Evolution of multi-stage dormancy in temporally autocorrelated environments.

Question: Under what circumstances does a capacity for multi-stage dormancy (i.e. dormancy
in more than one life-stage) evolve?
Mathematical methods: Optimization in stochastic environments. Results are derived both
analytically and by simulations.
Key assumption: There exists some trade-off between resources allocated to reproduction and
adult dormant survival. Different shapes of this trade-off are investigated.
Major conclusions: Multi-stage dormancy can evolve in an environment with low serial
autocorrelation. However, a slowly changing environment, with high positive autocorrelation,
will prevent the evolution of dormancy in several life-stages. In general, a high positive
environmental autocorrelation will separate the evolution of life parameters associated with
active life from that of parameters associated with dormant life.