Genomic Evidence for the Purging of Deleterious Genetic Variation in the Endangered North Atlantic Right Whale

Abstract

The reduced genetic diversity and frequent inbreeding associated with small population size may underpin the accumulation and expression of deleterious mutations (mutation load) in some declining populations. However, demographic perturbations and inbreeding coupled with purifying selection can also purge declining populations of deleterious mutations, leading to intriguing recoveries. To better understand the links between deleterious genetic variation and population status, we assess patterns of genetic diversity, inbreeding, and mutation load across the genomes of three species of Balaenidae whale with different demographic histories and recoveries following the end of commercial whaling in the 1980s. Unlike bowhead (BH) and Southern right whales (SRW), which show signs of recent recovery, reproductive rates of the endangered North Atlantic right whale (NARW) remain lower than expected. We show that the NARW is currently marked by low genetic diversity, historical inbreeding, and a high mutation load. Still, we reveal evidence that genetic purging has reduced the frequency of highly deleterious alleles in NARW, which could increase chances of future population recovery. We also identify a suite of mutations putatively linked to congenital defects that occur at high frequencies in nulliparous NARW females but are rare in NARW with high reproductive success. These same mutations are nearly absent in BH and SRW in this study, suggesting that the purging of key variants may shape the probability of population recovery. As anthropogenic disturbances continue to reduce the sizes of many populations in nature, resolving the links between population dynamics and mutation load could become increasingly important.

Keywords: North Atlantic right whale; genetic purging; inbreeding; mutation load.

FIGURE 1

Estimates of genetic diversity and evolutionary relationships across Balaenidae whales. Species are colored as depicted in the legend. (A) Mitochondrial genome phylogeny estimated using BEAST. Posterior probabilities > 0.95 are denoted with a filled circle. (B) Estimates of nucleotide diversity within 100‐Kb windows across autosomes and the X chromosome. (C) Estimates of inbreeding given by F for autosomes and the X chromosome. Note that the order of species is consistent across (B) and (C). The distribution of the lengths of runs of homozygosity for each species for (D) the autosomes and (E) X chromosome. Error bars are (±) one standard error.

FIGURE 2

Estimates of mutation load across Balaenidae whales for autosomes and the X chromosome. Species are colored as depicted in the legend and genomic region is labeled along the X axis for (A) Total mutation load, (B) homozygous mutation load, and (C) heterozygous mutation load. Error bars are (±) one standard error.

FIGURE 3

Genetic purging across Balaenidae whales. (A) Mean values of Rxy for the autosomes and (B) for the X chromosome. Note that posterior probability estimates are shown in Figure S10 in Supporting Information. The predicted mutation impact is colored as depicted in the legend. (C) The relative mutation load (homozygous mutation load/F) for both autosomes and the X chromosome for each population. The proportion of deleterious homozygotes found within runs of homozygosity for the autosomes and (D) for the X chromosome (E) for each predicted mutation impact. Species are colored as depicted in the legend. Error bars are (±) one standard error.

FIGURE 4

Deleterious mutations across the genomes of adult female NARW for the autosomes and the X chromosome. Violin plots summarizing variation are shown to the left and associated Manhattan plots for each outlier approach are shown to the right. Genomic region and position are labeled accordingly along the X axes. The predicted impact of nonsynonymous mutations is colored as depicted in the legend. Note that only the 41 variants found to be outliers across all three approaches are highlighted in Manhattan plots. Gray points in Manhattan plots reflect nonsynonymous mutations that were not classified as candidate variants. Median values are indicated with solid circles or lines and the assigned outlier thresholds are shown as dotted lines. (A) The distribution of −log₁₀ P‐values for variant alignment with Principal Components and (B) Manhattan plot of the −log₁₀ P‐values for variant alignment with Principal Components. (C) The distribution of F _ST values per site between the genomes of high fecund and nulliparous NARW. (D) Manhattan plot of F _ST values per site between the genomes of high fecund and nulliparous NARW. (E) The distribution of differences in total mutation load per site between the genomes of high fecund and nulliparous NARW females and (F) Manhattan plot of differences in total mutation load per site between the genomes of high fecund and nulliparous NARW females. Positive values indicate a greater mutation load in nulliparous NARW.

FIGURE 5

Results of gene enrichment analyses showing only weighted terms to account for redundancy. (A) Nonredundant terms with the highest gene enrichment values from over‐representation analysis using a Gene Ontology background. (B) Nonredundant terms with the highest gene enrichment values from over‐representation analysis using the KEGG background. GO‐Slim summary terms for biological process (C), cellular component (D), and molecular function (E).