Abstract
The stability and the performances of microalgae production systems are challenging issues, particularly in the context of the use of microalgae for the bioremediation of effluents and the production of biomass ofindustrial interest. One of the challenging axes of research is to take advantage of the diversity of microalgae through the use of assemblages (polycultures) for improving the overall production performances. However, theexploitation of these complex assemblages in an open system is subject to various stresses : biotic (undesirable microalgae or competing bacteria) and abiotic (limitation of resources, especially nitrogen and light). Thesestressing factors could influence interactions within assemblages, making thereby difficult the prediction and the optimization of the overall production. The objective of this thesis is to propose methods and tools forunderstanding, predicting, and optimizing the biomass production of an algal assemblage (natural or artificially designed microbial consortium) under the fluctuation of resources and culture conditions. More specifically,we have developed mathematical models based on dynamical systems that can be compared to laboratory and pilot-scale experiments. Firstly, an experimental method was defined for characterizing the specific growth rate of photosynthetic microorganisms, according to the limiting resource. Furthermore, a new functional estimation method is proposed. This method provides two growth curves that envelop scattering growth rate data and thus allow dynamic estimates of the state variables with guaranteed intervals. Secondly, the type of the interactions between two microalgae Chlorella sorokiniana and Scenedesmus pectinatus, which typically have successional trends in the outdoor ponds used for the treatment of urban wastewater, has been characterized ; first according to the fluctuation of the different forms of nitrogen (NH+4/NH3), and secondly to the light availability. The mathematical models associated with the experiments carried out during this thesis demonstrate that the initial development of an opportunistic microalgal species, which is more resistant to high levels of NH+4/NH3 was required for the subsequent development of a more efficient microalgal species under low light availability in these turbid processes. Third, we explored the paradoxical interactions occurring between microalgae and heterotrophic bacteria. Through the exudation of carbon, microalgae stimulate its competitors for common resources : nitrogen or phosphorus. We have thus studied the influence of this phenomenon by proposing a four-dimensional model. The mathematical analysis revealed that the equilibrium of coexistence is unique, and the installation of bacteria is robust, as it has been shown that the equilibrium without bacteria is unstable. We show that when the concentration of the mineral resource that continuously feeds the bioreactor is large enough, there are renewal rate values for which there is coexistence, while bacteria could not grow in the absence of microalgae under these same operating conditions. Finally, it has been shown, using simulations, that during coexistence their biomasses can oscillate. These results demonstrate the complexity of biotic interactions, provide methods applicable to other model organisms, and raise promising application possibilities for the optimization and the control of bioprocess dynamical systems for future work.