TRACE METAL SPECIATION AND BIOAVAILABILITY IN URBAN CONTAMINATED SOILS

YINC GE

Department of Natural Resource Sciences, McCill University Montreal, Canada, May, 1999

A Thesis Submitted to the Faculty of Graduate Studies and Research in Partial Fulfilment of the Requirement for the Degree of Master of Science

O 1999 YING GE

of Canada

National Library

Bibliothèque nationale du Canada

Acquisitions and Bibliographie Sewices

Acquisitions et services bibliographiques

395 Wellington Street Ottawa ON K 1 A W Canada

395, rue Wellington OttawaON K1AON4

Canada

The author has granted a nonexclusive Iicence allowing the National Library of Canada to reproduce, loan, distribute or sel1 copies of this thesis in microform, paper or electronic formats.

L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la fome de microfiche/film, de reproduction sur papier ou sur format électronique.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fiom it may be printed or otherwise reproduced without the author's permission.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.

ABSTRACT

Trace Metal Speciation and Bioavailability in Urban Contaminated Soils

Ying Ge. Department of Natural Resource Sciences, McGill University M. Sc. in Environmental Soil Chemistry. May 1999

Urban soils are ofien contaminated with trace metals and the toxicity of the rnetals depends. in part, on their speciation in soil solutions. The objectives of this project were to estimate the metal speciation in urban soils and to evaluate the predictability of soil metal pools on plant uptake. The chernicd speciation of Cd, Cu, Ni, Pb and Zn was estimated by using the Windennere Humic Aqueous Mode1 (WHAM). In soil solutions, Cd, Ni and Zn

were present m d y as fiee ions when the solutions were acidic and their organic complexes were dominant as the pH was over 7.5. The other two metals mostiy fomed complexes with

organic ligands.The activities of Cd'*. Cu'*. Ni2-,Pb2-and Zn2'were affected by soil pH and total soil metal burdens. Ail five rnetals were under-saturated with respect to the minerais which could po tentiaüy control the metal solubility.

Metal upt ake by plants in the contaminatedrailway yards was generallynot correlated with k e , dissolved and total soi1 metal pools. A pot experiment demonstrated better

conelations between the metal pools and the metal content in wild chicory. Multiple regression analysis showed that the metals in the leaves and roots of wild chicory could be adequately predicted by the soi1 total met& and soil propaies such as pH and exchangeable Ca.

Spéciation et biodisponibilité des métaux-traces dans les sols urbains contaminés

Ying Ge, Département des sciences des ressources naturelles. Université McGill

M. Sc. Chimie de l'environnement. Mai 1999

Les sols urbains sont souvent contaminés par des métaux-traces et leur toxicité dépend en partie de leur spéciation dans les solutions de sol. Les objectifs de ce projet sont

i) d'estimer la spéciation des métaux pour certains sols urbains et ii) d'évaluer s'il est possible de prédire les formes de métaux du sol qui contribuent a l'absorption des métaux par les

plantes. La spéciation des métaux-traces Cd2-,Cu2-.Ni2*. ~ b ' *et 2n2*dans les solutions de sol a été déterminée en utilisant le logiciel WHAM. Le Cd, le Ni et le Zn étaient présents surtout sous leur fome ionique simple lorsque la solution de sol était acide.

Les autres

métaw se retrouvaient surtout sous leur forme complexée avec des ligands organiques ou inorganiques. L'activité des métaux Cd, Cu. Ni. Pb et Zn dépend du pH du sol et de la quantité totale de métaux dans les sols. Dans la solution de sol, les cinq métaux se trouvaient sous-sanués par rapport aux minéraux qui pouvaient contrôler leur solubilité.

Des sols et des plantes ont été échanhllomés dans des cours de triage ferroviaire et l'analyse chimique a été complétée. L'absorption des métaux par les plantes n'était comilée

avec aucune forme de métal en solution: libre. dissoute ou totale. Par ailleurs. une expérience de croissance en pot a démontré de meilleures corrélations entre les formes de métal dans la solution de sol et le contenu total dans la chicorée sauvage. Par une analyse

de régression multiple. on observe que la quantité de métal dans les feuilles et les racines de la chicorée sauvage peut être prédite par une combinaison de la quantité totale de métaux dans la solution de sol et de cenaines propriétés du sol telles le pH et la concentration de Ca échangeable.

CONTRIBUTION OF AUTHORS

This thesis coasists of two manuscnpts (Chapters 2 and 3) to be submitted for

publication. The manuscripts were CO-authoredby the candidate. Ms.Patricia Mmay and Prof. William H. Hendershot. The candidate was responsible for conducting eqeriments, collecting and analysing data and preparing the manuscripts. Ms.Patricia Murray, who is also a M.Sc. candidate. shared the responsibilities in sample collection. chemical analysis in the research project and helped in editing the manuscripts. Prof. Wiiliam H. Hendershot supervised the research project, provided technical assistance in the research process and contributed editorial comments on the manuscnpts.

In keeping with the Guidehes Concerning Thesis Preparation (McGill University, 1999) the followinç paragraph is reproduced in full:

in general, when CO-authoredpapers are included in a thesis the candidate must have made a substantial contribution to all p a p a Uicluded in the thesis. In addition, the candidate

is required to make an explicit statement in the thesû as to who contributed to such work and to what extent. This statement should appear in a single section entitled "Contributions of Authors" as a preface to the thesis. The supexvisor must attest to the accuracy of this statment at the doctoral oral defence. Since the task of the examiners is made more H c u l t

in these cases, it is in the candidate's interest to clearly spece the responsibility of ail the authors of the CO-authoredPapen.

ACKNOWLEDGMENTS

First of ali. 1 would k e to express deep gratitude to my supervisor. Prof. William H. Hendershot. for his inspiration. encouragement and support in my M.Sc. studies and research. 1 am very gratehl to Prof. W

i D. Marshall and Prof. Laurie Chan. who provided the use

of laboratory equipment and helpfbi suggestions in my thesis witing. in addition. 1gratefully acknowledge the help fkom Prof. Angus F. MacKenzie in my M. Sc. program. Special thanks are also given to Dr. Sébastien Sauvé for his comrnents on the manuscripts of my thesis.

1 greatly appreciate the assistance boom my colleague. Patricia Murray. in sample

analysis, data collection and thesis correction. 1 am thaakful to Hélène Lalande for her

assistance in chernical anaiysis and translation of the abstract of my thesis hto French. 1 &O

thank Donna Leggee and Ying Shi. who helped in the use of Iaboratory equipment. sample preparation and analysis.

The financiai suppon for this project, which was fiom the Natural Sciences and

Engineering Research Councii (NSERC) of Canada, is greatly acknowledged.

My gratitude also extends to my parents for their conthuous understanding and suppon.

TABLE OF CONTENTS

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i Résumé . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ü

Contribution of Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iv

\I

Chapter 1 . Literature Review on Metal Speciation and Bioavailability in Soils . . . . . . . . 3 1 . 1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2. Metal fractions in soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

.............................................6 1.3.1. Theoiy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.3.2. Methods ofdetermination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.3. Metal speciation

1.3.3. Controhg factors on metal speciation . . . . . . . . . . . . . . . . . . . . 9 1.4. Metal bioavdabüty

........................................ 11

........................... 11 1.4.2. Effects ofsoilproperties onmetal uptake . . . . . . . . . . . . . . . . . . 14 15. Research ne& . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 1.4.1. Metal pools and plant uptake

Chapter 2. Metai Speciation and Availability to Indigenous Plants in Three Railway Yards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.1. Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.2. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.3. Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . tO 1

Site description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.3.2. Soi1 and plant sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.3 .3 . Soil characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

--

2.3.4. Soi1 solution analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ? 3

-. . . . . . . . . . . . . . --

3.3.5. Plant analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

39

2.3.6. Speciation calculatioos and statisticai anaiysis

77

3.4. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23

2.4.1. Soil characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.4.2. Totai metais in soils and soi1 solutions . . . . . . . . . . . . . . . . . . . . 33

2.4.3. Meta1 speciation and mineral control . . . . . . . . . . . . . . . . . . . . . 28 2.4.4. Soi1 controls on fiee metal activities . . . . . . . . . . . . . . . . . . . . . . 31

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.5. Conclusiom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 4.4.5. Metai uptake by plants

Comecting Statement between Chapter 2 and Chapter 3 . . . . . . . . . . . . . . . . . . . . . . . 36 Chapter 3 . Interrelationsbips between Trace Metals in Urban Soils and Their Uptake by Wild Chicory (Cichoriurn intybus) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.2. Inuoduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3.3. Material and rnethods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.3.1. Site description a d sampling procedures . . . . . . . . . . . . . . . . . . 39 3.3.2. Pot experirnent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.3.3. Soi1 and plant anaiysis ................................ 40

3.1. Abstract

3.3.4. Soil solution analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.3.5. Speciation calculations and statistical analysis . . . . . . . . . . . . . . 40 3.4. Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

. . . . . . . . . . . . . . . . . . . . . . . . . 41 3.4.2. Metal speciation in saturation extracts . . . . . . . . . . . . . . . . . . . . 42 3.4.3. Effects of pH and DOC on metal speciation . . . . . . . . . . . . . . . . 44 3.4.4. Comparison of metal speciation with other studies . . . . . . . . . . 46 3.4.1. Soil properties and total metals

3 .4.5. Equilibria of metal in solution and mineral phase . . . . . . . . . . . . 17 3.4.6. Metal uptake of wild chicory . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 3.4.7. Prediction of the chicory uptake . . . . . . . . . . . . . . . . . . . . . . . . 50

3.5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Chapter 4 . A General Summary of the Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

LIST OF TABLES

Table 2.1. Selected soi1 chemical characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Table 2.2. Total metals in soi1 samples and Québec soil critena (mean~standard

.

deviation mg kg-') . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Table 2.3. Selected chemical propaies of soi1 solutioas and Quebec guideline of water quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Table 2.3. (Cont.) Selected chemical properties of soil solutions and Québec guideline of waterquality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Table 2.4. Metal speciation in extracted soil solutions

. . . . . . . . . . . . . . . . . . . . . . .29

Table 2.5. Trace metal concentrations in plant tissue (mg kg-'). . . . . . . . . . . . . . . . . . 33 Table 2.6. Correlation coefficient (r) of metal in plant and different metai pools (n=23)

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34 Table 3.1. Selected soi1 chemical characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41 Table 3.2. Total metal concentrations in soil samples and Québec soi1 criteria (mg kg")

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Table 3.3. Selected chemical properties of saturation extracts and Québec guideline of waterquality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Table 3.4. Metal speciation of extracted soi1 so1utions . . . . . . . . . . . . . . . . . . . . . . . . 44 Table 3.5. Total metal content and BCF in leaf and root tissue of wild chicory (mgkg")

............................................................

51

Table 3.6. Correlation coefficient (r) of metal in wild chicory and different metal pools

(Le& n = 58; Root: n = 52) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Table 3.7. Equations to predict plant met& in leaf tissue (n = 58) . . . . . . . . . . . . . . . 56 Table 3.8. Equations to predict plant met& in root tissue (n = 52) . . . . . . . . . . . . . . 57

LIST OF FIGURES Figure 1.1. Equilibria of metals in soi1 solid and iiquid phases . . . . . . . . . . . . . . . . . . . . . 5

Figure 2.1. The log activities of free ions venus the pH of the saturation extracts. The triangles and h e s represent the free ion activities in solution and on rnineral surface. respectively. Both activities were calculated by the WHAM . . . . . . . . . . . . . . 30 Figure 3.1. Distribution of metal species as affected by pH (5 - 9) and W C (column a: 9.58

mg C L*'and column b: 79.5 mg C L-'). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Figure 3.2. Cornparisons between activities of cd2-. Cu'- and Pb'- calculated by the WHAM (senes 1 ) and those predicted by Sauvé's equations (series 2) . . . . . . . . . . . . . . 48 Figure 3.3. The log activities of fhe ions versus the pH of the saturation extracts. The

triangles and iines represent the tiee ion activities in solution and on rnineral surface, respectively. Both activities were calculated by the WHAM . . . . . . . . . . . . . . 49

GENERAL INTRODUCTION

D h g the past two decades. trace metal contamination has received considerable attention due to its deleterious impact on the environment and human health (Adriano, 1986).

In urban regions. soils may contain excessive levels of metal contaminants. which are mainly the result of anthropogenic factors such as vehicular traffic, industrial activities and dumping of municipal waste or sewage sludge (Kabata-Pendias and Pendias. 1992). As a result of

incteasinç demands for environmental protection and land restoration. research interest has

been growing on chernical speciation and bioavailability of trace metals in urban soils (AUoway. 1995).

In this thesis, metal speciation is defhed as the distribution of physico-chernical forms that comprise the total concentration of a given metal in soil solutions. It is presumed that metal speciation has relationships with the uptake of that element by plants (Parker et al., 1 995). However, there is a lack of references on the Lioik between metal speciation and plant

availability in urban soils.

Thus the objectives of this thesis are to: 1) estirnate metal speciation in the solution of urban soils; 2) identify those factors that connol metal speciation; 3) establish the statistid relationships benveen met al pools (free ions, total dissolved and total soil metals) and plant uptake.

The first chapter of the thesis. the literature review, provides an overview on the theories, research met hods and Uiterrelationships between met al speciation and availabihy t o plants. Chapter 2 describes the experiments conducted on three railway yards in 1997 and sumniarizes the data for the metal content of soil, solution and plant samples. in Chapter 3. we attempted to link metal uptake by wild chicory, which was grown in a pot experiment

during 1998. with soil met al pools. The results, however, showed that metal bioavailability was difficult to predict by the different metal pools in urban sods.

CHAPTER 1. LITERATURE REVIEW ON METAL SPECiATION AM) BIOAVAUABILITY IN SOUS

1.1. Introduction

In urban soils. trace metals such as Cd. Cu. Ni. Pb and Zn ofien exceed guidelines of soil quality (e.g.. CCME, 1991) as results o f industrial activities. waste discharge. fuel combustion and sludge applications to arable lands (Adriano. 1986). Due to their excessive concentrations. these metals may cause both environmental degradation and toxic effects on human or animais ( Kabata-Pendias and Pendias, 1992).

Metal uptake by higher plants from soils is a major pathway for metals to enter the food chain (Loneragan. 1975). It is g e n d y accepted that plants mainly take up metals from

soi1 solutions and the level of dissoived metals can be buffered by various metal forms in the soi1 sotid phase (Chaney. 1988). Thus the accumulation of metals in plants. to some extent.

reflects the availability of metals in the soil pools.

Metal availability to plants can be assessed by ushg selective extraction and chemical speciation. The first approach employs chernical reagents to dissolve cenain metal pools, which are used to predict plant metal content (Haq et al., 1980; Xian. 1989; Jackson and Moway, 1992; Qian et al., 1996). Selective extraction is useful to characterizethe available

pools in the solid phase. However, it does not speciQ which metal form(s) are taken up by plants and provides little insight mto the process of metal tramfer from soi1solutions to plant

root surfaces (Tack and Verloo, 1995).

On the other hand in aquatic systems such as natucal waters or synthetic solutions, there is considerable evidence that plant response depends on metal speciation (Parker et al.. 1995). Most solution culture experiments have demonstrated that metal content in plants

generally correlates best with the activity of fkee. uncomplexed metal ions. However, unlilce aquatic systems, soil solution chemistry is usually more compiicated because of solid-liquid

phase interactions. Their effects on metal speciation have not been hily understood as t hennodynamic and kinetic descriptions of some of these interactions are not available (Tack

and Verloo. 1995). Therefore. the relationships between plant uptake and metal speciation in soil solution remain to be established,

1.2. Metal fractions in soi1

In soils, trace met& partition into a range of soil components and they may be in the foUowing forms ( Shuman. 1991 ): 1. Dissolved: as fiee ions or as soluble complexes with inorganic anions and organic

ligands.

2. Exchangeable: held by predorninantly electrostatic forces on negatively charged sites on the surface of clay, organic matter and amorphous materials. 3. Organicdy bound: complexeci, chelated or adsorbeci by organic materials either recently synthesized or decomposed by microbial activity. 4. horganicaily bound: specifically adsorbeci on surface of clay minerais, carbonates,

phosphates or oxides of iron. aluminurn or mmganese.

5. Residual: occluded in crystal lattice of minerais. Metals in soi1 solutions are assurnecl to be buffered by the various metal f o m in the soil solid

phase through reactions such as adsorption/desorption and dissolutioa~precipitation ( McBnde.

1989). The equitibria between these forms are shown in Figure 1.1.

Exchangeable

Inorgaaically bound

0 Dissolved

Figure 1.1. Equilibria of met& in soil solid and iiquid phases

The above met al fiactions cm be quantified by single or sequential extractions (Lake et ol., 1984). For example, the dissolved hction is usually extracted by distiiled or deioaized

water (Miller and McFee, 198 3 ;Miller et ai., 1986a). The exchangeable forms are estimated

by extraction with simple salt solutions such as KNO, or CaCll (Shuman, 1979;Emmerich et ai.,1982a). A chelating agent, EDTA,has shown its effectivmess in dissohmg inorganicaily

and organically bound met& (Emmerichet of., 1 982a; Miller and McFee, 1983; Miiler et al., l986b). The residual met& c m be estimated by digestion m strong acids such as HNO,, HF

and HCIO, (Sims and Pamck, 1978 ;Tessier et al., 1979; Emmerich et al., 1982a). Although

the extraction method is usehl to estimate the size of available metal pools. it remahs operationally-defhed because metal kactions may redistribute in the solid phase during the process of extraction (Beckett, 1989). Furthemore, it is difficult to compare resuits from different studies as a result of the variability in reagents and procedures of extractions (Ross, 1994).

1.3. Metal speciation

1.3.1. Theory in natural water, such as soi1 solutions. trace metals may exist as either bec

(uncompiexed) ions or complexes with inorganic and organic ligands (Evans, 1989). The formation of cornpiexes between metal ions (W)and ligands (L"')can be represented by: m w

+

n..

L

M,,,L;mi-ln':hL

b,= [&L/'mz"']/([M=-]m [L"'])

(1)

(2)

The total concentration of the metal in solution, [MlT. is the sum of ûee ions and all individual complexed metal species. [WT =

where

bL is the

W I L K , [W"[Ln-]' +

(3)

formation constant of the metd complex M,L:"'-ln'; rn and 1 are

stoichiometrk coefficients.

1.3.2. Methods of determination

There are two categorks ofmethods available to determine speciation oftrace metals.

nie f h t one, the expeimental approach, hcludes various andytical techniques such as ion-

selective electrode (ISE) (Hirsch and Banin. 1990: Sauvé et al.. 1997a). anodic stripping voltammetry (ASV) (Sauvé et al., 1997b), competitive chelation (Workman and Lindsay. 1!NO), Doman dialysis membranes (Cox et ai., 1984; Fitch and Helmke. 1989) and ion

exchange resins (Chakrabartiet al.. 1994; Holm et al.. 1995). These techniques measure the concentrations (or activities) of fiee and labile metals. They also cm differentiate strongly and weakly labile complexes. Although progress has been made in sensitivity and accuracy, these techniques may be limited by the detection limits of analytical equipments and interferences h m other elements when they are applied in soil system (Nordstrom. 1996).

When the analytical techniques are not available, computer programs provide an alternativeapproach to estimate met al speciation. Examples include MPIEQL- (Schecher and McAvoy. 1994). GEOCHEM (Mattigod and Sposito. 1979). MINTEQA? (Ailison and Brown. 1995) and WHAM (Tipping. 1994). These programs contain themodynamic databases of metal-ligand reactions. Soi1 solution components. including pH, electrical conductivity ( EC),cations, anions and dissolved organic carbon ( DOC),constitute data input to the programs to calculate metal speciation. The underlying assumption of this approach is that reactions among soil solid and liquid phases are at equilibrium. However. some reactions such as dissolution and precipitation of soil minerais are so slow that true equilibrium is not attained and kmetic factors have to be taken into account (Lindsay, 1979). Consequently, the validity of the assumption should be Mewed with caution.

Another major weahess of the computerprograms is that many complexes, especially

the organo-metallic compounds. are not calculated accurately due to the lack of thermodynamic data for metal cornplexation reactions with naturally occurring organic acids. Several models used to solve this problem are discussed as follows:

1. MLrnrre mode!. More1 et al. ( 1975) introduced the concept of the 'mixture model'

in a study of trace metal behavior in sewage sludge. This model contains a set of organic acids whose acidic functional groups are known to e.kt in the sewage sludge and whose reactions with metal cations are well characterized. The concentrations of these organic acids are proportional to that of DOC in a hlvic acid solution to be modeled. The advantage of this model is that it c m be applied in a wide range ofpH and ionic strengths. Also, this model does not require the measurement of stability constants for metal-organic acid compounds.

However, the results are questionable because the selection of the organic acids is arbitrary and may not represent the real soil conditions.

2. Quasiparticle model (Sposito et ai., 1982). This model assumes that the actual soil organic acids are an assemblage of hypothetical macromolecules. Their trace metal complexation reactions can be described ushg stability constants detemiined by poteatiometric titration (Sposito, 198 1). Then the stability constants for metal-fulvic acid compounds are incorporated into the thermodynamic equilium programs. However, the utilization of these constants is restricted to the solution conditions (e.g., pH and ionic strength) in which the constants have been determined. This is because the resuhs were fiom experiments performed at 6xed pH by mcremental additions of metai cations to a solution of

hlvic acids.

3. Nonelecnostatic sir@xe cornpieration mode!. Van Benschoten et al. ( 1 998) took Pb as a example to develop a model which described the Pb binding at minera1 surfaces. It is

assumed that the binding take place at single diprotic surface site. Only 3 fitting parameters are required: total site density (S,). surface-binding constant (&,)

and =. which is the dope

in the plot of log~bd'pb'' vs. pH (Pb, meam Pb adsorbed on mineral surfaces). This mode1

is simple; however, it is intended for engineering purposes rather than explmation of metal sorption on minerai surfaces.

4. WNidmereHmic Aquatic Model (WHAM) (Tipping, 1994) The WHAM Sicludes

a metal binding model (Model V) and inorganic chemistry models. In Model V metal complexation is assumed to be at discrete sites (Tipping and Hurley, 1992). Two types of

.

binding sites. namely 'monodentate' and 'bidentate' are involved in met al binding. The

descriptive parameters such as i n h i c constants of fulvic acid groups are derived kom published data for fùlvic-type material on proton-dissociation and metal binding . This model

is more sophisticated than the above ones as it assumes that proton-dissociation takes place at eight sites (Tipping and Hurley. 1992). Therefore, in this thesis we chose the WHAM to estirnate the metai organic complexation.

1.3.3. Controllhg factors on metal speciation

The partition of dissolved metals is primarily affecteci by the acidity of soil solutions

(Ross, 1994). M e n soil solutions are acidic. metals predomlliantly speciated as k e metal ions. For example, Dang et al. ( 1996) reponed that the proportion of fke Zn ions was up to 7 1.8Y0 when solution pH was 6.4. The experiments conducted by Reddy et al. ( 1995)also demonstrated that Cu2-constituted 30% of total dissolved Cu. As the solution pH inmeases, more met& fonn complexes with OH', CO:" and HC0,-. In a study of metal speciation in

alkaiine soi1 leachates, Lamy et al. (1994) found that the proportions of hydroxide or carbonate species (MOHS', MCO,"' or MHCO,"')exceeded 90%.

Organic matter in soi1 solutions may also influences metal speciation. Some metals such as Fe and Cu have a strong tendency to form soluble compounds with oxygen-containhg functional groups of dissolved oqanic matter. It has been s h o w that these two met& cm exist almost completely in organic complexes (McBride and Blasiak, 1979; James and Bouldin, 1986). In addition, the dissolution of soil organic matter is enhanced at neutral or

alkaline pH and therefore more dissolved metds are organicaily complexed. For example, increased concentrations of DOC lead to more Cu in the soi1 solution as Cu-organic complexes (Berggreu. 1992).

Metal speciation is also controlled by the reactions between soii solid and liquid

phases (Ross, 1994). At a low level of dissolved metais, adsorption of metal ions ont0 oxides and organic matter may control the metal solubility. When metal loadings are high and the soi1 is alkaline, precipitation processes may begh to control the concentrations of met& in soil

solution (McBride, 1989).

Soil minerais have been proposed as the ultmiate conmolhg factor of metal solubility

(Lindsay, 1979). Based on this assumption Emmerich et of.( l982b) attempted to identi& the solid phase which controls the activities of metal ions in soil solutions. The results. however. indicated that the activities of fke met& in solution were under-saturated with respect to the least soluble minerais. So it is unWrely to find pure minerais which control the concentrations of metal ions. McBride ( 1989) ascribed the difficulty of mineral identification to the complex soil matrix, in which other mechanisms such as chemisorption and coprecipitation are probably more important in controlling the metal solubility.

1.4. Metal bioavailability

The term 'metal bioavailability ' refers to the quantity of met Jwhich can be released into soil solution and taken up by soil biota. Since plants are able to adapt to variable soil properties and accumulate trace metals, they are considered as intermediatereservoirs in soilp l a n t - h a 1 (or human) systems ( Kabata-Pendias and Pendias. 1993). Therefore, plants have been used to test the metal bioavailability in soils.

1.4.1. Metal pools and plant uptake It has been accepteci that plants mainly t l e up met& fiom soil solutions (Mengel and

Kirkby, 1987). Moreover, the metal fiactions associated with carbonates, oxides, organic matter and mllierals may become available after they are released fiom the soil solid phase (Kabata-Pendias and Pendias, 1992). Consequently, metal bioavailabdity cm be characterized

by concentrations of dissolveci merals and the buffering capacity of metal pools in the solid

phase (Loneragan, 1975).

in order to quanti& available metal pools and to evaluate their predictability on plant

uptake, selective extractions have b e n performed on a variety of soils (Haq et al., 1980;

Valdares et al., 1983). The method used in these studies is to extract certain metal fraction(s) ushg chemical reagents and to link the extractable metals with plant uptake. Dilute acids.

weak saits and chelating agents have been employed m rnetai extraction. By using simple or multiple linear regression analysis, it has been shown that the concenvations of the extractable metals are correlated with the rnetai content of plants (Walsh et al.. 1972; Symeonides and McRae, 1977; Xian, 1989;Qian et al., 1996). However. the selective chemical extraction approach has k e n criticized because of its disturbance of the soil properties and possible redistributions of metal fractions during the process of extractions (Beckett, 1989). Furthemore. t hese results are only usehl in particular soil-plant systems and their extrapolation to a wide range of soil conditions is not straightfonvard.

On the other hand, the relationships berween metal speciation in the soil solution and plant uptake have been another focus in metal bioavailability research. Evidence fkom literature of aquatic to xicology has demonstrated that responses of single-ceiled plants like algae are correlated with concentrations of fiee metal ions (Tessier et al., 1994). In

hydroponic solutions where chelates were added, metai content in higher plants could be predicted by the concentrations (or activities) of fiee (uncomplexeci) metal ions (Checkai et

al., 1987). These results codormed to a 'kee ion activity mode1 (HAM)'(Parker et al.,

1995). Therefore. it would be plausible to try to apply this model to plant roots in soi1 solutions.

The theoretical foundations of the FïAM have been reviewed by Parker and Pedler ( 1 997).

Free metal ions ( M) react with not only ligands in solution ( L) but also binding sites

on ceii membrane (-X).The HAM is developed based on these two hypothetical equilibnum reactions: M+L*MkkIL

(4)

L=[

(5)

~ ~ I / ~ ~ ~ [ L I )

M + -X* M-X &x

= [M-XI/([W[-Xl)

(6) (7)

The key assumptions underlying the FIAM include (Campbell, 1995): a. Free ion is the only metal form to cross the celi membrane on plant roots. b. The plant response to metal is proportional to the quantity of the M-X cornplex.

c. Concentration of the fkee metal ion, [Ml, depends on the reaction between M and L. d. Concentration of the ce1 binding site, [-XI, remains constant afier reactions with M.

Although the FiAM has k e n supponed by numerous studies (e.g .,Pavan et ai., 1982; Checkai et ai., 1987; Bell et ai., 1 99 1 ;Parker et al., 1995;Sauvé et ai., 1W6),there remah some uncenainties with this model, especialiy when it is applied in soi1 systems. Fîrst, it is assumed that plam response is dependent on [W].However, ML may also play an important

role in the process of metal uptake. For example, the amount Fe3- in soi1 solutions is fm

below that required by plants. ûrgano-Fe complexes. mainly cheiates. are the dominate species difhsed to plant root cell to supply Fe'- (Mengel and Kirkby, 1987). So. the ratio [Fe3*]/[Fe-chelate]is close to zero. In this case the FIAM may be questionable because plants may not respond to srnall concentration of Fe-.

Second, fiee metai ion rnay not be the only f o m to be taken up by plants. In solution culture experiments, Smolder and McLaughlin (19964b) found that, other than Cd2*, Cd chloride complexes (CdCL"") also contributeci to plant uptake. From the viewpoint of plant

physio logy, the above observation can be attributed to breaks m the endodemal banier at the root surface and uptake of intact metal-ligand complexes (Smolders and McLaughlin, 1996 a, b).

Third the FIAM may not be applicable without taking rhizosphere effeft into consideration since plant root exudates rnay change chernical composition of rhizosphere solutions (Merckx et al., 1986). For instance, Lorenz et al. ( 1997)reported that less Cd and

Zn were present as Cd2*and Zn2+in rhizosphere solutions due to an increase of DOC.As a result of the modified metal speciation, they concluded that the correlation between Cd cûntent in radish and Cd speciation in non-rhizosphere solutioos was not statistically significant.

1.4.2. Eff'ects of soi1 properties on meril uptake

Metai supply to a plant's roois may not simply be a fiinction of the concentration of

some particular chemicai fiaction in soils because many physico-chemical processes affect the ilismbution of trace metals. It is recognized that soil chernical properties such as pH. organic matter and exchangeable cations exen strong influences on metal availability to plants (Davies. 1992;Smith, 1996).

Soii pH has been considered as the most imponant factor because it controls metal solubility. i.e., supply of metals from the soiid phase to the soil solution. As soi1 pH increases trace metals are ofien adsorbed or precipitated on the surface of carbonates, phosphates and insoluble hydroxides (Kiekens. 1984). Futhermore, concentrations of freemetal ions in soi1 solutions also decrease as the result of formation of inorganic complexes such as MCO,"' (Reddy et al., 1995). Hence, rnetal availability to plants can be reduced by treatments to increase soil pH (e.g., W g . ) .

Besides soil pH. organic maner is ais0 important to metai solubiliiy in soils. In general, the composition of soil organic matter is dominated by large molecular weight humin

and hurnic acid (HA) compounds and lower molecular weight fulvic acid (FA). nie metalorganic matter interactions affect metai solubility in two ways. On one hami, organic matter in the solid phase, especialiy humic compounds of high molecular weight, snongly retains

metals in soil and thus reduces the metal availability to plants (Ross, 1994). ûn the o t h a

hand the formation of Organo-metallic compounds in solutions may increase the availability of trace metals to plants as these compounds can enhance the diffusion of metals through the unstirred zone arouad the plant roots (Checkai et al., 1987).

In addition, the concentrations of exchangeable or dissolved major cations (Ca. Mg and K)in soil are much higher than that oftrace metals. They may compete withtrace metah for absorption sites on root surfaces and thus reduce metal uptake. For example. in an experirnent conducted in nutnent solution, John (1976) pointed out that an Bicrease in solution Ca reduced Cd concentrations in lettuce roots.

The effects of soi1 properties on metal uptake by plants can be evaluated by multiple

regression analysis. Usualiy the prediction of plmt metal content is improved when soil factors such as pH. organic matter or exchangeable cations are included as independent variables ( Davies, t 992).

1.5. Research needs

In urban regions, soils may contain considerable amount of trace met& as a result of industrialization (Chen et cil., 1997; Lavado et al.. 1998). Although the total metal burdens in urban soiis are usuaîiy high, in rnost situations these metals are unavailabie because of the

alkaline soil pH, which is ofien caused by the release of carbonate from construction materials (Jim, 1998). Shce higher plants mainly take up metals which are in available f o m (e.g., dissolved metah), it would be reasonable to study metal availability to plants in order to

ascertain the potential ihreat associateci wit h the met al contamination. Unfortunately, not much work has been carried out in this area.

Soi1solution is considered as a medium kom which higher plants absorb trace met&

(Ewber, 1984). Hence, characterihg chemical composition of the soil solution is a primary concem when metal uptake by plants is assessed ( Kmght et al.. 1998). Free metal activity has been shown to be correlated weU with plant response in most cases of solution culture experiments. It would be worthwhile to test this notion in the urban soils in order to identify the metal pools which are significantly correlated with plant uptake.

Furthemore, metal content in plant tissues is largely a function of binding of the metais in soil soiid and liquid phase. Therefore, in addition to identifjmg the metal pools and

linking them with the plant uptake, it is essential to have a Fundamental understanding of the mechanism~of metal binding on various soil components such as clay miner&. Fe or Mn oxides. solid oqanic matter and dissolved organic ligands in order to assess the bio availability

of uace metals (McBride, 1989).

CHAPTER 1. METAL SPECIATION AM) AVAILABILITY TO INDICENOUS PLANTS IN THREE RAILWAY YARDS

2.1. Abstract

Urban soils O fien contain concentrations of trace metals that exceed regulatory levels. However, the threat posed by trace metals to human health and the environment is thought to be dependent on their speciation in the soi1 solution rather than the total concentration.

Three inactive railway yards in Montreal, Quebec, were sampled to mvestigate the speciation and bioavailability of Cd, Cu, Ni, Pb and Zn. Soil solutions were obtained by cennifuging

saturated soil pastes. The metal speciation in soi1 solutions was estimated using the Windemere Humic Aqueous Mode1 (WHAM). In the saturation extracts. up to 59% of the dissolved Cd was in a fke-ionic fom. For Cu, Pb and Zn,organic complexes were the predominant species. Over 40% of Ni was present as inorganic complexes if the solution pH exceeded 8.1. Multiple regression analysis showed that pH and total metals in soil were significantly correiated with the activities of fke metals, except for Cd2-, which only had a weak correlation with soil pH. Free, clissolved and total soil rnetals were tested for their ability

to predict metal uptake by plants in the field. However, none of these metal pools were satisfactory predicton. The resuhs mdicated that in these urban soils, trace metals were

mainly in stable f o m and bioavailability was exnemely low.

K w r & : contaminateci soiis; metal speciation;fiee ion acuvity; soil property;bioavailability

2.2. Introduction

Trace met& such as Cd, Cu. Ni, Pb and Zn are cornmon pollutants in urban industrial soils (Dudka et al.. 1996). These elements are ofien adsorbed or occluded by carbonates. organic matters. Fe-Mn ondes and primary or secondary minerals (Adnano. 1986; Ross, 1994). Since plants take up most nunients fiom the soi1 solution, it is ofien assumed that the

dissolved trace met& are readily available to organisms (Barber, 1981). Determination of total dissolved met&

and their chernical speciation may provide usefùl information on rnetal

bioavailability and toxicity ( Knight et al., 1998).

Trace metals are distributed in the soi1 solution as Bee ions (M2+).inorganic and organic complexes. It is generaily accepted that, in the short tem. plant growth and rnetal uptake are related to fiee ion activities (Sparks. 1983; McBnde, 1994). Ln experiments conducted using soi1 or nutnent solution, the free ion mode1 \vas able to predict metal uptake by plants (Pavan et al.. 1982; Bell et al., 1991).

Both analytical and computational approaches are available to quant@ M" activities.

Although improvement has been achieved in M2+detennination using an ion selective electrode (ISE) or anodic stripping voltammeay (ASV), these techniques are not available for all trace metals. An alternativemethod is to measure pH, concentrations of major cations,

anions, DOC and trace metais in soil solutions. Then the M2-activities are estimated using chernical e q u i l i i models such as the WHAM (Tipphg, 1994).

So far. Little attention has been paid to trace metal speciatioa and bioavailability in

urban soils. especially the relationships between the amount of metal taken up by indigenous plants and fiee ion activities. Thus the objectives of this paper are to: 1) estimate metal speciation in urban soils; 2) identify the factors controhg M2- activities: and 3) test the applicability of the kee ion modei to metal uptake by plants in urban industrial soils.

2.3. Materials and methods 2.3.1. Site description

The experknent was carried out at three inactive railway yards (Yards 1. 2 and 3) close to the downtown of Montreal in 1997. Yards 1.3 and 3 had been inactive since the late 1960's. mid 1970's and late 1980's. respectively. Yard 1 had the greatest variety of plant

species including trees, grasses and herbaceous species. The rail tracks had been removed but the wooden ties and cernent platforms were still visible. Vegetation on the second yard was mai*

composed of grasses and herbaceous species. Several rail tracks were O bserved in

a small part of the sampling area. in yard 3. a large amount of construction materials such

as cement and bricks were found in the soil profle and the whole site was sparsely covered

in g r a s and herbaceous vegetation.

2.3.2. Soü and phnt sampüng

In each site. eight sampling points were chosen dong a transect of approxmiately 100 m with 12 m intervais. Soi1 samples were taken from the surface layen (0-15 cm), sieved to 2 mm on site, and transporteci to the laboratory. One sub-sarnple was stored at 5 "C for

saturation extracts while the other was air-dried and ground to 0.5 mm for standard soi1 and total rnetal analysis.

To detemine the metal content of plants grown under the field conditions. bulk above-ground vegetation sarnples were haxvested using a 50 cm x 50 cm wooden grid at the 23 points due to the lack of vegetation at one point of Yard 2. The samples were oven-dned

(60°C).weighed to determine the biomass and ground to 0.5 mm for analysis of total metals.

2.3.3. Soi1 characteristics

Electrical conductivity (EC) and pH were measured with a soil:water ratio of 1 :2 (Hendershot et al., 1993a). A 0.1 M BaCl, extraction was used to analyse cation exchange capacity and exchangeable cations (Hendershot et al., 1993b). The CaCOt content was determined by the method of Bundy and Bremner ( 1972). Organic carbon was measured by the wet oxidation method (Tiessen and Moir, 1993).

Total metals (Cd, Cu. Ni, Pb, Zn) in soils were determined folowing a HN0,-HCIO, digestion (Cook, 1998). One gram of air-dried soil ( ~ 0 . 5mm) was digested with 10 mL

HNO,and 5 mL HClO, at 150°C for at least 16 hours. The digests were then diluted to 100 mL and analysed on a flame atornic absorption spectrometer ( A M ) for total maals. A quality-contro1 soil sample (NIST 27 10) was andysed using this method and the average recoveries for the duplicate samples were 95.0%, 96.3%, 122%, 98.2% and 88.5% for Cd,

Cu,Ni, Pb and Zn,respective&.

2.3.4. Soil solution analysis Soil solutions were obtained ushg the centrifugemethod modified korn Adams et al. ( 1980). Saturated soil

pastes were prepared using 100 g of field moist soi1 and deionLed

distdied water. The pastes were equilibrated for 2 houn and centrifuged ( 130 g) for 1 hour to extract the soi1 solution, which was then filtered through 0.4 Pm polycarbonate membranes. Analysis on the fihered solution were as follows: major cations (Caz*.Mg2- and

KT)by flame AAS, anions (Cl-, NO,', SO,") by ion chromatography. DOC by colorùnetric absorbance at 254 nm, pH and EC by a glass electrode pH meter and conductivity meter, respectively. The concentration of DOC was calculated fiom the colorimetnc absorbance at

254 nm with the foliowing equation: C (mg Li)=22.1 1 a absorbance (n=141 ;R' = 0.860***). This equation was developed 6rom the cornparison between the absorbance at 254 nm and the K2C-O. oxidation method (Moore, 1985). To prevent precipitation of the trace metals, reagent grade Na,EDTA was added to sub-samples of the saturation extracts giving a final concentration of 0.00 1 M. The total concentration of dissoived metals (Cd. Cu, Ni, Pb, Zn) was analysed on a graphite h a c e AAS.

2.3.5. Piant aaalysis Plant samples were digested by HNO, and 30% H20,(Jones. 1991). The extracts were analysed by flame AAS for total Zn. Total Cd, Cu, Ni and Pb were determined by graphite h a c e AAS due to their low concentrations.

2.3.6. Speciation cakuhtionr and statistical malysis

Metal speciation was calculated by using the WHAM, which consists the Humic IonBinding Mo del V and models of inorganic solution chemistly (Tipping. 1994). In the Mode1 V, it is assumed that the metal binding takes place at a number of discrete binduig sites on

dissolved organic matter (Tipping and Hurley, 1992). Data of soil solution chemistq constituted input files of the model. The concentrations of fiee metal ions were convened to activities by using the Davies equation (Davis, 1962). Statistical analpis were made with the

SYSTAT sofnvare (Wilkinson, 1992). The ievel of statistical siçnificance is represented by

* for 0.05. ** for 0.0 1, *** for 0.00 1 and ns for >O.10 (not significant).

2.4. Results and discussion

2.4.1. Soi1 characteristics The pH of the soils ranged from 6 -39 to 7.88 in the first yard and was more alkaline

throughout the other two yards (Table 2.1 .). This c m be attributed to the hiph content of carbonate gravel. ash and cinden. ûrganic carbon concentration exceeded 10% at points 1 to 5 o f yard 1. The cation exchange capacity (CEC)varied in the range of 6-22 crnol(+)kg"

with exchangeable Ca king the dominant cation throughout the yards.

2.4.2. Total metals in soils and soil solutions Québec soil criteria establish acceptable b i t s of potentially toxic substances for residential, industrial or commercial use and recommead soil remediation work if necessary (Ministère de l'Environnement du Québec, 1990). According to these criteria, level 'A'

represents the background concentration of trace metals in sob. Level 'B' soils need

Table 2.1. Selected soi1 chernical characteristics -

Yard

Point

8

CaCO;

pH

CEC

Ex. Ca*

Organic Carbon

O h

(H@)

cmoI(+1 kg"

cmol(+) kg-'

9;

18.1

8 -29

*: Ex. Ca means exchangeable Ca

11.7

10.1

1.39

Table 2.2. Total metals in soil sarnples and Québec soi1 criteria (mean~standard deviation.

Cu

Ni

Pb

Zn

.

40W35

84.5k2.1

973135

853110

3

10.5*1.2

41 7130

72.317.7

9661 1 1

589113

3

13.3k0.3

38315.5

57.8*1 .O

68Oî22

9341 16

4

9.8i0.7

303* 1.1

116118

505123

527k27

5

7.110.1

19915.9

64.118.6

521k9.0

797129

6

1.910.3

58.711.5

48.815.4

90.417.5

23b20

7

1.7510.05

64.1123

46.315.2

14611.9

240146

8

1.910.3

66.8k2.2

25.6*4.1

157110

20312 1

1

3.4î0.2

1 181 1.4

68.î+2.7

236k8.3

3091 16

I

3

3.8k0.2

92.810.7

41.610.7

23316.2

21516.2

3

2.510.7

120~5.3

55.8k2.5

24416.2

239147

4

1 .&O.?

3 14172

129110

678149

2761 19

5

5.410.05

476k6.0

24312.6

76416.0

135&5

6

1.810.5

246k5.2

75.2I2.1

6561113

497137

7

2.8i0.1

310I9.6

69.8k1.0

456k9.3

66218

8

2.610.05

28512.5

92.312.9

51311 1

6261 18

1

10.611.0

181*14

1 19W192

351114

-

73.610.3

7

2.8i0.4

1 1816.6

5 3 . M .9

484~87

366*37

3

5.910.1

342I19

1 1416 .O

646*94

701k16

4

5.1k0.5

145k9.8

56.4I4.3

313133

294k 14

5

3.9*0.3

19617.0

61.1î1.7

373S

447123

6

9.21.6

1122.5

56.911.8

409*36

379*19

7

7.0I0.5

186*12

64.t k0.2

474*51

54ik40

8

5.0I0.5

8 1.312.9

3 7 . a 1.9

234*2 1

203* 18

1.5

50

50

50

100

Soil

A B

5

100

1O0

500

500

Criteria

C

20

500

500

1000

1500

Yard

Point

1

1

1O

2

3

Québec

Cd

25

thorough analysis and may require restoration work. Rompt remediation is necessary if the metd content in soi1 is above the 'C' level. In the railway yards studied, ail five met& (Cd. Cu. Ni, Pb, Zn) were in the 'A-C' range except for one samphg point in yard 3 where Pb exceeded the 'C' level (Table 2.2.). These results indicate that the three sites have not beeu severely contaminated by the five metals. For comparison. in the survey conducted by Holmgren et al. (1 993) in agricultural soils of the USA. the arithmetic mean concentrations of Cd Cu, Ni, Pb and Zn were 0.265, 29.6.23.9. 12.3 and 56.5 mg kg-'.respectively. which are rnuch lower than those in the present study. in addition, Lorenz et al. ( 1997) studied ten contaminated European soils and found average concentrations of Cd, Cu, Ni, Pb and Zn were 5.0 1, 55.3, 25.6, 1 107, and 10% mg kg". respectively. In comparison with the mean

concentrations of Cd. Cu, Ni, Pb and Zn obtained here (5.43, 2 17, 75.3, 497, 489 mg kg", respectively), their soils have been more severely contaminated with Pb and Zn.

The concentrations of dissolved trace metals (Table 2.3.) were generally below the

'B' level according to the Québec guideline of water quality (Ministère de l'Environnement

du Québec, 1990), however, at three points in yard 1, the values for Pb were in the 'B-C' range. Holm et al. (1998) reponed that the concentrations of dissolved Cd and Zn in 15 contaminated soils were up to 17.0 and 3600 pg L" , respectively, which are much greater

than the values in this study. Interestingly, total Cd and Zn content in their soils (0.18-1 6.5 and 36-1 320 mg kg',respectively) are gerierally lower than those in the present study. This

implies that the solubility and mobüity of Cd and Zn are lower in the railway yards under investigation.

Table 2.3. Selected chernical propaies of soii solutions and Québec guidehe of water

Point

1

1

(H,O) 7.78

mg L-'

15.3

---1.49

----49.0

pg

L-1

7.80

-

- ----42.5

93.5

Table 2.3. (Cont.) Selected chernical properties of soi1 solutions and Québec guideline of water quality

Québec

A

1

25

10

10

50

rruideline of

B

5

500

350

50

1000

1O0

5000 I O000

C

water quality C 20 IO00 *: hl means mol L-' ; (i: MDmeans dissolved m e t h

2.4.3. Meta1 speciition and minera1 control The activity of Cd2*in the soi1 solution varied from 10">-'to 1

mol L" (Table 2.4.).

The percentage of Cd2*to total dissolved Cd was fkom 3 1.4 to 59.2%. Sunilarly, Hirsch and

Banin ( 1990) reported that Cd2' constituted 4045% of the total Cd in the solutions 60m the top 5 cm of forest soiis. In the current study almost no Cu'- existeci in the soi1 solutions and orçano-Cu complex (CuFA) was predombant. This pattern of Cu distribution tvas also pointed out by McBride and Bouldin ( 1984), who found increasing the pH to approximately 6 in peat extracts resulted in alrnost cornplete complexation of Cu. The major species for Ni

in soil solutions was NEO; when pH was over 8.1 and reached a maximum of 59.4% of total dissolved Ni at pH 8.4. Similar to Cu, Pb and Zn were strongly bound by FA and the percentages of PbFA and ZnFA were up to 99.3% and 92.6%, respectively. The significant metal organic complexation may be atmbuted to the high concentration of MX: (maximum 30.4 mg C L-',Table 2.2.) in the saturation exnacts.

Soi1 miner* are believed to be the ultimate connol of the solubility of trace met& (Lindsay, 1979); however, the complex soil matrix often rnakes identification of these mînerals difncult (Manigod et al., 198 1). Solubility diagram (Figure 2.1 .) showed that the

Table 2.4. Metal speciation* in extracted soil solutions

Species

Yard 1

Yard 3

PM

PM

8.63-9.50 40.6-6?.3

CdL,

9.81-11.3

PM

9.01-9.70 3 1.9-49.1

8.00-10.1 21.2-43.0 8.91-9.59

0.95-6.42 9.49-10.7 2.68-14.3 10.2-10.9

---

?/o

mol L-'

8.74-9.80 3 1.4-56.6 8.71-10.1 39.0-59.2

CdFd4

CU"

Y0

mol L"

mol L-'

Cd"

Yard 3

36.3-48.1

1.88-4.30

. .

9.55-1 1.4