The biased evolution of generation time

Abstract

Many life-history traits, like the age at maturity or adult longevity, are important determinants of the generation time. For instance, semelparous species whose adults reproduce once and die have shorter generation times than iteroparous species that reproduce on several occasions. A shorter generation time ensures a higher growth rate in stable environments where resources are in excess, and is therefore a positively selected feature in this (rarely met) situation. In a stable and limiting environment, all combination of traits (or strategies) that produce the same number of viable offspring on average are strictly neutral even when their generation times differ. We first study the evolution of life-history strategies with different generation times in this context, and show that those with the longest generation time represent the most likely evolutionary outcomes. Indeed, strategies with longer generation times generate fewer mutants per time unit, which makes them less likely to be replaced within a given time period. This `turnover bias' inevitably exists and favors the evolution of strategies with long generation times. Its real impact, however, should depend on the strength and direction of other evolutionary forces; selection for short generation times, for instance, may oppose turnover bias. Likewise, the evolutionary outcome depends on the strength of such selection and population size, comparably to other biases acting on the occurrence of mutations.