Intra-specific evidence to support the drift-barrier hypothesis of mutation rate evolution

Rates of de novo mutations ({micro}) are highly variable among different taxa but less is known about their variability at the intraspecific level. Using a large data set (n=364 trios) from eight nine-spined stickleback (Pungitius pungitius) populations differing in their effective population sizes (Ne), we tested the prediction of the drift-barrier hypothesis (DBH) that {micro} scales negatively with increasing Ne. Indeed, {micro} was a negative function of Ne, also after correcting for phylogenetic non-independence of populations. While the range of variation in mutation rates across populations spanned more than a 4-fold range ({micro} = 1.60 - 6.99 x 10-9), we discovered one highly mutable family with a five-times higher mutation rate than the population average. Evidence was also found (i) for reduced efficiency of selection in small freshwater populations subject to strong genetic drift, (ii) that maternal age associates positively with {micro}, and shorter generation time elevates per-year mutation rates, (iii) that both replication errors and DNA repair efficiency contributed to {micro}, and that (iv) mutation rate variation has a polygenic basis. In general, the results provide support for the DBH identifying elevated mutation rates in populations subject to strong genetic drift.