How fast can cells evolve novel behaviors, what types of mutations lead to these
behaviors, and how many different mechanisms can lead to a single type of
behavior? I will discuss our progress in analyzing the evolution of behavior in the
budding yeast, Saccharomyces cerevisiae. Our basic approach is to apply strong
selection for interesting traits in laboratory populations with high mutation rates,
characterize the behaviors that appear, track down the mutations that cause these
behaviors, and see what we can learn about how our populations evolved and then
speculate about how natural populations have evolved over the much longer time
scales of the non-laboratory world. The behaviors we have evolved include forming
multicellular aggregates as a better way of exploiting public goods, using regularly
fluctuating selection to produce a circadian oscillation in cells' ability to stick to each
other, and the genetic equivalent of associative learning: the evolution of the ability
to use an earlier innocuous signal to induce a protective response that will keep a
later noxious environment from killing cells.