Fig. 4
Distribution of aerobes and anaerobes as a function of spatial positioning within the different pore networks. Results show combined experimental cell distribution and simulation results from the mathematical model for all pore network connectivities (a–c), using counter-gradients of citrate (maximum concentration at center port) and oxygen (maximum concentration at peripheral ports). Experimental results are represented in boxplots (whiskers indicate minimum to maximum value of data) of three, four, and five experimental replicates for the 100%, 70%, and 40% pore networks, respectively. In the mathematical model results, thick line indicates the mean and shaded area includes 95% of all cells. In the experiments, bacterial cells in fluorescence microscopy images were identified using digital image analysis, mapped onto the pore network, and attributed to the closest nodes. These were subsequently grouped into bins with respect to their shortest path from the central port. A pore-scale segregation of the two species is visible within each connectivity. In the 40% lattice, the distribution is less clear due to perturbations of the pore network geometry. To test the role of counter-gradients, the experimental results in d depict bacterial distributions for carbon and oxygen collocated at the peripheral ports (eliminating counter-gradients) resulting in dominance of facultative anaerobes at both the central and peripheral ports. Simulations of this scenario would result in the dominance of aerobes at the peripheral port due to the (simplified) representation of P. veronii as an obligate anaerobe
