This Ph. D. thesis presents results and conclusions from studies 1) investigating the interaction between metal and metalloid ions and thermophilic bacteria, and 2) characterizing microbial populations in a geothermally active habitat with relatively high concentrations of metalloid ions and compounds. In initial cadmium ion toxicity assays, the minimal inhibition concentration for 46 thermophilic bacteria of the genera Aneurinibacillus, Anoxybacillus, Bacillus, Brevibacillus, Geobacillus, and Thermus were determined. The highest tolerances to cadmium ions (Cd2+) in the range of 400 to 3200 micro;M were observed for species belonging to the genus Geobacillus. The thermophilic Gram-positive bacteria Geobacillus stearothermophilus and G. thermocatenulatus were selected to describe further biosorption reactions between cadmium ions and chemically reactive functional groups (potential ligands) within and onto the bacterial cell walls. Data obtained from electrophoretic mobility, potentiometric titration and cadmium ion adsorption experiments were used to quantify the number and concentrations of ligands and to determine the thermodynamic stability constants for the ligand-cation complexes. The first reported surface complexation models (SCMs) quantifying metal ion adsorption by thermophilic microorganisms predicted cadmium adsorption and desorption by both studied Geobacillus strains over a range of pH values and for different biomasses. The results indicated the functional group, with a deprotonation constant pK value of approximately 3.8, to be more dominant in cation biosorption accounting for 66 and 80% of all titrable groups for G. thermocatenulatus and G. stearothermophilus, respectively. The generated SCMs are different from model parameters obtained from mesophilic species that have been studied to date and might indicate a different biosorption behavior for both studied Geobacillus strains. Another objective of this thesis was to characterize microbial populations in the hot spring Champagne Pool, located in Waiotapu, New Zealand. The thermal spring is approximately 65 m in diameter and discharges water at 75eg; C and pH 5.5, which is oversaturated with arsenic and antimony compounds that precipitate and form orange deposits. Recovered nucleic acids and adenosine 5'-triphosphate (ATP) concentrations obtained for Champagne Pool water samples indicated low microbial density and were in good agreement with relatively low cell numbers of 5.6 plusmn; 0.5 x10^6 cells per ml. Denaturing gradient gel electrophoresis (DGGE) and 16S rRNA gene clone library analyses revealed the abundance of Sulfurihydrogenibium, Sulfolobus and Thermofilum-like populations in Champagne Pool. Two novel bacteria and one novel archaeon were successfully isolated with a distant phylogenetic relationship to Sulfurihydrogenibium, Thermoanaerobacter, and Thermococcus, respectively. Genotypic and metabolic characteristics differentiated isolate CP.B2 from described species of the genus Sulfurihydrogenibium. CP.B2 represents a novel genus within the Aquificales order, for which the name Venenivibrio stagnispumantis gen. nov., sp. nov. is proposed. V. stagnispumantis is a thermophilic, chemolithothrophic bacterium, that utilizes molecular hydrogen as electron donor and oxygen as electron acceptor and displayed growth in the presence of up to 8 mM NaAsO2 (As3+) and more than 20 mM Na2HAsO4.7H2O (As5+). However, growth was not observed when Na2HAsO4.7H2O and NaAsO2 were provided as the sole electron acceptor and donor pair. Arsenic resistance was conferred by the genes arsA and arsB

Les processus rhizosphériques sont reconnus comme une des stratégies majeures élaborées par les plantes afin d'augmenter la disponibilité en phosphore (P) et ainsi améliorer leur nutrition phosphatée. Pourtant, l'effet exact de ces processus est encore mal caractérisé et quantifié. L'objectif de ces travaux a été d e déterminer le rôle exercé par les modifications chimiques induites par les racines, particulièrement la modification de pH, dans les changements de disponibilité en P dans la rhizosphère. Pour ce faire, nous avons utilisé des modèles mécanistes géochimiques (« triple plane », échange d'ion et Nica-Donnan) en considérant une approche additive pour simuler l'effet de l'activité racinaire sur la disponibilité en P. Dans une première étape, nous avons caractérisé l'effet du pH sur la disponibilité en P dans plusieurs sols, un Cambisol et un Luvisol. Le Luvisol présentait deux concentrations en P inorganique contrastées en raison d'un essai de fertilisation phosphatée longue durée. Dans la rhizosphère du blé dur (Triticum turgidum durum L.) cultivé sur les mêmes sols, nous avons caractérisé qu'en plus de l'alcalinisation, le prélèvement en P et surtout en calcium (Ca) sont les processus rhizosphériques responsables du changement de disponibilit é en P observé. Le prélèvement du Ca favorise l'augmentation de la disponibilité en P dans la rhizosphère, en diminuant l'effet promoteur du processus d'adsorption-désorption du Ca sur celui du P. L'influence relative de ces trois processus rhizosphériques dépend toutefois de la composition chimique de la solution du sol (concentration en Ca et pH en particulier). Nos simulations mettent également en évidence la relation entre les changements de disponibilité en P est la distribution du P adsorbé sur les différentes phases minérales. La minéralogie du sol, spécialement l'abondance relative d'illite vs. les oxydes de fer, contrôle l'influence des processus rhizosphériques en déterminant les minéraux impliqués dans l'adsorption du P. A travers l'identification d'un nouveau processus rhizosphérique découlant du prélèvement en Ca et de ses effets sur la disponibilité en P, nos résultats démontrent la validité des modèles géochimiques pour prédire l'influence des processus rhizosphériques déterminant la disponibilité en P. / Root-induced chemical processes are recognized as a major strategy developed by plants to enhance phosphorus (P) availability and thus to promote P acquisition. However, the exact influence of these root-induced chemical processes is still poorly understood and quantified. The present study aimed at investigating the influence of root-induced chemical processes, especially root-induced pH changes, on P availability in the rhizosphere. In this work, we used a set of mechanistic adsorption models (« 1-pK triple plane », ion-exchange and Nica-Donnan) within the framework of the component additive approach in order to simulate the effects of root activity on P availability. First, we described the effects of pH on P availability in several soils unaffected by roots, a Chromic Cambisol and a Luvisol. The Luvisol showed different concentrations in inorganic P because of a long-term fertilisation trial. In the rhizosphere of durum wheat (Triticum tu rgidum durum L.) grown on these two soils, we found that calcium (Ca) uptake, in addition to P uptake and root-induced alkalisation, controlled to various extents the changes of soil P availability. Calcium uptake markedly increased P availability by decreasing the promoting effect of Ca adsorption on P adsorption. The relative influence of these three root processes depended on the solution composition (especially concentration of Ca and pH). Our simulations showed the relationship between changes in P availability and the speciation of adsorbed P onto the different soil minerals. Soil mineralogy, especially the relative abundance of illite vs. Fe oxides, controlled the influence of root processes by regulating the contribution of soil minerals to P adsorption. By identifying a novel root-induced processes, namely the Ca uptake, and describing its influence on P availability, our results demonstrate the ability of surface complexation models to predict the effects of root-i nduced processes on P availability in soils.

Diffuse Layer Modeling was used to describe single and multi-solute adsorption of Pb(II), Cu(II), Zn(II) and Cd(II) to ferrihydrite and As(V), V(V) Si and Ca(II) on granular ferric hydroxide, a commercially available iron oxide. Macroscopic data were used in conjunction with x-ray adsorption spectroscopy (XAS) data to evaluate the diffuse layer surface complexation model (DLM) for predicting sorption over a range of conditions. A self-consistent database was created for each of the adsorbents. The DLM provided excellent fits to the single solute data for the ferrihydrite system with the incorporation of spectroscopic evidence. Little competition was seen in the bisolute systems, except under very high coverages. While the DLM captured the lack of competition under low and medium coverages, competitive effects were not adequately modeled by the updated DLM for high coverages. Challenges remain in adequately describing metal removal when sorption may not be the primary mechanism of removal. The capabilities of the DLM were then evaluated for describing and predicting multisolute sorption to granular ferric hydroxide (GFH). The model can adequately describe anion competition, but the electrostatic effects due to outer sphere sorption were overpredicted, leading to an inadequate model fit for As(V) and Ca²⁺ systems. Despite the limitations of the DLM, it may be an appropriate compromise between goodness of fit and number of parameters for future integration into dynamic transport models and thermodynamic databases. / text

Glyphosate (N-(phosphonomethyl)glycine, PMG, H3L) is a widely used organophosphorous herbicide. It interacts with metal ions and mineral surfaces, which may affect its mobility, degradation and bioavailability in the environment. However, these interactions are far from fully understood. This thesis is a summary of five papers discussing the complexation of PMG with metal ions in aqueous solution and the adsorption of PMG and/or Cd(II) on different mineral surfaces. The complexation of PMG with the metals Cd(II) or Al(III) in aqueous solution was investigated with macroscopic and molecular scale techniques. Potentiometric titration data were combined with EXAFS, ATR-FTIR and NMR spectroscopic data to generate solution equilibrium models. In the PMG-Cd(II) system, only mononuclear complexes were formed, while both mono and binuclear complexes were observed in the PMG-Al(III) system. EXAFS, ATR-FTIR, and XPS measurements showed that PMG adsorbs to the surfaces of goethite (α-FeOOH), aged γ-alumina (γ-Al2O3) and manganite (γ-MnOOH) through one oxygen of its phosphonate group to singly-coordinated surface sites. Surface complexation models consistent with these spectroscopic results were fit to adsorption data using the 1pK reaction formalism. Electrostatic effects were accounted using either the Extended Constant Capacitance Model (ECCM) or the Basic Stern Model (BSM), and the charge of the surface complexes was distributed over the different planes. The formation of the surface complexes was described according to the following reactions: ≡MeOH(0.5-) + H3L <=> ≡MeHL(1.5-) + H2O + H+ ≡MeOH(0.5-) + H3L <=> ≡MeL(2.5-) + H2O + 2H+ The coadsorption of PMG and Cd(II) on the surfaces of goethite and manganite results in the formation of ternary mineral-PMG-Cd(II) surface complexes, as suggested from EXAFS results. Previous EXAFS measurements have also established the coordination geometries for the binary goethite-Cd(II) and manganite-Cd(II) surface complexes. In addition to the surface reactions in the binary mineral-Cd(II) and mineral-PMG systems, a single ternary complex with the stoichiometry ≡MeLCd(OH)(1.5-) was sufficient to explain coadsorption data: ≡MeOH(0.5-) + H3L + Cd2+ <=> ≡MeLCd(OH)(1.5-) + 3H+ It was concluded that the affinity of PMG for the three mineral systems decreases within the series: goethite > aged γ-Al2O3 > manganite. The formation of the ternary surface complex is more significant on goethite surfaces than on manganite surfaces.

The goal of this study was to characterize the binding sites on the surface of wheat roots, Triticum turgidum, involved in the adsorption of protons and metals, and quantify the thermodynamic constants needed for a surface complexation model to predict metal binding. The adsorption of protons, Cd(II), Cu(II), and Ni(II) to the root surface as a function of pH and ionic strength in single metal exposure scenarios was quantitatively described using potentiometric titrations, batch metal adsorption experiments, and the least squares fitting program FITEQL. Model predictions from single metal exposures were compared to measured metal adsorption concentrations when roots were exposed to binary and ternary combinations of the metals. Proton dissociation was a function of three discrete monoprotic acid sites on the root surface with log proton dissociation constants of -4.50, -6.23, and -7.37 respectively, upon which varied ionic strength had no effect. The total proton binding capacities for the three sites were 2.58 x 10-4, 1.29 x 10-4, and 2.58 x 10-4 M, respectively. Metal complexation was best described by a two-site model having conditional stability constant log values of 3.04 and 3.30 for Cd(II), 3.21 and 3.25 for Cu(II), and 2.83 and 2.84 for Ni(II) at ionic strength 0.01M. At ionic strength 0.1 M the conditional stability constants log values were 2.37 and 3.36 for Cd(II), 3.11 and 2.56 for Cu(II), and 2.18 and 3.00 for Ni(II). When roots were exposed to binary or ternary mixtures of the metals, the two monoprotic acid single metal model did not provide ideal fits to the data indicating that adsorption in a metal mixture scenario cannot be considered additive and is dependent on the combination of metals present in the exposure environment. The experimentally determined proton dissociation constants and metal stability constants could be used in commercial geochemical speciation programs such as Visual MINTEQ to predict metal adsorption to plants. / Natural Sciences and Engineering Research Council of Canada, The Mining Association of Canada, Ontario Power Generation, Environment Canada.