Genetic recombination as DNA repair

Maintenance of sexual reproduction and genetic recombination imposes physiological costs when compared to parthenogenic reproduction, most prominently: for maintaining the corresponding (molecular) machinery, for finding a mating partner, and through the decreased fraction of females in a population, which decreases the reproductive capacity. Based on principles from information theory, we have previously developed a new population genetic model, and applying it in simulations, we have recently hypothesized that all species maintain the maximum genomic complexity that is required by their niche and allowed by their mutation rate and selection intensity. Applying this idea to the complexity overhead of recombination maintenance, its costs must be more than compensated by an additional capacity for complexity in recombining populations. Here, we show a simple mechanism, where recombination helps to maintain larger biases of alleles frequencies in a population, so the advantageous alleles can have increased frequency. This allows recombining populations to maintain higher fitness and phenotypic efficiency in comparison with asexual populations with the same parameters. Random mating alone already significantly increases the ability to maintain genomic and phenotypic complexity. Sexual selection provides additional capacity for this complexity. The model can be considered as a unifying synthesis of previous hypotheses about the roles of recombination in Muller's ratchet, mutation purging and Red Queen dynamics, because the introduction of recombination both increases population frequencies of beneficial alleles and decreases detrimental ones. In addition, we suggest simple explanations for niche-dependent prevalence of transient asexuality and the exceptional asexual lineage of Bdelloid rotifers. Introduction In comparison with parthenogenic (or clonal or asexual) reproduction, recombination and sexual reproduction require the increase of complexity in a number of phenotypic features for mating partner choice, mating itself and recombination, e.g. by providing the corresponding (molecular) machinery. This creates additional costs for the organisms and must therefore provide some compensating advantages if it is not to be lost through selection. To explain these advantages, numerous hypotheses have been proposed (Maynard Smith 1978; Kondrashov 1993; Hartfield and Keightley 2012). They largely fall into two big groups (West et al. 1999; Meirmans and Strand 2010): The first one focuses on the effect of (deleterious) mutations at individual loci accumulated over time and is proposed by population geneticists (Mutational Deterministic hypothesis and Fisher-Muller hypothesis). These models were motivated by the mathematical formulations of classical population genetics and have emphasized respective parameters like mutation rate, population size, allele frequencies and interactions between (deleterious) mutations at different loci in the face of selection (epistasis). The