In the multicellular yeast, the division of labor was more subtle. Although the experiment's artificial selection favored large clusters, a large cluster required more time to grow before it could reproduce. That meant that smaller clusters, which divide in half more quickly, could soon outnumber the larger clusters. But after many generations of selection, the large clusters evolved a solution: nonreproductive cells which served as points where offspring could break away from the parent cluster. By providing more break points, these specialized cells allowed the clusters to break into more pieces, to produce a greater number offspring quickly.
“The discovery that there are cells specialized to die in order for the structure to reproduce is suggestive of the first steps toward cellular differentiation,” Grosberg says.
Although researchers agree that the yeast clusters could indeed be considered multicellular organisms, they remain relatively simple. "The researchers are not going to evolve sponges with this approach, but it's amazing what they’re able to do so quickly," Simpson says.
The fast evolution was not all that surprising to Grosberg, who has written papers arguing that multicellularity should be relatively easy to evolve; other researchers have estimated that multicellularity has arisen independently on at least 25 different occasions throughout the history of life. Yet nobody really knew how it originated, or what steps were involved in the process. By watching evolution in progress, the new research uncovered experimental evidence for these theories and revealed one possible scenario of how multicellularity may have evolved.
"We had hypotheses about how multicellularity could evolve, but until now, no one has really been able to test them,” Ratcliff says. "Now that we have this experimental system, we can ask lots of really exciting questions."