The roles of adventitious and basal roots were studied experimentally during investigations into the nutritional physiology of watercress. Absorption and translocation experiments with 32p, 86Rb and 59 Fe showed adventitious roots absorb the major proportion of Ions, basal roots a smaller proportion with little direct foliar uptake of nutrients. Field trials demonstrated basal roots could absorb supplementary nutrients from the bed substrate, Improving crop uniformity and production. Watercress cultured in bed simulation tanks at different levels of nitrogen, phosphorus and potassium, responded with increased yields to levels of phosphorus and potassium higher than normally occurs In natural water supplies. Yields In the higher nitrogen solutions were similar to those obtained from natural irrigating waters.

Salt stress is one of the most threatening environmental stresses reducing the global food production. Understanding mechanisms of salt tolerance in halophytic plants is a requirement for developing crop species with increased salt tolerance. This study focused on investigating ion transplant features in a halophytic relative of Arabidopsis, both at physiological and transcriptional level. A comparative approach was adopted in this study using the glycophytic model plant Arabidopsis thaliana, and its halophytic close relative, Thellungiella halophila. Net ion uptake and unidirectional Na fluxes during salt stress were analyzed in the two species. Furthermore, transcriptional profiles of ion transporters under control and high-salt conditions were compared between the two species. The considerable amount of data produced in this study provide important information for future physiological and molecular studies of both Arabidopsis and Thellungiella. The main results can be summarized thus: 1. After salt stress Thellungiella accumulates less Na in the shoots than Arabidopsis. Net uptake of Na into both roots and shoots was slower in Thellungiella than in Arabidopsis. 2. Lower unidirectional Na influx into root cells is the main reason for the lower Na accumulation in Thellungiella than in Arabidopsis. 3. Voltage-independent cation channels (VICs) are likely to be the Na uptake pathway in both Thellungiella and Arabidopsis. 4. Microarray analysis showed that after salt stress both species showed a tendency to reduce Na uptake by decreasing the expression of possible pathways for Na influx. However, transcriptional control of putative Na transporters occurred in Arabidopsis in the shoots, whereas it occurred in Thellungiella in the roots. 5. CNGC8 is a likely candidate for a Na uptake pathway in both Arabidopsis and Thellungiella. Transcript levels of CNGC8 decreased during salt stress in Thellungiella roots and Arabidopsis shoots.

Hyperaccumulator plant species grow in metalliferous soils and accumulate exceedingly high concentrations of metals. They are increasingly studied because of their potential for cleaning up land contaminated with heavy metals, but another aspect of study relates to the reason for hyperaccumulation. The most accepted hypothesis over the last few decades is the ‘elemental defence’ hypothesis, which states that high levels of metals defend the plant against herbivores. Whilst some of the literature is contradictory, some is supportive. An added complication is that many hyperaccumulators belong to the Brassicaceae and produce glucosinolates as organic defences against herbivory. The question to be answered is whether metals or glucosinolates act as the primary defence in these plants and the most recent suggestion is the ‘joint effects’ hypothesis, which states that both classes of chemical work together to benefit the plant and protect it from herbivores. This study investigates these hypotheses and utilized three experimental systems. The hyperaccumulator studied was Thlaspi caerulescens (Gange ecotype) which hyperaccumulates zinc. Plants were grown in a series of glasshouse experiments at a range of soil zinc amendments. There was a positive relationship between soil and foliar zinc; optimum growth occurred at 2000 mg Zn kg-1 soil and this equated to approximately 8000 mg Zn kg-1 shoot, although plants took up as much as 14000 mg Zn kg-1 shoot tissue at higher levels of soil amendment. The herbivore systems studied were generalist thrips (Franklinella occidentalis) and the specialist cabbage whitefly (Aleyrodes proletella). In addition, artificial damage caused by clipping served as a positive control. Four aromatic glucosinolates were extracted from T. caerulescens and two were identified as benzyl and p-OH-benzyl. Glucosinolates were synthesized 32 hours after damage occurred and reached a maximum concentration after 48 hours. Generally, lower concentrations of glucosinolates were observed in plants with higher foliar Zn concentrations and vice versa. However, when plants were subjected to a sustained and heavy herbivore attack, as was the case when thrips infested the plants, glucosinolate production occurred irrespective of foliar Zn concentration. This observation supports the ‘joint effects’ hypothesis, which states that both defences work in tandem and enhance overall defence. Nitrogen was an important component that directed herbivore response. Thrip feeding damage was negatively correlated with foliar nitrogen whilst cabbage whitefly (CWF) benefitted from higher N. Nitrogen was positively correlated with glucosinolate concentrations and glucosinolate content negatively affected the generalist thrips but not the specialist CWF. Data were analysed by accumulated general linear regression and the explanatory model for thrip feeding was C/N ratio + GS + Zn whilst the explanatory model for CWFs was C/N ratio + Zn. Use of the specialist feeder (CWF) allowed for study of the effects of zinc without glucosinolates confounding the results since the CWF was unaffected by foliar glucosinolates. Zinc acted as a defence against CWF but only at high concentrations. The data taken together show that zinc acts as a defence against herbivores that are unaffected by glucosinolates, but only at high concentrations. Zinc also defends the plant against generalist thrips, but glucosinolates are more influential in this case. This might be because of the severe and sustained damage that these plants suffered and systemic effects (i.e. higher concentrations of glucosinolates in undamaged leaves relative to attacked leaves) suggests flexibility in the Zn-glucosinolate relationship. The overall conclusion is in support of the joint effects hypothesis.

Noccaea caerulescens est la principale Brassicacée hyperaccumulatrice de Cd, Ni et Zn, candidate pour la phytoremédiation des sols contaminés. La distribution de l'espèce se caractérise par une importante hétérogénéité des facteurs environnementaux, et plus particulièrement en ce qui concerne la nature des sols. En outre, des variations importantes de la capacité d'hyperaccumulation entre les populations de l'espèce ont déjà été observées. Dès lors, à partir d'un vaste échantillonnage en France et ses régions limitrophes, l'objectif de cette thèse était de : i) mieux appréhender l'écologie de N. caerulescens notamment les composantes édaphiques de son habitat, iii) comparer les capacités d'accumulation en Cd, Ni et Zn et, iii) déterminer les relations génétiques entre les populations. Nos résultats montrent que N. caerulescens est largement répandue dans les massifs montagneux et que l'espèce n'est pas inféodée aux sites métallifères vu le nombre de stations en milieu non minier. Par ailleurs, les stations se caractérisent par une très large amplitude des paramètres édaphiques conduisant à proposer une nouvelle classification des stations. Concernant l'accumulation des éléments en traces in situ, nous avons démontré que la biodisponibilité ne permettait pas d'expliquer seule les concentrations observées dans la plante. Par ailleurs, des cultures en hydroponie ont mis en évidences un compromis entre l'allocation du carbone et l'accumulation des éléments en traces en particulier chez les populations sur serpentines. Enfin, l'analyse de la structure génétique neutre a fait ressortir trois zones géographiques fortement différenciées, sans cohérence avec le type de milieu / Noccaea caerulescens (Brassicaceae) is the main hyperaccumulator of Cd, Ni and Zn, candidate for phytoremediation of contaminated soils. The distribution of the species is characterized by a high degree of heterogeneity of environmental factors, especially concerning the soil composition. In addition, significant variations in the hyperaccumulation ability between populations of the species have been observed. Therefore, from a broad sampling in France and its neighboring regions, the aim of this thesis was: i) a better understanding of the ecology of N. caerulescens including soil habitat components, iii) a comparison of the ability of populations to accumulate Cd, Ni and Zn and iii) the assessment of the genetic structure among populations. Our results show that N. caerulescens is widespread in the French mountains and not restricted to the metalliferous sites given the large number of non-metalliferous stations explored. In addition, the prospected stations are characterized by a wide range in the soil composition leading to propose a new classification of N. caerulescens stations. On the accumulation of trace elements in situ, we have shown that the bioavailability of trace elements was not the only factor explaining the observed concentrations in the plant. Moreover, hydroponic cultures highlighted a compromise between carbon allocation and accumulation of trace elements especially in serpentine populations. Finally, the analysis of the neutral genetic structure highlighted three geographic regions highly differentiated, but not consistent with the edaphic type

Des modèles prédictifs de prélèvement d’éléments traces métalliques (ETM) par des plantes hyperaccumulatrices sont à développer pour rendre la phytoextraction opérationnelle. Ce travail a pour objectif de développer, calibrer et valider un modèle biophysique combiné d’accumulation foliaire et de mise en solution du nickel lors de cultures de l’hyperaccumulateur Leptoplax emarginata sur un sol fertilisé et contaminé en Ni. Une partie de ce modèle intègre un facteur de bioconcentration lié à la transpiration (TSCF) qui caractérise le mode de transport principal du Ni à travers la racine et jusqu’aux feuilles, lors d’une cinétique couplée de production de biomasse foliaire et de transpiration. Sur des plantes intactes et transpirantes, nous avons déterminé un TSCFNi supérieur à 1 du fait : (i) d’une grande perméabilité des racines à la fois au nickel et à l’eau et (ii) d’un transport actif du Ni largement prédominant. A l’opposé, le TSCFNi du blé de Printemps, plante exclusive, était inférieur à 0,02, et le coefficient de réflection correspondant proche de 1, ce qui caractérise des racines perméables à l’eau mais quasiment pas au nickel. L’exceptionnelle capacité de L. emarginata à accumuler et à tolérer le nickel dans ses feuilles, et plus précisément dans ses épidermes, serait également attribuable à ses transpiration et production de protéines soufrées très élevées, tout particulièrement au niveau de ses feuilles les plus jeunes. Enfin, après avoir couplé notre modèle biophysique d’accumulation foliaire du nickel au modèle de mise en solution des ETM développé par Ingwersen et al. (2006), nous avons optimisé les paramètres du modèle, notamment les paramètres physico-chimiques, et avons validé notre modèle sur des données cinétiques conjointes de quantités de nickel accumulé dans les feuilles de l’hyperaccumulateur et de concentration en nickel dans la solution du sol. Les perspectives de ce travail sont (i) un approfondissement des relations entre l’accumulation foliaire du nickel (ou d’un autre ETM) par un hyperaccumulateur, la transpiration et la production de protéines soufrées permettant une complexation de l’ETM et (ii) une adaptation du modèle pour le terrain, ce qui nécessite notamment une meilleure utilisation du couplage production de biomasse foliaire/transpiration et une prise en compte des cinétiques d’humectation et de dessiccation du sol (équation de Richards de transport d’eau en conditions non saturées), ce qui conduira à la mise au point d’un modèle 1D (la profondeur du sol) d’accumulation foliaire et de mise en solution d’ETM / To make phytoextraction practically feasible, predictive models of metal uptake by hyperaccumulators need to be developed. The aim of this work was to design, calibrate and validate a biophysical combined model of nickel foliar accumulation and availability in soil solution during cultures of the hyperaccumulator Leptoplax emarginata on a fertilized and Ni-contaminated sandy topsoil. We succeed in this. Part of the model integrates a transpiration bioconcentration factor (TSCF) which characterized the main Ni transport through the root and to the leaves. We determined a TSCF value greater than 1 for L. emarginata, which was attributed to (i) a high root permeability to both Ni and water and (ii) a predominant Ni active transport. By contrast, Spring wheat was characterized by a TCSF value less than 0.02 and a reflection coefficient very near 1, indicating that its roots are permeable to water but quite unpermeable to nickel. The high capacity of L. emarginata to tolerate and accumulate Ni in their leaves should also be attributed to its large transpiration and sulfur accumulation, particularly in their youngest leaves. Perspectives of this work are (i) a detailed study on relations between Ni accumulation, transpiration and production of sulphur proteins and (ii) a field adaptation of the model taken into account water transport in unsaturated conditions, leading to design a combined 1D model of nickel foliar accumulation and availability in soil solution