Sampling error and sterility. Unigenerational models of budding viscosity and kin assortment

The stability of sterility, or irreversible reproductive suicide, is explored by simulation in haploid (asexual and sexual) unigenerational models of budding viscosity and kin assortment, incorporating, variously, sampling error within groups, sampling error at the finite population level, stochastic mutation and/or migration, stochastic group selection, and viscous competition among groups. Sterility herein is opportunistic, so a genotype capable of sterility is not always phenotypically so; rather, the number of sterile individuals per group is always less than group size, so a group homogeneous for the genotype always has a reproductive. Under budding viscosity it is possible for sterility to appear stable under benefits precluded by Hamilton's rule; under kin assortment huge benefits can be insufficient to maintain sterility. Sterility may appear stable for up to 98,000 generations, with the genotype then eliminated in less than 40; in some simulations sterility is stable for 100,000 generations for multiple runs of the same parameters, but eliminated in another such run. Under kin assortment usually only single parent groups without sex or with inbreeding are favourable to sterility; even where stable under kin assortment, sterility may be made unstable by decreasing group size, holding pedigree relatedness constant. Sterility's trajectory is not a consequence of average selection effects, but rare population states which place it in jeopardy of stochastic elimination. Such population-wide sampling effects are independent of Hamilton's rule but, to some extent, nonetheless predictable. While stability criteria for sterility become somewhat a matter of taste, different group formation rules can be rank-ordered by relative stability.