By continually supplying nutrients, a turbidostat presents a steady environment to analyze bacterial growth dynamics. This makes it possible to model microbial community succession in a simple and more realistic way than is possible with batch culture. The problem with current commercial turbidostats are their industrial size and price. With the use of 3D printed parts, printed circuit boards and laser cut pieces, all readily obtained online and in collaboration with laboratories here at McMaster University, I've created a relatively cheap custom turbidostat, ideally suited for longitudinal studies of microbial consortia that can accommodate eight separate experiments simultaneously. I have modified the design to enable microbial growth at 37°C under anoxic conditions, by changing how the growth media and gas is handled. I have also improved sample collection to make it more convenient and flexible. Cell growth dynamics were interrogated separately with one strain of a facultative anaerobe (E. coli) and one strain of an obligate anaerobe (Bacteroides thetaiotaomicron) bacteria. For the individual strains, real-time optical density and dilution rate vs time graphs were created showing that these microbes can be reproducibly cultured, holding steady optical density rates for extended periods of time. Future directions include culturing a complex community without contamination by inoculating the system with microorganisms from an infant stool sample. Community composition and metabolite dynamics could then be analyzed by sampling over time. / Thesis / Master of Science (MSc)

Ce travail de thèse porte sur l’élaboration d’un modèle prédictif de la croissance d’une microalgue eucaryote en photobioréacteur, basé sur une analyse du métabolisme de croissance de l’algue verte modèle Chlamydomonas reinhardtii en conditions photoautotrophes. Le modèle proposé appartient à la famille des modèles biochimiquement structurés. Outre la détermination prédictive des rendements de conversion des substrats en produits assurée par la représentation du métabolisme énergétique cellulaire, le modèle permet de décrire le comportement dynamique des microalgues en photobioréacteurs, en incluant l’adaptation de la teneur en pigments aux conditions de lumière fluctuantes, pour une large plage de conditions opératoires. Le modèle a été identifié et validé sur une base de données expérimentales obtenues pour des cultures continues et discontinues, élaborée dans le cadre de ce travail de thèse, comportant des mesures intra- et extracellulaires. Cette modélisation du comportement d’une microalgue eucaryote en photobioréacteur a été étendue à une description de la phase abiotique du bioréacteur, reposant sur une analyse et une caractérisation du transfert de matière gaz-liquide (O2 / CO2) dans les conditions étudiées. Cette modélisation de la phase abiotique a conduit au développement d’un estimateur de la concentration en biomasse basé sur la mesure de la vitesse nette de production d’oxygène. Cet outil a été validé expérimentalement, et a servi au suivi de cultures en temps réel, ainsi qu’au pilotage d’un photobioréateur en mode turbidostat. / The present work deals with the elaboration of a predictive model for the growth of eukaryotic microalgae in photobioreactors, built on an analysis of the growth metabolism of the model green alga Chlamydomonas reinhardtii in photoautotrophic conditions. The proposed model belongs to the family of biochemically-based structured models. Besides a comprehensive representation of the conversion yields of substrates into products based on the description of cellular energetic metabolism, the model allows to predict the dynamic behaviour of microalgae growth in photobioreactors for a wide range of operating conditions, including pigment adaptation to fluctuating light conditions. The model has been identified and validated on an experimental dataset obtained for continuous and batch cultures carried out as part of the PhD, including intraand extracellular measurements. Modelling of microalgal growth behaviour in photobioreactors has been extended to a description of the reactor abiotic phase, relying on the analysis and characterization of gas-liquid mass transfer (O2 / CO2) under the conditions of the study. The abiotic phase modelling has led to the development of a biomass concentration estimator based on the measurement of the net oxygen production rate. This tool has been validated experimentally and has been applied to the real-time monitoring of cultures, as well as the control of a photobioreactor in turbidostat mode.