Stressed bacteria spawn elegant colonies.

Colonies of bacteria under stress form striking patterns. Put them on an inhospitable surface and a lean diet, and they spread out into elaborate networks, presumably in arrangements that enhance their survival.

But exactly how and why bacteria make these extraordinary patterns remains unexplained. How do they signal each other? By what mechanism do they respond to attractants or repellents? In what way does clumping together in rings help them use available resources more efficiently?

Lev Tsimring and Herbert Levine, physicists at the University of

California, San Diego, and their colleagues propose a model to explain this bacterial behavior. By means of computer graphics, their model--based on diffusion processes in nonliving chemical systems--produces patterns quite similar to those observed in live bacteria.

The physicists detail their results in the Aug. 28 Physical Review Letters.

When deprived of nutrients, colonies of Escherichia coli spawn stripes and rings as the microorganisms react to each other and to their environment. Presumably, they move toward food and neighboring bacteria and away from biological waste, yielding regular spacings, the researchers believe.

To simulate this phenomenon, the physicists invoke a chemical diffusion model first proposed by mathematician Alan Turing in the 1950s. Applied to bacteria, the model emphasizes feedback mechanisms, based on the interplay of chemical attractants and repellents. The fact that the computer simulations mimic patterns observed in live colonies of bacteria leads the physicists to conclude that "generic mechanisms" may be at work.

"Not much is known about how cells communicate with each other chemically," Levine says. "So in these biological structures, we're using reverse logic. We're working backwards from observed patterns in living systems to those seen in nonliving systems in an effort to determine what physical mechanisms must be at work." But do these models actually represent bacterial biochemistry? "It's hard to say," says Howard C. Berg, a biologist at Harvard University who, with biologist Elena O. Budrene, first reported such bacterial patterns in 1991 (SN: 3/4/95, p.136).

"It's definitely worthwhile to look at these patterns from a chemical systems point of view. But whether this model has anything to do with how cells organize and develop themselves is still a matter in question.

"It's possible that they're right," Berg continues. "But we don't know yet. We have to do more experiments in the laboratory with bacteria to test their hypothesis."