Microbiology (1998), 144, 3275–3287 Printed in Great Britain BacSim, a simulator for individual-based modelling of bacterial colony growth Jan-Ulrich Kreft, 1 Ginger Booth 2 and Julian W. T. Wimpenny 1 Author for correspondence : Jan-Ulrich Kreft. Tel : 44 1222 874000 ext. 6036. Fax : 44 1222 874305. e-mail : Kreftcardiff.ac.uk 1 School of Pure and Applied Biology, Cardiff University, PO Box 915, Cardiff CF1 3TL, UK 2 Center for Computational Ecology, Yale Institute for Biospheric Studies, PO Box 208104, New Haven, CT 06520-8104, USA The generic, quantitative, spatially explicit, individual-based model BacSim was developed to simulate growth and behaviour of bacteria. The potential of this approach is in relating the properties of microscopic entities – cells – to the properties of macroscopic, complex systems such as biofilms. Here, the growth of a single Escherichia coli cell into a colony was studied. The object-oriented program BacSim is an extension of Gecko, an ecosystem dynamics model which uses the Swarm toolkit for multi-agent simulations. The model describes bacterial properties including substrate uptake, metabolism, maintenance, cell division and death at the individual cell level. With the aim of making the model easily applicable to various bacteria under different conditions, the model uses as few as eight readily obtainable parameters which can be randomly varied. For substrate diffusion, a two-dimensional diffusion lattice is used. For growth-rate-dependent cell size variation, a conceptual model of cell division proposed by Donachie was examined. A mechanistic version of the Donachie model led to unbalanced growth at higher growth rates, whereas including a minimum period between subsequent replication initiations ensured balanced growth only if this period was unphysiologically long. Only a descriptive version of the Donachie model predicted cell sizes correctly. For maintenance, the Herbert model (constant specific rate of biomass consumption) and for substrate uptake, the Michaelis–Menten or the Best equations were implemented. The simulator output faithfully reproduced all input parameters. Growth characteristics when maintenance and uptake rates were proportional to either cell mass or surface area are compared. The authors propose a new generic measure of growth synchrony to quantify the loss of synchrony due to random variation of cell parameters or spatial heterogeneity. Variation of the maximal uptake rate completely desynchronizes the simulated culture but variation of the volume-at-division does not. A new measure for spatial heterogeneity is introduced : the standard deviation of substrate concentrations as experienced by the cells. Spatial heterogeneity desynchronizes population growth by subdividing the population into parts synchronously growing at different rates. At a high enough spatial heterogeneity, the population appears to grow completely asynchronously. Keywords : individual-based modelling, colony growth, growth synchrony, spatial heterogeneity INTRODUCTION The fundamental unit of bacterial life, encapsulating ................................................................................................................................................. Abbreviations : 2D, two-dimensional ; IbM, individual-based model/ modelling. action, information storage and processing, as well as variability, is the cell. It therefore seems appropriate to construct ecological models in terms of individual cells and their behaviour. This paper introduces spatially explicit individual-based modelling (IbM) to microbial ecology. The great potential of IbM lies in addressing 0002-2653  1998 SGM 3275