Key points are not available for this paper at this time.
A dynamic analysis is presented of the manner in which survival rates and reproductive schedules evolve in variable environments. This approach accounts for (1) age structure, (2) density-independent fluctuations in vital rates, (3) life histories of arbitrary length, (4) variance and covariance of vital rates, and (5) an incorporation of genotypic differences in life history. The results indicate that there is no general advantage to a more iteroparous life history in a variable environment and that the direction of evolution depends on the absolute amount of environmental variation and the correlation structure of the vital rates. Analysis of "indifference" curves also indicates that there is considerable potential for neutral evolution with respect to life history, given genotypic differences in environmental sensitivity. We suggest two nonexclusive partial answers to Cole's paradox. One is that the magnitude of environmental variability and the correlation structure of vital rates cause short-lived, early-reproducing semelparous life histories to be disadvantageous within populations. The other answer is that, if stochastic growth rates are nonnegative, such life histories have higher extinction rates when their stochastic growth rates equal those of more-iteroparous life histories, given that all the life histories share a constraint on the average life history.
Orzack et al. (Thu,) studied this question.