Maintenance of Adaptive Dynamics and No Detectable Load in a Range-Edge Outcrossing Plant Population

Abstract

During range expansion, edge populations are expected to face increased genetic drift, which in turn can alter and potentially compromise adaptive dynamics, preventing the removal of deleterious mutations and slowing down adaptation. Here, we contrast populations of the European subspecies Arabidopsis lyrata ssp. petraea, which expanded its Northern range after the last glaciation. We document a sharp decline in effective population size in the range-edge population and observe that nonsynonymous variants segregate at higher frequencies. We detect a 4.9% excess of derived nonsynonymous variants per individual in the range-edge population, suggesting an increase of the genomic burden of deleterious mutations. Inference of the fitness effects of mutations and modeling of allele frequencies under the explicit demographic history of each population predicts a depletion of rare deleterious variants in the range-edge population, but an enrichment for fixed ones, consistent with the bottleneck effect. However, the demographic history of the range-edge population predicts a small net decrease in per-individual fitness. Consistent with this prediction, the range-edge population is not impaired in its growth and survival measured in a common garden experiment. We further observe that the allelic diversity at the self-incompatibility locus, which ensures strict outcrossing and evolves under negative frequency-dependent selection, has remained unchanged. Genomic footprints indicative of selective sweeps are broader in the Northern population but not less frequent. We conclude that the outcrossing species A. lyrata ssp. petraea shows a strong resilience to the effect of range expansion.

Keywords: adaptation; deleterious mutations; negative frequency-dependent selection; range expansion; selective sweeps; self-incompatibility locus.

Fig. 1.

Demographic analysis of three Arabidopsis lyrata ssp. petraea populations. (a) Folded site frequency spectrum of synonymous sites for PL and SP. (b) Schematic representation of the best-fit demography model. Shown within the boxes are the effective number of diploid individuals (N_e), divergence times (horizontal black lines) are indicated in thousands (k) of generations, with the exception of the final bottleneck in PL. This bottleneck is inferred to have occurred only 143 years ago but it must be noted that, in contrast to the other demographic events, it is not supported by other methods. The time since migration ended (horizontal red lines and numbers in red) is also indicated in thousands of individuals or generations. Width of the elements represents relative differences in N_e (in logarithmic scale), whereas time-differences in logarithmic scale are represented by the height of the elements.

Fig. 2.

Evidence of a strong bottleneck along the SP genome. (a) Tajima’s D distribution for AUS, PL, and SP calculated along the chromosomes in 10-kb nonoverlapping windows. (b) Linkage disequilibrium decay in SP and PL given by SNP pairwise r² as a function of the distance between the SNPs. For comparison, both populations were down-sampled to 12 individuals each.

Fig. 3.

Comparative efficacy of selection and genomic burden in SP and PL. (a) Ratio of PL/SP of the proportion of variants for each s category and each allele frequency bin. Values below 1 indicate that mutations of a given size effect are less abundant in PL than in SP, within each frequency bin. This estimate is based on the joint estimate of the gamma distribution of the DFE using the Poisson optimization and the expected SFS in each category of s. As a proportion of the total number of variants at each count, PL has more slightly neutral and nearly neutral mutations (orange lines) at low frequency and considerably less strongly deleterious mutations (purple lines). (b) Difference in per-individual cumulative derived allele burden between PL and SP. The cumulative derived allele burden is based on the contribution of deleterious variants depending on their count in the population considering the point mass s estimate of deleterious mutations of −1.2, which was shown to fit the data well. Low-frequency mutations contribute more to the burden in PL—negative values indicate that an excess of up to 10,000 deleterious mutations with count 10 or less in the population accumulate in each individual in PL-, whereas fixed mutations (count 28 in the population) play an important role in SP. The net difference, given by the end of the line, is 185. (c) Comparison of genomic load in PL and SP, for synonymous, nonsynonymous, and high impact mutations. For each population, the genomic load was calculated as the mean number of nonsynonymous corrected by the total number of genotyped sites for each sampled individual. The ratio of mean per individual genomic load of PL versus SP is given. The distribution was established by bootstrap of the genome (see Materials and Methods).