Modelling Microbial Populations in Variable Environments

Tolla, C. 2006. Modelling Microbial Populations in Variable Environments. PhD-thesis, Vrije Universiteit, Amsterdam (2006/11/07) & Université de la Méditerrannée (2006/11/10)
Nederlandse versie


Natural environments are frequently complex and many factors can interplay in the performance of organisms; food (substrate) availability and quality is very important. Furthermore, these environments are forced by perturbations in time and space that influence the fate of populations. This introduces variability. This complexity is hard to capture in experiment, and mathematical models that are based on explicit assumption can help to fill the gap in understanding. Classical models for microbial dynamics typically deal with single substrates and equilibria, and are inappropriate for natural environments. These empirical formulations do not respect thermodynamical contraints on biochemical transformations and their mathematical and numerical analysis is typically awkward.

This thesis expands on a realistic characterisation of the links between microbial populations and their environment through a mechanistic modelling approach. This approach is based on the Dynamic Energy Budget (DEB) theory which accounts for the mass conservation law, stoichiometric constraints and difference between the organisms. This theory for individual cells differentiates between mobile (reserve) and non-mobile (structural biomass) metabolic pools and uses synthesizing units (SUs) to describe metabolic transformations. The dynamics of these units is a generalisation of classic enzyme kinetics. I (i) specified models for the degradation of organic matter, using these more realisic formulations for bacterial physiology; (ii) improved the model by incorporation a new inhibition module that better captures phenomena (.g. shrinking of biomass) during starvation; (iii) compared this new model with classical empirical models via application to a set of data; (iv) applied the model to understand the partial inhibition of denitrification by dioxygen to study the impact of the macro-benthos on the bacterial activity.

My work yields a number of general ideas (i) on the interactions between microbial populations and their environments by considering their functionality; (ii) on the balance between simplicity and realism of models; (iii) on the use of models to obtain biological insight. Classical and mechanistic models can show very similar results for stationary cases, but they show noticeable differences if environmental conditions are perturbed. Being more stable, the mechanistic model can better handle variable environments.

The models developed here can now to desing new experiments: which parameters must be measured, and how frequent, to obtain particular knowledge. Furthermore, the mechanistic formulations can help to understand the physiological state of the individual and have a wide applicability in modelling metabolic processes.

Keywords: modelling population dynamics, interactions between substrates, DEB theory, mechanisms, enzymatic kinetics.

Full text in pdf format

This is the symposium that concludes my project

Caroline's project page