N° d’ordre : 4691

L’UNIVERSITÉ BORDEAUX 1

ÉCOLE DOCTORALE SCIENCES

POUR OBTENIR LE GRADE DE

SPÉCIALITÉ : Écologie évolutive, fonctionnelle et des communautés

Phytoremédiation en zo

d'eaux contaminées

Directeur de recherche: Dr. Michel MENCH

Soutenue le: 10 décembre 2012 Devant la commission d’examen formée de: M. ALARD, Didier -M. SCHWITZGUEBEL, Jean-Paul -M. VANGRONSVELD, Jaco -M. FLETCHER, Tim -M. HUNEAU, Fréderic -Mme. RAVETON, Muriel -M. LE COUSTUMER, Philippe -M. MENCH, Michel -

THÈSE

PRÉSENTÉE À

L’UNIVERSITÉ BORDEAUX 1

ÉCOLE DOCTORALE SCIENCES ET ENVIRONNEMENTS

Par Lilian, MARCHAND

POUR OBTENIR LE GRADE DE

DOCTEUR

cologie évolutive, fonctionnelle et des communautés

diation en zones humides construites

d'eaux contaminées en cuivre

Directeur de recherche: Dr. Michel MENCH

Soutenue le: 10 décembre 2012

Devant la commission d’examen formée de:

- Professeur, Université Bordeaux 1, France Président du jury- Directeur de recherche, EPFL, Lausanne, Suisse Rapporteur- Professeur, Universiteit Hasselt, Belgique Rapporteur- Professeur, Melbourne university, Australie Examinateur- Professeur, Université de Corse Examinateur- Maître de conférences, Univ. J. Fourier, Grenoble Examinateur- Maître de conférences, Université Bordeaux 3 Examinateur- Directeur de recherche, INRA-Univ. Bordeaux 1 Directeur de thèse

L’UNIVERSITÉ BORDEAUX 1

ET ENVIRONNEMENTS

cologie évolutive, fonctionnelle et des communautés

nes humides construites

Président du jury Rapporteur Rapporteur Examinateur Examinateur Examinateur Examinateur Directeur de thèse

1

Résumé

Ces travaux contribuent à caractériser des compartiments environnementaux (i.e. eau, sol et

solution du sol, substrat, macrophytes à l’échelle individuelle et des communautés) et leur

fonctionnement pour in fine améliorer l’efficacité de zones humides construites (CW) à décontaminer

une masse d’eau contaminée en cuivre. Les connaissances sur le maintien de l’homéostasie de Cu chez

les végétaux ainsi que sa phytotoxicité aux expositions élevées sont résumées. Les principaux

mécanismes physico-chimiques et biologiques intervenant en phytoremédiation d’eaux contaminées

en Cu en CW sont également discutés. Plusieurs solutions de phytoremédiation de type

phytostabilisation aidée ont été évaluées en lysimètres in situ sur un site de traitement du bois

contaminé au Cu, afin d’établir le potentiel de certains amendements à sorber Cu dans le substrat des

CW. Les concentrations en éléments traces potentiellement toxiques (PTTE, dont Cu) et

macroéléments des lixiviats migrants vers les horizons aquifères ont été quantifiées. Un laitier

sidérurgique de type Linz-Donawitz enrichi en P (LDS, 1%) a permis le meilleur développement de

Lemna minor L., utilisé ici comme bioindicateur, exposée aux lixiviats. En parallèle, les communautés

de macrophytes ont été suivies le long du parcours de la Jalle d’Eysines, une rivière urbaine

contaminée en Cu et autres PTTE. Les concentrations en PTTE ont été déterminées dans le sol, l’eau,

l’eau interstitielle et les feuilles de 7 espèces de macrophytes. Un modèle statistique multivarié

(analyse discriminante linéaire, LDA) a ensuite été élaboré sur la base des concentrations foliaires en

PTTE pour biosurveiller l’exposition des macrophytes. Des populations de macrophytes ont aussi été

prélevées sur des zones humides de contamination croissante en Cu en Europe (France, Espagne,

Portugal et Italie), Biélorussie et Australie. La production de racines chez les macrophytes exposées

pendant 3 semaines à des concentrations croissantes en Cu (0,08 ; 2,5 ; 5 ; 15 et 25 µM Cu) montre

une variabilité intra-spécifique de la tolérance au Cu pour des populations de Juncus effusus,

Schoenoplectus lacustris , Phalaris arundinacea, Phragmites australis et Iris pseudacorus. A

l’inverse, une réponse similaire à une tolérance constitutive a été obtenue chez Typha latifolia, espèce

à forte production de rhizomes. L’importance des rhizomes est discutée. En CW, à l’échelle du

mésocosme (110 dm3), jusqu’à 99% du Cu de la masse d’eau (concentration initiale: 2.5 µM Cu) ont

été éliminés dans les trois modalités plantées de Juncus articulatus, P. arundinacea et P. australis,

ainsi que dans le contrôle non planté. Les rôles du biofilm microbien, du substrat et des macrophytes

en CW ainsi que leurs interactions sont discutés. La sélection d’écotypes de macrophytes tolérants aux

PTTE pour leur utilisation en zone humide construite ainsi que les mécanismes moléculaires impliqués

dans la variabilité intra-spécifique de cette tolérance, notamment chez P. australis, sont deux thèmes

de recherche à promouvoir.

2

Summary

This work aims at characterizing environmental compartments (i.e. water, soil and soil pore

water, substrate, macrophytes at the individual and community scale) and their functioning to in fine

improve the effectiveness of constructed wetlands (CW) for cleaning Cu-contaminated waters.

Knowledge on the homeostasis of Cu in plants and its phytotoxicity at medium and high exposures are

summarized. The main physico-chemical and biological mechanisms involved in the phytoremediation

of Cu-contaminated water in CW are discussed. Several aided-phytostabilisation options were in situ

evaluated in lysimeters at a Cu-contaminated wood preservation site to assess the potential of four

amendments to sorb Cu in a CW substrate. Concentrations of potentially toxic trace elements (PTTE,

including Cu) and macronutrients of leachates migrating from the root zone to the aquifers were

quantified. Based on the responses of Lemna minor L. used as a bioindicator, exposed to the leachates,

Linz-Donawitz slag spiked with P (LDS, 1%) best performed to sorb labile Cu in the root zone. In

parallel, macrophyte communities were monitored along the Jalle Eysines River, an urban river

slightly contaminated by Cu and other PTTE. The PTTE concentrations were determined in the soil,

water, soil pore water, and in the leaves of seven macrophyte species. A multivariate statistical model

was developed based on the foliar PTTE concentrations for biomonitoring macrophyte exposures.

Populations of macrophytes were also collected in wetlands displaying an increasing Cu

contamination in Europe (France, Spain, Portugal, and Italy), Belarus and Australia. Root production

of macrophytes exposed for 3 weeks at increasing Cu concentrations (0.08, 2.5, 5, 15 and 25 µM Cu)

shows an intra-specific variability of Cu tolerance in populations of Juncus effusus, Schoenoplectus

lacustris, Phalaris arundinacea, Phragmites australis and Iris pseudacorus. In contrast, a similar

response to constitutive tolerance occurred for Typha latifolia, species with high production of

rhizomes. The rhizome influence is discussed. In a CW at mesocosm scale (110 dm3), up to 99% of Cu

in water (initial concentration: 2.5 µM Cu) was removed after 2 weeks in the three modalities planted

with Juncus articulatus, P. arundinacea and P. australis, and in the unplanted control. The

influences of microbial biofilms, the substrate, and the macrophyte species and their interactions in

CW are discussed. The selection of PTTE-tolerant macrophytes for their used in CW and the

understanding of molecular mechanisms underlying the intra-specific variability in PTTE- tolerance,

i.e for P. australis, require further investigations.

3

Qu’est ce que vous nagez bien chef !

Avant c’était mieux que ça, mais vous savez…c’est toujours pareil.

Pithiviers et Chautard (Mais où est donc passée la 7e compagnie, 1973)

Robert Lamoureux

4

Remerciements

Cette thèse a été financée par le Fond AXA pour la Recherche

L’ensemble des laboratoires et personnes listées ci-dessous ont contribué – i.e. par leur aide lors de campagnes d’échantillonnage ou au travers de partenariat lors de l’interprétation des données et/ou le traitement de l’information pour les publications – à cette thèse. Je les en remercie.

All laboratories and people listed hereafter have contributed – i.e. in helping me during sampling campaigns or through partnerships during the data analysis and interpretation and paper drafting – in this thesis. I am very grateful to them.

Wet Ecosystem Research Group, Department of Biological Sciences, NDSU Dept. 2715, P.O. Box 6050, Fargo, ND 58108-6050, USA M. Otte, D. Jacob

EA4592 Géoressources & Environnement, ENSEGID, Université de Bordeaux 1, 1 allée F. Daguin, F-33607 Pessac, France P. Lecoustumer, F. Huneau, Y. Vystavna

Department of Civil Engineering, Monash university, Room 118, Building 60, Clayton Campus, Clayton Victoria 3168, Melbourne, Australia. T. Fletecher, A. Deletic, K. Lizama, C. Schang

Minnesota State University, Department of Biological Sciences, Mankato, Minnesota, 56001, USA B. Cook

Dipartimento di Biologia Vegetale, Laboratorio di Fisiologia Vegetale, Universit `a di Firenze, via Micheli 1, I-50121 Firenze, Italy C. Gonelli, I. Colzi

Departamento de Tecnologias e Ciências Aplicadas, Escola Superior Agrária - Instituto Politécnico de Beja, Rua Pedro Soares - Campus do IPB, Apartado 6155, 7801-295 BEJA, Portugal P. Alvarenga

Instituto de Investigaciones Agrobiológicas de Galicia, CSIC, Apdo. 122, Santiago de Compostela, 15780, Spain P. Kidd et al.

GRESE, Université de Limoges, 123 Avenue Albert Thomas, 87060 Limoges, France. F. Bordas

Department of Environment and Conservation, BSc (Chem)(EnvSc) Dip Public Safety (Forensic Inv) MRACI CCHEM Contaminated Sites Branch, Locked Bag 104, Bentley DC 6983, Australia P. Newell

ISTO, UMR 6113 CNRS - Université d'Orléans. Campus Géosciences, 1A Rue de la férollerie 45100 Orléans, France. M. Motelica

Mes Remerciements vont aussi à S. Branchu, J.M. Carnus et à toute l’équipe impliquée dans le programme TRANZFOR, programme m’ayant donné l’opportunité d’effectuer trois mois de ma thèse en Australie.

Mes remerciements vont enfin aux membres de mon jury : le Dr. J.P. Schwitzguebel, le Pr. J. Vangronsveld, le Dr. M. Raveton, le Pr. D. Alard, le Pr. T.Fletcher, le Pr. F. Huneau et le Pr. P. Lecoustumer.

5

Remerciements (suite)

Un remerciement particulier à Michel Mench pour son partage des connaissances, la liberté d’entreprendre qu’il a su me laisser…et les milliers de kilomètres que l’on aura parcouru ensemble !!

Et un énorme merci à vous tous !!

BaptisteBaptisteBaptisteBaptiste

Marie CarolineMarie CarolineMarie CarolineMarie Caroline

Un merci spécial pour ceux de toujours : Aline/Jean-Yves/Tibodrey

Et des bisous à la meilleure : Reyitas

À mon père…

6

Liste des publications

Articles à Comité de lecture international

Marchand, L ., Mench, M., Jacob, D.L., Otte, M.L. 2010. Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants and standardized measurements: a review. Environmental Pollution, 158, 3447-3461.

Marchand, L ., Mench, M., Marchand, C., Le Coustumer, P., Kolbas, A., Maalouf, J.P. 2011. Phytotoxicity testing of lysimeter leachates from aided phytostabilized Cu-contaminated soils using duckweed (Lemna minor L.). Science of the Total Environment 410,146-153.

Maalouf, J.P., Le Bagousse-Pinguet, Y., Marchand, L ., Touzard, B., Michalet, R. 2012. The interplay of stress and mowing disturbance for the intensity and importance of plant interactions in dry calcareous grasslands. Annals of Botany. doi: 10.1093/aob/mcs152

Maalouf, J.P., Le Bagousse-Pinguet, Y., Marchand, L ., Bâchelier, E., Touzard, B., Michalet, R. 2012. Integrating climate change into calcareous grassland management. Journal of Applied Ecology, 49, 795-802.

Nsangawimana, F., Marchand, L ., Mench, M. 2013. Arundo donax L., a candidate for phytomanaging water and soils contaminated by potentially toxic trace elements and producing plant-based feedstock. International Journal of Phytoremediation.

Articles soumis Marchand, L ., Mench, M., Nsanganwimana, F., Vystavna, Y., Huneau, F., Le Coustumer, P., Lamy, J.B, Cook, B.J. 2012. Macrophytes as biomonitors of trace element exposure along an urban river using a multimetric approach (Jalle d’Eysines River, France). Freswater Biology.

Communications orales Marchand L ., Mench M., Jacob D.L., Otte M.L. 2010. Are monocots more efficient than dicots for removing trace elements in constructed wetlands? 7th International conference on phytotechnologies "Phytotechnologies in the 21st century: challenges after Copenhagen 2009. Remediation-Energy-Health-Sustainability". University of Parma. Parma, Italy. September 26-29th. Marchand L ., Mench M., Marchand C., Le Coustumer P., Kolbas A., Maalouf J.P. 2012. Phytotoxicity testing of lysimeter leachates from aided phytostabilized Cu-contaminated soils using duckweed (Lemna minor L.). IWA Regional Conference on Wastewater Purification & Reuse. Heraklion, Crete, Greece. March 28-30th. Marchand L ., Mench M., Nsangwimana F., Oustrière N., Fletcher T. 2012. Copper tolerance in macrophyte populations: Innate tolerance and/or phenotypic plasticity? Panamerican Conference on Wetland Systems for water quality improvement, management and treatment. Technological University of Pereira, Pereira (Colombia). February 26th - March 1st. Mench M., Bert V., Vangronsveld J., Kolbas A., Marchand L., Puschenreiter M., Kidd P., Kumpiene J., Cundy A. 2012. Phytotechnologies appliquées aux sites pollués: Etat de l’art à l’international. Ademe, Bioindicateurs et phytotechnologies, des outils biologiques pour des sols durables, Journées techniques nationales 16-17 octobre, Paris. 11 p. Posters Marchand L ., Mench M., Vystavna Y., Kolbas A., Huneau F. 2010. Potential use of macrophytes as bio-indicators of trace element-contamination along the Jalle river (France). 12th International conference on wetland systems for water pollution control. Venice, Italy, October 4-8th. Kolbas A., Mench M., Herzig R, Nehnevajova E., Marchand L . 2010. Copper phytoextraction using mutants lines of sunflower and tobacco. Sakharov readings 2010, Minsk, Belarus. p. 208. May 21-22th.

7

Marchand L ., Mench M., Kolbas A. 2010. Are monocots more efficient than dicots for removing trace elements in constructed wetlands? International conference on environmental pollution and clean/bio phytoremediation. Pisa, Italy, June 16-19th. pp 91. http://EPCR2010.azuleon.org. Marchand L ., Mench M., Vystavna Y., Kolbas A., Huneau F. 2010. Potential use of macrophytes as bio-indicators of trace element-contamination along the Jalle river (France). p. 79 in: Int. Conf. Environmental Pollution and Clean Bio/Phytoremediation. R. Izzo et al. (Eds.), June 16-19, Pisa, Italy, pp 79. http://EPCR2010.azuleon.org. Marchand L ., Mench M., Marchand C., Le Coustumer P., Kolbas A., Maalouf J.-P. 2011. Phytotoxicity testing of lysimeter leachates from aided phytostabilised Cu-contaminated soils using duckweed (Lemna minor L.). Int. Phytotechnology Conference, Portland, USA. September 13-16th. Kolbas A., Mench M., Marchand L ., Herzig R., Nehnevajova E. 2011. Copper phytoextraction using mutant lines of sunflower. Int. Phytotechnology Conference, Portland, USA. September 13-16th. Kolbas A., Mench M., Marchand L ., Herzig R., Nehnevajova E. 2011. Copper phytoextraction using mutant lines of sunflower. 2011. 11th International Conference on the Biogeochemistry of Trace Elements (ICOBTE). Firenze , Italy, July 3-7th. Part. II, p. 97-98. Marchand L ., Mench M., Jacob D.L., Otte M.L. 2011. Metal and metalloid removal in constructed wetlands: Emphasis on the Relative Treatment Efficiency Index to quantify the importance of macrophytes. 2011. 11th International Conference on the Biogeochemistry of Trace Elements (ICOBTE). Firenze, Italy, July 3-7th. Part. II, p.417-418. Kolbas A., Mench M., Marchand L ., Herzig R., Nehnevajova, E. 2011. Biomonitoring of copper contaminated soils using a mutant line of sunflower. Doctoral student conference: Next generation insights into geosciences and ecology. Tartu (Estonia). May 12-13th. Mench M., Kolbas A., Atziria A., Herzig R., Marchand L ., Maalouf J.-P., Ricci A. 2012. Field evaluation of one Cu-resistant tobacco variant and its parental lines for copper phytoextraction at a wood preservation site. Proc. 4th Int. Congress Eurosoil 2012, Soil Science for the Benefice of Mankind and Environment, 12. Soil Pollution and Remediation, S12.08-Potentially harmful elements in soils, Bari, Italy. July 2-6. p. 2542. Marchand L ., Mench M., Kolbas A. Vystavna Y., Huneau F., Motelika-Heino M. 2012. Bioindicating properties of macrophytes for monitoring trace element exposure along the Jalle River (France). Proc. 4th Int. Congress Eurosoil 2012, Soil Science for the Benefice of Mankind and Environment, 12. Soil Pollution and Remediation, S12.03-Biogeochemistry of contaminants in wetland, Bari, Italy. July 2-6. p. 2395. Marchand L ., Mench M., Nsangawimana F . 2012. Copper tolerance in macrophyte populations: Innate tolerance and/or Phenotypic plasticity? 9th International conference on phytotechnologies "Plant based strategies to clean water, soil, air and provide ecosystem services". University of Hasselt. Diepenbeek, Belgium. September 11-14th. Rapports Mench M, A. Kolbas, L. Marchand, E. Hego, N. Oustrière, A. Atziria, F. Nsanganwimana, C. Bes, J. Guinberteau, M. Monmarson, C. Bouquet, F. Le Pierres, V. Lozano, F. Schneider. 2011. 12th month progress report, Gentle remediation of trace element contaminated land – GREENLAND, Project FP7-KBBE-266124, UMR BIOGECO INRA 1202, Talence, France. 51 p. Mench M, Kolbas A, Marchand L , Bes C, Atziria A, Oustrière N, Hego E, Sureau M-C, Guinberteau J, Monmarson M, 2011. Solutions de phytoremédiation et pilotes de démonstration sur la plate-forme de phytoremediation BIOGECO (PHYTODEMO). 1er rapport intermédiaire, convention ADEME 09 72 C0076 / INRA 22000460, Ademe, Département Friches industrielles et sols pollués. Angers, France. 131 p. Mench, Marchand L , Nsanganwimana F, Kolbas A, Oustrière N 2011 Phytoremédiation en zone humide construite d’eaux contaminées d’un site de traitement du bois (plate-forme de phytoremédiation BIOGECO) (CWDEMO). 1er rapport intermédiaire, convention: ADEME 1072C0081 / INRA 22000495, Ademe, Département Friches Urbaines et Sites Pollués (SFUSP). Angers, France. 31 p.

8

Mench M., Kolbas A., Marchand L ., Hego E., Oustrière N., Rousseau F., Roumier P-H, Nibaudeau M., Boechat C., Sinchuk A., Nsanganwimana F., Guinberteau J., Monmarson M. 2012 18th month progress report, Gentle remediation of trace element contaminated land – GREENLAND, Project FP7-KBBE-266124, UMR BIOGECO INRA 1202, Talence, France. 33 p. Bouchardon J-L, Faure O, Lespagnol G, Moutte J, Guy B, Mimoun D, Graillot D, Paran F, Croze V, Athénol J-D, Ferrando B, Boisson J, Perret S, Hitmi A, Verney P, Moussard Gauthier C, Ledoigt G, Goupil P, Sac C, Mench M, Oustrière N, Marchand L , Kolbas A, 2012. Physafimm: la phytostabilisation - méthodologie applicable aux friches industrielles métallurgiques et minières, 1er rapport, convention: ADEME 0872C0119, Ademe, Département Sites et Sols Pollués, École des Mines, St Etienne, France. 142 p.

Reviewer pour:

"Bioresource Technology" "Ecological Engineering" "ESPR : Environmental Science and Pollution Research" "KMAE : Knowledge and Management of Aquatic Ecosystems" "Fresenius Environmental Bulletin" "La terre et la vie"

9

Préambule

Les travaux de thèse se sont focalisés dès l’origine sur le développement de zone humides construites

(CW), en mésocosmes, avec pour finalité la décontamination d’une matrice eau contaminée en cuivre.

Ce projet, financé par la Fondation AXA pour la recherche, s’inscrit à l’interface de l’écologie des

communautés, de l’écologie de la restauration et de l’ingénierie écologique, dans le cadre des

phytotechnologies appliquées à la remédiation de matrices polluées (phytoremédiation).

Les mécanismes régissant la décontamination des eaux en CW relèvent de la physico-chimie et de la

biologie (activités des plantes et microorganismes). Ils reflètent des interactions biotiques et abiotiques

dans l’eau, le sol (pour plus de facilité dans le texte, ce terme regroupe les autres substrats tels que

sédiments, graviers, etc.), la rhizosphère (ici, volume de sol ou substrat directement influencé par les

racines et les micro-organismes associés, i.e. bactéries, actinomycètes et champignons (Hinsinger et

al., 2009) et la biomasse souterraine (racines et rhizomes). La réponse de ces compartiments à une

exposition aux Éléments Traces Potentiellement Toxiques (en anglais PTTE) est conditionnée par leur

nature, leurs interactions avec les autres compartiments, et leur variabilité potentielle (i.e. la possibilité

d’un compartiment à modifier de lui-même son état premier en réponse à une variation de son

environnement). La plasticité phénotypique des végétaux, comme leur plasticité génotypique, est un

exemple de cette variabilité potentielle. L’évaluation de cette plasticité phénotypique chez les

macrophytes – les espèces végétales visibles à l’œil nu et vivant en zone humide - en réponse à une

exposition aux PTTE (ici Cu) est considérée comme prioritaire par nos travaux.

Les mécanismes physico-chimiques et biologiques régissant le comportement des compartiments

environnementaux listés ci-dessus sont complexes. Cependant de grands sujets de recherches se sont

dessinés en termes de compréhension des processus biologiques et physico-chimiques impliqués dans

la phytoremédiation au cours des vingt dernières années. Pilon-Smith. (2005) résume les bases de ce

thème de recherche dans un article intitulé Phytoremediation, publié dans l’annual review of plant

biology. Au cours de cette thèse, j’ai contribué à ces sujets de recherches pour mieux caractériser les

pollutions aux PTTE dans les compartiments eau et sols ainsi que leurs impacts sur le vivant (ici les

macrophytes). Fort des résultats et conclusions de la première moitié de la thèse, et en utilisant les

données de la littérature, nous avons développé notre propre CW dont nous présentons les

performances. Ce CW a été réalisé dans le but de décontaminer une eau à forte concentration en Cu,

PTTE souvent en excès dans les sols des vignobles en Aquitaine, France et d’autres pays, mais aussi

dans des effluents d’origine variée (ex: traitements du bois, rejets domestiques, lisiers, etc.).

10

Les chapitres s’articulent autour de six grandes parties. La première Le cuivre chez les végétaux

supérieurs: de l’homéostasie à la phytotoxicité constitue la première moitié de l’introduction générale.

Elle résume les connaissances sur la place de l’élément cuivre dans le cycle végétatif de la plante,

depuis sa nécessité en tant que cofacteur enzymatique jusqu’à sa phytotoxicité, notamment via sa

contribution à la production de radicaux libres lors d’un stress oxydant. Les transporteurs

membranaires impliqués dans le transport de Cu sont détaillés, avec un aperçu des techniques

d’imagerie permettant de localiser les PTTE dans la plante, notamment les organites. L’accent est

aussi mis sur les interactions entre Cu et les macrophytes. La seconde moitié de l’introduction, publiée

sous forme d’un état-de-de l’art en 2010 sous le titre Metal and metalloid removal in constructed

wetlands, with emphasis on the importance of plants and standardized measurements: a review

résume les connaissances sur la phytoremédiation d’eaux contaminées en PTTE en CW. Les

principaux mécanismes physico-chimiques et biologiques intervenant en CW y sont présentés et

discutés.

Avant de décontaminer une matrice eau contaminée au Cu, des travaux ont été menés pour

comprendre comment celle-ci se chargeait en Cu sur site après percolation à travers une matrice sol

contaminée en Cu. La source de Cu était liée à une activité industrielle: le traitement de bois. Plusieurs

solutions de phytoremédiation de type phytostabilisation aidée ont été testées in situ et la concentration

en Cu des lixiviats migrants de la zone explorée par les racines vers les horizons aquifères a été

quantifiée. Ceci a renseigné sur la composition des eaux à traiter. Un biotest avec Lemna minor L.

comme bioindicateur a permis d’évaluer l’impact des solutions de phytostabilisation, aidée ou non, sur

la phytotoxicité des lixiviats. Ces premiers travaux ont été publiés sous le titre Phytotoxicity testing of

lysimeter leachates from aided phytostabilized Cu-contaminated soils using duckweed (Lemna

minor L.) en 2011 et forment le premier chapitre post-introduction de ce manuscrit. Après cette

première phase pour comprendre les voies d’entrée du Cu dans l’eau et leurs impacts, des travaux de

biosurveillance (biomonitoring) de communautés de macrophytes ont été menés sur des sites

potentiellement contaminés en Cu (et autres PTTE) d’après l’étude des activités anthropique et des

dangers. Les concentrations en PTTE dans le sol, l’eau, l’eau interstitielle et les feuilles de

macrophytes ont été déterminées le long du parcours de la Jalle d’Eysines, une rivière urbaine au nord

de l’agglomération de Bordeaux. Selon nos travaux, elle présente une contamination croissante en

PTTE de sa source à son embouchure. L’hypothèse à valider était le lien entre l’exposition des

macrophytes à une contamination polymétallique et leurs concentrations foliaires en PTTE et

macroéléments. Ces travaux sont soumis sous le titre Macrophytes as biomonitors of trace element

exposure along an urban river using a multimetric approach (Jalle d’Eysines River, France) et

11

constituent le second chapitre de la thèse. Ils proposent un modèle statistique multivarié pour

biosurveiller l’exposition des macrophytes aux PTTE au cours du temps. Parallèlement, la plasticité

phénotypique des macrophytes a été une hypothèse de travail. Ces plantes de zone humide sont

connues comme plus tolérantes aux PTTE que les plantes terrestres. Cette tolérance est définie comme

constitutive dans la littérature, mais elle n’est pas validée par des jeux de données suffisant.

Cependant, pour hiérarchiser les facteurs permettant d’optimiser la conception de CW, il était

nécessaire de savoir si cette hypothèse de tolérance, supposée constitutive, était validée ou bien si une

variabilité intra-spécifique avait échappée aux investigations de la littérature: tous les individus d’une

même espèce de macrophyte répondent-ils de la même manière à une forte exposition au cuivre, et ce

indépendamment de leur site de croissance originel? Cette question a été abordée dans la littérature,

mais à ce jour et à notre connaissance la plasticité phénotypique attendue n’a jamais été observée. Les

travaux relatifs à cette question sont le sujet du troisième chapitre post-introduction : Does phenotypic

plasticity explain Cu tolerance in macrophyte populations? Contrairement à l’hypothèse dominante,

ils montrent une variabilité intra-spécifique de la tolérance au Cu pour des populations de Juncus

effusus, Schoenoplectus lacustris et Phalaris arundinacea.

Sur la base des connaissances acquises lors de la première moitié du doctorat, notamment via l’état de

l’art Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants

and standardized measurements: a review et en s’appuyant sur les résultats obtenus sur la tolérance au

Cu et la plasticité phénotypique des macrophytes, un CW a été conçu et évalué durant la seconde

partie de la thèse. Il a permis de travailler la question première de cette thèse: peut-on décontaminer

une matrice eau contaminée au Cu dans un CW? Et si oui, quel est le rôle des acteurs que sont le

substrat, les macrophytes, les microorganismes et leurs biofilms. Les résultats afférents sont présentés

dans le quatrième chapitre post-introduction Copper removal from water using Bio-Racks planted with

Phragmites australis, Juncus articulatus and P. arundinacea. Une discussion générale synthétisant

l’ensemble des résultats conclut ce manuscrit.

12

Sommaire

Préambule 9

Introduction : partie I Le cuivre: un oligo-élément essentiel; homéostasie, carence et phytotoxicité selon l’intensité de l’exposition de la plante 16

I. Le cuivre: un oligo-élément essentiel pour la plante 17

1.1 Facing the challenges of Cu (Fe and Zn) homeostasis in plants 17 1.2 Acidification de la rhizosphère pour faciliter le prélèvement du cuivre 19 1.3 Prélèvement du Cu basé sur la réduction (Stratégie I) 20 1.4 Prélèvement de Fe et Cu chez les graminées (Stratégie II) 22 1.5 Transporteurs impliqués dans le prélèvement et le transfert du cuivre vers les parties aériennes 22 1.6 Transport intracellulaire du cuivre 23

II. Phytotoxicité et détoxification des effets délétères de l’excès de Cu 27

2.1 Les ROS et le système anti-oxydant 27 2.2 Impacts sur le développement racinaire 30

III. Le cuivre et les macrophytes 36

Introduction : partie II La phytoremédiation, les éléments traces, l’eau, les macrophytes et les zones humides construites 38

I. Rationale for using constructed wetlands for improvement of water quality 40

II. Metal removal processes in wetlands 41

2.1 Adsorption 41 2.2 Co-precipitation and redox reactions 43 2.3 Metal carbonates 45 2.4 pH 45 2.5 Erosion, sedimentation 46 2.6 The role of micro-organisms 46

2.6.1 Micro-organism mediated oxidation 48 2.6.2 Micro-organism mediated reduction 48

III. The roles of plant species 50

3.1 Emergent plants 50 3.2 Floating plants 54 3.3 Submerged plants 54 3.4 Adaptability of macrophytes to metal stress 55

IV. Importance of the plant ecotype 56

V. Selection of a data set for assessment of metal removal efficiency 57

VI. Removal of metals 60

6.1 Differences between systems planted in monoculture 60 6.2 Planted versus unplanted systems 63 6.3 Monoculture versus mixed stands 63 6.4 System size 63 6.5 Seasonal effects 65

VII. A new index: the relative treatment efficiency index (RTEI) 66

13

Conclusion 67

References 69 Chapitre I De la nécessité de traiter les sols contaminés au Cu afin d’en limiter le transfert vers les horizons aquifères…Le cas de la phytostabilisation aidée : vers une nouvelle perspective de zones humides construites 77

I. Introduction 81

II. Materials and Methods 83

2.1 Soils and amendments 83 2.2 Plant material and toxicity test 86 2.3 Statistical analyses 88

III. Results and Discussion 88

3.1 Composition of lysimeter leachates 88

3.1.1 UNT 90 3.1.2 LDS 90 3.1.3 DL 91

3.1.4 OM, OMDL, OMZ 91 3.1.5 PHYTO 92

3.2 Phytotoxicity and L. minor growth 93

Conclusion 96

References 97

Chapitre I : Take home message 101

Chapitre II Bio-surveillance de l’évolution de l’exposition des macrophytes aux PTTE…Le cas d’une rivière urbaine la Jalle d’Eysines 103

I. Introduction 108

II. Materials and methods 110

2.1 Description of the studied area 110 2.2 Water, soils and macrophytes sampling 111 2.3 Water and soil analysis 112 2.4 Mineral analysis of leaf samples 113 2.5 Statistical analysis 113

III. Results 114

3.1 Freshwater, soil and soil pore water 114 3.2 Foliar element concentrations of macrophytes 116 3.3 Modelling of macrophyte exposure by discriminant analysis of foliar PTTE concentrations 118

IV. Discussion 121

4.1 Copper and Zn in soils, soil pore water and macrophytes 121 4.2 Chromium, Fe, Mo, Cd and Pb in soils, soil pore water and macrophytes 121 4.3 Transfers from roots and rhizomes to leaves in hemicryptophyte and rhizomatous-geophytes 122 4.4 Biomonitoring of plant exposure to PTTE by using foliar PTTE concentrations

in macrophytes in a linear discriminant analysis (LDA) 123

14

References 126

Chapitre II : Take home message 129 Chapitre III Variabilité intra-spécifique de la tolérance au Cu parmi les populations de 6 espèces de macrophytes 131

I. Introduction 136

II. Materials and methods 138

2.1 Sites 138 2.2 Plant sampling, vegetative reproduction and Cu testing 149 2.3 Statistical analysis 141

III. Results 145

3.1 Copper concentrations and physico-chemical characteristics of the growing medium 145 3.2 Crossed effect of groth location and Cu exposure on root biomass production 146

IV. Discussion 153

4.1 Copper speciation in the growing medium 153 4.2 Inter/Intra specific variability vs constitutive tolerance to cope with excess Cu 153 4.2.1 Inter-specific variability of root biomass production across a Cu-gradient exposure 153 4.2.2 Intra-specific variability of root biomass production across a Cu-gradient exposure 154

Conclusion 157

References 158

V. Chapitre III Informations complémentaires 162

5.1 Matériel et Méthode 162 5.2 Résultats et discussion 163

5.2.1 Activité photosynthétique 163 5.2.2 Activité de la Gaïacol peroxydase 165

Chapitre III Take home message 168

Chapitre IV Traitement d’eaux contaminées par Cu en zones humides construites de type Bio-Racks plantées de Phragmites australis, Phalaris arundinacea et Juncus articulatus 170

I. Introduction 173

II. Material and Method 174

2.1 Pilot plant setup 174

2.1.1 Plants 175 2.1.2 Sampling and analysis 176 2.1.3 Statistical analysis 179

III. Results and discussion 179

3.1 Physico chemical parameters and Cu removal in bio-racks during a 14-day exposure period in alkaline conditions 180

3.2 Physico chemical parameters and Cu removal in bio-racks during a 14-day exposure period in acidic conditions 186

15

Conclusion 187

References 189 Supplementary material 192

Chapitre IV Take home message 192

Synthèse générale 194

I – L’analyse statistique multivariée : une approche nécessaire en phytoremédiation pour l’interprétation des points finaux de mesure (dont le ionome). 196

II – Vers une standardisation nécessaire à la sélection des macrophytes et des substrats en

CW: Nécessité d’un Biotest. 198

III – Quel est le rôle des microorganismes en CW? 200

IV – Influence de la population chez les macrophytes utilisés en CW. 200

V – Apport d’amendements au substrat d’un CW pour décontaminer une masse d’eau à forte concentration en Cu. 202

Références 205

Annexes 212 Liste des macrophytes en zones humides construites 213 Publications 256 Posters 261 Communications orales 266

Liste des abbreviations

ABCC ATP binding-cassette C Transporter, ACC 1-aminocyclopropane-1-carboxylate, ACCD ACC deaminase, AHA Arabidopsis H+ ATPase, AMD Acid Mine Drainage, ANOVA Analysis of Variance, APX Ascorbate Peroxidase, AsA Ascorbate, ATX Antioxidant, BR Batch Reactor, CAT Catalase, CCA Chromated Copper Arsenate, CCH Cu Chaperonne, CCS Cu Chaperonne for Cu/Zn SOD, CDF Cation Diffusion Facilitator, CDS3 Peroxisomal Cu/Zn Superoxide Dismutase, COPT Copper Transporter, COX Cytochrome Oxydase, CW Constructed Wetland, CYS Cystein, DBO5 Biological Oxygen Demand after 5 days, DL dolomitic limestone, DMA deoxy-mugineic acid, DOM Dissolved Organic Matter, ECx effective concentration, EC Electrical Conductivity, ECS glutamylcystein synthase, Eh Redox Potential, epi-DMA 3-epihydroxy-mugineic Acid, EPR Electron Paramagnetic Resonance, EXAFS extended X-ray absorption fine structure spectroscopy, FDR3 Ferric Reductase Defective 3, FDRL1 Ferric Reductase Defective Like 1, FRO2 Ferric Chelate Reductase, GLU Glutamate, GLY Glycin, GPOD Guaiacol peroxidase, GPX Glutathione Peroxidase, GR Glutathione Reductase, GSH Monomeric Glutathione (reduced form), GSSG Glutathione Disulfide (oxidized form), GST Glutathione-S-Transferase, HAVA hydroxyavenic Acid, HM Heavy Metal, HMA Heavy Metal ATPases, HMWC High Molecular Weigh Compound, HSSF Horizontal Sub Surface Flow, ICP-AES Inductively coupled plasma atomic emission spectroscopy, ICP-MS Inductively coupled plasma mass spectrometry, ITR1 Iron Transporter, LDA Linear Discriminant Analysis, LDS Linz-Donawitz slag, LMWOA Low Molecular Weight Organic Acid, MA Mugineic Acid, MDA malondialdehyde, MDHA Monodehydroascorbate, MDHAR MDHA reductase, µ-CT micro computed tomography, µ-XRF micro X-ray fluorescence mapping, MT Metallothionein, NA nicotianamine, NADP(H) Nicotinamide Adenine Dinucleotide Phosphate (Reduced), NOEC No observed effect concentration, OM Organic Matter, OMDL OM + dolomitic limestone, OMZ OM + zerovalent iron grit, (Os)YSL (Oryza sativa) Yellow Stripe Like, P5CS delta-1-pyrroline-5-carboxylate synthetase, PAA P-type ATPase Transporter, PC Phytochelatin, Pc Plastocyanin, PCS Phytochelatin Synthase, PHYTO untreated modality planted, POD Peroxidase, POP Persistent Organic Pollutant, PRO L-Proline; PRX Peroxyredoxin, PS phytosiderophores, PSI Photosystem I, PTTE Potentially Toxic Trace Element, QDA Quadratic Discriminant Analysis, RAN Responsive to Antagonist, RB Retarding Basin, ROS Reactive Oxygen Species, RTEI Relative Treatment Efficiency Index, SF Surface Flow, SOD Superoxide dismutase, SPL7 Squamosa Promoter Binding Protein Like7, S-XRF synchrotron-based x-ray fluorescence microscopy, TF Translocation Factor, TOM1 Transporter Outer Membrane 1, UNT untreated modality without plants, VSSF Vertical Sub Surface Flow, WHC Water Holding Capacity, WTP Wastewater Treatment Plant, XANES micro X-ray absorption near edge spectroscopy, XAS X-ray absorption spectroscopy, YS Yellow Stripe, ZIP Zinc/Iron-Regulated Transporter Like Protein

16

Introduction : partie 1

Le cuivre: un oligo-élément essentiel; homéostasie, carence et phytotoxicité

selon l’intensité de l’exposition de la plante

17

I. Le cuivre: un oligo-élément essentiel pour la plante

Si l’exposition excessive au Cu peut conduire à une phytotoxicité pour les macrophytes, Cu

doit d’abord être considéré comme un oligo-élément nécessaire aux fonctions métaboliques et

au développement de la plante (Palmer and Guerinot., 2009, Burkhead et al., 2009).

1.1 Facing the challenges of Cu (Fe and Zn) homeostasis in plants

Sous ce titre Palmer et Guerinot (2009) discutent les stratégies mises en place dans le règne

végétal pour maintenir l’homéostasie du Cu (et d’autres métaux) dans la plante. Le Cu est un

métal de transition - un élément chimique du bloc d du tableau périodique qui n'est ni un

lanthanide ni un actinide (www.IUAPC.org) – qui est essentiel pour les plantes supérieures

comme Fe, Zn, Mn, Mo et Ni (Tableau 1). Il existe sous de multiples formes redox grâce à sa

faculté d’échanger des électrons depuis son orbite d (Palmer and Guerinot, 2009). Ce métal

est un cofacteur dans la chaîne de transport des électrons dans la mitochondrie et le

chloroplaste (Marschner, 2011). La plus fréquente protéine associée au Cu dans la cellule est

la plastocyanine, chargée du transfert des électrons depuis le complexe b6f du cytochrome

vers le photosystème I (PSI) (Tableau 2). Cu est aussi cofacteur des protéines impliquées

dans la protection contre les dérivés réactifs de l’oxygène (ROS: reactive oxygen species).

Les ROS sont des espèces chimiques de l’oxygène telles que l'oxygène singulet 1O2, les

hydroxyles (HO�), les ions superoxydes (O2-) et les peroxydes (H2O2), rendus chimiquement

réactifs par la présence d'électrons de valence – les électrons de la couche superficielle de

l’atome - non appariés. Les ROS sont d'origine exogène, e.g. produits par des rayonnements

ionisants - ou bien endogène, sous-produits du métabolisme de l'oxygène, et jouent un rôle

dans la communication entre les cellules. Leur concentration peut croître en période de stress -

à cause d’une forte exposition aux éléments traces potentiellement toxiques (en anglais

PTTE). Lorsque les mécanismes pro-oxydants dépassent les mécanismes antioxydants, il y a

un stress oxydant. Une des oxydases clefs est la cytochrome oxydase (COX) (Yruela, 2009).

Cu est impliqué dans les processus de lignification des parois cellulaires. Il est le cofacteur de

l’amine oxydase, une enzyme associée aux parois cellulaires qui catalyse l’oxydation de la

putrescine et participe de fait à la production d’H2O2 impliqué dans la lignification (Ramocha

et al., 2002). Il participe également à la formation du pollen, au métabolisme des hydrates de

18

carbone et la synthèse de composés phénoliques en réponse aux attaques de pathogènes

(Marschner, 2011; Sancenon et al., 2004 ; Palmer and Guerinot, 2009).

Tableau 1 : Oligo-éléments essentiels aux plantes (Palmer and Guerinot, 2009)

Les deux matrices sources d’exposition racinaires des plantes au Cu sont l’eau, mais surtout le

sol ; une exposition foliaire est possible, en lien avec une contamination diffuse ou ponctuelle,

surtout chez les macrophytes immergés. Bien qu’abondamment présents dans une matrice, les

métaux ne sont pas toujours faciles à prélever par les racines s’ils sont sous forme insoluble.

Le cuivre, à l’image du zinc, est un métal principalement présent sous forme insoluble car il

est adsorbé sur les argiles, les carbonates et les matières organiques. Cette insolubilité est

marquée dans les sols alcalins, qui représentent 30% de la surface terrestre émergée (Palmer

and Guerinot, 2009). Les plantes ont adopté diverses stratégies afin de pallier à cette

disponibilité limitée. L’ensemble des plantes, si l’on excepte les Graminées, ont recours à des

stratégies de prélèvement basées sur la réduction du Cu présent dans le sol (Figure 1). Les

Graminées basent leur stratégie de prélèvement, celle du Fe notamment, sur la chélation via la

production de phytosidérophores (Figure 2). Une troisième stratégie, commune à l’ensemble

des plantes, est l’acidification de la rhizosphère via l’excrétion de protons H+.

19

Tableau 2 : Fonctions et localisation des protéines associées au Cu chez les plantes

(Burkhead et al., 2009).

1.2 Acidification de la rhizosphère pour faciliter le prélèvement du cuivre

Sur sols alcalins, les plantes peuvent exsuder des protons dans la rhizosphère via l’activité de

l’ATPase. Le pH diminue alors légèrement au niveau de la rhizosphère et passe sous le seuil

limitant la solubilité et le prélèvement des oligoéléments sous forme cationique (Palmgren et

al., 2008). Ce mécanisme est fréquent lors de carences ferriques, le fer non soluble sous la

forme d’oxydes Fe(OH)3 pouvant être plus facilement solubilisé et prélevé sous sa forme Fe3+

avec formation concomitante de trois molécules d’eau. Les ATPases responsables de

l’acidification ne sont à ce jour pas formellement identifiées mais sont présupposées

appartenir aux AHA (Arabidopsis H+ ATPases), notamment aux AHA1, 2 et 7 (Palmgren,

2001; Dinneny et al., 2008). L’acidification du sol peut induire une meilleure solubilisation

d’autres cations métalliques, dont Cu, en facilitant leur désorption des phases porteuses et des

particules du sol. L’acidification opérée dans la rhizosphère par les ATPases établit un

potentiel membranaire racinaire négatif [-100 to -250 mV] qui favorise le prélèvement des

cations, dont Cu (Palmgren et al., 2008).

20

Figure 1 : Transport de Cu, Fe et Zn à travers les cellules chez les dicotyledones (Palmer and Guerinot., 2009). Cu, Fe et Zn sont prélevés dans l’apoplasme par leurs transporteurs spécifiques localisés dans l’épiderme. La réduction de Fe et éventuellement de Cu par FRO2 et l’acidification du sol par les Arabidopsis H+ ATPases (AHA) contribuent à augmenter le prélèvement. Le Cu peut emprunter la voie symplastique jusqu’aux vaisseaux conducteurs en passant la bande de Caspari au niveau de l’endoderme. Le transport dans le xylème n’est à ce jour pas totalement caractérisé mais impliquerait des transporteurs de la famille des Heavy Metal ATPases (HMA) et la pompe à citrate FRD3. Dans le xylème, Cu est transporté vers les parties aériennes via le flux généré par l’évapotranspiration. Le Cu est déchargé dans les parties aériennes par une protéine de la famille des YSL. Les transporteurs YSL sont potentiellement à la source de la translocation de Cu vers le phloème. Les carrés sombres représentent la bande de Caspari

1.3 Prélèvement du Cu basé sur la réduction (Stratégie I)

Désorbés des phases porteuses ou plus en compétition avec les protons pour les sites de

sorption, les ions libre Cu2+ dans la solution du sol sont plus accessibles à la fonction de

prélèvement des racines. Les transporteurs membranaires du rhizoderme, et plus généralement

des cellules, ont une affinité spécifique pour un état donné d’oxydation du métal. Le Cu est

prélevé par la racine sous sa forme Cu+ par le transporteur COPT1 (Sancenon et al., 2004 ;

Puig et al., 2007). L’expression de la protéine COPT1 est renforcée en conditions de carence

21

en Cu via l’action du facteur de transcription SPL7 (Squamosa Promoter Binding Protein

Like7) (Sancenon et al., 2004; Yamasaki et al., 2009). Cependant, la forme Cu2+ est plus

présente que la forme Cu+ dans le sol (Puig et al., 2007). Il est possible que Cu2+ soit réduit

par FRO2, la réductase du fer. L’enzyme FRO2 (Ferric chelate reductase) réduit Fe3+ en Fe2+

afin qu’il soit prélevé par la racine (Robinson et al., 1999). La forme Cu2+ peut aussi être

prélevée via un transporteur de la famille des ZIP (Zinc/Iron-Regulated Transporter Like

Protein), celle-ci étant connue pour le transport des cations divalents (Eide et al., 1996). Les

transporteurs ZIP2 et ZIP4 s’expriment en conditions de carence en Cu (Wintz et al., 2003),

notamment via l’action de SPL7 (Yamasaki et al., 2009), mais leur implication dans le

prélèvement de Cu n’est pas clairement établie.

Figure 2 : Transport de Cu, Fe et Zn à travers les cellules chez les monocotyledones (Palmer and Guerinot., 2009). Le Fe et Zn sont chélatés par les phytosidérophores puis prélevés via les transporteurs Yellow Stripe-Like (YSL) localisés dans l’épiderme. Le Fe peut aussi être prélevé par OsIRT1. Les métaux traversent l’espace apoplastique jusqu’à la bande de Caspari au niveau de l’endoderme avant d’être chargés dans le xylème. La pompe à citrate FRDL1 intervient dans le relargage du citrate dans le xylème et le transport du fer dans le xylème sous forme citrate-Fe3+. Le Fe est alors transféré vers les parties aériennes via le flux généré par l’évapotranspiration. Les carrés sombres représentent la bande de Caspari.

22

1.4 Prélèvement de Fe et Cu chez les graminées (Stratégie II)

Les Graminées prélèvent Fe par l’exsudation dans la rhizosphère d’acides aminés non

protéinogéniques appartenant à la famille des acides muginéiques (MA), les

phytosidérophores (PS). Les PS se déclinent en trois types: les MA, les acides déoxy-

muginéiques (DMA) et les acides 3-épihydroxy-muginéiques (epi-DMA) (Tsednee et al.,

2012). Les acides aveniques (AVA) et les acides hydroxyavéniques (HAVA) sont deux autres

types de PS répertoriés par Fushiya et al. (1980) chez le genre Avena et plus récemment chez

Poa pratensis cv Baron (Ueno et al., 2007) et Hordeum vulgare cv Himalaya (Tsednee et al.,

2012). L’expression des gènes impliqués dans la biosynthèse des MA est induite par la

carence en Fe et est régulée par l’homéostasie cellulaire de Fe (dont les ferritines sont une

composante essentielle) (Vansuyt et al., 2000 ; Sappin-Didier et al., 2005). Pour chélater

Fe(III), les graminées exsudent les PS via le transporteur TOM1 (Nozoye et al., 2011), puis le

complexe PS-Fe(III) est prélevé par les cellules du rhizoderme via 2 familles de

transporteurs: Yellow-stripe 1 (YS1) et YS1-like (Curie et al., 2009). Le transporteur

OsYSL15 prend en charge le complexe Fe(II)-DMA. Le transporteur OsYSL16 interviendrait

dans le transport du complexe Fe(III)-DMA. Outre Fe, les phytosidérophores mobilisent

plusieurs métaux tels Ni, Zn, Cd, Mn et Cu (Mench and Fargues, 1994 ; Dell’mour et al.,

2010; Tsednee et al., 2012). La nicotianamine (NA), un précurseur des PS, participe au

prélèvement de Fe dans la rhizosphère des graminées sous la forme Fe(II)-NA, qui est prélevé

via le transporteur OsYSL12.

1.5 Transporteurs impliqués dans le prélèvement et le transfert du cuivre vers les parties

aériennes

Une fois prélevé et transféré via la voie apoplastique jusqu’à la bande de Caspari, Cu est

transféré via la voie symplastique (cytoplasme) dans le péricyle l’endoderme (Figures 1,2,3).

A ce niveau, il est chargé dans le xylème par des protéines de la famille des HMA (Heavy

Metal ATPase) dont HMA5. La HMA5 est exprimée et régulée par la concentration en Cu

dans le péricyle (Andrés-Colas et al., 2006 ; Kobayashi et al., 2008 ; Waters et al., 2011). Le

transporteur HMA9 interviendrait également dans le transport de Cu dans le xylème, voire

dans la phloème (Lee et al., 2007). Les ligands de Cu pendant sa translocation dans le xylème

vers les parties aériennes seraient principalement des acides aminés (histidine, nicotianamine -

NA) (Pich and Scholz 1996 ; Liao et al. 2000) et des acides organiques (citrate) (Puig et al.,

23

2007 ; Curie et al., 2009). La NA participe à la mobilisation de Cu dans la phloème depuis les

plus vieilles feuilles jusqu’aux feuilles plus jeunes et aux graines via le transporteur OsYSL16

(Zheng et al., 2012).

.

Figure 3 : Coupe transversale de racine © Beauchamp's Tenth Horse CC by-nc-nd 2.0

1.6 Transport intracellulaire du cuivre

Le Cu est requis dans le chloroplaste, notamment dans la membrane des thylakoïdes en tant

qu’atome central de la plastocyanine, une protéine transferant les électrons depuis le

complexe b6f du cytochrome vers le photosystème I (PSI). Le Cu est aussi dans le

chloroplaste un cofacteur de la superoxyde dismutase (SOD), qui convertit les superoxydes

(O2-) en peroxyde d’hydrogène (H2O2) pour diminuer les dommages cellulaires causés par la

24

libération de ces superoxydes au cours de la chaine de transport d’électrons (Alscher et al.,

2002). Les transporteurs chloroplastiques de Cu sont de la famille P-type ATPase ; il s’agit de

PAA1 (HMA6) et PAA2 (HMA8) (Shikanai et al., 2003 ; Abdel-Ghany et al., 2005)

(Figures 4, 5). PAA1 est localisé à l’intérieur de l’enveloppe interne du chloroplaste et PAA2

sur la membrane des thylakoïdes. D’autres transporteurs seraient impliqués car le transport de

Cu est actif chez des mutants ppa1paa2. Parmi ces transporteurs, HMA1 - une pompe

Ca2+/métal localisée dans l’enveloppe du chloroplaste - interviendrait pour le transport de Cu

dans le chloroplaste (Moreno et al., 2008). Un autre transporteur de type HMA, HMA7,

nommé aussi RAN1 pour Responsive to Antagonist, transfert Cu dans le réticulum

endoplasmique où il participe à la formation de récepteurs fonctionnels à l’éthylène (Chen et

al., 2002). L’éthylène est un signal du stress abiotique et du stress résultant de l’action des

pathogènes. Une Cu-chaperonne (COX 17) livrerait du Cu à la mitochondrie (Maxfield et al.,

2004), mais aucun transporteur n’est formellement identifié pour cet organite. Un transporteur

de Cu, COPT5 a été caractérisé dans le tonoplaste – membrane qui sépare la vacuole du

cytoplasme – d’A. thaliana (Jaquinod et al., 2007 ; Pilon, 2011 ; Klaumann et al., 2011 ;

Garcia-Molina et al., 2011). Le COPT5 exporterait Cu de la vacuole au cytoplasme où il peut

être réalloué aux autres organites. Selon Klaumann et al. (2011), COPT5 sert à la réallocation

inter-organes de Cu, depuis la racine jusqu’aux parties aériennes. La forme Cu+ est à la fois

peu soluble et hautement réactive dans la cellule. Des protéines chaperonnes sont impliquées

dans son transport. La Copper chaperone (CCH) permet son entrée dans le cytoplasme et son

transport dans le réticulum endoplasmique (Yruela, 2009). L’ATX1 délivre Cu+ au

transporteur P-type ATPase HMA5 (Shin et al., 2012). La protéine CCS (Copper chaperone

superoxyde) transfert Cu+ aux Cu/Zn superoxyde dismutases CSD1 (Cu/Zn superoxyde

dismutase du cytosol), CDS2 (Cu/Zn superoxyde dismutase du chloroplaste) et CDS3 (Cu/Zn

superoxyde dismutase du peroxysome) via une interaction protéine-protéine (Yruela, 2009).

L’expression des protéines CSS et des superoxydes dismutases 1, 2 et 3 diminue lors d’une

carence en Cu ; leur régulation permet une utilisation optimale de Cu pour la photosynthèse

(Abdel-Ghany et al., 2005).

25

Figure 4 : Schéma du transport de Cu dans une cellule de plante générique. Les proteines de transport membranaires sont indiquées en orange, les Cu-Chaperonne en violet et les protéines liées au cuivre en bleu. CCH, Cu chaperone ; ATX, antioxydant 1 ; CCS, Cu chaperone pour la Cu/Zn superoxyde dismutase ; CDS3, Cu/Zn superoxyde dismutase peroxisomale ; COPT, transporteur de Cu; COX, cytochrome-c oxydase ; ER, réticulum endoplasmique ; FRO, ferric reductase oxydase ; HMA, Heavy metal P-type ATPase ; MT metallothionines ; NA, nicotianamine ; PAA, P-type ATPase d’ arabidopsis ; Pc, plastocyanine ; RAN1, responsive to antagonist 1 ; SOD, superoxyde dismutase ; YSL, yellow stripe-like proteine ; ZIP, IRT-like protéine. (Yruela, 2009)

26

Figure 5 : Prédiction de la topologie de la membrane pour certains transporteurs des familles COPT et HMA. Cette topologie est basée sur des prédictions qui n’ont pas été vérifiées expérimentalement (Yruela, 2009).

27

II. Phytotoxicité et détoxification des effets délétères de l’excès de Cu

2.1 Les ROS et le système anti-oxydant

Le cuivre est défini comme un Élément Trace Potentiellement Toxique (en anglais PTTE)

(Adriano, 2001); comme tout élément essentiel au développement de la plante, il peut induire

des symptômes de toxicité (e.g. réduction de la biomasse, inhibition de la croissance racinaire,

bronzing, chlorose, réduction du prélèvement de Fe, Zn et P, perte de l’intégrité du

chloroplaste, etc.) à des expositions supérieures à l’homéostasie cellulaire de Cu (5-20 µg Cu

g-1 MS) (Marschner, 2011). Sa toxicité première vient de sa contribution à la production des

ROS tels les superoxydes (O2-), les radicaux hydroxyles (HO•) et le peroxyde d’hydrogène

(H2O2) notamment via la catalyse des réaction de Fenton, i.e. Fe2+(aq) + H2O2 → Fe3+

(aq) + OH-

(aq) + HO•, et d’Haber-Weiss, i.e. •O2- + H2O2 → HO•+ OH- + O2 (Figure 6)

Figure 6 : Voie de synthèse des ROS lors d’une forte exposition aux métaux (HM). Le prélèvement d’HM et sa distribution aux organites par les transporteurs est suivi d’une production de ROS stimulée par l’activité rédox des HM et leur impact sur les mécanismes sub-cellulaires. L’activation de la NADPH oxidase localisée dans la membrane plasmique par les HM contribue à la production de ROS. Les points rouges symbolisent la distribution des HM dans la cellule et l’apoplaste.

Les ROS produits peuvent peroxyder des lipides et oxyder des protéines et des domaines de

l’ADN, notamment au niveau de la guanine (Drazkiewicz et al., 2004 ; Sharma and Dietz.,

28

2009). En conditions non stressantes, la production de ROS a lieu principalement via les

transferts d’électrons dans le chloroplaste et la mitochondrie. Dans ces conditions, le système

de défense antioxydant maintient la teneur en ROS en dessous du seuil de toxicité. Le stress

provoqué par une exposition supérieure à l’homéostasie du Cu dans les tissus rompt cet

équilibre et stimule l’activité d’enzymes impliquées directement ou non dans l’élimination des

ROS. Ce réseau d’enzymes se base sur un large réseau enzymatique (Figure 7). Les SOD sont

impliquées dans l’élimination d’O2- via les réactions Cu2+/SOD +O2

- → Cu+/SOD+O2 et

Cu+/SOD+O2- → Cu2+/SOD+H2O2. Les catalases (CAT) catalysent la dismutation des

peroxydes, 2H2O2 → O2 + 2H2O. Les peroxyredoxines (PRX) sont également impliquées

dans la dégradation de H2O2, et les ascorbate peroxydases (APX) le dégradent en utilisant

l’ascorbate (AS) comme substrat au cours de la réaction AS + H2O2 → MDHA+ 2 H2O, où

MDHA représente le monodéhydroascorbate (Dietz et al., 2006). AS est régénéré à partir de

MDHA via la MDHA réductase (MDHAR) en parallèle avec l’oxydation de NADPH en

NADP+. Des isoformes de SOD, APX et PRX sont localisés dans l’ensemble des

compartiments cellulaires. A l’inverse, CAT se retrouve majoritairement dans le peroxysome

(Sharma and Dietz., 2009). D’autres enzymes participent au maintien de l’homéostasie

cellulaire. Les glutathion peroxydases (GPX) dégradent les peroxydes via la réaction

2GSH + H2O2 → GS–SG + 2H2O, où GSH représente le glutathion monomérique et GS-SG

le glutathion disulfure (Navrot et al., 2006). Le GSSG (glutathion oxydé) est à nouveau réduit

en GSH via la glutathion réductase (GR) en parallèle de l’oxydation de NADPH en NADP+.

(Figure 7). Les glutathion-S-transferases (GST) ainsi que les syringaldazine peroxydases

(SPOD) et les Guaiacol peroxydases (GPOD) sont également impliquées dans l’élimination

d’H2O2 (Cuypers et al., 2002 ; Hung et al., 2005), tout comme la proline et l’α-tocophérol

(Sharma and Dietz 2009). La L-proline, libre (PRO), s’accumulerait dans les plantes soumises

aux fortes expositions aux PTTE (Sharma and Dietz, 2006) et aurait un rôle d’antioxydant.

Sous une forte exposition de Chlorella reinhardtii au Cd, le pool de GSH est 4 fois plus

oxydé chez les individus non transgéniques, à faible teneur en proline dans la cellule, que

chez les individus transgéniques surexprimant la delta-1-pyrroline-5-carboxylate synthétase

(P5CS), l’enzyme de synthèse de PRO (Siripornadulsil et al., 2002). In vitro, la résonnance

paramagnétique électronique (EPR) révèle que la proline piège les ROS •OH et 1O2 (Kaul et

al., 2008). L’α-tocopherol réagit principalement avec 1O2 et au dernier stade de la chaine de

réaction lors de la peroxydation des lipides (Maeda et al., 2007). La concentration en α-

tocophérol augmente lorsque l’exposition au Cu et au Cd s’accroit chez A. thaliana (Collin et

al., 2008).

29

Figure 7 : Localisation des voies d’élimination des ROS dans la plante (Mittler et al., 2004).

Les réponses antioxydantes ne sont pas les seules que la plante développe lors d’expositions

excessives au Cu. Les phytochélatines (PC), métallothionines (MT) et la compartimentation

vacuolaire ont également un rôle prépondérant en évitant d’avoir des ions Cu2+ libres. Les

phytochélatines sont des peptides de type γ-Glu-(Cys)n-Gly, où n=2-11, qui sont exprimés en

réponse aux fortes expositions de certains métaux, dont Cu, et métalloïdes (Morelli and

Scarano., 2004). Elles sont synthétisées via la PC synthase en utilisant le glutathion comme

substrat (Hall, 2002). Les PCs sont des ligands dans le cytosol, la chélation ayant lieu entre les

unités sulfhydryles (-SH) (= thiol) de la cystéine et l’ion métallique. Les PC sont aussi des

ligands pour la forme As(III). Le complexe PC

transporteurs ABCC1 et ABCC2 ou bien les ions métalliques

par des acides organiques tandis que les PC sont dégradés en acides aminés. (

MT contiennent des cystéines et possèdent une capacité de chélation des

similaires aux PC et de stockage dans la vacuole.

restaurer la tolérance au Cu de levures (

and Goldsbrough, 1994) et la tolérance au Cu chez

élevés de MT (Van Hoof et al.,

Figure 8 : Synthèse et fonctionnement des phytochélatines (ABCC Transporteur ATP binding-synthase, GLU Glutamate, GLY Glutathion-S-Transferase, GSH Glutathion MonomériqueDisulfure, HMWC High Molecular Weigh Compound

2.2 Impacts sur le développement

La croissance racinaire, comme l’organisation du système racinaire, sont considérés comme

des traits morphologiques pouvant intégrer des mécanismes de tolérance dont

place aux expositions en excès de Cu. Une exposition à 50 µM Cu

racines primaires et secondaires chez

production de biomasse racinaire est aussi réduite par l’excès de

(Alaoui-Sossé et al., 2004), Zea mays

2003). Les impacts sur la fonction de prélèvement des racines s’accompagnent souvent d’une

30

SH) (= thiol) de la cystéine et l’ion métallique. Les PC sont aussi des

ligands pour la forme As(III). Le complexe PC-ion est transporté dans la vacuole via des

transporteurs ABCC1 et ABCC2 ou bien les ions métalliques (Cd, Zn, Cu, etc.)

par des acides organiques tandis que les PC sont dégradés en acides aminés. (

MT contiennent des cystéines et possèdent une capacité de chélation des

similaires aux PC et de stockage dans la vacuole. Des gènes de MT d’A

restaurer la tolérance au Cu de levures (Saccharomyces cerevisae) déficientes en MT

a tolérance au Cu chez Silene vulgaris est associée à des niveaux

al., 2001).

Synthèse et fonctionnement des phytochélatines (d’après Mendoza--cassette C, CYS Cystéine, EC glutamylcystéine, Glycine, GPX Glutathion Peroxydase, GR Glutathion Réductase, Glutathion Monomérique, GS Glutathion synthase

High Molecular Weigh Compound, PC Phytochelatine, PCS Phytochelatine Synthase

2.2 Impacts sur le développement racinaire

La croissance racinaire, comme l’organisation du système racinaire, sont considérés comme

des traits morphologiques pouvant intégrer des mécanismes de tolérance dont

place aux expositions en excès de Cu. Une exposition à 50 µM Cu2+ réduit la production de

racines primaires et secondaires chez A. thaliana (Figure 9) (Lequeux

production de biomasse racinaire est aussi réduite par l’excès de Cu chez

Zea mays (Ouzounidou et al., 1995), et A.thaliana

). Les impacts sur la fonction de prélèvement des racines s’accompagnent souvent d’une

SH) (= thiol) de la cystéine et l’ion métallique. Les PC sont aussi des

ion est transporté dans la vacuole via des

(Cd, Zn, Cu, etc.) sont chélatés

par des acides organiques tandis que les PC sont dégradés en acides aminés. (Figure 8). Les

MT contiennent des cystéines et possèdent une capacité de chélation des ions métalliques

A. thaliana peuvent

) déficientes en MT (Zhou

est associée à des niveaux

-Cozatl et al., 2010). ECS glutamylcysteine

Glutathion Réductase, GST Glutathion synthase, GSSG Glutathion

Phytochelatine Synthase.

La croissance racinaire, comme l’organisation du système racinaire, sont considérés comme

des traits morphologiques pouvant intégrer des mécanismes de tolérance dont ceux mis en

réduit la production de

Lequeux et al., 2010). La

Cu chez Cucumis sativus

A.thaliana (Wojcik et al.,

). Les impacts sur la fonction de prélèvement des racines s’accompagnent souvent d’une

31

modification du ionome, avec réduction du transfert de Ca2+ des racines aux parties aériennes

et de la concentration en K+ des racines (Maksymiec et al., 1998 ; Murphy et al., 1999 ;

Lequeux et al., 2010). L’efflux de K+ est lié à l’efflux de citrate induit par Cu2+. Cet efflux de

citrate est nécessaire car Cu inhibe une forme cytosolique de l’aconitase, conduisant à une

accumulation de citrate dans la cellule (Murphy et al., 1999).

Figure 9 : Adaptation morphologique du système racinaire d’Arabidopsis thaliana face à une exposition hétérogène au cuivre. Lorsque les deux premières racines apparaissent, la plantule est mise en culture pendant 5 jours en conditions de contamination hétérogènes, chaque racine dans un compartiment différent (A 0, 50 µM Cu) ou en conditions homogènes (B 0 µM, C 50 µM Cu). Les pointillés correspondent à l’endroit ou les racines ont été disposées à t = 0. (barres = 1 cm) (Lequeux et al., 2010).

L’excès de Cu2+ peut entrainer une accumulation de Fe et Zn dans les racines (Lequeux et al.,

2010). Les transporteurs de type CDF (Cation Diffusion Facilitator) transférant des métaux

dans le cytoplasme ont en général une sélectivité faible (Krämer et al., 2007), permettant une

compétition entre Fe, Zn et Cu, et à priori de faibles concentrations en Fe et Zn dans les

racines. De fait, l’activité enzymatique de la peroxydase induite par l’exposition au Cu

32

entraine une polymérisation de composés phénoliques qui favoriserait la soprtion de cations à

la surface des racines (Mench et al., 1987; Lavid et al., 2001 ; Jung et al., 2003).

La division cellulaire dans la racine est affectée par une exposition en excès de Cu (Figure

10). Chez A. thaliana, l’ajout de 50 µM Cu au milieu de culture entraine une mort des cellules

de l’apex des racines primaires et secondaires (Lequeux et al., 2010). Ceci est du à une

conséquence de l’augmentation des ROS dans les cellules (Pasternak et al., 2005) entrainant

un arrêt de la division mitotique (Jiang et al., 2001). Le statut hormonal des plantes est

modifié par l’excès de Cu. L’auxine, la cytokinine et l’éthylène sont des hormones

fondamentales dans l’établissement de l’architecture racinaire (Aloni et al., 2006). L’auxine

est principalement synthétisée dans les jeunes feuilles puis transportée via le phloème vers les

méristèmes apicaux des racines. Elle favorise l’émergence des racines secondaires (Nibau et

al., 2008). Son accumulation, en excès de Cu, juste au dessus de la zone apicale est impliquée

dans la formation de racines latérales (Lequeux et al., 2010). La cytokinine est surtout formée

dans la racine principale et a un rôle inhibiteur de l’activité mitotique, dans la racine

principale et les racines secondaires (Nibau et al., 2008 ; Ruzicka et al., 2009). Cette hormone

est produite en grande quantité à 50 µM Cu (Lequeux et al., 2010). L’éthylène est une

hormone pouvant inhiber la croissance racinaire. Son rôle en excès de Cu n’est pas élucidé car

de fortes expositions n’entrainent pas automatiquement une sur-prodution de cette hormone.

L’éthylène ne serait pas impliqué dans la réorganisation de l’architecture racinaire aux fortes

expositions au Cu (Lequeux et al., 2010). L’excès de Cu a aussi pour conséquence une

production de lignine (Figure 11) (Chen et al., 2002; Lin et al., 2005; Ali et al., 2006). La

production des précurseurs de la lignine est catalysée par les peroxydases et les laccases, des

glycoprotéines contenant Cu (Lequeux et al., 2010) dont l’activité est corrélée à l’exposition

au Cu (Chen et al., 2002). Cette rhizodéposition de lignine, au niveau de l’endoderme, a pour

probable conséquence une limitation des efflux d’éléments (ex : Ca) depuis le cylindre

vasculaire vers les parties aériennes (Van de Mortel et al., 2008) ainsi qu’une limitation de la

croissance cellulaire (Sasaki et al., 1996).

33

Figure 10 : Diagramme schématique de l’impact d’une forte exposition au Cu sur le remodelage de l’architecture racinaire à travers les changements dans le statut hormonal, la déposition de lignine et dans l’activité (Lequeux et al., 2010)

Figure 11 : Effet d’un excès de Cu sur la rhizodéposition de lignine chez A. thaliana cultivé sur Agar. A coloration au Phloroglucinol de la racine principale 7 jours après transfert dans des milieux à concentration croissante en cuivre [0 à 50 µM]. (barre = 100 µm), B Zoom sur la racine traitée avec 50 µM Cu (barre = 30 µm) (Lequeux et al., 2010).

34

Bref aperçu des techniques d’imagerie actuellement utilisées pour la caractérisation fonctionnelle de

l’homéostasie des PTTE dans les plantes

L’activité enzymatique des APX, CAT et SOD est souvent utilisée pour quantifier la réponse

au stress induit par l’exposition en excès de la plante aux PTTE (Thounaojam et al., 2012).

L’imagerie par rayon X et détection de fluorescence permet de compléter cette approche par

la caractérisation fonctionnelle des gènes gérant l’homéostasie cellulaire des PTTE (Donner et

al., 2012). Le micro X-ray fluorescence mapping (µ-XRF), la micro computed tomography

(µ-CT), l’extended X-ray absorption fine structure spectroscopy (EXAFS) et le micro X-ray

absorption near edge spectroscopy (XANES), apportent également des informations sur la

localisation, les atomes proches voisins et la spéciation chimique des PTTE, par l’image ou

l’analyse de spectres d’absorption des rayons X dans des tissus hydratés, in vivo, à des

résolutions plus fines que le micromètre. Ces méthodes indiquent une large gamme de

transporteurs pour le transport du Cu dans la cellule (Klaumann et al., 2011).

• µ-XRF (micro X-ray fluorescence mapping) Lorsque l'on bombarde de la matière avec des rayons X,

la matière réémet de l'énergie sous la forme, entre autres, de rayons X. Leur spectre est caractéristique

de la composition de l'échantillon. En analysant ce spectre, on peut en déduire la composition

élémentaire, c'est-à-dire les concentrations massiques en éléments.

• EXAFS (Extended X-ray absortption fine structure spectroscopy) et XANES (X-ray absorption

near edge spectroscopy): L’EXAFS est une technique d'analyse de spectrométrie d'absorption des

rayons X utilisant principalement le rayonnement synchrotron – le rayonnement électromagnétique

émis par une particule chargée qui se déplace dans un champ magnétique et dont la trajectoire est

déviée par ce champ magnétique. Elle apporte des informations sur l'environnement atomique d'un

élément donné. Elle est généralement couplée avec la technique XANES (X-Ray Absorption Near Edge

Structure)..

• µ-CT (micro computed tomography). Technique non destructive permettant la reconstruction (2D

et 3D) de l’échantillon. Les images 3D de l'intérieur d'un objet sont obtenues en réalisant une série de

radiographies 2D sous de nombreux angles de vue. Un algorithme de reconstruction permet de

recalculer, à partir de l'ensemble des projections 2D, le coefficient d'absorption de l'objet en chaque

point du volume de l'objet. On obtient ainsi une représentation quantitative des variations de densité

au sein de l'objet.

Avec l’utilisation conjointe de la synchrotron-based X-ray fluorescence microscopy (S-XRF)

et de l’X-ray absorption spectroscopy (XAS), Kopittke et al. (2011) ont identifié la

35

distribution et la spéciation de Cu (et de Zn) in situ, dans des tissus hydratés de racines de

Vigna unguiculata (Figure 12). Le Cu s’accumule principalement dans le rhizoderme –

l’assise pilifère qui recouvre la région absorbante de la racine – et dans les tissus de l’apex

racinaire, mais très peu dans le cylindre central. Ces résultats s’accordent avec la sorption de

Cu sur les parois cellulaires (Nishizono et al., 1987). En XAS, après 24h d’exposition, une

quantité importante de Cu est retrouvée fixée sur les acides polygalacturoniques (l’un des

composants principaux de la pectine) (Kopittke et al. 2011). Sur ces résultats, Kopittke et al.

(2008) formule l’hypothèse selon laquelle en se sorbant sur les parois cellulaires du

rhizoderme, Cu rend la couche cellulaire externe de la racine plus rigide que la couche

cellulaire interne, moins exposée (Figure 9). Ceci provoque des zones de ruptures lorsque les

cellules du cylindre central – peu rigides - se développent plus vite que les cellules du

rhizoderme – plus rigides.

Figure 12 Distribution de Cu, Ni et Zn dans des tissus racinaires hydratés de Vigna unguiculata examinée par S-XRF. L’intensité du signal (concentration) est indiquée par l’échelle de couleur. (Barre d’échelle : 1mm) (Kopittke et al., 2011)

36

III. Le cuivre et les macrophytes

Les macrophytes – l’ensemble des plantes aquatiques, immergées, émergées, racinées ou non,

et se développant soit directement dans l’eau soit sur un substrat – sont reconnues pour leur

tolérance plus élevée aux PTTE que les plantes terrestres (Sinha and Pandey, 2003; Marchand

et al., 2010). Leur participation à la décontamination d’une masse d’eau contaminée sera

détaillée dans la seconde partie de l’introduction, ainsi que dans les chapitres I et IV.

Les macrophytes diffèrent peu des plantes terrestres pour les mécanismes de tolérance à une

exposition excessive au Cu ; elles mettent en place un réseau d’enzymes similaire impliquant

APX, GPX, Cu-SOD ainsi que l’utilisation des MT et PC (Srivastava et al., 2006 ; Huang and

Wang, 2010 ; Delmail et al., 2011). Elles présentent cependant un stockage facilité dans les

vacuoles de leurs racines et rhizomes (Caldelas et al., 2012). Avec une exposition croissante

de 0,1 à 25 µM Cu, Hydrilla verticillata accumule jusqu’à 770 µg Cu g-1 MS dans ses tissus,

alors que le seuil de toxicité est défini aux alentours de 20-30 µg Cu g-1 MS (Srivastava et al.,

2006). Cette accumulation de Cu s’accompagne d’une diminution du ratio de chlorophylle a/b

et de l’ensemble des teneurs en pigments, se traduisant par une chlorose foliaire et une

réduction de la biomasse de la plante entière. La teneur en malondialdehyde (MDA)

augmente, traduisant un stress oxydant et la production de ROS. En réponse, les activités

SOD et APX s’accroissent sur l’intervalle de concentrations en Cu testées, pour éliminer les

radicaux superoxydes dès leur genèse, alors que celles des GPX et CAT ne sont stimulées

qu’aux faibles expositions (0,1 - 1 µM Cu). Ces résultats recoupent les données chez

Ceratophyllum demersum L (Devi and Prasad, 1998). Les MTs ont un rôle chez les

macrophytes. L’excès de Cu augmente la production des MTs de type 2 dans les feuilles de

Bruguiera gymnorrhiza (Huang and Wang, 2010), d’Avicennia germinans (Gonzalez-

Mendoza et al., 2007) et d’Avicennia marina (Huang and Wang., 2010). Tous les

macrophytes ne disposent pas du même arsenal pour faire face à une forte exposition au Cu

(et aux PTTE en règle générale). Selon Larue et al. (2010), l’activité et l’exsudation de

peroxydases sont plus élevées chez Phragmites australis que chez Typha latifolia et Iris

pseudacorus. A l’inverse, notamment au début de la période de croissance, I. pseudacorus a

des concentrations en phénols dans les racines plus importantes que T. latifolia et P. australis.

Ces phénols complexeraient des PTTE, dont Cu, dans les cellules mais aussi dans la

rhizosphère lorsqu’ils sont exsudés. L’exsudation de composés (ex: acides organiques,

mucilage) permettant la sorption du Cu dans la rhizosphère est communément admise, avec

37

des nuances. Aux fortes expositions en Cu, l’exsudation d’oxalate par Juncus maritimus est

stimulée mais pas celle d’autres acides organiques tels le citrate, le malate, le malonate ou le

succinate. Scirpus maritimus n’exsude pas d’acide organique en réponse à une forte

exposition au Cu (Mucha et al., 2010). Un grand nombre de macrophytes, notamment ceux de

la classe des géophytes rhizomateuses (Raunkiaer, 1934) produisent des rhizomes (Figure

13). Lors de fortes expositions aux PTTE, dont Cu, une première étape de détoxification est

leur sorption sur les membranes cellulaires du rhizoderme, conjointement avec la cascade de

réactions enzymatiques impliquées dans la détoxification du Cu décrite ci-dessus. Lors d’une

forte exposition chronique, ce système ne suffit pas à contenir le contaminant sur les parties

externes de la racine. Les PTTE pénètrent alors dans les espaces intercellulaires et le

cytoplasme des rhizomes et des racines. Il sont piégés dans les vacuoles, par formation de

granules de stockage (Caldelas et al., 2012). Cette compartimentation dans les vacuoles

rhizomatiques confère aux macrophytes une capacité à mieux tolérer les expositions aux

PTTE, dont Cu.

Figure 13 : Vue en microscopie à transmission d’électrons de cellules de parenchyme cortical de rhizomes d’Iris pseudacorus. (a) Plantes control, (b, c) plantes exposées à 0.75 mM Cr(III). am: amyloplates, g: granule, vac: vacuole. (a, b) grossissement x 3000, (c) grossissement x 4500. (Caldelas et al., 2012).

Références : Les références de cette première partie de l’introduction sont présentées

conjointement avec celles de la synthèse générale, à la suite de la synthèse générale.

38

Introduction : partie II

La phytoremédiation, les éléments traces, l’eau, les macrophytes et les zones

humides construites

Cette partie a été publiée sous forme d’une review dans Environmental Pollution

(158. pp 3447-3461) en 2010.

39

Environmental Pollution, 158. (3447-3461).

Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants and standardized measurements: a review

L. Marchand1, M. Mench1, D.L. Jacob2, M.L. Otte2

1UMR BIOGECO INRA 1202, Écologie des communautés, Université Bordeaux 1, Bat B8 RDC Est, Avenue des facultés, F-33405 Talence, France. lilian.marchand@hotmail.fr

2Wet Ecosystem Research Group, Department of Biological Sciences, NDSU Dept. 2715, P.O. Box 6050, Fargo, ND 58108-6050, USA

Abstract

This review integrates knowledge on the removal of metals and metalloids from contaminated

waters in constructed wetlands and offers insight into future R&D priorities. Metal removal

processes in wetlands are described. Based on 21 papers, the roles and impacts on efficiency

of plants in constructed wetlands are discussed. The effects of plant ecotypes and class

(monocots, dicots) and of system size on metal removal are addressed. Metal removal rates in

wetlands depend on the type of element (Hg>Mn>Fe=Cd>Pb=Cr>Zn=Cu>Al>Ni>As), their

ionic forms, substrate conditions, season, and plant species. Standardized procedures and data

are lacking for efficiently comparing properties of plants and substrates. We propose a new

index, the relative treatment efficiency index (RTEI), to quantify treatment impacts on metal

removal in constructed wetlands. Further research is needed on key components, such as

effects of differences in plant ecotypes and microbial communities, in order to enhance metal

removal efficiency.

Keywords: metal, pollution, ecotype, macrophyte, micro-organism, phytoremediation,

wastewater

40

I. Rationale for using constructed wetlands for improvement of water quality

Aquatic ecosystems are used directly or indirectly as recipients of potentially toxic liquids and

solids from domestic, agricultural and industrial wastes (Demirezen et al., 2007; Peng et al.,

2008). Pollutants may accumulate in surface waters, groundwater, substrates and plants

(Aksoy et al., 2005), and can be divided into four classes: nutrients (P, N), organic

contaminants (e.g. polycyclic aromatic hydrocarbons, polychlorinated biphenyls and

pesticides), xenobiotics derived from the pharmaceutical industry (personal care products,

hormones, etc.) and metals and metalloids (e.g. Cu, Zn, Fe, Cd, Ni, Pb, Hg, Cr, Sr, Al, Ba, Se

and As). For convenience, we will refer to both metal and metalloids as 'metals' throughout

the paper. Constructed wetlands have been used to successfully improve the quality of

contaminated waters and wastewaters for at least two decades (Murray-Gulde et al., 2005a;

Maine et al., 2009; Zhang et al., 2010) while natural ‘volunteer’ wetlands have been

improving water quality over millions of years. Of special interest in this respect are volunteer

wetlands associated with mining activities (Beining and Otte, 1996, 1997). The functions of

macrophytes include production of organic matter, pollutant uptake and bioengineering of the

rhizosphere, e.g. maintenance of habitats for micro-organisms.

Some macrophytes have accumulator phenotypes for one or several metals (Kamal et al.,

2004). These plants can accumulate metals in concentrations 100,000 times greater than in the

associated water and therefore have been used for metal removal from a variety of sources

(Mishra et al., 2008). Hyperaccumulators can tolerate, take up and translocate high levels of

certain metals that would be toxic to most organisms. They are defined as plants that complete

their life cycle with foliar metal concentrations exceeding (mg kg-1 dry weight, DW) Cd >100,

Ni and Cu >1,000, and Zn and Mn > 10,000 (Zavoda et al., 2001). However, most of works

on metal (hyper) accumulators has been done on dryland plants. To date no emergent wetland

plants have been identified as hyperaccumulators. Metal removal through uptake by

macrophytes in wetlands is therefore relatively minor compared to other processes. The main

functions of these plants are to provide organic matter needed to perpetuate the

biogeochemical processes in the substrate through die-back, and to provide organic

compounds via exudation from the roots (Jenssen et al., 1993). These characteristics make

them essential in constructed wetlands, which are considered effective, cost-efficient and

41

environmentally friendly bio-processes for quality improvement of polluted water (Brix,

1997; Stottmeister et al., 2003).

Constructed wetlands can be classified into types based on their characteristics. One system

defines four types based on the dominant plant species (Arias and Brix, 2003): (1) floating

macrophytes (e.g. Eichhornia crassipes, Lemna minor), (2) floating-leaf macrophytes (e.g.

Nymphea alba, Potamogeton gramineus), (3) submersed macrophytes (e.g. Littorella uniflora,

Potamogeton crispus), and (4) emerged rooted macrophytes (e.g. Phragmites australis, Typha

latifolia). Another common classification divides wetlands according to their hydrology: (1)

surface flow wetlands, (2) horizontal subsurface flow wetlands, (3) vertical subsurface flow

wetlands, and (4) hybrid systems (Arias and Brix, 2003).

II. Metal removal processes in wetlands

Four mechanisms affect metal removal in wetlands (Lesage et al., 2007a): (1) adsorption to

fine textured sediments and organic matter (Gambrell, 1994), (2) precipitation as insoluble

salts (mainly sulphides and oxyhydroxides), (3) absorption and induced changes in

biogeochemical cycles by plants and bacteria (Kadlec and Knight, 1996), and (4) deposition

of suspended solids due to low flow rates. All these reactions lead to accumulation of metals

in the substrate of wetlands. The efficiency of systems depends strongly on (i) inlet metal

concentrations and (ii) hydraulic loading (Kadlec and Knight, 1996).

2.1 Adsorption

Sorption, the transfer of ions from a soluble phase to a solid phase, is an important mechanism

for removal of metals in wetlands. It may result in short-term retention or long-term

stabilization. Sorption describes a group of processes which includes adsorption

(physisorption, i.e. physical processes with weak bindings, chemisorption, i.e. chemical

processes with strong bindings), absorption (e.g. with biochemical processes when a

compound from the external media is entering into a living organism) and precipitation

42

reactions. Metals are adsorbed to particles by either ion exchange depending upon factors

such as the type of element and the presence of other elements competing for adsorption sites

(Seo et al., 2008) or chemisorption. Retention of Pb, Cu, and Cr by adsorption is greater than

Zn, Ni, and Cd (Sheoran and Sheoran, 2006).

Freundlich and Langmuir models (Bohn et al., 1979) may be used to determine maximum

metal immobilization and their retention capacity over time (Lesage et al., 2007b; Seo et al.,

2008), as follows:

Freundlich, q = KCe(1/n), to describe adsorption onto a heterogeneous surface, where q =

adsorption capacity (metal concentration on adsorbing material, mg kg-1), K = constant linked

to adsorption capacity, Ce = concentration (mg L-1), and n = empiric parameter linked to

sorption intensity.

Langmuir, q = abCe/(1+bCe), for monolayer adsorption on a homogeneous surface with a

finite number of identical sorption sites, where q = mass of metal adsorbed to the substrate

(mg kg-1), Ce = concentration at the equilibrium (mg L-1), a = adsorption capacity maximum

to the substrate (mg kg-1), and b = a constant linked to the metal fixation force.

Another useful parameter to quantify adsorption capacity of a material for an ion is the

distribution coefficient Kd (Alloway, 1995):

To quantify the retention capacity of substrates over time, column experiments may be used.

In this case, the Thomas model is relevant (Seo et al., 2008):

Where Co = metal concentration at the influent (mg L-1), C = metal concentration at the

effluent (mg L-1), Kth = Thomas constant (L D-1 mg-1), Q = flow rate (L D-1), qo = maximum

43

adsorption capacity (mg g-1), M = total mass of sorbing (g), and V = Total volume of solution

passed through the column (L).

2.2 Co-precipitation and redox reactions

Some metals, e.g. Fe, Al, and Mn, can form insoluble compounds through hydrolysis and/or

oxidation. This leads to formation of a range of oxides, oxyhydroxides and hydroxides

(Sheoran and Sheoran, 2006). Fe removal depends on pH, redox potential and the presence of

anions (ITRC, 2003). The amounts and forms of Fe in solution strongly affect metal removal.

Fe(II) is soluble and represents an important bioavailable fraction. It can be oxidized to Fe(III)

in conjunction with H+ ion consumption under aerobic conditions (Jönsson et al., 2006).

Fe(III) can deposit onto root surfaces of aquatic macrophytes (Weiss et al., 2003), forming

plaques with a large capacity to adsorb metals (Doyle and Otte, 1997; Cambrolle et al., 2008),

aided also by the action of (Fe(II)-oxidizing bacteria (Emerson et al., 1999). Fe(III) can

precipitate to produce oxides, hydroxides and oxyhydroxides with which other metals may co-

precipitate. Fe(II) can also precipitate as oxides (Jönsson et al., 2006) or co-precipitate with

other metals such as Zn, Cd, Cu or Ni (Matagi et al., 1998). Iron oxides have a particularly

strong affinity for cations with a similar size compared to Fe(III) and Fe(II), e.g. Zn, Cd, Cu

and Ni (Dorman et al., 2009). Therefore, those TE may combine with Fe forming metal-oxide

complexes (Benjamin and Leckie, 1981). This co-precipitation is limited when Fe(II) forms

complexes with for example SO42- (Sung and Morgan, 1980), thus reducing the potential of

metal removal. Arsenic may also be removed from the water column by adsorbing onto

amorphous iron hydroxides or by co-precipitating with iron oxyhydroxides (Manning et al.,

1998).

Metals can also form insoluble compounds through reduction. Under chemically reducing

conditions (Eh<50 mV) sulfates can be reduced to sulfides. These can combine with various

elements, i.e. As, Hg, Se and Zn, to co-precipitate in relatively insoluble forms (Murray-

Gulde, 2005b). Some macrophytes, e.g. Schoenoplectus californicus, contribute to reductive

conditions into the substrate (Dorman et al., 2009).

44

A constructed wetland based on a matrix with exclusively reducing conditions, however,

cannot be efficient. These conditions promote massive ion release, particularly of Fe and Mn,

into the water by reduction of the oxides and oxyhydroxides trapped in the substrate (Goulet

and Pick, 2001). Most metals in the porewater precipitate as metal oxides or adsorb onto

organic matter at redox potentials higher than 100 mV. Between 100 mV to -100 mV, metal

oxides are reduced resulting in a release of dissolved metals. These metals can still adsorb

onto organic matter if adsorption sites are available. Below –100 mV metals may be mainly

associated with sulfides (Goulet and Pick, 2001). Most macrophytes play a role in maintaining

oxidizing conditions by shoot-to-root oxygen transport (Armstrong, 1978). Such conditions

promote formation of iron oxides, hydroxides and oxyhydroxides, such as the iron-plaques,

and consequently result in metal removal by adsorption and co-precipitation. Considering the

processes essential in metal removal it may be useful to design a system with two separate

compartments, i.e. a first compartment to promote sulphate reduction and induce their

combination with As, Hg, Se, and Zn ions, and a second, oxidizing compartment to enhance

metal co-precipitation with iron oxides (Dorman et al., 2009). Oxidation can be promoted by

an aeration cascade between both compartments (Schwarz et al., 2009). Adequate physico-

chemical substrate conditions offer an efficient matrix for metal removal. However, without

any plants the substrate will become devoid of organic matter, thus decreasing the capacity of

substrates to maintain sulphate reduction and metal immobilization (Jacob and Otte, 2004).

Arsenic, being a redox-sensitive metalloid, is potentially toxic and of major concern with

respect to its accumulation in waters and soils and so deserves a discussion separate from

metals. Arsenate, As(V), is the main As species in aerobic soils whereas arsenite, As(III), is

the dominant species in reducing environments such as flooded paddy soils (Zhao et al.,

2009). Since As(III) is more mobile and toxic than As(V), active As remediation may require

conversion of As(III) to As(V) in the rhizosphere and subsequent immobilization of As(V) by

adsorption or co-precipitation (Guan et al., 2009). Fe oxides in the rhizosphere have a strong

adsorptive capacity for arsenate. Concentrations of As in iron plaques of rice are about 5-fold

higher than those in root tissues (Liu et al., 2006). However, arsenic is not the only inorganic

ion present in natural waters and adsorption of As(V) and As(III) oxyanions by ferric

hydroxide may be adversely affected by anions such as carbonate, sulphate, phosphate,

silicate, and also by organic matter (Meng et al., 2000). Some terrestrial plants (for example

Pteris vittata and Miscanthus sinensis) show a high As removal efficiency (Ma et al., 2001),

45

but such accumulators have not been described for wetlands. Some reports suggest that

arsenic is more difficult to remove from wastewater, citing efficiencies of 5% by Eleocharis

acicularis (Ha et al., 2009) and of less than 2% for Schoenoplectus acutus (Crites et al.,

1997). On the other hand, Beining and Otte (1996, 1997) report removal rates of 60% in a

natural ‘volunteer’ wetland dominated by the grass Molinia caerulea.

2.3 Metal carbonates

Metals may also form metal-carbonates. Although carbonates are less stable than sulphides,

they can contribute to initial trapping of metals (Sheoran and Sheoran, 2006). Carbonate

precipitation is especially effective for the removal of Pb and Ni (Lin, 1995). According to

Maine et al. (2006) the incoming wastewater composition containing high pH, carbonate and

calcium concentrations favoured the metal retention in the sediment. Calcium carbonate

precipitation represents an important pathway governed by the incoming water pH. Metals are

removed adsorbed to carbonates.

2.4 pH

The pH strongly affects the efficiency of metal removal in wetlands. Ammonium conversion

into nitrites during nitrification leads to proton production. These hydrogen ions are then

neutralized by bicarbonate ions. Macrophytes, in releasing oxygen, promote the nitrification

process. Protons produced due to nitrification may not all be neutralized by HCO3- ions,

resulting in a pH decrease (Lee and Scholz, 2007). The overall mean surface charge of ferric

(oxyhydr)oxides changes from a positive to a negative value as pH increases. Hence, to

promote adsorption and removal of oxyanions of, for example, As, Sb, and Se, iron co-

precipitation must occur under acidic conditions (Sheoran and Sheoran, 2006). Conversely,

alkaline conditions are necessary to promote co-precipitation of cationic metals, such as Cu,

Zn, Ni, and Cd. A high rate of nitrification can therefore reduce the efficiency of a constructed

wetland in terms of cationic metal removal (Lee and Scholz, 2007).

In the special case of acid mine drainage, the water and substrates are characterized by high

metal concentrations and a low pH. When sulfide minerals contained in mine drainage are

46

exposed to atmospheric and dissolved oxygen they are oxidized (Holmstron, 2000). For

example, pyrite (FeS2) is oxidized to form soluble Fe(II), SO4 2- and H+. This leads to release

of dissolved iron and protons, which, in turn, leads to the release of other metallic ions, such

as Mn, Ni, Zn, Cu and Cd. Because of the extreme conditions of acid mine drainage, wetlands

may have a low potential for water treatment (Nyquist and Greger, 2009). Yet, to date more

than a thousand wetlands had been constructed specifically for that purpose (Skousen and

Ziemkiewicz, 1995). The formation of acid mine drainage can be prevented by limiting

contact between mining wastes and oxygen. One attractive and efficient solution for reducing

O2 diffusion is to construct a wetland as a cover over mine waste (Stoltz, 2005).

2.5 Erosion, sedimentation

Macrophytes, such as Phragmites australis, promote sedimentation of suspended solids and

prevent erosion by decreasing water flow rates by increasing the length per surface area of the

hydraulic pathways through the system (Lee and Scholz, 2007). Surface flow systems may

exhibit either static, in which there is virtually no flow, or dynamic behaviour, in which water

is passing through at relatively high flow rates. Under static conditions, the wetland behaves

like a stagnant pond in which displacement effects caused by submerged plant mass decrease

retention times. Under dynamic conditions active flow-through and stem drag is increased and

is more important than displacement of volume. Retention times increase with increasing

vegetation density, thus enabling better sedimentation. For particles less dense than water,

sedimentation is possible only after floc-formation. Flocs may adsorb other types of

suspended materials, including metals. Flocculation is enhanced by high pH, high

concentrations of suspended matter, high ionic strength and high algal densities (Matagi et al.,

1998).

2.6 The role of micro-organisms

The vicinity of plant roots, the rhizosphere, is a preferred environment for many soil micro-

organisms. Approximately 1.2x1011 cells per cm3 live within a distance of less than 1 mm

from the roots, whereas the numbers at a distance of 2 cm are at least one order of magnitude

lower. The root surface is covered with bacteria, and growing roots may transport bacteria

through the soil (Trapp and Karlson, 2001). Rhizosphere bacteria are predominantly gram-

47

negative (Chaudhry et al., 2005). This may be due to their ability to utilize efficiently growth

substrates available in the rhizosphere and to cope with toxic environments due to the

presence of detoxifying enzymes (Chaudhry et al., 2005). The genera commonly found in

rhizospheres include Rhizobium, Azotobacter and Pseudomonas (Hoagland, 1985).

Besides forming a habitat for micro-organisms, plant roots also provide substrates such as

sugars in exchange for phosphate or nitrogen (N2-fixation). Organic compounds exuded by

the roots, fungi and bacteria, e.g. saponines proteins, and enzymes, may mobilize soil-born

pollutants, including metals (Trapp and Karlson, 2001). Plants offer nutrition and protection

against the physical environment and act as hosts for endophytic bacteria.

Free-living, plant growth-promoting rhizobacteria, as well as symbiotic bacteria can improve

plant nutrition and growth, and plant competitiveness, as well as responses to external stress

factors such as contaminant exposure (Doty, 2008). Symbiotic bacteria complement certain

metabolic properties of their host, e.g. atmospheric N fixation, protection against pathogens

and contaminant detoxification (Mastretta et al., 2009). The main mechanisms involved in the

growth-promoting effects of bacteria are associated with the production of hormones and

siderophores. Excessive ethylene production promoted by stress can depress growth. Many

bacteria facilitate plant growth by metabolizing 1-aminocyclopropane-1-carboxylate (ACC),

the immediate precursor of ethylene, through the synthesis of ACC deaminase (ACCD).

Bacterial strains with ACCD activity can prevent reduction in root and shoot growth resulting

from stressful conditions. Microbial siderophores can interact with metals, reducing their

toxicity or increasing labile metal pools and uptake by roots (Lemanceau et al., 2009; Mench

et al., 2009).

Roots also live in symbiosis with fungal mycorrhizae, and their mycelia are also covered with

bacteria. Fine feeder roots of Phragmites australis or Juncus effusus are also sites for

colonisation by arbuscular mycorrhizal fungi, whether they emerged from the roots of new

seedlings or subterranean rhizomes (Oliveira et al., 2001). This depends on soil moisture

content and plant phenology. For example, arbuscule presence peaked in P. australis during

the spring and autumn prior to flowering, but it peaked during the winter months for J.

48

effusus. Intraradical and extraradical mycelia of adapted fungal isolates are capable of

sequestering metals (Griffioen, 1994) and of promoting plant growth by increasing nutrient

absorption by the roots (Berta et al., 2002). Arbuscular mycorrhizal fungi can thus enhance

plant tolerance to environmental stresses, including metals (Leyval et al., 2002). High metal

concentrations in contaminated soils have also been reported to reduce, delay or even

eliminate fungal colonization (Lingua et al., 2008).

2.6.1 Micro-organism mediated oxidation

Macrophytes transport oxygen to their rhizosphere. This, coupled with the action of nitrifying

bacteria such as Nitrosomas spp. and Nitrobacter spp., enables ammoniacal N removal of the

soil environment (Lee and Scholz, 2007). Furthermore, oxidizing soil conditions promote

formation of iron oxides, hydroxides, and oxyhydroxides and consequently result in metal

removal by co-precipitation. However, maintaining aerobic conditions also promote the action

of bacteria such as Thiobacillus spp. These bacteria participate in oxidation of sulfides to

sulfites, and then to sulfates. Because sulfides take part in metal co-precipitation in the

substrate (Murray-Gulde et al., 2005b) oxidation may lead to metal mobilization.

2.6.2 Micro-organism mediated reduction

Under reducing conditions, sulphate-reducing bacteria such as Desulfovibrio spp. take part in

the reduction of sulphates to sulphites and subsequently to sulphides. Then these sulphides

react with metals such as Cu, Zn and Fe to form insoluble precipitates (Murray-Gulde et al.,

2005b). Optimal conditions for sulphate-reducing bacteria are redox potentials lower than -

100 mV and pH greater than 5.5 (Garcia et al., 2001). Precipitation of metal sulphides in an

organic substrate improves water quality by decreasing the mineral acidity without causing an

increase in proton acidity. Protons released by dissociation of H2S are neutralized by an equal

release of HCO3 during sulphate reduction (Sheoran and Sheoran, 2006).

49

Table 3 : Details of selected studies.

Study Authors Metal selected Location Type of CW (a) Siz e (b) Number Unit replication ( c)* Statistical tests C ontrol (d)

1 Kamal et al. 2004 Fe, Zn, Cu, Hg N. Amer. SF Meso 504 3 1 N N

2 Europe HSSF (and SF) Large ? 3 1 N

3 Peng et al. 2008 Cd, Pb, Mn, Zn, Cu Asia BR Micro 2 2 3 Y

4 Zhang et al. 2006 Cr, Pb, Cu, Cd, Mn, Fe Asia BR Micro 360 6 4 ANOVA + LSD test Y

5 Nelson et al. 2006 Hg, Cu, Pb, Zn N.Amer. SF Large 48 1 4 N N

6 Ha et al. 2009 Cb, As, Cu, Zn Asia BR Micro 24 1 3 N N

7 Lim et al. 2003 Zn, Pb, Cd Asia HSSF Meso ? 1 1 N N

8 Crites et al. 1997 N.Amer. SF Large ~120 2 10 (?) N Y

9 Maine et al. 2009 Cr, Ni, Fe,(Zn) S. Amer SF Large 168-288 3 1 N

10 Rai et al. 1995 Cu, Cr, Fe, Mn, Cd, Pb Asia BR Micro 48-360 8 3 N Y

11 Nyquist et al. 2009 Fe, Zn, Cu, Cd Europe SF Large ? 1 Y

12 Mishra et al. 2009 Zn, Cr Asia BR Micro 264 1 3 Regression Y

13 Megatelli et al. 2009 Cu, Cd, Zn North-Africa BR Micro 240 1 3 T-test Y

14 Mishra et al. 2008 Fe, Zn, Cu, Cr, Cd Asia BR Micro 360 3 1

15 Hadad et al. 2006 Cr, Zn, Ni S. Amer SF Large 168 1 N

16 Mantovi et al. 2006 Cu, Ni, Pb, Zn Europe HSSF Large 240 1 1 N N

17 Maine et al. 2007 Cr, Ni, Fe S.Amer. SF Large 168-288 1 N

18 Chagué-Goff. 2005 Fe, Cu, Pb, Zn Oceany SF Large 2400 4 N N

19 Khan et al. 2009 Cd, cr, Fe, Pb, Cu, Ni Asia SF Large 40 1 N

20 Dorman et al. 2009 Zn, Hg, Cr, As, Se N. Amer. SF Meso 120 2 N N

21 Cheng et al. 2002 Cd, Cu, Mn, Zn, Al, Pb Europe VF Meso ? 1 N N

Retention time (hours)

Samecka-Cymerman et al. 2003

Al, Ba, Mn, Ni, Sr, V, Zn, Cd, Cu, Pb, P, N, Cl,Ca,Mg, Fe, K

ANOVA (pseudo-replication)

ANOVA+Tukey post hoc test

Sr, As, Cd, Cr, Cu, Pb, Hg, Ni, Ag, Zn

ANOVA +Duncan's test (pseudo-réplication)

3 (mixed stand)

ANOVA + Tukey post hoc test (pseudo-

replication)

Linear regression (+ ANOVA ,

pseudoreplication)

Y (but data not shown)

11 (mixed stand)

T-test (pseudoreplication)

2 (mixed stand)

ANOVA (pseudoreplication)

7 (mixed stand)

11 (mixed stand)

ANOVA (pseudoreplication)

2 (dissociated mixed stand)

2 (dissociated mixed stand)

(a) : Type of CW. VSSF : vertical subsurface flow; HSSF : horizontal subsurface flow; SF : free water, surface flow; BR: Batch reactor

(b) : size of experimental units (surface area): Micro: microcosms (columns, buckets)<0.5 m²; Meso: mesocosms, from 0.51 to 5m², Large: pilot scale and full-size CW>5m² (from Chazarenc and Brisson, 2009)

( c) : number of units per species treatment. One means no replication

(d) : Control : presence of unplanted control (yes/no)

* : datas concern exclusively removal experiments (and no others experiments when there are others. e.g. Phytoaccumulation)N.Amer. : North AmericaS.Amer. : South America

50

III. The roles of plant species

Macrophytes are key players in wetlands, both natural and constructed. Studies comparing

planted and unplanted systems often lead to conflicting results regarding the importance of

plants (Lee and Scholz, 2007). A key point that is often overlooked is the supply of organic

matter by plants. This is partly due to the short time span out of most studies – too short for

the initial organic matter to be consumed in the chemical processes of removing metals from

water. Plant-derived organic matter in wetlands over time continuously provides sites for

metal sorption, as well as carbon sources for bacterial metabolism, thus promoting long-term

functioning (Beining and Otte, 1996, 1997; Jacob and Otte, 2003, 2004).

3.1 Emergent plants

Several emergent plants have been tested in constructed wetlands, achieving variable metal

removal rates (Table 3, 4, 5). Phragmites australis is the species used most. Phalaris

arundinacea displays capacities similar to Phragmites australis (Vymazal et al., 2007), as do

Typha domingensis (Maine et al., 2009), Typha latifolia, and Phragmites karka (Juwarker,

1995). Suspended organic matter and metals are efficiently removed in the presence of

plants(Lesage et al., 2007a; Scholz and Hedmark, 2010) mainly by immobilization in the

rhizosphere and storage in the belowground biomass (Baldantoni et al., 2009; Zhang et al.,

2009). The highest plant metal concentrations occur in the winter in rhizomes (Baldantoni et

al., 2009), but overall, less than 2% of the trapped metals are stored in the plant biomass (Lee

and Scholz, 2007). Therefore macrophytes are not important sinks for metal removal (Mays

and Edward, 2001; Lee and Scholz, 2007). Generally, only a small amount of metals taken up

by roots is transported to the shoots. Long-distance translocation of metal ions between roots

and shoots is summarized elsewhere (Lu et al., 2008). Poor translocation may be due to

sequestration of most of the metals in the vacuoles of root cells, which may be a natural

response to alleviate potentially toxic effects (Shanker et al., 2005).

51

Table 4 : Efficiency removal (%) in constructed wetlands with plants grown in monocultures.

Cu

30.8

9

44.9

2

42.5

8

W :

43

S

: 56

, 79.

4

W :

16

S :

36

W :

49

S :

60

74 65 98.5

98.4

98.2

98.7

99.1

*

97.9

97.6

80 52 -

55.0

6

-, -

, 86-

95

- 55

90,

76-

91

37 90 55 82 37 10 77

22-3

6

88-9

6

Rem

ova

l rat

e (%

)

Zn

34.7

7

32.6

3

34.4

2

W :

59

S

: 6

8, 8

5.7

W :

84

S

: 92

W :

38

S

: 73

66 67 - - - - - - - 47 50

84-9

9

81.0

9

-, 8

8-9

4, 8

5-95

,-

- -

-, 8

2-90

- - - - - -

100

10

82-9

2

Fe

92.9

2

63.6

8

75.6

5

W :

86

S :

79,

-

W :

82

S

: 9

7

W :

90

S

: 9

1

- -

99.7

*

98.7

98.7

99.1

*

99.6

*

98.5

97.7 - - - -

72,

-, 7

8.6-

90.1

, 79

73, 8

0

95

65, 7

7.5-

83.5

60 95 38 50 40 40 - -

87-9

5

Stu

dy *

1 1 1

2, 1

6

2 2 3 3 4 4 4 4 4 4 4 5 6 7 8

9, 1

2, 1

4, 1

7

9, 1

7

10

10,1

4

10 10 10 10 10 10 13 13 14

Cod

e

Men

Lud

Myr

Phr

Sav

Poc

Po

p

Pom

Acg

Aco

Acc

Lsa

Irp

Rca C Sca

Nsp

Tla

Sac

Ecr

Tdo Hre

Sp

o

Cco

Cd

e

Vsp

Bm

o

Ase

Har

Lgi

C Pst

Com

mon

nam

e

Aqu

atic

min

t

Cre

epin

g pr

imro

se

Par

rot

feat

her

Ree

d

Com

mon

osi

er

Car

olin

a po

plar

Sag

o po

ndw

eed

-

Dw

arf s

edge

?

Sw

eet

flag

Pur

ple

loos

estr

ife

Pal

eyel

low

iris

? -

Gia

nt b

ulru

sh

Nee

dle

spik

erus

h

Cat

tail

Bul

rush

Wat

er h

yaci

nth

Sou

ther

n ca

ttail

Wat

er n

et a

lga

Gia

nt d

uck

wee

d

Wat

er h

orse

tail

Hor

n ta

il

Cha

nnel

gra

ss

Wat

er h

ysso

p

Alli

gato

r w

eed

Wild

ric

e

Gib

bous

duc

kwee

d

-

Wat

er le

ttuce

Spe

cies

nam

e

Men

tha

aq

ua

tica

Lu

dw

igia

pa

lust

ris

Myr

iop

hyl

lum

aq

ua

ticu

m

Ph

rag

mite

s co

mm

un

is/a

ust

ralis

Sa

lix v

imin

alis

Po

pu

lus

can

ad

ensi

s

Po

tam

og

eto

n p

ectin

us

Po

tam

og

eto

n m

ala

ian

us

Aco

rus

gra

min

eus

Aco

rus

ori

enta

le

Aco

rus

cala

mu

s

Lyt

hru

m s

alic

ari

a

Iris

pse

ud

aco

rus

R(?

).ca

rnea

Con

trol

Sh

oen

op

lect

us

calif

orn

icu

s

Ele

och

ari

s a

cicu

lari

s**

Typ

ha

latif

olia

Sh

oen

op

lect

us

acu

tus

Eic

hh

orn

ia c

rass

ipes

Typ

ha

do

min

gen

sis

Hyd

rod

icty

on

ret

icu

latu

m

Sp

iro

del

a p

oly

rrh

iza

Ch

ara

co

ralli

na

Cer

ato

ph

yllu

m d

emer

sum

Va

llisn

eria

sp

ira

lis

Ba

cop

a m

on

nie

ri

Alte

rna

nth

era

ses

silis

Hyg

rorh

yzza

ari

sta

ta

Lem

na

gib

ba

Con

trol

Pis

tia s

tra

tiote

s

W :

Win

ter

S :

Su

mm

er**

our

own

calc

ulat

ion

52

Cr - - - -

-, 5

1.6

- - -

91.8

**

78.

1

93.3

**

81.3

*

95.8

*

79.

9

76.9 - - -

70.

6

66,

63-8

4, 8

1-8

9, 6

2

65,

58

90

80,

62-8

3

50 90 40 50 25 25 - -

70-8

1

Ba - - -

W :

70

S

: 9

5, -

W :

13

S

: 4

4

W :

59

S

: 6

2

- - - - - - - - - - - - - - - - - - - - - - - - - -

Rem

ova

l rat

e (%

)

Mn - - -

W :

99

S

: 9

9

W :

98

S

: 9

9, -

W :

99.6

S :

99.

5

89 83 99.3

99 99.3

99.6

99.6

98.7

99.1 - - - - - - 82 62 55 90 20 60 82 60 - - -

Ni - - -

W :

33

S

: 55

W :

30

S

: 46

, 58

.6

W :

55

S

: 67

- - - - - - - - - - - -

14.

6

48,

48

52,

48

- - - - - - - - - - -

Sr - - -

W :

24

S

: 5

1

W :

13

S :

43,

-

W :

17

S

: 2

3

- - - - - - - - - - - - - - - - - - - - - - - - - -

Stu

dy *

1 1 1

2, 1

6

2 2 3 3 4 4 4 4 4 4 4 5 6 7 8

9, 1

7,15

10

10,1

4

10 10 10 10 10 10 13 13 14

Cod

e

Men

Lu

d

Myr

Ph

r

Sav

Po

c

Pop

Pom

Acg

Aco

Acc

Lsa Irp

Rca C Sca

Nsp

Tla

Sac

Ecr

Td

o

Hre

Spo

Cco

Cd

e

Vsp

Bm

o

Ase

Har

Lgi C Pst

Com

mon

nam

e

Aq

uatic

min

t

Cre

epin

g p

rim

rose

Par

rot

feat

her

Re

ed

Co

mm

on o

sier

Car

olin

a p

opla

r

Sa

go p

ond

wee

d

-

Dw

arf s

edg

e

?

Sw

eet

flag

Pu

rple

loos

est

rife

Pal

eye

llow

iris

? -

Gia

nt b

ulr

ush

Nee

dle

spik

eru

sh

Cat

tail

Bul

rush

Wat

er h

yaci

nth

So

uthe

rn c

atta

il

Wat

er n

et a

lga

Gia

nt d

uck

wee

d

Wat

er h

orse

tail

Hor

n ta

il

Cha

nne

l gra

ss

Wat

er h

ysso

p

Alli

gato

r w

eed

Wild

ric

e

Gib

bou

s du

ckw

eed

-

Wat

er le

ttuce

Spe

cies

nam

e

Men

tha

aqu

atic

a

Lud

wig

ia p

alu

stris

Myr

iop

hyl

lum

aq

ua

ticu

m

Phra

gm

ites

com

mu

nis

/aust

ralis

Salix

vim

inalis

Po

pu

lus

can

aden

sis

Pota

moge

ton p

ectin

us

Pota

moget

on m

ala

ian

us

Aco

rus

gra

min

eus

Aco

rus

orie

nta

le

Aco

rus

cala

mu

s

Lyt

hru

m s

alic

aria

Iris

pse

ud

aco

rus

R(?

).ca

rnea

Con

trol

Sh

oen

ople

ctu

s ca

lifo

rnic

us

Ele

och

aris

aci

cula

ris**

Typ

ha la

tifolia

Sh

oen

ople

ctu

s a

cutu

s

Eic

hh

orn

ia c

rass

ipes

Typ

ha d

om

ingen

sis

Hyd

rod

icty

on

re

ticu

latu

m

Spiro

del

a p

oly

rrhiz

a

Chara

cora

llina

Ce

rato

ph

yllu

m d

eme

rsum

Valli

sner

ia s

pira

lis

Baco

pa

mo

nn

ieri

Alte

rnanth

era s

essi

lis

Hyg

rorh

yzza

aris

tata

Lem

na g

ibb

a

Con

trol

Pis

tia s

tratio

tes

W :

Win

ter

S :

Su

mm

er

**ou

r ow

n c

alcu

latio

n

9, 1

2, 1

4,

17,1

5

Table 4 (continued)

53

Sb - - - - - - - - - - - - - - - - 0.7 -

66.1 - - - - - - - - - - - - -

As - - - - - - - - - - - - - - - - 5 -

1.1

4

- - - - - - - - - - - - -

Pb - - -

W : 6

4 S

: 8

1, 69

.6

W : 4

4

S

: 7

2

W : 5

2

S

: 6

9

79

78

91

**

84.1

89.1

*

87*

93.4

**

82

77

70 -

99-8

9

87.2

6

- - 75

50

50

80

45

17

20

80 - - -

Rem

ova

l rate

(%

)

Cd - - -

W : 1

7 S

: 56, 2

3.7

W : 5

8

S

: 7

1

W : 6

6

S

: 5

5

96

88

95.2

**

92.9

*

90.5

*

92.2

*

96.1

**

95.2

**

83.3 - -

99-9

6

54.5

5

-, -

, 7

7-8

5, -

- 80

80

, 63-7

1

58

60

60

90

60

90

90 5

70-8

2

Hg

99.9

9

99.7

4

99.9

7

- - - - - - - - - - - - 80 - -

55.6

1

- - - - - - - - - - - - -

Al - - -

W : 8

1

S

: 9

7, -

W : 5

1

S

: 6

4

W : 3

3

S

: 4

7

- - - - - - - - - - - - - - - - - - - - - - - - - -

Stu

dy *

1 1 1

2, 16

2 2 3 3 4 4 4 4 4 4 4 5 6 7 8

9, 12

, 14,

17

9, 17

10

10,1

4

10

10

10

10

10

10

13 13 14

Cod

e

Men

Lud

Myr

Phr

Sav

Poc

Pop

Po

m

Acg

Aco

Acc

Lsa

Irp

Rca C Sca

Nsp

Tla

Sac

Ecr

Tdo

Hre

Spo

Cco

Cde

Vsp

Bm

o

Ase

Har

Lgi

C Pst

Com

mo

n na

me

Aquatic

min

t

Cre

epin

g p

rimro

se

Par

rot fe

ath

er

Reed

Com

mon o

sier

Caro

lina

po

pla

r

Sago p

ondw

eed

-

Dw

arf

sedge

?

Sw

eet

flag

Purp

le lo

ose

strif

e

Pal

eyel

low

iris

? -

Gia

nt b

ulru

sh

Nee

dle

spik

eru

sh

Cat

tail

Bulru

sh

Wat

er

hya

cinth

Sou

ther

n c

atta

il

Wat

er n

et a

lga

Gia

nt duck

weed

Wat

er

hors

eta

il

Horn

tai

l

Chan

nel

gra

ss

Wat

er h

ysso

p

Alli

gat

or

weed

Wild

ric

e

Gib

bous

duck

weed

-

Wate

r le

ttuce

Spe

cies

nam

e

Menth

a a

quatic

a

Ludw

igia

pa

lust

ris

Myr

ioph

yllu

m a

qu

atic

um

Phra

gm

ites

com

mun

is/a

ust

ralis

Salix

vim

inalis

Populu

s ca

nadensi

s

Pota

mogeto

n p

ect

inus

Pota

mo

get

on

mala

ianus

Aco

rus

gra

min

eus

Aco

rus

orien

tale

Aco

rus

cala

mus

Lyt

hru

m s

alic

aria

Iris

pse

udaco

rus

R(?

).ca

rnea

Con

tro

l

Shoeno

ple

ctus

calif

orn

icus

Ele

och

aris

aci

cula

ris*

*

Typ

ha la

tifolia

Sho

eno

ple

ctu

s acu

tus

Eic

hhorn

ia c

rass

ipes

Typ

ha

dom

ing

ensi

s

Hyd

rodic

tyo

n r

etic

ula

tum

Spiro

del

a p

oly

rrhiz

a

Cha

ra c

ora

llina

Cera

toph

yllu

m d

em

ers

um

Valli

sneria s

piralis

Ba

copa m

onnie

ri

Alte

rnanth

era

sess

ilis

Hyg

rorh

yzza

arist

ata

Lem

na

gib

ba

Con

tro

l

Pis

tia s

tratio

tes

W : W

inte

r

S : S

um

mer

**ou

r o

wn c

alcu

latio

n

Table 4 (continued)

54

However, other studies have shown that metals such as Cr are efficiently stored into the

whole plant (Cheng et al., 2002; Southichak et al., 2006; Zhang et al., 2010). Plants certainly

contribute to metal trapping into the substrate via rhizodeposition (Kidd et al., 2009) and act

as a catalyst for biochemical reactions involving organic acids. The main proportion of root-

exuded organic acids occurs in soil as anions such as citrate, oxalate, malate, malonate,

fumarate, and acetate (Ryan et al., 2001). These anions can chelate metallic ions to varying

degrees, and thus decrease their phytotoxicity. Tricarboxylate anions (e.g. citrate) chelate

metallic cations more strongly than dicarboxylate anions (e.g. oxalate), which in turn chelate

more strongly than monocarboxylate anions (e.g. acetate) (Chaudhry et al., 2005).

3.2 Floating plants

Wetlands with floating aquatic plants were discussed in detail by Vymazal et al. (1998) and

provide good metal absorption, for example with Eichhornia crassipes (Dhote et al., 2009),

Pistia stratiotes (Maine et al., 2001, Suné et al., 2007), and Salvinia herzogii (Maine et al.,

2004; Miretzki et al., 2005). In contrast to emergent rooting plants, floating plants do not

actively promote metal adsorption to the substrate, but store them into their biomass.

Eichhornia crassipes doubles its biomass in six days under favourable conditions (Mitchell,

1976), which, in conjunction with its strong absorption capacity, makes it a suitable species in

combination with surface flow systems (Zhu et al., 1999). Moreover, E. crassipes takes up

high amounts of P and N in the roots (Brix, 1993). This support micro-organisms that degrade

organic matter and release oxygen into the water (Dhote et al., 2009) and allows an important

P removal rate in a short time after harvesting (Maine et al., 2007 b).

3.3 Submerged plants

Although submerged aquatic plants such as Potamogeton spp, Ceratophyllum demersum,

Myriophyllum spicatum and Hydrilla verticillata have been recognized as species with a high

potential for water decontamination (Bunluesin et al., 2007; Dosnon-Olette et al., 2008), their

use is still at an experimental stage (Lesage et al., 2007). Batch studies have been

encouraging, but their usefulness in large scale constructed wetlands is uncertain, particularly

due to their low winter performance and the necessity for harvesting biomass in order to

maintain an efficient system (Kivaisi, 2001). Nyquist and Greger (2009) proposed to use

55

submerged plants to stabilize acid mine drainage because these plants take up more metals,

using their whole biomass, than just the roots of emerged macrophytes. On the other hand, the

role of plants as suppliers of organic matter is far more important, and so differences in metal

accumulation between the two types is minor. Submerged macrophytes are probably not

suitable for the conditions prevailing in acid mine drainage because of excessive Fe

precipitation onto their surfaces, which inhibits light penetration and photosynthesis (Nyquist

and Greger, 2009).

3.4 Adaptability of macrophytes to metal stress

Adaptability to metal stress is multigenic, as are the mechanisms for detoxification (Palmer

and Guerinot, 2009; Sharma and Dietz, 2009; Pal and Rai, 2010). Plants resist excessive

exposure through several routes. One mechanism is biomineralisation onto roots leading to

metal precipitation (Lanson et al., 2008). Another is formation of complexes with glutathione

(GSH) and transport into the vacuole (e.g. unidentified ATP-binding cassette transporter of

As(III)-GS3 or Cd(II)-GS2 (Verbruggen et al., 2009) where high molecular weight compounds

(HMWC) may be formed that contain sulphides (S2-). A third is production of organic ligands

rich in cysteine and non-protein thiols (NP-SH), such as phytochelatins (PC) and

metallothioneins (MT). Phytochelatins chelate metals and complexes to be transported into

the vacuoles (Pal and Rai, 2010). Metallothioneins and metallochaperones contribute to

maintain the homeostasis, bind metals, and protect against oxidative stress (Palmer and

Guerinot, 2009). A fourth mechanism is hyperactivity of antioxidant systems to minimize

reactive oxygen species (ROS) (Sharma and Dietz, 2009).

A small percentage of dryland plants has the innate capability to develop metal-tolerant

populations, perhaps one out of 100 species, and an even smaller percentage (perhaps one in a

thousand) shows constitutive tolerance (Verkleij et al., 2009; Memon and Schroeder, 2009;

Verbruggen et al., 2009). In contrast, numerous wetland plants have constitutive metal

tolerance and mechanisms of resistance (McCabe et al., 2001; Matthews et al., 2005; Deng et

al., 2006, 2009). Potamogeton pusillus responds positively to short Cu exposure by inducing

activity of antioxidant enzymes like glutathione peroxidase (GPX), glutathione reductase

(GR), and peroxidase (POD) (Monferran et al., 2009). In Najas indica exposed to Pb the

56

activities of antioxidant enzymes such as superoxide dismutase (SOD), ascorbate peroxidase

(APX), guaiacol peroxidase (GP), catalase (CAT) and glutathione reductase (GR) were

elevated along with the induction of the antioxidants GSH, cysteine, ascorbic acid and proline

(Singh et al., 2010). Iris pseudacorus also responds to Pb and Cd exposure by enhancing POD

activities and proline (PRO) concentrations in roots and shoots. These mechanisms enable

most cells in plants experiencing Pb toxicity to continue normal activities while sacrificing a

few cells that accumulate large amounts of Pb (Zhou et al., 2010). Exposure of Egeria densa

to Cd resulted in both a formation of thiol-enriched Cd-complexing peptides and a synthesis

of low molecular weight, unidentified metal chelators in shoots (Malec et al., 2009). Cd and

Cu exposure also lead to coordinated responses of PC synthase (PCS) and MT genes in black

mangrove Avicennia germinans. Both MT and PCS gene expression increased in A.

germinans leaves in response to metal exposure, which supports the hypothesis that MTs and

PCS are part of the metal tolerance mechanism in this species (González-Mendoza et al.,

2007).

IV. Importance of the plant ecotype

Intraspecific polymorphism can affect ecosystem functions and services. Ecotypes of Spartina

alterniflora grown in the same area remained morphologically closer to the ecotypes of their

original site than to other ecotypes transplanted in the same site, their development depending

more on genetic than on environmental variation (Seliskar et al., 2002). At some level,

ecosystem function and services may be determined not just by environmental factors, but

also by the genetic make-up of the species within that ecosystem, particularly for species such

as S. alterniflora and P. australis, which form natural monocultures. Genetic regulation of

ecosystem functioning and services occurs at several levels. It affects (1) biomass production,

because the amounts of organic matter released in the environment depend on ecotype

(Seliskar et al., 2002); (2) root structure and depth, thus affecting substrate oxygenation

(Armstrong, 1978); (3) production of root exudates implicated in altering the level of soluble

ions and molecules in the rhizosphere (e.g. organic acids, siderophores, saponines, amino-

acids and proteins); (4) provision of habitat for rhizosphere microbes and rhizosphere fungi;

and (5) the ability of plants to absorb and accumulate contaminants.

57

The sum of these inter-ecotype effects cannot be neglected in terms of phytoremediation of

polluted waters. For terrestrial plants, the choice of metallicolous or non-metallicolous

ecotypes may modify environmental conditions and influence metal removal efficiency, as

well as trophic transfer (Mench et al., 2009). For wetland macrophytes, the importance of

variation between ecotypes is unclear, mainly because of lack of knowledge on intra-specific

variation in relation to metal content in substrates. McCabe et al. (2001) and Otte et al. (2004)

suggested that metal tolerance is a constitutive feature of wetland plants, as populations of

Glyceria fluitans displayed tolerance to Zn regardless of the Zn levels in their respective

habitats. Constitutive Zn tolerance has been found also in E. angustifolium, T. latifolia and P.

australis (Matthews et al., 2005). Conversely, Plantago aquatica and Phalaris arundinacea

show less resistance to Zn (Matthews et al., 2005). Constitutive tolerance to metals, not just

Zn, appears to be the rule, rather than the exception, as this trait has also been observed by

McNaughton et al. (1974), Ye et al. (1997a, b, c, 1998) and Deng et al. (2004, 2006, 2009). It

appears therefore that differences between ecotypes in sensitivity to metals are less important

in wetland plants than in dryland plants. This means that as far as tolerance to metals is

concerned, wetland plants can be selected for use in constructed wetlands for removal of

metals from polluted water regardless of their origin. However, variation in sensitivity to

other environmental factors may be important when considering which plants to use. For

example, Phragmites australis shows inter-population variability to salinity (Guo et al., 2003)

and Kuhl and Zemlin (2000) found that P. australis stands differed persistently in

morphology and stand structure, depending on the environment and genotype. Such variation

between ecotypes has the potential to affect performance of constructed wetlands, but, to our

knowledge, no studies specifically addressing this issue have been published.

V. Selection of a data set for assessment of metal removal efficiency

Few papers except, to our knowledge, Brisson and Chazarenc (2009) emphasized the

importance of species selection in constructed wetlands for optimum performance. Here we

reviewed the effects of wetland plants on metal removal by addressing two questions: (1) are

the current experimental designs appropriate to determine constructed wetland efficiency, and

58

(2) what experimental options are available to assess the role of plants? We focused mainly on

constructed wetlands with emergent rooted macrophytes, but we also reported on results from

other types of systems. The removal by plants alone can be determined by measuring plant

growth and contaminant content stored in plant tissues (Debusk et al., 1995), but plant

accumulation is one mechanism of relatively minor importance to pollutant removal in

wetlands (Mander et al., 2003). Here we therefore focused on studies based on the following

criteria: (1) the study concerned metal removal, (2) differences between influent and effluent

concentrations were used to estimate removal efficiency, (3) plants were grown in mono- or

mixed cultures in identical units (from microcosm to large scale), with or without replication,

with or without unplanted units, and (4) each treatment unit received similar hydraulic loading

rates and contaminant concentrations. The second criterion, using differences in influent and

effluent concentrations for assessment of efficiency, assumes that the systems in the studies

reviewed here were hydrologically isolated and that evapotranspiration was a minor factor,

thus ensuring that changes in concentrations truly related to removal processes, not

dilution/concentration effects. These criteria resulted in 21 experimental studies, 19 of which

were published after 2000 (Table 3).

Studies with monocultures are listed in Table 4 while those with mixed stands are listed in

Table 5. Less than 40% of the studies included unplanted, control units. The total number of

plant species used was 54, of which 30 were used in monocultures and 24 in mixed stands.

The majority of macrophytes (except P. australis, E. crassipes, S. polyrrhiza, L. gibba, and S.

californicus) were tested in only one study.

59

As - - - -

96*

Cr - - - 89

64*

Ni - 69 - 41 -

Hg - - - -

90*

Cd 0 - -

91.9 -

Rem

oval r

ate

(%

)

Pb - - 50 -

Fe

0.4 83

74 -

Zn 8.5 55 -

80*

Cu 57 -

48.

3

-

Stu

dy

11 15 18 19 20

Cod

e

Ean +

Cro

+ P

au

Cbu +

nca

Com

mon

nam

eS

peci

es n

ame

W : W

inte

r

S : S

umm

er

*our

ow

n c

alc

ula

tion

W =

9.5

S

=

26.

7W

= -

48

S

= 2

0.6

W =

83.3

S

= 6

1.5

W =

82.3

S =

78.3

Pst

+E

cr+

She

+E

eb+

Cal

+P

el+

Tge+

Ppu

+P

co+P

ro+

Tdo

Cse

+ P

te +

Jpa

+

Ssp

+ B

sp +

Jef +

Jgr +

Lhy

Tla

+ S

cy +

Caq

+

Pau

+ J

ar +

Cde

+

Lgi +

Ecr

+ P

gl +

Apa

+ P

st

Cotton

gras

s + B

ottle

se

dge +

reed

Wat

ter le

ttuc

e+W

ater

hyac

ynth

+W

ater

fern

+S

ea

holly

+Yello

w

nuts

edg

e+

Ele

phan

t pa-

nicg

rass

+Arr

ow

root+

Sm

artw

eed+

Pic

kere

lweed+

Pic

kere

lweed+

cattai

l

Ora

nge tu

ssock

sedg

e +

bu

sh fl

ax +

Pal

e r

ush

+

Bul

rush

+ A

lkal

i bul

rush

+ C

om

mon

rush

+ N

Z

Com

mon

rush

+ G

rass

po

ly

Cat

tail

+ W

ool g

rass

+

Wat

er

sedg

e +

Reed

+

Join

ted

rush

+ c

oont

ail +

D

uck

weed

+ W

ater hy

a-cy

nth

+ K

notw

eed

+ W

a-te

r pl

anta

in +

Wat

er

weed

Cal

iforn

ia b

ulru

sh +

Nar

-ro

wle

af c

atta

il

Eriophoru

m a

ngust

ifoliu

m +

C

are

x ro

stra

ta +

Phra

gm

ites

aust

ralis

Pis

tia s

tratio

tes

+E

ichhorn

ia

crass

ipes

+ S

alv

ina h

erz

ogii

+ E

ryngiu

m e

burn

eum

+ C

y-peru

s alte

rnifo

lius

+ P

anic

um

ele

phantip

es

+ T

halia

genic

u-

lata

+ P

oly

gonum

punct

atu

m

+ P

onte

deria c

ord

ata

+ P

on-

tederia r

otu

ndifo

lia +

Typ

ha

dom

ingensi

s

Care

x se

cta +

Phorm

ium

te-

nax

+ J

uncu

s palli

dus

+ S

choenople

ctus

sp +

Bolb

o-

schoenus

sp +

Juncu

s effusu

s + J

uncu

s gre

gifl

oru

s +

(Ly-

thru

m h

ysso

pifo

lia)

Typ

ha la

tifolia

+ S

cirp

us

cy-

periniu

s +

Care

x aquatil

is +

P

hra

gm

ites

aust

ralis

+ J

uncu

s art

icula

tus

+ C

era

tophyl

lum

dem

ers

um

+ L

em

na g

ibba +

E

ichhorn

ia c

rass

ipes

+ P

oly

-gonum

gla

bru

m +

Alis

ma

pla

nta

go-a

quatic

a +

Pis

tia

stra

tiote

s

schoenople

ctus

calif

orn

icus

+ Typ

ha a

ngust

ifolia

Table 5 : Efficiency removal (%) in constructed wetlands with plants grown in mixed stands.

60

The majority of studies (47%) were carried out on a large scale, while 33% were performed in

microcosms and 20% in mesocosms. Using microcosm experiments is cheaper and facilitates

including a large number of species, as well as adequate replication. However, their results

must be interpreted with care due to container edge effects (Fraser and Keddy, 1997). Plants

in microcosms do not experience the effects of neighbouring plants on light interception and

growth. Root dispersion is often affected, with proportionally more roots crowded along the

inner surface of containers. Consequently, microcosms are useful in determining broad

patterns and investigating mechanisms, but, for application purposes, results must be

validated in large-scale systems (Brisson and Chazarenc, 2009). Mesocosms provide more

realistic conditions, but they are also not without some level of container effects, mainly

affecting the roots. Large-scale experiments deliver reliable results but their cost rarely makes

it possible to have (1) sufficient control units, (2) sufficient replication, and (3) a large number

of plants tested. Only 20% of the studies reviewed here included a control unit when

conducted at a large scale, and 40% of the overall studies did not have replication (Table 3).

In several studies, the statistical comparisons of removal efficiency across plant species were

based on repeated measures within the same units (pseudo-replication) rather than between

units with the same plant species. While significant differences between units solely based on

pseudo-replication are often interpreted as being caused by different plant species, this is not

as strong evidence as statistically significant differences based on replicated units (Brisson

and Chazarenc, 2009).

VI. Removal of metals

6.1 Differences between systems planted in monoculture

In monoculture, regardless of system size, Cu removal rates ranged between 10% for wetlands

vegetated with H. aristata and 99% with I. pseudacorus; Zn between 34% with M. aquaticum

and 100% with L. gibba; Fe between 38% with Vallisneria spiralis and >95% with

Hydrodictyon reticulatum, Pistia stratiotes, C. demersum and Salix viminalis; Pb between

17% with B. monnieri and 99% with T. latifolia; Cd between 17% with M. aquaticum and

99% with T. latifolia and Cr between 25% with H. aristata and A. sessilis and 90% with H.

reticulatum and C. demersum (Tables 3, 4). With P. australis, a removal rate of 99% for Mn

61

in the summer was obtained, 97% for Al, and 95% for Ba. In the winter, this rate remains at

99% for Mn, but was lower at 81 % for Al and 70% for Ba. With Salix viminalis a removal

rate of >98% of Mn was obtained both in the winter and in the summer, and up to 64% of Al

in summer. Removal rates with Populus canadensis reached 99% for Mn and >60% for Ba

throughout the year (Samecka-Cymerman et al., 2003). Mercury was efficiently removed

from the medium, with rates reported greater than 99%, in systems with monocultures of

Mentha aquatica, Ludwigia palustris and Myriophyllum aquaticum (Kamal et al., 2004). If

we define an efficient system as one that has a metal removal rate of greater than 70%, then

73% of the studies involving monocultures were efficient in removing Mn. 52% of the studies

reported efficient removal of Cu. For Zn, this value is 56%, 73% for Fe, 72% for Pb, 71% for

Cd and 67% for Cr (Table 4).

Monocultures of systems planted with monocots tended to show more efficient removal rates

than those planted with dicots for Zn and Cu as tested by a Wilcoxon test. The probabilities of

significance, P, were, for Cu 0.02 and Zn 0.07. Probabilities were clearly not significant for

Fe 0.49, Pb 0.2, Mn 0.38, Cd 0.36, and Cr 0.33. The trend for Zn and Cu was strong, even

though not clearly significant at the conventional α=0.05 level of significance for Zn (Figure

14). However, considering the wide variation in the systems this trend is quite convincing. It

is unclear why planting a wetland with monocots would lead to better removal of metals than

dicots. Because uptake in plants is not an important factor in metal removal in wetlands,

differences between monocots and dicots with respect to metal acquisition cannot account for

the differences in removal. However, the answer may lie in differences in morphology and

exudation of organic compounds. Monocot root systems are fibrous and/or adventitious and

numerous roots develop from the stem generating high root density.

This contrasts with dicots, which present a tap root system. Their primary root grows

vertically down into the substrate and lateral roots grow at an acute angle outwards and

downwards. Monocot root systems offer more surface area, thus more habitats for micro-

organisms than do dicots. Monocots also produce phytosiderophores (PS), for example

mugineic acids, which efficiently chelate metals such as ferric iron due to their amine and

carboxyl groups (Kidd et al., 2009).

62

Figure 14 : Metal removal depending on plant class (monocots and dicots). For each class, the box plot represents the 25th and 75th percentiles (with the median contained therein). In cases where there are sufficient data points, the whiskers represent the largest and smallest values which are not outliers and, where shown, the dots represent the largest and the smallest outliers. Boxes are drawn with widths proportional to the square-roots of the number of observations in the groups.

63

6.2 Planted versus unplanted systems

Zhang et al. (2007) reported differences in metal removal rates between planted and

unplanted controls (i.e. values for planted minus unplanted) of 0% for Cu and Fe, 5% for Cd,

and 2% for Pb and Cr (Table 3, 4). This suggests that at least for this combination of metals,

plants and experimental time spans the presence of plants did not enhance the metal removal

rates of the systems. This again emphasizes that the processes in the substrate involved in

binding of metals from the overlying water are more important than are the immediate plant-

related processes (such as root exudation and uptake), at least over relatively short periods of

time. The characteristics in terms of adsorption capacity and redox environment, particularly

when rich in organic matter make wetlands particularly suitable for removal of metals

(Scholz, 2003). However, without any plants, as organic matter is used up in microbially

mediated redox reactions and over time the substrate will become devoid of binding sites, thus

decreasing the capacity of the substrate to maintain metal immobilizing capacity (Jacob and

Otte, 2004).

6.3 Monoculture versus mixed stands

In systems with mixed stands, removal rates ranged between 48% and >80% for Cu, 9 and

>80% for Zn, <1% and 83% for Fe, <10% and 50% for Pb, 0% and 92% for Cd, and between

64 and 89% for Cr. These values do not differ from those for systems with monocultures,

suggesting there is no community effect on metal removal.

6.4 System size

In large-scale systems 40% have a Cu removal rate higher 70%. This proportion is 25% for

Cd, 30% for Cr, 100% for Fe, 66% for Zn, and 75% for Pb. Cu removal significantly depends

on wetland size (Kruskal test, p < 0.05), as does Cd removal (Kruskal test, p < 0.01). In

microcosms, root density per volume unit is greater than in large scale systems, and root

dispersion is affected by the edge and container effects (Fraser and Keddy, 1997). This might

64

Figure 15 : Metal removal depending on system size (microcosm, mesocosm, and large scale). For each class, the box plot represents the 25th and 75th percentiles (with the median contained therein). In cases where there are sufficient data points, the whiskers represent the largest and smallest values which are not outliers and, where shown, the dots represent the largest and the smallest outliers. Boxes are drawn with widths proportional to the square-roots of the number of observations in the groups.

65

lead to an overestimate of removal rates. In mesocosms, the ratio of ‘effluent concentration /

potential binding sites’ may be higher than at larger scales, leading to an underestimate of the

removal rate. The differences in removal rates between systems of different size were not

significant for Zn, Fe, Pb, Mn, or Cr but they generally follow the same trends as for Cu and

Cd (Figure 15). Thus, microcosms and mesocosms provide information about differences in

efficiency between plant species, but they do not really mimic what happens at larger scale s

in full-size applications (Brisson and Chazarenc, 2009).

6.5 Seasonal effects

Plant metabolism is more active in summer than in winter. This leads to (i) a higher evapo-

transpiration, which means less water-filled and more gas-filled pores and therefore a much

faster flux of gases, like oxygen, through substrates (Trapp and Karlson, 2001), and (ii) a

greater carbon resource provided by plants to micro-organisms. Both phenomena, coupled

with accelerated biochemical processes catalysed by higher temperatures in summer lead

potentially to better metal removal. In the studies reviewed here (Table 4), Cu removal was

30% higher in summer than in winter for a P. australis dominated system, 125% for a system

dominated by S. viminalis, and 22% for P. canadensis. Values for Zn removal are 18%, 14%,

and 92% higher in summer than in winter for the same species. For other elements values are:

Fe removal 15%, 18%, and 1%; Pb removal 26%, 86%, and 33%; Cd removal 229%, 22%,

and 17% (Samecka-Cymerman et al., 2003).

VII. A new index: the relative treatment efficiency index (RTEI)

We propose a Relative Treatment Efficiency Index (RTEI), based on the Relative Interaction

Index (RII) of Armas et al. (2004), to quantify the efficiency of constructed wetlands for

removal of pollutants. The efficiency of metal removal (e.g. with or without plants, with or

without organic matter, with or without gravel, and differences in retention times) is typically

derived from differences in metal removal between influent and effluent for the treatment

being considered (T, %) and compared with the control (C, %). Treatment, e.g. the presence

of macrophytes compared to unplanted controls, may increase metal removal, thus providing

66

benefits (benefits are defined as ∆ta, where 0% < ∆ta < 100%) but can also be a source of

disadvantages, e.g. by decreasing the retention time because of the displacement of the

volume by biomass (disadvantages are defined as ∆tb, where 0 < ∆tb < C). We assume that

both phenomena are independent and have additive and antagonistic effects on metal removal.

T = C + ∆ta - ∆tb (a)

Because it is not possible to separate the potential effect of treatment benefits and treatment

disadvantages in the observed, final removal rate of a system, we assign the observed, actual

value to ∆tab, so that :

∆tab = ∆ta - ∆tb (b)

This index is an interaction index, between benefits and disadvantages of the treatment, and

thus is relative and non-dimensional (Armas et al., 2004). We propose a relative treatment

efficiency index (RTEI) defined as follows:

���� =∆���∆��

�∆��� �∆��� � (c )

= ∆���

∆����� (d)

where the absolute value of the denominator is always greater than the absolute value of the

numerator and hence has a finite range. This index represents the net metal removal due to the

interaction of benefits and disadvantages of the treatment considered (numerator) relative to

the removal affected only by benefits and only by disadvantages of this treatment,

simultaneously. RTEI has values ranging from -1 to 1, is symmetrical around zero and is

negative for disadvantages and positive for benefits. Values approaching 1 indicate strong

benefits for metal removal, values around 0 indicate no effect of the treatment and values

approaching -1 indicate strong inhibition of metal removal (Figure 16). Taking into account

equations (a) and (c), RTEI can also be expressed as:

67

To illustrate the RTEI concept we used the data of Megatelli et al. (2006). In that study, L.

gibba cultivated with contaminated water provided high removal rates for Zn (100%), Cu

(77%) and Cd (90%), whereas controls without plants showed low removal (Zn 10%, Cu 22-

36%, and Cd 5%). Using the RTEI index, the treatment impact, the presence of plants, can be

visualized and quantified (Figure 16). In Zhang et al. (2009), removal rates were high for

both treatment and control. In this case, RTEI indicates the presence of plants did not greatly

enhance efficiency relative to the unplanted control (Figure 16). The RTEI index can be used

to compare different treatments within and among experiments using statistical comparative

methods with more complex statistical models, such as factorial analysis of variance.

a) b)

Figure 16 : Examples of utilization of the Relative Treatment Efficiency Index, RTEI a) data

from Megatelli et al. (2006) (Lemna gibba) b) data from Zhang et al. (2009) (Acorus

calamus).

Conclusion

Metals and metalloids are removed to varying degrees depending on the constructed wetland

type. This review confirms previous findings by Kropfelova et al. (2009) that the removal

rates of metals and metalloids in constructed wetlands decrease as follows:

Hg>Mn>Fe=Cd>Pb=Cr>Zn=Cu>Al>Ni>As. Generally, the removal rate is higher than 70%

Fe Cu Mn Cd Pb Cr

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

RTEIZn Cu Cd

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

RTEI

68

particularly in systems dominated by P. australis, P. canadensis, Potamogeton spp., Acorus

spp., L. salicaria, I. pseudacorus, Schoenoplectus spp., E.crassipes, H. reticulatum, C.

demersum, and P. stratiotes. These results may be biased by the dominance of certain species,

e.g. P. australis, not necessarily because they are the best choice, but because they are widely

available and easy to grow. Few, if any, systems have been planted with particular species

because of a deliberate choice regarding their efficacy.

The field of phytoremediation needs to set up standardized protocols for validation and better

quantify the roles of substrate, microorganisms, macrophytes, and their interactions in

constructed wetlands. In order to optimize designs for constructed wetlands for metal removal

one of the first stages should be the selection of the plant species and/or ecotype selection, in

conjunction with a substrate suitable for plant growth and a balanced supply of organic matter

and pH buffering capacity. Studies to underpin such selection can be carried out in a micro- or

mesocosm scale experiment. These would not aim at exactly mimicking natural ecosystems

but allow visualization of plant responses to the conditions prevailing in constructed wetlands,

such as metal concentrations, redox conditions and pH. Experiments aimed at testing metal

removal should use the following criteria: (1) measurement of system efficiency via the

assessment of a mass balance of water and pollutants (for example by measuring

concentration differences between inlet and outlet, assuring the system is hydrologically

isolated, and taking into account effects of evapo-transpiration), (2) each treatment unit must

be properly replicated, and (3) an unplanted control must be included. Concerning the roles of

the substrate, studies using so-called Bio-Racks, which are completely soil free, offer new

avenues for research (Valipour et al., 2009). Proper experimental design is often neglected in

phytoremediation studies, but is essential because it makes it possible to accurately quantify

the effects of the various experimental conditions, including the choice of plants and

substrates. Adoption of the relative treatment efficiency index (RTEI) proposed here would

achieve quantification and standardisation of these various conditions and their effects.

Acknowledgements

This work was supported by Axa foundation (PhD grant of L. Marchand) and ADEME,

Department Polluted Site and Soils, Angers, France.

69

References (Format : Environmental Pollution)

Aksoy, A., Demirezen, D., Duman, F., 2005. Bioaccumulation, detection and analyses of heavy metal pollution in Sultan Marsh and its environment. Water Air and Soil Pollution 164, 241-255.

Alloway, B.J., 1995. Heavy Metals in Soils, 2nd ed. Chapman and Hall, London. 368 p.

Arias, C., Brix, H., 2003. Humedales artificiales para el tratamiento de aguas residuales. Ciencia e Ingeneria Neogranadina 13, 17-44.

Armas, C., Ordiales, R., Pugnaire, F.I., 2004. Measuring plant interactions: A new comparative index. Ecology 85, 2682-2686.

Armstrong, W., 1978. Root aeration in the wetland condition, in: Hook, D.D., Crawford, R.M.M. (Eds.) Plant Life in Anaerobic Environments. Processes in Anaerobiosis. Ann Arbor Science, Ann Arbor, MI, pp. 269-298.

Baldantoni, D., Ligrone, R., Alfani, A. 2009. Macro- and trace-element concentrations in leaves and roots of Phragmites australis in a volcanic lake in Southern Italy. Journal of Geochemical Exploration 101, 166-174.

Beining, B.A., Otte, M.L., 1996. Retention of metals originating from an abandoned lead-zinc mine by a wetland at Glendalough, Co. Wicklow. Biology and Environment 96B, 117–126.

Beining, B.A., Otte, M.L., 1997. Retention of metals and longevity of a wetland receiving mine leachate, in: Proceedings of 1997 National Meeting of the American Society for Surface Mining and Reclamation, Austin, Texas, May 10-16, pp. 43-46.

Benjamin, M.M., Leckie, J.O., 1981. Conceptual-model for metal-ligand-surface interactions during adsorption. Environmental Science & Technology 15, 1050-1057.

Berta, G., Fusconi, A., Hooker, J.E., 2002. Arbuscular mycorrhizal modifications to plant root systems: scale, mechanisms, and consequences, in: Gianinazzi, S., Schüepp, H., Barea, J.M., Haselwandter, K. (Eds.), Mycorrhizal Technology in Agriculture. Birkhäuser Verlag, Basel, Boston, Berlin. pp. 71–85.

Bohn, H., McNeal, B., O’Connor, G., 1979. Soil Chemistry. John Wiley & Sons. New York, 329 p.

Brisson, J., Chazarenc, F., 2009. Maximizing pollutant removal in constructed wetlands: Should we pay more attention to macrophyte species selection? Science of the Total Environment 407, 3923-3930.

Brix, H., 1993 Wastewater treatment in constructed wetlands: system design, removal processes and treatment performance, in: Moshiri, G.A. (Ed.), Constructed Wetlands for Water Quality Improvement, CRC Press, Boca Raton, FA, pp. 9-22.

Brix, H., 1997. Do macrophytes play a role in constructed treatment wetlands? Water Science and Technology 35, 11-17.

Bunluesin, S., Kruatrachue, M., Pokethitiyook, P., Upatham, S., Lanza, G.R., 2007. Batch and continuous packed column studies of cadmium biosorption by Hydrilla verticillata biomass. Journal of Bioscience and Bioengineering 103, 509-513.

Cambrolle, J., Redondo-Gomez, S., Mateos-Naranjo, E., Figueroa, M.E., 2008. Comparison of the role of two Spartina species in terms of phytostabilization and bioaccumulation of metals in the estuarine sediment. Marine Pollution Bulletin 56, 2037-2042.

Chaudhry, Q., Blom-Zandstra, M., Gupta, S., Joner, E.J., 2005. Utilising the synergy between plants and rhizosphere microorganisms to enhance breakdown of organic pollutants in the environment. Environmental Science and Pollution Research 12, 34-48.

Chague-Goff, C., 2005. Assessing the removal efficiency of Zn, Cu, Fe and Pb in a treatment wetland using selective sequential extraction: a case study. Water Air, and Soil Pollution 160, 161–179.

70

Cheng, S.P., Grosse, W., Karrenbrock, F., Thoennessen, M., 2002. Efficiency of constructed wetlands in decontamination of water polluted by heavy metals. Ecological Engineering 18, 317-325.

Crites, R.W., Dombeck, G.D., Watson, R.C., Williams, C.R., 1997. Removal of metals and ammonia in constructed wetlands. Water Environment Research 69, 132-135.

Debusk, T.A., Peterson, J.E., Reddy, K.R., 1995. Use of aquatic and terrestrial plants for removing phosphorus from dairy wastewaters. Ecological Engineering 5, 371-390.

Demirezen, D., Aksoy, A., Uruc, K., 2007. Effect of population density on growth, biomass and nickel accumulation capacity of Lemna gibba (Lemnaceae). Chemosphere 66, 553-557.

Deng, H., Ye, Z., Wong, M., 2004. Accumulation of lead, copper and cadmium by 12 wetland plant species thriving in metal-contaminated sites in China. Environmental Pollution 132, 29-40.

Deng, H., Ye, Z., Wong, M., 2006. Lead and zinc accumulation and tolerance in populations of six wetland plants. Environmental Pollution 141, 69-80.

Deng, H., Ye, Z., Wong, M. 2009. Lead, zinc and iron (Fe(II) tolerances in wetland plants and relation to root anatomy and spatial pattern of ROL. Environmental and Experimental Botany 65, 353–362.

Dhote, S., Dixit, S., 2009. Water quality improvement through macrophytes - a review. Environmental Monitoring and Assessment 152, 149-153.

Dorman, L., Castle, J.W., Rodgers, J.H., 2009. Performance of a pilot-scale constructed wetland system for treating simulated ash basin water. Chemosphere 75, 939-947.

Dosnon-Olette, R., Couderchet, M., Eullaffroy, P., 2009. Phytoremediation of fungicides by aquatic macrophytes: Toxicity and removal rate. Ecotoxicology and Environmental Safety 72, 2096-2101.

Doty, S.L., 2008. Enhancing phytoremediation through the use of transgenics and endophytes. New Phytologist 179, 318-333.

Doyle, M.O., Otte, M.L., 1997. Organism-induced accumulation of iron, zinc and arsenic in wetland soils. Environmental Pollution 96, 1-11.

Emerson, D., Weiss, J.V., Megonigal, J.P., 1999. Iron-oxidizing bacteria are associated with ferric hydroxide precipitates (Fe-plaque) on the roots of wetland plants. Applied and Environmental Microbiology 65, 2758-2761.

Fraser, L.H., Keddy, P., 1997. The role of experimental microcosms in ecological research. Trends in Ecology & Evolution 12, 478-481.

Gambrell, R.P., 1994. Trace and toxic metals in wetlands – a review. Journal of Environmental Quality 23, 883-889.

Garcia, C., Moreno, D. A., Ballester, A., Blazquez, M. L., Gonzalez, F., 2001. Bioremediation of an industrial acid mine water by metal-tolerant sulphate-reducing bacteria. Minerals Engineering 14, 997-1008.

Gonzalez-Mendoza, D., Quiroz-Moreno, A., Zapata-Perez, O., 2007. Coordinated responses of phytochelatin synthase and metallothionein genes in black mangrove, Avicennia germinans, exposed to cadmium and copper. Aquatic toxicology 83, 306-314.

Goulet, R.R., Pick, F.R., 2001. The effects of cattails (Typha latifolia L.) on concentrations and partitioning of metals in surficial sediments of surface-flow constructed wetlands. Water Air and Soil Pollution 132, 275-291.

Griffioen, W.A.J., 1994. Characterization of a heavy metal-tolerant endomycorrhizal fungus from the surroundings of a zinc refinery. Mycorrhiza 4, 197-200.

71

Guan, Xi., Dong, H., Ma, J., Jiang, L., 2009. Removal of arsenic from water: effects of competing anions on As(III) removal in KMnO4-Fe(II) process. Water Research (Oxford) 43, 3891-3899.

Guo, W.H., Wang, R.Q., Zhou, S.L., Zhang, S.P., Zhang, Z.G., 2003. Genetic diversity and clonal structure of Phragmites australis in the Yellow River delta of China. Biochemical Systematics and Ecology 31, 1093-1109.

Ha, N.T.H., Sakakibara, M., Sano, S., Hori, R.S., Sera, K., 2009. The potential of Eleocharis acicularis for phytoremediation: Case study at an abandoned mine site. Clean-Soil Air Water 37, 203-208.

Hadad, H.R., Maine, M.A., Bonetto, C.A., 2006. Macrophyte growth in a pilot-scale constructed wetland for industrial wastewater treatment. Chemosphere 63, 1744-1753.

Hoagland, R.E., Williams, R.D., 1985. The influence of secondary plant compounds on the associations of soil microorganisms and plant roots, in: Thompson, A.C. (Ed.). The Chemistry of Allelopathy: Biochemical Interactions Among Plants. American Chemical Society, Washington, DC, pp. 301-325.

Holmstrom, H., 2000. Geochemical processes in sulphidic mine tailings: field and laboratory studies performed in northern Sweden at the Laver, Stekenjokk and Kristineberg mine sites. PhD dissertation 2000:03. Lulea University of Technology, Lulea, Sweden.

ITRC, 2003. Technical and Regulatory Guidance Document for Constructed Treatment Wetlands. The Interstate Technology & Regulatory Council Wetlands. 199 p. Available at http://www.itrcweb.org/Documents/WTLND-1.pdf . Accessed on April 22, 2010.

Jacob, D.L., Otte, M.L., 2003. Conflicting processes in the wetland plant rhizosphere: metal retention or mobilization? Water, Air, and Soil Pollution Focus 3, 91-104.

Jacob, D.L., Otte, M.L., 2004. Long-term effects of submergence and wetland vegetation on metals in a 90-year old abandoned Pb-Zn mine tailings pond. Environmental Pollution 130, 337-345.

Jenssen, P.D., Maehlum, T., Krogstad, T., 1993. Potential use of constructed wetlands for waste-water treatment in Northern environments. Water Science and Technology 28, 149-157.

Jonsson, J., Jonsson, J., Lovgren, L., 2006. Precipitation of secondary Fe(III) minerals from acid mine drainage. Applied Geochemistry 21, 437-445.

Juwarker, A.S., Oke, B., Patnaik, S.M., 1995. Domestic wastewater treatment through constructed wetland in India. Water Science Technology 32, 291-294.

Kadlec, R.H., Knight, R.L. 1996. Treatment wetlands. Lewis publisher, Boca Raton, FI, 893 p.

Kamal, M., Ghaly, A.E., Mahmoud, N., Cote, R., 2004. Phytoaccumulation of heavy metals by aquatic plants. Environment International 29, 1029-1039.

Kidd, P., Barcelo, J., Pilar Bernal, M., Navari-Izzo, F., Poschenrieder, C., Shilev, S., Clemente, R., Monterroso, C., 2009. Trace element behaviour at the root-soil interface: implications in phytoremediation. Environmental and Experimental Botany 67, 243-259.

Kivaisi, A.K., 2001. The potential for constructed wetlands for wastewater treatment and reuse in developing countries - a review. Ecological Engineering 16, 545-560.

Kropfelova, L., Vymazal, J., Svehla, J., Stichova, J., 2009. Removals of trace elements in three horizontal sub-surface flow constructed wetlands in the Czech-Republic. Environmental Pollution 157, 1186-1194.

Kuhl, H., Zemlin, R., 2000. Increasing the efficiency of reed plantations on stressed lake and river shores by using special clones of Phragmites australis. Wetlands Ecology and Management 8, 415-424.

Lanson, B., Marcus, M.A., Fakra, S., Panfili, F., Geoffroy, N., Manceau, A., 2008. Formation of Zn-Ca phyllomanganate nanoparticles in grass roots. Geochimica Cosmochimica Acta 72, 2478-2490.

72

Lee, B. H., Scholz, M., 2007. What is the role of Phragmites australis in experimental constructed wetland filters treating urban runoff? Ecological Engineering 29, 87-95.

Lemanceau, P., Bauer, P., Kraemer, S., Briat, J.F., 2009. Iron dynamics in the rhizosphere as a case study for analyzing interactions between soils, plants and microbes. Plant Soil 321, 513-535.

Lesage, E., Rousseau, D.P.L., Meers, E., Tack, F.M.G., De Pauw, N., 2007a. Accumulation of metals in a horizontal subsurface flow constructed wetland treating domestic wastewater in Flanders, Belgium. Science of the Total Environment 380, 102-115.

Lesage, E., Mundia, C., Rousseau, D.P.L., Van de Moortel, A.M.K., du Laing, G., Meers, E., Tack, F.M.G., De Pauw, N., Verloo, M.G., 2007b. Sorption of Co, Cu, Ni and Zn from industrial effluents by the submerged aquatic macrophyte Myriophyllum spicatum L. Ecological Engineering 30, 320-325.

Leyval, C., Joner, E.J., del Val, C., Haselwandter, K., 2002. Potential of arbuscular mycorrhizal fungi for bioremediation, in: Gianinazzi, S., Schuepp, H, Barea, J.M., Haselwandter, K., Mycorrhizal Technology in Agriculture. Birkhauser-Verlag, Basel, Switzerland, pp. 175-186.

Lim, P.E., Mak, K.Y., Mohamed, N., Noor, A.M., 2003. Removal and speciation of heavy metals along the treatment path of wastewater in subsurface-flow constructed wetlands. Water Science and Technology 48, 307-313.

Lin, L.Y., 1995. Wastewater treatment for inorganics. Encyclopedia of Environmental Biology, vol. 3. Academic Press, pp. 479–484.

Lingua, G., Franchin, C., Todeschini, V., Castiglione, S., Biondi, S., Burlando, B., Parravicini, V., Torrigiani, P., Berta, G., 2008. Arbuscular mycorrhizal fungi differentially affect the response to high zinc concentrations of two registered poplar clones. Environmental Pollution 153, 137-147.

Liu, W.J., Zhu, Y.G., Hu, Y., Williams, P.N., Gault, A.G., Meharg, A.A., Charnock, J.M., Smith, F.A., 2006. Arsenic sequestration in iron plaque, its accumulation and speciation in mature rice plants (Oryza sativa L.). Environmental Science & Technology 40, 5730–5736.

Lu, L.L., Tian, S. K., Yang, X.E., Li, T.Q., He, Z.L., 2009. Cadmium uptake and xylem loading are active processes in the hyperaccumulator Sedum alfredii. Journal of Plant Physiology 166, 579-587.

McCabe, O.M., Baldwin, J.L., Otte, M.L., 2001. Metal tolerance in wetland plants? Minerva Biotechnologica 13, 141–149.

Ma, L.Q. Komar, K.M., Tu, C., Zhang, W., Cai, Y., Kennelley, E.D., 2001. A fern that hyperaccumulates arsenic: a hardy versatile, fast growing plant helps to remove arsenic from contaminated soils. Nature 409, 579.

Maine, M.A., Duarte, M.V., Suné, N.L., 2001. Cadmium uptake by Pistia stratiotes. Water research 35 2629-2634.

Maine, M.A., Suné, N.L., Lagger, S.C., 2004. Chromium bioaccumulation: comparison of the capacity of two floating aquatic macrophytes. Water Research 38, 1494-1501.

Maine, M.A., Suné, N., Hadad, H., Sanchez, G., Bonetto, C., 2007. Removal efficiency of a constructed wetland for wastewater treatment according to vegetation dominance. Chemosphere 68, 1105-1113.

Maine, M.A., Suné, N., Hadad, H., Sanchez, G., 2007b. Temporal and spatial variation of phosphate distribution in the sediment of a free surface water constructed wetland. Science of the Total Environment 380, 75-83.

Malec, P., Maleva, M.G., Prasad, M.N.V., Strzalka, K., 2009. Identification and characterization of Cd-induced peptides in Egeria densa (water weed): Putative role in Cd detoxification. Aquatic Toxicology 95, 213-221.

73

Mander, U., Teiter, S., Kuusemets, V., Lohmus, K., Oovel, M., Nurk, K., Augustin, J., 2003. Nitrogen and phosphorus budgets in a subsurface flow wastewater treatment wetland Water Resources Management II, 135-148.

Manning, B.A., Fendorf, S.E., Goldberg, S., 1998. Surface structures and stability of arsenic(III) on goethite: Spectroscopic evidence for inner-sphere complexes. Environmental Science & Technology 32, 2383-2388.

Mantovi, P., Marmiroli, M., Maestri, E., Tagliavini, S., Piccinini, S., Marmiroli, N., 2003. Application of a horizontal subsurface flow constructed wetland on treatment of dairy parlor wastewater Bioresource Technology 88, 85-94.

Mastretta, C., Taghavi, S., van der Lelie, D., Mengoni, A., Galardi, F., Gonnelli, C., Barac, T., Boulet, J., Weyens, N., Vangronsveld, J., 2009. Endophytic bacteria from seeds of Nicotiana tabacum can reduce cadmium phytotoxicity. International Journal of Phytoremediation 11, 251-267.

Matagi, S.V., Swai, D., Mugabe, R., 1998. A review of heavy metal removal mechanisms in wetlands. Afr. J. Trop. Hydrobiol. Fish. 8, 23-35.

Matthews, D.J., Moran, B.M., Otte, M.L., 2005. Screening the wetland plant species Alisma plantago-aquatica, Carex rostrata, and Phalaris arundinacea for innate tolerance to zinc and comparison with Eriophorum angustifolium and Festuca rubra Merlin. Environmental Pollution 134, 343-351.

Mays, P. A., Edwards, G.S., 2001. Comparison of heavy metal accumulation in a natural wetland and constructed wetlands receiving acid mine drainage. Ecological Engineering 18, 251-252.

McNaughton, S.J., Folsom, T.C., Lee, T., Park, F., Price, C., Roeder, D., Schmitz, J., Stockwell, C. 1974. Heavy metal tolerance in Typha latifolia without the evolution of tolerant races. Ecology 55, 1163–1165.

Megateli, S., Semsari, S., Couderchet, M., 2009. Toxicity and removal of heavy metals (cadmium, copper, and zinc) by Lemna gibba. Ecotoxicology and Environmental Safety 72, 1774-1780.

Memon, A.R., Schröder, P., 2009. Implications of metal accumulation mechanisms to phytoremediation. Environmental Science and Pollution Research 16, 162-175.

Mench, M., Schwitzguébel, J.P., Schröder, P., Bert, V., Gawronski, S., Gupta, S., 2009. Assessment of successful experiments and limitations of phytotechnologies: contaminant uptake, detoxification and sequestration, and consequences for food safety. Environmental Science and Pollution Research 16, 876-900.

Meng, X.G., Bang, S., Korfiatis, G.P., 2000. Effects of silicate, sulfate, and carbonate on arsenic removal by ferric chloride. Water Research 34, 1255-1261.

Miretzky, P., Saralegui, A., Fernandez, A., 2004. Aquatic macrophytes potential for the simultaneous removal of heavy metals. Chemosphere 57, 997-1005.

Mishra, V.K., Tripathi, B.D., 2008. Concurrent removal and accumulation of heavy metals by the three aquatic macrophytes. Bioresource Technology 99, 7091-7097.

Mishra, V.K., Tripathi, B.D., 2009. Accumulation of chromium and zinc from aqueous solutions using water hyacinth (Eichhornia crassipes). Journal of Hazardous Materials 164, 1059-1063.

Mitchell, D.S., 1976. The growth and management of Eichhornia crassipes and Salvinia spp. in their native environment and in alien situations. in Varshney, C.K., Rzoska, J. (Eds.), Aquatic weeds in Southeast Asia. The Hague, The Netherlands, pp. 167–175.

Monferran, M.V., Sanchez Agudo, J.A., Pignata, M.L., Wunderlin, D.A., 2009. Copper-induced response of physiological parameters and antioxidant enzymes in the aquatic macrophyte Potamogeton pusillus. Environmental Pollution 157, 2570-2576.

Murray-Gulde, C.L., Huddleston, G.M., Garber, K.V., Rodgers, J.H., 2005a. Contributions of Schoenoplectus californicus in a constructed wetland system receiving copper contaminated wastewater. Water Air Soil Pollution 163, 355-378.

74

Murray-Gulde, C. L., Bearr, J., Rodgers, J.H., 2005b. Evaluation of a constructed wetland treatment system specifically designed to decrease bioavailable copper in a wastestream. Ecotoxicology and Environmental Safety 61, 60-73.

Nelson, E.A., Specht, W.L., Knox, A.S., 2006. Metal removal from water discharges by a constructed treatment wetland. Engineering in Life Sciences 6, 26-30.

Nyquist, J., Greger, M., 2009. A field study of constructed wetlands for preventing and treating acid mine drainage. Ecological Engineering 35, 630-642.

Oliveira, R. S., Dodd, J.C., Castro, P.M.L., 2001. The mycorrhizal status of Phragmites australis in several polluted soils and sediments of an industrialised region of Northern Portugal. Mycorrhiza 10, 241-247.

Pal, R., Rai, J.P.N., 2010. Phytochelatins: Peptides involved in heavy metal detoxification. Applied Biochemistry and Biotechnology 160, 945-963.

Palmer, C., Guerinot, M.L., 2009. Facing the challenges of Cu, Fe and Zn homeostasis in plants. Nature Chemical Biology 5.5, 333-340.

Peng, K. J., Luo, C.L., Lou, L.Q., Li, X.D., Shen, Z.G., 2008. Bioaccumulation of heavy metals by the aquatic plants Potamogeton pectinatus L. and Potarnogeton malaianus Miq. and their potential use for contamination indicators and in wastewater treatment. Science of the Total Environment 392, 22-29.

Rai, U.N., Sinha, S., Tripathi, R.D., Chandra, P., 1995. Waste-water treatability potential of some aquatic macrophytes - removal of heavy-metals. Ecological Engineering 5, 5-12.

Ryan, P.R., Delhaize, E., Jones, D.L., 2001. Function of mechanism of organic acid exudation from plant roots. Annual Review Plant Physiology Plant Molecular Biology 52, 527-560.

Samecka-Cymerman, A., Stepien, D., Kempers, A.J., 2004. Efficiency in removing pollutants by constructed wetland purification systems in Poland. Journal of Toxicology and Environmental Health-Part A-Current Issues 67, 265-275.

Scholz, M., Hedmark, A., 2010. Constructed wetlands treating runoff contaminated with nutrients Water Air and Soil Pollution 205, 323-332.

Schwarz, R., Gerth, A., Morgenstern, S., Hebner, A., 2009. Strategies for managing environmental problems and water treatment in mining, in: Proceedings of the International Mine Water Conference. 19-23 October, Pretoria, South Africa, pp. 557-566

Seliskar, D.M., Gallagher, J.L., Burdick, D.M., Mutz, L.A., 2002. The regulation of ecosystem functions by ecotypic variation in the dominant plant: a Spartina alterniflora salt-marsh case study. Journal of Ecology 90, 1-11.

Seo, D.C., Yu, K., DeLaune, R.D., 2008. Comparison of monometal and multimetal adsorption in Mississippi River alluvial wetland sediment: Batch and column experiments. Chemosphere 73, 1757-1764.

Shanker, A.K., Cervantes, C., Loza-Tavera, H., Avudainayagam, S., 2005. Chromium toxicity in plants. Environment International 31, 739-753.

Sharma, S.S., Dietz, K.J., 2009. The relationship between metal toxicity and cellular redox imbalance Trends in Plant Science 14.1, 43-50.

Sheoran, A.S., Sheoran, V., 2006. Heavy metal removal mechanism of acid mine drainage in wetlands - a critical review. Minerals Engineering 19, 105-116.

Singh, R., Tripathi, R.D., Dwivedi, S., Kumar, A., Trivedi, P. K., Chakrabarty, D., 2010. Lead bioaccumulation potential of an aquatic macrophyte Najas indica are related to antioxidant system. Bioresource Technology 101, 3025-3032.

75

Skousen, J.G., Ziemkiewicz, P.F., 1995. Acid Mine Drainage Control and Treatment. National Mine Land Reclamation Publication. 2nd edition. 254 p.

Southichak, B., Nakano, K., Nomura, M., Chiba, N., Nishimura, O., 2006. Phragmites australis: A novel biosorbent for the removal of heavy metals from aqueous solution. Water Research 40, 2295-2302.

Stoltz, E., Greger, M., 2005. Effects of different wetland plant species on fresh unweathered sulphidic mine tailings. Plant Soil 276, 251-261.

Stottmeister, U., Wiessner, A., Kuschk, P., Kappelmeyer, U., Kastner, M., Bederski, O., Muller, R.A., Moormann, H. 2003. Effects of plants and microorganisms in constructed wetlands for wastewater treatment. Biotechnology Advances 22, 93-117.

Suné, N., Sanchez., Caffaratti, S., Maine, M.A., 2007. Cadmium and Chromium removal kinetics by two aquatic macrophytes. Environmental Pollution 145, 467-473.

Sung, W., Morgan, J.J., 1980. Kinetics and product of ferrous iron oxygenation in aqueous systems Environmental Science & Technology 14, 561-568.

Trapp, S., Karlson, U., 2001. Aspects of phytoremediation of organic pollutants. Journal of Soils and Sediments 1, 37-43.

Valipour, A., Raman, V.K., Ghole, V.S., 2009. A new approach in wetland systems for domestic wastewater treatment using Phragmites spp. Ecological Engineering 35, 1797-1803.

Verbruggen, N., Hermans, C., Schat, H., 2009. Mechanisms to cope with arsenic or cadmium excess in plants. Current Opinion in Plant Biology 12, 364-372.

Verkleij, J.A.C., Golan-Goldhirsh, A., Antosiewisz, D.M., Schwitzguébel, J.-P., Schröder P., 2009. Dualities in plant tolerance to pollutants and their uptake and translocation to the upper plant parts. Environmental and Experimental Botany 67, 10-22.

Vymazal, J., Brix, H., Cooper, P.F., Green, M.B., Haberl, R., 1998. Constructed Wetlands for Wastewater Treatment in Europe. Backhuys Publishers, Leiden, The Netherlands. 366 p.

Vymazal, J., Svehla, J., Kropfelova, L., Chrastny, V., 2007. Trace metals in Phragmites australis and Phalaris arundinacea growing in constructed and natural wetlands. Science of the Total Environment 380, 154-162.

Weiss, J.V., Emerson, D., Backer, S.M., Megonigal, J.P., 2003. Enumeration of Fe(II)-oxidizing and Fe(III)-reducing bacteria in the root zone of wetland plants: Implications for a rhizosphere iron cycle. Biogeochemistry 64, 77-96.

Whitehead, P.G., Prior, H., 2005. Bioremediation of acid mine drainage: an introduction to the Wheal Jane wetlands project. Science of the Total Environment 338, 15-21.

Ye, Z.H., Baker, A.J.M., Wong, M.H., Willis, A.J., 1997a. Zinc, lead and cadmium tolerance, uptake and accumulation by Typha latifolia. New Phytologist 136, 469 - 480.

Ye, Z.H., Baker, A.J.M., Wong, M.H., Willis, A.J., 1997b. Copper and nickel uptake, accumulation and tolerance in Typha latifolia with and without iron plaque on the root surface. New Phytologist 136, 481-488.

Ye, Z.H., Baker, A.J.M., Wong, M.H., Willis, A.J., 1997c. Zinc, lead and cadmium tolerance, uptake and accumulation by the common reed, Phragmites australis (Cav.) Trin, ex Steudel. Annals of Botany 80, 363-370.

Ye, Z.H., Baker, A.J.M., Wong, M.H., Willis, A.J., 1998. Zinc, lead and cadmium accumulation and tolerance in Typha latifolia as affected by iron plaque on the root surface. Aquatic Botany 61, 55-67.

76

Zavoda, J., Cutright, T., Szpak, J., Fallon, E., 2001. Uptake, selectivity, and inhibition of hydroponic treatment of contaminants. Journal of Environmental Engineering 127, 502-508.

Zhang, C.B., Wang, J., Liu, W.L., Zhu, S.X., Liu, D., Chang, S.X., Chang, J., Ge, Y., 2010. Effects of plant diversity on nutrient retention and enzyme activities in a full-scale constructed wetland. Bioresource Technology 101, 1686-1692.

Zhang, X.P., Liu, P., Yang, Y.S., Chen, W.R., 2007. Phytoremediation of urban wastewater by model wetlands with ornamental hydrophytes. Journal of Environmental Sciences 19, 902-909.

Zhao, S.J., Zhang, Z.C., Gao, X., Tohsun, G., Qiu, B.S., 2009. Plant regeneration of the mining ecotype Sedum alfredii and cadmium hyperaccumulation in regenerated plants. Plant Cell, Tissue and Organ Culture 99, 9-16.

Zhou, Y.Q., Huang, S.Z., Yu, S.L., Gu, J.G., Zhao, J.Z., Han, Y.L., Fu, J.J., 2010. The physiological response and sub-cellular localization of lead and cadmium in Iris pseudacorus L. Ecotoxicology 19, 69-76.

Zhu, Y.L., Zayed, A.M., Qian, J.H., de Souza, M., Terry, N., 1999. Phytoaccumulation of trace elements by wetland plants: II. Water hyacinth. Journal of Environmental Quality 28, 339-344.

77

Chapitre I

De la nécessité de traiter les sols contaminés au Cu afin d’en limiter le

transfert vers les horizons aquifères…Le cas de la phytostabilisation aidée :

vers une nouvelle perspective de zones humides construites

Cette partie a été publiée sous la forme d’un article dans Science of the Total Environment

(410-411. pp 146-153) en 2011

78

Figure 17 : Vue aérienne de la plateforme de phytoremédiation BIOGECO Aerial view of the BIOGECO phytoremediation platform

Figure 18 : Lysimètres placés sur la plateforme de phytoremédiation BIOGECO

Outdoor lysimeters located at the BIOGECO phytoremediation platform

79

Science of the Total Environment, 410-411. (146-153)

Phytotoxicity testing of lysimeter leachates from aided phytostabilized Cu-contaminated soils using duckweed

(Lemna minor L.)

L. Marchand1, M. Mench1, Charlotte Marchand1, Philippe Le Coustumer2, Aliaksandr Kolbas1, Jean-Paul Maalouf1

1 UMR BIOGECO INRA 1202, Écologie des Communautés, Université Bordeaux 1, Bât. B2

RDC Est, Avenue des facultés, 33405 Talence, France

2 EA4592 Géoressources & Environnement, ENSEGID, Université de Bordeaux 1, 1 allée F. Daguin, F-33607 Pessac, France

Abstract

Aided phytostabilization of a Cu-contaminated soil was conducted at a wood preservation site located

in south- west France using outdoor lysimeters to study leaching from the root zone and leachate

ecotoxicity. The effects of Cu-tolerant plants (Agrostis gigantea L. and Populus trichocarpa x

deltoides cv. Beaupré) and four amendments were investigated with seven treatments: untreated

soil without plants (UNT) and with plants (PHYTO), and planted soils amended with compost (OM,

5% per air-dried soil weight), dolomitic limestone (DL, 0.2%), Linz–Donawitz slag (LDS, 1%), OM

with DL (OMDL), and OM with 2% of zerovalent iron grit (OMZ). Total Cu concentrations (mg kg−

1) in lysimeter topsoil and subsoil were 1110 and 111–153, respectively. Lysimeter leachates

collected in year 3 were characterized for Al, B, Ca, Cu, Fe, Mg, Mn, P, K and Zn concentrations, free

Cu ions, and pH. Total Cu concentration in leachates (mg L− 1) ranged from 0.15 ± 0.08 (LDS) to

1.95 ± 0.47 (PHYTO). Plants grown without soil amendment did not reduce total Cu and free Cu ions

in leachates. Lemna minor L. was used to assess the leachate phytotoxicity, and based on its growth,

the DL, LDS, OM and OMDL leachates were less phytotoxic than the OMZ, PHYTO and UNT ones.

The LDS leachates had the lowest Cu, Cu2+, Fe, and Zn concentrations, but L. minor developed less in

these leachates than in a mineral water and a river fresh- water. Leachate Mg concentrations were in

decreasing order OMDL > DL > PHYTO = OM = LDS > UNT = OMZ and influenced the duckweed

growth.

Keywords : Compost, Dolomitic limestone, In situ stabilization, Linz–Donawitz slag,

Phytoremediation, Zerovalent iron grit

80

Capsule: a single incorporation of Linz-Donawitz slag into a Cu-contaminated soil planted

with two Cu-tolerant plants, Populus trichocarpa x deltoides cv Beaupré and Agrostis

gigantea, decreased Cu concentration in lysimeter leachates from the root zone and their

phytotoxicity for duckweed.

Abbreviations

LDS Linz-Donawitz slag, DL dolomitic limestone, OM compost, OMDL compost + 0.2%

w/w dolomitic limestone, Z zerovalent iron grit, OMZ compost + 2% w/w zerovalent iron

grit, PTTE potentially toxic trace elements, PHYTO untreated soil planted with Agrostis

gigantea and Populus trichocarpa x deltoides cv Beaupré, UNT untreated soil without plants,

CCA Chromated Copper Arsenate, DW dry weight, NOEC No observed effect concentration,

EC10 10% effective concentration, EC50 50% effective concentration.

81

I. Introduction

The soils at many current and former wood preservation sites are Cu-contaminated soils as

many Cu-based wood preservatives were used to control insects and fungi (Mills et al., 2006;

Mench and Bes, 2009; Karjalainen et al., 2009). Excessive Cu concentrations in topsoils

usually affect plant communities and performances (Bes et al., 2010; Verdugo et al., 2011). In

addition, due to runoff and leaching, Cu in dissolved and solid forms may migrate to aquatic

ecosystems where it accumulates in sediments and living organisms (Ma et al., 2003;

Kanoun-Boulé et al., 2009; Karjalainen et al., 2009). Such accumulation may cause

physiological and biochemical changes in macrophytes, animals and microbes (Megateli et

al., 2009; Cvjetko et al., 2010). Copper toxicity is mainly due to the existence of two readily

interconvertible oxidation states that making Cu highly reactive and a catalyst of the

formation of free radicals through the Haber-Weiss reaction. Free Cu ions can initiate

oxidative breakdown of polyunsaturated lipids (Kanoun-Boulé et al., 2009). Dispersion of

inorganic soil contaminants can be quenched by the in situ stabilization technique and/or the

restoration of a vegetation cover. Stabilization henceforth refers to the physico-chemical

stabilization of potentially toxic trace elements (PTTE) caused by addition of soil conditioners

(Mench et al., 2003; Kumpiene et al., 2008). Incorporation of amendments, e.g. lime, OM,

basic slag, activated carbon, and zerovalent iron grit (Z), into metal and metalloid-

contaminated topsoils induces changes in the physico-chemical state and/or chemical

speciation of metals such as Cu (Mench et al., 2003, 2006; Kumpiene et al., 2006; 2011; Bes

and Mench, 2008). Formation of insoluble, sorbed, or bound chemical species of metals such

as Cu may reduce their leaching from the root zone and their labile pool available for

biological action (Ruttens et al., 2006; Bes and Mench, 2008; Lagomarsino et al., 2010).

Increasing root uptake and storage in root system is another key step in removing metals from

the soil solution and keeping them in the rhizosphere. Phytostabilization uses tolerant plants

with excluder phenotypes and associated microbes for long-term containment of contaminants

such as PTTE in solid matrices. This works through mechanical and (bio)chemical

stabilization that either prevents or minimizes pollutant linkages such as PTTE transfer in the

food chain, leaching from the root zone and downward percolation to groundwater, migration

to aquatic systems, and re-entrainment of contaminated particulates for direct inhalation or

ingestion (Mench et al., 2010). In the case of aided phytostabilization, single or combined

amendments are incorporated into the soil to decrease the labile contaminant pool and

phytotoxicity by inducing various sorption and/or precipitation processes prior to planting

82

tolerant excluder plants (Adriano et al., 2004; Mench et al, 2010; Lizama et al., 2011a, b).

The vegetation cover minimizes wind dispersion of inorganic contaminants and limits their

leaching from the root zone through evapotranspiration and root uptake. Plant roots prevent

water erosion and store a fraction of the labile metal (Cu) pool during their lifespan.

Conversely, root channels may create pathways for enhanced leaching whereas decayed

products and dissolved organic matter from organic amendments such as compost may

change metal (Cu) speciation and mobility (Mench et al., 2003; Ruttens et al., 2006;

Marchand et al., 2010). Consequently, the composition and ecotoxicity of leachates from the

root zone of aided phytostabilized Cu-contaminated soils remain unclear.

The context of this study is the appraisal of aided phytostabilization at a former wood

preservation site through successive steps. Soil ecotoxicity was first assessed and stabilizing

amendments were selected in pot experiments (Bes and Mench, 2008; Mench and Bes, 2009;

Negim et al., 2009). In parallel, plants were assessed for their Cu-tolerance and potential

usefulness for Cu phytostabilization (Aulen et al., 2007; Bes et al., 2007; Mench et al., 2008;

Bes et al., 2010). Field plots and outdoor lysimeters were then established to assess several

aided phytostabilization options, notably using grassy and woody species (Bes et al., 2007;

Mench et al., 2009; Mench and Bes, 2009).

The aim of this study was to assess the phytotoxicity of lysimeter leachates from the root zone

of a Cu-contaminated soil with three options: (1) bare soil, (2) phytostabilization: untreated

soil planted with two non-native Cu-tolerant plant species, i.e. Agrostis gigantea L. and

Populus trichocarpa x deltoides cv. Beaupré, and (3) aided phytostabilization: incorporation

of one amendment into the soil, i.e. Linz-Donawitz slag (LDS), dolomitic limestone without

(DL) and with compost (OMDL), compost (OM), compost with zero-valent ion grit (OMZ),

followed by the transplantation of the two plants mentioned above.

Phytotoxicity of lysimeter leachates was assessed using Lemna minor L. (duckweed) as a

bioindicator of water quality (EPA, 1996). The test was also carried out on samples of a

mineral water and freshwater from an uncontaminated river (Jalle d'Eysine river, Gironde,

France) for the purpose of comparison. The questions asked were (1) What remediation

83

options minimize PTTE concentrations in lysimeter leachates? (2) Are the lysimeter leachates

phytotoxic, notably compared to a mineral water and a river freshwater?

II. Materials and Methods

2.1 Soils and amendments

The wood preservation site (6 ha) is located in southwest France (44°43’N; 0°30’O) and has

been used for over a century to preserve and store timbers, posts and utility poles (Figure 17)

(Mench and Bes, 2009). The industrial facility dates back to 1846. Creosote, Cu sulphate

(from 1913 to 1980), CCA (from 1980 to 2006), and Cu hydroxycarbonates with

benzylalkonium chlorides (since 2006) were successively used (Mench and Bes, 2009).

Established vegetation and site characteristics were previously assessed (Mench and Bes,

2009; Bes et al., 2010). Anthropogenic soils developed on an alluvial soil (Fluviosol). Soil

investigation pits (0-1.5 m) revealed major contamination of topsoils by Cu with spatial

variation (65 to 2400 mg Cu kg-1 soil DW), whereas total soil As and Cr, i.e. 10-53 mg As and

20-87 mg Cr kg-1 in topsoils, remained relatively low in all soil layers. In February 2007, a

sandy soil was collected with a steel spade in a trench at the P3 sub-site (for site details, see

Mench and Bes, 2009). Soil samples made of six independent sub-samples were taken from

the soil layers, air-dried, and sieved at 2 mm prior to analysis (Table 6). All soil analyses

were performed at the INRA Laboratoire d’Analyses des Sols (LAS, Arras, France) using

standard methods (INRA LAS, 2007).

In March 2007, large vats (75 dm3, 0.5 m diameter) were filled with three successive layers: 5

cm of coarse gravels (1-3 cm, diameter), 22 cm of sub-soil (from the 30-60 cm soil layer), and

25 cm of topsoil (from the 0-30 cm soil layer) (Table 6). Gravels and the sub-soil were

separated by a geotextile. Total Cu concentration (mg kg-1) was 1110 in the topsoil and 111-

153 in the subsoil (Table 6). Amendments were carefully mixed with the topsoil using a vat,

singly and in combination (% air-dried soil DW, w/w), before filling the lysimeters, to form

the seven soil treatments that were conducted in triplicate: bare untreated soil (UNT),

untreated soil with plants (PHYTO), 5% compost made of wood chips and poultry manure

84

(OM, Orisol, Cestas, France), 0.2% dolomitic limestone (DL, Bes and Mench, 2008), OM

with 2% zerovalent iron grit (Z, GH120, particle size <0.1 mm, Wheelbrator, Allevard,

France) (OMZ) and 1% P-spiked Linz-Donawitz slag (LDS, Centre Technique et de

Promotion des Laitiers Sidérurgiques, La Plaine Saint-Denis, France). LDS is mainly

composed of Ca (30.7 wt % CaO), Fe (21.4 wt % Fe2O3), Si (14.6 wt % SiO2), P (14.0 wt %

P2O5), Mg (9.5 wt % MgO), Al (5.5 wt % Al2O3) and Mn (2.5 wt % MnO2). Other

compounds such as TiO2 and K2O were detected at low concentrations, 1.09 and 0.53 wt %,

respectively (Negim et al., 2009). Agrostis gigantea L. (2 patches, 5 cm in diameter) and one

poplar (Populus trichocarpa x deltoides cv. Beaupré, initial shoot length: 30±5 cm) were

transplanted in all lysimeters except for the UNT treatment. Lysimeters (n=21) were placed in

situ (March, 2007) (Figure 18). Lysimeter leachates were periodically collected in plastic

bottles (1.5 dm3) from March 5, 2007 on after each major precipitation event (>30 mm,

leachate volume > 1.5 L). They were collected in year 3 (March, 2010) for this experiment,

and kept at 4°C for no more than 48 h prior to the test. Both leachate and soil pH (1:1

soil:water suspension, Jackson, 1967) were determined (Hanna instruments, pH 210,

combined electrode Ag/AgCl – 34).

85

Table 6 : Main characteristics of the soil layers used to fill the lysimeters and main soil characteristics at the P3 sub-site (0-25 cm soil layer).

a Background values in French sandy soils are median and upper whisker values, except b Frequent total As concentrations for all French soil types; c Threshold values are concentrations above which a negative impact can be observed on plants and animals (Baize, 1997). pCu2+ =3.20 +1.47pH -1.84 log10(total soil Cu) (Sauvé, 2003)

0-15 cm 15-30 cm 30-60 cm 60-100 cm P3 (0-25 cm)

Sand % - - - - 85.8 - -

- - - - 8.3 - -

- - - - 5.9 - -Organic carbon (g C/kg) 13.6 4.4 3.13 1.57 - - -

23.5 7.62 5.42 2.71 16 - -Total nitrogen (g N/kg) 0.78 0.44 0.36 0.21 - - -

C/N 17.3 9.95 8.63 7.5 17.2 - -EC (dS/m) 0.13 0.44 0.39 0.29 - - -

CEC cmol/kg 3.36 1.39 1.17 1.09 3.5 - -pH 7.05 4.69 4.04 4.09 7 - -

As (mg/kg) 15.5 3.53 4.71 6.11 9.8 -- - - - 0.12 0.03-0.24 -

1.66 1.87 2.28 2.67 <2 1.4-6.8 -1110 772 153 111 1460 3.2-8.4 35

Cr (mg/kg) 34.1 16.3 18.4 22.5 23 14.1-40.2 1006300 6800 7900 8700 6090 6000-14300 -172 238 185 107 181 72-376 -

Ni (mg/kg) 5.22 5.35 7.97 10.6 5 4.2-14.5 50- - - - 27 16.4-58.7 -

Tl (mg/kg) - - - - 0.24 0.29 -54.9 30.5 28.7 29.3 46 17-48 1507.96 4.78 5.12 5.45 7.66 - -

background values in French sandy soils a Threshold values c

Silt %

Clay %

Organic matter (g/kg)

1.0-25 b

Cd (mg/kg)Co (mg/kg)Cu (mg/kg)

Fe (mg/kg)Mn (mg/kg)

Pb (mg/kg)

Zn (mg/kg)pCu2+

86

2.2 Plant material and toxicity test

The Lemna minor L. population used was collected in the Jalle d'Eysines River (Bordeaux,

Gironde, southwest France) in April 2010. Plants were washed with sterile distilled water,

placed in a plastic container (10 dm3) filled with modified Hoagland medium (Hewitt, 1966)

(pH = 4.9±0.04), and grown for six weeks in a greenhouse at 22 ±2 °C with a 12 h

photoperiod before the tests. Container solution was replaced once a week toprovide nutrients

and oxygen for the Lemna fronds and to prevent the development of root fungal diseases

(Kamal et al., 2004). Fronds are defined here as single Lemna “leaf-like” structures (EPA,

1996).

In the second step, Lemna minor fronds (2 colonies with 3 fronds and 2 colonies with

4 fronds = 14 fronds in all), randomly selected from the storage container, were cultivated in

triplicate for 12 days in 250 mL Erlenmeyer flasks containing 150 mL of one of the following

solutions (adapted from EPA, 1996): lysimeter leachates (triplicates corresponding to the

three lysimeter replicates, for all treatments described above), the freshwater from the Jalle

d'Eysine river and Evian mineral water (France). An additional experiment was carried out in

the same way, but 30 mL of Hoagland solution were added to 120 mL of each treatment

solution. Leachates and waters were analyzed at the INRA Usrave laboratory, Villenave

d’Ornon, France (ICP-AES, Varian Liberty 200). The free Cu ion concentration in leachates

and waters was measured with an mV-meter (Hanna instruments, pH 210) and a selective ion

electrode (Fischer Bioblock Cupric Ion Electrode, N83921) (Table 7).

Fronds were counted every three days, as the number of fronds has been used as a relevant

surrogate for biomass (Radic et al., 2010), to assess the phytotoxicity of the growth media.

Relative growth rate (RGR) was then determined after 3, 6, 9 and 12 days of experimentation

using the following equation: RGR = [ln(final frond number)–ln(initial frond number)]/days,

the initial number of fronds being n=14 at the beginning of the experiment.

87

Table 7 : Elemental composition (mg L-1), pH, and EC (µS cm-1) of leachates in year 3 and of fresh- and mineral waters.

Values are means ± SD (n=3), letters indicate significant differences between treatments for each parameter at p < 0.05 (post hoc Tukey HSD test). Stars (*) indicate values

below the detection limit. aSalpeteur et al. (2006).

pH B Mg P K EC

PHYTO 6.58±0.27b 31.70±5.13c 0.02±0.01ab 4.82±1.63b 2.53±0.47 b 19.00±2.86c 1.85±0.33ab 0.42±0.07a 0.94±0.16bc 3.77±0. 67ab 0.10±0.01c 0.44±0.06b 70.7±34.9bcUNT 6.55±0.09b 20.47±3.33c 0.02±0.0a 4.70±1.33b 1.40±0.16ab 12.10±1.97b 1.21±0.17a 0.32±0.02a 0.60±0.11c 3.07±0.34a b 0.07±0.02b 0.33±0.1b 45.2±4.6cOM 6.77±0.28b 24.97±3.07c 0.03±0.0b 5.15±0.57b 1.56±0.12b 1 3.37±2.39bc 1.85±0.2ab 0.24±0.06a 0.76±0.29bc 8.51±1.75 c 0.06±0.01b 0.32±0.17b 89.5±0.7bDL 6.80±0.5b 10.85±3.26b 0.02±0.0a 8.78±4.39b 0.82±0.28ab 5 .96±1.53ab 2.66±0.66bc 0.11±0.04a 0.28±0.08ac 1.69±0.22 ab 0.03±0.01a 0.19±0.1b 92.4±20.1b

OMZ 6.76±0.15b 2.70±0.72ab 0.02±0.0a 4.13±1.76b 0.25±0.13ab 1.64±0.46a 0.80±0.29a 0.12±0.04a 0.20±0.0ac 10.73±2.98c 0.02±0.01a 0.14±0.09b 79.9±7.2bOMDL 6.73±0.25b 4.06±4.58ab 0.02±0.01ab 10.25±2.21b 0.40±0.1 5ab 3.19±4.2a 3.14±0.21cd 0.49±0.69a 0.23±0.05ac 7.23±4. 77bc 0.02±0.01a 0.023±0.017ab 112.5±19.2abLDS 7.62±0.1a 0.05±0.0a 0.02±0.0a 31.80±4.1a 0.15±0.08a 0.02 ±0.01a 1.91±0.2ab 0.55±0.57a 0.20±0.0a 1.01±0.77a 0.01±0 .0a 0.012±0.002a 158.6±11.6a

6.90±0.48b < 0.05* 0.032±0.01b 29.47±7.09a < 0.008* 0.12±0.09a 4.55±0.23d < 0.02* < 0.2* 4.24±1.32ab < 0.007* 0.006±0.002a 380±80d7.2 < 0.05* - 78 < 0.008* - 24 < 0.02* < 0.2* 1 < 0.007* - 571

8 6.6 µg/L 24.1 µg/L 100 mg/L 0.88 µg/L 24.1 µg/L 4.6 mg/L 12.8µg/L - 2.3 mg/L 2.27 µg/L - -

Treatments Al Ca Cu Fe Mn Zn Cu 2+

Jalle freshwaterMineral water

Median values in surface water for French rivers a

88

2.3 Statistical analyses

The effects of treatments on the leachate element concentrations, leachate pH and the RGR of

L. minor were tested using one-way ANOVAs. Normality and homoscedasticity of residuals

were met for all tests. Post hoc Tukey HSD tests were performed to assess multi-comparison

of means. Differences were considered statistically significant at p<0.05. A Principal

Component Analysis (PCA) was conducted on leachate composition and pH. Pearson

correlation coefficients were calculated between growth (RGR) and each of the trace- and

macro-element concentrations in leachates and leachate pH. All statistical analyzes were

performed using R software (version 2.12.0, R Foundation for Statistical Computing, Vienna,

Austria).

III. Results and Discussion

3.1 Composition of lysimeter leachates

Trace element concentrations and leachate pH depended on soil treatments, notably on the

amendments used to stabilize the contaminated soil (Table 7). The first axis (53.5%) of the

PCA corresponds to the PTTE concentrations and the second axis (14.5%) corresponds to Ca

and Mg concentrations and pH (Figure 19). Based on the PCA, leachates can be divided into

four groups. Group 1 (PHYTO, UNT and OM) included leachates with the highest P and

PTTE concentrations but with low Ca concentrations. Group 2 (OMDL and DL) included

leachates with lower PTTE concentrations, low Ca concentration, and high Mg concentration.

Group 3 (OMZ) had low PTTE concentrations, the highest K concentration, and the lowest Ca

and Mg concentrations. Group 4 (LDS) had low PTTE concentrations, the highest pH and Ca

concentrations, but the lowest P and K concentrations. The leachate Mg concentration

differed significantly across treatments (Table 7). Its value (in mg L−1) peaked in the OMDL

treatment (3.1), was the lowest in the OMZ treatment (0.8), and varied between 1.2 (UNT)

and 2.6 (DL) in the other treatments. This agrees with the results of Kumpiene et al. (2008)

who reported that soil treatments with zerovalent iron, e.g. OMZ, immobilized macro-

elements such asMg, whereas the use of dolomitic limestone ensured adequate Mg

concentrations (Riggs et al., 1995). However, leachate Mg concentrations in the amended

soils were all below the Jalle freshwater value (4.55 mg L−1, p<0.05).

89

(a)

(b) (c)

pH

Al B

Ca

total.Cu Fe

Mg Mn

P

K

Zn

Free.Cu.ion

Dimension 1 (53.47%)

Factor Correlation P value

Al 0.98 <0.0001

Fe 0.97 <0.0001

Zn 0.93 <0.0001

Total Cu 0.92 <0.0001

P 0.92 <0.0001Free Cu ion 0.86 <0.0001

B 0.49 <0.0001

pH -0.68 <0.0001

Ca -0.67 <0.0001

Dimension 2 (14.52%)

Factor Correlation P value

Ca 0.63 0.002

pH 0.51 0.01

Mg 0.45 0.04

K -0.69 <0.001

Dim 1 (53.5%)

Dim 2 (14.5%)

d = 2

DL

LDS

OM OMDL

OMZ

PHYTO

UNT

Figure 19 : Principal Component Analysis (PCA) of the treatments accounting for Al, B, Ca, total Cu, free Cu ion, Fe, Mg, Mn, P, K, and Zn concentrations and pH in lysimeter leachates: (a) PCA, (b) variable contributions to axes, and (c) correlation circle.

90

3.1.1 UNT

Leachates from the UNT soil (Table 7) presented high Cu and Zn concentrations compared to

the upper threshold values used by the French water agency (Adour-Garonne region) for

superficial freshwaters (respectively 10-15 and 43-98 µg L-1). Total Cu, Cu2+, Fe, Al, and Zn

concentrations would render UNT leachates toxic for the aquatic ecosystems (Karjalainen et

al., 2009). Total Cu concentration in UNT leachates largely exceeded frequent values reported

for European waters (in µg Cu L-1: minimum 0.08, median 0.88, maximum 14.6) (INERIS,

2010), the NOEC/CE10 value for algae in freshwater (0.01 mg Cu L-1, INERIS, 2011), the

chronic ecotoxicity parameters for L. minor, i.e. 14-day NOEC value (0.06 mg Cu L-1, Jenner

and Janssen-Mommen, 1993), the EC10 value (0.057 mg Cu L-1 based on frond number,

Naumann et al., 2007), and all the EC50 values cited in the literature for Lemna sp. – except

Lakatos et al. (1993) and Ince et al. (1999) (Table 7). Leachate Zn concentration in UNT was

four times higher than the EC10 value for L. minor (0.017 mg Zn L-1 based on frond number,

Naumann et al., 2007).

3.1.2 LDS

The lowest total Al, Cu, Fe and Zn concentrations occurred in the LDS leachates (Table 7),

which led to their individualization from the other leachates (Figure 19). As an alkaline by-

product of steel-making, LDS mainly consists of Al, Ca, Fe, Mg, Mn and Si oxides, but may

also contain relatively high Cr, Mn and V concentrations (Negim et al., 2009; Sjöberg et al.,

2010). It provides a high neutralizing capacity at low cost, which likely induces metal

hydrolysis reactions and/or co-precipitation with carbonates and acts as a precipitating agent

for metals in the solution (Bes and Mench, 2008). Al, Cu, Mn, Ni and Zn are co-precipitated

in Fe (hydr)oxides and Co, Fe, Ni and Zn are co-precipitated in Mn (hydr)oxides, dissolved

from the slag, whose overall mean surface charge switches from positive to negative value as

pH increases (Sheoran and Sheoran, 2006; Marchand et al., 2010). Here, metals co-

precipitated with Fe and Mn (hydr)oxides, and LDS leachates presented the highest pH across

the treatments (pH ranged from 6.5 for UNT to 7.6 for LDS, Tukey HSD test, p<0.05). Bes

and Mench (2008) reported an increase in pH from 6.25 (untreated soil) to 8.1 in a

contaminated soil from the same site amended with 0.1% calcium oxide. High pH and Ca

concentrations decreased Cu mobility in the soil its lowest value being at slightly alkaline pH

91

(Figure 19) (Kumpiene et al., 2008). Across treatments, the lowest total Cu and Cu2+

concentrations (0.15 and 0.012 mg L-1, respectively) occurred in LDS leachates (Table 7).

However, these concentrations remained higher than concentrations measured in the Jalle

freshwater (total Cu<0.008 mg L-1 and Cu2+ = 0.006 mg L-1). Total Cu in LDS leachates

exceeded the EC10 value for L. minor (Table 7). The LDS treatment was thus the most

efficient in reducing PTTE concentrations in lysimeter leachates, but did not decrease them to

the levels of an uncontaminated freshwater. The LDS leachates had the lowest K

concentration (1.0 mg L-1, Tukey post hoc test, p<0.05) and reduced P leaching (Table 7)

which may indicate an increase in phosphate sorption. Application of slag can decrease

exchangeable soil K content due to a reduction in the percentage of K saturation in the

exchange complex (Negim et al., 2009). Release of LDS-born vanadium and its leaching from

the root zone may constitute a potential environmental threat that needs to be further

investigated (Sjöberg et al., 2010).

3.1.3 DL

Dolomitic limestone has been widely used to reduce PTTE mobility, mainly Cu and Pb, by

raising soil pH. In this study, it increased pH to 6.8 and Mg concentration in leachates

(Table 7). Leachate Ca concentration increased less in DL than in LDS. The DL treatment

reduced PTTE mobility (e.g. Cu total = 0.82 mg L-1 in DL leachates) but it was less efficient

than the LDS treatment. As initial soil pH (6.5) was nearly neutral, liming may be not

sufficient to reduce PTTE mobility as it does in an acid soil, and an alkaline amendment

containing Fe/Mn oxides such as LDS was shown to be more appropriate. As for LDS, DL

incorporation led to a low leachate K concentration (Table 7).

3.1.4 OM, OMDL, OMZ

Incorporation of OM into the soil, singly or combined with another amendment, can be used

to reduce PTTE leaching (Ruttens et al., 2006). The Al, total Cu, Cu2+, Fe and Zn

concentrations were higher in the OM leachates than in the OMDL and OMZ ones (Table 7).

Except for Cu2+, which was lower in OMDL than in OMZ (respectively 0.023 and

0.14 mg L-1), PTTE concentrations in OMDL and OMZ leachates were similar (Table 7).

Insoluble high molecular weight organic acids can retain Cu in soil upon soil acidification

92

(Chirenje and Ma, 1999) and the OM-bound Cu fraction can account for 96% of the total Cu

in a CCA-contaminated soil (Balasoiu et al., 2001). However, the OM treatment did not

reduce Cu and Zn leaching from the root zone in the soil tested here. Dissolved OM may

partly promote Cu mobility (Kumpiene et al., 2008). The combination of OM with zerovalent

iron grit (OMZ) was more efficient than OM alone, especially for Cu stabilization (Figure

20). Newly formed Fe and Mn oxides after Z corroded in the soil likely enhance metal

sorption by the poorly crystalline Fe oxyhydroxides and crystalline Fe-Mn oxides of the OMZ

soil (Bes and Mench, 2008; Kumpiene et al., 2011). The OMDL treatment numerically

reduced Cu leaching more than the DL and OM (Figure 20, Table 7) confirming previous

findings with acid and contaminated soils (Mench et al., 2000; Tlustos et al., 2006). The

OMDL and OMZ treatments were less efficient than LDS (Figures 19, 20) in reducing metal

concentrations in lysimeter leachates.

Figure 20 : Total copper and free copper ion (Cu2+) concentrations in lysimeter leachates. Values are means ± SD (n=3); different letters stand for statistical significance at the 0.05 level with the Tukey HSD test.

3.1.5 PHYTO

The presence of A. gigantea and poplar may reduce metal concentrations in leachates by

sorbing metals onto the root plaque and root uptake (Doyle and Otte, 1997; Marchand et al.,

ab

ab ab

ab

a

b

b

α αβ β

β

β

β β

2+

DL LDS OM OMDL OMZ PHYTO UNT

93

2010) and increasing soil CEC (Vangronsveld et al., 1996). Conversely, through root

exudation of organic acids (Ryan et al., 2001) and grass phytosiderophores (Kidd et al., 2009)

metal chelation may maintain metal solubility. Indeed, Fe and Zn concentrations were

significantly higher in PHYTO leachates (Table 7, Tukey HSD test, p<0.05), and a similar

numerical trend was observed for total Cu, Cu 2+ and Al (Figure 19, Table 7). Root-induced

porosity and rhizodeposition, notably phytosiderophores from A. gigantea roots, may promote

leaching of these metals.

3.2 Phytotoxicity and L. minor growth

Duckweed is a ubiquitous floating freshwater monocotyledon and one of the world’s smallest

flowering plants. It is widely used in water quality tests to monitor contaminants such as

PTTE because of its physiological traits, i.e. small size, rapid growth between pH 5 - 9 with a

doubling time of 1-4 days or less, and vegetative reproduction (Horvat et al., 2007; Kanoun-

Boulé et al., 2009). An excess of PTTE, such as Cu, inhibits duckweed growth (Kanoun-

Boulé et al., 2009; Megateli et al., 2009). On day 3 of this experiment, the RGR of plants

growing in DL, LDS, OM, OMDL, OMZ, PHYTO, and UNT leachates were in the same

range (from -0.03 in UNT and DL to 0.02 in PHYTO and OMZ). These RGR values were

significantly lower (Tukey HSD test, p<0.05) than those of plants growing in the Jalle and

mineral waters (respectively 0.19 and 0.20, Figure 21). The same effect was found at days 6

and 9. On day 6, RGR of plants growing in lysimeter leachates ranged from -0.02 (OMDL) to

0.03 (OMZ), whereas RGR of plants growing in the Jalle and mineral waters reached 0.1. On

day 9, RGR of plants growing on lysimeter leachates ranged from 0-0.01 (UNT, OMDL, OM)

to 0.03 (LDS), while RGR of plants growing on the Jalle and mineral waters increased to 0.08

and 0.09. On day 12, RGR of plants growing in the Jalle and mineral waters reached 0.09 and

0.12, respectively, whereas the other treatments split in two groups (Tukey HSD, p<0.05). In

the first group, i.e. OM, DL, OMDL and LDS, duckweed growth remained stable or increased

slightly in the leachates (RGR values 0, 0.02, 0, and 0.02, respectively). In the second group,

i.e. PHYTO, UNT and OMZ, duckweed growth was inhibited (leaf discoloration, and RGR

respectively = -0.05, -0.1 and -0.04). Changes in plant growth were significantly negatively

correlated with total Cu, Cu2+, Al, Fe, and Zn concentrations in the solutions tested (Table 8).

Excessive Cu exposure, notably free Cu ions, induces oxidative stress in plants leading to

peroxidation of polyunsaturated lipids (Mocquot et al., 1996; Kanoun-Boulé et al., 2009). The

94

free radicals produced damage the photosynthetic apparatus (Kanoun-Boulé et al., 2009) and

may catalyze protein degradation through oxidative modification and increased proteolytic

activity (Romero-Puertas et al., 2002). Copper accumulation in plant tissues can also cause

changes in nitrogen metabolism and increase free amino acid concentrations (Llorens et al.,

2000; Megateli et al., 2009).

Figure 21: Relative growth rate (RGR) of Lemna minor across treatments over the 12-day period (without addition of Hoagland solution).Values are means ± SD (n=3), different lower case letters stand for statistical significance at the 0.05 level with the Tukey HSD test.

Table 8 : Correlations based on Pearson coefficients between the Relative Growth Rate (RGR) of Lemna minor (leachates without Hoagland solution) and the chemical parameters of leachates (total Al, B, Ca, Cu, Fe, Mg, Mn, P, K, and Zn concentrations, free Cu ion concentration, and pH) across the treatments (n=21)

*statistical significance at the 0.05 level and ** at the 0.01 level; Cu tot: total Cu concentration; Cu2+: free Cu ion concentration

3 6 9 12

b b b b b b b b b b b b b b

a a

b b b b b b b

a a

b b b b c c c

a a

a a

Day

pH Al B Ca Cu tot Fe Mg Mn P K Zn

RGR 0.31** -0.24* 0 0.37** -0.17** -0.30** -0.28* 0.41** 0 0.21* 0 -0.38**

Cu 2+

95

Previous studies on Lemna sp. exposed to Cu showed a wide range of EC50, i.e. from 2.52 µM

(0.16 mg L-1) to 23.6 µM (1.5 mg L-1) (Table 7). Babu et al. (2003) reported that duckweed

growth decreased above 4 µM Cu (0.25 mg L-1). Our results (Figure 19, 21, Table 7) agree

with the low EC50 values. However, the OMZ leachate, which had the lowest Cu

concentration after LDS (Cu = 3.94 µM, 0.25 mg L-1), inhibited duckweed growth at day 12

(Figure 21). Al, Fe, and Zn concentrations in the leachate were relatively low in this

treatment (respectively 2.70, 1.64, and 0.02 mg L-1). Therefore additional factors may be

involved. On one hand, Z may increase Ni exposure (Mench et al., 2000), while on the other

hand, several interactions between Mg and Ca homeostasis and metal exposure may drive

duckweed growth. Magnesium limited duckweed growth in the leachates tested (Table 8,

Pearson coefficients: 0.41). In plant cells, Mg2+ is vital for membrane stabilization, ATP

utilization, and nucleic acid biochemistry. It is a cofactor for many enzymes (including

ribosome enzymes) and the coordinating ion in the chlorophyll molecule (Schaul, 2002). With

excessive metal accumulation, Mg2+ ions associated with the tetrapyrrole ring of chlorophyll

molecules can be replaced (Radic et al., 2010), especially by Cu2+ ions (Kupper et al., 2002).

In addition, in plants exposed to PTTE, protein levels are often lower due to Mg and K

deficiencies, which cause loss of protein synthesis (Hou et al., 2007; Cvjetko et al., 2010). In

Mg2+ deficient plants, other secondary effects include carbohydrate immobility and loss of

RNA transcription. In particular, in our study, Mg could have been immobilized in the OMZ

soil, decreasing Mg concentration in the OMZ leachates (Table 7) and duckweed growth.

Conversely, duckweed growth was higher in the Jalle and mineral waters, richer in Mg, than

in leachates. In the additional experiment, Mg was added with the Hoagland solution to the

OMZ leachates and their negative effect was completely neutralized, OMZ being the most

efficient treatment at day 6 and day 12 (Table 9). High Al concentration in leachates,

especially in PHYTO and OM, may impact duckweed growth. Al3+ is a strong inhibitor of

Mg2+ uptake (Rengel, 1990). In our data, Cu2+/Mg ratio in lysimeter leachates from the root

zone was a relevant indicator (y = -251.51 x +21.14, R²=0.59, p<0.05) of duckweed growth

(RGR).

Growth of L. minor was slightly correlated with Ca concentration (Table 8). Calcium

increases soil and leachate pH and hence playing a role in metal immobilization, but it is also

implicated in cell wall development and cell membrane integrity of duckweeds (Huebert et

al., 1991), Ca sub-cellular homeostasis and the detoxification of oxidative stress (Cuin, 2006).

96

Calcium can influence Cu uptake and rhizotoxicity (Wang et al., 2010). This makes this

macro element, which was more present in the DL, OMDL, and LDS leachates, essential to

enhance L. minor growth in leachates from contaminated soils. Duckweed growth depends on

interactions between trace elements and between PTTE and macro-elements. This cannot be

explained by simple antagonism and/or synergism but has to be considered as a whole

(Cvjetko et al., 2010; Upadhyay and Panda, 2010).

Table 9 : Relative growth rate (RGR) of Lemna minor on lysimeter leachates, with addition of Hoagland solution, and on freshwater and mineral waters over the 12-day period.

Values are means (n=3), different lower case letters stand for statistical significance at the 0.05 level with Tukey HSD test.

Conclusion

Composition of lysimeter leachates from the root zone of a Cu-contaminated soil, collected in

year 3 after the incorporation of soil conditioners, depended on the amendments used for in

situ metal stabilization and, to a lesser extent, on the presence of Cu tolerant plants. The

growth of Agrostis gigantea L. and Populus trichocarpa x deltoides cv. Beaupré – Cu tolerant

plants - did not reduce total Cu and free Cu ions in leachates compared to the bare soil.

Conversely, aided phytostabilization based on the incorporation of Linz-Donawitz slag (LDS,

1% w/w) led to the lowest leachate Cu, Cu2+, Fe, and Zn concentrations. At similar soil pH

(~ 6.7), compost with zerovalent iron grit (OMZ), and dolomitic limestone with (OMDL) and

without compost (DL) numerically reduced total Cu and free Cu ions in leachates, but

differences were not significant compared to the unamended soil with (PHYTO) and without

(UNT) plants. The Mg concentration was lower in the OMZ leachate than in the other

RGR0.186 a 0.101 a 0.083 a 0.089 a0.197 a 0.099 a 0.092 a 0.123 a

PHYTO 0.009 b 0.045 b -0.052 c -0.072 cUNT -0.056 c 0.037 b -0.088 c -0.109 cOM -0.027 b 0.040 b 0.025 b 0.030 bDL 0.002 b 0.065 ab 0.024 b 0.021 b

OMZ 0.008 b 0.068 ab 0.025 b 0.050 bOMDL -0.036 b 0.043 b 0.001 b 0.017 bLDS -0.019 b 0.049 b 0.031 b 0.043 b

Hoagland solution ModalityJalle River

Mineral Water

Leachates withHoagland solution

97

leachates, and this negatively affected duckweed growth. Further investigation of the

enrichment of the OMZ treatment with Mg2+ would be necessary to enhance its efficacy. The

duckweed growth in leachates in increasing order was UNT, PHYTO, OMZ < OM, OMDL,

DL, LDS. However, duckweed developed less in the LDS leachates than in a mineral water

and a river freshwater.

Acknowledgements

This work was financially supported by AXA foundation (PhD grant of L. Marchand),

ADEME, Department Polluted Site and Soils, Angers, France, and the European Commission

under the Seventh Framework Program for Research (FP7-KBBE-266124, GREENLAND).

Authors thank Florien Nsangwimana for his technical assistance. We are grateful to Daphne

Goodfellow for improving the English.

References (Format : Science of the Total Environment)

Adriano DC, Wenzel WW, Vangronsveld J, Bolan NS. Role of assisted natural remediation in environmental cleanup. Geoderma 2004;122: 121-42.

Aulen M, Bes C, Mench M, Gasté H, Taberly J, Laporte-Cru J, Touzard B. Plant communities at a wood preservative treatment site, and Cu tolerance patterns of plant species. In: Zhu YG, Lepp N, Naidu R, editors. Biogeochemistry of Trace Elements: Environmental Protection, Remediation and Human Health. Tsinghua University Press, Beijing, PR China; 2007. p. 152-154.

Babu TS, Akhtar TA, Lampi MA, Tripuranthakam S, Dixon DG, Greenberg BM. Similar stress responses are elicited by copper and ultraviolet radiation in the aquatic plant Lemna gibba: Implication of reactive oxygen species as common signals. Plant Cell Physiol 2007; 44: 1320-9.

Baize D. Teneurs Totales en Éléments Traces Métalliques dans les Sols (France). Références et Stratégies d'Interprétation. Paris : INRA éditions; 1997.

Balasoiu CF, Zagury G, Deschenes L. Partitioning and speciation of chromium, copper, and arsenic in CCA-contaminated soils: influence of soil composition. Sci Total Environ 2001; 280: 239-55.

Bes CM, Mench M, Guinberteau J, Garcia San Martin S, Monmarson M. Sustainable management of a Cu-contaminated soil by phytostabilization: Effect of ectomycorrhization on the growth of 3 Salicaceae species in field plots. In: Baltrenas P, Baltrenaite E, editors. Fate of Pollutants in the Plant/Rhizosphere System: Fundamental Aspects and their Significance for Field Applications – Prospects and Research Needs. Cost Action WG2+WG4 Workshop, 30th May – 1st June, VGTU Press Technica, Vilnius, Lithuania, ISBN 978-9955-28-123-8; 2007. p. 18-19. Available at: http://w3.gre.ac.uk/cost859/documents/WG2_WG4Workshop_Vilnius%202007_Abstracts.pdf

Bes CM, Mench M. Remediation of copper-contaminated top soils from a wood treatment facility using in situ stabilization. Environ Pollut 2008; 156: 1128-38.

Bes CM., Mench M, Aulen M, Gaste H, Taberly J. Spatial variation of plant communities and shoot Cu concentrations of plant species at a timber treatment site. Plant Soil 2010; 330: 267-80.

98

Chirenje T, Ma LQ. Effects of acidification on metal mobility in a papermill-ash amended soil. J Environ Qual 1999; 28: 760-6.

Cvjetko P, Tolic S, Sikic S, Balen B, Tkalec M, Vidakovic-Cifrek Z, Pavlica,M. Effect of copper on the toxicity and genotoxicity of cadmium in duckweed (Lemna minor L.). Arhiv Za Higijenu Rada I Toksikologiju 2010; 61: 287-96.

Doyle MO, Otte ML. Organism-induced accumulation of iron, zinc and arsenic in wetland soils. Environ Pollut 1997; 96: 1-11.

Drost W, Matzke M, Backhaus T. Heavy metal toxicity to Lemna minor: studies on the time dependence of growth inhibition and the recovery after exposure. Chemosphere 2007; 67: 36-43.

EPA. Ecological Effects Test Guidelines. OPPTS 850.4400 Aquatic Plant Toxicity Test Using Lemna spp., Tiers I and II. EPA 712-C-96-156, Prevention, Pesticides and Toxic Substances 7101. Washington: US EPA; 1996.

Hewitt EJ. Sand and Water Culture Methods used in the Study of Plant Nutrition. London : The Eastern press Ltd; 1966.

Horvat T, Vidakovic-Cifrek Z, Orescanin V, Tkalec M, Pevalek-Kozlina B. Toxicity assessment of heavy metal mixtures by Lemna minor L. Sci. Total Environ 2007; 384: 229-38.

Hou WH, Chen X, Song GL, Wang QH, Chang, CC. Effects of copper and cadmium on heavy metal polluted waterbody restoration by duckweed (Lemna minor). Plant Physiol Biochem 2007; 45: 62-9.

Huebert DB, Shay JM. The effect of cadmium and its interaction with external calcium in the submerged aquatic macrophyte Lemna trisulca L. Aquat Toxicol 1991; 20: 57-71.

Ince NH, Dirilgen N, Apikyan IG, Tezcanli G, Ustun B. Assessment of toxic interactions of heavy metals in binary mixtures: A statistical approach. Arch Environ Contam Toxicol 1999; 36: 365-72.

INERIS. Données technico-économiques sur les substances chimiques en France: cuivre, composés et alliages, DRC-10-102861-01255A. Verneuil-en-Halatte : INERIS; 2010.

Available at: http://rsde.ineris.fr/ and http://www.ineris.fr/substances/fr/

INERIS. Fiches de données toxicologiques et environnementales, updated 1/02/11. Verneuil-en-Halatte : INERIS; 2011.

Available at: http://www.ineris.fr/substances/fr/substance/683

Kamal M, Ghaly,A.E, Mahmoud N, Cote R. Phytoaccumulation of heavy metals by aquatic plants. Environment International 2004; 29: 1029-39.

Jackson ML. Soil Chemical Analysis. New Dehli : Prentice Hall of India Pvt. Ltd; 1967.

Jenner HA, Janssen-Mommen JPM. Duckweed Lemna minor as a tool for testing toxicity of coal residues and polluted sediments. Arch Environ Contam Toxicol 1993; 25: 3-11.

Kanoun-Boule M, Vicente JAF, Nabais C, Prasad MNV, Freitas H. Ecophysiological tolerance of duckweeds exposed to copper. Aquat Toxicol 2009; 91: 1-9.

Karjalainen A, Kilpi-Koski J, Vaisanen A, Penttinen S, Gestel C, Penttinen O. Ecological risks of an old wood impregnation mill: application of the Triad approach. Integrated Environmental Assessment and Management (IEAM) 2009; 5: 379-89.

Khellaf N, Zerdaoui M. Growth response of the duckweed Lemna minor to heavy metal pollution. Iran. J Environ Health Sci Eng 2009; 6: 161-6.

Kidd P, Barcelo J, Bernal MP, Navari-Izzo F, Poschenrieder C, Shilev S, Clemente R, Monterroso C. Trace element behaviour at the root-soil interface: Implications in phytoremediation. Environ. Exper Bot 2009; 67: 243-59.

99

Kumpiene J, Ore S, Renella G, Mench M, Lagerkvist A, Maurice C. Assessment of zerovalent iron for stabilization of chromium, copper, and arsenic in soil. Environ Pollut 2006; 144: 62-9.

Kumpiene J, Lagerkvist A, Maurice C. Stabilization of As, Cr, Cu, Pb and Zn in soil using amendments - A review. Waste Management 2008; 28: 215-25.

Kumpiene J, Mench M, Bes CM, Fitts JP. Assessment of aided phytostabilization of copper-contaminated soil by X-ray absorption spectroscopy and chemical extractions. Environ Pollut 2011; 159: 1536-42.

Kupper H, Setlik I, Spiller M, Kupper FC, Prasil O. Heavy metal-induced inhibition of photosynthesis: Targets of in vivo heavy metal chlorophyll formation. J Phycol 2002; 38: 429-41.

Lagomarsino A, Marabottini R, Grego S, Stazi SR. Copper distribution among physical and chemical fractions in a former vineyard soil. Agrochimica 2010; 54: 167-78.

Lizama A K, Fletcher TD, Sun G. Enhancing the removal of arsenic, boron and heavy metals in subsurface flow constructed wetlands using different supporting media. Water Sci Technol 2011(a); (accepted, doi: wst.2011.533).

Lizama A K, Fletcher TD, Sun G. Removal processes for arsenic in constructed wetlands.. Chemosphere 2011(b); (In press, doi:10.1016/j.chemosphere.2011.01.022)

Llorens N, Arola L, Blade C, Mas A. Effects of copper exposure upon nitrogen metabolism in tissue cultured Vitis vinifera. Plant Sci 2010; 160: 159-63.

Ma M, Zhu WZ, Wang ZJ, Witkamp GJ. Accumulation, assimilation and growth inhibition of copper on freshwater alga (Scenedesmus subspicatus 86.81 SAG) in the presence of EDTA and fulvic acid. Aquat Toxicol 2003; 63: 221-28.

Marchand L, Mench M, Jacob DL, Otte ML. Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants and standardized measurements: A review. Environ Pollut 2010; 158: 3447-61.

Megateli S, Semsari S, Couderchet M. Toxicity and removal of heavy metals (cadmium, copper, and zinc) by Lemna gibba. Ecotoxicol Environ Saf 2009; 72: 1774-80.

Mench M, Vangronsveld J, Clijsters H, Lepp NW, Edwards R. In situ metal immobilization and phytostabilization of contaminated soils. In: Terry N., Banuelos G, editors. Phytoremediation of Contaminated Soil and Water. Boca Raton : Lewis Publishers, CRC Press; 2000. p. 323-58.

Mench M, Bussiere S, Boisson J, Castaing E, Vangronsveld J, Ruttens A, De Koe T, Bleeker P, Assuncao A, Manceau A. Progress in remediation and revegetation of the barren Jales gold mine spoil after in situ treatments. Plant and Soil 2003; 249: 187-202.

Mench M, Vangronsveld J, Lepp N, Bleeker P, Ruttens A, Geebelen W. Phytostabilisation of metal-contaminated sites. In: Echevarria G, Morel JL, Goncharova N, editors. Phytoremediation of Metal-Contaminated Soils. Dordrecht : Nato Science Series: IV: Earth and Environmental Sciences 68. Springer; 2000. p. 109-90.

Mench M, Gasté H, Bes C. Phenotypic traits of metallicolous and non-metallicolous Agrostis capillaris exposed to Cu. In: Liskova D, Lux A, Martinka M, editors. Contaminants and Nutrients: Availability, Accumulation/Exclusion and Plant-Microbia-Soil Interactions, COST 859 – WG1 meeting, Smolenice, May 22 -24. Bratislava : Mgr. P Cibulka, Copycentrum, PACI, ISBN 978-80-969950-0-4; 2008. p.19.

Available at: http://w3.gre.ac.uk/cost859/WG1/WG1_Smolenice2008_Abstracts.pdf

Mench M, Bes C. Assessment of ecotoxicity of topsoils from a wood treatment site. Pedosphere 2009; 19: 143-55.

100

Mench M, Renella G, Bes C, Herzig R, Gawronski S, Migliaccio A, Jaunatre R, Aulen M, Gasté H, Taberly J, Garcia-Martin S, François J, Nehnevajova E, Guinberteau J. Phytoremediation of Cu-contaminated soils at a timber impregnation site. In: Alarcon-Herrera MT, editor. Frontiers in Trace Elements Research and Education, 10th International Conference on the Biogeochemistry of Trace Elements. July 13-16. Chihuahua : CIMAV; 2009.

Available at: http://icobte2009.cimav.edu.mx/participant/xo3erif/search.php

Mench M, Lepp N, Bert V, Schwitzguébel JP, Gawronski SW, Schroder P, Vangronsveld J. Successes and limitations of phytotechnologies at field scale: outcomes, assessment and outlook from COST Action 859. J Soils Sed 2010; 10: 1039-70.

Mills T, Arnold B, Sivakumaran S, Northcott G, Vogeler I., Robinson BH, Norling C, Leonil D. Phytoremediation and long-term site management of soil contaminated with pentachlorophenol (PCP) and heavy metals. J Environ Manage 2006; 79: 232-41.

Mocquot B, Vangronsveld J, Clijsters H, Mench M. Copper toxicity in young maize (Zea mais L.) plants: effect on growth, mineral and chlorophyll contents and enzyme activities. Plant and Soil 1996; 182: 287–300.

Naumann B, Eberius M, Appenroth KJ. Growth rate based dose response relationships and EC-values of ten heavy metals using the duckweed growth inhibition test (ISO 20079) with Lemna minor L. clone St. J. Plant Physiol 2007; 164: 1656-64.

Negim O, Eloifi B, Mench M, Bes C, Gaste H, Motelica-Heino M, Le coustumer P. Effect of basic slag addition on soil properties, growth and leaf mineral composition of beans in a Cu-contaminated soil J Soil Sed Contam 2009; 19: 174-87.

Radic S, Stipanicev D, Cvjetko P, Mikelic IL, Rajcic MM, Sirac S, Pevalek-Kozlina B, Pavlica M. Ecotoxicological assessment of industrial effluent using duckweed (Lemna minor L.) as a test organism. Ecotoxicology 2010; 19: 216-22.

Rengel Z. Competitive Al inhibition of net Mg uptake by intact Lolium multiflorum roots: II. Plant age effects. Plant Physiol 1990; 93: 1261-67.

Riggs KS, Rimmer DL, Syers JK. Effect of liming on the chemical-composition of soils. J Sci Food Agric 2000; 69: 159-67.

Romero-Puertas MC, Palma JM, Gomez M, Del Rio LA, Sandalio LM. Cadmium causes the oxidative modification of proteins in pea plants. Plant Cell Environ 2002; 25: 677-86.

Ruttens A, Mench M, Colpaert JV, Boisson J, Carleer R, Vangronsveld J. Phytostabilization of a metal contaminated sandy soil. I: Influence of compost and/or inorganic metal immobilizing soil amendments on phytotoxicity and plant availability of metals. Environ Pollut 2006; 144: 524-32.

Ruttens A, Colpaert JV, Mench M, Boisson J, Carleer R, Vangronsveld J. Phytostabilization of a metal contaminated sandy soil. II: Influence of compost and/or inorganic metal immobilizing soil amendments on metal leaching. Environ Pollut 2006; 144: 533-9.

Ryan PR, Delhaize E, Jones DL. Function and mechanism of organic anion exudation from plant roots. Annu. Rev. Plant Physiol Plant Mol 2001; 52: 527-60.

Salpeteur I, Angel JM. Valeurs de références pour les teneurs en éléments traces dans les eaux de rivières et les sédiments, obtenues en France dans le cadre du nouvel atlas géochimique Européen (FOREGS). Environnement Risque et Santé 2010; 9: 121-35.

Sauvé S. Modelling trace element exposure and effects on plants. In: Mench M, Mocquot B, editors. Risk Assessment and Sustainable Land Management using Plants in Trace Element-Contaminated Soils. Villenave d'Ornon, Centre INRA Bordeaux-Aquitaine; 2003. p. 69-70.

Available at: http://w3.gre.ac.uk/cost859/book/Livre_COST837_2003_session_1.pdf.

Schaul O. Magnesium transport in plants: the tip of the iceberg. Biometals 2002; 15: 309-23.

101

Sheoran AS, Sheoran V. Heavy metal removal mechanism of acid mine drainage in wetlands: A critical review. Miner Eng 2006; 19: 105-16.

Sjöberg V, Karlsson S, Sartz L. Release of vanadium from LD-slag by exposure to ARD. In: Wolkersdorfer C, Freund A, editors. International Mine Water Association Symposium - Mine Water and Innovative Thinking. Proceedings IMWA 2010, Sydney, Nova Scotia, Canada, p. 399-402.

Available at: http://www.imwa.info/docs/imwa_2010/IMWA2010_Sjoberg_423.pdf

Sudhakar Babu T, Akthar TA, Lampi MA, Tripuranthakam S, Dixon DG, Greenberg BM. Similar stress responses are elicited by copper and ultraviolet radiation in the aquatic plant Lemna gibba: implication of reactive oxygen species as common signals. Plant Cell Physiol 2003; 44: 1320–29.

Teisseire H, Couderchet M, Vernet G. Toxic responses and catalase activity of Lemna minor L. exposed to folpet, copper, and their combination. Ecotoxicol Environ Saf 1998; 40: 194-200.

Upadhyay R, Panda SK. Zinc reduces copper toxicity induced oxidative stress by promoting antioxidant defense in freshly grown aquatic duckweed Spirodela polyrhiza L. J Hazard Mater 2010; 175: 1081-84.

Vangronsveld J, Colpaert JV, Van Tichelen KK. Reclamation of a bare industrial area contaminated by non-ferrous metals: Physicochemical and biological evaluation of the durability of soil treatment and revegetation. Environ Pollut 1996; 94: 131-40.

Verdugo C, Sánchez P, Santibáñez C, Urrestarazu P, Bustamante E, Silva Y, Gourdon D, Ginocchio R. Efficacy of lime, biosolids and mycorrhiza for the phytostabilization of sulfidic copper tailings in Chile: a greenhouse experiment. Int J Phytorem 2011; 13: 107-25.

Wang WC. Toxicity tests of aquatic pollutants by using common duckweed. Environ Pollut 1986; 11: 1-14.

Wang P, Zhou DM, Peijnenburg WJGM, Li LZ, Weng NY. Evaluating mechanisms for plant-ion (Ca2+, Cu2+, Cd2+or Ni2+) interactions and their effectiveness on rhizotoxicity. Plant and Soil 2010; 334: 277-88.

Chapitre I : Take home message

La phytostabilisation aidée conduite sur un site de traitement du bois au sol contaminé

en Cu a permis de réduire les concentrations en Cu dans les lixiviats de lysimètres. Le couvert

végétal utilisé, tolérant à Cu (Agrostis gigantea L. et Populus trichocarpa x deltoides cv.

Beaupré) ne réduit pas les concentrations en Cu2+ dans les lixiviats. L’utilisation

d’amendements du sol a en revanche permis de diminuer les teneurs en PTTE dans ces

lixiviats. Tous les amendements n’ont pas contribué à cette immobilisation des PTTE dans le

sol avec la même efficacité. L’impact de la contamination sur la qualité des lixiviats, modulé

par l’utilisation d’amendements, a été évalué par un biotest basé sur la croissance de Lemna

minor L. Deux traitements n’ont aucun effet sur la qualité des eaux en comparaison avec le

sol contaminé non traité, il s’agit de la modalité sol cultivé avec des plantes tolérantes au Cu,

mais non amendé (PHYTO, pour phytostabilisation), et de la modalité sol cultivé et amendé

avec un mélange de grenailles d’acier (Z, 2%, donnant par corrosion des oxydes de Fe et Mn)

et de compost (OM). Les deux amendements permettant un meilleur développement de L.

minor dans les lixiviats sont une dolomie (DL, 0,2%) et un laitier sidérurgique de type Linz-

102

Donawitz (LDS, 1%). La qualité des lixiviats dépend de leur concentration en Cu2+, pH, et des

concentrations en autres éléments tel Mg. L’effet délétère de OMZ sur le développement de L.

minor est corrélé avec une complexation de Mg sur les (oxy)hydroxydes de Fe, d’où une

concentration moins importante dans les lixiviats. Les concentrations des contaminants

principaux du site ne doivent pas être les seules suivies dans les lixiviats après mise en place

d’un dispositif de phytostabilisation aidée ; l’ensemble des éléments (macro et

micronutriments, métaux et métalloides) doit être suivi.

Aucun des amendements utilisés n’a permis d’atteindre une eau de qualité similaire à celle

d’une eau de rivière. L. minor s’est mieux développée sur un échantillon d’eau de la Jalle

d’Eysines, une rivière urbaine localisée au nord de Bordeaux, que sur l’ensemble des lixiviats.

Les techniques de phytostabilisation aidée utilisées sur ce site contaminé au Cu nécessitent

donc d’être améliorées. La présence de zones humides construites en aval proche de ce type

de site, de même que le traitement des eaux de ruissellement et de lessivage en CWs, apparait

comme l’une des phytotechnologies éco-innovatives envisageables dans ce contexte.

Le suivi des contaminants dans les compartiments de l’écosystème est une nécessité pour

savoir quand, où et comment intervenir dans le cadre de programmes de phytoremédiation et

de restauration des milieux. Ce suivi est souvent réalisé sur les matrices eau, sol ou eau

intersticielle. Si l’analyse chimique de ces matrices renseigne sur leur état de contamination,

elle n’informe pas sur l’intensité de l’exposition des organismes, qui dépend d’une cascade

d’interactions biotiques et abiotiques. Le bio-monitoring (dont une traduction est

biosurveillance) peut intégrer sur des temps longs l’intensité de l’exposition et ses variations.

Les macrophytes comptent parmi les bio-moniteurs de la qualité des zones humides. Les

concentrations en PTTE dans leurs parties aériennes et souterraines sont utilisées en

biomonitoring. Cependant, l’exposition aux PTTE est souvent quantifiée par leur suivi chez

chaque espèce végétale, pour un élement indépendamment des autres. Or l’exposition à un

contaminant impacte pourtant l’ensemble du ionome au sein de la plante. Dans le chapitre

suivant, nous proposons une approche multivariée – via l’utilisation d’une analyse

discriminante – pour évaluer l’intensité de l’exposition des macrophytes à l’ensemble des

PTTE. Ce travail a été réalisé sur les macrophytes des berges de la rivière Jalle d’Eysines.

103

Chapitre II

Bio-surveillance de l’évolution de l’exposition des macrophytes aux

PTTE…Le cas d’une rivière urbaine la Jalle d’Eysines

Cette partie est soumise à Freshwater Biology

104

Figure 22 : La Jalle d’Eysines, station J1

The Jalle d’Eysines River, sampling site J1

Figure 23 : La Jalle d’Eysines, station J2

The Jalle d’Eysines River, sampling site J2

105

Figure 24 : La Jalle d’Eysines, station J3

The Jalle d’Eysines River, sampling site J3

Figure 25 : La Jalle d’Eysines, station J4

The Jalle d’Eysines River, sampling site J4

106

Macrophytes as biomonitors of trace element exposure along an urban

river using a multimetric approach (Jalle d’Eysines River, France)

L. Marchand a, M. Mencha, F. Nsanganwimanaa,b, Y. Vystavnac,d, F. Huneauc, P. Le Coustumerc, J.B. Lamya, B.J. Cooke

a Université de Bordeaux, UMR 1202 BIOGECO INRA, Bât. B2, RDC Est, Avenue des Facultés, F-33405 Talence, France

b Laboratoire Génie Civil et géo-Environnement (LGCgE-EA 4515), Equipe Sols et Environnement, Groupe ISA, 48 Boulevard Vauban, 59046 Lille Cedex, France

c Université de Bordeaux, EA4592 Géoressources & Environnement, ENSEGID, 1 Allée F. Daguin, F-33607 Pessac, France

d National Academy of Municipal Economy at Kharkiv, Department of Environmental Engineering and Management, vul. Revolutsii 12, Kharkiv, 61002, Ukraine

e Minnesota State University, Department of Biological Sciences, Mankato, Minnesota, 56001, USA

Abstract

1. The concentrations of potentially toxic trace elements (PTTE) in the leaves of seven

macrophytes (Phragmites australis, Phalaris arundinacea, Ranunculus acris, Carex riparia,

Juncus effusus, Iris pseudacorus and Lythrum salicaria) and in the corresponding water, soil

and soil pore water samples were investigated to assess the suitability of macrophyte leaves

for biomonitoring of PTTE exposure along an urban river, the Jalle d’Eysines River

(Bordeaux, France).

2. Analyses were performed using a Linear Discriminant Analysis (LDA) based on foliar

concentrations of PTTE (Cu, Zn, Cr, Cd, As, Pb, Mo and Ni), macro nutrients (P, Mg, K and

Ca), Fe and Mn for determining which elements in which macrophyte leaves best

discriminated the sampling sites along the river after cross validation.

3. Macrophytes of the Jalle d’Eysines River were grouped in two life forms, the rhizomatous-

geophytes and the hemicryptophytes and hypothesized that PTTE concentrations in leaves of

macrophytes are life-form dependent. Three LDA were performed. The first LDA considered

foliar element concentrations of the whole macrophyte dataset as the macrophyte community.

The second LDA only considered the foliar element concentrations in the rhizomatous-

107

geophytes while the third was based on the foliar element concentrations in the

hemicryptophytes.

4. Foliar PTTE accumulation in macrophytes of the Jalle d’Eysines River as the expression of

the total PTTE concentration in riverbank soils often occurred but it was not the rule. The

global PTTE exposure at the Jalle d’Eysines River results from total PTTE concentrations in

the soil, PTTE concentrations in the soil pore water, soil texture, and the kinetics of reactions

between soil bearing phases and the soil pore water.

5. The LDA models classified after external cross-validation respectively 70% of the whole

macrophyte community, 80% of the rhizomatous-geophytes and 89% of the

hemicryptophytes. Hemicryptophytes such as L. salicaria, P. arundinacea and R. acris

emerged as best biomonitors when building a LDA model for assessing PTTE exposure along

the Jalle d’Eysines River. Hemicryptophytes transferred more PTTE in their leaves than

rhizomatous geophytes which trapped them more into the roots and rhizomes.

6. A multivariate analysis such as LDA better assesses macrophyte exposure to PTTE over

space and time than linear regressions because it integrates synergies and antagonisms

occurring in the ecosystems in terms of PTTE transfer. Further investigations are needed to

assess the reliability of these results on other rivers.

Keywords : Biomonitoring, Geophyte, Hemicryptophyte, Linear Discriminant Analysis,

Metal, Riverbank, Soil pore water

108

I. Introduction

The ecological status of many wetlands is affected by anthropogenic activities, some being

major sinks and sources of Potentially Toxic Trace Elements (PTTE – here essential and non-

essential metal(loid)s with common shoot concentrations below 100 mg kg-1 dry weight, DW)

(Adriano, 2001). As a result, PTTE may accumulate in surface waters, groundwater, soils, and

macrophytes (Aksoy et al., 2005). Root uptake and foliar accumulation of PTTE by

macrophytes depend on plant species, abiotic factors such as soil and water pH, temperature,

salinity, redox potential, and bearing phases such as (oxy)hydroxides, organic matter, and

clays, in the soil matrix, and on PTTE chemical speciation, notably free ions and organic

complexes (Bonanno, 2011). Due to their dense fibrous root system with large surface areas,

and well developed rhizome tissues, rooted macrophytes mainly accumulate PTTE in their

belowground part (Cardwell et al., 2002; Bonnano et al., 2010; Romero Nuñez et al., 2011).

Rhizome cortex generates vacuoles where PTTE are sequestered by metal-binding proteins

such as metallothioneins (MT) and phytochelatins (PC) (Caldelas et al., 2012). Thus, PTTE

accumulate in lesser concentrations in stems and leaves than in roots and rhizomes (Clemens

2002; Baldantoni et al., 2004; Bragato et al., 2006)

The European Water Framework Directive (EU WFD, 2000) recommends PTTE

contamination assessment along riverbanks. Physico-chemical analysis of environmental

matrices such as water and soils is a typical direct option to assess their PTTE contamination

status, but it cannot afford powerful evidence of the biological actions of such environmental

contamination, PTTE behavior in living organisms, and notably phytotoxicity risks (Zhou et

al., 2008). Biomonitoring and bioindication, using the sampling and analysis of living

organisms in a given ecosystem, are tools for assessing environmental effect due to xenobiotic

exposures (Markert, 2007). Numerous biotic indices and multimetric procedures based on the

sampling of living organisms have been developed to assess the status of freshwater

ecosystems (Lenoir and Coste, 1996; Schneider and Melzer, 2003). Aguiar et al. (2009) and

Feio et al. (2012) proposed two predictive models based on macrophyte community

compositions for ecological assessment of rivers. Archaimbault et al. (2010) examined the

accuracy of a multimetric approach based on biological and ecological traits of benthic

macroinvertebrate communities to assess toxic sediment pollution in streams. Multimetric

109

procedures such as the linear discriminant analysis (LDA) are powerful tools to assess

environmental contamination as they integrate the overall biotic and abiotic interactions in the

ecosystem. Identification of relevant biomonitors is thus needed to develop these models.

Markert et al. (2007) defined bioindicators as organisms and communities that contain

information on the quality of the environment, whereas biomonitors can provide quantitative

information on the quality of the environment and on pollutant linkages. In any case,

analytical data from biological samples do not provide information on ‘the state of the

environment’ directly, rather they provide information of a complex relationship combining

exposure levels, biological uptake and transfer processes with PTTE responding to both

abiotic and biotic diffusion barriers and dynamics when binding to biomass (Markert, 2007;

Fraenzle, 2007). Biomonitoring complements the chemical analysis of environmental

matrices, which can account for the subtle biological changes of organisms affected by

exogenous chemicals (Zhou et al., 2008). In wetlands, macrophytes have been used as

bioindicators (Klumpp et al., 2002; Bonanno and lo Giudice, 2010; Ladislas et al., 2012;

Bonanno, 2012) even if their potential use is discussed according to the considered

contaminant (Demars and Edwards, 2009). Zhou et al. (2008) reported that, for assessment of

PTTE in aquatic ecosystems, a relevant bioindicator is expected to: (1) accumulate high

contaminant levels without death; (2) be sessile, thus representing the local environmental

conditions; (3) be abundant and widely distributed for repetitive sampling and comparison;

(4) live long enough for the comparison between ages; (5) afford suitable target tissue or cells

for the further research at a microcosmic level; (6) be easy to sample; (7) be kept alive in

water (8) occupy an important position in food chain; and (9) have a well understood dose-

effect relationship. Macrophytes combine all these points except point (9). Metabolic activity

of exposed macrophytes and the root plaque effect (Kissoon et al., 2010) may obscure the

dose-effect relationship.

However for a given wetland and sampling time, and accounting for soil physico-chemical

parameters and their interactions with total and labile soil PTTE fractions, a relationship may

be expected between macrophyte exposure and PTTE accumulation in below- and/or

aboveground plant parts. This study focused on seven macrophytes classified as rhizomatous-

geophyte and hemicryptophyte life forms (Raunkiær, 1934). Hemicryptophytes exhibit shoot

110

apical meristems borne at or near soil level whereas rhizomatous-geophytes have shoot apical

meristems borne below the soil level and produce rhizomes. Although it is claimed that

macrophytes accumulate PTTE mainly in roots and rhizomes (Marchand et al., 2010;

Bonanno, 2011; Lizama et al., 2011) and that foliar PTTE concentrations of macrophytes

remain nearly unchanged along contaminated river courses (Klumpp et al., 2002), we

hypothesized that foliar PTTE concentrations of macrophytes are life-form dependant. Linear

discriminant analyses (LDA) based on foliar element concentrations were used to determine

which macrophyte leaves best described PTTE exposure along an urban river, the Jalle

d’Eysines River, in France.

II. Materials and methods

2.1 Description of the studied area

The Jalle d’Eysines River is located in southwest France (44°53′36″N; 00°40′40″O), North of

Bordeaux, and is a tributary of the Garonne River (Figure 26). From its headwaters to its

confluence with the Garonne River, it is 32 km long. Water depth typically varies from 0.8 to

2.5 m annually, and average water debit is 3 m3s-1. It receives PTTE-contaminated runoff

from industrial, agricultural and residential areas and effluents from two major municipal

wastewater treatment plants (WTP) that serve more than 100,000 inhabitants in the Bordeaux

suburbs. Treated effluents can account for up to 33% of the river flow (Labadie and

Budzinski, 2005). Anthropogenic inputs are reportedly high in metal(loid)s (La CUB 2006),

and organic xenobiotics such as steroidal hormones (i.e. estrone, estradiol, and estriol,

Labadie and Budzinski, 2005), xeno-estrogens (Hinfray et al., 2010), alkylphenols (Miege et

al., 2011), pesticides and pharmaceuticals (Vystavna et al., 2012). High levels of ammonium

perchlorate also have been identified in groundwater connected to this river (Basol 2012).

Four, 100-m reaches along the Jalle d’Eysines River were selected as sampling sites. Based

on preliminary studies (Vystavna et al., 2012), each site was selected to represent a

disturbance gradient from the river’s source to its confluence with the Garonne River. Sites 1

and 2 were located at 10 km and 20 km from the source. Site 1 was selected because of low

anthropogenic pressure (Figure 22). Site 2 was assumed to be more impacted by runoff from

adjacent lands with organic farming and was located 1 km downstream from the WTP

“Cantinolle” (near Eysines) (Figure 23). Site 3 and 4 were located at 27 km and 30 km from

111

the source and were located 1 km and 4 km downstream of the second WTP “L’Ile” (near

Blanquefort) (Figures 24, 25). Adjacent land use to Site 3 was organic farming and lands

adjacent to Site 4 were conventional maize crops.

Figure 26 : Sampling sites along the Jalle d’Eysines River (“Cantinolles” and “L’Ile”: Wastewater

treatment plants)

2.2 Water, soils and macrophytes sampling

Six freshwater samples were collected from the river at each site in May and June 2011. Each

sample was collected by submersing a plastic bottle (1.5 L) 10 cm below the water surface.

Bottles were previously rinsed with deionized water, and stored at 4°C prior to total fraction

analysis at the INRA USRAVE laboratory, Villenave d'Ornon, France with ICP-AES (Varian

liberty 200) and ICP-MS (Thermo X serie 200). Six soil samples (totaling 1.5 kg fresh weight,

FW) were collected in June 2011 at the riverbank of each site. Samples were collected using

an unpainted steel spade approximately ~1.5 m from the water, within the 0-25 cm soil layer

and 1 m apart.

112

On the same day in June 2011, the macrophytes Meadow buttercup (Ranunculus acris L.),

Common reed (Phragmites australis (Cav.) Trin. Ex Steud.), Common rush (Juncus effusus

L.), Greater pond sedge (Carex acutiformis Ehrh.), Purple loosestrife (Lythrum salicaria L.),

Yellow flag (Iris pseudacorus L.) and Reed canary grass (Phalaris arundinacea L.) were

sampled. These macrophytes were selected because they represent more than 80% of the

macrophyte biomass along the banks of the Jalle d’Eysines River (personal data). Each

species was grouped by life forms. Phragmites australis, I. pseudacorus, J. effusus and C.

riparia are helophytes/rhizomatous-geophytes (C. riparia can also be considered as

hemicryptophyte) whereas L. salicaria, P. arundinacea and R. acris are hemicryptophytes (L.

salicaria can also be considered as helophyte) (Raunkiær, 1934). At most sites, six leaf

samples from six individuals (totaling ~250g FW) were collected. Exceptions occurred for R.

acris where three individuals were pooled to make each leaf sample and for J. effusus at Site 4

and I. pseudacorus at Site 2 where individuals of these species were absent. Leaf samples

were collected immediately adjacent to soil collection sites. Individual leaf samples were kept

in paper bags. In preparation for mineral analysis, leaf samples were carefully washed with

tap water, rinsed with deionized water, blotted with filter paper, placed in paper bags, and

oven-dried for 48h at 55°C to a constant mass.

2.3 Water and soil analysis

The six soil samples taken from a site were mixed. A 0.5 kg sample was air-dried and sieved

at 5 mm (nylon mesh) prior to analysis and 1 kg was potted into a 1.3L plastic pot. Total

PTTE and macro-element concentrations, soil pH, EC, and Eh, texture were measured on air-

dried soil at the INRA Laboratoire d'Analyses des Sols (LAS, Arras, France) using standard

methods (INRA LAS, 2007). To determine soil pH, EC, and Eh, 25 mL of distilled water was

mixed to 10 g of air-dried soil and the mixture was allowed to react for 2h before

measurements. Potted soils were watered with deionized water and daily maintained at 70%

of their water holding capacity (WHC, Table 10). After one week, a single rhizon MOM

moister sampler (Eijkelkamp, The Netherlands) was inserted at 45° angle into each potted

soil. Soil pore water samples (20 mL) were collected after 7 days and analyzed with ICP-AES

(Varian liberty 200, USA) and ICP-MS (Thermo X serie 200, USA) at the INRA USRAVE

laboratory. Electrical conductivity, redox-potential (EC and Eh, WTW Multiline P4 meter,

113

Germany), and pH (Hanna instruments, pH 210, combined electrode Ag/AgCl, USA) were

measured in water, soil samples and soil pore water from all potted samples.

2.4 Mineral analysis of leaf samples

Dried leaf samples were ground (<1.0 mm particle size, Retsch MM200) and stored in plastic

containers (100 mL) at room temperature, in dark conditions, prior analysis. Weighted

aliquots (0.5g DW) were wet digested under microwaves (180°C, CEM Marsexpress, USA)

with 5 mL supra-pure 14 M HNO3, 2 mL 30% (v/v) H2O2 not stabilized by phosphates, and 1

mL milli-Q water. Certified reference material (BIPEA maize V463) and blank reagents were

included in all series. Macro elements and PTTE concentrations in digests were determined

by ICP-AES (Varian liberty 200) and ICP-MS (Thermo X serie 200) at the INRA USRAVE

laboratory, Villenave d'Ornon, France. All elements were recovered (>95%) according to the

standard values and standard deviation for replicates (n=3) was <5%. All element

concentrations in plant and soil samples are presented on DW basis.

2.5 Statistical analysis

When normality and homoscedasticity of data were met, one-way ANOVAs were performed

to compare differences in soil, freshwater, soil pore water, and leaf parameters according to

sites (1 to 4). The post-hoc Tukey HSD test was then used to assess multi-comparison of

means between sites. When assumptions were not met, Wilcoxon pairwise tests adjusted with

a Bonferroni correction were used. To determine the elements in which macrophyte leaves

best discriminated the four sites, linear discriminant analysis (LDA) was performed after

checking multivariate normality of data (http://www.statmethods.net/stats/anovaAssumptions.html,22/06/11)

and multivariate variance homogeneity (Tabachnick and Fidell, 2007). A preliminary LDA

was performed on the whole macrophyte community (all species). Plant species were then

grouped into two life forms (rhizomatous-geophytes and hemicryptophytes) and additional

LDA were performed on both groups. For each model, 75% of the dataset was defined as the

original training dataset while the remaining 25% was used as the testing dataset to assess

the accuracy of the model by cross-validation. For all LDA, multivariate homoscedasticity

was met. Multivariate normality of data was not fully met, however multivariate normality of

data is rarely fully met and the linear model is robust to violations of the normality

114

assumption when the assumption of homoscedasticity is met (Demirtas et al., 2008). The

LDA was preferred to the quadratic discriminant analysis (QDA) since sensitivity for QDA is

generally the same as that obtained by LDA, but specificity is generally slightly lower

(https://onlinecourses.science.psu.edu/stat857/book/export/html/80). All statistical analyses were performed

using R software (version 2.12.0, R Foundation for Statistical Computing, Vienna, Austria).

III. Results

3.1 Freshwater, soil and soil pore water

Fresh water pH and EC increased from Site 1 (respectively 6.8 and 0.28 mS cm-1) to Site 4

(7.6 and 0.94 mS cm-1) (Table 10). The increase in EC from freshwater samples corresponded

with changes in element concentrations along the Jalle d’Eysines River (Table 11).

Freshwater Mg, K, Mo, Ca, and As concentrations also increased from Site 1 to Site 4,

whereas Ni, Fe and P concentrations remained constant, and total Zn, Cu, Cd, Cr, Pb and Mn

concentrations in freshwater samples were below the detection limits (Table 11).

Concentrations of PTTE in Jalle d’Eysines River freshwater samples were in compliance with

the standards defined by the French Water Agency (SEQ EAU, 2003) for good water quality

(e.g. Cu <10 µg L-1and Zn <43 µg L-1) and similar to the median PTTE concentrations in

French surface waters except for Mo concentration below the L’Ile WTP which reached 8.6

µg L-1 (Table 11).

Soil texture was sandy at Sites 1 and 2 with low clay proportions while soils from Sites 3 and

4 contained more clays than sands (Table 10). Similarly soils from Sites 1 and 2 did not differ

in C, N and OM contents and had lower C, N, and OM contents than in soils from Sites 3 and

4. Soil CEC and EC were lower at Site 1 than in Site 2 but remained within the same

magnitude order, however both parameters were greater in soils from Sites 3 and 4

(Table 10)..Mean soil pH did not differ among sites but mean WHC increased from 17.3 to

24.8% along the river course. For these physico-chemical parameters, soils from Sites 1 and 2

are similar but differed from soils from Sites 3 and 4.

115

Table 10 : Main characteristics of riverbank soils (n=6), freshwaters (n=6) and soil pore waters (n=4) along the Jalle d’Eysines River. Values are

means ± SD.

soil

siteCoarse sand (200/2000µm)

(g kg -1)*

clays (<2 µm) (g kg -1)*

C (g kg -1)* N (g kg -1)* OM (g kg -1)* C/N** CEC (cmol kg -1)** pH*Conductivity (mS.cm -1)**

w ater holding

capacity (%)*1 667±184.3a 44±15a 6.8±7.6a 0.43±0,44a 11.9±13.1a 12.9±3. 7ab 1.3±0.7a 7.6±0.3ab 0.06±0.01a 17.3±0.4a2 779.3±24.9a 81±19b 11.3±2.9a 0.93±0.22a 19.5±5.1a 12.0±0 .6a 5.2±1.7b 7.5±0.3ab 0.15±0.06b 20.1±0,7b3 71.6±75b 466±94c 28.9±6.8b 2.6±0.5b 50±11.7b 11.0±1.1ab 2 6.3±2.9c 7.1 ±0.5a 0.34±0.17bc 23.8±1,2c4 58.3±48.3b 383±28c 26.3±6.8b 2.4±0.53b 45.5±11.7b 10.7±0 .8b 26.7±4.6c 7.9±0.3b 0.47±0.03c 24.8±1c

freshwater pore w ater

siteConductivity (mS.cm -1)**

pH* Eh(mV)* pH*Conductivity (mS.cm -1)*

1 0.28±0.02a 6.8±0.23a 286±2,3a 7.5±0.22a 0.46±0.03ab2 0.41±0.06b 7.0±0.05a 288±11,7a 7.6±0.32a 0.77±0.27ac3 0.39±0.06b 7.5±0.17b 269±2,1b 7.3±0.48a 0.33±0.05b4 0.94±0.54b 7.6±0.13b 264±2,1b 7.9±0.18a 1.06±0.20c

* : The different letters stand for statistical significance at the 0.05 level with Tukey HSD test, ** : The different letters stand for statistical significance at the 0.05 level with Wilcoxon pairwise test adjusted with a Bonferroni correction.

CEC = cationic exchange capacity; Eh=redox potential; OM = organic matter

Table 11 : Total element concentrations in freshwater samples from the Jalle d’Eysines River (n=6). Values are means ± SD.

site Zn Cu Mo Cd Mg Cr Ni Pb Fe Ca As P Mn Kµg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1 µg L -1

1 <7 <8 <0.4 <0.1 5.2±0,4 <0.3 1.3±0.9 <0.8 28.2±7.1 30.6±0.4 0 .7±0.05 <0.2 <20 3.9±0.042 <7 <8 0.5±0.09 <0.1 5.2±0.1 <0.3 2±0.3 <0.8 35.7±11.4 50.4±6 .1 0.8±0.1 0.3±0.01 <20 6.8±0.033 <7 <8 8.6±0.6 <0.1 4.9±0.5 <0.3 2.2±0.2 <0.8 36.1±5.9 46.5±2 .1 1.2±0.3 0.3±0.06 <20 7.8±0.114 7.1±0.5 <8 7.2±0.6 <0.1 10±0.03 <0.3 2.6±0.2 <0.8 36,6±1.1 5 4.3±0.18 1.8±0.2 0.2±0.01 <20 9.6±0.08

Median values in surface water for

French rivers *2.3-3.3 0.9-1.1 0.12-0.14 0.01-0.02 4.6-5.1 0.22-0.25 1.6-2.5 0.09-0.2 24-118 11.9-100 0.7-1.8 - 11.8-12.8 1.7-2.3

*Salpeteur and Angel (2010)

116

Total soil element concentrations were also generally higher in soils from Sites 3 and 4 than

in soils from Sites 1 and 2 (Table 12). Concentrations of PTTE in soils from 1 and 2 are in

compliance with standard PTTE values defined by the French Water Agency for a good

riverbank soil quality (e.g. <31 mg Cu and <120 mg Zn kg-1 soil DW, SEQ EAU, 2003) and

similar to the median PTTE concentrations in French riverbank soils (Salpeteur and Angel.,

2010). Total soil PTTE at Sites 3 and 4 generally exceeded both standard and median PTTE

concentrations for French riverbank soils (Table 12). Soils increased in PTTE concentrations

from the source to the confluence and soils from Sites 1 and 2 were relatively unpolluted

compared to soils from Sites 3 and 4.

In soil pore water, mean pH did not differ among sites whereas mean EC values were site

dependent. Mean Eh was higher at Sites 1 and 2 (286 and 288 mV) than at Sites 3 and 4 (269

and 264 mV). Element concentrations in pore water did not follow the same pattern as

reported for total element concentrations in soils. Soil pore water had the highest

concentrations of Cu, Cr, Ni, Mn, Fe, P, K and As at Site 2, not Site 4 (Table 12). However,

Site 4 had the highest concentrations of Ca and Mg. The lowest element concentrations in soil

pore water were found at Site 1 for Cu, Ni, As and Mo and at Site 3 for Ca, Mg, P, K

(Table 12). Zinc concentrations in soil pore water remained below detection limits (<7 µg L-1)

at all sites. No increase in element concentration was observed in soil pore water similar to the

gradient reported for total elements in soils along the river course (Table 12).

3.2 Foliar element concentrations of macrophytes

No symptoms of elemental toxicity such as spotted necrosis and chlorosis were visible on any

macrophytes samples. Foliar element concentrations of macrophytes were generally the

highest in the hemicryptophytes L. salicaria, R. acris and P. arundinacea. Copper

concentrations ranged between 3.85 mg kg-1 at Site 1 for J. effusus and 20.4 mg kg-1 at Site 1

for L. salicaria. Zinc concentrations were generally the lowest in C. riparia and I.

pseudacorus and were highest in L. salicaria at Site 1. Molybdenum peaked in

P.,arundinacea below the WTP L’Ile (up to 2.55 mg kg -1 at Site 4) while the overall highest

concentrations were found in R. acris at Site 2 (3.35 mg kg-1).

117

Table 12 : Total element concentrations in riverbank soils (n=6) and soil pore waters (n=4) along the Jalle d’Eysines River. Values are means ±

SD.

site Total TE concentration in soil

Cu** Zn** Cr* Ni* Co** Pb* Cd** Mn** Mo*

mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1

1 2.5 ±1.8 a 19.3 ± 13.4 a 10.9 ± 4.6 a 2.6 ± 1.2 a 1.7 ± 1 a 6.9 ± 4.5 a 0.11 ± 0.09 a 73.6±37.9a 0.12±0.09a2 8.6 ± 6.7 a 16.1 ± 5.9 a 17.4 ± 7.1 a 3.9 ± 2.3 a 1.6 ± 0.8 a 11.5 ± 1.8 a 0.09 ± 0.03 a 84.4±5a 0.27±0.21a3 32.6 ± 3.6 b 171.3 ± 18.1 b 79.6 ± 10.9 b 40.1 ± 2.7 b 18.9 ± 2.5 b 54.9 ± 7.3 b 0.47 ± 0.08 b 804.7±345.1b 1. 6±0.26b4 39.8 ± 4.4 b 274.2±52.8c 85.3 ± 9.1 b 39.1 ± 1.6 b 16.1 ± 0.4 b 70.1 ± 5.1 c 1.6 ± 0.54 c 767.8±107.5b 0.95±0. 1c

Al* Fe** Ca** K** Mg** Na* P* pcu 2+**

g kg -1 g kg -1 g kg -1 g kg -1 g kg -1 g kg -1 g kg -1

1 9 ± 5.4 a 3.7 ± 1.8 a 1.2 ± 0.8 a 3.9 ± 1.9 a 0.4 ± 0.2 a 1±0.5a 0.05±0.04a 12.4±0.6a2 10.5 ± 5.3 a 4.6 ± 2.5 a 2.9 ± 1.3 b 3.2 ± 1.1 a 0.4 ± 0.1 a 0.6±0.1a 0.08±0.03ab 11.4±0.6a3 85 ± 5.9 b 48 ± 10 b 6.4 ± 1.5 c 21.7 ± 1.5 b 8.6 ± 0.6 b 4.3±0.4b 0.14±0.05bc 10.1±0.09b4 82 ± 5.1 b 43 ± 1.7 b 16 ± 4.5 d 22.5 ± 1.6 b 10.5 ± 1. 3 b 4.5±0.6b 0.18±0.06c 9.9±0.08b

TE Concentration in pore water

Cu** Zn Cr Ni Cd Mn Fe Ca Mgµg l -1 µg l -1 µg l -1 µg l -1 µg l -1 µg l -1 µg l -1 mg l-1 mg l-1

1 11,1 ± 1,8 a < 7 1.9 ± 0.7 a 3,1 ± 0.07 a 0,17 ± 0.01 < 20 85 ± 53 58,3 ± 0.6 a 6,3 ± 0.5 a2 19,6 ± 7.4 a < 7 2 ± 0,6 a 5,3 ± 1,7 b < 0,1 576 ± 404 2012 ± 1689 79,6 ± 34 ac 8,3 ± 4,9 a3 12,4 ± 3.1 a < 7 0,8 ± 0,6 b 4,3 ± 1,3 ab < 0,1 576,5 ± 284 < 20 32,3 ± 5,8 b 4,9 ± 1,8 a4 15,9 ± 4.2 a < 7 0,7 ± 0,3 b 3,5 ± 0,9 a 0,12 ± 0,01 < 20 < 20 92,8 ± 16,9 c 16,4 ± 3,6 b

P** K** As** Mo**mg l-1 mg l-1 µg l -1 µg l -1

1 0,2 ± 0,18 a 8,1 ± 1,9 a 2,4 ± 0,3 a 6,6 ± 5 a2 1,1 ± 0.3 b 72,7 ± 31,2 b 17,9 ± 17,1 ab 10,5 ± 11,1 a3 0,11 ±0,18 a 2,9 ± 1,9 c 3,3 ± 3,1 a 14,9 ± 16 a4 0,42 ± 0,05 a 10,4 ± 3,1 a 6,5 ± 0,6 b 7,4 ± 2,6 a

* : The different letters stand for statistical significance at the 0.05 level with Tukey HSD test, ** : The different letters stand for statistical significance at the 0.05 level with

Wilcoxon pairwise test adjusted with a Bonferroni correction. Total soil elements were determined after wet digestion in fluorhydric acid NF X 31147); pCu2+ =3.20 +1.47

pH -1.84 log10 (total soil Cu) (Sauvé, 2003)

118

Foliar Mo concentration in C. riparia gradually increased along the river course (Table 13).

Juncus effusus, I. pseudacorus and L.salicaria had the highest Cd concentrations

(respectively, 0.79 mg kg-1 at Sites 2 and 3, 1.13 mg kg-1 at Site 1 and 0.65 mg kg-1 at Site 1).

Foliar Mn concentrations generally were the highest in L. salicaria (439 mg kg-1 at Site 2) but

did not differ across sites. Increases in foliar Fe concentrations in C. riparia, L. salicaria and

R. acris and in foliar Pb concentrations in P. arundinacea reflected similar increases of total

soil Fe and Pb concentrations along the river course (Table 13). Concentrations of K and P

were higher in R. acris and L. salicaria (32 g K and 4.5 g P kg-1 for R. acris and 37.3 g K and

5.1 g kg-1 for L. salicaria, respectively). Calcium and Mg concentrations peaked for I.

pseudacorus and L. salicaria at Site 1.

3.3 Modelling of macrophyte exposure by discriminant analysis of foliar PTTE

concentrations

To assess environmental contamination and exposure we ran three LDA models for this study

(Table 14). The first LDA was based on foliar PTTE concentrations of the whole macrophyte

dataset as the macrophyte community. The second only considered the foliar PTTE

concentrations in the rhizomatous-geophytes while the third was based on the foliar PTTE

concentrations in the hemicryptophytes. The first model well identified the sampling

locations, and consequently the PTTE exposure in the soil and soil pore water for 70.3% of

the 116 individuals of the original training dataset. The external cross validation, based on 38

individuals, correctly identified the sampling location of macrophytes for 63% of the testing

dataset. The second LDA model based on rhizomatous-geophytes successfully identified the

sampling location for 82.5% of the 65 individuals in the original training dataset. After

external cross validation, the sampling site of 80% of the test dataset (20 individuals) was

correctly identified. The best discrimination occurred for the LDA model of the

hemicryptophyte group, which well predicted the sampling location of 92.3% of the 51

individuals of the original dataset and is validated for 88.8% of the 18 individuals of the test

dataset by the external cross validation. The first axis of the LDA, when considering the

hemicryptophytes, was mainly driven by Zn, P, Mn and Mg foliar concentrations, while the

second axis was mostly driven by Cd, As, Cr and also Mg foliar concentration

119

Table 13 : Foliar element concentrations of macrophytes along the Jalle d’Eysines River (n=6). Values are means ± SD.

* : The different small letters stand for statistical significance at the 0.05 level with Tukey HSD test, ** : The different small letters stand for statistical significance at the 0.05 level with Wilcoxon pairwise test adjusted with a Bonferroni correction .

As Cd Zn Cr Cu Fe Pb Mn Mo Ni Mg Ca K P

mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW g kg -1 DW g kg -1 DW g kg -1 DW g kg -1 DW

site ** * ** ** ** ** ** ** ** ** * * * *

Juncus effusus 1 0.21±0.09 ab 0.37±0.11 a 33±6.2 a 0.46±0.18 a 3.8±1.2 a 36.7±8.9 a 0.98±0.35 a 73.1±20.1 a 1.16±0.35 a 1.24±0.71 a 0.9±0.2 a 2.2±0.4 a 17.6±4 a 1.8±0.4 a

2 0.10±0.03 a 0.79±0.21 b 37.1±2.6 a 0.58±0.13 a 12.4±3.2 b 62.9±39.8 ab 1.17±0.87 a 124.7±22.1 b 0.44±0.09 b 0.96±0.23 a 0.9±0.2 a 2.9±1.1 ab 21.3±2.3 a 2.2±0.2 a

3 0.17±0.04 b 0.79±0.17 b 42.1±10.8 a 0.62±0.05 a 8.2±1.1 c 84.5±36.6 b 1.04±0.61 a 262.9±69.7 c 1.68±0.35 c 0.83±0.16 a 1.4±0.2 b 3.4±0.7 b 16.8±2.5 a 1.8±0.2 a

4

** ** * ** ** ** ** ** * ** ** ** * **

Phragmites

australis1 0.18±0.01 a 0.05±0.01 a 43.4±2.1 a 0.62±0.05 a 19.4±0.7 a 134.5±5.2 a 1.11±0.27 a 86.8±3.3 a 0.68±0.30 ab 1.39±0.99 a 1.4±0.07 a 3.9±0.2 a 22.1±0.6 a 1.9±0.06 a

2 0.19±0.04 a 0.05±0.04 a 24.3±3.7 b 0.75±0.33 a 11.1±2.4 b 101.8±23.4 a 1.30±0.62 a 93.5±.1 a 1.29±0.76 b 0.92±0.21 a 1.1±0.8 a 3.3±1.6 a 20.8±2.9 ab 2.3±0.2 ab

3 0.37±0.33 a 0.21±0.35 a 29.2±6.8 b 0.92±0.41 a 16.1±3.8 a 140.1±23.4 a 1.02±0.56 a 119.2±35.8 a 2.6±0.53 c 0.92±0.43 a 1±0.1 a 4.9±1.1 a 17.7±2.9 b 1.9±0.3 a

4 0.22±0.07 a 0.22±0.14 a 34.2±10.8 ab 0.73±0.16 a 10.8±1.2 b 143.4±24.8 a 0.92±0.27 a 124.3±65.1 a 1.66±0.43 bc 0.59±0.09 a 1.5±0.3 a 4.9±1.1 a 19.5±2.9 ab 2.5±0.5 b

** ** * ** ** ** ** ** ** ** ** * * **

Phalaris

arundinacea1 0.12±0.01 a 0.08±0.01 a 53.6±2.3 a 0.66±0.01 a 7.4±0.5 a 73±1.8 a 0.29±0.02 a 92.5±7.3 a 0.39±0.01 a 1.09±0.01 a 4.2±0.2 a 6.6±0.5 a 14.5±0.9 a 2.8±0.02 ab

2 0.15±0.08 a 0.01±0.01 b 29.2±6.2 b 0.64±0.07 a 10.2±2.1 b 97.8±7.2 b 0.77±0.2 ab 63.7±23.2 a 0.42±0.03 a 0.51±0.01 b 1.2±0.2 b 4.2±0.7 a 29.2±7.1 b 3.2±0.7 a

3 0.20±0.02 a 0.02±0.02 b 24.9±3.9 b 0.85±0.32 a 6.3±0.5 a 102.1±21.8 b 1.42±1.1 b 67.6±17.1 a 2.03±0.04 a 0.67±0.01 c 1.3±0.1 b 3.7±0.4 b 26.3±1.2 b 2.8±0.2 ab

4 0.22±0.10 a 0.07±0.06 ab 31±2.4 b 0.60±0.14 a 11.9±2.2 b 98.7±10.2 b 1.62±0.44 b 93.1±57.4 a 2.55±1.8 a 0.53±0.01 bc 1.1±0.2 b 4.7±0.7 b 17.8±4.2 a 2.2±0.2 b

* ** ** ** ** ** ** ** ** ** ** * ** **

Carex riparia 1 0.09±0.03 a 0.12±0.01 a 19.2±3 a 0.81±0.2 a 10.2±1.1 a 65.4±3.8 a 0.83±0.21 a 57.9±2.4 ab 0.45±0.1 a 0.69±0.2 a 1.4±0.05 a 3.7±0.5 ab 14.9±0.5 a 1.1±0.04 a

2 0.13±0.02 a 0.10±0.01 a 14.5±1.2 a 0.53±0.1 b 8.2±0.5 a 83.7±2.7 a 0.86±0.72 a 21.6±0.6 a 0.41±0.01 a 0.71±0.1 a 0.9±0.03 b 2.8±0.1 a 15.8±0.6 a 1.3±0.04 a

3 0.11±0.03 a 0.13±0.04 ab 16.1±2.6 a 0.78±0.2 a 8.6±2.1 a 122.1±78 a 0.83±0.35 a 40.1±34.7 ab 0.69±0.5 a 0.75±0.1 a 1±0.2 bc 3.6±0.8 ab 16.3±2.9 a 1.3±0.3 a

4 0.24±0.03 b 0.18±0.03 b 22.2±8.7 a 1.18±0.1 c 13.3±2.5 b 345.8±78.6 b 1.39±0.35 a 59.2±18.5 b 1.65±0.1 b 1.4±0.8 b 1.2±0.1 ac 4.5±0.9 b 22.3±2.8 b 2±0.2 b

** ** ** ** ** ** ** ** ** ** ** * ** *

Ranunculus

acris1 0.16±0.03 a 0.25±0.02 a 54.9±2.2 a 0.67±0.1 a 16.4±0.4 a 95.2±3.6 a 1.06±0.32 a 29.3±0.7 a 2.25±0.1 ab 0.97±0.1 a 2.7±0.06 a 12.4±0.3 a 29.2±0.6 a 2.7±0.07 a

2 0.13±0.02 a 0.09±0.05 b 44.6+±8 a 0.83±0.2 ab 15.8±4.9 a 145.4±53.6 a 1.43±1.33 a 81.7±44.4 a 3.35±1.4 a 1.31±0.6 ab 1.6±0.6 b 12±3.1 a 30.8±4.9 a 3.9±0.7 ab

3 0.22±0.04 a 0.30±0.05 a 37.5±6.7 a 0.89±0.2 ab 16.5±4.2 a 184.9±71.5 a 1.06±0.14 a 83.5±28 a 1.87±0.5 bc 2.06±0.4 c 2.2±0.3 ab 10.6±1.6 a 32.3±5.9 a 3.8±0.9 ab

4 0.37±0.15 b 0.24±0.1 a 44.6±2.9 a 1.29±0.5 b 16.9±7.4 a 398.1±247.5 b 1.34±0.66 a 51.2±19 a 1.23±0.5 c 1.69±0.5 bc 3.8±1 c 9.4±2.5 a 30.6±3.6 a 4.5±0.4 b

* ** ** * ** ** ** ** ** ** ** ** ** **

Iris

pseudacorus1 0.07±0.01 a 1.13±0.04 a 23.5±1 a 0.56±0.1 a 8.1±0.5 a 75.1±5.2 a 1.06±1.01 a 37.6±2.3 a 0.38±0.01 a 1.09±0.01 a 2±0.09 a 21.1±0.8 a 28.9±0.6 a 2.9±0.09 a

2

3 0.09±0.03 a 0.15±0.05 b 13.6±4.7 b 0.63±0.1 ab 6.1±1 b 67.1±19.5 a 1.32±1.25 a 182.6±46.9 b 2.08±0.6 b 0.52±0.1 b 2.4±0.2 b 16.4±1.1 b 16.1±4 b 1.8±0.3 b

4 0.11±0.03 a 0.29±0.16 b 15.5±2.4 b 0.67±0.1 b 8.9±1.7 a 78.2±18.1 a 0.54±0.32 a 76.7±26.1 c 1.41±0.6 c 0.57±0.1 b 2.9±0.2 c 18.3±2.7 b 20.6±5.3 b 2.3±0.4 b

** ** ** * ** ** ** ** ** ** ** ** ** **

Lythrum

salicaria1 0.08±0.01 a 0.65±0.02 a 96.4±2.8 a 0.68±0.1 a 20.4±0.5 a 126.6±2.3 a 0.51±0.08 a 304.3±19 a 2.54±0.1 a 1.29±0.2 a 7.9±0.2 a 17.3±0.3 a 25.7±0.6 a 5.1±0.1 a

2 0.09±0.03 ab 0.05±0.02 b 54.4±6.8 ab 0.91±0.3 ab 12.8±0.7 b 132±36.3 a 0.96±0.36 a 439.6±117.8 a 0.43±0.1 b 0.51±0.01 b 3.5±0.9 b 8.7±1.7 b 37.3±2.4 b 4.2±0.5 b

3 0.18±0.04 bc 0.08±0.05 b 39±7.1 b 0.86±0.1 ab 12.1±2.4 b 168.5±50.6 ab 1.43±1.36 a 337.7±81.7 a 2.87±1.4 a 0.65±0.1 b 4.6±1 bc 9.8±1.7 b 30.4±3.9 bc 3.5±0.7 b

4 0.30±0.08 c 0.05±0.02 b 37.9±4 b 0.98±0.1 b 11.1±1.6 b 214.2±74.1 b 0.67±0.09 a 374±107.7 a 1.83±0.2 a 0.71±0.2 b 5.7±0.7 c 10.2±1.5 b 25.7±4.1 ac 3.6±0.5 b

120

Table 14 : Discriminant analysis of studied macrophyte communities with standardized canonical function coefficients, the number of observations (n)

and percent correctly classified of original grouped cases and of cross validated cases (%). LD1, 2 and 3 are the three discriminant functions retained in

this LDA.

Macrophyte Community Rhizomatous-Geophytes Hemicrypto phytes

Approx Wilks’ lamba 0,21 0.088 0,032

p-value < 0,001 *** < 0,001 *** < 0,001 ***

LD1 LD2 LD3 LD1 LD2 LD3 LD1 LD2 LD3

coefficients of Linear discriminants

0,62 0,28 0,09 0,56 0,26 0,17 0,52 0,29 0,19

As -0,24 0,93 0,19 -0,49 0,94 0,48 0,08 0,8 1,44

Cd 0,09 0,15 -0,54 1,04 0,25 -0,28 0,67 1,39 -1,27

Ca -0,33 0,47 0,35 -1,43 0,98 1,3 0,16 -0,35 1,47

Cr 0,52 -1,09 -0,52 0,35 -0,95 -1,02 0,33 -0,72 -0,21

Cu 0,04 -0,19 0,29 0,53 -0,02 -0,73 0,07 -0,51 1,09

Fe 0,34 0,75 0,52 0,74 -0,27 0,96 0,03 0,37 0,22

Mg -0,43 1,06 -0,21 1,2 -0,62 0,36 -0,93 1,02 -0,81

Mn 0,79 -0,27 -0,07 -0,06 0,63 -0,61 1,05 0,39 -0,26

Mo 0,74 0,18 -0,76 1,08 0,57 0,26 0,3 0,15 -1,11

Ni -0,14 0,14 0,01 -0,13 -0,18 0,04 -0,45 0,17 -1,38

P 1,28 -0,98 0,83 1,02 -1,74 -1,38 1,13 -0,63 0,97

K -0,7 -0,14 0,46 -1,31 0,89 0,69 -0,59 -0,27 -1,13

Pb -0,05 0,11 0,24 0,01 -0,17 0,07 0,03 0,24 0,31

Zn -1,53 0,08 -0,18 -1,22 0,28 1,52 -2,37 -0,69 0,63

n 116 38 65 20 51 18

Correctly classif ied Correctly classif ied Correctly classif ied

Original testing dataset Original testing data set Original testing dataset

% 70,4 63 82,5 80 92,3 88,8

121

IV. Discussion

4.1 Copper and Zn in soils, soil pore water and macrophytes

Foliar PTTE concentrations of investigated macrophytes along the Jalle d’Eysines River depended on

both plant species and soil physico-chemical properties such as total PTTE concentrations in soils,

PTTE concentrations in soil pore water and the soil texture at the sampling site. Total soil Cu and Zn

increased in riverbank soils between the source and the confluence. However, none of the seven

macrophytes studied had a similar increase in foliar Cu and Zn concentrations along the river.

Translocation of both of these metals from soils to the leaves of macrophytes was not evident.

Conversely, species such as P. australis, P. arundinacea, R. acris and especially L. salicaria, had the

highest foliar Zn concentrations at the relatively uncontaminated Site 1. Similarly, foliar Cu

concentrations did not directly reflect Cu concentration in pore water along the Jalle d’Eysines River,

except for P. arundinacea (Table 13). Carex riparia, I. pseudacorus, J. effusus, R. acris, P. australis,

P. arundinacea and L. salicaria may act as Cu and Zn excluders along the Jalle d’Eysines River. It is

known that species tolerant to PTTE contamination such as P. australis are generally able to segregate

PTTE in the root cortical tissue outside the endodermis, thereby preventing or reducing translocation

to other plant parts (Verkleij and Schat, 1990; Ye et al., 1997; Baldantoni et al., 2009). As a result,

leaves of these seven macrophytes cannot be considered good biomonitors of total Cu and Zn

accumulation in soils. Such weak correlations between foliar Cu and Zn concentrations and their

contents in the soil agreed with Romero-Nùñez et al. (2011) who concluded that metal concentrations

in plants depend on several interacting biotic and abiotic factors (i.e. element phytoavailability,

translocation rates, element affinity for adsorption sites in both soils and plant tissues, synergies and

antagonisms during root uptake and translocation, and endophytic activity) rather than only on soil

concentrations. However, these results contrast with the findings of Bonnano and Lo Giudice (2010),

which described a high correlation (R²=0.93) between foliar Cu concentrations in P. australis and total

Cu concentrations in riverbank soils.

4.2 Chromium, Fe, Mo, Cd and Pb in soils, soil-porewater and macrophytes

In this study, total soil Cr increased from the source to the confluence. Conversely, Cr concentration in

soil pore water decreased along the river course. These results may be explained by the soil texture,

which progressively increased in clay contents from a sandy-loam soil to silty-clay textures between

122

the source and the confluence. Clay minerals form organo-mineral complexes with metal oxides and

humic substances, which present peculiar sorption capacities different from those of each single soil

constituent (Violante and Pigna, 2008). In macrophytes, no or slight variation in foliar Cr

concentration was detected across sites. Caldelas et al. (2012) reported that roots and rhizomes of I.

pseudacorus have distinct functions in response to Cr concentrations. The rhizodermis limits Cr uptake

by means of Si deposition and cell wall thickening while the rhizome cortex generates vacuoles where

Cr is sequestered by metal-binding compounds such as metallothioneins and phytochelatins. Along the

Jalle d’Eysines River, Cr concentrations in soil bearing phases were low and macrophytes, whether

hemicryptophytes or rhizomatous geophytes, evenly trapped Cr in their belowground biomass. A

contradictory observation was made for R. acris, C. riparia and L. salicaria whose foliar Fe

accumulation along the Jalle d’Eysines River reflected that of total soil Fe but not the strong Fe

decrease in soil pore water along the river course. These results may also be explained by changes in

the soil texture and adaptation of plant metabolism and rhizosphere functioning. Carex riparia, P.

australis and P. arundinacea transferred more Mo in their leaves at sampling sites located below the

second WTP “L’Ile”, which reflect total Mo accumulation in soil below the WTP while Mo

concentrations in soil pore water remained stable between the river source and the confluence. These

results are in compliance with Bonanno (2011) who reported that Mo is readily taken up by P.

australis (root/soil ratio TF=0.87). Phalaris arundinacea and P. australis respectively accumulated Cd

and Pb in theirleaves along the river, which present a Cd and Pb accumulation between the source and

the confluence. These results contrast with Baldantoni et al. (2009) but agree with Peltier et al. (2003),

i.e. Pb might translocate from belowground to aboveground biomass of P. australis. Foliar element

accumulation in macrophytes of the Jalle d’Eysines River as the expression of total element

concentration in riverbank soils occurred for some PTTE but it is not a rule for all. We suggest the

global PTTE exposure at the Jalle d’Eysines River results from total PTTE concentration in the soil,

PTTE concentrations in the soil pore water, soil texture, and the kinetics of reactions between soil

bearing phases and the soil pore water.

4.3 Transfers from roots and rhizomes to leaves in hemicryptophyte and rhizomatous-geophytes

Romero-Nùñez et al. (2011) summarized that PTTE translocation from roots into the leaves of

macrophytes depends on physiological strategy to cope with PTTE exposure. Mechanisms driving

PTTE transport in plants have been previously reviewed by Verbruggen et al. (2009); Palmer and

Guerinot (2009); Sharma and Dietz (2010) and Pal and Rai (2010). Among them, the belowground

123

biomass of macrophytes retains PTTE and buffers their transfer to aboveground biomass by binding

them to the cell wall apoplast and trapping them into the vacuole. Rhizomatous- geophytes, which

produce high belowground biomass, notably rhizomes, may be more efficient in limiting PTTE

transfer to aboveground biomass than hemicryptophytes. Our results support these findings. At each

site, the highest foliar PTTE concentrations were generally found in the hemicryptophytes L. salicaria

and R. acris whereas the lowest ones occurred in rhizomatous-geophytes I. pseudacorus and J. effusus.

4.4 Biomonitoring of plant exposure to PTTE by using foliar PTTE concentrations in macrophytes

in a linear discriminant analysis (LDA)

The discrepancy between Bonanno and Lo Giudice (2010) and our results suggest the necessity to first

establish a relevant model of exposure-plant responses when using macrophytes as biomonitors. A

linear regression model between foliar PTTE concentrations of macrophytes and total element

concentrations in bearing phases is more likely at homogenous sites with small changes in soil

parameters and contamination. Along rivers, and more generally, in large wetlands with high biotic

and abiotic heterogeneity, a multivariate approach might be preferred to the study of each element

taken singly for biomonitoring PTTE exposure. Nummelin et al. (2007) successfully used this

approach with a LDA based on PTTE concentrations in predatory insects. For routine assessments of

the overall PTTE exposure of organisms living in contaminated soil riverbanks we suggest the use of a

LDA model based on foliar element concentrations in several macrophyte species. First, well

developed and not senescent leaves of macrophytes at the peak of the growing season are sampled.

Then, foliar macronutrient and PTTE concentrations are determined. Data are standardized and split in

two datasets. The original dataset is used to build a LDA for discriminating the sampling sites while

the training dataset is used for testing model reliability by correctly assigning macrophyte sampling

location in a cross validation. Model reliability depends on the selected biomonitor, which has to be

able to perfectly assign macrophyte location after the external cross validation. The model built for our

study allowed to correctly classify 70% of the whole macrophyte community after external cross

validation. When macrophytes were grouped according to Raunkiær classification, the model

classified after external cross validation respectively 80% of the rhizomatous-geophytes and 89% of

the hemicryptophytes. Hemicryptophytes such as L. salicaria, P. arundinacea and R. acris are the best

biomonitors when building such LDA model for assessing PTTE exposure along the Jalle d’Eysines

River. Further investigations are needed to assess the reliability of these results on other rivers.

124

Linear discriminant analysis can also be used to follow changes in PTTE exposure in the riverbank

soils in relate to elapsed time. The percentage of correct macrophyte assignation obtained at each

sampling site, at the same sampling period but at two different years may not significantly vary if the

exposure does not change over time along the river. In this case the model is built at T0 (Figure 27). A

first cross validation is conducted at T0 and P0 is the percentage of correct macrophyte assignation

along the river at T0. A second cross validation is conducted at T+n and Pn is the percentage of correct

macrophyte assignation along the river at T+n. If at all sites P0=Pn, this means that the exposure along

of the river did not change over time. If Pn significantly differs from P0, this means that the site

exposure changed between sampling years. In this case, if Pn ≥ P0 at sites previously defined as site

with low PTTE exposure while Pn ≤ P0 at sites defined as sites with high PTTE exposure, this means

exposure generally decreased along the river. Conversely, if Pn ≤ P0 at sites previously defined as sites

with low PTTE exposure while Pn ≥ P0 at sites defined as sites with high PTTE exposure, this means

that macrophyte exposure to PTTE has generally increased along the river (Figure 27). In the case

where the overall exposure changed at all sites, the LDA model built at the first sampling year is not

valid. New data are thus needed to build a new LDA for re-discriminating sites. A multivariate

approach allows better assessing macrophyte exposure to PTTE over space and time because it

integrates overall synergisms and antagonisms occurring in the ecosystems in terms of PTTE transfer.

Figure 27 : Tree decision : Biomonitoring of plant exposure to PTTE over time by using

foliar PTTE concentrations in mac

percentage of correct macrophyte assignation at each sampling site at T0,

macrophyte assignation at each sampling site at T+n

125

Tree decision : Biomonitoring of plant exposure to PTTE over time by using

concentrations in macrophytes in a linear discriminant analysis (LDA).

percentage of correct macrophyte assignation at each sampling site at T0, Pn = percentage of correct

macrophyte assignation at each sampling site at T+n

Tree decision : Biomonitoring of plant exposure to PTTE over time by using

rophytes in a linear discriminant analysis (LDA). P0 =

= percentage of correct

126

Acknowledgements

This work was financially supported by AXA foundation (PhD grant of L. Marchand),

ADEME, Department Urban Landfills and Polluted Sites, Angers, France, and the European

Commission under the Seventh Framework Programme for Research (FP7-KBBE-266124,

GREENLAND). The authors thank Jean Paul Maalouf for his advice in statistics.

References (Format: Freshwater Biology)

Adriano, D.C. (2001) Trace elements in terrestrial environments: biogeochemistry, bioavailability, and risks of metals. 2nd ed. Springer. Verlag.

Agence de l’eau (2003) Systeme d’evaluation de la qualité de l’eau et des cours d’eau. Grilles d’évaluation Seq-eau (version 2). France. http://sierm.eaurmc.fr/eaux-superficielles/fichiers-telechargeables/grilles-seq-eau-v2.pdf

Aguiar F.C., Ferreira M.T., Albuquerque A., Rodríguez-González P. & Segurado P. (2009) Structural and functional responses of riparian vegetation to human disturbance: performance and spatial-scale dependence. Fundamental and Applied Limnology 175, 249-267.

Aksoy, A., Demirezen, D. & Duman, F. (2005) Bioaccumulation, detection and analyses of heavy metal pollution in Sultan Marsh and its environment. Water Air and Soil Pollution, 164, 241-255.

Archaimbault, V., Usseglio-Polatera, P., Garric, J., Wasson, J.G., Babut., M. (2010) Assessing pollution of toxic sediments in streams using bio-ecological traits of benthic macroinvertebrates. Freshwater Biology, 55, 1430-1446.

Baldantoni, D., Alfani, A., Di Tommasi, P., Bartoli, G. & De Santo, A.V. (2004) Assessment of macro and microelement accumulation capability of two aquatic plants. Environmental Pollution, 130, 149-156.

Baldantoni, D., Ligrone, R. & Alfani, A. (2009) Macro- and trace-element concentrations in leaves and roots of Phragmites australis in a volcanic lake in Southern Italy. Journal of Geochemical Exploration, 101, 166-174.

Basol (2012) Base de données BASOL sur les sites et sols pollués appelant à une action des pouvoirs publics à titre préventif ou curatif. France. http://basol.environnement.gouv.fr/fiche.php?page=1&index_sp=33.0086

Bonanno, G. (2011) Trace element accumulation and distribution in the organs of Phragmites australis (common reed) and biomonitoring applications. Ecotoxicology and Environmental Safety, 74, 1057-1064.

Bonanno, G. & Lo Giudice, R. (2010) Heavy metal bioaccumulation by the organs of Phragmites australis (common reed) and their potential use as contamination indicators. Ecological Indicators, 10, 639-645.

Bonanno, G. (2012). Arundo donax as potential biomonitor of trace element contamination in water and sediments. Ecotoxicology and Environmental Safety, 80, 20-27.

Bragato, C., Brix, H. & Malagoli, M. (2006) Accumulation of nutrients and heavy metals in Phragmites australis (Cav.) Trin. ex Steudel and Bolboschoenus maritimus (L.) Palla in a constructed wetland of the Venice lagoon watershed. Environmental Pollution, 144, 967-975.

Caldelas, C., Bort, J., Febrero, A. (2012) Ultrastructure and subcellular distribution of Cr in Iris pseudacorus L. using TEM and X-ray microanalysis. Cell Biology and Toxicology, 28, 57-68.

Cardwell, A.J., Hawker, D.W. & Greenway, M. (2002) Metal accumulation in aquatic macrophytes from southeast Queensland, Australia. Chemosphere, 48, 653-663.

127

Clemens, S., Palmgren, M.G. & Kramer, U. (2002) A long way ahead: understanding and engineering plant metal accumulation. Trends in Plant Science, 7, 309-315.

Demars, B.O.L. & Edwards, A.C. (2009) Distribution of aquatic macrophytes in contrasting river systems: A critique of compositional-based assessment of water quality. Science of the Total Environment, 407, 975-990.

Demirtas, H., Freels, S.A., Yucel, R.M. (2008) Plausibility of multivariate normality assumption when multiply imputing non-Gaussian continuous outcomes: a simulation assessment. Journal of Statistical Computation and Simulation, 78, 69-84.

Deng, H., Ye, Z.H. & Wong, M.H. (2004) Accumulation of lead, zinc, copper and cadmium by 12 wetland plant species thriving in metal-contaminated sites in China. Environmental Pollution, 132, 29-40.

European Framework Water Directive. (2000) Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the field of water policy. Official Journal of the European Communities. 43, L327, 1-73.

Feio, M.J., Aguiar, F.C., Almeida, S.F.P., Ferreira, M.T. (2012) AQUAFLORA: A predictive model based on diatoms and macrophytes for streams water quality assessment. Ecological Indicators, 18, 586-598.

Franzle, S. & Markert, B. (2007) Metals in biomass - From the biological system of elements to reasons of fractionation and element use. Environmental Science and Pollution Research, 14, 404-413.

Hinfray, N., Palluel, O., Piccini, B., Sanchez, W., Ait-Aissa, S., Noury, P., Gomez, E., Geraudie, P., Minier, C., Brion, F. & Porcher, J.M. (2010) Endocrine disruption in wild populations of chub (Leuciscus cephalus) in contaminated French streams. Science of the Total Environment, 408, 2146-2154.

Institut National de la Recherche Agronomique Laboratoire d’Analyses des Sols (INRA LAS). 2007. Méthodes employées au laboratoire (in French).

Kissoon, L.T., Jacob, D.L. & Otte, M.L. (2010) Multi-element accumulation near Rumex crispus roots under wetland and dryland conditions. Environmental Pollution, 158, 1834-1841.

Klumpp, A., Bauer, K., Franz-Gerstein, C. & De Menezes, M. (2002) Variation of nutrient and metal concentrations in aquatic macrophytes along the Rio Cachoeira in Bahia (Brazil). Environment International, 28, 165-171.

Kropfelova, L., Vymazal, J., Svehla, J. & Stichova, J. (2009) Removal of trace elements in three horizontal sub-surface flow constructed wetlands in the Czech Republic. Environmental Pollution, 157, 1186-1194.

Labadie, P. & Budzinski, H. (2005) Determination of steroidal hormone profiles along the Jalle d'Eysines River (near Bordeaux, France). Environmental Science & Technology, 39, 5113-5120.

LA CUB (2006) PLU.1. Plan locale d’urbanisme de la communauté urbaine de Bordeaux. PLU approuvé par délibération du conseil CUB du 21/07/2006. Rapport de présentation – A1: L’état initial de l’environnement (in French). Bordeaux, France.

Ladislas, S., El-Mufleh, A., Gerente, C., Chazarenc, F., Andres, Y. & Bechet, B. (2012) Potential of aquatic macrophytes as bioindicators of heavy metal pollution in urban stormwater runoff. Water Air and Soil Pollution, 223, 877-888.

Lanson, B., Marcus, M.A., Fakra, S., Panfili, F., Geoffroy, N. & Manceau, A. (2008) Formation of Zn-Ca phyllomanganate nanoparticles in grass roots. Geochimica et Cosmochimica Acta, 72, 2478-2490.

Lenoir, A & Coste, M. (1996) Development of a practical diatom index of overall water quality applicable to the French national water Board network. - In: Whitton, B.A. and Rott, E. (eds.), Use of Algae for Monitoring Rivers II, Institut für Botanik, Universität Innsbruck, 29-43.

Lizama, A.K., Fletcher, T.D. & Sun, G.Z. (2011) Removal processes for arsenic in constructed wetlands. Chemosphere, 84, 1032-1043.

128

Manceau, A., Nagy, K.L., Marcus, M.A., Lanson, M., Geoffroy, N., Jacquet, T. & Kirpichtchikova, T. (2008) Formation of metallic copper nanoparticles at the soil-root interface. Environmental Science & Technology, 42, 1766-1772.

Marchand, L., Mench, M., Jacob, D.L. & Otte, M.L. (2010) Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants and standardized measurements: A review. Environmental Pollution, 158, 3447-3461.

Markert, B. (2007) Definitions and principles for bioindication and biomonitoring of trace metals in the environment. Journal of Trace Elements in Medicine and Biology, 21, 77-82.

Miege, C., Budzinski, H., Jacquet, R., Soulier, C., Pelte, T., Coquery, M. (2011). L’échantillonnage intégratif par POCIS—application pour la surveillance des micropolluants organiques dans les eaux résiduaires traitées et les eaux de surface. Techniques Sciences et Methode, 1–2, 80–94.

Nummelin, M., Lodenius, M., Tulisalo, E., Hirvonen, H. & Alanko, T. (2007) Predatory insects as bioindicators of heavy metal pollution. Environmental Pollution, 145, 339-347.

Nuñez, S.E.R., Negrete, J.L.M., Rios, J.E.A., Hadad, H.R. & Maine, M.A. (2011) Hg, Cu, Pb, Cd, and Zn accumulation in macrophytes growing in tropical wetlands. Water Air and Soil Pollution, 216, 361-373.

Pal, R. & Rai, J.P.N. (2010) Phytochelatins: Peptides involved in heavy metal detoxification. Applied Biochemistry and Biotechnology, 160, 945-963.

Palmer, C.M. & Guerinot, M.L. (2009) Facing the challenges of Cu, Fe and Zn homeostasis in plants. Nature Chemical Biology, 5, 333-340.

Peltier, E.F., Webb, S.M. & Gaillard, J.F. (2003) Zinc and lead sequestration in an impacted wetland system. Advances in Environmental Research, 8, 103–112.

Raunkiær, C. (1934) The life forms of plants and statistical plant geography. Oxford university Press. Oxford.

Salpeteur, I. & Angel, J.M. (2010) Geochemical baseline data for trace elements in surface water and active sediment from French rivers collected by the FOREGS Geochemical Atlas of Europe (I). Environnement Risques & Sante, 9, 121-135.

Sauvé, S. (2003) Modelling trace element exposure and effects on plants. In: Mench M, Mocquot B, editors. Risk assessment and sustainable land management using plants in trace element-contaminated soils. Villenave d'Ornon, France: Centre INRA Bordeaux-Aquitaine, 69-70.

Schneider, S. & Melzer, A. (2003) The trophic index of macrophytes (TIM) – a new tool for indicating the trophic state of running waters. International Review of Hydrobiology, 88, 49–67.

Sharma, S.S. & Dietz, K.J. (2009) The relationship between metal toxicity and cellular redox imbalance. Trends in Plant Science, 14, 43-50.

Soto, D.X., Roig, R., Gacia, E. & Catalan, J. (2011) Differential accumulation of mercury and other trace metals in the food web components of a reservoir impacted by a chlor-alkali plant (Flix, Ebro River, Spain): Implications for biomonitoring. Environmental Pollution, 159, 1481-1489.

Tabachnick, B.G & Fidell, L. (2007) Using multivariate statistics. 5th ed. Allyn & Bacon. Boston

Verbruggen, N., Hermans, C. & Schat, H. (2009) Mechanisms to cope with arsenic or cadmium excess in plants. Current Opinion in Plant Biology, 12, 364-372.

Verkleij, J.A.C. & Schat, H. (1990) Mechanisms of metal tolerance in higher plants. In : Schaw, A.J. (Ed.), Heavy Metal Tolerance in Plants : Evolutionary Aspects. CRC Press, Boca Raton.

Violante, A., Cozzolino, V., Perelomov, L., Caporale, A.G., Pigna, M. (2010) Mobility and bioavailability of heavy metals and metalloids in soil environments. Journal of Soil Science and Plant Nutrition, 10, 268-292.

Vystavna, Y., Huneau, F., Grynenko, V., Vergeles, Y., Celle-Jeanton, H., Tapie, N., Budzinski, H., & Le Coustumer, P. (2012) Pharmaceuticals in rivers of two Regions with contrasted socio-economic

129

conditions: Occurrence, accumulation, and comparison for Ukraine and France. Water, Air, & Soil Pollution, 223 2111–2124.

Ye, Z.H., Baker, A.J.M., Wong, M.H., & Willis, A.J. (1997) Zinc, lead and cadmium tolerance, uptake and accumulation in populations of Phragmites australis (Cav.) Trin. ex Steud. Annals of Botany, 80, 363-370.

Zhou, Q.F., Zhang, J.B., Fu, J.J., Shi, J.B. & Jiang, G.B. (2008) Biomonitoring: An appealing tool for assessment of metal pollution in the aquatic ecosystem. Analytica Chimica Acta, 606, 135-153.

Chapitre II : Take home message

Une approche multivariée – ici une analyse discriminante linéaire – basée sur les

concentrations foliaires en PTTE a permis d’établir un modèle de biomonitoring de

l’exposition des macrophytes aux PTTE le long de la rivière urbaine la Jalle d’Eysines. Les

macrophytes du type hémicryptophytes ont une biomasse souterraine plus faible que les

géophytes. Le stockage des PTTE dans la biomasse souterraine y est donc moindre et le

transfert vers les parties aériennes plus important. Si l’on considère les parties aériennes, les

hémcryptophytes sont de meilleurs biomoniteurs de l’exposition aux PTTE que les géophytes.

L’exposition n’est pas seulement la résultante des concentrations totales en PTTE dans le sol.

De nombreuses interactions entre ces concentrations totales et les paramètres physico-

chimiques du sol (texture, pH, teneur en matière organique, teneur en oxydes de Fe, potentiel

redox du compartiment) conditionnent l’exposition racinaire. Cette exposition est également

régie par l’ensemble des interactions biotiques dans la rhizosphère et la plante. Les micro-

organismes de la rhizosphère, les endophytes, les exsudats racinaires, les mécanismes de

stockage intra-cellulaires, les transporteurs membranaires et les systèmes de défense (cascade

des mécanismes anti-oxydants, compartimentage sub-cellulaire) sont autant de paramètres qui

conditionnent l’homéostasie du ionome de la plante et du potentiel rédox des cellules. Les

macrophytes auraient une tolérance constitutive aux expositions excessives en PTTE. Elles

s’appuieraient sur une capacité de stockage élevée dans les vacuoles des racines et des

rhizomes. Cependant, la variabilité intra-spécifique de cette tolérance aux PTTE est une

question à considérer, notamment pour utiliser des macrophytes en zones humides construites

et phytoremédier des eaux contaminées par des PTTE.

Cette question est abordée dans le chapitre suivant: les macrophytes présentent-elles une

plasticité phénotypique de la réponse à une contamination aux PTTE (ici Cu) en fonction de

130

l’intensité d’exposition aux PTTE à leur site d’origine ? La variable étudiée pour répondre à

cette question est un trait phénotypique intégrateur: la croissance racinaire.

131

Chapitre III

Variabilité intra-spécifique de la tolérance au Cu parmi les populations de 6

espèces de macrophytes

Cette partie est soumise à Environmental and Experimental Botany

132

a b

c d

e f

Figure 28 : Sites de prélèvement de macrophytes a.Ruisseau de Lagnet (Saint Christophe des Bardes (33), France), b. Ruisseau de Basilique (Saint Christophe des Bardes (33), France), c. Rivière Jalle d’Eysines, station J3 (Blanquefort (33), France), d.Zone humide du Cordon d’Or (Saint Médard d’Eyrans (33), France), e. Zone humide de La Cornubia (Bordeaux (33), France), f. Etang de Sanguinet (Sanguinet (33), France).g. Zone humide de la mine de Touro (Espagne).

g Macrophyte sampling sites, a. Lagnet Creek (Saint Christophe des Bardes (33), France), b. Basilique Creek (Saint Christophe des Bardes (33), France), c. Jalle d’Eysines River, sampling site J3 (Blanquefort (33), France), d. Cordon d’Or wetland (Saint Médard d’Eyrans (33), France), e. La Cornubia wetland (Bordeaux (33), France), f. Sanguinet Lake (Sanguinet (33), France), g. Touro mine wetland (Spain).

133

h i

j k

l m

Figure 28 (suite) : Sites de prélèvement de macrophytes h. Ruisseau du Palais (Le Palais sur Vienne (87), France), i. Zone humide de la mine de Fenice Capanne (Massa Maritima, Italie), j . Rivière Ribeira de Agua Forte (Beja, Portugal), k. Bassin de retardement de Avoca (Melbourne, Australie), l. Bassin de retardement de Cheltenham (Melbourne, Australie), m. Zone humide du lac Gwelup (Perth, Australie). n. Zone humide de Kozlovichi (Belarus).

n

h. Le palais Creek (Le Palais sur Vienne (87), France), i. Fenice Capanne mine wetland (Massa Maritima, Italy), j. Ribeira de Agua Forte Stream (Beja, Portugal), k. Avoca retarding basin (Melbourne, Australia), l. Cheltenham retarding basin (Melbourne, Australia), m. Lake Gwelup wetland (Perth, Australia), n. Kozlovichi wetland (Kozlovichi, Belarus).

134

Intra-specific variability of Cu tolerance in populations of six rooted

macrophytes

*L. Marchand a, F. Nsangawimana a, C. Gonnelli b, I. Colzi b, T. Fletcher c,d, N.

Oustrière a, A. Kolbas a, e, P. Kidd f, F. Bordas g, P. Newell h, P. Alvarenga i, J.B.

Lamy a, A. Deletic b, M. Mench a

a INRA, UMR 1202 BIOGECO, 69 route d’Arcachon, FR-33612, Cestas cedex, France ; University of Bordeaux 1, UMR 1202 BIOGECO, Bât B2, Avenue des facultés, FR-33405, Talence, France.

b Dipartimento di Biologia, Laboratorio di Ecologia e Fisiologia Vegetale, Università degli Studi di Firenze, via Micheli 1, IT-50121, Firenze, Italy.

c Department of Civil Engineering, Monash University, Room 118, Building 60, Clayton Campus, Clayton Victoria 3168, Melbourne, Australia.

d Melbourne School of Land & Environment, the University of Melbourne, 500 Yarra Boulevard, Burnley, 3121 and 221 Bouverie St, Parkville, Vic, 3010, Australia.

e Brest State University named after A.S. Pushkin, 21, Boulevard of Cosmonauts, 224016, Brest, Belarus

f Instituto de Investigaciones Agrobiológicas de Galicia, Consejo Superior de Investigaciones Científicas (CSIC), Santiago de Compostela, Spain.

g GRESE, Université de Limoges, 123 Avenue Albert Thomas, FR-87060, Limoges, France.

h Department of Environment and Conservation, Contaminated Sites Branch, Locked Bag 104, Bentley DC, 6983, Australia.

i Departamento de Tecnologias e Ciências Aplicadas, Escola Superior Agrária - Instituto

Politécnico de Beja, Rua Pedro Soares - Campus do IPB, Apartado 6155, PT-7801-295, Beja, Portugal.

*Corresponding author. marchand.lilian@gmail.com

Tel: +33 (0) 5 40 00 31 14

Fax: + 33 (0)5 40 00 36 57

135

Abstract

Intra-specific variability of root biomass production of six rooted macrophytes, i.e. Juncus

effusus, Phragmites australis, Schoenoplectus lacustris, Typha latifolia, Phalaris

arundinacea, and Iris pseudacorus grown from clones, in response to Cu exposure was

investigated. Macrophytes were sampled at sites with increasing total soil Cu (from 2.9 to

1750 mg kg-1) in France, Spain, Portugal, Australia, Italy, and Belarus. Clones were cultivated

in controlled conditions and then exposed to a range of Cu concentrations (from 0.08 to 25

µM) in hydroponics. Root biomass production varied widely for all the 6 studied macrophytes

in control conditions (0.08 µM) according to the plant origin. Root biomass production of

T.latifolia across the 2.5-25 µM Cu gradient depended on the sampling location (p<0.001) but

not on the Cu dose in the growth medium (p>0.05). For this rhizomatous geophytes rooted

macrophyte data support the hypothesis of a constitutive-like tolerance to Cu-exposure. Such

constitutive-like tolerance was not reported for I.pseudacorus, P.australis, J. effusus, S.

lacustris, and P. arundinacea. An intra-specific variability of root biomass production

depending on both the sampling site (p<0.001) and the Cu-dose (p<0.05) was reported for

these three species producing either low or no rhizomatous biomass. Root biomass production

depended on the plant origin for all the six investigated species, but no impact of the Cu-

exposure level at the sampling site on this trait was evidenced. Such clues of an intra-specific

variability of root biomass production depending on the location of origin and variation in Cu-

tolerance according to the species among rooted macrophyte species provide new hints in

choosing plant materials in constructed wetlands (CW).

Keywords: Constructed wetland, Hemicryptophytes, Rhizomatous geophytes, Root biomass,

Trace elements

136

I. Introduction

Aquatic ecosystems are used, directly and indirectly, as recipients of potentially toxic

effluents and wastes from domestic, agricultural and industrial activities (Demirezen et al.,

2007; Peng et al., 2008). Copper is one of the Potentially Toxic Trace Elements (PTTE),

which may migrate in dissolved and solid forms from urban areas and (agro)ecosystems to

surface waters, groundwater and wetland substrates, and its excess may accumulate in living

organisms (Kamal et al., 2004; van der Ent et al., 2013). Copper acts as a cofactor in the

electron transport chain in the mitochondria and the chloroplast (Palmer and guerinot, 2009;

Marschner, 2011). The most abundant protein associated with Cu in the plant cell is the

plastocyanin responsible for the electron transfer from the cytochrom b6f complex to the

photosystem I (PSI) (Yamasaki et al., 2009; Burkhead et al., 2009). However, at an exposure

higher than the cellular Cu homeostasis (5-20 µg Cu g-1 DW), Cu induces phytotoxicity

symptoms (e.g. biomass reduction, root growth inhibition, bronzing, chlorosis, reduced Fe, Zn

and P uptake, chloroplast integrity loss, etc.) (Kopittkke et al., 2010; Marschner, 2011).

Excessive free Cu ions can induce the formation of Reactive Oxygen Species (ROS) such as

superoxides (O2-), hydroxyl radicals (HO•) and hydrogen peroxide (H2O2) through Fenton and

Haber-Weiss reactions, which can peroxidize lipids and oxidize proteins and guanine

(Drazkiewicz et al., 2004; Sharma and Dietz, 2009, Kanoun-Boulé et al., 2009).

Natural ‘volunteer’ wetlands can improve water quality and in particular those associated with

mining activities by trapping PTTE in the rhizosphere (Beining and Otte, 1996; Narhi et al.,

2012). Constructed wetlands (CW) have also been used to improve the quality of

contaminated waters for at least two decades (Marchand et al., 2010, Lizama et al., 2011).

Rooted macrophytes are key players in both natural and constructed wetlands through radial

oxygen loss (ROL) and organic matter production which provide habitats for microorganisms

(Marchand et al., 2010; Cheng et al., 2009). Such macrophytes mainly accumulate PTTE in

roots, because of their fibrous system with large contact areas, rhizome tissues (Cardwell et

al., 2002; Bonnano and Lo Giudice, 2010; Romero Núñez et al., 2011), and to a lesser extent

in stems and leaves (Clemens, 2002; Baldantoni et al., 2004; Bragato et al., 2006). The

formation of Fe plaque deposits in the vicinity of wetland plant roots may contribute to

greater metal accumulation in the rhizosphere (McCabe et al., 2001; Otte et al., 2004). Thus,

rooted macrophytes may have been exposed to higher Cu concentrations than most dryland

plants over their evolution. Kissoon et al. (2010) reported for example greater Cu

137

concentrations under wetland compared to dryland conditions in the whole plant for Rumex

crispus. Consequently it is commonly admitted that selection may have shaped for a

constitutive PTTE tolerance for macrophytes (Ye et al., 1997, 2003; McCabe et al., 2001;

Matthews et al., 2004a,b; 2005; Kissoon et al., 2010).

However, in drylands, a small percentage of plants has the innate capability to develop metal-

tolerant populations at PTTE-contaminated sites (Verkleij et al., 2009; Memon and Schroeder,

2009; Verbruggen et al., 2009). When a species establishes on a soil with a too high PTTE

supply, adjustments will take place within the limits of phenotypic plasticity followed by

adaptation and evolution of efficiency or tolerance in populations over time (Schat 1999;

Pollard et al., 2002; Ernst, 2006; van der Ent et al., 2013). Such genetic adaptation may

generate distinct populations. Similarly, the presence of rooted macrophytes at PTTE-

contaminated sites already raised the question of whether wetland plants may evolve PTTE

tolerance likewise dryland plants (Deng et al., 2006). In other words, do the PTTE

constitutive tolerance for all rooted macrophytes is a species-wide trait or does it still exhibit

variability for some species? Investigations are needed to provide new insights into choosing

plant material in CW, since root biomass in CW determines the system efficiency and

promotes its long-term functioning (Marchand et al., 2010). Knowledge is currently lacking

on the intra-specific variability of macrophytes in response to PTTE exposure in wetland

communities (Brisson and Chazarenc, 2009; Marchand et al., 2010). Moreover, as suggested

by Deng et al. (2006), full and correct understanding of the nature of PTTE tolerance in

wetland plants should be based on study of a wide range of populations. Beside, Matthews et

al. (2004a) reported that in some studies populations used were from locations within

relatively close proximity to each other, and so could have originated from the same PTTE-

tolerant ancestors. Last, but not least, there is also a need to work simultaneously not on a

single species across a PTTE gradient, but more on several species at the same time to really

assess the site effect at field scale since each species may react differently when exposed to

high PTTE concentrations.

Therefore, we examined the root production of a wide range of macrophyte populations in

imbibed perlite with an increasing Cu exposure (from 0.08 to 25 µM). Six species, Juncus

effusus L., Phragmites australis (Cav.) Trin.ex Steud., Phalaris arundinacea L., Typha

138

latifolia L., Iris pseudacorus L., and Schoenoplectus lacustris L. were sampled in both metal-

contaminated and uncontaminated sites located in France, Portugal, Italy, Belarus, Spain, and

Australia. Root growth was considered here in the context of the trait-based approach

reflecting tolerance mechanisms to Cu exposure. The integrative, trait-based option is a

relevant tool to understand how organims face fast changing environmental conditions (Berg

and Ellers, 2010).

II. Materials and methods

2.1 Sites

Fourteen wetland sites were investigated between 2009 and 2012 (Table 15, Fig. 29).

Sampling sites were located in both the Northern hemisphere (i.e. seven in France, one in

Spain, one in Portugal, one in Italy, and one in Belarus) and the Southern hemisphere (three in

Australia). These sites had wide PTTE concentration ranges in soils, with total soil Cu (in mg

kg-1 DW) varying from 2.9 (Kozlovichi, Belarus) to 1750 (Ribeira de Agua Forte, Portugal)

and pH ranging from 3 (Gwelup Lake, Australia) to 7.4 (Lagnet, France and Fenice Capanne,

Italy) (Fig. 29). High soil PTTE concentrations were a result of either the soil geochemical

background or industrial and agricultural effluents. The La Cornubia site (Gironde, France) is

a CW which collected effluents and runoff from a former Cu sulphate plant and was in use for

over a century (Basol, 2012). The Cordon d’Or site (Gironde, France) is a natural wetland

receiving runoff from an adjacent former wood preservation site, which has been used over a

century and where creosote, Cu sulphate, chromated copper arsenate (CCA), and Cu

hydroxycarbonates with benzylalkonium chlorides have been successively used (Mench and

Bes, 2009; Marchand et al., 2011). The Lagnet and Basilique sites (Gironde, France) are both

draining ditches located in the vineyards of Saint-Emilion (Gironde, France), annually treated

with Cu sulphate. The Jalle d’Eysines River flows into the Garonne next to Bordeaux

(Gironde, France) and receives both PTTE-contaminated runoffs from industrial, agricultural

and residential areas and effluents from two major municipal wastewater treatment plants

(WTP) of the Bordeaux suburbs, serving more than 100,000 inhabitants. At Le Palais sur

Vienne (Haute-Vienne, France), macrophytes were sampled on the riverbanks of Le Palais

creek, a tributary of the Vienne River near Limoges, downstream from a former copper

electro refinery whose runoffs and discharges have occasionally contaminated the river. The

139

riverbank sandy soil of the uncontaminated Sanguinet Lake (Landes, France) has an acid soil

pH and its soil PTTE concentrations are generally lower than the background levels defined

by Blum et al. (2012). The Avoca and Argus Street sites (Victoria, Australia) are both

retarding basins (RB) built by Melbourne Water where stormwater runoff from a drainage

catchment is temporarily stored. Due to high urban pressure, these RB are contaminated by

PTTE (Bourgues et al., 2004; Marshall, 2004). The polymetallic (Zn, Cu, Pb, Fe, and Ag)

sulphide deposit of Fenice Capanne (Massa Marittima, Italy) was mined for 25 centuries up

until 1985 and the alteration of mine waste materials has produced pollution in superficial

waters and sediments (Mascaro et al., 2001). The Touro site (Galicia, Spain) is an abandoned

opencast mine under restoration, whose tailings mainly consist of oxidized materials such as

amphibolites, chalcopyrite, limonite, garnet and mainly Fe and Cu sulphides (Vega et al.,

2004; Asencio et al., 2013a,b). Samples were collected from a constructed wetland, which

was established as a means of controlling acid mine drainage. The Ribeira de Agua Forte

(Beja, Portugal) is a tributary stream to the Roxo stream, which receives acid mine drainage

(AMD) from the Aljustrel mine, a polymetallic (As, Zn, Cu, and Pb) sulphide deposit of the

Iberian pyrite belt (Alvarenga et al., 2008; Candeias et al., 2011). ). At Gwelup Lake,

Australia, high PTTE concentrations, mainly for As and Zn, result from soil geochemical

background. The bottom of the lake is constituted by mono-sulfidic black ooze (MBO), a

concentrated organically derived iron sulfide containing material. In summer, when it dries or

saturation is reduced, this produces highly acid conditions (acid sulfate material). Kozlovichi

(Belarus) is an uncontaminated wetland, located next a former beer production plant. Physico-

chemical parameters and PTTE concentrations of soils at the Avoca, Gwelup, Argus Street,

Fenice Capanne, Touro, and Ribeira de Agua Forte sites were previously reported (Table 15).

For other sites, three soil samples (0.5 kg fresh weight, FW) were collected with an unpainted

steel spade from the 0-25 cm soil layer. Samples were air-dried at the laboratory and sieved (5

mm, nylon mesh) prior to analysis. Total element concentrations and soil properties were

determined on air-dried soil at the INRA Laboratoire d'Analyses des Sols (LAS, Arras,

France) using standard methods (INRA LAS, 2007), (Table 15).

2.2 Plant sampling, vegetative reproduction and plant exposure to Cu

Emergent and rooted monocot macrophytes, i.e. P. australis, P. arundinacea, T. latifolia, J.

effusus, S. lacustris, S. holoschoenus, and I. pseudacorus were mainly collected at the

beginning of the growing season in 2009, 2010 and 2011. A few populations were also

140

sampled at the end of the growing season (Table 16). A total of 34 populations were sampled

with 20-30 individuals per population (below- and aboveground biomasses). For all

populations, plant samples were kept separate in buckets and as soon as possible placed in

water in a greenhouse at the Centre INRA-Bordeaux Aquitaine, Villenave d’Ornon, France.

The days after collection, rhizomes and/or stem-bearing buds were cut into small pieces (10-

20 cm). They were then grown in separate polyethylene containers (volume: 60x40x15 cm3)

containing perlite imbibed with a quarter-strength Hoagland nutrient solution (HNS,

Hoagland and Arnon, 1950): KNO3 (1.62 mM), Ca (NO3)2 (0.69 mM), NH4H2PO4 (0.25 mM),

MgSO4 (0.5 mM), H3BO3 (1.16 mM), MnCl2 (2.29 µM), CuSO45H2O (0.08 µM),

(NH4)6Mo7O24 (0.13 µM), ZnSO47H2O (0.19 µM) and FeSO4 (48.6 µM). Water volume was

maintained constant by adding tap water. Water was renewed and nutrients were added every

month during the growing season and every two months during winter to avoid anoxia and

nutrient depletion in the growth medium. After 6-10 months, in late winter, 24 standardized

tillers (stem and root size and volume were equal) of each population were isolated from the

sprouting rhizomes. These were grown in a new culture medium (imbibed perlite with HNS as

described above) in 9x8x9 cm3 pots for 6-8 weeks, and thereafter, 20 individual plants were

selected. At the beginning of the test (Table 16), their roots were stained with activated plant

coal (concentration: 1.5 %) according to Schat and Ten Bookum (1992). Thereafter,

individuals were transferred into plastic containers (1L) filled with 500 mL of a quarter-

strength HNS prepared with ultra-pure water (MiliQ sytem) (Hoagland and Arnon, 1950), and

perlite (50 g). The growth medium was spiked with Cu (CuSO45H2O) to achieve five

treatments (four replicates treatment-1): 0.08, 2.5, 5, 15, and 25 µM Cu (Table 17). All plants

were randomly placed on a bench in the same greenhouse (day (9-21h) 1911±1232 µM

photons m-2s-1, 28±5°C, night (21h-9h) 19±3°C). Nutrient solutions were changed every five

days to maintain Cu concentrations and avoid depletion of oxygen and nutrients. According to

Kopittke et al. (2010), redox-potential (Eh, WTW Multiline P4 meter, Germany), pH (Hanna

instruments, pH 210, combined electrode Ag/AgCl, USA) and Cu2+ concentrations (Cupric

ion electrode, Fischer Bioblock, USA) were measured in the solution imbibing perlite just

after changing the solution and after a 5-day exposure (n=6 for each concentration of the Cu

gradient, Table 17). The speciation of metal elements in the bulk solutions was determined

using the speciation program MINEQL+4.6. The standard databases included with the

software were applied (Schecher and McAvoy, 2003).

141

Plants were harvested after a 3-week exposure to Cu. Roots were thoroughly washed with

running tap water to remove unwanted perlite particles and blotted with a paper towel. White

roots, i.e. root biomass formed during Cu exposure, were isolated from black-stained roots

and their FW yield was determined. Roots were oven-dried at 60°C for 48h and then weighed.

2.3 Statistical analysis

Physico-chemical parameters in the growth medium of the distinct Cu levels were analyzed

just after changing the solutions (T0) and after a 5-day exposure (T5). For each parameter and

at a given Cu exposure differences between (T0) and (T5) were analyzed using a Student’s

t-test (Table 17). Differences were also analyzed for each parameters accross the Cu range

(0.08-25 µM Cu) at respectively (T0) and (T5) using a one-way ANOVA. Homoscedasticity

and normality were met for all tests. Post Hoc Tukey HSD tests were then performed to assess

multi-comparison of mean values (Table 17). Effect of the sampling site location on root DW

biomass production for each plant species in uncontaminated conditions (0.08 µM Cu) was

analyzed using a Kuskal-Wallis test (Table 18) and a Pairwise Wilcoxon sum ranks test (Fig.

30). Root DW yield depending on both Cu concentration in the growth medium (2.5-25 µM)

and the plant origin were analyzed for each plant species using two-way ANOVAs.

Homoscedasticity and normality were met for all tests (Table 19). Plots illustrating the two-

way interactions for each plant species are presented in Fig. 31. Finally, for each species and

within populations, differences in root production between each Cu dose (2.5-25 µM Cu) and

the uncontaminated modality (0.08µM Cu) were assessed using a Pairwise Wilcoxon sum

ranks test (Fig. 30). All analyses were carried out using R software (version 2.14.1 R

foundation for Statistical Computing, Vienna, Austria).

142

Table 15 : Metal concentrations (mg kg-1 DW) and physico-chemical parameters in soils at sampling sites.

sites geographic coordi nates Country C N Soil pH CEC Cu Zn Cr Ni

g kg -1 g kg -1 cmol kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1

Cornubia 44°54'26"N;0°32'46"W France 30.9 1.5 5.7 5.4 205 306 14 7.0

Cordon d'Or 44°43'27"N;0°30'56"W France 129.0 9.8 6.7 43.3 89.7 176 88 35.2 Lagnet 44°54'54"N;0°08'23"W France 5.5 0.5 7.4 2.4 27 22.9 14 4.6

Basilique 44°53'59"N;0°06'32"W France 19.9 1.6 7.1 13.7 71.2 60.5 67.5 31.9 Jalle 44°54'34"N;0°34'56"W France 28.9 2.6 7.5 26.3 32.7 171.3 79.6 40.2

Le palais 45°62'31"N;1°19'24"W France 6.1 0.5 6.5 2.33 21.2 46.9 16.9 6.8 Sanguinet 44°30'20"N;1°08'01"E France 173 9.1 5.0 13.4 3.3 11.9 5.6 1.5 Avoca** 37°07'56"S;145°01'53"E Australia nd nd nd nd 80 1000 28 13 Gwelup 31°52'40"S; 115°47'30"E Australia nd nd 3.0 nd nd nd nd Nd

Argus street* 37°47'42"S;145°04'24"E Australia nd nd nd nd 94 820 65 37 Capanne 43°00'39"N;10°55'04"W Italy 0.9 0.1 7.4 <1 375 720.0 35.7 29.7

Touro 42°52'34"N;8°20'40"W Spain 0.5-7.5 0.16-0.7 3.6-4.8 25.7-37.8 200-1200 80-110 100-150 60-70 Ribeira de Agua Forte 37°53'56"N;8°08'12"W Portugal nd nd 3.0 nd 1750 2000 nd Nd

Kozlovichi 52°06'48"N; 23°37'54"E Belarus 27.3 2.0 6.2 13.7 2.9 16.1 7.6 3 Background metal concentrations

in soils *** 10-40 20-200 10-50 10-50

sites Co Pb Cd Mo Mn Fe Ca Al Mg

mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 g kg -1 g kg -1 g kg -1 g kg -1

Cornubia 3.4 59.7 1.1 0.4 154 8.8 6.7 27.8 1.7 Cordon d'O r 11.1 62.0 0.55 3.1 195 28.3 14.6 57.8 3.3

Lagnet 2.1 15.6 0.07 0.2 139 5.7 9.8 22.6 0.9 Basilique 9.2 24.4 0.17 4 570 22.3 16.1 49.4 3.3

Jalle 19.0 54.9 0.50 1.6 804.7 45.1 6.3 90.4 9.3 Le palais 4.0 28.5 0.48 0.9 393 11.3 1.17 54.2 2.4 Sanguinet <1 13.0 0.06 0.2 46.5 5.1 3.1 7.5 0.7 Avoca** nd 110 1.0 nd nd 8 nd nd nd Gwelup nd nd nd nd nd nd nd nd nd

Argus street* nd 210 6.0 nd nd nd nd nd nd Capanne 10.5 66.0 1.03 0.5 3180 38.9 39.3 35.6 5.9

Touro nd nd 1.5-2.0 nd nd nd nd nd nd Ribeira de Agua Forte nd 540 nd nd nd nd nd nd nd

Kozlovichi <1 7.97 0.05 0.133 72.6 3.5 3.5 9.9 0.7 Background metal concentrations

in soils *** 1-10 10-50 0.05-1 0.5-2 300-1000 10-50 nd nd nd

(*Marshall, 2004; **Bourgues et al., 2004; ***Blum et al., 2012); nd: not determined

143

Figure 29 Site mapping according to total soil Cu and Zn concentrations (mg kg-1 DW). (Gwelup Lake is not reported due to lack of data).

5 10 20 50 100 200 500 1000 2000

10

20

50

100

200

500

1000

2000

Cu mg kg-1

(soil DW)

Zn mg kg-1

(soil DW)

Cornubia

Cordon d'O r

Lagnet

Basilique

Jalle

Le palais sur Vienne

Sanguinet

Avoca Argus street

Capanne

Touro

Ribeira de Agua Forte

Kozlovichi

144

Table 16 : Summary of the sampling time (first date) and the start of exposure to Cu (second date) for each macrophyte populations.

site J. effusus P. australis P. arundinacea T. latifolia S. lacustris I. pseudacorus

Juncaceae Poaceae Poaceae Typhaceae Juncaceae Iridaceae

Cornubia 05/2010-06/2011 04/2010-05/2011 05/2009-05/2010

Cordon d'Or 06/2009-05/2010 04/2010-04/2011 04/2009-04/2010

Lagnet 04/2010-05/2011 04/2009-05/2010 05/2009-05/2010 04/2010-06/2011

Basilique 06/2009-05/2010

Jalle 04/2009-05/2010 03/2010-07/2011 05/2009-05/2010 05/2009-05/2010

Le Palais sur Vienne 11/2010-06/2011 11/2010-05/2011 11/2010-05/2011

Sanguinet 04/2011-06/2011 04/2011-04-2012 04/2011-05/2012 Avoca

02/2011-04/2011

Gwelup 02/2011-04/2012 Argus Street 02/2011-04/2011 02/2011-04/2011

Capanne 11/2010-04/2012 Touro

04/2010-05/2011

Ribeira de Agua Forte 04/2011-04/2012 Kozlovichi 04/2010-06/2011

145

Table 17 Mean comparison on copper concentrations and physico-chemical parameters in growth medium, in

the 0.08-25 µM Cu range, at both day 0 and day 5 (n=6).

Total Cu

added Exposure time pH EC Cu 2+ Cu2+

(µM) Days (µS cm -1) (µM) (µg L -1)

0.08 0 6.6 ± 0.2a 555 ± 34a 0.08± 0.01a 5.8 ± 0.8a

2.5 0 6.9 ± 0.3a 542 ± 25a 0.11± 0.07a 7.1 ± 5.4a

5 0 6.8 ± 0.1a 566 ± 25a 0.30± 0.09b 19.4 ± 6.0b

15 0 6.6 ± 0.2a 562 ± 35a 1.20± 0.20c 74.3 ± 13.1c

25 0 6.6± 0.2a 549 ± 44a 2.80± 0.31d 175.4 ± 52.0d

0.08 5 7.3 ± 0.5a 442 ± 27a, * 0.05 ± 0.05a 3.5± 3.0a

2.5 5 7.3 ± 0.5a 453 ± 22a, * 0.06 ± 0.06a 3.9 ± 4.2a

5 5 7.1 ± 0.7a 458 ± 26a, * 0.16 ± 0.15a 10.4 ± 10.1a

15 5 7.2 ± 0.7a 460 ± 27a, * 0.15 ± 0.12a, * 10.1 ± 8.0 a, *

25 5 7.1± 0.4a 459 ± 34a, * 0.24 ± 0.21a, * 15.4 ± 13.4a, *

In a column and for each exposure time, mean values followed by the same letter did not statistically differ between

treatments at the 0.05 level with the Tukey HSD test. Stars (*) indicate for each Cu concentration significative diffferences

between exposure time at the 0.05 level with the Student’s t-test.

III. Results

3.1 Copper concentrations and physico-chemical characteristics of the growth medium

At T0, free Cu2+ concentrations significantly increased (p<0.05) across the Cu gradient from 5.8±0.8 to

175±52 µg L-1 (0.08 to 2.8 µM) (Table 17). For 0.08 µM Cu added in the bulk solution, speciation

calculations show that Cu was also present under both forms CuOH+ and CuSO4 (respectively 0.008

and 0.007 µM). When 25 µM Cu were added, Cu was mainly under the precipitate form CuO (24.6

µM). At the same time, pH and EC at 20 °C did not significantly change across the Cu gradient

(respectively, 6.6 to 6.9 and 542±25 to 562±35 µS cm-1) (p>0.05). Free Cu2+ cconcentrations did not

significantly vary between T0 and T5 across the 0.08-5 µM Cu gradient (p>0.05) (Table 17). For 15

and 25 µM Cu added in the growth medium, free Cu2+ decreased between T0 and T5 repectively from

74.3 to 10.1 µg L -1 (1.2 to 0.15 µM) and 175.4 to 15.4 µg L -1 (2.8 to 0.24 µM) (p<0.05). No variation

was reported for pH between T0 and T5 across the whole Cu gradient (p>0.05) whereas EC

significantly decreased (p<0.05) (Table 17). At T5, free Cu2+ concentrations were similar across the

Cu gradient (p>0.05), from 3.5±3 µg L-1 to 15.4±13.4 (0.05 to 0.24 µM) (p>0.05). For 0.08 µM Cu

146

added, and after five days, speciation calculations show that Cu was also present under both forms

CuOH+ and CuSO4 (respectively 0.03 and 0.005 µM). For 25 µM Cu added, Cu was mainly under the

CuO form (21.5 µM). At the same time pH and EC did not significantly change across the Cu gradient

(respectively, 7.3 to 7.1 and 442±27 to 459±34 µS cm-1). Decrease in free Cu2+ ions at T5 was

relatively higher at high Cu levels than at the low Cu level, leading to similar values of labile Cu2+ pool

in the growth medium across the Cu gradient after a 5 days exposure (Table 17).

Table 18 Kruskal-Wallis test for analyzing the effect of sampling site location on root biomass production of six

macrophytes in uncontaminated conditions (0.08 µM Cu) (n = 4 ind. population-1)

Root biomass production

Df KW Chi-Squared p

Juncus effusus 4 12.3 **

Schoenoplectus

lacustris 4 6.3 **

Phalaris arundinacea 5 16.5 ***

Iris pseudacorus 4 9 **

Phragmites australis 6 12.9 **

Typha latifolia 2 6.9 **

Significance levels: ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1; Df: Degree of freedom, p: p-value, KW Chi-squared:Kruskal-Wallis Chi Square

3.2 Crossed effect of growth location and Cu exposure on root biomass production

The lowest mean root biomass productions after a three weeks exposure at 0.08 µM Cu were

respectively 30 µg DW for S.lacustris and 40 µg DW for J.effusus. Mean root biomass production of I.

pseudacorus was 150 µg DW after three weeks, which was the highest value reported among the six

studied species, followed by T.latifolia (120 µg DW) and

147

Table 19 :Two-way ANOVA for analyzing the effects of sampling site location and total Cu concentration in

the growth medium (dose) on root growth rate and Cu-tolerance (expressed by the relative treatment efficiency

index: RTEI) of six macrophytes.

Root growth rate RTEI

Df Mean Sq F value p(>F) Df Mean Sq F value p(>F)

Juncus effusus site 4 4200 21.77 *** 4 0.252 3.523 *

dose 4 2236 11.59 *** 3 0.529 7.389 ***

site*dose 16 627 3.25 *** 12 0.173 2.423 *

residuals 75 193 52 0.0716

Typha latifolia site 2 62747 16.561 *** 2 0.156 1.77

dose 4 6322 1.669 3 0.067 0.758

site*dose 8 4726 1.247 6 0.058 0.665

residuals 30 3789 24 0.088

Phragmites australis site 6 104873 7.40 *** 6 0.338 4.681 ***

dose 4 31129 3.29 * 3 0.625 8.654 ***

site*dose 24 56798 1.003 18 0.106 1.472

residuals 105 247799 84 0.072

Iris pseudacorus site 4 234637 14.411 *** 3 0.422 3.642 *

dose 4 26929 1.654 4 0.413 3.566 *

site*dose 16 5508 0.338 12 0.088 0.761

residuals 75 16282 60 0.115

Schoenoplectus site 2 7335 26.84 *** 2 0.65 14.06 ***

lacustris dose 4 3413 12.49 *** 3 1.805 38.81 ***

site*dose 8 1164 4.26 *** 6 0.268 5.768 ***

residuals 45 273 36 0.0465

Phalaris arundinacea site 5 55125 20.31 *** 5 0.763 11.108 ***

dose 4 19910 7.33 *** 3 0.402 5.861 **

site*dose 20 7868 2.89 *** 15 0.738 2.011 *

residuals 90 2714 72 0.068

Significance levels: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1; Df: Degree of freedom, p: p-value, F: Fisher value, Means sq: Mean of squares.

P.arundinacea (116 µg DW) (Fig. 31). Across the Cu-gradient (2.5-25 µM Cu), root biomass

production of J.effusus and S.lacustris depended on both the total Cu concentration in the growth

medium and the sampling site location (growth location) (Table 19, p<0.01). Interaction between both

factors was also significant (p<0.001). Mean root production gradually decreased following the Cu

gradient from 28 to 18 µg DW for J.effusus and from 29 to 0 µg DW for S.lacustris after three weeks

(Fig. 31). Root biomass production of P.arundinacea also depended on both the Cu exposure

148

(p<0.001) and the sampling site location (p<0.001). Interaction between both factors was significant at

the 0.05 level (Table 19). Mean root biomass production of P.arundinacea after a three week exposure

remained stable aroud 100 µg DW when Cu exposure was in the range 2.5-5 µM (respectively 94 and

115 µg DW) but fall to 56 µg DW when Cu exposure rose to 15 µM (Fig. 31). In the case of

I.pseudacorus and P.australis root biomass production across the Cu gradient was greatly influenced

by the sampling site location (p<0.001) and less by the Cu concentration (p<0.05). Interaction between

both factors was also significant at the 0.05 level (Table 19). Mean root biomass production of

I.pseudacorus across the whole Cu gradient after three weeks was in the range 125-165 µg DW. Mean

root biomass production of P. australis remained stable (63-69 µg DW) when Cu exposure was in the

range 15-25 µM Cu, but fall to 42 µg DW when Cu exposure rose to 25 µM (Fig. 31). Root biomass

production of T.latifolia was in the range 95-120 µg DW after a three week exposure to 2.5-25 µM Cu

and was only influenced by the sampling site origin (p>0.001).

Root biomass production of J.effusus, S.lacustris, P.arundinacea, P.australis, I.pseudacorus and

T.latifolia growing in controlled and uncontaminated conditions (0.08 µM Cu) significantly differed

within species, depending on sampling site locations (p<0.01) (Table 18, Fig.30). In uncontaminated

conditions (0.08 µM Cu), the root biomass production of J.effusus was the lowest for the population

Argus Street and the highest for both the population La Jalle and Cordon d’Or (Fig. 30). Four of the

five populations of J.effusus displayed a root production decrease when exposed at 15-25 µM Cu in

comparison to uncontaminated conditions, except the population Argus Street whose root biomass

production remained stable across the whole Cu gradient. In the case of S.lacustris, root biomass

production was also the lowest for the population Argus Street when exposed at 0.08 µM Cu, whereas

the highest root biomass production was reported for the population Gwelup. All the three populations

of S.lacustris exhibited a significant root biomass production decrease when exposed at 15-25 µM Cu

in comparison to uncontaminated conditions. Among the six populations of P.arundinacea studied,

both populations from Sanguinet and Le Palais sur Vienne produced the lower root biomass when

exposed at 0.08 µM Cu, whereas populations from La Jalle, Cordon d’Or and Lagnet displayed the

highest root biomass production. Populations of P.arundinacea from La Jalle, Cordon d’Or and Lagnet

produced less root biomass under the 15-25 µM Cu gradient whereas no significant differences in root

biomass production was reported across the whole Cu gradient for the three populations Sanguinet, Le

Palais sur Vienne and Cornubia. Root biomass production did not significantly vary in uncontaminated

conditions between five of the seven populations of P.australis studied, but populations of Lagnet and

149

F.Capanne produced higher root biomass. Except for the population from Lagnet where root biomass

produced was lower at 25 µM Cu than at 0.08 µM Cu, no difference was reported within populations

in terms of root biomass production across the Cu gradient for P.australis. The same pattern occurred

for I.pseudacorus and T.latifolia, where no differences in root production was noted within populations

across the Cu gradient, except for T.latifolia growing at the location Basilique, which displayed lower

root biomass production when exposed at 5-15 µM Cu in comparison to uncontaminated conditions.

Root biomass production of T.latifolia under uncontaminated conditions was the highest for the

population Basilique. For I.pseudacorus, root biomass production in such conditions was

homogeneous for four of the five populations studied but was higher for the population Sanguinet (Fig.

30).

150

Root biomass production of J. effusus (µg DW) Root biomass production of S.lacustris (µg DW)

Root biomass production of P.australis (µg DW) Root biomass production of P.arundinacea (µg DW)

Figure 30 Root biomass production (µg DW plant-1) of six macrophytes after a 3-week exposure in the 0.08-25 µM Cu range. Bold letters indicate in each species differences

in root biomass production between populations in the 0.08 µM Cu conditions at p<0.1 (Pairwise Wilcoxon sum ranks test). For each species, stars (*) indicate within a

population values different from the mean root biomass production in the 0.08 µM Cu conditions at p<0.1 (Pairwise Wilcoxon sum ranks test).

Sanguinet Le Palais sur Vienne Cordon d'or Jalle Argus street

020

4060

8010

0

0.082.551525 µM

* * * * * * * * * * *BC

B

BC C

A

*

*

*

**

**

*

*

*

*

Argus street Avoca Gwelup

020

4060

8010

012

014

0 0.082.551525 µM

* * * * * *

* *

* * * * * *

* * A

B

B

Sanguinet Kozlovichi Jalle Lagnet Cornubia F. Capanne R. de Agua Forte

050

100

150

200

250

0.082.551525 µM

* *

AB

AA

B

AB

B

AB

Sanguinet Le Palais sur Vienne Cordon d'Or Jalle Lagnet Cornubia

050

100

150

200

250

0.082.551525 µM

* *

* *

* *A

A

B

C BC

BC

151

Root biomass production of T.latifolia (µg DW) Root biomass production of I.pseudacorus (µg DW)

Figure 30 (continued) Root biomass production (µg DW plant-1) of six macrophytes after a 3-week exposure in the 0.08-25 µM Cu range. Bold letters indicate in each

species differences in root biomass production between populations in the 0.08 µM Cu conditions at p<0.1 (Pairwise Wilcoxon sum ranks test). For each species, stars (*)

indicate within a population values different from the mean root biomass production in the 0.08 µM Cu conditions at p<0.1 (Pairwise Wilcoxon sum ranks test).

Basilique Cornubia Touro

050

100

150

200

250

300

350

0.082.551525 µM

* * *

* * *B

AA

Sanguinet Le Palais sur Vienne Cordon d'Or Jalle Lagnet

020

040

060

080

0 0.082.551525 µM

* *

C

AB B

ABC

A

152

Root biomass production of J. effusus (µg DW) Root biomass production of S. Lacustris (µg DW) Root biomass production of P. arundinacea(µg DW)

Cu (µM)

Root biomass production of I. pseudacorus (µg DW) Root biomass production of T. latifolia (µg DW) Root biomass production of P. australis (µg DW)

Cu (µM)

Figure 31 Root biomass production (µg DW plant-1) of six macrophytes after a 3-week exposure in the 0.08-25 µM Cu range. Values are means for individuals collected at different growing location. Error bars represent the intra-specific variability of root biomass production depending of sampling locations.

4060

8010

012

014

0

0 2.5 5 15 25

n=23 n=22 n=23 n=24 n=24

1020

3040

50

0 2.5 5 15 25

n=17 n=17 n=16 n=17 n=17 010

2030

4050

0 2.5 5 15 25

n=10 n=9 n=11 n=11 n=10

100

150

200

250

300

0 2.5 5 15 25

n=16 n=18 n=18 n=18 n=16

3040

5060

7080

90

0 2.5 5 15 25

n=23 n=24 n=25 n=25 n=26

5010

015

020

0

0 2.5 5 15 25

n=9 n=8 n=8 n=9 n=8

153

IV. Discussion

4.1. Copper speciation in the growth medium

As shown in Table 2, the six investigated species in the present study grew along an increasing Cu

gradient. Three forms of dissolved Cu were found in the ranking order Cu2+ (75%) > CuSO4 (9%) >

CuOH+ (1%) in the bulk solution just after 0.08 µM Cu was added. The rest of Cu (15%) may have

immediately adsorbed onto roots and perlite. After five days, the ranking order was Cu2+ (50%) >

CuOH+ (38%) > CuSO4 (6%). A slight pH increase across time promoted the formation of CuOH+

from Cu2+ (Brookins, 1988). Free Cu2+ may also have adsorbed onto roots and perlite and/or may have

been absorbed by the rhizosphere. These three phenomena led to Cu2+ decrease in the growth medium

after a five days exposure. When 25 µM of Cu were added in the bulk solution, almost all of Cu was

found to precipitate as CuO (Tenorite), whilst the remaining part was under the form Cu2+. After five

days, proportion of CuO slightly increased whilst Cu2+ concentrations strongly decreased. Both

phenomena resulted from Cu trapping onto roots and perlite, as well as Cu absorption by the

rhizophere.

4.2. Inter/Intraspecific variability vs constitutive tolerance to cope with excess Cu

4.2.1 Inter-specific variability of root biomass production across a Cu-gradient exposure

The six macrophytes investigated in our study exhibited an inter-specific variability of root biomass

production when grown in the same uncontaminated conditions in hydroponics (Fig. 31). In such

conditions, rhizomatous geophytes I.pseudacorus, T.latifolia (Raunkiær, 1934), but also the

hemicryptophyte P.arundinacea, produced the highest root biomass whilst the rhizomatous geophytes

J.effusus and S.lacustris produced the lowest one. The root biomass production of the six macrophytes

under 0.08 µM Cu exposure did not depend on the presence/absence of rhizomes. Across the Cu

gradient exposure, mean root biomass production of J.effusus gradually decreased from 2.5 µM Cu

whilst for S.lacustris and P.arundinacea it started to decrease from 15 µM Cu (Fig. 31). Phragmites

australis coped to Cu exposure till 15 µM Cu, then root production started to decrease at 25 µM Cu.

Conversely, mean root biomass production of the rhizomatous geophytes T. latifolia and I.

pseudacorus remained stable across the Cu gradient. An inter-specific variability of Cu-tolerance was

actually evidenced for the six species studied. Kopittke et al. (2011) reported Cu initially accumulates

154

with Cys, or ligands such as phytochelatins (PC) and metallothioneins (MT) possessing thiol groups,

due to its affinity for S, and that once these sites are saturated, Cu then accumulates within the cell

wall with polygalacturonic acids in roots of Vigna unguiculata (L.) Walp. The inter-specific

variability of Cu-tolerance in our Cu exposure range may be due to differences in PC and MT content

and in number of fixation sites in the cell wall among species. Another way to cope with Cu exposure

may be a limitation of root Cu uptake related to Si supply (Li et al., 2008). The Si deposition in cell

walls of the rhizodermis and/or the endodermis may provide additional metal binding sites and reduce

their apoplastic bypass flow. Caldelas et al. (2012) showed Si accumulation in rhizodermis to cope

with Cr exposure in I. pseudacorus. In our culture conditions, Si was likely provided by perlite.

Inequal root biomass production may also generate variations in Fe/Mn root-plaque content in the

rhizosphere, contributing to inequal Cu trapping onto the roots between species. Last but not least, Cu-

resistant endophytic bacteria can influence the dry weights of roots and aerial parts (Sun et al. 2010)

and their presence may differ between macrophyte species (Li et al., 2011). Further investigations are

needed to elucidate molecular and histological mechanisms underlying the inter-specific variability of

macrophyte root growth in response to Cu exposure.

4.2.2 Intra-specific variability of root biomass production across a Cu-gradient exposure

Among the six macrophytes investigated, all exhibited an intraspecific variability of root biomass

production when grown in the same uncontaminated conditions in hydroponics (Table 18). This

agreed with previous findings on macrophytes (Seliskar et al., 2002): ecotypes of Spartina alterniflora

grown at the same site remained morphologically closer to the ecotypes of their original site than to

other ecotypes planted at the same site, their development depending more on genetic variability than

on environmental conditions. In our study, in absence of environmental stress, such as high Cu

exposure, an intra-specific variability of root development of rooted macrophytes was actually found,

that was dependent on genetic variability among populations.

Across the Cu-gradient, the greatly significant interaction between Cu exposure and sampling site for

J. effusus and S. lacustris, suggested that the Cu-dose effect on root growth depended on the

population origin (Table 19). Consequently, J. effusus and S. lacustris displayed also here an intra-

specific variability of root production in response to Cu exposure. Root biomass production of P.

arundinacea also strongly depended on both the Cu-dose and the population origin, even if interaction

155

between both factors was surprisingly less significant. Thus, P.arundinacea also exhibited an intra-

specific variability of Cu-tolerance across the Cu gradient exposure. For the rhizomatous geophytes

P.australis and I. pseudacorus, the root biomass production in reponse to increasing Cu-exposure was

mainly driven by the sampling site of the population, but the Cu concentrations in the growing medium

also had a significative influence, although weaker, on this trait. The Cu-tolerance of P.australis and

I.pseudacorus displayed here an intra-specific variability depending on the population origin.

Conversely, the Cu-dose had no effect on the root biomass production within populations of T.latifolia.

Root growth of this species was only driven by the sampling location, not by the Cu exposure level

across the gradient. Thus, T.latifolia exhibited here a Cu constitutive-like tolerance, at least for the Cu

concentrations used.

In the literature, comparisons of wetland plant populations from PTTE-enriched sites with populations

from uncontaminated sites have shown that they were equally tolerant to high PTTE-exposure

(Matthews et al., 2004a). This led to the suggestion that wetland plants have an innate tolerance to

PTTE (McCabe et al., 2001). An early study by McNaughton et al. (1974) found that populations of T.

latifolia from PTTE-contaminated and uncontaminated environments grew equally well in elevated

PTTE conditions. This was confirmed by Taylor and Crowder (1983, 1984) and more recently by Ye

et al. (1997) for Zn, Pb and Ni tolerance in T.latifolia. The same researchers (Ye et al., 2003) found no

differences in Cu tolerance of populations of P.australis from PTTE-contaminated and

uncontaminated sites. Research on Glyceria fluitans also indicated that it has an innate tolerance to Zn

and Pb (McCabe and Otte, 2000, Matthews et al., 2004a). Constitutive Zn tolerance has also been

found in Eriophorum angustifolium L., T. latifolia, and P. australis (Matthews et al., 2004b, 2005).

Deng et al. (2006) reported no higher Zn and Pb tolerance in Alternanthera philoxeroides (Mart.)

Griseb, Beckmannia syzigachne (Steud.) Fernald, Leersia hexandra Swartz., Neyraudia reynaudiana

(Kunth) Keng, Oenanthe javanica (Bl.) DC and Polypogon fugax Steud for populations from

contaminated growing locations compared with populations from uncontaminated sites. Metal sorption

on the Fe/Mn plaque around macrophyte roots and with organic matter built up contributes to higher

metal accumulation in macrophyte rhizosphere (McCabe et al., 2001; Otte et al., 2004). However,

through root exudation and acidification, metals can be mobilized for uptake (Kissoon et al., 2010).

Consequently, wetland plants are frequently more exposed to higher metal concentrations than dryland

plants: e.g. metal uptake in Rumex crispus L. was 2.5 times higher under wetland compared to dryland

conditions (Kissoon et al., 2010). McCabe et al. (2001) and Otte et al. (2004) hypothesized high metal

exposure in wetlands may promote selection for constitutive metal tolerance.

156

As often reported in the literature, the absence of PTTE-tolerant populations at contaminated sites

often suggests the idea of an absence of intra-specific variability of PTTE tolerance within a species.

Our findings support the hypothesis according to which selection may have promoted constitutive

metal tolerance, but only for some wetland species. In this study, the only species that exhibited a

constitutive-like Cu tolerance was T.latifolia. The absence of significant effect of the Cu-dose on the

root biomass production of this species confirms previous findings of McNaughton et al. (1974),

Taylor and Crowder (1983, 1984), Ye et al. (1997,) and Matthews et al. (2005) on its PTTE-

constitutive-like tolerance. However, the so-called innate tolerance might be a buffer effect of the

rhizome on a relatively short-term exposure, and should be confirmed on a longer excess Cu exposure.

The low number of populations of T.latifolia investigated in our study might also be implied in the

absence of intra-specific variability reported. Further investigations are needed with a wider Cu

exposure range and a higher number of populations tested to really know to what Cu-level

macrophytes, especially T.latifolia, are tolerant.

The hypothesis of an innate tolerance was not obvious for J.effusus, S.lacustris, P.arundinacea and

even I.pseudacorus and P.australis. Actually, all these species displayed an intra-specific variability of

root development in response to an increasing Cu-exposure. But, no correlation was found between the

Cu-exposure at the sampling site and the Cu-tolerance when grown in hydroponics. As noted by Deng

et al. (2006), other mechanisms such as a different nutrient uptake might be possible for variance of

root biomass production in populations. Thus, the plant origin strongly influenced the root biomass

production of wetland plants across a Cu-gradient, whatever the species considered. Further

investigations are needed to assess which parameters drive the genetic selection leading to such

differences in root biomass production between distinct populations of macrophytes. Such insights are

needed for the selection of rooted macrophyte populations producing the highest root biomass for their

use in CW.

Conclusion

This study supports the hypothesis of a constitutive-like tolerance to Cu-exposure for I. T.latifolia.

However, the root biomass production depended on the plant origin for this species. Such constitutive-

like tolerance was not reported for I.pseudacorus, P.australis, J. effusus, S. lacustris, and P.

157

arundinacea since an intra-specific variability of root biomass production depending on both the

sampling site and the Cu-dose when grown in hydroponics was evidenced. Root biomass production

depended on the plant origin for all the six investigated species, but no impact of the Cu-exposure level

at the sampling site on this trait was evidenced. Thus, other mechanisms such as different nutrient

uptake might be possible for variance of root biomass production among populations. Such intra-

specific variability of root biomass production and Cu-tolerance among rooted macrophyte species

deserves more attention for designing innovative CW more efficient for cleaning Cu-contaminated

water and effluents.

Acknowledgements

This work was financially supported by AXA foundation (PhD grant of L. Marchand), ADEME, Department of

Urban Brownfield and Polluted Soils, Angers, France, the European Commission under the Seventh Framework

Programme for Research (FP7-KBBE-266124, GREENLAND) and the Aquitaine Region Council (Phytorem

project), Bordeaux, France. The authors thank Dr. Jean Paul Maalouf for his advice in statistic analysis, Dr.

Yohan Lebagouse-Pinguet for constructive discussion on phenotypic plasticity at the community level and Loic

Prudhomme for technical support.

158

References

Alvarenga, P., Palma, P., Goncalves, A.P., Baiao, N., Fernandes, R.M., de Varennes, A., Vallini, G., Duarte, E., Cunha-Queda, A.C., 2008. Assessment of chemical, biochemical and ecotoxicological aspects in a mine soil amended with sludge of either urban or industrial origin. Chemosphere. 72, 1774-1781. Asencio, V., Vega, F.A., Singh, B.R., Covelo, E.F., 2013a. Effects of tree vegetation and waste amendments on the fractionation of Cr, Cu, Ni, Pb and Zn in polluted mine soils. Sci Tot Environ. 443, 446-453. Asencio, V., Vega, F.A., Andrade, M.L., Covelo, E.F., 2013b. Tree vegetation and waste amendments to improve the physical condition of copper mine soil. Chemosphere. 90, 603-610. Baldantoni, D., Alfani, A., Di Tommasi, P., Bartoli, G., De Santo, A.V., 2004. Assessment of macro and microelement accumulation capability of two aquatic plants. Environ Pollut. 130, 149-156. Beining, B.A., Otte, M.L., 1996. Retention of metals originating from an abandoned lead-zinc mine by a wetland at Glendalough, Co Wicklow. Biol Environ. 96B, 117-126. Berg, M.P., Ellers, J., 2010. Trait plasticity in species interactions: a driving force of community dynamics. Evol Ecol. 24, 617-629. Bonanno, G., Lo Giudice, R., 2010. Heavy metal bioaccumulation by the organs of Phragmites australis (common reed) and their potential use as contamination indicators. Ecol Ind. 10, 639-645. Bragato, C., Brix, H., Malagoli, M., 2006. Accumulation of nutrients and heavy metals in Phragmites australis (Cav.) Trin. ex Steudel and Bolboschoenus maritimus (L.) Palla in a constructed wetland of the Venice lagoon watershed. Environ Pollut. 144, 967-975. Burkhead, J.L., Reynolds, K.A.G., Abdel-Ghany, S.E., Cohu, C.M., Pilon, M., 2009. Copper homeostasis. New Phytol. 182, 799-816. Basol., 2012. Pollution des sols. Ministère de l’écologie, du développement durable et de l’énergie. In French. Available at http://www.basol.environnement.gouv.fr/ Blum, W.E.H., Horak, O., Mentler, A., Puschenreiter, M., 2012. Environmental and Ecological chemistry – Vol.II – Trace elements. UNESO-EOLSS http://www.eolss.net/Sample-Chapters/C06/E6-13-0304.pdf (verified on September, 29, 2012) Bourgues, S., Hart, B., 2004. Sustainability of wetlands used for wastewater disposal. Final report. NSW, Environmental Trust. Water studies center, Monash University. Melbourne, Australia. Brisson, J., Chazarenc, F., 2009. Maximizing pollutant removal in constructed wetlands: Should we pay more attention to macrophyte species selection? Sci Tot Environ. 407, 3923-3930. Brookins, D.J., 1988. Eh-pH Diagrams for Geochemistry. Springer-Verlag Berlin Heidelberg, New York, NY, 176 pp.

Caldelas, C., Araus, J.L., Febrero, A., Bort, J., 2012. Accumulation and toxic effects of chromium and zinc in Iris pseudacorus L. Acta Physiol Plant. 34, 1217-1228. Candeias, C., Ferreira da Silva, E., Salgueiro, A.R., Pereira, H.G., Reis, A.P., Patinha, C., Matos, J.X., Avila, P.H., 2011. The use of multivariate statistical analysis of geochemical data for assessing the spatial distribution of soil contamination by potentially toxic elements in the Aljustrel mining area (Iberian Pyrite Belt, Portugal). Environ Earth Sci. 62, 1461-1479. Cardwell, A.J., Hawker, D.W., Greenway, M., 2002. Metal accumulation in aquatic macrophytes from southeast Queensland, Australia. Chemosphere. 48, 653-663. Cheng, X.Y., Chen, W.Y., Gu, B.H., Liu, X.C., Chen, F., Chen, Z.H., Zhou, X.Y., Li, Y.X., Huang, H., Chen, Y.J., 2009. Morphology, ecology and contaminant removal efficiency of eight wetland plants with differing root systems. Hydrobiol. 623, 77-85.

159

Clemens, S., Palmgren, M.G., Kramer, U., 2002. A long way ahead: understanding and engineering plant metal accumulation. Trends Plant Sci. 7, 309-315. Demirezen, D., Aksoy, A., Uruc, K., 2007. Effect of population density on growth, biomass and nickel accumulation capacity of Lemna gibba (Lemnaceae). Chemosphere. 66, 553-557. Deng, H., Ye, Z.H., Wong, M.H., 2006. Lead and zinc accumulation and tolerance in populations of six wetland plants. Environ Pollut. 141, 69-80. Drazkiewicz, M., Skorzynska-Polit, E., Krupa, Z., 2004. Copper-induced oxidative stress and antioxidant defense in Arabidopsis thaliana. BioMetals. 17, 379-387. Ernst, W.H.O., 2006. Evolution of metal tolerance in higher plants. For Snow Landsc Res. 80, 251-274. Hoagland, D.R., Arnon, D.L., 1950. The water-culture method for growing plants without soil. California Agricultural Experiment Station. Circular 347. Institut National de la Recherche Agronomique Laboratoire d’Analyses des Sols (INRA LAS)., 2007. Méthodes employées au laboratoire (in French). Available online at : http://www5.lille.inra.fr/las/methodes_d_analyse/Sols/10_elements_totaux_hf (Verified on January, 28, 2013) Kamal, M., Ghaly, A.E., Mahmoud, N., Cote, R., 2004. Phytoaccumulation of heavy metals by aquatic plants. Environ Int. 29, 1029-1039. Kanoun-Boule, M., Vicente, J.A.F., Nabais, C., Prasad, M.N.V., Freitas, H., 2009. Ecophysiological tolerance of duckweeds exposed to copper. Aquat Toxicol. 91, 1-9. Kissoon, L.T.T., Jacob, D.L., Otte, M.L., 2010. Multi-element accumulation near Rumex crispus roots under wetland and dryland conditions. Environ Pollut 158, 1834-1841. Kopittke, P.M., Blamey, F.P.C., Asher, C.J., Menzies, N.W., 2010. Trace metal phytotoxicity in solution culture: a review. J Exp Bot. 61, 945-954. Kopittke, P.M., Menzies, N.W., de Jonge, M.D., McKenna, B.A., Donner, E., Webb, R.I., Paterson, D.J., Howard, D.L., Ryan, C.G., Glover, C.J., Scheckel, K.G., Lombi, E., 2011. In situ distribution and speciation of toxic copper, nickel, and zinc in hydrated roots of cowpea. Plant Physiol. 156, 663-673. Li, J., Leisner, S.M., Frantz, J., 2008. Alleviation of copper toxicity in Arabidopsis thaliana by silicon addition to hydroponic solutions. J Am Soc Hortic Sci. 133, 670-677. Li, Y.H., Liu, Q.F., Liu, Y., Zhu, J.N., Zhang, Q., 2011. Endophytic bacterial diversity in roots of Typha angustifolia L. in the constructed Beijing Cuihu Wetland (China). Res Microbiol. 162, 124–131. Lizama, K.A., Fletcher, T.D., Sun, G., 2011. Removal processes for arsenic in constructed wetlands. Chemosphere. 84, 1032-1043. Marchand, L., Mench, M., Jacob, D.L., Otte, M.L., 2010. Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants and standardized measurements: A review. Environ Pollut. 158, 3447-3461. Marchand, L., Mench, M., Marchand, C., Le Coustumer, P., Kolbas, A., Maalouf, J.P., 2011. Phytotoxicity testing of lysimeter leachates from aided phytostabilized Cu-contaminated soils using duckweed (Lemna minor L.). Sci Tot Environ. 411, 146-153. Marschner, P., 2011. Mineral nutrition of higher plants, 3rd ed. (London: Academic press). 684 pp. Marshall, S.. 2004. Sediment toxicity to wetland plants at Argus Street retarding basin, Cheltenham. Report to Melbourne Water. University of Melbourne. Melbourne, Australia. Mascaro, I., Benvenuti, B., Corsini, F., Costagliola, P., Lattanzi, P., Parrini, P., Tanelli, G., 2001. Mine wastes at the polymetallic deposit of Fenice Capanne (southern Tuscany, Italy). Mineralogy, geochemistry, and environmental impact.

160

Environ Geol. 41, 417-429. Matthews, D.J., Moran, B.M., McCabe, P.F., Otte, M.L., 2004a. Zinc tolerance, uptake, accumulation and distribution in plants and protoplasts of five European populations of the wetland grass Glyceria fluitans. Aquat Bot. 80, 39-52. Matthews, D.J., Moran, B.M., Otte, M.L., 2004b. Zinc tolerance, uptake, and accumulation in the wetland plants Eriophorum angustifolium, Juncus effusus, and Juncus articulatus. Wetlands. 24, 859-869. Matthews, D.J., Moran, B.M., Otte, M.L., 2005. Screening the wetland plant species Alisma plantago-aquatica, Carex rostrata and Phalaris arundinacea for innate tolerance to zinc and comparison with Eriophorum angustifolium and Festuca rubra Merlin. Environ Pollut. 134, 343-351. McCabe, O.M., Baldwin, J.L., Otte, M.L., 2001. Metal tolerance in wetland plants? Minerva Biotecnol. 13, 141-149. McNaughton, S.J., Folsom, T.C., Lee, T., Park, F., Price, C., Roeder, D., Schmitz, J., Stockwell, C., 1974. Heavy metal tolerance in Typha latifolia without the evolution of tolerant races. Ecology. 55, 1163–1165. Memon, A.R., Schroder, P., 2009. Implications of metal accumulation mechanisms to phytoremediation. Environ Sci Pollut Res. 16, 162-175. Mench, M., Bes, C., 2009. Assessment of ecotoxicity of topsoils from a wood treatment site. Pedosphere. 19, 143-155. Narhi, P., Raisanen, M.L., Sutinen, M.L., Sutinen, R., 2012. Effect of tailings on wetland vegetation in Rautuvaara, a former iron-copper mining area in northern Finland. J Geochem Explor. 116, 60-65. Otte, M.L., Matthews, D.J., Jacob, D.L., Moran, B.M., Baker, A.J.M., 2004. Biogeochemistry of metals in the rhizosphere of wetland plants - An explanation for "innate" metal tolerance? Wetlands Ecosystems in Asia: Function and Management, 1, 87-94. Palmer, C.M., Guerinot, M.L., 2009. Facing the challenges of Cu, Fe and Zn homeostasis in plants. Nat Chem Biol. 5, 333-340. Peng, K., Luo, C., Lou, L., Li, X., Shen, Z., 2008. Bioaccumulation of heavy metals by the aquatic plants Potamogeton pectinatus L. and Potarnogeton malaianus Miq. and their potential use for contamination indicators and in wastewater treatment. Sci Tot Environ. 392, 22-29. Pollard, A.J., Powell, K.D., Harper, F.A., Smith, J.A.C., 2002. The genetic basis of metal hyperaccumulationin plants. Crit Rev Plant Sci. 21, 539-566. Raunkiær, C., 1934. The life forms of plants and statistical plant geography. Oxford University Press. Oxford. Romero Nunez, S.E., Marrugo Negrete, J.L., Arias Rios, J.E., Hadad, H.R., Maine, M.A., 2011. Hg, Cu, Pb, Cd, and Zn accumulation in macrophytes growing in tropical wetlands. Water Air Soil Poll. 216, 361-373. Schat, H., Ten Bookum, W.M., 1992. Genetic control of copper tolerance in Silene vulgaris. Heredity. 68, 219-229. Schat, H., 1999. Plant responses to inadequate and toxic micronutrient availability: General and nutrient –specific mechanisms. In: Gissen-Nielsen G, Jensen A (eds) Plant nutrition – molecular biology and genetics. Kluwer, Dordrecht, pp 311-326. Schecher, W.D., McAvoy, D.C., 2003. MINEQL+: A chemical equilibrium modeling system, version 4.6 for Windows, user’s manual, v2.00, Environmental Research Software: Hallowell, Maine. Seliskar, D.M., Gallagher, J.L., Burdick, D.M., Mutz, L.A., 2002. The regulation of ecosystem functions by ecotypic variation in the dominant plant: a Spartina alterniflora salt-marsh case study. J Ecol. 90, 1-11. Sharma, S.S., Dietz, K.J., 2009. The relationship between metal toxicity and cellular redox imbalance. Trends in Plant Sci. 14, 43-50. Sun, L.N., Zhang, Y.F., He, L.Y., Chen, Z.J., Wang, Q.Y., Qian, M., Sheng, X.F., 2010. Genetic diversity and

161

characterization of heavy metal-resistant-endophytic bacteria from two copper-tolerant plant species on copper mine wasteland. Bioresource Technol. 101, 501–509. Taylor, G.J., Crowder, A.A., 1983. Uptake and accumulation of heavy metals by Typha latifolia in wetlands of the Sudbury, Ontario region. Can J Bot. 61, 63-73. Taylor, G.J., Crowder, A.A., 1984. Copper and nickel tolerance in Typha latifolia clones from contaminated and uncontaminated environments. Can J Bot. 62, 1304-1308. van der Ent, A., Baker, A.J.M., Reeves, R.D., pollard, A.J., Schat, H., 2013. Hyperaccumulators of metal and metalloid trace elements: Facts and fiction. Plant Soil. 362, 319-334. van Hoof, N., Koevoets, P.L.M., Hakvoort, H.W.J., Ten Bookum, W.M., Schat, H., Verkleij, J.A.C., Ernst, W.H.O., 2001. Enhanced ATP-dependent copper efflux across the root cell plasma membrane in copper-tolerant Silene vulgaris. Physiol Plant. 113, 225-232. Vega, F.A., Covelo, E.F., Andrade, M.L., Marcet, P., 2004. Relationships between heavy metals content and soil properties in minesoils. Anal Chim Acta. 524, 141-150. Verbruggen, N., Hermans, C., Schat, H., 2009. Mechanisms to cope with arsenic or cadmium excess in plants. Curr Opin Plant Biol. 12, 364-372. Verkleij, J.A.C., Golan-Goldhirsh, A., Antosiewisz, D.M., Schwitzguebel, J.P., Schroeder, P., 2009. Dualities in plant tolerance to pollutants and their uptake and translocation to the upper plant parts. Environ Exp Bot. 67, 10-22. Ye, Z.H., Baker, A.J.M., Wong, M.H., Willis, A.J., 1997. Zinc, lead and cadmium tolerance, uptake and accumulation by Typha latifolia. New Phytol. 136, 469-480. Yamasaki, H., Hayashi, M., Fukazawa, M., Kobayashi, Y., Shikanai, T., 2009. SQUAMOSA Promoter Binding Protein-Like7 is a central regulator for copper homeostasis in Arabidopsis. Plant Cell. 21, 347-361. Ye, Z.H., Baker, A.J.M., Wong, M.H., Willis, A.J., 2003. Copper tolerance, uptake and accumulation by Phragmites

australis. Chemosphere. 50, 795-800.

162

V. Chapitre III Informations complémentaires

Marchand Lilian, Grebenshchykova Zhanna, Oustrière Nadège, Mench Michel

UMR BIOGECO INRA 1202, Écologie des Communautés, Université Bordeaux 1, Bât. B2, RdC Est, Avenue des Facultés, 33405 Talence, France, and INRA, 69 Route d’Arcachon, 33610 Cestas, France.

En parallèle aux mesures de plasticité phénotypique de la croissance racinaire des 6 espèces de

macrophytes exposées au Cu, une série de mesures complémentaires a été effectuée sur la seule espèce

Phragmites australis.

Ces mesures ne se focalisant que sur une des 6 espèces considérées dans le chapitre, elles sont exposées ici,

indépendamment du corps principal du chapitre.

� L’activité photosynthétique dans les feuilles de P. australis a été suivie pendantles trois semaines

d’exposition au Cu.

� L’activité de la Gaïacol peroxydase dans les racines de P. australis a été mesurée au bout de trois

semaines d’exposition au Cu.

� Les concentrations en Cu total dans les racines formées au cours des trois semaines d’exposition ont

été mesurées.

5.1 Matériel et Méthodes

Le protocole d’exposition des individus de P. australis est le même que celui détaillé dans le matériel

et méthode du chapitre III (section Plant sampling, vegetative reproduction and plant exposure to Cu).

Six individus (clones produits en serre à l’INRA de Villenave d’Ornon) de quatre populations (Jalle,

Cornubia, Capanne et Lagnet) on été testés pendant trois semaines en Juin 2012. Pour chaque

population, trois individus ont été exposés à 0,08 µM Cu (contrôle non contaminé) et trois autres à

25 µM Cu.

Mesures de fluorescence de la chlorophylle

Pendant les trois semaines d’exposition, de la semaine 27 à la semaine 29, l’activité photosynthétique

de chaque individu a été mesurée une fois par semaine par fluorométrie (fluorimètre: Portable

163

Chlorophyll Fluorometer PAM - 2100 (Walz) • Effeltrich • Germany). Elle a également été mesurée en

semaine 26 chez les individus non exposés (Figure 32). La nuit précédant la mesure, les plantes sont

placées à l’obscurité dans une pièce aux stores fermés. La première mesure a lieu à 9h du matin, à

l’obscurité (50 µmol m-2 s-1 photons). Tous les centres photosynthétiques sont alors fermés. Le niveau

minimal de fluorescence à l'obscurité (F0) est mesuré à l'aide d'une impulsion modulée

(<0,05 µmol m-2 s-1) trop petite pour induire des changements physiologiques importants dans la

plante. La fluorescence maximale (Fm) est ensuite mesurée en appliquant une impulsion de lumière

actinique saturante de 15000 µmol m-2s-1 de 0,7 s. Tous les centres photosynthétiques sont alors

saturés. L’efficience maximale (potentielle) du photosystème est obtenue par le rapport

Fm-F0/Fm (Cambrollé et al., 2011). La valeur Fm-F0 est notée Fv, il s’agit de la fluorescence variable. La

série de mesures est répétée à 14h afin d’obtenir F0’et Fm’ pour en déduire l’efficience réelle du

photosystème via le rapport Fm’-F0’/Fm’. Cette valeur est dénommée Yield (Figure 32).

Mesures de l’activité de la Gaïacol peroxydase

Le protocole de mesure de l’activité gaïacol peroxydase est détaillé dans la thèse d’A.Lagriffoul

(1998).

Mesures de la concentration en Cu total dans les racines

Les racines sont prélevées après trois semaines d’exposition à Cu. Elles sont rincées à l’eau du robinet

pour éliminer les particules de perlite et nettoyées avec du papier absorbant. Les racines blanches, i.e.

la biomasse formée pendant l’exposition à Cu, sont isolées des racines noires. Elles sont ensuite

séchées à l’étuve pendant 48h à 60 °C. Un échantillon moyen de 0,1 g de racines sèches est réalisé par

mélange des trois réplicats pour chaque modalité et les concentrations en Cu sont mesurées en suivant

la procédure décrite dans la section 2.4 du chapitre II Mineral analysis of leaf samples.

5.2 Résultats et discussion

5.2.1 Activité photosynthétique

Pour une exposition à 0,08 µM Cu, la valeur minimale du rapport Fv/Fm (sans unité) était de

0,56±0,04 pendant la semaine 29 (Lagnet) et la valeur maximale de 0,75±0,05 pendant la semaine 28

(Jalle). Ce rapport reste stable autour de 0,7 cours du temps et entre les populations (Figure 32). Lors

164

d’une exposition à 25 µM Cu, la valeur minimale de ce rapport Fv/Fm était de 0,58±0,04 (Capanne) en

semaine 28 et la valeur maximale de 0,69±0,04 (Cornubia) en semaine 27. Aucune variation

importante, au cours du temps ou entre les populations, n’a été observée. La valeur de ce rapport reste

stable autour de 0,7 (Figure 32).

Les valeurs du Yield pour 0,08µM Cu sont restées stables au cours du test. La valeur minimale

(0,73±0,06) a été mesurée en semaine 28 pour la population Capanne etla valeur maximale (0,80±0,01)

en semaine 26 pour les deux populations (Cornubia et Capanne). A 25 µM Cu dans le milieu, les

valeurs du Yield ont varié entre 0,74±0,01 (Jalle, Cornubia, Capanne) en semaine 29 et 0,79±0,02 en

semaine 28 pour la population Jalle. Ce paramètre n’a pas varié significativement au cours du temps

ou entre les populations.

a b

c d

Figure 32 : Activité photosynthétique dans les feuilles de quatre populations de Phragmites australis.

a. Fv/Fm à 0,08 µM Cu ; b. Fv/Fm à 25 µM Cu ; c. Yield à 0,08 µM Cu ; d. Yield à 25 µM Cu.

0,4

0,5

0,6

0,7

0,8

0,9

Semaine

26

Semaine

27

Semaine

28

Semaine

29

Fv/Fm (0,08 μM Cu)

0,4

0,5

0,6

0,7

0,8

0,9

Semaine 27 Semaine 28 Semaine 29

Fv/Fm (25 μM Cu)

0,4

0,5

0,6

0,7

0,8

0,9

Semaine

26

Semaine

27

Semaine

28

Semaine

29

Yield (0,08 μM Cu)

0,4

0,5

0,6

0,7

0,8

0,9

Semaine 27 Semaine 28 Semaine 29

Yield (25 μM Cu)

Jalle Capanne Cornubia Lagnet

165

L’activité photosynthétique de P. australis n’est pas influencée par une exposition à 25 µM Cu, et ce

quelle que soit l’exposition en Cu sur le site d’origine. Cambrolle et al. (2011) rapportent une

diminution progressive de l’activité photosynthétique dans les feuilles d’Halimione portulacoides –

une halophyte supportant une inondation temporaire – lorsque l’exposition s’accroît de 0 à 130 mM

Zn, avec un seuil à partir de 70 mM Zn. Cette stabilité de l’activité photosynthétique

d’H .portulacoides lorsque les concentrations en Zn sont inférieures à 70 mM Zn est similaires à nos

résultats. Une exposition de P. australis à Cu et d’H. portulacoides à Zn n’a pas de conséquence

directe sur l’activité photosynthétique tant que les concentrations restent de l’ordre du µM. Les

mécanismes de détoxification au niveau des racines et rhizomes décrits dans la première partie de

l’introduction et dans le chapitre III permettent de limiter efficacement les effets négatifs de Cu sur les

parties aériennes.

5.2.2 Activité de la Gaïacol peroxydase et Concentrations en Cu dans les racines

µmol min-1

0,08 µM Cu

25 µM Cu

Jalle Capanne Lagnet Cornubia

Figure 33 : Activité des gaïacol-peroxydases dans les racines de quatre populations de

Phragmites australis formées pendant trois semaines d’exposition à Cu. Les valeurs moyennes des

barre-graphes surmontées d’une même lettre ne diffèrent pas statistiquement au seuil de 5% (test de comparaison

par paire de Wilcoxon assorti d’une correction de Bonferroni).

166

Table 20 : Concentrations en Cu dans les racines de quatre populations de Phragmites australis

formées pendant trois semaines d’exposition à Cu.

Population

Concentrations en Cu ajoutées

dans le milieu de culture

Concentrations en Cu dans les racines

après trois semaines

d'exposition

(µM) (mg kg -1DW) Cornubia 0 30

25 1450

Capanne 0 12

25 980

Lagnet 0 36

25 2201

Jalle 0 Na 25 Na

Na : Mesure non obtenue

Les valeurs minimales de l’activité des gaïacolperoxydases(GPOD) dans les racines jeunes ont été

obtenues pour la population Capanne avec 312±51 µmol min-1 chez les individus exposés à 25 µM Cu

et 461±104 µmol min-1chez les individus exposés à 0,08 µM Cu. Pour les autres populations, l’activité

GPOD varie entre 1119±188et 1388±461 µmol min-1. Les valeurs maximales ont été déterminées chez

la population Jalle: 1388±461 (0,08 µM Cu) et 1119±188 µmol min-1 (25 µM Cu) et pour la population

Cornubia: 1254±113 (0,08 µM Cu) et 1335±445 µmol min-1 (25 µM Cu). Les valeurs moyennes de

l’activité GPOD pour les populations Jalle, Lagnet et Cornubia ne différent pas statistiquement

(p<0,05) quelle que soit l’intensité de l’exposition au Cu (0,08 ou 25 µM Cu) (Figure 33). L’activité

GPOD des racines de la population Capanne ne dépend pas de l’intensité de l’exposition au Cu (0,08

ou 25 µM Cu) (p<0,05), mais elle est en revanche plus faible que pour les trois populations

précédentes, aux deux expositions en Cu.

Les concentrations minimales en Cu dans les racines jeunes ont été obtenues pour la population

Capanne avec 12 mg kg-1 DW chez les individus exposés à 0.08 µM Cu et 980 mg kg-1 DW chez les

individus exposés à 25 µM Cu (Table 20). La population Cornubia présente des valeurs intermédiaires

de 30 et 1450 mg kg-1 DW pour des expositions respectives à 0.08 et 25 µM Cu. La population Lagnet

présente les concentrations en Cu dans les racines jeunes les plus élevées, respectivement 36 et 2201

mg kg-1 DW pour des expositions à 0.08 et 25 µM Cu. Les concentrations pour la population Jalle

n’ont pas été obtenues.

167

Il existe une plasticité phénotypique de l’activité GPOD dans les racines de P. australis en fonction du

site d’origine pour au moins une population. Pour une population donnée, il n’y a cependant pas de

variation significative de l’activité GPOD en fonction de l’intensité d’exposition, i.e. 0,08 ou 25 µM

Cu. La population avec l’activité GPOD la plus faible est la population Capanne. Parmi les quatre

populations, elle provient du site le plus contaminé en Cu. Cette population présente également des

concentrations en Cu dans les racines plus faibles que les populations Cornubia et Lagnet après

exposition à 0,08 et 25 µM Cu. La faible activité GPOD reflèterait plusieurs hypothèses à valider dans

des études complémentaires: e.g. (1) une possible faible activité de l’ensemble du système antioxydant

en relation avec un compartimentage cellulaire efficace du Cu, (2) une efficacité accrue d’autres

composantes de la cascade antioxydante, (3) une plasticité des mécanismes moléculaires régulant

l’homéostasie cellulaire de Cu, sans doute en amont de la cascade du système antioxydant, etc.

(Castruita et al., 2011; Shin et al., 2012). Comme dans la discussion du chapitre III, nous proposons un

compartimentage cellulaire dans les racines de la population Capanne de P. australis, dont une

complexation de Cu dans les vacuoles (Huang and Wang, 2010). Une autre hypothèse favorisant le

maintien de l’homéostasie consisterait en une capacité accrue à utiliser le Si présent dans le milieu par

la plante afin d’augmenter son dépôt et ceux de subérine et lignine sur les parois cellulaires et le

rhizoderme pour fournir plus de sites de fixation aux ions Cu2+ et obliger un transport symplastique

même à proximité de l’apex (Li and Leisner, 2008; Caldelas et al., 2012). Nous préconisons d'étudier

l'alcalinisation de la rhizosphère, via l’excretion d’ions OH- et HCO3-, chez les populations de

macrophytes provenant de sites à exposition relativement élevée en Cu. De même, la régulation de

l’homéostasie cellulaire de Cu, via la possible expression modulée des transporteurs COPT1, ZIP

et OsYSL doit être approfondie chez ces populations. Enfin, la contribution de la biominéralisation de

nanoparticules de Cu dans la rhizopshère à la diminution de l'efflux de Cu depuis la rhizosphère vers

les racines doit être envisagée comme participant aux mécanismes de tolérance chez ces populations

de macrophytes.

Références complémentaires

Cambrolle J, Mancilla-Leytón JM, Muñoz-Vallés S, Luque T, Figueroa ME. 2011. Zinc tolerance and accumulation in the salt-marsh shrub Halimione portulacoides. Chemosphere 86, 867–874.

Castruita M, Casero D, Karpowicz SJ, Kropat J, Vieler A, Hsieh S, Yan W, Cokus W, Loo JA, Benning C, Pellegrini M, Merchant SS. 2011. Systems Biology Approach in Chlamydomonas Reveals Connections between Copper Nutrition and Multiple Metabolic Steps. Plant Cell. 23, 1273-1292.

Lagriffoul A. 1998. Biomarqueurs métaboliques de toxicité du cadmium chez le mais (Zea mays L.): mécanismes de tolérance, relations dose-effet et précocité de la réponse. Thèse de doctorat. Université Bordeaux I. 322 pp.

Valipour A, Raman VK, Ghole VS. 2009. A new approach in wetland systems for domestic wastewater treatment using Phragmites sp. Ecological Engineering, 35:1797-1803.

168

Chapitre III Take home message

Les macrophytes d’une même espèce présentent une plasticité de la production de racines en fonction

de la population considérée. Les populations de Juncus effusus, Phalaris arundinacea et

Schoenoplectus lacustris ont aussi une plasticité de la tolérance des racines à l’excès de Cu selon

l’intensité de l’exposition aux PTTE à leur site d’origine. En revanche, les populations d’Iris

pseudacorus et Typha latifolia ont une réponse de la croissance de leurs racines à l’excès de Cu qui

s’apparente à une tolérance constitutive, indépendante du site d’échantillonnage, mais sans pouvoir

conclure définitivement. Le cas de P. australis est plus complexe et nécessite aussi des études

complémentaires pour statuer sur une platicité phénotypique de cette espèce en termes de production

de racines sur un gradient d’exposition à Cu. Les espèces présentant une plasticité phénotypique dans

cette éude ne produisent pas (hémicryptophytes, ici P. arundinacea) ou peu (géophytes à faible

production de biomasse rhizomateuse, ici S. lacustris et J. effusus) de rhizomes. Celles présentant une

réponse similaire à tolérance constitutive appartiennent au groupe des géophytes à forte production de

biomasse rhizomateuse (ici I. pseudacorus et T. latifolia). Les populations provenant de sites aux

faibles concentrations en Cu dans le sol sont plus tolérantes (RTEI plus élevé aux fortes expositions)

que celles échantillonnées sur les sites à forte exposition au Cu. Ces résultats suggèrent une

implication de mécanismes maintenant l’homéostasie au début d’une forte exposition à Cu chez les

macrophytes provenant de sites à faible teneur en Cu dans le sol, malgré la période de pré-culture. Des

hypothèses sont à formuler à partir des mécanismes activés en situation de déficience en Cu (Shin et

al., 2012). Une plasticité d’une composante enzymatique de la cascade antioxydant émerge de nos

travaux, avec une activité moindre de la gaïacol peroxydase chez la population Capanne de P. australis

provenant d’un site contaminé au Cu. Un système de protection et/ou détoxification (e.g.

compartimentage sub-cellulaire, changement histologique dans le rhizoderme, etc.) en amont du

système antioxydant est suggéré chez cette population.

Les résultats sur la plasticité phénotypique des macrophytes exposés à un excès de Cu conduisent à

sélectionner avec plus d’attention les populations utilisées en zones humides construites (CW),

certaines étant plus tolérantes que d’autres au Cu en excès. Après la réalisation d’un état de l’art sur la

phytoremédiation en zone humide construite (partie II), l’accent s’est porté sur la réalisation d’un CW

permettant de décontaminer des eaux à fortes concentrations en Cu. Le système d’étude s’est inspiré

des Bio-Racks développés par Valipour et al. (2009). Les Bio-Racks sont un ensemble de colonnes

perforées (avec ou sans substrat) et plantées de macrophytes, permettant un contact direct des racines

169

(passant au travers des perforations) avec l’eau, mais aussi le développement de micro-sites pour les

microorganismes, dans le substrat ou sous forme de biofilms. Dans notre étude, les Bio-Rack sont été

plantés de Juncus articulatus, Phalaris arundinacea et Phragmites australis, trois espèces dont les

populations ont été échantillonnées sur un site à forte contamination au Cu (i.e. zone humide de La

Cornubia, Bordeaux). La contribution de chaque espèce de macrophytes pour décontaminer une eau

additionnée de sulfate de Cu a été quantifée, et comparée à un CW de type Bio-Racks mais sans plante.

NB: Le choix des macrophytes utilisés pour les pilotes de CW type Bio-Racks a été réalisé en début de thèse,

avant d’avoir obtenu les résultats sur la plasticité phénotypique de P. arundinacea, S. lacustris et J. effusus.

170

Chapitre IV

Traitement d’eaux contaminées par Cu en zones humides construites de type

Bio-Racks plantées de Phragmites australis, Phalaris arundinacea et Juncus

articulatus

Cette partie sera soumise à Ecological Engineering

171

a b

c

Figure 34 : Zones humides construites de type Bio-Racks mises en service en serres (INRA,Villenave d’Ornon, Gironde, France). a. Début de la saison de végétation (Mars) ; b. Pic de la saison de végétation (Juin): au premier plan, un des trois contrôles non plantés, suivi par les 3 mésocosmes plantés de Juncus articulatus et les 3 mésocosmes plantés de Phalaris arundinacea. Les 3 mésocosmes plantés de Phragmites australis sont visibles à la droite de la photo c. Pilot constructed wetlands (Bio-rack type) under a greenhouse (INRA, Gironde, France): a. start of the growing season (March) ; b. top of the growing season (June) : in front, unplanted CW, then three CW planted with Juncus articulatus and three CW planted with Phalaris arundinacea. The three CW planted with Phragmites australis are located on the right of the pic c.

172

Copper removal from water using a bio-rack system planted with Phragmites australis, Juncus articulatus and Phalaris arundinacea

Marchand Lilian, Nsanganwimana Florien, Oustrière Nadège, Grebenshchykova Zhanna, Mench Michel

UMR BIOGECO INRA 1202, Ecologie des Communautés, Université Bordeaux 1, Bât. B2, RdC Est, Avenue des Facultés, 33405 Talence, France, and INRA, 69 Route d’Arcachon, 33610 Cestas, France.

Abstract A bio-rack system was developed for treating Cu-contaminated freshwaters. Twelve mesocosms, i.e.

pilot constructed wetlands (CW) (110 dm3), each containing 15 perforated vertical pipes filled with a

mixture of gravel (diorite; 80%) and perlite (20%) and assembled as a rack, were operated between

March and June. The CW (in triplicates) were planted with Phragmites australis, Phalaris

arundinacea and Juncus articulatus, and unplanted as control. Phragmites australis, P. arundinacea

and J. articulatus were sampled at a Cu-contaminated site. The CW were filled with a mix of

freshwater from the Jalle d’Eysines river (30%) and tap water (70%). Water was spiked with Cu (2.5

µM, 158.5 µg L-1). Three CW batches were carried out, i.e. in early spring (March, CW#1), beginning

of the growing season (May, CW#2), and peak growing season (June, CW#3). For each batch, water

was recirculated in the CW during 14 days. Physico-chemical parameters (pH, electrical conductivity,

redox potential, BOD5 and Cu2+ concentrations) were measured every three days. Water pH of both

CW#1 and #2 ranged between 7.8 and 8.5 in all the modalities across the experiment. The CW#3 water

was acidified to 6 at day 0. The pH rose to 7.8 after 6 days in all the modalities. Total Cu concentration

was analyzed at the start and end of each experiment. Free Cu2+ removal in CW#1 was <10% for all

treatments and increased to 77% in CW#2 for P. arundinacea. In acid conditions (CW#3), Cu2+

removal was 99% for all treatments. In March and May, highest total Cu removal occurred in CW

planted with P. arundinacea (respectively 52 and 68%). For CW#3, total Cu removal peaked up to

90% in the unplanted CW. The RTEI calculation suggested no beneficial effect of macrophytes on Cu

removal at short term. Conversely, the CW planted with J. articulatus generally displayed a lower

efficiency. The lowest value for total Cu concentration in water after the 14-day period was 13 µg L-1

in CW#3 unplanted and planted with P. arundinacea.

Keywords : biofilm, constructed wetland, decontamination, macrophyte, phytoremediation

173

I. Introduction

Water quality issues are a major challenge faced by mankind in the 21st Century. Treatment of

municipal wastewater streams aims at eliminating nutrients (N and P), pathogenic microbes, persistent

organic pollutants (POPs, e.g. polycyclic aromatic hydrocarbons, polychlorinated biphenyls and

pesticides), xenobiotics derived from the pharmaceutical industry (personal care products, hormones

and pharmaceuticals) and potentially toxic trace elements (PTTE, e.g. Cu, Zn, Fe, Cd, Ni, Pb, Hg, Cr,

Sr, Al, Ba, Se and As) from wastewater streams (Schwartzenbach et al., 2010). In industrialized

countries, connection to municipal wastewater treatment plants ranges from 50% to 95%, whereas

more than 80% of the municipal wastewaters in low-income countries are discharged without any

treatment, polluting rivers, lakes, and coastal sea areas (UNESCO, 2009). Potentially toxic trace

elements and POPs are reduced through implementation of internal water recycling and recovery

systems and end-of-pipe treatment using advanced technologies, such as activated carbon, advanced

oxidation, and membrane processes in industrialized countries. Such processes however are expensive

in complex maintenance operations and energy, and sometimes ineffective, mainly in developing

countries. In India, hardly 10% of the sewage generated is treated effectively, while the rest finds its

way into natural ecosystems and contaminates large scale aquatic ecosystems (Trivedy and Natake,

2001) where pollutants may accumulate in surface waters, groundwater, substrates and plants (Aksoy

et al., 2005; Demirezen et al., 2007; Lizama et al., 2011). To face these contaminations, water

deserves a relevant imagination and concrete actions (Braga, 2012). Constructed wetlands (CW)

planted with macrophytes are an emerging phytotechnology frequently used as an efficient and cost-

effective alternative for treating wastewater streams due to its low energy requirements, its convenient

operation, and its weak maintenance (Marchand et al., 2010; Hsu et al., 2011; Kuschk et al., 2012).

Five mechanisms affect metal removal in natural and constructed wetlands (Sheoran and Sheoran,

2006; Lesage et al., 2007; Marchand et al., 2010): (1) sorption to fine textured sediments and organic

matter (2) (co)precipitation as insoluble salts, mainly sulphides in reducing conditions and Fe/Mn/Al

(oxy)hydroxides in oxidative conditions, (3) carbonate (co)precipitation, (4) absorption and induced

changes in biogeochemical cycles by plants and associated micro-organisms (fungi and bacteria of the

rhizosphere as well as endophytes) and (5) deposition of suspended solids due to low flow rates. All

these reactions lead to metal accumulation in the wetland substrate. The CW efficiency depends on

inlet metal concentrations, hydraulic loading, and pH and redox conditions (Kadlec and Knight, 1996).

Macrophytes are key-players in CW. Plant-derived organic matter over time continuously provides

sites for metal sorption, as well as carbon sources for bacterial metabolism, thus promoting long-term

174

functioning (Jacob and Otte, 2004). Macrophytes drive PTTE uptake and storage in roots and rhizomes

(Caldelas et al., 2012) as well as PTTE immobilization onto the root plaque (McCabe et al., 2001) and

into the sediment by releasing organic metal ligands (Ryan et al., 2001). They also contribute to

maintain oxidative conditions in the rhizosphere (Armstrong, 1978). Valipour et al. (2009) have

proposed a new CW design, the so called bio-rack, planted with Phragmites australis. This system

provides a high plant density and a high surface area for microbial growth. It also reduces the problem

of clogging, has a low space requirement, and exhibits a high rate of organic degradation at low

hydraulic retention time (Valipour et al., 2009). To our knowledge, no experiments have been yet

reported on PTTE removal in bio-backs. This study aimed at assessing the efficiency of such bio-rack

system to remove Cu from freshwater.

II. Materials and Methods

2.1 Pilot plant setup

A pilot plant was built in a greenhouse located at the National Institute for Research in Agronomy

(INRA, Villenave d’Ornon, 44° 46′ 50″ N 0° 33′ 57″ W, France) and started its operation in January

2011. Experiments were carried out in twelve independent polyethylene tanks (32x56 cm² surface

opening, tapering to 36 cm depth; containing a 60 dm3 volume at 34 cm operational water depth). Each

tank was connected to a storage tank (total volume: 60 dm3, 50 dm3 at operational water depth). For

convenience, we will refer to the polyethylene tanks as “unit A”, storage tank as “unit B” and the

whole as a pilot constructed wetland (CW) throughout the paper. The scheme of the pilot plant is given

in Figure 35. In December 2010, each unit A was filled with 15 vertical PVC pipes (diameter: 10 cm,

H: 35 cm, V: 2.75 dm3) assembled as a rack termed as “bio-rack”. All vertical pipes were perforated

every five cm in height and width (hole diameter: 5-10 mm) to enable liquid transport and root

development out of the pipe, and filled with a homogeneous mix of gravels (diorite, 1-5 mm, 3.9 kg)

and perlite (0.035 kg) (respectively 80% v/v and 20% v/v). Porosity of the substrate made of gravels

and perlite was 36%, thus the water volume into each pipe was 1 dm3. Each Bio-Rack occupied 41 dm3

of the unit A and contained 15 dm3 of water and 26 dm3 of substrate. The 19 dm3 remaining in each

unit A were occupied by free water. Total water volume in each unit A was thus 34 dm3. Water

volume in each unit B was 50 dm3. The units A were connected to the units B using lift pumps

(Vc400ech, 7500 dm3 h-1, Leroy Merlin) (Figure 35). Total water volume in each CW was 84 dm3.

175

Figure 35: Scheme of the constructed bio-rack wetland (adapted from Valipour et al., 2009).

2.1.1 Plants

Plants were collected in 2010 at the beginning of the growing season (April-May), at the La Cornubia

site (44°54'26"N; 0°32'46"W, Bordeaux, France), a former chemical plant producing Cu sulphate and

fungicides, dating back to a century, closed in 2004. Phalaris arundinacea L. and Juncus articulatus

L. were sampled in an abandoned constructed pond colonized by macrophytes, connected to a pipe

collecting storm water and effluents from the plant. Total soil Cu and soil pH at this sampling site are

respectively 205 mg kg-1 DW and 5.7. Phragmites australis (Cav.) Trin. ex Steud. were sampled on

the riversides of a small creek that borders this chemical plant, contaminated by effluents, run off,

storm water and dust fallout. Phragmites australis and J. articulatus are rhizomatous geophytes; they

have shoots borne from buds in the soil and resting buds lying beneath the soil surface as rhizomes.

Phalaris arundinacea is a hemicryptophyte; it exhibits buds either at or near the soil surface

Phragmites australis/Juncus articulatus/Phalaris arundinacea

Inlet

Roots and rhizomesBio-rack

Bio-Racks

outlet

Lift pump

Unit B (60 dm3, , operational water volume 50 dm3 )

Levelregulator

Unit A ( 60 dm 3). 41 dm 3 : Racks. 19 dm 3: Free water

32 cm

36 cm

56 cm

35 cm

10 cm

Op

era

tio

na

lwa

ter

de

pth

: 34

cm

30 cm

61 cm

36 cm

176

(Raunkiær, 1934). Samples of plant populations were separately kept in buckets and immediately

transported to a greenhouse (INRA – Centre Bordeaux Aquitaine, Villenave d’Ornon (Day (9-21h)

1911±1232 µM photons m-2s-1, 28±5°C, Night (21h-9h) 19±3°C). The next day, rhizomes and/or stems

bearing buds were cut into small pieces (10-20 cm). They were then individually grown in plastic pots

placed in polyethylene bowls (volume: 60x40x15 cm3) containing perlite imbibed with a quarter

Hoagland nutrient solution (HNS, Hoagland and Arnon, 1950): KNO3 (1.62 mM), Ca (NO3)2 (0.69

mM), NH4H2PO4 (0.25 mM), MgSO4 (0.5 mM), H3BO3 (1156 µM), MnCl2 (2.29 µM), Cu.SO4 (0.08

µM), (NH4)6Mo7O24 (0.13 µM), ZnSO4 (0.19 µM), and FeSO4 (48.6 µM). Water volume was

maintained constant and monthly changed to avoid anoxia. 1 dm3 of a quarter HNS was added every

month during the growing season and every two months during winter to avoid nutrient depletion in

the growing medium. In January 2011, rhizomes and stems bearing buds were again cut into small

pieces (5-10 cm). Three unplanted CW were used as controls. Other CW were planted (in triplicates),

respectively with J. articulatus, P. arundinacea and P. australis. Two plant pieces were transplanted 5

cm beneath the substrate surface of each vertical pipe. Units A were then filled with tap water (34

dm3). Between January 2011 and March 2012, the water was changed every six weeks to avoid the

system anoxia and 2 dm3 of a quarter HNS were added every six weeks to avoid nutrient depletion.

Sodium and Fe salt of EDDHA (Sequestrene, Syngenta) was added in units A once at mid-June (10 mg

dm-3, 23 µM). Water level was maintained in the system by tap water addition. In December 2011, the

dried aboveground biomass was harvested at 20 cm above the Bio-Rack surface. These CW were

carried out from January 2011 to March 2012 for allowing plants to produce sufficient belowground

biomass in such greenhouse conditions.

2.1.2 Sampling and analysis

In February 2012 (weeks 8 and 9), CW were filled with tap water and a quarter HNS, and spiked with

1 µM Cu (63.5 µg Cu dm3 by addition of 21 mg CuSO4 in 84 dm3) for allowing progressive

macrophyte adaptation to Cu contamination and Cu2+ sorption on binding sites. Then, CW were filled

again with tap water and 2 dm3 of a quarter HNS during weeks 10 to 12. The day prior the experiment,

CW were emptied, and filled with 70% of tap water and 30% of freshwater from the Jalle d’Eysine

river. It is an urban river located in southwest France (44°53′36″N; 00°40′40″O), North of Bordeaux,

and a tributary of the Garonne River. From its headwaters to its confluence with the Garonne River, it

is 32 km long. Water depth typically varies from 0.8 to 2.5 m annually, and average water flow is

3 m3 s-1. It receives PTTE-contaminated runoff from industrial, agricultural and residential areas and

effluents from two major municipal wastewater treatment plants (WTP) that serve more than 100,000

177

inhabitants in the Bordeaux suburbs. Treated effluents can account for up to 33% of the river flow

(Labadie and Budzinski, 2005). Electrical conductivity (EC) in this freshwater ranges from 0.48 to

0.94 mS cm-1, pH from 6.8 to 7.6, and Cu concentrations are below the detection limit (<8 µg dm3)

(Chapter II). Addition of freshwater in the CW provided nutrients and microorganisms and avoided to

supply HNS. Conditions in the CW were thus closer to contaminated freshwaters. Experiments were

carried out during 14 days, and water was daily recirculated 10 h day-1, from 8h30 am to 18h30 pm.

Inlet and outlet water flows were 5.3 dm-3 min-1. For the 1st experiment (CW#1, weeks 13 and 14), the

1st day (23/03/2012), CW were spiked with 2.5 µM Cu (158.5 µg Cu dm-3 by addition of 52.4 mg

CuSO4 in 84 dm3). Immediately (T0), four water samples (100 cm3) were collected in all CW

(respectively, two in unit A and two in unit B), resulting in a total of 48 samples (12 per modality).

Such water sampling was repeated on days 3, 6, 10 and 14. The O2 concentration, EC and redox

potential were measured with a WTW Multiline P4 meter (Germany) in 24 samples, i.e. six for each

modality, at T0 as well as pH (Hanna instruments, pH 210, combined electrode Ag/AgCl, USA) and

Cu2+ concentrations (Fischer Bioblock Cupric Ion Electrode, USA) after addition of 2 cm3 of NaNO3

(5 M). All measurements were performed on the day of water sampling. The remaining 24 samples,

i.e. six for each modality, were kept at 20 °C, in dark conditions, for 5 days, then O2 concentration (mg

dm-3) was measured at T5. The 5-day biological oxygen demand (BOD5) was calculated as

BOD5 = [O2] T0 – [O2] T5. Total Cu concentration was analyzed in these water samples by ICP-AES

(Varian liberty 200) and ICP-MS (Thermo X serie 200) at the INRA USRAVE laboratory, Villenave

d'Ornon, France. At day 15, CW were emptied, and rinsed two times with tap water. Units B and

pumps were rinsed a third time, wiped with absorbing paper and put to dry during two days. Then, CW

were filled again with tap water and 2 dm3 of a quarter HNS. This experiment was duplicated during

weeks 19 and 20 (CW#2), when the growing season peaked under the greenhouse conditions. The

third experiment (CW#3) was carried out during weeks 26 and 27, but the water (i.e. 1/3 freshwater +

2/3 tap water) was acidified with 1M HNO3 the day prior the addition of 2.5 µM Cu. Nitric acid was

progressively added into all CW to reach a pH of 5.5. After one reaction day, pH at T0 rose up to 6.1

(Table 21).

178

Table 21: Changes in physic-chemical parameters, i.e. pH, Eh (mV), Ec (µS cm-1) and DBO5 (mg dm-3) in CW during a 14-day period (n=6).

pH Eh

mV

CW#1 (Cu =2.5 µM)

Control J. articulatus P. arundinacea P. australis Control J. articulatus P. arundinacea P. australis

day 0 8.0±0.07 8.0±0.02 7.8±0.07 7.9±0.05 218.6±1.9 220.2±1.2 224.6±2.6 222.8±1.5

day 3 8.4±0.03 8.4±0.03 8.2±0.2 8.3±0.07 193.5±10.7 200±5.06 201.2±5.5 194.2±4.9

day 6 8.3±0.03 8.2±0.05 8.1±0.2 8.1±0.1 179.2±3.1 186±2.5 185.6±5.6 190.8±4.5

day 10 8.5±0.1 8.6±0.02 8.5±0.09 8.5±0.05 203.3±3.6 204.5±2 206.6±4.0 205.2±2.1

day 14 8.4±0.04 8.3±0.04 8.3±0.03 8.3±0.06 172.8±1.5 174.8±1.9 176.3±2.6 182.8±3.6

CW#2 (Cu = 2.5 µM)

Control J. articulatus P. arundinacea P. australis Control J. articulatus P. arundinacea P. australis

day 0 7.9±0.08 7.9±0.2 7.9±0.1 7.9±0.05 232.8±3.6 232.3±3.4 237.2±5.3 235.3±3.9

day 3 8.4±0.1 8.4±0.03 8.2±0.1 8.4±0.02 221±4.1 219.2±1.0 227.6±3.9 222.6±0.5

day 6 8.3±0.03 8.2±0.04 8.1±0.1 8.1±0.1 228.6±1.5 234.2±1.8 231.3±3.2 231.3±1.7

day 10 8.3±0.04 8.3±0.06 8.1±0.06 8.2±0.1 233±2.7 240.3±3.4 236.0±3.5 234±3.6

day 14 8.4±0.08 8.4±0.05 8.4±0.04 8.4±0.03 210.5±2.2 213.6±1.9 206.2±2 210.3±1.5

CW#3

(Cu = 2.5 µM) Control J. articulatus P. arundinacea P. australis Control J. articulatus P. arundinacea P. australis

day 0 6.1±0.2 6.1±0.1 6.2±0.2 6.1±0.1 215.1±12.6 218.5±5.8 229.0±6.1 235.2±24.5

day 3 7.4±0.3 6.8±0.2 7.1±0.1 6.9±0.1 233.7±15.9 246.8±4.8 238.8±3.5 238.2±2.4

day 6 7.8±0.1 7.8±0.1 7.8±0.3 7.8±0.1 222.8±1.9 224.7±2.7 224.5±2.6 225.3±8.9

day 10 7.9±0.1 7.9±0.1 7.9±0.1 8.0±0.1 217.4±2.8 218.5±2.5 215.5±3.7 210.3±5.8

day 14 7.8±0.1 7.7±0.1 8.1±0.1 7.9±0.1 218.6±6.2 224.3±2.9 218.8±3.9 212.8±2.8

EC BOD5

µS cm -1 mg O2 dm-3

CW#1 (Cu =2.5 µM

Control J. articulatus P.arundinacea P. australis Control J. articulatus P.arundinacea P. australis

day 0 441±1.5 448±2.8 447.2±4.5 441.6±1.7 1.2±0.1 1.4±0.18 1.4±0.2 1.4±0.3

day 3 445.2±1 463.5±11.2 456.3±6.9 454±4.9 <1 <1 <1 <1

day 6 461.3±2.9 487.3±6.6 468.1±5.3 470.3±4.4 0.2±0.7 <1 <1 <1

day 10 477±7.6 502.5±8.9 478.3±6.2 487.5±3.8 <1 <1 1±0.2 <1

day 14 487.6±5.7 521.5±14.3 487.3±7.4 501.8±5 <1 <1 <1 <1

CW#2 (Cu = 2.5 µM)

Control J. articulatus P.arundinacea P. australis Control J. articulatus P.arundinacea P. australis

day 0 425.3±3.9 439±20.7 429.5±3.4 430±6.4 <1 <1 <1 <1

day 3 439.5±2.2 469.3±19.3 441.3±1.8 457.2±14.7 <1 <1 <1 <1

day 6 447±8.7 470.5±12.6 448.3±16.0 463.8±12.7 <1 <1 <1 <1

day 10 442±1.4 469.2±5.7 439.3±2.9 459.6±16.3 <1 1.0±0.5 <1 <1

day 14 451.3±2.7 485.5±9.8 449.5±3.9 471±20.1 <1 <1 <1 <1

CW#3

(Cu = 2.5 µM) Control J. articulatus P.arundinacea P. australis Control J. articulatus P.arundinacea P. australis

day 0 706.7±67.8 562.8±8.1 554.8±16.7 574.6±6.9 2.01±0,1 2.1±0.3 2.5±0.2 2.23±0.4

day 3 756.1±85.3 577.7±0.9 573.5±15.8 581.5±7.9 <1 <1 <1 <1

day 6 796.2±90.3 617.0±9.6 598.8±7.6 598.8±15.3 <1 <1 <1 <1

day 10 826.2±91.5 654.3±11.5 623.8±14.8 628.8±14.8 <1 1.0±0.5 <1 <1

day 14 850.2±96.5 681±14.5 653.2±10.7 642.8±6.4 <1 <1 <1 <1

179

2.2.3 Statistical analysis

The Relative Treatment Efficiency Index (RTEI) was used to assess the macrophyte effect on Cu

removal in CW planted with P. arundinacea, J. articulatus and P. australis compared to unplanted

CW (Marchand et al., 2010).

where T (%) is the Cu removal percentage in a planted CW and C (%) the Cu removal in an unplanted

CW (Table 22).

The effect of macrophytes planted in Bio-Racks during the 14-day period on Cu2+ (Figure 36) and

total Cu (Figure 37) removal was tested using a one-way analysis of variance (ANOVA). Normality

and homoscedasticity of residuals were met for all tests. Post-hoc Tukey HSD tests were performed to

assess multi-comparisons of means. Differences were considered statistically significant at p<0.05. All

analyses were carried out using R software (version 2.14.1 R foundation for Statistical Computing,

Vienna, Austria).

III. Results and discussion

Conventional root zone system in CW using multilayered bed from stone to mud may result in

clogging of interstices (Valipour et al. 2009). The roots of P. australis can poorly grow in such bed

conditions due to perforation difficulties (Davison et al., 2005). A modified root zone system with a

distinguished surface providing sites for the growth of microbial bio-films can facilitate either the

sorption or degradation of many pollutants through (bio)chemical reactions (Valipour et al., 2009). In

our bio-racks (Figure 35), vertical pipes filled with a mix of gravels and perlite provided potential

sorption sites for Cu onto the substrate, free roots and rhizomes (including their Fe/Mn plaques)

growing through the holes, microorganisms and their biofilms. This system allowed removing up to

99% of the Cu2+ and 90% of the total Cu from Cu-contaminated waters under acid conditions (Figure

36, 37, Table 22).

180

Table 22: Rates of Cu 2+ and total Cu removals (%), and RTEI of bio-racks during a 14-day period (from

23/03 to 05/04: CW#1, 03/05 to 17/05: CW#2, and 22/06 to 06/07: CW#3 in 2012).

Treatments Day % Removal RTEI % Removal RTEI % Removal RTEI

CW#1 CW#2 CW#3

Cu 2+ Control 3 73 - 81 - 77 -

Juncus articulatus 3 41 -0.3 71 -0.1 82 0

Phragmites australis 3 64 -0.1 84 0 87 0.1

Phalaris arundinacea 3 41 -0.3 83 0 87 0.1

Control 6 46 - 85 - 86 -

Juncus articulatus 6 12 -0.6 69 -0.1 97 0.1

Phragmites australis 6 23 -0.3 84 0 98 0.1

Phalaris arundinacea 6 36 -0.1 84 0 98 0.1

Control 10 48 - 71 - 99 -

Juncus articulatus 10 31 -0.2 54 -0.1 99 0

Phragmites australis 10 57 0.1 67 0 99 0

Phalaris arundinacea 10 57 0.1 64 -0.1 99 0

Control 14 10 - 62 - 99 -

Juncus articulatus 14 <0 -1 47 -0.1 99 0

Phragmites australis 14 10 0 72 0.1 99 0

Phalaris arundinacea 14 2 -0.7 77 0.1 99 0

Treatments Day % Removal RTEI % Removal RTEI % Removal RTEI

CW#1 CW#2 CW#3

Total Cu Control 14 40 - 49 - 90 -

Juncus articulatus 14 11 -0.5 11 -0.6 82 -0.1

Phragmites australis 14 7 -0.7 35 -0.2 85 0

Phalaris arundinacea 14 52 0.1 68 0.2 87 0

3.1 Physico-chemical parameters and Cu removal in bio-racks during a 14-day period in alkaline

conditions

The CW#1 and CW#2 experiments were carried out at pH=8, in early spring and the beginning of the

growing season. At day 0, for both experiments, pH was in the range [7.8±0.07 - 8±0.07] in all CW

(Table 21). Then, it slightly increased at day 3 to [8.2±0.2 - 8.4±0.07] and remained stable until the

exposure end. At the end of both experiments (day 14), pH varied between [8.3±0.06 - 8.4±0.08].

Therefore, pH conditions were similar for all CW during CW#1 and CW#2. Redox potential ranged

respectively between [172.8±1.5 - 218.6±1.9 mV] during the CW#1 and [210.5±2.2 - 233±2.7 mV]

during the CW#2. Redox conditions were oxidative in all CW treatments and slightly increased as a

function of time (Table 21). Electrical conductivity also slightly increased; it respectively ranged from

181

441±1.5 to 487.6±5.7 µS cm-1 for the CW#1 and between 425.3±3.9 and 451.3±2.7 µS cm-1 for the

CW#2. Total Cu was respectively 0.7 µM (44±9 µg dm-3) and 0.8 µM (48±7 µg dm-3) in CW at the

beginning of both first experiments (Figure 37). The major part (70%) of the 2.5 µM Cu added in

CW#1 and CW#2 was immediately sorbed as insoluble form onto the perlite, Fe oxides of diorite, the

microbial biofilms, macrophyte roots and PVC layers. Consequently, both Cu removal (in %) and Cu2+

removal (in %) were calculated respectively based on the total Cu and the free Cu2+ in water at day 0.

In CW#1, total Cu removal at day 14 reached 40% in the unplanted CW, 11% in the CW planted with

J. articulatus (RTEI=-0.5), 7% in the CW planted with P. australis (RTEI = -0.7) and 52% in CW

planted with P. arundinacea (RTEI = 0.1) (Table 22). Similar trend occurred for the CW#2, the lowest

and the highest total Cu removals being respectively obtained in CW planted with J. articulatus (11%,

RTEI = -0.6) and P. arundinacea (68%, RTEI = 0.2). Total Cu removal in the CW planted with P.

australis increased during the CW#2 (35%, RTEI = - 0.2). Highest total Cu concentrations in water at

day 14 were found for both CW#1 and CW#2 in CW planted with J. articulatus (respectively, 0.8 µM,

48.3±9 µg Cu dm-3 and 0.75 µM, 47.5±5 µg Cu dm-3) while lowest concentrations in water occurred in

CW planted with P. arundinacea (respectively 0.3 µM, 19.7±18 µg Cu dm-3 and 0.2 µM, 15.5±7 µg

Cu dm-3) (Figure 37). Similar patterns were quantified for free Cu2+ removal. At the beginning of

CW#1 and CW#2, free Cu 2+ concentration in water was 0.025 µM (1.6±0.13 µg dm-3). At day 14 of

CW#1, water Cu2+ concentrations did not differ across treatments (from 0.02 µM, 1.23±0.11 µg Cu

dm-3 in control to 0.03 µM, 1.81 ±0.17 µg Cu dm-3 in CW planted with J. articulatus). These values

were similar to the initial concentrations (day 0) (Figure 36). Thus, Cu2+ removal ranged from 0% in

CW planted with J. articulatus (RTEI =-1) to 10% in both unplanted CW and CW planted with P.

australis (RTEI=0) (Table 22). For CW#2, Cu2+ removal was more efficient in all treatments. At

day14, free Cu2+ removal reached 77% in CW planted with P. arundinacea (RTEI=0.1, final

concentration: 0.01 µM, 0.6±0.05 µg Cu2+ dm-3), while its lowest value was found for CW planted

with J. articulatus (17%, RTEI=-0.1, final concentration: 0.014 µM, 0.9±0.1 µg Cu2+ dm-3) (Table

22).

The pH influences the efficiency of metal removal in wetlands (Sheoran and Sheoran, 2006; Marchand

et al., 2010). When pH falls in the [8-9] range, Cu coprecipitation with sulphides and/or hydrogen

sulphides may occur under reducing conditions in the presence of OM (Brookins, 1988; Sheoran and

Sheoran, 2006).Under oxidative conditions, Cu-coprecipitation with (oxyhydr)oxides takes place in

CW at pH>7 (Brookins, 1988; Sheoran and Sheoran, 2006) (Supplementary material, Figure 38).

The overall mean surface charge of ferric (oxyhydr)oxides changes from a positive to a negative value

as pH increases (Sheoran and Sheoran, 2006). Here, water pH values of CW#1 and CW#2 were 8, and

182

redox conditions were oxidative due to water recirculation. Therefore Cu-coprecipitation with

sulphides could not explain Cu removal, but it may occur with (oxyhydr)oxides.

Copper in our CW may also sorb onto Fe oxides from diorite and react with silicates derived from the

perlite to form hydrated Cu silicates (CuSiO3.nH2O). Perlite was used for removing free Cu2+ ions

from aqueous solutions (Alkan and Dogan, 2001; Sari et al., 2007). The amount of Cu2+ adsorbed rose

as pH increased, whereas it diminished as the ionic strength, temperature and acid activation increased.

Sari et al. (2007) reported a monolayer adsorption capacity of 8.62 mg Cu2+g-1 on perlite. Here, the

perlite in each Bio-Rack amounted to 534 g and thus 4 603 mg Cu2+ could adsorb on it in slightly

alkaline conditions. All the Cu2+ added in CW#1 and CW#2 was not removed at day 14, and the Bio-

Rack design might be involved. Diffusion through the substrate and contact with perlite may be limited

by the PVC pipes. However, such limitation occurred also in CW#3, whereas Cu removal increased,

and can be likely ruled out. Other inorganic compounds may form such as Cu3(PO4)2 and CuCO3 since

phosphates and carbonates were present in the mixture of freshwater and tap water and in the HNS.

Phragmites australis and P. arundinacea produce a large belowground biomass (Adams and

Galatowitsch, 2005; Asaeda et al., 2006). Hence, they may take up in roots a significant Cu amount

(Bonanno and Lo Giudice, 2010) and provide binding sites for Cu sorption onto the Fe/Mn plaque

(Otte et al., 2004; Kissoon et al., 2010). Copper could also be immobilized under metallic

nanoparticles in and near roots with assistance of endomycorrhizal fungi (Manceau et al., 2008).

Additionally, graminaceous macrophytes such as P. arundinacea and P. australis excrete

phytosiderophores which may complex free Cu2+ (Tsednee et al., 2012). However, for CW#2 the

removal rates of total Cu and free Cu2+ were similar in the unplanted CW and CW planted with P.

australis and P. arundinacea (Table 22). Moreover, on the whole 14-day period, Cu2+ concentration

was low (<10%) compared to total Cu concentration in all treatments. Consequently, Cu in recirculated

water would mainly occur as CuO, Cu2O and Cu bound with dissolved organic matter (DOM). This

adds to conflicting data regarding plant influence on metal removal from water in studies comparing

planted and unplanted CW (Lee and Scholz, 2007; Marchand et al., 2010). Alternatively, similar rates

of Cu removal in the unplanted CW and CW planted with P. australis and P. arundinacea in CW#2

may be due to the microbial biofilm established in the bio-racks before the experiments with Cu

(Huang et al., 2000; Toner et al., 2005; Tanner and Headley, 2011; Valipour et al. 2009). Microbial

biofilms can trap Cu oxides such as Cu2O and CuO (Paradies et al., 1996). Several months prior the

experiments, unplanted and planted bio-racks were supplied with a quarter HNS for allowing the

macrophyte growth. Microbial biofilm may better develop in the unplanted CW without root

183

competition for nutrients. Conversely, macrophyte roots can provide organic compounds such as

polysaccharides to promote bacteria growth (Marchand et al., 2010). The DOM and microbial biofilms

in unplanted CW may bind Cu2+ and trap Cu oxides. In planted CW, the DOM, microbial biofilms,

macrophyte roots and root exudates can have similar effects. The higher Cu2+ removal rate of CW#2

compared to CW#1 may be due to higher plant and microbial activities.

Juncus articulatus may maintain Cu in solution through Cu complexation with LMMOA (low

molecular mass organic acid) such as oxalate released by Juncus maritimus with rising Cu exposure

(Mucha et al., 2010) and consequently may decrease total Cu removal. Indeed, total Cu but also Cu2+

removals were lower in CW planted with J. articulatus than in other treatments. In addition, J.

articulatus can release allelopathic compounds (Dakora and Phillips, 2002; Ervin and Wetzel, 2003)

which may constrain the biofilm development (Zhang et al., 2009). Moreover, J. articulatus produces

less belowground biomass than P. australis and P. arundinacea, which may result in less sorption sites

and root Cu uptake.

Macrophytes used in this study were sampled at a Cu contaminated site. Previous finding reported in

chapter III showed that the root production of P. arundinacea, P. australis and Juncus spp. is not

constrained by a 2.5 µM Cu exposure. However Cu exposure may increase and at 25 µM Cu root

growth rate differs between populations, depending on Cu exposure at sampling site for

hemicryptophytes such as P. arundinacea and rhizomatous geophytes with low belowground biomass

such as Juncus spp. (Chapter III). Therefore, as suggested by Brisson and Chazarenc (2009) and

Marchand et al. (2010), CW designs for metal removal may consider the selection of macrophyte

species and ecotypes. However it is not the only criterion. In our study, P. australis suffered aphid

attacks in both 2011 and 2012, leading to a massive death of plants in CW at the end of growing

season (weeks 35-40) (data not shown) whereas P. arundinacea and J. articulatus were not affected by

aphids. Such P. australis low resistance to aphids at the end of growing season in temperate climates

could be included in plant selection strategies.

184

0 3 6 10 14

Day

0.0

0.5

1.0

1.5

2.0

2.5

3.0

ControlJuncus articulatusPhalaris arundinaceaPhragmites australis

Cu2+ (µg dm-3)

B

0 3 6 10 14

Day

0.0

0.5

1.0

1.5

2.0

2.5

3.0

ControlJuncus articulatusPhalaris arundinaceaPhragmites australis

Cu2+ (µg dm-3) A

0 3 6 10 14

Day

0

20

40

60

80

100

120

140

ControllJuncus articulatusPhalaris arundinaceaPhragmites australis

C

a a a a b b b b b a b a b b b b a a a a a a a a d d d d d c d d d b c c c b d c

b a a a c c c c c c c c c c c c c c c c

Figure 36: Water Cu2+ concentrations (µg dm-3) in CW during a 14-day period. (A) CW#1, (B) CW#2 and (C) CW#3 (n=6). The different letters stand for

statistical significance at the 0.05 level with the Tukey HSD test.

Cu2+ (µg dm-3)

185

0 14

Day

0

20

40

60

80

100 Control Juncus articulatus Phalaris arundinaceaPhragmites australis

B

Total Cu (µg dm-3)

0 14

Day

0

20

40

60

80

100 Control Juncus articulatus Phalaris arundinaceaPhragmites australis

A

0 14

Day

0

50

100

150

200

250 ControlJuncus articulatusPhalaris arundinaceaPhragmites australis

C

a b ab a c b c d a a a a b a b b

a a b ab c d c c

Figure 37: Total Cu concentrations (mg dm-3) in CW water at day 0 and day 14 (n=4): (A) CW#1,

(CW#2) and (CW#3). The different letters stand for statistical significance at the 0.05 level with the Tukey

HSD test.

Total Cu (µg dm-3)

Total Cu (µg dm-3)

186

3.2 Physico-chemical parameters and Cu removal in bio-racks during a 14-day period in

acid conditions

In CW#3, pH was set to 6.1 at day 0. It progressively increased – similarly in all treatments –

from 6.8±0.2 (J. articulatus) to 7.4±0.3 (control) at day 3, reached 7.8±0.3 at day 6, and was

in the [7.8±0.1 - 8.1±0.1] range for all treatments at day 14 (Table 21). Simultaneously, redox

potential did not significantly vary, staying in the 210.3±5.8 mV (P. australis at day 10) -

238.2±2.4 mV (P. australis at day 3) range (Table 21). Cu in water would be thus mainly in

Cu2+ form for CW#3 (supplementary material: Figure 38). Electrical conductivity increased

in all CW, from 554.8±16.7 to 850.2±96.5 µS cm-1 in CW respectively planted with P.

arundinacea at day 0 and unplanted at day 14. Unplanted CW presented the highest EC

values during the whole CW#3 experiment. This likely resulted in metal desorption under

acid conditions. Total Cu concentrations in water increased in all treatments, from 1.6 µM

(103.5±26.8 µg dm-3) to 2.85 µM (181.4±48 µg dm-3) respectively in CW planted with P.

australis and J. articulatus at day 0. Free Cu2+ concentrations in CW water also increased

compared with CW#1 and CW#2 to reach 1.4 µM (90.1±48 µg dm-3) in CW planted with J.

articulatus, P. arundinacea and P. australis while this concentration was 0.9 µM (57.5±8 µg

dm-3) in the unplanted control at day 0. At CW#3 start, free Cu2+ represented 73% and 86% of

the total Cu in CW planted with P. australis and P. arundinacea, 49% in CW planted with J.

articulatus and 43% in unplanted CW (Table 22). Thus, Cu desorption from roots and root

exudates may be higher than Cu desorption from the microbial biofilm. Increase in pH for all

CW during the CW#3 may result from H+ competition with Cu previously sorbed on the

various CW components, i.e. perlite and Fe oxides of diorite, DOM, roots, microbial biofilms,

and CW walls. In parallel, roots may excrete OH− and HCO3− to buffer water pH and to take

up anions such as nitrates and sulfates (Dakora and Phillips, 2002). However, water pH was

similar in all CW (Table 22) and thus roots may less influence it compared to other CW

components.

After 6 days, once excess H+ was sorbed on the perlite, Fe oxides, DOM, microbial biofilms,

CW walls and roots in planted CW, the overall mean surface charge of ferric (oxyhydr)oxides

changed from a positive to a negative value as pH started to increase (Sheoran and Sheoran,

2006). Thus Cu2+ may adsorb again on diorite Fe oxides and silicates provided by the perlite.

Similarly, Ferris et al. (1989) reported that biofilm metal uptake at neutral pH was enhanced

by up to 12 magnitude orders over acid conditions and that adsorption strength values were

usually higher at elevated pH. At day 14 in CW#3, total Cu concentration in water decreased

187

to 0.5 (31.4±7), 0.3 (18.5±1.8), 0.2 (13.2±7), and 0.2 (13.5±4) µM (µg dm-3) respectively in

CW planted with J. articulatus, P. australis, P. arundinacea and the unplanted CW (Figure

36, Table 22). Free Cu2+ concentration in water decreased to 0.025 (1.6±0.5), 0.012

(0.80±0.5), 0.010 (0.65±0.2), and 0.015 (0.98±0.5) µM (µg dm-3) respectively in CW planted

with J. articulatus, P. arundinacea, P. australis and unplanted CW (Figure 37, Table 22). In

all CW, the removal rate of Cu2+ reached 99%, and consequently the RTEI was 0 for the three

macrophytes. Such Cu removal rates for both Cu2+ and total Cu are in line with previous

findings (Table 23). For CW planted with P. australis, the removal rate of Cu ranged from -

11% (Tromp et al. 2012), through 43% in winter and 56 % in summer (Samecka-Symerman

et al., 2004), to 79 % (Mantovi et al., 2003). The higher total Cu and Cu2+ removal rates of

CW#3 compared to CW#1 and CW#2 may be due to higher plant and microbial activities

(Faulwetter et al., 2009; Chazarenc et al., 2010).

Conclusion

Pilot constructed wetlands (CW) using bio-racks for cleaning synthetic Cu-contaminated

wastewaters (2.5 µM Cu) were carried out in alkaline (CW#1 and CW#2 experiments) and

acid (CW#3) conditions for assessing the influences of three macrophytes on the removal rate

of Cu. All macrophytes well developed in the Bio-Rack system and provided a high root

surface for the growth of microbial populations in oxidative conditions. Various

compartments, i.e diorite Fe oxides, perlite, microbial biofilms, PVC walls, roots and

rhizomes are likely key-players for water Cu-decontamination in such CW. However, the

influence of the three macrophytes on the Cu removal rate was low, and even negative for J.

articulatus, compared to the unplanted CW. Macrophytes generally provide habitats and root

exudates to bacterial communities. Their influence on Cu behavior in water is higher at mid-

and long-term. Supply of organic matter (OM) by plants in CW is often overlooked in short

term studies and initial OM just starts to either react or be used in (bio)chemical processes

underlying metal removal from water. Further investigations should determine at what time

macrophytes have an influence as a major OM source in CW. In CW#3, pH switched from 6

to 8 after 6 days. All CW compartments likely react with H+ ions. The French Water Agency

(FWA) has defined an upper critical threshold value (<10 µg Cu dm-3) for a good freshwater

quality (SEQ EAU, 2003). Recirculation of Cu-contaminated water (2.5 µM, 158.5 µg Cu dm-

3) in our bio-rack based-CW allowed reaching 0.8 µM (48 µg Cu dm-3) in early spring and 0.2

µM (13 µg Cu dm-3) as the growing season peaked. The FWA standards may be reached by

188

extending the residence time during the growing season, but in winter low microbial and root

activities are of concern. The selection of rhizosphere bacteria tolerant to cold conditions is an

option (Gilbert et al., 2012). Further investigations are needed to characterize sorption and

biological reactions as well as microbial biofilms and their interactions with Cu in our CW.

Table 23 : Removal rates of Cu (%) in constructed wetlands.

Authors Plant species Cu inlet (µg dm-3) Removal rate of Cu (%)

Crites et al. 1997 Schoenoplectus acutus 7 55

Mantovi et al. 2003 Phragmites australis 81 79

Kamal et al. 2004 Menta aquatica 5556 30.9

Ludwigia palustris

44.9

Myriophyllum aquaticum 42.5

Samecka-Cymerman et al. 2004 Phragmites australis - W 43; S 56

Salix viminalis

W 16; S 36

Populus canadensis W 49; S 60

Nelson et al. 2006 Schoenoplectus californicus 30 80

Zhang et al. 2007 Acorus gramineus 2031 98.5

Acorus orientale

98.4

Acorus calamus

98.2

Lythrum salicaria

98.7

Iris pseudacorus 99.1

Peng et al. 2008 Potamogeton pectinatus 4960 74

Potamogeton malaianus 65

Mishra and Tripathi 2008 Eicchornia crassipes 1000-5000 86-95

Spirodela polyrrhiza

76-91

Pistia stratiotes 88-96

Megateli et al. 2009 Lemna gibba 100 77

Sekomo et al. 2012 Lemna minor 110-250 27

Tromp et al. 2012 Phragmites australis 103.5 -11

W : winter, S : summer

189

Acknowledgements

This work was financially supported by AXA foundation (PhD grant of L. Marchand), ADEME,

Department Urban Brownfield and Polluted Soils, Angers, France, the Aquitaine Region Council

(Phytorem project), and the European Commission under the Seventh Framework Programme for

Research (FP7-KBBE-266124, GREENLAND). The authors thank P. Lamy for his technical help.

References (Format : Ecological Engineering)

Adams, C.R. and Galatowitsch, S.M., 2005. Phalaris arundinacea (reed canary grass): Rapid growth and growth pattern in conditions approximating newly restored wetlands. Ecoscience, 12:569-573.

Aksoy, A., Demirezen, D. and Duman, F., 2005. Bioaccumulation, detection and analyses of heavy metal pollution in Sultan Marsh and its environment. Water Air and Soil Pollution, 164:241-255.

Alkan, M. and Dogan, M., 2001. Adsorption of Copper(II) onto perlite. Journal of Colloid and Interface Science, 243:280-291.

Armstrong, W., 1978. Root aeration in the wetland condition. In: Hook, D.D., Crawford, R.M.M. (Eds.), Plant Life in Anaerobic Environments. Processes in Anaerobiosis. Ann Arbor Science, Ann Arbor, MI, pp. 269-298.

Asaeda, T., Rajapakse, L., Manatunge, J. and Sahara, N., 2006. The effect of summer harvesting of Phragmites australis on growth characteristics and rhizome resource storage. Hydrobiologia, 553:327-335.

Bonanno, G. and Lo Giudice, R., 2010. Heavy metal bioaccumulation by the organs of Phragmites australis (common reed) and their potential use as contamination indicators. Ecological Indicators, 10:639-645.

Braga, B. 2012. 6th World Water Forum. Marseille (France). March 12-17th. Available at http://www.worldwaterforum6.org

Brisson, J. and Chazarenc, F., 2009. Maximizing pollutant removal in constructed wetlands: Should we pay more attention to macrophyte species selection? Science of the Total Environment, 407:3923-3930.

Brookins, D.J., 1988. Eh-pH Diagrams for Geochemistry. Springer-Verlag Berlin Heidelberg, New York, NY, 176 pp.

Chazarenc, F., Brisson, J. and Merlin, G., 2010. Seasonal and spatial changes of microorganism communities in constructed wetlands: A community level physiological profiling analysis. International Journal of Chemical Engineering. doi:10.1155/2010/490240

Crites, R.W., Dombeck, G.D., Watson, R.C. and Williams, C.R., 1997. Removal of metals and ammonia in constructed wetlands. Water Environment Research, 69:132-135.

Dakora, F.D. and Phillips, D.A., 2002. Root exudates as mediators of mineral acquisition in low-nutrient environments. Plant and Soil, 245:35-47.

Davison, L., Headley, T. and Pratt, K., 2005. Aspects of design, structure, performance and operation of reed beds - eight years' experience in northeastern New South Wales, Australia. Water Science and Technology, 51:129-138.

Demirezen, D., Aksoy, A. and Uruc, K., 2007. Effect of population density on growth, biomass and nickel accumulation capacity of Lemna gibba (Lemnaceae). Chemosphere, 66:553-557.

190

Ervin, G.N. and Wetzel, R.G., 2003. An ecological perspective of allelochemical interference in land-water interface communities. Plant and Soil, 256:13-28.

Faulwetter, J.L., Gagnon, V., Sundberg, C., Chazarenc, F., Burr, M.D., Brisson, J., Camper, A.K. and Stein, O.R., 2009. Microbial processes influencing performance of treatment wetlands: A review. Ecological Engineering, 35:987-1004.

Ferris, F.G., Schultze, S., Witten, T.C., Fyfe, W.S. and Beveridge, T.J., 1989. Metal interactions with microbial biofilms in acidic and neutral pH environments. Applied and Environmental Microbiology, 55:1249-1257.

Gilbert, N., Fulthorpe, R. and Kirkwood, A.E., 2012. Microbial diversity, tolerance and biodegradation potential of urban wetlands with different input regimes. Canadian Journal of Microbiology, 58:887-892.

Hoagland, D.R. and Arnon, D.L., 1950. The water-culture method for growing plants without soil. California Agricultural Experiment Station. Circular 347.

Hsu, C.B., Hsieh, H.L., Yang, L., Wu, S.H., Chang, J.-S., Hsiao, S.C., Su, H.C., Yeh, C.H., Ho, Y.S. and Lin,

H.J., 2011. Biodiversity of constructed wetlands for wastewater treatment. Ecological Engineering, 37:1533-1545.

Huang, Y.B., Wang, W.H. and Peng, A., 2000. Accumulation of Cu(II) and Pb(II) by biofilms grown on particulate in aquatic systems. Journal of Environmental Science and Health Part a-Toxic/Hazardous Substances and Environmental Engineering, 35:575-592.

Jacob, D.L. and Otte, M.L., 2004. Long-term effects of submergence and wetland vegetation on metals in a 90-year old abandoned Pb-Zn mine tailings pond. Environmental Pollution, 130:337-345.

Kissoon, L.T.T., Jacob, D.L. and Otte, M.L., 2010. Multi-element accumulation near Rumex crispus roots under wetland and dryland conditions. Environmental Pollution, 158:1834-1841.

Kamal, M., Ghaly, A.E., Mahmoud, N. and Cote, R., 2004. Phytoaccumulation of heavy metals by aquatic plants. Environment International, 29:1029-1039.

Kuschk, P., Wiessner, A., Seeger, E.M., Kaestner, M., Kappelmeyer, U., Paredes, D. and Shtemenko, N.I., 2012. The Status of Research on Constructed Wetlands. In: Vitale, K. (Hrsg.) Environmental and food safety and security for South-East Europe and Ukraine NATO Science for Peace and Security Series C: Environmental Security, Springer, Berlin, S. 155 – 171 http://dx.doi.org/10.1007/978-94-007-2953-7_15

Labadie, P. and Budzinski, H., 2005. Determination of steroidal hormone profiles along the Jalle d'Eysines River (near Bordeaux, France). Environmental Science and Technology, 39:5113-5120.

Lee, B.-H. and Scholz, M., 2007. What is the role of Phragmites australis in experimental constructed wetland filters treating urban runoff? Ecological Engineering, 29:87-95.

Lesage, E., Rousseau, D.P.L., Meers, E., Tack, F.M.G. and De Pauw, N., 2007. Accumulation of metals in a horizontal subsurface flow constructed wetland treating domestic wastewater in Flanders, Belgium. Science of the Total Environment, 380:102-115.

Lizama, K.A., Fletcher, T.D. and Sun, G., 2011. Removal processes for arsenic in constructed wetlands. Chemosphere, 84:1032-1043.

Manceau, A., Nagy, K.L., Marcus, M.A., Lanson, M., Geoffroy, N., Jacquet, T., Kirpitchikova, T. 2008. Formation of metallic Copper nanoparticles at the soil-root interface. Environmental Science and Technology, 42: 1766-1772.

Mantovi, P., Marmiroli, M., Maestri, E., Tagliavini, S., Piccinini, S. and Marmiroli, N., 2003. Application of a horizontal subsurface flow constructed wetland on treatment of dairy parlor wastewater, Bioresource Technology, 88:85-94.

191

Marchand, L., Mench, M., Jacob, D.L. and Otte, M.L., 2010. Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants and standardized measurements: A review. Environmental Pollution, 158:3447-3461.

McCabe, O.M., Baldwin, J.L. and Otte, M.L., 2001. Metal tolerance in wetland plants? Minerva Biotecnologica, 13:141-149.

Mishra, V.K. and Tripathi, B.D., 2009. Accumulation of chromium and zinc from aqueous solutions using water hyacinth (Eichhornia crassipes). Journal of Hazardous Materials, 164:1059-1063.

Megateli, S., Semsari, S. and Couderchet, M., 2009. Toxicity and removal of heavy metals (cadmium, copper, and zinc) by Lemna gibba. Ecotoxicology and Environmental Safety, 72:1774-1780.

Mucha, A.P., Almeida, C.M.R., Bordalo, A.A. and Vasconcelos, M.T.S.D., 2010. LMWOA (low molecular weight organic acid) exudation by salt marsh plants: Natural variation and response to Cu contamination. Estuarine Coastal and Shelf Science, 88:63-70.

Nelson, E.A., Specht, W.L. and Knox, A.S., 2006. Metal removal from water discharges by a constructed treatment wetland. Engineering in Life Sciences, 6:26-30.

Otte, M.L., Matthews, D.J., Jacob, D.L., Moran, B.M. and Baker, A.J.M., 2004. Biogeochemistry of metals in the rhizosphere of wetland plants – an explanation for “innate” metal tolerance? In: Wong, M.H. (Ed.), Developments in Ecosystems, vol. 1. Elsevier, pp. 87-94 (Chapter 6).

Paradies, H.H., Thies, M. and Hinze, U., 1996. In Situ X-Ray absorption study of copper-surface complexes: microbial influenced oxidation of metallic copper. The Rigaku Journal. 13:16-22.

Peng, K. J., Luo, C.L., Lou, L.Q., Li, X.D. and Shen, Z.G., 2008. Bioaccumulation of heavy metals by the aquatic plants Potamogeton pectinatus L. and Potarnogeton malaianus Miq. and their potential use for contamination indicators and in wastewater treatment. Science of the Total Environment, 392:22-29.

Raunkiær, C., 1934. The life forms of plants and statistical plant geography. Oxford University Press. Oxford.

Ryan, P.R., Delhaize, E. and Jones, D.L., 2001. Function and mechanism of organic anion exudation from plant roots. Annual Review of Plant Physiology and Plant Molecular Biology, 52:527-560.

Samecka-Cymerman, A., Stepien, D. and Kempers, A.J., 2004. Efficiency in removing pollutants by constructed wetland purification systems in Poland. Journal of Toxicology and Environmental Health-Part A-Current Issues, 67:265-275.

Sari, A., Tuzen, M., Citak, D. and Soylak, M., 2007. Adsorption characteristics of Cu(II) and Pb (II) onto expanded perlite from a aqueous solution. Journal of Hazardous Materials, 148:387-394.

Schwartzenbach, R.P., Egli, T., Hofstetter, T.B., Von Gunten, U. and Wehrli, B., 2010. Global water pollution and human health. Annual Review of Environmental Resources, 35:109-136.

SEQ EAU. 2003. Agence de l’eau Rhône Méditerranée Corse. In French. Available at http://sierm.eaurmc.fr/eaux-superficielles/fichiers-telechargeables/grilles-seq-eau-v2.pdf

Sekomo, C.B., Rousseau, D.P.L., Saleh, S.A. and Lens, P.L.N., 2012. Heavy metal removal in duckweed and algae ponds as polishing step for textitle wastewater treatment. Ecological Engineering, 44:102-110.

Sheoran, A.S. and Sheoran, V., 2006. Heavy metal removal mechanism of acid mine drainage in wetlands: A critical review. Minerals Engineering, 19:105-116.

Tanner, C.C. and Headley, T.R., 2011. Components of floating emergent macrophyte treatment wetlands influencing removal of stormwater pollutants. Ecological Engineering, 37:474-486.

Toner, B., Manceau, A., Marcus, M.A., Millet, D.B. and Sposito, G., 2005. Zinc sorption by a bacterial biofilm. Environmental Science and Technology, 39:8288-8294.

Trivedy, R.K., 2001. Use of aquatic plants in wastewater treatment. In: Trivedy, R.K., Kaul, S. (Eds.), Low Cost Wastewater Treatment Technologies. ABD, Jaipur, India, pp. 164–183.

192

Tsednee, M., Mak, Y.W., Chen, Y.R. and Yeh, K.C., 2012. A sensitive LC-ESI-Q-TOF-MS method reveals novel phytosiderophores and phytosiderophore-iron complexes in barley. New Phytologist, 195:951-961.

Tromp, K., Lima, A.T., Barendregt, A. and Verhoeven, J.T.A., 2011. Retention of heavy-metals and poly-aromatic hydrocarbons from road water in a constructed wetland and the effect of de-icing. Journal of Hazardous Materials, 203-204:290-298.

UN Educ. Sci. Cult. Organ. (UNESCO), 2009. The United Nations World Water Development Report 3: Water in a Changing World. Paris/New York: UNESCO/Berghahn Books

Valipour, A., Raman, V.K. and Ghole, V.S., 2009. A new approach in wetland systems for domestic wastewater treatment using Phragmites sp. Ecological Engineering, 35:1797-1803.

Zhang, N., Wang, Z. and Li, F., 2009. Hydrophytes may play an important role in sewage disinfection in constructed wetlands. Journal of Water and Environment, 7:75-81.

Zhang, X.P., Liu, P., Yang, Y.S. and Chen, W.R., 2007. Phytoremediation of urban wastewater by model wetlands with ornamental hydrophytes. Journal of Environmental Sciences, 19:902-909.

Supplementary material

Figure 38: Eh-pH diagram for part of the system Cu-C-S-O-H, modified from Brookins (1988).

Chapitre IV Take home message

Ce chapitre concerne l’objectif majeur de cette thèse: la décontamination d’eaux contaminées

par Cu en zone humide construite (CW). En pleine saison de végétation, le CW développé

élimine 99% du Cu de l’eau en conditions initiales acides (pH=6). Le compartiment

biologique le plus impliqué dans l’élimination du Cu serait les biofilms microbiens. Son

193

action a été quantifiée de manière indirecte ici et mérite des études complémentaires sur la

fraction de Cu piégée par le biofilm pour valider nos données. Quelles bactéries et autres

microorganismes sont impliqués dans les réactions aboutissant à l’immobilisation du Cu?

Cette immobilisation est elle physico-chimique – le Cu est adsorbé sur les microorganismes

ou les composés du biofilm – ou biologique – le Cu est stocké à l’intérieur des

microorganismes. Dans notre pilote avec recirculation de l’eau sur 2 semaines, l’influence des

macrophytes est minime, et même négative lors de l’utilisation de Juncus articulatus. Une

hypothèse serait que les racines de J. articulatus émettent des composés allélopathiques dans

le milieu qui limitent le développement et l’action du biofilm microbien. Des données

complémentaires sont nécessaires pour valider cette hypothèse. La présence de Phragmites

australis et Phalaris arundinacea n’a pas optimisé l’efficience du CW dans nos 3

expériences. Cependant, comme la grande majorité des travaux menés sur l’assainissement

d’eaux contaminées par des éléments traces en CW, nos essais ont été réalisés en

mésocosmes, avec un pas de temps court (2 semaines). La matière organique initiale du

milieu, apportée notamment par l’eau de la rivière la Jalle d’Eysines, n’aurait pas le temps

d’être totalement utilisée par les microorganismes, et ce facteur ne limiterait pas leur

développement. Ce point est souvent négligé dans les études effectuées en mésocosmes. Les

racines des macrophytes et les parties aériennes sénescentes apportent continuellement de la

matière organique au CW, et participent ainsi à son entretien. On peut supposer que sur une

période plus longue, si le développement du biofilm diminue, le Cu dans l’eau serait moins

éliminé dans un CW non planté que dans un CW planté. Selon nos données pour les CW

plantés de J. articulatus, le choix de l’espèce est important. Nos résultats montrent l’efficacité

d’une zone humide construite sur la base d’un Bio-Rack (Valipour et al., 2009) pour éliminer

Cu pendant la période de fort développement de la végétation. Cette efficacité est moindre en

hiver, probablement à cause d’une activité du biofilm plus faible (Gilbert et al., 2012). Il

faudrait à l’avenir optimiser les processus biologiques impliqués dans l’élimination du Cu (et

des éléments traces en général) en zones humides construites et conditions hivernales via

notamment la sélection de microorganismes de la rhizosphère des macrophytes adaptés au

froid.

194

Synthèse générale

195

Biotest

Effect of treatment

PTTE contentMacronutrient

content

Water quality

Relative Treatment

Efficiency Index

(RTEI)

Macrophytes

Ferric-plaque

Root exudation

Mechanisms

Biomineralisation

Adsorption

Absorption

Translocation

Anti-oxidant system activity

Volatilization

Substrate

Δ pH

Δ Redox-potential

Δ(hydr)oxides content

Δ OM content

Mechanisms

Adsorption

(de)Nitrification

Sulfide production

Oxido/reduction

Microorganisms

Quantity

Diversity

Activity

Mechanisms

Adsorption

Absorption

Translocation

Anti-oxidant system

Rhizosphere

Chemical properties Microorganisms

Quantity

Diversity

Activity

Mechanisms

Adsorption

Absorption

Translocation

Anti-oxidant system

ΔpH

Δ Redox-potential

Δ(hydr)oxides content

Δ OM content

Δ Sulphide

Δ(Hydrogeno)carbonates

Δ DOM

Δ Macronutrients

Δ PTTE content

Δ Cu

Mechanisms

(co)Precipitation

Complexation

Adsorption

Oxido/reduction

Water

(Co

nst

ruct

ed

) W

etl

an

d

Intra-specific

Phenotypic

plasticity

(populations)

Inter-specific

phenotypic

variability

Biofilm

Driver 1 Driver 2

Effect of

amendments

Cu

Figure 39: Facteurs et variables impliqués dans la décontamination de masses d’eaux contaminées au Cu en zones humides (construites) - Factors and variables involved in the sanitation of Cu-contaminated

water bodies in (Bio-Rack) constructed wetland.

Contribution des travaux de thèse au sujet

Mise en lumière du sujet par les travaux de thèse

Biomonitoring (multivariate statistics)

196

Peut-on traiter une eau contaminée au cuivre en zone humide construite ? A cette

question initiale de la thèse, la réponse est OUI. Connait-on et maitrise-t-on totalement

l’ensemble des mécanismes physico-chimiques et biologiques régissant la séquestration de Cu

ou d’autres éléments traces dans ce filtre biologique qu’est une zone humide construite? Sur

la base des travaux menés et de la littérature, la réponse est OUI, partiellement… mais pas

totalement.

I – L’analyse statistique multivariée : une approche nécessaire en phytoremédiation

pour l’interprétation des points finaux de mesure (dont le ionome).

� Cas des amendements du substrat utilisés dans le chapitre Phytotoxicity testing of lysimeter

leachates from aided phytostabilized Cu-contaminated soils using duckweed (Lemna minor

L.).

Une question récurrente dans nos travaux concerne l’interprétation des points finaux de

mesure dans les compartiments de l’écosystème. En écotoxicologie, le suivi par des

techniques physiques et physico-chimiques de la distribution et de la spéciation chimique du

ou des contaminants dans la matrice à évaluer – avant et après la mise en œuvre d’une

solution de (phyto)remédiation – est insuffisant à lui seul pour évaluer l’efficacité de l’option

conventionnelle ou phytotechnologique choisie. Il doit s’accompagner d’une évaluation

biologique des impacts sur les composantes biotiques et abiotiques des compartiments de

l’écosystème (e.g. eau, sol, plantes, microorganismes, etc.) ainsi que sur les services

écosystémiques (e.g. filtration de l’eau, cycle de la matière organique, biomasse produite et C

séquestré, habitat et communautés animales, etc.). Dans le chapitre I Phytotoxicity testing of

lysimeter leachates from aided phytostabilized Cu-contaminated soils using duckweed

(Lemna minor L.) il était nécessaire de connaître (1) la composition en Cu et autres éléments

des compartiment sol et eau (2) la phytotoxicité des eaux d’infiltration traitées en CW installé

sur un site industriel de traitement du bois et (3) l’impact des amendements potentiellement

utilisables pour des sols contaminés en Cu et des substrats de CW devant accumuler le Cu

issu de la masse d’eau à assainir. Ce dispositif a permis de renseigner sur les phases porteuses

les plus efficaces pour assainir ces eaux d’infiltration contaminées en Cu. L’utilisation

conjointe d’amendements et de plantes tolérantes au Cu, i.e. Agrostis gigantea L. et Populus

trichocarpa x deltoides cv. Beaupré, limite le transfert de Cu depuis la zone racinaire vers les

197

lixiviats. Le meilleur assainissement vis-à-vis de la concentration en Cu des lixiviats a été

obtenu avec les amendements LDS (Linz Donavitz Slag) et OMZ (compost et grenailles

d’acier se corrodant en oxydes de Fe/Mn) (Figure 20). En revanche, l’utilisation seule des

plantes Cu-tolérante n’est pas efficace (Figure 20). Une analyse statistique multivariée

(i.e. une Analyse en Composantes Principales, ACP) a discriminé les amendements LDS et

OMZ sur la base des concentrations en Mg des lixiviats. Ceux de la modalité OMZ ont des

concentrations faibles en Mg, cinq fois inférieures à la concentration moyenne dans les

rivières françaises. Cette faible concentration en Mg pénalise la croissance de L. minor

(Figure 21) tout autant que les concentrations trop élevées en Cu. La qualité de la masse

d’eau issue d’un traitement en CW ne doit donc pas être évaluée par le suivi du seul

contaminant, mais aussi de l’ensemble des synergies et antagonismes entre les paramètres

biotiques et abiotiques de la masse d’eau. L’analyse multivariée permet ce suivi. L’utilisation

systématique de ces statistiques descriptives multivariées est utile pour juger de la qualité de

la masse d’eau ou du sol après un traitement de (phyto)remédiation.

� Cas du biomonitoring des contaminations en PTTE le long d’une rivière urbaine décrit dans le chapitre

II : Macrophytes as biomonitors of trace element exposure along an urban river using a multimetric

approach (Jalle d’Eysines River, France)

La biosurveillance d’une exposition au Cu à l’aide des parties aériennes de macrophytes est

un sujet qui s’intègre à la thématique plus large de la recherche d’indice de la qualité des

rivières et des berges (Bonanno and Lo Giudice, 2010). Phragmites australis – une espèce

végétale largement répartie sur le globe – est un des macrophytes le plus utilisé pour ce type

de biosurveillance. Cependant, l’exposition à un contaminant tel que Cu est souvent suivie par

une régression linéaire entre la concentration de Cu dans le sol (ou mieux dans la solution du

sol) et celle des feuilles. Dans le chapitre II nos données soutiennent que cette option n’est pas

valable chez les macrophytes, dont les géophytes à biomasse rhizomateuse, car ils stockent en

priorité les PTTE dans leurs racines et rhizomes (Figure 13). Ceci est un biais pour une

relation de type régression linéaire souvent proposée comme modèle de biosurveillance. De

plus, l’exposition à un élément comme Cu ne modifie pas seulement les concentrations en Cu

dans les organes de la plante, mais aussi le ionome des parties aériennes par l’intermédiaire

des synergies et antagonismes entre éléments au sein de la plante, des modifications du

fonctionnement des racines, et de la dilution dans la biomasse. L’empreinte d’une exposition

au Cu ne peut donc pas se résumer à la seule concentration en Cu dans les tissus de la plante,

198

mais doit considérer l’ensemble des concentrations en oligo- et macroéléments. Une analyse

multivariée est plus adaptée pour quantifier l’impact de cette exposition. Une Analyse

Discriminante Linéaire (LDA) basée sur les macrophytes du type Hémicryptophytes – ne

produisant pas de biomasse rhizomateuse – a permis d’intégrer l’exposition à un ensemble de

PTTE le long d’une rivière urbaine, la Jalle d’Eysines, dont les berges sont légèrement

contaminée en Cu et Zn. Ce modèle est à valider dans le temps pour vérifier s’il permet de

prédire l’évolution de l’exposition. Il doit être également construit et testé sur un large panel

de rivières urbaines afin de vérifier son bien-fondé.

II – Vers une standardisation nécessaire à la sélection des macrophytes et des substrats

en CW : Nécessité d’un Biotest.

Préciser qu’une zone humide construite (CW) est une synergie entre trois acteurs du

traitement d’une eau contaminée, i.e. le substrat, les macrophytes et les microorganismes, est

intuitive. La littérature regorge d’études sur le rôle du substrat dans un CW; celui des biofilms

microbiens dans le piégeage des PTTE dont Cu a été largement décrit et le débat sur

l’importance des macrophytes en CW est régulièrement alimenté dans les publications (Lee

and Scholz, 2007; Chazarenc and Brisson, 2009; Marchand et al., 2010). Cependant, les

études conduites en CW, notamment celles sur les macrophytes, continuent en majorité à

découpler le rôle de ces trois acteurs fondamentaux. Ce point a été soulevé par l’état-de-l’art

Metal and metalloid removal in constructed wetlands, with emphasis on the importance of

plants and standardized measurements: a review (Marchand et al., 2010). 72% des études

évaluant la contribution des macrophytes à l’assainissement d’eaux contaminées en PTTE

sont conduites en absence d’un contrôle non planté. Dès lors, comment valider si les résultats

en terme d’élimination du contaminant dépendent de l’espèce végétale considérée (en

partenariat avec le consortium bactérien et fongique de la rhizosphère, les biofilms

microbiens, et les endophytes) ou bien a-t-on mesuré seulement l’action du substrat (et des

microorganismes associés)? Ceci est crucial pour les travaux de recherche à mener sur le rôle

des macrophytes en CW. Nous avons proposé un indice, le Relative Treatment Efficiency

Index (RTEI, %) permettant de découpler l’effet des couples substrat/microorganismes et

macrophytes/microorganismes dans un CW assorti d’un contrôle non planté. Nous suggérons

199

de mesurer ce RTEI lors des études sur la contribution d’une espèce de macrophyte à

l’élimination d’un contaminant en CW.

Standardiser les procédures d’évaluation du rôle des compartiments acteurs dans un CW est

nécessaire. L’action d’une espèce de macrophyte dépend du substrat du CW. Les macrophytes

participent aux processus d’assainissement de la masse d’eau via leurs exsudats racinaires

(Mucha et al., 2010), la plaque d’oxydes de Fe/Mn déposée sur les racines (Otte et al., 2004),

l’oxygénation du milieu (Armstrong et al., 1990), l’absorption des contaminants (Bonanno

and Lo Giudice, 2010; Lizama et al., 2011) et l’apport régulier de matières organiques (e.g.

parties aériennes sénescentes, mat racinaire et éventuellement rhizomes). Mais ils fournissent

d’autres services écosystémiques dont un habitat aux microorganismes contribuant à la

décontamination de la masse d’eau (Hallberg et al., 2005; Valipour et al., 2009). Cet habitat

dépend des caractéristiques du macrophyte, mais aussi de celles du substrat. Le couple

macrophyte/substrat définit la nature des communautés bactériennes et fongiques de la

rhizosphère, des biofilms microbiens, et des endophytes, et par conséquent une partie de

l’efficience de la décontamination. On propose d’évaluer le fonctionnement des macrophytes

en conditions standards avec une gamme de substrats types (ex: perlite, sable, graviers) et de

concentrations en contaminant à tester (les gammes de substrats et de concentrations seraient

à définir lors de workshops). Ces biotests standardisés permettraient de sélectionner des

macrophytes selon les conditions recherchées en CW. Ils aideraient à comparer l’action des

macrophytes, le rôle des populations et des assemblages notamment sur l’élimination de Cu

(et d’autres PTTE) de la masse d’eau, en conditions similaires. La taille du CW est aussi à

uniformiser pour ces études comparatives. Selon Marchand et al. (2010), 47% des études

répertoriées sont à l’échelle de la parcelle, 20% en mésocosmes et 33% en microcosmes. Or,

le développement et la distribution des racines varient selon la taille du milieu d’étude. Ceci

est connu sous le terme d’Hedge effect (effet de bord). Pour limiter les biais de cet artefact, la

taille du milieu dans le protocole de sélection des macrophytes doit être standardisée. A

l’instar de Brisson et Chazarenc (2009), nous proposons un biotest réalisé en mésocosme (sa

taille serait à définir lors d’un workshop), le test à l’échelle de la parcelle étant couteux en

ressources et celui en microcosme augmentant les artefacts du Hedge effect.

200

III – Quel est le rôle exact des microorganismes en CW?

Nos travaux cherchent à découpler le rôle de la paire macrophytes/microorganismes associés

de celui de la paire substrat/microorganismes. Les résultats du chapitre IV (Copper removal

from water using a bio-rack system planted with Phragmites australis, Juncus articulatus and

Phalaris arundinacea) confirment de manière indirecte que les microorganismes du CW,

localisés dans la rhizosphère, le substrat, l’eau libre et les macrophytes (endophytes)

interviennent dans l’élimination de Cu (Figures 36, 37). L’effet des microorganismes est

décrit dans la littérature, mais les données ne permettent pas de quantifier la proportion de Cu

(ou d’autres PTTE) retenue par les microorganismes en CW par rapport à celle accumulée

par le substrat, le mat des racines et les rhizomes. Si l’action des communautés microbiennes

sur la spéciation des PTTE (dont Cu) est présumée forte en CW, elle demande une forte

expertise pour la caractériser et quantifier de manière directe. La caractérisation des

communautés microbiennes présentes en CW peut être réalisée par électrophorèse sur gel des

gènes ARNr 16S (Gilbert et al., 2012) et par PCR BOX (Becerra-Castro et al., 2012). Ce rôle

peut être aussi estimé par modélisation, comme dans le Chapitre IV. La quantification directe

de la fraction de Cu retenue par les microorganismes est nécessaire afin de savoir (1) Quelles

familles microbiennes participent à l’élimination de Cu en CW? et (2) Quel est le

pourcentage de Cu sorbé sur les cellules par rapport à celui stocké à l’intérieur des

microorganismes? Ces recherches existent (Hallberg et al., 2005; Koschorreck, 2008; Gilbert

et al., 2012; Becerra-Castro et al., 2012), mais les publications sur ce sujet sont peu

nombreuses au regard du nombre sur le choix des espèces de macrophytes à utiliser en CW.

Les techniques d’imagerie, i.e. la µ-XRF, l’EXAFS, le XANES et la µ-CT pourraient fournir

des éléments de réponse à ces questions.

IV – Influence de la population chez les macrophytes utilisés en CW.

Une question centrale dans nos travaux était existe t-il une plasticité phénotypique intra-

spécifique qui répond à une exposition élevée en Cu chez les macrophytes utilisées dans les

CW? Le paradigme dominant est que les macrophytes ont une tolérance constitutive à l’excès

de PTTE, due à (1) leur capacité à sorber les PTTE sur les racines via la plaque

201

d’oxy(hydroxyde) de Fe/Mn, (2) la formation de nanoparticules métalliques (Cu) autour et

dans les racines (Manceau et al. 2008), (3) un stockage dans les vacuoles des racines et

rhizomes (Kissoon et al., 2010; Caldelas et al., 2012) et (4) la liaison du Cu dans le

rhizoderme avec divers ligands (e.g. ceux possédant un ou plusieurs groupes thiols, les

groupements carboxyles dont ceux des acides polygalacturoniques des parois cellulaires, les

membranes des endophytes bactériens potentiellement présents dans l'apoplasme, etc.)

(Kopttike et al., 2011; Cestone et al., 2012). Peu d’études ont recherché une plasticité

phénotypique des macrophytes en réponse à une exposition élevée au Cu (et aux PTTE en

général). Ces études concluent en majorité à son absence (Ye et al., 1997, 2003; Matthews et

al., 2004, 2005). Notre échantillonnage du chapitre III Does phenotypic plasticity explain

copper tolerance in macrophyte populations? a fourni des populations pour six espèces de

macrophytes provenant de sites géographiques éloignés et aux niveaux de contamination en

Cu variant du fonds pédogéochimique à fortement contaminé (Figure 29, Tableau 15). Les

hémicryptophytes et géophytes à faible production de biomasse rhizomateuse répondent par

une plasticité de la croissance racinaire à une exposition élevée en Cu (Tableau 19). Cette

plasticité s’exprime aussi dans l’activité d’une enzyme du système antioxydant, la gaïacol

peroxydase, chez P. australis (Figure 33). Nos résultats remettent en cause le paradigme

dominant sur le potentiel des macrophytes à éliminer du Cu d’une masse d’eau contaminée.

Ce potentiel pour au moins 3 macrophytes n’est plus seulement imputable à l’espèce, mais

aussi à leur population chez des hémicryptophytes et géophytes à faible production de

biomasse rhizomateuse telles que J. effusus, S. lacustris et P. arundinacea. Par conséquence,

la question suivante est: les géophytes à forte production de biomasse sont elles plus

intéressantes – car plus tolérantes – en termes d’élimination du Cu en CW que les

hémicryptophytes et les géophytes à faible production de biomasse rhizomateuse?

Cette question mérite d’être abordée à l’avenir. Une tolérance à Cu des géophytes à forte

production de biomasse rhizomateuse, telles que T. latifolia, I. pseudacorus ou dans une

moindre mesure P. australis, est démontrée dans nos travaux. Cependant les résultats du

chapitre IV Copper removal from water using a bio-rack system planted with Phragmites

australis, Juncus articulatus and Phalaris arundinacea montrent que l’efficacité de P.

arundinacea (hémicryptophyte) est équivalente à celle de P. australis (géophyte à forte

production de biomasse rhizomateuse) et plus importante que celle de J. articulatus (géophyte

a faible production de biomasse rhizomateuse) (Figures 36, 37) en terme d’élimination de Cu

202

d’une masse d’eau. Pour valider ou non certaines hypothèses de mécanismes, nous suggérons

de quantifier en CW: (1) les phytosidérophores exsudés par les monocotylédones dont P.

arundinacea et (2) les composés allopathiques, car ceux exsudés par J. articulatus pourraient

limiter le développement du biofilm microbien (Zhang et al., 2009, Gilbert et al., 2012).

La plasticité phénotypique des racines de certains macrophytes étudiés n’est pas corrélée avec

l’exposition au Cu sur le site d’origine. Des populations échantillonnées en milieu non

contaminé ont pourtant une tolérance au Cu importante sur 3 semaines (Figure30, 31).

L’analyse préliminaire des racines de ces populations laisse supposer leur sub-carence en Cu

et leur réponse comme une conséquence des mécanismes de maintien de l’homéostasie

cellulaire du Cu. Les mécanismes physiologiques expliquant la plasticité phénotypique des

racines de macrophytes lors de forte exposition au Cu devront être étudiés de même que les

voies métaboliques mises en jeu. Des hypothèses sont : une activité différentielle de la

cascade anti-oxydante, une synthèse accrue de protéines chaperonnes, une variabilité dans la

quantité de Si déposée sur les parois de l’endoderme, une biominéralisation de nanoparticules

de Cu à proximité directe des racines et la présence différentielle (nature et quantité) de

transporteurs membranaires (COPT, ZIP). Des études en protéomique (e.g. gels 2D suivis de

l’identification des protéines solubles, identification des protéines membranaires) et

transcriptomique sur des macrophytes exposées à Cu (et aux PTTE en général) feraient

émerger des éléments de réponse à ces questions. La plasticité phénotypique des racines dans

nos travaux peut aussi résulter de la diversité intra-spécifique des communautés de

symbiontes et d’endophytes. Caractériser la plasticité de ces microorganismes et

communautés est un axe de recherche à développer, la littérature sur ce thème étant à ma

connaissance faible.

V – Apport d’amendements au substrat d’un CW pour décontaminer une masse d’eau à forte concentration en Cu.

Le traitement d’une masse d’eau contaminée en Cu en CW de type Bio-Racks a éliminé

jusqu’à 99% du Cu (Tableau 22). Il est difficile de confronter ces résultats à la littérature à

cause de la diversité des protocoles et dispositifs testés (Marchand et al., 2010). La

phytoremédiation en CW manque de standards (cf supra) pour comparer l’efficacité des

203

systèmes. La taille du pilote, le temps de rétention de la masse d’eau et l’âge des macrophytes,

sauf si ce sont les facteurs testés, devraient être similaires pour les tests en mésocosme.

L’utilisation de clones de macrophytes, mis en culture en prévision du test et plantés dans les

pilotes l’année le précédant, est préconisée.

Les processus et mécanismes de phytoremédiation en CW différent légèrement de ceux pour

phytoremédier des sols contaminés. Les CW ont des spécificités dont les (micro)variations du

potentiel rédox, le développement conséquent de la plaque Fe/Mn sur les racines, la présence

de rhizomes où le contaminant peut être stocké et celle de biofilms microbiens (Sheoran and

Sheoran, 2006; Lizama et al., 2011). Cependant, les mécanismes premiers influant sur l’éco-

dynamique du Cu sont les mêmes en CW que dans le sol: sorption (e.g. aux phases porteuses

du substrat, aux matières organiques, au biofilm, aux exsudats racinaires), absorption

racinaire, sorption aux (oxydr)oxydes de Fe et Mn couvrant la surface des racines et aux

alumino-silicates, précipitation avec les carbonates en milieu alcalin (Sheoran and Sheoran,

2006). L’apport d’amendements dans un sol contaminé en PTTE en complément de l’action

des racines pour diminuer les pools labiles de PTTE – la phytostabilisation aidée – est pour le

moment relativement peu étendu aux CW (Mench et al., 2010). Les lysimètres du chapitre I

peuvent être considérés comme des CW de type VF (vertical flux). Dans nos travaux, l’ajout

d’amendement a contribué de manière efficace à stabiliser Cu dans le sol et a diminué sa

concentration dans les lixiviats issus de la couche 0-65 cm. L’utilisation d’amendements en

CW pourrait augmenter leurs performances.

204

Sta

nd

ard

ize

dB

iote

st(s

)

Mu

ltiv

ari

ate

sta

tist

ics

Figure 40: Perspectives de recherche sur les zones humides construites -

Research perspectives on CW.

Rhizosphere

Chemical properties Microorganisms

Water

(Co

nst

ruct

ed

) W

etl

an

d Cu

Cu

Substrate Microorganisms

Macrophytes

Improvement of

Phytoremediation

in CW by adding

amendments in

the substrate

1. Characterization

2. Quantification

3. Activity

adsorption vs

absorption ?

1. Characterization

2. Quantification

3. Activity

adsorption vs

absorption ?

?

?

?

?

Intra-population phenotypic plasticity ?

Allelopathic compounds ?

Phenotypic plasticity ?

Rhizomatous geophytes vs Hemicryptophytes ?

Graminaceous vs non graminaceous ? (Phytosiderophores ?)

205

Références : (ces références englobent celles de la partie I de l’introduction et celles de la synthèse générale)

Abdel-Ghany, S.E., Burkhead, J.L., Gogolin, K.A., Andres-Colas, N., Bodecker, J.R., Puig, S., Penarrubia, L., Pilon, M., 2005. AtCCS is a functional homolog of the yeast copper chaperone Ccs1/Lys7. Febs Letters 579, 2307-2312.

Alaoui-Sosse, B., Genet, P., Vinit-Dunand, F., Toussaint, M.L., Epron, D., Badot, P.M., 2004. Effect of copper on growth in cucumber plants (Cucumis sativus) and its relationships with carbohydrate accumulation and changes in ion contents. Plant Science 166, 1213-1218.

Ali, M.B., Singh, N., Shohael, A.M., Hahn, E.J., Paek, K.-Y., 2006. Phenolic metabolism and lignin synthesis in root suspension cultures of Panax ginseng in response to copper stress. Plant Science 171, 147-154.

Aloni, R., Aloni, E., Langhans, M., Ullrich, C.I., 2006. Role of cytokinin and auxin in shaping root architecture: Regulating vascular differentiation, lateral root initiation, root apical dominance and root gravitropism. Annals of Botany 97, 883-893.

Alscher, R.G., Erturk, N., Heath, L.S., 2002. Role of superoxide dismutases (SODs) in controlling oxidative stress in plants. Journal of Experimental Botany 53, 1331-1341.

Andres-Colas, N., Sancenon, V., Rodriguez-Navarro, S., Mayo, S., Thiele, D.J., Ecker, J.R., Puig, S., Penarrubia, L., 2006. The Arabidopsis heavy metal P-type ATPase HMA5 interacts with metallochaperones and functions in copper detoxification of roots. Plant Journal 45, 225-236.

Becerra-Castro, C., Monterroso, C., Prieto-Fernández, A., Rodríguez-Lamas, L., Loureiro-Viñas, M., Acea, M.J., Kidd, P.S. 2012. Pseudometallophytes colonising Pb/Zn mine tailings: a description of the plant-microorganism-rhizosphere soil system and isolation of metal tolerant bacteria. Journal of Hazardous Materials 217-218, 350-359.

Brisson, J., Chazarenc, F. 2009. Maximizing pollutant removal in constructed wetlands: Should we pay more attention to macrophyte species selection? Science of the Total Environment 407, 3923-3930.

Burkhead, J.L., Reynolds, K.A.G., Abdel-Ghany, S.E., Cohu, C.M., Pilon, M., 2009. Copper homeostasis. New Phytologist 182, 799-816.

Caldelas, C., Bort, J., Febrero, A., 2012. Ultrastructure and subcellular distribution of Cr in Iris pseudacorus L. using TEM and X-ray microanalysis. Cell Biology and Toxicology 28, 57-68.

Cestone, B., Vogel-Mikus, K., Quartacci, M.F., Rascio, N., Pongrac, P., Pelicon, P., Vavpetic, P., Grlj, N., Jeromel, L., Kump, P., Necemer, M., Regvar, M., Navari-Izzo, F. 2012. Use of micro-PIXE to determine spatial distributions of copper in Brassica carinata plants exposed to CuSO4 or CuEDDS. Science of The Total Environment 427-428, 339-346.

Chen, E.L., Chen, Y.A., Chen, L.M., Liu, Z.H., 2002. Effect of copper on peroxidase activity and lignin content in Raphanus sativus. Plant Physiology and Biochemistry 40, 439-444.

Cheng, Y.F., Randlett, M.D., Findell, J.L., Schaller, G.E., 2002. Localization of the ethylene receptor ETR1 to the endoplasmic reticulum of Arabidopsis. Journal of biological chemistery 277, 19861-19866.

Collin, V.C., Eymery, F., Genty, B., Rey, P., Havaux, M. 2008. Vitamin E is essential for the tolerance of Arabidopsis thaliana to metal-induced oxidative stress. Plant Cell and Environment 31, 244-257.

Cuypers, A., Vangronsveld, J., Clijsters, H. 2002. Peroxydase in roots and primary leaves of phaseolus vulgaris. Copper and zinc phytotoxicity : a comparison. Journal of Plant Physiology 162, 383-392.

Curie, C., Cassin, G., Couch, D., Divol, F., Higuchi, K., Le Jean, M., Misson, J., Schikora, A., Czernic, P., Mari, S. 2009.Metal movement within the plant : contribution of nicotianamine and yellow-stripe 1 like transporters. Annals of Botany 103, 1-11.

Darko, E., Ambrus, H., Stefanovits-Banyai, E., Fodor, J., Bakos, F., Barnaba, B., 2004. Aluminium toxicity, Al tolerance and oxidative stress in an Al-sensitive wheat genotype and in Al-tolerant lines developed by in vitro microspore selection. Plant Science 166, 583-591.

206

Delmail, D., Labrousse, P., Hourdin, P., Larcher, L., Moesch, C., Botineau, M., 2011. Differential responses of Myriophyllum alterniflorum DC (Haloragaceae) organs to copper: physiological and developmental approaches. Hydrobiologia 664, 95-105.

Dell’mour, M., Koellensperger, G., Quirino, J.P., Haddad, P.R., Stanetty, C., Oburger, E., Puschenreiter, M., Hann, S. 2010. Complexation of metals by phytosiderophores revealed by CE-ESI and CE-ICP-MS. Electrophoresis 31, 1201-1207.

Devi, S.R., Prasad, M.N.V., 1998. Copper toxicity in Ceratophyllum demersum L. (Coontail), a free floating macrophyte: Response of antioxidant enzymes and antioxidants. Plant Science 138, 157-165.

Dietz, K.J., 2003. Plant peroxiredoxins. Annual Review of Plant Biology 54, 93-107.

Dietz, K.J., Jacob, S., Oelze, M.L., Laxa, M., Tognetti, V., de Miranda, S.M.N., Baier, M., Finkemeier, I., 2006. The function of peroxiredoxins in plant organelle redox metabolism. Journal of Experimental Botany 57, 1697-1709.

Dinneny, J.R., Long, T.A., Wang, J.Y., Jung, J.W., Mace, D., Pointer, S., Barron, C., Brady, S.M., Schiefelbein, J., Benfey, P.N., 2008. Cell identity mediates the response of Arabidopsis roots to abiotic stress. Science 320, 942-945.

Donner, E., Punshon, T., Guerinot, M.L., Lombi, E., 2012. Functional characterisation of metal(loid) processes in planta through the integration of synchrotron techniques and plant molecular biology. Analytical and Bioanalytical Chemistry 402, 3287-3298.

Drazkiewicz, M., Skorzynska-Polit, E., Krupa, Z., 2004. Copper-induced oxidative stress and antioxidant defence in Arabidopsis thaliana. Biometals 17, 379-387.

Eide, D., Broderius, M., Fett, J., Guerinot, M.L., 1996. A novel iron-regulated metal transporter from plants identified by functional expression in yeast. Proceedings of the National Academy of Sciences of the United States of America 93, 5624-5628.

Finkemeier, I., Goodman, M., Lamkemeyer, P., Kandlbinder, A., Sweetlove, L.J., Dietz, K.J., 2005. The mitochondrial type II peroxiredoxin F is essential for redox homeostasis and root growth of Arabidopsis thaliana under stress. Journal of Biological Chemistry 280, 12168-12180.

Fushiya, S., Sato, Y., Nozoe, S., Nomoto, K., Takemoto, T., Takagi, S. 1980. Avenic acid A, a new amino acid possessing an iron-chelating activity. Tetrahedron Letters 201, 3071-3072.

Garcia-Molina, A., Andres-Colas, N., Perea-Garcia, A., del Valle-Tascon, S., Penarrubia, L., Puig, S., 2011. The intracellular Arabidopsis COPT5 transport protein is required for photosynthetic electron transport under severe copper deficiency. Plant Journal 65, 848-860.

Gilbert, N., Fulthorpe, R., Kirkwood, A.E. 2012. Microbial diversity, tolerance and biodegradation potential of urban wetlands with different input regimes. Canadian Journal of Microbiology. 58, 887-892.

Gonzalez-Mendoza, D., Moreno, A.Q., Zapata-Perez, O., 2007. Coordinated responses of phytochelatin synthase and metallothionein genes in black mangrove, Avicennia germinans, exposed to cadmium and copper. Aquatic Toxicology 83, 306-314.

Hall, J.L., 2002. Cellular mechanisms for heavy metal detoxification and tolerance. Journal of Experimental Botany 53, 1-11.

Hinsinger, P., Bengough, A.G., Vetterlein, D., Young, I.M., 2009. Rhizosphere: biophysics, biogeochemistry and ecological relevance. Plant Soil 321, 117–152.

Huang, G.-Y., Wang, Y.-S., 2010. Expression and characterization analysis of type 2 metallothionein from grey mangrove species (Avicennia marina) in response to metal stress. Aquatic Toxicology 99, 86-92.

Hung, S.H., Yu, C.W., Lin, C.H. 2005. Hydrogen peroxide functions as a stress signal in plants. Bot. Bull. Acad. Sin 46, 1-10.

207

Jaquinod, M., Villiers, F., Kieffer-Jaquinod, S., Hugouvieu, V., Bruley, C., Garin, J., Bourguignon, J., 2007. A proteomics dissection of Arabidopsis thaliana vacuoles isolated from cell culture. Molecular & Cellular Proteomics 6, 394-412.

Jiang, W., Liu, D., Liu, X., 2001. Effects of copper on root growth, cell division, and nucleolus of Zea mays. Biologia Plantarum 44, 105-109.

Jung, C., Maeder, V., Funk, F., Frey, B., Sticher, H., Frossard, E., 2003. Release of phenols from Lupinus albus L. roots exposed to Cu and their possible role in Cu detoxification. Plant and Soil 252, 301-312.

Kaul, S., Sharma, S.S., Mehta, I.K., 2008. Free radical scavenging potential of L-proline: evidence from in vitro assays. Amino Acids 34, 315-320.

Kidd, P., Barcelo, J., Pilar Bernal, M., Navari-Izzo, F., Poschenrieder, C., Shilev, S., Clemente, R., Monterroso, C., 2009. Trace element behaviour at the root-soil interface: Implications in phytoremediation. Environmental and Experimental Botany 67, 243-259.

Klaumann, S., Nickolaus, S.D., Fuerst, S.H., Starck, S., Schneider, S., Neuhaus, H.E., Trentmann, O., 2011. The tonoplast copper transporter COPT5 acts as an exporter and is required for interorgan allocation of copper in Arabidopsis thaliana. New Phytologist 192, 393-404.

Kobayashi, Y., Kuroda, K., Kimura, K., Southron-Francis, J.L., Furuzawa, A., Kimura, K., Iuchi, S., Kobayashi, M., Taylor, G.J., Koyama, H., 2008. Amino acid polymorphisms in strictly conserved domains of a P-type ATPase HMA5 are involved in the mechanism of copper tolerance variation in Arabidopsis. Plant Physiology 148, 969-980.

Kopittke, P.M., Menzies, N.W., de Jonge, M.D., McKenna, B.A., Donner, E., Webb, R.I., Paterson, D.J., Howard, D.L., Ryan, C.G., Glover, C.J., Scheckel, K.G., Lombi, E., 2011. In situ distribution and speciation of toxic copper, nickel, and zinc in hydrated roots of cowpea. Plant Physiology 156, 663-673.

Kramer, U., Talke, I.N., Hanikenne, M., 2007. Transition metal transport. Febs Letters 581, 2263-2272.

Larue, C., Korboulewsky, N., Wang, R., Mevy, J.-P., 2010. Depollution potential of three macrophytes: Exudated, wall-bound and intracellular peroxidase activities plus intracellular phenol concentrations. Bioresource Technology 101, 7951-7957.

Lavid, N., Barkay, Z., Tel-Or, E., 2001. Accumulation of heavy metals in epidermal glands of the waterlily (Nymphaeaceae). Planta 212, 313-322.

Lequeux, H., Hermans, C., Lutts, S., Verbruggen, N., 2010. Response to copper excess in Arabidopsis thaliana: Impact on the root system architecture, hormone distribution, lignin accumulation and mineral profile. Plant Physiology and Biochemistry 48, 673-682.

Lee, S., Kim, Y.Y., Lee, Y., An, G. 2007. Rice P1B-type heavy metal ATP-ase, OsHMA9, is a metal efflux protein. Plant Physiology 145, 831-842.

Liao, M.T., Hedley, M.J., Woolley, D.J., Brooks, R.R., Nichols, M.A., 2000. Copper uptake and translocation in chicory (Cichorium intybus L. cv Grasslands Puna) and tomato (Lycopersicon esculentum Mill. cv Rondy) plants grown in NFT system. II. The role of nicotianamine and histidine in xylem sap copper transport. Plant and Soil 223, 243-252.

Lin, C.C., Chen, L.M., Liu, Z.H., 2005. Rapid effect of copper on lignin biosynthesis in soybean roots. Plant Science 168, 855-861.

Maeda, H., Sage, T.L., Isaac, G., Welti, R., DellaPenna, D., 2008. Tocopherols modulate extraplastidic polyunsaturated fatty acid metabolism in Arabidopsis at low temperature. Plant Cell 20, 452-470.

Maksymiec, W., Baszynski, T., 1998. The role of Ca ions in changes induced by excess Cu2+ in bean plants. Growth parameters. Acta Physiologiae Plantarum 20, 411-417.

Manceau, A., Nagy, K.L., Marcus, M.A., Lanson, M., Geoffroy, N., Jacquet, T., Kirpitchikova, T. 2008. Formation of metallic Copper nanoparticles at the soil-root interface. Environmental Science and Technology 42, 1766-1772.

208

Marchand, L., Mench, M., Jacob, D.L., Otte, M.L., 2010. Metal and metalloid removal in constructed wetlands, with emphasis on the importance of plants and standardized measurements: A review. Environmental Pollution 158, 3447-3461.

Marschner, P. 2011. Mineral nutrition of higher plants, 3rd ed. (London: Academic press). 684 pp.

Maxfield, A.B., Heaton, D.N., Winge, D.R., 2004. Cox17 is functional when tethered to the mitochondrial inner membrane. Journal of Biological Chemistry 279, 5072-5080.

Mench, M., Morel, J.L., Guckert, A., 1987. Metal binding properties of high-molecular-weight soluble exudates from maize (Zea mays –L) roots. Biology and Fertility of Soils 3, 165-169.

Mench, M., Fargues S., 1994. Metal uptake by iron efficient and inefficient oats. Plant and Soil 165, 227-233. Mench, M., Lepp, N., Bert, V., Schwitzguebel, J.P., Gawronski, S., Schröder, P., Vangronsveld, J. 2010.

Successes and limitations of phytotechnologies at field scale: outcomes, assessment and outlook from COST Action 859. Journal of soils and sediments 10, 1039-1070.

Mendoza-Cozatl, D.G., Zhai, Z.Y., Jobe, T.O., Akmakjian, G.Z., Song, W.Y., Limbo, O., Russell, M.R., Kozlovskyy, V.I., Martinoia, E., Vatamaniuk, O.K., Russell, P., Schroeder, J.I., 2010. Tonoplast-localized Abc2 transporter mediates phytochelatin accumulation in vacuoles and confers cadmium tolerance. Journal of Biological Chemistry 285, 40416-40426.

Mittler, R., Vanderauwera, S., Gollery, M., Van Breusegem, F., 2004. Reactive oxygen gene network of plants. Trends in Plant Science 9, 490-498.

Morelli, E., Scarano, G., 2004. Copper-induced changes of non-protein thiols and antioxidant enzymes in the marine microalga Phaeodactylum tricomutum. Plant Science 167, 289-296.

Moreno, I., Norambuena, L., Maturana, D., Toro, M., Vergara, C., Orellana, A., Zurita-Silva, A., Ordenes, V.R., 2008. AtHMA1 is a thapsigargin-sensitive Ca2+/heavy metal pump. Journal of Biological Chemistry 283, 9633-9641.

Mucha, A.P., Almeida, C.M.R., Bordalo, A.A., Vasconcelos, M.T.S.D., 2010. LMWOA (low molecular weight organic acid) exudation by salt marsh plants: Natural variation and response to Cu contamination. Estuarine Coastal and Shelf Science 88, 63-70.

Murphy, A.S., Eisinger, W.R., Shaff, J.E., Kochian, L.V., Taiz, L., 1999. Early copper-induced leakage of K+ from Arabidopsis seedlings is mediated by ion channels and coupled to citrate efflux. Plant Physiology 121, 1375-1382.

Navrot, N., Collin, V., Gualberto, J., Gelhaye, E., Hirasawa, M., Rey, P., Knaff, D.B., Issakidis, E., Jaquot, J.P., Rouhier, N. Plant glutathione peroxidases are functional peroxiredoxins distributed in several subcellular compartments and regulated during biotic and abiotic stresses. Plant Physiology 142, 1364-1379.

Nibau, C., Gibbs, D.J., Coates, J.C., 2008. Branching out in new directions: the control of root architecture by lateral root formation. New Phytologist 179, 595-614.

Nishizono, H., Ichikawa, H., Suziki, S., Ishii, F., 1987. The role of the root cell-wall in the heavy-metal tolerance of Athyrium yokocense. Plant and Soil 101, 15-20.

Nozoye, T., Nagasaka, S., Kobayashi, T., Takahashi, M., Sato, Y., Uozumi, N., Nakanishi, H., Nishizawa, N.H. 2011. Phytosiderophore efflux transporters are crucial for iron acquisition in graminaceous plants. Journal of Biological Chemistry 286, 5446-5454.

Ouzounidou, G., Ciamporova, M., Moustakas, M., Karataglis, S., 1995. Responses of maize (Zea mays L) plants to copper stress. 1. Growth, mineral content and ultrastructure of roots. Environmental and Experimental Botany 35, 167-176.

Palma, J.M., Gomez, M., Yanez, J., Delrio, L.A., 1987. Increased levels of peroxisomal active oxygen-related enzymes in copper-tolerant pea plants. Plant Physiology 85, 570-574.

209

Palmer, C.M., Guerinot, M.L., 2009. Facing the challenges of Cu, Fe and Zn homeostasis in plants. Nature Chemical Biology 5, 333-340.

Palmgren, M.G., 2001. Plant plasma membrane H+-ATPases: Powerhouses for nutrient uptake. Annual Review of Plant Physiology and Plant Molecular Biology 52, 817-845.

Palmgren, M.G., Clemens, S., Williams, L.E., Kraemer, U., Borg, S., Schjorring, J.K., Sanders, D., 2008. Zinc biofortification of cereals: problems and solutions. Trends in Plant Science 13, 464-473.

Pasternak, T., Rudas, V., Potters, G., Jansen, M.A.K., 2005. Morphogenic effects of abiotic stress: reorientation of growth in Arabidopsis thaliana seedlings. Environmental and Experimental Botany 53, 299-314.

Pich, A., Scholz, G., 1996. Translocation of copper and other micronutrients in tomato plants (Lycopersicon esculentum Mill): Nicotianamine-stimulated copper transport in the xylem. Journal of Experimental Botany 47, 41-47.

Pilon, M., 2011. Moving copper in plants. New Phytologist 192, 305-307.

Puig, S., Andres-Colas, N., Garcia-Molina, A., Penarrubia, L., 2007. Copper and iron homeostasis in Arabidopsis: responses to metal deficiencies, interactions and biotechnological applications. Plant Cell and Environment 30, 271-290.

Ramocha P., Chabannes, M., Chamayou, S., Danoun, S., Jauneau, A., Boudet, A.M., Goffner, D., 2002. Laccase down regulation causes alterations in phenolic metabolism and cell wall structure in poplar. Plant Physiology 129, 145-155.

Raunkiær, C., 1934. The life forms of plants and statistical plant geography. Oxford university Press. Oxford.

Robinson, N.J., Procter, C.M., Connolly, E.L., Guerinot, M.L., 1999. A ferric-chelate reductase for iron uptake from soils. Nature 397, 694-697.

Ruzicka, K., Simaskova, M., Duclercq, J., Petrasek, J., Zazimalova, E., Simon, S., Friml, J., Van Montagu, M.C.E., Benkova, E., 2009. Cytokinin regulates root meristem activity via modulation of the polar auxin transport. Proceedings of the National Academy of Sciences of the United States of America 106, 4284-4289.

Salt, D.E., Smith, R.D., Raskin.I., 1998. Phytoremediation. Annual Review of Plant Physiology and Molecular Biology 49, 643-668.

Sancenon, V., Puig, S., Mateu-Andres, I., Dorcey, E., Thiele, D.J., Penarrubia, L., 2004. The Arabidopsis copper transporter COPT1 functions in root elongation and pollen development. Journal of Biological Chemistry 279, 15348-15355.

Sasaki, M., Yamamoto, Y., Matsumoto, H., 1996. Lignin deposition induced by aluminum in wheat (Triticum aestivum) roots. Physiologia Plantarum 96, 193-198.

Sharma, S.S., Dietz, K.J., 2009. The relationship between metal toxicity and cellular redox imbalance. Trends in Plant Science 14, 43-50.

Sharma, S.S., Dietz, K.J., 2006. The significance of amino acids and amino acid-derived molecules in plant responses and adaptation to heavy metal stress. Journal of Experimental Botany 57, 711-726.

Shikanai, T., Muller-Moule, P., Munekage, Y., Niyogi, K.K., Pilon, M., 2003. PAA1, a P-type ATPase of Arabidopsis, functions in copper transport in chloroplasts. Plant Cell 15, 1333-1346.

Sinha, S., Pandey, K.., 2003. Nickel induced toxic effects and bioaccumulation in the submerged plant, Hydrilla verticillata (LF) Royle under repeated metal exposure. Bulletin of Environmental Contamination and Toxicology 71, 1175-1183.

Siripornadulsil, S., Traina, S., Verma, D.P.S., Sayre, R.T., 2002. Molecular mechanisms of proline-mediated tolerance to toxic heavy metals in transgenic microalgae. Plant Cell 14, 2837-2847.

Srivastava, S., Mishra, S., Tripathi, R.D., Dwivedi, S., Gupta, D.K.., 2006. Copper-induced oxidative stress and responses of antioxidants and phytochelatins in Hydrilla verticillata (L.f.) Royle. Aquatic Toxicology 80, 405-415.

210

Suzuki, M., Takahashi, M., Tsukamoto, T., Watanabe, S., Matsuhashi, S., Yazaki, J., Kishimoto, N., Kikuchi, S., Nakanishi, H., Mori, S., Nishizawa, N.K.., 2006. Biosynthesis and secretion of mugineic acid family phytosiderophores in zinc-deficient barley. Plant Journal 48, 85-97.

Suzuki, M., Tsukamoto, T., Inoue, H., Watanabe, S., Matsuhashi, S., Takahashi, M., Nakanishi, H., Mori, S., Nishizawa, N.K.., 2008. Deoxymugineic acid increases Zn translocation in Zn-deficient rice plants. Plant Molecular Biology 66, 609-617.

Sappin-Didier, V., Vansuyt, G., Mench, M., Briat, J.F., 2005. Cadmium availability at different soil pH to transgenic tobacco overexpressing ferritin. Plant and Soil 270, 189-197.

Ueno, D., Rombola, A.D., Iwashita, T., Nomoto, K., Ma, J.F., 2007. Identification of two novel phytosiderophores secreted by perennial grasses. New Phytologist 174, 304-310.

Thounajam, T.C., Panda, P., Mazumdar, P., Kumar, D., Sharma, G.D., Sahoo, L., Panda, S.K.., 2012. Excess copper induced oxidative stress and response of antioxidants in rice. Plant Physiology and Biochemistry 53, 33-39.

Tsednee, M., Mak, Y.W., Chen, Y.R., Yeh, K.C. 2012. A sensitive LC-ESI-Q-TOF-MS method reveals novel phytosiderophores and phytosiderophore-iron complexes in barley. New Phytologist 195, 951-961.

van de Mortel, J.E., Schat, H., Moerland, P.D., Ver Loren van Themaat, E., van der Ent, S., Blankestijn, H., Ghandilyan, A., Tsiatsiani, S., Aarts, M.G.M., 2008. Expression differences for genes involved in lignin, glutathione and sulphate metabolism in response to cadmium in Arabidopsis thaliana and the related Zn/Cd-hyperaccumulator Thlaspi caerulescens. Plant, Cell & Environment 31, 301-324.

van Hoof, N., Koevoets, P.L.M., Hakvoort, H.W.J., Ten Bookum, W.M., Schat, H., Verkleij, J.A.C., Ernst, W.H.O., 2001. Enhanced ATP-dependent copper efflux across the root cell plasma membrane in copper-tolerant Silene vulgaris. Physiologia Plantarum 113, 225-232.

Vansuyt, G., Mench, M., Briat, J.F. 2000. Soil-dependent variability of leaf iron accumulation in transgenic tobacco overexpressing ferritin. Plant Physiology and Biochemistry 38, 499-506.

Waters, B.M., Sankaran, R.P., 2011. Moving micronutrients from the soil to the seeds: Genes and physiological processes from a biofortification perspective. Plant Science 180, 562-574.

Wintz, H., Fox, T., Wu, Y.Y., Feng, V., Chen, W.Q., Chang, H.S., Zhu, T., Vulpe, C., 2003. Expression profiles of Arabidopsis thaliana in mineral deficiencies reveal novel transporters involved in metal homeostasis. Journal of Biological Chemistry 278, 47644-47653.

Wojcik, M., Tukiendorf, A., 2003. Response of wild type of Arabidopsis thaliana to copper stress. Biologia Plantarum 46, 79-84.

Yamasaki, H., Hayashi, M., Fukazawa, M., Kobayashi, Y., Shikanai, T., 2009. SQUAMOSA Promoter Binding Protein-Like7 is a central regulator for copper homeostasis in Arabidopsis. Plant Cell 21, 347-361.

Ye, Z.H., Baker, A.J.M., Wong, M.H., Willis, A.J. 1997. Zinc, lead and cadmium tolerance, uptake and accumulation by Typha latifolia. New Phytologist 136, 469-480.

Ye, Z.H., Baker, A.J.M., Wong, M.H., Willis, A.J. 2003. Copper tolerance, uptake and accumulation by

Phragmites australis. Chemosphere 50, 795-800. Zhang, N., Wang, Z., Li, F. 2009. Hydrophytes may play an important role in sewage disinfection in constructed

wetlands. Journal of Water and Environment 7, 75-81.

Zheng. L., Yamaji, N., Yokosho, K., Ma, J.F. 2012. YSL 16 is a phloem-localised transporter of the copper-nicotianamine complex that is responsible for copper distribution in rice. The Plant Cell DOI 10.1105/tpc.112.103820

Zhou, J.M., Goldsbrough, P.B., 1994. Functional homologs of fungal metallothionein genes from Arabidopsis. Plant Cell 6, 875-884.

211

212

Annexes

• Liste des macrophytes en zones humides construites

• Publications

• Posters

• Communications orales

213

Liste des macrophytes en zones humides construites

214

Phragmites australis (Reddy et al., 1989, Maltby et al., 1995; Mungur et al.,1995, 1997; Scholes et al., 1998; Gray et al., 2000; Balizon et al., 2000; Kharatanasis, 2003; Scholz and Xu, 2002; Nielsen, 2003, 2005; Scholz, 2006; Lesage et al., 2007; Vymazal et al., 2007; Lesage et

al., 2007; Lee and Scholz, 2007; Tzerakis et al., 2008; Baldantoni et al., 2009; Zhang et al., 2009; Nyquist and Greger, 2009; Khan. 2009)

Phragmites karka (Juwarker, 1995)

214

Phalaris arundinacea (Vyzamal et al., 2007)

Typha domingensis (Maine et al., 2009)

215

Typha latifolia (Juwarker, 1995; Khan et al., 2009)

Eichhornia crassipes (Boyd, 1976; Mitchell, 1976; Mac Donald, 1976; Stowell et al., 1981; Muramoto and Oti, 1983; Salati, 1987; Moorhead and Reddy, 1988; Nor, 1990; Brix, 1993;

Dhote et al., 2009; Maine et al., 2009; Mishra et al., 2009; Khan. 2009)

216

Acorus sp (Zhang et al., 2007) . Ici Acorus calamus

Iris pseudacorus (Zhang et al., 2007)

217

Scirpus validus (Fitch and Burken,?; Weiss et al., 2006)

Saggitaria latifolia (Fitch and Burken,?)

218

Cyperus alternifolius (Qian et al., 1999; Cheng et al., 2002)

Lemna minor (Zayed et al., 1998; Miretzky et al., 2004)

219

Pistia stratiotes (Qian et al., 1999; Maine et al., 2001, 2004 ; Mitetzky et al., 2004; Khan et al., 2009)

Salvinia herzogii (Maine et al., 2001, 2004)

220

Carex rostrata (Nyquist and Greger, 2009)

Eriophorum angustifolium (Nyquist and Greger, 2009)

221

Ceratophyllum demersum (Rai et al., 1995; Khan et al., 2009)

Myriophyllum spicatum (Keskinkan et al., 2003 ; Lesage et al., 2007)

222

Azolla filiculoides (Zhao et al., 1999; Keskinkan et al., 2003)

Hydrilla verticillata (Elankumaran et al., 2003, Bunluesin et al., 2007)

223

Polygonum hydropiperoides (Qian et al., 1999)

Hippuris vulgaris (Qian et al., 1999)

224

Myriophyllum brasiliense (Qian et al., 1999)

225

Spartina alternifolia (Qian et al., 1999)

226

Mimulus guttatus (Qian et al., 1999)

Cyperus pseudovegetus (Qian et al., 1999)

227

Marsilea drumondii (Qian et al., 1999)

Juncus xyphioides (Qian et al., 1999)

228

Ludwigia palustris (Kamal et al., 2004)

Mentha aquatica (Kamal et al., 2004)

229

Myriophyllum aquaticum (Kamal et al., 2004)

Spartina pectinata (Weiss et al., 2006)

230

Glyceria grandis (Weiss et al., 2006)

Potamogeton pectinatus (Peng et al., 2008)

231

Potamogeton malaianus (Peng et al., 2008)

Lythrum salicaria (Zhang et al., 2008)

232

Schoenoplectus californicus (Nelson et al., 2006, Chague-Goff, 2002)

Eleocharis acicularis (Sakakibara et al., 2009)

233

Pontederia cordata (Maine et al.. 2009, Hadad et al.. 2009)

Hydrodictyon reticulatum (Rai et al., 1995)

234

Spirodela polyrrhiza (Rai et al., 1995)

Chara corallina (Rai et al., 1995)

235

Vallisneria spiralis (Rai et al., 1995)

Bacopa monnieri (Rai et al., 1995)

236

Alternanthera sessilis (Rai et al., 1995)

Hygrorrhiza aristata (Rai et al., 1995)

237

Spirodela intermedia (Miretzky et al., 2004)

Lemna gibba (Megatelli, 2009; Khan et al., 2009)

238

Eryngium eburneum (Hadad et al., 2006)

Panicum elephantipes (hadad et al., 2006)

239

Thalia geniculata (Hadad et al., 2006)

Polygonum punctatum (Hadad et al., 2006)

240

Carex aquatilis (Khan et al., 2009)

Scirpus cyperinius (Khan et al., 2009)

241

Juncus articulatus (Khan et al., 2009)

Polygonum glabrum (Khan et al., 2009)

242

Alisma plantago-aquatica (Khan et al., 2009)

Villarsia exalta (Cheng et al., 1999)

243

Sparganium erectum (Misyuta et al., 2009)

244

Espèces potentielles

245

Molinia Cerulea

Miscanthus sinensis

246

Najas minor

Potamogeton polygonifolius

Herbiers enracinés des eaux douces dormantes pauvres en éléments nutritifs

247

Nuphar lutea

Herbiers enracinés des eaux douces dormantes relativement riches en éléments nutritifs

Nymphea alba

248

Nymphoides peltata

Trapa natans

249

Luronium natans

Potamogeton sp (ex P.natans)

250

Herbiers flottants des eaux douces dormantes pauvres ou moyennent riches en éléments nutritifs

Hydrocharis morsus-ranae

Lemna trisulca

251

Mares tourbeuses

Utricularia minor

Utricularia vulgaris

252

Sparganium minimum

Herbiers d'eaux dormantes saumâtres

Ruppia maritima

253

Ranunculus baudotii

Espèces dites envahissantes....à tester ?

Ludwigia sp (ex : L.repens)

254

Fallopia japonica

Elodea canadensis

255

Arundo donax

256

Publications

257

258

259

Journal of Applied Ecology doi: 10.1111/j.1365-2664.2012.02151.x

Integrating climate change into calcareous grassland management

Jean-Paul Maalouf*, Yoann Le Bagousse-Pinguet, Lilian Marchand, Emilie Bachelier, Blaise Touzard and Richard Michalet

UMR BIOGECO INRA 1202, Ecologie des Communaute s, Universite Bordeaux 1, Ba t. B2 RDC Est, Avenue des faculte s, 33405 Talence, France; and UMR BIOGECO INRA 1202, INRA, 69 route d’Arcachon, FR-33612 Cestas cedex, France

Summary

1. Climate change is rarely taken into consideration in conservation management strategies aimed at protecting biodiversity from other threats. We examined the implications of this perspective in European calcareous grasslands, which are among the richest herbaceous systems of the continent and are therefore of high nature conservation interest. These systems are currently undergoing species loss because of the abandonment of agro-pastoral practices. Classic ecological theory assumes that conservation management activities (such as regular mowing) and drought events should increase diversity through decreased plant competition in abandoned mesic communities. In turn, this could reduce diversity in xeric communities although positive plant interactions (facilitation) might buffer these negative effects and maintain diversity.

2. We studied the effects of regular mowing and experimentally induced drought on diversity and biotic interactions between two transplanted species in mesic and xeric calcareous grasslands. The study sites in south-western France have not been subjected to any management for the last 30 years.

3. Drought did not affect mesic systems although mowing increased plant

diversity through decreased competition. By contrast, mowing had no significant effect in xeric systems although drought decreased diversity. Interestingly, transplants were subject to neither competition nor facilitation in the xeric systems.

4. Synthesis and applications. Regular mowing and drought events impact plant diversity of mesic and xeric calcareous grassland communities in different ways. We recommend regular mowing of mesic grasslands, even in the context of climate change. By contrast, we recommend less-frequent mowing of xeric grasslands together with specific interventions such as assisted migration for species with poor drought tolerance. Similar studies in other ecosystems on larger spatial and temporal scales should examine the dual effects of management and climate change to identify appropriate management programmes.

Annals of Botany Page 1 of 8 doi:10.1093/aob/mcs152, available online

The interplay ofintensity and

Jean-Paul Maalouf1,

Blaise1UMR BIOGECO INRA 1202,

33405 Talence, France and 2

Received:

† Background and Aimsplant interactions under interknown on how disturbanceenvironments, althoughmay induce a collapseassessing such questionsnot their importance, for understanding thestudy was to assessinteractions in dry calc † Methods A field experimentthe importance and intensitymeasured along a water † Key Results The importanceway along treatments.from competition tospecies, whereas theinteraction along the gspecific; for two species, mowing in the wetteswitched to facilitainteractions for any species † Conclusions At veryto allow either competitioncollapse of interactionsand direction of inteenvironments.

260

online at www.aob.oxfordjournals.org

of stress and mowing disturbance for importance of plant interactions in

calcareous grasslands

1,2,*, Yoann Le Bagousse-Pinguet1,2, Lilian MarBlaise Touzard1,2, and Richard Michalet1,2

Ecologie des Communautes, Universite Bordeaux 1, Avenue des facultes,

2UMR BIOGECO INRA 1202, INRA, 69 route d’ArcaCestas cedex, France

* For correspondence. E-mail jeanpaul.maalouf@gmail.com

ed: 29 February 2012 Returned for revision: 4 April 2012 Accepted: 4 May 2012

Aims There is still debate regarding the direction and strctions under inter- mediate to high levels of stress. Furthermore,

turbance may interact with physical stress in unpralthough recent theory and models have shown that this

a collapse of plant interactions and diversity. The fewtions have considered the inten- sity of biotic interactions although this latter concept has been shown to be very

the role of interactions in plant communities. The objectivassess the inter- play between stress and disturbance for

calcareous grasslands.

xperiment was set up in the Dordogne, southern France,intensity of biotic interactions undergone by four speciester stress gradient, and with and without mowing distu

importance and intensity of interactions varied in a verytments. Under undisturbed conditions, plant interactions

to neutral with increasing water stress for three of the fourth species was not subject to any significantthe gradient. Responses to disturbance were more

species, competition disappeared with ettest conditions, whereas for the two other species, competition

ation with mowing. Finally, there were no significantspecies in the disturbed and driest conditions.

ery high levels of stress, plant performances become toocompetition or facilitation and disturbance may accelections in dry conditions. The results suggest that the importanceinteractions are more likely to be positively related in

for the in dry

rchand1,2,

Ba t. B2 RDC Est,

achon, FR-33612

rength of e, little is roductive interplay w studies ctions but ery useful ve of this for plant

ance, where species were

turbance.

ery similar switched the four

significant biotic species-

competition significant

too weak ccelerate the

importance stressful

261

Posters

262

263

264

References Ladislas, S., El-Mufleh, A., Gerente, C., Chazarenc, F., Andres, Y. & Bechet, B. (2012) Potential of Aquatic Macrophytes as Bioindicators of Heavy Metal Pollution in Urban

Stormwater Runoff. Water Air and Soil Pollution, 223, 877-888.

Acknowledgements This work was supported by AXA foundation for L. Marchand (PhD grant) and by EU Erasmus MUNDUS Lot 6 for Y. Vystavna and A. Kolbas (PhD grant).

Macrophytes as biomonitors of trace element exposur e along the Jalle riverbanks (France)

Mench M.1, Marchand L.

1, Nsangawimana F

1., Vystavna Y.2,3, Le coustumer P.

3, Huneau F.3

1UMR BIOGECO INRA 1202, Ecologie des Communautés, University Bordeaux 1, Talence, France. mench@bordeaux.inra.fr 2Dept. Urban Environmental Engineering & Management, National Academy

of Municipal Economy Kharkiv, Kharkiv, Ukrain 3UMR GHYMAC, University Bordeaux 1, Talence, France

.

Introduction We focused on the uptake of potentially toxic trace elements (PTTE) by macrophytes growing along the Jalle d'Eysines river (Bordeaux, Southwest France) in the

vicinity of contaminating activities, i.e. golf using pesticides, vegetable cultures using CuSO4, a former military base and two municipal wastewater treatment plant serving more

than 100,000 inhabitants. This aims at identifying macrophytes as potential biomonitors of the river contamination by PTTE (Ladislas et al., 2012).

Materials & Methods 7 macrophyte species (Ranunculus acris, Phalaris arundinacea, Phragmites asutralis, Lythrum salicaria, Iris pseudacorus, Juncus effusus, Carex acutiformis)

were sampled on 4 sites along the Jalle d'Eysine river (n=6). Leaves were washed in distilled water, oven-dried and ground. Foliar TE concentrations of macrophytes were

determined. At each sites sampled sediment samples were collected (n =6), air-dried, sieved, and total PTTE content was analyzed. Water samples (n=4 site -1) and pore water

samples (n=4 site -1) were also collected. Data were analyzed by Linear Dscriminant Analysis (LDA) to assess and predict the sampling location of macrophytes, and

consequently the PTTE exposure in the soil and soil pore water at the sampling time .

Results & Discussion

• Total concentrations of PTTE in sediments increase along the riverbanks of the Jalle d'Eysine but no such increase was reported in pore water (tab 1)

• Foliar element concentration are species dependant, but generally hemicryptophytes (L.salicaria, C.acutiformis, R.acris and P.arundinacea) accumulate more PTTE in leaves than rhizomatous geophytes (I.pseudacorus, J.effusus, P.australis) (tab 2)

• The combined use of hemicryptophytes as biomonitors with a multimetric approach such as the LDA allows to correctly assess and predict PTTE exposure of macrophytes along the Jalle d’Eysine riverbanks (table 3)

Macrophyte Community Rhizomatous-Geophytes Hemicrypto phytes

Approx Wilks ’ lam ba 0,21 0.088 0,032

Pvalue 3.247e-16 *** 1.904e-10 *** 1.2e-12 ***

LD1 LD2 LD3 LD1 LD2 LD3 LD1 LD2 LD3

Proportion of trace 0,62 0,28 0,09 0,56 0,26 0,17 0,52 0,29 0,19

Coefficients

As -0,24 0,93 0,19 -0,49 0,94 0,48 0,08 0,8 1,44

Cd 0,09 0,15 -0,54 1,04 0,25 -0,28 0,67 1,39 -1,27

Ca -0,33 0,47 0,35 -1,43 0,98 1,3 0,16 -0,35 1,47

Cr 0,52 -1,09 -0,52 0,35 -0,95 -1,02 0,33 -0,72 -0,21

Cu 0,04 -0,19 0,29 0,53 -0,02 -0,73 0,07 -0,51 1,09

Fe 0,34 0,75 0,52 0,74 -0,27 0,96 0,03 0,37 0,22

Mg -0,43 1,06 -0,21 1,2 -0,62 0,36 -0,93 1,02 -0,81

Mn 0,79 -0,27 -0,07 -0,06 0,63 -0,61 1,05 0,39 -0,26

Mo 0,74 0,18 -0,76 1,08 0,57 0,26 0,3 0,15 -1,11

Ni -0,14 0,14 0,01 -0,13 -0,18 0,04 -0,45 0,17 -1,38

P 1,28 -0,98 0,83 1,02 -1,74 -1,38 1,13 -0,63 0,97

k -0,7 -0,14 0,46 -1,31 0,89 0,69 -0,59 -0,27 -1,13

Pb -0,05 0,11 0,24 0,01 -0,17 0,07 0,03 0,24 0,31

Zn -1,53 0,08 -0,18 -1,22 0,28 1,52 -2,37 -0,69 0,63

N 116 38 65 20 51 18

Correctly classif ied Correctly classif ied Correctly classif ied

Original testing dataset Original testing data set Original testing dataset

% 70,4 63 82,5 80 92,3 88,8

As Cd Zn Cr Cu Fe Pb Mn Mo Ni Mg Ca K P

mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW mg kg -1 DW g kg -1 DW g kg -1 DW g kg -1 DW g kg -1 DW

site ** * ** ** ** ** ** ** ** ** * * * *

Juncus effusus 1 0.21±0.09 ab 0.37±0.11 a 33±6.2 a 0.46±0.18 a 3.8±1.2 a 36.7±8.9 a 0.98±0.35 a 73.1±20.1 a 1.16±0.35 a 1.24±0.71 a 0.9±0.2 a 2.2±0.4 a 17.6±4 a 1.8±0.4 a

2 0.10±0.03 a 0.79±0.21 b 37.1±2.6 a 0.58±0.13 a 12.4±3.2 b 62.9±39.8 ab 1.17±0.87 a 124.7±22.1 b 0.44±0.09 b 0.96±0.23 a 0.9±0.2 a 2.9±1.1 ab 21.3±2.3 a 2.2±0.2 a

3 0.17±0.04 b 0.79±0.17 b 42.1±10.8 a 0.62±0.05 a 8.2±1.1 c 84.5±36.6 b 1.04±0.61 a 262.9±69.7 c 1.68±0.35 c 0.83±0.16 a 1.4±0.2 b 3.4±0.7 b 16.8±2.5 a 1.8±0.2 a

4

** ** * ** ** ** ** ** * ** ** ** * **

Phragmites

australis1 0.18±0.01 a 0.05±0.01 a 43.4±2.1 a 0.62±0.05 a 19.4±0.7 a 134.5±5.2 a 1.11±0.27 a 86.8±3.3 a 0.68±0.30 ab 1.39±0.99 a 1.4±0.07 a 3.9±0.2 a 22.1±0.6 a 1.9±0.06 a

2 0.19±0.04 a 0.05±0.04 a 24.3±3.7 b 0.75±0.33 a 11.1±2.4 b 101.8±23.4 a 1.30±0.62 a 93.5±.1 a 1.29±0.76 b 0.92±0.21 a 1.1±0.8 a 3.3±1.6 a 20.8±2.9 ab 2.3±0.2 ab

3 0.37±0.33 a 0.21±0.35 a 29.2±6.8 b 0.92±0.41 a 16.1±3.8 a 140.1±23.4 a 1.02±0.56 a 119.2±35.8 a 2.6±0.53 c 0.92±0.43 a 1±0.1 a 4.9±1.1 a 17.7±2.9 b 1.9±0.3 a

4 0.22±0.07 a 0.22±0.14 a 34.2±10.8 ab 0.73±0.16 a 10.8±1.2 b 143.4±24.8 a 0.92±0.27 a 124.3±65.1 a 1.66±0.43 bc 0.59±0.09 a 1.5±0.3 a 4.9±1.1 a 19.5±2.9 ab 2.5±0.5 b

** ** * ** ** ** ** ** ** ** ** * * **

Phalaris

arundinacea1 0.12±0.01 a 0.08±0.01 a 53.6±2.3 a 0.66±0.01 a 7.4±0.5 a 73±1.8 a 0.29±0.02 a 92.5±7.3 a 0.39±0.01 a 1.09±0.01 a 4.2±0.2 a 6.6±0.5 a 14.5±0.9 a 2.8±0.02 ab

2 0.15±0.08 a 0.01±0.01 b 29.2±6.2 b 0.64±0.07 a 10.2±2.1 b 97.8±7.2 b 0.77±0.2 ab 63.7±23.2 a 0.42±0.03 a 0.51±0.01 b 1.2±0.2 b 4.2±0.7 a 29.2±7.1 b 3.2±0.7 a

3 0.20±0.02 a 0.02±0.02 b 24.9±3.9 b 0.85±0.32 a 6.3±0.5 a 102.1±21.8 b 1.42±1.1 b 67.6±17.1 a 2.03±0.04 a 0.67±0.01 c 1.3±0.1 b 3.7±0.4 b 26.3±1.2 b 2.8±0.2 ab

4 0.22±0.10 a 0.07±0.06 ab 31±2.4 b 0.60±0.14 a 11.9±2.2 b 98.7±10.2 b 1.62±0.44 b 93.1±57.4 a 2.55±1.8 a 0.53±0.01 bc 1.1±0.2 b 4.7±0.7 b 17.8±4.2 a 2.2±0.2 b

* ** ** ** ** ** ** ** ** ** ** * ** **

Carex riparia 1 0.09±0.03 a 0.12±0.01 a 19.2±3 a 0.81±0.2 a 10.2±1.1 a 65.4±3.8 a 0.83±0.21 a 57.9±2.4 ab 0.45±0.1 a 0.69±0.2 a 1.4±0.05 a 3.7±0.5 ab 14.9±0.5 a 1.1±0.04 a

2 0.13±0.02 a 0.10±0.01 a 14.5±1.2 a 0.53±0.1 b 8.2±0.5 a 83.7±2.7 a 0.86±0.72 a 21.6±0.6 a 0.41±0.01 a 0.71±0.1 a 0.9±0.03 b 2.8±0.1 a 15.8±0.6 a 1.3±0.04 a

3 0.11±0.03 a 0.13±0.04 ab 16.1±2.6 a 0.78±0.2 a 8.6±2.1 a 122.1±78 a 0.83±0.35 a 40.1±34.7 ab 0.69±0.5 a 0.75±0.1 a 1±0.2 bc 3.6±0.8 ab 16.3±2.9 a 1.3±0.3 a

4 0.24±0.03 b 0.18±0.03 b 22.2±8.7 a 1.18±0.1 c 13.3±2.5 b 345.8±78.6 b 1.39±0.35 a 59.2±18.5 b 1.65±0.1 b 1.4±0.8 b 1.2±0.1 ac 4.5±0.9 b 22.3±2.8 b 2±0.2 b

** ** ** ** ** ** ** ** ** ** ** * ** *

Ranunculus

acris1 0.16±0.03 a 0.25±0.02 a 54.9±2.2 a 0.67±0.1 a 16.4±0.4 a 95.2±3.6 a 1.06±0.32 a 29.3±0.7 a 2.25±0.1 ab 0.97±0.1 a 2.7±0.06 a 12.4±0.3 a 29.2±0.6 a 2.7±0.07 a

2 0.13±0.02 a 0.09±0.05 b 44.6+±8 a 0.83±0.2 ab 15.8±4.9 a 145.4±53.6 a 1.43±1.33 a 81.7±44.4 a 3.35±1.4 a 1.31±0.6 ab 1.6±0.6 b 12±3.1 a 30.8±4.9 a 3.9±0.7 ab

3 0.22±0.04 a 0.30±0.05 a 37.5±6.7 a 0.89±0.2 ab 16.5±4.2 a 184.9±71.5 a 1.06±0.14 a 83.5±28 a 1.87±0.5 bc 2.06±0.4 c 2.2±0.3 ab 10.6±1.6 a 32.3±5.9 a 3.8±0.9 ab

4 0.37±0.15 b 0.24±0.1 a 44.6±2.9 a 1.29±0.5 b 16.9±7.4 a 398.1±247.5 b 1.34±0.66 a 51.2±19 a 1.23±0.5 c 1.69±0.5 bc 3.8±1 c 9.4±2.5 a 30.6±3.6 a 4.5±0.4 b

* ** ** * ** ** ** ** ** ** ** ** ** **

Iris

pseudacorus1 0.07±0.01 a 1.13±0.04 a 23.5±1 a 0.56±0.1 a 8.1±0.5 a 75.1±5.2 a 1.06±1.01 a 37.6±2.3 a 0.38±0.01 a 1.09±0.01 a 2±0.09 a 21.1±0.8 a 28.9±0.6 a 2.9±0.09 a

2

3 0.09±0.03 a 0.15±0.05 b 13.6±4.7 b 0.63±0.1 ab 6.1±1 b 67.1±19.5 a 1.32±1.25 a 182.6±46.9 b 2.08±0.6 b 0.52±0.1 b 2.4±0.2 b 16.4±1.1 b 16.1±4 b 1.8±0.3 b

4 0.11±0.03 a 0.29±0.16 b 15.5±2.4 b 0.67±0.1 b 8.9±1.7 a 78.2±18.1 a 0.54±0.32 a 76.7±26.1 c 1.41±0.6 c 0.57±0.1 b 2.9±0.2 c 18.3±2.7 b 20.6±5.3 b 2.3±0.4 b

** ** ** * ** ** ** ** ** ** ** ** ** **

Lythrum

salicaria1 0.08±0.01 a 0.65±0.02 a 96.4±2.8 a 0.68±0.1 a 20.4±0.5 a 126.6±2.3 a 0.51±0.08 a 304.3±19 a 2.54±0.1 a 1.29±0.2 a 7.9±0.2 a 17.3±0.3 a 25.7±0.6 a 5.1±0.1 a

2 0.09±0.03 ab 0.05±0.02 b 54.4±6.8 ab 0.91±0.3 ab 12.8±0.7 b 132±36.3 a 0.96±0.36 a 439.6±117.8 a 0.43±0.1 b 0.51±0.01 b 3.5±0.9 b 8.7±1.7 b 37.3±2.4 b 4.2±0.5 b

3 0.18±0.04 bc 0.08±0.05 b 39±7.1 b 0.86±0.1 ab 12.1±2.4 b 168.5±50.6 ab 1.43±1.36 a 337.7±81.7 a 2.87±1.4 a 0.65±0.1 b 4.6±1 bc 9.8±1.7 b 30.4±3.9 bc 3.5±0.7 b

4 0.30±0.08 c 0.05±0.02 b 37.9±4 b 0.98±0.1 b 11.1±1.6 b 214.2±74.1 b 0.67±0.09 a 374±107.7 a 1.83±0.2 a 0.71±0.2 b 5.7±0.7 c 10.2±1.5 b 25.7±4.1 ac 3.6±0.5 b

site Total TE concentration in riverbank sediments

Cu** Zn** Cr* Ni* Co** Pb* Cd** Mn** Mo*

mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1 mg kg -1

1 2.5 ±1.8 a 19.3 ± 13.4 a 10.9 ± 4.6 a 2.6 ± 1.2 a 1.7 ± 1 a 6.9 ± 4.5 a 0.11 ± 0.09 a 73.6±37.9a 0.12±0.09a2 8.6 ± 6.7 a 16.1 ± 5.9 a 17.4 ± 7.1 a 3.9 ± 2.3 a 1.6 ± 0.8 a 11.5 ± 1.8 a 0.09 ± 0.03 a 84.4±5a 0.27±0.21a3 32.6 ± 3.6 b 171.3 ± 18.1 b 79.6 ± 10.9 b 40.1 ± 2.7 b 18.9 ± 2.5 b 54.9 ± 7.3 b 0.47 ± 0.08 b 804.7±345.1b 1.6±0.26b4 39.8 ± 4.4 b 274.2±52.8c 85.3 ± 9.1 b 39.1 ± 1.6 b 16.1 ± 0.4 b 70.1 ± 5.1 c 1.6 ± 0.54 c 767.8±107.5b 0.95±0. 1c

Al* Fe** Ca** K** Mg** Na* P* Pcu2+**

g kg -1 g kg -1 g kg -1 g kg -1 g kg -1 g kg -1 g kg -1

1 9 ± 5.4 a 3.7 ± 1.8 a 1.2 ± 0.8 a 3.9 ± 1.9 a 0.4 ± 0.2 a 1±0.5a 0.05±0.04a 12.4±0.6a2 10.5 ± 5.3 a 4.6 ± 2.5 a 2.9 ± 1.3 b 3.2 ± 1.1 a 0.4 ± 0.1 a 0.6±0.1a 0.08±0.03ab 11.4±0.6a3 85 ± 5.9 b 48 ± 10 b 6.4 ± 1.5 c 21.7 ± 1.5 b 8.6 ± 0.6 b 4.3±0.4b 0.14±0.05bc 10.1±0.09b4 82 ± 5.1 b 43 ± 1.7 b 16 ± 4.5 d 22.5 ± 1.6 b 10.5 ± 1. 3 b 4.5±0.6b 0.18±0.06c 9.9±0.08b

TE Concentration in pore water

Cu** Zn** Cr Ni Cd Mn Fe Ca Mgµg l -1 µg l -1 µg l -1 µg l -1 µg l -1 µg l -1 µg l -1 mg l-1 mg l-1

1 11.1 ± 1.8 a 0.45 ± 0.32 a 1.9 ± 0.7 a 3.1 ± 0.07 a 0.17 ± 0.01 < 20 85 ± 53 58.3 ± 0.6 a 6.3 ± 0.5 a2 19.6 ± 7.4 a 0.09 ± 0.02 b 2 ± 0.6 a 5.3 ± 1.7 b < 0.1 576 ± 404 2012 ± 1689 79.6 ± 34 ac 8.3 ± 4.9 a3 12.4 ± 3.1 a 0.09 ± 0.03 b 0.8 ± 0.6 b 4.3 ± 1.3 ab < 0 .1 576.5 ± 284 < 20 32.3 ± 5.8 b 4.9 ± 1.8 a4 15.9 ± 4.2 a 0.08 ± 0.01 b 0.7 ± 0.3 b 3.5 ± 0.9 a 0.12 ± 0.01 < 20 < 20 92.8 ± 16.9 c 16.4 ± 3.6 b

P** K** As** Mo**mg l-1 mg l-1 µg l -1 µg l -1

1 0.2 ± 0.18 a 8.1 ± 1.9 a 2.4 ± 0.3 a 6.6 ± 5 a2 1.1 ± 0.3 b 72.7 ± 31.2 b 17.9 ± 17.1 ab 10.5 ± 11.1 a3 0.11 ±0.18 a 2.9 ± 1.9 c 3.3 ± 3.1 a 14.9 ± 16 a4 0.42 ± 0.05 a 10.4 ± 3.1 a 6.5 ± 0.6 b 7.4 ± 2.6 a

Distance (km) Source 10 19 20 27 28 31 confluence

Site 1 *WTP Cantinolle 2 *WTP L’Ille 3 4

Anthropogenic activities urban area urban area & urban & maizeorganic farms rearing areas crop

Figure 1 : Sampling sites along the Jalle river (WT P: Wastewater treatment plant)

Table 1 : Total element concentrations in riverbank sediments (n=6) and soil pore waters (n=4). Values are means ±±±± SD.

Table 2 : Foliar element concentrations of macrophy tes along the Jalle riverbanks (n=6). Values are m eans ±±±± SD

. Table 3 Discriminant analysis of studied macrophyte communities with standardized canonical function discriminant function coefficients, number of observations (N), percent correctly classified

of original grouped cases and of cross-validated cases

265

266

Communications orales

267

268

269

270

That’s all folksThat’s all folksThat’s all folksThat’s all folks

271