Developing an individual-based model to study the bacterial denitrification process

Resumen

It is crucial to study the denitrification process driven by bacteria as it is one of the most important environmental processes for several reasons: i) it has an application in the removal of nitrogen (N) from high-N waste materials ii), it is one of the mechanisms to N-fertilizer¿s loss, iii) it contributes to the emissions of gasses with large global warming potential, and iv) it is the mechanism by which the global nitrogen cycle is balanced. Many models have been developed in the framework of continuous models to deal with the complexity of the denitrification process in order to become predictive models, but some of the assumptions contained in them are not realistic enough in those contexts, and also they have their own constraints and limitations. On the other hand, the researchers have paid more attention to the role of microbial activity with the advance of experimental techniques. Discrete models, such as individual-based models (IBMs), can be developed and applied to microbial systems due to the fact that they allow representation of some intracellular characteristics regarding the complexity of the micro-organisms, which constitutes a key advantage of this modelling approach in the study of the different biotechnological processes. The IBM is able to incorporate and accommodate the behaviour of denitrifying bacteria, and investigate their metabolism from different and attractive perspectives. A key factor in modelling the microbial activity is the methodology followed to represent metabolic pathways. A cellular metabolic model could be based on a non-equilibrium thermodynamic approach such as the Thermodynamic Electron Equivalents Model (TEEM), which is developed for biomass yield prediction using the associated standard Gibbs free energies and the bioenergetics growth efficiency between cell anabolism and catabolism. The main objective of this doctoral thesis is to develop an IBM to study denitrification processes driven by denitrifying bacteria, using TEEM to write microbial metabolic reactions (MMRs) which represent the metabolic pathways as the center of the individual sub-model. Two new computational models of the INDISIM family, INDISIM-Paracoccus and INDISIM-Denitrification have been designed, implemented on the NetLogo platform, parameterized and calibrated with experimental data to analyze the system dynamics in a bioreactor in batch and continuous culture with denitrifying bacteria growing in it. The bioreactor conditions can be aerobic and/or anaerobic, and the growing media is liquid medium with an electron donor, C-source, N-source, and oxygen and all N-oxides as electron acceptors. An open access and open source tool has been developed to write MMRs based on TEEM. It is called MbT-tool which stands for Metabolism-based on Thermodynamics. Using MbT-Tool three sets of MMRs have been written, which are the centre of the individual sub-model for INDISIM-Paracoccus and INDISIM-denitrification, representing reactions involved in: i) cellular maintenance, ii) individual mass synthesis, and iii) individual mass degradation to reduce cytotoxic products. The simulation results obtained with INDISIM-Paracoccus and INDISIM-Denitrification have been compared with experimental data published by Felgate et al. (2012) regarding the growth of Paracoccus denitrificans and Achromobacter xylosoxidans in a bioreactor. According to the statistical analysis of the simulations results, for both denitrifying bacteria tested, the IBMs developed show better adjustments in the assays with electron donor limited than in the assays with electron acceptor limited.