PDF《Tokyo 169, Japan》

14 samples from the same strain which look more or less similar, we obtained a fractal dimension M. Matsushita and H. Fujikawa / Diffusion-limited bacterial colony growth

501 b ~i ? i , Fig. 1. The effect of nutrient concentrations on the colony growth. The photographs were taken three weeks after inoculation. The initial peptone concentrations were (a)

0 g/l, (b) 0.25 g/I and (c) 0.5 g/l (the colony size is about 2.5 cm). Note in particular that no growth occurs without nutrient.

502 M. Matsushita and H. Fujikawa / Diffusion-limited bacterialcolony growth Fig. 2. A typicalexample of DLA-like colony patterns incubated at 35°C for three weeks after inoculation on the surface of agar plates containing initially 1g/l of peptone as nutrient. This pattern has a fractal dimension of D ~ 1.72. D = 1.73 _+0.02. This is in good agreement with that of two-dimensional DLA patterns [8]. As described above, this result does not immediately mean that bacterial colonies are formed by the DLA mechanism;

more evidence is needed. The next problem is whether these colony patterns are really growing through DLA processes or not. In order to examine this, we first tried to observe the existence of the screening effect during the colony growth. This effect is characteristic for the pattern formation in a Laplacian field. As clearly seen in fig. 3, many interior branches were found to stop growing afterwards in spite of their open neighborhood during the incubation period. This evidences the existence of the screening effect of protruding main branches against interior ones in a colony. We next observed the behavior of two neighboring colonies. In fig.

4 the colony pattern is shown incubated three weeks after having inoculated at two points

1 cm apart. As clearly seen in the figure, two neighboring colonies repel each other and never fuse together. This behavior is also another feature of the pattern formation in a ........