O material particulado (MP) atmosférico é associado a vários agravos e doenças, sendo recentemente classificado como carcinogênico para humanos (Grupo 1) pela Agência Internacional de Pesquisa em Câncer (IARC). Embora sua toxicidade, incluindo a genotoxicidade, esteja reconhecidamente ligada ao tamanho das partículas, a contribuição da sua composição química para esses efeitos ainda não está completamente elucidada. O tamanho e a composição das partículas são influenciados por características meteorológicas e climáticas, com destaque para a temperatura e o período de radiação solar. O teste Salmonella/microssoma é o mais utilizado para avaliação da mutagenicidade de amostras de MP, contudo um número reduzido de linhagens é usado. Assim, o objetivo desse estudo foi investigar e comparar a influência das diferentes condições atmosféricas e climáticas das cidades de Limeira (Brasil), Estocolmo (Suécia) e Quioto (Japão) nos perfis de mutagenicidade, utilizando 11 linhagens com diferentes seletividades, e nos perfis de composição química de amostras compostas de partículas totais em suspensão (PTS) coletadas durante o inverno dessas cidades. Para que os resultados pudessem ser diretamente comparados foi adotada a mesma metodologia, incluindo o procedimento de amostragem, o método de preparo de amostra, o protocolo do teste de mutagenicidade e as técnicas de análises químicas para identificação de hidrocarbonetos policíclicos aromáticos (HPA) e de seus derivados alquilados. Limeira apresentou a maior concentração de PTS (99,0 µg/m3), seguida por Quioto (28,0 µg/m3) e Estocolmo (6,2 µg/m3). Apesar da concentração de PTS em Limeira ser 16 vezes maior que em Estocolmo e 3,5 vezes maior que em Quioto, as porcentagens de material orgânico extraído (MOE) obtidas foram 9, 15 e 5%, respectivamente. Os extratos das amostras de PTS coletadas nas três cidades apresentaram atividade mutagênica para todas as linhagens, tanto na ausência quanto na presença de S9, com exceção da TA102 que não detectou a atividade mutagênica em nenhum dos extratos, e para YG7108 na presença de S9, apenas para a amostra de Limeira. Apesar das diferenças nas condições meteorológicas e climáticas dessas cidades, seus perfis de mutagenicidade foram semelhantes. A menor potência mutagênica para a YG7108 na ausência de S9 e a ausência de mutagenicidade na presença de S9 revela a menor contribuição de agentes alquilantes para a mutagenicidade da amostra de Limeira em relação as demais cidades. Considerando a seletividade das linhagens utilizadas, foi observado maior contribuição de compostos que causam danos oxidativo ao DNA e de agentes alquilantes para a mutagenicidade da amostra de Quioto. Porém, de acordo com as respostas das linhagens que possuem atividade aumentada das enzimas nitroredutase (NR) e/ou O-acetiltransferase (OAT), YG1021, YG1024 e YG1041, revelam a maior contribuição de nitroarenos e aminas aromáticas para a mutagenicidade das três cidades, com destaque para aqueles mutágenos que dependem da ativação via OAT. As potências mutagênicas expressadas em função da massa de MOE (rev./µg MOE) foram similares para todas as amostras. Quando expressas por volume de ar (rev./m3), as potências mutagênicas foram proporcionais às concentrações de PTS das três cidades (Limeira > Quioto > Estocolmo). Quando aplicadas as razões diagnósticas foi possível verificar uma mistura de fontes de poluição em Limeira e Estocolmo. As razões diagnósticas dos HPA indicam que a amostra de Limeira é composta por partículas frescas, com um leve predomínio de fontes de combustão. A amostra de Estocolmo apresenta partículas envelhecidas e com predomínio de fontes petrogênicas. Esse resultado foi inesperado, uma vez que as condições atmosféricas de Limeira são mais favoráveis ao envelhecimento das partículas que as de Estocolmo. O estudo do transporte das massas de ar explica, pelo menos em parte, a presença de partículas envelhecidas na amostra de Estocolmo. Sugere também que a redução das concentrações de HPA em Limeira depende do controle das fontes locais, enquanto em Estocolmo o controle das fontes locais não seria eficiente na diminuição desses poluentes. Em paralelo a avaliação das amostras de PTS, foi desenvolvido um novo protocolo miniaturizado do teste Salmonella/microssoma em microssuspensão. O protocolo utilizando microplacas de 12 poços foi validado empregando treze compostos mutagênicos testados com três linhagens de Salmonella selecionadas baseadas em suas diferentes frequências de reversão espontânea (baixa, média e alta). Tanto o teste miniaturizado em microplacas de ágar (MPA) quanto o em microssuspensão apresentaram sensibilidades semelhantes, concluindo que o MPA é uma ferramenta promissora e pode ser particularmente adequado para estudos ambientais como Análise Dirigida pelo Efeito (EDA) ou programas de monitoramento. / Atmospheric particulate matter (PM) is associated with various diseases and has recently been classified as carcinogenic to humans (Group 1) by the International Agency for Research on Cancer (IARC). Although the PM toxicity, including genotoxicity, is known to be related to particle size, the contribution of its chemical composition to the observed effects has not yet been fully elucidated. The size and composition of the particles are influenced by meteorological and climatic characteristics, especially the temperature and the period of solar radiation. The Salmonella/microsome assay is the most used for the evaluate of the mutagenicity of PM samples, however a small number of strains are used. The objective of this study was to investigate and compare the influence of the different atmospheric and climatic conditions of the cities of Limeira (Brazil), Stockholm (Sweden) and Kyoto (Japan) on mutagenicity profiles, using 11 strains with different selectivity, and chemical composition profiles of total particles suspended (TPS) composed samples collected during the winter of these cities. For the results to be directly compared, the same methodology was adopted, including the sampling procedure, the sample preparation protocol, the mutagenicity test protocol, and the chemical analysis techniques for the identification of polycyclic aromatic hydrocarbons (PAH) and their alkylated derivatives. Limeira presented the highest concentration of TPS (99.0 µg/m3), followed by Kyoto (28.0 µg/m3) and Stockholm (6.2 µg/m3). Although the TPS concentration in Limeira was 16-fold higher than in Stockholm and 3.5-fold higher than in Kyoto, the percentage of extracted organic material (EOM) obtained was 9, 15 and 5%, respectively. The extracts from the TPS samples collected in the three cities presented mutagenic activity for all strains, both in the absence and presence of S9, except for TA102 that did not detect the mutagenic activity in any of the extracts, and for YG7108 in the presence of S9, only for the Limeira sample. Despite the differences in meteorological and climatic conditions of these cities, their mutagenicity profiles were similar. The lower mutagenic potency for YG7108 in the absence of S9 and the non-mutagenicity in the presence of S9 reveals the lower contribution of alkylating agents to the mutagenicity of the Limeira sample in relation to the other cities. Considering the selectivity of the strains used, we can also observe a greater contribution of compounds that cause DNA oxidative damage and of alkylating agents for the mutagenicity of the Kyoto sample. However, according to the responses of the strains that have increased activity of the nitroreductase (NR) and/or O-acetyltransferase (OAT) enzymes, YG1021, YG1024 and YG1041, show the highest contribution of nitroarenes and aromatic amines for the mutagenicity of the three cities, with emphasis on those mutagens that depend on OAT activation. The mutagenic potencies expressed as a function of EOM mass (rev./µg EOM) were similar for all samples. When expressed by air volume (rev./m3), the mutagenic potencies were proportional to TPS concentrations of the three cities (Limeira> Kyoto> Stockholm). When applying the diagnostic ratios, it was possible to verify a mixture of sources of pollution in Limeira and Stockholm. The diagnostic ratios of the PAH indicate that Limeira sample is composed of fresh particles with a slight predominance of combustion sources. The Stockholm sample present aged particles with a predominance of petrogenic sources. This result was unexpected, since Limeira\'s atmospheric conditions are more favorable to particle aging than Stockholm\'s. The study of the transport of air masses explains, at least in part, the presence of aged particles in the Stockholm sample. It also suggests that the reduction of PAH concentrations in Limeira depends on the control of local sources, while in Stockholm control of local sources would not be efficient in reducing these pollutants. In parallel to the evaluation of the TPS samples, a new miniaturized protocol of the Salmonella/microsome assay in microsuspension was also developed. The protocol using 12-well microplates was validated by employing thirteen mutagenic compounds tested with three selected Salmonella strains based on their different spontaneous reversion frequencies (low, medium and high). Both the miniaturized microplate agar (MPA) and the microsuspension assay presented similar sensitivities, concluding that MPA is a promising tool and can be particularly suitable for environmental studies such as effect-directed analysis (EDA) or monitoring programs.

L’anticipation et l’adaptation de nos sociétés aux bouleversements résultants du changement climatique sont aujourd’hui des questions majeures guidant les activités humaines et l’action publique. Néanmoins, la prévision reste un défi essentiel en raison des fortes incertitudes existantes et il est primordial de continuer à progresser dans la compréhension des mécanismes à l’origine de ces bouleversements. Au sein du cycle hydrologique, le réservoir de surface (incluant les lacs, les rivières et les plaines d’inondation) occupe une place importante car il est l’une des principales ressources en eau des écosystèmes et des populations. Or, la dynamique des stocks d’eau de surface est toujours mal connue aux échelles régionale et globale, du fait de l’absence de mesures pluriannuelles d’extension et de hauteur d’eau des zones inondées, et de la prise en compte limitée de ces variables dans les modèles hydrologiques et hydrodynamiques. La télédétection spatiale offre désormais la possibilité d’effectuer un suivi des stocks d’eau de surface en utilisant la complémentarité entre l’imagerie multi-spectrale, permettant de cartographier les étendues inondées, et l’altimétrie radar fournissant des séries temporelles de hauteur d’eau des hydro-systèmes continentaux. L’objectif de ma thèse est de tirer le meilleur parti de la complémentarité entre les différents types d’observations spatiales pour évaluer les ressources en eau du réservoir de surface et mesurer la dynamique des transferts d’eau des continents aux océans, et son impact sur la zone côtière en utilisant les concentrations de matières en suspension comme traceur des masses d’eau. Depuis le milieu des années 1990, la multiplication des missions multi-spectrales à moyenne résolution (< 1 km de résolution spatiale) et des altimètres radars à haute précision (comprise entre 10 et 30 cm) permet de réaliser un suivi hebdomadaire à mensuel des volumes d’eau de surface dans les grands bassins fluviaux. / The anticipation and adaptation of our societies to the upheavals resulting from climate change are today major issues guiding human activities and public action. Nevertheless, the forecast remains a key challenge because of the strong uncertainties that exist and it is essential to continue to progress in understanding the mechanisms behind these upheavals. Within the hydrological cycle, the surface reservoir (including lakes, rivers and floodplains) occupies an important place as it is one of the main water resources of ecosystems and populations. However, the dynamics of surface water stocks are still poorly known at the regional and global scales, due to the absence of multi-year measures of extension and water depth of the flooded areas, and the limited consideration of these variables in hydrological and hydrodynamic models. Satellite remote sensing now offers the possibility of monitoring surface water stocks by using the complementarity between multispectral imagery, allowing to map flooded areas, and the radar altimetry providing time series of water depths of continental hydro systems. The aim of my thesis is to make the most of the complementarity between the different types of spatial observations to evaluate the water resources of the surface reservoir and to measure the dynamics of water transfers between land and ocean, and its impact on the coastal zone using suspended particulate matter as tracer of water bodies. Since the mid-1990s, the multiplication of multispectral missions with medium resolution (<1 km of spatial resolution) and high-precision radar altimeters (between 10 and 30 cm) makes it possible to carry out weekly-to-monthly monitoring of volumes surface water in large river basins.

<div>Agricultural headwater streams in the Midwestern United States are subject to contaminants from fields, increased sedimentation, and degradation of natural habitat. Previous research has shown that physical instream habitat degradation better explained variation in fish community structure than water chemistry. However, these studies did not include total suspended solids (TSS), which are considered a major freshwater contaminant. The objective of this study is to determine whether total suspended solids better explains fish community structure than other variables in agricultural headwater streams. Mixed linear effects modeling was used to determine the set of independent variables that best predicts each of the fish response variables of species richness, Shannon diversity index, fish density, and index of biotic integrity. Standardized coefficients were used to determine which independent variable in each of the models had the largest influence on fish response metrics. The set of independent variables that best explained species richness were mean total suspended solids, imidacloprid, discharge, and substrate richness. Shannon diversity index was explained best by the combination of maximum total suspended solids, mean total suspended solids, atrazine, total nitrogen, and discharge. Fish density was explained best by the percentage of silt and clay, dissolved oxygen, the percentage of canopy cover, cover type richness, and discharge. IBI was explained best by the combination of the percentage of silt and clay, total phosphorus, mean total suspended solids, and dissolved oxygen. Total suspended solids was the most influential independent variable for fish species richness and Shannon diversity, however the percentage of silt and clay in benthic sediments was the most influential independent variable for fish density and IBI. Results also indicate discharge and total phosphorus as being influential to fish community metrics. The results from this study suggest that models containing a combination of different types of independent variables best explain fish community structure. This study supports the use of conservation and restoration practices that reduce total suspended solids and the amount of silt and clay present in bed sediments to increase fish community integrity of agricultural headwater streams of the Midwestern United States.</div>

Estuarine circulation and its associated transport processes drive the environmental integrity of many near-shore habitats (the coastal ocean, rivers, estuaries and emergent wetlands). A thorough understanding and consideration of this circulation is, therefore, vital in the proper management of these habitats. The aim of this study is to bring together theory and new satellite observations in the Columbia River Estuary to increase our understanding of estuarine circulation and transport. Surface reflectance measurements gathered by the Moderate Imaging Spectroradiometer (MODIS) are first compared to in situ observations to develop an empirical model for remotely derived surface turbidity. Results indicate that MODIS data significantly correlate with in situ measurements of turbidity throughout the CRE (R2 = 0.96). Remote estimates of turbidity are then used to explore the physical processes that drive their spatial distribution. Although the response to different hydrodynamic conditions varies throughout the system, global levels of turbidity are most sensitive to fluvial and tidal inputs and increase during spring tides and high river flow. As a result, the turbidity field has temporal cycles that are consistent with the frequency of these processes. The location of the estuarine turbidity maximum (ETM) is highly dynamic and typically migrates downstream as the tidal velocity or river flow increases. The ETM becomes trapped near the Megler Bridge (river kilometer 20), however, and the presence of strong topography in this region suggests there exists an interaction between bottom topography and sediment transport. A 2-D semi-analytical model, developed herein from the simplified Navier-Stokes equations, confirms that topographic features exhibit substantial influence on longitudinal turbidity distributions. The model considers the coupled, tidally-averaged velocity (composed of gravitational circulation, internal tidal asymmetry, and river flow) and salinity fields and assumes a condition of morphodynamic equilibrium to estimate the distribution of sediment for arbitrary channel configurations. Model simulations demonstrate that topographic highs tend to increase local seaward sediment fluxes, and that topographic lows increase local landward sediment fluxes. Sediment flux convergence near topographic highs compresses the local turbidity distribution, whereas flux divergence near topographic lows dilates the distribution and, under appropriate conditions, produces multiple ETMs. In summary a combination of the model and satellite data has given valuable new insights into the sediment dynamics of estuarine environments; in particular, both show that turbidity distribution and ETM location vary considerably with tidal and river flow conditions, fluctuating on a variety of timescales, and are heavily influenced by bottom topography.

La connaissance de la dynamique des matières en suspension (MES) et des turbidités en milieu côtier est un élément clé pour comprendre les écosystèmes marins. Les processus de floculation/défloculation représentent des mécanismes essentiels contrôlant la dynamique des caractéristiques physiques des MES et, par conséquent, leur devenir dans l’environnement. Cette thèse a pour objectif de mieux comprendre la caractérisation des MES et leur variabilité de l’échelle tidale à l’échelle annuelle en mer côtière, en réponse aux différents forçages hydrodynamiques, hydrologiques et biologiques ayant lieu à l’interface entre estuaire et baie. À cette fin, des capteurs optiques et acoustiques ont été déployés in situ, lors de 6 campagnes en mer réalisées en 2016, en vue de caractériser et de quantifier les MES dans la colonne d’eau, à la sortie de l’embouchure de l’estuaire de Seine (France). Une analyse critique des mesures granulométriques issues du LISST-100X est proposée. Un travail de quantification des incertitudes de mesure associées à l’estimation de la concentration en MES a mis en avant le risqué lié à un rinçage insuffisant des filtres et a conduit à définir une masse minimum à filtrer, de l'ordre de 10 mg, quelle que soit la concentration. Ce travail a notamment permis de proposer une méthode d’estimation d’un volume optimal de filtration basée sur une mesure préalable de turbidité de référence. Les mesures à haute fréquence ont montré que la dynamique des caractéristiques physiques des MES en termes de concentration et de taille médiane sont contrôlées par le cycle advection-floculation-sédimentation-remise en suspension. Ces résultats ont permis de hiérarchiser les paramètres de contrôle des processus de floculation. La turbulence a été identifiée comme le principal paramètre de contrôle à l’échelle tidale et semi-lunaire, la taille médiane maximale des flocs étant inversement corrélée à l’intensité des courants. À l’échelle saisonnière, la variation des caractéristiques des MES (taille, densité, vitesse de chute) est liée à la variabilité du contenu en matière organique (MO), avec une floculation favorisée quand le taux de MO particulaire augmente. Ces flocs plus grands et moins denses sont également plus résistants à la fragmentation induite par cisaillement turbulent. Ces travaux ont également permis de discuter le concept de structuration fractale des flocs.La dynamique de la dimension fractale, déterminée en combinant des données in situ de concentration en MES et de distribution en classe de taille, peut traduire des variations de composition des MES, mais peut aussi résulter d'incertitudes de mesure associées aux instruments. Ces dernières sont discutées dans cette thèse. La variabilité saisonnière des caractéristiques des MES est plus prononcée au large que dans la zone sous influence de l’estuaire de Seine. À partir des observations ponctuelles en Baie de Seine et d’une série de mesure long terme dans la zone côtière belge acquise par le laboratoire RBINS, les méthodes acoustiques et optiques sont combinées. Ces dernières ont mis en avant une augmentation de la rétrodiffusion acoustique lorsque la densité moyenne des flocs diminue. / The knowledge of suspended particulate matter (SPM) and turbidity dynamics in coastal waters is essential for studying marine ecosystems. Flocculation/deflocculation processes are crucial mechanisms controlling the dynamics of SPM physical characteristics and thus, the fate of these SPM in the environment. This PhD thesis focuses on the SPM characteristics and their dynamics in coastal waters, from tidal to annual scales, in response to hydrodynamic, hydrological and biological forcing that take place at the interface between estuaries and coastal seas. To this end, optical and acoustic sensors were deployed in situ through 6 field campaigns in 2016, in order to characterize and quantify SPM in the water column, near the mouth of the Seine estuary (France). A critical analysis of LISST-100X measurements in coastal waters is presented. The quantification of uncertainties on SPM concentration measurements is investigated: it highlights the crucial effect of salt retention, and the need for a minimum mass to filter, around 10 mg. A method for estimating an optimal filtration volume based on a reference turbidity measurement is proposed. High frequency measurements show that the dynamics of SPM and median diameter are controlled by the advection-flocculationsedimentation-resuspension cycle.These results allow to classify the factors controlling flocculation processes. Turbulence is identified as the main factor at the semi-diurnal and semi-lunar tidal scales, as the maximum median size of flocs decreases when the tidal currents intensify.At the seasonal scale, the variation of SPM characteristics (size, density, settling velocity) is correlated to the variability of the organic matter (OM) content: in particular, flocculation is enhanced by an increase of the particular OM fraction. The resulting larger and lesser dense flocs are also more resistant to the fragmentation induced by shear. This work also investigates the fractal approach applied to flocs. The fractal dimension variability, calculated by combining in situ data of SPM concentration and particle size distribution, can be associated to variations of the SPM composition, but can also result from uncertainties linked to instrument limitations. The latter are discussed in this PhD thesis. The seasonal variability of SPM characteristics is more pronounced offshore than at the mouth of the Seine estuary. From short-term observations in the Seine Bay and from long-term series in the Belgian coastal zone provided by the RBINS, optical turbidity and acoustic backscatter measurements are combined. They highlight an increase of the acoustic backscatter intensity when mean floc density.

<p>This thesis presents models applicable for aquatic ecosystems. Geographical Information Systems (GIS) form an important part of the thesis. The dynamic mass balance models focus on nutrient fluxes, biotic/abiotic interactions and operate on different temporal and spatial scales (site, local, regional and international). The relevance and role of scale in mass balance modelling is a focal point of the thesis.</p><p>A mesocosm experiment was used to construct a model to estimate the nutrient load of phosphorus and nitrogen from net cage fish farming (i.e., the site scale). The model was used to estimate what feeding conditions that are required for a sustainable aquaculture scenario, i.e., a zero nutrient load situation (a linkage between the site scale and the regional scale). </p><p>A dynamic model was constructed for suspended particulate matter (SPM) and sedimentation in coastal areas (i.e., the local scale) with different morphometric characteristics and distances to the Sea. The results demonstrate that the conditions in the Sea (the regional and international scale) are of fundamental importance, also for the most enclosed coastal areas.</p><p>A foodweb model for lakes was transformed and recalibrated for Baltic Sea conditions (i.e., the international scale). The model also includes a mass balance model for phosphorus and accounts for key environmental factors that regulate the presuppositions for production and biomasses of key functional groups of organisms. The potential use of the new model for setting fish quotas of cod was examined.</p><p>For the intermittent (i.e., regional) scale, topographically complex areas can be difficult to define and model. Therefore, an attempt was made to construct a waterscape subbasin identification program (WASUBI). The method was tested for the Finnish Archipelago Sea and the Okavango Delta in Botswana. A comparison to results from a semi-random delineation method showed that more enclosed basins was created with the WASUBI method.</p>

<p>This thesis presents predictive models for important variables concerning eutrophication effects in lakes. The keystone is a dynamic phosphorus model based on ordinary differential equations. By calculating mass fluxes of phosphorus into, within and out from a lake, the concentrations of different forms of phosphorus in different compartments of the lake are estimated.</p><p>The dynamic phosphorus model is critically tested and several improvements are presented, including two new compartments for colloidal phosphorus, a sub-model for suspended particulate matter (SPM) and new algorithms for lake outflow, water mixing, diffusion, water content and organic content of accumulation sediments are implemented. Predictions with the new version show good agreement against empirical data in five tested lakes.</p><p>The sub-model for SPM uses the same driving variables as the basic phosphorus model, so the inclusion of this model as a sub-model does not require any additional variables. The model for SPM may also be used as a separate model giving monthly predictions of suspended particulate matter in two water compartments and one compartment with SPM available for resuspension in ET-sediments.</p><p>Empirical data from Lake Erken (Sweden) and Lake Balaton (Hungary) are used to evaluate the variability in settling velocity of SPM. It is found that the variability is substantial and may be accounted for by using a dimensionless moderator for SPM concentration. Empirical data from accumulation area sediments in Lake Erken are used to develop a model for the dynamics of phosphorus sedimentation, burial and diffusion in the sediments. The model is shown to provide reasonable monthly predictions of four functional forms of phosphorus at different sediment depths.</p><p>Simulations with the lake phosphorus model using two different climate scenarios indicate that lakes may respond very differently to climate change depending on their physical character. Lake Erken, with a water retention time of 7 years, appears to be much more sensitive than two basins of Lake Mälaren (Sweden) with substantially shorter retention times. The implication would be that in eutrophic lakes with long water retention times, eutrophication problems may become serious if the future becomes warmer. This will be important in contexts of lake management when remedial measures against lake eutrophication have to be taken.</p>

Les potentiels de nouveaux moyens d’investigation en sciences de l’environnement, tel que la géochimie des isotopes stables du Cu et du Zn et l’étude de la spéciation des éléments en trace métalliques (ETM) par titrage acidimétrique sont évalués dans le cadre de trois cas d’études, faisant intervenir des matériaux de matrices et de concentrations en ETM variables. Afin de mettre en évidence de très subtiles variations des compositions isotopiques du Cu et du Zn dans les matériaux particulaires variablement pollués, une attention particulière à été voué à la mise au point des méthodes de séparation chimique et d’analyse par MC-ICP-MS. La méthodologie analytique a été développée afin d’exploiter tout les avantages techniques offerts par le spectromètre de masse à l’ULB. Différents modes d’introduction (plasma humide vs plasma sec (Aridus/DSN), d’acquisition des mesures (mode statique ou dynamique) de correction du biais de masse (dopage au Cu, au Zn et au Ga; correction SSBM, SSBC et EEN)ont été évaluées, pour leurs effets sur la précision et la reproductibilité des mesures. D’autres travaux ont permis de quantifier les effets des interférences spectrales et non spectrales par les éléments de la matrice (Ti, Cr, Co,Fe, Ba, Si, Na, Mg et Ca) et par le rapport dopant/analyte en vue de comprendre les sources d’inexactitudes des mesures isotopiques et d’y remédier. Les données isotopiques acquises sur des minerais, sur des (plaques de) dépôts atmosphériques (usine Pb-Zn Metaleurop de Noyelles-Godault), sur des sédiments et des matières en suspension (estuaire de l’Escaut et mer du Nord), révèlent des variations significatives des compositions isotopiques du cuivre et du zinc. Dans l’estuaire de l’Escaut, les variations temporelles (carotte non perturbée, enregistrant la sédimentation sur 30 ans) et spatiales (matières en suspension échantillonnées sur 100 km, selon le profil de salinité) sont caractérisées par des signatures isotopiques en Zn anticorrélées à celle du Cu. Les données peuvent permettre de distinguer le fond géochimique naturel (le « réservoir naturel »), les pollutions diffuses (le « réservoir anthropique commun »- intégrant de multiples sources de contamination en provenance du bassin versant) et certaines pollutions ponctuelles associées à la métallurgie du Zn. Alors que les procédés utilisés en métallurgie sont eux,capable de produire des signatures isotopiques très fractionnées en Zn, ni la diagenèse précoce dans les sédiments, ni les gradients physicochimiques développés lors du mélange des eaux continentales et marines ne sont capables d’affecter significativement la signature isotopique du Cu et du Zn en phase particulaire. Dans de tels milieux, les signatures isotopiques en Cu et Zn (ainsi que celles en Pb) semblent n’être contrôlées que par le mélange conservatif entre le « réservoir naturel » et le « réservoir anthropique commun ». Les données en elles-mêmes constituent la seule base de données isotopiques en Cu, Zn, Pb dans des matériaux particulaires anthropisés estuariens et marins jamais produite à ce jour. Elles permettent de prédire que la signature isotopique en Cu du « réservoir naturel » devrait être légèrement plus enrichie en isotopes lourds que celle du réservoir « anthropique commun » (par analogie au Zn). Si la géochimie isotopique du Cu et du Zn a le potentiel de distinguer différentes sources de pollution, l’évaluation des risques liés à ces pollutions doit tenir compte des formes chimiques des ETM, typiquement mises en évidences par les méthodes d’extraction sélectives (séquentielles, dont le protocole BCR). La méthode alternative proposée pour l’étude de la spéciation, plus versatile (capable de mettre en évidence des phases non prédéfinies et applicable à une variété de matrices plus large) est capable de surmonter plusieurs limitations (dont les problèmes de sélectivité liées au mauvais contrôle de l’acidité du milieu réactionnel) caractéristiques des méthodes traditionnelles. Appliquée à des sédiments marins/estuariens et des sols, la méthode de titrage acidimétrique s’est révélée capable de quantifier les proportions relatives des phases acido-solubles (principal problème des méthodes d’extraction sélectives), ce qui permet d’établir univoquement la spéciation des éléments qui leurs sont associés, d’évaluer la réactivité des ETM vis-à-vis du pH, mais aussi d’évaluer la sélectivité des protocoles d’extraction séquentielle « BCR » et « Tessier ». La méthode de titrage acidimétrique apporte également des informations importantes sur la géochimie des éléments majeurs dans l’estuaire de l’Escaut et met en évidence les particularités minéralogiques des sols pollués en comparaison aux sédiments. Les études de cas démontrent que ces nouvelles techniques ont des applications dans le domaine de la géochimie de l’environnement, mais peuvent également être vouées, sur le moyen/long terme à une utilisation sortant du cadre des sciences exactes. En effet, l’identification/ discrimination des sources de pollutions et l’évaluation des risques de remobilisation des polluants métalliques font partie des nombreuses questions posées par la société aux scientifiques de l’environnement.

This thesis presents models applicable for aquatic ecosystems. Geographical Information Systems (GIS) form an important part of the thesis. The dynamic mass balance models focus on nutrient fluxes, biotic/abiotic interactions and operate on different temporal and spatial scales (site, local, regional and international). The relevance and role of scale in mass balance modelling is a focal point of the thesis. A mesocosm experiment was used to construct a model to estimate the nutrient load of phosphorus and nitrogen from net cage fish farming (i.e., the site scale). The model was used to estimate what feeding conditions that are required for a sustainable aquaculture scenario, i.e., a zero nutrient load situation (a linkage between the site scale and the regional scale). A dynamic model was constructed for suspended particulate matter (SPM) and sedimentation in coastal areas (i.e., the local scale) with different morphometric characteristics and distances to the Sea. The results demonstrate that the conditions in the Sea (the regional and international scale) are of fundamental importance, also for the most enclosed coastal areas. A foodweb model for lakes was transformed and recalibrated for Baltic Sea conditions (i.e., the international scale). The model also includes a mass balance model for phosphorus and accounts for key environmental factors that regulate the presuppositions for production and biomasses of key functional groups of organisms. The potential use of the new model for setting fish quotas of cod was examined. For the intermittent (i.e., regional) scale, topographically complex areas can be difficult to define and model. Therefore, an attempt was made to construct a waterscape subbasin identification program (WASUBI). The method was tested for the Finnish Archipelago Sea and the Okavango Delta in Botswana. A comparison to results from a semi-random delineation method showed that more enclosed basins was created with the WASUBI method.

This thesis presents predictive models for important variables concerning eutrophication effects in lakes. The keystone is a dynamic phosphorus model based on ordinary differential equations. By calculating mass fluxes of phosphorus into, within and out from a lake, the concentrations of different forms of phosphorus in different compartments of the lake are estimated. The dynamic phosphorus model is critically tested and several improvements are presented, including two new compartments for colloidal phosphorus, a sub-model for suspended particulate matter (SPM) and new algorithms for lake outflow, water mixing, diffusion, water content and organic content of accumulation sediments are implemented. Predictions with the new version show good agreement against empirical data in five tested lakes. The sub-model for SPM uses the same driving variables as the basic phosphorus model, so the inclusion of this model as a sub-model does not require any additional variables. The model for SPM may also be used as a separate model giving monthly predictions of suspended particulate matter in two water compartments and one compartment with SPM available for resuspension in ET-sediments. Empirical data from Lake Erken (Sweden) and Lake Balaton (Hungary) are used to evaluate the variability in settling velocity of SPM. It is found that the variability is substantial and may be accounted for by using a dimensionless moderator for SPM concentration. Empirical data from accumulation area sediments in Lake Erken are used to develop a model for the dynamics of phosphorus sedimentation, burial and diffusion in the sediments. The model is shown to provide reasonable monthly predictions of four functional forms of phosphorus at different sediment depths. Simulations with the lake phosphorus model using two different climate scenarios indicate that lakes may respond very differently to climate change depending on their physical character. Lake Erken, with a water retention time of 7 years, appears to be much more sensitive than two basins of Lake Mälaren (Sweden) with substantially shorter retention times. The implication would be that in eutrophic lakes with long water retention times, eutrophication problems may become serious if the future becomes warmer. This will be important in contexts of lake management when remedial measures against lake eutrophication have to be taken.