a compound locus, if not a polygenic complex. The most obvious effects of the 4 s factor(s) from teosinte upon corn, as described by Mangelsdorf and Reeves (1939), Rogers (1950), Sehgal (1963), and Galinat (1963), include an upward inclination of a short rachilla, cupule enlargement, and an induration of both outer glume and cupule, all of which are important to the teosinte fruitcase. The effects of the counterparts to these traits from all or a portion of corn 4s upon kernel exposure and threshability of teosinte have not been studied, despite the fact that the teosinte background is essential to analyze a teosinte-to-corn transformation. An assumption that corn 4 s alone will transfer the corn glume character and rachilla elongation and relaxation into teosinte is probably not valid. The teosinte background would still carry other genes for glume induration on chromosomes 1and 7 spikelet inclination on chromosome 1, and other modifying genes that alter the expression of corn 4s (W. C. Galinat, unpublished data). There remain numerous questions about 4s to resolve. Did its importance evolve just because there happened by chance to be a cluster of floral genes in this area? Did the block inheritance result from recent postdomestication evolution as a means to protect teosinte against harmful corn introgression that would destroy its fruitcase? In addition to tight linkage, the integrity of the complexes may be enhanced by inhibitors to crossing-over, such as cryptic rearrangements and chromosome knobs, or ,by close linkage with gametophyte genes. In the latter case, the Guatemalan teosintes are known to carry a gametophyte (Ga) allele within their chromosome 4 complex (D. Duvick, personal communication, 1984). More recently other teosintes (2. mays ssp. rnexicana, races Central Plateau, Chalco, Amecameca, and Los Reyes, but not Nobogame, as well as one accession of ssp. huehuefenangensis) were all found to carry some type of Ga system on 4 s which would serve to protect its linked teosinte segment against corn introgression. Kermicle and Allen (1990) note that one of these, the TIC-CP (teosinte incompatibility to corn-Central

EVOLUTION OF CORN

211

Plateau) system, is unknown in corn and, therefore, it may belong solely to teosinte as a protector of its associated segment if not of genomic integrity in general. It is of interest that the miniature teosinte of ssp. parviglumis, race Balsas, which Doebley (1990) places closest to corn, does not carry a 4s Ga system for protection against corn. Perhaps its small size is adequate. According to the linkage data of Rogers (1950), the key traits of the Guatemalan (Florida) teosinte do not have as tight a control by block inheritance as do the Mexican teosintes. In the case of this teosinte, it would seem, therefore, that evolution for block inheritance may still be underway. Investigations into the nature and evolution of teosinte 4s into corn 4s involves many aspects. Corn 4 s or various known components of it could be transferred to teosinte in order to determine their domestic value. The teosinte 4s could be dissected and then its parts reconstructed in a corn background with suitable marker genes on either side in order to reveal its components as reciprocal cross-overs. The breaking of tight linkages requires large populations when studied with conventional marker genes segregating at the whole plant level, which makes the analysis difficult. Such a study may be more amenable to analysis with molecular markers. I have isolated the chromosome 4 difference several times with classical marker genes and some of these have different effects, suggesting that crossing-over had occurred between independent factors within the segment for glume induration and spikelet inclination. This may happen unconsciously with small populations. Intercrosses between them are being made. Other essential traits controlled on the short arm of chromosome 4 include rind and pith abscission near the end of the arm (Galinat, 1975b) and apparently genes involved with rachilla elongation and cupule development. The historic tunicate locus lies down on the long arm of the same chromosome. Even the strongest allele (Tu) at this locus, which usually has monstrous effects, was once considered as representing wild corn (Mangelsdorf and Reeves, 1939; Mangelsdorf, 1948), a viewpoint to which Mangelsdorf adhered for over 30 years (Mangelsdorf, 1974).

B. ISOLATING MECHANISMS The various isolating mechanisms that protect teosinte against harmful introgression by corn gennplasm have already been described in detail (Galinat, 1988a). Briefly, they are (1) isolation by differences in flowering time, (2) isolation by geographic separation, and (3) isolation by block inheritance, including tight linkage and gametophyte factors.

212

WALTON C. GALLNAT

C. THE TEMPORARY ROLEOF THETUNICATE Locus Apparently a weak tunicate allele (tu w) played a temporary role as an attractant to domestication during at least one of corn’s origins, but it was usually discarded in favor of the chromosome 4s factors for reduced induration and rachilla elongation and downward relaxation. This tu w allele still lingers on in at least one primitive race, Chapalote, which has evolutionary roots tracing back to the oldest Tehuacan corn. The tu w allele derived from Chapalote does have value in vestigial glume corn by its modifying action in restoration of the tassel glumes to Vg plants, thereby allowing pollen production on VgVg plants with glumeless ears (Galinat, 1966). The tu w allele produces soft, slightly elongate female glumes together with slightly elongate rachillae. They are broader but not as long as the male glumes borne on the same plant. They could not be a product of transmutation by a male to female secondary sex trait as implied in the Iltis hypothesis. The tu w allele is known to have similar effects upon teosinte in producing “soft-shelled” fruitcases with partly emergent kernels that are freely threshable, as envisioned by Emerson and Beadle (Beadle, 1972). Genetic cross-over analysis with su gl 3 show that this allele is at the tunicate locus (Mangelsdorf and Edwardson, 1953).

III. TIME REQUIRED FOR TRANSFORMATION UNDER DOMESTICATION Any disagreements that remain over the origin of corn seem to revolve around the time frame available and that required for the transformation of teosinte into corn.

A. THE POWER OF DOMESTICATION The power of domestication in rapidly selecting new types is described by Darwin (1859) as “that which enables the agriculturist, not only to modify the character of his flock but to change it altogether. It is the magician’s wand, by means of which he may summon into life whatever form and mold he pleases.” Darwin (1859) elaborates Domestic races often have a somewhat monstrous character; by which I mean, that, although differing from each other, and from other species in the same genus, in several trifling respects, they often differ in an extreme degree in some one part, both when compared one with another, and more especially when compared with the species under nature to which they are nearest allied.

EVOLUTION OF CORN

213

The sudden appearance of a new exotic nitch such as domestication provides an opening for extremely rapid evolution, especially for annual plants such as teosinte, where there is (1) a short generation time, (2) abundant variability, and (3) isolated places under the eye, hand, and mind of farmers to direct the evolution. But some traditional taxonomists and even some evolutionists, in unexpected contradiction with Darwin’s observations and the basic understanding of both plant breeders and geneticists, do not accept the rapidity of change in the case of the origin of corn by domestication of teosinte. For example, Mangelsdorf (1947) states “If maize has originated from teosinte, it represents the widest departure that comes within man’s purview,” and then suggests that “One must indeed allow a considerable period of time for its accomplishment, or one must assume cataclysmic changes, of a nature unknown, have been involved.”

B. A CONSIDERABLE PERIOD OF TIME, AN UNNECESSARY ASSUMPTION The assumption of a considerable period of time as one of the Mangelsdorf options for a teosinte-corn relationship was picked up by Goodman (1988) as “involving multiple events long before the era of plant domestication” and earlier by Randolph (1972, 1976), who would even retain the obsolete generic name of Euchlaena for teosinte. Doebley (1990) would like to have a considerable period of time for “a series of improbable mutations” as if he were thinking of natural evolution instead of his stated support for domestication. He would gain extra time by removing the time frame set by the known archaeological record. Accordingly, Doebley (1990) accepts a reevaluation of the age of the oldest Tehuacan specimens by Bruce Benz, who claims they are 2000 years younger than previously thought (Benz and Iltis, 1990).

C. CATACLYSMIC CHANGES OF AN UNKNOW NATURE, A FALSE TENET The assumption of a cataclysmic change, of a nature unknown, as an option for a teosinte origin of corn by Mangelsdorf was picked up by Iltis (1983) in his CSlT (cataclysmic sexual transmutation theory) theory by which a macromutation would transmute the teosinte tassel into the corn ear in one giant step. The theory was intended to show instant corn production and thereby explain the lack of connecting links, in the archaeological records, that are older than the oldest known corn, and the fact

2 14

WALTON C . GALINAT

that most of the earliest archaeological specimens already had the key traits of corn (Mangelsdorf, 1974). The CS?T theory has been shown to be invalid, although the basic idea of Iltis, of a change in sexual timing, is correct, just the direction is wrong. Domestic transformation does not take long, especially when under the direction of a hungry, desperate people (Beadle, 1939, 1972, 1980; Galinat, l971,1975a, l977,1978,1983,1985a,b,c, 1988b). Many farmers on isolated farms would have precorn or protocorn simultaneously. At harvest time, the exchange of the best ear types would be as much a matter of pride as the exchange of cooking recipes, recipes for both evolution and cooking. Trade takes many forms, such as salt for grain and even as plunder from warfare. The time required to transform teosinte into corn would probably take less than 400 years and perhaps as few as 100 years. The variability was present, especially in parviglumis teosinte. It is an interesting fact that, in collecting this teosinte at Mazatlan, Mexico, in November, 1971, L. F. Randolph identified a spike with kernels elevated out of the fruitcase by rachilla elongation (unpublished data). This would be a promising first step in the origin of corn, but the reaction of Randolph was that “it must be from corn introgression.”

N.MULTIPLE DOMESTICATIONS IN THE ORIGIN OF CORN The concept of multiple domestications during corn’s origin is old (McClintock, 1959; Randolph, 1959; Mangelsdorf, 1974). Kato (1984) added the idea that there were at least five separate domestications from different teosintes on a basis of five different chromosome knob complexes, each of which is a common fingerprint between a given race of corn and its sympatric race of teosinte (Kato, 1976). A modified interpretation of the five different knob complexes has been suggested by Galinat (1988a) whereby the common knob patterns stemmed from adaptive introgression after primitive corn was moved into the teosinte’s habitat. Not everyone agrees that there were multiple domestications of different teosintes. Doebley (1990) interprets his data on molecular comparisons of corn with the different teosintes as evidence that there was just a single teosinte progenitor, although this concept does not exclude repeated domestications of this single teosinte. He identifies this common ancestor to all races of corn as the miniature teosinte, 2. mays ssp. parviglumis (Iltis and Doebley, 1980) or, according to Wilkes (1967), 2. mexicana, race Balsas. Although I accept this teosinte as a wild ancestor of the first corn on the grounds of its miniature fruitcase in comparison to the rachis segments of the oldest know archaeological corn, as I discussed in 1979

EVOLUTION OF CORN

215

(Galinat, 1979) in the Maize News Letter, I do not now think it is the only teosinte ancestor of corn. The study of Doebley does not include South American corn, of which the highland corn such as Confite Morocho appears to me to be related to Palomero Toluqueiio of Mexico in cob morphology (reduced cupules and short rachillae) and cytology (reduced knobs), with the geographic divergence occurring at some early time before the former lost its hairiness and the latter acquired its fasciation and its extreme pubescence. The hairy sheath trait does occur in Chalco teosinte but it is not as strong as in Palomero Toluqueiio. The isozyme studies of Bolivian corn by Goodman and Stuber (1983) were able to separate the lowland Coroico types from the highland Andean complex of corn, although they did not include teosinte. Doebley (1990) places all of the races of Mexican corn off to the right of 2. mays ssp. mexicana and partly above but mostly to the right of ssp. parviglumis, with a few accessions of corn intermediate between the two teosintes. Although none of the corn tested is currently within the range of ssp. mexicana, the shift of most corn to the right of even parviglumis suggests that shift is a product of domestication and at one early time there was considerable overlap with mexicana. Unless more molecular data excluding a double domestication of different teosintes become available, one must continue to take my data on cob and plant morphology as well as that of Kato on chromosome knob complexes seriously. My data may not exclude independent domestications of the same teosinte, but they are consistent with domestication from these two different subspecies of teosinte. Because almost everyone seems to agree that corn did not originate in a single macromutation, it also seems probable that many independent farmers were involved and they had various abilities to manipulate the few key domestic genes within teosinte into producing the corn phenotype. Thus, many steps and pathways probably existed simultaneously. This scenario is different from Doebley (1990), who concluded “the conversion of teosinte into maize is so improbable that it is difficult to imagine that it happened several times.” The evidence from both cob morphology and from plant habit of at least two independent domestications of teosinte has been described by Galinat (1988a, 1991a). A review of this new evidence is presented here.

A. THE EVIDENCE FROM COBMORPHOLOGY Apparently two very different morphological systems evolved independently to expose the kernel from within its teosinte fruitcase during the domestic origin of corn. Although the two exposure systems are structurally different, they both achieve the same outcome of releasing the kernel for easy shelling (Fig. 2).

2 16

WALTON C . GALJNAT

400

New England: 6 0 0

- ---

OOO?

New M e x i c o : 1200 Peru:

1500

Colombia: Mexico:

2000 2500

-- - x l c o : 2500

New M e x i c o : 3 0 0 0

----

7 c

Mexico: 3000

I eticall

Tehuacen, Mex.: 7 2 0 0

Oaxeca, M e x . : 8 0 0 0

--- - - --

!x.: 8000

f Figure 2. Cross-sections from the ears of corn representing about 8OOO years of evolution after domestication. Exposure of the kernel from the fruitcase of teosinte appears to have been accomplished independently during two domestications, probably from different teosintes, as indicated by the pattern of heterosis. The age is given in years before present (B.P.). (a) Teosinte (Z. mays ssp. parviglumis, race Balsas) is ancestral to most corn; kernel exposure is by rachilla elongation, with the cupule remaining unreduced. (b) The oldest known corn from Tehuacan, Mexico (7200 years B. P.) usually had paired female spikelets, sometimes in four ranks (eight rowed). (c) An increase in cob diameter and the level of ranking, as with this 12-rowed ear of Chapalote, increased productivity. (d) When the larger vascular supply evolved in Chapalote, recombined with the less demanding eight-rowed ear, at first the surplus photosynthate became deposited as cob induration. (e) Following selection for increased kernel size, the photosynthate was redirected into larger endosperms as in the

EVOLUTION OF CORN

217

1. Kernel Exposure by Rachilla Elongation with the Cupule Unreduced

The cobs of most modern corn are thick and round in cross-section due to greatly elongated rachillae, sometimes erroneously called pedicels, borne in many ranks. The rachillae extend along the floor of welldeveloped cupules, which are collapsed down upon them and fused in a very hard and strong configuration (Fig. 3). The long female rachillae are due to the combined action of two recessive or incompletely dominant genes (sg I and sg 2), depending on the background (Galinat, 1969). They not only occur in most modern corn but also in an evolutionary pathway extending back to the oldest known corn cobs from Tehuacan, Mexico (Fig. 2). It has been proposed that this rachilla elongation represents one of two different systems for kernel emergence from the teosinte fruitcase that were basic to the origin of corn by the domestication of teosinte (Galinat, 1988a). 2. Kernel Exposure by Cupule Reduction with the Rachilla Remaining Short

Instead of exposing the kernel by thrusting it up out of the teosinte fruitcase by rachilla elongation, the remains of another system to accomplish the same ends are apparent in another racial grouping of corn, namely, reducing the enclosing fruitcase (Fig. 2). The best examples of this system remain in Confite Morocho of Peru and Palomero Toluqueno from the valley of Mexico. With a greatly reduced or sometimes a totally suppressed cupule, productivity in grain was achieved by evolving a long narrow kernel as in Gourd Seed or Shoepeg of the southeastern United Northern Flints of New England (1300A.D.) and the giant Cuzco cornof Peru (1500 A.D.). (f) Teosinte ( Z . mays ssp. mexicana, Chalco) appears to be ancestral to another pathway, with kernel exposure by cupule reduction and the rachilla remaining short. (g) A hypothetical race of Mexican eight-rowed corn, almost without cupules and with short rachillae, that is similar to the Peruvian race, Confite Morocho, known from archaeological remains about 4000 B. P. at Ayacucho (see k). (h) Palomero Toluquefio is an ancient indigenous race from the Mexico City area that is considered here to have evolved soon after domestication of Chalco teosinte. It has reduced cupules, short rachillas, and elongate kernels. (i) Pepitilla is clearly related to Palomero Toluqueno. It is ancestral in varying degrees to the Southern Dents, especially to Shoe Peg, with its narrow, deep kernels, and to Gourd Seed, with its broad, deep kernels. These races all have reduced cupules and short rachillas. (j) The Corn Belt Dent is known to be a hybrid between the Northern Flint (e) and Southern Dent (1) pathways. (k) Confite Morocho of Peru has some similarities to both the earliest corn at Tehuacan (b) and the hypothetical domesticate of Chalco (f). Adapted from Galinat (1988b), with permission.

218

WALTON C . GALINAT

Figure 3. Longitudinal sections of the female spike showing two exposure systems of the teosinte kernel during two domestications and increasing levels of condensation. (a) Tripsacurn floridanurn showing elongate internodes with interspaces about equal to internode length. (b) Nobagame teosinte showing reduced interspace causing a slight triangulation of the fruitcase. (c) Confite Morocho with a loss of interspace and cupule reduction resulting in kernel exposure by one system. (d) Inbred A158 with condensation resulting in a folding down of the cupule apex, together with rachilla elongation, thrusting the kernel outward in another, different system of kernel exposure. (e) Inbred B2 with cupules starting to close. (f) Inbred B2 with cupules fused floor to roof.

EVOLUTION OF CORN

219

States. Because of the near absence of the cupule and the short rachillae attached to it, these cobs are slender, hence the name “string cob.” By extracting Sg I and Sg 2 genes from Confite Morocho and transferring them to eight-row sweet corn, I developed a slender cob inbred, W401, which had been important in slender cob sweet corn hybrids such as “Candystick.”

B. THE EVIDENCE FROM PLANT HABIT When the plant habits from the two pathways of corn races based on cob differences in cupule and rachilla development are arranged in the same sequences, supportive evidence for this double origin of corn becomes evident. Starting from two different subspecies of teosinte, pawiglumis, race Balsas (Fig. 4a), and mexicana, race Chalco (Fig. 4d), there are two different branching patterns that evolve stepwise toward the Corn Belt Dent (Fig. 4g). The branching pattern starting with parvigfumis teosinte is basal. It has links represented in Fig. 4 as the indigenous flour corn of the American Southwest and upper Missouri areas (Fig. 4b) followed by the New England flint corn from Rhode Island (Fig. 4c). In this sequence the ear appears as if it were derived from an elevated tiller. This branching habit can be duplicated by the grassy tiller (gt) gene in a primitive background. It is described by Shaver (1967) in a modern corn background as causing the production of numerous small tillers at the base of the main culm that terminate in small female inflorescences (sometimes called “ground ears”). In the presence of teosinte germplasm, the gt gene produces elongate tillers at the base of the plant that terminate in male tassels (W. C. Galinat, unpublished data). In a parallel pathway starting with mexicana teosinte, the branching pattern is lateral. The links in Fig. 4 are represented by Palomero Toluqueiio (Fig. 4e) followed by the United States’ Southern Dents, represented by Gourd Seed (Fig. 4f). In this sequence the ear is either high or centrally located on the plant. The Corn Belt Dent (Fig. 4g), as a hybrid between these two pathways, has productivity refinements indicated here as reduced tassels and erect upper leaves. The mexicana branching habit is duplicated by the teosinte branched (tb) gene in a modern corn background. It is located on the long arm of chromosome 1 between the f and bm 2 loci (Burnham, 1961). It is described as producing many slender lateral branches usually ending in a one-spike tassel. Although it produces abundant pollen, it rarely develops seed. The reason for this is that the primary effect of the tb gene is to

Figure 4. The plant habits from the two pathways of corn races based on cob morphology lend suport for the double origin of corn.They form two series of plant habits that may be termed the basal branching type and the lateral branching type, both of which lead to the Corn Belt Dent (g). (a) Teosinte subspecies pawiglumis, race Balsas, under good growing conditions. Note the proliferation of tillers at the base of the plant. (b) The indigenous flour corn of the upper Missouri area is close to one type of the first corn, pre- or proto-Chapalote. Its ear is near the ground, where it comes from an elevated and condensed tiller. (c) The Northern Flint (Rhode Island Flint). The upper ear-bearing node has elevated to a central position on the plant, but a trail of its ascent remains. (d) Teosinte subspecies rnexicana, race Chalco. Note that the branching is lateral and each branch terminates in a tassel with ears borne below as on the main stalk of corn. (e) Palomero Toluqueiio. This multieared popcorn has a barren zone both above and below the ear-bearing region. (f) Southern Dent (Gourd Seed). The ear is high on the plant and tillers are suppressed. (g) The Corn Belt Dent. This hybrid between the Northern Flint and Southern Dent pathways shows extreme heterosis. It has productivity refinements indicated here as reduced tassels and erect upper leaves. These line drawings are adapted from my painting showing the same sequences in dramatic color.

EVOLUTION OF CORN

22 1

increase the level of maleness and from this flows the free elongation of the lateral branches. The tb or the gt gene may have a practical use to increase maleness in the pollen parent of crossing fields (W. C. Galinat, unpublished data). The question of how a single gene mutation such as the tb gene discovered by Burnham (1961) can undercut in one step a polygenic trait of increased femaleness, reduced lateral branches, and evolution of the husk enclosure of the ear will be treated later.

V. INTERPATHWAY HETEROSIS If the high productivity of the Corn Belt Dent is at least partly a product of the ancient double domestications of two different teosintes during independent origins of corn, as the morphological evidence from the cob and plant type suggest, then this insight would help in breeding for optimum heterosis . Interpathway heterosis has long been recognized and used systematically to increase productivity in the breeding of hybrid corn. It is just that the ancient roots of the heterosis have never been traced all the way back to a double domestication from different teosintes. According to the traditional system for breeding hybrid corn, various inbred lines were developed by repeated self-pollination, at first within the various open-pollinated varieties of Corn Belt Dent. Then the different inbreds were intercrossed and tested for maximum productivity in the Fl hybrids. At first the inbreds were so unproductive that the commercial production of hybrid seed had to be based on the double-cross system involving four inbreds, as invented by Jones (1917, 1918). But now with second or higher cycles of inbred development, the inbreds themselves have been upgraded in productivity to a point that a switch to the even more productive and uniform single-cross hybrids has been made. But the varieties of Corn Belt Dent were already known to be derived from hybrids between the Northern Flints and Southern Dents (Wallace and Brown, 1956). As might be expected, it was discovered that the most productive hybrids were coming from crosses between the most Northern Flint-like inbreds such as Mo 17 and C 103 out of a Lancaster background crossed by the most Southern Dent-like inbreds such as B 73 out of Iowa Stiff Stalk Synthetic, or before this Wf 9 out of Reid’s Yellow Dent. The parviglurnis to Northern Flint and the rnexicana to Southern Dent pathways have diverse derivatives now widely scattered from Canada to southern Argentina. The corn at these two hemispheric extremes had to

222

WALTON C. GALINAT

adapt to short growing seasons by evolving early flowering. Yet the corn from these extremes in geographic separation is different in other respects. In Southern Argentina, the early-flowering flour corn called “Coya” has reduced cupules and short rachillae, whereas its early-flowering counterpart from the Gasp6 Peninsula in Canada has pronounced cupules and elongate rachillae. These are the same two traits attributed to the two independent domestications of different teosintes. If the proposed interpretation that the heterosis of modern hybrid corn relates to these two different domestications, then, with this understanding, we are able to exploit these two sources of early flowering while still capturing the interpathway heterosis in breeding short-season dent hybrids. If the predicted changes in rainfall patterns in the United States follow global warming, then the corn belt may be moving further north where such early-flowering dent hybrids would be adapted and important.

VI. HUSK ENCLOSURE OF THE EAR The husk system is a condensed lateral branch that permanently encloses the ear terminating that branch. The development of these lateral branches is characteristic of the mexicana subspecies of teosinte and is reflected in the phenotype of the tb gene. The entrapment of the ear within the leaves of the branch that it terminates results from precocious female development (protogyny) that is associated with a high level of feminization, as explained later.

A. THE PRODUCT OF PROTOGYNY That the husk enclosure of the corn ear is a female secondary sex trait resulting from precocious female development has not always been readily understood or obviously apparent. The common but mistaken assumption has been that corn is protandrous based on the late date of silk emergence, which is delayed by about a week while the silks elongate in their effort to reach above their husk enclosure for pollen exposure. The result is a false type of protandry that is only an artifact of artificial selection for the unnatural husk enclosure that characterizes the corn ear. The inheritance of this false protandry cannot be treated as if it were a real sexual difference independent from husk length and style (silk) elongation or ability for style emergence, as in the studies of Landi and Frascaroli

EVOLUTION OF CORN

223

(1986). Nor can false protandry be used as if it were a real thing resulting from the suppression of teosinte ears during an origin of the corn ear from the teosinte tassel (Iltis, 1983). False observations can only lead to false conclusions. The husk enclosure is a female secondary sex trait that is expressed in the vegetative phase just below the ear. Its completeness as a protective device for the entire ear depends upon the prococious onset and duration of female development. For example, if the ear switches its sex expression from a lower female region to an upper male region, then at that point in time the internodes in the shank below elongate and tend to elevate the ear out of its husk enclosure. This appears to have been the situation with some of the oldest tassel-tipped ears from Tehuacan and in certain primitive lving races (e.g., Confite Morocho, Pollo, and Nal Tel). But when the ear is exclusively female, as in the advanced races of modern corn, then the shank and all of its internodes remain short as if precocious female development were producing an early and continuous feedback of an inhibitory hormone acting primarily on internode elongation during the vegetative phase. The result is that the ear becomes trapped within its own bud, which then forms a tight permanent enclosure of husk leaves. Genetic factors determine whether female expression will be prococious as in corn or latent as in toesinte, and then this has feedback that regulates internode elongation in the plant habit below.

B. THE HORMONELIKE CONTROL The apparent hormonelike nature of the female-generated feedback that produces the husk enclosure of the corn ear is suggested by the restoration of internode elongation in the shank following removal of the immature (baby) ear or by the genetic masculinizing of the primary inflorescencesteosinte branched (tb) gene. The external application of gibberellic aid (GA) to four genetically recessive dwarfs of corn was found to restore normal internode elongation (Phinney , 1956) and the same hormone spray was found to restore the normal monoecious condition to genetically gyneocious squash plants (Shifriss, 1985). An increase in the level of femaleness as measured by the frequency of female flowers in squash may be produced by spraying with Ethrel(2-chloroethylphosphoric acid), which apparently changes to ethylene in the plant. It not only increases the yield in cucumber and butternut squash but it may be used as a chemical aid in producing F1 hybrids in these crops (Coyne, 1970). This flexibility in shifting from one sex to the other in squash suggests that their flowers are potentially hermaphroditic. It is just a matter of activating or deactivating

224

WALTON C . GALINAT

the primordina of one sex or the other. The situation is similar to that in corn, as described next. The perfect-flowered primordial inflorescence common to both tassel and ear of corn is sometimes manifestly carried over as staminodes (abortive or vestigial stamens) in the ear and more frequently as pistillodes (rudimentary pistils) in the tassel. An enhanced form of pistillodes is called “tassel seed,” which is produced by mutant genes at six or more loci ( t s l , ts2, Ts3, ts4, Ts.5, and Ts6), whereas staminode control is only known at one locus, anther ear ( a d ) . Modern corn breeders, usually being more concerned with the ear than the tassel, select strongly against staminodes, although small pistillodes may slip by selective elimination. Pistillodes occur in the normal tassel, terminating the main stalk, when grown under short days and/or low light intensity. All of this only indicates the perfectflowered primordial inflorescence and does not indicate that one unisexual condition was derived from the other. In an evolutionary sense, the perfect-flowered primordial inflorescence of maize and its relatives also reflects the perfect-flowered mature inflorescence in most other grasses and especially in the related tribe, Andropogoneae (Weatherwax, 1935, 1954; Bonnett, 1953). According to 0. Shifriss (unpublished data, 1990), cucumbers, which are like corn in having both unisexual mature flowers on the same plant, also exhibit the initials of both sexes within all flowers at the primordial level. This agrees with the old general dogma that hermaphroditism was the ancestral condition in plants (Durand and Durand, 1984). Sex reversal may occur in either tassel or ear depending on the genetics and internal or external environment that enhance or inhibit one sex or the other. When complete sex reversal does occur, giving a mixed inflorescence such as in the upper ear or the lower tassel of both maize and teosinte, each unisex expression carries with it the typical secondary sex traits with no evidence of “transmutation” of secondary sex traits between the sexes. When basal and lateral branches in maize are terminated by mixed-sex inflorescences, they carry normal 100% female ears below, contrary to the expectations of CSTT,as has been repeatedly pointed out by many (e.g., Bill Tracy, unpublished data).

VII. CURRENT DIRECTION OF CORN EVOLUTION AND WHERE IT IS GOING It is well known that the future is an extension of the past, but there is reluctant appreciation of how the present is dependent upon the past for its raw material to build the future, and how insights gained from past experience will help in directing the future.

225

EVOLUTION OF CORN

A. THE ROLEOF INCREASED FEMALENESS The more advanced and productive races of corn have evolved a higher level of femaleness in which the optimal productivity comes from increases in both the number of female spikelets per ear and increases in kernel size. This increased female productivity comes at the expense of male productivity in that the number of male spikelets and amount of pollen production are reduced. This is a reversal from that of primitive corn and the wild populations of most teosintes, although a single teosinte plant may have 100 or more small inflorescences that are male or female or mixed. In corn as in other monoecious crops such as in the Cucurbitaceae, the increased femaleness was associated with selection for increased productivity. As higher levels of femaleness increased productivity in corn, the ear became increasingly precocious, with the consequence that it also became more deeply entrapped within husk leaves and, thereby, was unable to disperse its own seed without assistance from humans. Although the key traits of corn, including paired female spikelets and many ranks of spikelets, are factors for increased productivity in that each one doubles the yield of grain per spike, they are inherited independently from increased feminization, rather than being a product of it, as proposed by Iltis (1983). According to the hypothesis of 11th (1983), the genes pd-Pd for paired versus single female spikelet had no role in the origin of the corn ear because the toesinte tassel has paired spikelets and, according to his hypothesis, the first ear of corn was a sexually transmuted tassel. But teosinte tassels feminized by either tassel seed genes or just by terminating short tillers carry only single female spikelets, not paired ones as expected by Iltis (1983). Increased feminization has been important to the origin and productivity of corn but not in the form suggested by Iltis (1983). But to know how to extend this part evolution into the future, we must discover and appreciate the pathways of corn’s history, including corn’s origins and the genetics of ear morphology that resulted in the present situation. It may be advantageous with some of the elite material now in commercial use to recycle some of corn’s ancestral germplasm in order to recover some lost trait or to enhance an existing trait, for example, the combining ability to produce productive hybrids. To extend the increased productivity of the large ears of modern corn, which is associated with increased levels of femaleness, we may have to devise genetic techniques to cope with reduced tassels in the male parent of hybrid crossing fields, such as certain ramosa genes or the tb gene previously mentioned.

B. PARTITIONING OF PHOTOSYNTHLTE BETWEEN

THE SEXES

Before the origin of corn, the teosinte progenitor, then as now, made a larger reproductive investment of resources in male inflorescences (tassels)

226

WALTON C. GALINAT

and their pollen production than it did in female inflorescences and their kernel development. The tassels of teosinte terminate not only the main stalk but also all primary branches, which become elongate and efficient pollen dispensers. The high male/female ratio is adaptive to teosinte because seed set depends upon wind for cross-pollination between scattered and sometimes widely separated plants. In many wild species, evolution favors out-crossing between widely separated and more masculine plants because such pollinations tend to produce more vigorous and successful hybrid offspring. Under domestication, the hybrid vigor of outcrossing is ensured and controlled by the hybrid seed corn companies, first by selective inbreeding and then by selective hybridization of two good combining parents, usually in the 4/1 female/male planting arrangements of a crossing field, which is sometimes modified to 6/2 female/male arrangement, this being more amenable to mechanization. Teosinte has two general branching patterns, one lateral and the other basal, and each results from a higher level of maleness than that of corn, as described earlier, under biphyletic domestications. As with other wild monoecious plants that came under domestication, human selection for increased productivity brought an increased feminization and an increase in female inflorescence size and a reduction in male inflorescence number and/or size. Factors that favor the evolution of one large ear per corn plant include (1) the free intraplant movement of resources (photosynthate), (2) a decrease in total construction costs for both the husk systems and cobs per plant, with fewer but larger ears, and (3) a decrease in tassel size and pollen costs per unit seed set, with the conserved resources partitioned into constructing larger cobs. These factors agree with one of the theoretical models for gamete packaging strategies for plants, in general as set forth in mathematical terms by Schoen and Dubuc (1990). With decreased expenditures on the husk system, protection for the longer ear may be inadequate under factor (2), indicating that a compromise of two or more shorter ears with better protection may be a more productive strategy. Under factor (3), a sudden one-step reduction in tassel size with corresponding increases in ear productivity may be achieved by the introduction of the unbranched (ub) tassel gene. With the extreme expression of ub, the tassel is completely without branches, but usually ub tassels have one or two short branches at the base of the central spike. Reduced tassels may be inadequate to function as males in crossing fields, but solutions are possible. Both factors (2) and (3) are fixed costs that are drawn from one initial pool of resources. A secondary pool of resources that accumulates and becomes available over the next 2-3 weeks contributes to kernel development and cob induration.

EVOLUTION OF CORN

227

VIII. SUMMARY AND CONCLUSIONS Research to segregate and isolate the key trait differences between teosinte and corn in isogeneic backgrounds has already enabled us to (1) determine the minimum number of genetic changes that are essential to convert teosinte into corn, (2) determine the inheritance and chromosomal location of these genes, (3) determine the modifying effects of background genes in shaping the expression of the key trait genes, and (4) determine from the past and current direction of corn’s evolution as to where it may be going from here, and which loci and genes will be useful in directing the future extension of this evolution. That only four or five inherited key trait units separate teosinte from corn is based on the recovery rate of parental types in F2 segregations (Collins and Kempton, 1920; Mangelsdorf and Reeves, 1939; Beadle, 1980). One viewpoint holds that these inherited units are clusters of linked genes that eventually evolved their partial isolating mechanisms through tight linkage, cross-over suppressors such as cryptic rearrangements, and chromosome knobs, or by close linkage with gametophyte genes. The Ga silks of teosinte would have selective protection against ga corn pollen. The evolution of some mechanism for segmental isolation would occur within teosinte in order to conserve certain essential combinations of genes from breaking up through crossing-over and recombination with the domestic alleles that would reduce natural survival (Galinat, 1988a). The corn counterparts to these blocks do not require such protection because the fitness of the corn ear is in the eye and mind of humans. The best example of such a teosine segment is on the short arm of chromosome 4, where there are genes controlling spikelet inclination, rachilla elongation, outer glume and cupule induration, and, near the end of the short arm, rachis and pith abscission. In addition to this partial segmental isolation, other mechanisms include differences in flowering time and geographic isolation. The location of the two-ranking (rr) gene compared with its manyranking allele in a background of eight-rowed corn is on chromosome 2, but tr expression may be suppressed at higher kernel row numbers. The p d (single female spikelets) gene is probably located on chromosome 3, but its expresssion is unstable in the presence of a thick and/or condensed cob. The expression of abscission layer development is both unstable and complex in a background of modern corn. In addition to the genes on the short arm of chromosome 4, condensation controlled on chromosome 1 may result in nondisarticulation by a fusion of cupule apex to glume cushion just above. The near failure of internodes to elongate in the cob of most modern corn results in an unstable expression of the p d and rr genes. With the tight

228

WALTON C. GALINAT

juxtaposition of spikelets, small fluctuations in the internal-external environment result in either a proliferation or a reduction in spikelet pairing and/or ranking. The exception to this instability is found in certain corn that became isolated in South America at an early time. These are Coroico and related corns from the Bolivian lowlands, which carry the primitive interspace (is) gene, and Confite Morocho of Peru in the highland Andes, which carry the string cob (Sg I and Sg 2) genes. These genes remove all tight condensation from the female spikelets, and thereby allow stabilization of p d and tr expression. It is significant that these background genes occur in teosinte and apparently in the oldest known archaeological corn. In this respect, the oldest corn cobs are connecting links to teosinte. In developing key trait stocks for genetic and molecular analysis, the background should contain the is, Sg I and Sg 2 genes in order to provide stable expression of the key trait genes. Because the hypothesis under analysis is that corn evolved out of teosinte, ideally we would be studying the key trait genes of corn in a teosinte background. But the traditional genetic and molecular markers are all in corn and the plant habit of corn is easier to manipulate in controlling pollinations, and so the modification of corn with these primitive genes is a reasonable compromise. Increased femaleness started with the domestication of teosinte and has continued throughout the millennia of corn improvement in productivity. The consequence in recent years has been reductions in tassel development to the extent that they tend to be inadequate as pollinators in seed production crossing fields. The increased femaleness has not only resulted in larger, more productive ears, but the earlier precocious thrust into rapid pistil development (protogyny) has resulted in a deeper, more permanent entrapment within the husk leaves of the branch that the ear terminates. When this husk system for ear protection functions in cooperation with man as the agent for seed harvesting and dispersal, it is man who usually gets the food as his reward instead of the birds and worms. Because the increased femaleness, productivity, and adequate husk coverage of the corn ear all appear to have a hormone basis involving feedback of a female-generated signal to terminate further development in the underlying vegetative phase, it behooves us to obtain a better understanding of its genetic and physiological basis for the continued improvement of corn both by standard plant breeding techniques and by the laboratory techniques of biotechnology. The possibility of a double domestication of two different teosintes cannot be dismissed yet on the basis of the molecular comparisons made so far. In particular, the corn of the Andean highlands has to be compared directly to 2. mays ssp. mexicana, race Chalco, as well as ssp. parviglumis. The latter, according to interpretations of the molecular data of Doebley (1990), is the only teosinte progenitor of all corn.

EVOLUTION OF CORN

229

ACKNOWLEDGMENTS This paper is from the Massachusetts Agricultural Experiment Station, University of Massachusetts at Amherst. The research was supported in part from Experiment Station Hatch Project No. 566 (NE-124), from CRGO Grant 88-37261-3541 (Hatch-8801056), and a grant from Pioneer Hi-Bred International. The typing was done by Mrs. Lorraine Daley.

REFERENCES Anderson, E., and Brown, W. L. (1948). A morphological analysis of row number in maize. Ann. Mo. Bot. Gard. 35, 323-336. Beadle, G. W. (1939). Teosinte and the origin of maize. J. Hered. 30, 245-247. Beadle, G. W. (1972). The mystery of maize. Field Mus. Nat. Hist. Bull. 43(10), 2-11. Beadle, G. W. (1980). The ancestry of corn. Sci. A m . 242(1), 112-119. Beadle, G. W., and Galinat, W. C. (1991). The origin of maize. In “Environment, Origins and Population,” Vol. 3. Sm’thsonian Inst. Publ., Washington, D. C. (in preparation). Benz, B. F., and Iltis, H. H. (1990). Studies in archaeological maize. I. The “wild” maize from San Marcos Cave re-examined. Am. Anat. 55, 500-511. Bonnett, 0. T. (1953). Developmental morphology of the vegetative and floral shoots of maize. Ill., Agric. Exp. Stn., Bull. 568. Burnham, C. R. (1961). Linkage relations of teosinte branched. Maize Genet. Coop. Newsl. 35, 87. Collins, G. N . , and Kempton, J. H. (1920). Teosinte-maize hybrid. J . Agric. Res. 19, 1-37. Coyne, D. P. (1970). Effect of 2-chloroethylphosphoric acid on sex expression and yield in Butternut squash and its usefulness in producing hybrid squash. HortScience 5,227-228. Darwin, C. (1859). “The Origin of Species.” Mentor Book. (Reprinted, 1958, The New American Library, New York). Doebley, J. (1983). The maize and teosinte male inflorescence: A numerical taxonomic study. Ann. Mo. Bot. Gard. 70, 32-70. Doebley, J. (1990). Molecular evidence and the evolution of maize. Econ. Bot. 44, Suppl. 3, 6-27. Doebley, J . , Stec, A. , Wendcl, J., and Edwards, M. (1990). Genetic and morphological analysis of a maize-teosinte F2 population: Implications for the origin of maize. Proc. Natl. Acad. Sci. U.S.A. 87, 9888-9892. Durand, R., and Durand, B. (1984). Sexual differentiation in higher plants. Physiol. Plant. 60, 267-274. Galinat, W. C. (1963). Form and function of plant structures in the American Maydeae and their significance for breeding. Econ. Bot. 17, 51-59. Galinat, W. C. (1966). The evolution of glumeless sweet corn. Econ. Bor. 20, 441-445. Galinat, W. C. (1969). The evolution under domestication of the maize ear: String cob maize. Mass.,Agric. Exp. Stn., Bull. 577, 1-19. Galinat, W. C. (1971). The origin of maize. Annu. Rev. Genet. 5 , 447-478. Galinat, W. C. (1973). Intergenomic mapping of maize, teosinte and Tripsacum. Evofution (Lawrence, Kans.) 27, 644-655. Galinat, W. C. (1975a). The evolutionary emergence of maize. Bull. Torrey Bot. Club 102, 3 13-324. Galinat, W. C. (1975b). Abscission layer development in the rachis of Zea: Its nature, inheritance and linkage. Maize Genet. Coop. Newsl. 49, 100-102. Galinat, W. C. (1977). The origin of corn. Agronomy 18, 1-47. Galinat, W. C. (1978). The inheritance of some traits essential to maize and teosinte. In “Maize Breeding and Genetics” (D. B. Walden, ed.), pp. 93-111.

230

WALTON C . GALINAT

Galinat, W. C. (1979). A miniature fruitcase type of teosinte as the wild ancestor of the first maize. Maize Genet. Coop. Newsl. 53, 99-100. Galinat, W. C. (1983). The origin of maize as shown by key morphological traits of its ancestor, teosinte. Maydica 28, 121-138. Galinat, W. C. (1985a). The domestication and diffusion of maize. In “Prehistoric Food Production in North America” (R. I. Ford, ed.), Anthropol. Pap. No. 75, pp. 245-278. University of Michigan, Ann Arbor. Galinat, W. C. (1985b). The missing links between teosinte and maize: A review. Maydica 30, 137-160. Galinat, W. C. (1985~).Teosinte, the ancestor of maize: Perspectives for its use in maize breeding for the tropics. In “Breeding Strategies for Maize Production Improvement in the Tropics” (A. Brandolini and F. Salamini, eds.), Monogr. 100, pp. 1-11. FAO, UN, and 1st Agron. per L‘Oltremar, Firenze, Italy. Galinat, W. C. (1988a). The origin of corn. Agronomy 18, 1-31. Galinat, W. C. (1988b). The teosinte progenitors of corn as tools for its improvement. Proc. 43rd Annu Corn Sorghum Res. Conf. pp. 180-193. Galinat, W. C. (1991a). Plant habit as evidence of a biphyletic domestication of teosinte in the origins of maize. In “Proceedings Symposium of the Corn and Culture in the Prehistoric New World.” University of Minnesota, Minneapolis (in press). Galinat, W. C. (1991b). Interspace (is)and string cob (Sg 1 Sg 2) as stabilizing factors for the expression of key trait genes (tr, p d ) . Maize Genetics Coop Newsf. 65, 116-117. Goodman, M. M. (1988). The history and evolution of maize. CRC Crit. Rev. Plant Sci. 7(3), 197-220. Goodman, M. M., and Brown, W. L. (1988). Races of corn. Agronomy 18, 33-79. Goodman, M. M., and Stuber, C. W. (1983). Races of maize. VI. Isozyme variation among races of maize in Bolivia. Maydica 28, 169-187. Harlan, J. R., and deWet, J. M. J. (1972). Origin of maize: The tripartite hypothesis. Euphytica 21,271-279. Iltis, H. H. (1972). The taxonomy of Zea mays L. (Gramineae). Phytologia 23, 248-249. Iltis, H. H. (1983). From teosinte to maize: The catastrophic sexual transmutation. Science 22, 886-894. Iltis, H. H., and Doebley, J. F. (1980). Taxonomy of Zea (Gramineae). 11. Subspecific categories in the Zea mays complex and a generic synopsis. A m . J . Bot. 67, 994-1004. Jones, D. F. (1917). Dominance of linked factors as a means of accounting for heterosis. Genetics 2, 466-479. Jones, D. F. (1918). The effects of inbreeding and crossbreeding upon development. B U N . Conn., Agric. Exp. Sm., New Haven 207. Kato Y,T. A. (1976). Cytological studies of maize and teosinte in relation to their origin and evolution. Mass.,Agric. Exp. Stn., Bull. 635. Kato Y,T. A. (1984). Chromosome morphology and the origin of maize and its races. Evol. Biol. 17, 219-253. Kermicle, J. O., and Allen, J. 0. (1990). Cross-incompatibilitybetween maize and teosinte. Maydica 35, 399-408. Landi, P., and Frascaroli, E. (1986). Inheritance of protandry in maize (Zea mays L . ) and its relationship to yield. Genet. Agrar. 40, 57-64. Langham, D. G. (1940). The inheritance of intergeneric differences in Zea-Euchlaena hybrids. Genetics 25, 88-108. Mangelsdorf, P. C. (1947). The origin and evolution of maize. Adv. Genet. 1, 161-207. Mangelsdorf, P. C. (1948). The role of pod corn in the origin and evolution of maize. Ann. Mo. Bot. Gard. 35, 377-406. Mangelsdorf, P. C. (1974). “Corn: Its Origin, Evolution and Improvement.” Harvard Univ. Press (Belknap), Cambridge, Massachusetts. Mangelsdorf, P. C. (1986). The origin of corn. Sci. A m . (Aug.), pp. 72-78.

EVOLUTION OF CORN

23 1

Mangelsdorf, P. C., and Edwardson, J. R. (1953).Tests for weak alleles at the Tu-tu locus and effects of alleles at the Tu-ru locus. Maize Genet. Coop Newsl. 27, 24-27. Mangelsdorf, P. C., and Reeves, R. G. (1939).The origin of Indian corn and its relatives. Tex., Agric. Exp. Stn. [Bull.]574, 1-315. McClintock, B. (1959). Chromosome constitutions of Mexican and Guatemalan races of maize. Annu. Rep. Dep. Gen. Carnegie Inst. Washington 59, 461-472. Phinney, B. 0. (1956). Growth response of single-gene dwarf mutants in maize to gibberellic acid. Proc. Nad. Acad. Sci. U.S.A. 42, 185-189. Randolph, L. F. (1959). The origin of maize. Indian J. Genet. Plant Breed. 19, 1-12. Randolph, L. F. (1972). Retention of Euchlaena as a genus separate from Zea. Maize Genet. Coop. Newsl. 46, 8. Randolph, L. F. (1976). Contributions of wild relatives of maize to the evolutionary history of domesticated maize: A synthesis of divergent hypotheses. I. Econ. Bor. 30, 321-345. Rogers, J. S . (1950).The inheritance of inflorescence characters in maize-teosinte hybrids. Genetics 35, 541-558. Schoen, D. J., and Dubuc, M. (1990). The evolution of inflorescence size and number: A gamete-packaging strategy in plants. A m . Nut. 135, 841-857. Sehgal, S . M. (1963). Effects of teosinte and “Tripsacum” introgression in maize. Ph.D. Dissertation, Bussey Institute, Harvard University, Cambridge, Massachusetts. Shaver, D. L. (1967). Perennial maize. J. Hered. 58, 270-273. Shifriss, 0. (1985).Origin of gynoecism in squash. HortScience 20, 889-891. Vavilov, N. I. (1931). Mexico and Central America as the principal centre of origin of cultivated plants of the New World. Bull. Appl. Bot., Genet. Plant Breed. 26, 135-199. Wallace, H. A,, and Brown, W. L. (1956). “Corn and Its Early Fathers.” Michigan State Univ. Press, East Lansing. Weatherwax, P. (1935). The phylogeny of Zea mays. A m . Midl. Nut. 16, 1-71. Weatherwax, P. (1954). “Indian Corn in Old America.” Macmillan, New York. Wellhausen, E. J., Roberts, L. M., Hernandex-Xolocotzi, E., and Mangelsdorf, P. C. (1952). “Races of Maize in Mexico.” Bussey Institute, Harvard University, Cambridge, Massachusetts. Wilkes, H. G. (1967).“Teosinte: The Closest Relative of Maize.” Bussey Institute, Harvard University, Cambridge, Massachusetts. Wilkes, H. G. (1986).Maize: Domestication, racial evolution and spread. In Plant Domestication and Early Agriculture. XI. World Archaeological Congress. Allen Unwin, London.

This Page Intentionally Left Blank

USEOF SURFACE COMPLEXATION MODELS IN SOIL CHEMICAL S y s m s Sabine Goldberg USDA-ARS, U.S. Salinity Laboratory, Riverside, California 92S01

I. Introduction

II. Description of Models A. Common Characteristicsof Surface Complexation Models B. Constant CapacitanceModel C. Triple-Layer Model D. Stem Variable Surface Charge-Variable Surface Potential Model E. Generalized Two-Layer Model F. One-pKModel 111. Application of Models to Protonation-Dissociation Reactions on Oxides, Clay Minerals, and Soils A. Constant Capacitance Model B. Triple-Layer Model C. Stern VSC-VSP Model D. Generalized Two-Layer Model E. One-pKModel N . Application of Models to Metal Ion Adsorption Reactions on Oxides, Clay Minerals, and Soils A. Constant Capacitance Model B. Triple-Layer Model C. Stern VSC-VSP Model D. Generalized Two-Layer Model E. One-pKModel V. Application of Models to Inorganic Anion Adsorption Reactions on Oxides, Clay Minerals, and Soils A. Constant Capacitance Model B. Triple-Layer Model C. Stern VSC-VSP Model D. Generalized Two-Layer Model E. One-pKModel VI. Application of Models to Organic Ligand Adsorption Reactions on Oxides k Constant Capacitance Model

233 Advances in A p m y , Volumr 47

234

SABINE GOLDBERG B. Triple-Layer Model C. Stern VSC-VSP Model VII. Applicationof Models to Competitive Adsorption Reactions on Oxides A. Metal-Metal Competition B. Anion-Anion Competition C. Metal-Ligand Interactions VIII. Incorporation of Surface ComplexationModels into Computer Codes A. Incorporation into Chemical Speciation Models B. Incorporation into Transport Models IX. Summary References

I. INTRODUCTION A model is a simplified representation of reality considering only the characteristics of the system important to the problem at hand. An empirical model is a description of data without theoretical basis. A chemical model provides a description of a chemical system consistent with its chemical properties and should be simultaneously as simple and as chemically correct as possible. The ideal model is effective, comprehensive, realistic, and predictive (Barrow and Bowden, 1987). An effective model closely describes observations, a comprehensive model applies to a wide range of conditions without modification, a realistic model conforms to accepted theories of behavior, and a predictive model can be applied to different conditions. Unlike empirical models, surface complexation models are chemical models that strive to satisfy the above characteristics and to give a general molecular description of adsorption phenomena using an equilibrium approach. The purpose of molecular theory is to derive thermodynamic properties such as activity coefficients and equilibrium constants from the principles of statistical mechanics (Sposito, 1981). The surface complexation models are designed to calculate values for the thermodynamic properties mathematically and constitute a family of models having similar characteristics. This model family includes the constant capacitance model (Stumm et al., 1980), the triple-layer model (Davis et al., 1978), the Stern variable surface charge-variable surface potential model (Bowden et al., 1980), the generalized two-layer model (Dzombak and Morel, 1990), and the one-pK model (van Riemsdijk et al., 1986). The major advancement of the surface complexation models is that they consider surface charge. Surface charge results from protonation and dissociation reactions as well as from surface complexation reactions of reactive surface hydroxyl

SURFACE COMPLEXATION MODELS

235

groups at mineral surfaces. The sign and magnitude of the mineral surface charge are dependent on the pH and the ionic strength of the electrolyte solution. The purpose of this article is to comprehensively review five common surface complexation models of the mineral-solution interface and their use in describing soil chemical systems. Common model characteristics and adjustable parameters will be discussed. For each model, surface species, chemical reactions, equilibrium constant expressions, and surface activity coefficients will be defined. Applications of the model to ion adsorption on soil minerals and soils will be presented. Incorporation of surface complexation models into computer codes will be discussed.

II. DESCRIPTION OF MODELS A. COMMON CHARACTERISTICS OF SURFACE COMPLEXATION MODELS 1. Balance of Surface Charge

The balance of surface charge on an oxide mineral in aqueous solution is (Sposito, 1984a)

where oH is the net proton charge, defined by aH = F ( r H - r O H ) , where r is a surface excess concentration, q, is the inner-sphere complex charge resulting from the formation of inner-sphere complexes between adsorbing ions (other than H+ and OH-) and surface functional groups, aosis the outer-sphere complex charge resulting from the formation of outer-sphere complexes between adsorbing ions and surface functional groups or ions in inner-sphere complexes, a d is the dissociated charge, equal to minus the surface charge neutralized by electrolyte ions in solution that have not formed adsorbed complexes with surface functional groups. All surface complexation models are based on a balance of surface charge expression. The surface functional group is defined as SOH, where S represents a metal ion of the oxide mineral surface bound to reactive hydroxyl group. The functional group can also be an aluminol or silanol group at the edge of a clay mineral particle. The balance of surface charge expression may be simplified depending upon the interfacial structure and the assumptions of the particular complexation model. The specific expressions will be provided in the detailed discussion of each model.

236

SABINE GOLDBERG

2. Electrostatic Potential Terms

All surface complexation models contain at least one coulombic correction factor to account for the effect of surface charge on surface complexation. These coulombic correction factors take the form of electrostatic potential terms, e-Fqi'RT where Tiis the surface potential (V) in the ith surface plane, F is the Faraday constant (C molil), R is the molar gas constant (J mol-' K-'), and T is the absolute temperature (K) in the conditional equilibrium constant expressions. Surface complexation models of the oxide-solution interface can be considered as special cases of the van der Waals model in statistical mechanics (Sposito, 1983). In this model, charged surface complex species create a long-ranged mean electric force field from screened coulomb forces by mutual interaction. Shortranged interactions are neglected. The mean field effect is responsible for the presence of electrostatic potential terms in the conditional equilibrium constant expressions. According to the van der Waals model, the activity coefficient differs from a value of one because the total potential energy changes when neutral surface hydroxyl groups are replaced by charged surface complexes (Sposito, 1983). Diffuse double-layer theory need not be invoked to lend chemical significance to the exponential terms (Sposito, 1983). Unfortunately, diffuse double-layer effects have been invoked almost universally in model applications to explain the presence of the exponential terms. The electrostatic potential terms should simply be considered as solid-phase activity coefficients correcting for the charges on the surface complexes. 3. Adjustable Parameters

The surface complexation models explicitly define equilibrium constant expressions for surface complexes. They contain mass balance equations for each type of surface site and charge balance equations for each surface plane of adsorption. Thus, all models contain the following adjustable parameters: Ki , the equilibrium constants; Ci, the capacitance density for the ith surface plane; and [SOHIT, the total number of reactive surface hydroxyl groups. Values of the equilibrium constants are almost always obtained with the help of a computer program. Details for each model will be discussed below. Values for some of the capacitance density parameters can be obtained experimentally (Sposito, 1984a). However, the capacitance density values have almost universally been taken as adjustable. Details for experimental determinations of capacitance densities will be provided in the model descriptions.

SURFACE COMPLEXATION MODELS

237

The total number of reactive surface hydroxyl groups, [SOHIT, is an important parameter in surface complexation models. The value of the surface site density has been determined either experimentally by tritium exchange (Davis and Leckie, 1978, 1980), potentiometric titration (Balistrieri and Murray, 1981; Hohl and Stumm, 1976; Kummert and Stumm, 1980), fluoride adsorption (Sigg, 1979), maximum adsorption (Goldberg and Sposito, 1984a; Goldberg, 1985, 1986a), calculated from crystal dimensions, or optimized to fit experimental adsorption data (Hayes et al., 1988). For goethite these determinations range from 4 sites/nm2 for potentiometric titration to 6-7 sites/nm2 for fluoride adsorption to 17 sites/nm2 for tritium exchange (Sigg, 1979). Crystallographic calculations of reactive surface hydroxyl groups on goethite provide a surface site density of 3 sites/nm2 (Sposito, 1984a). Various measurements of surface site density are described in detail by James and Parks (1982). These authors also provide values of this parameter obtained with diverse methods for many oxide minerals. Initial sensitivity analyses showed surface complexation models to be relatively insensitive to surface site density values for selenium adsorption on amorphous hydrous ferric oxide in the range of 3-12 sites/nm2 (Hayes et al., 1988) and for phosphate adsorption on a soil in the range of 1.252.5 sites/nm2 (Goldberg and Sposito, 1984b). Recently, sensitivity analyses for acid-base titration data on goethite, aluminum oxide, and titanium oxide showed the models to be relatively insensitive to surface site density in the range of 2-20 sites/nm2 (Hayes et al., 1991). However, the actual values of the equilibrium constants decreased with increasing surface site density. More detailed investigations of both the constant capacitance model and the triple-layer model have indicated that for adsorption of phosphate, arsenate, selenite, selenate, silicate, molybdate, and borate the ability of the models to describe adsorption using both inner-sphere and outer-sphere surface complexes was sensitively dependent on the value of the surface site density (Goldberg, 1991). Further research is needed to determine the most appropriate experimental measurement of surface site density for surface complexation modeling. The surface complexation models contain the assumption that ion adsorption occurs at only one or two types of surface sites. Clearly, soils are complex, multisite mixtures. However, experimental evidence suggests that even oxide mineral surfaces contain several sets of adsorption sites (Rochester and Topham, 1979a,b; Benjamin and Leckie, 1980, 1981). Thus, equilibrium constants determined for soils and even for pure mineral systems likely represent average composite values for all sets of reacting sites.

SABINE GOLDBERG

238

B. CONSTANT CAPAC~ANCE MODEL The constant capacitance model of the oxide mineral-aqueous solution interface was developed by the research groups of Schindler and Stumm (Schindler and Gamsjager, 1972; Hohl and Stumm, 1976; Schindler et al., 1976; Stumm ef al., 1976, 1980). The model is based on the following assumptions: (1) all surface complexes are inner-sphere complexes, with anion adsorption occurring via a ligand exchange mechanism, (2) the constant ionic medium reference state determines the activity coefficients of the aqueous species in the conditional equilibrium constants and therefore no complexes are formed with ions in the background electrolyte, and (3) a linear relationship exists between surface charge and surface potential: (T

= (CSU/F)*

where C is the capacitance density (F m-*), S is the specific surface area (m2g-'), a is the suspension density (g liter-'), and (T has units of mol, liter-'. These assumptions greatly simplify the balance of surface charge expression: (T=(TH+f7iis

(3)

The surface charge on a particle, (T,equals the net proton charge plus the charge resulting from formation of inner-sphere surface complexes. There is no strict balance of surface charge in the constant capacitance model. Equation (3) is a model expression for the total particle charge. Figure 1 provides a diagram of the surface-solution interface in the constant capacitance model. The following equations are general surface complexation reactions (Hohl et ul., 1980): SOH + H+

* SOH;

SOH a SO-

(4)

+ H+

SOH + M m + s SOM("-')

2SOH + M"' SOH + L'-

(5)

+ H+

(SO),M@"'*) + 2H'

SL('-')-

2SOH + L'- S S2L('-')-

+ OH+ 20H-

(6)

(7) (8)

(9) where SOH represents the surface functional group, M is a metal ion, r n f is the charge on the metal ion, L is a ligand, and 1- is the charge on the ligand. The intrinsic conditional equilibrium constants describing these

SURFACE COMPLEXATION MODELS

Cnarge

u

Addur&d

H'

Species

239

X

wM" L+

Figure 1. Placement of ions, potential, charge, and capacitance for the constant capacitance model. After Westall (1986), reproduced with permission from the American Chemical Society.

reactions are (Hohl et al., 1980)

KL(int) =

[SOM(m-l)][H'] exp[(m - l)FT/RT] [SOH][Mm+]

KL(int) =

[(SO)2M(m-2)][H']2 [SOH]2[Mmt] exp[(m - 2)FV/RT]

(13)

where square brackets represent concentrations (mol liter-'). The intrinsic conditional equilibrium constants are obtained by extrapolating the conditional equilibrium constants to zero net surface charge (Stumm et al., 1980). A detailed explanation of the procedure is provided in Section II1,A. Surface complexes exist in a chargeless environment in the standard state. To solve the equilibrium problem, two more equations are needed. The mass balance for the surface functional group is (Sigg and Stumm, 1981)

[SOH]== [SOH] + [SOH:]

+ [SO-] + [SOM("-')]

240

SABINE GOLDBERG

and the charge balance is (Sigg and Stumm, 1981)

- [SO-] + (rn - l ) [ S O l ~ l ( ~ -+~(m ) ] - 2)[(S0)2M(m-2)1 - (I - l)[SL''-l)-] - (I - 2)[S*L"-2'- 1 (17)

u = [SOH;]

This set of equations can be solved with a computer program using the mathematical approach outlined by Westall (1980).

C. TRIPLE-LAYER MODEL The original triple-layer model was developed as an extension of the site-binding model (Yates et al., 1974) by Davis and co-workers (1978; Davis and Leckie, 1978, 1980). In contrast to the constant capacitance model, the original triple-layer model assumes that all metals and ligands form outer-sphere surface complexes. Only the proton and the hydroxyl ion form inner-sphere surface complexes; all other surface complexes contain at least one water molecule between the ion and the surface functional group (Davis et al., 1978). The infinite dilution reference state determines the activity coefficients of the aqueous species in the conditional equilibrium constants. The surface charge is thus composed of the net proton charge plus the charge resulting from the formation of outersphere complexes: u = UH + u,,

(18) The model derives its name from the three planes of charge used to represent the oxide surface. The potential-determining ions at the surface, protons and hydroxyl ions, are placed in the surface o-plane. The outersphere complexes are situated a little further from the surface in the P-plane. The diffuse layer starts at the d-plane and extends into the aqueous solution phase. Figure 2 provides a diagram of the surfacesolution interface in the triple-layer model. Because the triple-layer model contains three surface planes, three charge-potential relations are needed (Davis et al., 1978):

%fo - 9 p = uJC1 "Yp - v d = - ud/c;? = -(8RTc eOD)lI2sinh(F%fd/2RT)

(19) (20)

(21) where u has units of C m-2, E~ is the permittivity of vacuum, D is the dielectric constant of water, and c is the concentration of a 1:1background electrolyte. In the case of 1:2 and 2: 1 background electrolytes, Eq. (21) gd

241

SURFACE COMPLEXATION MODELS

Charge Adsorbed Species

Cr,

%

5

X

H+

C+

cwnterions

OH(M"")

AM'"* Lg-

t-------*

(L*- 1

Figure 2. Placement of ions, potentials, charges, and capacitances for the triple-layer model. Parentheses represent ion placement allowed only in the modified triple-layer model. After Westall (1980),reproduced with permission from the American Chemical Society.

becomes much more complex (see, e.g., Sposito, 1984a). The reactions considered in the original triple-layer model include Eqs. (4), (9,and (6), although Eq. (6) is now written as SOH + M"+

SO- - M"+

+ H+

(22)

representing the formation of an outer-sphere surface complex (Davis et af.,1978). In the original triple-layer model the formation of the bidentate metal complex, Eq. (7), is replaced by the formation of a hydroxy-metal surface species: SOH

+ M"' + H20

SO- - MOH("-"

+ 2H+

(23) Davis and Leckie (1978) found that this surface reaction was more consistent with their experimental data. In the original triple-layer model the ligand exchange reactions, Eqs. (8) and (9), are now written as outersphere surface complex formations (Davis and Leckie, 1980): SOH + H+ + L' e SOH; - L'SOH + 2HC + L'-

(24)

(25) For anion adsorption reactions, Davis and Leckie (1980) found that they could obtain results more consistent with their experimental data by replacing the bidentate complex with a protonated surface species. Two SOH; - LH('-')-

242

SABINE GOLDBERG

additional reactions are needed to describe the outer-sphere surface complexes formed by ions of the background electrolyte (Davis et al., 1978): SOH + C + s S O - - C+ + H+

SOH + H+ + A- # SOH;

(26)

- A-

(27)

where C+ is the cation and A- is the anion. The intrinsic conditional equilibrium constants for the triple-layer model have a form similar to those for the constant capacitance model. The protonation and dissociation constants remain as written in Eqs. (10) and (11). The other intrinsic conditional equilibrium constants are (Davis et al., 1978; Davis and Leckie, 1978, 1980) KL(int) =

[SO- - Mm+][H+] [SOH][hirn+] exp[F(m*p - *O)/RTl

As in the constant capacitance model, the intrinsic conditional equilibrium constants are obtained by extrapolation (Davis et al., 1978). Surface complexes exist in a chargeless environment in the standard state. A detailed explanation is provided in Section II1,B.The mass balance equation for the surface functional group is [SOHIT = [SOH] + [SOH,']

+ [SO-] + [SO- - Mm+]

+ [SO- - MOHcrn-')]+ [SOH: - L'-] + [SOH: - LH('-')-] + [SO- - C'] + [SOH,'

- A-]

(34)

Last, three charge balance equations are needed: ao+Up+cTd=O

(35)

SURFACE COMPLEXATION MODELS

F u,,= -{[SOH:] Sa

+ [SOH,'

+ [SOH: - A-]

- L'-]

- [SO-]

+ [SOH:

- LH"-"-]

- [SO- - Mm+]

- [SO- - MOH('"-')] - [SO- - C']} F

+ (m- l)[SO-

UB = -{m[SO- - Mm+]

Sa

- f[SOHz

- L'-] - (I

243

- MOH("-"]

(36)

+ [SO- - C']

- 1)[SOHz - LH('-l)-] - [SOH: - A-1) (37)

This set of equations can be solved with a computer program using a mathematical approach similar to that for the constant capacitance model (Westall, 1980). Extensions of the original triple-layer model to describe ion adsorption as inner-sphere surface complexes in the o-plane have been carried out (Blesa et a f . , 1984a; Hayes and Leckie, 1986, 1987; Hayes et al., 1988). Blesa et al. (1984a) were the first to extend the triple-layer model to describe inner-sphere surface complexation via ligand exchange in modeling boron adsorption on the iron oxide, magnetite, and on zirconium dioxide. Later work by Hayes and co-workers described lead (Hayes and Leckie, 1986, 1987), cadmium (Hayes and Leckie, 1987), and selenite adsorption (Hayes et al., 1988) on iron oxides using inner-sphere surface complexes. In the modification of the triple-layer model developed by Hayes and Leckie (1986, 1987), the standard state for both solution and surface species is chosen as zero surface charge and without ionic interactions. The reference state for all species is infinite dilution and zero surface charge. The result of these changes is that the exponential terms become ratios of solution and surface activity coefficients and that the intrinsic conditional equilibrium constants are equivalent to the thermodynamic equilibrium constants (Hayes and Leckie, 1987). In the modified triple-layer model the surface charge is composed of the net proton charge plus the charge resulting from the formation of innersphere and outer-sphere surface complexes: u = U H + uiis+ uoos

(38)

Figure 2 indicates the placement of metal and ligand ions into the o-plane in the modified triple-layer model. In the modified triple-layer model, inner-sphere surface complexation by metal ions is described by Eq. (6), as in the constant capacitance model. Only this metal surface complexation reaction is considered. Inner-sphere surface complexation by anions via ligand exchange is described by

244

SABINE GOLDBERG

Eq. (8), as in the constant capacitance model. The intrinsic conditional equilibrium constants for these reactions are Eqs. (12) and (14), where q is replaced by qo.In addition to these reactions an additional surface complexation reaction between the cation of the background electrolyte and the adsorbed anion is considered: SOH + L'-

+ H+ + Cf

SL('-')- - C+ + HZO

(39) The intrinsic conditional equilibrium constant expression for this reaction is S

In the modified triple-layer model the mass balance is [SOHIT = [SOH] + [SOH:] + [SO-] + [SOM("-')] + [SO- - Mm+] + [SO- - MOH(m-1'1 + [sL('-l)-] + [SL([-I)- - C+]

+ [SOH: - L'-] + [SOH:

- LH('-')-]

+ [SO- - C']

+ [SOH: - A-]

(41) In the modified triple-layer model the charge balance in the o-plane is

F - -{[SOH:] -Sa

+ [SOH:

- L'-]

+ [SOH:

- LH('-''-]

+ (m- l)[SOM("-')] + [SOH:

- A-] - [SO-] - [SO- - M*+] - [SO- - MOH("-l)] - (I - l)[SL('-l)-] - (1 - l)[SL('+

- C']

-

[so-- C']}

(42)

and the charge balance in the /3-plane is

F

+ (m- 1)[SO- - MOH("-l)]

up= -{rn[SO- - Mm+]

Sa

+ [SL('-l)- - C'] + [SO- - C']

- I[SOH:

- L'-]

- (I - 1)[SOHz - LH"-l)-] - [SOH: - A-I}

(43)

D. STERNVARIABLE SURFACE CHARGE-VARIABLE SURFACE POTENTIAL MODEL The Stern variable surface charge-variable surface potential (VSCVSP) model was developed by Bowden, Barrow, and their co-workers (Bowden et al., 1977, 1980; Barrow et a / ., 1980a, 1981;Barrow, 1987). This

SURFACE COMPLEXATION MODELS

245

I

Adsorbed Species

H’

oxyontons

Mm’

OH-

C+ A-

counterions

Figure 3. Placement of ions, potentials, charges, and capacitances for the Stern VSCVSP model. After Bowden et al. (1980).

model is actually a mixture of the constant capacitance model and the original triple-layer model and might also be called the four-layer model. The Stern VSC-VSP model assumes that protons, hydroxyl ions, and “strongly adsorbed” oxyanions and metals form inner-sphere surface complexes. The balance of surface charge expression in this model is the same as in the modified triple-layer model, Eq. (38). The protons and hydroxyl ions reside in the o-plane close to the surface. The “strongly adsorbed” ions are placed in an a-plane a short distance away from the surface o-plane. The major cations and anions are assumed to form outer-sphere surface complexes and are placed in the p-plane a short distance away from the a-plane (Barrow et al., 1980a). The fourth plane, the d-plane, indicates the start of the diffuse double layer. Figure 3 provides a diagram of the surface-solution interface in the Stern VSC-VSP model. The surface functional group is defined as OH-S-OH2. This convention allows only one protonation or dissociation to occur for every two surface hydroxyl groups (Barrow er al., 1980a). The Stern VSC-VSP model emphasizes parameter optimization and defines no surface reactions, provides no equilibrium constant expressions for these reactions, and defines no specific surface species. The model equations are given in Barrow er al. (1980a) and Bowden et al. (1980). Because the Stern VSC-VSP model contains four surface planes, four charge-potential equations and four charge balance equations are needed: qo

- q a = Uolcoa

(44)

“a

- q p = (go + uJ/Cap

(45)

q p

-q

(46)

d =

- md/ c p d

246

SABINE GOLDBERG

where u has units of mol, m-2 and C has units of mol, V-’ m-2. The diffuse layer charge, u d , is calculated as in the triple-layer model, Eq. (21). However, the result from Eq. (21) is in C m-2 and must be divided by the Faraday constant to obtain mol, m-2. The remaining charge balance equations are

a, =

x

i

1+

i

Kiai exp(-ZjFYa/RT)

where Ns is the maximum surface charge density (mol, m-2), NT is the maximum adsorption of specifically adsorbed ions (mol, m-2), Ki is the binding constant, ai is the activity, and Ziis the charge of the ith specifically adsorbed ion. In the Stern VSC-VSP approach, no mass balance is carried out. Despite the fact that their values are available experimentally, the total number of surface sites and the maximum adsorption are adjustable parameters. In solving the set of above equations, values of N s , NT, the binding constants K j , and the capacitances Cj, are chosen to optimize model fit to the data. The charge densities uiand the electrostatic potentials Tisubsequently are calculated via a computer program (Barrow, 1979). Sposito (1984b) does not consider the Stern VSC-VSP model to be a chemical model because it lacks chemical reactions and equilibrium constant expressions. The model is not self-consistent chemically because it does not reduce to the triple layer model when “strongly adsorbed” ions are absent from the a-plane (Sposito, 1984b).

E. GENERALIZED TWO-LAYERMODEL The generalized two-layer model was developed by Dzombak and Morel (1990) as an expansion of the diffuse layer model proposed by Stumm and co-workers (1970; Huang and Stumm, 1973). Like the constant capacitance model, the generalized two-layer model assumes that all surface complexes are inner-sphere complexes. The surface charge expression is

247

SURFACE COMPLEXATION MODELS

the same as in the constant capacitance model, Eq. (2). The generalized two-layer model is like the triple layer model and the Stern VSC-VSP model in that it considers a diffuse layer. The model is named after the two planes of charge used to represent the surface. All surface complexes are placed into the surface plane. The diffuse layer commences at the dplane and extends into the solution phase. Figure 4 provides a diagram of the surface-solution interface in the generalized two-layer model. As in the triple-layer model, the surface charge, u,has units of C m-2. The relationship between the surface charge and the surface potential is defined by electric double-layer theory, Eq. (21). No capacitance parameters are required in the generalized two-layer model. The model employs the infinite dilution reference state for the solution and a reference state of zero charge and potential for the surface (Dzombak and Morel, 1990). The surface reactions considered in the generalized two-layer model include Eqs. (4) and (5). Metal ion adsorption is considered to occur on two types of sites: a small set of high-affinity "strong" sites and a large set of low-affinity "weak" sites. Adsorption on both sets of sites takes the form of Eq. (6):

+ M"+ SWOH+ M"+ S'OH

F2

SSOM("-') + H+

S"OM("-"

(51)

+ H+

(52) where s represents the high-affinity and w represents the low-affinity sites. In the generalized two-layer model no bidentate surface species [Eq. (7) F2

L1I

I I

I

I I 1

CT Charge

Hi

ad

X

b

counter ions

Adsorbed OHSpecies

Mm + L'-

Figure 4. Placement of ions,potential, and charges for the generalized two-layer model. After Dzombak and Morel (1990), reproduced with permission from John Wiley and Sons.

248

SABINE GOLDBERG

for metals or Eq. (9) for ligands] are defined. The ligand exchange reactions are

For ligand exchange, it is not necessary to specify two sets of binding sites. The intrinsic conditional equilibrium constants for the generalized twolayer model are similar to those for the constant capacitance model. Equations (10) and (11) describe protonation and dissociation. Metal surface complexation is described by two constants like Eq. (12), for the reaction of both strong sites, S'OH, and weak sites, S"OH. The intrinsic conditional equilibrium constants for ligand exchange are

Application of the generalized two-layer model to systems containing both metal and ligand species is difficult because the pool of binding sites for protonation-dissociation and ligand adsorption is split into two sets of binding sites of different affinity for metal adsorption. For this reason, mass balance and charge balance equations will be provided separately for metal and ligand adsorption. For metal surface complexation these equations are

+ [S'OH:] + [SO-] + [S'OM("-')] [S"OHIT = [S"OH] + [S"OH;] + [SwO-] + [S"OM("-')] = [S'OH]

F -{[S'OH:] Sa

(T=

(57) (58)

+ [S"OH$] + ( m - l)[SsOM(m-l)]

+ (m - l)[S"OM("-')]

- [SO-]

- [SwO-]}

(59)

For ligand surface complexation the equations are [SOHIT= [SOH] + [SOH;]

F

u=-

Sa

{[SOH:]

+ [SO-] + [SL('-')-] + [SHL('-2)-]

- [SO-] - (1 - l)[SL('-')-]

- ( I - 2)[SHL('-')-])

(60) (61)

These equations can be solved with a computer program or approximated using hand calculations as described by Dzombak and Morel (1990).

249

SURFACE COMPLEXATION MODELS

F. ONE-&MODEL The one-pK model was first proposed by Bolt and van Riemsdijk (1982) and was developed by van Riemsdijk and co-workers (1986, 1987; Hiemstra et al., 1987; van Riemsdijk and van der Zee, 1991). In the one-pK model the surface functional group is defined as a singly coordinated oxygen atom that carries either one or two protons, leading to surface sites SOH and SOH2, respectively. The constant capacitance model, the triple-layer model, and the generalized two-layer model could all be written based on the one-pK concept for singly coordinated oxygen atoms. The Stern model of the surface-solution interface has been chosen for the one-pK model (van Riemsdijk et al., 1987). This model assumes that protons and hydroxyl ions form inner-sphere surface complexes and reside in the o-plane close to the surface. Cations and anions form outer-sphere surface complexes at the Stern plane, or d-plane. The Stern plane indicates the start of the diffuse double layer. The surface charge on a particle is equal to the net proton charge: u = an

(62) Figure 5 provides a diagram of the surface-solution interface in the one-pK Stern model. Because the one-pK Stern model contains two planes, two

Charge

r&

Adsorbed Species

H' OH-

=d

X

counter ions

I

C' A-

M m* L'-

Figure 5. Placement of ions, potentials, charges, and capacitance for the one-pK Stern model. After Westall (1986), reproduced with permission from the American Chemical Society.

250

SABINE GOLDBERG

charge-potential relations are used (Hiemstra et al., 1987):

w, - wd)

a, = C(

(63)

where the units of a are C m-2. The relation between charge and potential in the Stern or d-plane is provided by diffuse double-layer theory, Eq. (21). The surface complexation reactions defined in the one-pK model are (van Riemsdijk et al., 1987; Hiemstra et al., 1987; van Riemsdijk and van der Zee, 1991) SOH'/'-

+ H+ d SOH:/2'

(64)

SOHIr'Z- + Mm+ e SOH'/'- - M"+

SOH~/+ ~ -~

m

++~ ' e 0 SOH1l2-- MOH(m-*)+ H+

SOHl/2++ Li-

e SOH:/2+

(66)

- L'-

(67)

- C+

(68)

SOH:"+ - A-

(69)

SOH"'-

+ C+ e SOH"'-

SOH:/'+

+ A-

Z?

(65)

The intrinsic equilibrium constants for the one-pK model are as follows (van Riemsdijk et al., 1987; Hiemstra et al., 1987; van Riemsdijk and van der Zee, 1991):

The mass balance of singly coordinated oxygen atoms Ns is N s = [SOH1/'-]

+ [SOH:/2+]

(76)

SURFACE COMPLEXATION MODELS

251

The surface charge balance equation is r.ro= 1/2[S0Hl/2’

-SOH‘/’-

1

(77)

The set of equations can be solved using a computer algorithm based in part on the method of Westall (1980).

III. APPLICATION OF MODELS T O PROTONATION-DISSOCIATION REACTIONS ON OXIDES, CLAY MINERALS, AND SOILS

A. CONSTANT CAPACITANCE MODEL The constant capacitance model has been used to describe the amphoteric acid-base behavior of inorganic surface hydroxyl groups on oxide minerals. The adsorption of protons and hydroxyl ions has been investigated on the following minerals: amorphous silica (Schindler and Kamber, 1968; Osaki et al., 1990b), anatase (TiOz) (Schindler and Gamsjager, 1972), aluminum oxide (yA1203) (Hohl and Stumm, 1976; Kummert and Stumm, 1980), bayerite [yAl(OH),] (Pulfer et al., 1984), boehmite ( y A100H) (Bleam et al., 1991), goethite (a-FeOOH) (Sigg, 1979; Lovgren et al., 1990), magnetite (Fe304) (Regazzoni et al., 1983), and amorphous iron oxide (Farley et al., 1985). In general, values of the intrinsic protonation and dissociation constants described by Eqs. (10) and (11) can be obtained from alkimetric or acidimetric titration curves carried out in the absence of specifically adsorbing metal or ligand ions (Stumm et al., 1980). In this analysis, the assumption is made that u,the net surface charge, is equal to [SOH;] below the zero point of charge, ZPC, and is equal to -[SO-] above the ZPC. The ZPC is defined as the pH value at which surface charge is zero. In the constant capacitance model the ZPC is also defined as ZPC = 1/2[log K+(int) - log K-(int)]

(78)

A plot of the logarithm of the conditional equilibrium constant, ‘ K , , versus surface charge, cr, will yield the logarithm of the intrinsic equilibrium constant, K,(int), upon linear extrapolation to zero surface charge. The conditional equilibrium constants for protonation-dissociation are defined as

‘K,

= [SOH;]/([SOH][H+])

‘K- = [SO-][H+]/[SOH]

(79)

(80)

252

SABINE GOLDBERG

By combining Eq. (10) with Eq. (79), and Eq. (11) with Eq. (go), one can relate the intrinsic protonation and dissociation constants and the conditional protonation and dissociation constants: (81) K,(int) = 'K, exp(+F1Zr/RT) where the positive sign represents the protonation constant and the negative sign represents the dissociation constant. Upon substituting for surface potential, q ,from Eq. (2), taking the logarithms of both sides, and solving for log 'K, , the following equation is obtained:

log 'K, = log K,(int)

-+ (+F2/[CSaRT(ln 101

(82) By plotting the titration data as log 'K, versus u,an estimate of log K,(int) is obtained from the y intercept, where (T = 0. The capacitance parameter can be obtained from the slope of such a plot. Values of log K,(int) have been obtained in the above fashion for anatase (Schindler and Gamsjager, 1972), aluminum oxide (Hohl and Stumm, 1976; Kummert and Stumm, 1980), goethite (Sigg, 1979), magnetite (Regazzoni et al., 1983), and amorphous iron oxide (Farley et al., 1985). Figure 6 provides an example of the linear extrapolation technique for the titanium oxide, anatase (Schindler and Gamsjager, 1972). Table I presents values of logK,(int) obtained by various authors using the linear extrapolation technique. The assumptions of the linear extrapolation tech-

5

3

0.2

0.1 II

9 7 -0.1

-0.2

-0.3

-0.4

[SO-](mol kg-') Figure 6. The logarithms of the conditional protonation and dissociation constants as a function of surface charge for anatase, TiOz, log K+(int) = 4.98, log K-(int) = -7.80, C+ = 1.10 F m-', C- = 2.22 F m-'. From Schindler and Gamsjager (1972).

253

SURFACE COMPLEXATION MODELS Table I

Values of Intrinsic Rotonation and Dissociation Constants Obtained with the Constant Capacitance Model Using Linear Extrapolation' Solid Aluminum oxides Y-Ab03 Y-A 1203 Y-AI(OHh y-AI00H Iron oxides Fe(OH)3(am) a-FeOOH a-FeOOH Fe304 Fe304 Fe304 Silicon oxides SiO,(am) SiO,(am) Si02(am) Si02(am) Si02(am) Si02(am) Quartz Titanium oxides Anatase Rutile Rutile Clay minerals Kaolinite Montmorillonite Mite

Ionic medium

logK+(int)

log K-(int)

0.1 M NaC10, 0.1 M NaClO, 1 M KN03 0.001 M K N 0 3

7.2 7.4 5.24 5.6

-9.5 -10.0 -8.08 -8.6

0.1 M NaNO, 0.1 M NaCIO, 0.1 M NaClO, 0.1 M KNO, 0.01 M KN03 0.001 M KN03

6.6 6.4 5.9 5.19 4.66 4.40

-9.1 -9.25 -8.65 -8.44 -8.81 -8.97

Farley er al. (1985) Sigg (1979) Sigg (1979) Regazzoni er al. (1983) Regazzoni er al. (1983) Regazzoni et al. (1983)

-

Schindler and Kamber (1968) Sigg (1973) Sigg (1973) Sigg (1973) Fiirst (1976) Gisler (1980) Osaki et al. (1990b) Schindler and Gamsjager (1972) Fiirst (1976) Gisler (1980)

0.1 M 1.0 M 1.0 M 1.o M 1 .O M 1.O M 0.1 M

NaCIO, LiCl NaC10, CSCl NaC10, NaC10, NaCIO,

-

-6.8 -6.57 -6.71 -5.71 -5.7 -6.56 -8.4

3.0 M NaC10,

4.98

-7.80

1.0 M NaCIO, 1.O M NaCIO,

4.46 4.13

-7.75 -7.39

0.01 M NaCl 0.01 M NaCl 0.01 M NaCl

2.4 2.0 7.5

-6.5 -5.6 -11.7

Reference Hohl and Stumm (1976) Kummert and Stumm (1980) Pulfer et al. (1984) Bleam ef al. (1991)

Motta and Miranda (1989) Motta and Miranda (1989) Motta and Miranda (1989)

"Adapted from Schindler and Stumm (1987) and expanded.

nique lead to large differences between the value of logK+(int) and the absolute value of log K-(int) (Westall, 1986). A weakness of the constant capacitance model is that the value of the capacitance, C+, obtained from linear extrapolations below the ZPC is usually not the same as the value, C- , obtained above the ZPC. Table I1 provides values of C, obtained from linear extrapolations of titration data for the iron oxide goethite and the aluminum oxide boehmite. It is clear that for two batches of goethite and for three experiments on the same batch of boehmite, the capacitance exhibits great variability even when variations in log K,(int) are small. Because of this variability, single values

2 54

SABINE GOLDBERG Table I1 Values of Capacitance Obtained with the Constant Capacitance Model Using Linear Extrapolation ~~

~

Solid Geothite‘ a-FeOOH(1) a-FeOOH(I1) Boehmite yAIOOHd Experiment 1 Experiment 2 Experiment 3

~

logK+(int)

C, (F rn-’)’

log K-(int)

C- (F m-2)b

5.9 6.4

1.5 2.7

-8.65 -9.25

3.5 4.4

5.8 5.5 5.5

1.3 2.8 2.9

-8.5 -8.7 -8.8

0.7 0.9 2.0

“C, is obtained from the slope of extrapolation for logK+(int).

b C - is obtained from the slope of extrapolation for logK-(int). ‘Calculated from Sigg (1979) data for two batches of goethite. dCalculated from Bleam et al. (1991) data for three experiments of one batch of boehmite.

of C,considered optimum [e.g., for goethite C = 1.8 F mV2(Sigg, 1979) and for aluminum oxide C = 1.06 F mP2 (Westall and Hohl, 1980)], have often been used in applications of the constant capacitance model. The ability of the constant capacitance model to describe potentiometric titration data on an oxide mineral using the protonation-dissociation constants obtained from the extrapolation method is indicated in Fig. 7. It is clear that the model represents the titration data very well (Schindler and Gamsjager, 1972). Values for the protonation-dissociation constants can also be obtained by optimization of potentiometric titration data using a computer program such as FITEQL (Westall, 1982). FITEQL uses a simultaneous, nonlinear, least-squares method to fit equilibrium constants to experimental data. FITEQL also contains surface complexation models, including the constant capacitance model, to describe surface complexation. Application of the FITEQL program to the titration data of Bleam et al. (1991) produced values of log K+(int) = 6.02 and log K-(int) = -8.45. It can be seen that the difference between the absolute value of log KJint) and the value of log K+(int), A log K,(int), is less for the nonlinear FITEQL optimization than for the linear extrapolation technique (values provided in Table I). The constant capacitance model has been modified and extended to describe potentiometric titration data on the clay mineral kaolinite (Schindler et al., 1987). In addition to the amphoteric surface hydroxyl group, SOH, Schindler et af. (1987) postulated a second surface functional group, XH, which is weakly acidic and can undergo ion exchange with cations from the background electrolyte. An additional assumption is made that cations

SURFACE COMPLEXATION MODELS

255

12 10 -

86 -

42-

0

0.1

0

-0.1

u

(rnol, kg

-0.2

-0.3

)

Figure 7. Fit of the constant capacitance model to potentiometric titration data on anatase, TiOz. Model result is obtained using logK,(int) from linear extrapolation (Fig. 6 ) and is represented by a solid line. From Schindler and Gamsjager (1972).

from the background electrolyte can bind with SOH, forming weak outersphere surface complexes. Thus, in addition to Eqs. (4) and (9, the following reactions are defined: SOH+ C ' Ft SO-- C+ + H+

(83)

X H + C+ $ XC + H'

(84)

In addition to Eqs. (10) and (ll),the equilibrium constants for this application are

The fit of the constant capacitance model to titration data on hydrogen kaolinite is indicated in Fig. 8. Values of K,(int), K c + , and Kxc+ were optimized using the computer program FITEQL (Westall, 1982). Schindler et al. (1987) considered the model fit acceptable but suggested that systematic errors might be due to extension of the Davies equation to ionic strength up to 1M and to the use of the same capacitance value for all ionic strengths. An alternative method for the description of potentiometric titration data on clay minerals was used by Motta and Miranda (1989). These authors used the same modeling approach on these heterogeneous systems as

256

SABINE GOLDBERG

4

6

0

10

-log [HI' Figure 8. Fit of the constant capacitance model to potentiometric titration data on hydrogen kaolinite. H*represents the number of hydrogen ions originating from the kaolinitewater interface; logK+(int) = 4.37, Iog K-(int) = -9.18, log&,+ = -9.84, logKXNa+ = -2.9, C = 2.2 F m-*. Model results are represented by a dashed line ( I = 0.01 M NaCIO,), dotted line (I= 0.1 M NaCIO,), and solid line (I= 1.0 M NaC10,). From Schindler et al. (1987).

had been used for oxide minerals. Values for protonation-dissociation constants were obtained by extrapolating to zero surface charge as described above. Values of log K,(int) for the clay minerals kaolinite, montmorillonite, and illite are given in Table I.

B. TRIPLE-LAYER MODEL The triple-layer model has been used to describe the amphoteric behavior of inorganic surface hydroxyl groups in inert background electrolytes. The adsorption of protons and hydroxyl ions and inert background electrolytes has been investigated on the following surfaces: goethite (aFeOOH) (Davis et al., 1978; Balistrieri and Murray, 1979, 1981; Hsi and Langmuir, 1985; Hayes, 1987), amorphous iron oxide (Davis and Leckie, 1978; Hsi and Langmuir, 1985), magnetite (Fe304)and zirconium dioxide (Regazzoni et al., 1983), titanium oxide (Davis et al., 1978;Sprycha, 1984), manganese oxide (6MnO2) (Balistrieri and Murray, 1982a; Catts and Langmuir, 1986), colloidal silica (MilonjiC, 1987), a-A1203 (Smit and Holten, 1980), y-A1203 (Sprycha, 1989a,b), boehmite (y-A100H) (Wood et af., 1990), and soils (Charlet and Sposito, 1987). Values of the intrinsic protonation and dissociation constants provided in Eqs. (10) and (11) and the intrinsic surface complexation constants for the background electrolyte provided in Eqs. (32) and (33) can be obtained from potentiometric titration curves carried out in the absence of specific

SURFACE COMPLEXATION MODELS

257

metal or ligand adsorption. The assumption is made that a,, the surface charge density in the surface o-plane, is equal to ( F / S a ) ([SOH:] + [SOH: - A-I) below the ZPC and is equal to (-F/Su)([SO-] + [SO- C’]) above the ZPC. Intrinsic equilibrium constant values are obtained by linear extrapolation (Davis et al., 1978) or by the double-extrapolation method (James et ul., 1978). The additional assumptions are made that at low ionic strength a . is equal to (F/Sa)[SOH:] below the ZPC and is equal to ( - F / S a ) [ S O - ] above the ZPC and that at high ionic strength, a, is equal to (F/Su)[SOHZ-A-] below the ZPC and is equal to (-F/Sa)[SO--C+] above the ZPC. A plot of the logarithm of the conditional equilibrium constant ‘ K , , ‘ K c + ,or ‘KA- versus surface charge will yield the logarithm of the intrinsic equilibrium constant K,(int), Kc+(int), or KA-(int) upon extrapolation. The conditional equilibrium constants for protonation-dissociation are defined in Eqs. (79) and (80) and are related to the intrinsic protonationdissociation constants by Eq. (81). The conditional surface complexation constants for the background electrolyte are

[SOH; - A-] cKA- = [SOH][H+][A-]

By combining Eq. (32) with Eq. (87), and Eq. (33) with Eq. (88), one can relate the intrinsic surface complexation constants and the conditional surface complexation constants: Kc+(int) = ‘Kc+ exp[ F(Wp- W o ) / R T ]

(89)

KA-(int) = ‘KA- exp[ F ( q o - W p ) / R T ]

(90) upon taking the logarithms of both sides of Eqs. (81), (89) and (90) and solving for log ‘Kithe following equations are obtained: log ‘ K ,

5

log K,(int)

* RTFWO In( 10)

log ‘Kc+= log Kc+(int) +

F W O-qp,

logE&-= log KA-(int) +

F(y!3 - WO) RT ln(10)

RT ln(10) (93)

Fractional surface charges are defined for a positive surface below the ZPC as a+ = f f , / N s

(94)

258

SABINE GOLDBERG

and for a negative surface above the ZPC as = -aO/Ns

(95)

where Ns = (F/Sa)[S0HlT is the surface mass balance in units of C m-* (Davis et al., 1978). By plotting the titration data of logc& versus a+ or a _ ,an estimate of logKi(int) is obtained from they intercept, where a+ or a- = 0 when a, = q o= 0. The capacitance parameter, C1,can be extracted from the slopes of such plots using Eqs. (92) and (93) (Smit and Holten, 1980; Sposito, 1984a; Blesa et al., 1984b). Values of log Ki(int) for protonation-dissociation and background electrolyte surface complexation have been obtained by linear extrapolation for goethite (Davis et al., 1978; Hayes, 1987), amorphous iron oxide (Davis and Leckie, 1978), magnetite and zirconium dioxide (Regazzoni et al., 1983), titanium oxide (Davis et al., 1978), and soils (Charlet and Sposito, 1987). Figure 9 provides an example of the linear extrapolation technique for amorphous iron oxide (Davis and Leckie, 1978). Using the definitions Eqs. (94) and (95) and the previous assumptions, expressions for a+ and a- can be calculated and are provided in Table 111. Equations (91), (92), and (93) can now be written in terms of a+ and a(Davis et al., 1978):

+:)

log K+(int) = pH + log (1 RT ln(10)

log K-(int) = -pH logKc+(int) = -pH

F*O

+

+ log (1- - RTFlIb ln(10) + log

log K,-(int) = pH + log

:;) (1";)

- log[C+] + F(% RT In(10) RT ln(10)

F P O

(97) (98)

(99)

In the double-extrapolation technique developed by James et al. (1978), two extrapolations are carried out. The intrinsic protonation-dissociation constants are obtained from extrapolation of Eqs. (96) and (97) to a+ or a- = 0 and zero electrolyte concentration: a, = 0, C = 0, qo= qp.The intrinsic surface complexation constants are obtained from extrapolation of Eqs. (98) and (99) to a+ or a- = 0 and infinite electrolyte concentration: a, = 0, C = 1M ,qo= Wp.Again, as in the linear extrapolation procedure, the capacitance parameter, C1,can be obtained from the slopes of the plots of Eqs. (98) and (99). Values of log Ki(int) for protonation-dissociation and background electrolyte surface complexation have been obtained by double extrapolation for goethite (Balistrieri and Murray, 1979, 1981;

259

SURFACE COMPLEXATION MODELS 81

I

$ Y

0

-8

6-

b

log K + (int)

+ Y 0

-8

Q 0.01 0.03 0.05 0.07

3 0

a+=o,/N, Figure 9. The logarithms of the conditional protonation constant and the anion surface complexation constant for the background electrolyte as a function of fractional surface charge for amorphous iron oxide; logK+(int) = 5.1, log KN,,(int) = 6.9. From Davis and Leckie (1978), based on experimental data of Yates (1975).

Hsi and Langmuir, 1985), amorphous iron oxide (Hsi and Langmuir, 1985), boehmite (Wood et al., 1990), manganese oxide (Balistrieri and Murray, 1982a; Catts and Langmuir, 1986), titanium oxide (James and Parks, 1982), and colloidal silica (MilonjiC, 1987). Figure 10 provides an example of the double-extrapolation procedure for manganese oxide (Balistrieri and Murray, 1982a). Table III Approximations for Estimating Intrinsic Protonation-Dissociation and Surface Complexation Constants by Extrapolation Ionic strength Low

pH < ZPC

pH > ZPC

a, = (F/Sa)[SOH;]

[SOH] = [SOHIT- [SOH:] a+ = [SOH;]/[SOH],

[SOH;]/[SOH] High

= a+/(l

- a+)

uo= (F/Sa)[SOH; - A-] [SOH] = [SOH], - [SOH: - A-] a+ = [SOH: - A-]/[SOH], [SOH; - A-]/[SOH] = a + / ( l - a+)

U" =

(-F/sa)[so-]

[SOH] = [SOH], - [SO-] a- = [so-]/[soH], [SO-]/[SOH] = a - / ( l - a-)

u, = ( - F / S a ) [ S O - - C+] [SOH] = [SOH,] - [so-- c'] a- = [so- - C']/[SOH], [SO- - C+]/[SOH]= a-/(l - a-)

SABINE GOLDBERG

2 60

h

I

0

0.1

0.2

or - 0.1 log CNaC, Figure 10. Determination of the intrinsic dissociation constant (top) and cation surface complexation constant for the background electrolyte (bottom) for manganese oxide. The fractional surface charge IY- is multiplied by an arbitrary constant solely to separate data of different concentrations; log K-(int) = -4.2, logK,,+(int) = -3.3. From Balistrieri and Murray (1982a).

An alternative method of determining intrinsic protonation-dissociation constants and surface complexation constants for the background electrolyte has been developed (Sprycha, 1983; 1989a,b; Sprycha and Szczypa, 1984). In this method the assumption is made that the zeta potential, 5, is equal to the diffuse layer potential, q d , at low ionic strength. The protonation-dissociation constants are determined from zeta potential data using a double straight-line extrapolation method (Sprycha and Szczypa, 1984). The assumption is made that a d , the diffuse layer charge, is equal to (F/Sa)[SOH:] below the ZPC and is equal to ( - F / S a ) [ S O - ] above the

SURFACE COMPLEXATION MODELS

261

ZPC. The intrinsic conditional protonation-dissociation constants are calculated from the following equations obtained by taking the logarithm of both sides of Eqs. (79) and (80) (Sprycha and Szczypa, 1984): log ' K + = pH + log[SOH;] - log[SOH] lOg'K- = -pH

(100)

+ 10g[S0-] - log[SOH]

(101) The straight lines are extrapolated to the pH of the ZPC and then to zero electrolyte concentration to obtain the intrinsic protonation-dissociation constants. The surface complexation constants for the background electrolyte are calculated from extrapolation of direct measurements of adsorption densities to zero surface charge (Sprycha, 1983, 1984, 1989a,b). In the Sprycha method the following equations, obtained by taking the logarithm of both sides of Eqs. (87) and (88), are used (Sprycha, 1989b): log"&+ = -pH - log[SOH] - 10g[C+] + lOg[SO- - C']

(102)

+ log[SOHg - A-]

(103)

logcKA- = pH - log[SOH] - log[A-]

The intrinsic surface complexation constants are obtained by extrapolating conditional surface complexation constants as a function of pH to the ZPC. The capacitance parameter, C1, can be obtained from the slope of charge versus potential curves [Eq. (19)] calculated using potential differences determined with Eqs. (92) and (93). The capacitance parameter, C2, can be obtained after determining the potential distribution within the electric double layer using electrokinetic data and Eq. (20). Values of logK,(int), log &+(int), log KA-(int), C1,and C2have been obtained with the Sprycha method for anatase, TiOz (Sprycha, 1984), and aluminum oxide (y-Al,O,) (Sprycha, 1989a,b). Figure 11 indicates the extrapolation techniques of the Sprycha method for aluminum oxide (Sprycha, 1989a,b). Table IV presents values of log&, logK,+(int), and logK,-(int) obtained by various researchers using all three extrapolation methods. As can be seen from Table IV for goethite and rutile, the intrinsic equilibrium constants for double extrapolation are almost identical to those obtained with the linear extrapolation technique. The intrinsic equilibrium constants obtained using the Sprycha method are also very similar to those obtained using the double-extrapolation method, although the latter constants are considered to be less accurate because of the asymptotic nature of the extrapolation (Sprycha, 1984, 1989a). A weakness of the triple-layer model is that, as in the constant capacitance model, the value of the capacitance, C1+, obtained from extrapolation below the ZPC is not equivalent to the capacitance value, C1-, obtained from extrapolation above the ZPC (Smit and Holten, 1980; Blesa

262 Figure 11. Determination of (a) the intrinsic protonation and dissociation constants (from Sprycha, 1989a), (b) the background electrolyte surface complexation constants, and (c) the capacitance, C1 (from Sprycha, 1989b); logK+(int) = 5.0, log K-(int) = -11.25, logKN,+(int) = -8.6, logKc.-(ht) ~7.5.

Table N

Values of Intrinsic Protonation-Dissociation and Background Electrolyte Surface Complexation Constants Obtained with the Triple-Layer Model by Extrapolation ~

~~

Solid

N

~

~

~

~

log K-(int)

~

Ionic medium

logK+(int)

log K,+(int)

wA1203 Y-Al203 a-FeOOH" a-FeOOH Fe@, Fe(OH)3(am) TiO,, rutile' Ti02, rutile" ZrO, Oxisol soil Oxisol soil Ultisol soilb

NaBr NaCl KNO, NaN0, KNo3 NaN0, KNO, LiNO, KNo3 NaCl KNo3 NaCl

4.9 4.2

5.8 4.4 5.1 2.6 2.6 4.2 2.33 2.07 1.8

Linear extrapolation -4.0 -11.3 -10.8 -8.9 -11.1 -8.8 -9.0 -7.2 -10.7 -9.0 -9.0 -7.1 -9.0 -7.1 -8.6 -7.6 -6.34 -2.06 -5.97 -2.0 -5.9 -

y-Al00H a-FeOOH a-FeOOH a-FeOOH

KN03 NaCl KCl NaNO,

5.4 4.9 4.9 4.5

Double extrapolation -11.7 -6.3 -10.4 -9.6 -10.4 -9.6 -12.0 -10.1

~~~~~

log&,-(int) 5.0

-

6.1 7.7 6.2 6.9 4.5 4.5 5.2 -1.80 2.4

10.9 5.5 5.5 7.0

Reference Smit and Holten (1980) Sprycha (1989a) Davis er al. (1978) Hayes (1987) Regazzoni et al. (1983) Davis and Leckie (1978) Davis ef al. (1978) Davis et al. (1978) Regazzoni et al. (1983) Charlet and Sposito (1987) Charlet and Sposito (1987) Charlet and Sposito (1987)

Wood et al. (19W) Balistrieri and Murray (1979) Balistrieri and Murray (1979) Hsi and Langmuir (1985) (continues)

Table IV (Continued) Solid

Ionic medium

log K,(int)

a-FeOOH" a-Fe20; Fe(OH),(am) TiO, , anatased TiOz, rutiled Ti02, rutiled TiO,, rutile" TiO,, rutile'l 6-Mn0, 6-Mn0, SiO,(am) Si02(am) SiO,(am) Si02(am)

KN03 KCI NaNO, NaCl NaC104 NaN0, KNo3 LiN03 NaCl NaN03 LiCl NaCl KCI CSCl

4.2 6.7 4.8 3.2 2.7 2.8 2.6 2.7

NaBr NaCl NaCl

5.0 3.1

log K-(int)

log K,-(int)

Double extrapolation -9.0 -10.3 -9.5 -11.1 -9.3 -8.7 -7.1 -9.1 -7.2 -9.1 -7.2 -9.0 -7.1 -9.1 -7.2 -4.2 -3.3 -6.2 -3.5 -8.2 -6.5 -8.2 -6.9 -8.2 -7.0 -8.2 -7.0 - 10.5

Sprycha electrokinetic extrapolation -8.7 -11.25 -8.6 -8.9 -5.8

"Based on experimental data of Yates (1975). 'Based on experimental data of Gallez el al. (1976). 'Based on experimental data of Breeuwsma and Lyklema (1971). dBased on experimental data of Btrubt and deBruyn (1968a,b)

log K,-(int)

Reference

6.2 7.5 7.0 4.6 4.5 4.5 4.5 4.0

James and Parks (1982) James and Parks (1982) Hsi and Langmuir (1985) James and Parks (1982) James and Parks (1982) James and Parks (1982) James and Parks (1982) James and Parks (1982) Balistrieri and Murray (1982a) Catts and Langmuir (1986) MilonjiC (1987) MilonjiE (1987) MilonjiC (1987) Milonjif (1987)

8.9 7.5 6.2

Sprycha (1983) Sprycha (1989a,b) Sprycha (1984)

265

SURFACE COMPLEXATION MODELS

et al., 1984b). Smit and Holten (1980) obtained C1+ = 1.5 F m-’ and C1- = 3.5 F m-2 for aluminum oxide in NaBr solution. Table V provides values of C1 annd C2obtained from the Sprycha electrokinetic extrapolation technique for three different ionic strengths and four solution pH values. It is clear that the capacitances are not completely constant, as assumed in the triple-layer model, but vary with changing pH and ionic strength (Sprycha, 1984,1989b). Capacitance values also vary as a function of solution temperature (Blesa et al., 1984b). With the exception of the above three studies, the capacitances have universally been taken as adjustable parameters in applications of the triple-layer model (Sposito, 1984a). The capacitance C, is fixed at 0.2 F m-2 and the capacitance C1is adjusted to optimize the fit of the triple-layer model to the experimental data. The ability of the triple-layer model to describe potentiometric titration data on an iron oxide mineral is indicated in Fig. 12. The model describes the titration data relatively well especially at high ionic strength. As for the constant capacitance model, use of the extrapolation procedures leads to large differences between log K+(int) and the absolute value of log K-(int) (Koopal et al., 1987). Application of the FITEQL program to the titration data of Yates (1975) on titanium oxide produced values of logK+(int) = 5.15 and log K-(int) = -6.61, giving a much smaller value of A log K,(int) (Westall and Hohl, 1980). However, it was pointed out by Koopal et al. (1987) that the numerical procedure can provide good fits with either a large or a small value of A logK,(int), depending on which is chosen as the starting value in the FITEQL optimization. Values of the intrinsic equilibrium constants for surface complexation of background electrolytes can also be obtained using the computer program FITEQL containing the triple-layer model (Westall, 1982). Table V Values of Capacitance Obtained with the Triple-Layer Model Using Sprycha Electrokinetic Extrapolation Above the ZPC“ __

~~

Capacitance (F m-’)

pH8

pH9

pH 10

CZ

0.001 M NaCl 1.20 1.02 0.14 0.20

1.05 0.20

1.08 0.19

c1

0.01 M NaCl 1.25 1.12

1.17

1.27

C,

0.1 M NaCl 1.35 1.35

1.47

1.58

c1

~

pH7

~~

“Adapted from Sprycha (1984).

266

SABINE GOLDBERG

4

6

8

10

PH Figure 12. Fit of the triple-layer model to potentiometric titration data on goethite for three ionic strengths of NaCl and KCI. Model results are represented by solid lines; logK+(int) = 5.57, logK-(int) = -9.52, and logK,,+(int) = -8.40 were obtained by double extrapolation; logKc,-(int) = 7.00 was adjusted to fit the data with C1= 1.4 F m-’. From Balistrieri and Murray (1981), reproduced with permission from the American Journal of Science.

The kinetics of adsorption/desorption on oxide surfaces can be investigated using relaxation techniques. Sasaki et al. (1983) used the electric field pulse technique to investigate adsorption/desorption kinetics of chloride and perchlorate anions on goethite. Surface complexation modeling results are necessary to analyze the kinetic data. Sasaki et al. (1983) were able to describe both their adsorption and their kinetic data using the triple-layer model. The magnitude of the Kcl-(int) value obtained was in good agreement with that previously found by Davis et d.(1978). The first application of the triple-layer model of heterogeneous natural samples was carried out on a Brazilian Oxisol, a variable-charge soil, by Charlet (1986; Charlet and Sposito, 1987). The authors simultaneously measured proton titration and background electrolyte adsorption. The intrinsic equilibrium constants obtained for protonation-dissociation on the Oxisol were similar to those for strongly acidic oxide minerals, whereas the intrinsic equilibrium constants obtained for surface complexation of the background electrolytes were similar to values found for oxide minerals. These results indicate that the Oxisol behaved similarly to strongly acidic oxides (Charlet and Sposito, 1987).

SURFACE COMPLEXATION MODELS

267

C. STERNVSC-VSP MODEL The Stem VSC-VSP model has been used to describe the surface charge densities of various minerals in inert background electrolytes. The charging behavior has been investigated for iron oxides, goethite (Barrow et al., 1980a; Madrid et al., 1983; Bloesch et al., 1987), lepidocrocite (Madrid et al., 1983; Madrid and Diaz-Barrientos, 1988), and hematite (Madrid et al., 1983), manganese oxides, cryptomelane and birnessite (McKenzie, 1981), clay minerals (Madrid and Diaz-Barrientos, 1988; Madrid et al., 1989), and soils (Bowden et al., 1977). Unlike in the constant capacitance model and the triple-layer model, where values of surface site density, equilibrium constants, and capacitances can be obtained from experiment, in the Stern VSC-VSP model these parameters are optimized to fit the charging data. Table VI presents values of Ns , log KH , log KOH, log K,,, , logK,,, Cop, and C,, obtained by computer optimization for various minerals. The ability of the Stern VSC-VSP model to describe potentiometric titration data on an iron oxide mineral is indicated in Fig. 13. The model described the titration data well. Application of the Stern VSC-VSP model to potentiometric titration data on the clay minerals illite and pyrophyllite was carried out (Madrid et al., 1989). The fit appeared good; however, evaluation of the fit of the model is difficult because the raw experimental titration data were not presented. Application of the Stern VSC-VSP model to charging data for a heterogeneous system has been studied for soil clays from a Spanish Eutrochrept soil (Madrid et al., 1983). The model fit was less satisfactory for the soil clays than for reference iron oxide minerals. However, again, the fit of the model is difficult to evaluate since no raw titration data were presented. A simplified Stern VSC-VSP model containing no P-plane and allowing no surface complexation for the background electrolyte was originally developed by Bowden et al. (1977). These authors fit titration data on an Acrohumox soil with this model (see Fig. 14). Madrid and Diaz-Barrientos (1988) fit titration data on the clay mineral montmorillonite using the simplified Stern VSC-VSP model. The Stern VSC-VSP model was applied to mixtures of iron oxides and clay minerals (Madrid and Diaz-Barrientos 1988; Madrid et al., 1989). For this application, the Stern VSC-VSP model was used for iron oxides and illite and the simplified Stern VSC-VSP model was used for montmorillonite. A common diffuse layer is considered. Additional parameters are up,the permanent charge on the clay surface, and R, the proportion of clay surface in the mixture. Each surface has its own surface charge densities, Ns , binding constants, Ki, and capacitances, Ci.The fit of the Stern

Table VI Values of Madmum Surface Charge Density, Binding Constants, and Capacitances Obtained with the Stern VSC-VSP Model by Computer Optimization of Charging Data Solid

Ionic medium

Ns

COS

cSd

(pmol m-2)

(F mP2)

(F m-2)

log KH

log KO"

log K,,

log K,,

-b

12.4 0.965 -b -b

-b

-b

7.59 7.94 7.2 7.5 7.9

5.76 3.98 4.2 7.3 8.4

-0.14 -0.086 -1.2 -1.5 -1.4

-0.96 -0.18 -0.92 -0.52 -0.52

Barrow er al. (1980aY Bloesch et al. (1987) Madrid et al. (1983) Madrid et al. (1983) Madrid et al. (1983)

1.3 0.97 4.3 1.4

1 2 1 4

12 12 12 12

-0.3 -0.3 -0.3 -0.3

-0.3 -0.3 -0.3 -0.3

McKenzie (1981) McKenzie (1981) McKenzie (1981) McKenzie (1981)

-b

-b -b -b

7.60 6.04 6.11

8.00 7.26 7.64

-0.40 -1.4 -2.0

-0.52 -0.097 -0.52

Madrid et al. (1983) Madrid er al. (1983) Madrid et 01. (1983)

-b

3.4 -0.081

5.8 7.7

2.7 -1.5

-2.0 4.3

Madrid et al. (1989) Madrid et al. (1989)

Iron oxides Goethite Goethite Goethite Lepidocrocite Hematite

NaCl NaCl NaCl NaCl NaCl

5.58 10.00 45 10 12

3.10 0.914

Manganese oxides Cryptornelane 41 Cryptomelane 42 Birnessite 40 Birnessite 44

NaN03 NaN03 NaN03 NaN03

7.6 10.5 28 22

2.0

soil clays Eutrochrept 3 Eutrochrept 4 Eutrochrept 5

NaCl NaCl NaCl

20 43

-b -b

Clay minerals Illite Pyrophyllite

NaCl NaCl

9.3 11

-b -b

17

~~

-b

2.5 7.7 5.2

~~

"Based on experimental data of Hingston (1970). bParameters required in model application whose values were not provided by the authors.

Reference

SURFACE COMPLEXATION MODELS

2 69

A 4.2M 0 I.OM

4-

A 0.1 M

0.01M 2-

0-

2L

,

,

4

5

6

7

8

9

1

0

PH Figure 13. Fit of the Stern VSC-VSP model to potentiometric titration data on goethite for four ionic strengths of NaCI. Model results are represented by solid lines. Model parameter values areprovidedinTable VI. After Barrow etal. (1980a), basedonexperimentaldataof Hingston (1970).

I

1

I

1.0 M 0.8 - A 0.1 M

I

0

N

‘E

0 E

I

O/’A

,I

0.01M

h

0.6 -

00.001M

0.4

,

.’ yy,I

I

v

0.2

-

o/

’ / /A

//

/,

q,&

/w

,’

,

/

,

/ ’’ 0”

,“.‘

-

,

,

-

,/

0

c

-

/‘

/

/ / / O / A

-

/

010.

I I

-

a

/ /

1

I

270

SABINE GOLDBERG

VSC-VSP model to mixtures could only be evaluated for lepidocrocitemontmorillonite; no model fits to data were presented for other mixtures. No raw titration data were presented for lepidocrocite-montmorillonite, but the model fit, as best could be evaluated, was good. Additional research is needed to evaluate the applicability of the Stern VSC-VSP model for describing charging behavior of oxide-clay mixtures.

D. GENERALIZED TWO-LAYER MODEL So far, application of the generalized two-layer model has been restricted to reactions occurring on the surface of hydrous ferric oxide (Dzombak and Morel, 1990). Dzombak and Morel much preferred computer optimization of equilibrium constants over graphical extrapolation techniques. They considered nonlinear least-squares optimization with the computer program FITEQL (Westall, 1982), containing the generalized two-layer model, to be less tedious and more precise than graphical methods. Additional advantages of computer optimization are that the parameters are considered bias free and that quality-of-fit criteria and parameter standard deviations are available. In obtaining best estimates for intrinsic conditional protonation-dissociation constants, Dzombak and Morel (1990) applied FITEQL to individual titration data sets at each ionic strength. The individual optimum values of logK,(int) were weighted to obtain the best estimate using the following equation:

where (qo,K(int))iis the standard deviation calculated by FITEQL for logK(int) of the ith data set. Table VII presents values of individual logK,(int) and best estimates of logK,(int) obtained by FITEQL computer optimization for hydrous ferric oxide. Values of logK,(int) obtained by Dzombak and Morel (1990) compare well with average values for a literature compilation of experimental log K , (int) values provided in Goldberg and Sposito (1984a) for iron oxide minerals. These authors found logK+(int) = 7.31 -+ 1.11 and logK-(int) = -8.80 f 0.80. The large standard deviations result because data for diverse iron oxides were averaged. The ability of the generalized two-layer model to describe potentiometric titration data on hydrous ferric oxide is indicated in Fig. 15. The model describes the data well using both individual data set values and best estimates of logK,(int).

271

SURFACE COMPLEXATION MODELS

Table VII Values of Intrinsic Rotonation and Dissociation Constants Obtained with the Generalized Two-Layer Model Using FITEQL Computer Optimization for Hydrous Ferric Oxide” fu

Ionic medium

log K+(int) -t (T

0.1 M NaNO, 0.01 M NaNO, 0.001 M NaN0, 0.1 M NaCIO, 0.25 M NaCIO, 0.5 M NaCI0, 0.01 M KNO, 0.1 M NaN0, 0.01 M NaN0, 0.001 M NaNO,

7.03 f 0.035 7.57 -t 0.038 8.20 f 0.045 7.69 T 0.048 7.17 f 0.032 6.63 2 0.033 7.10? 0.026 7.08 f 0.028 7.58 f 0.110 7.54 f 0.040

-8.74 2 0.065 -8.74 ir 0.15* -8.74k 0.15’ -9.25 2 0.118 -9.51 f 0.095 -9.10 0.065 -8.89 2 0.045 -9.01 2 0.034 -8.40k 0.126 -8.40 f 0.15’

7.29 f 0.10‘

-8.93

log K-(int)

Experimental data source ~

log K,(int)

*

~ _ _ _ _ ~

Davis (1977) Davis (1977) Davis (1977) Swallow (1978) Swallow (1978) Swallow (1978) Yates (1975) Hsi and Langmuir (1985) Hsi and Langmuir (1985) Hsi and Langmuir (1985)

-t 0.07‘

“From Dzombak and Morel (1990) blog KJint) and q,, K-(,nt) fixed at these values. Convergence was not possible unless one value of log K,(int) was fixed. ‘95% confidence intervals.

- 0.4

0

0.4

0.8

Acid added (mrnol liter-’) Figure 15. Fit of the generalized two-layer model to potentiometric titration data on hydrous ferric oxide for three ionic strengths of NaNO,. Model results are represented by solid lines for individual log K,(int) and by dashed lines for log K,(int). Model parameter values are providedinTable VII. After DzombakandMorel(1990), basedonexperimentaldataofHsi and Langmuir (1985), reproduced with permission from John Wiley and Sons.

272

SABINE GOLDBERG

E. ONE-& MODEL The one-pK model has been used to describe potentiometric titration data for various minerals in inert background electrolyte solution. The one-pK model was applied to titration data on titanium oxide (van Riemsdijk et al., 1986), hematite (a-Fe203), amorphous iron oxide (van Riemsdijk et al., 1987), A1(OH)3 gibbsite, and y-A1203 (Hiemstra et al., 1987). In the most simplified application of the one-pK model, no surface complexation reactions are considered for the background electrolyte. In this approach, the capacitance of the Stern layer is the only adjustable parameter. The association constant KH is obtained from the ZPC: log KH = PHo = ZPC

(105) where pH, is the proton concentration at the ZPC (van Riemsdijk et al., 1986). Titration data on titanium oxide (van Riemsdijk et al., 1986) and gibbsite (Hiemstra et al., 1987) were modeled successfully in this fashion. The ability of the simplified one-pK model to describe potentiometric titration data on titanium oxide is indicated in Fig. 16. An improvement in the model fit can be obtained by inclusion of surface complexation constants for the background electrolyte. Values of

-120

- 80 h

N

E

-40

0

E

Y

b0

O 40

80 -

2

-

1

0

I

2

3

4

pH-ZPC Figure 16. Fit of the simplified one-pK model to potentiometric titration data on titanium oxide for three ionic strengthsof KN03. Model results are represented by solid lines; log KH = 5.80 and C = 1.33 F m-’. From van Riemsdijk et al. (1986),based on experimental data of Yates (1975).

2 73

SURFACE COMPLEXATION MODELS

equilibrium constants and capacitance were obtained with a computer algorithm and are presented in Table VIII. Hiemstra et al. (1987) were able to describe potentiometric titration data well on gibbsite with two different approaches. The model fit using no surface complexation constants for the background electrolyte was good but produced a low value of capacitance. A larger value of capacitance was obtained using surface complexation constants for the background electrolyte and the assumption that only the edge surface area of gibbsite is reactive. The quality of the model fit was virtually identical and excellent in both cases. Figure 17 indicates the ability of the one-pK model to describe potentiometric titration data on gibbsite. The one-pK model was expanded to include surface heterogeneity (van Riemsdijk et al., 1986). It was found that the shape of the potentiometric Table VIII Values of Equilibrium Constants and Capacitance Obtained with the One-pK Model by Computer Optimization of Potentiometric Titration Data Solid

Ionic medium

log KH log Kc+ log KA-

C(F m-')

Reference van Riemsdijk et al. (1986) van Riemsdijk et al. (1986) van Riemsdijk et al. (1987) van Riemsdijk et al. (1987) Hiemstra et al. (1987) Hiemstra et al. (1987) Hiemstra et al. (1987) Hiemstra et al. (1987) Hiemstra et al. (1987) Hiemstra et a/. (1987)

KN03

5.8

-

-

1.33

KN03

5.8

- 1.02

-1.28

0.83

KN03

8.4

-

-

1.72

KNOj

7.9

-0.75

-0.89

1.54

NaCl

10

-

-

0.26

NaCl

10

-0.1

-0.1

1.40"

NaCl

8.5

-0.1"

-0.1'

1.2

NaCIO,

8.3

-0.1'

-0.1'

1.1

NaCIO,

8.3

-0.1'

-0.1'

1.o

NaCIO,

8.7

-0.1"

-0.1'

1.2

Considering only the singly coordinated groups at the particle edges to be reactive. bBased on experimental data of Huang and Stumm (1973). 'log Kc+ and log KA- fixed at the gibbsite values. dBased on experimental data of Westall and Hohl (1980). 'Based on experimental data of Hohl and Stumm (1976). fBased on experimental data of Kummert and Stumm (1980).

SABINE GOLDBERG

2 74 400

300 h

N

'E 0 200 E Y

bo 100 0 3

5

7

9

II

PH Figure 17. Fit of the one-pK model to potentiometric titration data on gibbsite for three ionic strengths of NaCI, assuming that only the edge surface area is reactive. Model results are represented by solid lines; log KH = 10,log KNa+= log Kc,- = -0.1, and C = 1.40 F m-'. After Hiemstra er al. (1987).

titration curves was insensitive toward the degree of heterogeneity, allowing a homogeneous model to be used to represent a heterogeneous oxide system (van Riemsdijk et al., 1986). The one-pK model is a special case of a more generalized model called the multisite complexation model, MUSIC, developed by Hiemstra et al. (1989a,b). This model considers equilibrium constants for the various types of surface groups on the various crystal planes of oxide minerals and was used to describe the charging behavior of gibbsite, goethite, hematite, rutile, and silica (Hiemstra et al., 1989b). The MUSIC model has not yet been applied to describe ion adsorption. Because of its increased complexity such an application may entail an unreasonably large number of adjustable parameters.

IV. APPLICATION OF MODELS TO METAL ION ADSORPTION REACTIONS ON OXIDES, CLAY MINERALS, AND SOILS A. CONSTANT CAPACITANCE MODEL Characterization of metal ion adsorption behavior as a function of solution pH results in curves termed adsorption edges. The constant capacitance model has been used to describe metal ion adsorption edges on silica

275

SURFACE COMPLEXATION MODELS

(Schindler et al., 1976; Osaki et af., 1990a,b), aluminum oxide (Hohl and Stumm, 1976), iron oxide (Sigg, 1979; Lovgren et al., 1990), titanium oxide (Furst, 1976; Gisler, 1980), and the clay mineral kaolinite (Schindler et al., 1987 Osaki et al., 1990b). The conditional equilibrium constants for metal adsorption in the constant capacitance model are 'KL = [SOM("-')] [H'] [SOH][Mm+] CK2

- [(S0)2M(m-2)][H+]2 M-

[SOH]2[Mm']

In order to graphically evaluate the surface complexation constants, the simplifying assumption is made that ly = 0. This assumption produces the result that the conditional intrinsic equilibrium constants are equal to the conditional equilibrium constants: Kh(int) = 'KL . For the special case of a divalent metal ion, M2+, this expression holds true universally even without any simplifying assumption because a neutral surface complex is formed. Excellent fits to metal adsorption data were obtained with the constant capacitance model despite the simplification. Figure 18 presents the ability of the constant capacitance model to describe metal adsorption on silica. The lack of dependence of the equilibrium constants on surface charge indicates a self-consistency problem in such applications of the constant capacitance model (Sposito, 1984a). This limitation can be overcome by computer optimization of the intrinsic surface complexation constants. This approach negates the requirement for simplifying assumptions and has been used by Lovgren et al. (1990). 100

80

8( e % 60

a

I 40

s

20 n "

0

1

2

3

4

5

6

7

8

9

-log (H+)

Figure 18. Fit of the constantcapacitancemodel to metal adsorptionon silica. Model results are represented by solid lines. Model parameters are provided inTable IX. From Schindleret a[. (1976).

276

SABINE GOLDBERG

These authors described aluminum complexation on the iron oxide, goethite, using the surface complexes SOAlOH+ and SOAl(OH);!, where log KAloH+(int)= - 1.49 and log KAl(oH),(int)= -9.10. Table IX provides values for metal surface complexation constants obtained with the constant capacitance model for various materials. Table IX Values of Metal Surface Complexation Constants Obtained with the Constant Capacitance Model Using Graphical Methods" Solid

Metal

Ionic medium

log 'KL

y-Al203 Y-AlzO3 y-A1203 Y-Ah03 a-FeOOH Fe304 SiO,(am) SiO,(am) SO2(am) Si02(am) SiO,(am) SiO,(am) Si02(am) Si02(am) Quartz Quartz Quartz TiOz, rutile Ti02, rutile TiO,, rutile Ti02, rutile TiO,, rutile 6-MnOZ Kaolin Kaolin Kaolin Particulates' Particulates' Particulates' Sediments' Sediments" Sediments'

Ca2+ Mg2+ Ba2+ Pb2+ Mg2+ coz+ Mg2+ Fe3+ Fe3+ cu2+ CdZ+ Pb2+

0.1 M NaN03 0.1 M NaN03 0.1 M NaN03 0.1 M NaC10, 0.1 M NaClO,

-6.1 -5.4 -6.6 -2.2 -6.2 -2.44 -7.7 -1.77 -0.81 -5.52 -6.09 -5.09 -5.83 -3.8 -0.97 -5.0 -4.4 -5.90 -4.30 -1.43 -3.32 0.44 -5.5 -0.683 -2.8 -2.1 -0.78 k 0.6 -3.5 f 0.5 -2.9 k 0.2 -0.74 k 0.8 -2.9 0.7 -2.5 2 0.4

co2+

Zn2+ Fe3+ co2+ Zn2+ Mg2+ co2+

cu2+ Cd2+ Pb2+ Ca2+ Fe3+ co2+ Zn2+ Fe3+ co2+ Zn2+ Fe3+ co2+ Zn2+

I=O 1 M NaC10, 3 M NaCIO, 0.1 M NaC10, 1 M NaCIO, 1 M NaC10, 1 M NaCIO, 0.1 M NaC10, 0.1 M NaCIO,

0.1 M NaCIO, 0.1 M NaCIO, 0.1 M NaCIO, 1 M NaC10, 1 M NaCIO, 1 M NaCIO, 1 M NaC10, 1 M NaC10, 0.1 M NaN03 0.1 M NaCIO, 0.1 M NaClO, 0.1 M NaC104 0.1 M NaC10, 0.1 M NaCIO, 0.1 M NaC10, 0.1 M NaCIO, 0.1 M NaC10, 0.1 M NaCIO,

*

log 'KLb -8.1 -14.7 -6.71 -17.15 -4.22 -11.19 -14.20 -10.68 -11.4 -9.2

-

-

-

-13.13 -10.16 -5.04 -9.00 -1.95

-

-7.9 -

Reference Huang and Sturnm (1973) Huang and Stumm (1973) Huang and Stumm (1973) Hohl and Stumm (1976) Sigg (1979) Tamura et al. (1983) Gisler (1980) Schindler et al. (1976) Osaki er al. (1990b) Schindler et al. (1976) Schindler et al. (1976) Schindler et al. (1976) Osaki et al. (1990b) Osaki et al. (1990b) Osaki er al. (1990b) Osaki et at. (1990b) Osaki er al. (1990b) Gisler (1980) Gisler (1980) Furst (1976) Fiirst (1976) Fiirst (1976) Stumm et al. (1970) Osaki et al. (1990b) Osaki et al. (1990b) Osaki et al. (1990b) Osaki et al. (1990b) Osaki et al. (1990b) Osaki et al. (1990b) Osaki et al. (1990b) Osaki et al. (1990b) Osaki et al. (1990b)

"Adapted from Schindler and Stumm (1987) and expanded. bFor divalent metal ions log KL(int) = log ' K L . Averages for three particulars or three sediments obtained from natural waters of Japan.

SURFACE COMPLEXATION MODELS

277

Application of the constant capacitance model to metal adsorption edges on the clay mineral kaolinite and particulate and sediment samples from natural waters was carried out by Osaki et al. (1990b). These authors used the same modeling approach on these heterogeneous systems as had been used for oxide minerals. The equilibrium constants for adsorption of Co2+ and Zn2+ exhibited good reproducibility, however, the standard deviation of logCKi, was great (see Table IX). An alternative model for the description of metal adsorption on kaolinite is the extended constant capacitance model introduced in Section II1,A (Schindler et al., 1987). In this application, reactions given by Eqs. (4), (3, (6), (7), (83), and (84) are defined. In addition, an ion exchange reaction of the surface functional group, XH, between the cation of the background electrolyte and the metal ion, M2', is defined:

*

2XC + Mz+ X,M

+ 2C'

(108) The equilibrium constants for this application are given in Eqs. (lo), ( l l ) , (12), (13), (85), (86), and (109):

[c' ]2/([xc]2[M2+1)

KXM = IX2M]

(109) The fit of the constant capacitance model to adsorption of lead on hydrogen kaolinite is indicated in Fig. 19. Despite the fact that values of the equilibrium constants were obtained using the computer program FITEQL, the assumption was made that T=O. As for potentiometric titration data (see Section III,A), Schindler et al. (1987) considered the model fit acceptable. 100

0

4

5

6

7

PH Figure 19. Fit of the constant capacitance model to lead adsorption on kaolinite; log KxPb= 2.98,10g'Kbb = -2.45,logCK$, = -8.ll.Modelresultsarerepresentedbyadashed line(Z= 0.01 M NaClO,),dottedline (I=0.1 M NaC104),andsolidline(Z= 1.0 M NaC10,). From Schindler et al. (1987).

2 78

SABINE GOLDBERG

“Adsorptive additivity” is a concept developed by Honeyman (1984) that multicomponent mixtures of oxides can be represented as a collection of pure solids. Upon testing this hypothesis for binary oxide mixtures, Honeyman (1984) found significant deviations. The constant capacitance model was used to test models of particle interaction of binary mixtures of amorphous silica-iron (Anderson and Benjamin, 1990a) and ironaluminum (Anderson and Benjamin, 1990b). “Adsorptive additivity” was not observed for these systems. In binary Si-Fe oxide suspensions, silica was considered to partially dissolve, and soluble silicate to adsorb onto iron oxide (Anderson and Benjamin, 1990a). Silver, zinc, and cadmium adsorption were virtually unaffected. In binary Fe-A1 suspensions, the reactive iron surface was considered blocked or replaced by reactive aluminum surface (Anderson and Benjamin, 1990b). Cadmium and silver adsorption were decreased and zinc adsorption was increased in the binary system. This divergent behavior was qualitatively described by the assumed mechanism. The surface precipitation model extends the surface complexation modeling approach by considering precipitation of ions on the solid. This model was developed by Farley et al. (1985) and incorporated into the constant capacitance model. Loss of ions from solution is described by surface complexation at low concentration and by surface precipitation as a solid solution at high concentration. The solid solution composition varies continuously between the original solid and the pure precipitate of the sorbing ion. The surface precipitation model can be incorporated into any surface complexation model, such as the generalized two-layer model (see Section IV,D). Reactions of the surface precipitation model for divalent metal sorption onto a trivalent oxide are (Farley et al., 1985) as follows: Adsorption of M2+onto S(OH)3(s):

+ M2+ + 2H20 Precipitation of M2+: =SOH

=MOH;

Precipitation of =SOH

S(OH),,) +=MOH:

+ H+

+ Mz+ + 2H20 * M(OH)Z,,) + =MOH; + 2H+

(110) (111)

s3+: + S3++ 3H20 $ S(OH),,,, + =SOH + 3H+

(112) The equilibrium constants for the reactions as written are: KadsMfor Eq. (110), 1/Ksp? for Eq. ( l l l ) , and l/Ksps for Eq. (112). The ability of the surface precipitation model to describe lead adsorption on amorphous iron hydroxide is indicated in Fig. 20. It can be seen that the model describes the data well although very few data points are available. The surface precipitation model has also been applied to cadmium, cobalt,

SURFACE COMPLEXATION MODELS

-8

-7

-6

-5

2 79

-4

log [Pb2'] Figure 20. Fit of the constant capacitancemodel containingthe surface precipitationmodel to a lead sorption isotherm on amorphous iron hydroxide at pH 4.5. Model results are represented by a solid line. r p b = ([=PbOH] + [Pb(OH,(,,])/TOTFe; logKadsPb= 5.0, log KspPb= 6.9, logKspFe= 2.6. From Farley et af. (1985), based on experimental data of Benjamin (1978).

manganese, and zinc adsorption on calcite (Comans and Middelburg, 1987) and to manganese adsorption on siderite (Wersin et al., 1989).

B. TRIPLE-LAYER MODEL The triple-layer model has been used to describe alkaline earth and metal ion adsorption edges on aluminum oxide, titanium oxide, amorphous iron oxide (Davis and Leckie, 1978; Benjamin and Bloom, 1981; Zachara et al., 1987; Cowan et al., 1991), goethite (a-FeOOH) (Balistrieri and Murray, 1981, 1982b; Hayes, 1987), manganese oxide (Balistrieri and Murray, 1982a; Catts and Langmuir, 1986), and soil (Charlet, 1986; Charlet and Sposito, 1989). Intrinsic conditional equilibrium constants for the triple-layer model have been obtained by using the computer programs MINEQL (Westall et al., 1976) and MICROQL (Westall, 1979), or by computer optimization using the FITEQL program (Westall, 1982). In general, in the application of the triple-layer model to metal adsorption, the reactions Eqs. (22) and (23), with their equilibrium constants Eqs. (28) and (29), are considered. Figure 21 presents the ability of the triple-layer model to describe silver adsorption on amorphous iron oxide. The ability of the model to fit the adsorption data is good. Table X provides values for intrinsic metal surface complexation constants obtained with the triple-layer model for various materials. Adsorption of the alkaline earth cations calcium and magnesium on manganese

80

J

6 4

f

w z

40

20 20 '

4 55

66

7

88

PH Figure 21. Fit of the triple-layermodel to silver adsorption on amorphousiron oxide. Model results are represented by a solid line; log Ka,(int) = -4.9, logKLg(int) = -12.1. From Davis and Leckie (1978).

Table X

Values of Intrinsic Metal Surface Complexation Constants Obtained with the Triple-Layer Model by Computer Optimization

Solid

Metal Ionic medium log KL(int)

log KL(int)

Reference

Original triple-layer model y-Al20;

Fe(OH),(am) Fe(OH),(am) Fe(OH),(am) Fe(OH),(am) Fe(OH),(am) Fe(OH),(am) a-FeOOH a-FeOOH a-FeOOH a-FeOOH a-FeOOH a-FeOOH TiOz 6-MnO, S-Mn02 S-Mn02 6-MnO, 6-Mn0, Oxis01 soil Oxisol soil a-FeOOH a-FeOOH a-FeOOH

Pb2+ Ag+ Cu2+ Cd2+ Co2+ Zn2+ Ca2+ Ca2+ Mg2+ Cu2+ Pb2+ Zn2+ Cd2+ Cd" Caz+ Mg2+ Pb2+ Zn2+

NaC104 NaN0, NaN0, NaN03 NaNO, NaN0, NaNO, NaCl NaCl NaCl NaCl NaCl NaCl KN03 NaCl NaCl NaNO, NaN0, NaN03

Ca2+ Mg2+ Pb2+ Cd2+ Ba2+

Ca(C10.J2 Mg(C10& NaN0, NaN0, NaNO,'

cu2+

-5.0 -4.9 -4.1 -4.8 -4.8 -4.8 -6.3 -5.00 -5.45 -3.0 -1.8

-

-1.3 -1.8 -0.1 1.8 -1.5

- 10.3 -12.1 -9.0 -11.25 -11.60 -10.50

-

-14.50 - 14.25 -7.0 -5.0 -9.15 -9.35 -8.7 -4.0' -3.3' -7.5 -6.5 -8.8

Modieed triple-layer model -1.26d -1.76d 2.30d -1.05d -5.10 -14.20

Davis and Leckie (1978) Davis and Leckie (1978) Davis and Leckie (1978) Benjamin and Bloom (1981) Benjamin and Bloom (1981) Benjamin and Bloom (1981) Zachara et al. (1987) Balistrieri and Murray (1981) Balistrieri and Murray (1981) Balistrieri and Murray (1982b) Balistrieri and Murray (1982b) Balistrieri and Murray (1982b) Balistrieri and Murray (1982b) Davis and Leckie (1978) Balistrieri and Murray (1982a) Balistrieri and Murray (1982a) Catts and Langmuir (1986) Catts and Langmuir (1986) Catts and Langmuir (1986) Charlet (1986) Charlet (1986) Hayes (1987) Hayes (1987) Hayes (1987)

aBased on experimental data of Hohl and Stumm (1976). bBased on experimental data of Stiglich (1976). 'Bidentate surface complexes as described by Eqs. (113) and (114). Inner-sphere surface complexes as described by Eqs. (6) and (12) where Y = Yo. 'Intrinsic surface complexation constants for Na+ and NO; adsorption were also adjusted.

SURFACE COMPLEXATION MODELS

281

oxide was proposed to occur via bidentate complex formation and not via hydrolysis complex formation (Balistrieri and Murray, 1982a). This reaction is written as

+ M2+

S(0)2M + 2H+

(113) Because a neutral surface complex is formed, the intrinsic conditional equilibrium constant for this reaction is equal to the conditional equilibrium constant and is given by S(OH),

KL(int) = ' K & = [S(0)2M][H']2/([S(OH)2][M2'l> (114) Values of these constants are provided in Table X. Adsorption reactions of actinide elements were investigated for uranium adsorption on goethite and amorphous iron oxide (Hsi and Langmuir, 1985), plutonium adsorption on goethite (Sanchez et al., 1985), thorium adsorption on goethite (LaFlamme and Murray, 1987; Hunter et al., 1988) and manganese oxide (Hunter et al., 1988), and neptunium adsorption on amorphous iron oxide (Girvin et al., 1991). In most of these applications of the triple-layer model a large number of surface complexation constants were fit to the adsorption data. Plutonium adsorption on goethite was described using four surface complexes: SO--PuOH3+, SO--Pu(OH):+, SO--Pu(OH)l , and SO---PU(OH)~ (Sanchez et al., 1985). Thorium adsorption was described using five complexes, SO--Th4+, SO--ThOH3+, SO--Th(OH)z+, SO--Th(OH):, and SO--Th(OH)4, for goethite (LaFlamme and Murray, 1987; Hunter et al., 1988) and three surface complexes, SO--Th(OH):+, SO--Th(OH):, and SO--Th(OH)4 for manganese oxide (Hunter et al., 1988). Uranium adsorption on iron oxides as the uranyl species was accurately described using two surface complexes: SO--U020H+ and SO--(U02),(OH)f (Hsi and Langmuir, 1985). Fit of the triple-layer model to plutonium adsorption on amorphous iron oxide as the plutonyl species was excellent using one surface complex: SOH-Np02( OH) (Girvin et al., 1991). Hsi and Langmuir (1985) observed that excellent fits to their data could also be obtained by adding additional uranyl surface complexes or by varying the combination of surface constants. This observation was also made by Catts and Langmuir (1986), who added surface complexes for SO---M(OH);? and SO--MNO: (in addition to SO--M2+ and SO--MOH+) to describe copper and zinc adsorption on manganese oxide. As has been observed previously in surface complexation modeling, good fits can be obtained for various combinations of surface complexes. For this reason it is necessary to limit the surface complexation reactions to a small number of the simplest and most chemically reasonable surface complexes. As the number of adjustable parameters is increased, the quality of the model fit improves. This does not

2 82

SABINE GOLDBERG

necessarily indicate any increased chemical insight and may compromise the representation of chemical reality. The modified triple-layer model was applied by Hayes and Leckie (1987) to describe ionic strength effects on cadmium and lead adsorption on goethite. These authors found that only by using an inner-sphere surface complex could the small ionic strength dependence of the adsorption reactions of these cations be accurately described. These authors assert that the modified triple-layer model can be used to distinguish between inner-sphere and outer-sphere surface complexes. Hayes (1987) used the pressure-jump relaxation technique to investigate lead adsorption/desorption kinetics on goethite The author was able to describe both his adsorption and his kinetic data using the modified triplelayer model. Based on the kinetic results, an inner-sphere surface complex between a lead ion and an adsorbed nitrate ion, SOHPb2+-N0;, was postulated in addition to the inner-sphere surface complex, SOPb+, obtained from equilibrium results. The magnitude of the log Kkb(int) value obtained from kinetics was identical to that obtained from equilibrium data. This result is expected because surface complexation model parameter values from equilibrium experiments are necessary to analyze the kinetic data. Therefore the kinetic approach is not independent. The first application of the triple-layer model to alkaline earth metal adsorption on heterogeneous systems was the study of Charlet (1986; Charlet and Sposito, 1989) on a Brazilian Oxisol soil. These authors found good fits of the triple-layer model to calcium and magnesium adsorption on the Oxisol using inner-sphere surface complexes. However, Charlet and Sposito (1989) suggested that these cations may also form outer-sphere surface complexes.

C. STERNVSC-VSP MODEL The Stern VSC-VSP model has been used to describe adsorption of the metals copper, lead, and zinc on the iron oxide (goethite) surface (Barrow et al., 1981). Application of the model to other metal ions or other oxide surfaces is not available. In the Stern VSC-VSP model values of surface site density, maximum adsorption, equilibrium constants, and capacitances were optimized to fit charge and adsorption data. Table XI presents values of N s , NT, log&, and Ciobtained by computer optimization for metal adsorption on goethite. For copper and zinc the adsorption of the species MOH+ and MC1+ is postulated; for lead the adsorption of Pb2+ and PbC1' is postulated based on goodness-of-fit criteria (Barrow et al., 1981). The ability of the Stern VSC-VSP model to describe zinc adsorption on goethite is indicated in Fig. 22. The model describes the data very well.

283

SURFACE COMPLEXATION MODELS Table XI

Values of Maximum Surface Charge Density, Maximum Adsorption Density, Binding Constants, and Capacitances Obtained with the Stern VSC-VSP Model by Computer Optimization for Metal Adsorption on Goethite"

Parameter

Copperb

Leadb

Maximum surface charge density (pmol m-') Maximum metal adsorption density (pmol m-') Capacitances c,, (F m-'1 Cop (Fm-') c p d (F m-') Binding constants 1% K H log KOH log KNa 1% KCl log KUOHt log KM1+ log K M C P

10.0 6.0

10.0

6.0

6.30 1.82 0.97

5.54

8.0 6.69 -0.7 -0.36 8.61

8.0 6.69 -0.7 -0.36

-

6.60

1.82 0.97

-

7.89 5.60

Zinc'

10.4 7.28 4.8 0.99 0.97 8.02 6.03 -0.96 -0.92 6.45

-

6.01

"From Barrow ef a&.(1981). bBased on experimental data of Forbes er al. (1976). 'Based on experimental data of Bolland er al. (1977).

- -4

6

8

lo

PH Figure 22. Fit of the Stern VSC-VSP model to zinc adsorption on goethite. Model results are represented by solid tines. Model parameter values are provided in Table XI. From Barrow et al. (1981), based on experimental data of Bolland et al. (1977).

2 84

SABINE GOLDBERG

The Stern VSC-VSP model was extended to describe ion adsorption by soil materials (Barrow, 1983) and then further extended to describe the rate of adsorption (Barrow, 1986a). This model was applied to a range of anions but generally was limited to one soil sample (details will be provided in Section V,C). The extended Stern VSC-VSP model has been called a mechanistic model and has been applied to describe zinc adsorption on several soils (Barrow, 1986~).The mechanistic Stern VSC-VSP model contains the following assumptions: (1) individual sites react with adsorbing ions as with sites on variable-charge oxides, (2) a range of sites exists whose summed adsorption behavior can be modeled by a distribution of parameters of the variable-charge model, and (3) the initial adsorption reaction induces a diffusion gradient into the particle interior and begins a solid-state diffusion process. The equations for the mechanistic Stern VSC-VSP model describe the following conditions (Barrow, 1986a): (A) Heterogeneity of the surface:

4 = l/(u/,/A277)

exp[-0.5(qaoj - qa0/a)2]

(115)

where is the probability that a particle has initial potential qaoi, qd is the average of Taoj, and u is the standard deviation of qaOj. (B) Adsorption on each component of the surface: (1) at equilibrium: 6. =

K p y c exp( -2,.Fqaj/RT) 1 + Kiayc exp(-ZiFqaj/RT)

(116)

where 6, is the proportion of the jth component occupied by the ith ion, Ki and Ziare the binding constant and valence for the ith adsorbing ion, qajis the potential of the jth component, (Y is the fraction of adsorbate present as the ith ion, y is the activity coefficient, and c is the total concentration of adsorbate. (2) rate of adsorption: 6.Jt =

KTc(1- 6,) - kz6, (1 - exp[-t(krc kfc + kz

+ kz)]}

(117)

where O, is the increment in 6, over time interval t , and kf = klayexp(&Fqaj/RT)

(118)

(119) kz = k2ayexp(-iiF'.Iaj/RT) where kl and k2 are rate coefficients and & and G are transfer coefficients. (C) Diffusive penetration:

SURFACE COMPLEXATION MODELS

285

where Mi is the amount of material transferred to the interior of the jth component on an area basis, Coj is the surface concentration of the adsorbed ion at time r, c k j is the value of Coj at time t k , b is the coefficient related to the diffusion coefficient via the thickness of the adsorbed layer, and f is the thermodynamic factor. (D) Feedback effects on potential: (1) for a single period of measurement: .-m10, a) a01 (121)

*.=*

where 'Paj is the potential of the jth component after reaction and ml is a parameter. (2) for measurement through time: a] a01. - m 1ej-m~MjINrnj (122)

*.=*

where Nmj is the maximum adsorption on component j and m2 is a parameter. (E) Effects of temperature :

b =Aexp(-E/RT)

(123)

where E is an activation energy and A is a parameter. These equations were incorporated into a computer program. The continuous distribution of Eq. (115) was divided into 30 discrete elements. The 30 sets of equations were solved by an iterative procedure using a computer program and the criterion of goodness of fit to ion sorption (Barrow, 1983). Because of the very large number of adjustable parameters, the use of the mechanistic Stern VSC-VSP model should be regarded as a curve-fitting procedure, although of course fit to the data is usually excellent. The mechanistic Stern VSC-VSP model has been used to describe the effects of time and temperature on zinc sorption on an Australian soil (Barrow, 1986b), the effect of pH on zinc sorption on several soils (Barrow, 1986c), and the point of zero salt effect for zinc sorption on an Australian soil (Barrow and Ellis, 1986b). The model was able to describe the data well in all cases using the assumption that the species ZnOH+ adsorbs on the surface. Serious difficulties in the use of the model result because calcium carbonate was added to raise the pH and calcium nitrate solutions were used as the background electrolyte. Unless calcium can be shown to act solely as an inert background electrolyte, the chemical significance of the parameters obtained in these modeling procedures is expected to be compromised by specific adsorption of calcium. The mechanistic Stern VSC-VSP model has also been used to describe zinc, nickel, and cadmium adsorption by a goethite slightly contaminated

2 86

SABINE GOLDBERG

with silicon (Barrow et al., 1989). As was the case for soil, excellent fits t f J the adsorption data were obtained. Although in these experiments pH w a i adjusted with additions of sodium hydroxide, the background electrolyte was still calcium nitrate.

D. GENERALIZED TWO-LAYER MODEL Application of the generalized two-layer model to metal adsorption has been restricted to the hydrous ferric oxide surface (Dzombak and Morel, 1990). As discuksed in Section II,D, adsorption of metal ions is postulated to occur on two types of sites of high or low affinity. Intrinsic conditional equilibrium constants for the generalized two-layer model for metal adsorption were obtained with the computer program FITEQL (Westall, 1982). Individual values of log K',(int) and best estimates of log Kh(int) were obtained as described for protonation-dissociation constants in Section II1,D. Metal adsorption is defined by reaction Eqs. (51) and (52) for silver, cobalt, nickel, cadmium, zinc, copper, lead, and mercury. For adsorption of alkaline earth cations, the reaction on the strong sites, Eq. (51), is replaced by: SSOH+ M2+

S'OHMZ+

+ Sr2+ + H 2 0

S"0SrOH

(124) For strontium adsorption an additional reaction on the weak sites is defined:

+ 2H+

(125) Adsorption of trivalent chromium metal occurs only on the strong sites by one reaction: S"OH

SsOH + Cr' t H 2 0

* S"OCrOH++ 2H+

(126) Table XI1 presents values of best estimates of logKh(int) obtained by FITEQL computer optimization. The ability of the generalized two-layer model to describe copper adsorption on hydrous ferric oxide is indicated in Fig. 23. The model describes the data very well using both individual data set values and the best estimates of log K',(int). The surface precipitation model for metal ions (see Section IV,A) has been incorporated into the generalized two-layer model (Dzombak and Morel, 1990). Adsorption of metal ions is assumed to occur on both the strong and the weak sites via the reactions

+ = MOH; + H+ + M2++ H 2 0 z= SW(OH),,,, + = MOHZ + H+

r S S O H+ M2+ + HzO ES"OH

S'(OH),,,,

(127) (128)

287

SURFACE COMPLEXATION MODELS Table XI1

Values of Intrinsic Metal Surface Complexation and Surface Precipitation Constants Obtained with the Generalized Two-Layer Model Using FITEQL Computer Optimization for Hydrous Ferric Oxide"

Metal

logKL(int)b

Ca2+ Sr2+ Ba2+

4.97 ? 0.10 5.01 f 0.03 5.46 f 0.12 -1.72 f 0.20 -0.46 f 0.12 0.37 f 0.58 0.47 f 0.03 0.99 f 0.02 2.89 f 0.07 4.65 It 0.14 7.76 f 0.02 2.06 f 0.15

&+

co2+ Ni2+ Cd2+ Zn2+ cu2+ Pb2+ Hg2+ Cr3+

log Kh(int)'

log Kh(int)d

Surface complexationf -5.85 f 0.19 -6.58 & 0.23 -17.60 f 3.06

Number of data sets' 9 12 6 6 13 2 24 21 10 4 12 4

Surface precipitationg

zn2+ Hg2+

log Kg(int)

log Ky(int)

1% KspM

3.49 10.26

0.51 8.95

11.7 3.88

1 11

"From Dzombak and Morel (1990). blogKh(int) is defined for reaction Eq. (124) for alkaline earth metals, for reaction Eq. (126) for chromium, and for reaction Eq. (51) for all other metals. "logK$(int) is defined for reaction Eq. (52). log KL(int) is defined for reaction Eq. (125). 'Experimental data sources provided in Chapter 6 of Dzombak and Morel (1990). r95% confidence intervals. glogKspFe= 2.5, logK$(int)= log Kh(int) + logKspFe, log Kl(int) = logKh(int) + 1% K s p k '

The intrinsic equilibrium constants for the reactions as written are Ki(int) for Eq. (127) and KT(int) for Eq. (128). The reactions for precipitation of M2+ and Fe3+ are as written in Eqs. (111) and (112). The surface precipitation model generally describes the metal sorption data well.

E. ONE-&MODEL Application of the one-pK model to metal adsorption has so far been restricted to cadmium adsorption on the iron oxides, hematite and amorphous iron oxide (van Riemsdijk et al., 1987). Adsorption of cadmium was

SABINE GOLDBERG

288 100

TOTFe = 1.WE -3 M TOTCu = 1.WE -7 M 0 TOTCU 2.WE -7 M A TOTCU= 5.00E -7M

-

a

u 60

a

-

3

c 40

Q,

2

a"

20 0 -

2.5

3.5

4.5

5.5

6.5

PH Figure 23. Fit of the generalized two-layer model to copper adsorption on hydrous ferric oxide. Model results are represented by a solid line for individual logK&,(int) and a dashed line for logK&(int); logK&,(int) = 2.91, logK$,(int) was not necessary; logK&,(int) is provided in Table XII. From Dzornbak and Morel (1990),based on experimental data of Benjamin (1978) and Leckie et al. (1980), reproduced with permission from John Wiley and Sons.

modeled as occurring in the Stern plane. Good model fits to data were obtained by considering only the formation of the hydrolysis surface complex as given in Eq. (66). The adsorption density of the metal surface complex formed by reaction Eq. (65) was negligible (van Riemsdijk et al., 1987). The equilibrium constant expression for the hydrolysis surface complex is provided by Eq. (72). To describe the adsorption of cadmium on iron oxides, values for log K H, log K c + , log KA- , and C were obtained from potentiometric titration data and have already been provided in Table VIII. To describe cadmium adsorption on hematite, no surface complexation constants for the background electrolyte are considered. The ability of the simplified one-pK model to describe cadmium adsorption on hematite is indicated in Fig. 24. The model describes the data quite well. To describe cadmium adsorption on amorphous iron oxide, values of log Kc+ and log KA- were included. The fit was similar in quality to that obtained on hematite (van Riemsdijk et al., 1987). The value of the equilibrium constant, logK& = -6.97, for amorphous iron oxide is very similar to that for hematite, log K d: = -6.41. The one-pK model was expanded to include surface heterogeneity (van Riemsdijk et al., 1987). As for surface charging behavior, the sensitivity of metal adsorption toward the degree of heterogeneity was low. The fit of the model could not be improved significantly by including surface heterogeneity, thus allowing a homogeneous model to be used.

SURFACE COMPLEXATION MODELS

log [ Cd

2 89

2c]

Figure 24. Fit of the simplified one-pK model to cadmium adsorption on amorphous iron oxide. Model results are represented by solid lines; log K&, = -6.41. From van Riemsdijk eral. (1987).

V. APPLICATION OF MODELS T O INORGANIC ANION ADSORPTION REACTIONS ON OMDES, CLAY MINERALS, AND SOILS A. CONSTANT CAPACWANCE MODEL Characterization of anion adsorption behavior as a function of solution pH results in curves termed adsorption envelopes. The constant capacitance model has been used to describe inorganic anion adsorption envelopes on iron oxides (Sigg and Stumm, 1981; Goldberg and Sposito, 1984a; Goldberg, 1985, 1986a,b; Goldberg and Glaubig, 1985), aluminum oxides (Hohl et af., 1980; Goldberg and Sposito, 1984a; Goldberg and Glaubig, 1985, 1988a; Goldberg, 1986a,b; Bleam et af., 1991), clay minerals (Goldberg, 1986b; Goldberg and Glaubig, 1986b, 1988b,c; Motta and Miranda, 1989), and soils (Goldberg and Sposito, 1984b; Goldberg, 1986b; Goldberg and Glaubig, 1986a, 1988b,c; Sposito ef al., 1988). In the application of the constant capacitance model to inorganic anion adsorption, the surface complexation reactions are usually written in terms of undissociated acids (Sigg and Stumm, 1981). The reactions Eqs. (8) and (9) are replaced by the expressions

+ HzO + (i - 1)H+ = SzH(x-j)L(2-i)+ 2Hz0 + ( j - 2)H"

SOH + H,L = SH(x-i)L('-i) 2SOH + H,L

(129) (130)

2 90

SABINE GOLDBERG

where x is the number of protons present in the undissociated form of the acid, 1 5i 5 n , and 2 sj I n , where n is the number of anion surface complexes and is equal to the number of dissociations undergone by the acid. The intrinsic conditional equilibrium constants describing these reactions are ( l - i ) H+ (i-1) KL(int) = [SH(x-i)L I[ exp[(l - i ) F T / R T ] (131) [SOHI[HxLl

Sigg and Stumm (1981) postulated bidentate reactions [Eq. (130)] for phosphate and sulfate adsorption on goethite; Hohl et al. (1980) postulated bidentate species for sulfate adsorption on aluminum oxide. However, Goldberg and Sposito (1984a) obtained good fits to the phosphate adsorption data of Sigg (1979) by considering only monodentate species. All other applications of the constant capacitance model have been restricted to monodentate anion adsorption reactions [Eq. (129)l. The fit of the constant capacitance model to sulfate adsorption was not good (Sigg, 1979; Sigg and Stumm, 1981). Sigg (1979) postulated that this poor fit could have been caused by the presence of a NaSO, ion pair that had not been included in model calculations. An alternative explanation is that sulfate adsorbs via an outer-sphere mechanism and that therefore use of the constant capacitance model is not appropriate. Intrinsic conditional equilibrium constants for anions in the constant capacitance model are obtained using the computer program MICROQL (Westall, 1979) or by computer optimization using the program FITEQL (Westall, 1982). Figure 25 presents the ability of the constant capacitance model to describe silicate adsorption on goethite. The ability of the model to describe the adsorption data is very good. Table XI11 provides values for intrinsic inorganic anion surface complexation constants obtained with the constant capacitance model for various materials. In the work of Goldberg and co-workers (Goldberg and Sposito, 1984a; Goldberg, 1985, 1986a,b; Goldberg and Glaubig, 1985, 1988a) values of log K,(int) were averages obtained from a literature compilation of experimental log K,(int) values. Values of the protonation-dissociation constants, the phosphate surface complexation constants (Goldberg and Sposito, 1984a), and the boron surface complexation constants (Goldberg and Glaubig, 1985) obtained in this fashion were not significantly different statistically for aluminum and iron oxide minerals. Applications of the constant capacitance model to anion adsorption edges on clay minerals have been carried out for boron (Goldberg and

SURFACE COMPLEXATION MODELS

..“

.-

0

A

0 3

-

U

I

4

5

6

7

0

9

1

0

291

1

PH Figure 25. Fit of the constant capacitance model to silicate adsorption on goethite. Model results are represented by solid lines. Model parameters are provided in Table XIII. From Sigg and Stumm (1981).

Glaubig, 1986b), selenium (Goldberg and Glaubig, 1988b), arsenic (Goldberg and Glaubig, 1988c), and molybdenum adsorption (Motta and Miranda, 1989). To describe boron, arsenic, and selenium adsorption on kaolinite and selenium adsorption on montmorillonite, log K,(int) values were based on averages for a literature compilation of aluminum oxides. The assumption was made that adsorption occurs via ligand exchange with aluminol groups on the clay mineral edges (Goldberg and Glaubig, 1986b). To describe boron and arsenic adsorption on montmorillonite and boron adsorption on illite, log K,(int) were optimized with the anion surface complexation constants (Goldberg and Glaubig, 1986b, 1988~).Although the fit to anion adsorption was generally good (see Fig. 26), in some cases the optimized value of log K+(int) was larger than the optimized absolute value for log K-(int) or the optimized value for log K-(int) was insignificantly small. These are chemically unrealistic situations that would potentially reduce the application of the model to a curve-fitting procedure. Additional research is needed. Alternatively, in the application of the constant capacitance model to molybdate adsorption on clays, log K,(int) values were obtained from potentiometric titration data (Motta and Miranda, 1989). Fit of the model to molybdate adsorption data was good, although the zero point of charge values for illite were surprisingly high. The first application of the constant capacitance model to adsorption on heterogeneous soil systems was the study of Goldberg and Sposito (1984b) of phosphate adsorption on 44 soils. These authors used log K,(int) values that were averages obtained from a literature compilation of log K,(int) values for aluminum and iron oxide minerals. The authors calculated a

Table XIII

Values of Intrinsic Inorganic Anion Surface Complexation Constants Obtained with the Constant Capacitance Model Using Computer Optimization" Solid y-AI2O3 Y-A1203b Y-A1203

Y-&o3 h,

UJ

N

6-A1203 &A1203 6-A1203 6-A1203 a-Numinab Hydrous aluminab Activated aluminab a-N(OH), a-AI(OH),' a-AI(OH), a-AI(OH), a-Al(OH), 7-AIOOH Pseudoboehmite Al(OH),(a& Al(oH),(a~n)~ AI(OH),(am)

Ionic medium 0.01 M NaCIO, 0.1 M NaCl 0.1 M NaCl 0.1 M NaCIO, 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.01 M NaCl 0.01 M KCI I=O 0.1 M NaCl NaCl KCI 0.1 M NaCl 0.1 M NaCl 0.001 M KCI 0.1 M NaCl 0.01 M NaCIO, 0.01 M NaCIO, 0.1 M NaCl

log K:(int) 10.34 8.50 9.78 -3.6' 4.14 5.13 5.56 2.87 8.69 7.79 5.09 9.46 11.11 9.01 9.72 9.74 7.28' 5.09 9.89 11.06 5.92

log Kt(int)

-

4.30 2.21 9.4' 3.17 2.89 2.19 5.26 3.98 2.56 3.41 3.64

-

3.32 3.55

-

log K:(int)

Reference

-2.81

Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Hob1 et al. (1980) Goldberg and Glaubig (1988a) Goldberg and Glaubig (1985) Goldberg and Glaubig (1985) Goldberg and Glaubig (1988a) Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Goldberg (1986b) Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Goldberg (1986a) Goldberg (1986b) Bleam ef al. (1991) Goldberg and Glaubig (1985) Goldberg (1986a) Goldberg (1986b) Goldberg and Glaubig (1985)

-3.61 -

-4.78 0.62 -3.75 -3.58

-

-1.05' -4.52

-3.19

$

a-FeOOH a-FeOOHb a-FeOOHb a-FeOOHb a-FeOOHb a-FeOOHb a-FeOOHb a-FeOOHb a-FeOOHb a-FeOOH a-FeOOH a-FeOOHb a-FeOOHb a-FeOOH a-FeOOH a-Fe203b a-Fe,03 Fe(OH)3(am)b Fe(OH)3(am)b Fe(OH),( am)b Fe(OH)3(am)b Fe(OH)3(am)b Fe(OH)3(am) Kaolinitesg Kaolinite Kaolinite Kaolinite

0.1 M NaCIO, 0.1 M NaCIO, 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.1 M NaCIO, 0.1 M NaClO, 0.1 M NaCIO, 0.1 M NaCIO, 0.1 M NaC10, 0.1 M NaCl I=0 0.1 M NaCl 0.01 M NaC10, 0.125 M NaC10, 1 M NaC104 0.1 M NaClO, 0.01 M NaCIO4 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.1 M NaCl 0.01 M NaCl

9.5' 10.54 10.43 10.49 11.22 10.10 10.87 10.02 11.10 -5.8' 4.1 3.82 4.48

-4.8' 5.25 7.43 4.88 11.84 10.78 11.75 10.72 -

5.63 5.28 f 0.2 11.04 11.07 4.95

5.1' 7.25 6.25 6.27 7.06 5.80 6.52 5.36 5.80 - 13.5' -3.3 -4.27 -3.43

-

2.06 5.60 3.86 4.22 4.63 3.39 8.24 3.34 0.95

-1.5' 2.94 0.17 0.17 0.99 -0.63 0.29

-

-4.23

-

-0.65 -3.67 -0.27 -1.63 -4.57 -

-3.21 -

-

Sigg and Stumm (1981) Goldberg and Sposito (1984a) Goldberg (1985) Goldberg (1985) Goldberg (1985) Goldberg (1986a) Goldberg (1986a) Goldberg (1985) Goldberg (1985) Sigg and Stumm (1981) Sigg and Stumm (1981) Goldberg (1985) Goldberg (1985) Sigg and Stumm (1981) Goldberg and Glaubig (1985) Goldberg and Sposito (1984a) Goldberg and Glaubig (1985) Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Goldberg and Sposito (1984a) Goldberg and Glaubig (1985) Goldberg and Glaubig (1986b) Goldberg and Glaubig (1988~) Goldberg and Glaubig (1988b) Motta and Miranda (1989) (continues)

Table XIII (Continued) Solid Montmorillonitesh Montmorillonite Montmorillonite Illites‘ Illite Soils’ Soils* Panoche soil Imperial soil

Ionic medium

log Kt(int)

0.1 M NaCl 0.1 M NaCl 0.01 M NaCl 0.1 M NaCl 0.01 M NaCl

6.37 f 1.3 10.92 5.23 5.39 ? 0.8

0.01 M NaCl 0.05 M NaCl 0.1 M NaCl

8.71 f0.6 5.48 f0.4 7.35‘ 9.94

log KE(int)

Reference

-

Goldberg and Glaubig (1986b) Goldberg and Glaubig (1988b) Motta and Miranda (1989) Goldberg and Glaubig (1986b) Motta and Miranda (1989) Goldberg and Sposito (1984b) Goldberg and Glaubig (1986a) Sposito et al. (1988) Goldberg and Glaubig (1988~)

3.40 2.75 2.41 f 2.3 0.85‘ 3.71

-5.14 f 1.7 -4.78

“The work of Goldberg and co-workers was based on logK+(int) = 7.38, log K-(int) = -9.09 for aluminum oxides and kaolinites, and logK+(int) = 7.31, logK-(int) = -8.80 for iron oxides. All soils work was based on logK+(int) = 7.35, logK-(int) = -8.95 unless indicated otherwise. bExperimental data source provided in the reference. ‘logK;(int) and l o g K $ n t ) are defined for reactions Eqs. (8) and (9), respectively. dThe aqueous solution species AlH,PO:+ and AIHPO: are also included. T h e bidentate constants for the formation of S2HP04and S,PO;, reaction Eq. (9), were also optimized; logK$(int) = 8.5 and logK;(int) =4.5. ’logK;(int) is defined for reaction Eq. (8). gAverage for four kaolinites. log K+(int) = 10.62 f 1.6 and log K-(int) = -10.46 f 1.3 were also optimized. Average for three montmonllonites. ‘log K-(int) = 9.30 f 1.3 and log K-(int) = -10.43 f 0.5 were also optimized. Average for three illites. ’Average for 44 soils. ‘logK+(int) = 9.34 f 0.8 and logK-(int) = -10.64 f 0.9 were also optimized. Average for 14 soils. ‘log K;,(int) = 20.05 defined for reaction Eq. (9) was also optimized.

295

SURFACE COMPLEXATION MODELS

-I

0

Y

-

z E

1.5-

z

I.0-

v

Q)

0

0.5-

a
X

X

7

8

9

10

II

PH Figure 26. Fit of the constant capacitance model to boron adsorption on Morris illite. Model results are represented by a solid line; log K+(int) = 8.49, log K-(int) = -10.16, log KB(int)= 5.04. From Goldberg and Glaubig (1986b).

phosphate-reactive specific surface area and obtained good fits of the constant capacitance model to phosphate adsorption data. A similar approach was used to describe arsenic adsorption on a soil (Goldberg and Glaubig, 1988b) and selenite adsorption on five alluvial soils (Sposito et al., 1988). Sposito et al. (1988) used logK,(int) values from Goldberg and Sposito (1984b) and assumed that two types of sites in soil were selenite reactive. Monodentate surface species are formed on one set of sites and bidentate surface species are formed on another set of sites. Using the intrinsic selenium surface complexation constants obtained for one of the soils, Sposito et al. (1988) were able to predict qualitatively the selenite adsorption envelopes for four other soils. To describe boron adsorption on 14 soils (Goldberg and Glaubig, 1986a) and selenium adsorption on a soil (Goldberg and Glaubig, 1988b), log K,(int) values were optimized with the anion surface complexation constants. As described previously for 2 :1 clay minerals, the optimized value of log K-(int) for some of the soils was insignificantly small. This unrealistic situation reduces the chemical significanceof the model application. Using an average set of intrinsic conditional surface complexation constants, the constant capacitance model predicted boron adsorption on most of the soil samples studied (Goldberg and Glaubig, 1986a). The ability of the constant capacitance model to describe boron adsorption on a soil is presented in Fig. 27. Additional research is needed on the application of the constant capacitance model for describing adsorption on natural materials such as clay minerals and soils.

2 96

SABINE GOLDBERG 08

I

I

I

I

I

I

x 0

I

-

h

07-

-0

0.6

-

-

d

0.5-

-

E

-0

a 04-

5

8

a

0.3 -

-

-

-

01-

-

c 0.2

2

8

3

0 -

I

I

I

I

I

I

Figure 27. Fit of the constant capacitance model to boron adsorption on Altamont soil. Squares represent 0- to 25-cm samples: logK+(int) = 8.72, Iog K-(int) = -8.94, logKB(int) = 5.57. Circles represent 25- to 51-cm samples: logK+(int) = 8.45, logK-(int) = -10.07, log K,(int) = 5.45. Model results are represented by solid lines. From Goldberg and Glaubig (1986a).

The constant capacitance model was used to describe ion adsorption in binary oxide mixtures (Anderson and Benjamin, 1990a,b; see Section IV,A for metal adsorption). In the binary Si-Fe oxide system, the presence of dissolved silicate reduced phosphate and selenite adsorption (Anderson and Benjamin, 1990a). The constant capacitance model qualitatively described this effect on phosphate adsorption but not on selenite adsorption. The surface precipitation model has been incorporated into the constant capacitance model to describe anion retention on oxide minerals (Farley et al., 1985). Reactions of the surface precipitation model for trivalent anion sorption onto a trivalent oxide are as follows: Adsorption of L3- onto S(OH),,,,: =SOH

+ L3- + 3H'

# =SH2L

+ H20

(133)

Precipitation of L ~ - : =SH,L

+ L3-+ S3+e SL,,, + &H2L

(134) The reaction for the precipitation of S3+ has already been defined for metal adsorption by Eq. (112). The equilibrium constants for the reactions

SURFACE COMPLEXATION MODELS

I -1

1

297 1

/--I

1.8 g/liter Gibbsite

-4Y

A 24hr 0 52hr

1

I

-7

-5

-3

-I

log [phosphate] Figure 28. Fit of the constant capacitance model containing the surface precipitation model to a phosphate adsorption isotherm on gibbsite at pH 5.0. Model results are represented by a solid line. rp= ([=AlH2P0,] + [A1PO4,,,])/T0T(=AlOH); log KadrP= 30.8, log KsPp= -16.6, log K,,, = 8.5. From Farley et al. (1985), based on experimental data of van Riemsdijk and Lyklema (1980a, triangles; 1980b, circles).

as written are KadsLfor Eq. (133), l/KspL for Eq. (134), and l/Ksps for Eq. (112). The ability of the surface precipitation model to describe phosphate adsorption on an aluminum oxide is indicated in Fig. 28. The model describes the data very well.

B. TRIPLE-LAYER MODEL The triple-layer model has been used to describe inorganic anion adsorption envelopes on amorphous iron oxide (Davis and Leckie, 1980; Benjamin and Bloom, 1981; Zachara et al., 1987; Hayes et al., 1988; Balistrieri and Chao, 1990), goethite (a-FeOOH) (Balistrieri and Murray, 1981; Hayes et al., 1988; Hawke et al., 1989; Ainsworth et al., 1989; Zhang and Sparks, l989,1990b,c; Goldberg, 1991), magnetite (Fe304)and zirconium oxide (Blesa et al., 1984b), aluminum oxide (Davis and Leckie, 1980; Mikami et al., 1983a,b), manganese oxide (Balistrieri and Chao, 1990), kaolinite (Zachara et al., 1988), and soils (Charlet, 1986; Charlet and Sposito, 1989; Zachara et al., 1989). Intrinsic conditional equilibrium constants for the triple-layer model have been obtained by using the computer programs MINEQL (Westall et al., 1976), MICROQL (Westall, 1979), and HYDRAQL (Papelis et al., 1988) or by computer optimization using the computer program FITEQL (Westall, 1982). In general, in the application of the triple-layer model to anion adsorption, the reactions Eqs. (24) and (25) with their equilibrium constants Eqs. (30) and (31) are considered.

SABINE GOLDBERG

298

5

4

6

7

0

PH Figure 29. Fit of the triple-layer model to selenate adsorption on amorphous iron oxide. Model results are represented by solid lines; log Kieo4(int) = 9.9, log KSco,(int) = 15.9. From Davis and Leckie (1980).

Figure 29 presents the ability of the triple-layer model to describe selenate adsorption on amorphous iron oxide. The model describes the adsorption data very well. Table XIV provides values for intrinsic inorganic anion surface complexation constants obtained with the triple-layer model for various surfaces. Adsorption of the trivalent anions PO:- and AsO$- is described by an additional reaction: SOH + 3H+ + L'- *SOH;

- LH$.-*)-

(135)

Table XIV Values of Intrinsic Inorganic Anion Surface Complexation Constants Obtained with the Triple-Layer Model by Computer Optimization Solid y-AlzO3 YmA12O3 a-AlZO; a-FeOOH a-FeOOH a-FeOOH Fe(OH),(am) Fe(OH),(am) Fe( OH),(am) Fe(OH),(am) Fe(OH),(am)

Anion

Po:Cr0:Cr0:SO:(30:FCr0;(3-0:Cr0;Cr0:SO:-

Ionic medium NaCl NaN0, NaNO, NaCl NaNO, NaCl NaNO, NaNO, NaNO, NaNO, NaN0,

log K[(int)

log Kt(int)

Original triple-layer model 25.8" 31.6' 10.1 16.8 11.3 18.1 9.10 14.40 9.8 19.4 11.85 18.1 10.6 19.3 10.1 11.90 18.00 16.80 14.40 15.9 9.9

Reference Mikami et al. (1983a) Mikami et al. (1983b) Ainsworth e? al. (1989) Balistrieri and Murray (1981) Ainsworth et al. (1989) Hawke et al. (1989) Davis and Leckie (1980) Zachara et al. (1987) Benjamin and Bloom (1981) Benjamin and Bloom (1981) Davis and Leckie (1980)

Table XIV (Continued) Solid

Anion

Ionic medium

log K:(int)

log Kt(int)

Reference

Original triple-layer model NaN0, 11.6 17.3 KCI 9.5 15.15 KCl 9.1 16.0 NaN03 9.9 15.9 NaN0, 11.75 15.60 NaN0, 15.00 19.00 NaN0, 12.80 20.75 NaN0, 19.90 22.00 NaN0, 27.70" 33.50b KCI 27.3 NaN03 10.0 NaCIO, 9.19 17.1 NaCIO, 9.42 16.9 NaC10, 9.42 16.3 NaCIO, 9.48 16.2 NaCIO, 9.49 15.6 NaClO, 9.37 15.9

Zachara et al. (1987) Balistrieri and Chao (1990) Balistrieri and Chao (1990) Davis and Leckie (1980) Benjamin and Bloom (1981) Benjamin and Bloom (1981) Benjamin and Bloom (1981) Benjamin and Bloom (1981) Benjamin and Bloom (1981) Balistrieri and Chao (1990) Davis and Leckie (1980) Zachara er al. (1988) Zachara et al. (1988) Zachara et al. (1988) Zachara et al. (1988) Zachara er al. (1988) Zachara et al. (1988)

Modified triple-layer model KN03 8.7' 9.7f KN03 6.5' 8.3f NaN0, 8.90 15.70 12.94 NaCl NaN03 15.1og 14.10h NaCl 15.4V 20.42' NaCl 29.W 33.40" NaN03 15.4 NaN03 9.60 14.50 NaNO, 14.4Y KC1 25.56 5 0.Sk KCI 18.0' KCI 18.28 5 0.5' KCI 18.7' Li2S04 6 9 -

Blesa et al. (1984a) Blesa et at. (1984a) Hayes et al. (1988) Zhang and Sparks (1990~) Hayes ef al. (1988) Zhang and Sparks (1990~) Hawke et al. (1989) Zhang and Sparks (1990b) Hayes et a!. (1988) Hayes el al. (1988) Balistrieri and Chao (1990) Balistrieri and Chao (1990) Balistrieri and Chao (1990) Balistrieri and Chao (1990) Charlet (1986)

"log Kt(int) is defined by Eq. (31). blog Kt(int) is defined by Eq. (136). 'Based on experimental data of Honeyman (1984). dAlso includes a surface complex SiOH-H,CrO, whose equilibrium constant is a fitting parameter. 'log K,(int) is defined for reaction Eq. (137). Inner-sphere surface complex. KBc+(int) is defined for reaction Eq. (138). Inner-sphere surface complex. gInner-sphere surface complexes as described by Eqs. (129) and (131), where Vr = Toand i = 2. hInner-sphere surface complex as described by Eqs. (39) and (40). 'Inner-sphere surface complexes as described by Eqs. (129) and (131), where T = Vro and i = 1. 'Optimization also included an inner-sphere complex as described by Eqs. (129) and (131) where Vr = Poand i = 3; log K$(int) = 24.00. kInner-sphere bidentate surface complex as defined for reaction Eq. (139). Average of four suspension densities. 'Average of five suspension densities. Inner-sphere surface complex.

3 00

SABINE GOLDBERG

The intrinsic conditional equilibrium constant for this reaction is KL(int) =

[SOH: - LHt-2’-] exp[F(*, - (I - 2)qP)/RT] [SOH][H’I3[ L’-]

(136)

The first extension of the triple-layer model to describe anion adsorption using a ligand-exchange mechanism and forming an inner-sphere surface complex was carried out by Blesa et al. (1984a) for boron adsorption on magnetite and zirconium dioxide. These researchers defined the following surface reaction for boron: SOH + B(OH)3 + OH- F? SOB(0H);

+ H20

(137)

An additional boron surface complex was formed by reaction with the cation from the background electrolyte: SOH + B(OH)3+ OH-

+ C+ * SOB(0H); - C+ + H 2 0

(138)

Blesa et al. (1984a) were well able to describe surface charge density data as a function of pH. It is impossible, however, to evaluate the ability of this approach to describe boron adsorption because the experimental adsorption data were not provided. The modified triple-layer model was applied to describe ionic strength effects on selenate and selenite adsorption on goethite and amorphous iron oxide (Hayes ef al., 1988). These authors could only describe the ionic strength dependence of selenite adsorption using an inner-sphere surface complex and that of selenate using an outer-sphere surface complex. In order to improve the fit of the model, reaction Eq. (39) was added to describe selenite adsorption on goethite. The modified triple-layer model has been successfully used to describe anion adsorption via a ligand-exchange mechanism. Inner-sphere surface complexes were used to describe phosphate (Hawke et al., 1989), molybdate (Zhang and Sparks, 1989), and selenite (Zhang and Sparks, 1990c) adsorption on goethite, and selenite and molybdate adsorption on amorphous iron oxide and manganese oxide (Balistrieri and Chao, 1990). Selenite adsorption on amorphous iron oxide is defined by the reaction 2SOH + Se0:-

+ 2H

S2Se03+ 2H20

(139)

The pressure-jump relaxation technique has been used to investigate anion adsorption on aluminum (Mikami et al., 1983a,b) and iron oxide surfaces (Zhang and Sparks, 1989, 1990b,c). Mikami et al. (1983a) described phosphate adsorption and Mikami er al. (1983b) described chromate adsorption on aluminum oxide. These authors used the triple-layer model and its resultant equilibrium constants to describe their kinetic data and concluded that phosphate and chromate ions adsorb as outer-sphere

SURFACE COMPLEXATION MODELS

301

surface complexes. Various types of experimental evidence support an inner-sphere adsorption mechanism for phosphate on oxides (Goldberg and Sposito, 1985). The modified triple-layer model was used to describe both kinetic and adsorption data for molybdate (Zhang and Sparks, 1989, 1990a), sulfate (Zhang and Sparks, 1990b), and selenium adsorption (Zhang and Sparks, 1990~).Based on the kinetic results, inner-sphere surface complexes were postulated for molybdate and selenite, whereas outer-sphere surface complexes were postulated for sulfate and selenate. The magnitudes of the log KL(int) values obtained from kinetics were similar to those obtained from equilibrium modeling. This result is not unexpected because an equilibrium surface complexation model is required to analyze the kinetic data. Application of the triple layer model to anion adsorption envelopes on the clay mineral kaolinite has been carried out (Zachara et al., 1988). Values used for the protonation-dissociation constants for the aluminol group were obtained by Davis (1977) for amorphous aluminum oxide from the data of Anderson et al. (1976). Values used for logK,(int) for the silanol group were obtained by Riese (1982) for kaolinite using the doubleextrapolation technique. In order to describe chromate adsorption on kaolinite, protonation-dissociation reactions for both aluminol and silanol groups were required. Anion surface complexation reactions for the alumino1 group were defined by Eqs. (24) and (25). For the silanol group, an additional complex, SiOH-H2Cr04, is postulated whose surface complexation constant is considered a fitting parameter (Zachara et al., 1988). To describe sulfate adsorption on kaolinite it was not necessary to invoke any reactions for the silanol group. The fit of the model to sulfate adsorption on kaolinite was good (see Fig. 30). Chromate adsorption on aluminum-substituted goethite, a-(Fe,Al)OOH, was investigated (Ainsworth et al., 1989). The authors considered two sets of reactive sites, partitioned into iron and aluminum sites based on a mole percentage of aluminum substitution. Good description of the adsorption data was obtained with the triple-layer model by considering protonationdissociation reactions for both aluminum and iron sites and chromate surface complexation reactions only on iron sites (Ainsworth et al., 1989). The first application of the triple-layer model to anion adsorption on a complex natural system was the study of sulfate adsorption on a Brazilian Oxisol soil (Charlet, 1986; Charlet and Sposito, 1989). Charlet (1986) found good fits of the triple-layer model to sulfate adsorption on the Oxisol using an inner-sphere surface complex. However, Charlet and Sposito (1989) suggested that both inner-sphere and outer-sphere surface complexes for sulfate are likely to form. Alternatively, the approach used by Ainsworth et al. (1989) for aluminum-substituted goethite was used by Zachara et al. (1989) to describe

3 02

SABINE GOLDBERG

I

"

'

"

'

'

I

-

'0

a

-

$ 20-

-

$ 40I

4

I

5

I

I

6

I

I

7

I

8

PH Figure 30. Fit of the triple-layer model to sulfate adsorption on kaolinite. Model results are represented by solid lines; log K: = 9.49; log K: = 15.6for 0.1 K M SO,; log K,' = 9.37, logK:= 15.9 for 0.5 &f SO4. After Zachara ef al. (1988).

Fipre 31. Fit of the triple-layer model to chromate adsorption on two soils. Model results are represented by a solid line (HC soil) and a dashed line (CP soil); logK!-,04(int) = 9.8, logK,&,(int) = 19.4.After Zachara ei al. (1989).

chromate adsorption on two soils. Aluminum-substituted goethite was assumed to be the chromate-reactive mineral in these soils. As for a(Fe,Al)OOH, protonation-dissociation reactions were considered for both aluminum and iron sites but chromate adsorption was restricted to iron sites. The surface site density was estimated based on clay content and surface area measurements (Zachara et al., 1989). Figure 31 presents the

SURFACE COMPLEXATION MODELS

303

ability of the triple-layer model to describe chromate adsorption on two soils. The fit is qualitatively correct.

C.STERNVSC-VSP MODEL The Stern VSC-VSP model has been successfully used to describe adsorption of the inorganic anions phosphate, selenite (Bowden et al., 1980; Barrow et al., 1980b), molybdate (McKenzie, 1983), and borate (Bloesch et al., 1987) on the iron oxide, goethite. Application of the model to describe molybdate adsorption on manganese oxide was unsuccessful (McKenzie, 1983). The model failed to describe the lower pH of maximum molybdate adsorption on manganese oxide. McKenzie (1983) attributed this failure to lack of knowledge of adsorbing molybdenum polymers. As for metal adsorption, in the Stern VSC-VSP model, values of surface site density, maximum adsorption, binding constants, and capacitances are optimized to fit charge and/or adsorption data. Table XV presents values of Ns , N T , log Ki, and Ciobtained by computer optimization for anion adsorption on goethite. For phosphate and selenite, adsorption of the divalent species HPOZ- and Se0:- is postulated (Bowden et al., 1980; Barrow et al., 1980b). For molybdate, adsorption of both HMo04 and MOO:- is considered (McKenzie, 1983). Four adsorbing species are considered for boron: B(0H); , B303(OH);, B405(OH)f, and B506(OH)4 (Bloesch et al., 1987). The ability of the Stern VSC-VSP model to describe boron adsorption on goethite is indicated in Fig. 32. The model describes the data well, as would be expected given the large number of adjustable parameters. The Stern VSC-VSP model was also able to describe ionic strength effects on phosphate adsorption by goethite (Barrow et al., 1980b). As discussed in Section IV,C, the Stern VSC-VSP model was extended to describe phosphate adsorption by soil (Barrow, 1983). This model was used to describe the effect of pH on phosphate sorption by various soils (Barrow, 1984) and was further extended to describe the rate of sorption (Barrow, 1986a). The equations and assumptions of this extended mechanistic Stern VSC-VSP model are provided in Section IV,C. As discussed in detail in Section IV,C, use of the mechanistic Stern VSC-VSP model should be regarded as a curve-fitting procedure. The mechanistic Stern VSC-VSP model has been used to describe the effects of time and temperature on fluoride, molybdate (Barrow, 1986b), selenite, and selenate adsorption on an Australian soil (Barrow and Whelan, 1989b), the effect of pH on fluoride (Barrow and Ellis, 1986a), phosphate (Barrow, 1986c), selenite, selenate (Barrow and Whelan, 1989a),

Table XV Values of Maximum Surface Charge Density, Maximum Adsorption Density, Binding Constants, and Capacitances Obtained with the Stem VSC-VSP Model by Computer Optimization for Anion Adsorption on Goethite ~~

~

Parameter

Phosphate"

Selenite"

Molybdateb

10.0

10.0

10.0

~~

Borate' 10.0

~

Citrate" 10.0

2.43

2.71

6.0

4.34

1.74

3.03 1.93 -d

4.06 1.93 -d

3.03 1.93

2.92 0.914 0.965

3.36 1.12

7.0 4.0 0 -1.2 0 6.86 0 -

8.3 5.0 0.08 -1.7 0 6.95

-

-

-d

6.0 4.6 0 -1.2

4.8

-

-

-

7.0 4.0 0 -1.2 0 0 4.79

1.21 4.69 1.12

-

7.94 3.98 -0.0862 -0.183 2.74

5.5

-d

"From Bowden et al. (1980). From McKenzie (1983). 'From Bloesch et al. (1987). dThis value was made very large so that the p and d planes virtually coincide.

I

I

5

6

7

8

9

10

II

PH Figure 32. Fit of the Stern VSC-VSP model to boron adsorption on goethite. Model result are represented by solid lines. Model parameters are provided in Table XV. From Bloesch et al. (1987).

-

3 05

SURFACE COMPLEXATION MODELS

and borate sorption on soils (Barrow, 1989b), the effect of ionic strength on phosphate, sulfate (Bolan et al., 1986), selenite, selenate (Barrow and Whelan, 1989a), and borate sorption on soils (Barrow, 1989b), and the point of zero salt effect for phosphate sorption on an Australian soil (Barrow and Ellis, 1986b). The model was able to describe the data well in all cases using one or two anion binding constants. As discussed previously in Section IV,C, difficulties in the use of the mechanistic Stern VSC-VSP model may arise from specific calcium adsorption because calcium carbonate was added to raise the pH. The mechanistic Stern VSC-VSP model has also been used to describe the effect of time on phosphate adsorption by iron and aluminum oxides (Bolan et al., 1985).

D. GENERALIZED TWO-LAYER MODEL Application of the generalized two-layer model to anion adsorption has been restricted to inorganic anions on the hydrous ferric oxide surface (Dzombak and Morel, 1990). As for metal adsorption, intrinsic conditional equilibrium constants for the generalized two-layer model for anion adsorption were obtained using the computer program FITEQL (Westall, 1982). Individual values of log Ki(int) and best estimates of log Ki(int) were obtained as described in Section II1,D for logK,(int). Anion adsorption is defined by the reactions given in Eqs. (53) and (54). For some anions the following anion surface complexation reactions are also considered: SOH + L'- +3H+ SOH + L'-

$ SH,L('-3)-

S0HL'-

+H20

(140) (141)

The intrinsic conditional equilibrium constants for these reactions are

K L(int) =

[S0HL'-] exp[ -lF?P/R TI [SOH][L'-]

(143)

For arsenite (H3As03) and borate (H3BO3) adsorption, surface complexation is defined by the reaction, Eq. (129), where i = 1. Table XVI presents values of best estimates of log Ki(int) obtained by computer optimization with FITEQL. Figure 33 depicts the ability of the generalized two-layer

3 06

SABINE GOLDBERG Table XVI

Values of Intrinsic Anion Surface Complexation Constants Obtained with the Generalized Two-Layer Model Using FITEQL Computer Optimization for Hydrous Ferric Oxide"

Anion

log K[(int)

logKt(int)

log KS(int)

*

17.72 0.51

-

31.29 f 2.88 29.31 k 1.02

7.78 7.73 2 0.08 12.69 10.85 0.07

-

-

-

* 5.41 * 0.15' 0.62 * 0.66'

Number of data setsb

logK;(int)

-

-

9 6 12 1 7 1 14 2 6 2

10.58 f 0.57 13.57 0.06 0.79 0.80 f 0.10 5.17

*

-

0.49 f 0.09

-

"From Dzombak and Morel (1990). 95% confidence intervals. bExperimental data sources provided in Chapter 7 of Dzombak and Morel (1990). 'LogK:(int) is defined for reaction Eq. (131), where i = 1.

TOTFe = 5.00E- 4 M

-

>

-

20 -

-7

n

I

1

8

9

I

6

II

10

12

I

I3

14

PH Figure 33. Fit of the generalized two-layer model to vanadate adsorption on hydrous ferric oxide. Model results are represented by a solid line for individual logK$(int) and a dashed line for logK$(int); logK$(int) = 13.34; logK$(int) is provided in Table XVI. From Dzombak and Morel (1990), based on experimental data of Leckie et ul. (1984), reproduced with permission from John Wiley and Sons.

307

SURFACE COMPLEXATION MODELS

model to describe vanadate adsorption on hydrous ferric oxide. The model describes the data well using both the individual data set value and the best estimate of log K4,(int). The incorporation of the surface precipitation model into the generalized two-layer model to describe anion sorption has been carried out and is described by Dzombak and Morel (1990). However, no applications to actual anion sorption data are available.

E. Om-& MODEL Application of the one-pK model to anion adsorption has so far been restricted to phosphate adsorption on the iron oxide, goethite (van Riemsdijk and van der Zee, 1991). Adsorption of phosphate is modeled as occurring at the d-plane and described by reaction Eq. (67) and equilibrium constant Eq. (73). The adsorption of potassium in the d-plane is also considered: reaction Eq. (68) and equilibrium constant Eq. (74). The equilibrium constant for potassium adsorption is optimized both with the phosphate adsorption equilibrium constant and for the charging data in the absence of phosphate. The ability of the one-pK model to describe phosphate adsorption on goethite is indicated in Fig. 34. The fit of the model is excellent over the entire, wide pH range investigated.

8 9 I0

0' 0

1

I

0.2

I

I

0.4

I

0.6

Solution P (rnrnol liter-')

Figure 34. Fit of the one-pK model to phosphate adsorption on goethite. Model results are represented by solid lines; log KHpo4= 6.65, log KK = -0.5. From van Riemsdijk and van der Zee (1991), based on experimental data of Bowden er af. (1980), reproduced with permission from Kluwer Academic Publishers.

308

SABINE GOLDBERG

M.APPLICATION OF MODELS TO ORGANIC LIGAND ADSORPTION REACTIONS ON OXIDES A. CONSTANT CAPACJTANCE MODEL The constant capacitance model has been used to describe organic ligand adsorption envelopes on aluminum oxide (Kummert and Stumm, 1980), silicon oxide, titanium oxide (Gisler, 1980), iron oxide (goethite) (Sigg and Stumm, 1981), and natural organic matter (Baccini et a / . , 1982). In the application of the constant capacitance model to organic anion adsorption, the surface complexation reactions are written in terms of undissociated acids as for inorganic anions (Kummert and Stumm, 1980; Sigg and Stumm, 1981). However, only monodentate surface complexes are considered as described by reaction Eq. (129) and intrinsic conditional equilibrium constant Eq. (131), where i = 1 or 2. In the application of the model to adsorption of amino acids on oxides, the following reactions leading to bidentate surface complexes are defined (Gisler, 1980): SOH + SOH,'

+ HL' e (SOH)HL(SOH;)

+ SOH: + HL' SO- + SOH + HL'

SO-

e (SO-)HL(SOH:)

(144) (145)

(SO-)HL(SOH)

(146) where HL' represents an amino acid, H,N+-CHR-COO-. The intrinsic conditional equilibrium constants are equal to the conditional equilibrium constants because the surface charge remains unchanged and are as follows: S

[(SOH)HL(SOH;)] K' = [SOH][SOH,+][HL']

(147)

[(SO-)HL(SOH:)] - [SO-][SOH:][HL']

K2 -

[(SO-)HL(SOH)] K' = [SO-][SOH][HL']

(149)

For adsorption on titanium oxide, all three of the above reactions are considered. For adsorption on silicon oxide, only reaction Eq. (146) is considered because no positive SOH: surface groups are found in the experimental pH range (Gisler, 1980). Figure 35 presents the ability of the constant capacitance model to describe phthalate adsorption on aluminum oxide. The model describes the data well at various total organic acid

SURFACE COMPLEXATION MODELS

3 09

Figure 35. Fit of the constant capacitance model to phthalate adsorption on aluminum oxide. Model results are represented by solid lines. Model parameters are provided in Table XVII.After Kummert and Stumm (1980).

concentrations. Table XVII provides values for intrinsic organic ligand surface complexation constants obtained with the constant capacitance model for oxide minerals.

B. TRIPLE-LAYER MODEL The triple-layer model has been used to describe organic ligand adsorption envelopes on the iron oxide, goethite (Balistrieri and Murray, 1987) and on amorphous iron oxide (Davis and Leckie, 1979). In the application of the triple-layer model to organic anion adsorption on goethite, the reactions Eqs. (24) and (25) and the equilibrium constants Eqs. (30) and (31) are considered. For adsorption of glutamate on amorphous iron oxide, reaction Eq. (24) is replaced by the formation of a neutral surface complex, SOH-H2L (Davis and Leckie, 1979). Figure 36 presents the ability of the triple-layer model to describe glutamate adsorption on amorphous iron oxide. The model describes the data well for three different total organic ligand concentrations. Unfortunately, values for the intrinsic surface complexation constants were not provided by the authors. Table XVIII provides values for intrinsic organic ligand surface complexation constants obtained with the triple-layer model for iron oxides.

Table X V n

Values of Intrinsic Organic Ligand Surface Complexation Constants Obtained with the Constant Capacitance Model

Solid

Amino acids Si02(am) Si02(am) Si02(am) Si02(am) TiOz, rutile TiOz, rutile

Ligand

Ionic medium

Benzoate Catechol Phthalate Salicylate Acetate

0.1 M NaC10, 0.1 M NaClO, 0.1 M NaClO, 0.1 M NaClO, 0.1 M NaClO,

Glycine a-Alanine P- Alanine y-Aminobutyric acid Glycine Glycine

1.0 M NaClO, 1.O M NaClO, 1.0 M NaClO, 1.0 M NaCIO, 1.0 M NaC10, 1.0 M NaClO,

log K:(int) 3.1

3.7 1.3

6.0 2.9

3.0 3.3

log Kt(int)

log K:(int)

<-5 2.4 -0.6 -

4.9

5.5

Reference Kummert and Stumm (1980) Kummert and Stumm (1980) Kummert and Stumm (1980) Kummert and Stumm (1980) Sigg and Stumm (1981)

0.04 -0.14 -0.26 -0.68 2.9 3.3

Gisler (1980) Gisler (1980) Gisler (1980) Gisler (1980) Gisler (1980) Gisler (1980)

311

SURFACE COMPLEXATION MODELS

PH Figure 36. Fit of the triple-layer model to glutamate adsorption on amorphous iron oxide. Model results are represented by solid lines. Model parameters were not provided by the authors. From Davis and Leckie (1979), reproduced with permission from the American Chemical Society. Table XVIII Values of Intrinsic Organic Ligand Surface Complexation Constants Obtained with the Triple-LayerModel

Solid

Ligand

Ionic medium

log KL(int)

a-FeOOH a-FeOOH a-FeOOH u-FeOOH

Oxalate Phthalate Salicylate Lactate

NaCl NaCl NaCl NaCl

10.8 9.7 9.2

Y

0

b

I

200

1

log K:(int)

Reference

15.5 15.7 22.9

Balistrieri Balistrieri Balistrieri Balistrieri

-

c

I

400

Citrate concentration (

600

and and and and

Murray Murray Murray Murray

(1987) (1987) (1987) (1987)

a

800

p o l liter ')

Figure 37. Fit of the Stern VSC-VSP model to citrate adsorption on goethite. Model results are represented by solid lines. Model parameters are provided in Table XV. From Bowden et al. (1980).

3 12

SABINE GOLDBERG

C. STERN VSC-VSP MODEL Application of the Stern VSC-VSP model to organic anion adsorption has been restricted to citrate adsorption on goethite (Bowden et al., 1980). As for inorganic anion adsorption, values of surface site density, maximum adsorption, binding constants, and capacitances are optimized to fit charge and adsorption data. Table XV presents parameter values obtained by computer optimization for citrate adsorption on goethite. The adsorption of the trivalent ion L3- is postulated. The ability of the Stern VSC-VSP model to describe citrate adsorption is good and is presented in Fig. 37.

VII. APPLICATION OF MODELS T O COMPETITIVE ADSORPTION REACTIONS ON OXIDES A. METAL-METAL COMPETITION 1. Constant Capacitance Model

The effects of metal-metal competition on the description of adsorption with the constant capacitance model have been investigated only preliminarily. The constant capacitance model containing the surface precipitation model was used to describe the effect of copper on cadmium adsorption by amorphous iron hydroxide (Farley et al., 1985). These authors overpredicted the competitive effect for the data of Benjamin (1978) obtained after 4 hr of reaction time. Using one competitive data point after 30 hr of reaction time, the authors concluded that the surface precipitation model is capable of predicting competitive adsorption of metal ions if slow kinetics are considered. Additional research is needed to substantiate the conclusion of Farley et al. (1985). 2. Triple-Layer Model

The competitive adsorption of alkaline earth cations (Balistrieri and Murray, 1981) and trace metal cations has been investigated on goethite (Balistrieri and Murray, 1982b), manganese oxide (Catts and Langmuir, 1986), and amorphous iron oxide (Cowan etal., 1991) using the triple-layer model. Using intrinsic surface complexation constants from single-cation systems, Balistrieri and Murray (1981) were able to quantitatively predict calcium and magnesium adsorption on goethite in a synthetic major-ion seawater solution. Using this same approach, Balistrieri and Murray (1982b) were able to predict decreases in lead, zinc, and cadmium adsorp-

SURFACE COMPLEXATION MODELS

313

tion on goethite in the presence of magnesium in major-ion seawater. The competitive effect of magnesium was quantitatively described for zinc adsorption over the entire pH range and for lead adsorption above pH 5 . Deviations from the experimental data occurred for lead adsorption below pH 5 and for cadmium adsorption over the entire pH range (Balistrieri and Murray, 1982b). Competitive adsorption of the trace metal ions copper, lead, and zinc on manganese oxide from a solution containing all three ions was predicted from single-ion systems (Catts and Langmuir, 1986). The adsorption of lead was predicted quantitatively, whereas the description of copper and zinc adsorption was qualitatively correct (see Fig. 38). In order to describe cadmium and calcium adsorption on amorphous iron oxides in single-ion systems, as well as to predict competitive adsorption, Cowan et al. (1991) hypothesized inner-sphere surface complexes for cadmium and a combination of inner- and outer-sphere surface complexes for calcium. This study represents the first time that both inner- and outer-sphere complexes have been postulated for a single adsorbing ion. Cowan er al. (1991) were able to describe competitive adsorption of cadmium in the presence of calcium, qualitatively. However, because a better fit was obtained using a nonelectrostatic model with fewer adjustable parameters, these authors suggested that competitive adsorption of cadmium and calcium on amorphous iron oxide is due to a mass-action effect.

Figure 38. Prediction of competitive trace metal adsorption from single-ion systems on manganese oxide using the triple-layer model. Model results are represented by solid lines. Model parameters are provided in Table X. From Catts and Langmuir (1986).

3 14

SABINE GOLDBERG

B. ANION-ANION COMPJXITION 1. Constant Capacitance Model

The ability of the constant capacitance model to predict competitive anion adsorption on goethite from solutions containing phosphate and selenite or phosphate and silicate using the intrinsic surface complexation constants obtained from single-anion systems has been tested (Goldberg , 1985). The model predicted anion competition qualitatively, reproducing the shapes of the adsorption curves; however, phosphate adsorption was overestimated and adsorption of the competing anion was underestimated. Competitive anion adsorption of phosphate and arsenate on gibbsite (Goldberg, 1986a) and goethite (Goldberg , 1986b),phosphate and selenite on goethite, and phosphate and silicate on gibbsite and goethite (Goldberg and Traina, 1987) could be described by direct optimization of the mixedligand data. The fit of the constant capacitance model to the adsorption data using the mixed-ligand approach was much better than that obtained by prediction from single-anion systems (see Fig. 39). In the mixed-ligand approach, the intrinsic surface complexation constants are conditional constants, dependent upon the surface composition of the oxide surface. Goldberg and Traina (1987) suggested that the composition dependence is due to heterogeneity of oxide surface sites. Goldberg (1986a) found that arsenate and phosphate surface complexation constants obtained using the

E

PH Figure 39. Prediction of competitive anion adsorption from single-ion systems and fit of the mixed-ligand approach on gibbsite using the constant capacitance model. Model results are represented by dashed lines for single-ion fits and solid lines for mixed-ligand fits. Singleion model parameters are provided in Table XIII. Mixed-ligand parameters: log Kh = 9.43, logK$ = 3.01, log K$ = -2.26, log K& = 3.53, logK& = -4.62. From Goldberg and Traina (1987).

SURFACE COMPLEXATION MODELS

315

mixed-ligand approach for one ternary gibbsite system could be used to predict competitive adsorption for other ternary gibbsite systems containing different amounts of arsenate and phosphate in solution.

2. Triple-Layer Model The application of the triple-layer model to describe competitive adsorption of anions has been investigated on goethite (Balistrieri and Murray, 1987; Hawke et al., 1989), amorphous iron oxide (Zachara et al., 1987; Balistrieri and Chao, 1990), manganese oxide (Balistrieri and Chao, 1990), and soils (Zachara et al., 1989). Using intrinsic surface complexation constants from single-anion systems, Balistrieri and Murray (1987) were able to predict oxalate adsorption but grossly underestimated salicylate adsorption in the presence of sulfate anions on goethite. Using this same approach, Hawke et al. (1989) were able to predict phosphate adsorption on goethite in the presence of sulfate; phosphate adsorption in the presence of fluoride was not well predicted. Chromate adsorption on amorphous iron oxide in the presence of the competing anions carbonate and sulfate was qualitatively predicted using anion surface complexation constants from single-anion systems (Zachara et al., 1987). Balistrieri and Chao (1990) were unable to predict selenite adsorption on amorphous iron oxide or manganese oxide in the presence of phosphate, silicate, or molybdate from single-anion systems. These researchers were able to describe selenite adsorption in the presence of molybdate by decreasing the magnitude of the selenite equilibrium constants with increasing molybdate concentration. The first application of a surface complexation model to describe competitive adsorption on soil was carried out by Zachara et al. (1989). These researchers used the triple-layer model to describe competitive adsorption of chromate in the presence of sulfate and dissolved inorganic carbon. As described in more detail in Section V,B, aluminum-substituted goethite was assumed to be the reactive mineral surface in the soil. Using the surface complexation constants obtained for amorphous iron oxide by Zachara et al. (1987), Zachara et al. (1989) were able to predict qualitatively chromate and sulfate adsorption on a soil from a solution containing both anions. 3. Stern VSC-VSP Model The mechanistic Stern VSC-VSP model, discussed in detail in Section IV,C, has been applied to competitive anion adsorption on soils (Barrow, 1989a). The author described phosphate-arsenate, phosphate-molybdate,

3 16

SABINE GOLDBERG

and arsenate-molybdate competition using the same parameters as for single-anion systems for two of the three soils studied. The mechanistic Stern VSC-VSP model described the competitive adsorption data very well. However, as discussed in Section IV,C, use of the mechanistic Stern VSC-VSP model should be regarded as a curve-fitting procedure.

C. METAL-LIGAND INTERACTIONS 1. Constant Capacitance Model Ternary surface complexes are formed when metal-ligand complexes, ML, react with the mineral surface. A type A ternary surface complex, S-0-M-L, is formed when attachment of the solution complex to the surface occurs through the metal ion (Schindler, 1990). The constant capacitance model has been used to describe adsorption of metal ligand complexes to form type A ternary surface complexes on silicon oxide (Bourg and Schindler, 1978; Gisler, 1980; Schindler, 1990) and on titanium oxide (Gisler, 1980). In the application of the constant capacitance model to metal-ligand complex adsorption, the following reactions are defined (Gisler, 1980): SOH + ML("-l)+ 2SOH + ML("-')+

+ H' & (SO)zML'"-3) + 2H+ SOML("-')

(150) (151)

The conditional equilibrium constants for these reactions are

K h L = [soML(~-~)][H+] [SOH][ML(m-l)+] [(SO)2ML(m-3)][H']2 K & =~ [SOH]2[ML("-1)+1

(153)

Table XIX provides values for type A ternary surface complexation constants obtained with the constant capacitance model. The fit of the model to the data was good over most of the pH range. 2. Triple-Layer Model

A type B ternary complex, S-L-M, is formed when the solution complex, ML, attaches to the mineral surface through the ligand (Schindler,

317

SURFACE COMPLEXATION MODELS Table XIX

Values of Type A Ternary Surface Complexation Constants Obtained with the Constant Capacitance Model

Metal

Ligand

Ionic medium

log K t

Reference

SiO,

cu2+

Ethylenediamine

1 M NaCIO,

-12.57

SiO, SiO, SiO, SiO,

Glycine a-Alanine P- Alanine -y- Aminobutyric acid cu2+ Glycine Mg2+ Glycine co2+ Glycine

Solid

Mg2+ Mg2+ Mgz+ Mg2+

SiO, TiO,, rutile Ti02, rutile

1 M NaC10, 1 M NaC10, 1 M NaCIO, 1 M NaCIOd

-8.24 -8.23 -8.23 -8.23

-17.21 -17.21 -17.21 -17.21

Bourg and Schindler (1978) Gisler (1980) Gisler (1980) Gisler (1980) Gisler (1980)

1 M KN03 1 M NaC10, 1 M NaCIO,

-5.64 -5.94 -4.40

-12.29 -13.62 -10.94

Schindler (1990) Gisler (1980) Gisler (1980)

1990). The triple-layer model has been used to describe adsorption of a silver-thiosulfate complex on amorphous iron oxide (Davis and Leckie, 1979). In the application of the triple-layer model to ligand-metal complex adsorption, the following reaction is defined: SOH + H+ + M"+

+ L'-

SOH: - LM('-"'-

(154)

The intrinsic conditional equilibrium constant for this reaction is KML(int)=

[SOH; - LM(lPm)-] exp[ F ( q o - (1 - rn)qp)/RT] (155) [SOH][H+][M"+] [LIP]

Figure 40 presents the ability of the triple-layer model to describe silver adsorption on amorphous iron oxide by postulating the formation of a type B ternary complex. The fit of the model to the data is qualitatively correct. Adsorption reactions of actinide elements on oxides have been described in the presence of carbonate and bicarbonate ion using the triple-layer model (Hsi and Langmuir, 1985; Sanchez et al., 1985). These authors postulated the existence and adsorption of actinide-carbonate ion pairs to describe their adsorption data for uranyl (Hsi and Langmuir, 1985) and plutonium adsorption on iron oxides (Sanchez et al., 1985). Carbonate adsorption was not verified in these experiments.

318

SABINE GOLDBERG

4 '

5

6

7

8

9

10

PH Figure 40. Fit of the triple-layer model to silver adsorption on amorphous iron oxide using a type B ternary silver-thiosulfate surface complex. Model results are represented by -(int) = -19.5. From Davis and Leckie (1979), reproduced with persolid lines; log KA8S203 mission from the American Chemical Society.

Metal-ligand interactions can occur without the formation of ternary surface complexes. The ability of the triple-layer model to predict competitive adsorption of cations and anions has been investigated on goethite (Balistrieri and Murray, 1981, 1982b) and amorphous iron oxide (Benjamin and Bloom, 1981; Zachara et d.,1987). Using intrinsic surface complexation constants from single-ion systems, Balistrieri and Murray (1981) were able to quantitatively predict calcium, magnesium, and sulfate adsorption on goethite from a synthetic seawater solution containing these ions. Using this same approach, Balistrieri and Murray (1982b) were able to predict the changes in lead, zinc, and cadmium adsorption on goethite from the presence of sulfate in major-ion seawater. Benjamin and Bloom (1981) were unable to predict competitive adsorption of the metals cadmium, cobalt, and zinc in the presence of the anions selenate, selenite, arsenate, arsenite, chromate, and phosphate on amorphous iron oxide from single-ion systems. The enhanced metal adsorption in the presence of anions was grossly overpredicted by the triple-layer model. Benjamin and Bloom (1981) suggested that either a new surface phase forms or that cations and anions bind to separate sets of sites. An alternative explanation may be that the outer-sphere adsorption mechanism is inappropriate to describe the adsorption behavior of some or all of these ions. Competitive adsorption of chromate on amorphous iron oxide in the presence of calcium and magnesium could be predicted using intrinsic surface complexation constants from single-ion systems (Zachara et al., 1987).

SURFACE COMPLEXATION MODELS

3 19

VIII. INCORPORATION OF SURFACE COMPLEXATION MODELS INTO COMPUTER CODES A. INCORPORATION INTO CHEMICAL SPECIATION MODELS The surface complexation models have been incorporated into various chemical speciation models. The first addition of a surface complexation model to a chemical speciation model was when Davis et al. (1978) added the triple-layer model to the computer program MINEQL (Westall et al., 1976). The computer program MINTEQ (Felmy et al., 1984) combines the mathematical framework of MINEQL (Westall et al., 1976) with the thermodynamic database of WATEQ3 (Ball et al., 1981). This program contains the surface complexation modeling approach. The constant capacitance model has been added to the computer speciation program GEOCHEM (Sposito and Mattigod, 1980) in the development of its successor, the chemical speciation program SOILCHEM (Sposito and Coves, 1988). The computer program HYDRAQL (Papelis et al., 1988) was developed from the computer program MINEQL (Westall et al., 1976) and contains the constant capacitance model, the diffuse layer model, the Stern model, the triple-layer model, and the Stern VSC-VSP model. The computer programs MICROQL (Westall, 1979) and FITEQL (Westall, 1982) do not contain thermodynamic data files. Instead, the equilibrium constants are entered by the user and specified for the problem under investigation.

B. INCORPORATION INTO TRANSPORT MODELS Surface complexation models have not yet been widely incorporated into transport models (Mangold and Tsang, 1991). Jennings et al. (1982) incorporated the constant capacitance model into a transport model and simulated competitive metal sorption at constant pH. Cederberg et al. (1985) have linked the program MICROQL (Westall, 1979) containing the constant capacitance model with the transport program ISOQUAD (G. F. Pinder, unpublished manuscript 1976) to produce the computer program TRANQL. This program was used to simulate cadmium transport in a one-dimensional laboratory column. The STEADYQL computer program (Furrer et at., 1989, 1990) is based on the MICROQL program (Westall, 1979) and calculates chemical speciation of a flow-through system

320

SABINE GOLDBERG

at steady state considering both fast, reversible processes described in terms of chemical equilibrium and slow processes described by kinetic equations. Although this program considers inflow and outflow for one box, the approach can form the basis for a transport model (J. C. Westall, personal communication 1991). The computer program FASTCHEM linking MINTEQ and transport modeling has been developed (Krupka et al., 1988). However, the adsorption model in the FASTCHEM program is nonelectrostatic. The computer program HYDROGEOCHEM links a MINEQL version containing surface complexation models with transport modeling (Yeh and Tripathi, 1990).

Ix. SUMMARY Surface complexation models provide a molecular description of adsorption phenomena using an equilibrium approach. Five such models, the constant capacitance model (Stumm et al., 1980), the triple-layer model (Davis et al., 1978), the Stern VSC-VSP model (Bowden et al., 1980), the generalized two-layer model (Dzombak and Morel, 1990), and the one-pK model (van Riemsdijk et al., 1986) were discussed. Unlike empirical models, surface complexation models define surface species, chemical reactions, equilibrium constant expressions, surface activity coefficients, mass and charge balance, and consider the charge of both the adsorbate and the adsorbent. Common model characteristics are surface charge balance, electrostatic potential terms, equilibrium constants, capacitances, and surface charge density. Each surface complexation model was discussed in detail and its application to protonation-dissociation reactions, metal ion, inorganic anion, and organic ligand adsorption reactions on soil minerals and soils was described. Extensive tables of surface equilibrium constants have been provided. It must be emphasized that because all five surface complexation models contain different basic assumptions for the mineralsolution interface, the chemical species defined and the equilibrium constants obtained with one model must never be used in any other model. Surface complexation models have been incorporated into chemical speciation programs and into transport programs. Application of these models to describe surface reactions on heterogeneous surfaces such as clay minerals and soils requires various simplifying approximations. Additional research is needed to develop consistent protocols for the use of surface complexation models in describing reactions of natural chemical systems such as soils.

SURFACE COMPLEXATION MODELS

32 1

ACKNOWLEDGMENTS Gratitude is expressed to Dr. G. Sposito and Dr. R. van Genuchten for peer reviews, Mr. H. Forster for library assistance, Ms. P. Speckman for typing, Ms. G. O’Donnell for drafting, and Ms. P. Howard for photography.

REFERENCES Ainsworth, C. C., Girvin, D. C., Zachara, J. M . , and Smith, S. C. (1989). Chromate adsorption on goethite: Effects of aluminum substitution. Soil Sci. SOC.A m . J . 53, 411-418. Anderson, M. A., Ferguson, J. F., and Gavis, J. (1976). Arsenate adsorption on amorphous aluminum hydroxide. J . Colloid Interface Sci. 54, 391-399. Anderson, P. R., and Benjamin, M. M. (1990a). Constant-capacitance surface complexation model. Adsorption in silica-iron binary oxide suspension. ACS Symp. Ser. 416,272-281. Anderson, P. R., and Benjamin, M. M. (1990b). Modeling adsorption in aluminium-iron binary oxide suspensions. Environ. Sci. Technol. 24, 1586-1592. Baccini, P., Grieder, E., Stierli, R., and Goldberg, S. (1982). The influence of natural organic matter on the adsorption properties of mineral particles in lake water. Schweiz. Z . Hydrol. 44, 99-116. Balistrieri, L. S., and Chao, T. T. (1990). Adsorption of selenium by amorphous iron oxyhydroxide and manganese dioxide. Geochim. Cosmochim. Acta 54, 739-751. Balistrieri, L. S., and MurrayJ. W. (1979). Surface of goethite (aFeOOH) in seawater. ACS Symp. Ser. 93, 275-298. Balistrieri, L. S., and Murray, J. W. (1981). The surface chemistry of goethite (aFeOOH) in major ion seawater. A m . J. Sci. 281, 788-806. Balistrieri, L. S . , and Murray, J. W. (1982a). The surface chemistry of i3Mn02 in major ion seawater. Geochim. Cosmochim. Acta 46, 1041-1052. Balistrieri, L. S . , and Murray, J. W. (1982b). The adsorption of Cu, Pb, Zn, and Cd on goethite from major ion seawater. Geochim. Cosmochim. Acra 46, 1253-1265. Balistrieri, L. S., and Murray, J. W. (1987). The influence of the major ions of seawater on the adsorption of simple organic acids by goethite. Geochim. Cosmochim. Acta 51, 1151-1160. Ball, J. W., Jenne, E. A., and Cantrell, M. W. (1981). WATEQ3: A geochemical model with uranium added. Geol. Surv. Open-File Rep. (U.S.) 81-1183. Barrow, N. J. (1979). “Computer Programmes for Calculating Charge and Adsorption of Ions in Variable Charge Surfaces,” Tech. Memo. 79/3. CSIRO, Div. Land Resour. Manage. Perth, Australia. Barrow, N. J. (1983). A mechanistic model for describing the sorption and desorption of phosphate by soil. J. Soil Sci. 34, 733-750. Barrow, N. J. (1984). Modelling the effects of pH on phosphate sorption by soils. J . Soil Sci. 35,283-297. Barrow, N. J. (1986a). Testing a mechanistic model. I. The effects of time and temperature on the reaction of fluoride and molybdate with a soil. J . Soil Sci. 37, 267-275. Barrow, N. J. (1986b). Testing a mechanistic model. 11. The effects of time and temperature on the reaction of zinc with a soil. J . Soil Sci. 37, 277-286. Barrow, N. J. (1986~).Testing a mechanistic model. IV. Describing the effects of pH on zinc retention by soils. J. Soil Sci. 37, 295-302.

322

SABINE GOLDBERG

Barrow, N. J. (1987). “Reactions with Variable-Charge Soils.” Martinus Nijhoff, Dordrecht, The Netherlands. Barrow, N. J. (1989a). Testing a mechanistic model. IX. Competition between anions for sorption by soil. J . Soil Sci. 40, 415-425. Barrow, N. J. (1989b). Testing a mechanistic model. X. The effect of pH and electrolyte concentration on borate sorption by a soil. J. Soil Sci. 40, 427-435. Barrow, N. J., and Bowden, J. W. (1987). A comparison of models for describing the adsorption of anions on a variable charge mineral surface. J. Colloid Interface Sci. 119, 236-250. Barrow, N. J . , and Ellis, A. S. (1986a). Testing a mechanistic model. 111. The effects of pH on fluoride retention by a soil. J . Soil Sci. 31, 287-293. Barrow, N. J., and Ellis, A. S. (1986b). Testing a mechanistic model. V. The points of zero salt effect for phosphate retention, for zinc retention and for acid/alkali titration of a soil. J. Soil Sci. 37, 303-310. Barrow, N. J . , and Whelan, B. R. (1989a). Testing a mechanistic model. VII. The effects of pH and of electrolyte on the reaction of selenite and selenate with a soil. J . Soil Sci. 40, 17-28. Barrow, N. J., and Whelan, B. R. (1989b). Testing a mechanistic model. VIII. The effects of time and temperature of incubation on the sorption and subsequent desorption of selenite and selenate by a soil. J . Soil Sci. 40, 29-37. Barrow, N. J., Bowden, J. W., Posner, A. M., and Quirk, J. P., (1980a). An objective method for fitting models of ion adsorption on variable charge surfaces. Aust. J . Soil Res. 18, 37-47. Barrow, N. J., Bowden, J. W., Posner, A. M., and Quirk, J. P. (1980b). Describing the effects of electrolyte on adsorption of phosphate by a variable charge surface. Aust. J . Soil Res. 18, 395-404. Barrow, N. J., Bowden, J. W., Posner, A. M., and Quirk, J. P. (1981). Describing the adsorption of copper, zinc and lead on a variable charge mineral surface. Aust. J . Soil Res. 19,309-321. Barrow, N. J., Gerth, J., and Briimmer, G. W. (1989). Reaction kinetics of the adsorption and desorption of nickel, zinc and cadmium by goethite. 11. Modelling the extent and rate of reaction. J . Soil Sci. 40, 437-450. Benjamin, M. M. (1978). Effects of competing metals and complexing ligands on trace metal adsorption at the oxide/solution interface. Ph.D. Thesis, Stanford University, Stanford, California. Benjamin, M. M., and Bloom, N. S. (1981). Effects of strong binding of anionic adsorbates on adsorption of trace metals on amorphous iron oxyhydroxide. I n “Adsorption from Aqueous Solutions” (P. H. Tewari, ed.), pp. 41-60. Plenum, New York. Benjamin, M. M., and Leckie, J. 0. (1980). Adsorption of metals at oxide interfaces: Effects of the concentration of adsorbate and competing minerals. In “Contaminants and Sediments” (R. A. Baker, ed.), Vol. 2, pp. 305-322. Ann Arbor Sci. Ann Arbor, Michigan. Benjamin, M. M., and Leckie, J. 0. (1981). Multiple-site adsorption of Cd, Cu, Zn, and Pb on amorphous iron oxyhydroxide. J . Colloid Interface Sci. 79, 209-221. BCrubC, Y . G., and deBruyn, P. L. (1968a). Adsorption at the rutile-solution interface. I. thermodynamic and experimental study. J . Colloid Interface Sci. 27, 305-318. BCrubC, Y. G., and deBruyn, P. L. (1968b). Adsorption at the rutile-solution interface. 11. model of the electrochemical study. J . Colloid Interface Sci. 28, 92-105. Bleam, W. F.,Pfeffer,P. E.,Goldberg, S . , Taylor, R. W., andDudley, R. (1991). A3’Psolidstate nuclear magnetic resonance study of phosphate adsorption at the boehmite/aqueoussolution interface. Langmuir 7 , 1702-1712.

SURFACE COMPLEXATION MODELS

323

Blesa, M. A., Maroto, A. J. G . , and Regazzoni, A . E. (1984a). Boric acid adsorption on magnetite and zirconium dioxide. J. Colloid Interface Sci. 99, 32-40. Blesa, M. A., Figliolia, N. M., Maroto, A. J.G., and Regazzoni, A. E. (1984b). The influence of temperature on the interface magnetite-aqueous electrolyte solution. J. Colloid Interface Sci. 101, 410-418. Bloesch, P. M., Bell, L. C., and Hughes, J. D . (1987). Adsorption and desorption of boron by goethite. Aust. 1. Soil Res. 25, 377-390. Bolan, N. S., Barrow, N. J., and Posner, A. M. (1985). Describing the effect of time on sorption of phosphate by iron and aluminum hydroxides. J. Soil Sci. 36, 187-197. Bolan, N. S., Syers, J. K., and Tillman, R. W. (1986). Ionic strength effects on surface charge and adsorption of phosphate and sulphate by soils. J. Soil Sci. 37, 379-388. Bolland, M. D., Posner, A. M., and Quirk, J. P. (1977). Zinc adsorption by goethite in the absence and presence of phosphate. Aust. J. Soil Res. 15, 279-286. Bolt, G . H., and van Riemsdijk, W. H. (1982). Ion adsorption on inorganic variable charge constituents. In “Soil Chemistry. Part B. Physico-Chemical Methods” ( G . H. Bolt, ed.), pp. 459-503. Elsevier, Amsterdam. Bourg, A. C. M., and Schindler, P. W. (1978). Ternary surface complexes. 1. Complex formation in the system silica-Cu(I1)-ethylenediamine. Chimia 32, 166-168. Bowden, J. W., Posner, A . M., and Quirk, J. P. (1977). Ionic adsorption on variable charge mineral surfaces. Theoretical-charge development and titration curves. Aust. J. Soil Res. 15, 121-136. Bowden, J. W., Nagarajah, S., Barrow, N. J., Posner, A. M., and Quirk, J. P. (1980). Describing the adsorption of phosphate, citrate and selenite on a variable charge mineral surface. Am!. J. Soil Res. 18, 49-60. Breeuwsma, A., and Lyklema, J. (1971). Interfacial electrochemistry of haematite (aFe,O,). Discuss. Faraday SOC.52, 324-333. Catts, J. G . , and Langmuir, D. (1986). Adsorption of Cu, Pb and Zu by 6Mn02: Applicability of the site binding-surface complexation model. Appl. Geochem. 1, 255-264. Cederberg, G . A., Street, R. L., and Leckie, J. 0. (1985). A groundwater mass transport and equilibrium chemistry model for multicomponent systems. Water Resour. Res. 21, 1095-1104. Charlet, L. (1986). Adsorption of some macronutrient ions on an oxisol. An application of the triple layer model. Ph.D. Thesis, University of California, Riverside. Charlet, L., and Sposito, G. (1987). Monovalent ion adsorption by an oxisol. Soil Sci. SOC. A m . J. 51, 1155-1160. Charlet, L., and Sposito, G. (1989). Bivalent ion adsorption by an oxisol. Soil Sci. SOC.Am. J . 53, 691-695. Comans, R. N. J . , and Middelburg, J. J. (1987). Sorption of trace metals on calcite: Applicability of the surface precipitation model. Geochim. Cosmochim. Acta. 51, 2587-2591. Cowan, C. E., Zachara, J . M., and Resch, C. T. (1991). Cadmium adsorption on iron oxides in the presence of alkaline-earth elements. Environ. Sci. Technol. 25, 437-446. Davis, J. A. (1977). Adsorption of trace metals and complexing ligands at the oxidelwater interface. Ph.D. Thesis, Stanford University, Stanford, California. Davis, J. A., and Leckie, J. 0. (1978). Surface ionization and complexation at the oxide/ water interface. 11. Surface properties of amorphous iron oxyhydroxide and adsorption of metal ions. J. Colloid Interface Sci. 67, 90-107. Davis, J. A,, and Leckie, J. 0. (1979). Speciation of adsorbed ions at the oxide/water interface. ACS Symp. Ser. 93, 299-317. Davis, J . A , , and Leckie, J. 0. (1980). Surface ionization and complexation at the oxide/ water interface. 3. Adsorption of anions. J. Colloid Interface Sci. 74, 32-43.

324

SABINE GOLDBERG

Davis, J. A., James, R. O., and Lecliie, J. 0. (1978). Surface ionization and complexation at the oxide/water interface. I. Computation of electrical double layer properties in simple electrolytes. J. Colloid Interface Sci. 63, 480-499. Dzornbak, D. A., and Morel, F. M. M. (1990). “Surface Complexation Modeling. Hydrous Ferric Oxide.” Wiley, New York. Farley, K. J., Dzombak, D. A,, and Morel, F. M. M. (1985). A surface precipitation model for the sorption of cations on metal oxides. J. Colloid lntetface Sci. 106, 226-242. Felmy, A. R., Girvin, D. C., and Jenne, E. A. (1984). “MINTEQ: A Computer Program for Calculating Aqueous Geochemical Equilibria,” EPA-600/3-84-032. Office of Research and Development, U.S. Environ. Prot. Agency, Athens, Georgia. Forbes, E. A., Posner, A. M., and Quirk, J. P. (1976). The specific adsorption of divalent Cd, Co, Cu, Pb and Zn on goethite. J. Soil Sci. 27, 154-166. Furrer, G., Westall, J., and Sollins, P. (1989). The study of soil chemistry through quasisteady-state models. I. Mathematical definition of model. Geochim. Cosmochim. Acta 53, 595-601. Furrer, G., Sollins, P., and Westall, J. C. (1990). The study of soil chemistry through quasi-steady-state models. 11. Acidity of soil solution. Geochim. Cosmochim. Acta 54, 2363-2374, Fiirst, B. (1976). Das Koordinationschemische Adsorptionsmodel: Oberflachenkomplexbildung von Cu(II), Cd(I1) und Pb(I1) an SiOz (Aerosil) und Ti02 (Rutil). Ph.D. Thesis, University of Bern, Bern, Switzerland. Gallez, A., Juo, A. S. R., and Herbillon, A. J. (1976). Surface and charge characteristics of selected soils in the tropics. Soil Sci. SOC. Am. J. 40,601-608. Girvin, D. C., Ames, L. L., Schwab, A. P., and McGarran, J. E. (1991). Neptunium adsorption on synthetic amorphous iron oxyhydroxide. J. Colloid Interface Sci. 141, 67-78. Gisler, A. (1980). Die Adsorption von Aminosauren an Grenzflachen Oxid-Wasser. Ph.D. Thesis, University of Bern, Bern, Switzerland. Goldberg, S . (1985). Chemical modeling of anion competition on goethite using the constant capacitance model. Soil Sci. SOC.Am. J . 49, 851-856. Goldberg, S. (1986a). Chemical modeling of arsenate adsorption on aluminum and iron oxide minerals. Soil Sci. SOC. Am. J. 50, 1154-1157. Goldberg, S. (1986b). Chemical Modeling of Specific Anion Adsorption on Oxides, Clay Minerals, and Soils, ENVIROSOFT 86, pp. 671-688. CML Publications, Ashurst, Southampton, U.K. Goldberg, S. (1991). Sensitivity of surface complexation modeling to the surface site density parameter. J. Colloid Interface Sci. 145,l-9. Goldberg, S . , and Glaubig, R. A. (1985). Boron adsorption on aluminum and iron oxide minerals. Soil Sci. SOC. Am. J . 49, 1374-1369. Goldberg, S., and Glaubig, R. A. (1986a). Boron adsorption on California soils. Soil Sci. SOC. Am. J. 50, 1173-1176. Goldberg, S., and Glaubig, R. A. (1986b). Boron adsorption and silicon release by the clay minerals kaolinite, montmorillonite, and illite. Soil Sci. SOC. Am. J. 50, 1442-1448. Goldberg, S., and Glaubig, R. A. (1988a). Boron and silicon adsorption on an aluminum oxide. Soil Sci. SOC. Am. J. 52, 87-91. Goldberg, S., and Glaubig, R.A. (1988b). Anion sorption on a calcareous, montmorillonitic soil-selenium. Soil Sci. SOC. Am. J . 52, 954-958. Goldberg, S., and Glaubig, R. A. (1988~).Anion sorption on a calcareous, montmorillonitic soil-Arsenic. Soil Sci. SOC. Am. J . 52, 1297-1300. Goldberg, S., and Sposito, G. (1984a). A chemical model of phosphate adsorption by soils. I. Reference oxide minerals. Soil Sci. SOC. Am. J. 48, 772-778.

SURFACE COMPLEXATION MODELS

325

Goldberg, S., and Sposito, G. (1984b). A chemical model of phosphate adsorption by soils. 11. Noncalcareous soils. Soil Sci. SOC.Am. J . 48, 779-783. Goldberg, S., and Sposito, G. (1985). On the mechanism of specific phosphate adsorption by hydroxylated mineral surfaces: A review. Commun. Soil Sci. Plant Anal. 16, 801-821. Goldberg, S., and Traina, S. J. (1987). Chemical modeling of anion competition on oxides using the constant capacitance model-mixed-ligand approach. Soil Sci. SOC.Am. J . 51,929-932. Hawke, D., Carpenter, P. D., and Hunter, K . A. (1989). Competitive adsorption of phosphate on goethite in marine electrolytes. Environ. Sci. Technot. 23, 187-191. Hayes, K. F. (1987). Equilibrium, spectroscopic and kinetic studies of ion adsorption at the oxide/aqueous interface. Ph.D. Thesis, Stanford University, Stanford, California. Hayes, K. F., and Leckie, J. 0. (1986). Mechanism of lead ion adsorption at the goethitewater interface. ACS Symp. Ser. 323, 114-141. Hayes, K. F., and Leckie, J. 0. (1987). Modeling ionic strength effects on cation adsorption at hydrous oxide/solution interfaces. J . Colloid Interface Sci. 115, 564-572. Hayes, K. F., Papelis, C., and Leckie, J. 0. (1988). Modeling ionic strength effects on anion adsorption at hydrous oxide/solution interfaces. J . Colloid Interface Sci. 125, 717-726. Hayes, K. F., Redden, G., Ela, W., and Leckie, J. 0. (1991). Surface complexation models: An evaluation of model parameter estimation using FITEQL and oxide mineral titration data. J . Colloid Interface Sci. 142, 448-469. Hiemstra, T., van Riemsdijk, W. H., and Bruggenwert, M. G. M. (1987). Proton adsorption mechanism at the gibbsite and aluminium oxide solid/solution interface. Neth. J . Agric. Sci. 35,281-293. Hiemstra, T., van Riemsdijk, W. H., and Bolt, G. H. (1989a). Multisite proton adsorption modeling at the solid/solution interface of (hydr)oxides: A new approach. I. Model description and evaluation of intrinsic reaction constants. J . Colloid Interface Sci. 133, 91-104. Hiemstra, T., de Wit, J. C. M., and van Riemsdijk, W. H. (1989b). Multisite proton adsorption modeling at the solid/solution interface of (hydr)oxides: A new approach. 11. Application to various important (hydr)oxides. J . Colloid Interface Sci. 133, 105-1 17. Hingston, F. J. (1970). Specific adsorption of anions on goethite and gibbsite. Ph.D. Thesis, University of Western Australia, Nedlands. Hohl, H., and Stumm, W. (1976). Interaction of Pb2+ with hydrous yA1,0,. J . Colloid Interface Sci. 55, 281-288. Hohl, H . , Sigg, L., and Stumm, W. (1980). Characterization of surface chemical properties of oxides in natural waters. Adv. Chem. Ser. 189, 1-31. Honeyman, B. D. (1984). Cation and anion adsorption at the oxide/solution interface in systems containing binary mixtures of adsorbents: An investigation of the concept of adsorptive additivity. Ph.D. Thesis, Stanford University, Stanford, California. Hsi, C.-K.D., and Langmuir, D. (1985). Adsorption of uranyl onto ferric oxyhydroxides: Application of the surface complexation site-binding model. Geochim. Cosmochim. Acta49,1931-1941. Huang, C.-P., and Stumm, W. (1973). Specific adsorption of cations on hydrous y-AI2O3. J . Colloid Interface Sci. 43, 409-420. Hunter, K. A., Hawke, D. J . , and Choo, L. K. (1988). Equilibrium adsorption of thorium by metal oxides in marine electrolytes. Geochim. Cosmochim. Acta 52,627-636. James, R. O . , and Parks, G. A. (1982). Characterization of aqueous colloids by their electrical double-layer and intrinsic surface chemical properties. Surf. Colloid Sci. 12, 119-216.

326

SABINE GOLDBERG

James, R. O., Davis, J. A,, and Leckie, J. 0. (1978). Computer simulation of the conductometric and potentiometric titrations of the surface groups on ionizable latexes. J . Colloid Interface Sci. 65, 331-343. Jennings, A. A., Kirkner, D. J., and Theis, T. L. (1982). Multicomponent equilibrium chemistry in groundwater quality models. Water Resour. Res. 18, 1089-1096. Koopal, L. K., van Riemsdijk, W. H., and Roffey, M. G. (1987). Surface ionization and complexation models: A comparison of methods for determining model parameters. J. Colloid Interface Sci. 118, 117-136. Krupka, K. M., Erikson, R. L., Mattigod, S . V., Schramke, J. A., and Cowan, C. E. (1988). “Thermochemical Data Used by the FASTCHEMTMPackage,” EPRI EA-5872. Electr. Power Res. Inst., Palo Alto, California. Kummert, R., and Stumm, W. (1980). The surface complexation of organic acids on hydrous yA1203. J . Colloid Interface Sci. 15, 373-385. LaFlamme, B. D., and Murray, J. W. (1987). Solid/solution interaction: The effect of carbonate alkalinity on adsorbed thorium. Geochim. Cosmochim. Actu 51,243-250. Leckie, J. O., Benjamin, M. M. Hayes, K. F., Kaufmann, G., and Altmann, S. (1980). “Adsorption/Coprecipitation of Trace Elements from Water with Iron Oxyhydroxide,” EPRI RP-910-1. Electr. Power Res. Inst. Palo Alto, California. Leckie, J. O., Appleton, A. R., Ball, N. B., Hayes, K. F., and Honeyman, B. D. (1984). “Adsorptive Removal of Trace Elements from Fly-Ash Pond Effluents onto Iron Oxyhydroxide,” EPRI RP-910-1. Electr. Power Res. Inst. Palo Alto, California. Liivgren, L., Sjoberg, S . , and Schindler, P. W. (1990). Acid/base reactions and AI(II1) complexation at the surface of goethite. Geochim. Cosmochim. Acta 54, 1301-1306. Madrid, L., and Diaz-Barrientos, E. (1988). Description of titration curves of mixed materials with variable and permanent surface charge by a mathematical model. 1. Theory. 2. Application to mixtures of lepidocrocite and montmorillonite. J. Soil Sci. 39, 215-225. Madrid, L., Diaz, E., Cabrera, F., and de Arambarri, P. (1983). Use of a three-plane model to describe charge properties of some iron oxides and soil clays. J . Soil Sci. 34, 57-67. Madrid, L., Diaz-Barrientos, E., and Sanchez-Soto, P. J. (1989). Description of titration curves of mixed materials with variable and permanent charge by a mathematical model. 3. Influence of the nature of the permanent charge mineral. J. Soil Sci. 40, 799-806. Mangold, D. C., and Tsang, C.-F. (1991). A summary of subsurface hydrological and hydrochemical models. Rev. Ceophys. 29, 51-79. McKenzie, R. M. (1981). The surface charge on manganese dioxides. Aust. J . Soil Res. 19, 41-50. McKenzie, R. M. (1983). The adsorption of molybdenum on oxide surfaces. Aust. J . Soil Res. 21, 505-513. Mikami, N., Sasaki, M., Hachiya, K., Astumian, R. D., Ikeda, T., and Yasunaga, T. (1983a). Kinetics of the adsorption-desorption of phosphate on the yA1203 surface using the pressure-jump technique. J . Phys. Chem. 87, 1454-1458. Mikami, N.,Sasaki, M., Kikuchi, T., and Yasunaga, T. (1983b). Kinetics of adsorptiondesorption of chromate on yAl2O3 surfaces using the pressure-jump technique. J. Phys. Chem. 81,5245-5248. MilonjiC, S . K. (1987). Determination of surface ionization and complexation constants at colloidal silica/electrolyte interface. Colloids Surf. 23, 301-312. Motta, M. M., and Miranda, C. F. (1989). Molybdate adsorption on kaolinite, montmorillonite, and illite: Constant capacitance modeling. Soil Sci. SOC.Am. J . 53, 380-385. Osaki, S., Kuroki, Y.,Sugihara, S., and Takashima, Y. (1990a). Effects of metal ions and organic ligands on the adsorption of Co(I1) onto silicagel. Sci. TotalEnviron. 99,93-103. Osaki, S . , Miyoshi, T., Sugihara, S., and Takashima, Y. (1990b). Adsorption of Fe(III),

SURFACE COMPLEXATION MODELS

327

Co(I1) and Zn(I1) onto particulates in fresh waters on the basis of the surface complexation model. I. Stabilities of metal species adsorbed on particulates. Sci. Total Environ. 99, 105-114. Papelis, C., Hayes, K. F., and Leckie, J. 0. (1988). “HYDRAQL: A Program for the Computation of Chemical Equilibrium Composition of Aqueous Batch Systems Including Surface-Complexation Modeling of Ion Adsorption at the Oxide/Solution Interface,” Tech. Rep. No. 306. Department of Civil Engineering, Stanford University, Stanford, California. Pulfer, K., Schindler, P. W., Westall, J. C., and Grauer, R. (1984). Kinetics and mechanism of dissolution of bayerite (rAI(OH)3) in HN03-HF solutions at 298.2 K. J. Colloid Interface Sci. 101, 554-564. Regazzoni, A. E., Blesa, M. A., and Maroto, A. J.G. (1983). Interfacial properties of zirconium dioxide and magnetite in water. J. Colloid Interface Sci. 91, 560-570. Riese, A. C. (1982). Adsorption of radium and thorium onto quartz and kaolinite: A comparison of solution/surface equilibria models. Ph.D. Thesis, Colorado School of Mines, Golden. Rochester, C. H., and Topham, S . A. (1979a). Infrared study of surface hydroxyl groups on goethite. J. Chem. SOC.,Faraday Trans. 175, 591-602. Rochester, C. H., and Topham, S. A . (1979b). Infrared study of surface hydroxyl groups on haematite. J. Chem. SOC., Faraday Trans. 1. 75, 1073-1088. Sanchez, A. R., Murray, J. W., and Sibley, T. H. (1985). The adsorption of plutonium IV and V on goethite. Geochim. Cosmochim. Acta 49, 2297-2307. Sasaki, M., Moriya, A. M., Yasunaga, T., and Astumian, R. D. (1983). A kinetic study of ion-pair formation on the surface of a-FeOOH in aqueous suspensions using the electric field pulse technique. J. Phys. Chem. 87, 1449-1453. Schindler, P. W. (1990). Co-adsorption of metal ions and organic ligands: Formation of ternary surface complexes. Rev. Mineral. 23, 281-307. Schindler, P. W., and Gamsjager, H. (1972). Acid-base reactions of the T i 0 2 (anatase)-water interface and the point of zero charge of TiOz suspensions. Kolloid-Z. 2. Polym. 250, 759-763. Schindler, P. W., and Kamber, H. R. (1968). Die Aciditat von Silanolgruppen. Helv. Chim. Acta 51, 1781-1786. Schindler, P. W., and Stumm, W. (1987). The surface chemistry of oxides, hydroxides, and oxide minerals. In “Aquatic Surface Chemistry” (W. Stumm, ed.), pp. 83-110. Wiley (Interscience), New York. Schindler, P. W., Fiirst, B., Dick, R., and Wolf, P. U. (1976). Ligand properties of surface silano groups. I. Surface complex formation with Fe3+, Cu2+,CdZ+,and Pb2+.J . Colloid Interface Sci. 55, 469-475. Schindler, P. W., Liechti, P., and Westall, J. C. (1987). Adsorption of copper, cadmium and lead from aqueous solution to the kaolinite/water interface. Neth. J. Agric. Sci. 35, 2 19-230. Sigg, L. (1973). Untersuchungen iiber Protolyse und Komplexbildung mit zweiwertigen Kationen von Silikageloberfiachen, M.Sc. Thesis, University of Bern, Bern, Switzerland. Sigg, L. M. (1979). Die Wechselwirkung von Anionen und schwachen Sauren mit aFeOOH (Goethit) in wassriger Losung. Ph.D. Thesis, Swiss Federal Institute of Technology, Zurich. Sigg, L., and Stumm, W. (1981). The interaction of anions and weak acids with the hydrous goethite (a-FeOOH) surface. Colloids Surf. 2, 101-1 17. Smit, W., and Holten, C. L. M. (1980). Zeta-potential and radiotracer adsorption measurement on EFG a-A1203 single crystals in NaBr solutions. J. Colloid Interface Sci. 78, 1-14.

328

SABINE GOLDBERG

Sposito, G. (1981). Cation exchange in soils: An historical and theoretical perspective. A S A Spec. Publ. 40, 13-28. Sposito, G. (1983). Foundations of surface complexation models of the oxide-aqueous solution interface. J. Colloid Interface Sci. 91, 329-340. Sposito, G. (1984a). “The Surface Chemistry of Soils.” Oxford Univ. Press, Oxford, U.K. Sposito, G. (1984b). Chemical models of inorganic pollutants in soils. CRC Crit. Rev. Environ. Control 15, 1-24. Sposito, G., and Coves, I. (1988). “SOILCHEM: A Computer Program for the Calculation of Chemical Speciation in Soils.” Kearney Foundation of Soil Science, University of California, Riverside. Sposito, G., and Mattigod, S. V. (1980). “GEOCHEM: A Computer Program for the Calculation of Chemical Equilibria in Soil Solutions and Other Natural Water Systems.” Kearney Foundation of Soil Science, University of California, Riverside. Sposito, G., deWit, J. C. M., and Neal, R. H. (1988). Selenite adsorption on alluvial soils. 111. Chemical modeling. Soil Sci. SOC.A m . J . 52, 947-950. Sprycha, R. (1983). Attempt to estimate up charge components on oxides from anion and cation adsorption measurements. J. Colloid Interface Sci. 96, 551-554. Sprycha, R. (1984). Surface charge and adsorption of background electrolyte ions at anatase/electrolyte interface. J . Colloid Interface Sci. 102, 173-185. Sprycha, R. (1989a). Electrical double layer at alumina/electrolyte interface. I. Surface charge and zeta potential. J. Colioid Interface Sci. 127, 1-11. Sprycha, R. (1989b). Electrical double layer at alumina/electrolyte interface. 11. Adsorption of supporting electrolyte ions. J. Colloid Interface Sci. 127, 12-25. Sprycha, R., and Szczypa, I. (1984). Estimation of surface ionization constants from electrokinetic data. J. Colloid Interface Sci. 102, 288-291. Stiglich, P. J. (1976). Adsorption of cadmium(I1) complexes at the oxide/water interface. M. S. Thesis, University of Melbourne, Melbourne, Australia. Stumm, W., Huang, C. P., and Jenkins, S. R. (1970). Specific chemical interaction affecting the stability of dispersed systems. Croat. Chem. Acra 42, 223-245. Stumm, W., Hohl, H., and Dalang, F. (1976). Interaction of metal ions with hydrous oxide surfaces. Croat. Chem. Acra 48, 491-504. Stumm, W., Kummert, R., and Sigg, L. (1980). A ligand exchange model for the adsorption of inorganic and organic ligands at hydrous oxide interfaces. Croar. Chem. Acta 53, 291-312. Swallow, K. C. (1978). Adsorption of trace metals by hydrous ferric oxide. Ph.D. Thesis, Massachusetts Institute of Technology, Cambridge, Massachusetts. Tamura, H., Matijevic, E., and Meites, L. (1983). Adsorption of Coz+ ions on spherical magnetite particles. J . Colloid Interface Sci. 92, 303-314. van Raij, B., and Peech, M. (1972). Electrochemicalproperties of some oxisols and alfisols of the tropics. Soil Sci. SOC. Am. Proc. 36, 587-593. van Riemsdijk, W. H., and Lyklema, J. (1980a). The reaction of phosphate with aluminum hydroxide in relation with phosphate bonding in soils. Colloids Surf. 1, 33-44. van Riemsdijk, W. H., and Lyklema, I. (1980b). Reaction of phosphate with gibbsite [Al(OH),] beyond the adsorption maximum. J . Colloid Interface Sci. 76, 55-66. van Riemsdijk, W. H., and van der Zee, S. E. A. T. M. (1991). Comparison of models for adsorption, solid solution and surface precipitation. I n “Interactions at the Soil ColloidSoil Solution Interface” (G. H. Bolt, ed.), pp. 241-256. Kluwer Academic Publishers, The Netherlands. van Riemsdijk, W. H., Bolt, G. H., Koopal, L. K., and Blaakmeer, I. (1986). Electrolyte adsorption on heterogeneous surfaces: Adsorption models. J. Colloid Interface Sci. 109, 219-228.

SURFACE COMPLEXATION MODELS

329

van Riemsdijk, W. H., dewit, J. C. M., Koopal, L. K., and Bolt, G. H. (1987). Metal ion adsorption on heterogeneous surfaces: Adsorption models. J . Colloid Interface Sci. 116, 511-522. Wersin, P., Charlet, L., Karthein, R., and Stumm, W. (1989). From adsorption to precipitation: Sorption of Mn2+ on FeC03(s). Geochim. Cosmochim. Acta 53, 2787-2796. Westall, J. C. (1979). “MICROQL. I. A Chemical Equilibrium Program in BASIC. 11. Computation of Adsorption Equilibria in BASIC,” Tech. Rep. Swiss Federal Institute of Technology, EAWAG, Diibendorf. Westall, J. (1980). Chemical equilibrium including adsorption on charged surfaces. Adv. Chem. Ser. 189, 33-44. Westall, J. C. (1982). “FITEQL: A Computer Program for Determination of Equilibrium Constants from Experimental Data,” Rep. 82-01. Department of Chemistry, Oregon State University, Corvallis. Westall, J. C. (1986). Reactions at the oxide-solution interface: Chemical and electrostatic models. ACS Symp. Ser. 323, 54-78. Westall, J . , and Hohl, H. (1980). A comparison of electrostatic models for the oxide/ solution interface. Adv. Colloid Interface Sci. 12, 265-294. Westall, J. C., Zachary, J. L., and Morel, F. M. M. (1976). “MINEQL: A Computer Program for the Calculation of Chemical Equilibrium Composition of Aqueous Systems,” Tech. Note 18. Ralph M. Parsons Laboratory, Department of Civil Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts. Wood, R., Fornasiero, D., and Ralston, J. (1990). Electrochemistry of the boehmite-water interface. Colloids Surf. 51, 389-403. Yates, D. E. (1975). The structure of the oxide-aqueous electrolyte interface. Ph.D. Thesis, University of Melbourne, Melbourne, Australia. Yates, D. E., Levine, S . , and Healy, T. W. (1974). Site-binding model of the electrical double layer at the oxide/water interface. J . Chem. Soc., Faraday Trans. I 70, 1807-1818. Yeh, G. T., and Tripathi, V. S . (1990). “HYDROGEOCHEM: A Coupled Model of HYDROlogic Transport and GEOCHEMical Equilibria in Reactive Multicomponent Systems,” Rep. ORNL-6371. Oak Ridge Natl. Lab., Oak Ridge, Tennessee. Zachara, J. M., Girvin, D. C . , Schmidt, R. L., and Resch, C. T. (1987). Chromate adsorption on amorphous iron oxyhydroxide in the presence of major groundwater ions. Environ. Sci. Technol. 21, 589-594. Zachara, J. M., Cowan, C. E., Schmidt, R. L., and Ainsworth, C. C. (1988). Chromate adsorption by kaolinite. Clays Clay Miner. 36, 317-326. Zachara, J. M., Ainsworth, C. C., Cowan, C. E., and Resch, C. T. (1989). Adsorption of chromate by subsurface soil horizons. Soil Sci. SOC. Am. J . 53, 418-428. Zhang, P. C., and Sparks, D. L. (1989). Kinetics and mechanisms of molybdate adsorption/ desorption at the goethite/water interface using pressure-jump relaxation. Soil Sci. SOC. Am. J . 53,1028-1034. Zhang, P. C., and Sparks, D. L. (1990a). Erratum-Kinetics and mechanisms of molybdate adsorption/desorptionon the goethite/water interface using pressure-jump relaxation. Soil Sci. SOC.Am. J . 54,624. Zhang, P. C., and Sparks, D. L. (1990b). Kinetics and mechanisms of sulfate adsorption/ desorption on goethite using pressure-jump relaxation. Soil Sci. SOC. Am. J 54, 12661273. Zhang, P., and Sparks, D. L. (1990~).Kinetics of selenate and selenite adsorption/ desorption at the goethite/water interface. Environ. Sci. Technol. 24, 1848-1856.

This Page Intentionally Left Blank

MODELING THE TRANSPORT AND RETENTIONOF INORGANICS IN SOILS H. M. Selim Agronomy Department, Louisiana State University, Baton Rouge, Louisiana 70803

I. Introduction II. Transport Equations III. Equilibrium Retention Models A, Freundlich B. Langmuir C. Two-Site Langmuir W . Kinetic Retention Models V. Multiple-Reaction Models A. Two-Site Models B. Multireaction Models C. Second-Order Models VI. Transport and Ion Exchange A. Classical Approach B. Mobile-Immobile Approach C. Variable Selectivities D. Specific Sorption and Kinetic Ion Exchange W.Transport in Layered Soil A. Saturated Soils under Constant Flux B. Water-Unsaturated Soils References

I. INTRODUCTION The interactions of dissolved chemicals and their transport in the soil profile play a significant role in their leaching losses beyond the root zone, availability for uptake by plants, and the potential contamination of groundwater supplies. The ability to predict the mobility of dissolved

33 1 Advance, in Agronomy, Volume 47 Copright 0 1992 by Academic Press, Inc. All rights ofreproduction in any form reserved.

332

H. M. SELIM

chemicals in the soil is of considerable value in managing land disposal of wastes and in fertilizer applications. Such predictive capability requires knowledge of the physical, chemical, and biological processes influencing solute behavior in the soil environment. Over the last three decades, a number of theoretical models for solute transport in porous media have been proposed. One group of models deals with solute transport in well-defined geometrical systems wherein it is assumed that the bulk of solute moves in pores and/or cracks of regular shapes or through interaggregate voids of known geometries. Examples of such models include those dealing with a soil matrix of uniform spheres, rectangular or cylindrical voids, and discrete aggregate or spherical size geometries. Solutions of these models are analytic, often complicated, and involve several numerical approximating steps. In contrast, the second group of transport models consists of empirical models that do not consider well-defined geometries of the pore space or soil aggregates. Rather, solute transport is treated on a macroscopic basis, with the water flow velocity, hydrodynamic dispersion, soil moisture content, and bulk density being treated as the associated parameters that describe the soil system. Refinements of this macroscopic approach are the “mobile-immobile” transport models, wherein local nonequilibrium conditions are due to diffusion or mass transfer of solutes between the mobile and immobile regions. The mobile-immobile models have been used to describe transport of several solutes in soils. These empirical or macroscopic models are widely used and are far less complicated than the above-mentioned more exact approach for systems of well-defined porous media geometries. To predict the transport of reactive solutes in the soil, models that include retention and release reactions of solutes with the soil matrix are needed. Retention and release reactions in soils include ion exchange, adsorption/desorption, precipitation/dissolution, and other mechanisms such as chemical or biological transformations. Retention and release reactions are influenced by several soil properties, including bulk density, soil texture, water flux, pH, organic matter, and type and amount of dominant clay minerals. Adsorption is the process whereby solutes bind or adhere to soil matrix surfaces to form outer- or inner-sphere solute-surface site complexes. In contrast, ion exchange reactions represent processes whereby charged solutes replace ions on soil particle surfaces. Adsorption and ion exchange reactions are related in that an ionic solute species may form a surface complex and may replace another ionic solute species already on surface sites. The term “retention,” or the commonly used term “sorption,” should be used when the mechanisms of solute removal from soil solution are not known, and the term “adsorption” should be used only to describe the formation of solute-surface site complexes.

MODELS OF INORGANICS IN SOILS

333

Solute retention proceses in soils have been quantified by several scientists along two different lines. One represents equilibrium reactions and the second represents kinetic or time-dependent types of reactions. Equilibrium models are those for which solute reaction is assumed to be fast or instantaneous in nature and “apparent equilibrium” may be observed in a relatively short reaction time. Langmuir and Freundlich models are perhaps the most commonly used equilibrium models for the description of fertilizer chemicals, such as phosphorus, and for several heavy metals. These models include the linear and Freundlich (nonlinear) and the oneand two-site Langmuir type. Kinetic models represent slow reactions whereby the amount of solute sorption or transformation is a function of contact time. Most common is the first-order kinetic reversible reaction for describing time-dependent adsorption/desorption in soils. Others include linear irreversible and nonlinear reversible kinetic models. Recently, a combination of equilibrium and kinetic-type (two-site) models and consecutive and concurrent multireaction-type models have been introduced. In this article, major features of retention models that govern retention reactions of solutes in the soil are presented. Single-reaction models of the equilibrium type are first discussed, followed by models of the kinetic type. Retention models of the multiple-reaction type, including the twosite equilibrium-kinetic models, the concurrent- and consecutive-multireaction models, and the second-order approach will be derived. This is followed by multicomponent or competitive-type models wherein ion exchange is considered the dominant retention mechanism. Selected experimental data sets will be described for the purpose of model evaluation and validation, and necessary (input) parameters are discussed. For convenience, we first present formulations of the transport equations that govern the transport of solutes in a water-saturated and water-unsaturated porous medium. Retention reactions of the reversible and irreversible types are incorporated into the transport formulation. Boundary and initial conditions commonly encountered under field conditions are also presented.

II. TRANSPORT EQUATIONS Dissolved chemicals present in the soil solution are susceptible to transport through the soil subject to the water flow constraints in the soil system. At any given point within the soil, the total amount of solute y, (pg/cm3) for a species i may be represented by XI = oci

+ psi

(1)

334

H. M. SELIM

where S is the amount of solute retained by the soil (pg/g soil), C is the solute concentration in solution (pg/ml), 0 is the soil moisture content (cm3/cm3), and p is the soil bulk density (g/cm3). The rate of change of x for the ith species with time is subject to the law of mass conservation such that (omitting the subscript i) d(0C + pS) = -div J - Q dt

(2)

or

where t is time (hr) and J,, J y , and J, represent the flux or rate of movement of solute species i in the x , y , and z directions (pg/cm2.hr), respecitvely. The term Q represents a sink (Q positive) or source that accounts for the rate of solute removal (or addition) irreversibly from the bulk solution (pg/cm3 * hr). If we restrict our analysis to one-dimensional flow in the z direction, the flux J, , or simply J, in the soil may be given by

where Dm is the molecular diffusion coefficient (cm2/hr), DLis the longitudinal dispersion coefficient (cm2/hr), and q is Darcy’s flux (cm/hr). Therefore, the primary mechanisms for solute movement are due to diffusion plus dispersion and by mass flow or convection with water as the water moves through the soil. The molecular diffusion mechanism is due to the random thermal motion of molecules in solution and is an active process regardless of whether there is net water flow in the soil. The result of the diffusion process is the well-known Fick’s law of diffusion wherein the flux is proportional to the concentration gradient. The longitudinal dispersion term of Eq. (4)is due to the mechanical or hydrodynamic dispersion phenomena, which are due to the nonuniform flow velocity distribution during fluid flow in porous media. According to Fried and Combarnous (1971), nonuniform velocity distribution through the soil pores is a result of variations in pore diameters along the flow path, fluctuation of the flow path due to tortuosity effect, and variation in velocity from the center of a pore (maximum value) to zero at the solid surface interface (Poiseuille’slaw). The effect of dispersion is that of solute spreading, which is a tendency opposite to that of piston flow. Dispersion is effective only during fluid flow; i.e., for a static water condition or when water flow is near zero, molecular diffusion is the dominant process for

MODELS OF INORGANICS IN SOILS

335

solute transport in soils. For multidimensional flow, longitudinal dispersion coefficients ( D L )and transverse dispersion coefficients (DT)are needed to describe the dispersion mechanism. Longitudinal dispersion refers to that in the direction of water flow and that for the transverse directions for dispersion perpendicular to the direction of flow. Apparent dispersion D is often introduced to simplify the flux Eq. (4) such that I3C

J=-OD-+qC d.2

where D now refers to the combined influence of diffusion and hydrodynamic dispersion for dissolved chemicals in porous media. Incorporation of flux Eq. (5) into the conservation of mass Eq. (3) yields the following generalized form for solute transport in soils in one dimension,

The above equation is commonly known as the convective-dispersive equation for solute transport, which is valid for soils under transient and unsaturated soil-water flow conditions. For conditions wherein steady water flow is dominant, D and 0 are constants; i.e., for uniform 0 in the soil, we have the simplified form of the convection-dispersion equation as

where v (cm/hr) is known as the pore water velocity and is given by (4/0). Solutions of the above convection-dispersive Eqs. (6) and (7) yield the concentration distribution of the amount of solute in soil solution C and that retained by the soil matrix S with time and depth in the soil profile. In order to arrive at such a solution, the appropriate initial and boundary conditions must be specified. Several boundary conditions are identified with the problem of solute transport in porous media. The simplest is that of a first-order type of boundary condition such that a solute pulse input is described:

C=C,,

z=O,

t
(8)

C=O,

z=O,

trT

(9)

where C, (pg/cm3) is the concentration of the solute species in the input pulse. The input pulse application is for a duration T , which is then followed by a pulse input that is free of such a solute. Such a boundary condition was used by Lapidus and Amundson (1952) and Cho (1971),

3 36

H. M. SELlM

among others. The more precise third-type boundary condition at the soil surface was considered by Brenner (1962) in his classical work, wherein advection plus dispersion across the interface was considered. A continuous solute flux at the surface can be expressed as,

and a flux-type pulse input as

O = - D d2 ~+vC,

z=O,

trT

(12)

Advantages of using the third-type boundary conditions have been discussed by Selim and Mansell (1976) and Kreft and Zuber (1978). The boundary conditions at some depth L in the soil profile are often expressed as (Danckwerts, 1953; Brenner, 1962; Lindstrom et al., 1967; Kreft and Zuber, 1978), z=L, t r o (13) which is used to deal with solute effluent from soils having finite lengths. However, it is often convenient to solve the dispersion-convection equation wherein a semi-infinite rather than a finite length ( L ) of the soil is assumed. Under such circumstances, the appropriate condition for a semiinfinite medium is dCldz=O,

ac/dz=o,

tzo (14) Analytical solutions to the convection-dispersion equation subject to the appropriate boundary and initial conditions are available for a number of situations whereas the majority of the solute transport problems must be solved using numerical approximation methods. In general, whenever the form of the retention reaction is a linear one, a closed-form solution is obtainable. A number of closed-form solutions are available from Crank (1956), Ozisik (1968), Kreft and Zuber (1978), and van Genuchten and Alves (1982). However, most retention mechanisms are nonlinear and time dependent in nature and analytical solutions are not available. As a result, a number of numerical models using a finite difference or finite element have been utilized to solve nonlinear retention problems of multireaction and multicomponent solute transport for one- and twodimensional geometries (Rubin and James, 1973; Valocchi et al., 1981; Miller and Benson, 1983; Cederberg et al., 1985; Selim et at., 1987; Mansell et al., 1988). z+a,

MODELS OF INORGANICS IN SOILS

337

III. EQUILIBRIUM RETENTION MODELS The form of solute retention reaction in the soil system must be identified if prediction of the fate of reactive solutes in the soil using the convection-dispersion Eq. (7) is sought. The reversible term (&/at) is often used to describe the rate of sorption or exchange reactions with the solid matrix. Sorption or exchange has been described by either instantaneous equilibrium or a kinetic reaction whereby concentrations in solution an sorbed phases vary with time. Reviews of various forms of equilibrium and kinetic models are given by Murali and Aylmore (1983), Selim (1989), Selim et al. (1990a). Recently, Nielsen et al. (1986) presented a comprehensive discussion of significant features of sorption exchange reactions of the equilibrium and kinetic type. Linear, Freundlich, and one- and two-site Langmuir equations are perhaps most commonly used to describe equilibrium reactions. Freundlich and Langmuir reactions and their use in describing equilibrium retention are discussed in subsequent sections. This is followed by kinetic-type reactions and their implication for single and multireaction retention and transport models.

A. FREUNDLICH The Freundlich equation is perhaps the simplest approach for quantifying the behavior of retention of reactive solute with the soil matrix. It is certainly one of the oldest of the nonlinear sorption equations and has been used widely to describe solute retention by soils (Helfferich, 1962; Travis and Etnier, 1981; Murali and Aylmore, 1983; Sposito, 1984). The Freundlich equation is S = KdCb (15) where S is the amount of solute retained by the soil in pg/g, C is the solute concentration in solution in pg/ml, Kd is the distribution coefficient in cm3/g, and the parameter b is dimensionless and typically has a value of b < 1. The distribution coefficient describes the partitioning of a solute species between solid and liquid phases over the concentration range of interest and is analogous to the equilibrium constant for a chemical reaction. For b equals unity, the Freundlich equation is often referred to as the linear retention equation (see Table I). There are numerous examples for solute retention, which was described successfully by use of the Freundlich equation (see Sposito, 1984; Travis and Etnier, 1981; Murali and Aylmore, 1983; Sparks, 1989). Selected examples of linear and Freundlich (or nonlinear) retention are shown in Fig. 1 for phosphate (P) sorption from batch studies for Al and A2

H. M. SELIM

338

Table I Selected Equilibrium and Kinetic-Type Models for Solute Retention in Soils" Model

Formulationb

~

~~

Equilibrium type Linear Freundlich (nonlinear) Langmuir Langmuir with sigmoidicity Kinetic type First order n-th order Irreversible (sinklsource) Langmuir kinetic Elovich Power Mass transfer

S = KdC S = KdCb

s = WCS,,/(l+ S = oCS,,,/(l

OC)

+ OC+ u/C)

d S / d t = k f ( O / p ) C - kbS

a s / d t = k f ( @ / p ) C " -kbs dS/Jt = k,(O/p)(C - C,) JS/Jt= kf(O/p)C(S,,, - S ) - kbs

dS/dt=Aexp(-BS) as/at = K(O/p)C"Sm d S / d t = K(O/p)(C- C * )

"Adapted from Selim er al. (1990a), with permission. 'A, B , b, C', C,, K, Kd, kb, k f , k,, n , m, Lax, 0, and adjustable model parameters.

ff

are

horizons of an Oldsmar fine sand (Mansell et al., 1977). For both isotherms, it appears that the P isotherms can be adequately described by the Freundlich equation. Logarithmic representation of the Freundlich equation is frequently used to represent the data as illustrated in Fig. 2. Here the slope of the best-fit curve provides the nonlinear parameter b and 1 0 0 -,

Figure 1. Equilibrium phosphate adsorption isotherms in surface A, and subsurface A2 soils of an Oldsmar fine sand. From Mansell et al. (1977). with permission.

MODELS OF INORGANICS IN SOILS

-

1000

I

I

I

339

Chromium

loot-

a

1

1 %

1.0-

-.

.

0.01

A

I.o

0.I

100

10

C (rng/liter)

Figure 2. Retention isotherms for chromium on three surface soils,Alligator (Al), Kula (Ku), and Windsor (Wi). From Buchter et al. (1989), with permission.

the intercept as & according to log(S) = & + b log(C) as long as a linear representation of the data in the log form is achieved. In Fig. 2, we illustrate the use of the Freundlich equation for Cd retention for three different soils, whereas Fig. 3 shows Pb, Cu, Cd, and Co isotherms for one (Alligator) soil (Buchter et al., 1989). Although the Freundlich equation has been rigorously derived (Sposito, 1980), the goodness of fit of the Freundlich equation to solute retention data does not provide definitive information about the actual processes involved, because the equation is capable of describing data irrespective of the actual retention mechanisms. Often complex retention processes can at 1000

100

-

f-

10

v1

1.0

0.I 0.001

0.01

0.I

1.0

10

100

C (mg/liter)

Figure 3. Retention isotherms Co, Cd, Cu, and Pb on Alligator soil. From Buchter et al. (1989), with permission.

340

H. M. SELIM

least in part be described by relatively simple models such as the Freundlich equation. Therefore, the Freundlich parameters Kd and b are best regarded as descriptive parameters in the absence of independent evidence concerning the actual retention mechanism.

The Langmuir isotherm is the oldest and most commonly encountered isotherm in soils. It was developed to describe the adsorption of gases by solids when a finite number of adsorption sites in the surface is assumed (Langmuir, 1918). As a result, a major advantage of the Langmuir equation over linear and Freundlich types is that a maximum sorption capacity is incorporated into the formulation of the model, which may be regarded as a measure of the amount of available retention sites on the solid phase. The standard form of the Langmuir equation is S --

, ,s

--

wc

1+wc

where w and S, are adjustable parameters. Here w (cm3/kg) is a measure of the bond strength of molecules on the matrix surface and S ,, (pg/g soil) is the maximum sorption capacity or total amount of available sites per unit soil mass. In an attempt to classify the various shapes of sorption isotherms, it was recognized that the Langmuir isotherm is the most commonly used and is referred to as the L-curve isotherm (Sposito, 1984). The Langmuir sorption isotherm has been used extensively by scientists for several decades. Travis and Etnier (1981) provided a review of studies in which the Langmuir isotherm to describe P retention for a wide range of soils was used. Moreover, Langmuir isotherms were used successfully to describe Cd, Cu, Pb, and Zn retention in soils. Figure 4 shows experimental and fitted isotherm examples of use of the Langmuir equation to describe Cr(V1) retention for three soils (Selim and Amacher, 1988).

Based on several retention data sets, the presence of two types of surface sites responsible for the sorption of P in several soils was postulated. As a consequence, the Langmuir two-surface isotherm was proposed (Holford

MODELS OF INORGANICS IN SOILS

341

c, rng liter -1

Figure 4. Chromium sorption isotherms for Cecil, Windsor, and Olivier soils after 14 days of reaction. Solid curves are calculated isotherms using equilibrium two-site Langmuir model. From Selim and Amacher (1988), with permission.

el

al., 1974) such that

where f (dimensionless) is considered as a fraction of type 1 sites to the total sites and w1 and 0, are the Langmuir coefficients associated with sites 1 and 2, respectively. The above equation is an adaptation of the original equation proposed by Holford et al. (1974) and was used to describe P isotherms by Holford and Mattingly (1975) for a wide range of soils. A more recent adaptation of the two-surface Langmuir equation is the incorporation of a sigmoidicity term, where -= S

s,

fOlC 1+ W l C + (a1/C)

+

(1 - f b 2 C 1 + w,c + (a,/C)

(18)

The terms al and a, are the sigmoidicity coefficients (pg/cm3) for type 1 and 2 sites, respectively. Schmidt and Sticher (1986) found that the introduction of this sigmoidicity term was desirable in order to adequately describe sorption isotherms at extremely low concentrations. Examples of

342

H. M. SELIM 5.5o= 1 n = 2 A = 3

0.000

0.005

0.015 0.020 0.025 Equilibrium Concentration (mg/ml)

0.010

0.030

O.(

Figure 5. Sorption isotherms for lead (l), cadmium (2), and copper (3) in the Ah horizon of a Luvisol. Solid curves are calculated isotherms using equilibrium two-site Langmuir model with sigmoidicity. From Schmidt and Sticher (1986), with permission.

the two-surface Langmuir with sigmoidicity are shown in Fig. 5 for Pb, Cd, and Cu in the Ah horizon of a Luvisol (Schmidt and Sticher, 1986). Although the Langmuir approach has been used to model P retention and transport from renovated wastewater, we are not aware of studies wherein the two-surface Langmuir with sigmoidicity has been used to describe solute retention during transport in soils.

IV.KINETIC RETENTION MODELS For several solutes, retention reactions in the soil solution have been observed to be strongly time dependent (e.g., phosphorus, several heavy metals, and organics). Selected examples of kinetics retention for Cd are given in Fig. 6. Here, the kinetic dependence of Cd retention, carried out in batch experiments, is shown for various soils (Selim, 1989). The amount of cadmium retained varied among soils, with Cecil soil exhibiting the lowest retention, whereas Sharkey soil showed maximum Cd sorption from soil solution. The fast decrease in Cd concentration (with time) indicates a fast-type sorption reaction that was followed by slower type reactions. It is

MODELS OF INORGANICS IN SOILS

343

Figure 6. Cadmium (Cd) concentration versus time of reaction for five soils with initial Cd concentration (G) = 1 pg/ml). From Selim (1989), with permission.

also apparent that after 300 hr of reaction time, quasi-equilibrium conditions were not attained. The results of Fig. 7 illustrate the influence of kinetic reactions on the shape of sorption isotherms of P for an Oldsmar fine sand (Fiskell etal. 1979). The amount of P removed from solution increased with solution concentration as well as time. It is obvious that when dealing with such data sets the use of equilibrium type models will probably provide inadequate prediction of the fate of solutes for these soils. A number of empirical models have been proposed to describe kinetic retention reactions of solutes in the solution phase. The earliest model is the first-order kinetic reaction, which was first incorporated into the classical (convective-dispersive) transport equation by Lapidus and Amundson (1952). dS

p-=

at

k f O C - k,pS

(19)

344

H. M. SELIh4

i 2 3 4 5 6 7 Phosphorus Concentration in Solution (pglml), C

O:,

Phosphorus Concentration in Solution (pg/ml),C Figure 7. Relationship of P sorbed versus solution concentration with time in shallowtilled (ST)and deeptilled (DT) Oldsmar fine sand. From Fiskell er &. (1979), with permission.

where kf and kb represent the forward and backward rates of reactions (hr-'), respectively. This type of reaction is fully reversible when the magnitudes of the rate coefficients dictate the extent of the kinetic behavior of the reaction. For small values of kf and kb , the rate of retention is slow and strong kinetic dependence is anticipated. In contrast, for large values

MODELS OF INORGANICS IN SOILS

345

of kf and kb, the retention reaction is a rapid one and should approach quasi-equilibrium in a relatively short time. In fact, at large times (t -+ a), when the rate of retention approaches zero, the above equation yields

which results in a linear equation, where Kd = Okf/pk,. This is similar to that for linear isotherms wherein equilibrium conditions were assumed. In addition, as discussed earlier, nonlinear retention has been commonly observed for several solutes (see also Fig. 7). As a result, the first-order reversible kinetic reaction has been extended to include the nonlinear kinetic type (Mansell et al., 1977),

where rn is a dimensionless parameter commonly less than unity and represents the order of the nonlinear reaction. In a similar formulation to the linear case, at large times (t+ m), when the rate of retention approaches zero, the above equation reduces to

which is similar in form to the Freundlich equation. Therefore, for linear or Freundlich isotherms, one may regard the distribution coefficient Kd as the ratio of the rate of sorption (forward reaction) to that for desorption or release (backward reaction). To illustrate solute retention behavior (adsorption-desorption) when the first-order or the nonlinear kinetic reactions are used, we present several simulations in Figs. 8 and 9 (Selim et al., 1976). As shown in Fig. 8 (top), the linear kinetic adsorption isotherms are strongly time dependent and even after more than 50 hr only 90% of equilibrium was achieved. The slow attainment of equilibrium can be attributed primarily to the magnitude of the retention rate coefficients kf and kb. Figure 8 (bottom) also shows simulated adsorption curves for 10 and 50 hr and desorption (dashed) curves initiated after 10 and 50 hr of adsorption. The simulated desorption isotherms shown were obtained by reducing the solute solution concentration one-half successively, every 10 (or 50) hr until the solution concentration was less than 5 pg/cm3. This procedure is similar to that used in desorption studies in the laboratory. As seen from the family of (dashed) curves, desorption did not follow the same path (i.e., nonsingularity) as the respective adsorption isotherm (solid curves). Obviously,

H. M. SELIM

346

LINEAR

200 -

6 0

o

100

200

300

400

500

600

- ADSORB

I

_ _ - - DESORB -

0

100

200

sOLUT ION

300

400

500

COO

CONCENT RATION (pg/g)

Figure 8. Simulated adsorption-desorption isotherms using linear kinetic retention. Desorption was initiated after 10 and 50 hr for each successive sorption step. From Selim er al. (1976),with permission.

this singularity or hysteresis results from failure to achieve equilibrium adsorption prior to desorption. If adsorption as well as desorption were carried out for times sufficient for equilibrium to be attained, or the kinetic rate coefficients were sufficiently large, such hysteretic behavior would be minimized. For the example shown in Fig. 9, reaction times greater than 200 hr were required to achieve >90% equilibrium; the adsorption and desorption times for curves at 200 hr were approximately the same. The time-dependent adsorption and the nonsingularity or hysteretic behavior of the adsorption-desorption processes are illustrated in Fig. 9, when nonlinear retention [see Eq. (21)] was considered. It is apparent that the hysteretic property is for linear or nonlinear kinetic reactions. Hysteretic behavior has been observed by several scientists (Munns and Fox,

MODELS OF INORGANICS IN SOILS

80

.

347

1 NONLINEAR 4

1-

h

5

-0

200

400

600

800

1000

0

- ADSORB

0

200

SOLUTION

400

GOO

800

1000

CONCENTRATION (w/crn3)

Figure 9. Simulated adsorption-desorption isotherms using nonlinear kinetic retention. Desorption was initiated after 10 and 50 hr for each successive sorption step. From Selim et al. (1976), with permission.

1976; Fluhler et al., 1982). Examples of hysteretic behavior for fluoride adsorption-desorption isotherms for two calcareous Swiss soils are given in Fig. 10. The nonlinear behavior of fluoride sorption isotherms is clearly shown from this data set. First-order kinetic reactions have been utilized to quantify the irreversible retention processes in soils. Specifically, the irreversible term Q in the convection-dispersion Eq. (6) or (7) is commonly used to account for various (sink/source) reactions, including precipitation/dissolutions, mineralization, immobilization, biological transformations, volatilization, and radioactive decay, among others. Models that account for Q as first-order kinetic (sink/source) and sequential first-order (irreversible) decay reactions include those of Cho (1971), Selim and Iskandar (1981), Rasmuson and Neretienks (1981), and Amacher et al. (1988). The

H. M. SELLM

348

zoo00

15 WO

?i fl

SCHITTERWALD 0-30 cm

10 OOO

A- - -A m. ........

- 5000

7

P

m-

.-..-.."

G L.

o

5

t Y

Dewrption

*-.-.-t

i! .-s

- ---m

Adsorption

I

500

I

I

1500

lo00

20000

I

2ooo

2! 0

SCHITTERWALD 3 0 - 5 O ~ n

LL

18

2 15000 10 OOO

-

Adwrption

.---A

..........

5000

.-----m

Desorption

0%

500

loo0

1500

2000

2500

Dissolvedconccntrrtion C,[pgF *rnl-']

Figure 10. Fluoride adsorption-desorption hysteresis isotherms tor a Swiss calcareous soil. From Fliihler et al. (1982), with permission.

MODELS OF INORGANICS IN SOILS

349

first-order irreversible retention form is

Q = k&C

(23) where kirris the irreversible rate of reaction (hr-I). On the other hand, description of precipitation reactions that involve secondary nucleation is not an easy task and it is often difficult to distinguish between precipitation and adsorption. In fact, Sposito (1984) stated that the problem of differentiating adsorption from precipitation is made more severe by the facts that new solid phases can precipitate homogeneously onto the surface of existing solid phases and that weathering solids may provide host surfaces for the more stable phases into which they transform chemically. First-order kinetic reactions have also been used as an approximate method for describing the retention reaction due to ion exchange on pure clays. Jardine and Sparks (1984) studied the kinetics of potassium retention on kaolinite, montmorillonite, and vermiculite and found that the adsorption process is kinetic in nature, as shown in Fig. 11. They also found that a single first-order decay-type reaction described the data adequately for kaolinite and montmorillonite, whereas two first-order reactions were

0

20

40

60

80

100

Time (min) 120 140

160

180

200

220 240

260 -l

-.2

-.4 -.6

Y

=m -1.0 -1.2 -1.4

-1.6 -1.8

Figure 11. First-order plots of potassium adsorption on clay minerals, where K,is quantity adsorbed at time t and K, is quantity adsorbed at equilibrium. From Jardine and Sparks (1984),with permission.

350

H. M. SELZM

necessary to describe potassium retention on vermiculite. Jardine and Sparks (1984) suggested that deviations of experimental data from firstorder kinetics at larger times (when equilibrium is approached) are likely due to the fact that potassium retention is not an irreversible but rather a reversible mechanism. At large times (or large amounts of potassium adsorbed) the contribution of the reverse or backward retention process becomes significant and thus should not be ignored. Other cations and anions that exhibited kinetic ion exchange behavior include Al, NH4, and several heavy metals. An extensive list of studies that illustrate kinetic behavior has been recently compiled by Sparks (1989). According to Ogwada and Sparks (1986), observed kinetic ion exchange behavior in soils is probably due to mass transfer (or diffusion) and chemical kinetic processes. They stated that for chemical sorption to occur, ions must be transported to active (fixed) sites of the soil particles. The film of water adhering to and surrounding the particles, and water within the interlayer spaces of the particles, are both zones of low concentrations due to depletion by adsorption of ions onto the exchange sites. The decrease in concentration in these two interface zones may be compensated by diffusion of ions from the bulk solution. Other kinetic models that are used to describe the rate of reversible and irreversible reactions, such as the Elovich model and the diffusioncontrolled model, are given in Table I. A variety of other kinetic reactions are given by Travis and Etnier (1981) and Murali and Aylmore (1983) and a detailed discussion of the characteristics of several reactions is available (Sparks, 1989).

V. MULTIPLE-REACTION MODELS The problem of identifying the fate of solutes in soils must include accounting for retention reactions and transport of the various species in the soil environment (Theis, 1988; Barrow, 1989). In fact, Barrow (1989) stated that the use of single-reaction models, such as those described above, is not adequate because such models describe the fate of only one species with no consideration to the simultaneous reactions of others in the soil system. This is supported by the work of Amacher et al. (1986), who showed that sorption-desorption of Cd, Cr, and Hg from batch studies on several soils were not described by use of single-reaction equations of the equilibrium Langmuir or Freundrich type. They also found that a firstorder kinetic reaction was not capable of describing Cd, Cr, and Hg con-

MODELS OF INORGANICS IN SOILS

351

centrations in the soil solution with contact time. Aringhieri et al. (1985) showed that retention of Cd and Cu on an organic soil was highly kinetic. A description of Cd and Cu reactions using a second-order approach gave adequate predictions of their behavior provided that the reaction rate coefficients were time dependent. Multisite or multireaction models deal with the multiple interactions of one species in the soil environment. Such models are empirical in nature and are based on the assumption that a fraction of the total sites are highly kinetic whereas the remaining fraction of sites interact slowly or instantaneously with those in the soil solution (Selim er al., 1976; Jardine et al., 1985). Nonlinear equilibrium (Freundlich) and first- or nth-order kinetic reactions were the associated processes. Such a two-site approach proved successful in describing observed extensive tailing of breakthrough results. Another two-site approach was proposed by Theis et al. (1988) for Cd mobility and adsorption on goethite. They assumed that the nature of reactions for both sites was governed by second-order kinetic reactions. The reactions were assumed to be consecutive, with the second reaction being irreversible in nature. Amacher et al. (1988) developed a multireaction model that includes concurrent and concurrent-consecutive processes behavior of Cd and Cr(V1) with time for several soils. In addition, the model predicted that a fraction of these heavy metals were irreversibly retained by the soil. Recently, Amacher et al. (1990) concluded that the multireaction model was also successful in describing adsorption of mercury for several soils.

A. T w o - S m M o ~ ~ u One of the earliest multireaction models is the two-site model proposed by Selirn et al. (1976). This model was developed in order to describe observed batch results that showed rapid initial retention reactions followed by slower types of reactions. The model was also developed to describe the excessive tailing of breakthrough results obtained from pulse inputs in miscible displacement experiments. The two-site model is based on several simplifying assumptions. First, it is assumed that a fraction of the total sites (referred to as type I sites) react rapidly with the solute in soil solution. In contrast, we assume that type I1 sites are highly kinetic in nature and react slowly with the soil solution. The retention reactions for both types of sites were based on the nonlinear (or nth-order) reversible kinetic approach as discussed previously. Therefore, the convectivedispersive transport equation with the two-site retention mechanisms may

352

H. M. SELIM

be expressed as

ST = s 1 + s,

(27)

where S1and Sz are the amount retained by sites I and sites 11, respectively, S, is the total amount of solute retained by the soil matrix (pg/g soil), and kl ,k 2 , k 3 , and k4 are the associated rate coefficients (hr-'). The nonlinear parameters n and m were considered less than unity and n f m . For the case m = n = 1, the retention reactions are of the first-order type and the problem becomes a linear one. This two-site approach was also considered for the case when type 1sites were assumed to be in equilibrium with the soil solution, whereas type 2 sites were considered of the kinetic type. Such conditions may be attained when the values for the forward and backward (or kl and k 2 )rate coefficients are extremely large in comparison to the water flow velocity (4).Under these conditions, the solute convective-dispersive transport equation for a combination of equilibrium and kinetic retention is (Selim et al., 1976)

as, - k3-o C" -at

P

- k&

Here Kd is the Freundlich distribution coefficient associated with type 1 sites for nonlinear equilibrium reaction (S, = KdC"). The term R is the retardation factor, which for this nonlinear case is a function of C . Jardine et al. (1985) found that the use of the equilibrium and kinetic two-site models provided good predictions of breakthrough curves (BTCs) for A1 from kaolinite at different pH values (see Fig. 12). Selim et al. (1976) found that the two-site model yields improved predictions of the excessive tailing of the desorption or leaching side and the sharp rise of the sorption side of the BTCs in comparison to predictions using single-reaction equilib-

MODELS OF INORGANICS IN SOILS

353

Figure 12. Breakthrough curve for A1 in a kaolinite column. Solid curve is calculated using the kinetic-equilibrium two-site model. From Jardine et at. (1985), with permission.

rium or kinetic models. The two-site model has been used by several scientists, including De Camargo et al. (1979), Nkedi-Kizza et al. (1984), Jardine ef al. (1985), and Parker and Jardine (1986), among others. The model proved successful in describing the retention and transport of several dissolved chemicals, including aluminum, 2,4-D, atrazine, phosphorus, potassium, cadmium, chromium, and methyl bromide. However, there are several inherent disadvantages of the two-site model. First, the reaction mechanisms are restricted to those that are fully reversible. Moreover, the model does not account for possible consecutive-type solute interactions in the soil system.

B. MULTIREACTION MODELS A schematic representation of the multireaction model is shown in Fig. 13. In this model we consider the solute to be present in the soil solution phase (C) and in four phases representing solute retained by the soil matrix as S, , S1,SZ, S 3 , and Sin. We further assume that S, , S1,and S2 are in direct contact with the solution phase and are governed by concurrent-type reactions. Here we assume S, as the amount of solute that is sorbed reversibly and is in equilibrium with C a t all times. The governing equilibrium retention/release mechanism was that of the nonlinear Freundlich type, as discussed previously.

3 54

H. M. SELIM

Figure 13. A schematic representation of the multireaction model. From Selim et al. (1990a), with permission.

The retention/release reactions associated with S1and S2 were considered to be in direct contact with C and reversible processes of the (nonlinear) kinetic type govern their reactions; S, = KdCb

(31)

where kl to k4 are the associated rate coefficients (hr-I). These two phases ( S , and S2) may be regarded as the amounts sorbed on surfaces of soil particles and chemically bound to A1 and Fe oxide surfaces or other types of surfaces, although it is not necessary to have a priori knowledge of the exact retention mechanisms for these reactions to be applicable. Moreover, these phases may be characterized by their kinetic sorption and release behavior to the soil solution and thus are susceptible to leaching in the soil. In addition, the primary difference between these two phases not only lies in the difference in their kinetic behavior but also on the degree of nonlinearity as indicated by the parameters n and m. The multireaction model also considers irreversible solute removal via a retention sink term Q in order to account for irreversible reactions such as precipitation/ dissolution, mineralization, and immobilization, among others. We expressed the sink term as a first-order kinetic process

where kirris the associated rate coefficient (hr-I).

MODELS OF INORGANICS IN SOILS

355

The multireaction model also includes an additional retention phase (S3) that is governed by a consecutive reaction with S 2 . This phase represents the amount of solute strongly retained by the soil that reacts slowly and reversibly with S2 and may be a result of further rearrangements of the solute retained on matrix surfaces. Thus, inclusion of S, in the model allows the description of the frequent observed very slow release of solute from the soil (Selim, 1981). The reaction between S2 and S3 was considered to be of the kinetic first-order type, i.e., d&dt = k5S2 - k&

(35) where k5 and k6 (hr-I) are the reaction rate coefficients. If a consecutive reaction is included in the model, then Eq. (33) must be modified to incorporate the reversible reaction between S2 and S,. As a result, Eq. (36)

must be used in place of Eq.(33). The above reactions are nonlinear in nature and represent initial-value problems that were solved numerically using finite-difference approximations (explicit-implicit). The initial conditions were that of a given initial solute concentration and assumed no solute retained at time zero, as is the case for kinetic batch experiments (see Amacher et al., 1988). Details of the numerical scheme and computer model code are documented in the multireaction model (MRM) program and are in Selim et al. (1990a). In addition, the above retention mechanisms were incorporated, in a separate model, into the classical convection-dispersion equation in order to predict solute retention as governed by the multireaction model during transport in soils (Selim et al., 1989). The initial and boundary conditions used were those described earlier [Eqs. (11)-(13)], where it is assumed that a solute solution of a concentration C, is applied to a soil column having a length L for a given duration T and is thereafter followed by a solute-free solution. The numerical solution for the multireaction and transport model (MRTM) is documented in program MRTM and is given in Selim et al. (1990a). To test how well the multireaction model is capable of predicting the kinetics of solute retention patterns, several MRM variations were examined (Amacher etal., 1988). This ranged from models wherein all solute reactions (see Fig. 13) were included to model variations wherein only three phases (C, S1,and Sir,)were considered. All model variations were tested using Cr(V1) retention data for a Windsor soil at an initial Cr(V1) concentration in solution of 1.0 pg/ml. Amacher et al. (1988) showed that the MRM version that accounted for two concurrent, nonlinear reversible reactions and one concurrent, first-order irreversible reaction provided the

H. M.SELIM

356

best overall prediction (lowest root mean square) of this data set. However, they concluded that a number of model variations can produce simulations of the data that are indistinguishable. A similar conclusion was made by Skopp (1986). For example, it was not possible to determine whether the irreversible reaction is concurrent or consecutive, because both variations provided similar overall fit of the batch data. Consequently, due to its simplicity, the concurrent reaction version was utilized for futher predictions of other data sets, as shown by the solid curves of Fig. 14. Amacher et al. (1988) found that the magnitude of the rate coefficients that provided the predictions shown were highly dependent on the initial (input) concentration Co. This was an indication that although the model is successful in describing kinetic data for a given Co, the same rate coefficients cannot be used to describe data for substantially different initial concentrations. Thus, the model may be considered as an oversimplification and does not provide a complete description of the actual processes that occur during sorption/desorption of solutes in soils. MRM is best considered as a representation of an apparent rate law, rather than a mechanistic rate law. The predictive capability of the multireaction and transport model was tested for Cr(V1) and P for several soils. Breakthrough curves for Cr(V1) CR I

-

WINDSOR

0

0 . I'

4b

0's

164

1b2

240

I

200

I

336

TIME, HR Figure 14. Experimental data of Cr(VI) retention with time for a Windsor soil for several initial concentrations (C& Predictions using the multireaction model are shown by the solid lines. From Amacher et al. (1988), with permission.

MODELS OF INORGANICS IN SOILS

357

from miscible displacement experiments are shown in Fig. 15 for Windsor soil (Selim et al., 1989). The solid and dashed curves are MRTM calculations using model parameters ( k , , k 2 , k 3 , k4, and kirr) derived from the batch studies of Amacher et al. (1988). Different MRTM model predictions were obtained because a unique set of values for the rate coefficients were not realized, rather a strong dependence on input concentration was observed. It is obvious that underestimation of Cr(V1) retention, and thus overestimation of potential mobility of Cr(V1) in Windsor soil, was consistently observed for all data sets of rate coefficients used. Because an excellent fit of a data set does not in itself constitute a proof of any specific retentionfrelease mechanism, no efforts were made to examine the capabilities of different model variations based on curve fitting alone. Instead, MRTM was utilized along with the optimization (curvefitting) scheme to test its capability for describing the transport data without reliance on parameter estimates from the batch experiments. The hypothesis was that a model gives an inaccurate representation of the reaction mechanism and should thus be discarded if it is incapable of describing the experimental data. As indicative by the curve shown in Fig. 14, adequate model predictions were obtained for the Cr(V1) data and the goodness of fit as measured by r2 exceeded 0.90. I

I

I

I

I

I

I

CR - Windsor I .o

0.a

e V

0.6

0.4

0.2

0

---_ 1

4

8

12

v/ v, Figure 15. Measured (closed circles) and predicted breakthrough curves for Cr(V1) in Windsor soil. Curves A, B , and C are MRTM model predictions using batch rate coefficients for C, = 2, 5, and 100 mg/liter, respectively. Curve D is a BTC using parameters obtained using least-squares best fit. From Selim ef al. (1989), with permission.

358

H. M. SELIM

For the above Cr(V1) simulations, model predictions were performed such that the consecutive reactions between S2 and S3 were ignored. The presence of a consecutive S3phase may occur as a result of further surface rearrangement of the adsorbed phase (see Fig. 13). In order to illustrate the versatility of the multireaction approach, we show phosphorus (P) results for which predictions were obtained when the consecutive sorbed phase was considered in the model (Fig. 16). Such predictions were obtained after several attempts were made for a Norwood soil using different model variations (Selim et al., 1990b). These simulations suggest that incorporation of the consecutive reaction improved the MRTM prediction capability of P in this soil.

C. SECOND-ORDER MODELS In this section, we present an analysis of a second-order kinetic approach for the description of multiple-solute retention reactions during transport in soils (Selim and Amacher, 1988). The basis for this approach is that it accounts for the sites on the soil matrix that are accessible for retention of

l.o

1 V

= 0.460

cm/hr

Q

0.60

0.4--

0.2--

0

I

0

4

a

I

I

I

12

16

20

,

24

1

I

2a

32

36

VlV,

Figure 16. Measured (closed circles) and predicted BTCs for P in Norwood soil. Curves

A and B are MRTM model predictions using nonlinear least-squares with and without S3,

respectively. From Selim et al. (1990b),with permission.

MODELS OF INORGANICS IN SOILS

3 59

the reactive solutes in solution. We incorporated this approach into two transport models; namely, the chemical nonequilibrium model and the diffusion-controlled mobile-immobile (two-region) model. Model evaluation was carried out using batch and miscible displacement data sets from different soils. 1. Second-Order Two-Site

In this model one assumes that there exist two types of retention sites on soil matrix surfaces.. The major difference between the two types of sites is based on the rate of the proposed kinetic retention reactions. We also assume that the retention mechanisms are site specific, e.g., the sorbed phase on type I sites may be characteristically different (in their energy of reaction and/or the identity of the solute-site complex) from that on type 2 sites. An additional assumption is that the second-order reactions associated with sites 1 and 2 may be considered as kinetically controlled, heterogeneous chemical retention reactions (Rubin, 1983). One can assume that these processes are predominantly controlled by surface reactions of adsorption and exchange. In this sense, the previousiy described model is along the same lines as the earlier two-site model of Selim etal. (1976). We denote F as the fraction of type 1 sites to the total amount of sites ST. We also denote 4 as the amount of unfilled or vacant sites in the soil, ( b l = ST1

- S1= FST - S1

= ST, - Sz= FST- Sz

(37)

(38) where 41and +z are amounts of vacant sites, S1and Sz are the amounts of solute retained on type 1 sites and type 2 sites, respectively. As the sites become occupied by the retained solute, the amount of vacant sites approaches zero [(41+ &) --f 01 and the amount of solute retained by the soil approaches that of the total capacity of sites [(Sl + Sz) + ST]. Therefore, we propose here that the retention mechanisms follow a second-order kinetic-type reaction whereby the forward process is controlled by the product of concentration (C) and the amount of unoccupied sites (4), by the reversible processes, 42

k

c+41+& and

3 60

H. M. SELIM

where kl to k4 are the associated rate coefficients. A result, the rate of retention may be expressed as p

as, = k1091C - k2pS1 as2

p-

dt

= k3O&C

- k4pS2

for type 1 sites

(39)

for type 2 sites

(40)

It is convenient to regard type 1 sites as those wherein equilibrium is rapidly reached. In contrast, type 2 sites are highly kinetic and may require several days or months for apparent local equilibrium to be achieved. As t + 00, i.e., when both sites achieve local equilibrium, the rates of retention become klO&C - k2pSl = 0

or

--

(41)

here w1 and w2 represent equilibrium constants for the retention reactions associated with type 1 and 2 sites, respectively. The formulations of Eqs. (41) and (42) are analogous to expressions for homovalent ion exchange equilibrium reactions. In this sense, the equilibrium constants o1 and w2 resemble the selectivity coefficients for exchange reactions and S, resembles the exchange capacity (CEC) of soil matrix surfaces (see Sposito, 1981, Chapter 5). However, a major difference between ion exchange and the proposed second-order approach is that no consideration for other competing ions in solution or matrix surfaces is incorporated in the rate of reactions. In a strict thermodynamic sense, the above equations should be expressed in terms of activities rather than concentrations. However, an implicit assumption is that solution-phase ion activity coefficients are constant in a constant ionic strength medium. Moreover, the solid-phase ion activity coefficients are assumed to be incorporated in the selectivity coefficients (wl and w z ) as in ion exchange formulations (Sposito, 1981). For the case wherein only one type of active site is dominant in the soil, the reaction can be described by the reversible kinetic Langmuir equation ,

Here k f and kb are the forward and backward retention rate coefficients and S is the total amount of solute retained by the soil matrix surfaces.

MODELS OF INORGANICS IN SOILS

361

Reaction (43) at equilibrium obeys the widely recognized Langmuir isotherm equation,

where K = Okf/pkb and K is equivalent to w of Eq. (16). [For a discussion of the formulation of the kinetic Langmuir equation, see Rubin (1983) and Jennings and Kirkner (1984) .] The above second-order two-site (SOTS) model along with a sink term of the first-order kinetic type such as that of Eq. (34) were incorporated into the classical convection-dispersion equation in a manner similar to that for the multireaction model (for details, see Selim and Amacher, 1988). The appropriate equations were solved numerically using finitedifference approximations (explicit-implicit). The resultant numerical scheme is documented in program SOTS. 2. Second-Order Mobile-Immobile

In this model, we consider the processes of retention to be controlled by two types of reactions-namely, a chemically controlled heterogeneous reaction and a physically controlled reaction (Rubin, 1983). We assume that the chemically controlled heterogenous reaction is governed according to the second-order approach. In the meantime, the physically controlled reaction is chosen to be described by diffusion or mass transfer of the mobile-immobile concept (Coats and Smith, 1964; Van Genuchten and Wierenga, 1976). To utilize this concept, we assume 0" as the mobile water content that is present inside large (interaggregate) pores, where solute transport occurs by convection and dispersion. The immobile water (Oim) is located inside aggregate pores (intraaggregate), where the solute transfer occurs by diffusion only. Moreover, C" and C'" are solute concentrations in the mobile and immobile phases, respectively. In addition, the soil matrix is assumed to be divided into two regions (or sites), namely, a dynamic or easily accessible region and a less accesible region. The dynamic region is located close to the mobile water phase whereas the stagnant region is in contact with the immobile phase. Analogies between the mobile-immobile concept and that of the two-site approach can be made. One may regard the dynamic and stagnant regions for solute retention analogous to sites 1 and 2 of the two-site concept. Nkedi-Kizza et al. (1984) presented a detailed discussion on the equivalence of the mobileimmobile and the equilibrium-kinetic two-site models. The form of the transport equation for the mobile-immobile water concept can be written

H. M. SELIM

362 as

acm dS " 0" dt +pf"-+@i"- d t

dC'" dt

dS'"

+ P ( W r n ) 7

where Dm is the hydrodynamic dispersion coefficient in the mobile water region and S" and Simare the sorbed amounts in the dynamic and stagnant regions, respectively. The associated transfer equation governing the interaction between solute in the mobile and immobile phases is dC'" dS'" 0'"+ p (1 - f") dt = (y (C" - Cim)- Q " dt

where a is the mass transfer coefficient between the mobile and immobile phases. The parameterf" represents the fraction of dynamic or active sites to the total sites S, and is analogous to F of the two-site model described above. Again, the terms Q" and Qimrepresent irreversible retention for the mobile and immobile regions, respectively, and were considered in a similar manner as discussed in the two-site model,

The retention mechanisms associated with the mobile and immobile phases were governed by our second-order kinetic approach. Specifically, the rate of reaction for S" and Simwere considered as

Here Cp" and represent the vacant or unfilled sites within the dynamic and the stagnant regions, respectively. In addition, the terms 4" and #m can be expressed as

#" = s,m- S"

=f"S,

- S"

= s p - s i m = (1 - f m ) s ,

(51)

- Sim

(52)

where ST,SF,and SF are the total amounts of the sites in the soil matrix, total sites in the dynamic region, and the total in the less accessible region,

MODELS OF INORGANICS IN SOILS

363

respectively. These terms are related by ST = SF + sirn =f"sT + (1 - f")sT

(53) The unfilled sites +rn and 4" are analogous to and & of the two-site concept [Eqs. (37) and (38)]. Again we assume that ST,which represents the total amount of filled and unfilled sites, remains time invariant. An important feature of the second-order retention approach [Eqs. (49) and (50)] is that similar reaction rate coefficients ( k , and k2) associated with the dynamic and stagnant regions were chosen. Specifically, it is assumed that the retention mechanism is equally valid for the two regions of the porous media. A similar assumption was made by van Genuchten and Wierenga (1976) for equilibrium linear reactions and by Selim et al. (1987) for selectivity coefficients for homovalent ion exchange reactions. The differential equations of the second-order mobile-immobile model (SOMIM) described above were solved using explicit-implicit finitedifference approximations subject to the appropriate initial and boundary conditions. The numerical scheme is documented in program SOMIM.

3. Evaluation In order to examine the capability of the SOTS and SOMIM models, Selim and Amacher (1988) utilized Cr(V1) transport data for three soils from miscible displacement experiments. The necessary parameters needed for model validation were either independently measured or estimated by indirect means. Specifically, they relied on parameter estimate for ST, F, kl , k 2 , k 3 , k 4 , and kirr,from kinetic batch studies. Estimates of STand F were based on the assumption that Cr(V1) in solution and the two types of sites attained quasi-equilibrium in 336 hr even though small amounts of Cr(V1) were still being retained by the soil. Moreover, if the magnitude of the irreversible term is small, as was the case here, then reliable estimates of S, and F can be made, although the actual S , is somewhat smaller. The statistical results from the parameter optimization scheme indicated a close approximation of the two-site Langmuir Eq. (17) to the experimental sorption isotherms (Fig. 6). Estimated ST and F were subsequently used with the SOTS model to describe the time-dependent retention of Cr(V1) by Cecil, Olivier, and Windsor soils. Estimates for kl , k 2 , k 3 , k4, and kirr, which provided best fits of the batch data, were obtained. Selim and Amacher concluded that, based on the high r2 values and low parameter standard errors, good predictions of the kinetic behavior of Cr(V1) using the SOTS model were obtained for all three soils. Moreover, SOTS model predictions of the batch data were indistinguishable from those obtained using the multireaction model described earlier.

364

H. M. SELIM

In fact, predictions using SOTS were similar to those shown in Fig. 14 for Windsor soil with MRM predictions. However, it should be emphasized that close approximation of the data by either model does not in itself constitute proof that the two types of sites actually exist nor that the retention reactions postulated in both models are the actual mechanisms. The predicted BTCs shown in Figs. 17 and 18 were obtained using different sets of parameter values for the rate coefficients ( k , , k2,k 3 , k4, and kirJ for the SOTS model. This is because a unique set of values for these rate coefficientswas not obtained from the batch data, rather a strong dependence of rate coefficients on input concentration was observed (Selim and Amacher, 1988). The use of batch rate coefficients at Co = 100 pg/ml, which is the concentration of Cr pulse inputs, grossly underestimated Cr retention by the predicted BTCs for the Windsor and Olivier soils. Reasons for this failure are not fully understood, with a most likely explanation being that the model is an apparent rather than a mechanistic rate law, which may not completely account for all reactions and reaction components. For both soils, closest predictions were realized using batch rate coefficients from Co = 10 to 25 pg/ml. The capability of the second-order mobile-immobile model to describe Cr miscible displacement results was also examined and predictions for OLlVlER

-

SOTS

1.0--

0.8,-

0

1

2

3

4

v I v,

5

6

7

8

Figure 17. Effluent concentration distributions for Cr(V1) in Olivier soil. Curves A , B, C, D, and E are predictions using the second-order two-site (SOTS)model with batch rate coefficients for C, of 100,25, 10, 5, and 1 pg/ml, respectively. From Selim and Amacher (1988),with permission.

365

MODELS OF INORGANICS IN SOILS

WINDSOR - S O T S

I

1.c

0.a 0

0.6 0 0.4

0.2

L

0

2

4

6

8

10

12

14

16

v / v, Figure 18. Effluent concentration distributions for Cr(V1) in Windsor soil. Curves A , B, C, D, and E are predictions using the second-order two-site (SOTS)model with batch rate coefficients for C, of 25, 10, 5, 2, and 1 pg/ml, respectively. From Selim and Amacher (1988), with permission.

Windsor and Olivier soils are shown in Figs. 19 and 20. To obtain the predicted BTCs shown, several assumptions were necessary for the estimation of model parameters. Values for S,, kl, k 2 , k 3 , k 4 , and kirrwere those used with the SOTS model as obtained from the kinetic batch studies. Estimates for the ratios of the immobile water fraction (@"/a) were obtained based on soil moisture retention relations and the fraction of sites f" was assumed to be equal to am/@ because independent measurement off" is not available. Estimates for a and D" were also obtained and were based on the assumption of uniform size aggregates (for details, see Selim and Amacher, 1988). Predicted BTCs were obtained using different sets of batch rate coefficients due to their strong dependence on input concentrations (Co values). Closest predictions to experimental Cr measurements were obtained from batch rate coefficients at low Co values (Co5 10 pg/ml). Moreover, the use of rate coefficients at higher Co values resulted in decreased tailing and reduced retardation of the BTCs. These observation are consistent with previous predictions using the SOTS model (Fig. 10). However, overall predictions of measured Cr using the SOMIM are considered less than adequate in comparison to SOTS model predictions. Reasons for the less than adequate predictions of BTCs for the three soils using SOMIM are not fully understood. It is conceivable that a

H. M. SELIM

3 66

OL l VIE R-SO M IM

0

1

2

3

5

4

6

7

8

v I v, Figure 19. Eftluent concentration distributions for Cr(V1) in Olivier soil. Curves A, B, C, and D are predictions using the SOTS model with batch rate coefficientsfor C, of 25, 10, 5 , and 1 pg/ml, respectively. From Selim and Amacher (1988), with permission.

0

2

4

6

8

VI

10

12

14

16

v,

Figure 20. Effluent concentration distributions for Cr(V1) in Windsor soil. Curves A, B, C, and D are predictions using the SOTS model with batch rate coefficients for C, of 25, 5 , 2, and 1 pg/ml, respectively. From Selim and Amacher (1988), with permission.

MODELS OF INORGANICS IN SOILS

3 67

set of applicable rate coefficients over the concentration range for the Cr transport experiments cannot be obtained simply by using the batch procedure described in this study. In addition, several parameters used in model calculations were estimated and not measured, for example, Om, a, and D m . Other possible factors responsible for these predictions may be due in part to lack of nonequilibrium conditions between the mobile and immobile fractions for the SOMIM model (Valocchi, 1985).

VI. TRANSPORT AND ION EXCHANGE The models described above were focused on the description of the transport and retention of one solute species only. Such an assumption implies that all other interactions that occur in the soil do not inflluence the behavior of the solute species under consideration. Although research on multicomponent interactions and transport in ion exchange columns has been an important subject in ion chromatography literature for several decades (see, e.g., Helfferich, 1962; Helfferich and Klein, 1970), this research area was addressed in the soil science literature only during the last two decades. One of the early works dealing with the transport and interaction of two species is that of Lai and Jurinak (1972). In their approach an equilibrium reaction was assumed to govern the exchange of two ions between the solution and the soil surfaces; this reaction was incorporated into the classical convective-dispersive equation. The work of Rubin and James (1973) may be considered one of the earliest classical studies describing competitive ion exchange during the transport of multiple cation species in the soil profile. They developed a transport model (convective-dispersive) that dealt with ion exchange for multiple cations present in the soil solution. Ion exchange reactions were assumed to be instantaneous, which implies local equilibrium conditions. Valocchi et al. (1981) extended this work to include multiple ion transport under conditions of varying ionic strength of the soil solution. Their model was used to simulate the transport of Na, Ca, and Mg ions in a shallow underground aquifer. The use of ion concentrations rather than activities was shown not to restrict the predictive capability of the model. Models that consider several processes, for example, ion exchange, complexation, and competitive adsorption, include FIESTA (Jennings er al., 1982), CHEMTRAN (Miller and Benson, 1983), and TRANQL (Cederberg et al., 1985), among others. A major disadvantage of such multicomponent models lies in the basic assumption of local equilibrium of

3 68

H. M. SELIM

the governing reactions. In addition, due to their complexity, several of these models have not been fully validated. CHEMTRAN and TRANQL models were used successfully to describe the Ca and Mg concentration data presented earlier by Valocchi et al. (1981). Kirkner et al. (1985) utilized the FIESTA model to describe Ni and Cd breakthrough results on a sandy soil. Model predictions provided higher retardation of Cd and lower retardation for Ni. However, improved predictions were obtained when a kinetic approach was used with parameters obtained from batch experiments. An overview of ion exchanged and transport models is presented below. The intent is to emphasize the importance of ion exchange as the retention mechanism during water flow in soil. Models for which nonequilibrium ion exchange behavior is described using physical (two-region) or chemical kinetic approaches using specific sorption are discussed.

A. CLASSICAL APPROACH We consider ion exchange as the governing mechanism for the retention of cations on soil matrix surfaces. In a standard mass-action formulation, the exchange reaction for two competing ions i and j may be written as (Sposito, 1981).

where TKii denotes the thermodynamic equilibrium constant and a and a* (omitting the subscripts) are the ion activity in soil solution and on the exchanger surfaces, respectively. Based on Eq. (54), one can denote the parameter vKi, as

where “K is the Vanselow selectivity coefficient and 6 is the activity coefficient on the soil surfaces. For binary homovalent exchange, i.e., vi= vj = v, and assuming similar ion activities in the solution phase, rearranging Eq. (55) yields

where Kij is a generic selectivity coefficient of ions i over j (Rubin and James, 1973) or a separation factor for the affinity of ions on exchange

MODELS OF INORGANICS IN SOILS

3 69

sites (Helfferich, 1962). In addition, ciand cj are the relative ion concentrations (dimensionless) such that ci= Ci/CT and c, = Ci/CT, where Ci, Cj, and CT [mmol( +)/ml] are the concentrations in the soil solution of ions i and j the total concentration. Also, si and sj are amounts retained on the solid matrix surfaces (dimensionless) and are expressed as equivalent fractions where si= SJR and sj = S,/R. Here, Si and S, are the amounts adsorbed [mmol(+)/g soil] and R is the cation exchange (or adsorption) capacity of the soil [mmol( +)/g soil]. Therefore, for homovalent exchange we have (Valocchi ef al., 1981)

rearrangement of Eq. (57) yields the following isotherm relation (for ion 1)

Because C, and R are assumed constant, the respective values for ion 2 (i.e., c2 and s2) can be easily obtained. This isotherm relation Eq. (58) indicates that for K12# 1 we have a nonlinear-type sorption isotherm. Upon incorporation of the above ion exchange sorption isotherm into the convective-dispersive transport Eq. (7), we have (for ion 1)

where the term R may be referred to as the retardation factor, expressed as

and R is a function of relative concentration when K12# 1. Breakthrough results for Ca and Mg leaching from Abist soil columns, packed uniformly with two aggregate sizes, are shown in Fig. 21. These results are from Selim et al. (1987), wherein a Mg pulse was introduced into Ca-saturated soil columns and the total concentration (CT) was maintained constant. The solid and dashed lines shown are model calculations of Ca and Mg results based on the classical convective-dispersive transport equation with ion exchange as the retention mechanism. In general, adequate model predictions for both ions were achieved for the two aggregate sizes for early times or pore volumes. Less than adequate predictions were obtained, however, for large times, i.e., during leaching of the Mg pulse by Ca. Selim et al. (1987) suggested that such model deviations may be due to lack of complete local equilibrium between the ions in solution and those

H. M. SELIM

370

-

1-2 rnm (aggregates)

e E 2 +

3

-

Y

IEz

-

Y

z

0 -

F

0

a

6

c

5

W

4

U

i

0

f 0 0

5

10

20

15

25

4-5 rnm

-

5 '

AA

3 2

0

0

5

1 Ca

10

15

20

V N ,

1

/ - -

1

-

30

25

30

Figure 21. Calcium and Mg breakthrough results from soil columns for two aggregate sizes of Abist soil. Predictions obtained using the classical ion exchange model are shown by the smooth curves. From Selim et al. (1987),with permission.

on the ion exchange surfaces. Similar findings were reported by Gaston and Selim (1990) for binary (Ca and Mg) and ternary (Ca, Mg, and Na) systems in a well-aggregated Sharkey clay soil.

B. M O B I L E - ~ O BAPPROACH ILE In an attempt to describe the transport and exchange reactions of cations in aggregated porous media, the physical nonequilibrium approach of the mobile-immobile (or two-region) concept was utilized. Van Eijkeren and Loch (1984) and Schulin et al. (1986) considered the exchange of cations to be governed by a general form of the exchange equation, which was then incorporated into the mobile-immobile, convective-dispersive transport equation. The capability of this approach was examined by Selim et al. (1987) for describing the mobility of Ca and Mg ions in soil columns,

MODELS OF INORGANICS IN SOILS

371

packed with 1- to 2-mm and 2- to 4-mm aggregates, under conditions of constant and variable ionic strength of the soil solution. Mansell et al. (1988) examined this mobile-immobile approach for the transport and exchange reactions for ternary systems (Na-Ca-Mg) in soil. In their analysis, Mansell et al. (1988) allowed cation exchange selectivities to vary with fractional coverage of the exchange sites, which provided improved predictions of cation breakthrough results. To describe ion retention in the dynamic and less accessible regions of the mobile-immobile concept, the sorption process is governed by the equilibrium ion exchange relationship of Eq. (57). We also assume that such a relationship is valid for the exchange sites of both regions, i.e., ion affinity for the dynamic region is the same as that for the less accessible sites, (61) Such an assumption was used by van Eijkeren and Loch (1984), Selim et al. (1987), and Gaston and Selim (1990). To test the capability of the mobile-immobile concept of describing the transport of cations, the breakthrough results shown previously (see Fig. 21) were utilized. Model predictions of Ca and Mg concentrations in the leachate for the two aggregate sizes are given by the solid and dashed lines shown in Fig. 22. Obvious improvements in model predictions, in comparison to the classical approach (Fig 21), were achieved and may be considered as evidence of lack of local equilibrium between the ions in the soil solution and those on the exchange surfaces. Improved predictions using this approach were achieved by Schulin et al. (1989) for Ca and Mg transport under conditions of variable ionic strength for two wellaggregated forest soils and by Mansell et al. (1988) for a ternary (Na-CaMg) system in a Yo10 loam soil. (K12Irn= (K12)'" = K12

C. VARIABLESELEOTWTES In the preceding classical and mobile-immobile concepts, the distribution of each pair of cations, i and j , can be described by a constant selectivity coefficient K i j . Such assumptions make it possible to arrive at recursion formulas for multiple ions, as were introduced by Rubin and James (1973). Generalized isotherms for multiple ions were based on binary exchange coefficients for all combinations of ions present in the soil system. Such generalized isotherms were used by Valocchi et al. (1981) and Mansell et al. (1988). However, such an assumption of constant exchange selectivity is often unfounded as evidenced by several isotherm data in the

H. M. SELIM

3 72 "

I-

U

6

I-

z

5

u

z

4

0

3

[I

w

0

I -2 mm (aggregates)

0

5

10

0

5

10

Z

0 -

I

15

20

25

:1 30

2 1

0

15

20

25

30

VIV, Figure 22. Calcium and Mg breakthrough results from soil columns for two aggregate sizes of Abist soil. Predictions obtained using the two-region ion exchange model are shown by the smooth curves. From Selim ef nl. (1987),with permission.

literature (Sposito, 1981; Jardine and Sparks, 1984; Parker and Jardine, 1986; Mansell et al., 1988). As a result, Kij coefficients are no longer constant but vary with the relative fraction of cations on the exchange surfaces. Mansell et al. (1988) utilized the sorption isotherm results of Lai et al. (1978) for Mg-Ca and Na-Ca in order to calculate binary selectivity coefficients Kij versus relative ion concentration (C/C,), shown in Fig. 23. The dashed curves represent least-squares best fit of the data to the empirical relation log(Kij) = a + bC. The results clearly show a strong Kij dependency over the concentration range with high affinity of Mg over Ca for C/CT less than 0.4. As expected, adsorption of Ca or Mg was preferred to adsorption of Na. However, the results show an increased Kij for Na --* Ca at low (C/CT < 0.4) and high (C/CT > 0.85) relative concentrations. The influence of variable selectivity coefficients on the predictions of cation transport was investigated by Mansell et al. (1988). The transport data sets used were those from miscible displacement experiments of Lai et af. (1978). Mansell et al. (1988) incorporated additional terms into the

373

MODELS OF INORGANICS IN SOILS

0

Mg --+ Ca

L

-

2

5 !2 w *

-*

0.3 -

.--

Na

I

3-

3

- - --

Ca

.-;---0

- \

0.1

-\

0

0.2I

*

I

.

I

-

,

I

,

classical (convective-dispersive) equation in order to account for variable Kji values for the ternary systems (Na-Ca-Mg) of Lai et al. (1978). They also utilized the mobile-immobile approach in conjunction with variable Kji values in order to describe the same data set. Mansell et al. (1988) found that good breakthrough predictions were obtained for the relatively noncompetitive Na when either constant or variable Kii values were used with the classical model (see Fig. 24). However, the use of constant Kij values underestimated the tailing of the Mg breakthrough data. Description of Mg tailing was improved when variable Kii values were used, but the extent of Mg retardation (peak location) was somewhat overestimated. The combined use of the mobile-immobile approach and variable K j j values provided the best overall description of Na and Mg breakthrough results.

D. SPECIFIC SORPTION AND KINETICIONEXCHANGE In this approach, two mechanisms were considered as the dominant retention processes in the soil, namely, ion exchange and specific sorption. We consider ion exchange as a nonspecific sorption/desorption process. Ion exchange is a fully reversible mechanism and is assumed here to be

H. M. SELIM

3 74 0.5

CONSTANT SELECTIVITIES

0 VARIABLE SELECTIVITIES

0

1

2

3

4

5

6

VIV, Figure 24. Sodium and Mg breakthrough results from soil columns of Yolo soil (Lai etal., 1978). Solid and dashed curves are predictions using the classical and two-region models, respectively. Exchange selectivity coefficients used in model predictions were considered constant (top figure) or variable (bottom figure). From Mansell et al. (1988), with permission.

either rapid (i.e., instantaneous) or may be considered as a kinetic process. Specific sorption is considered as a kinetic process wherein ions have high affinity for specific sites on matrix surfaces. Furthermore, retention of ions via specific sorption is regarded as an irreversible or weakly reversible process. The kinetic ion exchange reaction was analogous to mass transfer or between the solid and solution phase such that (for ion i) ds,/dt = a(si*- S i )

(62) where si is the amount sorbed (at time t ) on matrix surfaces, sT is the equilibrium sorbed amount, and a is an apparent rate coefficient (hr-'). Here s* (for ion 1) was calculated using the equilibrium isotherm Eq. (58). Obviously, as t + and/or large values of a,s* and s become equal and equilibrium conditions prevail, Expressions similar to the above equation have been used to describe mass transfer between mobile and immobile water and chemical kinetics (Parker and Jardine, 1986; Selim and Amacher, 1988). Studies that illustrate kinetic ion exchange behavior include those of Jardine and Sparks (1984) and Ogwada and Sparks (1986).

MODELS OF INORGANICS IN SOILS

375

It was postulated that in 2 :1 types of minerals, intraparticle diffusion as a possible rate-controlling mechanism governs the kinetics of adsorption of cations (Sparks, 1989). The specific sorption process was considered as a kinetic reaction whereby the rate of sorption is governed by a second-order mechanism such that

+

is where k f and kb are the forward and backward rate coefficients (hr-'), the amount of vacant specific sites, and I)is the amount specifically sorbed [mmol( +) g-'1, respectively. Vacant specific sites are not strictly vacant. They are assumed occupied by hydrogen, by hydroxyl, or by other specifically sorbed species. The role of specific sorption and its influence on the behavior of metal ions has been recognized by several investigators. Sorption/desorption studies showed that highly specific sorption mechanisms are responsible for metal ion retention for low concentrations (Tiller et al., 1979, 1984). The general view was that metal ions have a high affinity for sorption sites of oxide minerals surfaces in soils. In addition, these specific sites react slowly with heavy metals and are weakly reversible. In the absence of competing metal ions for specific sites (e.g., Ni, Co, and Cu), as is the case in this study, it is reasonable to consider specific sorption as an irreversible process. Therefore, the above second-order reaction was modified to describe irreversible or weakly reversible retention by setting the backward rate coefficient k b as zero,

p(a+/at) = k@+C (64) For several metal ions (e.g., Cd, Ni, Co, and Zn), specific sorption was shown to be dependent on time of reaction. Therefore, the use of kinetic rather than an equilibrium sorption mechanism is recommended. Although our model formulation is based on direct reaction between metal ions in soil solution and specific sorption sites, others have considered a consecutive-type approach for Cd sorption. According to Thesis et al. (1988), a set of two second-order reactions was considered; one fully reversible step was followed by an irreversible reaction. Ion exchange, as discussed above, was not considered. Theis et al. (1988) argued that the amount adsorbed on geothite surfaces was susceptible to migration (via surface diffusion) from primary to secondary surface sites. Other possible mechanisms may include formation of surface complexes, hydrolysis of sorbate at the surface, and surface complexation. Tiller el al. (1979) quantified specific sorption as the amount of sites that retain metal ions following several washings of the soil with high concentrations of a nonspecifically sorbed cation (0.01 M Ca2NO3). As a result, metal ions on specific

H. M. SELIM

3 76

sites are not easily replaceable by Ca ions, but can be replaced (exchanged) by competing (specifically sorbed) ions such as Ni, Cd, Co, and Zn. Figure 25 shows miscible displacement results of Cd in a Windsor sandy loam soil. A solution of 0.005 M Ca(N03)2 was applied to a packed soil column at a Darcy flow velocity of 0.271 m/day. A pulse of Cd(N03)2 dissolved in 0.005 M Ca(NO& solution was applied to the soil column and the effluent solution was collected and analyzed for Cd. The concentration of Cd in the input pulse was 100 mg/liter, therefore a small change in the total cationic concentration occurred from 10 mmol,/liter for the background solution to 11.786mmol,/liter for the Cd pulse. The use of the competitive transport model, assuming equilibrium conditions and ignoring specific sorption (kf= 0), resulted in an overestimation of the peak concentration and the BTC was more retarded than that experimentally measured (curve A, Fig. 25). For this model prediction, a selectivity coefficient (KCdCa)of 2, indicating a slight preference of Cd over Ca to the exchange surfaces, was used. This value was based on experimental measurements for a Cd-Ca sorption isotherm. All other parameters were experimentally measured or estimated using other experimental measurements, e.g., the dispersion coefficient was estimated using 3H20 BTCs. The use of the kinetic ion exchange mechanism along with specific sorption appears to provide an improved prediction of miscible displacements for Cd as shown by curves B-D in Fig. 25. Although the use of smaller values for a provided improved prediction of the observed tailing, BTCs were less retarded in comparison to the measured BTC. It is obvious that additional evaluation of this approach is needed for the prediction of the retention of other ions during transport in the soil profile.

h

u"

Cd Breakthrough Curve, 100 mglliter - Windsor Soil

0.8-

.-..-

5

PORE VOLUME (V/V,) Figure 25. Measured (closed circles) and predicted BTCs for Cd in Windsor soil. Prediction using equilibrium ion exchange model is shown by curve A. Curves B, C, and D are calculations using the kinetic ion exchange model with a of 2 day-' and specific sorption (kf) values of 0, 0.5, and 1.0 day-', respectively.

MODELS OF INORGANICS IN SOILS

377

VII. TRANSPORT IN LAYERED SOIL None of the mathematical models presented thus far deals with the problem of the fate of reactive or noreactive solutes in nonhomogeneous or layered soils. Because soil profiles are seldom uniform, it is essential to consider the fate of dissolved chemicals in stratified soil systems. The study of Shamir and Harleman (1967) is one of the earliest papers dealing with nonreactive solute transport through layered porous media having great depths ( z --* w). Others, including Rubin and James (1973), Selim et al. (1977), Selim (1978), and Barry and Parker (1987), considered the fate of reactive solutes in layered soils under various flow and boundary conditions. A schematic diagram of a soil of length L with three distinct layers I, 11, and I11 is shown in Fig. 26. Each layer has specific but not necessarily the same 0,p, and solute retention characteristics. First, we consider the case of solute transport in layered soils wherein fully saturated conditions under constant Darcy’s flux (steady flow) prevail. This is followed by cases for water-unsaturated layered soils under steady and transient water flow conditions.

A. SATURATED SOILSUNDER CONSTANT FLUX Simulations were carried out using equilibrium linear and nonlinear (Freundlich) [see Eq. (15)] as well as first-order kinetic (reversible and

.!

T

Ll

I

Figure 26. Schematic diagram of a three-layered soil.

378

H. M. SELIM

irreversible) retention [see Eqs. (19) and (23)] to describe solute behavior in each layer of a multilayered soil profile. Different simulations were conducted to evaluate the importance of soil layer stratification and adsorption characteristics on the shape and position of effluent concentration distributions layered _ _ from _ ~ ~ soil profiles. For the linear case, a retardation factor R [ R = 1 + p K d / O ;see Eq. (29)] represents the magnitude of retention for each layer and is represented by R 1 , R 2 , etc. (Selim et al., 1977). Solid lines in Fig. 27 are simulated BTC results from columns in which the solute passed first through L1 and then L 2 ; open circles are calculated results of solute flow in the opposite direction. The dashed lines are for the homogeneous cases wherein L = L1 or L = L2 and these cases have the appropriate retardation factors, R1 or R 2 , respectively. As expected, the BTCs for the two-layered cases lie between the homogeneous cases R1 and R2 (dashed lines). The retardation factor R1 equals one and represents a nonreactive solute whereas R2 equals 10. Increasing L2 (or decreasing L,) causes the breakthrough curves to move to the right toward the homogeneous R2 case. The most striking result in Fig. 27 is the failure of the order of soil layers to influence the shape or position of the effluent concentration distribution. Based on these results, a layered soil profile could be regarded as homogeneous with an average retardation factor used to calculate emuent concentration distributions. An average retardation factor R for N-layered soil can simply be obtained from

0

2

4

6

8

v

/

10

12

14

16

v.

Figure 27. Simulated effluent concentration distribution for two-layered soils with retardation factors R, and RZ (linear sorption) and varying lengths L , and Lz. Solid lines are simulations wherein layer 1 with R, is first encountered, whereas open circles are for flow in the reverse direction. Dashed curves are for homogeneous soils. From Selim et al. (1977), with permission.

MODELS OF INORGANICS IN SOILS

3 79

BTCs identical to those in Fig. 27 were obtained using the solution to the convention-dispersion Eq. (7) presented by Lindstrom et al. (1967) and an average retardation factor. This averaging procedure [Eq. (65)] can also be used to describe the BTCs from a soil profile composed of three or more layers (see Selim et al., 1977). However, if solute distribution within the profile is desired, the use of an average retardation factor is no longer valid and the problem must be treated as a multilayered case. The nonlinear (Freundlich) equilibrium and first-order kinetic retention were also considered for a two-layered soil column with L1= L2 = 0.5L (Selim ef al., 1977). The BTCs were identical regardless of the sequence of soil layers (not shown in Fig. 27). This is a significant conclusion and can be used to simplify field problems involving solute movement through nonhomogeneous field soils under steady water-saturated flow conditions. Unlike the linear adsorption cases, an average retardation factor R for nonlinear adsorption cannot be obtained, because of the dependency of R on the solute concentration C.

B. WATER-UNSATURATED SOILS Selim et al. (1977) simulated solute transport through water-unsaturated multilayered soil profiles, in which a steady, vertically downward water flow (q = constant) was considered. A soil profile was assumed to consist of two distinct layers, sand and clay, each having equal lengths, and was underlain by a water table at a depth L = 100 cm. The case for which the water table was at great depth (z 4 m) was also considered. When a constant flux was assumed, the steady state 0 and water suction (h) distributions for a sand-clay and a clay-sand soil profile were calculated (see Fig. 28 for the clay-sand case). Solute concentration versus pore volume of effluent (collected at a 100-cm depth) for a nonreactive and reactive solute having linear (equilibrium) retention is shown in Fig. 29. As expected, similar BTC results for the nonreactive solute for sand-clay or clay-sand soil profiles were obtained. In contrast, BTCs for the reactive solute show a distinct separation, with lower retardation factors for the soil profiles having a water table at z = 100 cm than at z + m. This observation is consistent for the sand-clay as well as clay-sand profiles. Due to the higher water contents in the soil profiles, wherein the water table was at x = 100, the retardation factor R is less in comparison to the case for which the water table was at great depth ( z + m). If the water content distributions were considered uniform, with an average water content within each individual layer (see Fig. 28), the problem of solute transport and retention through unsaturated multilayered soil profiles can be significantly simplified, as discussed in the

H. M. SELIM

3 80

Water Suction h, cm 40 60 80 100

A

6

20

Water Suction h, cm

0

5 2o 5 n

, I

,

/

/

/

__

100

40

#.-

60

53

80 100

0 0.1 0.2 0.3 0.4 0.5

I

0

.

0.1 0.2 0.3 0.4 0.5 Water Content e,cm3/cm3

Water Content 8,cm 31cm 3 Figure 28. Simulated water content 0 and water suction h versus depth in a clay-sand profile having a water table at (A) 100-cm depth and (B) great depth. From Selim et al. (1977), with permission.

previous section. The open circles in Fig. 29 are calculated results of concentration distributions for the reactive and nonreactive solutes when an average water content within each layer was used. These results show that, for all unsaturated profiles considered, the use of average water contents (open circles) provided concentration distributions identical to those obtained when the actual water content distributions were used (dashed and solid lines). Thus, when a steady water flux is maintained

1.0

0.8

0" 0.6

.

u

water table at x = 1 0 0 c m

0.4

0.2

0 0

1

2

3

4

5

6

7

8

v I v, Figure 29. Simulated effluent concentration distribution for reactive and nonreactive solute in an unsaturated clay-sand profile. Open circles are simulations based on average water content 0 for each soil layer. From Selim et al. (1977), with permission.

MODELS OF INORGANICS IN SOILS

381

through the profile, BTCs of reactive and nonreactive solutes at a given location in the soil profile can be predicted with average water contents within unsaturated soil layers. Based on the above results we can conclude that average microhydrologic characteristics for a soil layer can be used to describe the movement of solutes leaving a multilayered soil profile. This conclusion supports the assumption made earlier that uniform soil water content can be used to represent each soil layer in order to simplify the solute transport problem. However, such a simplifying approach was not applicable for the general case of transient water flow conditions of unsaturated multilayered soils. As illustrated by Selim (1978), the transport of reactive, as well as nonreactive, solutes through multilayered soils, for transient water flow, was significantly influenced by the order in which the soil layers were stratified.

REFERENCES Amacher, M. C., Kotuby-Amacher, J., Selim, H. M., and Iskandar, I. K. (1986). Retention and release of metals by soils-evaluation of several models. G e o d e m a 38, 131-154. Amacher, M. C., Selim, H. M . , and Iskandar, I. K. (1988). Kinetics of chromium (VI) and cadmium retention in soils: A nonlinear multireaction model. Soil Sci. SOC. Am. J . 52, 398-408. Amacher, M. C., Selim, H. M., and Iskandar, I. K. (1990). Kinetics of mercuric chloride retention in soils. J . Environ. Qual. 19, 382-388. Aringhieri, R., Carrai, P., and Petruzzelli, P. (1985). Kinetics of Cu and Cd adsorption by an Italian soil. Soil Sci. 139, 197-209. Barrow, N. J. (1989). Suitability of sorption-desorption methods to simulate partitioning and movement of ions in soils. Ecol. Stud. 74, 3-17. Barry, D. A., and Parker, J. C. (1987). Approximation of solute transport through porous media transverse to layering. Tramp. Porous Media 2, 65-82. Brenner, H. (1962). The diffusion model of longitudinal mixing in beds of finite length. Numerical values. Chem. Eng. Sci. 17, 220-243. Buchter, B., Davidoff, B., Amacher, M. C., Him, C., Iskandar, I. K., and Selim, H. M. (1989). Correlation of Freundlich Kd and n retention parameters with soils and elements. Soil Sci. 148, 370-379. Cederberg, G. A,, Street, R. L., and Leckie, 0. J. (1985). A groundwater mass transport and equilibrium chemistry model for multicomponent systems. Water Resour. Res. 21, 1095-1104. Cho, C. M. (1971). Convective transport of ammonium with nitrification in soil. Can. J . Soil Sci. 51, 339-350. Coats, K. H., and Smith, B. D. (1964). Dead-end pore volume and dispersion in porous media. SOC.Pet. Eng. 4, 73-84. Crank, J. (1956). “Mathematics of Diffusion.” Oxford Univ. Press. London. Danckwerts, P. V. (1953). Continuous flow systems. Distribution of residence times. Chem. Eng. Sci. 2, 1-13. De Camargo, 0. A., Biggar, J. W., and Nielsen, D. R. (1979). Transport of inorganic phosphorus is an Alfisol. Soil Sci. SOC.Am. J . 43, 884-890.

3 82

H. M. SELIM

Fiskell, J. G. A,, Mansell, R. S . , Selim, H. M., and Martin, G. G. (1979). Kinetic behavior of phosphate sorption by acid sandy soil. J . Environ. Qual. 8, 579-584. Fliihler, H., Polomski, J. and Blaser, P. (1982). Retention and movement of fluoride in soils. J . Environ. Qual. 11, 461-468. Fried, J. J., and Combarnous, M. A. (1971). Dispersion in porous media. A d v . Hydrosci. 7, 169-282. Gaston, L. A., and Selim, H. M. (1990). Transport of exchangeable cations in an aggregated clay soil. Soil Sci. SOC. Am. J. 54, 31-38. Helfferich, F. (1962). “Ion Exchange.” McGraw-Hill, New York. Helfferich, F., and Klein, G. (1970). “Multicomponent Chromatography,” pp. 1-147. Dekker, New York. Holford, I. C. R., and Mattingly, G. E. G. (1975). The high- and low-energy phosphate adsorption surfaces in calcareous soils. J . Soil Sci. 26,407-417. Holford, I. C. R., Wedderburn, R. W. M., and Mattingly, G. E. G. (1974). A Langmuir two-surface equation as a model of phosphate adsorption by soils. 1. Soil Sci. 25, 242-254. Jardine, P. M., and Sparks, D. L. (1984). Potassium-calcium exchange in a multireactive soil system. I. Kinetics. Soil Sci. SOC. Am. J . 48, 39-45. Jardine, P. M., Parker, J. C., and Zelazny, L. W. (1985). Kinetics and mechanisms of aluminum adsorption on kaolinite using a two-site nonequilibrium transport model. Soil Sci. SOC.Am. J . 49, 867-873. Jennings, A. A. (1987). Critical chemical reaction rates for multicomponent groundwater contamination models. Water Resour. Res. 23, 1775-1784. Jennings, A. A., and Kirkner, D. J. (1984). Instantaneous equilibrium approximation analysis. 1. Hydraul. Div., A m . SOC.Civ. Eng. 110, 1700-1717. Jennings, A. A., Kirkner, D. J., and Theis, T. L. (1982). Multicomponent equilibrium chemistry in groundwater quality models. Water Resour. Res. 18, 1089-1096. Kirkner, D. J., Jennings, A. A . , and Theis, T. L. (1985). Multisolute mass transport with chemical interaction kinetics. J . Hydrol. 76, 107-117. Kreft, A. and Zuber, A. (1978). On the physical meaning of the dispersion equation and its solutions for different initial and boundary conditions. Chem. Eng. Sci. 33, 1471-1480. Lai, S.-H., and Jurinak, J. J. (1972). Cation adsorption in one dimensional flow through soils. A numerical solution. Water Resour. Res. 8, 99-107. Lai, S.-H., Jurinak, J. J., and Wagenet, R. J. (1978). Multicomponent cation adsorption during convection-dispersive flow through soils. Experimental study. Soil Sci. SOC.Am. J . 42, 240-243. Langmuir, I. (1918). The adsorption of gases on plane surfaces of glass, mica and platinum, J . A m . Chem. SOC.40, 1361-1402. Lapidus, L., and Amundson, N. R. (1952). Mathematics of adsorption in beds. M.The effect of longitudinal diffusion in ion exchange and chromatographic column. J . Phys. Chem. 56, 984-988. Lindstrom, F. T., Hague, R. Freed, V. H., and Boersma, L. (1967). Theory on movement of some herbicides in soils. Linear diffusion and convection of chemicals in soils. Environ. Sci. Technol. 1, 561-565. Mansell, R. S., Selim, H. M., Kanchanasut, P., Davidson, J. M., and Fiskell, J. G. A. (1977). Experimental and simulated transport of phosphorus through sandy soils. Water Resour. Res. 13, 189-194. Mansell, R. S . , Bloom, S . A., Selim, H. M., and Rhue, R. D. (1988). Simulated transport of multiple cations in soil using variable selectivity coefficients. Soil Sci. SOC.A m . J. 52, 1533- 1540. Miller, C. W., and Benson, L. V. (1983). Simulation of solute transport in a chemically

MODELS OF INORGANICS IN SOILS

3 83

reactive heterogeneous system. Model development and application. Water Resour. Res. 19, 381-391. Munns, D. N., and Fox, R. L. (1976). The slow reactions which continues after phosphate adsorption; kinetics and equilibrium in some tropical soils. Soil Sci. SOC. Am. J . 40, 46-51. Murali, V., and Aylmore, L. A. G. (1983). Competitive adsorption during solute transport in soils. 1. Mathematical models. Soil Sci. 135, 143-150. Nielsen, D. R., van Genuchten, M. Th., and Biggar, J. W. (1986). Water flow and solute transport processes in the unsaturated zone. Water Resour. Res. 22, 89s-108s. Nkedi-Kizza, P., Biggar, J. M., Selim, H. M., van Genuchten, M. Th., Wierenga, P. J., Davidson, J. M., and Nielsen, D. R. (1984). On the equivalence of two conceptual models for describing ion exchange during transport through an aggregated soil. Water Resour. Res. 20, 1123-1130. Ogwada, R. A., and Sparks, D. L. (1986). Kinetics of ion exchange on clay minerals and soil. I. Evaluation of methods. Soil Sci. SOC.Am. J. 50, 1158-1162. Ozisik, M. N. (1968). “Boundary value Problems of Heat Conduction.” International Textbooks, Scranton, Pennsylvania. Parker, J. C., and Jardine, P. M. (1986). Effect of heterogeneous adsorption behavior on ion transport. Water Resour. Res. 22, 1334-1340. Rasmuson, A., and Neretienks, I. (1981). Migration of radionuclides in fissured rock. The influence of micropore diffusion and longitudinal dispersion. J . Geophys. Res. 86,37493758. Rubin, J. (1983). Transport of reactive solutes in porous media. Relation between mathematical nature of problem formulation and chemical nature of reactions. Water Resour. Res. 19, 1231-1252. Rubin, J., and James, R. V. (1973). Dispersion-affected transport of reacting solution in saturated porous media. Galerkin method applied to equilibrium-controlled exchange in unidirectional steady water flow. Water Resour. Res. 9, 1332-1356. Schmidt, H. W., and Sticher, H. (1986). Long-term trend analysis of heavy metal content and translocation in soils. Geoderma 38, 195-207. Schulin, R., Fliihler, H., Mansell, R. S . , and Selim, H. M. (1986). Miscible displacement of ions in aggregated soils. Geoderma 38, 311-322. Schulin, R., Papritz, A., Fluhler, H., and Selim, H . M. (1989). Calcium and magnesium transport in aggregated soils at variable ionic strength. Geoderma 44, 129-141. Selim, H. M. (1978). Transport of reactive solutes during transient unsaturated water flow in multilayered soils. Soil Sci. 126, 127-135. Selim, H. M. (1981). Modeling kinetic behavior of cadmium interaction in soils. I n “Summer Computer Simulation Conference, Washington, D. C., 1981,”pp. 385-390. Simulation Councils Inc., La Jolla, California. Selim, H. M. (1989). Prediction of contaminant retention and transport in soils using multireaction models. Environ. Health Perspect. 39, 69-75. Selim, H. M., and Amacher, M. C . (1988). A second order kinetic approach for modeling solute retention transport in soils. Water Resour. Res. 24, 2061-2075. Selim, H. M., and Iskandar, I. K. (1981). Modeling nitrogen transport and transformations in soils. 1. Theoretical considerations. Soil Sci. 131, 233-241. Selim. H. M., and Mansell, R . S. (1976). Analytical solution of the equation of reactive solutes through soils. Water Resour. Res. 12, 528-532. Selim, H. M., Davidson, J. M., and Mansell, R. S . (1976). Evaluation of a 2-site adsorptiondesorption model for describing solute transport in soils. I n “Summer Computer Simulation Conference, Washington, D. C., 1976” pp. 444-448. Simulation Councils Inc., La Jolla, California.

3 84

H. M. SELIM

Selim, H. M., Davidson, J. M., and Rao, P. S. C. (1977). Transport of reactive solutes through multilayered soils. Soil Sci. SOC.Am. J . 41, 3-10. Selim, H. M., Schulin, R., and Fliihler, H. (1987). Transport and ion exchange of calcium and magnesium in an aggregated soil. Soil Sci. SOC.Am. J . 51,876-884. Selim, H. M., Amacher, M. C., and Iskandar, I. K. (1989). Modeling the transport of chromium (VI)in soil columns. Soil Sci. SOC. Am. J . 53, 996-1004. Selim, H. M., Amacher, M. C., and Iskandar, I. K. (1990a). “Modeling the Transport of Heavy Metals in Soils,” CRREL-Monogr. 90-2. U. S. Army Corps of Engineers, Washington. D. C. Selim, H. M., Amacher, M. C., Persaud, N., and Mansell, R. S. (1990b). Retention and transport of phosphorus in soils; a multireaction approach. I n “Summer Computer Simulation Conference, Calgary, Alberta, 1990,” pp. 596-603. Simulation Councils Inc., La Jolla, California. Shamir, U. Y., and Harleman, D. R. F. (1967). Dispersion in layered porous media. J. Hydraul. Div., Am. SOC. Civ. Eng. 93, 237-260. Skopp, J. (1986). Analysis of time-dependent chemical processes in soils. J. Environ. Qunl. 15, 205-213. Sparks, D. L. (1989). “Kinetics of Soil Chemical Processes.” Academic Press, San Diego, California. Sposito, G. (1980). Derivation of the Freundlich equation for ion exchange reactions in soils. Soil Sci. SOC. Am. J . 44,652-654. Sposito, G. (1981). “The Thermodynamics of Soil Solutions.” Oxford Univ. Press, New York. Sposito, G. (1984). “The Surface Chemistry of Soils.” Oxford Univ. Press, New York. Theis, T. L. (1988). Reactions and transport of trace metals in groundwater. In “Metal Speciation: Theory, Analysis and Application” (J. R. Kramer and H. E. Allen, eds.), pp. 81-89. Lewis, Chelsea, Michigan. Theis, T. L., Iyer, R., and Kaul, L. W. (1988). Kinetic studies of cadmium and ferricyanide adsorption on goethite. Environ, Sci. Technol. 22, 1032-1017. Tiller, K. G., Nayyar, V. K., and Clayton, P. M. (1979). Specific and nonspecific sorption of cadmium by soil clays as influenced by zinc and calcium. Aust. J. Soil Res. 17, 17-28. Tiller, K. G., Gerth, J., and Briimmer, G. (1984). The relative affinitiesof Cd, Ni, and Zn for different soils clay fractions. Procedures and partitioning of bound forms and their interpretations. Geoderrna 34, 1-16. Travis, C. C., and Etnier, E. L. (1981). A survey of sorption relationships for reactive solutes in soil. J . Environ. Qual. 10, 8-17. Valocchi, A. J. (1985). Validity of the local equilibrium assumption for modeling sorbing solute transport through homogeneous soils. Water Resour. Res. 21, 808-820. Valocchi, A. J., Street, R. L., and Roberts, P. V. (1981). Transport of ion-exchange solutes in groundwater. Chromatographic theory and field simulations. Water Resour. Res. 17, 1517-1527. Van Eijkeren, J. C. M., and Loch, J. P. G. (1984). Transport of cation solutes in sorbing porous media. Water Resour. Res. 20, 714-718. van Genuchten, M. Th., and Alves, W. J. (1982). Analytical solutions of the onedimensional convective-dispersive solute transport equation. U.s., Depr. Agric., Bull. 1661, 1-151. van Genuchten, M. Th., and Wierenga, P. J. (1976). Mass transfer studies sorbing porous media I. Analytical solutions. Soil Sci. SOC. Am. J. 40,473-480.

Index A Abscission, corn, 209-210, 227 Absorption matrix, fingerprinting crop varieties, 89 Acid exchange, deposition on forested soils, 40 Acid-neutralizing capacity (ANC), deposition on forested soils, 4-6 forest soils, 12, 16, 18.24, 36,39 Acidic deposition on forested soils, 1-3. 65-66 chemical factors cation exchange, 19-29 hydrogen budgets, 38-40 mineral weathering, 35-38 nitrate retention, 29-32 soil organisms, 40-42 soil pH, 15-19 sulfate retention, 29, 32-35 future research, 63-65 physical factors canopy interactions, 7-9 hydrology, 11-15 soil horizons, 9-11 soil acidification, 3-6 soil change studies intensity-type changes, 44-48 leaching, 48-56,61-63 measurement, 42-43 seasonal variations, 59-61 uptake, 56-59,61-63 Acidification, soil, see Acidic deposition on forested soils Activity coefficient inorganics in soils, 368 surface complexation models, 234-235, 238 Adaptation, corn, 206-207,214,222 Adjustable parameters, surface complexation models, 236-237 Adsorption acidic deposition on forested soils, 33, 54-56,64

chemical transport through soil processes, 161-165.169, 171-175 solute transport, field studies of, 182-185, 188 inorganics in soils, 332, 340 ion exchange, 369,372-376 kinetic retention models, 345-347, 349-350 layered soil, 378-379 multiple-reaction models, 351, 358-359 surface complexation models competitive adsorption reactions, 312-318 description, 235,235-238.241. 243-248,320 inorganic anion adsorption, 289-307 metal ion adsorption, 274-289 organic ligand adsorption, 308-312 protonation-dissociation reactions, 5 1, 256-257,261,266,274 Adsorption density, surface complexation models, 283 Adsorption envelopes, surface complexation models, 289,301,308-309 Adsorption isotherm models, chemical transport through soil, 161 Adsorptive additivity, surface complexation models, 278 Air pollution, acidic deposition on forested soils, 63 Alleles corn, 208,211-212,227 fingerprinting crop varieties, 98, 105 Alumino hydroxy sulfate, acidic deposition on forested soils, 33-34 Aluminum acidic deposition on forested soils chemical factors, 16, 19-29.31.34-35 future research, 63 hydrogen budgets, 40 mineral weathering, 37-38 physical factors, 14-15 soil change studies, 43,45-46.58-59, 61,63

385

Aluminum (conrinued) soil organisms, 42 inorganics in soils, 352-353 surface complexation models, 276,278, 300-302,315 toxicity, acidic deposition on forested soils, 65 Aluminum oxide acidic deposition on forested soils, 32-34 inorganics in soils, 351 surface complexation models, 237 inorganic anion adsorption, 289-291, 297.300-301,305 metal ion adsorption, 275, 279 organic ligand adsorption, 308 protonation-dissociation reactions, 252-254,261,265 Aluminum trihydroxide, acidic deposition on forested soils, 22, 26 Amino acids fingerprinting crop varieties, 89 surface complexation models, 308,310 Amplification, fingerprinting crop varieties, 100,126 Anatase, surface complexation models, 252, 261 Anion exchange, acidic deposition on forested soils, 34 Anions acidic deposition on forested soils, 55-56, 61,64 inorganics in soils, 350 surface complexation models, 242,244, 249,266,320 competitive adsorption reactions, 314-316,318 inorganic anion adsorption, 289-307 Antibodies, fingerprinting crop varieties, 89 Antigens, fingerprinting crop varieties, 89 Apparent dispersivity, chemical transport through soil, 181-182 Apples, fingerprinting crop varieties, 94, 97 Arsenate, surface complexation models, 314-316,318 Arsenic, surface complexation models, 291, 295 Arsenite, surface complexation models, 305, 307,318 Artificial selection, corn, 222 Asymptotic dispersion models, chemical transport through soil, 151-154

Atrazine, chemical transport through soil, 184, 187-188 Autoradiograms, fingerprinting crop varieties, 103-104

B Back-crossing, fingerprinting crop varieties, 108, 112,121-122 Barley, fingerprinting crop varieties, 94, 113-114, 123 Base cations, acidic deposition on forested soils chemical factors, 28, 31.36-37 soil change studies, 54,57, 61 Base saturation (BS), acidic deposition on forested soils, 19-20.26 Bicarbonate salts, acidic deposition on forested soils, 20 Bidentate metal complex, surface complexation models, 241, 247,281, 290,295, 308 Biological transformation, chemical transport through soil, 168-169 Block inheritance, corn, 211 Boehmite, surface complexation models, 250, 253-254.256 Borate, surface complexation models, 304-305 Boron, surface complexation models, 243. 290-291,300,303 Boundary conditions, inorganics in soils, 335-336 Branching habit, corn, 219,221,226 Brassicas, fingerprinting crop varieties, 123, 125 Breakthrough curves, inorganics in soils ion exchange, 376,378-379 layered soil, 381 multiple-reaction models, 352, 356, 364-365 Bromide, chemical transport through soil, 176, 178,183-185,187 Buffer capacity, acidic deposition on forested soils, 18-19 Buffering mechanisms, acidic deposition on forested soils, 16, 35 Buffering ranges, acidic deposition on forested soils, 20, 26, 31,40, 66 Bulk density, inorganics in soils, 332, 334

INDEX C Cadmium inorganics in soils ion exchange, 368, 375-376 kinetic retention models, 342 multiple-reaction models, 350-351, 353 transport equations, 339-340 surface complexation models competitive adsorption reactions, 312-313,318 computer codes, 319 metal ion adsorption, 278,282,285, 287-288 Calcium acidic deposition on forested soils chemical factors, 17.20-21,24,26,28 mineral weathering, 36-37 physical factors, 7 , 9 soil change studies, 45, 52,54,56-62 deficiency, acidic deposition on forested soils, 29 inorganics in soils, 367-373, 376 surface complexation models, 282,285, 305,312-313,318 Calcium carbonate, surface complexation models, 285, 305 Calcium nitrate, surface complexation models, 285-286 Canopy, acidic deposition on forested soils, 7-9,28,31-32,41 Capacitance density, surface complexation models, 236,238 Capacity changes, acidic deposition on forested soils leaching, 48-56 uptake, 56-59 Capacity factors, acidic deposition on forested soils, 5-6, 66 Carbon acidic deposition on forested soils, 9, 30, 4 1-42 surface complexation models, 315 Carbon dioxide, acidic deposition on forested soils, 18 Carbonate, surface complexation models, 315, 317 Carbonic acid, deposition on forested soils, 3, 56,66 forest soils, 16, 18-19, 27-28

387

Cation exchange acidic deposition on forested soils, 66 chemical factors, 19-29, 31,38 soil change studies, 42,45,55.57, 59-63 inorganics in soils, 369-372 Cation exchange capacity, acidic deposition on forested soils, 16, 37 Cations inorganics in soils, 350,367-368, 375 surface complexation models, 242, 244, 249,254 competitive adsorption reactions, 312, 318 metal ion adsorption, 282,286 Celery, fingerprinting crop varieties, 124-125 Charge balance, surface complexation models, 236,245-246 Charge-potential relations, surface complexation models, 240,245, 250 Chemical mass flux laws, chemical transport through soil, 145-154 Chemical phase concentrations, chemical transport through soil, 144-145 Chemical speciation models, 319 Chemical transport through soil, 142-143, 191-192 field regime dispersion, 193-194 preferential Row, 192 rate processes, 193 history, 194-195 processes, 143-144 chemical mass flux laws, 145-154 chemical phase concentrations, 144-145 concentrations, 160-161 convection-dispersion, 154-156 dispersion transport, 169-171 interphase mass transfer laws, 161-167 mass conservation law, 144 react ions, 167- I69 solute adsorption, 171-175 transfer function models, 156-160 solute transport, field studies of, 175-176 indices of solute disperion, 189-191 mean solute velocity, 176-179 preferential flow, 185-188 scale of heterogeneity, 189 solute adsorption, 182-185 solute dispersion, 179-182 CHEMTRAN, inorganics in soils, 367-368

388

INDEX

Chloride chemical transport through soil, 176,181, 183-184, 186 surface complexation models, 266 Chromate, surface complexation models, 300-303,315,318 Chromatography fingerprinting crop varieties characters, 89,95-96,98 discrimination, 110 usage, 123,125 inorganics in soils, 367 Chromium, inorganics in soils, 340, 350-351,353,364,367 Chromosomes corn, 207-209,214,219,227 chromosome 4,210-212.227 fingerprinting crop varieties, 102, 124, 126 Citrate, surface complexation models, 304, 312 Clay minerals inorganics in soils, 332 surface complexation models, 320 inorganic anion adsorption, 289-291, 295 metal ion adsorption, 275,277 protonation-dissociation reactions, 254-256,267-268 Climate, acidic deposition on forested soils, 63 Clones, fingerprinting crop varieties, 100-101,103-104 Cloudwater, acidic deposition on forested soils, 7-8 Cobalt inorganics in soils, 339, 375-376 surface complexation models, 278, 286, 318 Coleoptile isozymes, fingerprinting crop varieties, 112-113, 115 Competitive adsorption reactions, surface complexation models, 312-318 Competitive transport models, inorganics in soils, 376 Complementary DNA, fingerprinting crop varieties, 99-100 Computer codes, surface complexation models, 319-320 Computer programs, surface complexation models, 279,290,297,319-320, see also FITEQL Concentration, inorganics in soils, 369, 380

Concentration outflow, chemical transport through soil, 172 Concentration-depth curves, chemical transport through soil, 169 Concentration-time curves, chemical transport through soil, 169 Condensation, corn, 209,228 Conditional equilibrium constants, surface complexation models, 236,238-240, 242-244,248 competitive adsorption reactions, 316-3 17 inorganic anion adsorption, 290,297,300, 305 metal ion adsorption, 275, 279 organic ligand adsorption, 308 protonation-dissociation reactions, 25 1, 257 Constant capacitance model, 234,320 competitive adsorption reactions, 312, 314-317 computer codes, 319 description, 238-240,242-249 inorganic anion adsorption, 289-297 metal ion adsorption, 274-279 organic ligand adsorption, 308-310 protonation-dissociation reactions, 251-256,267 Convection chemical transport through soil, 193-195 processes, 147-151,156,159-160, 175 solute transport, field studies of, 178, 187 inorganics in soils, 334 Convection-dispersion model, chemical transport through soil, 152-156,166, 176, 189 Convective-dispersive equation, inorganics in soils equilibrium retention models, 337 ion exchange, 367,369-370, 373 kinetic retention models, 347 layered soil, 379 multiple-reaction models, 351-352,355, 361 transport equations, 335-336 Convective-dispersive flux, chemical transport through soil, 154-156, 158, 161, 195 Convective-lognormal transfer function, chemical transport through soil, 159-160

INDEX Copper inorganics in soils, 339-340, 342, 351 surface complexation models, 281,283, 286,312-313 Corn, evolution of, 203-205, 224,227-228 husk enclosure of ear, 222-224 increased femaleness, 225 interpathway heterosis, 221-222 multiple domestications, 21 4-215 cob morphology, 215-219 plant habit, 219-221 partitioning of photosynthate, 225-226 transformation, 205-207 domestication, 212-214 isolation, 211 key traits, 207-211 tunicate locus, 212 Cross-overs, corn, 211-212.227 Crossing fields, corn, 226 CSTT theory, corn, 213-215, 224 Cucumbers, corn, 224 Cupules. corn, 209-210,217,219,222, 227 Cycling, nutrient, acidic deposition on forested soils, 10, 31-32, 34, 63 Cytoplasm, fingerprinting crop varieties, 100

D Deacidification, deposition on forested soils, 32,34 Degradation, chemical transport through soil, 168 Desorption acidic deposition on forested soils, 56 chemical transport through soil, 164, 172 inorganics in soils, 332 ion exchange, 373,375 kinetic retention models, 345-347 multiple-reaction models, 350, 356 surface complexation models, 266, 282 Diffuse double-layer theory, surface complexation models, 236, 245, 250 Diffuse layer, surface complexation models, 260, 267,319 Diffuse layer charge, surface complexation models, 246-247,260 Diffusion chemical transport through soil, 193 processes, 147-149, 151-152,155, 166 processes, analysis of, 173-174 inorganics in soils, 334-335,350,359, 361 surface complexation models, 284

389

Diffusion-controlled model, inorganics in soils, 350 Diffusion-dispersion coefficient, chemical transport through soil, 156 Diffusive mass transfer, chemical transport through soil, 174 Digital image analysis, fingerprinting crop varieties, 87-88, 103 Discrimination, fingerprinting crop varieties ability, 108-110 new techniques, 125, 127 usage, 123 Dispersion, see also Convection-dispersion model; Convective-dispersive equation chemical transport through soil field regime, 193-195 processes, 152, 154-156, 158, 161, 167 solute transport, field studies of, 176, 179-182, 187,189-191 inorganics in soils, 334-336, 361-362 Dispersion coefficient, chemical transport through soil, 181 Dispersion models, asymptotic, chemical transport through soil, 151-154 Displacement, inorganics in soils, 357, 359, 364,372,376 Dissociation acidic deposition on forested soils. 17 surface complexation models, 234, 320 inorganic anion adsorption, 290, 301-302 metal ion adsorption, 286 protonation-dissociation reactions, 25 1-274 Dissolution chemical transport through soil, 165-166, 171,193 inorganics in soils, 332, 347, 354 Distance measures. fingerprinting crop varieties, 104-105 Distinctness, fingerprinting crop varieties, 119-120 Distribution coefficient chemical transport through soil, 163 inorganics in soils, 337 DNA, fingerprinting crop varieties, 86 characters, 88-89, 98-108 discrimination, 110, 112-118 new techniques, 125-127 usage, 119-121,124-125 Domestication of corn, 226,228 interpathway heterosis, 221-222

390

INDEX

Domestication of corn (continued) origin, 214-215 cob morphology, 215-219 plant habit, 219-221 tqansformation,205-2 12 time required, 212-214 Drainage water, acidic deposition on forested soils, 65 chemical factors, 39-40 physical factors, 11-12, 14 Dry deposition, acidic deposition on forested soils, 7, 35,40,48 Dual velocity models, chemical transport through soil, 167

E Effective dispersion coefficient, chemical transport through soil, 155 Electric double-layer theory, surface complexation models, 247 Electrical neutrality, acidic deposition on forested soils, 30 Electrolytes, surface complexation models, 235 description, 235,238,240,242,244 inorganic anion adsorption, 300 metal ion adsorption, 277,286,288 protonation-dissociation reactions, 254-255,267,272-273 triple-layer models, 256-258, 260-261, 263-264,266 Electrophoresis, fingerprintingcrop varieties characters, 89,95-99, 101 discrimination, 110-115 new techniques, 125-126 usage, 119-121,123,125 Electrostatic potential terms, surface complexation models, 236,246,320 Elongation, corn, 221,223,226-227 Endonucleases, restriction, fingerprinting crop varieties, 98, 100-101, 126 Environment, fingerprinting crop varieties, 87-88,95-97,120 Enzymes, see also Restriction enzymes fingerprintingcrop varieties, 89,98, 102, 104,109, 113 Equilibrium acidic deposition on forested soils, 45 inorganics in soils, 333,379 equilibrium retention models, 337-340

ionexchange, 367-369,371,374,376 kinetic retention models, 343,345-346 multiple-reaction models, 350, 352-353,360,363 Equilibrium adsorption, chemical transport through soil, 162, 165,172,174,185, 187 Equilibrium concentrations, chemical transport through soil, 168 Equilibrium constants, surface complexation models, 234-235 competitive adsorption reactions, 316-317 computer codes, 319 description, 236-237,239-240,242-245. 248,250 inorganic anion adsorption, 290,297,300, 305,307 metal ion adsorption, 275,277-279,282, 287-288 organic ligand adsorption, 308-309 protonation-dissociation reactions, 25 1, 254-255,257,267,273 Equilibrium partitioning, chemical transport through soil, 171 Equilibrium pressure, chemical transport through soil, 146 Equilibrium sorption, chemical transport through soil, 183 External-internal proton ratio (EIPR), acidic deposition on forested soils, 39-40

F Faraday constant, surface complexation models, 236, 246 FASTCHEM, surface complexation models, 320 Feedback corn, 223,228 surface complexation models, 285 Femaleness, corn, 221-222 Feminization, corn, 222, 225 Ferric oxide, hydrous, surface complexation models, 270-271,286-287.305-307 Fertilizer, inorganics in soils, 332-333 Field regime, chemical transport through soil, 192-195 FIESTA, inorganics in soils, 367-368 Fingerprinting crop varieties, 85-86 characters DNA data, 98-108

391 morphology, 86-88 protein data, 88-98 discrimination, 108-1 18 new techniques, 125-127 usage, 119 genetic diversity, 122-123 genetic purity, 123 germplasm improvement, 123-124 minimum distance, 121-122 misappropriation, 124-125 plant variety protection, 119-121 FITEQL, surface complexation models computer codes, 319 inorganic anion adsorption, 290,297, 305-306 metal ion adsorption, 277,279,286-287 protonation-dissociation reactions, 254-255,265-266,270-271 Fluoride inorganics in soils, 347 surface complexation models, 237,303, 3 15 Flux concentration, chemical transport through soil, 157, 160-161 Flux of dissolved chemical in soil, 147-151 Forested soils, acidic deposition on, see Acidic deposition on forested soils Freundlich adsorption, chemical transport through soil, 171 Freundlich isotherm models, chemical transport through soil, 162-163, 171-172 Freundlich models, inorganics in soils, 333, 345 equilibrium retention models, 337-340 multiple-reaction models, 350, 352

G Gametophyte genes, corn, 210-211, 227 Gene expression, corn, 227-228 Generalized two-layer model, 234,320 description, 246-248,270-271 inorganic anion adsorption, 305-307 metal ion adsorption, 278, 286-287 Genes, corn, 227-228 husk enclosure of ear, 223-224 multiple domestications, 215,217,219,221 transformation. 206-210, 212 Genetic control, fingerprinting crop varieties, 96-97,105

Genetic distance, fingerprinting crop varieties, 87-88,102,104-105, 108 Genetic diversity corn, 207 fingerprinting crop varieties, 86 characters, 87, 96, 100-102 discrimination, 109-110 new techniques, 126-127 usage, 122-123 Genetic markers, fingerprinting crop varieties, 107, 125, 127 Genetic purity, fingerprinting crop varieties, 123 Genomic DNA, fingerprinting crop varieties, 101-102 Genotype, fingerprinting crop varieties characters, 87-88,96,101-102, 104-106, 108 discrimination, 109-1 15 usage, 119,121-124 GEOCHEM model, chemical transport through soil, 168 Geographic diversity, corn, 200,215 Geothite, inorganics in soils, 375 Germplasm corn, 207,211,219,225 fingerprinting crop varieties, 108-109, 123-124,126 Gibbsite acidic deposition on forested soils, 22, 25 surface complexation models, 314-315 Gliadin, fingerprinting crop varieties, 95, 111, 119-120, 124 Glitelins, fingerprinting crop varieties, 95, 111, 114-115 Globulin, fingerprinting crop varieties, 95-97, 114 Glume, corn, 210-212,227 Glutamate, surface complexation models, 308 Goethite, surface complexation models, 237 competitive adsorption reactions, 312, 314-315.318 inorganic anion adsorption, 290,297,300, 302-304,307 metal ion adsorption, 276, 279, 282-283, 285 organic ligand adsorption, 309,312 protonation-dissociation reactions, 251-254,256,258.266-268.274 Groundwater acidic deposition on forested soils, 13, 35

INDEX

392

Groundwater (conrinued) chemical transport through soil, 142, 154, 174, 187, 189 inorganics in soils, 331

H Harvesting acidic deposition on forested soils, 3, 31, 34,63 corn, 209,214 Hematite, surface complexation models, 288 Herbicides, chemical transport through soil, 182-183 Hermaphroditism, corn, 205,223-224 Heterogeneity chemical transport through soil, 143, 189, 191 inorganics in soils, 359,361 surface complexation models, 320 competitive adsorption reactions, 314 inorganic anion adsorption, 291 metal ion adsorption, 277,284,288 protonation-dissociation reactions, 255, 266-267,273-274 Heterosis fingerprinting crop varieties, 105, 122, 124 interpathway, corn, 221-222 High-performance liquid chromatography, reversed-phase, fingerprinting crop varieties, 95-96, 111-115 Hordeins, fingerprinting crop varieties, 113-114 Hormonelike control, corn, 223-224 Hormones, corn, 223,228 Humidity, acidic deposition on forested soils, 2, 16,18,63 Humus, acidic deposition on forested soils, 3, 64 Husk enclosure of corn ear, 222-224 Hybridization corn, 219,221-223.225-226 fingerprinting crop varieties characters, 97.99-101, 104-105 discrimination, 108-109, 113-114 new techniques, 126 usage, 122-124 HYDRAQL, surface complexation models, 297,319 Hydraulic conductivity, chemical transport through soil. 146, 154-155,158,191, 194

Hydrodynamic dispersion chemical transport through soil, 149 inorganics in soils, 334-335, 362 Hydrogen acidic deposition on forested soils, 1, 5 , 64-65 budgets, 38-40 chemical factors, 15-19,21-24,30, 33-34 mineral weathering, 36-37 physical factors, 7-9 soil change studies, 44,48,55 soil organisms, 40-42 inorganics in soils, 375 HYDROGEOCHEM, surface complexation models, 320 Hydrology acidic deposition on forested soils, 11-16, 27 chemical transport through soil, 149,191, 194 Hydrolysis acidic deposition on forested soils, 21, 24 chemical transport through soil, 168 inorganics in soils, 375 surface complexation models, 281,288 Hydrophobicity, fingerprinting crop varieties, 98 Hydrous ferric oxide, surface complexation models, 270-271,286-287.305-307 Hydroxyl groups, surface complexation models, 235-237,245,251,254,256 Hydroxyl ions, surface complexation models, 240,245,249,251,256 Hysteresis chemical transport through soil, 164, 172 inorganics in soils, 346-347

I Identical by descent, fingerprinting crop varieties, 105 Identity in state, fingerprinting crop varieties, 105 Illite, surface complexation models, 256, 267 Image analysis, fingerprinting crop varieties, 87-88,103 Immobile regions chemical transport through soil, 166, 173-174 inorganics in soils, 332 layered soil, 370-371, 373-374

INDEX multiple-reaction models, 359, 361-363, 365,367 Immobile water content, chemical transport through soil, 166,171 Immobilization, inorganics in soils, 347, 354 Infinite-time model of solute dispersion, chemical transport through soil, 152 Inflorescence, corn, 205,208,219,223-226 Inhibitors acidic deposition on forested soils, 45, 61 corn, 209-210,223-224 Inner-sphere surface complexes, surface complexation models, 238, 240,243, 245-246 competitive adsorption reactions, 313 inorganic anion adsorption, 300-301 metal ion adsorption, 282 Inorganics in soils, modeling of, 331-333 equilibrium retention models, 337 Freundlich, 337-340 Langmuir, 340 Langmuir two-site, 340-342 ion exchange, 367-368 classical approach, 368-370 mobile-immobile approach, 370-371 specific adsorption, 373-376 variable selectivities, 371-373 kinetic retention models, 342-350 layered soil, 377 constant flux, 377-379 water-unsaturated soils, 379-38 1 multiple-reaction models, 350-35 1 multireaction models, 353-358 second-order models, 358-367 two-site models, 351-353 transport equations, 333-336 Intensity acidic deposition on forested soils, 5-6, 44-48,66 Interpathway heterosis, corn, 221-222 Interphase transfer laws, chemical transport through soil, 161-167 Interspace trait, corn, 206-207 Introgression, corn, 210,214 Ion exchange, inorganics in soils, 332-333, 367-370 adsorption, 373-376 kinetic retention models, 349-350 mobile-immobile models, 370-371 multiple-reaction models, 360, 363 selectivities, 371-373

393

Ionic strength inorganics in soils, 360, 371 surface complexation models, 255,257. 260,265,270,300 Ions chemical transport through soil, 167, 184 inorganics in soils, 350 surface complexation models, 235, 320 competitive adsorption reactions, 312-313, 318 description, 235, 237-238, 240-242, 245-247 metal ion adsorption, 274-289 protonation-dissociation reactions, 25 1, 254,256,274 Iron acidic deposition on forested soils, 36 surface complexation models, 278, 287, 301-302 Iron oxide acidic deposition on forested soils, 33-34 inorganics in soils, 354 surface complexation models, 243 competitive adsorption reactions, 312-313, 315,317-318 inorganic anion adsorption, 289-291, 296-298,300,303,307 metal ion adsorption, 275-276, 278-279,281-282.287-288 organic ligand adsorption, 308-309 protonation-dissociation reactions, 251-253,267-268,270 triple-layer model, 256, 258-259, 265 Irrigation chemical transport through soil, 176, 178-179, 182-183, 186 corn, 205,207 Isoelectric focusing, fingerprinting crop varieties, 112, 114-115 Isoenzymes, fingerprinting crop varieties, 89-90,96-97 Isolation, corn, 205-207, 211 ISOQUAD, surface complexation models, 319 Isotherms chemical transport through soil, 161-165, 171-172 inorganics in soils equilibrium retention models, 339-341 ion exchange, 369,371-372,374,376 kinetic retention models, 343-345, 357 multiple-reaction models, 361, 363

INDEX

394 Isozymes corn, 215 fingerprinting crop varieties characters, 97-98, 103 discrimination, 109. 112-113, 115 usage, 120,122, 124-125

K Kaolinite acidic deposition on forested soils, 36 inorganics in soils, 349, 352 surface complexation models inorganic anion adsorption, 291,297, 301 metal ion adsorption, 275,277 protonation-dissociation reactions, 254-256 Kernels, corn, 207,212,225-227 exposure, 217-219 Key trait genes, corn, 207-211,227-228 Kinetic mass transfer, chemical transport through soil, 165 Kinetics inorganics in soils, 333,337-338,377 ion exchange, 368,373-376 kinetic retention models, 342-350 multiple-reaction models, 351-352, 354-456,358-364 surface complexation models, 266,282, 300-301.320

L Langmuir isotherms, chemical transport through soil, 162, 164 Langmuir models, inorganics in soils, 333, 337,340,350,360-361 Langmuir two-site model, inorganics in soils, 340-342.363 Layered soil, inorganics in soils, 377-381 Leaching acidic deposition on forested soils, 3-5 chemical factors, 18, 30, 32 future research, 64 physical factors, 13 soil change studies, 43,423-57,59, 61-63 chemical transport through soil, 167,176, 178,182-184 inorganics in soils, 331, 354,369

Lead inorganics in soils, 339-340 surface complexation models competitive adsorption reactions, 312-313,318 metal ion adsorption, 278,282-283, 286 Ligands, surface complexation models, 238, 240-241,243,248,320 competitive adsorption reactions, 314-318 inorganic anion adsorption, 291,300 organic ligand adsorption, 308-312 protonation-dissociation reactions, 25 1, 257 Lime, acidic deposition on forested soils, 55 Linear equilibrium adsorption, chemical transport through soil, 187 Linear equilibrium hypothesis, chemical transport through soil, 162 Linear isotherms, chemical transport through soil, 162-163, 172 Linear solute transport models, chemical transport through soil, 162 Linkage corn, 210-211,227 fingerprinting crop varieties, 103 Loam chemical transport through soil, 184, 187-188, 190 inorganics in soils, 371, 376 Loamy sand, chemical transport through soil, 186-188,190 Local dispersion, chemical transport through soil, 191 Lognormal frequency distribution, chemical transport through soil, 159 Longitudinal dispersion coefficient, inorganics in soils, 334-335 Lysimeters, acidic deposition on forested soils, 13

M Macropores, chemical transport through soil, 183, 187 Magnesium acidic deposition on forested soils chemical factors, 20-21, 36-37 physical factors, 7 soil change studies, 45,52,54,57-59, 61-63

395 inorganics in soils, 368-373 surface complexation models, 279,282, 312-313.318 Magnetites, surface complexation models, 256,258,297,300 Maize evolution of, see Corn, evolution of fingerprinting crop varieties, 94-97, 102-103 discrimination, 109- 113 new techniques, 125 usage, 120,122-123.125 Maleness, corn, 221,225-226 Manganese, surface complexation models, 279 Manganese oxide, surface complexation models, 259, 268,297 competitive adsorption reactions, 312-3 13, 315 inorganic anion adsorption, 300,303 metal ion adsorption, 279, 281 Mapping, fingerprinting crop varieties, 103-104, 126 Markers corn, 211 fingerprinting crop varieties, 103,125, 127 Masculinization, corn, 223,225-226 Mass balance chemical transport through soil, 160-161, 166, 181 surface complexation models, 238, 244 Mass balance equations, surface complexation models, 236, 242,250,258,320 Mass conservation law, chemical transport through soil, 144, 147,156 Mass flow, inorganics in soils, 334 Mass transfer chemical transport through soil, 187, 193 processes, 160-167 processes, analysis of, 172-175 inorganics in soils, 350,361-362, 374 Mass velocity, chemical transport through soil, 178 Mean convection rate, chemical transport through soil, 178, 187 Mean solute velocity, chemical transport through soil, 176-179 Mercury, surface complexation models, 286 Metal-ligand interactions, surface complexation models, 316-318

Metal-metal competition, surface complexation models, 312-313 Microbial population density, chemical transport through soil, 168-169 MICROQL, surface complexation models, 290,297, 319 MICTOQL, surface complexation models, 279 Migration, chemical transport through soil, 189, 193 MINEQL, surface complexation models, 279, 297,319 Mineral weathering, acidic deposition on forested soils, 35-40 Mineralization acidic deposition on forested soils, 5, 30-31,41,64 inorganics in soils, 347,354 Minerals clay, see Clay minerals inorganics in soils, 375 surface complexation models, 235,237, 309,316,320 protonation-dissociation reactions, 25 1, 254-256,267,270 MINETEQ, surface complexation models, 319-320 Minimum distance, fingerprinting crop varieties, 121-122, 126-127 Mobile regions chemical transport through soil, 173 inorganics in soils, 332,367,374 Mobile water content, chemical transport through soil, 166-167,174 Mobile-immobile models, inorganics in soils, 332 ion exchange, 370-371 multiple-reaction models, 359, 361-363, 367 Mobile-immobile water model, chemical transport through soil, 166,174 Modeling of inorganics in soils, see Inorganics in soils, modeling of Modified Roger's distance, fingerprinting crop varieties, 105 Molybdate, surface complexation models, 291, 300-301, 303-304, 315-316 Molybdenum, surface complexation models, 291,303 Montmorillonite inorganics in soils, 349

396

INDEX

Montmorillonite (continued) surface complexation models, 256,267, 270,29I Morphology corn, 215-219,221,225 fingerprinting crop varieties, 86-88, 108, 119-122 Multiple-reaction models, inorganics in soils, 350-351 multireaction models, 353-358 second order models, 358-367 two-site models, 351-353 Multireaction and transport model (MRTM), inorganics in soils, 355-358 Multireaction model (MRM), inorganics in soils, 355-356, 364 Multivariate analysis, fingerprinting crop varieties, 104-108 MUSIC program, surface complexation models, 274 Mutation corn, 205,213,224 multiple domestications, 215,221 transformation, 206,210 fingerprinting crop varieties, 98 Mycorrhizae, acidic deposition on forested soils, 64-65

N Napropamide, chemical transport through soil, 183,187-188 Neutralization, see also Acid-neutralizing capacity (ANC) acidic deposition on forested soils, 1 chemical factors, 24, 36, 38,40 future research, 64 physical factors, 8 surface complexation models, 235 Nickel inorganics in soils, 368, 375-376 surface complexation models, 285-286 Nitrates acidic deposition on forested soils, 1-3, 5-6.66 chemical factors, 17, 19,27 physical factors, 7, 14-15 retention, 29-32 soil change studies, 43.45-46,48,56 chemical transport through soil, 142, 167 surface complexation models, 282

Nitrification, acidic deposition on forested soils, 30,32,40,48,64 Nitrogen acidic deposition on forested soils, 65 chemical factors, 26,41 retention, 29-32 soil change studies, 45,48,52,56 fingerprinting crop varieties, 95 Nonaqueous phase liquid (NAF'L), chemical transport through soil, 145-147, 165 Nonlinear adsorption models, chemical transport through soil, 171-172 Novelty, fingerprinting crop varieties, 121 Nucleotides, fingerprinting crop varieties, 98 Nutrient cycling, acidic deposition on forested soils, 10,31-32,34,63 Nutrient uptake, acidic deposition on forested soils, 4,28,43,45,65

0 One-pKmodel, 234,249-251.307, 320 metal ion adsorption, 287-289 protonation-dissociation reactions, 212-214 Organic acids, acidic deposition on forested soils, 3, 28 Organic carbon, chemical transport through soil, 163, 168 Outer-sphere surface complexes, surface complexation models, 240-243 competitive adsorption reactions, 313 inorganic anion adsorption, 300-301 metal ion adsorption, 282 protonation-dissociation reactions, 255 Oxidation, acidic deposition on forested soils, 1,5, 30 Oxides inorganics in soils, 375 surface complexation models, 235-238, 240,243 competitive adsorption reactions, 312-318 inorganic anion adsorption, 296, 301 metal ion adsorption, 277-278.282, 284 organic ligand adsorption, 308-312 protonation-dissociation reactions, 25 1, 254,256,270 Oxisol, surface complexation models, 266, 282,301

397 Oxygen, surface complexation models, 249-250 Ozone, acidic deposition on forested soils, 41

P Pedigree, fingerprinting crop varieties, 105-106, 111, 122, 124-125 Permeability, chemical transport through soil, 175, 186, 189, 192 Pesticides, chemical transport through soil processes, 163-164, 171-172 solute transport, field studies of, 182-184, 186-187 PH acidic deposition on forested soils, 3-4,6, 65-66 chemical factors, 15-20.23-28.30-31 mineral weathering, 38 soil changes, 45,49.51,56-60.62 sulfate retention, 33-35 chemical transport through soil, 164, 168 fingerprinting crop varieties, 89 inorganics in soils, 332, 352 surface complexation models, 235,308, 313,319 inorganic anion adsorption, 300,303, 305 metal ion adsorption, 274,285-286 protonation-dissociation reactions, 25 1, 261,265 Phase concentration models, chemical transport through soil, 161 Phenotype corn, 215 fingerprinting crop varieties, 88, 105 Phosphate inorganics in soils, 337 surface complexation models competitive adsorption reactions, 314-315.318 inorganic anion adsorption, 290-291, 295-297,300-301.303-305,307 Phosphorus, inorganics in soils, 333, 340, 342,353,356,358 Photodecomposition, chemical transport through soil, 168 Photosynthate partitioning, corn, 225-226 Phylogenetic information, fingerprinting crop varieties, 98, 124

Pinus, acidic deposition on forested soils, 57-58 Piston flow velocity, chemical transport through soil, 176,178-179,184,187 Pith abscission, corn, 209, 211 Plant breeding corn, 203, 213,228 fingerprinting crop varieties, 119, 127 Plant habit, corn, 219-221.223 Plutonium, surface complexation models, 281, 317 Pollen, corn, 221-222,226-228 multiple domestications, 219,221 transformation, 207,212 Polymerase chain reaction, fingerprinting crop varieties, 100, 126 Polymers, acidic deposition on forested soils, 16.23-24 Polymorphism, see also Restriction fragment length polymorphisms fingerprinting crop varieties, 99, 101-102, 104, 110, 126 Pore water velocity, chemical transport through soil, 156 Porosity, chemical transport through soil, 146, 148, 175 Potassium acidic deposition on forested soils chemical factors, 21, 36 physical factors, 7, 9 soil change studies, 57-61 inorganics in soils, 349-350, 353 surface complexation models, 307 Potato chemical transport through soil, 186 fingerprinting crop varieties, 101, 115, 120 Potentiometric titration, surface complexation models, 237,291 metal ion adsorption, 277,288 protonation-dissociation reactions, 254-256,265,267,270-271.273 Precipitation acidic deposition on forested soils. 27, 33, 35,37-38,40 chemical transport through soil, 167 inorganics in soils, 332,347,349,354 surface complexation models, 278,287, 296,305, 307,319 Preferential flow of solute, chemical transport through soil field regime, 192-193

INDEX

398

Preferential flow of solute, chemical transport through soil (continued) field studies, 179, 181, 185-188 processes, 152,174-175 Probability density function (pdf), chemical transport through soil, 157-160, 162, 169-170,194 Probes, fingerprinting crop varieties, 100-104 Productivity, corn, 206,208,225-226,228 interpathway heterosis, 221 multiple domestications, 219 Prolamins, fingerprinting crop varieties, 114-115 Protandry, corn, 222-223 Protein, fingerprinting crop varieties, 86, 123-125 characters, 88-98, 104 discrimination, 109-110,112,114-118 Protogyny, corn, 222-223 Proton charge, surface complexation models, 235,238,243 Protonation, surface complexation models, 234,241-242,245,248,320 inorganic anion adsorption, 290, 301-302 protonation-dissociation reactions, 251-274 Protons, surface complexation models, 240, 245,249

Q Quantitative trait loci, fingerprinting crop varieties, 124 Quickflow, acidic deposition on forested soils, 13-15.35

R Rachilla, corn, 210,222 elongation, 212,217,227 multiple domestications, 217-219 Radial diffusion, chemical transport through soil, 166, 173 Radiolabeling, fingerprinting crop varieties, 89, 126 Rainfall acidic deposition on forested soils, 9, 13,27 chemical transport through soil, 167, 176, 181,184,187, 190 Rainwater, acidic deposition on forested soils, 7-9

Random dispersion, chemical transport through soil, 187 Rate coefficients chemical transport through soil, 173-174 inorganics in soils ion exchange, 375 kinetic retention models, 344 multiple-reaction models, 352, 354-357,360,363-365.367 surface complexation models, 284 Rate-limited adsorption, chemical transport through soil, 172-174 Rate-limited mass transfer model, chemical transport through soil, 166 Rate-limited sorption, chemical transport through soil, 164-165 Rate processes, chemical transport through soil, 193 Recombination corn, 227 fingerprinting crop varieties, 121 Reduction, acidic deposition on forested soils, 5 Relative permeability, chemical transport through soil, 146-147 Relative saturation, chemical transport through soil, 146 Resident concentration pulse, chemical transport through soil, 178 Resident concentrations, chemical transport through soil, 160-161 Restriction digests, fingerprinting crop varieties, 101-102 Restriction endonucleases, fingerprinting crop varieties, 98, 100-101, 126 Restriction enzymes, fingerprinting crop varieties, 102, 104, 126 Restriction fragment length polymorphisms ( R E P S ) , fingerprinting crop varieties characters, 98-100,103-108 discrimination, 110, 112-115 new techniques, 125 usage, 120, 122-125 Retardation factor chemical transport through soil processes, 162-164,173 solute transport, field studies of, 183-184, 187-199 inorganics in soils, 352, 365, 369, 378-379 Retention, inorganics in soils, 332-333, 336 computer codes, 368-369,371,373

INDEX equilibrium retention models, 337-342 kinetic retention models, 342-350 layered soil, 377-379 multiple-reaction models, 350-353,355, 357-365 Reversed-phase high-performance liquid chromatography, fingerprinting crop varieties, 95-96, 111-115 Reversing Acidification in Norway (RAIN), acidic deposition on forested soils, 27-28.37-38,66 Rhizosphere, acidic deposition on forested soils, 42, 64 Rhodamine, chemical transport through soil, 184,186-188 Rind abscission, corn, 209-210 RNA, fingerprinting crop varieties, 100 Root zone, inorganics in soils, 331 Rooting zones, acidic deposition on forested soils, 8, 10, 64 Roots, acidic deposition on forested soils, 4, 31, 65 Runoff, acidic deposition on forested soils, 27, 35.38

S Saturated hydraulic conductivity, chemical transport through soil, 146,154, 191 Saturation acidic deposition on forested soils, 13, 15, 30-33,35,49 chemical transport through soil, 154, 165, 189 inorganics in soils, 333, 377-379 Seasonal changes, acidic deposition on forested soils, 9, 30-31, 59-61 Second-order mobile-immobile model (SOMIM), inorganics in soils, 361-365, 367 Second-order models, inorganics in soils, 358-367.375 Second-order two-site model (SOTS), inorganics in soils, 359-361,363-365 Secondary sex traits, corn, 205,212, 222-224 Seed corn, 228 fingerprinting crop varieties, 8 9 , 9 5 , 9 7 , 110, 119 Segregation, corn, 208-209, 211,227

399

Selection, corn, 226 Selectivity coefficients, inorganics in soils, 360,363,368,371-372 Selenate, surface complexation models, 300-301,303,305,318 Selenite, surface complexation models competitive adsorption reactions, 314-315, 318 inorganic anion adsorption, 296,300-301, 303-305 Selenium, surface complexation models, 237, 291, 295 Sequences, fingerprinting crop varieties, 98-99, 101, 126 Serology, fingerprinting crop varieties, 89 Shattering cob, corn, 209 Sigmoidicity. inorganics in soils, 341-342 Silica, surface complexation models, 274, 278,286 Silicate acidic deposition on forested soils, 16, 36-38,40 surface complexation models, 290,296, 3 14-3 15 Silicon dioxide, surface complexation models, 308, 316 Silver, surface complexation models, 278-279.317 Sodium acidic deposition on forested soils, 21, 36 inorganics in soils, 367, 370-373 Soil chemical transport through, see Chemical transport through soil fingerprinting crop varieties, 96 forested, acidic deposition on, see Acidic deposition on forested soils inorganics in, see Inorganics in soils, modeling of Soil acidification, see Acidic deposition on forested soils Soil chemical systems, surface complexation models in, see Surface complexation models Soil coring, chemical transport through soil, 178-179,188 Soil horizons, acidic deposition on forested soils cation exchange, 19-20.24 future research, 64 physical factors, 9 , 17, 30-31, 33,40

Soil matrix chemical transport through soil, 148,152, 157,181,184,193 inorganics in soils, 332, 335, 337 ion exchange, 368 multiple-reaction models, 352-353, 358-362 Soil moisture, acidic deposition on forested soils, 13, 16,30 Soil organisms, acidic deposition on forested soils, 40-42 Soil pH, acidic deposition on forested soils, 15-19 Soil solution, acidic deposition on forested soils, 43, 52,54 SOILCHEM, surface complexation models, 319 Solute adsorption, chemical transport through soil, 161-165, 171-175, 182-185 Solute dispersion, chemical transport through soil, 151-153, 179-182,189-191, 194 Solute mass, chemical transport through soil, 187 Solute retention, inorganics in soils, 339, 342, 345,358-360 Solute transport chemical transport through soil, 154, 192-193,194 field studies, 175-191 inorganics in soils, 332-336 ion exchange, 367 layered soil, 377-378,381 multiple-reaction models, 350, 353, 355-356, 359,362 Solute velocity, chemical transport through soil, 176-179,189-191 Sorghum, fingerprinting crop varieties, 101, 115 Sorption chemical transport through soil, 164-165, 174,185 inorganics in soils, 332-333,368-369, 311-372.374-376 equilibrium retention models, 337, 340-341 kinetic retention models, 342-343, 345 multiple-reaction models, 350, 352-354,356,358-359,362-363 surface complexation models computer codes, 319

inorganic anion adsorption, 296,305, 307

metal ion adsorption, 278,285,287 Soybean, fingerprinting crop varieties, 101, 113, 125

Spikelets, corn, 205,207-211.225.227-228 Sprinklers, chemical transport through soil, 183,186-187,190-191

Squash, 223 Stability, fingerprinting crop varieties, 119-120

Stagnant water phase models, chemical transport through soil, 166-167, 173-174

STEADYQL, surface complexation models, 319

Stemflow, acidic deposition on forested soils, 7-10,31

Stem plane, surface complexation models, 249,288

Stem VSC-VSP model, 244-246,319-320 inorganic anion adsorption, 303-305 metal ion adsorption, 282-286 protonation-dissociation reactions, 267-270

Stochastic continuum model, chemical transport through soil, 153-154 Stochastic local solute transport model, chemical transport through soil, 154 Stochastic-convective flux, chemical transport through soil, 156,161, 189,194 Stochastic-convective transfer function, chemical transport through soil, 159 String cob trait, corn, 207,219,228 Sulfates acidic deposition on forested soils, 1-2, 5-6,66

chemical factors, 17, 19, 27 future research, 64 mineral weathering, 37 physical factors, 7-9, 13-15 retention, 29, 32-35 soil change studies, 45-46,52,54-56 surface complexation models, 290,301, 305, 315,318

Sulfur acidic deposition on forested soils, 65 chemical factors, 26, 29, 32-35 physical factors, 7 soil change studies, 52,54-56

INDEX fingerprinting crop varieties, 95,97 surface complexation models, 235 Surface charge, surface complexation models, 247,308,320 Surface charge density, surface complexation models, 246,267,283,300,320,341 Surface complexation models, 234-235, 320 competitive adsorption reactions anion-anion, 314-316 metal-ligand, 316-318 metal-metal, 312-313 computer codes, 319-320 description characteristics, 235-237 constant capacitance model, 238-240 generalized two-layer model, 246-248 one-pK model, 249-25 1 Stem VSC-VSP model, 244-246 triple-layer model, 240-244 inorganic anion adsorption constant capacitance model, 289-297 generalized two-layer model, 305-307 one-pK model, 307 Stern VSC-VSP model, 303-305 triple-layer model, 297-303 metal ion adsorption constant capacitance model, 274-279 generalized two-layer model, 286-287 one-pK model, 287-289 Stem VSC-VSP model, 282-286 triple-layer model, 279-282 organic ligand adsorption constant capacitance model, 308-310 Stern VSC-VSPmodel, 311-312 triple-layer model, 309, 311 protonation-dissociation reactions constant capacitance model, 251-256 generalized two-layer model, 270-27 1 one-pK model, 272-274 Stem VSC-VSP model, 267-270 triple-layer model, 256-266 Surface hydroxyl groups, surface complexation models, 235-237, 245, 251,254,256 Surface precipitation model, surface complexation models, 278, 287, 296-297,307, 312 Surface site density, surface complexation models, 237,267,282, 312

40 1

Surface-solution interface, surface complexation models, 238,240,247, 249

T Tassels, corn, 208,212,219,224-226,228 Taxonomy, fingerprinting crop varieties, 119-120 Tehuacan corn, 206-207,212-213,223 Temperature chemical transport through soil, 165-166, 168 surface complexation models, 236,265, 285,303 Teosinte, see Corn, evolution of Thermodynamics inorganics in soils, 360, 368 surface complexation models, 234,243, 285 Through fall, acidic deposition on forested soils, 7-10, 31 Tight linkage, corn, 210-211,227 Titanium oxide, surface complexation models, 237,308,316 metal ion adsorption, 275, 279 protonation-dissociation reactions, 252-253,256,258-259,265,271 Titration, see also Potentiometric titration acidic deposition on forested soils, 19 surface complexation models, 237,291 metal ion adsorption, 277, 288 protonation-dissociation reactions, 252-256,258,265-267, 270-274 Tortuosity, chemical transport through soil, 146, 148 Tracers, chemical transport through soil, 184, 191 TRANQL inorganics in soils, 367-368 surface complexation models, 319 Transcription, fingerprinting crop varieties, 89 Transfer coefficient models, chemical transport through soil, 165 Transfer coefficients, surface complexation models, 284 Transfer function, chemical transport through soil, 156-160,169,194

402 Transformation chemical transport through soil, 143-145 concentrations, 160- 161 convection-dispersion , 154- 156 interphase mass transfer laws, 161-167 mass flux laws, 145-154 reactions, 167-169 transfer function models, 156-160 corn, 205-214 inorganics in soils, 333,347 Transport equations, inorganics in soils, 333-336 Transport models inorganics in soils, 359,368, 370-372, 376 surface complexation models, 319-320 Transport through soil, chemical, see Chemical transport through soil Transport volume, chemical transport through soil, 158, 160 Travel time, chemical transport through soil processes, 156-157,159-160.162, 169-170 solute transport, field studies of, 190-191 Tricklers, chemical transport through soil, 178,190-191 Triple-layer model, surface complexation models, 234,240-247,320 competitive adsorption reactions, 312-313, 315-318 computer codes, 319 inorganic anion adsorption, 297-303 metal ion adsorption, 279-282 organic ligand adsorption, 309,311 protonation-dissociation reactions, 256-267 lhnicate locus, corn, 212 Two-site models, inorganics in soils, 351-353,361-362

U Uniformity, fingerprintingcrop varieties, 119-120

V Vapor-dissoved phase partitioning, chemical transport through soil, 165-166 Vapor flux, chemical transport through soil, 145-146

Variable charge model, metal ion adsorption, 284 Variable selectivities, inorganics in soils, 371-372 Variable surface charge model, see Stern VSC-VSP model Variable surface potential model, see Stern VSC-VSP model Velocity chemical transport through soil, 193 processes, 149-153, 167, 173,175 solute transport, field studies of, 176-179,181,189-191 inorganics in soils, 334-335, 352 Vermiculite, inorganics in soils, 349-350 Viscosity, chemical transport through soil, 174,192 Volume-averaged local solute transport model, chemical transport through soil, 153

W WATEQ3, surface complexation models, 319 Water flow chemical transport through soil chemical mass flux, 151,153 interphase mass transfer, 161-162, 166-167 processes, 155-156,158, 175 solute transport, field studies of, 183, 185-186 inorganics in soils, 333-335,368,377, 379,381 Water flow velocity, inorganics in soils, 332, 334,365 Water flux chemical transport through soil processes, 147-148.153-154, 156,175 solute transport, field studies of, 176, 179 inorganics in soils, 332, 380 Water-phase division model, chemical transport through soil, 166 Water-unsaturated soils, inorganics in soils, 379-381 Water velocity chemical transport through soil processes, 149,151-152, 156-157, 174 solute transport, field studies of, 176, 179

INDEX inorganics in soils, 335 Watershed, acidic deposition on forested soils chemical factors, 31, 34-35,40 physical factors, 11, 13-14 soil change studies, 54 Weathering, acidic deposition on forested soils, 35-40, 5 5 , 57 Wheat, fingerprinting crop varieties, 95, 125 discrimination, 110-111 usage, 123-125

Z Zea mays,evolution of, see Corn,evolution of Zeins, fingerprinting crop varieties, 112-113

403

Zero point of charge, surface complexation models, 251,253,257,261,272 Zero surface charge, surface complexation models, 243,256,261 Zero-time dispersion model, chemical transport through soil, 159 Zero-time model of solute dispersion, chemical transport through soil, 151-152 Zinc inorganics in soils, 340,375-376 surface complexation models competitive adsorption reactions, 312-313,318 metal ion adsorption, 279,281-286

This Page Intentionally Left Blank