The ideas presented here are compatible with some aspects of population genetic theory and are consistent with the substantial and growing corpus of empirical data. Admittedly, much of the scenario of the Sisyphean cycle is speculative, but I hope that this speculation will inspire empirical tests of hypotheses concerning the impact of protein polymorphisms on whole plant and animal physiology, the mechanism of fitness determination, the form of multilocus organization generated by truncation selection, and the intensity of natural selection. These tests must be conducted by population biologists and physiological ecologists, and they will have a profound impact on our perspectives on evolution.
A set of correlated axes describe some of the variation among species in fecundity, population density, geographic range, and genetic variation. At one end of this continuum are species with high fecundity, large populations, broad geographic ranges, and abundant genetic variation. At the other extreme are relatively rare, often endemic species with low fecundity and little genetic variation.
The opportunity for natural selection is highest at the highly fecund end of this set of correlated axes. In highly fecund species, selection is predicted to favor highly heterozygous individuals. Species at the high end of the fecundity axis have low rates of speciation, high rates of molecular evolution, long tenures in the fossil record, and stasis.