E. coli bacteria prefer to eat glucose, but once it is all gone, they can switch into a different metabolic mode to eat acetate. (Rocky Mountain Laboratories/NIAID/NIH)

Evolution is surprisingly predictable and its course seems relatively unaffected by chance events, a new study suggests.

Michael Doebeli, a mathematician and evolutionary biologist at the University of British Columbia, and postdoctoral researcher Matthew Herron conducted three separate experiments under the same laboratory conditions, observing 1,000 generations of E. coli bacteria as they evolved into two different strains — a process known as diversification that eventually gives rise to separate species.

They wanted to answer one of the big questions about evolution: Is diversification a predetermined process or is it partly driven by chance events?

'We have basically a genetic fossil record of what happened.'—Michael Doebeli, evolutionary biologist

That answer could in turn help answer a bigger question that some biologists have mused about.

That is, if you went back thousands of years and "replayed the tape of life," would you end up with humans and the species we know today, or would small differences caused by chance result in completely different plants and animals?

Doebeli was surprised to discover that in his study, all three bacterial populations evolved in almost exactly the same way, suggesting that chance or randomness doesn't play a big role, at least over a short period of time, such as 1,000 generations, and in laboratory conditions.

"It's a deterministic process — it unfolds in very similar ways in independent instances," said Doebeli. "We were surprised by just how parallel it actually is."

Not only did the bacteria go through similar mutations, but similar changes in the populations occurred at similar times.

Mathematical prediction

Doebeli said the process is consistent with mathematical predictions. The results were published this week in the journal PLoS Biology.

'For the same conditions, it will always occur in the same way.'—Michael Doebeli, evolutionary biologist

Doebeli's study involved growing E. coli bacteria in an environment with two nutrients, glucose and acetate.

The bacteria prefer to eat glucose, but once it is all gone, they can switch into a different metabolic mode to eat the acetate. Over 1,000 generations, the bacteria evolve via natural selection into two different strains:

• One that is better at eating glucose but bad at switching its system into acetate-eating mode.
• One that is not as good at eating glucose, but better at switching into acetate-eating mode.

That's because initially, when everyone is going after the same resource, an individual who can tap into a different resource has a distinct advantage and will be able to produce more descendants.

Throughout the process, which took about six months, the researchers removed and froze bacteria from different generations at regular intervals. At the end, they analyzed the changes to the DNA over time to see what changes occurred.

"We have basically a genetic fossil record of what happened," Doebeli said.

In all three cases, the diversification followed the same pattern, and each step occurred in the same order and took the same amount of time.

First, a strain that was unusually good at eating gluocose arose. Then a strain that was good at eating acetate came about. That led to more changes in the glucose specialist, which in turn led to more changes in the acetate specialist, and so on.

Key mutations 'not random'

Of course, Doebeli acknowledged, it's true that the mutations that allowed that process to take place occur randomly.

"But the ones that actually make it — the mutations that can actually proliferate — those are not random," he said. "They are determined by this environment."

Although the exact mutations differed between the three populations, they tended to occur in genes that had similar functions.

Doebeli said the study suggests this kind of diversification in an environment with multiple resources isn't something that happens "by chance every now and then."

"Under the right conditions, it's actually expected to occur," he said. "And for the same conditions, it will always occur in the same way."

He added that he thinks that finding is applicable to all organisms. However, he acknowledged that sexual reproduction — where mating between two strains can counteract divergence — does add some complications.

As for the question of whether this type of process played a big role in the evolution of humans, Doebeli said that process may have been "more deterministic than someone would have thought.

"But I don't think we can extrapolate from this 1,000-generation experiment to the emergence of humans."