TY - JOUR T1 - Genetic load may increase or decrease with selfing depending upon the recombination environment JF - bioRxiv DO - 10.1101/2021.05.20.445016 SP - 2021.05.20.445016 AU - Shelley A. Sianta AU - Stephan Peischl AU - David A. Moeller AU - Yaniv Brandvain Y1 - 2021/01/01 UR - http://biorxiv.org/content/early/2021/05/22/2021.05.20.445016.abstract N2 - Theory predicts that the ability for natural selection to remove deleterious mutations from a population, and prevent the accumulation of genetic load, is a function of the effective population size (Ne). Shifts from random mating to self-fertilization (“selfing”) are predicted to decrease Ne through a variety of genomic changes - including a reduction in effective recombination and an increase in homozygosity. While a long history of theory suggests that the efficacy of selection, particularly against non-recessive mutations, should decrease with selfing rate, comparisons of genomic-based estimates of the efficacy of selection between related outcrosser-selfer pairs have revealed conflicting results. We address this paradox by simulating the evolution of strongly deleterious recessive and weakly deleterious additive mutations across a range of recombination, mutation and selective parameter combinations. We find that the genetic load of a population can either increase, decrease, or not vary with selfing rate. Genetic load is higher in selfers only when recombination rates are greater than mutation rates. When recombination rates are lower than mutation rates, an accumulation of recessive mutations leads to pseudo-overdominance, a type of balancing selection, in outcrossing populations. Using both simulations and analytical theory, we show that pseudo-overdominance has strong negative effects on the efficacy of selection against linked additive mutations and that a threshold level of selfing prevents pseudo-overdominance. Our results show that selection can be more or less effective in selfers as compared to outcrossers depending on the relationship between the deleterious mutation rate and gene density, and therefore different genomic regions in different taxa could show differing results. ER -