The purpose of our study was to describe the dynamics of E. coli in the rivers of the Seine watershed. This watershed is characterized by a poor microbiological water quality due to an important population density, industrial activities and intensive agriculture. The main objectives of our study were to obtain field microbiological data describing the system and to understand the underlying mechanisms in order to build a model able to describe and predict the fate of the fecal bacteria at the scale of the whole watershed. This model was built in order to help management of microbiological water quality. To be able to describe and modelize the fate of fecal bacteria at the scale of such a large and complex system, it was first necessary to identify and quantify the sources of contamination and to be able to describe the processes affecting these bacteria once discharged to the rivers. The quantification of E. coli in the natural aquatic systems through the traditional culture-based techniques has been strongly questioned during the last years because they have a long response delay (24 to 48 h) and they do not take into account the viable but not culturable bacteria (VBNC) which nevertheless could represent a health thread. For this reason we developed in this study two alternative methods: the direct (without passing through a cultivation step) measurement of the â-D-glucuronidase activity and a fluorescent in situ hybridization with an E. coli specific probe coupled with a viability test. The first has the main advangate to give a very rapid evaluation of the E. coli content of the water while the second one was able to enumerate VBNC E. coli. The study of the sources of contamination of the rivers revealed the predominance, in this anthropogenicised watershed, of the point sources (effluents from wastewater tretament plants). Diffuse sources soil leaching and surface runoff) may nevertheless be locally important and have also be quantified. The influence of the land use on these diffuse sources has been quantified demonstrating that soil leaching and surface runoff of grazed areas was mainly responsible of fecal contamination in rural areas. The attachment of E .coli to particles in the river waters was studied as it can influence the fate of this kind of bacteria in the aquatic systems. A linear regression was found between the percentage of particle-associated E. coli and the SM content of the water. Settling velocities of particle-associated E. coli were determined. The mortality rate of E. coli was analysed following various perspectives: the river order, the attachment of E. coli to particles, the importance of the cultivability loss in the disappearance of these bacteria and the importance of protozoan grazing. The results confirmed the importance of grazing and showed that the abondance of free living E. coli bacteria decreased two times more rapidly than particle-associated E. coli. In the Seine river, we showed that the net loss of culturable E. coli was 1.5 time higher than the net loss of viable E. coli. Finally, no relationship was found between mortality rate and river order. These results have been used to build two models CF-SENEQUE and CF-SiAM-3D which describe E. coli dynamics respectively for the whole Seine river watershed and for the Seine for the estuary. Comparisons model calcultaions with field data of fecal coliforms abundances showed that these models correctly simulated the longitudinal distribution of fecal colifoms in the main rivers of the Seine watershed and in the estuary. These models were used to test the impact on the microbiological water quality of various scenarii of wastewater management .

La présente étude porte sur la modélisation à l'échelle du produit, puis à l'échelle du procédé, des transferts de chaleur et de matière, au cours du séchage solaire de viande de bœuf, et entre ce produit et son environnement. A l'échelle du produit, les isothermes de désorption et cinétiques de séchage de la viande de bœuf sont déterminées expérimentalement, puis modélisées afin de caractériser la viande sur le plan hygroscopique et de prédire son comportement au cours du séchage. L'étude et la modélisation du comportement hygroscopique permet de passer à une échelle du procédé par la mise en place d'un modèle de séchage solaire de la viande de bœuf adapté à un séchoir solaire, conçu et réalisé pour le séchage des produits carnés dans un contexte tropical (séchoir solaire de Kilichi). Le modèle est validé par des essais expérimentaux menés dans un contexte tropical (N'Gaoundéré-Cameroun). L'impact du séchage solaire sur la qualité du produit est étudié à travers des analyses microbiologiques et biochimiques effectuées sur des échantillons de viande de bœuf séchés, à l'échelle du laboratoire, dans un séchoir solaire conçu et réalisé pour une bonne maitrise des paramètres de séchage solaire. Les analyses microbiologiques montrent que l'impact du rayonnement solaire sur la destruction des microorganismes est plus significatif que celui de la température de séchage. Les analyses biochimiques montrent que le séchage induit une forte oxydation des lipides mais n'a pas d'impact sur la teneur en lipide de la viande. / This study deals with the modeling of heat and mass transfers during solar drying of beef, and between this product and its environment, at the scale of the product and thereafter at the scale of the process. At the scale of the product, desorption isotherm and drying kinetics of fresh beef were determined and modeled to characterize and predict the hygroscopic behavior of the beef during its drying. At the scale of the process, we proposed a model of solar drying of beef, adapted to a solar dryer that was designed and manufactured for the solar drying of beef in tropical environment (Solar dryer of Kilichi). This model was validated by experimental tests conducted in situ, in a tropical country (Cameroon- N'gaoundéré). The impact of solar drying on the quality of the product was investigated through microbiological and biochemical analyzes conducted on beef samples, dried at the laboratory with a solar dryer that was designed and manufactured to obtain a good control of solar drying parameters. These analyzes showed that the impact of solar radiation on the destruction of microorganisms is more significant than that of the drying temperature. Biochemical analysis showed that drying induces a severe lipid oxidation but has no impact on the lipid content of the meat.

L'objectif de ces travaux est de contribuer à augmenter la connaissance sur la réutilisation des eaux de ruissellement de toiture. Une maison individuelle située en zone rurale et équipée d'une toiture en tuiles, et un bâtiment collectif situé en zone périurbaine avec une toiture terrasse bitumée, ont été étudiés. Les eaux de pluie collectées sont réutilisées dans l'habitat pour l'alimentation des chasses d'eau. Un suivi analytique a permis l'étude de la qualité de l'eau en plusieurs points des systèmes. Les eaux stockées sont clairement non potables et révèlent une contamination microbiologique très variable. Des pathogènes ont été quantifiés à plusieurs reprises. La qualité de l'eau distribuée ne peut donc pas être garantie au cours de l'année, d'où la recommandation d'une désinfection de l'eau avant distribution. Un suivi hydraulique a permis de déterminer les volumes d'eau potable substitués et les taux de couverture de besoins en eau. Un modèle du comportement hydraulique a été développé puis utilisé pour discuter le dimensionnement des cuves. En présence d'une désinfection, les installations ne sont économiquement pas rentables. En l'absence de désinfection, le bilan économique est positif mais la période de retour sur investissement supérieure à la durée de vie de l'installation. Une première analyse de cycle de vie a été réalisée pour comparer les deux échelles étudiées. Dans les deux cas, les processus les plus impactants sont liés aux consommations électriques de l'éventuelle désinfection ultra-violet et des pompes. En présence d'une désinfection, l'échelle du bâtiment collectif présente des impacts plus faibles. Une deuxième analyse de cycle de vie a été réalisée pour étudier la pertinence de la substitution de l'eau du réseau d'eau public par de l'eau de pluie à l'échelle d'un particulier. Elle n'est pas probante d'un point de vue environnemental. / The aim of this work is to contribute to knowledge on roof runoff reuse. A private house located in rural area with a tiled sloping roof and a research building located in suburban zone with a bituminous flat roof were studied. The collected rainwater is reused for toilet flushing. An analytical monitoring allowed the study of water quality in several points of the systems. Stored rainwater is clearly non drinkable and reveals a very variable microbiological contamination. Pathogenic agents were quantified on several occasions. Distributed water quality cannot thus be guaranteed over the year and a disinfection before distribution is recommended. A hydraulic follow-up made it possible to determine substituted volumes of drinking water by collected rainwater and water saving efficiencies. A model of the hydraulic behaviour of systems was developed then used to discuss dimensioning of the tanks. In the presence of disinfection, the installations are economically not profitable. In the absence of disinfection, the economic assessment is positive but the payback period is longer than the lifespan of the installation. A first life cycle assessment was carried out to compare the two studied scales. In both cases, the more impacting processes are related to electricity consumptions of possible ultra-violet disinfection and the pumps. In the presence of disinfection, the scale of the building presents weaker impacts. A second life cycle assessment was carried out to study the relevance of the substitution of drinking water by rainwater on a private individual scale. It is not convincing of an environmental point of view.

Le bassin versant de l’Escaut (20 000 km²) est caractérisé par forte densité de population (plus de 500 habitants par km2) et une activité agro-pastorale intensive. Les rivières de ce bassin sont sévèrement affectées par les rejets d’eaux usées domestiques, les effluents d’élevage et les eaux de ruissellement des sols agricoles. Le but de cette étude est :(i) d’évaluer la qualité microbiologique de ces rivières ;(ii) d'identifier et de quantifier les différentes sources de contamination fécale à l’échelle du bassin de l’Escaut; (iii) d’étudier les processus qui contrôlent le devenir des bactéries fécales en rivière; (iv) de développer des modèles numériques sur la base des travaux expérimentaux permettant de prédire la concentration des bactéries indicatrices dans les rivières du bassin de l’Escaut.<p>L’évaluation de la qualité microbiologique des principales rivières du bassin est basée sur le dénombrement de deux indicateurs de contamination fécale (Escherichia coli et entérocoques intestinaux). Les abondances des deux indicateurs dans les principales rivières du bassin indiquent très clairement que les eaux et les sédiments de ces rivières sont fortement contaminés par des micro-organismes entériques. Les sources prédominantes de la pollution fécale de ces rivières sont les rejets des effluents des stations d’épuration. Les niveaux de contamination les plus élevés sont observés dans la Senne en aval de Bruxelles et s’expliquent par le faible débit de la Senne comparé aux débits des effluents des deux stations d’épuration de Bruxelles. Les niveaux de contamination atteignent leur maxima à l’aval de Bruxelles par temps de pluie en raison des surverses de réseaux unitaires.<p>Les connaissances acquises sur les apports des bactéries indicatrices par les sources ponctuelles et les sources diffuses et sur le devenir des bactéries indicatrices ont permis de développer un module décrivant la dynamique des E. coli dans les rivières. Ce module est original par le fait de considérer trois compartiments de bactéries fécales (libres, attachées aux particules dans la colonne d’eau et présentes dans les sédiments) qui sont affectés différemment par les processus de transport et de disparition. Ce module a été couplé à deux modèles décrivant l’hydrodynamique respectivement de l’ensemble des rivières du bassin (SENEQUE-EC) et de la partie fluviale de l’Escaut sous l’influence de la marée et son estuaire (SLIM-EC2). Ces deux modèles permettent de décrire la distribution temporelle et spatiale des E. coli dans les eaux de surfaces et de prévoir les modifications de la qualité microbiologique des eaux suite à des changements de gestion des eaux usées. <p> / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

The purpose of our study was to describe the dynamics of E. coli in the rivers of the Seine watershed. This watershed is characterized by a poor microbiological water quality due to an important population density, industrial activities and intensive agriculture. The main objectives of our study were to obtain field microbiological data describing the system and to understand the underlying mechanisms in order to build a model able to describe and predict the fate of the fecal bacteria at the scale of the whole watershed. This model was built in order to help management of microbiological water quality. <p>To be able to describe and modelize the fate of fecal bacteria at the scale of such a large and complex system, it was first necessary to identify and quantify the sources of contamination and to be able to describe the processes affecting these bacteria once discharged to the rivers. <p>The quantification of E. coli in the natural aquatic systems through the traditional culture-based techniques has been strongly questioned during the last years because they have a long response delay (24 to 48 h) and they do not take into account the viable but not culturable bacteria (VBNC) which nevertheless could represent a health thread. For this reason we developed in this study two alternative methods: the direct (without passing through a cultivation step) measurement of the â-D-glucuronidase activity and a fluorescent in situ hybridization with an E. coli specific probe coupled with a viability test. The first has the main advangate to give a very rapid evaluation of the E. coli content of the water while the second one was able to enumerate VBNC E. coli. <p>The study of the sources of contamination of the rivers revealed the predominance, in this anthropogenicised watershed, of the point sources (effluents from wastewater tretament plants). Diffuse sources soil leaching and surface runoff) may nevertheless be locally important and have also be quantified. The influence of the land use on these diffuse sources has been quantified demonstrating that soil leaching and surface runoff of grazed areas was mainly responsible of fecal contamination in rural areas.<p>The attachment of E .coli to particles in the river waters was studied as it can influence the fate of this kind of bacteria in the aquatic systems. A linear regression was found between the percentage of particle-associated E. coli and the SM content of the water. Settling velocities of particle-associated E. coli were determined. The mortality rate of E. coli was analysed following various perspectives: the river order, the attachment of E. coli to particles, the importance of the cultivability loss in the disappearance of these bacteria and the importance of protozoan grazing. The results confirmed the importance of grazing and showed that the abondance of free living E. coli bacteria decreased two times more rapidly than particle-associated E. coli. In the Seine river, we showed that the net loss of culturable E. coli was 1.5 time higher than the net loss of viable E. coli. Finally, no relationship was found between mortality rate and river order. <p>These results have been used to build two models CF-SENEQUE and CF-SiAM-3D which describe E. coli dynamics respectively for the whole Seine river watershed and for the Seine for the estuary. Comparisons model calcultaions with field data of fecal coliforms abundances showed that these models correctly simulated the longitudinal distribution of fecal colifoms in the main rivers of the Seine watershed and in the estuary. These models were used to test the impact on the microbiological water quality of various scenarii of wastewater management .<p> / Doctorat en sciences, Spécialisation biologie animale / info:eu-repo/semantics/nonPublished